INTRODUCTION

Overall tree topology resulting from neighbor-joining analysis of 4283 ingroup specimens based on ˜1044 bp of NADH2 sequence data. Numbers refer to figures 1-77 illustrating the respective subtrees; OG, outgroup.

Sharks, rays, and chimeras (chondrichthyan fishes) are widely recognized important constituents of vertebrate biodiversity. They constitute one of the two primary divisions of the gnathostome tree. Comparative anatomists, physiologists, and biochemists have long valued the basal phylogenetic position of chondrichthyans as a proxy for the primitive gnathostome condition. Yet, surprisingly, in spite of their phylogenetic importance, chondrichthyans remain poorly known in terms of their extant biodiversity relative to other vertebrate classes. This is reflected in the relative immaturity of chondrichthyan taxonomy. In the most recent comprehensive authoritative checklist of species of elasmobranchs, Compagno (2005a) recognized 1125 species. However, included in that list were tens of species that had not yet been described. Since 2005, more than 50 of these have been formally treated and named, while over 80 species not treated by Compagno (2005a) have also been described. This brings the total number of known species of elasmobranchs to 1221. The fact that so many new species have been discovered and described over such a short time span, suggests that more remain to be discovered. Much of this increase is a result of reassessment of geographic variation; some of the increase represents recognition of subtle morphological variants among congeneric forms that nevertheless exhibit substantial molecular sequence divergence.

Recent advances in elasmobranch taxonomy have had a major impact on parasitological studies involving elasmobranch hosts. The accurate identification of hosts is of paramount importance in studies of any parasitological system if the results are to be meaningful. However, as our understanding of elasmobranch diversity has grown, so too has the number of instances in which long-standing concepts of host associations have been challenged (e.g., Caira et al., 2007; Jensen and Caira, 2008; Fyler and Caira, 2010; Desjardins and Caira, 2011; Koch et al., 2012; Cielocha and Jensen, 2011; White et al., 2010c). The current project grew out of a global survey of elasmobranch tapeworms that was initiated in the mid-1980s by J.N.C. and was later substantially expanded through collaboration with K.J., in which a strategy to manage the issue of tracking the identities of novel (potentially undescribed) hosts was developed. Host tissue samples were routinely taken from each host individual examined, to serve as a voucher for subsequent sequence validation of host identity. The project was further expanded through collaboration with G.J.P.N. to include sequences from as diverse a set of chondrichthyans as possible, whether examined for parasites or not, with the goal of providing a baseline assessment of sequence variation among and within elasmobranch species in general. It is anticipated that this baseline will be useful to taxonomists, biogeographers, commercial fisheries, and conservation biologists. We hope this data set will serve as a valuable resource to aid investigators working on elasmobranch parasites in the identification of their host taxa.

The fast-evolving protein-coding gene NADH dehydrogenase subunit 2 (NADH2) was selected by G.J.P.N. to provide a means of distinguishing among very closely related species, cryptic species, or geographic variants, while still ensuring that sequence alignments could be rigorously validated at the amino acid level. NADH2 is one of the fastest, if not the fastest, evolving of the 13 mitochondrial protein-coding genes when measured at the amino acid level (Broughton and Reneau, 2006). In elasmobranchs, it is generally between 347 and 349 amino acids (1041–1047 bp) long. Patterns of nuclotide substitution in NADH2 show a high transition bias as is typical for most protein-coding genes. The base composition is often heterogeneous among taxa especially at third codon positions.

Taxon coverage

We have made every effort to include as much phylogenetic diversity as possible. However, the taxon coverage is most comprehensive at higher taxon levels. The analysis includes representation of 56 (98.3%) of the 57 families, 157 (81.4%) of the 193 genera, and as many as 574 species of elasmobranchs. Hypnidae is the only family for which no representatives were included. Also not included from Compagno (2005a) were representatives of the following 15 shark genera: Aculeola, Cephalurus, Chaenogaleus, Cirrhoscyllium, Ctenacis, Euprotomicroides, Gogolia, Heteroscymnoides, Miroscyllium, Mollisquama, new genus (Pseudotriakidae), Pentanchus, Scymnodalatias, Sutorectus, and Trigonognathus; and also the following 19 ray genera: Anacanthobatis, Benthobatis, Breviraja, Crassinarke, Dactylobatus, Diplobatis, Discopyge, Fenestraja, Gurgesiella, Heteronarce, Hypnos, n. gen. 1 (Potamotrygonidae), n. gen. 1 (Rajidae), n. gen. 2 (Potamotrygonidae), n. gen. 2 (Rajidae), Platyrhina, Plesiotrygon, Pseudoraja, and Temera. In total, the data set consists of NADH2 sequences for 4283 specimens; the number of replicate specimens of each species ranges from 1–176. Details of the taxon coverage are provided in table 1, as is a breakdown of elasmobranch species included by family of sharks versus rays. We anticipate that as researchers encounter additional species they will be able to contrast the sequences of their animals against those contained in our database and build on the initial data set to provide a more comprehensive understanding of the group.

TABLE 1

Taxon coverage relative to known genera and species of elasmobranchs, by family The number in parentheses after number of species included indicates number of species potentially new to science recovered from the analysis.

Choice of distance measure

Several distance measures have been proposed to summarize divergence among sequences. Most accommodate the fact that as more distant comparisons are made, the probability of multiple hits at a site increases. The various measures differ in the assumptions they make about the molecular evolutionary process. Jukes-Cantor distance (Jukes and Cantor, 1969), for example, assumes that the rate of change from every nucleotide to every other nucleotide is equal and the base composition is even and stationary. The Kimura 2–parameter model (K2P) (Kimura, 1980) allows for a different rate of transitional changes relative to transversional changes but assumes equal base frequencies. The choice of distance measure is critically important if the goal is to use the distances obtained to estimate phylogeny. Indeed, if a distance could be found that perfectly captures the amount of evolutionary change that has taken place, the tree deduced from the resulting distance matrix would accurately reflect phylogeny. To date no such distance has been found.

In our case, we were interested in exploring the within-species (i.e., intraspecific) versus within-genus (i.e., intrageneric) differences among closely related forms, and so have chosen to use p-distance (Jukes and Cantor, 1969). This simple measure represents the uncorrected distance, expressed as the number, or proportion when expressed as a percentage, of nucleotide differences between two sequences without estimating any unobserved changes (multiple hits) whatsoever. This has the effect of emphasizing differences among close relatives relative to differences among distant relatives. We also provide intrageneric p-distance summary statistics (table 2) as calculated in the Barcode of Life Data Systems (BOLD Systems;  http://www.boldsystems.org) using all specimens from all genera represented by more than one specimen and K2P distances as calculated in the BOLD Systems (v. 2.5;  http://www.boldsystems.org) for all intraspecific and intrageneric comparisons (appendix 1 and table 3, respectively).

TABLE 2

Summary p-distances and number of base pair differences in NADH2 within species and within genera of elasmobranchs

MATERIALS AND METHODS

Liver or muscle tissue was sampled in the field from elasmobranch specimens that were either speared by the authors or their collaborators, collected on research vessels, purchased in local fish markets, or obtained in conjunction with local fishermen. A number of the samples collected by G.J.P.N. in the initial part of the survey were put in cryotubes, and placed in liquid nitrogen until they could be taken to the laboratory for storage at −80° C. Samples collected later in the survey were stored in 95% alcohol or DMSO depending on availability. In early years, total DNA was extracted using phenol chloroform extraction (Sambrook et al., 1989), but in later years using High Pure PCR Template Preparation Kit by Roche Diagnostics (Indianapolis, IN). Extracted total DNA was stored at −20° C until used for PCR amplification. Samples were amplified using Fermentas Taq with primers designed to target the complete coding sequence for NADH dehydrogenase subunit 2 (NADH2). In most cases, a single set of universal primers (Naylor et al., 2005) designed to bind to the ASN and ILE tRNA regions of the mitochondrial genome were successful in amplifying the targeted fragment. However, for some taxa it was necessary to design taxon-specific primers. A complete list of primers for amplification and sequencing is shown in table 4. PCR products were purified by centrifugation through size-selective filters (Millipore, Bedford, MA) according to manufacturer's recommendations. Subsequent to 2008, samples were purified using ExoSAP-IT from USB (Cleveland, Ohio). The purified PCR products were sent off to commercial sequencing centers for sequencing (SeqWright, Houston, TX; Beckman-Coulter Genomics, Beverly, MA; Retrogen, San Diego, CA). The software packages Phred and Phrap (Ewing et al., 1998) were used to read sequence traces, assign quality values, make base calls, and produce output files for subsequent alignment. Sequences were translated to amino acids and aligned using the software package MUSCLE (Edgar, 2004). The aligned amino acid sequences were translated back, but in frame to their original nucleotide sequences, to yield a nucleotide alignment. The full protein-coding alignment was 1053 nucleotides long. Of the sequences that resulted, only those that were 1000 bp or greater were included in the alignment. To minimize the amount of missing data at the leading and trailing ends of some sequences, all sequences were truncated to a length of 1044 bp. Although our original goal was to generate 1044 bp of NADH2 sequence data for each of the 4283 specimens (of the 574 species) included in the analysis, in the end, we achieved this for 4220 (98.5%) of the specimens. Of the 63 remaining specimens, 45 had sequences that were three bases shorter and 18 had sequences that were up to 39 bases shorter; in such cases the missing data were mostly at the beginning and/or end of the sequence. We have not attempted to correct for this, given that the data generated for 99.5% of the specimens was between 1041 and 1044 bp in length.

Distance measures

Distance measures were employed to assess both intraspecific and intrageneric variation. For all species represented by two or more specimens in the analysis, p-distances were calculated and are given as total number of base pair differences in individual species treatments, and as a percent of the 1044 base pairs (along with mean, standard deviation, and range) in appendix 1. P-distances were also calculated for all genera represented by two or more species in the analysis; these are presented as a percent of the 1044 base pairs (along with mean, standard deviation, and range) in table 3. It should be noted that the p-distance data in table 3 represent the average p-distance among species within a genus, rather than among all constituent specimens of all species in each genus as, for example, calculated in BOLD Systems (v. 2.5;  http://www.boldsystems.org). We believe that the approach adopted here for summarizing intrageneric p-distances provides a more conservative and biologically reasonable assessment of intrageneric variation. Furthermore, comparisions made between or among clusters or subclusters are presented as the mean of the mean p-distance of each cluster or subcluster.

TABLE 3

Mean intrageneric distances (%) for NADH2 (n  =  85 genera) P-distance among species within a genus is given as the mean ± standard deviation, range, and number of congeners (n) included in the estimation. K2P distance is given as the mean ± standard deviation, range, and number of comparisons (c) of all constituent, nonconspecific specimens (as calculated in BOLD).

TABLE 4

Complete list of primers used for amplification and sequencing Location of first nucleotide of primer is given relative to first base of start codon in the multiple alignment. Negative numbers indicate that primer is upstream of start codon. Primer sequences are shown in the 5′ to 3′ direction even though their placement on the multiple alignment is numbered with respect to the coding strand.

Nonetheless, to facilitate comparisons with BOLD Systems-based studies (e.g., Ward et al., 2005; Ward and Holmes, 2007; Ward et al., 2007, 2008, 2009), intraspecific K2P distances (appendix 1) and intrageneric K2P distances (table 3), as calculated in BOLD, were also computed. Both of these distance measures are represented as a percent of the 1044 base pairs (along with mean, standard deviation, and range). As a consequence, these intrageneric K2P distances were calculated, as in BOLD, based on comparisons among all specimens of all species in a particular genus. Summary statistics for intraspecific and intrageneric p-distances are presented in table 2. Similar statistics for intraspecific and intrageneric K2P distances are presented in table 5.

Species boundaries and names

Specimens that formed distinct clusters in the neighbor-joining analysis were considered candidates for recognition as distinct species. However, only clusters that also exhibited distinct morphological and/or geographic differences relative to other clusters were assigned unique designations. For example, although Isurus oxyrinchus showed a substantial amount of genetic variation (figs. 35A and 91A), no geographic pattern to this variation was observed (fig. 91B), nor are we aware of any phenotypic differences consistent with this genetic variation. As a consequence, we have treated I. oxyrinchus as a single species. In contrast, specimens of Aetobatus collected from Vietnam (fig. 101B) are both genetically distinct from their congeners in other parts of the world (figs. 63 and 101A) and exhibit a distinct color pattern. Accordingly, these specimens have been given a distinct species designation (Aetobatus sp.).

Figs. 1–77

(following pages). Subtrees of elasmobranchs. 1. Carcharhinidae (requiem sharks) (1 of 18)

2.

same (2 of 18)

3.

same (3 of 18)

4.

same (4 of 18)

5.

same (5 of 18)

6a.

same (6 of 18)

6b.

Continued

7.

same (7 of 18)

8.

same (8 of 18)

9.

same (9 of 18)

10.

same (10 of 18)

11.

same (11 of 18)

12.

same (12 of 18)

13.

same (13 of 18)

14.

same (14 of 18)

15.

same (15 of 18)

16.

same (16 of 18)

17.

same (17 of 18)

18.

same (18 of 18)

19a.

Sphyrnidae (hammerhead sharks)

19b.

Continued

20.

Galeocerdo (tiger sharks)

21.

Hemigaleidae (weasel sharks)

22.

Leptochariidae (barbeled houndsharks)

23.

Triakidae (houndsharks) (1 of 5)

24.

same (2 of 5)

25.

same (3 of 5)

26.

same (4 of 5)

27.

same (5 of 5)

28.

Scyliorhinidae (catsharks) Group 1 (1 of 4)

29.

same (2 of 4)

30.

same (3 of 4)

31.

same (4 of 4)

32.

Pseudotriakidae (false catsharks) and Proscyllidae (finback catsharks)

33.

Scyliorhinidae (catsharks) Group 2 (1 of 2)

34.

same (2 of 2)

35a.

Lamniformes (mackerel sharks)

35b.

Continued

36a.

Hemiscylliidae (longtailed catsharks)

36b.

Continued

37.

Stegostomatidae (zebra sharks), Rhincodontidae (whale sharks), and Ginglymostomatidae (nurse sharks)

38.

Orectolobidae (wobbegongs) and Brachaeluridae (blind sharks)

39.

Parascylliidae (collared carpetsharks)

40.

Heterodontiformes (bullhead sharks)

41a.

Squalidae (dogfish sharks) (1 of 2)

41b.

Continued

42.

same (2 of 2)

43.

Centrophoridae (gulper sharks) (1 of 2)

44.

same (2 of 2)

45a.

Somniosidae (sleeper sharks) and Oxynotidae (roughsharks)

45b.

Continued

46a.

Etmopteridae (lanternsharks)

46b.

Continued

47.

Dalatiidae (kitefin sharks)

48.

Squatiniformes (angelsharks), Echinorhinidae (bramble sharks), and Pristiophoridae (sawsharks)

49.

Hexanchidae (sixgill and sevengill sharks) and Chlamydoselachidae (frilled sharks)

50.

Dasyatidae (whiptail stingrays) (1 of 10)

51.

same (2 of 10)

52.

same (3 of 10)

53.

same (4 of 10)

54.

same (5 of 10)

55.

same (6 of 10)

56.

same (7 of 10)

57.

same (8 of 10)

58.

same (9 of 10)

59.

same (10 of 10)

60.

Urotrygonidae (round stingrays), Potamotrygonidae (river stingrays)

61.

Rhinopteridae (cownose rays), Mobulidae (devilrays), and Pteromylaeus (duckbill ray)

62.

Myliobatidae (eagle rays) (1 of 2)

63.

same (2 of 2)

64.

Gymnuridae (butterfly rays) and Plesiobatidae (giant stingarees)

65.

Urolophidae (stingarees)

66.

Hexatrygonidae (sixgill stingrays)

67.

Zanobatidae (panrays)

68.

Pristidae (sawfishes) and Rhinobatidae (guitarfishes) Group 1

69.

Rhynchobatidae (wedgefishes), Rhinidae (shark rays), and Rhinobatidae (guitarfishes) Group 2

70.

Torpedinidae (torpedo rays), Narcinidae (numbfishes), and Platyrhinidae (thornbacks and fanrays)

71.

Rajidae (skates) (1 of 5)

72.

same (2 of 5)

73.

same (3 of 5)

74.

same (4 of 5)

75.

same (5 of 5)

76.

Arhynchobatidae (softnose skates) (1 of 2)

77.

same (2 of 2) and Anacanthobatidae (legskates)

Based on these criteria, the analysis yielded evidence of 79 potentially undescribed species (see table 1). In a number of cases, candidate species names are available but are not in common use. Rather than resurrecting these names, we have used numerical designations to recognize potentially undescribed species in both the figures and text but have given candidate species names in the species treatments in an effort to further taxonomic work on these species. Examples include Glyphis siamensis for G. sp. 1, Carcharhinus pleurotaenia for the Indo-Pacific C. cf. limbatus, C. japonicus for the Indo-Pacific C. cf. plumbeus, C. cerdale for the eastern Pacific C. cf. porosus, Rhizoprionodon crenidens for R. cf. acutus 2, Rhizoprionodon fissidens for R. cf. acutus 1, Rhizoprionodon walbeehmi for R. cf. acutus 3, Scoliodon muelleri for S. cf. laticaudus, Cephaloscyllium pictum for Cephaloscyllium sp. 1, and Scymnodalatias albicauda for Centroscymnus sp. 1, Galeocerdo arcticus for the Atlantic form of G. cuvier, and Okamejei meerdervoortii for O. cf. porosa. In order to be as conservative as possible, we have not included these taxa in the count of potentially novel species revealed by the analysis, for the existence of possible names for these species attests to the fact that they may not, in fact, be new to science.

Among the novel taxa, are several species that have been previously treated in the literature but never formally described. In such instances, we have attempted to adhere to designations used by previous authors (e.g., Compagno, 2005a; Compagno et al., 2005b). In cases in which entirely novel species were encountered, the taxon label used reflects the nominal species the undescribed species most closely resembles. In cases in which complexes of multiple undescribed species were encountered, each designation was assigned a unique number. The classification follows Compagno (2005a). Common names for species described up to 2005 are generally those formally presented by Compagno (2005a); for species described after 2005, common names were taken from original descriptions. Type locality data was taken from Eschmeyer and Fricke (2011).

RESULTS

The frontispiece provides a schematic overview of figures 1–77 as they relate to the topology of the full neighbor-joining tree. Given the taxonomic scope of this work and that, for some groups, our results suggest that congeners do not form monophyletic groups, we have treated taxa below in the order in which they appear in figures 1–77, regardless of their current generic or familial placements. This allows readers to use the figures as a guide to locate species treatments within the text (also see table 3 and appendix 1).

Carcharhiniformes (ground sharks)

Carcharhinidae (requiem sharks), in part

Carcharhinus falciformis (silky shark) (fig. 1)

A total of 48 specimens representing much of the circumtropical distribution of this species were analyzed. These consisted of 30 specimens taken from Pacific regions including the Philippines, Hawaii, Gulf of California, Borneo, and India, and 18 from Atlantic regions including Senegal, Trinidad, Florida, Gulf of Mexico as well as coastal Georgia. The analysis yielded two weakly divergent subclusters. One subcluster was comprised primarily of specimens from the Pacific localities and the other was comprised of specimens from Atlantic localities. One of the specimens from the Philippines (i.e., GN2242  =  RSE 001) was treated by Compagno et al. (2005a). The range in pairwise differences among specimens within the Pacific subcluster was 0–2, and among specimens within the Atlantic subcluster was 0–2. However, it should be noted that the Pacific cluster also included six specimens collected from Atlantic localities. The average of the pairwise differences between these two subclusters was 8.4. The range in pairwise differences among all specimens of C. falciformis was 0–10; the average of the pairwise differences among these specimens was 4.2. It should further be noted that a specimen (i.e., GN2214  =  BRU 023) from the Philippines considered of uncertain identity but tentatively identified by Compagno et al. (2005a) as “Hemitriakis cf. japanica (var PP),” grouped among the specimens in the Pacific subcluster of C. falciformis.

Prionace glauca (blue shark) (fig. 1)

In total 23 specimens were included. These were collected from across much of the global distribution of this species, consisting of samples from specimens from the western North Atlantic, Gulf of California, Hawaii, and Tasmania. One sample in this cluster came from a specimen from Tasmania in the Australian National Fish Collection (GN4917  =  ANFC H 4223-01). The range in pairwise differences seen among specimens was 0–5; the mean was 2.5. No geographic structure was seen among specimens for this species. This species grouped among species of Carcharhinus, supporting the contention of previous authors (e.g., Compagno, 1988; Naylor, 1992; Dosay-Akbulut, 2008) that the monophyly of Carcharhinus is challenged by recognition of the monotypic Prionace as an independent genus.

Carcharhinus amblyrhynchos (gray reef shark) (fig. 1)

A total of 18 specimens, coming from Borneo, Philippines, Egypt, and Madagascar, and thus spanning much of the Indo-West and central Pacific distribution of this species, were analyzed. The analysis yielded a single cluster; one of the samples from Sarawak, Malaysia, is represented by a museum specimen (GN3672  =  IPPS BO461). The range in pairwise differences seen among specimens in this cluster was 0–10; the mean was 3.8. Essentially no geographic structure was seen among specimens for this species.

Carcharhinus wheeleri (blacktail reef shark) (fig. 1)

A single specimen that was morphologically consistent with and collected from the type locality of C. wheeleri (i.e., the Red Sea) was included in the analysis. This specimen clustered most closely with but still outside the specimens of C. amblyrhynchos. The average of the pairwise differences between the specimen of C. wheeleri and those in the C. amblyrhynchos cluster was 16.8. This result provides support for the recognition of C. wheeleri as a species distinct from C. amblyrhynchos despite suggestions to the contrary (e.g., Compagno, 2005b).

Carcharhinus albimarginatus (silvertip shark) (fig. 1)

The analysis included five specimens, four from Taiwan and one from the Philippines, and thus represented only a subset of the Indo-Pacific distribution of this species. The analysis yielded two subclusters; one subcluster consisted of some specimens from Taiwan and the single specimen from the Philippines, the other subcluster consisted of the remaining specimens from Taiwan. The range in pairwise differences among specimens within these two subclusters was 0–1 and 2, respectively. The mean of the pairwise differences between subclusters was 19.3. The range in pairwise differences among all specimens of this species was 0–21; the mean was 12.

Carcharhinus borneensis (Borneo shark) (fig. 2)

A total of 13 specimens, all collected from Mukah on the island of Borneo, were analyzed. The range in pairwise differences seen among specimens was 0–5; the mean was 2.4. Six of these samples are represented by museum specimens (GN3638  =  IPPS BO426, GN3639  =  IPPS BO427, GN3660  =  IPPS BO449, GN3667  =  IPPS BO456, GN3670  =  IPPS BO459, and GN3640  =  ANFC H 6212-01). This species grouped most closely with the specimens in the C. macloti cluster.

Carcharhinus macloti (hardnose shark) (fig. 2)

A total of 13 specimens were analyzed. These came from the Gulf of Oman, India, Borneo, and northern Australia and thus spanned much of the Indo-West Pacific distribution of this species. The analysis yielded essentially three subclusters: one consisting of the specimens from India and the Gulf of Oman, one consisting of the specimens from Borneo, and one consisting of the Australian specimens. The range in pairwise differences within these subclusters was 0–4, 1–5 and 1, respectively. The mean of the pairwise differences between the Borneo and Australian cluster was 11.8; between the Gulf of Oman/India and Borneo cluster was 10.0, and between the Gulf of Oman/India and Australian cluster was 12.8. The range in pairwise differences among all 13 specimens was 0–14; the mean was 7.8. These results suggest that some consideration should be given to the existence of several distinct species of hardnosed sharks.

Carcharhinus sealei (blackspot shark) complex (fig. 3)

A total of 51 specimens originally identified as C. sealei were included in the analysis. These were collected from the more eastern portions of the distribution of this species and included the Philippines, Borneo, Singapore, and a diversity of localities in western, northern, and eastern Australia. Two distinct subclusters resulted from the analysis, one consisting of all 16 specimens from Australia, and the other consisting of the 35 specimens from the remaining areas in the Indo-West Pacific, including Malaysian Borneo. The range in pairwise differences within the latter subcluster was 0–8; the mean was 1.8. The range in pairwise differences within the Australian subcluster was 0–6; the mean was 1.6. The average of the pairwise differences between the two subclusters, however, was 20.8. Given that the type locality of this species is in eastern Malaysia, we have provisionally referred to the specimens in the Australian subcluster as Carcharhinus cf. sealei, reserving the name Carcharhinus sealei for those belonging to the subcluster that includes the type locality. Four samples of C. sealei from Borneo have voucher specimens (GN4454  =  CAS 229028, GN2962  =  IPPS HBO40, GN2960  =  IPPS HBO35, and GN4188  =  MZB 15.503). In addition, three samples of C. cf. sealei from Western Australia are vouchered (GN4904  =  ANFC H4009-01, GN4905  =  ANFC H6582-06, and GN4906  =  ANFC H6582-09).

Further support for the recognition of these as separate species comes from the haplotype maps. These reveal two distinctly different tight clusters of haplotypes, one representing each species (fig. 79A), with no overlap in the geographic distribution of the haplotypes in these clusters (fig. 79B).

Carcharhinus dussumieri (whitecheek shark) complex (fig. 3)

The analysis included seven specimens initially identified as C. dussumieri from Borneo and the Persian Gulf. Two distinct subclusters resulted from the analysis, one consisting of the specimens from Borneo, and the other of the specimens from the Persian Gulf. The range in pairwise differences observed within the Borneo subcluster was 0–1 and within the Persian Gulf subcluster was 1. However, the mean of the pairwise differences between these widely divergent subclusters was substantial, at 54.7. In recognition of this difference we have designated specimens from the Persian Gulf as Carcharhinus cf. dussumieri, and referred to the specimens from Borneo as C. dussumieri. However, given that the type locality (Pondichéry, India) falls between these two regions, these designations are essentially arbitrary. One of the specimens from Borneo was vouchered (GN4597  =  CAS 229042).

The haplotype maps also support the recognition of two species of whitecheek sharks. The haplotype map colored by phenotype reveals two distinctly different tight clusters of haplotypes, one representing each species (fig. 79A), with no overlap in the geographic distribution of the haplotypes in these clusters (fig. 79B).

Carcharhinus obscurus (dusky shark) (fig. 4) and Carcharhinus galapagensis (Galapagos shark) (fig. 4)

The analysis yielded a cluster comprised of 46 specimens with evidence of two subclusters. Somewhat unexpectedly, the first subcluster consisted of a combination of specimens considered to represent two distinct species. Specifically, these were four specimens identified as C. galapagensis, all collected from Hawaii, as well as 35 specimens of Carcharhinus obscurus collected primarily from Indo-West Pacific localities (i.e., Taiwan, the northern half of Australia, including both west and east coasts, and South Africa), as well as a single specimen from Senegal. Two of the samples from Australia were taken from specimens in the Australian National Fish Collection (GN4908  =  ANFC H 4778-01 and GN4909  =  ANFC H 6358-01). This result suggests that, either the gene used here failed to resolve these two species, or it questions the distinctiveness of these two species. The defining characteristics for distinguishing these two species are habitat and precaudal vertebral counts (Garrick, 1982). The animals identified as C. galapagensis were fully consistent with these criteria. The range in pairwise differences among specimens identified as C. galapagensis was 1–4. However, the average of the pairwise differences between the four specimens identified as C. galapagensis and the specimens of C. obscurus in the first (i.e., Indo-West Pacific and Senegal) subcluster was 3; the average of the pairwise differences between the Atlantic and Indo-West Pacific/Senegal clusters of C. obscurus was 9.3 (see below). If the NADH2 data are correct, it is possible that C. galapagensis may merely represent the oceanic form of C. obscurus.

The second subcluster in this cluster was comprised of seven specimens, representing the majority of those collected from Atlantic localities including the Gulf of Mexico and the western Atlantic. The exception was that the specimen from Senegal clustered with the Indo-Pacific specimens. The ranges in pairwise differences seen within the primarily Indo-West Pacific (excluding specimens identified as C. galapagensis) and Atlantic subclusters were 0–15 and 0–2, respectively. The range of pairwise differences between the two subclusters of C. obscurus (i.e., excluding the specimens identified as C. galapagensis) was 7–21; the average of the pairwise differences (see above) was 9.3. The range in pairwise differences among all 46 specimens was 0–21; the average was 4.7. The range in pairwise differences among all 42 specimens of C. obscurus 0–21; the average was 4.8.

The haplotype map generated for C. obscurus and C. galapagensis (fig. 80A) underscores the lack of distinction between these two species. Not only were two of the three haplotypes exhibited by the four specimens of C. galapagensis shared by specimens of C. obscurus, but the third haplotype of C. galapagensis clustered among haplotypes of specimens identified as C. obscurus. Although the haplotypes of the Atlantic cluster of C. obscurus were divergent from those of the Indo-Pacific specimens, one of the latter specimens (collected from Taiwan) exhibited a haplotype that was conspicuously divergent relative to the other specimens in its cluster, (fig. 80B).

Carcharhinus longimanus (oceanic whitetip shark) (fig. 4)

Our analysis included seven specimens collected from Taiwan and Hawaii. Thus, these specimens represent only a very small portion of the global distribution of this species. These specimens grouped in a single cluster, with a range in pairwise differences among specimens of 0–3 and an average of 1.7. They grouped most closely with the cluster comprised of the specimens of C. obscurus and C. galapagensis. The average of the pairwise differences between specimens in the C. longimanus cluster and those in the C. obscurus and C. galapagensis cluster was 34.8.

Carcharhinus perezi (Caribbean reef shark) (fig. 4)

A total of 14 specimens, from Belize, the Bahamas, and the Gulf of Mexico, were included. These were found to comprise a single cluster with a range in pairwise differences of 0–10 among specimens. One of the specimens from Belize was responsible for much of this difference. If this specimen is excluded, the range in pairwise differences among specimens within the cluster was 0–1. The average of pairwise differences among all 14 specimens was 1.4.

Carcharhinus sorrah (spottail shark) complex (fig. 5)

The analysis included 46 specimens of this species that came from Borneo, India, Singapore, Philippines, Thailand, Vietnam, and northern Australia and thus span much of the eastern regions of the Indo-West Pacific distribution of this species. Three samples from Borneo are represented by vouchers (GN4449  =  CAS 229026, GN2957  =  IPPS HBO32, and GN4185  =  MZB 15.504). The analysis yielded two subclusters, one consisting of the four specimens from Australia, and the other consisting of the remaining 42 specimens. The range in pairwise differences within the Australian subcluster was 1–5; the mean was 3. Within the second subcluster, the range in pairwise differences was 0–17, and the mean 3.8. The mean of the pairwise differences between the two clusters was 15.9, suggesting that this cluster may include two distinct taxa. Given the type locality of C. sorrah is Java, the specimens comprising the Australian subcluster have been provisionally referred to as C. cf. sorrah, reserving C. sorrah for the subcluster that includes specimens from Indonesia. The existence of genetic differences between Australian and Indonesian specimens of C. sorrah was also noted by Ovenden et al. (2009).

The haplotype map colored by phenotype (fig. 81A) shows two distinct clusters consistent with the existence of two species of spottail sharks. The haplotype map colored by geography (fig. 81B) shows interesting contrasting patterns of variation in C. sorrah. While several specimens collected from throughout Borneo and Vietnam share the same haplotype, specimens from Indian waters exhibit substantial variation.

Carcharhinus limbatus (blacktip shark) complex (fig. 6)

A total of 98 specimens identified as C. limbatus were analyzed. These specimens span much of the reported global distribution of this species and include a diversity of localities throughout the Indo-Pacific (i.e., Philippines, Taiwan, Borneo, Vietnam, northern Australia, India, Madagascar, Gulf of California, and Hawaii), as well as several localities in the eastern Atlantic (i.e., Sierra Leone and South Africa) and a diversity of localities in the western Atlantic (including the Gulf of Mexico, Belize, and Puerto Rico). The analysis yielded two distinct clusters, one consisting of the specimens collected from the western Atlantic, and the other consisting of the specimens collected throughout the Indo-Pacific Ocean and the eastern Atlantic. This result is fully consistent with that of Keeney and Heist (2006). The range in pairwise differences among the 39 specimens within the western Atlantic cluster was 0–5, and the average was 0.8; the range in pairwise differences among the 59 specimens within the second cluster was 0–18, and the average was 2.8. The specimens in the two clusters exhibited an average of pairwise differences of 22.2. However, of particular note was the fact that the specimens identified as C. limbatus did not comprise a single cluster independent of the other species of Carcharhinus. In fact, the two clusters of blacktip sharks were separated from one another by a cluster of the 36 specimens of Carcharhinus amblyrhynchoides included in the analysis. Given that the type locality of the blacktip shark is assumed to be Martinque Island, West Indies, specimens comprising the western Atlantic cluster have been given the provisional designation Carcharhinus limbatus, and those comprising the Indo-Pacific and eastern Atlantic cluster have been provisionally designated as Carcharhinus cf. limbatus. We note that one of the specimens from the Philippines was treated by Compagno et al. (2005a) as C. limbatus (i.e., GN2260  =  JPAG 180). A detailed taxonomic revision of this group is required. If the Indo-Pacific C. cf. limbatus is deemed to be a valid species, Carcharhinus pleurotaenia (Bleeker, 1852) might need to be resurrected.

A haplotype map was generated that included C. limbatus, C. cf. limbatus, C. tilstoni, as well as C. amblyrhynchoides given their similarities and the historical confusion surrounding these taxa. The haplotype map for phenotype (fig. 82A) shows four distinctive, relatively tight clusters of haplotypes that correspond to each of these four species. The greatest amount of variation within a species was seen in C. cf. limbatus. For example, that cluster included a specimen from the South China Sea near Borneo that was conspicuously divergent from its conspecifics. The haplotype map for geography (fig. 82B) illustrates the allopatric nature of C. limbatus and C. cf. limbatus. It also shows that three of these species (i.e., all but C. limbatus) cooccur in Australia.

Carcharhinus amblyrhynchoides (graceful shark) (fig. 6)

The analysis included a total of 36 specimens collected from northern Australia (17 specimens), Borneo (14 specimens), Vietnam (1 specimen), and India (4 specimens); these represent much of the eastern half of the Indo-West Pacific distribution of this species. The range in pairwise differences among the 36 specimens was 0–9; the average was 4.3. The analysis can be interpreted to yield three weak subclusters, one for each of these three regions. The range in pairwise differences within the subclusters was 0–4 for Australia, 0–4 for Borneo and Vietnam and 0 for India. The average of the pairwise differences between subclusters was 5.7 between the Australia and Borneo-Vietnam subclusters, 7.9 between the Borneo and India clusters, and 7.8 between the Australia and India clusters. One sample from northern Australia (GN1235  =  NTM S.04689-006) and one from Borneo (GN2959  =  IPPS HBO34) are represented by vouchers. As noted above, the specimens of this species grouped among specimens of the C. limbatus complex. The average of the pairwise differences between C. amblyrhynchoides specimens and those of C. limbatus was 20; and between C. amblyrhynchoides specimens and those of C. cf. limbatus was 17.3.

Carcharhinus tilstoni (Australian blacktip shark) (fig. 6)

Eleven specimens taken from the Timor Sea and the Arafura Sea, off northern Australia, were included in the analysis. A single cluster consisting of these specimens was found. The range in pairwise differences within this cluster was 0–5; the mean was 2. The existence of this cluster, independently from that of the C. limbatus complex, supports C. tilstoni as a valid species (Stevens and Wiley, 1986; Lavery and Shacklee, 1991). However, the vertebral count data collected for these specimens failed to support this as a valid key characteristic for distinguishing between C. tilstoni and C. limbatus. The average of the pairwise differences between C. tilstoni and C. limbatus was 20.6, between C. tilstoni and C. cf. limbatus 24.2, and between C. tilstoni and amblyrhynchoides 19.1.

The haplotype map colored by phenotype in figure 82A underscores the distinctness of C. tilstoni from C. limbatus. The haplotypes of the 11 specimens of C. tilstoni included in the analysis are very similar to one another, but conspicuously different from those of specimens of both C. limbatus and C. cf. limbatus. Haplotypes consistent with C. limbatus appear to be restricted to Australia (fig. 82B). Our results suggest that additional work is required to identify morphological features that allow for the reliable distinction between C. tilstoni and the other species of blacktip sharks.

Carcharhinus fitzroyensis (creek whaler) (fig. 6)

A total of three specimens, all collected from Fog Bay, Australia, were included in the analysis. These comprised a single cluster; the sequences for these three specimens were identical. One of these samples is represented by a voucher in the Northern Territories Museum (GN1267  =  NTM S.14690-002).

Carcharhinus melanopterus (blacktip reef shark) complex (fig. 7)

The analysis included a total of 26 specimens originally identified as this species. These were collected from the Philippines, Thailand, Borneo, the Timor Sea, the Gulf of Carpentaria in northern Australia, and Egypt. Despite our specimens being biased toward the eastern sector of the reported distribution of this species, the analysis yielded two subclusters; one consisting of the two specimens from Egypt, and the other consisting of the 24 specimens from a diversity of localities throughout the Indo-Pacific. The specimens from Egypt differed from one another by one base; the range in pairwise differences within the larger subcluster was 0–9; the average was 4.3. The mean of the pairwise differences between the two subclusters was 15.3. In recognition of this difference, the subcluster from Egypt has been provisionally referred to as Carcharhinus cf. melanopterus, reserving C. melanopterus for the subcluster comprised of specimens more proximal to the type locality (i.e., eastern Indonesia).

Carcharhinus cautus (nervous shark) (fig. 7)

The five specimens of this species included in the analysis were collected from Buffalo Creek in northern Australia and were all identical in sequence. Two of these samples are represented by vouchers in the Northern Territory Museum (GN1233 and GN1234  =  NTM S.14689-005). These specimens grouped most closely with specimens in the C. melanopterus complex. The average of the pairwise differences between C. cautus and those in the latter complex was 25.7.

Carcharhinus leucas (bull shark) complex (fig. 8)

A total of 24 specimens originally identified as Carcharhinus leucas were included in the analysis. These came from a diversity of localities that emphasize the Atlantic, rather than Pacific, elements of the distribution of this species, including the western North Atlantic (e.g., Alabama and Florida) as well as Belize, Sierra Leone, Senegal, and South Africa. In addition, seven of the specimens were collected in Borneo. The analysis yielded three potential subclusters, one consisting of the specimens from the western Atlantic, Belize, Senegal, and Sierra Leone, a second consisting of the three specimens from South Africa, and a third consisting of the seven specimens from Borneo. These results suggest that C. leucas may represent a complex of closely related species. Given that the type locality of C. leucas is in the western North Atlantic (i.e., the Antilles), we have used the provisional designation C. leucas for specimens comprising the primarily Atlantic subcluster. The specimens in the Borneo subcluster are referred to as C. cf. leucas 1. The three specimens in the South African subcluster are referred to as C. cf. leucas 2. The range of pairwise differences among specimens in the C. leucas subcluster was 0–1, among specimens the C. cf. leucas 1 subcluster was 0–12, and among specimens in the C. cf. leucas 2 subcluster was 1–7. The average of pairwise differences within the C. cf. leucas 1 subcluster was 4.8, and within the C. cf. leucas 2 subcluster was 4.7. The average of the pairwise differences between specimens of C. leucas and C. cf. leucas 1 was 17.3, between specimens of C. leucas and C. cf. leucas 2 was 17.1, and between specimens of C. cf. leucas 1 and C. cf. leucas 2 was 13.1.

The haplotype map colored by phenotype (fig. 83A) shows that, although there is notable variation in haplotypes within each cluster, there is no overlap in haplotypes among the three potential species of bull sharks. The haplotype map for geography (fig. 83B) illustrates that the haplotypes of all three bullshark species are allopatrically distributed. Members of this complex would benefit greatly from further investigation.

Carcharhinus amboinensis (pigeye shark) complex (fig. 8)

The 10 specimens of this species included here were collected from India, South Africa, and several localities in western and northern Australia and thus represent much of the reported distribution of this species. The analysis yielded a single cluster consisting of two conspicuous subclusters. One subcluster consisted solely of specimens taken from western and northern Australia, whereas the other subcluster consisted of specimens from India, South Africa, and northern Australia. What makes this pattern somewhat puzzling is the occurrence of specimens from northern Australia in both subclusters, particularly given that some of these specimens were collected from the same exact locality on the same day. The range in pairwise differences within the solely northern Australian subcluster was 0–1, whereas range for the other subcluster was 0–10, with an average of 4. We note that one of the Australian samples came from a specimen deposited in the Australian National Fish Collection (GN4903  =  ANFC H 6655-01). Given that the mean of the pairwise differences between these two subclusters was 26.6, it seems appropriate to nominally recognize both subclusters. In the absence of specimens collected from the type locality (i.e., Ambon Island, Indonesia), we have used the designation C. amboinensis 2 for the solely Australian subcluster, and C. amboinensis 1 for the second subcluster.

The haplotype map for phenotype (fig. 83A) supports the notion that there are two distinct species of pigeye sharks. There is no overlap in haplotypes between specimens of the two putative species. The haplotype map for geography (fig. 83B), illustrates that, although C. amboinensis 1 also occurs in South Africa, haplotypes of both species cooccur in Australia. This complex needs to be explored in more detail.

Carcharhinus plumbeus (sandbar shark) complex (fig. 9)

A total of 74 specimens originally identified as C. plumbeus were included in the analysis. These came from a diversity of localities in the western North Atlantic, as well as Hawaii, Borneo, Vietnam, the Philippines, and Taiwan. However, with respect to the overall reported distribution of C. plumbeus, our sample was conspicuous in its lack of representation from the Indian and eastern Atlantic oceans. Nonetheless, the analysis yielded two distinct clusters of sandbar sharks, one of which included the 22 specimens of Carcharhinus altimus in the analysis. The first cluster of sandbar sharks consisted solely of the 17 specimens from the Indo-Pacific. The second cluster consisted of 57 specimens originally identified as C. plumbeus from the western Atlantic localities, in addition to all of the specimens of C. altimus. The range in pairwise differences observed among the 17 specimens comprising the Indo-Pacific cluster of sandbar sharks was 0–8; the average within this cluster was 1.7. The range in pairwise differences seen among the 57 sandbar shark specimens comprising the western Atlantic cluster (excluding C. altimus) was 0–13; the average was 1.2. The average of the pairwise differences between the Indo-Pacific and the western Atlantic clusters of sandbar sharks (again, excluding C. altimus) was 14.9. Given this result, and the fact that the type locality of C. plumbeus is the Adriatic Sea, we have provisionally given specimens in the cluster collected from the western Atlantic the designation of C. plumbeus and those from the Indo-Pacific the designation of C. cf. plumbeus. If further taxonomic investigation reveals this latter species to be valid, Carcharhinus japonicus (Temminck and Schlegal, 1850) might need to be resurrected. The affinities between Atlantic sandbar sharks and C. altimus, to the exclusion of the Pacific sandbar sharks, have been observed previously by a number of authors (e.g., Heist and Gold, 1999; Greig et al., 2005).

The haplotype map of phenotypes for all 74 sandbar shark specimens and the specimens of C. altimus (fig. 84A) supports recognition of the two sandbar shark species and C. altimus as distinct taxa. However, in the Atlantic C. plumbeus cluster, one specimen exhibited a particularly divergent haplotype relative to its conspecifics (fig. 84A). The geography haplotype map (fig. 84B) illustrates that C. plumbeus and C. altimus are sympatric in the western Atlantic and its environs, while C. cf. plumbeus appears to be restricted to the Indo-Pacific.

Carcharhinus altimus (bignose shark) (fig. 9)

The analysis included 22 specimens of this species, collected from the western North Atlantic, as well as the Gulf of Mexico, Hawaii, and Taiwan. These specimens, which admittedly lack representation of the eastern Atlantic, eastern Pacific, and Indian Ocean elements of the distribution of this species, were found to comprise a single cluster, exhibiting a range in pairwise differences among specimens of 0–9; the average was 1.8. However, as noted above, the C. altimus cluster was nested among the Atlantic specimens of C. plumbeus. Given the substantial morphological differences between C. altimus and the Atlantic form of C. plumbeus, we have treated them as distinct here.

Carcharhinus brevipinna (spinner shark) (fig. 10)

In total, 35 specimens of this species were included. Collectively, these specimens represent much of the distribution of this species, coming from the western Atlantic, Gulf of Mexico, Senegal, Sierra Leone, India, Philippines, Borneo, Vietnam, and Taiwan. The analysis yielded only a single cluster. The range in pairwise differences within the cluster was 0–20 and the average was 6.2.

Carcharhinus brachyurus (bronze whaler) complex (fig. 10)

In total, 10 specimens initially identified as this species were included in the analysis. These were collected from South Africa, Madeira (Portugal), southern Australia, and Taiwan, and thus span much of the distribution of this species, with the exception of the Americas. The analysis yielded two subclusters, one consisting of the specimen from Madeira and the seven specimens from South Africa, and the other consisting of the single specimen from each of Australia and Taiwan. The pairwise difference between the two specimens in the latter subcluster was 2; the range in pairwise differences among specimens within the African subcluster was 0–7, with an average of 2.4. The mean of the pairwise differences between these two subclusters was 23.3. Given that the type locality of C. brachyurus is New Zealand, we have provisionally referred to the smaller subcluster of specimens from Taiwan and southern Australia as C. brachyurus, and have given the specimens in the larger subcluster the designation C. cf. brachyurus. One specimen of C. cf. brachyurus from Portugal is represented by a voucher (GN6628  =  MMF 39543). If further taxonomic investigation reveals the latter taxon as a distinct species, one of the species currently placed in the synonymy of C. brachyurus might need to be resurrected.

Carcharhinus acronotus (blacknose shark) (fig. 10)

All 11 specimens of this species were collected from the western Atlantic and the Gulf of Mexico. The analysis yielded a single cluster; the range in pairwise differences within this cluster was 0–5, with an average of 2.1.

Nasolamia velox (whitenose shark) (fig. 10)

The analysis included a single specimen of this species, collected from Panama. This specimen represents a southern element of the distribution of this species, which occurs from Baja to Peru. It grouped most closely with the specimens of C. acronotus; the average of the pairwise differences between specimens of these two species was 23.7.

Carcharhinus isodon (finetooth shark) (fig. 11)

All 16 specimens of this species included here were collected from the western North Atlantic and the Gulf of Mexico and thus represent only the northeastern elements of the distribution of this species. The analysis yielded a single tight cluster. The range in pairwise differences within this cluster was 0–4, with a mean of 1.3.

Isogomphodon oxyrhynchus (daggernose shark) (fig. 11)

The five specimens included in the analysis were all collected from Maranhao, Brazil. As this species is apparently restricted in distribution to the northeastern regions of South America these specimens are representative of its range. A single cluster resulted from the analysis. Sequences of the five specimens were essentially identical (i.e., the maximum difference seen among specimens was 1).

Carcharhinus porosus (smalltail shark) complex (fig. 11)

Our analysis included a total of 18 specimens, 15 from Trinidad, and three from the Gulf of California. As this species is reported from much of the western Atlantic seaboard, and also from throughout the west coasts of Mexico, Columbia, and Ecuador, our sample is somewhat limited with respect to the overall distribution of this species. Nonetheless, the analysis resulted in two subclusters, one representing the Atlantic and one the Pacific localities. The range in pairwise differences within the Trinidad subcluster was 0–10, with an average of 4.4; the sequences of the three specimens comprising the Gulf of California subcluster were identical. The mean of the pairwise differences between the two subclusters was 18.5. Given these differences, and the fact that the type locality is Brazil, we have provisionally designated the specimens from Trinidad as Carcharhinus porosus, and refer to the specimens from Baja as Carcharhinus cf. porosus. One specimen of the latter taxon is represented by a voucher (GN1107  =  IBUNAM PE9494). A detailed taxonomic revision of this species is required and if the eastern Pacific population is deemed to be a valid species, Carcharhinus cerdale Jordan and Evermann, 1898, might need to be resurrected.

Carcharhinus signatus (night shark) (fig. 12)

Our analysis included a total of six specimens, all taken from the western North Atlantic and the Gulf of Mexico. However, the south and eastern Atlantic elements of the distribution of this species were not sampled. The analysis yielded a single cluster; the range in pairwise differences among the six specimens of this cluster was 0–3, with a mean of 1.7.

Triaenodon obesus (whitetip reef shark) (fig. 13)

In total, the analysis included nine specimens of this species. These came from Hawaii (or possibly Christmas Island), Sulawesi, Borneo, and the Red Sea and Gulf of Aqaba, and thus span much of the distribution of this species with the exception of the eastern Pacific. The analysis resulted in essentially a single cluster. The range in pairwise differences among the specimens comprising this cluster was 0–9; the mean was 4.3

Lamiopsis tephrodes (Borneo broadfin shark) and Lamiopsis temmincki (broadfin shark) (fig. 14)

Lamiopsis tephrodes was originally collected by Fowler (1905), from the Baram River in Sarawak, Malaysia. Until recently, it has been considered to be a synonym of Lamiopsis temmincki (e.g., Compagno, 1984a, 1988), which was originally collected from India. However, L. tephrodes has recently been resurrected by White et al. (2010b) for the Indo-Malay form. Our sample included 26 specimens collected from Borneo (both Malaysian and Indonesian regions), as well as a single specimen collected from India; the analysis showed the 26 specimens from Borneo to cluster outside the specimen from India. Four of the Borneo specimens are represented by vouchers (GN4241  =  CAS 229045, GN4240  =  ANFC H 7083-01, GN4803  =  ANFC H 7084-01, and GN3476  =  IPPS BO259). The range in pairwise differences within the Borneo cluster was 0–3; the average of the pairwise differences between members of the Borneo cluster and the specimen from India was 23.9. Thus, our results support the decision to resurrect L. tephrodes.

Glyphis species (fig. 14)

Specific designations within this genus have recently come under much scrutiny. In these treatments (i.e., Compagno et al., 2008, 2010; Fahmi and Adrim, 2009), the genus has been considered to include a minimum of five valid species: Glyphis garricki recently described by Compagno et al. (2008) from northern Australia and Papua New Guinea ( =  Glyphis sp. C of Compagno and Niem, 1998) and confirmed as distinct from G. glyphis by Wynen et al. (2009); Glyphis glyphis, recently redescribed by Compagno et al. (2008) from material from northern Australia and Papua New Guinea (and includes Glyphis sp. A of Compagno and Niem, 1998, and Last and Stevens, 1994); Glyphis gangeticus from the Ganges River, India (and possibly Pakistan see Compagno et al., 2005b); Glyphis siamensis (Steindachner, 1896) from the Irrawaddy River in Burma; and the recently described Glyphis fowlerae Compagno, White and Cavanagh, 2010 (Glyphis sp. B of Compagno and Niem, 1998), from the Kinabatangan River in Sabah, Malaysian Borneo. In addition, Fahmi and Adrim (2009) reported, as Glyphis sp., a specimen from Sampit, Kalimantan, Borneo, which may represent another undescribed taxon.

Our analysis included a total of 15 specimens of Glyphis: five from the West Alligator River, Australia, collected and identified by Louise McMahon of Charles Darwin University, Darwin, and consisting of three specimens of G. garricki and two of Glyphis glyphis; four specimens of Glyphis fowlerae consisting of the holotype (GN3376  =  IPMB 38.14.02; also Fahmi and Adrim, 2009: fig. 7), a paratype (GN3377  =  IPMB 38.14.03), and a third specimen, all from the Kinabatangan River in Sabah, Borneo, as well as one from Java, Indonesia; three specimens from Pakistan and India, which we have provisionally identified as G. gangeticus, primarily because Pakistan is included among the localities of this species; one specimen from the Bay of Bengal, off Bangladesh, and two from Malaysian Borneo, tentatively identified as Glyphis sp. 1.

The analysis yielded four primary clusters of specimens. One cluster consisted of a subcluster of the four specimens of Glyphis fowlerae and a subcluster of the three specimens identified as Glyphis gangeticus. A second cluster consisted of the three specimens of Glyphis sp. 1. A third cluster consisted of the three specimens of G. garricki. The fourth cluster consisted of the two specimens of G. glyphis. The range in pairwise differences among specimens of G. gangeticus was 0–3, with an average of 2. The range in pairwise differences among specimens of Glyphis fowlerae was 0–6, with an average of 3. The range in pairwise differences among specimens of Glyphis sp. 1 was 5–7, with an average of six bases. The two specimens of G. glyphis differed by two bases, and the three specimens of G. garricki were identical in sequence. The means of the pairwise differences between clusters ranged from 7.5 (G. gangeticus vs. G. fowlerae) to 102.7 (G. glyphis vs. G. sp. 1). Pairwise differences among the five nominal species are as follows. The difference between G. fowlerae and G. gangeticus is 7.5, and Glyphis sp. 1 is 83.3, and G. garricki is 89, and G. glyphis is 85; the difference between G. gangeticus and Glyphis sp. 1 is 81.3, and G. garricki is 88, and G. glyphis is 83; the difference between Glyphis sp. 1 and G. garricki is 100.3, and between Glyphis sp. 1 and G. glyphis is 102.7; the difference between G. garricki and G. glyphis is 54.

In summary, our results support the distinction between G. garricki and G. glyphis despite their sympatry. While the correct name to apply remains uncertain, our results suggest that a species (perhaps G. gangeticus) occurs in Pakistan and India, and is very close to G. fowlerae from Borneo. Furthermore, our results suggest that there exists a species, distinct from all of those included here, that occurs in Bangladesh and Malaysian Borneo. Whether this represents a species new to science or is conspecific with one of those not included in our analysis (e.g., G. siamensis) remains to be determined. We have designated this specimen Glyphis sp. 1 to distinguish it from other recognized but as yet unnamed species of the genus.

Negaprion acutidens (sharptooth lemon shark) (fig. 14)

The analysis included 14 specimens of this lemon shark species, all collected from northern Australia, either the Gulf of Carpentaria or the Timor Sea. The analysis yielded a single tight cluster. The range in pairwise differences among specimens within this cluster was 0–5, with an average of 0.9. Unfortunately, our samples were relatively restricted in distribution and thus did not allow us to assess the variation among localities seen in this species by Schultz et al. (2008).

Negaprion brevirostris (lemon shark) (fig. 14)

In total, six specimens were included in the analysis. These all came from the Gulf of Mexico, the east coast of Florida, and the Caribbean Sea. The analysis yielded a single cluster with a range in pairwise differences among specimens within the cluster of 0–4, with an average of 1.3. Once again, our samples represent only a small portion of the distribution of this species and thus did not allow us to assess variation among localities in this species reported by Schultz et al. (2008). The mean of the pairwise differences between specimens of N. brevirostris and N. acutidens was 85.8.

Rhizoprionodon acutus (milk shark) complex (fig. 15)

In total, 61 specimens originally identified as this species were included in the analysis. These samples were relatively representative of the complete distribution of this species, consisting of 14 specimens from the west coast of Africa, 7 from the Gulf of Oman, 7 from India, 20 from Borneo, 3 from the Philippines, and 10 from northern Australia. The analysis yielded four distinct subclusters. The first cluster consisted of the specimens from west coast of Africa. The second cluster consisted of the specimens from the Gulf of Oman and the majority of the specimens from India. The third subcluster consisted of the specimens collected from Australia, and the fourth of all the specimens from Borneo and the Philippines as well as two of the specimens from India. The ranges in pairwise differences within these subclusters were 0–2 (western Africa subcluster), 0–6 (Gulf of Oman etc. subcluster), 0–8 (Australia subcluster), and 0–8 (Borneo, etc.); the means were 0.6, 1.5, 2.4, and 3.2, respectively. The means of the pairwise differences among subclusters were as follows: 10 (western Africa vs. Gulf of Oman), 16.6 (western Africa vs. Australia), 17.6 (western Africa vs. Borneo), 15.6 (Gulf of Oman vs. Australia), 18.8 (Gulf of Oman vs. Borneo), and 15.3 (Australia vs. Borneo). Based on these differences we have given the four subclusters separate designations. Since the type locality of R. acutus is the Red Sea, the specimens in the cluster from the Gulf of Oman and India have been given the designation R. acutus. The remaining three subclusters have been referred to as: Rhizoprionodon cf. acutus 1 (western Africa cluster), Rhizoprionodon cf. acutus 2 (Australia cluster), and Rhizoprionodon cf. acutus 3 (Borneo, etc., cluster). We note that two of the specimens of R. cf. acutus 2 came from specimens in the Australian National Fish Collection (GN4918  =  ANFC H 4559-01 and GN4919  =  ANFC H 6582-12) and one specimen of R. cf. acutus 3 came from the Institut Penyelidikan Perikanan Sarawak in Kuching, Malaysia (GN2955  =  IPPS HBO30). This potential complex of species is in need of further investigation. If a taxonomic revision reveals that the three subclusters listed above are distinct from R. acutus, the following species currently placed in synonymy with R. acutus might need to be resurrected as valid species: R. crenidens (Klunzinger, 1880) for the Australian species; R. fissidens (Bennett, 1831) for the western Africa species; R. walbeehmi (Bleeker, 1856) for the western central Pacific species.

A haplotype map colored by phenotype (fig. 85A) supports the idea that R. acutus is a complex of four species. It shows four relatively tight haplotype clusters consistent with the four species treated above. The haplotype map colored by geography (fig. 85B) indicates that all four species are allopatric, with the exception of the cooccurrence of R. acutus and R. cf. acutus 3 in India.

Rhizoprionodon terraenovae (Atlantic sharpnose shark) (fig. 16)

Our analysis included a total of 24 specimens, which came from the western North Atlantic, Gulf of Mexico, Caribbean Sea, and Belize, and thus represent much of the distribution of this species. The range in pairwise differences among specimens within this relatively divergent cluster was 0–16; the mean was 4.9.

Rhizoprionodon porosus (Caribbean sharpnose shark) (fig. 16)

Two specimens were included. Both came from Trinidad, located in the northern region of the distribution of this species, which extends throughout the eastern coast of South America. These two specimens differed in sequence by two bases. They were found to cluster most closely with the specimens of R. terraenovae. The average of the pairwise differences between specimens of R. porosus and those of R. terraenovae was 17.9.

Rhizoprionodon longurio (Pacific sharpnose shark) (fig. 16)

The eight specimens of this species included in the analysis were all collected from the Gulf of California and thus represent only the more northern region of the distribution of this species, which extends south from Baja along the western coast of North and Central America to Peru. These specimens comprised a relatively divergent cluster; two of the specimens (both collected from Bahia de Los Angeles) grouped together to the exclusion of the remaining six specimens (5 from Puertecitos and 1 from San Jose del Cabo). The range in pairwise differences among all eight specimens in the cluster was 1–8, with an average of 4.2; the two specimens from Puertecitos differed by 1; the range in pairwise differences among the remaining six specimens was 0–7.

Rhizoprionodon lalandii (Brazilian sharpnose shark) (fig. 16)

The six specimens of this species included in the analysis came from Trinidad. They represent the more northern regions of the distribution of this species. The analysis yielded a single cluster with the range in pairwise differences observed among specimens in this cluster being 0–2.

Rhizoprionodon taylori (Australian sharpnose shark) (fig. 16)

A total of 14 specimens, collected from the Timor Sea and the Gulf of Carpentaria off northern Australia, was included. The analysis yielded a single cluster, with a range in pairwise differences of 0–9 among specimens, with a mean of 4.4. One sample in this cluster was taken from a specimen in the Australian National Fish Collection (GN4920  =  ANFC H 6655-02) and two from samples from specimens in the Northern Territory Museum (GN1254 and GN1255  =  NTM S.14690-001).

Rhizoprionodon oligolinx (gray sharpnose shark) (fig. 17)

In total, 17 specimens were included; these came from Borneo and India and thus are generally representative of the distribution of this species. These specimens comprised a loose cluster with some structure reflecting these two regions. The range in pairwise differences among all 17 specimens was 0–15, with an average of 5.8. Two specimens are represented by vouchers (GN3685  =  IPPS BO474 and GN3686  =  IPPS BO475).

Scoliodon species (fig. 18)

Our analysis included 61 specimens of Scoliodon: 29 from Borneo, 17 from Vietnam, 2 from Taiwan, 11 from India, and 2 from the Bay of Bengal, east of Thailand. Up until recently, this genus was considered monotypic (e.g., Compagno et al., 2005a), containing only Scoliodon laticaudus. However, White et al. (2010a) resurrected Scoliodon macrorhynchos from Java, Indonesia as a valid species. Our analysis yielded three distinct clusters of Scoliodon. The first cluster consisted of the S. macrorhynchos specimens from Borneo, Vietnam, and Taiwan; the range in pairwise differences among these specimens was 0–8, with an average of 2.2. Two of these specimens are deposited in the IPPS (GN3449  =  IPPS BO222 and GN3450  =  IPPS BO223). The second cluster consisted of the S. laticaudus specimens from India, which had a range in pairwise differences among specimens of 0–2. The third cluster consisted of the two specimens from the Bay of Bengal west of Thailand; their sequences differed by 5. Between-cluster differences were found to be substantial and not only supported S. laticaudus and S. macrorhynchos as separate species, but suggested the specimens from Thailand represent a distinct species. The latter have been given the designation Scoliodon cf. laticaudus, but taxonomic investigation is required to determine whether they represent an undescribed species or whether S. muelleri (Müller and Henle, 1839) should be resurrected as a valid species. The average of the pairwise differences between the S. laticaudus and S. macrorhynchos clusters was 32.1, between the S. laticaudus and S. cf. laticaudus clusters 33.8, and between the S. macrorhynchos and S. cf. laticaudus clusters 40.4.

The haplotype map colored by phenotype (fig. 86A) supports recogniton of three species of Scoliodon; in each case haplotypes of replicate specimens within a species are tightly clustered and conspicuously different from specimens in other species clusters. The haplotype map colored by geography (fig. 86B) illustrates that, although our specimes of S. macrorhynchos were collected from a diversity of localities in the central Indo-Pacific, they exhibit relatively little haplotype variation in NADH2.

Loxodon macrorhinus (sliteye shark) complex (fig. 18)

In total, 21 specimens of Loxodon were included in the analysis. These came from Borneo (14 specimens), India (4 specimens), the Philippines (2 specimens), and Madagascar (1 specimen). At present this genus is considered to include only the single valid species, Loxodon macrorhinus, with the type locality unknown (probably Indian Ocean). Our analysis yielded two clusters, one consisting of two of the specimens from India and all the specimens collected from Borneo and the Philippines (with a range in pairwise differences among specimens of 0–11, and an average of 4.6), and the other consisting of the remaining two specimens from India and a specimen from Madagascar (with a range in pairwise differences among specimens of 0–1). We have provisionally referred to specimens in the former cluster as Loxodon macrorhinus, whereas those in the latter cluster have been given the designation L. cf. macrorhinus; the average of the pairwise differences between these clusters was 29.9. Our results support the existence of a second species in this currently monotypic genus. Further taxonomic research is required to determine which species represents the true L. macrorhinus, and whether the second species truly represents an undescribed taxon.

Sphyrnidae (hammerhead sharks)

Our analysis includes representation of seven of the eight described species of hammerhead sharks. However, we believe it also includes specimens of three undescribed species of hammerheads. To allow comparison of genetic variation across species, the haplotype maps presented for hammerheads (fig. 87A, B) include all 10 of these species. The implications of these maps for the three species complexes (S. lewini, S. zygaena, and S. tiburo) are each treated separately below. However, from the standpoint of the family in general, it is interesting to note that the haplotype map colored by geography (fig. 87B) illustrates that hammerheads are among the most widely distributed species of elasmobranchs. For example, the same haplotype of S. lewini 1 is found in animals from the western Indian Ocean, India, and South China Sea. Similarly, the same haplotype of S. zygaena is found in animals from Taiwan, Borneo, and northern Australia. Little variation is seen in the haplotypes of individuals of S. lewini 2 that occur in Borneo, Taiwan, and the Gulf of California. Little variation is seen in the haplotypes of individuals of S. zygaena from the Gulf of Mexico, western North Atlantic, Senegal, Japan, Taiwan, Vietnam, and the Gulf of California. Nevertheless, it seems there is a great deal of sympatry. For example, the western Atlantic (including the Gulf of Mexico and Trinidad) is home to six species of hammerheads (S. lewini 2, S. mokarran 1, S. tiburo, S. cf. tiburo, S. tudes, and S. zygaena).

Sphyrna lewini (scalloped hammerhead) complex (fig. 19)

Our analysis included a total of 45 specimens initially identified as the scalloped hammerhead S. lewini. These came from the western North Atlantic (11 specimens), the Gulf of Mexico (6 specimens), Senegal (4 specimens), Madagascar (3 specimens), India (5 specimens), Borneo (10 specimens), Gulf of California (2 specimens), and Taiwan (4 specimens). Given that the type locality of this species is southern Australia, the unavailability of specimens from Australia was unfortunate. Our analysis yielded considerable structure among S. lewini. First, the analysis yielded two strongly divergent clusters, each with some substructure. However, at this time we have recognized only the two main clusters. The 32 specimens comprising the first cluster, from the western Atlantic, Gulf of Mexico, Senegal, Madagascar, India, and Malaysian Borneo, have been designated Sphyrna lewini 1, and the 13 specimens comprising the second cluster, from the Gulf of California, Borneo, and Taiwan, have been designated Sphyrna lewini 2; one specimen from the latter cluster is vouchered (GN4187  =  CAS 229024). Within the S. lewini 1 cluster, the specimens from the Gulf of Mexico and the western North Atlantic comprised a weak subcluster; the specimens from Senegal comprised a second weakly supported subcluster, and the specimens from India, Madagascar, and those from Malaysian Borneo comprised a third weakly supported subcluster. In addition, the specimens from Madagascar and India and three of the specimens from Borneo comprised a subcluster. Within the Sphyrna lewini 2 cluster, there was evidence of a subcluster consisting of a specimen from Taiwan and one from Malaysian Borneo. The range in pairwise differences among all 45 specimens of S. lewini was 0–78. The range in pairwise differences within the S. lewini 1 cluster was 0–12, with an average of 4.4. The range within the S. lewini 2 cluster was 0–21, with an average of 7.2. The average of the pairwise differences between specimens of S. lewini 1 and S. lewini 2 was 64.2. These results suggest that undescribed diversity exists among scalloped hammerheads, which may include sympatric species. This result is consistent with those of a number of previous authors who also reported genetic diversity within S. lewini (e.g., Abercrombie et al., 2005; Duncan et al., 2006; Quattro et al., 2006; Zemlak et al., 2009).

The haplotype map colored by phenotype (fig. 87A) supports recognition of S. lewini 1 and S. lewini 2 as they exhibit distinct, but relatively tight clusters of haplotypes. The haplotype map colored by geography (fig. 87B) illustrates the relatively broad distributions of both of these clusters of haplotypes (fig. 87B).

Sphyrna tiburo (bonnethead shark) complex (fig. 19)

All 14 specimens originally identified as S. tiburo were collected from the Gulf of Mexico (12 specimens) and Trinidad (2 specimens). As a consequence, our sample represents only a portion of the distribution of this species along the western Atlantic seaboard, and we have no representation of the eastern Pacific portions of the distribution of this species. The analysis yielded some geographic structure in that the specimens from the Gulf of Mexico clustered together as did those from Trinidad. The range in pairwise differences among bonnetheads overall was 0–28. The range in pairwise differences within the Gulf of Mexico cluster was 0–3, with an average of 0.6, and the two specimens from Trinidad differed by 4. The average of the pairwise differences between the two clusters was 25.3. In recognition of the fact that the type locality of S. tiburo is America, we have given the Trinidad specimens the designation Sphyrna cf. tiburo. However, the type locality of S. tiburo is imprecise so the nomenclature of this group needs to be examined more thoroughly.

The haplotype map colored by phenotype (fig. 87A) supports the distinction between S. tiburo and S. cf. tiburo.

Sphyrna tudes (smalleye hammerhead) (fig. 19)

Our analysis included four specimens, all collected from Trinidad. Thus, our sample comes from a relatively northerly locality within the distribution of this species, which extends along the east coast of South American to Argentina. The analysis yielded a single cluster. The range in pairwise differences within the cluster was 0–6, with an average of 3.

Sphyrna corona (mallethead shark) (fig. 19)

The six specimens included in the analysis all came from the western coast of Panama. Thus, they represent the central region of the distribution of this species, which extends along the western seaboard of the Americas from the Gulf of California to Peru. The analysis yielded a single cluster. The range in pairwise differences within this cluster was 0–7, and the average was 3.3.

Sphyrna mokarran (great hammerhead) complex (fig. 19)

In total, 22 specimens were included. These were collected from the Gulf of Mexico (9 specimens), the western North Atlantic coast from Massachusetts to Florida (7 specimens), Malaysian Borneo (1 specimen; GN3471  =  IPPS BO254), and northern Australia (5 specimens). Given that this species has been reported to occur in a global band (Last and Stevens, 2009), on continental shelves throughout the tropics and subtropics, our sample largely underrepresents the distribution of this species, and does not include the type locality (Red Sea). Nonetheless, the analysis yielded two distinct clusters: one comprised of the specimens collected from the Atlantic, which we refer to as Sphyrna mokarran 1, and a second consisting of specimens from Australia and Borneo, which we refer to as Sphyrna mokarran 2. The range in pairwise differences among S. mokarran specimens overall was 0–16. The range in pairwise differences among specimens of S. mokarran 1 was 0–3 (with an average of 0.5) and the range in pairwise differences among specimens of S. mokarran 2 was 0–4 (with an average of 1.9). However, the average of the pairwise differences between these two clusters was 14.1.

Although they are not among the most divergent of the hammerhead species complexes treated here, there is no haplotype overlap between specimens of S. mokarran 1 and S. mokarran 2 (fig. 87A), which supports recognition of these as distinct allopatric species (fig. 87B).

Sphyrna zygaena (smooth hammerhead) (fig. 19)

The 16 specimens of this species were collected from the Gulf of California (4 specimens), western North Atlantic (6 specimens), Senegal (1 specimen), Vietnam (1 specimen), Taiwan (3 specimens), and Japan (1 specimen), and thus represent at least the longitudinal (if not the latitudinal) distribution of this species. The analysis yielded essentially a single cluster, with a range in pairwise differences among specimens of 0–8, with an average of 2.5. One of the specimens from the Gulf of California was vouchered (GN1097  =  IBUNAM PE9519).

Eusphyra blochii (winghead shark) (fig. 19)

All nine of our specimens of this species were all collected from Fog Bay, in the Timor Sea of northern Australia and thus represent only a small portion of the distribution of this species, which includes much of the coastal regions of the Indo-West Pacific. The analysis yielded a single cluster of specimens with a very low range in pairwise differences among specimens (i.e., 0–2); one of these samples was vouchered (GN1256  =  NTM S.14689-004).

Carcharhinidae (requiem sharks), continued

Galeocerdo cuvier (tiger shark) complex (fig. 20)

Our 29 specimens of this species represent much of the longitudinal distribution of this species, having come from the western North Atlantic and Gulf of Mexico (18 specimens), as well as the Gulf of California (1 specimen), Hawaii (1 specimen), Borneo (1 specimen), the Timor Sea off the coast of northern Australia (7 specimens), and the Red Sea (1 specimen). The specimen from the Gulf of California was deposited in the Texas Cooperative Wildlife Collection (GN5271  =  TCWC 7574.01). The analysis yielded two clusters of tiger sharks. One of these consisted of specimens from the Atlantic localities, and the second consisted of the specimens from Australia, Hawaii, Borneo, Gulf of California, and Red Sea (i.e., the Pacific and Indian Ocean localities). The range in pairwise differences among all 29 tiger shark specimens was 0–17; the average was 6.2. The range in pairwise differences among specimens in the Atlantic cluster was 0–5 (with an average of 1.1), and among specimens in the Pacific and Indian Ocean cluster was 0–7 (with an average of 2.9). The average of the pairwise differences between the two clusters was 11.1. Given that the type locality of G. cuvier is northwestern Australia, we have given the 11 specimens comprising this cluster the designation G. cuvier and the 18 specimens in the second cluster the designation G. cf. cuvier.

Both haplotype maps for this genus (figs. 88A and B) support the above conclusions. There is no haplotype overlap among specimens of the two potential species of tiger sharks, and the haplotype map colored by geography clearly shows that the two species are allopatrically distributed. The notion that multiple species may exist within this genus needs to be further explored. If taxonomic investigation reveals that Atlantic populations are not conspecific with G. cuvier, G. arcticus (Faber, 1829), described from Iceland and Norway, might need to be resurrected.

Hemigaleidae (weasel sharks)

Hemigaleus microstoma (sicklefin weasel shark) (fig. 21)

In total, 31 specimens of this species were included. They consist of one specimen collected from Singapore, six from the Philippines, and 24 from Borneo and thus represent only some of the more central elements of the Indo-West Pacific distribution of this species. Two of the specimens from the Philippines were treated by Compagno et al. (2005a) (i.e., GN4324  =  BRU 123 and GN4325  =  JPAG 216) as H. microstoma. One of the specimens from Borneo was also vouchered (GN3694  =  IPPS BO483). These 31 specimens comprised a single cluster, including a weakly supported subcluster comprised of a subset of specimens from Borneo. The range in pairwise differences among all 31 specimens of this species was 0–13 and the average was 4.9.

Hemigaleus australiensis (Australian weasel shark) (fig. 21)

Our analysis included six specimens of this newly described species (see White et al., 2005) from Western Australia, and the Timor and Arafura seas off the coast of northern Australia. Two of these specimens came from the Australian National Fish Collection (the holotype GN4913  =  ANFC H 5949-01 and GN4914  =  ANFC H 5949-02). Within this cluster, two specimens, one from Western Australia and one from the Arafura Sea, grouped together in a subcluster distinct from the other four specimens. The range in pairwise differences among all six specimens of this species was 0–11, with an average of 6.5. The two specimens in the former subcluster differed by one base, and the range in pairwise differences among the remaining four specimens was 0–5 (with an average of 3.2). The mean of the pairwise differences between the two subclusters of H. australiensis was 9.8. The mean of the pairwise differences between specimens of H. australiensis and those of H. microstoma was 94.6; this result strongly supports the distinction between these two congeners.

Paragaleus sp. (fig. 21)

A single specimen of Paragaleus from Phuket, Thailand, grouped outside both of the known species of Paragaleus, along with the two species of Hemigaleus. The average of the pairwise differences between this specimen and those of P. randalli was 128.9. The average of the pairwise differences between this specimen and those of P. pectoralis was 122.7. We have given this specimen the designation Paragaleus sp. It possibly represents an undescribed Paragaleus species, however, comparisons with the two described species of Paragaleus not included here would also be valuable. Given its substantial difference from the other two Paragaleus species, it may, however, be a case of incorrect generic identification, and the possibility that it represents a species of Chaenogaleus should be explored.

Paragaleus randalli (slender weasel shark) (fig. 21)

Our analysis included a total of 17 specimens belonging to this species, two of which were vouchered (GN4182  =  MZB 15.506 and GN4191  =  MZB 15.507). The analysis yielded a single cluster. The specimens in this cluster included 16 samples from Borneo and a single specimen from the Philippines. The range in pairwise differences among specimens in this cluster was 0–9; the average of pairwise differences among specimens of this species was 2.9. All of the localities sampled in this study fall well outside the originally described range of Paragaleus randalli (i.e., Bahrain and western Indian Ocean), but within the range of the similar looking congener Paragaleus tengi (straighttooth weasel shark) which might lead to the suspicion that these tissue samples were actually derived from specimens of P. tengi. However, unpublished sequence data from NADH2 indicate no significant differences in sequence between specimens identified as P. randalli from the Persian Gulf and those of the 17 specimens from Southeast Asia used in the current study. Furthermore, vertebral counts examined (by PL and WW) from three specimens originally identified as Paragaleus tengi from Borneo, range in number between 164 and 171 which is outside the reported range for P. tengi (131–135) but consistent with the range given for P. randalli (165–186). Taken together, these observations suggest that P. randalli has a more extensive distribution than suggested in its original description (Compagno et al., 1996) extending from the western Indian Ocean to Southeast Asia. They also suggest that P. randalli, rather than P. tengi, is the dominant form of Paragaleus in Borneo.

Paragaleus pectoralis (Atlantic weasel shark) (fig. 21)

The analysis yielded a cluster consisting of six specimens of this species, all from western Africa (i.e., from Mauritania, Senegal, and Sierra Leone). These specimens are generally representative of the distribution of this species, which extends throughout the northern half of the western coast of Africa. The range in pairwise differences among specimens within this cluster was 0–5, with an average of 2.9. The average of the pairwise differences between specimens of P. pectoralis and P. randalli was 118.

Hemipristis elongata (snaggletooth shark) (fig. 21)

Our analysis included a total of 14 specimens of this monotypic genus, nine from Borneo and five from the Arafura Sea off northern Australia; these represent only a subset of the distribution of this species which includes much of the Indo-West Pacific, with the Red Sea as the type locality. One of these specimens is represented by a voucher (GN4195  =  CAS 229035). The analysis yielded a single cluster. The range in pairwise differences among specimens was 0–5; the average was 2. There is some evidence of geographic substructure within this cluster with the Australian specimens grouping apart from the Borneo specimens, however the signal is weak as the average of the pairwise differences between specimens from the two localities was only 3.5.

Leptochariidae (barbeled houndsharks)

Leptocharias smithii (barbeled houndshark) (fig. 22)

All three of our specimens were collected off of Senegal and represent the northern region of the distribution of this species which extends south from there along the coast of western Africa to Angola. The analysis yielded a single cluster; the range in pairwise differences among these three specimens was 0–1.

Triakidae (houndsharks)

Mustelus (smoothhounds) (figs. 23, 24)

This was among the most problematic of genera included in our analysis. At present, the genus includes ∼31 species of which five have been described since 2005; clearly a substantial amount of diversity in this genus remains to be explored. Analysis of 129 specimens yielded 18 clusters, each of which we believe represents a distinct species of Mustelus. However, assigning correct names to clusters without vouchers was problematic; particularly challenging were the specimens collected from the Gulf of California. For nine clusters, photo vouchers are available for one or more of the included specimens. For three additional clusters, identifications were based on one or more museum specimens. In the remaining six cases we have relied on geographical range and/or the taxonomic expertise of individuals providing samples. Ultimately, while we are confident in the existence of 18 species among these specimens, the specific identifications of many of the clusters remain to be confirmed. We hope that the generation of comparable sequence data for additional specimens of known identity will serve to place the data presented here into a broader perspective. Each of the 18 species clusters is treated separately below.

Mustelus widodoi (whitefin smoothhound) and Mustelus ravidus (Australian gray smoothhound) (fig. 23)

In total, 31 specimens were analyzed, consisting of 30 specimens of M. widodoi from Borneo and one specimen of M. ravidus from a specimen deposited in the Australian National Fish Collection (GN4898  =  ANFC H 5947-01). The analysis yielded a cluster consisting of the 30 M. widodoi specimens from Borneo, with the M. ravidus specimen from Australia grouping outside. The specimens from Borneo were consistent morphologically with M. widodoi of White and Last (2006); however, they were collected further north than the type specimens and represent a distributional extension for this species (see Last et al., 2010c). The range of pairwise differences among the M. widodoi specimens was 0–5; the average was 1. The average of the pairwise differences between M. widodoi and M. ravidus was 8.5. We note that M. ravidus was referred to as Mustelus sp. A by Last and Stevens (1994) and Gardner and Ward (2002).

Mustelus mosis (Arabian smoothhound) (fig. 23)

The three specimens included in the analysis were collected from India and thus represent the eastern region of the distribution of this species, which extends from India, westward to Somalia, and possibly as far southwest as South Africa. The analysis yielded a single cluster. The range in pairwise differences within this cluster was 3–9; the average was 6.7. This is one of only two species of Mustelus known to occur in India. These specimens were found to group well outside those of the other Indian species, M. manazo, the identity of which is grounded by specimens.

Mustelus mustelus (smoothhound) (fig. 23)

In total, nine specimens of this species were included in our analysis, three from Senegal, one from Angola, and five from South Africa. Thus, these specimens come from much of the known distribution of this species, which extends from the North Sea along the coast of Europe and the western coast of Africa to South Africa. The analysis yielded a single cluster, with some evidence of substructure. The range in pairwise differences among all nine specimens was 0–8; the average was 3.4. The range in pairwise differences among the specimens from Senegal was 1–3 (with an average of 2). The range among the other five specimens was 0–1 (with an average of 0.3). The average of the pairwise differences between specimens from Senegal and those from southern Africa was 6.2. The identities of a number of specimens in this cluster are grounded with photographs.

Mustelus cf. lunulatus (sicklefin smoothhound) (fig. 23)

The analysis yielded a cluster of eight specimens that we have tentatively identified as M. cf. lunulatus. The range in pairwise differences among specimens was 0–7, with an average of 2.9. The identity of this cluster is problematic. All eight specimens were collected from the Gulf of California and all possess teeth with low, rounded cusps. They were initially identified as M. lunulatus, but dissections and x-rays revealed a number of inconsistencies. For example, while two of the specimens (BJ-802 and BJ-796) have precaudal vertebral counts (77 and 81, respectively) that are within the range for this species (74–82) presented by Heemstra (1973), their monospondylic vertebral count (37 in both cases) is outside the range for M. lunulatus (28–34) presented by Heemstra (1973). Moreover, the cluster includes a specimen (BJ-803) that has a precaudal vertebral count (87) that falls within the range for M. intermedius (83–93) of Heemstra (1973) but has a monospondylous vertebral count (41) that falls within the range (41–44) for Heemstra's “M. platyrhinus.” The identity of this cluster should be considered tentative until confirmed specimens of M. lunulatus can be included in the analysis.

Mustelus canis (dusky smoothhound) (fig. 23)

The analysis included eight specimens identified as this species; seven were collected from the western North Atlantic coast from Rhode Island to North Carolina and one from the Bahamas. Only some of the Northern Hemisphere elements of the distribution of this species are represented and the Southern Hemisphere elements are unrepresented. The analysis yielded a single cluster, with the specimen from the Bahamas grouping outside those from the more northern latitudes. The range in pairwise differences among the eight specimens within this cluster was 0–23; the average was 7.9. Some consideration should be given to the possibility that additional species of Mustelus may be represented by these specimens. Heemstra's (1997) recognition of Mustelus canis insularis from several islands of the Caribbean, may, for example be relevant to the identity of the specimen included here from the Bahamas.

Mustelus sp. 2 (fig. 23)

The analysis yielded a cluster of four specimens from the Gulf of California with a range in pairwise differences among specimens of 0–5 and an average of 2.5. We have been unable to definitively identify this cluster to species. Like candidate species from the Gulf of California, these specimens possess teeth with low rounded cusps. However they lack the white posterior margins of the dorsal, pectoral, pelvic, and anal fins described by Castro-Aguirre et al. (2005) in M. albipinnis and also described by Pérez-Jiménez et al. (2005) in M. hacat, a potential junior synonym of the former species (Eschmeyer and Fricke, 2011). Moreover, all four specimens were collected by artisanal fishermen using bottom gill nets and thus, unlike specimens of M. albipinnis, came from waters substantially shallower than 100 m. While one of the specimens (GN1565  =  BJ-671) in this cluster has a monospondylic vertebral count (41) that is generally consistent with that of M. albipinnis (40), it has a precaudal vertebral count (97) that is above that reported by Castro-Aguirre et al. (2005) for M. albipinnis (i.e., 92), and slightly lower than the range (101–102) reported in the original description of M. hacat. The possibility that this species represents one of the two undescribed species from the Gulf of California treated by Heemstra (1973) should not be ruled out.

Mustelus henlei (brown smoothhound) (fig. 23)

Material from the Gulf of California also included six specimens of Mustelus henlei. These specimens represent only a small portion of the distribution of this relatively easy to identify species, which has been reported from as far north as Washington state and as far south as Peru. Analysis of these specimens yielded a single cluster with a range in pairwise differences of 2–9, and an average of 5.6. The identity of this cluster is grounded with images.

Mustelus norrisi (narrowfin smoothhound) (fig. 23)

A single specimen, collected and identified by NOAA Fisheries biologist Lisa Jones, was included in the analysis. This specimen grouped independently of all of the other specimens of Mustelus included in the analysis. However, clearly this specimen represents only one point in the distribution of this species which extends along the eastern coast of North and South America, from Florida to Venezuela, and Argentina.

Mustelus californicus (gray smoothhound) (fig. 23)

Our analysis included 15 specimens of this species, all collected from the Gulf of California. Given that this species occurs along coastal California, the Gulf of California, and possibly as far south as Puerto Vallarta, our specimens represent the center of the relatively limited overall distribution of this species. Three of these specimens were deposited in the Texas Cooperative Wildlife Collection (GN2284  =  TCWC 7561.03, GN5291  =  TCWC 7561.05, and GN5292  =  TCWC 7561.04). The analysis yielded a single tight cluster. The range in pairwise differences among specimens within this cluster was 0–2 and the average of the pairwise differences among specimens was 1.4. The identity of this cluster is grounded not only with photographs, but also in that dissections and/or x-rays revealed that specimens had an extra symphyseal cartilage on the palatoquadrate, which other sympatric Mustelus species lack (Compagno, 1984a).

Mustelus sp. 1 (fig. 23)

The analysis yielded an additional cluster of two specimens from the Gulf of California, which were identical in sequence. We have been unable to definitively identify this cluster to species. These specimens, like those of Mustelus sp. 2, possess teeth with low, rounded cusps and lack the white posterior margins of the dorsal, pectoral, pelvic and anal fins described by Castro-Aguirre et al. (2005) in M. albipinnis as well as in its potential junior synonym M. hacat (see Pérez-Jiménez et al., 2005). Like the four specimens of Mustelus sp. 2, both specimens were also collected by artisanal fishermen using bottom gill nets and thus, unlike M. albipinnis, came from waters substantially shallower than 100 m. The possibility that these specimens represent one of the two undescribed species from the Gulf of California treated by Heemstra (1973) can also not be ruled out. These specimens clustered together, independent of the remaining Mustelus species. They clustered most closely with M. californicus; the average of pairwise differences between these two species was 40.1.

Mustelus antarcticus (gummy shark) (fig. 24)

A total of nine specimens of this species were included in the analysis. All nine came from southeastern and southwestern Australia and are representative of much of the range of this species. The analysis yielded a single cluster, with a range in pairwise differences of 0–4 and an average of 1.9. The identity of this cluster is based on a specimen in the Australian National Fish Collection (GN4902  =  ANFC H 6571-03).

Mustelus lenticulatus (spotted estuary smoothhound) (fig. 24)

This species occurs throughout New Zealand, which was the site of collection of all three specimens included in the analysis. The range in pairwise differences seen within the cluster that resulted from the analysis was 0–1. The identity of this cluster is based on two specimens in the Australian National Fish Collection (GN4896  =  ANFC H 5551-01 and GN4897  =  ANFC H 5551-02).

Mustelus stevensi (whitespotted gummy shark) (fig. 24)

Two specimens of this recently described species (see White and Last, 2006) were included in the analysis. Both specimens were collected from the type locality, from the Dampier Archipelago off northwestern Australia. These clustered together, independently of the remaining Mustelus species. The sequences of the two specimens differed by 2. The identity of this cluster is confirmed given that one of the samples comes from a paratype (GN4900  =  ANFC H 4650-01); the second sample comes from a voucher (GN4899  =  ANFC H 4649-08). The average of the pairwise differences between M. stevensi and M. lenticulatus was 14.3, between M. stevensi and M. antarcticus was 17.2, and between M. stevensi and M. ravidus was 83. Our results support the clear distinctions reported by Gardner and Ward (2002) between M. stevensi ( =  Mustelus sp. B of Last and Stevens, 1994, and Gardner and Ward, 2002) and each of M. lenticulatus, M. antarcticus, and M. ravidus ( =  Mustelus sp. A of Last and Stevens, 1994, and Gardner and Ward, 2002).

Mustelus manazo (starspotted smoothhound) (fig. 24)

In total, 13 specimens of this species were included in the analysis. These consist of seven specimens from Taiwan, two from Vietnam, three from Japan, and one specimen that was landed in Borneo. The analysis yielded a single cluster. The range in pairwise differences within this cluster is 0–6 and the average was 2.8. However, our specimens represent only the eastern central elements of the distribution of this species. Four samples from this cluster are represented by specimens (GN972, GN989, GN1019 and GN1020  =  UMMZ 231357).

Mustelus palumbes (whitespot smoothhound) (fig. 24)

Six specimens of this species, all collected from South Africa, were included. These are representative of the distribution of this species which occurs from Namibia to central Natal. The range in pairwise differences of this cluster was 0–4, and the average was 2.6.

Mustelus asterias (starry smoothhound) (fig. 24)

The analysis included two specimens, both collected from the eastern North Atlantic. The sequences of these two specimens differed by one base. However, our specimens represent only the northernmost region of the distribution of this species, which extends along the coast as far south as Senegal. These specimens clustered most closely with those of M. palumbes. The average of the pairwise differences between specimen of these two species was 8.3.

Mustelus schmitti (narrownose smoothhound) (fig. 24)

In total, seven specimens of this species, all collected and identified by Gustavo Chiaramonte of the Museo Argentino de Ciencias Naturales, Buenos Aires, were included. The analysis yielded a single cluster, with a range in pairwise differences among specimens of this species of 0–4 and an average of 1.9. These specimens are fairly representative of the distribution of this species, which extends along the coasts of southern Brazil, Uruguay, and Argentina.

Scylliogaleus quecketti (flapnose houndshark) (fig. 24)

The analysis included four specimens of this species, all from South Africa; one donated by Mark Harris and three others by Geremy Cliff of the KwaZulu-Natal Sharks Board. The analysis resulted in a single cluster of all four specimens of this monotypic genus. These specimens are representative of the distribution of this species, which is known only from the southeastern coast of South Africa. The range in pairwise differences among specimens within this cluster was 0–2 and the average was 1.2.

Triakis megalopterus (spotted gulley shark) (fig. 24)

Two specimens of this species were included in the analysis; both from South Africa and are representative of the distribution of this species, which occurs only along the coast of South Africa. These specimens clustered together, well away from the other two species of Triakis included in the analysis. The two specimens of T. megalopterus differed by two bases. The average of the pairwise differences between T. megalopterus and T. semifasciata (fig. 27) was 129, and the average of the pairwise differences between T. megalopterus and T. scyllium (fig. 27) was 139.5. Differences between T. megalopterus and T. semifasciata relative to T. scyllium were noted earlier by Compagno (1988) who suggested that Triakis may constitute two subgenera: T. (Triakis) containing T. semifasciata and T. scyllium, and T. (Cazon) containing not only T. megalopterus but also T. maculata and T. acutipinna, neither of which is represented here. Our results suggest that these two clades are likely not each other's closest relatives and thus should be considered as independent genera, rather than subgenera of Triakis.

Iago omanensis (bigeye houndshark) complex (fig. 25)

A total of 19 specimens was included in the analysis, which yielded three clusters, each of which corresponds to the geographic origins of the included specimens: Gulf of Oman (11 specimens), the Red Sea (6 specimens), and India (2 specimens). Unfortunately, the identities of these clusters are difficult to assign to nominal taxa. At present, only two species of Iago have been formally described (Compagno, 1984a). These are Iago omanensis, occurring from the Red Sea and Gulf of Oman to Pakistan, and Iago garricki from the western central Pacific. However, additional forms, mostly from the northeastern Indian Ocean, but also one from the Philippines, have been reported in the literature (e.g., Compagno, 1988). Based on available morphological data (i.e., images), the specimens comprising the Gulf of Oman subcluster appear to be consistent with I. omanensis, so we have used this designation for the members of this cluster. Although several additional informal names have been applied to additional forms from the Gulf of Oman (e.g., Compagno, 1988: 240, referred to “Iago sp. nov., low fins”; Compagno et al., 2005b, included an Iago sp. A from the Gulf of Aden to India), we are uncomfortable applying any of these names to the Indian or Red Sea clusters in the absence of additional morphological data. Thus, specimens within these clusters have been referred to as Iago cf. omanensis 1 and Iago cf. omanensis 2, respectively. The range in pairwise differences among specimens in the I. omanensis cluster was 0–8 (with an average of 3.5); the two specimens of I. cf. omanensis 1 differed by seven bases, and the range in pairwise differences among the specimens of the I. cf. omanensis 2 cluster was 0–3 (with an average of 1.5). The averages of the pairwise differences between clusters were as follows: 12.5 (between I. omanensis and I. cf. omanensis 1), 13 (between I. omanensis and I. cf. omanensis 2), and 17.2 (between I. cf. omanensis 1 and I. cf. omanensis 2).

The haplotype map for Iago species colored by phenotype (fig. 89A) shows no overlapping haplotypes among specimens of the three forms of Iago omanensis. The haplotype map colored by geography (fig. 89B) confirms that haplotypes are restricted to the Gulf of Oman, the Red Sea, or India. Also shown on these maps is the clearly distinctive I. garricki from the Sulu and China Seas.

Iago garricki (longnose houndshark) (fig. 25)

Samples from three specimens collected from the Philippines and treated by Compagno et al. (2005b), were included in the analysis (i.e., GN2224  =  JPAG 083, GN2228  =  JPAG 152, and GN4330  =  JPAG 346). These specimens clustered together, grouping most closely with the other specimens of Iago species. We have followed the identification indicated by Compagno et al. (2005b) for these specimens, who considered the Philippine specimens to be conspecific with those of I. garricki from Australia and New Caledonia. The range of pairwise differences among these specimens was 4–6 (with an average of 4.7). The average of the pairwise differences between I. garricki and those of the other Iago species are as follows: 124.5 (between I. garricki and I. omanensis), 123.5 (between and I. garricki and I. cf. omanensis 1), and 116.7 (between I. garricki and I. cf. omanensis 2).

Galeorhinus galeus (topeshark) (fig. 26)

Our analysis included 18 specimens of this species, which were found to comprise a single cluster. These came from South Africa (5 specimens), the Tasman Sea off Australia (1 specimen), New Zealand (7 specimens), California (3 specimens), the Azores (1 specimen), and the eastern North Atlantic (1 specimen); as such they represent much of the reported global distribution of this species, with the exception of South America and the western coast of Africa. The range in pairwise differences among all 18 specimens was 0–13, with an average of 4.9. The analysis yielded a group composed of three weak subclusters. The first subcluster consisted of the specimens from New Zealand, Australia, and South Africa with a range in pairwise differences among specimens of 0–4. The second consisted of the specimens from California with a range in pairwise differences of 0–1, and the third subcluster consisted of the specimens from Atlantic Ocean localities that were identical in sequence. At present only the single species, Galeorhinus galeus, with the type from European waters, is recognized in this genus. However, our results are consistent with those of others, for example, Compagno (1988) and Chabot and Allen (2009), who found variation in vertebral counts and genetic structure, respectively, among populations. Nonetheless, we have used the designation G. galeus for specimens from all three subclusters but note that the average of the pairwise differences between the New Zealand/Australia and California clusters is 6.7, between the New Zealand/Australia and Atlantic clusters 10.4, and between the California and Atlantic clusters 8.3. It seems likely that further investigation will result in the resurrection of synonyms of G. galeus for some of these regional subclusters.

Hypogaleus hyugaensis (blacktip topeshark) (fig. 26)

Thirteen specimens of this species, collected and identified by Adrian Kitchingman of Western Australian Fisheries, were included in the analysis. These all came from southwestern Australia and thus represent only a small region of the distribution of this species, which also includes Japan, Taiwan, the Persian Gulf, and the east coast of Africa. The analysis yielded a single tight cluster. The range in pairwise differences within the cluster was 0–2, with an average of 0.9. These belong to what is currently considered a monotypic genus.

Hemitriakis japanica (Japanese topeshark) (fig. 26)

In total, our analysis included nine specimens of this species collected from Japan, Taiwan, and Vietnam, thus well representing the western North Pacific distribution of this species. Many of these specimens were sent to us by Kazuhiro Nakaya from the Hokkaido University Museum (GN2597  =  HUMZ 162467, GN2598  =  HUMZ 162468, GN2599  =  HUMZ 162469, and GN2600  =  HUMZ 176993). The four specimens from Taiwan are deposited at the University of Michigan Museum of Zoology (GN1000, GN1008, GN1009, and GN1010  =  UMMZ 231964). The analysis yielded essentially a single cluster with the range in pairwise differences within the cluster being 0–7, with an average of 3.6.

Hemitriakis leucoperiptera (whitefin topeshark) (fig. 26)

The analysis included three specimens of this species collected from the Philippines and treated by Compagno et al. (2005b). The analysis yielded a single cluster; the range in pairwise differences among these specimens was 0–1. This result may help to eliminate concerns over the identification of Philippine H. leucoperiptera raised by Compagno et al. (2005b). Their samples included a specimen (GN2232  =  JPAG 169) with a low vertebral count identified as H. leucoperiptera, a specimen (GN2225  =  JPAG 161) of unknown vertebral count and thus included by those authors among a series of specimens of uncertain identity, tentatively considered to be H. leucoperiptera, and also a specimen (GN4366  =  JPAG 139) of unknown vertebral count, tentatively identified by those authors as “Hemitriakis cf. japanica (var PP).” Our results suggest that all three specimens are conspecific. These specimens grouped most closely with those of H. japanica. The average of the pairwise differences between specimens of H. leucoperiptera and those of H. japanica was 8.3.

Hemitriakis complicofasciata (ocellate topeshark) (fig. 26)

Our analysis was based on five samples, all taken from museum specimens (GN2592  =  HUMZ 165233, GN2593  =  HUMZ 165255, GN2594  =  HUMZ 162464, GN2595  =  HUMZ 162465, and GN2596  =  HUMZ162466) cited by Takahashi and Nakaya (2004) in the original description of this species from the Ryuku Islands, Japan. The analysis yielded a single cluster; the range in pairwise differences among these specimens was 0–3, with an average of 1.4.

Hemitriakis falcata (sicklefin houndshark) (fig. 26)

The single specimen of this species that was included in our analysis was taken from a specimen (GN4894  =  ANFC H 5946-01) from Western Australia. This specimen grouped most closely with those of H. complicofasciata. The average of the pairwise differences between these two species was 10.8.

Hemitriakis sp. (fig. 26)

Interestingly, one of the specimens originally identified as Hemitriakis complicofasciata and also referenced in the original description of the species by Takahashi and Nakaya (2004), was found to group outside the cluster consisting of the specimens of H. complicofasciata and H. falcata. The average of the pairwise differences between this sample and those from the five specimens in the H. complicofasciata cluster was 29.8. The difference between this specimen and that of H. falcata was 27. The sample was taken from a specimen deposited in the Hokkaido University Museum (GN2591  =  HUMZ 165225); this specimen needs to be more closely examined in light of these findings. It is possible this specimen represents an undescribed species.

Furgaleus macki (whiskery shark) (fig. 26)

Two specimens, both collected from Western Australia, were included in the analysis. These are fairly representative of the distribution of this species, which is known only from the western and southern regions of Australia. These two specimens comprised a single cluster; they were identical in sequence.

Triakis scyllium (banded houndshark) (fig. 27)

Six specimens of this species, all collected from the Izu Peninsula in Japan, and identified by Sho Tanaka, were included. The analysis yielded a single cluster with the range in pairwise differences among specimens in this cluster being 0–4, with an average of 2.1.

Triakis semifasciata (leopard shark) (fig. 27)

In total six specimens, one from the Gulf of California and five from coastal California, were included in the analysis. These represent much of the distribution of this species. The analysis yielded a single cluster; the range in pairwise differences within the cluster was 0–1. The average of the pairwise differences between T. scyllium and T. semifasciata was 91. Given that our specimens from western California all came from the same locality, our data did not allow us to examine the differences seen by Lewallen et al. (2007) between northern and southern specimens of this species.

Scyliorhinidae (catsharks): group 1

Apristurus melanoasper (black roughscale catshark) complex (fig. 28)

The analysis included 10 specimens initially identified as Apristurus melanoasper; two from the western North Atlantic, three from Australia, and five from New Zealand. The analysis yielded two clusters, one consisting of the two Atlantic specimens (which differed from one another by three bases, one of which was deposited in the Yale Peabody Museum [GN1076  =  YPM ICH.010136]) and a second cluster comprised of the eight specimens from Australia and New Zealand (all of which were identical in sequence). The average of the pairwise differences between members of these two clusters was 19.5. Given that A. melanoasper was described from the North Atlantic, we have given the specimens in the Atlantic cluster the designation of A. melanoasper. Although recent work extended the known distribution of this species to include Australia and New Zealand (Nakaya et al., 2008), the results in this paper do not fully support this decision, so the Australasian cluster has been given the provisional designation of Apristurus cf. melanoasper. Several of the samples in the latter cluster came from specimens in the Australian National Fish Collection (GN4868  =  ANFC H 1391-01 and GN4869  =  ANFC H 1391-03), and several from specimens in the Museum of New Zealand, Te Papa Tongarewa (GN6723  =  NMNZ P.041310, GN6738  =  NMNZ P.042336, GN6740  =  NMNZ P.042569, GN6754  =  NMNZ P.045140). Detailed taxonomic revision of this complex is required in the future.

Apristurus brunneus (brown catshark) (fig. 28)

The two specimens of this species were collected from California and thus generally represent the northern, but not the potential southern hemisphere elements of this species, which has been reported from Panama, Ecuador, and Peru. These specimens clustered together, independently from the other species of Apristurus included in the analysis; pairwise difference between these two specimens was three bases.

Apristurus laurussonii (Iceland catshark) (fig. 28)

A total of seven specimens, all collected from the North Sea off the coast of Scotland, were included in the analysis. These specimens represent only one of the eastern components of the rather disjunct distribution of this species, which also includes isolated localities off northwestern Africa, Iceland, Massachusetts, and the Gulf of Mexico. The seven specimens were found to comprise a single cluster. The range in pairwise differences seen among these specimens was 0–3, with a mean of 1.4.

Apristurus cf. sinensis (South China catshark) (fig. 28)

A total of six samples taken from specimens collected from New Zealand, mostly from the Museum of New Zealand, Te Papa Tongarewa (GN6728  =  NMNZ P.042126, GN6749  =  NMNZ P.045142, GN6757  =  NMNZ TMP004690, GN6745  =  NMNZ P.044309, and GN6752  =  NMNZ P.045139), were included in the analysis. They yielded a single cluster with a range in pairwise differences among specimens of 0–3 (with an average difference of 1). These specimens were originally identified as A. sinensis by Kazuhiro Nakaya. However, no specimens from near the type locality of this species (South China Sea) were included in this analysis. Given that a thorough taxonomic revision of this complex is required, the New Zealand form is provisionally referred to here as A. cf. sinensis.

Apristurus sp. 1 (fig. 28)

This sample, collected from Western Australia, came from a museum specimen (GN4863  =  ANFC H 6411-02). It clustered most closely with A. cf. sinensis; the range in pairwise differences between these two taxa was 21.2. This result suggests that it represents either an undescribed species, or a known species that was not otherwise represented in the analysis.

Apristrus platyrhynchus (spatulasnout catshark) (fig. 28)

The analysis included three specimens collected from deepwater near Bass Strait, Australia. It yielded a single cluster with a pairwise difference of 1–2. Although recent taxonomic work on this species extended its known distribution to include Australia (Kawauchi et al., 2008), it would be ideal to obtain specimens from the type locality for this species, i.e., Japan, for comparison with the Australian specimens. Thus, the specific designation used here is provisional until such time as its identity can be examined in more detail.

Apristurus sp. 2 (fig. 28)

This specimen was collected from near Bass Strait, Australia, and clustered most closely with A. platyrhynchus; the range in pairwise differences between these two taxa is 33.3. This result suggests that it too represents either an undescribed species, or a known species that was not otherwise represented in the analysis.

Apristurus exsanguis (flaccid catshark) (fig. 28)

The analysis included nine samples, five of which came from specimens at the Museum of New Zealand, Te Papa Tongarewa (GN6753  =  NMNZ P.045130, GN6743  =  NMNZ TMP004689, GN6736  =  NMNZ P.042520, GN6733  =  NMNZ P.042176, and GN6732  =  NMNZ P.042519). These specimens are representative of the distribution of this species, which appears to be endemic to New Zealand. Essentially a single cluster resulted from the analysis. The range in pairwise differences within the cluster was 0–4, with an average of 1.9.

Apristurus macrorhynchus (flathead catshark) (fig. 28)

Our sample included a total of four samples all taken from specimens in the University of Michigan Museum of Zoology (GN1013, GN1014, GN1015, and GN1016  =  UMMZ 231973), and all of which were collected from Taiwan. These specimens are generally representative of the distribution of this species, which is known only from Taiwan and southern Japan. The analysis yielded a single cluster; the range in pairwise differences among specimens within this cluster was 2–6; the average of pairwise differences among specimens was 4.7.

Apristurus ampliceps (roughskin catshark) complex and Apristurus manis (fig. 28)

The relatively newly described species A. ampliceps from Australia and New Zealand (see Sasahara et al., 2008), was represented by eight New Zealand samples, all of which came from specimens deposited in the Museum of New Zealand, Te Papa Tongarewa. Also included in the analysis was a single specimen from the Atlantic Ocean, tentatively identified as Apristurus manis. The analysis yielded two clusters, one consisting solely of five New Zealand specimens of A. ampliceps (with a range of pairwise differences among specimens of 0–3 and an average of 1.6), the other consisting of three specimens from New Zealand (with a range of pairwise differences among these three specimens of 4–6 and an average of 5.5) and the specimen of A. manis. The average of the pairwise differences between members of the two clusters of A. ampliceps (excluding the A. manis specimen) was 21.5 suggesting that conspecificity of specimens in these two clusters is doubtful. However, it is unclear which of the two clusters represents the true A. ampliceps. Similarly, the identity of the A. manis specimen remains to be confirmed. Until such time as this subgroup of Apristurus can be examined in more detail, we have designated specimens comprising the first cluster as A. ampliceps 1 and those comprising the second cluster as A. ampliceps 2. We have, however, provisionally retained the designation A. manis for the Atlantic specimen in the second cluster. The average of pairwise differences between A. manis and A. ampliceps 1 was 27.2; the average of pairwise differences between A. manis and A. ampliceps 2 was 20.7. All five specimens of Apristurus ampliceps 1 were vouchered (GN6726  =  NMNZ P.041688, GN6724  =  NMNZ P.041689, GN6725  =  NMNZ P.041994, GN6735  =  NMNZ P.042385, and GN6751  =  NMNZ P.045206); this is also the case for the three specimens of Apristurus ampliceps 2 (GN6727  =  NMNZ P.041993, GN6742  =  NMNZ TMP004687, GN6744  =  NMNZ TMP004691).

Apristurus profundorum (deepwater catshark) (fig. 28)

Our material included two specimens that we have tentatively identified as A. profundorum. Both specimens were collected from the western North Atlantic but neither is vouchered or represented by images and thus the identity of this cluster remains to be verified; these specimens differed by 2.0.

Galeus sauteri (blacktip sawtail catshark) (fig. 28)

The five specimens of this species included in our analysis were all collected from Taiwan and thus represent the center of the distribution of this species, which is also known from southern Japan and the Philippines. All five samples came from specimens deposited in the University of Michigan Museum of Zoology (GN1021, GN1022, and GN1023  =  UMMZ 231966 and GN991 and GN993  =  UMMZ 231974). The analysis yielded a single cluster, with a range of pairwise difference among specimens of 0–5, with an average difference of 3.2 bases. It is interesting to note that this cluster occurred well outside those containing the other species of Galeus included in the analysis (see fig. 29). The average pairwise differences between these were as follows: 173 (between G. sauteri and G. melastomus), 154.4 (between G. sauteri and G. arae), 149.4 (between G. sauteri and G. murinus), and 166.0 (between G. sauteri and G. polli).

Parmaturus xaniurus (filetail catshark) (fig. 28)

The five specimens of Parmaturus xaniurus included in the analysis are generally representative of the distribution of this species given that they were collected from Monterey Bay, California, and this species occurs only in the eastern Pacific from California to Mexico. The analysis yielded a single cluster. The range in pairwise differences among members of this cluster was 1–5, with an average of 3.4.

Haploblepharus edwardsii (puffadder shyshark) (fig. 29)

The analysis included 19 specimens, all collected from South Africa and identified as Haploblepharus edwardsii. The range in pairwise differences among specimens within this cluster were 0–12, with an average of 4.4. This identification requires confirmation.

Halaelurus buergeri (blackspotted catshark) (fig. 29)

The analysis included five specimens from the Philippines (GN2219  =  JPAG 005, GN2220  =  JPAG 008, GN2222  =  JPAG 115, GN2234  =  JPAG 114, and GN2252  =  RSE 003), examined by Compagno et al. (2005b). It yielded a single cluster; the range in pairwise differences among specimens was 0–4; the average was 2. Compagno et al. (2005b) noted that, while these specimens resembled Halaelurus buergeri, they differed somewhat in body shape and spotting pattern from those collected elsewhere (e.g., Taiwan, Indonesia, and Japan) and thus gave their specimens the designation H. cf. buergeri. However, as a result of work conducted in conjunction with the description of Halaelurus maculosus from Indonesia by White et al. (2007b), W.W. now considers these Philippine specimens to represent H. buergeri.

Halaelurus sellus (speckled catshark) (fig. 29)

This species was represented in the analysis by only a single specimen collected from Western Australia. This specimen was deposited in the Australian National Fish Collection (GN4893  =  ANFC H 6367-01). It was identified by P.L. as H. sellus, a species recently described from northwestern Australia by White et al. (2007b). This specimen clustered most closely with those of H. maculosus; the average of the pairwise differences between specimens of these two species was 60.6.

Halaelurus natalensis (tiger catshark) (fig. 29)

The analysis included three specimens of this South African endemic species. The range in pairwise differences among specimens in this cluster was 3–5, with an average of 4.

Halaelurus lineatus (lined catshark) (fig. 29)

Both of the specimens of this species included in the analysis were collected from KwaZulu-Natal in South Africa and identified by Geremy Cliff. They were found to comprise a single cluster; their sequences differed from one another by 4. These specimens are representative of the relatively limited distribution of this species. This species clustered most closely with H. natalensis. The average of the pairwise differences between these two species was 80.3.

Holohalaelurus regani (Izak catshark) (fig. 29)

All 17 specimens of this species included in the analysis were collected off the coast of Western Cape Province and the western regions of Eastern Cape Province, South Africa. The range in pairwise differences among specimens was 0–6, with an average of 2. Our analysis did not include specimens, formerly considered as the “Natal” form of this species, which were described as Holohalaelurus favus by Human et al. (2006).

Galeus melastomus (blackmouth catshark) (fig. 29)

A total of eight specimens identified as this species were included in this analysis. These were collected from a diversity of localities in the eastern Atlantic including Ireland, Scotland, and Madeira and thus represent the northwestern elements of the distribution of this species. The sample from Madeira was taken from a specimen in the Museu de Hist?ria Natural e Aqu?io in Funchal (GN6627  =  MMF 36798). The analysis yielded a single cluster; the range in pairwise differences among specimens in this cluster was 0–4; the average of pairwise differences was 2.

Galeus polli (African sawtail catshark) (fig. 29)

The six specimens of this species, all collected from western South Africa were included. These represent only a small portion of the distribution of this species, which occurs throughout the west coast of Africa, and also in the Mediterranean Sea. The range in pairwise differences among specimens of this species was 0–3. This species clustered most closely with G. melastomus. The average of the pairwise differences between these two species was 51.4.

Galeus arae (roughtail catshark) (fig. 29)

The analysis included a single Atlantic specimen identified provisionally as this species by Jose Castro. It was divergent from all other included Galeus species. For example, the average pairwise differences between the species with which it clustered most closely were 72.8 (A. melastomus) and 81.5 (A. polli).

Galeus murinus (mouse catshark) (fig. 29)

Both specimens of this species included in the analysis were collected from the central to eastern Atlantic; one of these was identified by Neils Roar Hareide. The specimens clustered together independent of the other species of Galeus. The sequences of these two specimens differed by 2.

Apristurus species (second major cluster) (fig. 30)

Seven additional specimens belonging to the genus Apristurus were included in the analysis. These clustered together, but well outside the 56 specimens considered to represent ∼14 other species of Apristurus (see fig. 28), questioning the integrity of this genus as currently circumscribed.

Three of the seven specimens in this cluster were identified as one of the long-snouted catsharks, Apristurus australis (pinocchio catshark). These specimens are representative of the known distribution of this Australian endemic species as they were collected from Tasmania and Western Australia. Both samples from Western Australia came from museum specimens (GN4877  =  ANFC H 2573-01 and GN4878  =  ANFC H 2600-04) and clustered with a specimen from Tasmania. The range of pairwise differences among these three specimens was 2–8 (with an average of 6).

Also included in the analysis was a single specimen from New Zealand. While this specimen clustered most closely with the specimens identified as Apristurus australis, the average of the pairwise differences between it and the specimens in the latter subcluster was 62.3. Thus, we have given this specimen the distinct designation Apristurus sp. 3 as it may represent an undescribed species. Although there are no photographs or retained specimens of this catshark, it could be another long-snout species; it could be the long-snouted Apristurus sp. A of Paulin et al. (1989), which has also been informally referred to as A. cf. herklotsi in unpublished checklists of New Zealand fishes. This cluster is most likely referable to the A. longicephalus group defined by Nakaya and Sato (1999).

Two specimens collected from the northeastern Atlantic clustered together (with a sequence difference of 5), independently from all four of the above specimens. The average of the pairwise differences between these Atlantic specimens and those of A. australis was 153.8, and between the Atlantic specimens and Apristurus sp. 3 was 161.5. Thus, specimens in the Atlantic cluster have been given the provisional designation Apristurus sp. 4; it is possible they also represent an undescribed species. Clustering with, but well outside these two specimens was a single specimen from the University of Kansas Ichthyology Collection (GN2533  =  KUI 29258), collected from California and identified as Apristurus kampae. The average of the pairwise differences between the specimen of A. kampae and the two specimens of Apristurus sp. 4 was 69.

Asymbolus rubiginosus (orange spotted catshark) (fig. 31)

Three specimens from Ken Graham, identified as Asymbolus rubiginosus, from off New South Wales, Australia, were included in the analysis. These specimens clustered together and had a range in pairwise differences of 0–4, with an average of 2.7.

Asymbolus parvus (dwarf catshark) (fig. 31)

Both samples of this species included in the analysis came from museum specimens (GN4879  =  ANFC H 6415-01 and GN4880  =  ANFC H 6415-02) collected from Western Australia. These specimens are representative of the distribution of this species, which is endemic to Western Australia. The sequences of these specimens differed from one another by 4. The specimens of this species grouped most closely with those of A. rubiginosus; the average of the pairwise differences between these two species was 35.

Asymbolus analis (gray spotted catshark) (fig. 31)

All five specimens of this Australian endemic species included in the analysis were collected in the central part of its range. The analysis yielded a single tight cluster with a range in pairwise differences among specimens of 0–2.

Figaro boardmani (Australian sawtail catshark) complex (fig. 31)

The nine specimens included in the analysis are generally representative of the distribution of this Australian endemic; six come from southeastern Australia and three from Western Australia. The analysis yielded two distinct clusters: one consisting of specimens from southeastern Australia and one consisting of specimens from Western Australia. The range in pairwise differences among specimens within the southeastern cluster was 0–6, with an average of 2.8; the range for the Western Australian cluster was 1–2, with an average of 1.3. The average of the pairwise differences between clusters was 19.3 suggesting that, as noted by Gledhill et al. (2008), regional variation occurs within this species. In recognition of this variation, we have given the specimens from Western Australia the designation Figaro cf. boardmani, reserving the designation Figaro boardmani for those in the southeastern cluster. The three samples of F. cf. boardmani were taken from specimens in the Australian National Fish Collection (GN4890  =  ANFC H 6414-07, GN4891  =  ANFC H 6414-08, and GN4892  =  ANFC H 6414-10). Taxonomic revision of this species complex is currently being undertaken by researchers at the Australian National Fish Collection.

Bythaelurus dawsoni (New Zealand catshark) (fig. 31)

Six specimens of this New Zealand endemic were included in the analysis and they yielded a single tight cluster. The range in pairwise differences among specimens in the cluster was 0–1. The specimens from which five of these six samples were taken are deposited in the Museum of New Zealand, Te Papa Tongarewa (GN6731  =  NMNZ P.042162, GN6737  =  NMNZ P.042731, GN6739  =  NMNZ P.042162, GN6746  =  NMNZ P.044374, and GN6748  =  NMNZ P.044375).

Pseudotriakidae (false catsharks)

Pseudotriakis microdon (false catshark) (fig. 32)

Our analysis included three specimens of this species, all from the mid-Atlantic ridge. Thus, our material represents only one element of the wide and patchy distribution of this monotypic genus. The analysis yielded a single cluster with the three specimens being identical in sequence.

Gollum species (fig. 32)

At present this genus formally includes only the New Zealand endemic, Gollum attenuatus. However, Compagno et al. (2005b) noted that a second, potentially undescribed species (Gollum sp. 1), occurs in the Philippines (treated as Gollum sp. A by Compagno et al., 2005a). Our analysis included one specimen of G. attenuatus from New Zealand, and one specimen (GN2440  =  JPAG 229) of the potentially undescribed species, Gollum sp. 1, from the Philippines. These two Gollum specimens were found to group together, but their sequences differed by 60, thus this supports the notion of these being separate species.

Proscylliidae (finback catsharks)

Proscyllium habereri (graceful catshark) (fig. 32)

Four specimens of this species collected and identified by Kazuhiro Nakaya from Okinawa and the East China Sea were included in the analysis. The specimen from Okinawa is deposited in the Hokkaido University Museum (GN2601  =  HUMZ 175853). These specimens represent the more northern elements of the distribution of this species, which extends as far south as Java. The analysis yielded a single cluster. The sequences of the three specimens from the East China Sea were identical. The average of the pairwise differences between these three specimens and the one from Okinawa was 6.

Eridacnis sp. 1 (Philippine ribbontail catshark) (fig. 32)

A single specimen, provisionally identified as belonging to the genus Eridacnis, from the Philippines was included in the analysis. This specimen was identified as “?Eridacnis sp. 1” by Compagno et al. (2005b) and who considered it (GN2212  =  BRU 004) potentially represented a new species. This specimen grouped potentially most closely with the specimens of Proscyllium habereri, but the average pairwise difference between specimens of these two species was 164.5.

Scyliorhinidae (catsharks): group 2

Atelomycterus marmoratus (coral catshark) (fig. 33)

In total, 10 specimens of this species were included in the analysis: nine were collected from Malaysian Borneo and one from the Philippines. These specimens represent eastern elements of the distribution of this species, which extends as far west as Pakistan. The analysis yielded a single cluster. The range in pairwise differences within the cluster was 0–11, with an average of pairwise differences of 3.8. There was some evidence of geographic structure within the cluster in that the specimens from Borneo grouped together. The average of the pairwise differences between specimens in the Borneo cluster and the specimen from the Philippines was 9.9. We note that the sample from the Philippines, which came from a museum specimen (GN2235  =  JPAG 044), was treated by Compagno et al. (2005b); one of the specimens from Borneo was deposited in the IPPS (GN3705  =  IPPS BO495).

Aulohalaelurus labiosus (blackspotted catshark) (fig. 33)

The three specimens of this species were all collected from Western Australia and are thus representative of the distribution of this Western Australian endemic. Two samples came from specimens deposited in the Western Australian Museum (GN2268  =  WAM P 31670-001 and GN2269  =  WAM P 31671-001). The analysis yielded a single cluster. The range in pairwise differences among specimens in this cluster was 4–7, with an average of 5.3.

Atelomycterus marnkalha (eastern banded catshark) (fig. 33)

All three specimens of this species were collected from the Torres Strait, off Australia, and represent the central elements of its distribution, which extends eastward to New Guinea. All three samples were taken from specimens in the Australian National Fish Collection (GN4881  =  ANFC H 6144-01, GN4882  =  ANFC H 6145-01, and GN4883  =  ANFC H 6146-01). The analysis yielded a single tight cluster, with a range in pairwise differences among members of this cluster of 2–4, and an average of 2.7. It was interesting that this species clustered with Aulohalaelurus labiosus. The average of the pairwise differences between these two species was 158.3 and between A. marnkalha and its congener A. marmoratus 155.6.

Parmaturus sp. (fig. 33)

Six samples, all collected from New Zealand and deposited in the Museum of New Zealand, Te Papa Tongarewa (GN6734  =  NMNZ P.045528, GN6730  =  NMNZ P.042517, GN6741  =  NMNZ P.044582, GN6747  =  NMNZ P.044583, GN6750  =  NMNZ P.044578, and GN6755  =  NMNZ P.042524), were identified only to genus. The analysis yielded a single cluster with the range in pairwise differences among specimens being 0–3, with an average of 1.2. This taxon grouped well away from the cluster comprised of its congener, Parmaturus xaniurus (see fig. 28); the average of pairwise differences between these two species was 190.5. This suggests that the identity of these specimens needs to be examined more closely, both at the specific and generic levels. The only other species of Parmaturus currently known from New Zealand is P. macmillani. This taxon is currently under taxonomic investigation by P.L., Bernard Séret, and Keiichi Sato.

Schroederichthys bivius (narrowmouth catshark) (fig. 33)

Both specimens of this species were collected from Argentina and thus represent the eastern elements of the distribution of this species, which extends throughout much of coastal Chile and Argentina. The sequences of these two specimens were identical.

Poroderma species

Given they exhibited the least amount of intrageneric variation seen in the analysis, a haplotype map was generated for the two sympatric species of Poroderma. This map (fig. 90) shows two relatively tight haplotype clusters, with no overlap between species, suggesting that although divergence is low the species are distinct.

Poroderma pantherinum (leopard catshark) (fig. 34)

In total, 16 specimens of this species were included in the analysis. These specimens came from a diversity of localities throughout South Africa and thus represent much of the distribution of the species. The analysis yielded a single cluster. The range in pairwise differences among specimens was 0–6, with an average of 3.

Poroderma africanum (striped catshark) (fig. 34)

The 12 specimens of this species were collected from South Africa and thus are representative of the distribution of this South African endemic. The analysis yielded a single cluster, which was tightly allied with the cluster of specimens of P. pantherinum. The range in pairwise differences among specimens in the P. africanum cluster was 0–2, with an average of 1.3. The average of the pairwise differences between specimens of the two Poroderma species was 6.3.

Scyliorhinus capensis (yellowspotted catshark) (fig. 34)

A total of 15 specimens of this southern Africa endemic were included. The range of pairwise differences among these specimens was 0–2; the average was 0.6.

Scyliorhinus canicula (smallspotted catshark) (fig. 34)

The single specimen of this species was collected from the eastern Atlantic. It grouped most closely with the specimens of the southern Africa endemic S. capensis. The average of the pairwise differences between specimens of these two species was 57.5.

Scyliorhinus retifer (chain catshark) (fig. 34)

Also included in the analysis were three specimens preliminarily identified as Scyliorhinus retifer. One of these came from a specimen in the University of Kansas Ichthyology Collection (GN2530  =  KUI 26984) that was collected from the mid-Atlantic bight. The others were collected from the northwestern Atlantic and Gulf of Mexico. The analysis yielded a single, divergent cluster. However, the range in pairwise differences among these specimens was substantial, at 9–22, and the average was 14.7.

Scyliorhinus stellaris (nursehound) (fig. 34)

A single specimen of S. stellaris, collected from the eastern Atlantic, was included. This specimen grouped along with, but outside its congeners. The average pairwise difference between the specimen of S. stellaris and those in the S. retifer cluster was 73.3, between S. stellaris and the specimen of S. canicula 73, and between S. stellaris and S. capensis 66.5.

Cephaloscyllium variegatum (saddled swellshark) (fig. 34)

The analysis included six specimens of this species from Australia, one of which was included among the other material examined by Last and White (2008c) in the original description of this species (GN4889  =  ANFC H 3580-01). The range in pairwise differences among specimens in this cluster was 0–2, with an average of 0.9.

Cephaloscyllium albipinnum (whitefin swellshark) (fig. 34)

The three samples of this species included in the analysis were taken from the holotype (GN4887  =  ANFC H 5314-11), a paratype (GN4888  =  ANFC H 5940-01), and an additional examined specimen (GN4886  =  ANFC H 3588-01) included in the original description of this relatively new species from New South Wales and Tasmania, Australia (see Last et al., 2008b). The range in pairwise differences among these three specimens was 1–2, with an average of 1.3.

Cephaloscyllium hiscosellum (Australian reticulate swellshark) (fig. 34)

The specimen of this species included in our analysis was a paratype of this Western Australian species (GN4884  =  ANFC H 6419-01), described by White and Ebert (2008). It clustered outside but along with the specimens of C. albipinnum. The average of the pairwise differences between this specimen and those of C. albipinnum was 44.3.

Cephaloscyllium umbratile (Japanese swellshark) (fig. 34)

The four samples for this species included in the analysis come from specimens in the University of Michigan Museum of Zoology (GN981  =  UMMZ 231960 and GN982, GN1017, and GN1018  =  UMMZ 231967) originally identified as C. isabellum. These specimens were all collected from Taiwan. However, Schaaf-Da Silva and Ebert (2008) clarified the status of C. umbratile as a valid species having been previously placed in synonymy with C. isabellum. Thus, the specimens included here have been provisionally identified as C. umbratile. As such, they represent the southern portion of the western North Pacific distribution of this species. The range in pairwise differences among specimens of was 0–4, with an average of 2.7. Schaaf-Da Silva and Ebert (2008) also described two new species of Cephaloscyllium from Taiwan, so the identity of these specimens should be confirmed.

Cephaloscyllium sp. 1 (Philippine swellshark) (fig. 34)

Our analysis included two specimens of Cephaloscyllium collected from the Philippines, both of which were treated by Compagno et al. (2005b) (GN4361  =  JPAG 231 and GN4352  =  JPAG 232) as Cephaloscyllium sp. 1. These specimens clustered together and their sequences differed from one another by three bases. However, this species needs to be critically compared with C. pardelotum and C. maculatum both described by Schaaf-DaSilva and Ebert (2008) from Taiwan, as well as with C. pictum, recently described from Indonesia by Last et al. (2008c). These specimens clustered mostly closely with specimens of C. umbratile. However, the average of the pairwise differences between these two clusters was 47.8.

Cephaloscyllium laticeps (Australian swellshark) (fig. 34)

In total, six specimens, all collected from southeastern Australia were included in the analysis. One of these samples was taken from a museum specimen (GN4885  =  ANFC H 3581-01). The range in pairwise differences among specimens in this cluster was 0–5, with an average of 1.9.

Cephaloscyllium ventriosum (swellshark) (fig. 34)

One sample of this species, taken from a specimen in the University of Kansas Ichthyology Collection (GN2529  =  KUI 28129) collected from California, was included in the analysis. This specimen was grouped with all other Cephaloscyllium species, but was by far the most divergent species. The average of the pairwise differences between this specimen and those of the six other Cephaloscyllium species was 109.

Lamniformes (mackerel sharks)

Lamnidae (mackerel sharks)

Isurus oxyrinchus (shortfin mako) (fig. 35)

In total, 24 specimens of I. oxyrinchus were included in the analysis, which yielded a well-defined cluster. The range in pairwise p-differences among specimens was 0–21, with an average of 10.6. This cluster consisted of two fairly well-defined subclusters, one of which exhibited more sequence heterogeneity than the other; this is also reflected in the haplotype map for phenotype (fig. 91A). However, as illustrated by the haplotype map of geography (fig. 91B) there was no apparent geographic pattern to these subclusters, each of which was comprised of specimens from the western North Atlantic, Gulf of Mexico, California, Gulf of California, and Taiwan; one subcluster also included a specimen from South Africa and the other also included a specimen from Vietnam. The range in pairwise differences among specimens in the subcluster including the specimen from South Africa was 0–4, with an average of 2.4. The range in pairwise differences among specimens in the other subcluster was 0–14, with an average of 6.8. The average of the pairwise differences between specimens in the two subclusters was 17.2. In the absence of morphological or geographic data to support this distinction, we have tentatively identified all 24 members of the cluster as I. oxyrinchus. Nonetheless, these results are interesting in light of previous workers (e.g., Moreno and Morón, 1992; Heist et al., 1996; Schrey and Heist, 2003), all of who provided evidence of population structure among makos globally.

Isurus paucus (longfin mako) (fig. 35)

A total of six specimens of this species were included in the analysis, four from the western North Atlantic, one from the Caribbean Sea, and one from the Gulf of Mexico. These specimens comprised a single cluster; the range in pairwise differences among specimens in this cluster was 0–8, with an average of 3.3. The average of the pairwise differences between the six specimens of this species and the 24 specimens of I. oxyrinchus was 110.3.

Carcharodon carcharias (great white shark) (fig. 35)

The 17 specimens of this species included in the analysis were collected from South Africa, the western North Atlantic, California, and South Australia. The analysis yielded a single cluster, with two subclusters. The range in pairwise differences among all 17 specimens was 0–20, with an average of 4.2. One subcluster was comprised of the specimens from South Africa and the western North Atlantic, the other comprised of specimens from California and South Australia (i.e., from localities in the Pacific Ocean). The range in pairwise differences among specimens in the former subcluster was 0–2; the two specimens in the latter subcluster differed from one another by seven bases. On average, pairwise difference between specimens of the two subclusters was 16.3. This intriguing result, which is consistent at least in part with the findings of Pardini et al. (2001) and Jorgensen et al. (2009), suggests that some consideration should be given to the potential lack of conspecificity of C. carcharias from the Atlantic and Indian oceans, and those from the Pacific Ocean.

Lamna nasus (porbeagle shark) (fig. 35)

The analysis of 10 specimens yielded a single relatively divergent cluster with some evidence of two subclusters within. The range in pairwise differences among all 10 specimens was 0–15, with an average of 6.8. One subcluster consisted of the two specimens from Tasmania, both vouchered (GN2261  =  AMS I32756-002 and GN2262  =  AMS I32756-001), and the other consisted of specimens collected from the northeastern and northwestern Atlantic. The two specimens from Tasmania differed from one another by nine bases. The range in pairwise differences among specimens in the other subcluster was 0–8. The average of the pairwise differences between specimens from the Tasmanian subcluster and those from the Atlantic Ocean subcluster was 14. These results suggest that some consideration should be given to the possibility of the existence of taxonomic variation among porbeagles between ocean basins.

Lamna ditropis (salmon shark) (fig. 35)

Two specimens of this north Pacific–dwelling species, both from Japan, were included. They differed by 4. The average of the pairwise differences between L. ditropis and L. nasus was 68.

Cetorhinidae (basking sharks)

Cetorhinus maximus (basking shark) (fig. 35)

Two specimens of this widely distributed species were included in the analysis. These samples differed from one another by five bases. As the locality of one of the two specimens is unknown, our results do not allow us to expand on the work of Hoelzel et al. (2006), who found relatively low genetic diversity among specimens from different ocean basins.

Odontaspididae (sand tiger sharks): group 1

Carcharias taurus (sand tiger shark) (fig. 35)

All six specimens of this species included in the analysis were collected from the western North Atlantic. The range in pairwise differences among these specimens was 0–4, with an average of 1.7. As our specimens represented only a very small portion of the range of this widely distributed species, they were not conducive to examining the genetic variation seen, for example, by Ahonen et al. (2009) in this taxon.

Alopiidae (thresher sharks)

Alopias pelagicus (pelagic thresher) (fig. 35)

The analysis yielded a cluster with weak evidence of two subclusters. The range in pairwise differences among all 20 specimens was 0–12; the average was 6. One subcluster consisted of a total of 10 specimens collected from Borneo, Taiwan, the Philippines, and India. The range in pairwise differences among these specimens was 0–3. The second subcluster consisted of 10 specimens from the Gulf of Mexico and the Gulf of California, as well as one specimen from Taiwan. One of the samples from the Gulf of California was vouchered (GN5402  =  IBUNAM PE9512). The range in pairwise differences among these specimens was 0–7. The average of the pairwise differences between specimens in the two subclusters was 9.9. Although these results do not support the potential existence of a cryptic species of Alopias off Baja California as suggested by Eitner (1995), they do provide some support for the existence of greater diversity in this genus than currently thought. We note that included here is a specimen (GN4309  =  JPAG 117B) treated by Compagno et al. (2005b) as the first record of A. vulpinus from the Philippines, which clusters among specimens of A. pelagicus. This result suggests that the identity of this specimen should be reconsidered.

Alopias vulpinus (thresher shark) (fig. 35)

The 11 specimens of this species included in the analysis were collected from both the western North Atlantic and the eastern Pacific oceans. While our analysis lacked representation from throughout much of the distribution of this species, it revealed no evidence of structure between the eastern and western regions of North America. The range in pairwise differences among specimens in this cluster was 0–2, with an average of 0.9. However, the average of the pairwise differences between specimens of A. vulpinus and those identified as A. pelagicus was 97.

Alopias superciliosus (bigeye thresher) (fig. 35)

Little difference was seen among the nine specimens in the analysis, which included specimens from Virginia, Florida, Taiwan, Senegal, Gulf of California, and the Philippines, and thus represented much of the global distribution of this species. The range in pairwise differences among these specimens was 0–8, with an average of pairwise differences of 2.8. For comparative purposes we note that the average of the pairwise differences between specimens of A. superciliosus and those of A. pelagicus was 120.1 and between specimens of A. superciliosus and those of A. vulpinus 118.7.

Megachasmidae (megamouth sharks)

Megachasma pelagios (megamouth shark) (fig. 35)

The six specimens included in the analysis came from California, Indonesia, Japan, and Taiwan, and thus represent essentially only the Pacific elements of the distribution of this species. The range in pairwise differences among these specimens was 0–10, with an average of pairwise differences of 3.5.

Odontaspididae (sand tiger sharks): Group 2

Odontaspis ferox (smalltooth sand tiger) (fig. 35)

Both of the included specimens of this species were collected from the Azores and thus represent only a very small portion of the distribution of this species. Nonetheless, these specimens were identical in sequence.

Odontaspis noronhai (bigeye sand tiger) (fig. 35)

Only a single specimen of this species, collected from Brazil, was included in the analysis. The average of the pairwise differences between the specimens of this species, and those of Odontaspis ferox was 93.

Pseudocarchariidae (crocodile sharks)

Pseudocarcharias kamoharai (crocodile shark) (fig. 35)

The four specimens of this species included in the analysis were collected from Taiwan and Hawaii, and thus represent only the Pacific elements of the distribution of this species. The range in pairwise differences among all four specimens was 1–18 (with an average of 9 bases). The analysis yielded a subcluster comprised of both specimens from Taiwan and a specimen from Hawaii. The range in pairwise differences among these specimens was 1–2. However, the second specimen collected from Hawaii, grouped outside this subcluster, with an average pairwise difference of 16.7 relative to the members of the cluster.

Mitsukurinidae (goblin sharks)

Mitsukurina owstoni (goblin shark) (fig. 35)

All three specimens of this species included in the analysis were collected from localities in the Pacific Ocean (i.e., California, Tasmania, and Japan), and thus they represent only a relatively small portion of the distribution of this species. These specimens differed from one another by only 0–2, with an average of 1.3 bases. One of the samples was taken from a specimen at the Los Angeles County Museum (GN1798  =  LACM 47362-1).

ORECTOLOBIFORMES (carpet sharks)

Hemiscylliidae (longtailed carpetsharks)

Chiloscyllium punctatum (brownbanded bambooshark) complex (fig. 36)

A total of 56 specimens originally identified as Chiloscyllium punctatum were included in the analysis. These were collected from Thailand, Singapore, Malaysian and Indonesian Borneo, Vietnam, and Australia. The analysis yielded two distinct clusters, one consisting of the two specimens from Australia; the other consisting of the 54 specimens from the remaining localities. One sample from Borneo was vouchered (GN4446  =  CAS 229025). The two specimens in the Australian cluster differed from one another by a single base. The range in pairwise differences among the 54 specimens in the second cluster was 0–10, with an average of 3.3. While some structure within the second cluster was observed, the averages of these differences (e.g., 5 between a group comprised of 3 of the specimens from Sabah and a group comprised of specimens primarily from Sarawak) were determined to be too small to warrant further mention at this time. The average of the pairwise differences between specimens of the two main clusters was 27.8. Given that the type locality of C. punctatum is Jakarta, the cluster comprised of specimens collected from Thailand, Singapore, and Borneo was given that provisional designation.

The haplotype map colored by phenotype (fig. 92A) supports the distinction between these two species in that there is no overlap in haplotypes among specimens of these two species. The haplotype map colored by geography (fig. 92B) illustrates the allopatric nature of the distribution of these two taxa. Although Last and Stevens (2009) provisionally referred to Australian populations as C. punctatum, this analysis suggest these specimens may represent an undescribed species. Thus, the specimens in the Australian cluster have been referred to here as Chiloscyllium cf. punctatum. Taxonomic revision of this species is currently being undertaken by W.W., P.L., and Gordon Yearsley.

Chiloscyllium indicum (slender bambooshark) (fig. 36)

The 42 specimens of this species included in the analysis were all collected from either Malaysian or Indonesian Borneo and thus represent only a portion of the Indo-West Pacific distribution of this species. Eleven of the specimens from Borneo were vouchered (GN4467  =  CAS 229031, GN4471  =  CAS 229032, GN3440  =  ANFC H 6123-01, GN3676  =  ANFC H 6213-01, GN3692  =  ANFC H 6214-01, GN3693  =  ANFC H 6214-02, GN3511  =  IPPS BO294, GN3470  =  IPPS BO253, GN3488  =  IPPS BO271, GN3679  =  IPPS BO468, and GN4276  =  MZB 15.505). The analysis yielded essentially a single cluster, with a range of pairwise differences among specimens in this cluster of 0–11, with an average of 1.5. Significant morphological variation exits among specimens of C. indicum, even from the same collection locality, and a more thorough investigation of the material is needed. The average of the pairwise differences between C. indicum and C. punctatum was 132.1, and between specimens of C. indicum and those of C. cf. punctatum 132.2.

Chiloscyllium plagiosum (whitespotted bambooshark) (fig. 36)

The analysis included three specimens of this species, which were collected from Singapore and Malaysian and Indonesian Borneo, and were identical in sequence. The average of the pairwise differences seen between these specimens and those of C. indicum was 115.9, those of C. punctatum 128.2, and those of C. cf. punctatum 128.5.

Hemiscyllium ocellatum (epaulette shark) (fig. 36)

This genus was relatively poorly represented in our analysis. Only three specimens of a single species were included and all three of these were of H. ocellatum collected from Cairns, Australia. These were identical in sequence and of particular interest is the fact that this cluster grouped among species of Chiloscyllium.

Chiloscyllium hasseltii (Indonesian bambooshark) (fig. 36)

The analysis included 13 specimens of this species, 12 from Borneo, and one from Singapore. The analysis yielded a single tight cluster. Three of the specimens from Borneo were vouchered (GN4458  =  CAS 229029, GN3416  =  ANFC H 6122-01, and GN3704  =  IPPS BO494.) The range in pairwise differences among members of this cluster was 0–4, with an average of 0.7. The average of the pairwise differences between specimens of this species and those of C. punctatum was 142.3, between this species and C. cf. punctatum 143.5, between those of C. indicum 157.6, and between those of C. plagiosum 150.9.

Chiloscyllium griseum (gray bambooshark) (fig. 36)

Two specimens collected from Maharastra, India, and preliminarily identified as C. griseum were included in the analysis. These specimens differed from one another by one base and clustered most closely with the specimens of C. hasseltii. The averages of the pairwise differences between this species and its congeners were as follows: 137.8 from C. punctatum, 140 from C. cf. punctatum, 162.1 from C. indicum, 147.5 from C. plagiosum, and 119.6 from C. hasseltii. We have no images or retained specimens of this taxon, so their identity is somewhat uncertain. However, given that two of the three other described species of Chiloscyllium not included here occur only in localities well outside India (i.e., Madagascar for C. caeruleopunctatum and Burma for C. burmensis), and that the specimens from India lacked the dorsal ridges typical of C. arabicum and lacked the dark marks on the fins seen in C. burmensis, this identification is likely to be correct. Alternatively, these specimens may represent an as yet undescribed species of Chiloscyllium. NADH2 data for the three remaining described species of the genus would do much to help confirm this identification.

Stegostomatidae (zebra sharks)

Stegostoma fasciatum (zebra shark) (fig. 37)

Eleven of the 12 specimens included in the analysis were collected from Borneo; the remaining specimen came from Mozambique, and thus our specimens span much of the distribution of this species. The analysis yielded essentially a single cluster. The range in pairwise differences among specimens in this cluster was 0–7, with an average of 3.8. The relatively low level of variation seen here is interesting in view of the results of Dudgeon et al. (2009), given that our specimens included a specimen from Mozambique, as well as the Indo-West Pacific.

Rhincodontidae (whale sharks)

Rhincodon typus (whale shark) (fig. 37)

A total of six specimens were included in the analysis. These came from Taiwan, California, Borneo, and Mozambique and thus represent much of the Indo-Pacific distribution of this species, but do not include any representatives from the Atlantic Ocean. The analysis yielded a single very tight cluster with all six specimens identical in sequence. The comparatively small size of our sample makes our results difficult to compare with those of previous studies on genetic variation of this species (e.g., Schmidt et al., 2009; Castro et al., 2007).

Ginglymostomatidae (nurse sharks)

Pseudoginglymostoma brevicaudatum (shorttail nurse shark) (fig. 37)

The three specimens of this eastern Africa endemic species included here all came from Kenya. They were identical in sequence. This species clustered most closely with Rhincodon typus, but a model-based formal phylogenetic analysis is required to resolve their relationships more definitively. The average of the pairwise differences among specimens of these two species was 148.

Ginglymostoma cirratum (nurse shark) complex (fig. 37)

The 12 specimens originally identified as this species included in the analysis were collected from the Gulf of Mexico, the western Atlantic off the coasts of Florida and Virginia, and the Gulf of California. The analysis yielded two distinct clusters: one consisting of the Atlantic specimens and one consisting of the three specimens from the Gulf of California. The nine specimens in the Atlantic cluster exhibited a range in pairwise differences of 0–2, with an average of 1. The cluster comprised of the specimens from the Gulf of California had a range of pairwise differences of 0–1. However, the average of the pairwise differences between the members of these two clusters was 15.9. This finding lends some support to the suggestion that the nurse sharks occurring off of Baja may not be conspecific with those occurring along the eastern seaboard of North America. One of the specimens from the Gulf of California was vouchered (GN3561  =  TCWC 7585.01  =  IBUNAM PE9492). Given that the type locality of G. cirratum is Jamaica, we have referred to specimens in the Atlantic cluster as G. cirratum and to those in the Gulf of California cluster as Ginglymostoma cf. cirratum. Both the haplotype map for phenotype (fig. 93A), which shows no haplotype overlap among specimens of the two taxa, and that for geography (fig. 93B) support the distinction between these two allopatric species.

Nebrius ferrugineus (tawny nurse shark) (fig. 37)

Both specimens of this species included in the analysis were collected from northern Australia, and thus are not fully representative of the Indo-West to central Pacific distribution of this species. These specimens differed from one another by three base pairs.

Orectolobidae (wobbegongs)

Orectolobus hutchinsi (western wobbegong) (fig. 38)

The 10 specimens of this relatively newly described Western Australian endemic species (see Last et al., 2006) included three paratypes (GN4847  =  ANFC H 6189-01, GN4848  =  ANFC H 6189-02, and GN4849  =  ANFC H 6189-03). All 10 specimens were identical in sequence but differed substantially from all five of the other included species of this genus.

Orectolobus parvimaculatus (dwarf spotted wobbegong) (fig. 38)

All three specimens included in the analysis were paratypes (GN4854  =  ANFC H 5633-07, GN4855  =  ANFC H 6172-01, and GN4856  =  ANFC H 6192-01) of this recently described species found only off southwestern Australia (see Last and Chidlow, 2008). These three specimens were identical in sequence and they clustered most closely with specimens of O. hutchinsi, but the average of the pairwise differences among specimens of these two species was 30.

Orectolobus halei (gulf wobbegong) (fig. 38)

Seven specimens of this Australian species, which was recently resurrected and redescribed by Huveneers (2006), an action supported by Corrigan et al. (2008), were included in the analysis; all seven were collected from New South Wales and Western Australia. The two samples from New South Wales came from voucher specimens deposited in the Australian National Fish Collection (GN4845  =  ANFC H 6278-01, and GN4846  =  ANFC H 6278-02). The seven specimens formed a tight cluster with a range in pairwise differences of 0–2, with an average of 1.

Orectolobus maculatus (spotted wobbegong) (fig. 38)

The three samples of this species included in the analysis came from voucher specimens deposited in the Australian National Fish Collection (GN4851  =  ANFC H 5765-01, GN4852  =  ANFC H 5766-01, and GN4853  =  ANFC H 5766-02). These specimens were identical in sequence. They clustered most closely with those of O. halei; the average of the pairwise differences between specimens of these two species was 15.4.

Orectolobus ornatus (ornate wobbegong) (fig. 38)

A single sample of this species, taken from a voucher specimen deposited in the Australian National Fish Collection (GN4857  =  ANFC H 5763-01) was included. The average of pairwise differences between O. ornatus and O. hutchinsi was 58, O. parvimaculatus 60, O. halei 54.4, and O. maculatus 50.

Orectolobus floridus (floral banded wobbegong) (fig. 38)

All three of the specimens of this southwestern Australian endemic included in the analysis were paratypes (GN4858  =  ANFC H 5936-01, GN4859  =  ANFC H 6168-01, and GN4860  =  ANFC H 6490-01). Like most other wobbegong species, these specimens were identical in sequence. They clustered along with, but outside their five congeners included in the analysis; the average of the pairwise differences between this species and the others were as follows: O. hutchinsi 59, O. parvimaculatus 59, O. halei 50.4, O. maculatus 46, and O. ornatus 72.

Eucrossorhinus dasypogon (tasselled wobbegong) (fig. 38)

The two specimens included in the analysis, both of which were collected in Australia, were identical in sequence. They represent the more southern elements of the distribution of this species, which is also known from Indonesia, Papua New Guinea, and possibly Malaysia. Specimens of this species grouped along with, but outside the six Orectolobus species.

Brachaeluridae (blind sharks)

Brachaelurus colcloughi (bluegray carpetshark) (fig. 38)

A total of four specimens of this Australian endemic species were included in the analysis. All came from Queensland, Australia; two were specimens deposited in the Australian National Fish Collection (GN6782  =  ANFC H 6849-01 and GN6785  =  ANFC H 6849-04). A remarkable amount of variation was seen among these specimens. The range in pairwise differences was 0–21, with an average of 10.5. Thus, this species is worthy of further investigation.

Brachaelurus waddi (blind shark) (fig. 38)

The single sample of this Australian endemic species was taken from a specimen collected from New South Wales, Australia, and deposited in the Australian Museum, Sydney (GN2265  =  AMS I31253-005). The average of the pairwise differences between this specimen and those of B. colcloughi was 113.5. These two species are considered congeneric by Last and Stevens (2009) and distinct genera (i.e., Heteroscyllium and Brachaelurus) by Compagno (1984b, 2005a). The placement of these species requires further attention.

Parascylliidae (collared carpetsharks)

Parascyllium collare (collared carpetshark) (fig. 39)

A total of six specimens of this Australian endemic, collected from Victoria and New South Wales, were included in the analysis. They were found to comprise a tight cluster. The range in pairwise differences among specimens was 0–2, with an average of pairwise differences of 0.9.

Heterodontiformes (bullhead sharks)

Heterodontidae (bullhead sharks)

Heterodontus mexicanus (Mexican hornshark) (fig. 40)

The 22 samples of this species included in the analysis were all collected from the Gulf of California, and thus represent only the more northern elements of an eastern Pacific distribution of this species. Three of these specimens were deposited in the Texas Cooperative Wildlife Collection (GN5178  =  TCWC 7572.01, GN5224  =  TCWC 7576.02, and GN5231  =  TCWC 7583.01) and two in the Instituto de Biologia, Universidad Nacional Autonóma de México (GN5390  =  IBUNAM PE9509 and GN5397  =  IBUNAM PE9511). The range in pairwise differences among these specimens was 0–6, with an average of 1.8.

Heterodontus portusjacksoni (Port Jackson shark) (fig. 40)

The seven specimens of this species included in the analysis were collected from Australia in New South Wales and Western Australia and thus span the western and eastern elements of the Australian distribution of this species. The sequences of all seven specimens were identical. Three of these specimens are deposited in the Australian National Fish Collection (GN4841  =  ANFC H 6340-01, GN4842  =  ANFC H 6354-09, and GN4843  =  ANFC H 6354-11).

Heterodontus zebra (zebra bullhead shark) complex (fig. 40)

A total of three specimens originally identified as H. zebra were included in the analysis. These were collected from Australia, Malaysian Borneo, and Japan. The sample from Australia, taken from a specimen in the Australian National Fish Collection (GN4844  =  ANFC H 6581-01), was found to cluster with the specimens of H. portusjacksoni, away from the other two specimens identified as H. zebra. Whereas the average of the pairwise differences between the Australian specimen and those in the H. portusjacksoni cluster was 13, the average of the pairwise differences between the Australian specimen and those of H. zebra from Japan and Malaysian Borneo was 24. Specimens from Japan and Malaysian Borneo differed from one another by two bases. The average of the pairwise differences between the two specimens from Japan and Malaysian Borneo and those of H. portusjacksoni was 19. These results suggest that the specimen from Australia probably represents an undescribed species distinct from both H. zebra and H. portusjacksoni. Given that the type locality of H. zebra is China, we have referred to specimens from Japan and Malaysian Borneo as H. zebra, and used the designation H. cf. zebra for the Australian form. A taxonomic revision of this group is currently being undertaken by P.L. and W.W.

Heterodontus galeatus (crested bullhead shark) (fig. 40)

The analysis included four specimens of this eastern Australian endemic, all collected from New South Wales, Australia. The sequences of these four specimens were identical. The average of their pairwise differences from H. portusjacksoni was 59, and from H. zebra was 64.

Heterodontus francisci (horn shark) (fig. 40)

The nine specimens of this species included in the analysis were all collected from the Gulf of California, and thus represent only the more northern elements of the disjunct eastern Pacific distribution of this species. Two of these specimens were deposited in the Texas Cooperative Wildlife Collection (GN5225  =  TCWC 7576.01 and GN5254  =  TCWC 7567.06). The range in pairwise differences among these specimens was 0–1. The average of the pairwise differences between this species and the other included described species were as follows: between H. mexicanus was 82.9, between H. portusjacksoni was 76.3, between H. zebra was 74.4, between H. cf. zebra was 77.3, and between H. galeatus was 95.3.

Squaliformes (dogfish sharks)

Squalidae (dogfish sharks)

Squalus species

The genus Squalus has recently undergone intensive scrutiny and revision with, for example, 11 new species being described in 2007 alone (e.g., see Last et al., 2007d), and with multiple species having been resurrected in the last several years (e.g., see Last et al., 2007d; Ebert et al., 2010). In our treatment of species in this genus, we have attempted to follow this new taxonomy as closely as possible.

Squalus acanthias (piked dogfish) (fig. 41)

The analysis included a total of 176 specimens of this species, collected from the western North Atlantic off several states including Rhode Island and Maryland, the western South Pacific off New Zealand, Tasmania, and the eastern South Pacific off Chile. The three specimens from Tasmania were vouchered (GN4956  =  ANFC H 2921-03, GN4957  =  ANFC H 4266-01, and GN4958  =  ANFC H 4876-01). The analysis yielded a single cluster with a range in pairwise differences among specimens of only 0–10, and an average of 2.2. This relatively low variation is one of the most striking results of our analysis given the small size of this essentially benthic species, which belongs to a genus that otherwise shows rather extensive regional diversification.

Squalus suckleyi (spotted spiny dogfish) (fig. 41)

A total of eight specimens, collected from the eastern North Pacific were included in the analysis. These specimens clustered independently from the specimens of S. acanthias. The specimens of S. suckleyi had a range of pairwise differences of 2–15, with an average pairwise difference of 6.5. The average of the pairwise differences between the specimens of S. suckleyi and those of the S. acanthias was 11.9. This result is consistent with the work of Ward et al. (2007), Hauser (2009), and Verissimo et al. (2010) regarding the distinct nature of specimens from the North Pacific, and thus also supports the resurrection of S. suckleyi for specimens previously identified as S. acanthias collected from localities in the North Pacific as implemented by Ebert et al. (2010).

Squalus sp. (fig. 42)

In total, eight specimens collected from the western South Atlantic off the coast of Brazil were included in the analysis. The range in pairwise differences among these specimens was 0–7, with an average of 3.1. These specimens have been referred to here preliminarily as Squalus sp. This identity remains to be explored further in the context of the species known to occur off the coast of Brazil (e.g., see Gadig, 2001).

Squalus cf. mitsukurii (shortspine spurdog) (fig. 42)

Six specimens, originally identified as the Squalus mitsukurii, a species described from Japan, all collected from South Africa, were included in the analysis. The range in pairwise differences among these specimens was 0–2, with an average of 0.9. Conversations with Dave Ebert lead us to believe these specimens represent an undescribed species, which resembles but is distinct from S. mitsukurii. Thus, we have designated this specimen as S. cf. mitsukurii at this time. A taxonomic treatment of this species is currently underway by Dave Ebert. This species clustered with Squalus sp., Squalus cubensis, and Squalus montalabani. The average of the pairwise differences between S. cf. mitsukurii and these three species was 8.5, 13, and 15 bases, respectively.

Squalus cubensis (Cuban dogfish) (fig. 42)

Two specimens, collected by John Morrissey from Jamaica and identified as S. cubensis, were included in the analysis. These specimens differed from one another by five bases. They represent the more northern elements of the range of this species, which is distributed as far south as the Falkland Islands. The average of the pairwise differences between S. cubensis, and Squalus sp., also from the western Atlantic Ocean, was 12.

Squalus montalbani (Philippine spurdog) (fig. 42)

The analysis included two samples of this recently resurrected (see Last et al., 2007c) Indo-Pacific species, both of which were collected from voucher specimens deposited in the Australian National Fish Collection (GN4981  =  ANFC H 2609-07 and GN4982  =  ANFC H 4623-05). These specimens differed from one another by one base.

Squalus chloroculus (greeneye spurdog) (fig. 42)

Four specimens of this relatively newly described (see Last et al., 2007c) Australian endemic species, collected from Western Australia, Victoria, and Tasmania, and thus from throughout much of its range, were included in the analysis. These samples all came from specimens deposited in the Australian National Fish Collection (GN4962  =  ANFC H 2564-24, GN4963  =  ANFC H 4775-01, GN4964  =  ANFC H 594-01, and GN4980  =  ANFC H 2606-06). The range in pairwise differences among these specimens was 0–3, with an average of 1.5.

Squalus grahami (eastern longnose spurdog) (fig. 42)

The two specimens of this newly described, Australian endemic species (see White et al., 2007c) included in the analysis consisted of one paratype (GN4973  =  ANFC H 4682-02) and one voucher (GN4972  =  ANFC H 4623-03) in the Australian National Fish Collection. Both specimens were collected from New South Wales, Australia; they were identical in sequence.

Squalus edmundsi (Edmund's spurdog) (fig. 42)

Three specimens of this newly described species (see White et al., 2007c) were included in the analysis. They consisted of a paratype (GN4969  =  ANFC H 2605-05) and two voucher specimens (GN4968  =  ANFC H 2591-17 and GN4970  =  ANFC H 3969-15) in the Australian National Fish Collection. All three specimens were collected from Western Australia and thus represent the more southern elements of the distribution of this species, which also occurs in Indonesia and Malaysia. These specimens clustered together; the range in pairwise differences among specimens was 1–4, with an average of 2.7.

Squalus cf. megalops (fig. 42)

A total of 17 specimens, all from southern Africa, and originally identified as S. megalops, were included in the analysis. The range of pairwise differences among these specimens was 0–7, with an average of 2.9. These specimens comprised a cluster distinct from the Australian endemic S. megalops. The average of the pairwise differences among specimens of this cluster and those of S. megalops was 13.2. At this time, these specimens are considered to represent a distinct, and possibly undescribed species, and have been given the designation Squalus cf. megalops. They clustered most closely with S. brevirostris. The average of the pairwise differences between specimens of S. cf. megalops and S. brevirostris was 10.1.

Squalus brevirostris (Japanese shortnose spurdog) (fig. 42)

Two samples of this species, both taken from specimens deposited in the Kagoshima University Museum (GN4996  =  KAUM I 186 and GN4995  =  KAUM I 187), were included in the analysis. Both specimens came from the waters off Japan and thus represent the more northern elements of the distribution of this species, which extends to the South China Sea. These two specimens were identical in sequence and clustered most closely with the specimens of S. cf. megalops.

Squalus megalops (shortnose spurdog) (fig. 42)

The analysis included seven specimens of this species, all but one of which is deposited in the Australian National Fish Collection (GN4974  =  ANFC H 2605-08, GN4975  =  ANFC H 3762-01, GN4978  =  ANFC H 6581-23, GN4976  =  ANFC H 4649-05, GN4977  =  ANFC H 6581-22, and GN4979  =  ANFC H 6581-24). These specimens were collected from Western Australia, Victoria, and New South Wales and thus represent much of the distribution of this species, which is now considered an Australian endemic (see Last and Stevens, 2009). The analysis yielded a single cluster of these specimens. The range in pairwise differences among specimens in this cluster was 0–6, with an average of 1.9.

Squalus formosus (Taiwan spurdog) (fig. 42)

Three specimens of this recently described species (White and Igl?sias, 2011) from Taiwan were included in the analysis. The range in pairwise differences among these specimens was 0–2, with an average of 1.3. All three specimens are deposited in the University of Michigan Museum of Zoology (GN969  =  UMMZ 231956 and GN976 and GN980  =  UMMZ 231963). These specimens clustered most closely with the specimens of S. albifrons. The average of the pairwise differences between specimens in these two clusters was 29.3. Ward et al. (2007) also found that specimens of this species (as “sp. Taiwan highfin”) clustered most closely with S. albifrons from eastern Australia.

Squalus albifrons (eastern highfin spurdog) (fig. 42)

The three specimens of this newly described (see Last et al., 2007e) eastern Australian endemic consist of a holotype (GN4960  =  ANFC H 4627-01), a paratype (GN4959  =  ANFC H 3589-01), and a voucher (GN4961  =  ANFC H 4704-01), all deposited in the Australian National Fish Collection. The range in pairwise differences among these specimens was 1–2, with an average of 1.3.

Squalus japonicus (Japanese spurdog) (fig. 42)

All three specimens of this species included in the analysis were collected from Taiwan and thus represent only the more northern elements of the distribution of this species. One of the specimens was deposited in the University of Michigan Museum of Zoology (GN975  =  UMMZ 231962). All three specimens were identical in sequence.

Squalus nasutus (western longnose spurdog) (fig. 42)

The two specimens of this newly described species (see Last et al., 2007b) included in the analysis were collected from Western Australia. One of these was deposited in the Australian National Fish Collection (GN4983  =  ANFC H 6413-01). These specimens differed from one another by one base. They clustered most closely to the S. japonicus cluster, as was also seen by Ward et al. (2007). The range of pairwise differences between specimens in these two species was 9.5.

Squalus crassispinus (fatspine spurdog) (fig. 42)

The analysis included three specimens of this newly described (see Last et al., 2007a) species, two of which are paratypes (GN4965  =  ANFC H 4649-03 and GN4966  =  ANFC H 4649-04) and one a voucher (GN4967  =  ANFC H 6412-01), all deposited in the Australian National Fish Collection. The analysis yielded a single cluster; the range in pairwise differences among these specimens was 1–8, with an average of 5.3.

Cirrhigaleus australis (southern Mandarin dogfish) (fig. 42)

The two specimens of this newly described species (see White et al., 2007a) included in the analysis consisted of the holotype (GN4944  =  ANFC H 5789-01) from Tasmania, Australia, and a voucher collected from New Zealand. These specimens differed by four bases. We believe this result supports the suggestion of White et al. (2007a) that C. australis occurs in New Zealand.

Cirrhigaleus asper (roughskin spurdog) (fig. 42)

A single specimen of C. asper collected off Florida was included in the analysis. This specimen clustered with the two specimens of C. australis. The average of the pairwise differences between this specimen and those comprising the cluster of C. australis was 58.5.

Centrophoridae (gulper sharks)

Centrophorus squamosus (leafscale gulper shark) (fig. 43)

The analysis included 50 specimens of Centrophorus squamosus. These came from several localities in the North Atlantic including the mid-Atlantic ridge, the Azores, Madeira, and Scotland, as well as New Zealand, Australia, and the southwest Indian Ocean, and thus represent much of the global distribution of this species. Only two of these samples come from museum specimens (GN4926  =  ANFC H 5343-06 and GN6614  =  MMF 36123). The range in pairwise differences among these 50 specimens was 0–9, with an average of 1.3. It is interesting to note that 28 of the specimens were identical in sequence and these came from the Azores, Scotland, the mid-Atlantic ridge, Australia, New Zealand, and the southwest Indian Ocean.

Centrophorus granulosus (gulper shark) (fig. 43)

In combination, the 13 specimens of this species included in the analysis were collected from the western North Atlantic, the Gulf of Mexico, Caribbean Sea, Madeira, the coast of mainland Portugal, and New South Wales, Australia. The overall distribution of this species is uncertain pending a taxonomic revision of this group. The two samples from Australia came from specimens deposited in the Australian National Fish Collection (GN4924  =  ANFC H 5343-06 and GN4925  =  ANFC H 5343-07); the two samples from mainland Portugal came from specimens in the museum in Funchal (GN6613  =  MMF 36122 and GN6615  =  MMF 36124). The range in pairwise differences seen among these 13 specimens was 0–2, with an average of 0.6. The mean of the pairwise differences between specimens of C. granulosus and those of C. squamosus was 37.8. A similar lack of divergence between these two species was reported by Moura et al. (2008).

Centrophorus cf. lusitanicus (fig. 43)

Four specimens taken from the Mozambique Channel off Madagascar were included in the analysis. The images for these specimens most closely resemble the illustration of C. lusitanicus from southern Africa in Bass et al. (1986). However, since no specimens from near the type locality for this species are included (i.e., Portugal) this species is referred to as C. cf. lusitanicus. The range in pairwise differences among these specimens was 1–5, with an average of pairwise differences of 3.2. Taxonomic revision of this complex is being undertaken by W.W. and Dave Ebert.

Centrophorus sp. 1 (fig. 43)

Two specimens collected from Jamaica by John Morrissey were included in the analysis and were identical in sequence. They clustered most closely with, but independently from, the four specimens of Centrophorus cf. lusitanicus collected from Madagascar. The average of the pairwise differences between these specimens and those of C. cf. lusitanicus was 8.8. Unfortunately, images are not available for these specimens. We have referred to them as Centrophorus sp. 1 until their identity can be examined in more detail. It is possible these specimens represent an undescribed species.

Centrophorus harrissoni (longnose gulper shark) (fig. 43)

Three specimens of this Australian/New Zealand endemic were included in the analysis. The range in pairwise differences among these three specimens was 1–5; the mean was 3.3. All three came from specimens collected from Tasmania (see White et al., 2008) and deposited in the Australian National Fish Collection (GN4941  =  ANFC H 6307-07, GN4942  =  ANFC H 6309-05, and GN4943  =  ANFC H 6500-01).

Centrophorus isodon (blackfin gulper shark) (fig. 43)

Two specimens identified as C. isodon were included in the analysis. Both specimens (GN4338  =  JPAG 227 and GN4392  =  JPAG 225) were collected from the Philippines and were treated by Compagno et al. (2005b) as new records of this species from this region. These specimens differed in sequence by only one base. The average of the pairwise differences between these specimens and those of C. harrissoni, with which they were closely clustered, was 11.5.

Centrophorus sp. 2 (fig. 43)

The analysis included three specimens, all from Taiwan, two of which came from vouchers (GN973  =  UMMZ 231976 and GN974  =  UMMZ 231959). This species belongs to the C. lusitanicus complex but likely represents an undescribed species. The range in pairwise differences among these specimens was 0–3, with an average of 2. Taxonomic revision of this complex is being undertaken by W.W. and Dave Ebert.

Centrophorus sp. 3 (fig. 43)

A single specimen collected from the Philippines that resembles C. lusitanicus was included in the analysis. This specimen (GN4348  =  JPAG 226) was considered by Compagno et al. (2005b) to represent a new record of C. lusitanicus for the Philippines. The average of the pairwise differences between this specimen and the three specimens of Centrophorus sp. 2 from Taiwan was 13.7. This species is possibly conspecific with specimens off Indonesia that were tentatively referred to as C. cf. lusitanicus by White et al. (2006). However, this species likely represents an undescribed species. Taxonomic revision of this complex is being undertaken by W.W. and Dave Ebert.

Centrophorus zeehaani (southern dogfish) complex (fig. 43)

The analysis yielded a single cluster, consisting of a total of five specimens, three of which were paratypes of the newly described species C. zeehaani deposited in the Australian National Fish Collection (GN4932  =  ANFC H 6628-01, GN4933  =  ANFC H 6628-03, and GN4934  =  ANFC H 6628-07) (see White et al., 2008). The remaining two specimens were collected from Angola and Madeira. The range in pairwise differences among all five specimens was 0–1. As C. zeehaani is currently considered an endemic of southern Australia, we have used the provisional designation C. cf. zeehaani for the specimens from Madeira and Angola until this unusual finding can be explored further. We note that photographs of the specimen from Madeira are available in the database; it is superficially similar morphologically to the Australian specimens.

Centrophorus moluccensis (smallfin gulper shark) (fig. 43)

Ten specimens, which were morphologically consistent with C. moluccensis, were included in the analysis: one specimen came from the southwestern Indian Ocean, one from Malaysian Borneo, six specimens were deposited in the Australian National Fish Collection (GN4921  =  ANFC H 3599-04, GN4922  =  ANFC H 6410-01, GN4923  =  ANFC H 6410-02, GN4927  =  ANFC H 5343-07, GN4928  =  ANFC H 4873-03, and GN4929  =  ANFC H 2575-26), five of these were collected from Australian and one from New Zealand localities, while three samples came from the Philippines. These specimens thus represent a large portion of the Indo-West Pacific distribution of this species. Two of the three specimens from the Philippines (GN4329  =  JPAG 257 and GN4351  =  JPAG 258) were among those treated by Compagno et al. (2005b) as Centrophorus cf. moluccensis. The analysis yielded a cluster consisting of two potential subclusters. The range in pairwise differences among all 10 specimens was 0–14, with an average of 6.4. One of the subclusters consisted of the specimens collected from Borneo and Australia; the other consisted of the three specimens from the Philippines. The range in pairwise distances among specimens in the former subcluster was 0–8 and among specimens in the latter subcluster was 5–11. The average of the pairwise distances among specimens in the two subclusters was 9.9.

Deania calcea (birdbeak dogfish) (fig. 44)

A total of 26 specimens of this species were included in the analysis. These came from the mid-Atlantic ridge, Ireland, Scotland, southern Africa, New Zealand, Tasman Sea, and southwest Indian Ocean, and thus represent much of the distribution of this species. Three of these specimens were from the Australian National Fish Collection (GN4945  =  ANFC H 4873-02, GN4946  =  ANFC H 5343-08, and GN4947  =  ANFC H 5365-01). The analysis yielded essentially a single cluster. The range of pairwise differences among specimens in the cluster was 0–11, with an average of 3.3.

Deania cf. profundorum (fig. 44)

The analysis included five specimens preliminarily identified as D. cf. profundorum, all taken from the coast of mainland Portugal and the Azores, but the identities of which remain to be confirmed. The range in pairwise differences among specimens in this cluster was 0–10, with an average of 5.4. Images are available only for the specimen from the Azores and they suggest that at least this specimen is not fully consistent with existing descriptions of Deania profundorum (e.g., Compagno, 1984b). For example, this specimen has a second dorsal fin that differs substantially in shape from that of the true D. profundorum. Further taxonomic work, including specimens definitively identified as D. profundorum, and preferably taken from throughout the extensive distribution of that species is required.

Deania quadrispinosa (longsnout dogfish) complex (fig. 44)

Three specimens identified as D. quadrispinosa were included in the analysis. These were all collected from New South Wales, Australia, and represent only the eastern portion of the distribution of this species, which extends from east of New Zealand to the west coast of South Africa. However, while two of these specimens exhibited sequences that differed from one another by only a single base, the average of the pairwise differences between these two specimens and the third specimen was 22.5, suggesting that they may represent more than a single species. We have given these specimens separate designations (i.e., D. quadrispinosum 1 and D. quadrispinosum 2) in order to call attention to this interesting result.

Somniosidae (sleeper sharks), in part

Centroscymnus coelolepis (Portuguese dogfish) (fig. 45)

A total of 18 specimens of this species were included in the analysis. These came from the western North Atlantic, mid-Atlantic ridge, Madeira, Scotland, and Tasmania, and thus represent all but the South Atlantic and North Pacific elements of the distribution of this species. The three specimens from Madeira are deposited in the Museu de Hist?ria Natural e Aqu?rio in Funchal (GN6616  =  MMF 36129, GN6626  =  MMF 36784, and GN6606  =  MMF 36090). Despite the disparate localities from which these specimens came, they were remarkably homogeneous in NADH2 sequence; the range in pairwise differences among these 18 specimens was 0–2, with an average of 1.1.

Centroscymnus owstonii (roughskin dogfish) (fig. 45)

The 17 included specimens of this species were collected from the mid-Atlantic Ridge, Madeira, coastal mainland Portugal, and New Zealand. The analysis yielded a single cluster. The range of pairwise differences among these 17 specimens was 0–6, with an average of 1.8. A number of specimens taken from Portugal and New Zealand were identical in sequence. The six specimens collected from Madeira are deposited in the Museu de Hist?ria Natural e Aqu?rio in Funchal (GN6598  =  MMF 36021, GN6599  =  MMF 36044, GN6600  =  MMF 36058, GN6623  =  MMF 36739, GN6619  =  MMF 36142, and GN6625  =  MMF 36741). This species clustered most closely with C. coelolepis. The average of the pairwise differences between C. owstonii and C. coelolepis was 74.7 (or 7.15%). This result is somewhat more conservative than that of Moura et al. (2008) who found the genetic divergence between these two species to be 12.8%.

Oxynotidae (roughsharks)

Oxynotus bruniensis (prickly dogfish) (fig. 45)

A total of 14 specimens of this species were included in the analysis. These were collected from New Zealand and New South Wales in Australia and thus are representative of the temperate Australasian distribution of this species. The analysis yielded one cluster. The range in pairwise differences seen among the 14 specimens was 0–6, with an average of 1.6.

Oxynotus paradoxus (sailfin roughshark) (fig. 45)

The four specimens of this species included in the analysis were all collected from the eastern North Atlantic, including off Scotland, and were all identical in sequence. These specimens represent the northern elements of this species, which is distributed southward to Senegal. The average of the pairwise differences between this species, and O. bruniensis was 55.3. It is of note that both of these oxynotid species grouped within the cluster containing the somniosids.

Somniosidae (sleeper sharks), continued

Proscymnodon plunketi (plunket shark) (fig. 45)

The 12 specimens of this species included in the analysis were all collected from New Zealand and thus represent the more eastern elements of the distribution of this species. Most of these specimens were identical in sequence, but one specimen differed from the others by one base.

Scymnodon ringens (knifetooth dogfish) (fig. 45)

All five specimens identified as this species included in the analysis were collected from Ireland and thus represent the northeastern elements of the distribution of this species. These specimens were identical in sequence. They clustered most closely with specimens of P. plunketi; the average of the pairwise differences between these two species was 59.1.

Centroselachus crepidater (longnose velvet dogfish) (fig. 45)

The analysis included 27 specimens of this currently monotypic genus. These specimens were collected from the mid-Atlantic ridge, Madeira, Scotland, southern Africa, Tasman Sea, and New Zealand and thus represent only the western Pacific and eastern Atlantic portions of the distribution of this species. The specimen collected from the Tasman Sea came was from the Australian National Fish Collection (GN4940  =  ANFC H 4873-01). The two specimens from Madeira are deposited in the Museu de Hist?ria Natural e Aqu?rio in Funchal (GN6617  =  MMF 36134 and GN6618  =  MMF 36136). The analysis yielded are a single cluster and, in fact, several specimens from Madeira, Scotland, and New Zealand were identical in sequence. However, the range of pairwise differences among all 27 specimens was 0–17, with an average of 2.4.

Zameus squamulosus (velvet dogfish) (fig. 45)

Four specimens identified as Zameus squamulosus were included in the analysis. Two were collected from southern Africa and two from Tasmania and thus represent only a portion of the southern hemisphere distribution of this essentially cosmopolitan species. The range in pairwise differences among specimens was 0–6 bases, with an average of 4.6. They clustered most closely with C. crepidater. The average of the pairwise differences between specimens of these two species was 107.3.

Centroscymnus sp. 1 (fig. 45)

The analysis included a specimen collected from New Zealand, and preliminarily identified as Centroscymnus sp. by Di Tracey. This specimen clustered provisionally along with, but outside the group comprised of the two specimens of Zameus squamulosus and the 27 specimens of Centroselachus crepidater (well away from the specimens of the two species of Centroscymnus included in the analysis). The average of the pairwise differences between this specimen and those of Z. squamulosus was 166. The average of the pairwise differences between this specimen and those of C. crepidater was 160.7. We have referred to this specimen here as Centroscymnus sp. 1. However, it should be noted that results from barcode (COI) data, which show similar results within the Somniosidae, include a clade consisting of Z. squamulosus, C. crepidater, and a single specimen of Scymnodalatias albicauda from New Zealand. Thus, it is highly likely that this species actually refers to S. albicauda, but additional samples are required to confirm this.

Somniosus microcephalus (Greenland shark) (fig. 45)

Four of the five specimens of this species included here came from the Northwest Territories in northern Canada, the remaining specimen was from the eastern Atlantic. Thus, our sample represents much of the breadth of the distribution of this species. The range in pairwise differences among specimens was 0–1.

Somniosus pacificus (Pacific sleeper shark) (fig. 45)

All four specimens of this species included in the analysis were collected from Alaska and thus represent only the eastern Pacific elements of the distribution of this species, which also includes the western Pacific. The range in pairwise differences among these specimens was 0–2, with an average of 1.2. The average of the pairwise differences between specimens of this species and those of S. microcephalus was 9.2. This result is consistent with that of Murray et al. (2008) with respect to the distinction between the latter two species.

Somniosus rostratus (little sleeper shark) (fig. 45)

The single specimen of this species included in the analysis was collected from the coast of France, and thus is representative of the eastern North Atlantic distribution of this species. The average of the pairwise differences between this specimen and the specimens comprising the cluster of S. microcephalus was 32.4, and between those comprising the cluster of S. pacificus was 32.8.

Etmopteridae (lantern sharks)

Etmopterus spinax (velvet belly) (fig. 46)

All but one of the 21 specimens of this species included in the analysis were collected from the Azores; the remaining specimen came from Scotland. Thus, our sample represents only some of the more northern elements of the distribution of this species, which occurs throughout the eastern Atlantic from Norway to Gabon. The analysis yielded essentially a single cluster. The range in pairwise differences among specimens in this cluster was 0–8, with an average of pairwise differences of 2.5.

Etmopterus princeps (great lanternshark) (fig. 46)

The seven specimens of this species included in the analysis were collected from throughout much of the distribution of this species, consisting of Connecticut in the western North Atlantic, Madeira, and the eastern North Atlantic. The two specimens from Madeira were deposited in the Museu de Hist?ria Natural e Aqu?rio in Funchal (GN6608  =  MMF 36093 and GN6611  =  MMF 36103). The analysis yielded essentially a single cluster. The range in pairwise differences among specimens within this cluster was 0–6, with an average of 3.4.

Etmopterus cf. unicolor complex (fig. 46)

Four specimens originally identified as Etmopterus unicolor were included in the analysis. Two of these were collected by Peter Smith from New Zealand. The other two specimens are from the Australian National Fish Collection (GN4952  =  ANFC H 5673-02 and GN4954  =  ANFC H 5674-08), both of which were collected from the eastern Indian Ocean. These four specimens grouped together but in two distinct clusters, one consisting of the specimens from New Zealand, the other of specimens from the eastern Indian Ocean. The specimens from New Zealand differed from one another by base; those in the latter cluster differed from one another by seven bases. The average of the pairwise differences between specimens in these two clusters was 14. Given the type locality of this species is Japan, a locality not represented by our specimens, we have referred to the specimens from New Zealand as E. cf. unicolor 1 and those from the eastern Indian Ocean as E. cf. unicolor 2. However, some consideration should be given to the specimens in the first of these clusters being conspecific with the newly described Etmopterus viator of Straube et al. (2011). The second cluster could represent one of the many species of this genus not nominally represented in the analysis.

Etmopterus baxteri (New Zealand lanternshark) (fig. 46)

In total, 11 specimens of this possibly southern global species, all from New Zealand, were included in the analysis, which yielded a single cluster. Ten of the specimens were almost identical in sequence, but one of the specimens differed from the remaining nine by seven or eight bases. The average of pairwise differences among the 11 specimens was 1.5.

Etmopterus virens (green lanternshark) (fig. 46)

A single specimen of this species, collected from the western North Atlantic, was included in the analysis.

Etmopterus gracilispinis (broadband lanternshark) (fig. 46)

A single specimen of this species, collected from the western North Atlantic was included in the analysis. This specimen clustered with the specimen of E. virens. However, the difference between the specimens of these two species was 126.

Etmopterus pusillus (smooth lanternshark) (fig. 46)

The analysis included 20 specimens of this species. These were collected from a diversity of localities in the eastern North Atlantic, including the Azores, Madeira, and the coast of mainland Portugal, as well as from New South Wales in Australia. However, given the extent of the reported distribution of this species, these localities represent only a small portion of its patchy global distribution. Three specimens from Portugal (GN6603  =  MMF 36075, GN6620  =  MMF 36530, and GN6624  =  MMF 36740), and one from Australia (GN4951  =  ANFC H 5956-01) were vouchered. The analysis yielded essentially a single cluster. The range in pairwise differences among these specimens was 0–11, with an average of 3.5.

Etmopterus bigelowi (blurred smooth lanternshark) (fig. 46)

The four specimens of this species included in the analysis represent only a very small portion of the patchy global distribution of this species, having come from the Gulf of Mexico and the mid-Atlantic Ocean. The analysis yielded a single cluster and the range in pairwise differences among specimens in this cluster was 0–2, with an average of 1.

Etmopterus splendidus (splendid lanternshark) (fig. 46)

Both samples of this species included in the analysis were collected from Taiwan, from specimens deposited in the University of Michigan Museum of Zoology (GN994 and GN995  =  UMMZ 231969). These specimens differed from one another by two bases. They clustered together, most closely allied with the E. pusillus and E. bigelowi clusters. The average of the pairwise differences between E. splendidus and E. pusillus was 128.1 and between E. splendidus and E. bigelowi was 105.5.

Etmopterus lucifer (blackbelly lanternshark) (fig. 46)

All 10 specimens of this species were collected from the eastern portions of the relatively extensive, but poorly defined, distribution of this species, having come from New South Wales in Australia and also from New Zealand. The analysis yielded a single cluster. The range in pairwise differences among these specimens was 0–6, with an average of 2.

Etmopterus molleri (slendertail lanternshark) (fig. 46)

The four specimens of this species included in the analysis were collected from Taiwan and thus represent only a very small portion of the distribution of a possible species complex (G. Burgess, personal commun.). All four specimens are deposited in the University of Michigan Museum of Zoology (GN996, GN997, GN998, and GN999  =  UMMZ 231971). The analysis yielded a single cluster. The range in pairwise differences among these specimens was 0–6, with an average of 3. These specimens clustered most closely with those of E. lucifer, with an average of the pairwise differences among specimens of these two species was 110.5. However, given that E. burgessi is also known from Taiwan (Schaaf-DaSilva and Ebert, 2006) and was not included in our analysis, the identity of the E. molleri cluster should be confirmed by careful comparison of the deposited specimens with verified specimens of E. burgessi.

Centroscyllium fabricii (black dogfish) (fig. 46)

The 10 specimens of this species included here represent the North Atlantic elements of the distribution of this Atlantic species having come from Connecticut, the Azores, and Scotland. The southern Atlantic elements of the distribution were not represented. The analysis yielded a single cluster; the range of pairwise differences among specimens in this cluster was 0–8, with an average of 3.6.

Dalatiidae (kitefin sharks)

Dalatias licha (kitefin shark) (fig. 47)

The analysis included 17 specimens of this broadly distributed monotypic genus, taken from the eastern North Atlantic (i.e., the Azores and Scotland), the western North Pacific (i.e., Taiwan) and the western South Pacific (i.e., New South Wales, Australia, and New Zealand). However, much of its distribution (e.g., the Indian Ocean elements), was not represented. One of the specimens from Taiwan is deposited at the University of Michigan Museum of Zoology (GN985  =  UMMZ 231958). The analysis yielded essentially a single relatively homogeneous cluster given the breadth of geographic regions from which these specimens came. The range in pairwise differences among specimens in this cluster was 0–5, with an average of 1.9.

Isistius brasiliensis (cookiecutter shark) (fig. 47)

The analysis included a total of three specimens of this remarkably broadly distributed species. These were collected from the central Pacific and Atlantic oceans. The analysis yielded essentially a single cluster with the specimen from the Pacific Ocean clustering outside the two specimens collected from the Atlantic Ocean. The range in pairwise differences among these three specimens was 4–10, with an average of 8.

Squaliolus aliae (smalleye pygmy shark) (fig. 47)

The three specimens included in the analysis were collected from Western Australia and Taiwan and thus are relatively representative of the distribution of this species. The specimen from Western Australia is in the Australian National Fish Collection (GN4955  =  ANFC H 6416-02). The analysis yielded essentially a single cluster comprised of these specimens. However, the specimen from Australia was relatively divergent from the two specimens from Taiwan. The range in pairwise differences among all three specimens was 5–14, with an average difference of 10 bases.

Euprotomicrus bispinatus (pygmy shark) (fig. 47)

The analysis included a single specimen of this broadly distributed species collected from the North Pacific. This species clustered among the two species of Squaliolus, most closely allied with S. aliae, suggesting that the generic placement of this, the only species in its genus, might warrant closer scrutiny. The average of the pairwise differences between this specimen and those of S. aliae was 89.3.

Squaliolus laticaudus (spined pygmy shark) (fig. 47)

The two specimens included in the analysis were collected from the Azores, and thus represent only a very small portion of the distribution of this almost circumtropical species. These specimens clustered together and differed from one another by eight bases. The average of the pairwise differences between specimens of S. laticaudus and those of S. aliae was 140.8 and between E. bispinatus was 187.7.

Squatiniformes (angel sharks)

Squatinidae (angel sharks)

Squatina oculata (smoothback angelshark) (fig. 48)

The 12 specimens included in the analysis were all taken from Senegal and Sierra Leone and thus represent the center of the distribution of this species, which extends coastally from France to Namibia. The analysis yielded a single cluster; all 12 specimens were identical in sequence.

Squatina formosa (Taiwan angelshark) (fig. 48)

The analysis included three specimens of this species which is possibly endemic to Taiwan. These specimens comprised a single cluster with a pairwise difference among specimens of 0–1. However, this genus includes several similar species from Taiwan not included in the analysis (e.g., S. japonica and S. nebulosus) (see Walsh and Ebert, 2007). The identity of this cluster requires confirmation in the context of these other taxa.

Squatina tergocellatoides (ocellated angelshark) (fig. 48)

The analysis included two specimens collected from Malaysian Borneo that represent the distinctive S. tergocellatoides. These specimens were identical in sequence to one another and clustered most closely with the specimens of S. formosa. The average of the pairwise differences between this species and S. formosa from Taiwan was 88.3.

Squatina aculeata (sawback angelshark) (fig. 48)

The 11 specimens of this species included in the analysis were all collected from Senegal and thus represent the center of the distribution of this species, which extends coastally from the western Mediterranean to Namibia. The analysis yielded a single cluster. The range in pairwise differences seen among specimens was 0–3, however, the majority of these specimens were identical in sequence. The average of the pairwise differences among the 11 specimens was 0.7.

Squatina dumeril (sand devil) (fig. 48)

The analysis included eight specimens representing the more northern elements of the distribution of this species including New Jersey and the northern Gulf of Mexico. The analysis yielded a single cluster, with pairwise differences seen among these specimens of 0–12, with an average of 5.6.

Squatina californica (Pacific angelshark) (fig. 48)

The seven specimens included were collected from California and the Gulf of California and thus represent much of the distribution of this species. The range in pairwise differences among specimens was 0–6, with an average of pairwise differences of 3.8. This species clustered most closely with S. dumeril, a result consistent with the findings of Stelbrink et al. (2010). The average of the pairwise differences among specimens of these two species was 46.

Squatina albipunctata (eastern angelshark) (fig. 48)

A single specimen of this newly described eastern Australian endemic species (see Last and White, 2008b) was included in the analysis. It clustered most closely with S. dumeril and S. californica. The mean of the pairwise differences between this specimen and those of the former species was 112.5, and between this specimen and those of the latter species 109.1.

Echinorhinidae (bramble sharks)

Echinorhinus brucus (bramble shark) (fig. 48)

Three specimens of this species, all collected from the western North Atlantic (i.e., Louisiana and North Carolina), were included in the analysis. One of these specimens came from the University of Florida Fish Collection (GN1983  =  UFFC 103000), and the other a specimen deposited in the Tulane University Museum of Natural History (GN1067  =  TU 172379). These specimens represent only a very small portion of the distribution of this species. The range in pairwise differences among these specimens was 2–5, with an average of 4.

Echinorhinus cookei (prickly shark) (fig. 48)

In total, two specimens of this species were included in the analysis. One came from the Australian National Fish Collection (GN4998  =  ANFC H 6115-01) collected from Queensland, and the other was collected from California. These specimens thus represent only some of the more southern elements of the Pacific distribution of this species. These specimens differed by one base, clustering most closely with those of E. brucus. The average pairwise difference between specimens of these species was 59.3.

Pristiophoriformes (sawsharks)

Pristiophoridae (sawsharks)

Pliotrema warreni (sixgill sawshark) (fig. 48)

Two specimens of this southern Africa endemic were included in the analysis. The specimens were identical in sequence.

Pristiophorus japonicus (Japanese sawshark) (fig. 48)

The single specimen of this species was collected from Japan. It clustered most closely with the specimens of P. warreni; the average of the pairwise differences between these two species was 130.

Pristiophorus cirratus (longnose sawshark) (fig. 48)

The two specimens of this southern Australian endemic species were both collected from New South Wales and differed by a single base. The average of the pairwise differences between this species and P. warreni was 152.5, and between this species and P. japonicus 137.5.

Hexanchiformes (cow and frilled sharks)

Hexanchidae (sixgill and sevengill sharks)

Hexanchus nakamurai (bigeye sixgill shark) and Hexanchus vitulus (fig. 49)

The analysis included three specimens collected from Madagascar, the Philippines (GN2246  =  JPAG 053 of Compagno et al., 2005b), and New South Wales in Australia as well as one specimen from the Bahamas in the Caribbean Sea. It yielded a cluster consisting of the three specimens from the Indo-Pacific localities grouping with the specimen from the Bahamas. The range of pairwise differences among the three Indo-Pacific specimens was 5–15, with an average of 11.3; the average of the pairwise differences between the specimen from Australia and those from Madagascar and the Philippines was 14.5. However, the average of the pairwise differences between the specimen from the Bahamas and the three Indo-Pacific specimens was 80.7. Hexanchus nakamurai and H. vitulus have been considered to be synonyms (e.g., Compagno, 1984b; Taniuchi and Tachikawa, 1991) with a relatively broad distribution, occurring throughout the tropics and subtropics globally (Last and Stevens, 2009). However, our results suggest that both may actually represent valid species. Given the type locality of H. vitulus is Bimini in the Bahamas and the type locality of H. nakamurai is Taiwan, the specimen from the Bahamas has been identified as H. vitulus, and those from Indo-Pacific localities have been given the designation H. nakamurai. Clearly, the conspecificity of the bigeye sixgill sharks should be examined in more detail.

Hexanchus griseus (bluntnose sixgill shark) (fig. 49)

Specimens of this species included here were both collected from the eastern and central sectors of the North Atlantic. They differed from one another by two bases and clustered most closely with their two congeners. The average of pairwise differences between H. griseus and H. nakamurai was 103, and between H. griseus and H. vitulus 94.

Heptranchias perlo (sharpnose sevengill shark) (fig. 49)

All four specimens of this monotypic genus included in the analysis came from Taiwan and thus represent only essentially a single locality of this global species. These four specimens were identical in sequence. All but one of these specimens are deposited in the University of Michigan Museum of Zoology (GN977, GN978, and GN979  =  UMMZ 231961).

Notorhynchus cepedianus (broadnose sevengill shark) (fig. 49)

The four specimens of this species included in the analysis were all collected from localities in the Pacific Ocean (e.g., California and Australia) and thus represent only a portion of the distribution of this essentially global species. The range in pairwise differences among these specimens was 0–6, with an average of 3.8.

Chlamydoselachidae (frilled sharks)

Chlamydoselachus anguineus (frilled shark) (fig. 49)

The three specimens of this species included in the analysis were collected from the mid-Atlantic ridge, Scotland, and Japan. Despite the geographic breadth of our samples, they represent only a relatively small portion of the distribution of this species. The range in pairwise differences among these three specimens was 2–5, with an average of 4.

BATOIDS

Rajiformes (batoids)

Dasyatidae (whiptail stingrays)

Himantura gerrardi (whitespotted whipray) complex (fig. 50)

As has been suggested by a number of authors (e.g., Manjaji, 2004; White et al., 2006; Ward et al., 2008), there is reason to believe that Himantura gerrardi represents a complex of species that urgently requires taxonomic revision. Our analysis included 71 specimens exhibiting the basic color pattern of H. gerrardi. However, these rays display considerable morphological variability within forms, without obvious differences between forms. In the absence of a formal taxonomic treatment and also material from the type locality of India, we have assigned numerical designations to the five clusters of these specimens that we believe, based in large part, on their genetic distances, may represent distinct species. The haplotype map colored by phenotype (fig. 94A) generally supports the distinct identity of each cluster, with no haplotypes shared among specimens with different species designations. However, there is considerable haplotype variation among specimens of H. cf. gerrardi 1, and also among the specimens of H. cf. gerrardi 2. The geographic haplotype map (fig. 94B) supports the notion that at least two species of H. cf. gerrardi (H. cf. gerrardi 1 and H. cf. gerrardi 2) cooccur in the Indo-Pacific.

Himantura cf. gerrardi 1 (fig. 50)

The analysis included a cluster of 39 specimens that have been given the designation H. cf. gerrardi 1. These specimens were collected from throughout the coasts of Malaysian and Indonesian Borneo as well as Thailand. One of the Borneo specimens has been deposited in the California Academy of Sciences (GN4521  =  CAS 229036). Given that the range in pairwise differences seen among specimens in this cluster was 0–22, with an average of 5.5, this designation is very conservative. It is possible that this cluster includes representation of perhaps two species. For example, within this cluster there was a distinct subcluster comprised of two of the specimens collected from Sabah in eastern Malaysian Borneo.

Himantura cf. gerrardi 5 (fig. 50)

One specimen, collected from Vietnam, clustered along with but outside the 39 specimens of Himantura cf. gerrardi 1. The average of the pairwise differences between specimens of these two taxa was 34.8.

Himantura cf. gerrardi 2 (fig. 50)

The analysis yielded a second cluster comprised of, at least conservatively, 29 specimens. However, as circumscribed here, this cluster includes one specimen (GN3022) that is relatively more divergent from the remaining specimens in the cluster. The range in pairwise differences among the 29 specimens in this cluster was 0–19 if this specimen is included (with an average of 3.4); the range is 0–8 if this specimen is excluded. Like H. cf. gerrardi 1, specimens of H. cf. gerrardi 2 were also collected from throughout the coasts of Malaysian and Indonesian Borneo; two of these specimens were vouchered (GN3426  =  IPPS BO158 and GN3431  =  IPPS BO163). At least two sympatric species of whitespotted whiprays may exist on the island of Borneo, given that the average of the pairwise differences between specimens comprising the clusters of H. cf. gerrardi 1 and H. cf. gerrardi 2 was 64.4.

Himantura cf. gerrardi 3 (fig. 50)

A single specimen collected from the Gulf of Oman was included in the analysis. This specimen grouped independently from the other clusters of whitespotted whiprays. Thus, this specimen has also been given a distinct numerical designation. The averages of the pairwise differences between this specimen and those comprising the clusters of H. cf. gerrardi 1, H. cf. gerrardi 2, and H. cf. gerrardi 5 were 74.7, 20.4, and 70, respectively. Additional specimens of whitespotted whiprays from the Gulf of Oman need to be investigated using both morphological and molecular methods.

Himantura cf. gerrardi 4 (fig. 50)

The analysis also included a specimen collected from the Mozambique Channel off Madagascar. This specimen grouped outside all three of the above clusters of whitespotted whiprays. The averages of the pairwise differences between this specimen and those of H. cf. gerrardi 1, H. cf. gerrardi 2, H. cf. gerrardi 3, and H. cf. gerrardi 5 were 67.2, 35.7, 47, and 65, respectively. Additional collections of whitespotted whiprays from the Mozambique Channel are required.

Himantura astra (blackspotted whipray) (fig. 50)

The analysis included nine specimens identified as this newly described Australasian species (see Last et al., 2008a). Six of these specimens were collected from the Arafura Sea, two from the Gulf of Carpentaria, and one from Moreton Bay. The range in pairwise differences among all nine specimens was 0–32, with an average of 7.9. One of the specimens collected from the Gulf of Carpentaria and the specimen from Moreton Bay clustered outside the six specimens from the Arafura Sea and one of the specimens from the Gulf of Carpentaria; the range in pairwise differences among the former six specimens was 0–3. The identities of the specimens in this cluster should be examined in more detail as some variation in morphology, such as dorsal surface spot pattern, were also observed among the nine specimens.

Himantura pastinacoides (round whipray) complex (fig. 50)

Five specimens originally identified as belonging to the Indo-Pacific species Himantura pastinacoides were collected from Malaysian Borneo. However, the analysis yielded two distinct, very divergent clusters of these specimens, with an average pairwise difference among specimens in the two clusters of 72. Given that the type locality of H. pastinacoides is Java, and none of our material came from that locality, and the clusters overlap in distribution, we have used numerical designations for specimens from both clusters. They have been referred to as Himantura pastinacoides 1 and H. pastinacoides 2 until such time as the identity of these clusters can be examined in more detail. The three specimens comprising the former cluster were identical in sequence. The two specimens comprising the latter cluster differed from one another by eight bases.

Himantura sp. B (Arabian banded whipray) (fig. 50)

The analysis included two specimens of whiprays collected from the Gulf of Oman that, while they had a banded tail, lacked spots on their disc; they appear conspecific with Himantura sp. B of Manjaji (2004). These specimens differed by three bases, and clustered most closely to but well outside specimens of Himantura pastinacoides 1 and H. pastinacoides 2. However, the average of the pairwise differences between these specimens and those of the former cluster was 136.3 and of the latter cluster was 111.5. While the Gulf of Oman form has been confused with H. gerrardi (see Randall, 1995), it clearly represents a distinct taxon. It is currently being described by Manjaji-Matsumoto, Last, and Moore (in prep.).

Himantura uarnacoides (whitenose whipray) complex (fig. 51)

In total, 32 specimens identified as the Indo-Malay species Himantura uarnacoides were included in the analysis, all collected from around the coast of Borneo. Two of these specimens were deposited in museums (GN4588  =  CAS 229041 and GN3418  =  IPPS BO149). The analysis yielded essentially a single cluster, although one of the specimens from Kalimantan (GN4781) was slightly divergent. The range of pairwise differences within this cluster was 0–20, with an average of 6.3, if this specimen was included.

An additional specimen (GN3366), collected from Sabah in Malaysian Borneo, resembling H. uarnacoides at least superficially, was also included in the analysis. This specimen clustered along with but well outside the main H. uarnacoides group. The average of the pairwise differences between this specimen and the 32 specimens in the H. uarnacoides cluster was 92.7. As this specimen is likely to represent a distinct species, it has been given the designation H. cf. uarnacoides until its identity can be examined in more detail.

Himantura jenkinsii (pointed-nose stingray) (fig. 51)

The analysis included a total of 13 specimens consistent with H. jenkinsii. The majority of these specimens were collected from Borneo and the Philippines; two specimens were collected in Vietnam. Four specimens collected from the Arafura Sea bear dark spots along the posterior margin of their disc which Last and Stevens (2009) noted is consistent with a color morph described as H. draco from South Africa (Compagno and Heemstra, 1984). The analysis yielded essentially a single cluster comprised of all 13 specimens. The range in pairwise differences among all 13 specimens was 0–16, with an average of 5.5. A specimen from the Philippines is deposited at the Philippines National Museum (GN2250  =  JPAG 038).

Himantura fai (pink whipray) (fig. 51)

Four specimens resembling H. fai were included in the analysis; two from Australia, one from the Philippines, and one from Malaysian Borneo. They represent the eastern portion of the distribution of this Indo-West Pacific ray, which has also been reported from the Arabian Peninsula and possibly as far south as South Africa. Although the specimens from Australia differed from one another by 11 bases, those from the other localities were identical to one another. Thus, the range in pairwise differences among all four specimens was 0–12, with an average of 9.5. The specimen from the Philippines (i.e., GN2231  =  BRU 083) was treated by Compagno et al. (2005b).

Himantura uarnak (reticulate whipray) complex (fig. 52)

The analysis yielded four distinct clusters of specimens that were generally consistent in morphology and color pattern with Himantura uarnak as defined by Last and Stevens (2009) based on Australian material (i.e., disc with a pale yellow background with a dense pattern of brown, fine reticulations). However, members of this species complex undergo multiple ontogenetic changes in coloration and these stages need to be described in detail for each of the forms. Unfortunately, such data were not available to inform the identifications of the specimens included here. Moreover, our analysis did not include specimens from the type locality of H. uarnak, the Red Sea, and thus it is unclear which, if any, of the four clusters truly represents H. uarnak. As a consequence, all four of the clusters have been given numerical designations of Himantura uarnak from 1 to 4; each of these is treated separately below.

The haplotype map of phenotypes (fig. 95A), which for comparative purposes also includes the similarly colored H. undulata and H. leoparda, shows no overlap in haplotypes among any of these six species. The haplotype map colored by geography (fig. 95B) indicates that H. undulata, H. leoparda, and as many as three of the four species in the Himantura uarnak complex (i.e., all but H. uarnak 2) may cooccur in Borneo. The H. uarnak species complex is currently being revised by P.L. and B. Mabel Manjaji-Matsumoto.

Himantura uarnak 1 (fig. 52)

The 12 specimens comprising the cluster referred to as Himantura uarnak 1 were collected from Malaysian and Indonesian Borneo and the Philippines. They exhibited a disc with a yellow background and fine brown reticulations comprised of brown lines and some spots. The two specimens (i.e., GN2934 and GN3006) collected from Lahad Datu in Malaysian Borneo were most divergent among the specimens in this cluster; these individuals exhibited a somewhat darker disc and spots. The range of pairwise differences among all 12 of the specimens of this cluster was 0–25, with an average of 8. We note that the Philippine specimens (GN2237  =  JPAG 035, GN2238  =  JPAG 036, and GN2249  =  BRU 027) were treated by Compagno et al. (2005b) as Himantura uarnak.

Himantura uarnak 2 (fig. 52)

The 12 specimens comprising the second cluster, referred to as Himantura uarnak 2, were collected from several localities off northern Australia. Their disc exhibits dark, crisp lines and some spots that may resemble joined hexagons. The range of pairwise differences among specimens of this cluster was 0–6, with an average of 2.6. The average of the pairwise differences between specimens in the H. uarnak 1 and H. uarnak 2 clusters was 54.9. These two clusters grouped together, well away from the third cluster of specimens with this general color pattern.

Himantura undulata (honeycomb whipray) (fig. 52)

A total of 11 specimens of this species were included in the analysis and were found to comprise a single tight cluster. These rays were collected from Malaysian and Indonesian Borneo. Our specimens include juveniles with discs of light background and brown spots (rhomboids) that were light/open internally, to very large, mature animals with discs that exhibit a light background and brown reticulations rather than brown spots. Manjaji-Matsumoto and Last (2008) and Manjaji (2004) considered this to be a senior synonym of H. fava. The range of pairwise differences among specimens in this cluster was 0–3, with an average of 1.5. These specimens were found to group with two of the clusters of Himantura uarnak, i.e., H. uarnak 1 and H. uarnak 2.

Himantura leoparda (leopard whipray) (fig. 52)

In total, four specimens of this newly described Indo-Pacific species (see Manjaji-Matsumoto and Last, 2008) were included in the analysis. These were collected from the Arafura Sea off northern Australia, Singapore, and Indonesian Borneo. These specimens comprised a single cluster. The range in pairwise differences among specimens in this cluster was 1–12, with an average of 7.8.

Himantura uarnak 3 (fig. 52)

The third H. uarnak cluster, which is referred to here as Himantura uarnak 3, was comprised of 15 specimens from Singapore and Malaysian and Indonesian Borneo. These specimens had a disc with a yellow background with fine brown reticulations or small brown spots. The range in pairwise differences among specimens in this cluster was 0–5, with an average of 1.8. However, two other species of Himantura (namely, H. undulata and H. leoparda) grouped more closely with the clusters comprised of specimens identified as H. uarnak 1 and H. uarnak 2 than did specimens of H. uarnak 3. The averages of the pairwise differences between specimens of H. uarnak 3 and those of H. uarnak 1 and H. uarnak 2, were 122.9 and 121.3, respectively.

Himantura uarnak 4 (fig. 52)

One of the specimens, also bearing a disc with dorsal spots and/or reticulations, clustered along with but outside all of the other specimens of the H. uarnak complex. This specimen also came from Malaysian Borneo. The average of the pairwise differences between this specimen and those of H. uarnak 1 was 140.7, from those of H. uarnak 2 was 148.3, and from H. uarnak 3 was 136.3. Thus, it was given the separate numerical designation of Himantura uarnak 4. This species also grouped outside the H. undulata and H. leoparda clusters.

Himantura oxyrhyncha (longnose marbled whipray) complex (fig. 53)

In total, 42 specimens identified as this species were included in the analysis. All of these were taken from the west coast of Indonesian Borneo and thus represent the more southern portion of the distribution of this species, which has also been reported from Cambodia and Thailand. Two specimens from Kalimantan were deposited in the California Academy of Sciences (GN4534  =  CAS 229037 and GN4540  =  CAS 229038). The analysis yielded a single cluster with a range in pairwise differences among members of this cluster was 0–6, with an average of 0.6. It is interesting to note, however, that included among these specimens was one animal (GN4483) that differed conspicuously from the remaining specimens in form, denticle pattern, and color pattern. In recognition of these morphological distinctions, we have referred to this specimen as Himantura cf. oxyrhyncha.

Himantura signifer (white-edge freshwater whipray) (fig. 53)

The 21 specimens of this species included in the analysis were all collected from the Kapuas River in West Kalimantan, Indonesia, and thus represent only a portion of the distribution of this species, which has also been reported from freshwaters in Thailand. One sample has been deposited in the California Academy of Sciences (GN4552  =  CAS 229039). The analysis yielded a single cluster with a range of pairwise differences among specimens in this cluster of 0–7, with an average of pairwise differences of 1.6.

Himantura cf. kittipongi (fig. 53)

One specimen, collected from the Pawan River in West Kalimantan, Indonesia, was included in the analysis. This specimen is generally consistent in morphology with the relatively recently described but morphologically variable H. kittipongi (see Vidthayanon and Roberts, 2005). However, as H. kittipongi is known only from Thailand, this specimen has been given the designation H. cf. kittipongi until additional material from Borneo can be examined in more detail. This specimen clustered most closely with specimens of H. signifer; however, the average of the pairwise differences between this specimen and those in the H. signifer cluster was 45.3.

Himantura imbricata (scaly whipray) (fig. 53)

In total, seven specimens of this species were included in the analysis. One of these was collected from the west coast of India and the remaining six came from the Persian Gulf and Gulf of Oman. Thus, our specimens represent the western and central elements of the distribution of this species, the range of which is not well defined in the Indo-West Pacific. The analysis yielded essentially a single cluster. The range in pairwise differences among specimens in this cluster was 0–11, with an average of 5.3.

Himantura walga (dwarf whipray) (fig. 53)

The six specimens of this species included in the analysis were all collected from western Malaysian Borneo and thus represent only a small portion of the distribution of this species, which is more widespread in the Indo-West Pacific. These six specimens were found to comprise a single cluster, however, the range in pairwise differences among specimens in this cluster was relatively broad at 2–17, with an average of 8.5. These specimens clustered most closely with those of H. imbricata, but the average of the pairwise differences among specimens of these two species was 95.8.

Himantura lobistoma (tubemouth whipray) (fig. 54)

The analysis included 20 specimens of this relatively newly described (see Manjaji-Matsumoto and Last, 2006) Borneo endemic species. These specimens include the holotype (GN2972  =  SMEC 369), a paratype (GN2965  =  SMEC 370), and four vouchers (GN4609  =  CAS 229043, GN3691  =  ANFC H 6214-03, and GN3465  =  IPPS BO248, and GN4211  =  MZB 15.508). The analysis yielded a single tight cluster with a range in pairwise differences among these specimens was 0–4, with an average of 0.9.

Himantura granulata (mangrove whipray) (fig. 54)

In total, three specimens of this Indo-West Pacific species were included in the analysis, from Malaysian and Indonesian Borneo and the Gulf of Carpentaria in northern Australia. The analysis yielded a single cluster. The range of pairwise differences among these specimens was 1–3; the average of the pairwise differences was 2.

Himantura sp. 1. (fig. 54)

The analysis included a single specimen (GN2103) of a very large species of Himantura (161 cm disc width), collected from the Arafura Sea off northern Australia, the identity of which is uncertain. While this animal clustered most closely with the specimens of H. granulata, the average of the pairwise differences between this specimen and those of H. granulata was 65.3. This specimen has been referred to here as Himantura sp. 1. It should be noted that this specimen was host to five new cestodes described by Fyler et al. (2009) and was referred to by those authors as Himantura sp. The NADH2 sequence data for this animal were deposited in GenBank (No. FJ896004) by those authors as an aid to establishing the identity, if not the name, of this host species.

Himantura polylepis (giant freshwater whipray) (fig. 54)

All seven of the specimens of this species included in the analysis were collected from eastern Borneo, specifically from the Kinabatangan River in Malaysian Borneo, and the Mahakam River and the Sulawesi Sea in Indonesian Borneo. The analysis yielded a single cluster of specimens with a range of pairwise differences among specimens of 0–6, with an average of 1.7. However, it would be interesting to include specimens collected from elsewhere in the distribution of this species, which includes the Cho Phraya River basin in Thailand and the Mekong River basin in Cambodia. As noted by Last and Manjaji-Matsumoto (2008), it is likely that records from New Guinea actually refer to Himantura dalyensis, specimens of which were not included in the analysis.

Urogymnus asperrimus (porcupine ray) complex (fig. 54)

The analysis included six specimens originally identified as belonging to this relatively broadly distributed species. However, the analysis yielded two distinct clusters. One consisted of four specimens collected from northern and Western Australia that were identical in sequence. The second cluster consisted of two specimens collected from the Philippines. These specimens differed from one another by two bases. However, the average of the pairwise differences among specimens in the two clusters was 25, suggesting that more than one species may be involved. The genus currently includes only two species. Our analysis did not include representation of the second described species, Urogymnus ukpam, which is known only from rivers in western Africa. Given that our analysis also did not include representation from the type locality of U. asperrimus (i.e., India), we have referred to the Australian form as Urogymnus asperrimus 1 and the Philippines form as Urogymnus asperrimus 2 until the identity of Urogymnus from these regions has been studied in more detail. We note that both specimens from the Philippines (GN2259  =  JPAG 191 and GN4385  =  JPAG 170) were treated by Compagno et al. (2005b) as U. asperrimus. Also of interest is the fact that all specimens of Urogymnus clustered among species of Himantura.

Dasyatis margarita (daisy stingray) (fig. 55)

The 11 specimens of this species included in the analysis were identical in sequence. All were collected from the coast of Senegal and thus represent the more northern elements of the distribution of this species, which extends as far south as Angola.

Dasyatis margaritella (pearl stingray) (fig. 55)

The analysis included five specimens of this species, all of which were collected from Senegal, and all of which comprised a single cluster. The range in pairwise differences among specimens in this cluster was 0–6, with an average of 2.6. These specimens also represent only the more northern elements of the distribution of this species, which extends south to Angola. This species grouped most closely with the sympatric species D. margarita. The average of the pairwise differences between specimens of these two species was 75.4.

Taeniura grabata (round fantail stingray) (fig. 55)

The two specimens of this species included in the analysis were identical in sequence. Both were collected from Senegal, and thus represent only the northwestern elements of the distribution of this species, which may occur throughout much of the coast of Africa, and as far east as the Red Sea. These specimens clustered among Dasyatis species, most closely with D. microps. However, the average of the pairwise differences between specimens of these two species was 161.5. It should be noted that the generic placement of this species is questionable.

Dasyatis microps (thickspine giant stingray) (fig. 55)

Two specimens of this species, one from the Arafura Sea off northern Australia and one from Mozambique, were included in the analysis. These specimens represent the eastern and western margins of the Indo-West Pacific distribution of this species. The specimens clustered together, differing by only a single base.

Dasyatis zugei (pale-edged stingray) complex (fig. 56)

The analysis included 24 specimens from Borneo and 12 from Vietnam and thus represents only the central elements of the distribution of this species, which is considered to occur widely throughout the Indo-West Pacific. The analysis yielded two clusters. One consisted of the specimens from Borneo with a range of pairwise differences among specimens of 0–7, with an average of 3.2; the other consisted of the specimens from Vietnam with a range of pairwise differences among specimens of 0–5, and an average of 1.5. These clusters differed substantially from one another, with an average of the pairwise differences among specimens in the two clusters of 33.7 and likely represent two distinct taxa. Given that the type locality of D. zugei is Macau, China, the specimens from Vietnam have been designated as D. zugei. The specimens in the cluster from Borneo have been designated as D. cf. zugei until such time as this complex can be examined in more detail. Two specimens of the latter taxon have been deposited (GN4450  =  CAS 229027 and GN3437  =  IPPS BO169).

Dasyatis centroura (roughtail stingray) (fig. 56)

Two specimens of this Atlantic species were included in the analysis. These were collected from Virginia and differed from one another by one base.

Dasyatis ushiei (cow stingray) (fig. 56)

A single specimen identified as D. ushiei was included in the analysis. This species was also reported in Indonesia, as D. cf. ushiei, by White et al. (2006) and has since been validated as conspecific with specimens from Taiwan (Last et al., 2010c). This specimen clustered most closely with the specimens of D. centroura, with an average of the pairwise differences between specimens of these two species of 20.5.

Dasyatis brevicaudata (shorttail stingray) (fig. 56)

The single specimen of this species included is deposited in the Australian National Fish Collection (GN4652  =  ANFC H 6346-25). It was collected from Western Australia and thus represents only a small portion of the distribution of this southern temperate Indo-West Pacific species. This specimen grouped most closely with D. ushiei and D. centroura. The averages of the pairwise differences between D. brevicaudata and the latter two species were 49 and 58.5, respectively.

Pteroplatytrygon violacea (pelagic stingray) (fig. 56)

The specimen of this species included in the analysis was collected off California and thus represents a single locality from the global distribution of this oceanic species. It clustered most closely with a grouping consisting of D. brevicaudata, D. ushiei, and D. centroura with an average of the pairwise differences between the P. violacea and specimens in the latter grouping of 81.8. It should be noted that this species clustered among Dasyatis species. Its placement has alternated between Pteroplatytrygon and Dasyatis in the last few decades and based on our findings, this needs to be be revisited.

Dasyatis marmorata (marbled stingray) (fig. 56)

A single specimen, collected from Senegal, of this African species was included in the analysis. It clustered along with several other species of this genus. Following Eschmeyer and Fricke (2011), we have referred to this Senegalese specimen as D. marmorata.

Taeniurops meyeni (fantail stingray) (fig. 56)

Two specimens of this species were included in the analysis. These came from the Arafura Sea off northern Australia and the eastern coast of Malaysian Borneo, and thus represent only a small portion of the Indo-West Pacific distribution of this species. These specimens differed from one another by one base. It is of note that this species clustered among species of Dasyatis suggesting that its generic placement should be revisited. The average of the pairwise differences between this species and the other five with which it clustered was 108.

Dasyatis dipterura (diamond stingray) (fig. 56)

All four specimens of this species included in the analysis were collected from the Gulf of California and thus represent a more northern element of the distribution of this species, which extends coastally from southern California to Chile. These specimens differed from one another by 1–2 bases; the average of the pairwise differences was 1.5. We have followed Ebert (2003) here in referring to these specimens as D. dipterura, rather than D. brevis, because the latter is a junior synonym of the former.

Dasyatis say (bluntnose stingray) (fig. 56)

The three specimens of this species included in the analysis were all collected from the Gulf of Mexico, and thus represent only the northern elements of the distribution of this species, which extends south to Brazil. The range in pairwise differences among these specimens was 0–1. The cluster comprised of specimens of this species grouped most closely with those of D. dipterura. The average of the pairwise differences among specimens of these two species was 52.6.

Dasyatis longa (longtail stingray) (fig. 56)

Four specimens of this species, all from the Gulf of California, were included in the analysis. They were identical in sequence. However, they represent only the most northern elements of the distribution of this species, which occurs as far south as Ecuador.

Dasyatis sp. (fig. 56)

Two adults and one juvenile specimen from Senegal, which were identical in sequence, grouped close to but independently from the Pacific-dwelling D. longa with an average pairwise difference of 26. The adult of this unidentified species exhibited a median row of conspicuous thorns on the midline of the disc that extended onto the tail and a shorter row of smaller spines on either side of the median row. In this respect, this species appears to be inconsistent with species known from this region. Its identity remains uncertain and it possibly represents an undescribed species.

Dasyatis americana (southern stingray) (fig. 56)

The analysis included three specimens of this western Atlantic species, two collected from the western Atlantic Ocean and one from the Gulf of Mexico. Although these specimens grouped together, the range of pairwise differences among them was high at 15–29, with an average of 21.7. Unfortunately no specimens were retained; clearly the identity of D. americana from the western Atlantic seaboard would be interesting to explore in more detail.

Dasyatis sabina (Atlantic stingray) (fig. 56)

The two specimens of this species included in the analysis differed by 15 bases. Both were collected from the Gulf of Mexico at the center of the distribution of this species, which occurs coastally from North Carolina to Belize. They comprised a single cluster and differed substantially from all other closely grouped taxa. For example, the ranges of pairwise differences between this species and its sympatric congeners D. say and D. americana were 149 and 163.3, respectively.

Pastinachus species

The fact that this genus has recently been expanded from one or two species to include five valid species led us to generate phenotype and geography haplotype networks for the 29 specimens in the analysis that we believe constitute five distinct species. The phenotype haplotype map (fig. 96A) illustrates that, collectively, these specimens exhibit five distinct, tight (except for P. stellurostris, which is represented by only a single specimen) clusters of haplotypes that correspond to the species recognized here. The haplotype map showing the geographic origin of the specimens (fig. 96B) emphasizes that four of these species cooccur throughout the island of Borneo. Pastinachus atrus appears widely distributed, with a relatively tight clustering of haplotypes among specimens in different parts of its distribution.

Pastinachus atrus (cowtail stingray) (fig. 57)

In total, 14 specimens identified as this species were included in the analysis, which yielded a single cluster. These specimens came from eastern Malaysian and Indonesian Borneo, the Philippines, Madagascar, and Australia. The range of pairwise differences among specimens in this cluster was 0–14, with an average of 5.8. This species has long been confused with P. sephen, originally described from Saudi Arabia and the Red Sea. However, Last and Stevens (2009) resurrected P. atrus and it is now considered to have a broad Indo-Pacific distribution (Last et al., 2010c). The average of the pairwise differences between specimens of P. atrus and those of P. solocirostris was 93, P. stellurostris 101.1, P. gracilicaudus 95, and P. cf. sephen 84.1. We note that the specimen from the Philippines (GN2226  =  JPAG 167) was treated by Compagno et al. (2005b) as P. cf. sephen.

Pastinachus gracilicaudus (narrowtail stingray) (fig. 57)

The analysis yielded a cluster comprised of nine specimens of this recently described species (see Last and Manjaji-Matsumoto, 2010), collected from throughout the island of Borneo. These included three paratypes (GN4498  =  ANFC H 7107-01, GN4503  =  ANFC H 7108-01, and GN4456  =  MZB 18227). The range of pairwise differences among the specimens in this cluster was 1–8, with an average of 4.1.

Pastinachus cf. sephen (fig. 57)

Three specimens collected from the Gulf of Oman off the coast of Iran were included in the analysis; the range of pairwise differences among specimens in this cluster was 1–4, with an average of 2.7. These specimens exhibited a deep ventral fin fold consistent with that seen in P. sephen and also in P. atrus. The specimens from the Gulf of Oman grouped well away from the specimens of P. atrus, and thus are not likely conspecific with the latter species. Given that no specimens definitely identified as P. sephen were included here, this cluster has provisionally been designated as P. cf. sephen until the required morphological and molecular comparisons with P. sephen can be made. These specimens clustered most closely with P. gracilicaudus. The average of the pairwise differences between specimens in these two clusters was 86.8.

Pastinachus solocirostris (roughnose stingray) (fig. 57)

The analysis included 12 specimens identified as this relatively newly described species (see Last et al., 2005). These specimens were collected from Malaysian and Indonesian Borneo and represent the core parts of its distribution; it is also known to occur elsewhere in Indonesia. Three of these specimens are paratypes (GN3432  =  ANFC H 6123-03, GN3433  =  ANFC H 6123-02, and GN3441  =  ANFC H 6219-01) and one is a voucher deposited in the California Academy of Sciences (GN4612  =  CAS 229044). The analysis yielded a single cluster comprised of these specimens. The range of pairwise differences among 11 of these specimens was 0–8. However, one specimen from western Malaysian Borneo was particularly divergent; this specimen extended the upper boundary of the range of pairwise differences within this cluster to 18. However, this specimen did not differ conspicuously morphologically from the other specimens in the cluster. The average of pairwise differences among all 12 specimens in this cluster was 6.2.

Pastinachus stellurostris (starrynose stingray) (fig. 57)

One specimen of this recently described species (see Last et al., 2010b) was collected from western Indonesian Borneo. It clustered most closely with specimens of P. solocirostris; the average of the pairwise differences between this specimen and those of P. solocirostris was 81.2.

Neotrygon kuhlii (bluespotted maskray) complex (fig. 58)

In total, 47 specimens of maskrays with blue spots were included in the analysis. Conservatively, the resulting group includes four distinct clusters, suggesting that N. kuhlii actually represents a complex of species as reported by Last et al. (2010c). Unfortunately, our specimens do not include the localities of any of the syntype material of N. kuhlii (i.e., India, Solomon Islands, or New Guinea). As a consequence, we have referred to these clusters as N. kuhlii 1 through 4. These species are treated separately below. A taxonomic revision of this complex is being undertaken by P.L. and W.W. and a detailed molecular phylogenetic study is being undertaken by Melody Puckridge.

The phenotype and geography haplotype networks generated for Neotrygon include these four forms as well as N. picta and a specimen we have identified as N. cf. ningalooensis (see below). The haplotype network colored by phenotype (fig. 97A) shows no overlap in haplotypes among any of these six forms. Neotrygon kuhlii 1 showed the greatest amount of haplotype variation, but also was represented by the greatest number of specimens. The sympatry of three species (N. picta, N. cf. ningalooensis, and N. kuhlii 4) in Australia, and two species in Borneo (N. kuhlii 1 and N. kuhlii 2) is clearly illustrated in the geography network (fig. 97B).

Neotrygon kuhlii 1 (fig. 58)

The 33 specimens given this designation were collected from throughout Borneo and also the Philippines. In general, these specimens are typical of members of the N. kuhlii complex, bearing a dark mask and blue spots of several sizes. Six of the specimens from Borneo were vouchered (GN4475  =  CAS 229033, GN3445  =  IPPS BO218, GN3446  =  IPPS BO219, GN3684  =  IPPS BO473, GN3697  =  IPPS BO486, and GN3698  =  IPPS BO487). The range of pairwise differences among specimens comprising this cluster was 0–16, with an average of 7. While some structure was seen within this cluster, particularly with respect to four specimens collected from South and Central Kalimantan, to be conservative we have refrained from using a different designation for these specimens. We note that the two Philippine specimens (GN2216  =  BRU 029 and GN2211  =  BRU 001) were treated by Compagno et al. (2005b) as Dasyatis kuhlii.

Neotrygon kuhlii 2 (fig. 58)

This designation was employed for a cluster comprised of eight specimens collected from western Borneo, Thailand, Vietnam, and Taiwan. These animals either lacked spots entirely, or possessed only a few small blue spots. One of the specimens from Borneo was vouchered (GN3621  =  IPPS BO409). We note that the relatively large range in pairwise differences among specimens in this cluster (i.e., 0–19, with an average of 6.6) can be accounted for by the specimen from Thailand. Nonetheless, we have again taken the conservative approach of assigning only a single designation to the members of this cluster at this time. These specimens grouped most closely to the cluster consisting of specimens of N. kuhlii 1. However, the average of the pairwise differences between specimens of N. kuhlii 2 and those of N. kuhlii 1 was 20.6.

Neotrygon kuhlii 3 (fig. 58)

Also unique, was a single specimen collected from the western coast of Madagascar in the Mozambique Channel. This specimen also exhibited the color pattern typical of Neotrygon kuhlii, with a dark mask and blue spots of several sizes. The averages of the pairwise differences between this specimen and N. kuhlii 1 and N. kuhlii 2, were 29.3 and 30.4, respectively.

Neotrygon kuhlii 4 (fig. 58)

This designation was employed for the cluster comprised of the five specimens collected from northern Australia. These specimens also exhibited the color pattern typical of Neotrygon kuhlii. The range of pairwise differences among specimens in this cluster was 0–10, with an average of 5.8. The averages of the pairwise differences between specimens in this cluster and those of species Neotrygon kuhlii 1, N. kuhlii 2, and N. kuhlii 3 were 30.3, 32.4, and 36.4, respectively.

Neotrygon picta (speckled maskray) (fig. 58)

The specimens of this newly described Australian endemic species (see Last and White, 2008a) that were included in the analysis consisted of the holotype (GN4654  =  ANFC H 5771-01) and five additional specimens. The analysis yielded essentially a single cluster with a range in pairwise differences among specimens of 0–9, with an average of pairwise differences of 3.8. The average of the pairwise differences between specimens of this cluster and the 46 specimens in the Neotrygon kuhlii complex overall was 95.1.

Neotrygon cf. ningalooensis (fig. 58)

A single specimen, collected from the Gulf of Carpentaria in northern Australia, which is similar in appearance to the recently described N. ningalooensis from northwestern Australia (see Last et al., 2010d) was included in the analysis. This specimen clustered along with but outside all the other Neotrygon species. The average of the pairwise differences between this specimen and all other Neotrygon taxa included in the analysis was 111.4. Until this specimen can be compared with the types of N. ningalooensis, it has been given the designation N. cf. ningalooensis.

Taeniura lymma (bluespotted fantail ray) complex (fig. 59)

In total, 33 specimens of bluespotted fantail rays were included in the analysis. The analysis yielded a group exhibiting a substantial amount of substructure indicating that more than a single species is represented by these specimens. However, because much of this molecular variation was not associated with any immediately apparent phenotypic variation or geographic distribution, we have given separate designations to only the two most conspicuous clusters until bluespotted fantail rays can be examined in more detail. Given that the type locality of T. lymma is the Red Sea and our analysis included no specimens from that region, we have given specimens in the two clusters the numerical designations T. lymma 1 and T. lymma 2 as it is unclear which, if either, represents the original T. lymma.

The haplotype maps reflect the uncertainty associated with specific designations in this complex. While there is no haplotype overlap among specimens identified as T. lymma 1 and T. lymma 2 (fig. 98A), there is substantial haplotype variation among specimens within both of these clusters. The geography haplotype map (fig. 98B) suggests that these two species are not entirely allopatric with respect to one another for one or more specimens of both species came from Kalimantan.

Taeniura lymma 1 (fig. 59)

The 28 specimens comprising this cluster were collected from a diversity of localities throughout Borneo, Vietnam, and the Philippines. While many of the specimens from Sabah in northeastern Borneo are identical in sequence, the range in pairwise differences among all the specimens in this cluster was 0–21, with an average of 3.8 bases. The upper end of this range can be accounted for by a specimen from Sabah and also by several specimens from West Kalimantan. We note that the four specimens from the Philippines (GN2243  =  JPAG 033, GN2255  =  RSE 007, GN2257  =  RSE 008, and GN2215  =  BRU 028) were treated by Compagno et al. (2005b) as Taeniura lymma.

Taeniura lymma 2 (fig. 59)

The seven specimens comprising the second cluster of bluespotted fantail rays were collected from northern Australia, Sulawesi, and Indonesian Borneo. The range in pairwise differences among the specimens in this cluster was 0–17, with an average of 9.5. However, within this cluster, the four specimens from Australia grouped independently from the remaining specimens. The average of the pairwise differences between the Australian specimens and those from the Sulawesi and Borneo cluster was 14.7. The average of the pairwise differences between specimens of the T. lymma 1 cluster and those of the T. lymma 2 cluster was 22.8.

Urotrygonidae (round stingrays)

Urobatis concentricus (bullseye stingray) (fig. 60)

Five specimens of this species were included in the analysis, which yielded a single cluster. These specimens were collected from the Gulf of California and are thus representative of the distribution of this species. One of these specimens was deposited in the Texas Cooperative Wildlife Collection (GN2275  =  TCWC 7580.01). The range in pairwise differences among these specimens was 0–4, with an average of 1.6.

Urobatis maculatus (Cortez round stingray) (fig. 60)

All five specimens of this species included in the analysis were collected from the Gulf of California and thus are representative of the distribution of this species in general. The analysis yielded a single cluster. The range of pairwise differences among specimens in this cluster was 0–6, with an average of 2.8. This cluster grouped closely with that comprised of the specimens of U. concentricus. The average of the pairwise differences among specimens of these two species was 11.1.

Urobatis halleri (round stingray) (fig. 60)

The nine specimens included in the analysis all came from the Gulf of California and thus are representative of the more northern elements of the distribution of this species, which occurs as far south as Panama. The range in pairwise differences among these specimens was 0–8, with an average of 4. This cluster grouped outside that consisting of specimens of U. concentricus and U. maculatus. The average of the pairwise differences between specimens of this species and those of U. concentricus was 53.2 and between this species and U. maculatus was 58.8.

Urotrygon rogersi (lined round stingray) (fig. 60)

Two specimens of this species were included in the analysis, both collected from the Gulf of California and thus represent the northern portion of the distribution of this species, which occurs as far south as Ecuador. One of these specimens was deposited in the Texas Cooperative Wildlife Collection (GN5250  =  TCWC 7567.04). The specimens differed by five bases.

Urotrygon cf. simulatrix (fig. 60)

One specimen of this species was included in the analysis. The specimen clustered along with, but outside the two specimens of U. rogersi. The average of the pairwise differences between specimens of these two species was 185.5.

Urobatis jamaicensis (yellow stingray) (fig. 60)

Both specimens of this species included in the analysis were collected from the Cayman Islands and thus come from essentially the center of the distribution of this species, which occurs throughout much of the subtropical and tropical western seaboard of the Atlantic Ocean including the Caribbean Sea and Bahamas. These specimens were identical in sequence. They clustered most closely with the specimens of U. rogersi and U. cf. simulatrix; the average of the pairwise differences between specimens of U. jamaicensis and the latter two species was 180.5 and 184, respectively.

Potamotrygonidae (river stingrays)

Potamotrygon species (fig. 60)

Much confusion surrounds the identities of the potamotrygonid stingrays from the rivers of South America. In total, the analysis included nine specimens of the South American freshwater stingray genus Potamotrygon, five of which were collected from the Madre de Dios River in Peru. The results of the analysis suggest that these nine specimens represent perhaps as many as six distinct lineages. However, the molecular results are not necessarily congruent with the color patterns of the specimens involved. Given this, in combination with the fact that only two of the nine specimens have been deposited in a museum (GN5881  =  MZUSP 95411 and GN5880  =  MZUSP 107670), and in light of the concerns raised by Toffoli et al. (2008), we have only limited confidence in the identifications of the specimens included here.

The analysis yielded a group comprised of all nine specimens, with a cluster comprised of five specimens collected from the Madre de Dios River in Peru. Within the Madre de Dios cluster, the genetic identities of three specimens were much closer to one another than to either of the other two specimens in the cluster. These three specimens had a range of pairwise differences of 2–3 (with an average of 2.7). However, the color patterns of these five specimens are inconsistent with their genetic identities. Given the issues raised by Toffoli et al. (2008), the identities of these specimens have been assigned here are based in large part on color pattern. We are confident that GN5880 is P. tatianae because it was among the voucher specimens examined by da Silva and de Carvalho (2011) in their description of P. tatianae, which is known only from the Madre de Dios River. Two specimens have been identified as P. cf. tatianae because, although they are consistent with the color pattern of that species, they differ from it by 26.5 bases. The color pattern of the remaining two specimens resembles that of P. motoro. However, in the absence of definitively identified specimens of P. motoro, in combination with the fact that these specimens differ from one another by 37 bases, we have given these specimens the designations Potamotrygon cf. motoro 1 and P. cf. motoro 2. The remaining four specimens were purchased from pet stores and, as a consequence, are of unknown provenance and their identities are uncertain. These have been given the designation Potamotrygon sp. 1 and Potamotrygon sp. here, but it is important to note that these specimens may represent described species that were not included in the analysis. The two specimens of Potamotrygon sp. 1 differed from one another by one base and the two specimens of Potamotrygon sp. 2 differed from one another by one base. The average of the pairwise differences between specimens of Potamotrygon sp. 1 and those of Potamotrygon sp. 2 was 104. Clearly, much work remains to be done to resolve the identities of these specimens relative to confirmed material.

Paratrygon aiereba (discus ray) (fig. 60)

A single specimen of this currently monotypic genus, also collected from Peru, was included in the analysis. This grouped along with but outside the Potamotrygon specimens. This specimen is deposited in the Zoology Museum in Sao Paulo, Brazil (GN5874  =  MZUSP 95406).

Himantura schmardae (Chupare stingray) (fig. 60)

A single specimen of this species, collected from Guyana, was included in the analysis. This specimen clustered along with, but outside those of Potamotrygon and Paratrygon. This specimen was deposited at the Royal Ontario Museum (GN6488  =  ROM 66845). The grouping of this species with the Potamotrygonidae rather than with the Dasyatidae warrants further taxonomic investigation.

Rhinopteridae (cownose rays)

Rhinoptera species

In total, 52 specimens of cownose rays were included in the analysis. The identities of most of these specimens were difficult to determine using either morphological or molecular criteria, because the taxonomy of this genus remains so relatively poorly known. Even key characters such as tooth shape and number appear to be intraspecifically variable. Hence, in assigning names to taxa we have taken a conservative approach. Morphology, molecular differences, and geographic location were used in combination to make these determinations. In general, only in instances in which at least two of these three criteria were congruent, were the specimens comprising a cluster assigned a unique name. In total we believe our specimens represent as many as eight species of Rhinoptera.

The phenotype and geography haplotype maps for the 52 specimens of Rhinoptera included here are illustrated in figure 99A and 99B, respectively. The phenotype map supports the validity of at least seven of these species in that their haplotypes are distinct and, in the cases of species represented by multiple specimens, the haplotypes of their members are tightly clustered. However, it also serves to highlight issues with R. steindachneri and R. bonasus, which are addressed in the sections treating these species below.

Rhinoptera steindachneri (Pacific cownose ray) complex (fig. 61)

The analysis yielded a cluster comprised of seven specimens of Rhinoptera steindachneri from the Gulf of California, seven specimens from the Gulf of Mexico, and one from North Carolina all originally identified as Rhinoptera bonasus. The range of pairwise differences among the 15 members of this cluster was 0–9 (with an average of 3.8). However, these specimens differ from one another morphologically in a number of respects. For example, the specimens from the Gulf of California appear to be fully consistent with the eastern Pacific species, Rhinoptera steindachneri, the type locality of which is in the Gulf of California. In contrast, the configuration of the tooth plates of the specimens from the Gulf of Mexico and North Carolina resemble those of Rhinoptera brasiliensis, even more than they do those of R. bonasus. Furthermore, these specimens clustered well outside specimens from the western Atlantic considered to be morphologically consistent with R. bonasus (see below). Unfortunately, no confirmed specimens of the southwestern Atlantic R. brasiliensis were included in the analysis. Until such time as the New World members of this genus can be examined in more detail, despite the mixed nature of this cluster, we have retained the designation Rhinoptera steindachneri for the specimens from the Gulf of California and have adopted Rhinoptera cf. steindachneri for the members of this cluster that occur in the other localities.

However, the haplotype map colored by phenotype (fig. 99A) does not support this line of reasoning. Not only is one haplotype shared by specimens identified as R. steindachneri and R. cf. steindachneri, but the haplotypes of specimens of these two taxa are very similar and, in fact, comprise a single group. This suggests the alternative interpretation that R. steindachneri occurs not only in the eastern Pacific, but also in the Gulf of Mexico, and possibly also the western Atlantic Ocean, where it is sympatric with Rhinoptera bonasus.

Rhinoptera sp. 1 (fig. 61)

Clustering along with but outside the specimens in the Rhinoptera steindachneri complex was a specimen collected from Senegal (GN5978). The average of the pairwise differences between this specimen and those in the former cluster of 15 specimens was 20.8. Given its degree of genetic divergence and disparate locality, this specimen has been given a separate designation until such time as additional specimens from the eastern Atlantic can be examined in more detail.

Rhinoptera jayakari (Oman cownose ray) (fig. 61)

The analysis also yielded a cluster comprised of 10 specimens collected from a diversity of localities including Borneo, Mozambique, and the Gulf of Oman. The range of pairwise differences among specimens in this cluster was 0–17, with an average of 8.3. While there was some evidence of two subclusters within this cluster, the specimens in these two subclusters overlapped in geography and had an average of the mean pairwise differences of 13.7. At this time, all specimens in this cluster have been referred to as Rhinoptera jayakari following Last et al. (2010c).

Rhinoptera bonasus (cownosed ray) complex (fig. 61)

The analysis yielded a cluster comprised of 18 specimens from a diversity of localities in the western Atlantic including Virginia, North Carolina, and South Carolina. The range in pairwise differences among members of this cluster was 0–8, with an average of 1. The fact that these specimens grouped away from those comprising the R. cf. steindachneri cluster, which also included a specimen from the western Atlantic, as noted above, suggests that this region may be home to two sympatric species of Rhinoptera. These specimens have been given the designation R. bonasus, based on their morphology and because the type locality of this species is New York. The average of the pairwise differences between R. bonasus and its sympatric congener R. cf. steindachneri was 86.

Grouping along with but outside the members of the former cluster was a second specimen collected from Senegal. The average of the pairwise differences between this specimen and those of the R. bonasus cluster was 19.4. Thus, while this specimen from Senegal exhibited a tooth pattern similar to that of R. bonasus, given its geography and genetic difference, it has been referred to here as R. cf. bonasus. The difference between this specimen and that of the other from Senegal (Rhinoptera sp. 1) was 80. This suggests that Senegal may also be home to two sympatric species of Rhinoptera.

The haplotype map (fig. 99A) supports the contention that R. cf. bonasus from Senegal is distinct from R. bonasus. However, it also confirms that a second species (Rhinoptera sp. 1) occurs sympatrically with R. cf. bonasus in Senegal.

Rhinoptera neglecta (Australian cownose ray) (fig. 61)

The analysis also yielded a cluster comprised of five specimens of R. neglecta from northern Australia. One of these specimens came from the Australian National Fish Collection (GN4662  =  ANFC H 3915-01). The range of pairwise differences among these specimens was 0–3, with an average of 1.2.

Rhinoptera javanica (Javanese cownose ray) (fig. 61)

Two specimens identified as R. javanica (see Last et al., 2010c), one from eastern Malaysian Borneo and one from Vietnam, grouped immediately outside the R. neglecta cluster. These specimens differed from one another by nine bases. The average of the pairwise differences between these specimens and those of R. neglecta was 22.6.

Mobulidae (devil rays)

Mobula japanica (spinetail devilray) (fig. 61)

In total, 12 specimens of this species were included in the analysis. Given that 10 of these specimens were collected from the Gulf of California and two from Vietnam, they represent two disparate elements of the distribution of this widely distributed species. One of the specimens was deposited in the Texas Cooperative Wildlife Collection (GN5273  =  TCWC 7568.01). The analysis yielded a single cluster; the range of pairwise differences among specimens in this cluster was 0–5, with an average of 2.3.

Manta birostris (manta) (fig. 61)

The two specimens of this species included in the analysis were collected from Indonesia and the Philippines and thus represent only a very small portion of the distribution of this global species. The sequences of these specimens differed by seven bases. It is interesting that this species clustered among Mobula species; most closely with Mo. japanica. The average of the pairwise differences among specimens of Ma. birostris and Mo. japanica was 125.8. The recent resurrection of Manta alfredi by Marshall et al. (2009) has caused doubt over the identifications of manta species in published literature and other records. The specimen from Indonesia can be accurately confirmed as M. birostris based on photographic evidence. There is a possibility that the specimen from the Philippines may represent M. alfredi; acquisition of more specimens confirmed as M. alfredi in the future will be required to confirm this. The Philippine specimen is deposited in the Philippines (GN4356  =  BRU 043). At this stage we retain M. birostris as the designation for both specimens.

Mobula thurstoni (smoothtail devilray) (fig. 61)

The analysis included eight specimens of this relatively widely distributed species, all collected from the Gulf of California and thus representing only a very small portion of its distribution. One of these specimens was deposited in the Texas Cooperative Wildlife Collection (GN5284  =  TCWC 7565.01). The range in pairwise differences among these specimens was 0–7, with an average of pairwise differences of 2.5.

Mobula kuhlii (shortfin devilray) (fig. 61)

Three specimens, two from the Philippines and one from eastern Malaysian Borneo, were included in the analysis. The specimens from the Philippines (GN4337  =  BRU 031 and GN4327  =  JPAG 303) were treated by Compagno et al. (2005b) as representing a new record of M. kuhlii for that region. The range of pairwise differences among all three specimens was 0–1. The analysis yielded a single cluster that grouped most closely with the specimens of M. thurstoni. The average of the pairwise differences between specimens of these two species was 46.8.

Mobula munkiana (pygmy devilray) (fig. 61)

Eight specimens of this species, all collected from the Gulf of California, were included in the analysis. These specimens represent the northern part of the eastern Pacific distribution of this species, which has been reported as far south as Ecuador. One of these specimens was deposited in the Texas Cooperative Wildlife Collection (GN2286  =  TCWC 7589.03). The range of pairwise differences among these specimens was 0–3, with an average of 0.8.

Mobula hypostoma (Atlantic devilray) (fig. 61)

Both specimens of this species included in the analysis were collected from the Gulf of Mexico and thus represent one of the more northern parts of the distribution of this species, which occurs from New Jersey to Argentina. These specimens differed by four bases and they clustered most closely with M. munkiana, with an average of the pairwise differences among specimens between these two species of 31.4.

Myliobatidae (eagle rays)

Pteromylaeus bovinus (duckbill ray) (fig. 61)

A single specimen of this species was included in the analysis. This specimen was collected from Senegal and thus represents a northeastern element of the distribution of this species, which occurs from Portugal, throughout the western coast of Africa, around the cape and as far north as Mozambique. It clustered away from other myliobatids and along with, but outside all of the species of Rhinoptera and Mobula.

Myliobatis californica (bat ray) (fig. 62)

The 18 specimens of this species were collected from throughout the Gulf of California and thus represent the southern parts of the distribution of this species, which has been reported to occur from Oregon to Baja, with a rare report from the Galapagos Islands (Grove and Lavenberg, 1997). One of these specimens was deposited in the Texas Cooperative Wildlife Collection (GN5203  =  TCWC 7564.03). The analysis yielded a single cluster with a range in pairwise differences among the 18 specimens of 0–12, with an average of 5.6.

Myliobatis aquila (common eagle ray) (fig. 62)

The seven specimens of this eastern Atlantic-dwelling species were all collected from South Africa. The range of pairwise differences among specimens in this cluster was 0–2, with an average of 0.7.

Myliobatis tobijei (kite ray) (fig. 62)

The analysis included three specimens of this species, which occurs from Japan to Indonesia, one of which came from Taiwan and two from the Philippines. The specimens from the Philippines (GN4384  =  JPAG 130 and GN4357  =  JPAG 147) were treated by Compagno et al. (2005b) as M. cf. tobijei. However, given that the range in pairwise differences among specimens from both localities was only 1–7, with an average of 4.7, we have used the designation M. tobijei for all three specimens in this cluster. These specimens grouped most closely with those of M. aquila. The average of the pairwise differences among specimens in these two clusters was 48.9.

Myliobatis longirostris (longnose eagle ray) (fig. 62)

Sixteen specimens of this Gulf of California endemic were included in the analysis, which yielded a single cluster. Five of these specimens were deposited in the Texas Cooperative Wildlife Collection (GN5200  =  TCWC 7564.04, GN5201  =  TCWC 7564.06, GN5241  =  TCWC 7589.04, GN5269  =  TCWC 7587.01, and GN1570  =  IBUNAM PE9517). Although many of the specimens were identical in sequence, the range in pairwise differences among all specimens in this cluster was 0–11, with an average of 2.5.

Myliobatis freminvillei (bullnose ray) (fig. 62)

The two specimens identified as M. freminvillei included in the analysis differed by a single base. They were both collected from the western North Atlantic and thus represent more northern elements of the amphitropical distribution of this species, which has been reported as far south as Argentina. This species grouped most closely with M. longirostris; the average of the pairwise differences among specimens of these two species was 52.8.

Myliobatis australis (southern eagle ray) (fig. 62)

The analysis included four specimens of this species, all collected from the Tasman Sea in Australia. They thus represent a more eastern element of the distribution of this species, which has been reported from southern Australia and possibly New Zealand (Last and Stevens, 2009). The range in pairwise differences among these specimens was 0–4, with an average of pairwise difference of two bases.

Aetomylaeus species

The analysis included specimens of six nominal species of Aetomylaeus. A haplotype map for phenotype (fig. 100A) and also one for geography (fig. 100B) was generated for the five most similar of these species (i.e., all but A. vespertilio). The phenotype map supports recognition of three distinct species in the A. nichofii complex because it shows no overlap in haplotypes among specimens of the different species and also illustrates the substantial amount of divergence between the sympatric A. nichofii and A. maculatus. However, additional specimens from the Persian Gulf would help to confirm the distinction between A. cf. nichofii 1 and A. nichofii. The haplotype map for geography illustrates the allopatric nature of these three species.

Aetomylaeus nichofii (banded eagle ray) complex (fig. 62)

Eighteen specimens that were generally consistent with the color morph of Aetomylaeus nichofii were included in the analysis. The majority of these came from a diversity of localities around the island of Borneo, but they also included one specimen from the Persian Gulf and two from the Arafura Sea off northern Australia. The analysis yielded a group consisting of these 18 specimens. However, the 15 specimens from Borneo represented a cluster within this group; five of these specimens were vouchered (GN4264  =  CAS 229046, GN2968  =  ANFC H 6209-01, GN2969  =  ANFC H 6209-02, GN3444  =  IPPS BO180, and GN3696  =  IPPS BO485). The range of pairwise differences among the 15 specimens in the Borneo cluster was 0–10, with an average of 3.3. The Persian Gulf and Arafura Sea specimens grouped along with but outside the Borneo specimens. The two specimens from the Arafura Sea differed by one base. The average of the pairwise difference between the specimens from Borneo and the specimen from the Persian Gulf was 17.7, and between the specimens from Borneo and those from the Arafura Sea was 74. Furthermore, pairwise difference between the specimens from the Persian Gulf and Arafura Sea was 72. These results suggest that our sample may include three distinct taxa. Until this issue can be addressed in more detail, we have provisionally referred to the specimens in the Borneo cluster as Aetomylaeus nichofii as the type locality for this species is Indonesia. The specimen from the Persian Gulf has been referred to as Aetomylaeus cf. nichofii 1, and the specimen from the Arafura Sea referred to as Aetomylaeus cf. nichofii 2. A taxonomic revision of this complex is currently being undertaken by P.L. and W.W.

Aetomylaeus maculatus (mottled eagle ray) (fig. 62)

In total, 10 specimens consistent with the color pattern of A. maculatus were included in the analysis. The analysis yielded a single cluster, but with evidence of two subclusters; one comprised primarily of specimens from Malaysian Borneo and one of specimens from Indonesian Borneo. Three of the samples from Borneo come from museum specimens (GN2993  =  ANFC H 6220-01, GN3423  =  ANFC H 6123-04 and H 6122-02, and GN3442  =  ANFC H 6219-02). The range in pairwise differences among all 10 specimens in the cluster was 0–17, with an average of 8.3. The range of pairwise differences among specimens within the Malaysian subcluster was 0–5; the three specimens in the Indonesian subcluster were identical in sequence. The average of the pairwise differences among specimens in the two subclusters was 15.2. However, in the absence of consistent morphological or geographic differences, we have made no formal distinction between the two subclusters. It should be noted that our specimens represent only a small portion of the distribution of this species, which has been reported from throughout much of the Indo-West Pacific.

Aetomylaeus milvus (ocellate eagle ray) (fig. 62)

Two specimens collected from the Persian Gulf were included in the analysis. These specimens were identical in sequence to one another. They grouped most closely with specimens of Aetomylaeus maculatus. However, the average of the pairwise differences between specimens of these two species was 95.2. This result supports the validity of this species, which has been questioned by some previous authors (e.g., Compagno, 1999).

Aetomylaeus vespertilio (ornate eagle ray) (fig. 62)

The five specimens of this species included here were collected from the Arafura Sea and Gulf of Carpentaria, off northern Australia, and from the Philippines. These specimens represent eastern elements of the distribution of this Indo-West Pacific species, which has been reported from as far west as the Mozambique Channel. The Philippine specimen (GN4344  =  JPAG 324) was treated as A. vespertilio by Compagno et al. (2005b). The analysis yielded a single cluster of these specimens, which had a range of pairwise differences of 0–5, with an average of pairwise differences of 2.8 bases.

Aetobatus species

Until recently, a great deal of confusion has existed with respect to the identity of the various color morphs of “spotted eagle rays.” Several authors (e.g., Last and Stevens, 2009; Richards et al., 2009) have noted that the genus may include more species than currently recognized. The taxonomy of the spotted eagle ray group was recently partially revised by White et al. (2010c), who resurrected a number of existing names for the various color morphs that appear to represent distinct valid taxa. The treatment of this genus here follows the taxonomy proposed by White et al. (2010c). The haplotype maps for our Aetobatus specimens provide support for the recognition of all seven of the nominal species treated below. Specimens of these species share no haplotypes (fig. 101A). Furthermore, the majority of these species exhibit relatively restricted and allopatric geographic distributions (fig. 101B).

Aetobatus ocellatus (whitespotted eagle ray) complex (fig. 63)

The analysis yielded a cluster comprised of 34 specimens from throughout the Indo-West Pacific (i.e., Malaysian and Indonesian Borneo as well as from Taiwan, Thailand, Vietnam, Singapore, the Philippines, and northern Australia). The range of pairwise differences among specimens in this cluster was 0–16, with an average of 5.8. These specimens are morphologically consistent with previous accounts of A. ocellatus (e.g., see White et al., 2010c), the type locality of which is Java in Indonesia. One of the specimens from the Philippines (GN4364  =  JPAG 314) was treated by Compagno et al. (2005b) as Aetobatus cf. narinari. Two of the specimens from Borneo were also deposited (GN3550  =  IPMB 38.01.08 and GN3513  =  IPPS BO296).

However, three additional specimens, one from the Mozambique Channel and two from off Qatar in the Persian Gulf, clustered along with but outside these 34 specimens. Specimens from the Mozambique Channel and Qatar are genetically different from one another, and also from those from the Indo-West Pacific. The average of the pairwise differences between the specimen from the Mozambique Channel (A. cf. ocellatus 1) and those from the Indo-West Pacific (A. ocellatus) was 20.2; the average of pairwise differences between the specimen from the Mozambique Channel and the two from Qatar (A. cf. ocellatus 2) was 29. The average of the pairwise differences between the specimens from Qatar (A. cf. ocellatus 2) and the Indo-West Pacific (A. ocellatus) was 26.8. Unfortunately, photographs are unavailable for any of the three specimens from the Mozambique Channel and Qatar. Nonetheless, they have been referred to here as Aetobatus cf. ocellatus 1 and Aetobatus cf. ocellatus 2, respectively, based on their genetic differences. The two specimens of A. cf. ocellatus 2 differed by two bases.

Aetobatus narinari (spotted eagle ray) (fig. 63)

Fourteen specimens identified as this species were included in the analysis. These specimens came from the Florida Keys and Gulf of Mexico, as well as from Puerto Rico. Three of the specimens from the Florida Keys came from vouchers (GN5675 and GN5676  =  AMNH 251703 and GN5678  =  AMNH 251704). The analysis yielded essentially a single cluster with a range in pairwise differences of 0–18, and an average of 4.4. However, one of the specimens from Puerto Rico (GN2118) differed conspicuously from the other 13 specimens in the cluster. The average of the pairwise differences between this specimen and the other 13 was 14.6. The specimens in this cluster are morphologically generally consistent with A. narinari, the type locality of which is St. Barthélemy in the Caribbean Sea and Brazil. We have included the more divergent Puerto Rican specimen under this name, but note that it differs, for example, from the other Puerto Rican specimen (GN2119), in that many more of its white spots are ocellate, rather than solid.

Aetobatus laticeps (Pacific whitespotted eagle ray) (fig. 63)

The analysis also yielded a cluster comprised of four specimens collected from Loreto in the Gulf of California. These specimens are generally consistent with Aetobatus laticeps, described from California by Gill (1867). The range of pairwise differences among these specimens was 0–2, with an average of 1. These specimens clustered most closely with those of A. narinari; the average of the pairwise differences between specimens of these two species was 14.9.

Aetobatus sp. (fig. 63)

Nine specimens from Vietnam, that differed conspicuously morphologically from the five other species of eagle rays included here, were also found to comprise a cluster that was genetically distinct from the five other species. The range of pairwise differences among specimens in this cluster was 0–2, with an average of 0.4. Unlike other species in the genus, these specimens lacked white spots from the dorsal surface and exhibited a relatively short head. These specimens likely represent an undescribed species. Two of these specimens were deposited in the Vietnam Natural Museum of Nature (GN7014  =  VN-z-v.000394 and GN7050  =  VN-z-v.000309).

Aetobatus flagellum (longheaded eagle ray) (fig. 63)

The two specimens of this species included in the analysis were collected from Indonesian Borneo and Maharastra, India, and they differed from one another by 11 bases. These specimens are consistent with A. flagellum, whose type locality is the coast of Coramandel off India. They clustered most closely with Aetobatus sp. from Vietnam. The average of the pairwise differences between A. flagellum and those of Aetobatus sp. was 93.7.

Gymnuridae (butterfly rays)

Gymnura species

The analysis included a total of 50 specimens of Gymnura. This was among the most problematic genera with respect to specific identifications of specimens. We have treated these specimens as representing nine distinct species. The haplotype map for phenotype (fig. 102A) supports this level of diversity, in that there are no overlapping haplotypes among the specimens of any of these nine species. The haplotype map for geography (fig. 102B) provides support for the notion that two species of Gymnura (G. crebripunctata and G. marmorata) cooccur in the Gulf of California, two species cooccur in the Indo-Pacific (G. zonura and G. cf. poecilura 1), and that potentially three species cooccur in northern Australia (G. australis, G. micrura, and Gymnura sp. 1). Furthermore, our analyses are consistent with the suggestion that butterfly rays in the Persian Gulf are distinct from those in the other localities sampled here. However, clearly this genus would benefit greatly from additional taxonomic work.

Gymnura cf. poecilura complex (fig. 64)

The analysis, which included 17 specimens showing a resemblance to Gymnura poecilura, yielded two clusters. Given that no specimens from India, the type locality of this species, were included, we have given the clusters numerical designations. One, designated Gymnura cf. poecilura 1, contained 12 specimens collected from throughout Borneo; two of these specimens were vouchered (GN3641  =  IPPS BO430 and GN4507  =  CAS 229034. The range of pairwise differences among specimens in this cluster was 0–13, with an average of 5.3. The second cluster, designated Gymnura cf. poecilura 2, consisted of the other five specimens, all of which were collected from the Gulf of Oman. Specimens comprising the latter cluster had a range of pairwise differences of 1–8, with an average of 4.8. The average of the pairwise differences between specimens the two clusters was 120.5. It is likely that at least one of these forms is an undescribed species.

Gymnura zonura (zonetail butterfly ray) (fig. 64)

Seven specimens identified as this species were included in the analysis. These came from Borneo, Singapore, and the Philippines and thus represent eastern elements of the Indo-West Pacific distribution of this species. The Philippine specimen (GN4369  =  BRU 105) was treated by Compagno et al. (2005b) as Aetoplatea zonura. These specimens were generally consistent with G. zonura; however, they varied in dorsal color pattern from the typical patterned form of G. zonura (GN4838) to plain (GN4830). The analysis yielded a single cluster comprised of these specimens. The range in pairwise differences among specimens in this cluster was 1–14, with an average of 6; the upper end of this range was accounted for by a specimen from the Philippines. Specimens in this cluster grouped among other Gymnura species, supporting the placement of this species in Gymnura (see White et al., 2006; Last et al., 2010c) rather than in the separate genus Aetoplatea. It is closest to Gymnura sp. 1; the average of the pairwise differences between specimens of G. zonura and Gymnura sp. 1 was 129.2; between those of G. cf. poecilura 1 was 159.5, and between those of G. cf. poecilura 2 was 143.6.

Gymnura sp. 1 (fig. 64)

The analysis included three specimens from the western North Atlantic that had preliminarily been identified as Gymnura altavela, but that grouped well outside specimens identified as G. altavela from the eastern Atlantic. Given that the type locality of G. altavela is in the eastern Atlantic, specimens comprising that cluster (see below) have been provisionally identified as G. altavela, and those comprising the genetically divergent western Atlantic cluster, as Gymnura sp. 1. Unfortunately, images of material from the western Atlantic are not available. In the interim, these have been identified as Gymnura sp. 1, but it should be recognized that they may refer to a Gymnura species not represented in this analysis. These specimens differed from one another by 0–1 bases. They clustered along with G. zonura. The average of the pairwise differences between Gymnura sp. 1 and those of G. zonura is 129.2, between Gymnura sp. 1 and G. cf. poecilura 1 150.3, and between the G. cf. poecilura 2 cluster 141.3.

Gymnura australis (Australian butterfly ray) (fig. 64)

A single specimen of this, the only species of its genus reported from Australia, was included in the analysis. This specimen was collected from the Arafura Sea off northern Australia. It represents a northwestern element of the distribution of this species. It clustered most closely with G. zonura and Gymnura sp. 1, with average pairwise differences of 158.4 and 131.3, respectively.

Gymnura crebripunctata (longsnout butterfly ray) (fig. 64)

In total, seven specimens of this species were included in the analysis. These comprised a single cluster, and exhibited a range of pairwise differences of 1–10, with an average of 5. These specimens, which were collected in the Gulf of California, are representative of the northern portion of the distribution of this species, which is known to occur coastally as far south as Panama.

Gymnura marmorata (California butterfly ray) (fig. 64)

The four specimens of this species included here comprised a single cluster and exhibited a range of pairwise differences of 1–3, and an average of pairwise differences of 1.5. These specimens, collected in the Gulf of California, are representative of the northern elements of the distribution of this species, which has been reported as far south as Peru. The average of the pairwise differences between specimens of this species, and those of their sympatric congener, G. crebripunctata, was 180.3. This is consistent with the work of Smith et al. (2009) in supporting the recognition of G. marmorata as a taxon distinct from G. crebripunctata.

Gymnura micrura (smooth butterfly ray) (fig. 64)

One specimen identified as this species, collected from the western Atlantic, was included in the analysis. Unfortunately, images of this specimen are not available, and thus the identity of this specimen has not been confirmed. However, field notes (K. Cleason, personal commun.) indicate that this specimen lacked a tail spine. This specimen clustered most closely with the specimens of G. marmorata, but the average of the pairwise differences between specimens of these two species was 160.8. This specimen represents only a portion of the western Atlantic distribution of this species, which occurs throughout much of the central western and eastern seaboards of the Atlantic Ocean.

Gymnura altavela (spiny butterfly ray) (fig. 64)

Ten specimens that are generally morphologically consistent with this species were included in the analysis. All 10 specimens came from Senegal and thus represent only a small portion of the distribution of this species, which has been reported from throughout much of the central western and eastern seaboards of the Atlantic Ocean. The range of pairwise differences among these specimens was 0–5, with an average of 1.9 bases. Specimens in this cluster and those of their sympatric congener Gymnura sp. 1 had an average pairwise difference of 196.5.

Plesiobatidae (giant stingarees)

Plesiobatis daviesi (giant stingaree) (fig. 64)

The three specimens of this species included in the analysis were collected from Malaysian Borneo and the Philippines and thus represent only a central portion of the distribution of this species, which has been reported from southeastern Africa to Hawaii. The specimen from the Philippines was vouchered (GN4346  =  MMLM 017). The sequences of all three specimens were identical.

Urolophidae (stingarees)

Urolophus paucimaculatus (sparsely spotted stingaree) (fig. 65)

Four specimens of this southern Australian endemic were included in the analysis. The analysis yielded a single tight cluster; the range in pairwise differences among these specimens was 1–5, with an average of pairwise differences of three bases.

Urolophus cruciatus (crossback stingaree) (fig. 65)

The analysis included three specimens identified as this distinctively colored southeastern Australian endemic species. The range of pairwise differences among these specimens was 1–2, with an average of 1.3. These specimens grouped most closely with those of U. paucimaculatus; the average of the pairwise differences between these two species was 76.5.

Urolophus westraliensis (brown stingaree) (fig. 65)

A single specimen of this Australian endemic was included. It is deposited in the Australian National Fish Collection (GN4637  =  ANFC H 4649-23).

Urolophus expansus (wide stingaree) (fig. 65)

The single specimen of this southwestern Australian endemic species included in the analysis is deposited in the Australian National Fish Collection (GN4656  =  ANFC H 6414-12).

Urolophus viridis (greenback stingaree) (fig. 65)

A single specimen of this species, a southeastern Australian endemic, was included in the analysis. This specimen is deposited in the Australian National Fish Collection (GN4661  =  ANFC H 2444-04). This species clustered most closely with the specimen of U. expansus, but the sequences of these specimens differed by 44 bases.

Urolophus kapalensis (Kapala stingaree) (fig. 65)

A single specimen of this relatively newly described Australian endemic (see Yearsley and Last, 2006) was also included. It was taken from a specimen (GN4658  =  H6153-02) listed among “other material examined” in the original description of this species.

Urolophus lobatus (lobed stingaree) (fig. 65)

The analysis included one specimen of this Australian endemic, which is in the Australian National Fish Collection (GN4659  =  ANFC H 6346-15).

Urolophus flavomosaicus (patchwork stingaree) (fig. 65)

Two specimens of this Australian endemic, both in the Australian National Fish Collection (GN4636  =  ANFC H 1036-6 and GN4657  =  ANFC H 1036-37) were included. These specimens were identical in sequence.

Urolophus bucculentus (sandyback stingaree) (fig. 65)

The analysis included a single specimen of this southeastern Australian endemic. This specimen is deposited in the Australian National Fish Collection (GN4655  =  ANFC H 1269-01). It clustered most closely with U. flavomosaicus. Although these two species were considered to bear a strong morphological resemblance to one another and to be sympatric in Southern Queensland by Last and Stevens (1994), the distinction between these two species both morphologically and geographically is made clear by Last and Stevens (2009). The average of the pairwise differences between the two species was 15.

Trygonoptera imitata (eastern shovelnose stingaree) (fig. 65)

The four specimens of this newly described southeastern Australian endemic species (see Yearsley et al., 2008) included in the analysis were collected from New South Wales. The range in pairwise differences among these specimens was 1–3, with an average of 1.7. The analysis yielded a single tight cluster comprised of these specimens.

Trygonoptera testacea (common stingaree) (fig. 65)

A single specimen of this eastern Australian endemic species was included in the analysis. This specimen grouped most closely with the cluster of T. imitata specimens. The average of the pairwise differences between this specimen and those of T. imitata was 141.

Trygonoptera personata (masked stingaree) (fig. 65)

The analysis included one specimen of this western Australian endemic; this specimen is in the Australian National Fish Collection (GN4635  =  ANFC H 6347-20). This species clustered along with but outside two of its three congeners included in the analysis. The average of the pairwise differences between this species and T. imitata was 131.5, and between T. testacea was 162.

Trygonoptera ovalis (striped stingaree) (fig. 65)

A single specimen of this western Australian endemic was included in the analysis. This specimen is in the Australian National Fish Collection (GN4634  =  ANFC H 6347-13). It grouped most closely with but outside its three congeners included in the analysis. The average of the pairwise differences between this species and T. imitata was 155, T. testacea 170, and T. personata 134.

Hexatrygonidae (sixgill stingrays)

Hexatrygon bickelii (sixgill stingray) (fig. 66)

Two specimens of this unusual ray were included in the analysis. Both specimens came from Taiwan. They differed from one another by one base. These specimens represent only a small portion of the distribution of this species, which is patchily distributed from southern Africa to Hawaii.

Zanobatidae (panrays)

Zanobatis schoenleinii (striped panray) (fig. 67)

Eight specimens of this species, all collected from Senegal, were included in the analysis. These specimens thus represent a more northern element of the distribution of this species, which extends from Morocco to the Gulf of Guinea. The range of pairwise differences among specimens of this species was 0–10, with an average of 3.6.

Pristidae (modern sawfishes): genus Pristis

Pristis pectinata (smalltooth sawfish) (fig. 68)

In total, 20 specimens of this species were included in the analysis. These were collected from the Gulf of Mexico off Florida and the Bahama Islands in the Caribbean Sea. The analysis yielded a single cluster and the range of pairwise differences among these specimens was 0–12, with an average of pairwise differences of 2.2. Given the reported breadth of the distribution of this species, these specimens represent only the northwestern elements of its distribution.

Pristis zijsron (green sawfish) (fig. 68)

All six specimens of this species included in the analysis were collected from northern Australia and thus represent only a small part of the original Indo-West Pacific distribution of this species, which has been reported from fresh, brackish, and marine habitats. Two of these specimens are vouchered (GN3155  =  NTM S.14689-001 and GN3159  =  NTM S.14689-003). The range of pairwise differences among these specimens was 0–5, with an average of pairwise differences of 1.7. The analysis yielded a single cluster that grouped most closely with the cluster of P. pectinata specimens. The average of the pairwise differences between specimens of these two species was 108.8.

Pristis clavata (dwarf sawfish) (fig. 68)

Four specimens of this species, which is now considered restricted to northern Australia, were included here. One of these specimens is vouchered (GN3158  =  NTM S.14689-002). All four of these specimens were collected from Buffalo Creek in northern Australia. The analysis yielded a single cluster and the range in pairwise differences among these specimens was 0–4, with an average of pairwise differences of 2. The average of the pairwise differences between specimens of this species and those of P. pectinata was 119.6, and between those of P. zijsron was 116.7.

Pristis perotteti (largetooth sawfish) (fig. 68)

The 17 specimens of this coastal Atlantic-dwelling species all came from Belem in Brazil. The range of pairwise differences among these specimens was 0–11, with an average of pairwise differences of 3.7. The average of the pairwise differences between specimens of this species and those of P. pectinata, which were also collected from the western Atlantic, was 123.1

Pristis microdon (freshwater sawfish) (fig. 68)

The analysis yielded a single cluster comprised of two specimens of this species, both from the Norman River in Queensland, Australia. The sequences of these specimens were identical and they grouped most closely with specimens of Pristis perotteti. The average of the pairwise differences between specimens of these two species was 16.6.

Rhinobatidae (guitarfishes): group 1

Rhinobatos productus (shovelnose guitarfish) and Rhinobatos glaucostigma (speckled guitarfish) (fig. 68)

The analysis included 11 specimens identified as R. productus and eight identified as R. glaucostigma, all but one of which came from the Gulf of California; one of the specimens of R. productus was collected from the coast of California. The analysis yielded a single cluster comprised of specimens of both species with no clear subclustering of species within this cluster. The range in pairwise differences among the 19 specimens was only 0–7, with an average of 3.6. Nonetheless, these species differ markedly in morphology, most conspicuously in maximum length and color pattern. Thus, we have retained the specific identities of specimens within this cluster in order to call attention to the discrepancy between the molecular and morphological results in this case. The average of the pairwise differences between specimens identified as R. productus and R. glaucostigma was 4.9.

Pristidae (modern sawfishes): genus Anoxypristis

Anoxypristis cuspidata (narrow sawfish) (fig. 68)

The seven specimens of this species included in the analysis all came from northern and eastern Australia and the range in pairwise differences was 0–1. Given the relatively broad distribution of this species, which includes much of the Indo-West Pacific, these specimens represent the only the eastern parts of its distribution. It is interesting to note that these specimens grouped conspicuously away from the five other species of sawfishes of the genus Pristis included in the analysis. Furthermore, the range of pairwise differences between specimens of A. cuspidata and P. pectinata, P. zijsron, P. clavata, P. perotteti, and P. microdon was 159.2, 155.3, 151.4, 168.7, and 166.1, respectively.

Rhynchobatidae (wedgefishes)

Rhynchobatus species

The taxonomy of the genus Rhynchobatus is in need of full revision. Recent descriptions of two new Indo-West Pacific species (Compagno and Last, 2008, 2010) has partly resolved some of the issues but further taxonomic work is required. We have attempted to employ the revised taxonomy in naming the specimens in the clusters resulting from the analysis conducted here. Our treatment was further informed by the morphological information presented for these Indo-West Pacific species by Last and Stevens (2009) and Last et al. (2010c).

Rhynchobatus australiae (whitespotted shovelnose ray) (fig. 69)

The analysis yielded a cluster comprised of 12 specimens collected from a diversity of localities throughout the island of Borneo, as well as from northern Australia, Singapore, Vietnam, and Thailand. One specimen is in the Australian National Fish Collection (GN2996  =  ANFC H 6221-01). These 12 specimens were generally morphologically consistent with R. australiae (see Last and Stevens, 2009). Most conspicuously, the majority of these specimens exhibited the single well-defined black pectoral fin spot, surrounded by four smaller white spots on either side of their body and lacked dark spots behind their eyes. The one exception was a relatively large specimen from Sarawak (GN2893) in which the pectoral fin spot configuration was not as distinct. The range in pairwise differences among these specimens was 0–7, with an average of 3.3.

Rhynchobatus cf. laevis and Rhynchobatus palpebratus (eyebrow wedgefish) (fig. 69)

The analysis yielded a cluster comprised of five Australian specimens identified as R. cf. laevis and two specimens identified as R. palpebratus. One of the specimens of R. cf. laevis is vouchered in the Australian National Fish Collection (GN4627  =  ANFC H 6319-01). The range of pairwise differences among the seven specimens in this cluster was only 0–4, with an average of 2.3. Within the cluster there was no evidence of the existence of subclusters of specimens of either species. Nonetheless, specimens identified as R. cf. laevis are morphologically consistent with that species as defined by Last and Stevens (2009). For example, not only do these specimens possess a diffuse black pectoral fin spot surrounded by scattered small white spots on each side of their body, but they also lack dark spots or markings near their eyes, whereas the specimens identified as R. palpebratus exhibit a well-defined black pectoral fin spot surrounded by four smaller white spots on each side of the body and also have dark spots behind their eyes. Thus, we have retained the different species designations of these specimens in order to call attention to the apparent discrepancy between color pattern and molecular data in this instance. The average of the pairwise differences between R. palpebratus and R. cf. laevis was 3. This cluster grouped within a larger cluster containing R. australiae. The average of pairwise differences between specimens identified as R. cf. laevis and those of R. australiae was 41.6 and between specimens identified as R. palpebratus and R. australiae was 41.

Rhynchobatus laevis (smoothnose wedgefish) (fig. 69)

Another specimen (GN3004  =  ANFC H 6221-02) of this genus included in the analysis, collected from western Borneo, grouped along with but outside the cluster containing R. cf. laevis and R. palpebratus. The average of the pairwise differences between this specimen and those of R. australiae was 44.3, from those of R. cf. laevis 25 and from those of R. palpebratus 23.

Rhinidae (sharkrays)

Rhina ancylostoma (sharkray) (fig. 69)

The six specimens included here came from northern Australia and northern Borneo and thus represent only a small portion of the distribution of this widespread Indo-West Pacific species. They were found to comprise a single cluster, which grouped along with but outside the specimens of Rhynchobatus. The range of pairwise differences among specimens in this cluster was 0–4, with an average of 2.1.

Rhinobatidae (guitarfishes): group 2

Glaucostegus typus (giant shovelnose ray) complex (fig. 69)

Fifteen specimens identified as G. typus were included in the analysis. These were collected from northern Australia and eastern and southern Borneo and thus represent the more eastern parts of the distribution of this species, which occurs as far west as India. One of these specimens was vouchered (GN4214  =  ANFC H 7085-02). These specimens comprised a single cluster; the range of pairwise differences among specimens in this cluster was 0–3, with an average of 1.

Two additional specimens from Borneo clustered along with, but outside the specimens of G. typus. Although morphologically indistinguishable, these specimens exhibited an average pairwise difference from the specimens of G. typus of 20.5. In recognition of this molecular result, these two specimens, which were identical in sequence, have been given the designation G. cf. typus. Taxonomic investigation of these specimens needs to be undertaken.

Glaucostegus thouin (clubnose guitarfish) (fig. 69)

The six specimens of this species were collected from localities in western and southern Borneo. These specimens represent the central portion of the distribution of this species, which has been reported as far west as the Red Sea and eastward to New Guinea. The analysis yielded a single tight cluster with the range of pairwise differences among these specimens being 0–1. These specimens clustered most closely with those of G. typus and G. cf. typus; the average of the pairwise differences between these specimens and those of G. typus was 27.9, and between those of G. cf. typus was 36.3.

Rhinobatos cemiculus (blackchin guitarfish) (fig. 69)

The two specimens of this species included here, both collected from Senegal, were identical in sequence. They clustered along with but outside G. typus with an average of the pairwise differences relative to G. typus of 106.9, relative to G. cf. typus 106, and relative to G. thouin 116.3. Our specimens represent only a relatively small portion of the distribution of this species, which is distributed throughout much of the northern and western coasts of Africa.

Rhinobatos rhinobatos (common guitarfish) (fig. 69)

All 20 specimens included in the analysis were collected from Senegal and thus represent the central part of the distribution of this species, which is known from the Mediterranean Sea to Angola. The analysis yielded a single tight cluster and the range of pairwise differences among these specimens was 0–6, with an average of 2.

Rhinobatos annulatus (lesser guitarfish) (fig. 69)

The five specimens of this southern African species all came from South Africa, and their sequences were identical. This species grouped most closely with R. rhinobatos, but the average of the pairwise differences among specimens of these two species was substantial at 122.2.

Rhinobatos cf. schlegelii (fig. 69)

The analysis included four specimens from the Philippines (GN4388  =  BRU 071, GN2244  =  BRU 073, GN2253  =  MMLM 001, and GN2254  =  MMLM 012), identified by Compagno et al. (2005b) as Rhinobatos cf. schlegelii and considered to represent an undescribed species. However, the cluster also included a specimen (GN4326  =  JPAG 310) collected from the Philippines, and originally identified by Compagno et al. (2005b) as Rhinobatos sp. 1, in recognition of the fact that it was a specimen of uncertain identity. Since these specimens form a single cluster, we have used the designation of R. cf. schlegelii for this Philippine species. The range of pairwise differences among these five specimens was 0–3, with an average of pairwise differences of 1.8.

Rhinobatos sp. 1 (fig. 69)

Two specimens collected from Malaysian Borneo clustered along with but outside their congeners from the Philippines. These differed from one another by one base. This species is very similar morphologically and in color to R. formosensis from Taiwan and was considered to be conspecific with that species by Last et al. (2010c), but this analysis suggests that it likely represents an undescribed species. The average of the pairwise differences between these specimens and those of R. cf. schlegelii was 24.9.

Rhinobatos formosensis (Taiwan guitarfish) (fig. 69)

One specimen from Taiwan clustered along with, but outside the specimens from the Philippines and Malaysian Borneo. The color pattern and morphology of this specimen is consistent with R. formosensis. The average of the pairwise differences between this specimen and those of R. cf. schlegelii was 105.2, and between this specimen and those of Rhinobatos sp. 1 was 109.5.

Aptychotrema rostrata (eastern shovelnose ray) (fig. 69)

Two specimens of this eastern Australian endemic were included in the analysis. These specimens differed from one another by four bases.

Aptychotrema vincentiana (southern shovelnose ray) (fig. 69)

The analysis included one specimen of this southwestern Australian endemic species, from the Australian National Fish Collection (GN4625  =  ANFC H 6348-06). It grouped along with but outside those of A. rostrata. The average of the pairwise differences between this specimen and those of A. rostrata was 51.

Zapteryx exasperata (banded guitarfish) (fig. 69)

Two specimens of this species, both collected from the Gulf of California, were included; they differed from one another by a single base.

Trygonorrhina dumerilii (southern fiddler ray) (fig. 69)

A single sample of this southwestern Australian endemic species was included in the analysis. The specimen is in the Australian National Fish Collection (GN4626  =  ANFC H 6346-22).

Torpedinidae (Torpedo rays)

Torpedo fuscomaculata (blackspotted torpedo) (fig. 70)

A total of 11 specimens of this species, which is considered to occur off the southern regions of Africa and possibly elsewhere in the Indian Ocean, were included. The range of pairwise differences among the specimens of this species, which were all collected off South Africa, was 0–6, with an average of pairwise differences of 2.6.

Torpedo torpedo (ocellate torpedo) (fig. 70)

A single specimen from Senegal provisionally identified by Marcelo de Carvalho (personal commun.) as this species was included. This specimen grouped most closely with T. fuscomaculata, with an average of the pairwise differences between species of 45.4.

Torpedo sinuspersici (gulf torpedo) (fig. 70)

Two specimens of this western Indian Ocean species, both collected in the Persian Gulf off Iran, were included. These clustered together, but differed by 22 bases. They grouped most closely with but outside T. fuscomaculata and T. torpedo; the average of the pairwise differences between T. sinusperscici and the latter two species was 52.7 and 64.5, respectively.

Torpedo mackayana (ringed torpedo) (fig. 70)

The single specimen of this species, which has been reported from Senegal to Angola, was included in the analysis. This specimen was collected from Senegal and thus represents only a northern element of the distribution of this species. The identification of this specimen was provided by M. de Carvalho (personal commun.).

Torpedo marmorata (spotted torpedo) (fig. 70)

The specimen included was collected from Senegal and thus represents only one element of the distribution of this species, which occurs throughout much of the western coast of Europe, the Mediterranean, and also much of Africa. Although this specimen resembles T. bauchotae, it has tentatively been identified as T. marmorata by M. de Carvalho (personal commun.) at this time.

Torpedo nobiliana (Atlantic torpedo) (fig. 70)

The analysis included eight specimens of T. nobiliana, all collected from the western Atlantic. These specimens represent only the western part of the distribution of this species, which, at present, is considered to occur in both the western and eastern Atlantic Ocean and Mediterranean Sea. However, the western Atlantic form has been considered by some to represent the distinct species T. occidentalis (see Storer, 1843). It is unfortunate that our analysis did not include specimens from the eastern Atlantic and thus did not allow comparison across this distribution. Until this issue can be investigated in more detail, we have used the name in common use for this species here. However, given that syntypes of T. nobiliana come from Italy, the name T. occidentalis may be more appropriate for the western Atlantic cluster if detailed study confirms the nonconspecificity of the western and eastern Atlantic Ocean forms. The range in pairwise differences among these eight specimens was 0–5, with an average of pairwise differences of 1.7.

Torpedo macneilli (Australian torpedo) and Torpedo cf. nobiliana (fig. 70)

The analysis included four specimens of the Australian endemic T. macneilli and a single specimen from South Africa initially identified as T. nobiliana. These specimens were found to group together well away from the specimens of Torpedo nobiliana collected from western Atlantic localities. The specimens of T. macneilli differed from one another by 0–2 bases, with an average of 1. This supports the work of Dave Ebert suggesting that the South African form, referred to here as T. cf. nobiliana, is not conspecific with its western Atlantic counterpart. The average of the pairwise differences between T. macneilli and T. cf. nobiliana was 5.5. The average of the pairwise differences between the South African specimens of T. cf. nobiliana and those of T. nobiliana from the western Atlantic was 27.6. The formal description of this species is currently underway (D. Ebert, personal commun.).

Narkidae (sleeper rays): genus Typhlonarke

Typhlonarke aysoni (blind legged torpedo) (fig. 70)

Two specimens of this New Zealand endemic were included in the analysis; both were vouchered (GN6758  =  NMNZ P.041329 and GN6759  =  NMNZ P.042187). They differed from one another by three bases. These specimens were found to comprise a cluster that grouped most closely with the Torpedo species.

Narcinidae (numbfishes)

Narcine tasmaniensis (Tasmanian numbfish) (fig. 70)

The four specimens of this Australian endemic included here all came from New South Wales, Australia. The range of the pairwise differences among these specimens was 0–9, with an average of 4.5.

Narcine lasti (western numbfish) (fig. 70)

A single specimen of this western Australian species was included in the analysis. It is in the Australian National Fish Collection (GN4628  =  ANFC H 6418-01). It grouped most closely with the specimens of N. tasmaniensis. The average of the pairwise differences between this specimen and those of the latter species was 83.8.

Narcine entemedor (Cortez electric ray) (fig. 70)

The analysis included a single specimen of this species, which was collected from the Gulf of California and thus represents the more northern part of the distribution of this species, which has been reported as far south as Peru. This specimen grouped with Narke capensis outside all the other species and genera of electric rays included in the analysis. The average of the pairwise differences between this species and N. tasmaniensis was 282 bases, and between N. entemedor and N. lasti is 281 bases.

Narkidae (sleeper rays): genus Narke

Narke capensis (Cape sleeper ray) (fig. 70)

A single specimen of this southern African endemic was included. It grouped most closely with the specimen of Narcine entemedor. However, the specimens of these two species differed by 245 bases.

Platyrhinidae (thornbacks and fanrays)

Platyrhinoidis triseriata (thornback) (fig. 70)

Both of the specimens of this monotypic genus included in the analysis were collected from California, and thus are generally representative of the distribution of this species, which also includes the Gulf of California. The sequences of these two specimens were identical.

Rajidae (skates)

Dipturus innominatus (New Zealand smooth skate) (fig. 71)

In total, 11 specimens of this New Zealand endemic were included in the analysis. They were found to comprise a single cluster and the range in pairwise differences among specimens in this cluster was 0–6, with an average of 2.2.

Dipturus gudgeri (bight skate) (fig. 71)

The differences between sequences of the three specimens of this Australian endemic that were included in the analysis ranged from 1–7, with an average of 4.7. These grouped most closely with the specimens of D. innominatus. The average of the pairwise differences between specimens of these two species was 29.2.

Dipturus sp. 4 (fig. 71)

The analysis included two skate specimens collected from the Philippines (GN4339  =  BRU 096 and GN4353  =  JPAG 091) that were treated by Compagno et al. (2005b) as Dipturus sp. 4. These specimens differed from one another by 13 bases. As noted by Compagno et al. (2005b) these specimens may represent an undescribed species in this genus.

Dipturus healdi (Heald's skate) complex (fig. 71)

Two specimens, both of which had provisionally been identified as Dipturus healdi, were included in the analysis. These specimens differed from one another by 56 bases. Two morphs referable to D. healdi were discussed by Last et al. (2008d), and a specimen of each morph was included in this analysis. The specimen of the northern morph (GN6788  =  ANFC H 6574-20) is consistent with D. healdi and clusters well away from the southern morph (GN6789  =  ANFC H 6419-04), which is provisionally referred to as Dipturus cf. healdi; these forms are indisputably separate species. Dipturus healdi clustered most closely with Dipturus sp. 4, with an average of the pairwise differences between these two taxa of 39.5.

Dipturus tengu (acutenose skate) (fig. 71)

The analysis included one specimen identified as this species, collected from Taiwan. If this identification is correct, this specimen is representative of a central element of the distribution of this species, which is known to occur from Japan to the Philippines.

Dipturus springeri (roughbelly skate) (fig. 71)

The four specimens of this species, all collected from South Africa, were identical in sequence. They represent the central southwestern elements of the distribution of this species, which occurs along the coast of Africa from Angola to Somalia as well as Madagascar.

Dipturus australis (Sydney skate) (fig. 71)

Six specimens of this species, which is endemic to eastern Australia, were included in the analysis. They formed a single cluster. The range of pairwise differences among these specimens was 0–2, with an average of 0.7.

Dipturus cerva (whitespotted skate) (fig. 71)

Three specimens of this South Australian endemic were included in the analysis. They were identical in sequence.

Dipturus confusus (longnose skate) (fig. 71)

Two specimens of this recently described (see Last, 2008) southeastern Australian endemic species were included here. They were identical in sequence and clustered most closely with the two specimens of D. cerva. The average of the pairwise differences among specimens of these two Australian endemic species was 18, and between this species and its other Australian congener, D. australis, was 30.3.

Zearaja chilensis (yellownose skate) (fig. 71)

The eight specimens of this species included here had a range of pairwise differences of 0–5, with an average of 2; they comprised a single cluster. These specimens all came from a single locality in Chile and thus represent the western part of the distribution of this species, which occurs throughout the coastal regions of southern South America and the Falkland Islands.

Zearaja nasuta (New Zealand rough skate) (fig. 71)

Five specimens of this New Zealand endemic were included in the analysis. These specimens had a range of pairwise differences of 1–5, with an average of 3.2. They comprised a cluster that grouped most closely with the specimens of Z. chilensis. The average of the pairwise differences between specimens of these two species was 10.3.

Zearaja flavirostris (Falklands skate) (fig. 71)

All three specimens identified as this species included in the analysis, were provided by Joost Pompert, of the Falkland Islands Fisheries Department, extending the range of this species from central Chile to the Falkland Islands. They comprised a single cluster, with a range of pairwise differences among specimens of 0–3 and an average of 2. They grouped most closely with the cluster consisting of Z. chilensis and Z. nasuta. The average of the pairwise differences between specimens of Z. flavirostris and those of Z. chilensis was 29.5, and those of Z. nasuta was 33.9. Given these differences, we have recognized Z. flavirostris as a species distinct from Z. chilensis, despite the suggestion that these two species may be synonyms (e.g., Last and Gledhill, 2007).

Dipturus leptocauda (thintail skate) (fig. 71)

Both specimens identified as this species were collected from the Falkland Islands, extending the range of this species from Brazil and Uruguay south to include the Falkland Islands. The sequences of these specimens were identical to one another.

Dipturus batis (gray skate) and Dipturus oxyrhinchus (sharpnose skate) complex (fig. 71)

The analysis included a total of 12 specimens originally identified as D. batis and two identified as D. oxyrhinchus. These specimens come from several localities in the eastern North Atlantic and were found to comprise a single group, exhibiting considerable internal structure. The largest cluster consisted of eight specimens from the eastern North Atlantic and three from the Azores. The range of pairwise differences among the 11 specimens in this cluster was 0–8, with an average of 4.3. However, there was evidence of structure within this cluster, with six of the specimens from the eastern North Atlantic (identified as D. batis by Bernard Séret) comprising a subcluster distinct from that consisting of the three specimens from the Azores identified locally as D. batis and two specimens from the eastern North Atlantic identified as D. oxyrhinchus, also by Bernard Séret. The average of the pairwise differences among specimens in these two subclusters was 7.5. In order to call attention to this discrepancy, we have retained the original identifications of these specimens. This result suggests that the identity of specimens of D. batis from the Azores needs to be investigated in more detail.

One specimen, also collected from the eastern North Atlantic, grouped outside this subcluster of 11 specimens. The average of the pairwise differences between this specimen and those in the D. batis and D. oxyrhinchus subcluster was 23.7. This specimen has been given the designation Dipturus cf. batis 1. Finally, two specimens, both collected from Norway, grouped outside all 12 other specimens in the cluster overall; these two specimens were identical with one another in sequence. The average of the pairwise differences between these specimens and those in the D. batis and D. oxyrhinchus subcluster was 16.7, and between D. cf. batis 1 was 37. These two specimens have been given the designation Dipturus cf. batis 2.

These results suggest that it is likely these specimens represent several distinct species, all of which bear a morphological resemblance to Dipturus batis. Recent published work by Iglésias et al. (2009) reveals that the D. batis complex likely consists of two species, both of which should be considered as valid, i.e., D. cf. flossada and D. cf. intermedia. Following these designations and morphological characters provided by Igl?sias et al. (2009), the specimens with images in the second subcluster (GN5160, GN5165 and GN6546) are consistent with D. cf. intermedia. Thus, this second subcluster, which includes the two D. oxyrhinchus specimens, may represent D. cf. intermedia. There is a possibility that one of the other subclusters represents D. cf. flossada and one represents D. oxyrhinchus. Another species which Iglésias et al. (2009) reported as very close to these species is the Norwegian skate D. nidarosiensis, and it is possible the two specimens from Norway may represent that species. Additional specimens, with designated vouchers and confirmed identifications based on the work of Iglésias et al. (2009), are required to resolve this issue. The lack of specific locality data for our eastern North Atlantic specimens makes it difficult for us to compare our results with those of Griffiths et al. (2010), except to note that, like these authors, we found specimens identified as D. oxyrhinchus to cluster among specimens identified as D. batis.

Dipturus laevis (barndoor skate) (fig. 71)

The three specimens of this unusually large skate included in the analysis all came from the western North Atlantic and are thus generally representative of the distribution of this species. One of these specimens is in the Harvard Museum of Comparative Zoology MCZ (GN2603  =  MCZ 159202 and TCWC 11021.01). The range of the pairwise differences among these specimens was 0–1. These specimens grouped along with, but outside the 14 specimens in the Dipturus batis and Dipturus oxyrhinchus complex.

Dipturus pullopunctatus (slime skate) (fig. 71)

Three specimens of this southern African endemic, all collected from South Africa, were included. These specimens were identical in sequence.

Spiniraja whitleyi (Melbourne skate) (fig. 71)

Two specimens of this southern Australian endemic species were included in the analysis. These differed from one another by two bases and they grouped along with but outside the Dipturus and Zearaja species.

Raja sp. 1 (fig. 71)

A single specimen collected from Taiwan clustered independently from all specimens of Dipturus, Zearaja, Spiniraja, and Raja. It is possible that this specimen, for which an image is available, represents a known species, but we have been unable to assign a specific identification with confidence at this time; its generic designation also requires confirmation.

Raja rhina (longnose skate) (fig. 71)

This northeastern Pacific species was represented by 17 specimens; the analysis yielded a single cluster. The range of pairwise differences among the specimens in this cluster was 0–8, with an average of 1.3. Two of these specimens are deposited in the University of Washington Fish Collection (GN6712  =  UW 47645 and GN6713  =  UW 49457). This species grouped well away from the specimens of most of the other Raja species included in the analysis (see figs. 73, 75). This is consistent with McEachran and Dunn (1998) and Compagno (2005b), who listed this species, among others, as comprising a “North Pacific Assemblage” that may represent an undescribed genus.

Raja binoculata (big skate) (fig. 71)

The analysis included two specimens of this northeastern Pacific species. These specimens differed by six bases. This species, like Raja rhina, grouped well outside specimens of the other Raja species included in the analysis (see figs. 73, 75). This is consistent with McEachran and Dunn (1998) and Compagno (2005b) who listed these two species, among others, as comprising a “North Pacific Assemblage” that may represent an undescribed genus. The average of the pairwise differences between specimens of this species and those of Raja rhina was 111.2.

Okamejei cairae (Borneo sand skate) (fig. 72)

Thirteen specimens of this recently described species (see Last et al., 2010a) were included in the analysis. These were all collected from Sarawak in Malaysian Borneo, Pontianak in West Kalimantan (Indonesia), or Vietnam. The specimens from Vietnam extend the northern boundary of the distribution of this species. They comprised a single cluster and the range of pairwise differences among them was 0–4, with an average of 2.1. All nine specimens from Kalimantan were vouchered (GN4730  =  ANFC H 7099-07, GN4731 through GN4735  =  ANFC H 7099-01 through ANFC H 7099-05, GN4737  =  ANFC H 7099-06, GN4729  =  MZB 17176, and GN4736  =  MZB 17177).

Okamejei cf. porosa (fig. 72)

Five specimens collected from Japan were included in the analysis. While the range of pairwise differences among these specimens was only 0–8 (with an average of 3.9) there was substantial morphological variation among these specimens (e.g., in color pattern, disc shape, etc.), based on images for four of the five specimens. These specimens, although they also resemble Okamejei meerdervoortii, are provisionally referred to here as Okamejei cf. porosa until their identities can be examined in more detail. These specimens grouped most closely with those of Okamejei cairae. However, the average of the pairwise differences between the two species was 96.

Okamejei hollandi (yellow-spotted skate) (fig. 72)

A total of 14 specimens, identified as O. hollandi by Last et al. (2010c), were included in the analysis. These specimens were all collected from Pontianak in West Kalimantan (Indonesia); one was vouchered (GN4738  =  ANFC H 7099-08). The range of pairwise differences among these 14 specimens was 0–13, with an average of six bases. The average of their pairwise differences relative to specimens of Okamejei cairae was 106.9; the average of their pairwise differences relative to specimens of Okamejei cf. porosa was 107.6.

Raja miraletus (brown skate) complex (fig. 73)

In total, 23 specimens that were generally consistent with the color pattern of R. miraletus (i.e., in their possession of small dark spots and a pair of distinctive ocelli on their discs) were included. These consisted of 17 specimens from Senegal and six from South Africa. The analysis yielded three distinct clusters. Two of these clusters consisted of specimens from Senegal and the third of specimens from South Africa. The average of the pairwise differences between specimens in the two Senegal clusters was 99.3. Morphologically, specimens of these two clusters differed most conspicuously in coloration; in one form the small dark disc spots were interspersed with numerous white spots, whereas specimens of the other form exhibited few, if any, small white spots. Specimens of the first Senegal cluster exhibiting the latter color pattern were designated R. miraletus. The more heavily spotted specimens comprising the second cluster of specimens from Senegal were given the designation Raja cf. miraletus 2. The range of pairwise differences among specimens in the R. miraletus cluster was 0–7 (with an average of 2) and among specimens in the R. cf. miraletus 2 cluster was 1–4 (with an average of 2.4).

The cluster of specimens from South Africa had a range of pairwise differences of 0–2 (with an average of 1.2). They most closely resembled R. cf. miraletus 2 in color pattern. However, they clustered more closely with the specimens of R. miraletus. These specimens have been given the designation Raja cf. miraletus 1 until such time as the taxonomy of this complex can be examined in more detail. The average of the pairwise differences between R. cf. miraletus 1 and R. cf. miraletus 2 was 85.2 and between R. cf. miraletus 1 and R. miraletus was 49.1. Our results suggest that the differences seen among these three forms may reflect the existence of several distinct species in these regions as was observed by McEachran et al. (1989). However, it is important to note that several known species from one or more of these regions (e.g., R. herwigi and R. ocellifera) were not included in the analysis.

Raja clavata (thornback skate) (fig. 73)

Eight of the specimens of this species were collected from the eastern North Atlantic and the Azores; the ninth specimen came from the Black Sea. The range in pairwise differences among all nine specimens was 0–11, with an average of 6.1. The analysis yielded some evidence of a subcluster comprised of the specimens from the Azores, and the other comprised of the specimens from the eastern North Atlantic and Black Sea. The average of the pairwise differences among specimens in these two subclusters was 9.5. The range within the Azores subcluster was 0–3, and within the North Atlantic subcluster 1–3. These results parallel the situation in the D. batis complex, which provides evidence that the fauna of the Azores may differ somewhat from that of their mainland counterparts. This result is consistent with that of Chevolot et al. (2006) who found regional differentiation between specimens from the European continental shelf and the Azores.

Raja straeleni (biscuit skate) (fig. 73)

The nine specimens of this species, which occurs throughout much of the western coast of Africa, were all collected off South Africa. The range in pairwise differences among specimens of this species was 0–3, with an average of 1.6. This species grouped most closely with Raja clavata. The average of the pairwise differences between these two species was 23.1.

Raja asterias (Atlantic starry skate) (fig. 73)

The analysis included three specimens of this species, all collected from Spain and generally representative of the eastern elements of this primarily Mediterranean-dwelling species. The range of pairwise differences among specimens was 1–4, with an average of 2.7. The analysis yielded a single cluster, which grouped most closely with that consisting of R. clavata and R. straeleni. The average of the pairwise differences between R. asterias and R. clavata was 62.1, and between R. asterias and R. straelini was 66.1.

Raja montagui (spotted skate) (fig. 73)

A single specimen of this species was included here. This specimen was collected from the coast of England and thus represents a northerly element of the distribution of this species, which extends throughout the Mediterranean Sea and northern coast of Africa. It grouped most closely with but outside the Raja clavata, Raja straeleni, and Raja asterias specimens. The average of the pairwise differences between this specimen and those of R. clavata was 66.1, between this specimen and those of R. straeleni 70.3, and between this specimen and those of R. asterias 78.3.

Rajella fyllae (round skate) (fig. 74)

The analysis included 18 specimens from the eastern North Atlantic (identified by Bernard Séret) and three from Norway. These were found to comprise a single cluster and the range of pairwise differences among these 21 specimens was 0–6, with an average of pairwise differences of 1.7. These specimens represent the more eastern parts of the distribution of this species, which occurs throughout the entire North Atlantic.

Rajella sp. (fig. 74)

Two specimens from the Harvard Museum of Comparative Zoology (GN4664  =  MCZ 167899 and GN4665  =  MCZ 167900), both collected on the continental shelf off New England, were included. These two specimens were identical in sequence. They grouped most closely with the two specimens of Rajella caudaspinosa; these two species were, in turn, most closely allied with Rajella fyllae. The average of the pairwise differences between Rajella sp. and R. caudospinosa was 33, and between the specimens of Rajella sp. and those of R. fyllae 33.3. We believe these specimens represent a western Atlantic relative of R. caudospinosa.

Rajella caudaspinosa (munchkin skate) (fig. 74)

Two specimens of this southern African endemic, all collected from South Africa, were included. These specimens were identical in sequence. They clustered most closely with Rajella sp. As noted above, the average of the pairwise differences between these two species was 33.

Rajella leopardus (leopard skate) (fig. 74)

Five specimens of this southern African endemic, all collected from South Africa, were included in the analysis. The range in pairwise differences among specimens in this cluster was 0–3, with an average of 1.2.

Rajella kukujevi (mid-Atlantic skate) (fig. 74)

The four specimens of this species included in the analysis, all collected from the North Atlantic, comprised a single cluster. The range of pairwise differences among these specimens was 0–3, with an average of 1.5. These specimens grouped most closely with those of Rajella leopardus. The average of the pairwise differences between these two species was 49.

Dipturus linteus (sailskate) (fig. 74)

A total of five specimens identified as this species, all collected from the eastern North Atlantic and off Norway, and thus representing the eastern portions of the distribution of this species, were included in the analysis. The range in pairwise differences among these specimens was 0–1. They comprised a single cluster, which grouped most closely with the clusters of the three Rajella species. This Dipturus species clustered well away from all 15 of its congeners included in the analysis (see fig. 69), lending support to Compagno (2005b) who listed this species as “Dipturus? linteus” and noted that its generic placement needs investigating.

Amblyraja radiata (thorny skate) (fig. 74)

In total, the analysis included 14 specimens identified as this species, one of which is in the Harvard Museum of Comparative Zoology (GN2602  =  MCZ 159184). These were found to comprise a single cluster. Although there was evidence of structure within this cluster, and the range of pairwise differences among the members of this cluster was 0–15, with an average of 8.5, there was no obvious pattern to this variation; all specimens within this cluster have thus been given the same designation. These specimens represent the more eastern part of the distribution of this species, which occurs throughout the North Atlantic.

Amblyraja hyperborea (Arctic skate), Amblyraja jenseni (Jensen's skate), Amblyraja badia (broad skate), and Amblyraja sp. complex (fig. 74)

The analysis yielded a somewhat confusing cluster of nine specimens, five of which are deposited in museums. This cluster consists of four specimens from the Barents Sea identified as A. hyperborea by Age Hoines, one specimen from California identified as A. badia deposited at the University of Washington (GN6681  =  UW 115021), one specimen from the mid-Atlantic identified as A. jenseni by Chip Cotton, and deposited at Virginia Institute of Marine Sciences (GN5040  =  VIMS 11757), as well as two specimens deposited at the Harvard Museum of Comparative Zoology (MCZ GN4666  =  MCZ 167944 and GN4667  =  MCZ 167945), both identified only as Amblyraja sp., and a specimen from the Tasman Sea, identified as Amblyraja hyperborea, from the Australian National Fish Collection (GN4650  =  ANFC H 5944-01). These specimens differ somewhat in morphology. The range of pairwise differences among the nine specimens comprising this cluster was 0–11, with an average of 5.7. As none of the specimens comprising this cluster have been definitively identified, we have retained the original identifications as they serve to draw attention to the uncertainty associated with the identity of the members of this cluster.

Amblyraja doellojuradoi (southern thorny skate) (fig. 74)

Three specimens of this species, all collected from the Falkland Islands, were included in the analysis. They represent the more eastern parts of the distribution of this species, which occurs throughout the southern regions of South America. The range of pairwise differences among specimens was 4–8, with an average of 5.3. These specimens grouped most closely with the A. hyperborea, etc., cluster. The average of the pairwise differences between the specimens of these two clusters was 18.7.

Leucoraja naevus (cuckoo skate) (fig. 75)

In total, 29 specimens of this species were included in the analysis. These were all collected from localities in the eastern North Atlantic and thus represent the northern portions of the distribution of this species, which has been reported from as far south as Senegal. The range of pairwise differences among the specimens in this species was 0–6, with an average of pairwise differences of 1.5.

Leucoraja wallacei (yellowspot skate) (fig. 75)

The 12 specimens of this southern African species, all collected from South Africa, were found to comprise a single cluster. The range in pairwise differences among specimens was 0–8, with an average of 3.7. This species grouped most closely with the eastern North Atlantic species Leucoraja naevus; the average of the pairwise differences between these two species was 37.5.

Leucoraja fullonica (shagreen skate) (fig. 75)

The analysis included two specimens collected from the Azores and two from the eastern North Atlantic. These specimens thus represent some of the southern and western elements of this species, which occurs as far eastward as Murmansk, Russia. The range in pairwise differences among these specimens was 0–3, with an average of 2. The analysis yielded a single cluster, which grouped most closely with the cluster consisting of L. naevus and L. wallacei. The average of the pairwise differences among specimens of L. fullonica and L. naevus was 72.4, and L. wallacei was 65.4.

Leucoraja erinacea (little skate) (fig. 75)

In total, four specimens of this western Atlantic species were included in the analysis, which yielded a single cluster with a range of pairwise differences among specimens of 3–8 and an average of pairwise differences of 5.3. One of these specimens is in the University of Kansas Ichthyology Collection (GN2505  =  KUI 26967).

Leucoraja ocellata (winter skate) (fig. 75)

The two specimens of this western Atlantic species included differed by two bases. They grouped most closely with the specimens of L. erinacea; the average of the pairwise differences between specimens of these two species was 45.5.

Leucoraja garmani (rosette skate) (fig. 75)

The single specimen of this species, collected from the western North Atlantic, represents a more northern element of the distribution of this species, which occurs as far south as Venezuela. It grouped outside all four other species of this genus included here. The averages of the pairwise differences between this specimen and its congeners were: L. naevus 130.3, L. wallacei 130.9, L. fullonica 132.5, L. erinacea 142.8, and L. ocellata 143.

Rostroraja alba (white skate) (fig. 75)

A total of six specimens of this species was included in the analysis. These specimens were all collected from South Africa and thus represent only a central element of the distribution of this species, which has been reported from Great Britain, the Mediterranean Sea, much of the coast of Africa, and the Red Sea. These specimens were identical in sequence. This species grouped most closely with the specimens identified as Raja velezi. The average of the pairwise differences between Ro. alba and Ra. velezi was 86. This grouping is generally consistent with the affinities reported by McEachran and Dunn (1998) and Compagno (2005b) who listed these species, among others, as comprising an “Amphi-American Assemblage.” However, these results suggest that some consideration should be given to Rostroraja as the generic name to apply to the species comprising this entire assemblage.

Raja velezi (rasptail skate) (fig. 75)

All three specimens of this species were identical in sequence. All three were collected from the Gulf of California and thus represent a northern distribution of this species, which has been reported from as far south as Peru. These specimens grouped most closely with those of Ro. alba.

Raja eglanteria (clearnose skate) (fig. 75)

Four specimens, identified as R. eglanteria, collected from the western North Atlantic and the Gulf of Mexico off Florida, were included in the analysis. These represent much of the distribution of this species, which occurs from Massachusetts to Florida. The range in pairwise differences among specimens was 0–12, with an average of 8.2. This species grouped with Ra. velezi and Rostroraja alba, well outside specimens of the other Raja species included in the analysis (see figs. 71, 73), which is consistent with McEachran and Dunn (1998) and Compagno (2005b) who listed Ra. eglanteria and Ra. velezi, among other species, as comprising an “Amphi-American Assemblage” that may represent an undescribed genus.

Okamejei jensenae (Philippine ocellate skate) (fig. 75)

The analysis included two specimens of the recently described Okamejei jensenae (see Last and Lim, 2010) from the Philippines, which were treated by Compagno et al. (2005b) (GN4343  =  BRU 171 and GN4382  =  JPAG 328) as Okamejei sp. 1. These differed from one another by one base. It is of note that this species grouped most closely with specimens in the R. eglanteria and R. velezi clusters, well away from its congeners (see fig. 70). The generic placement of this species requires additional taxonomic investigation when adult males of this species are collected as clasper cartilages are needed to assign this taxon to a genus.

Malacoraja senta (smooth skate) (fig. 75)

All three specimens of this species were found to be identical in sequence. These specimens were collected from the western North Atlantic and are largely representative of the distribution of this species, which occurs from Newfoundland to New Jersey.

Neoraja caerulea (blue pygmy skate) (fig. 75)

Two specimens of this eastern North Atlantic endemic were included in the analysis; they were identical in sequence. They clustered most closely with the specimens of M. senta and the average of the pairwise differences among specimens of these two species was 80.

Arhynchobatidae (softnose skates)

Bathyraja spp. (includes Bathyraja sp., B. interrupta, B. kincaidi, B. kincaidi/interrupta, B. mariposa, B. taranetzi, B. violacea) (fig. 76)

The analysis yielded a cluster comprised of 19 specimens whose identities are uncertain. The range of pairwise differences among these specimens was 0–24, with an average of pairwise differences of 7.6. These specimens were collected from a diversity of localities in the eastern North Pacific Ocean, including the Gulf of Alaska, Bering Sea, and Strait of Juan de Fuca. Their specific designations, assigned by their original collectors, varied wildly and include Bathyraja sp., B. interrupta, B. kincaidi, B. kincaidi/interrupta, B. mariposa, B. taranetzi, and B. violacea. The original field identifications have been retained for all specimens, regardless of current thoughts on generic placements and specific synonymies, because their identities of these specimens could not be confirmed. This draws attention to the need for voucher specimens and/or images to confirm species designations and problems created when they are unavailable. It should be noted that only three of these samples come from museum specimens (i.e., B. interrupta GN6689  =  UW 111883, B. violacea GN6707  =  UW 48757, and B. mariposa GN6694  =  UW 47201). In addition, as indicated in figure 76, at least some images are available for another eight specimens. We suspect that the mixed nature of this cluster may reflect either extensive hybridization among closely related forms with different color patterns or complex polymorphism that has hindered accurate identification in the field.

Bathyraja minispinosa (smallthorn skate) (fig. 76)

The three specimens of this North Pacific species included here were collected from the eastern North Pacific. All three are deposited at the University of Washington (i.e., GN6696  =  UW 47049, GN6697  =  UW 117948, GN6698  =  UW 117950). The range of pairwise differences among the specimens comprising this cluster was 0–4, with an average of 2.7. These grouped most closely with the specimens comprising the Bathyraja spp. cluster; the average of the pairwise differences between specimens of B. minispinosa and those in the Bathyraja spp. cluster was 24.1.

Bathyraja pallida (pallid skate) (fig. 76)

Four samples from skates identified by their collectors as B. pallida were included in the analysis. Two of these specimens are deposited in museums (GN4668  =  MCZ 167975 and GN5039  =  VIMS 11758). The analysis yielded a single cluster comprised of these four specimens, however, the range of pairwise differences among these specimens was 4–21, with an average of 17.3. We note, however, that while B. pallida is known only from the Bay of Biscay, the specimens included here come from a diversity of localities that include the western Atlantic off New England, the mid-Atlantic, and the Barents Sea. Further investigation of the identity of skates from these regions is required.

Bathyraja shuntovi (narrownose skate) (fig. 76)

A single specimen of this New Zealand endemic was included in the analysis. This specimen clustered most closely with those of Bathyraja pallida. The average of the pairwise differences between those comprising the latter cluster and the specimen of B. shuntovi was 30.5.

Bathyraja trachura (roughtail skate) (fig. 76)

The specimen of this eastern North Pacific endemic included in the analysis was collected from the Gulf of Alaska.

Bathyraja aleutica (Aleutian skate) fig. 76)

This cluster consisted of a specimen from the Gulf of Alaska and one from the Bering Sea; they differed from one another by 10 bases. The identity of these specimens was confirmed by D. Ebert (personal commun.) following examination of the images of GN5722.

Rhinoraja macloviana (Patagonian skate) (fig. 76)

Three specimens of this southeastern South American endemic skate, all from the Falkland Islands, were included in the analysis. The range of pairwise differences among the specimens in this cluster was 0–1.

Rhinoraja albomaculata (whitedotted skate) (fig. 76)

Three specimens of this southern South American endemic species, all collected from the Falkland Islands, were included here. The range of pairwise differences among these specimens was 1–4, with an average of 2.7; they clustered most closely with specimens of R. macloviana, with an average pairwise difference among specimens of 9.7. We note that this was one of several species of Rhinoraja that grouped among the clusters of species of Bathyraja.

Bathyraja brachyurops (broadnose skate) (fig. 76)

The four specimens of this species included in the analysis were found to comprise a single cluster with a range of pairwise differences among specimens of 0–10, and an average of pairwise differences of 7. All four of these specimens came from the Falkland Islands and thus represent a southeastern part of the distribution of this species, which is known from Chile to northern Argentina.

Rhinoraja magellanica (Magellan skate) (fig. 76)

A single specimen of this species, collected from the Falkland Islands, was included. This southern South American endemic clustered along with but outside the specimens of B. brachyurops, R. macloviana, and R. albomaculata.

Bathyraja sp. 1 (fig. 76)

Three unidentified Bathyraja specimens from the Falkland Islands clustered together. The range of pairwise differences among specimens in this cluster was 0–5, with an average of 3.3.

Bathyraja griseocauda (graytail skate) (fig. 76)

The analysis included two specimens, both from the Falkland Islands, of this southern South American and Antarctic endemic. These specimens differed from one another by a single base. They grouped most closely with the specimens of Bathyraja sp. 1. The average of the pairwise differences between specimens of what we believe are two distinct species was 22.2.

Bathyraja scaphiops (cuphead skate) (fig. 76)

The analysis included two specimens of this southwestern Atlantic endemic. These specimens differed from one another by two bases.

Bathyraja cf. taranetzi (fig. 76)

The analysis included a single specimen collected from the Bering Sea that was originally identified as B. taranetzi, a species currently known only from the Kuril Islands in the western North Pacific. This specimen grouped most closely with the Bathyraja scaphiops cluster; the average of the pairwise differences between these two species was 42. Several other specimens also identified by their collectors as B. taranetzi clustered elsewhere in the analysis (see above). Thus, we have given this specimen the tentative designation of B. cf. taranetzi. The grouping of this specimen among Bathyraja species is noteworthy in the context of the generic distinctions between Rhinoraja and Bathyraja.

Rhinoraja multispinis (multispine skate) (fig. 76)

Three specimens of this South American endemic species, all collected from the Falkland Islands, were included in the analysis. All three specimens had identical sequences.

Bathyraja parmifera (Alaska skate) (fig. 76)

Eight specimens tentatively identified as B. parmifera, two of which are deposited at the University of Washington (GN6699  =  UW 111889 and GN6721  =  UW 117306), were included in the analysis. These specimens were all collected from localities in the eastern North Pacific Ocean, including the Bering Sea, and are generally representative of the distribution of this species. The range of pairwise differences among these specimens was 0–12, with an average of 3. The upper end of this range was accounted for by the second specimen in the UW collection. Given that some of the specimens in this cluster were originally identified as B. aleutica, and none of the specimens in this cluster have been firmly established to represent B. parmifera, the identity of this cluster remains to be confirmed. Our results are consistent with those of Spies et al. (2011) and suggest that this species should be recognized as a member of the generic group level taxon Arctoraja.

Bathyraja smirnovi (golden skate) (fig. 76)

A single specimen of this western North Pacific Ocean species was included. It clustered most closely with the specimens of B. parmifera. The average of the pairwise differences between specimens of these two species was 17.5. Our results are consistent with those of Spies et al. (2011) and suggest that this species should also be recognized as a member of the generic group level taxon Arctoraja.

Bathyraja bergi (bottom skate) (fig. 76)

A single specimen of this western North Pacific species was included in the analysis. It clustered along with but outside those of B. parmifera and B. smirnovi; the averages of the pairwise differences between this species and the specimens comprising the clusters of each of the latter two species were 45.5 and 40, respectively. Our analysis suggests that it would also be interesting to explore the generic placement of this species relative to Arctoraja species.

Bathyraja maculata (whiteblotched skate) (fig. 76)

The analysis included four specimens of this Bering Sea endemic. The range of pairwise differences among these specimens was 1–5, with an average of 3.8.

Brochiraja asperula (prickly deepsea skate) complex (fig. 77)

The analysis yielded a group that included two clusters, one of which was comprised of 19 specimens of the New Zealand endemic, Brochiraja asperula; 13 of these came from the Museum of New Zealand, Te Papa Tongarewa (GN6803  =  NMNZ P.040505, GN6804  =  NMNZ P.040506, GN6809  =  NMNZ P.040512, GN6811  =  NMNZ P.040514, GN6812  =  NMNZ P.040515, GN6820  =  NMNZ P.041493, GN6818  =  NMNZ P.041491, GN6795  =  NMNZ P.040465, GN6798  =  NMNZ P.040467, GN6801  =  NMNZ P.040469, GN6802  =  NMNZ P.040470, GN6821  =  NMNZ P.041494, and GN6817  =  NMNZ P.041490). These specimens were collected from a diversity of localities throughout the waters of New Zealand, including the Chatham Islands, Chatham Rise, Campbell Plateau, and the South Island. The range of pairwise differences among the 19 specimens in this cluster was 0–2, with an average of 0.3. However, the analysis yielded a second smaller cluster in this group comprised of two specimens also in the National Museum of New Zealand, Te Papa Tongarewa (GN6807  =  NMNZ P.040510 and GN6808  =  NMNZ P.040511), collected off Westland on New Zealand's South Island. The sequences of these two specimens differed from one another by a single base and were somewhat divergent relative to those comprising the specimens of the first cluster. The average of the pairwise differences between specimens of these two clusters was 27.7. The specimens in the smaller cluster have been referred to as Brochiraja cf. asperula.

Brochiraja spinifera (spiny deepsea skate) (fig. 77)

Sixteen specimens of this New Zealand endemic, all collected from localities in and around New Zealand (e.g., North Island, South Island, Stewart Island, Campbell Island, Auckland Islands, Chatham Rise), were found to comprise essentially a single cluster. A total of 12 of these are from the Museum of New Zealand, Te Papa Tongarewa (GN6794  =  NMNZ P.040464, GN6826  =  NMNZ P.041752, GN6822  =  NMNZ P.041748, GN6827  =  NMNZ P.041753, GN6805  =  NMNZ P.040508, GN6836  =  NMNZ P.045193, GN6806  =  NMNZ P.040509, GN6810  =  NMNZ P.040513, GN6823  =  NMNZ P.041749, GN6824  =  NMNZ P.041750, GN6797  =  NMNZ P.040466, and GN6828  =  NMNZ P.041754). The range of pairwise differences among the 16 specimens in this cluster was 0–13, with an average of 6.1. While there was some evidence of substructure within this cluster, this remains to be explored in more detail. These specimens have all been referred to here as B. spinifera, but this identity needs to be confirmed. This cluster grouped most closely with the specimens of B. asperula and B. cf. asperula. The average of the pairwise differences between specimens of B. spinifera and those of B. asperula was 25.3, and those of B. cf. asperula 31.1.

Brochiraja leviveneta (smooth blue skate) (fig. 77)

Three specimens of this recently described New Zealand endemic (see Last and McEachran, 2006) were included in the analysis. All three specimens are deposited in the Museum of New Zealand, Te Papa Tongarewa (GN6835  =  NMNZ P.045192, GN6815  =  NMNZ P.040676, and GN6830  =  NMNZ P.041985). The range of pairwise differences among these three specimens was 4–12, with an average of 9.3.

Brochiraja albilabiata (whitelipped skate) (fig. 77)

This relatively newly described New Zealand endemic (see Last and McEachran, 2006) was represented by a single specimen from the Museum of New Zealand, Te Papa Tongarewa (GN6833  =  NMNZ P.042691). This specimen grouped most closely with the specimens of B. leviveneta. The average of the pairwise differences between specimens of these two species was 28.7.

Insentiraja subtilispinosa (velvet skate) (fig. 77)

A specimen of this species from the Australian National Fish Collection (GN4629  =  ANFC H 6417-03), collected from Western Australia, was included in the analysis. This specimen clustered outside those of the Brochiraja species. The generic placement of this species follows Last and Stevens (2009).

Notoraja azurea (blue skate) (fig. 77)

A paratype (GN4631  =  ANFC H 6409-02) of this recently described Tasmanian endemic (see McEachran and Last, 2008) was included in the analysis. This specimen clustered with the specimen of Insentiraja subtilispinosa, but the sequences of these two species differed by 34 bases.

Brochiraja microspinifera (small prickly skate) (fig. 77)

The single specimen of this relatively recently described New Zealand endemic (see Last and McEachran, 2006) was included in the analysis. This specimen is in the Museum of New Zealand, Te Papa Tongarewa (GN6816  =  NMNZ P.041321). It was found to belong to a cluster consisting of the other Brochiraja species, but also of the specimens of Insentiraja subtilispinosa and Notoraja azurea.

Pavoraja nitida (peacock skate) (fig. 77)

The analysis included three specimens of this Australian endemic species. These specimens differed from one another by 0–4 bases, with an average of pairwise differences of 2.7.

Pavoraja alleni (Allens skate) (fig. 77)

This Australian endemic was represented by a single specimen from the Australian National Fish Collection (GN4632  =  ANFC H 6419-03). This specimen clustered with those of its congener, P. nitida. The average of the pairwise differences between specimens of these two species was 53.7.

Irolita waitii (southern round skate) (fig. 77)

The analysis included a single specimen of this southwestern Australian endemic species (GN4630  =  ANFC H 6350-01). This sample clustered along with, but outside all of the specimens representing species of Brochiraja, Insentiraja, Notoraja, and Pavoraja.

Psammobatis sp. (fig. 77)

In total, 10 specimens of an as yet unidentified species of Psammobatis, all collected from the Falkland Islands, were included in the analysis. They were found to comprise a single cluster and the range of pairwise differences among these specimens was 0–6, with an average of 2.2. It is likely this represents one of the seven Psammobatis species occurring in the western South Atlantic.

Sympterygia bonapartii (smallnose fanskate) (fig. 77)

Two specimens of this southern South American endemic were included in the analysis. Both specimens came from Argentina and were identical in sequence.

Arhynchobatis asperrimus (longtailed skate) (fig. 77)

One specimen of this New Zealand endemic was included in the analysis. This specimen is in the Museum of New Zealand, Te Papa Tongarewa (GN6831  =  NMNZ P.042403). It grouped most closely with the specimens of S. bonapartii and the average of the pairwise differences between specimens of these two species was 130.

Sympterygia acuta (bignose fanskate) (fig. 77)

The two specimens of this southeastern South American species included in the analysis were both collected from Argentina, and are deposited in the Instituto Nacional de Investigación y Desarrollo Pesquero (GN4726  =  INIDEP T 0405 and GN4727  =  INIDEP T 0407). These differed from one another by six bases. They grouped most closely with but outside the specimens of S. bonapartii and A. asperrimus, which suggests that the distinctions between Sympterygia and the currently monotypic Arhynchobatis needs to be explored in more detail. The average of the pairwise differences between S. acuta and S. bonapartii was 129.5; the average of the pairwise differences between S. acuta and A. asperrimus was 137.

Atlantoraja cyclophora (eyespot skate) (fig. 77)

Three specimens of this species, which is known from Argentina and Brazil, were included in the analysis. These were all collected from Argentina and are deposited in the Instituto Nacional de Investigación y Desarrollo Pesquero (GN4717  =  INIDEP T 0469, GN4718  =  INIDEP T 0471, and GN4719  =  INIDEP T 0474). They were found to comprise a single cluster; the range of pairwise differences among these specimens was 1–4, with an average of 2.7.

Atlantoraja platana (La Plata skate) (fig. 77)

The analysis included two specimens of this southwestern Atlantic species, both collected from Argentina and both from the Instituto Nacional de Investigación y Desarrollo Pesquero (GN4721  =  INIDEP T 0425 and GN4722  =  INIDEP T 0426). They differed from one another by one base. They grouped most closely with the specimens of A. cyclophora and the average of the pairwise differences between these two species was 38.8.

Atlantoraja castelnaui (spotback skate) (fig. 77)

The analysis included a single specimen of this southeastern South American species (GN4720  =  INIDEP T 0406). It clustered along with but outside the specimens of its two congeners. The average of the pairwise differences between A. castalnaui and A. cyclophora was 90, and the average of the pairwise differences between A castalnaui and A. platana was 89.5.

Rioraja agassizii (Rio skate) (fig. 77)

This monotypic genus was represented by two specimens in the analysis, both collected from Argentina and deposited in the Instituto Nacional de Investigación y Desarrollo Pesquero (GN4724  =  INIDEP T 0404 and GN4725  =  INIDEP T 0408). These specimens thus represent a central element of the distribution of this species, which has been reported from throughout much of the eastern coast of South America. These specimens differed from one another by 10 bases and they grouped most closely with the three species of Atlantoraja.

Anacanthobatidae (Legskates)

Cruriraja hulleyi (roughnose legskate) (fig. 77)

Ten specimens of this southern African endemic, all collected from South Africa, were included in the analysis. The range in pairwise differences among specimens in this cluster was 0–8, with an average of 1.9.

Sinobatis bulbicauda (western legskate) (fig. 77)

Both specimens of this newly described skate (see Last and Séret, 2008) included here were collected from localities in Western Australia. One was a paratype (GN6775  =  ANFC H 6417-04), the other a voucher (GN6776  =  ANFC H 6578-02). These differed from one another by a single base.

DISCUSSION

The foregoing results give a first approximation of intraspecific and intrageneric differentiation seen across a broad spectrum of elasmobranchs. When these data are combined with other sources of information (morphological, distributional, ecological, and tagging data) they can enrich our understanding of species ranges, life histories, gene flow, and the incidence of philopatry, and thereby give us a better sense of the operational evolutionary units that characterize extant diversity. There are, for example, species such as Rhincodon typus that appear genetically homogeneous for NADH2 across their entire range (see fig. 37), while others, like Carcharodon carcharias, appear distinct in different parts of their range (see fig. 37). We might hypothesize that those that are genetically homogeneous are likely to exhibit more movement and gene flow across their range than those with regionally distinct populations. Tagging data can be invaluable to test such hypotheses.

Intraspecific p-distances (average of % pairwise differences)

The average of the intraspecific p-distances for the 468 species for which replicate specimens were included was 0.27% ± 0.28%; the range was 0%–2.12% (table 2). The distribution of mean % p-distances for these 468 species is illustrated in fig. 78. In terms of base pairs, on average, specimens of a species differed by 2.86 ± 2.93 (0–22) base pairs. For comparative purposes, the overall average of the intraspecific K2P distance for 468 taxa, as calculated in BOLD, was 0.27% ± 0.001%; with a range of 0%–3.14% (table 5).

Fig. 78.

Distribution of mean p-distances (%) for NADH2; within species (black); within genera (grey).s

Fig. 79.

Parsimony haplotype network for Carcharhinus sealei, C. cf. sealei, C. dussumieri, and C. cf. dussumieri color coded by phenotype (A) and geography (B). For this illustration and all successive illustrations, circle diameter corresponds to haplotype frequency; number of base pair differences between haplotypes is indicated in red; only differences ≥3 bp are indicated.

Fig. 80.

Parsimony haplotype network for Carcharhinus obscurus and C. galapagensis color coded by phenotype (A) and geography (B).

Fig. 81.

Parsimony haplotype network for Carcharhinus sorrah and C. cf. sorrah color coded by phenotype (A) and geography (B).

Fig. 82.

Parsimony haplotype network for Carcharhinus limbatus, C. cf. limbatus, C. amblyrhynchoides, and C. tilstoni color coded by phenotype (A) and geography (B).

Fig. 83.

Parsimony haplotype network for Carcharhinus leucas, C. cf. leucas 1, C. cf. leucas 2, C. amboinensis 1, and C. amboinensis 2 color coded by phenotype (A) and geography (B).

Fig. 84.

Parsimony haplotype network for Carcharhinus plumbeus, C. cf. plumbeus, and C. altimus color coded by phenotype (A) and geography (B).

Fig. 85.

Parsimony haplotype network for Rhizoprionodon acutus, R. cf. acutus 1, R. cf. acutus 2, and R. cf. acutus 3 color coded by phenotype (A) and geography (B).

Fig. 86.

Parsimony haplotype network for Scoliodon macrorhynchus, S. laticaudus, and S. cf. laticaudus color coded by phenotype (A) and geography (B).

Fig. 87.

Parsimony haplotype network for species of Sphyrindae color coded by phenotype (A) and geography (B).

Fig. 88.

Parsimony haplotype network for Galeocerdo cuvier and G. cf. cuvier color coded by phenotype (A) and geography (B).

Fig. 89.

Parsimony haplotype network for Iago omanensis, I. cf. omanensis 1, I. cf. omanensis 2, and I. garricki color coded by phenotype (A) and geography (B).

Fig. 90.

Parsimony haplotype network for Poroderma pantherimum and P. africanum color coded by phenotype.

Fig. 91.

Parsimony haplotype network for Isurus oxyrinchus color coded by phenotype (A) and geography (B).

Fig. 92.

Parsimony haplotype network for Chiloscyllium punctatum and C. cf. punctatum color coded by phenotype (A) and geography (B).

Fig. 93.

Parsimony haplotype network for Ginglymostoma cirratum and G. cf. cirratum color coded by phenotype (A) and geography (B).

Fig. 94.

Parsimony haplotype network for Himantura cf. gerrardi 1, 2, 3, 4, and 5 color coded by phenotype (A) and geography (B).

Fig. 95.

Parsimony haplotype network for Himantura uarnak 1, 2, 3, and 4, and H. undulata and H. leoparda color coded by phenotype (A) and geography (B).

Fig. 96.

Parsimony haplotype network for species of Pastinachus color coded by phenotype (A) and geography (B).

Fig. 97.

Parsimony haplotype network for Neotrygon kuhlii 1, 2, 3, and 4, and N. picta and N. cf. ningalooensis color coded by phenotype (A) and geography (B).

Fig. 98.

Parsimony haplotype network for Taeniura lymma 1 and 2 color coded by phenotype (A) and geography (B).

Fig. 99.

Parsimony haplotype network for species of Rhinoptera color coded by phenotype (A) and geography (B).

Fig. 100.

Parsimony haplotype network for species of Aetomylaeus color coded by phenotype (A) and geography (B).

Fig. 101.

Parsimony haplotype network for species of Aetobatus color coded by phenotype (A) and geography (B).

Fig. 102.

Parsimony haplotype network for species of Gymnura color coded by phenotype (A) and geography (B).

Of the 468 species for which greater than a single specimen was included in the analysis, replicate specimens of 60 species were identical in sequence (i.e., there was no intraspecific variation). Although 25 of these 60 species were represented by only two specimens and 19 by only three specimens, 16 were represented by four to 12 specimens. Those with 10 or more specimens were the western Australian endemic Orectolobus hutchinsi (10 specimens), as well as Dasyatis margarita (11 specimens) and Squatina oculata (12 specimens), the latter two represented by specimens from the eastern Atlantic elements of their distributions.

The mean p-distances calculated as the average of all pairwise comparisons among specimens within each recognized species (appendix 1) allow us to make the following observations. The most variable species (i.e., those with average p-distances of >1%) were Brachaelurus colcloughi (mean p  =  1.01%), Isurus oxyrinchus (mean p  =  1.02%), Aetobatus flagellum (p  =  1.05%), Hexanchus nakamurai (mean p  =  1.10%), Carcharhinus albimarginatus (mean p  =  1.15%), Dipturus sp. 4 (p  =  1.25%), Scyliorhinus retifer (mean p  =  1.41%), Dasyatis sabina (p  =  1.44%), Bathyraja pallida (mean p  =  1.66%), Dasyatis americana (mean p  =  2.08%), and Torpedo sinuspersici (p  =  2.12%). This result does not appear to be tied to the number of specimens examined because all but one of these species were represented by only 2–5 specimens. Isurus oxyrinchus was the exception in that 24 specimens were included in the analysis. In some cases the relatively substantial divergence seen among specimens may be attributable to geographic variation. For example, the three specimens of H. nakamurai came from Australia, Madagascar, and India; the 24 specimens of I. oxyrinchus came from the Atlantic, Pacific, and Indian Ocean basins; one specimen of A. flagellum came from Borneo, the other from India; the three specimens of S. retifer were collected from mid-Atlantic bight, the northwestern Atlantic and the Gulf of Mexico; the four specimens of B. pallida were collected from the Barents Sea, mid-Atlantic, and the western Atlantic off New England. However, in the cases of other species, the replicate specimens were collected from locations of relatively close proximity, and thus an alternate explanation for the divergence seen may need to be invoked. For example, all four specimens of B. colcloughi came from Queensland, both specimens of Dipturus sp. 4 were collected in the Philippines, both specimens of D. sabina came from the Gulf of Mexico, as did all three specimens of D. americana; both specimens of T. sinuspersici were collected in the Persian Gulf, and the five specimens of C. albimarginatus came from Taiwan and the Philippines. It is of further note that in the majority of these cases, the relatively substantial divergence seen among specimens of a species was the result of a single divergent specimen.

OUTLOOK AND FUTURE WORK

AUTHOR CONTRIBUTIONS

This project was originally conceived by J.N. Caira and K. Jensen in collaboration with G.J.P. Naylor as a result of discussions of how best to validate and verify host identifications for parasitological studies. G.J.P. Naylor was responsible for expanding its scope to include representation of as broad a spectrum of elasmobranchs as possible to serve the needs of both parasitologists and elasmobranch biologists. G.J.P. Naylor oversaw the sequencing and initial analysis of the distance data and wrote sections of the Introduction and sections of the Discussion. J.N. Caira (in close collaboration with K. Jensen) wrote the results section as well as portions of the Introduction and Discussion; she also supervised the design and population of the host specimen database. K. Jensen (in close collaboration with J.N. Caira) generated the distance data, and prepared the figures, tables, and appendices; she also took the majority of the photographs of elasmobranchs presented in the host specimen database. J.N. Caira and K. Jensen collected 53% of the tissue samples; G.J.P. Naylor collected and/or solicited tissues for the remainder. K.A.M. Rosana carried out 90% of the laboratory work involved in generating the sequence data. W.T. White and P.R. Last contributed key tissue samples from rare taxa and provided valuable input on the species treatments.

ACKNOWLEDGMENTS

Elizabeth Barbeau played an invaluable role in populating and assisting with the design of the online host specimen database. Clemens Lakner wrote scripts to automate the alignment process integrating scores from Phred base-calling software with amino acid translation and the MUSCLE alignment package. We are grateful to Dave Ebert and Marcelo de Carvalho for their assistance with identifications of particularly problematic genera. Some of the sequences were generated by Julie Ryburn, Rebecca Wetherby, and Britanny Burdelsky during their tenures as technicians in the lab of G.J.P.N. Other sequences were generated by Andres Lopez, Olivier Fedrigo, Vicente Faria, and Toni Ferarra during their time as graduate students in G.J.P.N.'s lab. Kendra Koch assisted with the generation of p-distance matrices. We thank Dave Swofford for providing access to development versions of his PAUP* software pakage (v40b106). We are extremely grateful to Dave Ebert, Ed Heist, and Kevin Feldheim for their extensive and thoughtful comments and recommended revisions, which led to substantial improvement of this manuscript.

Many people and institutions have been involved in collecting specimens and tissues for this project. We would particularly like to thank Larry and Ken Abele, Mohammed Ahmed, Hassan Ahmed, Alexandre Aires da Silva, Moonyeen Alava, Capt. Walter Allyn, Zafira Almeida, Salem Alyen, Aric, Greta Arie, Jose Azevado, Scott Bachman, Julian Baggio, Roger Bakey, Avi Baranes, Simon Beatson, Andrew Bentley, George Benz, T. Berden, Dana Bethea, Rob Bilby, Albert Buckley, Sr., Stephen Bullard, George Burgess, Loren Caira, Hilconida P. Calumpong, Juan Carvajal, Jack Casey, Jose Castro, Rachel Cavanagh, Gilles Charmillon, Patricia Charvet-Almedia, Justin Chidlow, Kerin Claeson, Eugenie Clark, Malcom Clark, Geremy Cliff, Nick Colyanis, April Cook, Brian Cooksley, Rob Cooper, Graca Costa, Chip Cotton, David Bay Courtney, Paul Cowley, Rold Czabayski, Juan M. Diaz de Astarloa, Eugene de Bruins, Don de Maria, Dharmadi, Kathy Dickson, John Dindo, Jaques Ngor Diouf, Vince Dodge, Matt Doty, Henriette Dubuit, Sheldon Dudley, Clinton Duffy, Dave Ebert, Scott Eckert, M. D. Ehrlich, Nick Elliot, Tim Enslow, Asni and Masni Etin, Salik Etin, Fahmi, Tracey Fairweather, Vicente Faria, Olivier Fedrigo, Iain Field, Rodney Fox, Mafalda Freitas, Kathy Frost, John Galbraith, J. Gaudiano, Jenny Giles, Daniel Gledhill, John Gordon, Alistair Graham, Ken Graham, Dean Grubbs, Mervyn Guevarra, Julian Hall, Nils Roar Hareide, Mark Harris, Mohamad Haseli, David Hayashi, Claire Healy, Ed Heist, Hamri bin Hgi-Kating, Barry Hitchins, Eric Hoffmayer, Age Hoines, Lisa Hollensead, Dave Holt, David Holts, Steve Hornby, Marie Horton, Janet Houston, Gordon Hubbell, Brett Human, Charlie Huveneers, Samuel Iglesias, Eric Johnson, Lisa Jones, Laura Jordan, Shoou-Jeng Joung, Stephen Kajiura, Devon Keeney Adrian Kitchingman, Brian, Lana, and Barbara Koenneke, Nancy Kohler, Peter Kyne, R. Lamothe-Argumedo, Rob Leslie, Rosangela Lessa, Annie Lim, Julie Lloyd, Andres Lopez, Richard Lord, Nathan Lovejoy, Lloyd Lowry, Martin Lucero, Hamish Macgibbon, Mabel Manjaji-Matsumoto, Masoumeh Malek, Loghman Maleki, Mark Marks, Andrea Marshall, Andrew Martin, R. McCauley, John McEachran, Pamela McGarr, R. Means, Gui Menezes, Toru Miyashita, Alec Moore, Jon Moore, J. Moyse, Catherine Morrish, John Morrissey, Richard Mounsey, William Mure, Jack Musick and Virginia Institute of Marine Science team, Scott Mycock, Kazuhiro Nakaya, Lisa Natanson, Geoff Oke, Peter Olson, Simon O'Reilly, James Orr, Cathy Painter, Brian Potvin, Roger Parker, Maria Pickering, Ted Pietsch, Joost Pompert, Steve Poston, R. Potter, Wes Pratt, Chris Reid, Florian Reyda, Getulio Rincon, Heidi Robeck, Tyson Roberts, Buddy Robey, Dawn Roje, Maria Ines Roldan, Norman and Eric Sanders, Ivan Schultz, Frank Schwartz, Latip Sait, Ed Schellenburger, M. Scott, Richard Seager, Bernard Seret, Russell Shortridge, Colin Simpendorfer, David Sims, Greg Skomal, Peter Smith, John Smythe, Lyle Squire, Julia Stace, John Stevens, Michael Stevenson, Duane Stevenson, Andrew Stewart, Ait Sulloman, Sho Tanaka, Steven Taylor, James Teater, Christine Thacker, J. Thornton, Di Tracey, Binh Tih Tran, Ted Turton, Gaines Tyler, Mugi Utomo, David Weller, Dana West, G. West, Warren Williams, Jamie Williamson, Andrew Williston, Sabine Wintner, Kazunari Yano, Gordon Yearsley, Kara Yopak, Forrest Young and the staff of Dynasty Marine, Diana Zaera-Perez, and Jiang Born Zen. In addition, we wish to thank the captains and crews of the FRS Africana, FV Amaltal Explorer, FV Amaltal Mariner, FV Clarabelle, FV Christmas Creek, FV K.K. Manchong, FV Ocean Harvest, FV San Waitaki, FV Seamount Explorer, FV Thomas Harrison, FV Sea Dog, RV Matthew Melissa, RV Tangaroa, as well as the various captains and crews of the NMFS northwest Atlantic shark long-lining survey crews from 1986–2002. We are also grateful to the Suao and Tahsi Fishery in Taiwan, Ningaloo Tagging Team, Sierra Fisheries (Sierra Leone), fishermen at Arena Blanco in Mindanao the fisherman of Dakar, Senegal, and the Jacksonville Jaycees.

Finally, we are grateful to the staff of the following museums for their willingness to provide tissues, specimen numbers and/or images: American Museum of Natural History, New York; Australian Museum, Sydney; Burke Museum, Washington; Cairns Marine in Cairns, Australia; CSIRO Australian National Fish Collection, Hobart, Tasmania; California Academy of Sciences, San Franciso; De Jong Marinelife in B.V., KG Spijk (Gem Lingewaal), Netherlands; Havforskningsinstituttet Institute of Marine Research in Bergen, Norway; Hokkaido University Museum, Sapporo, Japan; Instituto Nacional de Investigación y Desarrollo Pesquero, Mar del Plata, Argentina; Instituto de Biologia, Universidad Nacional Autonóma de México, Mexico City; Universiti Malaysia Sabah, Kota Kinabalu, Malaysia; IPPS Institut Penyelidikan Perikanan Sarawak, Kuching, Malaysia; Kagoshima University Museum, Kagoshima, Japan; University of Kansas Ichthyology Collection, Lawrence, Kansas; Los Angeles County Museum, Los Angeles, California; MMF Museu de História Natural e Aquário, Funchal, Madeira; Museum of New Zealand, Te Papa Tongarewa, Wellington, New Zealand; Museum Zoologicum Bogoriense, Bogor-Cibinong, Indonesia; Museum of Comparative Zoology–Harvard, Cambridge, Massachusetts; Northern Territories Museum, Darwin, Northern Territories, Australia; Royal Ontario Museum, Toronto, Canada; Zoology Department of the Sabah State Museum, Kota Kinabalu, Malaysia; Texas Cooperative Wildlife Collection, College Station, Texas; Tulane University Museum of Natural History, New Orleans, Louisiana; University of Florida Fish Collection, Gainesville, Florida; University of Michigan Museum of Zoology, Ann Arbor, Michigan; University of Washington Fish Collection, Seattle, Washington; Virginia Institute of Marine Sciences, Gloucester Point, Virginia; Vietnam Natural Museum of Nature, Hanoi; Western Australian Museum, Welshpool, Western Australia; Yale Peabody Museum, New Haven, Connecticut.

The more than 20 years of fieldwork underlying the majority of the tissue collection and DNA sequencing involved in this project has been supported by NSF grants DEB 8708121, DEB 9707145, DEB 0089533, DEB 0415486, DEB 1036500 to GJPN, DEB 9300796, DEB 9521943, DEB 0118882, DEB 0103640, DEB 0542846, DEB 0818696 to J.N.C and DEB 0542941, DEB 0818823, Kansas NSF EPSCoR First Award to K.J.

TABLE 5

Comparison of intraspecific and intrageneric K2P distances (%) for elasmobranchs between NADH2 (current study) and CO1 (Ward et al., 2008) sequences