Introduction

The purpose of this account is to establish that the Australasian shark Heptranchias dakini Whitley, 1931, is conspecific with H. perlo (Bonnaterre, 1788), the widely distributed Sharpsnouted Sevengill shark (alternatively named the Slender Sevengill shark or the Perlon shark). The main reported difference between these nominal species—the relative positions of the dorsal and anal fins—can be shown to be due to sexual dimorphism, and other listed differences are apparently derived from an inaccurate illustration of a single Australian specimen or cannot be substantiated.

Previous Accounts

Whitley (1931, p. 310) proposed the new species Heptranchias dakini on the basis of an earlier account (McCulloch, 1911, p. 2) of seven specimens from Victoria, Australia, which McCulloch identified as H. perlo. Whitley noted that the specimen in McCulloch's figure "appears to differ in having the head 4½ in total length and anal originating below middle of dorsal, with its base shorter than that of dorsal", and he selected as holotype of H. dakini the specimen figured by McCulloch (1911, pl. 1, fig. 1).

No definitive account of H. dakini has appeared since the original description. In fact, there have apparently never been any measurements taken from a specimen; those of Whitley (1931) were, as far as we can determine (see below), taken straight from McCulloch's illustration. Although this illustration has been republished several times, no new illustrations of the species have appeared.

Several Australian workers have referred briefly to the Sharpsnouted Sevengill shark. Whitley (1940, 1968), Munro (1956), Scott (1962) and Lynch (1963) treated it as H. dakini, while Stead (1963) treated it as H. perlo, stating, without supporting evidence, that "Whitley's Heptranchias dakini is identical with this species".

Heptranchias was first recorded from New Zealand in 1953 from one specimen (DM 2745) which was referred to H. dakini by McCann (1953) and to H. perlo (mis-spelt parlo in title) by Parrott (1953). The same specimen was discussed in Parrott (1958). No further specimens are recorded in the literature, although several were received and catalogued by the Dominion and Canterbury Museums from 1953 onwards. Whitley (1956, 1968) listed the New Zealand species as H. dakini. Richardson and Garrick (1953) mentioned that both dakini and perlo might occur in the New Zealand area.

Bigelow and Schroeder (1948) listed the two species as distinct, but with some reservations, viz. (p. 92): Heptranchias perlo "is represented in Australian waters by a relative (dakini) so close to perlo that it might finally prove identical". Garrick and Schultz (1963) implied in their account of H. perlo— "almost worldwide in distribution", with no mention of a related species H. dakini—that only one species existed.

Early brief accounts and listings of Heptranchias indicus, Notidanus indicus and Notorynchus indicus, sometimes named as the "Perlon Shark", in New Zealand literature (e.g. Hutton, 1872: 79; 1904: 54; Sherrin, 1886: 121; Waite, 1907: 6; Hamilton, 1908: 30; Thomson, 1913: 235; Thomson and Anderton, 1921: 68) are inadequate and confusing. We suspect, from the context of the brief accounts, that they refer to the more commonly caught Broadsnouted Sevengill Notorynchus. We have not checked the early Australian literature, but the same situation may well occur there.

Present Study

One of us (L.J.P.) observed that two female Heptranchias trawled in New Zealand waters during 1969 had the anal fin origin beneath the middle of the dorsal fin base, while a male also trawled in 1969 had the anal fin origin below the posterior end of the anal base (Plate 1). The females agreed, therefore, with one of Whitley's criteria for dakini, while the male fitted perlo as diagnosed and figured from a Cuban specimen in Bigelow and Schroeder (1948, p. 88, fig. 10).

The other of us (J.A.F.G.), who was studying in the Australian Museum at the time, took the opportunity of re-examining the material of Heptranchias in that institution, including the types of H. dakini. We later examined New Zealand Heptranchias material in the Dominion Museum collection.

Although Heptranchias was not recorded from New Zealand until 1953, and has generally been regarded as rare, we found that there were 14 specimens available, including the three 1969 specimens mentioned above. In addition, we had measurements for two other New Zealand specimens examined by one of us (J.A.F.G.) several years ago in the Canterbury Museum, Christchurch.

Information from these 16 New Zealand specimens, plus five Australian specimens examined in the Australian Museum, plus the account in Bigelow and page 3Schroeder (1948) of Atlantic specimens, was utilised in arriving at the present conclusion that only one species is involved. Details of the material examined are as follows.

Material Examined

Australia: AMS I.10825, female, 695 mm (possible holotype of H. dakini Whitley, 1931), Victorian coast, 1910, Endeavour, AMS IB.3277, female, 720 mm, New South Wales, off Barrenjoey, Jan. 1955, J. E. Smith; AMS I.13929, female, 778 mm, New South Wales, N.E. of Cape Solander, 56 fathoms, 1916, State Trawling Dept.; AMS I.10795, female, 822 mm (supposed holotype of H. dakini Whitley, 1931), Victorian coast, 60 miles S. of Cape Everard, 60–70 fathoms, 1910, Endeavour. AMS I.10794, male, 845 mm (paratype of H. dakini Whitley, 1931), same data as AMS I.10795.

New Zealand: DM 4951, female, 805 mm, Doubtless Bay, 80 fathoms, Feb. 2, 1969, L. J. Paul; DM 4952, male, 856 mm, between Plate Island and White Island, 65–110 fathoms, Feb. 9, 1969, L. J. Paul; DM 1947, male 857 mm, Castlepoint, 40–50 fathoms, Jul. 11, 1956, F. Abernethy; DM 3103, male, 860 mm, off Plate Island, 124–190 fathoms, Feb. 1961, W. Sampson; DM 3023, female, 964 mm, Palliser Bay, 15 fathoms, Mar. 15, 1954, F. Abernethy; DM 3022, female, 1016 mm, off Te Awaiti, 40–70 fathoms, Jul. 8, 1955, F. Abernethy; DM 1848, four females, 1044 mm, 1072 mm, 1078 mm and 1305 mm, Hawke Bay, 180 fathoms, Dec. 16, 1955, J. A. F. Garrick; DM 4953, female, 1100 mm, N.W. of Cape Farewell, 200 fathoms, Jul. 30, 1969, L. J. Paul; CM (no number), female, 1100 mm, probably from Kaikoura; DM 1870, female, 1104 mm, Kaikoura, off Oaro, 100 fathoms, Feb. 14, 1956, R. Baxter; DM 2180, female, 1340 mm, Palliser Bay, 30 fathoms, Mar. 14, 1957, J. Slater; CM (no number), female, 1365 mm, Kaikoura, off Oaro, 180 fathoms, Feb. 1954, R. Baxter; also DM 2745, female, damaged and with some fins removed, not measured by us because of this, but length given by Parrott (1953) as 52 in. (1320 mm), Tasman Bay, off Cape Susie (presumably Cape Soucis), Apr. 18, 1950, J. Cotton.

Atlantic: Measurements from Bigelow and Schroeder (1948, p. 89) of MC2 35897, male 698 mm and female 932 mm, Cuba.

Type Specimens of Heptranchias dakini

The annotations regarding holotypes of H. dakini in the list of material above stem from conflicting data in Whitley (1931, 1957) regarding type material of this species. In the original account of H. dakini, Whitley (1931, p. 310) stated that "The holotype of H. dakini is the specimen from sixty miles south of Cape Everard, Victoria, figured by McCulloch." McCulloch's illustration (1911, pl. 1, fig. 1) is undoubtedly of a female, although it is not labelled as such, and the illustration is stated to be two-sevenths natural size. The total length in the illustration is 200 mm, which means the specimen was about 700 mm long. Subsequently, Whitley (1957, p. 1) listed as "Types" of H. dakini two specimens in the Australian Museum, a male of 845 mm (AMS I.10794) and a female of 822 mm (AMS I.10795). The male cannot be the holotype, because Whitley (1931) nominated McCulloch's figured female as such. The female is thus the supposed holotype—and this supposition is supported by data from the file-card catalogue in the Fish Department of the Australian Museum. However, this female, 822 mm long, appears to be much too long—unless McCulloch's statement of "two-sevenths page 4natural size" was a very general approximation. If neither of these two specimens is the holotype, the possibility remains that a third specimen in the Australian Museum (AMS I.10825), a female of 695 mm labelled "Victorian Coast. 1910. Commonwealth Fish Bureau. Endeavour," was one of the seven specimens, all from the same haul, on which McCulloch based his account, but we are not able to establish this. The size of this female closely accords with our calculated size for the shark figured by McCulloch. Consequently we treat this specimen as the "possible holotype" in our list of material examined and in Tables 1 and 2. We do not know the fate of the remainder of McCulloch's specimens.

Identification of Heptranchias dakini as H. perlo

Whitley (1931, p. 310) gave three diagnostic characters for H. dakini. The first of these was that the head was "4 ½ in total length", but Whitley did not say how this differed from the situation in H. perlo. From our data (Table 1) we find no obvious difference in head length (measured to the posterior end of the 7th gill-opening) relative to total length between Australian, New Zealand and Atlantic Heptranchias. Nor do we find any specimens in which this proportion is 4½. The range for three Australian specimens, including the two "Types" of dakini, is 5.0 to 5.5 (mean 5.3); for 15 New Zealand specimens 4.8 to 5.6 (mean 5.2); and for two Cuban specimens the almost equivalent prepectoral lengths, as given by Bigelow and Schroeder, equal 4.8 and 5.2 in total length. We note that McCulloch's illustration shows a head length of about 4.5 in total length and assume that this was the basis of Whitley's statement.

Whitley's second diagnostic character for H. dakini was that the anal fin originates below the middle of the dorsal fin base. We have examined this character and find that it occurs only in females; it is present in New Zealand specimens and in a Cuban specimen of H. perlo as well as in Australian specimens. Males on the other hand have the anal fin origin further rearward, below the posterior end of the dorsal base or at least nearer to that level than to the middle of the dorsal base. There is some overlap between the sexes, with a minority of the females approaching or overlapping the male condition. Table 1 shows for each of our specimens the proportion (%) of the dorsal fin base which is anterior to the anal fin origin. In 13 of the 17 females there is from 50.0% to 71.7% of the dorsal base anterior to the anal fin origin. In other words, in these 13 females the anal fin origin is nearer to the middle of the dorsal base than to the posterior end. In the remaining four females there is from 75.4% to 100.0% of the dorsal base anterior; thus their anal fin origin is nearer to the posterior end of the dorsal fin base than to the middle. The condition in these four females is similar to that of the five males in which from 76.3% to 102.8% of the dorsal base is anterior.

It is clear from the above data that the position of the anal fin relative to the dorsal fin in Heptranchias predominantly reflects sexual dimorphism and offers no evidence for distinguishing a separate Australian species, H. dakini, from the world-wide H. perlo.

It is, perhaps, surprising that this possibility did not occur to Whitley when he described H. dakini, insofar as the "Types" of H. dakini which he designated (Whitley, 1957) included a male as well as a female specimen—and these particular specimens differ strongly in this feature.

Whitley's third diagnostic criterion for H. dakini was that the anal base is shorter than the dorsal base. We find that this is true for the Australian H. dakini material, but it also applies to our New Zealand specimens, as well as to the page 5Atlantic specimens of H. perlo described in Bigelow and Schroeder (1948, p. 89). There seems, therefore, to be no substance in this criterion for recognising H. dakini.

On the basis of the above data we regard H. dakini as conspecific with H. perlo.

Because there are relatively few published measurements of H. perlo we provide here (Table 2) measurements of five New Zealand specimens and three Australian specimens, plus two Cuban specimens (from Bigelow and Schroeder, 1948).

In the same table we give vertebral counts of six Australian and New Zealand specimens, which range from 90–93 precaudal and 52–59 caudal centra. Such counts closely agree with those published in Springer and Garrick (1964, p.83) for three western North Atlantic specimens (89–90 precaudal, 52–61 caudal) of H. perlo and provide supporting evidence that only one species is involved.

Sexual Dimorphism in Heptranchias perlo

In order to provide further information on the nature of sexual dimorphism in H. perlo we have listed in Table 1 various proportional dimensions which appear to be significant in this respect. Because of the small number of specimens involved, and the scarcity of males, we interpret the differences between the sexes with caution.

Expressed as a percentage of total length, the average predorsal length of females is slightly longer (0.8%) than that of males. Conversely, the average distance from snout tip to anal fin origin is slightly shorter (0.7% t.l.) in females than in males. The greater proximity of the dorsal and anal fins in the females compared to the males seems due, therefore, to equal differences in the positions of both of these fins between the sexes rather than to a more marked difference in the position of either one of the fins. In females the average prepelvic length is slightly longer (0.9% t.l.) than in males. This situation coupled with the shorter snout to anal fin length in females provides a more notable difference between the sexes in the distance between the pelvic fin origin and the anal fin origin. In females this distance averages 13.7% t.l., whereas in males it is 15.2% t.l.

Thorson, Watson and Cowan (1966, p. 395) discussed sexual dimorphism in the Bull shark, Carcharhinus leucas, from Lake Nicaragua and showed that in that species also the distance between the pelvic and anal fins was shorter in females than males. They concluded that "externally, the pelvic fins and cloaca are displaced posteriorly by about 2.5% of the body length in females, while other features are placed approximately as in the males". Our conclusions for H. perlo agree, in general, with those of Thorson et al., but the differences we have observed are much less trenchant.

Hubbs, Iwai and Matsubara (1967) have shown that in the dwarf pelagic shark Euprotomicrus bispinatus the length of the abdomen increases with size of fish, and that this increase is most marked in females. They consider this "an adaptation to fecundity in a dwarfed shark", but suggest, after noting the findings of Thorson et al., that a relatively larger abdomen in females "may well be a common characteristic of sharks".

The function of such sexual dimorphism is not well established, but it possibly hinges on two quite different requirements. For females a greater prepelvic length would permit a longer body cavity and hence be advantageous for a live-bearing species. For males a shorter prepelvic length, which in H. perlo is combined with a more posteriorly sited anal fin, provides for a longer pelvic-anal fin interspace and hence allows space for the large claspers.

Table 1. Selected dimensions of Australian, New Zealand and Atlantic specimens of Heptranchias perlo. Dimensions in columns A-B are those which according to Whitley (1931) should separate H. dakini from H. perlo. Dimensions in columns B-F illustrate sexual dimorphism in H. perlo.

Table 2. Heptranchias perlo: proportional dimensions as percentage of total length

We have not searched the literature on other notidanoid sharks to see if they show sexual dimorphism comparable with that in H. perlo, but we do note that illustrations of the male holotype and a female of Hexanchus vitulus, a new species of Sixgill shark described by Springer and Waller (1969), show the male with the anal fin origin towards the posterior end of the dorsal base, whereas in the female it is below the middle of the base. Through the kindness of Mr. Stewart Springer, co-author of Hexanchus vitulus, we have been given measurements of the type specimens. These establish that in the male holotype the proportion (%) of the dorsal fin base anterior to the anal fin origin is 81%, whereas in the female paratype it is 67% and in another smaller female it is 74%. The situation in Hexanchus vitulus, is, therefore, akin to that in H. perlo, and it seems likely that such sexual dimorphism is common to all notidanoids.

Distribution and Habitat

Bigelow and Schroeder (1948) record H. perlo from the eastern and western North Atlantic, the Mediterranean, the Cape of Good Hope and Japan. The present paper extends its known range to Australia and New Zealand. It is undoubtedly more widely distributed than has been recorded.

Little is known of its mode of life. Bigelow and Schroeder (1948) note that "It seems to be a bottom dweller chiefly, of coastal waters", and give it an extended depth range from about 400 m off Portugal and deep water off Cuba, to very shallow water and lagoons on the coast of Africa. The latter record is based on Rochebrune's (1882) account of Notidanus cinereus, but this possibly is referable to the Broadsnouted Sevengill Notorynchus, which is regularly taken in very shallow water in several parts of the world. There is considerable confusion between Heptranchias and Notorynchus in early accounts.

Australian accounts (of dakini) indicate a range of 100 to 400 m, with some captures in more shallow water. Most recorded New Zealand captures have been in moderately deep water, with some shallow-water captures (50 m) in areas where other deep-water fish are also sometimes taken.

We believe that H. perlo is a shelf-edge species, and that it may be rather more common than its reported captures indicate, its slenderness and agility allowing it to escape from trawl nets. When caught it is very active and aggressive, its flexible vertebral column allowing almost snake-like movements. Scott (1962) reports it as being frequently taken on long-lines in South Australia, and one of us (J.A.F.G.) took four fish on one long-line set in 330 m in Hawke Bay, New Zealand, in 1955. The lightly calcified skeleton of H. perlo, mentioned in passing by Springer and Garrick (1964), and shown here in Plate 2, may be an adaptation to a bathypelagic, or at least "above-bottom" mode of life, similar to that of the lightly-calcified "pelagic" shark Euprotomicrus bispinatus (see Hubbs et al., 1967). Bone and Roberts (1969) have tentatively classified H. perlo as pelagic, apparently on the basis of an earlier liver-oil density determination (Tsujimoto, 1932). Further work on fish density, liver-oil analysis (such as that of Shimma and Shimma, 1969) and studies on eye structure (Denton and Nicol, 1964) may provide some answers on the mode of life of this species.