Descriptive Notes on Some New Zealand Fishes

Transactions and Proceedings of the Royal Society of New Zealand, Volume 69, 1940, Page 228

Descriptive Notes on Some New Zealand Fishes By Gilbert P. Whitley, F.R.Z.S., Ichthyologist, The Australian Museum, Sydney and W. J. Phillips , Dominion Museum, Wellington. * By permission of the Trustees of The Australian Museum [Read before the Wellington Philosophical Society, May 26, 1937; received by the Editor, June 16, 1937; issued separately, September, 1939.] The present paper contains notes and descriptions of various species of New Zealand fishes which have been examined and discussed by us jointly. The holotype of Eleotris huttoni Ogilby = Gobiomorphus huttoni is now figured for the first time, and several new species are described. Family Myctophidae Genus Serpa Cloquet, 1827. Serpa peccatus sp.nov. This new name is necessary for Lampanyctus macropterus of Regan (Terra Nova Zool., 1, 4, March 25, 1916, p. 140, pl. 6, fig. 5) from Spirits Bay, North Cape, New Zealand. Regan's fish differs in the form of the suborbital, insertion of anal fin, fin-counts and probably also in myomere-counts from the true Myctophum (Lampanyctus) macropterum Brauer (Zool. Anzeiger, 28, Dec. 20, 1904, pp. 397 and 404, fig. 5) from the Indian Ocean. Family Paralepidae Genus Prymnothonoides nov. Genotype, P. regani sp.nov. Resembles a young Paralepis, but has only sixty myomeres and eleven anal rays. The adipose fin is, moreover, above the posterior end of the anal fin. Prymnothonoides regani sp.nov. “Prymnothonus” Regan, Terra Nova, Zool., 1, 4, March 25, 1916, p. 138, pl. 7, fig. 3. A remarkable fish from off Cape Maria van Diemen, caught in a depth of two metres, which has been left unnamed up to the present time. Family Muraenidae Genus Serranguilla nov. Genotype, Gymnothorax prionodon Ogilby (Proc. Linn. Soc. N.S. Wales (2), 9, 4, March 28, 1895, p. 720), said to have come from Port Jackson, N.S. Wales. Serranguilla prionodon (Ogilby). A new generic name is required for Ogilby's species, the sawteeth being a distinctive feature. The type, in the Australian Museum, has been examined anew, although the salient characters of

the specimen, given by McCulloch, were recently quoted by Griffin (Trans. N.Z. Inst., 56, 1926, p. 538, pl. 93) when figuring a New Zealand specimen. Genus Gymnothorax Bloch, 1795. Gymnothorax griffini sp.nov. Gymnothorax meleagris Griffin, Trans. N.Z. Inst., 58, Sept., 1927, p. 138, pl. 10, fig. 2. White Island, Bay of Plenty, New Zealand. This species, well described and figured by Griffin, is not the true G. meleagris of Shaw and Nodder (Nat. Miscell. 7, Sept. 1, 1795, pl. 220) from the “Southern Ocean,” which has less elevated head, wider gape, and is of much smaller size. The dentition, according to Kaup, is quite different from that figured by Griffin. Richardson (Voy. Erebus and Terror, Fish, 1848, p. 93) also redescribed Shaw and Nodder's original specimen. We rename the New Zealand species after our late friend Louis T. Griffin, author of several valuable papers on New Zealand fishes. Family Galaxiidae Genus Galaxias Cuvier, 1816. Galaxias castleae n.sp. (Plate 21, fig. 1.) The anal fin when laid back does not reach to procurrent caudals and pectoral is half distance from its base to base of ventral. A typical fin formula is: D.2/7; C.18 (true rays); A.2/10; P.14; V.7. The eye is 4 in head and anal origin nearer snout than caudal extremity. Fishes examined were 48 mm. in length. Locality: Lake Waikaremoana. This species agrees with G. brevipinnis in having 7–8 gill-rakers on lower part of anterior arch. It is interesting to note that G. koaro has 8 and G. huttoni has 9. The body is marked with darker bands. The species is very like G. attenuatus and might be directly descended from it. Although comparative measurements vary slightly with age, it is interesting to compare relative lengths of heads in this and other species:— attenuatus Length of head 5 to 6 ½ in the total length. castleae Length of head 5.7 to 6 in total length. brevipinnis 4 ⅔ to 5 ¼ as above. fasciatus 4 to 5 as above. alepidotus 3 ⅗ to 3 ¾ as above. huttoni Nearly 5. koaro 4.8 to 5.1. It will be seen that in size of head castleae approaches very close to attenuatus. The species is named after Miss Amy Castle, formerly Entomologist, Dominion Museum, in recognition of generous help accorded one of us in identifying insects from stomachs of fish. The Waikaremoana species of Galaxias has no previous mention in scientific literature and appears to have escaped the notice of our pioneer ichthyologists, Hector and Hutton.

The maehe, as this species is called, is a lake-dwelling species, the young of which are common in Waikaremoana in the summer. The late Elsdon Best informed us that unfortunately nothing is yet on record in regard to the life history of the maehe. The species was caught and eaten by the Maoris, and, before the introduction of trout, was an important food fish. It is now not so common as it was fifty years ago. This fish is known only from Lake Waikaremoana. It is possibly descended from a race of G. attenuatus which became land-locked there. Galaxias charlottae sp.nov. (Plate 21, fig. 2.) Br. 7 or 8. D?; A.13; P.14; V.1/6; C.14 branched rays. 10 gill-rakers on lower limb of first branchial arch. Head (40 mm.) 4, depth (36) 4.4, in standard length (161). Eye (7) 5.6, snout (13) 3, interorbital (18) 2.2, depth of caudal peduncle (24) 1.6 in head. General form very robust, and deep but strongly compressed posteriorly. Head longer than broad, with the snout tumid and the upper jaw the longer. Maxillary reaching backward nearly to below middle of eye. Teeth minute, none of them canines. A crescentic row of silt-like pores across the chin. “A few large pores above eyes, along preorbital, and over preoperculum. The left gill-membrane meets the isthmus slightly in advance of the right gill-membrane. Branchiositegals overlaid by adipose tissue. Nostril somewhat bell-shaped, lying in a depression. Pectorals and ventrals broadly rounded, more than half length of head. Ventral origin almost halfway between tip of snout and root of caudal. Anal origin behind the vertical of the anterior dorsal rays; when laid back, the anal fin reaches bases of caudal rays. Caudal bilobed, with a fin-like extension above and below the caudal peduncle. Lateral line a median series of simple pores. The fish is covered with thick mucus. General coloration in formalin, dark brownish grey, fairly uniform, lighter on the belly. A dark blue blotch behind the operculum near the origin of the lateral line. Fins yellowish with blackish margins; in the dorsal, anal, and caudal the membranes are blackish. Described and figured from the holotype of the species, a specimen 161 mm. in standard length or 7 ½ inches in total length. Austr. Mus. temporarily registered No. 1A. 5195. Preserved in Dominion Museum. Locality: Queen Charlotte Sound. Collected by Mr. B. Osborne. This species is allied to G. fasciatus (= argenteus Gmelin), but differs in having uniform coloration, a relatively, smaller upper jaw, and a deeper habitus. It differs from Regan's description of G. alepidotus in dentition and proportions. The correct name of the Mountain Trout or Black Kokopu is Galaxias argenteus Gmelin and the Barred Trout or Kokopu is apparently the young, with conspicuous

Fig. 1.—Galaxias castleae n.sp. —B. Osborne del. Fig. 2.—Galaxias charlottae n.sp. —Joyce K. Allan del. Fig. 3.—Nemadactylus concinnus Richardson. —Joyce K. Allan del.

Fig. 4.Gobiomorphus huttoni (Ogilby). —G. P. Whitley del. Fig. 5.—Oliverichtus melobesia, (Phillipps) (side view). —Joyce K. Allan del. Fig. 6.—Oliverichtus melobesia, (Phillipps) (ventral surface). —Joyce K. Allan del.

markings. The synonymy therefore is: Galaxias postvectis Clarke = G. kokopu Clarke = G. grandis Clarke (preocc. by Haast) = G. reticulatus Richardson = G. brocchus Richardson = G. forsteri Cuv. and Val. = G. fasciatus Gray = Esox alepidotus Bloch and Schneider = Esox argenteus Gmelin = Esox sp. Forster, Voy. World Resolut., 1, 1777, p. 159, from Dusky Bay. Family Berycidae Genus Paratrachichthys Waite, 1899. Paratrachichthys trailli (Hutton). Trachichthys trailli Hutton, Ann. Mag. Nat. Hist. (4), 16, Nov., 1875, p. 315, and Trans. N.Z. Inst., 8, May, 1876, p. 212. Stewart Island. Trachichthys trailli Arthur, Trans. N.Z. Inst., 17, 1885, p. 163, pl 14, fig. 2. Id. Gunther, Chall. Repts. Zool., 22, 1887, p. 23, pl. 55, fig. A. Paratrachichthys trailli (McCulloch), Zool. Res. Endearour, 1, pt. 1, 1911, p. 44. Through the courtesy of the late H. Hamilton two examples 8 inches and 6 ½ inches long were acquired by the Dominion Museum some years ago. The ventral scutes number 13 and 11 respectively. Arthur (1885, loc. cit.) found 12 in an example 7 ½ inches long; so it is possible that the number of these scutes increases as the fish grows. McCulloch (1911, loc. cit.) states that he found from 12 to 16 scutes, and the first either divided or single, with one median and two lateral spines, proving the identity of T. macleayi with Hutton's species. The Roughy or Sand-paper Fish, as the species is called, is not uncommon in trawler catches in New Zealand and southern Australia. An examination of the genital organs showed these were developing in the two above examples taken in Cook Strait in August, while Arthur stated that he found two well-advanced genital lobes in the species which he took in the Otago Harbour in September. Our examples were for the most part of a brilliant scarlet colour, being more silvery on the ventral surface. The fins were scarlet, with the exception of the ventrals, which were silvery. An excellent description of this fish has been already given by Arthur. Family Rhombosoleidae Genus Rhombosolea Gunther, 1862. Subgenus Adamasoma nov. Genotype, Rhombosolea retiaria Hutton. Sixteen to nineteen gill-rakers on lower part of first branchial arch. Body more or less coloured on blind side, general coloration darker than that of Rhombosolea, sensu stricto. Eyes small, about one-seventh length of head. Snout not produced. Rhombosolea (Adamasoma) retiaria Hutton. River or Black Flounder; Patikimohoao. Rhombosolea retiaria Hutton, Trans. N.Z. Inst., 6, 1874, p. 107. New name for R. tapirina Hutton, Trans. N.Z. Inst., 5, May, 1873, p. 268, pl. 12, fig. 83b, non Gunther. Wellington Harbour.

Rhombosolea (Adamasoma) retiaria adamas subsp.nov. Diamond Plaice. Rhombosolea retiaria Phillipps, N.Z. Jour. Sci. Tech., 7, 6, 1925, p. 368, fig. Hokitika. This South Island fish is subspecifically distinct from the North-Island Black Flounder, the characters of a Hokitika specimen having been given by Phillipps, loc. cit. Family Stromateidae The following genus and species have not yet been defined:— Crystaculeus dyscritus Thomson, Ann. Rept. Mar. Dept. N.Z., 1926, p. 19. Nomen nudum ex Young MS. Family Nomeidae Genus Hyperoglyphe Gunther, 1859. Hyperoglyphe porosa (Richardson). Diagramma porosa Richardson, Zool. Voy. Erebus and Terror, Fish, 1845, p. 26, pl. 16, figs. 5–6. Australia. It is of interest to place on record the fact that the species, figured by Thomson and Anderton under the name Eurumetopos johnstonii (History of the Portobello Marine Fish Hatchery, 1921, p. 73) is not the above species, but one of the Scorpaenidae, Helicolenus papillosus probably. H. porosa appears in Wellington markets at odd intervals. It is sold as bonito and is caught by line fishermen in deep water in Cook Strait. The usual size is 2 ft. 6 in., though much larger examples are recorded. The following is a short description of an example examined on 16th February, 1924:— The dorsal has seven short spines, the fourth being the longest: second dorsal being 3/17; A.3/11; V.24; pectoral is long and slender, reaching to above anus; ventrals inserted beneath the pectoral a little further back, and reaching halfway from its base to anus. The lateral line contains about 90 scales and curves above pectoral; maxillary reaches to under posterior of eye; caudal deeply concave; diameter of eye 1.75 inches; length of head 8 inches; depth of ventral base 8.75 inches, length, of pectoral 7.75 inches, total length 2 ft. 6 ½ in. Family Cepolidae Genus Cepola Linné, 1764. Cepola aotea Waite. Cepola aotea Waite, Proc. N.Z. Inst., 1910, pt. 1, Sept. 10, 1910, p. 26. Bay of Plenty, New Zealand. At the end of August, 1931, some college boys found a fish on Wanganui Beach which has been identified with the above species. Mr G. Shepherd kindly supplies the following description:— Length from snout to hypural joint 45 cm.; greatest depth 28 mm.; depth of hypural joint 3mm.; point of snout to vent 73 mm.; teeth in upper jaw, 4 anterior on either side, large and hooked, the anterior pair the longest, widely spaced, behind a row of closely set small teeth, lower jaw has 6 large

widely spaced hooked teeth on either side anteriorly and behind a closely set row of smaller teeth; all teeth in a single series; premaxillary extends to base of maxillary which is much dilated and square at base. Colour is uniform bright red including anal fin and eye, ventrals whitish. The whole fish is extremely slender, greatly compressed, and tapered almost to a point. D.72; A.66;V.1 + 5; Br.5. Gill-rakers numerous and comb-like. Hypolycodes haastii resembles this species; but has a larger number of dorsal fin rays. Family Histiopteridae Genus Griffinetta nov. Genotype, Griffinetta nelsonensis sp.nov. Fourteen dorsal spines and four anal spines. Base of soft dorsal much shorter than that of the spinous dorsal, none of the fins falcate. Head and scapular regions with exposed, rugose, bony surfaces radially striated. A few small teeth form a band in each jaw, and others are present on the vomer, but the palatines are edentulous. Snout not excavate or produced. Named after the late Louis T. Griffin of Auckland. Griffinetta nelsonensis sp.nov. Pseudopentaceros richardsoni McCulloch and Phillipps, Rec. Austr. Mus., 14, 1, Feb. 28, 1923, p. 18, pl. 4, fig. 1. Nelson, New Zealand. Not Pentaceros richardsoni Smith, Illustrations of the Zoology of South Africa, Pisces, 1844, pl. 21, from South Africa. Family Cheilodactylidae We consider that records of the Chilean Mendosoma lineatum Guichenot (Gay's Chili, Peces, 1849, p. 213, pl. 5, fig. 2) from New Zealand, actually refer to young specimens of the Trumpeter, Latris lineata (Bloch and Schneider). Genus Nemadactylus Richardson, 1839. Nemadactylus Richardson, Proc. Zool. Soc. London, 7, Nov., 1839, p. 97. Orthotype, N. concinnus Richardson. Nematodactylus Gill, Proc. Acad. Nat. Sci. Philad., 1862, pp. 114 and 121. Id. Boulenger. Zool. Record, 1880 (1881) Pisces, p. 8. Emendations. This genus may perhaps be founded on some incompletely developed Cheilodactylid, but is evidently not the young of any well-known species, such as the Terakihi (Sciaenoides macropterus), since its fin-formulae are so different. The relationships of Nemadactylus have always been a puzzle and until now it has only been known from around Tasmania and the Island of St. Paul. We are now able to record it from New Zealand on the basis of the interesting little specimen described below.

Nemadactylus concinnus Richardson. (Plate 21, fig. 3.) Nemadactylus concinnus Richardson, Proc. Zool. Soc. Lond., 7, Nov., 1839, p. 97;Trans. Zool. Soc. Lond., 3, June 16, 1842, p. 116, pl 4, fig. 2; Rept. Brit. Assn. Adv. Sci., 11, 1841 (1842), p. 71; Tasm. Journ. sci., 1, 1842, p. 63. Id. Gunther, Cat. Fish. Brit. Mua, 2, 1860, p. 85 (? and p. 415 as Mene maculata from “Open sea”). Id. Kner, Sitzungsb. Akad. Wiss. Wien, 49, 1, 1864, p. 486 (no loc.);Reise Novara, Zool., 1865, p. 94. Id. Sauvage, Comptes Rendus, 81, 1875, p. 988 (St. Paul) and Arch. Zool. Exper., 8, 1879, pp. 3 and 22. Id. Johnston, Proc. Roy. Soc. Tasm., 1882 (1883), p. 113 et ibid. 1890 (1891), p. 31. Id. McCulloch, Austr. Mus. Mem. 5, 2, 1929, p 257. ? Platystethus huttonii Gunther, Rept. Voy. Challenger, Zool., 31, 78, 1889, p. 13. Part only, not type, but young specimens nos. 1 and 2 from between Australia and New Zealand and between Australia and South Africa. Not Platystethus huttonii Gunther, Ann. Mag. Nat. Hist. (4), 17, 1876, pp. 390 and 395, a species of Evistius. D.18/28; A.3/16; P.15; V.⅕ C.13 branched rays. L.lat.55. Head 3 ½, height of body 2 ½ in standard length. Eye nearly 34 in head. First dorsal fin commences very slightly in advance of the vertical of opercular edge. Anal commences below second dorsal's origin. Pectoral commences below first dorsal and reaches to origin of second. Anal ceases below a point about two-thirds of the way along the second dorsal. Total length 78 mm. Snout to first dorsal 14. Diameter of eye 5. Eye to end of snout 4. Depth at dorsal origin 24. Snout to anal origin 35. Length of caudal 16. Longest ray of pectoral 16. Length of head 16. The angle of the mouth reaches over half the distance from the snout to the eye and inclines obliquely downward at about 45° to the horizontal line of the body. The suboperculum is striated and the eyes and front of the head are covered by a transparent membrane. Described and figured from a specimen 78 mm. long or about three inches overall. Temporarily registered No. 1A. 5196, Austr. Mus. Now preserved in the Dominion Museum. Locality: Mouth of the Waikanae River, New Zealand; 17th November, 1930. Collected by Captain Hayes. Relationships: This is evidently a pelagic species referable to the Cheilodactylidae, yet in some respects it recalls Evistius and Platystethus. Barnard, in South Africa, has identified some of Gunther's “Challenger” specimens of “Platystethus huttonii” as young Cheilodactylidae, near, if not identical with, the genus Palunolepis Barnard, 1927. Again, on comparing descriptions and figures of fishes identified by authors as Evistius huttonii, discrepancies in formulae and proportions are noticeable, and it is evident that more than one species has been confused. We think we are correct in identifying our New Zealand specimen as N. concinnus, because it agrees well with Richardson's detailed description and figure published in 1842; the only differences are the deeper thoracic profile and the smaller size, but these might be accounted for by growth-changes. In our detailed analysis of descriptions and records of Nemadactylus, Evistius, and Platystethus, one fact emerges: the Tasmanian form of Evistius huttonii recorded by McCulloch (Rec. Austr. Mus., 14, 2, 1923, p. 121) actually represents a new subspecies, which we name Evistius huttonii tasmaniensis. n.subsp.

Family Parapercichthyidae nov. Genus Parapercichthys nov. Genotype, Enchelyopus colias Bloch and Schneider (Syst. Ichth., 1801, p. 54. Type-loc, Queen Charlotte Sound, South Island, New Zealand). This is the New Zealand Blue Cod which has been erroneously placed in the genera Percis or Parapercis by modern authors. It has, however, no affinity whatever with the typical members of those genera, so the above slight nomenclatural change is made and the species may henceforth be called Parapercichthys colias. The genotype of Parapercis is cylindrica, a relatively large scaled species with 50 scales in the lateral line, while colias has 66. Vomerine and palatine teeth present in cylindrica are absent from colias. Percis gilliesii Hutton (Ann. Mag. Nat. Hist. (5), 3, Jan. 1, 1879, p. 53) from Brighton, near Dunedin, belongs to the genus Neopercis Steindachner, in the family Parapercidae. Family Uranoscopidae. Genus Gnathagnoides nov. Genotype, Gnathagnus innotabilis Waite (Rec. Austr. Mus., 5, 4, June 16, 1904, p. 238, pl. 26, fig. 1), from New South Wales. This new genus is allied to the Japanese Gnathagnus Gill, 1861, whose genotype is Uranoscopus elongatus Temminck and Schlegel, but differs from the latter in having a pronounced humeral spine, rounded plectroid mental dilatations, differently formed fins, and apparently different skull-structure. The New Zealand form is larger in size than the Australian, and may be named Gnathagnoides innotabilis grandior subsp.nov. Family Eleotridae. Genus Gobiomophus Gill, 1863. Gobiomorphus huttoni (Ogilby). (Plate 22, fig. 4.) “Gobius n.sp” Krefft List Rept. Fish Austr. Mus., 1862, p. 13. Eleotris huttoni Ogilby, Proc. Linn. Soc. N.S Wales (2), 9, 1894, p. 369. Waikato River, North Island of New Zealand. This species was excellently described by Ogilby, but has not been featured in New Zealand ichthyological literature. The holotype, here figured for the first time, is an old specimen (regd. no. 1. 3162) in the Australian Museum, Sydney. The species appears to be valid, being distinguished from Gobiomorphus gobiodes (Cuvier and Valenciennes) by having less, extensive jaws, a narrowed interorbital, comparatively smaller scales, and fewer dorsal rays. Family Blenniidae. Genus Gilloblennius nov. Genotype, Blennius tripennis Bloch and Schneider, Syst. Ichth., 1801, p. 174. Ex Forster MS. South Island, New Zealand. Now Gilloblennius tripennis. This genus is distinguished by the very long lateral line, which is more primitive in structure than in other New Zealand blennies, in which it has atrophied. Dorsal spines three. Head pointed, its

profile much less steep than in other Three-fin Blennies. Other characters as given by Waite, Rec. Canterb. Mus., 2, 1913, p. 3. Genus Forsterygion nov Genotype, Blennius varius Bloch and Schneider, Syst. Ichth., 1801, p. 178. Ex Forster MS. South Island of New Zealand. Cuvier and Valenciennes (Hist. Nat. Poiss., 11, July, 1836, p. 413, pl. 339) figure this New Zealand blenny as Tripterygion nigripenne beside the typical Tripterygion (nasus). Though both forms have since been retained in Tripterygion, they should be separated generically, since the New Zealand species differs in being of much larger size, has an increased number of dorsal spines, and a quite differently shaped head. The lateral line of Wellington specimens which we have examined ends sometimes below the second and sometimes below the third dorsal fin. It may be mentioned here that Enneapterygius mortenseni Rendahl, 1925, is evidently a synonym of Helcogramma medium (Gunther), originally described in 1861 as another Tripterygium. Family Linophrynidae. Genus Haplophryne Regan, 1912. Haplophryne triregium sp.nov. New name for Haplophryne mollis of Regan (Terra Nova, Zool., 1, 4, March 25, 1916, p. 148, pl. 10, fig. 2) from surface at Three Kings Islands, not Aceratias mollis Brauer (Zool. Anzeiger, 25, April 7, 1902, p. 297) from the “Mitte des Indischen Oceans, Tiefe 2200 m.” This type-locality was between New Amsterdam and Cocos Island, fide Brauer, Wissen. Ergebn. Deutschen Tiefsee-Exped. Valdivia, 1906, p. 324, pl. 16, fig. 10. Brauer s specimen is larger than Regan's, and the New Zealand fish differs in having much smaller nostrils, a different bodily outline, anal origin in advance of that of dorsal, and a hidden rudimentary illicium, the pectoral fin is differently situated and there are spot-like chromatophores on the flanks. Family Gobiesocidae. Genus Oliverichtus nov. Genotype, Oliverichtus melobesia(Phillipps, 1927). Plate 22, figs. 5 and 6. A species, Trachelochismus melobesia, was described by Phillipps, Trans. N.Z. Inst., 58, p. 131, pl. 5. Numerous specimens had been collected by Dr. W. R. B. Oliver under rocks in beach pools on the westerly coasts of Wellington. Its differences from Trachelochismus were realised at the time. The dorsal and anal are much larger and the number of ventral rays smaller than in Trachelochismus. We therefore designate Trachelochismus melobesia Phillipps as the genotype of Oliverichtus gen.nov. The body is held to the rocks by a disc quite unlike that of related genera. The following additional material on this interesting species may now be added:—Length of head 9 mm., interorbital space 2 mm., eye 1 mm., eye to snout 1.8 mm., total length 28 mm., length to caudal peduncle 24 mm., coralline patch 12 × 7 mm., sucker 4 mm., height of body 3 mm. D.9; P.10; C.10; A.9; V.3 fused rays.