A New Freshwater Fish of the Genus Philypnodon.
[Read before the Canterbury Branch, July 6, 1938; received by the Editor, December 7, 1938; published separately, June, 1939.]
Taxonomy And Nomenclature.
The Gobioid fishes of Australia, with which the New Zealand forms show close affinities, were revised in 1917–21 by McCulloch and Ogilby, who separate them into three sub-families, Periophthalminae in which the eyes are erectile and the base of the pectoral fin is very muscular, Gobiinae which has the ventral fins more or less united by a membrane, and Electrical in which none of the above-mentioned features is present. In the subject of the present paper the eyes are not erectile, the base of the pectoral is not unduly muscular and the ventral fins are separate; it belongs, therefore, to the sub-family Electrical or, as it is classed by some zoologists, the family Electrode. McCulloch and Ogilby recognize 13 Eleotrid genera, the more distinct of which are separated by the number of rays in the ventral fins, the presence or absence of bony crests on the head, the presence or absence of spines on the preoperculum and whether the number of scales in the longitudinal series exceeds or falls short of 50. The present fish has 6 rayed ventrals, it has neither bony crests on the head nor spines on the preoperculum, and the number of scales is less than 50; on these grounds it is disqualified for admission to 8 of the genera recognised. The remaining 5 are separated largely by the disposition of scales on the head, Ophiocara and Mogurnda, which are very closely related, having the head almost entirely covered with scales, Carassiops being scaled on the cheeks, opercles and occiput, but having the snout and interorbital space naked, while in Gobiomorphus the arrangement is very similar except that the cheeks are naked or carry only a few rudimentary scales. As the subject of the present paper has no scales on the cheeks, opercles, snout or interorbital space, and the remainder of the head is naked or sub-naked, it cannot be admitted to any of the foregoing genera. Of the remaining genus, Philypnodon, there is no further description in the paper referred to beyond the statement that the cheeks and opercles are naked, but reference to synonymous descriptions and comparison of the local fish with specimens of Philypnodon grandiceps, kindly forwarded by the Australian Museum and by the National Museum, reveals a close agreement in the characters mentioned and also in the arrangement of the body scales, which are smaller than those on the tail and deeply embedded. There is further agreement in the fin ray formula, the presence of papillae on the head and the absence of open pores, as well as in the vertebral count and in the circumstance of the caudal vertebrae being more numerous than those of the abdominal section. The principal differences between the two forms are that in P. grandiceps the head is longer, the gill-openings extend much further forward below, the maxillary is freer posteriorly, the genital papilla is smaller and the ventral surface is more completely sealed. These differences, although perhaps greater than those separating certain genera such as Ophiocara
and Mogurnda, do not seem sufficient to justify an addition to the existing list of Eleotrid genera, which, to the present writer, appears to have been unduly extended, and the local fish is here regarded as con-generic with the form known as Philypnodon grandiceps. The validity of the genus Philypnodon and the propriety of applying this, name to any of these fishes are, however, open to question. This genus, was founded by Bleeker (1874) to accommodate Eleotris nudiceps Castelnau (1872), which is described as being naked-headed and having teeth on the vomer, palatines and the tongue in addition to those on the jaws. As Bleeker's paper appears to be rather inaccessible, and the present writer is fortunate in possessing a copy, the definition of Philypnodon is reprinted below.
Dentes palatini et linguales. Dentes inframaxillares. intermaxillaribus longiores. Caput alepidotum. Nucha squamata. Squamae ctenoideae 47 circ. in serie longitudinali. D. 7–11. A. 11.
Spec. typ. Eleotris nudiceps Cast.
Waite (1903-05) has adopted this genus for the species nudiceps and grandiceps, which have no teeth on any part of the mouth except the jaws, on the grounds that Castelnau's description of the dentition of E. nudiceps was incorrect and that his species was the common Eleotrid of the Yarra River to which the name Philypnodon nudiceps is now applied. According to Ogilby (1897) no type or other specimen of E. nudiceps had been located, and the identification of this species with the common Eleotrid of the Yarra was based principally on the circumstances of locality and exclusive occurrence. Dr. Léon Bertin, of the Paris Museum National D'Histoire Naturelle, now informs the writer that three of Castelnau's types of E. nudiceps are in that institution and that they have no teeth on the vomer, palatines and tongues. Specifications of other structural characters, kindly’ supplied by Dr. Bertin agree with those of the common. Eleotrid of the Yarra known as Philypnodon nudiceps. The identification of this fish with Castelnau's species is therefore complete, and Castelnau's description as published is shown to be incorrect, but there seems no reason to suppose that the error was anything-more than an accidental mistake in transcription.
The most complete description of these sub-naked headed Eleotrids is that of Ogilby (1897), who created for them the genus Ophiorrhinus, nominating as genotype Eleotris grandiceps Krefft (1864), a form so closely related to nudiceps that its specific distinctness seems open to question. Ogilby's genus, however, has no standing, as Bleeker had in 1874 created the genus Gymnobutis for E. gymnocephalus Steindachner (1866), a synonym of E. grandiceps. The definition of Gymnobutis is given below:—
Dentes maxillis pluriseriati subaequales, canini nulli. Caput prismaticum, aeque altum circ. ac latum, valde acutum, ubique alepidotum. Squamae 40 circ. in serielongitudinali. D. 7–10. A. 10.
Sp. typ. Eleotris gymnocephalus Steind.
This genus was defined in the same paper as Philypnodon but on a later page, the only claim to recognition that can be made on behalf of Philypnodon being based solely on this page precedence The present writer submits that Gymnobutis has the better claim. The International Rules of Zoological Nomenclature include the following recommendation regarding the application of the law of priority:—“Other things being equal, that name is to be preferred which stands first in the publication (page precedence).”
In the present instance other things are by no means equal. They are equal in so far that both genera were based on published descriptions without reference to specimens, both have recognisable genotypes, and these genotypes, if not identical, are certainly congeneric, but there is no equality in the accuracy of the generic definitions. While the definition of Gymnobutis forms a reasonably accurate description of the fishes referrable to this genus, Philypnodon is based expressly on characters that are non-existent in its component species. The anomalous position exists that if an Eleotrid possessing the dentition indicated in the definition of Philypnodon were discovered it could not be admitted to this genus; it would be excluded by the possession of those very characters upon which the genus is founded.
It is proposed to make the above circumstances the basis of an application for a ruling by the International Council, and, pending this decision, the name Philypnodon is tentatively retained.
The author wishes to express his thanks to Mr. G. P. Whitley, of the Australian Museum, and Mr. G. Mack, of the National Museum, Melbourne, for specimens of P. grandiceps, and to Dr. Léon Bertin, of the Paris Museum National D'Histoire Naturelle, for information regarding the types of Eleotris nudiceps.
Philypnodon breviceps n.sp. (Plate 15, figs. 1–3.)
B. 6. D. vi-vii i 8–10. A. i 8–9. P. 14–15. V. 6. C. viii 15 ix. Vertebrae 29–31.
Caudal vertebrae the more numerous. Body somewhat depressed anteriorly, compressed posteriorly. Head 3.5-3.8 in length (without caudal fin), depth before ventrals 4.1-4.6 in same. Eye 3–9-5 in head, interorbital space 0.7-1.37 times the diameter of the eye. Depth of head 1.1-1.26 in width, snout rounded, mouth steeply inclined, the lower jaw usually slightly the longer, maxillary reaching to, or scarcely to, anterior of eye. Nostrils wide apart, the anterior tubular. Teeth multiserial, hooked, outer row definitely enlarged throughout in upper jaw, in lower jaw enlarged anteriorly. No teeth on vomer, palatines and tongue, pharyngeal bones with villiform and fine hooked teeth. Gill-opening not much extended below, usually narrower than the isthmus. Pseudobranchiae poorly and irregularly developed, gill-rakers obsolete. Rows of minute papillae on cheeks, opercles and around eyes, no open pores. Cheeks, opercles and inter-orbital space scaleless, top of head scaleless or with a few irregularly-scattered imperfectly developed scales in the region of the nape of neck, sub-naked dorsal area usually extending to within a short distance of the origin of the first dorsal fin. A few scales extending
forward above the base of pectoral fin usually to about mid-way between cheek and rear of opercle. Base of pectoral scaleless, ventral surface usually scaleless to within a short distance of anus. Scales ctenoid laterally and dorso-posteriorly, becoming cycloid ventrally and dorso-anteriorly, fully imbricate on tail, but on the anterior part of the body steeply inclined in deep pockets with only a narrow edge protruding. The irregularity with which the scales are placed renders their enumeration a matter of uncertainty, but so far as may be determined there are from 43 to 48 counted longitudinally from above the base of the pectoral fin to the base of the caudal. The first dorsal fin is inserted at 0.36-0.38 of the length (without caudal) its form more or less rounded in some specimens, in others the top almost parallel with the base; aberrant forms with 5 developed rays and rudimentary ray articulated to either the first or last developed ray of fin. Height of second dorsal contained 1.3-1.55 in its length at base, rays about equal in length except the first two, which are shorter, last ray cleft to base. Origin of anal about perpendicular from fourth ray of second dorsal, its height contained 0.95-1.04 times in its length, last ray cleft to base. Caudal rounded, pectorals very broad and rounded. Caudal peduncle 1.52-2.05 times as long as its last depth. Genital papilla prominent, that in the female the larger.
Colour variable with the locality, in some specimens greyish with fins speckled darker, in others merging to yellow with reddish brown spots. A band of dull yellow across pectoral fins at base, top of first dorsal usually dark yellow or orange.
Maximum total length observed, 109 mm.
Types: Holotype and two paratypes in Canterbury Museum, two paratypes in Australian Museum, two paratypes in British Museum. Type locality Kowai River, tributary of the Waimakariri, Canterbury, New Zealand. Altitude 1800 feet.
Differs from P. grandiceps as already indicated and from P. nudiceps in most of the same characters, these two forms being very similar.
The present species is named breviceps on account of its comparatively short head.
The variation in the number of vertebrae (29-31) is greater than has been observed in other New Zealand Eleotrids, and the extent of the scale covering is notably inconstant. In most secimens the ventral surface in naked to within a short distance of the anus, but in occasional ones it is covered with small cycloid scales from the rear of the ventral fins. The latter arrangement closely approximates that of P. grandiceps. There is also variation in the extent to which the dorsal surface is scaled anterior to the first dorsal fin; in examples of maximum scaling the complete dorsal covering commences about the perpendicular from the rear of the opercle. The depth to length ratio of the caudal peduncle varies, to some extent, with the size of
the fish, the larger specimens having the comparatively greater depth. Other characters such as fin ray formula show only normal variation. Distribution and Habit.
This fish has been collected from the Ashburton, Rakaia, Selwyn, Waimakariri, and Ashley Rivers and is probably widely distributed in the South Island. It occurs most plentifully in streams and lakes at altitudes exceeding 1000 feet where Gobiomorphus gobioides, for which it has hitherto been mistaken, is rare or absent. The latter fish, which has a more coastal habit, is readily distinguishable from Philypnodon by the difference in the scale covering on the head and nape shown in figs. 2 and 3. P. breviceps has not been taken in tidal water, but in rapid shingly rivers its range extends to within a short distance of tidal influence. Its food in alpine streams has been found to consist of larvae of Diptera, Plectoptera, Trichoptera, and Perlaria, and its spawning season appears to be about December. Specimens taken from the Harper River at the end of November were distended with ova measuring approximately 1.4 mm. in diameter, and apparently on the point of spawning.
Bleeker, P., 1874. Esquisse d'un Systeme naturel des Gobioides, Archives Néerlandaises des Sciences Exactes et Naturelles, pp. 289—331.
Castelnau, P. DE, 1872. Contribution to the Ichthyology of Australia, Pro. Zoo Soc. Vict., I, p. 126.
Krefft, G., 1864. Notes on Australian Freshwater Fishes and Descriptions of Four New Species, Pro. Zoo. Soc. Land., p. 183.
Mcculloch, A. R, and Ogilby, J. D., 1917–21. Some Australian Fishes of the Family Gobiidae, Rec. Aust. Mus., V., xii, pp. 193–291.
Ogilby, J. D., 1897. On some Australian Electrical, Pro. Linn. Soc., N.S.W., vol. xxi, p. 740.
Steindachner, F., 1866. Zur Fischfauna von Port Jackson in Australien, Sitzungsber Akad. Wissenchaften, liii, pp. 1–58.
Waite, E. R., 1904. A Review of the Eleotrids of New South Wales, Rec. Aust. Mus., V, p. 283.