25 September 1991 PROC. BEX. SOC. WASH. 104(3), 1991, pp. 468-473 MATSUBARICHTHYSINUSITATUS, A NEW AND OF (: APLOACTINID AE) FROM THE GREAT BARRIER REEF

Stuart G. Poss and G. David Johnson

Abstract.—Matsubarichthys inusitatus is described from a single specimen, taken from One Tree Island, Queensland, . It is readily distinguished from other scorpaenoids by its unique dorsal-fin formula (II, IV, 6) and ver- tebral formula (9 + 12 = 21), both with fewer elements than previously reported for a scorpaenoid; a low anal fin-ray count (11,4); a pelvic fin-ray count of 1,1; and the gill opening restricted to a pore dorsal to the opercle. Although the holotype is small (8.5 mm SL) and study of its internal anatomy thus limited, Matsubarichthys inusitatus is a relatively primitive aploactinid that can not be readily assigned to either of the two currently recognized aploactinid subfam- ilies, but may be a sister-group of the Bathyaploactininae.

Neuston tow sampling conducted at One using Olympus SZH and Leitz Stereo TS Tree Island at the southern end of the Great binocular dissecting microscopes. B amer Reef resulted in the capture of a sin- Radiographic techniques suggested by gle predemersal-staged specimen represent- Tucker and Laroche (1984) were followed. ing a new genus and species of Aploactini- Radiographs were taken on a Koizumi Sof- dae (). tex SE X-ray unit, using Kodak ultrafine Aploactinids, of which 36 have been pre- grain SB single emulsion diagnostic radio- viously described, are widely distributed graphic film and Kodak Industrex R type throughout the Indo-Pacific. However, film. Exposure times of 6 to 12 sec at 50 to many remain extremely rare in collections 70 kv and 3 mA were employed. and thus poorly understood. Although it is The specimen was lightly stained with evident that velvetfishes are closely allied alizarin red S in 70% ethanol to show more with the scorpionfishes, particularly the clearly critical osteological features not vis- stonefishes and waspfishes (Matsubara, ible in the radiographs. The bones stained 1943a), the combination of presumptively only weakly, probably the result of pro- uniquely derived characters used to define longed storage in unbuffered formalin prior the family presents conflicting hypotheses to transfer to ethanol. of relationship. This paper describes an The abbreviation AMS designates the aploactinid scorpaenoid whose features un- Australian Museum, Sydney, where the ho- derscore this conflict. lotype and only specimen is deposited. Materials and methods.— Counts and head spine terminology used in this paper Matsubarichthys, new genus follow those of Eschmeyer (1969). Mea- Figs. 1, 2 surements were taken as specified in Poss Type species.—Matsubarichthys inusita- (1982), but using an ocular micrometer in- tus. stead of dial calipers. Illustration of the axial Diagnosis.—An aploactinid scorpaenoid skeleton was based on multiple radiographs fish with: II, IV, 6; anal fin II, 4; VOLUME 104, NUMBER 3 469 pelvic fin I, 1; 15 pectoral fin-rays; fin rays ible in radiograph, probably absent. First in all fins unbranched; no enlarged preor- two dorsal-fin pterygiophores enlarged and bital spines; no scales (except for ); closely associated with cranium, both in- branchiostegal membranes connected to serting in first interneural space; first dorsal isthmus along entire length, opening to spine in supernumerary and second in serial opercular chamber restricted to a small pore association with first pterygiophore, both above tip of opercle; no slit behind posteri- spines enlarged and robust, the second no- ormost hemibranch; no pterygiophore in- tably so; third pterygiophore inserting in serting between neural spines of sixth and fourth interneural space and bearing no su- seventh vertebrae. pernumerary spine (Fig. 2). Neural spines Matsubarichthys inusitatus can be distin- of first two vertebrae small, directed for- guished readily from all other scorpaenoid ward, but not clearly visible in radiographs. fishes by its dorsal-fin, pelvic-fin, and ver- Neural spines of sixth and seventh vertebrae tebral counts, which are lower than those of without interdigitating dorsal-fin pterygio- any other species in the suborder. phore. Caudal skeleton with parhypural and Description.—Head large (Fig. 1), without hypurals 1,2,3, and 4 fused into single plate extensive spination. Lacrimal (infraorbital and to urostyle (hypural 5 not visible in 1), immovable, its posteroventral border radiograph); neural spine of preural cen- with 2 small, weak spines. Infraorbital bones trum wide and long; number of epurals and 2 and 3 strongly arched, with large infra- uroneurals not determinable. orbital canal pores opening as distinctly Etymology.—Named in honor of the late raised tubes. Small, tubular, fourth infra- Professor Kiyomatsu Matsubara, in recog- orbital bone present. Nasal bone tubular, nition of his pioneering and inspiring re- without spine. Anterior nostril, an attenuate search on scorpaenoid fishes. The gender is tube near tip of snout, extends over pre- masculine. maxilla. Interorbit wide, with ascending processes of premaxillae extending between Matsubarichthys inusitatus, anterior half of orbits. Mouth strongly up- new species turned. Teeth on upper and lower jaws; none Figs. 1, 2 on palatines. Presence of teeth on vomer uncertain. No slit behind posteriormost Holotype.-AMS 1.29395-001, 8.5 mm hemibranch. One infrapharyngobranchial standard length (SL). Australia, Queens- tooth plate. Epibranchial of first gill arch land, Great Bamer Reef, Capricorn Group, with a narrow, notably short uncinate pro- One Tree Island, 23°30'S, 152°05'E, neus- cess extending posteromedially from dorsal ton tow, Patti D. Schmitt, date, time, and ramus of bone at an angle of about 90°. depth of capture unknown. Three lateral-line pores on each side on low- Description.—Dorsal fin-rays II, IV, 6. er jaw, a fourth between lower jaw and pre- Anal fin-rays II, 4. Pectoralfin-rays 15 right, opercle on each side. Interhyal narrow 15 left. Caudal fin damaged, but relatively throughout its length, not expanded poste- long and rounded, without branched rays; riorly. Branchiostegal rays 6. Posterior 10 segmented rays (5 ventral; 5 dorsal), and opening of opercular chamber restricted to 4 unsegmented, procurrent rays (2 dorsal;.2 a small pore above tip of opercle, with bran- ventral). Lateral line with 6 long, tube-like chiostegal membranes firmly attached to scales, the first 3 very close together, the last isthmus. Ventral surface of urohyal narrow. extending over base of caudal fin. Gill rakers No pseudobranch. 1+4 = 5 (right). Vertebrae: 9 precaudal + Body covered with smooth skin, naked, 12 caudal = 21. except for lateral line. S wimbladder not vis- Head extremely large; notably depressed 470 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON

Fig. 1. Holotype of Matsubarichthys inusitatus in lateral view (AMS 1.29335-001, 8.5 mm SL). Broken tip of second dorsal fin spine reconstructed.

(specimen with buccal cavity strongly flared of orbit just dorsal to third infraorbital bone; laterally on preservation). Lacrimal with another smaller spot deep and immediately small spinous points, but without enlarged ventral to opercular-hyomandibular hinge; spines or strong ventral lobes. Preopercle a smaller inclusion just dorsal to opercular with 2 short, somewhat blunt spines near opening, and a still smaller spot anterior to posterior margin of third infraorbital bone, this; two small spots on each side posterior the more dorsal the larger; additional, more to symphysis of premaxillae; large spot on ventrally located spines, if present, not vis- left side of isthmus; several minute dark ible through skin. Cheek with a few fleshy spots over ventral surface of iris, one lateral protuberances. Similar warty structures to urostyle. Small, densely melanic spots at scattered over ventral surface of head. bases of dorsal-fin spines, one between base Body scales absent, except for lateral line. of first and second, one posterior to fourth, Lateral-linescales extremely long, distinctly and a more conspicuous one posterior to tubular. The first 3 touch each other and fifth spine; spot anterior to anal-fin origin; appear, without high magnification, as a sin- another near base of last dorsal-fin ray (right gle elongate scale, except for presence of side only); one in axil of left (but not right) pores demarking their ends. pectoral fin. Dorsal fin originates above dorsal arm of Measurementsfor the holotype in mm are preopercle, with anteriormost two spines as follows (percent SLin parentheses): stan- well separated from remainder of fin; the dard length 8.5, head length 3.7 (44); snout first locked and directed anteriorly, the sec- 0.8 (9); orbit 1.0 (12); interorbit 1.0 (12); ond more robust, also locked, but directed jaw 1.4 (16); postorbit 2.0 (24); body depth vertically. Second spine broken (recon- 3.1 (36); predorsa!2.2 (26); anal fin 2.9 (34); structed in Fig. 1). Third and succeeding caudal fin 1.6 (19); pectoral fin 3.1 (36); pel- spines relatively narrow. vic fin 0.6 (7); dorsal spines, first 0.7 (8), Color in fife unknown. Color of specimen, second 0.9 (11), third 0.3 (4), fourth 0.6 (7), now in 70%ethanol, uniformly buff tan with fifth 0.7 (8), sixth 0.7 (8); anal spines, first several conspicuous, densely melanic inclu- 0.8 (9), second 1.1(13); width ofinterorbital sions that form distinct, deep-to-superficial ridge0.4 (5); caudal peduncle depth 0.9 (11); spots, most bilaterally symmetric: largest, snout to second dorsal spine 2.3 (27), to about V* to Vi size of pupil in ventral well third dorsal spine 4.0 (47), to fourth dorsal VOLUME 104, NUMBER 3 471

Fig. 2. Axial skeleton of holotype of Matsubarichthys inusitatus as reconstructed from multiple radiographs. Neural and haemal spines are not autogenous. Shape of anterior pterygiophores and epural region of caudal skeleton (indicated by hatching) cannot be determined from radiographs. Bar represents 2.0 mm. spine 4.4 (52), to fifth dorsal spine 4.7 (55), cannot be verified until larger specimens are to pelvic-fin insertion 3.0 (35); width of first collected. Large tubular lateral-line scales dorsal spine at midlength 0.2 (2); depth of occur commonly in -stem scor- interorbitat midorbit 0.1 (1); incision of fin paenoids. Those of Matsubarichthys inusi- membrane at fourth dorsal spine from tip tatus are particularly large, but this exag- to membrane 0.1 (1); operculartip to dorsal gerated size may reflect a juvenilecondition. fin 0.7 (8); uppermost preopercular spine In the tetrarogine Coccotropsis gymnoder- 1.0(12); first dorsal spine to fifth dorsal spine ma, which also has very large tubular lat- 3.1 (36), to pelvic-fin insertion 2.8 (33), to eral-line scales, those of very small speci- anal-fin origin 4.5 (53); fifth dorsal spine to mens appear disproportionately large. pelvic-fin insertion 4.0 (47), to last dorsal Conspicuous cephalic lateral-line pores are spine 0.9 (11), to anal origin 2.2 (26); last present in the aploactinid Prosoproctuspa- dorsal spine to last dorsal ray 2.3 (27), to taecus (see Poss & Eschmeyer, 1979). Dis- pelvic insertion 4.0 (47), to last anal ray 2.4 tinctly tubed head pores occur in a number (28), to anal origin 2.0 (24); last dorsal ray of aploactinids, for example Xenaploactis to last anal ray 1.2 (14), to anal origin 2.4 anopta (see Poss & Eschmeyer, 1980). (28); anal-fin origin to last anal ray 1.9 (22); Discussion.—The precise limits of the pelvic-fin insertion to anal-fin origin 3.2 (38). Aploactinidae and the details of their phy- Distribution.—Known only from the ho- logeny are uncertain due to the high degree lotype taken near the southern end of the of morphological differentiation and con- Great Barrier Reef at One Tree Island, Cap- siderable character incompatibility among ricorn Group, Queensland, Australia. specialized features otherwise useful in Etymology. —The Latin specific epithet, characterizing velvetfishes. Matsubara inusitatus, means unusual or rare. (1943a, 1943b) defined the aploactinids (as Comparisons.—A pelvic fin with only one a subfamily Aploactinae [sic] of the Scor- spine and a single soft ray (l)js known to paenidae) based on the close association of occur among scorpaenoids only in atypical the anteriormost two dorsal spines with the specimens of the aploactinids pot- cranium, their separation from succeeding tii and Kanekonia florida. dorsal spines, a highly reduced suspenso- Distinct, densely melanic spots on the rium, and lack of a supraoccipitalcrest (nei- head are unusual among scorpaenoids. It is ther of the latter two features readily visible likely that their presence in our specimen is from radiographs of Matsubarichthys inu- a transitory larval/juvenile character but this sitatus). These features probably are not 472 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON uniquely derived among scorpaenoids and to be the sister-group of the bathyaploac- do not characterize all aploactinids. Among tines. If so, the shapes of the highly mobile, scorpaenoids it is typical for a single pte- complex, and variously spined lacrimals rygiophore to insert between each succes- seen in both aploactinines and bathyaploac- sive pair of neural spines, except the ante- tines, and especially useful in diagnosing the riormost. However, among all species genera, may be quite plastic phylogeneti- assigned to the Aploactinidae as delimited cally. by Poss & Eschmeyer (1978), a single pair As Poss & Eschmeyer (1979) pointed out, of neural spines, invariably of vertebrae 6- it is difficult to argue that the presence of 8, have no interdigitating pterygiophore. restricted gill slits resulting from fusion of A combination of presumably special- the branchiostegal membranes to the isth- ized, but not uniquely derived, features mus is uniquely derived among scorpae- readily distinguishes aploactinids from oth- noids (these authors erroneously included er scorpaenoid fishes. These include the to- Gnathanacanthus goetzeei among scorpae- tal absence of branched fin rays, notably noid species with restricted gill openings). tubular and often highly ornamented lat- Nonetheless, the combination of features eral-line scales, non-spinous, tubular nos- associated with this fusion that occurs in all trils, pelvic fins with 1-3 soft rays, no slit members of the Synanceiinae (sensu Esch- behind the 4th gill arch, a single infrapha- meyer & Rama Rao 1973 and including ryngobranchial tooth plate, no palatine teeth, Erosinae of Matsubara 1943a, 1943b), the typically blunt as opposed to pungent head Minoinae, the Choridactylinae(sensu Esch- spines, and atypical, tack-like, pseudocy- meyer et al. 1979), and some aploactinid cloid or ctenoid scales, when scales are pres- scorpaenoids, may indicate common an- ent. These features also place Matsubarich- cestry. Many species of these lineages pos- thys inusitatus within the Aploactinidae. sess extremely large heads (greater than 40% Matsubarichthys inusitatus can not, how- SL), enlarged and variously mobile lacrimal ever, be assigned unequivocally to either spines, anterior neural spines in the region aploactinid subfamily. Matsubarichthys in- of precaudal vertebrae 5 to 8 that are no- usitatus is unlike members of the Aploac- tably thinner and more delicate than the tininae in lacking a unique papillose exten- others, a long and expanded neural spine on sion near the isthmus that forms a seal the first preural centrum, a strongly con- between the otherwise separate isthmus and solidated caudal skeleton with a reduced branchiostegal membranes (Poss & Esch- number of caudal fin-rays, palatines with- meyer 1978). Although M. inusitatus shares out teeth, a single bilaterally symmetric pair a highly restricted gill slit and a narrow pos- of infrapharyngobranchialtooth plates (ex- terior margin of the interhyal with members cept Inimicus spp. and Erosa erosa, which of the Bathyaploactininae,it lacks the great- have two bilaterally symmetric pairs, albeit ly enlarged and highly movable first infra- the anterior pair substantially reduced in the orbital bone, rather spatulate head spines, latter genus), and loose, virtually naked skin and characteristic orientation, shape, and (lateral-line scales and vestigial scales in a reduced size of the opercle that characterize few species excepted). the two bathyaploactine genera Acantho- Like other aploactinids, M. inusitatus is sphex and Bathyaploactis. Interestingly, the more specialized than the synanceiines, first epibranchial of Matsubarichthys inu- choridactylines, and minoines, at least in sitatus bears a much reduced posterome- having one pair of consecutive neural spines dially directed uncinate process similiar to (6 and 7 in M. inusitatus) of the precaudal that of Acanthosphex leurynnis and Kane- vertebrae that lacks an intervening dorsal konia jlorida. Matsubarichthys may prove pterygiophore. Like some aploactinids, the VOLUME 104, NUMBER 3 473 anterior nostril is far forward on the snout. ceedings of the California Academy of Sciences The combination of character states in M. 41(21):475-500. Matsubara, K. 1943a. Studies on the scorpaenoid inusitatus results in incompatibility among fishes of . Anatomy, phylogeny and tax- shared, derived characters that would oth- onomy (I).—Transactions Sigenkagaku Ken- erwise be useful in establishing the cladistic kyusyo 1:1-170. relationships of the relatively advanced . 1943b. Studies on the scorpaenoid fishes of members of the Cocotropus stem. Another Japan. Anatomy, phylogeny and (II).—Transactions Sigenkagaku Kenkyusyo 2: rare scorpaenoid, Eschmeyer nexus, which 171-486. exhibits a number of features discussed Poss, S. G. 1982. A new species of the aploactinid above (Poss & Springer 1983), poses similar fish genus Kanekonia from Halmahera, Indo- problems. Additional specimens of Mat- nesia and a redescription of Kanekonia flori- subarichthys inusitatus and other extremely da.—Japanese Journal of Ichthyology 28(4): 1-6. , & W. N. Eschmeyer. 1978. Two new Aus- rare basal aploactinids are needed for more tralian velvetfishes, genus Paraploactis (Scor- rigorous analyses. paeniformes: Aploactinidae), with a revision of the genus and comments on the genera and spe- Acknowledgments cies of the Aploactinidae.—Proceedings of the California Academy of Sciences41(18):401-426. We thank J. Leis of the Australian Mu- ,& . 1979. Prosoproctuspataecus.anew seum for kindly making this unique speci- genus and species of velvetfish from the (Aploactinidae: Scorpaeniformes).— men available to us and L. Knapp for de- Japanese Journal of Ichthyology 26:11-14. livering it. We further thank V. G. Springer , & . 1980. Xenaploactis, a new genus and W. N. Eschmeyer for review of this for Prosopodasys asperrimus Gunther (Pisces: manuscript. JoEllen Trecartin is acknowl- Aploactinidae), with description of two new spe- edged for illustrating the holotype and Sara cies. —Proceedings of the California Academy of Sciences 48:287-243. LeCroy for rendering its axial skeleton. This , & V. G. Springer. 1983. Eschmeyer nexus, a work was supported, in part, through an NSF new genus and species of scorpaenid fish from grant (BSR 8705373) to SGP. Fiji.—Proceedings of the Biological Society of Washington 96:309-316. Tucker, J. W, & J. L. Laroche. 1984. Radiographic Literature Cited techniques in studies of young fishes. Pp. 37-39 Eschmeyer, W. N. 1969. A systematic review of the in H. G. Moser et al., eds., Ontogeny and sys- scorpionfishes of the Atlantic Ocean (Pisces: tematics of fishes. Special Publication No. 1, Scorpaenidae).-Occasional Papers of the Cal- American Society of Ichthyologists and Her- ifornia Academy of Sciences 79:1-130. petologists. , & K. V. Rama-Rao. 1973. Two new stone- fishes (Pisces, Scorpaenidae) from the Indo-West (SGP) Gulf Coast Research Laboratory, Pacific, with a synopsis of the subfamily Syn- P.O. Box 7000, Ocean Springs, Mississippi anceiinae.—Proceedings of the California Acad- 39564; (GDJ) Department of Vertebrate emy of Sciences 39(18):337-382. Zoology (Fishes), National Museum of Nat- , , & L. E Hallacher. 1979. Fishes of the scorpionfish subfamily Choridactylinae from ural History, Smithsonian Institution, the Western Pacific and . —Pro- Washington, DC. 20560.