450
NOAA Technical Report NMFS Circular 450 The Utility of Developmental Osteology in Taxonomic and Systematic Studies of Teleost Larvae: A Review
Jean R. Dunn
June 1983
U.S. DEPARTMENT OF COMMERCE National Oceanic and Atmospheric Administration National Marine Fisheries Service E RT" NOAA Technical Report NMFS Circular 450
The Utility of Developmental Osteology in Taxonomic and Systematic Studies of Teleost Larvae: A Review
Jean R. Dunn
June 1983
U.S. DEPARTMENT OF COMMERCE Malcolm Baldrige, Secretary National Oceanic and Atmospheric Administration John V. Byrne, Administrator National Marine Fisheries Service William G. Gordon, Assistant Administrator for Fisheries The National Marine Fisheries Service (NMFS) does not approve, rec ommencl or endorse any proprietary product or proprietary material mentioned in this publication. No reference shall he made to NMFS, or to this publication furnished by NMFS, in any advertising or sales pro motion which would indicate or imply that N;\lFS approves, recommends or endorses any proprietary product or proprietary material mentioned herein, or which has as its purpose an intent to cause directly or indirectly the advertised product to be used or purchased because of this NMFS publication. CONTENTS
Introduction ...... Methods of preparing larval fishes for skeletal study...... 2 Skeletal structures of tel~ost larvae as taxonomic aids and systematic characters...... 3 Vertebral column and associated bones ...... 3 Fins and their supports...... 6 Caudal fin ...... " 6 Dorsal and anal fins ...... 6 Pectoral fins and their supports ...... 11 Pelvic fins and their supports ...... 12 Other skeletal structures ...... 12 Branchiostegal rays ...... 12 Gill rakers and pharyngobranchial apparatus ...... 12 Oromandibular teeth ...... 12 Predorsal bones...... 13 Head spines and sculpturing ...... 13 Scales and lateral line pores ...... 13 Cranial bones ...... 13 Examples of the use of developmental osteology in systematic studies of teleosts ...... 13 Acknowledgments ...... 15 Literature cited ...... 16
Figures
1. Skeletal structure of Sebastes sp...... 4 2. Structure of (A) parts of a teleost vertebral column and (B) precaudal and caudal vertebrae...... 5 3. Left lateral view of the caudal complex of a 38 mm SL Argentina silas...... 7 4. Left lateral view of the caudal complex of a 68.1 mm SL Scombrolabrax heterolepis from the Atlantic Ocean . . . .. 7 5. A. Lateral and anterior views of the serial association between pterygiophore and fin ray from an 85.0 mm SL Thun nus atlanticus. B. Lateral view of pterygiophores of the dorsal finlets from an 85.0 mm SL Thunnus atlanticus ... . 10 6. A. Lateral view of ontogeny of left pectoral girdle from Scombrolabrax heterolepis. B. Ventral view of left and right basi pterygia from Scombrolabrax heterolepis ...... 11 7. Numerically constructed dendogram depicting larval morphological relationships among 12 genera of the subfamily Scombrinae ...... 14
Tables
1. Characters of pelagic life history stages that aid in identification to order or suborder...... 3 2. Characters of fins that aid in identification of larvae to order or suborder ...... 6 3. Larval characters presumed phylogenetically important as coded state for comparison of 12 genera of the subfamily Scombrinae ...... 14 4. Comparisons of 12 genera of the subfamily Scombrinae ...... 14 5. Potential characters for analysis of phyletic affinities of teleost larvae ...... 15
III FOCU ) The Utility of Developmental Osteology in Taxonomic and Systematic Studies of Teleost Larvae: A Review
JEAN R. DUNN I
ABSTRACT
Numerous examples from the literature indicate that cartilaginous and bon} ; tructure; aid in the ta\ on· omy of teleost larvae. The vertebral column, neural and haemal spines, ribs, predorsal bones, ba;al ,> upport ... and constituent bones of the median and paired fins , and the bones of the skull ha ve all been u;ed to va r y ing degrees in larval fish taxonomy. The size, number, and morphology of cartilaginous or bony struct ure; and their sequence of ossification have been employed to elucidate identification of larvae. Such tructure .. haH been used as characters at every taxonomic level, although no single structure has been found to be useful at a ll levels. Various methods for preparing and studying cartilaginous or bony structures a re available (e.g., clearing and staining, and radiography). The use of biological stains for detecting cartilage a nd/o r bone, hOl.eHr, ;., the technique most commonly used for larval fishes. Studies of developmental osteology have been used only recently to infer phylogenetic relati on .. hip,> of teleosts. Morphological characters and their development have great systematic utility at all hierarchical lev· els. Osteological studies of fish larvae and rigorous documentation and analyses of ; uch studies in the primary literature have the potential to notably increase our understanding of teleost phylogen}.
INTRODUCTION relati onships of organisms, their distributIOn. c l a~~ificatltm. and evolution (Bl ackwelder 1967; see abo Slmp~on 1961 and ~la)r Due to similarities in appearance, fish larvae of some groups are 1969 for alternate definitions). Phylogen} (fnllov.lng la) r 1t)ot) difficult to identify- using either pigment patterns or morphology. is the study of the lines of evolution of a group of organI~m~ the Examples of such difficult groups occurring in the marine environ o ri gin and evolution of higher taxa. ment include (but are not limited to) clupeids, scorpaenids, gadids, Much of th e activity of larval fish taxonomlqs lOda) IS stH:alkd scombrids, cottids, and hexagrammids. Because of the problems alpha taxonomy (Mayr et a!. 1953), IdentifYing fj,h lanae to pI.' encountered when attempting to identify larvae of these groups, cies. In the eastern subarctiC PaCific. on I} about 50"'< of the Ian ..Ie detailed studies of developmental osteology are often required . coll ected can be Identified to species (Kendall et ..II 19XO) duc Ossified structures may be defined as those skeletal structures primaril y to such abundant. pccles-nch. and poorl} kno\\ n tanll composed of bone (e.g., skull and trunk bones); in practice they are li es as Scorpaenl dae. Cottlda" SlIchaeldae. and AgonIdac distinguished in the laboratory by their uptake 01 biological stains. Increased examinati on of osteological structurcs and their dtlL'U usually Alizarin Red . Endochondral bone is preformed in cartilage, mentation in th e formalltterature may expand our undeNandIng nt whereas dermal bones ossify directly from membrane (Wake the re lationships ofteleosts as well as enable the pecitic Identltl a 1979). Cartilaginous structures may also be distinguished by their tion of a larger proportion of larvae. uptake of vital stains, usually Alcian or Toluidine Blue, allowing The util it} of developmental osteology In taxOntlnllc and S) tcm one to count or examine the structure of bones which may not ossify ati c tudies of te leost larvae has not pre\lousl) been rL'V IL'W d until late in the larval period, until well into the juvenile stage, or Therefore. in addition to descnbIng laboratof) mL'thods oInlllonly never (e.g., epurals, predorsal bones, radials of pterygiophores, used tll facilitate the stud} of larttiagInous and bony tructure and etc.). Meristic segments, those structures commonly defined as in revieW ing pertinent structurcs of the a\lal skeleton, mcdlan and countable structures normally occurring in a series (e.g. , vertebrae, paired fins. and craniu m. I have attemptcd to ,) nthL'SIIl' thc \alla neural and haemal spines, fin spines and rays and their supporting ble knowledge regarding the \ alue of the,e ,tructun:s In t xonoml bones), are often a powerful tool for identifying teleost larvae (e. g., and systematic studies of fish Ian ae. In prepanng thl re\ le\\, I Matsumoto et a1. 1972; Richards and Potthoff 1974; Berrien 1978; have drav. n freely upon previousl) publl,hed papc~ on tdxonOnl\~ Potthoff 1980; Fritzsche and Johnson 1980) . Analysis of osteologi aids In larval fish studies (Serf) and Richard, I t)7. hi trom and cal tructures can also help place unknown fish larvae in the correct Moser 1976) as well as on unpubli. hed notes trom cia e (;on order, suborder, and . sometimes, family (Ahlstrom and Moser dueted by the late E. H . Ahbtrom~ 1976). Osteological structures are of demonstrable valu e in phylo genetic studies and can be of great utility in elucidating systematic relationships based on studie of teleost larvae (Moser and Ahl strom 1970; Ahlstrom et a!. 1976; Kendall 1976, 1979: Potthoffet a!. 1980). Taxonomy is here taken as the study of groups of organi ms. their identification and variation; systematic as the tudv of the
I orth"e, t and AIa , ~a Fi,herie, Center. "'auonal ~I a nn e Fi, he ri e, er\lce, OAA.2725 lontlake Boule, ard Ea,t. eattle. \\A 9 11 2 11 n. ofth' ~ rm r pn.. ' cdure The. e pmbJem mdude the deo .. itt- Lr R\ L tlon of bone . du to e ce ,,3cldu) dunng p.:cllllen fi attOn. and IlDY th po Iblht:- thaI \\ a hmg of pc'imen. follo\\ ing fixation and pnor I ) ~<1ntia.!e tainin." ma) render cJrtllage un\lamabJe. tore Ii? can.:h L nt:cd<:d. ixlth on cartibge and bone laIn and on melh o
Table l.--Characters of pelagic life history stages that aid in identification to order or suborder (from Ahlstrom and Moser 1976).
Character Pleuronecti Myctophiformes Berycifonnes Perciformes Scorpaeniformes C lupei formes Argentinoidcl StOll' atoidcl AII&Ulll1J o,..e. adl( .,..", fo rme s
I,.arvae Predomi nant body Various, Various, Slender Varlous, Various. Elongate. Elongate. Elongate, Leptocephoro1u '.1' ou • shape markedly often elongate to stubby usually stubby usually stubby slender slender slender eic iah l Compressed dl!"ep-bod I red
Snout to anus length Usually 40-70%5L ca. 30-60%.5L Various I 35-60%5L 65-95%5l. 70-90%51. 30-95%5l. 40 9~"-S1. Ua.u i <4O%SL 20-60%5L (usually long) ",. Character of gut coiled Straight, Coiled Various, Coiled Straight Strught .. variously usually coiled shaped
Trail ing gut Not trailing, Seldom trailing, Not trailing Not trailing Not trai ling Not tralling Not tralllng Often lra111n&. Seld<;41 tr.. ,1 i but gu t can be reverse can sometlmes {_rlu~dly distended apply; gut urkedly traLlu.. g on frolll body gradually so.e cor'ir.dl increases in relative length on larvae
Number of vertebrae 25 to 65 /olyctophids Usually ca. 20 to 100+, ca. 2S to 65 ca. 40 to 60 ca. 40 to 85 c. 30 to 100. HI to 400 .a 40 .. 28 to 45, 23 to 33 often 24 to 28 • Olt ...... others 2 9 to 121 )Q to
l.arval stage c haracters Larval eyes Round Round to ca. Round Usually round, ca. Round ca Round Round or Round to Round r markedly can be narrowed, aanr.:edly .aderate y narrowed; narrowed sometlmes narf'("cd \.Ir .... d often choroid wi th choroid st.alked (sulkl."d n t. '" tissue under eye, tissue ~ infreq. stalked
l.arval head spi.nation frequently- Varlous various various Usually Sone various, none to none to none to useful In useful In i dent.. markedly markedly markedly ident heavy heavy heavy
Early fO l'lD ing fin Often: Occasionally Often V Sometimes No. No ccallHI!\a Y rays or spines 1 to 12 ant. P fin rays &. ant. 0 1 or ClOre but P fln C:U\ P n (often ornate) Drays 1 st 0 sp. and be qUl te l&r&e Id'thjocOC~'" Sometillles: V sp. &. rays 2 or 3 V rays
TransfOMU.bon stage Marked Various. Usually Usually Gradual Marked ....u .... e'd ..... (1 eye shi fts to often marked. gradual gradu.l D. It. , \ photophore right or left) sometlmes ftns .a.... e. _t delayed ancho\') .. sometimes snout (onu pro led prolonged
Early Juvemle stage No, In soee foras Sometimes SOIIeti.cs Pelailc (preJuvenile of but larval stage (ex. Nacristiella marked (ex. (In V1I.r Juvenile sUge HUbbs , 1958 ) can be IUrkedl)" stage) Holocentrids ) f_lliesl (ex. soae prolonged scorpaenlds 1
3 (Hubbs et al. 1974; Williams and Bond 1980) . The sometimes ture. Vertebral counts for about 553 species of fishes occurring in diagnostic surface features of centra have been used by archeolo the eastern Pacific Ocean were listed by Clothier (1950) and Cloth 6 gists and paleontologists for taxonomic identification (Casteel ier and Baxter (1969 ). Hotta (1961) listed vertebral counts for 266 1976). However, like other structures, centra undergo ontogenetic species in 106 families of fishes in Japanese waters, while Taka changes in shape even after ossification has occurred (Clothier hashi (1962) listed such counts for 256 species in 100 families. 1946, 1950; Bond and Uyeno 1981). Meristic characters (including precaudal and caudal vertebral The length of individual centra may vary according to stage of counts) for 642 species of marine fishes of the northwest Atlantic development (Kramer 1960); in Scomber japonicus, centra near the Ocean were listed by Miller and Jorgenson (1973), and means (and middle of the vertebral column initially outgrew those more ante range of means) of vertebral counts of3, 137 fish species belonging rior or posterior. In some fishes sexual dimorphism occurs in the to 118 families were presented by Lindsey (1975). Lists also exist structure of vertebral elements, such as in the relative size of the for certain specific groups such as North Pacific blennioids 7 haemal canal of the first caudal vertebra in Labrus mixtus (Ford (Makushok 1958) and cottids (Howe and Richardson 1978 ). 1937). Other variations in structure, such as the size and shape of The shape and structure of projecting processes of the vertebrae, the first haemal spine and the modifications of neural and haemal termed "apophyses" (Wake 1979), can be of taxonomic signifi spines, were included by Clothier (1950) in his key to adult fishes cance. Neurapophyses are dorsal projections from the centrum and noted by Matsumoto (1963) and Potthoff (1974) for the first which join to form the neural arch; haemapophyses are ventro haemal spine in Thunnus alalunga. lateral projections from the centrum which join to form the haemal The separation of vertebral counts into precaudal (abdominal) arch (Fig. 2A). Parapophyses are bony projections on each side of and caudal components (Fig. 1) has useful application in the taxon the precaudal centra to which pleural ribs are attached; zygapophy omy of a number of groups of larvae, even at the species level (Pot ses are projections of the centra which , in some fishes, may inter thoff 1974; Sumida et al. 1979). As commonly defined, the caudal lock with each other and give rigidity to the vertebral column (Fig. centra lack pleural ribs and possess medial haemal spines (Hubbs 2B). These may be separated into neural and haemal pre- and post and Lagler 1949; Berry and Richards 1973) . In various taxa, the zygapophyses (Clothier 1950) . The size and shape of the apophyses haemal spine of the first caudal vertebra is anterior to, or articulates were used as a taxonomic character by Clothier (1950) and Clothier with, the first anal pterygiophore (Fig. 1, 2A). In some species of and Baxter (footnote 6) in their keys to adult California fishes based nomeid fishes, from one to five caudal vertebrae bear both a haemal on the vertebral column. Potthoff (1974) described the ontogeny of arch and a pair of pleural ribs (Ahlstrom et al. 1976) . These authors apophyses in tunas (Thunnus) and used the position of the first defined caudal vertebrae in nomeids as those possessing a haemal haemal postzygapophyses as one of several characters to separate spine regardless of the presence of ribs . They noted that precise three species of Thunnus. Bond and Uyeno (1981) used the remark determination of precaudal and caudal vertebrae could easily be able changes in parapophysis shape which occur during growth in made in cleared and stained specimens by following the sequence Scombrolabrax heterolepis as one character justifying placement of of ossification of samples because the haemal spines ossify earlier this monotypic species in a separate suborder and family. than the ribs. The number, size, and shape of neural and haemal spines may be Vertebral counts are of great value in separating larvae which are useful taxonomic characters in larval fishes as they have been superficially similar but phylogenetically remote, such as clupeids and pholids (Russell 1976) ; vertebral counts also aid in identifica 6C IOlhier, C. R., and 1. L. Baxter. 1969. Vertebral characters of some Californian tion of closely related taxa such as species of clupeids (Houde et al. fishes with notes on other Eastern Pacific species. Unpubl. manuscr., 264 p. Calif. 1974). The intraspecific variation in vertebral counts must be Dep. Fish Game, Sacramento. known when using this character to separate closely related taxa 7Howe, K. M ., and S. L. Ri chardson. 1978. Taxonomic review and meristic vari ati on in marine sculpins (Osteichthyes: Cottidae) of the Northeast Pacific Ocean. (Berry and Richards 1973). Unfortunately, few lists of vertebral Unpubl. manuscr., 142 p. School of Oceanography, Oregon State Univ., Corvallis, counts and their intraspecific variation are available in the Iitera- OR 97331.
Predorsal bone Dorsal fin spines Neural spines Dorsal fin rays
Supraoccipital Dorsal fin pterygiophores
Haem.1 spines
Epipleural ribs Anal fin pterygiophores
~ Abdominal vertebrae ---'------Caudal vertebrae ______--'
Figure I.-Skeletal structure of Sebastes sp.
4 A Neural zygapophyses
p"apo ~yse s / First haemal arch
B
Precaudal
I ntermuscu lar bone
Fi!(ure 2.- \. Structure of par" uf " a teleo,t Hrltbral column (aft" Clothier 1951)); B. "'tructurt of pH Haemal spine caudal and caudal Hrtthrae faftt r Bond 1979).
shown to be for adult fishes by Clothier (1950) and Clothier and same vertebra in some fishes (e.g., £Il~nlllil lIIordlLl. 5\/lod/l Baxter (footnote 6) from which the following has been extracted. In lucioceps, Merluccius produulIs), aCLOrdlng to CI(lthH!r (I q~OJ adults, the most anterior neural spines may differ in shape, length. but more commonl) the first haemal spme occurs po tenor 10 the and width from the more posterior spines. The neural and haemal first haemal arch (Fig. 2A). The po. Itlon of the,e 1\\0 (/1 ratl r, spines may be modified into plates (as in Gymnothorax lI1ordax), or was used in a ke) to adult fishe, ba,ed on ~eJctal element b) the first few anterior neural spines may be fused into a bony plate Clothier (1950) and Clothier and B,nter (lootnotc 6) Accordmg t (e.g., Verrunculus polylepis). The terminal three or so neural or Clothier (1950). the first haemal ,pine ma) be indeterminate \\ hen haemal spines may be truncated distall y and laterally broadened, it is onl) a minute process, \\ ith sub,equent pme b L(>flUng gradu whereas those anteriorly situated may be more spinous. In the cottid all) longer (Sarda, ThuIlI/US, '\('olhuIlIllO). or bee u c the ha mal [cel inus quadriseriarus, the last three neural and haemal spines are process fuse proximall) onl) J.nd remalm, blfurcatc dl tall) enlarged into wide plates. In bothid and pleuronectid flatfishe , the (Albula I'lIlpes) In some atherinids (4lhcf1IlOl'f tlhcnI {11/ and first haemal spine is distinctively strong and heavy, and its structure Leureslhes). the position of the fiN haemal spmc t ob ured b) the varie interspecifically. Whether or not the centra adjacent to the "haemal funnel" (Clothier (950). In ath nmd thl tru,,(U ural centrum (or centra) bears a neural spine is a taxonomic charac beginS 6 to 10 \crtebrae in lront \.11 the first hemal pm .md I ter of value at the ordinal I vel (Ahlstrom and Moser 1976). as e scntiall) a double an.h. the 10\\ amo t part of \ hi h d n t fu shown in Table 2. If reduced or lacking, thi IS often termed the until the fiN haemal pine form (ther \ anatJon 10 tru ture "rudimentary neural arch" (Greenwood and Rosen 1971) and. neural and haemal . pme. are do.:um nted b) rord th r when enlarged or variou Iy shaped and aok) losed to other bone. is (1950). and Clothier and Ba ter (footn te 61 called the "supraneurallamina" b) these authors. The po. teriormo. t one or mor~ ha rna pin rna) The first haemal spine and first haemal arch may occur on the nou. or ank) losed to the t:entra FI,mge rna) be p nt Table 2.--Characters of fins tho.lt Id In iucnLlflf.:ullQIl of I H'VIt~ l onJr.r- 01 lubfJrtir.r (fro. Ahlalro. ntJ M Ic t 1 1 ).
~.,.,.... plC'uf"O""o;-tl lItjctc'f'hj 10".... fl.,.)' "o"e' 'enll '0,.... "" , M .. ch" ... acte'" (0,.. •• ,, ..,. Ipl"'" , "'1' .. ·"1' Typ. 0' fH\ elc.... nt. R>loY, ".y' ."Ine. " r-,.. Poctoral ,. "''',Y'' Not I.t. hh .. Varl0,n .. l.al(' ... Sequ.nc. of 1"0"'" '0' r' tl0n )(Ien ..... 1
v.ntral (tn. S.que"". of (O~ $I ... tl.". ".'-1)' ta,.,y to la t ..
POultOn on bud.)' Thora, tt At>.I,. ~ to JUI"I.r
fo ....ub • ~!I. " .. Vanou" . ~ , .. I, . , 00 ..... 1 hn{.) , fl_ ". , ( Ana' ( .. t nn, o •• .. o ••
o , A t .... tnal roy '0 blrurcatt- Adlpo,. ". '0 'Wllly pr••• nt Ca"d.l fi. 'YP' ."" ho.ot...... ereel
Prlnclp.ll C AY' ", no b, ,..". 111 •• 1_ 1'10 e:pural. '0 un.i cent .... ., , (ae. te.t . .. !-Ieun on vertebra _ .. ·' ..... 1 ...... 11. r h 110 .dJ "01.1,...1 ., I.' Ipu2 of M",nod, ... tenor three or four neuml and haeTllal 'pine, (,I '0111<' II hc (c g . are of rnmc u c tn Id nuri aU n to ordtnal and ubordtna) level , argentinoids) and have been termed "preural 1I.lng<' .. b} (lI"\~cn "hcre.J th lime 01 ~ rmaU n of pc lor.J) ftn I considered an Ordi wood and Rosen (1971). The~e "nclOge .. \\er Uscu hj thc ~ nal ch.Jra I r Jabl 2) I urth r. thc caudal fin od 11 uppontn authors as a character to infer ph) logcnctic rel..ltlOn\hlp In cpar.lt bones are u ful tn Id OUf) 109 lar.:t I filOlIl}, w hll the median ing adult argentinoid fi~hes fmm alcp 6 I I PUF , .0 rnrn for some fishes , e.g., 6- 8 dorsal, 6-8 ventral in Hemilepidotus; 11 The number of hypural bones is generally consistent among dorsal, 7 ventral in Chitonotus; and 10- 12 dorsal, 9- 10 ventral in related phylogenetic groups, and variation from the norm is a useful Scorpaenichthys. The related agonids and cyclopterids appear to systematic character. A high number of hypural bones is often taken possess few secondary caudal rays, but the ventral count is often 1 as a primitive character (i.e., seven in Elops and salmonids); reduc or 2 less than the dorsal count (footnote 8) . In the scorpaenids (a nd tion is advanced (i.e., four in pleuronectids) as shown in Table 2. hexagranunids) , the dorsal and ventral secondary caudal counts are The hypural bones may be va riously fu sed , ultimately into a single usually equal (foot~ote 8). The distribution and variation of sec plate in adults of some species, e.g., Coryphaena (Potthoff 1980). ondary rays have not been investigated in most taxa. When fusion of hypural bones occurs, it may be observed in ontog The homocercal caudal fin is generally considered a conservative eny in some fishes (e.g., Thunnus atlanticus, Potthoff 1975 ; Cor structure, and the distribution of upper and lower principal caudal yphaena , Potthoff 1980) but not in others (e.g., Ophichthus rays is usually consistent within broad phylogenetic groups gomesi, Leiby 1979; Microgadus proximus, Matarese et al. 1981). (Gosline 1960; Berry and Richards 1973 ; Ahlstrom and Moser This reduction of constituent elements by fusion during ontogeny 1976). The count of 10 (upper) and 9 (lower) principal caudal rays may be useful in anal yzi ng relationships at the generic level (Moser occurs in almost all clupeiform and salmoniform fishes, as well as and Ahlstrom 1970). in most berycoids. Perciforms have 9 + 8 principal caudal rays and The centra supporting the hypural bones in teleosts are generally scorpaenifo rms vary but have 17 or fewer principal rays, whereas believed to undergo reduction in numbers by fu sion during the pleuronectifo rms have highly variable counts ranging from 10 to course of evolution. Again , a hi gh number is usually considered a 23 principal caudal rays. Exceptions occur, however. In the cottid primitive character (e.g. , three in salmonids, occasionally four in genus Ascelichthys, the count of principal caudal rays is 6 + 7 clupeoids); a single centrum is considered advanced (e.g., Perci (Matarese\ and a greater number of lower than upper principal formes , pleuronectids) as indicated in Table 2. These are termed caudal rays may occur in other cottid genera (footnote 8) as it also " ural centra" (Nybelin 1963) or the "urostyle" when they are fu sed occurs in ceratiods (footnote 3) and in some catfishes (Lundberg into one (Gosline 1971). The fusio n of ural centra can often be and Baskin 1969). The distribution of principal caudal rays on observed in ontogenetic series of cleared and stained ·Iarvae. For hypural bones may vary among genera, as in some cottids (footnote example, in myctophids, the two ural centra usually fu se into a sin 8), or among species, as in the Hexagrammidae (Kendall 1<) . gle centrum before the early juvenile stage, although the two free Detailed examination of the distribution of principal caudal rays in centra may persist in some species well into the juvenile stage various groups of teleosts may result in useful taxonomic charac before fusing (Moser and Ahlstrom 1970). ters. Nybelin (1963) and Monod (1968) suggested that the urostyle Departures from the normal homocercal caudal fin ofteleosts are consists of a fusion of a number of ural centra and preural centrum taxonomically helpful both in hypural structure and the distribution of the associated rays. The isocercal (gadoid) caudal fin is unique one. Hence, in certain groups of fishes some investigators consider the ultimate preural centrum as number two and the urostyle to be a among teleosts in that the hypural bones support relatively few fusion of preural centrum one and of the ural centra (Rosen and Pat branched rays, some of which may occur dorsad of the hypurals terson 1969; Houde and Potthoff 1976; Potthoff 1975 , 1980) as (Ahlstrom and Counts 1955). Leptocercal caudal fins in eels are shown in Figure 4. Other investigators (Moser and Ahlstrom 1970; markedly reduced , as are those in some blennies. Molids possess a unique "pseudocaudal," termed a " clavus" characterized by fail Ahlstrom and Moser 1976; Sumida et al. 1979; Markle 1980) con ure of the notochord to fl ex, a lack of hypural bones, and posses sider the ultimate preural centrum as number one (Fig . 3). sion of bones posterior to the last centrum which do not articulate Epurals, which support epaxial fi n rays, are small median bones directly with fin ray bases (Leis 1977). Much descriptive caudal-fin dorsal to the ultimate vertebra and posterior to the posteriormost morphology remains to be completed, although the works on adult complete neural spine or specialized neural arch (Patterson 1968). fishes provide a good basis (Hollister 1936; Gosline 1961 ; Monod Considered to be neural spines detached from their neural arches 1968; Patterson 1968; Rosen and Patterson 1969; Nybelin 1963 , (Gosline 1960; Nybelin 1963; Patterson 1968), they are of taxo 1971; Rosen 1973). nomic and phylogenetic significance in some groups of fishes and Hypural bones lie ve ntrad to the ural centrum or centra (or uro are generally constant in number within species. There are usually style if fused) and support principal caudal rays in most teleosts one to three epurals, but they may be reduced in number or size dur (Figs. 3, 4). Counts of hypural bones are made from ventral to dor ing ontogeny (Richardson et al. 1980), fusion can occur (Potthoff sal, but are usually expressed as superior plus inferior (e.g., 2 + 2, 1980), they may not ossify until well into the juvenile stage as in 4 + 3, etc.). The anteriormost hypural bone in some fishes is called some bathylagids (Dunn 1983), or they may be lacking. the "parhypural" by some authors (Monod 1968; Rosen and Patter Uroneurals, paired bones generally occurring dorsolaterally to son 1969; Potthoff 1975, 1980) but not others (Moser and Ahlstrom the ultimate vertebra, are often the first bones to ossify in the caudal 1970; Ahlstrom et al. 1976; Richardson et al . 1980; Markle 1980). complex. According to Patterson (1968) they are modified ural As defined by Nybelin (1963), the parhypural differs from the fol neural arches. They frequently fuse to the terminal centrum , and lowing hypural bones in that it contains a haemal canal and more the anterior uroneural may expand during ontogeny. The presence closely resembles a typical haemal spine than do the followi ng of three pairs of uroneurals is considered a primitive character, hypural bones (see also Monod 1968). In some phylogeneticall y whereas reduction in numbers or even loss (Pleuronectidae) is con diverse groups of fis hes, the pa rh ypural possesses a lateral flange sidered an advanced character. The ontogeny of uroneurals in usuall y terminating in a point posteriorly (Fig. 4)-the parhypura Hiodon, Elops, and Sa lmo was described by Cavender (1970), in pophy i of Nursall (1963). Archosargus by Houde and Potthoff (1976) , and in Coryphaena by Potthoff (1980). "A C l\!ataresc. North west anti Alaska Fish. Cent. , Na tl. Mar. Fi sh. Serv., NOAA, Seattle, \VA 98112, per;. commun. December 1980. The presence of urodermals or stegurals is considered a primitive In", \ \ Kentlall, J r, NO lthwest and Alaska Fb h. Cent., Natl. Mar. Fish. Serv., character. Urodermals (Nybelin 1963) are medial bones posterior to OA \, cattle, WA 98 1 12, per.;. commun . ovember 1980. the urostyle (Greenwood and Rosen 1971) and are thought to be 8 modified scales (Patterson 1968). Stegurals may be defined as the position (Table 2). The number of splOes 10 dor-al or anal flO I incorporation of the first ural neural arch and usually the first pre often constant in many familie or genera, but the number of ra) ural neural arch into the first uroneural . The definition of stegural may vary inter- or intraspeclfically. In fishes po, es 109 fin \\ Ith differs among authors (Monod 1968; Greenwood and Rosen 1971), both spines and rays, care should be taken 10 di nmlO:llin and questions remain as to the hQmology of "stegurals" among dif between the two. Separation of the two types based on the adult ferent groups of fishes (Rosen and Patterson 1969). In some fishes complement does not always apply to larvae. pines and It ra) (e.g., some gadoids, cynoglossids), medial bones occur dorsally may be fimbriated at the termlflUS 10 larvae, but the e are repl,Ked and/or ventrally anterior to the last neural and haemal spines. These later in spine by blunt or pOlOted tip according to BerT) and are termed "dorsal (or ventral) accessory bones" (Marshall and Richards (1973). Segmentation of soft rays does not nClur until Cohen 1964; Rosen and Patterson 1969) or "X" and "Y" bones after the ray has formed. In certain fishes (e go, Marone I(mlli" ). (Monod 1968) , and may be taxonomically significant at the generic some soft rays may become spines during or after the Ian al ,tage or subfamily level. (Mansueti 1958) . In Sebastes, the po tenormost dor-al 'PIO and Other characteristics of the caudal fin may be of taxonomic or the third anal spine first form as soft rays and then transform to phylogenetic significance. The procurrent spur of the caudal fin of spines, beginning at the base and continulflg distally (Moser et al some perciform fishes was described by Johnson (1975) who dis 1977; Richardson and Laroche 1979); these were termed "pre cussed its phylogenetic implications (Fig. 4). Nursall (1963) spines" by Richardson and Laroche (1979) . In many fishes thl: ter defined the hypurapophysis as "a lateral process of the anterior minal soft ray of the dorsal and anal fins consist of an antenor and hypural bone associated with the terminal vertebra (of Gosline posterior part but is associated with a single pteryglOphore (Table 2) 1960), serving as the anterolateral portion of the proximal attach and should be counted as one ray (Berry and Richards 1971) ment of the hypochordal longitudinal muscle." This process was Transient ossified structures occurring in dorsal and anal flOS of noted by Ford (1937) to occur in 24 families; Nursall found this certain groups may aid in identification. Larvae of some epinephe process present in 11 additional families. Whether or not the ante line serranids possess extremely elongate dorsal SplOCS v.hlch riormost hypural (parhypural of some authors) is free proximally is become reduced during ontogeny (Kendall 1979). Such dlstinctl\.e considered of phylogenetic significance by some workers (Rosen elongate dorsal spines also are characteristic of a number of other and Patterson 1969). groups including holocentrids, acanthurids, and balistids DlstlOC The structure of the adult caudal fin in both recent and fossil tive elongate dorsal rays occur in bothids, lophlids, and bregma fishes has been widely used to indicate phyletic relationships cerotids, and elongate anal rays are present in some lophiids. (Greenwood et al. 1966; Patterson 1968; Rosen and Patterson When using the structure, shape, and positIOn of dorsal and anal 1969; Gosline 1971; Greenwood and Rosen 1971 ; Rosen 1973) . fins as taxonomic characters, one should keep in mind that in some Relatively few authors have analyzed caudal fin ontogeny, and even fishes (e.g., c1upeids) these fins may migrate antcriorly or poste fewer have attempted to interpret the phylogenetic significance of riorIy during ontogeny (Houde et al. 1974) . Additlonally, dor-al the development of this fin. and anal fins may form in the fin fold and attach to the body b) Several recent papers on the development of the caudal fin in "streamers," as in argentinoids (Ahlstrom 1969; Moser [1981 n. teleost larvae indicate the structure differs from that commonly In some phylogenetically distant groups of fishes, the dor-al and accepted for adult specimens. Contrary to accepted interpretation anal fins may form early (e.g., Engraulus, some stromatllds. of adult caudal structure, no evidence was observed during ontog Syacium), whereas in other groups these fins may not OSSify until eny of a two-part fusion of the ventral hypurd.! plate in Thunnus transformation to the juvenile tage (e.g., Leuroglossus IChtlllJri, atlanticus (Potthoff 1975). Matarese et al. (1981) noted the absence Dunn 1983). Dorsal fins may form before or after the anal fin or of uroneurals in the caudal fin development of Microgadus prox they may form simultaneously. In species v.ith more than one dor imus, the presence of which was considered by Rosen and Patterson sal or anal fin, the sequence of ossification of the vanous fins rna) (1969) as a characteristic of the Gadiformes. lsopsetta isolepis also differ among taxa. The second dorsal fin forms (simultaneousl) lacked uroneurals according to Richardson et al . (1980), a structure with the anal) before the first dorsal fin in Scomber japOI1ICIlI. Fro considered by some authors to be characteristic of Pleuronecti churus symmetricus, Scombrolabrax heterolepl!> , and Anholllr 'III formes (Rosen and Patterson 1969; Amaoka 1969). The two pairs rhomboidalis (Kramer 1960; Ahlstrom and Ball 1954; Houde and of uroneurals in Coryphaena fused during ontogeny into a single Potthoff 1976; Potthoff et aI. 1980). In Merlucciu~ proJUUII the pair (Potthoff 1980). Based on examination of adult caudal fin mor first dorsal fin forms before the second dorsal fin and the anal 110 phology, the absence (i.e., fusion or loss) of the posterior pair of (Ahlstrom and Counts 1955), as It does in most gempyhds and .tIl uroneurals could be considered an evolutionary advance (Fraser scombrids, except Scomber and Rastrel/iger (Potthoff I; . fat u 1972). moto 1959, 1962, 1967; Voss 1954). In MlcrogatilH pro Willi • thl: The reported absence ofuroneurals in Microgadus and lsopsetta first anal fin is the first to form (after larval pectoml fin ) lollov.l:d does not necessarily imply that fusion or absorption of these struc by the second anal fin. Nearly ~imultaneously, the thmJ. e( nd. tures into the ural centrum (or centra) did not occur during evolu and first dorsal fins develop (Matarese et al 1981). In :J related tionary history. Similarly, the reported absence of ontogenetic gadid, Theragra chalcograrmna, anal fin one and t\HI and du al fusion (as in the ventral hypural plate of T. atlanticus) does not fin three form together, followed b, the ewnd and then th first refute the hypothesis that this plate is derived from the fusion of two dorsal fin (Dunn. pers. ob ) The sequence of 0 IticatJon of plO elements. Absence of such structures in adult fishes is still an evo and/or rays of the dorsal and anal fin vane among dlff rt'llt lutionary advance, even if it occurs by ontogenetic fusion or dele groups of fishe . Ossification rna) begin 10 the center of th flO and tion. progress anteriorad andlor po. tenorad (e.g., Seha It rant n Richard on and Laroche 1979: COf}phaena, Potlh ff 19 0) r II Dorsal and anal fins.- The number of dorsal and anal fins , and whether the fins are composed of rays only or of spines and ray , liT. Ponhoff. Southeast F"h Cent 'ham I Labornt ry atl \u F h rv are useful taxonomic characters and can be indicators of phyletic OAA. Miami, FL 331-19, pen; mmun Januaf) 19 :! 9 A Figure 5.-A. Lateral and anterior views of the serial association - --- M -- between pterygiophore and lin ray ._ .._._ .._ p --_ ... _-- from an 85.0 mm SL ThullIlUs at/anticus. 1) Fifth spiny ray of the lirst dorsal lin and its serial ptery giophore, lateral view on left and anterior view on right; 2) Eleventh r-1 ray of the second dorsal lin and its serial pterygiophore, lateral view 1 2 on left and anterior view on right; 3 3) Third finlet and parts of its >-----< I------t serial pterygiophore, lateral view a .SCmm O.SOmm on left and anterior view on right. S ; spiny ray; 0 ; distal radial; p ; proximal radial; R ; ray; M ; middle radial (from Potthoff fl14 15 1975). B. Lateral view of ptery B giophores of the dorsallinlets from F4 F5 F6 F7 FB an 85.0 mm SL Thullnus at/anticus. Left to right: PRI4 ; Proximal radial serial with the 14th ray of the MF2 MF5 MF6 MF7 second dorsal fin; NS22 ; neural spine of the 22nd centrum ; PFl ; proximal radial serial with the lirst lin let; ORIS ; distal radial serial with the last (15th) ray of the second dorsal lin, partly hidden by the base of the ray; R15 ; last ray (15th) of the second dorsal fin; MFl ; middle radial serial with the first finlet; Fl ; first finlet; 1-----4 1 XF8 ; fourth radial or stay of the IOmm last (8th) finlet (from Potthoff PF3 : NS22 NS26 XFB 1975). may be initiated at the anterior (Merluccius productus, Ahlstrom TUapia macrocephala, Eaton 1945 ; Blenniidae and Labridae, and Counts 1955; Citharichthys, Ahlstrom and Moser 1975) or Lindsey 1955); a fourth radial , called a "stay, " may be present posterior (ldiacanthusfasciola, Beebe 1934) portions of the fins. In (Fig. 5), as in the last finlet of Thunnus (Potthoff 1975) . The num those species possessing spines in the dorsal and anal fins, develop ber of pterygiophores is usually smaller than the number of fin rays ment may begin at or near the origin of the fins (Berry and Richards or spines, in that the anteriormost or posteriormost pterygiophore 1973) as it does for gempylids and scombrids (footnote 11). may support one or more fin elements (Fig. 1) . In S. japonicus, the The ontogeny of dorsal and anal finlets in the carangid, Elagatus pterygiophores are continuous between the first and second dorsal bipinnulata, was discussed by Berry (1969), in scombrids, Thun fins, but some do not support fin spines (Kramer 1960); such inter nus atlanticus, by Potthoff (1974), and in Scomber japonicus by neurals are also found between fins in some gadids (Dunn pers. Kramer (1960). obs.). Some pterygiophores (usually the first or last) may be bifur The supporting structures of the dorsal and anal fins, called cate or variously expanded (e.g., tetraodontids, Tyler 1980) ; and "pterygiophores" (Eaton 1945) or, by some authors, " interneu fusion of adjacent pterygiophores may also occur (e.g. , T atlanti rals" (e.g., Kramer 1960), are generally composed of three parts: cus, Potthoff 1975) . The anteriormost dorsal and anal ptery Proximal, middle, and distal radials (also called pterygiophores; giophores represent a fusion of two in many Perciformes (Potthoff Wake 1979; Fink and Weitzman 1982) as shown in Figure 5. The 1975; Fritzsche and Johnson 1980) . presence of three components is considered a primitive character The number of pterygiophores in the first interneural space dif (e.g., salmonids, Norden 1961; bathylagids, Borodulina 1969); the fered between two species of Coryphaena , according to Potthoff middle and proximal radials, however, may be fused into a single (1980). Pterygiophore ontogeny has been described for S. japoni structure (e.g., Scomber japonicus, Kramer 1960; Coryphaena, cus (Kramer 1960), Thunnus atlanticus (Potthoff 1975), Archo Potthoff 1980)-considered an advanced character. In some cases sargus rhomboidalis (Houde and Potthoff] 976), Morone spp. the three components may be fused into a single structure (e.g. , (Fritzsche and Johnson 1980), Scombrolabrax heterolepis (Pot- 10 A 0~ CI 1 I------l 2 3 I 2.0mm 0.5mm 0.25 mm B Ct CP IDW ~ EVW 1\ EDW VW AX P PX 1 ~ 2 ~ 3 ~ 1.0mm o 10mm o 10mm thoffet al. 1980), and CoryphaellCl (Potthoff 1980). The ontogeny of dorsal and anal finlet~ 10 the carangid Elagufls blplIIllulalCl \\ a~ di cu sed by Berry (1969); in the scombnd , 77ll11l1l1/S ulllIllIiclO, by Potthoff (197.f); and in Scolllherjapoll/cus b) Kramer (1960). rn addition to its u e a a taxonomic character, pter} glOphore mor phology may be an important tool 10 ph) loge netic studies (Potthoff 1975; Potthoffet a1. 1980) Pectoral fin and their upport .-Generall) pectora! fin form later in the larval stage than median fins (Table 2) and. therefore. are somewhat less \ aluable a a taxonomic aid 10 the Identlficatlon of larval fihe . Inome myctophld larvae. pectoral fins ma) be large. con plCUOU appendage \\hlch become reduced in .Ize dur mgontogen) (Moser and hlstrom 1970. loser [19 l)l Pectoral buds are the first fins to form 10 mo t IJn ac. but dey elopment of rJ) usuall) lag behmd those of other fms (Table. I 2) . Ho\\c\ er. 10 the m )hd RlIlI::'lIl1ill the pectoral fin. are thc t rst to dc\ clop raJ (LCI, 1977). Earl) dey eloping pcctoral fin. are al. 0 found in m~ dlll.\(t'.1 It\.10 er et al 19 7) and 10 loph Id. 1(lotnLtc 3) In me anguilhform.. Ian al pt.'Ctor.l1 tins are b\.:kmg, \\ herea 10 7i.1 to \(Olllll. idiacOII(hlls, and om malaeo teld the Ian I PC't ral re II Coryphaena; Potthoff et al. 1980 for Scombrolabrax; and Potthoff constant in other groups, including argentinoids (Nansenia 3 to 4, and Kelley 1982 for Xiphias gladius). Bathylagidae 2), percoids (usually 6 to 7; Bond 1979), Lampridi The reduction or total atrophy during ontogeny of the cartilagi formes (5 to 7), Pleuronectiformes (6 to 8), Batrachoidiformes (6), nous posterior process of the coracoid apparently occurs in most etc. (McAll ister 1968) . In giganturids and Saccopharyngiformes, fishes (Swinnerton 1906; Starks 1930; Potthoff 1980) , but such a branchiostegal rays are wanting in adults (McAllister 1968; see process is lacking)n developing Theragra chalcogramma (Dunn, also Orton 1963) . pers. obs.). The number of supporting elements of the pectoral fin (radials) may be a generic character in some fishes. As an example, Gill rakers and pharyngobranchial apparatus.-Because gill they range from two to five in various genera of lophiiform fishes rakers usually form relatively late in the larval period and the adult (Pietsch 1981). complements are often not reached until well into the juvenile stage, this characteristic is sometimes of little taxonomic value. Pelvic fins and their supports.- Counts of pelvic fin spines The presence of rudimentary gill rakers in some fishes and the diffi and/or rays are taxonomically useful in many fishes (Table 2). In culty of distinguishing them from adjacent normal gill rakers have most perciform families, pelvic fin counts are stabilized at one led to inconsistency in the literature (Moser and Ahlstrom 1970). spine and five rays (I, 5) according to Regan (1913). In pleuronec Additionally, gill rakers are lacking in some fishes (Giganturoidea) tids and bothids the pelvic fin counts are stabilized at 6 rays (in each and extremely reduced in others (e.g., Psettodes). In some groups, fin) , whereas in soleids the count is 5- 5 and in cynoglossids the however, counts of gill rakers on the epibranchials or ceratobranch count is 0-4 (footnote 3). In myctophids the pelvic fin ray count is ials or the length of particular gill rakers may be of value in discrim usually 8, but in some genera the count is reduced to 7 (Gonichthys) inating among allied species (Potthoff 1974; Richardson and or 6 (Notolychnus) (Moser and Ahlstrom 1970) . Pelvic fins are Laroche 1979). Gill arch structure among adult fishes has been the absent in a number of fishes (e.g., trichiurids, stromateids, sti subject of extensive reviews by Nelson (1967, 1969, 1970) , Green chaeids, Ammodytes, and some lophiids); others have only a single wood and Rosen (1971), and other workers ir. a search for phyloge spine and no rays as seen in some balistids (Nelson 1976). In some netic affinities. fishes (e.g., clupeoids) the pelvic fins may undergo migration dur Pharyngobranchial teeth are often one of the first elements to ing ontogeny (Russell 1976). ossify in teleost larvae (Moser 1972) . The sequence of calcification Elongate pelvic rays are found in gadids, macrourids, and trachip of branchial elements was described for myctophids by Moser and terids, and in some serranids and ceratioids; whereas elongate pel Ahlstrom (1970) and for Sebastes macdonaldi by Moser (1972). vic spines are present in some serranids, lutjanids, trichiurids, and Potthoff et al. (1980) described the ontogeny of gill arches in Scom others (footnote 3). Pelvic fins often ossify in an inward direction brolabrax. (Scombrolabrax, Potthoff et al . 1980); in some fishes the ossifica tion of all pelvic fin rays occurs nearly simultaneously. Oromandibular teeth.- Teeth on the maxillary, premaxillary, The pelvic bone or basipterygium (Fig. 6B) has also been used as vomer, palatine, dentary, and glossohyal bones are often used as a systematic character in adult fishes (Collette and Chao 1975; taxonomic characters in adult fishes. In general, the adult and juve Tyler 1962, 1980) , but has rarely been investigated in larval fishes. nile complement of teeth differs markedly from those in larval Sewertzoff (1934) discussed the pelvic bone among various groups fishes in number, size, and shape (e.g., Ahlstrom and Counts offishes, while Gosline (1961) discussed the structure of the pelvic 1958), although in some species teeth added during metamorphosis bone in adult " lower" fishes. may be retained in the adults (Moser and Ahlstrom 1970). In many teleosts, radials are present between the pelvic rays and Tooth ontogeny in teleosts is poorly known and is a subject ripe the pelvic girdle (Gosline 1961) , but they may be lost in various for further study. Teeth in many larvae initially emerge in a single diverse groups (Gosline 1961 ; Potthoff 1980). Additionally the pel series, but as more teeth are added they tend to emerge between vic bones may be fused (e.g., some tunas, Collette and Chao 1975), existent teeth and often form multiple rows. Teeth apparently are they may be highly modified (some balistoids, Tyler 1962, 1980), added slowly during transformation in some species (Scomber or lost (e.g., Mola, Tyler 1980) . japonicus, Kramer 1960) and more rapidly in others (lsopsetta iso The ontogeny and structure of pelvic bones were described by lepis, Richardson et al. 1980). Leptocephalus larvae of some fishes Potthoff (1980) for Coryphaena and Potthoff et al. (1980) for (e.g., Elopiformes, Anguilliformes, and Notacanthiformes ) pos Scombrolabrax, while Fritzsche and Johnson (1980) briefly sess fang-like teeth in the upper and lower jaws; they are apparently described the ontogeny of the basipterygia in Morone. not homologous with teeth in adult eels and are shed or resorbed during metamorphosis (Smith 1979). In larvae of Scopelosaurus, Other Skeletal Structures teeth on the maxillary are apparently shed or resorbed, as they are not present in adults (Berry 1964). In the bathylagid Leuroglossus Teleosts possess a great number of other skeletal structures, par schmidti teeth on the glossohyal disappear during transformation; ticularly of the head and axial skeleton, which may be of taxonomic hence the common name "smoothtongue" (Dunn 1983). and systematic use when applied to fish larvae. It is generally believed (Scott and Symons 1964; Jollie 1973; Fink 1981) that teeth of teleosts are continuously replaced, but Branchiostegal rays.-Branchiostegal rays often form their questions apparently remain as to such replacement (Hildebrand adult complement relatively early in the larval period, and their 1974). The tooth-replacement mechanism may also differ among position and number are of value in separating some groups of kinds of fishes (Lawson and ManJy 1973; Holmbakken and Fosse fishes, particularly in taxa considered advanced (McAllister 1968). 1973; see also Edmund 1960). Numbers ofbranchiostegal rays may vary in some groups, whereas Fink (1981) described four major tooth-attachment modes in in other groups they may be relatively constant. For example, num actinopterygian fishes and concluded that paedomorphic tooth ber of branchiostegal rays varies from 7 to 19 in engraulids, 5 to 24 development can be associated with major phylogenetic groups' of in stomiatoids, and 17 to 51 in ophichthid eels. They are relatively teleost fishes. Suga et al. (1981) analyzed the fluoride concentra- 12 tion in teeth of tetradontiform fishes and compared the levels used in taxonomic and phylogenetic studIes of adult ft~hes A, observed with the phylogenetic tree proposed for the order by Tyler demonstrated most recently by Fritzsche and Johnson (1980), the (1980) . The authors concluded that the fluoride concentration in ontogeny of cranial bones can proVIde dIagnostic dIfference, at the the enameloid is closely related to the propo ed phylogeny. species level. The structure and shape of cramal bones (particularly the supra Predorsal bones.- Predorsal bones are usually slender splinter occipital, epiotic, prootic, alisphenoid. and parasphenoid) v.ere bones which precede the pterygiophores supporting the spines or used by Clothier (1950) as a taxonomic character In his key to rays of the first dorsal fin (Fig. I) . Thought to be derived from southern California adult fi he . The lack of an orbitosphenoid and pterygiophores (Smith and Bailey 1961), they insert in interneural the presence of antorbitals in adult Leuroglossus schmic/ri were spaces, and their number and arrangement are in some groups of noted by Borodulina (1969), who used these trenchant characters fishes a taxonomic aid as well as a character relating to phylogeny. for removing Leuroglossus from the synonomy of Burhylagus . Pax Smith and Bailey (1961) investigated the relationships of predorsal ton (1972) separated myctophids into two subfamilies based on bones, dorsal fin spines, and dorsal fin pterygiophores in adult photophore arrangement and adult osteology (includIng jaw length fishes in the families of the superfamily Percoidea and illustrated and unsculptured circumorbitals with an extensive orbital shelf). some hypothetical evolutionary trends in dorsal fin supports. Pre These two ubfamilies corre ponded closely to the classIfication dorsal bones and spine-bearing pterygiophores in adult serranoid independently proposed by Moser and Ahlstrom (1970) based on fishes were examined by Kendall (1976) who found phylogenetic larval morphology, particularly eye hape. Recently Fritzsche and relationships. Counts of predorsal bones and their relationship to Johnson (1980) compared the early 0 sification of two species of spine- or ray-bearing pterygiophores were used by Ahlstrom et al. Morone and noted osteological differences between them. Includ (1976) as a diagnostic aid in separating stromateoid larvae. The ing the shape of th e ethmoid cartilage. position and shape of predorsal bones differed between larvae of A number of workers have described the ontogeny of cramal Morone saxarilis and M. americana (Fritzsche and Johnson 1980). bones in teleosts; some examples include Weisel (1967) for Catos Predorsal bones may often ossify either in late larvae or well into rom us and Poecilia, Moser and Ahlstrom (1970) for myctophlds, the juvenile stage but may form earl y in cartilage. Use of a Aprieto (1974) for carangids, and Mook (1977) for Archosurgus. cartilage-bone tain will enable detection of predorsal bones when Study of the development of cranial bones In teleost larvae formed, as shown by Fritzsche and Johnson (1980). Discriminating appears to offer a valuable avenue of pursuit, and Its increased use predorsal bones by radiography is sometimes difficult, but Kendall could assist in differentiating closely related taxa. (1976) successfully used radiographs in his study of adult serranids. EXAMPLES OF THE USE OF Head spines and sculpturing.-Spines located on the head of DEVELOPMENTAL OSTEOLOGY IN teleost larvae may be an important diagnostic character (Table 1) . SYSTEMATIC STUDIES OF TELEOSTS Ridges, crests, or variou sculpturing of head bones may occur in some larvae and be useful in specific, generic, or fam ilial identifi Ahlstrom and Moser (1976) pointed out that a combinatIOn of cation. Some head spines are transient, whereas others (particularly morphological and meristic characters i not only of value in IdentI opercular spines) may persist into the juvenile or adult stage. These fying larval fishes to order, suborder, and often to family, but that such characters can be used to infer phylogenetic affinities. As spines, crests, or ridges may vary intra s pecificall~' in size, shape, or sculpturing. shown in Tables I and 2, morphological and meristtc characters can Spines, crests, or ridges are often associated with the opercle be used to assign unknown larvae to order or suborder. The charac bones, supraoccipital, pterotic, or other head bones; they may be teristics of fin position, development, and structure are particularly located on the nuchal, snout, or supraoccular regions and are found useful in taxonomy and systematics. Although ' number of authors have used characteristics of larval in a number of diverse groups of fishes such as melamphids, scor fishes as a means of analyzing phylogenetic affinities above the paenids, erranids, istiophorids, carangids, and pleuronectids. species level (e.g., for ceratioids, Bertleson 1951; for myctophlds, A number of authors have used head spination as a diagnostic Moser and Ahlstrom 1970, 1972, 1974; for serramds, Kendall character. Examples for scorpaenid larvae are Mo er et al. (1977), 1979; for coryphaenids, Potthoff 1980; and for Scombrolabraci Richardson and Laroche (1979) , and Laroche and Richardson dae, Potthoff et al. 1980), few workers have attempted to analyze (1980); for serranid larvae Kendall (1979) ; for carangid larvae phylogenetic affinities in a rigorou , quantitative manner. Even Aboussouan (1975); and for cottid larvae Richardson and Washing fewer have used developmental osteology in such analyses. A fev. ton (1980) . examples of such attempts follow. Employing an approach previously ued on adult fishes by Ebel Scales and lateral line pores.- Scales rarely develop until the ing and Weed (1973), Okiyama and Ueyanagl (1978) analyzed 12 juvenile stage and hence are of small value in identification of lar of the 14 genera of the subfamily Scombnnae (Scombridae) accord vae. A number of exceptions occur including some serranids (Ken ing to a sequence of primitive to advanced character state . A dall 1979), chiasmodontids (footnote II) , Xiphias gladius "character" may be defined as any attribute of an organism, and a (Potthoff and Kelley 1982), holocentrids, and branchiostegids "character state" a the qualitative and quantItative description of (Okiyama 1964) in which scales form during the larval period. The the character in question (Sneath and Sokal 1973) . Oki) ama and count of lateral line pores or of diagonal scale rows in juveniles, Ueyanagi (1978) a signed coded values to 13 characters (Table 3) however, may be an important tool in identification of juveniles and including head spination, ize and shape of premaxillary teeth, adults which can be traced back to the larvae (e.g., Laroche and pre ence or absence of a cartilaginou pad on lower jav., and myO Richardson 1980). mere counts. By summing the character- tate code \alue (Table 4), they arrived at an index of primitivene s v.hich allowed them to Cranial bones.- Osteology of the teleost skull has long been con truct a dendrogram depictIng morphological relationship 13 Table 3.-Larval characters (c hiefl) adnnced postlanal or earl) jmenile among the 12 genera studied (Fig. 7). Groups "A," "B," "C," and stages) presumed ph) logeneticall) important as coded states for comparison of " D " of Figure 7 correspond well to generic affinities previously 12 genera of the subfamily Scombrinae (fr om Oki)ama and Ue)anagi 1978). postulated by Richards (1973) on the bases of myomere counts and Coded state Char. pigment patterns, as well as to affinities postulated by Collette and index Character 3 Chao (1975) based on characters of adult fishes. Supraoccipital absent present Using larval as well as adult characters, Johnson (1974) analyzed spine generic relationships of Scopelarchidae. He analyzed primitive and 2 Head small; large; advanced states of six larval characters in addition to 15 adult char > 11:3 of SL < '13 ofSL acters, including several osteological characters: 1) extent of peri 3 Viscera compact; wide elongated; vent toneal pigment; 2) presence or absence of accessory pigment spots; and vent space from just in front 3) gradual or abrupt metamorphosis; 4) position of pelvic fin bud anal fin of anal fin in relation to the dorsal fin; 5) precocious or nonprecocious appear 4 Snout rounded pointed elongated ance of the pectoral fin; and 6) head length relative to standard Premaxillary minute large large; some 5 length . He also determined the length of the basihyal, and whethe teeth fang-like 6 Jaw equal size equal or unequal; distinct unequal size upper jaw projection Table 4.-Comparisons of 12 genera of the subfamily Scombrinae. For the 13 7 Preopercular absent present characters and their coded state, see Table 3 (from Okiyama and Ueyanag spine 1978). 8 Spiny supra absent present or present Coded character states by character number Tota orbital crest absent Genus 2 3 4 5 6 7 8 9 10 I I 12 13 score 9 Pterottc spine absent present Scomber 13 10 Cartilaginous absent present or present Rastrelliger I I 13 pad on lower absent Gralllmatorcynus I 3 2 I I 3 I I I 18 Jaw Scolllberolllorus 3 3 1 32 2 3 2 2 2 I 3 28 Acalllhocybiulll 332 3 3 I 3 3 3 3 32 II Dorsal body heavier lighter Sarda 3 3 2 2 3 3 3 I 3 I 3 29 pigmentatton GYlIlnosarda 3 3 3 3 3 3 3 3 3 2 34 12 Post yent present, absent or a absent Allothunnus 2 2 3 3 3 2 24 p'gmentatton extensive few dots Aluis 3 2 2 3 3 2 24 13 Myotome counts low; 30-31 middle; 38-4 1 Ihigh; 40-65 Euthynnus 3 2 2 3 3 2 24 KatsuH'onus 3 3 3 3 I 2 IExceptional 10" vertebral count of Scomberomorus sinensis is responsible for 2 2 24 Thunnlls 3 2 this undearly discreted coded state. 2 2 3 3 3 2 25 Figure 7.-Numerically con Class intervals of numbers of coincident character states structed dendogram depicting lar between pairs of genera ,al morphological relationships among 12 genera of the subfamily 2 4 5 9 10 13 Scombrinae (from Okiyama and Ueyanagi 1978). Groups A-D rep resent related genera, based on Scomber,' Rastrelliger data in Table 3 as coded in Table 4. 13 Group iii. Group A is considered the most primitive group; Group D the most 15 derived. Grammatorcynus Group B '"Q) '"c 20 Q) > : ~ .§ Allothunnus; Auxis,' a. Euthynnus,' Katsuwon us ( '+- 0 ~ Group x 25 Q) "0 - Thunnus C Scomberomorus 30 Sarda Group Acanthocybium 35 Gymnosarda 14 or not teeth were present. John on used these characters and char Table 5.-Pntenlial,harallcr, fur anal • olllll\llli acter states to infer relation hips among species and to group scope 'Ole; character u,uall) "I' ,alue (+, and OI11('llIn< uf, larchid species into genera. hierarch.callc-el. Certain larval characters were used by Zahuranec (1980) in his cladistic analysis of myctophlds of the genus Nallllobrachium. In O.J ~ addition to using a number of adult characters, he analyzed ontoge <;ubordcr netic characters to establ ish criteria of polarity of characters (i .e., Morpholog.cal P'gmcnl pallern from primitive to derived). Larval characters used by Zahuranec Predominant b"dy hare were: 1) Pectoral fin size and shape; 2) direction of orientation of Snoul·anu, lenglh pectoral fins; 3) pectoral ray thickness; 4) pectoral ray fuion to the Charactcri"t1c\ 01 gut edge of the cleithrum during ontogeny; 5) larval snout shape; Pel"c fin p'hll.on Lanai e)e shape 6) presence or absence of a swim bladder in larvae; and 7) presence Tr..ln,formatlon'tage: + of a fat-filled or gas-filled swim bladder. The adult and larval char PreJu'cnlle ,Iage prc,cnt .. + 1969; acters were used to construct a Wagner tree (Kluge and Farri O'leologlCal Farris 1970) to hypothesize phylogenetic relationships of pecies in ~umnL'r of \ertchrac Nanllobrachium. Number of bmnlhll"tcgal, + Washington (1981) used 11 larval characters in a Wagner tree Do"a".n(s': Number .. T) pc of elemenl, + analysis (Farris 1970; Farris et al. 1970) of larvae of three cottid Formula genera, Artedius. Clinocottus, and Oligoco((us. She used the fol Sequence of format um + lowing characters: I) Number of preopercular pines; 2) size of Anal fln(,' !\umber preopercular spines; 3) presence or absence of basal spines on pre T) pe of clement, .. Formula opercular spines; 4) number of auxiliary spines on inner helf of Sequence of formatIOn + preopercle; 5) presence or absence of bubble of skin at nape; La,t dnrsal and anal ra} 6) presence, absence, and size of gut diverticula; 7) number and b.furcate + modification of parietal spines ; 8) presence or absence of melano Peltoral rln m}, . phores on nape; 9) snout pointed or round; 10) presence or absence Sequence of forma linn ... Formula of trailing gut and, if present, the configuration and length of the PelVlL f.n m) " trailing gut; and II) number of pelvic fin ray. Sequence of formatlllO .. Based on these characters, Wa hington (1981) erected a clado Formula gram inferring systematic relationships of the three genera and 13 Caudal fin T) p~ Pnnllpal laudal ra), t species. She concluded that two sister groups exist within the gen MH.mum number h)pural, era examined: I) Those with multiple preopercular spines, and Ma\ll11um numberepumls 2) those sharing two derived characters-basal preopt;rcular pines Number or ural centra + and pointed snouts. eu ml 'p.ne on vertebl'lle adJa.:enllo ural> + The investigation of phylogenetic relationships following the cla Fin 'plOe, or m), elongale + + di tic approach of Hennig (1966) has stimulated an inten e debate Head ~plnc~ prc ... cnt In Ian ac in systematics ( ee, for example, the last 10 or more years of the Predor,al bone, pre,ent journal Systematic Zoology). This conceptual approach has been Oromandll'ular teeth little u ed with larval fishes, yet larval fishes possess a rich suite of characters and character states. In many groups of fishes, the larval stages exist in habitats di tinctly different from adults (e.g., plank and Eckardt 191'2: Estabrook ct al 1977; traugh \97 , Pre c.h tonic larval pleuronectids vs. demersal adult pleuronectids). 1980) such as uscd b} \\'a,hmgllln (19 I) ha c rarcl} been Hennig ( 1966) propo ed that ontogeny should be u ed as a crite attempted exclusi\ cl) \\ nh Ian al fishcs , rion for analyzing phylogenetic relationships and establIshmg the InLfcased usc of uc\c!opmcntal ostcolog} hould mlrea c ( ur polarity (direction) of character states from primitive to derived. unueNanuing of fishcs Jnu thcir rClJ1IOnship Eldredge and Cracraft (1980), in their pre entation of cladistic methods, restated Hennig's idea concerning ontogen} and stated ACK. 0\\ LEDGi\tE :TS that it is urprising that little attention has been directed toward using ontogenetic data in systematics. IIh Listed in Table 5 are a number of characters of potential value in analyzing phylogenetic affinities among teleo~t larvae. A number of other characters might also lend themselves to stud) and anal}- is, the most notable of which are bones of the neurocranIum (see Fra er 191'2 for examples of character state of adult apogonIds) A combination of o teo logical and other charaLter states \\ould be ueful in any such analysis. Analyse of character states ' uch as \\ ere done b) 01,,1) ama and manu ~'ript Ueyanagi t 197 ) can be of \ alue in eumatmg ph) logenetH: affinI tru~ tI\ eI) fC\ IC\\ cd th m nu ties at genenc Jnu higher level. depending upon the particular Flgurc, I and 3 D ugla B SUIlC of characters chosen and how thc) are Jnal) zed ~1ore rigor penultimate draft of thc p per. ou quantltatl\e approache. to ph) logcnetll anal) ,I. (klugc and hhng<:r allurat"'}. rapldl}. Farris 1969: mlth and Koehn 1971. Ma", and Rabb 1972: Baird draft ofthl manu lnpt Kevin M. Howe, College of Fisheries, University of Washington 1969 . Elagatis bipinnulata (Pisces: Carangidae): Morphology of the fins and (U W), shared his knowledge of fishes with the author and brought other characters. Copeia 1969:454-463. BERRY, F. H., and W 1. RICHARDS. numerous references to his attention. Howe gave unduly of his time 1973. Characters useful to the study of larval fishes. In A. L. Pacheco (edi and knowledge in the preparation of this review and constructively tor) , Proceedings of a workshop on egg, larval, and juvenile stages of fis h in criticized the manuscript. Atlamic coast estuaries, p. 48-65. Tech . Publ. I, Middle Atl. Coastal Fish. Leanne L. Stahl, (formerly of the Department of Oceanography, Cent., Natl. Mar. Fish. Serv., NOAA, Highlands, NJ07732. BERTELSEN, E. UW), Kathryn 1. Garrison, Fisheries Research Institute, UW, 1951. The Ceratioid fishes. Ontogeny, taxonomy, distribution an Sally L. Richardson , Gulf Coast Research Laboratory, Ocean biology. Dana-Rep. , Carlsberg Found. 7(39), 276 p. Springs, Miss., and Michael P. Fahay, National Marine Fisheries BLACKWELDER. R. E. Service CNMFS), Sandy Hook, N.1. , each reviewed various drafts 1967. Taxonomy, a text and reference book. Wiley, N. Y, 698 p. of the manuscript and made numerous helpful suggestions. BOND. C . E . 1979. Biology of fishes. W B. Saunders Co .. Phila., 514 p. Betsy B. Washington CU W) kindly made available to me a draft BOND, C. E., and T. UYENO. of her M.S. thesis, and read and commented on portions of the 1981 . Remarkable changes in the vertebrae of perciform fish ScombrolabrfU manuscript. Theodore W Pietsch CU W) read portions of the type with notes on its anatomy and systematics. lpn. 1. Ichthyol. 28:259-262. script and Daniel M. Cohen CNWAFC) read the whole of it. BORODULINA, O. D. 1969. Osteology of LeuroglossLls stilbius schmidti Rass (Bathylagidae) Thomas Potthoff, Southeast Fisheries Center, NMFS, Miami, Fla., Prob. Ichthyol. 9:309- 320. generously prov ided me with a copy of his unpublished manuscript CASTEEL, R. W on the development of Xiphias gladius. 1976. Fish remains in archaeology and paleo-environmental studies. Acad. One referee (Thomas Potthoff) made numerous helpful com Press, N. Y, 180 p. ments on a draft of the manuscript which enabled me to eliminate CAVENDER, T. M . 1970. A comparison of coregonines and other salmonids with the earlies errors and to expand the scope of certain portions of it. Any remain known teleostean fishes. In C. C. Lindsey and C. S. Woods (editors), Biolog ing errors in fact or interpretation remain the sole responsibility of of coregonid fi shes, p. 1- 32. Univ. Manitoba Press, Winnipeg. the author. CLOTHIER, C. R. 1946. Vertebral variation with size in Clevelandia ios. Copei, 1946: 113- 116. LITERATURE CITED 1950. A key to some southern California fishes based on vertebral characters ABOUSSOUAN. A. Calif. Dep. Fish Game, Fish Bull. 79, 83 p. 1975 . Oeufs et larves de Tel eosteens de l'Ouest africain . XIII. Contribution COHEN, D. M. , and 1. G. NIELSEN. a I'identification des larves de Carallgidae. Bull. Inst. Fond am. Afr. 1978. Guide to the Identification of genera of the fish order Ophidiiforme. Noire. Ser. A SCI. Nat. 37:899-938. with a tentative classification of the Order. U.S. Dep. Commer., NOA AHLSTROM. E . H. Tech . Rep. NMFS Circ. 417,72 p. 1969. Remarkable movements of oil globules in eggs of bathylagid smelts COLLETTE, B. B. , and L. N. CHAO. dunng embryonic development. J. Mar. Bi oI. Assoc. India 11 :206-2 17 . 1975 . Sys tematics and morphology of the bonitos (Sarda) and their relative AHLSTROM. E . H .. and O. P. BALL. (Scombridae, Sardini). Fish. Bull.. U.S. 73 :516- 625. 1954 Description of eggs and larvae of jack mac kerel (Trachurus symmetri DAWSON , A. B. cus) and distribution and abundance of larvae in 1950 and 195 1. U.S. Fish 1926. A note on the staining of the skeleton of cleared specimens with alizarit Wildl Serv. • Fish. Bull. 56:209- 245 . red S. Stain Technol . I : 123- 124 . AHLSTROM. E. H .• 1. L. BUTLER, and B. Y SUMIDA. DINGERKUS, G ., and L. D. UHLER. 1976 PelagIC stromateid fis hes (Pisces, Perciformes) of the eastern Pacific: 1977. Enzy me clearing of alcian blue stained whole small vertebrates fo kInds. dIstributions. and earl y lIfe his tori es and observati ons on fi ve of these demonstration of cartilage. Stain Technol. 52 :229- 232. from the northwest Atlantic. Bull. Mar. Sci. 26:285-402. DUNKELBERGER, D. G ., 1. M. DEAN, and N. WATABE. AHLSTROM. E. H., and R. C. COUNTS. 1980. The ultrastructure of otolithic membrane and otolith in the juveni 1955 Eggs and larvae of the Pacific hake. Merluccius productus. U.S. Fish mummichog, Fundulus heteroclitus. 1. Morphol. 163 :367-377. Wildl Serv.. Fish. Bull. 56' 295-329. DUNN. 1. R. 1958 Development and distributIon of Vi ll ciguerria (u cetia and related spe 1983. Development and distribution of the young of northern smooth tongue cies in the eastern Pacific. U.S. Fis h Wildl. Serv.. Fish. Bull. 58 :363-416. Lel/rog/ossus schmidti (Bathylagidae) , in the northeast Pacific, with com AHLSTROM. E. H . and H. G. MOSER. ments on the systematics of the genus LeuroglossLls Gilbert. Fish. Bull 1975 Dbtributional atlas of fish larvae in the California Current region: Flat U.S. 81 :23- 40. fishes 1955 through 1960. Calif. Coop. Oceanic Fish. Invest. Atlas 23. EATON , T. H ., Jr. 207 p. 1945 . Skeletal supports of the median fins of fishes . J. Morpho 1976 Eggs and larvae of fis hes and their role in systematic in vesti gations and 76: 193- 212 . in fisheries. Rev Trav. Inst. Peches Mant. 40:379-398. EBELING, A. W , and W H. WEED Ill. ;\MAOKA, K 1973. Order Xenoberyces (Stephanoberyciformes). In D. M. Cohen (editor 1969 Studies on the SInistral flou nders found in the waters around Japan Fishes of the Western North Atlantic. Mem. Sears Found. Mar. Res., Yal la,-onom). anatomy, and phylogeny. J. Shimonoseki Uni v. Fish. 18 :65- 340. Univ. I (Part 6) :397- 478 . \PRIETO, V L. EDMUND, A. G . I Y74 Earl) development of five carangid fis hes of the Gul f of Mexico and the 1960. Tooth replacement phenomena in the lower vertebrates. R. On south Atlantic coast of the UnI ted States. Fis h. Bull. , U. S. 72 :415- 443. Mus. , Contrib. 52. 190 p. BAIRD, R C , and M J ECKARDT ELDREDGE. N .. and 1. CRACRAFT. I'ln DIVergence and relatIonshIp in deep-sea hatchetfis hes (Sternoptychi 1980. Phylogenetic patterns and the evolutionary process. Method and theo .lac) Syst Zool.21 :80-90 . oc omparative bIology. Columbia Univ. Press. N. Y. 349 p. BLEBI- \\ EMELIANOV, S. W I 16 FARRIS. J. S. HUBBS. C. L.. and K. F. LAGLER. 1970 Methods fo r computing Wagner trees Sy I. Zoo I. 19 :83-92. 1949 Fishes of the Great Lakes Region Cranbrook Inst Sci. Bull 26. 186 p. FARRIS. J S .. A G . KLUGE. and M . 1. ECKARDT HUBBS. C. L.. R. R MILLER. and L. C HUBBS. 1970 A numerical approach to phylogenetic systematic,. Sysl. Zool. 1974. HydrographIC hlslOry and relict fishes of the North-Cemral Great 19: 172-1 9. Ba;1I1 Calif. Acad SCI. Mem . 7. 259 p. FARUQI. A J HUGHES. D. R. 1935 The developmem of the vertebral column 111 the haddock (Gadus aeglt 1981 . Developmem and organization of the posterior field of ctenOld scales 111 filiUS). Proc Zool Soc Lond. 1935 :3 13 -332 the Platycephal idae . Copeia 1981 596-606 FI NK. W. L. JOHNSO • G D. 198 1. Ontogeny and phylogeny of 100th allachment modes in actinopterygIan 1975. The procurrent spur: An undescribed perclform caudal character and liS fis hes. J. Morphol. 167 : 167- 174. phylogenellc Implications. Occas. Pap. Calif Acad Sci . 121 .23 p. FI K. W. L.. and S. H. WEITZMAN. JOHNSON . R. K. 1982 . Relation,hlps of the stomiiform fishes (Te leostei) . WIth a deSCription of 1974. A revision of the alepisaurold famdy Scopelarchidae (Pisces Myc Diplophos. Bull. Mus. Compo Zoo I. 150:31-93. IOph lformes). Fieldiana Zoo I. 66. 249 p. FORD. E. JOLLI . M . 1937. Vertebral variallon In teleostean fishes Mar BIOI. A"oc . U.K. 1973 Chordate morphology. R E Krieger Pub\. Co .. N. Y.. 478 p. 22: 1-60 KENDALL. A W. Jr. FOWLER. J. A 1976 Predorsal and associated bones In serranld and grammlslld fishes 1970. Comrol of vertebral number 111 teleosts-an embryologIcal prob Bull Mar SCI. 26:585-592 lem. Q. Rev. BIOI 45.148- 167 . 1979 MorphologIcal comparISons of North American sea bass larvae (Pis FRASER. T H ces Serranldae). U.S. Dep. Commer.. NOAA Tech. Rep. NMFS Circ 428. 1972 . ComparatlYe osteology of the shallow water cardll1al fi;he, (Percl 50 p formes Apogonldae) \\ IIh reference to the sy>!cmallc, and e\ olullon of the KLUGE. A G . and J S FARRIS. famIly. Rhode; Unlv.. Dep. lchthyol..lchthyol Bull . 34 1- 105. 1969 Quanlllatlve phyletics and the evolutIon of anurans. Sysl. Zoo I. 181-32 FRJTZSCHE. R. A .. and G. D. JOH SON. KNUDSEN. J W. 1980. Early osteologIcal developmem of "hlle perch and striped bas, wllh 1966. BIOlogIcal technIques: collecting. preserving and illustrating plams and empha>ls on Idenllficallon of their la"ae Tran, Am F"h Soc anImals Harper and Row. N. Y.• 525 P 109' 387 - 406 KONNERTH. A GERAUDIE. J 1965 Preparation of ligamentary artIculated fish skeletons Curator 1978 . Scanning electron mIcroscope study of the developIng Irout pelVIC fin 8:325-332. bud Anat Rec. 191.391-394. KRAMER. D GOSLI E. W A 1960 Development of eggs and larvae of PaCIfic mackerel and distribution 1960. COnlrlbUIIOnS toward a claSSIfication of modern isospondylous fishes. and abundance of larvae. 1952-56. U.S Fi h Wddl Servo Fish. Bull Bull Br Mu; ( at HJ;I) Zoo I. 6:325-365 60393-438. 1961 Some osteologIcal features of modern lower teleo>!ean fishes . LAROCHE. W L . and S. L. RICHARDSO Smithson Mlsc Collect 142 1-42. 1980 Development and occurrence of larvae and juveniles of the rockfishes 1971 . Functional morphology and cla"ification of teleo>!ean fl"hes. UnlV Seblwes jlol'idus and Sebastes melallops (Scorpaenldae) off Oregon. Fish Hawaii Press. Honolulu. 208 p. Bull. U.S. 77'901-924. GREENWOOD. PH. and D E. ROSEN LAWSON. R . and B. F J MANLY 1971 OIes on the structure and relallon,hlps of the alepocephalold fishes. 1973 Tooth growth and replacement in Ctenolucius hujeta. a neotropical Am. Mus OVlt 1971(2473).41 P characold fish J Morphol. 14\:383-394 GREE WOOD. PH. D E. ROSE . S H WEITZMA . and G S. MYERS. LEIS. J M 1966. Phylellc studIes ofteleostean fi,hes. ,,"h a pro",.onal cla"lficallon of 1977 Developmem of the eggs and larvae of the slender mola. Ranzall/a liVing forms Bull Am. Mus. Nat HI>t 131.341 -.. 5 laem (PIsces. Molldae). Bull Mar SCI 27 448-466 HEMPEL. G • and J H S BLAXTEP LEIBY. M M 1961 The experlmemal modification of meristic characters In herring (Clu 1979 Leptocephalus larvae of the eel family Ophichthidae. I. Ophichthus pea harengus L.) 1. Cons. Perm. Int Explor. Mer 26.336-346 gomesl Ca;telnau. Bull Mar. SCI 29329-343. HENNIG. W LI DSEY. C. C. 1966. Phylogenellc sy>!emallcs. UnlV. IllInOIS Pre,s. ChIcago. 251 p 1955. E, >lutlOn of mensllc relallons 10 the dorsal and anal fins of teleost HENSLEY. D. A. fi he Trans. R. Soc. Can. 49 :35-49. 1977. Larval developmem of Engyophrys senta (Bothldae). wllh commems 1975 Pleomerism . the widespread tendency among related fish specIes for on imermuscular bones in natfishes. Bull Mar Sci 27681 -703 vertebral number to be correlated WIth maXImum bod) length. J. Fish. Res. HILDEBRAND. M. Board Can. 32:2453-2469. 1974 . AnalYSIS of vertebrate structure Wiley. NY. 710 p. LUNDBERG . J G. and 1. N BASKIN . HOLLISTER. G 1969 The caudal skelelOn of the catfishes. order Silurlformes. Am Mus 1934 . Clearing and dyell1g fish for bone study Zoologlca 12 89-10 I Novlt 2398.49 p. 1936. Caudal skelelO n of Bermuda shallow water fishes I Order Isospon MAKUSHOK. V. M . dyli: Elopldae. Megalopidae. Albuhdae. Clupeldae. Dussumieriidae. 1958 The morphology and classification of the northern blenniOld fishes Engrauhdae. Zoologica 21 :257-296 (SlIchaeOldae. Blennioidei. PIsces) Akad. Nauk SSSR. Tr. Zoo\. Insl. HOLM BAKKEN. N .• and G FOSSE. 253-129. [In Russ.] . (Transl. by A R Gosline and W A Gosline. 1959. 1973. Tooth replacememll1 Gadus callanas. Z. Anal Emwlcklungsgesch. 105 p .. avaIl. U.S. Dep. Inter.. Fish. Wildl Serv. Ichthyo\. Lab .. Wash. 143 .65-79. D.C) HOTTA. H. MANAVALA RAMANUJAM , S. G. 1961. Comparallve study of the aXIal skelclOn of Japanese teleostei. Bull . 1929 The study of the development of the vertebral column in teleosts. as Tohoku Reg. Fish Res Lab .• 155 p. shown In the life history of the herring Proc Zool Soc Lond . HO UDE. E. D .. and T POTTHOFF. 1929 :365-414 . 1976. Egg and larval development of the sea bream Archosarglls rhomboida MANSUETI. R lis (L1nnaeus): PIsces. Sparidae. Bull. Mar. SCI 26:506- 529. 1958 Eggs. larvae and young of the striped bass. Roccus saxarilis. Chesa HOUDE. E. D .. W. 1. RICHARDS. and V. P. SAKSENA peake BioI. Lab. Comrib. 112.35 p. 1974 . DescriptIon of eggs a nd larvae of ,caled sardlOe. Harellgula MARKLE. D. F jaguana Fish. Bull.. U.S 72: 1106-1122. 1980. A new speCIes and a revIew of the deep-sea fish genus Asquanllceps HUBBS. C L. (Salmonlformes: Alepocephalidae) Bull Mar. SCI. 30:45-53. 1958 . DlkellorhYllchus and Kallaza,,,Q1chthys: nominal fish genera inter MARSHALL. N. B.. and D. M. COHEN preted as based on preJuveniles of Malacanthlts and Anlllennarius. respec 1964. Order AnacanthiOl (Gadlformes) Character and synopsis of families lively. Copeia 1958 :282-285. /11 D. M . Cohen (editor). Fishes of the Western North Atlanllc Mem. Sear; 17 Found Mar. Re, . Yale Untv .. I (Part 6) 479 495 I'LSON, G . J MARX. H ,and G B RABB. 1967 Fplbranchlal organs In lowcr teleostean fIshes. J Zool (Lond) 1972 . Phyletic analysIs of fifty characters of advanced snakes Fieldlana 153'71 89 Zool 63,321 p. 1969 Gill arc he; and thc phylogeny elf fi,he;, With note, on the cla"lficatlo MATARESE. A. C ,S L. RICHARDSON . and J R. DUNN of vertebrate, Bull . Am . Mus. ·al. HlSt 141 '475-552 1981 Larval development of Pacific tomcod. MICTogadu.1 I'ro.nmus. In the 1970 Gdl arches of some telcostcan fl~hes of the famllie Salangldae an northeast Pacific Ocean with comparative notes on larvae of walleye pollock Argentlnldae. Jpn J Ichthyol. 1761 66. Theragra chalcogramma. and PacifIC cod, Gadus mllcro«'I'/wlul (Gadl NELSON, J S dae). Fish. Bull, U.S. 78 923 940 1976 Fishes of the world Wiley, NY, 416 P MATSUMOTO, W M NORDEN.C R 1959. DescnptlOns of Eut/nnrlll.\ and ALLtlS larvae from the PaCific and Atlan 1961 Comparative osteology of rcpresentatlve \almonld fi~hes, with partlcu tic Oceans and adjacent seas. Dana-Rep.. arl,berg Found 9(50). 34 p. lar refercnce to the grayling (77l\mllllu\ ant/wI) and It; phylogeny. J FISh 1962 IdentificatIOn of larvae of four species of tuna from the Indo·PaCinc Res Board Can 18679 -79 1 region I. Dana-Rep. Carbberg Found 10(55). 16 P NLRSALL, J R 1963 Untque shape of the first elongate hac mal 'pine of albacore, Tilllnllu.1 1963 The hypurapophysls an Important element of the caudal ,keleto alalunga (Bonnaterre). Copela 1963:460464 Copela 1963.458 459 1967 Morphology and distribution of larval wahoo 4ulllt/wnblU/II HlIUlld" YBEL! 0 (Cuvler) In the central Pacific Ocean U.S. Fish Wildl Sen , Fish Bull 1963 fur Morpholo!pe und Termlnologle des Sch",an,kelene 66299-3 22 terygler Ark 1.001, cr.~ , 15485-516 MATSUMOTO, W M . E. H AHL TROM. S JONES. \\ L II.LAW , W J 1971 . On the caudal keleton In flops with remarks on other teleo\le· RICHARDS. and S UEYANAGI fIShes . Acta Reglae Soc SCI Lltt Gothob Zool 7 I-52 972 On the clanficatlon of Ian al tuna identlt'ication particularly In the OKIYAMA , M ge nus Thunnus . Fish Bull. U.S 701-12 1964 Earl) Itfe hIStory 01 the Japane'e blanqudlos, BranchlUstegu! Japon MAYR, E (·U>j(/I'"f1/(US (Hounuyn) Bull Jpn Sea Reg FISh Res. Lab 13 1-14 1969 Pnnclples of systematic zoology McGraw-HIli N. Y , 428 P OKIYAMA , M and S LEYAl'iAGI MAYR, E ,E G LI SLEY. and R L. USINGER 1978 lnterrelatwnshlp, of scomb,,"d fishes an a peet from larval morpho 1953 Methods and principles of systematic zoology. McGraw -HilI , . Y. ogy Bull Far Sea, FISh Res Lab (ShlmilU) 16 ' 103-113 336 P ONO, R. 0 McALLISTER. D. E. 1980. A ,dver ImpregnatIon techntque to demonstrate muscle·bone-cartda 1968 Evolution of branchlostegals and cla",ficatlon of teleoswme relationship, in fishes Stain Technol 55 67 - 70. fishes. atl Mus Can. Bull n I . 239 P ORTO , G L MEYER. 0 B • and R O·RAHILLY 1963 'otc, on larsal anatomy of fishe, 01 the Order Lyomen Cope 1958 ~lultiple techntques In the stud) 01 the onset 01 prenatal o",Ilca 19636 15 tion A nat Rec 132 18 1- I 83 OS IA , C R MEYER-ROCHO\>,. V B. 1970 PreparatIOn or dISartICulated skeleton, uSlOg enzyme-bd,ed laund 1972 A scanning electron mIcroscope slUdy of manne fISh lanae "pre ,oaker' C 'PCld 1970 19'1-200 Mlcrosc (Pans) 13 169- 172. PATTERSOl'i , C MILLER, G L., and S C JORGE SO 1968 The caudal ,keleton 10 Lower LIassic pholtdophonJ fishes Bull. 1973 Meristic characters of some manne fIShes of the western Atlantic \-1u, ( 'at HISt ) Geol 16201-239 Ocean Fish. Bull. U.S 71 301 312. PA"TO -, J R MILLER, J M ,and J W TUCKER. Jr 1972 O,teology anJ relalton,hlps 01 the lanternfishe (FamIly Myctoph 1979 X-RadIOgraphy of lanaI and Juvenile fishes. Copela 1979 5)5 538 dae) Bull Los Ang Cty .\-1u, at HIS!. (I 13, I p. MILLER, R. V, and J W VA LA 01 GHAM PIETSCH T W 1969 Additional procedures for effective enzyme cleanng and staining of 1974 Osteology and relatlon,hlps of ceraltold anglerfishes of the fami fishes. Copeia 1969 829-830. Oneirodldae. ""th a rev Ie'" of the genus OnelfCxies Lutlen Bull Los An MO 00, T Cly. Mus at HI t SCI. I . 113 p. 1968 Le complex urophore des poissons teleosteens Mem Insl Fondam. 1981 The osteology and relatIOnshIp, of the anglerfish genus Tetrabrachl d'Afnque o ire 81. 705 P w Ilh comments on lophllform claSSIfication . FI;h Bull . G. 79387-41 MOOK, D. POTTHOFF. T 1977 Larval and osteologIcal development of the sheepshead, ArdwslIrgus 1974 Osteological development and v ariallon in young tunas. genus Thunn probatocephalus (PISces Spar/dae) Copela 1977 126 133 (Pisces. Scombridae). from the Atlanltc Ocean Fish Bull , Uol MOOKERJEE, H K , G MITRA, and S R MAZUMDAR n 563 588 1940 The development of the vertebral column of a ""parou, teleost. 1975 Development and structure of the caudal complex. the verteb Lebistes reticulatus. J. Morphol . 67 241-269. column, and the pteryglOphores In the black fin tuna (Thunnus atlanllcus, P MOSER, H . G. ces, Scombndae) Bull Mar SCI 25 205-231 1972 . Development and geographiC d,stnbutlOn of the rockfish. Sebastes 1980. Development and tructure of fin and fin supports in dolphin fishl macdonaldi (Eigenmann and Beeson. 1893). family Scorpaenidae, off south Cor.lphaena hippurus and Conpltaena equiselis (Coryphaenidae). Fi ern Cahfornia and Baja Cahfornta. FISh Bull, U.S 70941-958 Bull , U. 78:277-312 [1981] . Morphological and functIOnal aspects of manne fish larvae. In R POTTHOFF, T , and SKELLEY Lasker (editor), Manne fi sh larvae: Morphology, ecology, and relation to 1982 Development of the vertebral column. fins and fin supports, branch I fisheries. p. 90-13\. Univ. Wash Press. Seanle. tegal rays and squamation in Ihe swordfish (Xiph/Os gladius) Fish. Bul MOSER, H G . and E. H AHLSTROM. U.S. 80 161 - 186. 1970. Development of lantern fishes (famIly Myctophldae) In Ihe CalifornIa POTTHOFF, T. W 1. RICHARDS. and S UEYANAGI. Current. Part I SpecIes with narrow-eyed larvae. Bull. Los Ang. Cty. Mus. 1980. Development of Scombrolabrax heterolepis (Pisces, Scombrolabra Nat Hlst. Sci 7, 145 p. dae) and comments on familial relallonships. Bull. Mar. SCI 30:329-35 1972 . Development of the lanternfish. Scopelopsis multipunctatus Brauer PRESCH , W 1906. with a diSCUS SIOn of ItS phylogenetIc positIOn In the family Myctophi 1980. Evolutionary hl,tory of the South American microtelld Itzards (Tellda dae and ItS role In a proposed mechanism for the evolutIOn of photophore pat Gymnophthalminae). Copeia 1980'36- 56. terns in lanternfishes. Fish . Bull., U.S. 70:541-564 REGAN, C. T. 1974 Role of larval stages In systematic Investigations of marine teleosts : 1913 . The classification of percoid fishes. Ann. Mag. at. HISt., Ser. j The Myctophldae, a case study Fish. Bull.. U.S. 72:391-413. 12 : 111 - 145 . MOSER, H. G., E. H . AHLSTROM, and E. M. SANDKNOP. RICHARDS, W. 1., R. V. MILLER, and E. D. HOUDE. 1977. Guide to the IdentificatIOn of scorpIOn fish larvae (family Scorpaenidae) 1974. Egg and larval development of the Atlantic thread herring, Opi. in the eastern Pacific \\ ilh comparative notes on species of Sebastes and Heli than em a ogliflum. Fish. Bull ., U.S. 72: 1123-1136. colen us from other oceans U.S. Dep. Commer., NOAA Tech . Rep. NMFS RICHARDS, W. 1., and T. POTTHOFF. Clrc. 402, 71 p. 1974. Analysis of the taxonomic characters of young scombrid fishes, gem 18 ThUll/IUs. III 1. H. S. Blaxter (editor), The early life history of fi sh. p. SWINNERTON, H H. 623-648. Springer-Verlag, N. Y. 1906. A contribution to the morphology and de\elopment of the peLtoml skel RlCHARDSON, S. L. , 1. R. DUNN. and N. A. NAPLI eton ofteleosteans. Q. J Mlcrosc. SCI. 49'363 383. 1980. Eggs atld larvae of butter sole, Isopsetta isolepis (Pleuronectidae). off TAKAHASHI , Y. Oregon and Washington. Fish Bull., U.S 78 :401 - 417 . 1962. Study for the identification of species based on the \ertebral column of RlCHARDSON, S. L. , and W. A. LAROCHE. Teleostei In the Inland sea and its adjacent waters. Bull Nan"-al Reg Fbh 1979. Development and occurrence of larvae and juvenIles of the rockfishes Res Lab 16:1-72. Sebasles crameri , Sebasles pinniger, and Sebasles he/,'omaculallls (famIl y TAYLOR , W R. Scorpaenidae) off Oregon. Fish. Bull. U.S. 77: 1- 46. 1967 An enzyme method of clearing and staining small \ertebmtes Pro.: RlCHARDSON, S. L. , and B. B. WASHINGTON. U.S. Natl Mus. 122 1-17 1980. Guide to the identification of some sculpin (Cottldae) larvae from 19 77. Observations on specimen fixation Proc Bioi Soc Wash marine and brackish waters off Oregon and adjacent areas in the Northeast 90:753-763. Pacific. U.S. Dep. Commer., NOAA Tech. Rep. NMFS CIrc. 430, 56 p. ROSEN , D. E. TYLER, 1. C 1973 . Interrelationships of higher euteleostean fishes. In P. H Greenwood, 1962. The pelvis and pelViC fin of plectognath "shes: a stud) In reduLlllln R. S. Miles. and C. Patterson (editors). Interrelationships of fishes, p Proc Acad at SCI. Phila. 114:207-250. 397-513. Acad. Press, N.Y. 1980 Osteology, phylogeny. and higher c1as"llcatlOn of the fishes of 'he ROSEN , D. E., and C . PATTERSO . Order Plectognathi (Tetraodontiform"l U.S Dep. Cammer. !'I;OAA Tech 1969. The tructure and relationships of the paracanthopterygian fishes Rep. NMFS Clrc. 434, 422 p. Bull. Am. Mu s. Nat. Hlst. 141 :357- 474. VOSS. N A RUSSELL. F. S. 1954 The postlarval development of the fishes of the famll) Gemp, IIdac 1976. The eggs and planktoniC stages of British marine fishes Acad . Press. from the Flonda Current. I. NesiardlUI Johnson and Gemprlus CU\ and . Y., 524 p. Val Bull Mar SCI. GulfCanbb 4 120-159 SCHULZTE, O. WAKE, M. H. (Editor). 1897 . Veber Hersteilung un Conservirung Durchich-tlger Embryonen zum 1979. Hyman'S comparative vertebrate anatomy 3d ed Ul11v. Chicago Studium der Skeletbildung. Anat. Anz. 13:3-5. Press, Chi cago, 788 p. SCOTT, 1. H., and N. B. B. SYMONS. 1964. Introduction to dental anatomy. E&S LI Vingstone, Ltd .. Lond ., 406 p. WASHINGTON , B B. SEWERTZOFF, A. N. 1981 Identification and systematics of lan ae of 4rledius, Clinocottu.l. and Oligocottlls (Scorpaeniformes Cottldae). M S TheSIS, Oregon State 1934. Evolution der Bauchflossen der Fische Zoo I. Jah rb . Anat. Ul11v .. Corvallis, 206 p. 58:415- 500. SIMO S. E. V., and 1. R. VA HOR . WASSERSUG, R. 1. 1971 . A new procedure for whole-mount aiclan blue staining of the cartilagi 1976 A procedure for differential staining of cartIlage and bone In \\ hole nous skeleton of chicken embryos, adapted to the cleanng procedure in potas formalin-fixed vertebrates. Stain Technol . 51 131 134 sium hydroxide. Acta Morphol. Neerl. -Scand . 8'281 -292 . WEISEL. G. F SIMPSON, G . G. 1967. Early ossification in the skeleton of the sucker (Caloslomus macro 1961 . PrinCiples of animal taxonomy. Columbia UnIv. Press. N. Y.. 247 p. chellus) and the guppy (Poecilia reliculara) J Morphol 121 I 18. SMITH, C. L. , and R. M. BAILEY. WEITZMAN, S. H. 1961. Evolution of the dorsal-fin su ppo rts of percold fishes. Pap. Mich. 1974 Osteology and evolutionary relation,hlps of the Sternoptychidae \\ Ith a Acad . Sci., Arts Lett. 46:345-363. new claSS ification of the stomiatOid families. Bull Am Mus Nat. H"t. SMITH , D. G . 15 3:329-478. 1979. GUide to the leptocephali (Eloplformes, Anguilliformes, and Notocan- thiformes). U.S. Dep. Commer., OAA Tech Rep MFS Circ. 424, 39 p. \IlLLIAMS. J E., and C E. BO D. SMITH, G . R., and R. K. KOEHN . 1980. Gila boraxobius, a new species ofcypnnld fish from southeastern Ore gon 'Hth a companson to G. ail'ordellsis Hubb and Miller Proc Bioi Soc 1971. Phenetic and cladistic studie of bIOchemical a n~ morphological char acteristics of Catoslomus. Syst. 7 aol. 20:2 82-297 . Wash . 93:29 1-298. STARKS. E. C. WILLIAMS. T. w., Jr. 1930. The pnmary shoulder girdle of the bony fi shes. Stanfo rd Univ. Publ 1941 AlIzarin red S and tolUidine blue for differentiating aduh and embry- BioI. Sci . 6: 149-239. onic bone and cartilage. Stain Technol 1623-25 STRAUGH,1. G., Jr. WINKELMANl\, R. K., and R. W. SCHMIT 1978. The phylogeny of the Charadriiformes (Aves): a new estimate using the 1957. A simple Si lver method for nerve axoplasm Proc. Staff Meet. Mayo method of character compatibility analysis. Trans. Zool. Soc. Lond. Clln .32:217-222. 34:263- 345. SUGA, S., K. WADA , and M. OGAWA . YAMAMOTO, M., and N. EGAMI 1981 . Fluonde concentration in teeth of tetraodontlform fishes and its phylo 1974. Fine stru cture of the surface of the anal fin and the processes on 11> fin genetic ignIficance. Jpn. 1. Ichthyol. 28:304-312. rays of male Ory~ias lalipes Copela 1974:262-265 SUMIDA, B. Y. , E. H. AHLSTROM . and H. G . MOSER. ZAHURANEC, B. J. 1979. Early development of seven natfishes of the eastern North PaCific With 1980 Zoogeography and systematics of the lantern fishes of the genus Nanno· heavily pigmented larvae (Pisces. Pleuronectiforme ). Fi h. Bull ., U.S. brachium (Lampanyctil11: Myctophldae). Ph D. TheSi . George Washing 77: 105-145. ton Univ., Wash ., D.C., 331 p. 19