the new actinopterygian order guildayichthyiformes from...

171GEODIVERSITAS • 2000 • 22 (2) © Publications Scientifiques du Muséum national d’Histoire naturelle, Paris. www.mnhn.fr/publication/

The new Actinopterygian orderGuildayichthyiformes from the LowerCarboniferous of Montana (USA)

Richard LUNDDepartment of Biology, Adelphi University, Garden City,

New York 11530 (USA)[email protected]

Lund R. 2000. — The new Actinopterygian order Guildayichthyiformes from the LowerCarboniferous of Montana (USA). Geodiversitas 22 (2) :171-206

The data matrix is available at http://www.mnhn.fr/publication/matadd/g00n2a2.html

ABSTRACTThe new order Guildayichthyiformes n. ord. (Actinopterygii) is erected forGuildayichthys carnegiei n. gen. n. sp. and Discoserra pectinodon n. gen. n. sp.from the Upper Chesterian Bear Gulch Limestone of Montana. These marinefish have highly compressed, discoidal bodies, tall rhombic “ganoid” scaleswith peg-and-socket joints, fins supported by well-jointed and well-spacedrays, small mouths and specialized bones of the operculo-gular series. Theskull roof features several large median bones extending from the rostrum tothe occiput, three bones in the otic canal series, and several longitudinal seriesof pre-frontal elements. The cheek and antorbital areas are equally unique.The cheek has highly variable numbers of suborbitals, tall narrow dorsal andventral preopercular lateral line canal bones, and a series of interopercular rays(Guildayichthys carnegiei n. gen. n. sp.) or bones (Discoserra pectinodo n. gen.n. sp.) between the ventral preopercular bone and the branchiostegal series.There are several small postspiracular bones and eight to ten sclerotic bones inthe orbit. The extensive development of median skull roofing bones in theGuildayichthyiformes n. ord. compares to patterns seen in Dipnoi andPlacodermi, that also concentrate biting forces anterior to the orbits, suggest-ing similar adaptive responses to similar types of cranial stress regimes.Cladistic analysis involving 72 characters results in a very stable sister grouprelationship between Polypterus and the Guildayichthyiformes n. ord. as acrown group within the tested Paleozoic Actinopterygii. The Cladistia arerediagnosed as a superorder to reflect this relationship. The Cladistia appearto be the sister group of Platysomus and Amphicentrum.

KEY WORDSGuildayichthys carnegiei n. gen. n. sp.,

Discoserra pectinodon n. gen. n. sp., Osteichthyes,

new order,Guildayichthyiformes n. ord.,

Cladistia, Namurian, Montana,

osteichthyan skulls.

INTRODUCTION

The Bear Gulch Limestone lens is contained withinthe Bear Gulch Limestone member of the HeathFormation (Upper Chesterian, Namurian E2b). Itwas deposited in a shallow, tropical marine baylacking any significant traces of fresh-water fossils(Ziegler et al. 1979; Williams 1981, 1983; Horner1985; Horner & Lund 1985; Lund et al. 1993;Feldman et al. 1994; Grogan & Lund ms.).The Bear Gulch Limestone contains over110 species of fish, including approximately45 osteichthyan species. Only three of the approx-imately 34 species of actinopterygians from theBear Gulch Limestone have strongly compresseddiscoidal body forms. The two species that are thesubjects of this article raise difficult phylogeneticquestions, because while they are specialized inpropulsive and feeding adaptations, they possessmany features of cranial osteology that are unique

among Paleozoic actinopterygians (Lund, Poplin& McCarthy 1995) or that can best be understoodby reference to sarcopterygian characters.

ABBREVIATIONS

Specimens have been deposited in the collections of:MV University of Montana Museum, Missoula,

Montana;CM Section of Vertebrate Fossils, Carnegie

Museum of Natural History, Pittsburgh,Pennsylvania;

ROM Royal Ontario Museum, Toronto, Ontario;FMNH Field Museum of Natural History, Chicago,

Illinois.

MATERIAL AND METHODS

There are 110 specimens collected that pertain tothe taxa described herein. All collection data are

Lund R.

172 GEODIVERSITAS • 2000 • 22 (2)

RÉSUMÉGuildayichthyiformes, un nouvel ordre d’ Actinopterygii du Carbonifère inférieurdu Montana (États-Unis).Le nouvel ordre des Guildayichthyiformes n. ord. (Actinopterygii) est créépour Guildayichthys carnegeiei n. gen. n. sp. et Discoserra pectinodon n. gen. n.sp. du calcaire de Bear Gulch, Chesterien supérieur, du Montana. Ces pois-sons marins ont le corps fortement comprimé et discoïde, les écailles hautes etrhombiques avec des articulations en tenon et mortaise, les rayons desnageoires segmentés et espacés, la bouche petite et la série operculo-gulaireavec des os spécialisée. Le toit crânien comporte plusieurs grands os médiansdepuis le rostre jusqu’à l’occiput, trois os dans la série du canal otique et plu-sieurs séries longitudinales d’éléments préfrontaux. Les régions préorbitaire etde la joue sont également singulières. La joue a un nombre très variable desousorbitaux, des os du canal préoperculaire, dorsal et ventral, hauts et étroitset une série de rayons (Guildayichthys carnegiei n. gen. n. sp.) ou d’os(Discoserra pectinodon n. gen. n. sp.) interoperculaires entre l’os préoperculaireventral et la série branchiostège. Il y a plusieurs os postspiraculaires et huit àdix os sclérotiques dans l’œil. Le grand développement des os médians du toitcrânien chez les Guildayichthyiformes n. ord. est comparable aux dispositionsrencontrées chez les Dipneustes et les Placodermes, qui concentrent égale-ment en avant de l’orbite les forces de préhension, suggérant ainsi desréponses adaptatives similaires à des types semblables de stress crâniens. Uneétude cladistique impliquant 72 caractères implique une relation de groupesfrères très stable entre Polypterus et les Guildayichthyiformes n. ord., quiconstituent un groupe coronal (crown group) au sein des Actinopterygii paléo-zoïques analysés. Les Cladistia sont redéfinis en un super ordre pour exprimercette parenté. Les Cladistia apparaissent comme le groupe frère de Platysomuset d’Amphicentrum.

MOTS CLÉSGuildayichthys carnegiei n. gen. n. sp.,

Discoserra pectinodon n. gen. n. sp., Osteichthyes, nouvel ordre,

Guildayichthyiformes n. ord., Cladistia,

Namurien, Montana,

morphologie crânienne.

New Actinopterygian order from Carboniferous of Montana

173GEODIVERSITAS • 2000 • 22 (2)

archived with the Section of Vertebrate Paleon-tology, Carnegie Museum of Natural History,Pittsburgh, Pa. (USA).Specimens were prepared manually and studiedeither from the original bones or from latex peels(Baird 1955). The courses of the calcite filled late-ral line canals were determined by viewing thebones under alcohol. Peels were shadowed withmagnesium oxide. Images were acquired using ahigh-resolution color flatbed scanner. Illustrationswere prepared with Adobe Photoshop and AdobeIllustrator software packages.Cladistic analyses were performed using theHennig86 and Clados programs.

BONE NOMENCLATURE

The bone terminology used in these descriptionsattempts to favor that used for Actinopterygii.Certain elements, series of elements and arrange-ments of elements, however, are more readilycomparable with those of primitive Sarcopterygii,and others with Neopterygii, and those names areused. The names in question are applied topogra-phically, but there are ample reasons to suggest,as have Arratia & Cloutier (1996), that probablehomologies could exist.

SYSTEMATIC PALAEONTOLOGY

Superorder CLADISTIA Cope, 1871

Order GUILDAYICHTHYIFORMES n. ord.

DIAGNOSIS. — Paired premaxillae not sutured in mid-line, maxilla does not extend posterior to midorbitallevel. Small rostral, paired lateral postrostrals followedby a large median second postrostral. Supraoccipitalprominent in skull roof. Two to three rows of pairedbones, including supraorbitals, over orbits. Threebones in anteroposterior otic series. Infraorbital seriesof six to ten bones lacking a specialized posterodorsalinfraorbital or “postorbital”. Suborbitals numerous,variable, extending under orbit. Narrow dorsal andventral preoperculars. Interopercular rays or interoper-cular bones present, that may carry a branch of thepreopercular lateral line canal. Single large opercularbone ventral to two to for smaller postspiraculars;branchiostegal rays few, not extending forward under

mandible. Braincase ossified as several bones; paras-phenoid extends entire length of braincase.Presupracleithrum large. Body laterally compressed,discoidal; squamation complete, deepened peg-and-socket scales on anterior flank. Caudal hemiheterocer-cal; fin rays well spaced.

Family GUILDAYICHTHYIDAE n. fam.

TYPE GENUS. — Guildayichthys n. gen. by original des-ignation.

DIAGNOSIS. — As for order, only family.

Genus Guildayichthys n. gen.

TYPE SPECIES. — Guildayichthys carnegiei n. gen. n. sp.

ETYMOLOGY. — Named after John Guilday, late Curatorof Fossil Mammals of Carnegie Museum, Pittsburgh, forhis unique appreciation of the beauty of life.

DIAGNOSIS. — Teeth short, and absent posterior tomid-maxillary and mid-dentary level. Posterior end ofmaxillary extending to midorbital level. Parietals meetin dorsal midline. Three rows paired bones over orbits.Five to six interopercular rays extending posterior toquadrate. Seven to eight branchiostegals, two dorsallyconcave, one expanded posteriorly and four to fiveventrally concave, plus two lateral gulars. Dorsal ridgescales from occiput to dorsal fin bear posteriorly pro-jecting spines; imperceptible transition from dorsalridge scales to dorsal fin rays. Rays of leading edges ofdorsal and anal fins closely spaced, following rays ofother fins well spaced; all fin rays jointed andunbranched. Anal fin with two to three short leadingspines, caudal fin slightly forked.

Guildayichthys carnegiei n. gen. n. sp.(Figs 1-4; 14A)

HOLOTYPE. — CM 41071.

REFERRED SPECIMENS. — MV 6045, 6046, 6932,7671, 7758, 7795. CM 27293, 27297, 35217, 37548,37549-37550, 41010, 46091, 46095, 46131, 46293,46294, 46295, 46296, 46297. ROM 36560, 41039,41042, 41047, 43012. FMNH PF10026.

ETYMOLOGY. — Named in honor of AndrewCarnegie, founder of Carnegie Museum of NaturalHistory, Pittsburgh, Pennsylvania.

HORIZON AND LOCALITY. — Bear Gulch Limestonelens, Bear Gulch Limestone member of the HeathFormation, Big Snowy Group, south of Becket, FergusCounty, Montana.

DIAGNOSIS. — For meristics see Table 1. Diagnosis asfor genus, only species.

Lund R.

174 GEODIVERSITAS • 2000 • 22 (2)

REMARKS

All skull bones have ganoine sculpturing of longridges; either the mesial or the anterior face ofeach ridge lies at a low angle to the plane of thebone while the opposite face meets the boneapproximately perpendicularly. Ganoine ridgeson the gulars are of the reverse orientation(Fig. 2). Scales of the anterior trunk have peg-and-socket joints and have ganoine sculpturing offaint marginal vertical lines; all trunk scales arepectinated and the deepened ventral abdominalscales are particularly strongly pectinated. Onlythe central flank scale rows are markedly deep-ened (rows 2-5 below the lateral line). The anteri-or scales of these rows average 3.85 times deeperthan wide. Squamation at the bases of dorsal andanal fins is of small thin scales, extending as lobesover and under the caudal peduncle.The statistics of Guildayichthys (Tables 1; 2) arebased upon an inadequate sample size and may

be misleading. Standard length-maximum heightcorrelation is strong, and caudal peduncle pro-portions also correlate. No other recorded para-meters show meaningful correlations.

DESCRIPTION

Lateral aspect of skull (Figs 2-4)The paired premaxillae (P, Figs 3; 4) are high,narrow and separated posteriorly in the midlinewhere they meet the rostral. They are firmly atta-ched only to the first infraorbital posteriorly.Short styliform to conical teeth are borne in anarrow band along the oral margin. The maxillais elongate and narrow (height one seventh oflength), extends to midorbital level and overlapsthe posterodorsal rim of the mandible. Shortteeth are borne only upon the anterior half of themaxilla.There are five to six bones in the infraorbitallateral line canal series (IO, Figs 3; 4). The most

FIG. 1. — Guildayichthys carnegiei n. gen. n. sp., holotype (CM 41071A), Lower Carboniferous Bear Gulch Limestone, HeathFormation, Fergus County, Montana (USA). Scale: 9 mm.



TABLE 1. — Meristics of Discoserra pectinodon n. gen. n. sp and Guildayichthys carnegiei n. gen. n. sp. Abbreviations: AFP, anal finposition; CFP, caudal fin position; CANG, caudal angle; BCL, braincase length; CPL, caudal peduncle length; CPW, caudalpeduncle width; DFP, dorsal fin position; GL, gape length; HDL, predorsal length; HMAX, maximum height; NANAL, number of analrays; NCAUD, number of caudal rays; NDORS, number of dorsal rays; NPECT, number of pectoral rays; NPELV, number of pelvicrays; PCR, precaudal ridge scales; PELPO, pelvic fin position; PDR, predorsal ridge scales; SL, standard length; SRALL, scale rowsfrom lateral line to dorsal fin origin; SRBLL, scale rows from lateral line to anal fin origin. Fin positions are in numbers of scale rowsalong the lateral line canal-bearing scales.

Discoserra pectinodon n. gen. n. sp.n range mean SD

DFP 19 17-23 19.95 1.62NDORS 16 41-54 46.38 2.8AFP 19 13-21 16.53 1.93NANAL 15 28-43 34.53 4.17CFP 21 34-48 43.05 2.87NCAUD 19 22-27 24.26 1.37NPECT 7 11-16 14.43 1.62PELPO 10 9-12 10.80 0.92NPELV 5 3-5 4.20 0.84PDR 18 19-29 24.00 3.27PCR 11 2-5 3.00 1.91SRALL 13 8-11 9.46 0.97SRBLL 19 8-13 9.89 1.19SL 26 46-57 51.50 3.63HDL 20 36-43 39.79 2.67GL 20 7-10 8.51 0.69BCL 26 16-20 17.97 0.99CPL 23 3-6 4.55 0.56CPW 23 7-10 8.27 0.65HMAX 25 34-44 38.04 2.45CANG 22 30-44 38.59 4.58

Guildayichthys carnegiei n. gen. n. sp.n range mean SD

8 16-20 18.38 1.517 33-40 37.29 2.698 11-18 14.63 2.449 26-34 29.78 2.788 33-38 35.75 1.677 20-21 20.14 0.38– – – –6 8-9 8.83 .044 13-16 13.75 2.227 14-21 16.29 2.295 4-7 5.40 1.528 10-15 11.88 1.648 9-12 10.63 1.067 49-56 55.89 2.946 36-43 39.42 2.865 8-11 9.39 1.057 17-19 17.57 0.698 2-7 5.08 1.688 8-10 9.27 0.608 34-40 36.64 2.208 28-34 31.25 2.76

PDR SL HDL GL BCL CPL CANG

AFP .9246

.004**CFP –.940

5.009*

NCAU .913 .9126 6

.005* .006*PDR –.908

5.016*

GL -.9524

.024*CPW .762 .712

6 8.762* .024*

HMAX .893 .8657 5

.003* .029*

TABLE 2. — Pearson correlation matrix, Guildayichthys carnegiei n. gen. n. sp. Numbers are correlation coefficient/number of speci-mens/significance. Abbreviations: AFP, anal fin position; BCL, braincase length; CANG, caudal angle; CFP, caudal fin position; CPL,caudal peduncle length; CPW, caudal peduncle width; GL, gape length; HDL, predorsal length; HMAX, maximum height; NCAU, num-ber of caudal; PCR, precaudal ridge scales; PDR, predorsal ridge scales; SL, standard length; *, significant at 5% level; **, significantat 1% level.

Lund R.

176 GEODIVERSITAS • 2000 • 22 (2)

anterior is narrow and forms the ventral edge ofthe posterior nostril. The second is large andmoderately expanded anteriorly. The remainingelements are quite narrow, the most posterodor-sal of which is tightly associated posteriorly witha thin and large bone. The latter two bones coverthe dilator opercular fossa and articulate with theskull roof. A series of thin suborbitals extendfrom the skull roof to midorbital level.There is a short dorsal and a taller ventral preop-erculum, both anteroposteriorly narrow, carryingthe preopercular canal from the margin of theskull roof to the posterior end of the lower jaw(POPD, POPV, Figs 3; 4). Two small postspirac-

ular bones (see Polypterus, Pehrson 1958) lie dor-sal to the tall, narrow, principal bone of the oper-cular series. The opercular is flanked ventrally bya series of branchiostegals, two dorsally concave,the next a posteriorly widened triangle, followedby 4-5 ventrally concave, deeply overlappingbones and a ventral plate (posterolateral gular).The plate is divided anteroposteriorly in onespecimen (Fig. 3); a narrow lateral gular lies ante-rior to the plate and mesial to the mandible.A series of five mobile, overlapping interopercu-lar rays (Fig. 2; I, Fig. 3) extends from the poster-ior margin of the mandible to the anteroventraledge of the opercular. There is strong evidence in

FIG. 2. — Guildayichthys carnegiei n.gen. n. sp., latex peel of the head (MV 6045B) of holotype, Lower Carboniferous Bear GulchLimestone, Heath Formation, Fergus County, Montana (USA). Scale: 6 mm.



MV 6045 of either lateral line pores or pit-organsin the two rays closest to the mandible. The late-ral surface of the mandible consists of a large den-tary and smaller, posterior, angular. Nosurangular is visible. Short conical teeth occupythe anterior half of the oral margin of the denta-ry. The mandibular lateral line canal traverses themandible near its ventral margin, and bendsabruptly dorsad in the angular. The articularfacet is located slightly below a projection of thedorsal margin of the jaw that presumably functio-ned as a coronoid process.

Skull roof (Figs 3; 4; 14A; 15C)The small median rostral (R, Figs 3; 4) is follo-wed by a paired postrostral that is notched for themesial margin of the dorsal nostril. An ethmoidcommissure has not been observed. A moderatelylarge median second postrostral extends past thelevel of the anterior edge of the parietals. Thesupratemporal commissure crosses the midlineanterior to the posterior margin of the supraocci-pital (Figs 3; 14A).The posterior border of the anterior (dorsal) nasalopening is formed by the most anterior of a series

M

P

RPR

AN

O

G

D

A

S

PA

E

SC

PSP

PC

C

POPVPOPD

F

PF1

PR

IO

SO

N1

N2

PF2

B

I

IT STT

FIG. 3. — Guildayichthys carnegiei n. gen. n. sp., latex peel of the head (MV 6045B) of holotype, lateral aspect of skull, LowerCarboniferous Bear Gulch Limestone, Heath Formation, Fergus County, Montana (USA). Abbreviations: A, angular; AN, presupra-cleithrum; B, branchiostegal rays; C, cleithrum; D, dentary; E, extrascapular; F, frontal; G, gular; I, interoperculars; IO, infraorbital; IT,intertemporal; M, maxilla; N1, N2, nasals; O, opercular; P, premaxilla; PA, partietal; PC, postcleithrum; PF1, PF2, prefrontals; POPD,POPV, dorsal and ventral preoperculums; PR, postrostral; PSP, postspiracular; R, rostral; S, supraoccipital; SC, supracleithrum; SO,supraorbital; ST, supratemporal; T, tabular. Sclerotic bones stippled; suborbital bones hatched. Scale: 3 mm.

Lund R.

178 GEODIVERSITAS • 2000 • 22 (2)

P

R

PRPR

S

M

FPAE

SC

PF

SO

N

IO

PSP

POPD

POPVD

A

I

ITST

T

PT

AN

FIG. 4. — Guildayichthys carnegiei n. gen. n. sp., dorsolateral aspect of skull, from latex peel (CM 41010), Lower Carboniferous BearGulch Limestone, Heath Formation, Fergus County, Montana (USA). Abbreviations: A, angular; AN, presupracleithrum; D, dentary; E,extrascapular; F, frontal; I, interoperculars; IO, infraorbital; IT, intertemporal; M, maxilla; N, nasals; P, premaxilla; PA, partietal; PF,prefrontals; POPD, POPV, dorsal and ventral preoperculums; PR, postrostral; PSP, postspiracular; PT, posttemporal; R, rostral; S,supraoccipital; SC, supracleithrum; SO, supraorbital; ST, supratemporal; T, tabular. Suborbital bones hatched. Scale: 3 mm.

of two to three narrow bones. This series extendslateral to the postrostrals, and mesial to the nasaland supraorbital, and abuts posteriorly againstthe frontal, the intertemporal, or both (Figs 3; 4).A thin nasal bone bridges the space between nos-trils. The course, or even the presence of thesupraorbital canal in the nasal region is uncertain;the canal can only be followed in the frontal andparietal and may lie under the suture between thepre-frontal and the more lateral series. Thesupraorbital canal in MV 6045 bears a short pos-teromesial branch in the frontal before conti-nuing on to the parietal; the canal does not reachthe transverse pit line. There are three bones inthe otic canal series, the largest and most anterior,the intertemporal (IT, Figs 3; 4), receiving the

infraorbital canal from the posterodorsal infraor-bital. Pores associated with a profundus branch ofthe otic canal (Poplin 1973) extend forwardwithin the intertemporal from the junction withthe infraorbital canal in MV 6045 (Fig. 3). Twoto three pores in a transverse row extend mesiallyacross the supratemporal and parietal toward thetransverse pitline but there is no evidence for anunderlying canal branch. The posterior margin ofthe skull roof in MV 6045 is bordered by onepair of bones bearing the canal intersection cha-racteristic of the extrascapular. In CM 41010,however, two pairs of bones are evident in thesame space, posttemporal and extrascapular(Fig. 4). In neither condition do these bonesextend to the dorsal midline.



The braincase and the floor of the olfactory cupare well ossified. Details, however, are obscuredby overlying bones. A short process, possibly apalatobasal (basal) process, is visible in the orbitalregion of CM 41010. Neither palate nor visceralskeleton is visible.

PostcranialShoulder Girdle. The posttemporal is abuttedanteroventrally by a presupracleithrum (Lund &Melton 1982) and posteroventrally by a largesupracleithrum. The supracleithrum-cleithrumcontact is masked by the operculum; a tall, thinpostcleithrum is present and may have served tostrengthen what appears to be a very limited areaof contact. The cleithrum is deeply notched forthe insertion of the pectoral fin. The scapulocora-coid is ossified in a thin sheet and supports oneradial for each pectoral fin ray. Clavicles areabsent.Scales have strong peg-and-socket articulations,smooth ganoine coatings, and are slightly pectina-

ted. Anterior flank and abdominal scales are mode-rated deepened. There is an abrupt caudal inver-sion, and the scaled body axis continues past thelast caudal fin ray. A complete series of median dor-sal scutes extends from the head to the origin ofthe dorsal fin; as the series approaches the dorsalfin, the scutes increase in height and the sharp pos-teriorly directed spine of each overlaps the spine ofthe following scute. The bases of the first jointedfin rays originate behind the last of the dorsal scutesand level with the bases of the scutes. There aretwo to three dorsal scutes between the end of thedorsal fin and the caudal lobe, followed by a seriesof dorsal caudal scutes that extend to the end ofthe caudal axis. There is a series of short, stout,strongly serrated abdominal scutes that extendfrom the shoulder girdle to the pelvic fins.All fins are composed of well-spaced and jointedrays. The pectoral, dorsal and anal fins are sup-ported on finely scaled lobes. The pelvic fin ismoderate in length, and the caudal fin is slightlyforked and slightly inequilobate. There are no fin

FIG. 5. — Discoserra pectinodon n. gen. n. sp., holotype (CM 30621A), Lower Carboniferous Bear Gulch Limestone, HeathFormation, Fergus County, Montana (USA). Scale: 8 mm.

Lund R.

180 GEODIVERSITAS • 2000 • 22 (2)

cc n p

NDORSALNANAL .648 13 .008*NCAU .602 14 .011*PCR .804 11 .001**NCAUNDORSAL .602 14 .011*CFP .542 17 .012*HMAX .564 18 .007*CFPNCAU .542 17 .012*PCR .653 11 .015*PELPO –.691 10 .013*SRALLSL .762 13 .001**HDL .784 13 .001**BCL .723 13 .003**SRBLLPCR .741 11 .003**NPECT –.905 07 .003**DFPPDR .710 17 .001**BCL .531 19 .01*SLHDL .854 20 .000**HMAX .807 25 .000**CPW .585 23 .002**HMAXSL .807 25 .000**HDL .748 20 .000**BCL .693 25 .000**CPW .723 22 .000**CPLCANG .515 21 .008*CPWSL .585 23 .002**BCL .518 23 .006HMAX .723 22 .000**

TABLE 3. — Pearson correlation coefficients, Discoserra pectino-don n. gen. n. sp. Abbreviations: BCL, braincase length; CANG,caudal angle; cc, correlation coefficient; CFP, caudal fin posi-tion; CPL, caudal peduncle length; CPW, caudal pedunclewidth; DFP, dorsal fin position; HDL, predorsal length; HMAX,maximum height; n, number; NANAL, number of anal rays;NCAU, number of caudal rays; NDORSAL, number of dorsalrays; NPECT, number of pectoral rays; p, Pearson coefficients;PCR, precaudal ridge scales; PELPO, pelvic fin position; PDR,predorsal ridge scales; SL, standard length; SRALL, scale rowsfrom lateral line to dorsal origin; SRBLL, scale rows from lateralline to anal origin; *, significant at 5% level; **, significant at 1%level. Fin positions are in numbers of scale rows along the lateralline canal-bearing scales, measurements in mm.

fulcrae. No postcranial endoskeletal detail is avai-lable.

Genus Discoserra n. gen.

TYPE SPECIES. — Discoserra pectinodon n. gen. n. sp. byoriginal designation.

ETYMOLOGY. — Discoserra, serrated disc, descriptiveof the appearance of the body in lateral view.

DIAGNOSIS. — Teeth of the premaxilla, maxilla and den-tary long, thin, and styliform. Posterior end of maxilladoes not extend back to level of anterior margin of orbit.Parietals excluded from contact in dorsal midline bypostrostral 2, which contacts supraoccipital. No trans-verse supratemporal commissure in supraoccipital. Tworows of paired bones over orbit. One to three inte-ropercular bones; two to three small postspiraculars anda presupracleithrum. Branchiostegals very variable insize, number and shape. Dorsal ridge scales with small,forwardly facing hooks; two to three small anal finhooks. Origin of anterior edge of dorsal fin set well belowdorsal margin of ridge scales. All fins with well spacedrays; pelvic fin reduced, caudal fin rounded.

Discoserra pectinodon n. gen. n. sp.(Figs 5-13; 14B)

HOLOTYPE. — CM 30621.

REFERRED SPECIMENS. — MV 2772, 2773, 2956,2984, 3579, 3810, 3811, 7669, 7670, 7756, 7757,7793, 7794. CM 27290-27292, 27294-27296,27298, 27333, 35206-35216, 35409, 35547,37545-37547, 37665, 41009, 44500, 46201-46206,48650, 48651, 48717-48720, 48841, 62794-62802.ROM 36562, 41030, 41169, 43003, 43115, 43976.

ETYMOLOGY. — Pectinodon, in reference to the longteeth.

DIAGNOSIS. — For meristics see Table 1. Other char-acters as for genus, only species.

REMARKS

The statistics of Discoserra (Table 3) are remar-kable. The number of dorsal fin rays correlates withthe numbers of anal and caudal rays as well as thoseof the precaudal ridge scales. The number of cau-dal rays also correlates with the number of lateralline scales and a size parameter, maximum height.Pelvic position in scale rows along the lateral linenegatively correlates with number of lateral linescales. Scale rows above the lateral line correlateswith size parameters, while scale rows below the

lateral line do not. Maximum height correlates wellwith most morphometrics, and excellently withcaudal peduncle width. Caudal peduncle lengthcorrelates only with caudal angle. Despite thesecorrelations, Discoserra pectinodon is morphologi-cally and statistically the most variable fish in the



Bear Gulch fauna. It has not proven possible toconsistently isolate variant morphologies thatwould suggest separate populations either throughspace or time; in fact several bone complexes suchas branchiostegals and suborbitals are not bilate-rally consistent within all individuals.

DESCRIPTION

GanoineSkull bones have ganoine sculpturing of coarse,irregular longitudinal to circumferential convexridges. The central areas of broad bones such asthe opercular may bear ganoine sculpturing thatgrades from tubercles centrally to ridges periphe-rally (Fig. 6).

ScalesLateral flank scales are up to 3.6 times deeperthan wide and bear strong peg-and-socket jointsas well as internal anterior thickenings. Mostlateral scale rows are deepened. Ganoine sculp-turing varies on the flank scales from few, relati-vely coarse circumferential grooves near thedorsum to finer circumferential striations atmidflank. The stout, deep ventral abdominalscales are strongly serrated in addition to bea-ring coarse circumferential grooves. Squamationat the bases of dorsal and anal fins is of small,thin scales, extended into lobes that project overand under the caudal peduncle. A significantbulk of dorsal and anal fin radial musculature isindicated.

Lateral aspect of skull (Figs 7; 8)The premaxillae (P, Figs 7; 8) are about fivetimes longer than wide, and are not firmly sutu-red either across the midline or to any otherbones. There appears to have been a tendency forthem to fuse, in some individuals, to the anterior-most infraorbital posteriorly. The premaxillaebear a single row of long, fine, styliform, closelyset teeth. The maxilla (M, Figs 7; 8) is most oftentriangular in shape, varying from 1: 4 to 1: 5 inlength: height ratio. It is not firmly held to anyother skull bones. A mesial view of the maxilla ofone specimen is available, that shows whatappears to be a slight anterior articular facet

(CM 27290). The single row of long, closely fitteeth diminishes abruptly in height and ends atone third of the distance to the posterior end ofthe bone.The infraorbital lateral line canal bones are veryvariable in number (I0, Figs 7-9). The firstinfraorbital is short and in close contact with thepremaxilla. The second element is wedge shaped,of variable length, and may either be absent orfused to the large third anterior infraorbital insome specimens. When present, it forms theanteroventral margin of the posterior nostril. Thethird anterior infraorbital is large and trapezoidal,with pores distributed irregularly and remotefrom the course of the infraorbital canal. As manyas five to seven additional infraorbitals, most onlysurrounding the infraorbital canal, rim the orbit.Pores from the infraorbital bones below the orbitare irregularly distributed upon the suborbitalbones ventrally. The most posterodorsal infraor-bital is firmly associated with a large, thin bonethat covers the dilator opercular fossa.Suborbitals (Figs 7; 8) on the posterodorsal cheekare scale-like. Under the orbit, the suborbitalbones vary in number and shape from one long element (CM 35215) to a series of seven(CM 27292). Normally, a trapezoidal element orelements lie dorsal to the articular region of themandible, followed by a variable number of smallposterodorsal bones; in several specimens ananterior triangular element intervenes betweenthe third infraorbital and the trapezoidal bone orbones (CM 27294, 35206, 35211).The dorsal of the two preopercular bones contactsthe skull roof; it is shorter and thinner than theventral element. The preopercular canal branchesinto an anteroventral and a ventral branch near themandibular end of the bone (Fig. 8). There is onetall, narrow principal opercular bone, ventral totwo to four considerably smaller bones and a largepresupracleithrum (PSP, Figs 7; 8). The bran-chiostegals are very variable and are not always bila-terally symmetrical. A typical pattern is one dor-sally concave ray below the opercular, followedventrally by one that widens posteriorly, four to sixventrally concave rays and a single ventral plate(posterolateral gular). Variations are detailed in

Lund R.

182 GEODIVERSITAS • 2000 • 22 (2)

Figure 10. One to three small interopercular bones(I, Figs 7; 8) are found between the posterior endof the mandible, the posteroventral end of the ven-tral preopercular and the anteroventral end of theopercular. There is clear evidence for a branch ofthe mandibular lateral line canal extending underthe anterior of these bones, and some indicationthat the ventral branch of the preopercular lateralline canal extended ventrally between the inter-operculars. This is best displayed in CM 35206.The external aspect of the lower jaw consists of alarge dentary and a small angular, with a singlerow of long, slim teeth along the anterior oralmargin of the dentary. A surangular seems to beabsent. The internal aspect of the mandibleconsists of two elements, an extensive prearticularthat seems to lack teeth, and an articular with astrongly elevated anterior wall and a dorsomesial

pit for adductor musculature insertion(CM 27290). The mandibular lateral line canallies near the ventral border of the dentary; thecanal is L-shaped in the angular but a posteriorcontinuation of the canal is evident in several spe-cimens.

Skull roof (Figs 7; 8; 14B)The long, narrow median rostral (R, Figs 7; 8)extends back to the level of the anterior border ofthe nostrils. While lateral line canals enter themedian rostral from either side, a complete eth-moid commissure has not been seen. The pairedpostrostrals extend along the dorsal midline bet-ween the nostrils, receiving the posteromedian endof the median rostral between them. The posteriormedian postrostral contacts the supraoccipital pos-teriorly and excludes the frontal and parietal bones

FIG. 6. — Discoserra pectinodon n. gen. n. sp., latex peel of skull (CM 35211B), Lower Carboniferous Bear Gulch Limestone, HeathFormation, Fergus County, Montana (USA). Scale: 4 mm.



from the dorsal midline (Fig. 14B). The supraoc-cipital is a significant bone of both the braincase,the skull roof, and the posterior cranial surface,where it bears a strong posteriorly projecting crest.The supratemporal commissure does not enter thesupraoccipital.The more anterior of two pre-frontal bones (PF,Figs 7; 8) forms the posterior border of the ante-rior nostril, and a small nasal bone (N, Figs 7; 8)extends between the anterior and posterior nasalopenings. The posterior prefrontal is narrow andflanked laterally by a single supraorbital bone.The supraorbital lateral line canal extends fromthe posterior pre-frontal through the frontal and

parietal; the canal cannot be traced anterior tothe posterior pre-frontal. The parietal bears ashort transverse pitline, which may be continuedlaterally onto the supratemporal, the second ofthe three bones of the otic series, in the form of afew pores or pits. Behind the skull roof the supra-temporal canal intersects with the extrascapularcanal in a bone lateral to two paired extrascapu-lars, but there is no evidence for a complete com-missure.

Visceral skeletonThe palate and suspensorium are displayed only inCM 27290 (Fig. 11). No palatal teeth can be seen.

P

R

M

PR

PRS

AN

O

G

NPF

F PA

ITST

T

E

IOPSP

PSP

SC

C

D

I

B

A

FIG. 7. — Discoserra pectinodon n. gen. n. sp., lateral aspect of skull (CM 35211B), Lower Carboniferous Bear Gulch Limestone,Heath Formation, Fergus County, Montana (USA). Abbreviations: A, angular; AN, presupracleithrum; B, branchiostegal rays; D, den-tary; E, extrascapular; F, frontal; G, gular; I, interoperculars; IO, infraorbital; IT, intertemporal; M, maxilla; N, nasal; O, opercular; P,premaxilla; PA, partietal; PF, prefrontals; PR, postrostral; PSP, postspiracular; R, rostral; S, supraoccipital; SC, supracleithrum; ST,supratemporal; T, tabular. Sclerotic bones stippled; suborbital bones hatched. Scale: 4 mm.

Lund R.

184 GEODIVERSITAS • 2000 • 22 (2)

The hyomandibula (H, Fig. 11) has a ventral pro-cess one half of the total height of the bone and hasa prominent central foramen for the hyomandi-bular branch of nerve VII. The metapterygoid(ME, Fig. 11) is attached only to the anterior edgeof the ventral process of the hyomandibula, leavingample room for a large spiracle. There is an indi-cation of a symplectic and a palatal groove for thisbone near the posterior edge of the quadrate. Thequadrate condyle is poorly preserved but faced for-ward; a prominent ventral facet posterior to thecondyle may have been for attachment of a strongquadrate-mandibular ligament. A stout ceratohyalis visible, with branchiostegal rays attached, inFMNH PF10207.

There are eight to ten sclerotic bones in the eye,the largest of which are the dorsal and anteriorbones.

BraincaseThe braincase is well ossified and is devoid ofsutures that would delineate individual bones. Nooccipital fissure is evident, unless the small fora-mina (OC, Fig. 12) are remnants of such a struc-ture. Structures are best displayed in CM 27295,although CM 27290 and CM 35547 have alsocontributed information (Figs 11; 12). The para-sphenoid (PS, Figs 11; 12) extends the entire lengthof the braincase. It has small lateral wings, noobvious palatal articulations, and is strongly V-sha-

FIG. 8. — Discoserra pectinodon n. gen. n. sp., dorsolateral aspect of skull (CM 35214), Lower Carboniferous Bear Gulch Limestone,Heath Formation, Fergus County, Montana (USA). Abbreviations: A, angular; AN, presupracleithrum; B, branchiostegal rays; D, den-tary; E, extrascapular; F, frontal; I, interoperculars; IO, infraorbital; IT, intertemporal; M, maxilla; O, opercular; P, premaxilla; PA, par-tietal; PC, postcleithrum; PF, prefrontals; PR, postrostral; PSP, postspiracular; R, rostral; S, supraoccipital; SC, supracleithrum; SO,supraorbital; ST, supratemporal; T, tabular; TE, tectal. Sclerotic bones stippled; suborbital bones hatched. Scale: 2 mm.

M

PR

PR

R

P

S

AN

D

A

PAF

E

SC

PC

O

SO

F

PF

ITSTT

B

IO

TEPSP

I



ped in transverse section through most of its length.The ethmoid region is sheathed by two thin ossi-fications, the dorsal one also forming the floor ofthe nasal capsule and enclosing a foramen for theolfactory nerve. The sphenoid region is stronglynotched for the exit of the optic nerve (II, Fig. 12).The sphenotic ossification area projects laterallybeyond the skull roofing bones and is excavated bya deep pit with an apparent narrow foramen in itsfloor. This pit clearly is of the form and positionof the dilator opercular fossa (DO, Fig. 12) ofteleosts and seems to shows no relationship to aspiracular canal as reported for some chondros-teans (Patterson 1975). Foramina for branches ofcranial nerves V, VII, and X are visible (Fig. 12).A strong posteriorly directed supraoccipital crestprojects beyond the posterior face of the braincase.

PostcranialClavicles are absent in the shoulder girdle. Thecleithrum is deeply notched for the insertion ofthe pectoral fin and has a strong peg-and-socketarticulation with the supracleithrum. The post-cleithrum is small.All scales have strong peg-and-socket articula-tions, smooth ganoine coatings, and have finelypectinated posterior borders (Fig. 13C, D). Theanterior flank and abdominal scales are tall and

narrow.There is no caudal inversion, and theunscaled body axis extends to the last caudal finray and bears only long, fine dorsal scutes. A com-plete series of median dorsal scutes extends fromthe head to the origin of the dorsal fin; each scutetaller than the preceding one and bearing a smallforwardly facing hook. The bases of the firstjointed fin rays originate behind the last of thedorsal scutes and level with the bases of the scutes.There are two to three dorsal and ventral scutesbetween the end of the dorsal fin and the caudallobe. There is a series of tall, thick, strongly ser-rated abdominal scutes that extend from theshoulder girdle to the pelvic fins, and there aretwo to three pre-anal scutes.All fins are composed of well-spaced and jointedrays. The pectoral, dorsal and anal fins are sup-ported on finely scaled lobes. The pelvic fin is verysmall, and the caudal fin is rounded but notdorsoventrally symmetrical. There are no finfulcrae.

Postcranial endoskeleton (Fig. 13)The pectoral fin base is lobed. Up to 11 to 12 radialsare visible supporting the pectoral fin rays. The pel-vic fin is very small and no endoskeleton is known.The axial skeleton contains 12 to 13 precaudal seg-ments, 13 to 14 caudal segments anterior to the

B C

D E

A

FIG. 9. — Discoserra pectinodon n. gen. n. sp., infraorbital bones, Lower Carboniferous Bear Gulch Limestone, Heath Formation,Fergus County, Montana (USA). A, CM 35206A; B, CM 35207A; C, CM 35209B; D, CM 35211B, from latex peel; E, CM 35214A, fromlatex peel. Arrow indicates anterior. Scale: 3 mm.

Lund R.

186 GEODIVERSITAS • 2000 • 22 (2)

caudal fin endoskeleton and 11 segments are visiblein the sharply upturned caudal lobe. Precaudalneural arches are paired, and while there are blockyprecaudal ventral arch elements, there are no ossi-fied interarcuals. Caudal neural arches are fusedacross the midline. Neural spines are not fused tothe arches but extend from occiput to the end ofthe caudal as a complete separate series. Dorsal tothe neural spines, a complete supraneural seriesextends from the occiput to the posterior end ofthe dorsal fin, the predorsal supraneurals each bea-ring a dorsal facet for articulation with the dorsalridge scales (Fig. 13A, B). Approximately two dor-sal fin rays are supported by each supraneural(baseost); axonosts are not evident. Haemal archesare fused across the midline and articulate withseparate haemal spines; these in turn support theinfrahaemal series. The origin of the anal fin is sup-ported by two haemal and infrahaemal elements.The anal fin ray: infrahaemal ratio is also approxi-mately 2: 1.

DISCUSSION

ADAPTATIONS

Guildayichthys and Discoserra are clearly sister taxa,separated from all other actinopterygians by thecommon possession of unique skull, opercular andcheek osteology. The two genera differ subtly inbody and fin form, and both are median and pai-red fin propulsive specialists for maneuvering ingeometrically complex environments (Webb1984). Propulsive system differences, includingbody and tail form, leading edges of dorsal and analfins, size of pelvics and height of anterior flankscales, indicate subtle differences in propulsive spe-cializations. Similarly, the two taxa differ in den-tition and mouth gape, indicating differing dietaryadaptations. D. prionodon, with a very attenuatedgape and long, gracile teeth was clearly a pluckingand grazing feeder, while the longer gape and shor-ter, well spaced teeth of G. carnegiei are somewhatcloser to a generalized condition.

A B C

D E F G

H I J

FIG. 10. — Discoserra pectinodon n. gen. n. sp., branchiostegal series, Lower Carboniferous Bear Gulch Limestone, HeathFormation, Fergus County, Montana (USA). A, MV 2956; B, MV 3811; C, CM 27292B; D, E, CM 44500A, B; F, CM 35211A;G, CM 35215B; H, FMNH PF 10207; I, J, CM 37547A, B. Scale: 3 mm.



The feeding mechanisms of the Guilday-ichthyiformes are highly specialized for pluckingand suction-feeding. The extensive alterations ofthe operculo-gular series are manifestly associatedwith great deepening of the pharyngeal region andrestriction of the posterior opening of the opercu-lar chamber and stabilizing or strengthening theopercular pumping system, and these changes arefunctionally associated with restrictions of the sizeof the gape, toward much more effective suctionfeeding. Loosening, extension and fragmentationof the bones of the cheek and snout, producinggreater snout mobility, may well have accompa-nied these alterations. It is possible that the inter-opercular elements are also associated with thesechanges; similar pressures evidently resulted in theevolution of the halecostome interopercular bone(Schaeffer & Rosen 1961).

OSTEOLOGY

The most anterior bone of the rostral series of theGuildayichthyiformes is a median element, iden-tified as the rostral because of the passage of theethmoid commissure through it. Paired lateralpostrostrals lie medial to the nasal, and a largeposterior median postrostral follows posteriorly.

PS

Q CH SY

O

H

ME

ITST

T

PAF

S

FIG. 11. — Discoserra pectinodon n. gen. n. sp., palate and neu-rocranium (CM 27290A), Lower Carboniferous Bear GulchLimestone, Heath Formation, Fergus County, Montana (USA).Abbreviations: CH, ceratohyal; F, frontal; IT, intertemporal;H, hyomandibula; ME, metapterygoid; O, opercular; PA, pari-etal; PS, parasphenoid; Q, quadrate; S, supraoccipital; ST,supratemporal; SY, symplectic; T, tabular. Scale: 2mm.

OLDO

II PSH7B

SC

IT

ST T

PA

F

S

PTF

SP

X

VOC

HPPS

FIG. 12. — Discoserra pectinodon n. gen. n. sp., neurocranium (CM 27295A), from latex peels, Lower Carboniferous Bear GulchLimestone, Heath Formation, Fergus County, Montana (USA). Abbreviations: B, basioccipital; DO, dilator opercular fossa; F, frontal;HP, hypophyseal foramen; H7, foramen for hyomandibular branch of facial nerve; IT, intertemporal; OC, foramina for occipital artery;OL, olfactory cup; PA, parietal; PS, parasphenoid; PTF, post-temporal fossa?; S, supraoccipital; SC, supraoccipital crest; SP, sphe-notic ossification; ST, supratemporal; T, tabular; V, foramen for trigeminal nerve; X, foramen for vagus nerve; II, optic foramen.Scale: 2 mm.

The posterior median postrostral may be thetopographical equivalent of a pineal plate (Arratia& Cloutier 1996; Pearson 1982) but is signifi-cantly larger and is not perforated for a pinealforamen.

Lund R.

188 GEODIVERSITAS • 2000 • 22 (2)

A small nasal bone is notched both medially andlaterally, presumably for anterior and posteriornostrils, in the usual pattern for Actinopterygii.Two or three anamestic “pre-frontal” bones arefound in anteroposterior series lateral to the late-ral postrostrals, and an additional series of two tothree anamestic bones lie posterior to the nasals,lateral to the “pre-frontals” and medial to thesingle large supraorbital. This series seems to bethe topographic equivalent of the serial nasals ofParatarrasius (Lund & Melton 1982) and thefrontonasals of Coelacanthiformes (Lund &Lund 1985), but no lateral line canal can bedetected in this area.Parietals, bearing transverse pit-lines, carry thesupraorbital canal forward onto paired frontals.

Lateral to the parietals and frontals are threeclearly delineated bones of the skull roof thatcarry the otic lateral line canal and accompa-nying pit-lines. Several Sarcopterygii have morethan two otic canal bones, the X-Y series of cer-tain Dipnoi, Powichthys , Panderichthys ,Barameda and Youngolepis (Campbell & Barwick1986; Chang 1982; Jessen 1975; Long 1989;Vorobyeva & Schultze 1991; Young et al. 1992).The relationships among the surrounding andsupporting structures, canals and pit-lines, allstrongly suggest that the anterior bone of theseries may be homologous to the dermospheno-tic, and the posterior two elements homologousto or the equivalent of the dermopterotic of acti-nopterygians. The terms intertemporal, supra-

A

B

C D

FIG. 13. — Discoserra pectinodon n. gen. n. sp., Lower Carboniferous Bear Gulch Limestone, Heath Formation, Fergus County,Montana (USA). A, origin of the dorsal fin, from MV 7756; B, axial skeleton, from CM 35547, with CM 27295; C, first enlargedabdominal scale, from ROM 43976; D, scale rows 9-11, from ROM 43976. Scale: A, 3 mm; B, 5 mm; C,D, 4 mm.

temporal and tabular are most appropriate forthese bones.A large median bone between the parietals, exten-ding to the posterior margin of the skull roof, iscalled the supraoccipital. It is in the same topo-graphical position as bone B of Dipnoi(Campbell & Bell 1982) and quite similar to thenuchal of most dolichothoracid arthrodires(Goujet 1975), but has distinctly different rela-tionships to underlying structures. There are noother Paleozoic Osteichthyes with this bone asidefrom Dipnoi, and no homology with any othernamed supraoccipital is intended.There is no distinct antorbital bone in theinfraorbital canal bone series, although there is anenlarged infraorbital-canal-carrying bone anteriorto the orbit. The infraorbital canal itself is conti-nuous with the ethmoid commissure; there is nodetectable lateral line canal in the nasal and novisible anterior communication with the supraor-bital canal. The anterior bones of the infraorbitalseries are variable in shape and number in theGuildayichthyiformes. Suborbitals are numerousand variable in number and shape, as inParatarrasius (Lund & Melton 1982).The single dorsal suborbital is in the same anato-mical relationship to the posterodorsal infraorbi-tal, and presumed spiracle, as the spiracular(postspiracular) of coelacanths. Several small,irregular bones are also positioned dorsal to thelargest bone of the opercular series, which is desi-gnated the operculum. These smaller bones are inthe position of the posterior spiraculars ofPolypterus (Pehrson 1958) and Bobasatrania(Lehman 1956).Tall, narrow, dorsal and ventral bones carry thepreopercular lateral line canal ventrally from themargin of skull roof. There is no sensory linecommunicating between the preopercular andinfraorbital lateral line canals.A series of overlapping rays to small plates occu-pies a prominent area ventral to the ventral preo-percular bone, posterior to the lower jaw, and ante-rior to the operculum and branchiostegals, andattach to the posterior edge of the palate and ven-tral preoperculum. A branch of the preopercularcanal enters this bony series in at least one speci-

men. These elements occupy the positions of thepreopercular canal-carrying bones “9” of earlyDipnoi such as Chirodipterus, Dipnorhynchus andGriphognathus (Westoll 1949; Miles 1977;Campbell & Barwick 1982, 1984, 1986; Schultze& Arsenault 1987), the preopercular bone of sar-copterygians such as Quebecius and Glyptolepis(Schultze & Arsenault 1987; Jarvik 1972) and coe-lacanths (Lund & Lund 1985), as well as the inter-opercular bone of Neopterygii. They are calledinteroperculars for topographic reasons. Severalalternative interpretations of these bones are pos-sible:1. The guildayichthiform interoperculars areunique neomorphs, autapomorphies of thisgroup, with no homology to the interopercular ofhalecostomes. They may be derived, for instance,by fragmentation of the anteroventral end of theoperculum or preoperculum for functional rea-sons associated with evolution of the feedingmechanism, as has been advocated for the evolu-tion of the neopterygian interoperculum.2. Alternatively, the guildayichthiform inter-operculars have homologies with other bones inother fishes. Two possibilities ensue:2A. the condition in the two Guilday-ichthyiformes represents a transformation seriesfrom many rays or bones to a single bone. Thisprospect leads to 3A, B or C.2B. the reverse is the case; fragmentation of anoriginal ventralmost preopercular bone resultedin the 9 series in the origin of the Dipnoi and inthe interopercular rays in G. carnegiei. This yieldsan origin for the structures, and can conceivablylead to alternative 3.3A. The guildayichthyiform interoperculars arehomologous with the 9 a-f series of preopercularbones in primitive dipnoans (Miles 1977). Thedipnoan condition of 9 a-f series is the most pri-mitive known state of these bones, and the guil-dayichthyiform condition is derived from that.3B. The guildayichthyiform interoperculars arederived from a portion of a mandibular opercu-lum, as in the submandibular series of Jarvik(1972); the presence of a mandibular operculumwould be the most primitive state known amongOsteichthyes.



3C. The elements are homologous with theinteroperculum of Neopterygii.Alternative 1 is rejected on the basis of the pro-bable homologies shown between these elementsand the series 9 bones of sarcopterygians.If alternative 2 is accepted, that there are possiblehomologies, alternative 2A is unsupported by anyother evidence and is unlikely, and two of the cha-racter associations it would lead to, 3A and 3B, arealso unlikely. Alternative 3A would require that thepresence of this series of elements is a symplesio-morphy of actinopterygians and dipnoans, a pros-pect unsupported by any other evidence. Theobviously imbricated rays of G. carnegiei seem torule out any explanation that might relate to thefunctions of either a preoperculum or an inter-operculum, and suggest only alternative 3B, bran-chiostegal rays of a mandibular operculum.Counter to 3B, however, Pearson (1982) hasconvincingly argued that the elements in questionamong crossopterygians are branchiostegal rays ofa hyoidean operculum. Further, no other firm evi-dence can be demonstrated that a functional man-dibular operculum or aphetohyoidean conditionexisted, or could have existed, among the gnatho-stomes.Thus alternative 2B seems to be both the mostparsimonious and the most functionally elegantthesis, that the plesiomorphous condition of thepreopercular canal bone series in the cheek ofOsteichthyes consisted of three bones as seen, forinstance, in most Sarcopterygii, and that frag-mentation of the ventral-most element resultedin both the dipnoan and the guildayichthyiformconditions.Alternative 3C would support the concept thatthe Guildayichthyiformes could be the primitivesister group of the Neopterygii; this explanationseems quite plausible on other osteologicalgrounds, and would not be excluded by theacceptance of a plesiomorphous three-preopercu-lar state.The acceptance of 2B as an operational hypothe-sis, with or without 3C, further indicates that thesingle squamous cleaver-shaped preopercularbone, cleaver-shaped maxilla and absence ofinteropercular ossifications are synapomorphies

of a monophyletic group of Paleozoic-MesozoicActinopterygii, the Paleoniscimorpha (seeKazantseva 1981).Extrascapulars, bearing the supratemporal com-missure, lie immediately posterior to the occiputand the skull roof. A prominent element of theshoulder girdle anterior to the supracleithrum linksthe girdle to the skull. It does not bear a canal andit is in the same anatomical relationship and topo-graphical position as the presupracleithrum ofcoelacanths, other crossopterygians, several otherpaleoniscoids, and tarrasiids (Lund & Melton1982; Lund & Lund 1985). It has been termed thepresupracleithrum in Howqualepis by Long(1988).

CRANIAL ARCHITECTURE

The cranial osteology of the Guildayichthyiformesis unique among Actinopterygii. Many of the sin-gular features of the skull roofs of these fish hadvery strong structural, and consequently adaptive,value when viewed in the context of the architec-ture of the cranium.Major differences in cranial architecture betweenthe guildayichthyids and other PaleozoicActinopterygii involve: extension of the para-sphenoid to the rear of the braincase, a strongsphenotic ossification and a dilator opercularfossa, the presence of three lateral otic series skullroofing bones, the presence of numbers of loosepre-orbital snout bones, a large median posterioroccipital crest and the enlargement of median skullroof bones relative to lateral bones (Figs 14; 15),to manifest, in D. prionodon, a virtually completeantero-posterior median shield. Moreover, themidline of the skull roof is elevated to a sharp peak,dramatically altering the cross-sectional profile ofthe neurocranium from the inverted isosceles tri-angle of typical paleoniscoids to somethingapproaching a diamond shape.Skull roofs of early Osteichthyes and Arthrodiraseem to assemble into two basic configurations, alateral pattern and a median pattern of pairedversus median bones (Fig. 15). A median ethmo-rostral element, or chain of elements, is commonto both patterns but is not universal. A lateralpattern is typical of paleoniscoids (including

Lund R.

190 GEODIVERSITAS • 2000 • 22 (2)



Platysomidae and other deep-bodied forms), andsarcopterygians (Fig. 15D-F), where median skullroofing bones may be seen only in the ethmoidregion, if present at all in the lateral pattern; thesupratemporal commissure is straight and post-cranial, and prominent laterally placed posttem-poral bones are present. A median pattern is seenin guildayichthyids, Dipnoi, and Arthrodira(Fig. 15A-C). Median elements include anethmo-rostral, a bone spanning the orbits, and abone in the position of a supraoccipital. Thesupratemporal commissure bends onto the skullroof in one or more places. A strong posteriormedian vertical cranial or postcranial skeletalstructure may be a feature of this pattern. It canbe theorized that the two patterns reflect marked-ly different stress regimes upon regions of the cra-nium and skull roof. Some basic observationslend credibility to this conjecture.The most extensive fusions of bones of the rostralarea form a structural dome and are associated with

a posteroventral olfactory capsule and a lack of firmposterior palatal support. The consequence of thisdome is that strong biting stresses could be trans-ferred to, and dissipated around, the snout froman anterior, ethmo-palatine, joint or support.Examples include the “pariea-autostylic” (Stensiö1963) placoderms and the autostylic Dipnoi, bothwith anteriorly braced dental apparatus; the “endo-cranial autostylic” porolepids, with fangs, jaws, andpalate, braced on an extensively excavated, reducedethmo-rostral moiety; and the limited methyosty-lic (DeBeer 1937) guildayichthyids, where mostbiting stresses were concentrated preorbitally. It isnoteworthy that the extent of median bonesincreases with the reduction of the gape fromG. carnegiei to D. prionodon. More limited rostralarea ossifications are associated with some poste-rior palatal bracing, such as an extensive methyos-tylic suspensorium, a basal process, or an antoticprocess; and the transfer of a significant propor-tion of feeding mechanism forces posteriorly. They

A B

FIG. 14. — Skull roofs reconstructed in dorsal view. A, Guildayichthys carnegiei n. gen. n. sp., from MV 6045 and CMNH 41010;B, Discoserra pectinodon n. gen. n. sp., from CMNH 35211 and CMNH 35214.

Lund R.

192 GEODIVERSITAS • 2000 • 22 (2)

A B C

FED

FIG. 15. — Skull roofs of fishes. A, Arctolepis decipiens, after Goujet (1975); B, Dipnorhynchus kurikae, after Campbell & Barwick(1986); C, Guildayichthys carnegiei n. gen. n. sp. ; D, Panderichthys rhombolepis, after Vorobyeva & Schultze (1991); E, Hadronectordonbairdi, after Lund & Lund (1985); F, Cheirolepis trailli, after Pearson (1982). Skull roofs scaled to equal snout-occiput lengths,positions and sizes of (left) orbits approximate, median elements shaded.

may also be associated with the introduction ofdorsal olfactory capsules in the Actinopterygii orrostral organs in the Coelacanthiformes. Examplesinclude paleoniscoids, where much of the force ofthe bite is applied toward the rear of the palate andcheek and posterolaterally braced via the neuro-cranium; coelacanths, with very little bite force, ahighly flexible snout, and posterolateral bracing viaan antotic process; and osteolepid and panderich-thyid sarcopterygians, with strong posteriorlydirected neurocranial and endoskeletal bracing.A paired or median supra-orbital bone acts as abridge to transmit stresses across the orbits of thefish skull, the most structurally excavated part of theneurocranium, from ethmo-rostral to otic region.

The interorbital bridge may be broadly supportedby paired walls, or only underlain at each end bynarrow ethmoid and sphenoid buttresses and amembranous septum. It is conceivable that maxi-mized ethmo-rostral stress regimes will be associa-ted with near or total fusion of the ethmo-rostralstructures, as in porolepids, or with a fused medianrostral dome followed by a median bone that contri-buted to the distribution of stresses into the oticbraincase and skull roof, as in Dipnoi. Examples ofa median pattern include the Dolichothoraci andDipnoi, wide interorbitally with vertical orbitalwalls; and the guildayichthyids, narrow interorbi-tally, with dorsal sphenoid and ethmoid buttressesand a ventral parasphenoid girder.

The origin and considerable expansion of asupraoccipital bone and occipital crest, as well asthe posterior extension of the parasphenoid of theGuildayichthyidae, suggest the development ofsignificant stresses upon the occiput and posteriormedian skull roof by trunk musculature. They alsorecall the development of bone B of Dipnoi(Campbell & Barwick 1986) and the nuchal boneof Arthrodira (Goujet 1975). Guildayichthyidsalso have a continuous series of enlarged thoracicsupraneurals articulated with deeply imbricated,immobilizing median dorsal scales; the supraneu-ral series would function as a median dorsal keelfor musculo-skeletal origins that insert on the occi-put. A sharply peaked roof and an otically wideskull also strongly suggests vertical and lateral com-plexity of trunk-cranial musculature. The occipi-tal surfaces of arthrodire skull roofs have a largemedian nuchal, followed posteriorly by a mediandorsal girdle element, usually with a ventral keelfor muscle origin; cranio-thoracic joints, withactive mobility; prominent posttemporal fossae fortrunk muscular insertion; and strong occipital arti-culations (see for instance Goujet 1975). The pos-terodorsal surfaces of lungfish skulls have a largemedian bone B; extensive lateral cranial vaults(adductor mandibular fossae); a median cranialvault (median temporal fossa, with a median inter-nal keel supporting the median dorsal roofingbone); and complex lateral trunk muscular inser-tions (Miles 1977; Campbell & Barwick 1982). Aposterodorsal median skull roof element, whethercalled supraoccipital, nuchal, or “B”, is associatedwith evidence of complex stress regimes transmit-ted from the trunk as well as from lateral and ante-rior directions.Instances of median skull roof bone patternsaccordingly involve areas of the cranium ofmembers of several classes of fish. The skulls ofthe osteichthyan fish with median skull roofbone patterns, such as the Guildayichthyiformes,have evidently been extensively altered to shiftstresses drastically away from the plesiomor-phous lateral bone pattern of, for instance, atypical chondrostean-level fish, for very strongmechanical, adaptive, and therefore positivelyselected ends.

CHARACTER ANALYSIS

The Gui ldayichthyi formes are c lear lyActinopterygii but differ strongly from otherPaleozoic actinopterygians in many cranial cha-racters. The cladistic characters utilized weredrawn from the following sources:– Polypterus: Allis (1922); DeBeer (1937);Pehrson (1947, 1958) and specimens of Polypteruspalmas.– Fusiform Actinopterygii: Arratia & Cloutier(1996); Gardiner (1984); Long (1989); Lund &Poplin (1997); Lund, Poplin & McCarthy(1995); Moy-Thomas & Dyne (1938); Pearson(1982); Pearson & Westoll (1979).– Platysomiformes: Campbell & Le Duy Phuoc(1983); Lehman (1956); Moy-Thomas & Dyne(1938); Traquair (1879); Zidek (1992) and speci-mens.– Additional information on coelacanths: Lund& Lund (1985).Comments on the interpretation of certain of thecharacters follow:3. Premaxillary condition and relationships withother bones. Paired, toothed, tightly sutured pre-maxillae, without an included lateral line canal,are plesiomorphous for Osteichthyes. Premaxillaewith an included ethmoid commissure of thelateral line canal system are a derived state repre-senting fusion with rostral bony components;premaxillae including a bony territory with tri-partite canal are a separately derived state repre-senting fusion with the antorbital bone.4, 6. Rostral. The outgroup condition in sarcop-terygians is a micromeric/mesomeric bone pat-tern, with the canal carried by a rostral bone orbones (Pearson 1982; Lund & Lund 1985). Thiscondition of multiple median and paired rostralsis not unusual among the Actinopterygii (Lund& Melton 1982; Pearson & Westoll 1979) and isconsidered plesiomorphous for the Osteichthyes(Lund & Lund 1985). The rostral series ofCheirolepis trailli (Pearson 1982; Arratia &Cloutier 1996) also consists of multiple bones,but a correspondence of bone pattern or numbersof bones with the Guildayichthyiformes cannotbe established. The extension of the rostral seriesover much of the skull roof, and the precise



Lund R.

194 GEODIVERSITAS • 2000 • 22 (2)

combination of paired first postrostrals andmedian second postrostral, is autapomorphousfor the Guildayichthyiformes.11, 13. The series of small pre-frontals, anteriorto the frontals and lateral to the postrostrals(Fig. 15), is quite similar to, if simpler than, thefrontonasal series of bones in coelacanths (Lund& Lund 1985), porolepids (Jarvik 1972; Younget al. 1992) and dipnoans (Campbell & Barwick1986; Miles 1977; Denison 1968). Paratarrasiushibbardi has three pairs of lateral-line-carryingbones anterior to the frontals (Lund & Melton1982); it has also tectals, absent in Guildayi-chthyidae. Cheirolepis trailli has a pair of anames-tic pre-frontals and serial nasals (Pearson 1982).Other Actinopterygii have only three pairs ofbones in the series; parietals, frontals, and nasals.The “mosaic” condition of many small pre-fron-tals is plesiomorphous for the Actinopterygii andfor the Osteichthyes.15. Bones carrying the otic portion of the infra-orbital canal. The presence of three lateral linecanal bones in the otic series, accompanied by thelateral transverse pit-line on the supratemporal, isnot known from any published actinopterygians.A trace of this condition can be inferred in the24 mm stage of Polypterus (Pehrson 1947).Primitive Dipnoi (Miles 1977; Campbell &

Barwick 1986; Campbell & Bell 1982; Denison1968) also show three bones (X, Y1, Y2) in the oticseries as do Powichthys, Panderichthys, Barameda(Jessen 1975; Vorobyeva & Schultze 1991; Younget al. 1992) and early tetrapods (Jarvik 1972) wherethey are named intertemporal, supratemporal andtabular. Coelacanths, osteolepids and other earlyactinopterygians have two canal-carrying bones inthe otic series, while porolepids have only one.Tarrasius and Paratarrasius change ontogeneticallyfrom two bones to one (Lund & Melton 1982).Commonality suggests that two bones would bethe plesiomorphous state. This is supported by theobservation that those fishes with three otic seriesbones share very few other synapomorphies. Threebones in the otic series would be considered as aderived condition, and the reduction to one insome Actinopterygii, later coelacanths, andporolepids, is a separately derived state that hasoccurred several times among the subclasses. Thederived condition in this character appears to bethe retention into adulthood of an early develop-mental condition that is common to the develop-ment of early Osteichthyes.38, 39. Preoperculum. Two bones commonlycarry the preopercular canal across the cheek of pri-mitive coelacanths to join the infraorbital canal(Lund & Lund 1985). Three bones carry the preo-

A B D

F E

C

G

FIG. 16. — Cheek bones of Osteichthyes. A, Chirodipterus australis, after Campbell & Barwick (1982); B, Eusthenopteron foordi, afterJarvik (1972); C, Glyptolepis leptopterus, after Schultze & Arsenault (1987); D, Hadronector donbairdi, after Lund & Lund (1985);E, Paleoniscus frieislebeni, after Pearson (1982); F, Cheirolepis trailli, after Pearson (1982); G, Discoserra pectinodon n. gen. n. sp.

percular lateral line canal in Glyptolepis leptopterus,other porolepids, and primitive lungfish (Jarvik1948; Schultze & Arsenault 1987; Young et al.1992; Campbell & Barwick 1984, 1986). Twobones also carry the preopercular canal inMoythomasia and Howqualepis, some platysomids,and bobasatraniids (Zidek 1992). The larger, ven-tral preoperculum of the Guildayichthiformes is ofthe same form and anatomical relationship to thepalate and hyomandibula as that of Acentrophorus,halecomorphs and their relatives (Moy-Thomas &Miles 1971; Daget 1964). Dorsal (suprapreoper-culum) and ventral preoperculae are well-knownamong teleosts (Daget 1964). They bear littleresemblance to the cleaver-shaped or diamond-shaped squamous preoperculum or preopercularbones of the Paleoniscimorpha or other PaleozoicActinopterygii (Lund & Melton 1982; Lowney1980; Kazantseva 1981). The latter condition isconsidered a synapomorphy of certain PaleozoicActinopterygii, the Paleoniscimorpha.44. Anteoperculum [dermohyal]. There is noindication of an element with the tight anatomi-cal relationship to the hyomandibular bone andoperculum that could be interpreted as a dermo-hyal or anteopercular bone in sarcopterygians orguildayichthyids. Polypterus has a single anteoper-cular element dorsal to the preoperculum as domost paleoniscoids; Paratarrasius has a series ofanteopercular elements dorsal to the preopercu-lum. An anteopercular or anteopercular series istherefore regarded as a derived condition amongActinopterygii.45-47. Suboperculum and branchiostegals. Theopercular-gular series of the guildayichthyids isautapomorphous in lacking a clear subopercular,in the discontinuous and complex branchiostegalseries, and in the branchiostegals not extendingbehind or below the lower jaw.53. Postspiraculars. Several small elements, thepostspiraculars, are located between the principalopercular bone and the rear of the skull in bothG. carnegiei and D. pectinodon. They are equiva-lent to, and probably homologous to, that por-tion of the spiracular series of Polypterus locatedposterior to the dermohyal (Pehrson 1947). Theprespiraculars, anterior to the dermohyal of

Polypterus, are probably homologous to thesuborbitals of paleoniscoids and evidently to thespiracular or postspiracular bone of coelacanthsand other crossopterygians.55. Sclerotic bones. The guildayichthyids havesclerotic numbers that are higher than the normalcondition of two to four found among otherActinopterygii. Higher numbers of sclerotics cha-racterize many arthrodires, Chondrichthyes (Lund1985, 1986) and Sarcopterygii (Lund & Lund1984). The plesiomorphous condition for the gna-thostomes is considered here to be many sclerotics,although precise numbers can not be arrived at.56. Parasphenoid. A parasphenoid extending thelength of the braincase is derived relative toSarcopterygii, most Paleoniscimorpha (Gardiner1973; Lowney 1980; Pearson 1982), Acanthodii(Miles 1973) and Placodermi (Denison 1975).Guildayichthyiformes share this derived characterwith Neopterygii.57. Suspensorium. An autodiastylic suspenso-rium is plesiomorphous for gnathostomes(DeBeer 1937; Grogan 1993; Lund & Grogan1997); all known Osteichthyes have a hyomandi-bular bone, an enlarged epihyal articulated to thebraincase. The sarcopterygian non-suspensoryamphistyly is plesiomorphous for Osteichthyes(Grogan 1993). In Polypterus, there is limitedosteological and ligamentous contact between thehyomandibula and the palate, although a portionof the adductor mandibularis musculature origi-nates upon the hyomandibula (Allis 1922). Theguildayichthyid palate contacts the hyomandibu-la only along its ventral process and by means of asymplectic element. The typical actinopterygiancondition is a methyostylic suspensorium, themetapterygoid and hyomandibula being in exten-sive osteological contact.Cladistic analysis was performed using the mhen-nig procedure of Hennig86. The data matrixused for the analysis employed 72 characters of13 fusiform Paleozoic Actinopterygii and platyso-moid fish plus Polypterus. Of the characters used,50 were multi-state and the remainder were two-state characters (Appendix 1). Cheirolepis wasused as the outgroup. The analysis resulted in onetree of length 196, ci 64, and ri 63 (Fig. 17). All



Lund R.

196 GEODIVERSITAS • 2000 • 22 (2)

treatments result in the placement of Polypterus asthe sister taxon of the Guildayichthyidae;node 16 is supported by 16 character states(Appendix 2; data matrix in Appendix 3). Theseresults were unaffected by alternative interpreta-tions of the condition of the antorbital area, preo-perculum, suborbital series, or the condition ofthe premaxilla, the numbers or identities ofincluded taxa, or the choice of outgroup or out-groups. The strength of the Polypterus-Guildayi-chthyidae association is remarkable consideringthe long history of debate on the relationships ofthe Polypteriformes. The resulting clade, theCladistia, is placed as a crown group among theactinopterygians tested in this specific analysis.Perhaps more importantly, the Cladistia emergesas the sister clade of the included members of thePlatysomidae and Amphicentridae at node 17,supported by 15 characters.

CONCLUSION

The Guildayichthyiformes are characterized bydiscoidal body forms, lobed pectorals set high onthe flank, and well-webbed fins, adaptations for

swimming in geometrically complex habitats.The feeding mechanisms of these fishes featurepointed snouts, reduced gapes, reduced maxillae,and loosened snout bones, as well as restrictedand modified opercular bones. They are alsocharacterized by a unique assemblage of charac-ters, including an extended rostral and postrostralseries, multiple prefrontal ossifications, threebones in the otic series, an interopercular series,and an opercular bone surmounted by postspirac-ulars and followed ventrally not by a subopercu-lum but by a short, highly modified and veryvariable branchiostegal series. There is also evi-dence that, in Discoserra pectinodon, the reducedmaxilla had a limited ability to passively rotatearound an anteromesial articulation in associa-tion with mandibular opening. There are eight toten sclerotics of unequal size.A brief architectural analysis supports the thesisthat the guildayichthyiform skull is median-boneoriented, and is highly modified for structuralreasons from the typical lateral-bone orientedactinopterygian skull.Aspects of the rostral-postrostral series, thesupraorbital bones, the interopercular series,preoperculae, and sclerotics are readily compa-

Cheirolepis

Howqualepis

Mansfieldiscus

Cyranorhis

Wendyichthys

Moythomasia

Paratarrasius

Discoserra

Guildayichthys

Polypterus

P. Gibbosus

Amphicentrum

P. Superbus

23

22

21

20

19

18

17

15

13

16

14

11

10

12

9

7

8

1

2

5

6

4

3

0

FIG. 17. — Relationships of the Guildayichthyiformes n. ord. Characters were treated non-additively using the procedure mhennig;72 characters resulted in one tree of length 196, ci 64, and ri 64. The characters and character states used in the analyses are givenas Appendix 1, the characters supporting each node are listed in Appendix 2, and the data matrix is given in Appendix 3. The char-acters have been arrived at from Allis (1922); Arratia & Cloutier (1996); Campbell & Le Duy Phuoc 1983; DeBeer (1937); Gardiner1984; Lehman 1956; Long 1988; Lund & Lund (1985); Lund & Poplin (1997); Lund, Poplin & McCarthy (1995); Moy-Thomas & Dyne(1938); Pearson (1982); Pehrson (1947, 1958) and Zidek 1992, as well as from unpublished Bear Gulch Limestone taxa.



rable to features of the osteology of Polypteri-formes, primitive Dipnoi, and Sarcopterygii.Cladistic analysis results in a strong alignment ofthe Guildayichthyiformes as the sister taxon ofPolypterus within the Cladistia. The Cladistia inturn are indicated by the cladistic analysis as thesister group of Platysomus and Amphicentrum.

AcknowledgementsThis work has been funded by the Center For FieldResearch (Earthwatch), by United States NationalScience Foundation grants BMS 75-02720,DEB 77-02548 and DEB 79-19492, and by grantsfrom Carnegie Museum of Natural History,Pittsburgh, Pa., and The Royal Ontario Museum,Toronto, Ontario. I am profoundly grateful to allthe field volunteers who have helped in the exca-vations in the Bear Gulch Limestone through thedecades, and particularly to Ron Bugosh, CarolDeBell, William Hopkins, Brian Kleinhaus, MyraMoore, Bryan Snyder and the late Stuart Smith.I thank the Muséum national d’Histoire naturellefor the support and library facilities necessary tocomplete this article.The cladistic characters and associated data matrixused in this article are part of an ongoing series ofpapers by Cécile Poplin and myself describing theactinopterygians of the Bear Gulch Limestone. Mydeepest thanks to Cécile Poplin for her intensivediscussions on actinopterygian characters, and toDaniel Goujet for his advice and wise counsel oncladistics. The manuscript has benefitted greatlyfrom the comments of John Long, the commentsof an anonymous reviewer, and from those of thelate Colin Patterson.

REFERENCES

Allis E. P. Jr. 1922. — The cranial anatomy ofPolypterus. Journal of Morphology 12: 489-809.

Arratia G. & Cloutier R. 1996. — Reassessment of themorphology of Cheirolepis canadensis(Cheirolepididae: Actinopterygii): 165-197, inSchultze H.-P., Cloutier R. & Veziná D. (eds),Devonian Fishes and Plants of Miguasha, Quebec,Canada. Verlag Dr. Pfeil., München.

Baird D. 1955. — Latex micro-molding and latex-plaster molding mixture. Science 122: 202.

Campbell K. S. W. & Barwick R. E. 1982. — A newspecies of the lungfish Dipnorhynchus from NewSouth Wales. Palaeontology 25: 509-527.

Campbell K. S. W. & Barwick R. E. 1984. —Speonesydrion, an early Devonian dipnoan withprimitive toothplates. Paleo Ichthyologica 2: 1-48.

Campbell K. S. W. & Barwick R. E. 1986. —Paleozoic lungfishes, a review. Journal of Morphologysupplement 1: 93-131.

Campbell K. S. W. & Bell M. W. 1982. —Soederberghia (Dipnoi) from the Late Devonian ofNew South Wales. Alcheringa 6: 143-149.

Campbell K. S. W. & Le Duy Phuoc 1983. — A latePermian actinopterygian fish from Australia.Palaeontology 26: 33-70.

Chang M.-M. 1982. — The Braincase of Youngolepis,a Lower Devonian Crossopterygian from Yunnan,South-Western China. Swedish Museum of NaturalHistory, The Hague, 113 p. + 13 pls.

Cope E. D. 1871. — Contributions to the ichthyolo-gy of the Lesser Antilles. Transactions of theAmerican Philosophical Society 14: 445-483.

Daget J. 1964. — Le crâne des téleostéens. Mémoiresdu Muséum national d’Histoire Naturelle, N. S.Zoologie 31: 1-341.

DeBeer G. R. 1937. — The Development of theVertebrate Skull. Oxford University Press, London,552 p.

Denison R. H. 1968a. — Early Devonian lungfishesfrom Wyoming, Utah and Idaho. Fieldiana, Geology17: 353-409.

Denison R. H. 1968b. — The evolutionary signifi-cance of the earliest known lungfish Uranolophus.Nobel Symposium 4: 247-257.

Denison R. H. 1975. — Evolution and classificationof placoderm fishes. Breviora 432: 1-24.

Feldman H., Lund R., Maples C. & Archer A. 1994. —Origin of the Bear Gulch beds (Namurian, Montana,USA). Geobios M.S. 16: 283-291.

Fierstine H. L. 1968. — Swollen dorsal fin elementsin living and fossil Caranx (Teleostei, Carangidae).Los Angeles County Museum, Contributions in Science137: 1-10.

Gardiner B. G. 1973. — Interrelationships of teleos-tomes: 105-135, in Greenwood P. H., Miles R. &Patterson C. (eds), Interrelationships of Fishes.Academic Press, London.

Gardiner B. G. 1984. — The relationships of thepalaeoniscid fishes, a review based on new speci-mens of Mimia and Moythomasia from the UpperDevonian of Western Australia. Bulletin of theBritish Museum of Natural History 37 (4): 173-428.

Goujet D. 1975. — Dicksonosteus, un nouvelarthrodire du Dévonien du Spitzberg. Remarquessur le squelette viscéral des Dolichthoraci : 81-99, inLehman J.-P. (éd.), Problèmes Actuels dePaléontologie : Évolution des Vertébrés, Actes du

Lund R.

198 GEODIVERSITAS • 2000 • 22 (2)

Colloque International CNRS, Paris 4-9 juin 1973.CNRS, Colloques internationaux du Centrenational de la recherche scientifique 218, Paris.

Grogan E. D. 1993. — The Structure of the Holoce-phalan Head and the Relationships of the Chondrich-thyes. Ph.D. dissertation, School of Marine Science,College of William and Mary, Gloucester Point,Virginia, USA.

Grogan E. D. & Lund R. — The Environment of theBear Gulch Limestone (Mississippian of Montana).31 pp, 3 figs; submitted ms for Annals of CarnegieMuseum.

Horner J. R. 1985. — The stratigraphic position ofthe Bear Gulch Limestone (Lower Carboniferous)of central Montana. Compte Rendu, IXe Congrèsinternational de Stratigraphie et de Géologie duCarbonifère 5 : 427-436.

Horner J. R. & Lund R. 1985. — Biotic distributionand diversity in the Bear Gulch Limestone of cen-tral Montana. Compte-Rendu, IXe Congrès interna-tional de Stratigraphie et de Géologie du Carbonifère5 : 437-442.

Jarvik E. 1948. — On the morphology and taxonomyof the Middle Devonian osteolepid fishes. Kunglicasvenska Vetenskapsakademiens, Handlingar Ser. 3,25: 1-301.

Jarvik E. 1972. — Middle and Upper DevonianPorolepiformes from East Greenland with specialreference to Glyptolepis groenlandica n. sp. and a dis-cussion on the structure of the head inPorolepiformes. Meddelelser om Gronland 187:1-307.

Jessen H. 1975. — A New Choanate fish, Powichthysthorsteinssoni, n. g., n. sp., from the Early LowerDevonian of the Canadian Arctic Archipelago.Colloques Internationaux du Centre national de larecherche scientifique 218 : 213-225.

Kazantseva A. R. 1981. — The phylogeny of the loweractinopterygians. Journal of Ichthyology 21: 579-594; English translation 1982, 21: 1-16, ScriptaPublishing, Silver Spring, Md.

Lehman J. P. 1956. — Compléments à l’étude desgenres Ecrinesomus et Bobasatrania de l’Éotrias deMadagascar. Annales de Paléontologie 42 : 67-94.

Long J. A. 1989. — New palaeoniscoid fishes from theLate Devonian and Early Carboniferous of Victoria:1-64, in Jell P. A. (ed.), Devonian andCarboniferous Fish Studies, Memoir 7 of theAssociation of Australasian Paleontologists.

Lowney K. A. 1980. — Certain Bear Gulch(Namurian A, Montana) Actinopterygii (Osteich-thyes) and a Reevaluation of the Evolution of thePaleozoic Actinopterygians. Ph.D. thesis, New YorkUniversity, New York, USA.

Lowney K. A. 1985. — The morphology of Falcatusfalcatus (St John & Worthen), a Mississippianstethacanthid chondrichthyan from the Bear GulchLimestone of Montana. Journal of VertebratePaleontology 5: 1-19.

Lowney K. A. 1986. — On Damocles serratus nov.gen. et sp. (Elasmobranchii: Cladodontida) fromthe Upper Mississippian Bear Gulch Limestone ofMontana. Journal of Vertebrate Paleontology 6:12-19.

Lund R., Feldman H., Lund W. & Maples C.1993. — The depositional environment of the BearGulch Limestone, Fergus County Montana: 87-96,in Energy and Mineral Resources of Central Montana,Montana Geological Society Guidebook 1993,Billings.

Lund R. & Grogan E. D.1997. — Relationships ofthe Chimaeriformes and the basal radiation of theChondrichthyes. Reviews of Fish Biology andFisheries 7: 1-59.

Lund R. & Lund W. 1984. — New genera and speciesof coelacanths from the Bear Gulch Limestone ofMontana. Geobios 17: 237-244.

Lund R. & Lund W. 1985. — Coelacanths from theBear Gulch Limestone (Namurian) of Montana andthe evolution of the Coelacanthiformes. Bulletin ofCarnegie Museum of Natural History 25: 1-74.

Lund R. & Melton W. G. Jr. 1982. — A newactinopterygian fish from the Mississippian BearGulch Limestone of Montana. Palaeontology 25:485-498.

Lund R., Poplin C. & McCarthy K. 1995. —Preliminary analysis of the interrelationships ofsome Paleozoic actinopterygians, in Lelièvre H.,Wenz S., Blieck A. & Cloutier R. (eds), PremiersVertébrés et Vertébrés inférieurs. Geobios M.S. 19 :215-220.

Lund R. & Poplin C. 1997. — The rhadinichthyids(paleoniscoid actinopterygians) from the BearGulch Limestone of Montana. Journal of VertebratePaleontology 17: 466-486.

Miles R. S. 1977. — Dipnoan (lungfish) skulls andrelationships of the group: a study based on a newspecies from the Devonian of Australia. ZoologicalJournal of the Linnean Society 61: 1-328.

Moy-Thomas J. A. & Dyne M. B. 1938. — Theactinopterygian fishes from the LowerCarboniferous of Glencartholm, Eskdale,Dumfriesshire. Transactions of the Royal Society ofEdinburgh 59: 437-480.

Moy-Thomas J. A. & Miles R. S. 1971. — PaleozoicFishes. Saunders, Philadelphia, 259 p.

Patterson C. 1975. — The braincase of pholidophoridand leptolepid fishes, with a review of theactinopterygian braincase. Philosophical Transactionsof the Royal Society of London, 269 (B): 275-579.

Pearson D. M. 1982. — Primitive bony fishes, with aspecial reference to Cheirolepis and paleonisciformactinopterygians. Zoological Journal of the LinnaeanSociety 74: 35-67.

Pearson D. M. & Westoll T. S. 1979. — TheDevonian actinopterygian Cheirolepis Agassiz.Transactions of the Royal Society of Edinburgh 70:337-399.



Pehrson T. 1947. — Some new interpretations of theskull in Polypterus. Acta Zoologica 28: 400-455.

Pehrson T. 1958. — The early ontogeny of the senso-ry lines and the dermal skull in Polypterus. ActaZoologica 39: 241-258.

Poplin C. 1973. — Survivance de la “ligne épi-branchiale”, en particulier du “canal dorso-latéral-antérieur” (ligne profonde) chez les vertébrésinférieurs. Bulletin du Muséum national d’Histoirenaturelle 175 : 117-140.

Schaeffer B. & Rosen D. E. 1961. — Major adaptivelevels in the evolution of theactinopterygian feedingmechanism. American Zoologist 1: 187-204.

Schultze H.-P. & Arsenault M. 1987. — Quebeciusquebecensis (Whiteaves), a porolepiformcrossopterygian (Pisces) from the Late Devonian ofQuebec, Canada. Canadian Journal of Earth Sciences24: 2351-2361.

Stensiö E. (1963). — Anatomical studies on theArthrodiran head. Part 1. Kunglica SvenskaVetenskapsakademiens 9: 1-419.

Traquair R. H. 1879. — On the structure and affini-ties of the Platysomidae. Transactions of the RoyalSociety of Edinburgh 29: 343-391, 6 pls.

Vorobyeva E. & Schultze H.-P. 1991. — Descriptionand systematics of panderichthyid fishes with com-ments on their relationships to tetrapods: 68-109,in Schultze H.-P. & Trueb L. (eds), Origins of theHigher Groups of Tetrapods, Cornell UniversityPress, Ithaca.

Webb P. W. 1984. — Body form, locomotion and for-aging in aquatic vertebrates. American Zoologist 24:107-120.

Westoll T. S. 1949. — On the evolution of theDipnoi: 121-184, in Jepsen G. L., Mayr E. &Simpson G. G. (eds), Genetics, Paleontology andEvolution. Princeton University Press, Princeton.

Williams L. A. 1981. — The Sedimentational Historyof the Bear Gulch Limestone (Middle Carboniferous,Central Montana). Ph.D. dissertation, PrincetonUniversity, Princeton, USA.

Williams L. A. 1983. — Deposition of the BearGulch Limestone: a Carboniferous Plattenkalkfrom central Montana. Sedimentology 30:843-860.

Young G. C., Long J. A. & Ritchie A. 1992. —Crossopterygian fishes from the Devonian ofAntarctica: systematics, relationships and biogeo-graphic significance. Records of the AustralianMuseum Suppl. 14: 1-77.

Zidek J . 1992. — Late Pennsylvanian Chon-drichthyes , Acanthodi i , and deep-bodiedActinopterygi i f rom the Kinney Quarry,Manzanita Mountains, New Mexico. New MexicoBureau of Mines and Mineral Resources Bulletin138: 145-182.

Ziegler A. M., Scotese G. R., McKerron W., Johnson M.& Barnbach R. K. 1979. — Paleozoic paleogeogra-phy. Annual Review of Earth and Planetary Sciences7: 473-502.

Submitted for publication on 2 December 1998;accepted on 31 August 1999.

Lund R.

200G

EO

DIV

ER

SITA

S• 2000 • 22 (2)

Character State 0 State 1 State 2 State 3 States 4 and 5

0 Body form 0 - fusiform/elongate 1 - high dorsum, 2 - round 3 - angulatedflatter venter

1 Highest point 0 - no specializations 1 - ridge scutes 2 - few specialized scalesof body profile

2 Premaxillary position 0 - sutured in midline 1 - lateral, under antorbital 2 - absent as distinct bone3 Premaxillary condition 0 - sutured 1 - loose 2 - premaxilla field fused 3 - absent as distinct

to other bones bone4 Rostral 0 - separate bone or bones 1 - fused to postrostral 2 - fused to premaxilla 3 - fused to premaxilla 4 - absent as distinct bone

and postrostral5 Ethmoid commissure 0 - across midline 1 - not across midline 2 - absent6 Postrostrals 0 - several median 1 - single median 2 - median and paired 3 - paired 4 - absent as distinct bone7 Mouth 0 - mouth subterminal 1 - mouth terminal8 Snout shape 0 - rounded 1 - sharp bump 2 - pointed snout9 Rostral notch 0 - absent 1 - vertical 2 - keyhole-shape

10 Nasals 0 - serial 1 - single bone11 Supraorbitals 0 - absent 1 - single series 2 - double series12 Tectals 0 - present 1 - absent13 Frontals 0 - two pairs 1 - one pair14 Parietal/frontal ratio 0 - subequal 1 - much smaller parietal15 Otic canal bones 0 - dermosphenotic and

dermopterotic 1 - dermopterotic only 2 - three bones16 Infraorbital canal 0 - to dermosphenotic 1 - to dermopterotic 2 - to fused dpt-parietal17 Dsphenotic/nasal contact 0 - absent 1 - present18 Supraoccipital 0 - absent 1 - only on braincase 2 - on skull roof19 Supratemporal

commissure 0 - across midline 1 - not across midline 2 - loop to skull roof20 Extrascapulars 0 - three 1 - one pair 2 - two pairs 3 - absent21 Posteroventral IO is # 0 - two 1 - one 2 - > two22 Posterodorsal IO is # 0 - three 1 - two (< three) 2 - > three23 Posterodorsal IO shape 0 - simple 1 - T-shaped 2 - rev-L shaped24 Antorbital presence 0 - absent as distinct bone 1 - differentiated 2 - fused25 Antorbital shape 0 - rhombic 1 - inverted L 2 - vertical pillar 3 - fused to premaxilla 4 - absent as distinct bone26 Suborbital numbers 0 - none 1 - one row 2 - more than one row 3 - large one with canal27 Suborbital fit 0 - absent 1 - sutured 2 - loose, gaps28 Suborbital thickness 0 - absent 1 - thick 2 - thin, no ganoine29 Suborbital fit 0 - not between maxilla 1 - between maxilla 2 - separate maxilla

and infraorbital and infraorbital and infraorbital30 Maxilla length 0 - posterior to orbit 1 - orbital 2 - preorbital 3 - absent

APPENDIX 1

Characters and character states for the cladogram of Fig. 17. The characters have been arrived at from Allis (1922); Arratia & Cloutier (1996); Campbell & Le Duy Phuoc (1983);DeBeer (1937); Gardiner (1984); Lehman (1956); Long (1988); Lund & Lund (1985); Lund & Poplin (1997); Lund, Poplin & McCarthy (1995); Moy-Thomas & Dyne (1938); Pearson (1982);Pehrson (1947, 1958) and Zidek (1992), as well as from unpublished Bear Gulch Limestone taxa.

New

Actinopterygian order from

Carboniferous of M

ontana

201G

EO

DIV

ER

SITA

S• 2000 • 22 (2)

31 Maxilla posterior end 0 - moderate rectangular plate 1 - high rounded 2 - high triangle 3 - narrow 32 Maxilla mandibularfossa 0 - rear of mandible 1 - absent 2 - most of mandible33 Maxilla posteroventral angle 0 - not projecting 1 - projecting34 Mandible coronoid process 0 - absent 1 - present35 Maxilla-preopercular 0 - circumferential ridges 1 - vertical ridges 2 - tuberculations 3 - absent

ganoine36 Marginal teeth 0 - vertical 1 - forward 2 - posteriad 3 - absent37 Dentition 0 - marginal 1 - marginal-palatal 2 - phyllodont38 Preoperculum numbers 0 - one bone 1 - two bones39 Preoperculum shape 0 - «hatchet-shape» 1 - trapezoidal 2 - tall, narrow 3 - polygonal plate40 Preopercular canal 0 - to mid-otic region 1 - to postorbital corner 2 - to postotic region41 Preopercular canal 0 - to mandible 1 - ends blindly 2 - absent

ventrally in preoperculum42 Vertical preopercular 0 - short on preoperculum 1 - on preoperculum, long 2 - absent

pit line43 Horizontal preopercular 0 - short on preoperculum 1 - absent

pit line 44 Anteoperculum [dermohyal] 0 - absent 1 - single 2 - series 45 Suboperculum 0 - shorter than opercular 1 - taller than opercular 2 - absent as distinct bone46 Anterior branchiostegals 0 - numerous 1 - absent below mandible 47 Posterior branchiostegals 0 - uniform 1 - short behind mandible 2 - complex 3 - absent48 Extralateral gular 0 - absent 1 - present49 Gular, median 0 - present 1 - absent50 Gular, lateral 0 - like branchiostegals 1 - absent 2 - extended51 Clavicles 0 - present 1 - rudimentary 2 - absent52 Clavicle length 0 - short 1 - elongate 2 - absent53 Postspiraculars 0 - absent 1 - present54 Presupracleithrum 0 - present 1 - absent55 Sclerotics 0 - many 1 - four 2 - two56 Parasphenoid 0 - short 1 - length of braincase57 Suspensorium 0 - limited methyostyly 1 - extensive methyostyly58 Dorsal fin length 0 - short 1 - slight extension 2 - rear close to caudal 3 - merged with caudal59 Dorsal fin shape 0 - Triangular 1 - sigmoidal 2 - long, acuminate 3 - very high and long 4 - long, low (> 60 rays)

5 - short rounded60 Dorsal fin base 0 - no specializations 1 - scaled lobe 2 - guard scales61 Pectoral extended lobe 0 - absent 1 - present62 Pelvic fin 0 - normal proportions 1 - reduced or absent 2 - elongate63 Caudal axis 0 - heterocercal past last ray 1 - hemiheterocercal 64 Caudal outline 0 - cleft equilobate 1 - strongly inequilobate 2 - rounded 3 - pointed65 Predorsal scutes 0 - absent 1 - few 2 - many 3 - complete to occiput66 Anal fin base length 0 - short 1 - anal ends close to caudal 2 - merged with caudal67 Caudal fin rays 0 - not webbed 1 - webbed68 Scale row inversion 0 - progressive 1 - abrupt 2 - absent69 Fin fulcrae 0 - present 1 - absent70 Flank scales 0 - same size as others 1 - very high 2 - high, cristate71 Ventrolateral flank scales 0 - same proportions as others 1 - very narrow

Lund R.

202 GEODIVERSITAS • 2000 • 22 (2)

APPENDIX 2

Characters and character states at nodes and terminals for Fig. 17. The characters have been arrived at from Allis (1922); Arratia &Cloutier (1996); Campbell & Le Duy Phuoc (1983); DeBeer (1937); Gardiner (1984); Lehman (1956); Long (1988); Lund & Lund (1985);Lund & Poplin (1997); Lund, Poplin & McCarthy (1995); Moy-Thomas & Dyne (1938); Pearson (1982); Pehrson (1947, 1958) and Zidek(1992), as well as from unpublished Bear Gulch Limestone taxa.

Node Character State

23 Basal node3 Premaxillary condition 2 - premaxilla field fused to other

bones6 Postrostrals 1 - single median

10 Nasals 1 - single bone12 Tectals 1 - absent20 Extrascapulars 1 - 1 pair23 Posterodorsal IO shape 1 - T-shaped33 Maxilla posteroventral angle 1 - projecting35 Maxilla-preopercular ganoine 2 - tuberculations44 Anteoperculum [dermohyal] 2 - series52 Clavicle 1 - elongate55 Sclerotics 1 - 457 Suspensorium 1 - extensive methyostyly65 Predorsal scutes 1 - few

0 Cheirolepis0 Body form 0 - fusiform/elongate

12 Tectals 1 - absent14 Parietal/frontal ratio 0 - subequal25 Antorbital shape 4 - absent as distinct bone40 Preoperculum 1 - to postorbital corner50 Gular, lateral 2 - extended54 Presupracleithrum 1 - absent

3 Howqualepis2 Premaxillary position 2 - absent as distinct bone4 Rostral 3 - fused to premaxilla & postrostral

25 Antorbital shape 3 - fused to premaxilla38 Preoperculum numbers 1 - 2 bones44 Anteoperculum [dermohyal] 1 - single47 Posterior branchiostegals 1 - short behind mandible62 Pelvic fin 2 - elongate

22 4 Rostral 1 - fused to postrostral

26 Suborbital numbers 1 - 1 row27 Suborbital fit 1 - sutured28 Suborbital thickness 1 - thick65 Predorsal scutes 3 - complete to occiput

4 Mansfieldiscus3 Premaxillary condition 0 - sutured

15 Otic canal bones 1 - dermopterotic only16 Infraorbital canal 1 - to dermopterotic24 Antorbital presence 1 - differentiated27 Suborbital fit 2 - loose, gaps40 Preopercular canal 1 - to postorbital corner

2120 Extrascapulars 2 - 2 pairs29 Suborbital fit 1 - between maxilla & infraorbital68 Scale row inversion 1 - abrupt

20 Rhadinichthyidae2 Premaxillary position 2 - absent as distinct bone3 Premaxillary condition 3 - absent as distinct bone

19 Supratemporal commissure 1 - not across midline




36 Marginal teeth 1 - forward44 Anteoperculum [dermohyal] 1 - single48 Extralateral gular 1 - present60 Dorsal fin base 2 - guard scales65 Predorsal scutes 1 - few71 Ventrolateral flank scales 1 - very narrow

5 Wendyichthys9 Rostral notch 2 - keyhole-shaped

17 Dermosphenotic/ nasal contact 1 - present25 Antorbital shape 1 - inverted L27 Suborbital 2 - loose, gaps35 Maxilla-preopercular ganoine 3 - absent47 Posterior branchiostegals 1 - short behind mandible

6 Cyranorhis8 Snout shape 1 - sharp bump

68 Scale row inversion 0 - progressive19

35 Maxilla-preopercular ganoine 0 - circumferential ridges69 Fin fulcrae 1 - absent

2 Moythomasia5 Ethmoid commissure 1 - not across midline7 Mouth 1 - terminal

14 Parietal/frontal ratio 0 - subequal24 Antorbital presence 2 - fused25 Antorbital shape 3 - fused to premaxilla38 Preoperculum numbers 1 - 2 bones60 Dorsal fin base 2 - guard scales64 Caudal outline 1 - strongly inequilobate

18 33 Maxilla posteroventral angle 0 - not projecting55 Sclerotics 0 - many58 Dorsal fin length 2 - rear close to caudal59 Dorsal fin shape 2 - long, acuminate

1 Paratarrasius3 Premaxillary condition 0 - sutured4 Rostral 0 - separate bone or bones

10 Nasals 0 - serial11 Supraorbitals 2 - double series12 Tectals 0 - present13 Frontals 0 - two pairs18 Supraoccipital 1 - only on braincase26 Suborbital numbers 2 - more than 1 row42 Vertical preopercular pit line 1 - long52 Clavicle length 0 - short58 Dorsal fin length 3 - merged with caudal61 Pectoral extended lobe 1 - present62 Pelvic fin 1 - reduced or absent63 Caudal axis 2 - diphycercal64 Caudal outline 3 - pointed65 Predorsal scutes 0 - absent66 Anal fin base length 2 - merged with caudal

17 7 Mouth 1 - terminal8 Snout shape 2 - pointed snout

15 Otic canal bones 1 - dermopterotic only22 Posterodorsal IO is # 1 - 2 (<3)23 Posterodorsal IO shape 0 - simple24 Antorbital presence 0 - absent as distinct bone25 Antorbital shape 4 - absent as distinct bone

Lund R.

204 GEODIVERSITAS • 2000 • 22 (2)


27 Suborbital fit 2 - loose, gaps30 Maxilla length 1 - orbital31 Maxilla posterior end 2 - high triangle39 Preoperculum shape 2 - tall, narrow44 Anteoperculum [dermohyal] 0 - absent46 Anterior branchiostegals 1 - absent below mandible57 Suspensorium 0 - limited methyostyly66 Anal fin base length 1 - anal ends close to caudal

16 Cladistia0 Body form 2 - round4 Rostral 0 - separate bone or bones6 Postrostrals 2 - median & paired

10 Nasals 0 - serial21 Posteroventral 10 is # 1 - 131 Maxilla posterior end 3 - narrow32 Maxilla mandibular fossa 1 - absent49 Gular, median 1 - absent50 Gular, lateral 2 - extended51 Clavicles 1 - rudimentary52 Clavicle length 2 - absent53 Postspiraculars 1 - present56 Parasphenoid 1 - length of braincase61 Pectoral extended lobe 1 - present64 Caudal outline 2 - rounded67 Caudal fin rays 1 - webbed

15 Guildayichthyidae1 Highest point of body profile 1 - ridge scutes3 Premaxillary condition 1 - loose

11 Supraorbitals 2 - double series13 Frontals 0 - two pairs14 Parietal/frontal ratio 0 - subequal15 Otic canal bones 2 - 3 bones18 Supraoccipital 2 - on skull roof19 Supratemporal commissure 1 - not across midline21 Posteroventral IO is # 2 - >222 Posterodorsal IO is # 2 - >326 Suborbital numbers 2 - more than 1 row32 Maxilla mandibular fossa 2 - most of mandible38 Preoperculum numbers 1 - 2 bones45 Suboperculum 2 - absent as distinct bone51 Clavicles 2 - absent59 Dorsal fin shape 4 - long, low (>60 rays)60 Dorsal fin base 1 - scaled lobe70 Flank scales 1 - very high

8 Discoserra5 Ethmoid commissure 1 - not across midline

30 Maxilla length 2 - preorbital42 Vertical preopercular pit line 2 - absent43 Horizontal preopercular pit line 1 - absent47 Posterior branchiostegals 2 - complex62 Pelvic fin 1 - reduced or absent

7 Guildayichthys17 Dermosphenotic/nasal contact 1 - present19 Supratemporal commissure 2 - loop to skull roof20 Extrascapulars 1 - 1 pair

9 Polypterus6 Postrostrals 4 - absent as distinct bone8 Snout shape 0 - rounded

16 Infraorbital canal 2 - to fused dermopterotic-parietal




29 Suborbital fit 0 - not between maxilla & infraorbital30 Maxilla length 3 - absent34 Mandible coronoid process 1 - present35 Maxilla-preopercular ganoine 2 - tuberculations39 Preoperculum shape 3 - polygonal plate44 Anteoperculum [dermohyal] 1 - single47 Posterior branchiostegal 3 - absent58 Dorsal fin length 3 - merged with caudal59 Dorsal fin shape 5 - polypteriform63 Caudal axis 1 - hemiheterocercal66 Anal fin base length 0 - short68 Scale row inversion 2 - absent

14 “Platysomidae”0 Body form 3 - angulated2 Premaxillary position 2 - absent as distinct bone3 Premaxillary condition 3 - absent as distinct bone

29 Suborbital fit 0 - not between maxilla & infraorbital36 Marginal teeth 3 - absent37 Dentition 2 - phyllodont43 Horizontal preopercular pit line 1 - absent70 Flank scales 2 - high, cristate

12 P. gibbosus1 Highest point of body profile 1 - ridge scutes4 Rostral 4 - absent as distinct bone5 Ethmoid commissure 2 - absent

21 Posteroventral IO is 2 - >222 Posterodorsal IO is 2 - >326 Suborbital numbers 3 - large one with canal30 Maxilla length 2 - preorbital34 Mandible coronoid process 1 - present35 Maxilla-preopercular ganoine 1 - vertical ridges38 Preoperculum numbers 1 - 2 bones40 Preoperculum 2 - to postotic region41 Preopercular canal ventrally 2 - absent42 Vertical preopercular pit line 2 - absent47 Posterior branchiostegals 3 - absent53 Postspiraculars 1 - present

13 22 Posterodorsal IO is # 0 - 326 Suborbital numbers 0 - none27 Suborbital fit 0 - absent28 Suborbital thickness 0 - absent45 Suboperculum 1 - taller than opercular46 Anterior branchiostegals 0 - numerous54 Presupracleithrum 1 - absent

10 Amphicentrum1 Highest point of body profile 2 - few specialized scales

14 Parietal/frontal ratio 0 - subequal20 Extrascapulars 1 - 1 pair23 Posterodorsal IO shape 1 - T-shaped30 Maxilla length 0 - posterior to orbit35 Maxilla-preopercular ganoine 2 - tuberculations37 Dentition 1 - marginal- palatal62 Pelvic fin 1 - reduced or absent70 Flank scales 1 - very high

11 “P.” superbus5 Ethmoid commissure 2 - absent

59 Dorsal fin shape 3 - very high & long65 Predorsal scutes 0 - absent69 Fin fulcrae 0 - present

Lund R.

206 GEODIVERSITAS • 2000 • 22 (2)

APPENDIX 3

Data matrix for cladogram of Fig. 17.

Character 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Cheirolepis 0 0 0 2 ? ? 0 0 0 0 ? 0 0 1 0 0 0 0 0 0 1 0 0 1Paratarrasius 0 0 0 0 0 0 1 0 0 0 0 2 0 0 1 0 0 0 1 0 1 0 0 1Moythomasia 0 0 0 2 1 1 1 1 0 0 1 0 1 1 0 0 0 0 0 0 2 0 0 1Howqualepis 0 0 2 2 3 0 1 0 0 0 1 0 1 1 1 0 0 0 0 0 1 0 0 1Mansfieldiscus 0 0 0 0 ? 0 1 0 0 0 1 0 1 1 1 1 1 0 0 0 1 0 0 1Wendyichthys 0 0 2 3 1 0 1 0 0 2 1 0 1 1 1 0 0 1 0 1 2 0 0 1Cyranorhis 0 0 2 3 1 0 1 0 1 1 1 0 1 1 1 0 0 0 0 1 2 0 0 1Guildayichthys 2 1 0 1 0 ? 2 1 2 0 0 2 1 0 0 2 0 1 2 2 1 2 2 0Discoserra 2 1 0 1 0 1 2 1 2 0 0 2 1 0 0 2 0 0 2 1 2 2 2 0Polypterus 0 0 0 2 0 0 4 1 0 0 0 0 1 1 1 1 2 0 0 0 2 1 1 0Amphicentrum 3 2 2 3 1 0 1 1 2 0 1 0 1 1 0 1 1 0 0 0 1 0 0 1P. superbus 3 0 2 3 ? 2 1 1 2 0 1 0 1 1 1 1 1 0 0 0 2 0 0 0P. gibbosus 3 1 2 3 4 2 1 1 2 0 1 0 1 1 1 1 1 0 0 0 2 2 2 0

Character 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47

Cheirolepis 0 4 0 0 0 0 0 0 0 0 0 ? 0 0 0 0 1 0 0 0 2 0 0 0Paratarrasius 1 0 2 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 1 0 2 0 0 0Moythomasia 2 3 1 1 1 1 0 0 0 1 0 0 0 0 1 0 0 0 ? 0 2 0 0 0Howqualepis 0 3 0 0 0 0 0 0 0 1 0 2 0 0 1 0 0 0 0 0 1 0 0 1Mansfieldiscus 1 0 1 2 1 0 0 0 0 1 0 ? 0 0 0 0 1 0 ? ? 2 0 0 0Wendyichthys 1 1 1 2 2 1 0 0 0 1 0 3 1 0 0 0 0 0 0 0 1 0 0 1Cyranorhis 1 0 1 1 1 1 0 0 0 1 0 2 1 0 0 0 0 0 0 0 1 0 0 0Guildayichthys 0 4 2 2 1 1 1 3 2 0 0 0 0 0 1 2 0 0 0 0 0 2 1 0Discoserra 0 4 2 2 1 1 2 3 2 0 0 3 0 0 1 2 0 0 2 1 0 2 1 2Polypterus 0 4 1 2 1 0 3 3 1 0 1 2 0 0 0 3 0 0 0 0 1 0 1 3Amphicentrum 0 4 0 0 0 0 0 2 ? 0 ? 2 3 1 0 2 0 0 ? ? 0 1 0 0P. superbus 0 4 0 0 0 0 1 2 ? 0 0 0 ? 2 0 2 0 0 ? 1 0 1 0 0P. gibbosus 0 4 3 2 1 0 2 2 0 0 1 1 3 2 1 2 2 2 2 1 0 0 1 3

Character 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71

Cheirolepis 0 0 2 0 ? 0 1 1 ? 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0Paratarrasius 0 0 0 0 0 0 0 0 0 1 3 2 0 1 1 2 3 0 2 0 1 1 0 0Moythomasia 0 0 0 0 1 0 0 1 0 ? 0 0 2 0 0 0 1 3 0 0 1 1 0 0Howqualepis 0 0 0 0 1 0 0 1 0 1 0 0 0 0 2 0 0 1 0 0 0 0 0 0Mansfieldiscus 0 0 0 0 ? 0 0 1 ? ? 0 0 0 0 0 0 0 3 0 0 0 0 0 0Wendyichthys 1 0 0 0 1 0 0 1 0 1 0 0 2 0 0 0 0 1 0 0 1 0 0 1Cyranorhis 1 0 0 0 1 0 0 1 0 1 0 0 2 0 0 0 0 1 0 0 0 0 0 1Guildayichthys 0 1 2 2 2 1 0 0 1 0 2 4 1 1 0 0 2 3 1 1 1 1 1 0Discoserra 0 1 2 2 2 1 0 0 1 0 2 4 1 1 1 0 2 3 1 1 1 1 1 0Polypterus 0 1 2 1 2 1 ? ? 1 0 3 5 0 1 0 1 2 3 0 1 2 1 0 0Amphicentrum 0 ? 0 0 1 0 1 ? ? ? 2 2 0 0 1 0 0 3 1 0 1 1 1 0P. superbus 0 0 0 0 ? 0 1 ? 0 ? 2 3 0 0 0 0 0 0 1 0 1 0 2 0P. gibbosus 0 ? ? ? ? 1 0 ? ? ? 2 2 0 0 0 0 0 3 1 0 ? ? 2 0

the new actinopterygian order guildayichthyiformes from...

Documents