Conodont affinity and chordate phylogeny

61 Pages·2000·2.18 MB·English

Biol. Rev. (2000),75, pp.191–251 Printed in the United Kingdom #Cambridge Philosophical Society 191
Conodont affinity and chordate phylogeny
" #
PHILIP C. J. DONOGHUE , PETER L. FOREY
$
and RICHARD J. ALDRIDGE
"School of Earth Sciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK (p.c.j.donoghue!bham.ac.uk)
#The Natural History Museum, South Kensington, London SW7 5BD, UK (plf!nhm.ac.uk)
$Department of Geology, University of Leicester, University Road, Leicester LE1 7RH, UK !le.ac.uk)
(Received 4 June 1999; revised 29 November 1999; accepted 29 November 1999)
ABSTRACT
CurrentinformationontheconodontsClydagnathuswindsorensis(Globensky)andPromissumpulchrumKova!cs–
Endro$dy,togetherwiththelatestinterpretationsofconodonthardtissues,arereviewedanditisconcluded
that sufficient evidence exists to justify interpretation of the conodonts on a chordate model. A new
phylogeneticanalysisisundertaken,consistingof17chordatetaxaand103morphological,physiologicaland
biochemical characters; conodonts are included as a primary taxon. Various experiments with character
coding, taxon deletion and the use of constraint trees are carried out. We conclude that conodonts are
cladistically more derived than either hagfishes or lampreys because they possess a mineralised dermal
skeleton and that they are the most plesiomorphic member of the total group Gnathostomata. We discuss
theevolutionofthenervousandsensorysystemsandtheskeletoninthecontextofouroptimalphylogenetic
tree. There appears to be no simple evolution of free to canal-enclosed neuromasts; organised neuromasts
withincanalsappeartohavearisenatleastthreetimesfromfreeneuromastsorneuromastsarrangedwithin
grooves.Themineralisedvertebrateskeletonfirstappearedasodontodesofdentineordentineplusenamel
intheparaconodont}euconodontfeedingapparatus.Boneappearedlater,co-ordinatewiththedevelopment
ofadermalskeleton,anditappears tohavebeen primitivelyacellular.Atubular dentineismoreprimitive
than tubular dentine. However, the subsequent distribution of the different types of dentine (e.g.
mesodentine, orthodentine), suggests that these tissue types are homoplastic. The topology of relationships
and known stratigraphic ranges of taxa in our phylogeny predict the existence of myxinoids and
petromyzontids in the Cambrian.
Key words: Agnatha, Gnathostomata, cladistics, conodont, chordate, craniate, evolution, phylogeny,
vertebrate, skeleton.
CONTENTS
I. Introduction............................................................................................................................ 192
(1) Poly-, para- and monophyly of the Conodonta................................................................ 194
II. Soft tissue anatomy.................................................................................................................. 195
(1) The head........................................................................................................................... 197
(2) The trunk.......................................................................................................................... 199
(3) The tail............................................................................................................................. 199
III. Hard tissues............................................................................................................................. 199
(1) The histological debate..................................................................................................... 201
(2) Tissue types....................................................................................................................... 201
(a) Lamellar crown tissue................................................................................................. 201
(b) White matter.............................................................................................................. 202
(c) Basal tissue.................................................................................................................. 202
(3) Relative growth of the tissues........................................................................................... 202
192 P. C. J. Donoghue, P. L. Forey and R. J. Aldridge
(4) Hard tissue homologies..................................................................................................... 202
(a) Lamellar crown tissue................................................................................................. 202
(b) White matter.............................................................................................................. 203
(c) Basal tissue.................................................................................................................. 203
(5) Conodont element growth................................................................................................. 203
(6) Histochemical studies........................................................................................................ 204
IV. Phylogenetic analysis............................................................................................................... 204
(1) Taxon sampling................................................................................................................ 204
(2) Character coding............................................................................................................... 207
(3) Character matrix............................................................................................................... 208
(a) Brain, sensory and nervous systems............................................................................ 208
(b) Mouth and branchial system...................................................................................... 211
(c) Circulatory system...................................................................................................... 211
(d) Fins and fin folds........................................................................................................ 212
(e) Skeletal....................................................................................................................... 213
(f) Physiological............................................................................................................... 215
(g) Miscellaneous.............................................................................................................. 215
(4) Results............................................................................................................................... 215
(a) Experimental analysis of the data set......................................................................... 219
(b) Cyclostome monophyly............................................................................................... 220
(c) Alternative hypotheses of chordate relationships........................................................ 220
(d) The effects of alternative interpretations of conodont anatomy and histology........... 221
(e) Testing alternative hypotheses of conodont affinity.................................................... 223
(f) Summary of conclusions drawn from experimental analyses of our data set.............. 225
(5) Character changes............................................................................................................. 225
(a) General....................................................................................................................... 225
(b) Nervous and sensory systems...................................................................................... 225
(c) Skeleton...................................................................................................................... 233
V. Conodont affinity..................................................................................................................... 236
VI. Directions for future research.................................................................................................. 238
(1) General.............................................................................................................................. 238
(2) The origin of the Conodonta............................................................................................ 238
(3) Euconodont phylogeny...................................................................................................... 239
(4) Histology........................................................................................................................... 239
(5) Microevolution.................................................................................................................. 240
(6) Conodont element function............................................................................................... 240
VII. Revised classification............................................................................................................... 240
VIII. Conclusions.............................................................................................................................. 240
IX. Acknowledgements.................................................................................................................. 241
X. References................................................................................................................................ 241
XI. Appendix................................................................................................................................. 249
(1) Character diagnostics........................................................................................................ 249
I. INTRODUCTION Klussendorf,1985a,b;Aldridgeetal.,1993;Gabbott,
Aldridge & Theron, 1995) has resulted in a
‘The problem of the zoological affinities of this group revolutioninourunderstanding; wenowknow that
remains…one of the most fascinating and perplexing conodonts are a group of eel-shaped animals with
problems of palaeozoology’ (Rhodes, 1954, p. 419). the phosphatic elements constituting a complex
feeding apparatus. Although the systematic position
When Frank Rhodes wrote his article on conodont of conodonts is still contentious, the diversity of
relationshipsforBiologicalReviewsin1954,conodonts opinionhasnarrowedgreatly.Justayearbeforethe
were known exclusively in the form of phosphatic first conodont fossil with preserved soft tissues was
tooth-like microfossils of unknown or uncertain found, Mu$ller (1981) compiled a list of groups to
affinity. The perception of conodonts is now very which conodonts had been attributed; his list
different.Thediscoveryofsofttissueremains(Briggs, includes at least three kingdoms and almost every
Clarkson & Aldridge, 1983; Mikulic, Briggs & major animal phylum. The debate is now much
Conodont affinity and chordate phylogeny 193
more tightly constrained (see Aldridge, 1987; a bilaterally disposed grasping array in what is
Aldridgeetal.,1993)andalmostallauthoritiesagree interpreted as the cephalic region. A longitudinal
that conodonts are chordates, although a few gut, possible ovaries and a laterally disposed ray-
workers still advocate a chaetognath affinity (e.g. supportedcaudal fin arealso preserved.The overall
Kasatkina & Buryi, 1996a,b, 1997). Among those anatomy of T. rokycanensis is distinctly chaetognath-
authorswhoacceptachordateassignment,however, like and quite dissimilar to preserved remains of
there is still controversy regarding the precise place conodonts.
of the conodonts within the chordate clade. The interpretation of early Cambrian conodonti-
To conduct a cladistic analysis involving both form elements known as protoconodonts as the
chordatesandchaetognathsiscurrentlynotfeasible, grasping spines of chaetognaths (Szaniawski, 1982)
as the affinity of chaetognaths themselves is highly provides another possible line of evidence, if the
contentious.Someauthorssuggestthattheyarebest evolutionary sequence from protoconodonts,
placed as a sister-group to the vertebrates through paraconodonts to euconodonts (the true
(Christoffersen & Arau!jo-de-Almeida, 1994), conodonts) proposed by Bengtson (1976; Fig. 1)
whereas others doubt even their deuterostome weresustained.Thislinkagewasfurtheremphasised
affinity (Halanych, 1996; Nielsen, 1995; Nielsen, byBengtson(1983a,b) andSzaniawski(1987), who
Scharff & Eibye-Jacobsen, 1996). A relationship noted that the hypothesis that chaetognaths and
between conodonts and chaetognaths was first (eu)conodonts shared a common ancestor did not
suggested by Rietschel (1973) and was resurrected necessarily conflict with conodonts being closely
by Briggs et al. (1983) as one of the possible related to the chordates. Andres (1988) has also
interpretations consistent with the features of the suggestedareconciliationbetweenthechordateand
first known conodont with preserved soft tissues chaetognathhypothesesofaffinity,proposingchaeto-
(IGSE [British Geological Survey, Keyworth] gnaths as ancestors to euconodonts, which are in
13822). This interpretation was more strongly turn ancestral to the vertebrates. This suggestion is
advocatedbyBengtson(1983b),andissupportedby consistent with a close relationship between chaeto-
Kasatkina & Buryi (1996a,b, 1997). However, the gnaths and chordates (Christoffersen & Arau!jo-de-
resemblances between conodont and chaetognath Almeida, 1994; Cavalier-Smith, 1998; Hall, 1998),
anatomy are vague; both are small and bilaterally butatoddswithrecentmolecularandmorphological
symmetrical, and both bear ray-supported fins, but work which supports the view that chaetognath
the caudal fin of conodonts is dorso-ventrally rather affinity lies amongst the protostomes (Telford &
than laterally disposed as in chaetognaths (contra Holland, 1993; Philippe, Chenuil & Adoutte, 1994;
Kasatkina & Buryi, 1996a). Furthermore, the V- Wada & Satoh, 1994; Nielsen, 1995; Nielsen et al.,
shapedstructuresapparentinthetrunkofconodonts 1996; Halanych, 1996; Zrzavy! et al., 1998; current
havenoparallelinthemusculatureofchaetognaths, evidence suggests that chaetognaths are closely
althoughthereis some resemblance totheovaries of related to the nematodes and gnathostomulids,
some taxa [e.g. Pterosagitta draco (Krohn, 1853)]. possibly forming a monophyletic group, see
Thepossibilityremainsthat thetaphonomic state Littlewood et al., 1998). Until the affinities of
of specimens preserving conodont anatomy masks chaetognaths are resolved, the relative position of
similarities between the two groups, but, as no conodonts and chaetognaths remains untestable,
taphonomic studies of chaetognaths have been andanapparent morphologicalgulfbetweenproto-
undertaken, this remains difficult to assess. Com- conodonts and paraconodonts limits the applica-
parisonoffossilconodontswithfossilchaetognathsis bility of the evolutionary model forwarded by
limitedbytherarityoffossilchaetognathspecimens. Bengtson (1976; Fig. 1).
Amiskwia sagittiformis Walcott from the Middle Aldridge & Purnell (1996), in a review of the
Cambrian Burgess Shale was originally described as debate, considered six competing hypotheses put
a chaetognath, but has been reinterpreted as a forward in the recent literature for the systematic
possible nemertine (Conway Morris, 1977). position of conodonts amongst the living chordates
Paucijaculum samamithion Schram from the mid- (Dzik,1995;Kemp&Nicoll,1995a,1996;Peterson,
Pennsylvanian Francis Creek Shale of Illinois has 1994; Aldridge et al., 1986, 1993; Gabbott et al.,
been similarly reassessed (Richardson 1982). More 1995; Janvier, 1995). Although most of these
recently, Kraft & Mergl (1989) have described a hypotheses have been couched in cladistic termin-
moreconvincingcandidate,Titerinarokycanensisfrom ology or expressed in the form of cladograms, not
the Lower Ordovician of Bohemia, which preserves one is the result of a numerical cladistic analysis.
194 P. C. J. Donoghue, P. L. Forey and R. J. Aldridge
A B C
Fig. 1. Proto- (A), para- (B) and eucondont (C) grades of organisation (after Bengtson, 1976, and Szaniawski &
Bengtson, 1993). (A) Different shadings distinguish three layers which have no homologues in B and}or C. (B, C)
Dark shading represents basal tissue of euconodonts and its putative homologue in paraconodonts; light shading
represents lamellar crown tissue of euconodonts.
Instead, each hypothesis represents the result of divided into three main groups on histological
placingconodontsinapre-existingcladogramusing criteria by Mu$ller & Nogami (1971, 1972) and
pre-established synapomorphies, or classifies them Bengtson(1976;Fig.1).Thestratigraphicallyoldest
according to unsubstantiated a priori assumptions of group, the protoconodonts, are represented by
character polarity in chordate phylogeny, or places dominantly organic spine-like fossils which are
themwithinaphylogenybasedonaconsiderationof multilayered and grew by adding new layers to the
limitedpartsoftheanatomy.Thisisunfortunate,as concave side of the elements only (Fig. 1A).
conodontsclearlyhaveanimportantcontributionto Protoconodont elements bear a strong histological
make to our understanding of chordate evolution. resemblance to chaetognath grasping spines
Their significance centres around the feeding (Szaniawski, 1982) and all available evidence indi-
elements, which represent the only biomineralised cates a probable link between these two groups
component of the conodont skeleton. Depending on (Szaniawski, 1983, 1987). Paraconodont elements
whether conodonts are acraniate or craniate are also relatively poorly mineralised; in growth,
chordates, their elements are either an interesting each successive phosphatic layer accreted onto the
but esoteric feature (in the former case), or an margins and base, but not the tip of its predecessor
importantcluetothedevelopmentofthemineralised (Szaniawski & Bengtson, 1993; Fig. 1B). These
skeleton in craniate phylogeny (in the latter case). elements bear a strong resemblance to the basal
The mainaim of this contribution is to determine body of euconodont elements (Lindstro$m, 1964;
the systematic position of the Conodonta through Bengtson, 1976). Euconodont elements are com-
formal cladistic analysis. To this end, we briefly posed of two structures: the crown, which is very
review interpretations of the anatomy and histology heavily mineralised and relatively coarsely crys-
of conodonts in order to explain our coding of talline,andthebasalbodywhichisfinelycrystalline
characters in the data matrix on which our analysis and containsmoreorganic material (Fig. 1C). Both
isbased.First,however,itispertinenttoconsiderthe components grew by external apposition of new
monophyly of the Conodonta and to establish layers of mineral on the outer surface, which were
precisely what constitutes a conodont. either added synchronously (Mu$ller & Nogami,
1971)oratleastinstep(Lindstro$m&Ziegler,1971).
The lack of demonstrated links between proto-
conodontsandparaconodontsandtheevidencethat
(1) Poly-, para- and monophyly of the
theformerarerelatedtochaetognathsmeansthatit
Conodonta
is likely that protoconodonts belong to a clade of
Microscopic phosphatic cones, or ‘conodontiform’ animalsdistinctfromeuconodonts(seealsoPeterson,
fossils,firstoccurintheuppermostPrecambrianand 1994). An evolutionary link between paraconodont
become diverse in Cambrian strata. They were and euconodont elements has been more firmly
Conodont affinity and chordate phylogeny 195
a
d a
Paraconodonti Proconodontid Teridontus unnamed group Panderodontida unnamed group Prioniodinida Ozarkodinida
Pr
ot
o a
p d
anderodonti Prioniodonti
d
a
? ? ?
Fig. 2. Attempted reconstruction of the relationships of the major groups of euconodont-grade taxa, based on the
suprageneric scheme and proposed relationships of Sweet (1988). This is not a cladogram but our interpretation of
the views expressed by Sweet (1988) in the form of a phylogenetic tree. The three orders of complex conodonts are
generally given equal taxonomic rank. However, it is implicitly accepted that the Ozarkodinida and Prioniodinida
are derived from taxa classified within the Prioniodontida. Therefore, we have chosen to express this relationship
explicitly in the figure and in the text.
established, primarily on the similarity in growth unclear. The Teridontus lineage includes the over-
pattern. Szaniawski & Bengtson (1993) have whelming majority of conodonts and, although the
attemptedtodemonstratesometransitionalformsin maximumdiversityofthiscladewasalsoattainedin
whichlayersofcrowntissuewereaddedtothebasal theOrdovician,itisknowntohavesurvivedthrough
bodyonlyinlateontogeny.Morerecently,Mu$ller& to the end of the Triassic Period. The three taxa
Hinz-Schallreuter (1998) have documented the known with preserved soft tissue features all belong
histology of a number of Cambrian paraconodonts to the Teridontus lineage, as do the vast majority of
which possess microstructural features comparable taxathathavebeenstudiedhistologically.Giventhe
with the basal tissue of euconodonts. The link uncertain and probable independent origins of
between paraconodonts and euconodonts points to Proconodontus and Teridontus, we restrict our phylo-
an origin for the whole group in the Early or Mid geneticanalysesofeuconodontstothemonophyletic
Cambrian and renders the aptly named Para- lineage first represented by Teridontus.
conodontida a paraphyletic assemblage unless it is
expanded to include the Euconodonta. However,
monophyly of the Euconodonta is far from certain
and not all taxa currently regarded as euconodonts II. SOFT TISSUE ANATOMY
necessarilyfallwithinaninclusiveParaconodontida.
The early evolutionary history of the Euco- The soft tissue anatomy of conodonts is known from
nodonta, or true conodonts, has been most clearly at least three euconodont taxa: Clydagnathus
documented by Miller (1969, 1980, 1984) who windsorensis (Globensky) from the Lower Carbon-
distinguished two distinct lineages that are first iferousGrantonShrimpBedofEdinburgh,Scotland
represented in the stratigraphic record by Procono- (Aldridgeetal.,1993);Panderodusunicostatus(Branson
dontus and Teridontus (Fig. 2). Both lineages are of & Mehl) from the Lower Silurian of Waukesha,
uncertain ancestry and both made their first ap- Wisconsin (Mikulic et al., 1985a,b; Smith, Briggs &
pearance in the late Cambrian. The Proconodontus Aldridge, 1987), and Promissum pulchrum Kova!cs-
lineage is known to have flourished in the Early Endro$dyfromtheUpperOrdovicianSoomShaleof
Ordovician but its subsequent evolution remains South Africa(Aldridge&Theron, 1993; Gabbottet
196 P. C. J. Donoghue, P. L. Forey and R. J. Aldridge
Fig. 3. For legend see opposite.
Conodont affinity and chordate phylogeny 197
al.,1995).Theremaybeadditionaltaxarepresented to distinguish between otic and optic capsules (see
in the specimens from the Granton Shrimp Bed. Donoghue, Purnell & Aldridge, 1998), and in
specimenIGSE13821}2fromGranton,thelobeslie
anterior to a smaller pair of round structures which
(1) The head
have themselves been interpreted as otic capsules
The most frequently preserved soft tissues are a (Aldridge et al., 1993; Aldridge & Donoghue, 1998;
single pair of lobate structures, which occur com- Donoghue et al., 1998). There is, furthermore, no
monly in association with natural assemblages of evidence that otic capsules are more likely to be
conodont elements in the Soom Shale and are also preserved than optic capsules; indeed, remnants of
evident in two of the specimens from Granton. The eyeshavebeenreportedinallfossilrepresentativesof
lobes occur as carbonised remains that are round to soft-bodied jawless vertebrates. In contrast, the
oval in outline with a thickened or darker rim, and remains of unmineralised otic capsules are rare and
have been reconstructed to three dimensions as a normally limited to moldic preservation (e.g.
pair of outwardly expanding cups (Aldridge et al., Myxinikela; Bardack, 1991).
1993;Aldridge&Theron,1993).FromtheGranton The presence of extrinsic eye musculature in
specimens it is evident that these cups were located specimenGSSAC721issupportedbyevidencefrom
slightly in front of and above the feeding apparatus additional, recently discovered specimens of
(Fig. 3A–C). Aldridge & Theron (1993) drew Promissum pulchrum that also display paired lobes
comparisonwithsimilarstructuresintheheadofthe composed of fibrous white clay minerals. Although
fossil jawless vertebrate Jamoytius kerwoodii White, no other soft tissue remains are preserved in these
where they have been interpreted as eye capsules specimens, the lobate patches occur in consistent
(Ritchie 1968). Similar structures, interpreted as topological association with natural assemblages of
tracesofretinalpigments,havealsobeenrecordedin elements, indicative of an original position above
a number of entirely soft-bodied fossil jawless and in front of the feeding apparatus. In a few
vertebrates including Myxinikela (Bardack 1991, specimens, the patches are preserved only on the
1998), Mayomyzon (Bardack & Zangerl, 1968, 1971) part, but are matched by lobate traces of organic
and Hardistiella (Janvier & Lund, 1983; Lund & film on the counterpart that are identical with the
Janvier, 1986). The most completely preserved structures interpreted as eye capsules on other
specimen of Promissum pulchrum (GSSA [Geological material. In extant lampreys, the extrinsic eye
Survey of South Africa, Pretoria] C721) includes muscles also match the eye capsules in their
oval patches of white tissue in a similar position distribution (see e.g. Fig. 3.6.I in Janvier, 1996a).
relativetothefeedingapparatustothatoccupiedby Scanningelectronmicrographyofthewhitepatches
the carbonised lobes in other specimens. This white reveals a fibrous microstructure similar to the
tissue has a fibrous texture comparable with preserved trunk musculature of specimen GSSA
preserved muscle tissue in the trunk of the same C721 (Fig. 4F–H).
specimen, and was interpreted as extrinsic eye Inourcodingsforconodontswehaveacceptedthe
musculature by Gabbott et al. (1995). growing evidence for the presence of extrinsic eye
The interpretation of these structures as eye musculature and we have taken the presence of
capsules and eye musculature has been criticised by paired sensory organs (eyes and, perhaps, otic
Pridmore, Barwick & Nicoll (1997), who argued capsules) to indicate the presence of a brain.
that the paired lobes more probably represent otic SpecimenIGSE13821}2alsoshowsfainttracesof
capsules. Pridmore et al. (1997) based their argu- approximately five paired rectangular box-like
mentsonthesize,shapeandpositionofthelobesand structures posterior to the paired capsules (Fig. 3A
on an assertion that otic capsules are more likely to in Briggs et al., 1983; Fig. 3C). Each of these
be preserved in the fossil record than optic capsules. structures is oriented with its long axis transverse to
However, size and shape are inadequate characters the body axis. These compare well in shape and
Fig. 3. (A–E) Clydagnathus windsorensis (Globensky). (A, B) Part and counterpart of whole animal (respectively IGSE
13821 & 13822); frame widths 16 mm and 18 mm respectively. (C) Head region of part (IGSE 13821); frame width
25mm. (D) Bilobed tail of IGSE 13821; frame width 7mm. (E) Representative portion of the trunk (RMS GY
1992.41.1); frame width 10.5mm. (A, C, D from Briggs et al. 1983; B, E from Aldridge et al. 1993). A, C, D are
reproduced from Briggs et al. (1983) with the permission of the Lethaia Foundation.
198 P. C. J. Donoghue, P. L. Forey and R. J. Aldridge
Fig. 4. (A–H) Promissum pulchrum Kova’ cs-Endro$dy. (A, B) Part and counterpart of only known specimen preserving
trunk musculature (GSSA C721b, a respectively); frame widths 106 mm and 112 mm respectively. (C) Head region
including feeding apparatus and preserved extrinsic eye musculature (GSSA C721a); frame width 16mm. (D, E)
Scanning electron micrograph of trunk musculature (GSSA C721a); D frame width 74lm; E frame width 22lm.
(F–H) Extrinsic eye musculature under progressively higher magnification (GSSA C1320); F frame width 2900lm;
G frame width 930lm; H frame width 360lm. B, D, E reproduced from Gabbott et al. (1995), with the permission
of Macmillan Magazines Ltd.
Conodont affinity and chordate phylogeny 199
position to gill pouches in the fossil lamprey themostextensivedevelopmentofthisfinwasonthe
Mayomyzon (Aldridge & Donoghue, 1998), and in dorsal or the ventral margin of the animal, but our
our coding we follow Aldridge et al. (1993) in their re-examination of IGSE 13822 indicates that the
tentative interpretation of these structures as gill bilobed portion occurs on the dorsal margin (Fig.
pouches. 3D). The second specimen (RMS [Royal Museum
ofScotland,Edinburgh]GY1992.41.3)preservesan
(2) The trunk unequallydevelopedray-supportedfinthatislonger
on the dorsal margin. The shape of the bilobed fin
Trunk muscle tissues occur as distinct anteriorly
comparestothatoflampreys,whereitistraditionally
directed chevrons in the Granton specimens; their
interpreted as hypocercal. Janvier (1998) has re-
discretenatureispresumablytheresultofpostmortem
centlycomparedthetailofRMSGY1992.41.3with
decay and shrinkage, a process recognised in
that of Myxine glutinosa, which he interprets as
decomposition experiments on Branchiostoma (Briggs
cryptically hypocercal. Here, we regard the tail of
& Kear, 1994). The architecture of the muscle
conodontstobehypocercalbecausethebilobedfinis
blocks appears to be V-shaped, although the possi-
restrictedtothedorsalmarginofthenotochord,but
bilityremainsthatthemyomereswereoriginallyW-
wedonotacceptJanvier’s(1998)argumentasthere
shapedwithonlythecentralVpreserved(Donoghue
is no evidence of a ventral bend of the conodont
et al., 1998). In modern ammocoetes the myomeres
notochord at the caudal termination.
are W-shaped, but the dorsal limbs of the muscle
blocksareextremelyshort(P.C.J.Donoghue,pers.
III. HARD TISSUES
obs.).Althoughthereissomeevidenceforindividual
fibre preservation in one of the Granton specimens
The mineralised skeleton of conodonts is limited to
(RMS GY 1992.41.3; Aldridge et al., 1993) muscle
the phosphatic elements that are generally inter-
ultrastructureisbestpreservedinP.pulchrum(GSSA
preted to have comprised a feeding apparatus. The
C721), in which each myomere is demonstrably
composition and architecture of this apparatus is
composed of fibril bundles (Fig. 4E), together with
only fully understood for a few taxa, almost all of
possible sarcolemmic membranes and collagenous
which are highly derived, within the context of
connective tissues (Gabbott et al., 1995). This
conodont phylogeny as it is understood currently
specimen also includes a black organic patch within
(Fig. 2). We know very little of the apparatuses of
themid-trunkregionthatisprobablytheremainsof
the earliest conodonts, with the simplest fully
avisceralorgan.Pairedaxiallinesinterpretedasthe
reconstructed apparatus being that of Panderodus
margins of a notochord (Aldridge et al., 1993), run
which has cone-like elements of similar morphology
the length of the trunk in all Granton specimens.
to those of primitive conodonts. The apparatus of
This interpretation has been corroborated by ex-
Panderodus was bilaterally arranged with eight pairs
perimental decay of Branchiostoma, in which the
of opposed elements and one symmetrical
notochord decomposes to leave the notochordal
‘symphysial’ element lying on the axis of symmetry
sheath, which then collapses to a pair of linear
(Smith et al., 1987; Sansom, Armstrong & Smith,
thickened margins (Briggs & Kear, 1994). Remains
1994). These elements are considered to have
of a notochord have not been found in association
performed a grasping function (Smith et al., 1987;
with Promissum pulchrum, but its position has been
Sansom et al., 1994a), although this hypothesis
consideredtoberepresentedbya2mmgapbetween
remains to be tested.
the dorsal and ventral limbs of the myomeres
Morederivedconodontgroupstypicallypossessed
(Gabbottetal.,1995).Apossiblenervecordhasbeen
elements of more complex morphology and their
identifiedstoppingshortofthepairedlobesintwoof
apparatuses were composed from morphologically
the Granton specimens (Aldridge et al., 1993).
distinct groups of elements. Almost all ‘complex
conodonts’ belong to the Prioniodontida and many
(3) The tail
of these fall within the Ozarkodinida, the most
Thetailispreservedinonlytwoofthe10specimens derived of all conodont groups (Fig. 2). The
from Granton, one of which (IGSE 13822) displays architecture of non-ozarkodinid prioniodontids has
an asymmetrical bilobed ray-supported fin. The fin only been fully resolved for Promissum pulchrum,
rays are closely set, without apparent branching, whichpossessednineteenelements,dividedintofour
andthereisnoevidenceforsupportingmusculature. pairsofrobust‘P’elements whichlay behindapair
Briggsetal.(1983)wereunabletodeterminewhether of‘M’elements,andaboveanarrayof‘S’elements
200 P. C. J. Donoghue, P. L. Forey and R. J. Aldridge
Fig. 5. For legend see opposite.