Evolutionary development of the neurocranium in Dissorophoidea (Tetrapoda: Temnospondyli), an integrative approach

Evolutionary development of the neurocranium in Dissorophoidea

(Tetrapoda: Temnospondyli), an integrative approach

Hillary C. Maddin,a,� Robert R. Reisz,b and Jason S. Andersonc

aDepartment of Biological Sciences, University of Calgary, 2500 University Drive, Calgary, AB, Canada T2N 1N4bDepartment of Biology, University of Toronto at Mississauga, 3359 Mississauga Road, Mississauga, ON, Canada L5L 1C9cDepartment of Comparative Biology and Experimental Medicine, University of Calgary, 3330 Hospital Drive, Calgary, AB,

Canada T2N 1N4�Author for correspondence (email: [email protected])

SUMMARY Ontogenetic data can play a prominent role inaddressing questions in tetrapod evolution, but such evidencefrom the fossil record is often incompletely consideredbecause it is limited to initiation of ossification, or allometricchanges with increasing size. In the present study, specimensof a new species of an archaic amphibian (280 Myr old),Acheloma n. sp., a member of the temnospondyl superfamilyDissorophoidea and the sister group to Amphibamidae, whichis thought to include at least two of our modern amphibianclades, anurans and caudatans (Batrachia), provides us withnew developmental data. We identify five ontogenetic events,enabling us to construct a partial ontogenetic trajectory(integration of developmental and transformation seq-uence data) related to the relative timing of completion ofneurocranial structures. Comparison of the adult amphiba-mid morphology with this partial ontogeny identifies a

heterochronic event that occurred within the neurocraniumat some point in time between the two taxa, which is consistentwith the predictions of miniaturization in amphibamids, pro-viding the first insights into the influence of miniaturization onthe neurocranium in a fossil tetrapod group. This study refineshypotheses of large-scale evolutionary trends withinDissorophoidea that may have facilitated the radiation ofamphibamids and, projected forward, the origin of the gen-eralized batrachian skull. Most importantly, this studyhighlights the importance of integrating developmental andtransformation sequence data, instead of onset of ossificationalone, into investigations of major events in tetrapod evolutionusing evidence provided by the fossil record, and highlightsthe value of even highly incomplete growth series comprisedof relatively late-stage individuals.

INTRODUCTION

Ontogenetic data have been playing an increasingly crucial

role in exploring limited fossil data for new characters to help

address phylogenetic problems. Under ideal conditions, initial

centers of ossification, and even preossification anlagen, are

determinable (Bystrow and Efremov 1940; Schoch and Car-

roll 2003; Witzmann 2006; Witzmann and Schoch 2006;

Frobisch et al. 2007). However, onset of ossification is but one

of a number of potential ontogenetic data that can be altered

by natural selection to lead to morphological change, and

thus generate ontogeny-based cladistic characters. In seminal

works on ontogeny and phylogeny, Gould (1977) and

Alberch et al. (1979) cited onset, offset (together deriving du-

ration), and rate of development as important developmental

factors that could be shifted in timing in ontogeny and

thereby produce evolutionary change (Fig. 1). By restricting

consideration of ontogeny to onset of ossification, these other

valuable data are overlooked.

The analysis of ontogeny requires the use of precise ter-

minology to describe different observable developmental phe-

nomena. The term Developmental Sequence (DS) is often

used interchangeably with the term Ontogenetic Sequence

(OS). A DS, as envisioned by Alberch et al. (1979) can de-

scribe both the sequence of discrete developmental events/

processes, and the sequence of stages within a single event/

process. Schulmeister and Wheeler (2004) made a valuable

distinction between these two types of sequences, referring the

former to DS, and the latter to OS or Transformation Se-

quence (TS). The term TS is preferred here to avoid confusion

with a more general application of the term OS and the very

different term, Ontogenetic Trajectory (OT), which Alberch

et al. (1979) used to describe the growth of the entire individual.

Typically, these types of data are depicted as staging tables,

which are descriptive tools that document the integration of

both the sequence of event/process onset (DS) and the TS of

each event/process. Staging tables are especially valuable

because they incorporate developmental data often excluded

EVOLUTION & DEVELOPMENT 12:4, 393 –403 (2010)

DOI: 10.1111/j.1525-142X.2010.00426.x

& 2010 Wiley Periodicals, Inc. 393

(e.g., duration, rate, and offset), and can thereby assist in the

identification of heterochronic events that extend beyond data

such as onset of ossification.

An understanding of the origin and evolution of modern

amphibian morphology has been a challenging endeavor due

to the disparate morphology between extant and fossil forms,

and the scarcity of fossil intermediates. Current research con-

tinues to implicate the importance of evolutionary change

related to heterochrony, such as miniaturization, in the ques-

tion of the origin and evolution of some or all of the modern

clades of amphibians (Schoch 1992, 2002; Schoch and Rub-

idge 2005; Frobisch and Reisz 2008; Frobisch and Schoch

2009). This has, in part, contributed to a general consensus

among the majority of tetrapod systematists that at least two

lineages of modern amphibians, anurans and caudatans, were

derived from within dissorophoid temnospondyls, a diverse

clade of animals positioned within one of the most stemward

tetrapod radiations (Fig. 2; i.e., Ruta et al. 2003; Schoch and

Milner 2004; Carroll 2007; Ruta and Coates 2007; Anderson

et al. 2008b; and references therein, but see Laurin and Reisz

(1997) for a different perspective). Despite this broad level of

agreement, there remains disagreement as to which of the

derived dissorophoids is the closest relative to anurans, ca-

udatans, or both (see Carroll 2007; Ruta and Coates 2007;

Anderson et al. 2008b), although recent suggestions that

branchiosaurids are deeply nested within amphibamids

(Schoch and Milner 2008; Frobisch and Schoch 2009) does

moderate this discrepancy. This disagreement is, at least in

part, due to an inconsistent picture of the evolutionary tran-

sitions that led up to the origin of anurans and caudatans, and

is considered by many to be due to the fragmentary

dissorophoid fossil record and the need for more character

data from other important sources, such as developmental

series, as suggested by Schoch and Milner (2004).

Most recent hypotheses for dissorophoid intrarelationships

are in agreement in the relatively basal position of Trematop-

idae within Dissorophoidea, where it may (Fig. 2; Olsoni-

formes sensu Anderson et al. 2008a; Frobisch and Schoch

2009) or may not form a clade with Dissorophidae (Schoch

and Rubidge 2005; Frobisch and Reisz 2008; Frobisch and

Schoch 2009). In contrast to derived dissorophoids, such as

amphibamids, trematopids represent some of the larger, more

generalized forms within Dissorophoidea, appearing in the

Time

Sha

pe

OffsetOnset

Duration

Rate

Fig. 1. Schematic representation of the multiple factors influencingontogeny. Onset or initiation of an event/process is commonlyused in studies of development; however, offset, rate and durationare additional informative factors of ontogeny that are oftenoverlooked.

Fig. 2. The pattern of relationships within Dissorophoidea (mod-ified from fig. 4 of Frobisch and Schoch 2009). In this hypothesis,Trematopidae (black arrow) forms a clade (Olsoniformes) withDissorophidae that is the sister taxon to Amphibamidae. Alterna-tively, Trematopidae does not form a clade with Dissorophidae, butrather falls out on the stem of Amphibamidae, basal to Dissoroph-idae (Schoch and Rubidge 2005). The amphibamid taxonGerobatrachus is the sister taxon to Batrachia (gray arrow) (Ander-son et al. 2008b), which includes modern anurans and caudatans.

394 EVOLUTION & DEVELOPMENT Vol. 12, No. 4, July--August 2010

Late Pennsylvanian and persisting into the Early Permian

(Dilkes 1990; Sumida et al. 1998; Berman et al. 2009). Trem-

atopidae is placed in a phylogenetically important position for

interpretation of the evolutionary processes that took place

within the dissorophoid lineage leading eventually to batra-

chians. Here we present a recently discovered, partial onto-

genetic series of the braincase of a large trematopid

dissorophoid, Acheloma n. sp. (Polley and Reisz in press).

Until now, OS data for trematopids were virtually nonexistent

(but see Olson 1985). Whereas much is known about the on-

togeny of the dermal skull in a variety of relevant extinct and

extant taxa, aspects of the development of the neurocranium

remain largely unknown, due to the fact that the former, in

most well preserved specimens, obscures the latter. The ex-

quisite preservation of the material presented here contributes

significantly to our understanding of dissorophoid braincase

anatomy and post-metamorphic ontogeny. The new informa-

tion is discussed within the context of dissorophoid phylogeny

and refines hypotheses of dissorophoid cranial evolution. We

demonstrate the importance of integrating DS and TS data in

addressing questions of heterochrony and its value in con-

tributing to the resolution of controversial evolutionary rela-

tionships, and that these data can be generated by limited

numbers of specimens, as long as they derive from different

points along the OT. This integrative approach permits the

identification of the evolutionary developmental events that

played a key role the evolution of the neurocranium of living

amphibians, and these may be incorporated into broader

considerations of the evolvability of the tetrapod cranium.

MATERIALS AND METHODS

The material used in the current study consists of four new spec-

imens referred to a new species of Acheloma (OMNH 73281; Polley

and Reisz in press), housed in the Sam Noble Museum of Natural

History, Oklahoma (OMNH 73493, 73494, 73509, and 73511). The

four new specimens are identical in morphology to Acheloma n. sp.

and to each other, differing only in size and proportions. The

sample is therefore interpreted as a partial ontogenetic series of

Acheloma n. sp. (Fig. 3; Table 1), consisting of one small (OMNH

73493; Fig. 4, A–D), one intermediate (OMNH 73494; Fig. 4,

E–H), and two large specimens (OMNH 73511 and 73509; Fig. 4,

I–L). The specimens preserve, to varying degrees of completeness,

the braincase and associated dermal skull roof. The holotype and all

referred specimens described here were recovered from the Dolese

Brothers Limestone quarry, near Richards Spur and Fort Sill, Co-

manche Co., Oklahoma. The site is designated V51 by the OMNH,

and referred to in the literature simply as ‘‘Richards Spur’’ or ‘‘Fort

Sill.’’ The fossiliferous fissure fill deposits are argued to represent a

Clear Fork Formation equivalent (Leonardian, Lower Permian,

TX, USA) horizon on the grounds of biostratigraphic data (see

Sullivan and Reisz 1999; Kissel et al. 2002; Maddin et al. 2006).

GENERAL ANATOMY OF THE NEUROCRANIUM

The preserved neurocraniumFparasphenoid, basisphenoid,

prootic, opisthotic, exoccipital, and basioccipitalFis exqui-

sitely preserved in three dimensions (Fig. 4). It is a well-

ossified structure roofed by the posterior portion of the

parietals and postparietals. The parasphenoid and basis-

phenoid are indistinguishably fused. A transverse ridge di-

vides the cultriform process from the basal plate of the

parasphenoid portion of the complex. A dental field extends

from this ridge almost to the level of the paired, posteriorly

projecting processes (basal tubera; Figs. 4K and 5L, asterisk).

Longitudinal grooves (vidian sulci) define the lateral borders

of the basal plate. At the anterior end of the groove a small

foramen for the entry of the internal carotid artery is present.

The ascending lateral walls of the parabasisphenoid complex

Fig. 3. Illustration comparing the relativesizes of the specimens referred to a partialontogenetic series of the trematopidAcheloma n. sp. (A) the smallest speci-men (OMNH 73493); (B) the intermedi-ately sized specimen (OMNH 73494); (C)the largest specimen (OMNH 73509). Aforth specimen (OMNH 73511) is com-parable in size to OMNH 73509. Linedrawings modified from Polley and Reisz(in press). Scale bar52 cm.

Dissorophoid neurocranial development 395Maddin et al.

form the anterior and dorsal borders of the large prootic

foramen. The prootic and opisthotic comprise the anterior

and posterior otic capsule, respectively. Both elements contact

the parabasisphenoid ventrally, and form the anterior and

posterior margins of fenestra vestibuli, respectively. In one

specimen (OMNH 73494), the stapes can still be found with

its footplate within the fenestra vestibuli. The opisthotic con-

tacts the exoccipital posteriorly, enclosing the jugular foramen

along their common suture. The opisthotic extends dorsolat-

erally, forming the robust paroccipital process. The occiput

and skull roof meet at a nearly right angle. A large portion of

the occiput is comprised of the paired occipital flanges of the

postparietals. The dorsal processes of the exoccipitals make

contact dorsal to the foramen magnum. The exoccipitals and

basioccipital are indistinguishably fused, and this complex

occupies the ventral portion of the occiput. The single median

basioccipital is a minor component in advanced temnospon-

dyls, but not in basal forms like Edops. The occipital condyle

is a bifaceted structure, formed by the exoccipitals. Two small

foramina are present on the lateral and ventrolateral surfaces

of the exoccipitals, and likely represent the exit point(s) for the

hypoglossal nerve (XII).

DESCRIPTION OF ONTOGENY-DEPEDENTFEATURES

Examination of the developmental series reveals a number of

ontogeny dependent features. The occipital plate in the small-

est specimen of Acheloma n. sp. (OMNH 73493) possesses a

gap between the dorsal processes of the exoccipitals and the

occipital flanges of the postparietals (Fig. 5, A and D, arrow).

The exoccipitals and postparietals grow towards each other to

make contact in the intermediately sized specimen (OMNH

73494). This growth closes the majority of the gap (Fig. 5, B

and E), and appears to be the product of both a dorsal ex-

tension of the exoccipitals and a ventral extension of the

postparietal flanges. In the largest specimens (OMNH 73509

and 73511) the exoccipitals have grown medially to make

contact dorsally, above the foramen magnum (Fig. 5, C and

F). It is interesting to note that presence of a gap, like that

observed in the smallest specimen here (Fig. 5D, arrow), has

been interpreted to indicate the presence of a cartilaginous

supraoccipital in other taxa (e.g., Phonerpeton, Dilkes 1990).

The specimens presented here show the gap is ontogeny

dependent, being closed in larger and presumably more ma-

ture specimens, and suggests a supraoccipital (even cartilag-

inous) was absent. This may also indicate that other taxa

possessing a gap in this region are immature individuals or

paedomorphic in this respect.

Two features of the basal plate of the parabasisphenoid

complex are revealed here to undergo change during the on-

togeny of the braincase. First, the relative dimensions of the

basal plate are modified through ontogeny. The extension of

the basal tubera throughout ontogeny results in a shape

change that progresses from a basal plate that is wider than

long in the most immature specimen (OMNH 73493; Fig. 5,

G and J) to a basal plate that is longer than wide in the most

mature specimens (OMNH 73509; Fig. 5, I and L). The basal

plate of the intermediately sized specimen (OMNH 73494) is

roughly square (Fig. 5, H and K). Second, the size and extent

of the dental field on the basal plate are also modified through

ontogeny. The adult specimens described here possess a well-

developed dental field extending from the base of the cultri-

form process to the level of the fenestra vestibuli (Fig. 5L);

however, in the most immature specimen, the dental field is

restricted to a small area between the basipterygoid articula-

tions (Fig. 5J). Both of these features are of particular interest

because they are used in both phylogenetic analyses at the

broader tetrapod and finer dissorophoid levels (Schoch and

Rubidge 2005; Ruta and Coates 2007; Anderson et al. 2008a;

Frobisch and Reisz 2008).

Table 1. Table of measurements of the four new specimens referred to Acheloma n. sp. (OMNH 73493, 73494, 73509,

73511) and the holotype specimen (OMNH 73281)

Measurement description (mm)

OMNH

73493

OMNH

73494

OMNH

73509

OMNH

73511

OMNH

73281

Total length 71.4 122.6 44.7 35.8 37.0

Total width 52.3 99.5 55.4 61.6 84.6

Length of parasphenoid from base of basal tubera

to base of cultriform process, cranial to basicranial articulations

15.9 24.2 25.2 25.2 24.2

Width of parasphenoid basal plate 19.8 29.9 31.4 31.6 31.8

Width of occipital condyles 11.5 24.2 25.2 25.2 23.7

Height of occipital condyles 4.9 10.3 10.3 10.0 10.0

Width of foramen magnum 5.0 8.1 7.6 9.0 8.4

Height of foramen magnum 6.3 14.1 12.1 12.7 13.8

Maximum width of postparietals on occipital surface 25.0 49.8 52.8 46.58 56.4

Width of exoccipitals 14.9 21.6 20.1 28.5 21.3

Height of exoccipitals 10.3 22.85 20.7 23.4 24.8

396 EVOLUTION & DEVELOPMENT Vol. 12, No. 4, July--August 2010

The otic capsule of the most immature specimen (OMNH

73493) is clearly comprised of two distinct ossifications

(separate prootic and opisthotic; Fig. 4C, gray). The prootic

and opisthotic are fused in the intermediately sized specimen

(OMNH 73494), forming the single, robust element of the

otic capsule commonly seen in dissorophoids. It had been

Fig. 4. Photographs and line drawings of the specimens representing the partial ontogenetic series of Acheloma n. sp. Ventral (A, C) andoccipital (B, D) views of the smallest specimen (OMNH 73493), showing the portion of the anterior and the complete posterior neuro-cranium. Ventral (E, G) and occipital (F, H) views of the intermediately sized specimen (OMNH 73494), showing portions of the anteriorbraincase, nearly complete posterior braincase, and the stapes of the middle ear. Ventral (I, K) and occipital (J, L) views of one of the largespecimens (OMNH 73509), showing the nearly complete posterior braincase. epi, epipterygoid; ex-bo, exoccipital–basioccipital complex; f,frontal; fm, foramen magnum; fv, fenestra vestibuli; jf, jugular foramen; op, opisthotic; p, parietal; pbs, parabasisphenoid; pof, postfrontal;pp, postparietal; pro, prootic; pro1op, fused prootic–opisthotic; pt, pterygoid; ptf, posttemporal foramen; st, supratemporal; sph,sphenethmoid/orbitosphenoid; s, stapes; t, tabular; II, optic nerve foramen; III, occulomotor nerve foramen; XII, hypoglossal nerveforamen. Scale bar52 cm.

Dissorophoid neurocranial development 397Maddin et al.

suggested that this single otic element in certain derived

temnospondyls is the product of a single ossification, as

evidence of more than one center of ossification is completely

absent in the adult forms (Schoch 1999; Robinson et al. 2005),

but the present study demonstrates the compound nature

of this ossification, and corroborates earlier observations

Fig. 5. Photographs and line drawings of the transformation series (TS) of the exoccipital–postparietal contact (A–F) and the para-basisphenoid complex (G–L) identified in the ontogenetic series of Acheloma n. sp. (A–F) Occipital views showing the TS of the exoccipitalsand postparietals. (A, D) The smallest specimen (OMNH 73493) showing a gap between the exoccipitals (gray) and postparietals (arrow).(B, E) Shows the gap closed in the intermediately sized specimen (OMNH 73494). (C, F) Shows that in the largest specimen the exoccipitalshave grown to meet dorsally, over the foramen magnum (OMNH 73509). (G–L) Ventral views showing the TS of the basal tubera anddental field of the basal plate. (G, J) Shows the presence of concavities (arrowheads) in the posterior margin of the basal plate of the smallestspecimen (OMNH 73493), where the basal tubera of the basal plate (gray) will grow, and the small patch of dentition (stippled area). (H, K)Shows the growth of the basal tubera and expansion of the dental field in the intermediately sized specimen (OMNH 73494). (I, L) Showsthe well-developed basal tubera (asterisks) and dental field (stippled area) in one of the largest specimens (OMNH 73511). Scale bar52 cm.

398 EVOLUTION & DEVELOPMENT Vol. 12, No. 4, July--August 2010

(Watson 1956, Carroll 1964). Additionally, the paroccipital

processes of the opisthotic extend in length laterally, through-

out the ontogeny of Acheloma n. sp. In the smallest specimen

(OMNH 73493) the processes are short and blunt, extending

only to the level of postparietal-tabular suture on the occipital

surface. In the intermediately sized specimen (OMNH 73494)

the paroccipital process extends laterally to just beyond the

postparietal-tabular suture. The largest specimens are incom-

plete laterally; however, the paraocciptial processes can be

seen in the holotype (OMNH 73281) extending well beyond

the postparietal-tabular suture (Polley and Reisz in press).

DISCUSSION

DSs are rich sources of information that can be used to both

generate hypotheses of phylogenetic relationship and explain

diverging morphologies within a clade. Studies that use DS

data overwhelmingly use the sequence of onset of ossification

as the primary source of information. The study of ossifica-

tion sequences in temnospondyl amphibians has yielded much

information about their development leading to comparisons

of DSs between taxa, and subsequent identification of poten-

tial heterochronic changes which can be incorporated into

discussions of evolution and relationships (Schoch 1992, 2004;

Boy and Sues 2000; Schoch 2003; Witzmann and Pfretschner

2003; Witzmann 2006; Witzmann and Schoch 2006). Con-

trary to these previous ontogenetic studies on temnospondyl

taxa, all osseous elements in the present sample, as well as

those of most fossils, are long past the stages of ossification

onset and therefore do not provide information on the onset

sequence. The present material does, however, provide valu-

able comparative information concerning the sequence of

later occurring ontogenetic events and their relative timing.

The obtained DS and TS data permit the construction of a

partial OT of the neurocranium of Acheloma n. sp.

Between the three post-metamorphic ontogenetic stages

represented by the material studied here, five events were

identified that appeared as incomplete in at least one specimen

and can be observed in an advanced stage in the successively

larger and presumably more mature specimen(s): (1) closure

of the exoccipital–postparietal gap, (2) ossification of basal

tubera of the basal plate (parabasisphenoid complex), (3) ex-

pansion of the basal plate dental field, (4) ossification of the

paroccipital processes, and (5) fusion of the prootic and opis-

thotic. At the stage represented by the smallest specimen all

bones have begun ossification but their ontogenies are in-

complete. By the time represented by the intermediately sized

specimen the exoccipital has contacted the postparietal and

the prootic and opisthotic have fused to form the otic capsule.

The basal tubera, dental field, and paroccipital processes have

progressed from the condition seen in the smallest specimen,

but have not yet reached the state of development seen in the

most mature specimens. In the largest specimens these latter

three events have achieved their adult condition (Fig. 5). The

exoccipitals have further progressed to meet dorsally over

the foramen magnum. The resulting partial OT is given in

Table 2.

The sequence of onset of ossification of the elements in-

volved (parasphenoid, prootic opisthotic, and exoccipital) is

variable among anurans, caudatans, and dissorophoid

temnospondyls for which there is ontogenetic data available

(Altig 1969; Bonebrake and Brandon 1971; Hanken and Hall

1984; Trueb and Hanken 1992; Schoch and Carroll 2003; and

others). However, a recurring sequence of onset, which is seen

in the basal anuran Ascaphus truei (Altig 1969) and was also

determined to represent the basal caudatan pattern (Rose

2003; Carroll 2007; Germain and Laurin 2009), is para-

sphenoid, exoccipital, opisthotic/prootic. This sequence of

onset does not correspond closely to the sequence of the dis-

crete TS documented here (Table 2). For example, the rela-

tively early onset of the parasphenoid ossification fails to

capture the relatively late-stage transformation described

above (basal tubera growth, dental field expansion). Con-

versely, relatively late onset of prootic and opisthotic ossifi-

cation is conflicted by the relatively early fusion of these

elements to form the otic capsule. It is therefore demonstrated

that consideration of integrated developmental data, like that

comprising the partial OT presented here, contains new in-

formation that considering only onset of events neglects.

Furthermore, this partial OT of Acheloma n. sp. can be

used to identify new potential heterochronic events that took

place in the evolution of the dissorophoid neurocranium by

permitting comparative analysis of derived amphibamid

temnospondyls (Fig. 6). In the smallest specimen of Ache-

loma n. sp. (T1; Fig. 6A) the exocciptals and postparietals are

separated by a gap, the basal plate, dental field, and the par-

occipital processes are weakly developed and in a state of

Table 2. Tabulation of the ontogenetic trajectory (OT)

of the neurocranium of Acheloma n. sp.

Stage Event

T1

OMNH 73493

All elements ossified

Basal tubera absent

Dental field small

Paroccipitals short

T2

OMNH 73494

Fusion of otic capsule

Closure of exo–pp gap

Extension of basal tubera

Extension of dental field

Extension of paroccipitals

T3

OMNH 73511

Exoccipitals meet

Completion of basal tubera

Completion of dental field

Completion of paroccipitals

Dissorophoid neurocranial development 399Maddin et al.

growth (events 1–4; Fig. 6A). At the same time the prootics

and opisthotics are unfused (event 5; Fig. 6A). Fusion of the

prootics and opisthotics occurs at some point between the

time represented by the smallest specimen (T1 in Fig. 6A) and

the intermediately sized specimen (T2 in Fig. 6A). By the time

fusion of the otic capsule has taken place, the gap between the

exoccipitals and postparietals is closed, the basal plate is

roughly as wide as it is long, the dental field has extended

posteriorly, and the paroccipital processes have extended lat-

erally (T2; Fig. 6A) from the conditions seen in the smallest

specimen (T1; Fig. 6A). The adult condition is represented by

the largest specimen (T3 [black]; Fig. 6A), wherein the exoc-

cipitals have made contact with each other dorsally, the basal

plate is longer than wide, the dental field covers half the plate,

and the paroccipital processes extend well beyond the tabular-

postparietal contact.

The presumed adult condition of the neurocranium of

amphibamids, in most regards, resembles the condition seen

in the most immature specimen of Acheloma n. sp. (lack of

black regions in Fig. 6, B and C). The presence of a gap

between the exoccipitals and postparietals is observed in sev-

eral derived amphibamid taxa including Amphibamus, Doles-

erpeton, and Eoscopus, as well as several more basal

amphibamids, such as Tersomius and Pasawioops (Carroll

1964; Bolt 1969; Daly 1994; Frobisch and Reisz 2008;

Sigurdsen 2008). The dimensions of the basal plate in derived

amphibamids also resemble those seen in the most immature

specimen of Acheloma n. sp. (wider than long; character 6 in

Schoch and Rubidge 2005; Frobisch and Reisz 2008). The

dental field is typically small and anteriorly restricted in

amphibamids (Frobisch and Reisz 2008), as in the immature

specimen of Acheloma n. sp. The condition of the paroccip-

ital processes is difficult to determine in most amphibamids

due to preservation limitations. Interestingly, the otic capsule

(prootic1opisthotic) is very well developed in all known

amphibamids, including most anurans and caudatans.

Fusion of the prootics and opisthotics to form the otic cap-

sule (a T2 event in Acheloma n. sp.) occurs in animals that

simultaneously exhibit morphologies of the other four events

that most resemble the conditions seen in the most immature

specimen of Acheloma n. sp. (T1 events in Acheloma n. sp.). It

appears, therefore, that in amphibamids the timing of otic

capsule fusion has changed relative to the timing of the other

four ontogenetic events when compared with those of Ache-

loma. The resulting morphology of the amphibamid neuro-

cranium is paedomorphic with respect to that of Acheloma,

with the exception of the otic capsules.

As described in previous studies (Jeffery et al. 2002, 2005)

the apparent shift in the timing of one event relative to an-

other may be the result of any one of five heterochronic events

(see fig. 3 in Jeffery et al. 2005). Jeffery et al. (2002, 2005)

argue that the location of the shift requiring the least number

of ad hoc hypotheses (or steps) to explain the new pattern

should be the one chosen (following the principle of parsi-

mony). Comparison of the possible locations of shifts that

would generate the observed change during the evolution of

Fig. 6. Schematic representation of the Ontogenetic Trajectory (OT)of the neurocranium of Acheloma n. sp. (A) derived from the partialontogenetic series in comparison with that of amphibamids (B andC). (A) Depiction of the progression (rates arbitrary) of the trans-formation series (TS) of the five events (1, exoccipital–postparietalcontact; 2, basal tubera; 3, dental field; 4, paroccipital processes;5, otic capsule fusion) relative to one another in the three ontogeneticstages (T1, smallest; T2, intermediate; T3, largest) of Acheloma n. sp.At the time represented by T1 (OMNH 73493) a gap is presentbetween the exoccipitals and postparietals, basal tubera are absent,dental field is restricted, parocciptial processes are short, and the oticcapsule is unfused. (B, C) Depiction of the modified OT hypoth-esized for amphibamids, wherein otic capsule fusion is completewhen events 1–4 resemble the state in T1 of Acheloma n. sp. (B)Depicts the acceleration of the timing of otic capsule fusion relativeto the other four events, which results in the observed combinationof morphologies observed in adult amphibamids. (C) Depicts the de-lay of the events 1–4 relative to otic capsule fusion in order to obtainthe combination of morphologies observed in adult amphibamids.

400 EVOLUTION & DEVELOPMENT Vol. 12, No. 4, July--August 2010

the neurocranium in dissorophoids reveals a single most par-

simonious hypothesis. In this hypothesis fusion of the otic

capsule has shifted to an earlier position relative to the re-

maining four events (Fig. 6B). This hypothesis invokes a sin-

gle step: acceleration (or predisplacement) of otic capsule

fusion. Although this is the most parsimonious hypothesis, it

involves a potentially risky modification to a highly functional

structure, the auditory apparatus. Virtually all known

dissorophoids possess anatomy suggestive of a tympanic-

stapedial auditory apparatus (Lombard and Bolt 1979; Bolt

and Lombard 1985; Reisz et al. 2009). The tympanic-stapedial

system works to convey air-borne sounds, such as intra- and

interspecific communication, to the inner ear (Mason 2007;

and references within). The development of this system in

extant anurans is precisely timed to coincide with emergence

to the terrestrial domain (Hetherington 1987). Presumably, a

shift to an earlier development of this system would not incur

an adaptive advantage.

An alternative hypothesis involves the delaying the other

four events relative to otic capsule fusion, and therefore, un-

der the method of Jeffery et al. (2002, 2005), invoking four

steps (Fig. 6C). However, it is possible that the four events are

not independent and that a single heterochronic event could

affect the entire neurocranium (except for otic capsule fusion),

such as somatic delay or developmental truncation. This in-

terpretation invokes an additional ad hoc hypothesis that is

not directly observable in the data at hand. However, recent

studies may provide insight into validity of this alternative.

Recent studies have implicated the heterochronic process

of miniaturization as having a major effect on the evolution of

our modern amphibians (Schoch and Rubidge 2005; Schoch

and Frobisch 2006; Frobisch and Reisz 2008; Frobisch and

Schoch 2009). Miniaturization may arise as a result of a

number of developmental modifications, and it can be difficult

to decipher the exact mechanism causing miniaturization

(Hanken and Wake 1993). The morphological consequences

of miniaturization are often reduced- or under-development,

and therefore can resemble paedomorphosis (e.g., neoteny or

progenesis), which may be due to truncated or condensed

development (Hanken and Wake 1993; Frobisch and Schoch

2009). However, miniaturization may also produce non-

paedomorphosis-like morphologies, such as hyperossification,

resulting in individuals that are chimeras of both pa-

edomorphosis-like and -unlike features (see Hanken and

Wake 1993). The observations made here are congruent with

predictions of miniaturization and reveals the neurocranium

was, in fact, undergoing morphological changes that had pre-

viously not been recognized.

We are provided with the first insights into how different

regions of the neurocranium of amphibamids were influenced

by miniaturization. It is interesting to note that the changes in

the neurocranium due to miniaturization appear superficially

less dramatic than those of the dermal skull, and that they are

possibly taking place at a slower rate. This may in part be due

to the very early appearance of the neurocranium in embry-

onic development, as well as the strong selection pressures

placed on the system by the central nervous system and au-

ditory apparatus. In general, miniaturization appears to have

had a paedomorphosis-like effect on the nature of the exoc-

cipital–postparietal contact, the basal tubera, basal plate den-

tal field, and paroccipital processes. These changes can be

broadly characterized as the loss of late-stage ontogenetic

events, a phenomenon well known to be a source of mor-

phological variation. In contrast, a well-developed fused otic

capsule has been retained in derived amphibamid temnos-

pondyls, despite an otherwise paedomorphic appearance of

the dermal skull and several other aspects of the neurocra-

nium, possibly due to the strong functional constraint placed

on the otic capsule during the evolution of stapedial-tympanic

auditory apparatus (Lombard and Bolt 1979; Bolt and Lom-

bard 1985). This conflict in morphology between the otic

capsules and the remainder of the neurocranium is even fur-

ther exaggerated in anurans and caudatans. These data sup-

port the hypothesis that the morphological change that is

observed in anurans and caudatans is a result of a shared

evolutionary process due to common ancestry with amp-

hibamids.

CONCLUSIONS

The discovery of four specimens referred to a partial onto-

genetic series of a new dissorophoid temnospondyl, Acheloma

n. sp., permitted the construction of a partial OT of the ne-

urocranium, integrating DS data and discrete TS data. It is

revealed that several features of the derived amphibamid

neurocranium, except for the well-developed otic capsule,

most closely resemble the conditions observed in the most

immature specimen of Acheloma. Consideration of alternative

hypotheses of heterochrony lends support to the hypothesis

that development of the neurocranium (excluding the otic

capsules) is delayed or truncated. The resulting morphology

of the amphibamid neurocranium is consistent with predic-

tions of miniaturization, and provides the first insights into

the nature of change in the neurocranium associated with

miniaturization within Dissorophoidea, which is continued in

extant members, such as anurans and caudatans. This obser-

vation supports a close evolutionary relationship between

these taxa and identifies a possible mechanistic source of

morphological evolution. Until now it was thought the

neurocranium remained relatively unaltered in comparison

with the dermal skull. It is apparent from this study that the

neurocranium was indeed being modified by the process min-

iaturization, though likely at a slower rate than the dermal

skull. This study emphasizes the value of developmental

data that extend beyond the sequence of ossification onset,

Dissorophoid neurocranial development 401Maddin et al.

especially since ontogenetic series preserving very early stages

are extremely rare in fossil taxa, and highlights the fact that

even incomplete series of specimens of late-stage individuals

of comparable size potentially hold important ontogenetic

data with phylogenetic implications.

AcknowledgmentsFirstly, we would like to thank the private collectors for graciouslydonating the material to the Sam Noble Museum of Natural History(OMNH), permitting this study. We are also grateful to the collec-tions staff at the OMNH for orchestrating the loan of this material.We thank B. Polley for numerous, co-operative discussions about theholotype specimen, and for permitting us to examine the holotypespecimen while under his study. We are indebted to Diane Scott forproviding photographs of the specimens figured here and to NicolaWong Ken for providing measurement for OMNH 73281 andOMNH 73511. The manuscript was improved by suggestions fromtwo anonymous reviewers. Funding for this study was provided by aNatural Science and Engineering Research Council C. G. S. to H. C.M., NSERC Discovery Grants to J. S. A. and R. R. R., and byAlberta Ingenuity to H.C.M.

REFERENCES

Alberch, P., Gould, S. J., Oster, G. F., and Wake, D. B. 1979. Size andshape in ontogeny and phylogeny. Paleobiology 5: 296–317.

Altig, R. 1969. Notes on the ontogeny of the osseous cranium of Ascaphustruei. Herpetol. Monogr. 25: 59–62.

Anderson, J. S., Henrici, A. C., Sumida, S. S., Martens, T., and Berman,D. S. 2008a. Georgenthalia clavinasica, a new genus and species ofdissorophoid temnospondyl from the Early Permian of Germany, andthe relationships of the family Amphibamidae. J. Vert. Paleo 28: 61–75.

Anderson, J. S., Reisz, R. R., Scott, D., Frobisch, N., and Sumida, S.2008b. A stem batrachian from the early Permian of Texas and the originof frogs and salamanders. Nature 453: 515–518.

Berman, D. S., Henrici, A. C., Brezinski, D. K., and Kollar, A. D. 2009. Anew trematopid amphibian (Temnospondyli: Dissorophoidea) from theupper Pennsylvanian of Western Pennsylvania: earliest record of terres-trial vertebrates responding to a warmer, drier climate. Ann. CarnegieMus. 78: 289–318.

Bolt, J. R. 1969. Lissamphibian originsFpossible protolissamphibian fromLower Permian of Oklahoma. Science 166: 888–891.

Bolt, J. R., and Lombard, R. E. 1985. Evolution of the amphibian tympanicear and the origin of frogs. Biol. J. Linn. Soc. 24: 83–99.

Bonebrake, J. E., and Brandon, R. A. 1971. Ontogeny of cranial ossificationin the small-mouthed salamander, Ambystoma texanum (Matthes).J. Morphol. 133: 189–204.

Boy, J. A., and Sues, H.-D. 2000. Branchiosaurs: larvae, metamorphosis,and heterochrony in temnospondyls and seymouriamorphs. InH. Heatwole and R. L. Carroll (eds.). Amphibian Biology: Paleontology.Surrey Beatty & Sons PTY Limited, Chipping Norton, pp. 1150–1197.

Bystrow, A. P., and Efremov, I. A. 1940. Benthosuchus sushkini Efremov.A labyrinthodont from the Eotriassic of Sharzenga River. Trav. I. Pa-leont. Acad. Sci. USSR 10: 1–152.

Carroll, R. L. 1964. Early evolution of the dissorophoid amphibians. Bull.Mus. Comp. Zool. 131: 161–250.

Carroll, R. L. 2007. The Paleozoic ancestry of salamanders, frogs and ca-ecilians. Zool. J. Linn. Soc. 150: 1–140.

Daly, E. 1994. The Amphibamidae (Amphibia: Temnospondyli), with adescription of a new genus from the Upper Pennsylvanian of Kansas.Univ. Kansas Misc. Pub 85: 1–59.

Dilkes, D. W. 1990. A new trematopsid amphibian (Temnospondyli:Dissorophoidea) from the Lower Permian of Texas. J. Vert. Paleo 10:222–243.

Frobisch, N. B., Carroll, R. L., and Schoch, R. R. 2007. Limb ossificationin the Paleozoic branchiosaurid Apateon (Temnospondyli) and the earlyevolution of preaxial dominance in tetrapod limb development. Evol.Dev. 9: 69–75.

Frobisch, N. B., and Reisz, R. R. 2008. A new Lower Permian amphibamid(Dissorophoidea, Temnospondyli) from the fissure fill deposits nearRichards Spur, Oklahoma. J. Vert. Paleo. 28: 1015–1030.

Frobisch, N. B., and Schoch, R. R. 2009. Testing the impact of miniatur-ization on phylogeny: Paleozoic dissorophoid amphibians. Syst. Biol 58:312–327.

Germain, D., and Laurin, M. 2009. Evolution of ossification sequences insalamanders and urodele origins assessed through event-pairing and newmethods. Evol. Dev. 11: 170–190.

Gould, S. J. 1977. Ontogeny and Phylogeny. Harvard University Press,Cambridge, 501pp.

Hanken, J., and Hall, B. K. 1984. Variation and timing of the cranialossification sequence of the Oriental Fire-bellied Toad, Bombina orien-talis (Amphibia, Discoglossidae). J. Morphol. 182: 245–255.

Hanken, J., and Wake, D. B. 1993. Miniaturization of body size: organ-ismal consequences and evolutionary significance. Ann. Rev. Ecol. Syst24: 501–519.

Hetherington, T. E. 1987. Timing of development of the middle-ear ofAnura (Amphibia). Zoomorphology 106: 289–300.

Jeffery, J. E., Bininda-Edmonds, O. R. P., Coates, M. I., and Richardson,M. K. 2005. A new technique for identifying sequence heterochrony.Syst. Biol. 54: 230–240.

Jeffery, J. E., Richardson, M. K., Coates, M. I., and Bininda-Edmonds, O.R. P. 2002. Analyzing developmental sequences within a phylogeneticframework. Syst. Biol. 51: 478–491.

Kissel, R. A., Dilkes, D. W., and Reisz, R. R. 2002. Captorhinus magnus, anew captorhinid (AmniotaEureptilia) from the Lower Permian ofOklahoma, with new evidence on the homology of the astragalus. Can.J. Earth Sci 39: 1363–1372.

Laurin, M., and Reisz, R. R. 1997. A new perspective on tetrapod phy-logeny. In S. S. Sumida and K. L. M. Martin (eds.). Amniote Origins.Academic Press, San Diego, pp. 9–59.

Lombard, R. E., and Bolt, J. R. 1979. Evolution of the tetrapod ear: ananalysis and reinterpretation. Biol. J. Linn. Soc. 11: 19–76.

Maddin, H. C., Evans, D. C., and Reisz, R. R. 2006. An Early Permianvaranodontine varanopid (Synapsida: Eupelycosauria) from the Rich-ards Spur locality, Oklahoma. J. Vert. Paleo 26: 957–966.

Mason, M. J. 2007. Pathways for sound transmission to the inner ear inamphibians. In P. Narins, A. S. Feng, R. R. Fay, and A. N. Popper(eds.). Hearing and Sound Communication in Amphibians. Springer Sci-ence1Business Media, New York, pp. 147–183.

Olson, E. C. 1985. A larval specimen of a trematopsid (Amphibia:Temnospondyli). J. Paleo 59: 1173–1180.

Polley, B., and Reisz, R. R. A new Lower Permian trematopid(Temnospondyli: Dissorophoidea) from Richards Spur, Oklahoma.Zool. J. Linn. Soc. in press.

Reisz, R. R., Schoch, R. R., and Anderson, J. S. 2009. The armoureddissorophid Cacops from the Early Permian of Oklahoma and the ex-ploitation of the terrestrial realm by amphibians.Naturwissenschaften 96:789–796.

Robinson, J., Ahlberg, P. E., and Koentges, G. 2005. The braincase andmiddle ear region of Dendrerpeton acadianum (Tetrapoda: Temnospon-dyli). Zool. J. Linn. Soc 143: 577–597.

Rose, C. S. 2003. The developmental morphology of salamander skulls. InH. Heatwole and M. Davies (eds.). Amphibian Biology: Osteology. Sur-rey Beatty & Sons, Chipping Norton, pp. 1684–1781.

Ruta, M., and Coates, M. I. 2007. Dates, nodes and character conflict:addressing the lissamphibian origins problem. J. Syst. Palaeo 5: 69–122.

Ruta, M., Coates, M. I., and Quicke, D. L. J. 2003. Early tetrapod rela-tionships revisited. Biol. Rev. 78: 251–345.

Schoch, R. R. 1992. Comparative ontogeny of early Permian branchiosau-rid amphibians from southwestern Germany. Palaeontographica, Abt. A222: 43–83.

Schoch, R. R. 1999. Studies on braincase of early tetrapods: structure,morphological diversity, and phylogenyF2. Dissorophoids, eryopids,and stereospondyls. Neues Jahrb. Geol. P-A 213: 289–312.

402 EVOLUTION & DEVELOPMENT Vol. 12, No. 4, July--August 2010

Schoch, R. R. 2002. The evolution of metamorphosis in temnospondyls.Lethaia 35: 309–327.

Schoch, R. R. 2003. Early larval ontogeny of the Permo-Carboniferoustemnospondyl Sclerocephalus. Palaeontology 46: 1055–1072.

Schoch, R. R. 2004. Skeleton formation in the Branchiosauridae: acase study in comparing ontogenetic trajectories. J. Vert. Paleo 24: 309–319.

Schoch, R. R., and Carroll, R. L. 2003. Ontogenetic evidence for thePaleozoic ancestry of salamanders. Evol. Dev. 5: 314–324.

Schoch, R. R., and Frobisch, N. B. 2006. Metamorphosis and neoteny:alternate pathways in an extinct amphibian clade. Evolution 60: 1467–1475.

Schoch, R. R., and Milner, A. R. 2004. Structure and implications of the-ories on the origin of lissamphibians. In G. Arratia, M. V. H. Wilson,and R. Cloutier (eds.). Recent Advances in the Origin and Early Radiationof Vertebrates. Verlag Dr. Friedrich Pfeil, Munich, pp. 345–377.

Schoch, R. R., and Milner, A. R. 2008. The intrarelationships and evo-lutionary history of the temnospondyl family Branchiosauridae. J. Syst.Palaeo 6: 409–431.

Schoch, R. R., and Rubidge, B. S. 2005. The amphibamidMicropholis fromthe Lystrosaurus assemblage zone of South Africa. J. Vertebr. Paleontol25: 502–522.

Schulmeister, S., and Wheeler, W. 2004. Phylogenetic analysis of develop-mental sequences. Evol. Dev 6: 50–57.

Sigurdsen, T. 2008. The otic region ofDoleserpeton (Temnospondyli) and its

impliactions for the evolutionary origins of frogs. Zool. J. Linn. Soc. 154:

738–751.Sullivan, C., and Reisz, R. R. 1999. First record of Seymouria (Vertebrata:

Seymouriamorpha) from Early Permian fissure fills at Richards Spur,

Oklahoma. Can. J. Earth Sci. 36: 1257–1266.Sumida, S. S., Berman, D. S., and Martens, R. 1998. A new trematopid

amphibian from the Lower Permian of central Germany. Palaeontology

25: 502–522.Trueb, L., and Hanken, J. 1992. Skeletal development in Xenopus laevis

(Anura: Pipidae). J. Morphol. 214: 1–41.Watson, D. M. S. 1956. The brachyopid labyrinthodonts. Bull Mus. Nat.

Hist. 2: 315–392.Witzmann, F. 2006. Developmental patterns and ossification sequence in

the Permo-Carboniferous temnospondyl Archegosaurus decheni (Saar-

Nahe Basin, Germany). J. Vert. Paleontol 26: 7–17.Witzmann, F., and Pfretschner, H.-U. 2003. Larval ontogeny of Micro-

melerpeton credneri (Temnospondyli, Dissorophoidea). J. Vert. Paleontol

23: 750–768.Witzmann, F., and Schoch, R. R. 2006. Skeletal development of the

temnospondyl Acanthostomatops vorax from the Lower Permian Dohlen

Basin of Saxony. Trans. Royal Soc. Edinburgh: Earth Sci. 96: 365–385.

Dissorophoid neurocranial development 403Maddin et al.