Cretaceous soft-shelled turtles (Trionychidae) of Mongolia: new data and a revision

This article was downloaded by: [Igor G. Danilov]On: 21 February 2014, At: 11:45Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Journal of Systematic PalaeontologyPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/tjsp20

Cretaceous soft-shelled turtles (Trionychidae) ofMongolia: new diversity, records and a revisionIgor G. Danilova, Ren Hirayamab, Vladimir B. Sukhanovc, Shigeru Suzukid, Mahito Watabed &Natasha S. Viteke

a Zoological Institute of the Russian Academy of Sciences, Universitetskaya Emb. 1, St.Petersburg, 199034, Russiab Waseda University, Nishiwaseda 1-7-14, Shinjuku-ku, Tokyo, 169-0051, Japanc Borisyak Paleontological Institute of the Russian Academy of Sciences, Profsoyuznaya 123,Moscow, 117997, Russiad Hayashibara Natural Sciences Museum, Okayama, 700-0907, Japane Department of Geological Sciences, The University of Texas at Austin, EPS RM 1.130, 1University Station C9000, Austin, TX 78712-0254, USAPublished online: 19 Feb 2014.

To cite this article: Igor G. Danilov, Ren Hirayama, Vladimir B. Sukhanov, Shigeru Suzuki, Mahito Watabe & Natasha S. Vitek ,Journal of Systematic Palaeontology (2014): Cretaceous soft-shelled turtles (Trionychidae) of Mongolia: new diversity, recordsand a revision, Journal of Systematic Palaeontology, DOI: 10.1080/14772019.2013.847870

To link to this article: http://dx.doi.org/10.1080/14772019.2013.847870

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

http://www.tandfonline.com/loi/tjsp20

http://www.tandfonline.com/action/showCitFormats?doi=10.1080/14772019.2013.847870

http://dx.doi.org/10.1080/14772019.2013.847870

http://www.tandfonline.com/page/terms-and-conditions

http://www.tandfonline.com/page/terms-and-conditions

Cretaceous soft-shelled turtles (Trionychidae) of Mongolia: new diversity,records and a revision

Igor G. Danilova*, Ren Hirayamab, Vladimir B. Sukhanovc, Shigeru Suzukid, Mahito Watabed and Natasha S. Viteke

aZoological Institute of the Russian Academy of Sciences, Universitetskaya Emb. 1, St. Petersburg, 199034, Russia; bWaseda University,Nishiwaseda 1-7-14, Shinjuku-ku, Tokyo, 169-0051, Japan; cBorisyak Paleontological Institute of the Russian Academy of Sciences,Profsoyuznaya 123, Moscow, 117997, Russia; dHayashibara Natural Sciences Museum, Okayama, 700�0907, Japan; eDepartment ofGeological Sciences, The University of Texas at Austin, EPS RM 1.130, 1 University Station C9000, Austin, TX 78712-0254, USA

(Received 10 April 2013; accepted 2 June 2013)

This paper is devoted to the description and revision of material of Cretaceous soft-shelled turtles (Trionychidae) ofMongolia. It includes the description of seven trionychid species, six of which are new, and two new genera: thecyclanorbine Nemegtemys conflata gen. et sp. nov. from the Nemegt Formation (Maastrichtian), and the trionychinesGobiapalone breviplastra gen. et sp. nov. from the Nemegt and Barungoyot (Campanian) formations, G. orlovi from theBaynshire Formation (Cenomanian–Santonian), ‘Trionyx’ baynshirensis sp. nov. from the Baynshire Formation, ‘T.’gilbentuensis sp. nov. from the Nemegt Formation, ‘T.’ gobiensis sp. nov. from the Nemegt Formation, and ‘T.’shiluutulensis sp. nov. from an unknown formation (Campanian). In addition, one shell from the ?Baynshire Formation ofKhermin Tsav is assigned to Gobiapalone sp. The type material of Amyda menneri is considered to be Trionychidae indet.and Amyda menneri to be a nomen dubium. Finally, we revise other available materials of Cretaceous trionychids from 45localities in Mongolia. Nemegtemys conflata, if correctly assigned, is the earliest known member of Cyclanorbinae. Thetwo species of the new genus Gobiapalone are included in two phylogenetic analyses of Trionychidae. In both analysesGobiapalone is monophyletic. In the first analysis, Gobiapalone is placed within Apalonina. In the second analysis,Gobiapalone is sister to Apalonina. Thus, the results of both analyses show that Apalonina, which is a rather advanced andwell-supported trionychid clade, or its closest sister taxon (stem-Apalonina), were present in the Late Cretaceous of Asia.These results suggest that most other supra-generic clades of modern trionychids had been established in Asia by the LateCretaceous. That suggestion is supported by the discovery of a cyclanorbine Nemegtemys conflata in the Late Cretaceousof Mongolia. Finally we summarize the latest data on temporal and geographical distributions of Cretaceous Trionychidaeof Asia and North America.

http://zoobank.org/urn:lsid:zoobank.org:pub:921DA1C5-C4B2-463D-A49D-608024C6036A

Keywords: Testudines; Trionychidae; Upper Cretaceous; Asia; Mongolia

Introduction

Trionychidae Grey, 1825, or soft-shelled turtles, are a

group of aquatic cryptodires (Meylan 1987). The phylog-

eny and taxonomy of extinct species within this group are

still not well understood (Meylan 1987; Gardner et al.

1995; Karl 1998; Joyce & Lyson 2011; Vitek 2011,

2012). The lack of understanding is especially problem-

atic for Cretaceous trionychids, which are important for

understanding the early diversification and evolution of

the family (Danilov & Vitek 2012a provided a review of

Cretaceous trionychids of Asia).

This paper continues a series of publications on Creta-

ceous trionychids of Asia (Danilov & Vitek 2009; Vitek

& Danilov 2010, 2012; Danilov & Vitek 2012a, b) and is

devoted to Cretaceous trionychids of Mongolia.

Mongolian specimens from that time are particularly

important because of their geographical and temporal loca-

tion in relation to specimens from North America and the

rest of Asia. That location and the perspective it provides

may help resolve standing questions of trionychid evolu-

tion. Long ghost lineages near the base of the trionychid

tree, especially in regard to the Cyclanorbinae Hummel,

1929, remain problematic. The gap in the cyclanorbine fos-

sil record has recently been filled to some extent by the

discovery that Plastomenidae Hay, 1902 are stem-

cyclanorbines (Joyce & Lyson 2010). Cretaceous plasto-

menids as they are currently known are restricted to North

America, but trionychids probably first evolved in Asia

and extant species of Cyclanorbinae are restricted to Africa

and Asia (Ernst & Barbour 1989; Hutchison 2000). Creta-

ceous representatives of Cyclanorbinae and Plastomenidae

*Corresponding author. Email: [email protected]

� The Trustees of the Natural History Museum, London 2014. All Rights Reserved.

Journal of Systematic Palaeontology, 2014

http://dx.doi.org/10.1080/14772019.2013.847870

Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

mailto:[email protected]

http://dx.doi.org/10.1080/14772019.2013.847870

are still unknown from the Cretaceous of Asia. The origin

of cyclanorbines remains unclear, as does the relationship

of plastomenids to the rest of cyclanorbine evolution in

Asia, for which the 80 million year long ghost lineage still

remains.

The example of cylanorbines and plastomenids can be

considered a specific example of the broader problem of

unknown biogeographical patterns within Trionychidae.

Current phylogenetic hypotheses for recently revised

taxa of Trionychidae support two potential hypotheses.

The hypothesis is that trionychids dispersed between

North America and Asia multiple times between the Cre-

taceous and today (Hutchison 2000; Vitek 2011). The

second hypothesis is that multiple species of trionychid

dispersed only once from Asia to North America, but

that most of the diversification leading to the extant gen-

era of trionychids had already occurred (Vitek 2011).

The composition and identity of the Mongolian triony-

chids may help support or refute those hypotheses,

because of the regions that are known to preserve rela-

tively complete Cretaceous fossil material, Mongolia is

closest to any potential land bridge between North

America and Asia.

The first report on Cretaceous trionychids of Mongolia

was published by Khosatzky & M»ynarski (1971), whodescribed ‘Trionyx sp. a’ based on an incomplete carapace

from the Nemegt Formation (Maastrichtian) of Tsagan

Khushu (see Fig. 1 for the position of this and other local-

ities of Cretaceous trionychids of Mongolia). M»ynarski &Narmandach (1972) mentioned a second carapace of

‘Trionyx sp. a’ and shell fragments of ‘Trionyx sp. b’, all

from the Nemegt Formation of Nemegt. Khosatzky

(1976) described Amyda orlovi Khosatzky, 1976 based on

an incomplete carapace from the lower part of the Bayn-

shire Formation (Cenomanian–early Turonian) of Bayn

Shire; the description of this species lacked illustrations.

Earlier, Sochava (1975) mentioned undescribed remains

attributed to Amyda orlovi (according to identifications by

L.I. Khosatzky) from the Baynshire Formation of Khara

Khutul. Undescribed records of trionychids from several

localities of the Baynshire (Cenomanian–Santonian) and

Nemegt formations of Mongolia were reported by

Figure 1. Map showing all known occurrences of Cretaceous Trionychidae in Mongolia: 1, Abdarain Nuru; 2, Abdrant Nuru; 3, AlakShand Khuduk; 4, Altan Ula, Altan Ula I, III and IV; 5, Amtgai; 6, Bayn Shire; 7, Bamba Khuduk; 8, Bayn Dzak; 9, Baishin Tsav; 10,Bugin Tsav and Bugin Tsav II; 11, Burkhant; 12, Char Teeg; 14, Dzun Shakhai; 16, Erdeni Ula; 17, Gilbentu; 18, Gurilin Tsav; 19,Ingeni Khobur and Ingeni Khobur III; 20, Ingeni Tsav; 22, Khara Khutul; 23, Khermin Tsav; 26, Khongil Tsav; 27, Khuren Tsav; 28,Nemegt; 29, Nogon Tsav and Nogon Tsav II; 30, Ongon Ulan Ula; 31, Shiluut Ula; 32, Shine Us Khuduk; 33, Shine Usny Tolgod; 34,Shiregin Gashun; 35, Tel Ulan Shalcha; 36, Tsagan Khushu; 37, Tsagan Teg; 39, Ulan Khushu; 40, Ulan Tsab Ula; 43, Ushyin Khuduk;44, Yagaan Khovil. Localities with uncertain position (13, 15, 21, 24, 25, 38, 41, 42, and 45) are not shown. See Online SupplementaryMaterial Appendix 1 for data on geology, age and material.

2 I. G. Danilov et al.

Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

Shuvalov & Chkhikvadze (1975, 1979). Chkhikvadze &

Shuvalov (1988) described Amyda menneri Chkhikvadze

in Chkhikvadze & Shuvalov, 1988 based on shell frag-

ments from the Nemegt Formation of Gurilin Tsav and

other localities. Khosatzky (1999) reported new triony-

chid shell material from several localities of the Lower

Cretaceous, and from the Baynshire, Barungoyot (Campa-

nian) and Nemegt formations of the Upper Cretaceous.

Sukhanov (2000) published illustrations of the holotype

of Amyda orlovi. Suzuki & Narmandakh (2004) reported

additional trionychid material from several Cretaceous

localities of Mongolia. Suzuki (2005) reported an almost

complete skeleton of Apalonina from an unreported local-

ity and formation of the Upper Cretaceous of Mongolia.

Data on Cretaceous trionychids of Mongolia were

included in reviews of published and/or figured material

of Cretaceous trionychids of Asia (Danilov & Vitek 2009,

2012a). Finally, Danilov et al. (2011a, b) reported prelim-

inary results of a revision of Cretaceous trionychids of

Mongolia.

Gilmore (1931) in his article entitled ‘Fossil turtles of

Mongolia’ mentioned fragmentary shell remains of trio-

nychids from the Cretaceous of the On Gong (Aptian–

Albian) and Iren Dabasu (?Campanian) formations.

Actually, these remains come from the territory of Inner

Mongolia, China, and are not considered in our study.

Here we revise, describe and illustrate material of Cre-

taceous trionychids of Mongolia (Figs 3–16), including

seven trionychid species, six of which are new, and two

new genera: the cyclanorbine Nemegtemys conflata gen.

et sp. nov. and the trionychines Gobiapalone breviplastra

sp. nov., G. orlovi, ‘Trionyx’ baynshirensis sp. nov., ‘T.’

gilbentuensis sp. nov., ‘T.’ gobiensis sp. nov. and ‘T.’ shi-

luutulensis sp. nov. In addition, one shell from the ?Bayn-

shire Formation of Khermin Tsav is assigned to

Gobiapalone sp. The type material of Amyda menneri is

considered to be Trionychinae indet. and Amyda menneri

to be a nomen dubium. Finally, we revised other available

material of Cretaceous trionychids from 45 localities of

Mongolia (Fig. 1, Table 1, Online Supplementary Mate-

rial Appendix 1).

The taxonomy of Trionychidae used here generally fol-

lows Meylan (1987), who presented the first comprehen-

sive phylogenetic analysis for living trionychids based on

skeletal morphology and provided diagnoses for the fam-

ily and all less inclusive taxa. Because Meylan’s (1987)

work relied on skeletal characters, it is especially useful

for dealing with fossil specimens and taxa. More recent

attempts to develop a phylogenetic taxonomy of triony-

chids (Engstrom et al. 2004; Joyce et al. 2004) are less

useful for Cretaceous taxa, because these studies also use

non-skeletal features and because few specimens of

Cretaceous trionychids have been included in cladistic

analyses. Anatomical terminology follows Meylan

(1987), Gardner & Russell (1994) and Karl (1999). Table1.ComparisonofshellcharactersofsomespeciesofCretaceoustrionychids.

Characters

‘Aspideretes’

maortuensis

Aspideretoides

riabinini

Gilmoremys

lancensis

Gobiapalone

breviplastra

Gobiapalone

orlovi

‘Trionyx’

gobiensis

‘Trionyx’

kansaiensis

‘Trionyx’

kyrgyzensis

‘Trionyx’

shiluutulensis

Maxim

um

carapacelength

(mm)

300�

500�

340

260�

335

130�

750�

150�

225

Nuchalem

argination

?Weak

Absent

Weakorabsent

Weakorabsent

Weakorabsent

Strong

Absent

Weak

Preneural

?Present

Present

Absent

Absent

Absent

Absent

?Present

Number

ofneurals

87

77or8or9

7or8

7or8

7or8

88

Neuralreversal

55or6

65

55

55

6Costals8

Small

Small

Large

Smallorabsent

Small

Large

Large

Large

Small

Epiplastralnotchonhyoplastron

Absent

Absent

Present

Absent

Absent

?Absent

Absent

?Medialprocesses

ofhyoplastron

Present

Present

Absent

Present

Present

?Present

Present

?Lateralhyoplastronlobein

relation

tomedialhyoplastronlobe

?Shorter

Alm

ostequal

Shorter

Shorter

?Longer

Longer

?

Ratio

ofminim

albridgelength

tomaxim

alhypoplastronlength

?About50%

About100%

About50%

About50%

?50–60%

About50%

?

Extensivemedialcontactofhyo-and

hypoplastra

No

No

No

No

No

?No

No

?

Medialhypoplastralprocesses

Clustered

Divided

?Divided

Divided

?Clustered

Clustered

?Number

ofplastralcallosities

Atleast2

Atleast4

Atleast4

55

?Atleast2

Absent

?

� Estim

ation

Cretaceous soft-shelled turtles (Trionychidae) of Mongolia 3

Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

Material and methods

We examined, either first hand or based on photographs,

most of the available material of Cretaceous trionychids

of Mongolia collected by the Mongolian Palaeontological

Expedition of Academy of Sciences of the USSR (MPE,

1946, 1948, 1949), the Polish-Mongolian Palaeontologi-

cal Expedition (PMPE, 1963–1971), the Joint Soviet-

Mongolian Palaeontological Expedition (JSMPE, 1969–

1991), and the Japan-Mongolia Joint Palaeontological

Expedition (JMJPE, 1993–2008), as well as material col-

lected by Russian and Mongolian geologists and biolo-

gists working in Mongolia in the twentieth century: V. N.

Chaykovskiy, R. Barsbold, L. Ya. Borkin, G. G. Martin-

son, I. Yu. Neustroeva, V. F. Shuvalov, A. V. Sochava

and N. N. Verzilin. However, we did not manage to exam-

ine material reported and described by Shuvalov &

Chkhikvadze (1975, 1979) and Chkhikvadze & Shuvalov

(1988), including type material of Amyda menneri, because

this material is not accessible for study (V. M. Chkhik-

vadze pers. comm. to I.G.D., 2012). Specimens collected

by JMJPE belong to the MPC, but are temporary stored in

the HMNS, and have both MPC and HMNS numbers.

The shell material described in our paper demonstrates

two main types of sculpturing on the shell surface, which

are referred to as type A and type B. The sculpturing of

type A is a pattern of thin, connected ridges forming a

honeycomb or netlike pattern. The sculpturing of type B

is a pattern of wide disconnected ridges and tubercles

raised above the plate. These types of sculpturing were

previously noted for Aspideretoides riabinini (Kuznetsov

& Chkhikvadze, 1987) (type A) and ‘Trionyx’ kansaiensis

Vitek & Danilov, 2010 (type B) from the Late Cretaceous

of Kazakhstan and Tajikistan (Fig. 2; see Vitek & Danilov

2010).

In order to determine the phylogenetic position of spe-

cies of Gobiapalone we performed two phylogenetic anal-

yses of trionychids, using the character/taxon matrices of

Joyce & Lyson (2011), and Vitek (2012), hereinafter anal-

yses 1 and 2 respectively. Characters coded for Gobiap-

alone breviplastra sp. nov. and G. orlovi in each of the

analyses are given in Online Supplementary Material

Appendices 2 and 3. The final data matrices include 83

characters for 28 taxa in analysis 1, and 88 characters for

41 taxa in analysis 2. In each analysis, Gobiapalone brevi-

plastra is 46% (analysis 1) and 49% (analysis 2) com-

plete, and G. orlovi is 80% (analysis 1) and 81% (analysis

2) complete. Each matrix was assembled using NDE 0.5.0

(Page 2001) and analysed with NONA ver. 2 and

Winclada ver. 1.00.08 by Ratchet algorithm with 1000

iterations. All characters were run unordered, unweighted,

and with no topological constraints. Support for each

node was measured by calculating Bremer supports

(Eriksson 1998).

Institutional abbreviationsHMNS: Hayashibara Museum of Natural Sciences,

Okayama, Japan; IPGAS: Institute of Paleobiology, Geor-

gian Academy of Sciences, Tbilisi, Georgia; MNHN:

Mus�eum national d’Histoire naturelle, Paris, France;

MPC: Mongolian Palaeontological Centre, Ulan Bator,

Mongolia; PIN: Borissyak Palaeontological Institute of

the Russian Academy of Sciences, Moscow, Russia; ZIN

PH: Paleoherpetological collection, Zoological Institute

of the Russian Academy of Sciences, St. Petersburg, Rus-

sia; ZPAL: Institute of Palaeobiology, Polish Academy of

Sciences, Warsaw, Poland. The term ‘Collection’ immedi-

ately preceding an institutional catalogue number

(e.g. Collection ZIN PH 156) indicates that multiple

specimens are accessioned under that number.

Systematic palaeontology

Testudines Batsch, 1788

Cryptodira Cope, 1868

Trionychidae Gray, 1825

Cyclanorbinae Hummel, 1929

Nemegtemys gen. nov.

Derivation of name. Nemegt, for the Nemegt locality,

and emys, Greek for turtle.

Type species. Nemegtemys conflata sp. nov.

Diagnosis. Same as for the type species.

Nemegtemys conflata sp. nov.

(Fig. 3)

1999 ‘Plastomenus’?; Khosatzky: 147.

2009 Trionychinae indet. 6; Danilov & Vitek: 54 (part.).

2012a Trionychidae indet. 5; Danilov & Vitek: 429

(part.).

Figure 2. Two main types of shell sculpturing of Cretaceoustrionychids of Asia. A, type A (ZIN PH 822/64, Aspideretoidesriabinini, costal 1); B, type B (ZIN PH 917/64, ‘Trionyx’kansaiensis, partial costal).


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

Derivation of name. Species name, conflata, Latin for

fused, in reference to fused hyo- and hypoplastra.

Holotype. ZIN PH 1/157, lateral fragment of left hyo-

and hypoplastra, from Nemegt, Umunugovi Aimag, Mon-

golia, Nemegt Formation, Maastrichtian (collected by

MPE in 1948).

Referred material. MPC 25/161 (HMNS 2006-4-472),

lateral fragment of right hypoplastron, from Bugin Tsav,

Umunugovi Aimag, Mongolia, Nemegt Formation, Maas-

trichtian (collected by JMJPE in 2006).

Diagnosis. A cyclanorbine which can be differentiated

from all other cyclanorbines in that, at the stage when the

hyo- and hypoplastra are completely ossified (i.e. when

all their processes are covered by metaplastic bone), the

hyoplastral lappet is small, forming an almost straight

anterior margin of the hyoplastron. In all other cyclanor-

bines, at the similar stage of ossification, the hyoplastral

lappet is larger (more extensive anteroposteriorly), form-

ing the rounded anterior margin (shoulder) of the

hyoplastron.

Description. ZIN PH 1/157 is a lateral fragment of the

left hyo- and hypoplastra. It is 4.5 cm long (from the ante-

rior margin of the lateral hyoplastron to the posterior mar-

gin of the lateral hypoplastron). The estimated length of

the plastron (based on comparison with Cycloderma

aubryi, MNHN 1930-362) is about 130 mm. The hyo-

and hypoplastra are fused and well-ossified. There are two

lateral processes on both the hyo- and hypoplastra (visible

on the upper surface of the fragment). These processes are

almost completely covered by metaplastic bone and proj-

ect only slightly beyond it (this is visible on the processes

of the hypoplastron and reconstructed for the processes of

the hyoplastron, which are broken off). The lateral

margins of the hyo- and hypoplastra are almost straight

and were probably parallel to the midline of the plastron.

The anterior margin of the hyoplastron (and hyoplastral

lappet) is straight and perpendicular to the lateral margin

of the hyo- and hypoplastra. The hyoplastral lappet makes

up about half of the bridge length and is shorter anteropos-

teriorly than in other cyclanorbines. The plastral callosity

covers most of the external surface of the fragment, except

for a narrow band along the anterior and lateral margins of

the specimen. The sculpturing on the external surface is

made of a fine net of ridges and tubercles; the latter are

strongly expressed at the posterior margin of the

specimen.

MPC 25/161 is a lateral fragment of the right hypoplas-

tron from a slightly larger individual than ZIN PH 1/157.

The fragment is well ossified, probably had a straight lat-

eral margin. It has a pair of lateral processes, which pro-

trude only slightly beyond the metaplastic bone. The

plastral callosity completely covers the external surface of

the specimen. The sculpturing of the external surface is

similar to ZIN PH 1/157, although tubercles are absent in

MPC 25/161.

Remarks. ZIN PH 1/157 was mentioned by Khosatzky

(1999, p. 147) as a “plastron fragment of a young triony-

chid individual, characterized by a rapid extension of the

plastron in the ontogenesis, and by a sculpturing of the

external surface with tubercles (‘Plastomenus’?)”. Dani-

lov & Vitek (2009, 2012a) erroneously listed

‘Plastomenus’? amongst the synonymies of their Triony-

chinae indet. 6 and 5 respectively, which are herein con-

sidered to be synonymies of ‘Trionyx’ gilbentuensis sp.

nov. (see below). The most peculiar character of ZIN PH

1/157 is the fusion of the hyo- and hypoplastra at a small

size. According to Meylan (1987), such a fusion (just after

hatching) is a synapomorphy of recent cyclanorbines. In

Figure 3. Shell material of Nemegtemys conflata gen. et sp. nov. A, B, ZIN PH 1/157 (holotype), a lateral fragment of hyo- andhypoplastra from the Nemegt Formation of Nemegt in A, dorsal and B, ventral views; C, D, MPC 25/161, a lateral fragment of the righthypoplastron from Bugin Tsav in C, ventral and D, dorsal views. Reconstructed outlines are indicated by dashed lines.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

trionychines, the fusion of the hyo- and hypoplastra, if

present, occurs only in adults (at large size) of some spe-

cies. In the clade Plastomenidae, which is considered to

be stem Cyclanorbinae, the fusion of the hyo- and hypo-

plastra is not known (Hutchison 2009; Joyce et al. 2009;

Joyce & Lyson 2010, 2011).

Trionychinae Gray, 1825

Trionychini Gray, 1825

Gobiapalone gen. nov.

Derivation of name. Gobi-, for the Gobi Desert, and

Apalone Rafinesque, 1832, a genus of North American

trionychids.

Type species. Gobiapalone orlovi (Khosatzky, 1976)

comb. nov.

Other species. Gobiapalone breviplastra sp. nov.

Diagnosis. A member of Trionychini which is most simi-

lar to members of the clade Apalonina and to Oliveremys

uintaensis (Leidy, 1872) in the presence of a vomer that

divides the maxillae and reaches the intermaxillary fora-

men, and small or absent costals 8. It can be differentiated

from all other members of Trionychini and other Creta-

ceous trionychids with known shell morphology by the

presence of five plastral callosities. In addition, it can be

differentiated from Apalone spp. by an apertura narium

externa weakly emarginated laterally and an unfused hyo-

and hypoplastra; from Aspideretoides spp. by a narrower

and longer crista supraoccipitalis, a wider dentary sym-

physis, absence of a dentary pocket and absence of a pre-

neural; from Axestemys spp. by smaller size, a vomer that

does not divide the maxillae and reach the intermaxillary

foramen, absence of a dentary pocket, absence of a pre-

neural and two lateral hyoplastral processes; from

Oliveremys uintaensis by a larger intermaxillary foramen,

a smaller carapace size, less ossified hyo- and hypoplastra,

and xiphiplastra with sculptured callosities; from Trionyx

triunguis (Forska�l, 1775) by a larger intermaxillary fora-

men, and smaller costals 8. See Table 1 for further differ-

ences between Gobiapalone spp. and Cretaceous

trionychids.

Gobiapalone orlovi (Khosatzky, 1976), comb. nov.

(Figs 4–9)

1975 Amyda orlovi Khos.; Sochava: 115 (nomen

nudum).

1976 Amyda orlovi Khosatzky; Khosatzky: 6.

1988 Amyda orlovi Khosatzky; Chkhikvadze & Shuvalov:

200.

1999 Amyda sp. cf. A. orlovi; Khosatzky: fig. 1.

1999 Trionychina indet.; Khosatzky: 143, fig. 2.

1999 Trionychidae indet.; Khosatzky: 144, fig. 3.

2000 Amyda orlovi Khosatzky; Sukhanov: 345,

fig. 17.27.

2000 ‘Amyda’ orlovi Khosatzky; Sukhanov: 345.

2005 Apalonina; Suzuki: 27.

2009 ‘Amyda’ orlovi Khosatzky; Danilov & Vitek: 54.

2009 Trionychinae indet. 3; Danilov & Vitek: 54

2009 Trionychinae indet 4; Danilov & Vitek: 54.

2010 ‘Amyda’ orlovi Khosatzky; Vitek & Danilov: table 1,

fig. 1.

2011a Amyda orlovi Khosatzky; Danilov et al.: 15.

2011b ‘Amyda’ orlovi Khosatzky; Danilov et al.: 96.

2011b Trionychinae indet.; Danilov et al.: 96.

2012a Apalonina indet. 3; Danilov & Vitek: 421.

2012a ‘Amyda’ orlovi Khosatzky; Danilov & Vitek: 423,

fig. 23.2b.

2012a Trionychinae indet. 3; Danilov & Vitek: 429,

fig. 23.3e.

2012a Trionychinae indet 4; Danilov & Vitek: 429,

fig. 23.3f.

Holotype. PIN 557-132/1 (formerly PIN 557-1/1),

incomplete carapace, from Bayn Shire, Dornogovi Aimag,

Mongolia, lower part of the Baynshire Formation,

Cenomanian–early Turonian (collected by MPE in 1948).

Referred material. MPC 25/162 (HMNS 94-10-9), car-

apace (collected by JMJPE in 1994); MPC 25/160

(HMNS 96-12-21-25), partial skeleton (collected by

JMJPE in 1996), all from Burkhant, Dornogovi Aimag,

Mongolia, Baynshire Formation, Cenomanian–Santonian;

PIN 557-133, fragments of carapace on plaster, from

Bayn Shire, Dornogovi Aimag, Mongolia, lower part of

Baynshire Formation, Cenomanian–early Turonian (col-

lected by MPE); ZIN PH T/M46-1, partial hyo- and

hypoplastra, from Bayn Shire, Dornogovi Aimag, Mon-

golia, lower part of Baynshire Formation, Cenomanian–

early Turonian (collected by MPE in 1946); ZIN PH T/

M67-5, anterior part of carapace, from Unegetu Ula,

Dornogovi Aimag, Mongolia, Baynshire Formation, Cen-

omanian–Santonian (collected by A.V. Sochava in

1967); ZIN PH T/M71-1, incomplete carapace, from

Khongil, Dornogovi Aimag, Mongolia, Baynshire For-

mation, Cenomanian–Santonian (collected by V. F.

Shuvalov in 1971).

Diagnosis. A species of Gobiapalone with a carapace

length up to 335 mm, which can be differentiated from

G. breviplastra by the closed suprascapular

fontanelles in specimens larger than 275 mm in cara-

pace length, slender epiplastra and entoplastron, a

smaller (about 80�) angle between arms of entoplastron,

more numerous (five) medial processes of hyoplastron,

a bigger gap between anterior and posterior groups of

medial processes of hypoplastron, and slender

xiphiplastra.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

Figure 4. Skull of Gobiapalone orlovi (Khosatzky, 1976) (MPC 25/160) from the Baynshire Formation of Burkhant. A, dorsal view; B,ventral view; C, right lateral view; D, left lateral view; E, partial right ventrolateral view; F, anterior view; G, posterior view; H, partialposteroventral view; I, partial posterodorsal view.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

Figure 5. Skull of Gobiapalone orlovi (Khosatzky, 1976) (MPC 25/160) from the Baynshire Formation of Burkhant, explanation draw-ings: A, dorsal view; B, ventral view; C, right lateral view; D, left lateral view; E, partial right ventrolateral view; F, anterior view;G, posterior view; H, partial posteroventral view; I, partial posterodorsal view. Bones are filled with grey (foreground) and dark grey(background). Matrix or gypsum is stippled. Broken surfaces are hatched. Reconstructed sutures and outlines are indicated by dashedlines. Abbreviations: ane, apertura narium externa; bo, basioccipital; bs, basisphenoid; co, condylus occipitalis; ex, exoccipital; fi,foramen intermaxillaris; fjp, foramen jugulare posterius; fm, foramen magnum; fnh, foramen nervi hypoglossi; fpcci, foramen posteriuscanalis carotici interni; fpo, fenestra postotica; fpp, foramen palatinum posterius; fr, frontal; fst, foramen stapedio-temporalis; ica,incisura columellae auris; ina, internal narial opening; ju, jugal; mx, maxilla; op, opisthotic; pal, palatine; pf, prefrontal; po, postorbital;pro, prootic; pt, pterygoid; qj, quadratojugal; qu, quadrate; so, supraoccipital; sq, squamosal; vo, vomer.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

Description

Skull. The skull is known from a single specimen (MPC

25/160; Figs 4, 5). It is almost complete, but damaged due

to partial deformation and poor preservation. As a result, the

skull is covered with numerous cracks filled with matrix;

some breaks were restored with gypsum. Many sutures are

not distinct or barely visible. The premaxillae are absent, as

well as part of the left jugal arch. From the anteriormost tip

of the maxillae to the posteriormost end of the supraoccipi-

tal, the skull is about 14 cm long. The condylobasal length

of the skull is about 10 cm. The width of the skull at the

Figure 6. A–F, Lower jaw, and G, left cornu branchiale I of Gobiapalone orlovi (Khosatzky, 1976) (MPC 25/160) from the BaynshireFormation of Burkhant. A, D, left lateral view; B, E, left dorsolateral view; C, F, dorsal view; G, dorsal view. Bones are filled withgrey. Matrix or gypsum is stippled. Broken surfaces are hatched. Reconstructed sutures and outlines are indicated by dashed lines.Abbreviations: an, angular; ar, articular; cor, coronoid; den, dentary; fai, foramen alveolare inferius; fme, fossa meckelii; pra, prearticu-lar; sur, surangular.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

level of the quadrates is about 7.5 cm. Thus, the width/con-

dylobasal length ratio is 0.75. The face of the skull is rela-

tively wide and short. The orbits are relatively large and

face mainly dorsally. The ratio of the interorbital distance to

the length of the orbit is about 0.30.

The prefrontal contacts the frontal posteriorly and the

maxilla laterally, makes up the anterodorsal part of the

margin of the orbit between the frontal and maxilla, and

forms the dorsal edge of the apertura narium externa. A

descending process of the prefrontal was probably

Figure 7. Shell material of Gobiapalone orlovi (Khosatzky, 1976). A–D,MPC 25/160 from the Baynshire Formation of Burkhant: A, B,carapace in A, dorsal and B, ventral views; C, D, plastron in C, dorsal and D, ventral views. E, MPC 25/162, carapace from the Bayn-shire Formation of Burkhant in dorsal view. F, PIN 557-132/1, carapace (holotype) from the lower part of the Baynshire Formation ofBayn Shire in dorsal view. G, PIN 557-133, fragments of carapace on plaster from the lower part of the Baynshire Formation of BaynShire in dorsal view. H, ZIN PH T/M46-1, partial hyo- and hypoplastra from the lower part of the Baynshire Formation of Bayn Shire inventral view. I, ZIN PH T/M71-1, carapace from the Baynshire Formation of Khongil in dorsal view. J, ZIN PH T/M67-5, anterior partof the carapace from the Baynshire Formation of Unegetu Ula in dorsal view. Reconstructed sutures and outlines are indicated by dashedlines.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

present within the orbit, but it is not discernible from the

surrounding bones due to poor preservation. The

anterior margin of the prefrontal (i.e. the dorsal edge of

the apertura narium externa) is weakly emarginated

laterally as in many trionychids, but not at all medially

emarginated.

The frontal contacts the prefrontal anteriorly, the parie-

tal posteriorly, and the postorbital posterolaterally, and

forms a majority of the dorsal border of the orbit between

the prefrontal and postorbital. The suture between the

frontal and prefrontal is slightly oblique and the prefrontal

therefore projects slightly between the frontals. The suture

between the frontal and parietal is perpendicular to the

midline and partially interdigitated. The suture between

the frontal and postorbital is clearly visible only on the

right side of the skull, whereas on the left side that area is

damaged. The ventral surface of the frontals is not visible

and its condition (fused or unfused; see Joyce & Lyson

2011) cannot be established.

The dorsal plate of the parietal contacts the frontal ante-

riorly, the postorbital anterolaterally on the skull roof,

the jugal anterolaterally within the fossa temporalis, and

the supraoccipital posteriorly. The parietal does not con-

tribute to the orbital margin or walls. The anterior contacts

of the ventral plate of the parietal are unclear, whereas

posteriorly the parietal contacts the prootic and supraocci-

pital. The contribution of the ventral plate of the parietal

to the margin of the foramen trigemini is not clear. The

contribution of the ventral plate of the parietal to the

formation of the processus trochlearis oticum is about

one-fourth of the total width.

The postorbital contacts the frontal medially, the parie-

tal posteriorly and the jugal laterally. The postorbital con-

tributes to the posterior rim of the orbit and to the anterior

rim of the upper temporal emargination. The postorbital

bar is about a quarter of the orbit diameter.

On the external surface of the skull, the jugal contacts

the maxilla anteriorly and the postorbital anterodorsally,

and the quadratojugal posteriorly. The ventral contact

between the jugal and the parietal is clearly visible within

the fossa temporalis. Whether the jugal contributes to the

orbital rim is not clear.

The quadratojugal contacts the jugal anteriorly, and the

squamosal and quadrate posteriorly, and contributes to the

anterior rim of the cavum tympani. The quadratojugal

does not participate in the processus trochlearis oticum.

The squamosal contacts the quadratojugal anteriorly,

the quadrate ventrally and anteromedially, and the

Figure 8. Cervical vertebrae of Gobiapalone orlovi (Khosatzky, 1976) (MPC 25/160) from the Baynshire Formation of Burkhant, in A,ventral, B, left lateral and C, dorsal views. Reconstructed outlines are indicated by dashed lines.

Figure 9. Girdles of Gobiapalone orlovi (Khosatzky, 1976)(MPC 25/160) from the Baynshire Formation of Burkhant. A,right pectoral girdle in posterodorsal view; B, right pelvic girdlein dorsal view. Reconstructed outlines are indicated by dashedlines.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

opisthotic posteromedially. In lateral view, the squamosal

forms the roof of the cavum tympani as well as much of

the antrum postoticum.

The premaxillae are not preserved.

In dorsal view, the maxilla contacts the prefrontal ante-

romedially and the jugal posteriorly, and forms the lateral

margin of the apertura narium externa and the ventral mar-

gin of the orbit. The maxilla does not contact the frontal or

the quadratojugal. In ventral view, the maxilla forms the

primary palate and probably contacted the premaxilla ante-

riorly. The maxilla contacts the vomer anterior and the pal-

atine and pterygoid posterior to the choana. The foramen

intermaxillaris is present and makes up about 60% of the

length of the primary palate. The maxilla forms most of the

margin of that foramen. The maxilla does not participate in

the formation of the foramen palatinum posterius. The trit-

urating surface formed by the maxilla is triangular-shaped,

narrowed anteriorly and expanded posteriorly.

The vomer is shaped peculiarly, with an expanded ante-

rior plate and a narrow posterior process. The vomer con-

tacts the maxillae laterally and the palatines posteriorly. It

fully divides the maxillae and contributes to the formation

of the foramen intermaxillaris anteriorly and internal

narial openings posteriorly.

The palatines are only observable in ventral view. Each

palatine contacts its counterpart medially, the vomer ante-

romedially, the maxilla anterolaterally, the pterygoid lat-

erally and the basisphenoid posteriorly. The palatine

forms the posterior rim of the choana and, together with

the pterygoid, contributes to the formation of the small

foramen palatinum posterius. Other contacts and contribu-

tions of the palatine are not observable.

In ventral view, each pterygoid contacts the maxilla

anteriorly, the palatine, basisphenoid and basioccipital

medially, and the quadrate posterolaterally. The foramen

palatinum posterius is positioned between the anterior

portion of the pterygoid and the palatine. The foramen

posterius canalis carotici interni is positioned below the

lateral crest of the basioccipital tubercle and is completely

enclosed within the pterygoid. In posterior view, the pter-

ygoid contacts the opisthotic and quadrate and contributes

to the formation of the fenestra postotica. The contact of

the pterygoid with the exoccipital is not clear. The con-

tacts and contributions of the pterygoid to the lateral wall

of the braincase are not observable.

The epipterygoid and the trigeminal foramen are not

discernable due to poor preservation of the specimen.

The basisphenoid is trapezoidal in shape, narrowed ante-

riorly and widened posteriorly. It is not medially con-

stricted. The basisphenoid contacts the palatines anteriorly,

the pterygoids laterally and the basioccipital posteriorly.

The prootic forms much of the processus trochlearis oti-

cum. It contacts the parietal medially, the supraoccipital

posteromedially, the quadrate laterally and the opisthotic

posteriorly. The foramen stapedio-temporalis is

positioned between the prootic and the quadrate. Other

contacts and contributions of the prootic are not

observable.

In dorsal view, the opisthotic forms the posterior mar-

gin of the otic capsule and contacts the prootic anteriorly,

the quadrate and squamosal laterally and the supraoccipi-

tal posteromedially. In posterior view, the opisthotic con-

tacts the supraoccipital and exoccipital medially, and the

squamosal and quadrate laterally. The descending process

of the opisthotic excludes the foramen jugulare posterius

from the fenestra postotica only on the right side of the

skull.

In dorsal view the quadrate contacts the quadratojugal

anteriorly, the prootic anteromedially, the opisthotic post-

eromedially and the squamosal posteriorly. The quadrate

contributes minimally to the formation of the processus

trochlearis oticum and to the lateral rim of the foramen

stapedio-temporalis. In lateral view, the quadrate forms

most of the subtriangular cavum tympani and the articular

surface for the lower jaw. Within the cavum tympani, the

quadrate contacts the squamosal dorsally. Anterior to the

cavum tympani, the quadrate contacts the quadratojugal.

In ventral view, the quadrate contacts the pterygoid medi-

ally. In posterior view, the quadrate contacts the squamo-

sal dorsally, the opisthotic dorsomedially, and the

pterygoid ventromedially.

In ventral view, the basioccipital contacts the basisphe-

noid anteriorly and the pterygoids laterally. The basiocci-

pital forms a pair of distinct tubercula basioccipitale. In

posterior view, the basioccipital forms part of the condy-

lus occipitalis and contacts the exoccipitals laterally.

The exoccipitals form the dorsal part of the condylus

occipitalis. In posterior view, they contact the supraocci-

pital and opisthotic dorsally and the basioccipital ven-

trally. The single foramen nervi hypoglossi is visible on

the left exoccipital.

The supraoccipital projects from the posterior end of

the skull in a long process (crista occipitalis). The crista

occipitalis is T-shaped in cross section. In dorsal view, the

supraoccipital contacts the parietals anteriorly, the prootic

anterolaterally and the opisthotic laterally. In posterior

view, the supraoccipital makes up the dorsal margin of

the foramen magnum and contacts the exoccipitals ven-

trally and the opisthotic ventrolaterally.

Mandible. The mandible is almost complete, but like the

skull, is covered with a net of cracks. In spite of these

cracks, most sutures are clearly visible.

The anterior margin of the dentary is rounded in dorsal

view. The triturating surface of the dentary is flat, lacks

lingual or labial ridges, and is expanded (the length of its

symphyseal region is about 25% of the total length of the

mandible). In dorsal view, the dentary contacts the coro-

noid posterolaterally. In lateral view, the dentary produces

an ascending process that covers the anterior aspect of the


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

coronoid process. Posterior to the coronoid process, the

dentary contacts the coronoid anterodorsally, the surangu-

lar posterodorsally and the articular posteriorly. In medial

view, the dentary contacts the prearticular and angular

below the coronoid process. There are a number of rather

small foramina penetrating the dentary below the coro-

noid process, one of which probably represents the fora-

men dentofaciale majus.

The angular, as seen in the medial view of the right

ramus of the mandible, is broken and represented only by

its anterior half. It contacts the dentary ventrally, and

probably had a dorsal contact with the prearticular.

The surangular is clearly visible on the lateral side of

the mandible. It contacts the dentary anteriorly and the

coronoid anterodorsally. Contact between the surangular

and the angular is not observable due to breakage of the

latter. The foramen nervi auriculotemporalis is visible on

the surangular just anterior to the area articularis mandibu-

laris. In dorsal view, the surangular frames the lateral mar-

gin of the fossa Meckelii and forms the lateral part of the

area articularis mandibularis and the retroarticular process.

The coronoid forms most of the coronoid process, but

does not contribute to the triturating surface. It is over-

lapped anteriorly by the ascending process of the dentary.

The coronoid contacts the surangular lateral to and the

prearticular medial to the fossa Meckelii, respectively.

The prearticular is broken along the probable ventral

contact with the angular. It contacts the coronoid antero-

dorsally, the dentary anteriorly and the articular posteri-

orly, and forms the medial wall of the fossa Meckelii. The

contribution of the prearticular in the formation of the

foramen intermandibularis caudalis is unclear.

The articular contacts the surangular and dentary anteri-

orly and forms the medial part of the area articularis man-

dibularis and retroarticular process.

Hyoid. The hyoid is represented by a single element, the

left cornu branchiale I. In its outline it is most similar to

those of Aspideretes hurum (Gray, 1831) (see Meylan

1987, fig. 18C).

Carapace. The disc of the carapace is slightly longer

than wide in PIN 557-132/1 and about as long as wide in

other specimens. The anterior border of the carapace is

rounded, either with a weak nuchal emargination (MPC

25/160 and ZIN PH T/M71-1) or with no emargination

(MPC 25/162). The lateral borders of the carapace are

either smooth, or sometimes slightly scalloped. The poste-

rior margin of the carapace can be rounded (MPC 25/162

and MPC 25/160), straight (ZIN PH T/M71-1), or slightly

emarginated (PIN 557-133). The free end of the ribs,

when preserved, project far beyond the carapace disc even

in large specimens. Suprascapular fontanelles are present

in small and medium specimens (MPC 25/160, PIN

557-132/1 and PIN 557-133), and absent in the large

specimens (MPC 25/162, ZIN PH T/M71-1 and probably

ZIN PH T/M67-5). All specimens have type A sculptur-

ing. See Table 2 for variation in some carapace characters

of Gobiapalone orlovi.

The nuchal is more than four times wider than long (the

nuchal length/width ratio varies from about 7.0 in MPC

25/160 to about 5.0 in MPC 25/162). Its anterior margin is

not emarginated, and more (MPC 25/162) or less (MPC

25/160) rounded. As is visible in ventral view of MPC 25/

160, the costiform processes are united. The posterolateral

aspects of the nuchal are covered by the expansions of the

free ribs of costals 1.

The neural series is fully preserved in all specimens

except PIN 557-133, in which only the presence and shape

of neurals 1, 3, 5–7 can be reconstructed based on the

shape of adjacent costals. The preneural is absent in all

specimens. In most specimens neurals 1–4 are hexagonal

short-sided posteriorly, neural 5 is tetragonal (isometric),

neural 6 is hexagonal short-sided anteriorly, and neural 7

is pentagonal. A small diamond-shaped neural 8 is present

in MPC 25/162, where it is separated from neural 7 by the

short midline contact between costals 7. In MPC 25/160,

neural 9 is located at the posterior margin of the carapace.

It is semilunate in shape (rounded anteriorly) and is sepa-

rated from neural 7 by the midline contact of costals 7. In

ZIN PH T/M67-5, only neurals 1–3 are hexagonal short-

sided posteriorly, and neural 4 is isometric. In MPC 25/

160 and MPC 25/162, the neural series seems to be

slightly wider than in the holotype. In all specimens,

where observable, neural 1 is the longest. Its lateral bor-

ders are almost straight in the holotype and MPC 25/162,

but they are distinctly concave in MPC 25/160 and ZIN

PH T/M67-5. Neural 7 varies in size and shape, being the

largest and the longest in MPC 25/160, relatively small

Table 2. Variation of some carapace characters in Gobiapalone orlovi.

Characters PIN 557-133 PIN 557-132/1 ZIN PH T/M67-5 MPC 25/160 MPC 25/163 ZIN PH T/M71-1

Carapace length (mm) 155� 240� ? 275 290 335Suprascapular fontanelles Present Present Probably absent Separated Absent AbsentNumber of neurals 7 At least 7 ? 8 8 8Costals 8 Present Present ? Present Present PresentIsometric neural 5 5 4 5 5 5

�Estimation.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

and short in MPC 25/162, and even smaller in the

holotype.

Eight pairs of costals are present in specimens where

the posterior part of the carapace is preserved. The poste-

rior borders of costals 1 are almost perpendicular to the

midline in the holotype and MPC 25/162, but they are

curved anterolaterally in MPC 25/160 and ZIN PH T/

M67-5. Costals 6 are the longest laterally (at the free mar-

gin) of the costals. In the holotype, only the left costal 8 is

present. It is small, wider than long and probably con-

tacted its counterpart at the midline. In MPC 25/160 cos-

tals 8 are much smaller than in the holotype and are

separated from each other by the midline contact of cos-

tals 7. In MPC 25/162, costals 8 are larger than in the

holotype, and almost as long as wide. Depressions on cos-

tals 8 for contact of ilia are absent.

Plastron. Plastral remains consist of an almost complete

plastron (with a broken left hyo- and hypoplastra) of MPC

25/160, and a fragment of the left hyo- and hypoplastra

(ZIN PH T/M46-1). The bridge is short (the ratio of mini-

mal bridge length to maximal hypoplastron length is about

50%). The sculpturing of the plastron is less distinct than

those of the carapace. In MPC 25/160, it is most strongly

expressed in the lateral part of the hyo- and hypoplastra

and fades away in other parts of the plastron. In ZIN PH

T/M46-1, the sculpturing is similar to that on the carapace

in the lateral part of the specimen, but changes and con-

sists of tubercles and ridges in its central and medial parts.

There are five plastral callosities: one on the entoplastron,

one on each hyo- and hypoplastra, and one on each xiphi-

plastra (MPC 25/160).

The epiplastron is J-shaped and slender. The length of

the anterior process is about 50% of the length of the pos-

terior process and about 0.2 times the total width of the

hypoplastron. The epiplastron bears no callosity.

The entoplastron is V-shaped and slender. Each arm of

the entoplastron contacts the medial processes of the hyo-

plastron, but they are not sutured to the hyoplastron and

there is no hyoplastral shoulder to accommodate any

extensive entoplastron contact. The arms meet at about an

80� angle at a rounded anterior tip. The entoplastral cal-

losity is very small, rounded, and restricted to the middle

part of the plate.

The hyoplastron and hypoplastron are connected by a

suture and are not fused. The medial hypoplastral pro-

cesses could have met at the midline, but there was no

extensive midline contact between the hyoplastra or hypo-

plastra. The callosity covers part of the external surface of

the hyoplastron and hypoplastron. The medial, lateral and

posterior processes are left exposed. As is visible in MPC

25/160, the lateral hyoplastral lobe is almost equal to the

medial hyoplastral lobe in length. There are two lateral

hyoplastral processes in both specimens and five medial

hyoplastral processes (visible in MPC 25/160). The

hypoplastron has two lateral processes in both specimens.

As is visible in MPC 25/160, the medial hypoplastral pro-

cesses are divided into two groups (anterior and posterior)

separated by a short gap. The anterior group includes two

processes on the right hypoplastron and probably had

three processes on the left one. The anterior processes are

orientated roughly perpendicular to the midline. The pos-

terior group includes four processes directed posterome-

dially. The hyoplastron and hypoplastron of MPC 25/160

are different from those of ZIN PH T/M 46-1 in their

more emarginated lateral margin (between lateral pro-

cesses) and, as a result, in their slender lateral lobes, and

sculpturing (see above).

At the hypo-xiphiplastral contact, the xiphiplastral pro-

cess is lateralmost. There is no extensive midline contact

between the xiphiplastra. The right xiphiplastron of MPC

25/160 has three anteromedial processes, whereas the left

one has only two such processes. The posteromedial pro-

cesses of the xiphiplastra are singular. The lateral margin

of the xiphiplastra has a slight emargination. Most of the

external surfaces of both xiphiplastra, except for the pro-

cesses, are covered by callosities.

Nonshell postcranials. MPC 25/160 has preserved most

cervical, thoracic, sacral and caudal vertebrae, pectoral

and pelvic girdles, and limb bones.

The cervical vertebrae of MPC 25/160 are represented

by isolated fragments of cervical 1, complete cervicals

2–4, a partially preserved cervical 6 and complete cervi-

cals 7 and 8. Similar to other trionychids (see Meylan

1987), cervical centra 2–7 are opisthocoelous, cervical

centrum 8 has no posterior articular surface for contact

with thoracic vertebra 1 and a ventral process on cervical

8 is absent. There are no strong dorsal processes on the

cervicals, although there are small ridges along the central

axis. A ventral keel on cervical 8 is absent.

The first nine thoracic vertebrae are firmly attached to

the carapace in MPC 25/160. One more vertebral centrum

(10), connected with neural 9 (see above) is preserved

posterior to thoracic vertebra 9. There are two sacral cen-

tra and two pairs of isolated sacral ribs that became

detached during preparation of the specimen.

MPC 25/160 has a continuous series of caudal vertebrae

1–13 from the anterior part of the tail. They are procoe-

lous and steadily decrease in size towards the posterior

end of the tail.

The pectoral girdle of MPC 25/160 is represented by an

articulated right scapula (with the acromion process bro-

ken off) and partial coracoid, and fragments of the left

scapula and coracoid. It is not clear if the coracoid is the

longest of three pectoral processes, because the coracoid

is missing its distal end and the acromion process is bro-

ken off. The coracoid is flat and wide, as in other triony-

chids, with a slightly concave anterior margin and convex

posterior margin. The angle between the acromion process


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

and the main body of the scapula and the angle between

the coracoid and the acromion process cannot be mea-

sured because the acromion process is broken off.

Only the right humerus is preserved in MPC 25/160. It

has a morphology typical for trionychids, including an

open ectepicondylar groove and the lesser trochanter of

the humerus lying in an anteroposterior plane that runs

through the main axis of the humerus (see Meylan 1987).

A single radius and ulna are preserved in MPC 25/160. It

is unclear whether they are from the right or the left forelimb.

The pelvic girdle is represented in MPC 25/160 by a

complete right half of the girdle and an isolated left ilium.

The epipubic region is not present and presumably did not

ossify. The pubis and ischium do not divide the thyroid

fenestra. The pectineal process is about the same width as

the medial border of the pubis (interpubic suture); they lie

in a single plane. The ilia curve posteriorly but not medi-

ally. The distal ends are not expanded. Both the thelial

process and the ilioischial notch are absent. The ischium

does not extend into the thyroid fenestra. A metischial

process is present.

Both femora are preserved in MPC 25/160, although

the distal part is broken off the left femora. Their mor-

phology is typical for trionychids in that both trochanters

lie at a wide angle to the anteroposterior plane that runs

through the main axis of the femur (Meylan 1987).

A single tibia and fibula are preserved in MPC 25/160.

It is unclear whether they are from the right or the left

hind limb.

MPC 25/160 has a number of disarticulated elements of

the autopodium of the fore- and hind limbs. It is unclear

how many clawed digits were in the manus or pes, or if

manus or pes digits display hyperphalangy.

Remarks. All specimens referred here to Gobiapalone

orlovi are similar to the holotype and/or to the most com-

plete specimen (MPC 25/160) in the following shell char-

acters: type A shell sculpturing, carapace length up to

335 mm, suprascapular fontanelles closed in specimens

larger than 275 mm in carapace length, and general mor-

phology of the hyo- and hypoplastra. In addition, all these

specimens come from the Baynshire Formation. Differen-

ces between these specimens can be explained through

intraspecific (number of neurals, position of an isometric

neural, shape of costals 8), and ontogenetic (presence/

absence of the suprascapular fontanelles) variation that

characterizes other species of Trionychidae (Gardner &

Russell 1994). In addition to ZIN PH T/M46-1 (consisting

of the lateral portion of the hyo- and hypoplastra and two

cervical vertebrae), which was primarily reported as

Amyda sp. cf. A. orlovi (Khosatzky 1999), and later

assigned to ‘Amyda’ orlovi (Danilov & Vitek 2009,

2012a), here we attribute several other specimens to this

species. Some of those specimens (MPC 25/162, PIN

557-133 and carapace fragment of ZIN PH T/M46-1)

have not been reported previously, whereas others (MPC

25/160, ZIN PH T/M67-5 and ZIN PH T/M71-1), have

been reported under different names. MPC 25/160 was

reported as Apalonina (Suzuki 2005) and Apalonina indet

3. (Danilov & Vitek 2012a). ZIN PH T/M67-5 was origi-

nally reported as “Trionychidae possibly representing a

new species and characterized by having a rather long

neural 1 that is widened anteriorly and posteriorly, costals

1 that considerably turn distally to the anterior, and an iso-

metric neural 4” (Khosatzky 1999, pp. 144–5). Later it

was mentioned as Trionychinae indet. 4 (Danilov & Vitek

2009; 2012a; Danilov et al. 2011a) or Trionychinae indet.

(Danilov et al. 2011b). ZIN PH T/M71-1 was originally

reported as Trionychina indet. (Khosatzky 1999), and later

mentioned as Trionychinae indet. 3 (Danilov & Vitek

2009, 2012a).

Gobiapalone breviplastra sp. nov.

(Figs 10–12)

1972 Trionyx sp. b; M»ynarski & Narmandach: 100.

1978 Platypeltis sp.; Merkulova: 156.

1999 Trionychidae indet.; Khosatzky: 145, fig. 4.

1999 Apalonini; Khosatzky: 147, fig. 5.

2009 Trionychinae indet. 5; Danilov & Vitek: 54.

2009 Trionychinae indet. 6 (part.); Danilov &

Vitek: 54.

2009 Apalonina indet.; Danilov & Vitek: 55.

2011a ‘Amyda’ menneri: Danilov et al.: 15 (part.).

2011b ‘Amyda’ menneri: Danilov et al.: 96 (part.).

2012a Apalonina indet. 1; Danilov & Vitek: 421,

fig. 23.2a.

2012a Apalonina indet. 2; Danilov & Vitek: 421.

2012a Trionychini indet. 2; Danilov & Vitek: 425,

fig. 23.2f.

2012a Trionychinae indet. 5; Danilov & Vitek: 429

(part.).

Derivation of name. Brevis, Greek for short, and

plastra, in reference to short anterior processes of

epiplastra.

Holotype. PIN 4694-3, partial shell, from Ulan Khushu

(¼ Ulan Bulak), Umunugovi Aimag, Mongolia, Nemegt

Formation, Maastrichtian (collected by JSMPE).

Referred material. MPC 25/163 (HMNS 94-10-24), par-

tial disarticulated skeleton, from Bugin Tsav, Umunugovi

Aimag, Mongolia, Nemegt Formation, Maastrichtian (col-

lected by JMJPE in 1994); MPC 25/164 (HMNS 98-16-

499), partial skeleton, from Bugin Tsav II, Umunugovi

Aimag, Mongolia, Nemegt Formation, Maastrichtian (col-

lected by JMJPE in 1998); PIN 551-461, partial juvenile

shell with fragments of non-shell postcrania, from Nem-

egt, Umunugovi Aimag, Mongolia, Nemegt Formation,

Maastrichtian (collected by MPE in 1949); PIN 4692-301,


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

partial juvenile skeleton (collected by JSMPE in 1991),

PIN 4692-302, partial disarticulated skeleton (collected

by JSMPE in 1974), PIN 4692-303, partial shell (collected

by A. V. Sochava in 1968), all from Bugin Tsav, Umunu-

govi Aimag, Mongolia, Nemegt Formation, Maastrich-

tian; PIN 4694-1, partial entoplastron (collected by

JSMPE in 1970), PIN 4694-2, partial carapace (collected

by MPE), all from Ulan Khushu ( ¼ Ulan Bulak), Umunu-

govi Aimag, Mongolia, Nemegt Formation, Maastrich-

tian; PIN 5505-1, partial carapace, from Alak Shand

Khuduk, Mongolia, Nemegt Formation, Maastrichtian

(collected by JSMPE in 1974); ZIN PH 1/158, partial car-

apace on a cast of the internal cavity of the shell, from

Altan Ula I, Umunugovi Aimag, Mongolia, lower part of

the Nemegt Formation, Maastrichtian (collected by N. N.

Verzilin in 1976); ZIN PH T/M67-6, partial carapace,

from Bugin Tsav, Umunugovi Aimag, Mongolia, Barun-

goyot Formation, Campanian (collected by A. V. Sochava

in 1967); ZIN PH T/M68-1, a partial carapace, from

Nogon Tsav, Bayankhongor Aimag, Mongolia, upper part

Figure 10. Shell material of Gobiapalone breviplastra gen. et sp. nov. from the Nemegt Formation of: A–E, Nemegt; F–I, Ulan Khushu(¼ Ulan Bulak); J, Altan Ula; I, K, Alak Shand Khuduk; and L, from the upper part of the Barungoyot Formation of Nogon Tsav. A, B,PIN 551-461, a partial juvenile shell with fragments of non-shell postcrania in: A, dorsal and B, ventral views. C, ZPAL MgCh/80, pos-terior carapace fragment in dorsal view. D, ZPAL MgCh/81, left hypoplastron in ventral view. E, ZPAL MgCh/83, partial left hyo- andhypoplastra in ventral view. F, G, PIN 4694-3, a partial shell (holotype): F, carapace in dorsal view; G, plastron in ventral view. H, PIN4694-1, partial entoplastron in ventral view. I, PIN 4694-2, carapace in dorsal view. J, ZIN PH 1/158, partial carapace on a cast of theinternal cavity of the shell. K, PIN 5505-1, partial carapace in dorsal view. L, ZIN PH T/M68-1, partial carapace in dorsal view.Reconstructed sutures and outlines are indicated by dashed lines.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

Figure 11. Shell material of Gobiapalone breviplastra gen. et sp. nov. from the: A–D, F–M, Nemegt; and E, Barungoyot formations ofBugin Tsav. A–D, MPC 25/163; A, B, carapace in A, dorsal and B, ventral views; C, D, plastron in C, ventral and D, dorsal views. E, ZINPH T/M67-6, carapace in dorsal view. F, G, MPC 25/164; F, carapace in dorsal view; G, plastron in ventral view. H, I, PIN 4692-301,juvenile shell with fragments of non-shell postcrania in H, dorsal and I, ventral views; J, K, PIN 4692-302; J, carapace in dorsal view; K,plastron in ventral view. L, M, PIN 4692-303: L, carapace in dorsal view; M, right hyo- and hypoplastra in ventral view. Reconstructedsutures and outlines are indicated by dashed lines.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

of Barungoyot Formation, Campanian (collected by A. V.

Sochava in 1968); ZPAL MgCh/80, posterior carapace

fragment, including neurals 6 and 8 and medial parts of

costals 6 and 7 and additional costal fragment, ZPAL

MgCh/81, left hypoplastron, ZPAL MgCh/83, partial left

hyo- and hypoplastra missing most of their medial por-

tions, additional shell fragments and bones of non-shell

postcrania, all from Nemegt, Umunugovi Aimag, Mongo-

lia, Nemegt Formation, Maastrichtian (collected by

PMPE).

Diagnosis. A species of Gobiapalone with estimated car-

apace length up to 260 mm, which can be differentiated

from G. orlovi by the closed suprascapular fontanelles in

specimens larger than 224 mm in carapace length, more

robust epiplastra and entoplastron, a larger (about 100�)angle between the arms of the entoplastron, fewer (two or

three) medial processes of the hyoplastron, a smaller gap

between the anterior and posterior groups of the medial

processes of hypoplastron, and more robust xiphiplastra.

Description

Carapace. The carapace disc is longer than wide in MPC

25/164 and PIN 4692-301, as long as wide in PIN 4694-3

and, probably, PIN 551-461 and ZIN PH T/M68-1, and

wider than long in MPC 25/163, PIN 4692-302, PIN

4694-2 and ZIN PH T/M67-6. The anterior border of the

carapace is rounded and some specimens have a weak

nuchal emargination. The lateral borders of the carapace

are straight, or sometimes slightly scalloped. The poste-

rior margin of the carapace is rounded (PIN 551-461),

straight (MPC 25/163, PIN 4692-301, PIN 4692-302, PIN

4694-3, ZIN PH T/M67-6 and ZIN PH T/M68-1) or more

or less emarginated (PIN 4694-2 and PIN 5505-1). The

free ribs of the costals, when preserved, project far beyond

the carapace disc even in large specimens. The suprascap-

ular fontanelles are present and confluent in smaller speci-

mens (PIN 551-461, PIN 4692-301, PIN 4694-2, PIN

4694-3 and PIN 5505-1), present but separated from each

other in a larger one (MPC 25/164) and absent in the larg-

est specimens (MPC 25/163 and PIN 4692-303). The

exception is PIN 4692-302, which has a medium carapace

length, but has no suprascapular fontanelles. This may

indicate some variation in rates of shell growth potentially

connected to sexual dimorphism. All specimens have type

A shell sculpturing. See Table 3 for variation of some car-

apace characters of Gobiapalone breviplastra.

The nuchal is more than four times wider than long (the

nuchal length/width ratio varies from about 7.0 in MPC

25/163 to about 5.0 in PIN 5505-1). Its anterior margin is

either not emarginated (MPC 25/164, PIN 4692-303, PIN

4694-3, and ZIN PH T/M67-6) or weakly emarginated

(PIN 4692-302 and PIN 5505-1). As is visible in ventral

view of MPC 25/163, the costiform processes are united.

The posterolateral aspects of the nuchal are covered by

the expansions of costals 1, and free ribs of costals 1 lie

posterior to the nuchal/costals 1 suture. In other speci-

mens, the posterolateral aspects of the nuchal are covered

by the expansions of the free ribs of costals 1.

A complete neural series is present in MPC 25/163, PIN

551-461, PIN 4692-301, PIN 4692-302, PIN 4694-2, PIN

4694-3, ZIN PH T/M67-6 and ZIN PH T/M68-1. In MPC

25/164, only neurals 1–7 are preserved, and the shape of

neural 8 can be reconstructed based on adjacent elements.

In PIN 5505-1, all neurals are preserved, except 4 and 5,

Figure 12. Non-shell postcrania of Gobiapalone breviplastra gen. et sp. nov., MPC 25/163, from the Nemegt Formation of Bugin Tsav.A, cervical vertebrae in dorsal view; B, left scapula in lateral view; C, left coracoid in dorsal view; D, pelvic girdle in dorsal view.Reconstructed outlines are indicated by dashed lines.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

whose shape can be reconstructed based on adjacent ele-

ments. In PIN 4692-303, only neurals 1–4 are preserved.

In ZPAL MgCh/80, part of neural 6 and complete neural

8 are preserved, and the shape of neural 7 is reconstructed

based on adjacent elements. In total, specimens may have

seven (PIN 4692-302, PIN 4694-2, ZIN PH T/M67-6 and

ZIN PH T/M68-1), eight (MPC 25/163, MPC 25/164, PIN

4692-301, PIN 4694-3, PIN 5505-1 and ZPAL MgCh/80)

or nine (PIN 551-461) neurals. The preneural is absent in

all specimens. In all specimens, where observable, neurals

1–4 are hexagonal short-sided posteriorly, neural 5 is

tetragonal (isometric), and neural 6 is hexagonal short-

sided anteriorly. Neural 1 is longer than the other neurals

in most specimens, except PIN 4692-301 and ZIN PH T/

M68-1. The lateral borders of neural 1 are straight (PIN

4692-301, PIN 4694-3 and ZIN PH T/M68-1) or more or

less concave (other specimens). Neural 1 is either nar-

rowed anteriorly (most specimens) or as wide anteriorly

as posteriorly (largest specimens: MPC 25/163 and PIN

4692-303). Neural 7 is either pentagonal (MPC 25/163,

PIN 551-461, PIN 4692-301, PIN 4692-302, PIN 4694-2,

PIN 4694-3, ZIN PH T/M68-1) or hexagonal short-sided

anteriorly (MPC 25/164, PIN 5505-1, ZPAL MgCh/80).

Neural 8 is the most variable in shape and position. In

PIN 551-461, neural 8 is hexagonal and located between

costals 7 and 8, separated from neural 7 by short midline

contact of costals 7, and contacts neural 9 posteriorly. In

MPC 25/163, neural 8 is small and rounded, located

between costals 7 and 8, separated from neural 7 by long

midline contact between costals 7 and contributes to the

free margin of the carapace posteriorly. In PIN 4694-3,

neural 8 is large and almost triangular, located between

costals 7, separated from neural 7 by short midline contact

between costals 7 and contributes to the free margin of the

carapace posteriorly. In PIN 5505-1, neural 8 is irregularly

tetragonal in shape, located between costals 7, contacts

neural 7 anteriorly and was probably separated from the

free margin of the carapace by the midline contact of cos-

tals 7 or 8. In ZPAL MgCh/80, neural 8 is elongated and

tetragonal in shape, located between costals 7, contacts

neural 7 anteriorly and contributes to the free margin of the

carapace posteriorly. In PIN 4692-301, neural 8 is very

small and rounded, located between costals 7,

contacts neural 7 anteriorly and is separated from the

free margin of the carapace by the short midline contact of

costals 7. In MPC 25/164, neural 8 is not preserved, but

was probably tetragonal, located between costals 7, and

contacted neural 7 anteriorly and costals 8 posteriorly.

Seven (PIN 4692-301, PIN 4692-302, PIN 4694-3 and

ZIN PH T/M68-1) or eight (MPC 25/163, PIN 551-461

and ZIN PH T/M67-6) pairs of costals are present. Costals

6 and 7 make up the posterior margin of the carapace of

ZPAL MgCh/80. Costals 8 are reconstructed in MPC 25/

164, PIN 4694-2 and PIN 5505-1. The posterior border of

costals 1 curve anterolaterally in all specimens. Costal 1 is

either almost as long laterally as it is medially (most speci-

mens) or considerably shorter laterally than it is medially

(MPC 25/164 and PIN 4694-2). All other costals are usu-

ally longer laterally than medially. However, in PIN 4692-

303, costal 2 is very long laterally, whereas costal 3 is as

long laterally as medially. Costals 6 are much longer lat-

erally than medially, form the posterolateral edge of the

carapace and usually contribute to its posterior edge. Cos-

tals 7 are either large triangular-shaped elements which

form most of the posterior carapace margin (PIN 4692-

301, PIN 4692-302, PIN 4694-3, ZIN PH T/M68-1 and

ZPALMgCh/80), or smaller elements that are notched pos-

teriorly by costals 8, and make a smaller contribution to the

posterior carapace margin (MPC 25/163, MPC 25/164,

PIN 551-461, PIN 4694-2, PIN 5505-1 and ZIN PH T/

M67-6). Costals 7 usually contact each other along the

midline posterior to neurals 7 or 8, except MPC 25/164

and ZIN PH T/M67-6, in which costals 7 are separated by

neural 8 and costals 8 respectively. Costals 8, when pres-

ent, are always small, triangular-shaped elements, which

contribute to the posterior carapace margin. They are either

separated from each other by neurals 8 and 9 (PIN 551-

461), or only by neural 8 (MPC 25/163), or by costals 7

(PIN 5505-1 and PIN 4694-2), or contact each other along

the midline (MPC 25/164 and ZIN PH T/M67-6).

Table 3. Variation of some carapace characters in Gobiapalone breviplastra gen. et sp. nov.

CharactersPIN

4692-301PIN

551-461PIN

4692-302PIN

5505-1ZIN PHT/M67-6

PIN4694-3

PIN4694-2

MPC25/164

ZIN PHT/M68-1

MPC25/163

ZIN PH4692-303

Carapacelength (mm)

115 147� 185 193 196 197 207� 224� 230� 241 260�

Suprascapularfontanelles

Confluent ? Absent Confluent ? Confluent Confluent Separated ? Absent Absent

Number ofneurals

8 9 7 8 7 8 7 Probably 8 7 8 ?

Costals 8 Absent Present Absent Present Present Absent Probablypresent

Present Absent Present ?

�Estimation.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

Plastron. Plastral remains are represented by an almost

complete plastron missing the left epiplastron and lateral

portions of the entoplastron (MPC 25/163), a partial plas-

tron represented by the left hyo- and hypoplastra, a medial

portion the of right hyo- and hypoplastra, and central parts

of both xiphiplastra (MPC 25/164), a partial plastron rep-

resented by the left epiplastron, the entoplastron, medial

portions of both hyo- and hypoplastra, and fragments of

both xiphiplastra (PIN 4694-3), a partial plastron repre-

sented by the entoplastron, medial portions of both hyo-

and hypoplastra, and a fragment of the left xiphiplastron

(PIN 4692-301), a partial plastron represented by the left

epiplastron, a partial entoplastron, the right hyoplastron

and medial portion of the left hyo- and hypoplastra (PIN

4692-302), a partial plastron represented by the posterior

portion of the right epiplastron, the entoplastron, both

hyo- and hypoplastra and both xiphiplastra (PIN 551-

461), a partial entoplastron missing the left lateral portion

(PIN 4694-1), a partial right hyo- and hypoplastra (PIN

4692-303), a partial left hypoplastron (ZPAL MgCh/81),

and a partial right hyo-hypoplastron (ZPAL MgCh/83).

The bridge is short (the ratio of minimal bridge length to

maximal hypoplastron length is about 50%). The sculptur-

ing of the plastron is less distinct than that of the carapace

and is most strongly expressed on the hyo- and hypoplas-

tra. There are five plastral callosities: one on the entoplas-

tron, two on the hyo- and hypoplastra and two on the

xiphiplastra.

The epiplastron is J-shaped and more robust than in

Gobiapalone orlovi. The length of the anterior process is

about 45% of the length of the posterior process and about

0.2 times the total width of the hypoplastron. The epiplas-

tron bears no callosity.

The entoplastron is V-shaped, and more robust than in

Gobiapalone orlovi. Each arm of the entoplastron contacts

the medial processes of the hyoplastron, but they are not

sutured to the hyoplastron and there is no hyoplastral

shoulder to accommodate any extensive entoplastron con-

tact. The arms meet at about a 100� angle with a rounded

(PIN 551-461 and PIN 4694-3), pointed (PIN 4692-302

and PIN 4694-1) or truncated (MPC 25/163) anterior tip.

The entoplastral callosity is very small, rounded, and

restricted to the middle part of the bone.

The hyoplastron and hypoplastron are connected by a

suture and are not fused. The medial hypoplastral pro-

cesses could have met at the midline, but there was no

extensive midline contact between the hyoplastra or hypo-

plastra. The callosity covers part of the external surface of

the hyoplastron and hypoplastron, except for the medial,

lateral and posterior processes, as is visible in MPC 25/

163. In small specimens (PIN 551-461 and PIN 4692-

301), the callosity is relatively small and covers only the

central part of the hyoplastron and hypoplastron. As is vis-

ible in MPC 25/163, the lateral hyoplastral lobe is slightly

shorter than the medial hyoplastral lobe. There are two

lateral hyoplastral processes and two or three medial

hyoplastral processes. The hypoplastron has two lateral

processes. As is visible in MPC 25/163 and PIN 4694-3,

the medial hypoplastral processes are divided into two

groups (anterior and posterior) separated by a very short

gap (shorter than in Gobiapalone orlovi). The anterior

group includes two (MPC 25/163 and PIN 4694-3) or

three (PIN 551-461) processes, which are almost perpen-

dicular to the midline. The posterior group includes from

three to four processes directed posteromedially. One

hypoplastron (ZPAL MgCh/81; Fig. 10D), about the same

size as MPC 25/163, demonstrates more complete cover-

age of its medial and lateral lobes by the callosity.

At the hypo-xiphiplastral contact, the xiphiplastron is

lateralmost. There is no extensive midline contact

between the xiphiplastra. The right xiphiplastron of MPC

25/163 has one anteromedial processes, whereas the left

one has two such processes. The posteromedial processes

of the xiphiplastra are singular. The lateral margin of the

xiphiplastra has a very slight emargination (smaller than

in Gobiapalone orlovi). The majority of the xiphiplastra,

except the processes, is covered by a callosity.

Nonshell postcranials. MPC 25/163 has complete cervi-

cal vertebrae IV–VI and an anterior fragment of cervical

vertebra VIII. It also has thoracic, sacral and several cau-

dal vertebrae, the pectoral and pelvic girdles, both humeri

and part of the right(?) hind limb in articulation in matrix.

MPC 25/164 has cervical vertebrae IV–VI and VIII, frag-

ments of pectoral and pelvic girdles, a complete left

humerus and fragments of other limb bones. PIN 4692-

301 has the pectoral and pelvic girdles and a fragment of

the right humerus. PIN 4692-302 has cervical vertebra IV,

the pectoral girdle, and two humeri. PIN 551-461 has the

anterior caudal vertebrae, the pectoral and pelvic girdles,

and bones of the hind limb autopodium preserved.

The morphology of the available cervical vertebrae is

similar to that of Gobiapalone orlovi.

The first 11 thoracic vertebrae are attached to the cara-

pace in MPC 25/163. There is one sacral centrum with a

pair of sacral ribs that became detached during the prepa-

ration of the specimen.

MPC 25/163 has five anterior caudal vertebrae pre-

served, one of which is isolated and four of which are in

articulation; four more caudal vertebrae of the same speci-

men have been preserved in the piece of matrix.

As is visible in MPC 25/163, the coracoid and the main

body of the scapula are broken off distally and it is not

clear what pectoral process is the longest. The coracoid is

flat and wide, as in other trionychids, with a slightly con-

cave anterior margin and convex posterior margin. The

angle between the acromion process and the main body of

the scapula is about 60�, but the angle between the cora-

coid and the acromion process cannot be measured.

The morphology of the available humeri is similar to

that of Gobiapalone orlovi and other trionychids.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

The morphology of the pelvic girdle and hind limb

(as seen in MPC 25/163) is similar to that of Gobiapalone

orlovi, as well as most other trionychids.

Remarks. All specimens referred here to Gobiapalone

breviplastra are similar to the holotype and the most com-

plete specimen (MPC 25/163) in the following shell char-

acters: type A shell sculpturing, carapace length up to

260 mm, suprascapular fontanelles usually closed in

specimens larger than 224 mm in carapace length, and

morphology of the plastral elements (see Diagnosis). In

addition, all these specimens come from the Barungoyot

and Nemegt formations. Differences between the speci-

mens are explained by intraspecific (shape of the cara-

pace, number of neurals, shape of neurals 1, 7 and 8,

number of costals, shape and contacts of costals 7 and 8,

shape of the anterior tip of entoplastron, number of medial

processes of the hyo- and hypoplastra) and ontogenetic

(presence/absence of the suprascapular fontanelles and

degree of ossification of hyo- and hypoplastra) variation

noted previously (Gardner & Russell 1994). Some of

these specimens (MPC 25/163, MPC 25/164, PIN 4692-

301, PIN 4692-302, PIN 4692-303, PIN 4694-1, PIN

4694-2, PIN 5505-1, ZIN PH 1/158) have not been

reported previously, whereas others (PIN 551-461, ZIN

PH T/M67-6, ZIN PH T/M68-1, ZPAL MgCh/80, ZPAL

MgCh/81 and ZPAL MgCh/83) have been reported under

different names. PIN 551-461 was originally reported as

Platypeltis sp. (Merkulova 1978), and later mentioned as

Apalonina indet. (Danilov & Vitek 2009) and Apalonina

indet. 2 (Danilov & Vitek 2012a). ZIN PH T/M67-6 was

originally reported as Trionychidae indet. (Khosatzky

1999), and later mentioned as Trionychinae indet. 5 (Dan-

ilov & Vitek 2009) and Trionychini indet. 2 (Danilov &

Vitek 2012a). ZIN PH T/M68-1 was originally reported

as Apalonini (Khosatzky 1999), and later mentioned as

Apalonina indet. (Danilov & Vitek 2009) and Apalonina

indet. 1 (Danilov & Vitek 2012a). ZPAL MgCh/80,

ZPAL MgCh/81 and ZPAL MgCh/83 were originally

reported as ‘Trionyx sp. b’ (M»ynarski & Narmandach

1972), and later erroneously considered to be part of Trio-

nychinae indet. 6 (Danilov & Vitek 2009) and Trionychi-

nae indet 5 (Danilov & Vitek 2012a). Other specimens

assigned to ‘Trionyx sp. b’ (ZPAL MgCh/88, about 20

fragments of neurals and costals; ZPAL MgCh/89, six

plastral fragments; M»ynarski & Narmandach 1972) can-

not be identified below the family level and should be

considered Trionychidae indet.

Gobiapalone sp.

(Fig. 16A)

Referred material. MPC 25/165 (HMNS 97-21-55),

incomplete carapace, from Khermin Tsav, Umunugovi

Aimag, Mongolia, Lower White Beds, ?Baynshire Forma-

tion, Cenomanian–Santonian.

Description. This carapace is poorly preserved and dam-

aged. Its length is estimated to be about 230 mm. The car-

apace is slightly wider than long. Its anterior border is

rounded and has no nuchal emargination. Suprascapular

fontanelles are present and separated by contact between

the nuchal and neural 1. The number and shape of the neu-

rals is unclear. Seven pairs of costals are observable, but

the presence or absence of costals 8 is unclear. The speci-

men has type A shell sculpturing.

Remarks. This specimen is assigned to Gobiapalone

based on its similar size, carapace shape, presence of

suprascapular fontanelles and sculpture pattern. The spec-

imen is missing characters that would allow for identifica-

tion to species level.

Trionychinae incertae sedis

‘Trionyx’ baynshirensis sp. nov.

(Fig. 13A, B)

Derivation of name. Species name is from Bayn Shire

locality.

Holotype. PIN 557-134 (formerly PIN 557-130), medial

fragment of right hyo- and hypoplastra, from Bayn Shire,

Dornogovi Aimag, Mongolia, Baynshire Formation, Cen-

omanian–Santonian (collected by MPE in 1948).

Diagnosis. A trionychine, with an estimated carapace

length of about 500 mm, that is most similar amongst Cre-

taceous trionychids to ‘Trionyx’ kansaiensis in the shape

of the hyo- and hypoplastra and the shell sculpturing of

type B. It can be differentiated from ‘Trionyx’ kansaiensis

in that its medial hyoplastral and hypoplastral processes

are completely covered by a callosity at smaller size, and

the hyo- and hypoplastra has straight medial borders. It

can be differentiated from Gobiapalone orlovi (another

trionychid known from the Baynshire Formation) by its

larger size, shell sculpturing, callosity that completely

covers the medial hyoplastral and hypoplastral processes,

and straight medial borders of the hyo- and hypoplastra.

Description. PIN 557-134 belongs to an individual with

an estimated carapace length of about 500 mm, based on

comparison with ‘Trionyx’ kansaiensis. The specimen has

type B sculpturing. The suture between hyo- and hypo-

plastra is not discernable due to poor preservation, and

may be absent. The hyo- and hypoplastra have straight

medial borders. They could probably contact their coun-

terparts across the midline, but there is no evidence of a

midline suture. There were two medial fontanelles

between the hyo-and hypoplastra: the anterior one

between anterior halves of the hyoplastra, and the poste-

rior one between posterior halves of the hypoplastra. The

medial hyoplastral and hypoplastral processes are almost

completely covered by a callosity and their precise num-

ber is unclear.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

‘Trionyx’ gilbentuensis sp. nov.

(Fig. 13C, D)

1999 ‘Plastomenus’; Khosatzky: 147, fig. 6.

2009 Trionychinae indet. 6; Danilov & Vitek: 54 (part.).

2011b Trionychinae indet.; Danilov et al.: 96.

2012a Trionychidae indet. 5; Danilov & Vitek: 429 (part.),

fig. 23.3g.

Derivation of name. Species name is from Gilbentu

locality.

Holotype. ZIN PH T/M46-2, incomplete left hyo- and

hypoplastra, from Gilbentu, Umunugovi Aimag, Mongo-

lia, Nemegt Formation, Maastrichtian (collected by MPE

in 1946).

Diagnosis. A trionychine, with an estimated carapace

length of about 500 mm, that is most similar, amongst

Cretaceous trionychids, to ‘Trionyx’ baynshirensis in the

shape of the hyo- and hypoplastra, shell sculpturing and in

that its medial hyoplastral and hypoplastral processes are

completely covered by a callosity. It can be differentiated

from ‘T.’ baynshirensis by the convex medial borders of

Figure 13. A, B, ‘Trionyx’ baynshirensis sp. nov., PIN 557-134 (holotype), a medial fragment of right hyo- and hypoplastra from theBaynshire Formation of Bayn Shire in: A, ventral and B, dorsal views. C, D, ‘Trionyx’ gilbentuensis sp. nov., ZIN PH T/M46-2(holotype), incomplete left hyo- and hypoplastra from the Nemegt Formation of Gilbentu in: C, ventral and D, dorsal views.Reconstructed outlines are indicated by dashed lines.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

the hyo-and hypoplastra. It can be differentiated from

other trionychids of the Nemegt Formation by its larger

size and shell sculpturing. In addition, it can be differenti-

ated from Nemegtemys conflata by a sutural connection

between the hyo- and hypoplastra and from Gobiapalone

breviplastra by a callosity that completely covers the

medial hyoplastral and hypoplastral processes, a more

acute angle between the medial and lateral lobes of the

hypoplastron and a more complete covering of the lateral

hypoplastral lobe by a callosity.

Description. ZIN PH T/M46-2 (previously ZIN PHT

M46-2; Khosatzky 1999) consists of the medial part of a

left hyo- and hypoplastra and an additional posterolateral

fragment of the hypoplastron (not reported previously);

the posteromedial part of the hypoplastron is broken off.

The specimen belongs to an individual with an estimated

carapace length of about 500 mm, based on comparison

with ‘Trionyx’ kansaiensis. The specimen has type B

sculpturing. The hyo- and hypoplastra are connected by a

suture and are not fused. They could contact their counter-

parts across the midline only at the medialmost border of

the hyoplastron. The medial fontanelle between the hyo-

and hypoplastra was not hourglass-shaped, as opposed to

most other trionychines. The ratio of the minimum bridge

length to the maximum hypoplastron length is impossible

to estimate due to incompleteness of the specimen. The

medial hyoplastral processes are completely covered by a

callosity and their number is unclear. In ‘Trionyx’ kan-

saiensis, there are three large medial hyoplastral processes,

which are not covered by callosities in both small and large

sizes (Vitek & Danilov 2010). The medial hypoplastral

processes are completely covered by a callosity. Only

three processes, which project slightly beyond the medial

margin of the hypoplastron, are observable. The angle

between the medial and lateral hypoplastral lobes is acute

(75–80�). The posterolateral fragment of the hypoplastron,

if correctly identified, indicates that the lateral hypoplastral

lobe was almost entirely covered in sculpturing.

Remarks. Khosatzky (1999) described this specimen as

a relatively large trionychid with tuberculate sculpture

and assigned it to ‘Plastomenus’. Danilov & Vitek

(2009, 2012a) assigned this specimen to Trionychinae

(as Trionychinae indet. 6 and 5 respectively) based on its

similarities with the trionychine ‘Trionyx’ kansaiensis.

Those authors also noted that ‘Trionyx sp. b’, based on

shell fragments of a large trionychid that came from the

Nemegt Formation of the Nemegt locality (M»ynarski &Narmandach 1972), probably belonged to the same taxon

as ZIN PH T/M46-2. Here, this taxon (‘Trionyx’ gilben-

tuensis) is restricted to the holotype, whereas material of

‘Trionyx sp. b’ is referred to Gobiapalone breviplastra

and Trionychidae indet. (see account of G. breviplastra).

Danilov & Vitek (2012a) mentioned ‘Plastomenus’?

of Khosatzky (1999) amongst the synonymies of

Trionychinae indet. 5. Here, the material on which this

determination was based (a plastron fragment) is consid-

ered as the holotype of Nemegtemys conflata gen. et sp.

nov. (see above).

‘Trionyx’ gobiensis sp. nov.

(Fig. 14)

1971 Trionyx sp. a; Khosatzky & M»ynarski: 141,

pl. 24, fig. 2, text-fig. 7.

1972 Trionyx sp. a; M»ynarski & Narmandach: 100.

1999 ‘Trionyx’ sp.; Khosatzky: 144.

2009 Trionychinae indet. 2; Danilov & Vitek: 54.

2012a Trionychinae indet. 2; Danilov & Vitek: 429,

fig. 23.3d.

Derivation of name. Species name is from Gobi Desert.

Holotype. PIN 4064-2, incomplete carapace, from

Bamba Khuduk (Eastern Sayr), Umunugovi Aimag, Mon-

golia, Nemegt Formation, Maastrichtian (collected by

JSMPE in 1976).

Referred material. ZPAL MgCh/59 (formerly ZPAL

MgCh/52) and ZPAL MgCh/76, two incomplete carapa-

ces, from Tsagan Khushu and Nemegt, respectively;

Umunugovi Aimag, Mongolia, Nemegt Formation, Maas-

trichtian; collection ZPAL MgCh/121, shell fragment,

consisting of the lateral parts of the nuchal and both cos-

tals 1, from Altan Ula III, Umunugovi Aimag, Mongolia,

Nemegt Formation, Maastrichtian (all specimens col-

lected by PMPE); ZIN PH 2/24, left costals 1 and 2, from

Tsagan Khushu, Umunugovi Aimag, Mongolia, Nemegt

Formation, Maastrichtian (collected by MPE in 1948; as

“two very small costals with tuberculate sculpturing”;

Khosatzky, 1999, p. 147).

Diagnosis. A trionychine that can be differentiated from

all other Cretaceous trionychids with known shells

(Table 1) by its smaller size (up to 130 mm). In addition,

it can be differentiated from ‘Aspideretes’ maortuensis

Yeh, 1965 by large costals 8; from Aspideretoides spp. by

the absence of a preneural and the presence of large cos-

tals 8; from Gilmoremys lancensis (Gilmore, 1928) by the

absence of a preneural and the presence of neural reversal

at neural 5; from Gobiapalone spp. by the absence of

suprascapular fontanelles and large costals 8; from

‘Trionyx’ kansaiensis by weak or absent nuchal emargina-

tion; from ‘T.’ shiluutulensis sp. nov. by neural reversal at

neural 5 and large costals 8.

Description. The carapace length is up to 130 mm long.

The carapaces are roughly circular, almost as wide as

long, with a weak or absent nuchal emargination and a

small posterior caudal emargination (visible in ZPAL

MgCh/59). The carapace is widest at costals 4. The free

ribs project far beyond the carapace disc in the smallest


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

specimen (ZPAL MgCh/59) and project very little in

larger specimens (PIN 4064-2, ZPAL MgCh/76). The

suprascapular fontanelles are not present in any specimen.

The sculpturing on the surface of all specimens is similar

to those of Gobiapalone spp. and Aspideretoides spp.

(see above).

The nuchal is more than four times wider than long

(slightly more than four times in PIN 4064-2 and ZPAL

MgCh/59; more than five times in ZPAL MgCh/76). Its

anterior margin is either straight or has a weak emargina-

tion. As is visible in the ventral view of PIN 4064-2, the

costiform processes are united. The posterolateral aspects

of the nuchal are covered by the expansions of costal ribs 1.

The neural series is complete in all three relatively

complete carapaces and consists of seven neurals in PIN

4064-2 and ZPAL MgCh/59 and eight neurals in ZPAL

MgCh/76. The preneural is absent. Neurals 1–4 are hexag-

onal short-sided posteriorly, neural 5 is tetragonal

(isometric), and neural 6 is hexagonal short-sided anteri-

orly. Neural 7 is pentagonal in PIN 4064-2 and ZPAL

MgCh/59 and hexagonal short-sided anteriorly in ZPAL

MgCh/76. Neural 8 is pentagonal and present only in

ZPAL MgCh/76. Neural 1 is variable in shape, either nar-

rowed anteriorly (PIN 4064-2) or not (ZPAL MgCh/76).

Eight pairs of costals are present in three specimens,

although some of them are broken off or damaged. The

costals which are longest at the lateral margin are costals

2 and 5 in PIN 4064-2 and, probably, also in ZPAL

MgCh/76, and costals 3 and 5 in ZPAL MgCh/59. In PIN

4064-2 and ZPAL MgCh/59, costals 7 and 8 contacted

each other at the midline, whereas in ZPAL MgCh/76

only costals 8 have midline contact. Costals 8 are rela-

tively large and are wider than long. As is visible in

ZPAL MgCh/59, they formed most of the posterior cara-

pace margin.

Remarks. Khosatzky & M»ynarski (1971) described

ZPAL MgCh/59 (as ZPAL MgCh/52) as ‘Trionyx sp. a’.

In the schematic drawing of this specimen (Khosatzky &

M»ynarski 1971, text-fig. 7) the carapace had no preneu-

ral, eight neurals and eight pairs of costals. Other speci-

mens attributed to ‘Trionyx sp. a’ (ZPAL MgCh/60,

Figure 14. Shell material of ‘Trionyx’ gobiensis sp. nov. A, B, PIN 4064-2, an incomplete carapace from the Nemegt Formation ofBamba Khuduk in A, dorsal and B, ventral views. C, ZPAL MgCh/59, incomplete carapace from the Nemegt Formation of TsaganKhushu in dorsal view. D, ZPAL MgCh/76, incomplete carapace from the Nemegt Formation of Nemegt in dorsal view. E, ZPALMgCh/121, shell fragment, consisting of lateral parts of the nuchal and both costals 1, from the Nemegt Formation of Altan Ula III indorsal view. F, ZIN PH 2/24, left costals 1 and 2 in dorsal view from the Nemegt Formation of Tsagan Khushu. Reconstructed suturesand outlines are indicated by dashed lines.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

ZPAL MgCh/76 and unnumbered) have never been

described (Khosatzky & M»ynarski 1971; M»ynarski &

Narmandach 1972). Later, Danilov & Vitek (2012a)

reported that judging from a photograph of ZPAL MgCh/

59 (as ZPAL MgCh/52) it is impossible to confirm the

presence of a preneural and neural 8. These authors fur-

ther suggested that if this specimen has no preneural and

only seven neurals, it could be attributed to ‘Amyda’ men-

neri. However, new observations of this and other speci-

mens with similar morphology show that there was no

preneural, and the number of neurals varies between seven

and eight. All these specimens differ from other Creta-

ceous trionychids in multiple features (see Diagnosis

above) and, for that reason, we identify them as a new

species within Trionyx Geoffroy Saint Hillaire, 1809

sensu lato. Isolated costal fragments (ZPAL MgCh/60 and

unnumbered) from Tsagan Khushu, also attributed to

‘Trionyx sp. a’ by Khosatzky & M»ynarski (1971), cannotbe identified below the family level. The statement of

Khosatzky & M»ynarski (1971, p. 142) that “the greatest

number of the remains of these small soft-shelled turtles

are in the collections of Zoological Institute in Leningrad”

(ZIN PH) cannot be confirmed.

‘Trionyx’ shiluutulensis sp. nov.

(Fig. 15)

2011a Aspideretoides; Danilov et al.: 15.

2011b Aspideretoides; Danilov et al.: 96.

Derivation of name. Species name is from the Shiluut

Ula locality.

Holotype. MPC 25/166 (HMNS 97-21-56), carapace,

from Shiluut Ula, Umunugovi Aimag, Mongolia;

unknown formation, Campanian (collected by JMJPE in

1997; for age see Watabe et al. 2010).

Diagnosis. A trionychine with a carapace length of

225 mm that can be differentiated from all other Creta-

ceous trionychids with known carapaces (Table 1) by a

combination of two characters: the presence of both a pre-

neural and eight neurals. In addition, it can be differenti-

ated from ‘Aspideretes’ maortuensis by smaller size and

neural reversal at neural 6; from Aspideretoides spp. by

smaller size; from Gilmoremys lancensis by smaller size

and reduced costals 8; from Gobiapalone spp. by neural

reversal at neural 6; from ‘T.’ gobiensis by larger size,

neural reversal at neural 6 and small costals 8; and from

‘T.’ kyrgyzensis Nessov, 1995 by larger size, neural rever-

sal at neural 6 and small costals 8.

Description. The carapace length is 225 mm long. It is

almost as wide as long, and narrowed anteriorly due to the

strong emargination of its lateral borders between the free

ribs of costals 1 and 2. Posterior to these emarginations,

the lateral borders of the carapace are scalloped due to

smaller emarginations between the free ribs of costals 2

and 3, 3 and 4, 4 and 5, and 5 and 6. The free ribs of cos-

tals project beyond the carapace disc. Suprascapular fon-

tanelles are absent.

The nuchal is more than seven times wider than long.

Its anterior margin is not emarginated. As is visible in

ventral view, the costiform processes are united. The pos-

terolateral aspects of the nuchal are covered by the expan-

sions of the ribs of costals 1.

The neural series included a preneural and eight neurals,

of which the preneural and neurals 1, 2 and 7 are complete,

and neurals 3–6 are partially preserved. Neural 8 is not pre-

served, but its shape can be reconstructed based on adja-

cent bones. The preneural is roughly pentagonal and wider

than long. Its anterior point extends into the nuchal. Neu-

rals 1–5 are hexagonal short-sided posteriorly, neural 6 is

reconstructed as isometric (tetragonal), neural 7 is hexago-

nal short-sided anteriorly, and neural 8 is reconstructed as

pentagonal. Neurals 1 to 7 steadily decrease in width

towards the posterior end of the carapace.

Eight pairs of costals are present. The medial border of

costal 1 is divided into two parts, a short one for contact

with the preneural and a long one for contact with neural

1. The suture between costals 1 and 2 curves anterolater-

ally. The lateral border of costal 2 is much longer on the

right side of the carapace than on the left side. The lateral

margins of costals 6 are as long as the free, lateral margin

of left costal 2. Costals 8 are small and are wider than

long. Depressions on costals 8 for contact of the ilia are

absent.

Remarks. Previously, this specimen was attributed to the

genus Aspideretoides Gardner et al., 1995 (Danilov et al.

2011a, b), based on the presence of a preneural. However,

a preneural is characteristic of other Cretaceous triony-

chids such as the genus Gilmoremys Joyce & Lyson,

2011, which is considered to be a member of the Plasto-

menidae (Joyce & Lyson 2011). In addition, new observa-

tion shows that this specimen had eight neurals, a

character that is also observed in some Cretaceous triony-

chines, but is absent in Aspideretoides spp. and Gil-

moremys, which have seven neurals. For this reason, we

decide to establish a new species within Trionyx sensu

lato based on this specimen.

Trionychidae nomen dubium

Amyda menneri Chkhikvadze in Chkhikvadze &

Shuvalov, 1988

1988 Amyda menneri Chkhikvadze in Chkhikvadze &

Shuvalov; Chkhikvadze & Shuvalov: 200, fig. 1.

1999 Amyda menneri Chkhikvadze in Chkhikvadze &

Shuvalov; Khosatzky: 147.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

2000 ‘Amyda’ menenri (sic.) Chkhikvadze in

Chkhikvadze & Shuvalov; Sukhanov: 345.

2009 ‘Amyda’ menneri; Chkhikvadze in Chkhikvadze &

Shuvalov; Danilov & Vitek: 54.

2010 ‘Amyda’ menneri; Chkhikvadze in Chkhikvadze &

Shuvalov; Vitek & Danilov: table 1, fig. 1.

2011a Amyda menneri Chkhikvadze in Chkhikvadze &

Shuvalov; Danilov et al.: 14, 15 (part).

2011b Amyda menneri Chkhikvadze in Chkhikvadze &

Shuvalov; Danilov et al.: 95, 96 (part).

2012a ‘Amyda’ menneri; Chkhikvadze in Chkhikvadze &

Shuvalov; Danilov & Vitek: 421.

Holotype. IPGAS 11-5-1, incomplete postcrania of one

individual, including a nuchal, the proximal part of costal

1, a fragment of the right hyo- and hypoplastron, a frag-

ment of a xiphiplastron, and assorted non-shell bones,

from Gurilin Tsav, Umunugovi Aimag, Mongolia, Nem-

egt Formation, Maastrichtian.

Paratypes. IPGAS 11-5-2, nuchal fragment, IPGAS 11-

5-3, anterior part of carapace, IPGAS 11-5-4, distal part

of right hyoplastron, IPGAS 11-5-5, right costal 7, and

IPGAS 11-5-6, right hyoplastron, all from holotype

locality; IPGAS 11-13-11, medial part of right hyoplas-

tron, from Bugin Tsav, Umunugovi Aimag, Mongolia,

Nemegt Formation, Maastrichtian; IPGAS 11-14-2, left

posterior part of carapace, IPGAS 11-14-3, medial part

of left hypoplastron, and IPGAS 11-14-4, posterior part

of carapace, all from Ingeni Khobur, Umunugovi Aimag,

Mongolia, Nemegt Formation, Maastrichtian; IPGAS 11-

17-1, left half of carapace, from Bamba Khuduk, Umunu-

govi Aimag, Mongolia, Nemegt Formation, Maastrichtian.

Remarks. Amyda menneri was described based on iso-

lated shell fragments from several localities of the

Nemegt Formation of Mongolia (Chkhikvadze & Shu-

valov 1988; see Danilov & Vitek 2012a for the English

diagnosis of this species). Chkhikvadze & Shuvalov

(1988) suggested that material called ‘Trionyx sp. a’ by

Khosatzky & M»ynarski (1971) from a different locality

(Tsagan Khushu) in the Nemegt Formation might

belong to A. menneri. Danilov et al. (2011a, b) sug-

gested that most forms of trionychids known from the

Barungoyot and Nemegt formations of Mongolia are

attributable to Amyda menneri. Danilov & Vitek

(2012a, p. 423) reported that “figures in the published

type description [of Amyda menneri] are of such low

quality that we have found it difficult to meaningfully

compare the above-listed specimens [see Referred mate-

rial] with those of other trionychids”. Those authors

mentioned some similarities between A. menneri and

Apalonina (in the reduced costal 8), but conservatively

assigned the species to Trionychini. They also sug-

gested that some of the materials mentioned by Kho-

satzky (1999, p. 147) from the Nemegt Formation may

also belong to this species. However, results of the

present study show that the diversity of trionychids of

the Nemegt Formation is higher than considered previ-

ously. Given that material of Amyda menneri is very

fragmentary and comes from different localities, it may

belong to multiple trionychid taxa. Some of this mate-

rial (IPGAS 11-14-4; Chkhikvadze & Shuvalov 1988,

fig. 1E) shows similarities with Gobiapalone spp. in the

shape of costals 6–8, whereas other figured specimens,

including the holotype (IPGAS 11-5-1, IPGAS 11-5-6,

IPGAS 11-14-3; Chkhikvadze & Shuvalov 1988, fig.

1“A–)”), can be identified only as Trionychidae indet.

For this reason, we consider all material of Amyda men-

neri to be Trionychidae indet. and Amyda menneri to be

a nomen dubium.

Figure 15. Carapace of ‘Trionyx’ shiluutulensis sp. nov., MPC 25/166, from the unknown formation (Campanian) of Shiluut Ula in: A,dorsal and B, ventral views. Reconstructed outlines are indicated by dashed lines.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

Figure 16. Shell material of Late Cretaceous Trionychidae of Mongolia from different localities. A, Gobiapalone sp., MPC 25/165,incomplete carapace from the Lower White Beds of Khermin Tsav in dorsal view. B–X, Trionychidae indet: B, C, material from theDjadokhta Formation of Abdarain Nuru; B, ZIN PH 1/159, partial nuchal in dorsal view; C, ZIN PH 2/159, partial right hyo- andhypoplastra in ventral view; D, ZPAL MgCh, unnumbered, costal fragment from the Djadokhta Formation of Bayn Dzak in dorsal view;E, PIN 3458-4, partial carapace from the upper part of the Baynshire Formation of Bayshin Tsav in dorsal view; F, ZPAL MgCh/71,fragment of the carapace on a cast of the internal cavity of the shell from the ?Nemegt Formation of Char Teeg in dorsal view; G, ZINPH 1/163, neural from the Nemegt Formation of Gilbentu; H–J, material from unknown formation of Khamdin Khural; H, ZIN PH 1/162, shell fragment; I, ZIN PH 2/162, lateral fragment of left hyoplastron in ventral view; J, middle part of hyoplastron in ventral view;K, L, material from the Baynshire Formation of Khara-Khutul; K, ZIN PH 52/80, medial part of right hypoplastron in ventral view; L,ZIN PH 53/80, middle part of hypoplastron in ventral view; M, MPC 25/167 (HMNS 2006-04-477), shell fragment from the UpperWhite Bed of Khermin Tsav II; N, ZIN PH 1/164, lateral costal fragment from the Nemegt Formation of Khermin Tsav in dorsal view;O, ZIN PH 2/88, partial nuchal from the Baynshire Formation of Khongil Tsav in dorsal view; P, ZIN PH 1/166, lateral hyoplastronfragment from the lower part of the Baynshire Formation of Ongon Ulan Ula in ventral view; Q, ZIN PH 2/81, partial costal from theBaynshire Formation of Shine Us Khuduk in dorsal view; R, ZIN PH 1/167, lateral costal fragment from the upper part of the BaynshireFormation of Shiregin Gashun; S, ZIN PH 1/168, costal fragment from the unknown formation of Tel Ulan Shalcha in dorsal view; T,ZIN PH 3/24, partial right hyo- and hypoplastra from the Nemegt Formation of Tsagan Khushu in ventral view; U, V, material from theBaynshire Formation of Tsagan Teg; U, ZIN PH 2/91, costal fragment in dorsal view; V, ZIN PH 3/91, posterior neural in dorsal view;W, ZIN PH 1/169, lateral costal fragment from the lower part of the Baynshire Formation of Tugrikin Ula in dorsal view; X, ZIN PH 8/97, lateral fragment of hyo- or hypoplastron from the unknown locality in ventral view. Reconstructed outlines are indicated by dashedlines.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

Discussion

Cretaceous trionychids of MongoliaThe material described in this study belongs to seven trio-

nychid species: the cyclanorbine Nemegtemys conflata,

and the trionychines Gobiapalone breviplastra, G. orlovi,

‘Trionyx’ baynshirensis, ‘T.’ gilbentuensis, ‘T.’ gobiensis

and ‘T.’ shiluutulensis.

Nemegtemys conflata is known from two plastral frag-

ments from the Nemegt Formation. This taxon is assigned

to Cyclanorbinae based on the fusion of the hyo- and hypo-

plastra at a small size (just after hatching) (see Meylan

1987). Nemegtemys conflata, if correctly assigned, is the

first Cretaceous and the earliest known cyclanorbine. Previ-

ously, fossil cyclanorbines were known from a series of

African and Arabian localities that date as far back as the

Early Miocene (see Lapparent de Broin (2000) for sum-

mary of cyclanorbine fossil record). Plastomenidae, known

from the Campanian to Eocene of North America, are now

considered to be stem-cyclanorbines (Joyce et al. 2009;

Joyce & Lyson 2010, 2011). According to Danilov & Vitek

(2012a), no diagnosable specimens of Plastomenidae are

known from the Cretaceous of Asia.

Gobiapalone breviplastra and G. orlovi are members of

the genus Gobiapalone. This genus is assigned to Triony-

chinae based on a combination of the following charac-

ters: the dorsal edge of the apertura narium externa is

slightly to strongly emarginate, the foramen jugulare pos-

terius is isolated from the fenestra postotica by a descend-

ing arch of the opisthotic, the nuchal bone is at least three

times wider than long, and the neural series always con-

tains at least one reversal in neural orientation (synapo-

morphies of Trionychinae sensu Meylan 1987). In

addition, it lacks the following synapomorphies of Plasto-

menidae (Joyce & Lyson 2011): the presence of a preneu-

ral, contact between the xiphiplastra along their entire

length, the formation of an extensive, infolded secondary

palate by the maxillae, and an extremely elongated mandi-

ble. Gobiapalone is assigned to Trionychini based on the

following characters: the parietal makes up nearly one-

quarter of the processus trochlearis oticum, and the pres-

ence of seven neurals (Meylan 1987).

Gobiapalone breviplastra is represented by abundant

shell material and some non-shell postcrania from the

Nemegt Formation. We refer historical specimens that

had been reported under different names to this species

(see Systematic palaeontology section for complete list of

synonymies). The material of this species demonstrates a

high degree of variation in shell morphology (see Descrip-

tion). This species differs from other trionychine species

of the Nemegt Formation (‘Trionyx’ gilbentuensis and

‘T.’ gobiensis) in having a medium carapace length (up to

260 mm), suprascapular fontanelles (at carapace length

up to 224 mm), and type A shell sculpturing, as well as

differing in the shape of the hyo- and hypoplastra.

Gobiapalone orlovi is the best known trionychid from

the Late Cretaceous of Mongolia. Remains include a com-

plete skeleton, several mostly complete carapaces, and a

plastral fragment from the Baynshire Formation. Gobiap-

alone orlovi has a long list of synonymies and material of

this species demonstrates a high degree of variation in the

shell morphology (see the account of this species), similar

to G. breviplastra. This species differs from ‘Trionyx’

baynshirensis (the second trionychid of the Baynshire

Formation) in its smaller size (up to 335 mm), type A shell

sculpturing and the shape of the hyo- and hypoplastra.

‘Trionyx’ baynshirensis is known from a single plastral

fragment from the Baynshire Formation. That species is

assigned to Trionychinae based on its similarity with the

trionychine ‘Trionyx’ kansaiensis. This species differs

from the other trionychid taxon known from the Baynshire

Formation, Gobiapalone orlovi, in its larger size (up to

500 mm), type B shell sculpturing, and the shape of the

hyo- and hypoplastra. These characters are also found in

‘Trionyx’ gilbentuensis from the Nemegt Formation.

‘Trionyx’ gilbentuensis is also known from a single

plastral fragment from the Nemegt Formation. This spe-

cies is assigned to Trionychinae based on its similarity

with the trionychine ‘Trionyx’ baynshirensis. It differs

from other trionychine species of the Nemegt Formation

(Gobiapalone breviplastra and ‘Trionyx’ gobiensis) in its

larger size (up to 500 mm), type B shell sculpturing, and

the shape of the hyo- and hypoplastra.

‘Trionyx’ gobiensis is known from three incomplete

carapaces and additional carapace fragments from the

Nemegt Formation. This species is assigned to Trionychi-

nae based on a combination of the following characters:

the nuchal bone is at least three times wider than long, the

neural series always contains at least one reversal in neu-

ral orientation (synapomorphies of Trionychinae sensu

Meylan 1987), and the preneural is absent (difference

from Plastomenidae which have a preneural; Joyce &

Lyson 2011). We refrain from assigning ‘T.’ gobiensis to

Trionychini based on the presence of seven neurals

(Meylan 1987), because this character is variable in ‘T.’

gobiensis and no other characters of Trionychini are

known in this species. ‘Trionyx’ gobiensis differs from

other trionychine species of the Nemegt Formation

(Gobiapalone breviplastra and ‘Trionyx’ gilbentuensis) in

its smaller size (up to 130 mm), absence of suprascapular

fontanelles, type A sculpturing, and large costals 8.

‘Trionyx’ shiluutulensis is known from a carapace from

an unknown formation of the Campanian. This species is

assigned to Trionychinae based on the combination of the

following characters: the nuchal is at least three times

wider than long, the neural series contains at least one

reversal in neural orientation (synapomorphies of Triony-

chinae sensu Meylan 1987), and eight neurals are present

(as compared to Plastomenidae which have seven neurals;

Joyce & Lyson 2011). This species differs from all other


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

trionychines of the Late Cretaceous of Mongolia in its

medium size (225 mm), type A sculpturing, and the pres-

ence of a preneural.

In addition to these species, one shell from the ?Bayn-

shire Formation of Khermin Tsav is assigned to Gobiap-

alone sp. (see Systematic palaeontology section).

The type material of Amyda menneri, a species previously

recognized from the Nemegt Formation (Chkhikvadze &

Shuvalov 1988), is considered by us to be Trionychidae

indet. and Amyda menneri to be a nomen dubium. That

material may belong to multiple trionychid species of the

Nemegt Formation and needs re-examination.

Only one Cretaceous species of trionychid from Mon-

golia is known from both relatively complete skull and

shell material. All of the other species have unknown skull

morphology. In addition, five of the seven described spe-

cies are known from only the carapace or the plastron, but

not both. Until more complete material of these species is

found and described, our ability to use them to infer evo-

lutionary patterns within Trionychidae is limited.

Our reanalysis of published data and presentation of

original data about fragmentary material of Cretaceous tri-

onychids from 45 Mongolian localities (Fig. 1, Online

Supplementary Material Table 4 and Appendix 1) shows

varying levels of trionychid diversity in different locali-

ties. Three trionychid taxa can be identified at each of

three localities (Bugin Tsav, Khermin Tsav and Nemegt),

and two trionychid taxa can be identified at each of eight

localities (Altan Ula, Bayn Shire, Bamba Khuduk, Gil-

bentu, Khamdin Khural, Khara Khutul, Nogon Tsav and

Tsagan Khushu). All other localities have only one triony-

chid taxon. Material from 15 localities can be identified to

species level, whereas material from other known locali-

ties can be identified only as Trionychidae indet.

To summarize, we recognize one trionychid taxon (Tri-

onychinae indet.) in the Early Cretaceous of Mongolia,

two trionychid species (Gobiapalone orlovi and ‘Trionyx’

baynshirensis) in the Baynshire Formation (Cenomanian–

Santonian), one trionychid taxon (Trionychidae indet.) in

the Djadokhta Formation (Campanian), one trionychid

species (Gobiapalone breviplastra) in the Barungoyot

Formation (Campanian), and four trionychid species

(Gobiapalone breviplastra, Nemegtemys conflata,

‘Trionyx’ gilbentuensis and ‘T.’ gobiensis) in the Nemegt

Formation (Maastrichtian). One more trionychid species

(‘Trionyx’ shiluutulensis) comes from an unknown forma-

tion (Campanian). The diversity of Late Cretaceous trio-

nychids of Mongolia is summarized in Fig. 17.

Figure 17. Trionychids from the Late Cretaceous formations of Mongolia. A–C, Gobiapalone breviplastra gen. et sp. nov.; A, carapace(MPC 25/163) in dorsal view; B, left hyo- and hypoplastra (MPC 25/163) in ventral view; C, left hypoplastron (ZPAL MgCh/81) in ven-tral view. D, partial left hyo- and hypoplastra (ZPAL MgCh/83) in ventral view (mirror image). E, ‘Trionyx’ gobiensis sp. nov., carapace(ZPAL MgCh/76) in dorsal view. F, Nemegtemys conflata sp. nov., lateral fragment of left hyo- and hypoplastra (ZIN PH 1/157, holo-type) in ventral view. G, ‘Trionyx’ gilbentuensis sp. nov., incomplete left hyo- and hypoplastra (ZIN PH T/M46-2) in ventral view. H–J,Gobiapalone orlovi; H, carapace (MPC 25/160) in dorsal view; I, hyo- and hypoplastra (MPC 25/160) in ventral view; J, partial hyo-and hypoplastra (ZIN PH T/M46-1) in ventral view. K, ‘Trionyx’ baynshirensis sp. nov., a medial fragment of righthyo- and hypoplastra (PIN 557-134, holotype) in ventral view. L, ‘Trionyx’ shiluutulensis sp. nov., carapace (MPC 25/166, holotype) indorsal view.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

The presence of four species of trionychids in the

Nemegt Formation of Mongolia marks it as amongst the

most diverse localities in the entire known record of trio-

nychid evolution, including the modern day. The only

other known localities with comparable diversity are the

Campanian Judith River Formation in Canada (five spe-

cies), the Eocene Bridger Formation in the western USA

(at least four species), and the modern day fauna of

Java, Indonesia (four species: Pritchard 2001; Vitek

2012).

Phylogenetic position of GobiapalonePhylogenetic analysis 1 resulted in four shortest trees of

248 steps each (CI ¼ 0.45, RI ¼ 0.63). The strict consen-

sus tree is shown in Fig. 18A. Phylogenetic analysis 2

resulted in 24 most parsimonious trees of 294 steps each

(CI ¼ 0.39, RI ¼ 0.61). The strict consensus tree is shown

in Fig. 18B.

In both analyses Gobiapalone is monophyletic. In anal-

ysis 1, Gobiapalone is placed within Apalonina Meylan,

1987 and forms a polytomy with Rafetus euphraticus

(Daudin, 1802) and Apalone Rafinesque, 1832. The mono-

phyly of Gobiapalone is supported by three characters: a

nuchal width-to-length ratio greater than four, the pres-

ence of seven or eight neurals, and the presence of five

plastral callosities. Apalonina is supported by five charac-

ters: the vomer divides the maxillae, the vomer reaches

the intermaxillary foramen, the average ratio of the inter-

maxillary foramen length to the length of the primary pal-

ate is about 60%, seven or more ossifications in cornu

branchiale II are present, and costals 8 are reduced or

absent (Meylan 1987).

In analysis 2, Gobiapalone is united with Oliveremys

uintaensis (Leidy, 1872) in a clade that is sister to Apalo-

nina. The monophyly of Gobiapalone is supported only

by the presence of five plastral callosities. The clade

Gobiapalone þ Oliveremys is supported by two charac-

ters: a nuchal width-to-length ratio greater than four, and

a postorbital bar with a width about equal or a third of the

diameter of the orbit. This clade is united with Apalonina

by one character: the posterior edge of neural 5 (not

counting the preneural) lies anterior to the posterior edge

of costal 5. The clade Apalonina þ (Gobiapalone þ Oli-

veremys) forms a polytomy with ‘Trionyx’ egregius Hay,

1908 and a clade uniting extinct species of the genera

Aspideretoides and Axestemys Hay, 1899. The results of

analysis 2 are different from the strict consensus tree in

Figure 18. Strict consensus trees resulting from A, analysis 1, and B, analysis 2 of fossil and recent Trionychidae based on morphologicaldata (see Discussion for description of the trees). Numbers below nodes are Bremer support indices (0 indexes are not shown).


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

the analysis of Vitek (2012, fig. 2.1) with respect to the

better resolution within Trionychinae, and are more simi-

lar to a tree resulting from the same data matrix but delet-

ing Nilssonia formosa (Grey, 1869) and the fossil

‘Trionyx’ egregius (see Vitek 2012, fig. 2.2). However,

unlike the analysis of Vitek (2012), analysis 2 does not

support monophyly of Axestemys.

Thus, the results of analyses 1 and 2 show that Apalo-

nina, which is a relatively derived and well-supported trio-

nychid clade (Meylan 1987; Gardner et al. 1995;

Engstrom et al. 2004; Joyce et al. 2009; Joyce & Lyson

2011; Vitek 2011, 2012; this study), or its closest sister

taxa (stem-Apalonina) were already present in the Late

Cretaceous of Asia. Their presence suggests that most

other supra-generic clades of modern trionychids had

already evolved in Asia by the Late Cretaceous (Danilov

et al. 2011a, b) and should be present in the fossil record.

This hypothesis is supported by the discovery of a cycla-

norbine, Nemegtemys conflata, in the Late Cretaceous of

Mongolia. Another trionychid clade, formed by the sister

relationship of the extinct genera Aspideretoides and

Axestemys (Vitek 2012; this study), is also present in the

Late Cretaceous of Asia and represented by Aspidere-

toides spp. (Vitek & Danilov 2010; Danilov & Vitek

2012a, b). Members of other crown-clades that should

have been present given the hypotheses discussed above,

such as Chitrini, Pelodiscini, Aspideretini and Gigantae-

suarochelys, are still unknown in the published

Cretaceous fossil record (for a review see Danilov &

Vitek 2012a). Their absence and long ghost lineages in

general highlight a need for further collection, description

and revision of Cretaceous–Holocene fossil material of

trionychids. In particular, it is possible that some stem- or

crown-clade members of these clades are already recorded

in the literature, but unidentified as members of modern

clades because of a general lack of recent re-examination

of known fossils from Europe and Asia.

Distribution of Cretaceous Trionychidae in AsiaSince the Cretaceous trionychids of North America and

Asia were last reviewed (Danilov & Vitek 2012a), newly

published data on the subject can refine or revise our

understanding of the biogeography and evolutionary pat-

terns of trionychids (Liu et al. 2011; Danilov & Vitek

2012b; Hirayama et al. 2012; Vitek 2012; Vitek &

Danilov 2012).

Liu et al. (2011) reported a new pan-trionychid (¼ stem-

trionychid) from the Jiufotang Formation (Aptian) of west-

ern Liaoning, China. Danilov & Vitek (2012b) described

trionychid shell material from the Bissekty Formation (late

Turonian) of Uzbekistan. This material was assigned to

two shell-based taxa, Aspideretoides cf. riabinini and

‘Trionyx’ cf. kansaiensis. Skull-based taxa from the same

formation are Khunnuchelys kizylkumensis and Trionychini

indet. Hirayama et al. (2012) described the stem-

Figure 19. Temporal and geographical distributions of Cretaceous Trionychidae of Asia and North America. Gaps in record are filledwith grey. �denotes skull-only taxa. See text for the Mongolian record. ‘Stem-trionychid’ denotes an occurrence from Liu et al. (2011).Other data are after Danilov & Vitek (2012a), with additions and modifications from Danilov & Vitek (2012b), Hirayama et al. (2012),Vitek (2012) and Vitek & Danilov (2012) (see Discussion for details).


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

trionychid, Kappachelys okurai Hirayama et al., 2012,

based on two isolated carapacial elements from the Early

Cretaceous (?late Neocomian) Akaiwa Formation of west-

central Honshu, Japan. Vitek (2012) published a revision

of North American trionychids of the genus Axestemys.

According to this revision, Axestemys was present in the

Late Cretaceous (Campanian and Maastrichtian) of North

America. Vitek & Danilov (2012) described trionychid

shell material from the Santonian–?middle Campanian

Kyrkkuduk I locality in southern Kazakhstan. This material

was assigned to two taxa, ‘Trionyx’ kansaiensis and Trio-

nychidae indet. ‘Trionyx’ kansaiensis from Kyrkkuduk I

corresponds to Trionychinae indet. 1 of Danilov & Vitek

(2012a). An updated summary of the temporal and geo-

graphical distributions of Cretaceous Trionychidae of Asia

and North America is provided in Fig. 19.

In general, new data supports the recognition of two

patterns. First, the earliest and only known stem-triony-

chids are from the Early Cretaceous of Japan and China,

and support the hypothesis that trionychids evolved some-

where in Asia (Danilov 2005; Danilov & Vitek 2012a).

Second, two lineages of trionychids persisted in Middle

Asia and Kazakhstan throughout the Late Cretaceous.

Danilov & Vitek (2012a) identified three patterns of

distribution of Cretaceous trionychids in Asia: (1) repre-

sentatives of the genus Aspideretoides (Trionychini with a

preneural) are known only from Middle Asia and Kazakh-

stan; (2) Apalonina indet. and other Trionychini without a

preneural are known only from Mongolia; and (3) mem-

bers of Trionychinae with eight neurals have the broadest

geographical distribution and are present in Middle Asia,

Kazakhstan, Mongolia and China. Our study updates these

patterns. We show that Trionychini with a preneural had a

broader distribution and were also present in Mongolia

(‘Trionyx’ shiluutulensis).

Conclusions

A revision of the available material of Cretaceous triony-

chids of Mongolia allows us to recognize seven species,

six of which are new. Amongst these species are the oldest

known cyclanorbine, Nemegtemys conflata, two species of

the genus Gobiapalone, which is a close relative of mod-

ern Apalonina, and four other species (‘Trionyx’ baynshir-

ensis, ‘T.’ gilbentuensis, ‘T.’ gobiensis, and ‘T.’

shiluutulensis). Some of these species are based on speci-

mens collected by the Mongolian Palaeontological Expe-

dition (1946, 1948, 1949) that remained undescribed and

unrecognized for more than 60 years. Most of these spe-

cies are based on incomplete material. Some of the revised

material, including several skulls, needs additional prepara-

tion and study. In spite of the recent contributions to our

knowledge of Cretaceous trionychids of Asia and North

America, there are still many gaps in the record, and the

phylogenetic position of most taxa of Cretaceous triony-

chids remains poorly understood. In particular, the relation-

ship of Nemegtemys conflata to other cyclanorbines and

plastomenids remains an open question, as does the rela-

tionship of ‘Trionyx’ kansaiensis and similar Mongolian

taxa to other trionychids. Early representatives of crown

clades other than Apalonina are still missing from the

known fossil record. New discoveries of more complete

material of Cretaceous trionychids of Mongolia are needed

to further improve our knowledge about diversity of Creta-

ceous trionychids of Mongolia and Asia in general and

understanding of the early diversification of this family.

Acknowledgements

We thank Magdalena Borsuk-Bia»ynicka for access to tri-

onychid specimens in ZPAL and E. V. Syromyatnikova

(ZIN) for providing photographs of these specimens. We

thank Walter Joyce and an anonymous reviewer for useful

comments and suggestions. This study was supported by

the grants of the President of Russian Federation to the

Leading Scientific Schools [No. NSh-6560.2012.4],

Russian Foundation for Basic Research [No. 14-04-

00416_a], and Hirayama Ikuo Silk Road Museum to I.G.D.

Supplemental material

Supplemental material for this article can be accessed

here: http://dx.doi.org/[DOI: 10.1080/14772019.2013.

847870]

References

Batsch, A. J. G. C. 1788. Versuch einer Anleitung, zur Kenntnißund Geschichte der Thiere und Mineralien. AkademischeBuchhandlung, Jena, 528 pp.

Chkhikvadze, V. M. & Shuvalov, V. F. 1988. A new species ofa trionychid from the Upper Cretaceous deposits of Mongo-lia. Izvestiya Akademii Nauk Gruzinskoi SSR, 14, 198–204[In Russian].

Cope, E. D. 1868. On the origin of genera. Proceedings of theAcademy of Natural Sciences of Philadelphia, 20, 242–300.

Danilov, I. G. 2005. Die fossilen Schildkr€oten Europas. Pp.329–441 in U. Fritz (ed.) Handbuch der Reptilien undAmphibien Europas. Band 3/IIIB: Schildkr€oten (Testudines)II. Aula-Verlag, Wiebelsheim.

Danilov, I. G. & Syromyatnikova, E. V. 2008. New materialson turtles of the family Nanhsiungchelyidae from the Creta-ceous of Uzbekistan and Mongolia, with a review of thenanhsiungchelyid record in Asia. Proceedings of the Zoolog-ical Institute of the Russian Academy of Sciences, 312, 3–25.

Danilov, I. G. & Vitek, N. S. 2009. Cretaceous trionychids ofAsia: a review of record and biogeography. Pp. 52–58 inGaffney Turtle Symposium. Abstract Volume. Royal TyrrellMuseum, Drumheller.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

Danilov, I. G., Hirayama, R., Sukhanov, V. B.,Watabe, M. &Suzuki, S. 2011a. Turtles-trionychids of the Cretaceous ofMongolia: new data and a revision. Pp. 14–16 in Otchetnayanauchnaya sessiya po itogam rabot 2010 g. Tezisy dokladov.Zoological Institute of the Russian Academy of Sciences, St.Petersburg. [In Russian].

Danilov, I., Hirayama, R., Sukhanov, V., Watabe, M. &Suzuki, S. 2011b. Cretaceous softshelled turtles (Trionychi-dae) of Mongolia: new data and a revision. Journal of Verte-brate Paleontology, 31 (Supplement 2), 95–96.

Danilov, I. G. & Vitek, N. S. 2012a. Chapter 23. Cretaceous tri-onychids of Asia: an expanded review of their record andbiogeography. Pp. 419–438 in D. B. Brinkman, P. A.Holroyd & J. D. Gardner (eds)Morphology and Evolution ofTurtles. Springer, Berlin.

Danilov, I. G. & Vitek, N. S. 2012b. Soft-shelled turtles (Trio-nychidae) from the Bissekty Formation (Late Cretaceous:late Turonian) of Uzbekistan: Shell-based taxa. CretaceousResearch, 41, 55–64.

Daudin, F. M. 1802. Histoire Naturelle, G�enerale et Partic-uli�ere des Reptiles (Vol. 2). F. Dufart, Paris, 432 pp.

Engstrom, T. N., Shaffer, H. B. & McCord, W. P. 2004. Mul-tiple data sets, high homoplasy, and the phylogeny of soft-shell turtles (Testudines: Trionychidae). Systematic Biology,53, 693–710.

Eriksson, T. 1998. AutoDecay. 4.0.1. Department of Botany,Stockholm University, Stockholm.

Ernst, C. H. & Barbour, R. W. 1989. Turtles of the World.Smithsonian Institution Press, Washington, D.C., 313 pp.

Forska�l, O. 1775. Descriptiones Animalium. Avium, amphibio-

rum, piscium, insectorum, vermium. Quae in itinere orientalobservavit. Hauniae, M€olleri, 164 pp.

Gardner, J. D. & Russell, A. P. 1994. Carapacial variationamong soft-shelled turtles (Testudines: Trionychidae), andits relevance to taxonomic and systematic studies of fossiltaxa. Neues Jahrbuch f€ur Geologie und Pal€aontologie, 193(2), 209–244.

Gardner, J. D., Russell, A. P. & Brinkman, D. B. 1995. Sys-tematics and taxonomy of soft-shelled turtles (Family Trio-nychidae) from the Judith River Group (mid-Campanian) ofNorth America. Canadian Journal of Earth Sciences, 32,631–643.

Geoffroy Saint Hillaire, M. 1809. Sur les tortues molles,nouveau genre sous le nom de Trionyx et sur la formationdes carapaces. Annales du Mus�eum d’Histoire Naturelle, 14,1–20.

Gilmore, C. W. 1928. Description of two new species of fossilturtles, from the Lance Formation of Wyoming. Proceedingsof the US National Museum, 50, 641–646.

Gilmore, C. W. 1931. Fossil turtles of Mongolia. Bulletin of theAmerican Museum of Natural History, 59, 213–257.

Gray, J. E. 1825. A synopsis of the genera of reptiles andamphibia, with a description of some new species. Annals ofPhilosophy, 10, 193–217.

Gray, J. E. 1831. Synopsis Reptilium; or short descriptions ofthe species of reptiles. Part I. Cataphracta. Tortoises, croco-diles, and enaliosaurians, London: Treuitel, Wiirtz and Co.,78 pp.

Gray, J. E. 1869. Notes on the families and genera of tortoises(Testudinata), and on the characters afforded by the study oftheir skulls. Proceedings of the Zoological Society of Lon-don, 1869, 165–225.

Hay, O. P. 1899. On the nomenclature of certain American fos-sil vertebrates. American Geologist, 24, 345–349.

Hay, O. P. 1902. Bibliography and catalogue of fossil verte-brates of North America. United States Geological SurveyBulletin, 179, 1–449.

Hay, O. P. 1908. The fossil turtles of North America. CarnegieInstitute of Washington Publications, 75, 1–568.

Hirayama, R., Isaji, S. & Hibino, T. 2012. Kappachelys okuraigen. et sp. nov., a new stem soft-shelled turtle from the EarlyCretaceous of Japan. Pp. 179–185 in D. B. Brinkman, P. A.Holroyd & J. D. Gardner (eds) Morphology and evolution ofturtles. Springer, Dordrecht.

Hummel, K. 1929. Die fossilen Weichschildkr€oten (Triony-chia). Eine morphologisch-systematische und stammesge-schichtliche studie. Geologische und Palaeontologische,Abhandlungen, 16, 359–487.

Hutchison, J. H. 2000. Diversity of Cretaceous turtle faunas ofeastern Asia and their contribution to the turtle faunas ofNorth America. Paleontological Society of Korea SpecialPublication, 4, 27–38.

Hutchison, J. H. 2009. New soft-shelled turtles (Plastomeninae:Trionychidae: Testudines) from the Late Cretaceous andPaleocene of North America. Paleobios, 29, 36–47.

Joyce, W. G., Parham, J. F. & Gauthier, J. A. 2004. Develop-ing a protocol for the conversion of rank-based taxon namesto phylogenetically defined clade names, as exemplified byturtles. Journal of Paleontology, 78, 989–1013.

Joyce, W. G., Revan, A., Lyson, T. R. & Danilov, I. G. 2009.Two new plastomenine softshell turtles from the Paleoceneof Montana and Wyoming. Bulletin of the Peabody Museumof Natural History, 50, 307–325.

Joyce, W. G. & Lyson, T. R. 2010. A neglected lineage ofNorth American turtles fills a major gap in the fossil record.Palaeontology, 53, 241–248.

Joyce, W. G. & Lyson, T. R. 2011. New material of Gil-moremys lancensis nov. comb. (Testudines: Trionychi-dae) from the Hell Creek Formation and the diagnosisof plastomenid turtles. Journal of Paleontology, 85,442–459.

Karl, H.-V. 1998. Zur Taxonomie der k€anozoischen Weich-schildkr€oten €Osterreichs und Deutschlands (Trionychidae:Trionychinae). Mitteilungen der Abteilungs f€ur Geologieund Pal€aontologie der Landesmuseum Joanneum, 56, 273–328.

Karl, H.-V. 1999. Die Zoogeographie der k€anozoischenWeichschildkr€ote Trionyx triunguis Forsk€al, 1775. JoanneaGeoogie und Pal€aontologie, 1, 27–60.

Khosatzky, L. I. & M»ynarski, M. 1971. Chelonians from theUpper Cretaceous of the Gobi Desert, Mongolia. Palaeonto-logia Polonica, 25, 131–144.

Khosatzky, L. I. 1976. A new representative of trionychids fromthe Late Cretaceous of Mongolia. Pp. 3–19 in Gerpetolo-giya. Kuban State University, Krasnodar. [In Russian].

Khosatzky, L. I. 1999. Turtles – trionychids of the Cretaceousof Mongolia. Pp. 141–149 in Voprosy Paleontologii. T. 11.St. Petersburg State University, St. Petersburg. [In Russian].

Kuznetsov, V. V., & Chkhikvadze, V. M. 1987. The Late Cre-taceous trionychids from Shakh-Shakh locality in Kazakh-stan. Materialy po Istorii Fauny i Flory Kazakhstana, 9, 33–39. [In Russian].

Lapparent de Broin, F. 2000. African chelonians from theJurassic to the Present. A preliminary catalog of the Africanfossil chelonians. Palaeontologica Africana, 36, 43–82.

Leidy, J. 1872. Notice of donation of fossils etc. fromWyoming.Proceedings of the Academy of Natural Sciences of Phila-delphia, 1872, 267–268.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

Liu, J., Joyce, W. & Tong, H. 2011. The oldest morphologi-cally modern soft-shelled turtle (Testudines, Pan-Trionychi-dae) from the Jehol biota of China. Journal of VertebratePaleontology, 31 (Supplement 2), 144.

Merkulova, N. N. 1978. A new Trionyx from the Nemegt(MPR). Byulleten’ Moskovskogo Obshchestva IspytateleyPrirody. Otdel Geologicheskiy, 53, 156. [In Russian].

Meylan, P. A. 1987. The phylogenetic relationships of soft-shelled turtles (Family Trionychidae). Bulletin of the Ameri-can Museum of Natural History, 186, 1–101.

M»ynarski, M. & Narmandach, P. 1972. New turtle remainsfrom the Upper Cretaceous of the Gobi Desert, Mongolia.Palaeontologia Polonica, 27, 95–102.

Page, R. D. M. 2001. NDE: NEXUSData Editor 0.5.0.Univer-sity of Glasgow, Glasgow, U.K.

Pritchard, P. C. H. 2001. Observations on body size, sympatry,and niche divergence in softshell turtles (Trionychidae).Chelonian Conservation and Biology, 4, 5–27.

Rafinesque, C. S. 1832. Description of two new genera of softshell turtles of North America. Atlantic Journal and Friendof Knowledge, 1, 64–65.

Sochava, A. V. 1975. Stratigraphy and lithology of the UpperCretaceous deposits of southern Mongolia. Trudy Sovmest-noy Sovetsko-Mongol’skoy Nauchno-Issledovatelskoy Geo-logicheskoy Ekspeditsii, 13, 113–182. [In Russian].

Shuvalov, V. F. & Chkhikvadze, V. M. 1975. New data on theLate Cretaceous turtles of South Mongolia. Trudy Sovmest-noy Sovetsko-Mongol’skoy Paleontologicheskoy Ekspeditsii,2, 214–229. [In Russian].

Shuvalov, V. F. & Chkhikvadze, V. M. 1979. On stratigraphicand systematic position of some freshwater turtles from newCretaceous localities in Mongolia. Trudy Sovmestnoy Sovet-sko-Mongol’skoy Paleontologicheskoy Ekspeditsii, 8, 58–76. [In Russian].

Sukhanov, V. B. 2000. Mesozoic turtles of Middle and CentralAsia. Pp. 309–367 in M. J. Benton, M. A. Shishkin, D. M.Unwin & E. N. Kurochkin (eds) The Age of Dinosaurs in Rus-sia and Mongolia. Cambridge University Press, Cambridge.

Suzuki, S. & Narmandakh, P. 2004. Change of the Cretaceousturtle faunas in Mongolia. Hayashibara Museum of NaturalSciences Research Bulletin, 2, 7–14.

Suzuki, S. 2005. An almost complete skeleton of the trionychidturtle (Trionychidae: Trionychinae) from the Upper Creta-ceous of Mongolia. P. 27 in Abstracts with Programs of the154th Regular Meeting of the Palaeontological Society ofJapan. Yamagata, Japan. [In Japanese].

Vitek, N. S.& Danilov, I. G. 2010. New material and a reassess-ment of soft-shelled turtles (Trionychidae) from the LateCretaceous of Middle Asia and Kazakhstan. Journal of Ver-tebrate Paleontology, 30, 383–393.

Vitek, N. S. 2011. Insights into the taxonomy and systematics ofNorth American Eocene soft-shelled turtles from a well-pre-served specimen. Bulletin of the Peabody Museum of Natu-ral History, 52(2), 189–208.

Vitek, N. S. 2012. Giant fossil soft-shelled turtles of NorthAmerica. Palaeontologia Electronica, 15(1), 1–43.

Vitek, N. S. & Danilov, I. G. 2012. New data on the soft-shelledturtles from the Upper Cretaceous of Kyrkkuduk I localityof Southern Kazakhstan. Proceeding of the Zoological Insti-tute of the Russian Academy of Sciences, 316(1), 50–56.

Watabe, M., Tsogtbaatar, Kh., Suzuki, S. & Saneyoshi, M.2010. Geology of dinosaur-fossil-bearing localities (Jurassicand Cretaceous: Mesozoic) in the Gobi Desert: Results ofthe HMNS-MPC Joint Paleontological Expedition. Hayashi-bara Museum of Natural Sciences Research Bulletin, 3, 41–118.

Yeh, H.-K. 1965. New materials of fossil turtles of Inner Mon-golia. Vertebrata PalAsiatica, 9, 47–69.


Dow

nloa

ded

by [

Igor

G. D

anilo

v] a

t 11:

45 2

1 Fe

brua

ry 2

014

Table 4. Localities, formations and identifications of Cretaceous Trionychidae of Mongolia. “*” denotes identifications based on published data. No. Locality Formation(s) Identification(s) 1 Abdarain Nuru Djadokhta Trionychidae indet. 2 Abdrant Nuru Djadokhta Trionychidae indet.* 3 Alak Shand Khuduk Nemegt Gobiapalone breviplastra 4 Altan Ula, Altan Ula I, III

and IV Nemegt Gobiapalone breviplastra, ‘Trionyx’

gobiensis 5 Amtgai Baynshire Trionychidae indet.* 6 Bayn Shire Baynshire Gobiapalone orlovi, ‘Trionyx’

baynshirensis 7 Bamba Khuduk Nemegt ‘Trionyx’ gobiensis, Trionychinae indet. 8 Bayn Dzak Djadokhta Trionychidae indet. 9 Bayshin Tsav Baynshire (upper part) Trionychidae indet.

10 Bugin Tsav Barungoyot and Nemegt Gobiapalone breviplastra (Barungoyot), G. breviplastra, Nemegtemys conflata

(Nemegt) 11 Burkhant Baynshire Gobiapalone orlovi 12 Char Teeg ?Nemegt Trionychidae indet. 13 Dersnii Khuduk Baynshire (lower part) Trionychidae indet.* 14 Dzun Shakhai Unknown (Early Cretaceous) Trionychidae indet.* 15 Ekhin Tukhum Baynshire (lower part) Trionychidae indet.* 16 Erdeni Ula Khulsangol Trionychidae indet.* 17 Gilbentu Nemegt ‘Trionyx’ gilbentuensis, Trionychidae

indet. 18 Gurilin Tsav Nemegt Trionychidae indet.* 19 Ingeni Khobur and Ingeni

Khobur III Nemegt Trionychidae indet.*

20 Ingeni Tsav Barungoyot and Nemegt Trionychidae indet.* 21 Khamdin Khural Unknown (Late Cretaceous) Trionychidae indet. (two forms) 22 Khara Khutul Baynshire Trionychidae indet. (two forms) 23 Khermin Tsav and

Khermin Tsav II ?Baynshire and Nemegt Gobiapalone sp. (?Baynshire),

Trionychidae indet (two forms; Nemegt) 24 Khongil Baynshire Gobiapalone orlovi 25 Khongil Obo Baynshire (upper part) Trionychidae indet.* 26 Knongil Tsav Baynshire Trionychidae indet. 27 Khuren Tsav Barungoyot (lower part) Trionychidae indet.* 28 Nemegt Nemegt Gobiapalone breviplastra, Nemegtemys

conflata, ‘Trionyx’ gobiensis 29 Nogon Tsav and Nogon

Tsav II Barungoyot (upper part) and

Nemegt Gobiapalone breviplastra (Barungoyot),

Trionychidae indet. (Nemegt) 30 Ongon Ulan Ula Baynshire (lower part) Trionychidae indet. 31 Shiluut Ula Unknown (Campanian) ‘Trionyx’ shiluutulensis 32 Shine Us Khuduk Baynshire Trionychidae indet. 33 Shine Usny Tolgod Unknown (Early Cretaceous) Trionychidae indet.* 34 Shiregin Gashun Baynshire (upper part) Trionychidae indet. 35 Tel Ulan Shalcha Unknown (probably, early Late

Cretaceous) Trionychidae indet.

36 Tsagan Khushu Nemegt ‘Trionyx’ gobiensis, Trionychidae indet. 37 Tsagan Teg Baynshire Formation Trionychidae indet. 38 Tugrikin Ula Baynshire (lower part) Trionychidae indet. 39 Ulan Khushu Nemegt Gobiapalone breviplastra 40 Ulan Tsab Ula Probably, Baynshire Trionychidae indet.* 41 Unegetu Ula Baynshire Gobiapalone orlovi 42 Uryl’b Usu Baynshire (lower part) Trionychidae indet.* 43 Ushyin Khuduk Baynshire (upper part) Trionychidae indet.* 44 Yagaan Khovil Djadokhta Trionychidae indet.* 45 Unknown locality Unknown (?Late Cretaceous) Trionychidae indet.

Appendix 1. Review of records of Cretaceous Trionychidae in Mongolia. See Fig. 1 for map of

occurrences.

1. Abdarain Nuru, Northern Gobi.

Geology and age. Djadokhta Formation, Campanian.

Material and references. Partial nuchal, partial right hyo- and hypoplastra, and other shell

fragments of Trionychidae indet. (Fig. 16B, C; collection ZIN PH 159; collected by A.V.

Sochava in 1970; as “part of plastron”; Khosatzky 1999: 145).

Remarks. The proportions of the nuchal (that was at least four times wider than long)

correspond to Trionychinae or Plastomenidae. The outline of the preserved part of the hyo- and

hypoplastra and type A sculpturing are similar to those of Gobiapalone spp. However, because

plastral morphology is not known in most other Mongolian trionychids of the Cretaceous, this

material is identified as Trionychidae indet.

2. Abdrant Nuru, Umunugovi Aimag.


Material and references. Disarticulated shell fragments and several skulls of Trionychidae

indet. (as Trionychidae; Suzuki & Narmandakh 2004: 9).

3. Alak Shand Khuduk (SW from Alak Shand Khuduk spring), Umunugovi Aimag.

Geology and age. Nemegt Formation, Maastrichtian.

Material and references. Partial carapace of Gobiapalone breviplastra (see Systematic

palaeontology section).

Remarks. Alak Shand Khuduk spring is situated 30 km W from Bugin Tsav.

4. Altan Ula, Altan Ula I, III and IV, Umunugovi Aimag.


Material and references. Partial carapace on a cast of the internal cavity of the shell of

Gobiapalone breviplastra from Altan Ula I (see Systematic palaeontology section); shell

fragment of ‘Trionyx’ gobiensis from Altan Ula III (see Systematic palaeontology section); shell

fragments of Trionychidae indet. from Altan Ula (PIN 553-66, collected by MPE in 1948; as

“shell fragments of trionychids”; Khosatzky 1999: 147), Altan Ula III (collection ZPAL

MgCh/121, collected by PMPE) and Altan Ula IV (HMNS 2006-04-474, collected by JMJPE in

2006; as Trionychidae; Suzuki & Narmandakh 2004: 10).

Remarks. All shell fragments of Trionychidae indet. from Altan Ula, Altan Ula III and IV have

type A sculpturing.

5. Amtgai, Umunugovi Aimag.

Geology and age. Baynshire Formation, Cenomanian – Santonian.

Material and references. Shell fragments of Trionychidae indet. (as Trionychidae; Suzuki &

Narmandakh 2004: 8).

6. Bayn Shire (= Bayn Shire), Dornogovi Aimag.


Material and references. Two incomplete carapaces and hyo- and hypoplastra of Gobiapalone

orlovi from the lower part of the Baynshire Formation, Cenomanian – early Turonian (see

Systematic palaeontology section); medial fragment of hyo-and hypoplastron of ‘Trionyx’

baynshirensis (see Systematic palaeontology section); shell fragments and two cervical vertebrae

of Trionychidae indet. (collection ZIN PH 156; collected by MPE in 1948 and G.G. Martinson in

1967; Khosatzky 1999: 142); shell fragments of Trionychidae indet. (collection PIN 557,

collected by MPE in 1948; collection ZPAL MgCh/149, collected by PMPE in 1963).

Remarks. The shell fragments of Trionychidae indet. from Bayn Shire have type A sculpturing.

7. Bamba Khuduk (= Bambu Khuduk), Umunugovi Aimag.


Material and references. Incomplete carapace of ‘Trionyx’ gobiensis (see Systematic

palaeontology section); left half of carapace of Trionychidae indet. (IPGAS 11-17-1, paratype of

Amyda menneri; see Systematic palaeontology section); unprepared skull of Trionychidae indet.

in matrix (PIN 4064-3); numerous shell fragments of Trionychidae indet. (collection PIN 4064,

collected by JSMPE in 1976).

Remarks. The shell fragments of Trionychidae indet. from Bamba Khuduk have type A

sculpturing. All these specimens may belong to a single taxon. The skull specimen needs a

preparation.

8. Bayn Dzak (= Shabarakh Usu), Umunugovi Aimag.


Material and references. Costal fragment of Trionychidae indet. (Fig. 16D; ZPAL MgCh,

unnumbered, collected by PMPE in 1970).

Remarks. The fragment has type A sculpturing.

9. Bayshin Tsav, Umunugovi Aimag.

Geology and age. Upper part of the Baynshire Formation, late Turonian – Santonian.

Material and references. Partial carapace of Trionychidae indet. (Fig. 16E; PIN 3458-4,

collected by JSMPE in 1972); shell fragments of Trionychidae indet. (as Trionychidae; Suzuki &


Remarks. PIN 3458-4 is represented by the anterior two-thirds of a carapace, which needs

additional preparation. The length of the preserved part of the carapace is about 22 cm, and its

estimated total length was about 25 cm. The outline of the carapace is slightly rounded laterally

and weakly scalloped, similar to Gobiapalone spp. The proportions of the nuchal (about five

times wider than long) correspond to Trionychinae or Plastomenidae. The neurals are not

observable. Only parts of costals 1–6 are preserved, and their total number is not clear. The free

ribs project slightly from the costals, similar to Gobiapalone spp. The specimen has type A

sculpturing. In spite the similarities of PIN 3458-4 with Gobiapalone spp. some important

characters of this specimen are not observable and, for this reason, it is determined only as

Trionychidae indet.

10. Bugin Tsav and Bugin Tsav II, Umunugovi Aimag.

Geology and age. Barungoyot Formation, Campanian, and Nemegt Formation, Maastrichtian.

Material and references. Four partial skeletons and a partial shell of Gobiapalone breviplastra

from the Nemegt Formation of Bugin Tsav and Bugin Tsav II, and partial carapace of

Gobiapalone breviplastra from the Barungoyot Formation of Bugin Tsav (see Systematic

palaeontology section); lateral fragment of right hypoplastron of Nemegtemys conflata from the

Nemegt Formation of Bugin Tsav (see Systematic palaeontology section); partial carapace and

skull of Trionychidae indet., which need a preparation, from the Nemegt Formation of Bugin

Tsav II (HMNS 2006-04-470, collected by JMJPE in 2006); fragmentary shell remains of

Trionychidae indet. from the Nemegt Formation of Bugin Tsav II (HMNS 2006-04-471,

collected by JMJPE in 2006); nuchal of Trionychidae indet. from the Nemegt Formation of

Bugin Tsav (HMNS 2006-04-473, collected by JMJPE in 2006; as Trionychidae; Suzuki &

Narmandakh, 2004); medial part of right hyoplastron of Trionychidae indet. from the Nemegt

Formation of Bugin Tsav (IPGAS 11-13-11, paratype of Amyda menneri; see Systematic

palaeontology section); shell fragments of Trionychidae indet. from the Nemegt Formation of

Bugin Tsav (collection ZIN PH 160, collected by A.V. Sochava in 1967).

11. Burkhant, Dornogovi Aimag.


Material and references. Partial skeleton and carapace of Gobiapalone orlovi (see Systematic


12. Char Teeg (= Khar Teg), Dornogovi Aimag.

Geology and age. ?Nemegt Formation, ?Maastrichtian.

Material and references. Fragment of the carapace on a cast of the internal cavity of the shell of

Trionychidae indet. (Fig. 16F; ZPAL MgCh/71, collected by PMPE in 1963; Khosatzky and

Młynarski 1971).

Remarks. This specimen was mentioned by Khosatzky & Młynarski (1971) as “an imprint of a

large fragment of carapace together with pleural plate”. Actually, it consists of the complete left

costal 5 and the medial fragment of the left costal 6. In addition, imprints of neurals 5-7 and left

costals 6-8 are visible on a cast of the internal cavity of the shell. The estimated length of the

shell was about 40 cm. Type A sculpturing. There were seven neurals and costals 7 and 8

contacted at the midline. Neural 5 was isometric, neural 6 was hexagonal short-sided anteriorly,

and neural 7 was pentagonal. Costal 5 was longer laterally than medially. The free rib of costal 5

slightly projects from the border. Given the incompleteness of this specimen and the uncertainty

with the formation, we can identify it only as Trionychidae indet.

13. Dersnii Khuduk, Dornogovi Aimag.

Geology and age. Lower part of the Baynshire Formation, Cenomanian – early Turonian.

Material and references. Shell fragments of Trionychidae indet. (as Trionyx sp.; Shuvalov &

Chkhikvadze 1975: 217, 222, 229).

14. Dzun Shakhai, Dornogovi Aimag.

Geology and age. Early Cretaceous.

Material and references. Incomplete nuchal and costals of Trionychidae indet. (as

Trionychidae; Suzuki & Narmandakh 2004: 8, pl. 1, pl.-fig. 2; as Trionychinae indet. 7; Danilov

& Vitek, 2009: 54; Trionychinae indet. 6; Danilov & Vitek 2012: 430).

Remarks. Danilov & Vitek (2009, 2012) identified this material as Trionychinae indet. based on

the nuchal which is is at least three times wider than long. However, recently Joyce & Lyson

(2011) demonstrated that the presence of a nuchal at least four times wider than long is also

characteristic of Plastomenidae. For this reason, we reidentify the material from Dzun Shakhai

as Trionychidae indet.

15. Ekhin Tukhum, Dornogovi Aimag.



Chkhikvadze 1979: 70).

16. Erdeni Ula (= Erdene Uul), Dundgov Aimag.

Geology and age. Khulsangol Svita, Aptian–Albian.

Material and references. Shell fragments of Trionychidae indet. (as trionychid; Khosatzky

1999: 141).

Remarks. We did not manage to find any trionychid specimens from Erdeni Ula in ZIN PH

collections studied by L.I. Khosatzky. It is probable, that specimens which Khosatzky identified

as trionychids are actually specimens of Nanhsiungchelyidae, which were reported from this

locality by Danilov & Syromyatnikova (2008).

17. Gilbentu, Umunugovi Aimag.


Material and references. A partial hyo- and hypoplastra of ‘Trionyx’ gilbentuensis (see

Systematic palaeontology section); shell fragments of Trionychidae indet. (Fig. 16G; collection

ZIN PH 163, collected by MPE in 1946).

Remarks. The shell fragments of Trionychidae indet. from Gilbentu are characterized by type A

sculpturing, whereas ‘Trionyx’ gilbentuensis has type B sculpturing.

18. Gurilin Tsav, Umunugovi Aimag.


Material and references. Incomplete postcranium of one individual, including nuchal, proximal

part of costal 1, fragment of right hyo- and hypoplastron, fragment of xiphiplastron, and assorted

non-shell bones of Trionychidae indet. (IPGAS 11-5-1, holotype of Amyda menneri; see

Systematic palaeontology section).

19. Ingeni Khobur and Ingeni Khobur III, Umunugovi Aimag.


Material and references. Shell fragments of Trionychidae indet. from Ingeni Khobur (IPGAS

11-14-2, IPGAS 11-14-3, and IPGAS 11-14-4, paratypes of Amyda menneri; see Systematic

palaeontology section); shell fragments of Trionychidae indet. from Ingeni Khobur (Khosatzky,

1999: 147); shell fragments of Trionychidae indet from Ingeni Khobur III (as Trionyx sp.;

Shuvalov & Chkhikvadze 1979: 66).

20. Ingeni Tsav, Bayankhongor Aimag.

Geology and age. Barungoyot Formation, Campanian, and Nemegt Formation, Maastrichtian.


Chkhikvadze 1979: 63, 64).

21. Khamdin Khural.

Geology and age. Unknown formation, Late Cretaceous.

Material and references. Shell fragments of Trionychidae indet. (Fig. 16H–J; collection ZIN

PH 162; collected by G.G. Martinson in 1971).

Remarks. The shell fragments of Trionychidae indet. belong to two forms, which demonstrate

two types of sculpturing: the first form (Fig. 16H) has type A sculptruing, whereas the second

one (Fig. 16I, J) has shell sculpturing which is a pattern of more dense net of ridges and smaller

pits. The second type of sculpturing is present on two plastral fragments. One of them (Fig. 16I),

probably, from a lateral part of hyoplastron, demonstrates high degree of ossification.

22. Khara Khutul (= Khara Khutul Ula; = Khar Hötöl Uul), Dornogovi Aimag.


Material and references. Shell fragments of Trionychidae indet. (Fig. 16K, L; collection ZIN

PH 80, collected by G.G. Martinson in 1971 and 1978, V.F. Shuvalov in 1971 and N.N. Verzilin

in 1978); one fragmentary shell remain of Trionychidae indet. (as Trionychidae; Suzuki &


Remarks. The shell fragments of Trionychidae indet. from Khara Khutul belong to two forms

wich demonstrate different types of shell sculpturing. The first form, represented by the medial

part of a hypoplastron and other fragments, has type A sculpturing and is most similar to

Gobiapalone orlovi. It is probable, that this material was mentioned as Amyda orlovi by Sochava

(1975: 115) and Khosatzky (1999: 142). The second form, represented by a fragment of a

hypoplastron has type B sculpturing and, most probably, was mentioned by Khosatzky (1999:

142) as “middle-sized trionychids with tuberculated relief of the plastron between axillary and

inguinal notches”.

23. Khermin Tsav (= Khermeen Tsav) and Khermin Tsav II, Umunugovi Aimag.

Geology and age. Lower White Beds, ?Baynshire Formation, Cenomanian – Santonian, and

Upper White Beds, Nemegt Formation, Maastrichtian.

Material and references. Shell fragments of Trionychidae indet. from the Nemegt Formation of

Khermin Tsav (as Trionyx sp.; Shuvalov & Chkhikvadze 1975: 219, 222, 229); MPC 25/165,

nearly complete carapace of Gobiapalone sp. from the Lower White Beds of Khermin Tsav (see

Systematic palaeontology section); shell fragments of Trionychidae indet. from the Upper White

Beds of Khermin Tsav, shell fragments of Trionychidae indet. from the Lower White Bed of

Khermin Tsav II and ten specimens of Trionychidae indet. from the Upper White Bed of

Khermin Tsav II (Fig. 16M; MPC 25/167, collected by JMJPE in 2006; as Trionychidae; Suzuki

& Narmandakh 2004: 9); shell fragments of Trionychidae indet from Khermin Tsav II

(collection ZPAL MgCh/160, collected by PMPE in 1971); shell fragments of Trionychidae

indet. from the Nemegt Formation of Khermin Tsav (Fig. 16N; collection ZIN PH 164, collected

by I. Yu. Neustroeva in 1977).

Remarks. The ten specimens of Trionychidae indet. from the Upper White Bed of Khermin

Tsav II are shell fragments with sculpturing similar to the type B. Other shell fragments from

Khermin Tsav and Khermin Tsav II that we examined have type A sculpturing. Therefore,

trionychids of Khermin Tsav are represented by one form (Gobiapalone sp.) in the ?Baynshire

Formation and probably by two forms (Trionychidae indet.) in the Nemegt Formation.

24. Khongil, Dornogovi Aimag.


Material and references. Incomplete carapace of Gobiapalone orlovi (see Systematic


Remarks. Khosatzky (1999) mentioned that Khongil is situated 2 km N from Tsagan Teg Obo

(=Tsagan Teg), which is very close to Khongil Tsav, whereas Sochava (1975, fig. 24) indicated

that Khongil was about 50 km SW from Khongil Tsav.

25. Khongil Obo, Dornogovi Aimag.


Material and references. Trionychidae indet. (as Trionyx sp.; Shuvalov & Chkhikvadze 1979:

67).

Remarks. According to Shuvalov & Chkhikvadze (1979), Khongil Obo is situated in the

northern part of Bayshin Tsav Depression, 3 km E from the Khongil Obo Mountain.

26. Khongil Tsav (= Khangil Tsav), Dornogovi Aimag.


Material and references. Shell fragments of Trionychidae indet. (Fig. 16O; collection ZIN PH

88, collected by R. Barsbold and G.G. Martinson in 1967; as “a fragment of the nuchal of the

trionychid of the Turonian-Santonian type and a lateral part of the plastron, which can be

assigned to “Plastomenus””; Khosatzky 1999: 144); five specimens of Trionychidae indet. (as

Trionychidae; Suzuki & Narmandakh 2004: 8).

Remarks. The shell fragments of Trionychidae indet. from Khongil Tsav have type A

sculpturing.

27. Khuren Tsav, Bayankhongor Aimag.

Geology and age. Lower part of the Barungoyot Formation, Campanian.


221, 222, 229).

28. Nemegt, Umunugovi Aimag.


Material and references. Lateral fragment of hyo- and hypoplastra of Nemegtemys conflata (see

Systematic palaeontology section); partial juvenile shell with fragments of non-shell postcrania,

posterior carapace fragment, left hypoplastron, and partial left hyo- and hypoplastra of

Gobiapalone breviplastra (see Systematic palaeontology section); incomplete carapace of

‘Trionyx’ gobiensis (see Systematic palaeontology section); shell fragments of Trionychidae

indet. (ZPAL MgCh/88 and ZPAL MgCh/89, collected by PMPE; as Trionyx sp. b; Młynarski &

Narmandach 1972: 100; collection ZIN PH 157, collected by MPE in 1948; Khosatzky 1999:

147); two specimens of Trionychidae indet. (as Trionychidae; Suzuki & Narmandakh 2004: 10).

Remarks. The shell fragments of Trionychidae indet. from Nemegt have type A sculpturing.

29. Nogon Tsav and Nogon Tsav II, Bayankhongor Aimag.

Geology and age. Upper part of the Barungoyot Formation, Campanian, and Nemegt Formation,

Maastrichtian.

Material and references. Partial carapace of Gobiapalone breviplastra from the upper part of

the Barungoyot Formation of Nogon Tsav (see Systematic palaeontology section); shell

fragments of Trionychidae indet. from the Nemegt Formation of Nogon Tsav (collection ZIN PH

165, collected by MPE, by A.V. Sochava in 1966 and 1968, and by L.Ya. Borkin in 1980 and

1981; as “bones of trionychids”; Khosatzky 1999: 147); shell fragments of Trionychidae indet.

from the Nemegt Formation of Nogon Tsav II (as Trionyx sp.; Shuvalov & Chkhikvadze 1979:

66).

Remarks. The shell fragments of Trionychidae indet. from the Nemegt Formation of Nogon

Tsav have type A sculpturing.

30. Ongon Ulan Ula, Umunugovi Aimag.

Geology and age. Lower part of the Baishire Formation, Cenomanian – early Turonian.

Material and references. Shell fragments of Trionychidae indet. (Fig. 16P; collection ZIN PH

166).

Remarks. The shell fragments of Trionychidae indet from Ongon Ulan Ula have type A

sculpturing.

31. Shiluut Ula, Umunugovi Aimag.

Geology and age. Unknown formation, Campanian (see Watabe et al., 2010).

Material and references. Carapace of ‘Trionyx’ shiluutulensis (see Systematic palaeontology

section).

32. Shine Us Khuduk (=Shine Us Khudag), Dornogovi Aimag.


Material and references. Shell fragments of Trionychidae indet. (Fig. 16Q; collection ZIN PH

81; collected by G.G. Martinson in 1968 and 1971, and by N.N. Verzilin in 1978; Khosatzky,

1999: 141; three specimens of Trionychidae indet. (as Trionychidae; Suzuki & Narmandakh

2004: 8).

Remarks. The shell fragments of Trionychidae indet from Shine Us Khuduk have type A

sculpturing.

33. Shine Usny Tolgod, Dornogovi Aimag.

Geology and age. Early Cretaceous.

Material and references. One specimen of Trionychidae indet. (as Trionychidae; Suzuki &


34. Shiregin Gashun, Bayankhongor Aimag.



220, 222, 229); shell fragments of Trionychidae indet. (Fig. 16R; collection ZIN PH 167,

collected by G.G. Martinson and N.N. Verzilin in 1978).

Remarks. The shell fragments of Trionychidae indet from Shiregin Gashun have type A

sculpturing.

35. Tel Ulan Shalcha (= Tel Ulan Chaltsai), Dornogovi Aimag.

Geology and age. Unknown formation, probably, early Late Cretaceous.

Material and references. Shell fragments of Trionychidae indet. (Fig. 16S; collection ZIN PH

168, collected by V.F. Shuvalov in 1971; as “shell fragments of rather large trionychids”;

Khosatzky 1999: 144).

Remarks. The shell fragments of Trionychidae indet. from Tel Ulan Shalcha have type A

sculpturing.

36. Tsagan Khushu (= Tsagan Ula), Umunugovi Aimag.


Material and references. Incomplete carapace and left costals 1 and 2 of ‘Trionyx’ gobiensis

(see Systematic palaeontology section); shell fragments of Trionychidae indet. (Fig. 16T;

collection ZIN PH 24, collected by MPE in 1948; as “parts of the plastron with considerably

projecting gastralia and vermiform details of the external relief”; Khosatzky 1999: 147); shell

fragments of Trionychidae indet. (ZPAL MgCh/60, collected by PMPE in 1964); four specimens

of Trionychidae indet. (as Trionychidae; Suzuki & Narmandakh 2004: 10).

Remarks. The shell fragments of Trionychidae indet from Tsagan Khushu (collection ZIN PH

24) have type A sculpturing and correspond in size to Gobiapalone spp. Among these fragments,

a partial right hyo- and hypoplastra (Fig. 16T; ZIN PH 3/24) is similar in its outline to

Gobiapalone spp. That specimen was probably mentioned by Khosatzky (1999) as “parts of the

plastron with considerably projecting gastralia and vermiform details of the external relief”.

Thus, we can recognize two forms of trionychids in Tsagan Khushu: ‘Trionyx’ gobiensis and

Trionychidae indet. (similar to Gobiapalone spp.).

37. Tsagan Teg (= Tsagan Teg Obo), Dornogovi Aimag.


Material and references. Shell fragments of Trionychidae indet. (Fig. 16U, V; collection ZIN

PH 91, collected by V.F. Shuvalov in 1971, G.G. Martinson and N.N. Verzilin in 1978; as “large

trionychid of the Turonian-Santonian type”; Khosatzky 1999: 144); one specimen of

Trionychidae indet. (as Trionychidae; Suzuki & Narmandakh 2004: 8).

Remarks. The shell fragments of Trionychidae indet. from Tsagan Teg demonstrate type A

sculpturing (Fig. 16U; ZIN PH 2/91) and B (Fig. 16V; ZIN PH 3/91). Thus, probably, there are

two forms of trionychids in this locality.

38. Tugrikin Ula (near Khamarin Khural), Dornogovi Aimag.


Material and references. Shell fragments of Trionychidae indet. (Fig. 16W; collection ZIN PH

169, collected by V.F. Shuvalov in 1971; as “remains of medium-sized trionychids”; Khosatzky

1999: 144).

Remarks. The shell fragments of Trionychidae indet. from Tugrikin Ula have type A

sculpturing.

39. Ulan Khushu (= Ulan Bulak), Umunugovi Aimag.


Material and references. Partial shell, partial carapace and partial entoplastron of Gobiapalone

breviplastra (see Systematic palaeontology section); one specimen of Trionychidae indet. (as

Trionychidae; Suzuki & Narmandakh 2004: 9); shell fragments of Trionychidae indet.

(collection PIN 4694, collected by JSMPE); carapace fragment of Trionychidae indet. (ZIN PH

1/170); shell fragments of Trionychidae indet. (collection ZPAL MgCh/69, collected by PMPE

in 1963).

Remarks. The shell fragments of Trionychidae indet. from Ulan Khushu have type A

sculpturing.

40. Ulan Tsab Ula (9 km NW from Ulan Tsab Ula), Bayankhongor Aimag.

Geology and age. Probably, Baynshire Formation, Cenomanian – Santonian.

Material and references. Shell fragments of Trionychidae indet. (as “fragments of medium-

sized trionychids”; Khosatzky 1999: 144).

41. Unegetu Ula, Dornogovi Aimag.


Material and references. Anterior part of carapace of Gobiapalone orlovi (see Systematic


42. Uryl’b Usu, Dornogovi Aimag.



69).

43. Ushyin Khuduk, Dornogovi Aimag.



71).

44. Yagaan Khovil, Umunugovi Aimag.


Material and references. Six specimens of Trionychidae indet. (as Trionychidae; Suzuki &


45. Unknown locality, probably, north-eastern regions of Mongolia.

Geology and age. ?Late Cretaceous.

Material and references. Shell fragments of Trionychidae indet. (Fig. 16X; collection ZIN PH

97, collected by V.N. Chaykovskiy in 1936 from unknown locality, probably, in north-eastern

regions of Mongolia, where he worked as a geologist (Khosatzky 1976)).

Remarks. The probable Late Cretaceous age of the trionychid remains is confirmed by presence

of Nanhsiungchelyidae in the same collection (Danilov & Syromyatnikova 2008). The shell

fragments of Trionychidae indet. from this locality have type A sculpturing.

Appendix 2. Characters coded for Gobiapalone breviplastra and G. orlovi and added to the

matrix of Joyce and Lyson (2011).

Gobiapalone breviplastra: 42224 12211 12124 21122 11??? ????? ????? ????? ????? ?????

????1 ????? ?121? ?0000 00000 02??? ??0

Gobiapalone orlovi: 42224 12211 12124 21222 11?11 ?2111 11211 a31?1 ?1??? ????2 21111

????1 ?1211 ?0000 00000 02000 000

Appendix 3. Characters coded for Gobiapalone breviplastra and G. orlovi and added to the

matrix of Vitek (2012).

Gobiapalone breviplastra: 31113 01200 0a0b1 00110 0???? ????? ????? ????? ????? ????? ???0?

????? 010?? 00000 00000 2???? ?0110 000

Gobiapalone orlovi: 31113 01200 0a0b1 01110 0?00? 10000 0100a 20?0? 0???? ???11 0000?

???0? 0100? 00000 00000 20000 00110 000

Cretaceous soft-shelled turtles (Trionychidae) of Mongolia: new data and a revision

Documents

Transcript of Cretaceous soft-shelled turtles (Trionychidae) of Mongolia: new data and a revision