Prévia do material em texto
See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/311900191
Reevaluation of the largest published morphological data matrix for
phylogenetic analysis of Paleozoic limbed vertebrates
Article · December 2016
DOI: 10.7287/peerj.preprints.1596v2
CITATIONS
0
READS
304
2 authors:
Some of the authors of this publication are also working on these related projects:
Convergent jumping locomotion in terrestrial vertebrate tetrapods View project
Accidental misscores in datasets and their consequences View project
David Marjanović
Museum für Naturkunde - Leibniz Institute for Research on Evolution and Biodiver…
50 PUBLICATIONS 347 CITATIONS
SEE PROFILE
Michel Laurin
French National Centre for Scientific Research
228 PUBLICATIONS 4,360 CITATIONS
SEE PROFILE
All content following this page was uploaded by David Marjanović on 25 December 2016.
The user has requested enhancement of the downloaded file.
https://www.researchgate.net/publication/311900191_Reevaluation_of_the_largest_published_morphological_data_matrix_for_phylogenetic_analysis_of_Paleozoic_limbed_vertebrates?enrichId=rgreq-ae675f05887cbc2b4b6f3eff3795c8ad-XXX&enrichSource=Y292ZXJQYWdlOzMxMTkwMDE5MTtBUzo0NDMyOTIxMTk3NjkwODlAMTQ4MjcwMDQ4NTU3NA%3D%3D&el=1_x_2&_esc=publicationCoverPdf
https://www.researchgate.net/publication/311900191_Reevaluation_of_the_largest_published_morphological_data_matrix_for_phylogenetic_analysis_of_Paleozoic_limbed_vertebrates?enrichId=rgreq-ae675f05887cbc2b4b6f3eff3795c8ad-XXX&enrichSource=Y292ZXJQYWdlOzMxMTkwMDE5MTtBUzo0NDMyOTIxMTk3NjkwODlAMTQ4MjcwMDQ4NTU3NA%3D%3D&el=1_x_3&_esc=publicationCoverPdf
https://www.researchgate.net/project/Convergent-jumping-locomotion-in-terrestrial-vertebrate-tetrapods?enrichId=rgreq-ae675f05887cbc2b4b6f3eff3795c8ad-XXX&enrichSource=Y292ZXJQYWdlOzMxMTkwMDE5MTtBUzo0NDMyOTIxMTk3NjkwODlAMTQ4MjcwMDQ4NTU3NA%3D%3D&el=1_x_9&_esc=publicationCoverPdf
https://www.researchgate.net/project/Accidental-misscores-in-datasets-and-their-consequences?enrichId=rgreq-ae675f05887cbc2b4b6f3eff3795c8ad-XXX&enrichSource=Y292ZXJQYWdlOzMxMTkwMDE5MTtBUzo0NDMyOTIxMTk3NjkwODlAMTQ4MjcwMDQ4NTU3NA%3D%3D&el=1_x_9&_esc=publicationCoverPdf
https://www.researchgate.net/?enrichId=rgreq-ae675f05887cbc2b4b6f3eff3795c8ad-XXX&enrichSource=Y292ZXJQYWdlOzMxMTkwMDE5MTtBUzo0NDMyOTIxMTk3NjkwODlAMTQ4MjcwMDQ4NTU3NA%3D%3D&el=1_x_1&_esc=publicationCoverPdf
https://www.researchgate.net/profile/David_Marjanovic2?enrichId=rgreq-ae675f05887cbc2b4b6f3eff3795c8ad-XXX&enrichSource=Y292ZXJQYWdlOzMxMTkwMDE5MTtBUzo0NDMyOTIxMTk3NjkwODlAMTQ4MjcwMDQ4NTU3NA%3D%3D&el=1_x_4&_esc=publicationCoverPdf
https://www.researchgate.net/profile/David_Marjanovic2?enrichId=rgreq-ae675f05887cbc2b4b6f3eff3795c8ad-XXX&enrichSource=Y292ZXJQYWdlOzMxMTkwMDE5MTtBUzo0NDMyOTIxMTk3NjkwODlAMTQ4MjcwMDQ4NTU3NA%3D%3D&el=1_x_5&_esc=publicationCoverPdf
https://www.researchgate.net/institution/Museum_fuer_Naturkunde-Leibniz_Institute_for_Research_on_Evolution_and_Biodiversity?enrichId=rgreq-ae675f05887cbc2b4b6f3eff3795c8ad-XXX&enrichSource=Y292ZXJQYWdlOzMxMTkwMDE5MTtBUzo0NDMyOTIxMTk3NjkwODlAMTQ4MjcwMDQ4NTU3NA%3D%3D&el=1_x_6&_esc=publicationCoverPdf
https://www.researchgate.net/profile/David_Marjanovic2?enrichId=rgreq-ae675f05887cbc2b4b6f3eff3795c8ad-XXX&enrichSource=Y292ZXJQYWdlOzMxMTkwMDE5MTtBUzo0NDMyOTIxMTk3NjkwODlAMTQ4MjcwMDQ4NTU3NA%3D%3D&el=1_x_7&_esc=publicationCoverPdf
https://www.researchgate.net/profile/Michel_Laurin2?enrichId=rgreq-ae675f05887cbc2b4b6f3eff3795c8ad-XXX&enrichSource=Y292ZXJQYWdlOzMxMTkwMDE5MTtBUzo0NDMyOTIxMTk3NjkwODlAMTQ4MjcwMDQ4NTU3NA%3D%3D&el=1_x_4&_esc=publicationCoverPdf
https://www.researchgate.net/profile/Michel_Laurin2?enrichId=rgreq-ae675f05887cbc2b4b6f3eff3795c8ad-XXX&enrichSource=Y292ZXJQYWdlOzMxMTkwMDE5MTtBUzo0NDMyOTIxMTk3NjkwODlAMTQ4MjcwMDQ4NTU3NA%3D%3D&el=1_x_5&_esc=publicationCoverPdf
https://www.researchgate.net/institution/French_National_Centre_for_Scientific_Research?enrichId=rgreq-ae675f05887cbc2b4b6f3eff3795c8ad-XXX&enrichSource=Y292ZXJQYWdlOzMxMTkwMDE5MTtBUzo0NDMyOTIxMTk3NjkwODlAMTQ4MjcwMDQ4NTU3NA%3D%3D&el=1_x_6&_esc=publicationCoverPdf
https://www.researchgate.net/profile/Michel_Laurin2?enrichId=rgreq-ae675f05887cbc2b4b6f3eff3795c8ad-XXX&enrichSource=Y292ZXJQYWdlOzMxMTkwMDE5MTtBUzo0NDMyOTIxMTk3NjkwODlAMTQ4MjcwMDQ4NTU3NA%3D%3D&el=1_x_7&_esc=publicationCoverPdf
https://www.researchgate.net/profile/David_Marjanovic2?enrichId=rgreq-ae675f05887cbc2b4b6f3eff3795c8ad-XXX&enrichSource=Y292ZXJQYWdlOzMxMTkwMDE5MTtBUzo0NDMyOTIxMTk3NjkwODlAMTQ4MjcwMDQ4NTU3NA%3D%3D&el=1_x_10&_esc=publicationCoverPdf
Reevaluation of the largest published morphological 1
data matrix for phylogenetic analysis of Paleozoic 2
limbed vertebrates 3
4
David Marjanović1, Michel Laurin2 5
6
1 Science Programme “Evolution and Geoprocesses”, Museum für Naturkunde, Leibniz 7
Institute for Evolutionary and Biodiversity Research, Humboldt-Universität zu Berlin, Berlin, 8
Germany 9
2 Centre de Recherches sur la Paléobiologie et les Paléoenvironnements (CR2P), Centre 10
national de la Recherche scientifique (CNRS)/Muséum national d’Histoire naturelle 11
(MNHN)/Université Pierre et Marie Curie (UPMC; Sorbonne Universités), Paris, France 12
13
Corresponding author: 14
David Marjanović1 15
Museum für Naturkunde, Invalidenstraße 43, D-10115 Berlin, Germany 16
E-mail address: david.marjanovic@gmx.at 17
1
1
Abstract 18
19
The largest published data matrix for phylogenetic analysis of Paleozoic limbed vertebrates 20
(Ruta M, Coates MI. 2007. Journal of Systematic Palaeontology 5:69–122) supported some 21
hypotheses that are, to varying degrees, controversial; e.g., it recovered Seymouriamorpha, 22
Diadectomorpha and (in some trees) Caudata as paraphyletic and found the “temnospondyl 23
hypothesis” on the origin of Lissamphibia (TH) to be more parsimonious than the 24
“lepospondyl hypothesis” (LH) – though only, as we show, by one step. 25
We report thousands of suboptimal scores due to typographic and similar errors and to 26
questionable coding decisions: logically linked (redundant) characters, others with only one 27
described state, even characters for which most taxa were scored after presumed relatives. 28
Even continuous characters were unordered, the effects of ontogeny were not sufficiently 29
taken into account, and data published after 2001 were mostly excluded. 30
After these issues are improved – we document and justify all changes to the matrix –, 31
but no characters are removed or added, we find (unconstrained Analysis R1) much longer 32
trees with e.g. monophyletic Caudata, Diadectomorpha and (in some trees) 33
Seymouriamorpha; Ichthyostega rootward of Acanthostega; Anthracosauria rootward of 34
Temnospondyli which includes Caerorhachis; the LH is 9 steps shorter than the TH (R2; 35
constrained) and 12 steps shorter the “polyphyly hypothesis” (PH – R3; constrained). 36
We then added 48 OTUs to the original 102. This destabilizes some parts of the tree, 37
e.g. the positions of Anthracosauria, Temnospondyli and Caerorhachis. Yet, many added taxa 38
have well-resolved positions, ranging from the well known Chroniosaurus (Chroniosuchia), 39
which lies just crownward of Temnospondyli and Gephyrostegidae, to isolated lower jaws. 40
Even though Gerobatrachus, Micropholis and Tungussogyrinus and the extremely 41
peramorphic salamander Chelotriton are added, the difference between LH (R4) and TH (R5) 42
rises to 12 steps, that between LH and PH (R6) to 17 steps; the TH also requires several more 43
regains of lost bones than the LH. Brachydectes (Lysorophia) is not found next to 44
Lissamphibia. 45
We duplicated all analyses after coding losses of bones as irreversible. The impact on 46
the results is modest. Anthracosauria is always rootward of Temnospondyli. With 102 OTUs, 47
the LH (R7) is 10 steps shorter than the TH (R8) and 11 steps shorter than the PH (R9); with 48
150, the LH (R10) is 14 steps shorter than the TH (R12)– and 13 steps shorter than the PH 49
(R11). 50
Bootstrap values are mostly low, and plummet when taxa are added. Statistically, the 51
TH (R2, R5, R8, R12) is not distinguishable from the LH or the PH, but the LH (R1, R4, R7, 52
R10) and the PH (R3, R6, R9, R11) may be distinguishable from each other under both taxon 53
samples and both reversibility settings. A reliable test is not available. 54
We discuss the relationships of certain taxa, approaches to coding, some character 55
complexes, and prospects for further improvement of this matrix. 56
57
2
2
Table of contents 58
59
Introduction 5 60
Background 7 61
Aims 10 62
Abbreviations 11 63
Nomenclature 12 64
Taxonomic nomenclature 12 65
Phylogenetic nomenclature 13 66
Anatomical nomenclature 13 67
Material and methods 14 68
Treatment of characters 14 69
Ordering of multistate characters 14 70
Continuous characters 15 71
Inapplicable characters and mergers 15 72
“Ontogeny discombobulates phylogeny” (Wiens, Bonett & 73
Chippindale, 2005) 17 74
Deleted, recoded and split characters 18 75
Reversibility 18 76
The tail fin skeleton (character CAU FIN 1) 19 77
Modifications to individual cells 20 78
The albanerpetid neck 25 79
Ignored information 25 80
Treatment of OTUs 26 81
Paleothyris/Protorothyris 26 82
Dendrerpetidae 26 83
Rhynchonkos 27 84
The skull roof of Brachydectes 28 85
Taxa added as parts of existing OTUs 28 86
Taxa added as new OTUs for a separate set of analyses 32 87
Taxa that were not added 45 88
Phylogenetic analyses 47 89
Why we only used maximum parsimony 49 90
Robustness analyses 50 91
Results 51 92
Reanalysis of the matrix of Maddin, Jenkins & Anderson (2012) 51 93
Reanalyses of the matrix of Ruta & Coates (2007) 56 94
Analyses of our modified matrix 57 95
Analysis R1 59 96
Analysis R2 61 97
Analysis R3 64 98
Analysis R4 66 99
Analysis R5 68 100
Analysis R6 72 101
Bootstrap analyses B1 and B2 72 102
Analysis R7 75 103
Analysis R8 78 104
Analysis R9 78 105
Analysis R10 82 106
Analysis R11 83 107
3
3
Analysis R12 86 108
Bootstrap analyses B3 and B4 87 109
Discussion 87 110
Phylogenetic relationships 90 111
Devonian taxa and Whatcheeriidae 90 112
Mississippian mysteries 92 113
Colosteidae 93 114
The interrelationships of Anthracosauria, Silvanerpeton, 115
Gephyrostegidae, Caerorhachis and Temnospondyli 95 116
The “Parrsboro jaw” 98 117
Anthracosaurian phylogeny 99 118
Temnospondyl large-scale phylogeny 100 119
Stereospondylomorpha 106 120
Dissorophoidea 107 121
Other added temnospondyls 109 122
Chroniosuchia 111 123
Solenodonsaurus 113 124
Seymouriamorpha 113 125
Amniota and Diadectomorpha 114 126
Westlothiana 115 127
The lepospondyl problem 115 128
The interrelationships of the “microsaurs” 121 129
Added “microsaurs” 122 130
Lissamphibian origins 125 131
Lissamphibian phylogeny 126 132
Characters 127 133
Surprising reversals 127 134
Effects of coding surprising reversals as impossible 131 135
Other recently discussed characters that are included in this 136
matrix 131 137
Preaxial polarity in limb development 132 138
Other characters that are potentially important for the origin of 139
lissamphibians but not considered in this matrix 135 140
Conclusions 135 141
142
Errata in some of our previous publications 137 143
Errata caused by the lack of page proofs at PLOS ONE 139 144
145
Acknowledgments 139 146
147
References 140 148
149
Legends of supplementary information 169 150
151
Appendix 1 171 152
153
Appendix 2 320 154
155
4
4
Introduction 156
157
This ancient inhabitant of the coal swamps of Nova Scotia, was, in short, as we often find to be 158
the case with the earliest forms of life, the possessor of powers and structures not usually, in the 159
modern world, combined in a single species. It was certainly not a fish, yet its bony scales, and 160
the form of its vertebræ, and of its teeth, might, in the absence of other evidence, cause it to be 161
mistaken for one. We call it a batrachian, yet its dentition, the sculpturing of the bones of its 162
skull, which were certainly no more external plates than the similar bones of a crocodile, its ribs, 163
and the structure of its limbs, remind us of the higher reptiles; and we do not know that it ever 164
possessed gills, or passed through a larval or fish-like condition. Still, in a great many important 165
characters, its structures are undoubtedly batrachian. It stands, in short, in the same position with 166
the Lepidodendra and Sigillariæ under whose shade it crept, which though placed by palæo-167
botanists in alliance with certain modern groups of plants, manifestly differed from these in 168
many of their characters, and occupied a different position in nature. In the coal period, the 169
distinctions of physical and vital conditions were not well defined—dry land and water, 170
terrestrial and aquatic plants and animals, and lower and higher forms of animal and vegetable 171
life, are consequently not easily separated from each other. 172
– Dawson (1863: 23–24) about Dendrerpeton acadianum 173
174
Homoplasy Is even More Common than I, or perhaps Anyone, Has ever Imagined 175
– section headline in Wake (2009: 343) 176
177
What is required is a more complete discussion of the character coding of previous data matrices, 178
and a thorough reanalysis based on those matrices. This would enable a well-founded discussion 179
of lissamphibian origins in light of a supermatrix based on all the current and pertinent data. 180
– Sigurdsen & Green (2011: 459) 181
182
Giant morphological data matrices are increasingly common in cladistic analyses of vertebrate 183
phylogeny, reporting numbers of characters never seen or expected before. However, the concern 184
for size is usually not followed by an equivalent, if any, concern for character 185
construction/selection criteria. Therefore, the question of whether quantity parallels quality for 186
such influential works remains open. 187
– Simões et al. (2016a: abstract) 188
189
Not too surprisingly, as it is yet a youthful paradigm shift, modern phylogenetic systematics is 190
still evolving to improve on the lack of precision, rigour and objectivity it inherited from the pre-191
cladistic period. Furthermore, transforming a descriptive science (morphological description) 192
bounded by language as a means of outlining empirical observations into hopefully objectively 193
delimited characters and character-states is a difficult task; every effort to do so is to be 194
commended, while at the same time rigorously scrutinized and improved upon. 195
– Simões et al. (2016a: 18th page) 196
197
Many aspects of biology rely on phylogenies. Phylogenetic trees have allowed biologists to 198
generate and test hypotheses about the evolution of characters and character complexes (e.g. 199
Witzmann, 2007, 2010, 2013, 2015; Fröbisch et al., 2014; Mondéjar-Fernández, Clément & 200
Sanchez, 2014; Miyashita, 2015), including evolutionary changes in habitat (e.g. Laurin & 201
Soler-Gijón, 2010; Laurin, 2011; Laurin & de Buffrénil, 2015); about evolutionary radiations 202
into morphospace (e.g. Ruta, 2009; Fortuny et al., 2011; Anderson, Friedman & Ruta, 2013; 203
Ruta & Wills, 2016); about evolutionary trends or their absence (e.g. Laurin, 2004); about 204
mode and models of evolution (Laurin et al., 2011) as well as rates of diversification and 205
extinction (Ruta, 2011; Bernardi et al., 2016); and about the dates of origin of cladeswhose 206
fossil record is not dense enough to be read literally (Marjanović & Laurin, 2007, 2013a, b, 207
and references therein). Phylogenies also have applications in conservation biology (Faith, 208
1992) and have even been used to solve crimes (Scaduto et al., 2010, and references therein). 209
Phylogenetic trees, however, are themselves hypotheses that attempt to explain a data 210
matrix. Much work has gone, to great success, into the methods for generating and testing 211
5
5
phylogenetic hypotheses from a given data matrix; and a matrix of molecular data can, apart 212
from the problem of alignment, be largely taken for granted as a set of observed facts. But 213
molecular data are not always available. In many cases phylogeneticists have to rely on 214
morphological data – and a matrix of morphological data is a matrix of hypotheses. The 215
characters, their states, and the relationships between the states are hypotheses that rely on 216
hypotheses about homology, about independent evolution from other characters, about 217
ontogeny and even preservation (especially in the case of fossils); the terminal taxa (OTUs) 218
are hypotheses that rely on hypotheses of monophyly, ontogeny and again preservation; and 219
even given all these, each cell in a data matrix is still a hypothesis that relies on hypotheses of 220
homology, ontogeny and preservation – some are close enough to observed facts, others less 221
so. In addition, morphological data matrices can only be compiled by hand – there is no 222
equivalent to sequencer machines or alignment programs. This makes human error inevitable. 223
Consequently, morphological data matrices must not be taken for granted as sets of objective 224
data; the hypotheses of which they consist must be identified and carefully tested. 225
Here we reevaluate the largest published data matrix for the early evolution (Late 226
Devonian to Cisuralian, with a few younger taxa) of the limbed vertebrates (Ruta & Coates, 227
2007 – hereinafter RC07), which has played a large role in shaping current ideas on the 228
phylogeny of these animals. 229
230
231
232
Figure 1: Hypotheses on the origin(s) of the extant amphibians, and their compatibility 233
with molecular and morphological data. Names of extant taxa in boldface in this and all 234
following tree figures. The names Amphibia and Lissamphibia do not apply in E. A: 235
Consensus of all recent molecular studies (e.g. Pyron, 2014: supplementary file amph_shl.tre). 236
B: Part of the consensus of all recent morphological studies (e.g. our results, RC07, and 237
references in both); modern amphibians not shown. C: Extant amphibians added to B 238
according to the “temnospondyl hypothesis”; note compatibility with A. D: “Lepospondyl 239
hypothesis” mapped to B; note compatibility with A. E: “Polyphyly hypothesis” mapped to 240
B; note lack of compatibility with A. In earlier versions of the “polyphyly hypothesis”, 241
Caudata often lies next to Gymnophionomorpha instead of Salientia; in that case, the name 242
Batrachia does not apply (or becomes a synonym of the tetrapod crown-group). 243
6
6
Background 244
245
The early phylogeny of the limbed vertebrates contains a number of open questions on which 246
there is either no consensus, or the existing consensus is weakly supported. Most famously, 247
the origin of the modern amphibians (Lissamphibia and its possible member or sister-group 248
Albanerpetidae – see below on that name) remains a vexing problem (Fig. 1). From the late 249
19th century to now, the modern amphibians have been considered temnospondyls by some 250
(Fig. 1C – “temnospondyl hypothesis”, abbreviated as TH below; most recently found by: 251
RC07; Sigurdsen & Green, 2011; Maddin, Jenkins & Anderson, 2012), lepospondyls by 252
others (Fig. 1D – “lepospondyl hypothesis”, abbreviated as LH below; Vallin & Laurin, 2004; 253
Pawley, 2006: appendix 16; Marjanović & Laurin, 2008, 2009, 2013a), and polyphyletic by 254
yet others (Fig. 1E – “polyphyly hypothesis”, abbreviated as PH below; Carroll, 2007; 255
Huttenlocker et al., 2013), with Salientia being nested among the temnospondyls, 256
Gymnophionomorpha among the lepospondyls, and Caudata either in the lepospondyls (all 257
early works) or in the temnospondyls (works published in the 21st century). 258
Among other more or less open questions are the phylogenies of Temnospondyli (e.g. 259
Pawley, 2006; Ruta, 2009; Dilkes, 2015a), “Lepospondyli” (RC07; Anderson, 2007b; 260
Maddin, Jenkins & Anderson, 2012; Marjanović & Laurin, 2013a) and Devonian limbed 261
vertebrates (e.g., Ahlberg & Clack, 1998; Clack et al., 2012a), as well as the relative positions 262
of Anthracosauria and Temnospondyli (Laurin & Reisz, 1999; RC07) and the position of 263
Diadectomorpha inside or next to Amniota (Berman, Sumida & Lombard, 1992; Berman, 264
2013), not to mention the positions of confusing (Andrews & Carroll, 1991; Smithson et al., 265
1994; Clack, 2001; Ruta, Milner & Clack, 2002; Vallin & Laurin, 2004; RC07) or 266
fragmentary Carboniferous taxa (Smithson, 1980; Paton, Smithson & Clack, 1999; Bolt & 267
Lombard, 2006; Clack et al., 2012b; Sookias, Böhmer & Clack, 2014). 268
Molecular data are of limited use for tackling these questions: of all tetrapodomorphs 269
(tetrapods and everything closer to them than to lungfish), only frogs, salamanders, caecilians 270
and amniotes still have living members (Fig. 1). Thus, molecular data cannot test, say, 271
whether Amniota is closer to Anthracosauria or to Temnospondyli, or if the many disparate 272
taxa classified as “Lepospondyli” constitute a clade, a grade, or a wastebasket. When it comes 273
to the origins of the extant amphibians, molecular data can distinguish the PH (Fig. 1E) from 274
the other hypotheses, because the PH predicts that the extant amphibians are paraphyletic with 275
respect to Amniota, while the other two hypotheses of course predict monophyly under recent 276
conceptions of the affinities of temnospondyls, lepospondyls and amniotes (Fig. 1B; Laurin, 277
2002; see also Marjanović & Laurin, 2007, 2013a); but molecular data cannot distinguish 278
between the two monophyly hypotheses, because too many relevant taxa have been extinct for 279
too long. Finding the extant amphibians monophyletic with respect to Amniota (Fig. 1A), 280
analyses of molecular data support both monophyly hypotheses equally (Fig. 1A, C, D); only 281
paleontological data can distinguish them. Existing analyses of paleontological data, however, 282
disagree greatly on this question (see above; Fig. 1C–E) as well as on others. To some extent, 283
no doubt, this is due to the many differences in their taxon and character samples. Another 284
possible reason, however, are mistakes in datasets. Analogously to the problem of alignment 285
in molecular analyses, morphological analyses begin with the construction of a dataset, where 286
characters need to be defined in ways that prevent them from being redundant. (When the 287
state of a character is predictable from the state of another character, the characters are 288
redundant for the purposes of phylogenetic analysis; to use redundant characters amounts to 289
counting the same character twice, doubling its influence on the results.) Observations need to 290
be interpreted in terms of these characters, and then these interpretations need to be inserted in 291
the data matrix by hand. In our own practice, we have every once in a while caught ourselves 292
making typographic errors, being momentarily confused about which state is 0 and which is 1 293
7
7
(because of faulty memory as well as conflicting conventions on how to assign such numbers 294
– 0 can for instance mean “absent” or “presumably plesiomorphic”), inserting the right value 295
in the wrong column or line, and committing similar blunders; additionally, we have on 296
occasion misinterpretedthe descriptive literature and its illustrations (line drawings, but even 297
photographs, can give misleading three-dimensional impressions), overlooked poorly known 298
publications, had language barriers or conflicting terminologies prevent us from being sure if 299
a published sentence said one thing or the opposite, or simply relied on the then current state 300
of research that was later overturned when the next publication came out. It stands to reason 301
that these things also happen to other people. When such mistakes are removed from their 302
matrices – without, as far as possible, changing the taxon or character sample –, do the results 303
change? 304
The largest published morphological data matrix that has been applied to the problems 305
of the phylogeny of limbed vertebrates in general and the origin of the modern amphibians in 306
particular is that by RC07; it supported the TH and is often cited for this result. (Further, it 307
was used unchanged by Bernardi et al., 2016, to study diversification patterns.) Here we 308
reevaluate this matrix, as we have done (Marjanović & Laurin, 2008, 2009) for those by 309
McGowan (2002) and Anderson et al. (2008a), in order to test if this result – and others that 310
together constitute the consensus tree of RC07 (fig. 5, 6; our Fig. 2) – continues to follow 311
from a presumably improved version of their dataset. 312
Our changes to the matrix are documented and justified in Appendix 1. The revised 313
matrix is presented in human-readable form as Appendix 2, where the changes are highlighted 314
in color, and in NEXUS format as Data S1 and S2 (these two files differ in analysis settings 315
as explained below). 316
This work began as chapter V of the doctoral thesis of Damien Germain supervised by 317
M. L. (Germain, 2008a). D. Germain did not have time to complete this large task, so 318
additional work came to be done as chapter 5 of the doctoral thesis of D. M., again supervised 319
by M. L. (Marjanović, 2010; parts were presented in Marjanović & Laurin, 2013a). This, too, 320
remained incomplete for lack of time. Further work was later done by D. M.; this is presented 321
here for the first time. 322
323
324
325
↓ Figure 2: Strict consensus of the MPTs (length: 1621 steps including polymorphisms) 326
found by RC07 and by our unconstrained reanalysis of their unchanged data matrix 327
(Analysis O1). The topology is identical to RC07: fig. 5, 6. In this figure and all following 328
ones, Albanerpetidae and Dendrerpetidae are each a single OTU (called “Albanerpetontidae” 329
and “Dendrerpeton” by RC07). The name Edopoidea (and, in some other figures, 330
Eutemnospondyli) is placed under the assumption that **Mastodonsaurus (the two asterisks 331
indicate that this taxon is not included in any of our analyses) would be closest to Eryops as 332
found by Schoch (2013); “Dissorophoidea (S13)” is placed according to Schoch’s (2013) 333
definition, “Dissorophoidea (YW00)” is placed according to the definition by Yates & 334
Warren (2000), and “Dissorophoidea (content)”, only shown where different from 335
Dissorophoidea (YW00), is the smallest clade that has the traditional contents of that taxon 336
(trematopids, Broiliellus, amphibamids/branchiosaurids, Micromelerpeton). The origin of 337
digits cannot be narrowed down to a single internode in this and several other figures. For 338
easier orientation, Ichthyostega, Baphetes and Eryops are written in purple. The color scheme 339
of the background boxes is consistent across all figures. Occasional abbreviations: 340
Gymnophio., Gymnophionomorpha; Liss., Lissamphibia; micr. or microbr., 341
microbrachomorphs; Ph., Pholiderpeton. All “microsaurs” are underlain in the same shade of 342
gray, but some (in some figures) are not labeled as such due to lack of space. 343
8
8
Eusthenopteron
Panderichthys
Ventastega
Acanthostega
Tulerpeton
Colosteus
Greererpeton
Baphetes
Megalocephalus
Eucritta
Edops
Chenoprosopus
Cochleosaurus
Eryops
Capetus
Balanerpeton
Dendrerpetidae
Acheloma
Phonerpeton
Ecolsonia
Broiliellus brevis
Amphibamus
Doleserpeton
Albanerpetidae
Karaurus
Valdotriton
Triadobatrachus
Notobatrachus
Vieraella
Eoscopus
Platyrhinops
Micromelerpeton
Apateon
Leptorophus
Schoenfelderpeton
Isodectes
Trimerorhachis
Neldasaurus
Caerorhachis
Eocaecilia
Acherontiscus
Adelospondylus
Adelogyrinus
Dolichopareias
Ptyonius
Urocordylus
Sauropleura
Lethiscus
Oestocephalus
Phlegethontia
Batropetes
Brachydectes
Scincosaurus
Keraterpeton
Diceratosaurus
Batrachiderpeton
Diplocaulus
Diploceraspis
Tuditanus
Asaphestera
Pantylus
Stegotretus
Saxonerpeton
Hapsidopareion
Rhynchonkos/Aletrimyti/Dvellecanus
Micraroter
Pelodosotis
Cardiocephalus
Euryodus
Odonterpeton
Microbrachis
Hyloplesion
Westlothiana
Diadectes
Orobates
Tseajaia
Limnoscelis
Captorhinus
Paleothyris/Protorothyris
Petrolacosaurus
Kotlassia
Discosauriscus
Ariekanerpeton
Leptoropha
Microphon
Utegenia
Seymouria
Solenodonsaurus
Bruktererpeton
Gephyrostegus
Silvanerpeton
Eoherpeton
Proterogyrinus
Archeria
Pholiderpeton scutigerum
Pholiderpeton attheyi
Anthracosaurus
Crassigyrinus
Whatcheeria
Pederpes
Ossinodus
Ichthyostega
Whatcheeriidae
Baphetoidea
Anthracosauria
tetrapod crown-group
Amphibia
Batrachia
Pantylidae
Aïstopoda
Urocordylidae
Diplocaulidae
Salientia
Gymnarthridae
Ostodolepididae
Microbrachomorpha
Sauropsida
“diadectomorphs”
“caudates”
“microsaurs”
“seymouriamorphs”
“gephyrostegids”
“amphibamids”
Colosteidae
adelospondyls Adelogyrinidae
Branchiosauridae
Gymnophionomorpha
Lysorophia
Holospondyli
Embolomeri
Lepospondyli
Edopoidea
Dvinosauria
Dissorophoidea (S13)
Dissorophoidea (YW00)
Trematopidae
origin of digits
Lissamphibia
In the meantime, Sigurdsen & Green (2011) performed their own reevaluation of the 344
matrix of RC07. That work had a very limited scope (see Marjanović & Laurin, 2013a, for 345
discussion). We have incorporated most, though not all, of the changes to individual cells 346
suggested in it (as discussed under the respective characters in Appendix 1). Unlike Sigurdsen 347
& Green (2011), we have not deleted characters of unclear value. 348
349
Aims 350
351
The present work cannot pretend to solve the question of lissamphibian origins or any other of 352
the controversies in the phylogeny of early limbed vertebrates (of which there are many, as 353
we will discuss). It merely tries to test, and explain within the limitations of the dataset, to 354
what degree the trees found by RC07 still follow from their matrix – the largest published one 355
that has been applied to those questions – after a thorough effort to improve the accuracy of 356
the scoring and reduce character redundancy has been carried out to the best of our current 357
knowledge. However, we think this effort forms a necessary step towards solving any of those 358
problems. Further progress may come from larger matrices that will contain more taxa and 359
many more characters – if and only if the increase in the number of cells is not accompanied 360
by a proportional decrease in the care that goes into scoring them (Simões et al., 2016a). 361
Originally we did not intend to add any taxa to the matrix, just as we have not added 362
any characters. However, soon after 2007, the intriguing amphibamid temnospondyl 363
Gerobatrachus was published and was argued to add strong support to the PH (Anderson et 364
al., 2008a). Phylogenetic analyses that included Gerobatrachus in different versions of the 365
same matrix have supported the PH (Anderson et al., 2008a), the LH (Marjanović & Laurin, 366
2009), or more recently the TH (Maddin & Anderson, 2012; Maddin, Jenkins & Anderson, 367
2012); the latter work even found Gerobatrachus to be nested within Lissamphibia. We 368
therefore anticipated that, should our reappraisal of the matrix of RC07 support the LH (or 369
any other hypothesis, for that matter),readers and reviewers could judge such a result useless 370
and outdated because the matrix necessarily lacked Gerobatrachus. Following the example of 371
Marjanović & Laurin (2008), we have therefore performed a separate series of analyses for 372
which we added Gerobatrachus to the matrix; at that opportunity, for the same series of 373
additional analyses, we added a further 47 OTUs as detailed in the Methods section, bringing 374
the total from 102 to 150. 375
In the Discussion section, we explore the relationships of certain taxa and the 376
distributions of certain character states in the light of our findings and other recent 377
publications (including published conference abstracts that were peer-reviewed before 378
presentation). By presenting the current areas of uncertainty (some expected, some 379
unexpected) in detail, we hope to highlight opportunities for future research. 380
A quick look at matrices such as those of Ruta (2009), Sigurdsen & Green (2011), 381
Maddin, Jenkins & Anderson (2012), Clack et al. (2012b), Sookias, Böhmer & Clack (2014) 382
or Dilkes (2015a), or at reinvestigations of anatomy such as those of Witzmann (2007, 2010, 383
2013), Bolt & Lombard (2010), Mondéjar-Fernández, Clément & Sanchez (2014) or Dilkes 384
(2015a) – see also Pardo (2014), Pardo, Szostakiwskyj & Anderson (2015a) and Pardo & 385
Anderson (2016) – will demonstrate that many characters that are known to carry a 386
phylogenetic signal for the present taxon sample remain absent from this matrix; this even 387
includes some of the only 41 characters used by McGowan (2002). Adding them (as well as 388
yet more taxa) will be part of future work, and may well lead to trees with a different 389
topology. However, the present work is a necessary first step. 390
391
392
393
10
10
Abbreviations 394
395
AMNH: properly AMNH DVP or AMNH FARB – Department of Vertebrate Paleontology or 396
Collection of Fossil Amphibians, Reptiles and Birds at the American Museum of Natural 397
History, New York. 398
BEG: Vertebrate Paleontology Laboratory, The University of Texas at Austin (formerly 399
Bureau of Economic Geology). 400
BPE: bootstrap percentage given the expanded taxon sample and full reversibility of bone 401
losses (bootstrap analysis B2). 402
BPIE: bootstrap percentage given the expanded taxon sample and irreversibility of bone 403
losses (bootstrap analysis B4). 404
BPIO: bootstrap percentage given the original taxon sample and irreversibility of bone losses 405
(bootstrap analysis B3). 406
BPO: bootstrap percentage given the original taxon sample and full reversibility of bone 407
losses (bootstrap analysis B1). 408
CG78: Carroll & Gaskill (1978). 409
CM: Carnegie Museum of Natural History, Pittsburgh. 410
LH: “lepospondyl hypothesis” on the origin of lissamphibians. 411
MB.Am.: “amphibian” in the vertebrate paleontology collection of the Museum für 412
Naturkunde, Berlin. 413
MCZ: Museum of Comparative Zoology, Harvard University, Cambridge (Massachusetts). 414
MGUH: Geologisk Museum, Statens Naturhistoriske Museum, Københavns Universitet, 415
Copenhagen (“Museum Geologicum Universitatis Hafniensis”). 416
MNHN: Muséum national d’Histoire naturelle, Paris. 417
MPT: most parsimonious tree. 418
NHMW: Naturhistorisches Museum Wien, Vienna. 419
NMS: National Museum of Scotland, Edinburgh. 420
NSM: Nova Scotia Museum, Halifax. 421
OTU: operational taxonomic unit (a line in a data matrix). 422
PH: “polyphyly hypothesis” on the origin of “lissamphibians”. 423
PIN: Paleontological Institute of the Academy of Sciences of the Russian Federation, 424
Moscow. 425
PU-ANF: “amphibian” specimens from Puertollano (Spain) in the vertebrate collection of the 426
Department of Paleontology, Geology division, Universidad Complutense, Madrid; formerly 427
kept at the Museum für Naturkunde, Berlin. 428
RC07: Ruta & Coates (2007). 429
TH: “temnospondyl hypothesis” on the origin of lissamphibians. 430
TMM: Vertebrate Paleontology Laboratory, The University of Texas at Austin (formerly 431
Texas Memorial Museum). 432
USNM: National Museum of Natural History, Smithsonian Institution, Washington (DC). 433
YPM: Yale Peabody Museum, New Haven (Connecticut). 434
435
Abbreviations not listed here that consist of at least three letters, at least one space and a 436
number designate characters, following the practice of RC07; see Appendix 1. Note that 437
merged characters keep the abbreviations of all their constituents: e.g., PREMAX 1-2-3 438
(character 1) is built from the three characters PREMAX 1, PREMAX 2 and PREMAX 3 of 439
RC07, and MAX 5/PAL 5 (character 22) consists of MAX 5 and PAL 5 of RC07. 440
441
442
443
11
11
Nomenclature 444
445
A few remarks are necessary to explain our use of certain terms. 446
447
Taxonomic nomenclature 448
449
Without mentioning the fact that they were doing so, Averianov & Sues (2012: 466) corrected 450
the spelling of Albanerpetontidae Fox & Naylor, 1982, to Albanerpetidae. We follow this in 451
analogy to several corrections by Martín, Alonzo-Zarazaga & Sanchiz (2012) as discussed by 452
Marjanović & Laurin (2013b: 543), who did not know of Averianov & Sues (2012) and 453
therefore incorrectly claimed that “no other spelling [than Albanerpetontidae] has ever been 454
used”, as well as in analogy to several further corrections by Schoch & Milner (2014). 455
Assuming that this correction is a “justified emendation”, the name Albanerpetidae must 456
continue to be attributed to Fox & Naylor, 1982 (International Commission on Zoological 457
Nomenclature, 1999: articles 19.2, 29). 458
Likewise, the spelling Hapsidopareiontidae must be corrected to Hapsidopareiidae 459
Daly, 1973: there is no basis for -ont- in Homeric Greek παρήϊον (“cheek”). This name has 460
been used so rarely that the question of common usage does not arise. 461
There is little if any room for doubt about the fact that the species name Trihecaton 462
howardinus is automatically emended to T. howardinum Vaughn, 1975 (International 463
Commission on Zoological Nomenclature, 1999: article 31.2 = 34.2), because Trihecaton 464
(meaning “300”; a numeral, not a noun) is – if anything – neuter. 465
There is no room for doubt about the fact that the same Article emends 466
Madygenerpeton pustulatus to Madygenerpeton pustulatum Schoch, Voigt & Buchwitz, 2010. 467
Previously, Tryphosuchus contained T. paucidens (type species) and T. kinelensis, and 468
Konzhukovia contained K. vetusta (type species) and K. tarda. Pereira Pacheco et al. (2016) 469
found T. paucidens and K. vetusta as sister-groups, which together form the sister-group of K. 470
tarda. Consequently, Pereira Pacheco et al. (2016) wondered if T. paucidens should be 471
referred to Konzhukovia. This is not possible: Tryphosuchus Konzhukova, 1955, has priority 472
over Konzhukovia Gubin, 1991, which only the International Commission on Zoological 473
Nomenclature could overturn. The fact that Pereira Pacheco et al. (2016) considered T. 474
paucidens a nomen dubium has no bearing on this: first, to declare a name a nomen dubium is 475
a taxonomic (subjective) act, not a nomenclatural (objective) one; second, a taxon name does 476
not automatically become a nomen dubium if the name of its type subtaxon is one. 477
Konzhukovia and Tryphosuchus were previously included in Tryphosuchinae Golubev, 478
1995 (type genus: Tryphosuchus), within Melosauridae Fritsch, 1885 (type genus: 479
Melosaurus). Pereira Pacheco et al. (2016: fig. 5) confirmed that Konzhukovia and 480
Tryphosuchus form a clade to which Melosaurus does not belong. However, because they 481
found this clade to lie closer to Stereospondyli than to Melosaurus, they did not refer it to 482
Melosauridae; instead, “we erect the new family Konzhokoviidae [sic!] fam. nov. to replace 483
Tryphosuchinae” (11th page). This, too, is not possible without an Opinion by the 484
Commission.By means of the Principle of Coordination (Article 36.1), when the name 485
Tryphosuchinae was created at the subfamily rank, the name Tryphosuchidae at the family 486
rank was automatically created at the same time; it must be attributed to Golubev, 1995, and 487
therefore will have priority over Konzhukoviidae once the latter is validly published. (We 488
should note that Konzhukoviidae is so spelled in the title, the abstract, and five times in the 489
text, of which two are marked “fam. nov.”, including the one at the beginning of the 490
“Systematic palaeontology” section; the misspelling Konzhokoviidae occurs five times in the 491
text as well, three of which are marked “fam. nov.”.) Because Konzhukoviidae is (trivially) 492
not in prevailing usage, the exception constituted by Article 35.5 will not apply. – Valid 493
12
12
publication of the names Konzhukoviidae and Konzhukovia sangabrielensis will occur when 494
the paper by Pereira Pacheco et al. (2016) will appear in print; the online-early version does 495
not contain evidence of having been registered in ZooBank and is therefore not considered 496
published for purposes of nomenclature (Article 8.5.3 as amended in 2012). For the same 497
reason, these names are not available from the present paper either. 498
499
Phylogenetic nomenclature 500
501
Because of the length of this paper and because the International Code of Phylogenetic 502
Nomenclature (“PhyloCode”) has not yet taken effect, we refrain from proposing new clade 503
names or definitions. Having, however, noticed that the names Adelogyrinidae and 504
Adelospondyli currently refer to the same taxon, we follow RC07: 81 (but not fig. 5) and 505
Coates, Ruta & Friedman (2008: fig. 2: “Adelospondyli”) in informally referring to a clade 506
composed of Acherontiscus and Adelogyrinidae as “adelospondyls” for brevity. 507
As in a previous paper (Marjanović & Laurin, 2013a), where we failed to make this 508
explicit, we use “modern amphibians” for Lissamphibia and its possible member or sister-509
group Albanerpetidae. 510
Several names for temnospondyl clades, including Temnospondyli itself, have been 511
given very different definitions by Yates & Warren (2000) and Schoch (2013). We have 512
generally applied the definitions by Yates & Warren (2000). The important exceptions are 513
Dissorophoidea and Stereospondylomorpha, which we use in the text for clades which contain 514
all those OTUs that have traditionally been regarded as members of clades with these names 515
and exclude most or all OTUs that have traditionally not been regarded as members. For 516
Dissorophoidea, we indicate both definitions and our usage in the tree figures (our usage 517
usually, but not always, coincides with the definition by Yates & Warren, 2000). 518
519
Anatomical nomenclature 520
521
Consistently or nearly so, RC07 (as also Pardo, Szostakiwskyj & Anderson, 2015b) 522
exchanged the terms “skull roof” and “skull table”. The former refers to the dermatocranium 523
except its ventral side (the palate); the latter refers to the dorsal side of the skull roof, often 524
demarcated from the lateral sides by distinct edges. 525
RC07 (and Ruta, Coates & Quicke, 2003) also almost consistently used “mesial” when 526
they were aiming at “medial”. This appears to be a common British practice; internationally, 527
however, it is “medial” that means “toward the sagittal plane at a right angle to it”, while 528
“mesial” belongs to toothrow nomenclature and means “toward the symphysis along the 529
curvature of the jaw”, the opposite of the poorly chosen term “distal”. Many instances of 530
“mesial” apply to the lower jaw and actually mean “lingual”, “proximal at a 90° angle to the 531
curvature of the jaw”, the opposite of “labial”; only at the symphysis is mesial medial. 532
We have tried to rectify these issues in the annotated character list (Appendix 1). 533
However, we use the pairs “anterior – posterior” and “cranial/rostral – caudal” 534
interchangeably (despite a preference for the latter) because there is, in our taxon sample, no 535
danger of confusion. 536
Our ambiguous usage of “orbit” (Marjanović & Laurin, 2008, 2013a) confused Pardo 537
& Anderson (2016), who claimed that we had first thought that Brachydectes had very large 538
eyes which filled its orbitotemporal fenestrae (ascribed to our 2008 paper) and that we had 539
later amended this view (ascribed to 2013a); in reality, we explicitly stated (2008: 194–195) 540
that we considered the “orbits (or ‘orbitotemporal fenestrae’ as they are sometimes called in 541
salientians and caudates)” to have “presumably” contained jaw muscles in Brachydectes, not 542
just the eyeball, and we implied no changes in the later paper (Marjanović & Laurin, 2013a: 543
13
13
239, 241). – In the present work we consistently speak of “orbitotemporal fenestrae” to mean 544
skull openings that appear to have contained the eyes as well as jaw muscles, regardless of the 545
inferred homologies of these openings. 546
547
Material and methods 548
549
The data matrices were edited in successive versions of Mesquite up to 3.03 (Maddison & 550
Maddison, 2015); this program was also used to display and visually compare trees and to 551
optimize characters on them. 552
553
Treatment of characters 554
555
Character-taxon matrices and their accompanying character lists should be viewed as formatted 556
data, and not just a table of observations. That is, they should be constructed with an 557
understanding of how that information will be interpreted by the algorithm that is receiving 558
them. For many multistate characters, authors should consider how character state information is 559
(or is not) distributed to other transformation series. The problem with ‘0’ being used as a catch-560
all for anything that simply ‘isn’t 1’ should be borne in mind when using binary characters (see 561
discussion below). 562
– Brazeau (2011: 494) 563
564
Characters in a data matrix for phylogenetic analysis are interdependent when a state of a 565
character (other than “unknown”) is predictable – without prior knowledge of phylogeny – 566
from the state of another character. Because phylogenetic analysis operates on the assumption 567
that all characters are independent of each other, the presence of interdependent characters in 568
a matrix amounts to counting the same apomorphy at least twice, which can distort the 569
resulting tree topology or, in less extreme cases, its support values. While this fact seems to 570
be universally acknowledged in principle, we find (Marjanović & Laurin, 2008, 2009, 2013a; 571
and below) that it is underappreciated in practice. 572
Different kinds of character interdependence require different amounts of effort to 573
detect. O’Keefe & Wagner (2001: 657; and references therein) distinguished “logical 574
correlations among characters” from “[b]iological correlations”; Pardo (2014: 52–60) 575
distinguished four kinds of interdependence in his unpublished MSc thesis. We call Pardo’s 576
first three kinds, which include logical interdependence, “redundancy” and biological 577
interdependence “correlation” hereinafter. 578
It can be very difficult to determine whether characters are correlated; studies of 579
development genetics are sometimes, perhaps often, required (Kangas et al., 2004; Harjunmaa 580
et al., 2014). We expect, therefore, that all of our best efforts will be unable to completely 581
eliminate character interdependence from any morphological matrix. However, many cases of 582
redundant characters are much more obvious; and although RC07 noticed and removed 583
several cases from the preceding version (Ruta, Coates & Quicke, 2003), we present a 584
considerable number of additional cases in Appendix 1. 585
586
Ordering of multistate characters 587
588
RC07 (p. 78), like Ruta, Coates & Quicke (2003: 271), followedthe widespread practice (e.g. 589
Sigurdsen & Green, 2011; Schoch, 2013; stated but unexplained preference of Simões et al., 590
2016a: 14th page) of treating all multistate characters as unordered. Following another 591
widespread practice, they did not spell out any reasons for this decision. Presumably they 592
adhered to the common assumption (already identified and discussed by Slowinski, 1993) that 593
ordering a character is to make an assumption, while “leaving” it unordered means not to 594
14
14
make an assumption – which is incorrect (Slowinski, 1993). In particular, potentially 595
continuous (clinal) characters should be ordered, because the basic assumption behind 596
ordering – that it is easier to change from any state to a similar state than to a less similar one 597
– has already been used to partition the observed spread of data into discrete states in the first 598
place; it would be incoherent to reject this assumption in one place but not the other 599
(Slowinski, 1993; Wiens, 2001). This also holds for certain meristic characters (Wiens, 2001). 600
As advocated by Slowinski (1993) and Wiens (2001), we have ordered many, but not all, 601
multistate characters; see Appendix 1 for discussion of each case. Four characters (numbers 602
32, 91, 133, 275: PAR 2/POSFRO 3/INTEMP 1/SUTEMP 1, PIN FOR 2, EXOCC 2-3-4-603
5/BASOCC 1-5, DIG 1-2-3-4) have part of their state range ordered (e. g. Slowinski, 1993: 604
fig. 1a, d); this is accomplished by creating stepmatrices (Appendix-Tables 2, 5, 8, 9). We 605
have marked these decisions in the name of each multistate character by adding “(ordered)”, 606
“(unordered)” or “(stepmatrix)” to its end. 607
The consequences of ordering potentially continuous characters are largely 608
unpredictable. Empirically, ordering such characters can reveal additional signal and thus 609
increase the resolution of the consensus tree (Slowinski, 1993; Fröbisch & Schoch, 2009a; 610
Grand et al., 2013; Simões et al., 2016a: fig. 2b, 3b, 4); on the other hand, and even at the 611
same time, it can also reveal previously hidden character conflict and thus decrease the 612
resolution, showing that the original resolution was not supported by the data (Slowinski, 613
1993; Marjanović & Laurin, 2008). Both of these results are congruent with the finding of the 614
simulation studies by Grand et al. (2013) and Rineau et al. (2015) that ordering clinal 615
characters decreases the rate of artefactual resolution and increases the power to detect real 616
clades. 617
618
Continuous characters 619
620
Due to the limits of available computation power, we have not been able to analyze 621
continuous characters “as such” (as recommended by Simões et al., 2016a; Brocklehurst, 622
Romano & Fröbisch, 2016; and references therein), but have broken them up into a small 623
number of discrete states. Such characters were ordered if multistate (see above). Where 624
RC07 had defined reproducible state limits that did not render the characters parsimony-625
uninformative, we have not changed them; in the cases of PREMAX 7 and SKU TAB 1, 626
which we have wholly recoded (see below: “Deleted, recoded and split characters”), we 627
divided the observed range into several equal parts (not many – Bardin et al., 2014) at round 628
numbers because there are no gaps in the distribution. This is an obvious opportunity for 629
future improvement. See Brocklehurst, Romano & Fröbisch (2016) for further discussion. 630
631
Inapplicable characters and mergers 632
633
Perhaps the eventual solution will be to write new algorithms for computer programs that will 634
allow the characters to be coded independently but that will consider interactions between 635
characters and count steps in some characters only on those portions of the tree on which they 636
are applicable. 637
– Maddison (1993: 580) 638
639
It is a common situation that a character is inapplicable to part of a taxon sample: when, say, 640
the ectopterygoid bone in the palate is absent, it does not have a size or a shape, and it is not 641
toothed or toothless. Strong & Lipscomb (1999) reviewed the methods for dealing with such 642
cases: 643
Reductive coding: inapplicable scores are treated as missing data, using the symbols 644
“?” or “-”. (The latter designates a gap in a molecular sequence, which is treated either 645
15
15
as missing data – the default setting in available software – or as a fifth nucleotide/21st 646
amino acid – which is absence coding, see below.) 647
Composite coding: characters that are inapplicable to any OTU are merged with the 648
characters they depend on, producing multistate characters. For instance, 649
“ectopterygoid toothed (0); toothless (1)” might be merged with “ectopterygoid 650
present (0); absent (1)” as the unordered multistate character “ectopterygoid toothed 651
(0); toothless (1); absent (2)”. 652
Nonadditive binary coding: the presence or absence of each character state is treated 653
as a binary character of its own that can be scored for all OTUs. For example, “toothed 654
ectopterygoid: present (0); absent (1)” and “toothless ectopterygoid: present (0); 655
absent (1)” can be scored for all OTUs – taxa that lack an ectopterygoid have state 1 of 656
both of these characters. 657
Absence coding: inapplicability is coded as an additional state of each concerned 658
character. 659
As Strong & Lipscomb (1999) pointed out, absence coding creates redundant characters by 660
counting each condition that makes characters inapplicable several times, once for each 661
character that it makes inapplicable. Absence coding must therefore be avoided. Nonadditive 662
binary coding runs the same risk (see “Errata in some of our previous publications” below). It 663
also treats nonhomologous conditions as the same state, which runs the risk of creating 664
spurious apomorphies which may increase artefactual resolution (Brazeau, 2011, and 665
references therein); Strong & Lipscomb (1999: 368–369) went so far as to state that “non-666
additive binary coding denies homology and the hierarchical relationships between states. The 667
result are cladograms and character interpretations that are absurd and inaccurate 668
representations of our observations.” Brazeau (2011: 495) further performed a reductio ad 669
absurdum by asking what would happen if molecular data were coded as, e.g., “adenine at site 670
121: absent (0); present (1)”, where state 0 would treat guanine, cytosine, thymine, and the 671
absence of site 121 (a deletion) as identical and primarily homologous for no defensible 672
reason. 673
Composite coding has the advantage over reductive coding that it prevents 674
contradictory optimization of ancestors. With reductive coding, an ancestor may be optimized 675
as having a large ectopterygoid and at the same time as having no ectopterygoid at all; 676
composite coding makes this impossible and may thus make arguably nonsensical trees less 677
parsimonious. However, composite coding is not always feasible to its full extent. In our 678
example, if the character that represents the presence of the ectopterygoid is merged with the 679
character that describes its dentition, what is to become of size and shape? Merging all of the 680
affected characters would often yield a huge multistate character that would, in many cases, 681
require a complex stepmatrix and be difficult to interpret in evolutionary terms (as already 682
noted by Maddison, 1993). 683
RC07 used reductive, composite and nonadditive binary coding in different cases. The 684
numerous occurrences of nonadditive binary coding may be a side effect of taking characters 685
from diagnoses and synapomorphy lists (Bardin, Rouget & Cecca, 2014), where they are 686
usually presented in the form of a single (presumedly derived) state, while the other states are 687
not mentioned. We have greatly reducedthe amount of nonadditive binary coding. There are a 688
few cases we have left for the time being; these are cases where the effort needed to rectify 689
the situation would probably be out of proportion to the gain in phylogenetic signal. Examples 690
are PREMAX 4, “Premaxilla with flat, expanded anteromedial dorsal surface and elongated 691
along its lateral margin but not along its medial margin, when observed in dorsal aspect: 692
absent (0); present (1)”, PASPHE 11, “Basipterygoid processes of the basisphenoid shaped 693
like anterolaterally directed stalks, subtriangular to rectangular in ventral view and projecting 694
anterior to the insertion of the cultriform process: absent (0); present (1)”, TIB 6, “Outline of 695
16
16
tibia medial margin shaped like a distinct, subsemicircular embayment contributing to 696
interepipodial space and the diameter of which is less than one-third of bone length: absent 697
(0); present (1)”, and TRU VER 30, “Transverse processes stout and abbreviated, the length 698
of which is less than 30% of neural arch height: absent (0); present (1)”. We would need to 699
survey the range of morphologies and in some cases their considerable ontogenetic 700
transformations – see below – in detail in order to determine how many characters, let alone 701
how many states, should be distinguished within state 0 of each of these. All these characters 702
should, however, be reinvestigated in the future. 703
In a few cases we have changed fully reductive to partially composite coding. This 704
includes, for instance, the following characters of RC07: 705
140. ECT 1: Separately ossified ectopterygoid: present (0); absent (1). 706
141. ECT 2: Ectopterygoid with (0) or without (1) fangs […]. 707
142. ECT 3: Ectopterygoid without (0) or with (1) small teeth (denticles) […]. 708
143. ECT 4: Ectopterygoid longer than/as long as (0) or shorter than (1) palatine. 709
144. ECT 5: Ectopterygoid with (0) or without (1) row of teeth (3+) […]. 710
145. ECT 6: Ectopterygoid/maxilla contact: present (0); absent (1). 711
146. ECT 7: Ectopterygoid narrowly wedged between palatine and pterygoid: no (0); 712
yes (1). 713
The presence (ECT 1) and size of the ectopterygoid (ECT 4) are now a single character, to 714
which we have added an additional state (ultimately from McGowan, 2002): 715
114. ECT 1-4: Ectopterygoid at least as long as palatine (0); at least about a third as 716
long as but shorter than palatine (1); at most about a third as long as palatine (2); 717
absent (3) (ordered). 718
The interpretation of ECT 1 and ECT 4 as parts of a potentially continuous character fits the 719
observation that taxa with very small ectopterygoids tend to be the closest relatives of those 720
with no ectopterygoids and nested among taxa with middle-sized ones: absence, in this 721
particular case, seems to be in effect a length of zero. ECT 2, ECT 3 and ECT 5, however, are 722
unaffected; they continue to be scored as unknown for OTUs that lack ectopterygoids. 723
Restudy of the literature and a specimen has shown that states ECT 6(1) and ECT 7(1) can 724
only be scored for at most one OTU each (Appendix 1); this makes the characters ECT 6 and 725
ECT 7 parsimony-uninformative, so we have deleted them (see below). 726
To make our mergers more transparent, we have created abbreviations for merged 727
characters from those of all their constituents: e.g., PREMAX 1-2-3 (character 1) is built from 728
the three characters PREMAX 1, PREMAX 2 and PREMAX 3 of RC07, and MAX5/PAL5 729
(character 22) consists of MAX 5 and PAL 5 of RC07. The extreme case is VOM 5-10/PTE 730
10-12-18/INT VAC 1 (character 104), assembled from the six characters VOM 5, VOM 10, 731
PTE 10, PTE 12, PTE 18 and INT VAC 1 of RC07. 732
733
“Ontogeny discombobulates phylogeny” (Wiens, Bonett & Chippindale, 2005) 734
735
Heterochrony can result in misleading scores if only morphologically immature (juvenile or 736
paedomorphic) individuals of some taxa are known, which can result in large-scale character 737
correlation that can strongly distort phylogenetic trees (Wiens, Bonett & Chippindale, 2005). 738
We have tried to deal with this problem as described by Marjanović & Laurin (2008) and in 739
more detail by Marjanović & Laurin (2013a), modified from the recommendation of Wiens, 740
Bonett & Chippindale (2005: 96) and independently proposed by Tykoski (2005: 276): in 741
OTUs known only from apparently immature or paedomorphic individuals, observed states 742
are scored as unknown if they are restricted to immature stages in non-paedomorphic close 743
relatives of the OTU in question. The ontogeny of most of our OTUs and their closest 744
relatives is insufficiently well known; despite this, and in spite of the additional problems 745
17
17
discussed by Marjanović & Laurin (2013a), we think that this approach offers the greatest 746
chance to escape character correlation caused by heterochrony. 747
748
Deleted, recoded and split characters 749
750
Our matrix has only 276 characters, a strong decrease from the 339 of RC07. For the most 751
part, this is due to our mergers of redundant characters and does not entail a loss of 752
information (detailed in Appendix 1). The characters IFN 1, ILI 10, ISC 1, DOR FIN 1 and 753
BAS SCU 1, however, were parsimony-uninformative in the matrix of RC07 and remain so 754
after our corrections (all of them documented and justified in Appendix 1); we have deleted 755
them rather than carrying these currently useless characters along and inflating the character 756
count (see Marjanović & Laurin, 2013a, for discussion). The characters PREFRO 6, PREFRO 757
9, LAC 6, NOS 1, VOM 11, PAL 6, ECT 6, ECT 7, BASOCC 6 and DEN 1 were parsimony-758
informative as scored by RC07, but this is no longer the case after our corrections (likewise 759
documented and justified in Appendix 1), so we have deleted them as well. (For some of 760
them, like LAC 6 and VOM 11, one of the two states turns out not occur in the original taxon 761
sample at all, and only once or never in the expanded taxon sample.) Conversely, the 762
character ANG 3 was parsimony-uninformative in RC07, but is parsimony-informative in our 763
matrix even for the original taxon sample (see Appendix 1); we have kept it (though merged it 764
with ANG 2). 765
Characters INT FEN 1 and TEETH 3 were composites of two independent characters 766
each; we have split INT FEN 1 into the redefined INT FEN 1 and the new MED ROS 1, and 767
TEETH 3 into the redefined TEETH 3 and the new TEETH 10. We have also partitioned 768
several characters into more states than before. 769
Character PREMAX 7, a ratio of two measurements, was parsimony-uninformative as 770
defined (one of the two original states is limited to an OTU that was scored as unknown by 771
RC07), but not as described or scored; we have measured all OTUs, defined new state limits 772
(as discussed in Appendix 1), and rescored the entire taxon sample. SKU TAB 1, another 773
ratio, was defined and described in contradictory ways, neither of which matched the scores; 774
we have treated it the same way (see Appendix 1). The measurements, sources, calculations 775
and state changes are presented and compared to the original and revised scores in Data S4; 776
the ratios, sources and state changes are also shown in Appendix-Tables 1 (for PREMAX 7) 777
and 6 (for SKU TAB 1). 778
PTE 15 is not reproducible; we have not figured out, either from the description or 779
from the scores, what the difference between the two states was meant to be (see Appendix 1). 780
The most likely meaning would be a duplicate of PTE 14; we have deleted PTE 15. 781
Similarly, TAB 4 was described in a confusing way (see Appendix 1). Most likely, it 782
is either a duplicate of PAR 7 or is scored in a way that strongly contradicts its definition. (It 783
concerns the position of the suture between the tabularand the squamosal – most OTUs do 784
not have such a suture, but were scored as known nonetheless.) We have deleted it as well. 785
786
Reversibility 787
788
RC07 assumed, as is usual, that all transitions between the states of any character were freely 789
reversible. Several characters in their matrix concern the loss of skull bones or other features 790
that may be expected to be rarely (though perhaps not never: Wiens, 2011; Diaz & Trainor, 791
2015; Ossa-Fuentes, Mpodozis & Vargas, 2015) reversed. In the resulting trees, surprising 792
reversals occur; notably, the reappearance of the lost supratemporal is a synapomorphy of 793
Urocordylidae and Aïstopoda. In order to test the impact of such phenomena on the results, 794
we duplicated all analyses after recoding the binary characters 12 (PREFRO 1), 18 (LAC 1), 795
18
18
45 (POSFRO 1), 53 (TAB 1/SQU 4), 60 (POSORB 1), 71 (JUG 1), 77 (QUAJUG 1), 81 796
(PREOPE 1), 113 (PAL 8), 146 (PSYM 1), 156 (POSPL 1), 159 (ANG 1), 162 (SURANG 1), 797
166 (ANT COR 1), 170 (MID COR 1), 174 (POST COR 1) and 276 (CAU FIN 1) as 798
irreversible. For the multistate characters 32 (PAR 2/POSFRO 3/INTEMP 1/SUTEMP 1), 79 799
(POSPAR 1-2), 114 (ECT 1-4) and 275 (DIG 1-2-3-4), some transitions were coded as 800
irreversible for this additional set of analyses while others were not; this required new or 801
altered stepmatrices (Appendix-Tables 3, 4, 7, 10). These settings are contained in Data S2, a 802
NEXUS file otherwise identical to Data S1. For characters that may be considered cases of 803
doubt, the decision not to code them as irreversible is explained in Appendix 1. 804
Limb loss – transitions to state 5 of character DIG 1-2-3-4 – was coded as irreversible 805
in all analyses (Appendix-Tables 9, 10) in an attempt to avoid bias against the conservative 806
hypothesis that the adelospondyls and the aïstopods (the two limbless groups in this matrix) 807
are sister-groups as found e.g. by Vallin & Laurin (2004). 808
809
The tail fin skeleton (character CAU FIN 1) 810
811
The plesiomorphic osteichthyan tail fin skeleton, as retained e.g. by Acanthostega (Coates, 812
1996), contains supraneural and infrahemal radials (endochondral bones) and lepidotrichia 813
(dermal bones) around the vertebrae; the supraneural radials articulate with the neural spines, 814
the infrahemal radials with the hemal spines. The presence or absence of lepidotrichia in the 815
tail fin is coded as character 276 (CAU FIN 1) in this matrix. 816
There was no clear evidence for any component of the tail fin skeleton in any post-817
Devonian limbed vertebrate before Clack (2011a) described CM 34638, a part of an 818
anthracosaur (embolomere) tail that cannot be referred to any named genus with reasonable 819
certainty. (The tail is unknown for several anthracosaurs, including the one or two taxa from 820
the same site.) CM 34638 (meanwhile seen by D. M.) preserves supraneural radials and 821
lepidotrichia; the area where infrahemal radials would be is not preserved. 822
If added to our matrix, CM 34638 could influence the phylogenetic position of the 823
anthracosaurs: lepidotrichia (as well as endochondral radials) are clearly absent in all 824
temnospondyls with preserved tail tips, as well as in the possible anthracosaur Silvanerpeton 825
(Ruta & Clack, 2006), while tails are insufficiently well known to score this character in all of 826
the presumably more stemward Carboniferous taxa (baphetoids, colosteids, Crassigyrinus, 827
whatcheeriids) and even the Devonian Tulerpeton. Unfortunately, CM 34638 would be 828
indistinguishable from all other anthracosaurs in our matrix. It would differ from any of them 829
only in the distribution of missing data; at every mathematically possible position in or around 830
Anthracosauria (if not beyond), it would form a zero-length branch. Rather than greatly 831
increasing calculation times and memory needs by adding such a poorly known OTU, we 832
have therefore arbitrarily scored one of the existing anthracosaur OTUs as possessing 833
lepidotrichia. 834
For this purpose we have chosen Archeria because the more distal ones of its known 835
tail vertebrae bear hemal spines with rectangular ends (Holmes, 1989: fig. 26B) which might 836
have articulated with infrahemal spines, and because the more distal ones of its neural spines 837
are so extremely long and preserved in so many fragments (Holmes, 1989: fig. 24, 25) that we 838
wonder if some of the breaks actually represent flat joints between neural and supraneural 839
spines, or if the neural and the supraneural spines had fused. We stress, however, that this 840
decision is an “emergency solution”; CM 34638 clearly does not belong to Archeria (Clack, 841
2011a). 842
Squared-off ends of hemal and neural spines may be preserved in Proterogyrinus 843
(Holmes, 1984: fig. 22f). 844
845
19
19
Modifications to individual cells 846
847
We have changed scores that contradict the descriptive literature or our observations of 848
specimens. While we have not compared every single score to the literature or specimens, our 849
attention was not limited to particular taxa or characters; we ended up making changes to the 850
scoring of all taxa, and of all characters that remain in the matrix (as well as several that we 851
deleted as parsimony-uninformative, see above) except FONT 1, PSYM 3, PSYM 4, TEETH 852
6, TEETH 7 and TRU VER 29. All modifications (including those made by Damien Germain 853
[2008a: chapter V], except where we found them unjustified) are documented and justified in 854
Appendix 1, with citations of the literature and/or specimens we used to rescore each cell. For 855
the sake of brevity, the original scores (RC07) are usually not mentioned there, only our 856
modified ones are; scores we did not change are not mentioned in Appendix 1 except where 857
they could be controversial. 858
“This has not been a literature-based exercise”, wrote Ruta, Coates & Quicke (2003: 859
292) in their conclusions; and while they and RC07 did have to rely on the literature to score 860
some taxa (if only because the taxon sample is simply too large to see all relevant specimens 861
within realistic time and budget constraints), both of their publications contain lists of 862
specimens they had studied firsthand. Strangely, however, the results of almost none of these 863
observations are mentioned in the papers; there is no way of extracting information on which 864
characters are described or illustrated incorrectly in the literature or which ones are visible in 865
the fossils but not presented in a publication (so that scoring from the literature would 866
produce spurious missing data). We must insist that a data matrix is not an appropriate place 867
to publish new observations: in a matrix there is no way of telling if a surprising score is a 868
new observation or just a typographic error, a momentary confusion of contradictory 869
conventions about which state is called 0 and which is called 1, or one of a number of other 870
very common problems (see above and Marjanović & Laurin, 2013a, for discussion). For 871
each of the scores we have changed based on our observations of specimens, we have 872
therefore documented the change and referred to the numbers of specimens that show the 873
states we have scored. Only the specimens that have led us to changing a score (or several) 874
are listed in Table 1; specimens that agree with the scores of RC07 are therefore only listed 875
when other specimens of the same OTU show another state and we have consequently scored 876
polymorphism. 877
In the end we have probably compared almost all cells in the matrix to the literature 878
and/or to specimens. Shying away from the inordinate amount of time this work ended up 879
taking, we initially concentrated our efforts on characters and OTUs relevant to lissamphibian 880
origins, as determined,mostly, by optimizations on trees in Mesquite (Maddison & Maddison, 881
2015). Soon, though, we branched out to characters with subjectively suspicious state 882
distributions (it goes without saying that many turned out to be entirely or partially correct), 883
characters we did not immediately understand, characters we recoded while merging them 884
with others (see above), and taxa that had been redescribed (recently or sometimes not so 885
recently). Furthermore, we investigated characters that supported conflicting or unusual 886
hypotheses of relationships apart from lissamphibian origins, and the taxa implicated in these; 887
examples are Seymouriamorpha and Diadectomorpha, both surprisingly found to be 888
paraphyletic by RC07, and Adelogyrinidae and Acherontiscus, found for the first time to 889
clade with colosteids rather than “lepospondyls” by RC07. Finally, we checked the scores of 890
OTUs that temporarily did strange things in preliminary analyses of ours, like Tulerpeton, 891
Edops, Trimerorhachis or Ossinodus (unsurprisingly, some of this strangeness has remained). 892
Whenever we investigated a character, we usually verified its scores for all OTUs, and 893
whenever we investigated an OTU, we verified its scores for all characters except as restricted 894
below. 895
20
20
Table 1: List of specimens used by D. M. to change scores in the matrix (as cited in 896
Appendix 1) or to score added taxa. The OTUs are in the same order as in the matrix, which 897
is unchanged from the non-alphabetical version of the matrix file of RC07 for the shared taxa 898
(Acanthostega through Tseajaia). Specimens that do not contradict any scores of RC07 or the 899
literature on the added taxa (Iberospondylus through “St. Louis tetrapod”) are not listed 900
except in cases of polymorphism; a few such specimens are mentioned in the text. *, type 901
specimen of the type species; **, type and only known specimen of the only known species. 902
903
OTU Specimens
Acanthostega TMM 41766-1 (cast of MGUH VP 6033*, formerly “A. 33”)
Ichthyostega AMNH 23100; MCZ 3361; TMM 41224-2 (all of them casts of MGUH
VP 6055, formerly “A. 55”)
Greererpeton TMM 41574-1
Edops MCZ 1378, 1769, 1781, 1782, 6489, 6493, 7126, 7128, 7136, 7143,
7158, 7162, 7197, 7258, 7259, 7264, 7274; USNM 23309
Chenoprosopus CM 34909; USNM 437646
Isodectes CM 81430, 81512; MCZ 6044 (cast of USNM 4481); unnumbered MCZ
cast of AMNH 6935 before etching; USNM 4471, 4474, 4555
Trimerorhachis AMNH 4565*, 4572; TMM 40031-80, 40031-81, 40998-39
Eryops AMNH 4180, 4183, 4186, 4189*, 4673, 23529; MCZ 1126, 1129, 1536,
2588, 2638, 2682, 2766; TMM 31225-33, 31226-12, 31227-11, 31227-
14, 40349-20; USNM uncatalogued “Texas ’84 #40”, “Texas ’86 #77”
Phonerpeton AMNH 7150; MCZ 1414, 1419*, 1485, 1548, 1771, 2313, 2474; USNM
437796
Ecolsonia CM 38017, 38024
Broiliellus brevis MCZ 3272
Doleserpeton AMNH 24969, 29466, 29470; BEG 40882-25
Platyrhinops AMNH 2002
Eocaecilia MNA V8066* (Museum of Northern Arizona; formerly MCZ 9010)
Karaurus unnumbered MNHN cast1 of PIN 2585/2**
Triadobatrachus MNHN MAE 126** (natural mold and cast)1
Caerorhachis Casineria: NMS G 1993.54.1** (part and counterpart; Fig. 3–5)
Archeria MCZ 2049, other MCZ specimens2
Gephyrostegus MB.Am.641; TMM 41733-1 (cast)
Seymouria BEG 30966-176
Diadectes AMNH 4352, 4839; BEG 31222-56
Paleothyris TMM 45955-2 (cast of MCZ 3482)
Batropetes B. palatinus: MB.Am.1232 (including casts)
Tuditanus CM 29592
Stegotretus CM 34901 (all others were on loan)
Microbrachis MB.Am.840
Hyloplesion NHMW 1983/82/54, other NHMW specimens
Odonterpeton USNM 4465+4467**3
Scincosaurus MB.Am.29
Diceratosaurus AMNH 6933*, CM 25468, 26231, 29593, 29876, 34617, 34656, 34668,
34696, 34670, 67157, 67169, 72608, 81504, 81507, 81508; MB.Am.776,
MB.Am.778
Ptyonius MCZ 3721 (cast of “AMNH 6871 (85466)”)
Sauropleura CM 25312
21
21
Lethiscus MCZ 2185
Phlegethontia USNM 17097
Tseajaia CM 38033
Iberospondylus PU-ANF 2, 14*, 154
Caseasauria Eothyris: MCZ 1161**; Oedaleops: unnumbered MCZ cast of UCMP
35758*
Sparodus NHMW 1899-0003-0006 (Fig. 6, 7)
Carrolla TMM 40031-54**
Sclerocephalus MB.Am.1346
Chelotriton MB.Am.45 (natural mold and cast)
Trihecaton CM 47681*, 47682 (probably part of the same individual)
St. Louis tetrapod MB.Am.1441** (natural mold and cast)
904
1 Observed together with M. L. 905
2 All were used to score the same characters. 906
3 The apparently unpublished specimens CM 81525 and CM 81526 are also labeled 907
Odonterpeton, and are kept together with a note by Donald Baird (dated 1991) which says 908
that further small skeletons attributed to Brachydectes may belong to Odonterpeton as well. 909
More likely, CM 81525 – a limbless skeleton without a trace of limbs or girdles, whose skull 910
is only preserved as an indistinct impression – is instead a juvenile aïstopod with ribs that are 911
not yet k-shaped, and the disarticulated CM 81526 is a juvenile Brachydectes after all (D. M., 912
pers. obs.; J. Pardo, pers. comm. after seeing photos taken by D. M.); a more detailed study is 913
in preparation. 914
4 All three were observed together with Rodrigo Soler-Gijón; additionally, they had been 915
studied earlier by M. L. (Laurin & Soler-Gijón, 2001, 2006). 916
917
918
For the following OTUs – and quite possibly others – we have compared most or all 919
cells to the literature (see below for specimens): 920
Panderichthys (Boisvert, 2005, 2009; Brazeau & Ahlberg, 2006; Boisvert, Mark-921
Kurik & Ahlberg, 2008; Ahlberg, 2011 – occiput, extremities and girdles) 922
Ventastega (Ahlberg, Lukševičs & Lebedev, 1994; Ahlberg et al., 2008; Lukševičs, 923
Ahlberg & Clack, 2003) 924
Acanthostega (Coates, 1996; Porro, Rayfield & Clack, 2015) 925
Ichthyostega (skull roof in dorsal view: Clack, 2007; lower jaw: Ahlberg & Clack, 926
1998; Clack et al., 2012a; humerus: Jarvik, 1996; Callier, Clack & Ahlberg, 2009; 927
Ahlberg, 2011; Pierce, Clack & Hutchinson, 2012; Pierce et al., 2013) 928
Tulerpeton (Lebedev & Coates, 1995) 929
Crassigyrinus (Panchen, 1985; Panchen & Smithson, 1990; Clack, 1998) 930
Whatcheeria (Lombard & Bolt, 2006 – lower jaw and teeth only) 931
Baphetes (Milner & Lindsay, 1998; Milner, Milner & Walsh, 2009 – only previously 932
missing data and explicitly stated contradictions to older literature) 933
Eucritta (Clack, 2001) 934
Edops (Romer & Witter, 1942) 935
Chenoprosopus (Hook, 1993; Reisz, Berman & Henrici, 2005 – previously missing 936
data, including all postcranial material, and data mentioned as having been hitherto 937
misinterpreted) 938
Cochleosaurus (skull: Sequeira 2004; postcranium: Sequeira 2009) 939
Neldasaurus (Chase, 1965) 940
22
22
Trimerorhachis (postcranium: Pawley, 2007; skull: Milner & Schoch, 2013) 941
Dendrerpetidae (Carroll, 1967; A. R. Milner, 1980, 1996; Godfrey, Fiorillo & Carroll, 942
1987; Holmes, Carroll & Reisz, 1998; Robinson, Ahlberg & Koentges, 2005 – 943
especially to make sure no polymorphisms were overlooked; see below and Schoch & 944
Milner 2014) 945
Eryops (appendicular skeleton: Pawley & Warren, 2006) 946
Acheloma (Maddin, Reisz & Anderson, 2010; Polley & Reisz, 2011 – mostly filled in 947
previously missing data) 948
Broiliellus (Carroll, 1964; Schoch, 2012) 949
Platyrhinops (Clack & Milner, 2010; remaining missing data filled in from Hook & 950
Baird, 1984, and from Werneburg, 2012b, where not likely ontogeny-dependent) 951
Doleserpeton (Sigurdsen, 2008; Sigurdsen & Bolt, 2009, 2010; Sigurdsen & Green, 952
2011: appendix 2) 953
Micromelerpeton (Boy, 1972, 1995; Schoch, 2009b) 954
Apateon, Leptorophus, Schoenfelderpeton (Boy, 1972, 1986, 1987; Werneburg, 1991; 955
Schoch & Fröbisch, 2006; Schoch & Milner, 2008; Fröbisch & Schoch, 2009b) 956
Albanerpetidae (Estes & Hoffstetter, 1976; Fox & Naylor, 1982; Gardner, 2001; 957
McGowan, 2002; Gardner, Evans& Sigogneau-Russell, 2003; Venczel & Gardner, 958
2005; Maddin et al., 2013a; note that we follow this latter reference in interpreting the 959
“cultriform process of the parasphenoid” [McGowan, 2002: fig. 13] as a hyobranchial 960
element, contra Marjanović & Laurin, 2008) 961
Eocaecilia (Jenkins, Walsh & Carroll, 2007) 962
Triadobatrachus (Rage & Roček, 1989; Roček & Rage, 2000; Sigurdsen, Green & 963
Bishop, 2012; Ascarrunz et al., 2016) 964
Valdotriton (Evans & Milner, 1996) 965
Karaurus (Ivachnenko, 1978) 966
Caerorhachis (Ruta, Milner & Coates, 2002) 967
Bruktererpeton (Boy & Bandel, 1973) 968
Gephyrostegus (skull: Klembara et al., 2014; presacral axial skeleton: Godfrey & 969
Reisz, 1991; rest: Carroll, 1970) 970
Eoherpeton (Panchen, 1975; Smithson, 1985) 971
Proterogyrinus (Holmes, 1980, 1984) 972
Archeria (Romer, 1957; Holmes, 1989) 973
Solenodonsaurus (Laurin & Reisz, 1999; Danto, Witzmann & Müller, 2012 – mostly 974
filled in previously missing data) 975
Kotlassia (dermatocranium: Bulanov, 2003; rest: Bystrow, 1944) 976
Diadectes (Case, 1910, 1911; Case & Williston, 1912; Olson, 1947; Moss, 1972; 977
Berman, Sumida & Lombard, 1992; Berman, Sumida & Martens, 1998; Berman et al., 978
2004) 979
Limnoscelis (Fracasso, 1983; Berman & Sumida, 1990; Reisz, 2007; Berman, Reisz & 980
Scott, 2010; Kennedy, 2010) 981
Westlothiana (Smithson et al., 1994) 982
Batropetes (Glienke, 2013, 2015) 983
Rhynchonkos (CG78; Szostakiwskyj, Pardo & Anderson, 2015) 984
Microbrachis (CG78; Vallin & Laurin, 2004; Olori, 2015) 985
Hyloplesion (CG78; Olori, 2015) 986
Brachydectes (Wellstead, 1991; Pardo & Anderson, 2016) 987
Acherontiscus (Carroll, 1969b) 988
23
23
Adelospondylus, Adelogyrinus, Dolichopareias (Andrews & Carroll, 1991) 989
Scincosaurus (Milner & Ruta, 2009) 990
Diplocaulus (postcranium and lower jaw: Williston, 1909; Douthitt, 1917) 991
Diploceraspis (mostly postcranium and lower jaw: Beerbower, 1963) 992
Lethiscus (Wellstead, 1982; Anderson, Carroll & Rowe, 2003) 993
Oestocephalus (Carroll, 1998; Anderson, Carroll & Rowe, 2003; Anderson, 2003a) 994
Phlegethontia (Anderson, 2002, 2007a) 995
Notobatrachus (Estes & Reig, 1973; Báez & Basso, 1996; Báez & Nicoli, 2004, 2008) 996
Vieraella (Estes & Reig, 1973; Báez & Basso, 1996) 997
Ossinodus (Warren & Turner, 2004; Warren, 2007; Bishop, 2014) 998
Pederpes (Clack & Finney, 2005) 999
Orobates (Berman et al., 2004; Nyakatura et al., 2015: digital model) 1000
Ariekanerpeton (Laurin, 1996a; Klembara & Ruta, 2005a, b) 1001
Leptoropha (Bulanov, 2003) 1002
Microphon (Bulanov, 2003, 2014) 1003
Tseajaia (Moss, 1972; Berman, Sumida & Lombard, 1992) 1004
Utegenia (Laurin, 1996b; Klembara & Ruta, 2004a, b) 1005
Parts of the above list have puzzling implications. For instance, Ruta, Coates & Quicke 1006
(2003) cited Ahlberg, Lukševičs & Lebedev (1994) as their source for scoring Ventastega, but 1007
only used that publication to code not much more than half of the bones that are described and 1008
illustrated in that work; large parts of the skull, e.g. almost the entire palate, were scored as 1009
entirely unknown by Ruta, Coates & Quicke (2003) and RC07 for reasons that we could not 1010
determine. This goes so far that in some instances some of the bones belonging to a skull 1011
fragment were scored while others belonging to the same fragment were not. 1012
RC07 cited the paper by Sequeira (2004), but only for its phylogenetic analysis, not 1013
for its redescription of the skull of Cochleosaurus based on “several large, presumably adult, 1014
skulls” (Sequeira, 2004: abstract) – all previous descriptions “were based almost entirely on 1015
small, subadult […] specimens” (ibid.). RC07 thus perpetuated the outdated scores of 1016
Cochleosaurus by Ruta, Coates & Quicke (2003). 1017
Ruta, Coates & Quicke (2003) used Boy & Bandel (1973) as their source for the 1018
scoring of Bruktererpeton. That publication is written in German, which may explain some of 1019
the differences between it and the scoring by Ruta, Coates & Quicke (2003) and RC07. 1020
(Fortunately, German is D. M.’s native language.) Some of the figures, however, contradict 1021
the matrix by RC07 as well. 1022
Despite the overlap in authors, many discrepancies exist between the matrix by RC07 1023
and the detailed redescriptions of Caerorhachis (Ruta, Milner & Coates, 2002 – not cited by 1024
RC07, but cited as “2001”, the year of publication originally intended by the journal, by Ruta, 1025
Coates & Quicke, 2003), Silvanerpeton (Ruta & Clack, 2006 – not cited, not even as “in 1026
press”), Ariekanerpeton (Klembara & Ruta, 2005a, b – not cited, although personal 1027
observations are cited) and Utegenia (Klembara & Ruta, 2004a, b – cited). 1028
Ruta, Coates & Quicke (2003) scored Kotlassia after the description by Bystrow 1029
(1944); RC07 did not change this. Bystrow explicitly synonymized Kotlassia and 1030
Karpinskiosaurus and did not document for all parts of his description on which specimen(s) 1031
they were based. Bulanov (2003) disagreed with Bystrow and separated the two taxa again, 1032
considering them rather distant relatives, but he merely mentioned the existence of postcrania 1033
of Kotlassia (the entire monograph describes only the cranial anatomy of a wide range of 1034
seymouriamorphs). Similarly, Klembara (2011) redescribed only the skull of 1035
Karpinskiosaurus. A useful description of the postcrania of either taxon does not exist as far 1036
as we know. We have accepted the scores by RC07 at face value for the time being and scored 1037
24
24
all postcranial characters of Karpinskiosaurus, a taxon we have added to the matrix (see 1038
below), as unknown, except for those few that Bystrow (1944) explicitly commented on. 1039
For the following taxa we have compared most or all cells to specimens (fossils or 1040
casts) which are, where necessary, mentioned under the respective characters in Appendix 1: 1041
Ichthyostega (palate only) 1042
Edops (as Romer & Witter [1942] noted, the sutures on the only known adult skull 1043
roof – that of MCZ 1378 – are for the most part extremely difficult to trace, so we 1044
have accepted their interpretations wherever D. M. was unable to confirm them; D. M. 1045
did not detect any contradictions) 1046
Isodectes (postcranium only) 1047
Ecolsonia (previously missing data only) 1048
Triadobatrachus 1049
Karaurus (everything except the palate, which is not reproduced by the cast we had 1050
access to) 1051
Casineria (part of the Caerorhachis OTU, see below; from photographs taken by D. 1052
M.) 1053
Diceratosaurus is about to undergo redescription by Angela Milner, D. M. and Florian 1054
Witzmann; we have not preempted that work, but only changed the scores of characters of 1055
particular interest based on D. M.’s observations of specimens, mostly filling in missing data 1056
– and often confirming the redescription by Jaekel (1903). The complete list of specimens we 1057
have used to change scores in the matrix forms Table 1. We cannot provide a list of all 1058
specimens we have seen; however, whenever D. M. visited a collection, he saw all or almost 1059
all specimens that are broadly relevant to this study. The exception to this is the NMS, where 1060
the collection is kept in a building on the other side of the city and specimens that are not on 1061
exhibit are brought to the museum building only on request; the only NMS specimen D. M. 1062
has seen is Casineria. 1063
1064
The albanerpetid neck 1065
1066
Though this concerns very few characters, our coding of the albanerpetids (a single OTU) 1067
assumes our reinterpretation of their unique atlas-axis complex. As in mammals, this complex 1068
accommodates dorsoventral and lateral movements of the head at separate joints. 1069
Traditionally (e.g. Estes & Hoffstetter,1976; Fox & Naylor, 1982), this complex is 1070
considered to consist of the atlas (a complete vertebra consisting of a centrum and fused, fully 1071
formed, full-size neural arch), the axis (a centrum that lacks any trace of a neural arch), and 1072
the third vertebra (again a complete vertebra consisting of a centrum and fused, fully formed, 1073
full-size neural arch). The “axis” is commonly sutured to the “third vertebra”. Dorsoventral 1074
movements of the head occurred between the skull and the atlas, lateral ones between the atlas 1075
and the “axis”. By comparison to amniotes and “microsaurs”, we think it is more 1076
parsimonious to interpret the “axis” as only the intercentrum of the axis, even though 1077
intercentra are otherwise unknown in albanerpetids, while the “third vertebra” would be the 1078
pleurocentrum and neural arch of the axis. 1079
1080
Ignored information 1081
1082
In a conference abstract, Anderson et al. (2013) reported having μCT-scanned the skulls of 1083
several aïstopods, including Lethiscus which is present in this matrix, and found several 1084
surprises that will have an impact on the phylogenetic position of Aïstopoda: “When analyzed 1085
using parsimony in a matrix of lower jaws of early tetrapods, Lethiscus and Coloraderpeton 1086
fall stemward, near Crassigyrinus.” Of course, that abstract only mentioned a few isolated 1087
25
25
features, and Anderson et al. have not investigated the other aïstopods in this matrix 1088
(Oestocephalus and Phlegethontia), opting instead for the much less crushed Coloraderpeton. 1089
Updating the scores of Lethiscus would thus have exposed us to the risk of creating a self-1090
contradictory chimera – if not of Lethiscus by itself, then likely of Aïstopoda as a whole. 1091
Thus, we prefer to wait for the full publication of these exciting results; in the meantime, we 1092
note that these may affect aïstopod intra- and interrelationships. 1093
The abstract by Porro, Clack & Rayfield (2015) promises previously unknown 1094
character states of Crassigyrinus, but the only new information it makes explicit is that “all 1095
three coronoids bear teeth”; we are unable to use this information, because our matrix 1096
contains separate characters for the presence or absence of tusks, denticles and toothrows. 1097
We updated our scores of Brachydectes following Pardo & Anderson (2016) before 1098
we ran the phylogenetic analyses, but forgot to repeat the measurements of the skull required 1099
to score characters 3 (PREMAX 7) and 94 (SKU TAB 1) in time. For SKU TAB 1, our score 1100
(2) is correct. For PREMAX 7, it should be 0 rather than 1; judging, in hindsight, from the 1101
scores of the closest relatives of Brachydectes, the correct state 0 would most likely just add 1102
one step to each tree, regardless of analysis settings and constraints. 1103
1104
Treatment of OTUs 1105
1106
Paleothyris/Protorothyris 1107
1108
One OTU of RC07 is (explicitly) a composite of Paleothyris acadiana and Protorothyris 1109
archeri. Protorothyris is, according to the only phylogenetic analysis that has included it so 1110
far (Müller & Reisz, 2006), more closely related to Petrolacosaurus – another OTU in this 1111
matrix – than to Paleothyris; Paleothyris could even be more closely related to Captorhinus – 1112
also included in this matrix – than to either Protorothyris or Petrolacosaurus. 1113
Still, these two species are morphologically similar, so we have not made a special 1114
effort to weed out scores based on Protorothyris. Our changes to this OTU, however, are 1115
exclusively based on the description of Paleothyris by Carroll (1969a) and D. M.’s 1116
observations of Paleothyris specimens, especially TMM 45955-2 (a cast of MCZ 3482, which 1117
is on display and was not accessible at the time), cursorily confirmed against MCZ 3473, 1118
MCZ 3475, MCZ 3477, MCZ 3481 (the holotype), MCZ 3483, MCZ 3486, MCZ 3487, MCZ 1119
3488, MCZ 3490, MCZ 3492 and the apparently uncatalogued unlabeled specimen B-17 (also 1120
kept at the MCZ). Some of these changes are from known to unknown, so they likely concern 1121
characters whose state is known in the more completely preserved Protorothyris but not in 1122
Paleothyris. 1123
1124
Dendrerpetidae 1125
1126
Milner (1996) revised the specimens from Joggins (Nova Scotia, Canada) that were at that 1127
time included in Dendrerpeton acadianum (following Carroll, 1967), but had originally been 1128
described as several genera with many species. “The majority of specimens are still 1129
attributable to Dendrerpeton acadianum but three other forms are represented by one or two 1130
specimens each. These are Dendrerpeton confusum sp. nov., based on a single large skull, 1131
Dendrerpeton helogenes (Steen) comb. nov., based on two specimens, and an indeterminate 1132
cochleosaurid, based on a set of cranial fragments” (Milner, 1996: abstract). One of the 1133
specimens newly referred to D. helogenes was the well preserved skull that had been 1134
described by Godfrey, Fiorillo & Carroll (1987) as D. acadianum. “Subsequently HOLMES et 1135
al. [ = Holmes, Carroll & Reisz] (1998) described an exquisitely preserved skeleton of a 1136
temnospondyl from Joggins and referred it to Dendrerpeton acadianum. By doing so, they 1137
26
26
lumped all the Joggins material together again. In this work, it is argued that the Joggins 1138
material includes two distinct genera, Dendrerpeton sensu MILNER 1996 and Dendrysekos 1139
(Dendrerpeton helogenes of MILNER 1996), which differ in a number of characters. The 1140
specimens of GODFREY et al. [ = Godfrey, Fiorillo & Carroll] (1987) and HOLMES et al. 1141
(1998) are referable to Dendrysekos. This effectively means that Dendrysekos rather than 1142
Dendrerpeton has formed the outgroup in many recent cladistic analyses of temnospondyls.” 1143
(Schoch & Milner, 2014: 25) 1144
Ruta, Coates & Quicke (2003) did not cite Milner (1996). Their “Dendrerpeton 1145
acadianum” OTU, kept unchanged by RC07 (who did not cite Milner [1996] either), is in part 1146
based on their personal observations of specimens, of which at least some – notably the 1147
lectotype – really do belong to D. acadianum according to Milner (1996) and Schoch & 1148
Milner (2014). However, Ruta, Coates & Quicke (2003) also used and cited the descriptions 1149
by Godfrey, Fiorillo & Carroll (1987) and Holmes, Carroll & Reisz (1998). Assuming that the 1150
personal observations were in fact used to score the taxon and not merely to cursorily confirm 1151
the literature, this OTU is thus a chimera of Dendrerpeton acadianum and Dendrysekos 1152
helogenes. 1153
Schoch & Milner (2014) continued to maintain that Dendrerpeton and Dendrysekos 1154
are sister-groups; for this reason, we have not split the OTU for the time being, but merely 1155
renamed it Dendrerpetidae, the name recommended by Schoch & Milner (2014), and 1156
compared it to the literature to make sure no polymorphisms had been overlooked (see 1157
above). In future work, however, it may be a good idea to treat all dendrerpetid species 1158
separately (including Dendrerpeton rugosum from Jarrow in Ireland: A. R. Milner, 1980; 1159
Schoch & Milner, 2014): Ruta (2009) added Dendrerpeton confusum to a temnospondyl 1160
matrix derived from the matrix of RC07 and found that it was not the sister-group of the “D. 1161
acadianum” OTU. (Whether the latter included any other dendrerpetids is unclear to us. The 1162
matrix of Ruta [2009] is a slight modification of that of Ruta & Bolt [2006]. The latter’s 1163
appendix 1 lists all OTUs, except Dendrerpeton confusum, and the literature sources used to 1164
score them. For “Dendrerpeton acadianum”, these are Carroll, 1967; A. R. Milner, 1980, 1165
1996; Godfrey, Fiorillo & Carroll, 1987; and Holmes, Carroll & Reisz, 1998. In addition, this 1166
OTU is marked with an asterisk for having “been examineddirectly by one or both authors 1167
(using either casts or original specimens)”; which specimens those are is not mentioned, but 1168
presumably they are the same as those consulted for Ruta, Coates & Quicke [2003]. However, 1169
given the absence of D. confusum from this appendix, we wonder if the appendix was written 1170
before Ruta & Bolt decided to separate the two Dendrerpeton OTUs; if so, it is possible that 1171
the “D. acadianum” OTU really is just D. acadianum and not a chimera.) 1172
Balanerpeton has sometimes (Pawley, 2006; Clack et al., 2012b; Dilkes, 2015a), but 1173
not always (Milner & Sequeira, 1994; Ruta, 2009), been found as the sister-group of 1174
“Dendrerpeton”; we therefore follow Schoch & Milner (2014) in not including it in the 1175
Dendrerpetidae OTU. 1176
1177
Rhynchonkos 1178
1179
Szostakiwskyj, Pardo & Anderson (2015) have shown that previous conceptions of 1180
Rhynchonkos (e.g. CG78; considered to contain a single species) were chimeric. They 1181
restricted Rhynchonkos to its holotype, a skull with lower jaw. The other skulls were referred 1182
to the new taxa Aletrimyti and Dvellecanus; a further lower jaw turned out to belong to none 1183
of the three. Szostakiwskyj, Pardo & Anderson (2015) did not mention the postcranial 1184
material that had been referred to Rhynchonkos (e.g. CG78). A poorly preserved but largely 1185
articulated presacral skeleton (CG78: fig. 67) belongs to the same specimen as the referred 1186
skull of Aletrimyti (Szostakiwskyj, Pardo & Anderson, 2015: 5). We therefore wondered 1187
27
27
whether to restrict our Rhynchonkos OTU to Rhynchonkos or instead to Aletrimyti (the type 1188
material of which is also slightly better preserved than that of Rhynchonkos). J. Pardo (pers. 1189
comm. 2015) has, however, pointed out that the postcranial material needs to be revisited. We 1190
have therefore restricted the Rhynchonkos OTU to the type material of Rhynchonkos and 1191
scored all postcranial characters as unknown. 1192
RC07 had scored a few characters that remain unknown for all of the specimens 1193
referred to Rhynchonkos by CG78. An example is CAU FIN 1, for which they scored state 1, 1194
the absence of lepidotrichia in the tail – only the most proximal tail vertebrae are represented 1195
in the previously attributed material (CG78: 103–104, 109–110, fig. 67–72)! This is, 1196
incidentally, a common issue that affects several other OTUs (Appendix 1). 1197
1198
The skull roof of Brachydectes 1199
1200
Wellstead (1991) reconstructed several unusual features in the skull roof of Brachydectes. We 1201
have previously tried (Marjanović & Laurin, 2008, 2013a) to resolve these problems by 1202
reinterpreting the homologies of the bones that Wellstead (1991) identified. For example, the 1203
supposed postparietals do not meet in Wellstead’s reconstruction, but are separated by a 1204
contact between the suproccipital and the parietals; we have suggested that they are tabulars 1205
instead, while the supposed tabulars are postorbitals (thought absent by Wellstead), and 1206
postparietals are absent. As a byproduct, this would restore normal spatial relationships to the 1207
posttemporal foramen identified by Wellstead (1991). 1208
Based on new material and μCT scans, Pardo & Anderson (2016) have now shown 1209
that Wellstead (1991) overlooked a few sutures and misjudged certain 3D relationships; our 1210
reinterpretation of Wellstead’s reconstruction is therefore moot. In particular, Wellstead 1211
correctly identified the postparietals, which do, however, meet in the midline in most or all 1212
specimens for a variable part of their length, as D. M. can now confirm from personal 1213
observation of MCZ 3329 (a cast of AMNH 6925) and MCZ 3330 (a cast of AMNH 6861); 1214
there is no posttemporal foramen at all; and postorbitals are, like in lissamphibians, absent. 1215
We have updated the scores of Brachydectes following Pardo & Anderson (2016). – Note also 1216
that the reconstruction of Batropetes by Carroll (1991), which Marjanović & Laurin (2008, 1217
2013a) used for comparison to Brachydectes, differs in several respects from the ones by 1218
Glienke (2013, 2015), which we have used here to score Batropetes. 1219
1220
Taxa added as parts of existing OTUs 1221
1222
We have interpreted almost all OTUs as genera (or larger taxa in the cases of Albanerpetidae 1223
and Dendrerpetidae) rather than species. This has allowed us to fill in some missing data and 1224
to approach more plesiomorphic morphotypes. We do not think the polymorphisms this has 1225
occasionally introduced are a problem; quite the opposite – they are a better representation of 1226
the scope of the matrix. With the exceptions of Megalocephalus (Milner, Milner & Walsh, 1227
2009), Broiliellus (Schoch, 2012; Holmes, Berman & Anderson, 2013; Dilkes, 2015b) and of 1228
course Pholiderpeton, the monophyly of the few genera in this matrix that are not 1229
monospecific is fairly obvious, at least with respect to the other OTUs in this matrix, and has 1230
not been disputed in the literature. Pholiderpeton was already coded as two separate OTUs 1231
(Ph. scutigerum, Ph. attheyi) by RC07 and indeed Ruta, Coates & Quicke (2003). 1232
Megalocephalus is, if at all, only paraphyletic with respect to Kyrinion (Milner, Milner & 1233
Walsh, 2009) which is not included in this matrix (we have refrained from adding it because it 1234
would, given the present character sample, provide almost no new information on baphetid 1235
morphodiversity and thus baphetid relationships). We have not used information from any 1236
species referred to Broiliellus other than the one used by RC07, B. brevis – which is, 1237
28
28
incidentally, not the type species of Broiliellus and might therefore end up outside that taxon. 1238
In the figures and Appendix 2, we explicitly call this OTU “Broiliellus brevis”. Similarly, we 1239
have ignored “Euryodus” bonneri (Proxilodon: Huttenlocker et al., 2013). 1240
1241
1242
1243
Figure 3: Caudal trunk vertebrae of NMS G 1993.54.1cp, the holotype of Casineria kiddi 1244
(counterpart plate), in partly right lateral, partly interior view. The neural arches end 1245
dorsally in an ossification front marked by black material (pyrite or another iron sulfide?) just 1246
dorsal to the level of the postzygapophyses, indicating that the individual was immature; the 1247
pleuro- and intercentra are reminiscent of those of the temnospondyl Neldasaurus (Chase, 1248
1965), although their precise shapes are difficult to determine because the vertebrae are split 1249
lengthwise and probably recrystallized. Abbreviations: ic, intercentrum; nsp, neural spine; pc, 1250
pleurocentrum; poz, postzygapophysis; prz, prezygapophysis. The scale bar is approximate. 1251
Photo taken by D. M. 1252
1253
1254
29
29
Milner & Schoch (2013) found that all recognized species of Trimerorhachis except T. 1255
sandovalensis form a clade which may or may not be the sister-group of T. sandovalensis to 1256
the exclusion of Neldasaurus. Because Neldasaurus is an OTU in this matrix, we have not 1257
used information from ?T. sandovalensis to score the Trimerorhachis OTU. We considered 1258
adding ?T. sandovalensis as an OTU to our analysis with added taxa, but it would hardly 1259
differ from the Trimerorhachis OTU at all. – Because Milner & Schoch (2013) presented a 1260
phylogeny of the four species in the undoubted Trimerorhachis clade, we have avoided 1261
scoring Trimerorhachis as polymorphic and instead tried to identify the plesiomorphic 1262
condition (as we have done with Albanerpetidae and Batropetes); this concerns very few 1263
characters, however. 1264
Pawley (2006: 207) added the headless skeleton known as Casineria kiddi to the 1265
matrix by Ruta, Coates & Quicke (2003) as well as to expanded versions with larger character 1266
samples, and found that it scoredidentically – except for the distribution of missing data – to 1267
Caerorhachis bairdi which comes from a similar but not identical age and locality. Pawley 1268
(2006: 195, 239) went so far as to call them “indistinguishable based on the available 1269
evidence”. This is a considerable surprise, because Casineria has been suggested to be a close 1270
relative or even member of Amniota (Paton, Smithson & Clack, 1999; Clack et al., 2012b), 1271
while Caerorhachis is variously considered to be close to the origin of temnospondyls and/or 1272
anthracosaurs (Ruta, Milner & Coates, 2002, and references therein; RC07; arguably Clack et 1273
al., 2012b). Yet, in our matrix, Casineria and Caerorhachis again differ only in their 1274
distribution of missing data; indeed, based on the descriptive literature, we confirm Pawley’s 1275
(2006: 239; see also 195) remark that, “[a]s in Caerorhachis bairdi, none of the postcranial 1276
characteristics claimed to be ‘reptiliomorph’ in Casineria kiddi are truly apomorphic for the 1277
amniote lineage. All are present in temnospondyls […], or potentially may be present in basal 1278
temnospondyls (including the five[-]digit manus), because they are plesiomorphic for early 1279
tetrapods.” For the time being, we refrain from synonymizing the two taxa formally, because 1280
we have not seen the only known certain specimen of Caerorhachis. D. M. has, however, 1281
seen the Casineria specimen (part and counterpart: NMS G 1993.54.1p, NMS G 1993.54.1cp) 1282
and taken a large number of photographs through a microscope with a mounted camera, of 1283
which we here present Figures 3–5; we confirm the absence of differences to the thorough 1284
redescription of Caerorhachis by Ruta, Milner & Coates (2002) and have used these 1285
observations, in addition to the description by Paton, Smithson & Clack (1999), to score 1286
Casineria as part of the Caerorhachis OTU (filling in missing data, documented for each 1287
character in Appendix 1 and highlighted in Appendix 2). 1288
Pawley (2006: 239) also reported to have found the “microsaurs” Tuditanus and 1289
Asaphestera (both included by Ruta, Coates & Quicke, 2003, and RC07) to likewise score 1290
identically except for missing data. They do differ in our matrix, so we have kept them 1291
separate instead of merging them into a Tuditanidae OTU (which would follow the 1292
classification by CG78). Indeed, the added “tuditanid” Crinodon (see below) has more in 1293
common with Tuditanus than Asaphestera does, despite being easy to distinguish from both. 1294
Bolt & Rieppel (2009) redescribed Llistrofus and several characters of 1295
Hapsidopareion, a “microsaur” included in this matrix. According to them, there is only one 1296
character in which Llistrofus and Hapsidopareion are known to differ (and which cannot be 1297
ascribed to differential preservation or preparation): in the former, the frontals participate in 1298
the orbit margins (FRO 4(1) in this matrix), while the prefrontals of the latter contact the 1299
postfrontals, excluding the frontals from the orbit margins (FRO 4(0)). Because Llistrofus 1300
preserves many characters that are unknown in Hapsidopareion, we have used it to fill in 1301
missing data in the Hapsidopareion OTU. In Appendices 1 and 2, we indicate which of the 1302
scores of the resulting compound Hapsidopareion/Llistrofus OTU are only known for 1303
30
30
Llistrofus; FRO 4 is scored as polymorphic. Unfortunately, Llistrofus does not add to the very 1304
poorly known postcranium of Hapsidopareion. 1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
Figure 4: Left 1316
forelimb of NMS 1317
G 1993.54.1, the 1318
holotype of 1319
Casineria kiddi, 1320
in plantar view. 1321
The carpus is 1322
unossified except 1323
for what may be a 1324
poorly ossified 1325
distal carpal 1. 1326
The inset shows 1327
phalanx IV-5 at a 1328
higher magnifi-1329
cation (length 1330
approximately 1 1331
mm); note that the 1332
tip, although 1333
curved plantarly, 1334
is blunt, unlike in 1335
amniotes and 1336
contrary to Paton, 1337
Smithson & Clack 1338
(1999). 1339
Abbreviation: mc, 1340
metacarpal. The 1341
scale bar is 1342
approximate. 1343
Photos taken by 1344
D. M. 1345
1346
1347
1348
1349
1350
1351
1352
1353
31
31
1354
1355
Figure 5: Right cleithrum of NMS G 1993.54.1, the holotype of Casineria kiddi, in caudal 1356
view, showing postbranchial lamina (compare fig. 70-2.4 of Pawley, 2006). Total length 1357
about 11.5 mm. Photo taken by D. M. 1358
1359
Taxa added as new OTUs for a separate set of analyses 1360
1361
Several theoretical considerations suggest that taxon exemplars should be as diverse as possible 1362
[…] [five references]. Importantly, a recent study based on simulations of true phylogenies 1363
(Salisbury & Kim, 2001) indicates that dense and random taxon sampling increases the 1364
probability of retrieving correctly the plesiomorphic condition of characters as well as the 1365
ancestral state near the tree root. Furthermore, Salisbury & Kim’s (2001) simulations show that 1366
in the analysis of small clades, estimates of ancestral states are strongly affected by cladogram 1367
topology and by the number of descendent [sic] branches in progressively more distal internal 1368
nodes. 1369
– Ruta, Coates & Quicke (2003: 255) 1370
1371
Also, taxon removal because of incomplete preservation and missing character scores may be 1372
undesirable, because such taxa may have a positive effect on cladogram resolution […] [five 1373
references]. 1374
– Ruta, Coates & Quicke (2003: 257) 1375
1376
We recommend the inclusion of as many fossils as possible in any phylogenetic analysis. 1377
– Conrad & Norell (2015) 1378
1379
Analogously to that of Marjanović & Laurin (2008b, 2009), the main goal of the present work 1380
is to find out which hypothesis on the origin of the modern amphibians the matrix by RC07 1381
supports if we keep, as far as possible, its taxon and character sample intact. However, the 1382
Early Permian (Cisuralian) Gerobatrachus was described too late (Anderson et al., 2008a) to 1383
be included in the matrix of RC07, yet it is highly relevant because the phylogenetic analysis 1384
included in its description supported the PH. Being a temnospondyl with many similarities to 1385
lissamphibians, Gerobatrachus may also be considered to support the TH over the LH; 1386
indeed, the analysis by Maddin, Jenkins & Anderson (2012) found it to be the oldest known 1387
lissamphibian while supporting the TH. However, this is based on a further development of 1388
the matrix by Anderson et al. (2008a) that did not take the reevaluations by Marjanović & 1389
Laurin (2009) and Sigurdsen & Green (2011) into account. Following the example of 1390
Marjanović & Laurin (2008), we have therefore added Gerobatrachus to the present matrix in 1391
order to test if it changes the results. 1392
At this opportunity we also added the following taxa: 1393
Chroniosaurus is by far the most thoroughly described representative (Clack & 1394
Klembara, 2009; Klembara, Clack & Čerňanský, 2010) of the enigmatic 1395
Chroniosuchia, with most of the skeleton being preserved. These animals have 1396
occasionally (e.g. Laurin, 2000; Klembara, Clack & Čerňanský, 2010) been 1397
considered embolomeres (thus anthracosaurs) mainly because of their embolomerous 1398
centra, but their confusing mosaic of character states is compatible with a number of 1399
other phylogenetic positions as well; indeed, Klembara et al. (2014) recovered them as 1400
the sister-group to a gephyrostegid-seymouriamorph clade, while Schoch, Voigt & 1401
32
32
Buchwitz (2010) and Buchwitz et al. (2012) even found them to be lepospondyls. In 1402
any case, Chroniosaurus may influence the positions of the embolomeres, 1403
Gephyrostegus, Bruktererpeton, Silvanerpeton,perhaps Caerorhachis, and possibly 1404
Solenodonsaurus in our tree. – The specimens described and figured by Clack & 1405
Klembara (2009) did not preserve much of the lower jaws; Clack & Klembara (2009: 1406
21) therefore stated that they had scored the reconstruction by Ivachnenko & 1407
Tverdochlebova (1980) in their phylogenetic analysis. This reconstruction 1408
(Ivachnenko & Tverdochlebova, 1980: drawing 1б) would allow unambiguous scoring 1409
of our characters 146 (PSYM 1) and 154 (SPL 2), which Clack & Klembara (2009) 1410
scored as unknown, and 155 (SPL 3-4), which they scored as having our state 1 or 2. 1411
However, we have kept the scores of Clack & Klembara (2009) because all of the 1412
drawings in Ivachnenko & Tverdochlebova (1980) are reconstructions which often do 1413
not indicate which parts are actually known; the characters in question concern those 1414
parts of the lower jaw that are most likely to be incompletely preserved or crisscrossed 1415
by fractures (as chroniosuchian dermal bones usually are). We should note that the 1416
reconstruction of the skull proper in all three views (dorsal, ventral, lateral: drawings 1417
1а, б, в) differs appreciably from those of Clack & Klembara (2009) and Klembara, 1418
Clack & Čerňanský (2010) in the shape and proportions of the skull, the course of 1419
some sutures, and the size, shape and relative positions of all openings, in two cases 1420
even their presence. The text (Ivachnenko & Tverdochlebova, 1980: 22) only briefly 1421
describes the chroniosuchian lower jaw in general, without mentioning which 1422
information is based on Chroniosaurus or Chroniosuchus, and does not refer to the 1423
characters in question. Specimen drawings or photos of the lower jaw of any 1424
chroniosuchian have never been published to our knowledge, except for part of the 1425
labial side of that of Chroniosaurus in Clack & Klembara (2009: fig. 2) which does 1426
not provide further information. 1427
Further temnospondyls other than stereospondylomorphs: 1428
o Micropholis is an amphibamid, which means that the TH and the PH consider 1429
it close to the ancestry of some or all extant amphibians. Within 1430
Amphibamidae, it is usually found to lie far away from Doleserpeton, 1431
Amphibamus and Gerobatrachus (e.g. Maddin et al., 2013b); this is probably 1432
why it has never (to our knowledge) been included in a phylogenetic analysis 1433
together with lissamphibians. However, it is the only dissorophoid known to 1434
have two widely spaced occipital condyles (like the modern amphibians), and 1435
it shares the Early Triassic age of the oldest known lissamphibians (the stem-1436
salientians Triadobatrachus and Czatkobatrachus), unlike the Cisuralian 1437
Doleserpeton and Gerobatrachus or the Pennsylvanian Amphibamus. Growth 1438
series spanning most of the skeleton are known (Schoch & Rubidge, 2005). – 1439
It is particularly surprising that RC07 did not add Micropholis to their matrix; 1440
they speculated that it might “turn out to occupy a more derived position than 1441
Doleserpeton on the amphibian stem group” (RC07: 86) because of its 1442
stratigraphic position. 1443
o Nigerpeton was described (Steyer et al., 2006; Sidor, 2013) as a cochleosaurid 1444
edopoid temnospondyl. The edopoids are thought to be close to the base of 1445
Temnospondyli – indeed, some analyses of temnospondyl phylogeny have 1446
used one or more edopoids as the outgroup or part thereof (e.g. Schoch & 1447
Witzmann, 2009a, b) – so we expected an influence on the position of 1448
Temnospondyli and on the interrelationships of its largest constituent groups. 1449
The skull and lower jaw have been described, though their surface is mostly 1450
badly preserved. 1451
33
33
o Saharastega, of which likewise the entire skull is known, though the 1452
preservation of the surface is again bad (Damiani et al., 2006), has had an 1453
unstable phylogenetic position close to the root of Temnospondyli. Based on a 1454
few character states that are rather odd for a temnospondyl, it has even been 1455
suggested to be a seymouriamorph, though not in the peer-reviewed literature 1456
(Yates, 2007). It almost certainly is a temnospondyl, but, as for Nigerpeton, we 1457
expected an influence on the relationships and the large-scale phylogeny of the 1458
temnospondyls. In addition, this is of course an opportunity to clarify the 1459
position of Saharastega itself. 1460
o Iberospondylus is a rather early temnospondyl known from most of the skull 1461
and various postcranial remains, preserving an unusual combination of 1462
plesiomorphies and apomorphies such as the dissorophoid-like dorsal process 1463
on the quadrate. The six phylogenetic analyses that have included it so far 1464
(Laurin & Soler-Gijón, 2001, 2006; Pawley, 2006: fig. 44; Schoch & 1465
Witzmann, 2009a, b; Dilkes, 2015a) have given six different results, and two 1466
of them were based on very small matrices. The present matrix, large as it is, is 1467
an opportunity to clarify the position of Iberospondylus; as the latter does not 1468
seem to belong to any of the large recognized temnospondyl clades, the 1469
interrelationships of these clades could also be influenced. – Both of us have 1470
studied all three specimens (PU-ANF 14, 2 and 15; the first is the holotype). 1471
o Tungussogyrinus has occasionally been considered a caudate, which would be 1472
highly interesting considering its Early Triassic age. Werneburg (2009) 1473
redescribed it as the sister-group to all other branchiosaurids, but noted 1474
similarities to lissamphibians, especially one apomorphy shared with Salientia 1475
that is coded in this matrix. It might thus bolster the TH or certain versions of 1476
the PH. – RC07 (p. 86) mentioned Tungussogyrinus together with Micropholis. 1477
o Acanthostomatops (Witzmann & Schoch, 2006a) is a Cisuralian zatracheid 1478
temnospondyl thought to lie close to Dissorophoidea and/or Eryops. As 1479
pointed out but not tested by Witzmann & Schoch (2006a), it may thus 1480
influence temnospondyl phylogeny as well as the position of Lissamphibia. 1481
o Palatinerpeton is a temnospondyl known from an incompletely preserved, 1482
immature skeleton (Boy, 1996). Endowed with an unusual combination of 1483
characters and suggested to be a stereospondylomorph (Schoch & Milner, 1484
2000) or in a trichotomy with Stereospondylomorpha and an 1485
Eryopidae/Zatracheidae/Parioxys clade (Boy 1996; see “Taxa that were not 1486
added”, below, for Parioxys), it has the potential to influence temnospondyl 1487
phylogeny and thus perhaps even the distance (in number of steps) between the 1488
hypotheses on the origin(s) of the modern amphibians. 1489
o Erpetosaurus, redescribed by Milner & Sequeira (2011) on the basis of many 1490
flattened specimens that range from skull fragments to incomplete articulated 1491
skeletons, is a dvinosaurian temnospondyl, thought to be most closely related 1492
to Isodectes in the taxon sample of RC07. We have added it in order to confuse 1493
things – in other words, to test the robusticity of this matrix against homoplasy: 1494
Erpetosaurus shows a confusing mix of unexpected plesiomorphies and 1495
unexpected similarities to colosteids (among others). It further has some 1496
potential to clarify the intra- and interrelationships of the few dvinosaurs 1497
included here. 1498
o Mordex was neglected for a long time because it is mostly known from larvae 1499
that were confused with Branchiosaurus, Platyrhinops and other dissorophoids 1500
from the same site. Recently, however, it has been understood as the oldest 1501
34
34
known trematopid (Milner & Sequeira, 2003; Milner, 2007; Werneburg, 1502
2012b); Milner (2007) called it “most similar to Ecolsonia” (tentatively 1503
confirmed by Schoch, 2012, and Schoch & Milner, 2014), which implies that 1504
adding Mordex to the presentanalysis will influence the unstable phylogenetic 1505
position of Ecolsonia. Although it is not very well known – the largest 1506
specimen is an incomplete skull roof, the others are all much smaller and 1507
incompletely ossified (Werneburg, 2012b) –, it differs from all other taxa in 1508
this matrix. 1509
o Branchiosaurus used to be the name given to almost all skeletally immature 1510
temnospondyls from the Czech Pennsylvanian, as well as occasionally the 1511
Pennsylvanian and Cisuralian of other places. Building on and greatly 1512
expanding the work of Milner & Sequeira (2003), Werneburg (2012b) tried to 1513
sort out this confusion and provided a preliminary redescription of the type 1514
species, B. salamandroides from the Czech Republic. This redescription 1515
confirms that Branchiosaurus differs in several features from all other 1516
branchiosaurids (see Milner & Schoch, 2008); some of these features are coded 1517
in the present matrix. Branchiosaurus is further necessary to test the 1518
phylogenetic position of Tungussogyrinus (see above). – As Werneburg 1519
(2012b) noted, B. fayoli from Commentry and “B. aff. fayoli” from Montceau-1520
les-Mines (both France) need to be redescribed before it can be assessed 1521
whether they form a clade with B. salamandroides; we have therefore not used 1522
any information from them. 1523
“Microsaurs”: 1524
o Utaherpeton was a surprising omission by RC07, given the facts that it is 1525
among the oldest known “microsaurs” and has been considered a basal 1526
“microsaur” (Carroll, Bybee & Tidwell, 1991; Carroll & Chorn, 1995; 1527
Anderson, 2001; Anderson et al., 2008a) or the sister-group of Microbrachis 1528
and thus a basal member of the “microsaur”-lysorophian-lissamphibian clade 1529
(Vallin & Laurin, 2004); in a more recent analysis it fell out close to a 1530
“nectridean”-aïstopod-lysorophian-lissamphibian clade (Marjanović & Laurin, 1531
2009: Electronic Supplementary Material 2). Similarities to the “nectrideans” 1532
were already noted in the original description (Carroll, Bybee & Tidwell, 1533
1991). A variety of interesting effects on lepospondyl intra- and perhaps even 1534
interrelationships could be expected from its addition to the present matrix 1535
based on the two descriptions (Carroll, Bybee & Tidwell, 1991; Carroll & 1536
Chorn, 1995). 1537
o The “Goreville microsaur”, too, is among the oldest known “microsaurs”. 1538
Although it was deliberately not named by Lombard & Bolt (1999), it differs 1539
from all other OTUs in this matrix, so we do not see a reason not to include it; 1540
its somewhat unusual combination of plesio- and apomorphic character states 1541
(noted in its description) could change the topology of the tree. Eight badly 1542
preserved specimens, amounting to most of the skeleton, are known. 1543
1544
1545
↓ Figure 6: NHMW 1899-0003-0006 (formerly 1899-III-6), a specimen referred to 1546
Sparodus validus. For an interpretative drawing, see Carroll (1988: pl. XII), which, however, 1547
makes the vertebrae appear much flatter than they are and omits the unusually well preserved 1548
scales. Some of the scales are only visible as striations in the matrix (see Fig. 7), which is 1549
usual in other “microsaurs” and in micromelerpetid temnospondyls; but most retain thick 1550
bone. Photo taken by D. M. 1551
35
35
1552
1553
36
36
1554
1555
Figure 7: Ventral view of the palate and lingual view of the left lower jaw of the 1556
Sparodus specimen shown in Fig. 6. Carroll (1988: fig. 1A) provided an interpretative 1557
drawing of a latex cast. The photograph was taken by holding a digital camera to an ocular of 1558
a binocular microscope. Abbreviations: l., left; r., right; .c, largest caniniform tooth (on the 1559
maxilla); .t, tusk (one of two per bone); cor, unidentified coronoid; d, dentary; hu, humerus; 1560
j, jugal; mx, maxilla; orb, orbit; pal, palatine; pmx, premaxilla; po, postorbital; pof, 1561
postfrontal; sc, scales preserved as striations. Photo taken by D. M. 1562
1563
1564
o Sparodus is another “microsaur” that is not obviously deeply nested in one of 1565
the universally recognized clades, showing affinities to both Pantylidae and 1566
Gymnarthridae but having fewer apomorphies than the undisputed members of 1567
both (and being older than most of them). Thus, it may clarify the relationships 1568
of one or both of these clades. Furthermore, “microsaur” phylogeny is so 1569
37
37
unstable that any addition of a taxon may drastically change it once again; 1570
changes in “microsaur” phylogeny moreover affect the other lepospondyls, and 1571
with them the positions of the modern amphibians. – Personal observation by 1572
D. M. of NHMW 1899-0003-0006 (Fig. 6, 7), the referred specimen described 1573
by Carroll (1988), shows that Carroll’s (1988: fig. 1) drawing and 1574
reconstruction of the skull are probably optimistic, unless the sutures are better 1575
visible on the latex cast Carroll used (which we have not seen) than on the 1576
fossil itself. However, this concerns very few characters. 1577
o Carrolla is a brachystelechid “microsaur” known only from a skull with lower 1578
jaw. Few differences between it and Batropetes (the brachystelechid that was 1579
already included in the matrix) are known, and few of those concern characters 1580
in the present matrix, but the redescription by Maddin, Olori & Anderson 1581
(2011) based on high-resolution computed tomography has made a large 1582
amount of information available. This is especially important because it allows 1583
us to better test the results by Vallin & Laurin (2004), who found 1584
Brachystelechidae to be the sister-group of Lysorophia + Lissamphibia, as well 1585
as the results by Maddin, Jenkins & Anderson (2012), who found the 1586
brachystelechids to be nested in a clade (Recumbirostra) that contains all the 1587
other “microsaurs” with a snout tip that protrudes beyond the premaxillary 1588
toothrow (state PREMAX 8(1) in the present matrix). – D. M. has seen the 1589
only known specimen (TMM 40031-54) and was consequently able to score 1590
some characters that concern externally visible anatomy and were 1591
insufficiently addressed by Maddin, Olori & Anderson (2011); these are 1592
described in Appendix 1. 1593
o Crinodon is a “microsaur” that is rich in plesiomorphies. Known from a well-1594
preserved dermatocranium and partial lower jaws (CG78) – parts of the body 1595
that the present craniocentric matrix is well equipped to deal with –, it may 1596
help shed light on “microsaurian” intra- and interrelationships. CG78 classified 1597
it as a tuditanid, but the two “tuditanids” in the present matrix (Tuditanus and 1598
Asaphestera) were not found as sister-groups by RC07, leaving us wondering 1599
whether the diagnostic characters of “Tuditanidae” might all be 1600
symplesiomorphies or homoplasies. 1601
o Trihecaton is a large “microsaur” known from an articulated skeleton that 1602
lacks the skull (apart from the maxilla, the dorsal edge of which is not 1603
prepared), part of the lower jaw, most of the distal limbs and much of the tail – 1604
although further preparation would probably reveal additional bones. 1605
Presumably because most of the skull is unknown, Trihecaton has never 1606
featured in a phylogenetic hypothesis since CG78 gave it its own family in 1607
Tuditanomorpha; yet, it differs from all other OTUs in this matrix, so we see 1608
no reason to continue its exclusion from phylogenetic analyses. D. M. coded 1609
Trihecaton directly from CM 47681 and CM 47682 (which most likely belong 1610
to the same individual); it is possible that additional preparation has been done 1611
since the description by CG78. 1612
o Quasicaecilia, known from an isolated skull without lower jaws (except an 1613
articular fragment), was redescribed by Pardo, Szostakiwskyj & Anderson1614
(2015b). It is a very small brachystelechid that may contribute to resolving the 1615
relationships of brachystelechids, other burrowing “microsaurs”, lysorophians 1616
and lissamphibians; additionally or alternatively, it could highlight the effects 1617
of miniaturized taxa on phylogenetic analyses. 1618
38
38
The diadectomorphs (in the present matrix: Diadectes, Orobates, Tseajaia, 1619
Limnoscelis) are usually thought to lie on the amniote stem. It has, however, been 1620
proposed (Berman, Sumida & Lombard, 1992; Berman, 2013) that they are actual 1621
amniotes – in particular, the closest known relatives of Synapsida, thus part of 1622
Theropsida, the sister-group of Sauropsida (Goodrich, 1916). Surprisingly, all 1623
unquestioned amniotes in the matrix by RC07 – Petrolacosaurus, Paleothyris, 1624
Captorhinus – are sauropsids; the lack of unambiguous theropsids means that the 1625
matrix was unable to test the mentioned hypothesis, even though its large amount of 1626
non-amniote OTUs would have made it very well suited for such a purpose. The two 1627
synapsid OTUs we have added help to solve this problem. Furthermore, they might 1628
further influence diadectomorph monophyly – which was not found by Ruta, Coates & 1629
Quicke (2003), Pawley (2006), RC07, or even Germain (2008a). 1630
o We have included Eothyris, Oedaleops and Eocasea as a single OTU called 1631
Caseasauria. Eothyris is known from an isolated, almost complete skull and 1632
lower jaw (Reisz, Godfrey & Scott, 2009); Oedaleops is known from a less 1633
complete skull and lower jaw, additional skull fragments, vertebrae, ribs, and 1634
girdle and limb elements (Reisz, Godfrey & Scott, 2009; Sumida, Pelletier & 1635
Berman, 2014; we follow the latter in not including the ilium tentatively 1636
referred by Langston, 1965); Eocasea is known from a skeleton lacking the 1637
shoulder girdle, the forelimbs, the distal part of the tail and most of the skull 1638
(Reisz & Fröbisch, 2014). The overlapping parts of these three taxa score 1639
almost identically in our matrix; where they disagree, we have simply scored 1640
polymorphism (and noted this in Appendix 1), because the relationships of 1641
these taxa to each other and to the herbivorous members of Caseidae, which 1642
are not considered here, remain unclear (Reisz, Godfrey & Scott, 2009; 1643
Sumida, Pelletier & Berman, 2014; Reisz & Fröbisch, 2014; Brocklehurst, 1644
Romano & Fröbisch, 2016; though see Brocklehurst et al., 2016).– Some data 1645
of Eothyris not presented in the cited publications were filled in by D. M. by 1646
observation of the holotype, MCZ 1161; they agree with the conditions in 1647
Oedaleops (Reisz, Godfrey & Scott, 2009) and Eocasea (Reisz & Fröbisch, 1648
2014). We did not consider Callibrachion or Datheosaurus, which are poorly 1649
preserved and incompletely ossified (Spindler, Falconnet & Fröbisch, 2015), or 1650
Vaughnictis, which was (re)described too late for us (Brocklehurst et al. 2016); 1651
they should be taken into account in future analyses, however, because all the 1652
analyses by Brocklehurst et al. (2016: fig. 10, 11) fully resolved the 1653
phylogenetic positions of all taxa mentioned in this paragraph. 1654
o Supplementing the Caseasauria OTU is Archaeovenator, the oldest known 1655
varanopid and sister-group to all other varanopids, known from a largely 1656
complete, mostly articulated skeleton described by Reisz & Dilkes (2003) and 1657
recently found nearly as close to the origin of Synapsida as Caseasauria (Reisz 1658
& Fröbisch, 2014: appendix S5; Brocklehurst et al., 2016). 1659
Lissamphibians: 1660
o Liaobatrachus is an Early Cretaceous salientian known from plentiful, very 1661
well preserved articulated skeletons (Dong et al., 2013). Although it was found 1662
to form a trichotomy with the two component clades of the crown-group of 1663
frogs, which means it is not very close to the root of Salientia, it may still 1664
influence the position of Salientia in particular and Lissamphibia in general, 1665
because few of the more stemward salientians are well preserved. 1666
o Beiyanerpeton is a recently described (Gao & Shubin, 2012) neotenic 1667
salamandroid of apparently Late Jurassic age, in any case older than its Early 1668
39
39
Cretaceous relative Valdotriton which was already included in the matrix. It is 1669
the only known lissamphibian, extant or extinct, to possess grooves for the 1670
lateral-line organ on the skull. Among the lissamphibians in the matrix, it is the 1671
only one known to possess opisthotics that are not fused to the prootics or 1672
exoccipitals (or otherwise absent). These factors make it important enough to 1673
add it to this matrix (which contains both of these characters), even though its 1674
paedomorphosis makes some characters difficult to interpret for purposes of 1675
phylogenetics (see Gao & Shubin, 2012: supplementary information). 1676
o The even older (Late or Middle Jurassic) Pangerpeton was found by Jia & Gao 1677
(2016a) to be a stem-hynobiid salamander, similarly close to the root of 1678
Urodela as Beiyanerpeton and Valdotriton. Several features of the only known 1679
skeleton (which lacks most of the tail, but is accompanied by a body outline) 1680
suggest that it had undergone metamorphosis, although it remained aquatic 1681
(Wang & Evans, 2006). Indeed, in having the jaw joints and the occiput in the 1682
same plane, it is more peramorphic than Beiyanerpeton, Valdotriton and even 1683
Karaurus. The skeleton is preserved in ventral view, so that most of the skull 1684
roof remains unknown, and it is unfortunate that the photos (even in the PDF 1685
version) have such low resolution that, for example, the ribs are entirely 1686
invisible and had to be scored from sparse remarks in the text that were not 1687
meant for comparison with our taxon sample; still, Pangerpeton does not score 1688
redundantly with any other OTU in this matrix. 1689
o Even more peramorphic than Pangerpeton is Chelotriton. Scoring of this late 1690
Oligocene pleurodeline salamandrid (newt) is based mostly on MB.Am.45 1691
(Marjanović & Witzmann, 2015). This specimen exhibits peramorphic traits 1692
that are otherwise rare or even entirely unknown in salamanders (in a few 1693
cases even in lissamphibians as a whole), such as very long ribs, some of 1694
which are ventrally curved, and jaw joints that lie far caudal of the occiput. In 1695
1981, it was misidentified as a “Dissorophid temnospondyl” “?Amphibamus” 1696
on a label. MB.Am.45 is indeed strikingly amphibamid-like in appearance, 1697
with some similarity also to certain diplocaulid “nectrideans”. We have added 1698
Chelotriton to the matrix to see the effects, if any, of its unexpected 1699
peramorphic features on caudate and lissamphibian intra- and 1700
interrelationships. For this purpose, we took MB.Am.45 (scored mostly from 1701
the silicone cast MB.Am.45.3) at face value wherever possible rather than 1702
scoring polymorphism; missing data were filled in from the less extremely 1703
peramorphic Chelotriton individuals described and figured by Roček & 1704
Wuttke (2010) and Schoch, Poschmann & Kupfer (2015). 1705
Undoubted colosteids: 1706
o Deltaherpeton is one of the oldest known colosteids, represented by a skull 1707
roof and lower jaw (Bolt & Lombard, 2001, 2010). It shows a few 1708
plesiomorphies as well as several unique features that are absent or not 1709
preserved in Colosteus and Greererpeton. Bolt & Lombard (2010) drew 1710
attention to this fact, but then undertook only a comparison instead of a 1711
phylogenetic analysis to justify their conclusion that “[t]he morphology of 1712
Deltaherpeton and the revised data presented for colosteids do not clarify the 1713
relationship of colosteids to other early tetrapods” (Bolt & Lombard, 2010: 1714
abstract). Indeed, there are characters that indicatea very rootward position of 1715
Colosteidae (as found by RC07), but others support a position more 1716
crownward than those of Whatcheeriidae and Crassigyrinus (as found by Ruta, 1717
2009, if we assume that the temnospondyls would be close to or within the 1718
40
40
crown), and a few have led to the traditional concept of colosteids as 1719
temnospondyls (the sister-group to all other members of that clade). 1720
o Pholidogaster has been known for over 150 years; it was redescribed by 1721
Romer (1964) and, after additional preparation of the skull and shoulder girdle, 1722
Panchen (1975). Its identity as a colosteid has not been questioned since the 1723
latter publication; and yet, to the best of our knowledge, it has never been 1724
included in a phylogenetic analysis. Although it is generally less well known 1725
than Colosteus, Greererpeton and Deltaherpeton, it scores differently from all 1726
three in our matrix; unlike in any known specimen of the other three colosteids 1727
(Bolt & Lombard, 2010), the region around the external naris and the 1728
septomaxilla is preserved. Its age, close to that of Deltaherpeton, makes it a 1729
particularly interesting potential source of information on the phylogeny and 1730
affinities of Colosteidae. 1731
Karpinskiosaurus is a seymouriamorph that may help clarify seymouriamorph 1732
phylogeny and the affinities of Seymouriamorpha. The skulls of the type species, K. 1733
secundus, including an adult, non-paedomorphic one, were redescribed by Klembara 1734
(2011); according to the same work, the type specimen of K. ultimus requires revision, 1735
so this species is not considered here. Much of the axial skeleton was illustrated and 1736
commented on by Bystrow (1944). 1737
Anthracosaurs: 1738
o Holmes & Carroll (2010) have described a small but adult (or nearly so) 1739
incomplete skeleton of an anthracosaur (embolomere) from Joggins (Late 1740
Carboniferous). Unfortunately, NSM 994 GF 1.1 cannot be assigned to a 1741
named genus, because it does not overlap with diagnostic parts of the only 1742
known anthracosaur from the same site, Calligenethlon; however, because it is 1743
considerably more complete than any specimen which can be assigned to 1744
Calligenethlon (all of which are slightly smaller and possibly less mature), we 1745
have added it to this matrix. Being an unusually small anthracosaur, it 1746
improves our representation of anthracosaur diversity and may contribute to 1747
the clarification of anthracosaurian intra- and interrelationships. 1748
o Palaeoherpeton is less well known than the originally included anthracosaurs, 1749
but the material was very thoroughly described by Panchen (1964) and scores 1750
differently from any other OTU in our matrix. We have therefore included it in 1751
the hope that it will contribute to resolving the rather unstable phylogeny of 1752
anthracosaurs. 1753
o Neopteroplax, an anthracosaur described by Romer (1963) from a largely 1754
complete skull with lower jaw, has been mostly ignored ever since. In spite of 1755
having a skull shape quite similar to that of Pholiderpeton attheyi, however, 1756
Neopteroplax shows a number of features that are unusual for embolomeres; it 1757
makes an interesting addition to the question of anthracosaur phylogeny and 1758
affinities. We consider the referral of isolated jaw fragments and centra (not 1759
clear if pleuro- or intercentra) from a younger, very distant site to 1760
Neopteroplax (Romer, 1963: 451) highly doubtful; for the centra in particular 1761
it requires a rather long chain of inference from a single potentially diagnostic 1762
feature (the distance between the teeth). The jaw fragments would not add any 1763
information to our scoring of Neopteroplax; we have ignored the centra to 1764
avoid creating a chimera. 1765
Mississippian enigmas: 1766
o Sigournea is known only from an isolated lower jaw of mysterious affinities 1767
(Bolt & Lombard, 2006). Although similarities to the baphetoids, especially 1768
41
41
Spathicephalus (see below), and to the equally mysterious jaw material called 1769
Doragnathus (see below) have been noted (Bolt & Lombard, 2006; Milner, 1770
Milner & Walsh, 2009), these conjectures have only ever been tested in a 1771
phylogenetic analysis by Sookias, Böhmer & Clack (2014), who found it in the 1772
whatcheeriid region of the tree (more rootward than Baphetes, the only 1773
included baphetoid). That analysis did not include Doragnathus and had a 1774
generally insufficient taxon and, likely, character sample to address this 1775
question. Our matrix contains enough lower-jaw and tooth characters to show 1776
that Sigournea is distinct from any other taxon for which more than a few such 1777
characters can be scored. 1778
o Doragnathus is another mysterious animal known only from lower jaws and 1779
parts of upper jaws. It has never before been included in a phylogenetic 1780
analysis, even though it was described quite some time ago (Smithson, 1980), 1781
and even though it differs from Sigournea, Spathicephalus (see below) and all 1782
other potential relatives in characters that are included in the present matrix. 1783
As recommended by Smithson & Clack (2013), we have not scored the 1784
isolated postcranial material that was found at the same site and may or may 1785
not belong to the same taxon (described and illustrated by Smithson & Clack, 1786
2013). 1787
o Spathicephalus is a baphetoid known from skull and lower-jaw material with 1788
several strange autapomorphies (Beaumont & Smithson, 1998). Still, in a few 1789
characters it may be more plesiomorphic than at least Baphetes and 1790
Megalocephalus, which may help test baphetoid relationships (see Milner, 1791
Milner & Walsh, 2009); it may also influence the positions of Sigournea and 1792
Doragnathus. 1793
o MB.Am.1441, the “St. Louis tetrapod”, is a natural mold of a skull with lower 1794
jaws in ventral view that was described by Clack et al. (2012b). The 1795
description noted several similarities to colosteids, but also to temnospondyls, 1796
and ultimately did not commit to either hypothesis of relationships; the 1797
phylogenetic analysis in that paper, which used a very small taxon sample and 1798
a rather repetitive character sample, found it to be the sister-group of a novel 1799
(Ptyonius, (Adelogyrinus, Greererpeton)) clade, which together with the “St. 1800
Louis tetrapod” formed the sister-group to the only two included 1801
temnospondyls – an unusual arrangement as well. The matrix presented here 1802
may be better suited to resolving the relationships of MB.Am.1441, which D. 1803
M. was able to study firsthand (both the natural mold, MB.Am.1441.1, and the 1804
silicone cast, MB.Am.1441.2). Despite the small number of characters that can 1805
be scored, the “St. Louis tetrapod” differs from all other OTUs in this matrix. 1806
Devonian enigmas: 1807
o Ruta, Coates & Quicke (2003: 262) listed Metaxygnathus among Devonian 1808
taxa that are “known mainly from lower jaw rami and/or incomplete 1809
postcranial remains” and “are omitted” from their matrix, presumably because 1810
they are so incompletely known. As with Sigournea and Doragnathus, 1811
however, the isolated lower jaw ramus called Metaxygnathus differs from all 1812
other lower jaws (isolated or not) in this matrix. Indeed, one character, ANG 3, 1813
seems to have been deliberately included to potentially hold Acanthostega and 1814
Metaxygnathus together in “future, expanded versions of our matrix” (RC07: 1815
103) which have not so far been published. Including Metaxygnathus may 1816
influence the positions of Ventastega, Acanthostega and Ichthyostega, perhaps 1817
42
42
even Tulerpeton and the whatcheeriids. Our source was the redescription by 1818
Ahlberg & Clack (1998). 1819
o Ymeria is an “Ichthyostega-grade” animal known from premaxillae,a palate in 1820
dorsal view, and lower jaws (plus shoulder girdle remains that are too 1821
fragmentary to be coded in this matrix). It was found as just one node less 1822
crownward than Ichthyostega in those analyses by Clack et al. (2012a) and 1823
Sookias, Böhmer & Clack (2014) that had enough resolution to tell. Its 1824
inclusion may influence the positions of Ventastega, Acanthostega, 1825
Ichthyostega and Metaxygnathus. 1826
o Densignathus, much like Metaxygnathus, is known only from lower-jaw 1827
material (Daeschler, 2000; Ahlberg, Friedman & Blom, 2005) and was listed 1828
among the excluded taxa by Ruta, Coates & Quicke (2003: 262), but differs 1829
from all other lower jaws in the matrix and may influence the positions of its 1830
fellow Devonian OTUs as well as perhaps the Mississippian whatcheeriids. 1831
o Elginerpeton has been described, from successively greater amounts of 1832
isolated material and jaw fragments, as a Devonian animal close to the origin 1833
of limbs (Ahlberg, 1995, 1998; Ahlberg & Clack, 1998; Ahlberg, Friedman & 1834
Blom, 2005). It could serve to establish character polarities for the other 1835
Devonian OTUs. Following Ahlberg, Friedman & Blom (2005) and Ahlberg 1836
(2011), we have included the postorbital bone and the postcranial material 1837
described by Ahlberg (1998), except for the supposed humerus. 1838
Temnospondyl phylogeny is a vexing question; most mathematically possible 1839
permutations of Edopoidea (which may not be a clade; Pawley, 2006: chapter 6), 1840
Capetus, Dendrerpetidae + Balanerpeton (together mono- or paraphyletic), 1841
Dvinosauria, Eryopidae, Stereospondylomorpha and Dissorophoidea have been 1842
supported by recent phylogenetic analyses, and a few taxa have been found close to 1843
Eryopidae or as early stereospondylomorphs. This is, in part, discussed by Ruta, 1844
Coates & Quicke (2003) (as well as by Pawley, 2006: chapter 5; Ruta, 2009; Schoch, 1845
2013; Dilkes, 2015a; and references in all five). It is strange, then, that the matrices of 1846
Ruta, Coates & Quicke (2003) and RC07 do not contain a single possible 1847
stereospondylomorph despite the presence of representatives of all other clades listed 1848
above. We have therefore added Sclerocephalus, Cheliderpeton, Glanochthon, 1849
Archegosaurus, Platyoposaurus, Konzhukovia, Lydekkerina and Australerpeton. 1850
Cheliderpeton, Platyoposaurus, Konzhukovia, Lydekkerina and Australerpeton were 1851
not included even in the analysis of temnospondyl phylogeny by Ruta (2009); Schoch 1852
(2013; followed by Dilkes, 2015a), who included a comparatively large number of 1853
stereospondyls, still did not include Cheliderpeton, Platyoposaurus or Konzhukovia. 1854
o Sclerocephalus, variously considered an eryopid relative and/or a 1855
stereospondylomorph, is known from complete articulated skeletons and was 1856
redescribed by Meckert (1993; shoulder girdle and forelimb only) and Schoch 1857
& Witzmann (2009a); D. M. compared some scores to the specimen 1858
MB.Am.1346 (a skull roof in dorsal view, a left lower jaw in medial and a 1859
right lower jaw in lateral view). Because Sclerocephalus is a weakly supported 1860
clade of four species (Schoch & Witzmann, 2009a) – ignoring the poorly 1861
known ?S. stambergi, whose phylogenetic position is even less clear 1862
(Klembara & Steyer, 2012) –, we have only considered the type species, S. 1863
haeuseri, which also happens to be the best-known one. It is known from 1864
several borderline amphibious and eryopid-like, as well as strictly aquatic, 1865
vaguely embolomere-like morphs (Schoch, 2009a). 1866
43
43
o Cheliderpeton vranyi, the type species of Cheliderpeton, was redescribed by 1867
Werneburg & Steyer (2002). Like Sclerocephalus, it has variously been 1868
considered an eryopid relative or a stereospondylomorph. The present matrix 1869
may be able to decide this question. Further, Werneburg considered 1870
Cheliderpeton an intasuchid, Steyer preferred to consider it an archegosaurid 1871
(Werneburg & Steyer, 2002); this question is testable in this matrix insofar as 1872
Cheliderpeton may form an exclusive clade with Archegosaurus (see below) or 1873
not. Growth series of the skull roof and a few other bones are known. We have 1874
not been able to use information from the putative second species, Ch. 1875
lellbachae, which “[i]n most respects […] resembles Sclerocephalus haeuseri” 1876
and needs to be (re)described (Schoch & Witzmann, 2009b: 122). 1877
o Glanochthon was variously referred to Archegosaurus, Actinodon, 1878
Sclerocephalus or Cheliderpeton till Werneburg & Steyer (2002) recognized it 1879
as distinct from all four and Schoch & Witzmann (2009b) redescribed it. (For 1880
example, it appears as “Cheliderpeton latirostre” in the analysis by Ruta, 1881
2009.) Found to be closer to Archegosaurus and Stereospondyli than 1882
Cheliderpeton and Sclerocephalus by Schoch & Witzmann (2009b), the two 1883
species (G. latirostris, G. angusta; Schoch & Witzmann, 2009b) are known 1884
from growth series of many skulls (including lower jaws: Boy, 1993) and most 1885
of the postcranium except, apparently, the tail. 1886
o Archegosaurus is known from complete articulated skeletons and was 1887
redescribed by Witzmann (2006: head skeleton) and Witzmann & Schoch 1888
(2006b: postcranium). Together with the other potential archegosaurids, all of 1889
which are much less completely known, it is considered a close relative of 1890
Stereospondyli. We have not used information from Memonomenos, which 1891
was often referred to Archegosaurus in earlier times but belongs elsewhere in 1892
the tree (Schoch & Milner, 2000: fig. 52; Schoch & Witzmann, 2009b). 1893
o Platyoposaurus, known from rich cranial and postcranial material (Efremov, 1894
1932; Konzhukova, 1955; Gubin, 1991), is superficially very similar to 1895
Archegosaurus, but may actually be closer to Stereospondyli (Schoch & 1896
Witzmann, 2009b). 1897
o Konzhukovia has been allied to Archegosaurus and to Stereospondyli in 1898
various sources. Gubin (1991: drawings 6, 15) illustrated and described the 1899
well preserved type skull of the type species, K. vetusta; we have not used 1900
other information except for two comments on the same species by Pereira 1901
Pacheco et al. (2016: appendix 2). First, further information on K. vetusta is 1902
not available (Gubin [1991: fig. 2 of plate II] only showed a small photo of an 1903
additional skull fragment of K. vetusta and merely mentioned the existence of 1904
a lower-jaw fragment and postcranial fragments); second, the referral of K. 1905
tarda and a fortiori K. sangabrielensis to Konzhukovia is doubtful (as 1906
discussed above in the Nomenclature section); third, K. sangabrielensis is only 1907
known from a partial skull that hardly seems to differ from K. vetusta in 1908
characters included in this matrix, while Gubin (1991: fig. 1 of plate II) only 1909
showed small photos of a skull of K. tarda and merely mentioned the existence 1910
of another skull. The diagnoses of the three species do not differ in characters 1911
that the present matrix contains, and the illustrations apparently do not either 1912
(except for the lateral-line grooves of K. sangabrielensis, which K. vetusta 1913
actually shares: Pereira Pacheco et al., 2016: appendix 2). 1914
o Lydekkerina is within a few internodes of the origin of Stereospondyli 1915
according to all recent analyses. It is known from complete skeletons that are 1916
44
44
unusually well ossified for a stereospondyl. The postcranium was recently 1917
described by Pawley & Warren (2005) and Hewison (2008), and the skull and 1918
lower jaw by Jeannot, Damiani & Rubidge (2006) and Hewison (2007). 1919
Shishkin, Rubidge & Kitching (1996) showed that various proportions of the 1920
skull roof of Lydekkerina are paedomorphic, a hypothesis further supported for 1921other characters by Hewison (2007); this probably does not affect any 1922
characters in this matrix. Miniaturization by progenesis may explain the well 1923
ossified endochondral bones. 1924
o Australerpeton was found to lie even closer to the origin of Stereospondyli in 1925
the analysis by Eltink & Langer (2014; also Pereira Pacheco et al., 2016) after 1926
being thought to lie close to Archegosaurus by Schoch & Milner (2000). 1927
Almost the entire skeleton is known (Barberena, 1998; Dias & Schultz, 2003; 1928
Eltink & Langer, 2014). 1929
The “Parrsboro jaw” is an incomplete impression of a lower jaw of Pennsylvanian age, 1930
consistently called NSM 987GF65 in the original description (Godfrey & Holmes, 1931
1989) but NSM 987GH65 in the partial redescription (Sookias, Böhmer & Clack, 1932
2014). It has a unique combination of characters and differs from all other OTUs in 1933
this matrix, which may therefore have the potential to resolve the relationships of this 1934
mysterious fragment that has warranted redescription but defied the phylogenetic 1935
analysis of Sookias, Böhmer & Clack (2014: fig. 5A). 1936
1937
Taxa that were not added 1938
1939
It goes without saying that complete sampling of the clade represented by the matrix, however 1940
desirable for theoretical reasons, is impossible. Thus, we have not added taxa that likely could 1941
not be handled by the character sample of the present matrix (e.g. too deeply nested taxa, such 1942
as any further baphetoids, cochleosaurids, eryopids, dvinosaurs, dissorophids, stereospondylo-1943
morphs, amniotes or lissamphibians beyond those we did add, or taxa too far outside the clade 1944
of limbed vertebrates). Still, a few taxa would have been intuitive candidates for addition, but 1945
we have left them out for the following reasons: 1946
Because of the state of preparation of the specimens, the original description of the 1947
enigmatic and interesting “anthracosaur” Eldeceeon (Smithson, 1994) is very preliminary and 1948
contains little information; Eldeceeon is being redescribed according to Ruta & Clack (2006). 1949
Although the abovementioned Madygenerpeton is only the second chroniosuchian for 1950
which a thorough description exists, we have not included it because it seems to lie at the tip 1951
of a long branch. According to the first phylogenetic analysis of Chroniosuchia ever 1952
performed (Schoch, Voigt & Buchwitz, 2010), it is highly nested within that clade as the 1953
sister-group of Chroniosaurus, so that its seemingly plesiomorphic aquatic adaptations are 1954
probably reversals and would likely, perhaps together with the many other autapomorphies 1955
that Madygenerpeton has, distort the tree in general and the position of Chroniosaurus in 1956
particular – as long as other chroniosuchians cannot be added for lack of information (or lack 1957
of overlap with the character sample). Indeed, Madygenerpeton lacks traces of the lateral-line 1958
organ (Schoch, Voigt & Buchwitz, 2010; D. M., pers. obs.) which would be expected to be 1959
present in a primarily aquatic vertebrate; the dorsal rims of the eye sockets are raised high 1960
above the skull table, implying a crocodile-like predator that looked for terrestrial prey rather 1961
than staying underwater, an interpretation further corroborated by the nostrils which are 1962
apparently raised above the roof of the snout. In temnospondyls, for example, the nostrils are 1963
always well below eye level, and the eye sockets are never noticeably raised (D. M., pers. 1964
obs., and see below). The second analysis that included Madygenerpeton (Buchwitz et al., 1965
2012) found it as the sister-group to all other chroniosuchians, but – much like its predecessor 1966
45
45
– had a very small sample of outgroups. Conversely, the present matrix undersamples 1967
characters relevant to chroniosuchian phylogeny. 1968
The temnospondyl Parioxys ferricolus, variously allied to Eryops and/or 1969
Dissorophoidea and more specifically found as the sister-group of Iberospondylus by Pawley 1970
(2006: chapter 5), is most likely a chimeric taxon (J. Pardo, pers. comm. 2012). The lectotype 1971
(AMNH 4309) is covered by a crust that would be difficult to prepare away and completely 1972
obscures the sutures as well as most other details (D. M., pers. obs.; Schoch & Milner, 2014: 1973
77), and referred specimens (AMNH 2445 and almost all of the MCZ material) are being 1974
restudied by M. Ruta (D. M., pers. obs. of loan slips in 2013). “Preparation of USNM material 1975
has revealed many similarities to Cacops” (a dissorophid like Broiliellus; Schoch & Milner, 1976
2014: 77); it remains to be seen if the same will hold true of the lectotype. – AMNH 7118, the 1977
type and only known specimen of P. bolli, consists only of encrusted (D. M., pers. obs.) 1978
assorted postcrania, and whether it (or any other material currently assigned to Parioxys) can 1979
be referred to Parioxys awaits investigation. 1980
Onchiodon is well known, but so similar to the even better known Eryops (Werneburg, 1981
2007a) that it would most likely not add any new information to this matrix. Pretty much the 1982
same holds for the less well known Glaukerpeton (Werneburg & Berman, 2012). The rest of 1983
Eryopidae – and also the diversity within Eryops – is very poorly understood and currently 1984
undergoing revision (Werneburg, 2007a, 2012a; Werneburg & Berman, 2012; Schoch & 1985
Milner, 2014; Rasmussen, Huttenlocker & Irmis, 2016). 1986
Comparison to character lists in Fröbisch & Reisz (2012), Schoch (2012), Maddin et 1987
al. (2013b) and Holmes, Berman & Anderson (2013) shows that many characters relevant to 1988
dissorophoid phylogeny are lacking from this matrix. Adding these characters and a better and 1989
more even sample of dissorophoid temnospondyls – micromelerpetids (in our matrix 1990
represented only by Micromelerpeton), dissorophids (Broiliellus brevis), trematopids 1991
(Ecolsonia, Acheloma, Phonerpeton; Mordex) and amphibamids (Amphibamus, Eoscopus, 1992
Doleserpeton, Platyrhinops, three “branchiosaurids”; Gerobatrachus, Micropholis, two more 1993
“branchiosaurids”) – will be part of future work. As Schoch (2012), Werneburg (2012b), 1994
Maddin et al. (2013b) and Holmes, Berman & Anderson (2013) pointed out, some of these 1995
taxa need to be (or are being) redescribed, and some material probably even needs additional 1996
preparation, which is in many cases as difficult to do as for Parioxys (see above). In 1997
particular, as mentioned above, we have not used information from species currently referred 1998
to Broiliellus other than B. brevis. 1999
Kirktonecta (Clack, 2011b), “the oldest known microsaur”, is so poorly preserved (and 2000
prepared – it is split through the bone like Eldeceeon) that it would not differ from several 2001
other OTUs in this matrix if we added it. 2002
The characters in which Czatkobatrachus differs from Triadobatrachus (Evans & 2003
Borsuk-Białynicka, 2010) are not represented in this matrix. It would be possible to code 2004
Czatkobatrachus for a few characters that are unknown in Triadobatrachus, but in all of 2005
these, Czatkobatrachus has the state expected (under all hypotheses) for the ancestral 2006
lissamphibian and the ancestral salientian, so it is not relevant here. 2007
“Several other taxa [of Mesozoic salamanders] (e.g., Laccotriton, Sinerpeton, and 2008
Jeholotriton) are excluded from this analysis because they are anatomically uncertain and are 2009
currently under taxonomic revision” (Gao & Shubin, 2012: supplementary information: 3). 2010
This revision is ongoing (Jia & Gao, 2016b). 2011
Very recently, Clack et al. (2016) described several intriguing taxa from the middle 2012
Tournaisian of southeastern Scotland. We hope to include at least some of them in future 2013
versions of the present matrix. 2014
2015
2016
46
46
Phylogeneticanalyses 2017
2018
All our analyses used maximum parsimony (see below) and were conducted in PAUP* 2019
4.0b10 (Swofford, 2003) on a computer with an Intel® Core™2 Duo processor (2.67 GHz) 2020
and 3.87 GB of usable RAM. 2021
Twelve analyses (Table 2) of our modified matrix were conducted with (Analyses R4–2022
R6 and R10–R12) and without the added taxa (Analyses R1–R3 and R7–R9), and with and 2023
without backbone constraints. The first constraint (Fig. 8A), used in Analyses R2, R5, R8 and 2024
R11, forced the dissorophoid temnospondyl Doleserpeton to be closer to the three salientians 2025
(Triadobatrachus, Notobatrachus, Vieraella; their monophyly with respect to Doleserpeton 2026
was specified, but not their relationships to each other) than the lysorophian lepospondyl 2027
Brachydectes, making the LH impossible but allowing both the TH and the PH. The second 2028
constraint (Fig. 8B), used in Analyses R3, R6 and R9, only allowed the PH by additionally 2029
forcing the gymnophionomorph Eocaecilia to be closer to Brachydectes than to Doleserpeton. 2030
Because Analysis R11 unexpectedly recovered the PH rather than the TH, we performed 2031
Analysis R12 under a third constraint (Fig. 8C), which forced Eocaecilia, the salientians and 2032
Doleserpeton to be closer to each other than to Brachydectes. This constraint was not used 2033
elsewhere. For Analyses R7–R12, unlike R1–R6, many characters were scored as irreversible 2034
(see Treatment of characters: Irreversibility). In order to find all optimal islands and all 2035
optimal trees within each island, each heuristic search used 10,000 addition-sequence 2036
replicates (with random addition sequences), each of which was restricted to 25 million 2037
rearrangements by tree bisection and reconnection. Total calculation time was about four 2038
weeks; Analysis R1 took 29:18:43, Analysis R12 lasted for 90:03:30. 2039
2040
2041
2042
Figure 8: Topological constraints used in Analyses R2, R3, R5, R6, R8, R9, R11 and 2043
R12. These are backbone constraints: the OTUs not mentioned in the constraints were free to 2044
be positioned anywhere within the constraint tree (including at its root), and the polytomies 2045
were allowed to be resolved in all mathematically possible ways. A: Constraint against the 2046
LH, allowing both TH and PH; used in Analyses R2, R5, R8 (all found the TH) and R11 2047
(found the PH). Whether Lepospondyli is extinct depends on the unconstrained positions of 2048
the caudates. B: Constraint for the PH, used in Analyses R3, R6 and R9. C: Constraint for the 2049
TH, used in Analysis R12 (after R11 unexpectedly found the PH). Lissamphibian monophyly 2050
with respect to Doleserpeton is not constrained, but was found in all MPTs. Whether 2051
Lepospondyli is extinct technically depends on the unconstrained positions of the caudates. 2052
2053
2054
Possibly because some of the stepmatrices are asymmetric, PAUP* 4.0b10 ignores the 2055
outgroup specification by default, roots the trees it finds on the OTU with the largest number 2056
of question marks, and finds enormous numbers of MPTs after a very long time. Manually 2057
rooting the trees afterwards makes many of them identical. To circumvent these problems, we 2058
included an ancestor identical to the Eusthenopteron OTU in the analyses mentioned above. 2059
We also reanalyzed (Analysis O1; Table 2) the original, unmodified matrix of RC07 2060
without any constraints to test whether their procedure for reducing calculation time (the 2061
“parsimony ratchet”; see Ruta, Coates & Quicke, 2003) had overlooked any MPTs. This was 2062
deemed important because, when Skutschas & Gubin (2012) applied the same procedure to 2063
47
47
their own matrix which was small enough for a branch-and-bound analysis, it found a single 2064
tree that was no less than 35 steps longer than the MPTs found by branch-and-bound – a 2065
method which is guaranteed to find all MPTs. (To our surprise, Skutschas & Gubin [2012] 2066
reported the length difference without commenting on it and without discarding the crassly 2067
suboptimal tree or offering a reason why it should be retained in consideration.) Calculation 2068
time with 5,000 addition-sequence replicates and up to a hundred million rearrangements per 2069
replicate was 21:48:07. 2070
2071
2072
Table 2: Overview of analysis settings and results concerning lissamphibian origins. See 2073
Table 3 for results concerning other questions. Steps were counted in PAUP* (Swofford, 2074
2003), partial uncertainty was distinguished from polymorphism. Analyses whose numbers 2075
(also used in the text and Tables 3, 4) begin with O were performed on the original matrix of 2076
RC07, those with R and B on our modified version; O and R were parsimony analyses, B 2077
were parsimony bootstrap analyses. Constrained reversibility means that losses of entire 2078
bones were coded as irreversible. 2079
2080
Reversibility Taxon
sample
Analysis Constraint Steps Result Figures
free RC07 O1 none 1621 TH 2
(102 O2 for LH1 1622 LH 9
OTUs) R1 none 2157 LH 10
R2 against LH2 2166 TH3 11
R3 for PH 2169 PH 12
B1 none 2174 LH 17
expanded R4 none 2901 LH 13
(150 R5 against LH2 2913 TH 14, 15
OTUs) R6 for PH 2918 PH 15, 16
B2 none 2948 LH 18
constrained RC07 R7 none 2168 LH 10, 19
(102 R8 against LH2 2178 TH 20
OTUs) R9 for PH 2179 PH 12, 21
B3 none 2188 LH 25
expanded R10 none 2913 LH 22
(150 R11 against LH2 2926 PH 23
OTUs) R12 for TH 2927 TH 24
B4 none 2963 LH 26
2081
1 Technically, this constraint would also have allowed the “inverse polyphyly hypothesis”, 2082
where the salientians would be lepospondyls and the gymnophionomorphs (represented by 2083
Eocaecilia) would be temnospondyls. 2084
2 This constraint is aimed at enforcing the TH, but it is also compatible with the PH, as 2085
Analysis R11 demonstrates. 2086
3 A highly unusual version of the TH, see text and Table 3. 2087
2088
2089
On p. 85, RC07 reported that constraining their analysis to find a topology fully 2090
congruent with that of Laurin (1998a) – a version of the LH – required “the addition of a mere 2091
nine steps” and yielded 60 suboptimal trees. On the next page, however, they wrote: “Using 2092
48
48
our own data, the 720 suboptimal trees placing lissamphibians with lepospondyls are 15 steps 2093
longer than the most parsimonious trees from the original analysis.” To resolve this 2094
contradiction, we reanalyzed (Analysis O2; Table 2) the original, unmodified matrix of RC07 2095
under a backbone constraint that required Brachydectes to be closer to the salientians than 2096
Doleserpeton. (RC07 did not publish the constraints they used.) The salientians are 2097
Triadobatrachus, Notobatrachus and Vieraella; their monophyly with respect to Doleserpeton 2098
was specified, but not their relationships to each other (Fig. 8A with the positions of 2099
Brachydectes and Doleserpeton exchanged). Because this constraint is unrooted, it is 2100
compatible with the MPTs, which support the TH; we therefore rooted the trees by including 2101
an ancestor identical to the Eusthenopteron OTU in the analysis. (Without this ancestor, the 2102
constraint had no effect at all.) Calculation time was 23:48:00. 2103
Maddin, Jenkins & Anderson (2012) reported the majority-rule consensus of their 2104
MPTs, but not the strict consensus, and did not explicitly discuss any MPTs that disagreed 2105
with the majority. Out of curiosity, we therefore repeated their analysis without modifying 2106
their matrix or applying any constraints. Because Maddin, Jenkins & Anderson (2012) did not 2107
report what settings they had used, we ran 1,000 unlimited addition-sequence replicates. 2108
Calculation time was 03:58:40.9. 2109
2110
Why we only used maximum parsimony 2111
2112
Parametric methods of phylogenetics – maximum likelihood and especially Bayesian 2113inference – are increasingly used on morphological datasets. They are generally less sensitive 2114
to long-branch attraction than the nonparametric method called “maximum parsimony” is, and 2115
Bayesian inference provides a measure of robusticity as part of every phylogenetic analysis. 2116
Much as in the case of Marjanović & Witzmann (2015), however, we cannot use parametric 2117
methods on our dataset for the following reasons. 2118
First, our dataset contains 19 unordered multistate characters, 38 ordered multistate 2119
characters and 4 characters with more complex stepmatrices even in Analyses R1–R6 where 2120
(see above) bone losses are not scored as irreversible. To the best of our knowledge, the 2121
available programs in which parametric methods are implemented cannot handle such a 2122
combination. This practical obstacle is, unfortunately, enough on its own to restrict us to 2123
using maximum parsimony. 2124
Second, our matrices contain large amounts of nonrandomly distributed missing data. 2125
In the studies of Simmons (2011a, b) on contrived, simulated and empirical datasets, this 2126
situation routinely led parametric methods to find strong support for wholly spurious clades – 2127
even under ideal conditions (no homoplasy, a perfectly fitting model of evolution, a perfect 2128
alignment, no rate heterogeneity). Moreover, which spurious clades were found was very 2129
sensitive to small changes in the scores of a matrix, despite the high support values. The size 2130
of a matrix did not make a difference to the performance of maximum likelihood in Simmons 2131
(2011a); Bayesian inference performed about as badly unless the dataset was increased to 2132
100,000 total and about 2500 parsimony-informative characters (we have 276). While Wright 2133
& Hillis (2014) did simulate nonrandomly distributed missing data and found them not to 2134
cause serious problems, they did so by scoring all characters that evolved at the same rate as 2135
unknown for selected taxa; this may be realistic for molecular data, where different genes may 2136
have been sequenced for different taxa in a supermatrix, but does not make sense for 2137
morphological data. (Yet, Wright & Hillis [2014] simulated only binary characters and 2138
explicitly intended to study the performance of Bayesian inference with morphological data. 2139
Simmons [2011a, b] studied DNA characters.) The most recent study, O’Reilly et al. (2016), 2140
did not mention the problem of missing data (and likewise restricted itself to binary 2141
characters). – As expected, maximum parsimony is immune to this problem: it simply 2142
49
49
represents a lack of data as a lack of resolution. Given datasets with missing data that are 2143
distributed as in his studies, Simmons (2011a, b) recommended that nodes and support values 2144
from parametric analyses in parts of the tree that are not resolved in the strict consensus of a 2145
parsimony analysis of the same dataset should only be accepted after special scrutiny of their 2146
causes. 2147
Third, for matrices with high average evolutionary rates (therefore large amounts of 2148
noise, which translates to low consistency indices of MPTs), the performance of Bayesian 2149
inference and maximum parsimony converges when the number of characters increases far 2150
enough (Wright & Hillis, 2014: fig. 6; O’Reilly et al., 2016). As shown below, the 2151
consistency indices are below 0.22 in all our analyses; some characters change states 30 to 40 2152
times on every tree. 2153
Last but not least, parametric analyses are inherently much more time-consuming than 2154
nonparametric ones; our maximum-parsimony analyses took about four weeks in total, not 2155
including the bootstrap tests which also used maximum parsimony and took another four 2156
weeks. 2157
We are looking forward to further developments of parametric analyses and further 2158
tests of their performance with paleontological datasets. In the meantime, it should be kept in 2159
mind that long-branch attraction is not unknown in morphological data; in particular, our 2160
results concerning the placement of adelospondyls, aïstopods and perhaps urocordylids may 2161
have to be taken with caution. However, for the third reason given above, we are optimistic 2162
about the performance of maximum parsimony on datasets like ours – as long as such datasets 2163
are not loaded with redundant characters and misscores. 2164
2165
Robustness analyses 2166
2167
We performed four bootstrap analyses on our modified matrix (Table 2), one each for the 2168
original (Analyses B1, B3) and for the expanded taxon sample (Analyses B2, B4) with bone 2169
losses coded as reversible (Analyses B1, B2) or irreversible (Analyses B3, B4). All used 2170
1,000 bootstrap replicates consisting of 200 addition-sequence replicates which were limited 2171
to 2.5 million rearrangements each; Analysis B4 took 175:29:41. To maximize the amount of 2172
visible information, we included not only clades with bootstrap values of 0.5 and above in the 2173
bootstrap trees, but also clades with lower bootstrap values that do not contradict any of those 2174
with higher ones (PAUP* setting: bootstrap keepall=yes). 2175
In order to test whether the differences between MPTs from different analyses with the 2176
same taxon sample and the same reversibility coding are statistically significant, we ran the 2177
tests implemented in PAUP*: a Kishino-Hasegawa test, a Templeton test and a winning-sites 2178
test. The p value we report for each is the probability (Kishino-Hasegawa) or approximate 2179
probability (Templeton, winning sites) of finding a more extreme test statistic under the null 2180
hypothesis that the differences between the two tested trees are indistinguishable from 2181
random. The MPTs used for these tests are part of the data files (Data S1, S2) and were 2182
chosen to differ as little as possible from each other topologically (given the same taxon 2183
sample). 2184
Eduardo Ascarrunz (pers. comm.) and the reviewer Michael Buchwitz have kindly 2185
notified us of problems with our previous (2015) applications of these tests. Goldman, 2186
Anderson & Rodrigo (2000) and Planet (2006) have shown that all three tests, as 2187
implemented in PAUP* and elsewhere, are inappropriate to use for our questions and indeed 2188
for the vast majority of the questions they have been applied to in the phylogenetics literature. 2189
To the best of our knowledge, better tests are not available for practical purposes. Thus, we 2190
cannot reject the abovementioned null hypothesis in any case. The best we can do is to look at 2191
the problem from the other side and distinguish the cases where the null hypothesis can 2192
50
50
certainly not be rejected from those where it remains rejectable. This is possible because the p 2193
values found by better tests are always greater than the p values found by the inappropriately 2194
used one-tailed tests – greater by an amount that cannot be estimated in advance (Goldman, 2195
Anderson & Rodrigo, 2000). PAUP* performs two-tailed versions of all three tests; we report 2196
p values from the one-tailed versions, derived by halving the p values put out by PAUP* 2197
(Goldman, Anderson & Rodrigo, 2000; Planet, 2006). 2198
2199
Results 2200
2201
A very brief overview and comparison of our results is presented by Table 3; the results of 2202
statistical tests for lack of distinguishability of topologies are reported below and shown in 2203
Table 4. 2204
2205
Reanalysis of the matrix of Maddin, Jenkins & Anderson (2012) 2206
2207
Maddin, Jenkins & Anderson (2012) reported having found 34 MPTs of 1450 steps, and 2208
stated that 274 of the 308 characters in their matrix were parsimony-informative. We find that 2209
274 is instead the number of MPTs; PAUP* reports that 280 of the 308 characters are 2210
parsimony-informative.The MPTs have 1449 steps, and their indices, too, are slightly better 2211
than reported (consistency index = 0.3527, homoplasy index = 0.7191, retention index = 2212
0.6544, rescaled consistency index = 0.2308); the consistency and homoplasy indices 2213
excluding parsimony-uninformative characters, not reported by Maddin, Jenkins & Anderson 2214
(2012), are 0.3252 and 0.6748. The number of trees that are 1450 steps long, incidentally, is 2215
not 34, but 159,650. 2216
However, the identical strict and Adams consensus we find is practically identical to 2217
the majority-rule consensus of Maddin, Jenkins & Anderson (2012: fig. 4). The only 2218
contradictions are the following: Utaherpeton emerges as the sister-group to all other 2219
lepospondyls rather than to Holospondyli (a clade that includes Brachydectes, the aïstopods 2220
and the nectrideans); Microbrachis and Adelogyrinus are not sister-groups forming the sister-2221
group to the clade that contains all remaining “microsaurs” except Utaherpeton, but are 2222
successively closer to Holospondyli, within which the “nectrideans” no longer form a clade 2223
and Scincosaurus is the sister-group to (Urocordylidae (Diplocaulidae (Brachydectes, 2224
Aïstopoda))) rather than to Diplocaulidae alone; within Diplocaulidae, Diceratosaurus is 2225
closer to (Diplocaulus + Diploceraspis) than to (Keraterpeton + Batrachiderpeton). All other 2226
differences concern lack of resolution in our strict consensus: Tuditanus and Asaphestera 2227
form a trichotomy with the clade formed by all “microsaurs” not mentioned above; Pantylus, 2228
Stegotretus, Gymnarthridae, Ostodolepididae and (Rhynchonkos + Batropetes) form a 2229
polytomy; Temnospondyli has a basal polytomy between Eryops, (Balanerpeton + 2230
Dendrerpeton), Trematopidae (itself unresolved) and the clade of the remaining 2231
dissorophoids, and finally Greererpeton forms a trichotomy with Temnospondyli and the rest 2232
of the ingroup. The position of Gerobatrachus within Lissamphibia is not affected. 2233
The majority-rule consensus of our trees resolves these polytomies identically to 2234
Maddin, Jenkins & Anderson (2012: fig. 4), except that the polytomy encompassing most 2235
“microsaurs” remains unresolved. 2236
2237
2238
2239
51
51
2240
Table 3: Summary and comparison of analysis results. See Table 2 for the settings of these analyses and for the numbers of the figures that 2241
represent the trees. The “modern amphibians” column gives the closest one or two relatives of the clade of modern amphibians, or the two clades 2242
of modern amphibians given the PH; the Batrachia column indicates whether Caudata and Salientia form a clade to the exclusion of Eocaecilia; 2243
the Kotlassia column states whether Kotlassia is crownward of, rootward of or a member of Seymouriamorpha; the Limnoscelis column states 2244
whether Limnoscelis is crownward of, rootward of or a member of Diadectomorpha; in the adelospondyl column, “Colosteidae” without further 2245
comment means that a clade of colosteids and adelospondyls lies rootward of Baphetoidea, well outside of Temno- and Lepospondyli; the last 2246
column states whether Ichthyostega is rootward or crownward of Acanthostega and Ventastega. Note that whenever Amniota is closer to 2247
Temnospondyli than to CAS, the Caerorhachis OTU is found as a temnospondyl, except in some trees of Analyses R8 and R10–R12. 2248
Brachystelechidae contains Batropetes, *Carrolla and *Quasicaecilia. OTUs unique to the expanded taxon sample are marked with an asterisk. 2249
Abbreviations: #, number of the analysis (as used in the text and in Tables 2, 4); CAS, Caerorhachis OTU (including Casineria), Anthracosauria 2250
and Silvanerpeton; Diad., Diadectomorpha; Seym., Seymouriamorpha; T, Temnospondyli; (U, A), a clade formed exclusively by Urocordylidae 2251
and Aïstopoda. Numbers in curly braces are bootstrap percentages, shown in boldface if 50 or higher. 2252
2253
# Modern amphibians Batrachia Albanerpetidae
closest to:
Amniota
closer to:
Kotlassia Limnoscelis adelospondyls
closest to:
Ichthyostega
O1 TH: Doleserpeton yes Eocaecilia CAS crownward rootward Colosteidae crownward
O2 LH: Holospondyli yes Lissamphibia CAS crownward rootward Colosteidae crownward
R1 LH: Brachydectes yes 1) Lissamphibia;
2) Batrachia
T 1) rootward;
2) Seym.
Diad. (U, A) rootward
R2 TH: Phlegethontia1 yes Lissamphibia T 1) rootward;
2) Seym.
Diad. 1 rootward
R3 PH: Doleserpeton;
Brachydectes2
no2 1) Procera; 2)
Caudata
T 1) rootward;
2) Seym.
Diad. (U, A) rootward
B1 LH {80}3:
(Brachydectes +
Batropetes) {18}4
yes {68} Batrachia {45} T {17} Seym. {26} Diad. {59} Solenodonsaurus
+ crownward {38;
32}5
rootward {67}
R4 LH: (Brachydectes +
Brachystelechidae)
yes Batrachia CAS Seym. crownward 1) (U, A); 2) (U,
A) +
*Utaherpeton
rootward
R5 TH: Doleserpeton or
*Gerobatrachus
no Eocaecilia T Seym. unresolved (U, A) rootward
52
52
R6 PH: Doleserpeton;
(adelospondyls (U,
A))6
yes6 Eocaecilia 1) CAS; 2)
T
1) rootward;
2) Seym.
crownward (U, A) rootward
B2 LH {253, 267}:
(Brachydectes +
Brachystelechidae)
{17}4
yes {49} Lissamphibia
{64}
T {8} Seym. {29} rootward
{27}
(U, A) +
*Utaherpeton{27;
16}5
rootward of
Ventastega
{11} and
Acanthostega
{13}
R7 LH: Brachydectes yes Batrachia T 1) rootward;
2) Seym.
Diad. (U, A) rootward
R8 TH: Doleserpeton yes 1) Batrachia; 2)
Caudata; 3)
Valdotriton; 4)
Salientia
T 1) rootward;
2) Seym.
Diad. (U, A) rootward
R9 PH: Doleserpeton;
Brachydectes2
no2 Caudata T 1) rootward;
2) Seym.
Diad. (U, A) rootward
B3 LH {763}:
(Brachydectes +
Batropetes) {28}4
yes {64} Batrachia {52} T {17} Seym. {25} Diad. {71} Solenodonsaurus
+ crownward {37,
21}5
rootward {65}
R10 LH: (Brachydectes +
Brachystelechidae)
yes Batrachia T Seym. crownward 1) (U, A); 2)
*Utaherpeton
rootward
R11 PH: Doleserpeton or
*Gerobatrachus;
(Brachydectes +
Brachystelechidae)9
yes9 Eocaecilia T Seym. crownward 1) (U, A); 2)
*Utaherpeton
rootward
R12 TH: Doleserpeton or
*Gerobatrachus
yes 1) Valdotriton; 2)
(Valdotriton +
Pangerpeton)
T Seym. crownward 1) (U, A); 2)
*Utaherpeton
rootward
B4 LH {233, 7}:
(Brachydectes +
Brachystelechidae)
{18}4
yes {53} Batrachia {45} T {10} Seym. {26} Diad. {31} *Utaherpeton
{25; 11}5
rootward {11}
53
53
1 An entirely novel arrangement where Lissamphibia and the strangest “lepospondyls” are nested among the dvinosaurian temnospondyls, closest 2254
to Isodectes; see text. 2255
2 Salientia (as constrained) is nested within Temnospondyli (next to Doleserpeton), Eocaecilia (as constrained), Caudata and Albanerpetidae 2256
within Lepospondyli. 2257
3 Dissorophoidea. 2258
4 18 (B1, B4), 17 (B2) and 28 (B3) are the bootstrap percentages of the smallest clade that contains Batropetes, Brachydectes and Lissamphibia; 2259
(Batropetes + Brachydectes) respectively (Brachystelechidae + Brachydectes) have 39 (B1), 17 (B2), 42 (B3) or 20 (B4). 2260
5 The highest bootstrap percentage that keeps the adelospondyls away from Colosteidae is 38 (B1), 27 (B2), 37 (B3) or 25 (B4), the highest that 2261
keeps them away from (U, A) is 32 (B1), 16 (B2), 21 (B3) or 11 (B4). 2262
6 Salientia (as constrained) and Caudata are nested within Temnospondyli (next to Doleserpeton or *Gerobatrachus), Eocaecilia (as constrained) 2263
and Albanerpetidae within Lepospondyli. 2264
7 Amphibamidae (not including *Micropholis). 2265
8 The highest boostrap percentage that keeps the adelospondyls away from Colosteidae is 27. 2266
9 Salientia (as constrained) and Caudata are nested within Temnospondyli (next to Doleserpeton or *Gerobatrachus), Eocaecilia and 2267
Albanerpetidae within Lepospondyli. 2268
2269
2270
2271
2272
↓ Figure 9: Strict consensus of the MPTs (length: 1622 steps including polymorphisms) found by our constrained reanalysis of the 2273
unchangedmatrix of RC07 (Analysis O2). The constraint forced the lysorophian “lepospondyl” Brachydectes to be closer to the three 2274
salientians (Triadobatrachus, Notobatrachus, Vieraella) than the dissorophoid temnospondyl Doleserpeton; this allowed only the lepospondyl 2275
hypothesis (Fig. 8A with the positions of Brachydectes and Doleserpeton exchanged). 2276
2277
54
54
Eusthenopteron
Panderichthys
Ventastega
Acanthostega
Tulerpeton
Colosteus
Greererpeton
Baphetes
Megalocephalus
Eucritta
Edops
Chenoprosopus
Cochleosaurus
Eryops
Capetus
Balanerpeton
Dendrerpetidae
Acheloma
Phonerpeton
Ecolsonia
Broiliellus brevis
Amphibamus
Doleserpeton
Albanerpetidae
Karaurus
Valdotriton
Triadobatrachus
Notobatrachus
Vieraella
Eoscopus
Platyrhinops
Micromelerpeton
Apateon
Leptorophus
Schoenfelderpeton
Isodectes
Trimerorhachis
Neldasaurus
Caerorhachis
Eocaecilia
Acherontiscus
Adelospondylus
Adelogyrinus
Dolichopareias
Ptyonius
Urocordylus
Sauropleura
Lethiscus
Oestocephalus
Phlegethontia
Batropetes
Brachydectes
Scincosaurus
Keraterpeton
Diceratosaurus
Batrachiderpeton
Diplocaulus
Diploceraspis
Tuditanus
Asaphestera
Pantylus
Stegotretus
Saxonerpeton
Hapsidopareion
Rhynchonkos/Aletrimyti/Dvellecanus
Micraroter
Pelodosotis
Cardiocephalus
Euryodus
Odonterpeton
Microbrachis
Hyloplesion
Westlothiana
Diadectes
Orobates
Tseajaia
Limnoscelis
Captorhinus
Paleothyris/Protorothyris
Petrolacosaurus
Kotlassia
Discosauriscus
Ariekanerpeton
Leptoropha
Microphon
Utegenia
Seymouria
Solenodonsaurus
Bruktererpeton
Gephyrostegus
Silvanerpeton
Eoherpeton
Proterogyrinus
Archeria
Pholiderpeton scutigerum
Pholiderpeton attheyi
Anthracosaurus
Crassigyrinus
Whatcheeria
Pederpes
Ossinodus
Ichthyostega
Whatcheeriidae
Baphetoidea
Anthracosauria
tetrapod crown-group
Amphibia
BatrachiaLiss-
amphibia
Pantylidae
Aïstopoda
Urocordylidae
Diplocaulidae
Salientia
Caudata
Gymnarthridae
Ostodolepididae
Microbrachomorpha
Sauropsida
“diadectomorphs”
“microsaurs”
“seymouriamorphs”
“gephyrostegids”
“amphibamids”
Colosteidae
adelospondyls Adelogyrinidae
Branchiosauridae
Gymnophionomorpha
Lysorophia
Holospondyli
Embolomeri
Temnospondyli
Eutemnospondyli
Edopoidea
Dvinosauria
Dissorophoidea (S13)
Dissorophoidea (YW00)
Trematopidae
modern
amphibians
origin of digits
Table 4: Summary of support for hypotheses on lissamphibian origins. See Table 2 for 2278
analysis settings. The second column subtracts the MPT length found by the second analysis 2279
in each line from that found by the first. The rightmost column summarizes the p values from 2280
the one-tailed Kishino-Hasegawa, Templeton and winning-sites tests (listed in the text; our 2281
use of the tests is discussed in Materials and methods: Robustness analyses). Bootstrap values 2282
of 0.5 or higher are given in boldface, as are p values that do not certainly fail to reject the 2283
null hypothesis at the (arbitrary) 0.05 level; the null hypothesis is that the topology 2284
differences cannot be distinguished from random variations. Note that a cutoff value of 0.1 2285
instead of 0.05 would have given the same results, except that the winning-sites test 2286
distinguishes R7 and R8 at a minimum of p = 0.0975. Conversely, a value of 0.04 would have 2287
rejected the comparison of R7 and R9 according to the Kishino-Hasegawa (p = 0.0468) and 2288
Templeton (p = 0.0480) tests as well as that of R10 and R11 by the Templeton test (p = 2289
0.0411). 2290
2291
Analyses Difference
in number
of steps
Highest bootstrap
support for difference in
topology
Difference in topology
indistinguishable from
random at p = 0.05?
O1 (TH), O2 (LH) 1 not tested certainly (p > 0.44)
R1 (LH), R2 (TH) 9 0.80 (Dissorophoidea) certainly (0.31 > p >
0.062)
R1 (LH), R3 (PH) 12 0.80 (Dissorophoidea) not certainly (0.033 > p >
0.027)
R2 (TH), R3 (PH) 3 0.68 (Batrachia) certainly (p > 0.32)
R4 (LH), R5 (TH) 12 0.40 (Doleserpeton +
*Gerobatrachus)
certainly (0.273 > p >
0.172)
R4 (LH), R6 (PH) 17 0.64 (modern amphibians) not certainly (0.016 > p >
0.014)
R5 (TH), R6 (PH) 5 0.64 (modern amphibians) certainly (p > 0.39)
R7 (LH), R8 (TH) 10 0.76 (Dissorophoidea) certainly (0.188 > p >
0.097)
R7 (LH), R9 (PH) 11 0.76 (Dissorophoidea) not certainly (0.048 > p >
0.024)
R8 (TH), R9 (PH) 1 0.64 (Batrachia) certainly (0.47 > p > 0.32)
R10 (LH), R11 (PH) 13 0.51 (Lissamphibia) not certainly (0.042 > p >
0.027)
R10 (LH), R12
(TH)
14 0.35 (Doleserpeton +
*Gerobatrachus)
certainly (0.157 > p >
0.144)
R11 (PH), R12 (TH) 1 0.51 (Lissamphibia) no (p > 0.39)
2292
Reanalyses of the matrix of Ruta & Coates (2007) 2293
2294
Our unconstrained reanalysis of the original, unmodified matrix (Analysis O1) found 324 2295
MPTs, as RC07 did; their strict consensus (Fig. 2) is identical to that reported by RC07 (fig. 2296
5, 6). Their length is 1621 steps when we distinguish polymorphism from partial uncertainty 2297
(both occur in the matrix; PAUP* setting: pset mstaxa=variable), but 1584 steps – as reported 2298
by RC07 – when we treat both as partial uncertainty (pset mstaxa=uncertain). The indices we 2299
find are slightly different from those reported by RC07, and differ again depending on the 2300
treatment of polymorphism: when polymorphism is treated as uncertainty, the consistency 2301
index excluding the parsimony-uninformative characters is 0.2281 rather than 0.22, the 2302
retention index is 0.6768 rather than 0.67 and the rescaled consistency index is 0.1564 rather 2303
56
56
than 0.15; when polymorphism and uncertainty are distinguished, the retention index is 2304
unaffected, but the consistency index excluding the parsimony-uninformative characters rises 2305
to 0.2458 and the rescaled consistency index becomes 0.1683. 2306
Our analysis constrained for the LH (Analysis O2) found 60 most parsimonious trees. 2307
Their strict consensus (Fig. 9) is quite unlike what RC07 reported on p. 86, and also unlike 2308
that of Laurin (1998a) which they used as a constraint on p. 85: we find Lissamphibia as the 2309
sister-group of Albanerpetidae, both of which together form the sister-group to Holospondyli 2310
(“nectrideans” including aïstopods). All of these together are found as the sister-group of 2311
Brachydectes; in other words, lepospondyl topology is identical to that in the shortest trees, 2312
except for the addition of Lissamphibia + Albanerpetidae next to Holospondyli, and except 2313
for the loss of resolution in Urocordylidae. The resolution of Temnospondyli is greatly 2314
improved. The length of these suboptimal trees is 1622 steps (1585 when polymorphism is 2315
treated as uncertainty); consistency index excluding the parsimony-uninformative characters 2316
= 0.2280 (polymorphism treated as uncertainty) or 0.2457 (polymorphism and uncertainty 2317
distinguished), retention index = 0.6766, rescaled consistency index = 0.1562 or 0.1681. 2318
In short, the matrix of RC07 supports the temnospondyl hypothesis over the 2319
lepospondyl hypothesis by one single step, rather than either nine or fifteen steps as originally 2320
claimed (RC07: 85, 86). 2321
To test this result, we manually drew a most parsimonious rooted tree and a rooted tree 2322
that fulfills the constraint in Mesquite (Maddison & Maddison, 2015), which does not take 2323
ancestors into consideration. Mesquite reported the exact same lengths (distinguishing 2324
polymorphism from partial uncertainty, as Mesquite always does). 2325
This result is perhaps not as surprising as it may seem. Although they only published 2326
the values for seven nodes, RC07 (appendix 4 – p. 122) did conduct a bootstrap and a decay 2327
analysis. While Lissamphibia (bootstrap percentage = 67, Bremer index = 8) and most of the 2328
nodes within it are well supported, the sister-group relationship of Lissamphibia and 2329
Doleserpeton has a Bremer index of 1 and “no boots[t]rap support compatible witha 50% 2330
majority-rule consensus”, and the sister-group relationship of both of these to Amphibamus 2331
has a Bremer index of 2 and “no bootstrap support in a 50% majority-rule consensus”. Values 2332
for more rootward nodes were not reported, but are not likely to be any higher given the 2333
limited resolution of this part of the tree (Fig. 2). 2334
Unsurprisingly, unconstrained and constrained trees are statistically indistinguishable: 2335
p = 0.4500 (Kishino-Hasegawa and Templeton tests), p = 0.5000 (winning-sites test). 2336
Data S3 contains the unmodified matrix, the constraint, our other analysis settings, a 2337
most parsimonious tree (1621 steps) and a tree that fulfills the constraint (1622 steps). 2338
Executing it in PAUP* repeats our constrained analysis. 2339
2340
Analyses of our modified matrix 2341
2342
Note that the tree lengths we report throughout are those given by PAUP*. Mesquite 2343
consistently reports one fewer step for trees from Analyses R1–R3 (original taxon sample, 2344
free reversibility; see Table 2) and two to three fewer steps for trees from Analyses R4–R6 2345
(expanded taxon sample, free reversibility); apparently, the programs handle the complex 2346
stepmatrices differently. (As mentioned, PAUP* and Mesquite report the same lengths for the 2347
MPTs of Analyses O1 and O2.) Mesquite cannot optimize irreversible characters at all and 2348
therefore ignores them altogether when calculating the length of trees from Analyses R7–R12, 2349
resulting in much too low values. 2350
2351
2352
57
57
Eusthenopteron
Panderichthys
Ventastega
Acanthostega
Tulerpeton
Colosteus
Greererpeton
Baphetes
Megalocephalus
Eucritta
Edops
Chenoprosopus
Cochleosaurus
Eryops
Capetus
Balanerpeton
Dendrerpetidae
Acheloma
Phonerpeton
Ecolsonia
Broiliellus brevis
Amphibamus
Doleserpeton
Eoscopus
Platyrhinops
Micromelerpeton
Apateon
Leptorophus
Schoenfelderpeton
Isodectes
Trimerorhachis
Neldasaurus
Caerorhachis
Albanerpetidae
Karaurus
Valdotriton
Triadobatrachus
Notobatrachus
Vieraella
Eocaecilia
Batropetes
Brachydectes
Scincosaurus
Keraterpeton
Diceratosaurus
Batrachiderpeton
Diplocaulus
Diploceraspis
Ptyonius
Sauropleura
Urocordylus
Lethiscus
Oestocephalus
Phlegethontia
Pantylus
Stegotretus
Tuditanus
Asaphestera
Saxonerpeton
Hapsidopareion/Llistrofus
Micraroter
Pelodosotis
Rhynchonkos
Cardiocephalus
Euryodus
Microbrachis
Hyloplesion
Odonterpeton
Westlothiana
Diadectes
Orobates
Tseajaia
Limnoscelis
Captorhinus
Paleothyris
Petrolacosaurus
Kotlassia
Discosauriscus
Ariekanerpeton
Leptoropha
Microphon
Utegenia
Seymouria
Solenodonsaurus
Bruktererpeton
Gephyrostegus
Silvanerpeton
Eoherpeton
Proterogyrinus
Archeria
Pholiderpeton scutigerum
Pholiderpeton attheyi
Anthracosaurus
Acherontiscus
Adelospondylus
Adelogyrinus
Dolichopareias
Crassigyrinus
Whatcheeria
Pederpes
Ossinodus
Ichthyostega
Batrachia
Pantylidae
Tuditanidae
Salientia
Caudata
Gymnarthridae
Ostodolepididae
Sauropsida
Diadectomorpha
“microsaurs” 1
“microsaur” 2
adelospondyls
Holospondyli
“microbrachomorph” 2
“microbrachomorphs” 1
Seymouriamorpha
Gymnophionomorpha
Lysorophia
Lissamphibia
Aïstopoda
Urocordylidae
Diplocaulidae
Adelogyrinidae
tetrapod crown-group
Amphibia
Temnospondyli
Whatcheeriidae
Baphetoidea
Anthracosauria
Embolomeri
Colosteidae
Dvinosauria
Branchiosauridae
Edopoidea
Gephyrostegidae
Amphibamidae
Dissorophoidea (S13)
Dissorophoidea
(YW00)
Trematopidae
origin of digits
↑ Figure 10: Representation of the MPTs (length: 2157 steps) from Analysis R1, 2353
performed on our modified matrix. The taxon sample was unchanged from RC07, bone 2354
losses were coded as reversible; no constraint was enforced. In this figure and all following 2355
ones, Caerorhachis includes Casineria (see text). 2356
This tree, like those in the following figures, aims to represent all of the information 2357
contained in the MPTs, more than any consensus tree can. The uninterrupted part of the tree 2358
(black lines), polytomies excepted, is present in all MPTs as a backbone on which the 2359
branches connected only by colored underlays have varying positions; all nodes underlain in 2360
color are absent from the strict consensus. The cyan underlay connects the three positions of 2361
Tulerpeton that occur in different MPTs (one node more rootward than Whatcheeriidae, one 2362
node more crownward than Whatcheeriidae, or sister to Crassigyrinus). 2363
In this and the following figures, branches with only two neighboring positions in 2364
different MPTs are indicated with blue lines: for example, the clade of all seymouriamorphs 2365
except Kotlassia is the sister-group of Kotlassia or of the tetrapod crown-group. Trichotomies 2366
that are resolved in all three possible ways in different MPTs (none in this figure) are shown 2367
in the usual way (without a blue line). 2368
2369
2370
Analysis R1 2371
2372
The unconstrained analysis of our presumably improved matrix without added taxa and 2373
without irreversible characters (except limb loss) yielded nine MPTs with a length of 2157 2374
steps (a drastic increase over the 1621 steps required by the unmodified matrix of RC07), a 2375
consistency index of 0.2109 (likewise revealing increased character conflict), a homoplasy 2376
index of 0.8336, a retention index of 0.6298, and a rescaled consistency index of 0.1328. 2377
Figure 10 presents the similarities and differences between the MPTs, based on 2378
inspection of all of them in Mesquite. It shows that, in all MPTs, Panderichthys is the most 2379
rootward member of the ingroup as in RC07, followed crownwards by Ichthyostega, then 2380
Ventastega, and only then Acanthostega. This rootward position of Ichthyostega compared to 2381
more traditional ideas (e.g. Coates, 1996; RC07) appears to be mostly or entirely due to new 2382
information on Ichthyostega (Clack, 2007; Callier, Clack & Ahlberg, 2009; Ahlberg, 2011; 2383
Pierce, Clack & Hutchinson, 2012; Pierce et al., 2013) and has been considered a strong 2384
possibility for several years (e.g. Ahlberg et al., 2008: supplementary information; Clack et 2385
al., 2012a). 2386
In different MPTs, the next more crownward branch is either the Devonian Tulerpeton 2387
or the Mississippian Whatcheeriidae. Contrary to the findings of Clack & Klembara (2009) 2388
and Ruta (2009), these two taxa never fall out as sister-groups; making Tulerpeton the sister-2389
group of either Whatcheeriidae or Crassigyrinus (as in Marjanović & Laurin, 2015) requires 2390
one extra step in Mesquite. 2391
Crownward of these is a Hennigian comb of – from rootward to crownward – 2392
Crassigyrinus, Colosteidae, Baphetoidea, Anthracosauria, Silvanerpeton, Gephyrostegidae, 2393
Temnospondyli, Solenodonsaurus and Kotlassia (Fig. 10). Much of this arrangement 2394
contradicts the findings of RC07 (Fig. 2); the position of Solenodonsaurus is altogether novel. 2395
Whatcheeriidae is resolved as (Ossinodus (Whatcheeria, Pederpes)), unchanged from 2396
RC07. Baphetoidea is (Eucritta (Baphetes, Megalocephalus)) – we thus contribute to the 2397
recent consensus (RC07; Ruta, 2009; Milner, Milner & Walsh, 2009) that Eucritta is a non-2398
baphetid baphetoid. Except for the departure of Silvanerpeton, Anthracosauria is unchanged 2399
from RC07. A clade of Gephyrostegus and Bruktererpeton is new compared to RC07, but has 2400
been found elsewhere (e.g. Klembara et al., 2014). 2401
59
59
The sister-group to all other temnospondyls is Caerorhachis, replicating the result 2402
found by Pawley (2006 – separately for both Caerorhachis and Casineria) and suggested 2403
earlier by Godfrey, Fiorillo & Carroll (1987). The clade of the remaining temnospondyls (the 2404
four-fingered ones) consists of Dvinosauria – resolved, unlike in RC07, as (Neldasaurus 2405
(Trimerorhachis, Isodectes)) – and a Hennigian comb in which Dissorophoidea, 2406
(Balanerpeton + Dendrerpetidae),Capetus and Eryops are successively closer to Edopoidea 2407
(which is unchanged from RC07); this is a novel result, though similar to that of Ruta (2009). 2408
Within Dissorophoidea, Ecolsonia emerges as a trematopid as in several recent 2409
analyses (Ruta, 2009; Maddin et al., 2013b; Holmes, Berman & Anderson, 2013; Dilkes, 2410
2015a). Trematopidae remains the sister-group of the other dissorophoids, among which 2411
Amphibamidae is the sister-group of Broiliellus to the exclusion of a “branchiosaur” clade 2412
(Micromelerpeton as the sister-group of Branchiosauridae, resolved as in RC07). 2413
Amphibamidae is resolved as (Platyrhinops (Eoscopus (Amphibamus, Doleserpeton))). 2414
The seymouriamorphs other than Kotlassia form a fully resolved clade in all MPTs; 2415
Seymouria is found as the sister-group of Discosauriscidae, within which Utegenia is the 2416
sister-group to (Leptoropha + Microphon), unlike in RC07. In different MPTs, this whole 2417
clade forms the sister-group of either Kotlassia or the tetrapod crown-group. 2418
The tetrapod crown-group consists of Amphibia (Lissamphibia nested among the 2419
“lepospondyls”) and the sister-groups Diadectomorpha and Sauropsida. Sauropsida is 2420
resolved as in RC07; Diadectomorpha is resolved as (Limnoscelis (Tseajaia (Orobates, 2421
Diadectes))) as in Reisz (2007), Kissel (2010) and Berman (2013) and further supported by 2422
Berman, Reisz & Scott (2010). 2423
Westlothiana is the most rootward amphibian, unchanged from RC07, even though we 2424
have changed many of its scores. Much like in RC07, it is followed crownward by two 2425
“microsaur” clades, which comprise all “microsaurs” in the taxon sample except Batropetes; 2426
the composition and especially the topologies of these clades are quite different from those 2427
found by RC07, however. The larger and more rootward clade has a basal split into 2428
(Microbrachis + Hyloplesion) on one side and (Odonterpeton (Saxonerpeton (Gymnarthridae 2429
(Ostodolepididae (Rhynchonkos, Hapsidopareion/Llistrofus))))) on the other; the smaller and 2430
more crownward one is composed of Pantylidae (as in RC07) and Tuditanidae (Tuditanus + 2431
Asaphestera). 2432
The next more crownward clade contains, one could say, all the strangest 2433
“lepospondyls”: the adelospondyls form the sister-group to Urocordylidae + Aïstopoda. 2434
Unlike in RC07, Adelospondylus is resolved as the sister-group to the other adelogyrinids, 2435
Urocordylus as the sister-group to the other urocordylids and Lethiscus as the sister-group of 2436
the other aïstopods. To move the adelospondyls out of the “lepospondyls” and next to the 2437
colosteids, as in RC07 (but not Ruta, Coates & Quicke, 2003), requires two extra steps in 2438
Mesquite. 2439
Next more crownward lie the sister-groups Scincosaurus and Diplocaulidae. Their 2440
relationship is reminiscent of Milner & Ruta (2009) or Ruta, Coates & Quicke (2003). 2441
Diplocaulidae is resolved as in RC07 or Germain (2010), except that Keraterpeton and 2442
Diceratosaurus are sister-groups. Two extra steps are needed in Mesquite to make 2443
(Scincosaurus + Diplocaulidae) the sister-group of (adelospondyls (Urocordylidae, 2444
Aïstopoda)); two more are required to make the latter clade the sister-group of Diplocaulidae 2445
alone as in RC07. 2446
The mentioned brachystelechid “microsaur” Batropetes and the lysorophian 2447
Brachydectes lie successively closer to the modern amphibians, as in Vallin & Laurin (2004); 2448
two extra steps are needed in Mesquite to render them each other’s sister-group, three steps 2449
are required to move Batropetes closer to Lissamphibia than Brachydectes. Lissamphibia is 2450
resolved as (Eocaecilia (Caudata, Salientia)), unlike in RC07 or Vallin & Laurin (2004) but 2451
60
60
like in some other recent works; Albanerpetidae is the sister-group of either Lissamphibia or 2452
of Batrachia, both unlike in RC07 – see the Discussion section and Marjanović & Laurin 2453
(2013a). 2454
This analysis thus supports the LH: Lissamphibia is closer to the “lepospondyls” 2455
Batropetes and Brachydectes than to any temnospondyl; within Temnospondyli, 2456
Amphibamidae is monophyletic and highly nested. 2457
Among the “microsaurs”, neither Tuditanomorpha nor Microbrachomorpha of the 2458
classification by CG78 are found as monophyletic, and the name Recumbirostra Anderson, 2459
2007b, cannot be applied (see Discussion). Still, the “microbrachomorphs” other than the 2460
tentatively (by CG87) referred Batropetes form a paraphyletic series: Odonterpeton is only 2461
one internode away from (Hyloplesion + Microbrachis), and two extra steps in Mesquite 2462
allow these two clades to be sister-groups.. 2463
2464
Analysis R2 2465
2466
Constraining the analysis of the original taxon sample against the lepospondyl hypothesis 2467
yielded 261 MPTs with a length of 2166 steps – 9 more than in the unconstrained Analysis R1 2468
– and a consistency index of 0.2101, a homoplasy index of 0.8343, a retention index of 2469
0.6278, and a rescaled consistency index of 0.1319. Figure 11 presents the similarities and 2470
differences between the MPTs, based on comparison of the strict consensus tree to the first, 2471
the last, and every tenth MPT (1, 10, 20 … 250, 260, 261) in Mesquite. All MPTs have a 2472
bundle of highly unusual features. 2473
Specifically, when the salientians move closer to Doleserpeton than to Brachydectes, 2474
they drag not merely the other lissamphibians along, but also the aïstopods, urocordylids and 2475
adelospondyls; and this assemblage nests not within the amphibamids or the dissorophoids 2476
more generally, but within the dvinosaurs. The modern amphibians form the sister-group to 2477
the aïstopod Phlegethontia; all together are joined on the outside by a clade formed by the 2478
remaining two aïstopods (Lethiscus, Oestocephalus), then Urocordylidae (where Ptyonius is 2479
resolved as the sister-group to the rest), then (Acherontiscus (Adelospondylus (Adelogyrinus, 2480
Dolichopareias))), and then the dvinosaurian temnospondyl Isodectes. The clade that contains 2481
all non-dvinosaurian temnospondyls except Caerorhachis is still recovered as in Analysis R1, 2482
but now has three different resolutions: some of the MPTs (inset in Fig. 11) are entirely 2483
compatible with the strict consensus of the results of RC07 (fig. 5; our Fig. 2), others (also 2484
inset in Fig. 11) nearly so, while yet others (main tree in Fig. 11) replicate the topology from 2485
Analysis R1. In some trees, the (Microbrachis + Hyloplesion) clade falls out as the sister-2486
group to all lepospondyls except Westlothiana. 2487
2488
2489
2490
↓ Figure 11: Representation of the MPTs (length: 2166 steps) from Analysis R2, 2491
performed on our modified matrix. The taxon sample was unchanged from RC07, bone 2492
losses were coded as reversible. A constraint forced the three salientians (Triadobatrachus, 2493
Notobatrachus, Vieraella) to be closer to the dissorophoid temnospondyl Doleserpeton than 2494
to the lysorophian “lepospondyl” Brachydectes; this allowed both the temnospondyl and the 2495
polyphyly hypotheses. The inset at the lower left shows the two equally parsimonious 2496
alternatives to the temnospondyl topology shown in the main tree. 2497
Cyan underlay as in Fig. 10; dark green underlay for the two positions of Edops in the 2498
inset. 2499
2500
61
61
Eusthenopteron
Panderichthys
Ventastega
Acanthostega
Tulerpeton
Colosteus
Greererpeton
Baphetes
Megalocephalus
Eucritta
Edops
Chenoprosopus
Cochleosaurus
Eryops
Capetus
– see below for
alternatives –
Balanerpeton
Dendrerpetidae
Acheloma
Phonerpeton
Ecolsonia
Broiliellus brevis
Amphibamus
Doleserpeton
Eoscopus
Platyrhinops
Micromelerpeton
Apateon
Leptorophus
Schoenfelderpeton
Edops
Chenoprosopus
Cochleosaurus
Eryops
Capetus
Balanerpeton
Dendrerpetidae
Acheloma
Phonerpeton
EcolsoniaBroiliellus brevis
Amphibamus
Doleserpeton
Eoscopus
Platyrhinops
Micromelerpeton
Apateon
Leptorophus
Schoenfelderpeton
Isodectes
Albanerpetidae
Eocaecilia
Karaurus
Valdotriton
Triadobatrachus
Notobatrachus
Vieraella
Phlegethontia
Lethiscus
Oestocephalus
Ptyonius
Sauropleura
Urocordylus
Acherontiscus
Adelospondylus
Adelogyrinus
Dolichopareias
Trimerorhachis
Neldasaurus
Caerorhachis
Solenodonsaurus
Kotlassia
Discosauriscus
Ariekanerpeton
Leptoropha
Microphon
Utegenia
Seymouria
Diadectes
Orobates
Tseajaia
Limnoscelis
Captorhinus
Paleothyris
Petrolacosaurus
Westlothiana
Batropetes
Brachydectes
Scincosaurus
Keraterpeton
Diceratosaurus
Batrachiderpeton
Diplocaulus
Diploceraspis
Tuditanus
Asaphestera
Pantylus
Stegotretus
Saxonerpeton
Hapsidopareion/Llistrofus
Rhynchonkos
Micraroter
Pelodosotis
Cardiocephalus
Euryodus
Odonterpeton
Microbrachis
Hyloplesion
Bruktererpeton
Gephyrostegus
Silvanerpeton
Eoherpeton
Proterogyrinus
Archeria
Pholiderpeton scutigerum
Pholiderpeton attheyi
Anthracosaurus
Crassigyrinus
Whatcheeria
Pederpes
Ossinodus
Ichthyostega
Pantylidae
Tuditanidae
Salientia
Caudata
Gymnarthridae
Ostodolepididae
Sauropsida
Diadectomorpha
“microsaurs”
“temnospondyls”
“edopoids”
adelospondyls
“microbrachomorph” 2
“microbrachomorphs” 1
Seymouriamorpha
Gymnophionomorpha
Lysorophia
BatrachiaLissamphibia
“aïstopods”
“dvinosaurs”
Urocordylidae
Diplocaulidae
Adelogyrinidaetetrapod crown-group
Amphibia
Whatcheeriidae
Baphetoidea
Anthracosauria
Embolomeri
Colosteidae
Branchio-
sauridae
Edopoidea
Gephyrostegidae
Amphibamidae
Dissorophoidea (S13)
Dissorophoidea
(YW00)
Dissorophoidea (S13)
Dissorophoidea
(YW00)
Trematopidae
Branchiosauridae
Amphibamidae
Trematopidae
origin of digits
Eusthenopteron
Panderichthys
Ventastega
Acanthostega
Tulerpeton
Colosteus
Greererpeton
Baphetes
Megalocephalus
Eucritta
Edops
Chenoprosopus
Cochleosaurus
Eryops
Capetus
Balanerpeton
Dendrerpetidae
Acheloma
Phonerpeton
Ecolsonia
Broiliellus brevis
Amphibamus
Doleserpeton
Albanerpetidae
Karaurus
Valdotriton
Triadobatrachus
Notobatrachus
Vieraella
Eoscopus
Platyrhinops
Micromelerpeton
Apateon
Leptorophus
Schoenfelderpeton
Isodectes
Trimerorhachis
Neldasaurus
Caerorhachis
Eocaecilia
Acherontiscus
Adelospondylus
Adelogyrinus
Dolichopareias
Ptyonius
Urocordylus
Sauropleura
Lethiscus
Oestocephalus
Phlegethontia
Batropetes
Brachydectes
Scincosaurus
Keraterpeton
Diceratosaurus
Batrachiderpeton
Diplocaulus
Diploceraspis
Tuditanus
Asaphestera
Pantylus
Stegotretus
Saxonerpeton
Hapsidopareion/Llistrofus
Rhynchonkos
Micraroter
Pelodosotis
Cardiocephalus
Euryodus
Odonterpeton
Microbrachis
Hyloplesion
Westlothiana
Diadectes
Orobates
Tseajaia
Limnoscelis
Captorhinus
Paleothyris
Petrolacosaurus
Kotlassia
Discosauriscus
Ariekanerpeton
Leptoropha
Microphon
Utegenia
Seymouria
Solenodonsaurus
Bruktererpeton
Gephyrostegus
Silvanerpeton
Eoherpeton
Proterogyrinus
Archeria
Pholiderpeton scutigerum
Pholiderpeton attheyi
Anthracosaurus
Crassigyrinus
Whatcheeria
Pederpes
Ossinodus
Ichthyostega
Salientia
Caudata
adelo-
spondyls
Gymnophionomorpha
“aïstopods”
Urocordylidae
Adelogyrinidae
tetrapod crown-group
Temnospondyli
Amphibamidae
Lepospondyli
Holospondyli
Procera
Whatcheeriidae
Baphetoidea
Anthracosauria
Embolomeri
Colosteidae
Dvinosauria
Branchiosauridae
Edopoidea
Gephyrostegidae
Dissorophoidea
(S13)
Dissorophoidea
(YW00)
Trematopidae
origin of digits
Pantylidae
Tuditanidae
Gymnarthridae
Ostodolepididae
“microsaurs” 1
“m.” 2 “microbrachomorph” 2
“microbrachomorphs” 1
Lysorophia
Diplocaulidae
Sauropsida
Diadectomorpha
Seymouriamorpha
↑ Figure 12: Representation of the MPTs (length: 2169 steps) from Analysis R3, 2501
performed on our modified matrix. The taxon sample was unchanged from RC07, bone 2502
losses were coded as reversible. A constraint forced the dissorophoid temnospondyl 2503
Doleserpeton to be closer to the three salientians (Triadobatrachus, Notobatrachus, Vieraella) 2504
than the lysorophian “lepospondyl” Brachydectes, and additionally forced the 2505
gymnophionomorph Eocaecilia to be closer to Brachydectes than to Doleserpeton; this 2506
allowed only the polyphyly hypothesis. 2507
Cyan underlay as in Fig. 10; dark brown underlay for the two positions of 2508
Albanerpetidae. Abbreviation: m., microsaurs. 2509
2510
2511
Other than this, the topology is consistent with the trees from Analysis R1 (Fig. 10). 2512
The clade formed by Diplocaulidae and Scincosaurus does not follow the other holospondyls 2513
into Temnospondyli, but stays behind as the sister-group of (Batropetes + Brachydectes). 2514
Tulerpeton and Kotlassia have the same positions as in Analysis R1; Lissamphibia and 2515
Albanerpetidae are sister-groups in all MPTs. Moving Albanerpetidae back into Lepospondyli 2516
(as the sister-group of Brachydectes) cannot be done in Mesquite without adding at least six 2517
steps. 2518
Statistical tests find the difference between a tree from Analysis R2 and a tree from 2519
Analysis R1 insignificant at a p level of 0.05: p is 0.1465 under the Kishino-Hasegawa test, 2520
0.1531 under the Templeton test and 0.0622 under the winning-sites test (Table 4). 2521
2522
Analysis R3 2523
2524
An analysis of the original taxon sample that was constrained to be compatible with the 2525
polyphyly hypothesis yielded eight MPTs with a length of 2169 steps – only three more than 2526
those from Analysis R2, 12 more than those from Analysis R1 – and a consistency index of 2527
0.2098, a homoplasy index of 0.8345, a retention index of 0.6271 and a rescaled consistency 2528
index of 0.1316. 2529
In the different MPTs (Fig. 12), Tulerpeton and Kotlassia occur in the same positions 2530
as in Analyses R1 and R2. Salientia – (Triadobatrachus (Notobatrachus, Vieraella)) as in R1 2531
and R2 – forms the sister-group to Doleserpeton and is nested within the amphibamid 2532
dissorophoid temnospondyls (which are resolved as in R1), not within the dvinosaurian 2533
temnospondyls as it is in R2. All other modern amphibians have the same positions as in R1; 2534
specifically, in different MPTs, Albanerpetidae is the sister-group of Procera (Eocaecilia + 2535
Caudata) or of Caudata alone. 2536
This version of the PH, where Salientia alone belongs to the temnospondyls while 2537
Caudata and Gymnophionomorpha are lepospondyls, has not been proposed since Carroll & 2538
Holmes (1980). However, to move Caudata into Temnospondyli (as the sister-group of 2539
Salientia) requires three extra steps in Mesquite; to move Caudata + Albanerpetidae into 2540
Temnospondyli is cheaper (two extra steps), but comes at the cost of nesting this clade within 2541
Salientia (next to Notobatrachus + Vieraella) and turning Batropetes, Brachydectes and 2542
Eocaecilia into successively closer relatives of the adelospondyl-urocordylid-aïstopod clade.2543
The difference between a tree from Analysis R3 and a tree from Analysis R1 is 2544
potentially significant at the level of p = 0.05 according to the statistical tests: p = 0.0320 2545
(Kishino-Hasegawa), 0.0321 (Templeton), 0.0274 (winning sites). The difference between 2546
Analyses R3 and R2 is far from significance: p = 0.3852 (Kishino-Hasegawa), 0.3847 2547
(Templeton), 0.3221 (winning sites). 2548
2549
64
64
Eusthenopteron
Panderichthys
Ventastega
Acanthostega
Tulerpeton
Colosteus
Greererpeton
*Deltaherpeton
*Pholidogaster
*St. Louis tetrapod
Baphetes
Megalocephalus
*Sigournea
*Doragnathus
Eucritta
*Spathicephalus
Edops
Chenoprosopus
Cochleosaurus
Isodectes
Trimerorhachis
*Erpetosaurus
*Saharastega
Neldasaurus
*Nigerpeton
*Archegosaurus
*Australerpeton
*Platyoposaurus
*Konzhukovia
*Lydekkerina
*Cheliderpeton
*Glanochthon
*Sclerocephalus
Balanerpeton
Dendrerpetidae
Capetus
*Acanthostomatops
Acheloma
Phonerpeton
Ecolsonia
Broiliellus brevis
Amphibamus
Platyrhinops
Doleserpeton
*Gerobatrachus
EoscopusMicromelerpeton
Leptorophus
Schoenfelderpeton
*Tungussogyrinus
Apateon
*Micropholis
*Branchiosaurus
Broiliellus brevis
Amphibamus
Platyrhinops
Doleserpeton
*Gerobatrachus
Eoscopus
Micromelerpeton
Leptorophus
Schoenfelderpeton
*Tungussogyrinus
Apateon
*Micropholis
*Branchiosaurus
*Mordex
*Iberospondylus
*Palatinerpeton
Eryops
Albanerpetidae
Eocaecilia
Karaurus
*Beiyanerpeton
Triadobatrachus
Notobatrachus
Vieraella
*Liaobatrachus
*Chelotriton
Valdotriton
*Pangerpeton
Batropetes
Brachydectes
*Carrolla
*Quasicaecilia
Tuditanus
*Crinodon
Asaphestera
*Goreville microsaur
Pantylus
Stegotretus
*Sparodus
Saxonerpeton
Hapsidopareion/Llistrofus
Rhynchonkos
Odonterpeton
Micraroter
Pelodosotis
Cardiocephalus
Euryodus
Microbrachis
Hyloplesion
Acherontiscus
Adelospondylus
Adelogyrinus
Dolichopareias
Ptyonius
Urocordylus
Sauropleura
Oestocephalus
Phlegethontia
Lethiscus
*Utaherpeton
Scincosaurus
Keraterpeton
Diceratosaurus
Batrachiderpeton
Diplocaulus
Diploceraspis
*Trihecaton
Westlothiana
Diadectes
Orobates
Tseajaia
Limnoscelis
Captorhinus
*Caseasauria
*Archaeovenator
Paleothyris
Petrolacosaurus
Kotlassia
*Karpinskiosaurus
Seymouria
Discosauriscus
Ariekanerpeton
Leptoropha
Microphon
Utegenia
Solenodonsaurus
*Chroniosaurus
Bruktererpeton
Gephyrostegus
Eoherpeton
Proterogyrinus
Archeria
Pholiderpeton scutigerum
Pholiderpeton attheyi
Anthracosaurus
*NSM 994 GF 1.1
*Palaeoherpeton
*Neopteroplax
Silvanerpeton
Caerorhachis
*Parrsboro jaw
Crassigyrinus
Whatcheeria
Pederpes
Ossinodus
Ichthyostega
*Densignathus
*Metaxygnathus
*Ymeria
*Elginerpeton
Whatcheeriidae
Baphetoidea
Colosteidae
Dvinosauria
“edopoids”
Dissorophoidea (S13)
Dissorophoidea
(content)
Temnospondyli
Stereospon-
dylomorpha/
Limnarchia
“trematopids”
Amphi-
bamidae
“bran-
chio-
saurids”
Dissorophoidea (S13)
“amphi-
bamids”
“bran-
chio-
saurids”
Sauropsida
Synapsida
Amniota
Anthracosauria
Embolomeri
Seymouriamorpha
tetrapod
crown-
group
Amphibia
Pantylidae
Tuditanidae
Salientia
Gymnarthridae
Ostodolepididae
“microsaurs” 3
Diadecto-
morpha
“gephyro-
stegids”
“microsaurs” 4
“m.” 2
“microsaurs” 1
adelospondyls
Holo-
spondyli
“microbrachomorphs” 1
Brachystelechidae
“microbrachomorphs” 4
“microbrachomorph” 3
“microbrachomorph” 2
Gymnophionomorpha
Lysorophia
Lissam-
phibia Batra-
chia
Urocordylidae
Diplocaulidae
Adelogyrinidae
“caudates”
“aïstopods”
origin of digits
E
utem
nospondyli
E
ryopiform
es D
issorophoidea (Y
W
00)
↑ Figure 13: Representation of the MPTs (length: 2901 steps) resulting from Analysis R4 2550
(modified matrix, expanded taxon sample, no constraint, bone losses reversible). The 2551
inset at the left shows the two remaining parsimonious alternatives to a part of the 2552
temnospondyl topology shown in the main tree. 2553
In this and the following figures, the names Stereospondylomorpha and Limnarchia 2554
are placed according to prevailing usage, not according to the definitions by Yates & Warren 2555
(2000) or Schoch (2013). According to the former, Stereospondylomorpha would in this 2556
analysis probably exclude *Sclerocephalus, *Glanochthon and *Cheliderpeton, which were 2557
included in the different topology of Yates & Warren (2000: fig. 1), while Limnarchia would 2558
be a synonym of Dissorophoidea (YW00) – and so would Stereospondylomorpha be 2559
according the definition by Schoch (2013). The same holds for Dvinosauria, which would 2560
under both definitions include *Platyoposaurus in some MPTs from this analysis, 2561
*Platyoposaurus, *Archegosaurus and *Australerpeton in others, and none of these (as 2562
shown) in yet others. 2563
The cyan underlay connects the two positions of *Densignathus (sister to 2564
Ichthyostega, or one node more crownward than Whatcheeriidae); the violet one connects the 2565
several positions of Ymeria, which can be one node more rootward than *Metaxygnathus, 2566
form a clade with it, form a clade with Ichthyostega and *Densignathus, or form a clade with 2567
Ventastega and Acanthostega. The seven positions of the “Parrsboro jaw” (see text) are 2568
connected by the red underlay. The lavender underlay connects the three positions of 2569
Dvinosauria, the bright blue one the two of *Saharastega, the brownish green one the two of 2570
Proterogyrinus, the dark blue one the two of *Neopteroplax, the brown one the two of 2571
*Utaherpeton. 2572
2573
2574
Analysis R4 2575
2576
From here on, the OTUs we have added are marked with an asterisk; taxa absent even from 2577
the expanded taxon sample are marked with two asterisks. 2578
An unconstrained analysis of the increased taxon sample found 786 MPTs with a 2579
length of 2901 steps, a consistency index of 0.1824, a homoplasy index of 0.8762, a retention 2580
index of 0.6222, and a rescaled consistency index of 0.1135. 2581
Figure 13 presents the similarities and differences between the MPTs, based on the 2582
strict and Adams consensus trees and on the first, the last and every 25th tree. 2583
The Devonian area of the tree contains a stable pattern (Fig. 13): *Metaxygnathus, 2584
Ichthyostega, Ventastega, Tulerpeton and Whatcheeriidae are successively more crownward. 2585
This pattern is obscured by the following unstable OTUs: Acanthostega forms a clade with 2586
Ventastega and optionally *Ymeria (see below), or lies one node more crownward than 2587
Ventastega. *Densignathus can be more crownward than even Whatcheeriidae, but can also 2588
form a clade with Ichthyostega and optionally *Ymeria. *Ymeria, finally, has a range of 2589
positions from one node more rootward than *Metaxygnathus to a clade with Ventastega and 2590
Acanthostega. The most rootward member of the non-whatcheeriid post-Devonian clade is 2591
Crassigyrinus, followed by a fully resolved Colosteidae; the *“St. Louis tetrapod” 2592
(MB.Am.1441; Clack et al., 2012b) is the sister-group to all other colosteids, among which 2593
*Pholidogaster is the sister-group to (*Deltaherpeton (Colosteus, Greererpeton)). The next 2594
more crownward branch is Baphetoidea; it, too, is fully resolved, in that *Spathicephalus, 2595
Eucritta, Baphetes and Megalocephalus are successively closer to *Doragnathus and 2596
*Sigournea. 2597
The *Parrsboro jaw (Godfrey & Holmes, 1989; Sookias, Böhmer & Clack, 2014) is 2598
found in no less than seven distinct positions: one node more crownward than Baphetoidea, or 2599
66
66
as the sister-group to all other temnospondyls, or to Caerorhachis, or to the 2600
stereospondylomorph temnospondyls *Archegosaurus and/or *Australerpeton, or nested 2601
within Anthracosauria (which is then rearranged internally). 2602
As in RC07, Temnospondyli is more rootward than Anthracosauria; depending on the 2603
position of the *Parrsboro jaw, it is only one or two nodes more crownward than Baphetoidea, 2604
as found by RC07 except for the position of Lissamphibia. Temnospondyl phylogeny is 2605
resolved as an unusual Hennigian comb where Edops, Cochleosauridae, Eryops, 2606
Stereospondylomorpha, a novel clade, *Palatinerpeton and *Iberospondylus are successively 2607
closer to Dissorophoidea. The novel clade contains *Acanthostomatops and Capetus as 2608
successively closer to Balanerpeton and Dendrerpetidae. 2609
It is particularly notable that Dvinosauria is highly nested within 2610
Stereospondylomorpha (in three different positions). The phylogeny of Dvinosauria is 2611
unusual, too: in some trees, it is a Hennigian comb of *Nigerpeton, Neldasaurus, 2612
*Saharastega, Isodectes and (Trimerorhachis + *Erpetosaurus), while in others *Nigerpeton 2613
and *Saharastega together form the sister-group of a more conventional dvinosaur clade 2614
(Neldasaurus (Isodectes (Trimerorhachis, *Erpetosaurus))). Within Dissorophoidea, the 2615
trematopids form a grade at the base, and the remainder has three resolutions: (1) 2616
*Branchiosaurus, Broiliellus, the remaining “branchiosaurs”,Eoscopus and Platyrhinops are 2617
successively closer to a symmetric ((Amphibamus, Doleserpeton) (*Gerobatrachus, 2618
*Micropholis)) clade, where the clade of the “branchiosaurs” other than *Branchiosaurus 2619
contains *Tungussogyrinus as the sister-group of Schoenfelderpeton and is otherwise resolved 2620
as in Analyses R1–R3; (2) (Broiliellus + *Branchiosaurus, the remaining “branchiosaurs”, 2621
*Micropholis, Apateon, (Amphibamus + Platyrhinops) and Eoscopus are successively closer 2622
to (Doleserpeton + *Gerobatrachus), where the clade of the “branchiosaurs” other than 2623
*Branchiosaurus and Apateon contains *Tungussogyrinus as the sister-group of Leptorophus; 2624
(3) identical to (2), except that *Tungussogyrinus is the sister-group of Leptorophus + 2625
Schoenfelderpeton. 2626
Caerorhachis (and optionally the *Parrsboro jaw) is one node more crownward than 2627
Temnospondyli. It is followed crownward by Silvanerpeton and Anthracosauria (whose 2628
mutual positions are fully unresolved), then by Gephyrostegus and Bruktererpeton (in this 2629
order or as a clade), then by *Chroniosaurus (the only chroniosuchian in the matrix) and then 2630
Solenodonsaurus. Just outside the tetrapod crown-group lies a largely resolved 2631
Seymouriamorpha, which consists of a (Seymouria (*Karpinskiosaurus, Kotlassia)) clade and 2632
an incompletely resolved Discosauriscidae. Within the crown, the amniote + diadectomorph 2633
clade is fully resolved; curiously, however, the added OTUs *Caseasauria and 2634
*Archaeovenator emerge as the sister-group to Limnoscelis, followed by the clade of the other 2635
diadectomorphs and then by Sauropsida – in other words, the diadectomorphs are found as 2636
amniotes, specifically as synapsids that lack the synapsid condition (Berman, 2013), but they 2637
do not form a clade (unlike in Berman, 2013, or Analyses R1–R3). 2638
On the amphibian side of the crown – all MPTs support the LH –, Westlothiana 2639
remains the sister-group to the remainder, which has a topology somewhat reminiscent of 2640
Vallin & Laurin (2004). Instead of having a “microsaur” grade at the base as in Analyses R1–2641
R3, this clade shows a basal dichotomy: (*Trihecaton (Holospondyli (Microbrachis, 2642
Hyloplesion))) lies on one side, while the other branch groups the remaining “microsaurs” and 2643
Brachydectes (which is the only lysorophian even in the expanded taxon sample) with the 2644
fully resolved Lissamphibia. 2645
The basal dichotomy of Holospondyli is into a (Scincosaurus + Diplocaulidae) clade, 2646
which is fully resolved as in Analyses R1–R3, and a clade with a backbone of (adelospondyls 2647
(Urocordylidae, Aïstopoda)). The latter clade has three resolutions: (1) Lethiscus lies next to 2648
the other aïstopods as usual, while *Utaherpeton is the sister-group to the rest of the whole 2649
67
67
clade; (2) Lethiscus joins *Utaherpeton in the latter position; (3) (Lethiscus + *Utaherpeton) 2650
is the sister-group to (Urocordylidae + Aïstopoda). 2651
On the other side of the amphibian branch, Pantylidae and Tuditanidae are sister-2652
groups like in Analyses R1–R3; Tuditanidae is augmented by *Crinodon and the *Goreville 2653
microsaur. A (Gymnarthridae + Ostodolepididae) clade lies next to the remainder, in which 2654
Saxonerpeton, an unexpected (Odonterpeton + *Sparodus) clade, Rhynchonkos, 2655
Hapsidopareion/Llistrofus and finally a clade composed of Brachydectes and 2656
Brachystelechidae (composed of Batropetes + (*Carrolla + *Quasicaecilia); named after a 2657
synonym of Batropetes) are successively closer to Lissamphibia; unlike in R1, Lissamphibia 2658
is not the sister-group of Brachydectes alone. Within Lissamphibia, Albanerpetidae emerges 2659
as the sister-group of Batrachia. The closest relative of Salientia is the caudate *Chelotriton; 2660
together they are the sister-group to a clade formed by all other caudates, in which Karaurus 2661
is as highly nested as possible. 2662
2663
Analysis R5 2664
2665
An analysis of the increased taxon sample with a constraint against the LH found 51885 2666
MPTs which have 2913 steps – 12 more than the MPTs of Analysis R4 – as well as a 2667
consistency index of 0.1816, a homoplasy index of 0.8768, a retention index of 0.6203, and a 2668
rescaled consistency index of 0.1126. 2669
It is striking how strong an effect the constraint has on the entire tree. As shown by 2670
comparison of the strict and Adams consensus trees and the first, the last and every 2000th 2671
MPT (Fig. 14, 15), all of which support the TH, the temno- and lepospondyls are drastically 2672
rearranged (the former losing much resolution in the process), and they come closer to each 2673
other, excluding on the one side the anthracosaurs (though not the gephyrostegids, unlike in 2674
Analyses R1–R3) from what is now the tetrapod crown-group and rendering on the other side 2675
Seymouriamorpha the sister-group of the very poorly resolved amniote/diadectomorph clade 2676
to the exclusion of Lepospondyli. Even anthracosaurs and baphetoids are affected; 2677
Gephyrostegus and Bruktererpeton are now sister-groups in all MPTs, and Tulerpeton can lie 2678
crownward as well as rootward of Whatcheeriidae. While the *Parrsboro jaw loses its position 2679
just rootward of Temnospondyli, it gains a new one within it – next to Trimerorhachis + 2680
*Erpetosaurus – in trees where the dvinosaurs remain monophyletic. 2681
The clade of non-trematopid dissorophoids can now be resolved in additional ways; in 2682
particular, (*Tungussogyrinus + *Branchiosaurus) is sometimes the sister-group of 2683
*Acanthostomatops, followed by the remaining dissorophoids except *Mordex, while 2684
*Acanthostomatops is also found next to Dendrerpetidae or Eryops. All MPTs contain a clade 2685
of Amphibamus + Platyrhinops. Unlike in Analysis R2, which was performed under the same 2686
constraint, Lissamphibia is not nested among the dvinosaurs, and no “lepospondyls” are found 2687
among the temnospondyls; instead, either Doleserpeton or *Gerobatrachus, but never both 2688
together, is the sister-group of Lissamphibia. Lissamphibia is resolved with a Procera 2689
topology, where an (Albanerpetidae + Eocaecilia) clade is nested within a caudate grade; 2690
*Chelotriton is the sister-group to all other procerans. Holospondyli is fully resolved 2691
internally, except for the positions of Ptyonius within Urocordylidae and Batropetes within or 2692
next to the clade consisting of the other brachystelechid “microsaurs” and Brachydectes; this 2693
clade forms the sister-group of (Diplocaulidae + Scincosaurus). *Utaherpeton is the sister-2694
group to an (adelospondyl (urocordylid, aïstopod)) clade; *Trihecaton has five tightly 2695
clustered positions in the middle of Lepospondyli. 2696
The statistical tests find the difference between a tree from Analysis R5 and a tree 2697
from Analysis R4 insignificant: p = 0.2722 (Kishino-Hasegawa), 0.1725 (Templeton), 0.2057 2698
(winning sites). 2699
68
68
Eusthenopteron
Panderichthys
Ventastega
Acanthostega
Tulerpeton
Colosteus
Greererpeton
*Deltaherpeton
*Pholidogaster
*St. Louis tetrapod
Baphetes
Megalocephalus
*Sigournea
*Doragnathus
Eucritta
*Spathicephalus
Edops
Chenoprosopus
Cochleosaurus
Isodectes
Trimerorhachis
*Erpetosaurus
*Saharastega
Neldasaurus
*Nigerpeton
*Archegosaurus
*Australerpeton
*Platyoposaurus
*Konzhukovia
*Lydekkerina
*Cheliderpeton
*Glanochthon
*Sclerocephalus
Balanerpeton
Dendrerpetidae
Capetus
*Acanthostomatops
Acheloma
Phonerpeton
Ecolsonia
Broiliellus brevis
Amphibamus
Platyrhinops
Doleserpeton
Albanerpetidae
Karaurus
*Beiyanerpeton
Valdotriton
*Pangerpeton
*Chelotriton
Eocaecilia
*Mordex
*Tungussogyrinus
*Branchiosaurus
*Acanthostomatops
Acheloma
Phonerpeton
Ecolsonia
*Iberospondylus
Triadobatrachus
Notobatrachus
Vieraella
*Liaobatrachus
*Gerobatrachus
Eoscopus
*Micropholis
Micromelerpeton
Apateon
LeptorophusSchoenfelderpeton
*Tungussogyrinus
*Branchiosaurus
*Mordex
*Iberospondylus
*Palatinerpeton
Eryops
Caerorhachis
Eoherpeton
Proterogyrinus
Archeria
Pholiderpeton scutigerum
Pholiderpeton attheyi
Anthracosaurus
*NSM 994 GF 1.1
*Palaeoherpeton
*Neopteroplax
Bruktererpeton
Gephyrostegus
Solenodonsaurus
Kotlassia
*Karpinskiosaurus
Seymouria
Discosauriscus
Ariekanerpeton
Leptoropha
Microphon
Utegenia
Diadectes
Orobates
Tseajaia
Limnoscelis
Captorhinus
*Caseasauria
*Archaeovenator
Paleothyris
Petrolacosaurus
Westlothiana
Batropetes
Brachydectes
*Carrolla
*Quasicaecilia
Scincosaurus
Keraterpeton
Diceratosaurus
Batrachiderpeton
Diplocaulus
Diploceraspis
Acherontiscus
Adelospondylus
Adelogyrinus
Dolichopareias
Ptyonius
Urocordylus
Sauropleura
Oestocephalus
Phlegethontia
Lethiscus
*Utaherpeton
Microbrachis
Hyloplesion
Tuditanus
*Crinodon
Asaphestera
Pantylus
Stegotretus
Saxonerpeton
Hapsidopareion/Llistrofus
Micraroter
Pelodosotis
Rhynchonkos
Cardiocephalus
Euryodus
Odonterpeton
*Goreville microsaur
*Sparodus
*Trihecaton
*Chroniosaurus
Silvanerpeton
*Parrsboro jaw
Crassigyrinus
Whatcheeria
Pederpes
Ossinodus
Ichthyostega
*Densignathus
*Metaxygnathus
*Ymeria
*Elginerpeton
Whatcheeriidae
Baphetoidea
Colosteidae
Dvinosauria
Dissorophoidea
(S13)
Dissorophoidea (S13)
Dissorophoidea
(YW00)
Dissorophoidea (YW00)
Lepospondyli
Pantylidae
Tuditanidae
Salientia
Gymnarthridae
Ostodolepididae
Sauropsida
SynapsidaAmniota
“microsaurs” 1
Diadectomorpha
“microsaurs” 2
“temnospondyls”
“amphibamids” 2
“amphibamid” 1
adelo-
spondyls
Holo-
spondyli
E
ryopiform
es
Edopoidea
“microbrachomorph” 2
“microbrachomorphs” 1
“microbrachomorphs” 3
Anthracosauria
“aïstopods”
Stereo-
spondylo-
morpha/
Limnarchia
Trematopidae
Trematopidae
Caudata/
Procera
“branchiosaurid” 1
“branchio-
saurids” 2
“branchiosaurids” 1
Seymouria-
morpha
Gephyro-
stegidae
Gymnophionomorpha
Lysorophia
Lissam-
phibiaUrocordylidae
Diplocaulidae
tetrapod
crown-
group
Amphibia
Adelogyrinidae
origin of digits
↑ Figure 14: Representation of some MPTs (length: 2913 steps) resulting from Analysis 2700
R5 (modified matrix, expanded taxon sample, constraint against the LH as in Analysis 2701
R2, bone losses reversible). The inset at the left shows an equally parsimonious alternative to 2702
a part of the temnospondyl topology shown in the main tree. The remaining MPTs from R5 2703
are represented in Fig. 15; the differences are again limited to temnospondyl phylogeny. 2704
Cyan, violet, red, brownish green and dark blue underlays as in Fig. 13. Further 2705
underlay colors: magenta: Tulerpeton (2 positions); orange: *Palatinerpeton (2); very dark 2706
purple: Balanerpeton (2); gray-brown: *Acanthostomatops (2); reddish brown: *Glanochthon 2707
(4); yellow-green: Dvinosauria (3); lavender: *Branchiosaurus (2); reddish purple: 2708
*Tungussogyrinus (2); bright green: *Gerobatrachus (2); dark green: (Eocaecilia + 2709
Albanerpetidae) (2); medium brown: Limnoscelis (2); pastel green: Diadectomorpha 2710
excluding Limnoscelis (3); reddish orange: Paleothyris + Petrolacosaurus (4); dark brown: 2711
*Trihecaton (5); light brown: Batropetes (4). 2712
Note that Caerorhachis and *Palatinerpeton “keep a constant distance”: in some 2713
MPTs, Caerorhachis is a temnospondyl while *Palatinerpeton lies next to the Capetus-2714
Edopoidea-Eryopiformes clade, and in the other MPTs Caerorhachis lies outside the tetrapod 2715
crown-group while *Palatinerpeton is the sister-group to all other temnospondyls together. 2716
2717
Figure 15: 2718
Temnospondyl 2719
phylogeny from 2720
those inspected 2721
MPTs from 2722
Analysis R5 that 2723
are not repre-2724
sented in Fig. 14. 2725
Apart from the 2726
absence of 2727
Eocaecilia and 2728
Albanerpetidae, one 2729
inspected MPT from 2730
Analysis R6 is 2731
almost entirely 2732
compatible with this 2733
figure, differing only 2734
in *Sclerocephalus 2735
and *Cheliderpeton 2736
being sister-groups, 2737
*Tungussogyrinus 2738
being the sister-2739
group of (Micromel-2740
erpeton + “branchio-2741
saurids” 2), and 2742
caudate relation-2743
ships, including 2744
*Chelotriton, being 2745
compatible with Fig. 16. See the legend of Fig. 14 for more information. The remaining MPTs 2746
from Analysis 6 are presented in Fig. 16. 2747
Underlays: red: *Parrsboro jaw (2 positions); lavender: *Branchiosaurus (2); reddish 2748
purple: *Micropholis (2); bright green: *Gerobatrachus (2). 2749
70
70
Eusthenopteron
Panderichthys
Ventastega
Acanthostega
Tulerpeton
Colosteus
Greererpeton
*Deltaherpeton
*Pholidogaster
*St. Louis tetrapod
Baphetes
Megalocephalus
*Sigournea
*Doragnathus
Eucritta
*Spathicephalus
Edops
Chenoprosopus
Cochleosaurus
Balanerpeton
Dendrerpetidae
Capetus
*Acanthostomatops
Acheloma
Phonerpeton
Ecolsonia
Broiliellus brevis
Amphibamus
Platyrhinops
Doleserpeton
*Gerobatrachus
Eoscopus
Micromelerpeton
Leptorophus
Schoenfelderpeton
*Tungussogyrinus
Apateon
*Micropholis
*Branchiosaurus
Capetus
Ecolsonia
Acheloma
Phonerpeton
Balanerpeton
*Palatinerpeton
Dissorophoidea (S13)
Dendrerpetidae
*Acanthostomatops
*Mordex
*Iberospondylus
Stereospondylomorpha/
Limnarchia
*Mordex
*Iberospondylus
*Palatinerpeton
Eryops
Albanerpetidae
Eocaecilia
Batropetes
Brachydectes
*Carrolla
*Quasicaecilia
Tuditanus
*Crinodon
Asaphestera
*Goreville microsaur
Pantylus
Stegotretus
*Sparodus
Saxonerpeton
Hapsidopareion/Llistrofus
Rhynchonkos
Odonterpeton
Micraroter
Pelodosotis
Cardiocephalus
Euryodus
Microbrachis
Hyloplesion
Acherontiscus
Adelospondylus
Adelogyrinus
Dolichopareias
Ptyonius
Urocordylus
Sauropleura
Oestocephalus
Phlegethontia
Lethiscus
*Utaherpeton
Scincosaurus
Keraterpeton
Diceratosaurus
Batrachiderpeton
Diplocaulus
Diploceraspis
*Trihecaton
Westlothiana
Diadectes
Orobates
Tseajaia
Limnoscelis
Captorhinus
*Caseasauria
*Archaeovenator
Paleothyris
Petrolacosaurus
Kotlassia
*Karpinskiosaurus
Seymouria
Discosauriscus
Ariekanerpeton
Leptoropha
Microphon
Utegenia
Solenodonsaurus
*Chroniosaurus
Bruktererpeton
Gephyrostegus
Eoherpeton
Proterogyrinus
Archeria
Pholiderpeton scutigerum
Pholiderpeton attheyi
Anthracosaurus
*NSM 994 GF 1.1
*Palaeoherpeton
*Neopteroplax
Silvanerpeton
Caerorhachis
*Parrsboro jaw
Crassigyrinus
Whatcheeria
Pederpes
Ossinodus
Ichthyostega
*Densignathus
*Metaxygnathus
*Ymeria
*Elginerpeton
Karaurus
*Beiyanerpeton
Valdotriton
*Pangerpeton
*Chelotriton
Triadobatrachus
Notobatrachus
Vieraella
*Liaobatrachus
Kotlassia
*Karpinskiosaurus
Leptoropha
Microphon
Seymouria
Discosauriscus
Ariekanerpeton
Utegenia
Amniota
Lepospondyli
Solenodonsaurus
Kotlassia
*Karpinskiosaurus
Seymouria
Discosauriscus
Utegenia
Microphon
Leptoropha
Ariekanerpeton
Amniota
Lepospondyli
Solenodonsaurus
Whatcheeriidae
Baphetoidea
Colosteidae
“edopoids”
Dissorophoidea (S13)
Dissor-
ophoidea
(content)
Dissor-
ophoidea
(content)
Temnospondyli
Amphibamidae
Stereospondylomorpha/Limnarchia as in Fig. 12
“trematopids”
“bran-
chio-
saurids”
Dissorophoidea (YW00)
Sauropsida
Synapsida
Amniota
Anthracosauria
Embolomeri
tetrapod
crown-
group
Lepospondyli
Pantylidae
Tuditanidae
Gymnarthridae
Ostodolepididae
“microsaurs” 3
Diadecto-
morpha
“gephyro-
stegids”
Seymouria-
morphaDiscosaur-
iscidae
“microsaurs” 4
“m.” 2
“microsaurs” 1
adelospondyls
Holo-
spondyli
“microbrachomorphs” 1
Brachystelechidae
“microbrachomorphs” 4
“microbrachomorph” 3
“microbrachomorph” 2
Gymnophionomorpha
Lysorophia
Batrachia
Urocordylidae
Diplocaulidae
Adelogyrinidae
“aïstopods”
origin of digits
Eutem
nospondyli
Eryopiform
es
D
issorophoidea (Y
W
00)
Salientia
Caudata
Trematopidae
Seymouria-
morpha
Seymouria-
morpha
↑ Figure 16: Representation of the inspected MPTs (length: 2918 steps) resulting from 2750
Analysis R6 (modified matrix, expanded taxon sample, constraint for the PH as in 2751
Analysis R3, bone lossesreversible). Note that only those positions of Silvanerpeton and 2752
Anthracosauria and only those phylogenies of non-batrachian temnospondyls and 2753
seymouriamorphs are represented that are not compatible with Figures 14 or 15; most 2754
(perhaps all) arrangements of these taxa found in those figures occur in MPTs from Analysis 2755
R6 as well. The insets at the left show the remaining equally parsimonious alternatives to 2756
parts of the temnospondyl and seymouriamorph topologies shown in the main tree. 2757
Cyan, violet and red underlays as in Fig. 13. Magenta: Tulerpeton (2 positions); 2758
orange: *Spathicephalus (3 positions); greenish yellow: *Branchiosaurus (3); reddish purple: 2759
*Tungussogyrinus (3); dark green: *Pangerpeton (2); reddish brown: (Leptoropha + 2760
Microphon) (4 positions in the inset at the upper left). 2761
2762
2763
Analysis R6 2764
2765
With a constraint for the PH, an analysis of the increased taxon sample found 4966 MPTs 2766
with 2918 steps – 5 more than those from Analysis R5, 17 more than those from Analysis R4. 2767
The MPTs have a consistency index of 0.1813, a homoplasy index of 0.8770, a retention 2768
index of 0.6195 and a rescaled consistency index of 0.1123. 2769
Inspection of the first, the last and every 250th MPT (we later examined every 50th 2770
MPT for the positions of Caerorhachis and the *Parrsboro jaw) shows that *Ymeria, 2771
*Densignathus and the *Parrsboro jaw have the same positions as in Analyses R4 and R5, 2772
except that the *Parrsboro jaw does not seem to occur at the root of Temnospondyli; 2773
Tulerpeton has the same two as in R5 (Fig. 16). Batrachia is nested as the sister-group to 2774
Doleserpeton or *Gerobatrachus, but never both together. Within Batrachia, *Chelotriton is 2775
the sister-group of either the remaining caudates or of Salientia; it appears that only the 2776
former resolution occurs when the sister-group of Batrachia is *Gerobatrachus. 2777
Albanerpetidae and Eocaecilia form the sister-group to (Brachydectes + Brachystelechidae) in 2778
all trees. 2779
Lepospondyli is resolved very much like in Analysis R4, with (Albanerpetidae + 2780
Eocaecilia) as the sister-group of (Brachydectes + Brachystelechidae). *Trihecaton has the 2781
same position as in R4 in all MPTs. 2782
Trees from Analyses R6 and R4 are potentially different at the level of p = 0.05: p = 2783
0.0146 (Kishino-Hasegawa), 0.0149 (Templeton), 0.0155 (winning sites). Unsurprisingly, 2784
trees from Analyses R6 and R5 are hardly distinguishable: p = 0.3968 (Kishino-Hasegawa), 2785
0.4958 (Templeton), 0.5000 (winning sites). 2786
2787
Bootstrap analyses B1 and B2 2788
2789
Figures 17 and 18 present the bootstrap trees for the original and the expanded taxon samples, 2790
respectively (Analyses B1 and B2). They are fully resolved because clades with bootstrap 2791
values below 0.5 (50%) are included if they do not contradict the majority-rule consensus (see 2792
above). In Fig. 17 (Analysis B1), support is skewed towards peripheral nodes, while the 2793
2794
2795
↓ Figure 17: Bootstrap tree from Analysis B1 (modified matrix, original taxon sample, 2796
no constraint, bone losses reversible; compare Fig. 10). Clades with a bootstrap percentage 2797
below 50 are included if they are compatible with those above 50; percentages of 50 and 2798
above are in boldface. 2799
72
72
Eusthenopteron
Panderichthys
Ventastega
Acanthostega
Tulerpeton
Colosteus
Greererpeton
Baphetes
Megalocephalus
Eucritta
Edops
Chenoprosopus
Cochleosaurus
Eryops
Capetus
Balanerpeton
Dendrerpetidae
Acheloma
Phonerpeton
Ecolsonia
Broiliellus brevis
Amphibamus
Doleserpeton
Eoscopus
Platyrhinops
Micromelerpeton
Apateon
Leptorophus
Schoenfelderpeton
Isodectes
Trimerorhachis
Neldasaurus
Caerorhachis
Albanerpetidae
Karaurus
Valdotriton
Triadobatrachus
Notobatrachus
Vieraella
Eocaecilia
Batropetes
Brachydectes
Scincosaurus
Keraterpeton
Diceratosaurus
Batrachiderpeton
Diplocaulus
Diploceraspis
Ptyonius
Sauropleura
Urocordylus
Lethiscus
Oestocephalus
Phlegethontia
Pantylus
Stegotretus
Tuditanus
Asaphestera
Saxonerpeton
Hapsidopareion/Llistrofus
Micraroter
Pelodosotis
Rhynchonkos
Cardiocephalus
Euryodus
Microbrachis
Hyloplesion
Odonterpeton
Westlothiana
Diadectes
Orobates
Tseajaia
Limnoscelis
Captorhinus
Paleothyris
Petrolacosaurus
Kotlassia
Discosauriscus
Ariekanerpeton
Leptoropha
Microphon
Utegenia
Seymouria
Solenodonsaurus
Bruktererpeton
Gephyrostegus
Silvanerpeton
Eoherpeton
Proterogyrinus
Archeria
Pholiderpeton scutigerum
Pholiderpeton attheyi
Anthracosaurus
Acherontiscus
Adelospondylus
Adelogyrinus
Dolichopareias
Crassigyrinus
Whatcheeria
Pederpes
Ossinodus
Ichthyostega
Batrachia
Pantylidae
Tuditanidae
Salientia
Caudata
Gymnarthridae
Ostodolepididae
Sauropsida
Diadectomorpha
“microsaurs”
adelospondyls
Holospondyli
“microbrachomorph” 3
“microbrachomorphs” 1
Seymouriamorpha
Gephyrostegidae
Gymnophionomorpha
Lysorophia
Lissamphibia
Aïstopoda
Urocordylidae
Diplocaulidae
Adelogyrinidae
tetrapod crown-group
Amphibia
Temnospondyli
Whatcheeriidae
Baphetoidea
Anthracosauria
Embolomeri
Colosteidae
Dvinosauria
Branchiosauridae
Edopoidea
AmphibamidaeDissorophoidea (S13)
Dissorophoidea (content)
Dissorophoidea (YW00) Trematopidae
origin of digits
100 100
67
48
87
47
36
38
17
14
41
79
33
37
28
21
11
38
79
31
88
73
77
60
71
43
44
53
31
28
68
96
36
39
54
58
21
33
61
64
59
18
68
63
42
63
21
37
28
16
37
20
25
18
52
26
68
58
45
16
24
20
7
8
3
27
21
30
13
80
58
19
57
81
53
51
26
24
59
90
84
78
94
73
96
45
100
79
70
65
75
52
39
25
35
43
42
90
54
72
Eusthenopteron
Panderichthys
Ventastega
Acanthostega
Tulerpeton
Colosteus
Greererpeton
*Deltaherpeton
*Pholidogaster
*St. Louis tetrapod
Baphetes
Megalocephalus
*Sigournea
*Doragnathus
Eucritta
*Spathicephalus
Edops
Chenoprosopus
Cochleosaurus
Isodectes
Trimerorhachis
*Erpetosaurus
*Saharastega
Neldasaurus
*Nigerpeton
*Archegosaurus
*Australerpeton
*Platyoposaurus
*Konzhukovia
*Lydekkerina
*Cheliderpeton
*Glanochthon
*Sclerocephalus
Balanerpeton
Dendrerpetidae
Capetus
*Acanthostomatops
Acheloma
Phonerpeton
Ecolsonia
Broiliellus brevis
Amphibamus
Platyrhinops
Doleserpeton
*Gerobatrachus
Eoscopus
Micromelerpeton
Leptorophus
Schoenfelderpeton
*Tungussogyrinus
Apateon
*Micropholis
*Branchiosaurus
*Mordex
*Iberospondylus
*Palatinerpeton
Eryops
Albanerpetidae
Eocaecilia
Karaurus
*Beiyanerpeton
Triadobatrachus
Notobatrachus
Vieraella
*Liaobatrachus
*Chelotriton
Valdotriton
*Pangerpeton
Batropetes
Brachydectes
*Carrolla
*Quasicaecilia
Tuditanus
*Crinodon
Asaphestera
*Goreville microsaur
Pantylus
Stegotretus
*Sparodus
Saxonerpeton
Hapsidopareion/Llistrofus
Rhynchonkos
Odonterpeton
Micraroter
Pelodosotis
Cardiocephalus
Euryodus
Microbrachis
Hyloplesion
Acherontiscus
Adelospondylus
Adelogyrinus
Dolichopareias
Ptyonius
Urocordylus
Sauropleura
Oestocephalus
Phlegethontia
Lethiscus
*Utaherpeton
Scincosaurus
Keraterpeton
Diceratosaurus
Batrachiderpeton
Diplocaulus
Diploceraspis
*Trihecaton
Westlothiana
Diadectes
Orobates
Tseajaia
Limnoscelis
Captorhinus
*Caseasauria
*Archaeovenator
Paleothyris
Petrolacosaurus
Kotlassia
*Karpinskiosaurus
Seymouria
Discosauriscus
Ariekanerpeton
Leptoropha
Microphon
Utegenia
Solenodonsaurus
*Chroniosaurus
Bruktererpeton
Gephyrostegus
Eoherpeton
Proterogyrinus
Archeria
Pholiderpeton scutigerum
Pholiderpeton attheyi
Anthracosaurus
*NSM 994 GF 1.1
*Palaeoherpeton
*Neopteroplax
Silvanerpeton
Caerorhachis
*Parrsboro jaw
Crassigyrinus
Whatcheeria
Pederpes
Ossinodus
Ichthyostega
*Densignathus
*Metaxygnathus
*Ymeria
*Elginerpeton
Whatcheeriidae
Baphetoidea
Colosteidae
Dvinosauria
Edopoidea
Dissorophoidea (S13)
Dissorophoidea
(content)
Dissoro-
phoidea
(YW00)
Temnospondyli
Gephyrostegidae
Limnarchia
Eryopi-
formesStereo-
spondylo-
morpha
Trematopidae
“branchiosaurids” 1
“amphi-
bamids”
“branchi-
osaur-
ids” 2
Saur-
opsida
Synapsida
Amniota
Anthracosauria
Seymouria-
morpha
tetrapod
crown-
group
Amphibia
Pantylidae
Tuditanidae
Salientia
Gymnarthridae
Ostodolepididae
“microsaurs” 1
Diadectomorpha
“micro-
saurs” 3
“microsaur” 2
adelospondyls
“microbrachomorphs” 1
“microbrachomorph” 2
Brachystelechidae
“microbrachomorphs” 3
Gymnophionomorpha
Lysorophia
Lissamphibia
Batrachia
modern
amphi-
bians
Urocordylidae
Diplocaulidae
Adelogyrinidae
“caudates”
“aïstopods”
origin of digits
100
97
41
33
11
22
13
28
42
21
34 61
30
35
22
27 22
18
34
86
59
54
28
50
37
24
29
73
54
13
14
12
12
22
16
16
14
14
25
21
39
29
20
12
67
65
17
64
46
49
3786
81
23
25
30
32
17
70
46
5
38
73
20
23
47
18
25
16
27
29
22
29
72
27
42
26
6
5
4
9
8
8
9
3
3
1
1
8
10
8
8
8 1
2
4
16
36
56
26
46
43
22
58
18
7
8
71
28
19
30
14
37
26
15
12
47
23
20
12
3
12
23
12
16
50
36
44
28
83
27
40
18
40
74
80
44
52 68
59
60
66
37
47
100
66
61
28
↑ Figure 18: Bootstrap tree from Analysis B2 (modified matrix, expanded taxon sample, 2800
no constraint, bone losses reversible; compare Fig. 13). See legend of Fig. 17. 2801
Adding taxa without adding characters (Analysis B2) slightly increases the bootstrap 2802
support of a few nodes, but – unsurprisingly – generally depresses the support of almost the 2803
entire tree (Fig. 18). Of its potentially controversial branches, only the sister-group 2804
relationship of Microbrachis and Hyloplesion (66%), that of Scincosaurus and Diplocaulidae 2805
(67%), the adelospondyl clade (65%) and the clade of modern amphibians (Albanerpetidae + 2806
Lissamphibia: 64%) are found by more than 50% of bootstrap replicates; Batrachia is found in 2807
49% of the replicates, Lissamphibia in 46%, Karaurus + *Beiyanerpeton likewise in 46%, 2808
Gephyrostegidae in 47%, *Archegosaurus + *Australerpeton in 48%, Doleserpeton + 2809
*Gerobatrachus in 40%, the *“St. Louis tetrapod” forms a clade with the other colosteids in 2810
43%, and Tulerpeton belongs to the otherwise post-Devonian clade in 42%. The addition of 2811
two synapsids (Synapsida: 59%) has made Limnoscelis the sister-group of Diadectomorpha + 2812
Amniota (27%) and reduced the support for the monophyly of Sauropsida to 28%, although 2813
support for Amniota is considerably higher at 42%, and all amniotes and “diadectomorphs” 2814
including Limnoscelis continue to form a well-supported (72%) exclusive clade. 2815
2816
2817
“trunk” of the tree has bootstrap percentages well below 50. Still, together with many 2818
uncontroversial results, the position of Ichthyostega rootward of Ventastega and Acanthostega 2819
is supported (67%), as are the position of Whatcheeriidae rootward of Crassigyrinus and 2820
Colosteidae (54% and 72%, respectively), the position of Eucritta as a baphetoid (79%), 2821
Balanerpeton + Dendrerpetidae (51%), Dissorophoidea sensu lato excluding Lissamphibia 2822
(80%), Ecolsonia as a trematopid (53%), Trematopidae as the sister-group to Dissorophoidea 2823
sensu stricto (52%), Amphibamidae excluding Branchiosauridae (58%), Micromelerpeton as 2824
the sister-group of Branchiosauridae (61%), Bruktererpeton and Gephyrostegus as sister-2825
groups (53%), the adelospondyl clade (64%), the monophyly of Diadectomorpha (59%) and 2826
Sauropsida (79%) as well as their sister-group relationship (68%), Microbrachis and 2827
Hyloplesion as sister-groups (70%), Lissamphibia including Albanerpetidae (58%), Batrachia 2828
excluding Albanerpetidae (68%), Caudata (73%) and the sister-group relationship of 2829
Scincosaurus and Diplocaulidae (58%). Further, Acanthostega as crownward of Ventastega 2830
(48%), Whatcheeriidae as rootward of Tulerpeton (47%) and Batrachia + Albanerpetidae 2831
(45%) are each almost certainly better supported than any single alternative. In contrast, 2832
Kotlassia is found within Seymouriamorpha only in 26% of the replicates. 2833
2834
Analysis R7 2835
2836
Analysis R7 differs from Analysis R1 (which both used the original taxon sample of RC07) in 2837
that all losses of entire bones were coded as irreversible. The 104 resulting MPTs (Fig. 19) 2838
have 2168 steps, a modest increase from the 2157 of Analysis R1; correspondingly, the 2839
indices are similar (consistency index = 0.2053, homoplasy index = 0.8386, retention index = 2840
0.6479, rescaled consistency index = 0.1330). 2841
2842
2843
↓ Figure 19: Representation of only those MPTs (length: 2168 steps) from Analysis R7 2844
(modified matrix, original taxon sample, no constraint, bone losses coded as irreversible) 2845
that are not identical to MPTs from Analysis R1 (Fig. 10) in topology; all topologies from 2846
R1 are also found in R7, except that R7 never recovers Albanerpetidae in a place other than 2847
the one shown here. Cyan underlay for the two positions of Tulerpeton as in Fig. 10. 2848
75
75
Eusthenopteron
Panderichthys
Ventastega
Acanthostega
Tulerpeton
Colosteus
Greererpeton
Baphetes
Megalocephalus
Eucritta
Edops
Chenoprosopus
Cochleosaurus
Eryops
Capetus
Balanerpeton
Dendrerpetidae
Acheloma
Phonerpeton
Ecolsonia
Broiliellus brevis
Amphibamus
Doleserpeton
Eoscopus
Platyrhinops
Micromelerpeton
Apateon
Leptorophus
Schoenfelderpeton
Isodectes
Trimerorhachis
Neldasaurus
Caerorhachis
Albanerpetidae
Karaurus
Valdotriton
Triadobatrachus
Notobatrachus
Vieraella
Eocaecilia
Batropetes
Brachydectes
Tuditanus
Asaphestera
Pantylus
Stegotretus
Acherontiscus
Adelospondylus
Adelogyrinus
Dolichopareias
Ptyonius
Sauropleura
Urocordylus
Lethiscus
Oestocephalus
Phlegethontia
Scincosaurus
Keraterpeton
Diceratosaurus
Batrachiderpeton
Diplocaulus
Diploceraspis
Saxonerpeton
Hapsidopareion/Llistrofus
Rhynchonkos
Micraroter
Pelodosotis
Cardiocephalus
Euryodus
Odonterpeton
Microbrachis
Hyloplesion
Westlothiana
Diadectes
Orobates
Tseajaia
Limnoscelis
Captorhinus
Paleothyris
Petrolacosaurus
Kotlassia
Discosauriscus
Ariekanerpeton
Leptoropha
Microphon
Utegenia
Seymouria
Solenodonsaurus
Bruktererpeton
Gephyrostegus
Silvanerpeton
Eoherpeton
Proterogyrinus
Archeria
Pholiderpeton scutigerum
Pholiderpeton attheyi
Anthracosaurus
Crassigyrinus
Whatcheeria
Pederpes
Ossinodus
Ichthyostega
Batrachia
Pantylidae
Salientia
Caudata
Gymnarthridae
Ostodolepididae
Sauropsida
Diadectomorpha
“microsaurs” 1
“edopoids”
adelospondyls
“microsaur” 2
Tuditanidae
“microbrachomorph” 2
“microbrachomorphs” 1
Seymouriamorpha
Gymnophionomorpha
Lysorophia
Lissamphibia
Aïstopoda
Urocordylidae
Diplocaulidae
Adelogyrinidae
tetrapod crown-group
Amphibia
Temnospondyli
Gephyrostegidae
Whatcheeriidae
Baphetoidea
Anthracosauria
Embolomeri
Colosteidae
Dvinosauria
Branchiosauridae
AmphibamidaeDissorophoidea (S13)
Eutemnospondyli
Dissorophoidea
(content)
Dissorophoidea
(YW00)
Trematopidae
origin of digits
Holospondyli
Eusthenopteron
Panderichthys
Ventastega
Acanthostega
Tulerpeton
Colosteus
Greererpeton
Baphetes
Megalocephalus
Eucritta
Edops
Chenoprosopus
Cochleosaurus
Eryops
Capetus
Balanerpeton
Dendrerpetidae
Acheloma
Phonerpeton
Ecolsonia
Broiliellus brevis
Amphibamus
Doleserpeton
Eoscopus
Platyrhinops
Micromelerpeton
Apateon
Leptorophus
Schoenfelderpeton
Isodectes
Trimerorhachis
Neldasaurus
Caerorhachis
Batropetes
Brachydectes
Tuditanus
Asaphestera
Pantylus
Stegotretus
Acherontiscus
Adelospondylus
Adelogyrinus
Dolichopareias
Ptyonius
Sauropleura
Urocordylus
Lethiscus
Oestocephalus
Phlegethontia
Scincosaurus
Keraterpeton
Diceratosaurus
Batrachiderpeton
Diplocaulus
Diploceraspis
Saxonerpeton
Hapsidopareion/Llistrofus
Rhynchonkos
Micraroter
Pelodosotis
Cardiocephalus
Euryodus
Odonterpeton
Microbrachis
Hyloplesion
Westlothiana
Diadectes
Orobates
Tseajaia
Limnoscelis
Captorhinus
Paleothyris
Petrolacosaurus
KotlassiaDiscosauriscus
Ariekanerpeton
Leptoropha
Microphon
Utegenia
Seymouria
Solenodonsaurus
Bruktererpeton
Gephyrostegus
Silvanerpeton
Eoherpeton
Proterogyrinus
Archeria
Pholiderpeton scutigerum
Pholiderpeton attheyi
Anthracosaurus
Crassigyrinus
Whatcheeria
Pederpes
Ossinodus
Ichthyostega
Albanerpetidae
Karaurus
Valdotriton
Triadobatrachus
Notobatrachus
Vieraella
Eocaecilia
Pantylidae
Gymnarthridae
Ostodolepididae
Sauropsida
Diadectomorpha
“microsaurs” 1
“edopoids”
adelospondyls
“microsaur” 2
Tuditanidae
“microbrachomorph” 2
“microbrachomorphs” 1
Seymouriamorpha
Lysorophia
Aïstopoda
Urocordylidae
Diplocaulidae
Adelogyrinidae
tetrapod crown-group
Lepospondyli
Amphi-
bia
Gephyrostegidae
Whatcheeriidae
Baphetoidea
Anthracosauria
Embolomeri
Colosteidae
Dvinosauria
Branchiosauridae
Amphibamidae
Dissorophoidea (S13)
Eutemnospondyli
Dissorophoidea
(content)
Dissorophoidea
(YW00)
Trematopidae
origin of digits
Holospondyli
Batrachia
Salientia
Caudata
Gymnophionomorpha
Lissamphibia
“temnospondyls”
↑ Figure 20: Representation of the MPTs (length: 2178 steps) from Analysis R8 (modified 2849
matrix, original taxon sample, constraint against the LH as in Analyses R2 and R5, bone 2850
losses irreversible). 2851
Cyan underlay as in Fig. 10; the dark green underlay connects the two positions of 2852
Broiliellus brevis, the brown one those of Albanerpetidae. 2853
2854
2855
The MPTs (we have inspected the first, the last and every fifth) are generally very 2856
similar to those of Analysis R1; some are even identical in topology to some from R1 (despite 2857
being 11 steps longer as some state distributions are interpreted as multiple events of 2858
convergence rather than single reversals). All support the LH. As in R1, Tulerpeton lies one 2859
node more rootward or one node more crownward than Whatcheeriidae, and Kotlassia occurs 2860
one node more rootward than Seymouriamorpha or as the sister-group to all other 2861
seymouriamorphs. The amniote-diadectomorph-amphibian clade is fully resolved; unlike in 2862
R1, Albanerpetidae is always the sister-group of Batrachia. Silvanerpeton can be one node 2863
more crownward than Anthracosauria like in R1, but it can also be an anthracosaur (the sister-2864
group to all others). The internal and external relationships of Temnospondyli can be as in R1, 2865
but in other trees Anthracosauria (including Silvanerpeton), Caerorhachis, Temnospondyli 2866
and Gephyrostegidae are successively more crownward. In these latter trees, Temnospondyli 2867
is rearranged internally as well: Dvinosauria, Cochleosauridae, Edops, Eryops, Capetus, 2868
Dendrerpetidae and Balanerpeton are successively closer to Dissorophoidea, within which 2869
Broiliellus lies either next to Amphibamidae alone (as in R1) or next to Amphibamidae and 2870
the “branchiosaurs” together (as in RC07). When Amphibamidae lies next to Broiliellus, it is 2871
resolved as in R1; otherwise, Eoscopus is the sister-group to the rest, and Amphibamus can be 2872
the sister-group of Doleserpeton or Platyrhinops. 2873
2874
Analysis R8 2875
2876
Adding a constraint against the LH results in 35 MPTs which have 2178 steps, 10 more than 2877
those from Analysis R7. Their indices are worse than those from Analysis R7, but not by 2878
much (consistency index = 0.2043, homoplasy index = 0.8393, retention index = 0.6458, 2879
rescaled consistency index = 0.1320). Apart from the position of Lissamphibia (the TH is 2880
supported), the MPTs (Fig. 20) are almost identical to some MPTs from Analysis R7 (Fig. 2881
19); unlike in Analysis R2, but like in R5 (ignoring *Gerobatrachus in some trees), 2882
Lissamphibia lies next to Doleserpeton, and no “lepospondyls” are found among the 2883
temnospondyls. 2884
Inspection of all MPTs finds only the following differences among them: Tulerpeton 2885
and Kotlassia each occur in the same two positions as in Analyses R1–R3 and R7; Broiliellus 2886
is found as the sister-group of Amphibamidae alone or of the entire amphibamid-2887
“branchiosaur” clade; Albanerpetidae can be the sister-group of Batrachia, of Caudata, of 2888
Valdotriton or even of Salientia alone, an entirely novel arrangement. 2889
A tree from Analysis R8 and one from Analysis R7 are not significantly different at p 2890
= 0.05: p = 0.1870 (Kishino-Hasegawa), 0.1734 (Templeton), 0.0975 (winning sites). 2891
2892
Analysis R9 2893
2894
A constraint for the PH results in 327 MPTs of 2179 steps, only one more than those from 2895
Analysis R8 and 11 more than those from Analysis R7. Their indices are: consistency index = 2896
0.2042, homoplasy index = 0.8394, retention index = 0.6456, rescaled consistency index = 2897
0.1318. 2898
78
78
Eusthenopteron
Panderichthys
Ventastega
Acanthostega
Tulerpeton
Colosteus
Greererpeton
Baphetes
Megalocephalus
Eucritta
Edops
Chenoprosopus
Cochleosaurus
Eryops
Capetus
Balanerpeton
Dendrerpetidae
Acheloma
Phonerpeton
Ecolsonia
Broiliellus brevis
Amphibamus
Doleserpeton
Eoscopus
Platyrhinops
Micromelerpeton
Apateon
Leptorophus
Schoenfelderpeton
Isodectes
Trimerorhachis
Neldasaurus
Caerorhachis
Albanerpetidae
Karaurus
Valdotriton
Triadobatrachus
Notobatrachus
Vieraella
Eocaecilia
Batropetes
Brachydectes
Tuditanus
Asaphestera
Pantylus
Stegotretus
Acherontiscus
Adelospondylus
Adelogyrinus
Dolichopareias
Ptyonius
Sauropleura
Urocordylus
Lethiscus
Oestocephalus
Phlegethontia
Scincosaurus
Keraterpeton
Diceratosaurus
Batrachiderpeton
Diplocaulus
Diploceraspis
Saxonerpeton
Hapsidopareion/Llistrofus
Rhynchonkos
Micraroter
Pelodosotis
Cardiocephalus
Euryodus
Odonterpeton
Microbrachis
Hyloplesion
Westlothiana
Diadectes
Orobates
Tseajaia
Limnoscelis
Captorhinus
Paleothyris
Petrolacosaurus
Kotlassia
Discosauriscus
Ariekanerpeton
Leptoropha
Microphon
Utegenia
Seymouria
Solenodonsaurus
Bruktererpeton
Gephyrostegus
Silvanerpeton
Eoherpeton
Proterogyrinus
Archeria
Pholiderpeton scutigerum
Pholiderpeton attheyi
Anthracosaurus
Crassigyrinus
Whatcheeria
Pederpes
Ossinodus
Ichthyostega
Pantylidae
Salientia
Caudata
Gymnarthridae
Ostodolepididae
Sauropsida
Diadectomorpha
“microsaurs” 1
“edopoids”
adelospondyls
Tuditanidae
“microbrachomorph” 2
“microbrachomorphs” 1
Seymouriamorpha
Gymnophionomorpha
Lysorophia
Procera
Aïstopoda
Urocordylidae
Diplocaulidae
Adelogyrinidae
tetrapod crown-group
Lepospondyli
Temno-
spondyli
Gephyrostegidae
Whatcheeriidae
Baphetoidea
Anthracosauria
Embolomeri
Colosteidae
Dvinosauria
Branchiosauridae
Amphibamidae
Dissorophoidea (S13)
Eutemnospondyli
Dissorophoidea
(content)
Dissorophoidea
(YW00)
Trematopidae
origin of digits
Holospondyli
↑ Figure 21: Representation of only those MPTs (length: 2179 steps) from Analysis R9 2899
(modified matrix, original taxon sample, constraint for the PH as in Analyses R3 and 2900
R6, bone losses irreversible) that are not identical to MPTs from Analysis R3 (Fig. 12) in 2901
topology; all topologies from R3 are also found in R9, except that R9 never recovers 2902
Albanerpetidae in a place other than the one shown here. 2903
Cyan underlay as in Fig. 10; the dark green underlay connects the two positions of 2904
Broiliellus brevis, the blue one the two of Eoscopus. 2905
2906
2907
Inspection of the first ten and every tenth MPT reveals two classes of topologies, 2908
much like in Analysis R7. One is identical to some trees from Analysis R3 (Fig. 12). The 2909
other (Fig. 21) is mostly identical to the class from Analysis R7 shown in Fig. 19: 2910
Caerorhachis lies just outside the tetrapod crown-group, Temnospondyli is rearranged and 2911
slightly destabilized, Gephyrostegidae lies just inside the tetrapod crown-group (on the non-2912
temnospondyl side), and Silvanerpeton is an anthracosaur. In both classes, Salientia alone 2913
forms the sister-group of Doleserpeton in all MPTs; an (Eocaecilia (Albanerpetidae, 2914
Caudata)) clade – not sharedin this form with the remaining trees from Analysis R3 – lies 2915
next to Brachydectes, followed by Batropetes; Tulerpeton and Kotlassia each keep their two 2916
usual positions. 2917
2918
2919
↓ Figure 22: Representation of the MPTs (length: 2913 steps) from Analysis R10 2920
(modified matrix, expanded taxon sample, no constraint, bone losses irreversible). For 2921
space reasons, some clades that are resolved exactly as in Fig. 23 are not shown here. The 2922
insets in this and the next two figures show the remaining equally parsimonious alternatives to 2923
parts of the temnospondyl topologies shown in the main tree. (The topology shown in the 2924
Stereospondylomorpha inset was only found in one inspected MPT; it is possible that it 2925
should look like the corresponding inset in Fig. 24, for which we sampled the MPTs more 2926
densely.) Some resolutions allowed by this representation do not occur in any inspected MPT; 2927
e.g., Dissorophoidea and Dvinosauria are never both closer to Dendrerpetidae than to Eryops 2928
in the same tree. 2929
The name Eryopiformes applies to two different nodes on this tree depending on the 2930
position of Eryops; the name Dissorophoidea (S13) applies to two different nodes depending 2931
of the main “branchiosaur” clade that contains Micromelerpeton. These nodes are pointed out 2932
by gray lines. 2933
Underlays: reddish purple: *Ymeria (several positions); cyan: *Densignathus (2); dark 2934
blue: Tulerpeton (2); bluish teal: Silvanerpeton (2); magenta: *Parrsboro jaw (3); dark green: 2935
Proterogyrinus (2); bright red: *Neopteroplax (2); yellow-green: Gephyrostegidae (2); 2936
purplish red: *Palatinerpeton (4 positions in main tree, note another in the 2937
Stereospondylomorpha inset); light orange: Balanerpeton (3); dark purplish blue: 2938
Dissorophoidea (YW00 – in some trees, Dendrerpetidae, Balanerpeton, *Palatinerpeton 2939
and/or *Acanthostomatops would be included under this definition) (6); saturated orange: 2940
Dvinosauria (4); yellow: *Erpetosaurus (2); beige: *Acanthostomatops (3 in main tree – note 2941
one next to (*Branchiosaurus + *Tungussogyrinus) in the main tree, and two more in the left 2942
two Dissorophoidea insets); middle green: *Iberospondylus (2 in main tree, note others in the 2943
Dissorophoidea insets); light green: *Micropholis (2 in main tree, 4 more in total in the 2944
Dissorophoidea insets); violet: *Tungussogyrinus (4 in main tree, 3 more in total in the 2945
Dissorophoidea insets); medium brown: clade of Micromelerpeton, Apateon, Leptorophus, 2946
Schoenfelderpeton (2); dark brown: Apateon (2 in top left inset). 2947
80
80
Eusthenopteron
Panderichthys
Ventastega
Acanthostega
Tulerpeton
Colosteus
Greererpeton
*Deltaherpeton
*Pholidogaster
*St. Louis tetrapod
Edops
Cochleosaurus
Chenoprosopus
Eryops
*Sclerocephalus
*Cheliderpeton
*Archegosaurus
*Konzhukovia
*Lydekkerina
*Glanochthon
*Platyoposaurus
*Palatinerpeton
*Australerpeton
Acheloma
Phonerpeton
Ecolsonia
*Mordex
Broiliellus brevis
Amphibamus
Platyrhinops
Doleserpeton
*Gerobatrachus
Eoscopus
*Tungussogyrinus
*Branchiosaurus
Micromelerpeton
Leptorophus
Schoenfelderpeton
Apateon
*Micropholis
*Iberospondylus
Capetus
*Acanthostomatops
Isodectes
Neldasaurus
*Saharastega
*Nigerpeton
Trimerorhachis
*Erpetosaurus
Balanerpeton
Dendrerpetidae
*Palatinerpeton
Albanerpetidae
Karaurus
Valdotriton
*Pangerpeton
*Beiyanerpeton
Triadobatrachus
Notobatrachus
Vieraella
*Liaobatrachus
*Chelotriton
Eocaecilia
Batropetes
*Quasicaecilia
Brachydectes
*Carrolla
Hapsidopareion/Llistrofus
Rhynchonkos
Odonterpeton
*Sparodus
Saxonerpeton
Micraroter
Pelodosotis
Cardiocephalus
Euryodus
Tuditanus
*Crinodon
Asaphestera
Pantylus
Stegotretus
Microbrachis
Hyloplesion
*Trihecaton
Acherontiscus
Adelospondylus
Adelogyrinus
Dolichopareias
Ptyonius
Sauropleura
Urocordylus
Oestocephalus
Phlegethontia
Lethiscus
*Utaherpeton
Scincosaurus
Keraterpeton
Diceratosaurus
Batrachiderpeton
Diplocaulus
Diploceraspis
*Goreville microsaur
Westlothiana
Neldasaurus
*Saharastega
Isodectes
Trimerorhachis
*Erpetosaurus
Edops
Cochleosaurus
Chenoprosopus
*Nigerpeton
Apateon
Micromelerpeton
other “amphibamids”
Leptorophus
Schoenfelderpeton
*Tungussogyrinus
Ecolsonia
Acheloma
Phonerpeton
Broiliellus brevis
*Branchiosaurus
*Micropholis
*Mordex
*Iberospondylus
*Balanerpeton
Micromelerpeton
*Tungussogyrinus
Apateon
Leptorophus
Schoenfelderpeton
Ecolsonia
Broiliellus brevis
*Branchiosaurus
*Micropholis
Eryops
Acheloma
Phonerpeton
*Mordex
*Iberospondylus
*Acanthostomatops
Micromelerpeton
*Tungussogyrinus
Apateon
Leptorophus
Schoenfelderpeton
Ecolsonia
Broiliellus brevis
*Branchiosaurus
*Micropholis
*Gerobatrachus
Acheloma
Phonerpeton
Platyrhinops
Eoscopus
Amphibamus
Doleserpeton
*Mordex
*Acanthostomatops
Solenodonsaurus
*Chroniosaurus
Bruktererpeton
Gephyrostegus
Caerorhachis
*Parrsboro jaw
Eoherpeton
Proterogyrinus
Archeria
Pholiderpeton scutigerum
Pholiderpeton attheyi
Anthracosaurus
*NSM 994 GF 1.1
*Palaeoherpeton
*Neopteroplax
Silvanerpeton
Crassigyrinus
Whatcheeria
Pederpes
Ossinodus
Ichthyostega
*Densignathus
*Metaxygnathus
*Ymeria
*Elginerpeton
Whatcheeriidae
Baphetoidea as in Fig. 23
Euskelia
other “amphibamids”
Colosteidae
Edopoidea
Edopoidea
Dissoroph-
oidea (S13)
Eryopi-
formes
Dissoro-
phoidea
(content)
Dissoro-
phoidea
(YW00)
D
issorophoidea (Y
W
00)
Dissorophoidea (YW00)
D
issorophoidea (content)
D
issorophoidea (content)
D
issoroph-
oidea (S
13)
D
issoroph-
oidea (S
13)
D
issorophoidea (content)
D
issoroph-
oidea (S
13)
Temnospondyli
Stereospondylomorpha
as in inset or Fig. 23
Stereospondylomorpha
Dvinosauria
Dvinosauria
Trematopidae
(Amphi-
bamidae)
“branchio-
saurids”
Amniota incl. Diadectomorpha
as in Fig. 23
Gephyrostegidae
Anthracosauria
Embolomeri
Seymouriamorpha as in Fig. 23
tetrapod
crown-group
Amphibia
Pantylidae
Tuditanidae
Salientia
Gymnarthridae
Ostodolepididae
“microsaurs” 3
“microsaur” 1
“microsaur” 2
adelospondyls
Holospondyli
“microbrachomorphs” 2
Brachystelechidae
“microbrachomorphs” 4
“microbrachomorph” 1
“microbrachomorph” 3
Gymnophionomorpha
Lysorophia
Lissamphibia
Batra-
chia
Urocordylidae
Diplocaulidae
Adelogyrinidae
“caudates”
Aïstopoda
origin of
digits
Trees from Analyses R9 and R8 are statistically indistinguishable from each other: p = 2948
0.4636 (Kishino-Hasegawa), 0.4547 (Templeton), 0.3221 (winning sites). Trees from 2949
Analyses R9 and R7, however, are potentially different according to all three tests: p = 0.0468 2950
(Kishino-Hasegawa), 0.0480 (Templeton), 0.0243 (winning sites). 2951
2952
Analysis R10 2953
2954
An unconstrained analysis of the expanded taxon sample, with all losses of bones coded as 2955
irreversible, found 11890 MPTs that are 2913 steps long, 12 steps longer than those from 2956
Analysis R4 (its analog with free reversibility). The indices are comparable to those of 2957
Analysis R4: consistency index = 0.1775, homoplasy index = 0.8798, retention index = 2958
0.6511, rescaled consistency index = 0.1156. 2959
Inspecting the first, the last and every 500th MPT shows (Fig. 22) the same resolutions 2960
as in Analyses R5 and R6 (Fig. 14, 16) for the relationships between Devonian taxa, 2961
whatcheeriids and more crownward taxa. The resolution of Baphetoidea is different and 2962
slightly worse than in Analysis R4: instead of being the sister-group to all other baphetoids, 2963
*Spathicephalus is the sister-group of either *Sigournea and *Doragnathus or the latter alone; 2964
both of them stay next to Megalocephalus. There is always an anthracosaur clade which lies 2965
one or two nodes more rootward than Temnospondyli. The immediate surroundings of the 2966
anthracosaur clade can be resolved as follows: (1) Silvanerpeton lies one node more rootward 2967
than Anthracosauria, (Caerorhachis + *Parrsboro jaw) one node more crownward, followed 2968crownward by Temnospondyli; (2) Silvanerpeton is one node more rootward than 2969
Anthracosauria, Caerorhachis alone is one node more crownward (followed crownward by 2970
Temnospondyli), the *Parrsboro jaw is nested within Anthracosauria; (3) Silvanerpeton lies 2971
one node more crownward than Anthracosauria, while Caerorhachis forms the sister-group to 2972
a clade composed of the *Parrsboro jaw and the clade of all other temnospondyls. 2973
That latter clade hides a rudimentary backbone which may be represented as 2974
(Balanerpeton, Dendrerpetidae, (Capetus (Edopoidea, Eryops, Stereospondylomorpha))), 2975
where Eryops is closer to either Edopoidea or Stereospondylomorpha. Everything else, it 2976
seems, is highly variable. *Palatinerpeton is perhaps particularly unstable, occurring 2977
variously at the base, two nodes more highly nested, next to Balanerpeton (together forming 2978
the sister-group of Dendrerpetidae) or within a Stereospondylomorpha clade. Unlike in 2979
Analyses R1–R9 and B1–B4, Dvinosauria may form the sister-group of Chenoprosopus 2980
within Edopoidea, with *Nigerpeton as an intermediate (Fig. 22: inset). Contrary to Analyses 2981
R4–R6, Edopoidea is never very close to the temnospondyl root, and Stereospondylomorpha 2982
is never closer to Dissorophoidea than Eryops is (Fig. 22). *Iberospondylus is the sister-group 2983
of Dissorophoidea, of *Mordex alone, or of (Capetus (Edopoidea, Eryopiformes)).Within 2984
Dissorophoidea, a Trematopidae clade may emerge, or the other dissorophoids may be nested 2985
within it, next to Ecolsonia as in Analysis R4 (Fig. 13), or *Mordex alone may break out (Fig. 2986
22: insets). *Branchiosaurus never forms a clade with any other “branchiosaurids” except 2987
sometimes *Tungussogyrinus, which is particularly unstable, as is *Micropholis. 2988
*Acanthostomatops may lie just outside of Dissorophoidea, just inside, or well inside, 2989
forming in the latter case a clade with *Tungussogyrinus and *Branchiosaurus that lies next 2990
to a clade formed by Broiliellus and all amphibamids except *Micropholis. An 2991
(*Acanthostomatops (*Iberospondylus, Dissorophoidea)) clade can be the sister-group of 2992
Eryops. In yet other trees, however, *Acanthostomatops becomes the sister-group of 2993
Dendrerpetidae. 2994
In trees with resolution (3) as described two paragraphs above, Gephyrostegidae lies 2995
one node more crownward or more rootward than Temnospondyli, remaining one node 2996
82
82
crownward of Silvanerpeton in the latter case. Otherwise it remains one node crownward of 2997
Temnospondyli. 2998
Diadectomorpha is nested within Amniota, on the synapsid side, as in R4 and R6. 2999
The amphibian side – all MPTs support the LH – is almost identical to Analysis R4. 3000
Notable differences are the positions of (Microbrachis + Hyloplesion) at the base of the 3001
“microsaur” branch and of *Trihecaton next to Micraroter, within Ostodolepididae. In other 3002
words, all “microsaurs” except the *Goreville microsaur and *Utaherpeton form a grade 3003
which is ancestral to Brachydectes and, separately, to Lissamphibia (Fig. 22). The *“Goreville 3004
microsaur” is the sister-group of Holospondyli or of (Scincosaurus + Diplocaulidae) alone; 3005
*Utaherpeton is the sister-group of either an adelospondyl-urocordylid-aïstopod clade or of 3006
the adelospondyls alone.*Beiyanerpeton is the sister-group of either Karaurus or (Valdotriton 3007
+ *Pangerpeton). 3008
3009
Analysis R11 3010
3011
Adding a constraint against the LH results in 1566 MPTs with 2926 steps (13 more than those 3012
from Analysis R10) and slightly worse indices (consistency index = 0.1767, homoplasy index 3013
= 0.8804, retention index = 0.6492, rescaled consistency index = 0.1147). Inspection of the 3014
first, the last and every 50th MPT shows that the increased resolution is concentrated in 3015
Dissorophoidea and its surroundings; elsewhere, the resolution is at least as high (Fig. 23) as 3016
in Analysis R10. 3017
Surprisingly, all MPTs from this analysis support the PH, with a topology much like in 3018
Analysis R6: Salientia and Caudata together (Batrachia) lie next to either Doleserpeton or 3019
*Gerobatrachus; Albanerpetidae and Eocaecilia form the sister-group to Brachydectes + 3020
Brachystelechidae. Interestingly, *Chelotriton is always a caudate when Batrachia lies next to 3021
*Gerobatrachus, while it is a caudate or a salientian when Batrachia lies next to 3022
Doleserpeton. Other than that, Caudata is fully resolved. 3023
3024
3025
↓ Figure 23: Representation of the MPTs (length: 2926 steps) from Analysis R11 3026
(modified matrix, expanded taxon sample, constraint against the LH as in Analyses R2, 3027
R5 and R8, bone losses irreversible). Rootward of Baphetoidea, the MPTs are identical to 3028
those of R10; this part is not shown for space reasons. Likewise, Anthracosauria and 3029
Dvinosauria are resolved as in R10 and not shown here. In this and the next figure, optional 3030
placements (underlays) at the same position as each other do not occur together in the same 3031
inspected MPT. 3032
The name Eryopiformes applies to two different nodes on this tree depending on the 3033
position of Eryops. These nodes are pointed out by gray lines. 3034
Underlays: bluish teal: Silvanerpeton (2 positions); magenta: *Parrsboro jaw (3); 3035
yellow-green: Gephyrostegidae (2); red: Dissorophoidea (YW00) (5); yellow: 3036
*Palatinerpeton (3 in main tree, 1 more in insets); dark purplish blue: *Acanthostomatops (2 3037
in main tree, more in insets); gray-brown: Balanerpeton (1 in main tree, more in insets); 3038
orange: Dvinosauria (4); brown: *Glanochthon (4); beige: *Konzhukovia (3); violet: 3039
*Tungussogyrinus (2 in main tree, more in insets); dark blue: *Branchiosaurus (2); bright 3040
green: *Gerobatrachus (2); reddish purple: Dissorophoidea (S13) + trematopids other than 3041
*Mordex (2 in top left inset). 3042
Eryopiformes and Euskelia are set in bold parentheses; whether they are extant 3043
depends on the volatile position of Dissorophoidea (YW00). Further, whether Euskelia is 3044
redundant with Eryops depends on the positions of Dissorophoidea (YW00) and 3045
*Acanthostomatops. 3046
83
83
Baphetes
Megalocephalus
*Sigournea
*Doragnathus
*Spathicephalus
Eucritta
Edops
Chenoprosopus
Cochleosaurus
Eryops
*Sclerocephalus
*Cheliderpeton
*Archegosaurus
*Konzhukovia
*Lydekkerina
*Glanochthon
*Platyoposaurus
*Australerpeton
Capetus
Balanerpeton
Dendrerpetidae
Acheloma
Phonerpeton
Ecolsonia
Broiliellus brevis
Amphibamus
Platyrhinops
Doleserpeton
*Gerobatrachus
Karaurus
Valdotriton
*Pangerpeton
*Beiyanerpeton
Triadobatrachus
Notobatrachus
Vieraella
*Liaobatrachus
*Chelotriton
Eoscopus
Micromelerpeton
Apateon
Leptorophus
Schoenfelderpeton
*Micropholis
*Tungussogyrinus
*Branchiosaurus
*Iberospondylus
*Mordex
*Acanthostomatops
*Palatinerpeton
Albanerpetidae
Eocaecilia
Batropetes
Brachydectes
*Carrolla
*Quasicaecilia
Hapsidopareion/Llistrofus
Rhynchonkos
Odonterpeton
*Sparodus
Saxonerpeton
Micraroter
Pelodosotis
Cardiocephalus
Euryodus
Tuditanus
*Crinodon
Asaphestera
Pantylus
Stegotretus
Microbrachis
Hyloplesion
Acherontiscus
Adelospondylus
Adelogyrinus
Dolichopareias
Ptyonius
Sauropleura
Urocordylus
Oestocephalus
Phlegethontia
Lethiscus
*Utaherpeton
Scincosaurus
Keraterpeton
Diceratosaurus
Batrachiderpeton
Diplocaulus
Diploceraspis
*Goreville microsaur
*Trihecaton
Westlothiana
Diadectes
Orobates
Tseajaia
Limnoscelis
Captorhinus
*Caseasauria
*Archaeovenator
Paleothyris
Petrolacosaurus
Kotlassia
*Karpinskiosaurus
Seymouria
Discosauriscus
Ariekanerpeton
Utegenia
Leptoropha
Microphon
Solenodonsaurus
*Chroniosaurus
Bruktererpeton
Gephyrostegus
Caerorhachis
*Parrsboro jaw
Silvanerpeton
*Tungussogyrinus
Micromelerpeton
Apateon
Leptorophus
Schoenfelderpeton
“amphibamids” 2
*Micropholis
Ecolsonia
Broiliellus brevis
*BranchiosaurusPhonerpeton
*Mordex
Acheloma
Micromelerpeton
Apateon
Leptorophus
Schoenfelderpeton
“amphibamids” 2
*Micropholis
*Tungussogyrinus
Broiliellus brevis
*Branchiosaurus
Trematopidae
*Iberospondylus
Eryops
*Acanthostomatops
*Iberospondylus
*Tungussogyrinus
Micromelerpeton
“branchiosaurids” 2
“amphibamids” 2
*Micropholis
Ecolsonia
Broiliellus brevis
*Branchiosaurus
Phonerpeton
*Mordex
Acheloma
Balanerpeton
*Acanthostomatops
*Palatinerpeton
Dendrerpetidae
*Tungussogyrinus
Micromelerpeton
“branchiosaurids” 3
other amphibamids
*Micropholis
Broiliellus brevis
*Branchiosaurus
Trematopidae
Balanerpeton
*Acanthostomatops
*Palatinerpeton
Dendrerpetidae
*Tungussogyrinus
Micromelerpeton
“branchiosaurids” 2
“amphibamids” 2
*Micropholis
Ecolsonia
Broiliellus brevis
*Branchiosaurus
Phonerpeton
*Mordex
Acheloma
*Acanthostomatops
*Iberospondylus
Lepospondyli
Pantylidae
Tuditanidae
Gymnarthridae
“microsaur” 1
“micro-
saurs” 3
Holo-
spondyli
“microbracho-
morphs” 2
Brachystelechidae
“microbrachomorphs” 4
“microbrachomorph” 1
“microbrachomorph” 3
Lysorophia
Urocordylidae
Diplocaulidae
Adelogyrinidae
Gymnophionomorpha
Salientia
Sauropsida
Synapsida
Amniota
Diadectomorpha
Gephyrostegidae
“amphibamids” 2
“amphibamid” 1
Aïstopoda
Ostodolepididae
“branchio-
saurids”
“branchiosaurids” 3
Anthracosauria as in Fig. 22
Dvinosauria as in Fig. 22
“caudates”
Seymouria-
morpha
Dissoroph-
oidea (S13)
Dissoro-
phoidea
(YW00)
Dissorophoidea (YW00)
Dissorophoidea (YW00)
Dissorophoidea (YW00)
Dissorophoidea (YW00)
Batrachia
Baphetoidea
Temno-
spondyli
Edopoidea
(Eryopiformes)
(Euskelia)
Euskelia
D
issorophoidea (Y
W
00)
D
issorophoidea (content)
D
issorophoidea
(content)
D
issoroph-
oidea (content)
Dissorophoidea
(content)
D
issorophoidea (S
13)
D
issorophoidea (S
13)
Dissoroph-
oidea (S13)
Dissoroph-
oidea (S13)
D
issoroph-
oidea (S
13)
adelo-
spondyls
Stereo-
spondylo-
morpha
Trematopidae
Trematopidae
tetrapod
crown-group
as in
Fig. 22
*Tungussogyrinus
Micromelerpeton
“branchiosaurids” 2
“amphibamids” 2
*Micropholis
Ecolsonia
Broiliellus brevis
*Branchiosaurus
Phonerpeton
*Mordex
Acheloma
*Acanthostomatops
*Iberospondylus
Balanerpeton
Dendrerpetidae
*Acanthostomatops
*Palatinerpeton
Caerorhachis
*Parrsboro jaw
Silvanerpeton
Albanerpetidae
Valdotriton
*Pangerpeton
*Beiyanerpeton
Karaurus
Triadobatrachus
Notobatrachus
Vieraella
*Liaobatrachus
*Chelotriton
Solenodonsaurus
*Chroniosaurus
Bruktererpeton
Gephyrostegus
Batropetes
Brachydectes
*Carrolla
*Quasicaecilia
Scincosaurus
Keraterpeton
Diceratosaurus
Batrachiderpeton
Diplocaulus
Diploceraspis
Acherontiscus
Adelospondylus
Adelogyrinus
Dolichopareias
Ptyonius
Sauropleura
Urocordylus
Oestocephalus
Phlegethontia
Lethiscus
*Utaherpeton
Microbrachis
Hyloplesion
Hapsidopareion/Llistrofus
Rhynchonkos
Odonterpeton
*Sparodus
Saxonerpeton
Tuditanus
Pantylus
Stegotretus
Asaphestera
Micraroter
Pelodosotis
Cardiocephalus
Euryodus
*Goreville microsaur
*Crinodon
*Trihecaton
Westlothiana
*Sclerocephalus
*Cheliderpeton
*Archegosaurus
*Konzhukovia
*Lydekkerina
*Glanochthon
*Platyoposaurus
*Palatinerpeton
*Australerpeton
Acheloma
Phonerpeton
Balanerpeton
*Iberospondylus
*Palatinerpeton
*Micromelerpeton
*Mordex
Ecolsonia
Broiliellus brevis
Amphibamidae
*Branchiosaurus
“branchiosaurids” 3
*Acanthostomatops
*Tungussogyrinus
Edops
Chenoprosopus
Cochleosaurus
Capetus
*Iberospondylus
Eryops
Micromelerpeton
Apateon
Leptorophus
Schoenfelderpeton
*Tungussogyrinus
*Branchiosaurus
Acheloma
Phonerpeton
Ecolsonia
*Mordex
Broiliellus brevis
Amphibamus
Platyrhinops
Doleserpeton
*Gerobatrachus
Eoscopus
*Micropholis
Eocaecilia
Micromelerpeton
Balanerpeton
*Tungussogyrinus
Apateon
Leptorophus
Schoenfelderpeton
Ecolsonia
Broiliellus brevis
*Branchiosaurus
Amphibamidae
Acheloma
Phonerpeton
*Mordex
*Acanthostomatops
*Palatinerpeton
Dendrerpetidae
*Tungussogyrinus
Micromelerpeton
Apateon
Leptorophus
Schoenfelderpeton
“amphibamids” 2
*Micropholis
Ecolsonia
Broiliellus brevis
*Branchiosaurus
Phonerpeton
*Mordex
Acheloma
Eryops
*Acanthostomatops
*Iberospondylus
(Eryopi-
formes)
(Euskelia)
Stereospondylomorpha as in inset or Fig. 23
Dissoro-
phoidea
(YW00)
Dissoro-
phoidea
(content)
Dissoroph-
oidea (S13)
Dissoro-
phoidea
(S13)
tetrapod
crown-
group
Amphibia
Anthracosauria as in Fig. 22
Liss-
am-
phibia Batrachia
Lepospondyli Amniota incl. Diadectomorpha as in Fig. 23
Salientia
“caudates”
Pantylidae
Tuditanidae
Gymnarthridae
Ostodolepididae
“micro-
saurs” 2
“microbrach-
omorphs” 2
“microbrachomorphs” 4
Aïstopoda
Lysorophia
(Brachy-
stelechidae)
Urocordylidae
Diplocaulidae
Adelogyrinidae
“microbr.” 3
“microbr.” 1
Holo-
spondyli
adelo-
spondyls
Seymouriamorpha as in Fig. 23
Trematopidae
Amphibamidae
“branchio-
saurids”
Gymnophio.
Dvinosauria as in Fig. 22
Edopoidea
Gephyrostegidae
Baphetoidea as in Fig. 23
as in
Fig. 22
Stereospondylomorpha
Dissorophoidea (YW00)
Dissorophoidea (YW00)
Dissorophoidea (YW00)
Dissoroph-
oidea (S13)
D
issorophoidea (content)
Dissoroph-
oidea (S13)
Euskelia
D
issorophoidea (Y
W
00)
D
issorophoidea (content)
D
issorophoidea (S
13) “branchiosaurids” 2
“trematopids”
↑ Figure 24: Representation of the MPTs (length: 2927 steps) from Analysis R12 3047
(modified matrix, expanded taxon sample, constraint for the TH not used in other 3048
analyses, bone losses irreversible). As indicated in the figure, much of the topology is 3049
identical to R10 and R11 (Fig. 22, 23) and not shown here. 3050
The name Eryopiformes applies to two different nodes on this tree depending on the 3051
position of Eryops. These nodes are pointed out by gray lines. 3052
Underlays: cyan: Silvanerpeton (2 positions); magenta: *Parrsboro jaw (3); yellow-3053
green: Gephyrostegidae (2); yellow: *Palatinerpeton (4 in main tree, more in the three top left 3054
insets); dark purplish blue: Dissorophoidea (YW00) (4 in main tree, arguably 1 more in 3055
Euskelia inset); orange: Dvinosauria (4); light blue: Balanerpeton (2); dark green: 3056
*Iberospondylus (3 in main tree, arguably 1 more in lowermost inset); red: 3057
*Acanthostomatops (2 in main tree, 3 more in the three left insets); violet: *Tungussogyrinus 3058
(5 in main tree, 4 more in insets); reddish purple: *Micropholis (3); bright green: 3059
*Gerobatrachus (2); yellowish beige: *Goreville microsaur (3); dark brown: *Trihecaton (2); 3060
gray-brown: Rhynchonkos (3); medium brown: Saxonerpeton (2); light beige: *Quasicaecilia 3061
(3). 3062
3063
3064
Unlike in R6 or R10, Amphibamus and Platyrhinops are always recovered as sister-3065
groups which lie next to the Doleserpeton-*Gerobatrachus-Batrachia clade, all together 3066
forming the sister-group of Eoscopus. *Acanthostomatops is sometimes the sister-group of 3067
Eryops as in some MPTs from R5 and R6, but never enters Dissorophoidea in the traditional 3068
sense as it does in some MPTs from R10; *Micropholis and *Branchiosaurus have fewer 3069
positions than in R10, while a rearranged Trematopidae sometimes appears in a novel place 3070
next to *Micropholis, Micromelerpeton and the “branchiosaurids” from the original taxon 3071
sample; one fewer topology than in R10 is found for Stereospondylomorpha; and 3072
Hapsidopareion/Llistrofus has the same two positions as in R6. The entire rest of the tree is 3073
resolved in the same ways as in R10 (Fig. 22). 3074
Trees from Analyses R11 and R10 are potentially distinguishable: p = 0.0398 3075
(Kishino-Hasegawa), 0.0411 (Templeton), 0.0277 (winning sites). 3076
3077
Analysis R12 3078
3079
A special constraint for the TH yields 5500 MPTs with 2927 steps (one more than those from 3080
Analysis R11, 14 more than in Analysis R10). Naturally, the indices are almost identical to 3081
those from Analysis R11 (consistency index = 0.1766, homoplasy index = 0.8804, retention 3082
index = 0.6399, rescaledconsistency index = 0.1130). Inspecting the first, the last, and every 3083
50th MPT (Fig. 24) does not reveal many differences from the MPTs from Analyses R10 or 3084
R11. 3085
*Chelotriton forms the sister-group of Salientia; Albanerpetidae is highly nested 3086
within Caudata; Eocaecilia and Batrachia are sister-groups. Analogously to Analysis R11, 3087
either Doleserpeton or *Gerobatrachus, but never both, form the sister-group of 3088
Lissamphibia. When Doleserpeton occupies this position, *Micropholis sometimes nests just 3089
rootward of *Gerobatrachus; otherwise *Micropholis lies next to Eoscopus or next to the 3090
usual “branchiosaur” clade. The resolutions of the remaining temnospondyls are mostly a 3091
subset of those from R10 and R11 combined (Fig. 24).In addition to their positions in R10 3092
and R11, *Trihecaton sometimes forms the sister-group of Holospondyli, and the *“Goreville 3093
microsaur” occasionally nests next to *Sparodus. Saxonerpeton, Rhynchonkos and 3094
*Quasicaecilia are destabilized in more limited ways (Fig. 24). 3095
86
86
Trees from Analyses R12 and R10 are statistically indistinguishable: p = 0.1566 3096
(Kishino-Hasegawa), 0.1259 (Templeton), 0.1449 (winning sites). Unsurprisingly, trees from 3097
Analyses R12 and R11 are also indistinguishable: p = 0.4708 (Kishino-Hasegawa), 0.3977 3098
(Templeton), 0.5000 (winning sites). 3099
3100
Bootstrap analyses B3 and B4 3101
3102
The bootstrap trees and values found when all bone losses are coded as irreversible (Analyses 3103
B3 and B4 of the original and the expanded taxon sample, respectively; Fig. 25, 26) are 3104
largely identical to those found otherwise (Analyses B1 and B2; Fig. 17, 18). Differences in 3105
topology between B1 and B3 are limited to the position of Seymouriamorpha (next to 3106
Diadectomorpha + Amniota in B3) and to “microsaur” phylogeny, none of them supported by 3107
bootstrap percentages above 16. B4 differs from B3 in the positions of *Neopteroplax within 3108
Anthracosauria, (Caerorhachis + *Parrsboro jaw) inside or rootward of Temnospondyli, 3109
temnospondyl phylogeny, the position of Limnoscelis inside or rootward of Diadectomorpha, 3110
and large parts of amphibian phylogeny; however, the only difference supported by a 3111
bootstrap percentage above 31 is the position of Albanerpetidae, which lies outside of 3112
Lissamphibia in B2 (where Lissamphibia, excluding Albanerpetidae, has a value of 0.46) but 3113
inside in B4 (where Lissamphibia, including Albanerpetidae, has a value of 0.51). 3114
3115
Discussion 3116
3117
Taxon and character sets are now large enough to be mined for large-scale evolutionary trends 3118
[…]. 3119
– Coates, Ruta & Friedman, 2008: 572 3120
3121
Our mergers of correlated characters are the largest factor in the decrease of the number of 3122
parsimony-informative characters from 333 to 276. This is less than three times the number of 3123
taxa (which is 102 in the original sample, 150 in the expanded sample). Many of the 3124
characters now have multiple states, so the total number of apomorphic character states has 3125
decreased much less than the number of characters suggests; still, the low number of 3126
characters (compared with the number of taxa) explains why the present matrix does not 3127
provide a fully resolved, robust large-scale phylogeny of the limbed vertebrates. It has 3128
become common for phylogenetic analyses of a similarly large or smaller sample of extinct 3129
amniotes to have far larger numbers of characters (Ezcurra, 2012: 55 OTUs, 1097 characters; 3130
Godefroit et al., 2013: 101 OTUs, 992 characters; Lee et al., 2014: 120 OTUs, 1251 3131
characters; Tschopp, Mateus & Benson, 2015: 81 OTUs, 477 characters; Ezcurra, 2016: 96 3132
OTUs, 600 characters; Simões et al., 2016b: 193 OTUs, 610 characters; Shelley, Williamson 3133
& Brusatte, 2016: 169 OTUs, 693 characters). Indeed, the matrices of McGowan (2002), 3134
Vallin & Laurin (2004), RC07, and Anderson et al. (2008a) each contain characters that the 3135
three others lack; the supermatrix by Sigurdsen & Green (2011), compiled from the last three, 3136
has 335 characters that are parsimony-informative even though the taxon sample is restricted 3137
to 25. There is no reason to suppose that the present character sample could not be greatly 3138
expanded. Therefore, in particular because our trees are so weakly supported, we strongly 3139
doubt the statement quoted above. 3140
3141
3142
↓ Figure 25: Bootstrap tree from Analysis B3 (modified matrix, original taxon sample, 3143
no constraint, bone losses irreversible; compare Fig. 19). See legend of Fig. 17 for more 3144
information. 3145
87
87
Eusthenopteron
Panderichthys
Ventastega
Acanthostega
Tulerpeton
Colosteus
Greererpeton
Baphetes
Megalocephalus
Eucritta
Edops
Chenoprosopus
Cochleosaurus
Eryops
Capetus
Balanerpeton
Dendrerpetidae
Acheloma
Phonerpeton
Ecolsonia
Broiliellus brevis
Amphibamus
Doleserpeton
Eoscopus
Platyrhinops
Micromelerpeton
Apateon
Leptorophus
Schoenfelderpeton
Isodectes
Trimerorhachis
Neldasaurus
Caerorhachis
Albanerpetidae
Karaurus
Valdotriton
Triadobatrachus
Notobatrachus
Vieraella
Eocaecilia
Batropetes
Brachydectes
Scincosaurus
Keraterpeton
Diceratosaurus
Batrachiderpeton
Diplocaulus
Diploceraspis
Ptyonius
Sauropleura
Urocordylus
Lethiscus
Oestocephalus
Phlegethontia
Pantylus
Stegotretus
Tuditanus
Asaphestera
Saxonerpeton
Hapsidopareion/Llistrofus
Micraroter
Pelodosotis
Rhynchonkos
Cardiocephalus
Euryodus
Microbrachis
Hyloplesion
Odonterpeton
Westlothiana
Diadectes
Orobates
Tseajaia
Limnoscelis
Captorhinus
Paleothyris
Petrolacosaurus
Kotlassia
Discosauriscus
Ariekanerpeton
Leptoropha
Microphon
Utegenia
Seymouria
Solenodonsaurus
Bruktererpeton
Gephyrostegus
Silvanerpeton
Eoherpeton
Proterogyrinus
Archeria
Pholiderpeton scutigerum
Pholiderpeton attheyi
Anthracosaurus
Acherontiscus
Adelospondylus
Adelogyrinus
Dolichopareias
Crassigyrinus
Whatcheeria
Pederpes
Ossinodus
Ichthyostega
Batra-
chia
Pantylidae
Tuditanidae
Salientia
Caudata
Gymnarthridae
Ostodolepididae
Sauropsida
Diadectomorpha
“microsaurs”
adelospondyls
Holospondyli
“microbrachomorph” 2
“microbrachomorphs” 1
Seymouriamorpha
Gephyrostegidae
Gymnophionomorpha
Lysorophia
Lissamphibia
Aïstopoda
Urocordylidae
Diplocaulidae
Adelogyrinidae
tetrapod crown-group
Amphibia
Temnospondyli
Whatcheeriidae
Baphetoidea
Anthracosauria
Embolomeri
Colosteidae
Dvinosauria
Branchiosauridae
Edopoidea
AmphibamidaeDissorophoidea (S13)
Dissorophoidea (content)
Dissorophoidea (YW00) Trematopidae
origin of digits
100 100
65
46
87
40
36
37
17
14
42
74
36
32
26
24
12
34
78
26
85
69
75
60
69
45
48
52
28
31
69
95
38
40
51
52
23
30
55
66
71
7
64
70
42
61
21
33
21
18
43
20
25
28
47
22
68
66
52
13
16
12
7
16
16
9
16
22
12
76
55
19
53
80
49
48
25
20
56
87
75
82
91
73
96
52
100
76
75
62
77
61
42
33
38
42
44
89
54
68
Eusthenopteron
Panderichthys
Ventastega
Acanthostega
Tulerpeton
Colosteus
Greererpeton
*Deltaherpeton
*Pholidogaster
*St. Louis tetrapod
Baphetes
Megalocephalus
*Sigournea
*Doragnathus
Eucritta
*Spathicephalus
Edops
Chenoprosopus
Cochleosaurus
Isodectes
Trimerorhachis
*Erpetosaurus
*Saharastega
Neldasaurus
*Nigerpeton
*Archegosaurus
*Australerpeton
*Platyoposaurus
*Konzhukovia
*Lydekkerina
*Cheliderpeton
*Glanochthon
*Sclerocephalus
Balanerpeton
Dendrerpetidae
Capetus
*Acanthostomatops
Acheloma
Phonerpeton
Ecolsonia
Broiliellus brevis
Amphibamus
Platyrhinops
Doleserpeton
*Gerobatrachus
Eoscopus
Micromelerpeton
Leptorophus
Schoenfelderpeton
*Tungussogyrinus
Apateon
*Micropholis
*Branchiosaurus
*Mordex
*Iberospondylus
*Palatinerpeton
Eryops
Albanerpetidae
Eocaecilia
Karaurus
*Beiyanerpeton
Triadobatrachus
Notobatrachus
Vieraella
*Liaobatrachus
*Chelotriton
Valdotriton*Pangerpeton
Batropetes
Brachydectes
*Carrolla
*Quasicaecilia
Tuditanus
*Crinodon
Asaphestera
*Goreville microsaur
Pantylus
Stegotretus
*Sparodus
Saxonerpeton
Hapsidopareion/Llistrofus
Rhynchonkos
Odonterpeton
Micraroter
Pelodosotis
Cardiocephalus
Euryodus
Microbrachis
Hyloplesion
Acherontiscus
Adelospondylus
Adelogyrinus
Dolichopareias
Ptyonius
Urocordylus
Sauropleura
Oestocephalus
Phlegethontia
Lethiscus
*Utaherpeton
Scincosaurus
Keraterpeton
Diceratosaurus
Batrachiderpeton
Diplocaulus
Diploceraspis
*Trihecaton
Westlothiana
Diadectes
Orobates
Tseajaia
Limnoscelis
Captorhinus
*Caseasauria
*Archaeovenator
Paleothyris
Petrolacosaurus
Kotlassia
*Karpinskiosaurus
Seymouria
Discosauriscus
Ariekanerpeton
Leptoropha
Microphon
Utegenia
Solenodonsaurus
*Chroniosaurus
Bruktererpeton
Gephyrostegus
Eoherpeton
Proterogyrinus
Archeria
Pholiderpeton scutigerum
Pholiderpeton attheyi
Anthracosaurus
*NSM 994 GF 1.1
*Palaeoherpeton
*Neopteroplax
Silvanerpeton
Caerorhachis
*Parrsboro jaw
Crassigyrinus
Whatcheeria
Pederpes
Ossinodus
Ichthyostega
*Densignathus
*Metaxygnathus
*Ymeria
*Elginerpeton
Whatcheeriidae
Baphetoidea
Colosteidae
Dvinosauria
Edopoidea
Dissorophoidea (S13)
Dissorophoidea
(content)
Dissoro-
phoidea
(YW00)
Temno-
spondyli
Eryopiformes
Stereo-
spondylo-
morpha
Trematopidae
“branchio-
saurids” 1
“amphi-
bamids” 2
“amphibamid” 1
“branchi-
osaur-
ids” 2
Saur-
opsida
Synapsida
Amniota
Diadectomorpha
Gephyrostegidae
Anthracosauria
Seymouria-
morpha
tetrapod
crown-
group
Amphibia
Pantylidae
Tuditanidae
Salientia
Gymnarthridae
Ostodolepididae
“microsaurs” 2
“microsaurs” 3
“microsaur” 1
adelospondyls
Holospondyli
“microbrachomorphs” 2
“microbrachomorph” 1
Brachystelechidae
“microbrachomorphs” 3
Gymnophionomorpha
Lysorophia
Lissamphibia
Batrachia
Urocordylidae
Diplocaulidae
Adelogyrinidae
“caudates”
“aïstopods”
origin of digits
99
96
44
36
11
26
11
26
40
20
30 58
27
32
16
25
21
5
34
84
60
59
24
45
36
26
26
80
43
15
20
10
10
17
12
17
11
14
27
20
42
26
16
9
72
62
18
51
45
53
38
84
87
26
28
36
20
30
70
28
1
43
70
14
7
43
19
23
15
23
23
23
26
71
31
32
23
6
10
15
4
8
7
7
1
3
3
7
11
3
7
10
8
1
2
6
14
37
55
21
48
39
21
55
20
9
8
69
29
27
29
13
38
26
15
12
42
24
12
15
4
11
23
20
26
52
37
50
27
82
27
35
19
35
66
76
43
53 67
58
57
62
38
47
100
73 58
28
↑ Figure 26: Bootstrap tree from Analysis B4 (modified matrix, expanded taxon sample, 3146
no constraint, bone losses irreversible; compare Fig. 22). See legend of Fig. 17 for more 3147
information. 3148
3149
3150
Unfortunately, the character sample is not only small, it is not random. We think that 3151
postcranial characters in particular are underused in the present matrix. This is implied by its 3152
craniocentrism – 189 of the 276 characters, a bit more than two thirds, describe the skull, 3153
lower jaw or teeth. Further, the endochondral braincase (Maddin, Jenkins & Anderson, 2012; 3154
Pardo, 2014; Szostakiwskyj, Pardo & Anderson, 2015; Pardo, Szostakiwskyj & Anderson, 3155
2015a, b; Ascarrunz et al., 2016; Pardo & Anderson, 2016) is represented by only three 3156
characters (which all concern the occiput), and the hyobranchial skeleton (Witzmann, 2013) – 3157
including even the much-discussed stapes (e.g. Lombard & Bolt, 1979, 1988; Bolt & 3158
Lombard, 1985, 1992; Laurin, 1998a; Schoch, 2002a; Clack, 2003b; Robinson, Ahlberg & 3159
Koentges, 2005; Sigurdsen, 2008) – has not been considered at all. The analyses of Pawley 3160
(2006), who added a net total of 53 postcranial characters to the 95 of the matrix of Ruta, 3161
Coates & Quicke (2003 – the preceding version of RC07) and found different results, appear 3162
to confirm our suspicion that the present matrix and the resulting trees are affected by 3163
sampling bias. For an overlapping taxon sample, Ruta (2011) found 157 characters from the 3164
appendicular skeleton alone, where the present matrix has 53. 3165
Reevaluating the matrix has revealed additional character conflict and polymorphism 3166
and greatly increased the length of the most parsimonious trees. Evidently, the MPTs found 3167
by RC07 painted an oversimplified picture of the evolution of the limbed vertebrates. 3168
The addition of taxa to the analysis has had unexpected effects that may have 3169
improved the reliability of the tree. As previously demonstrated (Mortimer, 2006; Butler & 3170
Upchurch, 2007), every OTU in a data matrix can influence the position of every other OTU 3171
in the resulting cladogram. 3172
Unstable areas of the tree and other phenomena highlight promising areas for future 3173
research. These include redescription of Westlothiana (currently being undertaken; M. Ruta, 3174
pers. comm. 2015), **Eldeceeon (see “Taxa that were not added” above), the “microsaurs” 3175
Asaphestera, Tuditanus, Odonterpeton and *Trihecaton (see below), **Sauravus (the 3176
presumed sister-group of Scincosaurus), Casineria (especially in order to determine whether 3177
it is distinguishable from Caerorhachis), perhaps *Utaherpeton (see below), and quite 3178
possibly others. 3179
3180
Phylogenetic relationships 3181
3182
In this section we mention the bootstrap percentages (BPO, BPE, BPIO, BPIE from Analyses 3183
B1, B2, B3, B4, shown in Fig. 17, 18, 25, 26, respectively) that support or contradict clades 3184
found by our analyses. Most of them are below 50, especially for the expanded taxon sample 3185
(Fig. 18, 26); this highlights the suboptimal ratio between number of characters and number 3186
of taxa, as well as character conflict and wildcard taxa. As in Figures 17, 18, 25 and 26 as 3187
well as Tables 3 and 4, bootstrap percentages of 50 and above are presented in boldface. 3188
3189
Devonian taxa and Whatcheeriidae 3190
3191
Despite many changes to its scores, Whatcheeriidae is found as a clade unchanged from RC07 3192
– (Ossinodus (Pederpes, Whatcheeria)) – in all analyses (BPO = 41 for the whole and 68 for 3193
(Pederpes + Whatcheeria); BPE = 34 and 61, respectively; BPIO = 42 and 69; BPIE = 30 and 3194
58). This contrasts with the topology (Pederpes (Ossinodus, Whatcheeria)) found by Pawley 3195
90
90
(2006) before the recent descriptions of further Ossinodus material (Warren, 2007; Bishop, 3196
2014). 3197
3198
Figure 27: Hypotheses on the 3199
relationships of Devonian limbed 3200
and possibly limbed vertebrates. 3201
(Whatcheeriidae, which has limbs, is 3202
only known from the Carboniferous 3203
with certainty.) Equally parsimoni-3204
ous positions of the same OTU are 3205
highlighted in color in this and the 3206
following figures. A: RC07 and 3207
references therein. B: our results 3208
from the same taxon sample as RC07 3209
(Analyses R1–R3, R7–R9). Numbers 3210
below inter-nodes are BPO/BPIO, in 3211
boldface if 50 or higher. C: our 3212
results from the expanded taxon 3213
sample (R4–R6, R10–R12); Analysis 3214
R4 limits Tulerpeton to the more 3215
rootward of its two positions. Num-3216
bers below internodes are BPE/BPIE, 3217
in boldface if 50 or higher; neither of 3218
the shown positions of *Densigna-3219
thus is supported by Analyses B2 and 3220
B4, which instead place 3221
*Densignathus where Tulerpeton is 3222
in R4 (BPE: 28/BPIE: 26). 3223
3224
3225
At the base of the ingroup, RC07 found the “textbook” Hennigian comb (Fig. 2, 27A): 3226
(Panderichthys (Ventastega (Acanthostega (Ichthyostega (Tulerpeton, post-Devonian 3227
clade))))). However, the position of Ichthyostega soon began to be questioned (Ahlberg et al., 3228
2008: supplementary information; Clack et al., 2012a) as more information about it was 3229
discovered and published (Clack, 2007; Callier, Clack & Ahlberg, 2009; Ahlberg, 2011; 3230
Clack et al., 2012a; Pierce, Clack & Hutchinson, 2012; Pierce et al., 2013).In all of our 3231
analyses (Fig. 27B, C), Ichthyostega is more rootward than Ventastega (BPO = 67; BPE = 3232
BPIE = 11; BPIO = 65) and Acanthostega (BPO = 67; BPE = 13; BPIO = 65; BPIE = 11). We 3233
predict that this result will be upheld by analyses with larger character and outgroup samples; 3234
in this part of the tree, the character sampling by RC07 was denser than in many others. The 3235
picture of Ichthyostega as the Godzilla of mudskippers (Pierce, Clack & Hutchinson, 2012; 3236
see also Coates & Clack, 1995), an animal that acquired an amphibious lifestyle wholly 3237
independently from more crownward walking and limbless tetrapodomorphs that lived after 3238
Romer’s Gap, may well be accurate. 3239
With the addition of taxa (Fig. 27C), the position of Acanthostega is somewhat 3240
destabilized: in Analyses R4–R6 and R10–R12 it remains more crownward than Ventastega 3241
in some trees (BPE = 13, BPIE = 11) but becomes its sister-group or that of Ventastega + 3242
*Ymeria (a clade with BPE = 22, BPIE = 26) in others. 3243
The added Devonian taxa are all poorly known. Yet, the stability of their positions is 3244
not proportional to the amount of scored data. While the comparatively well known 3245
91
91
*Elginerpeton has a stable position (BPE = 97 rootward, 41 crownward; BPIE = 96 rootward, 3246
44 crownward), so does the isolated lower jaw called *Metaxygnathus (BPE = 41 rootward, 3247
33 crownward; BPIE = 44 rootward, 36 crownward); the isolated lower jaw called 3248
*Densignathus has two clusters of positions (both contradicted by the bootstrap trees at BPE 3249
= 28 rootward, 42 crownward; BPIE = 26 rootward, 40 crownward), and *Ymeria, known 3250
from a lower jaw and a few other fragments, ranges through the rootward half of the new 3251
backbone of (Panderichthys (*Elginerpeton (*Metaxygnathus (Ichthyostega (Ventastega, 3252
Acanthostega (Tulerpeton, Whatcheeriidae, non-whatcheeriid post-Devonian clade)))))). 3253
The more crownward cluster of positions of *Densignathus – as the sister-group of the 3254
non-whatcheeriid post-Devonian clade, or as the sister-group of that clade + Tulerpeton – is 3255
particularly intriguing in the light of the “whatcheeriid-grade” skull bones that were found 3256
with *Densignathus; we are much less certain than Daeschler, Clack & Shubin (2009) that 3257
they do not belong to *Densignathus themselves. Alternatively, however, we find 3258
*Densignathus as the sister-group of Ichthyostega or of (Ichthyostega + *Ymeria), far from 3259
Whatcheeriidae (against a BPE of 28 and a BPIE of 26). 3260
In all analyses, Tulerpeton can be one node more rootward than Whatcheeriidae as in 3261
RC07 and Pawley (2006) (against a BPO = BPIO of 87, a BPE of 42 and a BPIE of 40); in all 3262
except R4, it can also lie one or two nodes more crownward (BPO = 47; BPE = 21; BPIO = 3263
40; BPIE = 20). Interestingly, Tulerpeton and Whatcheeriidae never emerge as sister-groups 3264
(Fig. 27B, C), unlike in Ruta & Bolt (2006) and the almost identical Ruta (2009) or in Clack 3265
& Klembara (2009); creating such a topology in Mesquite adds one extra step to Analysis R1, 3266
two to R4, at least one to R7 and at least two to R10 (the exact number is unknown for R7 and 3267
R10 because Mesquite cannot model irreversible characters). 3268
We eagerly await the forthcoming redescription of **Elpistostege based on a whole 3269
articulated CT-scanned skeleton (Cloutier & Béchard, 2013; Cloutier et al., 2016); the 3270
improved sample of Devonian stem-tetrapods and probably characters in the accompanying 3271
phylogenetic analysis will likely contribute to resolving this part of the tree. 3272
3273
Mississippian mysteries 3274
3275
RC07 found Colosteidae, Crassigyrinus and Whatcheeriidae to be successively more 3276
crownward (Fig. 2, 28A). Like Ruta (2009), we find them successively more rootward in all 3277
analyses (Fig. 27B, 28B; BPO = 72, BPE = 35, BPIO = 68 and BPIE = 32 for Colosteidae + 3278
more crownward taxa). This agrees with other recently published analyses; we attribute this to 3279
our corrections. For example, RC07 (and Ruta, Coates & Quicke, 2003) had scored 3280
Whatcheeria as lacking the preopercular bone in the skull (state PREOPE 1(1), which is 3281
shared by the colosteids and Crassigyrinus), so that its presence (PREOPE 1(0)) in Pederpes 3282
appeared as an isolated reversal. This was apparently due to a misreading of Clack (1998), 3283
who explicitly confirmed that the preopercular is present in Whatcheeria as described by 3284
Lombard & Bolt (1995) and Bolt & Lombard (2000); see character 81 (PREOPE 1) in 3285
Appendix 1 for discussion. The third sampled whatcheeriid, Ossinodus, was scored as 3286
unknown, which was correct at the time; it is now known to have retained the preopercular as 3287
well (Warren, 2007), and we have scored it accordingly. 3288
RC07 also found Eucritta as the sister-group of Baphetes + Megalocephalus. We 3289
confirm this (BPO = 79; BPE = 36; BPIO = 74; BPIE = 37), as did Ruta (2009), Milner, 3290
Milner & Walsh (2009) and the original full description by Clack (2001). 3291
Like Eucritta, *Spathicephalus has been considered a non-baphetid baphetoid 3292
(Beaumont & Smithson, 1998; Milner, Milner & Walsh, 2009; following the nomenclature of 3293
Milner, Milner & Walsh, 2009). *Spathicephalus is consistently a baphetoid in our analyses 3294
as well (BPE = 16, BPIE = 14), and in Analyses R4, B2 and B4 and some trees from R6 it is 3295
92
92
found outside the smallest clade that contains Eucritta and Baphetidae (BPE = 36, BPIE = 3296
40), which is compatible with the results by Milner, Milner & Walsh (2009); in Analyses R5, 3297
R10–R12 and the remaining trees from R6, however, it is nested within Baphetidae, closer to 3298
Megalocephalus than to Baphetes. 3299
The discrepancy between our various analyses cannot be resolved by appealing to that 3300
of Milner, Milner & Walsh (2009). That analysis sampled baphetids and baphetoid-related 3301
characters more densely; however, it is possible that these advantages of Milner, Milner & 3302
Walsh (2009) are outweighed by the disadvantages of their much smaller outgroup sample 3303
(only Acanthostega and Crassigyrinus), the fact that they did not order any characters (of the 3304
24 characters, seven are continuous or meristic multistate characters), redundancy between 3305
characters 3 and 4, and the scoring of juvenile morphology as adult (affecting at least 3306
character 19). 3307
A further factor that may affect the position of *Spathicephalus may be the enigmatic 3308
*Doragnathus and *Sigournea, the latter known from an isolated lower jaw, the former from 3309
fragments of lower and upper jaws. Both are found as baphetoids in all analyses that include 3310
them (BPE = 21; BPIE = 26). In Analyses R4–R6 and R10–R12, these two taxa consistently 3311
nest as baphetids closer to Megalocephalus than Baphetes is (against a BPE of 56 and a BPIE 3312
of 55); in Analyses R5, R10–R12, B2 and B4 and some trees from R6, *Spathicephalus forms 3313
a clade with them (BPE = 26, BPIE = 21). Similarities between all these taxa have long been 3314
noted, but never tested in a phylogenetic analysis (see “Taxa added as new OTUs for a 3315
separate set of analyses” in the “Material and methods” section for *Sigournea). Although our 3316
weakly supported results remain to be tested with a larger sample of characters and of 3317
baphetoids, we think that *Doragnathus and *Sigournea will likely be upheld as baphetoids, 3318
possibly as baphetids (against a BPE of 56 and a BPIE of 55) and likely as sister-groups (BPE 3319
= 46, BPIE = 48). It is noteworthy in this respect that Smithson & Clack (2013) noted 3320
similarities to baphetoids in the isolated postcranial material from the site where 3321
*Doragnathus was found; perhaps it belongs to *Doragnathus after all. 3322
3323
Colosteidae 33243325
Colosteidae is consistently found in a position one node more rootward than Baphetoidea and 3326
one node more crownward than Crassigyrinus (Fig. 28B), as found e.g. by Pawley (2006) and 3327
Ruta (2009) and discussed immediately above. 3328
Among the added taxa are two uncontroversial colosteids. Analyses R4–R6 and R10–3329
R12, the first phylogenetic analyses to include them, unsurprisingly find *Pholidogaster (BPE 3330
= 58; BPIE = 55) and *Deltaherpeton (BPE = 71; BPIE = 69) as successively closer to 3331
Colosteus + Greererpeton (a clade with a BPE of 50 and a BPIE of 52). 3332
Also included in the enlarged taxon sample are *Erpetosaurus and the unnamed *St. 3333
Louis tetrapod (MB.Am.1441). The latter was described (Clack et al., 2012b) as having 3334
several similarities to colosteids, but also some to temnospondyls, while lacking several 3335
features otherwise found in colosteids; the former was most recently redescribed as a 3336
dvinosaurian temnospondyl with convergent similarities to colosteids, including features that 3337
the *St. Louis tetrapod lacks (Milner & Sequeira, 2011). In our analyses, the *St. Louis 3338
tetrapod consistently emerges as the sister-group to all other colosteids (BPE = 43, BPIE = 3339
39), far from the temnospondyls, while *Erpetosaurus is nested within the dvinosaurs (BPE = 3340
12, BPIE = 11; BPE = 27 and BPIE = 25 for not lying next to Colosteidae, BPE = 43 and 3341
BPIE = 39 for not being nested within it) when a dvinosaur clade is found; even otherwise 3342
(i.e. in some MPTs from R5 and R6), it still forms the sister-group of Trimerorhachis (BPE = 3343
12; BPIE = 11). 3344
93
93
3345
3346
Figure 28: Hypotheses on the relationships of post-Devonian limbed vertebrates, and 3347
distribution of several character states. A: RC07; the entire clade shown here equals the 3348
“post-Devonian limbed vertebrates” of Fig. 27A. Anthr., Anthracosauria. B: our results. The 3349
entire clade shown here equals the “other post-Devonian limbed vertebrates” of Fig. 27B. 3350
Added taxa, where applicable, are not shown. The extant amphibians are shown indirectly as 3351
the three positions of the tetrapod crown-group; although not rendered in boldface, “(rest of) 3352
Temnospondyli” is extant in R2, R3, R5, R6, R8, R9, R11 and R12, and “Lepospondyli” is 3353
extant in R1, R3, R4, R6, R8 and R11. Parentheses show which of these positions are found in 3354
which analyses. The numbers of these analyses are written in black if these analyses find a 3355
single position; for any position found as the single position by any analyses, the parentheses 3356
themselves are also black. Numbers below internodes are BPO/BPE/BPIO/BPIE, in boldface 3357
if 50 or higher. 3358
The rectangles indicate state 0 (filled) or 1 (empty) of characters 31 (PAR 1 – cyan), 3359
146 (PSYM 1 – teal), 191 (CLE 2 – blue), 211 (HUM 10 – purple), 213 (HUM 12-15 – red), 3360
259 (TRU VER 8 – orange) and 276 (CAU FIN 1 – green); absence of a rectangle means 3361
missing data. Where known, Devonian limbed vertebrates have state 0 of each of these 3362
characters. 3363
31: supratemporal/postparietal suture (0) or tabular/parietal suture (1); 146: presence 3364
(0) or absence (1) of parasymphysial; 191: presence (0) or absence (1) of postbranchial lamina 3365
on cleithrum; 211: humerus not (0) waisted (1); 213: humerus L-shaped (0) or not (1); 259: 3366
absence (0) or presence (1) of fusion between left and right pleurocentra in ventral midline; 3367
276: presence (0) or absence (1) of tail fin skeleton (supraneural radials, lepidotrichia). See 3368
Appendix 1 for more precise definitions and discussion. Note that, of all adelospondyls, 3369
character 276 is only known in Acherontiscus; the adelogyrinids share state 1 if their single-3370
piece centra are pleurocentra, but they have state 0 if their centra are intercentra, a question 3371
we cannot presently decide. 3372
94
94
Further derivations occur within some of the composite clades shown here: the highly 3373
nested temnospondyl Micromelerpeton is polymorphic for 31, and the diadectomorph 3374
Tseajaia has state 31(0); Pederpes (nested within Whatcheeriidae) has state 191(1); 3375
Urocordylidae and the diplocaulid Keraterpeton (“lepospondyls”) have state 211(0); 3376
Eusthenopteron, Panderichthys and the embolomeres Archeria and Pholiderpeton scutigerum 3377
have state 213(1), while Limnoscelis and Orobates (Diadectomorpha) as well as the 3378
diplocaulids Keraterpeton and Diceratosaurus have state 213(0); and the highly nested 3379
temnospondyls Doleserpeton and *Gerobatrachus have state 259(1). State 2 of the ordered 3380
character 213 is not shown; all the OTUs that have this state (an extra-long humerus) are 3381
nested well within clades for which 213(1) is plesiomorphic. 3382
3383
3384
One feature that may be relevant to the relationships of the *St. Louis tetrapod is not 3385
coded in this matrix: as pointed out by Clack et al. (2012b), denticle-covered platelets of 3386
dermal bone that fill the interpterygoid vacuities are only known in that specimen and in 3387
temnospondyls. We speculate that such platelets appear more or less automatically when the 3388
interpterygoid vacuities are wide enough (the *St. Louis tetrapod has the proportionally 3389
widest ones of any colosteid) and denticles are present throughout the roof of the mouth (they 3390
are entirely absent in Batropetes, *Carrolla, apparently *Quasicaecilia, Diplocaulus, 3391
Diploceraspis and lissamphibians except Eocaecilia). This character deserves further 3392
investigation. It may be relevant that Eocaecilia, which fairly clearly lacks such platelets, has 3393
denticles on the pterygoids and the parasphenoid but not on the vomers or palatines (Jenkins, 3394
Walsh & Carroll, 2007). 3395
Whether the *St. Louis tetrapod can claim the title of oldest known colosteid depends 3396
on three older specimens: the **“Type 3 humerus” from Blue Beach in Nova Scotia (middle 3397
Tournaisian), **Aytonerpeton from Burnmouth Ross in (Old) Scotland (middle Tournaisian), 3398
which strongly reminds us of the *St. Louis tetrapod, and possibly **“Ribbo” from the early 3399
middle Tournaisian site of Willie’s Hole in Scotland (Anderson et al., 2015; Clack et al., 3400
2016). 3401
3402
The interrelationships of Anthracosauria, Silvanerpeton, Gephyrostegidae, Caerorhachis and 3403
Temnospondyli 3404
3405
Although [anthracosaurs have] long [been] associated with the amniote stem, this link appears to 3406
be increasingly tenuous. 3407
– Coates, Ruta & Friedman, 2008: 579 3408
3409
Ever since the early 20th century, Anthracosauria (which traditionally encompassed at least 3410
Embolomeri and the more recently discovered Eoherpeton) has generally been considered 3411
closer to Amniota than Temnospondyli is. The analysis by RC07 supported this “textbook” 3412
consensus (Fig. 28A). However, both the anthracosaurs and the temnospondyls share features 3413
with Amniota, Diadectomorpha, “Lepospondyli”, Seymouriamorpha and/or Solenodonsaurus 3414
that the other taxon lacks (Fig. 28). Consequently, Temnospondyli has been found closer to 3415
Amniota than Anthracosauria is in a few phylogenetic analyses (Ahlberg & Clack, 1998; 3416
Laurin & Reisz, 1999: bootstrap tree and discussion; Pawley, 2006; Klembara et al., 2014, 3417
with poor support). Furthermore, both temnospondyls and anthracosaurs may lie outside the 3418
tetrapod crown, as found in several of our previous studies (e.g. Laurin, 1998a; Laurin & 3419
Reisz, 1999; Vallin & Laurin, 2004; Marjanović & Laurin, 2009), as well as by Pawley 3420
(2006) and in all unconstrained analyses of the present matrix (R1, R4, R7, R10; B1–B4; Fig. 3421
28B). 3422
95
95
Our results are inconclusive on the first point (relationships between temnospondyls, 3423
anthracosaurs and amniotes), though they unambiguously support (withonly weak bootstrap 3424
indices; see below) a stem-tetrapod position of anthracosaurs and temnospondyls. Analyses 3425
R1–R3, R5 and R7–R12 as well as some MPTs from R6 and all bootstrap analyses find 3426
Anthracosauria rootward of Temnospondyli (BPO = 17; BPE = 9; BPIO = 17; BPIE = 7), 3427
while the “textbook” version occurs in R4 and the remaining MPTs from R6. 3428
Part of the reason may be character conflict: not only Anthracosauria and 3429
Temnospondyli have conflicting combinations of character states, but so do other taxa found 3430
in the same area of the tree (Fig. 28). 3431
Silvanerpeton, found as an anthracosaur (the sister-group to all others) by RC07, by 3432
our Analysis R8 and by some MPTs from R7 and R9, forms a fully unresolved trichotomy 3433
with Anthracosauria in Analyses R4 and R6, lies one node more crownward or one node more 3434
rootward than Anthracosauria in different MPTs from Analyses R5 and R10–R12, and is 3435
resolved as one node more crownward (BPO = 13; BPE = BPIE = 6; BPIO = 12) in Analyses 3436
R1–R3 and the remaining trees from R7 and R9. Its position in many of our analyses shows 3437
an odd correlation to that of Caerorhachis, in that Silvanerpeton and Caerorhachis “keep a 3438
constant distance” – whenever Silvanerpeton is an anthracosaur or more rootward than 3439
Anthracosauria, Caerorhachis is more rootward than Temnospondyli, and whenever 3440
Silvanerpeton is more crownward than Anthracosauria, Caerorhachis is a temnospondyl. 3441
The instability of Caerorhachis, which RC07 found one node closer to Amniota than 3442
Temnospondyli and one node farther away than Anthracosauria, appears to be unaffected in 3443
its extent or range by our addition of Casineria as part of the same OTU (see “Taxa added as 3444
parts of existing OTUs” in the Methods section; Fig. 3–5 and 28). In Analyses R1–R3 as well 3445
as some trees from Analyses R5–7 and R9–R12, Caerorhachis is the sister-group to all other 3446
temnospondyls (BPO of Temnospondyli including Caerorhachis: 16; BPE = 8; BPIO = 18), 3447
as found by Pawley (2006 – separately for Caerorhachis and Casineria) and suggested earlier 3448
by Godfrey, Fiorillo & Carroll (1987); Analysis R8, the remaining trees from R5, R7 and R9–3449
R12 as well as some from R6 find it outside the smallest clade that contains Amniota and 3450
Temnospondyli (BPIE = 10 for Caerorhachis being rootward of Temnospondyli), one node 3451
more crownward than Anthracosauria. Analysis R4 and the remaining MPTs from R6 find the 3452
same position as RC07 (except that Silvanerpeton may intervene between it and 3453
Anthracosauria as described above). It remains to be seen whether restudy of the specimens 3454
will be able to solve this problem; our Caerorhachis OTU is scored after the detailed 3455
redescription of Caerorhachis by Ruta, Milner & Coates (2002), which contains much 3456
information that RC07 did not consider, and personal observations of Casineria by D. M., 3457
which do not contradict the brief description by Paton, Smithson & Clack (1999) for 3458
characters in this matrix but have yielded additional information – see also below (and Fig. 5). 3459
Paton, Smithson & Clack (1999) emphasized that Casineria had gastrocentrous 3460
vertebrae and claws, highlighting especially the latter as an amniote-like feature. Figure 3 3461
shows that the vertebrae are no more amniote-like than those of Caerorhachis, from which 3462
they differ at most in a slightly lesser degree of ossification; Figure 4 shows that the terminal 3463
phalanges, while curved ventrally, have squared-off ends – neither pointed as expected of a 3464
claw, nor rounded as incorrectly shown in the interpretative drawing by Paton, Smithson & 3465
Clack (1999: fig. 3b). These two character complexes, thus, are not obstacles to placing 3466
Casineria far from amniote origins or even to referring it to Caerorhachis. 3467
RC07 found Gephyrostegus and Bruktererpeton as a paraphyletic series. In Analyses 3468
R1–R3, R5 and R7–R12, as in that by Klembara et al. (2014), they consistently form a well-3469
supported clade (BPO = 53; BPE = 47; BPIO = 49; BPIE = 43). Analyses R4 and R6, 3470
however, find both results to be equally parsimonious. As found by RC07, Gephyrostegus and 3471
Bruktererpeton are more crownward than Anthracosauria in all analyses; they appear one 3472
96
96
node more rootward than Temnospondyli in Analyses R1–R3, but more crownward (as in 3473
RC07) in Analyses R4–R6 and R8 (BPO = BPIO = 14; BPE = 12; BPIE = 10), while R7 and 3474
R9–R12 allow for both options. A sister-group relationship of Gephyrostegus to 3475
Anthracosauria, as found by Pawley (2006), does not occur (and is contradicted by BPO = 3476
BPIO = 17; BPE = 12; BPIE = 10); it may be relevant that Pawley’s (2006) taxon sample 3477
lacked Silvanerpeton and Bruktererpeton. 3478
Character conflict in this part of the tree concerns, among others, features that are 3479
probably relevant to the origin of terrestriality (Fig. 28). At least one anthracosaur (CM 3480
34638, discussed above: Clack, 2011a; see also Holmes & Carroll, 2010) retained a dermal 3481
and endochondral tail fin skeleton (state CAU FIN 1(0) in this matrix, see character 276 in 3482
Appendix 1); Silvanerpeton lacks it (Ruta & Clack, 2006), as do Bruktererpeton, 3483
Gephyrostegus, and all temnospondyls that are well enough known to tell. The anthracosaur 3484
Archeria is known (Pawley, 2006: chapter 6) to retain the postbranchial lamina on the 3485
cleithrum (CLE 2(0)); due to preservation and publication bias (the lamina is practically only 3486
visible in cranial or caudal view), this character is unknown in many OTUs in this matrix, but 3487
the lamina is absent in Gephyrostegus and all sufficiently well known unquestioned 3488
temnospondyls (e.g. Pawley, 2006: chapter 6). Because D. M. (pers. obs.) has identified what 3489
seems to be a postbranchial lamina quite similar to that of Archeria (depicted by Pawley, 3490
2006: fig. 70-2.4) on the cleithrum of Casineria, presented here in Figure 5, we have scored 3491
Caerorhachis as possessing the lamina; this will require further investigation. Anthracosauria, 3492
Silvanerpeton and Gephyrostegidae retain flat humeri (HUM 10(0)), while Caerorhachis 3493
(Casineria; D. M., pers. obs.) shares a waisted (HUM 10(1)), twisted humerus shape with all 3494
skeletally mature temnospondyls; the anthracosaurs Eoherpeton, Proterogyrinus and *NSM 3495
994 GF 1.1 additionally retain an L-shaped humerus (HUM 12-15(0)) where the entepicon-3496
dyle is about as long (measured along its proximal margin) as the part of the humerus 3497
proximal to it, which is not seen in Archeria or Pholiderpeton scutigerum (unknown in Ph. 3498
attheyi, Anthracosaurus, *Palaeoherpeton and *Neopteroplax) or in the other taxa listed in 3499
this paragraph. Finally, pleurocentra that fuse in the ventral midline (TRU VER 8(1)) are 3500
found in all the taxa listed in this paragraph except the unquestioned temnospondyls. (Among 3501
the latter, TRU VER 8(1) does occur in Doleserpeton and *Gerobatrachus, as well as in 3502
**brachyopoid stereospondyls and in unsampled dvinosaurs like **Tupilakosaurus, but these 3503
are too highly nested to be relevant here.) However, while state TRU VER 8(1) presumably 3504
increases the resistance of the vertebral column to compression, this may be dorsoventral 3505
compression from bearing weight just as well as mediolateral compression from e.g. 3506
anguilliform swimming. Various amphibious and terrestrial temnospondyls, on the other 3507
hand, elaborated the neural arches and (in Dissorophidae) neomorphic osteoderms into a 3508
weight-bearing apparatus that evidently did not need extensively ossified centra; furthermore, 3509
the vertebral columns of Gephyrostegus and Bruktererpeton in particular look rather supple in 3510
reconstructions despite showing state TRU VER 8(1). 3511A postbranchial lamina on the clavicle, not coded here, has been identified in the 3512
dvinosaur **Thabanchuia (Witzmann, 2013) and in a number of *stereospondyls: Schoch & 3513
Witzmann (2010) described it in **Plagiosuchus and **Trematolestes; D. M. has seen it on a 3514
clavicle of **Metoposaurus that will be catalogued by the University of Opole (Poland); 3515
Yates & Warren (2000: fig. 7) depicted it in **Benthosuchus, **Paracyclotosaurus, 3516
**Lyrocephaliscus and **Koskinonodon – they only called it “anterior flange (prescapular 3517
process)”, but Florian Witzmann (pers. comm. 2015) confirms that this structure is most 3518
likely identical to the postbranchial lamina. The homology of the postbranchial laminae on the 3519
cleithrum of non-temnospondyls and the clavicle of temnospondyls remains unclear at the 3520
moment. 3521
97
97
Interestingly, the parasymphysial bone in the lower jaw (PSYM 1) has the same 3522
distribution as the postbranchial lamina, missing data excepted (Fig. 28). 3523
Noting the presence of anthracosaurs and Tulerpeton-like animals soon after the 3524
beginning of the Carboniferous, Anderson et al. (2015) drew attention to the old idea that all 3525
or almost all limbed vertebrates are part of the smallest clade that contains Anthracosauria and 3526
Temnospondyli. “If Tulerpeton represents the earliest occurrence of embolomeres (as appears 3527
to be suggested by the close similarity between humeral and femoral morphology)” 3528
(Anderson et al., 2015: 24), the last common ancestor of anthracosaurs and temnospondyls 3529
would have lived in the Devonian. However, Anderson et al. (2015) provided no support for 3530
this complex of ideas which contradicts the current consensus (including RC07) beyond citing 3531
Lebedev & Coates (1995); they did not mention the reanalysis and refutation of the latter by 3532
Laurin (1998b). Unsurprisingly, then, these historical hypotheses are not supported by our 3533
analyses; as discussed by Pawley (2006), the similarities between Tulerpeton and 3534
Anthracosauria are synapomorphic at that level, but were further modified by Temnospondyli 3535
and various other branches – they represent intermediate states. A fortiori, even if some or all 3536
extant amphibians are temnospondyls, there is currently no strong reason to think that the 3537
origin of the tetrapod crown-group happened in the Devonian as Anderson et al. (2015) 3538
speculated. The highest bootstrap percentages that keep Tulerpeton away from 3539
Anthracosauria are BPO = 72, BPE = 35, BPIO = 68 and BPIE = 32 (all for the clade of 3540
colosteids and everything more crownward; Fig. 17, 18, 25, 26); the lack of support in the 3541
analyses of the extended taxon sample is likely due to the unstable positions of 3542
*Densignathus, Whatcheeriidae, Silvanerpeton and Temnospondyli which can all intervene 3543
between Tulerpeton and Anthracosauria. 3544
A topology somewhat similar to the above scenario was found by Klembara et al. 3545
(2014), where a weakly supported (Crassigyrinus (Whatcheeria, Anthracosauria)) clade and a 3546
weakly supported clade composed of Baphetidae and Temnospondyli were found. However, 3547
the second clade came out – likewise with weak support – as closer to Amniota than the first 3548
clade, which appeared just one node crownward of Ichthyostega. Containing 36 ingroup taxa 3549
(a quarter of them seymouriamorphs; Tulerpeton was not sampled), an all-zero outgroup 3550
(Klembara & Ruta, 2004b: 86 – see Marjanović & Laurin, 2008, for discussion of this 3551
practice) and 156 characters (many of course focused on seymouriamorphs), their matrix 3552
seems at least a priori less well suited to resolving this problem than ours. 3553
At present, the oldest known anthracosaurs date (as mentioned) from the middle 3554
Tournaisian (Anderson et al., 2015), while the oldest known temnospondyls appear only after 3555
Romer’s Gap (Balanerpeton, Caerorhachis/Casineria: Milner & Sequeira, 1994; Paton, 3556
Smithson & Clack, 1999; Ruta, Milner & Coates, 2002; no temnospondyls have been 3557
identified in the middle Tournaisian material reported by Clack et al., 2016). This is 3558
congruent with the hypothesis that Temnospondyli and the tetrapod crown-group, which is 3559
likewise currently unknown before the end of Romer’s Gap, are more closely related to each 3560
other than to Anthracosauria. However, current understanding of the diversity of limbed 3561
vertebrates during Romer’s Gap is still very poor; the current absence of evidence may turn 3562
out not to be evidence of absence. 3563
3564
The “Parrsboro jaw” 3565
3566
The position of the *Parrsboro jaw, an incomplete natural mold of a partly crushed lower jaw, 3567
is not fully resolved. However, the range of most parsimonious positions is remarkably 3568
narrow (Fig. 29). In Analyses R4–R6 the jaw either forms part of the temnospondyl/anthraco-3569
saur polytomy as described below, or it is the sister-group of one or both of the stereospondy-3570
lomorphs *Archegosaurus and *Australerpeton, or – in some of the trees from R5 and R6 3571
98
98
where Dvinosauria is nested within Stereospondylomorpha – it lies either next to the 3572
dvinosaurs Trimerorhachis and *Erpetosaurus or next to the latter alone. The latter two sets 3573
of positions require that the toothed parasymphysial bone was lost (state PSYM 1(1)) near the 3574
root of Temnospondyli and then regained (PSYM 1(0)) in the *Parrsboro jaw (see Fig. 28 for 3575
its distribution); consequently, the *Parrsboro jaw stays out of Stereospondylomorpha and 3576
Dvinosauria entirely in Analyses R10–R12, where loss of the parasymphysial was coded as 3577
irreversible. 3578
As part of the temnospondyl/anthracosaur polytomy, the *Parrsboro jaw sometimes 3579
falls out as the sister-group to all other temnospondyls, whether followed more distantly by 3580
Caerorhachis (in some MPTs from R5 and R10–R12), or together with Caerorhachis (in 3581
other MPTs from R5, R10–R12 and in B2, where (Caerorhachis + *Parrsboro jaw) has a BPE 3582
of 22 and is held together with the other temnospondyls by a BPE of 8), or alone (in some 3583
MPTs from R4); otherwise, it may lie one node more rootward than Temnospondyli (together 3584
with Caerorhachis in the remaining MPTs from R5, R6, R10–R12 and in B4: BPIE = 21), or 3585
one node more crownward than Temnospondyli together with Caerorhachis (only in R4 and 3586
the remaining MPTs from R6), or nested within the fully resolved Anthracosauria as 3587
described below. 3588
Immediately, a character comes to mind that is absent from the present matrix but 3589
would influence the position of the *Parrsboro jaw: the presence/absence of denticles on the 3590
prearticular. Presence, a plesiomorphy shared with e.g. colosteids (Bolt & Lombard, 2001), is 3591
found in the *Parrsboro jaw and in Caerorhachis but not, to our knowledge, in (other) 3592
temnospondyls or anthracosaurs. 3593
3594
3595
3596
Figure 29: The positions of the *Parrsboro jaw (NSM 987GF65 or NSM 987GH65 in 3597
different sources) and Caerorhachis (compare Fig. 28B) in otherwise simplified trees. A., 3598
Anthracosauria; T., Temnospondyli. A: positions found in all MPTs from Analysis R4 and 3599
some from R6. B: positions found in all MPTs from R5 and R10–R12 as well as some from 3600
R6. Note that, when not containing the *Parrsboro jaw, Dvinosauria often has other positions 3601
or is (in some MPTs from R5 and R6) scattered over several. D., Dvinosauria (only in some 3602
MPTs from R5 and R6). Numbers below internodes are BPE/BPIE. 3603
3604
3605
Anthracosaurian phylogeny 3606
3607
Despite many changes to the scores of all anthracosaurs, their phylogeny according to the 3608
analyses with the original taxon sample (R1–R3, R7–R9, B1 and B3) is unchanged from 3609
RC07, except that Silvanerpeton is an anthracosaur exclusively in some MPTsfrom R7–R9 3610
(see above). Bootstrap support greater than 50% exists for three nodes: Anthracosauria (BPO 3611
99
99
= 63; BPIO = 61), Embolomeri (BPO = 54; BPIO = 51) and Pholiderpeton attheyi + 3612
Anthracosaurus (BPO = 57; BPIO = 53). 3613
Adding taxa produces four different resolutions (with or without irreversibility of bone 3614
losses), only two of which are congruent with that found without the added taxa. In all of 3615
them, however, Eoherpeton remains the sister-group of all other anthracosaurs (BPE = BPIE 3616
= 20), and the smallest clade that contains Pholiderpeton attheyi and Anthracosaurus 3617
continues to exclude all other members of the original taxon sample (against a BPE of 18 and 3618
a BPIE of 20). The four resolutions are similar in that Proterogyrinus, Archeria and 3619
Pholiderpeton scutigerum form a clade (BPE = 28; BPIE = 29), a Hennigian comb or (as in 3620
RC07) a grade of two clades. *Palaeoherpeton and *NSM 994 GF 1.1 are always highly 3621
nested and close to Anthracosaurus (BPE = 23; BPIE = 24), which may be the sister-group of 3622
the latter (despite the great difference in body sizes) or of both together (which form a clade 3623
with BPE = BPIE = 28); *Neopteroplax can be the sister-group of *Palaeoherpeton or lie 3624
outside the smallest clade that contains the above and Ph. attheyi; and in the latter case the 3625
*Parrsboro jaw can come in and form the sister-group of that clade. In the bootstrap trees, 3626
*Neopteroplax is the sister-group (Anthracosaurus + (*Palaeoherpeton + *NSM 994 GF 1.1)) 3627
(BPE = 12) or of Embolomeri as a whole (which has a BPIE of 12). 3628
Like RC07, we consistently find that the anthracosaurs Pholiderpeton scutigerum and 3629
Ph. attheyi are not sister-groups (BPO = 57; BPE = 28; BPIO = 53; BPIE = 29). A logical 3630
consequence (if para- or polyphyletic genera are to be avoided) would be to reinstate the 3631
genus name Eogyrinus Watson, 1926, for Ph. attheyi. However, we refrain from performing a 3632
taxonomic act, because Pawley (2006) did find these two species as sister-groups using a 3633
different and larger character sample (but the same sample of possible anthracosaurs as RC07, 3634
minus Silvanerpeton). More characters and probably more taxa might be useful to test the 3635
phylogenetic position of Ph. attheyi, such as the less well known anthracosaurs **Pteroplax, 3636
**Calligenethlon (assuming *NSM 994 GF 1.1 does not belong to it), **Carbonoherpeton, 3637
**Leptophractus or **Aversor. 3638
We are more confident that the traditional taxon Eogyrinidae, comprising all 3639
embolomeres except Proterogyrinus, Archeria and Anthracosaurus, is paraphyletic (BPO = 3640
57; BPE = 28; BPIO = 53; BPIE = 29). These taxa were only found to form a clade in the 3641
much smaller analysis of Buchwitz et al. (2012), which was focused on chroniosuchians. 3642
3643
Temnospondyl large-scale phylogeny 3644
3645
I shall refrain here from discussion of the temnospondyls, although work being done at present, 3646
by Baird and Carroll, for example, suggests progress toward sorting out true phyletic lines 3647
among the Rhachitomi in preference to the somewhat artificial grouping which I used in my 3648
1947 classification. 3649
– Romer (1964: 154) 3650
3651
More than fifty years after this hopeful statement, temnospondyl phylogeny remains poorly 3652
understood (Fig. 30). RC07, fittingly, found the base of Temnospondyli to be a polytomy of 3653
Edopoidea ( = Edops + Cochleosauridae), Dvinosauria, Balanerpeton, Dendrerpetidae and an 3654
Eryops-Dissorophoidea clade (Fig. 30A); *stereospondylomorphs were not included in the 3655
matrix. Most of the mathematically possible resolutions of this polytomy have been found in 3656
more focused but still large analyses: Pawley (2006: chapter 5; Fig. 30B) found Edops alone 3657
as the sister-group to all other temnospondyls (except Caerorhachis, see above and below), 3658
among which a (Capetus (Dendrerpetidae, Balanerpeton)) clade, Cochleosauridae, 3659
Dvinosauria, paraphyletic *stereospondylomorphs and Eryops were successively closer to 3660
Dissorophoidea. Ruta (2009; Fig. 30C) found Edopoidea as a whole to be the sister-group to 3661
all other temnospondyls, among which Capetus, (Eryops + *Stereospondylomorpha), 3662
100
100
Dendrerpeton confusum and “D. acadianum” were successively closer to a trichotomous 3663
clade formed by Balanerpeton, Dvinosauria and Dissorophoidea; the only two 3664
*stereospondylomorphs in the matrix were *Sclerocephalus and *Glanochthon. (The OTU 3665
called “Dendrerpeton acadianum” most likely included Dendrysekos.) McHugh (2012; Fig. 3666
30D) found Neldasaurus, otherwise considered a dvinosaur, to be the sister-group of all other 3667
temnospondyls; those were divided into, on the one hand, a clade that contained the 3668
*stereospondyls and the neotenic dissorophoids inside a clade formed by the remaining 3669
dvinosaurs, and on the other hand a clade that contained (Dendrerpetidae + non-aquatic 3670
dissorophoids), (Balanerpeton + Capetus), Edopoidea and Eryops as successively closer 3671
relatives of the remaining *stereospondylomorphs. After deleting the colosteid Greererpeton 3672
from his taxon sample, Schoch (2013) recovered Edopoidea and (Balanerpeton + 3673
Dendrerpetidae) successively closer to a trichotomy of Dvinosauria, Dissorophoidea and 3674
Eryopiformes, the clade of (Eryops + Stereospondylomorpha) – when he used TNT; PAUP 3675
3.1 resolved the trichotomy, but failed to find any of the shortest trees (Schoch, 2013: 682). 3676
Using the dendrerpetid Dendrysekos as the outgroup, Eltink & Langer (2014; Fig. 30E) found 3677
the ingroup to be divided into (Eryops + Dissorophoidea) and (Dvinosauria + 3678
*Stereospondylomorpha); theirs was a *stereospondylomorph-focused analysis where 3679
Trimerorhachis was the only included dvinosaur. Most recently, Dilkes (2015a) analyzed an 3680
expanded and corrected version of Schoch’s (2013) dataset. Given the full sample of 73 3681
OTUs (Dilkes, 2015a: fig. 10), he found a polytomy of Capetus, *Iberospondylus, 3682
(Dendrerpetidae + Balanerpeton), Edopoidea and a clade which contained all other 3683
temnospondyls and was itself a polytomy of *Zatracheidae, Eryopidae, Dissorophoidea, a 3684
clade containing the *non-stereospondyl stereospondylomorphs and three separate clades of 3685
*stereospondyls. Deleting 25 OTUs, but not Greererpeton, resulted (Dilkes, 2015a: fig. 11A) 3686
in a single eryopiform clade that contained Eryops (the remaining eryopid) and the *non-3687
stereospondyl stereospondylomorphs as a grade within which a single *stereospondyl clade 3688
was nested, and also revealed a (Dvinosauria (*Zatracheidae, Dissorophoidea)) clade, but 3689
continued to resolve the relationships of these clades in two very different ways (Fig. 30F, G). 3690
Only the additional deletion of Capetus and *Iberospondylus (Dilkes, 2015a: fig. 11B) made 3691
Eryopiformes consistently the sister-group of the (Dvinosauria (*Zatracheidae, 3692
Dissorophoidea)) clade, followed on the outside by (Dendrerpetidae + Balanerpeton) and 3693
finally Edopoidea as in Schoch (2013) and in Fig. 30F. 3694
This diversity of topologies appears to result in large part from the use of different 3695
outgroups. Pawley (2006) used Caerorhachis as the outgroup in chapter 5 after finding it to 3696
be in a trichotomy with Casineria and all other temnospondyls in chapter 6 (her main analysis 3697
of the phylogeny of limbed vertebrates). Ruta (2009) included a variety of early limbed 3698
vertebrates, but entirely omitted the taxa that form the amniote total group in the trees of 3699
RC07 (see Fig. 2), including Gephyrostegidae, Anthracosauria and Caerorhachis – the closest 3700
relative of Temnospondyli in his trees is Baphetoidea. McHugh (2012) used Greererpeton as 3701
the outgroup. Schoch (2013) chose the anthracosaur Proterogyrinus;in analyses where he 3702
included Greererpeton in the ingroup for unstated reasons, Greererpeton either fell out as the 3703
sister-group to Temnospondyli or, “in some variant analyses of the large dataset (66 taxa, ‘no 3704
swapping’ option in PAUP), Greererpeton nests with the dvinosaurs, and in another variant 3705
Greererpeton and Neldasaurus form a ‘clade’ at the very base of the tree (66 taxa, ‘no 3706
swapping’ option, TNT). Whereas this reflects similarities in the postorbital skull shared by 3707
the two taxa, it also shows that aquatic taxa with particular features are sometimes attracted 3708
despite the obvious homoplastic status of these characters (indicated by multiple conflicting 3709
evidence).” This attraction between Greererpeton and Neldasaurus replicates the result of 3710
McHugh (2012), as does, to a lesser extent, the attraction of Greererpeton to the dvinosaurs as 3711
a whole. Unfortunately, however, Schoch (2013: 693) did not report the lengths of these trees; 3712
101
101
the lack of branch-swapping means that the local optima found in each tree-building replicate 3713
were not explored any further, which makes it likely that the reported trees are suboptimal. 3714
Notably, however, Greererpeton did not have such effects in the analyses of an improved 3715
version of that matrix by Dilkes (2015a). 3716
3717
3718
Figure 30: Hypotheses on temnospondyl phylogeny from 2006 to 2015. Taxa absent from 3719
our expanded taxon sample are omitted, those we added to the sample of RC07 are marked 3720
with an asterisk. Dvinosauria and Dissorophoidea are collapsed where possible. The colored 3721
rectangles (labeled in D) and the terms “Dissorophoidea (content)” and “Dissorophoidea 3722
(S13)” are consistent with Figures 2 and 10–26. The matrices of A–C are ultimately based on 3723
that of Ruta, Coates & Quicke (2003), so they – and ours (Fig. 31) – should not be considered 3724
fully independent of each other. Numbers below internodes are bootstrap percentages; those 3725
below 50 were, except in E, not published. A: strict consensus of RC07; non-temnospondyls 3726
102
102
omitted. Dissorophoidea is marked in boldface for being extant because it contains 3727
Lissamphibia; this is the only analysis shown here that included any lissamphibians. RC07 3728
(see appendix 4) performed a bootstrap analysis but did not publish the results for the part of 3729
the tree shown here. B: Pawley (2006: chapter 5). “*Cheliderpeton + *Glanochthon” was a 3730
single OTU called “Cheliderpeton spp.”; the name Glanochthon had not yet been coined. 3731
*Acanthostomatops was included as part of a Zatrachydidae OTU. C: Ruta (2009), strict 3732
consensus of analysis without reweighting; non-temnospondyls omitted. *Glanochthon 3733
latirostris was called by its previous name, “Cheliderpeton latirostre”. The “rest-dendrerpetid 3734
OTU” was called “Dendrerpeton acadianum”, but may have contained information from 3735
other dendrerpetids as well; see “Material and methods” > “Treatment of OTUs” > 3736
“Dendrerpetidae”. D: McHugh (2012); non-temnospondyl omitted. Dissorophidae (as two 3737
clades, one of which contains Broiliellus brevis) and Trematopidae (also as two clades) are 3738
nested in three different places within Amphibamidae. All bootstrap percentages of the nodes 3739
shown here are below 50 and were not published. E: Eltink & Langer (2014), focused on 3740
stereospondylomorphs. F, G: all equally parsimonious topologies (8 MPTs) from the analysis 3741
of 48 taxa in Dilkes (2015a); non-temnospondyls omitted. Note (in G) that Dendrerpetidae 3742
and Balanerpeton are sister-groups in all MPTs. To see the individual MPTs, we repeated the 3743
analysis (calculation time: 00:04:36.5), finding MPTs of the same number, length and indices 3744
as published by Dilkes (2015a). The bootstrap percentages are from Dilkes (2015a: fig. 11B), 3745
which shows an analysis where Capetus and *Iberospondylus were omitted; the topology is 3746
otherwise identical to F. 3747
3748
3749
Our analyses of the original taxon sample (R1–R3, R7–R9, B1, B3; Fig. 31A–C) 3750
consistently find Dvinosauria (which has a BPO of 28 and a BPIO of 31) as the sister-group 3751
to all other temnospondyls (BPO = 37; BPIO = 43) except, where applicable, Caerorhachis 3752
(see above). In R1, R3 and some MPTs from R2, R7 and R9 (Fig. 31A), this clade of all other 3753
temnospondyls (BPO = BPIO = 20) consists of a dichotomy of Dissorophoidea on the one 3754
side and an unusual clade on the other side in which (Balanerpeton + Dendrerpetidae), 3755
Capetus and Eryops are successively closer to a surprisingly highly nested Edopoidea. 3756
Analyses B1 and B3 find almost the same topology; the only difference is that they place 3757
(Balanerpeton + Dendrerpetidae) on the other side of the dichotomy, as the sister-group of 3758
Dissorophoidea (BPO = BPIO = 25). 3759
The remaining MPTs from Analysis R2 (Fig. 11: inset; Fig. 31B) find results more 3760
similar to those of RC07: while Caerorhachis and the dvinosaurs are positioned as above, the 3761
(Balanerpeton + Dendrerpetidae) clade is the sister-group of all remaining temnospondyls, 3762
among which Capetus, Cochleosauridae and Eryops are successively closer to a slightly 3763
differently resolved Dissorophoidea; Edops is the sister-group of either Cochleosauridae or all 3764
non-Caerorhachis, non-dvinosaurian temnospondyls except (Balanerpeton + 3765
Dendrerpetidae). Analysis R8 and the remaining trees from R7 and R9 similarly find 3766
Cochleosauridae, Edops, Eryops, Capetus, Dendrerpetidae and Balanerpeton successively 3767
closer to Dissorophoidea (Fig. 31C). 3768
Much like Pawley (2006: chapter 5; Fig. 30B), Analysis R4 and some MPTs from R6 3769
find Edops alone as the sister-group to all other temnospondyls (except, in some trees, the 3770
*Parrsboro jaw). The cochleosaurids and Eryops are successively closer to another novel 3771
dichotomy (Fig. 31D). The one side might be termed Limnarchia (see Yates & Warren, 2000), 3772
but Dvinosauria – always augmented by *Erpetosaurus, but unexpectedly also by 3773
*Nigerpeton and *Saharastega which have not been considered dvinosaurs before – is nested 3774
in various places within the stereospondylomorphs, not as their sister-group (discussed 3775
below). The other side consists of Dissorophoidea, its sister-group *Iberospondylus, their 3776
103
103
A
Isodectes
Caerorhachis
Eryops
Capetus
Edops
Cochleosaurus
Chenoprosopus
Neldasaurus
Dendrerpetidae
Balanerpeton
Dissorophoidea (content/YW00)
16/18
37/43
20/20
14/20
14/20
14/20
16/26
13/15
28/27
13/14
21/21 T.
T.
T.
T.
T.
T.
T.
T.
25/25
18/15
13/15
13/15
1/2
8/–
9/10
37/36
25/23
18/–
–/–
5/5
12/11
3/4
2/–
50/45
21/23
33/36
59/56
80/76
51/48
–/–
–/–
–/–
–/–
–/–
–/–
–/–
–/–
–/–
–/–
–/–
28/31
42/44
Trimerorhachis
B
C
Isodectes
Eryops
Capetus
Edops
Cochleosaurus
Chenoprosopus
Neldasaurus
Dendrerpetidae
Balanerpeton
Dissorophoidea (content)
Trimerorhachis
D
F G
E
Isodectes
Caerorhachis
Eryops
Capetus
Edops
Edops
Cochleosaurus
Chenoprosopus
Neldasaurus
Dendrerpetidae
Balanerpeton
Dissorophoidea (content/YW00)
Trimerorhachis
Isodectes
Eryops
Capetus
Edops
Cochleosaurus
Chenoprosopus
*Iberospondylus
*Acanthostomatops
*Palatinerpeton
*Nigerpeton
*Erpetosaurus
*Saharastega
*Saharastega
Neldasaurus
Dendrerpetidae
other stereospondylomorphs
Balanerpeton
Dissorophoidea (content)
Trimerorhachis
Isodectes
Eryops
Capetus
Edops
Cochleosaurus
Chenoprosopus
*Iberospondylus + Dissorophoidea (content) incl. *Acanthostomatops
*Acanthostomatops
*Acanthostomatops
*Palatinerpeton
Caerorhachis
*Palatinerpeton
*Nigerpeton
*Erpetosaurus
*Saharastega
Neldasaurus
Dendrerpetidae
other stereospondylomorphs
Balanerpeton
*Iberospondylus + Dissorophoidea (content) incl. *Acanthostomatops
Balanerpeton
Balanerpeton
TrimerorhachisEryops
Capetus
Edops
Cochleosaurus
Chenoprosopus
*Ib. + Diss. incl. *Ac.
*Iberospondylus
*Iberospondylus
Dissorophoidea (content)
*Ib. + Diss. incl. *Ac.
*Ib. + Diss. incl. *Ac.
*Acanthostomatops
*Acanthostomatops
*Acanthostomatops
*Palatinerpeton
*Palatinerpeton (R12)
Caerorhachis
*Palatinerpeton (R10, R12)
*Palatinerpeton
Dvinosauria
Dvinosauria
Dvinosauria
Dendrerpetidae
Stereospondylomorpha incl. *Pal.
Balanerpeton
Balanerpeton
*Ib. + Diss. incl. *Ac. (R10, R12)
*Palatinerpeton (R11)
*Ib. + Diss. incl. *Ac. (R11)
*Iberospondylus
Balanerpeton (R10, R12)
*Ib. + Diss. incl. *Ac. (R10, R11)
*Ib. + Diss. incl. *Ac. (R11)
*Ib. + Diss. incl. *Ac. (R11, R12)
Balanerpeton (R10, R11)
Dvinosauria
Isodectes
Eryops
Capetus
Edops
Cochleosaurus
Chenoprosopus
*Acanthostomatops
*Palatinerpeton
Caerorhachis
*Nigerpeton
*Erpetosaurus
*Saharastega
Neldasaurus
Dendrerpetidae
Dissorophoidea (content)
other stereospondylomorphs
Balanerpeton
*Iberospondylus
Trimerorhachis
Dvinosauria
Stereospondylomorpha
Edopoidea
–/6
4/–
↑ Figure 31: Temnospondyl phylogeny in our analyses, distribution of the intertemporal 3777
bone, and the application of the name Temnospondyli. Dissorophoidea and 3778
Stereospondylomorpha are collapsed; modern amphibians and various “lepospondyls” (see 3779
Fig. 11) are omitted, as is the *Parrsboro jaw (see Fig. 29). Filled rectangles indicate presence 3780
of the intertemporal (state 0 of character 32 – PAR 2/POSFRO 3/INTEMP 1/SUTEMP 1), 3781
empty rectangles its absence (state 1, or state 2 in the case of *Erpetosaurus; see Appendix 3782
1); the absence of a rectangle for *Palatinerpeton represents missing data. T., Temnospondyli 3783
as defined by Schoch (2013); the entire depicted clade is Temnospondyli as defined by Yates 3784
& Warren (2000). A–C: original taxon sample (numbers below internodes are BPO/BPIO); 3785
D–G: expanded taxon sample (added taxa marked with an asterisk, numbers below internodes 3786
are BPE/BPIE). A: topology found in Analyses R1–R3, R7 and R9. B: additional topologies 3787
found in R2. C: additional topology found in R7–R9; Caerorhachis is not a temnospondyl in 3788
these trees (see Fig. 28B, 29). D: topology found in R4 and R6. E: topology found in R5 and 3789
R6. The second position of Caerorhachis lies outside of Temnospondyli (as defined by Yates 3790
& Warren, 2000) and is therefore not shown. *Iberospondylus variably lies next to or inside 3791
Dissorophoidea. F: additional topology found in R5 and R6 (Fig. 15). G: topologies found in 3792
R10–R12; note that the topology of Dvinosauria depends on its position (Fig. 22). The second 3793
position of Caerorhachis lies outside of Temnospondyli and is therefore not shown. All 3794
abbreviations are spelled out elsewhere in the tree. “*Ib. + Diss.” indicates that 3795
*Iberospondylus variably lies next to or inside Dissorophoidea; in some trees from R12 where 3796
Dissorophoidea contains *Acanthostomatops, it also includes *Palatinerpeton (not shown 3797
here, see Fig. 24). 3798
3799
3800
common sister-group *Palatinerpeton and finally an unexpected clade where 3801
*Acanthostomatops, Capetus and Balanerpeton are successively closer to Dendrerpetidae. 3802
The corresponding bootstrap tree (Analysis B2: Fig. 18) differs in several respects, but the 3803
highest BPE throughout the base of the temnospondyl tree is only 16 (it belongs to 3804
Stereospondylomorpha, which excludes Dvinosauria); we suspect that the most important 3805
reason for these low values is the instability of the position of the *Parrsboro jaw (discussed 3806
above). 3807
Analysis R5 and the remaining MPTs from R6 instead recover Edopoidea and nest it 3808
close to Eryops and the “Limnarchia” assemblage, with Dissorophoidea placed in a variety of 3809
positions (Fig. 31E, F). Some MPTs from R5 and R6 fail to recover Dvinosauria; they scatter 3810
its contents across three positions (Fig. 15, 31F). 3811
Coding the losses of bones as irreversible is expected to rule out classes of topologies 3812
and thus increase resolution. Indeed, Analyses R10–R12 never find the *Parrsboro jaw highly 3813
nested within Temnospondyli (Fig. 29B). Yet, temnospondyl resolution decreases. Almost 3814
anything compatible with the backbone (Balanerpeton, Dendrerpetidae, (Capetus (Edopoidea, 3815
Eryops, *Stereospondylomorpha))) seems to be possible – with the notable exceptions that 3816
Dissorophoidea is never closer to the *stereospondylomorphs than Eryops is, and that Eryops 3817
is always closer to either *Stereospondylomorpha (Schoch, 2013; Dilkes, 2015a: fig. 11; 3818
BPIE = 6) or Edopoidea (Fig. 22–24, 31G). 3819
Dvinosauria or parts thereof being nested within Stereospondylomorpha (Analyses 3820
R4–R6; Fig. 31D–F) or Edopoidea (some trees from Analyses R10–R12; Fig. 31G) involves 3821
the long-snouted Neldasaurus and *Nigerpeton (see below) in crucial positions, apparently 3822
providing intermediates to the long-snouted cochleosaurid Chenoprosopus (Fig. 22: top 3823
middle inset) or to the long-snouted *stereospondylomorphs, especially – but not limited to – 3824
the extremely longirostrine *Archegosaurus, *Australerpeton and *Platyoposaurus. Several 3825
character states in this matrix are always but not only present in the OTUs with the longest 3826
105
105
snouts. In keeping with this, Neldasaurus is highly nested within Dvinosauria in those trees 3827
from R10–R12 where Dvinosauria lies outside of Edopoidea (and *Stereospondylomorpha), 3828
but stays close to *Nigerpeton; and in those MPTs from R5 and R6 where the dvinosaurs are 3829
scattered as three separate clades (Fig. 15), Neldasaurus and *Nigerpeton stay sister-groups 3830
and remain closest to *Platyoposaurus. 3831
The question of which taxa make up Edopoidea (only Edops, Cochleosaurus and 3832
Chenoprosopus in the bootstrap trees: BPO = 33; BPE = 18; BPIO = 36; BPIE = 15) will need 3833
to be tested by the addition of further cochleosaurids (**Adamanterpeton, 3834
**Procochleosaurus) and further characters. In this context, the dvinosaurs should not be 3835
neglected. To our knowledge, dvinosaurs and edopoids have never been considered close 3836
relatives, let alone one being nested within the other; both groups are generally considered 3837
“basal temnospondyls” that share only plesiomorphies like the intertemporal and the 3838
undivided occipital cotyle with each other (and with Balanerpeton and Dendrerpetidae). We 3839
think that the arrangement found in some MPTs from R10–R12 (Fig. 22: top middle inset), 3840
where “dvinosaurs are bottom-dwelling cochleosaurids” with *Nigerpeton as an intermediate, 3841
deserves further scrutiny; see below for additional discussion of *Nigerpeton. 3842
In Analysis R4 and some MPTs from R6 (Fig. 31D), Stereospondylomorpha and 3843
Dissorophoidea are closer to each other than to Eryops (against a BPE of 1). This is novel, but 3844
not especially likely to withstand further scrutiny; Eryops remains only one internode away 3845
from Stereospondylomorpha in this topology (two in B2) which we have not found in trees 3846
from Analyses R10–R12 or B4. 3847
We conclude that progress in resolving temnospondyl phylogeny may come from 3848
analyses that have not only dense sampling of temnospondyls and a long list of maximally 3849
independent characters, but also a large, dense sample of other taxa, because the closest 3850
relatives of Temnospondyli have not been identified. In other words, the problem of 3851
temnospondyl phylogeny cannot be solved in isolation. 3852
Schoch (2013: 689) defined the name Temnospondyli as applying to “[t]he least 3853
inclusive clade containing Edops craigi and Mastodonsaurus giganteus”. While 3854
**Mastodonsaurus is not included in our samples, this definition may be applied to our trees 3855
– as done in Fig. 31 – by assuming that **Mastodonsaurus would group with the other 3856
*stereospondylomorphsor, in their absence, with Eryops. In Analysis R4 and some MPTs 3857
from R6 (Fig. 31D), Temnospondyli would then have its usual contents, because Edops is 3858
found as the sister-group to all other traditional temnospondyls. This does not, however, occur 3859
in any of our other MPTs. In Analyses R1, R3, R10–R12, B1, B3, B4 and some MPTs from 3860
R2, R7 and R9–R12 (Fig. 31A, G), Temnospondyli would include nothing but Eryops, 3861
Stereospondylomorpha and Edopoidea (with Dvinosauria added in B2 and some MPTs from 3862
R10–R12: Fig. 31G); furthermore, Capetus, Balanerpeton and Dendrerpetidae fall outside of 3863
Temnospondyli under this definition in all inspected MPTs from Analysis R10–R12 and some 3864
more from R2 (Fig. 31B, G), as they did in some MPTs found by Dilkes (2015a: 69), and the 3865
latter two fall outside in the remaining MPTs from R2 as well (Fig. 31A). Therefore, we 3866
strongly recommend against using this definition, or any other minimum-clade definition with 3867
so few internal anchors, for this name. We have used the definition by Yates & Warren 3868
(2000), which applies the name Temnospondyli to the clade encompassing all organisms that 3869
are more closely related to Eryops than to the “microsaur” Pantylus. 3870
3871
Stereospondylomorpha 3872
3873
Throughout Analyses R10–R12 (Fig. 22–24, 31G), the stereospondylomorph temnospondyls 3874
– *Sclerocephalus, *Cheliderpeton, *Glanochthon, *Archegosaurus, *Platyoposaurus, 3875
*Konzhukovia, the stereospondyl *Lydekkerina, and finally *Australerpeton whose 3876
106
106
membership in Stereospondyli depends on how that name is defined (Eltink & Langer, 2014) 3877
– form a clade (alone – BPE = 16; BPIE = 26 – or with *Palatinerpeton, see below), even 3878
though no characters intended to resolve stereospondylomorph phylogeny were included in 3879
the matrix. In Analyses R4–R6, they are consistently found as a grade within which 3880
Dvinosauria (Fig. 13, 14, 16, 31D, E) or, in some MPTs from R5 and R6, at least 3881
*Neldasaurus is nested (Fig. 15, 31F). 3882
*Sclerocephalus or, in some trees from Analyses R10–R12, a clade formed by 3883
*Sclerocephalus and *Cheliderpeton (BPE = BPIE = 23) is consistently found as the sister-3884
group to the rest. Within this remainder, except in some MPTs from R10–R12, a clade or 3885
grade of *Cheliderpeton and *Glanochthon as well as a clade (BPE = 30; BPIE = 29) or grade 3886
of *Konzhukovia and *Lydekkerina are (in various configurations) closer to a clade (or grade, 3887
containing the dvinosaurs in some trees from Analyses R4–R6) formed by the extremely 3888
long-snouted *Platyoposaurus, *Archegosaurus and *Australerpeton; a clade of the latter two 3889
has a BPE of 44 and a BPIE of 50, the highest bootstrap values in all of 3890
Stereospondylomorpha by far. *Lydekkerina and *Australerpeton are never found as sister-3891
groups. As mentioned above, we suspect that this is due to convergence related to snout 3892
elongation; most likely we have exhausted the capability of the present character sample, and 3893
the stereospondylomorphs are sorted by noise more than by signal. Still, the position of 3894
*Sclerocephalus with respect to the others agrees – ignoring the dvinosaurs in Analyses R4–3895
R6 – with Eltink & Langer (2014), Dilkes (2015a) and arguably McHugh (2012). 3896
Cisneros et al. (2015) found Dvinosauria nested within their small sample of 3897
stereospondylomorphs. We do not interpret this result as corroborating any of our own, 3898
because it could easily be due to their small sample of non-dvinosaurs in general – or to the 3899
fact that Eryops was included in the outgroup so that Eryopiformes was forced to include 3900
Dvinosauria a priori. Furthermore, Cisneros et al. (2015) did not order any characters. 3901
3902
Dissorophoidea 3903
3904
Something of a consensus on dissorophoid phylogeny has been forming. Recent analyses 3905
(Ruta, 2009; Maddin et al., 2013b; Holmes, Berman & Anderson, 2013; Dilkes, 2015a) have 3906
generally found Dissorophidae (represented here by Broiliellus brevis alone) and 3907
Trematopidae (Acheloma, Phonerpeton, Ecolsonia, *Mordex – the latter here included in a 3908
phylogenetic analysis for the first time) as sister-groups (forming Olsoniformes Anderson et 3909
al., 2008b), and many have found the traditional “branchiosaur” Micromelerpeton to have no 3910
connection to the branchiosaurids (Apateon, Leptorophus, Schoenfelderpeton, 3911
*Tungussogyrinus, *Branchiosaurus); sometimes, Micromelerpeton is placed as the sister-3912
group to all other dissorophoids together (Maddin et al., 2013b; Dilkes, 2015a). RC07, 3913
however, upheld the more traditional topology where the trematopids lie outside a clade 3914
formed by the other dissorophoids while Micromelerpeton forms the sister-group of 3915
Branchiosauridae. Further, in the dispute over whether Ecolsonia is a trematopid or a 3916
dissorophid (e.g. Berman, Reisz & Eberth, 1985), RC07 found a sort of compromise position 3917
in that Ecolsonia came out one node closer to Broiliellus brevis and the remaining 3918
dissorophoids than to (Acheloma + Phonerpeton). 3919
This position remains a possibility for Ecolsonia in some trees from Analyses R2, R5, 3920
R6 and R10–12. The other trees, as well as all trees from Analyses R1, R3 and R7–R9, place 3921
Ecolsonia as a trematopid (BPO = 53; BPE = 37; BPIO = 52; BPIE = 38), but continue to find 3922
a clade of all non-trematopid dissorophoids (BPO = 52; BPE = 16; BPIO = 47; BPIE = 15), 3923
with exceptions in R10–R12. Within that clade, Broiliellus can be the sister-group to most or 3924
all of the rest (which has the values BPO = 37, BPIO = 33) or to Amphibamidae alone (BPE = 3925
10; BPIE = 8); the latter option is the only one in Analyses R1 and R3. Despite our best 3926
107
107
efforts to eliminate the effects of ontogeny on their scores, Micromelerpeton and 3927
Branchiosauridae remain sister-groups throughout the analyses with the original taxon 3928
sample. This does not change when taxa are added, except that Apateon moves out of this 3929
clade in some MPTs from R4, R10 and R11 and that *Tungussogyrinus is even closer to 3930
Micromelerpeton in some MPTs from R6, R10 and R11. 3931
As expected, *Mordex settles at the base of the trematopid clade (*Mordex (Ecolsonia 3932
(Acheloma, Phonerpeton))); BPE = BPIE = 26) or grade in R4–R6 and some MPTs from 3933
R10–R12. However, it nests even more rootward (sometimes next to *Iberospondylus) in the 3934
remaining trees from R10–R12. Unexpectedly, some MPTs from R11 recover a rearranged 3935
Trematopidae (Ecolsonia (*Mordex (Acheloma, Phonerpeton))) next to a clade of 3936
*Micropholis, Micromelerpeton and the “branchiosaurids” from the original taxon sample. 3937
Some trees from R12 find the other dissorophoids nested within this rearranged version of 3938
Trematopidae: (*Mordex (Acheloma, Phonerpeton)) lies next to (*Palatinerpeton 3939
(*Tungussogyrinus, *Acanthostomatops)), while Ecolsonia is the sister-group to the 3940
remaining dissorophoids. 3941
Interestingly, none of our trees, except for some from Analysis R6, find all 3942
branchiosaurids to form a clade; such a clade is contradicted by a BPE of 27 and a BPIE of 3943
23. *Branchiosaurus is sometimes (always in R4) found closer to the root of Dissorophoidea 3944
(BPE = 8), and Apateon sometimes nests with the conventional amphibamids in R4, yet 3945
Micromelerpeton always stays with the bulk of “branchiosaurids” as described above. 3946
*Tungussogyrinus, its somewhat froglike ilium (state ILI 9(1)) and Early Triassic age 3947
notwithstanding, nests with Leptorophus and/or Schoenfelderpeton in R4 and some MPTs 3948
from R5 and R6, but comes to lie in less highly nested positions in the “branchiosaur” clade 3949
or instead forms a clade or grade with *Branchiosaurus (BPE = BPIE = 15) in the remaining 3950
MPTs of these analysesand R10–R12. Some trees from R10 and R12 add the option of a 3951
sister-group relationship with Trematopidae. In short, the present matrix cannot test whether 3952
the branchiosaurids are a clade (Fröbisch & Schoch, 2009a; Maddin et al., 2013b) or a 3953
polyphyletic assemblage of larval and neotenic dissorophoids. 3954
RC07 did not determine if the conventional amphibamids in their sample 3955
(Amphibamus, Platyrhinops, Eoscopus, Doleserpeton) formed a clade with respect to 3956
Micromelerpeton and the branchiosaurids. The analyses with the original taxon sample (R1–3957
R3, R7–R9) find them as a clade (BPO = 58; BPIO = 52; also containing various modern 3958
amphibians in Analyses R3, R8 and R9 as the sister-group of Doleserpeton). This remains the 3959
case when taxa are added (BPE = 26 and BPIE = 23 for a clade that also contains 3960
*Gerobatrachus). *Micropholis, however, joins this clade only in some MPTs from R4, R6 3961
and R10–12 (often as the sister-group of *Gerobatrachus in R4, R6 and R10); elsewhere, it 3962
can be the sister-group of all other amphibamids plus Apateon (in some MPTs from R4) or of 3963
a “branchiosaur” clade (some MPTs from R5, R6 and R10–R12); more rootward positions 3964
occur in some MPTs from R10 and R11. In Analyses R4 and R10, Doleserpeton and 3965
*Gerobatrachus are sometimes sister-groups (BPIE = 40; BPIE = 35), while other trees show 3966
Doleserpeton next to Amphibamus and *Gerobatrachus next to *Micropholis; in Analyses 3967
R5, R6, R11 and R12, Doleserpeton and *Gerobatrachus are equally parsimoniously found as 3968
the sister-group of whichever modern amphibians are nested among the temnospondyls, with 3969
the other one of the two being the next more distant relative. 3970
While we are quite skeptical about a close relationship of Micromelerpeton to some or 3971
all branchiosaurids (compare with Wiens, Bonett & Chippindale, 2005), the alternative in the 3972
literature – Micromelerpetidae as the sister-group to all other dissorophoids (e.g. Schoch, 3973
2013; Maddin et al., 2013b; Dilkes, 2015a) – does not seem to be well supported either; it 3974
never occurs in our MPTs and contradicts the bootstrap trees as well (BPO = 80; BPE = 27; 3975
BPIO = 76; BPIE = 23). The name Dissorophoidea should certainly not be defined with 3976
108
108
Micromelerpeton as one of only two specifiers, as Schoch (2013) has done. That said, the 3977
older definition of Dissorophoidea by Yates & Warren (2000: 86) as all organisms closer to 3978
**Dissorophus than to Eryops (**Dissorophus being very closely related to Broiliellus) 3979
applies to a clade that contains, in the trees from some of our analyses, not only the traditional 3980
dissorophoids but also various other temnospondyls, like *Acanthostomatops, sometimes 3981
Balanerpeton and/or Dendrerpetidae and, in R4 and some MPTs from R6, even 3982
Stereospondylomorpha (Fig. 13). 3983
Schoch (2013) listed dvinosaur-like character states of micromelerpetids and used 3984
them to argue for a sister-group relationship of Micromelerpetidae to the rest of 3985
Dissorophoidea (accepted without comment by Fröbisch et al., 2015) together with a sister-3986
group relationship (or nearly so) of Dissorophoidea and Dvinosauria. A close relationship of 3987
Dvinosauria and Dissorophoidea is never found in our MPTs (the highest bootstrap 3988
percentages that contradict such a topology, however, are below 30 in all four bootstrap 3989
analyses). We wonder if the similarities listed by Schoch (2013) are instead homoplastic and 3990
indicate that the micromelerpetids are a “dissorophoid version of a dvinosaur” – a clade that 3991
generally had a dvinosaur-like lifestyle as aquatic predators with more or less anguilliform 3992
locomotion but, unlike dvinosaurs, was capable of dispersing over land in the adult stage (at 3993
least in the one species, M. credneri, that is known to have had non-neotenic, skeletally 3994
mature adults in one population: Lillich & Schoch, 2007; Schoch, 2009b). This might even 3995
explain why dvinosaurs and micromelerpetids have not been found in the same sites (see 3996
Schoch & Milner, 2014; Cisneros et al., 2015): maybe micromelerpetids were competitively 3997
excluded from bodies of water that dvinosaurs could reach, but were able to diversify in 3998
inland water bodies out of the reach of dvinosaurs. 3999
In the matrix of Dilkes (2015a), a revision and extension of that by Schoch (2013), 4000
Dvinosauria and Dissorophoidea are held together exclusively or almost exclusively by 4001
paedomorphic adaptations to an aquatic lifestyle (“Node D” in Dilkes, 2015a: 81): 4002
hypobranchials present in adults; lack of ossification of the glenoid facet of the scapula and of 4003
the carpals, pubis and tarsals; and weak torsion of the humerus – the remaining character 4004
state, jaw joints and occipital condyle(s) being at the same transverse level, may also belong 4005
here; it is in fact absent in the more mature known Micromelerpeton specimens (Boy, 1995; 4006
Schoch, 2009b: fig. 2a, b). While the sister-group of Dissorophoidea in the tree of Dilkes 4007
(2015a: fig. 13) is the terrestrial Zatracheidae (see below), followed by Dvinosauria on the 4008
outside, Dissorophoidea would be reconstructed as ancestrally aquatic because 4009
Micromelerpetidae is found as the sister-group to the other dissorophoids, among which 4010
Branchiosauridae is sister to an amphibious/terrestrial clade composed of Amphibamidae, 4011
Dissorophidae and Trematopidae; we consider this position of Branchiosauridae unlikely 4012
because of the evidence for a lack of a clear distinction between amphibamids and 4013
branchiosaurids (Vallin & Laurin, 2004; Fröbisch & Schoch, 2009a; Werneburg, 2012b; 4014
Maddin et al., 2013b; and references therein; these results are more or less congruent with 4015
some MPTs from R4–R6 and R10–R12, where such a clear distinction is likewise absent). 4016
4017
Other added temnospondyls 4018
4019
*Nigerpeton was described as a cochleosaurid (Steyer et al., 2006; Sidor, 2013). This was 4020
confirmed by Pawley (2006), Schoch (2013) and Dilkes (2015a); McHugh (2012) found it to 4021
form the sister-group of Cochleosauridae together with Edops (Fig. 30D). In our analyses 4022
(Fig. 31), it is always found as a dvinosaur (BPE = BPIE = 10), except arguably in those trees 4023
from Analyses R10–R12 where *Nigerpeton and the dvinosaurs (as sister-groups) are found 4024
as cochleosaurids, and in those MPTs from R5 and R6 where a clade of *Nigerpeton and the 4025
“dvinosaur” Neldasaurus lies within Stereospondylomorpha. Within a dvinosaur clade, 4026
109
109
*Nigerpeton is either the sister-group of Neldasaurus (BPE = 14; BPIE = 13) or the sister-4027
group of all other dvinosaurs together, among which Neldasaurus is the sister-group to the 4028
remainder; instead of the first position, *Nigerpeton emerges next to *Saharastega in some 4029
MPTs from R4, and R10–R12 allow all three positions. As mentioned above, the attraction of 4030
*Nigerpeton to Neldasaurus – and/or Chenoprosopus! – may be causally connected to its 4031
elongate snout; this remains to be tested with larger taxon and character samples. 4032
*Saharastega was described as a temnospondyl of rather uncertain position (Damiani 4033
et al., 2006). In dorsal view, its skull is unusual for a temnospondyl in some ways, leading 4034
Yates (2007) to suggest that it was a seymouriamorph. Its palate, however, is unremarkable 4035
for a temnospondyl and quite unlike that of any seymouriamorph. Pawley (2006; Fig. 30B) 4036
found it to be the sister-group of all other cochleosaurids; McHugh (2012; Fig. 30D) 4037
recovered it as the sister-group of Zatracheidae, in this matrix represented by 4038
*Acanthostomatops (see below); Schoch (2013), followed by Dilkes (2015a), explicitly 4039
excluded it as too incomplete. We, too, find it as a temnospondyl– and offer the novel 4040
suggestion that it is a dvinosaur: as *Nigerpeton is attracted to Neldasaurus, so the short-4041
snouted *Saharastega is attracted to the short-snouted Isodectes, forming a clade (BPE = 28; 4042
BPIE = 27) or a grade with it in every MPT from our analyses, except those MPTs from 4043
Analyses R4 and R10–R12 where *Saharastega and *Nigerpeton are sister-groups while 4044
Isodectes is more highly nested within Dvinosauria. 4045
*Iberospondylus is better preserved than the above two, but no less enigmatic. In both 4046
descriptions (Laurin & Soler-Gijón, 2001, 2006) it was recovered as the sister-group to a 4047
clade that contained Eryops, **Parioxys (on which see “Taxa that were not added” in the 4048
Materials and Methods section), **Zatrachys (*Zatracheidae; see below), the included 4049
dissorophoids and the stereospondylomorph *Sclerocephalus. Pawley (2006) found it as the 4050
sister-group of Dissorophoidea (together with **Parioxys); McHugh (2012) did not include or 4051
even mention it; Dilkes (2015a: 69) recovered it as an edopoid or an eryopid with a reduced 4052
taxon sample (Fig. 30F, G). Our Analysis R4, as well as some MPTs from Analyses R5, R6 4053
and R10–R12 (Fig. 31D–G), side with Pawley’s in finding *Iberospondylus as the sister-4054
group of Dissorophoidea (BPE = 18; BPIE = 15 for (*Iberospondylus + (Dissorophoidea + 4055
*Acanthostomatops))); the remaining MPTs from Analysis R5 as well as some from R6 and 4056
R10–R12 contain a clade of *Iberospondylus and either all trematopids or specifically 4057
*Mordex as the sister-group to the other traditional dissorophoids, while the remaining MPTs 4058
from R6 recover *Iberospondylus as part of a dissorophoid clade that also contains 4059
Balanerpeton, Dendrerpetidae and *Palatinerpeton (Fig. 16: inset). *Iberospondylus and 4060
Dissorophoidea share the dorsal process on the caudoventral tip of the quadrate (state QUA 4061
1(1)). In spite of this, however, the remaining trees from R10–R12 find *Iberospondylus next 4062
to a clade consisting of Capetus, Edopoidea, Eryops, Stereospondylomorpha and optionally 4063
*Palatinerpeton (Fig. 31G). 4064
*Acanthostomatops represents Zatracheidae, a clade with terrestrial adults long 4065
thought to lie close to Eryops and/or Dissorophoidea. Pawley (2006) found it next to 4066
(*Iberospondylus + Dissorophoidea), McHugh (2012) found it nested within Eryopidae 4067
together with *Saharastega, Dilkes (2015a: fig. 11; Fig. 30F, G) recovered (Dvinosauria 4068
(Zatracheidae, Dissorophoidea)) in his analyses of reduced taxon samples. In our Analysis R4 4069
and some MPTs from R6, unexpectedly, *Acanthostomatops is the sister-group of (Capetus 4070
(Balanerpeton, Dendrerpetidae)), together forming the sister-group to (*Palatinerpeton 4071
(*Iberospondylus + Dissorophoidea)). Some MPTs from Analyses R5, R6 and R10–R12 find 4072
*Acanthostomatops alone as the sister-group of (*Iberospondylus + Dissorophoidea), as does 4073
Analysis B2 (BPE = 13); others instead find a clade of *Acanthostomatops, *Tungussogyrinus 4074
and *Branchiosaurus nested next to or within Dissorophoidea, while B4 has 4075
*Acanthostomatops and Dissorophoidea as sister-groups (BPIE = 19). In some MPTs from 4076
110
110
Analyses R5, R6 and R10–R12, *Acanthostomatops gains a third position as the sister-group 4077
of Dendrerpetidae near the root of Temnospondyli; yet other MPTs from R5, R6, R11 and 4078
R12 place *Acanthostomatops next to Eryops. Evidently, *Acanthostomatops defies our 4079
character sample (Fig. 31D–G) despite being the least modified of all zatracheids. 4080
*Palatinerpeton is poorly known and has an unexpected combination of features. 4081
Unsurprisingly, then, it has never before been included in a large phylogenetic analysis. Boy 4082
(1996) presented a detailed phylogenetic hypothesis that put it close to Stereospondylomorpha 4083
and to an eryopid-*zatracheid-**Parioxys clade, and less close to Capetus; of all other 4084
temnospondyls, however, only Edopoidea was included in the investigation which polarized 4085
its characters with reference to non-temnospondyl outgroups. Schoch & Witzmann (2009a) 4086
found it as the sister-group of Eryopidae + Stereospondylomorpha, not close to 4087
*Acanthostomatops, which instead emerged as the sister-group of Dissorophoidea. Despite 4088
the dearth of data, Analysis R4 and some MPTs from R6 place *Palatinerpeton as the sister-4089
group to (*Iberospondylus + Dissorophoidea) (BPIE = 4; Fig. 31D); almost the same 4090
arrangement is found by the bootstrap analysis conducted under the same conditions (B2), 4091
where *Palatinerpeton emerges as the sister-group to (*Acanthostomatops (*Iberospondylus, 4092
Dissorophoidea)) with a BPE of 4. Analysis 5 and the remaining MPTs from R6, however, 4093
place it in a few positions close to the temnospondyl base (Fig. 31E, F). When bone losses are 4094
treated as irreversible (R10–R12; Fig. 31G), *Palatinerpeton may retain the latter positions 4095
or,surprisingly, enter Stereospondylomorpha. Positions within Dissorophoidea, too, are found 4096
in some MPTs from R10 and R12. *Palatinerpeton could thus become important for the 4097
question of temnospondyl phylogeny; a μCT scan could perhaps reveal the unexposed 4098
surfaces and fill in much of its missing data. 4099
*Erpetosaurus was originally thought to be a colosteid, but has long been recognized 4100
as an unusual dvinosaur with a few colosteid-like features and was redescribed as such by 4101
Milner & Sequeira (2011). We confirm this as discussed above under Colosteidae. Its odd 4102
attraction to the superficially quite dissimilar Trimerorhachis, however, should be 4103
investigated further; *Erpetosaurus is consistently the sister-group of Trimerorhachis (BPE = 4104
12; BPIE = 11) or only one node away. 4105
4106
Chroniosuchia 4107
4108
*Chroniosaurus has a fully resolved position (Fig. 32G, H) one node more crownward than 4109
Gephyrostegidae, Bruktererpeton or Temnospondyli (BPE = 12; BPIE = 10) and one node 4110
more rootward than Solenodonsaurus (BPE = 22; BPIE = 17). This forms a kind of 4111
compromise between the positions found in earlier phylogenetic analyses, while disagreeing 4112
with all of them: Clack & Klembara (2009) and Clack et al. (2012b) found *Chroniosaurus in 4113
a polytomy with Silvanerpeton, Anthracosauria and a “gephyrostegid”-seymouriamorph-4114
amniote-“microsaur” clade; Klembara, Clack & Čerňanský (2010) resolved it as an 4115
anthracosaur, agreeing with earlier suggestions (Laurin, 2000, and references therein) that the 4116
chroniosuchians were the geologically youngest anthracosaurs; Schoch, Voigt & Buchwitz 4117
(2010) and Buchwitz et al. (2012) found it to be the sister-group of the only included 4118
lepospondyl – which Buchwitz et al. (2012) and table 1 of Schoch, Voigt & Buchwitz (2010) 4119
stated to be the poorly known Asaphestera, while the text of Schoch, Voigt & Buchwitz 4120
(2010) mentioned Pantylus instead –; Klembara et al. (2014) recovered it as the sister-group 4121
of a gephyrostegid-seymouriamorph clade, though their bootstrap tree has it as the sister-4122
group of Gephyrostegidae alone. All these analyses relied on quite small data matrices; 4123
however, unlike us, Schoch, Voigt & Buchwitz (2010) and Buchwitz et al. (2012) included 4124
more than one chroniosuchian in their analyses (the latter’s matrix building on the former’s). 4125
111
111
A
B
C G
D
E H
Anthracosauria
Temnospondyli
D.
45
44
78
78
58
81
62
92
55
92
66
63
74
51
70
79
92
68
100
100
100
63
64
67
72
A.
D.
D.
Lepospondyli
–/–
–/–
17/17
14/14
30/22
21/16 24/16
68/68
79/78
59/71
60/60
78/82
84/75
16/–
27/–
53/49
16/18
Gephyrostegidae (composite OTU)
Westlothiana
Seymouriamorpha
(rest of) Amphibia (Lepospondyli)
Limnoscelis
Solenodonsaurus
Synapsida (composite OTU)
Diadectes
Captorhinus
SeymouriamorphaTemnospondyli
rest of Lepospondyli
Limnoscelis
Orobates
Diadectes (2 OTUs)
Petrolacosaurus
Captorhinus
Paleothyris
Westlothiana
Saur.
Seymouriamorpha
Temnospondyli
Baphetoidea
2 of 4 “lepospondyls”
2 of 4 “lepospondyls”
Limnoscelis
Tseajaia
Diadectes
Petrolacosaurus
Captorhinus
Chroniosaurus
Gephyrostegus
Bruktererpeton
Lepospondyli
Sauropsida
Solenodonsaurus
Kotlassia
Seymouria
Gephyrostegus
Bruktererpeton
Caerorhachis
“seymouriamorphs” not listed below
Amphibia (Temnospondyli)
Anthracosauria
rest of Lepospondyli
Limnoscelis
Tseajaia
Orobates
Diadectes
Petrolacosaurus
Captorhinus
Paleothyris
Westlothiana
(rest of) Anthracosauria
Colosteidae
incl. 2 of 4 “lepospondyls”
Temnospondyli
Caerorhachis
Lepospondyli?
Saur.
Saur.
Chroniosaurus
Gephyrostegus
Silvanerpeton
Bruktererpeton
Paleothyris
Seymouriamorpha
Westlothiana
Casineria
2 of 4 “lepospondyls”
F
Tseajaia
Limnoscelis
Diadectes
Orobates
(rest of) Temnospondyli/Amphibia
(extant in R2, R3, R8, R9)
Caerorhachis (R1–R3, R7, R9)
Caerorhachis (R7, R8, R9)
(Lepospondyli)
(Temnospondyli)
Gephyrostegidae
Gephyrostegidae
Saur.
Gephyrostegus (R7, R8, R9)
Kotlassia
Bruktererpeton (R7, R8, R9)
Gephyrostegus (R1–R3, R7, R9)
Bruktererpeton (R1–R3, R7, R9)
Solenodonsaurus
Paleothyris
Captorhinus
(rest of) Seymouriamorpha
Westlothiana
Petrolacosaurus
rest of Lepospondyli/Amphibia
(extant in R1, R3, R7, R9)
26/25
–/–
D.
A.
Tseajaia
Limnoscelis
Caseasauria (composite OTU)
Anthracosauria
Archaeovenator
Diadectes
Orobates
Temnospondyli (extant in R6)
Caerorhachis
Lepospondyli
Saur.
Gephyrostegus
Kotlassia (R4, R6)
Kotlassia (R6)
Bruktererpeton
Solenodonsaurus
Chroniosaurus
Paleothyris
Captorhinus
(rest of) Seymouriamorpha
Westlothiana
Leptoropha + Microphon (R6)
Leptoropha + Microphon (R4, R6)
Petrolacosaurus
rest of Lepospondyli/Amphibia
D.
together: R5, R6, R10–R12
together: R5
Diadectomorpha
–/–
–/–
–/–
–/–
–/–
–/–
–/–
–/–
–
–/–
–/–
–/–
–/–
–/–
–/–
42/32
–/–
–/–
12/10
12/10
22/17
23/17
16/12
16/12
47/43
8/8
20/14
29/26
28/24
27/25
72/71
94
97
80
76
7777
59
60
69
63
84
79
72/71
73/80
59/60
54/59
54/43
Tseajaia
Caseasauria (composite OTU)
Limnoscelis (R5)
Limnoscelis (R5)
Limnoscelis (R5, R6, R10–R12)
Diadectes
Orobates
Tseajaia
Diadectes
Orobates
(rest of) Temnospondyli/Amphibia
(extant in R5, R6, R11, R12)
Caerorhachis
(Lepospondyli)
Seymouriamorpha
Gephyrostegidae
Gephyrostegidae
Sauropsida
A.
A.
Gephyrostegus (R5, R6, R10–R12)
Chroniosaurus
Bruktererpeton (R5, R6, R10–R12)
Gephyrostegus (R10–R12)
Bruktererpeton (R10–R12)
Solenodonsaurus
Archaeovenator
Captorhinus
Seymouriamorpha (R6, R10–R12)
Seymouriamorpha (R5)
Westlothiana
Paleothyris + Petrolacosaurus
(R5, R6, R10–12)
Paleothyris + Petrolacosaurus (R5)
Paleothyris + Petrolacosaurus (R5)
Paleothyris + Petrolacosaurus (R5)
rest of Lepospondyli/Amphibia
(extant in R6, R10, R11)
8/–
–/10
↑ Figure 32: Amniote relationships and distribution of the intertemporal bone. Taxa not 4126
included in our analyses are omitted. Several named clades are collapsed. The matrices of B, 4127
D and F–H are ultimately based on that of Ruta, Coates & Quicke (2003), so they should not 4128
be considered fully independent of each other. Numbers are bootstrap percentages. Filled 4129
rectangles indicate presence of the intertemporal (state 0 of character 32 – PAR 2/POSFRO 4130
3/INTEMP 1/SUTEMP 1), empty rectangles its absence (any other state); the absence of a 4131
rectangle (only Casineria in C) indicates missing data. A., Amniota; D., Diadectomorpha; 4132
Saur., Sauropsida. A: Vallin & Laurin (2004). B: Pawley (2006: chapter 6). C: Clack et al. 4133
(2012b); all bootstrap percentages in this part of the tree are below 50 and were not published. 4134
D: RC07; “rest of Lepospondyli” excludes the adelospondyls (see Fig. 2, 28A), bootstrap 4135
values were not published. E: Klembara et al. (2014). Bootstrap percentages below 50 were 4136
not published, but note that a sister-group relationship of Chroniosaurus and Gephyrostegidae 4137
is supported by a bootstrap value of 53%. F: our analyses of the original taxon sample (R1–4138
R3, R7–R9). Numbers are BPO/BPIO. G: topology found in Analyses R4 and R6; Parrsboro 4139
jaw and Silvanerpeton omitted (see Fig. 28, 29). Numbers below nodes are BPE/BPIE. H: 4140
topologies found in our Analyses R5, R6 and R10–12; Parrsboro jaw omitted (see Fig. 29). 4141
Numbers are BPE/BPIE. 4142
4143
4144
Our analyses thus place *Chroniosaurus as the sister-group to a clade with terrestrial 4145
adults, with which it shares various adaptations to walking and bearing weight that the 4146
gephyrostegids, as well as all more rootward clades, lack (with exceptions nested within the 4147
temnospondyls). 4148
4149
Solenodonsaurus 4150
4151
Solenodonsaurus was long thought to lie particularly close to Diadectomorpha + Amniota 4152
(Laurin & Reisz, 1999; Vallin & Laurin, 2004; Fig. 32A). RC07 found it just outside the 4153
smallest clade of diadectomorphs, amniotes and “lepospondyls” (Fig. 32D); we consistently 4154
find it just outside the smallest clade formed by the above and all seymouriamorphs (which 4155
has BPO = 21; BPE = BPIO = 16; BPIE = 12; Fig. 32F–H). 4156
Danto, Witzmann & Müller (2012), who redescribed Solenodonsaurus, unexpectedly 4157
found it to be a lepospondyl instead; however, their changes to the scores of Solenodonsaurus 4158
in the matrix of Ruta, Coates & Quicke (2003) clearly contradict their own text and/or 4159
illustrations in several cases, which are listed under the respective characters in Appendix 1. 4160
4161
Seymouriamorpha 4162
4163
Uniquely, RC07 found Seymouriamorpha to be a grade, with Seymouria and Kotlassia 4164
successively closer to a (Solenodonsaurus (Lepospondyli (Amniota, Diadectomorpha))) clade 4165
than a clade formed by the remaining seymouriamorphs (Fig. 2, 32D). In all trees from R1–4166
R3 and R7–R9, all traditional seymouriamorphs except Kotlassia form a clade; in each of 4167
these analyses, Kotlassia can be the sister-group of that clade or one node more rootward than 4168
Seymouriamorpha. The corresponding bootstrap trees show (Kotlassia + Seymouria), 4169
supported by a BPO of 26 and a BPIO of 25, as the sister-group to all other seymouriamorphs 4170
(BPO of Seymouriamorpha = 26; BPIO = 22). The addition of *Karpinskiosaurus, however, 4171
rearranges seymouriamorph phylogeny and makes it harder for Kotlassia to leave: throughout 4172
Analyses R4 and R5 as well as in some MPTs from R10–R12, (Seymouria (Kotlassia, 4173
*Karpinskiosaurus)) is the sister-group to the clade formed by all other seymouriamorphs 4174
(Discosauriscidae; BPE = 29; BPIE = 23), and the remaining MPTs from R10–R12 place 4175
113
113
(Seymouria + Discosauriscidae) as the sister-group of (Kotlassia + *Karpinskiosaurus). In 4176
some MPTs from Analysis R6, however, Kotlassia is one node more rootward than 4177
Seymouriamorpha, and in a subset of these trees the (Leptoropha + Microphon) clade is one 4178
node more crownward than Seymouriamorpha, a novel and implausible arrangement. 4179
The only seymouriamorph clade with good bootstrap support is (Leptoropha + 4180
Microphon) at a BPO of 90, a BPE of 86, a BPIO of 87 and a BPIE of 84. All other 4181
percentages, including those for Seymouriamorpha as a whole, are below 40 in all four 4182
analyses. 4183
Neither the classification of Bulanov (2003), which was not based on a phylogenetic 4184
analysis, nor the results of the analysis by Klembara et al. (2014: fig. 8) are congruent with 4185
any of our results. From this and the very low bootstrap values, we conclude that at least the 4186
relationships of Utegenia, Ariekanerpeton and (Leptoropha + Microphon) to each other and to 4187
Discosauriscus remain unclear; while the matrix by Klembara et al. (2014) has a denser 4188
sampling of seymouriamorphs andof characters that are relevant to their phylogeny, our 4189
matrix has a much larger sample of non-seymouriamorphs to root the seymouriamorph tree. 4190
Curiously, Analyses R5 and B2–B4 together with some MPTs from R6 find 4191
Seymouriamorpha inside the amniote-“lepospondyl” clade, as the sister-group to the 4192
diadectomorph-amniote clade (Fig. 14, 18, 25, 26, 32H); in other words, Seymouriamorpha 4193
and Amphibia switch places compared to all other MPTs and and B1. If corroborated by 4194
future studies, the topology found by Analyses R5 and B2–B4 would amount to a reversal to 4195
the textbook hypothesis of the mid-late 20th century (which was still assumed, for unstated 4196
reasons, by Fröbisch & Schoch, 2009a: fig. 1, and by Berman, 2013). However, the sample of 4197
characters where seymouriamorphs and diadectomorphs share the same apomorphic state may 4198
already be exhaustive or nearly so in the present matrix, and the bootstrap support for this 4199
topology is low (BPE = 20; BPIO = 13; BPIE = 14; contradicting a BPO of 16). 4200
4201
Amniota and Diadectomorpha 4202
4203
Uniquely, RC07 found the diadectomorphs to form a Hennigian comb with respect to the 4204
sauropsids (Fig. 2, 32C): (Tseajaia (Limnoscelis (Orobates (Diadectes (Captorhinus 4205
(Paleothyris/Protorothyris, Petrolacosaurus)))))). This was due to numerous misscores, at 4206
least some of which were evidently caused by reliance on ancient literature about badly 4207
preserved specimens. In Analyses R1–R3 and R7–R9 we find (Fig. 32F) a monophyletic 4208
Diadectomorpha (BPO = 59; BPIO = 71; BPIE = 31), resolved as (Limnoscelis (Tseajaia 4209
(Orobates, Diadectes))) like in the latest published analyses (Reisz, 2007; Kissel, 2010; 4210
Berman, 2013), as the sister-group of Sauropsida; the clade of diadectomorphs except 4211
Limnoscelis is found in all of our analyses and supported in all bootstrap analyses (BPO = 4212
BPIO = 60; BPE = 54; BPIE = 59), and the (Diadectes + Orobates) clade enjoys some of the 4213
highest values in the entire trees (BPO = 78; BPE = 73; BPIO = 82; BPIE = 80). 4214
As mentioned (Material and methods > Treatment of OTUs > Taxa added as new 4215
OTUs for a separate set of analyses), the three amniotes included by RC07 are all sauropsids, 4216
so RC07 were unable to test whether the diadectomorphs are amniotes. We therefore added 4217
the synapsid OTUs *Caseasauria (composite) and *Archaeovenator, which robustly come out 4218
as sister-groups (BPE = 59; BPIE = 60). Surprisingly, however, diadectomorph monophyly 4219
breaks down; more precisely, throughout the analyses with added taxa, Limnoscelis never 4220
emerges as a diadectomorph (Fig. 32G, H). Analyses R4, R6, R10–12 and some MPTs from 4221
R5 find both clades as amniotes, specifically recovering Amniota as (Sauropsida 4222
(Diadectomorpha (Limnoscelis (*Caseasauria, *Archaeovenator)))); in terms of the topology 4223
found by Analyses R1–R3 and R7–R9, *Caseasauria and *Archaeovenator are not placed 4224
next to Sauropsida, but next to Limnoscelis. The remaining MPTs from R5 allow most other 4225
114
114
permutations of these four clades, additionally permitting (Paleothyris + Petrolacosaurus) to 4226
lie not only next to Captorhinus, but also next to *Archaeovenator, next to (*Caseasauria + 4227
*Archaeovenator) or next to a clade of all other amniotes. 4228
Under the conditions of Analysis R4, only one additional step is required in Mesquite 4229
to restore Amniota and Diadectomorpha as sister-groups – at the cost of rendering 4230
Captorhinus a synapsid; moving it back to Sauropsida needs one step more. Analysis B2 finds 4231
(Limnoscelis (Diadectomorpha, Amniota)), where Amniota has a BPE of 42 and 4232
(Diadectomorpha + Amniota) a BPE of 27; B4 uniquely restores Limnoscelis as a 4233
diadectomorph (BPIE = 31) and also excludes Diadectomorpha from Amniota at a BPIE of 4234
32. All this is in stark contrast to the strong support for the smallest clade that contains all of 4235
the taxa mentioned in this section: BPO = BPIO = 68, BPE = 72, BPIE = 71. The support for 4236
Sauropsida depends very strongly on the taxon sample (BPO = 79; BPE = 28; BPIO = 78; 4237
BPIE = 24), as does that for (Paleothyris + Petrolacosaurus) to a lesser extent: BPO = 84; 4238
BPE = 54; BPIO = 75; BPIE = 43. 4239
Many of the characters presented by Berman (2013) as supporting the membership of 4240
Diadectomorpha in Amniota are not parsimony-informative given his taxon sample; the 4241
others are already included in our matrix. Any satisfactory resolution of amniote origins will 4242
require a larger sample of amniotes and possibly hitherto unrecognized characters. Several of 4243
those features of the description of Tseajaia by Moss (1972) that were not revisited by 4244
Berman, Sumida & Lombard (1992) seem unusual in the light of more recent work on 4245
diadectomorphs and may warrant redescription. 4246
4247
Westlothiana 4248
4249
Despite numerous changes to its scores, Westlothiana stays (Fig. 32) in the position found by 4250
RC07 and Pawley (2006): it is the sister-group to all other lepospondyls, whatever those are, 4251
although support is low (BPO = 24; BPE = 23; BPIO = 16; BPIE = 17). 4252
4253
The lepospondyl problem 4254
4255
[…] the fact that we find throughout the Carboniferous and early [sic] Permian varied series of 4256
small, non-labyrinthodont amphibians, which I have classed as lepospondyls in a broad use of 4257
that term, presents an evolutionary problem for which we have at present no solution. 4258
– Romer (1964: 158) 4259
4260
Recent research is beginning to shed new light on the anatomy and relationships of rare and 4261
problematic forms, such as lepospondyls […] [four references]. Several issues related to 4262
lepospondyl interrelationships are likely to undergo extensive revision in the near future. 4263
Published analyses of lepospondyls reveal a disconcerting lack of agreement, to the point that 4264
almost any pattern of relationships has been proposed […] [11 references]. 4265
– Ruta, Coates & Quicke (2003: 290) 4266
4267
And what is a lepo…? 4268
– Gary Johnson, candidate of the Libertarian Party for president of the USA, in a TV interview 4269
in 2016 4270
4271
For much of the 20th century, all limbed vertebrates (or nearly so) except Lissamphibia and 4272
Amniota were divided into Labyrinthodontia and Lepospondyli. When phylogenetic analysis 4273
began to be introduced in the mid-1990s, these two taxa had largely come to be seen as 4274
wastebaskets, Labyrinthodontia basically containing the large-bodied taxa and Lepospondyli 4275
the small-bodied ones (except small temnospondyls). It was expected that Labyrinthodontia 4276
would turn out to be paraphyletic and Lepospondyli to be polyphyletic. The first prediction 4277
has held up, if perhaps less radically than expected; the latter has fared much less well, in that 4278
115
115
analyses like those of Ruta, Coates & Quicke (2003), Vallin & Laurin (2004) or most recently 4279
Maddin, Jenkins & Anderson (2012) turned up a lepospondyl clade (Fig. 33) lying not close 4280
to Temnospondyli or Colosteidae as some had surmised, but right next to Diadectomorpha + 4281
Amniota (Fig. 32), reminiscent of the old idea of a close relationship between amniotes and 4282
“microsaurs” that Romer had so thoroughly excluded in the mid-20th century. (See, however, 4283
Pardo [2014], Pardo, Szostakiwskyj & Anderson [2015a] and Pardo & Anderson [2016].) 4284
4285
4286
4287
Figure 33: Hypotheses on “lepospondyl” relationships from 2004 to 2013. Various named 4288
clades are collapsed. A., Amphibia; HPSA13, Huttenlocker et al. (2013); L., Lepospondyli; 4289
MJA12, Maddin, Jenkins & Anderson (2012); ML09, Marjanović & Laurin (2009). A: Vallin 4290
& Laurin (2004). The other position of Westlothiana lies outside Amphibia and is not shown 4291
(see Fig. 32A). NumbersMais conteúdos dessa disciplina
Evolução e Origem Vida- Adaptação Visual em Drosophila
- AD1 Diversidade dos seres vivos
- Cálculo de Carga Máquina e Homem
- A CASINHA DA VOVO Matematica - Resumo
- TAXONOMIA GERAL DOS SERES VIVOS - Resumo
- Atividade Aula 7 - Resumo
- AD2-EVOLUCAO-VITOR SOUZA DA SILVA-20214020022-DCA-LIC BIO
- AD1_Evolucao_2025-VÍTOR SOUZA DA SILVA-20214020022-DUQUE DE CAXIAS-LI CIÊNCIAS BIOLÓGICAS
- Document (4)
- Aula prática camuflagem_evolucao
- Questão 02 (ENADE-2017) Duas espécies de besouros, Trobolium castaneum e Tribolium freemani, foram utilizadas em um experimento para o estudo de ba...
- A teoria da seleção natural foi proposta por: a. Watson e Crick. b. Charles Darwin. c. Mendel. d. Aristóteles. e. Lamarck.
- a seleção sexual entre plantas ocorre por meio da competição
- Visando à arrecadação dos impostos, o processo administrativo fiscal poderá (ou não) gerar uma ação fiscal a partir das não-conformidades encontrad...
- A seleção natural é hoje considerada uma das principais explicações da diversidade biológica observada no mundo natural. Sobre os elementos essenciais
- A geometria do substrato também afeta o mecanismo de nitretação, pois as temperaturas locais das amostras podem variar, assim como a taxa de bombardea
- Seleção natural, mutações e a deriva genética são importantes fatores evolutivos que atuam sobre a variação genética das populações ao longo do pro...
- Sistema Cardiovascular
- Classificação das Proteínas
