Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Integrated approaches to understanding diversification through time using sea urchins as a model system
(USC Thesis Other)
Integrated approaches to understanding diversification through time using sea urchins as a model system
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
INTEGRATED APPROACHES TO UNDERSTANDING
DIVERSIFICATION THROUGH TIME USING SEA URCHINS AS A
MODEL SYSTEM
by
Jeffrey Robert Thompson
______________________________________________________________________________
A Dissertation presented to the
Faculty of the USC Graduate School
University of Southern California
In Partial Fulfillment of the
Requirements for the Degree
Doctor of Philosophy
in Geological Sciences
August, 2018
i
ACKNOWLEDGEMENTS
This dissertation could not have been possible without the help from my colleagues,
mentors, friends, and family all across the world. First and foremost, I need to thank David J.
Bottjer for his support, encouragement, and guidance, in science and otherwise, and for bearing
with me for these past five years. His comment to me upon one of our first meetings that getting
a job in paleontology was “like training for the Olympics” set the stage for my work trajectory
throughout my time at USC. His career advice and our discussions of the “people of science”
have been invaluable, and I look forward to many more years of collaboration. I furthermore
need to thank the support staff in the Department of Earth Sciences at the University of
California and the Southern California Earthquake Center for providing support and facilitating
this research. In particular, Cindy Waite, Karen Young, Vardui Ter-Simonian, John McRaney,
Deborah Gormley and Barbara Grubb. A special thanks is due to John Yu, for introducing me to
the world of high-performance computing, and holding my hand as I traversed this world during
the course of the many analyses I ran on the USC Geosys Cluster.
In the Department of Earth Sciences, I want to thank my qualifying exam committee
members, Frank Corsetti and Will Berelson, for providing emotional support, incisive insight,
and many intellectually stimulating conversations about science and music. Rounding out my
committee, Dave Caron and Luis Chiappe provided valuable comments and feedback during the
early stages of this project. I also want to thank Julien Emile-Geay for his course on data
analysis, which provided me the confidence and tools to implement the statistical analyses and R
scripting on which much of this dissertation is based.
I also owe a great debt to the late Eric H. Davidson of the California Institute of
Technology for giving me the opportunity to work in his lab and expand my skillset to include
ii
molecular biology. In Eric’s lab, I want to thank Feng Gao, Jane Rigg, Deanna Thomas, Isabelle
Peter, Andy Cameron, Ping Dong, Erika Vielmas, Pat Leahy, Roberto Feuda, and Jon Valencia,
for training, support, and intellectually stimulating conversation. Without a doubt the most
valuable thing I took away from my time in Eric’s lab was my friendship and ongoing
collaboration with Eric Erkenbrack. I look forward to many more years collaborating, and
hopefully many successful grant applications, as we use echinoderms to understand what
evolution is really all about.
Thanks are due to all of my friends and colleagues in the Bottjer and Corsetti labs past
and present: Joyce Yager, Olivia Piazza, Scott Perl, Kathleen Ritterbush, Lydia Tackett, Carlie
Pietsch, Becky Wu, Hank Woolley, Nate Carroll, Yadi Ibarra, Reena Joubert, Claire Johnson,
Katya Larina and Amanda Godbold. Outside of these labs, I want to thank Kirstin Washington,
Erin McParland and Pieter-Ewald Share. All of these people provided invaluable support and fun
times, making these past five years a truly enjoyable experience.
I owe a special debt to Dylan Wilmeth, my best friend, colleague, and brother-in-arms
during these five years at USC. Whether in the good times or the bad, he has been a pillar of
support, and I am excited to see where our careers take us in the coming years. Furthermore, my
good friend Laura Cotton provided support and advice throughout much of these past five years,
and without her I would not have smiled nearly as much as I did.
I additionally want to thank my friend and closest collaborator, Liz Petsios, for the many
hours of stimulating conversation we shared in the Bottjer lab. Without her as a sounding board,
I would not be near as well versed in quantitative methods, R programming, and the Permian-
Triassic mass extinction as I am today. I look forward to our future collaboration.
iii
I would also like to thank my friends, colleagues, collaborators, and reviewers from
around the world for their feedback, stimulating commentary, and help throughout these past five
years. These include Davey Wright, Lena Cole, Imran Rahman, Samuel Zamora, Bernard
Mottequin, Georgy Mirantsev, Steve Donovan, George Sevastopulo, Julien Denayer, Paola
Oliveri, Johnny Waters, William Foster, Richard Twitchett, Alexa Sedlacek, Andreas Kroh,
James Nebelsick, Tobi Grun, Louis Zachos, Chris Schneider, Andrew Smith, Yves Candela,
Laura Cotton, Tim Ewin, Shixue Hu and Jin-Yuan Huang.
Chief amongst these collaborators, my undergraduate mentor Bill Ausich has continued
to be a friend and colleague, and I am indebted to him for always being available to discuss the
minutiae of taxonomic opinions, synonymy lists, and Carboniferous stratigraphy. His scientific
guidance and philosophy has shaped my early research career and will likely shape the scientist I
aim to become.
The majority of the data underlying this dissertation was collected from Museum
collections around the world and I owe a great debt to the countless curators and collections
managers who made their collections and time available to me throughout the course of this
study. These include K. Hollis, D. Levin, M. Florence, and J. Strotman at the Smithsonian
Institute United States National Museum, Washington D. C.; S Butts and J. Utrup at the Yale
Peabody Museum of Natural History, New Haven; J. Cundiff at the Museum of Comparative
Zoology, Harvard University, Cambridge; B. Hunda at the Museum of Natural History and
Science, Cincinnati Museum Center, Cincinnati; D. Gnidovec, Orton Geological Museum, The
Ohio State University, Columbus; B. Hussaini at the American Museum of Natural History, New
York; P. Shepherd at the British Geological Survey, Keyworth; M. Riley at the Sedgwick
Museum of Earth Sciences, Cambridge; E. Howlett, Oxford University Museum of Natural
iv
History, Oxford; Y. Candela at the National Museums Scotland, Edinburgh; P. Wyse-Jackson
and G. Sevastopulo, Trinity College Dublin, Dublin; M. Parkes at the Irish Geological Survey
and National Museum of Ireland, Dublin; M. Simms at the Ulster Museum, Belfast; S.
Charbonnier at the Muséum national d’Histoire Naturelle, Paris; J. Treguier at the Musée des
Sciences de Laval, Laval; J. Cuvelier at the Université Lille, Lille; J. Denayer at the
Paleontological Collections, University of Liége, Liége; A. Folie, A. Dreze, C. Prestianni, and B.
Mottequin at the Royal Belgian Institute of Natural Sciences, Brussels; P. Van Genabeek at the
Abbaye de Maredsous: Le Centre Grégoire Fournier, Dinée; A. Kroh at the Naturhistorisches
Museum Wien, Vienna; G. Mirantsev and S. Rozhnov at the Paleontological Institute, Russian
Academy of Sciences, Moscow; C. Neumann at the Museum für Naturkunde, Berlin; and M.
Reich at the Bayerische Staatssammlung für Paläontologie und Geologie, Munich. Finally, Tim
Ewin at the Natural History Museum, London not only gave me access to the NHM’s
echinoderm collection over the course of six visits in four years, but also provided support,
friendship and many laughs over many trips to London pubs. This time in London would also not
have been monetarily feasible without help from Mark Lawmon and the “Hotel Jives”. Mark
offered me not only his hospitality, and a place to sleep rent-free, but true and long-lasting
friendship during the months spent as roommates in London.
Financial support from this work has come from the USC Earth Sciences Department, US
National Science Foundation Grant IOS1240626, the American Association of Petroleum
Geologists, the Geological Society of America, the Paleontological Society, and the
Paleontological Association.
Finally, and most importantly, I would like to thank my parents, Bruce and Wendy, and
my brother Justin for their financial and emotional support and guidance over the past five years.
v
They have supported my dreams of being a paleontologist for as long as I can remember, and
instilled in me the values of hard work, independence, and ambition from an early age. I dedicate
this dissertation to them.
vi
Table of Contents
Acknowledgements……………………………………………………………………………...…i
Abstract…………………………………………………………………………………………..xii
Chapter 1: Introduction……………………………………………………………………………1
Purpose and Significance………………………………………………………………….1
Dissertation Outline……………………………………………………………………….2
Chapter 2: The Ordovician diversification of the Bothriocidaroida (Echinodermata:Echinoidea)
with a revision of all species and a description of a new species from the Silurian of Scotland….6
Introduction………………………………………………………………………………..6
Locality Details and Geological Setting…………………………………………………..9
Material and Methods……………………………………………………………………12
Institutional Abbreviations………………………………………………………12
Computed Tomography………………………………………………………….12
Phylogenetic Analyses…………………………………………………………...13
Systematic Paleontology………………………………………………………………..16
Results……………………………………………………………………………………64
Diversity of Ordovician Bothriocidaoids………………………………………...64
Phylogenetic Analyses…………………………………………………………...64
Inclusive Dataset…………………………………………………………64
Pruned Datasets…………………………………………………………..66
Discussion………………………………………………………………………………..68
vii
Phylogenetic Relationships of Bothriocidaroids…………………………………68
The Early Diversification of Bothriocidaroids and Echinoids…………………...71
Bothriocidaroids and the Ordovician-Silurian Transition………………………..75
Biogeography of Silurian Bothriocidaroids……………………………………...77
Conclusions………………………………………………………………………………78
Figures and Tables……………………………………………………………………….79
Chapter 3: Quantitative analysis of substrate preference in Carboniferous stem group echinoids..
…………………………………………………………………………………………………..111
Introduction……………………………………………………………………………..111
Substrate preference in post-Paleozoic echinoids………………………………112
Previous hypotheses of Carboniferous echinoid environmental distribution…..114
Methods…………………………………………………………………………………116
Echinoid Occurrence Database…………………………………………………116
Analytical Methods……………………………………………………………..119
Standardized Relative Affinity…………………………………………119
Binomial Test…………………………………………………………...121
Bayesian Approach……………………………………………………..123
Results…………………………………………………………………………………..124
Binomial Test…………………………………………………………………...125
Bayesian Approach……………………………………………………………..126
Standardized Relative Affinity…………………………………………………128
Discussion………………………………………………………………………………131
viii
Substrate affinity of Paleozoic echinoids……………………………………….131
Mineralogical preferences………………………………………………131
Preferences for Grain Size……………………………………………...135
Conclusions……………………………………………………………………………..139
Figures and Tables……………………………………………………………………...141
Chapter 4: Reorganization of sea urchin gene regulatory networks at least 268 million years ago
as revealed by oldest fossil cidaroid echinoid…………………………………………………..159
Introduction……………………………………………………………………………..159
Stratigraphy and Geologic Setting……………………………………………………...161
Systematic Paleontology………………………………………………………………161
Results………………………………………………………………………………….167
Discussion………………………………………………………………………………168
Conclusions……………………………………………………………………………..172
Methods…………………………………………………………………………………172
Figures…………………………………………………………………………………..174
Chapter 5: A diverse assemblage of Permian echinoids (Echinodermata: Echinoidea) and
implications for character evolution in early crown group echinoids ………………………….178
Introduction……………………………………………………………………………..178
Geologic Setting………………………………………………………………………...180
Materials and Methods………………………………………………………………….181
Repositories and Institutional Abbreviations…………………………………...182
ix
Systematic Paleontology…………………………………………………………….…182
Discussion………………………………………………………………………………194
Co-occurrence of Stem Group and Crown Group Echinoids…………………..197
The Acquisition of Characters Leading to Crown Group Echinoids…………...198
Conclusions……………………………………………………………………………..201
Figures…………………………………………………………………………………..203
Chapter 6: Phylogenetic analysis of the Archaeocidaridae (Echinodermata: Echinoidea) and the
origin of crown group echinoids………………………………………………………………..212
Introduction……………………………………………………………………………..212
The Family Archaeocidaridae…………………………………………………..213
Previous Hypotheses for the Relationships of the Archaeocidarids and Crown
Group Echinoids………………………………………………………………..214
Methods……………………………………………………………………………..….216
Taxon Choice……………………………………………………………….….216
Phylogenetic Analyses…………………………………………………………216
Comparing the Fit of Most Parsimonious Trees to the Stratigraphic Record…..218
Results…………………………………………………………………………………..219
Phylogenetic Analyses………………………………………………………….219
Fit to Stratigraphy………………………………………………………………221
Discussion………………………………………………………………………………222
Phylogenetic Relationships of the Archaeocidarids and Late Paleozoic Echinoid
Macroecology…………………………………………………………………..222
x
Fit to the Stratigraphic Record………………………………………………….225
Origin of the Echinoid Crown Group…………………………………………..225
Conclusions……………………………………………………………………………..227
Systematic Paleontology………………………………………………………………288
Figures and Tables……………………………………………………………………...241
Chapter 7: A new stem group echinoid from the Triassic of China leads to a revised
macroevolutionary history of echinoids during the end-Permian mass extinction……………..259
Introduction……………………………………………………………………………..259
Methods…………………………………………………………………………………261
Systematic Paleontology………………………………………………………………262
Results…………………………………………………………………………………..265
Discussion………………………………………………………………………………265
Figures …………………………………………………………………………………270
Chapter 8: Paleogenomics of echinoids reveals an ancient origin for the double-negative
specification of micromeres in sea urchins……………………………………………………..276
Introduction……………………………………………………………………………..276
The Double-Negative Gate……………………………………………………..279
Results…………………………………………………………………………………..282
Holothurians appear to lack the Double-Negative Gate………………………..282
Divergence Time Estimation…………………………………………………...283
Ancestral State Reconstruction…………………………………………………285
xi
Discussion………………………………………………………………………………288
The Antiquity of the Double-Negative Gate……………………………………288
The Double-Negative Gate and Echinoid Macroevolutionary Trends…………290
“Fossilized” Gene Regulatory Networks……………………………………….292
Methods…………………………………………………………………………………292
Phylogenetic Tree Construction………………………………………………...292
Divergence Time Estimation…………………………………………………...295
Ancestral State Reconstructions………………………………………………..297
Figures………………………………………………………………………….299
Chapter 9: Conclusions…………………………………………………………………………303
The Importance of Phylogenetics in Assessing Macroevolutionary Histories…………303
Integrating Multiple Data Types to Understand Macroevolution………………………304
The Importance of Systematics…………………………………………………………305
Future work……………………………………………………………………………..305
References………………………………………………………………………………………307
xii
Abstract
Understanding the patterns and processes that shape organismal biodiversity are key to
understanding the evolution of life on earth. These drivers of biodiversity are both intrinsic to
organisms, such as the genetic regulatory networks which direct animal development, and
extrinsic like global climatic shifts and environmental change. In order to truly understand biotic
diversification through evolutionary time, a multi-pronged approach, examining both these
extrinsic and intrinsic factors is necessary. I herein set out to understand the patterns of, and
mechanisms underlying diversification and extinction in an ideal model group: sea urchins. I
focus on sea urchins from the Paleozoic and Early Mesozoic eras, which have until now been
relatively understudied. I herein present a number of phylogenetic analyses of echinoid groups,
spanning the Ordovician and Silurian to the Late Paleozoic and Mesozoic, and provide
taxonomic descriptions for a number of important echinoid taxa which provide novel insight into
the macroevolutionary history of sea urchins. The first of these analyses indicates that the initial
diversification of echinoids in the Early Paleozoic appears to have been rapid, with early
morphological diversification of numerous bodyplans and genera. Using a large database from
museum collections, I also examine the paleoenvironmental preference of echinoids during the
Carboniferous period using Bayesian and frequentist statistics. Additionally, I show that the
crown group echinoids, all of the descendants of the last common ancestor of extant echinoids,
appear to have evolved from within the diverse and stem group echinoid family the
Archaeocidaridae. By describing a new fauna from the Permian of west Texas, I show that the
archaeocidarids and the crown group echinoids are not only closely related phylogenetically, but
also co-existed in some of the same environments in the Permian period. The oldest definitive
crown group echinoid, Eotiaris guadalupensis, also occurs in this fauna, which indicates that
xiii
many of the morphological changes differentiating the two clades of crown group echinoids, the
cidaroids and euechinoids, and their associated genetic regulatory underpinning, appears to have
diverged by at least 258 million years ago. The entire echinoid stem group has also long been
thought to have gone extinct during the end-Permian mass extinction, 252 million years ago.
Using a robust phylogenetic analysis of stem group and crown group echinoids, and a description
of a new species from the Triassic of China, I show that in fact at least one lineage of stem group
echinoids did survive the P-T mass extinction, and that in fact the stem group echinoid did not go
extinct at the end of the Permian. Finally, by using phylogenetic comparative methods, fossil-
calibrated divergence time estimation, and comparative analysis of gene regulatory networks
(GRN) in extant echinoids, I show that the Double Negative Gate, a key network subcircuit
specifying the micromeres of many euechinoid echinoids, is an ancient evolutionary novelty, at
least as old as the Triassic period. Through these analyses, I developed a novel approach to
rigorously identifying the evolutionary age of particular GRN novelties, which provides a
timeline for identifying the rate of GRN evolution. This multi-disciplinary approach to
understanding diversification highlights the importance of holistic analyses of diversification,
including both extrinsic and intrinsic factors.
1
Chapter 1. Introduction
PURPOSE AND SIGNIFICANCE
Throughout geologic and evolutionary time, patterns of metazoan diversification and
morphological diversity have waxed and waned (Alroy, 2010; Alroy et al., 2008). The processes
responsible for shaping this diversity, both morphological and taxonomic, are varied. In order to
gain a holistic understanding of the evolution of the history of life on Earth , it is important to
understand the intrinsic properties shaping organismal morphologies, such as developmental
gene regulatory networks and phylogenetically derived bauplans, and those factors extrinsic to
organisms, such as their ecology and the evolution of the broader Earth system. Echinoids, or sea
urchins provide an exceptional group of organisms in which to examine the interplay of these
extrinsic and intrinsic factors which have shaped their evolutionary history. I will therefore set
out to test hypotheses regarding the timing, setting, and developmental processes underlying the
morphological and taxonomic diversification of echinoids in the Paleozoic and Early Mesozoic
Eras.
Sea urchins, or echinoids are a diverse and abundant group of marine organisms in both
modern oceans and in the fossil record. They are keystone taxa and ecosystem engineers in many
of the environments they inhabit, and are found across numerous habitats from shallow seas to
the abyssal depths of the oceans. Echinoids have an excellent fossil record, which makes them an
ideal group of organisms to examine taxonomic diversity through time. In addition to their fossil
record, the genomic underpinning of echinoid development is amongst the best understood for
any organism (Davidson et al. 2002, Oliveri et al. 2008). This combination of an excellent fossil
record and a deep understanding of echinoid developmental gene regulatory networks makes
2
them the ideal model system to study the evolution of gene regulatory networks in the context of
deep time.
Furthermore, the Paleozoic and Early Mesozoic are a relatively understudied interval of
time with regard to sea urchin diversification. Most previous work has simply consisted of
taxonomic descriptions and until now, little analytical work has been done teasing apart the
phylogenetics, diversification trajectories, and paleoecology of echinoids in the Paleozoic and
early Mesozoic. By using Paleozoic and early Mesozoic echinoids as a model, I have been able
to speak to larger macroevolutionary questions such as the effects of mass extinctions, relative
timing of diversification events, the effect of environmental shifts on diversification and
extinction, and the relationships between environmental changes and molecular and genetic
evolution.
DISSERTATION OUTLINE
In order to gain an understanding of the varying patterns and processes underlying
organismal diversification, I have taken an interdisciplinary approach emphasizing systematics,
phylogenetics, taxonomy, paleoecology, mass extinctions and paleogenomics. Because the
phylogenetic relationships of many of the Paleozoic echinoids are not well constrained, I will
attempt to understand the relationships of many of these taxa and how those relationships speak
to the history of their diversification. One of these case studies (Chapter two) will focus on the
bothriocidaroid echinoids, a morphologically diverse clade of stem group echinoids, and the first
group of sea urchins to obtain significant levels of diversity in the geological past. Through
detailed taxonomy and parsimony and Bayesian phylogenetics, I have shown that each of the
genera of the bothriocidaroids forms a clade. Furthermore these clades achieved high levels of
3
morphological diversity early in their history, indicating that sea urchins as a whole were
morphologically diverse early in their evolutionary history.
While the bothriocidaroids achieved their diversity in the Ordovician, the Paleozoic peak
of echinoid diversity took place in the Carboniferous Period (Kier, 1965; Smith, 1984), which is
the focus of chapter three. This peak of echinoid diversity also coincides with a peak in
Phanerozoic carbonate accumulation, and thus the Carboniferous is an ideal interval of time in
which to examine the relationship between environmental shifts and echinoid paleoecology. In
order to understand paleoenvironmental preference in different clades of Paleozoic echinoids, I
use frequentist and Bayesian statistical approaches to test for preference in substrate affinity. I
found that the differential tolerances and affinities of different echinoid clades for particular
environments may have controlled the varying macroevolutionary history of these clades in the
Late Paleozoic.
Chapter four details the implications of a newly discovered Permian cidaroid echinoid,
Eotiaris guadalupensis, the oldest verifiable member of the crown group. This species has
important implications for understanding the timing of the origination of the echinoid crown
group, and the gene regulatory networks which shape the adult body plan of crown group
echinoids. Furthermore, this specimen was crucial for paleogenomic analyses of echinoid gene
regulatory networks detailed in Chapter eight.
Chapter five builds on Chapter three, and examines the other members of the echinoid
assemblage that co-occurred with Eotiaris guadalupensis. These taxa include two stem group
echinoids, one proterocidarid and one archaeocidarid, and suggests for the first time that
echinoids from the stem group and the crown group were co-occurring in the same environments
during the Paleozoic. Furthermore, the archaeocidarid species from this assemblage has crenulate
4
tubercles, indicating that this morphological feature, long associated with the crown group,
actually originated in the stem group echinoids, and was fairly widespread in the Permian.
The most diverse and abundant group of echinoids in the Paleozoic were the
archaeocidarids, which have long been thought to be the sister group to the crown group
echinoids. The archaeocidarids are the focus of chapter six. In order to understand the
relationships of the different genera and species of the archaeocidarids, and to uncover the
origins of the echinoid stem group, I ran a species level analysis of archaeocidarid and Paleozoic
crown group echinoids. This analysis elucidated many of the morphological innovations
associated with the origin of the stem group echinoids, and furthermore demonstrated that the
archaeocidarids are likely paraphyletic with respect to the echinoid crown group.
In chapter seven, I provide a revised interpretation for the macroevolutionary history of
echinoids over the Permian-Triassic mass extinction interval. The end-Permian mass extinction
represents the most significant loss of metazoan life in the Phanerozoic, and was particularly
devastating for echinoids. It has long been thought that the entire echinoid stem group went
extinct during the end-Permian event, however, my work has shown that this is not the case.
Using a newly described fossil from the Triassic of China, and detailed phylogenetic analyses, I
have shown that stem group echinoids did survive the end-Permian extinction, and went on to
populate the seas of the Triassic.
In chapter eight, I applied a paleogenomic approach to understand the evolutionary origin
of a gene regulatory network which specifies the micromeres in many crown group sea urchins.
In doing so, I developed a framework for rigorously determining the evolutionary age of
evolutionary novelties, which can be built on to determine the rates at which gene regulatory
networks evolve. I found that this particular gene regulatory network subcircuit likely originated
5
at the youngest in the Triassic period, and that it cannot be definitively associated with the
divergence of the euechinoids and the cidaroids, as has long been thought.
6
Chapter 2. The Ordovician diversification of the
Bothriocidaroida (Echinodermata:Echinoidea) with a revision of
all species and a description of a new species from the Silurian
of Scotland
INTRODUCTION
The Ordovician marks an exceptionally important interval of Earth’s history with the Great
Ordovician Biodiversification Event leading to a substantial rise in taxonomic richness (Miller,
1997; Miller and Foote, 1996; Miller and Mao, 1995; Sepkoski, 1988, 1993) and ecological
expansion (Droser and Finnegan, 2003; Miller and Connolly, 2001; Servais et al., 2009; Servais
et al., 2010) of metazoans associated with the diversification of the Paleozoic evolutionary fauna
(Sepkoski, 1981). Amongst echinoderms, the Ordovician represents the peak in taxonomic
diversity of blastozoans (Foote, 1992; Nardin and Lefebvre, 2010) and stylophorans (Lefebvre et
al., 2006) and the Phanerozoic peak diversity of echinoderm classes (Sprinkle, 1980). Though
the utility of taxonomic estimates of diversity over this interval has been criticized (Smith, 1988),
estimates of morphological diversity show a similar pattern, with Ordovician peaks in disparity
of blastozoans (Foote, 1992) and stylophorans (Lefebvre et al., 2006). There is therefore little
doubt that the Ordovician represents an important interval of geologic time with respect to the
diversification of echinoderms.
Slightly later in time, the Ordovician-Silurian transition is characterised by one of Earth’s
“Big Five” mass extinction events (Finnegan et al., 2012; Raup and Sepkoski, 1982; Sheehan,
7
2001). It has been associated with faunal turnover and extinction in a number of echinoderm
groups including crinoids (Ausich and Peters, 2005; Borths and Ausich, 2011; Deline and
Ausich, 2011; Donovan, 1988; Eckert, 1988), and blastozoans (Nardin and Lefebvre, 2010). This
extinction is widely thought to be linked to glaciation and global cooling (Finnegan et al., 2011;
Finnegan et al., 2012), though recent work has also demonstrated an abundance of aberrant,
teratological morphologies in fossil microplankton (acritarchs, chitinozoans) caused by heavy
metal poisoning linked to widespread oceanic anoxia (Vandenbroucke et al., 2015). This is
coincident with the global, positive δ
13
C excursion that marks the onset of extinction (Delabroye
et al., 2012; Munnecke et al., 2012). Because of these numerous changes in the Early Paleozoic
Earth system, a keen understanding of the taxonomic dynamics of echinoderms from the
Ordovician to Silurian is necessary to understand the macroevolutionary history of echinoderm
clades later in the Paleozoic.
Echinoids are amongst the most recognizable echinoderms, and are ubiquitous members
of many post-Paleozoic ecosystems ( e.g. Kier and Grant, 1965; Nebelsick, 1996; Smith, 1984).
Although they are more commonly known from Mesozoic and Cenozoic-aged rocks, the fossil
record of echinoids spans back to the Ordovician (Pisera, 1994; Smith and Savill, 2001). Though
they are much less abundant than their post-Paleozoic kin, the Paleozoic echinoids comprise the
great majority of the echinoid stem group. Stem group echinoids were rare but often present
members of many Paleozoic communities, especially during the Carboniferous and Permian
(Schneider et al., 2008; Thompson and Ausich, 2016; Thompson and Denayer, 2017; Thompson
et al., 2017b). Amongst the oldest echinoids are those belonging to the Order Bothriocidaroida
Eichwald, 1860 (Männil, 1962; Paul, 1967; Pisera, 1994). The bothriocidaroids displayed thick,
tessellated plates arranged into a variable number of columns of interambulacral and ambulacral
8
plates and had small spines and a primitive Aristotle’s lantern. They are unique amongst known
echinoids in that many species retained “plated” tubefeet (Männil, 1962), which contained
embedded calcitic ossicles similar to the spicules embedded in the podia of ophiocistioids
(Haude and Langenstrassen, 1976). Though since their initial discovery, there has been much
debate regarding their taxonomic affinities, with hypotheses ranging from cystoid (Mortensen,
1928a) to echinoid (Jackson, 1912) to stem-holothuroid (Smith, 1984), cladistic analyses now
convincingly show bothriocidaroids to be amongst the most basal stem group echinoids (Smith
and Savill, 2001).
The bothriocidaroids are known from the Pygodus serra conodont zone (Darriwilian,
Middle Ordovician; Cooper and Sadler, 2012; Pisera, 1994) to the Ludlow (Silurian; Reich and
Smith, 2009) and were the first clade of echinoids known to achieve fairly widespread
biogeographic distribution, and to exhibit relatively high degrees of taxonomic diversity
(comprising fourteen named species and three genera). In the Ordovician, the bothriocidaroids
occupied marine environments in Baltica and Laurentia, with a single occurrence from a peri-
Gondwanan terrane represented by deposits in the Carnic Alps (see recent review in Lefebvre et
al., 2013). Three genera of bothriocidaroids are known, Bothriocidaris Eichwald 1860,
Neobothriocidaris Paul 1967, and Unibothriocidaris Kier, 1982. Between these three genera, the
bothriocidaroids comprise over seventy-five percent of known echinoid species from the
Ordovician, and represent much of what is known regarding the initial diversification of
echinoids (Fig. 2.1). Though the bothriocidaroids are more abundant in the Ordovician, their
record in the Silurian remains less well known, and taxonomic dynamics leading up to their
eventual extinction remains unclear. Until now, Silurian bothriocidaroids were only known from
Llandovery to Ludlow strata in Gotland Sweden (Franzén, 1979; Kutscher and Reich, 2001,
9
2004; Reich and Smith, 2009). In addition to being poorly known from the Silurian, the
phylogenetic relationships of species within the bothriocidaroids have yet to be examined with
modern analytical techniques.
We herein undertake the first species level phylogenetic analysis of the Bothriocidaroida
and report on a new species, Neobothriocidaris pentlandensis n. sp. from the Telychian of the
Pentland Hills, Scotland. This is the first Silurian bothriocidaroid known from outside of Baltica,
and provides important new information regarding their global distribution during the Silurian.
Furthermore, we utilize micro-CT scanning to better visualize the morphology of this new
specimen, and report on its morphology in greater detail than otherwise possible. We
additionally assembled a database of bothriocidaroid specimens and species (available from the
author), based on occurrences recorded from the previously published literature. In our
phylogenetic analyses, we employed parsimony and Bayesian inference, to attempt to understand
the phylogenetic relationships within the bothriocidaroids to elucidate the details of their
diversification in the Ordovician and Silurian.
LOCALITY DETAILS AND GEOLOGICAL SETTING
The new specimen of Neobothriocidaris pentlandensis n sp. described here was collected
in the North Esk Inlier (NEI) (Fig. 2.2), which is part of the easternmost group of inliers located
in the Midland Valley terrane (MV). The NEI is the largest of the three Silurian inliers in the
Pentland Hills. It is located about 2.5 km NNW of the town of Carlops, which lies 13 km SSW of
the city of Edinburgh. The inlier forms a 6.8 km
2
elongate area, stretching NE-SW encompassing
the River North Esk and the North Esk Reservoir. Exposure is reasonably good but is often
restricted to streams and gullies. Within the inlier, the sedimentary beds are highly inclined to
10
vertical, striking in a NE–SW direction and younging to the NW (Clarkson and Taylor, 2007).
The NEI is the most abundantly fossiliferous inlier and although found in vertical beds the fossils
remain remarkably undistorted. The Silurian succession is 1500 m thick. The inlier is ringed by a
greywacke-conglomerate of early Old Red Sandstone age. However, a fault truncates the
northwestern part of the inlier and throws down Lower Carboniferous sediments (Clarkson and
Taylor, 2007).
The specimen was donated to the National Museum of Scotland (then Royal Scottish
Museum) in the mid 1970s, and associated data were limited. Data consisted of locality details
given as ‘head of the Gutterford Burn, Pentland Hills’; no stratigraphical data accompanied the
specimen. However, comparison of the lithology type with other selected specimens from the
NEI in the National Museums Scotland’s Paleobiology collection shows that it may have been
collected from the ‘Starfish Beds’ in the lower Deerhope Formation, which is the source of
Aptilechinus caledonensis Kier, 1973, until now the only known echinoid in the NEI. The
Deerhope Formation stratigraphically overlies the poorly fossiliferous Reservoir Formation. It is
390 m thick and consists mainly of fine-grained clastic sediments (Clarkson and Taylor, 2007).
Along the Gutterford Burn section, Peach and Horne (1899) have described two fossiliferous
beds, within what was then considered to belong to the top of their Bed A (Reservoir Formation),
characterised by “well-preserved specimens of starfishes” (op. cit. p. 593). These horizons have
since been included in the Deerhope Formation by Anderson et al. (2007). The Gutterford Burn
section of the Deerhope Formation is characterised by an important concentration of arthropods
within a horizon termed the ‘Eurypterid Bed’ exposed on the right bank of the Gutterford Burn
(locality R191) and first discovered by John Henderson, then curator of the Phrenological
Museum, Edinburgh, in 1880. The Deerhope Burn section exposes the younger beds of the
11
Deerhope Formation, including the ‘Deerhope Coral Beds’. These have revealed a rich, diverse
and exquisitely preserved fauna found in a number of bands in a yellow-brown siltstone
occurring within a sequence of fine green mudstones (Clarkson and Taylor, 2007). Clarkson and
Taylor (2002) described in great detail the faunal succession.
The Silurian succession in the NEI represents phases of continuous marine regression
(Clarkson, 2000), from the deep-water Reservoir Formation to the shallow-water, lagoonal
setting of the Whether Law Linn Formation, which has yielded abundant and relatively well-
preserved fossil faunas. At the top of the succession, the Henshaw Formation was deposited in a
semi-arid desert, however, high in the formation, thelodont scales and crinoid columnals have
been found, which testify to a brief return of marginal marine conditions (Clarkson and Taylor,
2007). In this depositional context, the Deerhope Formation represents a shallower water setting
than the underlying Reservoir Formation, however accumulated below wave base.
The Deerhope Formation is considered to be of Late Telychian (Late Llandovery) age
(Bull and Loydell, 1995; Clarkson et al., 2001; Robertson, 1989) but no precise level in the
graptolite biozonation has ever been proposed. Bull and Loydell (1995) described graptolites
from the underlying Reservoir Formation which they assigned to the middle part of the spiralis
biozone. Loydell (2005) described a very small and poorly preserved graptolite assemblage, from
sampling between localities R236 and R247 along the Deerhope Burn, from the Deerhope
Formation but noted a couple of important and better preserved taxa. He described
Diversograptus ramosus (locality R236), which is restricted to the middle Oktavites spiralis –
lower Cyrtograptus lapworthi biozones, and Oktavites possibly of the spiralis group, which is
not known from above the middle of the lapworthi biozone. Therefore Loydell (2005) suggested
an age between the middle spiralis and the middle lapworthi biozones, which implies a Late
12
Telychian (Late Llandovery) age. As noted by Molyneux et al. (2008), the study of the
palynological assemblage of the Deerhope Formation concurs with the interpretation based on
graptolite evidence.
MATERIALS AND METHODS
Institutional Abbreviations
Institutional abbreviations where specimens are housed are as follows: TMM: Texas Memorial
Museum collection housed at the Non-vertebrate Paleontology Lab at the University of Texas at
Austin; PIN: Paleontological Institute of the Russian Academy of Sciences, Moscow; UI and
SUI: University of Iowa Paleontology Repository, Iowa City; FGWG: FR Geowissenschaften,
Ernst-Moritz-Arndr-Universität Greifswald, Germany; GIT: Institute of Geology at Univeristy
of Tartu, Estonia; HM: Hunterian Museum, Glasgow; NMHUK: Natural History Museum,
London; NMS G: National Museum of Scotland, Edinburgh. GBA: Geologische Bundesanstalt
Vienna; ZPAL: Institute of Paleobiology, Polish Academy of Sciences. MBE: Museum für
Naturkunde, Berlin. PM SPU: Paleontological Museum, Institute of Earth Sciences, St.
Petersburg State University
Computed Tomography
Micro-computed tomography (Micro-CT or X-ray microtomography) was carried out using a
Bruker SkyScan 1172 micro-CT scanner at Naturalis Biodiversity Center, the Netherlands. The
two halves of specimen NMS G.1976.48.1, which is an external mould, were aligned and held
together using masking tape for the scanning process. 902 projections of 1020 ms exposure were
collected with a 2000x1336 detector, using a source voltage of 81 kV, and current of 124 µA. An
13
aluminium (0.5 mm) was employed, and the scan provided a reconstructed dataset with 13.17 µm
voxels. Digital visualisation was achieved by segmenting of the test and spines, followed by
surface rendering in Avizo.
Phylogenetic Analyses
For our phylogenetic analyses, we utilized a sequential workflow inspired by that of Ausich
(1998a, 1998b), where we first analysed a larger dataset encompassing more taxa and characters,
and then analysed subsets of this dataset which were resolved from the initial analysis using a
smaller set of characters. These datasets are referred to the “inclusive dataset” and “pruned
datasets” respectively. We utilized this methodology because many of the characters in the initial
analysis of the inclusive dataset were coded as non-applicable due to the reduction and absence
of interambulacral plates in Unibothriocidaris and Neobothriocidaris, and the absence of
imperforate ambulacral plates in Bothriocidaris. Because PAUP* and MrBayes treat non-
applicable characters as unknown, by using a pruned list of characters on smaller subsets of the
dataset, we reduced the ratio of characters scored as unknown to that of known characters in the
dataset.
Our initial analyses were run using a matrix of fifteen taxa and twenty-six characters.
Seventeen characters were binary while nine were multistate, and all characters were unordered
and equally weighted. A list of taxa included in the analysis and their stratigraphic ranges are
shown in Table 2.1 and the character matrix and list of characters are available from the author.
Analyses using the maximum parsimony optimality criterion were run in PAUP* (Swofford,
2003). Bromidechinus rimaporus Smith and Savill, 2001 was used as an outgroup, following its
placement as sister group to bothriocidaroids or sister group to all other echinoids in previous
14
cladistic analyses (Smith and Savill, 2001). Eighteen of the characters were parsimony
informative while eight were not. Tree searching was conducted using stepwise addition and
1000 random addition replicates with tree-bisection and reconnection (TBR). Bootstrap
proportions for nodes on the resulting strict consensus trees were evaluated using 10000 “fast”
bootstrap replicates. Analyses were also run with characters reweighted by their consistency
indices (CI), retention indices (RI) and rescaled consistency indices (RC).
Bayesian methods have recently been show to outperform parsimony-based phylogeny
reconstruction for discrete morphological data (Wright and Hillis, 2014), albeit at the expense of
phylogenetic precision (O’Reilly et al., 2016). Bayesian analyses were thus additionally run
using MrBayes 3.2 (Ronquist et al., 2012). We utilized the Mk model of Lewis (2001). In order
to account for ascertainment bias prevalent in the coding of morphological characters, such as
the exclusion of constant characters, character coding in MrBayes was set to variable to reflect
our exclusion of constant, but not parsimony uninformative, characters. A gamma distribution
was used to model rate variation among characters with prior of an exponential distribution with
parameter 1.0 on α. The prior used on unconstraind branch lengths was an exponential
distribution with parameter 1.0. Wright et al. (2016) recently demonstrated the importance of
allowing for asymmetrical rates of character change in morphological datasets using a symmetric
Dirichlet prior on character state frequencies. A symmetric discrete beta distribution with a
single shape parameter α is thus used as either a prior (binary characters) on state frequency or
hyperprior (multistate characters) with larger values of α corresponding to more symmetrical
transition rates among characters and smaller values of α resulting in more asymmetry (Wright
et al., 2016). Results using all six utilized values of α: ∞, 10, 2, 1, 0.2 and 0.05 are presented.
The posterior distribution of trees and branch lengths was estimated using Markov chain Monte
15
Carlo (MCMC). MCMC consisted of two runs with four chains per run for 1200000 generations.
Chains were sampled every 500 generations with the first 25% generations discarded as burn in.
Convergence was assessed using the average deviation of split frequencies, and chains were run
until this value was less than 0.01. Results of Bayesian analyses are presented as 50% majority
rule consensus trees, with clade credibility values, representing posterior probabilities (PP) of
clades shown at resolved nodes.
Based upon the results of these analyses, further analyses utilizing maximum parsimony
and Bayesian inference were run using two pruned datasets; one of which included only species
of Bothriocidaris, and one which included all species of Neobothriocidaris and
Unibothriocidaris. Parsimony analyses of both of these datasets were rooted on the outgroup,
Bromidechinus rimaporus. Characters which, upon pruning of taxa, were predominantly non-
applicable or constant, were removed from data matrices. Due to the small number of taxa,
exhaustive searches were employed using the maximum parsimony optimality criterion
(Alltrees) in PAUP*. This ensured that all of the most parsimonious trees which existed for each
data matrix were found. Bayesian analyses used the same priors and MCMC settings as were
utilized for the analysis of the inclusive dataset including all bothriocidaroids. All six considered
values of α are presented for both datasets. The dataset consisting of species of Bothriocidaris
included nine taxa and nineteen characters, which were a subset of the twenty-six used in
analysis of all taxa. Twelve characters were binary and seven were multistate and all were treated
as unordered and equally weighted. The data matrix for this analysis is available from the author.
For analyses of species of Neobothriocidaris and Unibothriocidaris, seven taxa and seventeen
characters were utilized in analyses. Thirteen of these characters were binary while four were
16
multistate and all characters were treated as unordered and weighted equally. This data matrix is
also available from the author.
SYSTEMATIC PALEONTOLOGY
Class Echinoidea Leske, 1778
Stem Group Echinoidea
Order Bothriocidaroida von Zittel, 1879
Diagnosis. Stem-group echinoids with small test. Plating thick, tessellate. Ambulacra arranged
into two columns or more, dependent upon the genus (Fig. 2.3). Some genera with an
imperforate column of plates separating each half ambulacra of perforate plates (Fig. 2.3C, D).
Plates irregularly hexagonal (Fig. 2.4). Ambulacral plates with single or double perforation per
tube foot (Fig. 2.5). Ambulacral pores are arranged diagonally on plates. Primary ambulacral
tubercles on small, hollow, raised platform. These tubercles are non-crenulate. In most species
these tubercles are merged with a peripodium surrounding the ambulacral pores. In all genera,
the radial water vessel is within the test, though in one genus, Neobothriocidaris, the lateral
vessels are enclosed within the coronal plates. In all taxa in which the apical system is known,
there are five radially located ocular plates (Fig. 2.4), which are imperforate and tuberculate. One
of these bears a series of furrowed indentations in the plate, which is interpreted to be the
madreporite. Numerous small periproctal plates are present at the most adapical pole of the test,
while five larger elongate periproctal plates are present interradially, and may bear tubercles. A
rudimentary Aristotle’s lantern is present in taxa in which a lantern has been found (see
17
Diagnosis of Bothriocidaris). Spines are small and striate, with non-crenulate acetabula and
without a milled ring.
Occurrence. Bothriocidaroids are known from the Middle Ordovician, (Darriwilian, Pygodus
serra conodont zone; Pisera, 1994) to the Ludlow, Silurian (Reich and Smith, 2009). They are
known from strata in North America, Estonia, Scotland, Poland, Gotland, Sweden, Norway and
glacial erratic boulders representing Baltic strata found across the north of Europe. They are also
possibly known from the Austrian Carnic Alps.
Remarks. The phylogenetic position of the order Bothriocidaroida within the echinoderms has
been arguably the most hotly debated topic within the study of Paleozoic echinoids over the
course of the past 150 years (Clark, 1932). The first known bothriocidaroid, Bothriocidaris
globulus Eichwald, 1860 was described from the Katian of Estonia. Jaekel (1894) acquired
another specimen of Bothriocidaris globulus Eichwald, also from the Katian of Hiiuma, Estonia
which very clearly displayed the structure of its adoral and adapical ends, showing the structure
of the peristomial plating and apical disc. He chose to retain bothriocidaroids within the
Echinoidea, although as a very primitive member, based on the presence of what he thought were
teeth (now recognized to be large radial plates bordering the peristome) and the placement of the
mouth and anus at ventral and dorsal ends of the animal respectively.
Jackson (1896) treated Bothriocidaris as the most basal echinoid. At the time of his
description, Bothriocidaris was known only from species with two columns of plates in each
ambulacrum and a single column of plates in each interambulacrum. Jackson’s justification for
the phylogenetic position of Bothriocidaris was mostly developmental. In early post-
18
metamorphosis development of extant echinoids, a single interambulacral plate and two
ambulacral plates are generated in each zone of the corona (Gao et al., 2015; Gordon, 1927,
1928; Lovén, 1892; Zachos and Sprinkle, 2011). That each interambulacral area in B. globulus
and B. pahleni were composed of a single column of plates, and that the earliest added
ambulacral plates were arranged into two columns, was Jackson’s (1896) justification for
treating Bothriocidaris as the most basal echinoid. Agassiz (1904) also commented on the adoral
ambulacral structure of B. globulus, comparing the two rows of adoral ambulacral plates to those
of juvenile echinothurioids. Later, Jackson (1912) continued to use his developmentally
informed line of reasoning to treat Bothriocidaris as the most basal echinoid.
The first to align Bothriocidaris with diploporite cystoids was Yakowlew (1922)
however, he still chose to treat Bothriocidaris as an echinoid, favoring two independent origins
and a polyphyletic nature of the group. The echinoid nature of Bothriocidaris was then
questioned by Mortensen (1928a), who argued that the genus was related to diploporite cystoids
and belonged within the Cystoidea. The position of the madreporite in Bothriocidaris is radial, as
opposed to interradial in all other echinoids, and is located on one of five large apical plates
interpreted to be oculars. The position of the madreporite, plus the radial position of the plates
that Jackson (1912) regarded as teeth, which in other echinoids are interradial (Gao et al., 2015;
Fig. B2), led Mortensen (1928a) to reject Bothriocidaris as not only the most primitive echinoid,
but an echinoid at all. This sparked a flurry of responses from leading echinodermologists of the
time. Hawkins (1929) vehemently disagreed with Mortensen, though pointed out that although
Mortensen had (in Hawkins’ view) correctly identified that the plates previously classified as
teeth were not in fact homologous with the teeth of other echinoids, but that no suitably
homologous structures had not yet been identified did not mean that true teeth were not present.
19
Jackson (1929b) additionally agreed that the plates he had formerly regarded as teeth, were in
fact non-ambulacral (radial) peristomial plates, but that Mortensen (1928a) failed to produce any
convincing evidence for the treatment of Bothriocidaris as a cystoid. Furthermore, Bather (1931)
also rejected Mortensen’s assertions that Bothriocidaris was a cystoid on the grounds that it
lacked food grooves and facets for brachiole articulation and Clark (1932) agreed with Jackson’s
developmentally informed argument that Bothriocidaris was the most ancestral echinoid..
Mortensen (1930), however, continued to treat bothriocidaris as a cystoid.
Männil (1962) provided important observations on the morphology of new specimens of
Bothriocidaris including the first description of the bothriocidaroid jaw apparatus. These new
specimens showed that the plates previously interpreted by Jackson (1912) to be teeth, were in
fact peristomial plates (as demonstrated by Mortensen, 1928a) and that the true teeth were
located interradially as in all other echinoids. The discovery of a true lantern provided the final
nail in the coffin that was necessary to disprove the diploporite cystoid hypothesis of Mortensen
(1928a, 1930) and Yakowlew (1922). Philip (1965), Kier (1966) and Durham (1966) agreed with
Männil (1962) and treated bothriocidaroids as echinoids based on the presence of a rudimentary
lantern. In discussing the phylogenetic placement of bothriocidaroids within the echinoidea, he
placed them as sister group to the Palaechinoida. He did not place them as the most basal
echinoids, choosing instead the Megalopoda (equivalent with the now utilized family
Eothuriidae, Macbride and Spencer, 1938), which includes only the genus Eothuria Macbride
and Spencer, 1938.
Smith (1984) chose to treat the bothriocidaroids as eleutherozoans, however, based on the
presence of a perradial ambulacral column, internal radial water vessels, reduction or loss of
interambulacra and the presence of plated tube feet similar to those present in some
20
ophiocistioids (Haude and Langenstrassen, 1976), he chose to treat bothriocidarioids and
ophiocistioids as stem group holothurians. With the discovery of the echinozoan Bromidechinus
Smith and Savill, 2001, Smith and Savill (2001) utilized cladistic analysis to tease apart the
relationships of the most basal echinoids, along with the Ophiocistioid Gillocystis Jell , 1983 and
rooted on the edrioasteroid Stromatocystites Pompeckj, 1896. Their analyses found the
bothriocidaroida to be monophyletic with fairly high bootstrap support (85%) in their four most
parsimonious trees. More importantly, they also demonstrated through the use of cladistic
methodology that the bothriocidaroids were in fact, stem group echinoids, as opposed to stem
holothurians. We follow Smith and Savill (2001) and thus treat the bothriocidaroids as stem
group echinoids.
The cladistic analysis of Smith and Savill (2001) also provided the first quantitative
analysis of the phylogenetic relationships of the genera within the Bothriocidaroida. Their
resultant phylogenetic topologies showed Bothriocidaris to be sister group to a clade comprised
of Neobothriocidaris+Unibothriocidaris. Although this analysis was key in estimating the
phylogenetic relationships of the genera within the Bothriocidaroida, and the placement of the
Bothriocidaroida as stem echinoids, many bothriocidaroid species were excluded from the
analysis, and the phylogentic relationships of all species of the Bothriocidaroida have, until now,
never been analysed using modern phylogenetic methods. The results of our parsimony and
Bayesian analyses are shown in Results.
Family Bothriocidaridae Klem, 1904
21
Type genus. Bothriocidaris Eichwald, 1860
Other genera. Neobothriocidaris Paul, 1967 and Unibothriocidaris Kier, 1982
Diagnosis. As for order Bothriocidaroida.
Occurrence. As for order Bothriocidaroida.
Genus Bothriocidaris Eichwald, 1860
Type species. Bothriocidaris globulus Eichwald, 1860
Other species. Bothriocidaris pahleni Schmidt, 1874, Bothriocidaris parvus, Männil, 1962,
Bothriocidaris eichwaldi Männil, 1962, Bothriocidaris solemi Kolata, 1975, Bothriocidaris
maquoketensis Kolata, Strimple, Levorson, 1977, Bothriocidaris kolatai Kier, 1982,
Bothriocidaris vulcani Guensburg, 1984.
Diagnosis. Bothriocidaroid with two columns of perforate ambulacral plates in each ambulacral
area at ambitus and without imperforate column of ambulacral plates. Each ambulacral plate
bears two pore-pairs, which are contained entirely within each plate. Ambulacral plates bearing
from two to six primary tubercles dependent upon the species. In all species with a peripodial
ring, these tubercles are merged with the peripodium. Ambulacral pores in most species
surrounded by peripodial ring. Interambulacrum ranging from one to three columns of regular or
22
irregularly plated polygonal interambulacral plates (Fig. 2.3A, B). Interambulacral plates are
with or without primary tubercles dependent upon the species, and most species bear small
secondary perforate or imperforate tubercles on their ambulacral and interambulacral plates. The
interambulacra are separated from the interradial plates bordering the peristome by one row of
ambulacral plates (Fig. 2.4). In the apical system above each ambulacral zone lies one large
radially located tuberculate and imperforate plate (Fig. 2.4). These plates are interpreted to be
ocular plates, though their homology to the ocular plates of the echinoid crown-group has been
doubted (Smith and Savill 2001). One of these plates bears a furrowed structure, which is the
madreporite. Located adapically of the interambulacra in each interradial series is an elongate
plate which in some specimens is in contact with the most adapical interambulacral plate, but
which in others is separated from the interambulacral plates by the ocular plates. Adapical of
these elongate plates and ocular plates, are numerous small periproctal plates
Most-oral plating of the corona consists of ten radially arranged ambulacral plates. Each
pair of two radially-most oriented plates is in continuity with more adapical columns of
ambulacral plates (Fig. 4). Situated adorally to these plates and radially between each pair of
ambulacral plates is a single radial plate, each being slightly smaller than the ambulacral plates.
There total five of these plates which are in turn separated from contact with each other by five
smaller and more elongate interradial plates (Männil 1962; Fig. 15). Adoral to these interradial
plates are ten smaller subinterradial plates, two of each which lie immediately adoral to each
interradial plate. Adoral of these radial and interradially located plates are numerous small,
conical, peristomial plates. The lantern of Bothriocidaris consists of rudimentary hemipyramids,
each of which is elongate and cylindrical in nature distal from the adoral end. The adoral end of
the hemipyramids curve such that the hemipyramids meet and form a circular process at the
23
aboral end of each pyramid. The teeth are small and appear to end in a pointed, nonserrate tip
proximal to the adoral end of the test. The spines of Bothriocidaris are small, striate. The tube
feet of many specimens of this genus bear spicules such that calcified tube feet are visible
protruding from the pore-pairs in many specimens.
Occurrence. Articulated specimens are known from the Sandbian and Katian of Estonia (Männil
1962) and the Katian Kiln Mudstone Member, Craighead Limestone Formation of Craighead
Girvan, Scotland. In North America, articulated specimens are known from the Katian Fort
Atkinson Formation of Iowa, the Sandbian Bromide Formation of Oklahoma (Kier 1982),
Lebanon Limestone of Tennessee (Guensburg 1984), and Walgreen Member of the Grand
Detour Formation of the Platteville Group of Illinois (Kolata 1975). Articulated specimens are
also known from Ordovician glacial erratic boulders in Germany (Rhebergen, 1990) and the
Netherlands (Eggink, 1991).
Disarticulated plates also occur, “throughout the Grand Detour Formation in [Northern
Illinois and Southern Wisconsin]” (Kolata 1975). Disarticulated plates, spines and pedicellariae
are known from a Katian or Hirnantian Öjlemyr Flint glacial erratic boulder on Gotland, Sweden
(Nestler, 1968), and disarticulated plates are known from Late Ordovician glacial erratics of the
Öjlemyr Flint and Backsteinkalk of Gotland and the Gdansk area (Schallreuter, 1989). In the
Silurian, disarticulated Bothriocidaris plates occur from the Ludlow Hemse beds of Gotland
(Kutscher and Reich, 2001, 2004).
Remarks. Eichwald named two species, B. globulus, Eichwald, 1860 from the Katian of the
island of Hiiuma, Estonia (referred to by Eichwald as Dago) and B. exilis Eichwald, 1860 from
24
Talkhof. B. globulus is a small tessellately plated echinoid with interambulacral columns in each
area arranged in a single column, and ambulacral areas arranged into two columns. B. exilis is,
according to Jackson, “A small unrecognizable mass” (Jackson 1912; Pg. 453), and is strictly
indeterminate, being treated herein as neither bothriocidaroid nor echinoid. B. globulus bears
pore-pairs surrounded by a raised peripodial ring on which lie four spine-bearing perforate
tubercles. A second species was described by Schmidt (1974), B. pahleni Schmidt, 1974 from the
Sandbian of Estonia, which differs from B. globulus in having two spine bearing perforate
tubercles as opposed to four. Jackson (1912) then erected a new species, B. archaica Jackson,
1912 based on differences in the apical plating between the specimen published by Jaekel (1894)
and other specimens of B globulus.
Männil (1962) revised the genus, following Mortensen (1928) in synonymizing Jackson’s
Bothriocidaris archaica Jackson, 1912 with Bothriocidaris globulus Eichwald, 1860 and
erecting two new species, Bothriocidaris eichwaldi Männil. 1962 and Bothriocidaris parvus,
Männil, 1962 from the Katian of Estonia. He also published the first phylogenetic hypothesis of
the species-level relationships within Bothriocidaris, with B. globulus as sister taxon to B.
parvus, and this clade sister taxon to B. eichwaldi. This clade of ((B. globulus + B. parvus) + B.
eichwaldi) descended from a lineage that earlier split from the older species B. pahleni (Männil
1962; Text fig. 22). Nestler (1968) expanded upon Männil’s (1962) phylogenetic hypothesis to
include his Bothriocidaris sp. n., aff. pahleni from the Katian Öjlemyr Chert of Gotland. Nestler
proposed the same topological relationship as Männil, however, he treated his species from
Gotland as a direct descendent of the older B. pahleni.
Bothriocidaroids were unknown from Laurentian strata in North America until Kolata
(1975) described three taxa from the Sandbian and possibly Katian of Northern Illinois,
25
Bothriocidaris solemi Kolata, 1975, Neobothriocidaris tempeltoni Kolata, 1975, and an unknown
genus and species (herein treated as Unibothriocidaris sp.). Kolata et al. then described a new
species of Bothriocidaris, B. maquoketensis, Kolata et al. 1977 from the Katian Fort Atkinson
Formation of Northeastern Iowa. This taxon differs from described Bothriocidaris from Baltica
in having all interambulacral areas composed of more than one column of interambulacral plates,
with rows composed of either one or two plates. Bothriocidaris vulcani Guensburg, 1984 from
the Sandbian Lebanon Limestone also has more than one column of interambulacral plates,
however, the shape of the interambulacral plates are irregular in this taxon, and the plating does
not occur in well-defined columns. Kier (1982) described two bothriocidaroids from the
Sandbian Pooleville member of the Bromide Formation of Oklahoma, including a species of
Bothriocidaris, B. kolatai Kier, 1982.
Bothriocidaris globulus Eichwald, 1860
1860 Bothriocidaris globulus Eichwald: 655, pl. 32, figs 22, 23
1868 Bothriocidaris globosus Eichw; Bigsby: 27.
1874 Bothriocidaris globulus Eichw; Schmidt: 36, 38, 40, pl. 4, figs 2A-E.
1879 Bothriocidaris globulus Eichw.; Zittel: 481.
1881 Botriocidaris globulus Eichw.;Neumayr: 10.
1883 Bothriocidaris globulus Eichwald; Pomel: 117.
v.1894 Bothriocidaris globulus Eichwald; Jaekel: 244, 245, 247, 249, 250, 254, figs 1,2.
1904 Bothriocidaris globulus Eichw; Agassiz: 80.
1904 Bothriocidaris globulus Eichwald; Klem: 15.
1910 Bothriocidaris globulus Eichwald; Lambert & Thiéry: 118.
26
v.1912 Bothriocidaris archaica Jackson: 28, 34, 42, 45, 52, 53, 62, 64, 69, 79, 87, 88, 148, 167,
173, 210, 239, 240, 244, 446, pl. 1, figs 1, 2, text-figs 2, 22, 40, 162.
1912 Bothriocidaris globulus Eichwald; Jackson: 34, 87, 88, 2102, 239, 240, 243, pl. 1, figs 7-9.
1920 Bothriocidaris archaica; Hawkins: pl. 61, fig. 1.
1922 Bothriocidaris globulus; Yakowlew: 327, fig. 1.
1922 Bothriocidaris archaica; Yakowlew: 329.
1928 Bothriocidaris globulus; Mortensen: 94, 97-99, 102, 107, 122, figs 3.2, 3.3, 4-6, 10.3.
1928 Bothriocidaris archaica; Mortensen: 94, 97, 98.
1929 Bothriocidaris archaica Jackson; Jackson: 482, 483, 491, 492, 495, 506-510, figs 1, 2.
1929 Bothriocidaris globulus Eichwald; Jackson: 485, 491, 506-509.
1930 Bothriocidaris globulus; Mortensen: 325, 326, 331, fig. 20.
1930 Bothriocidaris archaica Jackson; Mortensen: 325, 326, 331.
1931 Bothriocidaris globulus; Bather: 58.
1931 Bothriocidaris archaica; Bather: 58.
1938 Bothriocidaris globulus Eichwald; MacBride & Spencer: 93.
1962 Bothriocidaris globulus Eichwald; Männil: 143,152, 159, 160, 163, 164, 183, 185, 187,
189, text-figs 10.
1962 Bothriocidaris archaica Jackson; Männil: 160, 163, 164, 189.
1963 Bothriocidaris globulus Eichwald; Müller: fig. 641.
1966 Bothriocidaris globulus Eichwald; Durham: 372.
1966 Bothriocidaris globulus; Kier: U301.
1967 Bothriocidaris globulus Eichwald; Paul: 525, 527, 528, 531, pl. 85, figs 1-4.
1968 Bothriocidaris archaica Jackson; Raup: text-fig. 1D.
27
1969 Bothriocidaris globulus Eichwald; Nestler: 1224, fig. 5.
1975 Bothriocidaris globulus Eichwald; Kolata: 66.
1977 Bothriocidaris globulus Eichwald; Kolata et al.: 146, 149.
1978 Bothriocidaris globulus Eichwald; Müller: fig. 649.
1984 Bothriocidaris globulus Eichwald; Guensburg: 67.
1982 Bothriocidaris globulus Eichwald; Kier: 314.
1989 Bothriocidaris globulus Eichwald; Schallreuter: 9, pl. 7, figs 4D.
1990 Bothriocidaris globulus Eichwald; Rhebergen: 50, 51, figs 4D, 5.
1996 Bothriocidaris globulus Eichwald; Donovan et al.: 260, pl. 51, fig. 4.
1998 Bothriocidaris globulus Eichwald; Zuidema: fig. 4.
2001 Bothriocidaris globulus Eichwald; Smith & Savill: 138, fig. 2.
2009 Bothriocidaris globulus Eichwald; Solovjev: 1415, 1418, fig. 3D.
Type specimen. The Holotype is specimen number PM SPU 1627/1..
Diagnosis. Modified from Männil (1962). Bothriocidaris with four perforate, non-crenulate
tubercles on each ambulacral plate merged with a raised, well-developed peripodium. Numerous
imperforate secondary tubercles cover ambulacral plates, which are less than twice as wide as
high. Interambulacral plates with imperforate secondary tubercles and one, or rarely more,
perforate, non-crenulate, primary tubercles. Interambulacral plates are hexagonal and arranged
into a single, regular, column (Fig. 2.4).
28
Occurrence. The holotype was described from Katian outcrops on the Island of Hiiuuma,
Estonia. On Hiiuma, this species is known from the Katian outcrops at the Kõrgessaare and
Paluküla quarries. It is also known from the Kiln Mudstone Member of the Craighead Limestone
Formation, Ardwell Subgroup, Ardmillan Group at the Kilns in the Main Quarry at Craighead,
Girvan, Scotland.
Remarks. This was the first bothriocidaroid discovered (Eichwald, 1860). It bears four primary
tubercles on ambulacral plates, which is more than all species of Bothriocidaris except for B.
solemi. The tuberculate interambulacral plates also make this taxon similar to B. parvus, B.
solemi, B. maquoketensis, B. vulcani, and B. kolatai. The majority of specimens have a single
column of plates in each interambulacra (Fig. 4), similar to B. parvus, B. solemi, B. kolatai and
B. pahleni though Mortensen (1928) noted irregularities in the interambulacra of B. globulus
which are similar to the multi-columned ambulacra present in B. vulcani and B. maquoketensis.
Though the majority of specimens bear a single column, this aberrant specimen shows interesting
similarities between these three taxa. Männil (1962) and Nestler (1968) placed B. globulus as
sister group to B. parvus in their initial phylogenetic hypotheses. None of our phylogenetic
analyses could confidently resolve the position of B. globulus.
Bothriocidaris pahleni Schmidt, 1874
v.1874 Bothriocidaris Pahleni Schmidt: 1, 36, 38, pl. 4, figs 1A-G.
1879 Bothriocidaris Pahleni Schmidt; Zittel: 481, figs 339A-C.
1881 Botriocidaris Pahleni Schmidt; Neumayr: 10, pl. 1, fig. 6.
1883 Bothriocidaris Palhenii Schmidt; Pomel: 117.
29
1883 Botryocidaris Pahleni Fred. Schmidt; Lovén: 57.
1885 Bothriocidaris Pahleni; Schmidt: 97.
1892 Bothriocidaris Pahleni; Agassiz: pl. 29, fig. 1.
1894 Bothriocidaris Pahleni Schm; Jaekel: 244, 246, 249, 253, 254.
1896 Bothriocidaris pahleni Schmidt; Jackson: fig. 4.
1900 Bothriocidaris Pahleni Schmidt; Gregory: fig. 14.
1904 Bothriocidaris phaleni Schmidt; Klem: 15.
1910 Bothriocidaris Pahleni Schmidt; Lambert & Thiéry: 118.
1911 Bothriocidaris Pahleni; Mickwitz: 150, 163, fig. 18.
1912 Bothriocidaris pahleni Schmidt; Jackson: 34, 69, 77, 87, 88, 210, 239, 240, 242, 244, pl. 1,
figs 3-6, pl. 8, fig. 1.
1922 Bothriocidaris Pahleni; Yakowlew: 327.
1929 Bothriocidaris Pahleni; Mortensen: 94, 95, 97-99, 102, 103, 109, 121, 122, figs 1A, 2, 3.1,
8.
1929 Bothriocidaris pahleni Schmidt; Jackson: 483, 485, 491, 492, 500, 506-509, fig. 8.
1930 Bothriocidaris Pahleni; Mortensen: 314, 316, 317, 322, 324, 325-327, 331, 340, 351, figs
1-5, 12-14.
1931 Bothriocidaris pahleni; Bather: 58, 60.
1938 Bothriocidaris pahleni Fr. Schmidt; MacBride & Spencer: 93.
1962 Bothriocidaris pahleni Schmidt; Männil:143, 144, 147, 152, 155, 156, 161, 163, 164, 167,
170, 183, 185, 187, 189, pl. 1, figs 1-5, pl. 2, figs 1, 2, text-figs 3A, 4, 13, 14.
1963 Bothriocidaris pahleni F. Schmidt; Müller: figs 640, 642-645.
1966 Bothriocidaris pahleni Schmidt; Durham: 372.
30
1967 Bothriocidaris pahleni Schmidt; Fletcher et al.: 586.
1967 Bothriocidaris pahleni Schmidt; Paul: 225, 531.
1969 Bothriocidaris pahleni Schmidt; Nestler: 1223, 1224, fig. 5.
1970 Bothriocidaris pahleni Schmidt; Rõõmusoks: 230, 237, 241, 247.
1975 Bothriocidaris pahleni Schmidt; Kolata: 66.
1977 Bothriocidaris pahleni Schmidt; Kolata et al.: 149.
1978 Bothriocidaris pahleni F. Schmidt; Müller: figs 648, 650-653.
1982 Bothriocidaris pahleni Schmidt; Kier: 314.
1984 Bothriocidaris pahleni Schmidt; Guensberg: 67.
1984 Bothriocidaris pahleni; Smith: fig. 9.5.
1989 Bothriocidaris pahleni Schmidt; Schallreuter: 4, 5, 7, 9, pl. 6, figs 2, 3I-K, 4A.
1990 Bothriocidaris pahleni F. Schmidt; Rhebergen: 50-52, figs 1, 4A, 5, 6, 7, 8, 9.
1990 Bothriocidaris pahleni Schmidt; Schallreuter: 126.
1994 Bothriocidaris pahleni Schmidt; Eggink: 7, figs 1-7.
2001 Bothriocidaris pahleni Schmidt; Smith & Savill: 138, 146, figs 1, 2, 6C, 7.
2009 Bothriocidaris pahleni Schmidt; Solovjev: 1415, 1418, figs 3A.
Type specimen. The holotype is specimen number PIN 2039/1.
Diagnosis. Modified from Schmidt (1874) and Männil (1962). Bothriocidaris with two closely
spaced, perforate, and non-crenulate primary tubercles merged with a slightly raised peripodium
on each ambulacral plate. Ambulacral plates are less than twice as wide and covered with
numerous imperforate secondary tubercles. Interambulacral plates are hexagonal and arranged
31
into a single, regular column. Interambulacral plates covered with numerous imperforate
secondary tubercles. No primary tubercles present on interambulacral plates.
Occurrence. This species is known from Sandbian outcrops at Kukruse, Alliku, Rae, Nõmmise
and the Sandbian aged Kahula Formation at Aluvere Quarry, Estonia. Additionally, specimens
have been found in Ordovician aged glacial erratic boulders at Wilsumer Berge, Germany, and
Westerhaar, Twente, The Netherlands. Disarticulated plates attributed to this species have been
found in Backsteinkalk glacial erratics at Gotland Sweden, and in the Gdansk Region.
Remarks. This species is differentiated morphologically from B. eichwaldi only in the spacing
of its primary tubercles, with B. eichwaldi having more widely spaced primary tubercles on its
ambulacral plates. Männil (1962) chose to treat B. pahleni as the sister group to a clade
comprised of all other species of Bothriocidaris. Nestler (1968) then expanded upon Männil’s
hypothesis by treating B. pahleni as the direct ancestor of his B. sp. n., aff. Pahleni, which was
found in a younger Katian Öjlemyr Chert glacial erratic boulder. Our analyses did not resolve the
phylogenetic placement of this species.
Bothriocidaris sp. aff. pahleni Schmidt, 1874
1968 Bothriocidaris sp. aff. pahleni Schmidt; Nestler: 1224, figs 1-5.
Material. FGWG 46.
Description. This material consists of a collection of coronal plates, spines, and pedicellariae
that was described in detail by Nestler (1968). We will herein only treat the coronal plates. The
32
ambulacral plates collected display two morphologies. The first consists of ambulacral plates
with two perforate, non-crenulate primary tubercles merged with the peripodium and spaced
about the width of the peripodial ring apart. The second ambulacral plate morphology bears only
a single perforate, non-crenulate primary tubercle. The lone interambulacral plate collected bears
a single perforate, non-crenulate primary tubercle, and no secondary tubercles.
Occurrence. The material was collected from a Katian Öjlemyr glacial erratic boulder from
Visby, Gotland, Sweden.
Remarks. Nestler (1968) described this material as Bothriocidaris sp. n., aff. pahleni Schmidt,
1874. Though the material is similar to B. pahleni in bearing two primary ambulacral tubercles, it
is greatly differentiated from that species in bearing ambulacral plates with a single primary
tubercle, and in having interambulacral plates with a primary tubercle. This is the only known
Bothriocidaris with ambulacral plates bearing a single tubercle, and as such is very likely a
species separate from all other known bothriocidaroids. Because we have not seen this material
in person, we do not attempt to name it, though we agree with Nestler (1968) that this is a
species distinct from all other Bothriocidaris.
Bothriocidaris parvus Männil, 1962
1962 Bothriocidaris parvus Männil: 152, 156, 164, 165, 167, 168, 170, 183, 185, 187, 189, pl. 2,
figs 3, 4, pl. 3, fig. 5, text-figs 3B, 8, 9.
1966 Bothriocidaris parvus Mannil; Durham: 372, 373, fig. 4.
1968 Bothriocidaris parvus Männil; Nestler: 1224, fig. 5.
33
1975 Bothriocidaris parvus Männil; Kolata: 66.
1977 Bothriocidaris parvus Männil; Kolata et al.: 146, 149.
1978 Bothriocidaris parvus Männil; Kier & Lawson: 2.
1982 Bothriocidaris parvus Männil; Kier: 314.
1989 Bothriocidaris parvus Männil; Schallreuter:12, 13, pl. 8, fig. 4C.
1990 Bothriocidaris parvus Männil; Rhebergen: 50, 51, figs 4C, 5.
2001 Bothriocidaris parvus Männil; Smith & Savill: 138, fig. 2.
2009 Bothriocidaris parvus Männil; Solovjev: 1415, 1418, fig. 3C.
Type specimen. The holotype is specimen GIT 106-7.
Diagnosis. Modified from Männil (1962). Bothriocidaris with three perforate, non-crenulate
tubercles on each ambulacral plate. Tuberlces merged with raised, well-developed peripodial
rim. Ambulacral plates are less than twice as wide. Interambulacral plates bear a single perforate,
non-crenulate, primary tubercle. Interambulacra are arranged into a single column of regularly
arranged, hexagonal plates.
Occurrence. The holotype is from Katian strata at Lehtse, Lääne-Viru County, Estonia (Männil,
1962). Schallreuter (1989) has also recovered disarticulated plates of this species from Katian
Öjlemyr glacial erratic boulders on Gotland, Sweden.
Remarks. This species of Bothriocidaris is the only known taxon with three primary tubercles
on each ambulacral plate. The holotype is weathered and thus the presence or absence of
secondary tubercles is unknown. Männil (1962) and Nestler (1968) postulated that this taxon was
34
the sister species to Bothriocidaris globulus, however none of our phylogenetic analyses could
confidently resolve its position.
Bothriocidaris eichwaldi Männil, 1962
1962 Bothriocidaris eichwaldi Männil: 152, 156, 164-167, 169, 170, 183, 185, 187-189, pl. 1,
figs 6-9, pl. 3, figs 1-4, pls 4, 5, text-figs 3C, 6, 7, 11, 12, 14A-E, 15-17.
1967 Bothriocidars eichwaldi Männil; Paul: 539.
1966 Bothriocidaris eichwaldi Mannil; Durham: 372, 373 figs 2, 3, 5.
1967 Bothriocidaris eichwaldi Männil; Fletcher et al.: 586.
1968 Bothriocidaris eichwaldi Männil; Nestler: 1224, fig. 5.
1975 Bothriocidaris eichwaldi Männil; Kolata: 66.
1977 Bothriocidaris eichwaldi Mannil; Kolata et al.: 149.
1978 Bothriocidaris eichwaldi Männil; Kier & Lawson: 1.
1982 Bothriocidaris eichwaldi Männil; Kier: 314.
1984 Bothriocidaris eichwaldi Männil; Guensberg: 67, 68.
1989 Bothriocidaris eichwaldi Männil; Schallreuter: 5, 9, 12, 13, pl. 7, figs 3A-H, 4B.
1990 Bothriocidaris eichwaldi Männil; Rhebergen: 50, 51, figs 4B, 5.
1990 Bothriocidaris eichwaldi Männil; Schallreuter: 126, fig. 3.2.
2001 Bothriocidaris eichwaldi Männil; Smith & Savill: 138, fig. 2.
2009 Bothriocidaris eichwaldi Männil; Solovjev: 1415, 1417, 1418, figs 3B, 5, 6.
Type specimens. The type is specimen GIT 106-3.
35
Diagnosis. Modified from Männil (1962). Bothriocidaris with two widely spaced perforate,
non-crenulate tubercles merged with the raised, well-developed peripodium on each ambulacral
plate. Ambulacral plates are less than twice as wide as they are high. Ambulacral and
interambulacral plates are covered with numerous small, imperforate secondary tubercles.
Interambulacral plates are hexagonal, bear no primary tubercles, and are arranged into a single
column in each area.
Occurrence. This species is known from the Katian outcrops at Jootma Ditch (Kuru) and the
Paluküla quarrys, Hiiu County, Estonia and has been found as disarticulated coronal plates from
Katian Öjlemyr glacial erratic boulders on Gotland, Sweden (Schallreuter 1989).
Remarks. This species is the most-well known bothriocidaroid as this is the only bothriocidaroid
with a well-known Aristotle’s lantern (Männil 1962). This species is very similar to B. pahleni as
both species contain two primary tubercles on each ambulacral plate. The two species are only
differentiated morphologically based upon the distance between these two tubercles, with B.
eichwaldi having more widely separated tubercles, and B. pahleni having more closely spaced
primary tubercles. In their initial phylogenetic hypotheses for Bothriocidaris, Männil (1962) and
Nestler (1968) placed B. eichwaldi as sister to a clade of B. parvus + B. globulus. Our analyses
did not confidently resolve this species’ phylogenetic placement.
Bothriocidaris solemi Kolata, 1975
1975 Bothriocidaris solemi Kolata: 66, pl. 14, figs 14-15.
1977 Bothriocidaris solemi Kolata; Kolata et al.: 149.
36
1982 Bothriocidaris solemi Kolata; Kier: 314.
1984 Bothriocidaris solemi Kolata; Guensberg: 67.
1987 Bothriocidaris solemi Kolata; Kolata et al.: fig. 16.2.
2001 Bothriocidaris solemi Kolata; Smith & Savill: 138, fig. 2.
2009 Bothriocidaris solemi Kolata; Solovjev: 1418.
2010 Bothriocidaris solemi Kolata; Kroh: 204.
Type specimens. The holotype is UI X-4882 and the paratypes are UI-X-4942 to X-4945.
Diagnosis. Modified from Kolata (1975). Bothriocidaris with slightly raised peripodia
surrounding pore-pairs on ambulacral plates. Peripodia merged with two perforate non-crenulate
primary tubercles. Ambulacral plates less than twice as wide with five or six smaller perforate
secondary tubercles. Interambulacral plates arranged into a single column, with a single perforate
non-crenulate primary tubercle and five to ten smaller perforate secondary tubercles.
Occurrence. The holotype and paratypes are from the Walgreen Member of the Grand Detour
Formation, Platteville Group (Sandbian) at Kolata’s (1975) locality 22, which is the Medusa
Cement quarry on the north side of Route 2, 0.5 miles north of Dixon, Lee County, Illinois.
Kolata (1975) also states that coronal plates of this species occur, “throughout the Grand Detour
Formation in the study area”, with the study area being Northern Illinois and Southern
Wisconsin.
37
Remarks. This species is differentiated from B. parvus, B. pahleni, B. eichwaldi, B. solemi, and
B. globulus by its numerous perforate secondary tubercles, which cover its ambulacral and
interambulacral plates. This character makes it most similar to B. vulcani, though it is different
from this taxon in having only a single column of interambulacral plates and ambulacral plates
which are less than twice as wide as high. Our phylogenetic analyses could not discern any clear
relationships between this taxon and any other species of Bothriocidaris.
Bothriocidaris maquoketensis Kolata, Strimple, Levorson, 1977
1977 Bothriocidaris maquoketensis Kolata et al.: 146, pl. 1, figs 1-8, text-fig. 1.
1984 Bothriocidaris maquoketensis et al.; Guensberg: 67.
1987 Bothriocidaris maquoketensis et al.; Kolata et al.: fig. 16.2.
1989 Bothriocidaris maquoketensis et al.; Schallreuter: 3.
1999 Bothriocidaris maquoketensis et al.; Frest et al.: 654.
2009 Bothriocidaris maquoketensis Kolata; Solovjev: 1415, 1418, fig. 10.
2010 Bothriocidaris maquoketensis et al.; Kroh: 204.
Type specimens. The holotype is specimen SUI 42700 and the paratype is SUI 42701.
Diagnosis. Modified from Kolata et al. (1977). Bothriocidaris with two widely spaced,
perforate, non-crenulate, primary tubercles on all ambulacral plates. Ambulacral plates more than
twice as wide and peripodium slightly raised. The interambulacral plates bear between one and
three perforate, non-crenulate, primary tubercles. The interambulacra is composed of alternating
rows of one and two plates, which are irregularly pentagonal to heptagonal in outline. The plates
38
in rows with one plate are large, and either higher than wide or wider than high. Plates in rows
consisting of two plates are smaller, and usually equidimensional.
Occurrence. The holotype and paratype are from the Katian Fort Atkinson formation of the
Maquoketa group. Both specimens are from road fill exposures, and the holotype was collected
from Southeast of Fort Atkinson, Winneshiek County, Iowa, and the paratype from Southwest of
Eldorado, Fayette County, Iowa.
Remarks. B. maquoketensis is similar to B. pahleni, B. eichwaldi, B. vulcani, and B. solemi in
having only two primary tubercles on each ambulacral plate. It differs from all other species of
Bothriocidaris except for B. vulcani in having multiserial interambulacral areas, and in having
interambulacral plates that are not hexagonal in shape. Our phylogenetic analyses consistently
found this taxon to be sister group to B. vulcani, though with low, and variable, bootstrap support
and posterior probabilities.
Bothriocidaris kolatai Kier, 1982
(Figs 2.5C, D)
v*1982 Bothriocidaris kolatai Kier: 311, pl. 41, figs 1-2; figs 74A-B.
1984 Bothriocidaris kolatai Kier; Guensberg: 67.
1989 Bothriocidaris kolatai Kier; Schallreuter: 3.
2009 Bothriocidaris kolatai Kier; Solovjev: 1418.
2010 Bothriocidaris kolatai Kier; Kroh: 204.
39
Type specimen. The holotype is specimen TMM 1122TX.57.
Diagnosis. Modified from Kier (1982). A species of Bothriocidaris with five or six high,
pronounced perforate, non-crenulate primary tubercles on each ambulacral plate (Figs. 5C, D).
The peripodial ring is absent and ambulacral plates are less than twice as wide as high. There are
four to five high perforate, non-crenulate tubercles on each interambulacral plate.
Interambulacral plates are hexagonal and arranged into a single column in each area. There are
no secondary tubercles present on any ambulacral or interambulacral plates.
Occurrence. The holotype and only known specimen is from Zone 3 of the Pooleville Member,
Bromide Formation (Sandbian), 9 to 10 m below the Viola Limestone at Culley Creek, Criner
Hills, Carter County, Southern Oklahoma (Kier 1982).
Remarks. This species is the only known of Bothriocidaris which lacks an elevated periopdial
ring surrounding the ambulacral tubercles. Furthermore, this species has far more primary
tubercles on both its ambulacral (five to six) and interambulacral (four to five) plates than any
other bothriocidaroid (Figs. 5C, D). Our phylogenetic analyses were not able to consistently
determine the relationships of this species to any other species of Bothriocidaris.
Bothriocidaris vulcani Guensburg, 1984
(Fig 2.3B)
40
1984 Bothriocidaris vulcani Guensberg: 67, pl. 14, figs 4-10.
2001 Bothriocidaris vulcani Guensberg; Smith & Savill: 138, 144, 142, 146, figs 2, 7.
2001 “Bothriocidaris” vulcani Guensberg; Smith & Savill: fig. 6.
2009 Bothriocidaris vulcani Guensburg; Solovjev: 1418.
2010 Bothriocidaris vulcani Guensburg; Kroh: 204.
Type specimens. The holotype is UI X-5842, and the two paratypes are UI X-5841 and UI X-
5843.
Diagnosis. Modified from Guensburg (1984). Bothriocidaris with two perforate, non-crenulate
primary tubercles on each ambulacra plate. Ambulacral plates more than twice as wide as high,
and peripodia slightly raised. From zero to four perforate secondary tubercles on ambulacral
plates. Interambulacral plates arranged into numerous, irregular columns in each interambulacral
area. The shape of interambulacral plates is irregular, and larger plates bear a single, perforate,
non-crenulate primary tubercle. Interambulacral plates bear up to fourteen (average of 12) small,
perforate, secondary tubercles.
Occurrence. The holotype, specimen UI X-5842 is from the lower member of the Sandbian
Lebanon Limestone at Guensburg’s (1984) locality Z-656, which is a roadcut on interstate 65,
approximately 0.3 km south of the bridge over Duck River, Maury County, Tennessee. The
paratypes, UI X-5841 and UI X-5843, are from the lower member of the Sandbian Lebanon
Limestone at Guensburg’s (1984) locality Z-651, which is the Vulcan Materials Company
Quarry at Una on the southeast side of Nashville, Davidson County, Tennessee.
41
Remarks. This species is similar to B. maquoketensis, B. pahleni, B. eichwaldi, and B. solemi in
that it has two perforate, non-crenulate tubercles on each ambulacral plate. This is the only
bothriocidaroid with more than two columns of irregular interambulacral plates (Fig. 3B), and is
similar to only B. maquoketensis in having more than a single column of interambulacral plates.
It is also similar to this species in that it has ambulacral plates which are more than twice as wide
as high. B. vulcani also has numerous small, secondary tubercles covering the test, which is
similar to the condition found in B. solemi. Many of our phylogenetic analyses resolved this
taxon as the sister species of B. maquoketensis.
Genus Neobothriocidaris Paul, 1967
Type species. Neobothriocidaris peculiaris Paul, 1967
Other species. Neobothriocidaris minor Paul, 1967 and Neobothriocidaris tempeltoni Kolata,
1975, Neobothriocidaris pentlandensis n sp.
Diagnosis. Bothriocidaroid with greatly expanded ambulacra and reduced or absent
interambulacra. A single column of imperforate ambulacral plates separates each half-
ambulacrum of perforate ambulacral plates. One primary tubercle and one or two small
secondary tubercles on each of these plates. Pore-pairs and surrounding peripodial ring for each
ambulacral pore span onto two plates and pore-pairs arranged en chevra. A single primary
tubercle is present on each ambulacral plate and is merged with the peripodial ring. A few
42
imperforate secondary tubercles are present on perforate ambulacral plates. Lateral vessels are
enclosed entirely within ambulacral plates (Fig. 2.6A, 2.8E). In some species, more than one
column of perforate ambulacral plates are in contact with perradial series of imperforate plates.
Spines are small, striate. With non-crenulate acetabula and tapering distally (Fig. 2.8A, C).
Occurrence. Articulated or semi-articulated specimens are known from the Katian Lady Burn
and Craighead Limestone formations of Girvan, Scotland (Paul 1967), the Katian or Sandbian
Eagle Point Member of the Dunleith Formation of the Galena Group of Illinois, and now the
Telychian of the Pentland Hills, Scotland. Disarticulated coronal plates are known from the
Forreston Member of the Grand Detour Formation, Platteville Group of Illinois and, “on nearly
every bedding plane in the Grand Detour Formation through the Dunleith Formation of the
Galena Group in [Northern Illinois and Southern Wisconsin]” (Kolata 1975). Furthermore they
are known from the Katian Kalvsjøen Formation of the Oslo Region, Norway (Bockelie and
Briskeby, 1980), the Darriwilian Mójcza Limestone of the Holy Cross Mountains, Poland (Pisera
1994), glacial erratic boulders of the Katian or Hirtantian Öjlemyr Flint, from Gotland, Sweden
(Schallreuter 1989) and provisionally from the Sandbian Lebanon Limestone of Tennessee
(Guensburg 1984). In the Silurian, disarticulated coronal plates of Neobothriocidaris are known
from numerous horizons spanning the Llandovery to Ludlow on Gotland, Sweden (Franzén,
1979; Kutscher and Reich, 2001, 2004; Reich and Smith, 2009).
Remarks. Neobothriocidaris Paul, 1967 was the first genus described from within the
Bothriocidaroida other than Bothriocidaris. It was first described from the Katian strata of
Girvan, Scotland and this fauna, which included two species of Neobothriocidaris, N. peculiaris
43
Paul, 1967, and N. minor Paul, 1967 and B. globulus, was the first bothriocidaroid fauna known
from outside of Baltica. Neobothriocidaris is differentiated from Bothriocidaris by its enlarged
ambulacral areas, comprised of up to six columns of perforated plates in each half-ambulacra.
Furthermore, Neobothriocidaris, is unique amongst the bothriocidaroids (and all other Paleozoic
echinoids) in having pore-pairs of each tube foot which are not confined to a single plate. Lastly,
like Unibothriocidaris, Neobothriocidaris displays a column of imperforate ambulacral plates
located perradially between both half-ambulacra. Unlike Unibothriocidaris, the preservation of
in-filled lateral water vessels (Paul 1967) demonstrates that these structures existed enclosed
within the ambulacral plates. However, we note after Smith and Savill (2001) that given the
impression of the radial water vessel in a cast of the holotype (Fig. 6A), the radial water vessel
was not enclosed within this perradial series of plates. Our phylogenetic analyses found high
support for a clade of all species of Neobothriocidaris (see Results).
Neobothriocidaris minor Paul, 1967
1967 Neobothriocidaris minor Paul: 535, pl. 85, figs 5-8.
1975 Neobothriocidaris minor Paul; Kolata: 67.
1978 Neobothriocidaris minor Paul; Kier & Lawson, 2.
1979 Neobothriocidaris minor Paul; Franzén: 220.
1980 Neobothriocidaris minor Paul; Bockelie & Briskeby: 91.
1989 Neobothriocidaris minor Paul; Schallreuter: 12, fig. 9, pls. 10, 11.
1990 Neobothriocidaris minor Paul; Rhebergen: 50.
1994 Neobothriocidaris minor Paul; Pisera: 299.
1996 Neobothriocidaris minor Paul; Donovan et al.: 260, pl. 51, figs 2,3.
44
2001 Neobothriocidaris minor Paul; Smith & Savill: 138, 142, fig. 2.
2009 Neobothriocidaris minor Paul; Solovjev: 1418, fig. 8A.
Type specimens. The holotype is HM E 1427.
Diagnosis. Modified from Paul (1967). A species of Neobothriocidaris with perforate
ambulacral tubercles bearing one perforate, non-crenulate, primary tubercle attached to the
peripodium, and a smaller imperforate secondary tubercle on the plate surface slightly adapical
of the primary tubercle. Some perforate ambulacral plates also bear a second imperforate
secondary tubercle or small swelling merged with the peripodium located adorally to the other
secondary tubercle and either perradially or adradially to the primary tubercle. There are up to
seven columns of perforate ambulacral plates in each half-ambulacrum. Imperforate ambulacral
plates have one perforate, non-crenulate, primary tubercle, and a single smaller imperforate
secondary tubercle located adapically of the primary tubercle. A single non-contiguous column
of small diamond-shaped interambulacral plates are also present inter-radially.
Occurrence. This species was described by Paul (1967) from two specimens form Girvan,
Scotland. The holotype HM E 1427 was collected from the Katian Lady Burn Formation,
Drummock Subgroup, Ardmillan Group from Threave Glen. A second specimen, Paul’s (1967)
“Lund Specimen” is from the Katian Kiln Mudstone Member, Craighead Limestone Formation,
Ardwell Subgroup, Ardmillan Group at the Kilns in the Main Quarry at Craighead. In addition to
these two occurrences at Girvan, disarticulated coronal plates attributable to this genus have been
45
recovered from glacial erratic boulders of the Katian or Hirtantian Öjlemyr Flint, from Gotland,
Sweden (Schallreuter 1989).
Remarks. This species of Neobothriocidaris is characterized in part by having a single
secondary tubercle located adapical of the primary tubercle on plates. Paul (1967) additionally
noted the presence of small peripodial swellings on his material, which Schallreuter (1989)
showed were likely tubercles. Thus, although the presence of a single, as opposed to two,
secondary tubercles in part differentiates this from N. peculiaris, the two tubercles of N.
peculiaris are both separate of the peripodium, while a second secondary tubercle on N. minor is
merged with the peripodium. The location of these swellings or secondary tubercles, either
adradially or perradially of the primary tubercle appears to very from plate to plate (Paul, 1967;
Schallreuter, 1989). This species is also similar to N. tempeltoni from the Sandbian or Katian of
Illinois in having a series of diamond shaped interambulacral plates. This series of plates does
not appear to have been present in N. peculiaris or N. pentlandensis. Our phylogenetic analyses
consistently resolved N. minor as the sister species to N. tempeltoni.
Neobothriocidaris peculiaris Paul, 1967
(Fig 2.6)
v*1967 Neobothriocidaris peculiaris Paul: 535, pl. 84, figs 1-6, pl. 85, figs 9-11.
1975 Neobothriocidaris peculiaris Paul; Kolata: 67.
1978 Neobothriocidaris peculiaris Paul; Kier & Lawson: 2.
1979 Neobothriocidaris peculiaris Paul; Franzén: 220.
46
1980 Neobothriocidaris peculiaris; Bockelie & Briskeby: 91, fig. 2.
1989 Neobothriocidaris peculiaris Paul; Schallreuter: 12, 13.
1990 Neobothriocidaris peculiaris Paul; Rhebergen: 50.
1996 Neobothriocidaris peculiaris Paul; Donovan et al.: 260, pl. 51, fig. 1.
1998 Neobothriocidaris peculiaris Paul; Lewis & Donovan: fig. 1A.
2001 Neobothriocidaris peculiaris Paul; Smith & Savill: 138, 142, 144, fig. 2.
2009 Neobothriocidaris peculiaris Paul; Solovjev: 1418, fig 8A.
Type specimens. The holotype is NHMUK E 42523 a-d.
Diagnosis. Modified from Paul (1967). Neobothriocidaris with perforate ambulacral plates
bearing one perforate, non-crenulate primary tubercle merged with the peripodium and one to
three smaller imperforate secondary tubercles on the plate surface (Figs. 2.6E, F). The type
specimen contains five columns of perforate ambulacral plates in each half-ambulacrum. The
imperforate ambulacral plates yield a single, large perforate non-crenulate primary tubercle
located vertically between two smaller secondary tubercles (Fig. 2.6G). The secondary tubercle
above the primary tubercle may be imperforate or perforate. This species does not appear to bear
any interambulacral plates. Only a single column of perforate ambulacral plates in contact with
perradial column of imperforate ambulacral plates in each half-ambulacrum (Fig. 2.6A, B).
Occurrence. The holotype is known from the Katian Lady Burn Formation, Drummock
Subgroup, Ardmillan Group from Threave Glen, Girvan, Scotland (Paul 1967). Bockelie and
47
Briskeby (1980) described additional disarticulated ambulacral plates attributable to this species
from the Katian Kalvsjøen Formation of Hadeland, Oslo Region, Norway.
Remarks. This species has more tuberculate ambulacral plates than any other species of
Neobothriocidaris. It is similar to N. pentlandensis, and differs from N. minor and N. tempeltoni
in that it does not appear to have had any interambulacral plates. Additional material described
by Bockelie and Briskeby (1980) from the Katian of Norway displayed an additional third
imperforate secondary tubercle near to the peripodium and adoral to the other two imperforate
tubercles. Given that the presence or absence of some of these small imperforate ambulacral
tubercles appears to be variable from plate to plate across numerous species of Neobothriocidaris
(Bockelie and Briskeby, 1980; This study; Schallreuter, 1989) we favor treating the material
from Norway as N. peculiaris. Our Bayesian phylogenetic analyses and many of our most
parsimonious trees resolved this species as sister group to N. pentlandensis.
Neobothriocidaris templetoni Kolata, 1975
1975 Neobothriocidaris templetoni Kolata: 66, pl. 14, figs. 1-2, text-figs 18, 19.
1987 Neobothriocidaris templetoni Kolata; Kolata et al.: fig. 16.2.
2001 Neobothriocidaris templetoni Kolata; Smith & Savill: 138, 142, 144, fig. 2.
2009 Neobothriocidaris templetoni Kolata; Solovjev: 1418.
2010 Neobothriocidaris templetoni Kolata; Kroh: 205.
Type specimens. The holotype is specimen UI X-4883.
48
Diagnosis. Neobothriocidaris with a single perforate, non-crenulate primary tubercle on
ambulacral plates. Three columns of perforate ambulacral plates in each half ambulacrum. Half
ambulacra in each ambulacrum separated by column of imperforate ambulacral plates bearing a
single perforate, non-crenulate tubercle. Two or three small, diamond shaped interambulacral
plates present along interradial suture.
Occurrence. The holotype is known from the Eagle Point Member of the Dunleith formation of
the Galena Group (Katian or Sandbian) at Kolata’s (1975) locality 10, which is the Porter
Brother's quarry on the west side of old highway 2, one mile south of Rockton, Winnebago
County, Illinois.
Remarks. This is the only named species of Neobothriocidaris from North America, though
disarticulated plates occur abundantly in the Galena Group (Kolata 1975) and a single tentative
specimen of Neobothriocidaris was described from the Lebanon Limestone. N. tempeltoni
differs, however, from the disarticulated plates occurring in the Galena Group in lacking
secondary tubercles. N. tempeltoni is similar to N. minor, in that it also has small, diamond
shaped, interambulacral plates, and these are the only two species of Neobothriocidaris known to
have had such as series of plates. Our phylogenetic analyses also consistently resovle these two
species as sister species.
Neobothriocidaris sp.
1994 Neobothriocidaris sp. cf. N. minor Paul; Pisera: 299, pl. 61, fig. 15.
Material. This specimen is number ZPAL E.IX/MA-27 Pisera 1994: pl. 61: 15.
49
Occurrence. This specimen was recovered from Skala hill near Mojcza village south of Kielce
in the Holy Cross Mountains, southern Poland. It is from the Pygodus serra conodont zone of the
Mójcza Limestone and is thus Dariwillian in age.
Description. See Pisera (1994). This is a fragmentary perforate ambulacral plate with a
perforate, non-crenulate tubercle and a well-developed periopdium. The primary tubercle is
slightly located within the peripodium. The plate surface appears slightly irregular, and a single
imperforate, non-crenulate, tubercle is located near to the primary tubercle separated from the
peripodium.
Remarks. Pisera (1994) described a disarticulated and fragmentary Neobothriocidaris plate from
the Dariwillian of Poland, which he compared to N. minor. This specimen is the earliest
occurrence of a bothriocidaroid, and thus brings their range into the Dariwillian. Pisera’s N. sp.
cf. N. minor bore a single, perforate primary tubercle and a single secondary tubercle. Pisera
noted that although his material was superficially similar to Paul’s N. minor, the latter bore
secondary tubercles located to the right of the left of the primary tubercles, while the former’s
secondary tubercle was located to the right. Schallreuter (1989) described disarticulated coronal
plates attributed to N. minor, some of which had secondary tubercles located to the right of the
primary tubercle, and some of which had secondary tubercles to the left, this is probably simply a
function of whether or not the perforate ambulacral plates are from half-ambulacra to the left or
right of the perradial zone of plates. Pisera’s N. sp. cf. N. minor does, however, differ from all
other described species of Neobothriocidaris, including N. minor, in the placement of its primary
50
tubercle. While all other Neobothriocidaris have primary tubercles slightly above, yet merged
with, the peripodium, N. sp. cf. N. minor has a primary tubercle which appear to be slightly
below the adapical rim of the peripodium, and thus appears to be located slightly within the
peripodial ring. We feel this feature, plus its stratigraphic occurrence in the Dariwillian, are
enough to merit its distinct designation from N. minor until more complete material is recovered.
Neobothriocidaris sp.
1975 Neobothriocidaris sp. Kolata: 68, pl. 14, figs 9-13.
1987 Neobothriocidaris sp. Kolata; Kolata et al.: fig. 16.2.
Material. UI X-4946 to UI X-4948.
Occurrence. The type material is from the Forreston Member, Grand Detour Formation,
Platteville Group (Sandbian) from Kolata’s (1975) locality 7, which is the Quarry on the north
side of Illinois state highway 70, 12 miles northwest of Rockford, Winnebago County, Illinois.
Kolata also notes that there are disarticulated Neobothriocidaris plates on, “nearly every bedding
plane in the Grand Detour Formation through the Dunleith Formation of the Galena Group in the
study area.”
Description. This material consists of disarticulated coronal plates and was described in detail
by Kolata (1975). The coronal plates bear one perforate primary tubercle in contact with the
peripodium and two or three smaller secondary perforate or imperforate tubercles located around
the peripodium.
51
Remarks. The material figured by Kolata (1975) bears more secondary tubercles than any other
known species of Ordovician Neobothriocidaris. It is most similar to specimens of
Neobothriocidaris peculiaris, described by Bockelie and Briskeby (1980) from the Katian of
Norway which had three secondary tubercles, though this material differs in that it is the only
known Neobothriocidaris with perforate secondary tubercles. This material was also collected
from the same strata as N. tempeltoni. Although the morphology of the coronal plates, and
particularly the presence or absence of secondary tubercles, differs between these specimens, it is
possible that this material may represent disarticulated plates attributable to N. tempeltoni, the
holotype of which appears to be rather weathered.
?Neobothriocidaris sp.
1984 ?Neobothriocidaris species Guensburg: 69, pl. 14, fig. 11.
2001 ?Neobothriocidaris spp. Guensburg; Smith & Savill: 138.
2009 ?Neobothriocidaris spp. Guensburg; Solovjev: 1418.
Material. UI X-5839.
Occurrence. This specimen is from the Sowerbyella-Diplograptus Zone of the upper member of
the Sandbian Lebanon limestone at Guensburg’s (1984) locality Z-652, which is the M.C. West
Lime Company Quarry at Turney, 6 km south of Pulaski, Giles County, Tennessee.
52
Description. This is a single disarticulated ambulacral plate with a central pore and a single
sutural pore.
Remarks. Guensburg (1984) chose to treat this specimen as ?Neobothriocidaris sp. He noted the
presence of a single sutural pore, which indicates that if this is a plate of Neobothriocidaris, the
presence of only a single sutural pore (as opposed to two pores, which are present on all non-
adradial ambulacral plates) indicates that this is an adradial plate. We have not directly examined
this specimen, and thus choose to follow Guensburg (1984) in treating it as ?Neobothriocidaris
sp.
Neobothriocidaris pentlandensis n. sp.
(Figs 2.7, 2.8.)
Type Specimens. The holotype is specimen NMS G.1976.48.1.
Diagnosis. Neobothriocidaris with irregular ridges radiating out from primary tubercle on
ambulacral plate surface and small single mamelon on ambulacral plates.
Derivation of name. The species is named for the Pentland Hills, in Scotland, from where the
type material was collected.
53
Occurrence. The specimen is from the ‘head of the Gutterford Burn, Pentland Hills’ in Scotland.
The Gutterford Burn in the Pentland Hills is thus likely the lower Deerhope Formation, which is
Telychian in age.
Description. Test is small, regular. The test has been crushed, such that accurate determination
of its width is not possible. The plating is tessellate, and plates are thick, about 0.5 mm thick.
This species has expanded ambulacral areas, and numbers at least four columns of perforated
ambulacral plates in each half ambulacrum (Figs. 2.7A, 2.8A-D). No column of imperforate,
perradial plates is preserved (Fig. 2.8D), so the morphological details of this structure are
unknown. The interambulacral areas appear to be absent and there is no sign of the small,
diamond shaped interambulacral plates that are present in some species of the genus. Details of
the apical system are unknown.
Ambulacral plates are approximately as high as wide, with some plates being slightly
wider than high (Height/Width=1.16), and some being slightly higher than wide
(Height/Width=0.92), dependent upon their placement in the test. Each pore-pair spans onto
multiple plates, with the more perradial pore in each pair being located on the suture between the
perradial side of a plate and the next most perradial plate, and the more adradial being located
subcentrally on the plate (Fig. 2.8E). On the interior of the test, the more adradial pores have a
mean diameter of 0.38 mm. Each pore-pair is surrounded by a peripodial ring which appears to
surround both pores (Figs. 2.8F, G), though the presence of this peripodial ring around the more
perradial more is difficult to see. Each ambulacral plate bears a perforate, primary tubercle,
which bore a spine. This tubercle consists of a small, hollow tube extending distally from the
plate, which is, on average 0.4 mm in diameter. They are merged with the peripodial ring
54
adradially and adapically on each plate. On some plates, a single small, imperforate mamelon is
present on the plates adapical to the primary tubercle (Figs. 2.8E-G). The interior side of the
ambulacral plates appears smooth and without ornamentation, however the exterior surface bears
distinct sculpturing (Fig. 2.8E, G). The surfaces of these plates bear smooth, raised, ridges,
which appear to irregularly radiate out from the primary tubercle towards the edges of the plates.
These ridges may bifurcate distally from the tubercle towards the plate edges. Along the edge of
some plates, the openings of the lateral water vessels are visible, indicating that they are enclosed
within the plates of this species. The peristomial plating and Aristotle’s lantern are unknown.
Spines are small, tapering distally (Fig. 2.8A). The acetabulum is a small concave
indentation at the base of the spines. Spines are up to 2 mm long. There is no milled ring. There
appear to be no secondary spines, though we cannot exclude that there may have been such
spines which became separated from the test due to taphonomic processes.
Remarks. The occurrence of this taxon from the Gutterford Burn in the Pentland Hills marks the
second species of echinoid known from the locality. Kier (1973) additionally described the
lepidocentrid species Aptilechinus caledonensis Kier, 1973 from the “Starfish Bed Gutterborn
Burn, Pentland Hills”. While N. pentlandensis is only known from a single specimen, A.
caledonensis is known from at least twenty-one specimens, and may well have been more
abundant than N. pentlandensis. It is also, however, unknown if these taxa co-occur in the same
exact stratigraphic layers, as the precise stratigraphic distribution of both taxa is not known.
Morphologically, this taxon is most similar to the undescribed Neobothriocidaris sp. from the
Silurian of Gotland figured by Kutscher and Reich (2001). Both of these taxa are differentiated
from all other bothriocidaroids in having raised ridges radiating out from the tubercle on the
55
ambulacral plate surfaces. Because the Neobothriocidaris sp. from Gotland is only known from
disarticulated plates, it’s phylogenetic affinities related to N. pentlandensis are unknown.
Bayesian phylogenetic analyses and some of our most parsimonious trees found this species to
form a clade with N. peculiaris.
Genus Unibothriocidaris Kier, 1982
Type species. Unibothriocidaris bromidensis Kier, 1982
Other species. Unibothriocidaris kieri Guensburg, 1984
Diagnosis. Bothriocidaroid with enlarged ambulacral zones and no interambulacral zones.
Imperforate column of perradial ambulacral plates separates each half-ambulacrum. This column
of plates without primary tubercles and does not appear to reach the peristome (Guensburg 1984;
Plate 14, Fig. 12). Single opening for each ambulacral pore located entirely within each perforate
ambulacral plate and pores are arranged en chevra. Perforate ambulacral plates with a single
primary tubercle merged with peripodial ring. Ambulacral plates covered in numerous
imperforate secondary tubercles. More than one column of perforate ambulacral plates in contact
with perradial series of imperforate plates. Radial water vessel and lateral water vessels are
entirely within the corona, and are not enclosed within the coronal plates. Peristomial plating
consisting of larger radial and interradial plates with smaller peristomial plates present adorally.
Spines small, striate. With non-crenulate acetabula and tapering distally.
56
Occurrence. Unibothriocidaris is known from articulated or semi-articulated test material from
the Sandbian Bromide Formation of Oklahoma (Kier 1982), and Lebanon limestone of
Tennessee (Guensburg 1984), U.S.A. This genus is also known from disarticulated material from
the Sandbian Briton Member of the Mifflin Formation, Platteville Group, Illinois (Kolata 1975).
Remarks. This genus was first described from Unibothriocidaris bromidensis Kier, 1982 from
the Sandbian Bromide formation of Oklahoma. Unibothriocidaris is differentiated from
Bothriocidaris, and similar to Neobothriocidaris, in having expanded numbers of ambulacral
columns. Unibothriocidaris has numerous columns of ambulacral plates pierced by a single
perforation and a single column of imperforate perradial plates. The presence of a single
perforation in each ambulacral plate, however, is distinct from the double perforations of
Bothriocidaris and Neobothriocidaris. This perforation is identical to those of Kolata’s (1975)
“Genus and Species Unknown”, which is herein treated as Unibothriocidaris sp. Unlike
Neobothriocidaris, there is no evidence in Unibothriocidaris for the enclosure of the radial water
vessel and lateral vessels within the coronal plates.
The treatment of the imperforate columns of plates in Unibothriocidaris as
interambulacral or ambulacral have changed since the taxon was described. Kier (1982) initially
interpreted the imperforate column of plates as interambulacral, and used the lack of a perradial
column of imperforate ambulacral plates paired with the absence of an internal radial water
vessel and lateral vessels to infer that Unibothriocidaris was more closely related to
Bothriocidaris than to Neobothriocidaris. The interpretation of the imperforate column of plates
as interambulacral in Unibothriocidaris persisted with Guensburg’s (1984) description of
Unibothriocidaris kieri Guensburg, 1984 from the Sandbian Lebanon Limestone of Tennessee.
57
Smith and Savill (2001), however, interpreted this column of imperforate plates as a column of
imperforate perradial plates, with Unibothriocidaris thus lacking interambulacra. The pores on
the ambulacral plates in Unibothriocidaris are located on the side of each plate that is closest to
this column of imperforate plates (Guensburg 1984; Plate 4, Figs. 13, 14). Because this closely
matches the arrangement of pores on the ambulacral plates of Neobothriocidaris, which
unambiguously has an imperforate column of ambulacral plates (Paul 1967; Plate 85, Fig. 11),
we follow Smith and Savill (2001) and interpret this column of plates as a perradial column of
imperforate ambulacral plates. Our phylogenetic analyses found relatively high support for both
species of Unibothriocidaris forming a clade (see Results).
Unibothriocidaris bromidensis Kier, 1982
(Figs 2.5A, B)
1982 Unibothriocidaris bromidensis Kier: 311, pl. 41, fig. 3; fig 73.
1984 Unibothriocidaris bromidensis Kier; Guensberg: 68.
2001 Unibothriocidaris bromidensis Kier; Smith & Savill: 138, fig. 2.
2009 Unibothriocidaris bromidensis Kier; Solovjev: 1418, fig. 9.
2010 Unibothriocidaris bromidensis Kier; Kroh: 205.
Type specimen. The holotype is TMM 1122TX.1.
58
Diagnosis. Emended from Kier (1982). Unibothriocidaris with three columns of ambulacral
plates in each half-ambulacrum and four to six imperforate secondary tubercles present on each
ambulacral plate (Figs. 5A, B). Peripodia are well-developed and raised.
Occurrence. The holotype is from Zone 3 of the Pooleville Member, Bromide Formation
(Sandbian), 9 to 10 m below the Viola Limestone at Culley Creek, Criner Hills, Carter County,
Southern Oklahoma (Kier 1982).
Remarks. This species is similar to U. kieri except in the number of secondary tubercles present
on ambulacral plates, the number of ambulacral plate columns, and the degree to which its
peripodia are developed. U. bromidensis has three columns in each half-ambulacrum, as opposed
to the eight in U. kieri, and only four to six secondary tubercles, whereas U. kieri has an average
of twelve secondary tubercles on each ambulacral plate. U. bromidensis furthermore has much
more well-developed peripodia than U. kieri, which has faint to absent peripodia. Our
phylogenetic analyses generally found high support for these two taxa forming a clade.
Unibothriocidaris kieri Guensburg, 1984
1984 Unibothriocidaris kieri Guensburg: 68, pl. 14, figs 12-15, text-fig. 16.
2001 Unibothriocidaris kieri Guensburg; Smith & Savill: 138, fig. 2.
2009 Unibothriocidaris kieri Guensburg; Solovjev: 1418.
2010 Unibothriocidaris kieri Guensburg; Kroh: 205.
59
Type specimens. The holotype is UI X-5832, and the paratypes are UI X-5833 to UI X-5838 and
UI X-6036 to 6044. All of these specimens are presumed to be disarticulated portions of the
same individual.
Diagnosis. Updated from Guensburg (1984). Unibothriocidaris with four columns of ambulacral
plates in each half-ambulacrum and an average of twelve, and up to sixteen, imperforate
secondary tubercles present on each ambulacral plate. Peripodia are absent or faint.
Occurrence. The holotype and paratypes are from the Lebanon Limestone (Sandbian) at the
Vulcan Materials Company Quarry at Una on southeast the side of Nashville, Davidson County,
Tennessee.
Remarks. See Remarks. of U. bromidensis for a comparison of this taxon to the only other
named species of Unibothriocidaris.
Unibothriocidaris sp.
1975 Genus and species unknown Kolata: 68, pl. 14, figs 3-8.
1982 Unibothriocidaris sp.; Kier: 310.
1984 Bothriocidaridae, genus and species unknown Kolata; Guensberg: 68.
1987 Genus and species unknown Kolata; Kolata et al.: fig. 16.2.
2001 Unibothriocidaris sp. cf bromidensis Kier; Smith & Savill: 138.
2009 Unibothriocidaris sp. cf bromidensis Kier; Solovjev: 1418.
60
Material. UI X-4950 to UI X-4953, UI X-5096 and UI X-4954.
Occurrence. This material was colleceted from the Briton Member of the Mifflin Formation,
Platteville Group (Sandbian) from Kolata’s locality 22, which is the Medusa Cement quarry on
the north side of Route 2, 0.5 miles north of Dixon, Lee County, Illinois.
Description. A thorough description of this material is found in Kolata (1975), however, we
briefly describe the material herein. This material consists of polygonal plates with a single
ambulacral pore, a single perforate non-crenulate primary tubercle, and numerous, small
imperforate secondary tubercles. The associated spine is slender and striated with a flat base.
Remarks. Kolata (1975) named this material as “Genus and Species Unknown”, however this
was before the description of Unibothriocidaris by Kier (1982). The single perforation in
ambulacral plates diagnoses this material as belonging to Unibothriocidaris. The numerous
secondary tubercles, and faint peripodia are reminiscent of those present in U. kieri, however, as
the material consists of only disarticulated ambulacral plates and spines, we refrain from
assignment to the species level.
Unibothriocidaris sp.
1984 ?Unibothriocidaris species Guensberg, 1984; 69, pl. 13, fig. 15.
2001 ?Unibothriocidaris sp. Guensberg 1984; Smith & Savill: 138.
2009 ?Unibothriocidaris sp. Guensburg; Solovjev: 1418.
61
Material. UI X-5840.
Occurrence. This specimen is from the lower member of the Sandbian Lebanon Limestone and
was collected from Guensburg’s locality Z-655, which is the road cut on interstate 24
approximately 2.8 km north of Hoovers Gap exit (approximately 15 km southeast of
Murphreesboro), Rutherford County, Tennessee.
Description. This specimen consists of a number of articulated ambulacral plates, visible only
from the interior. The plates possess medial constrictions, such that it appears that the single pore
consists of two merged pores.
Remarks. Guensburg chose to treat this specimen as ?Unibothriocidaris sp. Given, however,
that the ambulacral plates only display a single pore, and that the hexagonal, tessellate, plates
indicate that this specimen is a bothriocidaroid, we choose to treat it more confidently as
Unibothriocidaris sp. Guensburg (1984) chose to assign this questionably to Unibothriocidaris
on the basis of the lack of visible exterior plates, and the visible medial constriction of the pore-
pair, however, this feature is also present in Unibothriocidaris sp. described by Kolata (1975)
and in U. kieri. We thus treat this material as a species of Unibothriocidaris.
Bothrioicidaroidea? indet.
1990 Bothriocidaris sp. aff. eichwaldi Männil; Schallreuter: 1990, 126, pl. 3, fig. 1.
62
Material. The specimen is stored in the paleontological collection of the GBA.
Occurrence. This specimen is known from the Hirnantian Plöcken Formation of the Carnic
Alps, Austria.
Description. This taxon was described from a single disarticulated, fragmentary putative
ambulacral plate. The plate bears two perforate protrusions towards one end, which Schallreuter
interpreted to be tubercles. These protrusions are unequal in size, with the rightmost larger than
the leftmost and the protrusions appear to be connected by a narrow raised area.
Remarks. Schallreuter (1990) noted that this specimen has two raised, perforate protrusions,
which were widely spaced apart. Because these two protrusions were widely spaced apart, like
the tubercles of B. eichwaldi, he referred to this specimen as Bothriocidaris sp. aff. eichwaldi
Männil. However, the state of preservation of this single disarticulated plate is such that it is not
definitely bothriocidaroid, or echinoid, as there is no evidence of a pore or pore-pairs on the
plate. The state of preservation of the two protrusions is also such that they cannot definitively be
identified as tubercles. As the state of preservation of this specimen precludes its assignment to
any bothriocidaroid genus with certainty, we choose to treat it as Bothrioicidaroidea? indet.
Bothriocidaroidea? Indet
1994 Bothriocidaris? sp.: Pisera, 298, pl. 61, fig. 16, pl. 62, fig 6.
63
Material. ZPAL E.IX/MA-82 Pisera 1994: pl. 62: 6.
Occurrence. Samples MA-71 and MA-82 of Pisera (1994), from the Amorphognathus tvaernis
Zone from Skala hill near Mojcza village south of Kielce in the Holy Cross Mountains, southern
Poland.
Description. This taxon consists of two disarticulated plates with small, raised circular
protrusions which are morphologically similar to the perforate tubercles of bothriocidaroids. The
first plate bears two of these raised tubercles while the other plate contains three of these raised
tubercles arranged in a very slight arc and numerous smaller bumps which may be secondary
tubercles or milliaries. The margins of both plates are irregular and may not reflect their original
shape.
Remarks. Though Pisera (1994) initially tentatively assigned this material to interambulacral
plates of Bothriocidaris, they could just as well be the interambulacral or perforate ambulacral
plates of any other bothriocidaroid. For instance, the perforate ambulacral plates of N. peculiaris
bear both primary and secondary tubercles (Paul, 1967) and species of Unibothriocidaris bear
secondary tubercles on their perforate ambulacral plates. Furthermore, given the stratigraphic
occurrences of these specimens, and the uncertainty surrounding the shape of the plates, they
could also belong to another, as of yet undescribed clade of basal eleutherozoans or echinoderms,
and are not definitely bothriocidaroid in nature.
64
RESULTS
Diversity of Ordovician Bothriocidaroids
The database of bothriocidaroid occurrences was tabulated to produce estimates of the number of
bothriocidaroid species known from all stages of the Ordovician. The raw diversity of
Ordovician bothriocidaroids, compared to that of other echinoids is shown in Figure 2.1. The
first occurrence of any bothriocidaroids, or echinoids, is in the Dariwillian, however peak
diversity is in the Sandbian and Katian. There is only a single occurrence of a bothriocidaroid or
echinoid from the Hirnantian.
Phylogenetic Analyses
Trees presented resulting from parsimony analyses are strict consensus trees, which show all
clades present in all most parsimonious trees. Bayesian trees shown are 50% majority rule
consensus trees. These summarize the posterior distribution of trees by showing all clades that
occurred in more than 50% of trees in the posterior distribution, and thus which have
probabilities of greater than 0.5 of being true clades given the model, priors, and data (e.g. Larget
and Simon, 1999; Yang, 2014). These have recently been shown to be the most accurate
summarization method for posterior distributions of trees generated from discrete morphological
data (O’Reilly and Donoghue, 2017).
Inclusive dataset
Following the initial analysis of all taxa using all twenty-six characters yielded 1260 most
parsimonious trees, all of length 50 steps with CI of .820 and RI equal to .827. The strict
consensus of these trees with bootstrap support is shown in Figure 2.9A, and bootstrap
proportions were evaluated using 10000 “fast bootstraps”. This tree contains one large polytomy
65
at the most basal node in the ingroup. Originating from this polytomy are branches leading to
each species of Bothriocidaris except for B. maquoketensis and B. vulcani, which form a clade, a
branch leading to this clade of B. maquoketensis and B. vulcani, and a clade consisting of all
species of Neobothriocidaris and Unibothriocidaris. The clade with species of Unibothriocidaris
and Neobothriocidaris is well supported (bootstrap proportions=82%) and within this clade all
species of Neobothriocidaris form a well-supported clade. The sister pairings of B.
maquoketensis and B. vulcani and N. minor and N. tempeltoni have bootstrap proportions lower
than 70%, and are thus not particularly well-supported by the data (Hillis and Bull, 1993).
Reweighting by the CI and rerunning the analysis resulted in 987 most parsimonious trees of
length 42.33333. Using the RI and applying the same tree searching parameters resulted in 462
most parsimonious trees of length 42.86667 while reweighting by the RC resulted in 462 most
parsimonious trees of length 38.46667. Bootstrap proportions were evaluated using 10000 “fast
bootstraps” and characters reweighted by their retention indices (WTS=Simple in PAUP*) which
is shown in Figure 2.9B. The strict consensus tree of each of these reweighted analyses
independently shows a tree with topology identical to that from the unweighted analysis, except
that two additional clades were resolved; one that includes all species of Bothriocidaris, and one
which contains all species of Unibothriocidaris and Neobothriocidaris. Reweighting by the RI
also resulted in high bootstrap support for the clades of Unibothriocidaris+Neoothriocidaris
(89.57%), and Neobothriocidaris (87.90%) (Fig. 2.9B).
Like our parsimony analyses, Bayesian analyses were not particularly well resolved. 50%
majority rule trees generated from the posterior distribution when the α parameter of the
symmetrical discrete beta distribution was set to infinity (Fig. 2.9C) show five resolved clades.
The best supported of these clades (PP=0.82) consists of both species of Unibothriocidaris.
66
Additionally, a clade consisting of all species of Neobothriocidaris was resolved (PP=0.76) and
within this clade are sister pairings of N. peculiaris and N. pentlandensis (PP=0.57), and N.
minor and N. tempeltoni (PP=0.76). Furthermore B. maquoketensis and B. vulcani were resolved
as sister taxa (PP=0.67). Dependent upon which value of α was used in the analysis support for
resulting clades, particularly which clades were the best and the least supported, was variable
between analyses. This is evident in the resulting majority rule trees (Fig. 2.9C,D, 2.13).
Furthermore, when α was set to infinity, 10, and .05 the clade consisting of B. maquoketensis
and B. vulcani was present in the resultant 50% majority rule tree (Fig. 2.9C,D), however, when
α was 0.2, 1, and 2 the clade consisting of B. maquoketensis and B. vulcani had a PP less than
0.5, and thus wasn’t present in the resulting tree (Fig. 2.13).
Pruned Datasets
Following analyses with all taxa, two separate pruned analyses were run, one of which included
all species of Bothriocidaris and the other included all species of Unibothriocidaris and
Neobothriocidaris. Parsimony analyses of species of Bothriocidaris resulted in nineteen most
parsimonious trees, all of which were 32 steps in length and had CI of 0.875, RI of 0.667 and RC
of 0.583. The strict consensus tree of these (Fig. 2.10A) was poorly resolved and recovered a
clade of B. solemni, B. maquoketensis and B. vulcani with bootstrap proportion of 22.4%. B.
maquoketensis and B. vulcani were also resolved as a sister pair with bootstrap proportion of
65%, a clade which was present in the initial parsimony analysis including all bothriocidaroids.
When characters were reweighted by their CI, RI, and RC there were eight most parsimonious
trees of length of 28.66667, when the RI was used the length of the most parsimonious trees was
27.66667 and when characters were reweighted by their RC the length of most parsimonious
trees was 26.27778. All reweighted consensus trees displayed the same topology as was resolved
67
prior to reweighting. The strict consensus tree resolved after reweighting with the RI is shown
with bootstrap proportions in Figure 2.10B. Bayesian analyses of species of Bothriocidaris was
also not well resolved (Figs. 2.10C, 2.10D). Analyses either failed to resolve any clades with PP
greater than 0.5, (α=0.05, 0.2, 1, 2) or only resolved the clade of B. maquoketensis and B.
vulcani, albeit with low PP (0.55 when α=10 and 0.59 when α=∞; Fig 2.14). This clade was also
resolved in the Bayesian analysis of the inclusive dataset.
Parsimony analyses of the dataset consisting of species of Neobothriocidaris and
Unibothriocidaris, rooted on B. rimaporus resulted in nine most parsimonious trees of length
twenty-three steps, CI of 0.957, RI of 0.909 and RC of 0.87. The strict consensus tree of these
analyses shows that all species of Neobothriocidaris were resolved as a clade with bootstrap
proportions of 92% (Fig. 2.11A). The clade of N. minor and N. tempeltoni was resolved as a
clade with bootstrap proportions of 68%. In six of the nine most parsimonious trees, the two
species of Unibothriocidaris branch in a sequential, pectinate manner, and are paraphyletic with
respect to all species of Neobothriocidaris . In three of the nine most parsimonious trees,
separate clades of Neobothriocidaris and Unibothriocidaris were resolved. Other topological
disagreement between the most nine parsimonious trees concerns the nature of relationships of
species of Neobothriocidaris. Reweighing of characters by their RI and RC resulted in nine most
parsimonious trees with the same strict consensus topology with length twenty-one steps (Fig.
2.11B). When the CI was used, tree length was twenty-two steps. Bayesian analyses using this
dataset and under all examined possible values of α (∞, 10, 2, 1, 0.2 and 0.05) returned 50%
majority rule trees with the same topology, albeit differing support values for each clade
dependent upon α (Fig. 2.11C, D, 2.15). Initial analyses resulted in sister pairing of B. rimaporus
and a clade of species of Neobothriocidaris. This is likely the result of long branch attraction, as
68
branches leading to B. rimaporus and the clade of Neobothriocidaris were much longer than
those leading to Unibothriocidaris species. Analyses were thus re-run with B. rimaporus
specified as an outgroup. The topology of the resultant majority rule trees resolved
Neobothriocidaris and Unibothriocidaris as monophyletic clades, but was unable to resolve the
nature of the relationships between the two genera (Fig. 2.15). U. bromidensis and U. kieri were
resolved as sister taxa with PP ranging from 0.85 to 0.73 dependent upon the α value used. All
four species of Neobothriocidaris form a clade with PP ranging from 0.51 to 0.70. In majority
rule trees resulting from analyses under all utilized α values, a clade of N. peculiaris and N.
pentlandensis (PP from 0.51 to 0.57) is sister to a clade of N. tempeltoni and N. minor (PP from
0.61 to 0.69).
DISCUSSION
Phylogenetic Relationships of Bothriocidaroids
Because initial parsimony analyses using the dataset comprised of all bothriocidaroids
resolved Neobothriocidaris and Unibothriocidaris as a clade with high bootstrap support (Fig.
2.9A), and upon reweighting resolved all species of Bothriocidaris as a separate clade (Fig.
2.9B), analyses using pruned taxon and character matrices for each of these two clades were also
run. This was done to further resolve the relationships of these two clades without inclusion of
characters coded as non-applicable, which results from the absence of imperforate ambulacral
plates in Bothriocidaris and the absence or reduction of interambulacra from Neobothriocidaris
and Unibothriocidaris. Clades present in analyses of all datasets under both Bayesian and
maximum parsimony frameworks are treated as those with the highest overall support. Analyses
using pruned datasets usually failed to further resolve relationships, and only in one case (Fig.
69
2.10) were strict consensus trees of parsimony analyses more resolved in the pruned analyses
than in analyses using the dataset of all bothriocidaroids. No clades present in Bayesian majority
rule trees resulting from analyses of the pruned datasets were not already present in analyses of
the dataset including all bothriocidaroids.
All most parsimonious trees using the pruned dataset consisting only of species of
Bothriocidaris rooted on B. rimaporus resolved a clade consisting of B. maquoketensis and B.
vulcani which was also resolved from Bayesian analyses with PP=0.55 and PP=0.59 when α was
set to favor more symmetric transition rates (α=10 and α=∞, respectively). This clade was also
present in parsimony analyses using the dataset including all bothriocidaroids and was present in
majority rule consensus trees resulting from Bayesian analysis of this dataset using three of the
six values of α with PP’s ranging from 0.62 to 0.67 (Fig. 2.10, 2.14). Bayesian analyses of the
pruned dataset of Bothriocidaris species using other values of α failed to resolve any clades with
PP > 0.5, in contrast to parsimony analyses, where a clade of B. solemi, B. maquoketensis, and B.
vulcani was found in all most parsimonious trees, albeit with very low bootstrap support (Fig.
2.10). The clade of B. solemi, B. maquoketensis, and B. vulcani was present in numerous most
parsimonious trees resulting from analyses of the dataset including all bothriocidaroids, but
again, wasn’t present in more than 50% of the posterior trees resulting from Bayesian analyses
using this dataset. The resolution of the B. maquoketensis and B. vulcani clade from analyses
using both maximum parsimony and Bayesian inference lends confidence to the hypothesis that
there exists a clade of species of Bothriocidaris with multiple columns of plates in their
interambulacral areas, and that the evolution of multi-columned interambulacral areas occurred
only once within the bothriocidaroida.
70
In the pruned analysis including only species of Neobothriocidaris and Unibothriocidaris
all Bayesian analyses resolved Neobothriocidaris and Unibothriocidaris as distinct clades with
PP > 0.5, while six of the nine most parsimonious trees found Unibothriocidaris to be
paraphyletic with respect to Neobothriocidaris. Bayesian analyses of the dataset comprised of all
bothriocidaroids resolved Unibothriocidaris as a distinct clade with PP of 0.66 to 0.82, amongst
the highest clade credibility values of any clades in all analyses, and additionally resolved
Neobothriocidaris as a distinct clade with PP of 0.57 to 0.74 (Fig. 2.15). Given that three of the
most parsimonious trees resolved species of Neobothriocidaris and Unibothriocidaris as separate
clades, and that all Bayesian analyses resolved distinct clades of Bothriocidaris and
Neobothriocidaris with PP > .05 supports the hypothesis that these two genera are distinct
clades, and that neither is paraphyletic with respect to the other.
All most parsimonious trees using both the inclusive and pruned datasets found a clade of
N. tempeltoni and N. minor. This clade was also resolved in all Bayesian analyses using the
pruned dataset with PP ranging from 0.61 to 0.69, dependent upon α and in the inclusive dataset
with PP from 0.76 to 0.81 (Figs. 2.13, 2.15). Despite being present in only three of the nine most
parsimonious trees resulting from analyses of the pruned dataset, a clade of N. peculiaris and N.
pentlandensis was present in analyses of the inclusive dataset with PP of 0.55 to 0.6 and 0.51 to
0.57 using the pruned dataset. The resolution of these clades from Bayesian analyses and in a
number of the most parsimonious trees supports the hypothesis that the clade of N. minor and N.
pecularis is sister group to the clade of N. peculiaris and N. pentlandensis. A phylogenetic tree
of hypothesized bothriocidaroid relationships, based on a composite of relationships inferred
from cladistic analyses and Bayesian inference is shown in Figure 2.12. The phylogenetic tree is
calibrated to stratigraphy using the stage-level stratigraphic range of each taxon.
71
In all Bayesian analyses, the effect of α on the resultant clade credibility trees was also
responsible for fairly high variation in PP node support (Figs. 2.13-2.15). It is also worth noting
that this variation was substantial enough to change the resulting topology of 50% majority rule
trees, as some trees contained the clade of B. maquoketensis and B. vulcani, while others did not.
Interestingly, in the case of B. maquoketensis and B. vulcani α values of 10 and infinity, which
correspond to more symmetrical transition rates, resolved the same topology as when α was
equal to .05, corresponding to high asymmetry in transition rates. When α was equal to .2, 1, and
2, this clade was not present in the resultant 50% majority rule trees (Fig. 2.14) This is indicative
of a rather non-straightforward effect of α on resultant clade credibility values. Although 50%
majority rule trees show clades which have a higher probability of being present than not (PP >
0.5), 50% is still a fairly arbitrary value with respect to visualizing clades on a consensus tree.
That different values of α result in variation in PPs of clades substantial enough to change their
presence or absence in the 50% majority rule tree indicates that running analyses under different
prior values of α should be requisite part of Bayesian phylogenetic analyses of discrete
morphological data (Wright et al., 2016). Furthermore, if no particular value of α emerges as
superior from model comparison tests, paying close attention to variable clade support under
different values of α provides a measure of the robustness of results to different parameter
values.
The Early Diversification of Bothriocidaroids and Echinoids
Raw species richness of bothriocidaroid species shows a peak of diversity in the
Sandbian (Fig. 2.1). Given that their first occurrence is from the Darriwilian, and the clade is
thus known from Darriwilian to Ludfordian-aged rocks seems to show a relatively early burst of
72
diversification in the clade’s history. Although phylogenetic analyses failed to unambiguously
resolve the phylogeny of most species of Bothriocidaris, the occurrence of numerous lineages of
bothriocidaroids in the Sandbian, including clades of all three genera, implies morphological
diversification and differentiation fairly early in the evolutionary history of the Bothriocidaroida
(Fig. 2.12). Early work hypothesized that the base of the Middle Ordovician was associated with
diversification of numerous animal clades, and may have been tightly linked to the broad
Ordovician diversification of metazoans (Droser and Finnegan, 2003). Conversely, more recent
work on brachiopods (Harper et al., 2015) has shown that during the Ordovician, diversification
followed three diversity acmes. The first, in the Tremadocian–Floian, is characterised by a
diachronous event spanning paleocontinents and groups of organisms. The second took place in
the Darriwilian where global marine richness increased threefold. The final spike in diversity
took place in the Katian and was more unified across taxonomic groups and terranes. Harper et
al. (2015) have also shown that the diversity trend continued into the Silurian and Devonian,
making the GOBE the most sustained marine biodiversity pulse of the Phanerozoic. They also
noted that key factors driving and enabling diversification may have included competition
(Servais et al., 2008), diversification of phyto- and zooplankton (Servais et al., 2009; Servais et
al., 2008; Servais et al., 2010) and predation (Servais et al., 2010). Furthermore, proxy evidence
indicates that the GOBE may have been fuelled by icehouse conditions (similar to Quaternary
glaciations) at the Dapingian–Darriwilian boundary (Rasmussen et al., 2016). This was
interpreted to create new viable ecospace in the form of nutrient-rich upwelling zones, and
initiated a revolution, not only for primary producers, but for the entire trophic chain (Rasmussen
et al. 2016).
73
Echinoids are first known from the Darriwilian and clearly show higher diversity in the
Sandbian and Katian (Fig. 2.1). This indicates that diversification of echinoids appears to have
happened later in the Ordovician than other metazoan clades, and between two of the diversity
peaks noted above for brachiopods (Harper et al., 2015). Sprinkle and Guensburg (2003) noticed
that the Mohawkian (Sandbian-Katian) showed a peak in raw species richness for numerous
echinoderm clades including crinozoans, blastozoans, echinozoans and blastozoans. This trend
was further verified for blastozoans by Nardin and Lefebvre (2010) who utilizing numerous
sample-standardization techniques. The complied raw species richness for echinoids tends to
follow this same trend (Fig. 2.1). Sprinkle and Guensburg (2003) attributed this peak in
echinoderm diversity to the evolution of new morphologies and ecologies across the
Echinodermata, increased preservation potential and favorable living conditions due to high sea
level and expansive shelf area, and high endemism resulting from biogeographic separation of
echinoderm faunas during the Sandbian-Katian. That Nardin and Lefebvre’s (2010) rarified
diversity curves show a similar trend as documented by Sprinkle and Guensburg (2003) for
blastozoans suggests that differences in sampling resulting from differential preservation cannot
alone explain this Sandbian-Katian peak in diversity. While both Neobothriocidaris and
Bothriocidaris appear to be cosmopolitan genera, Unibothriocidaris is currently known only
from Laurentian strata during the Sandbian (Guensburg, 1984; Kier, 1982). Furthermore, during
the Katian, the species B. globulus and N. minor are known from Laurentian strata of the
Midland Valley of Scotland (Paul, 1967) and Baltic strata of Estonia and Gotland (Männil, 1962;
Schallreuter, 1989), casting doubt on high endemism as a driver of bothriocidaroid
diversification. The Middle and Upper Ordovician are characterized by amongst the highest
levels of tropical shelf area known during the Phanerozoic (Walker et al., 2002) and the
74
Sandbian and early Katian had amongst the highest sea levels of the Cambrian and Ordovician
(Haq and Schutter, 2008). These high sea levels and expansive tropical shelves likely resulted in
expanded habitats in which bothriocidaroids, and other echinoderms, were able to thrive.
Another Phanerozoic peak of tropical shelf area is in the Mississippian (Walker et al., 2002),
which has yielded the highest diversity of echinoids in the Paleozoic (Kier, 1965). That these two
intervals of high echinoid diversity, the early Carboniferous and the Late Ordovician, are
coincident with widespread tropical shelves poses an intriguing avenue for future research.
Further analytical work will be necessary to test the hypothesized link between echinoid
diversity and shelf area through precise, statistical, comparisons.
Two distinct clades of bothriocidaroids, one composed of Neobothriocidaris and
Unibothriocidaris and one composed of species of Bothriocidaris, appear to have diverged prior
to the Sandbian (Fig. 2.12). This indicates that further interrogation of the Dariwillian rock
record is likely to reveal the earlier stages of bothriocidaroid evolution and diversification. The
presence of disarticulated plates attributed to Neobothriocidaris sp. from the Darriwilian of
Poland (Pisera, 1994) is a promising first piece of the puzzle that is the Darriwilian fossil record
of bothriocidaroids. The occurrence of Neobothriocidaris sp. marks not only the oldest
bothriocidaroid occurrence, but also the oldest occurrence of any echinoid. That the taxon
Bromidechinus rimaporus from the Sandbian has been demonstrated to be the most basal
echinoid indicates that there are also likely even more basal echinoid taxa in the Darriwilian, and
provides an instance in the fossil record where the stratigraphically oldest occurrence of a clade
is not of the most basal taxon.
75
Bothriocidaroids and the Ordovician-Silurian Transition
The phylogeny and stratigraphic ranges of bothriocidaroids show a marked decrease in
raw taxonomic richness from the Ordovician to the Silurian (Fig. 2.12). This decrease in richness
may also have begun prior to the Hirnantian glaciation (Finnegan et al., 2011). Given that
Unibothriocidaris and Neobothriocidaris appear to be distinct clades (Fig 2.9, Fig. 2.11), it is
evident that the Unibothriocidaris clade, consisting of U. kieri and U. bromidensis, likely went
extinct in the Sandbian. Species of Unibothriocidaris are so far only known from the Sandbian,
and given that the data herein most strongly support that Unibothriocidaris is not paraphyletic
with respect to Neobothriocidaris, it appears that the clade comprising Unibothriocidaris went
extinct before the Katian. That any species of Unibothriocidaris have yet to be found from the
Katian, which has yielded numerous specimens of Neobothriocidaris and Bothriocidaris (e.g.
Kolata et al., 1977; Männil, 1962; Paul, 1967; Schallreuter, 1989) supports this hypothesis.
Although species of both Neobothriocidaris and Bothriocidaris are known from the
Silurian, phylogenetic analyses show that a taxon in the N. pentlandensis + N. peculiaris clade
must have crossed the Ordovician-Silurian boundary, surviving the end-Ordovician mass
extinction (Fig. 2.12). Although we have been unable to include it in our analysis, a species of
Bothriocidaris from the Ludlow of Gotland (Kutscher and Reich, 2001, 2004) likely implies that
at least one lineage of Bothriocidaris also survived the Late Ordovician mass extinction.
Furthermore, the reported, though undescribed, ten species of Nebothriocidaris from the Silurian
of Gotland (Kutscher and Reich, 2004) indicate that diversity of Neobothriocidaris in the
Silurian may have remained relatively high. The presence of N. pentlandensis in the Telychian
also indicates that Neobothriocidaris wasn’t limited to a single paleocontinent in the Silurian.
Bothriocidaris, however, appears to have undergone a marked decrease in diversity in the
76
Silurian relative to the Ordovician (Fig. 2.12). The stratigraphic distribution of Bothriocidaris
species displays their differential species richness in the Ordovician and Silurian. At least one
clade, that consisting of B. maquoketensis and B. vulcani appears to have been extant through the
Sandbian and Katian, before going extinct prior to the Silurian. There furthermore appears to be
a dearth of bothriocidaroids recorded from the Hirnantian relative to the Sandbian and Katian,
though, the short duration of the Hirnantian, and lesser preserved Laurentian strata in the
Hirnantian compared to the Katian and Sandbian (Finnegan et al., 2012) could explain this
pattern. The Late Ordovician mass extinction is associated with vast Gondwannan glaciation, and
a resultant fall in sea level (Finnegan et al., 2011; Finnegan et al., 2012). A significant portion of
extinction amongst metazoans has been attributed to climatic cooling and reduction in habitat as
opposed to simply stratigraphic truncation related to this sea level fall (Finnegan et al., 2012).
This provides an example of the so called “common cause hypothesis” whereby both taxonomic
diversity and the extent of preservable rock strata are both driven by changes in sea level (Peters
and Foote, 2002). Given the high raw diversity of echinoids in the Sandbian and Katian, and
reduction in diversity thereafter (Fig. 2.1, Fig. 2.12), it seems probable that the reduction in
bothriocidaroid diversity from the Katian to Hirnantian may have been due to reduction in
habitat, or aversion to the cooler temperatures resulting from the glaciation. At the very least two
clades of bothriocidaroids appear to have survived the Late Ordovician mass extinction, and
gone on to populate Silurian seas. Nevertheless, the biotic and abiotic crises prevalent during the
Hirnantian do appear to have altered clade dynamics of bothriocidaroids, and may have been
responsible for the differential species richness of bothriocidaroids in the Ordovician and
Silurian.
77
Biogeography of Silurian Bothriocidaroids
The paleobiogeographic distribution of bothriocidaroids in the Ordovician has recently
been documented (Lefebvre et al. 2013), however, in the Silurian, their biogeographic
distribution is poorly understood. Prior to this study, the only documented occurrences of
bothriocidaroids were from the Silurian of Gotland, Sweden (Franzén, 1979; Kutscher and
Reich, 2001, 2004). In Gotland, bothriocidaroids are known from Llandovery through Ludlow
strata, and are known from at least two genera, Bothriocidaris, and Neobothriocidaris.
Unibothriocidaris is currently unknown from the Silurian, and furthermore has not been
recorded from outside of Laurentia in the Ordovician (Guensburg, 1984; Kier, 1982; Kolata,
1975). The Silurian strata from Gotland represent oceans of the paleocontinent Baltica (Torsvik
et al., 1992; Trench and Torsvik, 1991) while the strata of the North Esk Inlier and the Deerhope
Formation were deposited along the southeastern margin of Laurentia in the closing Iapetus
ocean (Armstrong and Owen, 2001; Torsvik et al., 1996). The new specimen of
Neobothriocidaris pentlandensis. thus indicates that the biogeographic range of
Neobothriocidaris in the Silurian spanned from at least Baltica to Laurentia. Populations of
species of Neobothriocidaris thus apparently survived the end-Ordovician mass extinction in
both of these paleocontinents, or underwent secondary migrations to one or both continents
following the extinction event. Because documentation of the fossil record of Silurian
bothriocidaroids is a work in progress (Kutscher and Reich, 2001, 2004), comparison of
Nebothriocidaris pentlandensis with still undescribed taxa from Gotland (Kutscher and Reich,
2001, 2004; Reich and Smith, 2009) will help to further resolve the details of Silurian
bothriocidaroid biogeography at the species level. Nebothriocidaris pentlandensis is recorded
from the ‘Starfish Beds’ in the lower Deerhope Formation in the Pentland Hills. This formation
78
is also known to yield numerous specimens of the echinoid Aptilechinus caledonensis (Kier,
1973). Reich and Smith (2009) and Franzén (1979) also report co-occurrence of
Neobothriocidaris and Aptilechinus from the Silurian strata of Gotland. That these two taxa are
known to occur in the same strata in both the Pentland Hills, and on Gotland indicates similarity
at the genus level of these Laurentian and Baltic echinoid assemblages.
CONCLUSIONS
The diversification of bothriocidaroids represents the initial diversification of echinoids,
occurring concomitantly with the rise of other echinoderms in the Late Ordovician (Figs. 1, 8;
Nardin and Lefebvre, 2010; Sprinkle and Guensburg, 2003). All three bothriocidaroid genera
appear to be clades, with Bothriocidaris being the sister group to a clade composed of
Neobothriocidaris and Unibothriocidaris. Unibothriocidaris and Neobothriocidaris appear to
have been sister clades, and represent two monophyletic groups, while the species level
relationships of Bothriocidaris are still largely unresolved. Phylogenetic analyses inform our
understanding of bothriocidaroid macroevolutionary dynamics, and though peak bothriocidaroid
diversity is in the Sandbian and Katian (Fig. 2.1), Unibothriocidaris, which is known only from
Laurentian strata, appears to have gone extinct prior to the Katian, while lineages of both
Neobothriocidaris and Bothriocidaris survived the Late Ordovician mass extinction.
Neobothriocidaris and Bothriocidaris went on to inhabit oceans of the paleocontinent Baltica in
the Silurian, with the Silurian bothriocidaroid record in this paleocontinent spanning Telychian
to Ludfordian (Kutscher and Reich, 2004; Reich and Smith, 2009). The new species,
Neobothriocidaris pentlandensis n sp., shows that the biogeographic range of
Nebobothriocidaris spanned from at least Baltica to Laurentia during the Telychian.
79
Furthermore, the faunal composition of Silurian echinoid localities in Laurentia and Baltica
appears to have been similar, at least at the genus level.
Figure 2.1. Number of species of Bothriocidaris known from each stage of the Ordovician
compared to that of other echinoids. Species named in open nomenclature were only included in
counts when there were no other named species within the same genus for a given stage. N.
templetoni is included in the Sandbian, though it could also be Katian in age. Bothriocidaroid
richness shown in black bars, non-bothriocidaroid echinoids shown in white bars and grey bars
indicate total echinoid diversity. No bars are shown for non-bothriocidaroids in the Darriwilian
because no non-bothriocidaroid echinoids have been collected from these strata, and likewise no
definitive echinoids are known from the Hirnantian. Age dates are from Cooper and Sadler
(2012).
80
467.3
470
Hirnantian
Katian Sandbian
Darriwilian
Late
Middle
Ordovician
458.4 453.0 445.2 443.8
ages
0 2 4 6 8 10
Number of species
Age (Ma)
Bothriocidaroids
Non-bothriocidaroids
Total Echinoids
81
Figure 2.2. Location map of the Pentland Hills in Scotland, and simplified geological map of the
North Esk Inlier (after Anderson et al. 2007), showing the Silurian rocks in relation to Devonian
and Carboniferous rocks. Position of the boundary between the Reservoir and Deerhope
formations after Anderson et al. (2007). Snowflake marks the locality where the specimen was
likely collected.
82
83
Figure 2.3. Different test structure arrangements found in the Bothriocidaroida modified from
Smith and Savill (2001). (A) The test structure of most species of Bothriocidaris, where there is
a single column of interambulacral plates flanked by two columns of ambulacral plates in each
adjacent area. (B) Plating arrangement of Bothriocidaris vulcani, where the ambulacral areas
consist of two columns of plates and interambulacra are arranged into columns of a variable
number. (C) Plating structure of Unibothriocidaris, where there are numerous (in this case eight)
columns of ambulacral plates with each half-ambulacrum separated by a single column of
imperforate, perradial ambulacral plates, and there are no interambulacral plates. (D) Plating
structure of Neobothriocidaris, where there are numerous columns of ambulacral plates in each
ambulacrum and each half-ambulacrum is separated by a single column of perradial plates.
Interambulacral plates are absent or reduced. Perforate ambulacral plates are white, perradial
ambulacral plates are dotted and interambulacral plates are shown in gray.
84
A.
B.
C.
D.
85
Figure 2.4. Plate diagram of specimen MBE 2329 of Bothriocidaris globulus showing the
arrangement of ambulacral and interambulacral plates and ocular plates. Modified from Jackson
(1912) and details of ambulacral pores and tuberculation have been removed. Each ambulacra
consists of two columns of plates, while interambulacral plates are arranged into a single column
in each area. Ambuacral plates are white, interambulacral plates are shown in gray.
86
87
Figure 2.5. Photographs of type specimens of Unibothriocidaris bromidensis Kier, 1982 (A-B)
and Bothriocidaris kolatai Kier, 1982 (C-D). (A) Specimen TMM 1122TX.1, holotype of U.
bromidensis. Note the single pore per each perforate ambulacral plate. (B) View of other side of
the same specimen. Note the single spines associated with each perforate primary tubercle and
the small sparse granules on each perforate ambulacral plate. (C) Specimen TMM 1122TX.57,
holotype of B. kolatai. (D) View of other side of same specimen. Note the multiple primary
tubercles on ambulacral plates and the lack of a well-developed peripodial ring. All scale bars are
5 mm.
88
A. B.
C. D.
89
Figure 2.6. Photographs and camera lucida drawings of casts made of specimen NHMUK E
42523 holotype of Neobothriocidaris peculiaris Paul, 1967. (A) View of disarticulated and semi-
articulated test interior of specimen NHMUK E 42523. In the center of the photo the impression
of the radial water vessel can be seen with ambulacral plates branching off from this en chevra.
Note that ambulacral plates bear a single perforation entirely within each plate and one or two
others along the more perradial or adradial sutures between plates. (B) Photograph of mould of
specimen NHMUK E 42523, showing the disarticulated and semi-articulated exterior of the test.
In the upper right-hand corner the column of non-perforate ambulacral plates, and flanking rows
of ambulacra arranged en chevra can be seen. Note the perforate primary tubercles on the
column of imperforate ambulacral plates and NHMUK E 42523. Peripodial rings on ambulacral
plates also span across two plates. (C) View of external surface of ambulacral plates of specimen
NHMUK E 42523. (D) View of internal surface of ambulacral plates and a few ambulacral
plates showing the external surface. In the bottom-center of the plates note the external surface of
ambulacral plates showing the peripodial ring, pore-pairs and primary tubercle. (E) Camera
lucida drawing of ambulacral plates of specimen NHMUK E 42523, showing the details of the
tuberculation, and peripodial ring of N. peculiaris. Compare with (D). (F) Camera lucida of
single ambulacral plate showing details of plate sculpturing and tuberculation. (G) Camera lucida
drawing of non-perforate ambulacral plate of NHMUK E 42523 showing details of
tuberculation. Compare with (B). Scale bars in A and D-G are 1 mm, bars in B and C are 5 mm.
90
E. F. G.
A. B.
C. D.
91
Figure 2.7. Photographs of type specimens of Neobothriocidaris pentlandensis n. sp. (A) Interior
view of specimen NMS G.1976.48.1, holotype of N. pentlandensis. Compare with Figure 8D.
The absence of interambulacral plates can clearly be seen between adradial columns in each
ambulacral area. (B) Counterpart to (B) showing exterior view of test of specimen NMS
G.1976.48.1, of N. pentlandensis. Scale bar is 5 mm.
92
A. B.
93
Figure 2.8. Micro-CT renderings of the holotype of Neobothriocidaris pentlandensis n sp.,
specimen NMS G.1976.48.1. (A) Test and spines of specimen NMS G.1976.48.1. The view
shown is an exterior view of the test surface, made from the rendering of an external mould. (B)
Same as A, but with spines removed. Note the tubercles and peripodial rings. (C) View of the
interior of the test and spines. Test plating is jumbled, resulting from differential disarticulation
of the test after death. (D) Same view as in (C), but with disarticulated plates removed to see
interior surface of the test. Note the absence of interambulacral plates between ambulacral areas.
(E) Jumble of disarticulated plates separated from the articulated portions of the test and spines.
Note the primary tubercles and plate sculpturing. Furthermore, lateral water vessel canals are
present. (F) Close up lateral view of a single ambulacral plate. Note the primary and secondary
tubercles. (G) Plan view of the same plate as shown in F. Note the primary and secondary
tubercles, peripodial ring, and plate sculpturing. Scale bars in A-E are 1 mm. Scale bar in F and
G is 0.5 mm. Test plating is shown in purple, while spines are shown in red.
94
A. B.
C. D.
E. F. G.
95
Figure 2.9. Strict consensus of cladistic analyses using maximum parsimony and 50% majority
rule consensus trees resulting from Bayesian inference on the dataset of all bothriocidaroids
(inclusive dataset). Parsimony analyses were rooted on the outgroup Bromidechinus rimaporus.
(A) Strict consensus of 1260 most parsimonious trees when analyses were run with characters
equally weighted. Tree length is 50 steps with CI=0.820, RI=0.827 and RC=0.678. (B) Strict
consensus of 462 most parsimonious trees when analyses were run with characters reweighted by
their retention indices. Tree length is 42.86667 steps with RI=0.879, CI=0.873, and RC=0.768.
(C) 50% Majority rule consensus of posterior distribution of trees from Bayesian analyses when
α was set to ∞, which corresponds to more symmetrical rates of character transitions. (D) 50%
Majority rule consensus of posterior distribution of trees from Bayesian analyses when α was set
to 0.05, corresponding to higher asymmetry in character transition rates. Bootstrap proportions
shown on cladistic trees derived from 10,000 “fast” bootstrap replicates. PP of nodes present in
the 50% Majority rule consensus trees are shown in bold, while branch lengths are not.
96
A. B.
Bothriocidaris
parvus
Bothriocidaris
maquoketensis
Bothriocidaris
vulcani
Bothriocidaris
globulus
Neobothriocidaris
minor
Unibothriocidaris
kieri
Bothriocidaris
pahleni
Unibothriocidaris
bromidensis
Neobothriocidaris
tempeltoni
Neobothriocidaris
pendeltoni
Bromidechinus
rimaporus
Neobothriocidaris
peculiaris
Bothriocidaris
kolatai
Bothriocidaris
solemi
Bothriocidaris
eichwaldi
82.46
84.35
66.95
56.66
Neobothriocidaris
peculiaris
Bothriocidaris
kolatai
Bromidechinus
rimaporus
Bothriocidaris
parvus
Bothriocidaris
eichwaldi
Neobothriocidaris
tempeltoni
Bothriocidaris
vulcani
Neobothriocidaris
minor
Bothriocidaris
pahleni
Bothriocidaris
solemi
Bothriocidaris
globulus
Unibothriocidaris
kieri
Bothriocidaris
maquoketensis
Unibothriocidaris
bromidensis
Neobothriocidaris
pentlandensis
87.9
89.57
69.29
66.85
Neobothriocidaris minor
Bothriocidaris eichwaldi
Bothriocidaris maquoketensis
Bothriocidaris globulus
Neobothriocidaris peculiaris
Unibothriocidaris bromidensis
Bromidechinus
rimaporus
Bothriocidaris solemi
Bothriocidaris parvus
Bothriocidaris vulcani
Bothriocidaris pahleni
Neobothriocidaris pentlandensis
Unibothriocidaris kieri
Bothriocidaris kolatai
Neobothriocidaris tempeltoni
0.7618
0.8259
0.6735
0.5639
0.7479
0.0391
0.173
0.0434
0.0669
0.0658
0.6517
0.2536
0.0721
0.1428
1.6391
0.0718
0.1057
0.076
0.0511
0.0519
0.0907
0.1543
0.8399
0.0409
0.3946
α=∞
α=0.05
Bothriocidaris parvus
Bothriocidaris maquoketensis
Bothriocidaris solemi
Bothriocidaris vulcani
Unibothriocidaris bromidensis
Neobothriocidaris tempeltoni
Bothriocidaris pahleni
Unibothriocidaris kieri
Bromidechinus
rimaporus
Neobothriocidaris sp.
Bothriocidaris kolatai
Neobothriocidaris peculiaris
Bothriocidaris eichwaldi
Bothriocidaris globulus
Neobothriocidaris minor
0.6505
0.5514
0.7682
0.7277
0.8162
0.1265
0.0748
0.7301
0.1578
0.5432
0.08
0.0992
0.0362
0.0706
0.0599
0.0842
0.2457
0.0446
1.5644
0.1387
0.3523
0.0606
0.0502
0.0653
0.0332
C. D.
97
Figure 2.10. Strict consensus of cladistic analyses using maximum parsimony and 50% majority
rule consensus trees resulting from Bayesian inference on the dataset consisting of all species of
Bothriocidaris. Parsimony analyses were rooted on the outgroup Bromidechinus rimaporus. (A)
Strict consensus of nineteen most parsimonious trees when analyses were run with characters
equally weighted. Tree length is 32 steps with RI=0.667, CI=0.875, and RC=0.583. (B) Strict
consensus of eight most parsimonious trees when analyses were run with characters reweighted
by their retention indices. Tree length is 27.66667 steps with RI=0.796, CI=0.934, and
RC=0.744. (C) 50% Majority rule consensus of posterior distribution of trees from Bayesian
analyses when α was set to ∞, which corresponds to more symmetrical rates of character
transitions. (D) 50% Majority rule consensus of posterior distribution of trees from Bayesian
analyses when α was set to 0.05, corresponding to higher asymmetry in character transition rates.
Bootstrap proportions shown on cladistic trees derived from 10,000 “fast” bootstrap replicates.
PP of nodes present in the 50% Majority rule consensus trees are shown in bold, while branch
lengths are not.
98
Bothriocidaris
kolatai
Bothriocidaris
pahleni
Bothriocidaris
eichwaldi
Bothriocidaris
globulus
Bothriocidaris
maquoketensis
Bothriocidaris
parvus
Bromidechinus
rimaporus
Bothriocidaris
solemi
Bothriocidaris
vulcani
64.4
22.4
Bothriocidaris
kolatai
Bothriocidaris
pahleni
Bothriocidaris
eichwaldi
Bothriocidaris
globulus
Bothriocidaris
maquoketensis
Bothriocidaris
parvus
Bromidechinus
rimaporus
Bothriocidaris
solemi
Bothriocidaris
vulcani
71.2
23.9
A. B.
Bothriocidaris parvus
Bothriocidaris kolatai
Bothriocidaris maquoketensis
Bothriocidaris pahleni
Bothriocidaris solemi
Bothriocidaris eichwaldi
Bothriocidaris globulus
Bromidechinus
rimaporus
Bothriocidaris vulcani
0.5997
0.1048
0.3972
0.6296
0.264
0.0646
0.1098
0.172
0.0554
0.0885
1.8181
α=∞
Bothriocidaris parvus
Bothriocidaris maquoketensis
Bothriocidaris kolatai
Bothriocidaris eichwaldi
Bothriocidaris vulcani
Bothriocidaris pahleni
Bromidechinus
rimaporus
Bothriocidaris solemi
Bothriocidaris globulus
0.285
0.6392
0.9567
0.1162
0.4911
0.1453
1.2669
0.35
0.2124
α=1
C. D.
99
Figure 2.11. Strict consensus of cladistic analyses using maximum parsimony and 50% majority
rule consensus trees resulting from Bayesian inference on the dataset consisting of species of
Neobothriocidaris and Unibothriocidaris. Parsimony analyses rooted on the outgroup
Bromidechinus rimaporus. (A) Strict consensus of nine most parsimonious trees when analyses
were run with characters equally weighted. Tree length is 23 steps with RI=0.909, CI=0.957, and
RC=0.870. (B) Strict consensus of nine most parsimonious trees when analyses were run with
characters reweighted by their retention indices. Tree length is twenty-one steps with RI=1,
CI=1, and RC=1. Bootstrap proportions derived from 10,000 “fast” bootstrap replicates. (C) 50%
Majority rule consensus of posterior distribution of trees from Bayesian analyses when α was set
to ∞, which corresponds to more symmetrical rates of character transitions. (D) 50% Majority
rule consensus of posterior distribution of trees from Bayesian analyses when α was set to 0.05,
corresponding to higher asymmetry in character transition rates. Bootstrap proportions shown on
cladistic trees derived from 10,000 “fast” bootstrap replicates. PP of nodes present in the 50%
Majority rule consensus trees are shown in bold, while branch lengths are not.
100
Unibothriocidaris
bromidensis
Neobothriocidaris
minor
Bromidechinus
rimaporus
Neobothriocidaris
pentlandensis
Unibothriocidaris
kieri
Neobothriocidaris
tempeltoni
Neobothriocidaris
peculiaris
68
92
Unibothriocidaris
bromidensis
Neobothriocidaris
minor
Bromidechinus
rimaporus
Neobothriocidaris
pentlandensis
Unibothriocidaris
kieri
Neobothriocidaris
tempeltoni
Neobothriocidaris
peculiaris
65
98
A. B.
Unibothriocidaris kieri
Neobothriocidaris tempeltoni
Neobothriocidaris pentlandensis
Bromidechinus
rimaporus
Unibothriocidaris bromidensis
Neobothriocidaris minor
Neobothriocidaris peculiaris
0.7024
0.5586
0.849
0.6941
0.0919
0.0479
0.0474
0.0712
1.8007
0.76
0.1505
0.0495
0.0693
0.1537
0.7044
α=∞
Neobothriocidaris tempeltoni
Bromidechinus
rimaporus
Neobothriocidaris peculiaris
Unibothriocidaris bromidensis
Neobothriocidaris minor
Unibothriocidaris kieri
Neobothriocidaris pentlandensis
0.518
0.7032
0.8537
0.6441
0.043
0.1252
1.697
0.0736
0.064
0.0443
0.651
0.5852
0.0754
0.0498
0.1269
α=0.05
C. D.
101
Figure 2.12. Stratigraphic ranges and hypothesized phylogenetic relationships of bothriocidaroid
echinoids in the Ordovician and Silurian. Top shows stratigraphic ranges of select
bothriocidaroids which were not included in our phylogenetic analyses. These species are either
described only from disarticulated coronal plates (Neobothriocidaris sp.) or not yet properly
described (Neobothriocidaris sp., Bothriocidaris sp.). Other bothriocidaroid taxa described in
open nomenclature have been excluded if they are known from the same stage as a named
bothriocidaroid species. Bottom shows Phylogenetic tree of hypothesized bothriocidaroid
relationships based on a composite of relationships inferred in Figures 9-11 plotted with the
stratigraphic range of each taxon at the stage level. The polytomy of species of Bothriocidaris
implies lack of phylogenetic resolution as opposed to a multifurcating evolutionary event. Figure
was created using the STRAP package in R (Bell and Lloyd, 2015).
102
420
430
440
450
460
Dapingian
Darriwilian
Sandbian
Katian
Hirnantian
Rhuddanian
Aeronian
Telychian
Sheinwoodian
Homerian
Gorstian
Ludfordian
Lochkovian
Middle Upper
Llandovery Wenlock Ludlow Pridoli
Lower
Ordovician Silurian De
Bothriocidaris pahleni
Bothriocidaris parvus
Bothriocidaris eichwaldi
Bothriocidaris globulus
Bothriocidaris solemi
Bothriocidaris maquoketensis
Bothriocidaris vulcani
Bothriocidaris kolatai
Unibothriocidaris bromidensis
Unibothriocidaris kieri
Neobothriocidaris peculiaris
Neobothriocidaris pentlandensis
Neobothriocidaris minor
Neobothriocidaris templetoni
Bromidechinus rimaporus
Neoothriocidaris sp.
Bothriocidaris sp.
Neobothriocidaris sp.
103
Figure 2.13. 50% majority rule consensus trees for Bayesian analyses of inclusive dataset, which
include all named species of bothriocidaroids. All analyses were run using the MK model
(Lewis, 2001) with priors of symmetric discrete beta distributions with six different values of the
shape parameter α (Wright et al. 2016). (A) 50% majority rule consensus tree when α was set to
∞. (B) 50% majority rule consensus tree when α was set to 10. (C) 50% majority rule consensus
tree when α was set to 2. (D) 50% majority rule consensus tree when α was set to 1. (E) 50%
majority rule consensus tree when α was set to 0.2. (F) 50% majority rule consensus tree when α
was set to 0.05. Clade credibility values for resolved nodes (Posterior probabilities) shown in
bold, branch lengths not bolded.
104
Neobothriocidaris minor
Bothriocidaris eichwaldi
Bothriocidaris maquoketensis
Bothriocidaris globulus
Neobothriocidaris peculiaris
Unibothriocidaris bromidensis
Bromidechinus
rimaporus
Bothriocidaris solemi
Bothriocidaris parvus
Bothriocidaris vulcani
Bothriocidaris pahleni
Neobothriocidaris pentlandensis
Unibothriocidaris kieri
Bothriocidaris kolatai
Neobothriocidaris tempeltoni
0.7618
0.8259
0.6735
0.5639
0.7479
0.0391
0.173
0.0434
0.0669
0.0658
0.6517
0.2536
0.0721
0.1428
1.6391
0.0718
0.1057
0.076
0.0511
0.0519
0.0907
0.1543
0.8399
0.0409
0.3946
A. B.
C. D.
E. F.
α=10
α=2 α=1
α=0.2 α=0.05
α=∞
Bothriocidaris eichwaldi
Bothriocidaris maquoketensis
Bothriocidaris globulus
Neobothriocidaris minor
Neobothriocidaris sp.
Bothriocidaris parvus
Bothriocidaris pahleni
Bothriocidaris kolatai
Unibothriocidaris bromidensis
Unibothriocidaris kieri
Bromidechinus
rimaporus
Neobothriocidaris peculiaris
Bothriocidaris solemi
Neobothriocidaris tempeltoni
Bothriocidaris vulcani
0.5597
0.7357
0.8293
0.7696
0.6294
0.0477
0.154
0.1975
0.0689
0.0387
0.0538
0.0824
0.1608
0.4693
0.0546
0.9293
0.0765
0.0986
1.5782
0.0723
0.6983
0.039
0.0815
0.2867
0.1127
Bothriocidaris eichwaldi
Bothriocidaris solemi
Bothriocidaris pahleni
Unibothriocidaris bromidensis
Neobothriocidaris minor
Neobothriocidaris pentlandensis
Bothriocidaris parvus
Bromidechinus
rimaporus
Bothriocidaris kolatai
Bothriocidaris maquoketensis
Bothriocidaris globulus
Bothriocidaris vulcani
Neobothriocidaris peculiaris
Unibothriocidaris kieri
Neobothriocidaris tempeltoni
0.8054
0.6974
0.7704
0.573
0.0637
0.1211
0.0689
0.1629
0.0841
0.1611
0.0344
0.0524
0.7834
0.946
0.1125
1.3357
0.7231
0.2741
0.0935
0.1699
0.0625
0.094
0.0345
Bothriocidaris eichwaldi
Neobothriocidaris
tempeltoni
Bothriocidaris parvus
Bothriocidaris globulus
Unibothriocidaris kieri
Bromidechinus
rimaporus
Bothriocidaris pahleni
Bothriocidaris solemi
Unibothriocidaris bromidensis
Neobothriocidaris
peculiaris
Neobothriocidaris
minor
Bothriocidaris maquoketensis
Bothriocidaris vulcani
Bothriocidaris kolatai
Neobothriocidaris
pentlandensis
0.5908
0.6535
0.751
0.7971
0.1597
0.082
0.0288
0.1745
0.9359
0.1333
0.0967
1.247
0.0923
0.0917
0.1866
0.0632
0.0311
0.3269
0.2311
0.8782
0.8125
0.0499
0.1504
Neobothriocidaris
peculiaris
Bothriocidaris kolatai
Unibothriocidaris kieri
Bothriocidaris parvus
Bothriocidaris maquoketensis
Neobothriocidaris
minor
Neobothriocidaris
sp.
Bothriocidaris solemi
Bothriocidaris pahleni
Bothriocidaris globulus
Bromidechinus rimaporus
Neobothriocidaris
tempeltoni
Unibothriocidaris bromidensis
Bothriocidaris eichwaldi
Bothriocidaris vulcani
0.8195
0.6005
0.5797
0.666
0.0521
0.1224
0.1962
0.7177
0.0266
0.1789
0.2932
0.0469
0.8554
0.1664
0.1011
0.142
0.1912
0.9826
0.026
1.0441
0.1134
0.0916
0.232
Bothriocidaris parvus
Bothriocidaris maquoketensis
Bothriocidaris solemi
Bothriocidaris vulcani
Unibothriocidaris bromidensis
Neobothriocidaris tempeltoni
Bothriocidaris pahleni
Unibothriocidaris kieri
Bromidechinus
rimaporus
Neobothriocidaris sp.
Bothriocidaris kolatai
Neobothriocidaris peculiaris
Bothriocidaris eichwaldi
Bothriocidaris globulus
Neobothriocidaris minor
0.6505
0.5514
0.7682
0.7277
0.8162
0.1265
0.0748
0.7301
0.1578
0.5432
0.08
0.0992
0.0362
0.0706
0.0599
0.0842
0.2457
0.0446
1.5644
0.1387
0.3523
0.0606
0.0502
0.0653
0.0332
105
Figure 2.14. 50% majority rule consensus trees for Bayesian analyses of pruned dataset including
only species of Bothriocidaris. All analyses were run using the MK model (Lewis, 2001) with
priors of symmetric discrete beta distributions with six different values of the shape parameter α
(Wright et al. 2016). (A) 50% majority rule consensus tree when α was set to ∞. (B) 50%
majority rule consensus tree when α was set to 10. (C) 50% majority rule consensus tree when α
was set to 2. (D) 50% majority rule consensus tree when α was set to 1. (E) 50% majority rule
consensus tree when α was set to 0.2. (F) 50% majority rule consensus tree when α was set to
0.05. Clade credibility values for resolved nodes (Posterior probabilities) shown in bold, branch
lengths not bolded.
106
Bothriocidaris parvus
Bothriocidaris pahleni
Bothriocidaris globulus
Bothriocidaris maquoketensis
Bothriocidaris solemi
Bothriocidaris vulcani
Bothriocidaris kolatai
Bothriocidaris eichwaldi
Bromidechinus
rimaporus
0.0787
0.0878
0.0663
0.281
0.1646
0.2276
0.3979
0.0588
1.7095
Bothriocidaris solemi
Bothriocidaris pahleni
Bothriocidaris vulcani
Bothriocidaris maquoketensis
Bothriocidaris kolatai
Bothriocidaris eichwaldi
Bothriocidaris globulus
Bothriocidaris parvus
Bromidechinus
rimaporus
0.5771
0.3397
0.7445
0.8197
0.8572
0.2667
0.363
0.4392
1.081
Bothriocidaris parvus
Bothriocidaris maquoketensis
Bothriocidaris kolatai
Bothriocidaris eichwaldi
Bothriocidaris vulcani
Bothriocidaris pahleni
Bromidechinus
rimaporus
Bothriocidaris solemi
Bothriocidaris globulus
0.285
0.6392
0.9567
0.1162
0.4911
0.1453
1.2669
0.35
0.2124
Bothriocidaris pahleni
Bothriocidaris maquoketensis
Bothriocidaris vulcani
Bothriocidaris parvus
Bothriocidaris solemi
Bromidechinus
rimaporus
Bothriocidaris globulus
Bothriocidaris kolatai
Bothriocidaris eichwaldi
0.0947
0.4862
0.3293
0.1733
0.1999
1.469
0.1253
0.9268
0.0741
Bothriocidaris maquoketensis
Bothriocidaris vulcani
Bothriocidaris pahleni
Bothriocidaris parvus
Bothriocidaris eichwaldi
Bromidechinus
rimaporus
Bothriocidaris globulus
Bothriocidaris kolatai
Bothriocidaris solemi
0.5505
0.3025
0.199
0.0717
0.1134
0.058
1.7408
0.0876
0.705
0.1202
0.4329
Bothriocidaris parvus
Bothriocidaris kolatai
Bothriocidaris maquoketensis
Bothriocidaris pahleni
Bothriocidaris solemi
Bothriocidaris eichwaldi
Bothriocidaris globulus
Bromidechinus
rimaporus
Bothriocidaris vulcani
0.5997
0.1048
0.3972
0.6296
0.264
0.0646
0.1098
0.172
0.0554
0.0885
1.8181
A. B.
C. D.
E. F.
α=10
α=2 α=1
α=0.2 α=0.05
α=∞
107
Figure 2.15. 50% majority rule consensus trees for Bayesian analyses of pruned dataset including
only species of Neobothriocidaris and Unibothriocidaris. All analyses were run using the MK
model (Lewis, 2001) with priors of symmetric discrete beta distributions with six different values
of the shape parameter α (Wright et al. 2016). (A) 50% majority rule consensus tree when α was
set to ∞. (B) 50% majority rule consensus tree when α was set to 10. (C) 50% majority rule
consensus tree when α was set to 2. (D) 50% majority rule consensus tree when α was set to 1.
(E) 50% majority rule consensus tree when α was set to 0.2. (F) 50% majority rule consensus
tree when α was set to 0.05. Clade credibility values for resolved nodes (posterior probabilities)
shown in bold, branch lengths not bolded.
108
Unibothriocidaris kieri
Neobothriocidaris tempeltoni
Neobothriocidaris pentlandensis
Bromidechinus
rimaporus
Unibothriocidaris bromidensis
Neobothriocidaris minor
Neobothriocidaris peculiaris
0.7024
0.5586
0.849
0.6941
0.0919
0.0479
0.0474
0.0712
1.8007
0.76
0.1505
0.0495
0.0693
0.1537
0.7044
α=∞
Unibothriocidaris kieri
Neobothriocidaris pentlandensis
Bromidechinus
rimaporus
Neobothriocidaris tempeltoni
Unibothriocidaris bromidensis
Neobothriocidaris peculiaris
Neobothriocidaris minor
0.5638
0.6951
0.8553
0.6806
0.0954
0.1504
0.0478
1.7357
0.0477
0.8
0.076
0.0701
0.7429
0.0479
0.1597
α=10
Unibothriocidaris bromidensis
Neobothriocidaris minor
Bromidechinus
rimaporus
Neobothriocidaris pentlandensis
Neobothriocidaris tempeltoni
Unibothriocidaris kieri
Neobothriocidaris peculiaris
0.6699
0.8073
0.6727
0.5711
0.0825
0.0473
0.8494
1.6052
0.1694
0.8375
0.0502
0.1631
0.0479
0.101
0.0702
α=2
Neobothriocidaris minor
Neobothriocidaris tempeltoni
Bromidechinus
rimaporus
Unibothriocidaris kieri
Neobothriocidaris peculiaris
Neobothriocidaris pentlandensis
Unibothriocidaris bromidensis
0.548
0.6008
0.6513
0.7718
0.0528
0.0527
0.2167
0.9577
0.2028
1.4642
0.1107
0.0822
0.8599
0.0629
0.1023
α=1
Neobothriocidaris minor
Neobothriocidaris
tempeltoni
Neobothriocidaris
peculiaris
Bromidechinus
rimaporus
Unibothriocidaris bromidensis
Neobothriocidaris
pentlandensis
Unibothriocidaris kieri
0.6194
0.7376
0.5183
0.5144
0.066
0.2915
0.0658
0.1165
1.3497
0.8747
0.1195
0.3126
0.0878
0.1259
1.0235
α=0.2
Neobothriocidaris tempeltoni
Bromidechinus
rimaporus
Neobothriocidaris peculiaris
Unibothriocidaris bromidensis
Neobothriocidaris minor
Unibothriocidaris kieri
Neobothriocidaris pentlandensis
0.518
0.7032
0.8537
0.6441
0.043
0.1252
1.697
0.0736
0.064
0.0443
0.651
0.5852
0.0754
0.0498
0.1269
α=0.05
A. B.
C. D.
E. F.
109
Table 2.1. All taxa included in phylogenetic analyses. The stratigraphic range of each taxon is
listed as the stage in which the taxon is recorded, and does not indicate that the taxon is known to
range throughout the entire stage. * indicates taxon used as outgroup for parsimony analyses.
110
Species
Stratigraphic
Range
Bothriocidaris pahleni Schmidt, 1874 Sandbian
Bothriocidaris parvus, Männil, 1962 Katian
Bothriocidaris eichwaldi Männil, 1962 Katian
Bothriocidaris globulus Eichwald, 1860 Katian
Bothriocidaris solemi Kolata, 1975 Sandbian
Bothriocidaris maquoketensis Kolata, Strimple, Levorson,
1977 Katian
Bothriocidaris vulcani Guensberg, 1984 Sandbian
Bothriocidaris kolatai Kier, 1982 Sandbian
Unibothriocidaris bromidensis Kier, 1982 Sandbian
Unibothriocidaris kieri Guensberg, 1984 Sandbian
Neobothriocidaris peculiaris Paul, 1967 Katian
Neobothriocidaris minor Paul, 1967 Katian
Neobothriocidaris templetoni Kolata, 1975 Katian or Sandbian
Neobothriocidaris pentlandensis n sp. Telychian
Bromidechinus rimaporus Smith and Savill, 2001* Sandbian
111
Chapter 3. Quantitative analysis of substrate preference in
Carboniferous stem group echinoids
INTRODUCTION
Echinoids, or sea urchins, are a diverse and abundant clade that has occupied a range of
environments in both modern and ancient ecosystems. In modern oceans, echinoids are abundant
in numerous marine habitats (e.g., Kier and Grant, 1965; Nebelsick, 1996) and make up an
important part of the ecosystems in which they inhabit (Chesher, 1969; Lohrer et al., 2004;
Steneck, 2013). Though members of the modern fauna (Sepkoski, 1981), echinoids first
originated in the Ordovician (Pisera, 1994; Smith and Savill, 2001) and were moderately diverse
and abundant in the late Paleozoic (Schneider, 2008), reaching their peak specific and generic
diversities in the Carboniferous (Kier, 1965; Smith, 1984). Paleozoic echinoid faunas differed
substantially from post-Paleozoic communities, in part because all fossil echinoids known from
before the Permian belonged to the echinoid stem group (Thompson et al., 2017b; Thompson et
al., 2015b). These stem group echinoids displayed a variable number of columns of
interambulacral and ambulacral plates in each area (Fig. 3.1), as opposed to the reduced two
ambulacral and interambulacral columns that characterize the echinoid crown group (Smith,
1984). Despite having existed throughout the majority of the Paleozoic, the paleoecology and
environmental distribution of stem group echinoids is very poorly known. We set out herein to
provide a quantitative assessment of the environmental distribution of different families of stem
group echinoids during the Carboniferous.
Extant echinoids occupy different substrates dependent upon their life modes (Nebelsick,
1992, 1996; Smith, 1984) and many echinoids have evolved to live on or in particular substrates.
112
Often associated with this environmental specialization is the evolution of novel morphologies
suited for life in these environments (Carter et al., 1989; Kanazawa, 1992; Saitoh and Kanazawa,
2012; Smith, 1978a). For example, the atelostomate echinoids are specially adapted to life in
fine-grained substrates, and have evolved specialized tube feet to allow for feeding in these
sediments (Barras, 2008; Smith, 1984). Little is known about echinoid substrate affinities in the
Paleozoic, however, and to date there has been no quantitative testing of hypotheses regarding
Paleozoic echinoid environmental distribution. The Carboniferous peak in stem group echinoid
diversity saw the peak Paleozoic diversity of echinoids at the family level, and was home to six
families including the archaeocidarids, palaechinids, proterocidarids, lepidesthids, lepidocentrids
and cravenechinids. These families display a wide array of disparate morphologies, and have
been postulated to represent different life modes and ecologies based upon functional
morphology and studies of local paleoecology (Chesnut and Ettensohn, 1988; Smith, 1984;
Thompson and Ausich, 2016). Despite the information gleaned from these local studies, a large-
scale analysis of Carboniferous echinoid environmental distribution has been wanting.
Substrate Preference in Post-Paleozoic Echinoids
The affinity for particular substrates amongst post-Paleozoic echinoids has long been
demonstrated based upon the environmental distribution of certain clades, and functional
morphological studies. Numerous morphological innovations including test shape (François and
David, 2006; Kanazawa, 1992; Saitoh and Kanazawa, 2012; Schlüter, 2016), tubercle and spine
shape and structure (Smith, 1980b; Walker and Gagnon, 2014), and the size, shape, and structure
of tube feet and their associated ambulacral pores (Smith, 1978b, 1980c) can be directly linked to
environmental correlates. Furthermore, the preference for particular substrates is closely related
113
to feeding strategies and food source. Although many post-Paleozoic regular echinoids display a
tendency for hard substrates (Nebelsick, 1995, 1996), many irregular echinoids deposit feed, and
rely on nutrients in the sediment (Lohrer et al., 2004). Grain size thus plays a key role in
determining the distribution of modern echinoids (eg., Ferber and Lawrence, 1976; Gladfelter,
1978; Hendler et al., 1995; Hill and Lawrence, 2003; Ivany et al., 1994; Kier and Grant, 1965;
Mooi, 1990; Nebelsick, 1992, 1995; Pawson and Miller, 1983; Poulin and Feral, 1995; Schinner,
1993; Stanley and James, 1971; Telford et al., 1983; Telford and Mooi, 1986; Walker and
Gagnon, 2014), and numerous field-based studies of echinoid-rich Cenozoic stratigraphic
successions have shown grain-size to be a determinant in the distribution of clypeasteroids,
atelostomates, and regular echinoids (Carter et al., 1989; Carter and Hamza, 1994; Gordon and
Donovan, 1992; Kroh and Nebelsick, 2003; Mancosu and Nebelsick, 2016, 2017). Aside from
study of differential facies distributions, functional morphology has long influenced the
interpretation of fossil echinoid life modes (Kanazawa, 1992; Nichols, 1959; Saitoh and
Kanazawa, 2012; Seilacher, 1979; Smith, 1978a, 1979, 1980b). Test shape, for instance, has
been shown to control the ability of different spatangoid echinoids to burrow and locomote in or
on different substrates (Kanazawa, 1992; Saitoh and Kanazawa, 2012). Furthermore tubercle and
spine morphology and distribution have adaptive significance, and in-part control the distribution
of echinoids amongst differing grain size regimes (Ferber and Lawrence, 1976; Smith, 1978a,
1980b).
In addition to grain size, substrate mineralogy has been shown to influence the
distribution of extant echinoids. Examination of the distribution of three Atlantic species of sand
dollars revealed that Leodia sexiesperforata occurred only on biogenic carbonate sediments,
while Mellita quinquiesperforata preferred terriginous siliciclastic sediments. Furthermore,
114
Encope michelini occurred on both siliciclastic and carbonate sediments, and showed no affinity
for one over the other (Telford and Mooi, 1986). Mooi (1990) furthermore discussed that
different species of cassiduloid echinoids in the Caribbean prefer substrates of differing
mineralogies, with Cassidulus caribaearum, Echinolampas depressa, and Conolampas sigsbei
preferring carbonate substrates, while Eurhodia relicta was found in association with siliciclastic
sediments. Despite this apparent preference in some species, there are counterexamples (Guidetti
et al., 2004), and though there is much less evidence for mineralogical preference in echinoids
than there is with respect to grain size, analyses of the fossil records of marine animals have
demonstrated that clear mineralogical preferences do exist throughout the Phanerozoic (Foote,
2006, 2014; Kiessling and Aberhan, 2007; Kiessling et al., 2007; Miller and Connolly, 2001).
Previous Hypotheses of Carboniferous Echinoid Environmental Distribution
Though there has been no rigorous examination of Paleozoic echinoid substrate affinities,
previous workers have speculated regarding their environmental preference based upon
functional morphology and local studies of distribution. Functional morphological interpretations
have long influenced interpretations of substrate affinity in the family Proterocidaridae.
Proterocidarid echinoids are characterized by their enlarged oral ambulacral pores, which are
surrounded by large peripodial rings and may have supported large, penicillate tube feet (Smith,
1984). Such penicillate tube feet are found in the extant atelostomate echinoids, and have been
associated with their ability to colonize fine-grained environments (Barras, 2008; Smith, 1980b,
1984). The presumptive penicillate tubefeet of the proterocidaridae may also indicate that they
were specialized for life on fine-grained substrates. The proterocidaridae also have a large,
115
expanded oral surface (Fig. 3.1E), which bears many more columns of ambulacral plates than the
adapical surface, and was presumably associated with food-gathering (Kier, 1965).
Studies of local paleoecology have also informed hypotheses regarding stem group
echinoid environmental distribution. Chesnut and Ettensohn (1988) in their study of the
echinoderm fauna of the Serpukhovian Pennington Formation of Kentucky proposed that
Lepidesthes was an epifaunal deposit feeder with affinities for marly, soft substrates. Thompson
and Ausich (2016) found the archaeocidarid echinoid Lepidocidaris sp. in the siliciclastic green
shale facies of the Viséan Fort Payne Formation of Kentucky. This genus is also fairly abundant
in the coarse carbonate sediments of the Tournaisian Burlington Limestone (Meek and Worthen,
1869). Furthermore, in the Fort Payne Formation, the archaeocidarid, ?Archaeocidaris sp. was
found preserved on the coarse-grained crinoidal flank beds, indicating that the archaeocidarids
likely had a wide tolerance for fine- and coarse-grained, and siliciclastic and carbonate,
sediments (Thompson and Ausich, 2016). They also found lepidesthid echinoids on the firm,
packstone and wackestone carbonate buildups of the Fort Payne Formation, and showed that
despite lepidesthids’ high abundance in fine-grained, siliciclastic environments (Lane, 1973) they
were also apparently capable of inhabiting coarser carbonate substrates (Thompson and Ausich,
2016). Donovan and Lewis (2016) examined the distribution of echinoids in the Carboniferous of
Derbyshire, U.K. and also commented on the limited occurrence of Lepidocidaris in shallow
environments relative to the wider occurrence of Archaeocidaris and the palaechinid
Melonechinus in shallow and deeper-water carbonate environments.
Despite the utility of these local studies in interpreting environmental distribution, they
generally rely on limited sample sizes and short intervals of geological time. Therefore we
examined Paleozoic echinoid environmental preferences using a global database of echinoid
116
occurrences spanning the entirety of the Carboniferous. Our aim was to test the hypothesis that
different echinoid clades showed a preference for particular substrates. Our database of echinoid
occurrences consists of echinoid specimens in museum collections in North America and
Europe, and is supplemented by occurrences from the published literature. We then used this
database to determine if different families showed preferences for particular substrate
mineralogies and grain sizes using three distinct analytical metrics.
METHODS
Echinoid Occurrence Database
In order to determine substrate affinities in Paleozoic echinoids, a comprehensive
database of echinoid occurrences was assembled. The species and genera comprising the six
known families of Carboniferous echinoids are known globally and, with the probable exception
of the Palaechinidae span all stages of the Carboniferous (Jackson, 1912; Kier, 1965, 1968a).
Due to poor sampling from Asia, South America, Africa, Antarctica and Australia, however,
Carboniferous echinoids are most well known from North America and Europe. Most of the
fossil record of echinoids in the Mississippian can be attributed to a number of localities in North
America and Western Europe that preserve numerous in situ echinoderm communities (Ausich,
2016). These localities have long been sampled for their diverse and abundant crinoid “gardens”
and have provided the data sets by which numerous hypotheses regarding crinoid diversity and
abundance in the Mississippian have been tested (Ausich and Kammer, 2006; Ausich and
Kammer, 2013; Kammer and Ausich, 2006, 2007).
Our echinoid occurrence database was compiled from visits to museum collections across
the United States and Europe. Because relatively little has been published regarding
117
Carboniferous echinoids since the 1960’s, numerous unpublished specimens are known in
museum collections that have not been reported in the literature. In order to increase the number
of specimens available for statistical analyses, and to utilize the most representative dataset
available, we catalogued and photographed all Paleozoic echinoid specimens found in each
museum. In total, twenty-five museum and university collections were visited. In addition to the
name of the taxon, we recorded stratigraphic age, locality information, lithology and rock
type/grain size. Rock type/grain size was determined using associated matrix of specimens
(Carter et al., 1989; Carter and Hamza, 1994), and specimens without associated matrix, or
which appeared obviously transported or reworked, were excluded from analyses. Specimens in
carbonate matrices were classified using the Dunham classification scheme (Dunham, 1962),
while those in siliciclastic matrices were classified using the Wentworth size scale (Wentworth,
1922). Taxonomic assessments were made by JRT, as many specimen label identifications were
incorrect and/or overly specific.
We utilized only unique occurrences in our analyses, so, for example, despite the
presence of 91 specimens of Melonechinus multiporus from wackestones of the St. Louis
Limestone at St. Louis, Missouri, in our database, for analytical purposes only one of these
occurrences was included. When locality details were vague, such as “Indiana, USA” these
occurrences were excluded to avoid including duplicate occurrences with more specific locality
names. Furthermore, due to the varying nature of locality and formational name spellings, we
standardized locality and formation names. This acted to keep multiple names for the same
locality such as “Miatschkowa”, “Mjatschkova”, and “Myachkova”, all of which represent the
classic Pennsylvanian locality of the Myachkovo quarry, Russian Federation, from representing
multiple occurrences in the database. Likewise, non-english formational names and descriptors
118
such as “Bergkalk” were standardized. A taxonomic occurrence in the database was treated as
unique if the combination of the species, locality, formation and lithology or grain size was
distinct. Much of the stem group echinoid fossil record comprises specimens which are
disarticulated and thus only assignable to the taxonomic rank of genus or family (Thompson and
Denayer, 2017). Occurrences that were only identifiable to the genus or family levels were
treated as unique occurrences, unless they occurred in the same locality, formation, lithology or
grain size combination as another member of their genus or family, in which case they were
excluded to avoid counting the same occurrence twice. Our initial database included 1491
specimens, however after standardizing names and accounting for non-unique occurrences, we
were left with 568 unique occurrences for analyses. The raw database is available from the
author.
Our analyses focused on two aspects of substrate character: substrate mineralogy and
substrate grain size. Both of these have been shown to affect echinoid distribution in modern and
ancient environments (see Section 1.1). For analyses of grain size preference, all specimens were
classified as either coarse-grained or fine-grained to allow for binary comparisons using
statistical methods. For specimens in siliciclastics, those with grain sizes greater than or equal to
medium-grained following the Wentworth (1922) scale were classified as a coarse-grained
specimen, while those with grain sizes smaller than medium-grained were classified as fine-
grained specimens. For specimens with carbonate matrices, those classified as packstones or
grainstones under the Dunham (1962) scale were grouped as coarse-grained, while wackestones
and mudstones were classified as fine-grained. Further analytical methods are described below,
and code used to run all analyses and generate figures 2 and 3 is available from the author.
119
Analytical Methods
To assess substrate affinity, we used three different analytical methods, which have been
proposed in the literature over the course of the past two decades (Foote, 2006, 2014; Miller and
Connolly, 2001; Simpson and Harnik, 2009). The details of these methods are described in more
detail below. We decided to analyze our data at three different hierarchical timescales, the stage,
subperiod, and period levels. Because stratigraphic resolution of some of our occurrences is not
fine enough to assign them to a particular stage or subperiod, by running analyses at coarser
timescales we were able to include occurrences whose ages are less well-resolved and thus
increase the number of occurrences used to infer affinity.
Standardized Relative Affinity
The first metric we used is a slightly modified version of the standardized relative affinity
(SRA) developed by Miller and Connolly (2001). Their approach makes use of bootstrap
resampling to compare empirically derived affinities to distributions of randomly generated
affinities. These empirically derived affinities are:
!
=
!
!"!
( 1)
!
=
!
!"!
( 2)
Where A
C
and A
S
are the affinities for each taxon for particular substrates, in the above cases
carbonates and siliciclastics. N
C
and N
S
are the number of carbonate and siliciclastic occurrences
respectively for a particular taxon over a sampled interval of time, and
!"!
is the total number
120
of occurrences in a given time bin,
!
+
!
. A
C
and A
S
are thus equivalent to the proportion of
occurrences of a taxon in a given substrate out of the total number of times it occurs over a given
interval of time.
Our approach differed from that of Miller and Connolly (2001) due to our limited sample
size for each genus. As opposed to calculating the mean of affinities of each genus within a
family (
!
for carbonates and
!
for siliciclastics) as implemented by Miller and Connolly
(2001), we instead calculated A
C
and A
S
for each family. We then generated 1000 synthetic
“families” by randomly sampling with replacement occurrences from the entire pool of
occurrences for each time bin. Each of these “families” has the same number of constituent
occurrences as the family of interest (
!"!
) and for each randomly generated “family”, affinities,
termed RND
C
and RND
S
were calculated. These 1000 affinities from randomly sampled sets thus
provided two distributions of affinities, all of which were calculated from a sample in equivalent
size to the number of occurrences in each family over the time period of interest. The mean
affinities of these distributions
!
and
!
are then subtracted from A
C
and A
S
respectively,
and the difference between these two values is divided by the standard deviation of the
distributions of affinities,
!"#
!
or
!"#
!
. This calculation describes the SRA, which for
carbonates is shown below:
=
−
!
!"#
!
( 3)
The SRA is, as the name implies, relative, meaning that it describes the propensity for a taxon to
occur on a given substrate relative to the occurrence of all other taxa in the sampling pool in a
121
given time bin (Miller and Connolly, 2001). When interpreting the SRA, a positive value
indicates that the given taxon has a relative affinity for the substrate in question, while a negative
value indicates a relative disinclination for the substrate. An SRA of 0 indicates no or mixed
affinity. Given that we only have two categories in each of our analyses (carbonate vs.
siliciclastic or fine-grained vs. coarse-grained) reported SRAs for opposing environmental
regimes are of the same magnitude but different sign. For more information on the SRA, see
Miller and Connolly (2001). All of the calculations described above in the context of carbonate
and siliciclastic affinity are analogous to the calculations performed for comparisons of fine-
grained and coarse-grained affinity.
Binomial Test
In addition to the resampling based SRA, we additionally utilized the simple method
devised by Foote (2006) which relies upon the binomial test. This method is based upon
statistical comparison of the number of occurrences of a given taxon, again in our case a given
family, in carbonate or siliciclastic environments, and the number of carbonate or siliciclastic
collections in the Paleobiology Database (PaleobioDB) spanning the interval of interest (Foote,
2014). The number of occurrences of a taxon in a given environment may simply represent the
amount of rock units representing that environment in the rock record (Foote, 2014; Miller and
Connolly, 2001; Smith and Benson, 2013; Smith et al., 2001; Smith and McGowan, 2011). This
method thus compares the distribution of taxonomic occurrences in either of two substrates to the
proportion of environments representing one of the substrates in the rock record. The numbers of
occurrences of a given family in carbonate or siliciclastic sediments are
!
and
!
(see above),
122
while the proportions of environments that are carbonates or siliciclastics in the fossil record
over a given duration are:
!
=
!
!
+
!
( 4)
!
=
!
!
+
!
( 5)
where
!
and
!
are the number of carbonate and siliciclastic collections from the
PaleobioDB over a given duration and
!
and
!
are the proportion of carbonates and
siliciclastics present over that same duration (Foote, 2014). Binomial tests compare the
distribution of a binary variable to the proportion of “successes” in a reference distribution,
which represents the null hypothesis (Howell, 2010). Foote’s method treats the proportion of
carbonate or siliciclastic environments,
!
and
!
as the null expectation for the distribution of
occurrences in different environments. If the taxon of interest has significantly more occurrences
in one substrate than the other at the α=0.05 level using a one-tailed binomial test, it is treated as
having an affinity for that substrate. Collections extracted from the PaleobioDB and used to
formulate
!
and
!
were classified as representing either carbonate or siliciclastic
paleoenvironments. Collections with “Chert”, “Floatstone”, “Framestone”, “Rudstone”,
“Grainstone”, “Reef Rocks”, “Bindstone”, “Bafflestone”, “Dolomite”, “Lime Mudstone”,
“Limestone”, “Carbonate”, “Packstone”, or “Wackestone” in the Lithology1 field were classified
as carbonates. Collections noted as, “Silty Sandstone”, “Siltstone”, “Mudstone”, “Shale”,
123
“Sandstone”, “Claystone”, and “Conglomerate” were classified as siliciclastic collections. In
total, there were 3363 carbonate collections 2022 siliciclastic collections spanning the
Carboniferous. The supplemental file with PaleobioDB collections is available from the author.
For analyses of grain size, our analytical approach was slightly different due to the relative lack
of PaleobioDB collections with associated grain size information. To formulate
!"
and
!"
, the
proportions of fine-grained and coarse-grained occurrences, instead of using the PaleobioDB, we
compared the distribution of occurrences of the family of interest to all echinoid occurrences in
the given time bin. Binomial tests for affinity in grain size using this metric are thus a more
relative metric than for mineralogical comparisons, as the distribution of the family of interest is
not compared to a reference distribution from the PaleobioDB.
Bayesian Approach
Finally, we assessed substrate preference using the Bayesian method developed by
Simpson and Harnik (2009) and used by Hopkins et al. (2014). Like Foote’s (2006, 2014)
method, their assessment also makes use of the Binomial distribution to compare the distribution
of a clade in different substrates to a reference distribution, though it does so using a different
statistical framework. This method relies on Bayes’ Theorem to assign posterior probabilities to
each clade representing affinity for a particular substrate:
(
!
|)=
(|
!
)(
!
)
(|
!
)(
!
)+(|
!
)(
!
)
( 6)
Where (
!
|) is the posterior probability that a clade has an affinity for a given substrate given
(), the proportion of all PaleobioDB collections (or total echinoid occurrences for grain size
124
analyses) that are of that substrate.
!
and
!
are the prior probabilities that the clade
occurs in one substrate or the other, and are both set equal to 0.5, representing no
affinity.
!
and (|
!
) are the probabilities that a taxon prefers or is adverse to a
particular substrate, respectively.
!
, the probability that a taxon prefers carbonates is
calculated as shown below:
!
=
!!"
!
!
!
!
(1−
!
)
!
!"!
!!
!
( 7)
Where
!"!
,
!
, and
!
are as defined above for carbonates. Calculation of
!
is the same
as for
!
, however,
!
takes the place of
!
. Posterior probabilities are thus scaled from 1
to 0, where 1 indicates preference for a substrate, and 0 is aversion to it. All analyses were run
with
!
representing carbonates for mineralogical comparisons or fine-grained sediments for
comparisons of grain size. As for analyses using the binomial test,
!
is formulated from
PaleobioDB collections, while
!"
, the proportion of fine-grained occurrences, is based on the
total number of echinoid occurrences in a given interval of time.
RESULTS
As previously mentioned, results using the binomial test and Bayesian inference for
mineralogical comparisons are based on comparisons to PaleoDB collections, while analyses of
grain size are based on comparisons of the distribution of occurrences of a given family to all
echinoid occurrences in a given time bin. Likewise, in all analyses the SRA as calculated herein
is relative to the database of Paleozoic echinoid occurrences. Because of these differences,
results from comparisons of substrate mineralogy using the binomial test (Foote, 2006, 2014)
125
and Bayesian approach (Simpson and Harnik, 2009) must be interpreted differently from results
of the SRA (Miller and Connolly, 2001). Additionally, due to the potentially confounding effects
of small sample size, we do not report analytical results for a particular family from time bins
where there were less than five unique occurrences of that family. Results are shown in Tables
3.1-3.6 and Figures 3.2-3.3..
Binomial Test
Foote’s method (2006, 2014) uses a binomial test to compare the distribution of
taxonomic occurrences of a family to the distribution of available environments from the
PaleobioDB (mineralogy) or of all echinoid families (grain size). Taxa are deemed to have an
affinity for a particular environment if they occur significantly more often in one environment
than the other at the α=0.05 level. When this method was used to calculate substrate affinities for
each family in each stage (Table 3.1), the palaechinids were found to have statistically significant
preferences for carbonate environments in both the Tournaisian and Viséan, and no preference
for siliciclastic environments. Archaeocidarids showed a statistically significant preference for
carbonate environments in the Tournaisian, Viséan, Moscovian and Kasimovian and no
preference either way during the Serpukhovian. In the Bashkirian, archaeocidarids did show a
statistically significant preference for siliciclastic environments, and were the only clade to show
any siliciclastic preference using the binomial test during any stage. In the Tournaisian and
Kasimovian, the Proterocidaridae showed a statistically significant preference for carbonate
environments, but in the Viséan showed no statistically significant preference either way. The
lepidesthids showed a statistically significant preference for carbonates during the Serpukhovian,
but during the Viséan and Moscovian there was no significant preference for carbonates or
126
siliciclastics. During the Tournaisian, the Lepidocentridae showed no statistically significant
preference either way.
At the subperiod level (Table 3.3), no family showed a statistically significant preference
for siliciclastic environments. The palaechinids and the lepidesthids showed a statistically
significant preference for carbonate environments during the Mississippian, while the
archaeocidarids and proterocidarids showed a significant preference during both the
Mississippian and Pennsylvanian. The Lepidocentridae, however, showed no statistically
significant preference for either carbonates or siliciclastics. This same result was found when
analyses were run over the entire Carboniferous (Table 3.5), as all clades except for the
Lepidocentridae showed a statistically significant preference for carbonates and no preference
for siliciclastics.
When we tested for preference for fine- or coarse-grained environments, we found that no
statistically significant preference was found for any clade during any stage (Table 3.2). This was
markedly different at the subperiod level, however, where every clade showed a statistically
significant preference for fine-grained sediments in both the Mississippian and Pennsylvanian
(Table 3.4). When analyses were run with all Carboniferous occurrences for each taxon grouped
together, however, there was again no statistically significant preference for either fine- or
coarse-grained environments (Table 3.6).
Bayesian Approach
The Bayesian method for calculation of substrate affinity from Simpson and Harnik
(2009) assigns a posterior probability (PP) between 1 and 0 to a given taxon where 1 indicates
high affinity for a particular substrate and 0 indicates for the opposing substrate. In our
127
comparisons of carbonate versus clastic affinity, 1 indicated a high affinity for carbonates while
0 represents an aversion for carbonates, and perhaps an affinity for clastics. Likewise in
comparisons of grain size 1 corresponds to an affinity for fine-grained environments while 0
represents an aversion for fine-grained environments, and perhaps affinity for coarse-grained
environments. All results at the stage level are shown in Figure 3.2A-B, the subperiod level in
Figure 3.2C-D and the period level in 3.2E-F.
At the stage level (Fig. 3.2A), the Palaechinidae show strong affinity for carbonate
environments during the Tournaisian and Viséan, both time bins that they are present in the
Carboniferous. The Archaeocidaridae also show a strong preference for carbonate sediments in
the Tournaisian and Viséan, but additionally in the Serpukhovian and the Moscovian.
Archaeocidarids, however, do not show a strong affinity for carbonates in the Kasimovian or
Bashkirian. The Proterocidaridae show a strong affinity for carbonates in the Tournaisian and
Viséan, but also do not show an affinity for carbonates in the Kasimovian. The Lepidesthidae
show a strong affinity for carbonates during the Viséan and Serpukhovian, and a slight affinity
for carbonates during the Moscovian while the Lepidocentridae show an affinity for carbonates
in the Tournaisian. When data were analyzed for each Carboniferous subperiod, all families
showed strong preference for carbonate environments in the Mississippian (Fig. 3.2C).
Furthermore, the Archaeocidaridae, Proterocidaridae and Lepidesthidae all showed strong
affinities for carbonates in the Pennsylvanian. When taxa were analyzed at the scale of the entire
Carboniferous, all clades showed an affinity for carbonate environments (Fig. 3.2E).
When affinity for grain size was analyzed at the stage level (Fig. 3.2B), the Palaechinidae
showed a preference for fine-grained sediments in the Tournaisian and Viséan which slightly
increased from the Tournaisian to the Viséan. This same trend of increasing affinity for fine-
128
grained environments was present in the Archaeocidaridae, Proterocidaridae and
Lepidocentridae. After this increase in affinity from the Tournaisian to Viséan, the
archaeocidarids continued to show strong affinity for fine-grained sediments until the
Kasimovian. The Proterocidaridae also showed a preference for fine-grained environments in the
Kasimovian, and the lepidesthids showed strong affinity for fine-grained sediments in the Viséan
and Serpukhovian. The Lepidocentridae showed a mixed affinity during the Tournaisian, which
increased in the Viséan to show a strong affinity for fine-grained sediments. When analyses were
run at the subperiod level, all analyzed taxa showed a preference for fine-grained environments
in both the Mississippian and Pennsylvanian (Fig. 3.2D). Likewise, when analyses were run at
the period level, all taxa showed a strong preference for fine-grained environments (3.2F).
Standardized Relative Affinity
Data for SRA are reported with respect to carbonates (Fig. 3.3A,C,E) and fine-grained
sediments (Fig. 3.3B,D,F). Since there are two categories for each analysis, carbonate or
siliciclastic and fine-grained or coarse-grained, the SRA for one environment will have the same
magnitude as its opposing environment, but with an opposite sign. When data were analyzed for
each stage (Fig. 3.3A), the Palaechinidae showed a fairly positive relative affinity for
siliciclastics in the Tournaisian. In the Viséan, however, their affinity switched drastically to
relatively preferring carbonate environments. Archaeocidarids showed an inverse trend, showing
a relative preference for carbonates in the Tournaisian that decreased in the Viséan then switched
to a relative affinity for siliciclastics in the Serpukhovian and Bashkirian. They then showed a
positive SRA for carbonates in the Moscovian, and their strongest relative affinity for
siliciclastics in the Kasimovian. The proterocidarids additionally showed a Tournaisian relative
129
affinity for carbonates and a Viséan relative affinity for siliciclastics. In the Kasimovian they
also appeared to have displayed a positive SRA for carbonates. Like the Proterocidaridae, in the
Viséan the lepidesthids showed a relative preference for siliciclastics, but in the Serpukhovian
they showed a relative affinity for carbonates. In the Moscovian they showed a strong relative
affinity for siliciclastics. In the Tournaisian, the Lepidocentridae showed a slight relative affinity
for carbonates.
When analyzed at the subperiod level (Fig. 3.3C), the Palaechinidae show the
strongest relative affinity for carbonates of any taxon in the Mississippian. The
Archaeocidaridae, however, show a relative affinity for siliciclastics in both the Mississippian
and Pennsylvanian. The Proterocidaridae show essentially mixed or no relative affinity in the
Mississippian, though in the Pennsylvanian show a relative affinity for carbonates. The
Lepidesthidae show a relative affinity for siliciclastics in both the Mississippian and
Pennsylvanian, but only slightly so. Lastly, in the Mississippian, the lepidocentrids show the
strongest relative affinity for siliciclastics of any clade. When analyzed for the entire
Carboniferous (Fig 3.3E), palaechinids show the strongest relative affinity for carbonates,
followed by the proterocidarids. The lepidesthids show a slight relative affinity for siliciclastics,
and the Lepidocentridae and archaeocidarids show stronger relative affinities for siliciclastics.
When we analyzed relative affinity for either fine- or coarse-grained sediments at the
stage level (Fig. 3.3B), the Palaechinidae showed a slight relative affinity for coarse-grained
sediments in the Tournaisian, followed by transition to a slight relative affinity for fine-grained
sediments in the Viséan. The Archaeocidaridae also showed a slight affinity for coarse-grained
sediments in the Tournaisian, though in the Viséan this affinity for coarse-grained environments
increased and maintained relatively constant until the Bashkirian, when the archaeocidarids
130
showed a relative affinity for fine-grained sediments. This transitioned back to a relative affinity
for coarse-grained environments in the Moscovian and a slight relative affinity for fine-grained
sediments in the Kasimovian. In the Tournaisian, the Proterocidaridae showed the highest
relative affinity for fine-grained sediments of any family, and this trend became even more
apparent in the Viséan, when their relative affinity for fine-grained environments increased. In
the Kasimovian, however, the Proterocidaridae showed a relative affinity for coarse-grained
environments. The Lepidesthidae showed a mixed or no relative affinity during the Viséan and
Serpukhovian, and the Lepidocentridae showed a slight relative affinity for coarse-grained
environments in the Tournaisian and Viséan.
When we analyzed relative affinity for grain size at the subperiod level (Fig. 3.3D), we
found that in the Mississippian, the Palaechinidae showed the highest relative affinity for coarse-
grained environments. The Archaeocidaridae similarly showed a relative affinity for coarse-
grained environments in the Mississippian, and the highest relative affinity for coarse-grained
environments in the Pennsylvanian. The Proterocidaridae showed a strong relative affinity for
fine-grained environments in the Mississippian, but in the Pennsylvanian showed a relative
affinity for coarse-grained environments. The Lepidesthidae showed a high relative affinity for
fine-grained environments in the Mississippian and the only relative affinity for fine-grained
environments of any family in the Pennsylvanian. The lepidocentrids also had a relative affinity
for coarse-grained environments in the Mississippian. When analyses were run at the period level
(Fig. 3.3F), the Palaechinidae showed the strongest relative affinity for coarse-grained
environments, followed by the Lepidocentridae and Archaeocidaridae. The Proterocidaridae,
however, showed a relative affinity for fine-grained environments and the Lepidesthidae showed
the strongest relative affinity for fine-grained substrates.
131
DISCUSSION
Substrate affinity of Paleozoic echinoids
Mineralogical Preferences
In general, Paleozoic echinoids tend to prefer carbonate substrates to siliciclastic
environments, and this preference seems to be fairly robust to variations in number of carbonate
and siliciclastic environments through the Carboniferous. This is evident in analyses at the stage
level, where the majority of families in the majority of analyzed stages prefer carbonates (Table
3.1, Fig. 3.3A), and even more at the more inclusive subperiod (Table 3.3, Fig. 3.3C) and period
(Table 3.5, Fig. 3.3E) levels. Only in one stage, did any family (the Archaeocidaridae) have a
statistically significant preference for siliciclastic environments, while all families except for the
Lepidocentridae showed statistically significant preference for carbonate environments at least
once. This resulted in a total of nine stage/family combinations in which a statistically significant
preference for carbonate environments was identified out of a total of fifteen stage/family
combinations that had more than five occurrences for that family. Additionally, when analyses
were run at the subperiod level, all taxa except for the lepidocentrids showed a statistically
significant preference for carbonates in the Mississippian, and in the Pennsylvanian
archaeocidarids and proterocidarids both showed preferences for carbonates (Table 3.3). When
analyses were run over the entire Carboniferous, all taxa except for the lepidocentrids showed a
preference for carbonates (Table 3.5). Furthermore, at the stage level only the archaeocidarids
and proterocidarids showed PPs indicative of preference for siliciclastic sediments (Fig. 3.2A),
and at the subperiod and period levels all taxa in all time bins had PP’s suggesting affinity for
carbonate substrates (Fig. 3.2C, E). A preference for carbonate environments has been seen in
132
some modern species of clypeasteroid and cassiduloid echinoids in the Caribbean (Mooi, 1990;
Telford and Mooi, 1986). Given that most families appear to have preferred carbonate
environments, it seems unlikely that resource partitioning, at least due to functional
morphological or ecological specialization at the family level, would have been responsible for
the affinity to carbonate environments. A possible reason for this affinity may be due to the
greater diversity of carbonate environments relative to siliciclastic environments. Amongst the
carbonate environments inhabited by echinoids in the Carboniferous are the oolitic inner shelf
Gilmore City Formation (Witzke and Bunker, 2005) , the reefal mudmounds of the Fort Payne
Formation (Thompson and Ausich, 2016) and the Clitheroe Limestone Formation (Donovan et
al., 2003; Miller and Grayson, 1972), and the deep basinal sediments of the Molignée Formation
(Mottequin, 2008; Mottequin et al., 2015). For epifaunal Paleozoic echinoids, siliciclastic
settings may lack the microhabitat differentiation present in carbonate environments such as
mudmounds thus allowing for diverse echinoid faunas.
In contrast to the binomial test and Bayesian approach, the SRA is a relative metric which
we used to compare affinity of a particular family to a random sample of all echinoids present
during a given time bin. Thus while the Bayesian and binomial test methods generally identified
strong affinities for carbonates relative to the distribution of carbonate substrates, the SRA
speaks to a family’s affinity for a given environment relative to all other echinoids in the
occurrence database over a given interval of time, and is useful for teasing out family-specific
differences in echinoid affinity. At the stage level, the SRA of carbonates showed that the
palaechinids were the only clade with a relative affinity for siliciclastics in the Tournaisian,
however in the Viséan, and when analyzed at the subperiod level and period level, they showed
the highest affinity for carbonates amongst any clade (Fig. 3.3). Their affinity for siliciclastics in
133
the Tournaisian is due to the fact that they were the only clade of echinoids to occur in
siliciclastic environments in the occurrence database in the Tournaisian, albeit only twice, so this
positive relative affinity for siliciclastics in the Tournaisian is not surprising, and is probably the
exception to the rule. There are currently taxa described in the literature assigned to the
Palaechinidae from the Pennsylvanian (Fischer Von Waldheim, 1848) and Permian (König,
1982), however, these taxa are likely proterocidarids (Thompson and Mirantsev, unpublished
data), and were analyzed as such in these analyses. The palaechinids thus likely went extinct
towards the end of the Viséan, as they are not definitely known from Serpukhovian-aged or
younger strata. The extinction of palaechinids may be linked to their tight affinity for carbonates.
It has been shown in crinoids, that from the Viséan to Serpukhovian there is an extinction of
primitive cladid and camerate groups and an increase in proportional diversity of the pinnulate
and siliciclastic tolerant advanced cladids (Kammer and Ausich, 2006). This reorganization of
crinoid faunas appears to have taken place over the course of the Viséan (Ausich and Kammer,
2013) and it is possible that the extinction of palaechinids may have been linked to the extinction
of many crinoid genera at this time.
The archaeocidarids show the highest relative affinity for carbonate environments in the
Tournaisian, but this decreased in the Viséan and in the Serpukhovian, Bashkirian and
Kasimovian they have a relative affinity for siliciclastics (Fig. 3.3A). The strong preference for
Archaeocidarids in the Kasimovian is due to the fact that they were the only family known to
inhabit siliciclastic environments during this time. In the Moscovian they show a relative affinity
for carbonates, which is evident in the very high abundance and diversity of archaeocidarids in
the Moscovian of the Moscow Basin, Russian Federation (Trautschold, 1879). When analyses
were run at the subperiod level, archaeocidarids show a relative affinity for siliciclastics in both
134
the Mississippian and Pennsylvanian (Fig. 3.3C), and thus appear to have been able to inhabit
siliciclastic substrates better than any other family in the Pennsylvanian. Furthermore, when
analyses were run over the entire Carboniferous, archaeocidarids showed the strongest relative
affinity for siliciclastic environments of any echinoid clade (Fig 3.3E). This ability to tolerate
siliciclastic settings may underlie the archaeocidarids longevity and abundance during the Late
Paleozoic. The Mississippian has long been recognized as a Phanerozoic, and particularly
Paleozoic peak in carbonate sedimentation (Peters, 2006; Ronov et al., 1980; Walker et al.,
2002), while the Pennsylvanian saw a relative increase in siliciclastic sedimentation that may
have begun in the Viséan (Smith and Read, 2000) and intensified in the Serpukhovian due to
onset and expansion of Gondwanan ice sheets during the Late Paleozoic Ice Age (Buggisch et
al., 2008; Miller and Eriksson, 1999; Stanley and Powell, 2003). That archaeocidarids show a
higher relative affinity for siliciclastics in the Mississippian may have allowed them to better
exploit new siliciclastic substrates in the Pennsylvanian, where they were the most abundant
clade of echinoids.
The proterocidarids show a relative affinity for carbonates in the Tournaisian and
Kasimovian, but for siliciclastics in the Viséan (Fig. 3.3A). When analyzed at the subperiod
level, they show an essentially mixed or no affinity in the Mississippian, but the largest relative
affinity for carbonates of any family during the Pennsylvanian (Fig. 3.3C ) and a strong affinity
for carbonates when analyses were run at the period level (Fig. 3.3E). Proterocidarids range from
the Devonian (Baird et al., 2009) to Triassic (Thuy et al., 2017), but despite their survival
through the Pennsylvanian and Permian, they remain rare relative to their Mississippian peak in
diversity and abundance. It is possible that their rarity may be due to their affinity for carbonate
135
environments in the Pennsylvanian, which as previously mentioned, are less common than in the
Mississippian.
The Lepidesthidae show relative affinity for siliciclastic settings in the Viséan and
Moscovian, and an affinity for carbonates in the Serpukhovian (Fig. 3.3A). At the subperiod
level though their relative affinity is slightly for clastics and they essentially have no or mixed
affinity when analyzed at the period level (Fig. 3.3C,E). They thus appear to have had a
relatively wide environmental tolerance. Furthermore when the lepidocentrids were analyzed at
the stage level, they show a slight relative affinity for carbonates, though at the subperiod and
period level, they show a relative affinity for siliciclastics (Fig. 3.3A,C,E). The lepidocentrids are
amongst the rarest of taxa in our analyses, and thus we are hesitant about making definitive
interpretations regarding their affinities, however, they do tend to have been relatively tolerant of
siliciclastics.
Preferences for Grain Size
When analyzed at the stage level no clades showed any statistically significant preference
for a particular grain size (Table 3.2), however, at the subperiod level all taxa analyzed in both
the Mississippian and Pennsylvanian showed a statistically significant preference for fine-
grained environments (Table 3.4). When analyzed at the scale of the entire Carboniferous,
however, there is again no statistically significant preference for substrates of any grain size
(Table 3.6). This is markedly different from the results obtained from the Bayesian method of
Simpson and Harnik (2009), which showed in almost every case, preference for fine-grained
environments. At the stage level, all analyzed taxa showed preference for fine-grained
environments in the Tournaisian, except for the Lepidocentridae, which showed equivocal
preference (Fig. 3.2B). After the Tournaisian, all analyzed taxa showed strong affinities (PP’s
136
above 0.9) for fine-grained sediments. This same preference for fine-grained environments was
also evident at the subperiod (Table 3.3D) and period (Table 3.3F) levels. Given the results of
the Bayesian approach, each family thus tends to occur more frequently in fine-grained
environments in such a way that is systematically different from the distribution of all echinoids
sampled at a given time. This preference for fine-grained environments may be in part
taphonomic, since many Carboniferous echinoderm occurrences are found in Lagerstätte, which
are often fine-grained (e.g (Seilacher et al., 1985)). However, numerous rock units which have
yielded echinoids included in our database, such as the Burlington Limestone, contain abundant
coarse-grained deposits, thus the results are not likely highly influenced by the presence of
lagerstätte. The SRA for fine-grained affinity shows that the palaechinids tend to have had a
higher relative affinity for coarse-grained environments than all other examined families when
examined at the subperiod and period levels while having a minor relative affinity for coarse-
grained environments in the Tournaisian and fine-grained environments in the Viséan (Fig.
3.3B,D,F). This tolerance for coarse-grained environments should not be surprising given the
relative diversity of palaechinids in the coarse-grained mobile substrates of the Burlington
Limestone (Jackson, 1912; Meek and Worthen, 1868) and in packstone facies at the Waulsortian
mud-mounds of Clitheroe (Donovan et al., 2003).
The archaeocidarids also showed relative affinities for coarse-grained environments
during numerous stages and at the subperiod and period levels (Fig. 3.3B,D,F). Many
archaeocidarids, particularly those of the genus Archaeocidaris, had large spines and primary
interambulacral tubercles (Fig. 3.1), which may have assisted in their mobility. This functional
morphological adaptation may have allowed archaeocidarids to inhabit mobile, coarser-grained
substrates. Archaeocidarids also showed affinity for fine-grained environments in the Bashkirian
137
and Kasimovian, and like all other echinoids analyzed using the Bayesian approach, did show an
affinity for fine-grained environments. As was the case with mineralogical preference, the
archaeocidarids appear to have been able to inhabit a wider range of substrate types than other
echinoids. This wide distribution was hypothesized by Thompson and Ausich (2016) based upon
the occurrence of archaeocidairds in the Viséan Fort Payne Formation. The archaeocidarids went
extinct sometime in the Late Permian, or during the end-Permian mass extinction (Thompson et
al., 2017b) and were not only the most abundant and diverse clade of echinoids during the
Pennsylvanian and Permian, but likely also the entire Paleozoic (Jackson, 1912). They may have
been able to reach this abundance in part due to their ability to inhabit a wide array of
environments and substrates.
In the Mississippian subperiod, Mississippian stages and at the period-level, the
proterocidarids display a relative affinity for fine-grained environments (Fig. 3.3B,D,F). In the
Pennsylvanian they show a relative affinity for coarse-grained environments (Fig. 3.3B,D).
Given their expanded oral ambulacra and large peripodial rings, proterocidarids have been
hypothesized to be well-adapted to fine-grained substrates (Smith, 1984). Furthermore,
proterocidarids are both diverse and abundant in the fine-grained, basinal sediments of the
Viséan Molignée Formation (Jackson, 1929a; Kier, 1962; Mottequin, 2008; Mottequin et al.,
2015). Our data support, through a non-functional morphological dataset, a relative affinity for
fine-grained sediments amongst Mississippian proterocidarids. In the Pennsylvanian,
proterocidarids show a relative affinity for coarse-grained sediments, indicating that later in their
evolutionary history, they were also likely capable of successfully inhabiting coarse-grained
environments.
138
The Lepidesthidae also appear to show a relative affinity for fine-grained environments
over coarse-grained environments. At the stage level, they tend to show no or equivocal
preference for fine-grained or coarse-grained environments in the Viséan and Serpukhovian (Fig.
3.3B). When analyzed at the subperiod and period levels, however, they show amongst the
highest affinities for fine-grained taxa of any of the examined families. Like the
Proterocidaridae, the Lepidesthidae also have greatly enlarged ambulacral areas (Fig. 3.1F) and
peripodial rings. Though the peripodial rings of the Lepidesthidae are dwarfed in size by those of
the Proterocidaridae, it is possible that these peripodia are associated with food-gathering tube
feet (though see Smith (1984) for an alternative viewpoint). That the Lepidesthidae show a
relative affinity for fine-grained sediments also supports the hypotheses of Chesnut and
Ettensohn (1988), who noted that the lepidesthids from the Serpukhovian Pennington Formation
were found predominantly in soft, marly sediments. They interpreted the lepidesthids as semi-
infaunal deposit feeders, and though there remains little evidence for an infaunal life mode for
lepidesthids, their large numbers of tube feet and relative affinity for fine-grained sediments may
support the deposit-feeding hypothesis. Of further note, in the Permian, the lepidesthid
Meekechinus elegans appears to have evolved a flattened test shape and enlarged peripodia
convergent upon those of the Proterocidaridae (Jackson, 1912; Kier, 1965) which may
additionally support the hypothesis that lepidesthids filled a similar niche to that of the
proterocidarids.
Finally, the Lepidocentridae show, at the stage, subperiod and period scales, a relative
affinity for coarse-grained environments (Fig. 3.3B,D,F). Though, again, sample size is small
with the lepidocentrids, this relative affinity for coarse-grained environments should not be
139
surprising given their abundance in coarse-grained, near shore settings like the Burlington
Limestone and the Gilmore City Formation (Kier, 1958a).
CONCLUSIONS
Much like many post-Paleozoic sea urchins, the stem group echinoids clearly showed
environmental preferences for both substrate mineralogy and for grain size. Most families of
echinoids show a broad preference for carbonate environments, which is robust to secular
variations in the abundance of carbonate and siliciclastic sediments in the rock record.
Furthermore, most of the families examined herein tend to prefer fine-grained environments to
coarse-grained environments. Relative to other echinoids, the palaechinids showed a relative
affinity for carbonate, coarse-grained environments, and this strong affinity for carbonate
environments may have been responsible for their extinction in the Mississippian concomitant
with the increased siliciclastic input linked to the Late Paleozoic Ice Age. The archaeocidarids,
on the contrary, showed a relative affinity for siliciclastic environments, and this tolerance for
siliciclastic settings may have facilitated their diversification and abundance in the
Pennsylvanian. The archaeocidarids also displayed a relative affinity for coarse-grained
environments, and were likely capable of inhabiting a wide variety of habitats. The
proterocidarids and lepidesthids showed a preference for fine-grained environments, and with
their expanded ambulacra may have thrived in these settings as deposit feeders. The
Lepidocentridae, despite their rarity, appear to have had a relative affinity for coarse-grained
environments. This study has focused only on the Carboniferous, due mostly to the relative
abundance of stem group echinoids during this time. Studies spanning the entire Paleozoic will
140
be necessary to further test the hypotheses put forward herein, and to better understand the
evolutionary paleoecology of the echinoid stem group.
141
Figure 3.1. Representatives of each of the five families of stem group echinoids analyzed herein.
A. The archaeocidarid echinoid Archaeocidaris marmorcataractensis from the Pennsylvanian
Marble Falls Formation of Texas (Thompson et al., 2015a). Specimen number TMM NPL64598.
B. The archaeocidarid Lepidocidaris squamosa from the Tournaisian Burlington Limestone of
Iowa (Jackson, 1912). Specimen number MCZ Cat 101858. C. The lepidocentrid Pholidechinus
brauni from the Viséan Edwardsville Formation of Crawfordsville, Indiana. Specimen number
MCZ Cat 101934. D. The palaechinid Oligoporus danae from the Viséan Keokuk Limestone of
Adams County, Illinois. Specimen Number MCZ Cat 101962 (Jackson, 1912). E . The
proterocidarid Proterocidaris belli from the Pennsylvanian Marble Falls Formation of Texas
(Kier, 1965). Note the expanded ambulacra on the oral surface. Specimen number USNM
144190. F. The lepidesthid echinoid Lepidesthes colletti from the Viséan of Montgomery
County, Indiana (Jackson, 1912). Specimen number MCZ Cat 101991 and scale bar is 5 mm.
Institutional abbreviations are as follows: TMM, Non-Vertebrate Paleontology Laboratory of the
Texas Natural Science Center, Austin, Texas, USA. MCZ, Museum of Comparative Zoology,
Harvard, Cambridge, Massachusetts, USA. USNM, Smithsonian Institution National Museum of
Natural History, Washington D. C., USA. All scale bars 10 mm unless otherwise stated.
142
C D
E
F
A B
143
Figure 3.2. Plots of posterior probabilities (PPs) showing substrate affinities for stem group
echinoids calculated using the method of Simpson and Harnik (2009) during the Carboniferous.
A. Preference for carbonate environments for each family when data was analyzed for each
stage. B. Preference for fine-grained environments for each family at the stage level. C.
Preference for carbonate environments for each family when data were analyzed at the subperiod
level. D. Preference for fine-grained environments for each family when data were analyzed at
the subperiod level. E. Preference for carbonate environments for each family when data were
analyzed at the period level. F. Preference for fine-grained environments for each family when
data were analyzed at the period level. For A., C. and E. PP of 1 indicates affinity for carbonates,
0 affinity for clastics, and 0.5 mixed or no affinity. For B., D. and E. PP of 1 indicates affinity for
fine-grained environments, 0 affinity for coarse-grained environments, and 0.5 indicates mixed
or no affinity. For each time bin, results are not presented for a family if there were less than five
occurrences of that family in the time bin. Dashed line plotted for PP of 0.5 indicating mixed or
no affinity. Symbols and colors shown for each family in the legend. Stage abbreviations are as
follows: Tour, Tournaisian; Vis, Viséan; Serp, Serpukhovian; Bash, Bashkirian; Mosc,
Moscovian; Kasi, Kasimovian; Gzh, Gzhelian.
144
300
310
320
330
340
350
360
370
Gzh
Kasi
Mosc
Bash
Serp
Vis
Tour
Pennsylvanian Mississippian
ages
Posterior Probability
0.0 0.4 0.8
Palaechinidae
Archaeocidaridae
Proterocidaridae
Lepidesthidae
Lepidocentridae
A
300
310
320
330
340
350
360
370
Pennsylvanian Mississippian
ages
Posterior Probability
0.0 0.4 0.8
Palaechinidae
Archaeocidaridae
Proterocidaridae
Lepidesthidae
Lepidocentridae
C
300
310
320
330
340
350
360
370
Gzh
Kasi
Mosc
Bash
Serp
Vis
Tour
Pennsylvanian Mississippian
ages
Posterior Probability
0.0 0.4 0.8
Palaechinidae
Archaeocidaridae
Proterocidaridae
Lepidesthidae
Lepidocentridae
B
300
310
320
330
340
350
360
370
Pennsylvanian Mississippian
ages
Posterior Probability
0.0 0.4 0.8
Palaechinidae
Archaeocidaridae
Proterocidaridae
Lepidesthidae
Lepidocentridae
D
300
310
320
330
340
350
360
370
Carboniferous
ages
Posterior Probability
0.0 0.4 0.8
Palaechinidae
Archaeocidaridae
Proterocidaridae
Lepidesthidae
Lepidocentridae
E
300
310
320
330
340
350
360
370
Carboniferous
ages
Posterior Probability
0.0 0.4 0.8
Palaechinidae
Archaeocidaridae
Proterocidaridae
Lepidesthidae
Lepidocentridae
F
Fig. 2
145
Figure 3.3. Plots of standardized relative affinities (SRA) showing substrate affinities for five
families of stem group echinoids calculated using the method Miller and Connolly (2001) during
the Carboniferous. A. Relative affinity for carbonate environments for each family when data
was analyzed for each stage. B. Relative affinity for fine-grained environments for each family at
the stage level. C. Relative affinity for carbonate environments for each family when data were
analyzed at the subperiod level. D. Relative affinity for fine-grained environments for each
family when data were analyzed at the subperiod level. E. Relative affinity for carbonate
environments for each family when data were analyzed at the period level. F. Relative affinity
for fine-grained environments for each family when data were analyzed at the period level. For A
C and E a positive SRA indicates affinity for carbonates, a negative SRA indicates affinity for
clastics, and SRA of 0 indicates mixed or no affinity. For B., D. and E. a positive SRA indicates
affinity for fine-grained environments, a negative SRA indicates affinity for coarse- grained
environments, and an SRA of 0 indicates mixed or no affinity. For each time bin, results are not
presented for a family if there were less than five occurrences of that family in the time bin.
Dashed line plotted for SRA of 0 indicating mixed or no affinity. Symbols and colors shown for
each family in the legend. Stage name abbreviations are as in Figure 2.
146
300
310
320
330
340
350
360
Kasi
Mosc
Bash
Serp
Vis
Tour
Pennsylvanian Mississippian
Standardized Relative Affinity
−3 −2 −1 0 1 2
Palaechinidae
Archaeocidaridae
Proterocidaridae
Lepidesthidae
Lepidocentridae
A
300
310
320
330
340
350
360
370
Pennsylvanian Mississippian
Standardized Relative Affinity
−3 −2 −1 0 1 2
Palaechinidae
Archaeocidaridae
Proterocidaridae
Lepidesthidae
Lepidocentridae
C
300
310
320
330
340
350
360
370
Gzh
Kasi
Mosc
Bash
Serp
Vis
Tour
Pennsylvanian Mississippian
Standardized Relative Affinity
−2 −1 0 1 2
Palaechinidae
Archaeocidaridae
Proterocidaridae
Lepidesthidae
Lepidocentridae
B
300
310
320
330
340
350
360
370
Pennsylvanian Mississippian
Standardized Relative Affinity
−2 −1 0 1 2
Palaechinidae
Archaeocidaridae
Proterocidaridae
Lepidesthidae
Lepidocentridae
D
F
300
310
320
330
340
350
360
370
Carboniferous
Standardized Relative Affinity
−3 −1 0 1 2 3
Palaechinidae
Archaeocidaridae
Proterocidaridae
Lepidesthidae
Lepidocentridae
E
300
310
320
330
340
350
360
370
Carboniferous
Standardized Relative Affinity
−3 −1 0 1 2 3
Palaechinidae
Archaeocidaridae
Proterocidaridae
Lepidesthidae
Lepidocentridae
Fig. 3
147
Table 3.1. p-values resulting from binomial tests comparing the distribution of occurrences of
five families of stem group echinoids in carbonate and siliciclastic sediments in each of the seven
stages of the Carboniferous to the proportion of carbonate and siliciclastic collections from the
PaleobioDB. Dashes indicate less than five unique occurrences for a family, and results are thus
not reported. Bold values indicate statistically significant preferences.
148
Carbonate Siliciclastic Carbonate Siliciclastic Carbonate Siliciclastic Carbonate Siliciclastic Carbonate Siliciclastic Carbonate Siliciclastic Carbonate Siliciclastic
Palaechinidae 1.47E-06 1.00 1.62E-09 1.00 - - - - - - - - - -
Archaeocidaridae 2.36E-04 1.00 1.68E-06 1.00 0.07 0.98 1.00 0.01 2.81E-04 1.00 3.69E-04 1.00 - -
Proterocidaridae 0.02 1.00 0.18 0.94 - - - - - - 3.71E-03 1.00 - -
Lepidesthidae - - 0.23 0.90 0.01 1.00 - - 0.65 0.70 - - - -
Lepidocentridae 0.13 1.00 - - - - - - - - - - - -
Gzhelian Tournaisian Visean Serpukhovian Bashkirian Moscovian Kasimovian
149
Table 3.2. p-values resulting from binomial tests comparing the distribution of stem group
echinoid occurrences for each of the five examined families in fine- or coarse-grained sediments
to the distribution of total echinoid occurrences in each stage of the Carboniferous. Dashes
indicate less than five unique occurrences for a family in a stage, and thus results are not
reported. None of the tests resulted in statistically significant preferences.
150
Coarse Fine Coarse Fine Coarse Fine Coarse Fine Coarse Fine Coarse Fine Coarse Fine
Palaechinidae 0.55 0.58 0.65 0.49 - - - - - - - - - -
Archaeocidaridae 0.53 0.63 0.26 0.82 0.44 0.80 1.00 0.40 0.29 0.87 0.66 0.58 - -
Proterocidaridae 0.81 0.43 1.00 0.11 - - - - - - 0.33 0.91 - -
Lepidesthidae - - 0.63 0.67 0.73 0.56 - - - - - - - -
Lepidocentridae 0.50 0.79 0.61 0.80 - - - - - - - - - -
Gzhelian Tournaisian Visean Serpukhovian Bashkirian Moscovian Kasimovian
151
Table 3.3. p-values from binomial tests comparing the distribution of stem group echinoid
occurrences for each family in carbonate and siliciclastic sediments in each subperiod of the
Carboniferous to the proportion of PaleobioDB collections of each sediment type in each
subperiod. Dashes indicate less than five unique occurrences in a time bin, and are thus not
reported. Bold values indicate statistically significant preferences.
152
Carbonate Siliciclastic Carbonate Siliciclastic
Palaechinidae 2.849E-18 1.000 - -
Archaeocidaridae 6.498E-10 1.000 0.002 0.999
Proterocidaridae 0.006 0.999 0.002 1.000
Lepidesthidae 0.008 0.998 0.110 0.973
Lepidocentridae 0.301 0.876 - -
Mississippian Pennsylvanian
153
Table 3.4. p-values resulting from binomial tests comparing the distribution of echinoid
occurrences for each family in fine- and coarse-grained substrates to the proportion of total
echinoid occurrences in each sediment type in each subperiod of the Carboniferous. Dashes
indicates less than five unique occurrences in a time bin, and are thus not reported. Bold values
indicate statistically significant preferences.
154
Mississippian Pennsylvanian
Fine Coarse Fine Coarse
Palaechinidae 1.326E-24 1.000 - -
Archaeocidaridae 1.074E-32 1.000 1.074E-12 1.000
Proterocidaridae 9.107E-09 1.000 0.001 1.000
Lepidesthidae 3.053E-10 1.000 1.822E-04 1.000
Lepidocentridae 0.001 1.000 - -
155
Table 3.5. p-values resulting from binomial test comparisons of echinoid occurrences for each
family of stem group echinoids in carbonate and siliciclastic to the distribution of PaleobioDB
collections of each of these sediment types in the Carboniferous. Bold values indicate
statistically significant comparisons.
156
Carbonate Siliciclastic
Palaechinidae 5.127E-21 1.000
Archaeocidaridae 2.565E-12 1.000
Proterocidaridae 1.030E-05 1.000
Lepidesthidae 0.001 1.000
Lepidocentridae 0.199 0.928
157
Table 3.6. p-values resulting from binomial test comparisons of echinoid occurrences for each
family of stem group echinoids in fine- and coarse-grained environments to the total proportion
of echinoids occurrences in fine- and coarse-grained sediments in the Carboniferous. None of the
comparisons showed statistically significant preferences.
158
Fine Coarse
Palaechinidae 0.811281 0.25153
Archaeocidaridae 0.655986 0.402045
Proterocidaridae 0.226244 0.880586
Lepidesthidae 0.072948 0.97337
Lepidocentridae 0.81573 0.386027
159
Chapter 4. Reorganization of sea urchin gene regulatory
networks at least 268 million years ago as revealed by oldest
fossil cidaroid echinoid
Originally Published as:
Thompson, J. R., Petsios, E., Davidson, E. H., Erkenbrack, E. M., Gao, F., and Bottjer, D. J.,
2015b, Reorganization of sea urchin gene regulatory networks at least 268 million years
ago as revealed by oldest fossil cidaroid echinoid: Scientific reports, v. 5.
INTRODUCTION
Living echinoids, members of the phylum echinodermata, belong to either the Cidaroidea
or Euechinoidea, and these two subclasses comprise the crown group echinoids (Kroh and Smith,
2010). The differential morphological diversity of these two subclasses is striking. Since their
emergence, euechinoids have diversified extensively from the bauplan of their earliest
representatives (Hopkins and Smith, 2015). For example, some euechinoid clades, such as the
Irregularia, which includes heart urchins, have secondarily gained anterior-posterior bilateral
symmetry (Hopkins and Smith, 2015; Kroh and Smith, 2010). In contrast, cidaroids have never
strayed far from the body plan of the earliest cidaroids. Neither the euechinoids nor cidaroids,
however, are known to be more basal than the other, and the Paleozoic Archaeocidaridae, from
which the euechinoids and cidaroids likely evolved, display synplesiomorphic characters of both
(Kroh and Smith, 2010; Lewis and Ensom, 1982).
The genetic and molecular developmental assembly of the echinoid bauplan is amongst
the best understood for any taxon (Davidson et al., 2002a; Davidson et al., 2002b; Oliveri et al.,
2008) and a large-scale reorganization of echinoid gene regulatory networks (GRNs) underlay
160
the initial divergence of cidaroids and euechinoids (Erkenbrack and Davidson, 2015).
Developmentally, cidaroids and euechinoids are also strikingly different. Cidaroid embryos
possess a variable number of micromeres, whereas those of euechinoids possess a characteristic
four (Schroeder, 1981; Wray and McClay, 1988). Embryonic cidaroids also lack primary
mesenchyme cells (Wray and McClay, 1988), from which the larval skeleton arises in
euechinoids (Amemiya and Emlet, 1992; Okazaki, 1975). Recent work has begun to explore the
genomic underpinning responsible for these morphological differences in early development
(Erkenbrack and Davidson, 2015). One of the key differences between the euechinoid and
cidaroid skeletogenic GRNs is the likely absence from the genome of the Pmar1 first repressor
in the double negative gate (Oliveri et al., 2008) of cidaroids (Erkenbrack and Davidson, 2015;
Yamazaki et al., 2014). The double negative gate is a regulatory circuit wiring design that is key
to the specification of skeletogenic mesenchyme in euechinoids and the use of which in
skeletogenesis is probably peculiar to this clade.
Echinoids are important and common constituents of modern ecosystems (Kier and
Grant, 1965; Linse et al., 2008; Nebelsick, 1996). Though they have a diverse and storied history
ranging back more than 400 myr to the Ordovician(Smith and Savill, 2001), echinoids do not
become abundant in the fossil record until 200 myr later in the Mesozoic (Hopkins and Smith,
2015; Kroh and Smith, 2010). Echinoids radiated in the Mesozoic after undergoing a bottleneck
at the Permo-Triassic mass extinction (252 Ma) where they experienced a severe reduction in
diversity (Erwin, 1994; Twitchett and Oji, 2005). The euechinoidea and cidaroidea clearly
diverged before this mass extinction at the end of the Permian (Smith and Hollingworth, 1990),
though the details of the timing of this divergence are not well constrained due to the rarity of
echinoids in Paleozoic strata. Apart from disarticulated spines, echinoids in the Paleozoic are
161
exceedingly rare. Most Paleozoic echinoids had poor preservation potential compared to post-
Paleozoic forms, with many clades displaying imbricate, overlapping, plating which presumably
lacked stereomic interlocking (Donovan, 1991; Smith, 1980a). Because of this non-rigid test
plating, Paleozoic echinoids presumably disarticulated rapidly following their death, and thus
well-preserved specimens in the Paleozoic are usually limited to Lagerstätte deposits (Schneider
et al., 2005). The stem-group cidaroid herein described from the Guadalupian of Texas, Eotiaris
guadalupensis n. sp., is the earliest putative crown group echinoid known in the fossil record,
and as such, provides new insight to the timing of the divergence of the euechinoids and
cidaroids, which must have preceded it, and the associated morphologic and developmental gene
regulatory changes that are the basis for this divergence.
STRATIGRAPHY AND GEOLOGIC SETTING
All new specimens of Eotiaris guadalupensis n. sp. are known from the Lamar Member
of the Bell Canyon Formation in the Guadalupe Mountains of west Texas (Thompson et al. 2015,
Supplementary Fig. S1). Specimens described by Kier (Kier, 1958b, 1965) are from the Word
and Road Canyon Formations of the Glass Mountains of west Texas (Thompson et al. 2015,
Supplementary Fig. S1). The Lamar Limestone is Lower Capitanian, about 264-263 Ma and the
Road Canyon Formation is, at its youngest, 268.8 Ma. Stratigraphy and geologic setting is herein
treated for only newly described material and detailed stratigraphic and locality information are
in supplementary information.
SYSTEMATIC PALEONTOLOGY
Class Echinoidea Leske, 1778
162
Subclass Cidaroidea Smith, 1984
Family Miocidaridae, Durham and Melville, 1957
Type genus. ⎯Miocidaris Döderlein, 1887
Other genera. ⎯Eotiaris Lambert, 1900, Couvelardicidaris Vadet, 1991, Procidaris Pomel,
1883
Genus Eotiaris Lambert, 1899
Type species. ⎯ Cidaris keyserlingi Geinitz, 1848, from the Wuchiapingian Zechstein of
Germany and England.
Diagnosis. ⎯Miocidarid with small test. Interambulacral plates imbricate adapically. Areoles
confluent only at and below ambitus. Spines with spinules, clavate to bulbous.
Occurrence. ⎯ Upper Permian of Germany, the U.K. and now Guadalupian of Texas, USA.
Remarks. ⎯The name Eotiaris is used instead of Miocidaris as the type material of the type
species of Miocidaris is indeterminate. The name Miocidaris was first used by Döderlein (1887)
who failed to explicitly name a type species for the genus. Bather (1909) then designated Cidaris
klipsteini Desor, 1855 as the type species, renaming it Miocidaris cassiani since it was
preoccupied by C. klipsteini Agassiz & Desor, 1847(Agassiz and Desor, 1847). M. cassiani,
itself, however, is a junior objective synonym of C. ampla Desor, 1858, a name proposed by
Desor in the Addendum to his synopsis when he realized that his C. klipsteini was preoccupied
(Kroh, 2015). Bather’s lectotype (Bather, 1909) consists of just fragmentary interambulacral
plates
(Smith and Kroh, 2011), which are indeterminate at the generic level, and are best left
restricted to the type material. Geinitz (1848) and King (1848) described the taxa Cidaris
keyserlingi Geinitz and Cidaris verneuiliana King from the Wuchiapingian of the UK and
163
Germany. King (1850) then placed Cidaris verneuiliana into Archaeocidaris, however this taxon
does not have the four interambulacral columns that characterize Archaeocidaris. Desor (1858)
furthermore placed Cidaris keyserlingi into Eocidaris however, this genus is strictly
indeterminate, being based solely off of disarticulated interambulacral plates. Lambert then
proposed the name Eotiaris keyserlingi for the material described by Geinitz. We follow Bather
(1909) and Smith and Hollingworth (1990) in synonymizing Cidaris keyserlingi Geinitz and
Cidaris verneuiliana King. Because the type of Miocidaris, however, is indeterminate, the genus
should only be restricted to the type species, Miocidaris ampla (Desor) from the Carnian St.
Cassian beds. Lambert’s name Eotiaris is thus the oldest available name for the material
described by Geinitz and King and is used herein.
Eotiaris guadalupensis Thompson n. sp.
1959 Spine Kier 1958a p. 889 Plate 114 Fig. 3.
1965 Miocidaris sp. Kier 1965 p. 456.
Type. ⎯Holotype is USNM 610600, paratypes are USNM 610601-610605.
Diagnosis. ⎯Eotiaris with straight, clavate and bulbous spines covered in numerous spinules
arranged helically around the shaft.
Derivation of name. ⎯ guadalupensis from the Guadalupe Mountains of west Texas, from
where the type material was collected.
Description. ⎯ Test regular and small, known only from disarticulated interambulacral columns.
Columns range in width from 4.2 mm to 9.3 mm (Fig. 4.1A-B, D-E, G). Modern cidaroids have
an interambulacral ambital width about 45% of their test diameter (Smith and Hollingworth,
164
1990), thus estimated E. guadalupensis test diameters about 9.4 mm to 20.6 mm. Apical system
unknown, and adapical interambulacral plates are not preserved articulated to the interambulacral
columns of the test. Adapical interambulacral plates likely imbricate whereas ambital and adoral
interambulacral plates rigidly sutured (Fig. 4.1E). Peristomial plates unknown, however
apophyses are present on most oral interambulacral plates (Fig. 4.1G,H). No buccal notches
present.
Lantern and teeth unknown. Ambulacra unknown, although likely beveling under
interambulacral plates as interior adradial interambulacral plate edges are denticulate.
Interambulacral plating arranged into two rows. First four to six plates usually rigidly
sutured with more adapical plates disarticulated (Fig. 4.1D-E). Plates pentagonal, about 1.3 to
1.6 times as wide as high. Primary tubercles large, sunken, and confluent below ambitus (Fig.
4.1A, E). Areoles at ambitus on specimen USNM 610601 about 2.6 mm wide and 2.6 mm high.
Boss crenulate with mamellons undercut and perforate. At ambitus, one row of secondary
tubercles on each plate separates tubercles. Above ambitus, multiple rows of secondary tubercles
separate ambitus on large specimens. On large specimens, about four rows of secondary
tubercles between the edge of each tubercle and the perradial suture at ambitus (Fig. 4.1E).
About three rows of secondary tubercles between primary tubercles and adradial suture at
ambitus. Adorally, this is reduced to two rows and eventually one row on the most adoral plates.
On smaller specimens, the number of secondary tubercles arranged laterally to the primary
tubercles are reduced to one. Interior of interambulacral plates slightly concave with seven or
eight denticles per plate at ambitus.
Spines ranging in morphology from straight (Fig. 4.1F) to clavate to bulbous (Fig 4.1C).
Proximal fourth to third of spine shaft smooth, ending in diagonally oriented ridge, which
165
contains the first row of spinules. Spinules oriented diagonally, along this raised ridge with more
distal rows parallel to first row. Spine morphology variable, with some maintaining constant
width and others tapering distally. Others ending in large clavate bulb covered in spinules. It is
likely that spines varied aborally to orally, as is present in some archaeocidarids(Schneider et al.,
2005) and recent cidaroids such as Eucidaris clavata
(Mortensen, 1928b). Although this
variability exists, all spines of differing morphologies contain diagonally oriented ridge bearing
first row of spinules. Acetabulum of spine bearing perforation and faint crenulations. A single
non-clavate spine is found associated with an interambulacral fragment which is 5.0 mm in
length (Fig. 4.1B). The interambulacral fragment is 7.6 mm wide indicating a probable test
diameter of 16.8 mm. This would indicate that the spines were likely less wide than the diameter
of the test. Spines have a prominent milled ring proximally. Bulbous spines hollow distally in
bulb and non-bulbous spines hollow distally. Secondary spines and pedicellariae unknown.
Remarks. ⎯ This taxon has been mentioned previously by Kier (Kier, 1958b, 1965) from the
Roadian and Wordian of west Texas, albeit as a single disarticulated interambulacral area and as
misidentified cidarid secondary spines respectively. The inclusion of more material, and the
association of the spines with the test of this species allow for a more thorough description
herein. All new specimens of this taxon are known from the Lamar Member of the Bell Canyon
Formation from the Guadalupe Mountains of west Texas, however, previously described
specimens, now assigned to this taxon, indicate its stratigraphic range expands into the Roadian.
The spines of this taxon are known from the Word Formation (Kier, 1958b) of the Glass
Mountains, however, they were originally incorrectly described as secondary spines of a larger
cidarid. These spines were collected from in between the Willis Ranch and Appel Ranch
members of the Word Formation, which are lower Wordian in age
(Wardlaw, 2000).
166
Furthermore, Kier (1965) attributed a specimen from the Road Canyon Formation of the Glass
Mountains to Miocidaris sp. This specimen (Figs. 4.1A, 2D) is herein assigned to Eotiaris
guadalupensis. This extends the stratigraphic range of this taxon into the Roadian, as the Road
Canyon Formation is Roadian to Kungurian in age (Lambert et al., 2010; Lambert et al., 2000;
Wardlaw, 2000). All of the material described herein has been silicified.
Morphologically, E. guadalupensis is very similar to Eotiaris keyserlingi from the
Zechstein of the UK and Germany, differing significantly only in the morphology of its spines.
Both bear rigidly sutured tests with plate imbrication adapically, sunken tubercles with multiple
rows of scrobicular tubercles and crenulate and perforate tubercles. The spines of E. keyserlingi,
which are well known, are smooth and have much smaller spinules than those of E.
guadalupensis(Smith and Hollingworth, 1990). They lack the clavate spine morphotype of E.
guadalupensis and are much shorter. E guadalupensis also differs significantly from E. connorsi
(Kier, 1965). The test of E. connorsi is composed entirely of imbricate, non-rigid plates, while
the tests of E. guadalupensis and E. keyserlingi are rigid except for adapically. The
interambulacral plates in E. connorsi are also much wider and do not display densely packed
scrobicular tubercles, as is the case in E. guadalupensis or E. keyserlingi.
Occurrence. ⎯ Specimens are known from the Lamar Member of the Bell Canyon Formation of
the Guadalupe Mountains, and the Road Canyon Formation and Word Formations of the Glass
Mountains of west Texas. They are thus Roadian-Capitanian in age.
Localities are USNM 725e, 728p, and 738b from Cooper and Grant (1972) see Thompson et al.
(2015) Supplementary information.
167
RESULTS
Phylogenetic analyses support the hypothesis that this taxon is a member of the
Cidaroidea (Fig. 4.2, Thompson et al. 2015, Supplementary Figs. S2, S3), and furthermore that it
is sister taxon to E. keyserlingi (Thompson et al. 2015, Supplementary Figs. S2, S3; See Methods
below). The euechinoid and cidaroid clades are confidently supported by bootstrap resampling
(Thompson et al. 2015, Supplementary Fig. S3) and Eotiaris guadalupensis is sister group to E.
keyserlingi with a bootstrapped confidence interval of 83%. Because Eotiaris guadalupensis had
apophyses and two columns of interambulacral plates, and plots as a cidaroid in the phylogenetic
analyses (Thompson et al. 2015, Supplementary Fig. S2, S3), then the strata from which it is
known must be younger than the divergence time of euechinoids and cidaroids. Furthermore this
provides a new basis upon which to obtain the hard minimum divergence date and thus is used to
date the gene regulatory changes associated with this divergence. Following the best practices
approach of Parham et al. (2011) a hard minimum divergence time was established for the
divergence of the euechinoids and cidaroids. The oldest known occurrence of Eotiaris
guadalupensis is the Road Canyon Formation of the Glass Mountains of west Texas. Based upon
the presence of the transitional form between the conodonts Jinogondolela idahoensis and J.
nankingensis and the presence of J. nankingensis, the Road Canyon Formation was determined
to be Kungurian to Roadian in age (Lambert et al., 2000; Lambert et al., 2002; Wardlaw, 2000).
Because the exact stratigraphic horizon within the Road Canyon Formation from which the
specimen of E. guadalupensis was collected is unknown, the top of the Roadian stage was
chosen as the hard minimum for the divergence of the cidaroids and euechinoids, following the
conservative practices for establishing hard minima set forth by Parham et al. (2011). The top of
the Roadian stage is set at 268.8 Ma based upon a smoothed cubic spline interpolation fit to the
168
existing radiometric age dates for the Carboniferous and Permian (Henderson et al., 2012), thus
making the hard minimum divergence time for the euechinoids and cidaroids 268.8 Ma (Fig.
4.2). The discovery of this new taxon extends the minimum divergence time of the euechinoids
and the cidaroids ten million years older than previously demonstrated (Smith and Hollingworth,
1990; Smith et al., 2006), shifting the minimum divergence time between these two taxonomic
groups from Wuchiapingian (Lopingian) to Roadian (Guadalupian) (Fig. 4.2) and establishing
that gene regulatory changes associated with this divergence must have also occurred by the
Roadian.
DISCUSSION
The Euechinoidea and Cidaroidea are differentiated, in part, because of the structure of
their Aristotle’s lanterns and perignathic girdles. The Aristotle’s Lantern operates as the “jaws”
of the echinoid, and contains numerous calcareous elements including the teeth. The perignathic
girdle comprises skeletal protrusions on the interior of the test that the retractor and protractor
muscles, which move the lantern in and out of the test, attach to. Based upon the lantern and
perignathic girdle structure of Eotiaris keyserlingi, Smith and Hollingworth (1990) determined
that the euechinoids and cidaroids must have diverged prior to the Wuchiapingian stage (259.8
Ma). The perignathic girdle structures in the euechinoids and cidaroids are developmentally
different, with the euechinoid auricles forming as protrusions from ambulacral plates and
cidaroid apophyses developing from interambulacral plates (Gao et al., 2015; Jackson, 1912;
Lovén, 1892). Although euechinoids and cidaroids have differing perignathic girdle structures,
neither structure is basal with respect to the other. This is known to be the case, because
archaeocidarids, from which both the cidaroids and euechinoids likely evolved (Lewis and
169
Ensom, 1982; Smith and Hollingworth, 1990), possessed the basal character state of having no
perignathic girdle. Eotiaris guadalupensis also has two columns of interambulacral plates, and,
through phylogenetic inference likely had two columns of ambulacral plates, as this character
had been fixed in Archaeocidaris and its predecessors for approximately 90 Myr, since the
Devonian
(Smith, 1984). These characters are synapomorphies of the crown group echinoids. As
demonstrated in Figure 4.2. and Thompson et al. (2015), Supplementary Figures S2 and S3, the
presence of apophyses, paired with two columns of interambulacral plates, indicates that Eotiaris
guadalupensis is definitively a cidaroid, and thus the cidaroid lineage and euechinoid lineage
must have already diverged prior to the appearance of this taxon in the rock record.
The presence of this taxon in Guadalupian rocks not only reinforces that the cidaroid-
euechinoid divergence happened prior to the Permo-Triassic mass extinction
(Smith and
Hollingworth, 1990), but indicates that it had occurred in the Roadian (268.8 Ma; Fig. 4.2) at
least 10 Myr earlier than previous estimates. Furthermore, the potential exists for new
discoveries to show that it may be even earlier, especially given that Eotiaris guadalupensis does
not plot as the most basal cidaroid in the phylogenetic analyses (Thompson et al. 2015,
Supplementary Fig. S2). In addition, this indicates that crown-group echinoids may have been
established by the Guadalupian and were certainly biogeographically widespread by the
Lopingian (Kier, 1965). The appearance of Eotiaris guadalupensis in the Roadian also extends
the inferred range of euechinoids prior to the Permian-Triassic boundary. The oldest definitive
euechinoids, Hemipedina hudsoni and Diademopsis heberti are not known until the Norian (Late
Triassic) (Kier, 1977b; Smith, 1994, 2007) thus making the implied fossil gap a minimum of 40
Myr. This new species also likely has profound impacts on the molecular clock divergence
dating for all echinoid clades. As the divergence of the cidaroids and euechinoids is the root
170
divergence node used for all divergence-dating analyses of echinoids (Nowak et al., 2013; Smith
et al., 2006), this new taxon has pushed back the basal node for divergence analyses 10 Myr.
Future work will attempt to incorporate this new basal divergence node into molecular clock
analyses.
Underlying this phylogenetic divergence must have been large-scale reorganization of the
developmental GRNs of cidaroids and euechinoids, with profound impacts on the differential
development of these clades. With regard to post-larval development, E. guadalupensis and other
basal stem-group cidaroids are morphologically very similar to even the most derived members
of the crown group Cidaroidea, due to the conserved nature of the cidaroid body plan.
Developmentally, this poses an interesting comparison with the Euechinoidea, which have a
much higher degree of post-larval morphological disparity relative to the cidaroids (Hopkins and
Smith, 2015; Kroh and Smith, 2010). New evidence has also shed light on the gene regulatory
development of juvenile skeletal structures, particularly with regard to the development of
apophyses and auricles. Both apophyses and auricles develop through the expression of specific
genes known to be required for skeletogenic specification in embryonic and post-embryonic
development: sm37, alx1, and vegfR
(Gao et al., 2015). In particular, sm37 is a well-understood
biomineralization gene (Lee et al., 1997; Livingston et al., 2006) the expression of which is
regulated by the upstream transcription factor alx1 (Ettensohn et al., 2003; Oliveri et al., 2008).
The differential spatial deployment of these genes during skeletogenesis is controlled by vegfR in
the embryo (Duloquin et al., 2007), and as such, this gene may be responsible for the differential
spatial expression of alx1 and sm37 during the formation of apophyses and auricles
(Gao et al.,
2015). Because of the presence of Eotiaris guadalupensis, which has definite apophyses, in the
171
Roadian, the fixation of the differential deployment of these biomineralization genes must have
at least begun by 268.8 Ma.
Additionally, there are a number of larval and embryonic developmental differences
between modern cidaroids and euechinoids that must have arisen with the divergence of these
two clades in the Permian. Euechinoid embryos possess four micromeres, and their larval
skeleton arises from primary mesenchymal cells, which ingress at the vegetal pole of the embryo
(Okazaki, 1975). Cidaroids, however, have a variable number of micromeres (Bennett et al.,
2012; Schroeder, 1981; Wray and McClay, 1988) and lack primary mesenchymal cells, instead
deriving their larval skeleton from skeletogenic cells emerging along with other mesodermal
cells from the tip of the archenteron (Emlet, 1988; Schroeder, 1981; Wray and McClay, 1988). In
euechinoids, the specification of skeletogenic mesenchyme is regulated by the double-negative
gate, whereby in the micromere lineage, Pmar1 represses HesC, which then allows for the
expression of downstream genes responsible for micromere specification such as Alx1, Ets1, and
tbr(Oliveri et al., 2008; Revilla-i-Domingo et al., 2007). The double negative gate appears to be
responsible for skeletogenic micromere specification across numerous phylogenetically diverged
euechinoid lineages, including the stomopneustoids, spatangoids, clypeasteroids and
camaradonts (Yamazaki and Minokawa, 2015) such that it is very likely present throughout all
indirect developing euechinoids. Contrary to euechinoids, it has been demonstrated that cidaroids
lack the HesC mediated double negative gate and that tbr plays no role in skeletogenesis
(Erkenbrack and Davidson, 2015). Many of the genes encoding transcription factors and
biomineralization genes responsible for micromere specification and embryonic skeletogenesis
in euechinoids are also involved in juvenile euechinoid skeletogenesis and were likely co-opted
by the skeletogenic micromere lineage (Gao and Davidson, 2008). As the euechinoids alone
172
possess a larval skeleton that is derived from primary mesenchymal cells, it is likely that this co-
option of juvenile skeletogenic genes occurred with the divergence of cidaroids and euechinoids.
It is unknown as to whether the euechinoid or cidaroid suites are ancestral, however, this new
fossil evidence indicates that the acquisition of one of these two differential character suites must
have occurred since the divergence of the euechinoids and cidaroids in the Roadian (268.8 Ma)
and is potentially very ancient.
CONCLUSIONS
Eotiaris guadalupensis, the geologically oldest cidaroid, is the oldest known probable
crown-group echinoid in the fossil record. This taxon pushes back the divergence of the crown-
group echinoids, the cidaroids and the euechinoids, to at least 268.8 Ma in the Roadian stage of
the Permian. It furthermore extends the inferred range of early euechinoids and establishes a new
hard minimum divergence for the basal node of all divergence dating studies regarding the
euechinoidea. In light of recent discoveries of differential cidaroid and euechinoid embryonic
and juvenile development, this taxon also provides strong evidence for fixation of disparate gene
expression systems by the Roadian. Eotiaris guadalupensis provides direct evidence for the
differential spatial expression of specific genes in euechinoid and cidaroid post-metamorphosis
skeletogenesis and indicates that this differential spatial expression must have been established
by at least 268.8 million years ago.
METHODS
Specimens of Eotiaris guadalupensis were analysed using dissecting microscopes and ESEM
microscopy was used to determine mineralogy of specimens. Measurements were taken with
173
calipers. Phylogenetic analyses were undertaken to rigorously demonstrate the phylogenetic
relationships of this species with respect to other Permian and Triassic echinoids. Permian and
Triassic euechinoids (three species; all from the family Pedinidae) and cidaroids (three species;
two from the family Miocidaridae and one from the Triadotiaridae) were included in the
analysis, in addition to E. guadalupensis. The outgroup of the analysis was Archaeocidaris
whatleyensis, a well-known, stem-group echinoid, which has been used as outgroup to all crown
group echinoids in previous analyses (Hopkins and Smith, 2015; Kroh and Smith, 2010; Smith,
2007). The characters used in the phylogenetic analysis in Thompson et al. (2015)
Supplementary Figures S2 and S3 consisted of 24 characters, 20 were binary and 4 were
multistate. Characters and character states are in supplementary information. All characters were
unordered and unweighted in original analyses and character matrix is listed in Thompson et al.
(2015) Supplementary Table S1. Corresponding Nexus file is in supplementary information.
Initial phylogenetic analysis was run in PAUP* version 4 (Swofford, 2003) and consisted of an
exhaustive search of all possible trees. This analysis resulted in 2 most parsimonious trees with
length 31 consistency index (CI) .806 and retention index (RI) .750. Characters were then
reweighted by their maximum retention indices and analyses were rerun.. This resulted in one
most parsimonious tree, equal to one of the two resultant trees from the unweighted search and
with length 22.5, CI .911 and RI .875 (Thompson et al. 2015, Supplementary Fig. S2). In order
to estimate branch support we ran a heuristic search with 1000 RASs and TBR with 1000
bootstrap replicates on the reweighted character matrix. Bootstrapped confidence intervals are
shown with appropriate branches in Thompson et al. (2015) Supplementary Figure S3.
174
Figure 4.1. Eotiaris guadalupensis n. sp. (A) Paratype USNM 610604. Interambulacral area
fragment first mentioned in Kier
25
from Roadian of the Glass mountains. Note two column
interambulacral area structure indicative of crown group echinoids. (B) Holotype USNM
610600. Interambulacral area fragment and associated spine. Note crenulate tubercles. (C)
USNM 610605a. displaying clavate, bulbous spine morphology. (D) Paratype USNM 610604.
Internal view of interambulacral fragment showing apophyses at adoral end. (E) Paratype USNM
610601. Interambulacral fragment of larger specimen. Note at least six plates in ambulacral
columns and crenulate tubercles with sunken areoles. Plates rigid at least below adapical plates.
(F) Paratype USNM 610605b. Spine displaying less clavate morphotype and spinules. (G)
Internal view of interambulacral area of paratype USNM 610602. Note apophyses, which
identify this species as a cidaroid, and denticulate adambulacral plate margin indicative of
beveling. (H) Close up of apophyses of USNM 610602. All scale bars represent 2.5 mm.
175
A B C
G
E
F
H
D
176
Figure 4.2. New divergence date of the divergence of cidaroid and euechinoid clades based on
the Roadian occurrence of Eotiaris guadalupensis n. sp. Thick lines represent fossil range and
thin lines represent inferred range based on phylogenetic relationships. The establishment of E.
guadalupensis as the oldest known cidaroid in the fossil record also extends the inferred range of
euechinoids, as the oldest known euechinoids, Diademopsis herberti, and Hemipedina hudsoni
are first found in the fossil record in the Norian, 40 Ma years later. Phylogenetic relationships are
from Kroh and Smith
1
and Kroh
60
modified with information regarding phylogenetic placement
of E. guadalupensis from Thompson et al. (2015) Supplementary Figure S2.
177
Archaeocidaris
Miocidaridae
Polycidaridae
Serpianotiaridae
Triadotiaridae
Echinothurioida
Other crown
euechinoids
Previous divergence estimate
(Smith and Hollingworth, 1990)
Cidaroidea
Diplocidaridae,
Rhabdocidaridae,
crown cidaroids
Euechinoidea
New divergence estimate
252.2
259.8
254.2
265.1
272.3
237.0
247.1
250.0
241.5
268.8
279.3
290.1
228.4
209.5
201.3
PERMIAN
Lopin-
gian
Guada-
lupian
TRIASSIC
Middle
Early
Late
Cis-
uralian
Artin-
skian
Kung-
urian
Roadian
Wordian
Capit-
anian
Wuchia-
pingian
Changh-
singian
Induan
Olenekian
Anisian
Ladinian
Carnian
Norian
Rhaetian
Fossil range
Inferred ranges
E. guadalupensis n.sp.
fossil range
Eotiaris guadalupensis n.sp.
Miocidaridae
Diademopsis heberti Smith, 1994
Pedinidae
178
Chapter 5. A diverse assemblage of Permian echinoids
(Echinodermata: Echinoidea) and implications for character
evolution in early crown group echinoids
Originally Published as:
Thompson, J. R., Petsios, E., and Bottjer, D. J., 2017b, A diverse assemblage of Permian
echinoids (Echinodermata, Echinoidea) and implications for character evolution in early
crown group echinoids: Journal of Paleontology, v. 91, no. 4, p. 767-780.
INTRODUCTION
Echinoids are members of the phylum Echinodermata and are important and common
constituents of modern ecosystems (Gagnon and Gilkinson, 1994; Kier and Grant, 1965; Linse et
al., 2008; Nebelsick, 1996). Though they encompass a wide morphological diversity in the Post-
Paleozoic (Hopkins and Smith, 2015), echinoids are first known from the Ordovician (Reich and
Smith, 2009; Smith and Savill, 2001) and following a maximum Paleozoic generic richness in
the Carboniferous (Kier, 1965; Smith, 1984) underwent a severe bottleneck at the Permo-Triassic
mass extinction (Erwin, 1993, 1994; Twitchett and Oji, 2005). Apart from disarticulated spines,
echinoids in Paleozoic strata are relatively rare. Prior to the Permian, most echinoids had flexible
tests, with many clades displaying imbricate plating presumably lacking the stereomic
interlocking present in post-Paleozoic clades (Smith, 1980a). Because of this non-rigid test
plating, Paleozoic echinoids disarticulated rapidly following their death, and thus articulated
179
echinoid material from the Paleozoic is often limited to Lagerstätte. Given their propensity to
disarticulate (Thompson and Ausich, 2016; Thompson and Denayer, 2017), echinoid diversity in
the Paleozoic is almost certainly underestimated, and thus all new taxa are important.
Furthermore, the late Paleozoic has been demonstrated to be the interval of time in which the
first crown group echinoids are known from the fossil record (Nowak et al., 2013; Smith et al.,
2006; Thompson et al., 2015b). As such, new faunas are important as they allow for a greater
understanding of the morphological innovations present in these earliest crown group, and latest
stem group echinoids, and the environments and communities in which the evolution of the first
crown group echinoids took place.
Examination of material in the collections of the National Museum of Natural History in
Washington D.C. revealed new specimens representing several taxa of at least three families, the
Miocidaridae, Archaeocidaridae, and Proterocidaridae, from the Guadalupian Bell Canyon
Formation of west Texas. The fauna herein described is the first well-preserved assemblage of
Permian echinoids comprising taxa of multiple families, and represents one of the most diverse
assemblages of echinoids known thus far from the Permian. The assemblage contains the first
known occurrence of the Proterocidaridae in the Permian of North America and increases the
number of stem cidaroid taxa known from the Permian to three. The discovery of this fauna also
indicates that echinoids were likely more geographically widespread in the Permian than
previously thought. Among these new specimens is the recently described Eotiaris
guadalupensis Thompson, 2017 representing the earliest known cidaroid and crown group
echinoid known in the fossil record (Thompson et al., 2015b). Furthermore, the echinoids
described herein demonstrate that some of the morphological innovations associated with the
echinoid crown group, were in fact present in numerous stem group taxa in the Permian.
180
GEOLOGIC SETTING
All specimens were collected from the Lamar Member of the Bell Canyon Formation from the
Guadalupe Mountains of west Texas. From 1939 to 1968, numerous expeditions were made by
G. A. Cooper and others to the Permian outcrops of west Texas. The specimens herein described
were collected during those excursions. In the Guadalupe Mountains, a number of microfossil
taxa have yielded good biostratigraphic control. Based on the concurrent presence of the
conodonts Jinogondolella postserrata and J. shannoni, the Lamar Member of the Bell Canyon
Formation has been determined to be Capitanian in age (Lambert, 2006; Lambert et al., 2002).
The transition from the J. postserrata Zone to the J. shannoni Zone takes place within the
uppermost Lamar Limestone Member (Lambert et al., 2010), thus indicating a lower Capitanian
age (~264 mya) (Henderson et al., 2012). The uppermost Lamar Limestone Member also marks
the transition from J. shannoni to J. altudaensis. The presence of both of these faunal transitions
allows for relatively precise biostratigraphic control and clarifies the age of the Lamar Limestone
Member to lower Capitanian (263-264 Ma).
The Lamar Member of the Bell Canyon Formation was deposited in the Delaware Basin
and is spatially located to the southeast of the Guadalupe Mountains and the Capitan Formation.
The Lamar Member contains carbonate debris flows transported from the reef edge sediments of
the Capitan Formation. The unit displays a wedge-shaped morphology, being over 90 m thick
near the shelf margin, where allochthonous sedimentation from the reefal sediments represented
by the Capitan Formation was greatest, and thinning basinward to only about 2 m (Babcock,
1977). The Capitan Formation and the Bell Canyon Formation are coeval (Lambert et al., 2010),
and merge towards the edge of the Delaware Basin. Close to the basin edge, at the type section of
181
the Reef Trail Member (which overlies the Lamar Limestone Member) the Lamar Limestone
Member was described as containing medium to dark grey organic-rich mudstones, and skeletal,
peloidal wackestones and packstones with interspersed carbonate debris flows containing
silicified fossils (Lambert et al., 2010). Babcock (1977) noted the presence of numerous
transported silicified reef fossils infilling channels in the zone proximal to the reef. Proximal to
the reef edge, the fauna of the Lamar Limestone Member consists of brachiopods, bryozoans and
crinoids (Babcock, 1977). Cooper and Grant (1972) furthermore noted that the brachiopod fauna
in the Lamar was similar to that occurring on the ‘reef slope’. Of special importance to this
paper, silicified echinoid spines and plates have been noted as common in these debris flows
(Babcock, 1977, p. 365, fig. 5). Basinward, the Lamar Limestone Member thins, and is
composed primarily of finely laminated mud lacking fossils and bioturbation (Babcock, 1977).
The specimens discussed in this study were collected from localities USNM 725e, 728p, and
738b near the Guadalupe Mountains (Cooper and Grant, 1972) which are interpreted as having
been deposited near a shelf margin.
MATERIALS AND METHODS
Following their collection, specimens were prepared out of bulk limestone blocks at the USNM
by using the hydrochloric acid dissolution method of Cooper and Grant (1972). Observations
were made using dissecting microscopes, and specimens were measured using calipers. Silicified
fossils are common in the Lamar Limestone (Babcock, 1977; Cooper and Grant, 1972) and all
Lamar echinoid specimens discussed in this study are silicified. Fine scale details of plate
structure and tuberculation are obscured by silicification and stereomic microstructure is lacking
from the surface of specimens. Cooper and Grant (1972) discussed two types of silicification
182
present amongst the fossils of the Glass Mountains. One of these preservation types results in a
thin coat of silica on the surface of the specimens, which, when treated with acid, protects the
calcite on the interior of the plate from disintegration. This is the non-pervasive silicification
discussed by (Butts and Briggs, 2011) and indeed, some echinoid specimens within this fauna are
preserved with only a thin layer of silica and thus contain calcitic interiors. The second type of
silicification mentioned by Cooper and Grant (1972) is complete replacement, where the entire
fossil has been recrystallized to silica. This is also common amongst the specimens described
herein.
Repositories and Institutional Abbreviations
Institutional abbreviations for specimen repositories are as follows, USNM=United States
National Museum, Washington D. C., USA; MGL=Musée d’Histoire Naturelle de Lille, France,
NMS G=National Museum of Scotland, Edinburgh, Scotland; RGM=Naturalis Biodiversity
Center, Leiden, The Netherlands.
SYSTEMATIC PALEONTOLOGY
Terminology and classification follows Smith (1984) and Kroh and Smith (2010). Methodology
follows Lewis and Donovan (2007).
Class Echinoidea Leske, 1778
Subclass Cidaroidea Smith, 1984
Family Miocidaridae Durham and Melville, 1957
Type genus. ⎯ Miocidaris Döderlein, 1887
Other genera. ⎯ Eotiaris Lambert, 1899, Couvelardicidaris Vadet, 1991, Procidaris Pomel,
1883
183
Genus Eotiaris Lambert, 1899
Type species. ⎯ Cidaris keyserlingi Geinitz, 1848, from the Wuchiapingian Zechstein of
Germany and England.
Diagnosis. ⎯ Miocidarid with small test. Interambulacral plates imbricate adapically. Aureoles
confluent only at and below ambitus. Spines with spinules.
Occurrence. ⎯ Wuchiapingian of Germany, the U.K. and Roadian and Guadalupian of Texas.
Eotiaris guadalupensis Thompson new species
urn:lsid:zoobank.org:act:6B7A2509-8B8D-4A48-9C18- 6BAF42BB9E51
Figure 5.1.1-1.5
1958 Spine Kier, p. 889, pl. 114, fig. 3.
v. 1965 Miocidaris sp.; Kier, p. 456.
v. 2015 Eotiaris guadalupensis; Thompson et al., p. 3, fig. 1. Unavailable name.
Holotype. ⎯The holotype is specimen USNM 610601 (Fig. 5.1.5)
Occurrence. ⎯ The specimens are known from the Lamar Member of the Bell Canyon
Formation of the Guadalupe Mountains. They are thus Capitanian in age.
Localities are USNM 725e, 728p, and 738b from Cooper and Grant (1972) with coordinates
from Wardlaw (2008). USNM 725e has latitude and longitude coordinates in decimal degrees of
31.9474, 104.7075. Type locality USNM 728p is located at 31.942, 104.701 and locality USNM
738b is at 31.981, 104.7497. Specimens of E. guadalupensis are also known from the Road
184
Canyon and Word formations of the Glass Mountains of west Texas. The road Canyon
Formation is, at the Youngest, Roadian, while the Word Formation is Wordian.
Materials. ⎯ USNM 610600 (Fig. 5.1.5), which is the holotype and USNM 610601 (Fig. 5.1.4)-
610605, which are paratypes.
Remarks. ⎯ The description of this species was published in the online only journal
Scientific Reports, and the name was not registered with ZooBank, making it unavailable.
The name is herein validated. Because this species was very recently described (Thompson et
al., 2015b) and as such, no further description is warranted here. We have figured, however, the
proximal spine shaft, milled ring and base of the spines of this taxon (Fig. 5.1.1-1.3), which
appears to be diagnostic due to the distinct diagonally oriented ridge and furthermore allows for
attribution of disarticulated spines to the coronas. In addition, two of the type specimens, the
holotype (USNM 610600; Fig. 5.1.5) and one of the paratypes (USNM 610601; Fig. 5.1.4) have
been figured for completeness of the fauna.
Stem Group Echinoidea
Family Archaeocidaridae M’Coy, 1844
Type genus.⎯Archaeocidaris M’Coy, 1844
Archaeocidaridae indet.
Figure 5.2.4-5.2.14
Occurrence. ⎯ Lamar Member of the Bell Canyon Formation of the Guadalupe Mountains, west
Texas. Localites 728p, 725e, 738b from Cooper and Grant (1972). See above description of
localities in Eotiaris guadalupensis for details.
185
Description. ⎯This taxon is only known from disarticulated interambulacral plates and spines.
Tubercles perforate and crenulate. Mamelons in the shape of an inverted cone. Crenulations
present in between parapet edge and mamelon and appear as extensions of the parapet projecting
radially inwards towards the mamelon (Fig. 5.2.6, 5.2.7). Some plates show diagenetic alteration
in the morphology of the mamelon, and as such, it is difficult to discern whether or not they are
crenulate (Fig. 5.2.4, 5.2.8). Radial plications present faintly (Fig. 5.2.4). Interior of plates
slightly concave (Fig. 5.2.5) with denticles present on adambulacral edges (Fig. 5.2.9).
Hexagonal plates are also present which lack denticles (Fig. 5.2.4, 5.2.5), indicating that the
interambulacral plates were arranged into more than two columns per area. There are two distinct
interambulacral plate morphotypes present in the assemblage and whether or not they represent
two distinct taxa, or plate variability within a species or individuals is unknown. We have chosen
to treat all disarticulated archaeocidarid ossicles together as one taxon until better material is
known, and there is justification for, or against, subdivision into different taxa. The first plate
morphotype consists of plates ranging from about 1.1 to 1.23 times as wide as high (Fig. 5.2.4,
5.2.5). Boss about 0.4 to 0.5 times as wide as plate and 0.5 to 0.6 times as high as plate. One ring
of scrobicular tubercles present adjacent to plate edge with about 19 scrobicular tubercles per
plate. Scrobicular tubercles on raised edge of plate such that area between tubercles is also raised
relative to the aureole. Tubercles sunken relative to scrobicular ring. Median (interradial)
interambulacral plates hexagonal, adambulacral (adradial) plates pentagonal with adambulacral
scrobicular ring slightly thicker than rest of plate. The second morphotype consists of plates that
are about equally high as wide to 1.2 times higher than wide (Fig. 5.2.6-5.2.9). Median
interambulacral plates hexagonal, adambulacral plates pentagonal. Plates have one row of
186
scrobicular tubercles on their lateral margins, but may have more than one on their adoral and
adapical edges. These additional rows of scrobicular tubercles contain smaller tubercles. About
20-22 scrobicular tubercles per plate. The primary tubercles are less sunken than the tubercles in
the first morphotype. Spines straight with alternating rows of spinules (Fig. 5.2.10, 5.2.13,
5.2.14). In cross section spines appear to be triangular to circular and appear to be hollow (Fig.
5.2.11, 5.2.12), though it is difficult to tell if this is a true morphological feature or taphonomic.
Materials. ⎯ Over 100 disarticualted plates and spines fragments were examined. Specimen
USNM 617187 is one lot of disarticulated spines of Archaeocidaridae indet. USNM 617188 is
one lot of interambulacral plates belonging to the first plate morphology while USNM 617189 is
one lot of disarticulated interambulacral plates belonging to the second plate morphotype.
Remarks. ⎯Although very few species of Archaeocidaris based off of articulated or semi-
articulated test material are known from the Permian, numerous plates and spines that have been
attributed to Archaeocidaris are known globally (Boos, 1929; Gortani, 1905; Hlebszevitsch and
Cortiñas, 2009; Kier, 1958b; Kittl, 1904; Leupke, 1976; Matthieu, 1949; Mihály, 1980; Prosser,
1895; Schneider, 2010; Simpson, 1976; Waagen, 1885; Wanner, 1941; Webster and Jell, 1992).
As cautioned by Kier (1958a, 1965) because different genera within the Archaeocidaridae have
differing numbers of interambulacral columns, it is best not to assign disarticulated
archaeocidarid plates to a particular genus. That being the case, these plates likely belong to
Archaeocidaris as the genus Polytaxicidaris Kier, 1958 is not known from outside of the
Mississippian, and the species of Polytaxicidaris for which external plate morphology is well
known, Polytaxicidaris lirata Kier, 1965 displays perforate secondary tubercles (Kier, 1965, figs.
187
4, 6), and are very unlike the plates described herein. However, because this determination is not
definitive, and because of its crenulate tubercles, the taxon is designated as Archaeocidaridae
indet. The shape and secondary tuberculation of the first plate morphotype, is similar to the
interambulacral plate morpholgy of numerous Archaeocidaris species, including Archaeocidaris
brownwoodensis Schneider, Sprinkle, Ryder, 2005, Archaeocidaris marmorcataractensis
Thompson et al., 2015 and Archaeocidaris wortheni Hall, 1858. The shape of the second plate
morphotype, with its extended flange adorally and aborally, is similar to the plates of
Archaeocidaris rossica (von Buch, 1842) from the Pennsylvanian of Russia. This plate
morphotype is also present in Archaeocidaris selwyni Etheridge, 1892 and archaeocidarids
described as “Cidaroid indet.” in Webster and Jell (1992) from the Permian of Australia,
Archaeocidaris manhattanensis Matthieu, 1949 (Fig. 5.2.2) from the Permian of Kansas and
Archaeocidaris aculeata Shumard in Shumard and Swallow, 1858 and from the Pennsylvanian
of Kansas. The plates of the indeterminate archaeocidarid from the Bell Canyon Formation,
however, are different from these aforementioned taxa in that they display crenulate tubercles
(Fig. 5.2.6, 5.2.7). The only definitive species of Archaeocidaris with crenulate tubercles are
Archaeocidaris apheles Schneider, Sprinkle, Ryder, 2005, however these are merely faint
indentations on the platform of the tubercle, and may not be homologous with true crenulate
tubercles. Archaeocidaris forbesiana (de Koninck, 1863), which was placed into Archaeocidaris
tentatively by Jackson (1912) was illustrated with crenulate tubercles by Waagen (1885).
Additionally, crenulate tubercles are known from echinoids from the Permian of Timor (Wanner,
1941; Fig. 2.1). That crenulate tubercles are clearly present in the indeterminate archaeocidarid
described herein indicates that the crenulate tubercles illustrated by Waagen (1885) were likely
truly present on the figured specimens and that crenulate tubercles were likely widespread
188
amongst numerous taxa in the Permian. Crenulate tubercles are also present in miocidarids such
as Eotiaris keyserlingi (Geinitz, 1848) and Eotiaris guadalupensis and the importance of
crenulate tubercles on Archaeocidaris? sp. will be discussed further below.
It is also necessary herein to address the genus Permocidaris Lambert, 1900, which has
been regarded within the Miocidaridae by Smith and Kroh (2011) and has also been described as
having crenulate tubercles. The type species of Permocidaris is Cidaris forbesiana de Koninck,
1863 and the type specimen is NMS G.1871.1.34. The type of Cidaris forbesiana consists of
several disarticulated spines that are clavate in morphology, bearing columns of spinules
arranged laterally from the proximal end, slightly above the milled ring, to the distal end (Fig.
5.2.3). Waagen (1885) attributed several disarticulated interambulacral plates to this species,
while transferring it to the genus Eocidaris Desor, 1856. He and Bather (1909) pointed out that
these plates were not definitively associated with the spines described by de Koninck (1863).
Bather (1909) rightly pointed out that, dependent upon the morphology of the interambulacral
plates, there seems little to distinguish Permocidaris from Archaeocidaris save for its crenulate
tubercles. However, because the material attributed by Waagen (1885) to Eocidaris forbesiana
and the spines similar to the type of Eocidaris forbesiana have not been found in direct
association, it is best to treat this material as indeterminate, and as such, we treat the genus
Permocidaris Lambert, 1900 as incertae sedis. Furthermore, as this genus is known solely from
disarticulated spines and plates, its familial level affiliation is uncertain. Smith and Kroh (2011)
choose to place it within the Miocidaridae, one of the diagnosable characters of which is the
presence of two columns of interambulacral plates. As the number of interambulacral columns of
this taxon is unknown, however, this seems unadvisable. Furthermore if the interambulacral
plates assigned by Waagen (1885) to Eocidaris forbesiana are similar in nature to the
189
interambulacral plates herein assigned to Archaeocidaridae indet., then they may well in fact
have been arranged into more than two columns of interambulacral plates, as is the case with the
latter. Permocidaris? timorensis Wanner, 1941 from the Permian of Timor also appears to have
interambulacral plates bearing crenulate tubercles arranged into more than two columns, as
plates both with and without a denticulate margin bear crenulations (Wanner, 1941, pl. 25, figs.
11-19). Because both of these taxa have more than two columns of interambulacral plates and
crenulate primary tubercles, they may be closely aligned. Because of the incomplete nature of
specimens of both taxa, however, we find it best to treat both as indeterminate.
Family Proterocidaridae Smith, 1984
Type genus.⎯Proterocidaris de Koninck, 1882
Genus Pronechinus Kier, 1965
Type species.⎯Pronechinus anatoliensis Kier, 1965 from the Changhsingian of southeastern
Turkey.
Other species.⎯Pronechinus cretensis König, 1982 from the Asselian of Crete.
Pronechinus? sp.
Figure 5.2.15-5.2.21
Occurrence. ⎯ Lamar Member of the Bell Canyon Formation of the Guadalupe Mountains, west
Texas. Capitanian in age. Locality 728p from Cooper and Grant (1972). See above description of
localities in Eotiaris guadalupensis for details.
Description. ⎯ This taxon is known only from disarticulated ambulacral and interambulacral
plates. Ambulacral plates variably polygonal in shape. Some plates bear well-defined
190
imperforate tubercles (Fig. 5.2.15) while others do not (Fig. 5.2.19). Additionally, some pore
pairs are surrounded by distinct peripodia (Fig. 5.2.15, 5.2.16). Peripodia about 0.5 to 0.8 times
as wide as plates. Ambulacral plates both with and without primary tubercle can bear smaller
secondary tubercles. Pore pairs perforating ambulacral plates at about half-way through their
thickest point. Ambulacral plates with adoral margin heavily beveled, and ambulacral plating
almost certainly imbricate. Interambulacral plates also imbricate, larger than ambulacral plates.
All bearing minute tubercles but with some additionally bearing distinct imperforate primary
tubercles.
Materials. ⎯USNM 617192 and USNM 617193 (Fig. 5.2.15) are ambulacral plates with well-
defined peripodial rims. USNM 617194 (Fig. 5.2.16) is an ambulacral plate with a well-defined
peripodial rim and imperforate tubercle. USNM 617195 (Fig. 5.2.18) is an ambulacral plate
without a well-defined peripodial rim and an imperforate tubercle. USNM 617196 (Fig. 5.2.19)
lacks both a well-defined peripodial rim and a primary tubercle. USNM 617198-617200 are lots
of ambulacral plates and USNM 617197 and 617201 (Fig. 5.2.20, 5.2.21) are lots of
interambulacral plates.
Remarks. ⎯This taxon is questionably assigned to Pronechinus because the details of the
arrangement of its test plating are incompletely known. The details of the interambulacral and
ambulacral plates, however, confidently allow placement within the Proterocidaridae.
Pronechinus is the only proterocidarid that has numerous ambulacral plates with small tubercles
lacking peripodia, yet which contain large tubercles. These plates are located within the more
perradial ambulacral columns of Pronechinus anatoliensis and this tuberculate morphology is
191
present among the ambulacral plates described herein (Fig. 5.2.18). Proterocidaris belli (Kier,
1965) from the Pennsylvanian Marble Falls Formation of Texas has ambulacral plates with
adoral tubercles, however, all of these plates bear peripodia (Kier, 1965). Additionally, it is
likely that the interambulacral plates with primary tubercles described herein are adambulacral in
origin, as the adambulacral plates of Pronechinus anatoliensis bear distinct primary tubercles
(Kier, 1965). Pronechinus is known from two species, Pronechinus cretensis and Pronechinus
anatoliensis from the Asselian and Changhsingian respectively (Fig. 5.4). This is the first
putative occurrence of this genus in the Capitanian, and the first occurrence of a proterocidarid in
the Permian of North America, indicating that the Proterocidaridae were likely biogeographically
widespread in the Permian.
Echinoidea indet.
Figure 5.3.1-5.3.3
Occurrence. ⎯Same as for Eotiaris guadalupensis.
Description. ⎯Numerous disarticulated fragments of Aristotle’s lanterns with associated teeth
are present among the described fauna. Most of the fragments are disarticulated hemipyramids,
though some articulated hemipyramids with teeth are present (Fig. 5.3.1-5.3.3). Maximum height
of hemipyramid 16.44 mm. Foramen magnum is 1.58 mm deep on this specimen and generally
sloping at about 50° to the horizontal. Depth and angle of foramen magnum is variable, however,
even in smaller specimens the foramen magnum never appears to exceed 0.2 times the height of
the hemipyramids. Protractor muscle scars about halfway down length of hemipyramid
terminating in deep perforation. Retractor muscle attachment scars below. The wing edge is a
faint ridge along the side of the hemipyramids, though may be faint due to preservational biases.
192
Wings about 0.3 times as long as hemipyramids are high. Dental slide present. Teeth U-shaped,
non-serrate.
Materials. ⎯Lanterns are in lots USNM 617190 (Fig. 5.3) and 617191.
Remarks. ⎯It is unknown whether or not the pyramids described here can be attributed to any of
the taxa described herein. The hemipyramids are much taller than the test of Eotiaris
guadalupensis, thus it is unlikely that the pyramids belong to this taxon. Smith and Hollingworth
(1990) described the lantern of Eotiaris keyserlingi, which is much smaller than the pyramids
discussed here. The morphology of lanterns of echinoids in the upper Paleozoic is rarely
described in great detail (though see (Lewis and Ensom, 1982) for a counterexample). The
indeterminate pyramids described here appear to be similar in morphology to those described as
Population A in Hoare and Sturgeon (1976). They attribute these lanterns to either
Archaeocidaris or Polytaxicidaris. The foramen magnum slopes at a gentler angle in the lanterns
attributed to Archaeocidaris? jacksoni Spreng and Howe, 1963 and Archaeocidaris? remotus
Spreng and Howe, 1963 than in the pyramids herein described. Additionally, the indeterminate
lanterns appear to have less sloping foramen magna then those of the lanterns described by Kier
(1958b) from the Wordian of west Texas. Proterocidarid lanterns are also not well enough
known to confidently assign the herein described taxon to Pronechinus? sp. The lantern
characteristics of particular clades of echinoids in the Paleozoic are not well known, as pointed
out by Spreng and Howe (1963). This is in part because lanterns are often assigned to genera and
species without the presence of any articulated test material, and are often given the names of
taxa that are associated with disarticulated plates and spines of preexisting genera. This results in
193
inconsistency with regard to the association of lanterns with particular genera and species, and
thus the lantern characteristics attributable to higher-level taxa are not well known. Because of
this, the characteristics of the lanterns described herein preclude assignment to a particular
taxonomic group, and thus they are left in open nomenclature.
The teeth associated with these lanterns are of interest primarily because they do not have
a serrated point, and thus are not like the teeth typically associated with archaeocidarids and
proterocidarids. Archaeocidaris, Polytaxicidaris and the proterocidarids Proterocidaris de
Koninck, 1882 and Fournierechinus Jackson, 1929 display serrated teeth constructed of multiple
rows of columns of primary and secondary plates. These are termed compound lamellar teeth
(Reich and Smith, 2009). That the teeth present herein do not display the serrated tip
characteristic of compound lamellar teeth suggests that they are not compound lamellar in origin.
The nature of echinoid tooth microstructure is best understood through scanning electron
microscopy analyses, however, due to the silicified nature of the material described herein, we
did not attempt to examine the microstructure of the teeth. That they do not appear to be
compound lamellar, however, leaves a few hypotheses regarding the nature of the teeth. They
could be from a taxon not present in the described fauna, however, this seems unlikely given the
nature of the material and its preparation, which exposed the entire silicified fauna of the
limestone blocks. Furthermore, these pyramids could belong to Archaeocidaridae indet. or
Pronechinus? sp. The non-serrate morphology may indicate either a primitive (e.g. simple
lamellar) or derived (e.g. cidaroid-type U-shaped) tooth. U-shaped teeth are known from the
Permian of the Ford Formation associated with the cidaroid Eotiaris keyserlingi and had clearly
evolved by the Permian (Smith and Hollingworth, 1990). The teeth present in this west Texas
194
fauna do have the U-shaped cross section characteristic of cidaroid teeth, however, without
microstructural analysis their affinities remain unclear.
DISCUSSION
The assemblage herein described is particularly diverse by Permian standards. That this
assemblage is so diverse relative to other Permian echinoid occurrences is likely due to the
interplay of a number of competing factors, including worker bias, paleoenvironmental setting,
and taphonomy.
This is the first study to methodologically identify disarticulated echinoid ossicles to the
taxonomic rank of family in the Permian, though the approach has been used with much success
in older strata (Kutscher and Reich, 2004; Reich and Smith, 2009; Thompson and Denayer,
2017). It has been demonstrated that the examination and identification of disarticulated
echinoid ossicles increases the number of taxa known from a given formation (Donovan, 2001;
Gordon and Donovan, 1992; Kroh, 2007; Thompson and Denayer, 2017). Additionally, it is
expected that when disarticulated echinoid ossicles are used to evaluate diversity estimates,
recorded species richness will be higher than when solely articulated specimens are used
(Nebelsick, 1996). Therefore, it is possible that the high diversity assemblage herein observed is
higher than that of other localities in the Permian due to the failure of previous studies to account
for disarticulated ossicles.
The silicified nature of fossils from the Permian of the Guadalupe Mountains also
undoubtedly plays a role in the higher reported diversity. Because the examined specimens were
silicified and later dissolved out of limestone blocks, a greater number of specimens were
available for study. Silicified faunas also often yield higher diversity aragonitic mollusk faunas
195
than those which are not silicified (Cherns and Wright, 2000, 2009; Wright et al., 2003),
however it is unknown if the same would be true of the High-mg calcite echinoderm faunas.
Regardless, if the specimens reported herein were not silicified, and therefore could not have
been easily dissolved from their interring matrix, then only specimens on bedding planes would
have been visible. To our knowledge, no other silicified Permian echinoid faunas from outside of
west Texas have been acquired through dissolution of CaCO
3
matrix. This technique has,
however, yielded exceptional preservation and increased estimates of diversity and abundance in
Pennsylvanian (Kier, 1965) and Triassic (Smith, 1994; Stanley, 1989, 1994) faunas. It is
probable that if other silicified Permian faunas are prepared through matrix dissolution
techniques, they will yield comparably diverse faunas.
It is also possible that the assemblage herein described has a relatively high diversity
because of the depositional environment from which it was collected. This assemblage herein
reported preserves a transported reefal community (Babcock, 1977). Reefal communities have
been shown to contain diverse regular echinoid assemblages in the Red Sea, and the sediments
associated with these reefal environments also preserve disarticulated components of diverse
regular echinoid assemblages (Nebelsick, 1996). A transported reefal community may also have
preserved a mixed assemblage representing the wide range of microhabitats in which echinoids
inhabit on a reef. Though reefal environments are unlikely to preserve a full suite of articulated
echinoid communities due to their high energy-setting (Nebelsick, 1996), as mentioned above,
when disarticulated material is accounted for, a more diverse assemblage is likely to be
preserved. Also of importance is the fact that this is not the first Permian reef to yield fossil
echinoids. The Wuchiapingian Zechstein reefs of the Ford Formation from Northern England
(Hollingworth and Pettigrew, 1988; Smith and Hollingworth, 1990) and the Zechstein Reefs of
196
Thuringia, Germany (Reich, 2007) are both known to contain the stem group cidaroid Eotiaris
keyserlingi. It is thus possible that reefal environments in the Late Paleozoic may have supported
diverse echinoid communities and that the assemblage described here is so diverse simply
because of the reefal nature of the sediments in which it was preserved. That two species of
Eotiaris, E. guadalupensis and E. keyserlingi are both reported exclusively from reefal
environments may also indicate that early crown group echinoids evolved in, or at least thrived
in, these environments. These Permian reefal environments are also not the oldest such Paleozoic
reefal settings to yield echinoid faunas. Although their framework differs from reefs of the
Permian, Mississippian mud mounds from the Fort Payne Formation of Kentucky (Thompson
and Ausich, 2016) and the Waulsortian mud mounds of Clitheroe, Lancashire (Donovan et al.,
2003; Hawkins, 1935) and Waulsort, Belgium (Jackson, 1929a) have also yielded diverse and
abundant echinoid faunas, and were likely favorable habitats for echinoids. It has also been
recently proposed that Triassic echinoids may have had an affinity for reefal environments
(Zonneveld et al., 2016), as much of the known Triassic echinoid fossil record is from reefal
settings (Smith, 1994; Stanley, 1979; Stanley, 1989, 1994; Zonneveld, 2001; Zonneveld et al.,
2007). Many of these Triassic taxa are stem group cidaroids belonging to the Miocidaridae
(Zonneveld et al., 2007) and the families Triadocidaridae and Paurocidaridae (Smith, 1994),
which are likely to be descendants of miocidarids such as Eotiaris guadalupensis (Smith, 2007).
Given the abundance of these stem cidaroids in reefal environments in the Permian and Triassic,
it is possible that the early diversification of stem group cidaroids may have taken place in reefal
environments, however, more data will be necessary to further test this hypothesis.
197
Co-occurrence of Stem Group and Crown Group Echinoids
From a paleoecological standpoint, this fauna is important because it demonstrates a depositional
environment in which archaeocidarids, proterocicarids and miocidarids coexisted. Permian
miocidarids have until now only been reported from assemblages where they are the only
echinoid constituent (Kier, 1965; Reich, 2007) and that miocidarids are herein found from the
same environments as archaeocidarids indicates that the most crown-ward stem group echinoids,
the archaeocidarids, and the earliest crown group echinoids, the miocidarids, were occupying the
same environments at the same time. This is particularly interesting given the survival of the
miocidarids through the Permian-Triassic mass extinction (Erwin, 1993, 1994; Smith and
Hollingworth, 1990), which appears to have been responsible for the extinction of the
archaeocidarids. Miocidarids outside of localities in west Texas are only known from the two
reefal localities described above. Archaeocidarids, however, were much more abundant and
apparently more geographically widespread as they have been described from test or
interambulacral plate material from localities in Texas (Kier, 1958b; Schneider, 2010), Australia
(Etheridge, 1892; Webster and Jell, 1992), Kansas (Boos, 1929; Matthieu, 1949), Oklahoma
(Boos, 1929), Pakistan (Waagen, 1885), Tunisia (Matthieu, 1949), Timor, (Wanner, 1941),
Argentina (Hlebszevitsch and Cortiñas, 2009), Hungary (Mihály, 1980) and Bosnia (Kittl, 1904).
This abundance of archaeocidarids relative to miocidarids makes their demise at the Permian-
Triassic all the more interesting, and currently there exists no good mechanism to explain the
differential survival of the miocidarids and archaeocidarids. Furthermore, there is no good
understanding of the temporal distribution of Permian archaeocidarid abundance or diversity at
the stage level or lower, which will be necessary to understand the dynamics of stem group
echinoid richness and abundance leading up to the Permian-Triassic boundary. For instance, the
198
end-Guadelupian extinction event (Stanley and Yang, 1994), which was responsible for major
extinctions in some clades (Groves and Wang, 2013; Stanley and Yang, 1994) and only slightly
elevated extinction rates in others (Clapham, 2015; Payne and Clapham, 2012) may have played
a role in extinction of the archaeocidarids, however, whether or not this is the case remains to be
seen.
The Acquisition of Characters Leading to Crown Group Echinoids
New fossil discoveries are key for establishing the sequence of character evolution associated
with the transition from stem group to crown group taxa. As specimens with new morphologies
are discovered, a clearer picture of the order of character changes leading from the stem group to
the crown group becomes available, and the true synapomorphies defining the crown group
become apparent (Donoghue, 2005). Basal crown group echinoids have previously been united
by a number of synapomorphies, which distinguish them from members of the echinoid stem
group. Among the most conspicuous of these synapomorphies include the reduction in coronal
plating to two columns of interambulacral plates and two columns of ambulacral plates, and the
evolution of the perignathic girdle (for a complete list of crown group echinoid synapomorphies
see (Kroh and Smith, 2010)). Additionally, though not demonstrably present in the most basal
euechinoids (for which there is little fossil evidence), the earliest, and most basal cidaroid taxa
also display crenulate tubercles and a rigid interambulacral area at and below the ambitus (Kier,
1965; Smith and Hollingworth, 1990; Thompson et al., 2015b). Previous to this study, neither a
perignathic girdle, an interambulacral area composed of two columns of plates, nor crenulate
tubercles, were known to be present in the most derived stem group echinoids, which belonged
to the genus Archaeocidaris. It is very unlikely, however, that the acquisition of these three
199
characters took place all at once, as evolutionary transitions marked by large numbers of
character acquisitions are often incremental (Donoghue and Purnell, 2005; Makovicky and
Zanno, 2011). Although it is well known that the origination of crown group echinoids took
place in the Late Paleozoic (Nowak et al., 2013; Smith and Hollingworth, 1990; Smith et al.,
2006; Thompson et al., 2015b), the order of character state transitions associated with, and
leading up to the origination of the crown group is not well understood.
The new specimens of Archaeocidaridae indet. from west Texas appear to shed light on
the order of some of the character changes associated with the morphological transition from
stem group to crown group echinoids (Fig. 5.4). The interambulacral plates of this indeterminate
archaeocidarid are comprised of pentagonal and hexagonal forms (Fig. 5.2.4-5.2.9). On one
interior edge, the pentagonal plates bear a distinct denticulate margin, where the plates of the
interambulacral plate imbricated over the adjacent ambulacral plates (Fig. 5.2.9). Furthermore,
the hexagonal interambulacral plates do not bear this denticulate margin (Fig. 5.2.5), therefore
assuredly not beveling over ambulacral plates and therefore belonging to median interambulacral
columns. This indicates that the test of this indeterminate archaeocidarid was composed of
interambulacral areas with more than two columns of plates, as was the case for Archaeocidaris,
with its four columns of interambulacral plates. In addition to having multiple columns of
interambulacral plates, the interambulacral plates of Archaeocidaridae indet. also bear crenulate
tubercles (Fig. 5.2.6, 5.2.7). In regular echinoids, crenulate tubercles interlock with crenulations
on the acetabulum of the spine. In Archaeocidaridae indet., these crenulate tubercles appear to
have acted as a restricted-pivot tubercle sensu Smith (1980b), which would have offered each
tubercle a lesser range of motion, but resulted in more sturdy support for the spines. Restricted-
pivot tubercles are often associated with extant taxa which use large primary spines for defense
200
against predators (Smith, 1980b), and this may have been the utility of crenulate tubercles in
Archaeocidaridae indet. Regardless of their purpose, their presence in the indeterminate
archaeocidarid described herein indicates that crenulate tubercles, a character previously first
definitively present in the fossil record from crown group echinoids such as Eotiaris was in fact
present in members of the stem group; the archaeocidarids. The most basal cidaroids (family
Miocidaridae) and some of the most basal euechinoids of the Echinothurioida display crenulate
tubercles (Kroh and Smith, 2010; Thompson et al., 2015b), and as such it is not entirely
surprising that this character appears to have preceded the euechinoid-cidaroid divergence. The
evolution of crenulate tubercles also appears to have preceded the evolution of the perignathic
girdle and the reduction in coronal plating from four columns of interambulacral plates, to two
(Fig. 5.4). Furthermore this character seems to have been widespread, as crenulate tubercles are
known in Permian archaeocidarids from Timor (Wanner, 1841, Fig. 2.1), Pakistan (Waagen,
1885), and Hungary (Mihály, 1980). The discovery of archaeocidarids with crenulate tubercles
allows for a better understanding of the probable last common ancestor of the euechinoids and
cidaroids. That the innovation of crenulate tubercles likely preceded the evolution of a
perignathic girdle and the reduction of interambulacral plating to two columns indicates that the
evolution of crenulate tubercles probably preceded the last common ancestor of euechinoids and
cidaroids, and thus this last common ancestor likely bore crenulate tubercles. It is, of course
possible that the Archaeocidaridae indet. described herein is a separate lineage of
archaeocidarids, which convergently evolved crenulate tubercles that are not homologous to
those of the miocidarids. The ideal way to test this hypothesis would be through phylogenetic
analyses. The material described herein, however, is too incomplete to incorporate into a
quantitative phylogenetic analysis as the specimens described herein are composed solely of
201
disarticulated material, with numerous characters that would need to be coded as unknown.
Furthermore, this unknown data is not random, given that the taxon’s interambulacral plates and
spines are the only morphological features that are preserved. Non-random preservation, and
subsequent non-random missing characters can introduce systematic bias into the topological
placement of taxa in cladograms (Sansom and Wills, 2013). This can be a particularly serious
problem with respect to determining the phylogenetic placement of taxa near the crown group-
stem group transition, as crown group taxa with many missing characters can appear to be
members of the stem group (Sansom et al., 2010). That the indeterminate archaeocidarid
described herein had more than two columns of interambulacral plates indicates it is likely a
member of the stem group. We hope, however, to test the hypotheses of character evolution put
forth herein in a rigorous phylogenetic context with the discovery of more complete specimens in
the future.
CONCLUSIONS
The Permian echinoid fauna described here expands upon previously known echinoid diversity
in the Permian and sheds light on the Paleozoic divergence of crown group echinoids. This is the
most diverse fauna of echinoids known from the Permian, and indicates that a number of major
families, the Archaeocidaridae, Proterocidaridae, and the Miocidaridae coexisted in reefal
environments adjacent to the Delaware Basin. The presence of Eotiaris guadalupensis in reefal
environments, similar to those inhabited by the European Eotiaris keyserlingi, may indicate that
stem group cidaroids originated and preferentially thrived in these reefal environments.
Furthermore, the presence of crenulate tubercles on the indeterminate archaeocidarid indicates
that crenulate tubercles evolved in archaeocidarids, likely before the reduction in interambulacral
202
columns from four to two, and the evolution of the first perignathic girdle. Crenulate tubercles
may thus be plesiomorphic with respect to the echinoid crown group.
203
Figure 5.1. Cidaroid echinoids from the Bell Canyon Formation. (1), Straight spine of Eotiaris
guadalupensis (USNM 610605c); (2), Clavate spine of Eotiaris guadalupensis (USNM
610605a); (3), Spine base and milled ring of USNM 610605 (Same as 2.2); (4), Paratype of
Eotiaris guadalupensis (USNM 610601). This specimen consists of a single interambulacral area
with at least six interambulacral plates in each column; (5), Holotype of Eotiaris guadalupensis
(USNM 610600). Note crenulate tubercles and spine, which is morphologically similar to those
in Figure 1.1-1.3. Scale bars 2.5 mm.
204
205
Figure 5.2. Permian archaeocidarids from Timor, Kansas, the Salt Range (presumably Pakistan)
and west Texas, and ambulacral and interambulacral plates of Pronechinus? sp. (1),
Interambulacral plate of Permocidaris? timorensis (Wanner, 1941) from the Permian of Timor
(RGM 835575). Specific locality unknown. Crenulate tubercle of this taxon is very similar to
those of the indeterminate archaeocidarid in Figure 2.6 and 2.7; (2), Interambulacral plate of
holotype of Archaeocidaris manhattanensis (MGL 206289). Note plate dimensions, which are
similar to those of Archaeocidaridae indet. in Figure 2.7-2.9; (3), Syntype of Archaeocidaris
forbesiana (NMS G.1871.1.34); (4), First plate morphotype of Archaeocidaridae indet. (USNM
617188a); (5), interior view of same plate, note lack of denticulate margin indicating median
location of plates; (6), Crenulate interambulacral plate of Archaeocidaridae indet. (USNM
617188b); (7), Crenulate interambulacral plate of second plate morphotype of Archaeocidaridae
indet. (USNM 617189); (8), Adambulacral second interambulacral plate morphotype of
Archaeocidaridae indet. (USNM 617189). Note plate dimensions which are much higher than
wide; (9), Interior side of the same; (10), Spine of Archaeocidaridae indet. (USNM 617187a);
(11), The same in cross section. Note triangular cross section; (12), Base and acetabulum of
spine (USNM 617187b); (13), The same spine in plan view; (14), Thin spine of
Archaeocidaridae indet. (USNM 617187c); (15), Ambulacral plate of Pronechinus? sp. with well
developed peripodial ring surrounding pore pairs (USNM 617193); (16), Ambulacral plate of
Pronechinus? sp. with peripodial ring and imperforate primary tubercle (USNM 617194); (17),
Same as 2.16 but in side view; (18), Ambulacral plate lacking peripodial ring and with
imperforate primary tubercle (USNM 617195), (19), Ambulacral plate lacking well-defined
peripodial ring and primary tubercle (USNM 617196); (20), Interambulacral plate of
Pronechinus? sp. with large imperforate primary tubercle (USNM 617197a); (21), Non-
206
tuberculate interambulacral plate of Pronechinus? sp. (USNM 617197b). All scale bars are 2.5
mm.
207
208
Figure 5.3. Pyramids of indeterminate echinoid. (1), Pyramid of indeterminate echinoid (USNM
617190a); (2), Pyramid of indeterminate echinoid (USNM 617190b). Tooth comes to a single
non-serrate point; (3), Pyramid of indeterminate echinoid (USNM 617190c). Scale bars are 2.5
mm.
209
210
Figure 5.4. Simplified cartoon showing hypothesized order of select character acquisitions along
the branch from stem group to crown group echinoids, and stratigraphic distribution of taxa
described herein and other named proterocidarid, archaeocidarid and miocidarid taxa from the
Permian (excluding Eotiaris connorsi [Kier, 1965], whose position within the Miocidaridae is
currently being tested [Thompson, unpublished data]). The oldest demonstrable euechinoid is
also shown for comparison. The lineage leading to archaeocidarids and miocidarids likely
diverged from that giving rise to the proterocidarids in the Devonian or earlier (Smith, 1984) and
is not figured. Only Archaeocidaris species known from relatively articulated test material are
included and branching events do not reflect time-scaling. Archaeocidaris is a stem group
echinoid, while Eotiaris guadalupensis Thompson, 2015 and Diademopsis ex. gr. heberti
(Agassiz and Desor, 1946) are the oldest fossil cidaroid and euechinoid respectively, and are
amongst the oldest members of the crown group. The crenulate tubercles of Archaeocidaridae
indet. indicate that crenulate tubercles likely evolved in the echinoid stem group before the
reduction in interambulacral columns from four to two, and before the acquisition of a
perignathic girdle. Consequently, crenulate tubercles appear to have preceded the acquisition of
rigid coronal plating in cidaroids. For full suite of characters defining the echinoid crown group,
see Kroh and Smith (2010). Occurrences of Archaeocidaris are from Boos (1929) and Etheridge
(1892). Proterocidarid occurrences are from Kier (1965), König (1982) and this study. The
indeterminate archaeocidarid is described herein. Crown group echinoids within gray box. Figure
plotted in STRAP (Bell and Lloyd, 2015).
211
210
220
230
240
250
260
270
280
290
300
Kasimovian
Gzhelian
Asselian
Sakmarian
Artinskian
Kungurian
Roadian
Wordian
Capitanian
Wuchiapingian
Changhsingian
Induan
Olenekian
Anisian
Ladinian
Carnian
Norian
Rhaetian
Pennsylvanian
Cisuralian
Guadalupian
Lopingian
Lower
Middle
Upper
Permian Triassic
Archaeocidaris
Indeterminate archaeocidarid
Euechinoidea
Proterocidaridae
Crown group echinoids
Rigid coronal plating
Perignathic girdle
Two column
interambulacra
Crenulate tubercles
Cidaroidea
Diademopsis ex. gr. heberti
Eotiaris guadalupensis
Pronechinus? sp.
Archaeocidaridae indet.
MYA
212
Chapter 6. Phylogenetic analysis of the Archaeocidaridae
(Echinodermata: Echinoidea) and the origin of crown group
echinoids
INTRODUCTION
Echinoids , or sea urchins, are important members of modern communities (Lohrer et al.,
2004; Tuya et al., 2004), and have been abundant in ancient ecosystems since the late Paleozoic
(Schneider et al., 2008). Though they are first known from the Ordovician (Pisera, 1994; Smith
and Savill, 2001), echinoids reach their peak Paleozoic diversity in the Mississippian subperiod
(Kier, 1965; Smith, 1984) and undergo an evolutionary bottleneck during the end-Permian mass
extinction at the end of the Paleozoic Era (Erwin, 1993, 1994; Hagdorn, 2017; Pietsch et al.,
2017; Thompson et al., 2018; Thuy et al., 2017; Twitchett and Oji, 2005). Although the
phylogenetic relationships amongst the post-Paleozoic echinoids are generally thought to be
well-constrained (Kroh and Smith, 2010; Smith et al., 2006; Thompson et al., 2017a), the
phylogenetic relationships of Paleozoic taxa, which make up the majority of the stem group, are
more poorly understood. A key unanswered question in the evolutionary history of echinoids
concerns the precise relationships of the echinoid crown group to the stem group echinoids of the
Paleozoic. While the Paleozoic was home to numerous families of echinoids with widely
disparate morphologies (Kier, 1965; Smith, 1984), the most taxonomically diverse and abundant
of these families is the Archaeocidaridae. The archaeocidarids, which are present in the fossil
record from the Devonian to the Permian, have long been proposed to be paraphyletic with
respect to the echinoid crown group (Kroh and Smith, 2010; Smith, 1984). In particular,
numerous authors have treated the genus Archaeocidaris, as the sister group to all crown group
213
echinoids (Kier, 1968a; Lewis and Ensom, 1982; Smith and Hollingworth, 1990; Smith, 1984),
and this genus has been used as the outgroup in numerous phylogenetic analyses of post-
Paleozoic echinoids (Kroh and Smith, 2010; Littlewood and Smith, 1995; Smith, 2007; Smith et
al., 2006).
We herein set out to determine the species level relationships amongst the
Archaeocidaridae and the Paleozoic members of the crown group family Miocidaridae. Until
now, the phylogenetic relationships of both the species and genera within the archaeocidarids
have not been examined using modern phylogenetic methods. Our goals were to understand the
relationships amongst the different species and genera of the archaeocidarids, while focusing on
the relationship of the echinoid crown group to the archaeocidarids. In particular, we set out to
determine if the echinoid crown group, represented by the miocidarids, was the sister taxon to a
particular genus of archaeocidarid (e.g Archaeocidaris), or if any genus of archaeocidarid is
paraphyletic with respect to the crown group. Additionally, we set out to determine the fit of
archaeocidarid cladograms to the stratigraphic record of archaeocidarid occurrences, to compare
the fossil record of archaeocidarids to that of other post-Paleozoic echinoid groups. Though we
do not herein set out to provide a systematic revision of all archaeocidarid species, in
undertaking this project, we felt it necessary to provide novel observations about previously
described archaeocidarids and miocidarids and to describe for the first time some new
archaeocidarid species.
The Family Archaeocidaridae
There are four well-known genera within the Archaeocidaridae: Nortonechinus Thomas,
1920, Lepidocidaris Meek & Worthen, 1869 , Polytaxicidaris Kier, 1958, and Archaeocidaris
M’Coy, 1844. There is also a fifth genus, Devonocidaris Thomas, 1921, which is herein
214
described in detail for the first time. Archaeocidarids were the most diverse and abundant
echinoids in the Paleozoic and are known to have inhabited numerous substrates and
environments. Their disarticulated spines and plates have a wide biogeographic distribution and
are a dominant bioclastic contributors to Late Paleozoic strata. This is especially the case in the
Pennsylvanian, where entire strata are composed entirely of their disarticulated plates and spines
(Schneider et al., 2008). In addition to their abundance, the archaeocidarids were also the first
echinoids to exhibit large, movable, spines and use these spines to attract encrusting organisms
(Borszcz, 2012; Schneider, 2003, 2010; Schneider et al., 2008; Schneider et al., 2005; Thompson
et al., 2015a).
Previous Hypotheses for the Relationships of the Archaeocidarids and Crown Group
Echinoids
The archaeocidarid echinoids have long been associated with the cidaroids of the crown
group (Desor, 1858; Lovén, 1875; M’Coy, 1844), a point which is typified by the fact that many
of the oldest described species of Archaeocidaris, were initially attributed to the extant cidaroid
genus Cidaris (Buch, 1842; de Koninck, 1842-1844; Fleming, 1828). Since the 20
th
century,
more work has focused on the phylogenetic relationships between the different genera of
archaeocidarids, and on their relationship to the crown group. Jackson (1912) proposed that the
archaeocidaridae and the cidaroida were not closely related. He suggested that the cidaroida, and
most other post-Paleozoic echinoids, were the descendants of the bothriocidaroids while the
archaeocidarids were more closely related to the lepidocentrids, palaechinids and lepidesthids.
Mortensen (1928b) thought Jackson (1912) incorrect, and based upon the imbrication present in
archaeocidarids and the miocidarid genus Eotiaris, regarded the archaeocidarids as the ancestors
215
of the cidaroids. He also regarded Nortonechinus and Archaeocidaris as the only two confident
archaeocidarids, choosing to treat Lepidocidaris as a lepidesthid or lepidocentrid.
Following the description of Polytaxicidaris (Kier, 1958a), Kier (1965), proposed that
Archaeocidaris and Polytaxicidaris were sister taxa, and this clade was in turn the sister group to
Lepidocidaris. He additionally agreed with Mortensen (1928b) that Eotiaris and the crown group
were the descendants of Archaeocidaris. Building on these earlier ideas, Kier (1968a) based
much of his evidence for the relationships of the archaeocidarids upon the number of columns of
interambulacral plates in each genus. He considered Nortonechinus, Polytaxicidaris,
Archaeocidaris, and Eotiaris to be a series of ancestors and descendants, each with a sequential
reduction in the number of columns of interambulacral plates. Based upon its sunken tubercles,
multiserial ambulacra, and the enlargement of every third ambulacral plate, Kier considered
Lepidocidaris to be a lineage separate from Archaeocidaris, Polytaxicidaris and the miocidarids,
that separately descended from Nortonechinus. Smith’s (1984) cladogram was similar in that he
treated Archaeocidaris and the crown group echinoids as sister taxa, and Lepidocidaris as sister
group to a clade of Polytaxicidaris, Archaeocidaris and the crown group echinoids. Smith,
however, considered the proterocidarid echinoids Perischodomus and Hyattechinus to be more
derived than Nortonechinus, and thus sister group to the clade of Lepidocidaris, Polytaxicidaris,
Archaeocidaris and the crown group. In order to tease apart the relationships amongst these
different genera, and their relationship to the crown group, we employed parsimony-based and
Bayesian phylogenetic methods to attempt to understand the species-level relationships of the
archaeocidarids and Paleozoic miocidarids.
216
METHODS
Taxon Choice
The majority of the archaeocidarid fossil record is comprised of disarticulated interambulacral
plates and spines. Numerous species of Archaeocidaris have been named, however, the majority
of these taxa were erected from solely disarticulated interambulacral plates and spines, with the
details of test plating, the Aristotle’s lantern, and the ambulacra unknown. Because of their
incompleteness, these taxa cannot be assigned confidently to the genus Archaeocidaris (Kier
1958, 1965). Although these taxa can be useful for reconstructing paleobiogeography or
environmental preference, they are of little use for phylogenetic studies. We have thus chosen to
include only taxa which have been described from relatively complete specimens, and that have
diagnoses which are consistent throughout all known specimens from that species. The species
included herein are thus all known from relatively complete test material including associated,
complete interambulacral columns, associated spines, and in most cases, ambulacral areas.
Because we were interested in understanding the relationship of the crown group echinoids to the
archaeocidarids of the stem group, in addition to the twenty species from the family
Archaeocidaridae, we also included the four Paleozoic species of the crown group echinoid
family Miocidaridae. A list of all taxa included in our analyses is found in Table 6.1 and select
genera are figured in Figure 6.1.
Phylogenetic Analyses
We set out to infer the phylogenetic relationships of twenty four species comprising six genera of
the archaeocidarids and miocidarids. We coded thirty characters from museum specimens and
the literature, twenty-three of which were binary and seven of which were multistate. Characters
and character states are available from the author. Character number fourteen, which details the
217
number of columns of interambulacral plates in each interambulacrum, was ordered following
justification of Kier (1968). To infer phylogenetic relationships, we employed both parsimony-
based and Bayesian approaches. Analyses using the maximum parsimony optimality criterion
were run in PAUP* Version 4 (Swofford, 2003) with trees rooted on the outgroup Nortonechinus
welleri. A heuristic search was run with 10000 random addition search replicates (RASs) using
tree bisection reconnection (TBR). The strict consensus of these trees is shown in Figure 6.2A.
Analyses were then re-run with characters re-weighted based upon their retention indices (RI)
with 10000 RASs and TBR. A strict consensus of these trees is show in Figure 3. Node support
for parsimony-based analyses was assessed using Bremer Support/Decay Indices (Bremer, 1994)
and bootstrap resampling (Felsenstein, 1985). Bremer Support values were calculated up to 2
with 10 RASs for unweighted data and nodes with values of ≥ 2 are indicated on Figure 6.2.
Bremer support values were only calculated up to 2 because further calculations proved too time-
intensive.
In addition to parsimony-based analyses, we also employed Bayesian approaches to tree
inference, which have recently been shown to be more accurate than parsimony when
reconstructing phylogenetic relationships from discrete morphological data (O’Reilly et al.,
2016; Puttick et al., 2017; Wright and Hillis, 2014). Bayesian analyses were run using MrBayes
version 3.2 (Ronquist et al., 2012) using the Mk model of character state change (Lewis, 2001).
Character coding was set to variable to reflect the acquisition bias often present in phylogenetics
using morphological datasets, where constant characters are typically excluded. The Brlens
parameter was set to be unconstrained and the prior on branch lengths was an exponential
distribution with parameter λ=1. The prior on the shape parameter α of the gamma distribution
of rate variation was also an exponential distribution with λ=1. We ran our analyses with a
218
number of different values for the symmetric Dirichlet hyperprior, α. The value of α reflects the
degree of character transition rate asymmetry, with larger values of α reflecting less asymmetry
of transition rates (more symmetrical transition rates) and smaller values of α corresponding to
more asymmetry in the model (Wright et al., 2016). We ran our analyses using six different
values of α: ∞, 10, 2, 1, 0.2, and 0.05.
Different combinations of topologies and parameters were sampled using Markov Chain
Monte Carlo (MCMC) in MrBayes. We ran two independent MCMC analyses, each with four
chains. We ran our analyses for 12,000,000 generations, sampling every 500 generations. The
burn in fraction was set to 25%. Convergence was assessed using the average standard deviation
of split frequencies, and all analyses were run until this value was at least below 0.01. The results
of our Bayesian analyses are displayed using 50% majority rule trees, which have recently been
shown to most accurately display the results of the posterior distribution in analyses of
morphological character data (O’Reilly and Donoghue, 2017). 50% majority rule trees show all
nodes found in greater than 50% of trees present the posterior distribution, and clade support for
each node is shown as the proportion of trees in the posterior that contained that node.
Comparing the Fit of Most Parsimonious Trees to the Stratigraphic Record
We compared the fit of parsimonious trees resulting from our reweighted parsimony analysis to
the occurrence of species in the fossil record. The fit of these trees to the stratigraphic record was
determined by calculating the stratigraphic consistency index (SCI) (Huelsenbeck, 1994),
relative completeness index (RCI) (Benton, 1994), and gap excess ratio (GER) (Wills, 1999) for
each tree using the R package STRAP (Bell and Lloyd, 2015). The significance of these values
was calculated with 1000 permutations. The SCI is a ratio of the number of stratigraphically
consistent nodes to the number of internal nodes on the tree excluding the root. A node is
219
stratigraphically consistent if the oldest taxon above the node is younger than or the same age as
the sister taxon of the node (Huelsenbeck, 1994). The SCI is sensitive to the balance of a given
tree, with more pectinate trees having a wider range of possible SCI values than more balanced
trees (Siddall, 1996, 1997; Wills, 1999). Both the RCI and GER make use of the minimum
implied gap (MIG), which is the difference in time between the oldest appearance of a taxon in
the fossil record and the oldest appearance of its sister taxon. The RCI is dependent upon the sum
of the ratio of the MIG and the simple range length (SLR: the observed stratigraphic range of a
taxon) for each taxon (Benton, 1994). The RCI is also sensitive to tree balance, with totally
pectinate trees capable of yielding the maximum and minimum MIG for the tree dependent upon
the stratigraphic ranges of the included taxa. The GER is dependent upon the ratio of the
difference between the MIG and the minimum amount of inferred range possible, and the
difference between the maximum gap (G
max
) and minimum (G
min
) gap possible for a given tree
fit to stratigraphy. We choose to focus on the GER and SCI for our trees, and base comparisons
to other previously formulated datasets on the GER. For calculation of SCI, RCI and GER,
stratigraphic ranges were binned at the stage level and stratigraphic ranges of all taxa are
presented in Table 1. Taxa occurred in fifteen time bins from the Frasnian (Devonian) to the
Wuchiapingian (Permian). Absolute ages for stage bases were from Gradstein et al. (2012).
RESULTS
Phylogenetic Analyses
When characters were equally weighted, the heuristic search found 817 most parsimonious trees.
The length of these trees was 83 steps, the CI was 0.48 and the RI was 0.62. The strict consensus
tree of these trees are shown in Figure 6.2. The topology of the strict consensus of these analyses
220
was largely unresolved. D. primaevus was identified as the sister group to all other
archaeocidarids except for N. welleri, and thus plots as the most basal archaeocidarid other than
Nortonechinus. This was with fairly high bootstrap and Bremer support (BP=81, DI≥2).
Additionally, we found all miocidarids, including E. meurevillensis formed a clade (BP=55,
DI=1), and identified a clade of A. hemispinifera+P. dyeri, albeit with low support (BP=43,
DI=0).
Re-running analyses following reweighting characters by their retention indices resulted
in 102 most parsimonious trees of length 56.55 steps. These trees had CI of 0.545 and RI of 0.69.
The strict consensus of these trees is shown in Figure 6.3 and is more resolved than that of the
unweighted analysis, while still containing the three clades that were present in the unweighted
strict consensus tree. Re-weighing by the RI increased node support for the clades resolved in the
unweighted analysis, and only one node containing all species of Archaeocidaris, Lepidocidaris,
Polytaxicidaris, and the miocidarids was resolved with BP≥75. Within the miocidarids, we
resolved a sister pairing of E. guadalupensis and E. keyserlingi (BP=63), which was sister to E.
meurevilleneis (BP=50). We additionally resolved a clade of A. marmorcataractensis, A. rossica,
A. brownwoodensis, and A. subwortheni (BP=44) and a clade of A. whatleyensis and the
miocidarids (BP=18), which was sister to the clade of A. whatleyensis and the miocidarids
(BP=10). This clade was also part of a larger clade with A. ivanovi, A. mosquensis, A. agassizi,
and A. ausichi (BP=6). In the strict consensus tree, this clade formed a polytomy with A. blairi,
A. legrandensis, A. apheles, and a clade of A. wortheni+A. imanis+L. squamosus (BP=16). All
species except for D. primaevus and the outgroup N. welleri thus formed a clade with BP support
of 96. Within this clade, a smaller sister pairing of A. hemispinifera and P. dyeri (BP=57) was
221
present, in addition to P. lirata and a much larger clade of all other taxa except for D. primaevus
and N. welleri.
Our Bayesian analyses resolved largely similar clades as our parsimony analyses,
however clade support varied dependent upon the value of α used for the symmetric Dirichlet
hyperprior (Fig. 6.8), and PPs were, in in all cases, lower than the 0.9 that is typically used to
indicate strong support. When α was large (∞ and 10) and very small (0.05), we resolved the
clade of A. rossica, A. brownwoodensis, A. subwortheni, and A. marmorcataractensis.
Additionally, when α was small (0.2 and 0.05), we resolved a clade of all species except for D.
primeavus and N. welleri. The sister pairing of E. keyserlingi and E. guadalupensis was present
with varying support under five of the six different values of α. A clade of the grown group
echinoids was present in 50% majority rule trees when α was equal to ∞ and 0.05, and when α
was equal to ∞, E. connorsi was identified as the most basal miocidarid, with E. meurevillensis
forming a clade with the sister pairing of E. keyserlingi and E. guadalupensis. Clearly varying
the hyperprior α affected posterior clade support across our analyses, and it is worth noting that
whether or not clades are displayed in the 50% majority rule trees is dependent upon this node
support.
Fit to Stratigraphy
The SCI, RCI, GER and their siginificance of the 102 most parsimonious trees resulting
from the reweighted parsimony analysis are discussed herein. SCI values range from 0.52 to 0.66
and GER values range from 0.67 to 0.71. All RCI are negative, because the MIG is greater than
the SRL in all cases. Using STRAP, we also calculated the uncertainty that observed trees differ
from random by holding the topology of each most parsimonious tree constant, and randomly
permuting the stratigraphic ranges of each taxon. The uncertainty that the observed SCI, RCI,
222
and GER values differ from random is then calculated by comparing the observed SCI, RCI and
GER to the distribution generated from these random permuatations (Wills, 1999). Histograms
showing the distribution of observed, and random SCI and GER values are shown in Figure 6.4.
The dashed line represents the critical value above which congruency indices values differ
significantly from random. As shown in Figure 4 and Supplementary Table 2, all SCI and GER
values are significantly higher than expected at random (p>0.05). The tree with the single highest
SCI, and the strict consensus of the topologies with the seven highest GER values are shown in
Figure 6.5.
DISCUSSION
Phylogenetic Relationships of the Archaeocidarids and Late Paleozoic Echinoid
Macroecology
In our parsimony analyses, the best supported clade found Devonocidaris as the most basal
archaeocidarid, and sister to a clade of all other archaeocidarids except for Nortonechinus. We
also found this topology when the α parameter was low in our Bayesian analyses. This topology
makes sense in light of the stratigraphic occurrence of Devonocidaris, and with respect to its
many columns of interambulacral plates. Interestingly, and in contrast to previous hypotheses
(Kier, 1968a; Smith, 1984), we did not consistently resolve archaeocidarids with multiple
columns of interambulacral plates as monophyletic genera and basal with respect to
Archaeocidaris. In both our unweighted and reweighted parsimony analyses, we found a clade of
A. hemispinifera and P. dyeri, which both have enlargement of ambulacral plates and lack
tubercles on all ambulacral plates. Additionally, in our reweighed parsimony analysis, we found
Lepidocidaris squamosa as nested within a clade with species of Archaeocidaris. Despite its
223
sister pairing with A. immanis, whose ambulacral plates are also not all in contact with the
perradius, L. squamosus has autapomorphies, such as multiserial ambulacra, which may indicate
a more basal placement in the archaeocidaridae and were not parsimony informative, or
included, in this archaeocidarid specific analysis. The sister pairing of P. dyeri and A.
hemispinifera, and the broader topological uncertainty surrounding the placement of P. lirata,
may indicate that the presence of greater than four columns of interambulacral plates within the
archaeocidaridae may have evolved again within the clade, and is not a strictly basal character.
Another clade we resolved in our reweighted parsimony analysis and in the majority rule
trees of three of our six Bayesian analyses was a clade of A. marmorcataractensis, A. rossica, A.
brownwoodensis, and A. subwortheni. These taxa are all known from the Pennsylvanian, and
share multiple rows of scrobicular tubercles, and spines covered with numerous spinules. In our
reweighted analyses, these two characters seem to play an important role in shaping the topology
of the strict consensus tree. Archaeocidarids with smooth spines and a single ring of scrobicular
tubercles, such as A. wortheni, A. blairi, A. legrandensis, and A. apheles tend to occupy earlier
branching positions within the most parsimonious trees (Fig. 6.3), while with taxa with
bilaterally symmetrical, spinule-covered spines, and multiple rings of scrobicular tubercles are
more crown-ward. This reflects the observation of Jackson (1912) who postulated that smooth-
spined archaeocidarids were more primitive than those with spinulose spines. This trend also
seems to reflect the stratigraphic distribution of archaeocidarid and miocidarid species, as most
species with extensively sculptured spines occur from the Pennsylvanian or Permian, and most
with smooth, striate, and radially symmetrical spines are from the Mississippian. This character
could thus in part underlie the relatively high GER values calculated for our most parsimonious
trees.
224
The proliferation of a clade of archaeocidarids and miocidarids with extensively
ornamented spines may reflect large-scale macroecological changes taking place during the late
Paleozoic. For instance, the first echinoids to host encrusting organisms on their spines are
archaeocidarid echinoids known from the Pennsylvanian (Borszcz, 2012; Schneider, 2003, 2010;
Schneider et al., 2008; Schneider et al., 2005; Thompson et al., 2015a). Epibionts are common on
the spines of extant cidaroids, which have an external cortex layer (David et al., 2009; Hopkins
et al., 2004; Linse et al., 2008). The presence of extensively ornamented spines, likely associated
with a cortex layer, may have allowed for the development of commensal relationships between
archaeocidarids and epibionts, and the diversification of archaeocidarids with more ornate and
spinulose spines.
A second hypothesis for the development of more extensive spinosity concerns the
advent of more enhanced predation in the Late Paleozoic. The Late Paleozoic saw turnover in a
number of vertebrate clades, which resulted in an abundance and diversity of durophagous
predators (Sallan and Coates, 2010). This enhanced predation from durophagous predators has
been tied to the evolution of anti-predatory strategies in crinoids (Sallan et al., 2011; Syverson
and Baumiller, 2014; Thompson and Ausich, 2015). The archaeocidarids were amongst the first
echinoids to evolve large spines and associated primary tubercles, perhaps as an anti-predatory
response (Smith, 2005). It is also possible that the enhanced spinulosity, and morphological
differentiation of spines seen in archaeocidarids in the Pennsylvanian also evolved as a result of
predation pressure. Both commensalism and predation driven evolutionary trends remain
hypotheses, and further work will be necessary to investigate the drivers underlying the
diversification of archaeocidarids with more spinulose and differentiated spines.
225
Fit to the Stratigraphic Record
The coronal plates of archaeocidarids are imbricate, and overlap overtop of adjacent plates. They
lack the stereomic interlocking present in Post-Paleozoic taxa (Smith, 1980a, 1984) and thus
disarticulate rapidly following soft tissue decay (Thompson and Ausich, 2016). As such, the
majority of the archaeocidarid record is composed of disarticulated plates and spines, with the
whole tests necessary for phylogenetic analyses, usually limited to lagerstätte (Schneider et al.,
2005; Thompson et al., 2015a). As such it is no surprise that the GER values for the
archaeocidarid trees, while better than the mean value for 1000 Phanerozioc cladograms (Benton
et al., 2000), are worse than those calculated for other echinoid groups. Jurassic irregular
echinoids (Saucéde et al. 2007, GER=.80), the cidaroids+autolodonts clade (Kroh and Smith,
2010, GER=.81), atelostomates (Kroh and Smith, 2010, GER=.84), neognathostomates (Kroh
and Smith, 2010, GER=.925), and the tree for all Post-Paleozoic taxa (Kroh and Smith, 2010,
GER=.8) have higher GER values than the archaeocidarid trees. These values for archaeocidarids
are better, however, than for the Echinacea+Calycina+basal irregulars (Kroh and Smith 2010,
(GER=.60), which live on hard substrates and thus experience a lower preservation potential than
clades inhabiting other environments (Greenstein, 1991; Kier, 1977a; Nebelsick, 1996). It is
worth noting, however, that the taxonomic scale and sampling intensity at which our analyses
and those of Kroh and Smith (2010) differ, and that these characteristics can influence analyses
of stratigraphic congruence (Benton et al., 2000; Wills et al., 2008).
Origin of the Echinoid Crown Group
The primary goal of our analyses was to determine the phylogenetic placement of the crown
echinoids of the Miocidaridae with respect to genera of the Archaeocidaridae. Though our results
were fairly unresolved and nodes were not often well-supported, the results of our reweighted
226
analysis do indicate that the crown group echinoids appear to be paraphyletic with respect to the
genus Archaeocidaris. The only clade present in all of our parsimony-based strict consensus
trees, and five of our Bayesian majority rule trees, was the sister pairing of the miocidarid
echinoids E. guadalupensis and E. keyserlingi. We also resolved these two taxa in a clade with
all other miocidarids in our reweighted and unweighted parsimony analyses, and in two of our
Bayesian analyses. These two taxa have plotted as sister taxa in other previous analyses
(Thompson et al., 2018; Thompson et al., 2015b). In our reweighted parsimony analysis, we
found the clade of miocidarids as sister group to the Viséan A. whatleyensis. This clade was in
turn sister group to the aforementioned clade of A. marmorcataractensis, A. rossica, A.
brownwoodensis, and A. subwortheni. All of these taxa have multiple rows of scrobicular
tubercles on their interambulacral plates, which differentiates them from archaeocidarids such as
A. wortheni, A. agassizi, A. blairi, and A. legrandensis. Furthermore the clade of archaeocidarids
and miocidarids is predominantly composed of taxa with greater sculpturing on their spines
(though this is not entirely the case given the presence of E. meurevillensis, which has smooth
striate spines).
It thus appears that the Miocidaridae may be closely related to archaeocidarids with more
sculpturing on their spines and with multiple rows of scrobicular tubercles on their
interambulacral plates. Thompson et al. (2017b) recently provided evidence to support the
hypothesis that crown group echinoids evolved from archaeocidarid ancestors with crenulate
tubercles. Though their material was too fragmentary to analyze using quantitative
phylogenetics, they showed that crenulate tubercles appear to have been widespread in Permian
archaeocidarids. The only non-miocidarid with crenulate tubercles in our analysis, A. apheles,
did not plot as sister group to the miocidarids in any of our analyses. This implies that crenulate
227
tubercles may have arisen multiple times within the Archaeocidaridae. The Permian
archaeocidarids mentioned by Thompson et al. (2017b) are of note, however, because some of
these specimens (though not all of them) also have multiple rows of scrobicular tubercles on
their interambulacral plates. This thus provides a secondary line of evidence in favor of the
hypothesis that the miocidarids evolved from species of Archaeocidaris with multiple rows of
secondary tubercles on their interambulacral plates.
Very recently, Hagdorn (2017) has proposed that at least some of the crown group
echinoids may have evolved from proterocidarid echinoids, which along with other stem-group
taxa have recently been found in post-Paleozoic strata (Thompson et al., 2018; Thuy et al.,
2017). In our analyses, we found strong support for the crown group echinoids nested within the
Archaeocidaridae, and in particular within Archaeocidaris, however we did not include any
proterocidarids to explicitly test Hagdorn’s hypothesis. Thompson et al. (2018) found strong
support for distinct clades of proterocidarids and archaeocidarid+crown group echinoids, which
provides little support for the hypothesis of Hagdorn (2017). Nevertheless, we are currently
working on a larger analysis of all stem group echinoid genera, which will allow us to directly
test the hypothesis that the crown group echinoids evolved from proterocidarid ancestors.
CONCLUSIONS
The results of the analyses herein indicate that the different genera of the
Archaeocidaridae do not seem to be monophyletic, as opposed to what has been proposed by
previous authors (Kier, 1968a; Smith, 1984). Furthermore, the crown group echinoids, as
exemplified by the family Miocidaridae, appear to be nested within the genus Archaeocidaris,
which is thus paraphyletic. The diversification of the archaeocidarids may also be linked to
228
broader macroecological shifts taking place in the late Paleozoic, including increasing predation
and the advent of epibiont encrustation on echinoid spines. Finally, the fit of our phylogenetic
hypotheses to the stratigraphic order of archaeocidarid occurrences is generally worse than other
post-Paleozoic echinoid groups, which may be due to the imbricate nature of the archaeocidarid
test.
SYSTEMATIC PALEONTOLOGY
Abbreviations for specimen numbers are as follows: USNM = United Stated National
Museum, LACIMP = Los Angeles County Museum of Natural History, SUI = University of
Iowa Paleontology Repository, BMNH=Natural History Museum, London.
Class Echinoidea LESKE, 1778
Family Archaeocidaridae M’COY, 1844
Type genus.⎯Archaeocidaris M’Coy, 1844
Other genera.⎯ Polytaxicidaris Kier, 1958, Lepidocidaris Meek and Worthen, 1869,
Nortonechinus Thomas, 1920, Devonocidaris, Thomas, 1921.
Diagnosis.⎯Stem group echinoids with large primary tubercles on all but the most adapical
interambulacral plates supporting long or fairly long spines. Plating imbricate. Interambulacral
areas composed of four or more columns of interambulacral plates at ambitus. All
archaeocidarids have two columns of ambulacral plates in each ambulacral area adapically.
Rotulae hinge jointed. No perignathic girdle.
Genus Devonocidaris Thomas, 1924
229
Type species.⎯ Devonocidaris primaevus Belanski, 1927 from the Frasnian of Iowa
Diagnosis.⎯Archaeocidarid with test radially oriented. Interambulacra arranged into 8 columns
of plates. Interambulacral plates thin, imbricating orally and with adjacent plates.
Interambulacral plates with single perforate, non-crenulate primary tubercle, lacking significant
vertical relief. Perforate, non-crenulate secondary tubercles present on interambulacral plates.
Secondary tubercles sparsely arranged on plates. Ambulacra arranged into two columns, each
plate bearing perforate tubercle with large spines. Interambulacral and ambulacral spines thin.
Occurrence.⎯ Devonian-Mississippian of North America and Belgium.
Devonocidaris primaevus Belanski, 1927
Fig. 6.6.
1927 Devonocidaris primaevus Belanski, 1927, p. 84-86, Pl 13, figs. 29-34
Material Studied.⎯ SUI 40246 from the Upper Devonian, Shellrock Formation, Mason City
Member, along the Shellrock River and SUI 53171A and B from the Devonian Cedar Valley
Formation, Johnson County, Iowa.
Diagnosis.⎯Devonocidaris with eight columns of interambulacral plates. Ambulacral plates
uniserial, simple, orally and to ambitus, with perforate tubercles. Interambulacral plates with
single perforate primary tubercle and multiple perforate secondary tubercles. Spines thin distally
tapering. Lacking ornamentation.
Occurrence.⎯Devonian (Frasnian) Shell Rock and Cedar Valley Formations of Iowa, USA.
Description.⎯ Belanski’s (1927) original description of the spines is adequate, however, it
lacked details of the test. A more comprehensive description is provided herein. Test regular,
circular in plan view. Plating is imbricate adorally, with more oral plates imbricating over top of
230
more adoral plates. Test likely compressed vertically. Apical system not preserved. Peristome
and perignathic girdle unknown. Adoral surface of test unknown. Lanterns are present on
specimens SUI 40246, and SUI 53171B, however, the compacted nature of the preservation has
obscured their details.
Ambulacra in two columns, imbricating under interambulacra. Ambulacral plates simple,
about three times wider than high. Single spine-bearing tubercle present on interambulacral
plates. Pore pairs are uniserial and con-conjugate. Interambulacral area about 3.5 times wider
than ambulacral area. Composed of eight columns of thin interambulacral plates. Adambulacral
plates between 1 and 1.5 times wider than high at ambitus. Median interambulacral plates
between 1.4 and 4 times wider than high at ambitus. Primary tubercle not well developed;
mamelon low, non-crenulate, perforate; situated on a very faint basal terrace. Secondary
tubercles perforate; not densely arranged on plate. Zero to five secondary tubercles per plate.
Spines straight, slender. Circular in cross section. Apparently less than half the length of
the test diameter. Milled ring present towards the base of the spine. Spines lack ornamanetation.
Secondary spines present on ambulacral tubercles and secondary tubercles. Shorter and more
slender than primary spines, but otherwise similar.
Discussion.⎯ Smith and Kroh (2011) assigned Devonocidaris primaevus Belanski, 1927 to the
genus Deneeechinus based on its similarity to Deneechinus tenuispinus Jackson, 1929. We have
examined the type of D. tenuispinus, which is not well preserved, and its taxonomic affinities are
not well-known. D. primaevus is similar to D. tenuispinus, in regard to the morphology of the
tuberculate plates and thin-ness of the plates. Additionally, the spines are long and slender in
both taxa. What differs between the two taxa, however, are their size, with D. primaevus being
much smaller than D. teunispinus, and the number of interambulacral columns present in each
231
area. D. primaevus contains eight interambulacral columns in each area, while D. tenuispinus
contains at least ten (Jackson, 1929, Smith and Kroh, 2011). Given the poor state of preservation
of D. tenuispinus, we find it best to treat the two taxa as separate genera, and thus retain D.
primaevus within Devonocidaris.
Genus Archaeocidaris M’Coy, 1844
Type species.⎯Cidaris urii Fleming, 1828, p. 478, from the Mississippian of northwestern
Europe.
Diagnosis.⎯ Archaeocidarid radial test. Four columns of interambulacral plates. Interambulacral
plates bearing a single, large imperforate tubercle and imbricating slightly orally. This tubercle is
non-crenulate in all species except for one. Interambulacral plates with a single or multiple rows
of scrobicular tubercles dependent upon the species, and varying along test in some species.
Some species with plates wider than high, some with plates higher than wide. Ambulacral plates
simple and uniserial in most species, but in one species not all plates reaching perradial suture.
Bearing large spines relative to body size. Spine morphology varying dependent upon the species
and along the test, but spine morphotypes range from straight to spinulate to clavate, and with
various combinations of each. Spines hollow in all species but one. In taxa where most adapical
plates are known, they appear to be nontuberculate and organized into up to six or seven
columns. Teeth of most species are serrated. Radial water vessel internal.
Occurrence.⎯ Mississippian-Permian of North America, Russia, Europe, Australia.
Archaeocidaris legrandensis Miller and Gurley, 1890
Fig. 6.1C
232
1890 Archaeocidaris legrandensis Miller and Gurley, p. 59, Pl 10, fig. 15
1895 Archaeocidaris legrandensis Keyes, p. 185
1904 Archaeocidaris legrandensis Klem, p. 50
1910 Archaeocidaris legrandensis Lambert and Thiéry, p. 124
1912 Archaeocidaris legrandensis (part) Jackson, p. 260 Pl 10, figs. 3,4.
1958 Archaeocidaris aliquantula Kier, p. 7,8, Pl. 3C
Diagnosis.⎯Diagnosis of Kier (1958) for A. aliquantula.
Occurrence.⎯Kinderhookian (Tournaisian) Gilmore City and Maynes Creek Formations of
Iowa.
Description.⎯The description of Kier (1958) for A. aliquantula is sufficient.
Discussion.⎯ A. legrandensis was first described from the Kinderhook of Iowa, however, it’s
precise stratigraphic affinities were not clear when the taxon was first described in 1890 (Miller
and Gurley, 1890). A. aliquantila is known from the Gilmore City Formation of Iowa, USA
(Kier, 1958), which is also Kinderhookian in age (Lower Tournaisian). Miller and Gurley’s
holotype of A. legrandensis does not show any striking difference between the A. aliquantila of
Kier (1958). Though the plate sculpturing of the type of A. aliquantula was described in detail
(Kier, 1958) Miller and Gurley’s holotype of A. legrandensis does not display well-preserved
plate morphology (Kier, 1958). Both specimens display small tests, with that of A aliquantila
having a diameter of about 13 mm (USNM 136451) and that of Miller and Gurley’s type for A.
legrandensis being about 15 mm in diameter (measured from Miller and Gurley, 1890). The
spines in both specimens are also smooth. Given the similarity of these two taxa
morphologically, and their similar stratigraphic ranges, they are best synonymized. Jackson
233
(1912) synonymized A. legrandensis with A. blairi from the Viséan Warsaw Formation,
however, Kier (1958) later separated the two taxa, based primarily upon the poor preservation of
the plate sculpturing of Miller and Gurley’s A. legrandensis holotype, and the differences in the
stratigraphic occurrences of the taxa (Lower Tournaisian of A. legrandensis and Viséan of A.
blairi). There are, however, differences between the sizes of these taxa, which are important in
differentiating species. Kier (1958) published on a new specimen of A. blairi from the same
locality as Miller’s (1891) holotype. This taxon was 23.56 mm in diameter and Miller’s (1891)
cotype is about 26 mm in diameter (measured from Jackson, 1912). These specimens are
considerably larger than the A. legrandensis of Miller and Gurley (1890) and thus the
designation of separate A. legrandensis and A. blairi is supported. Under this current revision
then, Jackson’s (1912) A. legrandensis constitutes two different taxa. His figures 7 and 8 on plate
8 and 12 and 13 on plate 9 are referred to A. blairi while his figures 3 and 4 on plate 10, one of
which (Fig. 4) depicts Miller and Gurley’s holotype, are referred to A. legrandensis.
Additionally, Jackson’s specimens 3198 (now MCZ Cat. 101899), 3199 (now MCZ Cat.
101900), and 2300 (now MCZ Cat. 101901) from the Keokuk Group of Indiana are referred to A.
blairi. The phylogenetic analysis herein thus considers two species, A. legrandensis and A. blairi.
Archaeocidaris ausichi n. sp.
Fig. 6.7A-D.
Diagnosis.⎯Small to medium sized Archaeocidaris with spines bearing very small spinules.
Basal terrace present on interambulacral plates. Mamelons perforate and non-crenulate. Only a
single row of scrobicular tubercles at edge of interambulacral plates.
234
Derivation of name.⎯Named for William Irl Ausich, expert on Paleozoic crinoids, and
undergraduate advisor of the lead author.
Material.⎯LACIMP 14515 is the holotype and LACIMP 14516 is the paratype. Both are on the
same slab.
Occurrence.⎯Mississippian, Meramecian (Viséan) St. Louis Limestone, near St. Louis,
Missouri, USA.
Description.⎯ Test, regular, small to medium sized, about 33 mm wide on largest individual.
Test plating is imbricate Details of the ambulacra, peristome, perignathic girdle, lantern and
apical system are not known. Four columns of interambulacral plates are present in each
interambulacral area. Median interambulacral plates are about 1.3 times wider than high and
hexagonal and adambulacral plates are pentagonal. Primary tubercles are perforate and non-
crenulate. The boss or basal terrace takes up about 1.7/4.3 of the width of the plate. The primary
tubercles are not sunken and a sunken ring on the parapet surrounds the mamelon. A single row
of scrobicular tubercles is present adjacent to plate edge and radial plications are absent. These
scrobicular tubercles are imperforate and non-crenulate.
Primary spines are shorter than full diameter of the test, though longer than half the
diameter of the test. Longest spine 16.8 mm in length. Spines bear numerous very small spinules
along the whole length of the shaft. A prominent milled ring is present, which is the widest part
of the spines. Primary spines are hollow and circular in cross section without any obvious
bilateral symmetry. Acetabulum is non-crenualte. Secondary spines are small, smooth, striated
and without ornamentation.
235
Discussion.⎯ The exact locality details of these specimens are unclear except for that they come
from near St. Louis and were collected from the St. Louis Limestone. This species is similar to
A. wortheni, A. legrandensis, and A. blairi in that its interambulacral plates bear only a single
ring of scrobicular tubercles. It differs from those species primarily in having small, slight
spinules adorning its primary spines. Furthermore, this species is known to occur on the same
bedding planes as specimens of that A. wortheni, as is the case with the holotype.
Archaeocidaris apheles Schneider et al. 2005
Fig. 6.7E.
2005 Archaeocidaris apheles Schneider et al., p. 753, Figs. 9.1-9.4
Types. ⎯The holotype is TMM 1967TX20, and paratypes are TMM 1967TX17, 1967TX18, and
1967TX55.
Diagnosis.⎯The diagnosis of Schneider et al. (2005) stands, however, following examination of
the type material we felt it necessary to note that the primary interambulacral tubercles of this
species are crenulate.
Occurrence.⎯Pennsylvanian Winchell Formation (Kasimovian) of Brown County, Texas, USA.
Description.⎯Schneider et al.’s (2005) description is very good, however some clarification on
one morphological feature is needed. Schneider et al. (2005) note that the bosses of the tubercles
of these specimens are striated. It should be noted that these striations appear to be crenulations,
and that A. apheles bears crenulate tubercles (Fig. 6.7E).
236
Discussion.⎯The crenulate tubercles of A. apheles are the only crenulate tubercles known from
a species of Archaeocidaris. Crenulate tubercles are also known from archaeocidarids from the
Permian, though the taxonomic affinities of these species are not well known (Thompson et al.,
2017b). Because these Permian taxa are known from only disarticulated interambulacral plates,
whether or not their crenulate tubercles are homologous with those of of A apheles and those of
the echinoid crown group remains unknown. We hope that with the discovery of further Permian
archaeocidarids this hypothesis can be tested.
Archaeocidaris brownwoodensis Schneider et al. 2005
Fig. 6.7F.
2005 Archaeocidaris brownwoodensis Schneider et al., p. 747-753, Figs. 1-8
Types. ⎯The holotype is TMM 1967TX60, and numerous paratypes are listed in Schneider et al.
(2005).
Diagnosis.⎯The diagnosis of Schneider et al. (2005) is sound, however, we herein add a few
notes to their diagnosis. Some larger adapical interambulacral plates on this species bear multiple
rows of scrobicular tubercles. Additionally, the most adapical interambulacral plates of the
corona are small and imbricate, without primary tubercles. These adapical plates appear to be
arranged into six or seven columns.
Occurrence.⎯ Pennsylvanian Winchell Formation (Kasimovian) of Brown County, Texas,
USA.
Description.⎯The description put forth by Schneider et al. (2005) is very accurate, and quite
detailed. However, further examination of the type material has allowed for the inclusion of
237
additional data. Some of the more adapical tuberculate interambulacral plates contain multiple
rows of scrobicular tubercles, as opposed to just a single row. This is particularly present on
adambulacral plates on the most adambulacral edge of the plates.
In A. brownwoodensis, the most adapical interambulacral plates of the corona (at least the
most six or seven adapical rows) also lack primary tubercles. In the most adapical portion of the
interambulacral area, the plates are arranged into six or seven columns of interambulacral plates
as opposed to the four columns present in the non-adapical rows. It should also be clarified that
the ambulacral plates, although arranged into two columns in each ambulacral area have uniserial
pore pairs.
Discussion.⎯ This presence of multiple rows of scrobicular tubercles on A. brownwoodensis is
similar to the condition present on other species of Archaeocidaris, such as A. rossica, A
marmorcataractensis, A. whatleyensis, and A. subwortheni. This may explain why in some of our
phylogenetic analyses A. rossica, A marmorcataractensis, A. brownwoodensis, and A.
marmorcataractensis were resolved as a monophyletic group.
The presence of greater than four columns of nontuberculate, imbricate interambulacral plates in
A. brownwoodensis is also interesting to note. This arrangement is similar to what is found in A.
immanis (Kier, 1958; Fig. 1), and A. hemispinifera (Chesnut and Ettensohn, 1988; Text-Fig. 28)
which have small, imbricate non-tuberculate interambulacral plates in their most adapical
interambulacra. In all three species these plates are arranged into five to seven columns. In A.
hemispinifera this area of non-tuberculate plates is enlarged, taking up about half of the corona,
while it is much smaller in A. immanis and A. brownwoodensis. In A. blairi, the most adapical
interambulacral plates are also small, imbricate and nontuberculate (Kier, 1958; Fig. 3),
238
however, the number of columns these plates are arranged into are unknown. Polytaxicidaris
lirata also has these small, imbricate, nontuberculate interambulacral plates (Kier, 1956; Plate
56, Fig. 2). It may be that the condition of having greater than four columns of small, imbricate,
nontuberculate interambulacral plates is the condition in all species of Archaeocidaris and
perhaps also Polytaxicidaris, however the state of the archaeocidarid fossil record is such that the
apical and adapical portions of tests are rarely preserved, and are thus unknown in most species.
Subclass ?Cidaroidea Smith, 1984
Family ?Miocidaridae, Durham and Melville, 1957
Type genus. ⎯Miocidaris Doderlein, 1887
Other genera. ⎯Eotiaris Lambert, 1900, Couvelardicidaris Vadet, 1991, Procidaris Pomel,
1883
Genus ?Eotiaris Lambert, 1900
Type species. ⎯ Cidaris keyserlingi Geinitz, 1848, from the Wuchiapingian Zechstein of
Germany and England.
?Eotiaris meurevillensis (Dehée, 1927)
1927 Archaeocidaris meuervillensis Dehée, p. 290-293, Pl. 7, figs. 1-10.
non
Diagnosis.⎯?Eotiaris with long thin, coarsely striate spines.
Occurrence.⎯Namurian or Westphalian (Pennsylvanian) near Merville, Nord, France.
239
Description. ⎯The description of Dehée is amended herein. Test small, regular, apical system
and peristome unknown. Lantern and perignathic girdle also unknown. Coronal plates rigidly
sutured at and below ambitus, with most adapical plates likely imbricate as they are not
preserved. Six rigidly sutured plates preserved on known specimens.
Ambulacra arranged into columns of two rows. Ambulacra straight, presumably beveling
under interambulacral. Pore pairs arranged slightly obliquely on plate. Further details of
ambulacra difficult to ascertain through photos. Interambulacra arranged in two columns.
Interambulacral plates pentagonal, between 1.5 and 2 times wider than high. Single primary
tubercle on each interambulacral plate. Scrobicular tubercles not particularly densely arranged,
and not directly adjacent to plate edge. At least two rows of scrobicular tubercles at interradial
edges of plates. Single row on adoral and adapical edges of plates. Aoreoles confluent on most
three or four adoral interambulacral plates. Primary tubercle perforate, crenulate. Mammelon not
undercut.
Spines long, slender, circular in cross-section. Spines are coarsely striate and the shaft is
about twice as wide as the base. Shaft of uniform width for entire length. Base tapering distally
from its widest to the most proximal shaft. Acetabulum bearing crenulations.
Discussion.⎯This is a potentially a very important species because it is the oldest known
putative miocidarid, and thus crown-group echinoid, in the fossil record. Unfortunately, the type
material and only known specimen is lost (J. Cuvelier, pers. Comm.). Furthermore, because this
specimen is lacking details of the perignathic girdle and lantern, it’s precise taxonomic affinities
remain unknown. Furthermore, the precise stratigraphic occurrence of these taxa is also poorly
known. The specimens were collected from 247 M below ground near Merville in the Nord-Pas-
240
de-Callais Province of France from a layer of bituminous black shale high in volatiles (Dehée
1927).
The associated mollusc fauna was used to correlate the age of the rocks to those of the
Westphalian Chokier Formation, in Belgium. Because of the depth at which the specimens of
?Eotiaris meurevillensis were collected, they are most likely from Namurian or Westphalian
strata, however, the specific horizons are unknown. Furthermore, because the specimens were
collected at great depth, and they and their associated matrix are lost, attempts to further
corroborate their age are not currently feasible. Thus, despite the potential importance of these
specimens, we caution placing too much import on their stratigraphic distribution, and because
the precise stratigraphic details are unknown, their age should be regarded as broadly Namurian-
Westphalian.
This species differs from the two known Permian species of Eotiaris, E. guadalupensis
and E. keyserlingi in having less densely spaced scrobicular tubercles. These taxa also have more
rows of scrobicular tubercles on the adambulacral and interradial edges of interambulacral plates.
This species is similar to E. connorsi in having less dense scrobicular tuberculation. Furthermore
the spines of ?E. meurevillensis are much thinner than the more robust spines of E.
guadalupensis and E. keyserlingi, being more similar to those of E. connorsi. E. connorsi differs
from this taxon, however, in having more flexible, and thinner coronal plating.
241
Figure 6.1. Images of representative archaeocidarid and miocidarid species and genera included
in analyses herein. A. Specimen USNM S3828 of Archaeocidaris blairi (Miller, 1891). Scale bar
is 5 mm. B. Specimen MCZ 101862 of Lepidocidaris squamosa (Meek and Worthen, 1869).
Scale bar is 10 mm. C. Specimen USNM 136453 of Archaeocidaris legrandensis Miller and
Gurley, 1980. Scale bar is 10 mm. D. Specimen USNM 158415 of Nortonechinus sp. Scale bar
is 5 mm. E. Specimen USNM 610601 of Eotiaris guadalpuensis Thompson, 2017. Scale bar is
2.5 mm. F. Specimen of USNM 144197 of Polytaxicidaris lirata Kier, 1965. Scale bar is 10 mm.
242
A
D C
B
F E
243
Figure 6.2. Phylogenetic trees resulting from (A) unweighted parsimony analysis and (B)
Bayesian analysis of archaeocidarid and Paleozoic miocidarid species. A. Strict concensus of 817
most parsimonious trees resulting from unweighted parsimony analysis of archaeocidarid and
miocidarid taxa rooted on Nortonechinus welleri. Bootstrap proportions are shown at the nodes
in regular font and Bremer support in Italics. Bremer support was only calculated for two steps
longer than the most parsimonious trees due to computational constraints. B. 50% majority rule
tree resulting from Bayesian analyses of archaeocidarid and miocidarid species with the
symmetric Dirichlet hyperprior, α=∞. Node support (PP) is equal to the proportion of trees in the
posterior distribution that contained the given node.
244
A. immanis
A. whatleyensis
A. ivanovi
A. wortheni
N. welleri
A. blairi
E. connorsi
E. meurevillensis
A. mosquensis
P. lirata
A. agassizi
E. guadalupensis
A. apheles
A. rossica
P. dyeri
D. primaevus
A. hemispinifera
A. legrandensis
A. subwortheni
E. keyserlingi
A. ausichi
A. brownwoodensis
L. squamosis
A. marmorcataractensis
81, ≥2
43, 0
55, 1
0.2
A. hemispinifera
E. keyserlingi
P. dyeri
E. guadalupensis
A. apheles
E. meurevillensis
A. ivanovi
P. lirata
E. connorsi
A. brownwoodensis
A. ausichi
L. squamosis
A. mosquensis
N. welleri
A. agassizi
A. blairi
D. primaevus
A. marmorcataractensis
A. rossica
A. legrandensis
A. subwortheni
A. wortheni
A. whatleyensis
A. immanis
0.5403
0.6449
0.603
0.6049
A B
245
Figure 6.3. Strict consensus of 102 most parsimonious trees resulting from parsimony analysis of
archaeocidarid and Paleozoic miocidarid species following reweighting of characters by their
retention indices following unweighted analysis and rooted on Nortonechinus welleri. Node
support shown are bootstrap proportions with values less than 5 not shown.
246
247
Figure 6.4. Histograms comparing the SCI and GER scores from the 102 most parsimonious
trees from the reweighted parsimony analysis (A, C) to 1000 randomly generated topologies
assigned to the same taxon and age data (C, D). A. Histogram showing frequency distribution of
SCI scores for 102 MPTs. Note that all scores are above the critical value. B. Frequency
histogram of SCI scores for 1000 randomly generated topologies. Note that the majority of
randomly generated topologies do not differ significantly from random.
C. Histogram showing frequency distribution of GER scores for 102 MPTs. Note that all scores
are above the critical value. D. Frequency histogram of GER scores for 1000 randomly generated
topologies. The majority of randomly generated topologies do not differ significantly from
random. In all plots, dotted lines represent the critical value above which SCI or GER values
differ significantly from random.
248
Input Trees
SCI
Frequency
0.0 0.2 0.4 0.6 0.8 1.0
0 5 10 15 20 25 30
Randomly Generated Trees
SCI
Frequency
0.0 0.2 0.4 0.6 0.8 1.0
0 50 100 150 200 250
Input Trees
GER
Frequency
0.0 0.2 0.4 0.6 0.8 1.0
0 10 20 30 40 50
Randomly Generated Trees
GER
Frequency
0.0 0.2 0.4 0.6 0.8 1.0
0 50 100 150
A
D C
B
249
Figure 6.5. Fit of the optimal most parsimonious trees from the reweighted parsimony analysis to
the stratigraphic record. A. The single most parsimonious tree with the highest SCI score plotted
against stratigraphy. B. A strict consensus of the seven most parsimonious trees with the highest
GER scores plotted against stratigraphy.
250
250
260
270
280
290
300
310
320
330
340
350
360
370
380
390
Eifelian
Givetian
Frasnian
Famennian
Tournaisian
Visean
Serpuk.
Bashkirian
Moscovian
Kasimovian
Gzhelian
Asselian
Sakmarian
Artinskian
Kungurian
Roadian
Wordian
Capitanian
Wuchiapingian
Chang.
Induan
Olenekian
Anisian
Devonian Carboniferous Permian Triassic
N. welleri
A. marmorcataractensis
A. rossica
A. brownwoodensis
A. subwortheni
E. connorsi
E. keyserlingi
E. guadalupensis
E. meurevillensis
A. whatleyensis
A. ivanovi
A. mosquensis
A. agassizi
A. ausichi
A. wortheni
A. immanis
L. squamosis
A. apheles
A. blairi
A. legrandensis
A. hemispinifera
P. dyeri
P. lirata
D. primaevus
250
260
270
280
290
300
310
320
330
340
350
360
370
380
390
Eifelian
Givetian
Frasnian
Famennian
Tournaisian
Visean
Serpukhovian
Bashkirian
Moscovian
Kasimovian
Gzhelian
Asselian
Sakmarian
Artinskian
Kungurian
Roadian
Wordian
Capitanian
Wuchiapingian
Changhsingian
Induan
Olenekian
Anisian
Devonian Carboniferous Permian Triassic
N. welleri
A. marmorcataractensis
A. rossica
A. brownwoodensis
A. subwortheni
E. connorsi
E. keyserlingi
E. guadalupensis
E. meurevillensis
A. whatleyensis
A. ivanovi
A. mosquensis
A. agassizi
A. ausichi
A. wortheni
A. immanis
L. squamosis
A. apheles
A. blairi
A. legrandensis
A. hemispinifera
P. dyeri
P. lirata
D. primaevus
A
B
251
Figure 6.6. Photographs of specimens of Devonocidaris primaevus Belanski, 1927. A. Test of
specimen SUI 53171B of D. primaevus. Note the eight columns of interambulacral plates in each
interambulacral area, perforate noncrenulate primary tubercles and perforate secondary tubercles
on interambulacral plates. B. Specimen 53171A, a smaller specimen of D. primaevus occurring
on the same slab as SUI 53171B. C. Close up of the interambulacrum of specimen SUI 53171B,
note the shape of interambulacral plates and the faint tubercle bosses. D. Close up of the
ambulacrum of specimen SUI 53171B. Ambulacra are arranged into two columns and bear fairly
large, perforate tubercles. E. Specimen SUI 40246 of D. primaevus. Note sparse secondary
tubercles on ambulacral plates and lack of milled rings on spines. F. Other side of Specimen SUI
40246 showing details of lantern, perforate ambulacral tubercles, and long, thin spines. Scale bar
in all photos is 10 mm.
252
A B
C D
E F
253
Figure 6.7. Photos of Archaeocidaris ausichi n. sp., Archaeocidaris apheles Schneider, Sprinkle,
Ryder, 2005, and Archaeocidaris brownwoodensis Schneider, Sprinkle, Ryder, 2005. A.
Specimen LACIMP 14515 the holotype of A. ausichi. Scale bar is 5 mm.
B. Close up of specimen LACIMP 14515 showing the details of the interambulacral plating and
the spine base. Scale is 5 mm. C. Specimens LACIMP 14515 and 14516 of A. ausichi in addition
to a juvenile specimen of A. wortheni. Scale bar is 10 mm. D. Close up of the spines of
specimen LACIMP 14516, paratype of A. ausichi, showing the ribbed spinules present on the
primary spine shaft. Scale bar is 5 mm. E. Close up of the interambulacral plates of specimen
TMM 1967TX18, paratype of A. apheles showing the crenulate primary tubercles. Scale bar is
1.25 mm. F. Specimen TMM 1967TB60, holotype of of A. brownwoodensis showing the
multiple rings of scrobicular tubercles on interambulacral plates and the six columns of non-
tuberculate interambulacral plates in the most adapical portion of an interambulacrum. Scale bar
is 10 mm.
254
B
E F
A B
C D
255
Figure 6.8. 50% majority rule trees for Bayesian analysis of evolutionary relationships of
archaeocidarid and miocidarid species. All analyses were run using the MK model (Lewis, 2001)
with priors of symmetric discrete beta distributions with six different values of the shape
parameter α (Wright et al. 2016). (A) 50% majority rule consensus tree when α was set to ∞. (B)
50% majority rule consensus tree when α was set to 10. (C) 50% majority rule consensus tree
when α was set to 2. (D) 50% majority rule consensus tree when α was set to 1. (E) 50%
majority rule consensus tree when α was set to 0.2. (F) 50% majority rule consensus tree when α
was set to 0.05. Clade credibility values for resolved nodes (Posterior probabilities) shown in
larger font while branch lengths are smaller.
256
0.2
A. hemispinifera
E. keyserlingi
P. dyeri
E. guadalupensis
A. apheles
E. meurevillensis
A. ivanovi
P. lirata
E. connorsi
A. brownwoodensis
A. ausichi
L. squamosis
A. mosquensis
N. welleri
A. agassizi
A. blairi
D. primaevus
A. marmorcataractensis
A. rossica
A. legrandensis
A. subwortheni
A. wortheni
A. whatleyensis
A. immanis
0.5403
0.6449
0.603
0.6049
0.0897
0.1024
0.1711
0.0846
0.0301
0.1617
0.0583
0.0868
0.0603
0.7361
1.3136
0.1355
0.2465
0.0764
0.1495
0.1433
0.1748
0.5963
0.0673
0.1787
0.0599
0.3565
0.097
0.3276
0.0682
0.0569
0.3439
0.2164
INF
0.2
P. dyeri
A. hemispinifera
A. blairi
A. ivanovi
E. keyserlingi
A. mosquensis
A. apheles
A. whatleyensis
E. guadalupensis
A. immanis
E. meurevillensis
N. welleri
D. primaevus
P. lirata
A. rossica
E. connorsi
A. ausichi
A. subwortheni
A. wortheni
A. legrandensis
A. brownwoodensis
L. squamosis
A. marmorcataractensis
A. agassizi
0.5274
0.5766
0.1676
0.1073
0.4398
0.4278
0.0725
0.0739
0.0589
0.1843
0.1055
0.1991
0.0649
0.034
0.1073
0.075
0.1039
0.1788
0.2832
0.1879
1.269
0.1476
0.0979
0.2692
0.0627
0.0913
0.9249
0.2999
10
0.2
P. dyeri
A. ausichi
A. wortheni
A. agassizi
L. squamosis
D. primaevus
A. hemispinifera
E. keyserlingi
A. marmorcataractensis
E. guadalupensis
A. apheles
A. mosquensis
A. whatleyensis
A. legrandensis
P. lirata
A. subwortheni
A. immanis
A. ivanovi
A. blairi
E. connorsi
N. welleri
E. meurevillensis
A. rossica
A. brownwoodensis
0.5199
0.3113
0.2909
0.3607
0.9732
0.4955
0.1574
0.3266
0.1632
0.7411
0.1509
0.1829
0.0811
0.6593
0.1265
0.4735
1.0021
0.268
0.0601
0.3123
0.2252
0.1204
0.141
0.3795
0.1044
0.1691
2
0.09
A. legrandensis
A. apheles
A. immanis
A. whatleyensis
A. marmorcataractensis
A. brownwoodensis
A. blairi
E. keyserlingi
N. welleri
P. dyeri
P. lirata
L. squamosis
E. guadalupensis
D. primaevus
A. ivanovi
A. agassizi
E. connorsi
A. rossica
A. ausichi
A. subwortheni
A. wortheni
E.meurevillensis
A. hemispinifera
A. mosquensis
0.5049
0.5432
0.7915
0.2981
0.1226
0.4256
0.1565
0.1886
0.5089
0.1093
0.2849
0.3583
0.3428
0.3557
0.7367
0.1818
0.0639
0.7487
0.2781
0.8955
0.3806
0.2424
0.5239
0.5114
0.3899
0.5294
1
0.2
E. keyserlingi
A. wortheni
L. squamosis
P. dyeri
A. subwortheni
E. guadalupensis
A. apheles
A. immanis
A. ivanovi
A. marmorcataractensis
A. mosquensis
E. meurevillensis
N. welleri
D. primaevus
A. whatleyensis
A. ausichi
A. legrandensis
A. rossica
E. connorsi
P. lirata
A. hemispinifera
A. blairi
A. agassizi
A. brownwoodensis
0.5733
0.1284
0.1481
0.4458
0.2403
0.2337
0.3439
0.2646
0.4478
0.0752
0.3686
0.7837
0.134
0.1658
0.7066
0.2707
0.3598
0.1817
0.8132
0.1017
0.233
0.1329
0.5893
0.1213
0.2563
0.5715
0.2
0.2
N._welleri
E. guadalupensis
A. marmorcataractensis
A. agassizi
A. legrandensis
A. whatleyensis
E. keyserlingi
A. blairi
A. hemispinifera
A. mosquensis
D._primaevus
A. wortheni
A. ausichi
A. immanis
L. squamosis
P. lirata
E. connorsi
A. subwortheni
E. meurevillensis
A. brownwoodensis
P. dyeri
A. rossica
A. ivanovi
A. apheles
0.7873
0.6215
0.5017
0.5609
0.1304
0.5978
0.1026
0.1632
0.1545
0.068
0.0644
0.1146
0.0756
0.0789
0.2908
0.0564
0.7938
0.1707
0.0555
0.1156
0.0349
0.1074
0.6899
0.4352
0.0779
0.7218
0.2601
0.1916
0.1097
0.0536
0.0992
0.1684
A. B.
C. D.
E. F.
257
Table 6.1. List of taxa and source of information for all taxa included in phylogenetic analyses
herein.
258
Species Author Source of Data Range
Nortonechinus welleri Thomas, 1921 Thomas Frasnian
Devonocidaris primaevus Belanski, 1928 Specimens Frasnian
Lepidocidaris squamosa (Meek & Worthen, 1869) Specimens and Jackson 1912 Tournaisian-Viséan
Polytaxicidaris dyeri Kier, 1958 Specimens Viséan
Polytaxicidaris lirata Kier, 1965 Specimens Serpukhovian
Archaeocidaris apheles Schneider et al. 2005 Specimens Kasimovian
Archaeocidaris brownwoodensis Schneider et al. 2005 Specimens Kasimovian
Archaeocidaris marmorcataractensis Thompson et al. 2015 Specimens Bashkirian
Archaeocidaris legrandensis Miller and Gurley, 1890 Specimens Tournaisian
Archaeocidaris wortheni Hall, 1858 Specimens Viséan
Archaeocidaris blairi (Miller, 1891) Specimens Viséan
Archaeocidaris immanis Kier, 1958 Specimens Kasimovian
Archaeocidaris whatleyensis Lewis & Ensom, 1982 Specimens Viséan
Archaeocidaris rossica (von Buch, 1842) Specimens and Jackson 1912 Moscovian-Gzhelian
Archaeocidaris mosquensis Ivanov, in litt, Fass, 1939 Specimens Moscovian-Kasimovian
Archaeocidaris subwortheni Faas, in litt, Fass, 1939 Specimens Kasimovian
Archaeocidaris agassizi Hall, 1858 Specimens Tournaisian
Archaeocidaris ausichi n. sp. Specimens Viséan
Archaeocidaris hemispinifera Chesnut & Ettensohn, 1988 Chesnut and Ettensohn, 1988 Serpukhovian
Eotiaris connorsi (Kier, 1965) Specimens Wordian
Eotiaris keyserlingi (Geinitz, 1848) Specimens and Smith and Hollingworth, 1990 Wuchiapingian
Eotiaris meurevillensis (Dehée, 1927) Dehée, 1927 Bashkirian-Kasimovian*
Eotiaris guadalupensis Thompson 2017 Specimens Roadian-Capitanian
Archaeocidaris ivanovi Thompson and Mirantsev, n. sp. Specimens Moscovian-Kasimovian
259
Chapter 7. A new stem group echinoid from the Triassic of
China leads to a revised macroevolutionary history of echinoids
during the end-Permian mass extinction
Originally published as:
Thompson, J. R., Hu, S.-x., Zhang, Q.-y., Petsios, E., Cotton, L. J., Huang, J.-y., Zhou, C.-y.,
Wen, W., and Bottjer, D. J., 2018, A new stem group echinoid from the Triassic of China
leads to a revised macroevolutionary history of echinoids during the end-Permian mass
extinction: Royal Society Open Science, v. 5, p. 171548.
INTRODUCTION
The effect of the end-Permian mass extinction on the macroevolutionary history of
echinoids has become a classic example of the extinction event’s devastating influence on the
macroevolutionary history of metazoans (Benton, 2003; Erwin, 1993, 1994; Twitchett and Oji,
2005). The long-accepted model proposed that only a single lineage, that of the genus Eotiaris
(previously called Miocidaris), belonging to the crown group echinoid family Miocidaridae was
the only lineage of echinoid to survive the extinction event, and all post-Paleozoic echinoids
could, thus, trace their origin back to Eotiaris (Erwin, 1994; Kier, 1977b; Smith, 1984). New
fossil finds (Thuy et al., 2017) are beginning to shift this paradigm, as echinoid taxa once
thought exclusive to the Paleozoic have been found in Triassic strata. Paleozoic echinoids, which
make up the majority of the clade’s stem group, have tests composed of multiple columns of
260
interambulacral and ambulacral plates which articulate flexibly and disarticulated rapidly
following death (Kier, 1965; Thompson and Ausich, 2016). This is in stark contrast to the
echinoid crown group, which has a test structure consisting of only two columns of ambulacral
plates, and two columns of interambulacral plates, and in many taxa displays test plating with
interlocking stereom (Smith, 1984; Thompson et al., 2015b). Further phylogenetic analyses and
analyses of ghost lineages have shown that, although the miocidarids were the only echinoids
with fossil representation in both the Paleozoic and the Mesozoic, one or two other lineages of
crown group echinoids probably also crossed the Permian-Triassic boundary (Kroh and Smith,
2010; Smith and Hollingworth, 1990; Smith, 2007). Nevertheless, it was accepted that the
Permian-Triassic extinction spelled the end for the echinoid stem group, which presumably never
survived into the Mesozoic.
The fossil record of echinoids in the Early Triassic is, however, notoriously poor (Smith,
2007), and only three published localities have produced articulated specimens globally
(Godbold et al., 2017; Kier, 1968b, 1977b; Linck, 1955). Stem group echinoids putatively
assigned to the family Proterocidaridae have also recently been recovered from Middle and
Upper Triassic strata, though not without controversy (Salamon and Gorzelak, 2017), and
indicate that stem group echinoids did, in fact survive the Late Permian mass extinction.
Additional fieldwork in the Middle Triassic of Yunnan Province, southwestern China (Fig.
7.1A), further indicates that this poor fossil record may have obscured the true
macroevolutionary history of echinoids throughout the Early Triassic. New fossil specimens
from the Anisian Luoping Biota (Hu et al., 2010) indicate that stem group echinoids were widely
distributed in the Triassic. Phylogenetic analyses of Paleozoic and early Mesozoic echinoids,
including this new species Yunnanechinus luopingensis n. sp. (Fig. 7.1B-G), indicate that it is a
261
stem group echinoid and re-affirm that multiple lineages of echinoids crossed the Permian-
Triassic boundary. Thus, multiple Paleozoic echinoid clades survived the biotic and
environmental turmoil characterizing the latest Permian and Early Triassic.
METHODS
In order to determine whether or not Yunnanechinus represented a stem group echinoid which
survived the end-Permian mass extinction event, we used parsimony-based and Bayesian
phylogenetic analyses to determine its phylogenetic placement with respect to thirteen stem and
crown group echinoids spanning the Tournaisian (Early Mississippian) to Rhaetian (Late
Triassic). The Silurian echinoid Echinocystites ponum was used as an outgroup. Our character
matrix consisted of sixty-nine characters coded for fourteen taxa. Fourteen characters were
multistate, while fifty-five were binary. All but one character were unordered. Parsimony
analyses were run in PAUP* (Swofford, 2003) using a heuristic search with 1000 random
addition sequence replicates using starting trees obtained using stepwise addition and branch
swapping through tree bisection and reconnection. Our Bayesian analyses were run in MrBayes
3.2 (Ronquist et al., 2012) using the Mk model of character change (Lewis, 2001), which, in a
Bayesian framework, has recently been demonstrated to estimate phylogenetic relationships
more accurately than parsimony based methods (Puttick et al., 2017; Wright and Hillis, 2014). In
order to account for ascertainment bias, as there were no constant characters included in the
dataset, character coding was set to variable. We used a gamma distribution with a prior of
exponential(1.0) to model rate variation and the prior on unconstrained branch lengths was also
exponential(1.0). We used a symmetric Dirichlet prior on character state frequencies with
parameter α=∞. The joint posterior distribution of trees, branch lengths, and parameters was
262
estimated using Markov Chain Monte Carlo (MCMC). Details of MCMC are shown in
Electronic Supplementary Material. We additionally utilized sensitivity analyses to determine the
robustness of our results to outgroup choice and model parameters. Further details of character
scoring and sensitivity analyses can be found in Thompson et al. (2018) Electronic
Supplementary Material. The institutional abbreviation LPI is the Invertebrate Paleontology
Collection, Chengdu Institute of Geology and Mineral Resources, Chengdu, China.
SYSTEMATIC PALEONTOLOGY
Echinoidea Leske, 1778
Stem group Echinoidea
Incertae familiae
Genus Yunnanechinus n. gen.
Etymology. Named for Yunnan, China from whence the type species is known.
Diagnosis. As for species
Type species. Yunnanechinus luopingensis n. sp.
Occurrence. As for species.
Yunnanechinus luopingensis n. sp.
urn:lsid:zoobank.org:act:19902A64-E79D-4E2C-ADB0-329CD565F2CB
Etymology. Named for the Luoping Biota, the fossil Lagerstätte from which the species is
described.
Diagnosis. Test with imbricate plating, at least adapically and ambitally. Genital plates
with one gonopore per plate (Fig. 7.1F). Plates of apical system covered with small, imperforate
263
non-crenulate tubercles. Interambulacral plates polygonal to subpentagonal in shape.
Interambulacral plates with a single imperforate non-crenulate tubercle, and sparse imperforate
non-crenulate secondary tubercles. Spines less than half the diameter of the test in length, finely
striate and without a milled ring (Fig. 7.1 E, G).
Material. The holotype is specimen LPI-32321, paratypes are specimens LPI-2638, LPI-
61163, and LPI-61701A,B.
Occurrence. All specimens from the Middle Anisian (Pelsonian) Guanling Formation of
the Luoping Biota of Yunnan Province, South China.
Description. Test small to very small. Specimen 32321 6.48 mm in diameter, specimen
32220 15.99 mm in diameter, specimen 32638 is 13.31 mm in diameter, specimen LPI-61163 is
22.44 mm in diameter, specimen LPI-61701B is 21.31 mm in diameter. Test plating is imbricate,
at least above the ambitus.
Peristomial plating and lantern unknown, but all specimens show a small bump located in
the centre of the test, which probably represents the lantern inside the collapsed test. The apical
disc is well preserved on specimen 32321 (Thompson et al. 2018, Fig. S4A, S4B). The
madreporite is clearly visible atop interambulacrum number 2, while gonopores are visible on
other genital plates. Each visible genital plate bears a single large gonopore. These gonopores are
centrally located, except above interambulacrum 2, where the gonopore appears to be located
adambulacrally to the madreporite (Thompson et al. 2018, Fig. S4A, S4B). At least one ocular
plate appears to be present and appears to be in contact with the periproctal plates, indicating that
at least some of the ocular plates are insert. These plates appear to be covered with numerous
small secondary, imperforate and non-crenulate tubercles, which appear to have borne small,
striate spines, morphologically similar to those on the rest of the test.
264
The interambulacral plating is largely obscured throughout the tests of all specimens
except for specimen 32321, which displays interambulacral plating adapically and in a few
places at and slightly above the ambitus (Thompson et al. 2018, Fig. S4A, S4B). The number of
interambulacral plates in each interambulacral area is thus unknown. The shape of the plates is
obscured near the ambitus, though adapically the plates are polygonal to subpentagonal in shape
(Thompson et al. 2018, Fig. S4C). These plates imbricate adapically, with more adoral plates
laying overtop of those more adapical plates. None of the plates appear to show any scrobicular
ring or well-defined scrobicule. A single non-crenulate and imperforate primary tubercle is
located subcentrally on the more adambital plates, while smaller primary tubercles are present on
the more adapical plates (Thompson et al. 2018, Fig. S4C). In a few places on the test, these
larger tubercles are associated with just barely disarticulated spines, indicating that there does
appear to be size differentiation amongst tubercles and spines located more adambitally. On one
plate, the presence of a smaller, imperforate and non-crenulate tubercle associated with a small
spine confirms that these plates bear smaller secondary tubercles, though the exact number is
unknown. A single ambulacral plate is present on specimen 32321 (Fig. 7.1E). The pore pair is
oriented obliquely on the plate, though it is not possible to tell if the pore pair is angled
adambulacrally or perradially. A small interporal partition separates the two pores of the plate.
Above the pore pair is a single small imperforate and non-crenulate tubercle.
Small spines cover the tests of all specimens (Thompson et al. 2018, Fig. S4D-S4F) and,
with the exception of specimen 32321, are all that is visible on the specimens. These spines are
less than half the diameter of the test and taper distally. There are both primary and secondary
spines, associated with primary and secondary tubercles, respectively, and both are identical in
morphology except for their size. There is no milled ring and the spines bear fine striations.
265
Proximally, all spines end in a small swelling, which is the base. The acetabulum is non-
crenulate.
RESULTS
Bayesian and Parsimony-based phylogenetic analyses are shown in Figure 7.2, while
results of sensitivity analyses are in Thompson et al. (2018) Electronic Supplementary Material
Figures S1-S3. Parsimony analyses resolved a single most parsimonious tree (MPT) with length
121 steps, consistency index (CI)=0.69, and retention index (RI)=0.74. Yunnanechinus plotted as
the most basal taxon in the ingroup, while two clades were more derived, one composed of
proterocidarid, palaechinid and lepidocentrid echinoids, and one composed of Archaeocidaris
and the echinoid crown group (Fig. 7.2). We resolved the members of the echinoid crown group
with fairly high bootstrap support (73) and the clade of Archaeocidaris plus the crown group
with even higher support (89). Our Bayesian analyses show similar results, with posterior
probability (PP) of 0.85 for the clade of Archaeocidaris plus the crown group echinoids, and 0.6
for the clade of crown group taxa (Fig. 7.2). When five additional sensitivity analyses were run
using varying values of the parameter α for the symmetric Dirichlet prior on character state
transitions (Thompson et al. 2018, Electronic Supplementary Material Figure S2), we found a
clade of Archaeocidaris and the crown group in all analyses. The PP for this clade was between
0.81 and 0.85 dependent upon which value of α was used. Furthermore, in 50% majority rule
trees using all but one (α=0.2) values of α, Yunnanechinus is resolved in the basal trichotomy,
with a clade of all other taxa supported with PPs between 0.54 to 0.71.
266
DISCUSSION
The end-Permian mass extinction, brought about by cataclysmic global climate change
linked to Siberian Traps volcanic outgassing (Svensen et al., 2009), represents the most
significant culling of metazoan life in the Phanerozoic (Stanley, 2016). Though the extinction is
widely accepted to have begun in the Changhsingian stage at the end of the Permian, the Early
Triassic aftermath was characterized by repeated extinction events (Song et al., 2013) and
ecosystem recovery does not appear to have come to fruition until the late Anisian,
approximately 8 Ma after the onset of extinction (Chen and Benton, 2012). The results of our
phylogenetic analyses show that throughout this interval of global turmoil, a clade of stem group
echinoids endured the onset of extinction during the latest Permian and persisted until at least the
middle Anisian. Furthermore, the Louping Biota was deposited in a low energy, intracratonic
basin (Hu et al., 2010), thus the risk of the fragile, imbricate-plated specimens of Yunnanechinus
being reworked (Salamon and Gorzelak, 2017) is low to nonexistent. Our phylogenetic analyses
further re-affirm recent work (Thuy et al., 2017) indicating that the extinction of stem group
echinoids did not take place in the Paleozoic, as has long been postulated. Thus, the long-utilized
distinction between Paleozoic and post-Paleozoic echinoids is not analogous to the distinction
between stem and crown group echinoids. Furthermore, our phylogenetic analyses consistently
resolved a clade of Archaeocidaris and the crown group echinoids (Fig. 7.2, Thompson et al.
2018, Electronic Supplementary Material Figs. S1-S3). The strong support for this clade
indicates that the lineage which gave rise to Yunnanechinus must have diverged from the
archaeocidarids and the crown group prior to the first occurrence of Archaeocidaris in the fossil
record in the Tournaisian (Early Mississippian). Our sensitivity analyses utilizing the alternative
outgroup Paleodiscus ferox (Thompson et al. 2018, Electronic Supplementary Material Fig. S1)
267
for parsimony analyses returned five MPTs. Therefore, although we are confident that
Yunnanechinus is a stem group echinoid, its phylogenetic placement within the stem group
remains tentative. We thus refrain from assigning it to a particular family, but note that it has
plotted with both echinocystid and lepidocentrid echinoids (Thompson et al. 2018, Electronic
Supplementary Material Figs. S1-S3). Yunnanechinus was briefly mentioned previously (Thuy et
al., 2017) and referred to as a proterocidarid; however, our results show no support for the
placement of this genus within the Proterocidaridae.
Our analyses also show that stem group and crown group echinoids, which inhabited the
same environments in the Permian (Thompson et al., 2017b), must have co-existed from the
Roadian, when the first crown group echinoid is known from the fossil record, until at least the
Anisian occurrence of Y. luopingensis (Fig. 7.3A). This implies an overlap in stratigraphic ranges
of approximately 23 Ma between stem group and crown group echinoids (Fig. 7.3B), and
potentially longer as suggested by disarticulated ossicles (Thuy et al., 2017). The presence of Y.
luopingensis in the Middle Triassic now opens the door to questions regarding the diversity and
distribution of stem group echinoids in the Mesozoic. With two Anisian occurrences, the stem
group echinoids were more diverse and widely distributed in the Mesozoic than has hitherto been
thought. This also begs the question: does the extinction of stem group echinoids have more to
do with competitive replacement by crown group echinoids than with the end-Permian mass
extinction? Although Yunnanechinus and the putative proterocidarid from the Muschelkalk
(Thuy et al., 2017) are the first stem group echinoids known from the Middle Triassic, numerous
crown group echinoids have been reported from Anisian and Ladinian strata of the Muschelkalk
Basin (Hagdorn, 1995), Turkey (Hagdorn and Göncüoglu, 2007) British Columbia (Zonneveld,
2001) and China (Stiller, 2002), all belonging to the subclass Cidaroidea. The diversity and
268
abundance of cidaroids compared to these two stem group occurrences indicates that crown
group echinoids attained much higher levels of diversity and a wider distribution than the stem
group echinoids in the Triassic. Thus, despite having endured the end-Permian mass extinction,
the stem group echinoids appear to have been ecologically minor members of Triassic
ecosystems and these occurrences of stem group echinoids probably represent the last vestiges of
a “dead clade walking” (Jablonski, 2002), headed for extinction.
The occurrence of Yunnanechinus from the Anisian of South China also has novel
implications for understanding the Permian-Triassic fossil record of echinoids. The youngest
verified stem group echinoid known from the Paleozoic is the Changhsingian proterocidarid
Pronechinus anatoliensis (Fig. 7.3A) (Kier, 1965). This implies a minimum gap in the fossil
record of approximately 7 Ma between the Anisian Y. luopingensis, the proterocidarid echinoid
from the Muschelkalk, and the next oldest stem group echinoid in the fossil record. Stem group
echinoids thus clearly exhibit the Lazarus effect (Jablonski, 1986) during the latest Permian and
Early Triassic, having disappeared from the fossil record for the entire duration of the Early
Triassic. This is in contrast to crown group echinoids, which have a Late Permian and Early
Triassic fossil record of both articulated tests and disarticulated ossicles. Given that no stem
group echinoids have been sampled from the fossil record of the Early Triassic and are now
known again only from the Anisian supports the explanatory model of Wignall and Benton
(Wignall and Benton, 1999), where Lazarus taxa disappear from the fossil record during the
interval of crisis, only to reappear during the phase of recovery. The Luoping Biota is thought to
represent a recovered ecosystem (Chen and Benton, 2012; Hu et al., 2010) and that stem group
echinoids have reappeared only when ecosystems appear to have stabilized is predicted by this
model. Conversely, the occurrence of disarticulated echinoderm ossicles (Thuy et al., 2017) and
269
relatively complete crown group echinoid tests (Godbold et al., 2017) in deeper water Triassic
environments has led some authors to speculate that these habitats may have been refugia for
stem group echinoids and other echinoderms during the Early Triassic. Whether or not the
disappearance from the fossil record of stem group echinoids is due to mass rarity and low
population size (Hull et al., 2015; Twitchett, 2001; Wignall and Benton, 1999) or to the existence
of habitable, but restricted, refugia in the Early Triassic (Beatty et al., 2008; Godbold et al.,
2017) will remain the subject of further work. Given that stem group echinoids are now known
from the Mesozoic, we urge paleontologists and geologists working on Triassic strata to survey
for stem group echinoids in the field to expand our collective knowledge of their diversity
leading to their eventual extinction.
270
Figure 7.1. Specimens and location of Yunnanechinus luopingensis n. sp. (A) Locality map
showing the location of the Luoping Biota marked as star. Modified from (Hu et al., 2010) (B)
specimen 61701; note the bulge in the centre of the test which likely indicates the Aristotle’s
lantern inside of the compressed test. (C) Specimen 32321 which shows an apical view of a
compressed test with apical disc with genital plates, an ocular plate and the madreporite. (D)
Specimen 61163 showing a compressed test with spines. (E) Close up of spines and ambulacral
plate on specimen 32321. Note the absence of a milled ring and the striate nature of the spines.
(F) Close up view of the madreporite, ocular plate, and adapical coronal plating of specimen
32321. Note the imbrication of the plates, with more adoral plating imbricating over more
adapical plates. (G) Close up of coronal plating and spines on specimen 32321. Spines and
tubercles are arranged in distinct rows with larger spines laying slightly below corresponding
imperforate and non-crenulate tubercles. Scale bars in (A, B, D) are 1 cm, bar in (C) is 2 mm,
and bars in (E-G) are 500 µm.
271
B.
C.
E. F. G.
0 40 80 km
Yunnan
Kunming
Luoping
Banqiao
25˚
26˚
103˚ 104˚
26˚
25˚
104˚ 103˚
China
A.
D.
Shizong
272
Figure 7.2. Results of phylogenetic analyses showing phylogenetic position of Yunnanechinus
luopingensis n. sp. (indicated in bold) relative to Late Paleozoic and Early Mesozoic stem and
crown group echinoid genera. (A) Single most parsimonious tree resulting from equally
weighted parsimony analyses using sixty-nine characters and rooted on Echinocystites ponum.
Tree length=121 steps, CI=0.69, RI=0.74. Numbers at nodes in bold represent bootstrap
proportions resulting from 10,000 “fast” bootstrap replicates while italicised numbers are decay
indices for each node. (B) 50% majority rule consensus tree summarizing posterior distribution
of trees resulting from Bayesian analyses using a symmetric Dirichlet prior with parameter α=∞.
Numbers at nodes represent PP of each node. Scale bar denotes scale for branch lengths, which
are the average branch lengths from the posterior distribution of trees.
273
Diademopsis
ex. gr. heberti
Archaeocidaris
whatleyensis
Lenticidaris
utahensis
Pronechinus
anatoliensis
Echinocystites ponum
Pholidechinus brauni
Maccoya sphaerica
Lovenechinus lacazei
Yunnanechinus
luopingensis n. sp.
Eotiaris
guadalupensis
Proterocidaris belli
Elliptechinus
kiwiaster
Diademopsis
serialis
Eotiaris keyserlingi
9 2, 3 8 9, 6
9 5, 4
7 3, 2
5 2, 1
9 3, 2
6 0, 0
5 9, 0
3 4, 1
3 0, 1
1 6, 1
Archaeocidaris whatleyensis
Diademopsis ex. gr. heberti
Diademopsis serialis
Eotiaris guadalupensis
Eotiaris keyserlingi
Lenticidaris utahensis
Pholidechinus brauni
Elliptechinus kiwiaster
Maccoya sphaerica
Lovenechinus lacazei
Pronechinus
anatoliensis
Proterocidaris belli
Echinocystites ponum
Yunnanechinus
luopingensis n. sp.
A. B.
0.2
0.85
0.70
0.70
0.92
0.60
0.96
0.98
274
Figure 7.3. Diversity of echinoids from the late Paleozoic to the early Mesozoic. (A)
Stratigraphic ranges of echinoid taxonomic groups included in phylogenetic analyses. The
phylogenetic relationship of Yunnanechinus to other stem group echinoids is uncertain, thus its
phylogenetic relationships are depicted with a dashed line. The Early Triassic, which
experienced significant environmental stress (Benton, 2003; Chen and Benton, 2012), lacks any
fossil representation of stem group echinoids, which are Lazarus taxa during this time. (B)
Diversity curve of stem group and crown group echinoids during the Permian and Triassic. The
curve was compiled using the range through method (Alroy, 2010). The dashed line for stem
group echinoids in the Induan and Olenekian represents the inferred presence of stem group
precursors to Yunnanechinus but which have yet to be sampled from the fossil record. Crown
group echinoids are first known from the Roadian, while stem group echinoids go extinct at the
oldest in the Carnian. Throughout the entire figure, stem group echinoids are indicated in red and
crown group echinoids in blue. Data is compiled from Tables S3 and S4 in Thompson et al.
(2018), which list Permian and Triassic echinoid taxa.
275
276
Chapter 8. Paleogenomics of echinoids reveals an ancient origin
for the double-negative specification of micromeres in sea
urchins
Originally published as:
Thompson, J. R., Erkenbrack, E. M., Hinman, V. F., McCauley, B. S., Petsios, E., and Bottjer, D.
J., 2017a, Paleogenomics of echinoids reveals an ancient origin for the double-negative
specification of micromeres in sea urchins: Proceedings of the National Academy of
Sciences, v. 114, no. 23, p. 5870-5877.
INTRODUCTION
The investigation of gene regulatory networks (GRNs) in modern taxa allows for the
understanding of evolutionary changes in the regulatory genome that have underpinned the
evolution of new morphological structures in deep time (Ettensohn, 2009; McClay, 2011; Peter
and Davidson, 2011, 2016). Establishing a timeline for the rates at which these novel structures
arise, and the rate at which the developmental GRNs that encode them evolve, lies at the heart of
evolutionary developmental biology (Davidson and Erwin, 2006). In recent years, identifying
genetic regulatory differences between diverse organisms is becoming more feasible with
broader phylogenetic sampling of developmental and gene expression data across Metazoa
(Hejnol and Lowe, 2015; Levin et al., 2016). These new data provide insight into genomically-
encoded developmental programs and the species-specific GRNs that direct animal development
277
in previously unexplored branches of the tree of life. Studies comparing GRN subcircuit wiring
in distantly diverged taxa (Erkenbrack and Davidson, 2015; Hinman et al., 2003) are thus paving
the way for the study of GRN evolution. However, the arrival of these new data have introduced
new problems. Importantly, as comparative studies of developmental GRNs become more
commonplace, it is critical to keep in mind that simple pairwise comparisons between taxa
violate statistical assumptions of independence and must be carried out in an explicit
phylogenetic framework (Hejnol and Dunn, 2016). Phylogenetic comparative methods thus
provide a rigorous methodology in which to examine gene expression datasets and ultimately
animal body plan evolution (Dunn et al., 2013), within the context of evolutionary time. After
genomic novelties underlying differential body plan development have been identified, we can
then consider the rates at which these novelties arise, and the rates at which GRNs evolve. To
achieve this end, however, an explicit timeline in which to explore GRN evolution is first
necessary.
In a phylogenetically informed, comparative framework, it is possible to infer where on a
phylogenetic tree and when, in deep time, GRN innovations are likely to have first arisen.
Utilizing a paleogenomic approach (Bottjer, 2017; Bottjer et al., 2006) it is possible to
incorporate deep time into an analysis of when GRN novelties arose and to infer the regulatory
interactions directing development of extinct organisms, thereby bringing forth a unique
understanding of the evolution of GRNs and the body plans they encode. Paleogenomics allows
for the dating of the appearance of apomorphic GRNs, their subcircuits, and particular network
linkages by using a combination of the fossil record, statistically derived divergence dates and
comparative analyses of robust experimental data from extant organisms. With reliable dates in
hand, the task of determining rates of GRN evolution is not far off. We set out to establish a
278
rigorous framework for determining the timeline of GRN evolution and test a recently proposed
hypothesis (Erkenbrack and Davidson, 2015) concerning the timing of the evolution of a key sea
urchin GRN novelty, the double negative specification of micromeres.
Echinoids, or sea urchins, are an ideal model system for understanding the mechanistic
basis of GRNs in development and for studying the evolution of development (Davidson et al.,
2002a; Oliveri et al., 2008; Peter and Davidson, 2009). Research on the early embryonic
development of echinoids has revealed the regulatory interactions that comprise the circuitry of
developmental GRNs driving early development of the purple sea urchin Strongylocentrotus
purpuratus (14). Importantly, echinoids also have an excellent fossil record that dates back to
Ordovician strata, more than 400 million years ago (Smith and Savill, 2001). The combination of
a robust fossil record and detailed understanding of the early developmental GRNs in numerous
species makes echinoids an opportune group in which to implement integrated approaches to
understanding GRN evolution. The echinoid crown group is comprised of two clades, the
euechinoids and the cidaroids (Kroh and Smith, 2010). The adult body plans of these two clades
provide a prime example of differential morphological disparity (Gao et al., 2015). Euechinoids
have evolved numerous diverse morphologies throughout their evolutionary history and
comprise morphologically distinct clades such as the bilaterally symmetrical sand dollars and
heart urchins. In contrast, cidaroids have shown remarkable morphological conservation, and the
earliest fossil cidaroids are almost morphologically identical to cidaroids living in the oceans
today (Hopkins and Smith, 2015; Thompson et al., 2015b). Comparative studies of GRN
architecture between cidaroids and euechinoids have revealed extensive differences in the wiring
of their early developmental GRNs (Erkenbrack, 2016; Erkenbrack et al., 2016; Erkenbrack and
Davidson, 2015), and have served to inform our understanding of the genomic underpinning of
279
the myriad morphological differences between their embryonic and larval development
(Schroeder, 1981; Wray and McClay, 1988).
The Double-Negative Gate
A striking example of a GRN subcircuit is the double-negative gate (DNG) that operates in the
early development of S. purpuratus (Oliveri et al., 2002; Oliveri et al., 2008; Revilla-i-Domingo
et al., 2007). This GRN subcircuit employs a double repression mechanism that spatially
localizes genes critical to primary mesenchyme cell (PMC) specification to the micromeres of
the 16-cell stage euechinoid embryo. PMC genes, e.g. alx homeobox 1 (alx1), delta, ets proto-
oncogene 1 (ets1), and tbrain (tbr), are kept silent in the embryo by Hesc, a globally expressed
repressor. Asymmetric localization of maternal regulatory factors at the vegetal pole results in
the upregulation of a repressor of Hesc, paired-class micromere anti-repressor 1 (pmar1),
specifically in the micromere lineage (Davidson et al., 2002a; Oliveri et al., 2002). Cells that
express pmar1 also express PMC-specific genes and are fated to become PMCs, which later
construct the larval skeleton of the embryo.
Although the DNG was first identified in S. purpuratus, its activity has widely been
regarded to occur throughout the euechinoid order Camarodonta (Yamazaki and Minokawa,
2015). The now classic experiment by which ectopic overexpression of pmar1 converts all
embryonic cells to a skeletogenic fate (Oliveri et al., 2002) has been duplicated in numerous
camaradont euechinoids, viz. Hemicentrotus pulcherrimus (Yamazaki et al., 2009),
Paracentrotus lividus (Duloquin et al., 2007) and Lytechinus variegatus (Wu and McClay,
2007), suggesting that DNG function and circuitry are conserved in these taxa. Furthermore,
whole mount in situ hybridization data from the camarodonts H. pulcherrimus (Yamazaki et al.,
280
2010) and L. variegatus (Sharma and Ettensohn, 2010) reveal spatial distributions of hesc and its
downstream targets that are consistent with the existence of the DNG in these euechinoids.
It has also been shown (Yamazaki et al., 2010; Yamazaki and Minokawa, 2015), that the
DNG appears to be present in a number of euechinoid echinoid taxa outside the camaradonts.
Micro1, a pmar1 ortholog, has been identified and localizes to the 16-cell stage micromeres of
the clypeasteroid sand dollar Scaphechinus mirabilis (Yamazaki et al., 2010) and the spatangoid
heart urchin Echinocardium cordatum (Yamazaki and Minokawa, 2015). Furthermore, hesc
shows complementary yet non-overlapping expression patterns with alx1, ets1, and delta at the
10
th
cleavage stage in S. mirabilis (Yamazaki et al., 2010), and roughly complementary
expression with alx1, tbr, and ets1 at the blastula stage in E. cordatum, and the stomopneustoid
Glyptocidaris crenularis (Yamazaki and Minokawa, 2015).
Recent work suggests that the DNG is absent in two species of cidaroid echinoids,
Eucidaris tribuloides (Erkenbrack and Davidson, 2015) and Prionocidaris baculosa (Yamazaki
et al., 2014). No evidence supporting the presence of pmar1 was found in the E. tribuloides
genome, and both organisms show variable, overlapping expression patterns of hesc and its
downstream targets in S. purpuratus, tbr and ets1 (Erkenbrack and Davidson, 2015). These data
strongly suggest that the DNG is absent in cidaroid echinoids. Additionally, both pmar1 and
DNG circuitry are absent from asteroids (McCauley et al., 2010) and ophiuroids (Dylus et al.,
2016), though the latter study identified a potential ortholog to pmar1, pplx, that lacked its
function as a repressor in the ophiuroid Amphiura filiformis. We present here for the first time
data regarding the expression of hesc from the holothurian Parastichopus parvimensis, which
also suggests the DNG is absent in the sister group to echinoids (Telford et al., 2014). The wide
phylogenetic sampling of data indicating the presence or absence of the DNG in numerous
281
euechinoid and cidaroid echinoids, as well as in other echinoderms, makes it an ideal candidate
subcircuit to study the tempo of GRN evolution.
Previously, the initial evolution of the DNG was associated with the divergence of
cidaroid and euechinoid echinoids, at least 268 million years ago (Erkenbrack and Davidson,
2015; Thompson et al., 2015b). However, this conclusion was reached out of the context of a
phylogenetic comparative framework and was based upon comparison of the differential GRN
circuitry responsible for micromere specification in regular euechinoid and cidaroid echinoids
given the most recent date at which euechinoid and cidaroid echinoids could have diverged
(Thompson et al., 2015b). There are multiple evolutionary scenarios pertaining to the timing of
the evolution of the DNG, however, which could explain the presence of this circuitry in regular
euechinoids like S. purpuratus and its absence in cidaroids. By analyzing the presence or absence
of the DNG within an explicit phylogenetic framework, it is possible to estimate the age of
particular nodes in deep time We then estimate the probability that the DNG was present or
absent at ancestral nodes using ancestral state reconstruction to determine support for particular
hypotheses explaining the course of GRN evolution. In order to determine when, on evolutionary
timescales, the DNG likely evolved, we set out to establish a framework for rigorously
investigating a timeline of GRN evolution. We utilized comparative phylogenetic estimates of
echinoid relationships, ancestral state reconstructions, the fossil record, and a wide array of
experimental data concerning the presence or absence of the DNG in numerous echinoid taxa
and other echinoderm outgroups. This paleogenomic approach allowed us to estimate the age of
the DNG and establish a timeline for GRN evolution in echinoids.
282
RESULTS
Holothurians appear to lack the Double-Negative Gate
The holothurian, Parastichopus parvimensis, has only recently been developed as an
experimental developmental system, after improved methods for spawning animals were
established (Le et al., 2013; McCauley et al., 2012). These earlier studies revealed that P.
parvimensis has an alx1-expressing PMC-like cell lineage that produces a small larval spicule.
This lineage arises in the vegetal pole, and, prior to ingression, alx1 is co-expressed with
holothurian tbr, ets1 and erythroblast transformation specific ets-related gene (erg) orthologs
(McCauley et al., 2012). The expression of hesc, and indeed any earlier mechanisms that might
lead to the specification of this lineage at the vegetal pole of P. parvimensis was, however, not
determined.
For the study herein, we blasted hesc and pmar1 sequences obtained from multiple species of
echinoderms against the P. parvimensis transcriptomes
(http://trace.ncbi.nlm.nih.gov/Traces/sra/?study=SRP012968). These transcriptomes were
obtained from RNAs of early gastrula and larval stages (McCauley et al., 2012). This identified a
single predicted hesc transcript (E-values of 9e-13 for SpHesc and 6e-15 for PmHesc).
Phylogenetic analyses (SI Appeidix) show that this transcript is the direct ortholog of the sea star
and sea urchin hesc genes, and we therefore named this gene Pp-hesc. No sequences with
significant similarity to pmar1 or the brittle star pplx were identified, but we cannot discount that
additional sequence coverage, especially of earlier embryonic stages, might reveal an ortholog.
We next used whole mount in situ hybridization to determine the spatial expression of Pp-hesC.
Transcripts are localized in the vegetal plate from hatching through to early invagination
(Thompson et al. 2017, SI Appendix, Fig S1), and in a spotted pattern throughout the animal
283
ectoderm. hesc continues to be expressed in the archenteron and, by the larval stage, transcripts
are localized to the top of the archenteron and diffusely across the ectoderm (Thompson et al.
2017, SI Appendix, Fig S1F.). Expression of Pp-hesC is very reminiscent of the expression of
the orthologous gene in the sea star Patiria miniata (shown for comparison in Thompson et al.
2017, SI Appendix, Fig S1). These data therefore cannot support a role for Hesc as a dominant
repressor of tbr or erg in P. parvimensis. Because Hesc in P. parvimensis does not repress tbr, a
key downstream target of Hesc repression in the DNG, we coded the DNG as absent in
holothurians in our ancestral state reconstructions.
Divergence Time Estimation
In order to constrain the age of the DNG with respect to evolutionary time, and the age of
particular nodes onto which we reconstructed ancestral states, we integrated fossil and molecular
data by using a fossil calibrated relaxed molecular clock (Yang and Rannala, 2006) to estimate
times of divergence. Because genome scale data is lacking for echinoids, we instead decided to
explicitly account for topological and branch length uncertainty in our downstream analyses. We
thus estimated divergence times using fifteen alternative topologies (thirteen of the fifteen
reached convergence and are discussed herein) resulting from the resolution of polytomies
existing in our consensus tree inferred from Bayesian phylogenetic analyses in Phylobayes V.
4.1 (Lartillot et al., 2009). We additionally estimated divergence times on the best maximum
likelihood (ML) tree, which we used to perform a number of sensitivity analyses (Thompson et
al. 2017, SI Appendix)
The 95% credible intervals (CI) and mean posterior divergence times for nodes which are
informative for tracing the evolution of the DNG on one of our thirteen alternative topologies are
284
shown in Figure 1, while mean ages and CIs for all other divergences are shown in Thompson et
al. 2017, SI Appendix, Figures S7-S19 and Tables S8-S20. Because some dated nodes are not
present across all trees due to topological variation, mean ages and CIs for nodes not explicitly
discussed in the main text can be found in Thompson et al. (2017) SI Appendix. Across all trees,
CIs on nodes which were directly calibrated by fossils are all relatively narrow, while 95% CIs
on nodes which were not calibrated by fossils, such as the Irregularia and Camarodonta for
instance, are wider. Of particular interest for questions regarding GRN evolution are the ages of
particular nodes onto which we have reconstructed the presence or absence of the DNG. All of
our analyses show a Carboniferous-Permian (approximately 265 to 342 MYA) divergence for the
euechinoids and cidaroids, and a Triassic to Early Jurassic diversification of basal euechinoid
lineages. The mean divergence time of the Camarodonta is at the oldest in the Triassic, and at the
youngest in the Late Cretaceous, dependent upon the topology used to estimate divergence times
(Thompson et al. 2017, SI Appendix, Figs. S7-S19). The mean divergence time of the Irregularia
is approximately Jurassic in most analyses, though the CI on this node is in all analyses wide
(>100 Ma). The node representing the most recent common ancestor (MRCA) of all analyzed
cidaroids has a 95% CI between ≈8 and ≈230 MYA putting the divergence of the analyzed extant
cidaroids in the Cenozoic or Mesozoic. This node was not directly calibrated by fossils, which
likely explains the wide 95% CI. The mean age for the node representing the most recent
common ancestor of all extant euechinoids is Middle Triassic in all alternative topologies, while
the CI ranges from Early Triassic or Latest Permian (≈252 Ma) to Late Triassic (≈220 Ma) on all
topologies. The mean estimated divergence times for the divergence between the echinothurioids
(Araeosoma+Asthenosoma) and (aspidodiadematoids+pedinoids)+diadematoids are in the Late
Triassic with CIs from the Late Triassic (≈211 Ma) to Middle Triassic (≈246 Ma) on all thirteen
285
topologies. Furthermore the mean ages and CIs for the divergences between
(aspidodiadematoids+pedinoids) and diadematoids and aspidodiadematoids and pedinoids reside
within the Late Triassic. Results from analyses utilizing the ML topology are broadly
comparable, and can be found in Thompson et al. (2017) SI Appendix, Figure S3 and are
generally insensitive to changes in model parameters and tree topology (Thompson et al. 2017,
SI Appendix and Figs. S4 and S6).
Ancestral State Reconstruction
Recent paleogenomic studies have used the fossil record to date the appearance of particular
GRNs in evolutionary time by determining the first appearance of the morphological structures
that are specified by these GRNs in the fossil record (Bottjer et al., 2006). There is no fossil
record of embryonic sea urchins and we thus used ancestral state reconstructions (Pagel et al.,
2004) to determine the probability that the double negative gate was present or absent in the last
common ancestor of particular clades. We analyzed two datasets for our ancestral state
reconstructions. One of these datasets maximized taxon sampling and included many more taxa
for which experimental evidence was unknown than known. This dataset is based upon 1300
time scaled trees (See Divergence Time Estimation), with 100 trees of differing branch lengths
randomly sampled from the posterior distribution of each of the thirteen divergence time
estimation analyses run using alternative topologies. In this way, we explicitly account for
phylogenetic uncertainty in our ancestral state reconstructions. The other dataset was limited to
taxa for which direct experimental evidence concerning the DNG was available, and the
topology onto which this ancestral state reconstruction was run reflected the ML estimates of
topological relationships (Fig. 8.2, Thompson et al. 2017, SI Appendix, Fig. S5).
286
Our Bayesian ancestral state reconstructions using the taxonomically inclusive dataset
were run under a variety of priors on the transition rates for each state. Given that the presence or
absence of the DNG is unknown for many of the taxa in our tree, with more uninformative priors
(Uniform distribution with bounds 0 and 100), the probability of reconstructed ancestral states at
all nodes in the phylogeny is 0.5, with an equal probability of presence or absence of the DNG at
each ancestral node. Using narrower priors, which may be more appropriate for modeling the
presence or absence of the DNG, which is unlikely to have evolved and unevolved numerous
times throughout the evolutionary history of echinoids, results in more informative
reconstruction of ancestral states. ML reconstruction of these ancestral states on all but four of
the 1300 trees result in transition rate parameters of less than 0.001 (Thompson et al. 2017,
Dataset S1), which is in line with the more informative reconstructed ancestral states resulting
from smaller priors on transition rates. Using a uniform prior from 0 to 0.01 on transition rates
results in mean posterior probabilities showing relatively high mean posterior probabilities (PPs)
in favor of the presence of the DNG in the last common ancestor of camarodonts (0.76),
irregularia (0.76), and in the last common ancestor of Strongylocentrotus and Paracentrotus
(0.97) (Fig. 8.1). The DNG was present in the MRCA of camarodonts and the Irregularia with a
mean PP of 0.79, the MRCA of diadematoids and camarodonts with a mean PP of 0.76, and that
of the MRCA of camarodonts and stomopneustoids with a mean PP of 0.78. The MRCA of
echinoids appears to have utilized the DNG with a mean PP of 0.65 while the MRCA of
euechinoids appears to have utilized the DNG with a mean PP of 0.76. The MRCA of extant
cidaroids, however, appears to have not utilized the DNG with a mean PP of 0.95. These
probabilities of the reconstructed presence and absence of the DNG at select nodes are plotted on
Figure 1. When using a slightly more diffuse prior (U(0, 0.05)), mean posterior probabilities of
287
presence or absence of the DNG at all reconstructed ancestral nodes approaches .5. Under this
prior, mean PP’s of presence or absence of the DNG of most reconstructed ancestral nodes are
between .45 and .55 The exception to these less informative reconstructions are the node
representing the MRCA of extant analyzed cidaroids, where the DNG was absent with a mean
PP of 0.72, and the node representing the MRCA of Strongylocentrotus and Paracentrotus,
which likely utilized the DNG with a mean PP of 0.75.
Many of our mean PP values using the dataset that maximized taxon sampling are
systematically close to 0.50, resulting from the large amount of taxa included in the analysis for
which no data are available concerning the DNG. This is especially the case when wider priors
are used. This is likely because the tips with missing data are treated as having an equal
probability of the presence or absence of the DNG in ancestral state reconstructions (Pagel et al.,
2004). Using a pruned dataset (Fig. 8.2), which included only taxa from orders for which direct
experimental evidence regarding the presence or absence of the double negative gate exists,
resulted in posterior probabilities farther from 0.50 and more representative of the current state
of knowledge with regard to taxonomic sampling of developmental data. This analysis was run
using the ML branch lengths, as opposed to a time scaled tree, which was able to more
adequately resolve ancestral states with a wider prior (U(0, 100). This dataset showed high mean
PPs (Fig 8.2, Thompson et al. 2017, SI Appendix, Table S3), and showed that the MRCA of the
camarodonts used the DNG with a mean PP of 0.91, while the DNG was present in the MRCA of
the Irregularia+Camadodonta with a mean PP of 0.82 and that of the Carinacea with a mean PP
of 0.83. The last common ancestor of echinoids shows mean PPs of 0.41 and 0.59 respectively in
favor of the absence or presence of the DNG. Using a tree topology which is based on
morphology (Kroh and Smith, 2010) with this pruned dataset also show similarly high posterior
288
mean PPs supporting the presence of the DNG in the MRCA’s of the camarodonts, the
Carinacea, the Irregularia, and the Camarodonta+Stomopneustidae (Thompson et al. 2017, SI
Appendix, Table S3 and Fig. S5).
DISCUSSION
The Antiquity of the Double-Negative Gate
Our data strongly favor a euechinoid origin of the DNG, which appears to have evolved at the
latest sometime before the Late Triassic (≈220-230 MYA depending on the topology utilized).
Support is highest for the presence of the DNG in the MRCA of camarodonts and at the MRCA
of Paracentrotus and Strongylocentrotus, which is in line with other recently published studies
showing marked conservation in the timing of gene expression between camarodonts (Gildor and
Ben-Tabou de Leon, 2015). Results also favor the presence of the DNG over its absence in the
last common ancestor of early branching euechinoid clades. This indicates that the DNG was
likely utilized in the MRCA of extant euechinoids. The pruned dataset in particular shows high
support for a pre-Carinacea age of the DNG; however, this dataset provides limited data
concerning the non-Carinacea euechinoids, as there is not yet direct experimental evidence from
these early diverging euechinoid taxa. Future experimental work in these euechinoids, such as
the diadematoids, will augment phylogenetic sampling and further resolve the age of the DNG in
the euechinoid lineage. In the last common ancestor of echinoids, our Bayesian analyses show a
mean PP of 0.65 for the presence of the DNG at this node under our smallest prior. Our analyses
using the ML topology also show mean PPs slightly favoring the presence of the DNG over its
absence at this node. Because mean PPs at this node show the lowest support among all nodes
examined for the presence of the DNG, we are tentative in interpreting that the DNG was utilized
289
in the MCRA of all extant echinoids. Given the much higher mean PPs at the MRCA of extant
euechinoids, however, it seems more likely that the DNG evolved some time between the
divergence of euechinoids and cidaroids in the Carboniferous-Permian and the MRCA of extant
euechinoids in the Triassic. It is also necessary to note that that our results at the more basal
euechinoid nodes are more sensitive to prior choice than those more recently branching nodes
(Thompson et al. 2017, SI Appendix, Table S21), and broader phylogenetic sampling in the
future will hopefully decrease the sensitivity of results at these nodes to prior choice
Because there is low support for either the presence (mean PP=0.65) or absence (mean
PP=0.35) in the node representing the divergence of euechinoids and cidaroids, we are unable to
confidently attribute the evolution of the DNG to the divergence of the euechinoids and cidaroids
as we have done in the past (Erkenbrack and Davidson, 2015; Thompson et al., 2015b). Previous
authors have stated that the DNG is an evolutionary novelty unique to all euechinoid echinoids
(Erkenbrack and Davidson, 2015; Ettensohn, 2009; Thompson et al., 2015b) and our ancestral
state reconstructions show mean PPs which strongly identify the DNG as a novelty unique to all
extant euechinoids. Our analysis also fails, however, to demonstrate unambiguously that the last
common ancestor of echinoids (including cidaroids) did not utilize the DNG, and instead utilized
the mechanism present in extant cidaroids (Erkenbrack and Davidson, 2015; Thompson et al.,
2015b). Our results instead slightly favor the alternative hypothesis, that the DNG was utilized in
the MRCA of extant echinoids. We are hesitant about this interpretation, however, given that PPs
for both the presence and absence of the DNG at this node are closer to 0.5 than at any other
nodes we examined.
Utilizing ancestral state reconstruction provides a sound statistical framework in which to
explore the evolution of GRNs. Although we ourselves have previously made statements
290
regarding GRN evolution from cross-species comparisons in a few taxa (Erkenbrack and
Davidson, 2015; Thompson et al., 2015b) , we herein demonstrate that statistical analyses of
GRN evolution within the context of a phylogeny provide a more rigorous framework in which
to examine their evolution. Whereas a most parsimonious interpretation of our data would have
led us to make binary statements regarding the presence or absence of the DNG at particular
ancestral nodes, our approach allows us to assign probabilities to our conclusions while taking
into account uncertainty in phylogenetic topology and branch lengths. Surprisingly, within this
statistical framework we cannot definitively determine whether or not the DNG was utilized in
the MRCA of all extant echinoids. This result indicates that inferences following from cross-
species comparisons of GRNs are at best one of many evolutionary scenarios and that caution
should be exercised with regard to definitive statements about evolutionary changes in gene
regulation associated with divergence of taxa. As still more data become available concerning
utilization of the DNG in diverse echinoid clades, and as the echinoid phylogeny is improved
upon with new phylogenomic techniques, our confidence regarding the phylogenetic origin, and
antiquity, of the DNG will improve. At the very least, we are able to say that the DNG is an
ancient GRN subcircuit, which likely evolved in or before the MRCA of extant euechinoids and
has been conserved and utilized throughout the euechinoid tree of life and across at least 220
million years of echinoid evolution.
The Double Negative Gate and Echinoid Macroevolutionary Trends
By obtaining age estimates for the DNG we are able to compare this GRN innovation
with other events in the macroevolutionary history of echinoids. It has been recently shown that
crown-group echinoids underwent heightened rates of evolution during the Early Jurassic
(Hopkins and Smith, 2015), which, given the results of our analysis, post-date the origin of the
291
DNG during or prior to the Late Triassic. This burst of diversification is tightly linked to the
diversification of irregular echinoids and other euechinoid clades that our analyses indicate likely
utilized the DNG. As the DNG functions to specify skeletogenic cells in developing embryos, its
correct spatiotemporal deployment is necessary for embryonic and larval skeletogenesis. That
such a critical function is dependent upon the DNG may have even resulted in the existence of
“fail-safe” regulatory wiring of blimp1 and hesc, wherein Blimp1 represses hesc in the absence
of Pmar1 (Smith and Davidson, 2009). Given that the DNG appears to have been conserved
across ≈220 million years of echinoid evolution, it is possible that the DNG is so critical for early
indirect development in euechinoids that it has undergone stabilizing selection, though this has
yet to be empirically demonstrated. The conserved role of the DNG in micromere specification,
which appears to predate the diversification of irregular euechinoids, also indicates that the
regulatory mechanism of micromere specification very early in development has been
constrained, while aspects of later larval and adult morphology, which show high degrees of
disparity (Hopkins and Smith, 2015; Wray, 1992) have been relatively free to vary. The
euechinoid double negative specification of micromeres and skeletogenesis thus offers a striking
example of genomically encoded developmental constraint early in development, while high
degrees of plasticity have prevailed in the evolution of later larval and adult morphology. This is
also in stark contrast to the cidaroids, which although displaying much more constraint in their
adult body plans (Hopkins and Smith, 2015) display high levels of morphological variability in
early development. Whereas euechinoids have an invariant four micromeres, cidaroids contain a
variable number, even within a single species (Erkenbrack and Davidson, 2015; Schroeder,
1981; Wray and McClay, 1988). This apparent differential constraint at different times in
development found in these two clades poses an intriguing avenue for future research.
292
“Fossilized” Gene Regulatory Networks
Furthermore, paleogenomics allows us to frame hypotheses regarding the nature of
embryonic micromere specification in fossil echinoids. Because it has been shown that direct
development in echinoids was not widespread until the Cretaceous (Jeffery, 1997), our analyses
show that most indirect developing euechinoids, which make up the majority of the echinoid
fossil record since the Triassic, likely utilized the DNG. When put into the context of
environmental and ecological change undergone by echinoids through the Cretaceous-Paleogene
mass extinction, which had profound effects on echinoid size, feeding strategy, and
biogeographic distribution (Smith and Jeffery, 1998) and further Cenozoic climatic changes
(Kroh, 2007), it suggests that the DNG has been very robust to evolutionary change over the
course of the past 220 million years. Paleogenomics offers the opportunity to compare
presumptive evolutionary changes in GRNs with paleoclimatic and paleoenvironmental changes.
We assert that this is a promising and intriguing avenue of future research, and is only one of the
many doors that can be opened by paleogenomic approaches to studying the evolution of GRNs.
METHODS
Phylogenetic Tree Construction
For our dataset, we chose to use previously published mitochondrial 16s, 18s and 28s small
subunit rRNA sequence data (Littlewood and Smith, 1995; Smith et al., 2006; Stockley et al.,
2005), which consists of the most comprehensive molecular dataset with respect to echinoid
taxonomic sampling known. Although using such a dataset for inference of topology has limited
use compared to more widely used phylogenomic datasets, we have accounted for uncertainty in
293
topology and branch length in our downstream ancestral state reconstructions. We used the
concatenated aligned dataset of (Smith et al., 2006) with gaps removed, leaving 3230 sites
(Thompson et al. 2017, Dataset S2-S3). For our ML analysis, we determined the best fitting
substitution model using Modeltest 3.7 (Posada and Crandall, 1998) in PAUP* (Swofford, 2003)
which amongst the models available in MCMCTREE (which we used to estimate divergence
times on our ML topology) found the GTR+Γ to be the best fit. The outgroup taxa were
constrained to represent the recently supported Asterozoa hypothesis (Telford et al., 2014).
Phylogenetic trees were constructed using RaxMLversion 8 (Stamatakis, 2014) where we ran 20
ML analyses, with the tree with the highest likelihood used for ancestral state reconstructions
and divergence time estimation. This tree, showing bootstrap support from 1000 replicates, is
shown in Thompson et al. (2017) SI Appendix, Figure S2.
Our ML topology differs from previously published estimates (Smith et al., 2006) and the
most up-to-date morphological phylogeny (Kroh and Smith, 2010) predominantly in the
placement of a clade consisting of the stomopneustoids and arbacioids plotting as sister group to
a clade of Camarodonta+Irregularia. Because this topology differs with morphology and
previous molecular estimates, which either put arbacioids as sister to the camarodonts or
Arbacioida+Stomopneustoida as sister group to the camarodonts respectively, we decided to run
20 additional ML analyses using a constrained topology which reflects most recent
morphological phylogeny (Fig. 5 in (Kroh and Smith, 2010)). This topology is shown in
Thompson et al. (2017) SI Appendix, Figure S4, and was used for sensitivity analyses.
In addition to our ML analyses, we performed a Bayesian analysis in Phylobayes V. 4.1
(Lartillot et al., 2009) using the CAT-GTR+ Γ model (Lartillot and Philippe, 2004). Two
independent Markov chains were run until both reached convergence. The burn in for each chain
294
was 2000 points and each chain was sampled every ten points up to 20000 points. Chains were
run until the effective sample size for each parameter was greater than 300, and the relative
difference between parameters of each run was less than .1. Convergence of MCMC was
assessed using the tracecomp and bpcomp commands in Phylobayes.
The majority rule consensus tree resulting from our Bayesian analyses (Thompson et al.
2017, SI Appendix, Fig. S20) resolved many traditional clades supported by morphology and
previous molecular analyses including the cidaroids, camarodonts and two clades of irregular
echinoids. This tree, however, contained a large polytomy where the relationships of many of
these major echinoid clades to each other were unresolved. An analysis in Phylobayes using the
same model parameters and convergence criteria was also run using only the 18s and 28s
sequence data (Thompson et al. 2017, Dataset S4). This topology retained most of the same
clades as the topology resulting from inference using all three genes, and was equally unresolved
(Thompson et al. 2017, SI Appendix, Fig. S21). We thus chose to use the topology resulting
from inference using all three genes as a backbone for topologies used in divergence time
estimation and ancestral state reconstruction (See Divergence Time Estimation and Ancestral
State Reconstruction). Given that we explicitly account for phylogenetic uncertainty in our
ancestral state reconstruction, we do not expect that inference using polytomy resolutions of the
two-gene consensus tree would seriously alter our results. Because the goals of this analysis were
not explicitly to reconstruct echinoid phylogeny, but rather to examine the evolution of the DNG
within a phylogenetic context, the phylogenetic uncertainty evident from our Bayesian analysis
is accounted for in our ancestral state reconstructions (see below).
295
Divergence Time Estimation
The ML tree with the highest log-likelihood was used for divergence time estimation in the
MCMCTREE package in PAML (Yang, 2007). After removal of the outgroup taxa, a global
clock model (clock=1 in BASEML) with the root of the tree (euechinoid-cidaroid divergence)
calibrated was used to estimate a suitable prior for the substitution rate µ in BaseML. We thus
used a gamma-Dirichlet distribution with shape parameter α=1.79 and scale parameter β=70
(rgene_gamma=1.79, 70 in MCMCTREE). Parameters α and β of the gamma Dirichlet prior on
σ
2
, which specified how variable rates are across branches, were set to 1 and 2.68 respectively.
The birth death priors were set to λ=1, µ=1, ρ=0 (BDParas= 1 1 0). Divergence time estimation
using the ML topology was done using MCMCTREE using the approximate likelihood
calculation of dos Reis and Yang (2011). We chose to use the independent rates model
(Drummond et al., 2006; Rannala and Yang, 2007). For our time priors, we used nine uniform
constraints, most of which focused on the earlier branching nodes of the tree. MCMCTREE uses
soft constraints (Yang and Rannala, 2006), and for all of our divergence time priors 2.5% of the
distribution was used as a soft tail on both the minimum and maximum constraints. Two
independent MCMC chains were run for 5,000,000 iterations with a sample drawn every 50
iterations. The first 50,000 runs were discarded as burn in. This resulted in a total of 100,001
samples in each chain. Following analyses, all resultant MCMC samples were examined in
Tracer v1.6 (Rambaut et al., 2013) to check for convergence.
Fossil calibrations for analyses in MCMCTREE were compiled and justified with up-to-
date absolute dates and the first attempt at using “Best Practices” (Parham et al., 2011) to
rigorously assemble well supported fossil calibrations for use in prior construction for divergence
time estimation in echinoids. Our compilation is not exhaustive, as our primary goal was to
296
rigorously calibrate nodes towards the base of the echinoid tree, where most discussion regarding
the origin of the DNG has been focused. A full list of fossil calibrations with justifications are
listed in Thompson et al. (2017) SI Appendix, Fossil Calibrations for Priors on Divergence Time
Estimates and Datasets S5-S6.
We additionally ran sensitivity analyses to check the robustness of our results to
variations in prior, model choice, and tree topology. Our results were, in general, robust to
changes in model parameters, and sensitivity analyses are shown in Thompson et al. (2017) SI
Appendix, Figure S6 and discussed in the SI Appendix.
Because our Bayesian majority rule consensus tree was unresolved, we estimated
divergence times using multiple phylogenetic topologies resulting from resolution of our
Bayesian majority rule consensus tree. Nodes in the consensus tree with PPs < .25 were
collapsed. Following this, all irregular echinoids were constrained to form a clade (following
morphological evidence (Kroh and Smith, 2010)) and divergence times were estimated on the 15
possible resolutions of the resulting tree. Divergence times were estimated using Phylobayes V.
4.1 (Lartillot et al., 2009) under the model of (Drummond et al., 2006) where the rate on each
branch is an independent draw from a gamma distribution and using the CAT-GTR+ Γ model for
the substitution process. The birth death priors were set to λ=1, µ=1, ρ=0. Fossil calibrations
were soft and used variable combinations of the same calibrations as used for divergence time
estimation using the ML tree plus three calibrations for divergences outside of echinoids
(Thompson et al. 2017, SI Appendix, Fossil Calibrations for Priors on Divergence Time
Estimates and Datasets S7-S20). Each analysis thus used from nine to twelve calibrations.
For each topology, two independent chains were run with a burn-in of 1000 points
sampling every 10 points. MCMC convergence was assessed using tracecomp and bpcomp in
297
Phylobayes. Chains were run until the effective sample size for each parameter was greater than
100, and the relative difference between parameters of each run was less than .3. Thirteen of the
fifteen analyses reached convergence within time available, while the two analyses which failed
to converge were excluded from ancestral state reconstructions. These two topologies displayed
relationships widely different from previous molecular and morphological inferences of echinoid
topology (Kroh and Smith, 2010; Smith et al., 2006), so their exclusion from ancestral state
reconstruction likely does not alter our inferences.
Ancestral State Reconstructions
Representatives from families (or orders in the case of the Stomopneustoida) and non-echinoid
echinoderm classes for which there is experimentally derived evidence in favor or in disfavor of
the presence of the double negative repression of hesc by Pmar1, and alx1, tbr, or ets1 by Hesc
were used to score extant taxa for the presence or absence of the double negative gate. For
instance, the cidaroid Prionocidaris baculosa expresses hesc in the same cells as tbr, alx1 and
ets1 at the late blastula stage, which indicates that Hesc could not be acting as a repressor of
these genes, a canonical feature of the DNG, and as such the DNG is not utilized in this taxon
(Yamazaki et al., 2014). When experimental data has shown precise non-overlapping expression
patterns, and experimentally tested GRN linkages that mirror those of S. purpuratus, in which
the DNG was first discovered, the DNG was scored as present in that taxon. Because we are
interested in modeling the evolution of the DNG as it was first described from S. purpuratus, we
feel that coding the DNG as present/absent is appropriate for this analysis because any difference
in experimentally derived linkages, especially evidence for non-repression of alx1, tbr, or ets1 by
298
Hesc or non-repression of hesc by Pmar1, allows us to code a taxon as absent, and thus
maximize phylogenetic coverage.
We utilized two datasets for our analyses, one which maximized phylogenetic coverage,
but had low proportional coverage of experimentally derived presence and absence data, and a
pruned dataset, which included only taxa from families (or the Stomopneustoida) for which
direct experimental evidence regarding the presence or absence of the double negative gate
exists. The dataset maximizing phylogenetic coverage was used to reconstruct ancestral states
across the 1300 time calibrated trees resulting from divergence time estimation. In this way,
phylogenetic uncertainty with respect to both branch length and topology were accounted for in
reconstruction (Pagel et al., 2004) (Thompson et al. 2017, Supplemental Dataset S23). We ran
these analyses under four differing sets of priors on transition rates (Thompson et al. 2017, SI
Appendix Table 21). This dataset was additionally run using the ML branch lengths from the
best ML tree inferred in RAxML and the ML estimates of branches constrained on the
morphological tree of (Kroh and Smith, 2010) (Thompson et al. 2017, SI Appendix, Tables S1
and S2). The topology of the pruned dataset was constrained in RaxML prior to branch length
estimation such that the topological relationships matched those of the ML tree inferred from the
more complete dataset. All analyses were run in BayesTraits V2.0 (Pagel et al., 2004) using the
MRCA command, which reconstructs probabilities of ancestral states at the node representing
the most recent common ancestor of the specified taxa (Thompson et al. 2017, Datasets S21-
S22). We ran our MCMC analyses for 10,000,000 iterations with a burn in of 2,000,000
iterations and sampling every 1,000 iterations. All analyses were run under varying priors to
assess sensitivity of analyses to prior choice.
299
Figure 8.1. Fossil calibrated time tree showing age of clades of crown group echinoids based
upon one of the fifteen alternative topologies resulting from the Bayesian phylogenetic analysis
and showing the probability of presence or absence of the double negative gate on nodes of
interest. Blue bars are 95% credible intervals on divergence times for nodes. Pie charts show
mean posterior probabilities that the double negative gate was present (blue) or absent (red) at
particular ancestral nodes in ancestral state reconstructions on 1300 alternative topologies under
priors on transition rates of U(0, 0.01). Taxon names that are blue have been demonstrated
experimentally to utilize the double negative gate for micromere specification, while red taxa do
not utilize the double negative gate. Taxonomic names represent higher-level echinoid
taxonomic groupings. Posterior probabilities for ancestral state reconstructions shown in
Thompson et al. (2017) SI Appendix, Table S8. * indicates Stomopneustoida+Arbacioida.
Geological timescale made in STRAP (62). Spatangoid silhouette illustrated by Hans Hillewaert
(Vectorized by T. Michael Keesey). Cidaroid silhouette illustrated from Didier Descouens
(vectorized by T. Michael Keesey). Regular euechinoid from Frank Förster (based on a picture
by Jerry Kirkhart; modified by T. Michael Keesey). These are available for use under a Creative
Commons Attribution-ShareAlike 3.0 Unported license
(https://creativecommons.org/licenses/by-sa/3.0/). Clypeasteroid image from Michael Site and is
available for use under a Creative Commons Attribution-NonCommercial 3.0 Unported license
(https://creativecommons.org/licenses/by-nc/3.0/). All are from Phylopic (www.phylopic.org).
300
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
180
190
200
210
220
230
240
250
260
270
280
290
300
310
320
330
340
350
360
370
380
390
400
410
420
430
440
450
460
470
480
490
500
510
520
530
540
Terreneuvian
Series2
Series3
Furongian
Early
Middle
Late
Llandovery
Wenlock
Ludlow
Pridoli
Early
Middle
Late
Mississippian Pennsylvanian
Cisuralian
Guadalupian
Lopingian
Early
Middle
Late
Early
Middle
Late
Early
Late
Paleocene
Eocene
Oligocene
Miocene
Cambrian Ordovician Silurian Devonian Carboniferous Permian Triassic Jurassic Cretaceous Paleogene Neogene
Antedon
Ophiocan
Asterias
Psychopetes
Cucumaria
Prionocidaris
Stereocidaris
Calocidaris
Diadema
Centrostephanus
Caenopedina
Aspidodiadema
Asthenosoma
Araeosoma
Stomopneustes
Coelopleurus
Arbacia
Lytechinus
Strongylocentrotus
Psammechinus
Paracentrotus
Anthocidaris
Tripneustes
Sphaerechinus
Genocidaris
Cytechinus
Pseudechinus
Salmacis
Salmaciella
Temnotrema
Temnopleurus
Microcyphus
Mespilia
Holopneustes
Amblypneustes
Echinoneus
Fellaster
Encope
Echinodiscus
Rumphia
Echinocyamus
Echinolampas
Conolampas
Cassidulus
Plexechinus
Spatangus
Echinocardium
Brissopsis
Meoma
Archaeopneustes
Paleopneustes
Abatus
Cidaroidea
Euechinoidea
95% credibile Interval on
divergence time estimate
Mean posterior probability of
divergence time estimates
Mean posterior probability of
DNG Absent or Present
Camarodonta Irregularia
Carinacea
*
301
Figure 8.2. Ancestral state reconstruction of the DNG. (A) Pruned Analyses showing
reconstructed ancestral presence or absence of the double negative gate at select nodes on the
ML tree using ML estimates of branch lengths. Pie charts show the mean posterior probabilities,
with red sections representing probability of absence of the DNG, and blue representing
reconstructed probability of the DNG being present at a given ancestral node. (B) Wiring
diagram for extant cidaroid echinoids showing the probable interactions of key genes involved in
the DNG in S. purpuratus. There is no double negative repression, as there is no Pmar1 gene and
hesc does not repress tbrain or ets1/2 as it does in euechinoids (C) Wiring diagram showing the
interactions of genes involved in the DNG in S. purpuratus and other examined euechinoids.
302
303
Chapter 9. Conclusions
The results of this dissertation highlight the important of multidisciplinary studies of
diversification. By combining data from the fossil record, geological record, gene sequences, and
gene expression patterns, I have elucidated both extrinsic and intrinsic processes driving sea
urchin diversification on a number of scales. I have herein shown that the initial diversification
of echinoids was rapid. All of the genera of bothriocidaroids appear to be clades, and these
clades all appear to have diversified early in the evolutionary history of echinoids. I have also
shown that different clades of echinoids had different environmental tolerances during the
Carboniferous period, which may have controlled the differing macroevolutionary history of
these groups in the Carboniferous and afterwards. Furthermore, the end-Permian extinction lead
to a decline of echinoid diversity in the late Paleozoic and perhaps caused the extinction of the
archaeocidarids. Nevertheless, the echinoid stem group did survive this event, and thus their
extinction cannot be tied definitively with the Permian-Triassic extinction. I have also
characterized the molecular underpinning of embryonic development in Early Mesozoic crown
group echinoids, and have shown that many of the earliest members of the euechinoidea likely
utilized the DNG. This indicates that this genetic regulatory novelty is apparently quite old, and
has remained evolutionarily constrained for over 200 million years. In addition to these insights,
a number of larger thematic issues have come from this work, which are detailed below.
THE IMPORTANCE OF PHYLOGENETICS IN ASSESSING
MACROEVOLUTIONARY HISTORIES
Throughout the course of this work, I have consistently used phylogenetics to attempt to better
understand the patterns of echinoid diversification. The echinoid crown group thus appears to
304
have evolved from species of Archaeocidaris with multiple rings of interambulacral scrobicular
tubercles, and with crenulate tubercles. This characterization of the late stem group and early
crown group echinoids helps to refine our idea of what these taxa may have looked like, and
further refines our understanding of morphological transitions in the history of sea urchins. This
also indicates that the Archaeocidaridae, and in particular the genus Archaeocidaris, is
paraphyletic. Furthermore, by using phylogenetics I was able to demonstrate analytically that the
Triassic echinoid from the Luoping Biota, Yunnanechinus luopingensis, was a stem group
echinoid. In this way I was able to demonstrate that a lineage of stem group echinoids survived
the extinction, and that the morphology of Yunnanechinus luopingensis wasn’t simply
convergent upon the morphology of other crown group echinoids. By using phylogenies to
examine survivorship over extinction intervals, it is possible to have a much clearer picture of the
history of life.
INTEGRATING MULTIPLE DATA TYPES TO UNDERSTAND MACROEVOLUTION
In order to rigorously constrain the evolution of gene regulatory networks, an integrated
approach is necessary which explicitly accounts for uncertainty. By using phylogenetic
comparative methods to explicitly test different evolutionary hypotheses it is possible to have a
clearer picture of when GRNs or particular GRN subcircuits evolved. Fossils are key to
understanding the timing and rate of GRN evolution. Because the fossil record affords us the
timing of evolutionary events, including key instances of cladogenesis in the history of life, by
using fossil-calibrated divergence time estimation, a more accurate view of the history of life
becomes apparent. Furthermore, with wide sampling from numerous different taxa, it is possible
to account for instances of convergent evolution.
305
THE IMPORTANCE OF SYSTEMATICS
Almost all of this work is grounded in systematic paleontology and taxonomy. Because
taxonomic and morphological data forms the raw data on which most paleobiological studies are
based, new fossil finds can pave the way for revised macroevolutionary interpretation. In the
studies presented herein, taxonomic descriptions allowed for a number of novel interpretations.
With regard to the Ordovician-Silurian transition, the new species Neobothriocidaris
pentalandensis indicated that a clade of Neobothriocidaris survived the end-Ordovician
extinction (Similar to Y. luopingensis and the end-Permian extinction; discussed above).
Furthermore, detailed systematic descriptions of the Guadalupian fauna from west Texas
indicated that stem group and crown group echinoids co-existed in the late Permian in the same
paleoenvironments. Within that fauna, Eotairis guadalupensis indicated that the crown group
echinoids evolved at the latest in the Roadian stage of the Permian, 10 million years older than
previously thought. This taxon also indicates that the GRNs underlying the perignathic girdle in
cidaroid and euechinoid echinoids, including expression of the genes VEGFR and ALX1,
diverged by at least the Roadian. These new fossils, and their corresponding description, have
consistently provided new insight into our macroevolutionary and macroecological
understanding of sea urchins.
FUTURE WORK
The results presented herein form a sound basis onto which numerous other studies of sea urchin
macroevolution will build. For instance, I have presented phylogenetic analyses of the
Bothriocidaroida and the Archaeocidaridae/Miocidaridae. These clades bracket the beginning
306
and end of the Paleozoic, respectively, and future work will aim to bridge these and provide a
comprehensive phylogeny of the stem group echinoids, and link them to the already well-known
relationships of the crown group (Kroh and Smith, 2010). Additionally, though the paleogenomic
approach herein applied was focused primarily on echinoids, novel data regarding embryonic
GRNS is becoming more widely available from other echinoderm groups, and cross-class
comparisons will soon be feasible using the methodological approaches detailed herein. The
dissertation presented herein will form a solid foundation and model for future interdisciplinary
research into sea urchins and other metazoan groups.
307
REFERENCES
Agassiz, A., 1892, Calamocrinus diomedae, a new stalked crinoid, with notes on the apical
system and the homologies of echinoderms: Memoirs of the Museum of Comparative
Zoölogy at Harvard College, v. 17, no. 2, p. 1-95.
Agassiz, A., 1904, The Panamic deep sea echini: Memoirs of the Museum of Comparative
Zoölogy at Harvard College, v. 31, p. 1-243.
Agassiz, L., and Desor, E., 1847, Catalogue Raisonné des Espéces, des Generes et des Families
D’Echinides: Annales des Sciences Naturelles, Troisiéme Série, Zoologie, v. 7, p. 129-
168.
Alroy, J., 2010, Fair sampling of taxonomic richness and unbiased estimation of origination and
extinction rates, in Alroy, J., and Hunt, G., eds., Quantitative Methods in Paleobiology,
The Paleontological Society, p. 55-80.
Alroy, J., Aberhan, M., Bottjer, D. J., Foote, M., Fürsich, F. T., Harries, P. J., Hendy, A. J.,
Holland, S. M., Ivany, L. C., and Kiessling, W., 2008, Phanerozoic trends in the global
diversity of marine invertebrates: Science, v. 321, no. 5885, p. 97-100.
Amemiya, S., and Emlet, R. B., 1992, The development and larval form of an echinothurioid
echinoid, Asthenosoma ijimai, revisited: Biological Bulletin, v. 182, p. 15-30.
Anderson, L. I., Clarkson, E. N. K., Stewart, S. E., and Mitchell, D., 2007, An Upper Llandovery
Konservat-Lagerstätte in a depositional context: the Pentland Hills Eurypterid Bed,
Midlothian: Scottish Journal of Geology, v. 43, p. 41-50.
Armstrong, H. A., and Owen, A. W., 2001, Terrane evolution of the paratectonic Caledonides of
northern Britain: Journal of the Geological Society, London, v. 158, p. 475-486.
308
Ausich, W., 1998a, Early phylogeny and subclass division of the Crinoidea (Phylum
Echinodermata): Journal of Paleontology, v. 72, no. 3, p. 499-510.
-, 1998b, Phylogeny of Arenig to Caradoc crinoids (Phylum Echinodermata) and suprageneric
classification of the Crinoidea: The University of Kansas Paleontological Contributions.
New Series, v. 9, p. 1-36.
-, 2016, Fossil Species as Data: A Perspective from Echinoderms, in Alllmon, W. D., and
Yacobucci, M. M., eds., Species and Speciation in the Fossil Record: Chicago, University
of Chicago Press, p. 301-311.
Ausich, W., and Kammer, T., 2006, Stratigraphical and geographical distribution of
Mississippian (Lower Carboniferous) Crinoidea from England and Wales: Proceedings of
the Yorkshire Geological Society, v. 56, no. 2, p. 91-109.
Ausich, W., and Peters, S. E., 2005, A revised macroevolutionary history for Ordovician-Early
Silurian crinoids: Paleobiology, v. 31, no. 3, p. 538-551.
Ausich, W. I., and Kammer, T. W., 2013, Mississippian crinoid biodiversity, biogeography and
macroevolution: Palaeontology, v. 56, no. 4, p. 727-740.
Babcock, L. C., 1977, Life in the Delaware Basin: The paleoecology of the Lamar Limestone,
Upper Guadalupian facies, Permian reef complex, Guadalupe Mountains, New Mexico
and west Texas, Permian Basin Section SEPM Publication, p. 357-389.
Baird, G. C., Gryta, G. C. J. J., McKenzie, S. C., Over, J. D., Pulawski, S., and Sullivan, J. S.,
2009, Deconvoluting the end-Devonian story in the “Oil Lands Region” of northwest
Pennsylvania. Guidebook, 74ᵗʰ annual field conference of Pennsylvanian Geologit,
Titusville, PA, in Harper, J. A., ed., History and geology of the oil regions of
northwestern Pennsylvania, p. 5-31.
309
Barras, C. G., 2008, Morphological innovation associated with the expansion of atelostomate
irregular echinoids into fine-grained sediments during the Jurassic: Palaeogeography,
Palaeoclimatology, Palaeoecology, v. 263, no. 1, p. 44-57.
Bather, F. A., 1909, On Eocidaris and some species referred to it: Annals and Magazine of
Natural History, Series 8, v. 3, p. 43-66.
-, 1931, What is Bothriocidaris?: Palaeontologische Zeitschrift, v. 13, no. 1, p. 55-60.
Beatty, T. W., Zonneveld, J., and Henderson, C. M., 2008, Anomalously diverse Early Triassic
ichnofossil assemblages in northwest Pangea: A case for a shallow-marine habitable
zone: Geology, v. 36, no. 10, p. 771-774.
Belanski, C. H., 1928, Descriptions of some typical fossils of the Shellrock stage: American
Midland Naturalist, v. 11, no. 5, p. 171-212.
Bell, M. A., and Lloyd, G. T., 2015, strap: an R package for plotting phylogenies against
stratigraphy and assessing their stratigraphic congruence: Palaeontology, v. 58, no. 2, p.
379-389.
Bennett, K. C., Young, C. M., and Emlet, R. B., 2012, Larval development and metamorphosis
of the deep-sea cidaroid urchin Cidaris blakei: Biological Bulletin, v. 222, p. 105-117.
Benton, M. J., 1994, Palaeontological data and identifying mass extinctions: Trends in Ecology
& Evolution, v. 9, p. 181-185.
-, 2003, When life nearly died: the greatest mass extinction of all time, London, Thames &
Hudson, 336 p.:
Benton, M. J., Wills, M. A., and Hitchin, R., 2000, Quality of the fossil record through time:
Nature, v. 403, p. 534-537.
310
Bigsby, J. J., 1868, Thesaurus Siluricus: The Flora and Fauna of the Silurian Period with
Addenda (from Recent Acquisitions), London, John van Voorst, 214 p.:
Bockelie, J. F., and Briskeby, 1980, The presence of a bothriocidarid (Echinoid) in the
Ordovician of Norway: Norsk Geologisk Tidsskrift, v. 1, p. 89-91.
Boos, M. F., 1929, Stratigraphy and Fauna of the Luta Limestone (Permian) of Oklahoma and
Kansas: Journal of Paleontology, v. 3, no. 3, p. 241-253.
Borszcz, T., 2012, Echinoids as substrates for encrustation- Review and quantitative analysis:
Annales Societatis Geologorum Poloniae, v. 82, p. 139-149.
Borths, M. R., and Ausich, W., 2011, Ordovician-Silurian Lilliput crinoids during the end-
Ordovician biotic crisis: Swiss Journal of Palaeontology, v. 130, p. 7-18.
Bottjer, D. J., 2017, Geobiology and palaeogenomics: Earth Science Reviews, v. 164, p. 182-
192.
Bottjer, D. J., Davidson, E. H., Peterson, K. J., and Cameron, R. A., 2006, Paleogenomics of
echinoderms: Science, v. 314, no. 5801, p. 956-960.
Bremer, K., 1994, Branch Support and Tree Stability: Cladistics, v. 10, p. 295-304.
Buch, L. v., 1842, Beiträge zur Bestimmung der Gebirgsformationen in Russland: Archiv für
Mineralogie, Geognosie, Bergbau und Hüttenkunde, v. 15, no. 1, p. 1-128.
Buggisch, W., Joachimski, M. M., Sevastopulo, G., and Morrow, J. R., 2008, Mississippian
δ13Ccarb and conodonts apatite δ18O records – Their relation to the Late Palaeozoic
Glaciation: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 268, p. 273-292.
Bull, E. E., and Loydell, D. K., 1995, Uppermost Telychian graptolites from the North Esk
Inlier, Pentland Hills, near Edinburgh: Scottish Journal of Geology, v. 31, p. 163-170.
311
Butts, S. H., and Briggs, D. E. G., 2011, Silicification Through Time, in Allison, P. A., and
Bottjer, D. J., eds., Taphonomy: Process and Bias Through Time, p. 411-434.
Carter, B., Beisel, T., Branch, W., and Mashburn, C., 1989, Substrate preferences of late Eocene
(Priabonian/Jacksonian) echinoids of the eastern Gulf Coast: Journal of Paleontology, v.
63, no. 04, p. 495-503.
Carter, B. D., and Hamza, F., 1994, Substrate preferences and biofacies distributions of Egyptian
Eocene echinoids: Palaios, p. 237-253.
Chen, Z.-Q., and Benton, M. J., 2012, The timing and pattern of biotic recovery following the
end-Permian mass extinction: Nature Geoscience, v. 5, no. 6, p. 375-383.
Cherns, L., and Wright, V. P., 2000, Missing molluscs as evidence of large-scale, early skeletal
aragonite dissolution in a Silurian sea: Geology, v. 28, p. 791-794.
-, 2009, Quantifying the impacts of early diagenetic aragonite dissolution on the fossil record:
Palaios, v. 24, p. 756-771.
Chesher, R. H., 1969, Contributionf to the biology of Meoma ventricosa (Echinoidea:
Spatangoida): Bulletin of Marine Science, v. 19, no. 1, p. 72-110.
Chesnut, D. R., and Ettensohn, F. R., 1988, Hombergian (Chesterian) echinoderm paleontology
and paleoecology, south-central Kentucky, Paleontological Research Institution.
Clapham, M. E., 2015, Ecological consequences of the Guadalupian extinction and its role in the
brachiopod-mollusk transition: Paleobiology, v. 42, p. 266-279.
Clark, H. L., 1932, The ancestry of echini: Science, v. 76, no. 1982, p. 591-593.
Clarkson, E. N. K., 2000, Pentland Odyssey: Scottish Journal of Geology, v. 36, p. 8-16.
Clarkson, E. N. K., Harper, D. A. T., and Taylor, C. M., 2001, Scottish Silurian shorelines:
Transactions of the Royal Society of Edinburgh: Earth Sciences, v. 91, p. 479-487.
312
Clarkson, E. N. K., and Taylor, C. M., 2002, The Deerhope Formation in the North Esk Inlier,
Pentland Hills, Scotland: Special Papers in Palaeontology, v. 67, p. 29-44.
-, 2007, Silurian rocks and fossils in the Pentland Hills, in Clarkson, E. N. K., Harper, D. A. T.,
Taylor, C. M., and Anderson, L. I., eds., Silurian Fossils of the Pentland Hills, Scotland:
London, The Palaeontological Association, p. 8-28.
Cooper, G. A., and Grant, R. E., 1972, Permian Brachiopods of West Texas, Part I.: Smithsonian
Contributions to Paleobiology, v. 14, p. 1-230.
Cooper, R. A., and Sadler, P. M., 2012, The Ordovician Period, in Gradstein, F. M., Ogg, J. G.,
Schmitz, M., and Ogg, G., eds., The Geologic Time Scale 2012: Amsterdam, Elsevier, p.
488-523.
David, B., Stock, S. R., De Carlo, F., Hétérier, V., and De Ridder, C., 2009, Microstructures of
Antarctic cidaroid spines: diversity of shapes and ectosymbiont attachment: Marine
Biology, v. 156, no. 8, p. 1559-1572.
Davidson, E. H., and Erwin, D. H., 2006, Gene regulatory networks and the evolution of animal
body plans: Science, v. 311, no. 5762, p. 796-800.
Davidson, E. H., Rast, J. P., Oliveri, P., Ransick, A., Calestani, C., Yuh, C.-H., Minokawa, T.,
Amore, G., Hinman, V., Arenas-Mena, C., Otim, O., Brown, C. T., Livi, C. B., Lee, P.
Y., Revilla, R., Rust, A., G., Pan, Z. j., Schilstra, M. J., Clarke, P. J. C., Arnone, M. I.,
Rowen, L., Cameron, R. A., McClay, D. R., Hood, L., and Bolouri, H., 2002a, A
genomic regulatory network for development: Science, v. 295, p. 1669-1678.
Davidson, E. H., Rast, J. P., Oliveri, P., Ransick, A., Calestani, C., Yuh, C.-H., Minokawa, T.,
Amore, G., Hinman, V., Arenas-Mena, C., Otim, O., Brown, C. T., Livi, C. B., Lee, P.
Y., Revilla, R., Schilstra, M. J., Clarke, P. J. C., Rust, A., G., Pan, Z., Arnone, M. I.,
313
Rowen, L., Cameron, R. A., McClay, D. R., Hood, L., and Bolouri, H., 2002b, A
provisional regulatory gene network for specification of endomesoderm in the sea urchin
embryo: Developmental Biology, v. 246, no. 1, p. 162-190.
de Koninck, L.-G., 1842-1844, Description des Animaux Fossiles qui se Trouvent des le Terrain
Carbonifére de Belgique, 650 p.:
-, 1863, Description of some fossils from India, discovered by Dr. A. Fleming, of Edinburgh:
The Quarterly Journal of the Geological Society of London, v. 19, p. 1-19.
-, 1882, Notice sur un échinoïde gigantesque de calcaire carbonifère de Belgique Association
française pour l’Avancement des Sciences, Compte Rendu de la 10ᵉ Session, Alger, p.
514-515.
Delabroye, A., Munnecke, A., Servais, T., Vandenbroucke, T. R. A., and Vecoli, M., 2012,
Abnormal forms of acritarchs (phytoplankton) in the upper Hirnantian (Upper
Ordovician) of Anticosti Island, Canada: Revue of Palaeobotany and Palynology, v. 173,
p. 46-56.
Deline, B., and Ausich, W., 2011, Testing the plateau: a reexamination of disparity and
morphologic constraints in early Paleozoic crinoids: Paleobiology, v. 37, no. 2, p. 214-
236.
Desor, E., 1855-1858, Synopsis des échinides fossils, Paris, Reinwald, 490 p.:
-, 1858, Synopsis des Échinides Fossiles, Paris, Chez Ch. Reinwald.
Döderlein, L., 1887, Die Japanischen Seeigel. I Thiel. Familie Cidaridae und Saleniidae,
Stuttgart, E Schweizerbartische Verlagshandlung.
Donoghue, P. C. J., 2005, Saving the stem group⎯a contradiction in terms?: Paleobiology, v. 31,
p. 553-558.
314
Donoghue, P. C. J., and Purnell, M. A., 2005, Genome duplication, extinction and vertebrate
evolution: Trends in Ecology & Evolution, v. 20, p. 314-319.
Donovan, S., 1988, The British Ordovician crinoid fauna: Lethaia, v. 21, p. 424.
-, 2001, Evolution of Caribbean echinoderms during the Cenozoic: moving towards a complete
picture using all of the fossils: Palaeogeography, Palaeoclimatology, Palaeoecology, v.
166, no. 1, p. 177-192.
Donovan, S., Paul, C. R. C., and Lewis, D. N., 1996, Echinoderms., in Harper, D. A. T., and
Owen, A. W., eds., Fossils of the Upper Ordovician. , Volume 7: London,
Palaeontological Association, p. 202-267.
Donovan, S. K., 1991, The taphonomy of echinoderms: calcareous multi-element skeletons in
the marine environment, in Donovan, S. K., ed., The processes of fossilization: London,
Belhaven Press, p. 241-269.
Donovan, S. K., Lewis, D. N., and Crabb, P., 2003, Lower Carboniferous Echinoderms of North-
West England, in The Natural History Museum, L., ed., Fold-Out Fossils, Volume 1, The
Palaeontological Association.
dos Reis, M., and Yang, Z., 2011, Approximate likelihood calculation on a phylogeny for
Bayesian estimation of divergence times: Molecular Biology and Evolution, v. 28, no. 7,
p. 2161-2172.
Droser, M. L., and Finnegan, S., 2003, The Ordovician radiation: A follow-up to the Cambrian
explosion?: Integrative and Comparative Biology, v. 43, p. 178-184.
Drummond, A. J., Ho, S. Y. W., Phillips, M. J., and Rambaut, A., 2006, Relaxed phylogenetics
and dating with confidence: PloS Biology, v. 4, no. 5, p. e88.
315
Duloquin, L., Lhomond, G., and Gache, C., 2007, Localized VEGF signaling from ectodetm to
mesenchyme cells controls morphogenesis of the sea urchin embryo skeleton:
Development, v. 134, p. 2293-2302.
Dunham, R. J., 1962, Classification of carbonate rocks according to depositional textures.
Dunn, C. W., Lou, X., and Wu, Z., 2013, Phylogenetic analysis of gene expression: Integrative
and Comparative Biology, v. 53, no. 5, p. 847-856.
Durham, J. W., 1966, Evolution among the echinoidea: Biological Reviews, v. 41, no. 3, p. 368-
391.
Durham, J. W., and Melville, R. V., 1957, A classification of echinoids: Journal of Paleontology,
v. 31, no. 1, p. 242-272.
Dylus, D. V., Czarkwiani, A., Stångberg, J., Ortega-Martinez, O., Dupont, S., and Oliveri, P.,
2016, Large-scale gene expression study in the ophiuroid Amphiura filiformis provides
insights into evolution of gene regulatory networks: Evodevo, v. 7, no. 1, p. 1.
Eckert, J. D., 1988, Late Ordovician extinction of North American and British Crinoids: Lethaia,
v. 21, p. 147-167.
Eggink, R. G., 1991, Vondstmelding van een Bothriocidaris te Westerhaar: Grondboor en
Hamer, v. januari, p. 7-9.
Eichwald, E., 1860, Lethaea Rossica ou Paléontologie de la Russia, Stuttgart, E. Schweizerbart.
Emlet, R. B., 1988, Larval form and metamorphosis of a “primitive" sea urchin, Eucidaris
thouarsi (Echinodermata: Echinoidea: Cidaroida), with implications for developmental
and phylogenetic studies: The Biological Bulletin, v. 174, no. 1, p. 4-19.
316
Erkenbrack, E. M., 2016, Divergence of ectodermal and mesodermal gene regulatory network
linkages in early development of sea urchins: Proceedings of the National Academy of
Sciences, v. 113, no. 46, p. E7202-E7211.
Erkenbrack, E. M., Ako-Asare, K., Miller, E., Tekelenburg, S., Thompson, J. R., and Romano,
L., 2016, Ancestral state reconstruction by comparative analysis of a GRN kernel
operating in echinoderms: Development genes and evolution, v. 226, no. 1, p. 37-45.
Erkenbrack, E. M., and Davidson, E. H., 2015, Evolutionary rewiring of gene regulatory network
linkages at divergence of the echinoid subclasses: Proceedings of the National Academy
of Sciences, v. 112, no. 30, p. E4075-E4084.
Erwin, D. H., 1993, The great Paleozoic crisis: life and death in the Permian, Columbia
University Press, 327 p.:
-, 1994, The Permo-Triassic extinction: Nature, v. 367, p. 231-236.
Etheridge, R., 1892, A monograph of the Permo-Carboniferous invertebrata of New South Wales
Part II.-Echinodermata, Annelida, and Crustacea: Memoirs of the Geological Survey of
New South Wales. Paleontology, v. 5, p. 1-132.
Ettensohn, C. A., 2009, Lessons from a gene regulatory network: echinoderm skeletogenesis
provides insights into evolution, plasticity and morphogenesis: Development, v. 136, no.
1, p. 11-21.
Ettensohn, C. A., Illies, M. R., Oliveri, P., and De Jong, D. L., 2003, Alx1, a member of the
Cart1/Alx3/Alx4 subfamily of Paired-class homeodomain proterins, is an essential
component of the gene regulatory network controlling skeletogenic fate specification in
the sea urchin embryo: Development, v. 130, p. 2917-2928.
317
Felsenstein, J., 1985, Confidence limits on phylogenies: An approach using the bootstrap:
Evolution, v. 39, no. 4, p. 783-791.
Ferber, I., and Lawrence, J. M., 1976, Distribution, substratum preference and burrowing
behaviour of Lovenia elongata (Gray)(Echinoidea: Spatangoida) in the Gulf of Elat
('Aqaba), Red Sea: Journal of Experimental Marine Biology and Ecology, v. 22, no. 3, p.
207-225.
Finnegan, S., Bergmann, K., Eiler, J. M., Jones, D. S., Fike, D. A., Eisenman, I., Hughes, N. C.,
Tripati, A. K., and Fischer, W. W., 2011, The magnitude and duration of Late
Ordovician-Early Silurian Glaciation: Science, v. 331, p. 903-906.
Finnegan, S., Heim, N. A., Peters, S. E., and Fischer, W. W., 2012, Climate change and the
selective signature of the Late Ordovician mass extinction: Proceedings of the National
Academy of Sciences, v. 109, no. 18, p. 6829-6834.
Fischer Von Waldheim, G., 1848, Notice sur quelques fossiles de la Russie: Bulletin de la
Société impériale des naturalistes de Moscou, v. 21, p. 237-249.
Fleming, J., 1828, History of British Animals Exhibiting the Characters and Systematical
Arrangement of the Genera and Species of Quadrupeds, Birds, Reptiles, Fishes,
Mollusca, and Radiata of the United Kingdom; Including the Indigenous, Extirpated, and
Extinct Kinds, Together with Periodical and Occasional Visitors, London, Duncan and
Malcolm, 565 p.:
Fletcher, B. N., Philip, G. M., and Wright, C. W., 1967, Chapter 22: Echinodermata:
Eleutherozoa, in Harland, W. B., Holland, C. H., House, M. R., F., H. N., Reynolds, A.
B., Rudwick, M. J. S., Satterthwaite, G. E., Tarlo, L. B. H., and Willey, E. C., eds., The
Fossil Record: London, Geological Society of London.
318
Foote, M., 1992, Paleozoic record of morphological diversity in blastozoan echinoderms:
Proceedings of the National Academy of Sciences of the United States of America, v. 89,
p. 7325-7329.
Foote, M., 2006, Substrate affinity and diversity dynamics of Paleozoic marine animals:
Paleobiology, v. 32, no. 3, p. 345-366.
-, 2014, Environmental controls on geographic range size in marine animal genera: Paleobiology,
v. 40, no. 3, p. 440-458.
François, É., and David, B., 2006, Variations morphologiques des Toxaster (Echinoida :
Spatangoida) en regard des fluctuations spatiales (Arc de Castellane, SE France) et
temporelles (Valanginien-Hauterivien) du milieu sédimentaire: expression d’un potentiel
adaptatif restreint: Geobios, v. 39, p. 355-371.
Franzén, C., 1979, Echinoderms, in Jaanusson, V., Laufeld, S., and Skoglung, R., eds., Lower
Wenlock faunal and floral dynamics - Vattenfallet section, Gotland: Uppsala, p. 216-244.
Gagnon, J.-M., and Gilkinson, K. D., 1994, Discrimination and distribution of the sea urchins
Strongylocentrotus droebachiensis (O. F. Müller) and S. pallidus (G. O. Sars) in the
northwest Atlantic: Sarsia, v. 79, no. 1, p. 1-11.
Gao, F., and Davidson, E. H., 2008, Transfer of a large gene regulatory apparatus to a new
developmental address in echinoid evolution: Proceedings of the National Academy of
Sciences, v. 105, no. 16, p. 6091-6096.
Gao, F., Thompson, J. R., Petsios, E., Erkenbrack, E., Moats, R. A., Bottjer, D. J., and Davidson,
E. H., 2015, Juvenile skeletogenesis in anciently diverged sea urchin clades:
Developmental biology, v. 400, no. 1, p. 148-158.
319
Geinitz, H. B., 1848, Die Versteinerungen des Zechsteingebirges und Rothliegenden oder des
permischen Systemes in Sachsen Dresden and Leipzig, Arnoldische Buchhandlung, 26
p.:
Gildor, T., and Ben-Tabou de Leon, S., 2015, Comparative study of regulatory circuits in two sea
urchin species reveals tight control of timing and high conservation of expression
dynamics: PloS Genetics, v. 11, no. 7, p. e1005435.
Gladfelter, W., 1978, General ecology of the cassiduloid urchin Cassidulus caribbearum: Marine
Biology, v. 47, no. 2, p. 149-160.
Godbold, A., Schoepfer, S., Shen, S., and Henderson, C. M., 2017, Precarious ephemeral refugia
during the earliest Triassic: Geology, v. 45, no. 7, p. 607-610.
Gordon, C. M., and Donovan, S. K., 1992, Disarticulated echinoid ossicles in paleoecology and
taphonomy: the last interglacial Falmouth Formation of Jamaica: Palaios, v. 7, no. 2, p.
157-166.
Gordon, I., 1927, The development of the calcareous test of Echinocardium cordatum:
Philosophical Transactions of the Royal Society of London. Series B, v. 215, p. 255-313.
-, 1928, Skeletal development in Arbacia, Echinarachnius and Leptasterias: Philosophical
Transactions of the Royal Society of London. Series B, v. 217, p. 289-334.
Gortani, M., 1905, Fossili carboniferi del M. Pizzul e de piano di Lanza nelle Alpi Carniche:
Bolletino Società Geologica Italiana, v. 24, p. 521-597.
Greenstein, B. J., 1991, An integrated study of echinoid taphonomy: Predictions for the fossil
record of four echinoid families: Palaios, v. 6, no. 6, p. 519-540.
Gregory, J. W., 1900, The Echinoidea, in Lankester, E. R., ed., A Treatise on Zoology: London,
A. & C. Black, p. 282-332.
320
Groves, J. R., and Wang, Y., 2013, Timing and size selectivity of the Guadalupian (Middle
Permian) Fusulinoidean extinction: Journal of Paleontology, v. 87, p. 183-196.
Guensburg, T. E., 1984, Echinodermata of the Middle Ordovician Lebanon Limestone, Central
Tennessee: Bulletins of American Paleontology, v. 86, no. 319, p. 1-100.
Guidetti, P., Bianchi, C. N., Chiantore, M., Schiaparelli, S., Morri, C., and Cattaneo-Vietti, R.,
2004, Living on the rocks: substrate mineralogy and the structure of subtidal rocky
substrate communities in the Mediterranean Sea: Marine Ecology Progress Series, v. 274,
p. 57-68.
Hagdorn, H., 1995, Die Seeigel des germanischen oberen Muschelkalks: Geologisch-
Paläontologische Mitteilungen Innsbruck, v. 20, p. 245-281.
-, 2017, Slipped through the bottleneck: Lazarechinus mirabeti gen. et sp. nov., a Paleozoic-like
echinoid from the Triassic Muschelkalk (late Anisian) of East France: PalZ, p. 1-16.
Hagdorn, H., and Göncüoglu, M. C., 2007, Early-Middle Triassic echinoderm remains from the
Istranca Massif, Turkey: Neues Jahrbuch für Geologie und Paläontologie-Abhandlungen,
v. 246, no. 2, p. 235-245.
Hall, J., 1858, Report on the Geological Survey of the State of Iowa: Embracing the results of
investigations made during portions of the years, 1855, 56 & 57, Geological Survey of
Iowa, 724 p.:
Haq, B. U., and Schutter, S. R., 2008, A chronology of Paleozoic sea-level changes: Science, v.
322, p. 64-68.
Harper, D. A. T., Zhan, R.-b., and Jin, J., 2015, The Great Ordovician Biodiversification Event:
Reviewing two decades of research on diversity’s big bang illustrated by mainly
brachiopod data: Palaeoworld, v. 24, p. 75-85.
321
Haude, R., and Langenstrassen, F., 1976, Rotasaccus dentifer n. g. n. sp., ein devonischer
Ophiocistioide (Echinodermata) mit “holothuroiden” Wandskleriten und “echinoidem”
Kauapparat: Paläontologische Zeitschrift, v. 50, no. 3/4, p. 130-150.
Hawkins, H. L., 1920, The morphology and evolution of the ambulacrum in the echinoidea
Holectypoida: Philosophical Transactions of the Royal Society of London. Series B,
Containing Papers of a Biological Character, v. 209, p. 377-480.
-, 1929, In defence of Bothriocidaris: Geological Magazine, v. 66, no. 2, p. 71-79.
-, 1935, Two genera of Carboniferous Echinoidea (Lepidocidaris and Hyattechinus) New to
Britain: Quarterly Journal of the Geological Society, v. 91, p. 239-249.
Hejnol, A., and Dunn, C. W., 2016, Animal Evolution: Are phyla real?: Current Biology, v. 26,
p. R408-R431.
Hejnol, A., and Lowe, C. J., 2015, Embracing the comparative approach: how robust
phylogenies and broader developmental sampling impacts the understanding of nervous
system evolution: Philosophical Transactions of the Royal Society B, v. 370, p.
20150045.
Henderson, C. M., Davydov, V. I., and Wardlaw, B. R., 2012, The Permian Period, in Gradstein,
F. M., Ogg, J. G., Schmitz, M., and Ogg, G., eds., The Geologic Time Scale 2012,
Elsevier, p. 653-679.
Hendler, G., Miller, J. E., Pawson, D. L., and Kier, P. M., 1995, Sea Stars, Sea Urchins, and
Allies: Echinoderms of Florida and the Caribbean, Washington and London.,
Smithsonian Institution Press.
322
Hill, S. K., and Lawrence, J. M., 2003, Habitats and Characteristics of the Sea Urchins
Lytechinus variegatus and Arbacia punctulata (Echinodermata) on the Florida Gulf ‐
Coast Shelf: Marine Ecology, v. 24, no. 1, p. 15-30.
Hinman, V. F., Nguyen, A. T., Cameron, R. A., and Davidson, E. H., 2003, Developmental gene
regulatory network architecture across 500 million years of echinoderm evolution:
Proceedings of the National Academy of Sciences of the United States of America, v.
100, no. 23, p. 13356-13361.
Hlebszevitsch, J. C., and Cortiñas, J. S., 2009, El Registro más antiguo de echinoidea en
Argentina y el contacto Pérmico-Jurásico inferior en la region del Río Geona, Chubut:
Revista de la Associación Geológica Argentina, v. 65, p. 479-486.
Hoare, R. D., and Sturgeon, M. T., 1976, Echinoid remains from the Pennsylvanian Vanport
Limestone (Allegheny Group), Ohio: Journal of Paleontology, v. 50, no. 1, p. 13-24.
Hollingworth, N., and Pettigrew, T., 1988, Zechstein reef fossils and their palaeoecology,
London, The Palaeontological Association, 75 p.:
Hopkins, M. J., Simpson, C., and Kiessling, W., 2014, Differential niche dynamics among major
marine invertebrate clades: Ecology letters, v. 17, no. 3, p. 314-323.
Hopkins, M. J., and Smith, A. B., 2015, Dynamic evolutionary change in post-Paleozoic
echinoids and the importance of scale when interpreting changes in rates of evolution:
Proceedings of the National Academy of Sciences, v. 112, no. 12, p. 3758-3763.
Hopkins, T. S., Thompson, L. E., Walker, J. M., and Davis, M., 2004, A study of epibiont
distribution on the spines of the cidaroid sea urchin, Eucidaris tribuloides (Lamarck,
1816) from the shallow shelf of the eastern Gulf of Mexico, in Heinzeller, T., and
Nebelsick, J. H., eds., Echinoderms: München: Proceedings of the 11th Internation
323
Echinoderm Conference, 6-10 October 2003, Munich, Germany: London, Taylor &
Francis, p. 207-211.
Howell, D. C., 2010, Statistical Methods for Psychology, Belmont, CA, Wadsworth, Cengage
Learning.
Hu, S.-x., Zhang, Q.-y., Chen, Z.-Q., Zhou, C.-y., Lü, T., Xie, T., Wen, W., Huang, J.-y., and
Benton, M. J., 2010, The Luoping biota: exceptional preservation, and new evidence on
the Triassic recovery from end-Permian mass extinction: Proceedings of the Royal
Society of London B: Biological Sciences, p. rspb20102235.
Huelsenbeck, J. P., 1994, Comparing the stratigraphic record to estimates of phylogeny:
Paleobiology, v. 40, no. 4, p. 470-483.
Hull, P. M., Darroch, S. A., and Erwin, D. H., 2015, Rarity in mass extinctions and the future of
ecosystems: Nature, v. 528, no. 7582, p. 345-351.
Ivany, L. C., Newton, C. R., and Mullins, H. T., 1994, Benthic invertebrates of a modern
carbonate ramp: a preliminary survey: Journal of Paleontology, v. 68, no. 03, p. 417-433.
Jablonski, D., 1986, Causes and consequences of mass extinctions: a comparative approach, in
Elliott, D. K., ed., Dynamics of extinction: New York, John Wiley & Sons, Inc., p. 183-
229.
-, 2002, Survival without recovery after mass extinctions: Proceedings of the National Academy
of Sciences, v. 99, no. 12, p. 8139-8144.
Jackson, R. T., 1896, Studies of Palaeechinoidea: Bulletin of the Geological Society of America,
v. 7, p. 171-254.
324
-, 1912, Phylogeny of the Echini: With a Revision of Palaeozoic Species, Boston, The Boston
Society of Natural History, Memoirs of the Boston Society of Natural History, v. 7, 491
p.:
-, 1929a, Palaeozoic Echini of Belgium: Mémoires du Musée Royal D’Histoire Naturelle de
Belgique, v. 38, p. 1-74.
-, 1929b, The status of Bothriocidaris: Bulletin of the Museum of Comparative Zoology at
Harvard College, v. 69, no. 13, p. 481-512.
Jaekel, O., 1894, Über die älteste Echininiden-Gattung Bothriocidaris unter Vorlegung eines
neuen Exemplares: Sitzungs-Bericht det Gesellschaft naturforschender Freunde zu
Berlin, v. 10, p. 243-255.
Jeffery, C. H., 1997, Dawn of echinoid nonplanktotrophy: Coordinated shifts in development
indicate environmental instability prior to the K-T boundary: Geology, v. 25, no. 11, p.
991-994.
Jell, P. A., 1983, Early Devonian echinoderms from Victoria (Rhombifera, Blastoidea and
Ophiocistioidea): Memoir of the Association of Australasian Palaeontologists, v. 1.
Kammer, T. W., and Ausich, W. I., 2006, The “Age of Crinoids”: a Mississippian biodiversity
spike coincident with widespread carbonate ramps: Palaios, v. 21, no. 3, p. 238-248.
-, 2007, Stratigraphical and geographical distribution of Mississippian (Lower Carboniferous)
Crinoidea from Scotland: Earth and Environmental Science Transactions of the Royal
Society of Edinburgh, v. 98, no. 02, p. 139-150.
Kanazawa, K., 1992, Adaptation of test shape for burrowing and locomotion in spatangoid
echinoids: Palaeontology, v. 35, no. 4, p. 733-750.
325
Kier, P. M., 1958a, New American Paleozoic Echinoids: Smithsonian Miscellaneous Collections,
v. 135, no. 9, p. 1-26.
-, 1958b, Permian echinoids from West Texas: Journal of Paleontology, v. 32, no. 5, p. 889-892.
-, 1962, Redescription of some Lower Carboniferous echinoids from Belgium: Institut royal des
Sciences naturelles de Belgique, Bulletin, v. 38, no. 5, p. 1-12.
-, 1965, Evolutionary trends in Paleozoic echinoids: Journal of Paleontology, p. 436-465.
-, 1966, Noncidaroid Paleozoic echinoids, in Moore, R. C., ed., Treatise on Invertebrate
Paleontology: Part U Echinodermata 3, Volume 1: Lawrence, Kansas, The Geological
Society of America, Inc. and The University of Kansas Press, p. U298-U312.
-, 1968a, Nortonechinus and the ancestry of the cidarid echinoids: Journal of Paleontology, p.
1163-1170.
-, 1968b, The Triassic echinoids of North America: Journal of Paleontology, v. 42, no. 4, p.
1000-1006.
-, 1973, A new Silurian echinoid genus from Scotland: Palaeontology, v. 16, no. 4, p. 651-663.
-, 1977a, The poor fossil record of the regular echinoid: Paleobiology, v. 3, no. 02, p. 168-174.
-, 1977b, Triassic echinoids, Washington, D. C., Smithsonian Institution Press, Smithsonian
Contributions to Paleobiology, v. 30, 88 p.:
-, 1982, Echinoids, in Sprinkle, J., ed., Echinoderm Faunas from the Bromide Formation (Middle
Ordovician) of Oklahoma, Volume 1, p. 310-314.
Kier, P. M., and Grant, R. E., 1965, Echinoid distribution and habits, Key Largo coral reef
preserve, Florida, Smithsonian Institution, v. 1-11.
Kier, P. M., and Lawson, M. H., 1978, Index of living and fossil echinoids 1924-1970:
Smithsonian Contributions to Paleobiology, v. 34, p. 1-182.
326
Kiessling, W., and Aberhan, M., 2007, Environmental determinants of marine benthic
biodiversity dynamics through Triassic–Jurassic time: Paleobiology, v. 33, no. 3, p. 414-
434.
Kiessling, W., Aberhan, M., Brenneis, B., and Wagner, P. J., 2007, Extinction trajectories of
benthic organisms across the Triassic–Jurassic boundary: Palaeogeography,
Palaeoclimatology, Palaeoecology, v. 244, no. 1, p. 201-222.
King, W., 1848, A catalogue of the organic remains of the Permien rocks of Northumberland and
Durham, London.
-, 1850, A monograph of the Permian fossils of England, London, Palaeontographical Society.
Kittl, E., 1904, Geologie der Umgebung von Sarajevo: Jarhbuch der k. k. geologischen
Reichsanstalt, v. 53, p. 516-748.
Klem, M. J., 1904, A revision of the Palaeozoic Palaëchinoidea, with a synopsis of all known
species: Transactions of the Academy of Science of St. Louis, v. 14, no. 1, p. 1-98.
Kolata, D. R., 1975, Middle Ordovician Echinoderms from Northern Illinois and Southern
Wisconsin: Memoir (Paleontological Society), v. 7, p. 1-74.
Kolata, D. R., Brower, J. C., and Frest, T. J., 1987, Upper Mississippi valley Champlainian and
Cincinnatian Echinoderms, in Sloan, R. E., ed., Middle and Late Ordovician
Lithostratigraphy and Biostratigraphy of the Upper Mississippi Valley: Saint Paul,
University of Minnesota, p. 179-181.
Kolata, D. R., Strimple, H. L., and Levorson, C. O., 1977, A new species of Bothriocidaris
(Echinoidea) from the Cincinnatian Maquoketa Group of Iowa: The Proceedings of the
Iowa Academy of Science, v. 84, no. 4, p. 146-149.
327
König, H., 1982, Unterpermische Seeigel aus Kreta (Griechenland): Neues Jahrbuch für
Geologie und Paläontologie Monatshefte, v. 1982, no. 3, p. 167-175.
Kroh, A., 2007, Climate changes in the Early to Middle Miocene of the Central Paratethys and
the origin of its echinoderm fauna: Palaeogeography, Palaeoclimatology, Palaeoecology,
v. 253, p. 169-207.
-, 2010, Index of living and fossil echinoids: Annalen des Naturhistorischen Museums in Wien,
Serie A, v. 112, p. 195-470.
-, 2015, Miocidaris ampla (Desor, 1858), World Echinoidea Database.
Kroh, A., and Nebelsick, J. H., 2003, Echinoid assemblages as a tool for palaeoenvironmental
reconstruction–an example from the Early Miocene of Egypt: Palaeogeography,
Palaeoclimatology, Palaeoecology, v. 201, no. 1, p. 157-177.
Kroh, A., and Smith, A. B., 2010, The phylogeny and classification of post-Palaeozoic
echinoids: Journal of Systematic Palaeontology, v. 8, no. 2, p. 147-212.
Kutscher, M., and Reich, M., 2001, Die Echiniden aus dem Silur der Insel Gotland: Greifswalder
Geowissenschaftliche Beiträge, v. 9, no. 3, p. 24-25.
-, 2004, Archaeocidarid and bothriocidarid Echinozoa from the Silurian of Gotland, Sweden, in
Heinzeller, T., and Nebelsick, J. H., eds., Echinoderms: München: Proceedings of the
11th International Echinoderm Conference, Munich, Germany, 6-10 October 2003:
London, Taylor & Francis, p. 457-458.
Lambert, J., 1899, Échinodermes: Revue critique de Paleozoologie, v. 3, p. 82-84.
-, 1900, Étude sur quelques échinides de l’infra-Lias et due Lias: Bulletin de la Société des
Sciences Historiques et Naturelles de l’Yonne, , v. 53, p. 3-57.
Lambert, J., and Thiéry, P., 1910, Essai de Nomenclature Raisonnée des Échinides, Chaumont.
328
Lambert, L. L., 2006, Taxonomic update for Reef Trail conodonts illustrated in Wilde et al.
1999, in Hinterlong, G., ed., Basinal Facies of the Uppermost Guadalupian: Applicability
to Exploration and Development Projects; Field Trip Guidebook Permian Basin Section,
SEPM Publication 2006-46: Midland, Society for Sedimentary Geology. Permian Basin
Section, p. 78-85.
Lambert, L. L., Bell Jr., G. L., Fronimos, J. A., Wardlaw, B. R., and Yisa, M. O., 2010,
Conodont biostratigraphy of a more complete Reef Trail Member section near the type
section, latest Guadalupian Series type region: Micropaleontology, v. 56, p. 233-256.
Lambert, L. L., Lehrman, D. J., and Harris, M. T., 2000, Correlation of the Road Canyon and
Cutoff Formations, West Texas, and its relevance to establishing an international Middle
Permian (Guadalupian) Series, in Wardlaw, B. R., Grant, B. R., and Rohr, D. M., eds.,
The Guadalupian Symposium. Smithsonian Contributions to the Earth Sciences 32:
Washington, Smithsonian Institution Press, p. 153-183.
Lambert, L. L., Wardlaw, B. R., Nestell, M. K., and Nestell, G. P., 2002, Latest Guadalupian
(Middle Permian) conodonts and foraminifers from West Texas: Micropaleontology, v.
48, p. 343-364.
Larget, B., and Simon, D. L., 1999, Markov Chain Monte Carlo Algorithms for the Bayesian
Analysis of Phylogenetic Trees: Molecular Biology and Evolution, v. 16, no. 6, p. 750-
759.
Lartillot, N., Lepage, T., and Blanquart, S., 2009, PhyloBayes 3: a Bayesian software package
for phylogenetic reconstruction and molecular dating: Bioinformatics, v. 25, no. 17, p.
2286-2288.
329
Lartillot, N., and Philippe, H., 2004, A Bayesian mixture model for across-site heterogeneities in
the amino-acid replacement process: Molecular Biology and Evolution, v. 21, no. 6, p.
1095-1109.
Le, H. S., Schulz, M. H., McCauley, B. S., Hinman, V. S., and Z, B.-J., 2013, Probabilistic error
correction for RNA sequencing: Nucleic Acids Research, v. 41, no. 10, p. e109.
Lee, Y., Britten, R. J., and Davidson, E. H., 1997, SM37, a skeletogenic gene of the sea urchin
embryo linked to the SM50 gene: Development, Growth & Differentiation, v. 41, p. 303-
312.
Lefebvre, B., Eble, G. J., Navarro, N., and David, B., 2006, Diversification of atypical Paleozoic
echinoderms: a quantitative survey of patterns of stylophoran disparity, diversity, and
geography: Paleobiology, v. 32, no. 3, p. 483-510.
Lefebvre, B., Sumrall, C. D., Shroat-Lewis, R. A., Reich, M., Webster, G. D., Hunter, A. W.,
Nardin, E., Rozhnov, S. V., Guensburg, T. E., Touzeau, A., Noailles, F., and Sprinkle, J.,
2013, Palaeobiogeography of Ordovician echinoderms, in Harper, D. A. T., and Servais,
T., eds., Early Palaeozoic Biogeography and Palaeogeography, Volume 38: London, The
Geological Society of London, p. 173-198.
Leske, N. G., 1778, Iacobi Theodori Klein Naturalis disposito Echinodermatum, Edita et aucta a
N. G. Leske, Leipzig, G. E. Beer, 278 p.:
Leupke, G., 1976, Archaeocidaris spines from the Permian Scherrer Formation, Southeastern
Arizona: Journal of the Arizona Academy of Science, v. 11, p. 87-88.
Levin, M., Anavy, L., Cole, A. G., Winter, E., Mostov, N., Khair, S., Senderovich, N., Kovalev,
E., Silver, D. H., Feder, M., Fernandez-Valverde, S. L., Nakanishi, N., Simmons, D.,
Simakov, O., Larsson, T., Liu, S.-Y., Jerafi-Vider, A., Yaniv, K., Ryan, J. F., Martindale,
330
M. Q., Rink, J. C., Arendt, D., Degnan, S. M., Degnan, B. M., Hashimshony, T., and
Yanai, I., 2016, The mid-developmental transition and the evolution of animal body
plans: Nature, v. 531, p. 637-641.
Lewis, D. N., and Donovan, S., 2007, A standardized method of describing fossils, using
Echinoidea as an example: Scripta Geologica, v. 134, p. 109-118.
Lewis, D. N., and Ensom, P. C., 1982, Archaeocidaris whatleyensis sp. nov. (Echinoidea) from
the Carboniferous Limestone of Somerset, and notes on echinoid phylogeny: Bulletin of
the British Museum (Natural History), Geology, v. 36, no. 2, p. 77-104.
Lewis, P. O., 2001, A likelihood approach to estimating phylogeny from discrete morphological
character data: Systematic biology, v. 50, no. 6, p. 913-925.
Linck, O., 1955, Ein bemerkenswerter Seeigel-Rest (Miocidaris pakistanensis n. sp.) aus der
Unter-Trias der Salt Range (Pakistan): Neues Jahrbuch für Geologie und Paläontologie
Monatshefte, v. 11, p. 489-495.
Linse, K., Walker, L. J., and Barnes, D. K. A., 2008, Biodiversity of echinoids and their epibions
around the Scotia Arc, Antarctica: Antarctic Science, v. 20, no. 3, p. 227-244.
Littlewood, D. T. J., and Smith, A., 1995, A combined morphological and molecular phylogeny
for sea urchins (Echinoidea: Echinodermata): Philosophical Transactions of the Royal
Society of London B: Biological Sciences, v. 347, p. 213-234.
Livingston, B. T., Killian, C. E., Wilt, F. H., Cameron, A., J., L. M., Ermolaeva, O., Sapojnikov,
V., Maglott, D. R., Buchanan, A. M., and Ettensohn, C. A., 2006, A genome-wide
analysis of biomineralization-related proteins in the sea urchin Strongylocentrotus
purpuratus: Developmental Biology, v. 300, p. 335-348.
331
Lohrer, A. M., Thrush, S. F., and Gibbs, M. M., 2004, Bioturbators enhance ecosystem function
through complex biogeochemical interactions: Nature, v. 431, no. 7012, p. 1092-1095.
Lovén, S., 1875, Études sur les Échinoidés: Kongliga Svenska Vetenskaps-Akademiens
Handlingar, v. 2, no. 7, p. 1-91.
-, 1883, On Pourtalesia a genus of echinoidea: Kongliga Svenska Vetenskaps-Akademiens
Handlingar, v. 19, no. 7, p. 1-95.
-, 1892, Echinologica: Bihang Till Kongliga Svenska Vetenskaps-Akademiens Handlingar, v. 18,
p. 1-74.
Loydell, D. K., 2005, Graptolites from the Deerhope Formation, North Esk Inlier: Scottish
Journal of Geology, v. 41, p. 189-190.
M’Coy, F., 1844, A Synopsis of the Characters of the Carboniferous Limestone Fossils of
Ireland, Dublin, M. H. Gill.
Macbride, E. W., and Spencer, W. K., 1938, Two new echinoidea, Aulechinus and Ectinechinus,
and an adult plated holothurian, Eothuria, from the Upper Ordovician of Girvan,
Scotland: Philosophical Transactions of the Royal Society of London B: Biological
Sciences, v. 229, no. 558, p. 91-136.
Makovicky, P. J., and Zanno, L. E., 2011, Theropod diversity and the refinement of avian
characteristics, in Dyke, G., and Kaiser, G., eds., Living Dinosarus: The Evolutionary
History of Modern Birds: Oxford, John Wiley & Sons, p. 9-29.
Mancosu, A., and Nebelsick, J. H., 2016, Echinoid assemblages from the early Miocene of
Funtanazza (Sardinia): A tool for reconstructing depositional environments along a shelf
gradient: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 454, p. 139-160.
332
-, 2017, Palaeoecology and taphonomy of spatangoid-dominated echinoid assemblages: A case
study from the Early-Middle Miocene of Sardinia, Italy: Palaeogeography,
Palaeoclimatology, Palaeoecology, v. 466, p. 334-352.
Männil, R. M., 1962, Taksonomiâ i morfologiâ roda Bothriocidaris (Echinoidea) [The taxonomy
and morphology of Bothriocidaris (Echinoidea)]: Eesti NSV Teaduste Akadeemia
Geoloogia Instituudi Uurimused/Trudy Instituta Geologii Akademii Nauk Èstonskoj
SSR, v. 9, p. 143-190.
Matthieu, G., 1949, Contribution a l’étude des monts troglodytes dans l’extreme Sud-Tunisien:
Geologie regionale des environs de Matmata Medenine et Foum-Tatahouine, Tunis,
Imprimerie S.A.P.I., 78 p.:
McCauley, B. S., Weideman, E. P., and Hinman, V. F., 2010, A conserved gene regulatory
network drives different developmental fates in the vegetal pole of highly divergent
echinoderm embryos: Developmental Biology, v. 340, p. 200-208.
McCauley, B. S., Wright, E. P., Exner, C., Kitazawa, C., and Hinman, V. F., 2012, Development
of an embryonic skeletogenic mesenchyme lineage in a sea cucumber reveals the
trajectory of change for the evolution of novel structures in echinoderms: EvoDevo, v. 3,
p. 17.
McClay, D. R., 2011, Evolutionary crossroads in developmental biology: sea urchins:
Development, v. 138, p. 2639-2648.
Meek, F. B., and Worthen, A. H., 1868, Remarks on some types of Carboniferous Crinoidea,
with descriptions of new genera and species of the same, and of one echinoid:
Proceedings of the Academy of Natural Sciences of Philadelphia, v. 20, p. 335-359.
333
-, 1869, Descriptions of new crinoidea and echinoidea, from the Carboniferous rocks of the
Western States, with a note on the genus Onychaster: Proceedings of the Academy of
Natural Sciences of Philadelphia, v. 21, p. 67-83.
Mickwitz, A. v., 1911, Archaikum, Kambrium, Silur., in Kupffer, K. R., ed., Baltische
Landeskunde: Riga, G. Löffler.
Mihály, S., 1980, Echinoidea maradványok a Bükk hegység felsöpermjéböl: A Magyar Állami
Földtani Intézet Évi Jelentése, v. 1978, p. 399-412.
Miller, A. I., 1997, Comparative diversification dynamics among paleocontinents during the
Ordovician radiation: Geobios, v. 20, p. 397-406.
Miller, A. I., and Connolly, S. R., 2001, Substrate affinities of higher taxa and the Ordovician
Radiation: Paleobiology, v. 27, no. 4, p. 768-778.
Miller, A. I., and Foote, M., 1996, Calibrating the Ordovician radiation of marine life:
Implications for Phanerozoic diversity trends: Paleobiology, v. 22, no. 2, p. 304-309.
Miller, A. I., and Mao, S., 1995, Association of orogenic activity with the Ordovician radiation
of marine life: Geology, v. 23, no. 4, p. 305-308.
Miller, D. J., and Eriksson, K. A., 1999, Linked sequence development and global climate
change: The Upper Mississippian record in the Appalachian basin: Geology, v. 27, no. 1,
p. 35-38.
Miller, J., and Grayson, R. F., 1972, Origin and structure of the Lower Viséan “reef” limestones
near Clitheroe, Lancashire Proceedings of the Yorkshire Geological Society, v. 38, no. 4,
p. 607-638.
334
Molyneux, S. G., Barron, H. F., and Smith, R. A., 2008, Upper Llandovery-Wenlock (Silurian)
palynology of the Pentland Hills inliers, Midland Valley of Scotland: Scottish Journal of
Geology, v. 44, no. 151-168.
Mooi, R., 1990, A new “living fossil” echinoid (Echinodermata) and the ecology and
paleobiology of Caribbean cassiduloids: Bulletin of Marine Science, v. 46, no. 3, p. 688-
700.
Mortensen, T., 1928a, Bothriocidaris and the origin of echinoids: Videnskabelige Meddeleser fra
Dansk Naturhistorisk Forening I Kiobenhavn, v. 86, p. 93-122.
-, 1928b, A Monograph of the Echinoidea. I. Cidaroidea, Copenhagen, C. A. Reitzel.
-, 1930, Bothriocidaris and the ancestry of echinoids. A reply: Videnskabelige Meddeleser fra
Dansk Naturhistorisk Forening I Kiobenhavn, v. 90, p. 313-352.
Mottequin, B., 2008, The “black marble” of Denée, a fossil conservation deposit from the Lower
Carboniferous (Viséan) of southern Belgium: Geological Journal, v. 43, p. 197-208.
Mottequin, B., Poty, E., and Prestianni, C., 2015, Catalogue of the types and illustrated
specimens recovered from the “black marble” of Denée, a marine conservation-
Lagerstätte from the Mississippian of southern Belgium: Geologica Belgica, v. 18, no. 1,
p. 1-14.
Müller, A. H., 1978, Lehrbuch der Paläozoologie: Band II Invertebraten, Tiel 3, Arthropoda 2-
Hemichordata, Jena, VEB Gustav Fischer Verlag, 748 p.:
Munnecke, A., Delabroye, A., Servais, T., Vandenbroucke, T. R. A., and Vecoli, M., 2012,
Systematic occurrences of malformed (teratological) acritarchs in the run-up of Early
Palaeozoic Palaeogeography, Palaeoclimatology, Palaeoecology, v. 367-368, p. 137-146.
335
Nardin, E., and Lefebvre, B., 2010, Unravelling extrinsic and intrinsic factors of the early
Palaeozoic diversification of blastozoan echinoderms: Palaeogeography,
Palaeoclimatology, Palaeoecology, v. 294, p. 142-160.
Nebelsick, J. H., 1992, Echinoid distribution by fragment identification in the northern Bay of
Safaga, Red Sea, Egypt: Palaios, p. 316-328.
-, 1995, Uses and limitations of actuopalaeontological investigations of echinoids: Geobios, v.
18, p. 329-336.
-, 1996, Biodiversity of shallow-water Red Sea echinoids: implications for the fossil record:
Journal of the Marine Biological Association of the United Kingdom, v. 76, no. 01, p.
185-194.
Nestler, H., 1968, Echinidenreste aus einem Öjlemyr-Gescheibe (Ordovizium, FII) von Gotland:
Zeitschrift für das Gesamtgebiet der Geologie und Mineralogie sowie der angewandten
Geophysik, v. 17, no. 10, p. 1219-1225.
Neumayr, M., 1881, Morphologische Studien über fossile Echinodermen: Sitzungsberichte der
Mathematisch-Naturwissenschaftlichen Classe der Kaiserlichen Akademie der
Wissenschaften, v. 84, no. 1, p. 143-176.
Nichols, D., 1959, Changes in the Chalk heart-urchin Micraster interpreted in relation to living
forms: Philosophical Transactions of the Royal Society of London B: Biological
Sciences, v. 242, no. 693, p. 347-437.
Nowak, M. D., Smith, A. B., Simpson, C., and Zwickl, D. J., 2013, A simple method for
estimating informative node age priors for the fossil calibration of molecular divergence
time analyses: PLoS One, v. 8, no. 6, p. e66245.
336
O’Reilly, J. E., and Donoghue, P. C. J., 2017, The efficacy of consensus tree methods for
summarizing phylogenetic relationships from a posterior sample of trees estimates from
morphological data: Systematic Biology, v. 0, no. 0, p. 1-9.
O’Reilly, J. E., Puttick, M. N., Parry, L. A., Tanner, A. R., Tarver, J. E., Fleming, J. F., Pisani,
D., and Donoghue, P. C. J., 2016, Bayesian methods outperform parsimony but at the
expense of precision in the estimation of phylogeny from discrete morphological data:
Biology Letters, v. 12, p. 20160081.
Okazaki, K., 1975, Spicule formation by isolated micromeres of the sea urchin embryo:
American Zoologist, v. 15, no. 3, p. 567-581.
Oliveri, P., Carrick, D. M., and Davidson, E. H., 2002, A regulatory gene network that directs
micromere specification in the sea urchin embryo: Developmental Biology, v. 246, p.
209-228.
Oliveri, P., Tu, Q., and Davidson, E. H., 2008, Global regulatory logic for specification of an
embryonic cell lineage: Proceedings of the National Academy of Sciences, v. 105, no. 16,
p. 5955-5962.
Pagel, M., Meade, A., and Barker, D., 2004, Bayesian estimation of ancestral character states on
phylogenies: Systematic Biology, v. 53, no. 5, p. 673-684.
Parham, J. F., Donoghue, P. C., Bell, C. J., Calway, T. D., Head, J. J., Holroyd, P. A., Inoue, J.
G., Irmis, R. B., Joyce, W. G., and Ksepka, D. T., 2011, Best practices for justifying
fossil calibrations: Systematic Biology, p. syr107.
Paul, C. R. C., 1967, New Ordovician Bothriocidaridae from Girvan and a reinterpretation of
Bothriocidaris Eichwald: Palaeontology, v. 10, no. 4, p. 525-541.
337
Pawson, D. L., and Miller, J. E., 1983, Systematics and ecology of the sea-urchin Genus
Centrostephanus (Echinodermata: Echinoidea) from the Atlantic and Eastern Pacific
Oceans: Smithsonian Contributions to the Marine Sciences, v. 20, p. 1-15.
Payne, J. L., and Clapham, M. E., 2012, End-Permian mass extinction in the oceans: an ancient
analog for the twenty-first century?: Annual Review of Earth and Planetary Sciences, v.
40, p. 89-111.
Peach, B. N., and Horne, J., 1899, The Silurian rocks of Britain, vol. 1, Scotland, Memoirs of the
Geological Survey of the UK.
Peter, I. S., and Davidson, E. H., 2009, Genomic control of patterning: The International Journal
of Developmental Biology, v. 53, p. 707-716.
-, 2011, Evolution of gene regulatory networks controlling body plan development: Cell, v. 144,
p. 970-985.
-, 2016, Implications of developmental gene regulatory networks inside and outside
developmental biology: Current Topics in Developmental Biology, v. 117, p. 237-251.
Peters, S. E., 2006, Macrostratigraphy of North America: The Journal of Geology, v. 114, p. 391-
412.
Peters, S. E., and Foote, M., 2002, Determinants of extinction in the fossil record: Nature, v. 416,
p. 420-424.
Philip, G. M., 1965, Classification of Echinoids: Journal of Paleontology, v. 39, no. 1, p. 45-62.
Pietsch, C., Ritterbush, K. A., Thompson, J. R., Petsios, E., and Bottjer, D. J., 2017, Evolutionary
models in the Early Triassic marine realm: Palaeogeography, Palaeoclimatology,
Palaeoecology.
338
Pisera, A., 1994, Echinoderms of the Mojcza limestone: Ordovician carbonate platform
ecosystem of the Holy Cross Mountains, Poland. Palaeontologia Polonica, v. 53, p. 283-
307.
Pomel, A., 1883, Classification méthodique et genera des échinides vivants & fossiles, Alger, A.
Jourdan.
Pompeckj, J. F., 1896, Die Fauna des Cambrium von Tejrovic und Skrej in Böhmen: Jahrbuch
der kaiserlich-königlichen geologischen Reichsanstalt, v. 45, p. 495-614.
Posada, D., and Crandall, K. A., 1998, Modeltest: testing the model of DNA substitution:
Bioinformatics, v. 14, no. 9, p. 817-818.
Poulin, E., and Feral, J.-P., 1995, Pattern of spatial distribution of a brood-protecting schizasterid
echinoid, Abatus cordatus, endemic to the Kerguelen Islands: Marine ecology progress
series. Oldendorf, v. 118, no. 1, p. 179-186.
Prosser, C. S., 1895, The classification of the Upper Paleozoic rocks of Central Kansas: The
Journal of Geology, v. 3, p. 764-800.
Puttick, M. N., O'Reilly, J. E., Tanner, A. R., Fleming, J. F., Clark, J., Holloway, L., Lozano-
Fernandez, J., Parry, L. A., Tarver, J. E., Pisani, D., and Donoghue, P. C. J., 2017,
Uncertain-tree: discriminating among competing approaches to the phylogenetic analysis
of phenotype data: Proceedings of the Royal Society B: Biological Sciences, v. 284, no.
1846.
Rambaut, A., Suchard, M. A., and Drummond, A. J., 2013, Tracer version 1.6 [computer
program].
Rannala, B., and Yang, Z., 2007, Inferring speciation times under an episodic molecular clock:
Systematic Biology, v. 56, no. 3, p. 453-466.
339
Rasmussen, C. M. Ø., Ullman, C. V., Jakobsen, K. G., Lindskog, A., Hansen, J., Hansen, T.,
Eriksson, M. E., Dronov, A., Frei, R., Korte, C., Nielsen, A. T., and Harper, D. A. T.,
2016, Onset of main Phanerozoic marine radiation sparked by emerging Mid Ordovician
icehouse: Scientific reports, v. 6, no. 18884.
Raup, D. M., 1968, Theoretical Morphology of Echinoid Growth: Memoir (Paleontological
Society), v. 2, p. 50-63.
Raup, D. M., and Sepkoski, J. J., 1982, Mass Extinctions in the Marine Fossil Record: Science,
v. 215, no. 4539, p. 1501-1503.
Reich, M., 2007, Linguaserra spandeli sp. nov. (Echinodermata: Ophiocistioidea) from the Late
Permian (Zechstein) of Thuringia, Germany: Annales de Paléontologie, v. 93, p. 317-330.
Reich, M., and Smith, A. B., 2009, Origins and biomechanical evolution of teeth in echinoids
and their relatives: Palaeontology, v. 52, no. 5, p. 1149-1168.
Revilla-i-Domingo, R., Oliveri, P., and Davidson, E. H., 2007, A missing link in the sea urchin
embryo gene regulatory network: hesC and hte double-negative specification of
micromeres: Proceedings of the National Academy of Sciences, v. 104, p. 12383-12388.
Rhebergen, F., 1990, Een Bothriocidaris in een Ordovicische zwerfsteen van Wilsum (B. R. D.):
Grondboor en Hamer, v. mei, p. 50-52.
Robertson, G., 1989, A palaeoenvironmental interpretation of the Silurian rocks in the Pentland
hills, near Edinburgh, Scotland: Transactions of the Royal Society of Edinburgh: Earth
Sciences, v. 80, p. 127-141.
Ronov, A. B., Khain, V. E., Balukhovsky, A. N., and Seslavinsky, K. B., 1980, Quantitative
analysis of Phanerozoic Sedimentation: Sedimentary Geology, v. 25, p. 311-325.
340
Ronquist, F., Teslenko, M., van der Mark, P., Ayres, D. L., Darling, A., Höhna, S., Larget, B.,
Liu, L., Suchard, M. A., and Huelsenbeck, J. P., 2012, MrBayes 3.2: efficient Bayesian
phylogenetic inference and model choice across a large model space: Systematic biology,
v. 61, no. 3, p. 539-542.
Rõõmusoks, A., 1970, Stratigraphy of the Viruan Series (Middle Ordovician) in Northern
Estonia, Tallinn, Valgas.
Saitoh, M., and Kanazawa, K. i., 2012, Adaptative morphology for living in shallow water
environments in spatangoid echinoids: Zoosymposia, v. 7, no. 1, p. 255-265.
Salamon, M. A., and Gorzelak, P., 2017, Paleozoic echinoderm hangovers: Waking up in the
Triassic: COMMENT: Geology, v. 45, no. 7, p. e419.
Sallan, L. C., and Coates, M. I., 2010, End-Devonian extinction and a bottleneck in the early
evolution of modern jawed vertebrates: Proceedings of the National Academy of
Sciences, v. 107, no. 22, p. 10131-10135.
Sallan, L. C., Kammer, T. W., Ausich, W., and Cook, L. A., 2011, Persistent predator-prey
dynamics revealed by mass extinction: Proceedings of the National Academy of
Sciences, v. 108, no. 20, p. 8335-8338.
Sansom, R. S., Gabbott, S. E., and Purnell, M. A., 2010, Non-random decay of chordate
characters causes bias in fossil interpretation: Nature, v. 463, p. 797-800.
Sansom, R. S., and Wills, M. A., 2013, Fossilization causes organisms to appear erroneously
primitive by distorting evolutionary trees: Scientific Reports, v. 3, p. 2545.
Saucéde, T., Mooi, R., and David, B., 2007, Phylogeny and origin of Jurassic irregular echinoids
(Echinodermata: Echinoidea): Geological Magazine, v. 144, no. 2, p. 333-359.
341
Schallreuter, R., 1989, Ordovizische Seeigel aus Geschieben: Geschiebekunde aktuell, v. 5, no.
1, p. 3-16.
Schinner, G. O., 1993, Burrowing behavior, substratum preference, and distribution of Schizaster
canaliferus (Echinoidea: Spatangoida) in the northern Adriatic Sea: Marine Ecology, v.
14, no. 2, p. 129-145.
Schlüter, N., 2016, Ecophenotypic variation and developmental instability in the Late Cretaceous
Echinoid Micraster brevis (Irregularia; Spatangoida): PLoS One, v. 11, no. 2, p.
e0148341.
Schmidt, F., 1874, Miscellanea Silurica II. Einige neue und Wenig Bekannte Baltisch-Silurische
Petrefacten: Mémoires de L’Académie Impériale des Sciences de St. Pétersbourg , VII
Série., v. 21, no. 11, p. 1-45.
-, 1885, Revision der Ostbaltischen Silurischen Trilobiten: Mémoires de L’Académie Impériale
des Sciences de St. Pétersbourg , VII Série., v. 33, no. 1, p. 1-127.
Schneider, C. L., 2003, Hitchhiking on Pennsylvanian echinoids: epibionts on Archaeocidaris:
Palaios, v. 18, no. 4-5, p. 435-444.
-, 2010, Epibionts on Late Carboniferous through Early Permian echinoid spines from Texas,
USA, in Harris, L. G., Böttger, S. A., Walker, C. W., and Lesser, M. P., eds.,
Echinoderms: Durham: Proceedings of the 12th International Echinoderm Conference, 7-
11 August 2006, Durham, New Hampshire, U.S.A.: Durham, CRC Press, p. 71-76.
Schneider, C. L., Ausich, W., and Webster, G., 2008, The importance of Echinoids in late
Paleozoic ecosystems: Echinoderm Paleobiology, v. 71.
342
Schneider, C. L., Sprinkle, J., and Ryder, D., 2005, Pennsylvanian (Late Carboniferous)
Echinoids from the Winchell Formation, North-Central Texas, USA: Journal of
Paleontology, v. 79, no. 4, p. 745-762.
Schroeder, T., 1981, Development of a “primitive” sea urchin (Eucidaris tribuloides):
Irregularities in the Hyaline layer, micromeres, and primary mesenchyme: Biological
Bulletin, v. 161, p. 141-151.
Seilacher, A., 1979, Constructional morphology of sand dollars: Paleobiology, v. 5, no. 03, p.
191-221.
Seilacher, A., Reif, W.-E., and Westphal, F., 1985, Sedimentological, ecological and temporal
patterns of fossil Lagerstätten: Philosophical Transactions of the Royal Society of
London B: Biological Sciences, v. 311, p. 5-23.
Sepkoski, J. J., 1981, A factor analytic description of the Phanerozoic marine fossil record:
Paleobiology, v. 7, no. 01, p. 36-53.
-, 1988, Alpha, beta, or gamma: Where does all the diversity go?: Paleobiology, v. 14, no. 3, p.
221-234.
-, 1993, Ten yeas in the library: New data confirm paleontological patterns: Paleobiology, v. 19,
no. 1, p. 43-51.
Servais, T., Harper, D. A. T., Li, J., Munnecke, A., Owen, A. W., and Sheehan, P. M., 2009,
Understanding the Great Ordovician Biodiversification Event (GOBE): Influences of
paleogeography, paleoclimate, or paleoecology?: GSA Today, v. 19, no. 4/5, p. 4-10.
Servais, T., Lehnert, O., Li, J., Mullins, G. L., Munnecke, A., Nützel, A., and Vecoli, M., 2008,
The Ordovician Biodiversification: revolution in the oceanic trophic chain: Lethaia, v. 41,
no. 99-109.
343
Servais, T., Owen, A. W., Harper, D. A. T., Kröger, B., and Munnecke, A., 2010, The Great
Ordovician Biodiversification Event (GOBE): The palaeoecological dimension:
Palaeogeography, Palaeoclimatology, Palaeoecology, v. 294, p. 99-119.
Sharma, T., and Ettensohn, C. A., 2010, Activation of the skeletogenic gene regulatory network
in the early sea urchin embryo: Development, v. 137, no. 7, p. 1149-1157.
Sheehan, P. M., 2001, The Late Ordovician Mass Extinction: Annual Review of Earth and
Planetary Sciences, v. 29, p. 331-369.
Shumard, B. F., and Swallow, G. C., 1858, Descriptions of New Fossils from the Coal Measures
of Missouri and Kansas: Transactions of the Academy of Science of St. Louis, v. 1, p.
198-227.
Siddall, M. E., 1996, Stratigraiphc Consistency and the Shape of Things: Systematic Biology, v.
45, no. 1, p. 111-115.
-, 1997, Stratigraphic Indices in the Balance: A Reply to Hitchin and Benton: Systematic
Biology, v. 46, no. 3, p. 569-573.
Simpson, C., and Harnik, P. G., 2009, Assessing the role of abundance in marine bivalve
extinction over the post-Paleozoic: Paleobiology, v. 35, no. 4, p. 631-647.
Simpson, R. D., 1976, Systematic Paleontology and Paleoenvironmental analysis of the Upper
Hueco Formation, Robledo and Dona Ana Mountains, Dona Ana County, New Mexico
[Master of Science in Geology: The University of Texas at El Paso, 256 p.
Smith, A., 2005, Growth and Form in Echinoids: The Evolutionary Interplay of Plate Accretion
and Plate Addition, in Briggs, D. E. G., ed., Evolving Form and Function: Fossils and
Development: New Haven, Connecticut, Yale Peabody Museum.
344
Smith, A., and Hollingworth, N., 1990, Tooth structure and phylogeny of the Upper Permian
echinoid Miocidaris keyserlingi: Proceedings of the Yorkshire Geological and
Polytechnic Society, v. 48, no. 1, p. 47-60.
Smith, A. B., 1978a, A comparative study of the life style of two Jurassic irregular echinoids:
Lethaia, v. 11, p. 57-66.
-, 1978b, A functional classification of the coronal pores of regular echinoids: Palaeontology, v.
21, no. 4, p. 759-789.
-, 1979, Peristomial tube feet and plates of regular echinoids: Zoomorphologie, v. 94, no. 1, p.
67-80.
-, 1980a, Stereom microstructure of the echinoid test: Special Papers in Palaeontology, v. 25, p.
1-85.
-, 1980b, The structure and arrangement of echinoid tubercles: Philosophical Transactions of the
Royal Society of London. Series B, Biological Sciences, p. 1-54.
-, 1980c, The structure, function, and evolution of tube feet and ambulacral pores in irregular
echinoids: Palaeontology, v. 23, no. 1, p. 39-83.
-, 1984, Echinoid palaeobiology, London, George Allen and Unwin, 190 p.:
-, 1988, Patterns of diversification and extinction in early Palaeozoic echinoderms:
Palaeontology, v. 31, no. 3, p. 799-828.
-, 1994, Trassic echinoids from Peru: Palaeontographica Abteilung A, v. 233, p. 177-202.
-, 2007, Intrinsic versus extrinsic biases in the fossil record: contrasting the fossil record of
echinoids in the Triassic and early Jurassic using sampling data, phylogenetic analysis,
and molecular clocks: Paleobiology, v. 33, no. 02, p. 310-323.
345
Smith, A. B., and Benson, R. B., 2013, Marine diversity in the geological record and its
relationship to surviving bedrock area, lithofacies diversity, and original marine shelf
area: Geology, v. 41, no. 2, p. 171-174.
Smith, A. B., Gale, A. S., and Monks, N. E., 2001, Sea-level change and rock-record bias in the
Cretaceous: a problem for extinction and biodiversity studies: Paleobiology, v. 27, no. 02,
p. 241-253.
Smith, A. B., and Jeffery, C. H., 1998, Selectivity of extinction among sea urchins at the end of
the Cretaceous period: Nature, v. 392, no. 6671, p. 69-71.
Smith, A. B., and Kroh, A., 2011, The Echinoid Directory.
Smith, A. B., and McGowan, A. J., 2011, The ties linking rock and fossil records and why they
are important for palaeobiodiversity studies: Geological Society, London, Special
Publications, v. 358, no. 1, p. 1-7.
Smith, A. B., Pisani, D., Mackenzie-Dodds, J. A., Stockley, B., Webster, B. L., and Littlewood,
D. T. J., 2006, Testing the molecular clock: molecular and paleontological estimates of
divergence times in the Echinoidea (Echinodermata): Molecular biology and evolution, v.
23, no. 10, p. 1832-1851.
Smith, A. B., and Savill, J. J., 2001, Bromidechinus, a new Ordovician echinozoan
(Echinodermata), and its bearing on the early history of echinoids: Earth and
Environmental Science Transactions of the Royal Society of Edinburgh, v. 92, no. 02, p.
137-147.
Smith, J., and Davidson, E. H., 2009, Regulative recovery in the sea urchin embryo and the
stabilizing role of fail-safe gene network wiring: Proceedings of the National Academy of
Sciences of the United States of America, v. 106, no. 43, p. 18291-18296.
346
Smith, L. B. J., and Read, J. F., 2000, Rapid onset of late Paleozoic glaciation on Gondwana:
Evidence from Upper Mississippian strata of the Midcontinent, United States: Geology,
v. 28, no. 3, p. 279-282.
Solovjev, A. N., 2009, Morphology, systematic position, and distribution of bothriocidaroid
echinoids: Paleontological Journal, v. 43, no. 11, p. 1415-1424.
Song, H., Wignall, P. B., Tong, J., and Yin, H., 2013, Two pulses of extinction during the
Permian-Triassic crisis: Nature Geoscience, v. 6, no. 1, p. 52-56.
Spreng, A. C., and Howe, W. B., 1963, Echinoid Jaws from the Mississippian and Pennsylvanian
of Missouri: Journal of Paleontology, v. 37, p. 931-938.
Sprinkle, J., 1980, An overview of the fossil record, in Broadhead, T. W., and Waters, J. A., eds.,
Echinoderms, Notes for a Short Course: Atlanta, Paleontological Society. Southeastern
Section., p. 15-26.
Sprinkle, J., and Guensburg, T. E., 2003, Major expansion of echinoderms in the early Late
Ordovician (Mohawkian, middle Caradoc) and its possible causes, in Albanesi, G. L.,
Beresi, M. S., and Peralta, S. H., eds., Ordovician from the Andes, Volume 17: San Juan,
Argentina, Instituta Superior de Correlatioen Geoloegica (INSUGEO), Serie Correlación
Geológica, p. 327-332.
Stamatakis, A., 2014, RAxML Version 8: A tool for phylogenetic analysis and post-analysis of
large phylogenies: Bioinformatics, v. 30, no. 9, p. 1312-1313.
Stanley, D. G., 1979, Paleoecology, structure, and distribution of Triassic coral buildups in
Western North America: The University of Kansas Paleontological Contributions, v. 65,
p. 1-58.
347
Stanley, D. J., and James, N. P., 1971, Distribution of Echinarachnius parma (Lamarck) and
associated fauna on Sable Island Bank, southeast Canada: Smithsonian Contributions to
the Earth Sciences, v. 6, p. 1-24.
Stanley, G. D., 1989, An Upper Triassic reefal limestone, southern Vancouver Island, B.C.:
Canadian Society of Petroleum Geologists, Memoir, v. 13, p. 766–776.
-, 1994, Upper Triassic Corals from Peru: Palaeontographica Abteilung A, v. 233, p. 75-98.
Stanley, S. M., 2016, Estimates of the magnitudes of major marine mass extinctions in earth
history: Proceedings of the National Academy of Sciences, v. 113, no. 42, p. E6325-
E6334.
Stanley, S. M., and Powell, M. G., 2003, Depressed rates of origination and extinction during the
late Paleozoic ice age: A new state for the global marine ecosystem: Geology, v. 31, no.
10, p. 877-880.
Stanley, S. M., and Yang, X., 1994, A double mass extinction at the end of the Paleozoic era:
Science, v. 226, p. 1340-1344.
Steneck, R. S., 2013, Sea urchins as drivers of shallow benthic marine community structure, in
Lawrence, J. M., ed., Sea Urchins: Biology and Ecology: London, Academic Press, p.
195-207.
Stiller, F., 2002, Echinoid test remains from the Anisian (Middle Triassic) of Qingyan, south-
western China: Münstersche Forschungen zur Geologie und Paläontologie, v. 93, p. 75-
92.
Stockley, B., Smith, A. B., Littlewood, T., Lessios, H. A., and Mackenzie ‐Dodds, J. A., 2005,
Phylogenetic relationships of spatangoid sea urchins (Echinoidea): taxon sampling
348
density and congruence between morphological and molecular estimates: Zoologica
Scripta, v. 34, no. 5, p. 447-468.
Svensen, H., Planke, S., Polozov, A. G., Schmidbauer, N., Corfu, F., Podladchikov, Y. Y., and
Jamtveit, B., 2009, Siberian gas venting and the end-Permian environmental crisis: Earth
and Planetary Science Letters, v. 277, no. 3, p. 490-500.
Swofford, D. L., 2003, PAUP*: phylogenetic analysis using parsimony, version 4.0 b10, Sinauer
Associates, Inc.
Syverson, V. J., and Baumiller, T. K., 2014, Temporal trends of predation resistance in Paleozoic
crinoid arm branching morphologies: Paleobiology, v. 40, no. 3, p. 417-427.
Telford, M., Harold, A. S., and Mooi, R., 1983, Feeding structures, behavior, and microhabitat of
Echinocyamus pusillus (Echinoidea: Clypeasteroida): Biological Bulletin, v. 165, p. 745-
757.
Telford, M., and Mooi, R., 1986, Resource partitioning by sand dollars in carbonate and siliceous
sediments: evidence from podial and particle dimensions: The Biological Bulletin, v. 171,
no. 1, p. 197-207.
Telford, M. J., Lowe, C. J., Cameron, C. B., Ortega-Martinez, O., Aronowicz, J., Oliveri, P., and
Copley, R. R., 2014, Phylogenomic analysis of echinoderm class relationships supports
Asterozoa: Proceedings of the Royal Society B: Biological Sciences, v. 281, p.
20140479.
Thomas, A. O., 1920, Echinoderms of the Iowa Devonian: Bulletin of the Geological Society of
America, v. 31, p. 211-213.
-, 1921, Echinoderms of the Iowa Devonian: Iowa Geological Survey Annual Report, v. 29, p.
385-552.
349
Thompson, J. R., and Ausich, W., 2015, Testing for escalation in Lower Mississippian camerate
crinoids: Paleobiology, v. 41, no. 1.
Thompson, J. R., and Ausich, W. I., 2016, Facies distribution and taphonomy of echinoids from
the Fort Payne Formation (late Osagean, early Viséan, Mississippian) of Kentucky:
Journal of Paleontology, v. 90, no. 2, p. 239-249.
Thompson, J. R., Crittenden, J., Schneider, C. L., and Bottjer, D. J., 2015a, Lower Pennsylvanian
(Bashkirian) echinoids from the Marble Falls Formation, San Saba, Texas, USA: Neues
Jahrbuch für Geologie und Paläontologie Abhandlungen, v. 276, no. 1, p. 79-89.
Thompson, J. R., and Denayer, J., 2017, Revision of echinoids from the Tournaisian
(Mississippian) of Belgium and the importance of disarticulated material in assessing
palaeodiversity: Geological Journal, v. 52, no. 4, p. 529-538.
Thompson, J. R., Erkenbrack, E. M., Hinman, V. F., McCauley, B. S., Petsios, E., and Bottjer, D.
J., 2017a, Paleogenomics of echinoids reveals an ancient origin for the double-negative
specification of micromeres in sea urchins: Proceedings of the National Academy of
Sciences, v. 114, no. 23, p. 5870-5877.
Thompson, J. R., Hu, S.-x., Zhang, Q.-y., Petsios, E., Cotton, L. J., Huang, J.-y., Zhou, C.-y.,
Wen, W., and Bottjer, D. J., 2018, A new stem group echinoid from the Triassic of China
leads to a revised macroevolutionary history of echinoids during the end-Permian mass
extinction: Royal Society Open Science, v. 5, p. 171548.
Thompson, J. R., Petsios, E., and Bottjer, D. J., 2017b, A diverse assemblage of Permian
echinoids (Echinodermata, Echinoidea) and implications for character evolution in early
crown group echinoids: Journal of Paleontology, v. 91, no. 4, p. 767-780.
350
Thompson, J. R., Petsios, E., Davidson, E. H., Erkenbrack, E. M., Gao, F., and Bottjer, D. J.,
2015b, Reorganization of sea urchin gene regulatory networks at least 268 million years
ago as revealed by oldest fossil cidaroid echinoid: Scientific reports, v. 5.
Thuy, B., Hagdorn, H., and Gale, A. S., 2017, Paleozoic echinoderm hangovers: Waking up in
the Triassic: Geology, p. G38909. 38901.
Torsvik, T. H., Smethurst, M. A., Meets, J. G., Van der Voo, R., McKerrow, W. S., Braisier, M.
D., Sturt, B. A., and Walderhaug, H. J., 1996, Continental break-up and collision in the
Neoproterozoic and Palaeozoic⎯ A tale of Baltica and Laurentia: Earth-Science Reviews,
v. 40, p. 229-258.
Torsvik, T. H., Smethurst, M. A., Van der Voo, R., Trench, A., Abrahamsen, N., and Halvorsen,
E., 1992, Baltica. A synopsis of Vendian-Permian palaeomagnetic data and their
palaeotectonic implications: Earth Science Reviews, v. 33, p. 133-152.
Trautschold, H., 1879, Die Kalkbrüche von Mjatschkowa. Eine Monographie des Oberen
Bergkalks: Nouveaux Mémoires de la Société Impériale des Naturalistes, v. 14, p. 1-82.
Trench, A., and Torsvik, T. H., 1991, The Loer Palaeozoic apparent polar wander path for
Baltica: Palaeomagnetic data from Silurian limestones of Gotland, Sweden: Geophysical
Journal International, v. 107, no. 373-379.
Tuya, F., Boyra, A., Sanchez-Jerez, P., Barbera, C., and Haroun, R. J., 2004, Relationship
between rocky-reef fish assemblages, the sea urchin Diadema antillarum and macroalgae
throughout the Canarian Archipelago: Marine Ecology Progress Series, v. 278, p. 157-
169.
Twitchett, R. J., 2001, Incompleteness of the Permian–Triassic fossil record: a consequence of
productivity decline?: Geological Journal, v. 36, no. 3 ‐4, p. 341-353.
351
Twitchett, R. J., and Oji, T., 2005, Early Triassic recovery of echinoderms: Comptes Rendus
Palevol, v. 4, no. 6, p. 531-542.
Vadet, A., 1991, Révision des Cidaris du lias et du Dogger Européens: Mémoires de la Société
Académique du Boulonnais, v. 10, p. 1-175.
Vandenbroucke, T. R. A., Emsbo, P., Munnecke, A., Nuns, N., Duponchel, L., Lepot, K.,
Ouijada, M., Paris, F., Servais, T., and Kiessling, W., 2015, Metal-induced malformations
in early Palaeozoic plankton are harbingers of mass extinction: Nature Communications,
v. 6, p. 1-7.
Waagen, W. H., 1885, Productus limestone fossils, Part V. Salt Range Fossils Palaeontologia
Indica, Series 13, v. 1, p. 771-834.
Walker, D. E., and Gagnon, J.-M., 2014, Locomotion and functional spine morphology of the
heart urchin Brisaster fragilis, with comparisons to B. latifrons: Journal of Marine
Biology, v. 2014, p. 1-9.
Walker, L. J., Wilkinson, B. H., and Ivany, L. C., 2002, Continental drift and Phanerozoic
carbonate accumulation in shallow-shelf and deep-marine settings: The Journal of
Geology, v. 110, no. 1, p. 75-87.
Wanner, J., 1941, Neue Beiträge sur Kenntnis der permischen Echinodermen von Timor, XV.
Echinoidea: Palaeontographica Supplement, v. 4, p. 297-314.
Wardlaw, B. R., 2000, Guadalupian conodont biostratigraphy of the Glass and Del Norte
Mountains, in Wardlaw, B. R., Grant, B. R., and Rohr, D. M., eds., The Guadalupian
Symposium. Smithsonian Contributions to the Earth Sciences 32: Washington,
Smithsonian Institution Press, p. 37-88.
352
Webster, G. D., and Jell, P. A., 1992, Permian Echinoderms from Western Australia: Memoirs of
the Queensland Museum, v. 32, p. 311-373.
Wentworth, C. K., 1922, A scale of grade and class terms for clastic sediments: The Journal of
Geology, v. 30, no. 5, p. 377-392.
Wignall, P., and Benton, M., 1999, Lazarus taxa and fossil abundance at times of biotic crisis:
Journal of the Geological Society, v. 156, no. 3, p. 453-456.
Wills, M. A., 1999, Congruence between phylogeny and stratigraphy: Randomization tests and
the gap excess ratio: Systematic Biology, v. 48, no. 3, p. 559-580.
Wills, M. A., Barrett, P. M., and Heathcote, J. F., 2008, The modified gap excess ratio (GER*)
and the stratigraphic congruence of dinosaur phylogenies: Systematic Biology, v. 57, no.
6, p. 891-904.
Witzke, B. J., and Bunker, B. J., 2005, Comments on the Mississippian stratigraphic succession
in Iowa, in Heckel, P. H., ed., Stratigraphy and Biostratigraphy of the Mississippian
Subsystem (Carboniferous System) it Its Type Region, the Mississippi River Valley of
Illinois, Missouri, and Iowa, Volume 34: Champaign, Illinois, Illinois State Geological
Survey, p. 63-75.
Wray, G. A., 1992, The evolution of larval morphology during the post-Paleozoic radiation of
echinoids: Paleobiology, v. 18, no. 03, p. 258-287.
Wray, G. A., and McClay, D. R., 1988, The origin of spicule-forming cells in a “primitive” sea
urchin (Eucidaris tribuloides) which appears to lack primary mesenchyme cells:
Development, v. 103, p. 305-315.
353
Wright, A. M., and Hillis, D. M., 2014, Bayesian analysis using a simple likelihood model
outperforms parsimony for estimation of phylogeny from discrete morphological data:
PLoS One, v. 9, no. 10, p. e109210.
Wright, A. M., Lloyd, G. T., and Hillis, D. M., 2016, Modeling character change heterogeneity
in phylogenetic analyses of morphology through the use of priors: Systematic biology, v.
65, no. 4, p. 601-611.
Wright, V. P., Cherns, L., and Hodges, P., 2003, Missing molluscs: Field testing taphonomic loss
in the Mesozoic through early large-scale aragonite dissolution: Geology, v. 31, p. 211-
214.
Wu, S., and McClay, D. R., 2007, The Snail repressor is required for PMC ingression in the sea
urchin embryo: Development, v. 134, p. 1061-1070.
Yakowlew, N., 1922, Bothriocidaris und die Abstammung der Seeigel: Zeitschrift der Deutschen
Geologischen Gesellschaft, v. 74, p. 325-330.
Yamazaki, A., Furuzawa, Y., and Yamaguchi, M., 2010, Conserved early expression patterns of
micromere specification genes in two echinoid species belonging to the orders
clypeasteroida and echinoida: Developmental Dynamics, v. 239, no. 12, p. 3391-3403.
Yamazaki, A., Ki, S., Kokubo, T., and Yamaguchi, M., 2009, Structure–function correlation of
micro1 for micromere specification in sea urchin embryos: Mechanisms of development,
v. 126, no. 8, p. 611-623.
Yamazaki, A., Kidachi, Y., Yamaguchi, M., and Minokawa, T., 2014, Larval mesenchyme cell
specification in the primitive echinoid occurs independently of the double-negative gate:
Development, v. 141, no. 13, p. 2669-2679.
354
Yamazaki, A., and Minokawa, T., 2015, Expession patterns of mesenchyme specification genes
in two distantly related echinoids, Glyptocidaris crenularis and Echinocardium cordatum:
Gene Expression Patterns, v. 17, no. 2, p. 87-97.
Yang, Z., 2007, PAML 4: Phylogenetic Analysis by Maximum Likelihood: Molecular Biology
and Evolution, v. 24, no. 8, p. 1586-1591.
-, 2014, Molecular Evolution: A Statistical Approach, Oxford, Oxford University Press, 492 p.:
Yang, Z., and Rannala, B., 2006, Bayesian estimation of species divergence times under a
molecular clock using multiple fossil calibrations with soft bounds: Molecular biology
and evolution, v. 23, no. 1, p. 212-226.
Zachos, L. G., and Sprinkle, J., 2011, Computational model of growth and development in
Paleozoic echinoids, in Elewa, A. M. T., ed., Computational Paleontology, Springer, p.
75-93.
Zittel, K. A. v., 1879, Handbuch der Palaeontologie. I. Abtheilung, Paleozoológia. Band I:
Protozoa, Coelenterata, Echinodermata und Molluscoidea. , Munich and Leipzig, R.
Oldenbourg.
Zonneveld, J.-P., 2001, Middle Triassic biostromes from the Liard Formation, British Columbia,
Canada: oldest examples from the Mesozoic of NW Pangea: Sedimentary Geology, v.
145, no. 3, p. 317-341.
Zonneveld, J.-P., Henderson, C. M., Stanley, G. D., Orchard, M. J., and Gingras, M. K., 2007,
Oldest scleractinian coral reefs on the North American craton: Upper Triassic (Carnian),
northeastern British Columbia, Canada: Palaeogeography, Palaeoclimatology,
Palaeoecology, v. 243, p. 421-450.
355
Zonneveld, J. P., Furlong, C. M., and Sanders, S. C., 2016, Triassic echinoids (Echinodermata)
from the Aksala Formation, north Lake Laberge, Yukon Territory, Canada: Papers in
Palaeontology, v. 2, no. 1, p. 87-100.
Zuidema, G., 1998, Paleozoïsche zee-egels: GEA, v. 31, no. 4, p. 118-125.
Abstract (if available)
Abstract
Understanding the patterns and processes that shape organismal biodiversity are key to understanding the evolution of life on earth. These drivers of biodiversity are both intrinsic to organisms, such as the genetic regulatory networks which direct animal development, and extrinsic like global climatic shifts and environmental change. In order to truly understand biotic diversification through evolutionary time, a multi-pronged approach, examining both these extrinsic and intrinsic factors is necessary. I herein set out to understand the patterns of, and mechanisms underlying diversification and extinction in an ideal model group: sea urchins. I focus on sea urchins from the Paleozoic and Early Mesozoic eras, which have until now been relatively understudied. I herein present a number of phylogenetic analyses of echinoid groups, spanning the Ordovician and Silurian to the Late Paleozoic and Mesozoic, and provide taxonomic descriptions for a number of important echinoid taxa which provide novel insight into the macroevolutionary history of sea urchins. The first of these analyses indicates that the initial diversification of echinoids in the Early Paleozoic appears to have been rapid, with early morphological diversification of numerous bodyplans and genera. Using a large database from museum collections, I also examine the paleoenvironmental preference of echinoids during the Carboniferous period using Bayesian and frequentist statistics. Additionally, I show that the crown group echinoids, all of the descendants of the last common ancestor of extant echinoids, appear to have evolved from within the diverse and stem group echinoid family the Archaeocidaridae. By describing a new fauna from the Permian of west Texas, I show that the archaeocidarids and the crown group echinoids are not only closely related phylogenetically, but also co-existed in some of the same environments in the Permian period. The oldest definitive crown group echinoid, Eotiaris guadalupensis, also occurs in this fauna, which indicates that many of the morphological changes differentiating the two clades of crown group echinoids, the cidaroids and euechinoids, and their associated genetic regulatory underpinning, appears to have diverged by at least 258 million years ago. The entire echinoid stem group has also long been thought to have gone extinct during the end-Permian mass extinction, 252 million years ago. Using a robust phylogenetic analysis of stem group and crown group echinoids, and a description of a new species from the Triassic of China, I show that in fact at least one lineage of stem group echinoids did survive the P-T mass extinction, and that in fact the stem group echinoid did not go extinct at the end of the Permian. Finally, by using phylogenetic comparative methods, fossil- calibrated divergence time estimation, and comparative analysis of gene regulatory networks (GRN) in extant echinoids, I show that the Double Negative Gate, a key network subcircuit specifying the micromeres of many euechinoid echinoids, is an ancient evolutionary novelty, at least as old as the Triassic period. Through these analyses, I developed a novel approach to rigorously identifying the evolutionary age of particular GRN novelties, which provides a timeline for identifying the rate of GRN evolution. This multi-disciplinary approach to understanding diversification highlights the importance of holistic analyses of diversification, including both extrinsic and intrinsic factors.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
The early Triassic recovery period: exploring ecology and evolution in marine benthic communities following the Permian-Triassic mass extinction
PDF
Ecological recovery dynamics of the benthic and pelagic fauna in response to extreme temperature events and low oxygen environments developed during the early Triassic
PDF
Paleoenvironmental and paleoecological trends leading up to the end-Triassic mass extinction event
PDF
Assessing the quality of the fossil record using a phylogenetic approach
PDF
Evolution & ecology of Mesozoic birds: a case study of the derived Hesperornithiformes and the use of morphometric data in quantifying avian paleoecology
Asset Metadata
Creator
Thompson, Jeffrey Robert
(author)
Core Title
Integrated approaches to understanding diversification through time using sea urchins as a model system
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Geological Sciences
Publication Date
08/02/2020
Defense Date
03/29/2018
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
diversification,echinoid,extinction,gene regulatory networks,OAI-PMH Harvest,Paleozoic,Permian-Triassic,sea urchin
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Bottjer, David J. (
committee chair
), Caron, David A. (
committee member
), Corsetti, Frank A. (
committee member
)
Creator Email
thompsjr@usc.edu,thompson.1983@buckeyemail.osu.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-51713
Unique identifier
UC11671711
Identifier
etd-ThompsonJe-6631.pdf (filename),usctheses-c89-51713 (legacy record id)
Legacy Identifier
etd-ThompsonJe-6631.pdf
Dmrecord
51713
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Thompson, Jeffrey Robert
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
diversification
echinoid
extinction
gene regulatory networks
Paleozoic
Permian-Triassic
sea urchin