Wi*^;W^'^i
Ay ^A'V
:<* 'M>.rf^\
fiT?o
:iWr-^^Mvili:?^^?^^?!^^^g: '
^ enz?1 \ Z (/> ? U> ? t/,SMITHS0NIAN"|NSTITUTI0N NOIiniliSNI NVINIOSHilWS S3lbVaan LIBRARIES SMITHSONIAN I
CO
I NviNiosHiit^s SBiavdan libraries Smithsonian institution NoiiniiiSNi nvinoshiiws
? CO ^ </> ^ '^ =: .- <^zCO
;^SMITHS0n7aN~'|NSTITUTI0N NOIini!iSNI~'NVINOSHilWS S3 I 8 VH 8 n~'LI B RAR I ES^SMITHSONIAN'2- r- z r-, z r- 2 'J^l^f^'-A, CD X^Jvmi^oSH-3
toCO = to ? CO ? CONVINOSHilkNS S3lbVdan LIBRARIES SMITHSONIAN INSTITUTION NlOlinillSNI NVINOSHilWS 'Z (/) Z , ^ ,^ 2 . CO z ^
;'^SMITHS0NIAN^INSTITUTI0N*^N0lini!iSNI_NVIN0SHillMs'^S3 I d VM 8 IT LI B RAR I ES*^SMITHSONlAN</i ~ <^ z \ <^ 2 *^ 5
cHO Xi^os?i^ ~ oZ ?| 2NVINOSHillMS S3IHVban LIBRARIES SMITHSONIAN INSTITUTION NOIiniliSNI NVINOSHilWS 3
01Z;^SMITHSONlAN~INSTITUTION NOIiniliSNrNVINOSHilNS S3 I H Vd 8 n~L I B RAR I ES SMITHSONIANCO 2 ,v </> z ^,^ ^ w z w g> 2
2 ^ '^ 2 (/) 2 </) * Z U\
I NVINOSHIIWS S3ldVdan LIBRARIES SMITHSONIAN INSTITUTION NOIiOiliSNI NVINOSHilWS
?
... CO 5 </> 5 ^ ^ <? ^ ^ 5 ,._,. (/5
c
S^SMITHSONlAN"'lNSTlTUTION NOIiniliSNI~'NVINOSHil/\IS S3lilVyan LIBRARIES SMITHSONIAN"2 "" z r-,2 r- 2 r-? ^ o
l~NVINOSHilWS S3IMVyan~LIBRARIES SMITHSONIAN~INSTITUTION NOIiniliSNl"NVINOSHilWSZ ^ j^ 5 ^ ^ ^- >^ 2 . CO 2 w
2 W
.
-^' Z W '^ Z W 2S SMITHSONIAN INSTITUTION NOIiniliSNI NVINOSHIIWS S3IUVdan LIBRARIES SMITHSONIAN<o ? </) ^ 2 \ *" ^^ ^ 5 t'* ^
luvyy {/> = w - w \ ? wTUTION NOIinillSNI NVINOSHilWS S3iavyan libraries SMITHSONIAN INSTITUTION NOIini
Vy a n^ J-' BRAR I ES*^SMITHSONIAN_ INSTITUTION N0linillSNI_NVIN0SHllfcMs'^S3 I M Vd 9 11^ LI B RAi
TUTION NoiiniiiSNi~NviNOSHims SBidvyaii LIBRARIES smithsonian"'institution NOIinii2 r-, 2 r- 2 r-_2
CO ? </> ? COBRARIES SMITHSONIAN INSTITUTION NOIiniliSMI NVINOSHiilMS S3ldVdan LIBRAICO ^ 2 , CO 2 CO 2 .2OCO
rUTlON "NOIinillSNI NVIN0SHllWS*^S3IHVaan LIBRARIES*^SMITHS0NIAN INSTITUTION Noiini</> ^ \ </5 ^^ ^ :; <n = CO
?| -S -J Z J Z -Jvyan libraries Smithsonian institution NoiiniiisNi nvinoshiiws S3iaviian libraZ r- 2 r- .. 2 ?~
</> = </) = CO \ ^ t/>TUTION NOIinillSNI NVINOSHlllNS S3iyvyan LIBRARIES SMITHSONIAN INSTITUTION NOIinjCO 2 ^ to 2 ....? (^ 2 .y COi ^^p i ^% i fCJi ^ ^% i '^'^vaan libraries Smithsonian institution NoiiniiiSNi NviN0SHims'^S3iyvdan libraCO z^ </> _ z: CO i;(/) CO
TUTION NOIinillSNl"'NVINOSHllWS S3 I d Vd 8 IT ""lI B RAR I ES^SMITHSONIAN'^INSTITUTION NOIiniz r^ . 2 n 2 r- 2
t/> = c/j E COvy a n_Li BRARIES Smithsonian institution NoiiniiisNi nvinoshiiiais S3iyvaan librai^ ^ 1I^^2:. CO 2 ^^5'
^v
>:TUTION *^N0liniIlSNI_NVIN0SHlllMS*^S3 I yV8 a I l^L I B R AR I ES*^SMITHSONlAN INSTITUTION '^NOIini^ CO ^ z \ ^'^ ^ 5 c'J 5 ^

QL66fiL25U41987XRept.
iiversity of California PublicationsZOOLOGYVolume 118
Phylogenetic Systematics ofIguanine LizardsA Comparative Osteological Study
by Kevin de Queiroz

PHYLOGENETIC SYSTEMATICS OF IGUANINE LIZARDSA COMPARATIVE OSTEOLOGICAL STUDY

KEPT.Phylogenetic Systematics ofIguanine Lizards/A Comparative Osteological Study
by Kevin de QueirozA Contribution from the Museum of Vertebrate Zoologyof the University of California at Berkeley
yti?%K?^*^
UNIVERSITY OF CALIFORNIA PRESSBerkeley ? Los Angeles ? London
UNIVERSITY OF CALIFORNIA PUBLICATIONS IN ZOOLOGYEditorial Board: Peter B. Moyle, James L. Patton,Donald C. Potts, David S. WoodruffVolume 118Issue Date: December 1987
UNIVERSITY OF CALIFORNIA PRESSBERKELEYAND LOS ANGELES, CALIFORNIAUNIVERSITY OF CALIFORNIA PRESS, LTD.LONDON, ENGLAND
ISBN 0-520-09730-0LIBRARY OF CONGRESS CATALOG CARD NUMBER: 87-24594
? 1987 BY THE REGENTS OF THE UNIVERSITY OF CALIFORNIAPRINTED IN THE UNITED STATES OF AMERICA
Library of Congress Cataloging-in-Publication DataDe Queiroz, Kevin.Phylogenetic systematics of iguanine lizards: a comparativeosteological study / by Kevin de Queiroz.p. cm.? (University of California publications in zoology:v. 118)Bibliography: p.ISBN 0-520-09730-0 (alk. paper)1. Iguanidae?Classification. 2. Iguanidae?Evolution.3. Iguanidae?^Anatomy. 4. Anatomy, Comparative. 5. Reptiles?Qassification. 6. Reptiles?Evolution. 7. Reptiles?Anatomy.I. Title. II. Series.QL666.L25D4 1987597.95?dc 19 87-24594CIP
Contents
Li^f ofIllustrations, viiList of Tables, xAcknowledgments, xiAbstract, xiiINTRODUCTION 1Historical Review, 1Goals of This Study, 10MATERIALS AND METHODS 1 3Specimens, 13Phylogenetic Analysis, 13Basic Taxa, 14The Problem of Variation, 14Construction of Branching Diagrams, 16IGUANINE MONOPHYLY 18COMPARATIVE SKELETAL MORPHOLOGY 2 1Skull Roof, 21Palate, 39Braincase, 44Mandible, 49Miscellaneous Head Skeleton, 59Axial Skeleton, 69Pectoral Girdle and Sternal Elements, 81Pelvic Gridle, 86Limbs, 89Osteoderms, 89NONSKELETAL MORPHOLOGY 92Arterial Circulation, 92Colic Anatomy, 93External Morphology, 94
vi Contents
SYSTEMATIC CHARACTERS 100Skeletal Characters, 100Nonskeletal Characters, 104CHARACTER POLARITIES AND THE PHYLOGENETIC INFORMATIONCONTENT OF CHARACTERS 106ANALYSIS OF PHYLOGENETIC RELATIONSHIPS 1 17PreHminary Analysis, 1 17Lower Level Analysis, 122PHYLOGENETIC CONCLUSIONS 130Preferred Hypothesis of Relationships, 130Character Evolution within Iguaninae, 130COMPARISONS WITH PREVIOUS HYPOTHESES 132DIAGNOSES OF MONOPHYLETIC GROUPS OF IGUANINES 135Iguaninae Bell 1825, 135Dipsosaurus Hallowell 1854, 141Brachylophus Wagler 1830, 143Iguanini Bell 1825, 145Ctenosaura Wxtgmonn 1828, 146Sauromalus T)\xvi\?n\ 1856, 157Amblyrhynchina, new taxon, 160Amblyrhynchus Bell 1825, 163Conolophus Fitzinger 1843, 165IguaninaBell 1825, 167Iguana Laurenti 1768, 168Odwra Harlan 1824,170Appendix I: Specimens Examined, 175Appendix II: Polarity Determination Under Uncertain OutgroupRelationships, 179Appendix III: Polarity Determinationfor Lower Level Analysis, 185Appendix IV: Polarity Reevaluation for Lower Level Analysis, 187Literature Cited, 191
List of Illustrations
FIGURES
1 . "The phylogeny and relationships of North American iguanid genera," after Mittleman(1942), 62. "Grouping and possible phylogeny of the genera of iguanids occurring in the UnitedStates," after H. M. Smith (1946), 73. "Phylogenetic relationships of the Madagascar Iguanidae and the genera of iguanineHzards," after Avery and Tanner (1971), 94. Etheridge's phylogeny of the Iguanidae, 1 15. Skull of Braehylophus vitiensis, 226. Skull and mandible of Braehylophus vitiensis, 237. Posteroventral views of iguanine premaxillae, 248. Dorsal views of the preorbital portions of iguanine skulls, 259. Dorsal views of the skulls of Cyclura cornuta and Sauromaliis obesus, 1110. Posterodorsal views of the anterior orbital regions oi Brachylophusfasciatm andConolophus pallidus, 2811. Dorsal view of the skull of Amblyrhynchus cristatus, 2912. Ventral views of iguanine frontals, 3113. Dorsal views of the parietals in an ontogenetic series of Iguana iguana, 3414. Lateral view of the skull of Ctenosaura similis, 3615. Lateral views of the posterolateral comers of iguanine skulls, 3816. Posterodorsal views of disarticulated right palatines of Iguana delicatissima andConolophus subcristatus, 4017. Posterodorsal views of the right orbits of five iguanines and Morunasaurus annularis,4118. Ventral view of the skull of Iguana delicatissima, 4319. Anterolateral views of the left orbitosphenoids in an ontogenetic series of Iguanaiguana, 4520. Ventral views of the posterior portion of the palate and anterior portion of the braincaseof Sauromalus varius and Amblyrhynchus cristatus, 4621. Ventral views of iguanine neurocrania, 4722. Lateral views of the right mandibles of Iguana delicatissima and Amblyrhynchuscristatus, 5023. Lingual views of the left mandibles of three iguanines, 5124. Lateral views of the right mandibles of Conolophus pallidus and Cyclura cornuta, 52
vu
viii List ofIllustrations
25. Lateral views of the right mandibles of Iguana delicatissima, Sauromalus obesus, andAmblyrhynchus cristatus, 5326. Lateral views of the right mandibles of Dipsosaurus dorsalis, Brachylophus vitiensis,and Iguana iguana, 5527. Medial views of the left mandibles of Iguana delicatissima and Conolophussubcristatus, 5628. Dorsal views of the posterior ends of the right mandibles in ontogenetic series ofCtenosaura hemilopha and Amblyrhynchus cristatus, 5729. Dorsal views of the posterior ends of the right mandibles in an ontogenetic series ofDipsosaurus dorsalis, 5830. Lingual views of left maxillary teeth of four iguanines and Basiliscus plumifrons, 6231. Hypothetical character phylogeny for the iguanine pterygoid tooth patch, 6532. Corneal view of the left scleral ring of Ctenosaura similis, 6733. Ventral views of the iguanine hyoid apparati, 6834. Twentieth presacral vertebra of Brachylophus vitiensis, 7035. Lateral views of the twentieth presacral vertebrae of Sauromalus obesus andCtenosaura pectinata, 1136. Dorsolateral views of the twentieth presacral vertebrae of Dipsosaurus dorsalis andSauromalus obesus, 7337. Dorsal views of caudal vertebrae of Dipsosaurus dorsalis from different regions of thetail, 7638. Lateral views of the ninth caudal vertebrae of Dipsosaurus dorsalis and Iguana iguana,7939. Presacral and sacral vertebrae and ribs of Dipsosaurus dorsalis in ventral view, 8040. Pectoral girdles of three iguanines, 8241. Dorsal views of the pelvic girdles of Sauromalus obesus and Ctenosaura pectinata, 8642. Bones of the anterior limb of Brachylophus fasciatus, 8743. Right hind limb skeleton of Brachylophusfasciatus, 8844. Right tarsal region of Brachylophusfasciatus, 9045. Anterodorsal views of pedal digit II of three iguanines, 9746. Minimum-step cladograms for eight basic taxa of iguanines resulting from apreliminary analysis of 29 characters, 11947. Alternative interpretations of character transformation for homoplastic characters on aminimum-step cladogram, 12148. Alternative interpretations of character transformation for homoplastic characters on aminimum-step cladogram, 12249. Minimum-step cladograms resulting from an analysis of 26 characters in a subset ofiguanines, 12750. Consensus cladogram for the three cladograms illustrated in Figure 49, 12851. Phylogenetic relationships within Iguaninae, according to the present study, 13152. Geographic distribution of Di/?^o^aMrM5, 14153. Geographic distribution of firacA}'/<9/p/zM5', 144
List ofIllustrations ix
54. Geographic distribution of CreAio5flMra, 14755. Cladogram illustrating phylogenetic relationships within Ctenosaura, 15456. Geographic distribution of Sawroma/t^, 15857. Geographic distribution of Amblyrhynchina {Amblyrhynchus and Conolophus), 16158. Geographic distribution of /^Mana, 16959. Geographic distribution of C}'c/Mra, 17160. All nine possible fully resolved cladogram topologies for four unspecified outgroupsand an ingroup, 17961. Dendrograms corresponding with the nine cladograms in Figure 60 after each isrerooted at the outgroup node, 18062. Examples of polarity inferences for different arrangements of outgroup character statedistributions, 18263. All possible cladogram topologies for two unspecified outgroups and an ingroupbefore and after rerooting at the outgroup node, 18564. All possible cladogram topol9gies for two unspecified near outgroups, one moreremote outgroup, and an ingroup before and after rerooting at the outgroup node, 186PLATE
1. Lateral and dorsal views of the skull oi Amblyrhynchus cristatus, 91
List of Tables
1 . The iguanine genera, 22. Position of the parietal foramen, 323. Numbers of premaxillary teeth, 604. Numbers of presacral vertebrae, 7 15. Distributions of character states of 95 characters among four outgroups to iguaninesand the polarities that can be inferred from them, 1086. Distributions of character states of 95 characters among eight iguanine taxa, 1127. Distributions of character states of 29 characters used in the preliminary analysis, 1188. Polarity inferences for lower-level analysis, using Brachylophus and Dipsosaurus asoutgroups, 1249. Distributions of character states of 26 characters among six taxa within Iguanini, 12510. Distributions of character states of 19 characters among basic taxa within Ctenosaura(in the broad sense) and three close and two more distant outgroups, 15311. Summary of polarity inferences for seven cases of character-state distribution amongfour outgroups of uncertain relationships to the ingroup, 18112. Summary of polarity inferences for four cases of character-state distribution amongtwo outgroups of uncertain relationships to the ingroup, 1 8513. Summary of polarity inferences for six cases of character-state distribution among twonear outgroups whose precise relationships to the ingroup are unresolved, and onemore remote outgroup exhibiting a fixed character state, 1 87
Acknowledgments
Many people have helped me toward the completion of this study in ways big and small.Over the years I have undoubtedly forgotten the contributions of some of them, and Iapologize for this. Of those I have not forgotten, I want to thank the following people forlending me specimens under their care: Pere Alberch, Walter Auffenberg, James Berrian,Robert Bezy, Steven Busack, Joseph Collins, Ronald Crombie, Mark Dodero, RobertDrewes, William Duellman, Anne Fetzer, George Foley, Harry Greene, L. Lee Grismer,W. Ronald Heyer, J. Howard Hutchinson, Charles Meyers, Peter Meylan, Mark Norell,Gregory Pregill, Jose Rosado, Albert Schwartz, Jens Vindum, Van Wallach, John Wright,George Zug, Richard Zweifel, and especially Jay Savage and Richard Etheridge whosecollections provided the majority of the specimens examined in this study.I am also grateful to various teachers, friends, and colleagues who helped my ideas onsystematics and iguanine biology unfold through countless discussions: Troy Baird, AaronBauer, Theodore Cohn, Michael Donoghue, Richard Estes, Richard Etheridge, JacquesGauthier, Eric Gold, David Good, George Gorman, Scott Lacour, Eric Lichtwardt, JamesMelli, Sheldon Newberger, Mark Norell, Michael Novacek, David Wake, and AndreWyss. Linda Condon-Howe, Charles Crumly, Sanae and John Moorehead, DouglasPreston, Doris Taylor, and the late Kenneth Miyata generously provided lodging while Iwas visiting museums. Richard Estes, Richard Etheridge, Darrel Frost, Gregory Pregill,David Wake, and Edward Warren, provided valuable comments on earlier versions of themanuscript. David Cannatella and Rose Anne White gready assisted in the preparation ofcamera-ready-copy.Finally, I want to give special thanks to Karen Sitton for providing emotional supportin her unique and charming way and to Richard Etheridge and Richard Estes for theirinfluence on both my academic and personal development.This study partially fulfilled the requirements of a Master's degree in Zoology at SanDiego State University, but was completed at the University of California, Berkeley. Theresearch and preparation of the manuscript were supported in part by a grants from theSociety of Sigma Xi, the San Diego State University Department of Zoology, the TheodoreRoosevelt Memorial Fund of the American Museum of Natural History, and the GraduateStudent Research Allocation Fund of the Department of Zoology, University of Californiaat Berkeley.
XI
Abstract
Iguaninae is a monophyletic taxon of tetrapodous squamates (lizards) that can bedistinguished from other iguanians by at least five synapomorphies. Skeletal variationwithin Iguaninae is described and forms the basis of systematic characters used todetermine phylogenetic relationships among eight basic taxa, the currendy recognizediguanine genera. Evolutionary character polarities are determined by comparison with fourclosely related taxa, basiliscines, crotaphytines, morunasaurs, and oplurines.The distributions of derived characters among iguanine taxa suggest that: (1) EitherBrachylophus or Dipsosaurus is the sister group of the remaining iguanines (Iguanini). (2)Dipsosaurus is a monophyletic taxon diagnosed by at least six synapomorphies. (3)Brachylophus is a monophyletic taxon diagnosed by at least eight synapomorphies. (4)Iguanini, containing Amblyrhynchus, Conolophus, Ctenosaura, Cyclura, Iguana, andSauromalus, is a new monophyletic taxon diagnosed by at least three synapomorphies. (5)vWithin Iguanini, the relationships among four t2Lxa.-Ctenosaura, Sauromalus,Amblyrhynchina, and Iguanina-are unresolved. (6) Ctenosaura is a monophyletic taxon
'diagnosed by at least three synapomorphies. (7) Enyaliosaurus is monophyletic, but it is asubgroup of Ctenosaura rather than a separate taxon. If Enyaliosaurus is separated fromCtenosaura, then Ctenosaura is not monophyletic. (8) Sauromalus is a monophyletic taxondiagnosed by at least 24 synapomorphies, many of which are convergent inAmblyrhynchus. (9) Amblyrhynchina is a new monophyletic taxon containing theGalapagos iguanas Amblyrhynchus and Conolophus, and is diagnosed by at least 1 1synapomorphies. (10) Amblyrhynchus is a monophyletic taxon diagnosed by at least 28synapomorphies and is perhaps the most divergent iguanine from the most recent commonancestor of all of them. Many of the unique features of Amblyrhynchus appear to berelated to its unique natural history. (11) Conolophus is a monophyletic taxon diagnosedby at least eight synapomorphies and cannot, therefore, be considered ancestral toAmblyrhynchus. (12) Iguanina is a new monophyletic taxon composed oi Iguana andCyclura and is diagnosed by at least three synapomorphies. (13) Iguana is a monophyletictaxon diagnosed by at least seven synapomorphies. (14) Monophyly of Cyclura is aproblem in need of further study. Although three ostensible synapomorphies supportmonophyly of Cyclura, other derived characters suggest that some Cyclura shared a morerecent common ancestor with Iguana than with other Cyclura.Summaries of Iguaninae and its monophyletic subgroups down to the level of the eightbasic taxa are provided; each summary includes the type of the taxon, etymology of thetaxon name, a phylogenetic definition, geographic distribution, a list of diagnosticsynapomorphies, the fossil record, and various comments.
xu
INTRODUCTION
Containing approximately 55 genera and more than 600 species, Iguanidae is one of thelargest families of lizards. Its members occur primarily in the New World, from southernCanada to austral South America including the Galapagos Archipelago and much of theWest Indies. Iguanids also occur on the island of Madagascar and in the ComoresArchipelago in the western Indian Ocean, and on the Fiji and Tonga island groups in thesouthwestem Pacific.For over 100 years, systematists have attempted to discover the pattern ofinterrelationships among the genera in the family Iguanidae, but, because of thebewildering morphological diversity within this family, the task is far from complete.Nevertheless, many systematists have recognized suprageneric groups of iguanids (e.g.,Wagler, 1830; Dumeril and Bibron, 1837; Fitzinger, 1843; Gray, 1845; Cope, 1886, 1900;Boulenger, 1890; H. M. Smith, 1946; Savage, 1958; Etheridge, 1959, 1964a). One of theearliest of these suprageneric groups to be recognized consists of the genera currentlyknown informally as iguanines. This assemblage is also one of the most readily diagnosedon the basis of uniquely derived features. As currently conceived, there are eight generaand 31 species of iguanines (Etheridge, 1982). The iguanine genera are listed in Table 1,which also gives the number of included species, their habits, and the geographicdistribution for each genus. HISTORICAL REVIEWThe concept of an iguanine group is remarkably old, predating the publication of Darwin'sOrigin ofSpecies (1859). This accomplishment is even more surprising when one realizesthat all iguanines are native to regions far from western Europe, where systematists weredeveloping the concept of an iguanine group. These systematists undoubtedly had fewspecimens at hand, and must have relied heavily on each others' character descriptions.Although I have been unable to see all of the potentially relevant literature, I attempt to traceand summarize the history of iguanine higher systematics.The Eighteenth Century. Although the eighteenth century was an important one forbiological systematics as a whole, it was not so important for iguanine systematics. Aconvenient date to begin a historical discussion of iguanine systematics is 1758, whenLinnaeus published the tenth edition of his Systema Naturae, the starting point ofzoological nomenclature. Linnaeus himself was neither interested in nor fond of the
"lower" tetrapods. He placed all tetrapodous squamates in two genera, one of which
1
University of California Publications in Zoology
TABLE 1. The Iguanine Genera
Genus(common name) Number ofSpecies Habits GeograpiiicDistribution
Amblyrhynchus Bell 1825(Marine Iguanas)
BrachylophusWagler 1830(Banded Iguanas)Conolophus Fi\zingcT 1843(Galapagos Land Iguanas)Clenosaura Wiegmann 1828(Spiny-tailed Iguanas)
1
Phylogenetic Systematics oflguanine Lizards
contained Lacerta iguana (=Iguana iguana), the single known iguanine, and animals nowplaced in at least 12 different families, including crocodilians and amphibians. Heconsidered them to be "foul and loathsome animals" (Linnaeus, 1758, translated in Goin et
al., 1978). At the close of the eighteenth century only three of the currently recognizediguanine species (now placed in two genera) had been described, giving the systematists ofthat century, such as Laurenti (1768) and Lacepede (1788), Uttle of a group to recognize.The Nineteenth Century. Major advances in iguanine systematics came during thenineteenth century. Many important natural histories and systems or classifications ofsquamates appeared during these years, and by 1856 all of the currently recognizediguanine genera had been described.The concept of a natural iguanine taxon emerged during the first half of the nineteenthcentury. Most of the authors of classifications published during this period recognized aclose relationship among at least some of the iguanine genera. Those that did not recognizea complete and exclusive group for the iguanines known at the time failed to do so for oneor both of two reasons. Brongniart (1805), Latreille (1825), Fitzinger (1826, 1843),Wagler (1830), and Dumeril and Bibron (1837) grouped all the known iguanines together,but included some noniguanines with them. Although all the iguanines were sometimesplaced together as part of a continuous list, it is not evident that they were considered toform their own subgroup within some larger group. Other authors such as Daudin (1805),Merrem (1820), Cuvier (1829, 1831), and Wagler (1830) failed to place all iguanines in asingle group. Daudin, Cuvier, and Wagler included Brachylophus with the agamids, whileMerrem did the same for Ctenosaura.At least three authors can truly be said to have recognized an iguanine group before1850. I have two criteria for determining the true recognition of an iguanine group. First,all of the iguanine taxa known to the author (or at least all those listed in the classification)were included in the group; and second, no other taxa were included. Cuvier's (1817)
"Les Iguanes proprement dits" consisted of what are now Iguana iguana, I. delicatissima,Cyclura cornuta, and Brachylophusfasciatus, although he later removed Brachylophus andplaced it among the agamids (Cuvier, 1829, 1831). Wiegmann (1834) placed only thegenera Cyclura, Ctenosaura, Iguana, Brachylophus, and Amblyrhynchus in his familyDendrobatae, Tribus II, b, ***, B. Like many of his contemporaries, Wiegmannconstructed his classification as a hierarchy of sets and subsets that would also function asa key.The most fully developed early concept of an iguanine group appears to have been thatof Gray (1831a, 1845). In 1831, Gray placed all known iguanines (equivalent to what arenow 10 species in five genera) by themselves in a single genus, Iguana. Fourteen yearslater, he recognized nine different iguanine genera. Because these nine genera (againequivalent to five modem genera) formed one entire set in his hierarchical classification, itis evident that Gray still recognized the unity of the iguanine group.Progress in iguanine systematics, though less rapid than in the previous fifty years,continued through the second half of the nineteenth century. The last two iguanine generathat are still recognized, Dipsosaurus and Sauromalus, were described, but at first they
University of California Publications in Zoology
were not explicitly included with the rest of the iguanines in an exclusive group. Theconcept of an iguanine group, exclusive of Dipsosaurus and Sauromalus, was refined withmore detailed anatomical descriptions. Beginning with Boulenger's (1885) monumentalCatalogue of the Lizards in the British Museum, I undertake here a more detailedchronological treatment of the history of iguanine higher systematics.Boulenger (1885) listed all of the genera that are now called iguanines in a nearlycontinuous sequence in his catalogue, reflecting their position in his key as those iguanidshaving femoral pores and the fourth toe longer than the third but lacking spines on the headand an enlarged occipital scale. Nevertheless, the distantly related Hoplocercus (Etheridgein Paull et al., 1976) breaks the continuity of the iguanines in the list, and, in terms ofBoulenger's characters, some iguanines are closer to certain non-iguanine iguanids than toother iguanines. Boulenger did not explicitly delimit subgroups within Iguanidae or anyother family, and we can only guess about his precise ideas concerning such relationships.Cope (1886) appears to have been the first to use the name Iguaninae as a formal taxonfor iguanine lizards. He further provided characters, both external and skeletal, by whichmembers of this group could be distinguished from other iguanids. Cope's Iguaninaeincluded Cyclura, Ctenosaura, Cachryx, Brachylophus, Iguana, Conolophus, andAmblyrhynchus, but failed to include Dipsosaurus and Sauromalus. The generaAloponotus and Metopoceros were synonymized with Cyclura.In response to Cope, Boulenger (1890) provided what he considered to be osteologicalevidence for the separation of Metopoceros and Cyclura, and briefly described the skulls of
"the iguanoid lizards allied to Iguana." Except for the recognition of Metopoceros and theomission of Cachryx, the genera included in this discussion were the same as Cope's(1886) Iguaninae. Dipsosaurus and Sauromalus were again left out of the group.Cope later (1900) greatly expanded his Iguaninae, and named two additional iguanidsubfamilies, Anolinae and Basiliscinae. This new Iguaninae was a catch-all group forthose iguanids that lacked midventrally continuous postxiphistemal inscriptional ribs, hadsimple clavicles, and lacked a left hepatopulmonary mesentery?in other words, thoseiguanids that lacked the distinctive features of anolines and basiliscines. Although this newIguaninae was almost certainly an unnatural group, Cope recognized a slightly expandedversion of his earlier (1886) Iguaninae as a discrete subset of his new and more inclusivegroup of the same name. This unnamed subset was characterized by the presence offemoral pores and of vertebrae with zygosphenal articulations. It contained Dipsosaurusand Sauromalus along with the genera included in his earlier Iguaninae; and it is thereforeidentical in generic content to the iguanine group as currently conceived.The Twentieth Century. During the first three-fourths of the twentieth century, theconcept of an iguanine group underwent considerable change. The efforts of nineteenth-century authors such as Cope and Boulenger seem to have been largely ignored, and atleast two authors envisioned the ancestry of most other North American iguanids withiniguanines. This idea seems to have resulted from the misconception that iguanines were
"primitive" iguanids and were, therefore, potential ancestors of other iguanid taxa; theintegrity of the group was deemphasized or completely overlooked. Nevertheless, by the
Phylogenetic Systematics oflguanine Lizards
mid-1960's the iguanines had been resurrected as a natural group, the same group thatCope (1900) had recognized at the turn of the century.In his landmark paper on squamate systematics. Classification of the Lizards, Camp(1923) dealt primarily with the interrelationships of the lizard families. Nevertheless, histreatise contains scattered but intriguing comments on relationships at lower taxonomiclevels. About the throat musculature of iguanines, he said:
In the "Cyclura group" comprising the genera Iguana, Amblyrhynchus,Ctenosaura, Brachylophus, Sauromalus, and Cyclura, the superficial bundle [of theM. mylohyoideus anterior] is very specialized and consists of definitely directedfibers not connected with the skin. Detailed resemblances are present in this groupwhich I have outlined in manuscript and which will not be repeated here. Suffice itto say that the group appears to be a natural one, on the basis of the musculaturewith close resemblances prevalent between Sauromalus and Cyclura, andCtenosaura and Brachylophus. (Camp, 1923:371)
Unfortunately, the whereabouts of the manuscript mentioned in this passage are unknowntome.Mittleman (1942) reviewed the genus Urosaurus and commented briefly on therelationships among the genera of North American iguanids, except Anolis. He impliedthat the North American iguanids formed a monophyletic group descended fromCtenosaura (Fig. 1) and that the similarities among Ctenosaura, Dipsosaurus, andSauromalus were retained primitive features:Dipsosaurus is probably the most primitive of the North American Iguanidae(excepting Ctenosaura, which is properly a Central and South American form), andpossesses several points in common with Ctenosaura, most easily observed ofwhich is the dorsal crest; the genera further show their relationship in the similarityof the cephalic scutellation which is essentially simple, and shows no particulardegree of differentiation. Sauromalus is considered a specialized offshoot of theCrotaphytus, or more properly, prQ-Crotaphytus stock, by reason of its solidsternum, as well as the five-lobed teeth; the simple type of cephalic scalationindicates its affinity with the more primitive Dipsosaurus-Ctenosaura stock.(Mittleman, 1942:112-113)H. M. Smith (1946:92) seemed to adopt a modified version of Mittleman's views onthe phylogeny of North American iguanids (Fig. 2). His herbivore section {Ctenosaura,Dipsosaurus, and Sauromalus) was considered to be ancestral to the other North AmericanIguanidae, save Anolis, with Sauromalus hypothesized to share a more recent commonancestry with these other iguanids than with either Ctenosaura or Dipsosaurus. Smith'ssubsequent comments (1946:101), however, indicate that he recognized affinities ofCtenosaura, Dipsosaurus, and Sauromalus to iguanids occurring outside of the United
University of California Publications in Zoology
StreptosaurusPetrosaurus -^Crotaphytus
CallisaurusUmaHolbrookia
Uta ,,y Urosaurus\ /"Sceloporus\Sator
Phrynosoma
Dipsosaurus
Ctenosaura
Primitive Iguanid Type
FIG. 1. "The phylogeny and relationships of North American iguanid genera," after Mittleman(1942:113).
States. In addition to the three genera found in or near the United States, Smith's herbivoresection contained other "large, primitive iguanids," namely Amblyrhynchus, Conolophus,Cyclura, and Iguana. Smith's Handbook dealt with the lizards of the United States andCanada; those iguanines whose ranges did not enter this area were apparently omitted fromhis phylogram for convenience. In any case. Smith could not have considered hisherbivore section to be monophyletic in the more restricted modem sense, since the groupwas considered to be ancestral to other North American iguanids.Savage (1958) explicitly challenged Mittleman's (1942) implication that the NorthAmerican iguanids formed a natural group:
Insofar as can be determined at this time, the so-called Nearctic iguanids form twodiverse groups that can only be distantly related. These two sections are
Phylogenetic Systematics oflguanine Lizards
FIG. 2. "Grouping and possible phylogeny of the genera of iguanids occurring in the United States,"after H. M. Smith (1946:92). Roman numerals apparently refer to the following: (I) leaf-toed section, (II)herbivore section, (III) sand-lizard section, (IV) rock-lizard section, (V) pored utiform section, (V) horned-lizard section, and (VII) poreless utiform section.
distinguished by marked differences in vertebral and nasal structures and includeseveral genera not usually recognized as being allied to Nearctic forms. (Savage,1958:48)
Savage's "iguanine line" contained Amblyrhynchus, Brachylophus, Conolophus,Crotaphytus, Ctenosaura, Cyclura, Dipsosaurus, Enyaliosaurus {=Ctenosaiira, part),Iguana, and Sauromalus. This group was distinguished from the "sceloporine line" by twoprimary characters: the presence of accessory vertebral articulations, the zygosphenes andzygantra, and the possession of a relatively simple, S-shaped nasal passage with a conchapresent (Dipsosaurus-lypt of Stebbins, 1948). Other osteological and integumentaryfeatures characteristic of the majority of the genera in each line were also given.
8 University of California Publications in Zoology
The currently recognized iguanine group is based on the work of Etheridge. In hispaper on the systematic relationships of sceloporine lizards, Etheridge (1964a) showed thatthe two primary characters used by Savage (1958) to diagnose the iguanines were actuallymore widespread within the Iguanidae, and were thus insufficient to diagnose the group.He listed four fundamental differences between Crotaphytus and Savage's other iguanines,and asserted that if Crotaphytus was considered to be an iguanine, no character orcombination of characters could be used to diagnose that group. Once he removedCrotaphytus from the group, the iguanines were readily diagnosed by their unique caudalvertebrae. Except for his recognition of Enyaliosaurus as a genus separate fromCtenosaura, Etheridge's (1964a) concept of the iguanines is identical to that held today(Etheridge, 1982).Despite the long history of iguanines as a recognized group and the great interest inmany aspects of iguanine biology (e.g., Burghardt and Rand, 1982; Troyer, 1983), theinterrelationships among the iguanine genera and the relationships of iguanines to otheriguanians remain largely unknown. Commonly held beliefs are that Ctenosaura andCyclura are closely related (Barbour and Noble, 1916; Bailey, 1928; Schwartz and Carey,1977), and that the same is true of the Galapagos iguanas Amblyrhynchus and Conolophus(Heller, 1903; Eibl-Eibesfeldt, 1961; Thornton, 1971; Higgins, 1978). As mentionedabove, Mittleman (1942) and H. M. Smith (1946) have offered dendrograms depictingtheir views on the relationships of the North American iguanines.Recent studies have examined diverse data for clues about the interrelationships amongthe iguanine genera, but have met with limited success. Zug (1971) studied the arterialsystem of iguanids. He published shortest-connection networks for more than 40 iguanidgenera, some based on his arterial characters and others based on characters obtained fromthe literature, most of which were osteological. Other shortest-connection networksconstructed from data on arterial variation within various suprageneric assemblages ofiguanids, including iguanines, were also presented. Nevertheless, Zug doubted theusefulness of his arterial characters in iguanid systematics, stating: "The arterial charactersemployed herein appear to be of minimal value in iguanid classification. At the intrafamiliallevel, they are disruptive and form groups of questionable zoogeographic unity" (Zug,1971:21).There has been but a single study in which the relationships among all known iguaninegenera were sought, that of Avery and Tanner (1971). These authors provideddescriptions of the iguanine skeleton, head and neck musculature, tongue, and hemipenes,and gave a number of osteological measurements. They based their hypothesis ofrelationships on mean length-width ratios of bones, assuming that "a difference of forty orless points between means of the same bone indicates a close relationship" (Avery andTanner, 1971:67). Large numbers of such similarities were taken to indicate closephylogenetic relationship among taxa and were used in some unspecified way to construct aphylogenetic diagram (Fig. 3). Avery and Tanner examined small series (never more thanfive individuals of a single species), giving no consideration to allometric changes in theratios that they used. I suspect that many of these ratios are correlated with a single
Phylogenetic Systematics oflguanine Lizards
Sauromalus Ctenosaura Cyclura
Iguana
Conolophus
Amblyrhynchus
Pre-Ctenosaura-lguana Stock
Opiurus
DipsosaurusBrachylophus
Chalarodon
Iguanid Ancestor
FIG. 3. "Phylogenetic relationships of the Madagascar Iguanidae and the genera of iguanine lizards,"after Avery and Tanner (1971:71).
variable, size, and should not therefore be used as independent evidence for relationship.Furthermore, these authors made no attempt to assess the evolutionary polarity of theircharacters by comparison with other iguanids.Karyological data on iguanines have been practically useless for systematic purposes.At the crude level of karyotypic analysis commonly applied to lizards, in which onlynumbers and sizes of chromosomes and their centromeric positions are determined,iguanines are conservative. All species of Conolophus, Cyclura, Ctenosaura,Dipsosaurus, and Sauromalus that have been studied possess a karyotype known to be
10 University of California Publications in Zoology
widespread within Iguanidae and found in several other lizard families as well (Paull et al,1976). Only Iguana iguana has been reported to differ from this seemingly primitivecondition in that this species supposedly lacks one pair of microchromosomes (Cohen et
al., 1967), but even this finding was contradicted in another study (Gorman et al., 1967;Gorman, 1973).Iguanine relationships have only been studied superficially with relatively new andincreasingly popular biochemical techniques. Gorman et al. (1971) presented evidence forclose relationship among iguanines based on immunological studies of lacticdehydrogenases and serum albumins in turtles and various diapsids. Higgins and Rand(1974, 1975) showed that the serum proteins and hemoglobins of Amblyrhynchus andConolophus were more similar to each other than to those of Iguana. Unfortunately, otheriguanines were not examined. Wyles and Sarich (1983) performed immunologicalcomparisons of the serum albumins of 10 species of iguanines including representatives ofall eight genera. However, antisera were prepared to the albumins of only four of thespecies, and comparisons with all others are given only for the antisera to the albumins ofAmblyrhynchus and Conolophus. Because of the incompleteness of the data, only verygeneral phylogenetic inferences can be drawn from them.The unique colon of iguanines was studied by Iverson (1980, 1982), who reported thatthe iguanine colon differed from that of all other iguanids and most other lizards in thepossession of transverse valves or folds. However, Iverson (1980) felt that the variation inthese structures within iguanines was of httle value for inferring phylogenetic relationships.Peterson (1984) has recently surveyed the scale surface microstructure of iguanids.Although some intergeneric variation in the morphology of the scale surface is known tooccur in iguanines, representatives of only three iguanine genera {Iguana, Dipsosaurus, andSauromalus) have been studied at this time.One final hypothesis about iguanine relationships deserves mention. At the promptingof a colleague (Ernest Williams) some twenty-five years ago, Richard Etheridge drew up aphylogenetic diagram depicting his views on the interrelationships among the iguanidgenera. The character basis for this diagram was not specified, and Etheridge (pers.comm., 1981) informs me that the relationships shown among the iguanine genera werestrongly influenced by his knowledge about the geographic distributions of these animals.Although he never intended the diagram to be published, it has been published in modifiedform (Paull et al., 1976; Peterson, 1984), and has also appeared in several graduate theses.I reproduce the original diagram here (Fig. 4), noting that its creator does not grant thehypothesis the conviction seemingly implied by a branching diagram.GOALS OF THIS STUDYA detailed study aimed at revealing the pattern of phylogenetic relationships among thevarious iguanine lizards is sorely needed. It would provide invaluable information for themany people studying other aspects of iguanine biology, particularly in an evolutionarycontext. I have attempted such a study here with the following as my goals: (1) to provide
Phylogenetic Systematics oflguanine Lizards 11
Uracentron Ophrv^oessoidesPlica 1 5tfob.\oru5 StenocercU5
PhrtjoosaoraCVcnoV)\cpV\Qris\_\o\acrt\us
Urarwscocion Le\ocet)V)a\os Sce\oporys
H"?.'*- , ' Urosaoros
, PWrij|noSoma 5auroma\us
^PetrosQurys
Phenacosaurus
CV\atT\ae\eoVis
Po^cVtTus CorijlViopVioinesLaetnanc^us
Op\uro5Cha\aro<ion
E nil aVio liesHopAocercosNNovunasouruS
FIG. 4. Etheridge's phylogeny of the Iguanidac.
12 University of California Publications in Zoology
a diagnosis and a description of the group, including the evidence for its monophyleticstatus; (2) to describe variation in the iguanine skeleton, as well as review certain aspects ofnonskeletal morphology; (3) to generalize characters based on this variation; (4) to assessthe polarity of these characters and, thus, their usefulness as evidence for closephylogenetic relationship; (5) to determine the phylogenetic relationships that mostreasonably account for the distributions of these characters among various iguanine lizards;and (6) to provide diagnoses and other pertinent information for monophyletic groupswithin iguanines. The pursuit of these goals has raised numerous questions, some ofwhich are also addressed.
MATERIALS AND METHODS
SPECIMENSThe great majority of the specimens examined in this study were partially intact, completeskeletons that had been prepared by hand or with dermestid beetles. Many skulls preparedby the above methods and some disarticulated skeletons were also examined. Additionally,I have studied radiographs of wet specimens and some whole wet specimens.Observations were made with the naked eye or with the aid of a binocular dissectingmicroscope. Drawings were made using the camera lucida on a Wild binocular dissectingmicroscope, from tracings of photographs and radiographs, and freehand. The specimensexamined are listed in Appendix I.PHYLOGENETIC ANALYSIS
I have used the phylogenetic method elaborated by Hennig (1966), and subsequently bymany others, to study the relationships of iguanine lizards. Reviews of this method can befound in Eldredge and Cracraft (1980) and Wiley (1981). Hennig's method wasformulated specifically for the purpose of evaluating the monophyletic status of groups oforganisms, and he stressed that only synapomorphies, shared features that have arisenwithin a group, logically provide evidence for the existence of its monophyletic subgroups(clades). The assessment of relative apomorphy (polarity) is thus crucial to phylogeneticanalysis. I have used the method of outgroup comparison (Watrous and Wheeler, 1981;Farris, 1982; Maddison et al., 1984) for determining character polarity. Although earlyfossil iguanids might be used as outgroups, I have avoided this practice until theirrelationships to extant forms are better understood. When ontogenetic transformations areadequately known, I have used the transformations themselves as characters (de Queiroz,1985).Outgroups used in this study were the members of four suprageneric groups ofiguanids thought to be outside of the iguanine clade (Etheridge and de Queiroz, 1988):basiliscines, crotaphytines, morunasaurs, and oplurines. Maddison et al. (1984) haveshown that the precise pattern of relationships among the ingroup and various outgroupshas a profound influence on the assessment of plesiomorphy and apomorphy.Unfortunately, relationships among the major groups of iguanids are poorly understood atpresent (Etheridge and de Queiroz, 1988). Therefore, I selected my four outgroups asmuch for convenience as for any notion that they were closely related to iguanines. Each of
13
14 University of California Publications in Zoology
the outgroups is relatively small, which enabled me to examine representatives of all oftheir included genera and most of their included species. Nevertheless, each one of theseoutgroups has at some time been proposed as the closest relative of iguanines: basiliscinesby Etheridge (1964a); crotaphytines by Savage (1958), who actually included them withiniguanines, morunasaurs by Etheridge (Fig. 4), who considered them to be part of a groupancestral to various iguanid lineages; and oplurines by Avery and Tanner (1971).Evidence about character polarity based on outgroups is not always unambiguous,especially in cases such as this one in which relationships of the various outgroups to theingroup are unclear (Maddison et al., 1984; Donoghue and Cantino, 1984). I made acompromise between using the maximum number of characters and using only thosecharacters whose polarities were completely unambiguous based on the outgroup evidence.When all four outgroups suggested the same polarity, that polarity was accepted. When allmembers of all four outgroup taxa did not suggest the same polarity, I considered thepolarity determinable only in those cases where all members of three of the four outgroupssuggested the same polarity. In all other cases I considered the polarity undeterminable.The reasoning behind this decision is given in Appendix 11.BASIC TAXA
I chose the iguanine genera recognized by Etheridge (1982) as the basic taxa among whichphylogenetic relationships were to be determined. Ideally, the basic taxa would all bemonophyletic; however, this information is currently unavailable for the iguanine genera.Although it would be preferable from a theoretical viewpoint to use less inclusive taxa(e.g., species) as basic taxa and then determine which genera are monophyletic, thisalternative has practical limitations. The characters used in this study are primarilyosteological, and many iguanine species are either poorly or not at all represented byosteological preparations. Furthermore, variation within iguanine genera generally appearsto be much less than variation among them. For these reasons, using genera rather thanspecies as basic taxa probably gives a more accurate estimation of the range of variationwithin basic taxa. Nevertheless, overall similarity should not be taken as evidence formonophyly, and I attempt here not only to determine relationships among the iguaninegenera but also to evaluate the evidence for the monophyletic status of each. These are notnecessarily independent issues.THE PROBLEM OF VARIATIONCharacter variation within basic taxa is a problem seldom addressed in the systematicliterature. Contrary to the impression given in many systematic studies, such variation isubiquitous, especially when the basic taxa are higher taxa that are themselves composed ofdiverse subgroups. Nevertheless, many systematic studies apparently ignore this variationor are based on such small samples that no variation within basic taxa is detected. Many ofthe characters used in this study vary within one or more of the iguanine genera, the basic
Phylogenetic Systematics ofIguanine Lizards 1 5
taxa of this analysis. To ignore these characters would be to throw away information; foralthough they contain ambiguities, they also suggest relationships among certain taxa. Onevery important reason for retaining characters that vary within basic taxa is that these are theonly characters that can reveal the paraphyletic status of a basic taxon. I have thus retainedcertain variable characters in my analysis. However, in cases where the variation withinbasic taxa was so great that it obscured any pattern of variation among them, the characterwas eliminated. Because characters form a continuum from those that vary within all basictaxa to those that vary only among them, the decision as to which characters were toovariable to be useful at this level of analysis was necessarily arbitrary.Variation within basic taxa is not itself a problem. For example, variation within basictaxa involving different characters from those that vary among them is simply irrelevant toan analysis of the relationships among these taxa. Even when variation within and amongbasic taxa involves the same characters , the situation is not necessarily problematic. If thevariation within basic taxa characterizes all recognizable subgroups that are units ofphylogenetic ancestry and descent (e.g., evolutionary species), or if the basic taxathemselves are such units, then the variable presence of a derived character in two or morebasic taxa may be attributable to a polymorphic ancestral population. In contrast, whenvariation within basic taxa occurs among one or more of their monophyletic subgroups,this variation represents homoplasy. Because homoplastic variation within a basic taxonmakes an assessment of its ancestral condition ambiguous, such variation is problematic.The conclusion that variation within a taxon represents homoplastic similarity with acondition occurring in another taxon rests on the assumption that the first taxon ismonophyletic. This is a critical point. The variable presence of a derived character in aparaphyletic taxon does not necessarily require homoplasy. In fact, the very characters thatgive evidence of a taxon's paraphyletic status necessarily vary within that taxon. For thisreason, the monophyletic status of basic taxa should be continually reevaluated in light ofvariation within them.Variation within basic taxa that is assumed to represent homoplasy can be handled invarious ways, none of which is without drawbacks. The most appropriate method for aparticular case will depend on the amount and nature of the variation, as well as on theassumptions that the investigator is reasonably able to make. I discuss below four methodsfor handling variation within basic taxa.
1. One means of handling variation within basic taxa is to abandon these taxa in favorof less inclusive basic taxa.: for example, one might use species instead of genera.Unfortunately, this practice carries no guarantee that variation will not exist within the newbasic taxa. A practical disadvantage of this method is that it necessarily increases thenumber of basic taxa. Furthermore, the number of specimens may be distributed in such away that choosing the more inclusive groups as basic taxa gives a more accurate picture ofthe range of variation. A lack of variation within less inclusive taxa may result simply fromsmall samples.
16 University of California Publications in Zoology
2. Kluge and Farris (1969) handled variation within basic taxa by subdividing eachtaxon into two taxa, each coded invariable for one of the alternative states of the variablecharacter. This method makes no assumptions about phylogenetic relationships other thanthose involved in the delimitation of the basic taxa, and it should be satisfactory as long asthere are few variable characters. As the number of variable characters increases for agiven taxon, however, the number of new "taxa" created by subdivision increasesgeometrically. Therefore, this method will be impractical if any of the basic taxa exhibitmore than a few variable characters.3. Information about relationships within a basic taxon can be used to assess theancestral condition of a character that varies within the taxon. The ancestral condition canthen be assigned to the taxon because it follows that the alternative condition bearshomoplastic rather than synapomorphic resemblance to similar conditions in other basictaxa. An obvious drawback of this method is that it requires information aboutrelationships within basic taxa, information that is not always available. An advantage ofthis method is that it does not increase the number of basic taxa.4. A fourth means of dealing with variation within basic taxa is to arbitrarily chooseone of the alternative conditions as the ancestral condition for the variable taxon. Althoughit might intuitively seem that the condition that appears to be ancestral on morphologicalgrounds (i.e., the condition that is found in outgroups) is the ancestral state, thisconclusion has a hidden bias. Given that the variation represents homoplasy, assigning themorphologically "ancestral" state to the taxon amounts to asserting that the homoplasy is aconvergence; assigning the morphologically "derived" state to the taxon amounts toasserting that the homoplasy is a reversal. Furthermore, the most reasonable interpretationof character transformation after construction of a cladogram or phylogenetic tree mayconflict with the original choice of an ancestral condition for a variable basic taxon.Therefore, the level at which characters that vary within basic taxa are considered to besynapomorphies should always be reevaluated after an initial analysis that does not take thevariation into consideration.
I have already presented reasons for not using less inclusive taxa than the iguaninegenera as my basic taxa. In this study, character variation within basic taxa was highenough that the method of Kluge and Farris (1969) would have been impractical. The thirdmethod for handling variation within basic taxa could not be used since not enough isknown about relationships within iguanine genera to use this information to assess theancestral condition of characters that vary within them. Therefore, I have been forced touse the last, and perhaps the least satisfactory, means of dealing with variation within basictaxa. CONSTRUCTION OF BRANCHING DIAGRAMSThe branching diagrams presented in this study are all intended to be cladograms ratherthan phylogenetic trees (Nelson, cited in Wiley, 1979); in other words, no attempt is made
Phylogenetic Systematics ofIguanine Lizards 17
to identify direct ancestors. Cladograms were constructed without the aid of computerprograms that minimize the number of instances in which shared characters appearing to besynapomorphic on morphological grounds have to be interpreted as homoplastic in light ofother characters. In most cases, these cladograms were checked using the Wagnerprograms in Farris's PHYSYS computer package installed on the California StateUniversity CYBER system; however, in the analysis of relationships within Ctenosaura,character incongruence was judged to be so low that such a check was unnecessary. In nocase did the computer analysis find a cladogram with fewer homoplasies than I was able tofind without it.
IGUANINE MONOPHYLY
The family Iguanidae is a heterogeneous assemblage of iguanian lizards distinguished fromother iguanians (agamids and chamaeleontids) primarily by their possession of pleurodontrather than acrodont teeth (e.g., Boulenger, 1884; Cope, 1900; Camp, 1923). Pleurodontyoccurs in nearly all other squamates (except trogonophid amphisbaenians) and in mostother lepidosauromorphs (except sphenodontidan rhynchocephalians), although the teeth ofearly lepidosauromorphs are unlike those of most squamates in that they are set in shallowdepressions (Gauthier et al., 1988). Thus, the primary diagnostic feature of Iguanidae isprobably plesiomorphic for Iguania and is not, therefore, evidence for the monophyleticstatus of the family. Other characters given in a recent diagnosis of Iguanidae (Moody,1980) all either appear to be plesiomorphic for Iguania or do not characterize all iguanids(Estes et al., 1988). There is presently no evidence that Iguanidae is monophyletic.Renous (1979) and Moody (1982) have even suggested that some iguanids may be moreclosely related to some or all of the acrodont iguanians than they are to other iguanids.Because Iguanidae is unlikely to be monophyletic, it seems advisable to break thisfamily into groups for which there is evidence of monophyly. A number of presumablymonophyletic groups of iguanids have been proposed and given informal names (Savage,1958; Etheridge, 1964a, 1976 in Paull et al.; Estes and Price, 1973; Etheridge and deQueiroz, 1988); however, it is beyond the scope of this paper to revise the taxonomy of allpleurodont iguanians. Furthermore, the relationships of these informal groups to agamidsand chamaeleontids are unclear (Estes et al., 1988). Therefore, I provisionally retain thetaxon Iguanidae, emphasizing that it may be paraphyletic and thus may have to beabandoned at some future time.I present evidence below for the monophyletic status of iguanine iguanids. Because Ifeel that it would be premature to disband Iguanidae, but because I also feel that themonophyletic groups of iguanians should be recognized, I resurrect the taxon Iguaninae.Traditionally, recognition of a taxon within a larger one is accompanied by theassignment of all members of the larger taxon to a taxon at the same categorical level as thenew one. This often leads to the recognition of some new paraphyletic group, especially incases such as this one where the author has adequate knowledge of only part of the largergroup being subdivided. Because paraphyletic taxa are undesirable in a phylogenetictaxonomy, I chose a logical alternative: to leave the remaining iguanids unassigned to taxaof rank equal to that of Iguaninae. Indeed, following Gauthier et al. (1988), I have more orless abandoned the use of categorical ranks. My use of "genera" as basic taxa is the resultof historical inertia, and I do not mean to imply that they are equivalent in any biologically
18
Phylogenetic Systematics ofIguanine Lizards 1 9
meaningful way. Similarly, the use of standard endings for new taxa proposed in thisstudy is an attempt to conform with nomenclatural rules rather than to designate categoricalranks. Although these practices go against tradition, they convey the current state ofknowledge precisely. If taxonomies are to serve as summaries of phylogenetic knowledge,conveying such information is more valuable than being able to pigeonhole all organismsequally.
Iguaninae Bell 1 825
Diagnosis. The following combination of characters is diagnostic for iguanines,separating the members of this group not only from all other iguanids but also from allother iguanians and probably from all other squamates:1. some caudal vertebrae with two pairs of transverse processes that diverge from oneanother (Etheridge, 1967);2. the presence of transverse valves or folds in the colon (Iverson, 1980, 1982);3. crowns of posterior marginal teeth laterally compressed, anteroposteriorly flared,and often polycuspate (Etheridge, 1964a);4. the supratemporal lies primarily on the posteromedial surface of the supratemporalprocess of the parietal;5. primarily herbivorous diet (H. M. Smith, 1946).This list does not include all features in which iguanines exhibit a derived conditionwithin Iguanidae; it consists of only those that I consider to be true iguaninesynapomorphies evidencing monophyly of the group. The first two are unique toiguanines (within iguanids) and are therefore unproblematic. Each can be used alone todistinguish an iguanine from any other iguanid, though colic anatomy is unknown formany members of the family. The other characters occur in some noniguanine iguanidswhere I consider them to be convergent, a conclusion necessitated by conflictingdistributions of other derived characters. Still other apomorphic characters of iguanines arenot included in the diagnosis. These characters are also more widely distributed and maybe iguanine synapomorphies (in which case, other convergences will be required) or mayserve to diagnose more inclusive clades within the Iguania. These characters are given inthe description.
Description. The description below is a list of the apomorphies, plesiomorphies, andcharacters of uncertain polarity shared by iguanines. In addition, some characters that varywithin the group are included. This list is meant to provide a basis for comparison withother iguanids and is intended to be a source of characters that may be useful for examiningthe relationships of iguanines to the rest of Iguanidae (and Iguania). For this reason, Iinclude only characters that vary within Iguanidae. Morphologies thought to be derivedwithin Iguanidae are indicated by an asterisk (*).HEAD SKELETON: premaxillary spine exposed dorsally or covered* between nasals;surface of dermatocranium smooth or with irregular rugosities; parietal roof trapezoidal, V-
20 University of California Publications in Zoology
shaped*, or with a posterior midsagittal crest (Y-shaped)*; parietal foramen present, onfrontoparietal suture or within frontal*; outUnes of osseous labyrinth indistinct (moderatelydistinct in Dipsosaurus); supratemporal situated primarily on posteromedial surface ofsupratemporal process of parietal*; lacrimal present; postfrontal present; septomaxilla witha large posterodorsal shelf*; epipterygoid present; Meckel's groove closed and fusedbetween anterior end of splenial and mandibular symphysis*; angular present; splenialpresent; coronoid with large lateral process overlapping dentary; three to elevenpremaxillary teeth (most species with a mode of seven); premaxillary teeth without lateralcusps, bicuspid, or tricuspid; crowns of posterior marginal teeth flared, with three tomany* cusps; palatine teeth absent*; pterygoid teeth present or absent*; secondceratobranchials long* or short, adpressed or separated* medially; 14 scleral ossicles.AXIAL SKELETON: vertebral neural spines tall or short*; zygosphenes and zygantrawell developed*; mode of 24 or 25* presacral vertebrae; caudal autotomy septa present orabsent*; less than 25* to more than 70 caudal vertebrae; some caudal vertebrae with twopairs of transverse processes that diverge from one another*; first rib usually borne on fifthcervical vertebra; usually four pairs of free (cervical) ribs on fifth through eighth presacralvertebrae; usually four pairs of sternal ribs on ninth through twelfth presacral vertebrae;one* or two pairs of xiphistemal ribs attached to 13th and usually 14th presacral vertebrae;all post-thoracic presacral vertebrae usually bear unfused ribs; postxiphistemal inscriptionalribs attached to corresponding bony ribs, with zero to three pairs meeting midventrally toform continuous chevrons.APPENDICULAR SKELETON: clavicles with or without* flattened lateral shelf; dorsalends of clavicles articulate with suprascapulae; clavicular fenestrae usually absent (except inConolophus); scapular fenestrae usually present* (variably reduced or absent inAmblyrhynchus and Sauromalus); posterior coracoid fenestrae present* or absent; lateralarms of interclavicle form angles of 45? to 90? with posterior process; posterior process ofinterclavicle extending posterior to lateral corners of sternum or not*; sternal fontanellepresent but small (approximately equal to interclavicle in width) or absent*; phalangealformula of manus 2:3:4:5:3; phalangeal formula of pes 2:3:4:5:4.NONSKELETAL ANATOMY: superciliary scales quadrangular and non-overlapping toelongate and strongly overlapping (Etheridge and de Queiroz, 1988); subocular scales allsubequal and quadrangular to one greatly elongate (Etheridge and de Queiroz, 1988);interparietal scale small; transverse gular fold present (weakly developed inAmblyrhynchus); pendulous dewlap present* or absent; gular crest present* or absent;middorsal row of enlarged scales present or absent*; subdigital lamellae keeled, withoutsetae; femoral pores present in one or two rows; preanal pores absent; scale surface withhoneycomb pattern (Etheridge and de Queiroz, 1988); scale organs lacking elongatedspines (Etheridge and de Queiroz, 1988); nasal passage S-shaped, with concha present(Stebbins, 1948); colon with transverse valves or folds (Iverson, 1980).
COMPARATIVE SKELETAL MORPHOLOGY
In this section I describe variation in the iguanine skeleton by region and compare iguanineswith basiliscines, crotaphytines, morunasaurs, and oplurines. These descriptionsemphasize variation that is relevant to an analysis of relationships among iguanine generaand are not intended to be exhaustive. For more detailed descriptions of the head skeletonthat include features common to all iguanines I refer the reader to Oelrich (1956), whoseterminology is followed below. SKULL ROOFThe iguanine skull roof, or the superficial dermatocranial elements of the skull proper,consists of the full complement of bones thought to be plesiomorphic for squamates. Otheriguanians may lack some of these bones, most commonly lacrimals and postfrontals. Thequadrate, a splanchnocranial bone, is also described in this section. Bones of the iguanineskull roof, palate, braincase, and lower jaw are illustrated in Figures 5 and 6.Premaxilla (Figs. 5A, 6A, 7, 8). The premaxilla is the anteriormost bone in the skull.Postembryonically, it is a median, unpaired bone that is sutured laterally with the maxillae,posterodorsally with the nasals, and posteroventrally with the vomers. The premaxillabears a ventrally directed incisive process near its posteroventral end on the midline (Fig.7). Foramina for the maxillary arteries (Oelrich, 1956) penetrate the premaxilla on eitherside of the incisive process. In most iguanines (Brachylophus, Cyclura, Ctenosaura,Iguana, Dipsosaurus, and Sauromaliis) and all outgroups examined, the ventral surface ofthe premaxilla bears well-developed posteroventral extensions where it sutures with themaxillae. A weak ventral crest runs along the posterior edge of the premaxilla from itsposterolateral comers to the base of the incisive process. This crest is generally not piercedby the foramina for the maxillary arteries. In Amblyrhynchiis, the posterolateral extensionsof the premaxilla are small and the posterolateral corners of this bone are concomitantlycloser to the base of the incisive process (Fig.7A) than are those of other iguanines (Fig.7B). Conolophiis differs from the typical iguanine pattern in having large posteroventralcrests that continue up the sides of the incisive process and are pierced or notched by theforamina for the maxillary arteries (Fig. 7C). In Sauromaliis slevini these foramina alsopierce the ventral crest of the premaxilla; however, the condition does not appear to behomologous with that seen in Conolophus. In Conolophus the crests are pierced becauseof their enlargement, while in S. slevini the foramina appear to have moved posteriorly.
21
22 University of California Publications in Zoology
pmx pmx
FIG. 5. Skull oi Brachylophus vitiensis (MCZ 160254): (A) dorsal view, (B) ventral view, (C)posterior view. Scale equals 1 cm. Abbreviations: bo, basioccipital; bs, parabasisphenoid; ect,ectopterygoid; eo, exoccipital-opisthotic; fr, frontal; ju, jugal; la, lacrimal; mx, maxilla; na, nasal; oc,occipital condyle; pal, palatine; par, parietal; pmx, premaxilla; prf, prefrontal; ps, parasphenoid rostrum;ptf, postfrontal; pto, postorbital; ptr, pterygoid; q, quadrate; soc, supraoccipital; sq, squamosal; st,supratemporal; vo, vomer.
Phylogenetic Systematics of/guanine Lizards 23
B
FIG. 6. Skull and mandible of Braehylophus vitiensis (MCZ 160254): (A) lateral view of skull, (B)lateral view of left mandible, (C) lingual view of left mandible. Scale equals 1 cm. Abbreviations: aiaf,anterior inferior alveolar foramen; amf, anterior mylohyoid foramen; an, angular; ap, angular process; ar,articular; bs, parabasisphenoid; cor, coronoid; den, dcntary; ect, ectopterygoid; ept, epipterygoid; fr, frontal;ju, jugal; la, lacrimal; mf, mental foramina; mx, maxilla; na, nasal; par, parietal; pmf, posterior mylohyoidforamen; pmx, premaxilla; pre, prearticular; prf, prefrontal; pro, prootic; ptf, postfrontal; pto, postorbital;ptr, pterygoid; ps, parasphenoid rostrum; q, quadrate; rap, retroarticular process; slf, supralabial foramina;smx, septomaxilla; sp, splenial; sq, squamosal; st, supratemporal; sur, surangular.
24 University of California Publications in Zoology
Bvo
fma
vo
FIG. 7. Posteroventral views of the premaxillae of (A) Amblyrhynchus cristatus (RE 1396), (B)Cyclura cornuta (RE 383), and (C) Conolophus pallidas (RE 439), showing the small posterolateralprocesses (arrows) of the premaxilla in Amblyrhynchus and the large lateral crests of the incisive processthat are pierced by foramina for the maxillary arteries in Conolophus. Scale equals 0.5 cm. Abbreviations:fma, foramen for maxillary artery; ip, incisive process; vo, vomer.
The premaxilla of Amblyrhynchus differs from that of basiliscines, crotaphytines,morunasaurs, oplurines, and all other iguanines in several other ways. Its anterior edgeforms a nearly flat plate rather than being arched, and its nasal process is nearly verticalinstead of sloping backwards.The nasal process of the premaxilla, or the premaxillary spine, meets the nasalsposterodorsally (Figs. 5A, 6A, 8). In cross section this process is roughly triangular, withthe apex of the triangle pointing posteroventrally. The shape of this triangle variesconsiderably, ranging from broad-based and low in Ctenosaura and Conolophus tonarrow-based in Amblyrhynchus, Dipsosaurus, and Sauromalus to nearly oval in Cycluracornuta. Differences between morphological extremes are great, but the extremes grademore or less continuously into one another through intermediates; thus, there are no gaps to
Phylogenetic Systematics oflguanine Lizards 25
FIG. 8. Dorsal views of the preorbital portions of the skulls of (A) Sauromalus varius (RE 308), (B)Ctenosaura hemilopha (RE 1964), and (C) Conolophus subcristatus (MVZ 77314), showing differences inthe degree to which the nasal process of the premaxilla is covered dorsally by the nasals. Scale equals 1.0cm. Premaxilla is shaded. Abbreviations: mx, maxilla; na, nasal; prf, prefrontal.
separate discrete character states. Furthermore, variation within Cyclura is greater than thatbetween many of the genera. For these reasons I have chosen not to use variation in thecross-sectional shape of the premaxillary spine as a character in phylogenetic analysis.When viewed dorsally, the posterior exposure of the nasal process is variable withiniguanines (Fig. 8). In Brachylophus, Ctenosaura, Cyclura, Dipsosaurus, Iguana, andSauromalus, the nasal process of the premaxilla is exposed dorsally or covered onlyslightly dorsolaterally by the nasals. Although differences in the extent of this overlap andthe length of the premaxillary spine yield strikingly different morphologies (Fig. 8A, B),intraspecific variation in these features is too great to permit their subdivision into characterstates. In Amblyrhynchus and Conolophus, however, one finds a condition not seen inany other iguanine. The nasals of these genera cover the premaxillary spine posteriorly sothat the posteriormost portion of the spine that is visible dorsally falls short of thetransverse plane at the posterior ends of the fenestrae exonarinae (bony external nares or
26 University of California Publications in Zoology
nasal fossae). In Cyclura cornuta, C. pingius, and Iguana delicatissima, the visible portionof the nasal process of the premaxilla also fails to reach this plane; however, this conditionresults from enlargement of the nasal fossae in these species rather than from increasedoverlap of the premaxillary spine by the nasals.Basiliscines, oplurines, and morunasaurs generally have the nasal process of thepremaxilla exposed dorsally, at least along the midline. In crotaphytines, more extensiveoverlap of the premaxillary spine by the nasals may occur, but never to the extent seen inAmblyrhynchus and Conolophus.Nasals (Figs. 5A, 6A, 8). This pair of bones lies along the midline of the skull roofimmediately posterior to the fenestrae exonarinae, forming the roof of the nasal capsule.The relative size of the nasals varies considerably among iguanines. Brachylophus,Conolophus, Ctenosaura, most Cyclura, Dipsosaurus, Iguana iguana, and Sauromalusexhibit a condition that is widespread outside of the iguanines in which the nasals are ofmoderate size. In Conolophus the nasals are subequal to the frontals in length, and in theremaining taxa the nasals are shorter than the frontals. The nasals of Amblyrhynchus aregreatly enlarged, accompanying the enlargement of the entire nasal capsule. The increasedsize of the nasal capsule probably enables it to house the large nasal salt glands of thisspecies (figured by Dunson, 1969, 1976). Because enlargement of the nasals is onecomponent of enlargement of the nasal capsule, I did not treat the two as separatecharacters.The relative size of the nasals is also correlated with the size of the fenestrae exonarinae(Fig. 9). In some species of Cyclura, most notably C. cornuta (Fig. 9A) and C. pinguis,and in Iguana delicatissima, enlargement of the fenestrae exonarinae results fromemargination at the anterior edges of the nasals, reducing the relative size of these bones.Because this apomorphic condition occurs only in some Iguana and Cyclura, I have notused it for analyzing intergeneric relationships. The feature may, however, provide usefulinformation for uncovering relationships within these genera.Septomaxillae (Fig. 6A). The septomaxillae are thin, paired bones lying in the anteriorportion of the nasal capsule on either side of the nasal septum. These bones rest atop thevomers and, except in Iguana delicatissima, they contact the roof of the nasal capsule(premaxilla or nasals) dorsally. In most iguanines, each septomaxilla is relatively flat on itsdorsal surface, though a low ridge often appears near the lateral edge of the bone. Unlikeall other iguanines, the septomaxillae of Amblyrhynchus bear sharp longitudinal crests ontheir anterodorsal surfaces. These crests appear to be apomorphic, since they are absent inall four outgroups. The septomaxillae of crotaphytines and oplurines are similar to those ofiguanines in that they each send a long shelf posterodorsally. Those of morunasaurs andBasiliscus are relatively small and are folded to form a transverse ridge. The septomaxillaeof Laemanctus are very small, and those of Corytophanes are possibly absent. A dorsalseptomaxillary crest is seen in some morunasaurs, but this crest is entirely different fromthat of Amblyrhynchus, extending transversely rather than longitudinally.Prefrontals (Figs. 5A, 6A, 8, 9). The prefrontals have roughly the shape of atriangular pyramid whose apex forms the posterolateral comer of the nasal capsule.
Phylogenetic Systematics of Iguanine Lizards 27
B
FIG. 9. Dorsal views of the skulls of (A) Cyclura cornuta (AMNH 57878) and (B) Sauromalus obesus(RE 467), showing differences in the relative sizes of the external nares and the presence or absence ofprefrontal contribution to the posterior margin of the nares. Prefrontal is shaded. Scale equals 1 cm.
Boulenger (1890) noted that the contribution of this bone to the posterior margin of thefenestra exonarina (Fig. 9A) distinguished Metopoceros cornutus (Cyclura cornuta) fromall other iguanines. In the other iguanines, maxilla and nasal meet anterior to the prefrontaland exclude it from the margin of the fenestra (Fig. 9B). This condition is undoubtedlyplesiomorphic, since it is found in all basiliscines, crotaphytines, morunasaurs, andoplurines, and in most other lizards (except varanoids, Shinisaurus, and chamaeleons). InC. cornuta, contribution of the prefrontal to the margin of the fenestra exonarina is relatedto enlargement of the latter. Although the prefrontals of other species with large bonyexternal nares do not enter the fenestra margin, none of these have bony external nares aslarge as those of C cornuta. Because the apomorphic condition is found in only onespecies of a genus containing eight, it provides no information about intergenericrelationships.A large lacrimal foramen, bounded by the prefrontal medially and the lacrimal laterally,pierces the anterior wall of the orbit (Fig. 10). Just posterior to the lacrimal foramen fourbones-lacrimal, jugal, palatine, and prefrontal-approach one another, and one of two
28 University ofCalifornia Publications in Zoology
B
FIG. 10. Posterodorsal views of the anterior orbital regions of (A) Brachylophus fasciatus (RE 1888)and (B) Conolophus pallidas (MCZ 19111), showing differences in the contacts of the bones in this region.In A the lacrimal (la) contacts the palatine (pal); in B the prefrontal (prf) contacts the jugal (ju). Scaleequals 0.5 cm.
different contacts may be established: between lacrimal and palatine (Fig. lOA), orbetween prefrontal and jugal (Fig. lOB). Occasionally all four bones meet at a single point.Because the lacrimal-palatine contact occurs in all outgroups except crotaphytines, thiscondition appears to be plesiomorphic for iguanines. The apomorphic prefrontal-jugalcontact is the common condition in Amblyrhynchus and Conolophus. The plesiomorphiccondition is characteristic of most species of all other iguanine genera, although Ctenosauraclarki, Cyclura carinata, C. cornuta, and C. ricordii appear to be characterized by theapomorphic condition. This character is also variable within some iguanine species andoccasionally even varies between right and left sides of a specimen thus decreasing its valueas a systematic character.Frontal (Figs. 5A, 6A, 11, 12). The frontal of postembryonic iguanines is an unpairedmedian bone that forms the dorsal borders of the orbits. It is sutured anteriorly with thenasals and anterolaterally with the prefrontals. Posteriorly the frontal meets the parietal,forming a transverse suture, and posterolaterally it contacts the postfrontal and variably thepostorbital bones. The proportions of the frontal vary within Iguaninae. Generally, thefrontal is about as wide (at the frontoparietal suture) as it is long (maximum exposed lengthdorsally), or is longer than wide. Four iguanines have frontals that are markedly widerthan long: Amblyrhynchus (Fig. 11) has the relatively widest frontal of all iguanines,
Phylogenetic Systematics oflguanine Lizards 29
pmx
FIG. 11. Dorsal view of the skull of Amblyrhynchus cristatus (MVZ 67721), showing the short, widefrontal and the wedge-shaped orbital borders. Scale equals 1 cm. Abbreviations: fr, frontal; ju, jugal; mx,maxilla; na, nasal; par, parietal; pmx, premaxilla; prf, prefrontal; ptf, postfrontal; pto, postorbital; sq,squamosal; st, supratemporal.
while Conolophus, Cyclura cornuta, and Iguana delicatissima are intermediate betweenAmblyrhynchus and the other iguanines. In C. cornuta and /. delicatissima reduction infrontal length may be related to enlargement of the fenestrae exonarinae, although Cyclurapinguis, which has large bony external nares, has a frontal that is about as wide as it islong.The width of the frontal between the orbits exhibits substantial intergeneric variation iniguanines, but the pattern of variation is obscured by correlation of this feature with size.Small species and juveniles of large species have relatively large orbits and concomitantly
30 University of California Publications in Zoology
narrow frontals. As members of large species grow, relative orbit size decreases andrelative frontal width increases correspondingly. Nonetheless, Brachylophus hasseemingly wider frontals than would be predicted on the basis of its body size andknowledge of frontal allometry in other iguanines. A detailed study of this allometry wasnot undertaken, and I did not use variation in frontal width as a systematic character.Along with the prefrontals and postfrontals, the frontal forms the dorsal borders of theorbits, which have a unique shape in Amblyrhynchus. In this genus, the dorsal orbitalborders are wedge-shaped, with the apices of the wedges pointing medially when viewedfrom above (Fig. 11). In all other iguanines and all outgroups examined, the dorsalborders of the orbits are more or less smoothly curved (Figs. 5A, 9), though the preciseshape varies among taxa.Dipsosaurus has a pair of relatively large openings at or near the suture between frontaland nasals, one on each side of the midline. No other iguanine nor any outgroup that Ihave examined has these, although some have much smaller foramina in roughly the samelocation that may or may not be homologous. An area just anterior to the frontal on themidline fails to ossify in certain other iguanids (e.g., Corytophanes and morunasaurs), butit is not divided into paired openings as in Dipsosaurus.On the anteroventral surface of the iguanine frontal at the posterior end of the nasalcapsule, one or two pairs of crests may develop (Fig. 12). The lateral pair are the cristaecranii, which form the dorsolateral walls of the olfactory tract and are invariably present.In most iguanines these cristae extend continuously from the frontal onto the prefrontals(Fig. 12A); however, in Conolophus and in Ctenosaura defensor, the frontal portions ofthe cristae project anteriorly, forming a step in each crista between its frontal and prefrontalportions (Fig. 12B). Basiliscines, crotaphytines, morunasaurs, and oplurines all possesscristae cranii resembling those in the majority of iguanines.A second pair of crests, located medial to the cristae cranii, is variably developed iniguanines. These crests extend posterolaterally from the anterior end of the frontal at themidline towards the cristae cranii. This medial pair of crests is absent in most iguanines(Fig. 12A). Weakly developed medial crests are seen in Brachylophus and someCtenosaura, and somewhat larger ones occur in Conolophus (Fig. 12B). Amblyrhynchusexhibits the strongest development of the medial crests and also undergoes considerableontogenetic change in this feature. At hatching, the medial crests of Amblyrhynchus aresimilar to those of other iguanines that possess the crests, but are more strongly developedand are united anteriorly to form a single median crest (Fig. 12C). During postembryonicdevelopment the median portion elongates and grows ventrally, while the posterior ends ofthe crests also grow ventrally and become continuous with the cristae cranii. The end resultis the formation of a pair of deep pockets separated by a median crest on the ventral surfaceof the frontal bone (Fig. 12D). Large nasal salt glands (Dunson, 1969, 1976) probably fillthese pockets in entire specimens.Continuous morphological variation and a high degree of intergeneric overlap in therange of this variation limit the usefulness of the medial pair of frontal cristae as charactersin phylogenetic analysis. The unique condition seen in Amblyrhynchus, however, is easily
Phylogenetic Systematics of Iguanine Lizards 31
mc
FIG. 12. Ventral views of the frontals of (A) Ctenosaura pectinata (RE 641), (B) Conolophussubcristatus (AMNH 165756), (C) near-hatching Amblyrhynchus cristatus (SDNHM 45157), and (D) adultA. cristatus (RE 1387), showing differences in the morphology of the cristae cranii and the development ofcrests medial to the cristae cranii. Scale equals 0.5 cm. Abbreviations: cc, crista cranii; fr, frontal; mc,medial crest; prf, prefrontal.
distinguished from that of all other iguanines and warrants recognition as a character. Alloutgroups either lack the medial cristae of the frontals or have them only weakly developed.The parietal foramen is a small hole that pierces the dermal skull roof above the anteriorportion of the brain. It serves as a window in the skull for the parietal eye, a photosensitiveorgan (Hamasaki, 1968, 1969). The location of the parietal foramen relative to the frontaland parietal bones is variable within iguanines (Table 2). Because the medial portion of the
32 University of California Publications in Zoology
TABLE 2. Position of the Parietal Foramen
Phylogenetic Systematics ofIguanine Lizards 33
skull roof in the vicinity of the frontoparietal suture ossifies in postembryonic ontogeny,the location of the parietal foramen cannot always be determined in young specimens.In Amblyrhynchus, Brachylophus, Conolophus, and Iguana, some species ofCtenosaura, and most species of Cyclura, this foramen almost always lies within thefrontoparietal suture, though it may notch either frontal or parietal more than the other. Theparietal foramen is entirely within the frontal in Cyclura carinata and Dipsosaurus; a sutureconnecting it with the frontoparietal suture is usually present in the former but usuallyabsent in the latter. The single skull of Ctenosaura defensor examined has the parietalforamen located entirely within the frontal. All species of Sauromalus and several speciesof Ctenosaura are variable in this character: the parietal foramen in these taxa is commonlyfound both at the suture between frontal and parietal or entirely within the frontal.Most members of the outgroups examined in this study have the parietal foramen at thefrontoparietal suture, suggesting that this condition is plesiomorphic for iguanines.Exceptions are Basiliscus and Corytophanes, which have the parietal foramen entirelywithin the frontal; Laemanctus, in which the parietal foramen may be either on thefrontoparietal suture or within the frontal; and Morunasaurus annularis, in which theparietal foramen appears to be absent. Assuming that fixation of an apomorphic feature ismore readily achieved through a polymorphic intermediate condition, I recognized thevariable condition as the intermediate stage in a three-state transformation series.Postfrontals (Figs. 5A, 6A). The postfrontals are small bones confined to theposterodorsal margins of the orbits. The posterior surface of each postfrontal is suturedmedially to the frontal and laterally to the postorbital. The postfrontal is invariably presentas a discrete element in iguanines, morunasaurs, and Laemanctus; it is indistinct (absent orfused) in crotaphytines, oplurines (rarely, a small separate bone is present), and thebasiliscines Corytophanes and Basiliscus.In iguanines, the lateral portion of the postfrontal may form part of a bony knob alongwith the postorbital (Figs. 6A), which serves as an attachment point for the skin (Oelrich,1956). The relative development of this knob varies among iguanine genera.Amblyrhynchus and Brachylophus have moderate-sized knobs directed mostly laterally.The knob is small or absent in Conolophus, Ctenosaura, Dipsosaurus, and Sauromalus.Iguana and especially Cyclura have large, anteriorly directed knobs (Fig. 9A). The relativesize of the postfrontal-postorbital knob increases with increasing body size, making itdifficult to compare those of animals differing gready in body size. Because of thisproblem, I have chosen not to use the variation in the development of this knob as asystematic character.Postorbitals (Figs. 5A, 6A). The postorbitals of iguanines are paired, triradiate bonessituated on the posterolateral sides of the skull just behind the orbits. Their lateral surfacesare often slightly concave. Like those of all iguanians, the postorbitals of iguanines form amajor part of the postorbital bar and articulate anteroventrally with the jugals andposteroventrally with the squamosals. Preliminary examination suggested that therelationship of postfrontal to parietal might be a useful systematic character. Thepostorbital generally ends medially where it contacts the parietal, or it may slightly overlap
34 University of California Publications in Zoology
B
FIG. 13. Dorsal views of the parietals of four Iguana iguana-{A) RE 454, condylobasal length =31.3mm; (B) JMS 713, 59.6mm; (C) RE 424, 71.6mm; (D) RE 489, 80.7mm-showing ontogeneticchange in the shape of the parietal roof. Scale equals 1 cm.
the posterior side of the anterolateral process of the parietal. In some iguanines, thisoverlap is much more extensive, and in others postorbital and parietal fail to contact.Because variation in this feature seems to be greater within the basic taxa than betweenthem, I did not use this variation as a systematic character.Parietal (Figs. 5A, 6A, 13). The posteriormost bone lying on the midline of the skullroof is the parietal, which is unpaired in postembryonic development. This bone forms anearly straight transverse suture with the frontal anteriorly, and meets the postorbitals andpostfrontals anterolaterally. It has a supratemporal process extending posterolaterally to thecomplex articulation for the cephalic condyle of the streptostylic quadrate. The adductormuscles of the jaw originate on the dorsolateral surfaces of the parietal. Their area ofattachment is set off by distinct crests from the medial and anterior portions of the bone, towhich the skin adheres. This latter area, the parietal roof, changes shape during thepostembryonic ontogeny of all iguanines. Changes in the shape of the parietal roof aremost pronounced in large species (Fig. 13), which undergo the greatest changes in sizeafter hatching. At hatching, the parietal roof is roughly trapezoidal, the lateral crests beingwidely separated. As the animal grows, the posterior portions of the crests moveprogressively closer together until they meet, forming a V-shaped roof. Further growthresults in elongation of the single median crest formed by the posterior union of the lateralcrests, so that the parietal roof takes on the shape of a Y with a growing leg. Different taxastop at different points along this pathway, and the point of termination is correlated withsize. Similar ontogenetic changes in the shape of the parietal roof have been noted in otheriguanids (Etheridge, 1959).The phylogenetic significance of differences in parietal roof shape cannot be adequatelyassessed without a detailed allometric study, for differences in shape are complicated bydifferences in the sizes of the organisms being compared. Furthermore, outgroup
Phylogenetic Systematics oflguanine Lizards 35
comparison has limited value here, because no other iguanids get as large as the largestiguanines; and the basiliscines, moderately large iguanids, have highly modified parietals.Nevertheless, a few observations warrant mention. The lateral parietal crests ofSauromalus never meet, despite the fact that some of its species (5. hispidus and S. varius)attain larger sizes than other iguanines in which the crests do eventually meet (e.g.,Brachylophus). In Amblyrhynchus, one of the largest iguanines, the parietal crests doeventually meet, but this occurs at a larger size than in other iguanines, and a Y-shaped roofdoes not develop. If the complete ontogenetic pathway described above is plesiomorphicfor iguanines, then Sauromalus and Amblyrhynchus exhibit derived conditions. Becauseof the ambiguities involved in this character, however, I omitted it from the phylogeneticanalysis.A unique feature of the parietal is seen within the genus Ctenosaura. In C. acanthuraand C. pectinata, the parietal extends posteriorly, forming a shelf above the braincase (H.M. Smith, 1949, and pers. comm. cited by Ray and Williams, unpublished). This featuresuggests that C. acanthura and C. pectinata form a clade within Ctenosaura, and it is usedas a systematic character only in an analysis of relationships within that taxon.Supratemporals (Figs. 5C, 6A). The supratemporals are small bones that form themajor part of the articulation for the dorsal end of the streptostylic quadrates. The posteriorend of each supratemporal is wedged between four bones: the quadrate ventrally, thesquamosal laterodorsally, the parietal dorsally, and the exoccipital medially. In alliguanines, the posterior end of the supratemporal wraps around the ventral edge of thesupratemporal process of the parietal. As it extends anteromedially, the supratemporalbecomes confined to the posteromedial surface of the supratemporal process. The greaterpart of the supratemporal lies on this posteromedial surface. In most other lizards, thesupratemporal lies primarily on the anterolateral surface of the supratemporal process of theparietal for its entire length. Sometimes it extends along the ventral edge of thesupratemporal process and bears medial and lateral portions of approximately equal size.As far as I am aware, oplurines, Enyalius, and mosasaurs are the only other lizards inwhich the greater portion of the supratemporal lies on the posteromedial surface of thesupratemporal process of the parietal. Therefore, I interpret this condition as an iguaninesynapomorphy.The anterior extent of the supratemporal varies within Iguaninae. In most, thesupratemporal extends forward at least halfway across the posterior temporal fossa. Thiscondition occurs also in basiliscines, crotaphytines, morunasaurs, and oplurines, and istherefore taken to be plesiomorphic for iguanines. The supratemporal of Conolophus hasapparently been reduced phylogenetically; it sometimes reaches halfway across theposterior temporal fossa, but generally falls short of this point.Maxillae (Figs. 5A,B, 6A, 14). The maxillae are paired bones that bear most of theupper marginal teeth. They are roughly triangular and lie on the anterior sides of the skull,where they meet the premaxilla anteriorly, the nasals and prefrontals dorsally, the lacrimalsand jugals posterodorsally, and the ectopterygoid posteriorly. A number of supralabialforamina pierce the maxillae in a row parallel to its ventral border. Compared to the
36 University ofCalifornia Publications in Zoology
FIG. 14. Lateral view of the skull of Ctenosaura similis (MCZ 21742), showing the dorsal curvature ofthe premaxillary process of the maxilla (arrow). Scale equals 1 cm. Abbreviations: fr, frontal; ju, jugal;la, lacrimal; mx, maxilla; na, nasal; par, parietal; pmx, premaxilla; prf, prefrontal; ptf, postfrontal; pto,postorbital; q, quadrate.
supralabial foramina of other iguanines, those of Amblyrhynchus seem to lie slightly higheron the maxillae above a rounded ridge that is not seen in any other iguanine or in anyoutgroup examined in this study. There is also variation in the relative size of thesupralabial foramina within Iguaninae, with large ones being found in some species ofCyclura, especially C. cychlura. This variation is not useful for examining relationshipsamong the basic taxa used in this study, since it occurs in only some Cyclura.The maxillae of Ctenosaura are unique in that the premaxillary processes in this genuscurve dorsally, so that the premaxilla and the anterior portions of the maxillae are higherthan the rest of the upper jaw margin, giving the skull a sneering appearance (Fig. 14).The teeth in this region form large, curved fangs. The dorsal displacement of the anteriorend of the tooth row increases allometrically both within and between species ofCtenosaura, being less pronounced in juveniles of large species and in adults and juvenilesof small species. In all other iguanines, the entire upper jaw margin lies in a singlehorizontal plane or is only slightly elevated anteriorly (Fig. 6A). Although no otheriguanines nor any of the outgroups exhibit as pronounced a curvature of the premaxillaryprocess of the maxilla as that seen in large Ctenosaura, many taxa are difficult to comparebecause of their small size and the allometric change seen in this feature.Lacrimals (Figs. 5A, 6A, 14). The lacrimals of iguanines are small bones situated atthe anterior corner of each orbit. In Amblyrhynchus these bones are relatively smallcompared to those of other iguanines. Conolophus also has relatively small lacrimals,intermediate in size between those of Amblyrhynchus and the smallest ones seen in otheriguanines. All other iguanines have relatively large lacrimals whose size and shape varyamong the genera. This variation ranges from the long, curved bones of Brachylophus
Phylogenetic Systematics oflguanine Lizards 37
(Fig. 6A) and Dipsosaurus to the almost square ones of Ctenosaura (Fig. 14), Cyclura, andIguana (those of Sauromalus are intermediate). Although the extreme lacrimalmorphologies in iguanines other than Amblyrhynchus and Conolophus are very different,the variation between them is more or less continuous.Basiliscines, crotaphytines, and oplurines all have relatively large lacrimals, but thoseof morunasaurs are relatively small and may even be absent in some Morunasaurus. Thus,it appears that a small lacrimal is apomorphic within iguanines, although the evidence is notcompletely unambiguous. If so, then the small lacrimals of morunasaurs must beconvergent.Jugals (Figs. 5A, 6A, 14). The iguanine jugals form the ventral margins of the orbitsand are sutured anteriorly with the lacrimals, anteroventrally with the maxillae, mediallywith the ectopterygoids, and posterodorsally with the postorbitals. Each jugal extendsposteriorly along the ventral border of the postorbital and variably contacts the squamosalon the ventral surface of the upper temporal arch. This contact appears to be too variablewithin genera to serve as a character for examining their interrelationships.Squamosals (Figs. 5A,C, 6A, 15). At the posterior end of each temporal arch, behindthe postorbitals, lie the squamosals. The shape of the squamosal is variable, ranging fromlong and thin in Ctenosaura and Sauromalus to short and wide in Amblyrhynchus; theremaining genera are intermediate. At its posterior end, the squamosal bears twoprocesses: a dorsal process that meets the supratemporal process of the parietal, and aventral process or peg directed towards the quadrate. The relative size of the ventralprocess is variable, being more strongly developed in Amblyrhynchus and Iguana than inthe other genera (Fig. 15). In these two genera, however, the relationship of the ventralprocess of the squamosal to the quadrate is different. As in most iguanines (Figs. 6A,15A), the ventral process of Amblyrhynchus (Fig. 15B) lies against the anterior edge of thecephalic condyle of the quadrate, projecting into a gap or hole between this condyle and thedorsal portion of the tympanic crest of the quadrate. In Iguana, the ventral process abutsdirectly against the top of the tympanic crest (Fig. 15C), presumably reducing the mobilityof the quadrate. Cyclura possesses a potentially incipient stage to the condition seen inIguana. Its squamosal also abuts against the tympanic crest of the quadrate, but does somore weakly because the ventral process is not as large (Fig. 15D).The relationship of the ventral process of the squamosal to the quadrate, and the relativesize of this process in basiliscines, crotaphytines, morunasaurs, and oplurines, are verysimilar to those of Amblyrhynchus, suggesting that this condition is plesiomorphic foriguanines. The unique articulation between squamosal and quadrate in Cyclura and Iguanasuggests that the large ventral process in Iguana may not be homologous with those ofAmblyrhynchus and non-iguanines. In order to maintain objectivity, however, I did notassume that such was the case.Quadrates (Figs. 5B,C, 6A, 15). The quadrates lie at the posteroventral comers of theskull. They are streptostylic and are important in jaw mechanics and feeding (Rieppel,1978; K. K. Smith, 1980). Ventrally, each quadrate forms the articulation of the skull withthe articular bone of the mandible, and it also articulates dorsally with the squamosal,
38 University of California Publications in Zoology
FIG. 15. Lateral views of the posterolateral comers of the skulls of (A) Conolophus pallidas (RE 439),(B)Amblyrhynchus cristatus (RE 1387), (C) Iguana iguana (RE 1006), and (D) Cyclura cornuta (T?E 383),showing differences in the size of the ventral process of the squamosal and the articulation betweensquamosal (sq) and quadrate (q).
Phylogenetic Systematics oflguanine Lizards 39
dorsomedially with the supratemporal, and ventromedially with the pterygoid. Thequadrate is bowed forward, and its lateral edge, the tympanic crest, supports the anterioredge of the external tympanic membrane. In small species and juveniles of larger species,the quadrates tilt backwards with their articular condyles lying anterior to the transverseplane at the posterior end of the occipital condyle. During ontogeny the articular condylesmove posteriorly, eventually coming to lie posterior to the occipital condyle and giving thequadrates a forward tilt. A related change occurs in the pterygoids, which must elongateposteriorly as the articular condyles move backwards if they are to remain connected to thequadrates. Variation in the relationship of the squamosal to the quadrate has been describedabove. PALATEThe iguanine palate (Fig. 5B) consists of four pairs of dermal bones that form the floor ofthe skull proper and the roof of the mouth: vomers, palatines, pterygoids, andectopterygoids.Vomers (Fig. 5B). The vomers are the anteriormost bones of the palatal complex.They are paired elements lying on either side of the midline and articulating with themaxillae and premaxilla anteriorly, the septomaxillae dorsally, and the palatines posteriorly.The shape of the vomers differs both among iguanine genera and among iguanines andvarious outgroups; however, I was unable to partition this variation into character statesand to assess its polarity.Palatines (Figs. 5B, 16, 17, 18). Just posterior to the vomers lie the paired palatines,which form the major portion of the palate. These bones also make up the anterior floor ofthe orbits and the posterior floor of the nasal capsules. Each palatine has three processes:the vomerine process anteriorly, the maxillary process laterally, and the pterygoid processposteriorly.In Brachylophus, Ctenosaura, Cyclura, Dipsosaurus, Iguana, and Sauromalus, as wellas in basiliscines, crotaphytines, oplurines and morunasaurs, the vomerine process of thepalatine bears a low ridge that extends longitudinally along its dorsomedial edge (Fig.16A). This ridge bends laterally at the posterior end of the nasal capsule. In place of thelow ridge, Amblyrhynchus and Conolophus have a high crest and thus greater bonyseparation of the nasal capsules (Fig. 16B).Behind the maxillary process, the palatine forms the medial border of the suborbitalfenestra (inferior orbital foramen of Oelrich, 1956). In this region, the palatines of someSauromalus differ from those of all other iguanines and all outgroups examined in thisstudy in that their lateral borders are fringed along the anterior margins of the suborbitalfenestrae. I have observed this feature in several species of Sauromalus, but its presencealways appears to be variable.The infraorbital foramen, which transmits the superior alveolar nerve and artery(Oelrich, 1956), pierces the anterior orbital wall in the region of the maxillary process ofthe palatine. The exact position of the foramen relative to this process varies within
40 University ofCalifornia Publications in ZoologyB
FIG. 16. Posterodorsal views of disarticulated right palatines of (A) Iguana delicatissima (MCZ 75388)and (B) Conolophus subcristatus (AMNH 110168), contrasting the low dorsomedial ridge in the formerwith the high dorsomedial crest in the latter. Abbreviations: c, dorsomedial crest; mp, maxillary process;ptp, pterygoid process; r, dorsomedial ridge; vp, vomerine process.
iguanines, and five categories can be recognized (Fig. 17): (1) entirely within the palatinewithout a suture connecting the foramen to the lateral edge of the palatine (Fig. 17A); (2)within the palatine with a suture running from the foramen to the lateral edge of the palatine(Fig. 17B) (individuals falling within this category vary considerably in the distance fromthe foramen to the lateral edge of the palatine and thus the length of the suture); (3) betweenthe palatine and the jugal (Fig. 17C) (sometimes the lacrimal also contributes to the borderof the foramen); (4) between palatine and maxilla with the portion of the palatine directlyposterior to the foramen large but not extending laterally to contact the jugal (Fig. 17D); (5)between palatine and maxilla, with the portion of the palatine directly posterior to theforamen small or absent (Fig. 17E).Intrageneric and intraspecific variation in the position of the iguanine infraorbitalforamen is great: many genera and species exhibit more than one of the five conditionsdescribed above. Nevertheless, sufficient differences exist among genera that some of thevariation can be used in phylogenetic analysis. Most iguanines (Amblyrhynchus,Conolophus, Ctenosaura, Cyclura, Iguana, and Sauromalus) commonly exhibit condition3, in which the infraorbital foramen is located at the suture between palatine and jugal.Conolophus, Cyclura, and Iguana are relatively constant in the position of this foramen:the great majority of individuals in these three genera exhibit condition 3. Amblyrhynchus,Ctenosaura, and Sauromalus are more variable. Though condition 3 occurs commonly inall three genera, in Amblyrhynchus and Ctenosaura the infraorbital foramen of manyindividuals is entirely within the palatine. A suture connecting the foramen to the lateral
CO
CQ c<3 ^
LU
4>
N
e3
42 University of California Publications in Zoology
edge of the palatine is generally present (condition 2). Sauromalus obesus is similar toAmblyrhynchus and Ctenosaura in this regard, but specimens of S. ater, S. hispidus, andS. varius exhibit condition 4, in which the maxilla contributes to the ventral rim of theforamen, rather than condition 2 (samples of other species of Sauromalus are too smallupon which to base generaUzations).Brachylophus and Dipsosaurus are unique among iguanines in the positions of theirinfraorbital foramina, though at different ends of the morphological spectrum. InBrachylophus, the infraorbital foramen is entirely within the palatine. A suture extendingfrom the foramen to the lateral edge of the maxillary process of the palatine (condition 2)was observed in all four B. vitiensis examined but was absent (condition 1) in over half ofthe specimens of B.fasciatus. Dipsosaurus is the only iguanine that commonly exhibitscondition 5, in which the infraorbital foramen emerges between palatine and maxilla. Insome specimens, a small posteriorly or laterally directed process is present at the medialedge of the foramen; in others it is absent. When present, the process is smaller than thatseen in other iguanines (some Sauromalus) in which this process fails to contact the jugallaterally.Because of the high intrageneric variation in the position of the infraorbital foramen, Irecognized three characters each with one apomorphic state rather than one character withfour or five: one for the size of the portion of the palatine immediately posterior (orposteromedial) to the infraorbital foramen, a second for the presence or absence of contactbetween this part of the palatine and the jugal, and a third for whether or not the infraorbitalforamen lies entirely within the palatine.The infraorbital foramina of the four outgroups examined in this study generally differfrom any of those seen in iguanines. Basiliscines and morunasaurs exhibit a conditionsimilar to that described above as condition 4, but the process of the palatine at the medialedge of the infraorbital foramen is directed posteriorly rather than laterally (Fig. 17F).Chalarodon and some Oplurus possess condition 5, while other Oplurus possess thecondition described for basiliscines and morunasaurs. Individual crotaphytines may alsoexhibit the basiliscine-morunasaur condition, but in other individuals the infraorbitalforamen is located between palatine and jugal as in some iguanines. In the latter case,however, the contact of the posteriorly directed process of the palatine with the jugal resultsfrom extensive medial development of the jugal, rather than from lateral extension of theprocess of the palatine as in iguanines.The differences between iguanines and the four outgroups indicate either that somemorphological change occurred between the most recent common ancestor of iguanines andtheir closest relatives among these four outgroups or that no living iguanine species ischaracterized by the condition that was present in the most recent common ancestor of thegroup (though some individual specimens may be). Nevertheless, differences betweeniguanines and the outgroups are minor enough that the polarities of all three characters canbe assessed. Because no iguanines possess the same morphology of the infraorbitalforamen seen in the outgroups, no iguanine is scored plesiomorphic for all three characters.
Phylogenetic Systematics oflguanine Lizards 43
FIG. 18. Ventral view of the skull oi Iguana delicatissima (MCZ 16157), showing the medial curvatureof the pterygoids and concomitant abrupt narrowing of the pyriform recess. Scale equals 1 cm.Abbreviations: pa, palatine; pt, pterygoid; vo, vomer.
Pterygoids (Figs. 5B,C, 6A, 18). These paired bones are the posteriormost palatalelements. Each pterygoid bears three processes: an anteriorly directed palatine process, ananterolaterally directed transverse process, and a longitudinally compressed and posteriorlydirected quadrate process. The ventral surface of the palatine process often bears smallteeth. Anterior to the pterygoid notches, where the basipterygoid processes of thebasisphenoid articulate with the pterygoids, the medial edges of the pterygoids of mostiguanines curve towards the midline, resulting in a sudden narrowing of the pyriformrecess (interpterygoid vacuity) (Fig. 18). In contrast, the medial edge of the pterygoids inBrachylophus is relatively straight, and the pyriform recess narrows more gradually fromposterior to anterior (Fig. 5B).Outgroup comparison suggests that the condition seen in Brachylophus isplesiomorphic. Among the four outgroups, only crotaphytines exhibit the strongly curved
44 University of California Publications in Zoology
medial borders of the pterygoids, though a moderate curvature occurs in some oplurines.Thus, depending upon the relationships among ingroup and outgroups, either the polarityof this character will be equivocal, or the interpretation that the relatively straight medialborder of the pterygoids is plesiomorphic will be favored.Ectopterygolds (Figs. 5A,B, 6A). Each ectopterygoid lies at the posterior margin ofthe suborbital fenestra forming a brace between the jugal and maxilla anterolaterally and thepterygoid posteromedially. Near the posteromedial comer of the suborbital fenestra, theectopterygoid may contact the palatine, usually on the dorsal surface of the palatal bones.Contact between ectopterygoid and palatine in this region is the common condition only inConolophus among iguanines, and occurs in about half of the Iguana delicatissimaexamined. This contact occurs rarely in some other iguanine species. Ectopterygoid-palatine contact in this region was not observed in any of the four outgroups and istherefore considered apomorphic.The ectopterygoid may also contact the palatine near the anterolateral comer of thesuborbital fenestra. This condition is clearly derived for iguanines on the basis of outgroupcomparison, but does not appear to be characteristic of any iguanine species. OnlyAmblyrhynchus exhibits the anterolateral ectopterygoid-palatine contact commonly, buteven here it occurs in less than half of the specimens examined. Because the apomorphicstate of this character is not characteristic of any iguanine species and because diagnosticapomorphies of Amblyrhynchus are plentiful, I have chosen to ignore this character in thephylogenetic analysis. BRAINCASEThe iguanine braincase (Figs. 5A,B, 6A), or neurocranium, is composed of four pairs ofendochondral bones-orbitosphenoids, prootics, opisthotics, and exoccipitals-and threeunpaired ones-basisphenoid, basioccipital, and supraoccipital. The parasphenoid, a dermalbone, is also described here because of its intimate association with the basisphenoid.Parasphenoid and basisphenoid as well as exoccipitals and opisthotics are fused to eachother even in juveniles, and all other elements except orbitosphenoids fuse withneighboring braincase elements late in ontogeny. In some very large specimens, even theorbitosphenoids are fused with one another. Although the stapes and epipterygoids aresplanchnocranial elements, they are included in this section because of their closeassociations with the braincase.Orbitosphenoids (Fig. 19). The orbitosphenoids are paired, crescent-shaped boneslying within the membranes that separate the brain cavity from the orbits. Eachorbitosphenoid is continuous with five orbital cartilages: the septal cartilage and planumsupraseptale anterodorsally, the pila accessoria and pila antotica posterodorsally, and thehypochiasmatic cartilage ventrally (Oelrich, 1956). Although consistent differences in theshapes of the orbitosphenoids exist between iguanine taxa, these differences seem to berelated to differences in body size. In large iguanines, the orbitosphenoids undergoconsiderable ontogenetic changes in shape resulting from progressive outward ossification
Phylogenetic Systematics ofIguanine Lizards 45
FIG. 19. Anterolateral views of the left orbitosphenoids of three Iguana iguana-(A) RE 454, (B) JMS245, (C) JMS 713-showing ontogenetic change in the shape of these bones resulting from progressiveossification outward along the orbital cartilages. Scale equals 1 mm. Abbreviations: he, hypochiasmaticcartilage; pac, pila accessoria; pan, pila antotica; pis, planum supraseptale.
along the orbital cartilages (Fig. 19). Thus, the posterodorsal edge of each orbitosphenoidfirst develops a posterior process where it joins the pila accessoria and pila antotica, andthis process later bifurcates following the two diverging orbital cartilages. The ventral andanterodorsal ends of the bone elongate by a similar process and, in the case of the latter, thetwo orbitosphenoids may eventually meet and fuse at the midline. Small iguaninesgenerally fail to develop the bifurcating posterodorsal processes of the orbitosphenoidsseen in adults of larger species, and I have never observed medial fusion of the two bonesat their anterodorsal ends in small iguanines.Epipterygoids (Fig. 6A). The epipterygoids are thin, rod-shaped bones extending fromthe palate to the skull roof. Ventrally, the epipterygoids sit in depressions in the dorsalsurfaces of the palatines, but dorsally their articulations with the parietal are either weak orlacking. I found no differences in epipterygoid morphology among iguanine genera thatmight serve as systematic characters.Prootics (Fig. 6A). The paired prootics form the lateral walls of the neurocranium.They are sutured to the supraoccipital dorsomedially, to the exoccipitals posteriorly, to thebasioccipital posteroventrally, and to the basisphenoid ventromedially. Although the
46 University of California Publications in ZoologyB
FIG. 20. Ventral views of the posterior portion of the palate and anterior portion of the braincase of (A)Sauromalus varius (RE 308) and (B) Amblyrhynchus cristatus (RE 1508), showing differences in the lengthof the parasphenoid rostrum. Scale equals 1 cm. Abbreviations: bptp, basipterygoid process; bs,parabasisphenoid; pr, pyriform recess; ps, parasphenoid rostrum; pt, pterygoid.
morphology of the prootics is complex, I have found no characters in these bones thatmight serve to elucidate relationships among the basic taxa used in this study.Parabasisphenoid (Figs. 5B, 6A, 20, 21). Because the parasphenoid and basisphenoidof iguanines are always fused postembryonically, I describe them as a single element. Theparasphenoid rostrum extends anteriorly like a thin, flat blade from the main body of theparabasisphenoid on the midline. Compared to those of all other iguanines as well as thoseof basiliscines, crotaphytines, morunasaurs, and oplurines, the parasphenoid rostrum ofAmblyrhynchus is relatively short (Fig. 20). Even the parasphenoid rostra of other short-skulled taxa, such as the basiliscine Corytophanes, are much longer.The main body of the parabasisphenoid is an unpaired median bone that forms theanterior floor of the brain cavity. It is sutured with the prootics laterally and with thebasioccipital posteriorly. Anterolaterally, two large basipterygoid processes meet theanteromedial surfaces of the quadrate processes of the pterygoids at the pterygoid notches,forming a movable joint between palate and braincase.Boulenger (1890) first noted variation in the form of the parabasisphenoid (Fig. 21)among different iguanines. In most iguanines, the ventrolateral edges of theparabasisphenoid, the cristae ventrolaterals, are strongly constricted behind the
Phylogenetic Systematics oflguanine Lizards 47
oc
FIG. 21. Ventral views of the neurocrania of (A) Sauromalus varius (RE 451), (B) Ctenosaurahemilopha (RE 325), (C) Iguana iguana (RE 1006), and (D) Cyclura nubila (RE 337), showing differencesin the width of the parabasisphenoid and the size of its posterolateral processes. Scale equals 1 cm.Abbreviations: bo, basioccipital; bs, parabasisphenoid; eo, exoccipital-opisthotic; oc, occipital condyle;pro, prootic; ps, parasphenoid rostrum; sot, spheno-occipital tubercle.
basipterygoid prcx;esses, giving the ventral outline of the braincase roughly the shape of anhourglass (Fig. 21A,B). In contrast, the cristae ventrolaterales of Iguana are widelyseparated, extending in almost straight lines from the basipterygoid processes posteriorly tothe spheno-occipital tubercles and giving the ventral outline of the braincase the shape of abox (Fig. 21 C). Cyclura is variable in this character, though all species have relativelybroad parabasisphenoids (Fig. 2 ID) compared to those of most other iguanines. C.carinata has the narrowest basisphenoid, while that of C. pinguis is at least as wide as thatof some Iguana delicatissima; other species are intermediate. In at least some of those
48 University of California Publications in Zoology
Cyclura with wide parabasisphenoids, this bone becomes relatively wider duringpostembryonic ontogeny. All basiliscines, crotaphytines, morunasaurs, and oplurines havethe parabasisphenoid strongly constricted behind the basipterygoid processes, indicatingthat this condition is plesiomorphic for iguanines.A second part of the iguanine parabasisphenoid exhibits two distinct morphologies thatare constant within genera. The parabasisphenoids of all iguanines except Ctenosaura bearlarge posterolateral processes that extend along the anterolateral edges of the lateralprocesses of the basioccipital, reaching or closely approaching the spheno-occipitaltubercles (Fig. 21A,C,D). In Cyclura (Fig. 21D) and especially in Iguana (Fig. 21C),widening of the parabasisphenoid obliterates the distinctness of its posterolateral processes;their existence is inferred from the lateral extent of the parabasisphenoid along the lateralprocesses of the basioccipital. Unlike other iguanines, the posterolateral processes of theparabasisphenoid are very short or absent in Ctenosaura (Fig. 2 IB), a condition that maybe related to the elongation of the skull in this taxon. Only Crotaphytus (but not Gambelia)among the outgroups examined exhibits a condition similar to that of Ctenosaura; therefore,I considered the possession of long posterolateral processes of the parabasisphenoid to beplesiomorphic.Basioccipital (Figs. 5B, 21). The basioccipital forms the posterior floor of the braincavity and makes up the large medial portion of the occipital condyle. It bears prominentventrolaterally directed lateral processes that are capped by the spheno-occipital tubercles.These tubercles fuse to the lateral processes late in ontogeny. The basioccipital is suturedto the exoccipitals dorsolaterally, to the prootics anterolaterally, and to the parabasisphenoidanteriorly. Although iguanine basioccipital morphology is variable, I found no obviouscharacters that bear on intergeneric relationships.Exoccipitals and Opisthotics (Figs. 5C, 21). The exoccipitals are indistinguishablyfused to the opisthotics in postembryonic developmental stages of all iguanines. Thesecompound bones form the posterior sides of the brain cavity and the lateral edges of theforamen magnum. They meet the supraoccipital dorsomedially, the prootics anteriorly, andthe basioccipital ventromedially. The paroccipital processes of the opisthotics extendlaterally to contact the supratemporals, bracing the posterolateral comers of the skull. Therelative length of the paraoccipital processes varies among iguanine genera, but differencesare complicated by positive allometry of this feature both within and among taxa (thoughthe correlation is less precise in the latter case). Apparently the braincase widens moreslowly than the skull as a whole. As the paraoccipital processes elongate, they also becomemore posteriorly oriented.Each exoccipital-opisthotic bears two prominent crests laterally: the cristainterfenestralis, which lies between the fenestra ovalis and the fenestra rotunda; and thecrista tuberalis, which bounds the antrum of the fenestra rotunda posteriorly. Variationexists in the degree to which the crista tuberalis slants inward dorsally and to which itobscures the crista interfenestralis in posterior view, but this variation is too great withiniguanine genera to be useful for inferring relationships among them. Dipsosaurus is uniqueamong iguanines in possessing a sharp, laterally directed point on each crista
Phylogenetic Systematics oflguanine Lizards 49
interfenestralis. Although this process is absent or very small in basiliscines,crotaphytines, morunasaurs, and oplurines, it is present in some sceloporines. I considerDipsosaurus and these sceloporines to have developed a pointed process on the cristainterfenestralis convergently.Stapes. The stapes, or columella, is a sound-transmitting bone that extends from thefenestra ovalis (foramen ovale of Oelrich, 1956) in the braincase to a point just behind theposterior crest of the quadrate. In Ufe it is attached to the external tympanic membrane via acartilaginous extracolumella, which is often damaged during skeletal preparation. Thestapes of Amblyrhynchus is robust compared to those of all other iguanines and most otheriguanids, although some sceloporines also have a thick stapes (Axtell, 1958; Earle, 1962).MANDIBLE
There are seven bones present in the mandibles of all iguanines (Fig. 6B,C); from anteriorto posterior these are: dentary, splenial, coronoid, angular, surangular, prearticular, andarticular. The articular is a splanchnocranial endochondral bone; the remaining bones aredermal. In some noniguanine iguanids, either splenial or both splenial and angular may beabsent (Etheridge and de Queiroz, 1988).Dentary (Figs. 6B,C, 22). The dentary is the anteriormost bone in the mandible andextends posteriorly to about the level of the apex of the coronoid. It is the only tooth-bearing bone in the lower jaw. Anterior to the splenial, Meckel's cartilage, which extendsfrom the articular bone to the anterior end of the mandible, is completely enclosed in a bonytube formed by the dentary. In some other iguanids (e.g., morunasaurs) the groove forMeckel's cartilage is completely open lingually, while in others (e.g., crotaphytines) thedorsal and ventral edges of the groove meet to close the tube but remain separated by asuture. In one late embryo of Amblyrhynchus (SDNHM 45156), Meckel's groove isclosed but retains a suture; however, in all postembryonic iguanines the upper and lowerdentary portions of Meckel's groove are closed and fused.A series of mental foramina are positioned along the labial face of the anterior half ofthe dentary. In all iguanines except Amblyrhynchus and in all outgroups examined, theseforamina lie in a line about halfway between the dorsal and ventral edges of the dentary,and the dorsal edge of the dentary where it meets the coronoid is approximately level withthe dorsal border of the surangular just posterior to the coronoid (Fig. 22A). The dorsalborder of the dentary in Amblyrhynchus is high, well above the level of the dorsal borderof the surangular, and the row of mental foramina lies more than halfway down the labialsurface of the dentary (Fig. 22B).Splenial (Fig. 6B,C, 23). The exposed portion of the splenial is roughly diamond-shaped and lies on the lingual face of the mandible wedged into the posterior end of thedentary. Posterodorsally, the splenial contacts the coronoid and the surangular;posteroventrally it is bounded by the angular. The relative size of the splenial is variable iniguanines, with that of Sauromalus being smaller than those of the other genera. Althoughthere is considerable variation in the size of the splenial among the four outgroups used in
50 University of California Publications in Zoology
FIG. 22. Lateral views of the right mandibles of (A) Iguana delicalissima {MCL 60823) and (B)Amblyrhynchus cristatus (RE 1396), showing differences in the relative heights of the dentary (den) andsurangular (sur) and in the position of the row of mental foramina (mf). Scale equals 1 cm.
this study, all have a relatively larger splenial than Sauromalus. Therefore, I consider asmall splenial to be apomorphic for iguanines.The anterior inferior alveolar foramen pierces the mandible on its lingual surface at apoint between one-third and one-half the way back from the anterior end of the jaw (Fig.23). In most iguanines, this foramen lies within the suture between the splenial and thedentary at the anterior end of the splenial or along its anterodorsal edge. The coronoid mayextend anteriorly between splenial and dentary so that it forms the posterior margin of theanterior inferior alveolar foramen (Fig. 23A) in some Brachylophus, Dipsosaurus, andSauromalus. Varying amounts of this anterior extension of the coronoid may be coveredby the splenial lingually, excluding the coronoid from the border of the foramen (Fig.23B). This condition occurs in Conolophus, Ctenosaura, Iguana, most Cyclura, and insome Brachylophus, Dipsosaurus, and Sauromalus. In Brachylophus, the splenial istruncated, and the anterior inferior alveolar foramen sometimes lies entirely within thedentary. In Amblyrhynchus, the coronoid extends far anteriorly, and the foramen liesbetween it, rather than the dentary, and the splenial (Fig. 23C).
Phylogenetic Systematics oflguanine Lizards 51
an amf
den
amf
amf
aiaf
FIG. 23. Lingual views of the left mandibles of (A) Sauromalus varius (RE 512), (B) Iguanadelicatissima (MCZ 60823), and (C) Amblyrhynchus crisiatus (RE 1091), showing differences in the bonesthat surround the anterior inferior alveolar foramen. Scale equals 0.5 cm. Abbreviations: aiaf, anteriorinferior alveolar foramen; amf, anterior mylohyoid foramen; an, angular; cor, coronoid; den, deniary; pre,prearticular; sp, splenial.
52 University of California Publications in Zoology
B
FIG. 24. Lateral views of the right mandibles of (A) Conolophus pallidas (RE 1382) and (B) Cycluracornuta (RE 383), showing differences in the size of the labial process of the coronoid (shaded). Scaleequals 1 cm.
Basiliscines, crotaphytines, morunasaurs, and oplurines have their anterior inferioralveolar foramina either between splenial and dentary or entirely within the splenial. Bothconditions are found in all four outgroups. The splenial is relatively larger in most of theseoutgroups than in any iguanine, which may account for the fact that the foramen ofiguanines does not lie entirely within this bone. Because location of the anterior inferioralveolar foramen between splenial and dentary is the only condition that occurs in bothingroup and outgroups, I considered this to be the plesiomorphic condition. The other twopositions of the foramen, entirely within the dentary and between coronoid and splenial,were considered to be separate modifications of the plesiomorphic condition.Coronoid (Figs. 6B,C, 23, 24). This bone forms a large dorsal process (coronoideminence) immediately posterior to the tooth row, which serves as the insertion for jawadductor muscles. It also bears one lateral and two medial ventrally directed processes thatstraddle the body of the lower jaw. Ahhough absent in many iguanids, the large process ofthe coronoid that extends over the labial surface of the mandible is present in all iguanines(Fig. 24). This labial extension of the coronoid is most strongly developed in adultConolophus, in which its ventral border reaches halfway or farther down the mandible andcovers the posterolateral end of the dentary (Fig. 24A). In most other iguanines, the labialprocess of the coronoid is relatively small (Fig. 24B), but in Amblyrhynchus andBrachylophus the size of the process is intermediate between that of Conolophus and thoseof other iguanines. In both Amblyrhynchus and Conolophus the labial process of thecoronoid is relatively small at hatching and increases in size during postembryonicontogeny. The labial process of the coronoid is very small in basiliscines, crotaphytines,and oplurines. Morunasaurs and other iguanids that possess a large labial process of the
Phylogenetic Systematics oflguanine Lizards 53
i?Ocm) (Tttirffma
> "
FIG. 25. Laterial views of the right mandibles of (A) Iguana delicatissima (MCZ 60823), (B)Sauromalus obesus (RE 467), and (C) Amblyrhynchus cristatus (RE 1396), showing differences in thelateral exposure of the angular (shaded). Scale equals 1 cm.
54 University of California Publications in Zoology
coronoid have a relatively slight ventral extension of this process compared toAmblyrhynchus, Brachylophus, and especially Conolophus.Angular (Fig. 6B,C, 25). The angular is located on the ventral surface of the mandible,forming sutures with the splenial anterodorsally and the prearticular posterodorsally on thelingual surface of the mandible and with the dentary anteriorly and the surangularposteriorly on the labial side. In Brachylophus, Ctenosaura, Cyclura, Dipsosaurus, andIguana, the angular extends far up the labial surface of the mandible so that it is easily seenin lateral view (Fig. 25A). The angulars of Amblyrhynchus, Conolophus, and Sauromalusare restricted labially so that they are barely visible from the lateral side (Fig. 25B,C).Compared to those of other iguanines, the angular of Sauromalus is relatively narrow.Because the angulars of basiliscines, crotaphytines, morunasaurs, and most oplurines arewide posteriorly and extend far up the labial surface of the mandible, I considered these tobe plesiomorphic conditions. In Oplurus, the width and labial exposure of the angular arevariable owing to varying degrees of reduction in this bone.Surangular (Fig. 6B,C, 26, 27). This bone forms the dorsal portion of the mandibleposterior to the coronoid and anterior to the articular facet. It fuses with the prearticular latein ontogeny. Dorsal to its suture with the angular on the labial surface of the jaw, theanterior extent of the iguanine surangular is variable (Fig. 26). In Amblyrhynchus,Brachylophus, and Dipsosaurus the exposed part of the surangular barely extends to thelevel of the apex of the coronoid, being covered by the dentary anterior to this level (Fig.26A,B). In Conolophus, it extends slightly farther, to the level of the anterior slope of thecoronoid eminence. The surangulars of Iguana and Cyclura extend far forward, wellbeyond the anterior slope of the coronoid eminence and often anterior to several of theposteriormost dentary teeth (Fig. 26C). Sauromalus and Ctenosaura are intermediate andvariable within species; the surangular in each of these genera usually extends beyond theanterior slope of the coronoid eminence, but falls short of the tooth row. Some membersof both genera exhibit a condition similar to that of Conolophus, and some Ctenosaura havea surangular that extends beyond the posteriormost dentary tooth.Although the outgroups used in this study are also variable in the anterior extent of thesurangular, in none does it extend as far forward as in Iguana and Cyclura. Therefore, inthe absence of other information, it seems that a great anterior extent of the surangular is asynapomorphy of these two taxa. If the basic taxa used in this study are monophyletic,then a similar condition seen in some Ctenosaura must either be convergent, or thecharacter may have arisen initially as a polymorphism, or some Ctenosaura have reverted tothe ancestral morphology.On the lingual side of the mandible, ventral to the apex of the coronoid in the archbetween the ventral feet of this bone, a small portion of the surangular is variably visible iniguanines (Fig. 27). In most iguanines, this part of the surangular is relatively large andhas the shape of a dome above the prearticular (Fig. 27A). In Amblyrhynchus,Conolophus, and Cyclura cychlura, the prearticular extends further dorsally, eithercompletely excluding the surangular from the lingual surface of the mandible (Fig. 27B) orleaving only a thin sliver of it exposed. Although few small specimens were examined.
Phylogenetic Systematics of Iguanine Lizards 55
B
FIG. 26. Lateral views of the right mandibles of (A) Dipsosaurus dorsalis (RE 359), (B) Brachylophusvitiensis (MCZ 160254), and (C) Iguana iguana (RE 453), showing differences in the anterior extent of thesurangular (shaded). Scale equals 0.5 cm.
there appears to be a transformation of this part of the surangular from exposed tounexposed during the postembryonic ontogenies of Amblyrhynchus and Conolophus.Some intraspecific variation exists in this feature; but other than the taxa in which theunexposed portion of the surangular is the common condition, only in Brachylophus
56 University of California Publications in Zoology
FIG. 27. Medial views of the left mandibles of (A) Iguana delicatissima (MCZ 16157) and (B)Conolophus subcristatus (MVZ 77314), showing differences in the exposure of the surangular (shaded)below the coronoid (cor). Scale equals 1 cm.
fasciatus, Cyclura nubila, and Sauromalus varius does this condition appear to be morethan a rare variant.Except for Corytophanes and Oplurus quadrimaculatus, all outgroups examined have arelatively large, dome-shaped portion of the surangular visible lingually between the ventralfeet of the coronoid. In Corytophanes, however, lingual restriction of the surangularresults from ventral extension of the coronoid rather than dorsal extension of theprearticular, the condition in iguanines. For this reason, as well as the hypothesis thatBasiliscus rather than Corytophanes is the sister group of the other two basiliscine genera(Etheridge and de Queiroz, 1988), I considered the superficially similar conditions seen inCorytophanes and in some iguanines to be nonhomologous. Thus, the large lingualexposure of the surangular between coronoid and prearticular is interpreted asplesiomorphic.Prearticular (Figs. 6B,C, 28, 29). This bone forms the ventromedial portion of theposterior end of the mandible. The prearticular bears two processes for the insertion ofjawadductor and abductor muscles, the posteriorly directed retroarticular process and themedially directed angular process. The retroarticular process is large in all iguanines, butthe relative size of the angular process is variable. In all iguanines except Amblyrhynchus,the angular process is small at hatching and increases in relative size as the animal grows(Fig. 28A-C). The angular process of Amblyrhynchus is very small in juveniles andincreases in relative size only slightly during postembryonic ontogeny (Fig. 28D-F); evenin large adults it has only about the same relative size as those of young of other iguaninegenera.Except for Corytophanes and Laemanctus, all outgroup taxa examined (including thosethat are small as adults) have relatively large angular processes. Thus, if basiliscines arethe sister group of iguanines, then the polarity of this character is equivocal; if not, then thedevelopment of a large angular process during ontogeny must be considered to beplesiomorphic. Because Amblyrhynchus exhibits the nontransforming ontogeny, strict
LU
o
m
58 University of California Publications in Zoology
B
tc ita\ mc
(^ Of
Phylogenetic Systematics oflguanine Lizards 59
Dipsosaurus, but in this taxon the posterior ends of the crests move apart during ontogenyso that the retroarticular process of large Dipsosaurus is quadrangular (Fig. 29).Most outgroups have a triangular retroarticular process, much like those seen in themajority of iguanines; however, I have observed quadrangular retroarticular processes inMorunasaurus annularis and Enyalioides praestabilis. Thus, either the quadrangularretroarticular process of Dipsosaurus is apomorphic or the polarity of this character isequivocal, but a quadrangular retroarticular process will never be considered to beplesiomorphic with the outgroups used in this study.The medial crest of the retroarticular process varies in size within Iguaninae. InAmblyrhynchus (Fig. 28D-F), Brachylophus, Conolophus, and Cyclura cornuta, thisstructure is but a low, rounded ridge, contrasting with the sharp crest seen in otheriguanines (Figs. 28A-C, 29). Intraspecific variability in Amblyrhynchus and Conolophus,but more important, variation within basiliscines, morunasaurs, and oplurines, preventedme from using the size of the medial crest as a character for phylogenetic analysis.Articular (Figs. 6C, 28, 29). The articular bone is the ossified posterior end ofMeckel's cartilage and forms the condyle that articulates with the quadrate of the skullproper. It sits in a groove in the dorsal surface of the jaw between the prearticularposteriorly and medially and the surangular anterolaterally. The articular of iguanines fusesto the prearticular around the time of hatching. I have not studied variation in the iguaninearticular. MISCELLANEOUS HEAD SKELETON
Marginal Teeth (Figs. 5B, 8, 30). The marginal teeth of iguanines exhibit a bewilderingdiversity of form and could easily be the subject of a study by themselves. Somedentitional features common to all iguanines are pleurodonty and the formation ofreplacement teeth directly lingual to the teeth being replaced (iguanid tooth-replacementpattern of Edmund, 1960). Although lizards are often stereotyped as being homodont, alliguanines exhibit some regional differentiation in the morphology of their marginal teeth.This differentiation is most pronounced, at least in terms of crown morphology, in Cycluraand Sauromalus, where the crowns of the anterior teeth are conical and usually lack lateralcusps while those of the posterior teeth are laterally compressed and polycuspate. Anotherfeature common to all iguanines is an allometric increase in tooth number within species, afeature that has been reported previously in iguanines (Ray, 1965; Montanucci, 1968) andin various other iguanids (Etheridge, 1962, 1964b, 1965a; Ray, 1965). This allometricincrease in tooth number results from the addition of teeth to the posterior ends of themaxillary and dentary tooth rows; the number of premaxillary teeth remains constant.Variation in the number of premaxillary teeth of iguanines is given in Table 3. Most orall species of Amblyrhynchus, Brachylophus, Conolophus, Ctenosaura, Dipsosaurus, andIguana have a statistical mode of seven premaxillary teeth. The species of Cycluragenerally have modes of greater than seven premaxillary teeth, and those of Sauromalushave modal numbers lower than seven. Ctenosaura defensor also has fewer than seven
60 University of California Publications in Zoology
TABLE 3. Numbers of Premaxillary Teeth
Phylogenetic Systematics oflguanine Lizards 61
premaxillary teeth. Two specimens of Cyclura pinguis have seven and eight premaxillaryteeth. I have assumed that C. pinguis actually has a modal number of premaxillary teethgreater than seven and that the bimodal distribution results from sampling error. It is alsopossible that a phylogenetic transformation has occurred within Cyclura and that thesynapomorphic condition applies to a subset of this taxon, or that the ancestral conditionwas polymorphic.Outgroup comparison yields equivocal results conceming the plesiomorphic number ofpremaxillary teeth in iguanines. Gambelia has the condition found in most iguanines, amode of seven premaxillary teeth. Other outgroup s have seven or more premaxillary teeth(basiliscines, Enyalioides); more than seven (Morunasaurus); fewer than seven (oplurines,Hoplocercus); or a range from fewer than seven to more than seven (Crotaphytus, mode ofsix). Because of this ambiguity, I withheld a decision on the primitive number ofpremaxillary teeth and used the character only at a level less inclusive than all iguanines.In most iguanines the premaxillary teeth, as well as the anterior maxillary and dentaryteeth, have fewer or smaller cusps than the posterior maxillary and dentary teeth. InCyclura and most species of Ctenosaura the premaxillary teeth and the dentary teeth withwhich they occlude lack lateral cusps. At least some of the premaxillary teeth of somespecimens have one or more lateral cusps in Brachylophus, Dipsosaurus, and Ctenosaurapalearis, although these lateral cusps are relatively small. Amblyrhynchus and Conolophusalmost invariably have two large lateral cusps on their premaxillary teeth. The premaxillaryteeth of basiliscines, crotaphytines, morunasaurs, and opliuines usually lack lateral cusps,though small ones may occasionally be present.Except in large Ctenosaura, in which the anterior maxillary teeth and the dentary teethoccluding with them are enlarged and recurved to form fangs, these teeth differ onlyslightly from the marginal teeth anterior to them. Moving posteriorly along the marginaltooth rows, the tooth crowns progressively become more laterally compressed, the size ofthe lateral cusps increases, and in most iguanines additional lateral cusps are added. Part ofthe progression is reversed abruptly at the posterior ends of the tooth rows. When stronglycompressed, the crowns of the teeth are much wider than their bases and overlap theirneighbors in a regular pattern: each tooth is twisted about its long axis so that its anterioredge is lingual to and its posterior edge is labial to the crowns of the adjacent teeth.Maximum cuspation is reached about three-fourths of the way back along the tooth row inadults, and here substantial differences exist among taxa (Fig. 30). The maximum numberof cusps on the marginal teeth oi Brachylophus, Conolophus, Dipsosaurus, and mostCtenosaura (C. acanthura, C. clarki, C. hemilopha, C. palearis, C. pectinata, and C.similis) is four: two anterior cusps, an apical cusp, and one posterior cusp (Fig. 30A).This crown morphology is seen in both maxillary and dentary teeth. The size andoccurrence of the anteriormost cusp, however, is variable, and it may be absent from allteeth in some specimens of some species.Greater cuspation is found in Ctenosaura defensor, Cyclura, and Sauromalus (Fig.BOB). The maximum number of cusps per tooth in these taxa ranges from as few as five inCyclura pinguis and some C. cychlura up to about 10 in C. cornuta and C. nubila.
62 University of California Publications in Zoology
FIG. 30. Lingual views of left maxillary teeth of (A) Conolophus pallidas (RE 1382), (B) Sauromalusvarius (RE 539), (C) Iguana iguana (JMS 1028), (D) Basiliscus plumifrons (RE 427), and (E)Amblyrhynchus cristatus (RE 1387), showing differences in cuspation. Scale equals 1 mm.
Increase in cuspation is accompanied by a difference in the morphology of the maxillaryversus dentary teeth: maxillary teeth bear more cusps along their anterior edges, whilecuspation of the dentary teeth is more or less symmetrical (Avery and Tanner, 1964:Fig.3). Within the tooth row of a single organism, increase in cuspation appears to result fromaddition of cusps to the anterior and posterior edges of the crowns.Still greater cuspation occurs in Iguana, reaching an extreme in /. iguana. In this genusthe teeth possess a large number of small cusps, giving them a serrated cutting edge (Fig.30C). The small cusps are difficult to count, especially when worn, but the maximumnumber is greater than 15 in /. delicatissima and greater than 20 in /. iguana. Cuspationincreases both ontogenetically and from anterior to posterior in a single tooth row by twomechanisms: addition of cusps and subdivision of the fields of preexisting ones. Theactual cusps of fully formed teeth are not subdivided, though their fields appear to be whenteeth are compared with their replacements; it is, of course, impossible to have actualsubdivision of cusps from one tooth to the next. Because cuspation increasesontogenetically, the teeth of young Iguana have about as many cusps as do those of some
Phylogenetic Systematics oflguanine Lizards 63
large Cyclura. The maximum number of cusps in mature Iguana, however, is greater thanin any Cyclura.Amblyrhynchus, Ctenosaura bakeri, and C quinquecarinata are the only iguanines thatcharacteristically have a maximum of only three cusps on their marginal teeth. Tricuspidteeth occur throughout the posterior half of the tooth row in juveniles of at least someiguanine species whose teeth later become four-cusped or polycuspate, and they arecommon outside of iguanines, occurring in basiliscines (Fig. 30D), crotaphytines,oplurines, and most morunasaurs (some Enyalioides are polycuspate). For these reasons,tricuspid posterior marginal teeth are judged to be plesiomorphic for iguanines. Themorphology of the tooth crowns in the outgroups, however, differs strikingly from that ofAmblyrhynchus, although it is similar to that of the tricuspid teeth found more anteriorly inthe tooth row or earlier in the ontogeny of other iguanines. In the tricuspid teeth of all thesetaxa, the apical cusp is much larger than each lateral cusp. In Amblyrhynchus, the lateralcusps are very large, each being nearly as large as the apical cusp (Fig. 30E). Theposterior marginal teeth of Ctenosaura quinquecarinata are similar to those seen in manyoutgroup taxa.Ontogenetic data relating to changes in iguanine tooth crown morphology are few, butwhat little are available suggest that the adult morphologies of the marginal tooth crownsrepresent stages in a single transformation series. Tricuspid teeth are judged to beplesiomorphic on the basis of outgroup comparison (see above), and they also occur in thefew hatchling specimens examined of those iguanines that, as adults, have four-cuspedteeth (Conolophus subcristatus, Ctenosaura hemilopha, C. pectinata, C. similis),polycuspate teeth (Cyclura carinata, C. cornuta, C. nubila), and serrate teeth {Iguanaiguana) as adults. Although I have never observed the replacement of four-cusped teeth bypolycuspate or serrate teeth, both Sauromalus and Cyclura (which are polycuspate asadults) normally possess four-cusped teeth in some portion of the tooth row. Thus, alliguanine tooth crown morphologies appear to be part of a single transformation series, withtricuspid teeth in the terminal stage at its plesiomorphic pole. I also propose thatontogenetic transformation to polycuspate teeth is a modification of a transformation tofour-cusped teeth, and that ontogenetic transformation to serrate teeth is a modification ofone to polycuspate teeth.Judging from the high numbers of replacement teeth in Amblyrhynchus, these animalsprobably replace their teeth at higher rates than other iguanines and the members of the fournoniguanine outgroups examined in this study. Presumably related to the high numbers ofreplacement teeth in Amblyrhynchus is a relatively wide alveolar margin on the bonesbearing the marginal teeth.Palatal Teeth (Fig. 31). Palatal teeth in iguanids may be present on the pterygoids andpalatines but never on the vomers. All iguanines lack palatine teeth, which are present(though not invariably) in crotaphytines and oplurines among the outgroups examined. Atleast some specimens of all iguanine species examined in this study have pterygoid teeth,the number and position of which vary considerably among genera. The pterygoid teethgenerally lack lateral cusps (in contrast with the tricuspid pterygoid teeth of some
64 University of California Publications in Zoology
basiliscines) and are directed posteroventrally; the tips of these teeth may also curveposteriorly. In most iguanines, the number of pterygoid teeth increases ontogenetically,though this increase is less conspicuous in species with small maximum numbers ofpterygoid teeth.Pterygoid teeth are present in all four outgroups examined and lie in a single row closeto the ventromedial edge of each pterygoid, next to the pyriform recess. The posterior endof the row may be displaced slightly laterally. This plesiomorphic condition is retained inBrachylophus, and is also seen in some Cyclura and Sauromalus as an individual variant.A modification of this condition seems to have occurred by lateral displacement of theposterior end of the tooth row toward the base of the transverse process of the pterygoid,with an accompanying tendency for this posterior portion of the tooth row to doubleontogenetically. Beneath the posterior end of the tooth row a bony mound may be raised.An ontogenetic transformation from the presumed plesiomorphic condition mirrors thehypothesized phylogenetic transformation of terminal morphologies based on outgroupcomparison. This apomorphic condition is seen in adult Ctenosaura and in some Cycluraand Sauromalus.Two independent phylogenetic transformations appear to have been derived from theapomorphic condition described above. The first, seen in Iguana, results ontogeneticallyand presumably was derived phylogenetically from an increase in the number of pterygoidteeth and a more extensive doubUng of the tooth row late in ontogeny. The second, seen inAmblyrhynchus, apparently resulted from loss of the anterior portion of the tooth row; theremaining teeth are located in a short, laterally displaced patch, even in juveniles.Pterygoid teeth are usually absent in Conolophus and Dipsosaurus (occasionally absentin individual specimens of Sauromalus), but their absence in these two taxa appears torepresent separate derivations from different antecedent conditions. In the rare specimensoi Dipsosaurus that have pterygoid teeth, these teeth are present in a single row near themedial edge of the bone, suggesting derivation from the plesiomorphic condition. Thisinference is complicated by the small size of Dipsosaurus combined with the large size atwhich lateral displacement of the row occurs in taxa that exhibit this derived condition.When pterygoid teeth are present in Conolophus they are located laterally, near the base ofthe transverse process. This suggests that lateral displacement of the posterior end of thetooth row (an apomorphic condition) preceded tooth loss; the reduction of the anterior endof the tooth row seen in Amblyrhynchus is a likely intermediate state.Figure 31 is a hypothetical character phylogeny for the iguanine pterygoid tooth patch.The three most speciose iguanine genera, Ctenosaura, Cyclura, and Sauromalus, exhibitmuch variation in their pterygoid teeth. They are all considered to exhibit one of the twoinitial modifications of the plesiomorphic condition in the diagram, although this treatmentignores much of the actual variation. Because of the complexity of this character, it isnecessary to subdivide it into three characters so that coding will accurately reflect thehypothesized phylogenetic transformations.The number of teeth on a single pterygoid is highly variable among iguanine taxa;however, allometric increase in this feature makes intertaxic comparison difficult among
Phylogenetic Systematics of Iguanine Lizards 65
Conolophus
Iguana
row doubles throughoutincrease in number of teeth
Amblyrhynchus
anterior part of row lost
CtenosauraCycluraSauromalus
row doubles posteriorlyposterior part of row moves laterally
BrachylophusZ' (Cyclura)(Sauromalus )
FIG. 31. Hypothetical character phylogeny for the iguanine pterygoid tooth patch. An asterisk indicatesthat pterygoid teeth are sometimes absent; parentheses indicate a rare condition in the enclosed taxon. Seetext for details.
66 University of California Publications in Zoology
taxa whose organisms reach different sizes. As stated above, Conolophus andDipsosaurus generally lack pterygoid teeth, although I have observed up to two and four ona single pterygoid in these genera, respectively. Amblyrhynchus, Brachylophus, andSauromalus generally have fewer than 10 pterygoid teeth and always have less than 15 (themaximum numbers that I have observed are seven, 11, and 12, respectively). Because ofthe wide range in body size of their included species, Ctenosaura and Cyclura exhibit awide range in pterygoid tooth number. Members of the large species of Ctenosaura (C.acanthura, C. pectinata, C. similis) usually have over 20 pterygoid teeth and sometimesexceed 30. Small species such as C. clarki, C. defensor, C. palearis, and C.quinquecarinata probably never have as many as 20 such teeth. Cyclura exhibits a range inthe number of pterygoid teeth similar to that of Ctenosaura, but I have few adequateontogenetic series for species in the former genus. The most teeth that I have seen on asingle pterygoid in Cyclura is 26 in a specimen of C. pingius that had not yet undergone thefusion of braincase elements indicative of the attainment of maximum size. If allometrictrends in this species are similar to those in Iguana and Ctenosaura, larger organismsprobably have upwards of 30 such teeth. Iguana is characterized by a high pterygoid toothnumber. Large /. delicatissima have a maximum of at least 30 pterygoid teeth, while thenumber exceeds 60 in /. iguana. I did not use variation in pterygoid tooth number as aseparate systematic character, though some of this variation is incorporated in the charactersthat were used.Scleral Ossicles (Fig. 32). The scleral ossicles are thin wafers of bone that overlap oneanother in such a way that they form a ring within the sclera on the corneal side of the eye.The number of scleral ossicles and their pattern of overlap is fairly constant withinsquamate species, and a standard terminology has been developed to describe and numberindividual ossicles for purposes of comparison (Gugg, 1939; Underwood, 1970). MostIguanidae characteristically possess 14 scleral ossicles per eye, with the following patternsof overlap: ossicles 1, 6, and 8 overlap both immediately adjacent ossicles; ossicles 4, 7,and 10 are overlapped by both immediately adjacent ossicles; and the remainder areoverlapped by one neighboring ossicle while overlapping the other (Underwood, 1970; deQueiroz, 1982). In a previous study (de Queiroz, 1982), I reported this pattern for alliguanine genera. I have now examined the following additional species and report the sameossicle configuration: Brachylophus vitiensis (one eye from one specimen examined);Ctenosaura bakeri (Roatan Island; 2, 1); C. clarki (4, 4); C. defensor (1, 1); C. palearis (1,1); C. quinquecarinata (1, 1); C. similis (8, 5); Cyclura carinata (4, 2); and C. rileyi (2, 1).Additional material of Amblyrhynchus (2, 1) also exhibits this pattern, supporting myprevious suggestion that two specimens with fewer than 14 ossicles are anomalous.Hyoid Apparatus (Fig. 33). The hyoid apparatus lies within the tissue between themandibles, where it serves as the skeletal framework for the tongue and throat muscles.This delicate structure is often lost or partially destroyed in dry skeletal preparations. Iniguanines, the hyoid apparatus consists of a median, anteriorly directed hypohyal (lingualprocess); the body of the hyoid, which is also a median element and is continuous with thehypohyal; and portions of three pairs of visceral arches. The hyoid arch is the most lateral
Phylogenetic Systematics oflguanine Lizards 67
FIG. 32. Corneal view of the left scleral ring of Ctenosaura similis (MCZ 9566). All iguanine speciestypically exhibit the pattern of scleral ossicles illustrated: a total of 14 ossicles, with numbers 1, 6, and 8positive (horizontal lines) and numbers 4, 7, and 10 negative (crosshatched). Scale equals 0.5 cm.
and consists of basihyals, projecting anterolaterally from the body of the hyoid, andceratohyals, which run posteriorly from the distal ends of the basihyals. Basihyals fuse tothe hyoid body late in postembryonic ontogeny. Separate epihyals are not evident. Medialto the ceratohyals lie the first ceratobranchials; these are the only bony elements of thehyoid apparatus, the remainder being composed of calcified cartilage. The firstepibranchials extend posteriorly and dorsally from the posterior ends of the firstceratobranchials. The second ceratobranchials lie medial to the first ceratobranchials andextend direcdy posteriorly. Camp (1923) reported the presence of second epibranchials inIguana. Although I have never observed discrete second epibranchials in iguanines, thedelicate nature of these elements may have resulted in their destruction during skeletalpreparation.Differences exist among iguanine taxa in the relative lengths and the orientations of thevarious hyoid elements (Fig. 33). The most obvious differences are seen in the secondceratobranchials. In Ctenosaura, Cyclura, Dipsosaurus, and Iguana delicatissima, thesecond ceratobranchials are of moderate size; they are generally more than two-thirds thelength of the first ceratobranchials, and never do they more than barely exceed the latter inlength (Fig. 33A). Although there is some overlap in the ranges of the relative lengths ofthe second ceratobranchials between Amblyrhynchus, Conolophus, and Sauromalus, onthe one hand, and members of the previously described group, the second ceratobranchials
m
?a o .3? ? >!a 2 pa u fc? "O <O O _?>^ ?^ S
-3 -? t/3O O
> c2 '^
"=
-S> c1)Od I
?a c
C3sa*
^&?:??3 ..
o c34) 4)
Phylogenetic Systematics oflguanine Lizards 69
of Amblyrhynchus, Conolophus, and Sauromalus are relatively short, often less than two-thirds the length of the first ceratobranchials (Fig. 33B). Iguana iguana and both species ofBrachylophus have long second ceratobranchials, invariably much longer than the firstceratobranchials (Fig. 33C). The long second ceratobranchials support the gular fans seenin these species.Another variable character in the hyoid skeletons of iguanines is the proximity of thetwo second ceratobranchials to one another. In all iguanines except Amblyrhynchus andSauromalus, these elements contact each other along the midline for most or all of theirlengths (Fig. 33A,C); sometimes they are separated by a small gap where they meet thebody of the hyoid. In Amblyrhynchus and Sauromalus the second ceratobranchials arelargely or entirely separated from one another (Fig. 33B).Most of the outgroup taxa examined in this study have second ceratobranchials ofintermediate size, these elements being slightly shorter than the first ceratobranchials.Some Basiliscus have slightly longer second ceratobranchials, but they are not nearly aslong as those of Brachylophus and Iguana iguana. Crotaphytus and Gambelia have shortsecond ceratobranchials, about half the length of their first ceratobranchials. Thus, verylong second ceratobranchials are almost certainly apomorphic for iguanines, and, unlesscrotaphytines are the sister group of iguanines, short ones are probably also apomorphic.Separation of the second ceratobranchials along the midline is unequivocally apomorphic,based on the outgroups used in this study.AXIAL SKELETON
Presacral Vertebrae (Figs. 34, 35, 36, 37). The presacral vertebrae (Fig. 34) of alliguanines are procoelous and possess supplementary articular surfaces, zygosphenes andzygantra, medial to the zygapophyses. Iguanine cervical vertebrae, defined as thosevertebrae anterior to the first one bearing a rib that attaches to the sternum (Hoffstetter andGasc, 1969) and including the atlas and axis, invariably number eight. From four to sevenventrally keeled intercentra are present on the atias, the axis, and between the centra of theanterior cervical vertebrae, decreasing in size posteriorly. The intercentrum of the axisfuses with its centrum late in postembryonic ontogeny. There is regional differentiation inthe shape of the presacral vertebrae: the anterior and posterior presacrals are relatively shortcompared to those in the middle of the column.The number of presacral vertebrae in iguanines ranges from 23 to 27 (Table 4). Mostspecies exhibit a strong statistical mode of 24 presacral vertebrae, with occasional variantshaving 23 or 25. I judge this to be the plesiomorphic condition because it is seen in allspecies of basiliscines, crotaphytines, morunasaurs, and oplurines that I have examined.Within the genus Ctenosaura, three species, C. clarki, C. defensor, and C.quinquecarinata, have a modal number of 25 presacral vertebrae. Because the apomorphiccondition occurs in only some Ctenosaura, this character reveals nothing aboutrelationships among my basic taxa. I used differences in modal numbers of presacralvertebrae as a character only in an analysis of relationships within Ctenosaura.
70 University of California Publications in Zoology
con
FIG. 34. Twentieth presacral vertebra of Brachylophus vitiensis (MCZ 160255) in (A) lateral (anteriorto left), (B) dorsal, and (C) ventral views. Scale equals 2 mm. Abbreviations: con, condyle; cot, cotyle;ns, neural spine; po, postzygapophysis; pr, prezygapophysis; s, synapophysis for articulation of rib; zy,zygosphene.
Phylogenetic Systematics oflguanine Lizards 71TABLE 4. Numbers of Presacral Vertebrae
72 University of California Publications in Zoology
FIG. 35. Lateral views of the twentieth presacral vertebrae of (A) Sauromalus obesus (RE 1578) and (B)Ctenosaura pectinata (RE 641), showing differences in the height of the neural spine. Scale equals 0.5 cm.Abbreviations: con, condyle; ns, neural spine; pz, postzygapophysis; s, synapophysis.
Sauromalus differs from other iguanines in the morphology of its presacral vertebrae.In this genus, the neural spines of the presacral vertebrae are short (Fig. 35A); from thebase of the postzygapophysial articular surfaces to the top of the neural spine they measureless than 50% of the total height of the vertebrae. In most other iguanines the neural spinesmake up more than 50% of the total vertebral height (Figs. 34A, 35B), though there isconsiderable variation in this category. This variation includes both interspecificdifferences in adult morphology and ontogenetic increase in neural spine height withinspecies. Ctenosaura bridges the morphological gap between the two categories, with somemembers (e.g., C. clarki) approaching the condition seen in Sauromalus. Outgroupcomparison yields equivocal results concerning the polarity of the different conditions of
Phylogenetic Systematics oflguanine Lizards 73
FIG. 36. Dorsolateral views of the tweniieth presacral vertebrae of (A) Dipsosaurus dorsalis (KdQ 22)and (B) Sauromalus obesus (RE 1578), showing absence and presence, respectively, of bony separation(arrows) between the prezygapophyses and the zygosphenes. Scale equals 1 mm.
74 University ofCalifornia Publications in Zoology
neural spine height. Crotaphytines, Hoplocercus, Chalarodon, and some Opiums haveshort neural spines; those of Laemanctus, Morunasaurus, and other Oplurus are roughlyintermediate; and those of Basiliscus, Corytophanes, and Enyalioides are tall, reachingextreme heights in adult male Basiliscus. Because of this ambiguous evidence, I did notuse neural spine height as a character at the first level of phylogenetic analysis withiniguanines, though it was used later at a lower hierarchical level.The zygosphenes oi Dipsosaurus differ from those of other iguanines (Fig. 36). In thistaxon, the articular surfaces of the zygosphenes are connected laterally to those of theprezygapophyses by a continuous arc of bone (Fig. 36A). All other iguanines have a deepanterior notch separating the articular surfaces of the zygosphenes from those of theprezygapophyses (Fig. 36B).In their weakest form, zygosphenes are mere out-tumings of the medial surfaces of theprezygapophysial facets that face dorsolaterally (Hoffstetter and Gasc, 1969). When morestrongly developed, the articular surfaces of the zygosphenes are oriented laterally orventrolaterally, eventually coming to face directly opposite those of the prezygapophyses.The final stage in the expression of the zygosphenal half of the accessory vertebralarticulation appears to be the separation of the zygosphenes from the prezygapophyses by anotch. Thus, Dipsosaurus is the only iguanine that does not exhibit full development of thezygospheneal articulations. Although the degree to which the zygosphene-zygantrumarticulation is developed may be positively correlated with size in iguanids (Etheridge,1964a), this fact alone cannot account for its relatively weak development in Dipsosaurus,the smallest iguanine. Outside of Iguaninae, Corytophanes, which is about the same size(snout-vent length) as Dipsosaurus, possesses the deep notch separating zygosphenes fromprezygapophyses, while Petrosaurus that are larger than Dipsosaurus do not.Outgroup comparison provides equivocal evidence concerning the plesiomorphiczygosphenal morphology for iguanines. Among the outgroups examined in this study, thevertebrae of basiliscines and some Enyalioides resemble those of most iguanines in havingstrongly developed zygosphenes and zygantra with deep anterior notches between thearticular surfaces of the zygosphenes and those of the prezygapophyses. Crotaphytinesand most morunasaurs have weakly developed accessory vertebral articulations: thearticular surfaces of the zygosphenes are continuous with the medial portions of those ofthe prezygapophyses, and, unlike those of all iguanines, they face dorsolaterally rather thanventrolaterally. The zygosphene-zygantrum articulations are very weakly developed inOplurus and Chalarodon. Therefore, some nonhomology between morphologically similarvertebrae is required under the assumption of iguanine monophyly. Either the notch in thebasiliscine accessory articulation (and that of some Enyalioides) is convergent with the onein iguanines, or its absence in Dipsosaurus is convergent (and possibly also a reversal)with a similar condition seen in other outgroups.Sacrum (Fig. 39). Like all tetrapodous squamates, iguanines characteristically havetwo sacral vertebrae, although some specimens have asymmetrical sacra of the formreported by Hoffstetter and Gasc (1969) involving three vertebrae (Table 4). I recognize
Phylogenetic Systematics of Iguanine Lizards 75
two characters in the sacra of iguanines, both involving the pleurapophyses of the posteriorsacral vertebra.The posterior edges of the pleurapophyses of the posterior sacral vertebrae of iguaninesmay or may not bear posterolaterally directed processes (Hofstetter and Gasc, 1969: Fig.50). These processes are usually present, though not invariably so, in Amblyrhynchus,Brachylophus, Conolophus, Dipsosaurus, and Sauromalus, and are present in the singlespecimen of Cyclura pinguis examined; they are absent in Ctenosaura, Iguana, and otherCyclura. When present, each process lies posteroventral to a foramen in the posteriorsurface of the second pleurapophysis. Occasionally, a process may develop dorsolateral tothe foramen; this process and the one described previously do not seem to be homologouson positional grounds.Given the outgroups used in this study and their uncertain relationships, outgroupanalysis is useless for assessing the plesiomorphic condition of this character. Theprocesses are absent in basiliscines and Hoplocercus, present in the Enyalioides, variablypresent in Oplurus, Chalarodon, Gambelia, and Morunasaurus, and present inCrotaphytus. Therefore, I did not employ this character in phylogenetic analysis at thelevel of all iguanines.The canal leading to the foramen that emerges alongside the posterior edge of eachposterior sacral pleurapophysis has its medial opening on the ventral surface of the samepleurapophysis. This ventral foramen is almost always present in all iguanines exceptConolophus. In Conolophus, the ventral foramen may also be present, but more often it isabsent, and an open groove is left in place of the enclosed canal. The condition seen inConolophus is almost certainly apomorphic, since all four outgroups generally possess theforamen and enclosed canal.Caudal Vertebrae (Figs. 37, 38). Iguanine caudal vertebrae are highly variable, butpossess many common structural features. The neural spines of the anterior caudalvertebrae are taller than their presacral counterparts, but they gradually decrease in sizeposteriad and increase their posterior orientation until they vanish toward the end of the tail.Complete haemal arches, positioned intercentrally, begin between the centra of the secondand third or the third and fourth caudal vertebrae. They are oriented posteroventrally and,like the neural spines, decrease in size and increase in posterior orientation, movingposteriorly, until they vanish near the end of the tail. The bases of the haemal arches mayform continuous basal bars or they may be separate. Small, paired elements, presumablyserially homologous with the bases of the haemal arches, or otherwise incomplete haemalarches, often precede the first complete arch.Four vertebral series (Fig. 37) can be recognized in the caudal sequence of iguanines(Etheridge, 1967). The anterior seven to fifteen caudal vertebrae bear a single pair oflaterally or posterolaterally oriented transverse processes (fused caudal ribs) and lackautotomy septa (fracture planes) (Fig. 37A). In the following series, each vertebra bearstwo pairs of transverse processes that are either parallel or diverge from one another (Fig.37B). The vertebrae in this second series and the remaining two series may or may nothave autotomy septa. Species that lack autotomy septa generally have a shorter double-
76 University of California Publications in Zoology
B D
FIG. 37. Dorsal views of caudal vertebrae ofDipsosaurus dorsalis (KdQ 22): (A) number 4, (B) number9, (C) number 15, and (D) number 28. Scale equals 1 mm. Abbreviations: fp, fracture plane; ns, neuralspine; prz, prezygapophysis; tp, transverse process.
process series and more frequently possess bilaterally asymmetrical transverse processes.The transverse processes decrease in size posteriorly and, although the members of theposterior pair are as large or larger than those of the anterior pair, it is usually the formerthat disappear first (although the alternative is not uncommon), resulting in a third serieswith a single pair of transverse processes (Fig. 37C). These processes, presumablyserially homologous with the anterior transverse processes of the second series, based ontheir anterior position on the vertebrae, continue to decrease in size until they vanish,leaving a fourth series whose vertebrae lack transverse processes (Fig. 37D). A variablenumber of vertebrae at the end of this last series are nonautotomic.
Phylogenetic Systematics oflguanine Lizards 77
The number of caudal vertebrae in iguanines varies from as few as 25 in Ctenosauradefensor to over 70 in Iguana iguana. Because this number varies considerably withinspecies, much of the variation is difficult to partition into character states nonarbitrarily.Nevertheless, an apparent gap exists between Sauromalus and some Ctenosaura, whichhave fewer than 40 caudal vertebrae, and all other iguanines, which have more than thisnumber.Outgroup comparison does not clearly indicate the plesiomorphic number of caudalvertebrae in iguanines. Most outgroup species have numbers of caudal vertebrae near orbridging the gap seen in iguanines. Hoplocercus is unique among outgroup taxa in havinga very short (fewer than 20 vertebrae), spiny tail, even more extreme than those of certainCtenosaura, and lacking any complete haemal arches. Because of this ambiguity, I usedthe number of caudal vertebrae as a systematic character only at a level less inclusive thanall iguanines.Unlike other iguanines, Amblyrhynchus, Brachylophus, Conolophus, and Iguanadelicatissima lack autotomy septa along their entire caudal sequences throughoutpostembryonic ontogeny, and thus presumably are unable to autotomize their tails. Thisdoes not mean, however, that these lizards cannot regenerate their tails, for caudalregeneration occurs in both Brachylophusfasciatus (Etheridge, 1967) and B. vitiensis. Inthese cases, regeneration was associated with a broken vertebra rather than intervertebralseparation, supporting Etheridge's (1967) suggestion that regeneration is a function oftrauma to the vertebra rather than autotomy itself (but see Bellairs and Bryant, 1985). It isnoteworthy that all iguanines that lack caudal fracture planes are insular forms. Caudalautotomy is generally thought to be an adaptation for escaping predators (Congdon et al.,1974; Turner et al., 1982), and the intensity of predation is often less severe on islands(Carlquist, 1974).I am unable to resolve the polarity of this character with the four outgroups used in thisstudy. The basiliscines Laemanctus and Corytophanes, the crotaphytine Crotaphytus, andthe morunasaur Hoplocercus lack autotomy septa, but in other members of all of thesegroups and in all oplurines examined, the septa are present. Thus, monophyly of each ofthe outgroups and of iguanines requires multiple homoplastic events no matter whichcondition, presence or absence of autotomy septa, is considered to be plesiomorphic foriguanines. Because of the ambiguity involved in this character, I withheld an initialdecision on its polarity and used it only at a hierarchical level below that of all iguanines.The beginning of the second series of caudal vertebrae varies both within and amongiguanine species. High overlap among species in the range of this character within speciesrenders much of this variation useless as systematic characters, but one character can berecognized for the purpose of comparisons among the basic taxa used in this study. InBrachylophus and Dipsosaurus, the series of caudal vertebrae with two pairs of transverseprocesses per vertebra begins at the eighth to the tenth caudal vertebra; in all otheriguanines, this series begins at the tenth or a more posterior vertebra. Because ofintraspecific variation in the beginning of this second series of caudal vertebrae, a given
78 University of California Publications in Zoology
specimen may not be assignable to one or the other group, but a species (sample) can be soassigned.Unfortunately, the pathway of character-state transformation cannot be analyzed byoutgroup comparison without making additional assumptions about the character. None ofthe four outgroups used in this study, nor any other iguanian, possesses caudal vertebraewith two pairs of transverse processes (Etheridge, 1967). Nevertheless, a closecorrespondence between the beginning of the series of caudal vertebrae with two pairs oftransverse processes and the beginning of the series of autotomic vertebrae in iguaninessuggests that the latter might be used as the character instead. Unfortunately, not alliguanines (nor all outgroup taxa) possess autotomic caudal vertebrae. Therefore, in orderto use this character I first must assume that the beginning of the series of caudal vertebraewith two pairs of transverse processes in taxa that lack autotomy septa corresponds withthe beginning of the autotomic series in those taxa that possess autotomy septa. Second, Imust assume that the beginning of the autotomic series in taxa that lack vertebrae with twopairs of transverse processes corresponds with the beginning of the series of vertebrae withtwo pairs of transverse processes.Under these assumptions, outgroup comparison can be used with those outgroupspossessing autotomic vertebrae, but it provides ambiguous evidence concerning theplesiomorphic condition of this character. The autotomic series of Basiliscus begins in arange that has the tenth caudal vertebra in its midst. That of Gambelia begins posterior tothe tenth vertebra, while those of Enyalioides, Morunasaurus, and oplurines begin anteriorto the tenth vertebra. The polarity decision for this character will thus vary depending uponthe relationships among iguanines and the four outgroups. Because these relationships areunknown, I withheld a decision on the polarity of this character in phylogenentic analysis atthe level of all iguanines.Lazell (1973:1-2) citing Etheridge (in litt.) distinguished Iguana from Cyclura by thepresence of "a low fmlike process above the neural arch of no more than six anterior caudalvertebrae" in the former, compared to the "high, fmlike processes above the neural archesof all the caudal vertebrae" in the latter. The processes in question are presumablyossifications of the dorsal skeletogenous septum. When the remaining iguanine genera areconsidered, there appears to be a continuum in the height of these processes rather than twodiscrete morphologies, low and high. Even within an organism, the morphology of theseprocesses differs among the caudal segments. In most iguanines, the processes on theanterior caudal vertebrae are merely thin, midsagittal extensions of the anterior edges of theneural spines. Moving posteriorly along the column, apices form on the processes, and theprocesses themselves are displaced anteriorly, sometimes becoming entirely separated fromtheir respective neural spines. The height of the processes increases, then graduallydecreases, moving anterior to posterior. Although the midsagittal processes generallydisappear short of the end of the tail, they are present (Fig. 3 8A) well beyond the anteriorthird of the caudal sequence (determined by vertebra number, not by distance from thebeginning of the tail) in all genera except Brachylophus and Iguana. The situation inBrachylophus and Iguana differs from the one described above in that the processes are
Phylogenetic Systematics oflguanine Lizards 79
con+ .. / . f/^ipv^.^^ con-
FIG. 38. Lateral views of the ninth caudal vertebrae of (A) Dipsosaurus dorsalis (KdQ 22) and (B)Iguana iguana (MVZ 78384), showing differences in the size of the dorsal midsagittal processes. Scaleequals 2 mm; anterior is to iJie right. Abbreviations: con, articular condyle; ns, neural spine; p, dorsalmidsagittal process.
relatively small and do not continue as far posteriorly in the caudal sequence (Fig. 38B).Although they may be present beyond the sixth caudal vertebra, I have never observedthem beyond the tenth. The caudal sequences oi Brachylophus and Iguana consist of morethan 55 vertebrae; thus, the processes are not present beyond the anterior fifth of thesequence.Although the evidence is somewhat equivocal, outgroup comparison favors theinterpretation that the condition of the midsagittal processes of the caudal vertebrae seen inBrachylophus and Iguana is apomorphic. The alternative condition occurs incrotaphytines, morunasaurs, and oplurines, but basiliscines are similar to Brachylophusand Iguana. In basiliscines, the small, fmlike processes are rarely found posterior to thefifth caudal vertebra. Basiliscines, Brachylophus, and Iguana are all arboreal, suggesting apossible functional relationship between the morphology of the caudal vertebrae and use ofthe tail in arboreality.Ribs (Fig. 39). Variation in the numbers and the morphology of various kinds of ribshas served as the basis for characters in previous systematic studies of iguanids (Etheridge,1959, 1964a, 1965b, 1966); but iguanines are conservative in most of these features. Likethose of all iguanids, iguanine ribs are holocephalous and most have two parts: a bonydorsal portion and a cartilaginous ventral portion, the inscriptional rib (Etheridge, 1965b).The length of the inscriptional ribs is highly variable from one region of the vertebralcolumn to another, and at the posterior end of the presacral series these elements are oftenlacking.Cervical ribs, those ribs anterior to the first ribs that are attached to the sternum,typically number four pairs in iguanines, beginning on the fifth presacral vertebra (very
80 University of California Publications in Zoology
atlasintercentra
sternum
sternal ribs
postxiphisternal ribs
cervical ribs
xiphisternal ribs
sacrum
FIG. 39. Presacral and sacral vertebrae and ribs oiDipsosaurus dorsalts in ventral view. The drawing isa composite.
rarely on the fourth) and ending on the eighth. The bony portions of the first two cervicalrib pairs are short, while the second two are much longer, about the same length as theanterior thoracic ribs. The next four (rarely three) rib pairs, on presacral vertebrae ninethrough twelve, are sternal ribs, attached ventromedially to the lateral borders of thesternum through their cartilaginous ventral portions. Two (rarely three; sometimes one inSauromalus) pairs of xiphisternal ribs follow the sternal ribs. These ribs articulate dorsally
Phylogenetic Systematics ofIguanine Lizards 8 1
with vertebrae 13 and 14, and their cartilaginous ventral portions unite with one anotherbefore attaching to the posterior end of the sternum. The remaining ribs are simply termedpostxiphisternal. The bony anterior postxiphistemal ribs are often as long as theirxiphisternal counterparts, but there is a progressive reduction in their length posteriorly.The posteriormost ribs are shorter than the sacral pleurapophyses. Lumbar vertebrae,posterior presacral vertebrae lacking ribs, are not found in iguanines. Very rarely, the ribsof the posteriormost presacral segment are fused to the vertebra.Etheridge (1965b) described variation in the abdominal skeleton (postxiphistemalinscriptional ribs) of iguanids. All iguanines were reported to exhibit a pattern in which allpostxiphistemal inscriptional ribs are attached to their corresponding dorsal bony ribs. Insome iguanines, all of these inscriptional ribs end free, while in others the members of oneor more of the anterior pairs may join midventrally to form continuous chevrons. Based onEtheridge's (1965b) findings and my own observations, the iguanine genera exhibit thefollowing morphologies in the abdominal skeleton: (1) continuous chevrons absent(Dipsosaurus, Sauromalus); (2) continuous chevrons present or absent (Amblyrhynchus,Conolophus, Ctenosaura, Cyclura, Iguana); and (3) continuous chevrons present(Brachylophus). The number of continuous chevrons and other enlarged postxiphistemalinscriptional ribs may exhibit taxon-specific pattems, but because the fragile abdominalskeleton is often destroyed in skeletal preparations, I have not been able to examine enoughspecimens to assess these pattems adequately.In the outgroups that I have examined, postxiphistemal inscriptional ribs that formcontinuous midventral chevrons are found only in momnasaurs; however, because theyshare the common feature of having at least some inscriptional ribs that bear no traces ofattachment to the bony ribs, Etheridge (pers. comm.) believes that the oplurine pattem is atransformation of that seen in momnasaurs. Basiliscines and crotaphytines are similar toDipsosaurus and Sauromalus in their lack of continuous chevrons. Thus, evidence bearingon the polarity of this character is equivocal, and I did not use it in my initial analysis ofrelationships among iguanine genera.PECTORAL GIRDLE AND STERNAL ELEMENTSThe iguanine pectoral girdle and stemal elements (Fig. 40) are closely associated and forma complex functional unit composed of six pairs of elements plus two median, unpairedones. Some of these elements are composed entirely of calcified cartilage, while others arebony. All iguanines possess all 14 elements: suprascapulae, scapulae, coracoids,epicoracoids, clavicles, interclavicle, sternum, ana xiphistema.Suprascapulae (Fig. 40). These are paired fan-shaped elements composed of calcifiedcartilage that extend continuously from the dorsal edges of the scapulae. The suprascapulaelie just extemal to the posterior cervical and the anterior thoracic bony ribs. They are notattached directly to the axial skeleton, but ride over the bony portions of the ribs. As inmost squamates, the only direct skeletal attachments between pectoral girdle and axialskeleton are through the stemum and cartilaginous portions of the anterior thoracic ribs. In
82 University ofCalifornia Publications in Zoology
FIG. 40. Pectoral girdles of (A) Brachylophus fasciatus (RE 1866), (B) Ctenosaura hemilopha (RE1341), and (C) Sauromalus obesus (RE 411). A is a lateral view; anterior is to the right. B and C areventral views. Calcified cartilage is stippled. Scale equals 1 cm. Abbreviations: acf, anterior coracoidfenestra; cf, coracoid foramen; cl, clavicle; cor, coracoid; epc, epicoracoid; gf, glenoid fossa; icl,interclavicle; pcf, posterior coracoid fenestra; sc, scapula; scf, scapulocoracoid fenestra; sf, scapular fenestra;sr, sternal ribs; ssc, suprascapula; st, sternum; stf, sternal fontanelle; xi, xiphistemum.
Phylogenetic Systematics oflguanine Lizards 83
most iguanines, the surfaces of the scapulae and suprascapulae form a continuous, laterallyconvex arc, but in Sauromalus the junction of these surfaces is angular and thesuprascapulae are oriented more horizontally than in other iguanines. The condition of thesuprascapulae in Sauromalus is presumably related to the depressed body form of theseanimals, and on the basis of outgroup comparison is almost certainly apomorphic.Scapulae, Coracoids, and Epicoracoids (Fig. 40). The scapula and coracoid of eachside are closely associated and function as a single unit. Although separated by a suturethroughout most of the period of growth, the two bones fuse to form a singlescapulacoracoid element near the attainment of maximum size. Prominent features of thescapulocoracoids are the glenoid fossae for the articulation of the humeri, which lie at thejunctions between scapulae and coracoids along their posterior edges, coracoid foraminaanteroventral to the glenoid fossae, and three or four (rarely two) scapulocoracoidfenestrations on each side of the girdle, the functional significance of which is discussed byPeterson (1973).The scapulocoracoid fenestrations pierce the pectoral girdle along the anterior marginsof the scapulae and coracoids, between these bones and the cartilagenous epicoracoids(Fig. 40). Following the terminology of Lecuru (1968a), from dorsal to ventral the fourpairs of fenestrations are: (1) scapular fenestrae, which lie anterodorsally within thescapulae; (2) scapulocoracoid fenestrae, situated at the junctions between scapulae andcoracoids; (3) anterior (primary) coracoid fenestrae, located within the coracoids; and (4)posterior (secondary) coracoid fenestrae, also located within the coracoids butposteroventral to the anterior coracoid fenestrae. All iguanines invariably possess thescapulocoracoid and the anterior coracoid fenestrae; the scapular fenestrae and the posteriorcoracoid fenestrae may be present or absent.Scapular fenestrae are invariably present in all iguanines except Amblyrhynchus andSauromalus, in which they are small or occasionally absent. Outgroup analysis yieldsequivocal results concerning the polarity of these character states. Scapular fenestrae arepresent in crotaphytines, the single Enyalioides oshaughnessyi examined, Chalarodon, andOplurus cuvieri; they are absent in basiliscines, other morunasaurs, and Oplurusquadrimaculatus (in which the large "scapulocoracoid" fenestrae may be homologous withthe scapular plus the scapulocoracoid fenestrae of other oplurines). Because of thisambigiuty, I used the presence or absence of scapular fenestrae as a systematic characteronly at a level less inclusive than all iguanines.The presence of posterior coracoid fenestrae is more variable intragenerically than thepresence of scapular fenestrae. Posterior coracoid fenestrae are invariably absent inBrachylophus (Fig. 40A); usually absent in Dipsosaurus; usually present inAmblyrhynchus, Ctenosaura (Fig. 40B), Cyclura, and Sauromalus (Fig. 40C); andinvariably present in Conolophus and Iguana. The amount of variability differs among thegenera in the third group. Posterior coracoid fenestrae are frequently absent inAmblyrhynchus and Sauromalus, in which all species are variable in the presence of thesefenestrae except S. australis and S. slevini, both of which are represented by small samples(n=2). The absence of a posterior coracoid fenestra is rare in Ctenosaura; it has been
84 University ofCalifornia Publications in Zoology
detected in only some members of three species, C. clarki, C. hemilopha, and C. similis.In Cyclura, the absence of a posterior coracoid fenestra was observed only in two out ofeight C. nubila, one of which lacked the fenestra unilaterally.According to Peterson (1973), the presence of a posterior coracoid fenestra isassociated with large size and/or the presence of a proximal belly of the M. biceps.Because a posterior coracoid fenestra is present in the species of Ctenosaura that reachsmaller maximum sizes than Braehylophus, in which the fenestra is absent, presence of thefenestra cannot be strictly size-dependent. The association of the fenestra with a proximalbelly of the M. biceps was not examined in the present study.Although the evidence is somewhat ambiguous, outgroup comparison favors theinterpretation that the absence of posterior coracoid fenestrae is plesiomorphic foriguanines. Basiliscines and oplurines invariably lack these fenestrae. Morunasaursgenerally lack posterior coracoid fenestrae, but in rare cases very small ones are present.Crotaphytines generally possess posterior coracoid fenestrae, although they areoccasionally absent in Gambelia. If the general rather than the invariable presence orabsence of posterior coracoid fenestrae is considered to be the systematic character, thenoutgroup comparison will either yield equivocal results or indicate that the absence ofposterior coracoid fenestrae is plesiomorphic, depending on the relationships amongiguanines and the four outgroups.Clavicles (Fig. 40). Iguanine clavicles are boomerang-shaped, paired bones lyingalong the anterior margin of the pectoral girdle. They articulate ventromedially with theanterior median end of the interclavicle and dorsolaterally with the anteroventral edges ofthe suprascapular Compared with those of certain other iguanids, the clavicles ofiguanines are relatively simple, generally lacking sharp, ventrally directed processes (hooksof Etheridge, 1964a) and ventromedial fenestrae, although small fenestrae are sometimespresent in Conolophus.Sauromalus differs from other iguanines in having slender clavicles, which are more orless elliptical in cross section. The clavicles of other iguanines have thin lateral shelves,making them wider when viewed anteriorly, although some Ctenosaura approach thecondition seen in Sauromalus. Because the clavicles of all outgroup taxa examined exceptOplurus quadrimaculatus are wide with thin lateral shelves, this condition must beconsidered plesiomorphic for iguanines.Interclavicle (Fig. 40). This median, unpaired bone is the ventralmost in the pectoralgirdle. In iguanines it bears the shape of a "T" or an arrow, formed by a lateral process atthe anterior end on each side and a median posterior process. The anterior process seen incertain other squamates (Lecuru, 1968b) is virtually absent.The extent of the posterior median process of the interclavicle varies among iguaninesand is here assessed by the location of the posterior tip of the bone relative to the lateralcomers of the sternum and the sternal attachments of the cartilaginous sternal ribs.Amblyrhynchus and Sauromalus (Fig. 40C) have short interclavicles that do not extendposteriorly beyond the lateral corners of the sternum, where the first pair of sternal ribsattaches. In all other iguanines except Conolophus pallidus and Cyclura nubila the
Phylogenetic Systematics oflguanine Lizards 85
posterior process of the interclavicle extends beyond this level (Fig. 40B) and, dependingon the taxon, it may extend beyond the points of attachment of the second or even the thirdsternal-rib pairs. Conolophus pallidus and Cyclura nubila have interclavicles ofintermediate length. In these taxa the interclavicle extends to about the level of the lateralcomers of the sternum or slightly beyond. The width of the posterior process appears to berelated to its posterior extent: short interclavicles are usually wider than long ones. Thecorrelation is not strict, however, for some Sauromalus have narrow posterior processes.Among the outgroups examined, only some Crotaphytm have an interclavicle that doesnot extend posteriorly beyond the lateral comers of the sternum. I therefore considered theshort interclavicle to be apomorphic for iguanines.Another variable feature of iguanine interclavicles is the angle between each lateralprocess and the posterior process. All species exhibit at least 10? of variation in this featurewith significant intertaxic overlap. For this reason I recognize only two categories ascharacter states. Amblyrhynchus and Sauromalus (Fig. 40C) have roughly T-shapedinterclavicles, with the angle between the lateral and posterior processes ranging from 75?to 90?. Other iguanines have arrow-shaped interclavicles (Fig. 40B); the angle formed bythe lateral and posterior processes is usually less than 75?. Although the angle in questionoverlaps the first category in some members of both species of Brachylophus andConolophus, as well as in some Cyclura nubila, the lower limits of the range of angles inthese species is well below that in Amblyrhynchus and Sauromalus. Outgroup comparisonindicates that the arrow-shaped interclavicle is plesiomorphic. Among basiliscines,crotaphytines, morunasaurs, and oplurines, I have found T-shaped interclavicles only inthe basiliscines Laemanctus serratus and Corytophanes hernandesii.Sternum and Xiphisterna (Figs. 37, 40). The sternum of iguanines is shaped like adiamond or a pentagon and is composed of calcified cartilage. In embryos and somehatchlings, the sternal plate is paired, but the two halves fuse in late embryonic or earlypostembryonic ontogeny to form a single median element. Anterolaterally, the sternummeets the epicoracoids in a tongue-in-groove articulation, the coracostemal joint, whichpermits posterolateral-anteromedial movements of the scapulocoracoid units relative to thesternum (Jenkins and Goslow, 1983). The posterolateral borders of the sternal plate arethe attachment sites for the cartilaginous ventral portions of four thoracic rib pairs (sternalribs) and two others that attach via the xiphistema. A sternal fontanelle may be present(Fig. 40B) or absent (Fig. 40C).In most iguanines, the sternal fontanelle is long and narrow and is covered partially orcompletely by the posterior process of the interclavicle. In Amblyrhynchus andSauromalus the sternal fontanelle is often small, and in the latter it may be subdivided intotwo or three small, round holes. In some specimens of both taxa the fontanelle is absent.Absence or small size of the sternal fontanelle is unequivocally apomorphic on the basis ofthe outgroups used in this study.Sternal shape is variable in iguanines and is partly related to another feature, theproximity of the two sternal-xiphistemal attachments to one another and the midline. Inmost iguanines the xiphistema attach to the sternum very close to the midline and to one
86 University ofCalifornia Publications in Zoology
B
-^ /?p
aip
FIG. 41. Pelvic girdles of (A) Sauromalus obesus (RE 467) and (B) Ctenosaura pectinata (RE 419) indorsal view. Scale equals 1 cm. Abbreviations: aip, anterior iliac process; ep, epipubis; hi, hypoischiaccartilage; il, ilium; is, ischium; it, ischial tuberosity; pi, proischiac cartilage; pu, pubis.
another, yielding a diamond-shaped sternum (Fig. 40B). In Sauromalus the xiphistema arewidely separated from one another, and the sternum is pentagonal (Fig. 40C).Amblyrhynchus is somewhat intermediate, having a small but distinct gap between itsxiphistema; however, the shape of its sternum is much closer to that of most otheriguanines than to that of Sauromalus.Most members of all outgroup taxa examined have diamond-shaped sterna with thexiphistema in close proximity to each other. The exceptions are Oplurus quadrimaculatusand Crotaphytus, which approach the condition seen in Sauromalus to a greater or lesserdegree, respectively. Although the pentagonal stemum with widely separated xiphistema isprobably apomorphic, the ambiguity is sufficient to force me to use this character only at aless inclusive level than that of all iguanines.PELVIC GIRDLEThe iguanine pelvic girdle (Fig. 41) consists of three pairs of bones: dorsal ilia, whicharticulate with the sacral pleurapophyses; posteroventral ischia; and anteroventral pubes.Cartilaginous epipubes, and proischiac and hypoischiac cartilages, are situated on themidline between the pubes and the anterior and posterior parts of the ischia, respectively.An obvious difference in the shape of the pelvic girdle separates Sauromalus (Fig. 41 A)from all other iguanines (Fig. 4 IB). Relative to those of other iguanines, the pelvis ofSauromalus is short and broad, clearly an apomorphic condition on the basis of theoutgroups examined.
oa ??
88 University ofCalifornia Publications in Zoology
ac
FIG. 43. Right hind limb skeleton of Brachylophusfasciatus: (A) femur; (B) tibia, fibula, and proximaltarsals; and (C) distal tarsals, metatarsals, and phalanges. Scale equals 1 cm. Abbreviations: ac,astragalocalcaneum; f, fibula; t, tibia; I-V, digits 1-5.
Phylogenetic Systematics oflguanine Lizards 89
Another unique feature occurs in some Sauromalus, notably S. varius. In theseanimals the ischium is excavated mesial to the posteriorly directed ischiac tubercle,enhancing the distinctness of this structure. Because this character varies within a singlegenus, it is uninformative about relationships among the basic taxa used in this study.I disagree with Lazell's (1973:1-2) statement that "In Dipsosaurus and Sauromalus theilial shaft tapers abruptly posteriorly and the anterior iliac process is rather weaklydeveloped." The ilial shaft of Sauromalus is narrower at its posterior terminus than thoseof other iguanines, but it does not taper abruptly. In Dipsosaurus the ilial shaft may taperabruptly, but it is broad near its posterior end like that of other iguanines exceptSauromalus. While the anterior iliac process of Sauromalus does appear to be relativelysmall, that of Dipsosaurus is not. LIMBS
Iguanine hmbs exhibit considerable variation, but I have chosen not to use this variation asthe basis for systematic characters. All iguanines possess the same bony elements in theirlimbs, but the proportions of the various limb bones vary considerably among iguaninetaxa. Nevertheless, these proportions seem to be very plastic features, so plastic that I wasunable to establish polarities with any confidence. Therefore, I give only a generaldescription of this variation and devote most of the section to the description of charactersthat do not vary among iguanines but that may be useful at higher levels of comparison.Compared to those of other iguanines, the limb bones of Brachylophus are relativelylong, while those of Amblyrhynchus and Sauromalus are relatively short. Theseproportional differences are most evident in the long bones, metapodials, and phalanges.Proportional differences in the carpal and tarsal elements (mesopodials) are less obvious.All iguanines possess the following bones in the forelimb (Fig. 42): humerus, radius,ulna, radiale, ulnare, pisiform, lateral centrale, five distal carpals, five metacarpals, and 17phalanges. According to Carroll (1977), the first distal carpal of modem lizards ishomologous with the medial centrale of other diapsids. As in other iguanids (Renous-Lecuru, 1973), the intermedium is absent. The phalangeal formula of the manus is2:3:4:5:3. An entepicondylar foramen is present in the humerus.The hind limbs of iguanines (Figs. 43, 44) consist of femur, tibia, fibula,astragalocalcaneum, two distal tarsals proximal to metatarsals three and four, fivemetatarsals, and 18 phalanges. The phalangeal formula of the pes is 2:3:4:5:4 which, likethat of the manus, is presumably plesiomorphic for squamates.OSTEODERMSTwo large Amblyrhynchus (JMS 126, 127) have dermal ossifications that apparentlyformed within the large, conical scales overlying the nasal, prefrontal, and frontal bones(PI. 1), confirming Camp's (1923:307) observation that osteoderms are present in thistaxon. Osteoderms, which differ from the rugosities that develop on various bones of the
90 University of California Publications in Zoology
tlV
FIG. 44. Right tarsal region of Brachylophus fasciatus. Scale equals 0.5 cm. Abbreviations: a,astragalus; c, calcaneum; f, fibula; ml-V, metatarsals 1-5; t, tibia; till and tlV, distal tarsals 3 and 4.
dermal skull roof in certain iguanids, are unknown in iguanids other than Amblyrhynchus(Etheridge and de Queiroz, 1988), and their presence is thus considered derived withiniguanines. Although Conolophus has enlarged, conical head scales overlying the nasal,prefrontal, and frontal bones similar to, yet smaller than, those seen in Amblyrhynchus, Ihave never observed osteoderms in Conolophus. The osteoderms of Amblyrhynchus areeasily removed along with the skin, judging from their absence in most skeletalpreparations of Amblyrhynchus, and it is therefore possible that Conolophus alsopossesses osteoderms. I will assume that osteoderms are absent in Conolophus until theirpresence is demonstrated.
Phylogenetic Systematics oflguanine Lizards 91J
Plate 1. Dorsal (above) and lateral (below) views of the skull oi Amblyrhynchus crisiatus (JMS 127),showing osteoderms.
NONSKELETAL MORPHOLOGY
Iguanines exhibit considerable morphological variation in functional systems other than theskeleton, and I have therefore used certain nonskeletal characters for which relativelycomplete data on variation, both among all iguanine genera and for the four outgroups,were easily obtained. Characters in this section were taken from diagnoses in revisions,reviews, and faunal accounts as well as from the few comparative studies of nonskeletalanatomy of iguanines. I also include some obvious characters that I noticed in the courseof this study. ARTERIAL CIRCULATION
Zug (1971) was pessimistic about the systematic utiUty of the variation that he found in thepatterns of the major arteries of iguanids. Nevertheless, I found at least three characters inhis descriptions, as well as one additional character, that suggest monophyletic groupswithin Iguaninae. Other arterial characters may also be useful for phylogenetic studieswithin this taxon, but have not yet been studied in sufficient detail. Still other charactersare either invariant among iguanines (e.g., branching pattern of the carotid arches,separation of the origins of dorsal aorta and subclavians) or variable within iguanine genera(e.g., separate origin of mesenteries versus origin from a common trunk), and thus cannotbe used for examining relationships among these genera. These characters may be useful atdifferent hierarchical levels.It should be noted that Zug (1971) surveyed nearly all genera of Iguanidae, whichlimited him to relatively small samples for each genus (a maximum of four specimens forany iguanine genus). Zug did not examine Conolophus; my data are based on dissection ofa single C. subcristatus (CAS 12058).Zug reported that the subclavians of Brachylophus and Dipsosaurus are coveredlaterally by a thin, flat ligament, while those of other iguanines pass laterally beneath(=dorsal to?) a muscle bundle. My own observations on Dipsosaurus reveal muscle fibersin the thin sheets of tissue that cover the subclavians just lateral to their origins from theright systemic arch. Furthermore, whether muscular or ligamentous, the structures thatcover the subclavians are the posterior portions of the paired M. rectus capitis anterior ortheir tendons, which originate on the ventral surfaces of the cervical vertebrae and insert onthe exoccipitals and basioccipital lateral to the occipital condyle. Thus, even if the reporteddifference exists, it is a difference in the muscles rather than in the subclavian arteries.
92
Phylogenetic Systematics oflguanine Lizards 93
The subclavians of Conolophus exhibit neither of the patterns described by Zug forother iguanines. In this taxon, the subclavians lie posterior and ventral to the origins of theM. rectus capitis anterior and are thus not covered by this muscle. For these reasons I useonly the difference between the subclavians of Conolophus and those of all other iguaninesas a systematic character.According to Zug (1971), in Dipsosaurus and Brachylophus the dorsal aorta originatesdorsal to the heart (by union of the left and right systemic arches), while in other iguaninesit originates posterior to the heart. My observations on Dipsosaurus (n=l) and Sauromalus(n=l) reveal a profound difference supporting this distinction. In Dipsosaurus the systemicarches unite to form the dorsal aorta about as far posterior as the middle of the heart and theanterior end of the ninth vertebra. In Sauromalus the systemic arches remain paired muchfurther posteriorly; they unite well behind the heart, near the middle of the 13th vertebra.Conolophus, however, is intermediate. The dorsal aorta in this taxon originates at aboutthe level of the posterior end of the heart and the anterior end of the 1 0th vertebra. BecauseZug did not discuss variation within his two categories, I arbitrarily placed Conolophuswith those iguanines in which the dorsal aorta originates posterior to the heart.Finally, I note minor exceptions to some of Zug's observations. In the singleDipsosaurus that I examined, the heart reaches the transverse axillary plane rather thanbeing entirely anterior to this plane. In the single Sauromalus that I examined, the coeliacoriginates between, but separate from, the two mesenteric arteries.COLIC ANATOMY
Iverson (1980) studied colic anatomy in iguanines. Variation within this group exists in thepresence of colic valves, irregular colic folds, circular valves, semilunar valves, and in thenumber of colic valves. Although Iverson considered iguanine colic anatomy to be oflimited phylogenetic value, at least two characters seem to be potentially useful for inferringphylogenetic relationships among iguanines. Nevertheless, because all of the colicmodifications that characterize subsets of iguanines appear to be transformations ofcharacters unique to iguanines, their polarity cannot be established by outgroup comparisonuntil certain phylogenetic relationships within iguanines are determined. For example, onecannot use noniguanine outgroups to infer that colic folds are plesiomorphic relative to colicvalves, or vice versa, because neither condition occurs in these outgroups.The fact that noniguanines possess neither of the conditions found in iguanines is onlya problem if these conditions are homologous members of a transformation series.Otherwise, each condition could be said to be lacking in the outgroups and therefore to be aseparate apomorphic state. If they are homologous, however, one is forced to detenninethe apomorphy of the alternative conditions relative to each other. I assume homologybetween the colic valves and colic folds, because they share the common property of beinginfoldings of the same tissue components of the colic wall (Iverson, 1980). I also assumehomology between circular and semilunar valves. The only difference between these twomorphologies is whether or not the infolded tissue extends around the entire perimeter of
94 University of California Publications in Zoology
the colon (Iverson, 1980). Because of the difficulties involved in outgroup comparisonwith the colic characters, I used them only at hierarchical levels less inclusive thanIguaninae as a whole.Although much variation exists in the modal number of colic valves among iguaninetaxa, this number is positively correlated with (maximum?) body size and does not changesignificantly during the postembryonic ontogeny of a given species (Iverson, 1980). Lackof a thorough study of the relationship between valve number and body size makescomparison of taxa that differ in body size problematic, and I have chosen not to use thenumbers of different types of colic valves as systematic characters.EXTERNAL MORPHOLOGY
Unlike the arterial and colic characters, which were obtained from comparative studies, thefollowing characters were taken primarily from generic diagnoses or are based on personalobservations. No adequate comparative descriptions of these characters exist in theliterature, and I therefore describe them in more detail than the arterial and colic characters.The scutellation of the iguanine head is complex and is potentially the source of manysystematic characters. I note here only some obvious intertaxic differences and charactersthat have been used by previous authors.Scales of the Snout and Dorsal Head. In most iguanines the snout terminates anteriorlyin a median, azygous rostral scale. Sauromalus differs from all other iguanines in that itusually lacks an unpaired, median rostral (H. M. Smith, 1946: Fig. 38); the anteriormostsnout scales above the lip are paired and separated by a median suture that meets the lipmargin. According to Gates (1968), this character occurs in about 78% of S. obesus. Allbasiliscines, crotaphytines, morunasaurs, and oplurines possess a median, azygous rostralscale, indicating that the condition seen in Sauromalus is apomorphic within iguanines.The other scales in the snout region also exhibit differences among iguanines. In mosttaxa they are relatively small, about the same size as the remaining dorsal cephalic scales.In Iguana and some Cyclura, however, these scales form large plates. Interspecificvariation in this character is great within Cyclura (figures in Schwartz and Carey, 1977),ranging from the small scales much like those of other iguanines in C carinata, C. pinguis,and C. ricordii to the large plates of C. cychlura and C. nubila. Cyclura collei and C. rileyiare intermediate, and the horns of C. cornuta are difficult to compare with the conditionsseen in other taxa. Because outgroup comparison suggests that enlarged rostral scales areapomorphic (only Lxiemanctus among the outgroups examined has enlarged snout scales),either (1) the occurrence of this feature in Iguana and some Cyclura is convergent; (2) itindicates that Iguana is the sister group of some part of a paraphyletic Cyclura; or (3)enlarged snout scales is a synapomorphy oilguana plus Cyclura, and some Cyclura haveevolved small snout scales secondarily. Only a consideration of other characters canresolve this question.Amblyrhynchus and Conolophus are similar to one another and differ from all otheriguanines in the scalation of the dorsal surface of the head. In these two genera the dorsal
Phylogenetic Systematics oflguanine Lizards 95
head scales are pointed and conical, giving the head a rugose texture. This condition ismore strongly developed in Amblyrhynchus than in Conolophiis. All other iguanines haveflat or only slighdy domed dorsal head scales. In Sauromalus hispidus these scales aremore strongly pointed than in the other taxa, but the condition is not nearly as extreme as inthe Galapagos iguanas.Like most iguanines, crotaphytines, oplurines, and most basiliscines have relatively flathead scales. Laemanctus serratus is the only basiliscine with conical head scales, but thesescales are confined to the casque on the back of the head and do not extend onto the frontaland nasal regions as in the Galapagos iguanas. The dorsal head scales of morunasaurs arevariable. In Hoplocercus and Morunasaurus these scales are convex but not pointed; inEnyalioides they are pointed and conical, but are relatively much smaller than those of theGalapagos iguanas. Thus, the condition of the dorsal head scales in Amblyrhynchus andConolophus is not seen in any of the outgroups and must be considered apomorphic.Superciliaries. Etheridge and de Queiroz (1988) noted variation in the superciliaryscales of iguanines. In Dipsosaurus these scales are elongate anteroposteriorly and overlapone another extensively, especially in the anterior portion of the row. Amblyrhynchus andSauromalus possess the opposite extreme in which the superciliaries are roughlyquadrangular and nonoverlapping. The remaining iguanines are intermediate, with onlymoderate overlap of the superciliaries. Outgroup comparison indicates that the condition ofthe superciliaries has been relatively plastic at this level of comparison, makingdetermination of its polarity ambiguous. Quadrangular, nonoverlapping superciliariesoccur in morunasaurs and the basiliscine Corytophanes. Elongate, strongly overlappingsuperciHaries occur in oplurines, and an intermediate condition occurs in crotaphytines andthe basiliscines Basiliscus and Laemanctus.Suboculars. The morphology of the subocular scales is also variable in iguanines(Etheridge and de Queiroz, 1988). Dipsosaurus and Ctenosaura have one long and severalshorter suboculars. In all other iguanines except Amblyrhynchus, which is intermediate,all of the suboculars are approximately equal in size. The condition of the suboculars in thefour outgroups is too variable to allow inference about the polarity of this character.Basiliscines, morunasaurs, and some Crotaphytus have suboculars that are subequal insize. Other Crotaphytus have one moderately elongate subocular. Gambelia and oplurineshave one very long subocular and several much shorter ones.Anterior Auricular Scales (Van Denburgh, 1922). Sauromalus differs from all otheriguanines in the scales that border the tympanum anteriorly, the anterior auricular scales.From two to five of these scales are enlarged relative to the neighboring scales and projectposterolaterally over the tympanum, offering protection to this delicate membrane. In allother iguanines except Dipsosaurus, the anterior auricular scales are small and thetympanum is completely exposed. Dipsosaurus possesses a row of slightly enku-gedanterior auricular scales. Outgroup comparison indicates that the enlarged anteriorauriculars of Sauromalus are apomorphic. Basiliscines, Crotaphytus, Hoplocercus,Morunasaurus, and some Enyalioides lack enlarged anterior auricular scales, while inGambelia and oplurines they are only slightly enlarged, roughly comparable to those of
96 University of California Publications in Zoology
Dipsosaurus. Some Enyalioides possess one or two seemingly nonhomologous large,pointed scales dorsal to the tympanum. Some sceloporines have anterior auriculars fully aslarge in proportion to their body size as those of Sauromalus; I consider this to beconvergent.Gular Region. All iguanines possess a transverse gular fold, although it is relativelyweakly developed in Amblyrhynchus compared to other iguanines. A midsagittal gularexpansion, or dewlap, is variably developed, but in no iguanine is it as highly extensible asin Anolis. A large dewlap is present in male Brachylophus fasciatus (Boulenger, 1885;Gibbons, 1981) and in both sexes of B. vitiensis (Gibbons, 1981), Ctenosaura palearis(Bailey, 1928), and Iguana. It is absent in Amblyrhynchus, Conolophus, mostCtenosaura, Dipsosaurus, and Sauromalus, but is weakly developed in Cyclura(Boulenger, 1885) and Ctenosaura bakeri (Bailey, 1928). The presence of a dewlap is nota simple dichotomy, as evidenced by the intermediate condition in Cyclura and Ctenosaurabakeri; nevertheless, a morphological gap exists between those taxa possessing a largedewlap and those in which it is weakly developed or absent.A prominent gular fold occurs in all outgroup taxa used in this study and is, therefore,inferred to be plesiomorphic for iguanines. Although the absence of a dewlap is the mostcommon condition among the outgroups, sufficient variation exists that this conditioncannot be inferred to be plesiomorphic for iguanines as long as higher-level relationshipsremain unresolved. The dewlap is absent in Basiliscus, Laemanctus, crotaphytines,Hoplocercus, Morunasaurus, and oplurines, but it is present in Corytophanes and maleEnyalioides (Boulenger, 1885).Although a dewlap is developed to varying degrees in different iguanines, only the twospecies of Iguana possess a gular crest, a midsagittal row of enlarged scales extendingbelow the throat along the edge of the dewlap. Because a gular crest is lacking in alloutgroup taxa examined except Corytophanes, its presence in Iguana is inferred to beapomorphic.Middorsal Scale Row. A row of scales aligned along the dorsal midline is present in alliguanines except Sauromalus. When present, the scales of the middorsal row aredifferentiated from the neighboring scales, although the degree of differentiation is highlyvariable. This variation ranges from the small, rounded knobs that form the row inDipsosaurus to the tall curved spikes of large Amblyrhynchus and Iguana. In someCyclura (Schwartz and Carey, 1977) and Ctenosaura (Bailey, 1928), the crest formed bythe series of modified middorsal scales is interrupted in the shoulder or the sacral region.The presence of a middorsal scale row in the outgroups is highly variable, making itimpossible to determine polarity at this level of analysis. A middorsal scale row is presentin most basiliscines, Enyalioides, Morunasaurus annularis, and Chalarodon; it is absent incrotaphytines, Laemanctus serratus, Morunasaurus groi, Hoplocercus, and Oplurus.Subdigital Scales of the Pes (Fig. 45). The conspicuous combs on the toes of Cyclurahave long been used to diagnose this genus and especially to separate it from Ctenosaura(Barbour and Noble, 1916; Bailey, 1928; Schwartz and Carey, 1977). Similar toedenticulations, however, are known to occur in other iguanines (Gibbons, 1981). These
Phylogenetic Systematics oflguanine Lizards 97
aks
FIG. 45. Pedal digit II of (A) Sauromalus obesus (MVZ 35978), (B) Brachylophus fasciatus (CAS54664), and (C) Cyclura carinata (CAS 54647) in anterodorsal view, showing differences in the morphologyof the subdigital scales. Scale equals 1 cm. Fused subdigital scales are shaded. Abbreviations: aks,anterior keels of subdigital scales.
98 University of California Publications in Zoology
denticulations are formed by enlarged keels on the anterior edges of the subdigital scales.Varying degrees of enlargement of these keels are seen in iguanines. In Sauromalus theanterior keels of the subdigital scales are nearly the same size as the posterior ones (thesubdigital scales are usually bi- or tricarinate), and the subdigital scales are roughlybilaterally symmetrical with respect to the long axis of the toe (Fig. 45A). In Dipsosaurusand Iguana the anterior keels of the subdigital scales are slightly larger than their posteriorcounterparts, and the subdigital scales are asymmetrical. Further enlargement of theanterior keels and a concomitant increase in the asymmetry of the pedal subdigital scales isseen in Amblyrhynchus, Conolophus, Brachylophus (Fig. 45B), and Cyclura (Fig. 45C)(increasing in size roughly in that order). Much of this variation can be seen withinCtenosaura.All subdigital scales do not exhibit equal enlargement of the keels, which are usuallylargest under the first phalanx of digit II and the first and second phalanges of digit HI.Cyclura and Ctenosaura defensor differ from other iguanines in that the scales bearing theselargest keels are fused at their bases, giving the scales the appearance of a comb whenviewed anteriorly (Fig. 45C). In Cyclura these combs are formed under the first phalanxof digit II and the first and second phalanges of digit III (illustrated in Barbour and Noble,1916: Plates 13-15); in Ctenosaura defensor they occur only under the first phalanx of digitIII. Enlargement of the anterior keels of the subdigital scales is present in all outgroupsexamined in this study except basiliscines, though the degree of enlargement is variable.Basiliscines cannot be compared with iguanines because they have but a single median keelon the subdigital scales. In oplurines and crotaphytines the keels are moderately enlargedas in Dipsosaurus, but in morunasaurs (especially Morunasaurus) they are very large.Thus it is not possible to determine the precise plesiomorphic size of the keels of iguanines.Nevertheless, two conditions seen in iguanines can be considered to be apomorphic.Because the subdigital scales of all outgroups (except basiliscines) bear large anterior keels,the small anterior keels and concomitant symmetry of the subdigital scales in Sauromalusare apomorphic. Fusion of the bases of the subdigital scales with enlarged anterior keels isnot seen in any outgroup and must also be considered apomorphic.Hands and Feet. The hands and feet of Amblyrhynchus are partially webbed(Boulenger, 1885), which is presumably related to the semi-aquatic habits of these lizardsand is unique among iguanids.Caudal Squamation. One of the supposedly diagnostic features of Ctenosaura is a tailarmed with strong, spinous scales (Bailey, 1928); however, similar caudal squamation alsooccurs in most Cyclura (Barbour and Noble, 1916; Schwartz and Carey, 1977). Withinthese two taxa the caudal squamation is highly variable among species. In some Cyclura(e.g., C. cornuta), the caudal scales in adjacent verticils are of similar size and are notspinous, a condition like that seen in most other iguanines. In the remaining Cyclura and inCtenosaura the tail bears whorls of enlarged, spinous scales at regular intervals along itslength. These whorls are separated by verticils of smaller scales that are smooth or muchless spinous (except the middorsal scale row). The number of verticils between the whorls
Phylogenetic Systematics of Iguanine Lizards 99
of enlarged, spinous scales is variable along the tail, generally decreasing posteriorly. Themaximum number of rows between whorls of enlarged scales ranges from none in someCtenosaura defensor (Bailey, 1928; Duellman, 1965) to about six in Cycliira nubila(Schwartz and Carey, 1977). Within Ctenosaura, there appears to be a negative correlationbetween the size of the scales in the enlarged whorls and both the number of scale rowsbetween them and the relative length of the tail.The evolution (or loss) of spinose tails appears to have occurred repeatedly withiniguanids. Like most iguanines, basiliscines, crotaphytines, Chalarodon, and someEnyalioides have more or less uniform caudal squamation without spinous scales. OtherEnyalioides, Morunasaurus, Hoplocercus, and Oplurus have whorls of enlarged spinousscales separated by smaller scales. The short, spinose tail of Hoplocercus is as extreme asanything seen in Ctenosaura. Although it seems likely that tails with whorls of enlarged,spinous scales are apomorphic within iguanines, this polarity is equivocal unlessassumptions are made about either the relationships among outgroups and ingroup or thosewithin morunasaurs and oplurines.Cross-sectional Body Shape. Sauromalus differs from all other iguanines in its crosssectional body shape. All other iguanines are either laterally compressed or cylindrical incross section, while Sauromalus is strongly depressed. The shape of the body ofSauromalus and several other of its distinctive skeletal features (e.g., low neural spines,horizontal orientation of the suprascapular short and broad pelvic girdle) are probablyredundant characters. They are treated separately here because (1) the correlation amongthem is only hypothesized, and (2) some of them are known to change withoutaccompanying changes in the others (e.g., not all depressed lizards have suprascapulae thatform sharp angles with the scapulae).Cross-sectional body shape in members of the four outgroups examined in this studyvaries in such a way that it is impossible to determine the plesiomorphic shape foriguanines. Basiliscines are laterally compressed. Some morunasaurs are compressed{Enyalioides) while others are depressed {Hoplocercus), and both crotaphytines andoplurines are depressed, though generally not as strongly as Sauromalus.
SYSTEMATIC CHARACTERS
Based on the descriptions of the iguanine skeleton and other anatomical features givenabove, I recognize the following systematic characters for use in phylogenetic analysis.SKELETAL CHARACTERSL Ventral surface of premaxilla (Fig. 7): (A) bears large posterolateral processes; (B)posterolateral processes absent.2. Posteroventral crests of premaxilla (Fig. 7): (A) small, do not continue up the sidesof incisive process and are not pierced by foramina for maxillary arteries; (B) large,continue up sides of incisive process and are pierced or notched by foramina for maxillaryarteries.3. Anterior surface of rostral body of premaxilla: (A) broadly convex; (B) nearly flat.4. Nasal process of premaxilla I (Figs. 6, 14, 45): (A) slopes backwards; (B) nearlyvertical.5. Nasal process of premaxilla II (Fig. 8): (A) wholly or partly exposed dorsallybetween nasals; (B) covered dorsally between nasals.6. Size of nasals and nasal capsule (Figs. 5, 9, 11): (A) nasal capsule of moderatesize, nasals relatively small; (B) nasal capsule enlarged, nasals relatively large.7. Bones in anterior orbital region (Fig. 10): (A) lacrimal contacts palatine behindlacrimal foramen; (B) prefrontal contacts jugal behind lacrimal foramen.8. Frontal (Figs. 5, 9, 11): (A) longer than wide, or length approximately equal towidth; (B) wider than long.9. Large paired openings at or near frontonasal suture: (A) absent; (B) present.10. Cristae cranii on ventral surface of frontal (Fig. 12): (A) extend in a smoothcontinuous curve from frontal onto prefrontals; (B) frontal portions project anteriorly,forming a step between frontal and prefrontal portions.11. Paired cristae on ventral surface of frontal medial to cristae cranii (Fig. 12): (A)absent or weakly developed; (B) strongly developed, united as a single median crestanteriorly and together with the cristae cranii forming pockets in the anteroventral surface ofthe frontal.12. Dorsal borders of orbits (Figs. 5, 9, 11): (A) more or less smoothly curved; (B)wedge-shaped.13. Position of parietal foramen (Figs. 5, 9, 11; Table 2): (A) on the frontoparietalsuture; (B) variable (either A or C); or (C) within the frontal bone.
100
Phylogenetic Systematics ofIguanine Lizards 101
14. Supratemporals: (A) extend anteriorly more than halfway across the posteriortemporal fossae; (B) extend anteriorly no more than halfway across the posterior temporalfossae.15. Maxilla I: (A) relatively flat or concave laterally; (B) flares outward ventral to therow of supralabial foramina.16. Maxilla II (Figs. 5, 14): (A) premaxillary process of maxilla lies roughly in thesame plane as the remainder of the maxilla; (B) premaxillary process of maxilla curvesdorsally.17. Lacrimal: (A) large; (B) intermediate; (C) small.18. Ventral process of squamosal (Fig. 15): (A) large; (B) small or absent.19. Squamosal (Fig. 15): (A) separated from or barely contacting dorsal end oftympanic crest of quadrate; (B) abuts against dorsal end of tympanic crest of quadrate.20. Septomaxilla: (A) flat, or with a weak ridge on anterolateral surface; (B) with apronounced longitudinal crest.21. Anterior dorsal surface of palatines (Fig. 16): (A) with a low medial ridge; (B)with a high medial crest.22. Infraorbital foramen I (Fig. 17), process of palatine projecting posterolaterally orlaterally behind the infraorbital foramen: (A) large; (B) small or absent.23. Infraorbital foramen II (Fig. 17), process of palatine projecting posterolaterally orlaterally behind the infraorbital foramen: (A) fails to contact jugal; (B) contacts jugal.24. Infraorbital foramen III (Fig. 17): (A) located on the lateral or posterolateral edgeof the palatine; (B) located entirely within the palatine (may or may not be connected by asuture to the lateral edge of the palatine).25. Pterygoids (Figs. 5, 18): (A) medial borders relatively straight anterior to thepterygoid notch, pyriform recess narrows gradually; (B) medial borders curve sharplytoward the midline anterior to the pterygoid notch, pyriform recess narrows abruptly.26. Ectopterygolds: (A) fail to contact palatines near posteromedial corners ofsuborbital fenestrae; (B) usually contact palatines near posteromedial corners of suborbitalfenestrae.27. Parasphenoid rostrum (Fig. 20): (A) long; (B) short.28. Cristae ventrolaterals of parabasisphenoid (Fig. 21): (A) strongly constrictedbehind basipterygoid processes; (B) intermediate; (C) widely separated.29. Posterolateral processes of parabasisphenoid (Fig. 21): (A) present and large; (B)small or absent.30. Laterally du-ected points on cristae interfenestrahs: (A) absent; (B) present.31. Stapes: (A) thin; (B) thick.32. Relative heights of dorsal borders of dentary and surangular on either side ofcoronoid eminence (Fig. 22): (A) approximately equal; (B) dorsal border of dentary wellabove that of surangular.33. Splenial: (A) large; (B) small.
102 University of California Publications in Zoology
34-35. Anterior inferior alveolar foramen (Fig. 23): (A) always between splenial anddentary, the coronoid may or may not contribute to its posterior margin; (B) entirely withinthe dentary in some specimens (others A); (C) between splenial and coronoid.36. Labial process of coronoid (Fig. 24): (A) small; (B) intermediate; (C) large.37. Angular I (Fig. 25): (A) extends far up the labial surface of the mandible and islargely visible in lateral view; (B) does not extend far up the labial surface of the mandibleand is barely visible in lateral view.38. Angular II: (A) wide posteriorly; (B) narrow posteriorly.39. Surangular (Fig. 26): (A) exposed laterally only about as far forward as the apexof the coronoid or the anterior slope of this bone, and never anterior to the last dentarytooth; (B) exposed laterally well anterior to the apex of the coronoid and often anterior tothe last dentary tooth.40. Lingual exposure of surangular between ventral processes of coronoid (Fig. 27):(A) a dome-shaped portion exposed; (B) largely or completely covered by prearticular.41. Angular process of prearticular (Fig. 28): (A) increases substantially in relativesize during postembryonic ontogeny, becoming a prominent structure in adults; (B)increases only slightly in relative size during postembryonic ontogeny, remaining relativelysmall even in adults.42. Retroarticular process (Figs. 28, 29): (A) tympanic and medial crests convergeposteriorly to give the process a triangular outline in both juveniles and adults; (B)tympanic and medial crests converge posteriorly in juveniles, but the posterior endsseparate during ontogeny so that the process assumes a quadrangular outline in adults.43-44. Modal number of premaxillary teeth (Table 3): (A) fewer than seven; (B)seven; (C) more than seven.45. Crowns of premaxillary teeth: (A) lateral cusps small or absent; (B) lateral cuspslarge.46. Crowns of posterior marginal teeth I (Fig. 30): (A) tricuspid; (B) four-cusped; (C)polycuspate (5 to 10 cusps); (D) serrate.47. Crowns of tricuspid posterior marginal teeth II (Fig. 30): (A) individual lateralcusps much smaller than apical cusp; (B) individual lateral cusps relatively large, subequalto apical cusp in size.48. Pterygoid teeth I (Fig. 31): (A) entire row lies along the ventromedial edge of thepterygoid adjacent to the pyriform recess; B) posterior portion of row displaced laterally.49. Pterygoid teeth II (Fig. 31): (A) entire row single throughout ontogeny; (B)posterior portion of row doubles ontogenetically; (C) entire row doubles ontogenetically.50. Pterygoid teeth III (Fig. 31): (A) anterior portion of tooth patch present; (B)absent (posterior end of suborbital fenestra used as reference point).51. Pterygoid teeth IV (Fig. 31): (A) usually present; (B) usually absent.52-53. Hyoid I (Fig. 33): (A) second ceratobranchials short, often less than two-thirds the length of the first ceratobranchials; (B) intermediate, from two-thirds the lengthof the first ceratobranchials to slightly longer than the first ceratobranchials; (C) long, muchlonger than the first ceratobranchials.
Phylogenetic Systematics ofIguanine Lizards 103
54. Hyoid n (Fig. 33): (A) second ceratobranchials in medial contact with one anotherfor most or all of their lengths; (B) separated from one another medially for most or all oftheir lengths.55. Neural spines of presacral vertebrae (Figs. 34, 35): (A) tall, making up more than50% of the total vertebral height; (B) short, making up less than 50% of the total vertebralheight.56. Zygosphenes (Fig. 36): (A) connected to prezygapophyses by a continuous arc ofbone; (B) separated from zygapophyses by a deep notch.57. Sacrum I: (A) posterolateral processes of second pleurapophyses (usually)present; (B) (usually) absent.58. Sacrum II: (A) foramina in the ventral surfaces of the second pleurapophyses(usually) present; (B) (usually) absent.59. Number of caudal vertebrae: (A) more than 40; (B) fewer than 40.60. Autotomy septa in caudal vertebrae: (A) present (Fig. 37); (B) absent.61. Beginning of the autotomic series of caudal vertebrae or beginning of the series ofcaudal vertebrae with two pairs of transverse processes (Fig. 37): (A) at or before the 10thcaudal vertebra; (B) at or behind the 10th caudal vertebra.62. Thin, midsagittal processes on the dorsal surface of the caudal centra anterior to theneural spines (Fig. 38): (A) relatively large and present well beyond the anterior third ofthe caudal sequence; (B) relatively small and confined to the anterior fifth of the caudalsequence.63. Postxiphistemal inscriptional ribs: (A) do not form continuous chevrons (Fig. 39);(B) variably form continuous chevrons; (C) invariably form continuous chevrons.64. Suprascapulae: (A) situated primarily in a vertical plane and forming a continuousarc with the scapulocoracoids; (B) situated primarily in a horizontal plane and forming anangle with the scapulocoracoids.65. Scapular fenestrae (Fig. 40): (A) large, invariably present; (B) small or absent.66. Posterior coracoid fenestrae (Fig. 40): (A) usually absent; (B) usually present.67. Clavicles: (A) wide, with a prominent lateral shelf; (B) narrow, the lateral shelfsmall or absent.68. Posterior process of the interclavicle (Fig. 40): (A) extends posteriorly beyond thelateral corners of the sternum; (B) does not extend beyond the lateral corners of thestemum.69. Lateral processes of the interclavicle (Fig. 40): (A) usually forming angles of lessthan 75? with the posterior process and giving the interclavicle the shape of an arrow; (B)forming an angle of between 75? and 90? with the posterior process and giving theinterclavicle the shape of a T.70. Sternal fontanelle (Fig. 40): (A) present and of moderate size; (B) small or absent.71. Stemum-xiphistemum (Fig. 40): (A) sternum diamond-shaped (quadrilateral), thexiphisterna in close proximity; (B) intermediate; (C) sternum pentagonal, the xiphisternawidely separated.72. Pelvic girdle (Fig. 41): (A) long and narrow; (B) short and broad.
104 University of California Publications in Zoology
12). Anterior iliac process: (A) large; (B) small.74. Osteoderms (PI. 1): (A) absent; (B) present.NONSKELETAL CHARACTERS
75. Heart (Zug, 1971): (A) does not extend posterior to the transverse axillary plane;(B) extends posterior to the transverse axillary plane.76. Subclavian arteries (Zug, 1971; present study): (A) covered ventrally by theposterior end of the M. rectus capitis anterior; (B) not covered by the M. rectus capitisanterior.11. Dorsal aorta (Zug, 1971): (A) right and left systemic arches unite to form thedorsal aorta above the heart; (B) origin of dorsal aorta posterior to heart.78. Coeliac artery (Zug, 1971): (A) arises from the dorsal aorta anterior to andseparate from the two mesenteric arteries; (B) arises posterior to the mesenteries, betweenthe mesenteries, or continuous with one or the other of the mesenteries.79. Colic wall (Iverson, 1980): (A) forms one or more transverse valves; (B) formsnumerous irregular transverse folds.80. Colic valves (Iverson, 1980): (A) all valves semilunar; (B) one or more valvescircular (semilunar valves may be present or absent).81. Rostral scale: (A) median and azygous; (B) subdivided by a median suture.82. Scutellation of snout region: (A) consists of many small scales subequal in size tothose of superorbital and temporal regions; (B) consists of relatively few large scales.83. Dorsal head scales: (A) flat or slightly convex; (B) pointed and conical.84. Superciliary scales (Etheridge and de Queiroz, 1988): (A) quadrangular and non-overlapping; (B) intermediate; (C) elongate and strongly overlapping.85. Subocular scales (Etheridge and de Queiroz, 1988): (A) all subequal in size; (B)one or two suboculars moderately elongate; (C) one subocular very long, the rest shorter.86. Anterior auricular scales: (A) all relatively small or one row slighriy enlarged; (B)one row of scales anterior to tympanum pointed and gready enlarged, extending posteriorlyover tympanum.87. Gular fold: (A) conspicuous; (B) weakly developed.88. Dewlap: (A) small or absent; (B) present and large.89. Gular crest: (A) absent; (B) present.90. Middorsal scale row: (A) present; (B) absent.91. Pedal subdigital scales I (Fig. 45): (A) anterior keels larger than posterior ones,scales asymmetrical; (B) anterior and posterior keels approximately equal in size, scalesroughly symmetrical with respect to the long axis of the toe.92. Pedal subdigital scales II (Fig. 45): (A) individual scales entirely separate; (B)scales with greatly enlarged anterior keels fused anteriorly at bases.93. Toes: (A) unwebbed; (B) partially webbed.
Phylogenetic Systematics ofIguanine Lizards 105
94. Caudal squamation: (A) caudal scales in adjacent verticils approximately equal insize, smooth or keeled but not spinous; (B) tail bears whorls of enlarged, strongly spinousscales.95. Cross-sectional body shape: (A) laterally compressed or cylindrical; (B) stronglydepressed.
CHARACTER POLARITIES AND THEPHYLOGENETIC INFORMATION CONTENTOF CHARACTERS
Character- State distributions for the 95 characters among the four outgroups and thepolarities inferred from these distributions are summarized in Table 5. Distributions of thecharacters among the basic taxa (genera) of iguanines are given in Table 6. Notsurprisingly, the number of characters that exhibit variation within a basic taxon iscorrelated with the number of recognized species in the taxon.Each character can be placed in one of four categories depending on its phylogeneticinformation content:
I. Unambiguous synapomorphies of basic taxa (characters 1, 2, 3, 4, 6, 9, 11, 12, 14,15, 16, 17-2, 20, 22, 26, 27, 29, 30, 31, 32, 33, 34, 35, 36-2, 38, 41, 42, 46-3, 47,49-2, 58, 64, 67, 72, 74, 75, 76, 81, 86, 87, 89, 91, 93). The derived condition ofeach of these characters is found in only one of the basic taxa and is characteristic of thetaxon in which it is found. These characters support the monophyly of particulariguanine genera but provide no information about relationships among them.
II. Ambiguous synapomorphies of basic taxa (characters 10, 13-2, 24, 28-2, 53, 78,82, 92). The derived condition of each of these characters is characteristic of one of thebasic taxa but is also variably present in one or more other basic taxa. These charactersare either (1) synapomorphies of one basic taxon that have arisen convergently in partof another one; (2) synapomorphies of one entire basic taxon plus part of another onethat are indicative of the paraphyletic status of the latter; or (3) synapomorphies of aclade consisting of two or more basic taxa that have subsequently reversed within someof them. These characters may or may not provide information about relationshipsamong basic taxa.
III. Derived characters shared by two or more basic taxa (characters 5, 7, 8, 13, 17,18, 19, 21, 23, 25, 28, 36, 37, 39, 40, 45, 46, 46-2, 48, 50, 51, 52, 54, 62, 66, 68,69, 70, 77, 83). The derived condition of each of these characters is characteristic ofmore than one of the basic taxa and may or may not occur variably in one or more ofthe others. These characters are the primary data relevant to an analysis of relationships
106
Phylogenetic Systematics oflguanine Lizards 107
among the basic taxa. Because of character incongruence, the interpretation of thesecharacters as synapomorphies is not always straightforward, and a reasonableinterpretation of any one character must take the others into consideration. Some of thesimilarities are undoubtedly homoplastic and must ultimately be interpreted as morethan one synapomorphy.
IV. Characters of undeterminable polarity (characters 43, 44, 55, 56, 57, 59, 60, 61,63, 65, 71, 73, 79, 80, 84, 85, 88, 90, 94, 95). These characters are too variableeither within or among the outgroups, or both, for any reasonable inference to be madeabout their polarity. Therefore, these characters cannot be used as evidence forphylogenetic relationships within Iguaninae until either the relationships of theoutgroups to iguanines are determined (Maddison et al., 1984) or some phylogeneticstructure within iguanines is established so that some iguanines can serve as outgroupsto others in an analysis of a less inclusive group (Watrous and Wheeler, 1981).
108 University of California Publications in Zoology
TABLE 5. Distributions of Character States of 95 Characters Among Four Outgroups toIguanines and the Polarities That Can Be Inferred From Them
Character
Phylogenetic Systematics of/guanine Lizards 109TABLE 5 (continued)
Character
no University of California Publications in Zoology
TABLE 5 (continued)
Character
Phylogenetic Systematics oflguanine Lizards 111
TABLE 5 (continued)
Character
o -^OOOOOOO
OS OOOO-^O-hO
oo O -H(SO?iOOOOO
^ OOO-hOOOO
lO ^ooooooooo-<ooooo
et o o
c (N ^OOOOOOO
wco6
(UC3UinONoOJc34~tCOV-c(U?4?(oJ3U?4-loco?d3
(U?*??oax:U Os
OO
^
>n
en(N
--oooooooOO'?i ^oooooo o o o o o oO ?I O - O ' oo oO >?I - - o o oo o
-hOOOOOOO
-hO-hOOOOO
-hOOOOOOOo o o o o o oOO-hOOOOO'-hOOOOOOOQwPQ o a-s;
3Cco
o
'si-ONCO
oo
c<-i
c<^
en
ocd OcdU OS
OO
tS
o
o - o o oo O O ?. T?I o >?I ooOOOOOOO-H
?'O-hOOOO-^^^(NOOOOO
?lOOOOOOOO?'OOOOOOooooooo^r-iOOOOOOO
-hOOOoOOOooooo-^ooOOO?iOOOOo o o o CN O CS o
?lOOOOOOOOO'-^OOOOO
-H O
I?I o - o o oo o oo ^OOOOO-hOO
-hO?lOOOOO
-s;s;
03 &u 5 Oc Co
c?acowm
o
0\
oc3 OU ON
oo
r^^
CO
oH
P3OO-hOOOOO
eQpQpqa3pQ<pQpQ<<<<<<<PQ^OOOOOO'-io^oooo-o
-hO-hOOOO-hOO^OO'-'OOr-HO^OO^OO
^ ^ z^ <=^ Z o o o
- o ^. -
o
(N O^ooooooo(N CN ^ C<^ <No o o o opQCQpQpqUPQPQ<cqmpQ03umpQ<ooooo^ooo o o o o o o
? ? Ito.9h ^ c^
ooo Z<CQCQmcQCQmON oq<<<<<<<
oo O O O O O O ?I
r-r^ ^O^-H^O'-H'-'oo-^ooooo
in OOOOOOO?I
'^r- -^ooooooo<<<<<<<CQOOOOOOO?I
cd O53 "^ PQ<<<<<<UO O O O O O -HU ON O O O O O O -H
oo -hOOOOOO?<
r-^ OOOOOOO-H
^^ ?' O -H ^CQ<<<<<<OQOOOOOOO-H
so pqUPQmcQ<m<O-hOOOO?'O
c?dcoowm
^
co
CQ<OQmpQ<CQpq
s:
03
tOco aU U U Q 03
(Ucacow
ON
OS
On
ON
ANALYSIS OF PHYLOGENETIC RELATIONSHIPS
Following character analysis, the phylogenetic relationships among the eight iguaninegenera (basic taxa) and the diagnoses of various monophyletic groups of iguanines,including the basic taxa, were determined by means of a three-step procedure: (1) First, thederived characters shared by the members of two or more basic taxa were used for apreliminary analysis of phylogenetic relationships. (2) Second, certain phylogeneticrelationships based on the preliminary analysis were used to identify new outgroups for thedetermination of polarities of characters that were undeterminable using basiliscines,crotaphytines, morunasaurs, and oplurines as outgroups. These characters were thenadded to the existing set for a second analysis of relationships within a subgroup ofiguanines. (3) Third, the results of the two analyses were combined to produce a finalestimate of phylogenetic relationships within Iguaninae, and the level at which eachcharacter exists as a synapomorphy was reanalyzed in light both of these relationships andof variation within the basic taxa.PRELIMINARY ANALYSISA total of 29 characters (numbers 5, 7, 8, 13, 17, 18, 19, 21, 23, 25, 28, 36, 37, 39, 40,45, 46, 48, 50, 51, 52, 54, 62, 66, 68, 69, 70, 77, 83), representing a minimum of 30phylogenetic transformations, were used in the preliminary analysis of relationships.These are the characters whose derived states are shared by two or more basic taxa(category III). Character 46 has multiple states, with two levels (states 1, 2, and 3; states 2and 3) that are shared by two or more taxa. This accounts for the difference between thenumber of characters and the minimum number of phylogenetic transformations. Table 7gives the 29 characters rescored to eliminate variation within basic taxa, as described underMaterials and Methods, above.The results of the preliminary analysis are summarized in Figure 46. Two differentcladograms (Fig. 46A,B) can account for the distribution of derived characters with aminimum number of phylogenetic character transformations. In terms of the phylogeneticrelationships suggested, the two cladograms differ only in the positions oi Brachylophiisand Dipsosaiirus. In one (Fig. 46A), Dipsosaurus is the sister group of all other iguanines;in the other (Fig. 46B), Brachylophiis occupies this position instead.The minimum-step cladograms require 46 phylogenetic character transformations(consistency index = 0.65), 16 more than the absolute minimum of 30, which would onlyobtain if all of the characters had mutually compatible distributions among the basic taxa.
117
118 University of California Publications in ZoologyTABLE 7. Distributions of Character States of 29 Characters Used in the PreliminaryAnalysis
Taxon Character5 7 8 13 17 18 19 21 23 25 28 36 37 39 40
AmblyrhynchusBrachylophusConolophusCtenosawaCycluraDipsosaurusIguanaSauromalusCI (Fig. 46A)CI (Fig. 46B)
1
Phylogenetic Systematics ofIguanine Lizards 119
^v-.<r .c^
.N<i .C^ X^s^.?^ ?"<>^v .OVV C^^ .<> ^^ c\c>
FIG. 46. Minimum-step cladograms for eight basic taxa of iguanines, resulting from a preliminaryanalysis of 29 characters (Table 7). Two different cladograms (A and B) account for the taxic distribution ofderived characters with 46 character transformations. Synapomorphies of the numbered nodes and basic taxaare given in the text.
120 University of California Publications in Zoology
The consistency indices (Kluge and Farris, 1969) for eacli of the characters on each of thetwo minimum-step cladograms are given in Table 7. The C-index is a measure of thedeviation of a character from a perfect fit (C-index of 1.00) to a given cladogram.Synapomorphies for the various nodes of the cladograms are given below by the number ofthe character and the letter of the character state as designated in the list of systematiccharacters. Convergent characters are underlined; characters involving reversal are markedwith an asterisk. Because only characters whose derived states are shared by two or moreof the basic taxa were used in this analysis, any character interpreted as a synapomorphy ofa basic taxon necessarily exhibits homoplasy.Figure 46A: Node 1: 18-B*, 25-B*; Node 2: 23-B*; Node 3: 48-B, 66-B, 77-B;Node 4: 46-B *. 46-C or-D*; Node 5: 37-B, 52-53-A; Node 6: 5-B, 7-B, 8-B, 17-B or-C,21-B, 36-B or-C . 40-B, 45-B, 46-B or-A*, 50-B . 83-B; Node 7: 19-B, 28-B or-C, 39-B;Amblyrhynchus: 18-A *. 46-A*, 54-B . 68-B . 69-B . 70-B : Brachylophus: 25-A*, 36-B .62-B : Conolophus: 51-B : Ctenosaura: none; Cyclura: none; Dipsosaurus: 13-C . 46-B .50-B . 51-B (last two characters are redundant); Iguana: 18-A *. 62-B ; Sauromalus: 13-B .23-A*, 54-B . 68-B . 69-B . 70-B .The synapomorphies of the second cladogram (Fig. 46B) are identical to those of thefirst (Fig. 46A), with the following exceptions: Node 1: 18-B*, 23-B*; Node 2: 25-B,46-B,-C, or-D*; Node 4: 46-C or-D*; Brachylophus: 36-B . 62-B ; Ctenosaura: 46-A*.Six of the homoplastic characters on the first minimum-step cladogram (Fig. 46A) canbe interpreted in more than one way, each involving the same number of phylogenetictransformations. These alternative interpretations are diagrammed in Figure 47. Character25-B can be interpreted as convergent synapomorphies of Dipsosaurus on the one hand andof all other iguanines except Brachylophus (node 3) on the other hand (Fig. 47A).Alternatively, it can be interpreted as a synapomorphy of all iguanines that has reversed inBrachylophus (Fig. 47B). Characters 54-B, 68-B, 69-B, and 70-B can be interpreted asconvergent synapomorphies of Amblyrhynchus on the one hand and of Sauromalus on theother (Fig. 47C). Alternatively, these characters can be interpreted as synapomorphies ofthe Galapagos iguanas plus Sauromalus (node 5) that have reversed in Conolophus (Fig.47D). Two alternative interpretations of character 46 are diagrammed in Figure 47E and F.Both interpretations require five phylogenetic transformations.Alternative interpretations of homoplastic characters on the second minimum-stepcladogram (Fig. 46B) are identical to those on the first (Fig. 46A), with the followingexceptions: Character 25-B has only one possible minimum-step interpretation; it is asynapomorphy of all iguanines except Brachylophus (node 2). Character 23-B can eitherbe interpreted as convergent synapomorphies of Brachylophus on the one hand and the taxaunited above node 3 (Fig. 48A) on the other hand, or it can be interpreted as asynapomorphy of all iguanines that has subsequently reversed in Dipsosaurus (Fig. 48B).The same alternative interpretations of character 46 are available for the second minimum-step cladogram as for the first, but two additional alternatives exist (Fig. 48C,D).Of the six subterminal nodes on each of the two minimum-step cladograms resultingfrom the preliminary analysis, three (nodes 3, 6, and 7) are well supported. That is, these
Phylogenetic Systematics oflguanine Lizards 121
Di Br Ct Am Co Sa Ig Cy Di Br Ct Am Co Sa Ig Cy
Am Co Sa Am Co Sa
Am Co
FIG. 47. Alternative interpretations of character transformation for homoplastic characters on aminimum-step cladogram (Fig. 47A). A and B are alternative interpretations for character 25; C and D forcharacters 54, 68, 69, and 70; E and F for character 46. Solid squares represent transformations to thederived condition; open squares represent reversals; half-solid squares represent intermediate slates.
nodes are diagnosed by more than two derived characters that are unique and unreversedand strongly outweigh conflicting characters. Node 1 is also well supported, but it issupported by the results of an analysis at a more inclusive hierarchical level. Node 2 is themost weakly supported, for it supports the monophyly of different groups of basic taxa onthe two minimum-step cladograms.
122 University of California Publications in Zoology
Am Co Am Co
FIG. 48. Alternative interpretations of character transformation for homoplastic characters on aminimum-step cladogram (Fig. 47B). A and B are alternative interpretations for character 23; C and D forcharacter 46. Solid squares represent transformations to the derived condition; open squares representreversals; half-solid squares represent intermediate states.
LOWER-LEVEL ANALYSIS
In an attempt to gain better resolution of iguanine phylogenetic relationships, I performedan analysis at a lower hierarchical level (node 3), using Brachylophus and Dipsosaurus asoutgroups in order to determine the polarities of characters that were undeterminable at thelevel of all iguanines. I chose node 3 for this analysis because it is the most inclusivegroup within iguanines whose monophyly is well supported.The precise relationships oi Brachylophus and Dipsosaurus to the rest of the iguaninesare problematical. One of the minimum-step cladograms resulting from the preliminaryanalysis has Dipsosaurus as the sister group of all other iguanines (Fig. 46A), while theother has Brachylophus in this position instead (Fig. 46B). The second hypothesis mightat first appear to be better supported, because Dipsosaurus shares two derived characterswith the other iguanines (characters 25-B and 46-B,-C, or-D), while Brachylophus sharesonly one derived character (23-B) with them. However, character 46 has four equallysimple alternative interpretations, and in only two of these (Fig. 48C,D) does it support asister-group relationship between Dipsosaurus and all iguanines other than Brachylophus.
Phylogenetic Systematics of Iguanine Lizards 1 23
Under the other two alternative interpretations, the presence of the first derived state inDipsosaurus is considered to be convergent, as in Figure 47E and F. For this reason, Ihave chosen to leave the relationships among Brachylophiis, Dipsosaurus, and the newingroup (node 3) unresolved in the assessment of polarities for the lower-level analysis,I used the same basic methodology for determining polarities in the lower-level analysis(Appendix III) that I used in the preliminary analysis, where the relationships of theoutgroups to the ingroup are uncertain (Appendix II). For reasons presented in AppendixIII, I considered polarity to be determinable only when both Brachylophus andDipsosaurus exhibit the same character state.Using Brachylophus and Dipsosaurus as additional outgroups for analysis at a lowerhierarchical level, I was able to determine polarities for 13 of the 20 characters whosepolarities could not initially be determined (Table 8). The number of premaxillary teethturns out to be two characters (hence the numbering in the character list as characters 43-44) representing transformations in opposite directions from the ancestral condition, amode of seven premaxillary teeth. Although character 84 (superciliary scales) differs inBrachylophus and Dipsosaurus, it seems reasonable to conclude that state A is derived,since it is found in neither Brachylophus nor Dipsosaurus and represents one end of acontinuum that has the conditions seen in these two taxa at the other end. BothBrachylophus and Dipsosaurus exhibit the same state for character 61, but this character isirrelevant to an analysis of relationships at the level in question because it does not varywithin the new ingroup. Characters 56 and 80 also do not vary within the ingroup, buttheir polarities are undeterminable because Brachylophus and Dipsosaurus exhibit differentconditions.The use oi Brachylophus and Dipsosaurus as additional outgroups for an analysis ofrelationships at a lower hierarchical level necessitates a reevaluation of the polarities ofthose characters whose polarities had already been determined using more remoteoutgroups. The reasoning behind polarity reevaluation is similar to that behind polarityassessment and is presented in Appendix IV. Under this reasoning, the only characterswhose polarity assessments needed to be changed after reevaluation were character 18(polarity reversed) and character 46 (changed to undeterminable). Character 46 is a four-state character, and what becomes undeterminable is whether state A or state B is ancestral.Therefore, I have lumped states A and B as state and consider states C and D to besuccessively more derived conditions (i.e., C = 1, D = 2).Eight of the characters used in the preliminary analysis of relationships among alliguanines cannot be used in the analysis of relationships of all iguanines other thanBrachylophus and Dipsosaurus, either because they must be interpreted as synapomorphiesof a basic taxon that are convergent with a condition found in Brachylophus or Dipsosaurus(characters 13, 51, and 62) or because they do not vary within the new ingroup (characters25, 48, 66, and 77). These characters were removed from consideration, and theremaining characters were combined with those whose polarities were newly determined,using Brachylophus and Dipsosaurus as outgroups, and whose derived states characterized
124 University of California Publications in ZoologyTABLE 8. Polarity Inferences for Lower-level Analysis Using Brachylophus andDipsosaurus as Outgroups
Character
Phylogenetic Systematics oflguanine Lizards 125TABLE 9. Distributions of Character States of 26 Characters Among Six Taxa Within aSubset of Iguaninae
Taxon Character5 7 8 17 18 19 21 23 28 36 37 39 40
Amblyrhynchus
126 University of California Publications in Zoology
Three fully resolved cladograms of equal and minimum length can be constructed fromthe 26 characters used in the lower-level analysis (Fig. 49). These cladograms differ onlyin the position of Ctenosaura, which in turn depends on the interpretation of character 57,the presence or absence of posterolaterally directed processes on the pleurapophyses of thesecond sacral vertebra. The derived absence of these processes occurs in Ctenosaura,Iguana, and some Cyclura, but was scored absent for the latter taxon in order to simplifyanalysis. This is one of only two derived characters out of the set of 26 that occursinvariably in Ctenosaura and is relevant to the placement of this taxon within the restrictedingroup. The only other derived character that occurs invariably in Ctenosaura (character23-B) also occurs in all ingroup taxa except some Sauromalus. Therefore, provided thatSauromalus is monophyletic, this character is most reasonably interpreted as asynapomorphy of the entire ingroup that has reversed in some Sauromalus. If the sister-group relationship between Iguana and Cyclura, based on other characters, is accepted,then character 57-B might be interpreted as convergent in Iguana on the one hand and inCtenosaura on the other. If so, Ctenosaura can have any of the relationships illustrated inFigure 49; given this information alone, there is no reason to prefer any one of thesealternative placements over the others. Alternatively, character 57-B might be interpreted asa synapomorphy of a clade consisting of Ctenosaura, Iguana, and Cyclura that hassubsequendy reversed within Cyclura. Because Cyclura is actually variable for thischaracter, the hypothesis of acquisition and reversal requires fewer phylogenetictransformadons than does that of convergence (two instances versus three). Although oneof the three cladograms (Fig. 49A) would be favored under such an interpretation, thedifference is so small that little importance can be attached to it in terms of resolving theplacement of Ctenosaura. Therefore, I consider the relationships of Ctenosaura within therestricted ingroup to be uncertain.Because the three minimum-step cladograms resulting from the lower-level analysisdiffer only in the placement of Ctenosaura, I present diagnostic synapomorphies for asingle consensus cladogram (Adams, 1972) that leaves the relationships of Ctenosauraunresolved (Fig. 50). This consensus cladogram is identical to the other three in terms ofevolutionary steps, requiring 37 phylogenetic character transformations out of the absoluteminimum of 26 (C-index = 0.70), which would only obtain if all characters had compatibledistribudons among basic taxa. The consistency indices (Kluge and Farris, 1969) for thecharacters on the consensus cladogram (Fig. 50) are identical to those on the threeminimum-step cladograms (Fig. 49A,B,C) from which it was derived. These are given inTable 9. Synapomorphies for the nodes of the consensus cladogram (Fig. 50) are givenbelow, with convergent characters underlined and characters involving reversal markedwith an asterisk.Node 1: 23-B*; Node 2: 37-B, 52-A; Node 3: 5-B, 7-B, 8-B, 17-B or-C, 21-B, 36-Bor-C, 40-B, 45-B, 50-B, 83-B; Node 4: 19-B, 28-B or-C, 39-B, 46-C or-D ;Amblyrhynchus: 18-A . 54-B . 65-B . 68-B . 69-B . 70-B . 71-B . 84-A : Conolophus: none;Ctenosaura: 57-B ; Cyclura: none; Iguana: 18-A , 57-B ; Sauromalus: 23-A*, 46-C . 54-B ,65-B . 68-B . 69-B . 70-B . 71-C . 84-A .
Phylogenetic Systematics ofIguanine Lizards m
FIG. 49. Minimum-step cladograms resulting from an analysis of 26 characters (Table 9) in a subset ofiguanines. Three different cladograms (A, B, and C) account for the taxic distribution of derived characterswith 37 character transformations.
128 University of California Publications in Zoology
.i-"
<^> s^^^
.V .^^'o^^ .<>^ ^^
'^ O^
A'
,<> > A'
.">S^^ c:^c>^
FIG. 50. Consensus cladogram for the three cladograms illustrated in Figure 49. The consensuscladogram is also a minimum-step cladogram in that it requires the same number of charactertransformations as do the three fully resolved cladograms upon which it is based. Synapomorphies for thenumbered nodes and the basic taxa are given in the text.
Eight of the eleven homoplastic characters can be interpreted in two different ways,each involving the same number of phylogenetic transformations on the minimum- stepcladograms. The alternative interpretations of character 57 have already been discussed.Its derived state is either convergent in Ctenosaura and Iguana, or it is a synapomorphy of amonophyletic group composed of Ctenosaura, Iguana, and Cyclura that has subsequentlyreversed in Cyclura. Characters 54, 65, 68, 69, 70, 71, and 84 are either convergent inAmblyrhynchus and Sauromalus or they are synapomorphies of a monophyletic groupcomposed of Amblyrhynchus, Conolophus, and Sauromalus that have subsequentlyreversed in Conolophus.Although all three of the subterminal nodes on the consensus cladogram (not includingnode 1, which is a conclusion of a higher-level analysis) are supported by at least twoderived characters, every one is contradicted by some other characters. Node 2, suggestinga sister-group relationship between Sauromalus and the Galapagos iguanas, is supportedby two characters: reduced labial exposure of the angular bone (37-B) and short secondceratobranchials (52-53-A). Nevertheless, the possession of polycuspate or serratemarginal tooth crowns (character 46-B or-C) suggests that Sauromalus is more closely
Phylogenetic Systematics oflguanine Lizards 129
related to Iguana and Cyclura, while the lack of lateral contact between palatine and jugalposterior to the infraorbital foramen (character 23-A) suggests that Sauromalus may be thesister group of all other iguanines in the lower-level analysis. However, this character isactually variable within Sauromalus and may have reversed within this taxon.Node 4, suggesting a sister-group relationship between Iguana and Cyclura, issupported by four characters: squamosal abuts against dorsal end of quadrate (19-B);cristae ventrolateralis of parabasisphenoid relatively widely separated (28-B or-C);surangular extends far forward on lateral surface of mandible (39-B); and polycuspate orserrate marginal tooth crowns (46-C or-D). One of these characters (46) actually suggestsmonophyly of a more inclusive group consisting of Sauromalus, Iguana, and Cyclura.Another character, absence of posterolateral processes on pleurapophyses of second sacralvertebra (character 57-B), suggests a sister-group relationship between Iguana andCtenosaura, although most Cyclura also lack the processes. Yet another character, largeventral process of the squamosal (18-A), suggests a sister-group relationship betweenAmblyrhynchus and Iguana (the homology of this character is dubious but cannot be ruledout on morphological grounds alone).Node 3, suggesting a sister-group relationship between Amblyrhynchus andConolophus, is the best-supported node. It is diagnosed by 10 derived characters: nasalprocess of premaxilla covered dorsally between nasals (5-B); prefrontal contacts jugalbehind lacrimal foramen (7-B); frontal wider than long (8-B); reduction of lacrimal (17-Bor-C); medial crest on anterior dorsal surface of palatine (21-B); enlarged labial foot ofcoronoid (36-B or-C); surangular covered lingually below coronoid (40-B); premaxillaryteeth with large lateral cusps (45-B); anterior portion of pterygoid tooth patch absent (50-B); and pointed, conical dorsal head scales (83-B). Nevertheless, seven derived characterssuggest a sister-group relationship between Amblyrhynchus and Sauromalus: medialseparation of second ceratobranchials (54-B); reduction or loss of scapular fenestrae (65-B); short posterior process of interclavicle (68-B); T-shaped interclavicle (69-B); reductionor loss of sternal fontanelle (70-B); medial separation of xiphisterna (71-B or-C); andquadrangular, nonoverlapping superciliary scales (84-A). Conolophus lacks all of thesederived characters. Therefore, if a sister-group relationship between Amblyrhynchus andConolophus is accepted, then the derived characters shared by Amblyrhynchus andSauromalus must either be convergent or reversed in Conolophus.
PHYLOGENETIC CONCLUSIONS
PREFERRED HYPOTHESIS OF RELATIONSHIPS
Figure 51 summarizes my conclusions about phylogenetic relationships among the generaof iguanine lizards, based on the two analyses discussed above as well as a considerationof variation within basic taxa. Synapomorphies of the various taxa are given in theDiagnoses section, below. Although this is not the most fully resolved cladogram that canbe obtained from the characters used in this study, it indicates the best-supportedmonophyletic groups. The differences between this cladogram and the most fully resolvedcladogram that can be obtained from these data are as follows: (1) Either Brachylophus orDipsosaurus can be considered the sister group of all other iguanines on a fully resolvedcladogram. Since both hypotheses are equally reasonable in terms of the charactersdiscussed here, I leave the relationships among Brachylophus, Dipsosaurus, and themonophyletic group composed of all other iguanines unresolved. (2) Although it ispossible to place Ctenosaura as the sister group of the clade composed of Iguana andCyclura, this conclusion is based on one of two possible interpretations of a singlecharacter, and this character must later be lost within the clade that it is supposed todiagnose. I prefer to leave the relationships of Ctenosaura to Sauromalus, Iguana andCyclura, and Amblyrhynchus and Conolophus unresolved. (3) Finally, a fully resolvedcladogram places Sauromalus as the sister group of the Galapagos iguanas, while I leavethe relationships of Sauromalus to Ctenosaura, the Galapagos iguanas, and Iguana andCyclura unresolved. The reasons for these differences are discussed more fully in thesections on phylogenetic analysis, above, and the diagnoses of the monophyletic groups ofiguanines, below. CHARACTER EVOLUTION WITHIN IGUANINAE
Although the primary goal of this study was to determine the relationships among thegenera of iguanine lizards, I was only partially successful in this endeavor. Other thanIguaninae as a whole, I recognize only three monophyletic groups composed of more thanone of the basic taxa, whereas a fully resolved dichotomously branching phylogeny wouldhave six such groups. Failure to resolve relationships cannot be attributed to a lack ofmorphological variation within Iguaninae, for derived characters-which are sometimesnumerous-support the monophyly of each of the basic taxa. Therefore, it seems that mostof the character evolution within iguanines occurred after the lineages leading to the extant
130
Phylogenetic Systematics oflguanine Lizards 131
Iguaninae
FIG. 51. Phylogenetic relationships within Iguaninae according to the present study.
genera had already diverged from one another. Accepting this proposition might lead oneto conclude that these lineages separated during a relatively brief time interval and that theyhave been evolving separately for a long time. Implicit in this conclusion, however, is theassumption that rates of character evolution are similar in separately evolving lineages.This assumption is contradicted by the distribution of derived characters among the basictaxa and the relationships that can be resolved by them. For example, Amblyrhynchuspossesses more obvious derived characters not found in Conolophiis than does eitherBrachylophus or Dipsosaurus, even though Conolophus apparendy shared a more recentcommon ancestor with Amblyrhynchus than it did with either Brachylophus orDipsosaurus. Given that the characters used in this study are representative of overallphenotypic evolution, one must conclude that the lineage leading to Amblyrhynchus hasevolved more rapidly than those leading to Brachylophus and Dipsosaurus.
COMPARISONS WITH PREVIOUS HYPOTHESES
Although a close relationship among some or all of the taxa currently placed in Iguaninaewas recognized by several nineteenth-century authors, no explicit hypotheses aboutphylogenetic relationships among the various iguanine genera appeared until the twentiethcentury. The phylogenetic relationships proposed here are both similar in some respectsand different in others when compared with previous hypotheses about iguaninerelationships. In this section, I evaluate these previous hypotheses in light of the results ofthe present study.Barbour and Noble (1916) and Bailey (1928) both hypothesized a close relationshipbetween Cyclura and Ctenosaura, and Schwartz and Carey (1977) further proposed thatCyclura originated from Ctenosaura. Neither of these hypotheses is supported by theresults of the present study. First, Ctenosaura possesses at least three characters that arederived relative to the condition seen in Cyclura (premaxillary process of maxilla curvesdorsally; short posterolateral processes of parabasisphenoid; elongate subocular scale), andthus cannot be considered ancestral to the latter. Second, Cyclura shares more derivedcharacters with Iguana than it does with Ctenosaura, implying that Cyclura shared a morerecent common ancestor with Iguana than with Ctenosaura. The relationships among thesethree taxa are discussed further in the comments on Cyclura in the Diagnoses section,below.Mittleman (1942) proposed a phylogenetic scheme for the North American iguanids,including Ctenosaura, Dipsosaurus, and Sauroma Ius (Fig. 1). This phylogeny wasmodified slighdy by H. M. Smith (1946), who removed Ctenosaura from a position ofdirect ancestry to all other North American iguanids and placed Dipsosaurus andSauromalus close to a group composed of what are now considered the sceloporines andcrotaphytines rather than to just part of this radiation (compare Figs. 1 and 2). AlthoughSmith did not include iguanines other than those occurring within or very near to the UnitedStates in his branching diagram, it is clear from his comments on the "herbivore section"(group II in Fig. 2) that he also considered Iguana, Amblyrhynchus, Conolophus, andCyclura to be part of this group.Common to the Mittleman (1942) and Smith (1946) phylogenies is the notion thatiguanines are ancestral to the other North American iguanids-that is, that some iguaninesshared a more recent common ancestor with these other iguanids than they did with otheriguanines. This idea seems to be related to another notion held by both Mittleman andSmith, namely that iguanines are "primitive" iguanids. According to Mittleman(1942:1 12), "Dipsosaurus is probably the most primitive of the North American Iguanidae
132
Phylogenetic Systematics ofIguanine Lizards 133
(excepting Ctenosaura, which is properly a Central and South American form)." H. M.Smith (1946:101) says of his herbivore section (iguanines), "this includes the large,primitive iguanids."The notions that iguanines are "primitive" iguanids and that they are ancestral tosceloporines and crotaphytines are false. While it is true that iguanines lack certain derivedfeatures seen in these other groups, this is simply a manifestation of the mosaic nature ofevolution, for the converse is also true. Sceloporines and crotaphytines lack derivedcharacters seen in iguanines. Iguanines are derived relative to sceloporines andcrotaphytines in numerous characters, among them the possession of caudal vertebrae withtwo pairs of transverse processes, the posterior location of the supratemporal bone,herbivory and associated morphological adaptations (flared tooth crowns, colic valves),and large body size. Because some of the derived characters of iguanines occur nowhereelse within Iguanidae, iguanines cannot be considered ancestral to any other iguanids.In the early 1960's, Etheridge constructed a phylogeny for iguanines as part of hisscheme of relationships for the entire Iguanidae (Fig. 4). This scheme was never intendedto be published (Etheridge, pers. comm.), and it is difficult to evaluate because the reasonsfor the various groupings were not specified. Other than differences in resolution, theresults of the present study differ from Etheridge's scheme in two primary ways: While Iconsider Dipsosaurus and Brachylophus to be outside of a monophyletic group formed bythe remaining iguanines, Etheridge considered Brachylophus to be the sister group of theGalapagos iguanas, and he considered Dipsosaurus to be the sister group of Sauromalus.Although the relationships proposed by Etheridge can be supported by particular shared,derived characters (e.g., lack of autotomy septa in caudal vertebrae oi Brachylophus andthe Galapagos iguanas; anterior position of parietal foramen in Sauromalus andDipsosaurus), the weight of the evidence suggests different relationships and necessitatesthat the distribution of these derived characters is partly the result of convergence. The fullevidence leading to this conclusion is given in the diagnoses of the various monophyleticgroups recognized in the present study and will not be repeated here.The only published study dealing with relationships among all the iguanine genera isthat of Avery and Tanner (1971). As I noted in the Introduction, these authors used anartificial system for assessing similarity, used many characters that are probably correlated,made no attempt to determine character polarity, and did not specify how their similaritydata were used to construct their phylogenetic tree. Furthermore, Avery and Tanner'sconclusions are obscured by self-contradictory, vague, and ambiguous statements. Forexample, they state (p. 69) that "the osteological characters . . . indicate that Oplurus andChalarodon are more closely related to each other than to the iguanines, and Oplurus is theMadagascarian genus most closely related to the Western Hemisphere iguanines." In oneplace (p. 68), Avery and Tanner claim that Ctenosaura is certainly ancestral to the WesternHemisphere iguanines, but their phylogenetic tree (Fig. 3) suggests that Dipsosaurus, aWestern Hemisphere iguanine, is not derived from Ctenosaura, and later (p. 73) they seemto consider Ctenosaura ancestral to only Cyclura and Sauromalus. One of Avery andTanner's 1 1 numbered conclusions is that Iguana and Ctenosaura evolved from a common
134 University of California Publications in Zoology
ancestral stock. This statement is uninformative, for they consider all iguanines to haveevolved from a common ancestor; it is also misleading when compared with theirphylogenetic tree (Fig. 3). For these reasons, I find it impossible to compare myconclusions with those of Avery and Tanner.Wyles and Sarich (1983) published the results of immunological comparisons for 10species of iguanines representing all eight genera. Given the limitations of these data, theirresults are in general agreement with the relationships proposed here. Wyles and Sarich'scomparisons are incomplete in that antisera were prepared to only four of the iguaninespecies, and immunological distances to all other iguanines in the study are given for theantisera to only two of the four, Amblyrhynchus and Conolophus. Assuming thatimmunological distance is roughly proportional to time of divergence, Wyles and Sarich'sdata suggest (1) that Amblyrhynchus and Conolophus are sister taxa; (2) that the Galapagosiguanas are roughly equally closely related to Ctenosaura, Cyclura, Iguana, andSauromalus; and (3) that they are more distantly related to Dipsosaurus and Brachylophus.All of these conclusions are in agreement with those of the present study.
DIAGNOSES OF MONOPHYLETIC GROUPSOF IGUANINES
In this section I provide discussions of the monophyletic groups of iguanines at and abovethe level of the basic taxa used in this study (traditional genera). For each taxon I include:(1) the type on which the taxon is based, (2) the etymology of the name, (3) a phylogeneticdefinition (de Queiroz, 1987; Gauthier et al, 1988), (4) the current distribution, (5) adiagnosis consisting of hypothesized synapomorphies, (6) fossil records, and (7) variouscomments. Synonyms are not provided; those of the basic taxa can be found in Etheridge(1982).
Iguaninae Bell 1825
Type genus: Iguana Laurenti 1768.
Etymology: Modification of Iguana, the name of its type genus.
Definition: The most recent common ancestor of Brachylophus, Dipsosaunis, andIguanini, and aU of its descendants.
Distribution: Southwestern United States southward through Mexico, Central America,and northern South America to southern Brazil and Paraguay; the West Indies; theGalapagos Islands; lies Wallis; and the Fiji and Tonga island groups.
Diagnosis: Iguanines are moderate to large iguanians that can be distinguished fromother iguanians by the following synapomorphies:1. Vertebrae in part of caudal sequence bear two pairs of transverse processes(Etheridge, 1967).2. Transverse colic folds or valves present (Iverson, 1980, 1982).3. Crowns of posterior marginal teeth laterally compressed, anteroposteriorly flared,often with four or more cusps (Etheridge, 1964a).4. Supratemporal lies primarily on posteromedial surface of supratemporal process ofparietal.5. Herbivorous (H. M. Smith, 1946; Iverson, 1982).
Fossil record: The diagnosis and description of iguanines presented here enable me toreject the possible iguanine relationships of certain fossil taxa. In their description of
135
136 University of California Publications in Zoology
Paradipsosaurus mexicanus. Fries et al. (1955:15) stated that this animal "would appear toapproach more closely to the northern crested lizard Dipsosaurus than to any of the otheriguanids that presently live in Mexico and the southwestern United States." However, thesimilarities they cite (broad, flat parietal table elevated well above level of supratemporalarch; unrestricted supratemporal fossa; deep, broad snout without pronouncednasolachrymal ridges; forward opening nares), provide no evidence for a close relationshipto Dipsosaurus, since they are all plesiomorphic for Iguania. Of the five diagnosticiguanine synapomorphies identified in this study, only the morphology of the tooth crownscan be assessed in Paradipsosaurus. Unlike the teeth of iguanines, those ofParadipsosaurus are said to be a little dilated and noncuspidate (Fries et al., 1955).Furthermore, while all postembryonic iguanines and various other iguanids have arelatively small splenial and have the dentary portion of Meckel's groove closed and fused,both derived features within Iguania, the splenial of Paradipsosaurus is relatively large andMeckel's groove is open (Estes, 1983). Therefore, although Paradipsosaurus andDipsosaurus share the derived condition of having the parietal foramen located within thefrontal bone, this similarity is convergent, since Paradipsosaurus is not an iguanine. Estes(1983) reached similar conclusions concerning the relationships of this fossil.Gilmore (1928) described Parasauromalus olseni based on a fragment of a right dentaryfrom the Eocene of Wyoming. Although he did not specifically propose that it was relatedto the iguanine Sauromalus, Gilmore considered the teeth of the fossil to resemble those ofSauromalus ater most closely, made his comparisons with this species only, and named thefossil as if to suggest a close relationship with Sauromalus (para means near). If newmaterial has been correctly referred to Parasauromalus (Estes, 1983), then this taxon is notan iguanine and therefore cannot be closely related to Sauromalus. Contrary to Gilmore's(1928) statements, the tooth crowns oi Parasauromalus are not particularly similar to thoseoi Sauromalus. They are only slightly flared and tricuspid (Estes, 1983), while those ofSauromalus are strongly flared and polycuspate. The supratemporal of Parasauromalus lieson the lateral surface of the supratemporal process of the parietal (figured by Estes, 1983),whereas the supratemporal of iguanines lies in a derived position on the medial surface.The splenial of Parasauromalus is relatively large and the Meckelian groove closed butunfused (Estes, 1983), primitive iguanian characters not retained by any iguanine.The oldest fossils referred to Iguaninae for which this reference cannot be rejected areLower Miocene in age: Tetralophosaurus (Olson, 1937), a fragment of a lower jaw fromNebraska referred to Dipsosaurus by Estes (1983); a fragment of a lower jaw and a sacralvertebra from Florida (Estes, 1963); and another fragment of a lower jaw from Texas,referred to either Ctenosaura or Sauromalus by Stevens (1977). Because of theirfragmentary nature, these specimens are not definitely referable to Iguaninae on the basis ofsynapomorphies. The oldest fossil that is clearly iguanine is a nearly complete skull fromthe Pliocene of southern California (Norell, 1983). These and other fossil records aregiven under the least inclusive taxon to which they belong or are most closely related.
Phylogenetic Systematics ofIguanine Lizards 1 37
Comments: Three of the five iguanine synapomorphies are presumably part of a single
"adaptive syndrome." Both the iguanine dentition (Hotton, 1955) and colic valves(Iverson, 1980, 1982) are thought to be adaptations for a third iguanine character,herbivory. However, because this correlation of form and function does not extend to allherbivorous lizards, dentition, diet, and colic anatomy are here treated as separatecharacters.Although Iguaninae was first used by Cope (1886), Bell (1825) is credited withauthorship under the principle of coordination (Article 36, third edition of the InternationalCode ofZoological Nomenclature). The content of Iguaninae as defined here differs fromthat of Cope's (1886) Iguaninae in that the former includes Dipsosaurus and Sauromaluswhile the latter does not. Iguaninae as defined here is identical in content to an unnamedsubset of Cope's (1900) more inclusive Iguaninae and to Etheridge's (1964a, 1982)informal "iguanines."In addition to the diagnostic iguanine characters given above, acceptance of thephylogenetic relationships proposed in this paper requires that the reduction or loss of theventral process of the squamosal (character 18-A) be interpreted as an iguaninesynapomorphy that has subsequently reversed in Amblyrhynchus and Iguana.In order to facilitate diagnosis of the monophyletic subgroups of iguanines, I havereconstructed a hypothetical ancestral iguanine. This hypothetical ancestor has the derivedcharacters of iguanines as a whole but lacks the derived characters of its monophyleticsubgroups. The reason for constructing a hypothetical ancestor is that my diagnoses forthe monophyletic subgroups of iguanines consist exclusively of synapomorphies, while itmay also be useful to know what primitive features are retained by members of particularmonophyletic subgroups. Members of any monophyletic subgroup of iguanines possessthe condition found in the hypothetical ancestor unless an alternative state of the samecharacter is listed as a diagnostic synapomorphy either of the taxon in question or of alarger monophyletic taxon of iguanines within which the taxon in question is included. Itshould be kept in mind that the presence of a primitive character properly indicates only thatthe specimen possessing it does not belong to the taxon diagnosed by the derivedalternative condition. It does not preclude the possibility that the specimen in question,perhaps some newly discovered fossil, is not most closely related to the taxon diagnosedby the derived condition.The hypothetical ancestral iguanine is thought to have possessed the followingmorphological features (numbers and letters correspond with those in the list of systematiccharacters):1-A. Ventral surface of premaxilla bears large posterolateral processes.2-A. Posteroventral crests of premaxilla small, not continuing up sides of incisiveprocess and not pierced by foramina for maxillary arteries.3-A. Anterior surface of premaxilla broadly convex.4-A. Nasal process of premaxilla slopes posteriorly.5-A. Nasal process of premaxilla exposed broadly between nasals.6-A. Nasal capsule of moderate size, nasals relatively small.
138 University of California Publications in Zoology
1-A.. Lacrimal contacts palatine, and prefrontal fails to contact jugal behind lacrimalforamen.8-A. Frontal longer than wide.9-A. Paired openings near frontonasal suture small or absent.10-A. Cristae cranii of frontal form a smooth, continuous curve from frontal toprefrontal.1 1-A. Frontal cristae medial to cristae cranii absent or weakly developed.12-A. Dorsal borders of orbits form a more or less smooth curve.13-A. Parietal foramen lies on frontoparietal suture.14-A. Supratemporal extends anteriorly more than halfway across posterior temporalfossa.15-A. Lateral surfaces of maxillae relatively flat or concave below supralabialforamina.16-A. Premaxillary process of maxilla not curving dorsally; maxillary and premaxillaryteeth lie in the same plane.17-A. Lacrimal relatively large.18-B. Ventral process of squamosal reduced or absent.19-A. Squamosal does not abut against tympanic crest of quadrate.20-A. Septomaxilla without pronounced longitudinal crest on anterolateral surface.2 1-A. Palatine without high crest on dorsomedial surface.22-A. Large posterolateral process of palatine behind infraorbital foramen present.23-A. Posterolateral process of palatine behind infraorbital foramen fails to contactjugal. Contact of this process with the jugal may be a synapomorphy of all iguanines thathas been lost secondarily in Dipsosaurus.24-A. Infraorbital foramen located on lateral or posterolateral edge of palatine.25-A. Medial borders of pterygoids relatively straight anterior to pterygoid notch,pyriform recess narrows gradually anteriorly. Sharply curved medial pterygoid bordersand a pyriform recess that narrows abruptly may be a synapomorphy of all iguanines thathas been secondarily lost in Brachylophus.26-A. Ectopterygoid fails to contact palatine at posteromedial comer of suborbitalfossa.27-A. Long parasphenoid rostrum.28-A. Cristae ventrolaterals of parabasisphenoid strongly constricted behindbasipterygoid processes.29-A. Posterolateral processes of parabasisphenoid large, extending far up anterioredges of lateral processes of basioccipital.30-A. Laterally directed pointed process of cristae interfenestralis absent.3 1-A. Stapes relatively thin.32-A. Dorsal edges of dentary and surangular on either side of coronoid eminenceapproximately equal in height.33-A. Splenial relatively large.
Phylogenetic Systematics ofIguanine Lizards 139
34-35-A. Anterior inferior alveolar foramen lies between splenial and dentary;coronoid may or may not contribute to its posterior margin.36-A. Labial process of coronoid present but relatively small.37-A. Angular extends far up lateral surface of mandible and is easily visible in lateralview.38-A. Angular wide posteriorly.39-A. Surangular does not extend anteriorly to last dentary tooth on labial surface ofmandible.40-A. Dome-shaped portion of surangular visible below coronoid on lingual surface ofmandible.41 -A. Angular process of prearticular increases substantially in relative size duringpostembryonic ontogeny, becoming a prominent structure in adults.42-A. Outline of retroarticular process triangular rather than quadrangular in allpostembryonic developmental stages.43-44-B. Mode of seven premaxillary teeth.45-A. Lateral cusps of premaxillary teeth small or absent.46-A. Posterior marginal teeth tricuspid. The presence of a fourth cusp may be asynapomorphy of all iguanines, with secondary loss in Amblyrhynchus and in someBrachylophus and Ctenosaura. Alternatively, the ancestral iguanine may have beenpolymorphic for the presence of a fourth cusp (again with secondary loss inAmblyrhynchus and some Ctenosaura).41-A. Individual lateral cusps of tricuspid marginal teeth much smaller than apicalcusp.48-A. Entire pterygoid tooth row lies close to ventromedial edge of pterygoid.49-A. Pterygoid tooth patch consists of a single row of teeth throughoutpostembryonic ontogeny.50-A. Pterygoid tooth patch extends anteriorly beyond level of posterior edge ofsuborbital fenestra.51 -A. Pterygoid teeth present.52-53-B. Second ceratobranchials from two-thirds length to slightly longer than firstceratobranchials.54-A. Second ceratobranchials in medial contact for most or all their lengths.55-A. Neural spines of presacral vertebrae tall, more than 50% of total vertebralheight.56-A. Zygosphenes connected to prezygapophyses by continuous arc of bone.57-A. Posterolateral processes present on pleurapophyses of second sacral vertebra.58-A. Foramina present in ventral surface of pleurapophyses of second sacral vertebra.59-A. More than 40 caudal vertebrae.60-A. Caudal autotomy septa present. The polarity of this character is questionable.61 -A. Autotomic caudal series (or series of caudal vertebrae with paired transverseprocesses) begins at or before 10th caudal vertebra. The polarity of this character isquestionable.
140 University of California Publications in Zoology
62-A. Dorsal midsagittal fins of caudal vertebrae anterior to neural spines relativelylarge and present well beyond anterior third of caudal sequence.63-A or -B. Postxiphisternal inscriptional ribs do not form continuous chevrons, oranteriormost pairs do only variably.64-A. Suprascapulae oriented primarily vertically and form a continuous arc with thescapulocoracoids.65-A. Scapular fenestrae present and large.66-A. Posterior coracoid fenestrae absent.67-A. Clavicles wide, with prominent lateral shelves.68-A. Posterior process of interclavicle extends well beyond lateral corners ofsternum.69-A. Interclavicle arrow-shaped, lateral processes forming angles of less than 75?with posterior process.70-A. Sternal fontanelle present and of moderate size.71-A. Sternum diamond-shaped, xiphisternal rods attach close to midline.72-A. Pelvic girdle relatively long and narrow.73-A. Large anterior iliac process.74-A. Cephalic osteoderms absent.75-A. Heart lies entirely anterior to transverse axillary plane.76-A. Subclavian arteries covered ventrally by posterior end of M. rectus capitisanterior.77-A. Right and left systemic arches unite to form dorsal aorta above heart.78-A. Coeliac artery arises from dorsal aorta anterior to and separate from mesentericarteries.79-A. Colic wall with one or more transverse valves.80-A. All colic valves semilunar. The polarity of this character is questionable.81-A. Median azygous rostral scale present.82-A. Snout scales small and numerous, approximately same size as those ofsupraorbital and temporal regions.83-A. Dorsal head scales flat or only slighdy convex.84-B. Superciliary scales moderately elongate and partially overlapping. It is alsopossible that the ancestral iguanine had elongate and strongly overlapping superciliaries.85-A or -B. Subocular scales subequal in size, or one or two moderately elongate.86-A. Anterior auricular scales small or only slighdy enlarged.87-A. Gular fold well developed.88-A. Dewlap small or absent. The polarity of this character is questionable.89-A. Gular crest of enlarged scales absent.90-A. Middorsal scale row present.91-A. Pedal subdigital scales asymmetrical, anterior keels larger than posterior ones.92-A. Pedal subdigital scales lack greatiy enlarged anterior keels fused at their bases toform combs.93-A. Toes unwebbed.
Phylogenetic Systematics of[guanine Lizards 141
FIG. 52. Geographic distribution oi Dipsosaurus (modified from Stebbins, 1966).
94-A. Caudal scales in adjacent verticils approximately equal in size, smooth or keeledbut not spinous.95-A. Body laterally compressed or roughly cylindrical.
Dipsosaurus Hallowell 1 854
Type species (by monotypy): Crotaphytus dorsalis Baird and Girard 1852.
Etymology: (Greek) Dipsa, thirst(y), + sauros, lizard. Dipsosaurus was first knownfrom the "Colorado Desert" of western North America, as Hallowell (1854:92) described it
"a country without water."
Definition: The most recent common ancestor of the populations of RecentDipsosaurus dorsalis and all of its descendants.
Distribution: Deserts of the southwestern United States in southeastern California,southern Nevada, southwestern Utah, and western Arizona, southward into Mexicothrough western Sonora and northwestern Sinaloa and into Baja California to its southernend, including various islands in the Gulf of California (Fig. 52).
142 University of California Publications in Zoology
Diagnosis: Members of this taxon can be distinguished from other iguanines by thefollowing synapomorphies (here and afterwards the parenthetical numbers and letterscorrespond with those in the list of systematic characters):1. Large, paired openings at or near frontonasal suture present (9-B).2. Parietal foramen located entirely within frontal bone (13-C). This character occursalso in Cyclura carinata and variably in some Ctenosaura, Sauromalus, and other Cyclura.3. Lateral process of palatine behind infraorbital foramen small or absent (22-B).4. Medial borders of pterygoids curve sharply toward midline anterior to pterygoidnotch; pyriform recess narrows abruptly (25-B). This character occurs in all otheriguanines except Brachylophus and may thus be a synapomorphy of Iguaninae that hasreversed in Brachylophus.5. Lateral pointed processes on cristae interfenestralis present (30-B).6. Posterior ends of lateral and medial crests of retroarticular process divergeontogenetically, so that outline of retroarticular process is quadrangular in large specimens(42-B).7. Crowns of posterior marginal teeth with four cusps (46-B). An increase in toothcuspation characterizes all other iguanines except Amblyrhynchus and some Brachylophusand Ctenosaura; therefore this character may be a synapomorphy of a more inclusive groupthat has reversed in certain taxa.8. Pterygoid teeth usually absent (50-B, 51-B), This character also occurs inConolophus. When present, the pterygoid teeth of Dipsosaurus lie along the medial edgeof the pterygoid, while those of Conolophus lie more laterally, supporting the conclusionthat the absence of pterygoid teeth in these two taxa is convergent.9. Colon with one or more circular valves (80-B). This condition occurs also in allother iguanines except Brachylophus and may be a synapomorphy of a more inclusivegroup.10. Superciliary scales greatly elongate and strongly overlapping (84-C). The derivedstatus of this character is questionable.11. One subocular scale much longer than others (85-C). The derived status of thischaracter is questionable.
Fossil record: Olson (1937) described Tetralophosaurus minutus based on a fragmentof a lower jaw from Lower Miocene deposits in Nebraska. The specimen was referred toDipsosaurus by Estes (1983), who stated that it was indistinguishable from D. dorsalis, butthis conclusion is based on overall similarity. Almost complete skulls and dentaries fromthe PUocene of southern California have been referred to Dipsosaurus by Norell (1983).Comments: Failure of the lateral palatine process to contact the jugal behind theinfraorbital foramen (character 23) suggests that Dipsosaurus is the sister group of all otheriguanines. However, the gently curving medial pterygoid borders and wide pyriformrecess of Brachylophus (character 25) suggest that this taxon, rather than Dipsosaurus, is
Phylogenetic Systematics ofIguanine Lizards 143
the sister group of all other iguanines. The weaker tendency of Brachylophus to developfourth cusps on the posterior marginal teeth might be taken as further evidence in favor ofthe latter hypothesis, but the character is variable in Brachylophus and has reversed severalother times within iguanines. At least three other characters might be used to support oneor the other of these alternative hypotheses, but these characters must be used with cautionbecause their polarities are unclear. These are: (1) the lack of a notch separatingzygosphenes from prezygapophyses in Dipsosaurus (character 56); (2) the absence ofcircular colic valves in Brachylophus (character 80); and (3) the low number of colic valvesin Dipsosaurus (Iverson, 1982). Camp (1923) noted another character in which alliguanines except Dipsosaurus share what appears to be a derived condition (Conolophuswas not examined): a high degree of separation of the M. mylohyoideus anteriorsuperficialis. Because of this contradictory information, I have chosen to leave therelationships among Dipsosaurus, Brachylophus, and the monophyletic group consistingof the remaining iguanines (Iguanini) unresolved. I am not aware of any characterssuggesting that Dipsosaurus and Brachylophus are sister taxa.
Brachylophus Wagler 1830
Type species (by monotypy): Iguanafasdata Brongniart 1800.
Etymology: (Greek) Brachys, short, + lophos, a crest. The name presumably refers tothe relatively short scales of the dorsal crest in B.fasciatus, the type species.
Definition: The most recent common ancestor of B.fasciatus and B. vitiensis and all ofits descendants.
Distribution: Numerous islands in the Fiji Islands group, Tongatapu in the TongaIslands group, and lies Wallis northeast of Fiji, all in the southwestern Pacific Ocean (Fig.53).
Diagnosis: Members of this taxon can be distinguished from other iguanines by thefollowing synapomorphies:1. Lateral process of palatine behind infraorbital foramen contacts jugal (23-B). Thischaracter occurs in all iguanines except Dipsosaurus and some specimens of Sauromalus,and may be a synapomorphy of a more inclusive group.2. Infraorbital foramen located entirely within palatine bone, may or may not beconnected to lateral edge of palatine by suture (24-B). This character also occurs in someAmblyrhynchus, some Ctenosaura, and some Sauromalus, in which it is interpreted asconvergent.3. Anterior inferior alveolar foramen located entirely within dentary (34-35-B). Thischaracter occurs only in Brachylophus within Iguaninae, but does not occur in allspecimens.
144 University of California Publications in Zoology
FIG. 53. Geographic distribution oi Brachylophus (from Gibbons, 1981; Etheridge, 1982).
4. Labial process of coronoid moderately large (36-B). The enlarged labial coronoidprocess of Amblyrhynchus and Conolophus is interpreted as convergent.5. Second ceratobranchials much longer than first ceratobranchials (52-53-C). Thelong second ceratobranchials oilguana iguana are interpreted as convergent.6. Zygosphenes separated from prezygapophyses by a deep notch (56-B). Thischaracter occurs in all iguanines except Dipsosaurus, and may be a synapomorphy of amore inclusive group.7. Caudal autotomy septa absent (60-B). Although the outgroup evidence isequivocal, I have assumed that the presence of caudal autotomy, and the intravertebral septathat facilitate it, are primitive for iguanines. The absence of caudal autotomy septa inAmblyrhynchus and Conolophus on the one hand and in Iguana delicatissima on the otherare interpreted as convergent.8. Midsagittal processes on dorsal surfaces of caudal centra anterior to neural spinerelatively small and confined to anterior fifth of caudal sequence (62-B). This characteralso occurs in Iguana, in which it is interpreted as convergent.9. Anterior postxiphistemal inscriptional ribs enlarged and members of at least one pairunited midventrally to form continuous chevrons (63-C). Midventrally continuouschevrons formed by the first pair of postxiphistemal inscriptional ribs occur in variousother iguanines but not invariably within species, as in Brachylophus. Unlike other
Phylogenetic Systematics ofIguanine Lizards 1 45
iguanines, Brachylophus also exhibits enlargement of the second and third postxiphisternalinscriptional ribs, which may also unite to form continuous chevrons.10. Large dewlap present (88-E). The two species oi Brachylophus differ in that alarge dewlap is present in both sexes of B. vitiensis but only in male B.fasciatus(Gibbons, 1981). The polarity of this character is uncertain. If presence of a large dewlapis derived, then the phylogenetic relationships proposed here require that it has evolvedconvergently in Iguana and in some species of Ctenosaura.In addition, the following derived character occurs in some Brachylophus:Posterior marginal teeth with a fourth cusp (46-B). This character occurs in all otheriguanines except Amblyrhynchus and some Ctenosaura; it may thus be a synapomorphy ofa more inclusive group, perhaps of all iguanines.
Fossil record: Bones thought to be remains of Brachylophus are known fromarchaeological sites on Tongatapu and Lifuka in the Tonga Islands group (approximately2000 years before present). If correcdy referred, these bones indicate that Brachylophusonce reached much larger sizes than they do today (Etheridge, pers. comm.; Pregill, pers.comm.).Comments: Gibbons (1981) discusses the authorship oi Brachylophus, crediting thename to Wagler (1830), since Cuvier (1829) had used the informal apellation lesBrachylophes. The relationships oi Brachylophus to Dipsosaurus and other iguanines arediscussed in the comments on Dipsosaurus, above.
IguaniniBell 1825Type genus: Iguana Laurenti 1768.
Etymology: Modification of Iguana, the name of its type genus.
Definition: The most recent common ancestor of Ctenosaura, Sauromalus,Amblyrhynchina, and Iguanina, and all of its descendants.
Distribution: Southwestern United States southward through Mexico, Central America,and northern South America to southern Brazil and Paraguay, the West Indies, and theGalapagos Islands.Diagnosis: Members of this taxon can be distinguished from other iguanines{Brachylophus and Dipsosaurus) by the following synapomorphies:1. Lateral process of palatine contacts jugal behind infraorbital foramen (23-B). Thischaracter does not occur in some Sauromalus, where it is interpreted as a reversal. It doesoccur in Brachylophus and may thus be a synapomorphy of a more inclusive group.
146 University of California Publications in Zoology
2. Medial borders of pterygoids curve sharply toward midline anterior to pterygoidnotch; pyriform recess narrows abrupdy (25-B). This character occurs also in Dipsosaurusand may be a synapomorphy of a more inclusive group.3. Crowns of posterior marginal teeth with four or more cusps (46-B,-C, or-D). Thischaracter occurs also in Dipsosaurus and some Brachylophus, and may be a synapomorphyof all iguanines. It has reversed in Amblyrhynchus and some Ctenosaura.4. Posterior portion of pterygoid tooth patch displaced laterally away from medialborder of pterygoid (48-B). Pterygoid teeth are absent in most Conolophus, but whenpresent they lie away from the medial pterygoid border. This character develops duringpostembryonic ontogeny and is not always evident in small specimens.5. Zygosphenes separated from prezygapophyses by a deep notch (56-B). Thischaracter occurs also in Brachylophus and may be a synapomorphy of a more inclusivegroup.6. Sequence of autotomic caudal vertebrae or that of vertebrae with two pairs oftransverse processes begins at or behind 10th caudal vertebra (61-B). The polarity of thischaracter is questionable.7. Posterior coracoid fenestra usually present (65-B). This character exhibits somevariation within basic taxa.8. Right and left systemic arches unite to form dorsal aorta posterior to heart (77-B).9. One or more circular colic valves present (80-B). This character occurs also inDipsosaurus and may be a synapomorphy of a more inclusive group.
Fossil record: The earliest fossils that are clearly referable to Iguanini are from thePliocene of southern California. Among extant Iguanini these fossils appear to be mostclosely related to Iguana (Norell, 1983). Stevens (1977) considered a dentary fragmentfrom the early Miocene of Texas to be either Ctenosaura or Sauromalus. If correctlyreferred, this would be the oldest record of Iguanini. These and other fossil records aregiven under the least inclusive taxa to which they belong or are most closely related.Comments: Although this is the first use of Iguanini, Bell (1825) is credited withauthorship under the principle of coordination (Article 36, third edition of the InternationalCode of Zoological Nomenclature). Iguanini contains all the really large iguanines, andlarge body size may be an additional synapomorphy of this taxon. Some Ctenosaura arerelatively small, but this probably represents a secondary reduction in size (see commentson Ctenosaura, below). Relationships among four recognizable monophyletic subgroupsof Iguanini are uncertain and are discussed in greater detail in the comments on Ctenosaura,Sauromalus, Amblyrhynchina, Iguanina, and Cyclura.Ctenosaura Wiegmann 1 828Type species (subsequent designation by Fitzinger 1843): Ctenosaura cycluroidesWiegmann 1828 = Lacerta acanthura G. Shaw 1802.
Phylogenetic Systematics ofIguanine Lizards 147
FIG. 54. Geographic distribution of Ctenosnura (from Peters and Donoso-Barros, 1970; H. M. Smith,1972; Etheridge, 1982).
Etymology: (Greek) Ktenos, comb, + sauros, lizard, referring to the dorsal crest ofenlarged scales.
Definition: The most recent common ancestor of the extant species of Ctenosaura(acanthura, baked, clarki, defensor, hemilopha, palearis, pectinata, quinquecarinata, andsimilis) and all of its descendants.
Distribution: Lowlands of Mexico and Central America from southeastern BajaCalifornia and the middle of Sonora in western Mexico and near the Tropic of Cancer ineastern Mexico southward through most of Central America to central Panama, as well asIsla de Providencia, Isla de San Andres, the Tres Marias Islands, and various offshoreislands in the eastern Pacific, the western Caribbean, and the Sea of Cortez (Fig. 54).
Diagnosis: Members of this taxon can be distinguished from other iguanines by thefollowing synapomorphies:1 . Premaxillary process of maxilla curves dorsally; premaxillary teeth set higher thanmaxillary teeth (16-B). This character is not present in small specimens.
148 University ofCalifornia Publications in Zoology
2. Posterolateral processes of parabasisphenoid absent or relatively small (29-B).3. Posterolateral processes on pleurapophyses of second sacral vertebra absent (57-B).This character also occurs in Iguana and most Cyclura, and may be a synapomorphy of amore inclusive group.4. One subocular scale very long (85-C). The polarity of this character isquestionable. An elongate subocular occurs also in Dipsosaurus, in which it is interpretedas convergent.5. Tail bears whorls of enlarged, spinous scales (94-B). This character occurs also inmost Cyclura, in which it is interpreted as convergent.Other derived characters occur only in some Ctenosaura and may provide usefulinformation concerning relationships within this taxon:1. Prefrontal contacts jugal behind lacrimal foramen (7-B). This character also occursin Amblyrhynchus, Conolophus, and some Cyclura; within Ctenosaura, prefrontal-jugalcontact is characteristic only ofC clarki and may be a synapomorphy of that taxon.2. Crista cranii forms step rather than smooth curve between frontal and prefrontal (10-B). This character also occurs in Conolophus; within Ctenosaura it occurs only in C.defensor and may be a synapomorphy of that taxon.3. Parietal foramen located entirely within frontal (13-B). This character occurs also inDipsosaurus and in some Cyclura and Sauromalus; within Ctenosaura it varies as muchwithin species as among them, and it is therefore uninformative about relationships amongthese species.4. Infraorbital foramen located entirely within palatine (24-B). This character alsooccurs in Brachylophus and in some Amblyrhynchus and Sauromalus; within Ctenosaura itvaries as much within species as among them, and it is therefore uninformative aboutrelationships among these species.5. Surangular extends anteriorly well beyond coronoid apex and sometimes beyondposteriormost dentary tooth (39-B). This character occurs also in Iguana and Cyclura; itspattern of variation within Ctenosaura needs further study.6. Crowns of posterior marginal teeth polycuspate (46-C). This character occurs alsoin Iguana, Cyclura, and Sauromalus; within Ctenosaura it occurs only in C. defensor andmay be a synapomorphy of that taxon.7. Crowns of posterior marginal teeth tricuspid (46-A). Within Ctenosaura thischaracter, a presumed reversal, occurs in C. bakeri and C. quinquecarinata.8. Posterior portion of pterygoid tooth patch doubles ontogenetically (49-B). Thischaracter, or a further modification of it, occurs also in Iguana and some Cyclura. Sincemembers of the small species of both Ctenosaura and Cyclura do not exhibit ontogeneticdoubling of the tooth row, and since small maximum size in these taxa is thought to bederived (see comments on Iguanini, above), it is likely that this character is asynapomorphy at a higher level and that failure to double the pterygoid tooth row is derivedwithin Ctenosaura.9. Fewer than 40 caudal vertebrae (59-B). This character also occurs in Sauromalus;within Ctenosaura it occurs in C. clarki and C defensor.
Phylogenetic Systematics ofIguanine Lizards 1 49
10. Large dewlap (88-B). The polarity of this character is questionable. Largedewlaps occur also in Brachylophus and Iguana; within Ctenosaura they occur only in C.palearis.
Fossil record: The oldest fossils referred to Ctenosaura are from the Holocene ofMexico (Langebartel, 1953; Ray, 1965; Estes, 1983). Stevens (1977) suggested that afragment of a left dentary from the early Miocene of Texas was probably close toCtenosaura.Comments: Bailey (1928:7) claimed that "it is impossible to distinguish between thegenus Ctenosaura and its near allies by means of skeletal characters." This is false.Osteological synapomorphies are identifiable not only in Ctenosaura but also in all of theother iguanine taxa that have traditionally been assigned the rank of genus. Even withinCtenosaura, monophyletic groups can be recognized on the basis of skeletal characters.At least three characters suggest a close relationship among Ctenosaura, Iguana, andCyclura: extension of the surangular well anterior to the coronoid apex (39-B); tendency ofthe pterygoid tooth row to double ontogenetic ally (49-B,-C); and absence of posterolateralprocesses on the pleurapophyses of the second sacral vertebrae (57-B). Nevertheless, Ihave left the relationships of Ctenosaura to other Iguanini unresolved because all three ofthese characters are ambiguous. The first is variably present in Ctenosaura, the third isvariable in Cyclura, and the second is variable in both Ctenosaura and Cyclura. Thus,provided that the monophyly of each of these taxa is accepted, every one of these charactersmust involve homoplasy. If the homoplasy is interpreted as acquisition of the derived stateof these characters in the most recent common ancestor of Ctenosaura, Cyclura, andIguana, with subsequent reversal in certain taxa, then the close relationship among thesethree taxa might still be advocated. At present, however, the homoplasy can just asreasonably be interpreted as convergence, in which case the close relationship is notsupported. I prefer to leave the relationships of Ctenosaura within Iguanini unresolveduntil additional evidence suggests that one of the alternative interpretations of homoplasy inthe characters that vary within basic taxa is more plausible. The relationship betweenCtenosaura and Cyclura is discussed further in the comments on Cyclura, below.The species bakeri, clarki, defensor, palearis, and quinquecarinata, here included inCtenosaura, are sometimes placed in a separate genus, Enyaliosaurus. Etheridge (1982)reviewed the history of the problem as follows:The most recent taxonomic revision and key for the genus Ctenosaura is that ofBailey (1928), but several important papers on individual species or groups ofspecies have appeared subsequentiy. Bailey recognized 13 species, including thoseforms with a relatively small body size and a short, strongly spinose tail referred bysome authors to Enyaliosaurus. Following Gray's (1845) description ofEnyaliosaurus the name was seldom used until its revival by Smith and Taylor(1950: 75). In this work the species clarki, defensor, erythromelas, palearis and
150 University of California Publications in Zoology
quinquecarinata were allocated to Enyaliosaurus, but no justification was providedfor the revival of the genus. Duellman (1965: 599), followed Smith and Taylor inrecognizing the validity of Enyaliosaurus, placed erythromelas in the synonymy ofdefensor, provided a key to the species, and suggested that: "Enyaliosaurusdoubtless is a derivative of Ctenosaura, all species of which are larger and haverelatively longer tails and less well-developed spines than Enyaliosaurus." Meyerand Wilson (1973) referred Ctenosaura bakeri to Enyaliosaurus, but Wilson andHahn (1973: 114-5) returned bakeri to Ctenosaura, commenting that: "John R.Meyer is currently studying the problems of the relationship of the species nowgrouped in Enyaliosaurus to those now grouped in Ctenosaura. He (pers. comm.)advised us that he considers the two genera inseparable, and that bakeri appears tobe closely related to both palearis (now in Enyaliosaurus) and similis (now inCtenosaura)." In addition, Ernest Williams of Harvard University has informed me(pers. comm.) that based on an unpublished study of the group by him and ClaytonRay, he does not believe the recognition of Enyaliosaurus is warranted. At thepresent time the problem of the relationships of Ctenosaura and Enyaliosaurus areunder study by Diderot Gicca of the Florida State Museum. (Etheridge, 1982:9-10)More recently, Gicca (1983) recognized the genus Enyaliosaurus.Evidence for the monophyly of Ctenosaura in the broad sense of Bailey (1928) hasbeen presented above. An evaluation of the monophyletic status of Ctenosaura in thenarrow sense, and of Enyaliosaurus, required a phylogenetic analysis using the species ofboth as basic taxa. In this analysis, I have used primarily characters recognized byprevious workers, in particular, Bailey (1928), Smith and Taylor (1950), and Ray andWilliams (unpubl.). When possible, all characters were checked on specimens. Myanalysis is based on the following 19 characters representing a minimum of 23phylogenetic transformations. The polarities of these characters were determined usingAmblyrhynchina, Iguanina, Sauromalus, Dipsosaurus, and Brachylophus as outgroups.The character-state codes are as follows: 0, ancestral; 1, derived; 2, further derived; etc.Letter codes are used for characters whose polarities were considered undeterminable.1. Maximum snout-vent length: (0) greater than 190 mm; (1) less than 190 mm.Maximum snout-vent lengths for the various taxa are as follows: acanthura = 215 mm(MCZ 16074, Bailey, 1928; 315 mm according to Ray and Williams, unpubl., but theyinclude pectinata in acanthura); bakeri = 210 mm (USNM 25324, Bailey, 1928); clarki =154 mm (UMMZ 112711, Duellman and Duellman, 1959); defensor = 155 mm (HM 3420,Bailey, 1928); hemilopha = approximately 400 mm (H. M. Smith, 1972; the largestspecimen that Bailey [1928] presents data for is AMNH 2073 with a snout-vent length of260 mm); palearis = 254 mm (CAS 69308, A. Bauer, pers. comm.); pectinata = 305 mm(MCZ 2726, Bailey, 1928); quinquecarinata = 169 mm (Hidalgo, 1980; Gicca, 1983); andsimilis = 489 mm (Fitch and Hackforth-Jones, 1983). A cutoff of 190 mm was chosen,partly because of an apparent gap and partly because all other species of Iguanini reachgreater maximum snout-vent lengths than this.
Phylogenetic Systematics ofIguanine Lizards 151
2. Modal number of presacral vertebrae (Table 4): (0) 24; (1) 25.3. Modal number of premaxillary teeth (Table 3): (0) seven; (1) five. AlthoughCtenosaura defensor is the only species with a mode of five premaxillary teeth (range 5-6),the occurrence of five. premaxillary teeth in some specimens of C. clarki and C.quinquecarinata, but in no other Ctenosaura, suggests that these three species form amonophyletic group.4. Anterior orbital region (Fig. 10): (A) lacrimal contacts palatine behind lacrimalforamen; (B) prefrontal contacts jugal behind lacrimal foramen.5. Cristae cranii (Fig. 12): (0) form smooth curve from frontal to prefrontal; (1)frontal portions protrude anteriorly forming a step from frontal to prefrontal.6. Parietal roof: (0) remains deeply notched posteriorly throughout ontogeny, so thatthe braincase is broadly exposed in dorsal view; (1) extends posteriorly as a flat shelfduring postembryonic ontogeny, so that the braincase comes to be largely covered in dorsalview. This character is partially correlated with character 1, body size.7. Ontogenetic convergence of lateral edges of parietal roof: (A) eventually meetposteriorly and form a midsagittal crest, giving the parietal roof a Y-shaped outline; (B) failto meet, or meet but fail to form a midsagittal crest, giving the parietal roof a trapezoidal ortriangular outline. This character is partially correlated with character 1, body size.8-9. Crowns of posterior marginal teeth: (AO) with a maximum of four cusps; (BO)with a maximum of five or more cusps; (Al) with a maximum of three cusps.10. Pendulous dewlap: (0) absent; (1) present but small; (2) present and large.11. Parietal eye: (0) conspicuous externally; (1) external signs inconspicuous orabsent. This character may also be manifested in a reduction in the parietal foramen in C.defensor, but my osteological sample of this taxon is small (N=l).12. Dorsal crest scales I: (0) conform in color and pattern to adjacent body scales;adjacent crest scales similar in size; (1) unicolored and differing from body color; large,flap-like crest scales separated by one or more smaller scales.13. Dorsal crest scales II: (0) high-keeled, large, and conspicuous, at least in neckregion; (1) low-keeled to flat, inconspicuous throughout length of crest.14. Middorsal scale row: (0) continuous from neck onto tail, or narrowly interruptedin sacral region; (1) broadly discontinuous in lumbosacral region.15. Scales of anterodorsal surface of leg: (0) not enlarged or spinous; (1) enlarged andspinous on shank but not on thigh; (2) enlarged and spinous on both shank and thigh. Anadditional state could be recognized, since C. clarki and C. quinquecarinata have largeanterodorsal thigh scales compared to those of most other Ctenosaura, but these scales arenot as large as in C. defensor, and they are not spinous.16. Subdigital scales at the base of pedal digit 111: (0) with relatively small anteriorkeels or with moderately large anterior keels that are separate from those of adjacent scales;(1) with relatively large anterior keels fused at their bases to form a comb.17. Tail: (0) strongly spinose proximally, but not distally, and always longer thanbody (snout-vent length/total length = 0.27-0.45), more than 30 caudal vertebrae; (1) tail
152 University of California Publications in Zoology
strongly spinose throughout its length and almost the same length as the body (snout-ventlength/total length = 0.48-0.56), fewer than 30 caudal vertebrae.18. Anterior (referring to first 10) whorls of strongly spinous caudal scales: (0)always separated by at least two rows of intercalary scales; (1) at least some separated byonly one intercalary scale row, others by two or more; (2) none (or only the first) separatedby two intercalary scale rows, but all separated by at least one; (3) intercalary scales ofproximal whorls greatly reduced or absent.19. Snout region: (0) not inflated, sloping gradually downward; (1) inflatedanteriorly, sloping abruptly downward.Height of the vertebral neural spines may also be a useful character, but I have chosennot to use it because I have no postcranial skeletons of C. defensor and C. palearis.The distributions of these character states among basic taxa within Ctenosaura (sensulato) and three near (Amblyrhynchina, Iguanina, Sauromalus) and two more distant(Dipsosaurus, Brachylophus) outgroups are given in Table 10. Ctenosaura bakeri fromIsla de Utila and those from Isla de Roatan are scored separately because they differ in atleast three of the characters used in this analysis. Only those from Utila, the type locality,are included in the analysis of relationships.The phylogenetic relationships suggested by the characters in Table 10 (except character19, the derived state of which occurs only in the Roatan population of C. bakeri) arediagrammed in Figure 55. Synapomorphies for the subterminal nodes and the basic taxaare given below. Characters whose polarities were initially undeterminable were placed onthe cladogram after it was constructed using only those characters whose polarities weredeterminable using other iguanines as outgroups. Ignoring the Roatan ctenosaurs andintraspecific variation, these relationships require a total of 25 character transformations,three more than the minimum number required by the characters themselves (C-index =0.88). C-indices for the individual characters are given in Table 10.Node 1: Ctenosaura Wi^gmdiXm 1828See above. The characters of the hypothetical ancestral Ctenosaura can bereconstructed by taking the first state of each of the 19 characters in the character list.
1).
Node 2 (unnamed)1. Parietal roof extends posteriorly over braincase during postembryonic ontogeny (6-
Ctenosaura acanthuraNo synapomorphies identified. Ctenosaura pectinataNo synapomorphies identified.
Phylogenetic Systematics ofIguanine Lizards 153
TABLE 10. Distributions of Character States of 19 Characters Among Basic Taxa WithinCtenosaura (in the broad sense) and Three Close and Two More Distant Outgroups
Taxon Character1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
acanthwa OOOAO lAAOOOOOOOOOOObakeriqj\i\2L) OOOI--OBAI 10100100 0,1te)ten (Roatan) OOOAOOBA 1 00000 1 00 0,1 1
clarki llOBOOBAOOOOl 0,1 10120(kfensor 111A10BB001010,121130hemilopha OOOAOOAAOOOOOIOOOIOpalearis 000A00BA020100 10020pectimta OOOAO lAAOOOOOOOOOOOquinquecarinata 1 1 OA,BOOBA 1 00000 1 0020
similis OOOAOOAAOOOOOOOOOOOCI 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 0.5 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 0.6 1.0
Amblyrhynchina B 0,1 A,B A 0,1 02 0^ N^Iguanina Of^ A,B A,B B 1,2 0,1 0^ 0,n3o,1Sauromalus 1^ A B B N^ N^ 1^ o'' O^'^ N^
Dipsosaurus 1 OOAOOBAOOOO 1 OOOO^N^OBrachylophus A A,B A 0,1 0,1 0^ N^
Note: Character-state codes correspond with those used in the character list. A dash indicates the lack ofdata. Consistency indices (CI) for each character on the minimum-step cladogram for these characters (Fig.55) are also given. The consistency indices were calculated ignoring intraspecific variation.^50% have seven and 50% have six (N = 2).
^Large, conical crest scales are separated by smaller ones in Conolophus.^Not spinose in Amblyrhynchus, Conolophus, Iguana, Sauromalus, Dipsosaurus, and Brachylophus.^Greater than seven in Cyclura.^Some species have modes of four or six.^Middorsal scale row entirely absent.^In S. hispidus the entire leg has enlarged, spinous scales.^Tail about same length as body but not spinose.
154 University of California Publications in Zoology
<v
Phylogenetic Systematics ofIguanine Lizards 155
Node 4: Enyaliosaurus Gvd.y \M51. Lateral edges of parietal roof fail to meet, or meet and fail to form a midsagittal crest;outline of parietal roof trapezoidal or triangular (7-B).2. Scales on anterodorsal surface of shank enlarged and spinous (15-1 or-2).Node 5 (unnamed)1. Pendulous dewlap present (10-1 or-2),2. Dorsal crest scales unicolored and differing in color from adjacent body scales;large, flap-like crest scales separated by one or more smaller scales (12-1).Ctenosaura baked1. Crowns of posterior marginal teeth with a maximum of three cusps (9-1). Thischaracter occurs also in C. quinquecarinata, where it is interpreted as convergent.In some: anterior whorls of strongly spinous caudal scales always separated by at leasttwo rows of smaller scales (18-0). This is interpreted as a case of character reversal.Ctenosaura palearis1. Large pendulous dewlap present (10-2).2. All anterior whorls of strongly spinous caudal scales (except sometimes the first)separated by one or no intercalary scale rows (18-2 or-3). This character occurs also in C.defensor, C. clarki, and C. quinquecarinata, in which it is interpreted as convergent;altematively, it may be a synapomorphy of a more inclusive group (Enyaliosaurus).Node 6 (unnamed)1. Maximum snout-vent length less than 190 mm (1-1).2. Mode of 25 presacral vertebrae (2-1).3. All anterior whorls (except sometimes the first) of strongly spinous caudal scalesseparated by one or no intercalary scale rows (18-2 or 3). This character occurs also in C.palearis, where it is interpreted as convergent; altematively, it may be a synapomorphy of amore inclusive group.The occurrence of five premaxillary teeth in at least some specimens, as well asenlargement of the scales on the anterodorsal surface of the thigh, may also besynapomorphies of this group. Ctenosaura quinquecarinata1. Crowns of posterior marginal teeth with a maximum of three cusps (9-1). Thischaracter occurs also in C bakeri, in which it is interpreted as convergent.In some: prefrontal contacts jugal behind lacrimal foramen (4-B). This characteroccurs also in C. clarki, in which it is interpreted as convergent.
156 University of California Publications in Zoology
Node 7 (unnamed)1. Dorsal crest scales low-keeled to flat; inconspicuous throughout length of crest (13-1). 2. Tail strongly spinose throughout its length, and about same length as body (snout-vent/tail length = 0.48-0.56); fewer than 30 caudal vertebrae (17-1).In some: middorsal scale row broadly discontinuous in lumbosacral region (14-1),This character occurs also in C. hemilopha.Ctenosaura clarki1. Prefrontal contacts jugal behind lacrimal foramen (4-B). This character occurs alsoin some C. quinquecarinata, in which it is interpreted as convergent.Ctenosaura defensor1. Mode of five premaxillary teeth (3-1).2. Frontal portion of crista cranii projects anteriorly to form a step from frontal toprefrontal bones (5-1).3. Crowns of posterior marginal teeth with a maximum of five or more cusps (8-B).4. External signs of parietal eye inconspicuous or absent (11-1).5. Scales on anterodorsal surface of thigh enlarged and spinous (15-2).6. Anterior keels of subdigital scales at base of pedal digit III enlarged and fused attheir bases to form a comb (16-1).7. Proximal rows of smaller scales between whorls of enlarged, spinous caudal scalessmall or absent (18-3).The results of the present analysis indicate that Enyaliosaurus (including bakeri,palearis, quinquecarinata, clarki, and defensor) is a monophyletic group, but thatCtenosaura in the narrow sense (acanthura, pectinata, similis, and hemilopha) is not.Ctenosaura hemilopha appears to have shared a more recent common ancestor withEnyaliosaurus than with the other Ctenosaura (in the narrow sense). However, thecharacter that suggests an exclusive common ancestry for hemilopha and Enyaliosaurus, areduction in the number of intercalary scale rows between the whorls of enlarged, spinouscaudal scales, is problematical, in that it does not occur in all bakeri. Nevertheless, ifbakeri and palearis are sister taxa, it is simpler to interpret the incongruence as a reversal insome bakeri rather than four separate acquisitions of the derived condition (1-in hemilopha,2-in some bakeri, 3-in palearis, 4-in the common ancestor of quinquecarinata, clarki, anddefensor). In any case, the monophyly of Ctenosaura in the narrow sense is doubtful evenif this character is rejected, for there are no derived characters found in acanthura, pectinata,similis, and hemilopha that are not also found in the other taxa. Rather than the two beingseparate taxa, Enyaliosaurus appears to be a subgroup of a more inclusive Ctenosaura.There currentiy exist several problems concerning species-level taxa within Ctenosaura.Smith and Taylor (1950) considered the specimens from the west coast of Mexico assignedto C. acanthura by Bailey (1928) to be C. pectinata. Based on a conflict between the
Phylogenetic Systematics ofIguanine Lizards 157
supposed geographic ranges of the species and the actual geographic distribution ofspecimens possessing the diagnostic characters of each species, Ray and Williams(unpubl.) considered pectinata to be a synonym of acanthura. The primary character usedby Bailey (1928) to distinguish between these two taxa was whether the middorsal scalerow was continuous (pectinata) or interrupted {acanthura) in the sacral region, but Hardyand McDiarmid (1969) claim that this character is variable within pectinata from westernMexico. Nevertheless, synonymizing pectinata with acanthura on the basis of such datarests on an assumption that the two taxa are not broadly sympatric.Stejneger (1901) described Ctenosaura bakeri from Utilla (Utila) Island, Honduras, andBailey (1928) surmised that it may also occur on Bonacca (Guanaja) and Ruatan (Roatan)islands. Specimens collected subsequently on Roatan have been considered to be C. bakeri(Wilson and Hahn, 1973; Meyer and Wilson, 1973), but they differ from the Utilaspecimens in several ways (Table 10), including characters suggesting that they may noteven be one another's closest relatives. The populations from the two islands are probablybest considered separate species.Sauromalus Dumeril 1856
Type species (by monotypy): Sauromalus ater Dumeril 1856.
Etymology: (Greek) Sauros, lizard, + omalos, flat.
Definition: The most recent common ancestor of the Recent species of Sauromalus(ater, australis, hispidus, obesus, slevini, and varius) and all of its descendants.
Distribution: Deserts of the southwestern United States in southeastern California,southern Utah and Nevada, and western and central Arizona, southward into Mexico inwestern Sonora and eastern Baja California as well as various islands in the Gulf ofCalifornia (Fig. 56).
Diagnosis: Members of this taxon can be distinguished from other iguanines by thefollowing synapomorphies:1. Parietal foramen located variably within frontal (13-B). This character occurs alsoin some populations of Ctenosaura and Cyclura; the parietal foramen is invariably locatedwithin the frontal in Dipsosaurus and Cyclura carinata.2. Splenial relatively small (33-B).3. Angular does not extend far up labial surface of dentary and is not visible, or is onlybarely visible in lateral view (37-B). This character also occurs in Amblyrhynchus andConolophus; thus, it is either convergent or a synapomorphy of a more inclusive taxon.4. Angular reduced and narrow posteriorly (38-B).5. Modal number of premaxillary teeth fewer than seven (absolute range 3-7; range ofmodes for species 4-6) (43-44-A). This character also occurs in Ctenosaura defensor.
158 University of California Publications in Zoology
FIG. 56. Geographic distribution of Sauromalus (from C. E. Shaw, 1945; Gates, 1968; Etheridge,1982).
6. Crowns of posterior marginal teeth with five or more cusps (46-C). This characteroccurs in Cyclura and Iguana and may thus be a synapomorphy of a more inclusive group.It also occurs in Ctenosaura defensor, in which it is interpreted as convergent.7. Second ceratobranchials of hyoid apparatus short, often less than two-thirds thelength of the first ceratobranchials (52-53-A). This character also occurs inAmblyrhynchus and Conolophus, in which it is either convergent or a synapomorphy of amore inclusive taxon.8. Second ceratobranchials not in contact medially for most or all of their lengths (54-B). This character also occurs in Amblyrhynchus, in which it is interpreted as convergent.9. Neural spines of presacral vertebrae short, less than 50% of total vertebral height(55-B).10. Fewer than 40 caudal vertebrae (59-B). This character occurs also in Ctenosauraclarki and C. defensor, in which it is interpreted as convergent.11. Postxiphistemal inscriptional ribs never form continuous midventral chevrons (63-A). The polarity of this character is questionable. It also occurs in Dipsosaurus, in which,if derived, it is interpreted as convergent.12. Suprascapular cartilages situated primarily in a horizontal plane, and each forms anangle rather than a smooth curve with the scapula (64-B).
Phylogenetic Systematics ofIguanine Lizards 1 59
13. Scapular fenestrae small or absent (65-B). This character occurs also inAmblyrhynchus, in which it is interpreted as convergent.14. Clavicles narrow, the lateral shelf small or absent (67-B).15. Posterior process of interclavicle does not extend beyond lateral comers of sternum(68-B). This character also occurs in Amblyrhynchus, in which it is interpreted asconvergent.16. Lateral processes of interclavicle form angles of between 75? and 90? with posteriorprocess, interclavicle roughly T-shaped (69-B). This character also occurs inAmblyrhynchus, in which it is interpreted as convergent.17. Sternal fontanelle small or absent (70-B). This character occurs also inAmblyrhynchus, in which it is interpreted as convergent.18. Sternum pentagonal; xiphistema widely separated (71-B). This condition isapproached in Amblyrhynchus.19. Pelvic girdle short and broad (72-B).20. Anterior iUac process small (73-B).21. Heart extends posterior to transverse axillary plane (75-B).22. Rostral scale divided by a median suture (81-B).23. Superciliary scales quadrangular and non-overlapping (84-A). This character alsooccurs in Amblyrhynchus, in which it is interpreted as convergent.24. Enlarged anterior auricular scales (85-B).25. Middorsal scale row absent (89-B).26. Anterior and posterior keels of subdigital scales approximately equal in size;subdigital scales roughly symmetrical with respect to long axis of toe (91-B).27. Body strongly depressed (95-B).Another possible synapomorphy of Sauromalus is the failure of the lateral edges of theparietal table to meet ontogenetically. This character occurs also in Dipsosaurus and insome Brachylophus, Ctenosaura, and Cyclura.In addition, the following derived characters occur in some Sauromalus:1. Lateral process of palatine behind infraorbital foramen fails to contact jugal (23-A).This reversal occurs variably within ater, hispidus, and varius.2. Infraorbital foramen located entirely within the palatine (24-B). This characteroccurs also in Brachylophus and in some Amblyrhynchus and Ctenosaura. WithinSauromalus it is known only in obesus, and its occurrence is variable in this taxon.3. Coeliac artery originates between mesenteric arteries (78-B). This character occursalso in Iguana. Its pattern of variation is poorly known, owing to small samples.Fossil record: Estes (1983) summarized information on fossil Sauromalus, andadditional material has been described subsequendy by Norell (1986). The oldest fossilsreferred to this taxon are from the Pleistocene of California, Nevada, and Arizona. Stevens(1977) referred a fragment of a left dentary from the lower Miocene of Texas to eitherCtenosaura or Sauromalus, but thought that it was probably closer to Ctenosaura.
160 University of California Publications in Zoology
Comments: Sauromalus has a large number of derived characters supporting itsmonophyly, many of which are unambiguous (i.e., they do not occur in any otheriguanine). Ahhough several of these characters, such as the height of the neural spines, theorientation of the suprascapulae, and the shape of the pelvic girdle, may be part of a singleadaptive complex manifested externally in a depressed body form, I have treated them asseparate synapomorphies. Because various combinations of the alternative states of thesemorphologies occur in certain noniguanine taxa, there is no reason to believe that they mustalways occur together.Sauromalus shares a large number of derived characters with Amblyrhynchus,particularly in the shoulder girdle but not confined to this structure. BecauseAmblyrhynchus shares even more derived characters with Conolophus, and becauseConolophus does not possess most of the derived characters shared by Amblyrhynchusand Sauromalus, I interpret the derived characters shared by Amblyrhynchus andSauromalus as convergences. Perhaps this convergence results from similar functionaldemands placed on the shoulder girdle by the saxicolous habits of the animals in both taxa.The situation is complicated by the fact that all three taxa share two other derived characters:limited lateral exposure of the surangular (37-B), and relatively short secondceratobranchials (52-53-B). If one were to accept a sister-group relationship between theGalapagos iguanas and Sauromalus based on these characters, then the many charactersshared by Amblyrhynchus and Sauromalus, but not Conolophus, could be interpreted asadditional synapomorphies of this hypothesized clade that have reversed in Conolophus.However, because Sauromalus also shares a derived tooth morphology with Cyclura andIguana that does not occur in the Galapagos iguanas, I see no compelling reason to accept asister-group relationship between Sauromalus and the Galapagos iguanas. Furthermore,even if such a relationship were accepted, interpreting the derived characters shared byAmblyrhynchus and Sauromalus as convergent requires no more evolutionary changes thanhypothesizing a single origin for the derived state of each character, with reversal inConolophus.The most recent revision of Sauromalus is that of C. E. Shaw (1945), although worksof taxonomic significance have appeared subsequendy (Cliff, 1958; Tanner and Avery,1964; Soule and Sloan, 1966; Robinson, 1972, 1974). The boundaries, monophyly, andrelationships of the species within Sauromalus need further study.
Amblyrhynchina, new taxonType genus: Amblyrhynchus B&\\ 1825.
Etymology: Modification of Amblyrhynchus, the name of its type genus.
Definition: The most recent common ancestor of the extant Galapagos iguanas,Amblyrhynchus and Conolophus, and all of its descendants.
Phylogenetic Systematics oflguanine Lizards 161
!=;::;?;'????*.?a
1 1 1 nT III'* 1 1 1 1 J.I '
COLOMBIA
Equator ECUADOR r
PERU
FIG. 57. Geographic distribution of Amblyrhynchina (Amblyrhynchus and Conolophus).
Distribution: Islands of the Galapagos Archipelago, Ecuador (Fig. 57).
Diagnosis: Members of this taxon can be distinguished from other iguanines by thefollowing synapomorphies:1. Nasal process of premaxilla covered dorsally between nasals (5-B).2. Prefrontal contacts jugal, and lacrimal fails to contact palatine behind lacrimalforamen (7-B). This character occurs also in some Ctenosaura (clarki) and Cyclura(carinata, cornuta, and ricordii), in which it is interpreted as at least two separate instancesof convergence with the condition seen in Amblyrhynchina.3. Frontal wider than long (8-B). This character occurs also in Cyclura cornuta andIguana delicatissima, which I interpret as two separate instances of convergence.4. Lacrimal relatively small (17-B,-C).5. Dorsal surface of vomerine process of palatine bears a high medial crest (21-B).6. Labial process of coronoid relatively large (36-B,-C). This character occurs also inBrachylophus, in which it is interpreted as convergent.7. Angular does not extend up lateral surface of mandible and is barely visible in lateralview (37-B). This character also occurs in Sauromalus and is either convergent or asynapomorphy of a more inclusive taxon.
162 University of California Publications in Zoology
8. Surangular not exposed, or only barely exposed on lingual surface of mandiblebetween ventral processes of coronoid (40-B). This character occurs also in Cycluracychlura, in which it is interpreted as convergent; it also occurs as a rare variant in severalother iguanines.9. Premaxillary teeth have large lateral cusps (45-B).10. Anterior portion of pterygoid tooth patch absent (50-B). The entire pterygoid toothpatch is absent in most Conolophus and Dipsosaurus; however, when present, thepterygoid teeth of Dipsosaurus lie along the medial border of the pterygoid, those ofConolophus are located more laterally.11. Second ceratobranchials relatively short, often less than two-thirds the length ofthe first ceratobranchials (52-53-A). This character also occurs in Sauromalus and is eitherconvergent or a synapomorphy of a more inclusive taxon.12. Caudal autotomy septa absent (60-B). This character also occurs in Brachylophusand in Iguana delicatissima, in which it is interpreted as two separate instances ofconvergence.13. Dorsal head scales pointed and conical (83-B).
Fossil record: Steadman (1981) referred to Conolophus fossils of undetermined agefrom a lava tube on Isla Santa Cruz, Galapagos.Comments: A close phylogenetic relationship between Amblyrhynchus andConolophus is widely accepted (Heller, 1903; Eibl-Eibesfeldt, 1961; Avery and Tanner,1971; Thornton, 1971; Etheridge in PauU et al., 1976) but supporting evidence other thangeographic distribution has been scarce. Often the proposed close relationship betweenthese taxa was merely asserted or based on unspecified similarities. Avery and Tanner(1971) did not distinguish between ancestral versus derived characters and used a highlyartificial system for assessing similarity (see Introduction). The immunological studies ofHiggins and Rand (1974, 1975; Higgins, 1977) compared the Galapagos iguanas onlywith Iguana iguana among iguanines. Wyles and Sarich (1983) performed more extensiveimmunological comparisons, including outgroups, but they prepared antisera to only fourof the ten iguanine taxa used in their study. The morphological data presented in this studysupport the view that Amblyrhynchus and Conolophus are one another's closest livingrelatives.The relationships of Amblyrhynchina to other Iguanini are uncertain. Althoughmembers of Amblyrhynchina share two derived characters with Sauromalus-rQducQd lateralexposure of the angular (37-B) and short second ceratobranchials (52-A)-i do not considerthis convincing evidence for a close relationship between these taxa. At least one othercharacter, highly cuspate marginal teeth (46-B,C), suggests a close relationship amongSauromalus, Iguana, and Cyclura. Convergences between Amblyrhynchus andSauromalus are discussed in the comments on Sauromalus, above.
Phylogenetic Systematics ofIguanine Lizards 1 63
Amblyrhynchus Bell 1 825
Type species (by monotypy): Amblyrhynchus cristatus Bell 1825.
Etymology: (Greek) Amblys, blunt, + rhynchos, snout.
Definition: The most recent common ancestor of the populations of RecentAmblyrhynchus cristatus and all of its descendants.
Distribution: Rocky coasts of islands in the Galapagos Archipelago, Ecuador (Fig.57).
Diagnosis: Members of this taxon can be distinguished from other iguanines by thefollowing synapomorphies:1. Posterolateral processes on ventral surface of premaxilla absent (1-B).2. Anterior surface of premaxillary rostral body nearly flat (3-B).3. Nasal process of premaxilla nearly vertical (4-B).4. Nasal capsule greatly inflated; nasal bones relatively large (6-B).5. Frontal develops deep, paired pockets on ventral surface (1 1-B).6. Dorsal orbital borders wedge-shaped (12-B).7. Maxilla flares outward below row of supralabial foramina (15-B).8. Lacrimal very small (17-C).9. Large ventral process of squamosal (18-A). Because a reduced ventral process isinterpreted as a synapomorphy of Iguaninae, this is a character reversal. A similarcharacter in Iguana is interpreted as convergent.10. Anterodorsal surface of septomaxilla bears pronounced longitudinal crest (20-B).11. Parasphenoid rostrum very short (27-B).12. Stapes relatively thick (31-B).13. Dorsal edge of dentary much higher than dorsal edge of surangular on either sideof coronoid (32-B).14. Anterior inferior alveolar foramen located between coronoid and splenial; dentarydoes not contribute to its border (34-35-C).15. Angular process of prearticular remains relatively small throughout ontogeny (41-B). 16. Crowns of posterior marginal teeth tricuspid (46-A). This is a reversal, since thepresence of four or more cusps on the posterior marginal teeth is interpreted as asynapomorphy of Iguanini or possibly a more inclusive group. The presence of tricuspidposterior marginal teeth in adult Ctenosaura bakeri and C quinquecarinata is interpreted asconvergent.17. Secondary cusps of tricuspid marginal teeth relatively large, only slightly smallerthan apical cusp (47-B).
1 64 University of California Publications in Zoology
18. Second ceratobranchials of hyoid apparatus separated from one another mediallyfor most or all of their lengths (54-B). This character occurs also in Sauromalus, in whichit is interpreted as convergent.19. Scapular fenestrae small or absent (65-B). This character occurs also inSauromalus, in which it is interpreted as convergent.20. Posterior process of interclavicle does not extend beyond lateral comers of sternum(68-B). This character occurs also in Sauromalus, in which it is interpreted as convergent.21. Lateral processes of interclavicle form angles of between 75? and 90? with theposterior process; interclavicle roughly T-shaped (69-B). This character also occurs inSauromalus, in which it is interpreted as convergent.22. Sternal fontanelle small or absent (70-B). This character also occurs inSauromalus, in which it is interpreted as convergent.23. Xiphistema separated from one another medially (71-B). The xiphisterna ofSauromalus are also separated medially but to a much greater extent. I consider thissimilarity to be convergent.24. Separable skull osteoderms develop over frontal, prefrontal, and nasal bones (74-B). 25. Colic wall without valves but with numerous irregular transverse folds (79-B).26. Superciliary scales quadrangular and nonoverlapping (84-A). This character alsooccurs in Sauromalus, in which it is interpreted as convergent.27. Gular fold weakly developed (87-B).28. Digits of manus and pes partially webbed (93-B).Other possible synapomorphies of Amblyrhynchus are a laterally compressed tail(Tracy and Christian, 1985) and a high rate of tooth replacement associated with widealveolar margins of the maxilla, premaxilla, and dentary.In addition, the following derived character occurs in some Amblyrhynchus:Infraorbital foramen located entirely within palatine bone (24-B). This character occursalso in Brachylophus and in some Ctenosaura and Sauromalus.
Fossil record: None.Comments: Monophyly of Amblyrhynchus is the best- supported phylogenetichypothesis within Iguaninae. Sauromalus has almost as many characters that areinterpreted as synapomorphies, but it has more that require convergence elsewhere withiniguanines and, in this sense, are ambiguous. In terms of a simple tally of derivedcharacters used in this study, Amblyrhynchus is the most highly modified iguanine relativeto the most recent common ancestor of them all. Amblyrhynchus not only possessnumerous synapomorphies supporting its own monophyly, but also possesses those ofAmblyrhynchina and Iguanini. This high degree of morphological modification is notsurprising, given the unique natural history of these animals; Amblyrhynchus are the onlyextant lizards that gain a major part of their sustenance from the sea (Darwin, 1835; Heller,1903; Carpenter, 1966; Dawson et al., 1977; Dee Boersma, 1983). Many of the unique
Phylogenetic Systematics ofIguanine Lizards 165
morphological features of Amblyrhynchus discussed in this paper are probably related tothis unique mode of existence. For example, the modifications of the teeth and colon maybe related to the unique diet of these lizards, which consists largely of marine algae(Darwin, 1835; Carpenter, 1966; Dee Boersma, 1983). Derived characters obviouslyassociated with aquatic locomotion include the webbed digits and the strongly compressedtail. The thickened stapes may be related to differences between the sound-transmittingproperties of water and air. The inflated nasal capsule and the deep pockets that develop onthe ventral surface of the frontal house enlarged nasal salt glands, which allow marineiguanas to excrete excess salt accumulated from ingesting food with a salt concentrationsimilar to that of seawater (Schmidt-Nielsen and Fange, 1958). Convergences betweenAmblyrhynchus and Sauromalus are discussed in the comments on the latter taxon, above.Conolophus Fitzinger 1843
Type species (by original designation): Amblyrhynchus demarlii Dumeril and Bibron1837 = Amblyrhynchus subcristatus Gray 1831b.
Etymology: (Greek) Konos, cone, + lophos, crest, presumably referring to the conicalscales of the dorsal crest.
Definition: The most recent common ancestor of Conolophus pallidus and C.subcristatus and all of its descendants.
Distribution: Islands of the Galapagos Archipelago, Ecuador (Fig. 57).
Diagnosis: Members of this taxon can be distinguished from other iguanines by thefollowing synapomorphies:1. Lateral crests of premaxillary incisive process large and pierced or notched byforamina for maxillary arteries (2-B).2. Crista cranii of frontal projects anteriorly forming a step rather than a smooth curvewhere it meets medial edge of prefrontal at dorsal margin of orbitonasal fenestra (10-B).3. Supratemporals relatively small, extend one-half or less the distance across posteriortemporal fossae (14-B).4. Ectopterygoid contacts palatine near posteromedial comer of suborbital fenestra (26-B). This character occurs also in about half of the Iguana delicatissima examined, in whichit is interpreted as convergent.5. Labial process of coronoid very large, extends more than two-thirds the way downlateral surface of mandible in large specimens (36-C).6. Pterygoid teeth usually absent (51-B). This character also occurs in Dipsosaurus;however, when present, the pterygoid teeth of Dipsosaurus lie along the medial border ofthe pterygoid while those of Conolophus are situated more laterally.
166 University of California Publications in Zoology
7. Foramina in ventral surface of second sacral pleurapophyses usually absent; theirplace taken by open grooves (58-B). This character exists as a polymorphism in bothspecies of Conolophus; that is, it does not characterize all specimens.8. Subclavian arteries not covered ventrally by M. rectus capitis anterior (76-B). Thischaracter needs to be checked in additional specimens.
Fossil record: Steadman (1981) reported Conolophus fossils of undetermined age froma lava tube on Isla Santa Cruz, Galapagos.Etheridge (1964b) reported a fragmentary braincase and a body vertebra from LatePleistocene cave deposits on the West Indian island of Barbuda and estimated that bothwere from animals about 400 mm snout-vent length. He stated that the body vertebra, withits robust neural spine and well developed zygosphenes and zygantra, is similar to those oflarge iguanines. Etheridge compared the braincase with those of various large iguaninesnoting, as pointed out by Boulenger (1890), that the parabasisphenoid "is much wider thanlong and sHghtly to moderately constricted behind the [basi]pterygoid processes in Iguanaand Cyclura, about as wide as long and strongly constricted in Amblyrhynchus andConolophus, and much longer than wide and strongly constricted in Ctenosaura"(Etheridge, 1964b:68). He also gave the following length-to-width ratios for theparabasisphenoid (length measured from posterior border to apex of indentation betweenbasipterygoid processes, width measured at narrowest point posterior to basipterygoidprocesses): Iguana .40-.65, Cyclura .64-.72, Amblyrhynchus .79-.91, Conolophus .86-1.10, Ctenosaura 1.45-1.96. Because the ratio of the fossil is 1.00, Etheridge concludedthat it most closely resembles Conolophus.Based on its large size and the presence of zygosphenes and zygantra, the vertebra isreasonably interpreted as belonging to an iguanine. Based on size, both vertebra andbraincase might tentatively be referred to Iguanini, The similar proportions of theparabasisphenoid in the fossil and Conolophus, however, provide no evidence that the twoare closely related. The wide parabasisphenoids of Iguana and Cyclura, and the long oneof Ctenosaura, are derived conditions, while Amblyrhynchus, Conolophus, and the fossilretain primitive proportions of this element. Furthermore, the proportions of theparabasisphenoid in the fossil fall within the range of variation not only of Conolophus butalso of Cyclura. Etheridge's (1964b) range of .64-.72 for the length/width of theparabasisphenoid in Cyclura is based on C. cornuta, C.figginsi (=cychlura), C. ricordii,and C. macleayi (=nubila), but the range is actually much greater when other Cyclura areincluded. Pregill (1981) reported a ratio of .50 for a Puerto Rican fossil that he referred toC. pinguis, and I have obtained a range of .52 (C. pinguis) to 1.10 (C. carinata). Althoughneither Conolophus nor Cyclura occurs on Barbuda today, Cyclura occurs in the WestIndies, while Conolophus is restricted to the Galapagos Islands. Nevertheless, currentknowledge does not permit me to refer the Barbuda fossil to either of these taxa.Comments: Although not as obviously modified from the ancestral Amblyrhynchina asits sister taxon, Amblyrhynchus, Conolophus has eight derived characters not seen in
Phylogenetic Systematics ofIguanine Lizards 167
Amblyrhynchus . These characters indicate that Conolophus is monophyletic and thus,contrary to one commonly entertained hypothesis about the relationships of the Galapagosiguanas (Thornton, 1971; Higgins, 1978), cannot be considered ancestral toAmblyrhynchus.
IguaninaBell 1825
Type genus: Iguana Laurenti 1768.
Etymology: Modification of Iguana, the name of its type genus.
Definition: The most recent common ancestor of Cyclura and Iguana and all of itsdescendants.
Distribution: Lowlands of the American mainland from Sinaloa and Veracruz, Mexico,southward through Central America and northern South America to southern Brazil andParaguay as well as various Caribbean islands, including both the Greater and LesserAntilles.
Diagnosis: Members of this taxon can be distinguished from other iguanines by thefollowing synapomorphies:1 . Squamosal abuts against dorsal end of tympanic crest of quadrate ( 19-B).2. Cristae ventrolaterales of parabasisphenoid only narrowly constricted behindbasipterygoid processes (28-B,-C). This character does not occur in Cyclura carinata.3. Surangular exposed laterally well anterior to apex of coronoid and often anterior tolast dentary tooth (39-B). This character also occurs in some Ctenosaura and may be asynapomorphy of a more inclusive group.4. Crowns of posterior marginal teeth with five or more cusps (46-C,-D). Thischaracter occurs also in Sauromalus and may be a synapomorphy of a more inclusivegroup.Another possible synapomorphy of Iguanina is the development of a dewlap. Althoughthe dewlap of Cyclura is relatively small compared with that of Iguana, it is larger than thatof other iguanines QxcQpt Brachylophus and some Ctenosaura.
Fossil record: The oldest known fossil referable to Iguanina is an almost completeskull from the Pliocene of southern California (Norell, 1983). This and other fossilIguanina are discussed further in the sections on the fossil records of Iguana and Cyclura,below.Comments: The name Iguanina is first used in this work; Bell (1825) is credited withauthorship under the principle of coordination (Article 36, third edition of the InternationalCode ofZoological Nomenclature).
168 University of California Publications in Zoology
Although there are fewer characters supporting a sister-group relationship betweenCyclura and Iguana than there are for some of the other relationships proposed in thispaper, the monophyly of Iguanina is reasonably well supported. This sister group ofIguanina is not obvious from the results of the present study, but the best candidates areCtenosaura and Sauromalus (or perhaps a group composed of both these taxa). OnlySauromalus shares a derived character with Iguanina that is not variable within either ofthese taxa, increased cuspation of the posterior marginal teeth (46-B,-C). The distributionof other derived characters among taxa within Iguanini requires either that one or more ofthe basic taxa are not monophyletic or that some kind of homoplasy is involved.
Iguana Laurenti 1768
Type species (by tautonomy): Lacerta iguana Linnaeus 1758.
Etymology: (Spanish) Iguana, a modification of the name given to these animals byWest Indian natives.
Definition: The most recent common ancestor oilguana delicatissima and /. iguana andall of its descendants.
Distribution: Lowlands of the American mainland from Sinaloa and Veracruz, Mexico,southward through Central America and northern South America to southern Brazil andParaguay; in the Caribbean northward through the Lesser Antilles to the Virgin Islands(Fig. 58).
Diagnosis: Members of this taxon can be distinguished from other iguanines by thefollowing synapomorphies:1. Large ventral process of squamosal (18-A) abuts against dorsal edge of tympaniccrest of quadrate. Because the reduction of the ventral process of the squamosal is aniguanine synapomorphy, its reelaboration in Iguana is a character reversal.2. Cristae ventrolaterales of parabasisphenoid barely constricted behind basipterygoidprocesses (28-C). This character also occurs in some Cyclura and, although I haveinterpreted this as convergence, a wide parabasisphenoid may be a synapomorphy of amore inclusive group.3. Crowns of posterior marginal teeth serrate, with numerous small accessory cusps(46-D).4. Entire pterygoid tooth patch doubles ontogenetically (49-C).5. Posterolateral processes of pleurapophyses of second sacral vertebra absent (57-B).This character occurs also in Ctenosaura and in most Cyclura, and may be a synapomorphyof a more inclusive group.
Phylogenetic Systematics ofIguanine Lizards 169
^ -^\
170 University of California Publications in Zoology
2. Second ceratobranchials of hyoid apparatus much longer than first ceratobranchials(52-53-C). This character also occurs in Brachylophus, in which it is interpreted asconvergent. Within Iguana it occurs only in /. iguana and appears to be a synapomorphyof that taxon.3. Caudal autotomy septa absent (60-B). This character occurs also inAmblyrhynchina and in Brachylophus, in which it is interpreted as two separate instancesof convergence. Within Iguana it occurs only in /. delicatissima and appears to be asynapomorphy of that taxon.
Fossil record: Fossils referred to Iguana have been reported from Antigua (Wing et al.,1968), Barbados (Swinton, 1937; Ray, 1964), Martinique (Hoffstetter, 1946), andMontserrat (Steadman et al., 1984) in the West Indies; Ecuador (Hoffstetter, 1970); andsouthern California (Norell, 1983). The oldest of these are the Pliocene specimens fromsouthern California. However, since Norell (1983) considers these to be outside of theclade consisting of /. iguana and /. delicatissima, they are not Iguana according to mydefinition of this taxon, although they are its closest relatives. All other fossils referred toIguana are either Upper Pleistocene or Holocene in age.Comments: I consider the monophyly of Iguana to be reasonably well supported.Nevertheless, three of the derived characters employed in this study occur in some Iguanaas well as in Qither Brachylophus (character 52-53-C), Amblyrhynchina (8-B), or both ofthese taxa (60-B). The reason that I have interpreted these characters as convergent isacceptance of the monophyly not only of Iguana but also of Iguanina and Iguanini, basedon other characters.Within Iguana, the two currently recognized species both appear to be monophyletic;therefore, neither can be considered to be ancestral to the other. Monophyly of I. iguana issupported by the extreme width of the parabasisphenoid, the enlarged subtympanic scale(Dunn, 1934; Lazell, 1973), and the elongated second ceratobranchials. Monophyly of I.delicatissima is supported by the short frontal bone, absence of autotomy septa in thecaudal vertebrae, enlarged bony extemal nares, and possibly the failure of the septomaxillaeto reach the roof of the nasal capsule. The possibility that Iguana is a subgroup of ofCyclura is discussed in the comments on the latter taxon, below.
Cyc/wra Harlan 1824Type species (subsequent designation by Fitzinger 1843): Cyclura carinata Harlan1824.
Etymology: (Greek) Kylclos, circle, -i- oura, tail, referring to the verticils of enlarged,spinous scales on the tails of most species.
Phylogenetic Systematics ofIguanine Lizards 171
FIG. 59. Geographic distribution of Cyclura (from Schwartz and Carey, 1977).
Definition: The most recent common ancestor of Recent Cyclura (carinata, collei,comma, cychlura, nubila, pinguis, ricordii, and rileyi) and all of its descendants.
Distribution: The Bahama Islands; Cayman Islands; Mona and Anegada islands; andCuba, Hispaniola, and Jamaica, and their nearby islets (Fig. 59). Cyclura is nearly extincton Jamaica (Woodley, 1980) and has become extinct on Navassa Island in historical times(Thomas, 1966).
Diagnosis: Members of this taxon can be distinguished from other iguanines by thefollowing synapomorphies:1. Modal number of premaxillary teeth greater than seven (43-44-C). This characterapplies only to populations; its presence or absence cannot always be inferred from thecondition in a single organism.2. Presence of toe combs formed by enlargement of anterior keels of subdigital scalesand fusion of their bases (92-B). Ctenosaura defensor also possesses enlarged and fusedsubdigital keels, which are interpreted as convergent; however, in this taxon, they occuronly under the first phalanx of digit III. In Cyclura the enlarged and fused subdigital keelsoccur under the first phalanx of digit II and the first and second phalanges of digit III.In addition, the following derived characters occur only in some Cyclura:
172 University of California Publications in Zoology
1 . Prefrontal contacts jugal and lacrimal fails to contact palatine behind lacrimalforamen (7-B). This character occurs also in Amblyrhynchina and in some Ctenosaura;within Cyclura, it occurs only in C. carinata, C. cornuta, and C. ricordii.2. Frontal wider than long (8-B). This character occurs also in Amblyrhynchina andIguana delicatissima, in which it is interpreted as convergent. Within Cyclura it occursonly in C. cornuta and appears to be a synapomorphy of this taxon.3. Parietal foramen located variably or invariably within frontal bone (13-B,-C). Thischaracter occurs also in Sauromalus, Dipsosaurus, and some Ctenosaura. Within Cyclura,invariable location of the parietal foramen within the frontal is characteristic only of C.carinata.4. Cristae ventrolaterals of parabasisphenoid barely constricted behind basipterygoidprocesses (28-C). This character occurs also in Iguana. Within Cyclura, it variesconsiderably among taxa; in some (e.g., C. pinguis) the ventral surface of theparabasisphenoid is as wide or wider than that of/, delicatissima, while in others (e.g., C.carinata) it is relatively narrow, though still wider than in most iguanines other than Iguanaand other Cyclura (see section on fossil record of Conolophus, above).5. Surangular not exposed or only barely exposed below coronoid on lingual surfaceofjaw (40-B). This character occurs also in Amblyrhynchina, in which it is interpreted asconvergent. Within Cyclura, it characterizes only C. cychlura, although it occurs at amoderate frequency in C. nubila.6. Posterior portion of pterygoid tooth patch doubles ontogenetically (49-B). WithinCyclura, this character occurs in C. cornuta, C. nubila, C. pinguis, and C. ricordii. Thischaracter, or a further modification of it, occurs only in Iguana and some Ctenosaura andCyclura. Because its expression seems to depend on size, posterior doubling of thepterygoid tooth patch may be a synapomorphy of a more inclusive group, in which casefailure to double would be a synapomorphy within Cyclura.7. Posterolateral processes of pleurapophyses of second sacral vertebra absent (57-B).This character occurs also in Ctenosaura and Iguana and may be a synapomorphy of a moreinclusive group. Within Cyclura, I have found the processes only in C. pinguis.8. Snout covered by large, platelike scales (82-B). This character occurs also inIguana. Within Cyclura, it occurs in all taxa except C. carinata and C. ricordii.Considerable variation in the size of these scales exists even among those Cyclurapossessing enlarged snout scales (figures in Schwartz and Carey, 1977).9. Tail bears verticils of enlarged, spinous scales (94-B). This character occurs also inCtenosaura. The degree of caudal spinosity exhibits considerable variation within Cyclura(figures in Barbour and Noble, 1916).
Fossil record: Fossil Cyclura have been reported from the Upper Pleistocene andHolocene of Puerto Rico (Barbour, 1919; Pregill, 1981), St. Thomas in the Virgin Islands(Miller, 1918), and New Providence Island in the Bahamas (Etheridge, 1965c; Pregill,1982). The specimens from Puerto Rico have been referred to the extant species C.pinguis (Pregill, 1981). A braincase and a body vertebra from the Late Pleistocene of
Phylogenetic Systematics ofIguanine Lizards 173
Barbuda may also be remains of Cyclura (Pregill, 1981; see section on the fossil record ofConolophus, above).Comments: Cyclura is often assumed to be closely related to Ctenosaura (Barbour andNoble, 1916; Bailey, 1928; Schwartz and Carey, 1977), which it resembles in generalbody form, terrestrial habits, and the verticils of enlarged, spinous caudal scales. Thesesimilarities were noticed at least as early as Harlan (1824), who erected Cyclura for speciesthat are now placed in both Cyclura (carinata) and Ctenosaura (teres = acanthura). Despitethe resemblance between Ctenosaura and Cyclura, Cyclura probably shared a more recentcommon ancestor with Iguana than it did with Ctenosaura. The similarities betweenCyclura and Ctenosaura in general body form and terrestriality probably represent primitivefeatures retained from the common ancestor of all three taxa; and since not all Cyclurapossess the verticils of enlarged, spinous caudal scales, some form of homoplasy in tailmorphology is required no matter which relationships are accepted. Furthermore, Cycluraand Iguana share at least three derived characters not seen in Ctenosaura: abutment of thesquamosal against the dorsal end of the tympanic crest of the quadrate (19-B); a wideparabasisphenoid (28-B,-C); and highly cuspate posterior marginal teeth (46-C,-D).Although the last character occurs also in Ctenosaura defensor, my analysis of relationshipswithin Ctenosaura indicates that this is convergent.In addition to the characters suggesting a close phylogenetic relationship betweenIguana and Cyclura, there are other characters suggesting that Iguana is actually a subgroupof Cyclura, as defined here. In other words, there are characters suggesting that the mostrecent common ancestor of all Cyclura was also an ancestor of Iguana. Iguana sharesderived features of the cephalic scutellation, such as the enlarged snout scales and the rowof enlarged sublabials, as well as a derived widening of the parabasisphenoid with some,but not all, species of Cyclura. There is a particularly close resemblance between Cycluracychlura and Iguana delicatissima in these features. Nevertheless, the toe combs, theverticils of enlarged, spinous caudal scales, and the high number of premaxillary teeth arederived features seen in Cyclura but not in Iguana. The morphology of the posteriormarginal teeth also varies within Cyclura, with some approaching the highly cuspatemorphology seen in Iguana much more closely than others. However, the high degree ofvariation in this character, at least some of which is ontogenetic, along with small samplesand ambiguities caused by wear, prevent me from making any definite statement about therelationships suggested by this character.In any case, the precise relationships between Iguana and Cyclura are unclear, becausethe distributions of derived characters among taxa contradict one another. I provisionallyaccept the monophyly of Cyclura, but consider the issue to be in need of further study. Ifthe most recent common ancestor of all Cyclura was also an ancestor of Iguana, then,according to the phylogenetic definitions of taxa adopted here. Iguana is a subgroup ofCyclura rather than a separate taxon, and Iguanina is a synonym of Cyclura.

Appendix ISpecimens Examined
All specimens listed below are partial or complete skeletons unless otherwise indicated asalcoholic specimen (A) or radiograph (R). Institutional abbreviations are as follows:AMNH, American Museum of Natural HistoryASFS, Collection of Albert Schwartz (Miami-Dade Community College North, Miami,Fla.)CAS, California Academy of SciencesJMS, Collection of Jay M. Savage (University of Miami, Coral Gables, Fla.)KdQ, Collection of the author (University of California, Berkeley)KU, University of Kansas Museum of Natural HistoryLACM, Los Angeles County Museum of Natural HistoryLSUMZ, Louisiana State University, Museum of ZoologyMCZ, Museum of Comparative Zoology, Harvard UniversityMVZ, Museum of Vertebrate Zoology, University of California, BerkeleyRE, Collection of Richard Etheridge (San Diego State University, San Diego, Calif.)SDNHM, San Diego Natural History MuseumUCMP, Museum of Paleontology, University of California, BerkeleyUF, Florida State Museum, University of FloridaUSNM, United States National Museum of Natural History.IGUANINES
Amblyrhynchus cristatus: JMS 126-7, 181, 222; LACM 127324; RE 338, 386, 1041,1091, 1095, 1196, 1387, 1396, 1508, 2239; SDNHM 45156-7, 47000, 55600.Brachylophus fasciatus: AMNH 17701; CAS 54664 (A); RE 1019, 1770, 1866, 1888,2089, plus two radiographs of specimens whose institutions of deposition are unknown;MCZ 5222, 5800, 15008-9; SDNHM 55289, 55601, 55603, 60429, 62341.B. vitiensis: MCZ 158238, 160253-5.Conolophus pallidas: JMS 61, 213-8; MCZ 79772; RE 439-40, 1382, 1446-7.C. siibcristatus: AMNH 50798, 71304, 110167-8, 114493; CAS 12058 (A); MCZ 2027;MVZ 77314; RE 327; SDNHM 33682, 47007, 47140; USNM 89992, 165756.C. sp.: AMNH 14494.
175
176 Appendix I
Ctenosaura acanthura: AMNH 46483; MCZ 2176, 5013-21, 11350; SDNHM 47004,59542-3; USNM 220217-8.C. baked: ; LSUMZ 22275, 22293, 22367-71, 22399 (all A,R); UF 28530-33 (A,R;28530 also skeletonized); USNM 25324 (skull drawing from Ray and Williams, unpubl.).C. clarki: JMS 1544; MCZ 22454; MVZ 76690 (A,R), 76694 (A,R), 79256, 79293,164865-66; RE 57, 184; USNM 21450.C. defensor: KU 70261-2, 75528 (all A,R); MCZ 7095 (skull. A, and R); UF 41534(A,R).C. hemilopha: JMS 287-9, 291, X366 (R), X631-2 (R), X634-5 (R); RE 325, 491, 497-8, 502, 1087, 1341, 1386, 1887, 1964; SDNHM 48480, 48976, 55290, 57114.C. palearis: CAS 69297, 69299, 69307, 69310 (all A,R); MCZ 22390, 22399; MVZ162073-5, 162305 (all A,R).C. pectinata: JMS 238, 242, 250, 269, 692, 696, 704, 1252; RE 56, 419-21, 490, 641;SDNHM 55291.C. quinquecarinata: AMNH 77640; CAS 73554-62 (A,R); MCZ 24903; MVZ 79294,128903 (A,R).C. similis: AMNH 38949; JMS 178; MCZ 5011, 5457, 5799, 9566, 10312, 21742,22662, 25993, 26968, 27207, 36830, 139421; RE 469, 2003, 2233, 2238.Cyclura carinata: CAS 54647 (A); JMS 98; MCZ 59255, 139424; RE 1969; USNM88819.C. collei: CAS 74731 (A); MCZ 9397 (R).C. cornuta: AMNH 57878, 114487-8; JMS 221; MCZ 9974 (R); RE 383, 1226, 1837,1841-2, 1858, 1962, 1981-2, 1991.C. cychlura: AMNH 74440, 76875, 76877-8; KdQ 47-8; MCZ 6915; RE 2073; USNM64650.C. nubila: JMS 180, 182, 273; MCZ 6915; RE 228, 337, 610; SDNHM 42957, 42960.C.pinguis: ASFS V21995.C. ricordii: JMS 272, 367; RE 435.C. rileyi: MCZ 38165-69 (A); RE uncatalogued (A); UF 40744 (A), 57741.Dipsosaurus dorsalis: JMS X320, X324, X331, X334, X336, X338, X607, X612, X618(all R); RE 33, 355-9, 484, 661, 667, 1497, 1572-7, 1848, 1868, 1980; SDNHM 47006,57107-9, 59538-9, 60424.
Iguana delicatissima: KdQ 21; MCZ 6097, 10975, 16157, 60823, 75388, 83228./. iguana: JMS 244-5, 268, 713, 1028, 1545, 1553; RE 89, 158, 424, 452-4, 468, 489,1006, 1850, 1886, 2232; SDNHM 47001, 47008, 47010, 59466, 59540-1.Sauromalus ater: JMS 39; KdQ 68-9; MCZ 31521; RE 1504; SDNHM 6865.S. australis: KdQ 71, 72.
Appendix I 177
S. hispidus: JMS 172-4, 219, 239, 401, 404, 436, 915, 983; LACM 127279; RE 317,514-5, 736, 803-5, 1042, 1384, 1927, 1974; SDNHM 6873, 47028-30, 57103-4, 59471.S. obesus: RE 244, 354, 380, 408-11, 426, 461-7, 1578-9, 1852, 1864; SDNHM 48483,59534.S. slevini: MCZ 85553; RE 1340, 1367.S. varius: JMS 175-6, 246-8; RE 308, 323, 451, 512-3, 539, 1043, 1404, 1928, 2084;SDNHM 47024-6, 59542. BASILISCINES
Basiliscus basiliscus: JMS 347, 362, 1449, 1567, 1577, 1583-4; RE 555.B. plumifrons: RE 427, 2014; SDNHM 57098-100, 59467, 60430-1.B. vittatus: RE 49, 637, 1601, 1729, 1757, 1759, 2015; SDNHM 60432.Corytophanes cristatus: JMS 1701; KdQ 55; SDNHM 62345.C. hernandesii: RE 1176, 1800.Laemanctus longipes: MVZ 137673; UCMP 129880.L. serratus: AMNH 44982; RE 619.CROTAPHYTINES
Crotaphytus collaris: RE 85, 370-1, 404-7, 683, 1213-4, 1570, 1797, 1823, 1836, 1857;SDNHM 60433-5.C. insularis: SDNHM 47002.Gambelia wislizenii: RE 425, 550, 810, 1029, 1172, 1571.MORUNASAURS
Enyalioides heterolepis: AMNH 18232, 18278-9 (all R); MCZ 8063, 24959, 39977 (allR).E. laticeps: AMNH 37561 (R); MCZ 37282, 37284, 37286, 50238 (all R); RE 76, plusthree radiographs of specimens whose institutions of deposition are unknown; SDNHM47003.E. microlepis: AMNH 37562, 60608-9 (all R).E. oshaughnessyi: AMNH 28869-70, 28874-6, 28894 (all R); MCZ 29297 (R); RE 1957.E. palpebralis: AMNH 56401, 57159-61 (all R); MCZ 84035 (R).E. praestabilis: ANMH 37554-5 (R); USNM 7796-8 (R), 222583.Hoplocercus spinosus: AMNH 90658, 93807; CAS 93081-94, 93804-5, 101443-5 (allR); RE 1263, 1502.
178 Appendix I
Morunasaurus annularis: AMNH 57178, 57180-2, 57199 (all R); MCZ 146375; RE 1956;USNM 616-7, 3782 (all R).M. groi: CAS 98001 (R). OPLURINESChalarodon madagascariensis: RE 455, 457, 547.Opiums cuvieri: JMS 330; MVZ 117597 (A,R), 128904 (A,R); RE 558, 620, 1835.O. quadrimaculatus: AMNH 71452; RE 658.
In addition, I have examined various alcoholic specimens in the collections of RichardEtheridge, at San Diego State University; and the Museum of Vertebrate Zoology,University of California, Berkeley.
Appendix IIPolarity Determination Under UncertainOutgroup Relationships
I used a modified version of the outgroup method described by Maddison et al. (1984) andM. J. Donoghue (pers. comm., 1982) to assess character polarities. This method assessesthe condition of the outgroup node (branch point linking the ingroup with its sister groupon a phylogenetic tree or a cladogram) in order to minimize character- state changes at allhierarchical levels. Briefly, the cladogram for the ingroup and various outgroups isrerooted at the outgroup node, and the conditions of the various subterminal nodes on thererooted cladogram are assessed using an optimization procedure similar to that of Farris(1970). First, the terminal nodes (ends of branches) are labeled according to the conditionfound in the outgroup occupying that position. Second, the subterminal nodes are assignedcharacter states according to the following rules: (1) If both nodes above the node inquestion have the same state, assign that state to the node in question. (2) If the two nodesabove the node in question have different states, the assignment of the node in question isequivocal (?). (3) If one node above the node in question is equivocal and the other is not,assign the node in question the state of the unequivocal node. The state assigned to theoutgroup node (basal node of the rerooted cladogram) is taken as plesiomorphic.Because the relationships among iguanines and the outgroups used in this study arepoorly understood, I was forced to consider all possible cladograms for four unspecifiedoutgroups and an ingroup, of which there are nine (Fig. 60). After these cladograms arererooted at the outgroup node (Fig. 61), it can be seen that not all of them need to beconsidered further, since many will yield identical assessments of the condition at theoutgroup node. Complete equivalence is seen between some of the rerooted cladograms:A = E = G, and C = F. By swiveling branches about nodes, which does not alter therelationships implied by the diagrams, rerooted cladograms A, B, and D are found to beequivalent. Finally, given only the distribution of character states in the ingroup and thesefour outgroups, the state assigned to the outgroup node in rerooted cladograms H and Imust be identical to that of the basal node in the clade formed by the four outgroups.Therefore, for the purposes of this analysis, rerooted cladograms H and I can beconsidered to be equivalent to A and C, respectively. Only two topologies need to beconsidered further, A and C (Fig. 61).For any given character, the conditions of the outgroups can be placed on the terminalbranches of the two rerooted cladograms (Fig. 61 A, C) in all possible combinations, andthe condition of the outgroup node (i.e., the character's polarity) can be assessed. Forcases in which all four outgroups suggest a single interpretation, that interpretation isaccepted. For cases in which more than one of the states found in the ingroup also occur inone or more outgroups, the polarity of the character is ambiguous. In such cases, I have
179
180 Appendix II
FIG. 60. All nine possible fully resolved cladogram topologies for four unspecified outgroups and aningroup (inverted triangle).
made a compromise between maximizing the total number of characters on the one handand using only those characters whose polarities are completely unambiguous on the other.Table 1 1 shows polarity inferences for all possible arrangements of four outgroups onthe two rerooted cladograms (Fig. 61A,C) for seven cases of character-state distribution.This exhausts the possible character- state distributions for two state characters, since it isthe occurrence of a given state rather than its alphabetic designation that is important (e.g.,A/A/A/B = B/B/B/A). The following is a case-by-case discussion of possible polarityinferences under different relationships of the four outgroups to the ingroup.Case I (A/A/Ai/A,B): For the case in which three outgroups have one condition and theother has both altemative conditions, all arrangements except one require that the commonstate be considered plesiomorphic. The lone exception is when the variable outgroupattaches direcdy to the basal node of the rerooted cladogram (Fig. 62A). If resolution ofrelationships within this outgroup requires that state B be considered plesiomorphic for thisgroup, the polarity will be equivocal.
Appendix II 181
FIG. 61. Dendrograms corresponding with the nine cladograms in Figure 60 after each is rerooted at theoutgroup node.
Case II (A/A/A/B): If the outgroup possessing state B attaches directly to the basalnode of the cladogram (Fig. 62B), the polarity is equivocal. For all other arrangements,state A must be considered plesiomorphic.Case III (A/A/A,B/A,B): Support for the interpretation that state B is plesiomorphic isonly possible, first, if the outgroups are arranged as in Figure 62C; and second, ifresolution of the relationships within the variable outgroups requires that either state B isplesiomorphic for the outgroup attaching directly to the basal node, while the other remainsequivocal, or state B is plesiomorphic for both. Many arrangements of the outgroups willnecessitate that state A be considered plesiomorphic, and resolution of relationships withinthe variable outgroups will make many arrangements equivocal. Potential determination ofthe plesiomorphic condition for the two variable outgroups upon resolution of relationshipswithin these outgroups makes this case very ambiguous. In fact, there is only onearrangement in which it is impossible for the polarity inference to be equivocal (Fig. 62D).Case rV (A/A/A,B/B): Four arrangements of four outgroups with these conditionsyield equivocal evidence for polarity (Fig. 62E-H). In two of these arrangements (Fig.
182 Appendix II
TABLE 11. Summary of Polarity Inferences For Seven Cases of Character-stateDistribution Among Four Outgroups of Uncertain Relationships to the Ingroup
Case
I
n
III
IV
VI
vn
OutgroupCondition
A/A/A/A,B
A/A/A/B
A/A/A,B/A,B
A/A/A,B/B
A/A/B/B
A/A,B/A,B/A,B
A/A,B/A,B/B
Possible Polarity InferencesA is plesiomorphicPolarity is equivocal*A is plesiomorphicPolarity is equivocalA is plesiomorphicPolarity is equivocal*B is plesiomorphic*A is plesiomorphicPolarity is equivocalB is plesiomorphic*A is plesiomorphicPolarity is equivocalB is plesiomorphicA is plesiomorphicPolarity is equivocal*B is plesiomorphic*A is plesiomorphic*Polarity is equivocalB is plesiomorphic*
Note: An asterisk (*) indicates that the conclusion in question can only be reached uponresolution of relationships within one or more variable outgroups. See text for details.
62E,F), resolution of relationships within the variable outgroup may necessitate that state Bbe considered plesiomorphic. In all other arrangements state A must be consideredplesiomorphic, although resolution of the relationships within the variable outgroup canmake the situation equivocal.Case V {hlkfQr^): In this case, only one arrangement requires that state A beconsidered plesiomorphic (Fig. 621). One other arrangement requires that state B beconsidered plesiomorphic (Fig. 62J). For all other arrangements the polarity must beconsidered equivocal.Case VI {Klh,Blk,'Qlk,B): Because every outgroup but one is variable, the condition
Appendix 11 183
A A A A,B A A A B A A A,B A.B
Aor? A,?, or B
A,B A,B A A A A A,B B A A B A,B
?orB A,?,orB
A A,B A B A A A,B B B B A A
Aor?
A A B B
FIG. 62. Examples of polarity inferences for different arrangements of outgroup character-statedistributions. See text for discussion.
184 Appendix II
of the invariable outgroup (A) must always be considered plesiomorphic. However,resolution of relationships within one or more of the variable outgroups may necessitatethat either the alternative condition be considered plesiomorphic or that the polarity beconsidered equivocal.Case VII (A/A,B/A,B/B): In this case, polarities are always equivocal and can only bedetermined by the resolution of relationships within one or more of the variable outgroups.Of course, polarities are always equivocal in the case in which all outgroups exhibit bothstates (A,B/A,B/A,B/A,B).To summarize, In cases I and II (Table 11), either the more common state must beconsidered plesiomorphic or the situation is equivocal; the interpretation that the lesscommon state is plesiomorphic will never be favored. In the remaining five cases there willbe at least some situations in which the less common state either may or must be consideredto be plesiomorphic. Therefore, I have considered the more common state to beplesiomorphic for characters with case I and II distributions, but have withheld polaritydecisions on characters with case III, IV, V, VI, and VII distributions, using them only atlower hierarchical levels when certain ingroup taxa can serve as functional outgroups(Watrous and Wheeler, 1981).Of course, not all characters fit into the cases mentioned above. For example, there aresome characters with more than two states in the ingroup, and some in which one or morestates found in an outgroup are not comparable to any of those seen in the ingroup. In thisstudy, such cases are relatively rare and are discussed individually.
Appendix IIIPolarity Determination for Lower Level Analysis
The polarities of 19 characters could not be determined using basiliscines, crotaphytines,morunasaurs, and oplurines as outgroups. Therefore, I attempted to determine thepolarities of these characters for a less inclusive ingroup (node 3, Fig. '46), usingBrachylophus and Dipsosaurus as outgroups. The problem of determining polarities forthese characters is similar to that described in Appendix II, except that there are twooutgroups instead of four. With only two outgroups whose relationships to the ingroup areuncertain, there are only two possible cladogram topologies (Fig. 63A,B). When rerootedat the outgroup node (Fig. 63C,D), the two resulting topologies are equivalent for thepurposes of this analysis. The assessment of the condition at the outgroup node in Figure63D must be the same as that for the node linking the two outgroups, since there are nointervening nodes. Thus, both rerooted cladograms effectively have the two outgroupsattached direcdy to the basal (outgroup) node.Table 12 summarizes polarity inferences for all possible arrangements of two outgroupson the rerooted cladogram (Fig. 63C) for four cases of character-state distributions amongthe outgroups. This exhausts the possible cases for a two-state character. Case I isunambiguous: the state found in both outgroups is plesiomorphic. In case II, state A isconsidered plesiomorphic, although resolution of relationships within the variable outgroupcan render the polarity equivocal. In case III, the polarity is equivocal, but resolution ofrelationships within one or both variable outgroups may require that either state A or state Bbe considered plesiomorphic. Case IV is completely ambiguous: no polarity inference canbe made. Because Cases I and 11 are the only cases in which only one of the two states canbe considered plesiomorphic under all possible arrangements of the outgroups on thererooted cladogram (Fig. 63C), I consider polarity to be determinable only for these twocases. However, none of the characters whose polarities were undeterminable at the levelof all iguanines exhibits a Case II distribution. Thus, polarities are only determinable forcharacters with Case I distributions.
185
186 Appendix III
FIG. 63. All possible cladogram topologies for two unspecified outgroups and an ingroup (invertedtriangle), before (A and B) and after (C and D) rerooting at the outgroup node.
TABLE 12. Summary of Polarity Inferences For Four Cases of Character-stateDistribution Among Two Outgroups of Uncertain Relationships to the Ingroup
Case
I
II
III
IV
OutgroupCondition
A/AA/A,B
A,B/A,B
A/B
Possible Polarity InferencesA is plesiomorphicA is plesiomorphicPolarity is equivocal*A is plesiomorphic*Polarity is equivocalB is plesiomorphic*Polarity is equivocal
Note: An asterisk (*) indicates that the conclusion in question can only be reached uponresolution of relationships within one or both variable outgroups.
Appendix IVPolarity Reevaluation for Lower Level Analysis
Using Brachylophus and Dipsosaurus as outgroups to a subset of iguanines permitsdetermination of polarities for certain characters whose polarities were undeterminable atthe level of all iguanines. However, it also requires that the polarities of other characters bereassessed, since character polarities for the less inclusive group may not be identical tothose for the more inclusive group. The procedure used for reassessment is similar to thatused for assessing the polarities of characters whose polarities were not initiallydeterminable (Appendix III), but differs in that a more distant outgroup must be considered(Fig. 64A,B). This more distant outgroup is actually the outgroup node from the analysisof polarities for Iguaninae as a whole. It must be considered because, unlike the case forcharacters whose polarities were undeterminable at the level of all iguanines, it has beenassigned a character state. This additional branch also has the effect of rendering the tworerooted cladograms (Fig. 64C,D) nonequivalent.
FIG. 64. All possible cladogram topologies for two unspecified near outgroups, one more remoteoutgroup, and an ingroup (inverted triangle), before (A and B) and after (C and D) rerooting at the outgroupnode.
187
188 Appendix IV
TABLE 13. Summary of Polarity Inferences For Six Cases of Character-state DistributionAmong Two Near Outgroups Whose Precise Relationships to the Ingroup Are Unresolved,and One More Remote Outgroup Exhibiting a Fixed Character State
Case
I
II
III
IV
VI
OutgroupCondition
0/00/0,1
0,1
0,1/0,1
0,1/1
1/1
Possible Polarity Inferences
is plesiomorphic
is plesiomorphicPolarity is equivocal*is plesiomorphicPolarity is equivocalis plesiomorphicPolarity is equivocal*1 is plesiomorphic*is plesiomorphic*Polarity is equivocal1 is plesiomorphic*
Polarity is equivocal1 is plesiomorphic
Note: An asterisk (*) indicates that the conclusion can only be reached upon resolution ofrelationships within one or both variable outgroups. Outgroup condition refers to the nearoutgroups only; the remote outgroup is always assigned state zero.
Table 13 lists the possible polarity inferences for all possible arrangements of the twonear outgroups on the rerooted cladograms (Fig. 64C,D) for six cases of character- statedistributions among the two outgroups. This exhausts the possible character- statedistributions for a two-state character. The remote outgroup is always assigned state 0,because this was inferred to be its condition based on polarity analysis at the level of alliguanines. Unlike the case for characters whose polarities were initially undeterminable(Appendix HI), the numerical designations for the two outgroups are significant (e.g., 0/0is not equivalent to 1/1), since they may or may not be identical with that of the moredistant outgroup, which is always assigned state 0. Thus, there are six cases of character-state distributions rather than only four.For characters with Case I, II, and III distributions, I have left the polaritiesunchanged, because evidence from the new outgroups suggests that either the characterpolarity for the less inclusive ingroup is identical with that for Iguaninae as a whole or thatthe polarity is equivocal, but in no arrangement will the reverse polarity be favored. For
Appendix rv 189
characters with Case IV and V distributions, I have changed the polarity assessment toundeterminable, since the new outgroup evidence is compatible with either polarityinference. For characters with Case VI distributions, I have reversed the polarity, becausethe new outgroups suggest that either the old polarity is incorrect for the less inclusiveingroup or the situation is equivocal.

Literature Cited
Adams, E. N.1972. Consensus techniques and the comparison of taxonomic trees. Syst. Zool.21(4):390-397.Avery, D. F., and W. W. Tanner1964. The osteology and myology of the head and thorax regions of the ohesusgroup of the genus Sauromalus Dumeril (Iguanidae). Brigham YoungUniv. Sci. Bull, Biol. ser. 5(3): 1-30.1971. Evolution of the iguanine lizards (Sauria: Iguanidae) as determined byosteological and myological characters. Brigham Young Univ. Sci. Bull.,Biol. ser. 12(3): 1-79.Axtell, R. W.1958. A monographic revision of the iguanid genus Holbrookia. Ph.D.dissertation, Univ. Texas, Austin, 222 pp.Bailey, J. W.1928. A revision of the lizards of the genus Ctenosaura. Proc. U.S. Natl. Mus.73(12):l-58.Baird, S. P., and C. Girard1852. Characteristics of some new reptiles in the museum of the SmithsonianInstitution. Proc. Acad. Nat. Sci. Philadelphia 6:125-129.Barbour, T.1919. A new rock iguana from Porto Rico. Proc. Biol. Soc. Wash. 32:145-148.Barbour, T., and G. K. Noble1916. A revision of the lizards of the genus Cyclura. Bull. Mus. Comp. Zool.,Harvard, 60(4): 139- 164 + 15 plates.Bell, T.1825. On a new genus of Iguanidae. Zool. J. 2:204-207 -i- 1 plate.Bellairs, A. d'A., and S. V. Bryant1985. Autotomy and regeneration in reptiles. Pp. 301-410 in Biology of theReptilia, Vol. 15, Development B (C. Gans and F. Billett, eds.). JohnWiley, New Yoric.Boulenger, G. A.1884. Synopsis of the families of existing Lacertilia. Ann. Mag. Nat. Hist., ser.5, 24:117-122.1885. Catalogue of the lizards in the British Museum (Natural History), Vol. 2.British Museum (Nat. Hist.), London.
191
192 Literature Cited
1 890. On the distinctive cranial characters of the iguanoid lizards allied to Iguana.Ann. Mag. Nat. Hist., ser. 6, 6(35):41 2-414.Brongniart, A.1800. Essai d'une classification naturelle des reptiles. Bull. Soc. Philom., Paris,2(36):89-91 + 1 plate.1805. Essai d'une classification naturelle des reptiles. Baudouin, Imprimeur derinstitut National, Paris.Burghardt, G. M., and A. S. Rand (eds.)1982. Iguanas of the world: Their behavior, ecology, and conservation. NoyesPubl., Park Ridge, N. J., xix + 472 pp.Camp, C. L.1923. Classification of the lizards. Bull. Amer. Mus. Nat. Hist. 48(11):289-481.Carlquist, S.1974. Island biology. Columbia Univ. Press, New York, 660 pp.Carpenter, C. C.1966. The marine iguana of the Galapagos Islands, its behavior and ecology.Proc. Calif. Acad. Sci. 34(6):329-376.Carroll, R. L.1977. The origin of lizards. Pp. 359-396 in Problems in vertebrate evolution(S. M. Andrews, R. S. Miles, and A. D. Walker, eds.). Linn. Soc.Symp., ser. 4.Cliff, F. S.1958. A new species of Sauromalus from Mexico. Copeia 1958(4):259-261.Cohen, M. M., C. C. Huang, and H. F. Clark1967. The somatic chromosomes of 3 lizard species: Gekko gekko. Iguanaiguana, and Crotaphytus collaris. Experientia 23:769-771.Congdon, J. D., L. J. Vitt, and W. W. King1974. Geckos: Adaptive significance and energetics of tail autotomy. Science184:1379-1380.Cope, E. D.1886. On the species of Iguaninae. Proc. Amer. Philos. Soc. 23(122):261-271.1900. The crocodilians, lizards, and snakes of North America. Ann. Rept. U.S.Nad. Mus., 1898, part 2:151-1270 + 36 plates.Cuvier, G.1817. Le regne animal distribue d'apres son organisation, pour servir de base aI'histoire naturelle des animaux et d'introduction a I'anatomie comparee.Tome 2. Chez Deterville, Libraire, Paris.1829. Le regne animal distribue d'apres son organisation, pour servir de base aI'histoire naturelle des animaux et d'introduction a I'anatomie comparee.Tome 2. Chez Deterville, Libraire, Paris.
Literature Cited 193
1831. The animal kingdom arranged in conformity with its organization, withadditional descriptions by Edward Griffith and others, Vol. 9, The ClassReptilia. Whittaker, Treacher, and Co., London.Darwin, C.1835. The journal of a voyage in H.M.S. Beagle. Genesis Publ., Surrey,England, 1979.1859. On the origin of species by means of natural selection. John Murray,London.Daudin, F. M.1805. Histoire naturelle, generale et particuliere des reptiles. Imprimerie de F.Dufart, Paris.Dawson, W. R., G. A. Bartholomew, and A. F. Bennett1977. A reappraisal of the aquatic specializations of the Galapagos marine iguana{Amblyrhynchus cristatus). Evolution 3 1(4):89 1-897.Dee Boersma, P.1983. An ecological study of the Galapagos marine iguana. Pp. 157-176 inPatterns of evolution in Galapagos organisms (R. I. Bowman, M. Benson,and A. E. Leviton, eds.). Pacific Div. Amer. Assoc. Adv. Sci.de Queiroz, K.1982. The scleral ossicles of sceloporine iguanids: A reexamination withcomments on their phylogenetic significance. Herpetologica 38(2):302-311.1985. The ontogenetic method for determining character polarity and its relevanceto phylogenetic systematics. Syst. Zool. 34(3):280-299.1987. Systematics and the Darwinian Revolution. Phil. Sci., in press.Donoghue, M. J., and P. Cantino1984. The logic and limitations of the outgroup substitution approach to cladisticanalysis. Syst. Bot. 9(2): 192-202.Duellman, W. E.1965. Amphibians and reptiles from the Yucatan Peninsula, Mexico. Univ.Kansas Publ. Mus. Nat. Hist. 15(12):577-614.Duellman, W. E., and A. S. Duellman1959. Variation, distribution, and ecology of the iguanid lizard Enyaliosaurusclarki of Michoacan, Mexico. Occ. Pap. Mus. Zool. Univ. Michigan598:1-11.Dumeril, A.1 856. Description des reptiles nouveaux ou imparfaitement connus de la collectiondu Museum d'Histoire Naturelle et remarques sur la classification et lescaracteres des reptiles. Arch. Mus. Hist. Nat., Paris 8:56-588.Dumeril, A. M. C, and G. Bibron1837. Erpetologie generale: Ou histoire naturelle complete des reptiles. LibrairieEncyclopedique de Roret, Paris.
194 Literature Cited
Dunn, E. R.1934. Notes on /^Mana. Copeia 1934(1): 1-4.Dunson, W. A.1969. Electrolyte excretion by the salt gland of the Galapagos marine iguana.Amer. J. Physiol. 216(4):995-1002.1976. Salt glands in reptiles. Pp. 413-445 in Biology of the Reptilia, Vol. 5,Physiology (C. Gans and W. R. Dawson, eds.). Academic Press, NewYork.Earle, A. M.1962. The middle ear of the genus Uma compared to those of other sand lizards.Copeia 1962(1):185-188.Edmund, A. G.1960. Tooth replacement phenomena in the lower vertebrates. Royal Ont. Mus.,Life Sci. Div. Contr. 52:1-190.Eibl-Eibesfeldt, I.1961. Galapagos: The Noah's Ark of the Pacific. Doubleday, New York, 192pp.Eldredge, N., and J. Cracraft1980. Phylogenetic patterns and the evolutionary process. Columbia Univ. Press,New York, 349 pp.Estes, R.1963. Early Miocene salamanders and lizards from Florida. Quart. J. FloridaAcad. Sci. 26(3):234-256.1983. Sauria terrestria, Amphisbaenia. Handbuch der Palaoherpetologie, TeillOA. Gustav Fischer, Stuttgart, xii + 249 pp.Estes, R., K. de Queiroz, and J. A. Gauthier1988. Phylogenetic relationships within squamate reptiles. In Proceedings of asymposium on the phylogenetic relationships of the lizard families (R. Estesand G. Pregill, eds.), Stanford Univ. Press, in press.Estes, R., and L, I. Price1973. Iguanid lizard from the Upper Cretaceous of Brazil. Science 180:748-751.Etheridge, R.1959. The relationships of the anoles (Reptilia: Sauria: Iguanidae) an interpretationbased on skeletal morphology. Ph.D. dissertation, Univ. Michigan, 236pp.1962. Skeletal variation in the iguanid lizard Sator grandaevus. Copeia1962(3):613-619.1964a. The skeletal morphology and systematic relationships of sceloporine lizards.Copeia 1964(4):610-631.1964b. Late Pleistocene lizards from Barbuda, British West Indies. Bull. FloridaState Mus., Biol. Sci. 9:43-75.
Literature Cited 195
1965a. Fossil lizards from the Dominican Republic. Quart. J. Florida Acad. Sci.28(1):83-105.1965b. The abdominal skeleton of lizards in the family Iguanidae. Herpetologica21(3):161-168.1965c. Pleistocene lizards from New Providence. Quart. J. Florida Acad. Sci.28(l):349-358.1966. The systematic relationships of West Indian and South American lizardsreferred to the iguanid genus Leiocephalus. Copeia 1966(1):79-91.1967. Lizard caudal vertebrae. Copeia 1967(4) ;699-721.1982. Checklist of iguanine and Malagasy iguanid Uzards. Pp. 7-37 in Iguanas ofthe world: Their behavior, ecology, and conservation (G. M. Burghardtand A. S. Rand, eds.). Noyes Publ., Park Ridge, N. J.Etheridge, R., and K. de Queiroz1988. A phylogeny of Iguanidae. In Proceedings of a symposium on thephylogenetic relationships of the lizard families (R. Estes and G. Pregill,eds.), Stanford Univ. Press, in press.Farris, J. S.1970. Methods for computing Wagner trees. Syst. Zool. 19(l):83-92.1982. Outgroups and parsimony. Syst. Zool. 31(3):328-334.Fitch, H. S., and J. Hackforth-Jones1983. Ctenosaura similis. Pp. 394-396 in Costa Rican natural history (D. H.Janzen, ed.). Univ. Chicago Press.Fitzinger, L. I.1826. Nue Classification der Reptilien. Verlage von J. G. Heubner, Vienna.1843. Systema Reptilium. Braumiiller et Seidel Bibliopolas, Vindobonae.Reprinted by Soc. Study Amphib. Rept., 1973.Fries, C, C. W. Hibbard, and D. H. Dunkle1955. Early Cenozoic vertebrates in the red conglomerate at Guanajuato, Mexico.Smith. Misc. Coll. 123(7): 1-25.Gates, G. O.1968. Geographical distribution and character-analysis of the iguanid lizardSauromalus obesus in Baja California, Mexico. Herpetologica 24(4):285-288.Gauthier, J. A., R. Estes, and K. de Queiroz1988. A phylogenetic analysis of Lepidosauromorpha. In Proceedings of asymposium on the phylogenetic relationships of the lizard families (R. Estesand G. Pregill, eds.), Stanford Univ. Press, in press.Gibbons, J. R. H.1981. The biogeography of Braehylophas (Iguanidae) including the description ofa new species, B. vitiensis, from Fiji. J. Herpetol. 15(3):255-273.Gicca, D.1983. Enyaliosaurus quinquecarinatus. Cat. Amer. Amphib. Rept. 329:1-2.
196 Literature Cited
Gilmore, C. W.1928. Fossil lizards of North America. Mem. Natl. Acad. Sci. 22:1-197.Coin, C. J., O. B. Coin, and G. R. Zug1978. Introduction to herpetology. Third ed. W. H. Freeman, San Francisco,378 pp.Gorman, G. C.1973. The chromosomes of the Reptilia, a cytotaxonomic interpretation. Pp. 349-424 in Cytotaxonomy and vertebrate evolution (A. B. Chiarelli and E.Capanna, eds.). Academic Press, New York.Gorman, G. C, L. Atkins, and T. Holzinger1967. New karyotypic data on 15 genera of lizards in the Family Iguanidae, with adiscussion of taxonomic and cytological implications. Cytogenetics 6:286-299.Gorman, G. C., A. C. Wilson, and M. Nakanishi1971. A biochemical approach to the study of reptilian phylogeny: Evolution ofserum albumin and lactic dehydrogenase. Syst. Zool. 20(2): 167-1 85.Gray, J. E.1831a. A synopsis of the species of the class Reptilia. Pp. 1-110, paginatedseparately, in G. Cuvier, The animal kingdom arranged in conformity withits organization, with additional descriptions by Edward Griffith and others.Vol. 9, The Class Reptilia. Whittaker, Treacher, and Co., London.1831b. Description of a new species oi Amblyrynchus of Mr. Bell, in the BritishMuseum. Zool. Misc. 10:6. Reprinted by Soc. Study Amphib. Rept.,1971.1845. Catalogue of the specimens of lizards in the collection of the BritishMuseum. Edward Newman, London, xxviii + 289 pp.Gugg, W.1939. Der Skleralring der plagiotremen Reptilien. Zool. Jb. (Anat.) 65:339-416.Hallowell, E.1854. Descriptions of new reptiles from California. Proc. Acad. Nat. Sci.Philadelphia 7:91-97.Hamasaki, D. I.1968. Properties of the parietal eye of the green iguana. Vision Res. 8:591-599.1969. Spectral sensitivity of the parietal eye of the green iguana. Vision Res.9:515-523.Hardy, L. M., and R. W. McDiarmid1969. The amphibians and reptiles of Sinaloa, Mexico. Univ. Kansas Publ. Mus.Nat. Hist. 18(3):39-252.Harlan, R.1824. Description of two species of Linnaean Lacerta, not before described, andconstruction of the new genus Cyclura. J. Acad. Nat. Sci. Philadelphia4:242-251-^2 plates.
Literature Cited 197
Heller, E.1903. Papers from the Hopkins Stanford Galapagos expedition, 1898-1899, XIV.Reptiles. Proc. Wash. Acad. Sci. 5:39-98.Hennig, W.1966. Phylogenetic systematics. Univ. Illinois Press, Urbana, 263 pp.Hidalgo, H.1980. Enyaliosaurus quinquecarinatus (Gray) and Leptodeira nigrofasciataGunther in El Salvador. Herpetol. Rev. ll(2):41-43.Higgins, P. J.1977. Immunodiffusion comparisons of the serum albumins of marine and landiguanas from different islands in the Galapagos Archipelago. Canadian J.Zool. 55:1389-1392.1978. The Galapagos iguanas: Models of reptilian differentiation. Bioscience28(8):512-515.Higgins, P. J., and C. S. Rand1974. A comparative immunochemical study of the serum proteins of severalGalapagos iguanids. Comp. Biochem. Physiol. 49A:347-355.1975. Comparative immunology of Galapagos iguana hemoglobins. J. Exp.Zool. 193:391-397.Hoffstetter, R.1946. Faune du gisement precolombien d'Anse-Belleville (Martinique). Mem.Mus. Nation. Hist. Nat., Paris 22:1-18.1970. Vertebrados cenozoicos de Ecuador. Actas IV Congr. Latinoamer. deZool., Caracas, 2:955-970.Hoffstetter, R., and J.-P. Gasc1969. Vertebrae and ribs of modem reptiles. Pp. 201-310 in Biology of theReptilia, Vol. 1, Morphology A (C. Gans, A. d'A. Bellairs, and T. S.Parsons, eds.). Academic Press, New York.Hotton, N.1955. A survey of adaptive relationships of dentition to diet in North AmericanIguanidae. Amer. Midi. Nat. 53(1):88-114.Iverson, J. B.1980. Colic modifications in iguanine lizards. J. Morph. 163:79-93.1982. Adaptations to herbivory in iguanine lizards. Pp. 60-76 in Iguanas of theworld: Their behavior, ecology, and conservation (G. M. Burghardt and A.S. Rand, eds.). Noyes Publ., Park Ridge, N. J.Jenkins, F. A., and G. E. Goslow1983. The functional anatomy of the shoulder of the savannah monitor lizard(Varanus exanthematicus). J. Morph. 175:195-216.Kluge, A. G., and J. S. Farris1969. Quantitative phyletics and the evolution of anurans. Syst. Zool. 18(l):l-32.
198 Literature Cited
Lacepede, B. G.1788. Histoire naturelle des quadrupedes ovipares et des serpens. 2 Vols. Hotelde Thou, Paris.Langebartel, D.1953. The reptiles and amphibians. Pp. 91-108 in Faunal and archaeologicalresearches in Yucatan caves (R. T. Hatt, ed.). Cranbroolc Inst. Sci. Bull.33.Latreille, M.1825. Families naturelles du regne animal, exposees succinctement et dans unordre analytique, avec indication de leurs genres. J.-B. Bailliere, Libraire,Paris.Laurenti, J. N.1768. Specimen medicum, exhibens synopsin reptilium emendatam cumexperimentis circa venena et antidota reptilium austriacorum. Joan. Thom.Nob. de Trattnern, Vienna.Lazell, J. D.1973. The lizard genus Iguana in the Lesser Antilles. Bull. Mus. Comp. Zool.,Harvard, 145(1): 1-28.Lecuru, S.1968a. Remarques sur le scapulo-coracoide des lacertiliens. Ann. Sci. Nat. Zool.,Paris, 10:475-510.1968b. Etude des variations morphologiques du sternum, des clavicules et deI'interclavicule des lacertiliens. Ann. Sci. Nat. Zool., Paris, 10:51 1-544.Linnaeus, C.1758. Systema naturae, per regna tria naturae, secundum classes, ordines, genera,species cum characteribus, differentiis, synonymis, locis. 10th ed.Laurentii Salvii, Holmiae, 824 pp.Maddison, W. P., M. J. Donoghue, and D. R. Maddison1984. Outgroup analysis and parsimony. Syst. Zool. 33(1):83-103.Merrem, B.1820. Versuch eines Systems der Amphibien. Johann Christian Krieger,Marburg.Meyer, J. R., and L. D. Wilson1973. A distributional checklist of the turtles, crocodilians, and lizards ofHonduras. Contr. Sci. Nat. Hist. Mus. Los Angeles Co. 244:1-39.Miller, G. S.1918. Mammals and reptiles collected by Theodoor de Booy in the Virgin Islands.Proc. U.S. Natl. Mus. 54:507-511.Mitdeman, M. B.1942. A summary of the iguanid genus Urosaurus. Bull. Mus. Comp. Zool.,Harvard, 91(2): 105-1 81 -f- 16 plates.
Literature Cited 199
Montanucci, R. R.1968. Comparative dentition in four iguanid lizards. Herpetologica 24(4):305-315.Moody, S. M.1980. Phylogenetic and historical biogeographical relationships of the genera inthe family Agamidae (Reptilia: Lacertilia). Ph.D. dissertation, Univ.Michigan, xv + 373 pp.1982. Cladistic relationships within the Iguania. Paper presented at a symposiumon the phylogenetic relationships of the lizard families. Annual meeting ofthe American Society of Zoologists, Louisville, Ky.Nelson, G., and N. I. Platnick1981. Systematics and biogeography: Cladistics and vicariance. Columbia Univ.Press, New York, 567 pp.Norell, M. A.1983. Late Neogene lizards from the Anza Borrego Desert, San Diego County,California. M.S. thesis, San Diego State Univ., 199 pp.1986. Late Pleistocene lizards from Kokoweef Cave, San Bernardino County,California. Copeia 1986(l):244-246.Oelrich, T. M.1956. The anatomy of the head of Ctenosaura pectinata (Iguanidae). Misc. Publ.Mus. Zool. Univ. Michigan 94:1-122 + 59 figs.Olson, E. C.1937. A Miocene lizard firom Nebraska. Herpetologica 1(4): 111-112.Patterson, C.1982. Morphological characters and homology. Pp. 21-74 in Problems ofphylogenetic reconstruction (K. A. Joysey and A. E. Friday, eds.), Syst.assoc. spec. vol. no. 21. Academic Press, New York.Paull, D., E. E. Williams, and W. P. Hall1976. Lizard karyotypes from the Galapagos Islands: Chromosomes in phylogenyand evolution. Breviora Mus. Comp. Zool., Harvard, 441:1-31.Peters, J. A., and R. Donoso-Barros1970. Catalogue of the Neotropical Squamata, part 2, Lizards andamphisbaenians. U.S. Natl. Mus. Bull. 297:viii -i- 293 pp.Peterson, J. A.1973. Adaptation for arboreal locomotion in the shoulder region of lizards. Ph.D.dissertation, Univ. Chicago, 584 pp.1984. The microstructure of the scale surface in iguanid lizards. J. Herpetol.18(4):437-467.Pregill, G.1981. Late Pleistocene herpetofaunas from Puerto Rico. Univ. Kansas Mus. Nat.Hist. Misc. Publ. 71:1-72.
200 Literature Cited
1982. Fossil amphibians and reptiles from New Providence Island, Bahamas. Pp.8-21 in Fossil vertebrates from the Bahamas (S. L. Olson, ed.). Smith.Contr. Paleobiol. 48.Ray, C. E.1964. A small assemblage of vertebrate fossils from Spring Bay, Barbados. J.Barbados Mus. Hist. Soc. 3 1(1): 11-22.1965. Variation in the number of marginal tooth positions in three species ofiguanid lizards. Breviora Mus. Comp. Zool., Harvard, 236:1-15.Ray, C. E., and E. E. WilliamsRemarks on the taxonomy of the iguanid genus Ctenosaura and a key to itsspecies. Unpublished.Renous, S.1979. Application des principes cladistiques a la phylogenese et la biogeographiedes lacertiliens. Gegenbaurs Morph. Jahrb., Leipzig, 125:376-432.Renous-Lecuru, S.1973. Morphologic comparee du carpe chez les lepidosauriens actuels(rhynchocephales, lacertiliens, amphisbeniens). Gegenbaurs Morph.Jahrb., Leipzig, 119:727-766.Rieppel, O.1978. Streptostyly and muscle function in lizards. Experientia 34:776-777.Robinson, M. D.1972. Chromosomes, protein polymorphism and systematics of insularchuckwalla lizards (genus Sauromalus) in the Gulf of California, Mexico.Ph.D. dissertation, Univ. Arizona, 71 pp.1974. Chromosomes of the insular species of chuckwalla lizards (genusSauromalus) in the Gulf of California, Mexico. Herpetologica 30(2): 162-167.Savage, J. M.1958. The iguanid lizard genera Urosaurus and Uta, with remarks on relatedgroups. Zoologica 43(2) :4 1-54.Schmidt-Nielsen, K., and R. Fange1958. Salt glands in marine reptiles. Nature 182:783-785.Schwartz, A., and M. Carey1977. Systematics and evolution in the West Indian iguanid genus Cyclura. Stud.Fauna of Curacao and Other Caribbean Islands 53:15-97.Shaw, C. E.1945. The chuckwallas, genus Sauromalus. Trans. San Diego Soc. Nat. Hist.10(15):269-306.Shaw, G.1802. General zoology, vol. 3, part. 1, Amphibia. G. Kearsley, London, 312 pp.
Literature Cited 201
Smith, H. M.1946. Handbook of lizards: Lizards of the United States and Canada. CornellUniv. Press, Ithaca, N. Y., 557 pp.1949. Miscellaneous notes on Mexican Hzards. J. Wash. Acad. Sci. 39(l):34-43.1972. The Sonoran subspecies of the lizard Ctenosaura hemilopha. Gt. BasinNat. 32(2): 104-1 11.Smith, H. M., and E. H. Taylor1950. An annotated checklist and key to the reptiles of Mexico exclusive of thesnakes. Bull. U.S. Natl. Mus. 199:1-253.Smith, K. K.1980. Mechanical significance of streptostyly in lizards. Nature 283:778-779.Soule, M., and A. J. Sloan1966. Biogeography and distribution of the reptiles and amphibians on islands inthe Gulf of California, Mexico. Trans. San Diego Soc. Nat. Hist.14(11):137-156.Steadman, D.1981. Vertebrate fossils in lava tubes in the Galapagos Islands. Proc. Eighth Int.Congr. Speleology 549-550.Steadman, D. W., D. R. Watters, E. J. Reitz, and G. K. Pregill1984. Vertebrates from archaeological sites on Montserrat, West Indies. Ann.Carnegie Mus. 53(l):l-29.Stebbins, R. C.1948. Nasal structure in lizards with reference to olfaction and conditioning of theinspired air. Amer. J. Anat. 83(2): 183-222.1966. A field guide to western reptiles and amphibians. Houghton Mifflin,Boston, 279 pp.Stejneger, L.1901. On a new species of spiny-tailed iguana from Utilla Island, Honduras.Proc. U.S. Natl. Mus. 23(1217):467-468.Stevens, M. S.1977. Further study of Castolon local fauna (early Miocene), Big Bend NationalPark, Texas. Pearce-Sellards ser., Texas Memorial Mus., 28:1-69.Swinton, W. E.1937. Iguana remains from Barbados. Ann. Mag. Nat. Hist. 19:306-307.Tanner, W. W., and D. F. Avery1964. A new Sauromalus obesus from the Upper Colorado Basin of Utah.Herpetologica 20(l):38-42.Thomas, R.1966. A reassessmant of the herpetofauna of Navassa Island. J. Ohio Herpetol.Soc. 5(3):73-89.
202 Literature Cited
Thornton, I.1971. Darwin's islands: A natural history of the Galapagos. Natural HistoryPress, New York, 322 pp.Tracy, R. C, and K. A. Christian1985. Are marine iguana tails flattened? Brit. J. Herpetol. 6:434-435.Troyer, K.1983. The biology of iguanine lizards: Present status and future directions.Herpetologica 39(3):3 17-328.Turner, F. B., P. A. Medica, R. I. Jennrich, and B. G. Maza1982. Frequencies of broken tails among Uta stansburiana in southem Nevada anda test of the predation hypothesis. Copeia 1982(4):835-840.Underwood, G.1970. The eye. Pp. 1-97 in Biology of the Reptilia, Vol. 2, Morphology B (C.Gans and T. S. Parsons, eds.). Academic Press, New York.Van Denburgh, J.1922. The reptiles of western North America, Vol. 1, Lizards. Occ. Pap. Calif.Acad. Sci. 10:1-611.Wagler, J.1830. Natiirliches System der Amphibien, mit vorangehender Classification derSaugthiere und Vogel. J. G. Cotta'schen Buchhandlung, Munich,Stuttgart, and Tubingen.Watrous, L. E., and Q. D. Wheeler1981. The outgroup comparison method of character analysis. Syst. Zool.30(1):1-11.Wiegmann, A. F. A.1828. Beitrage zur Amphibienkunde. Isis von Oken 21:364-383.1834. Herpetologia mexicana seu descriptio amphibiorum Novae Hispaniae.C. G. Lunderitz, Berlin. Reprinted by Soc. Study Amphib. Rept., 1969.Wiley, E. O.1979. Ancestors, species, and cladograms-remarks on the symposium. Pp. 211-225 in Phylogenetic analysis and paleontology (J. Cracraft and N.Eldredge, eds.). Columbia Univ. Press, New York.1981. Phylogenetics: The theory and practice of phylogenetic systematics. JohnWiley, New York, 439 pp.Wilson, L. D., and D. E. Hahn1973. The herpetofauna of the Islas de la Bahia, Honduras. Bull. Florida StateMus., Biol. Sci. 17(2):93-150.Wing, E. S., C. A. Hoffman, Jr., and C. E. Ray1968. Vertebrate remains from Indian sites on Antigua, West Indies. Caribbean J.Sci. 8(3-4):123-139.Woodley, J. D.1980. Survival of the Jamaican iguana, Cyclura collei. J. Herpetol. 14(l):45-49.
Literature Cited 203
Wyles, J. S., and V. M. Sarich1983. Are the Galapagos iguanas older than the Galapagos? Pp. 177-186 inPatterns of evolution in Galapagos organisms (R. I. Bowman, M. Benson,and A. E. Leviton, eds.). Pacific Div. Amer. Assoc. Adv. Sci.Zug, G. R.1971. The distribution and patterns of the major arteries of the iguanids andcomments on the intergeneric relationships of iguanids (Reptilia: Lacertilia).Smith. Contr. Zool. 83:1-23.


SMITHSONIAN INSTITUTION LIBRARIES3 9088 00564 4265
Other Volumes AvailableUniversity of California Publications in Zoology
Vol. 112. Ned K. Johnson. Character Variation of Evolution of Sibling Species in theEmpidonax Difficilis-Flavescens Complex (Aves: Tyrannidae). ISBN 0-520-09799-5.Vol. 1 1 3. Thomas R. Howell. Breeding Biology of the Egyptian Plover PluvianusArgyptius (Aves: Glareolidae). ISBN 0-520-03804-5.Vol. 1 14. Blair Csuti. Type Specimens of Recent Mammals in the Museum ofVertebrate Zoology, University of California, Berkeley. ISBN 0-520-09622-3.Vol. 1 1 5. Peter B. Moyle et al. Distribution and Ecology of Stream Fishes of theSacramento-San Joaquin Drainage System, California. ISBN 0-520-09650-9.Vol. 1 1 6. Russell Greenberg. The Winter Exploitation Systems of Bay-breasted andChestnut-sided Warblers in Panama. ISBN 0-520-09670-3.Vol. 11 7. Marina Cords. Mixed-Species Association of Cercopithecus Monkeys inthe Kakamega Forest, Kenya. ISBN 0-520-09717-3.
ISBN 0-520-09730-0

zoNVINOSHllWS S3ldVdan LIBRARIES SMITHSONIAN INSTITUTION NOIiniiiSNI NVINOSHimSr- z r- ^ 2 ,-
U> \ Z W ? O) ? (/,
> SMITHSONIAN INSTITUTION NOIiniiiSNI NVINOSHill/MS S3iyvaan LIBRARIES SMITHSONIANU> 2 .v </> Z ^_?^ W 2 or.' <'>
tq y^Xc/) ? - - - - o ^^^f X yj^fc..^^/ oZ </) * Z (/> Z (/) _ ,1 NviNOSHiiws S3iyvijan libraries Smithsonian institution NouniiiSNi NviNosHims? (A :^ CO :5 u) ? cen
5^SMITHS0NIAN"^INSTITUTI0N NOIiniliSNI~'NVINOSHilWS S3 I a Va a n "^LI B RAR I ES^SMITHSONIANzooi-r)
_ CO ? c/) ? (/) ? z jjNviNosHiiws S3iyvaan libraries Smithsonian institution NoiiOiiiSNi NViNosHiiwsz 52 ^ 5 " ^ .- z ? CO z tX X^f'^: oX
^^SMITHSONIAN INSTITUTION N0liniliSNI_NVIN0SHilWs'^S3 I b Vd 3 ll^LI B RAR I ES*^SMITHSONIANen ? Oi .^ 2 \ *^ ^ ^ ^ <^
ozNVINOSHilWS S3ldVy9n LIBRARIES SMITHSONIAN INSTITUTION NOIiniiiSNI NVINOSHimS3
- _ t^ - w ? ? (/, _. .SMITHSONIAN INSTITUTION NOIiniiiSNI NVINOSHilWS S3iavyan LIBRARIES SMITHSONIAN
-.- ,r._^ ^-^^ Jz Vv ^7 ? ^^^ o 'fM^iX X \-^^,.SS^\ o WyyZM. X '^i^, oz CO *? z..^ w z CO ^z cI^NVINOSHIIWS S3 I aVdan_ LIBRARIES SMITHSON!AN_ INSTITUTION NOIiniliSNI_NVINOSHilWS
g ^^ I ^%#^ I ^^^ \^W t ^^ ? ^^#^ I ^^ !,^SMITHS0NIAN"^INST1TUTI0N NOIiniliSNl"'NVINOSHilWS S3 I a Vd 8 n"^LI B R AR I ES^SMITHSONIAN*z r^ ^_^ z ^ f- , z r- z f
CO * 3{/>
? CO ? to _ (/J ? ? oNVINOSHillNS S3iaVdan libraries SMITHSONIAN INSTITUTION NOIiniiiSNI NVINOSHillMSZ CO _ Z , CO Z ? CO z S<oa>if> J^^. ^ >*. :)Q ?Kr>
ryan libraries Smithsonian institution NoiiniiiSNi nvinoshii/js saiavhan libr/iz r- 2 r- 2 f-
I/)zUTiON NoiiniiiSNi NViNosHiiws S3iavban libraries SMITHSONIAN institution' NOlinec^ CO
^aan libraries*^smithsonian institution NoiiniiiSNi NviN0SHiiws*^s3idvaan^LiBR></> ^ ,, <^ :5 .- to
c
UTION NoiiniiiSNi'^NViNOSHiiws S3idvaan libraries^smithsonian~institution NOlin
</> ? f/> E </> =Man libraries Smithsonian institution NouniiiSNi nvinoshiiws saiavdan libr/^ ^ ^ 2 ? </> 2 to 2 .
toX
JTION'^NOIinillSNI NVIN0SH1IWS*^S3 I ? VM 8 ll^LI B RAR I ES*^SMITHSONIAN^INSTITUTlON *^N0linCO _''?(/) r:^ m ? m(/i(/>
_ 2 _j 2 -J 2 _iian LIBRARIES SMITHSONIAN INSTITUTION NOIiHiliSNI NVINOSHlllMS S3lbVyan LIBR>r- 2 r- ^^^ 2 r- z r-
^' ir\ ^- .^ V Z
^\ ^
CD3}>
(/> _ CO ? t/3 \ ? toUTION NOIinXtiSNI NVINOSHilWS SSlMViian LIBRARIES SMITHSONIAN INSTITUTION NOIiOin 2 ^ ^ 2 -.? w 2 .x< </>
^"^.J^ o /j^"" "-a^A X '!/Jf^',//^ o ^^^,^ X /i-^''-^'^^ o ^r^v X r"^
UTION NoiiniiiSNi~'NviNOSHiiyMS S3iavaan libraries smithsonian~'institution Noiin.2 r- , z r- 2 i;;;^ 2
_ vv^* to v^u.^^vtv' ^ NQ^^^iX ^ m
_ CO "= c/> ? toMan libraries Smithsonian institution NoiiniiiSNi nvinoshiiws S3ibvban _libr/s2, t/J 2? t/> 2 t?
a a n_ LI BRAR I ES^SMITHSONIAN_ INSTITUTION N0liniliSNI_NVIN0SHilWS*^S3 I b Vb B n_ LI B R;
^^ < ..?!^. 2<^ ?