ISSN 0145-9058 BULLETIN OF CARNEGIE MUSEUM OF NATURAL HISTORY ■ PHYLOGENETIC SYSTEMATICS OF CROTAPHYTID LIZARDS (REPTILIA: IGUANIA: CROTAPHYTIDAE) JIMMY A. McGUIRE NUMBER 32 PITTSBURGH, 1996 BULLETIN of CARNEGIE MUSEUM OF NATURAL HISTORY PHYLOGENETIC SYSTEMATICS OF CROTAPHYTID LIZARDS (REPTILIA: IGUANIA: CROTAPHYTIDAE) JIMMY A. McGUIRE Department of Biology, San Diego State University, San Diego, California 92182-0057 Current address: Department of Zoology and Texas Memorial Museum, The University of Texas at Austin, Austin, Texas 78712-1064. NUMBER 32 PITTSBURGH, 1996 BULLETIN OF CARNEGIE MUSEUM OF NATURAL HISTORY Number 32, pages 1-143, 52 figures Issued 25 June 1 996 James E. King, Director Editorial Staff: John L. Carter, Editor-, Bradley C. Livezey, Editor-, David R. Watters, Editor Mary Ann Schmidt, ELS, Assistant Editor Cover illustration: An adult male Crotaphytus dickersonae photographed approximately 2 km north of Bahia Kino Nuevo, Sonora, Mexico (see Fig. 3 IB). BULLETINS OF CARNEGIE MUSEUM OF NATURAL HISTORY are published at irregular intervals by Carnegie Museum of Natural History, 4400 Forbes Avenue, Pittsburgh, Pennsylvania 15213-4080, by the authority of the Board of Trustees of Carnegie Institute. © 1996 by Carnegie Institute, all rights reserved. ISSN 0145-9058 THE CARNEGIE MUSEUM OF NATURAL HISTORY Contents Abstract Introduction Historical Review Materials and Methods 6 Frequency Coding Allozyme Data Set 8 Ingroup Monophyly 8 Choice of Terminal Taxa 9 Outgroup Taxa 9 Morphology and Character Descriptions 11 Skull Roof 11 Palate 18 Braincase 20 Mandible 20 Miscellaneous Features of the Head Skeleton 24 Axial Skeleton 27 Pectoral Girdle 30 Pelvic Girdle 31 Limbs 32 Squamation 32 Pockets and Folds 36 Additional Morphological Characters 40 Coloration t 42 Behavioral Characters 52 Character List 54 Results 57 Discussion 63 Comparison with Previous Hypotheses 63 Character Evolution 65 Taxonomic Accounts 67 Crotaphytidae 67 Crotaphytus 68 C. antiquus 69 C. bid net ores 72 C. collaris 75 C. dicker sonae 80 C. grismeri 83 C. insular is 84 C. nebrius 88 C. oligocenicus't 92 C. reticulatus 92 C. vestigium 94 Gambelia 97 G. copei 98 G. corona t 102 G. silus 102 G. wislizenii 106 Key to the Species of Crotaphytus and Gambelia Ill Acknowledgments 112 Literature Cited 113 iii Appendices 120 1. Specimens Examined 120 2. Data Matrix 126 3. Outgroup Data Matrix 128 4. Step Matrices for Manhattan Distance Frequency Approach 132 5. Character Transformations for Each Stem of the Single Most Parsimonious Tree 134 6. List of Character State Changes by Character 139 7. Scleral Ossicle Data 143 IV ABSTRACT A revision of the alpha taxonomy of Crotaphytidae revealed that there are at least 12 and probably 13 species. A data set including 88 characters drawn from osteology, squamation, soft tissues, color pattern, life history, and behavior was collected. In addition, an allozyme data set composed of ten phylogenetically informative characters was obtained from the literature. Analysis of these data resulted in the following hypothesis of relationships: (( Gambelia silus (G. corona\ ( G . copei + G. wislizenii ))) + ( Cro - taphytus reticulatus (C. collaris (C. antiquus (C. nebrius (C. dick- ersonae (C. grismeri ( C . bicinctores (C. insularis + C. vestig- ium))))))))). Although little character evidence in support of crotaphytid monophyly has been presented in the literature (Eth- eridge and de Queiroz, 1988; Frost and Etheridge, 1989), cro- taphytid monophyly was found to be strongly supported with five fixed, unambiguous synapomorphies. Strong support was also discovered for monophyly of Crotaphytus (12 fixed, un- ambiguous synapomorphies) and Gambelia (six fixed, unambig- uous synapomorphies). The hypothesis of relationships estimated here was used to address life history and morphological evolution within the group including the relationship between head mor- phology and diet, the evolution of display-oriented morphology in males, the evolution of bipedal locomotion, and a functional consideration of gravid coloration. A taxonomic account is pro- vided for Crotaphytidae, Crotaphytus, Gambelia, and each spe- cies. Each species account includes a synonymy, an etymology, a diagnosis for the species, a detailed description of scalation and coloration, a section describing maximum adult size as well as size dimorphism, a description of the species geographic distri- bution with a dot distribution map, an account of the known fossil record, a summary of available natural history information, and a listing of references that provide illustrations of the species. Separate dichotomous keys are provided for males, females, and juveniles of Crotaphytus and Gambelia. 1 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 ? INTRODUCTION Lizards of the family Crotaphytidae (collared and leopard lizards) are among the most familiar squa- mates of western North America. This familiarity probably stems from their relatively large size (com- pared to other North American lizards), often vi- brant coloration, predatory lifestyle, and pugnacious habits. Crotaphytidae, one of nine iguanian families proposed by Frost and Etheridge (1989), is currently comprised of two genera, Crotaphytus (seven or eight species) and Gambelia (three species), that range from southern Idaho in the northwestern United States, southward into southern Baja California and northern Mexico, and eastward into the states of Missouri, Arkansas, and Louisiana. They have been the subject of numerous studies of ecology, physi- ology, reproduction, hybridization, and historical biogeography, and many of these studies have ad- dressed questions of a historical nature (e.g., Savage, 1960; Montanucci, 1970; Ingram andTanner, 1971; Axtell, 1972; Smith and Tanner, 1974; Montanucci etal., 1975; Tanner and Banta, 1977; Tanner, 1978; Sanborn and Loomis, 1979; Tollestrup, 1979, 1983; Murphy, 1983; Welsh, 1988). However, despite sev- eral important systematic analyses of the group (Smith and Tanner, 1972, 1974; Montanucci et ah, 1975), phylogenetic relationships within Crotaphy- tidae remain largely unresolved. Although the monophyly of the group has never been questioned, few derived characters have yet been offered to sup- port this contention (Etheridge and de Queiroz, 1988; Frost and Etheridge, 1989). The same can be said for the monophyly of the genera. The phylogenetic relationships of the group have been addressed using cladistic methodology only once (Montanucci et ah, 1975), and that study predated important meth- odological advances in cladistics, such as outgroup analysis (Watrous and Wheeler, 1981; Maddison et ah, 1984). There are three primary goals of the present study. The first goal is to revise the alpha taxonomy of Crotaphytidae in order to provide a better under- standing of species diversity within the group as well as an appropriate selection of terminal taxa for phy- logenetic analysis. The second goal is to provide an estimate of the phylogenetic relationships of Cro- taphytidae. The third goal is to use this phylogeny to investigate morphological and life history evo- lution among crotaphytids and provide a taxonomy that is logically consistent with the evolutionary his- tory of the group. HISTORICAL REVIEW The earliest accounts of crotaphytid lizards were closely associated with the joint military-scientific exploratory expeditions of the American frontier. In fact, it was only shortly after the epic Lewis and Clark expedition of 1803-1806 that the first cro- taphytid species was described. As a member of a party headed by Major Stephen H. Long that was exploring the Great Plains, Thomas Say collected and later described Agama collaris (James, 1823). Agama collaris was later placed as the sole member of the newly erected genus Crotaphytus by Holbrook (1842) in his classic account of the North American herpetofauna. A second crotaphytid species, Crotaphytus wisli- zenii, was obtained at Santa Fe (New Mexico) by Dr. Wislizenus, an army surgeon, who made the collection during the Mexican-American War of 1 846-1848. This species was first described by Baird and Girard ( 1 852 a), and a more detailed description was given by the authors (1852c) shortly thereafter in Stansbury (1852). From their first formal descrip- tions, crotaphytid lizards have been thought to form a natural group, despite the difficulty that more re- cent students have had in discovering synapomor- phies. In August of the same year, Baird and Girard (18526) described two additional species of Crota- phytus'. C. dorsalis from the desert of Colorado, and C. gambelii, for which locality data was lacking, although it was thought to have been collected in California. In December, Hallowed (1852) de- scribed a fourth species, C. fasciatus (a junior syn- onym of G. wislizenii), from the sand hills at the lower end of Jornada del Muerte, New Mexico. Hal- lowed's specimens were part of Samuel Wood- house’s collections made during the early 1850s, again emphasizing the important role that the early expeditions of the American West played in crota- phytid taxonomy. In 1854, Hallowed proposed the genus Dipsosaurus for C. dorsalis. Dumeril (1856) transferred Crotaphytus collaris to the genus Leiosaurus, a decision that was very 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 3 likely influenced by the similarity in head mor- phology and squamation in these genera. Further- more, Dumeril (1856) suggested that C. fasciatus Hallowell was synonymous with Leiosaurus fascia- tus Dumeril and Bibron 1837 (= Pristidactylus fas- ciatus fide Etheridge and Williams, 1 985). However, he provided the substitute name L. hallowellii to be used in the event that they were not found to be the same species. This taxonomy was not addressed by North American herpetologists until Cope (1900). Baird (1858) described Crotaphytus reticulatus based on specimens collected by J. H. Clark and A. C. B. Schott of the Mexican Boundary Survey. Baird designated syntypes (both labeled as USNM 2692) taken from Ringgold Barracks, Texas (Fort Ringgold Military Reservation, Starr County). In his descrip- tion, Baird (1858), without comment, erected the subgenus Gambelia for Crotaphytus wislizenii. Yarrow ( 18826) described Crotaphytus copeii from La Paz, California (Baja California Sur, Mexico), based on a specimen collected by L. Belding. Stejneger (1890) described Crotaphytus baileyi from the Painted Desert, Little Colorado River, Ar- izona. This western form was recognized on the ba- sis of two rows of interorbital scales, compared with the single row found in C. collaris, as well as smaller supraoculars, and a narrower head with a longer snout. He did not believe that C. baileyi warranted more than subspecific recognition; however, no in- tergradation zone was known at the time, and fol- lowing the rules of the American Ornithologist’s Union, he felt obligated to describe the form as a distinct species. Stejneger (1890) also described Cro- taphytus silus from the San Joaquin valley of Cali- fornia. In 1899, Mocquard described Crotaphytus fascia- tus from Cerro Las Palmas, Baja California. It is clear from his description and the accompanying figure that this is a juvenile specimen of what is now referred to as Crotaphytus vestigium, and, as the name fasciatus predates that of vestigium by 73 years, the former name has priority (see the C. vestigium taxonomic account for an assessment of the no- menclatorial implications of this taxonomy). Cope (1900) resolved several long-standing tax- onomic problems within Crotaphytus when he syn- onymized C. gambelii, C. fasciatus (Hallowell), and Leiosaurus hallowellii (= C. fasciatus ), with C. wis- lizenii. He also synonymized C. copeii and C. silus with C. wislizenii, citing an absence or gradation of distinguishing morphological features. Citing the work of Stejneger (1890), Cope did not support the recognition of Crotaphytus baileyi at either the spe- cific or subspecific rank. Over the next 50 years, there would be considerable disagreement with re- spect to the proper taxonomic ranking of baileyi, with some authors recognizing baileyi as a subspe- cies of C. collaris, others as a distinct species, and still others choosing not to recognize it at any tax- onomic level. Mocquard (1903), apparently realizing that the name Crotaphytus fasciatus had already been ap- plied to a leopard lizard species by Hallowell ( 1852), provided a substitute name (C. fasciolatus ) for the Baja California species. However, Cope (1900) had already synonymized C. fasciatus Hallowell with C. wislizenii. Thus, C. fasciatus Mocquard remained the senior synonym for the Baja California species of collared lizard. Stone and Rehn ( 1 903), noting a series of 1 1 spec- imens collected in the Pecos region of Texas that displayed the diagnostic characteristics of both C. collaris and C. baileyi, recognized the western pop- ulations as a subspecies of C. collaris, Crotaphytes (sic) collaris baileyi. Meek (1905), citing the con- stancy with which the supraorbital semicircles were unfused in the specimens he examined from Baja California, California, Arizona, and Utah, again fol- lowed Stejneger (1890) in recognizing Crotaphytus baileyi at the specific level. Over the following few years the taxonomic rank of baileyi jumped back and forth between the species and subspecies level. Ruthven (1907) followed Stone and Rehn (1903) in recognizing baileyi as a subspecies. After 1907, the taxonomy of baileyi more or less stabilized, with most workers recognizing this form as a subspecies of C. collaris. Van Denburgh and Slevin (1921) provided a brief description of Crotaphytus insularis from Isla Angel de La Guarda in the Gulf of California, Mexico. Van Denburgh (1922) could find no differences be- tween C. copeii from Islas de Cerros (= Cedros) and Magdalena and C. wislizenii, and following Cope (1900), recognized only the latter. Also, Van Den- burgh (1922) incorrectly synonymized both C. fas- ciatus Mocquard and C. fasciolatus Mocquard with C. wislizenii. In 1922, Schmidt described Crotaphytus dicker- sonae from Isla Tiburon in the Gulf of California, Mexico. In the description, he correctly hypothe- sized that the species might be found on the adjacent Sonoran mainland as well. Schmidt agreed with Van Denburgh (1922) in not recognizing C. copeii, citing extreme variation in the color pattern of this species 4 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 throughout its range. Burt (1928/?) synonymized Crotaphytus collaris baileyi with C. c. collaris on the basis of extensive variation in the interorbital scale characteristics used to separate the two forms, a taxonomy that was not followed by subsequent workers. Allen (1933) reduced Crotaphytus dicker- sonae to a subspecies of C. collaris, citing intergra- dation in the hindlimb and tail length characters that Schmidt (1922) used to distinguish C. dicker- sonae from C. c. baileyi. Allen (1933) did not follow Burt’s (1928 b) synonymy of C. c. baileyi with C. c. collaris. Mittleman (1942) discussed the higher level phy- logenetic relationships within North American ig- uanian lizards. His diagrammatic representation of relationships placed Crotaphytus as the sister taxon of Petrosaurus and Streptosaurus. This group was in turn depicted as the sister group of the phryno- somatid sand lizards Uma, Callisaurus, and Hol- brookia. Despite the relationships implied by his tree, he appears to have considered Crotaphytus to be a relatively primitive iguanid ( senso lato), be- cause he suggested that the sand lizards were derived from Crotaphytus- like stock, as was Sauromalus. Smith (1946) separated Crotaphytus mslizenii from C. collaris and C. reticulatus by placing it in the genus Gambelia, thus elevating Baird’s (1858) subgenus to generic rank. This controversial deci- sion initiated much debate among various workers on the group. Furthermore, Smith (1946) reduced G. silus to a subspecies of G. mslizenii. With respect to higher taxonomic relationships within the Igu- ania, Smith followed Mittleman (1942) in placing Crotaphytus and Gambelia as the sister group of Streptosaurus plus Petrosaurus, and this group as the sister taxon of the phrynosomatid sand lizards. Smith and Taylor (1950) elevated dickersonae from a subspecies of Crotaphytus collaris to the rank of full species. Fitch and Tanner (1951), reinterpreting the data of Burt (19286), recognized Crotaphytus collaris baileyi as a subspecies distinct from C. c. collaris. This taxonomy had generally been followed in the literature despite the earlier synonymy of the two by Burt (19286). In addition, they described a new subspecies of Crotaphytus, C. c. auriceps, from the upper Colorado River basin. Returning to the higher-level relationships within the Iguania, Savage (1958) presented a phylogeny that differed radically from that of Mittleman ( 1 942) and Smith (1946). In his classification, Savage pro- posed a new subgrouping, the iguanines, that in- cluded Crotaphytus plus those genera later placed in the Iguanidae by Frost and Etheridge (1989). Cochran (1961) recognized Crotaphytus silus as a full species. Robison and Tanner (1962) attempted to resolve the Crotaphytus-Gambelia debate by ex- amining osteological and myological evidence. As a result, they chose not to recognize Gambelia as a genus distinct from Crotaphytus. Tanner and Banta (1963), in the first of a three- part series examining the systematics of leopard liz- ards, described a new subspecies, Crotaphytus wis- lizeni punctatus, from the upper Colorado River ba- sin of Utah and Colorado. Like Cochran (1961), those authors did not recognize the genus Gambelia. Etheridge ( 1 964) removed Crotaphytus from Sav- age’s (1958) iguanines because he was unable to find any character or combination of characters that would serve to diagnose the iguanines if Crotaphytus was included. Furthermore, he hypothesized that Crotaphytus may be the sister taxon to the scelo- porines (= Phrynosomatidae) plus tropidurines (= Tropiduridae). Leviton and Banta (1964) resurrected the name copei for the Baja California populations of Crota- phytus mslizenii, recognizing C. w. copei. Weiner and Smith (1965) attempted to resolve the Gambelia-Crotaphytus controversy by exam- ining the osteology of the group. They placed all members of Crotaphytus (including those that had been placed in the genus Gambelia ) into a grouping they referred to as the “crotaphytiform” lizards. They recognized only four species of crotaphytiform liz- ards: C. collaris, C. reticulatus, C. insularis, and C. mslizeni (again relegating silus to a subspecies of C. mslizeni). Thus, without presenting evidence, Wei- ner and Smith (1965) reduced C. dickersonae to the rank of subspecies within C. collaris. Those mem- bers of the genus with a superficial resemblance to C. collaris (C. collaris, C. reticulatus, and C. insu- laris) were further separated into the “collariform” group. Finally, with respect to the Gambelia-Cro- taphytus debate, they concluded that the subgeneric rankings, Crotaphytus ( Gambelia ) wislizeni first pro- posed by Baird (1858) and Crotaphytus ( Crotaphy- tus) were the lowest levels of taxonomic segregation that could be justified by the data. Soule and Sloan ( 1 966) followed Weiner and Smith (1965) in recognizing dickersonae as a subspecies of C. collaris and reduced insularis to a subspecies of C. collaris as well. Banta and Tanner (1968), in their second study of leopard lizard systematics, provided a redescrip- 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 5 tion of Crotaphytus wislizeni copei and described a new subspecies, C. w. neseotes, from Isla de Cedros off the west coast of Mexico. They did not follow Weiner and Smith (1965) with regard to the sub- generic groupings. In response to Weiner and Smith (1965), Mon- tanucci (1969) entered the Gambelia-Crotaphytus debate. Based on an examination of the osteology of C. wislizenii, C. silus, C. collaris, and C. reticu- latus, he concluded that there should be no generic or subgeneric segregation within the group. He also recognized C. silus as a species distinct from C. wis- lizenii, based on unpublished data. Crotaphytus dickersonae and C. insularis were again recognized as full species distinct from C. collaris. In a paper published the following year, Montanucci (1970) formally elevated Crotaphytus silus from a subspe- cies of C. mslizenii to a full species based on mor- phological, ecological, and behavioral differences. Ingram and Tanner (197 1) described Crotaphytus collaris fuscus from the Chihuahuan Desert region. The subspecies could not be diagnosed by discrete morphological characters and was proposed on the basis of a distinctive discriminant function. In 1972, Holman described Crotaphytus oligo- cenicus on the basis of a right dentary from the early Oligocene Cypress Hills Formation, Saskatchewan, Canada. Smith and Tanner (1972) were the first to rec- ognize that there were additional distinct Crotaphy- tus taxa occurring primarily west of the Colorado River. They described Crotaphytus collaris bicinc- tores from the Great Basin region and C. insularis vestigium from the peninsular ranges of Baja Cali- fornia, Mexico, and southern California. Using a Ward’s Minimum Variance Cluster Analysis, they found that there were two phenotypically distinct groups of collared lizards (excluding C. reticulatus), each comprised of four named forms. The “western complex” was found to include C. z. insularis, C. i. vestigium, C. c. bicinctores, and C. c. dickersonae, while the “ collaris complex” was found to include C. c. collaris, C. c. baileyi, C. c. auriceps, and C. c. fuscus. Despite these findings, they described bi- cinctores as a subspecies of C. collaris and chose to recognize C. dickersonae as a subspecies of C. col- laris, as well. Thus, their own classification did not follow the phylogenetic relationships they had pro- posed. Axtell (1972) considered Crotaphytus collaris bi- cinctores and C. c. baileyi to be distinct at the species level based on morphological differences and a nar- row hybrid zone between the two in the Cerbat Mountains of Arizona. He tentatively placed bi- cinctores as a subspecies of C. insularis. Smith and Tanner (1974) again recognized bi- cinctores as a subspecies of Crotaphytus collaris. They based this taxonomic decision on intergrade speci- mens between C. bicinctores and C. collaris in north- western Sonora, Mexico, and southwestern Arizona, as well as the hybrid specimens identified by Axtell (1972) from the Cerbat Mountains of Arizona. However, the presumed intergrade specimens were actually C. c. nebrius, subsequently described by Axtell and Montanucci (1977), with the character- istic features of this species. They substantiated their previous recognition of C. dickersonae as a subspe- cies of C. collaris on the basis of intergrades between dickersonae and collaris from the Guaymas region. However, these specimens are again C. nebrius. Based on the results of their cluster, canonical, and discriminant function analyses, they provided two potential phylogenetic hypotheses for Crotaphytus shown here in parenthetical form: ( mslizenii { reti- culatus + (( fuscus { collaris { baileyi + auriceps))) + (( dickersonae + bicinctores) + ( insularis + vestig- ium))))) or ( mslizenii ( reticulatus + (( insularis + vestigium) + (( dickersonae + bicinctores) + {fuscus {collaris { baileyi + auriceps))))))). These hypotheses of relationship differ in that the first recognizes a group that includes C. c. dickersonae, C. c. bicinc- tores, C. i. insularis, and C. i. vestigium, while the second recognizes all of the C. collaris subspecies as a group. In addition, Smith and Tanner (1974) again recognized silus as a subspecies of C. mslizenii. Montanucci et al. (1975) made the first attempt at a cladistic analysis of the group. As a result of their electrophoretic study, they recommended the recognition of Gambelia as a valid genus, elevated Crotaphytus wislizeni silus and C. collaris dicker- sonae to full specific status, and removed C. c. bi- cinctores from C. collaris (again recognizing C. z. bicinctores). They did not recognize C. c. auriceps, considering it to be a junior synonym of C. c. baileyi. They found the character states present in C. dick- ersonae to be confounding and proposed a possible hybrid origin for the species. Their proposed phy- logeny of the group was similar to those of Smith and Tanner (1972, 1974), except that C. dickersonae was included with Smith and Tanner’s (1972) “co/- laris- complex.” The soon-to-be-described C. c. ne- brius (included as C. collaris ssp.) was also included in this complex. Their data suggested the following phylogenetic relationships: {{bicinctores {insularis + 6 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 vestigium )) + [reticulatus [dickersonae (collaris ssp. (c. collaris (c. fuscus + c. baileyi )))))). Axtell and Montanucci (1977) described the new subspecies of Crotaphytus collaris , C. c. nebrius, from the Sonoran Desert of southeastern Arizona and Sonora, Mexico. In the same year, Tanner and Banta (1977) published the third paper in their three-part study of the systematics of leopard lizards. They did not follow Montanucci et al. (1975) in recognizing Gambelia as a valid genus, or Montanucci (1970) in recognizing G. silus as a species distinct from G. wislizeni. In addition, they described a new subspe- cies, Crotaphytus wislizeni maculosus, from the La- hontan basin of western Nevada and parts of north- eastern California, southern Oregon, and the Snake River basin of southwestern Idaho. Montanucci (1978) again recognized the genus Gambelia and the species G. silus as valid taxa, while he synonymized the subspecies G. w. neseotes from Isla de Cedros with G. w. copei of the adjacent Baja California peninsula. Sanborn and Loomis (1979) elevated Crotaphytus insularis bicinctores to full specific status on the basis of distribution, squamation, and male display pat- tern differences. Wyles (1980) studied albumin immunological distances between Gambelia wislizenii and the re- maining crotaphytine species recognized by Mon- tanucci et al. (1975). Wyles (1980) concluded that the immunological distance estimates were well within the range observed for other iguanid ( sensu lato) genera and thus recommended that Gambelia again be reduced to a subgenus. Smith and Brodie (1982) erected the subfamily Crotaphytinae for Crotaphytus and Gambelia, thus providing the first higher taxonomic name for the group. Montanucci (1983), citing relative phenotypic similarity between bicinctores and vestigium and discounting the significance of the behavioral dif- ferences proposed by Sanborn and Loomis (1979), again recognized bicinctores as a subspecies of Cro- taphytus insularis. Estes (1983) synonymized Gam- belia with Crotaphytus. This taxonomic decision ev- idently passed unnoticed by most neoherpetologists and was not followed by later authors. In any event, Cooper (1984) and all later authors have referred to Gambelia as a valid taxon. Etheridge and de Queiroz (1988) were the first to provide evidence that the Crotaphytinae formed a monophyletic group, which they referred to under the informal heading “crotaphytines.” However, they were unable to find any uniquely derived char- acter states for the group and hypothesized its monophyly based on a unique combination of de- rived yet homoplastic character states. Frost and Etheridge (1989) reaffirmed the findings of Etheridge and de Queiroz (1988), although they also were unable to find any unique derived char- acters for the group. They elevated the subfamily Crotaphytinae of Smith and Brodie (1982) to fa- milial status, recognizing Crotaphytidae as one of nine monophyletic iguanian families. Norell (1989) described an extinct species of Gambelia, G. corona t, from the Pliocene-Pleisto- cene boundary of the Anza-Borrego Desert, Cali- fornia. Collins (1991), citing the evolutionary species concept of Frost and Hillis (1990), elevated C. i. vestigium (and, consequently, C. i. insularis) to full species, although no evidence was presented indi- cating morphological or genetic differentiation be- tween the two taxa. McGuire ( 1 99 1 ), in a note sum- marizing a geographic range extension, again rec- ognized vestigium (and thus insularis) as a subspe- cies of Crotaphytus insularis. MATERIALS AND METHODS The characters used in this study were obtained primarily from the skeleton, squamation, and color pattern, with additional characters taken from the hemipenes, behavior, and life history (hereafter re- ferred to as the “morphology” data set). The allo- zyme data set of Montanucci et al. (1975) also was reanalyzed. A few specimens were cleared and stained using the method of Dingerkus and Uhler (1977). Most external anatomical characters were scored from formalin-preserved specimens stored in alcohol, although some color pattern characters (noted in the character descriptions) could be ob- served only on live animals or in photographs of live individuals (field observations were made on all crotaphytid taxa and photographs taken of all crotaphytid taxa except Gambelia silus). Characters were scored primarily from adults, although some juveniles were included when ontogenetic variation 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYT1D LIZARDS 7 was not evident in the characters in question. Unless otherwise stated, scale terminology follows Smith (1946), skull terminology follows Oelrich (1956), and postcranial skeletal terminology follows Eth- eridge (1964, 1965, 1967), Hofstetter and Gasc (1969), and de Queiroz (1987). Museum numbers of crotaphytid specimens examined and their lo- calities are listed in Appendix 1 , along with museum numbers of iguanian outgroup taxa examined. Hypotheses of phylogenetic relationships were es- timated using cladistic analysis (e.g., Hennig, 1966; Wiley, 1981). Character states were polarized using outgroup analysis (Watrous and Wheeler, 1981; Maddison et al., 1984), a procedure that was com- plicated by the lack of interfamilial resolution within Iguania (see discussion of outgroup taxa below). Many characters could not be polarized unequivo- cally and these were described as “unpolarized” or “not polarized” in the character descriptions. Once character polarities were obtained, a hypothetical ancestor was constructed summarizing the hypoth- esized ancestral states for each character. The hy- pothetical ancestor was included in the analysis in order to root the tree. The phylogenetic software employed here was a test version of PAUP (version 4.0.0d26, Swofford, 1995). Because the number of taxa is relatively small, the branch-and-bound al- gorithm of Hendy and Penny (1982) was employed, guaranteeing that all most parsimonious trees would be discovered. Logical incongruencies (e.g., trans- formations of the collar pattern in species that have no collar) were coded as missing or unknown data (“?”). Following the recovery of the most parsi- monious tree, tree stability and phylogenetic infor- mation content were tested using the nonparametric bootstrap (Felsenstein, 1985; 2000 bootstrap rep- licates), as well as analyses of tree length distribution skewness (g, statistic; Hillis, 1991; Huelsenbeck, 1991; Hillis and Huelsenbeck, 1992) and the decay index (Donoghue et al., 1992). Simulations indicate that a strongly left-skewed distribution of tree lengths (described by a negative gx value) is an indicator of phylogenetic information content of the data (Hillis, 1991; Huelsenbeck, 1991; Hillis and Huelsenbeck, 1 992). Hillis and Huelsenbeck ( 1 992) provided crit- ical gi values for data matrices composed of various numbers of binary and four state characters. Because this data set differs from the simulated data sets generated by Hillis and Huelsenbeck (1992) both in number of characters and in the numbers of states per character, new gl critical values were calculated that are specific to this data set using a computer Table 1 . — Recalculated g, critical values expected for random data for the morphology-only, allozyme-only, and morphology + allo- zyme (allozymes coded using Manhattan distance frequency ap- proach) data sets and the observed g, value for each. Number of informative characters Number of taxa P = 0.05 P = 0.01 Observed Morphology only: 88 13 -0.15 -0.16 -1.49 Allozymes: 10 7 -0.43 -0.45 -0.50 Morphology + Allozymes: 98 13 -0.15 -0.15 -1.45 program written by J. Huelsenbeck (Table 1). These values were generated by randomly reshuffling char- acter states among taxa in the original data set 100 times and recalculating the gx for each reshuffled matrix. Critical values at both 95 percent and 99 percent confidence intervals were then calculated from the distribution of gx values generated. Frequency Coding The character coding scheme applied to morpho- logical data in this analysis is a frequency approach developed by Wiens (19936, 1995). An approach wherein polymorphic characters are excluded from the analysis is rejected because it is clear that many characters will be found to be polymorphic given a sufficient sample size. This was especially evident in this analysis as large sample sizes were available for both preserved (up to 87 specimens per taxon) and osteological (as many as 55 specimens per tax- on) material. Under the frequency approach, each binary character is partitioned into 25 bins (a-y), each representing 4 percent of the total range of possible frequencies that may be observed in a poly- morphic or monomorphic character (i.e., bin a = 0-3%, bin b = 4-7%, and so on; Table 2). Note that it is necessary for one of the bins to have a range of 5 percent rather than 4 percent in order to encom- pass the entire range of possible frequencies (0- 100%); this bin was arbitrarily chosen as bin y (96- 100%). Twenty-five frequency bins were used be- cause this was the maximum number of whole num- ber bins (i.e., 4 percent vs. 3.26 percent per bin, etc.) that PAUP is able to include (although PAUP will allow up to 31 bins; Swofford, 1995). Those char- acters that were analyzed using frequency coding were treated as ordered, following the assumption that any character state transformation must pass through a polymorphic state, no matter how tran- sitory, before reaching fixation (Wiens, 1 993 b, 1 995). Frequency coding was not applied to the three mul- 8 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Table 2. — Frequency values for the 25 ordered bins employed in the frequency coding analyses (Wiens, 1995). a = 0-3.99 b = 4-7.99 c = 8-11.99 d = 12-15.99 e = 16-19.99 f = 20-23.99 g = 24-27.99 h = 28-31.99 i = 32-35.99 j = 36-39.99 k = 40-43.99 1 = 44-47.99 m = 48-51.99 n = 52-55.99 o - 56-59.99 P = 60-63.99 q = 64-67.99 r = 68-71.99 s = 72-75.99 t = 76-79.99 u = 80-83.99 v = 84-87.99 w = 88-91.99 X = 92-95.99 y = 96-100 tistate characters that also showed intraspecific polymorphism (characters 75, 84, and 85) because the raw frequency data were not obtained for these characters. For these three characters, the polymor- phic OTUs were assigned more than one character state and PAUP’s “interpret multiple states as un- certainty” option was invoked. Two additional mul- tistate characters were included (characters 28 and 68), but in these cases each terminal taxon was fixed for a particular character state. The frequency cod- ing approach is unnecessary with respect to these characters (or fixed binary characters) because fre- quency coding only behaves differently from stan- dard binary coding when at least one OTU exhibits more than one character state. For example, if taxa A, B, and C are fixed for the ancestral state and taxa D, E, and F are fixed for the derived state, then under frequency coding A, B, and C will be assigned state “a” (0-3.99%) and D, E, and F will be assigned state “y” (96-100%). The ordered transformation from “a” to “y” takes one step, the same number of steps that would be assigned to this transforma- tion using standard binary coding. As a result, a clade composed of taxa D, E, and F would be re- covered and it would be supported by a single com- plete character state transformation (= one step). All six multistate characters were treated as unor- dered because no a priori information was available that would suggest a particular sequence through which these character states most likely evolved. Allozyme Data Set An allozyme data set taken from Montanucci et al. (1975) was incorporated into this analysis. These data were analyzed using a modified version of the Mabee and Humphries (1993) coding approach. Step matrices were again used, but frequency information was incorporated using Manhattan distances (Wiens, 1995). This approach allowed polymorphic allo- zyme data to be analyzed in a manner analogous to the frequency coding approach used for the mor- phology data. Alternatives to the Manhattan dis- tance frequency approach employed in the analyses of the allozyme data include the use of polymorphic coding (terminology taken from Wiens, 1995), wherein the locus is the character and the allele is the character state, and the step matrix approach recommended by Mabee and Humphries (1993). Wiens (1995) found that the Manhattan distance frequency approach performed better than either of these alternatives (plus a number of additional al- ternative approaches as well). Nevertheless, com- bined analyses were also undertaken in which the allozyme data were analyzed using polymorphic coding and the Mabee and Humphries (1993) ap- proaches. The allozyme data were analyzed sepa- rately in order to test for phylogenetic signal (using the bootstrap and skewness statistic). These data were then analyzed together with the morphological data generated in this study in order to determine whether together they could provide additional res- olution or modify the topology produced by the morphological data alone. In the combined analy- ses, the multistate morphological characters were assigned a weight of 1 00 in order that they be weight- ed equally with the allozyme characters (because the Manhattan distance approach effectively weights characters 100 times more strongly than standard binary characters). For the same reason, the fre- quency bin characters were assigned weights of four because the frequency bin approach effectively weights characters by 24. Therefore, all of the char- acters were given approximately equal weight. Ingroup Monophyly The monophyly of crotaphytid lizards has never been questioned and, as Etheridge and de Queiroz (1988) pointed out, the most persistent taxonomic 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 9 debate concerning crotaphytids has been whether or not Gambelia should be synonymized with Crota- phytus (Smith, 1946; Robison and Tanner, 1962; Weiner and Smith, 1965; Montanucci, 1969, 1978; Montanucci et al., 1975; Tanner and Banta, 1977). Nevertheless, very little character evidence has been presented supporting the monophyly of Crotaphy- tidae. Etheridge and de Queiroz (1988) recognized crotaphytids as a monophyletic group on the basis of a unique combination of derived, yet highly ho- moplastic features: the presence of posterior cora- coid fenestrae and female gravid coloration, and the absence of postfrontal bones and a middorsal scale row. Frost and Etheridge (1989) considered crota- phytids to be monophyletic on the basis of three reversals: presence of palatine teeth, posterior cor- acoid fenestrae, and ribs on the third cervical ver- tebra (the last of which is only infrequently observed in crotaphytids). In each of these analyses, character support for Crotaphytidae was dependent upon its placement within the ingroup topology. The follow- ing is a list of synapomorphies of Crotaphytidae recognized in this study: presence of black oral pig- mentation (reversed within Crotaphytus ), presence of a posterolaterally projecting jugal-ectopterygoid tubercle immediately posterior to the maxillary tooth row, presence of posterior coracoid fenestrae, the tympanic crest of the retroarticular process of the mandible curves posterodorsally, the parietal and frontal strongly overlap the medial process of the postorbital, the supratemporal lies in a groove along the ventral or ventrolateral border of the supratem- poral process of the parietal (reversed in most G. silus or convergent in Crotaphytus and other Gam- belia), presence of palatine teeth, and contact of the prefrontal and jugal in the anterolateral border of the orbit. Choice of Terminal Tax a The terminal taxa utilized in this study include the currently recognized species of Crotaphytus (C. antiquus, C. bicinctores, C. collaris, C. dickersonae, C. grismeri, C. insularis, C. reticulatus, and C. ves- tigium) and Gambelia (G. corona f, G. silus, and G. wislizenii). Over the course of this study, it was de- termined that at least one and probably two addi- tional species should be recognized. These include two taxa currently recognized as subspecies, C. c. nebrius and G. w. copei (see taxonomic accounts for data supporting the elevation of these taxa to full species). These species were also included in the analysis. Another population of Gambelia that may even- tually prove to be a full species is the population of G. wislizenii on Isla Tiburon in the Gulf of Califor- nia. The four osteological specimens examined in this study lacked autotomic fracture planes in the caudal vertebrae. Fracture planes are present in all other G. wislizenii (n = 19) and G. silus (n = 5) specimens examined, although they appeared to be fused in three of ten G. copei. Unfortunately, no osteological specimens were available from adjacent Sonora and it could not be determined if the absence of fracture planes is confined to this insular popu- lation. If this population proves to be a separate species, it may be the only endemic reptile or am- phibian on Isla Tiburon, a land-bridge island that supports an extensive herpetofauna. The remaining subspecies of Crotaphytus collaris and Gambelia wislizenii were not treated as separate terminal taxa because no evidence has been pre- sented, nor has any been discovered over the course of this investigation, suggesting that these forms are discrete evolutionary entities. Rather, they are pat- tern or convenience classes (Frost et al., 1992), color morphs largely consistent over an extensive area, but grading smoothly into other color morphs at their boundaries. Outgroup Taxa Etheridge and de Queiroz (1988) and Frost and Etheridge (1989) provided evidence for the mono- phyly of nine suprageneric groups (elevated to fam- ilies in the latter study) within Iguania. Interfamilial resolution was elusive and their strict consensus tree (at the familial level) was an unresolved polytomy. However, they were able to substantially reduce the number of equally parsimonious interfamilial to- pologies as depicted in their 1 2 unrooted trees with rooting points (Fig. 1). Thus, despite the continuing lack of unambiguous interfamilial resolution, the outgroup situation has improved considerably. In this analysis, characters were considered to be po- larized only when the polarity assessment was con- sistent with all 12 unrooted trees. For each of the eight remaining iguanian families, exemplars were examined for the purpose of char- acter polarization. The choice of exemplars was based whenever possible on the results of recent intrafamilial phylogenetic analyses. Thus, basal lin- eages have been proposed for clades within the fam- ilies Phrynosomatidae (Presch, 1969; Montanucci, 1987; Etheridge and de Queiroz, 1988; de Queiroz, 1989, 1992; Wiens, 1993a, 19937>), Tropiduridae 10 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Ho Co Ch Po — Ig Cr Ph Tr 1 1 1 Op Po Co Ho — Ch B — Ig Cr Ph Tr I I I op Po Ho Ch Co D — Ig —Co Cr Ph Tr Ho -Ch Ig Cr Ph Tr 1 — A — J — L — 1 — Op Po Ho — Ch Co Ig — A-t E Cr Ph Tr I I I Op Po Ho F — Ip Co Ch Cr Ph Tr Jp I I I I I Op Po Ho Ch G Ch — Ig ^ — Co — Cr Ph Tr 1 1 Ho i Co i — Ig Cr Ph Tr Op III 1 JLJ L Co H Op Po Ig Ho — Ch Cr Ph Tr — L — LJ 1 1 1 Op Ho Tr Po Ch Co i — Ig Cr — Ph Op -J — ) 1 l_ Po Ch Ho K Ho Co — Ig Cr Ph Po Op Ch Co — !g Cr Ph Tr J-J r mi i - Tr Po — 1 (j) — ( 1 1 1 1 Op Fig. 1. — The 12 unrooted trees discovered by Frost and Etheridge (1989) in their phylogenetic analysis of iguanian lizards. The open circles represent the discovered rooting points for these unrooted trees. Ch = Chamaeleonidae, Co = Corytophanidae, Cr = Crotaphytidae, Ho = Hoplocercidae, Ig = Iguanidae, Op = Opluridae, Ph = Phrynosomatidae, Po = Polychrotidae, Tr = Tropiduridae. (Etheridge and de Queiroz, 1988; Frost and Ether- idge, 1989; Frost, 1992; Pregill, 1992; Etheridge, 1995), Corytophanidae (Etheridge and de Queiroz, 1988; Lang, 1989), Hoplocercidae (Etheridge and de Queiroz, 1988), Iguanidae (de Queiroz, 1987; Norell and de Queiroz, 1991), Polychrotidae (Guyer and Savage, 1986, 1992; Etheridge and de Queiroz, 1988; Cannatella and de Queiroz, 1989), and Cha- maeleonidae (Moody, 1980, 1987; Klaver, 1981; Klaver and Bohme, 1986; Hillenius, 1986, 1988; Rieppel, 1987; Frost and Etheridge, 1989). For the remaining family (Opluridae), only the phenetic analysis of Blanc et al. (1983) was available. For this lineage, I examined Chalaradon and as many species of Opiums as possible. A list of outgroup taxa ex- amined for this study is provided in Appendix 1 and a data matrix documenting the character states observed in these taxa is provided in Appendix 3. 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 1 1 MORPHOLOGY AND CHARACTER DESCRIPTIONS Skull Roof Premaxilla (Characters 1, 2; Fig. 2-5, 7). — The posterodorsally projecting nasal process is long and very slender in Gambelia wislizenii and G. copei (Fig. 4, 5) and broad in most Crotaphytus (Fig. 2, 7) and the single specimen of G. corona f. Gambelia situs (Fig. 3) occasionally has a slender but short nasal process (seven of 30) owing to its truncated snout. In C. insularis, the nasal process is also long and extremely narrow, which may be a consequence of elongation of the snout region. Some variation occurs in C. vestigium and C. bicinctores, both with two of 28 specimens having similarly slender nasal processes, and C. grismeri, with one of five having a slender nasal process, although not as extreme as that seen in C. insularis. Among the outgroup taxa, a narrow nasal process was observed only in Petro- Fig. 2. — Dorsal view of the skull of Crotaphytus dickersonae (REE 2777, adult male, SVL =116 mm). Scale = 5 mm. saurus mearnsi and occasional Uta stansburiana, Dipsosaurus dorsalis, Phymaturus pal/uma, and Ph. punae (although the condition observed was not as extreme as that observed in Gambelia and C. in- sularis). Therefore, an elongate, narrow nasal pro- cess is considered to be the derived state. In Gambelia, the anteromedial portion of the al- veolar shelf at the articulation of the premaxilla and vomers is in the form of a strong vertical ridge. This ridge is rarely present in Crotaphytus (three of 51 C. collaris, one of four C. antiquus). Among the outgroup taxa, a strong vertical ridge was observed only in Corytophanes hernandezi, Microlophus grayi, two of three Leiocephalus schreibersi, and one of three Phymaturus patagonicus zapalensis. There- fore, a strong vertical ridge at the alveolar shelf is considered to be the derived state. The premaxillary base is also subject to much variation in crotaphytids. In all Gambelia, plus many C. antiquus, C. collaris (primarily those formerly Fig. 3. — Dorsal view of the skull of Gambelia silus (CAS 227 1 3, adult male, SVL =101 mm). Scale = 10 mm. 12 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Fig. 4. — Dorsal view of the skull of Gambelia wislizenii (REE 2918, adult female, SVL =119 mm). Scale = 10 mm. referred to the subspecies C. c. auriceps and C. c. baileyi ), C. grismeri, C. nebrius, and C. reticulatus, the broad, laterally oriented maxillary processes give the base a rectangular shape as opposed to a trap- ezoidal shape (Fig. 2-4). This condition is either absent or appears rarely in C. bicinctores, C. dick- ersonae, C. insularis, and C. vestigium. Despite this trend, most Crotaphytus species display continuous variation in this feature with all intermediates be- tween the rectangular and nonrectangular condi- tions present. Therefore, this character was not in- cluded in the phylogenetic study. Nasals (Character 3; Fig. 2-4, 7). -In Crotaphytus dickersonae (Fig. 2), two of four C. antiquus, and one of 28 C. bicinctores, forward expansion of the nasals results in their overlap of the nasal process Fig. 5. — Anterior portion of the skull of Gambelia wislizenii { REE 2918, adult female, SVL =119 mm) depicting the saddle-shaped premaxillary-maxillary articulations. The premaxilla is vertically hatched. Max = maxilla, Nas = nasal. Scale = 5 mm. of the premaxilla well anterior to the posterior bor- der of the external nares (fenestrae exonarina of Oel- rich, 1956). This feature varies ontogenetically in C. dickersonae, with individuals of less than 81 mm snout-vent length (SVL) having incomplete contact of the nasals anteriorly (character scored only from adults). The nasals occasionally overlap the nasal process of the premaxilla anterior to the posterior extent of the external nares in Gambelia wislizenii and G. copei. However, this appears to be the result of posterior expansion of the nares rather than an anterior expansion of the nasals and is here consid- ered to be nonhomologous. The nasals only rarely overlap the nasal process of the premaxilla anterior to the posterior extent of the external nares in the outgroup taxa. This condition was observed in Broo- kesia stumpffi and in a number of tropidurid taxa ( Ctenoblepharys adspersus, some Phymaturus pa- tagonicus patagonicus, P. p. payuniae, and P. p. so- muncurensis [but not other Phymaturus ], many Lei- ocephalus species, and Microlophus grayi ). Pregill (1992) considered this feature to be absent from most basal extant Leiocephalus, including L. her- minieri, L. greenwayi, L. punctatus, L. inaguae, L. psammodromus, and some L. carinatus. Therefore, the conditions observed in Liolaeminae, Leioce- phalinae, and at least one member of Tropidurinae may be convergent. Nevertheless, nasals that over- lap the nasal process of the premaxilla may be an- cestral for Liolaeminae and Leiocephalinae, and, 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYT1D LIZARDS 13 Fig. 7. — Dorsal view of the anterior portion of the skull of Cro- taphytus grismeri (MZFC 665 1, adult male, SVL = 92 mm). The vertical hatching denotes the extra frontonasal bone present in two of five specimens examined. Fro = frontal. Max = maxilla, Nas = nasal, Prf = prefrontal, Prm = premaxilla. Scale = 5 mm. Fig. 6.— Anterior portion of the orbit showing contact of the prefrontal and jugal bones ( Crotaphytus dickersonae, adult male, REE 2777 , SVL =116 mm). Jug = jugal. Lac = lacrimal, Max = maxilla, Pal = palatine, Prf = prefrontal. Scale = 2 mm. therefore, for the entire Tropiduridae. Because B. stumpffi and U. acanthinurus are the only nontro- pidurid iguanian taxa examined here in which the nasals overlap the nasal process of the premaxilla, it is most parsimonious to code extensive overlap of the nasal process by the nasals as the derived state. Prefrontals (Character 4; Fig. 2-4, 6, 7). — In all crotaphytids, the palatine process of the prefrontal broadly contacts the jugal just posterior to the lac- rimal foramen (de Queiroz, 1987; Fig. 6). This con- dition was observed in Phrynosoma asio, Uma exsul, U. inornata, U. notata, U. scoparia, some Phyma- turus patagonicus payuniae, one of three Leioce- phalus macropus, Microlophus grayi, one of three Uranoscodon supercihosus, some Pristidactylus tor- quatus, Polychrus acutirostris, and some Po. mar- moratus. In Phrynosoma asio and Uma (as well as other sand lizards), this contact appears to be as- sociated with loss of the lacrimal bone, which usu- ally separates the prefrontal from the jugal in other iguanians. The contact of the prefrontal and jugal is considered to be the derived state and, thus, rep- resents a synapomorphy for Crotaphytidae. Although Norell (1989) stated that crotaphytids can be diagnosed by the derived loss of the pre- frontals, clearly (as he stated elsewhere in the paper), he was referring to the loss of the postfrontals. Frontal (Character 5; Fig. 2-4, 7, 8).— A separate, median frontonasal bone (Fig. 7) is present in two of five Crotaphytus grismeri (MZFC 6650, 6651). Although the sample size is small for this taxon, its presence in two specimens suggests that it is a poly- morphism rather than an aberration. A similar bone was observed only in one Phymaturus palluma (REE 2313). Although this feature sheds no light on the phylogenetic relationships within Crotaphytus, it appears to represent an additional autapomorphy for the species. The skulls of Gambelia wislizenii, G. copei, and G. corona f are more depressed than those of Cro- taphytus and G. silus. Although this variation ap- pears to be associated with several bones, it is per- haps best illustrated by comparing the shape and orientation of the frontal bone. In Crotaphytus and G. silus, this bone is more strongly convex, while in 14 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 G. wislizenii, G. copei, and G. corona t, this bone usually is relatively flat, providing little height to the midorbital region of the skull. A description of the frontal bone only partially explains the complex variation in skull height within crotaphytids. There- fore, the character is here defined as “skull de- pressed” or “skull vaulted.” Although a vaulted mi- dorbital region of the skull is the more common condition within Iguania, this character could not be polarized. Norell (1989) described the fossil taxon Gambelia corona t based in part on a broad frontal that is transversely concave with supraorbital ridges. Many Gambelia have broad frontals; however, the dorsal surface is usually flat. Only one of 53 G. wislizenii (REE 2792) had weakly developed supraorbital ridges with a slight concavity and no G. copei were examined with this condition. Gambelia silus also usually lack the supraorbital ridges; however, three of 3 1 had well-developed supraorbital ridges with strong transverse concavity. Although this condi- tion cannot be considered unique to G. corona f, it appears to be a useful diagnostic feature for the spe- cies. Additional fossil material will be required in order to determine if this character is variable as in other Gambelia. The frontal may bear supraorbital ridges that give it a concave appearance in some Crotaphytus, although, as in Gambelia, it is only infrequently present. Among the outgroup taxa, a transversely concave frontal was observed in Eny- alioides laticeps, Basiliscus basiliscus, B. plumifrons, B. vittatus, Corytophanes hernandezi, some Cory- tophanes cristatus, some Laemanctus longipes, some Phymaturus palluma, some Leiocephalus carinatus, Uranoscodon superciliosus, Uromastyx hardwickii, Physignathus lesueurii, Hydrosaurus amboiensis, Brookesia kersteni, Enyalius perditus, Polychrus marmoratus, and P. acutirostris. Therefore, this character could not be polarized. Norell (1989) considered the frontoparietal suture anterior to the posterior extent of the orbit to be an additional autapomorphy of Gambelia corona f. Al- though it is possible that this condition is an artifact resulting from damage to the fossil (dorsoventral compression), it does appear as though the fronto- parietal suture was indeed anterior to the posterior extent of the orbits. The postorbitals project more posteriorly in G. corona t than in other crotaphytids, which may play some role in the anterior placement of the suture. Although this character is not phy- logenetically informative, it provides a diagnostic autapomorphy for the species. In articulated skulls of some iguanians, the suture that binds the frontal with the nasals and prefrontals takes the form of a “ W.” However, this shape results from the extensive overlap of the frontal by the nasals and prefrontals. The underlying anterior bor- der of the frontal is often squared off or may possess two small lateral processes that project anteriorly. In all crotaphytids, the anterior border of the frontal bears three well-developed processes, two lateral projections and one medial projection, that extend forward equidistantly. This condition occurs spo- radically within Iguania and could not be polarized. Therefore, this feature was not considered in the phylogenetic analysis. Postfrontals. — The postfrontals are small bones that form part of the posterior border of the orbits in many iguanian species, but are absent or fused in all crotaphytids. Postfrontals are absent or oc- casionally present as minute elements in Phryno- soma and the phrynosomatid sand lizards, some Phymaturus ( Phymaturus palluma, some Phyma- turus punae ), oplurids, Polychrus (contra Frost and Etheridge, 1989; verified in P. acutirostris and P. marmoratus ), Basiliscus, Corytophanes, and Cha- maeleonidae. Although the absence or fusion of the postfrontal bones may eventually prove to be a syn- apomorphy for Crotaphytidae, the currently unre- solved nature of iguanian phylogeny prevents po- larization of this character. Postorbitals (Characters 6, 7; Fig. 2-4, 8). — The dorsal process of the postorbital is roughly triangular in cross section in all crotaphytids. The dorsomedial aspect is completely overlapped by the frontal and parietal while the ventral portion is exposed. Thus, in an articulated skull it appears as though a long process projects medially beneath the overlying frontal and parietal. This condition appears to be more extreme in Crotaphytus because the parietal and frontal overlap the postorbital more laterally in these lizards. However, the condition of the post- orbital does not vary significantly between Crota- phytus and Gambelia. In the outgroup taxa, the fron- tal and parietal usually meet the dorsomedial por- tion of the postorbital without overlapping it exten- sively; the only obvious exceptions are hoplocercids, corytophanids, one of two Uromastyx acanthinurus, basal agamines ( Physignathus and Hydrosaurus am- boiensis), and Enyalius iheringi. A strong degree of overlap at this joint, which appears to provide ad- ditional structural support, is tentatively recognized as a synapomorphy of Crotaphytidae. The angle of the dorsal process often differs be- 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 15 tween Crotaphytus and Gambelia. In Crotaphytus, the process is transversely oriented, while in Gam- belia it is often anteromedially oriented. In many cases this difference is very obvious. However, con- tinuous variation within Gambelia prevented the inclusion of this character in the phylogenetic anal- ysis. The postorbital meets the jugal and squamosal in a tongue-in-groove articulation. In crotaphytids, the postorbital bears the shallow groove in which the jugal and squamosal lie. This condition is more de- veloped in Gambelia, which bears a large flare that broadly overlaps the jugal and squamosal on the medial side of the joint. This feature is difficult to evaluate in the outgroups due to the paucity of dis- articulated skulls. However, it appears that this con- dition is widespread within Iguania and it was not included in the phylogenetic analysis. Finally, in Gambelia copei (eight of eight), rela- tively few G. wislizenii (four of 49; REE 425, 2792; UIMNH 43378-79), and four of 31 G. silus (KU 121753, 121761, 121766, 121768), there is a small projection or tubercle on the anterolateral surface of the postorbital at the posterior edge of the orbit (= character 7). In G. copei, it is usually larger and more robust than in other Gambelia. This small tubercle may function as an additional attachment point for the skin of the head as does the larger dorsal tubercle. The presence of this tubercle appears to be unique within Iguania and may be a synapomorphy for Gambelia, although its more developed state may be further derived in G. copei. Nevertheless, this feature is coded as a binary character with the absence of a tubercle coded as the ancestral condi- tion (state 0) and the presence of a tubercle as the derived condition (state 1). Because they were poly- morphic with respect to this character, G. wislizenii and G. silus were assigned states c and d respectively. Parietal (Characters 8, 9; Fig. 2-4, 8). — The pa- rietal is a median bone that represents the major element of the skull roof. Its complex shape includes a trapezoidal roof with short anterolateral processes and long posterolaterally projecting, laterally com- pressed supratemporal processes. This shape changes ontogenetically, although not to the extent seen in some iguanids, polychrotids, and Leiocephalus (Eth- eridge, 1959; de Queiroz, 1987; Pregill, 1992). In juveniles, the parietal roof is roughly square, the crests of the supratemporal processes are less robust and project nearly directly posteriorly. During on- togeny, the posterior edge of the parietal roof be- comes increasingly constricted such that the lateral borders of the roof converge. This gives the roof a trapezoidal shape with the supratemporal processes projecting posterolaterally rather than posteriorly. Late in ontogeny, ridges may form along the lateral and posterior borders of the parietal roof giving the central portion a depressed appearance. The degree of constriction of the posterior border of the parietal roof during ontogeny differs between Crotaphytus (Fig. 2) and Gambelia (Fig. 3, 4). In Gambelia, the roof remains relatively broad posteriorly through- out ontogeny and remains approximately twice the width of the narrowest portion of the frontal bone. In Crotaphytus (particularly males) the posterior border of the parietal shelf becomes more constrict- ed such that it is approximately equal in width to the frontal bone or slightly wider. This constriction is often most dramatic in adult male C. dickersonae, although enough overlap occurs between species of Crotaphytus that this was not considered as a sep- arate character state. There is much variation in the degree of constriction of the parietal roof within Iguania, with the basal lineages of all but three fam- ilies (Phrynosomatidae, not constricted; Coryto- phanidae, constricted; Hoplocercidae, constricted) having representatives with both states. Although the polarity of the character could not be deter- mined, Gambelia and Crotaphytus always differ in the degree of constriction of the parietal roof. There- fore, this feature was coded as an unpolarized binary character with the Gambelia condition coded as state 0 and the Crotaphytus condition coded as state 1 . The supratemporal processes are extremely ro- bust in Crotaphytus and, in lateral view, project well above the temporal arches (Fig. 8). The lateral faces of the processes are also concave. The robust char- acter of the processes gives broad surface area for the origin of the hypertrophied jaw adductor mus- cles that these lizards possess. In all Crotaphytus examined except some eastern C. collaris (13 of 51 specimens), the supratemporal processes are strong- ly inflected ventrad at their distal ends. The skulls of some eastern C. collaris tend to be more dorso- ventrally compressed, which may result in less in- flected supratemporal processes. Gambelia also pos- sess ventrally oriented processes, although of a dif- ferent character. The crests of the supratemporal processes are well developed anteriorly, but quickly taper posteriorly, usually terminating anterior to the articulation of the process with the squamosal. By contrast, in Crotaphytus, the crests of the supratem- poral processes continue posteriorly well beyond the squamosal to its terminus. As a result, in Gambelia, 16 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Fig. 8. — Lateral view of the skull of Crotaphytus dickersonae (REE 2777, adult male, SVL =116 mm). Scale = 5 mm. the processes appear less robust and do not arch as far above the plane of the parietal roof. This vari- ation, which can be used to quickly differentiate between skulls of these genera, could not be polar- ized due to variation in the outgroups. Supratemporals (Character 10; Fig. 9, 10).— The supratemporals are small paired bones that lie in ventrolateral grooves in the supratemporal process- es of the parietal in most crotaphytids (Etheridge and de Queiroz, 1988; Frost and Etheridge, 1989). The supratemporals are more exposed posterola- terally and form the major portion of the process at Fig. 9. — Supratemporal region of Crotaphytus vestigium (REE 2935, adult male, SVL = 125 mm). The vertical hatching denotes the exposed portion of the supratemporal bone. Jug = jugal. Par = parietal, Pte = pterygoid, Pto = postorbital. Qua = quadrate, Squ = squamosal. Scale = 5 mm. Fig. 10. — Supratemporal region of Gambelia silus (CAS 22713, adult male, SVL =101 mm). The vertical hatching denotes the exposed portion of the supratemporal bone. Jug = jugal. Par = parietal, Pte = pterygoid, Pto = postorbital. Qua = quadrate, Squ = squamosal. Scale = 5 mm. its articulation with the quadrate and squamosal. The tongue-in-groove articulation of each supra- temporal with the parietal is well developed in all crotaphytids except Gambelia silus (Fig. 9, 10). In- deed, in most G. silus that could be coded for this character (25 of 28), the supratemporal does not sit in a groove, but lies along the lateral surface of the supratemporal process (Fig. 10). This variation is occasionally observed in G. wislizenii (four of 49), C. antiquus (one of four), and C. collaris (one of 5 1). In iguanian lizards, the tongue-in-groove relation- ship between the supratemporal and supratemporal process is seen only in crotaphytids and the tropi- durid genus Liolaemus and therefore is here con- sidered to be derived within Crotaphytidae. The condition observed in G. silus may be a reversal because some individuals do possess the rare grooved condition seen in few iguanian lizards. Septomaxillae (Character 11; Fig. 2-5, 7). — The septomaxillae are paired sheets of bone situated in the anteromedial nasal capsule where they form the floor of the nasal passages and the roof over the Jacobson’s organ (Oelrich, 1956; Jollie, 1960). In Gambelia wislizenii and G. copei, the septomaxillae are slender and more elongate than in either G. silus or Crotaphytus. It is likely that this condition is 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 17 associated with the more elongate snout seen in this species. This hypothesis is corroborated by the rel- atively slender septomaxillae seen in C. bicinctores, C. dickersonae, C. grismeri, C. insularis, and C. ves- tigium, which also have relatively elongate snouts. However, these taxa do not have the extreme con- dition present in G. wislizenii and G. copei. Elongate, slender septomaxillae are rarely observed within Ig- uania. In Opiums ( O . cuvieri and O. quadrimacu- latus), they are extremely slender, almost splinter- like, while in certain other iguanians ( Phrynosoma asio, P. orbiculare, P. coronatum, some Dipsosaurus dorsalis ) they are slender, although to a lesser degree. Elongate, slender septomaxillae are considered to be the derived state. However, septomaxillae are often destroyed during the preparation of skeletons and many outgroup taxa are not represented here. Because this feature appears to be associated with the much more elongate snout that occurs in G. wislizenii and G. copei, this character is treated as a character complex (although all of the differences that appear to be associated with an elongate snout cannot be polarized as can the septomaxillae con- ditions). Maxillae (Characters 12, 13; Fig. 2-5, 7, 8).— The premaxillary process contacts the premaxilla ante- riorly by means of an overlapping sheet of bone. It includes a well-developed shelf that passes posterior to the nasal process of the premaxilla and acts as the anterior wall of the external naris. The septo- maxilla contacts the posterodorsal edge of this shelf while posteroventrally the shelf is contacted by the vomer. In Gambelia wislizenii, G. silus, and five of eight G. copei (absent in REE 2798, 2802, 2805), a protrusion of the premaxillary process overlaps the lateral edge of the premaxilla such that the suture is saddle-shaped (Fig. 3-5). This condition is only rarely observed in the outgroups (present in some Chalaradon madagascariensis, Petrosaurus mearn- si, Urostrophus vautieri, some Pristidactylus torqua- tus, Enyalius brasiliensis, E. pictus, Phymaturus punae, some P. palluma, Leiocephalus melanoch- lorus, and some L. carinatus) and is considered to be the derived state. The dorsally directed nasal process of the maxilla contacts the nasal, prefrontal, and lacrimal bones and forms the posterolateral wall of the external naris and the lateral wall of the nasal capsule. A canthal ridge is present on the nasal process and extends from the rugose protuberance of the pre- frontal to the base of the premaxillary process near the posterolateral corner of the external naris. The angle of the canthal ridge, as well as the posterior margin of the external naris, is much greater (greater than 45 degrees) in Crotaphytus, Gambelia corona f, and G. silus than it is in G. wislizenii and G. copei due to the elongate snout of the latter two species. Several potentially useful characters are associated with the longer snout of G. wislizenii and G. copei, including the more elongate septomaxillae and vo- mers. However, as each of these appears to be linked to rostral elongation, they are considered as one character (see septomaxillae) in this analysis. Ventromedially, a thickening of the maxilla forms a shelf-like process that overlaps the palatine. This shelf projects further medially in Crotaphytus (Fig. 11, 12) than in Gambelia and is more nearly tri- angular. In Gambelia, the shape of the process is in the form of a low, rounded arch. There is extensive variation in the outgroups with regard to this feature and it was left unpolarized. Jugals (Characters 14, 15; Fig. 2-4, 8, 11). — The general shape of the jugal varies little in crotaphytids although three potentially useful variations were ob- served. A ridge, or thickening, is found on the ex- ternal surface of the jugal, extending from its im- mediate anterior end posteriorly just beyond the jugal’s articulation with the postorbital. The ridge is thicker in Crotaphytus than in Gambelia and is most developed in C. reticulatus. The function of this ridge is uncertain, although it provides the sur- face for attachment of the subocular scales. A lateral ridge is present on the jugal in many iguanians, al- though it is usually less strongly developed than that of Crotaphytus. Although this may eventually prove to be a phylogenetically useful character, it was not considered in this analysis. All crotaphytids possess an enlarged tubercle pos- terior to the termination of the maxillary tooth row (Fig. 2, 8, 11). This tubercle is actually comprised of both the jugal, which forms the anterior portion, and the ectopterygoid, which forms the posterior portion. The function of the tubercle appears to be as an attachment site for the ligamentum quadra- tomandibulare. The size of the tubercle is interspe- cifically variable, with Crotaphytus antiquus, C. col- laris, C. dickersonae, C. nebrius, and C. reticulatus having very large tubercles and the remaining taxa having small ones. Despite this variation in size, the presence or absence of a tubercle was coded as a binary character. In the outgroups, a similar tubercle is present in the leiosaurs Pristidactylus, Diplolae- mus, and Leiosaurus and a less similar laterally com- pressed tubercle is present in some chamaeleonids 18 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Fig. 1 1. — Ventral view of the skull of Crotaphytus dickersonae (REE 2777 , adult male, SVL =116 mm). Scale = 5 mm. ( Leiolepis belliana, Physignathus lesueurii). There- fore, the presence of a tubercle is considered to be the derived state and represents a synapomorphy for Crotaphytidae. There is also variation in the angle of the jugal where it serves as the ventrolateral border of the orbit. In Crotaphytus, the medial face of the jugal is oriented dorsolaterally at about a 45-degree angle over most of its length. In G. wislizenii, G. copei, and 1 5 of 3 1 G. silus, the medial face becomes pro- gressively more vertical anteriorly until it articulates with the palatine, lacrimal, and prefrontal. As a re- sult, the region of articulation of the three bones in Gambelia wislizenii, G. copei, and some G. silus is box-like because the jugal meets the palatine and prefrontal at perpendicular angles. In Crotaphytus and some G. silus, the jugal meets the prefrontal in a smooth, rounded arc. The box-like condition of the ventrolateral border of the orbit was approached only in Petrosaurus mearnsi, Uta stansburiana, Uma (but not Callisaurus, Cophosaurus, or Holbrookia maculata), one of two Enyalioides laticeps, and Lei- olepis belliana and is therefore considered to be the derived state within Crotaphytidae. Palate Vomers (Character 16; Fig. 11, 12). — In Crota- phytus insularis and C. vestigium, a separate pair of small bones, here termed extravomerine bones, may be present posteromedially where the vomers and palatines meet (Fig. 12). These medially contacting bones appear to be the result of secondary ossifi- cation centers in the vomers. In many specimens, this additional bone is present on one side only and the region where the bone is absent is filled in by the vomer from that side. Extravomerine bones are present in all five C. insularis available for study, although it is found on the right side only in one specimen (REE 2797). It is also found on at least one side in ten of 27 C. vestigium. Extravomerine bones are not present in the outgroup taxa examined here and no evidence has been discovered docu- menting their presence in other lizard species. Therefore, the presence of either one or two extra- vomerine bones is considered to be the derived state. Palatines (Character 17; Fig. 6, 11, 12). — In Cro- taphytus, the dorsal surface of the maxillary process usually bears the palatine foramen (Fig. 6), which may be situated in the suture of the maxillary pro- cess and the prefrontal or completely within the palatine. In one C. collaris (USNM 220216), the foramina were located entirely within the palatine processes of the prefrontals. A well-developed, transversely oriented canal, associated with the in- termediate palatine branch of nerve VII (Oelrich, 1956), projects medially from the palatine foramen (Fig. 6). In Gambelia, a palatine foramen is only rarely evident (five of 43 G. wislizenii, zero of eight G. copei, two of 30 G. silus), although the canal, and presumably the intermediate palatine branch of nerve VII, are present. Instead of passing through the prefrontal and palatine bones, the tube passes through the connective tissue medial to the palatine process of the prefrontal along the lateral border of the orbitonasal fenestra. The absence of a palatine foramen in the great majority of Gambelia appears to be the result of the narrower palatine process of the prefrontal found in this taxon, rather than the absence or rerouting of the intermediate palatine branch of nerve VII. Some variation was observed 1996 McGUIRE— SYSTEM ATICS OF CROTAPHYTID LIZARDS 19 within Crotaphytus including C. collaris (foramen present in seven of 51), C. grismeri (three of five), and C. reticulatus (22 of 26). The outgroup taxa are also extremely variable with respect to this feature, preventing its polarization. Phylogenetically useful variation was also observed in the palatine foramina of Phymaturus. In all specimens of Phymaturus ex- amined, the foramina were much larger proportion- ally than those of any other iguanian taxon exam- ined. Pterygoids (Characters 18-20; Fig. 2-4, 8, 11).— The transverse process of the pterygoid of Crota- phytus bears a sharp vertical crest near its lateral end. This crest is very weak or absent in Gambelia. A strong vertical crest is present in many iguanian species and its absence may be a synapomorphy for Gambelia. However, this crest appears to be asso- ciated with a more easily definable character of the ectopterygoid and its description is given in the dis- cussion of that element. The transverse processes of Crotaphytus reticu- latus and C. dickersonae are more ventrally ex- panded in comparison to the other crotaphytids. This condition is especially extreme in adult male C. dickersonae, which bear a well-developed crest that extends along the ventral edge of the entire transverse process terminating at, or near, the in- terpterygoid vacuity. This crest descends ventrally to a degree seen in no other crotaphytid species. Although it is difficult to compare this feature across a broad range of taxa with very different pterygoid morphologies, a strongly developed crest appears to be present in many corytophanids, chamaeleonids, and polychrotids, as well as within large iguanids. Therefore, this character was left unpoiarized. In Gambelia, the quadrate processes are approx- imately one-third shorter as a percentage of skull length than they are in Crotaphytus. In Crotaphytus, the posterior part of the skull is clearly longer than that of Gambelia and this is best illustrated by com- paring the posterior extents of the quadrate pro- cesses of the pterygoids, the supratemporal pro- cesses, and the paraoccipital processes with the pos- terior extent of the occipital condyle. In adult Cro- taphytus, all three processes project well posterior to the occipital condyle (Weiner and Smith, 1965; Fig. 2, 11), while in Gambelia, they reach a point roughly equidistant with the condyle (Fig. 3, 4). This condition is subject to considerable ontogenetic variation, with juveniles of both genera having the three processes extending posteriorly to a point equidistant with the occipital condyle until they reach Fig. 12. — Ventral view of skull of Crotaphytus vestigium (REE 2826, adult male, SVL = 105 mm) showing the extravomerine bones (vertically hatched) present in C. insularis and many C. vestigium. Scale = 5 mm. an SVL of approximately 80-85 mm. At this point in ontogeny, the processes begin to project further posteriorly in Crotaphytus than in Gambelia. The condition observed in adult Crotaphytus appears to be apomorphic and was only observed in large male Basiliscus basiliscus, Pristidactylus (as well as Di- plolaemus and Leiosaurus), Uromastyx acanthinu- rus, U. benti, U. microlepis, and Physignathus le- sueurii. This condition may represent an adaptation for more powerful jaw adduction in these lizards. In Crotaphytus and Gambelia silus, the quadrate processes project posterolaterally at a greater angle (approximately 26-3 1 degrees) than in G. wislizenii and G. copei (approximately 18 degrees). Most of the outgroup taxa appear to be similar to Crota- phytus and G. silus with respect to this feature, al- though enough variation was observed that the char- acter was left unpolarized. Ectopterygoids (Character 21; Fig. 2-4, 11). — In Crotaphytus, the transverse process of the pterygoid bears a strong vertical crest just medial to its artic- ulation with the ectopterygoid. In Gambelia, this crest is weakly defined and this appears to be as- sociated with the morphology of the ectopterygoid. The dorsal surface of the ectopterygoid is in the form of a sharp transverse edge or ridge that extends to the termination of the medially projecting process. 20 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 This ridge bears a posterior projection in Crotaphy- tus that overlaps the strong vertical crest of the transverse process. The ridge does not bear a strong posterior projection in Gambelia, instead being straight or nearly so. Presence or absence of a pos- terior projection of this ridge are coded as separate character states. However, there is much variation in the outgroups and this character could not be polarized. The ectopterygoid also bears a strong posterolat- eral process that is sutured to a similar process of the jugal. Together they form the tubercle that pro- jects posterolaterally just beyond the termination of the maxillary tooth row (see description of jugal above). Braincase Parabasisphenoid (Character 22; Fig. 11). — Pro- jecting anteriorly from the basisphenoid is the long, blade-like parasphenoid process. Although this is a separate osseous element, it is fused with the basi- sphenoid in postembryonic crotaphytids and, fol- lowing Jollie (1960:fig. 3), they are here treated as a single element, the parabasisphenoid. The posterior suture of the parabasisphenoid with the basioccipital differs between Gambelia and Cro- taphytus. In Gambelia, the parabasisphenoid bears long posterolateral processes that extend to the sphenoccipital tubercles. These processes are absent or extend only slightly beyond the transverse plane of the parabasisphenoid-basioccipital suture in most Crotaphytus examined (Fig. 1 1), although they may occasionally reach the base of the lateral process of the basioccipital. The posterolateral processes never were observed to reach the sphenoccipital tubercles, although they nearly reached the tubercle in two of 29 C. collaris (LLG 62, REE 2948). The majority of the outgroup taxa have long pos- terolateral processes of the parabasisphenoid that reach or nearly reach the sphenoccipital tubercles. Exceptions occur within the families Phrynosoma- tidae, Chamaeleonidae, Tropiduridae, and Poly- chrotidae. In Phrynosomatidae, short processes are present in Petrosaurus, Uta, Urosaurus graciosus, and Sator grandaevus (but not Sceloporus, at least those examined here; Appendix 1), while in Phry- nosoma and the sand lizards they are long. There- fore, short processes may be an additional syna- pomorphy for Petrosaurus plus the Sceloporus group, with a reversal in Sceloporus. Within Chamaeleonidae, short processes are pres- ent in Leiolepis belliana, but not Uromastyx or the basal agamines Physignathus lesueurii and Hydro- saurus amboiensis. Within chamaeleonines, Broo- kesia stumpffi has short processes, while all of the remaining chamaeleonines examined (Appendix 1) except Chamaeleo kerstenii have long processes. In C. kerstenii, the basioccipital is displaced forward by the exoccipitals such that it does not form the ventral portion of the occipital condyle. As a result, the basioccipital tubercles are found on the exoc- cipitals rather than the basioccipital. Thus, the ho- mology of the posterolateral processes (or lack there- of) of this species is questionable. In tropidurids, the processes are short in Cten- oblephary’s, Liolaemus, and some Leiocephalus { short in L. barahonensis, L. carinatus, L. lunatus, L. ma- cropus, and L. psammodromus; long in L. green- way i, L. melanochlorus, L. personatus, L. schrei- bersi, L. stictigaster, and L. vinculum ), but long in all of the Stenocercini and Tropidurini examined (Appendix 1) except T. spinulosus and T. melano- pleurus, which are nonbasal taxa (Frost, 1992). Within polychrotids, the processes are short in Pristidactylus, Diplolaemus, Leiosaurus, the anoles, the para-anoles (intraspecifically variable in Uros- trophus vautieri), and some Polychrus acutirostris (but not P. marmoratus), but long in Enyalius. Long posterolateral processes represent the an- cestral condition in Hoplocercidae, Opluridae, Cor- ytophanidae, Iguanidae, and Chamaeleonidae, and the polarity of this character is equivocal for Phry- nosomatidae and Tropiduridae (but long processes may be ancestral for Tropiduridae). It is most par- simonious to assume that short posterolateral pro- cesses were present in the common ancestor of Po- lychrotidae. Thus, the presence of short posterolat- eral processes are treated as the derived state within Crotaphytidae. Additional intergeneric variation was also ob- served in the parabasisphenoid. At the anterodorsal end of the basisphenoid is a depression, the sella turcica, that houses the pituitary gland. In adult Cro- taphytus, the sella turcica usually is elevated such that in lateral view, it is visible above the quadrate process of the pterygoid. In Gambelia, the sella tur- cica is more depressed and is rarely visible above the quadrate process. However, continuous varia- tion exists in this characteristic and it was omitted from the phylogenetic study. Mandible Dentary ( Fig. 13, 14). — In many iguanian lizards, the dentary is tubular anterior to the splenial and 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYT1D LIZARDS 21 Fig. 13. — Lingual view of the right mandible of (A) Crotaphytus reticulatus (REE 2912, adult male, SVL = 122 mm) and (B) Gambelia copei (REE 2800, adult female, SVL = 123 mm). Ang = angular. Art = articular. Cor = coronoid, Den = dentary, Pmf = posterior mylohyoid foramen, Pre = prearticular, Spl = splen- ial, Sur = surangular. Scale = 5 mm. completely encloses Meckel’s cartilage. In crota- phytids, the tubular nature of the dentary is incom- plete. The anterior end of the dentary is open, while posteriorly the groove is closed, but not fused. In Crotaphytus, with the exception of C. grismeri, the groove is usually closed over less than one-half of its length anterior to the splenial and there is rela- tively consistent interspecific variation in this char- acteristic. In C. collaris, C. nebrius, and C. reticu- latus, Meckel’s groove is often open over its entire length anterior to the splenial, and most of the re- maining specimens have the groove closed over less than one-third of its length. In C. antiquus and C. dickersonae, the groove is not open over its entire length, but as in the above-mentioned taxa, it was nearly always closed over less than one-third of its length. In C. bicinctores, C. insu/aris, and C. vestig- ium, Meckel’s groove is usually closed over between one-third and one-half of its length anterior to the splenial and was only once observed to be open over its entire length (C. vestigium, REE 2811). Crota- phytus grismeri is unique among Crotaphytus in that Meckel’s groove is closed over between approxi- mately 50 percent and 70 percent of its length in all specimens examined (five of five). Norell (1989) noted that in Gambelia, the groove is usually closed over two-thirds of its length anterior to the splenial. Unfortunately, this condition is much more variable in Gambelia than in Crotaphytus, and although the groove in most specimens is closed over greater than one-half of its length anterior to the splenial, 12 of 30 G. silus, two of nine G. copei, and 12 of 45 G. Fig. 14. — Labial view of the right mandible of (A) Crotaphytus reticulatus (REE 2912, adult male, SVL = 122 mm) and (B) Gambelia copei (REE 2800, adult female, SVL = 123 mm). Cor = coronoid, Den = dentary, Pre = prearticular, Sur = surangular. Scale = 5 mm. wislizenii had a condition similar to that observed in Crotaphytus, with the groove closed over less than half of its length anterior to the splenial. Because of this variation, this character was not considered in the phylogenetic analysis. Norell (1989) also considered an elongate dentary (with a posterior process projecting posterior to the superior apex of the coronoid, Etheridge and de Queiroz, 1988) to be a synapomorphy for Crota- phytidae. Although this character state was found to be derived in their phylogenetic analysis of pleu- rodont iguanians (possibly a paraphyletic assem- blage with respect to acrodont iguanians [Chamae- leonidae]), this state is widespread within Iguania and may be a synapomorphy for a group more in- clusive than Crotaphytidae. The dentary bears between three and eight mental foramina anteriorly. In Crotaphytus, the mental fo- ramina are usually restricted to the distal end of the dentary, while in Gambelia they may extend pos- teriorly to the midpoint of the bone. Continuous variation in this feature prevented its inclusion in the phylogenetic analysis. Angular (Characters 23, 24; Fig. 13). — In Crota- phytus, the exposed portion of the angular extends further anteriorly than in Gambelia wislizenii and G. copei. Defining states for this character is com- plicated by the variation that exists in those struc- tures that may serve as reference points. For this reason, two points of reference are included in the description of this character. In adult Crotaphytus, with very few exceptions, the angular extends an- teriorly at least to the fourth tooth (counting from BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 22 the rear of the tooth row) and usually well beyond this point. Juveniles are not always comparable be- cause their teeth are relatively larger than those of adults and are often widely spaced. The angular also extends well beyond the anterior extent of the cor- onoid in both adults and juveniles. In G. wislizenii and G. copei, the angular was never observed to reach the fourth tooth (from the rear of the tooth row) and rarely reached beyond the first. In most specimens, the angular does not extend as far an- teriorly as does the coronoid. In C. bicinctores, C. grismeri, and G. silus, the anterior extent of the angular shows continuous variation with most spec- imens having an intermediate condition but others with character states similar to those observed in G. wislizenii and G. copei or the remaining species of Crotaphytus. Because of this continuous variation in these three taxa, I have coded each as unknown for this character. With respect to the outgroup taxa, the angular projects well anteriorly in chamaeleon- ids, hoplocercids, the corytophamds Basiliscus bas- iliscus, B. vittatus, B. plumifrons, some Coryto- phanes cristatus, C. percarinatus, some Laemanctus longipes, L. serratus, and many polychrotids, while it is short in tropidurids (except Uranoscodon su- perciliosus ), phrynosomatids, oplurids (except O. fierinensis), and iguanids. The angular bears the posterior mylohyoid fora- men. This foramen usually is positioned well pos- terior to the superior apex of the coronoid in Gam- belia (eight of eight G. copei, 26 of 29 G. silus, 50 of 5 1 G. wislizenii ), while it is equidistant with, or anterior to, the superior apex in most Crotaphytus (posterior to the superior apex in two of 49 C. col- laris, three of 1 5 C. dickersonae, two of 1 7 C. ne- brius, two of 23 C. reticulatus, one of 27 C. vestig- ium). Although most of the outgroup taxa exhibit the condition observed in Crotaphytus, the presence of the posterior mylohyoid foramen posterior to the apex of the coronoid in phrynosomatids, some tro- pidurids, and some polychrotids (Frost and Ether- idge, 1989) as well as some oplurids prohibits po- larization of this character. Coronoid (Character 25; Fig. 13, 14). — The angle of the posterolingual process of the coronoid is near- ly vertical in Crotaphytus, while it extends poster- oventrally at an angle of approximately 45 degrees in G. wislizenii, G. copei, and G. corona\ (Norell, 1989). Gambelia silus may be intermediate in this feature or may approach the conditions observed in Crotaphytus or G. wislizenii-G . copei. Therefore, G. silus was coded as unknown (“?”) for this character. Most outgroup taxa have a condition similar to Cro- taphytus (state 0) or occasionally the intermediate condition usually present in G. silus. The outgroup taxa with the G. wislizenii-G. copei condition in- clude only Petrosaurus mearnsi, Phrynosoma doug- lassi, P. coronatum, Uromastyx, Brookesia stumpffi, and Chamaeleo kersteni (chamaeleonines as a whole are variable with respect to this feature). Therefore, the angled posterolingual process of the coronoid (state 1) is considered to be derived and the vertical condition ancestral. Norell (1989) considered this feature to be an unambiguous synapomorphy of Gambelia, presumably because he did not examine specimens of G. silus. Surangular (Characters 26-28; Fig. 13-15). — Im- mediately anterior to the articular facet lies a me- dially oriented knob-like process here referred to as the medial process. A thin shelf of bone may extend anteriorly between the distal extremity of the medial process and the body of the surangular (Fig. 15). This shelf is usually much more strongly developed in Gambelia and, to a lesser degree, Crotaphytus insularis than in the remaining Crotaphytus species. Crotaphytus vestigium is variable with respect to this character with seven of 27 having a shelf present. A lesser amount of variation was observed with a smaller shelf present in C. bicinctores (one of 25), C. collaris (five of 50), C. dickersonae (two of 16), C. nebrius (one of 17), and C. reticulatus (one of 14). In Gambelia, the shelf may entirely fill this space such that its edge may be either straight or, more frequently, convex in shape. The strongly de- veloped condition present in Gambelia suggests that it may be a further modification or intensification of the condition observed occasionally in Crota- phytus. Thin shelves of bone between the medial process and the ramus of the mandible are present in a small number of iguanian taxa, including Lei- olepis belliana, Opiums cuvieri, some Brachylophus fasciatus, some Uta stansburiana, Urosaurus auri- culatus, Microlophus grayi, and most Phymaturus taxa (absent only in P. palluma and some P. punae). The shelves only approached the condition of Gam- belia in the four Phymaturus patagonicus subspe- cies. This character was coded as a binary character with the absence of a shelf coded as state 0 and its presence as state 1 (taxa with intermediate frequen- cies coded appropriately). The presence of thin shelves of bone between the medial process of the surangular and the ramus of the mandible is inter- preted as the derived state. An additional process of the surangular may be 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 23 present immediately anterolateral to the articular facet. In Crotaphytus, a large knob-like process is present (here referred to as the lateral process), pre- sumably to provide a large surface area for insertion of the jaw adductor musculature of these lizards. In most Gambelia, no obvious process is visible, al- though in some individuals, a small elevation is present. In the outgroup taxa, a large lateral process is present in JJromastyx acanthinurus, U. microle- pis, Opiums fierinensis, some Leiocephalus macro- pus, Phrynosoma coronatum, some P. doug/assi, some Uma inornata, Urosaurus auriculatus, the leiosaurs Pristidactylus, Diplolaemus, and Leiosau- rus, the para-anoles, and Polychrus (although in Polychrus, the process is displaced further anteri- orly). The lateral process was enlarged to the degree observed in Crotaphytus only in Pristidactylus, Di- plolaemus, and Leiosaurus. The presence of an en- larged lateral process of the surangular is interpreted as the derived state within Crotaphytidae. In crotaphytids, a ridge on the dorsolateral surface of the surangular extends between the lateral process and the labial process of the coronoid. This ridge provides a broader area for insertion of M. adductor mandibularis extemus on the dorsal surface of the surangular. In Gambelia, the ridge is either absent or only weakly developed. In Crotaphytus, the ridge and corresponding dorsal shelf are more strongly developed, and in C. reticulatus, the ridge is ex- tremely well developed providing a concave area for muscle insertion in adults (Fig. 1 5). This feature was coded as an unordered multistate character with Gambelia given state 0, Crotaphytus (except C. re- ticulatus) given state 1 , and C. reticulatus given state 2. All of the outgroup taxa either lacked this ridge or had a very weakly developed one (state 0), with the possible exception of Hydrosaurus amboiensis, in which a ridge is present near the ventrolateral border of the mandible. Opiums fierinensis, and some Phrynosoma (P. asio and some P. doug/assi and P. orbiculare), in which the ventrolateral portion of the mandible is greatly expanded. The absence of a ridge or the presence of a weakly developed one is con- sidered to be the ancestral state. Prearticular (Character 29; Fig. 1 3-1 5). — Poste- riorly, the prearticular bears two large processes that serve as insertion sites for jaw adductor and de- pressor muscles. The angular process projects ven- tromedially from a point just below the articular facet, while the retroarticular process projects pos- teriorly. In Gambelia, thin shelves of bone extend between the processes of the posterior portion of the A B Fig. 15. — Dorsal view of the posterior portion of the right man- dible in (A) Crotaphytus reticulatus (REE 29 1 2, adult male, SVL = 122 mm) and (B) Gambelia copei (REE 280CT, adult female, SVL = 123 mm). LP = lateral process, TC = tympanic crest. Arrow indicates the shelf that extends between the medial process and the ramus of the mandible in Gambelia. Scale = 3 mm. mandible and the ramus of the mandible. One such shelf was discussed above with the surangular. Two additional shelves may also be present, both of which are associated with the angular process. One extends between the angular process and the retroarticular process, while the other extends forward from the angular process to the body of the lower jaw. Shelves of bone that extend between the processes of the mandible and the ramus of the mandible were treat- ed as a single character (see surangular). The shape of the retroarticular process and its tympanic crest in crotaphytids is distinctive. In dor- sal view, the retroarticular process is roughly qua- drangular, while in lateral view it is more nearly triangular. The distal terminus of the process is ex- panded, giving it a bulbous appearance. The tym- panic crest is more broadly expanded in Gambelia than in Crotaphytus (Fig. 1 5), but was not scored as a separate character state. The tympanic crest in all crotaphytids is robust and its edge expands poste- riorly such that at the end of the process, it is nearly 24 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 as broad as the process itself. Furthermore, the tym- panic crest angles posterodorsally and, thus, does not form the lateral border of the retroarticular pro- cess as it does in most other iguanian taxa examined (illustrations of the ancestral condition of the tym- panic crest can be seen for Dipsosaurus dorsalis in de Queiroz, 1987:fig. 29; for Basiliscus vittatus and Corytophanes cristatus in Lang, 1989:fig. 31; and for Physignathus cocincinus in Moody, 1980:fig. 16). The angle of the tympanic crest gives the retroar- ticular process a twisted appearance. The orienta- tion of the tympanic crest appears to undergo an ontogenetic change from the standard position along the lateral border of the retroarticular process in juveniles to a more posterodorsal orientation in adults. The medial crest of the retroarticular pro- cess, discussed by de Queiroz (1987), is only vari- ably present in crotaphytids. A similar posterodor- sal curvature of the tympanic crest was observed only in Opiums cuvieri and one Polychrus acutiros- tris (REE 568). Therefore, this condition is inter- preted as a synapomorphy for Crotaphytidae. Miscellaneous Features of the Head Skeleton Marginal Teeth (Characters 30, 31; Fig. 8, 11- 14). — The marginal teeth of crotaphytids are char- acteristic of most pleurodont iguanians in that the anterior teeth are conical and the posterior maxillary and dentary teeth are compressed and tricuspid. The dentition of crotaphytids has been described as het- erodont or weakly subheterodont (Marx, 1950; Wei- ner and Smith, 1965) because the teeth sometimes grade from conical to bicuspid then tricuspid (the bicuspid state is often omitted). This transition usu- ally begins further anteriorly in Crotaphytus (mean maxillary tooth position x = 8.11 , n - 152) and Gambelia situs (x = 8.08, n = 30) than in G. wis- lizenii (x = 1 1.27, n = 43) or G. copei (x = 11.13, n = 8), although the ranges overlap extensively. Het- erodonty was considered to be more developed in Gambelia than Crotaphytus by Marx (1950) and Weiner and Smith (1965) and was used as a char- acter to distinguish between the genera. However, Montanucci (1969) found that the degree of heter- odonty was indistinguishable between adult G. wis- lizenii and many C. collaris, especially juveniles. The degree of cuspation is certainly more pro- nounced in Gambelia than Crotaphytus and, despite the ontogenetic variation discussed by Montanucci (1969), this subtle variation could probably be cod- ed into discrete character states. However, degree of cuspation varies continuously within iguanians and this character therefore may be added to the long list of currently unpolarizable differences be- tween Crotaphytus and Gambelia. As in many ig- uanian lizards, the number of maxillary and dentary teeth increases ontogenetically, at least early in on- togeny. The number of premaxillary teeth does not increase ontogenetically. In some individuals of both Crotaphytus (Ether- idge, 1960; personal observation) and Gambelia, the tooth rows of the mandibles and/or maxillae may be doubled for a short distance (two sets of teeth occurring side by side). Although Etheridge (1960) hypothesized that this variation may be restricted to males, it actually occurs in both sexes. The number of maxillary and dentary teeth tends to be greatest in Gambelia wislizenii, G. copei, and Crotaphytus dickersonae (Tables 3, 4). The large number of teeth in these Gambelia is not surprising given the elongate snout that is characteristic of these species. The large number of teeth observed in C. dickersonae is the result of very closely spaced den- tition. The small number of teeth present in G. silus is probably correlated with the truncated snout of this species and may therefore be a plesiomorphic retention. Discrete character states could not be as- signed describing numbers of maxillary and dentary teeth. Therefore, this variation was not considered in the phylogenetic analysis. The number of premaxillary teeth varies within Crotaphytidae (Tables 3, 4). Gambelia is character- ized by the strong statistical mode of seven pre- maxillary teeth, while most Crotaphytus taxa have a somewhat weaker statistical mode of six. How- ever, C. dickersonae and some populations of C. collaris (those formerly referred to the subspecies C. c. baileyi) have modes of seven. This variation was coded as a multistate character using a step matrix and the Manhattan distance frequency approach (see Appendix 4). This character was not polarized. All crotaphytids have recurved anterior maxillary and dentary teeth, a condition that is more devel- oped in Gambelia than Crotaphytus, which have broader, more peg-like teeth (especially evident in C. reticulatus). Long, slender, recurved maxillary and dentary teeth, as present in Gambelia, were not observed in any of the outgroup taxa and are there- fore treated as the derived state. Palatal Teeth (Characters 32, 33; Fig. 11).— At the base of the pterygoid process of each palatine, 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 25 Table 3 . — Tooth count data for Crotaphytus. Premaxillary teeth Maxillary teeth Dentary teeth Crotaphytus : antiquus ( n = 4) mean ± SD 5.8 ± 0.50 16.9 ± 0.44 22.0 ± 1.20 range (5-6) (15-19) (21-24) bicinctores (n = 24) mean ± SD 6.2 ± 0.53 16.9 ± 0.44 23.0 ± 2.25 range (5-7) (15-21) (18-28) collaris ( n — 49) mean ± SD 6.2 ± 0.76 17.2 ± 1.78 21.5 ± 2.48 range (5-8) (14-22) (16-26) dickersonae (n = 16) mean ± SD 7.1 ± 0.95 20.3 ± 2.60 24.8 ± 3.51 range (6-9) (16-25) (19-31) grismeri (n = 5) mean ± SD 6.4 ± 0.89 18.6 ± 1.96 23.1 ± 2.23 range (6-8) (16-21) (19-26) insularis (n = 5) mean ± SD 6.0 ± 0.00 18.1 ± 1.60 23.8 ± 1.99 range (6) (15-20) (22-28) nebrius (n = 17) mean ± SD 6.2 ± 2.17 18.3 ± 2.17 22.4 ± 2.73 range (5-7) (15-23) (19-30) reticulatus ( n = 25) mean ± SD 6.0 ± 0.64 17.7 ± 1.69 21.9 ± 1.75 range (5-7) (14-21) (17-25) vestigium ( n = 28) mean ± SD 6.2 ± 0.39 18.3 ± 1.61 23.3 ± 2.12 range (6-7) (15-22) (19-28) most crotaphytids have an enlarged ridge that may collaris, 19 of 45; C. dickersonae, 12 of 16; C. gris- support palatine teeth. This ridge is usually more meri, two of five; C. nebrius, 11 of 15; C. reticulatus, developed in Gambelia than Crotaphytus. Most 17 of 26; C. vestigium, ten of 25), although only C. Gambelia (G. wislizenii, 39 of 46; G. copei, eight of insularis (zero of five) always lacked palatine den- nine; G. silus, 17 of 3 1) have palatine teeth. Within tition. Among the outgroup taxa examined, palatine Crotaphytus, the palatine ridge is almost always teeth are present only in some Opiums ( O . quadri- present but the teeth are only variably present (C. maculatus) and most polychrotids (all but Poly- bicinctores, ten of 24; C. antiquus, three of four; C. chrus, although palatine teeth are also absent in all Table 4. — Tooth count data for Gambelia. Premaxillary teeth Maxillary teeth Dentary teeth Gambelia-. copei (n = 9) mean ± SD 7.0 ± 0.00 20.9 ± 1.09 26.2 ± 1.54 range G) (19-23) (23-29) silus (n = 31) mean ± SD 6.6 ± 0.57 17.7 ± 1.46 22.0 ± 1.56 range (5-7) (14-20) (19-25) wislizenii (n = 45) mean ± SD 6.9 ± 0.42 19.9 ± 2.26 24.9 ± 2.80 range (6-8) (15-24) (18-31) 26 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 anoles except Chamaeleolis). Because Frost and Etheridge (1989) found Polychrus to be the sister taxon of the anoles, the presence of palatine teeth is considered as the ancestral state for Polychroti- dae. Therefore, if palatine teeth are to be considered apomorphic for Crotaphytidae, it must be assumed that Crotaphytidae and Polychrotidae are not sister taxa. Such a relationship was not supported in the analysis of Frost and Etheridge (1989) as depicted in their 12 equally parsimonious unrooted trees. Therefore, palatine teeth are tentatively considered to be apomorphic for Crotaphytidae. All crotaphytids possess pterygoid teeth on the posteromedial border of the palatine process (Fig. 1 1). These teeth may form a single row or, late in ontogeny, exist as a patch. During ontogeny, the number of pterygoid teeth clearly increases, al- though there is not a perfect correlation between number of teeth and SVL and some very large in- dividuals have relatively few teeth. Additional teeth are usually added to the posterior portion of the patch, and in larger individuals, the majority of the teeth are found posteriorly. In some juvenile and most adult Crotaphytus, the posterior aspect of the pterygoid tooth row curves laterally away from the interpterygoid vacuity (Fig. 1 1), while in Gambelia the tooth row follows the margin of the vacuity. Polarization of this character is complicated by the absence of pterygoid teeth in the families Phryno- somatidae and Chamaeleonidae and in some Phy- maturus and Leiocephalus. Furthermore, pterygoid teeth are often intraspecifically variable and limited sample sizes for certain outgroup species probably did not allow them to be coded adequately for this character. However, in the remaining outgroup taxa examined, the pterygoid tooth patch was observed to curve posterolaterally only in Uranoscodon su- perciliosus, Corytophanes percarinatus, some C. cristatus, some Laemanctus serratus, Brachylophus fasciatus, and Pristidactylus casuhatiensis (see de Queiroz, 1 987, for additional iguamd taxa with pos- terolaterally curved pterygoid tooth patches). There- fore, the posterolateral curving of the pterygoid tooth patch was considered to be the derived state within Crotaphytidae. Scleral Ossicles.- The scleral ossicles are thin, overlapping platelets of bone that form a supportive ring within the anterior portion of the sclera of the eye. De Queiroz (1982) found that most iguanian taxa are characterized by a standard pattern con- sisting of 14 ossicles, with numbers one, six, and eight positive (overlapping both of the adjacent os- sicles), numbers four, seven, and ten negative (over- lapped by both of the adjacent ossicles), and the remaining ossicles imbricating (overlapping one of the adjacent ossicles, but itself overlapped by the other). He noted that this pattern is present in Cro- taphytus collaris, C. vestigium, and Gambelia wis- lizenii. I have verified his observations for these species, and report further that the remaining cro- taphytid taxa are also characterized by this appar- ently ancestral iguanian condition. A list of speci- mens for which the scleral ossicles have been ex- amined is provided in Appendix 7. Hyoid Apparatus (Characters 34-36; Fig. 16). — A number of differences in the morphology of the hy- oid apparatus exist between Crotaphytus and Gam- belia. In Crotaphytus, the ceratohyals may be greatly expanded proximally, such that a large hook or pro- cess is present (processes absent in one of four C. antiquus). Their development is subject to ontoge- netic variation and subadults did not have the hook; therefore, the character was scored only from adults. In Gambelia, the proximal portion of the ceratohyal may be somewhat compressed; however, well-de- veloped hooks are absent. This character varies ex- tensively in the outgroups and was therefore left unpolarized. In Gambelia, the second ceratobranchials are short, extending posteriorly for about half the length of the ceratohyals and first ceratobranchials, while in Crotaphytus they are longer, extending more than two-thirds the length of the ceratohyals and first ceratobranchials (Robison and Tanner, 1962; Fig. 16). The second ceratobranchials of C. dickersonae are often particularly long and in adult males usually extend as far posteriorly as do the ceratohyals and first ceratobranchials. However, this was not treated as a separate character state because of continuous variation between the extreme C. dickersonae con- dition and that present in other Crotaphytus, par- ticularly in C. collaris. The longer second cerato- branchials of Crotaphytus may function in the de- pression of their more strongly developed gular pouch. The outgroups vary continuously in the length of the second ceratobranchials ranging in relative length from very short in Phymaturus to extremely elongate in Polychrus, the anoles, and Brachylophus. Therefore, this character was left unpolarized. In Crotaphytus, the second ceratobranchials are in close contact, although they are not actually fused, whereas in Gambelia, they may be widely separated 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 27 (Fig. 16). They were separated in one of four C. antiquus, in all specimens of G. copei and G. silus, and in at least ten of 1 5 G. wislizenii. However, in those five specimens of G. wislizenii in which the second ceratobranchials were in contact, the contact may have been an artifact of preparation. Separated second ceratobranchials are relatively rare in igu- anians and were only observed in Uta stansburiana, some Petrosaurus mearnsi, Phrynosoma asio, some Uma exsul, some Brachvlophus fasciatus, Phyma- turus, some Leiolepis belliana, and Enyalius bili- neatus. Separated second ceratobranchials was con- sidered to be the derived state within Crotaphytidae. However, this character could not be evaluated in many outgroup taxa because the hyoid apparatus is often damaged in preparation and this polarity as- sessment should only be considered tentative. Skull Rugosity (Character 37). — Rugosity of the skull was considered to be a synapomorphy for Cro- taphytus by Frost and Etheridge (1989). Although rugosities may indeed be found in all Crotaphytus taxa (rugosities are not found in Gambelia ), there is much variation with respect to the ontogenetic pe- riod during which rugosities develop. For example, most C. collaris develop rugosities as subadults, while C. bicinctores , C. dickersonae, and C. nebrius con- sistently develop rugosities only after reaching adult size. In C. grismeri, C. insu/aris, C. reticulatus, and C. vestigium, rugosities may be lacking even in large adults. For example, an extremely large C. vestigium (REE 2935; SVL = 125 mm) completely lacks skull rugosity, while several much smaller individuals have them. This variation was coded as a binary character with the absence of skull rugosity as state 0, and the presence of skull rugosity at some point in ontogeny as state 1 . This character could not be polarized. Axial Skeleton Presacral Vertebrae (Character 38). — The presa- cral vertebrae of crotaphytids are procoelous and have supplemental articular facets, zygosphenes and zygantra, medial to the pre- and postzygapophyses. A large posterodorsally oriented suprazygapophy- sial process is present on the atlas. Crotaphytids retain the apparently plesiomorphic mode of eight cervical vertebrae and 24 presacral vertebrae, al- though individuals occasionally have nine cervicals and more frequently may have 23 or 25 total pre- sacrals. Four to seven ventrally keeled intercentra Fig. 16. — Hyoid skeletons of (A) Crotaphytus collaris ( REE 2952, adult male, SVL = 131 mm), (B) C. dickersonae (REE 2905, adult male, SVL = 1 06 mm), and (C) Gambelia copei (REE 2800, adult female, SVL = 1 23 mm). B = body of hyoid, Bh = Basihyal, Cbl = first ceratobranchial, Cb2 = second ceratobranchial, Ch = Ceratohyal, Hh = hypohyal. Scale = 10 mm. occur between the anteriormost cervical vertebrae and these decrease in size posteriorly. The zygosphenes and zygantra of all crotaphytid taxa except Gambelia silus are weakly to moderately developed, according to the criteria established by Hoffstetter and Gasc (1969) and modified by de Queiroz (1987). In the weak form, the facet of the zygosphene faces dorsolaterally, while in the mod- erately developed form, the facet faces either lat- erally or ventrolaterally. The most strongly devel- oped form of zygosphene is characterized by a ven- trolaterally facing facet with a notch separating this facet from the prezygapophysis. This condition is approached in four of five G. silus, in which either a notch is present or a very thin sheet of transparent bone fills the space. Although notched zygosphenes are present in several of the outgroup taxa, including corytophanids, iguanids exclusive of Dipsosaurus, Uranoscodon superciliosus, Polychrus marmoratus, and some Enyalius ( E . boulengeri, E. bilineatus), the condition of G. silus is considered to be the derived state within Crotaphytidae. Caudal Vertebrae (Characters 39, 40). — The num- ber of caudal vertebrae present in crotaphytids is remarkably consistent with all of the species having between 54 and 63. No gaps were observed sug- gesting that the number of caudal vertebrae is not phylogenetically informative within Crotaphytidae. Most of the caudal vertebrae bear neural arches, transverse processes, and haemal arches, all of which 28 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 decrease in size posteriorly and disappear before the caudal terminus. The first haemal arch or rudimen- tary haemal arch usually occurs between the second and third or third and fourth caudal vertebrae, al- though it may occasionally lie between the first and second caudal vertebrae. The number of transverse processes is highly variable. Relatively few trans- verse processes are present in C. insularis (14-18, x = 16.6), C. grismeri (16-22, x = 18.0), G. silus (14-24 + ,x= 18.0), G. wislizenii^ 13-26, x = 18.4), C. antiquus (19-22, x = 20.3), C. vestigium (17-30, x = 21.3), C. bicinctores (16-26, x = 21.9), and G. copei ( 1 7-26, x = 23.3), while an intermediate num- ber is present in C. dickersonae (24-35, x = 28.6), and a relatively large number are found in C. reti- culatus (29-38, x = 33.4), C. nebrius (23-42, x = 34.9), and C. collaris (27-47, x = 37.4). These num- bers may be complicated by ontogenetic variation as juveniles tended to have fewer transverse pro- cesses than adults. Although the data presented here are suggestive, the extensive interspecific overlap in ranges prevented the assignment of discrete char- acter states for each taxon. Therefore, this variation was not considered in the phylogenetic analysis. In many iguanian lizards, the transverse processes of the more anterior caudal vertebrae project pos- terolaterally but abruptly change to an anterolateral orientation over the span of a few vertebrae (Eth- eridge, 1967). As Etheridge (1967) pointed out, this condition is present in crotaphytids, although in two taxa unavailable to Etheridge at the time, C. grismeri (five of five) and C. insularis (four of five), this change in orientation usually does not occur. The shift in orientation did not occur in seven of 15 C. bicinc- tores, one of four C. antiquus, one of 1 5 C. dicker- sonae, three of 2 1 C. vestigium, and four of 2 1 G. wislizenii. The ranges and means for the caudal ver- tebra number at which the shift in orientation of the transverse processes occurs for each taxon follows: C. antiquus (8-15, x = 10.7), C. dickersonae (8-12, x = 1 1.3), C. insularis (12), C. nebrius (10-17, x = 12.5), C. collaris (10-18, x = 13.3), G. silus (13-16, x= 14.2), C. vestigium (9-22, x = 14.3), G. wislizenii (13-18, x= 15.4), C. reticulatus (14-20, x = 16.1), G. copei (16-23, x = 17.1), and C. bicinctores (17- 23, x = 19.9). Again, the extensive interspecific overlap in ranges limits the phylogenetic usefulness of this variation and it was not considered in the phylogenetic analysis. Adult male C. bicinctores, C. dickersonae, C. gris- meri, C. insularis, and C. vestigium are characterized by the presence of a strongly laterally compressed tail (Fig. 3 IB, 32A-D). In each of these species, the tail is not only compressed, but relatively taller than in other crotaphytids and this is reflected in the morphology of the caudal vertebrae. The neural and haemal arches are relatively longer and the trans- verse processes narrower. In the species with strong- ly compressed tails the neural spines are approxi- mately 2. 0-3.0 times longer than the transverse pro- cesses while in the remaining species of Crotaphytus and in Gambelia, the neural spines are shorter than the transverse processes, approximately equal in length, or, in the case of C. reticulatus, approxi- mately 1.5 times longer than the transverse pro- cesses. The tail of C. reticulatus may be weakly lat- erally compressed. However, the tail is never com- pressed to the degree observed in the species men- tioned above and in some individuals may not be compressed at all. Furthermore, the height of the laterally compressed tail of the other species is en- hanced by the presence of large fat bodies on the dorsal and ventral crests of the tail. These large fat bodies are not present in C. reticulatus or any other crotaphytid, although I have observed a minute line of fat on the dorsal surface of the tail of one C. collaris. Although several anatomical systems have been modified to produce the lateral tail compres- sion of C. bicinctores, C. dickersonae, C. grismeri, C. insularis, and C. vestigium, these modifications are clearly associated with one complex character and are treated as such in this analysis. Although lateral tail compression occurs in several iguanian families, I have not observed similar fat bodies in the tails of these taxa. Therefore, lateral tail com- pression with the presence of dorsal and ventral fat bodies is considered to be the derived state within Crotaphytidae. Autotomic fracture planes of the caudal vertebrae are widespread in squamates and rhynchocepha- lians and at the level of Iguania certainly represent a plesiomorphic retention (Etheridge, 1967; Hoffs- tetter and Gasc, 1969). While fracture planes are present in most Gambelia, fracture planes are absent from Crotaphytus (Etheridge, 1967). Fracture planes were present in five of five G. silus and seven of ten G. copei (and apparently fused in the remaining three). Fracture planes were present in 19 of 23 G. wislizenii; however, the four that lacked them were the only four specimens available from Isla Tiburon and, thus, may represent a derived feature for this insular population. Many iguanian taxa lack auto- 1996 McGUIRE— SYSTEM ATICS OF CROTAPHYTID LIZARDS 29 tomic fracture planes, including the hoplocercid Ho- plocercus, the phrynosomatid Phrynosoma, some tropidurids of the genus Tropidurus, the polychro- tids Phenacosaurus, Chamaeleolis, Leiosaurus, Po- lychrus, Urostrophus, Anisolepis, Chamaelinorops, and some Enyalius and Anolis, the corytophanids Corytophanes and Laemanctus, the iguanids Iguana delicatissima, Conolophus, Amblyrhynchus, and Brachylophus, and all chamaeleonids except occa- sional Uromastyx (Etheridge, 1967; de Queiroz, 1987; Frost and Etheridge, 1989; R. Etheridge, per- sonal communication, 1993). Thus, it is most par- simonious to assume that autotomic fracture planes were present in the common ancestors of the fam- ilies Opluridae, Hoplocercidae, Iguanidae, Phry- nosomatidae, and Tropiduridae, given the phylo- genetic relationships that have been proposed for these groups (Etheridge and de Queiroz, 1988; Frost and Etheridge, 1989; Norell and de Queiroz, 1991; Frost, 1992). The polarity of this character is equiv- ocal for Corytophanidae and Polychrotidae (given the relationships proposed by Frost and Etheridge, 1989). The absence of fracture planes is known to be the ancestral condition only with respect to the family Chamaeleonidae. Although this character cannot be unequivocally polarized given the out- group uncertainties, I have tentatively coded the absence of autotomic fracture planes as the derived state. Etheridge (1967) mentioned that iguanians with the autotomic version of the type one iguanid ( senso lato) vertebral pattern (vertebrae with single trans- verse processes and fracture planes, when present, that pass posterior to the transverse process), of which Gambelia is an example, usually have be- tween five and 1 5 nonautotomic vertebrae that pre- cede the first autotomic vertebra. Gambelia gener- ally fits this pattern with the first fracture plane oc- curring in G. wislizenii somewhere between the 1 4th and 22nd vertebrae, in G. copei between the 1 8th and 21st vertebrae, and in G. silus between the 13th and 1 5th vertebrae. Ribs (Character 41). — Crotaphytids are charac- terized by a generally plesiomorphic complement of ribs, although phylogenetically informative varia- tion is present. As in other iguanians, most of the ribs have a bony dorsal portion and a cartilaginous ventral portion, the inscriptional rib, that may either connect the bony portion with the sternum or xiphi- stemum or end free. The first rib-bearing cervical vertebra is usually the fourth, although the third vertebra supports ribs in numerous individuals, and in a few, the second vertebra supports ribs. Thus, there are usually five cervical ribs, although six or seven are not uncommon. The cervical ribs are fol- lowed by four sternal ribs that connect the vertebral column to the posterolateral border of the sternum (only three sternal ribs present in one of four C. antiquus). The sternal ribs are followed by either one ( Gambelia ) or two ( Crotaphytus ) xiphisternal ribs that connect the vertebral column with the xiphisternum. Finally, there may be a series of post- xiphisternal ribs that end freely. The ribs rapidly decrease in length posteriorly to a width roughly equal to that of the sacral pleuropophyses. The ter- minal presacral ribs are often smaller than those immediately anterior to them and are very rarely fused to the vertebra. Three xiphisternal patterns were observed and two of these appear to be quite consistent. Crotaphytus has a pattern of two xiphisternal ribs with an oc- casional free xiphisternal rod. Gambelia have just one xiphisternal rib and one free xiphisternal rod that curves anteromedially. Variation was observed in two specimens of Crotaphytus (C. bicinctores, REE 2934; C. collaris, REE 2948) and two specimens of Gambelia (G. silus, CAS 22742; G. wislizenii, REE 2918). Both Crotaphytus specimens had the con- dition characteristic of Gambelia, although REE 2934 varied on one side only. The apparently anom- alous specimens of G. silus and G. wislizenii had two xiphisternal ribs plus a free xiphisternal rod, a condition observed infrequently in Crotaphytus. Etheridge (1959) found two xiphisternal ribs to be present in oplurids, corytophanids, iguanids, hoplo- cercids, polychrotids, tropidurids (with the excep- tion of Phymaturus and Uracentron), and phryno- somatids (except Phrynosoma,, which have no xiph- isternal ribs, and Callisaurus,, with three). Frost (1992) listed several additional species of Tropi- durus and one Microlophus with three xiphisternal ribs. In chamaeleonids exclusive of chamaeleonines, one xiphisternal rib is the common condition and is present in the presumably basal lineages of aga- minae ( Physignathus , Hydrosaurus), while the ab- sence of xiphisternal ribs were characteristic of Uro- mastyx and Leiolepis (Moody, 1980). In the few chamaeleonines that I have examined ( Chamaeleo senegalensis, C.johnstoni), two xiphisternal ribs were present, although variation within chamaeleonines seems likely. Because two xiphisternal ribs is clearly the ancestral condition in all of the iguanian families 30 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 except Chamaeleonidae, the presence of two xiph- isternal ribs is assumed to be the ancestral state within Crotaphytidae. Therefore, two xiphisternal ribs was coded as state 0 and that of a single xiph- isternal rib as state 1 . The shape of the xiphisternal rod of Gambelia is similar to that described in Tropidurus semitaen- iatus (Frost, 1992) in that the free end of the car- tilaginous rod curves anteromedially, crossing su- perficially to the xiphisternal rib and posteriormost sternal ribs. The posterior xiphisternal rod serves as the origin for nearly the entire posterior portion of M. pectoralis major, although it does not serve as the entire origin as in T. semitaeniatus. Regardless of whether the posteriormost xiphisternal cartilage ends freely or is continuous with a bony rib, it ap- pears to serve as the site of origin for a portion of M. pectoralis major. This appears to be the case even in those taxa that have extremely short carti- laginous protuberances projecting posteriorly from a second xiphisternal rib, for example G. silus and certain phrynosomatids (Etheridge, 1964). Pectoral Girdle Suprascapulae (Character 42). — The suprasca- pulae are composed entirely of calcified cartilage and lie dorsal to the scapulae. In Crotaphytus and some Gambelia, a deep notch is present in the an- terior margin of the suprascapula giving it the ap- pearance of a hook. This notch is usually present in Crotaphytus and variably present in Gambelia (five of 23 wislizenii, one of seven copei, one of five G. silus). Most of the outgroup taxa lack a strongly developed notch in the suprascapula (present in one of one Corytophanes hernandezi and four of four Uma scoparia). Therefore, the presence of a supra- scapular notch is treated as the derived state. Scapulae, Coracoids, and Epicoracoids (Charac- ters 43, 44). — In crotaphytids, the posterior coracoid fenestrae are nearly always present (absent on one side only in one of five specimens of C. insularis, and on one side only in one of 23 G. wislizenii). In C. reticulatus, the posterior coracoid fenestrae were observed to be absent in three of nine individuals. Furthermore, they were either proportionally small- er or present unilaterally in the remaining large spec- imens, suggesting that the fenestrae are lost late in ontogeny in this species. Posterior coracoid fenes- trae are absent in the great majority of iguanians and among the outgroup taxa are present in Uro- mastyx, Liolaemus, Stenocercini, Tropidurini, ig- uanids exclusive of Dipsosaurus and Brachylophus, para-anoles, Enyalius, Pristidactylus, Leiosaurus, and Diplolaemus (Savage, 1958; Etheridge, 1959; Moody, 1980; de Queiroz, 1987; Frost and Ether- idge, 1989). The weakly developed posterior cora- coid fenestrae of the latter three taxa were consid- ered by Frost and Etheridge (1989) to represent a separate character state. The presence of posterior coracoid fenestrae are considered to be the derived state and may represent a synapomorphy for Cro- taphytidae. The ontogenetic loss of the posterior coracoid fenestrae in C. reticulatus may represent an autapomorphy for the species. However, addi- tional osteological material is required to evaluate this potentially distinct character state and it was not treated as such in the phylogenetic analysis. In Gambelia, a calcified extension of the epicor- acoid cartilage forms the anterior border of the scap- ular fenestra. The anterior border of the scapular fenestra was either absent or incomplete in all of the Crotaphytus specimens examined except three of 2 1 C. bicinctores, one of 12 C. collaris, one of five C. grismeri, one of five C. insularis, and three of 2 1 C. vestigium. However, in all of these specimens except two of the three C. vestigium and the one C. collaris, the border of the fenestra was not completed by calcified cartilage, but rather by a thin sheet of bone or connective tissue. In adult C. reticulatus, the cal- cified cartilage extends dorsally from the ventral border of the scapular fenestra approximately half way to the dorsal border of the fenestra, a condition that may represent an intermediate step between the condition observed in Gambelia and that observed in most other Crotaphytus. Because the cartilage was present in 34 of 35 specimens of Gambelia exam- ined, it seems unlikely that the variation observed was an artifact of preparation. Character polarity could not be evaluated in many of the outgroup taxa because they lack scapular fenestrae, including Cha- maeleonidae, Polychrotidae (variable in Polvchrus), Corytophanidae, Liolaeminae, Hoplocercidae (ex- cept Enyalioides laticeps), Petrosaurus, Uta, and Urosaurus (Frost and Etheridge, 1 989). In those out- group taxa that have scapular fenestrae, most have the calcified cartilage borders, including phrynoso- matids (except P. orbiculare), hoplocercids, oplur- ids, iguanids, tropidurids ( Leiocephalus and Ura- noscodon), and Hydrosaurus amboiensis (other aga- mines lack scapular fenestrae [Frost and Etheridge, 1 989]). Therefore, the absence of a calcified cartilage 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 31 anterior border of the scapular fenestra is tentatively coded as the derived condition. Clavicles (Character 45). — In Gambelia, the clav- icles usually (all Gambelia except two of 23 G. wis- lizenii ) bear extensive fenestrations. Fenestrations were also present in all Crotaphytus reticulatus ex- amined, although Montanucci (1969) found that they were absent in six of the 14 specimens he examined. These fenestrations were absent in all 39 C. collaris examined, as well as in the 14 C. nebrius and five C. insularis examined. However, in the remaining species of Crotaphytus, there was much variability in this character with four of 2 1 C. bicinctores, two of four C. antiquus, five of 1 6 C. dickersonae, three of five C. grismeri, and two of 2 1 C. vestigium having fenestrated clavicles. Although Weiner and Smith (1965) noted that clavicular fenestrations were ab- sent in the 54 specimens of C. collaris they exam- ined, Robison and Tanner (1962) observed them in 20 percent of their specimens (although they in- cluded specimens of the yet-to-be-described C. bi- cinctores in their sample, which at the time was considered to be C. c. baileyi ) and Montanucci (1969) observed them in one of 45 specimens collected from Kansas and Oklahoma. Thus, clavicular fen- estrations, although uncommon, are occasionally present in C. collaris and it seems likely that addi- tional specimens will reveal their presence in C. nebrius and C. insularis as well. Clavicular fenes- trations are rare in the basal lineages of the outgroup taxa, being found only in Basiliscus, Laemanctus, some Corytophanes hernandezi( REE 1800, SDSNH 68090, although considered absent from this species by Lang, 1989), some Uma inornata, Ctenoble- pharys adspersus, some Leiolepis belliana, Physig- nathus concincinus, some P. lesueurii, and Enyalius brasiliensis. Therefore, the presence of clavicular fenestrations is considered to be the derived state. Interclavicle.— The interclavicle is an unpaired median element that lies along the ventral margin of the pectoral girdle. It varies extensively in form, although it usually is in the shape of an anchor or arrow. Lateral processes, present anteriorly, are in close contact with the proximal ends of the clavicles, while a long, narrow posterior process is bordered laterally by the epicoracoid cartilages and the ster- num. In most Crotaphytus and some Gambelia, the interclavicle expands laterally becoming diamond- shaped just anterior to the sternum. Although Wei- ner and Smith (1965) considered this character to be phylogenetically informative, there is continuous variation in this feature and it was not included in the phylogenetic analysis. Sternum and Xiphisterna. — The sternum is a me- dian, diamond-shaped element composed entirely of calcified cartilage. Anterolaterally, the sternum thickens, forming grooves into which fit the epicor- acoid cartilages. These tongue-in-groove joints al- low for extensive mobility of the pectoral girdle el- ements during locomotion (Jenkins and Goslow, 1983). The sternum also articulates medially with the posterior process of the interclavicle. In the cen- ter of the sternum there may be a fontanelle that, when present, is usually invaded by the mterclavicle. Posterolaterally, the sternum bears four or five fac- ets that serve as attachment points for the sternal and xiphisternal ribs and the postxiphisternal rods. The posteriormost facets (those that give rise to the xiphisternal ribs) are separated slightly more widely in eastern Crotaphytus collaris than in other crota- phytids. A similar, albeit more extreme, condition is observed in Sauromalus (de Queiroz, 1987). This may be related to the more depressed habitus of eastern C. collaris and their greater propensity for crevice dwelling. This condition was not coded as a character. Weiner and Smith (1965) noted that the sternum of Crotaphytus is broader and shorter than in Gambelia. Although there does appear to be a trend in this direction, this character appears to vary continuously and was not included in the phyloge- netic portion of this analysis. No phylogenetically informative variation was discovered in the ster- num (but see above section titled “Ribs” for dis- cussion of xiphisternal rib variation). Pelvic Girdle Illium and Pubis (Character 46). — In Gambelia, the iliac blades are robust and roughly cylindrical at their distal termini, while in Crotaphytus, they are usually laterally compressed. However, in some C. collaris (primarily those formerly referred to C. c. collaris ), they may approach the cylindrical con- dition observed in Gambelia. The outgroup taxa are extremely variable with respect to this character and it could not be polarized. Weiner and Smith (1965) discuss ventrolateral curvature of the pubes and the angle at which the two halves of the pelvic girdle meet. There does not appear to be consistent interspecific variation in ei- ther of these features (in fact, I am unaware of any ventrolateral curvature of the pubes, although they 32 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Fig. 17. — Ventral view of the fifth metatarsal bone of Crotaphytus collaris showing the contact of the medial and lateral plantar tubercles forming an arch (redrawn from Snyder, 1954). may be referring to ventromedial curvature). They may have been referring to the presence of a pro- portionally shorter and broader pelvic girdle in east- ern populations of Crotaphytus collaris (the only representative of the “collariform” group that they examined) than in other Crotaphytus species or Gambelia. This difference appears to be related, at least in part, to modification of the pubic rami, which are nearly transverse in orientation, rather than acutely angled anteriorly. However, the condition in the remaining populations of C. collaris (formerly referred to C. c. fuscus, C. c. baileyi, and C. c. au- riceps) appears to be intermediate in each of these features. Coding of this variation is further com- plicated by individual variation in pelvic girdle structure, such that some individuals approach the eastern C. collaris condition, while others approach the condition of other Crotaphytus species. Short, broad pelvic girdles are often observed in crevice- dwelling species (e.g., Sauromalus) and the rela- tively short, broad, pelvic girdles of eastern C. col- laris may be related to the crevice-dwelling behavior observed in these lizards. Limbs (Character 47; Fig. 17) On the plantar surface of the fifth metatarsal are two large tubercles termed the medial and lateral plantar tubercles by Robinson (1975). These tuber- cles serve as attachment points for the tendons of M. gastrocnemius. In the majority of iguanian spe- cies, a groove runs between the two tubercles and a tendon of M. flexor digitorum longus passes within it (Robinson, 1975). In Crotaphytus, the medial plantar tubercle usually curves laterally such that it contacts the lateral plantar tubercle forming a com- plete arch (Fig. 1 7), through which passes the tendon of M. flexor digitorum longus (noted and figured by Snyder, 1954). The contact of the tubercles is usually extensive and in some individuals, the tubercles may fuse completely. The arch condition was absent in the entire available series of Gambelia (41 speci- mens) and, in adults, it was always present in the 20 C. bicinctores, four C. antiquus, 1 2 C. dickerson- ae, five C. grismeri, and 22 C. vestigium examined. It was complete on at least one pes in 28 of 36 C. collaris, three of five C. insularis, 1 1 of 12 C. nebrius, and six of seven C. reticulatus. The majority of spec- imens that lacked the complete arch were juveniles, and in most cases the gap between the medial and lateral plantar tubercles was narrow. Therefore, this character was scored only for adults. Among the outgroup taxa examined, the arched form of the medial and lateral plantar tubercles was present only in the phrynosomatid sand lizards ( Uma , Callisau- rus, Cophosaurus, and Holbrookia). This feature ap- pears to represent a synapomorphy for Crotaphytus, as well as providing additional character support for the monophyly of the phrynosomatid sand lizards. The hindlimb of Crotaphytus is much longer than that of Gambelia of similar SVL. A relatively long hindlimb is typical of lizard species that utilize bi- pedal locomotion (Snyder, 1952, 1954, 1962), al- though agamines provide an interesting exception. Much of the variation in hindlimb length between Crotaphytus and Gambelia is realized in the longer crus of the former, while the pes appears to be of relatively similar length. Although a greater relative hindlimb length appears to be a derived character- istic of Crotaphytus, there is great variation in the outgroup taxa and this feature was not included in the phylogenetic analysis. Squamation The dorsal body squamation of Crotaphytus and Gambelia is remarkably similar in that both genera are characterized by relatively undifferentiated head scales and fine homogeneous dorsal body squama- tion. However, despite many similarities in scale patterns and scale sizes on the various regions of the body, phylogenetically useful variation in squa- mation exists. A more detailed description of the squamation of crotaphytids is provided in the tax- onomic accounts of the family, genera, and species. 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 33 Fig. 1 8. — Squamation of the dorsal portion of the head of Cro- taphytus collaris (USNM 17183, adult male). Scale = 5 mm. Rostral Scale (Character 48). — In all crotaphytids except Crotaphytus dickersonae, the rostral scale is approximately four times wider than high. In C. dickersonae, the rostral is less elongate and approx- imately two times wider than high. There is much variation in the outgroups, although most taxa have a rostral that is much wider than high. Consequent- ly, this character was left unpolarized. Supraorbital Semicircles (Character 49; Fig. 18, 19 ). — Crotaphytus have supraorbital semicircles composed of scales that are much larger than the adjacent supraoculars. In Gambelia, obvious supra- orbital semicircles are absent, with the supraoculars tending to grade into the frontal series. The out- groups vary considerably in the presence of discrete supraorbital semicircles. They are present in all oplurids and polychrotids examined (except Cha- maeleolis ), and variable within the remaining fam- ilies. Within Hoplocercidae, they are absent in En- yalioides laticeps, but present in E. praestabilis and E. oshaugnessyi. Within Phrynosomatidae, they are present in Petrosaurus and the Sceloporus group, Uma notata, U. scoparia, and U. inornata, but ab- sent in Phrynosoma and Uma exsul. In tropidurids, they are present in some Phymaturus patagonicus, Leiocephalus, Liolaemus, Stenocercini, basal Tro- pidurini (except Uranoscodon superciliosus ), and absent in Ctenoblepharys adspersus, most Phyma- Fig. 19.— Squamation of the dorsal portion of the head of Gam- belia wislizenii (SDSNH 68662, adult female). Scale = 5 mm. turns, and Uranoscodon superciliosus. In iguanids, they are present in Dipsosaurus, absent in Brachy- lophus fasciatus, and generally absent in the re- maining taxa. In chamaeleonids, they are absent in Hydrosaurus pustulatus, Leiolepis be/liana, Uro- mastyx loricatus, and U. ocellatus, variable in U. geyrii, U. microlepis, and U. acanthinurus, and pres- ent in U. aegypticus, U. asmussi, U. hardwickii, U. macfadyeni, U. philbyi, and U. thomasi. In coryto- phanids, they are present in Basiliscus plumifrons, B. vitattus, Corytophanes hernandezi, absent in C. cristatus and C. percarinatus, and variable in Lae- manctus. Because of this extensive variation, this character was left unpolarized. Suboculars (Character 50; Fig. 20, 21). — In Cro- taphytus, the suboculars are subquadrate, with the third scale occasionally larger than the others, whereas in Gambelia, the second subocular is four to five times larger than the others. Assessing po- larity of this feature is difficult because both states are widespread within the Iguania. At least one sub- ocular is much longer than the others in phryno- somatids except Phrynosoma, the oplurids Opiums 34 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Fig. 20. — Squamation of the lateral portion of the head of Cro- taphytus collaris (USNM 17183, adult male). Scale = 5 mm. saxicola, O. fierinensis, and O. quadrimaculatus, the chamaeleonid Leiolepis bel/iana, the tropidurids Phymaturus patagonicus, Leiocephalus, Liolaemus, Stenocercini, Microlophus, Plesiomicrolophus, and all but terminal Tropidurus (Frost, 1992), the igua- nid Dipsosaurus, and the polychrotids Anisolepis, Pristidactylus, and Eny alius bilineatus. Multiple subequal suboculars are present in the oplurids Opiums cyclurus, O. cuvieri, and Chalaradon, the chamaeleonids Uromastyx and Hydrosaurus pus- tulatus, hoplocercids, the tropidurids Phymaturus punae, P. palluma, Ctenoblepharys, and Uranos- codon superciliosus, the polychrotids Urostrophus, Polychrus, Phenacosaurus, Chamaeleolis, Anolis, and Enyalius (except E. bilineatus), iguanids (except Dipsosaurus ), and corytophanids. An elongate sub- ocular appears to be the ancestral state in Phryno- somatidae, Tropiduridae, and Opluridae, and equivocal in Iguanidae, and Polychrotidae. The presence of subequal suboculars is the ancestral state for Corytophanidae, Hoplocercidae, and Chamae- leonidae. Therefore, this character could not be po- larized. Terminal Supradigital Scales (Character 5 1). — In Gambelia, C. collaris, and C. reticulatus, the ter- minal supradigital scales nearly always lie flat against the dorsal surface of the claws. In the remaining Crotaphytus, the terminal supradigitals project dor- sally such that each is elevated from the claw. A similar elevated condition occurs occasionally in various iguanians including the phrynosomatids Pe- trosaurus, Uta stansburiana (three of four), U. pal- meri (one of four), and Uta squamata (one of three), the tropidurids Plesiomicrolophus koepkeorum (one of four), Microlophus grayi (one of four), M. ther- esioides (one of four), M. tigris (one of four), and M. stolzmanni (three of four), and the hoplocercid En- yalioides laticeps (one of five). Despite this varia- tion, the presence of elevated terminal supradigital scales is most parsimoniously considered to be the derived state. Femoral Pores (Characters 52, 53; Fig. 22, 23).— In Gambelia, the femoral pores extend distally at least to the inferior angle of the knee. The femoral pore series of G. silus usually just reaches this point, while the femoral pore series of G. wislizenii and G. copei almost always extend beyond and may even arch posteriorly onto the lower leg. The femoral pore series of Crotaphytus does not reach the inferior angle of the knee and usually terminates well prox- imal to this point. Polarization of this character is complicated by the absence of femoral pores in the Tropiduridae, Opluridae, and Corytophanidae. However, in the remaining outgroups, the femoral pore series always terminates before reaching the inferior angle of the knee ( Phrynosoma coronatum is variable with re- spect to this character). Therefore, the condition observed in Gambelia is interpreted as the derived state. In Gambelia wislizenii and G. copei, the femoral pores of females are large and contain substantial quantities of exudate, although the pores are usually slightly larger in males. In G. silus, Crotaphytus, and all of the outgroup taxa examined that have femoral pores except Enyalioides laticeps, they are much larger and fuller in males than in females and, in- deed, in females the pores may be devoid of exudate. Therefore, the condition observed in G. wislizenii and G. copei is considered to be the derived state. 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 35 Fig. 22.— Ventral view of Gambelia wislizenii (TNHC 33200) showing the femoral pore series extending beyond the angle of the knee. Fig. 23.— Ventral view of Crotaphytus reticulatus (TNHC 28364) showing the jet black femoral pores present in males. AGF = antegular fold, GF = gular fold. 36 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Fig. 24. — Ventral view of the neck folds of Crotaphytus reticulatus (EL 3250). Fold terminology follows Frost (1992). AGF = an- tegular fold, GF = gular fold. Postanal Scales. — In some iguanian lizards, males can be differentiated from females by the presence of enlarged postanal scales. Within Crotaphytidae, the postanal scales are enlarged in all male Gam- belia, as well as C. grismeri, Crotaphytus nebrius, and most C. bicinctores and C. collaris. The con- dition of the postanal scales is more variable in C. vestigium and C. insularis, with roughly equal pro- portions of males having large or only slightly en- larged scales. The postanal scales are not enlarged or are only slightly enlarged in C. antiquus, C. re- ticulatus, and C. dickersonae, although they may be larger than in females. Attempts to code this char- acter were prohibited by continuous variation in the size of the postanal scales in C. bicinctores, C. col- laris, C. dickersonae, C. insularis, C. reticulatus, and C. vestigium. Furthermore, this character could not be polarized as enlarged postanal scales are present in phrynosomatids, oplurids, many anoles (Cha- maeleolis chamaeleonides, Phenacosaurus, and most Anolis), and some Leiocephalus (although Pregill [1992] found that enlarged postanal scales were de- rived within the genus). Tail Skin (Character 54). — In all crotaphytids, the skin of the tail is relatively weakly adherent to the underlying musculature such that the skin can be removed easily. This condition contrasts strongly with that observed in most iguanians with fracture planes, such as Dipsosaurus, Sceloporus, and Opiu- ms, in which the skin is bound to the underlying musculature by connective tissue and is nearly im- possible to remove in one piece. This condition is more strongly developed in Crotaphytus than in Gambelia, such that in the former, the skin of the posterior 40-50 mm of the tail easily slips off. Loosely adherent skin that is easily removed from the ter- minal portion of the tail appears to be unusual if not unique among iguanians and is therefore con- sidered to be the derived state (1) in this analysis. Pockets and Folds Crotaphytids, like many fine-scaled iguanian liz- ards, have extensive lateral neck and gular folding. Both Crotaphytus and Gambelia share a standard complement of folds that includes gular, antegular, antehumeral, postauricular, longitudinal neck, and supra-auricular folds (terminology follows Frost, 1992). None of these folds are unique to Crotaphy- tidae and most are similar to folds present in a wide range of iguanian lizards. For example, the gular fold is well developed, enclosing a region of reduced squamation, and is continuous with the antehu- meral fold. Also, the antegular fold is continuous with the oblique neck fold. However, phylogeneti- cally informative variation does occur in the folds. As is the case with most fine-scaled species, addi- tional folds are often present with varying degrees of consistency. Thus, I have referred to the above- mentioned complement of folds as the standard pat- tern and will restrict the discussion to this series. Gular Fold (Character 55; Fig. 24-27). — The gular fold of Crotaphytus differs from that of Gambelia in that there is a pair of skin folds that separate from the gular fold and project posteromedially. These folds, which may be ventromedial continuations of the antehumeral folds (R. Etheridge, personal com- munication, 1993) usually meet midventrally and form a single longitudinally oriented midventral fold that extends posteriorly for a short distance. In the triangular-shaped region between the folds, the scales are reduced in size. In Gambelia, a pair of similar 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 37 Fig. 25.— Ventral view of the neck folds of Gambelia wislizenii (SDSNH 68663). Fold terminology follows Frost (1992). AGF = antegular fold, GF = gular fold. folds may occur; however, they are shorter and ap- pear near the lateral borders of the gular fold. As a result, the area of reduced squamation seen in Gam- belia takes the form of a uniform band that extends across the width of the gular fold. Those outgroup taxa with gular folds examined here display both conditions of the fold with phrynosomatids, the oplurid Chalaradon madagascariensis, chamae- leonids (except Hydrosaurus pustulatus ), the cory- tophanids Basiliscus vittatus and Laemanctus, the hoplocercid Enyalioides laticeps (four of five), and polychrotids displaying the Gambelia form, and the hoplocercids Enyalioides praestabilis and E. os- haugnessyi, the corytophanid Basiliscus p/umifrons, the oplurid genus Oplurus ( O . fierinensis and O. saxicolus variable), and the iguanid Dipsosaurus displaying the Crotaphytus form. Most tropidurids have incomplete gular folds or lack them altogether; thus, the evaluation of this character for Tropidur- idae is difficult. Uranoscodon super ciliosus, which has a complete gular fold, displays the Gambelia form. The only other species within Tropidurini with Longitudinal neck Fig. 26. — Lateral view of the neck folds of Crotaphytus reticulatus (EL 3250). Fold terminology follows Frost (1992). complete gular folds are Tropidurus azureutn, T. flaviceps, and T. plica (Frost, 1992), species far re- moved from the basal lineages of the clade, and, thus, unable to shed light on this polarity decision. Because of ambiguity in the outgroup taxa, this char- acter was left unpolarized. Supra- auricular Fold (Character 56; Fig. 26, 27).— Frost (1992) defined the supra-auricular fold as a continuation of the dorsolateral fold that passes above the tympanum. In crotaphytids, a similar fold is present; however, it originates from the postaur- icular fold at a point roughly midway between the dorsal and ventral borders of the external auditory meatus. Without strong evidence to the contrary, I treat the crotaphytid fold as homologous with that described by Frost (1992) and therefore apply his standardized nomenclature. The condition of the supra-auricular fold, in which it originates midway between dorsal and ventral borders of the external auditory meatus, is present in many iguanian taxa Longitudinal neck Fig. 27. -Lateral view of the neck folds of Gambelia wislizenii (SDSNH 68663). Fold terminology follows Frost (1992). 38 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Fig. 28. — An antehumeral mite pocket in a juvenile Crotaphytus grismeri. and, therefore, could not be included in the phy- logenetic analysis. The supra-auricular fold differs between Crota- phytus and Gambelia. In Crotaphytus, the fold ex- tends posterodorsally at an angle of roughly 45 de- grees. In Gambelia, the fold extends posteriorly along a horizontal plane. In most of the outgroup taxa that have a supra-auricular fold, the fold either projects posteriorly along a horizontal axis, or occasionally, posteroventrally. However, some taxa may have a Crotaphytus- like supra-auricular fold (often vari- ably), including the phrynosomatids Petrosaurus re- pens, Uta stansburiana, U. squamata, U. palmeri, Urosaurus auriculatus, and Phrynosoma coronatum, the tropidurids Leiocephalus schreibersi, L. melan- ochlorus, and L. psammodromus, the hoplocercid Enyalioides oshaughnessyi, and the chamaeleonids Uromastyx acanthinurus and U. philbyi. Because of this variation and because many outgroup taxa can- not be scored for this feature, this character was left unpolarized. Antehumeral Fold (Fig. 26-28).— The antehu- meral fold of crotaphytids is strongly developed, curving posteriorly over the forelimb insertion. The deepest portion of the fold is directly dorsal to the forelimb, a condition rarely observed in the out- groups. Furthermore, the antehumeral fold often ex- tends posteriorly beyond the forelimb insertion, then continues posteroventrally or ventrally forming a complete arc. This condition is again uncommon in the outgroups. However, there is sufficient variation within Iguania that I have chosen not to code this as a character. The antehumeral fold of Crotaphytus dickersonae is unique among crotaphytids in ter- minating anterior to the forelimb insertion. Al- though this condition is probably derived within Crotaphytidae, another character, presence or ab- sence of an antehumeral mite pocket, is certainly not independent. Therefore, this character is con- sidered under the section dealing with the antehu- meral mite pocket. Antehumeral Mite Pocket (Character 57; Fig. 28). — In all Crotaphytus except C. dickersonae, the antehumeral fold is well developed (deep), with an area of reduced squamation dorsal to the forelimb insertion. The pocket almost always is inhabited by large numbers of trombiculid mite larvae. The pres- ence of a mite pocket in this portion of the ante- 1996 McGUIRE- SYSTEMATICS OF CROTAPHYTID LIZARDS 39 Fig. 29.— A postfemoral mite pocket in a juvenile Crotaphytus bicinctores. humeral fold was not observed in any of the out- group taxa examined and, thus, appears to be unique to Crotaphytus, excluding C. dickersonae. As dis- cussed above, the antehumeral fold of C. dicker- sonae terminates further anteriorly than in any other crotaphytid, usually failing to reach the forelimb insertion, which probably explains the absence of an antehumeral mite pocket in this species. Postfemoral Mite Pockets (Character 58; Fig. 29). — In most crotaphytids, subdermal mite pockets are present at the posterodorsal border of the hind- limb insertion where a patch of finely scaled or un- sealed skin dips inward between M. iliofibularis and M. iliofemoralis. These pockets usually are inhab- ited by trombiculid mite larvae and occasionally ticks. Arnold (1986) noted that mite pockets, which may occur in a variety of anatomical regions, often vary both intra- and interspecifically in terms of their presence, degree of development (e.g., depth), and in the nature of their squamation, and in this respect Crotaphytidae is no exception. However, pockets were absent only in Crotaphytus reticulatus and occasionally in C. col laris and C. nebrius. In Crotaphytus, the depth of the mite pocket may be correlated with the degree of development of the antehumeral mite fold. For example, in C. reticu- latus, which lacks the postfemoral pocket, the mite pockets of the antehumeral fold (discussed above) are strongly developed. In contrast, the mite pockets of the antehumeral fold are absent in C. dickersonae, while the postfemoral pockets are the most strongly developed (deepest) of all Crotaphytus. Postfemoral mite pockets are not unique to Cro- taphytidae. Smith (1939) noted that they are present in seven species of Sceloporus, including the five species in his S. variabi/is group, as well as S', ma- culosus and S. gadoviae. Shallow postfemoral pock- ets were also observed in Uta squamata and U. pal- med, but not other Uta. Although not observed here, shallow mite pockets are occasionally present in sev- eral species of Urosaurus (J. Wiens, personal com- munication, 1994). However, the absence of post- femoral pockets in Phrynosoma, the sand lizards, Petrosaurus, most Uta (in those species that lack pockets, mites may accumulate in the postfemoral region, but an obvious subdermal pocket is lacking), most Urosaurus, Sator, as well as most Sceloporus, suggests that the pockets observed in subsets of Uro- 40 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 saurus, Uta, and Sceloporus are not homologous with crotaphytid postfemoral pockets. Most Stenocercus and at least two species for- merly referred to Ophryoessoides ( S . ornatus and S'. trachycephalus ) have postfemoral mite pockets (Fritts, 1974; Arnold, 1986). However, the postfe- moral pocket of those Stenocercus species examined here ( S . trachycephalus, S. chrysopygus, S. guenth- eri, S. imitator, S. roseiventris) occurs as a vertical fold along the lateral body wall immediately pos- terior to the hmdlimb insertion and, thus, does not appear to be homologous with the postfemoral mite pocket of crotaphytids. Furthermore, postfemoral mite pockets appear to be absent from the basal lineages of Liolaeminae ( Phymaturus and Ctenob- lepharys: species examined include Ctenoblepharys adspersus, Phymaturus sp., P. palluma, P. patagon- icus, P. punae), Leiocephalinae (G. Pregill, personal communication, 1993; verified in Leiocephalus car- inatus, L. inaguae, L. macropus, L. melanochlorus, L. pratensis [folds present, but no reduction in squa- mation], L. psammodromus, L. schreibersi), and Tropidurini ( Uranoscodon superciliosus, Plesiomi- crolophus koepkeorum, Microlophus theresioides, M. tigris, M. stolzmani, personal observation). Thus, the postfemoral mite pockets of certain members of the Stenocercini are considered to be nonhomolo- gous with crotaphytid postfemoral mite pockets. Several oplurids have postfemoral mite pockets that appear to be structurally identical with those of crotaphytids. That is, the pocket occurs as an invagination between M. iliofibularis and M. iliofe- moralis. Arnold (1986) noted the presence of post- femoral mite pockets in Opiums cuvieri and O. cy- clurus and I have observed them in O. cyclurus, as well as in O. saxicola, O.ferinensis, and Chalaradon madagascariensis. Postfemoral mite pockets appear to be absent in O. quadrimaculatus. Because we have no hypothesis of phylogenetic relationships for oplurids, it is not possible to say whether the pockets are derived within the family or were present an- cestrally. Therefore, the possibility that postfemoral mite pockets were present in the common ancestor of Opluridae cannot be discounted. Among iguanids, Dipsosaurus dorsalis has a weak- ly developed postfemoral pocket that occurs in the same anatomical position as the postfemoral mite pocket of crotaphytids. Because Dipsosaurus (along with the fossil species Armandisaurus exploratory is the sister taxon of the remaining iguanids (de Queiroz, 1987; Norell and de Queiroz, 1991), the possibility that postfemoral pockets were present in the common ancestor of Iguanidae cannot be elim- inated. Postfemoral mite pockets appear to be absent from Corytophanidae, Hoplocercidae, Chamaeleonidae, and Polychrotidae, although all of their constituent species have not been examined. Although postfe- moral mite pockets may have been present in the common ancestors of the families Opluridae and Iguanidae, their presence is most parsimoniously treated as the derived state for Crotaphytidae. Additional Morphological Characters Hemipenes (Character 59). — Hemipenes were ex- amined for all of the crotaphytid species except Cro- taphytus reticulatus. The hemipenes of crotaphytids are bulbous and weakly bilobed with a short median fissure separating the two lobes apically. The sulcus spermaticus is covered by a large fleshy flap of in- tegument that folds over it from its lateral margin. This fold does not project directly toward the apex but rather extends laterally toward the outer margin of the lateral lobe. The sulcus spermaticus itself ap- pears to terminate in a broad, shallow depression at the base of the lobes. The entire sulcate surface of the hemipenis is only weakly ornamented with a fine papillate or dimpled texture. Immediately outside of the sulcus sper- maticus, the surface is ornamented with plicae that are continuous with those of the asulcate surface. Distally, the lateral surfaces of the lobes bear small knob-like processes that are covered with extremely fine calyculae. The base of the asulcate surface of the hemipenis is naked. More distally, ornamentation is present in the form of plicae and calyculae. The proximal lat- eral surfaces of the ornamented region of the hem- ipenis are covered by fine plicae. These plicae grade medially into calyculae and this calyculate zone ex- tends distally toward the apex of the hemipenis where it spreads laterally. As a result, the entire surface of the hemipenis distal to the median apical fissure is ornamented with minute calyces. The lateral surface of each lobe bears a shallow depression ornamented with extremely fine calyces. The calyces reach their smallest sizes here and in the apical region of the hemipenis. The only obvious difference between the hemi- penes of Crotaphytus and those of Gambelia is in their relative size. The hemipenes of Gambelia are roughly twice the size of those of similar-sized Cro- taphytus. Although the hemipenes of Crotaphytus 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 41 Fig. 30. — (A) Gambelia wislizenii (adult female), (B) G. copei (adult male), (C) Crotaphytus reticulatus (adult male), (D) C. antiquus (adult male) 42 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 appear to be unusually small, no attempt was made to polarize this character because adequate com- parative material was not available. Sexual Dimorphism (Character 60). — Most igu- anian lizards are sexually dimorphic with males reaching larger SVLs than females (Fitch, 1981). This condition is exhibited in all Crotaphytus (Burt, 1929; Axtell, 1972; Fitch, 1981; McGuire, 1994; personal observation) as well as Gambelia silus (Tol- lestrup, 1979, 1982), while females are much larger than males in G. wislizenii (Tollestrup, 1979, 1982) and G. copei (Banta and Tanner, 1968). Sexual di- morphism in which males are larger than females appears to be the ancestral state for the families Chamaeleonidae (Parcher, 1974; Fitch, 1981), Igua- nidae (Fitch, 198 1; Gibbons, 198 1; Carothers, 1984), Opluridae (Blanc and Carpenter, 1969), Phrynoso- matidae (Fitch, 1981), and Tropiduridae (Dixon and Wright, 1975; Fitch, 1981; Cadle, 1991; Etheridge, 1992, 1994, 1995; Pregill, 1992; R. Etheridge, per- sonal communication, 1994). The ancestral state is equivocal for Hoplocercidae (Duellman, 1978), Corytophanidae (Fitch, 1981), and Polychrotidae (Lazell, 1969; Fitch, 1981; Frost and Etheridge, 1989; Etheridge and Williams, 1991; Schwartz and Hen- derson, 1991). Although the data regarding sexual dimorphism in iguanians are somewhat fragmen- tary, the most parsimonious conclusion at this time is that the ancestral condition for Crotaphytidae is males larger than females. Therefore the character state present in Gambelia copei and G. wislizenii (females larger than males) is treated as the derived state. Coloration Gravid and Subadult Coloration (Characters 61, 62; Fig. 3 1C, D). — All female crotaphytids display red or orange dorsal banding or spotting when grav- id. Although Frost and Etheridge (1989) suggested that gravid coloration may be a synapomorphy for the family, the presence of gravid coloration in many phrynosomatids and tropidurids and several cha- maeleonids (Cooper and Greenberg, 1992; personal observation) suggests that this condition may rep- resent a synapomorphy for a more inclusive group than Crotaphytidae. Subadult male Crotaphytus collaris develop a col- or pattern of red or orange dorsal banding that is very similar to that of gravid females, both in terms of its anatomical position and chromatic qualities of the pigments (Rand, 1986). The author has also observed this coloration in C. bicinctores, C. dick- er so nae, C. grismeri, C. insular is, C. nebrius, C. re- ticulatus, and C. vestigium. Rand (1986) demon- strated that the subadult male coloration of C. col- laris is not induced by progesterone, as it is in fe- males, which suggests that subadult male and gravid female coloration are independent. The presence in subadult males (but not subadult females) of orange or red banding similar to that of gravid females appears to be unique to Crotaphytus. The only spe- cies (that I am aware of) that exhibits a similar sub- adult coloration is Microlophus delanonis (Werner, 1978). This species has gravid coloration and ju- veniles of both sexes develop coloration similar to that of gravid females. Therefore, the presence of ephemeral red banding in subadult males is treated as the derived state. Juvenile Gambelia are characterized by the pres- ence of paravertebrally arranged rows of blood-red spots that extend from the top of the head to the proximal portion of the tail and may be present on the limbs as well. Each row generally consists of four large spots, although smaller spots may be present further laterally. These blood-red spots gradually fade into solid brown spotting in adult Gambelia. This condition, which was not observed in the out- groups, is coded as a character independent of the subadult male coloration character described for Crotaphytus because it does not occur in the same anatomical position and because it occurs in both sexes. Tail Color (Characters 63-65; Fig. 3 IB, 3 1C, 32A- D; observable only in live individuals). — Adult Cro- taphytus dickersonae females exhibit a unique fea- ture among crotaphytids in that the hindlimbs and in particular the tail may be bright lemon yellow in comparison to other species in which the tail is the same general color as the rest of the body. This description is based on a sample of only two living females. An examination of preserved specimens suggests that many adult female C. dickersonae have a substantial blue component to their color pattern and, thus, the yellow pigmentation may be restricted to a particular season or age class. Because this type of yellow pigmentation in adult females was not observed in the outgroups, I consider it the derived state. However, bright coloration often fades in pre- servative and it is possible that this character state has been overlooked in other taxa. Gambelia situs juveniles have yellow pigmenta- tion in the form of a narrow strip along the posterior surface of the thigh and on the anteroventral surface of the tail. The pigmentation ends abruptly at the cloaca. Similar coloration was present in the only subadult female C. antiquus that was observed and 1996 McGUIRE— SYSTEM ATICS OF CROTAPHYTID LIZARDS 43 this taxon is tentatively coded as having the same character state as that observed in G. silus. In other crotaphytids, the coloration of the tail and hin- dlimbs does not differ from that of the rest of the body. The presence of this juvenile coloration is treated as the derived state. In those species with strongly laterally com- pressed tails (C. bicinctores, C. dickersonae, C. gris- meri, C. insularis, and C. vestigium), a pale white or cream stripe runs down the dorsal surface of the tail (Fig. 3 IB, 32A-D). Presumably, the laterally compressed tail serves a display function and this white pattern may somehow enhance this role. The presence of a pale dorsal caudal stripe appears to be unique to these lizards as it was not observed in any of the outgroup taxa and is therefore considered to be the derived state. Reticulate Pattern (Characters 66, 67; Fig. 30C, 30D, 31A-D, 32A-D, 33-35). — All male Crotaphy- tus, except some C. nebrius, have some form of white reticulation in the dorsal and/or gular pattern. Indeed, all Crotaphytus neonates have an extensive reticulated dorsal pattern, with some of the reticu- lations surrounding black pigment. This is a con- dition very similar to that seen in adult C. reticulatus and C. antiquus of both sexes. The extent and place- ment of the reticulated pattern varies considerably between species resulting in somewhat bewildering interspecific variation. Nevertheless, a pair of dis- crete characters were obtained from this aspect of the color pattern. The first character (66) describes the presence or absence of a reticulate pattern in neonates. This con- dition is present in all Crotaphytus neonates, and is absent from Gambelia and the outgroups (although the number of outgroup species for which juveniles were examined is relatively small). Therefore, the presence of a neonatal pattern of white reticulations enclosing dark pigments is treated as the derived state. A second character (67) is the presence of small, almost granular, reticulations on the ventrolateral surface of the abdomen. This condition is present only in C. bicinctores and C. antiquus, although the abdominal reticulations of C. antiquus are slightly larger than those of C. bicinctores. Ventrolateral ab- dominal reticulations were not observed in the out- group taxa; therefore, their presence is treated as the derived state. In Crotaphytus, there are two common dorsal pat- tern types, reticulation and spotting. It seems likely that spots are formed when reticulations have be- come fragmented. For example, in large C. vestig- ium, the typical reticulated pattern of the hindlimbs may be fragmented on the dorsal portion of the femoral region, resulting in spots. The anterior and posterior surfaces of the leg retain their reticulated pattern. Thus, the spotted pattern that occurs on the dorsum of all Crotaphytus except C. reticulatus and C. antiquus may be the derived condition. This same situation applies to additional characters associated with reticulation. However, the dangers of polar- izing characters using ontogenetic methods are well known (de Queiroz, 1985; Mabee, 1989, 1993) and I present this scenario as a hypothesis and nothing more. The reticulated versus spotted adult dorsal body patterns are considered in the discussion of the white component of the dorsal pattern (see be- low). White Component of Dorsal Pattern (Character 68; Fig. 30-32). — The white component of the dor- sal pattern of crotaphytids is quite variable between species, but within species there is little variation. The two main dorsal pattern types present in adult Crotaphytus are reticulated and spotted. Crotaphy- tus antiquus and C. reticulatus exhibit the reticulated pattern, while the remaining species of Crotaphytus have a pattern that incorporates white spots or dash- es. Crotaphytus vestigium and C. insularis (see be- low) each differ from the other spotted species in their own way. Crotaphytus vestigium has thin, white, transverse dorsal bands (Fig. 32C). Axtell (1972) noted the presence of similar banding in C. bicinc- tores from the northern portion of its range, which he attributed to the retention of the juvenile pattern. However, an examination of approximately 300 specimens of C. bicinctores in the California Acad- emy of Sciences collection revealed that the white bands present in juveniles change during ontogeny into the broad, pale orange bands characteristic of adults. In adults that are dark from preservative, these orange bands fade and sometimes appear to be broad white bands. Although females may oc- casionally retain the juvenile white bands until near adult size is attained, males do not and no adult C. bicinctores that were not dark from preservative had white transverse banding. Only C. insularis ap- proaches the condition of C. vestigium, with most specimens having broad, wavy dorsal lines or dashes and a few specimens having what appear to be wavy transverse dorsal bands (Fig. 32D). Although the wavy dashes present in C. insularis may be modified transverse dorsal bands, the C. insularis condition is treated as a separate character state and no a priori assumptions were made regarding the order of trans- formation. Because Crotaphytus and Gambelia are 44 NO. 32 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY with orange subadult male coloration). 1996 McGUIRE — SYSTEMATICS OF CROTAPHYTID LIZARDS 45 Fig. 32. — (A) Crotaphytus bicinctores (adult male), (B) C. grismeri (adult male), (C) C. vestigium (adult male), (D) C. insularis (adult male). 46 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 variable with respect to the white component of the dorsum, this variation was coded as an unordered multistate character. The Gambelia condition often consists of broad, white or cream-colored, offsetting transverse bars with large, brown dorsal spots and is coded as state 0; the C. reticulatus and C. antiquus condition of a white reticulum, some or all of which enclose black pigmentation, is coded as state 1; the pattern composed of numerous small white spots (present in C. bicinctores, C. collaris, C. dickersonae, C. grismeri, and C. nebrius ) is coded as state 2; the C. vestigium condition of white, transverse dorsal bands on a background of white spots and dashes is coded as state 3, and the C. insularis condition of wavy, white dorsal dashes is coded as state 4. This character was left unpolarized. Sexual Dichromatism (Character 69; Fig. 3 IB, C). — Sexual dichromatism is widespread within the Iguania (Cooper and Greenberg, 1992) and, thus, it is not surprising that most crotaphytids also display strong sexual dichromatism. However, Gambelia and Crotaphytus reticulatus generally lack sexual di- chromatism in their permanent dorsal patterns (al- though G. silus and C. reticulatus do have male breeding coloration). There is obvious sexual di- chromatism in the gular pattern and femoral pore coloration and a small amount of sexual variation in the collar of C. reticulatus. However, the re- maining species of Crotaphytus have much more obvious sexual dichromatism throughout the year, with males differing from females in most aspects of dorsal coloration (e.g., much more vibrant blue, green, and/or yellow dorsal coloration in C. collaris ), as well as in the gular pattern. Although sexual di- chromatism is present in many iguanian taxa, data could not be obtained for many of the more obscure and poorly known species. Therefore, this character was left unpolarized. Paired Melanie Keels on Scales of Ventral Caudal Extremity (Character 70). — All Crotaphytus species except C. reticulatus (50 specimens examined) and C. insularis (23 specimens examined) are charac- terized by the presence, in at least some individuals, of darkly pigmented obtuse keels on the scales of the ventral surface of the tail tip (noted as present in C. nebrius and some C. collaris by Axtell and Montanucci, 1977). These take the appearance of paired dark spots that may extend along the ventral surface of the tail over the distal 2-30 mm. This feature is fixed in some species, polymorphic in oth- ers, and the percentage of individuals with the pig- mented keels may vary extensively between popu- lations of the same species. Crotaphytus collaris is polymorphic with respect to this character and there is much geographic vari- ation in the percentage of individuals with the paired pigmented scales. Individuals from regions of Mex- ico generally referred to the subspecies C. c. fuscus and C. c. baileyi usually possess this character (21 of 33 specimens examined). It is less often present (six of 23) in specimens from midwestern and south- ern United States (generally referred to the subspe- cies C. c. collaris ). It was absent in all specimens of C. collaris examined from Arizona, eastern Utah, and western Colorado (generally referred to the sub- species C. c. baileyi and C. c. auriceps, n = 38). Although the percentage of individuals with pig- mented keels varies regionally, the observed fre- quency for C. collaris (29 of 94) was employed in the phylogenetic analysis. In Crotaphytus nebrius, this characteristic appears to be nearly fixed. The pigmented scales were ob- served in 48 of 49 specimens examined. The only specimen that lacked the scales (KU 121460) was from the Tucson Mountains, an isolated range in- habited by what may be a distinct species. Unfor- tunately, this is one of only two preserved specimens available from the Tucson Mountains (the other specimen, SDSNH 15208, had pigmented scales). The pigmented scales are much darker, and thus more obvious, in C. nebrius than in C. collaris. Crotaphytus bicinctores is another species in which this characteristic is polymorphic. It was present in 37 of 79 specimens examined. However, the per- centage of individuals with the pigmented scales varied considerably between populations. Speci- mens from southern populations (Palo Verde Mountains, California; Chocolate Mountains, Cal- ifornia; Kofa Mountains, Arizona; Sentinel, Ari- zona) have the scales in high frequency (26 of 32), while specimens from more northern populations (Idaho; Inyo County, California; Washington Coun- ty, Utah) usually lack them (present in three of 30 specimens examined). The pigmented scales appear to be fixed in Cro- taphytus dickersonae (present in 44 of 44 specimens examined), C. grismeri (present in ten of ten spec- imens examined), C. vestigium (present in 43 of 43 specimens examined), and in C. antiquus (present in 17 of 17 specimens examined). The presence of paired melanic keels on the distal caudal extremity is considered to be the derived state as they appear to be unique to Crotaphytus. Black Oral Mucosa (Character 71). — In all cro- taphytids except Crotaphytus bicinctores, C. gris- meri, C. insularis, and C. vestigium, black pigments 1996 McGUIRE— SYSTEM ATICS OF CROTAPHYTID LIZARDS 47 Fig. 33.— Ventral view of a series of Crotaphytus col laris. are deposited in the oral mucosa and at least some of the underlying fascia of the M. adductor man- dibulae complex. There is interspecific variation in the extent of the pigmentation as well. In Gambelia, C. collaris, C. nebrius, and C. reticulatus, the cov- erage and density of the oral melanin is extensive. The pigments are present on the floor of the buccal cavity as well as on the fauces of the roof of the cavity. In C. antiquus and C. dickersonae, black oral melanin is present but it is less extensive in both coverage and density. Stebbins (1954) noted that G. wis/izenii from the Painted Desert region of Arizona may lack this coloration. However, this observation has not been confirmed in the present study and Stebbins himself (personal communication, 1991) does not recall where he obtained this information. A black oral mucosa appears to be absent from all basal outgroup taxa outside of the family Poly- chrotidae (the throat lining is deep violet in Tro- pidurus umbra, Etheridge, 1970). Within Polychro- tidae, black oral melanin is present in some Poly- chrus(P. marmoratus, P. acutirostris), Pristidactylus volcanensis, Leiosaurus catamarcensis, Urostrophus vautieri (Etheridge and Williams, 1991), Anisolepis grilli (Etheridge and Williams, 1991), Phenacosau- rus heterodermis, and all three species of Chamae- leolis (Schwartz and Henderson, 1991). It is variably present in Pristidactylus torquatus. The absence of black oral melanin has been verified in Polvchrus liogaster, P. guttarosus, Pristidactylus acha/ensis, P. scapulatus, P. casuhatiensis, Leiosaurus belli, U. gal- lardoi, Enyalius bilineatus, E. brazi/ienesis, E. ca- tenatus, E. iheringii, E. perditus, and E. pictus. Al- though black oral melanin may prove to be the an- cestral condition for Polychrotidae, the family does not appear to be the sister taxon of Crotaphytidae (Frost and Etheridge, 1989). Therefore, the presence of black oral melanin is treated as the derived state. Collars (Characters 72-75; Fig. 30-36). — Al- though all Crotaphytus species are characterized by the presence of at least one collar, there is consid- erable interspecific variation. Most species have two well-developed black collars, with relatively thick white borders that encircle or partially encircle them. The collar configurations of Crotaphytus reticulatus and C. antiquus suggest that the transversely ar- ranged series of black spots (each of which is bor- dered with white) present in these species may have 48 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Fig. 34. — Ventral view of an adult male Crotaphytus nebrius. been the precursor to the black collars outlined in white that are found in all Crotaphytus species. This is especially evident in the posterior collar markings, which in C. reticulatus are usually little more than a few closely approximating black spots with white borders. Furthermore, in most individuals there are dark pigments bleeding into the intervening areas between the black spots. A similar situation is some- times present in the anterior collar as well. With respect to the outgroup taxa, it is unlikely that a white-bordered collar or pair of collars is the an- cestral state in all but Opluridae (collars present in O. cuvieri and O. cyclurus ). Therefore, the presence of white-bordered collars is treated as the derived state. Additional variation occurs in C. bicinctores, C. antiquus, C. collaris, C. dickersonae, C. grismeri, and C. nebrius, where the posterior collars are either in contact or only narrowly separated at their medial margins (Fig. 30-32, 36). In C. insularis and C. ves- tigium, the posterior collars are broadly separated (Fig. 32C, D). The condition in C. reticulatus is more difficult to interpret because of the weak develop- ment of the posterior collar and it is tentatively coded as widely separated. Because Gambelia, the nearest outgroup to Crotaphytus, lacks collars, this character was left unpolarized. In all adult male Crotaphytus except C. collaris, the anterior collar is complete ventrally by way of dark brown or black pigmentation within the trans- verse gular fold (Fig. 33-35). Because the nearest outgroup taxa lack collars, this character was left unpolarized. In Crotaphytus collaris and C. nebrius, the pos- terior collar passes through the antehumeral fold before reaching the proximal dorsal surface of the brachium. A less developed condition usually oc- curs in C. reticulatus, where the collar passes through the extensive antehumeral mite pocket and isolated black patches may extend a short distance onto the proximal dorsal surface of the brachium. In C. an- tiquus, the posterior collar marking of males either terminates at the forelimb insertion or melanic spots extend onto the brachium, while in females, the collar marking generally terminates before entering the antehumeral fold (although in one individual [MZFC 6755], the marking seems to continue through much of the underlying mite pocket). In C. dickersonae, the posterior collar just reaches the dor- sal surface of the forelimb insertion and may extend slightly onto the brachium as in C. reticulatus. How- ever, the collar marking does not pass through the antehumeral fold in this species because the ante- humeral fold terminates anterior to the posterior 1996 McGUIRE— SYSTEM ATICS OF CROTAPHYTID LIZARDS 49 Fig. 35. — Ventral view of a series of adult male Crotaphytus vestigium. collar. In C. bicinctores and C. grismeri, the poste- rior collar terminates within the antehumeral fold. In C. vestigium and C. insularis, the posterior collar almost always terminates before reaching the an- tehumeral fold. The extreme situation exists in C. insularis where, in the few individuals that have a posterior collar, it terminates just before reaching the antehumeral fold. This character is less consis- tent in females, especially with respect to C. reti- culatus, in which females either lack collars or have them poorly developed. The four conditions de- scribed above were coded as separate character states of an unordered multistate character (state 0 = collar extends well out onto dorsal surface of brachium, state 1 = collar just reaches forelimb insertion, state 2 = collar terminates within antehumeral fold, state 3 = collar terminates before entering antehumeral fold). Again, because the nearest outgroups lack col- lars, this character was left unpolarized. As stated above, all Crotaphytus species are char- acterized by the presence of at least one collar (but see C. insularis below). In fact, with few exceptions, all Crotaphytus species except C. insularis and fe- male C. reticulatus have two collars. Crotaphytus insularis almost always have only the anterior collar, the posterior collar having apparently been lost (Fig. 32D). The fact that five specimens (CAS 21948, 50879, 86754, 148652; SDSNH 53064) have an ex- tremely reduced, but visible, posterior collar is con- sistent with the hypothesis that collar reduction has occurred in this species. Males have a more densely pigmented anterior collar than females, which sometimes have no collar at all. This reduction in both the posterior and anterior collars appears to be derived and hence an autapomorphy for this in- sular species. In C. reticulatus females, the anterior collar marking may be lacking while the posterior collars remain. However, the posterior collar mark- ing in both sexes of this species is often little more than a slightly modified band of black-filled retic- ulations. This variation was not included in the phy- logenetic analysis because of the potential problem of lack of independence between this state and the wide separation of the posterior collars described above. Dark Nuchal Spots (Character 76; Fig. 36). — A pair of black or dark spots usually occurs between the dorsal extensions of the anterior collar markings in Crotaphytus reticulatus (39 of 51), C. antiquus (16 of 16), and C. collaris (58 of 75), and are oc- 50 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 casionally present in C. nebrius ( 1 2 of 5 1 ). The spots, which occur between the dorsal extensions of the anterior collar markings, appear to be homologous with the first transversely arranged row of black- filled hexagonal reticulations seen in C. reticulatus. In C. antiquus, the nuchal spots are always present, but often incompletely separated from the remain- der of the anterior collar markings. Black nuchal spots are not present in the outgroup taxa and their presence is coded as the derived state. Inguinal Patches (Characters 77, 78; Fig. 32C, 34, 35). — In several species of Crotaphytus, adult males develop dark brown or black ventral patches in the inguinal region. These patches vary considerably in size with C. bicinctores, C. dickersonae, C. grismeri, C. insularis, and C. vestigium having large patches and C. antiquus, C. nebrius, and C. collaris having smaller ones. All adult male C. bicinctores, C. an- tiquus, C. dickersonae, C. grismeri, C. insularis, C. nebrius, and C. vestigium develop these patches while only some C. collaris have them. Interestingly, only C. collaris from the western periphery of its range (in the area usually referred to the subspecies C. c. baileyi ) are known to have inguinal patches. Thus, there are at least two characters associated with in- guinal patches: size of the patches and the frequency with which they occur. Homology of the patches seems likely. Both large and small patches begin development as small ventral spots near the hind limb insertion and the large patches differ only in that they continue to become larger (and probably grow faster). Inguinal patches of the type present in some Crotaphytus are extremely rare in the outgroup taxa. Similar markings are present in Uma exsul and U. paraphygas (de Queiroz, 1989; although they oc- cur more laterally than in Crotaphytus), Uta nolas- censis, Uromastyx hardwickii (concentrated on the thigh), and Enyalius iheringii (again, more laterally oriented). This character has been coded two ways: first, as a binary character with the absence of in- guinal patches (of any size) as state 0 and the pres- ence of patches as state 1 ; and secondly, as a separate binary character with the presence of small patches as state 0 and the presence of large patches as state 1. Taxa without inguinal patches were scored as unknown (“?”) for this second character. Because the first character (77) considers the frequency in which patches are present, the second character (78) does not take frequency into consideration. For character 78, the presence of small patches in any frequency is assigned state 0 and the presence of large patches in any frequency is coded as state 1. Femoral Pore Secretions (Character 79; Fig. 22, 23, 33-35).— The femoral pore secretions of male Crotaphytus reticulatus and C. antiquus are jet black. Unlike other Crotaphytus species, such as C. ne- brius, which often have grayish secretions, the sub- cutaneous glands themselves are also jet black. This condition was not observed in other species of ig- uanian lizards and is treated as the derived state. Gular Pattern (Characters 80-82; Fig. 33—35). — There is much variation in the gular pattern of male Crotaphytus, especially in the wide-ranging species C. collaris. However, the general arrangement of the gular colors is similar in all of the species. For ex- ample, each has a relatively uniformly colored cen- tral gular region that is surrounded by a peripheral reticulated or spotted pattern superficial to the man- dibles. It is in the context of this general pattern that the following discussion of variation is based. Be- cause the pattern and extent of the gular coloration is sexually dichromatic, the following discussion pertains only to adult male Crotaphytus. Adult male Crotaphytus bicinctores, C. antiquus, C. dickersonae, C. grismeri, C. insularis, C. reticu- latus, and C. vestigium (Fig. 35) have a patch of black pigment in the posteromedial portion of the gular region. This pigmentation corresponds with that portion of the gular pouch that is depressed by the second ceratobranchials of the hyoid apparatus, and thus presumably increases the visibility of the depressed gular pouch during aggressive display. The black patch is continuous with the black pigmen- tation of the gular fold and the ontogenetic devel- opment of the gular patch suggests that it may be an extension of the gular fold coloration. However, the presence of black pigmentation in the gular fold and in the posteromedial portion of the central gular region are treated as separate characters because the presence of black pigmentation in the gular fold is not always associated with a black central gular patch (e.g., C. nebrius). Because the outgroups do not have a gular pattern that is similar to Crotaphytus, out- group analysis cannot be utilized to assess polarity. Therefore, this character was left unpolarized. There is considerable variation in the peripheral gular pattern of Crotaphytus as well. In all Crota- phytus except C. nebrius, the peripheral gular pattern is composed of white reticulations on a dark field. In most C. collaris, the dark pigmentation within each subquadrate reticulation has a light center, which results in a pattern reminiscent of the dorsal pattern of a jaguar. A similar pattern is sometimes evident in other species, such as C. bicinctores. In 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 51 C. nebrius, at least three easily recognizable patterns occur. One of two available preserved specimens from the Tucson Mountains (SDSNH 15208), as well as a photographic record of a specimen from this locality, have a peripheral gular pattern that is is very similar to that of C. collaris. More specifi- cally, the pattern is composed of white reticulations filled with dark pigment (in this case dark blue) with pale, light centers. This is in striking contrast to the peripheral gular pattern of other C. nebrius. A sec- ond pattern, which has been observed in individuals from the Gila and Mohawk mountains of Arizona and the volcanic mountains immediately adjacent to Mexican Highway 2, at least as far south and east as 30 mi west of Caborca, Sonora, Mexico, is com- posed of radiating, oblique, white and dark blue stripes. These localities represent the northwestern portion of the range of C. nebrius. The third pattern, which corresponds to the pattern that Axtell and Montanucci (1977) used in their diagnosis of the species, is composed of pale white spots on a light blue to slate blue field. This pattern is seen in in- dividuals from the Silverbell Mountains, the Es- trella Mountains, and from Why, Arizona, and in one of three specimens from the Tucson Mountains, as well as from 1 6 mi south of Nogales, the vicinities of Nacori Chico and Bacadehuachi, 30 mi west of Caborca, and Guaymas, Sonora, Mexico. The pres- ence of the second and third peripheral pattern types from identical localities, 30 mi west of Caborca and in the Tucson Mountains, suggests that these pat- terns may occur polymorphically. A similar situa- tion occurs 0.9 mi south of Why, Arizona, where one individual has the pattern of white spots on a pale blue field and a second has a pattern inter- mediate between the spotted pattern and the one composed of radiating blue and white stripes (SDSNH 68645-46). Therefore, a taxonomic deci- sion based on the differences between the spotted and striped gular patterns would certainly be pre- mature. A binary character associated with this variation in peripheral gular pattern is recognized. One state is the presence of a reticulated pattern in the pe- ripheral gular region, the other is the presence of a pattern of pale spots or of radiating obliquely ori- ented stripes extending outward from the edge of the central gular region. If future collecting shows that the spotted and obliquely striped patterns do not grade into one another, and thus represent phy- logenetically useful variations in gular pattern, then this a priori assessment of homology will have to be reevaluated. Neither of the two character states that I have described above are present in the out- group taxa and therefore this character is left un- polarized. The gular pattern of Gambelia is very different from that of Crotaphytus. The pattern is composed of longitudinally arranged black streaks or spots that extend from the posterior gular region to the man- dibular symphysis. This gular pattern is present in all age classes of Gambelia and in both sexes, which is in contrast to the Crotaphytus condition, in which only adult males have a fully developed gular pat- tern. A single character was formulated in which the alternative states are a fully developed gular pattern in all age classes and in both sexes or a gular pattern that is only fully developed in adult males. Variation in the outgroups prevented polarization of this char- acter. Enlarged Melanie Axillary Patches (Character 8 3). — Enlarged melanic axillary patches are variably present in Crotaphytus bicinctores, C. collaris, C. insularis, C. nebrius, and C. vestigium. They are absent from C. antiquus, C. dickersonae, C. gris- meri, and C. reticulatus, although in C. reticulatus and C. antiquus, black-filled reticulations may occur in the same axillary position as the melanic spots seen in other Crotaphytus. Axillary patches are not a fixed feature in any Crotaphytus species. Within C. collaris, they are present only in western popu- lations from Arizona (and potentially Utah). Among the outgroup taxa, axillary patches were observed only in Uta, Uma exsul, and Leiocephalus macropus (within Leiocephalus, axillary patches are variable within L. macropus, but present in male L. lunatus, and male and female L. greenwayi; G. Pregill, per- sonal communication, 1994), and, therefore, the presence of axillary patches is treated as the derived condition. Ventrolateral Coloration (Character 84; often unobservable in preserved specimens). — Conspic- uous ventrolateral coloration is present in adult male C. dickersonae, Crotaphytus insularis, and C. ves- tigium, as well as some male C. collaris and C. ne- brius. The coloration present in C. insularis, C. ne- brius, and C. vestigium appears to be ephemeral in nature, appearing only in the breeding season, whereas the coloration in C. collaris and C. dick- ersonae appears to be an extension of the normal adult male dorsal coloration onto the ventrolateral abdominal region. If this observation holds true, then it would appear unlikely that the ventrolateral coloration observed within all of these species is 52 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 homologous. However, a survey of the ventrolateral coloration over the entire activity season has not been completed for each species and an assessment of homology is not possible. Breeding male Crotaphytus insularis are charac- terized by olive green ventrolateral coloration that contrasts strongly with their brown dorsal colora- tion. Coloration that is identical in appearance oc- curs in C. vestigium males from the northern part of their range (north of Bahia de San Luis Gonzaga, Baja California, Mexico). Between Bahia de San Luis Gonzaga and Bahia de Los Angeles (a distance of approximately 120 km), a shift in ventrolateral col- oration from olive green to golden orange occurs. The golden orange coloration is present in C. ves- tigium at least from Bahia de Los Angeles south- ward. In C. nebrius, coloration similar to that ob- served in southern C. vestigium may be present. This coloration has been observed in specimens from the Mohawk Mountains (Yuma County, Arizona), the Tucson Mountains (Pima County, Arizona), and 66.6 mi W Sonoita along Mexican Highway 2, and suggests that orange ventrolateral breeding colora- tion is characteristic of the species. Crotaphytus dickersonae and some C. collaris (those with tur- quoise or green dorsal coloration) may have bluish ventrolateral coloration. Ventrolateral coloration was coded as an unor- dered multistate character with the absence of ven- trolateral coloration coded as state 0, the presence of olive green coloration coded as state 1, the pres- ence of orange coloration coded as state 2, and the presence of bluish coloration as state 3. Crotaphytus vestigium is polymorphic for this feature with states 1 and 2 present; C. nebrius is assigned state 2; C. dickersonae is assigned state 3; and C. collaris is assigned states 0 and 3. All other Crotaphytus and Gambelia are assigned state 0. No attempt was made to polarize this character. Dorsal Coloration (Character 85; Fig. 30-32; some character states are not observable in preserved specimens).— The dorsal coloration of adult male Crotaphytus is characterized by much interspecific variation. Crotaphytus reticulatus has a dorsal col- oration of golden tan, while C. nebrius has a similar straw yellow coloration that lacks the golden hue of C. reticulatus. Crotaphytus dickersonae is unique among Crotaphytus in that its coloration ranges from aquamarine to cobalt blue. The coloration of this species is generally dissimilar to that of C. collaris, although the aquamarine phase of C. dickersonae is occasionally approached by C. collaris. Crotaphytus bicinctores, C. antiquus, C. grismeri, C. insularis, and C. vestigium have a brown dorsal coloration. Crotaphytus collaris is extremely variable geograph- ically, with some populations characterized by a tur- quoise body pattern with a yellow head (eastern Ar- izona, eastern Utah, western Colorado, western New Mexico, as well as some Great Plains populations, for example Altus, Oklahoma, and Flint Hills, Kan- sas), others by a bright green coloration (many east- ern populations), others by a pale to dark brown coloration (Chihuahuan Desert populations in southern New Mexico, western Texas, and Chihua- hua, Mexico), and still others by a combination of olive green and/or gray (Coahuila, Durango, Zaca- tecas). Most populations of Gambelia are off-white to tan in coloration. However, G. copei may range from golden tan to dark brown. An unordered mul- tistate character was coded with the off-white to tan coloration of most Gambelia represented by state 0, the golden tan of C. reticulatus by state 1, the straw yellow coloration of C. nebrius by state 2, the blue coloration of C. dickersonae and some C. col- laris by state 3, a brown coloration by state 4, and green and/or gray coloration by state 5. Crotaphytus collaris is considered polymorphic with states 3, 4, and 5 present, as is G. copei with states 0 and 4. This character was not polarized. Behavioral Characters Saxicoly (Character 86). — Gambelia and Crota- phytus reticulatus generally occur in flatland desert habitats and have a generalized terrestrial lifestyle. Montanucci (1965, 1967, 1969, 1971) performed ecological investigations of Gambelia silus, G. wis- lizenii, and C. reticulatus and concluded that they are virtually ecological equivalents. Although each will utilize rocks as perching points when they are available, they often are found in areas quite re- moved from any rocky habitat. Also consistent with the assumption that the terrestrial lifestyles of Gam- belia and C. reticulatus are homologous is the com- mon utilization of “freeze behavior” in G. wislizenii (McCoy, 1967), G. copei, and C. reticulatus. Mon- tanucci ( 1 967) described a similar behavior in young G. silus, although he later suggested that this be- havior is rare in this species (Montanucci, 1978). When disturbed, these species often take refuge be- neath a nearby bush and remain motionless, ap- parently relying on crypsis to avoid detection. In many cases, the lizard can be approached within one or two meters without causing it to flee. In con- trast with the terrestrial lifestyles of Gambelia and C. reticulatus, the remaining species of Crotaphytus are saxicolous such that they appear to be extremely 1996 McGUIRE— SYSTEM ATICS OF CROTAPHYTID LIZARDS 53 Fig. 36.— A juvenile Crotaphytus nebrius exhibiting lateral tail coiling behavior. dependent on rocky habitats and are almost never observed in areas devoid of rocks. Montanucci (1974) noted that C. col laris may be found in and flatland desert in at least two localities in Coahuila, Mexico. However, this behavior is certainly atypical for the species and similar behavior has not been observed by me or discussed in the literature for any of the other Crotaphytus taxa. Although saxicoly certainly is not unique to Cro- taphytus, this particular form of saxicoly, in which the lizards are restricted to boulder-strewn hillsides, alluvia, canyons, etc., where they scamper bipedally from rock to rock, perch atop rocks, and scan the immediate vicinity for potential prey and predators, is rare in the outgroup taxa. Nevertheless, because there are a diversity of character states present in the outgroup taxa that are absent from either Cro- taphytus or Gambelia (such as arboreality, burrow- ing, and crevice-dwelling), a clear polarity decision was not possible for this character. Therefore, this character was left unpolarized. Territoriality (Character 87). — Territoriality is known to be absent in Gambelia wislizenii (McCoy, 1967; Montanucci, 1970; Tanner and Krogh, 1974a; Tollestrup, 1979, 1982, 1983). Crotaphytus as well as G. silus are known to be highly territorial (Fitch, 1956; Montanucci, 1965, 1971; Yedlin and Fergu- son, 1973;Moehn, 1976; Sanborn and Loomis, 1979; Tollestrup, 1979, 1982, 1983). It has not been de- termined whether territoriality is present or absent in G. copei, although the behavior of this species appears to be quite similar to that of G. wislizenii. Territoriality is widespread within Iguania, and is known to be present in all of the remaining iguanian families except Hoplocercidae (Carpenter, 1967; Stamps, 1977), a group for which data were un- available. Of the many outgroup taxa that have been studied. Stamps (1977) could list only two, Phry- nosoma and Anolis agassizi, that are known to lack territorial behavior. Therefore, the absence of ter- ritoriality is here treated as the derived state. Lateral Tail Coiling (Character 89; Fig. 36). — All Crotaphytus coil their tails laterally when taking ref- uge under stones or debris, while at rest, and while hibernating (Legler and Fitch, 1957). Presumably, this behavior assists in keeping the tail out of the reach of predators. Lateral tail coiling is also known in the members of the Anolis homolechis complex 54 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 of Cuba (Hardy, 1958; Ruibal and Williams, 1961) and in several species of Leiocephalus (C. A. Haas, S. B. Hedges, personal communication, 1994; K. de Queiroz, personal communication, 1995 — although they are described as coiling their tails vertically over their backs by Schwartz and Henderson, 1991). However, these groups are nested within Polychro- tidae and Tropiduridae, respectively, indicating that their behaviors are convergent with that observed in Crotaphytus. The presence of lateral tail coiling is considered to be the derived state. Consumption of Vertebrates.— All crotaphytids except Crotaphytus antiquus, C. grismeri, and C. nebrius have either been documented in the litera- ture to include vertebrates in their diets (C. bicinc- tores : Banta, 1960; Snyder, 1972; Nussbaum et al., 1983; C. collaris : Fitch, 1956; McAllister and Trauth, 1982; C. reticulatus: Klein, 1951; Montanucci, 1971; Gambelia copei : Banta and Tanner, 1968; Montan- ucci, 1965; G. wislizenii : Stejneger, 1893; McCoy, 1967; Montanucci, 1967; Snyder, 1972; Tanner and Krogh, 1974a; Parker and Pianka, 1976; Tollestrup, 1979, 1983; Pietruszka et al., 1981; Crowley and Pietruszka, 1983) or have been observed to do so by the author. The primary vertebrate prey is other lizards, although rodents and snakes also have been recorded. There appears to be variation in the rel- ative proportion of vertebrates included in the diets of the various species, with Gambelia wislizenii (Parker and Pianka, 1976; Tollestrup, 1979, 1982, 1983) and G. copei consuming a greater proportion of vertebrate prey than other species. Many other iguanian species are known to eat vertebrates, including the phrynosomatid genera Pe- trosaurus, Uma, Holbrookia, and Sceloporus, which are all known to include other lizards in their diets (Stebbins, 1985); the corytophanid Basiliscus (Van Devender, 1982); the polychrotid Anolis equestris (Ruibal, 1964); and the chamaeleonids Chlamydo- saurus kingii and Physignathus lesueurii (Cogger, 1992). I have not attempted to review the feeding habits of all of the potential outgroup taxa, but it is likely that many other species have similar feeding habits. Thus, the presence or absence of carnivory may not be a polarizable character, limiting its use- fulness in this analysis. Furthermore, since most lizards will eat anything palatable that they are able to overcome, the inclusion of vertebrates in the diet may be, at least in part, a function of maximum adult size. For these reasons, this characteristic was not included in this analysis. However, the carniv- orous predatory habits of Crotaphytus and Gam- belia are consistent with a hypothesis of crotaphytid monophyly. Vocalization.— The ability to vocalize is rare in squamates, with gekkotans being the only family in which it is known to occur commonly. Within ig- uanian lizards, vocalization is apparently limited to crotaphytids and certain polychrotids. A squealing sound is known to be emitted by Gambelia wislizenii (Jorgenson et al., 1963; Wever et al., 1966; Smith, 1974) and Crotaphytus bicinctores (Smith, 1974) during periods of stress. Similar vocalizations were discussed by Ruibal ( 1964) in three species of Cuban anoles, Anolis iso/epis, A. lucius, and A. vermiculatus and by Lynn and Grant (1940) in A. grahami and A. opalinus (also noted in A. grahami by Etheridge, 1955). Because vocalization data are lacking for the majority of crotaphytid species, I have not included this character in the phylogenetic analysis. How- ever, as with carnivory, the presence of vocalization in some species of Crotaphytus and Gambelia is consistent with the hypothesis of crotaphytid mono- phyly. CHARACTER LIST The following character list includes the morpho- logical characters (informative or uninformative) discussed in the text, as well as the nine informative allozyme characters (characters 89-98) that could be coded using the Manhattan distance frequency approach discussed in Wiens ( 1 995). One multistate morphological character (31) was also coded using the Manhattan distance frequency approach. Its step matrix is presented in Appendix 4 along with the step matrices for the allozyme characters. Character descriptions followed by (P) are polarized, those fol- lowed by (U) are unpolarized, and those followed by (UO) are unordered. Characters 28, 68, 75, 84, and 85 were not analyzed using frequency coding (see Materials and Methods). Skull and Mandible 1. Nasal process of the premaxilla (P): (0) broad, (1) narrow and elongate. 2. Ventral suture between vomers and premaxilla (P): (0) does not form a strong vertical ridge, (1) forms a strong vertical ridge. 3. Nasals (P): (0) do not overlap nasal process of the premaxilla anterior to posterior extent of 1996 McGUIRE— SYSTEM ATICS OF CROTAPHYTID LIZARDS 55 external nares, (1) overlap nasal process of the premaxilla anterior to posterior extent of ex- ternal nares. 4. Prefrontals (P): (0) not in contact with jugals, (1) contact jugals. 5. Cranium (U): (0) vaulted, (1) not vaulted. 6. Postorbitals (P): (0) weakly overlapped dorsally by frontal and parietal, (1) strongly overlapped dorsally by frontal and parietal. 7. Tubercle on anterolateral portion of postorbi- tals (P): (0) absent, (1) present. 8. Posterior border of parietal roof (P): (0) ap- proximately twice as wide as narrowest portion of frontal bone (unconstricted), (1) equal in width or only slightly wider than narrowest por- tion of frontal bone (constricted). 9. Supratemporal processes (in lateral view) (P): (0) tapered, rapidly narrowing dorsoventrally at their midpoints; (1) not tapered, remain broad over entire length. 10. Supratemporals (P): (0) broadly exposed on the lateral surface of the supratemporal process of the parietal, (1) lies in a groove on ventral sur- face of supratemporal process of parietal. 1 1. Septomaxillae (P): (0) wide, (1) narrow. 12. Suture of maxillae with premaxilla (P): (0) not saddle-shaped, no process of the maxilla over- laps the lateral border of the premaxillary base; (1) saddle-shaped, a process of the maxilla over- laps lateral border of premaxillary base. 13. Shape of maxilla-palatine articulation (U): (0) low arch, (1) triangular. 14. Jugal-ectopterygoid tubercle (P): (0) absent, (1) present. 15. Angle of jugal along antero ventral border of orbit (P): (0) approximately 45 degrees, (1) ap- proximately 90 degrees (box-like condition). 16. Extravomerine bones (P): (0) absent, ( 1 ) at least one present. 17. Palatine foramen (U): (0) present, (1) absent. 18. Transverse process of the pterygoid with (U): (0) weakly developed ventral process, (1) strongly developed ventral process. 19. Paraoccipital processes project posteriorly (P): (0) to level of occipital condyle, ( 1 ) well beyond occipital condyle. 20. Angle of the quadrate process of the pterygoid (U): (0) approximately 18 degrees, (1) approx- imately 26-3 1 degrees. 21. Posterior projection of ectopterygoid crest (U): (0) present, (1) absent. 22. Posterior projections of parabasisphenoid (P): (0) reach the sphenoccipital tubercles; (1) ter- minate at, or anterior to, the base of the sphen- occipital tubercles. 23. Anterior extent of angular (U): (0) never reaches the fourth dentary tooth (counting forward from the posteriormost tooth) and rarely extends an- teriorly beyond the posteriormost tooth, (1) ex- tends at least to the fourth tooth (counting for- ward from the posteriormost tooth) and usually beyond. 24. Posterior mylohyoid foramen (U): (0) equal with apex of coronoid, (1) posterior to apex of cor- onoid. 25. Posterolingual process of the coronoid (P): (0) oriented vertically, (1) angled posteroventrally at approximately 45 degrees. 26. Bony shelf extending between medial process of surangular and ramus of mandible (P): (0) absent, (1) present. 27. Lateral process of surangular (P): (0) absent or present as a weakly elevated ridge, (1) present as a large protuberance. 28. Ridge on lateral surface of surangular (P, UO): (0) absent, (1) moderately developed, (2) strong- ly developed such that the dorsal surface of the mandible is concave. 29. Tympanic crest (P): (0) forms lateral border of retroarticular process, (1) curves posteromedi- ally. 30. Maxillary and dentary teeth (P): (0) stout, either straight or slightly recurved; (1) long, slender, and more strongly recurved. 3 1 . Number of premaxillary teeth (U, UO): (0) five, (1) six, (2) seven, (3) eight, (4) nine. 32. Palatine teeth (P): (0) absent, (1) present. 33. Pterygoid tooth patch (P): (0) follows margin of interpterygoid vacuity, (1) curves posterolater- ally. Hyoid Apparatus 34. Ceratohyals (U): (0) without hook-like process- es on proximal, medial edge; (1) with hook-like processes on proximal, medial edge. 35. Length of second ceratobranchials (U): (0) ap- proximately one-half length of ceratohyals, (1) more than two-thirds length of ceratohyals. 36. Second ceratobranchials (P): (0) in contact me- dially, (1) widely separated. Miscellaneous Features of the Head Skeleton 37. Skull rugosity at some point in ontogeny (U): (0) absent, (1) present. 56 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 POSTCRANIAL SKELETON 38. Zygosphenes (P): (0) not separated from pre- zygapophyses by notch, (1) separated from pre- zygapophyses by notch. 39. Tail shape (P): (0) round or subcyhndrical with- out well-developed dorsal and ventral fat bod- ies, (1) laterally compressed with well-devel- oped dorsal and ventral fat bodies. 40. Autotomic fracture planes of caudal vertebrae (P): (0) present, (1) absent. 4 1 . Number of xiphisternal ribs (P): (0) two, ( 1 ) one. 42. Notch on the anterior edge of the suprascapular cartilage (P): (0) absent, (1) present. 43. Posterior coracoid fenestrae (P): (0) absent, (1) present. 44. Calcified cartilage anterior border of scapular fenestra (P): (0) present, (1) absent. 45. Clavicular fenestrations (P): (0) absent, (1) pres- ent. 46. Termini of iliac blades (U): (0) laterally com- pressed, (1) round. 47. Arch formed by contact of the medial and lat- eral plantar tubercles (P): (0) absent, ( 1 ) present. Squamation 48. Rostral scale (U): (0) broad, approximately four times wider than high; (1) narrow, approxi- mately two times wider than high. 49. Some of the prefrontal, frontal, interparietal, and parietal scales are (U): (0) enlarged relative to the surrounding scales in such a way as to form conspicuous supraorbital semicircles, (1) not enlarged relative to surrounding scales such that conspicuous supraorbital semicircles are not distinguishable. 50. Elongate scale in subocular series (P): (0) pres- ent, (1) absent. 51. Terminal supradigital scales (P): (0) not elevat- ed from dorsal surface of claws, (1) elevated from dorsal surface of claws. 52. Femoral pore series (P): (0) terminates before reaching inferior angle of knee, (1) extends be- yond inferior angle of knee. 53. Femoral pores (P): (0) much larger and more strongly developed in males than females, (1) roughly equal in size or only slightly larger in males than females. 54. Distal tail skin (P): (0) bound to underlying musculature, (1) loosely adherent to underlying musculature. 55. Posteromedially angled folds within gular fold (U): (0) present, (1) absent. 56. Angle of supra-auricular fold (U): (0) horizon- tal, (1) at 45-degree angle. 57. Antehumeral mite pockets (P): (0) absent, (1) present. 58. Postfemoral mite pockets (P): (0) absent, (1) present. Additional Structural Characters 59. Hemipenes (U): (0) large, (1) small. 60. Sexual dimorphism (P): (0) males larger than females, (1) females larger than males. Coloration 6 1 . Ephemeral orange coloration in subadult males (P): (0) absent, (1) present. 62. Paired, paravertebrally arranged, blood-red spots in juveniles of both sexes (P): (0) absent, (1) present. 63. Bright yellow tail coloration in adult females (P): (0) absent, (1) present. 64. Bright yellow tail coloration in juveniles of both sexes (P): (0) absent, (1) present. 65. Off-white stripe on dorsal crest of tail (P): (0) absent, (1) present. 66. Juvenile dorsal pattern with a white reticular component (P): (0) absent, (1) present. 67. Granular reticulations on ventrolateral surface of abdomen (P): (0) absent, (1) present. 68. White component of adult dorsal body pattern in the form of (U, UO): (0) broad, offset, trans- verse bars; (1) a reticulum over the entire dorsal surface; (2) spots; (3) spots along with slender transverse dorsal stripes; (4) spots along with wavy lines and dashes. 69. Sexual dichromatism of the dorsal pattern (U): (0) absent; (1) present, such that the dorsal col- oration of males and females is conspicuously different. 70. Paired melanic keels on ventral caudal extrem- ity (P): (0) absent, (1) present. 71. Black oral melanin (P): (0) absent, (1) present. 72. Black collar or collars with white borders (P): (0) collars absent, (1) collars present. 73. Posterior collar markings (U): (0) in contact dorsally or nearly so, (1) widely separated dor- sally. 74. Anterior collar (U): (0) incomplete ventrally (no dark pigments in gular fold), (1) complete ven- trally (dark pigments present in gular fold). 75. Ventral extent of the posterior collar (U, UO): (0) extends onto dorsal surface ofbrachium, (1) 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 57 just reaches forelimb insertion, (2) terminates within antehumeral fold, (3) terminates before entering antehumeral fold. 76. Dark nuchal spots (P): (0) absent, (1) present. 77. Inguinal patches (P): (0) absent, (1) present. 78. Inguinal patches (U): (0) small, (1) large (taxa without inguinal patches coded as unknown [”?”]) 79. Femoral pore secretions (P): (0) off-white to gray, (1) black. 80. Black pigmentation in central region of gular pattern (U): (0) absent, (1) present. 8 1 . Peripheral gular pattern (U): (0) reticulated, ( 1 ) pale spots or radiating oblique stripes. 82. Gular pattern (U): (0) present only in adults and well developed in males only, (1) well devel- oped in all age classes and in both sexes. 83. Enlarged melanic axillary patches (P): (0) ab- sent, (1) present. 84. Ventrolateral coloration (P, UO): (0) does not differ from ventral coloration (white), (1) olive green, (2) orange, (3) blue. 85. Dorsal coloration (U, UO): (0) off-white to tan; An initial analysis was performed on the mor- phological data set of 88 characters (allozyme data of Montanucci et al. [1975] not included). This re- sulted in the discovery of a single tree (Fig. 37) with a length of 12,334 (123.34 when the effect of weight- ing the characters is removed) and a consistency index (Cl; excluding uninformative characters) of 0.761, a retention index (RI) of 0.848, and a gx tree length frequency distribution skewness value of — 1.49 (the critical gx value for this data set when randomized is —0. 16 [ P < 0.01]); Table 1). This Cl is greater than that expected for an analysis of 1 3 taxa (expected Cl = 0.649; Sanderson and Dono- ghue, 1989), indicating that there is less homoplasy than expected in these data when compared with the 60 data sets reexamined by Sanderson and Don- oghue (1989). The gx value is strongly left skewed suggesting that the data are phylogenetically infor- mative. The tree generated in the bootstrap analysis is presented in Figure 38. Reanalysis of the allozyme data set of Montanucci etal. (1975) using the Manhattan distance frequency approach resulted in the discovery of a single most parsimonious tree (Fig. 39A). The tree length fre- quency distribution data, summarized by the gx sta- (1) golden tan; (2) straw yellow; (3) aquamarine to cobalt blue; (4) brown; (5) olive green, tur- quoise, or gray. Behavior 86. Saxicoly (including the use of saltatory bipedal locomotion in rocky habitats) (U): (0) absent, (1) present. 87. Territoriality (P): (0) present, (1) absent. 88. Lateral tail coiling (P): (0) absent, (1) present. Allozymes 89. H-LDH (U): four electromorphs. 90. aGPD (U): two electromorphs. 91. 6-PGD (U): three electromorphs. 92. ICDs (U): four electromorphs. 93. ICDm (U): four electromorphs. 94. GOTs (U): three electromorphs. 95. Pro (U): two electromorphs. 96. Estl (U): three electromorphs. 97. Hbpf (U): two electromorphs. 98. Tr (U): four electromorphs. tistic (Hillis and Fluelsenbeck, 1992), suggests that there is phylogenetically informative signal in this data set (observed gx value of —0.50; the critical gx value for this data set when randomized is —0.45 [P < 0.01]). The bootstrap tree for the analysis is given in Figure 39B (see comments below regarding interpretation of bootstrap P values). Because analysis of both the morphological and allozyme data sets suggests that they contain phy- logenetic signal, these data sets were combined and the larger data set was analyzed. Analysis of the combined morphology and allozyme data sets re- sulted in the same tree as did the analysis of the morphology data alone (Fig. 37). The tree length is 139.91, while the Cl (excluding uninformative char- acters) for the combined tree is 0.761, the RI is 0.848, and the gj value is —1.45 (critical gx value = —0. 1 5 [P < 0.01]). PAUP is unable to incorporate the step matrix characters into the Cl and RI cal- culations, which explains why the Cl and RI values are identical to those discovered in the analysis of the morphological data alone. The gx and Cl values indicate that the data harbor substantial phyloge- netic signal. The 50 percent majority-rule consensus tree generated in the bootstrap analysis of the com- 58 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 combined morphology + allozymes data analyses. bined data set is presented in Figure 40. The boot- strap analysis and values indicate that there is similar support for tree A with or without the allo- zyme data. Reanalysis of the allozyme data set using the poly- morphic coding and Mabee and Humphries (1993) approaches each resulted in different trees than that estimated using the step matrix approach. However, the combined analyses always resulted in the same tree as the morphology data alone, regardless of the coding approach employed with the allozyme data. Character support for each stem of the cladogram discovered in the combined analysis (Fig. 37) is pre- sented below. A complete listing of apomorphies, including the autapomorphies of the terminal taxa, is presented in Appendix 5. Transformations that are described as “unambiguous” are supported un- der both ACCTRAN and DELTRAN optimization. Therefore, when a node is described as “ambigu- ously” supported by a particular character state change, this means that the character in question Fig. 38.— The 50 percent majority-rule consensus tree generated from the bootstrap analysis of the morphology-only data set. supports this node under either ACCTRAN or DELTRAN optimization but not under both. “Fixed” transformations are those that involve a change from one fixed state to another (state “a” to state “y,” or vice versa). When a transformation is not fixed it may be referred to as “polymorphic.” Such transformations involve incomplete changes (for example from state “a” to state “m”) and re- ceive a reduced weight due to the frequency coding approach employed. Unambiguous fixed transfor- mations include those character state changes from one fixed state to another that are discovered under both ACCTRAN and DELTRAN optimization. Stem A (Crotaphytidae) is supported by 1 1 un- ambiguous transformations. Five of the synapo- morphies represent fixed character state changes: 4. 1 (prefrontals contact jugals), 6.1 (parietal and frontal strongly overlap the postorbital), 14.1 (jugal-ectop- terygoid tubercle present), 29.1 (tympanic crest of the retroarticular process curves posteroventrally), and 71.1 (black oral melanin present, reversed in 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 59 G. wislizenii C. vestigium C. bicinctores C. reticulatus C. nebrius C. collaris C. dickersonae G. wislizenii C. vestigium C. bicinctores C. reticulatus C. nebrius C. collaris C. dickersonae Fig. 39.— (A) The single most parsimonious tree discovered in the reanalysis of the allozyme data set of Montanucci et al. ( 1 9 7 5) employing the approach in which frequency values are encoded into step matrices using Manhattan distances. (B) The 50 percent majority-rule consensus tree generated in the bootstrap analysis of this data set. stem J). Six additional unambiguous yet polymor- phic apomorphies support this stem: 10.1 (supra- temporal lies in a groove in ventral surface of su- pratemporal process of parietal, 23/24 of one step under ACCTRAN optimization, 2/24 of one step under DELTRAN optimization), 26.1 (bony shelf extends between medial process of surangular and ramus of mandible, 2/24 of one step under both ACCTRAN and DELTRAN optimization), 32.1 (palatine teeth present, 16/24 of one step ACCT- RAN, 1 3/24 of one step DELTRAN), 42. 1 (scapular notch present, 10/24 of one step ACCTRAN, 5/24 of one step DELTRAN), 43.1 (posterior coracoid fenestrae present, fixed ACCTRAN, 16/24 of one step DELTRAN), and 45. 1 (clavicular fenestrations present, fixed ACCTRAN, 8/24 of one step DEL- TRAN). Finally, Crotaphytidae may also be sup- ported by two ambiguously placed synapomorphies: 40. 1 (autotomic fracture planes of caudal vertebrae absent, 4/24 of one step ACCTRAN) and 58.1 (post- femoral mite pockets present, fixed ACCTRAN). Stem B ( Gambelia ) is supported by 1 3 unambig- Fig. 40. — The 50 percent majority-rule consensus tree generated in the bootstrap analysis of the complete (morphology + allo- zymes) data set. uous synapomorphies, six of which represent fixed character state changes: 12.1 (saddle-shaped suture between premaxilla and maxilla), 30. 1 (slender, re- curved maxillary and dentary teeth), 44.1 (loss of the calcified cartilage border of the scapular fenes- tra), 46.1 (termini of the iliac blades round), 52.1 (femoral pore series extends beyond the inferior an- gle of the knee), and 62.1 (paired, paravertebrally arranged, blood-red spots present in juveniles of both sexes). This stem is also supported by seven unambiguous but polymorphic apomorphies: 2.1 (articulation between premaxilla and vomers in the form of a vertical ridge, 23/24 of one step ACCT- RAN, fixed DELTRAN), 15.1 (angle of jugal along anteroventral border of orbit approximately 90 de- grees, 1 2/24 of one step under both ACCTRAN and DELTRAN), 17.1 (palatine foramen absent, 19/24 of one step ACCTRAN, 18/24 of one step DEL- TRAN), 24.1 (posterior mylohyoid foramen pos- 60 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 tenor to apex of coronoid, 20/24 of one step under both ACCTRAN and DELTRAN), 26. 1 (bony shelf between the median process and ramus of the man- dible, 22/24 of one step under both ACCTRAN and DELTRAN), 36.1 (second ceratobranchials widely separated, fixed ACCTRAN, 16/24 of one step DELTRAN), and 41.1 (one rather than two xiph- isternal ribs, 20/24 of one step under both ACCT- RAN and DELTRAN). This stem also may be sup- ported by five ambiguously placed transformations, the placements of which depend upon the optimi- zation routine employed: 1 . 1 (nasal process of pre- maxilla narrow, 5/24 of one step ACCTRAN), 7.1 (tubercle present on anterolateral portion of post- orbital, 2/24 of one step ACCTRAN), 25.1 (poster- olingual process of coronoid angled posterolaterally at approximately 45 degrees, fixed ACCTRAN), 43. 1 (posterior coracoid fenestrae absent, 8/24 of one step DELTRAN), and 58. 1 (postfemoral mite pock- et present, fixed DELTRAN). Finally, this stem may be further supported by as many as 12 unpolarized characters: 8.0 (parietal roof not constricted poste- riorly), 9.0 (supratemporal processes tapered), 13.0 (maxilla-palatine articulation in the form of a low, rounded arch), 21.1 (posterior process of the ectop- terygoid crest absent), 23.0 (angular does not extend anteriorly beyond the fourth dentary tooth [counting forward] and rarely extends beyond the posterior- most tooth), 37.0 (skull nonrugose), 49.0 (supraor- bital semicircles absent), 50.0 (subocular scale series includes one very elongate scale), 55.1 (gular fold without closely approximating posteromedial folds), 56.0 (supra-auricular fold horizontal), 59.0 (hemi- penes large), and 82. 1 (fully developed gular pattern in females). It is equally parsimonious for each of these characters to support stem E ( Crotaphytus ) depending upon their true polarity assignments. Stem C is supported by four unambiguous syna- pomorphies, two of which are fixed: 5.1 (loss of a vaulted cranium) and 20.0 (angle of the quadrate processes of the pterygoid approximately 18 de- grees). The polymorphic apomorphies are charac- ters 31.1 (number of premaxillary teeth, 0.35 of one step under both ACCTRAN and DELTRAN) and 32. 1 (palatine teeth present, 6/24 of one step ACCT- RAN, 8/24 of one step DELTRAN). Nine more potential synapomorphies depend upon the partic- ular optimization routine employed. All but one of these (discovered during ACCTRAN optimization data runs) were coded as missing (“?”) for G. co- ronal and thus may actually represent synapomor- phies for stem D: 11.1 (septomaxillae slender and elongate, fixed), 15.1 (angle of jugal along antero- ventral border of orbit approximately 90 degrees, 12/24 of one step), 24.1 (posterior mylohyoid fo- ramen posterior to apex of coronoid, 2/24 of one step), 41.1 (one rather than two xiphisternal ribs, 3/24 of one step), 42.0 (scapular notch absent, 5/24 of one step), 53.1 (femoral pores of approximately equal size in males and females, fixed), 60. 1 (females attain larger adult SVL than males, fixed), and 87. 1 (territoriality absent, fixed). Finally, 25.2 (postero- lingual process of the coronoid angled posteroven- trally at approximately 45 degrees) may represent a fixed synapomorphy for this node (fixed DEL- TRAN). Stem D is supported by one unambiguous syna- pomorphy: 1 . 1 (nasal process of premaxilla long and slender, 19/24 of one step ACCTRAN, fixed DEL- TRAN). It may be further supported by as many as nine ambiguous (DELTRAN) characters including six of the characters (with the same frequency val- ues) discussed under stem C (1 1, 15, 24, 41, 53, 60) for which G. corona f was coded as missing (“?”). The three remaining potential synapomorphies in- clude: 7.1 (tubercle on anterolateral border of post- orbital, 2/24 of one step), 10.1 (supratemporal lies in a groove along ventral border of supratemporal process, 21/24 of one step), and 40,1 (autotomic fracture planes absent, 4/24 of one steb). Stem E ( Crotaphytus ) is supported by 24 unam- biguous synapomorphies, 12 of which are fixed in- cluding: 19.1 (paraoccipital processes extend pos- terior to the occipital condyle), 27. 1 (lateral process of the surangular present as a large protuberance), 33.1 (pterygoid tooth patch curls posterolaterally), 34.1 (ceratohyals with hook-like processes on prox- imal, medial edge), 35.1 (second ceratobranchials more than two-thirds the length of the ceratohyals), 54.1 (skin of the distal portion of tail weakly ad- herent to underlying musculature), 57.1 (presence of antehumeral mite pockets), 61.1 (subadult males acquire “gravid female” coloration), 66.1 (juvenile color pattern composed of white reticulations), 72. 1 (acquisition of a black collar or collars outlined in white), 88.1 (lateral tail coiling behavior), and 98.3 (electromorph Tr). This stem is also supported by 12 unambiguously placed yet polymorphic trans- formations depending upon the optimization rou- tine employed: 10.1 (supratemporal lies in a groove along ventral border of supratemporal process, 1/24 of one step ACCTRAN, 22/24 of one step DEL- 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 61 TRAN), 22.1 (posterior projections of the parabas- isphenoid terminate at, or anterior to, the base of the sphenoccipital tubercles, fixed ACCTRAN, 23/ 24 of one step DELTRAN), 31.1 (number of pre- maxillary teeth, 0.38 of one step under both ACCT- RAN and DELTRAN), 40. 1 (loss of autotomic frac- ture planes of the caudal vertebrae, 20/24 of one step ACCTRAN, fixed DELTRAN), 42.1 (supra- scapular notch present, 12/24 of one step ACCT- RAN, 13/24 of one step DELTRAN), 47.1 (medial and lateral plantar tubercles contact to form an arch, 21/24 of one step ACCTRAN, 19/24 of one step DELTRAN), 76.1 (dark nuchal spots present, 19/ 24 of one step under both ACCTRAN and DEL- TRAN), 90.6 (aGPD, 0.08 of one step under both ACCTRAN and DELTRAN), 9 1 .2 or 9 1 .6 (6-PGD, 0. 1 2 of one step ACCTRAN, 0.05 of one step DEL- TRAN), 94.2 (GOTs, 0.69 of one step under both ACCTRAN and DELTRAN), 95.2 or 95.7 (Pro, 0.58 of one step ACCTRAN, 0.56 of one step DEL- TRAN), and 96.6 (Estl, 0.7 1 of one step under both ACCTRAN and DELTRAN). Four additional po- tential transformations at this node are discovered only under ACCTRAN optimization including: 28.1 (moderately developed ridge present on lateral sur- face of the surangular, fixed), 68. 1 (white component of adult dorsal pattern composed of reticulations, fixed), 85.1 (dorsal coloration golden tan, fixed), and 89.2 (H-LDH, fixed). Finally, this stem may be sup- ported by as many as 1 2 transformations that could not be polarized. It is equally parsimonious for each of these characters to support stem B ( Gambelia ) and a complete listing is given under the discussion of stem B. Stem F is supported by nine unambiguously placed transformations, four of which are fixed: 68.2 (white portion of dorsal pattern in the form of spots), 69. 1 (sexual dichromatism of the dorsal color pattern), 85.4 (brown dorsal body coloration), 86. 1 (saxicoly). The five unambiguous yet polymorphic apomor- phies include: 24.0 (posterior mylohyoid foramen equal with apex of coronoid, 1/24 of one step under both ACCTRAN and DELTRAN), 31.0 (number of premaxillary teeth, 0.01 of one step under both ACCTRAN and DELTRAN), 51.1 (terminal su- pradigital scales elevated from dorsal surface of claw, 4/24 of one step under both ACCTRAN and DEL- TRAN), 70. 1 (paired melanic keels on ventral cau- dal extremity, 7/24 of one step under both ACCT- RAN and DELTRAN), and 77.1 (inguinal patches present, 7/24 of one step under both ACCTRAN and DELTRAN). This stem may also be supported by the following five ambiguously placed transfor- mations, depending upon the optimization routine employed: 28.1 (ridge on lateral surface of the sur- angular, fixed DELTRAN), 43. 1 (posterior coracoid fenestrae present, 8/24 of one step DELTRAN), 45.0 (clavicular fenestrations lost, 16/24 of one step ACCTRAN), 58.1 (postfemoral mite pockets pres- ent, fixed DELTRAN), and 89.2 (H-LDH, fixed DELTRAN). Stem G is weakly supported by eight unambigu- ously placed transformations, none of which are fixed. The unambiguous, yet polymorphic apomor- phies include 17.0 (palatine foramen present, 4/24 of one step under both ACCTRAN and DEL- TRAN), 26.0 (no bony shelf present between medial process of the surangular and ramus of mandible, 1/24 of one step under both ACCTRAN and DEL- TRAN), 32.1 (palatine teeth present, 2/24 of one step ACCTRAN, 5/24 of one step DELTRAN), 42. 1 (suprascapular notch present, 2/24 of one step ACCTRAN, 5/24 of one step DELTRAN), 47.1 (arch formed by contact of medial and lateral plantar tubercles, 3/24 of one step under both ACCTRAN and DELTRAN), 51.1 (terminal supradigital scales elevated from dorsal surface of claws, 20/24 of one step under both ACCTRAN and DELTRAN), 70. 1 (paired melanic keels present on ventral caudal ex- tremity, 17/24 of one step under both ACCTRAN and DELTRAN), and 77.1 (inguinal patches pres- ent, 17/24 of one step under both ACCTRAN and DELTRAN). Four ambiguously placed transfor- mations may also support this node: 22. 1 (posterior projections of parabasisphenoid terminate at, or an- terior to, the base of the sphenoccipital tubercles, 1/24 of one step DELTRAN), 24.0 (posterior my- lohyoid foramen equal with apex of coronoid, 1/24 of one step ACCTRAN), 90.1 (aGPD, 0.08 of one step ACCTRAN), and 95.5 (Pro, 0.43 of one step ACCTRAN). Stem H is weakly supported by one unambigu- ously placed polymorphic transformation: 76.0 (ab- sence of dark nuchal spots, 13/24 of one step under both ACCTRAN and DELTRAN). This node may also be supported by two ambiguously placed trans- formations: 2.0 (ventral suture between vomers and premaxilla does not form a vertical ridge, 1/24 of one step ACCTRAN) and 90.1 (aGPD, 0.08 of one step, DELTRAN). Stem I is supported by five unambiguously placed transformations, three of which are fixed: 39.1 (lat- 62 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 erally compressed tail with dorsal and ventral fat bodies), 65.1 (off-white dorsal caudal stripe present), and 78.1 (enlarged inguinal patches in adult males). The two unambiguous yet polymorphic transfor- mations are 31.1 (number of premaxillary teeth, 0.0 1 of one step under both ACCTRAN and DELTRAN) and 76.0 (dark nuchal spots lost, 6/24 of one step under both ACCTRAN and DELTRAN). Three ambiguously placed transformations may also sup- port this node: 24.0 (posterior mylohyoid foramen equal with the apex of coronoid, 1/24 of one step ACCTRAN), 47.1 (arch formed by contact of the medial and lateral plantar tubercles, 2/24 of one step ACCTRAN), and 75.1 (ventral extent of posterior collar marking just reaches forelimb insertion, fixed ACCTRAN). Stem J is supported by six unambiguously placed transformations, two of which represent fixed changes: 71.0 (loss of black oral melanin) and 75.2 (posterior collar terminates within antehumeral fold). The unambiguous yet polymorphic transformations include: 1.1 (nasal process of premaxilla narrow, 1/24 of one step under both ACCTRAN and DEL- TRAN), 32.0 (palatine teeth lost, 8/24 of one step under both ACCTRAN and DELTRAN), 44. 1 (cal- cified cartilage anterior border of scapular fenestra absent, 3/24 of one step under both ACCTRAN and DELTRAN), and 95.4 or 95.5 (Pro, 0. 1 1 of one step under ACCTRAN, 0.41 under DELTRAN). This stem may be further supported by two ambiguously placed transformations, both of which were discov- ered under ACCTRAN optimization: 89.3 (H-LDH, fixed) and 94.3 (GOTs, fixed). Stem K is weakly supported by two unambigu- ously placed transformations, neither of which is fixed: 45.1 (clavicular fenestrations present, 4/24 of one step under both ACCTRAN and DELTRAN) and 83.1 (enlarged melanic axillary patches present, 18/24 of one step ACCTRAN, 15/24 of one step DELTRAN). The stem may be further supported by four ambiguously placed transformations: 70.0 (paired melanic keels absent from ventral caudal extremity, 13/24 of one step ACCTRAN), 89.3 (H-LDH, fixed DELTRAN), 94.3 (GOTs, fixed DELTRAN), and 95.4 (Pro, 0.1 1 of one step DEL- TRAN). Stem L is supported by seven unambiguously placed synapomorphies, three of which are fixed: 73.1 (posterior collars widely separated), 75.3 (pos- terior collar terminates ventrally before entering the antehumeral fold), and 84. 1 (olive green ventrolat- eral coloration present). Olive green ventrolateral coloration is not a fixed state in C. vestigium as northern populations are characterized by burnt- orange ventrolateral coloration. This was an artifact of the multistate character coding scheme employed in this analysis and resulted at least in part because a satisfying estimate of the frequencies of the orange and green ventrolateral conditions in C. vestigium could not be obtained from preserved material. The four unambiguous but polymorphic transforma- tions include: 16.1 (acquisition of extravomerine bones, 9/24 of one step under both ACCTRAN and DELTRAN), 26.1 (bony shelf between medial pro- cess of surangular and ramus of the mandible pres- ent, 5/24 of one step under both ACCTRAN and DELTRAN), 31.1 (number of premaxillary teeth, 0.14 of one step under both ACCTRAN and DEL- TRAN), and 45.0 (clavicular fenestrations lost, 2/24 of one step under both ACCTRAN and DEL- TRAN). Finally, four ambiguously placed transfor- mations may also support this node: 68.3 (presence of slender, transversely arranged, white dorsal stripes, fixed ACCTRAN), 91.1 (6-PGD, 0.12 of one step ACCTRAN), 95.3 (Pro, 0.29 of one step ACCT- RAN), and 96.1 (Est 1,0.71 of one step ACCTRAN). In an attempt to assess the amount of character support for each clade, bootstrap and decay index analyses were performed for the combined data set. From each of these analyses it is clear that a number of clades are rather unstable. For example, in the bootstrap analysis (Fig. 40), nodes G and H were supported in less than 50 percent of the bootstrap replications. Stems D and F also were found to be relatively weakly supported with bootstrap propor- tion values of 66 and 60, respectively. Despite its low bootstrap P value, stem F is supported by four fixed, unambiguous synapomorphies. The amount of support for stem D may be underestimated be- cause a number of derived character states were coded as missing (“?”) for the fossil taxon G. co- rona|, thus rendering these transformations ambig- uous when both ACCTRAN and DELTRAN op- timization routines are considered. Notably, when G. corona t is excluded from the analysis, the boot- strap P value of stem D is 1 00. The remaining clades appear to be well supported, given that Hillis and Bull (1993) found bootstrap proportions to be con- servative estimates of stem support with P values > 70 corresponding to a probability > 0.95 that the corresponding clade is real. The results of the decay index analysis (Fig. 41) 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 63 in parentheses indicates the number of additional steps that must be considered before the node is no longer supported. The num- ber in parentheses represents the number of equally parsimonious trees discovered when the given number of additonal steps are permitted. agree with those of the bootstrap analysis in sug- gesting that a number of clades (nodes D, F, G, H, and K) are relatively unstable. Particularly well-sup- ported clades appear to be stems B ( Gambelia ), E ( Crotaphytus ), and L (C. insularis + C. vestigium). The allozyme and morphology data sets are not entirely consistent with one another in that the allo- zyme data suggest that C. dickersonae shares a com- mon ancestor with C. collaris, C. nebrius, and C. reticulatus, whereas the morphological data suggest that C. dickersonae is more closely related to C. vestigium and C. bicinctores. The much smaller allo- zyme data set (ten characters) seems to contain less phylogenetic signal than does the morphology data set. For example, the differential between the ob- served gi and the critical gx value for random data is substantially greater for the morphological data set than it is for the allozyme data set (criterion suggested by J. Huelsenbeck as noted in Wiens [1995]). Nevertheless, because the topology of the single most parsimonious tree is unaffected by the inclusion or exclusion of the allozyme data, the rel- ative phylogenetic informativeness of the allozyme data is not a critical issue. However, the bootstrap results for both the morphology-only and combined analyses should be considered when evaluating to- pology robustness for the single most parsimonious tree. DISCUSSION Comparison with Previous Hypotheses The results of this analysis agree with those of Etheridge and de Queiroz (1988), Frost and Ether- idge (1989), and virtually every other study that has considered the systematics of this group in that Cro- taphytus and Gambelia are found to be sister taxa. The intrageneric relationships also are largely con- sistent with previous hypotheses with some notable exceptions. A major distinction between this anal- ysis and all previous studies is the complete repre- sentation of species included here, several of which were undiscovered or were not known to be distinct lineages at the times of the previous analyses. A phenetic analysis (Ward’s Minimum Variance Cluster Analysis; Wishart. 1968) of unspecified morphological data performed by Smith and Tanner (1972) provided the first estimate of interspecific relationships within Crotaphytus (exclusive of C. reticulatus). They concluded that there were two clusters of taxa within their study group, the collaris complex, composed of C. collaris populations, and the western complex, composed of C. bicinctores, C. vestigium, C. insularis, and C. dickersonae. The two clusters are consistent with the results presented here, as both groups appear to be monophyletic. Smith and Tanner (1974) performed another phe- 64 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 C. insularis C. vestigium C. bicinctores C. reticulatus C. dickersonae C. nebrius C. collaris Fig. 42. — The single tree discovered by Montanucci et al. (1975) in their analysis of crotaphytid relationships. netic analysis of Crotaphytus relationships (again without considering Crotaphytus reticulatus ). The Ward’s Minimum Variance Cluster Analysis (Wis- hart, 1968) employed morphometric and color pat- tern data. Their results were consistent with those of their 1972 study, although they were more specific in their assessment of relationships in this later anal- ysis. They discussed the interspecific relationships of the western complex species and recognized two pairs of sister taxa, (C. dickersonae + C. bicinctores ) and (C. vestigium + C. insularis). Their tree indi- cates that they were unsure whether the western complex was monophyletic or if (C. bicinctores + C. dickersonae) was actually the sister taxon of C. collaris (= the collaris complex). Their phyletic tree suggested Gambelia (= G. wislizenii) to be the sister taxon of Crotaphytus, and C. reticulatus to be the sister taxon of the remainder of Crotaphytus. How- ever, data were not presented for these species and it is therefore unclear how these conclusions were reached. The phylogenetic conclusions of this anal- ysis agree in most respects with those of the present study except in the placement of C. dickersonae, which was found to be the sister taxon of C. grismeri, C. bicinctores, C. vestigium, and C. insularis in this analysis. Montanucci et al. (1975) performed the first cla- distic analysis of Crotaphytus, utilizing 1 2 allozyme, discrete morphological, and morphometric char- acters. Their analysis of these data (using the Wagner program, Kluge and Farris, 1969) resulted in the tree depicted in Figure 42. This tree is similar to those discovered here in the placement of C. bi- cinctores as the sister taxon of (C. vestigium + C. insularis). However, their tree differs from the trees discovered here in the placement of C. dickersonae as the sister taxon of (C. nebrius + C. collaris ), in the placement of C. reticulatus as the sister taxon of this group, and in placing (C. bicinctores (C. ves- tigium + C. insularis)) as the sister taxon of (C. reticulatus (C. dickersonae (C. nebrius + C. collar- is))). As with the previous analyses, several taxa could not be included, such as C. grismeri (not yet recognized as a distinct lineage) and C. antiquus (yet to be discovered). Few comparisons can be drawn between the re- sults of this analysis and those of previous studies regarding the phylogenetic relationships of Gam- belia. Those previous workers who recognized G. silus as a distinct species generally assumed it to be the sister taxon of G. wislizenii. Only Norell (1989) attempted to elucidate the phylogenetic relation- ships of Gambelia and he was primarily interested in the position of G. corona |. Although Norell (1989) described a number of useful characters, he was un- able to provide phylogenetic resolution. A distinc- tion between this analysis and several others relates to the evolution of G. silus. Some previous workers suggested that G. silus may have evolved as recently as 1 1,000 years ago by peripheral isolation (Mon- tanucci, 1967, 1970; Tollestrup, 1979), although Montanucci (1970) also entertained the possibility that G. silus entered the valley much earlier. Re- gardless of the timing of the event, Montanucci (1970) suggested that differences between G. silus and G. wislizenii are examples of derived character states in G. silus rather than derived characteristics of G. wislizenii. However, most of these features, such as the presence of territoriality, a truncated snout, and sexual dimorphism wherein males are larger than females, are more parsimoniously inter- preted as plesiomorphic retentions in G. silus. This interpretation is consistent with that of Tollestrup (1983), at least with respect to the loss of territori- ality in G. wislizenii. Thus, it appears that G. silus is a relatively plesiomorphic taxon and not a re- cently derived offshoot of G. wislizenii. It may seem counterintuitive that a narrowly dis- tributed peripheral species such as Gambelia silus would be relatively plesiomorphic in comparison with a wide-ranging taxon such as G. wislizenii (plus its sister taxon, G. copei). However, there are ex- 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 65 amples of this phenomenon discussed in the her- petological literature. For example, Lynch (1982) found that the widely distributed species Cerato- phrys cornuta exhibits numerous autapomorphies, while its close relatives, C. calcarata and C. stolz- manni are peripherally isolated and exhibit no known autapomorphies. Wiens (1993 b) discussed a similar situation in Urosaurus. Urosaurus gadovi has a very restricted distribution in the Balsas-Tepalcatepec valley, Michoacan, Mexico, in comparison with its widely distributed sister taxon, U. bicarinatus. Yet, U. bicarinatus is relatively derived with several au- tapomorphies, while U. gadovi is relatively plesiom- orphic and has no fixed autapomorphies. Character Evolution Several evolutionary trends in the morphology and ecology of crotaphytids can be addressed in the context of the recovered phylogeny. These include the correlation between head morphology and sau- rophagy, the evolution of sexual dichromatism and morphologies that appear to be display oriented, bipedalism and the evolution of morphologies as- sociated with this form of locomotion, and the func- tion of gravid coloration and the evolution of similar coloration in subadult males. Head Morphology and Dietary Correlates. — Head morphology and dietary preferences appear to be related in crotaphytids. Within Gambelia, G. copei and G. wislizenii share the derived condition of an elongate head, while G. situs retains the plesiom- orphic blunt-snouted condition. Several studies, particularly those of Tollestrup (1979, 1983), sug- gest that G. wislizenii preys on vertebrates much more heavily than does G. silus and, based on my observations of stomach contents both in the field and in museum specimens, I suggest that G. copei will prove to be just as reliant on vertebrates as is G. wislizenii. A similar correlation is apparent in Crotaphytus. Crotaphytus reticulatus, C. collaris, C. nebrius, and C. antiquus have relatively broad heads with blunt snouts in contrast with the narrower, more elongate heads of C. dickersonae, C. grismeri, C. bicinctores, C. vestigium, and C. insularis (which form a monophyletic group; Fig. 37). The majority of the published dietary studies related to Crota- phytus have been confined to C. collaris, which is primarily insectivorous (Fitch, 1956, plus numerous additional references). Examination of preserved specimens with slit bellies and the skeletal prepa- ration of preserved and fresh material has allowed for numerous observations of stomach contents, al- though precise records have not been maintained. These observations suggest that the “long snout” clade specializes in vertebrate prey to a greater de- gree than C. reticulatus, C. collaris, C. nebrius, and presumably C. antiquus. The saurophagous species may have elongate heads to allow for faster jaw adduction and predation on fast-moving prey, whereas the short-snouted condition might be as- sociated with more powerful jaw adduction for crushing hard-shelled prey, perhaps certain insect taxa. A detailed dietary analysis to confirm these anecdotal observations for Crotaphytus, followed by an analysis of the functional morphology of crota- phytids (using kinematic and strain gauge analyses to measure jaw speed and jaw adductor power) would shed much light on this situation. The Evolution of Display-oriented Morphologies in Males.— Gambelia and Crotaphytus reticulatus essentially lack sexual dichromatism outside of the breeding season, whereas the remaining species of Crotaphytus are characterized by the derived con- dition of strong sexual dichromatism throughout the year. This is the first in a series of evolutionary modifications presumably associated with an in- crease in display-oriented morphologies within males. There appears to have been selection for black coloration within a number of clades, the best ex- ample of which is associated with the evolution of inguinal patches in adult males. Inguinal patches appear to have passed through the following trans- formation series: absent — * small — ► large, with a reversal to the polymorphic condition observed in C. collaris. The common ancestor of Crotaphytus exclusive of C. reticulatus appears to have been fixed for the presence of small inguinal patches. This con- dition persists in C. antiquus and C. nebrius, and appears to have been elaborated upon to produce much larger inguinal patches in the common an- cestor of C. dickersonae, C. grismeri, C. bicinctores, C. vestigium, and C. insularis (Fig. 34, 35). The inguinal region is prominently displayed by male Crotaphytus regardless of whether or not they have inguinal patches and this may have led first to the acquisition and then enlargement of inguinal patch- es. If this scenario holds true, a reversion to the polymorphic state in C. collaris is puzzling. Two additional examples of derived black components of the color pattern are the jet-black femoral pores of C. reticulatus and C. antiquus (Fig. 23) and the enlarged, melanic axillary patches present poly- morphically in C. collaris, C. nebrius, C. bicinctores, C. vestigium, and C. insularis. 66 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 An additional series of evolutionary modifica- tions that is presumably associated with male dis- play behavior are associated with lateral tail com- pression in the common ancestor of C. dickersonae, C. grismeri, C. bicinctores, C. vestigium, and C. in- sular is (Fig. 3 IB. 32A-D). This character complex includes the derived acquisition of dorsal and ven- tral caudal fat bodies as well as modifications of the neural and haemal arches and transverse processes of the caudal vertebrae. Lateral tail compression presumably increases the apparent size of adult males in lateral view. The evolution of sexual dichroma- tism, the acquisition and modification of black color pattern components that are restricted to males, and the development of lateral tail compression in males each suggest an increase in the importance of male display within Crotaphytus. Bipedalism.— The form of bipedalism present in Crotaphytus appears to be unique among iguanians (see below). Several morphological modifications within the genus appear to be related to this behav- ior, including the loss of autotomic fracture planes of the caudal vertebrae (character 39), the modifi- cation of the skin of the distal portion of the tail such that the skin may easily slip free (character 52; Fig. 34), the acquisition of lateral tail coiling be- havior (character 87; Fig. 36), and the contact of the medial and lateral plantar tubercles of the fifth meta- tarsal such that they form an arch (character 45; Fig. 1 7). The reference to the last character requires some explanation. Snyder (1952, 1954, 1962) observed that M. gastrocnemius is usually slightly larger in bipedal lizards than in quadrupedal species. Al- though he emphasized that the differences in muscle mass between quadrupedal and bipedal lizards are not usually great, he noted that M. gastrocnemius was conspicuously larger in Crotaphytus than in any other quadrupedal or bipedal lizard that he exam- ined (Snyder, 1962). Because M. gastrocnemius in- serts on the medial and lateral plantar tubercles, it is possible that the arch structure found in Crota- phytus increases the surface area for insertion of this muscle. Crotaphytus utilizes a unique form of bipedal lo- comotion, wherein individuals jump bipedally from rock to rock on the boulder-strewn hillsides that they inhabit. This saltatory form of bipedalism al- lows them to move rapidly over a complex substrate and, presumably, an individual would be at a dis- advantage if it were not able to maintain a bipedal gait. Snyder (1949, 1954, 1962) found that the tail of Crotaphytus acts as a counterbalance during bi- pedal locomotion and that the removal of between 25 and 33 percent of the tail prohibits a bipedal gait for more than three to five strides, while the removal of more than 50 percent prevents bipedal locomo- tion for more than one step. This may have been the selective factor that lead to the loss of autotomic fracture planes in the common ancestor of Crota- phytus. However, the tail of Crotaphytus is very long and it seems likely that there would be strong se- lective pressure to prevent predators from capturing them by this appendage, especially given that the tail cannot be broken easily (tail breakage can still occur, but requires an intervertebral separation or a fracture of the caudal vertebra itself; Etheridge, 1967). At least two evolutionary modifications have occurred in Crotaphytus that appear to play a role in minimizing predation by “tail capture.” First, the lateral tail-coiling behavior utilized by Crotaphytus when taking refuge from predators beneath rocks or surface debris, during hibernation, and when resting beneath stones appears to function as a means of keeping the tail out of the reach of potential pred- ators. Second, the presence of loosely adherent skin over the distal approximately 20 percent of the tail allows the skin of the caudal terminus to slip off when grasped, thus providing an alternative to cau- dal autotomy over the portion of the tail the lizard can lose without hindering its ability to run bipe- dally. Once the skin is removed, the underlying ver- tebrae and soft tissues wither and are lost. This hy- pothesis for the function of the loosely adherent caudal skin is based on three separate instances in which I experienced this phenomenon while at- tempting to capture lizards, as well as on the ob- servation of numerous museum specimens that lack the skin of the distal portion of the tail. Gravid Coloration.— Gravid coloration occurs in all crotaphytid taxa and a similar color pattern de- velops in subadult male Crotaphytus (character 59). Although gravid coloration itself may be a plesiom- orphic retention of Crotaphytidae, the subadult male coloration is almost certainly derived (see above). The coloration of the subadult males, which devel- ops soon after hatching and fades just before ma- turity (Fitch, 1956; Rand, 1986) is virtually indis- tinguishable both in terms of its chromatic char- acteristics and in its anatomical placement and, therefore, suggests that the young males may be mimicking females in order to incur some selective benefit. In fact, the presence of this coloration in young males may provide a clue to its function both in females and in the subadult males themselves. 1996 McGUIRE— SYSTEM ATICS OF CROTAPHYTID LIZARDS 67 The presence of bright red or orange dorsal pig- mentation makes gravid Crotaphytus conspicuous at a time when crypticity presumably would be at a premium. Therefore, it is likely that the coloration provides some form of visual signal to predators or conspecihcs that provides a greater selective benefit than cost to gravid females. The behavior of repro- ductive females toward males changes dramatically from submissive to aggressive soon after copulation, and this corresponds with an intensification of the coloration (Fitch, 1956; Clarke, 1965; Cooper and Crews, 1988). Therefore, Clarke (1965) and Cooper (1988) suggested that gravid coloration may act as an inhibitor of male aggression. If this is the case, subadult males with red or orange coloration po- tentially could benefit by being allowed to forage within adult male territories without being attacked. Indeed, Gambelia are well known for their canni- balistic habits and such coloration in Crotaphytus may limit predation on subadults by adult males. Because females are generally allowed to set up ter- ritories within male territories in many territorial species (Stamps, 1977; noted in C. collaris by Fitch, 1956, and Yedlin and Ferguson, 1973), subadult females potentially would benefit less by bearing red or orange dorsal coloration. If this is the case in Crotaphytus, the presence of bright red or orange coloration in subadult females might more likely be selected against (assuming the presence of vibrant orange or red coloration leaves them more conspic- uous to visually oriented predators such as raptors and loggerhead shrikes). Although this hypothesis is highly speculative, it is consistent with the idea that gravid coloration has a functional value in fe- males on which subadult males could also capitalize. TAXONOMIC ACCOUNTS The following taxonomic accounts include: (1) synonymies for each taxon name, (2) phylogenetic definitions for the three clade names (Crotaphyti- dae, Crotaphytus, and Gambelia ) following the rec- ommendations of de Queiroz and Gauthier (1992), (3) an etymology for each taxon, (4) a general de- scription of squamation for Crotaphytidae, (5) a more specific description of squamation for each species, (6) general descriptions of coloration in life for Crotaphytus and Gambelia, (7) more specific de- scriptions of coloration for each species, (8) a de- tailed summary of geographic distribution for the genera and species (locality data used in producing the distribution maps are available from the author upon request), (9) a discussion of natural history where appropriate, and ( 1 0) a remarks section under each species account that includes references to il- lustrations, as well as various additional comments. The list of published illustrations may be complete for the rarer taxa, but is certainly incomplete for wide-ranging, common species such as C. collaris and G. wislizenii. Natural history observations that are not followed by a literature citation are my own. Crotaphytidae Smith and Brodie, 1982 Crotaphytinae Smith and Brodie, 1982:106. Type genus: Cro- taphytus Holbrook, 1842. Crotaphytidae Frost and Etheridge, 1989:36. Definition. —Crotaphytidae is here defined as a node-based name for the most recent common an- cestor of Crotaphytus and Gambelia and all of its descendants. Description.— A description of the squamation of crotaphytids is given here to provide a consolidated view of those features common to the family. To prevent an unnecessary duplication of information, only variable features will be discussed under the separate species accounts. General color pattern descriptions are provided under the generic accounts of Crotaphytus and Gam- belia, with more specific characterizations given under each spe- cies account. Dorsal cephalic scales smooth, convex, polygonal, occasionally with numerous inconspicuous surface irregularities. Rostral ap- proximately two to four times wider than high, usually rectan- gular in shape. Rostral bordered by two to eight postrostrals. Remaining snout scales irregularly arranged, an enlarged mid- dorsal series may be present. Nasals form a thin-walled ring, pierced centrally by external nares; nares face laterally at a slight dorsal angle; nasals separated by three to nine intemasals. Fron- tonasals occasionally enlarged. Three or four canthals, posterior one or two wider than high; four to ten scales separate canthals of left and right sides; canthus rostralis forms prominent ridge. Supraorbital semicircles present or absent; when present some scales may fuse to form azygous frontals. Interparietal small, approximately twice as long as wide, with opalescent “eye.” Pa- rietals generally small and irregular. Supraoculars small, flat or convex, smooth, becoming progressively larger medially such that medial scales are two to four times larger than lateral ones. Circumorbitals present or absent; when present not well differ- entiated from supraoculars. Superciliaries six to 15, extremely elongate medial scale present or absent; anterior scales with oblique sutures oriented posterodorsally, posterior scales with oblique sutures oriented anterodorsally. Palpebrals ovoid, slightly con- vex, may be interspersed with numerous interstitial granules. Inner ciliaries deeper than wide, outer ciliaries of upper eyelid usually projecting, anterior and posterior ones projecting slightly 68 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 further than medial ones; outer ciliaries of lower eyelid larger than those of upper lid, strongly projecting, conical, with anterior and posterior scales projecting slightly further than medial ones. Preoculars, suboculars, and postoculars form an arc of four to 13 rectangular scales, second, third, or fourth scale elongate or not, all with strong superior keel, strongly concave below keel. Supralabials 11 to 18, usually slightly longer than high except anteriormost scale, which is square or pentagonal. Supralabials followed posteriorly by a series of elongate postlabials. Lorilabials in one to four rows, ovoid to rectangular, juxtaposed, separating supralabials from suboculars and nasals. Loreals numerous, larg- er than adjacent lorilabials. Lower temporals small, convex, oval, often separated by interstitial granules; zone of less convex, po- lygonal or rounded, juxtaposed scales approximately 1.5 to two times larger than bordering upper and lower temporals, extending posteriorly from postoculars but not reaching external auditory meatus; corresponding to underlying postorbital bones. Aperture of external auditory meatus rectangular or ovoid, often constrict- ed at or above the midpoint, approximately two to four times higher than wide, with small, strongly convex, somewhat conical auricular scales lining anterior margin. Mental pentagonal, one to 1.5 times wider than high, bordered laterally by anterior in- fralabials and posteriorly by a pair of large postmentals. Post- mentals may or may not be separated from infralabials by sub- labials. Chinshields weakly differentiated or undifferentiated. In- fralabials ten to 18, square or wider than high, inferior border convex. Gulars granular, strongly convex and beadlike or flat, each scale may be separated from adjacent scales by numerous asymmetrically arranged interstitial granules. Gulars flattened and discoid in gular pouch region. Gulars within symphysial groove much smaller than surrounding scales that overlie man- dibles. Dorsal scales of neck and body very small, rounded, strongly convex, nonimbricate, each characteristically surrounded by six interstitial granules giving appearance of a six-pointed star. Me- dian dorsal scales 1 . 5 to two times larger than lateral dorsal scales. Dorsals grade smoothly into ventrals, approximately 136 to 224 rows encircle body midway between forelimb and hindlimb in- sertions. Ventrals smooth, flat, varying from oval to rhombic in shape, approximately three to four times larger than adjacent laterals, occasionally slightly imbricate. Tail long, cylindrical to oval over entire length or anterior one- half strongly compressed laterally. Caudals usually keeled over distal 85 percent, keeling more pronounced distally. Paired, me- dian row of subcaudals larger than adjacent subcaudals and lateral caudals present or absent; posteriorly subcaudals become pro- gressively more distinctly keeled and often mucronate. Enlarged postanal scales in males present or absent, scales between post- anal plates and cloaca extremely small compared to remaining subcaudals. Scales in immediate vicinity of forelimb insertion minute, ex- cept for a patch of large, discoid scales at anterior forelimb ar- ticulation. Suprabrachials discoid, separated by interstitial gran- ules, becoming larger and slightly imbricate distally; distal su- prabrachials approximately two times larger than dorsal body scales. Suprabrachials grade smoothly into smaller postbrachials. Prebrachials convex, beadlike, each surrounded by six symmet- rically arranged interstitial granules; prebrachials grade abruptly into smaller, convex infrabrachials. Supra-antebrachials and pos- tantebrachials small, discoid, nonoverlapping proximally, prean- tebrachials slightly imbricate proximally; supra-antebrachials, preantebrachials, and postantebrachials much larger and strongly imbricate adjacent to supracarpals. Infra-antebrachials convex, smaller than adjacent preantebrachials and slightly smaller than postantebrachials. Supracarpals large, strongly imbricate, contin- uous with large supradigital scales. Proximal supradigitals wider than long. Infracarpals strongly imbricate, usually with three strong mucrons. Subdigital lamellae moderately imbricate, each with three to six short mucrons. Deep postfemoral dermal mite pocket may or may not be present at hindlimb insertion. Suprafemorals small, convex, near- ly equal in size to lateral dorsals, separated by numerous inter- stitial granules, grading into prefemorals. Prefemorals becoming more discoid, slightly imbricate and larger distally; prefemorals at knee larger than surrounding scales, five to ten times larger than suprafemorals. Prefemorals grade into smaller infrafemo- rals; 1 5 to 3 1 femoral pores, femoral pores extend beyond angle of knee or not, separated medially by ten to 26 granular scales. Suprafemorals grade smoothly into minute, convex, oblong post- femorals, interspersed with interstitial granules. Supratibials small, convex, grade into larger, flattened, juxtaposed posttibials and larger, similarly shaped pretibials; pretibials granular where ad- jacent to supratarsals. Infratibials smooth, flat, juxtaposed or weakly imbricate proximally, becoming imbricate distally, much larger than adjacent pretibials and slightly larger than posttibials. Supratarsals large, imbricate anteriorly, slightly convex, granular posteriorly. Infratarsals strongly imbricate, one to three keels per scale. Supradigital scales smooth, large, strongly imbricate. Sub- digital scales imbricate, with three to seven keels, each with a terminal mucron; subdigital lamellae on fourth toe 15 to 25. Size. — All crotaphytid species are sexually di- morphic; however, males are larger than females in some species while the reverse relationship pertains in others. Maximum adult sizes range from ap- proximately 99 mm SVL in male Crotaphytus gris- meri to approximately 1 44 mm SVL in adult female Gambelia wis/izenii. Crotaphytus Holbrook Crotaphytus Holbrook, 1842:79. Type species (by original des- ignation): Agama collaris Say 1823. Leiosaurus, part— Dumeril, 1856:532. Crotaphytes— Stone and Rehn, 1903:30. Definition. — Crotaphytus is defined as a node- based name for the clade stemming from the most recent common ancestor of Crotaphytus collaris and all species that are more closely related to that spe- cies than to Gambelia. Etymology. — From the Greek krotaphos, referring to the side of the head or temple region; and phyton, a creature or animal. The name apparently refers to the hypertrophied jaw adductor musculature of these lizards. Coloration in Life. — Dorsal body coloration is ex- tremely variable within adult male Crotaphytus, ranging between cobalt blue, aquamarine, green, turquoise green, golden tan, straw yellow, brown, and gray. Females of all species except C. reticulatus are generally characterized by a more faded version of the color present in males of their species or by 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 69 pale tan or green hues. Head coloration in males may differ from that of the body, most notably in those populations of C. collaris characterized by a pale yellow to fluorescent yellow head. A pattern of white reticulations is a recurring phenomenon with- in the genus and may be present over the entire dorsal surface of the body and limbs, as well as on the temporal and superficial mandibular regions or some subset thereof. A number of species have a dorsal body and limb pattern composed of white spots or dashes rather than net-like reticulations, and narrow, transverse dorsal bars may be present. A broad white or off-white vertebral stripe may ex- tend from the base of the tail posteriorly for most of its length. The dorsal surface of the head may be pale-colored, with a more or less patternless surface. All Crotaphytus are characterized by a ventral col- oration of white, off-white, or pale yellow, although additional markings may be present. Olive green, golden orange, or burnt orange ventrolateral col- oration may be present in males as well. The tail may or not be bright lemon yellow in adult females or burnt orange in subadult females. Gular coloration in adult males is highly variable with olive green, gun-barrel blue, slate gray, dark brown, dark blue, turquoise blue, yellow, or orange all characterizing the adult males of certain popu- lations. The gular region of females is generally white or only faintly patterned. The gular coloration of adult males may or may not include a black central component. The pattern surrounding the gular re- gion of adult males is also variable and may be composed of pale reticulations, white spots on a sky blue background, or radiating, obliquely oriented, white lines. Black is an important color component within the genus with all species having some combination of black markings. All Crotaphytus except some female C. insularis and C. reticulatus have at least one pair of black collar markings and most have two pairs. The anterior and posterior collar markings are sep- arated by a broad white bar that may or may not be complete middorsally. The anterior pair of collar markings contact ventrally through the gular fold in adult males of some species. The posterior collar markings may contact middorsally in some species as well. A pair of black spots may be present mid- dorsally between the anterior collar markings. A pair of enlarged melanic axillary patches are variably present immediately posterior to the forelimb in- sertion in adult males of several species. Small or large melanic inguinal patches are also present in the adult males of several species. All Crotaphytus neonates are characterized by a pattern of white reticulations, some of which enclose black pigments. This pattern may or may not be retained into adult- hood with little modification. The femoral pores are generally off-white to gray in color but are black in males of two species (C. antiquus and C. reticulatus). Paired, melanic keels may or may not be present on the ventral surface of the caudal extremity. All Crotaphytus females develop “gravid color- ation” in the form of red or orange lateral bars or spots. A similar pattern develops in subadult males of all Crotaphytus species. Size. — All Crotaphytus exhibit sexual dimor- phism wherein males are larger than females. Max- imum adult sizes range from approximately 99 mm SVL in C. grismeri to approximately 1 3 1 mm SVL in C. collaris (C. reticulatus may reach 1 37 mm SVL; Montanucci, 1976). Distribution.— Western and southcentral United States from southern Idaho and eastern Oregon southward and eastward across the southern Great Plains into Missouri, northwestern Arkansas, and extreme northwestern Louisiana, southward into southern Baja California and northcentral mainland Mexico. Fossil Record. —Numerous Pleistocene fossils have been referred to the genus, all of which have been placed within C. collaris or listed as C. sp. (Estes, 1983). However, the localities from which some of these specimens have been collected suggest that a few of these fossils may be C. bicinctores and C. nebrius (Brattstrom, 1954; Van Devender et al., 1977; Van Devender and Mead, 1978). The frag- mentary nature of most of the material renders spe- cific identification on the basis of character evidence impossible. Crotaphytus antiquus Axtell and Webb (Fig. 30D) Crotaphytus antiquus Axtell and Webb, 1995:1; fig. 1, 2. Type locality: “2.1 km N-1.7 km E Vizcaya (25°46'04"N- 103°1 1 '48"W, el 1 100± m) in the Sierra Texas, Coahuila, Mex- ico” (Holotype: UTEP 15900). Etymology. — From the latin antiquus, meaning old or of an- tiquity. The name was chosen by the authors because it “incor- porates (their) interpretation regarding the probable ancientness of the lizard.” Diagnosis.— Crotaphytus antiquus can be distin- guished from all other Crotaphytus by the presence of gravid coloration that is limited to the anterior 15 to 50 percent of the portion of the abdomen between the forelimb and hindlimb insertions and a much larger total number of white reticulations 70 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 that enclose melanic pigments. Crotaphytus antiq- uus can be distinguished further from C. reticulatus on the basis of its postfemoral mite pockets, sexual dichromatism of the dorsal color pattern such that females have much more subdued coloration than males, and the presence of paired, melanic mucrons on the distal subcaudal scales. It can be distin- guished from all Crotaphytus except C. nebrius and western populations of C. collaris by its small me- lanic inguinal patches (patches absent in C. reticu- latus, patches much larger and extending onto the ventral surface of the abdomen in C. bicinctores, C. dickersonae, C. grismeri, C. insularis, and C. vestig- ium. It can be distinguished further from all Cro- taphytus except C. reticulatus by its dorsal pattern composed of a white, net-like reticulum, some of which enclose melanic pigments. It can be distin- guished further from all other Crotaphytus except C. reticulatus and C. insularis by the the weakly defined collar markings of females. It can be distin- guished further from C. collaris by its ventrally com- plete anterior collar marking in adult males. It can be distinguished further from C. collaris and C. ne- brius by the presence in adult males of black pig- ments extending from the gular fold anteriorly into the central gular area. From C. dickersonae, C. bi- cinctores, C. grismeri, C. insularis, and C. vestigium, it can be distinguished further by its round, rather than laterally compressed, tail that lacks a white dorsal vertebral stripe (present in adult males of the latter five species). Finally, from C. bicinctores, C. grismeri, C. insularis, and C. vestigium, C. antiquus can be distinguished by its black buccal lining. In addition to the characters listed above, C. an- tiquus can usually be distinguished from all other Crotaphytus (with the possible exception of C. dick- ersonae) on the basis of a series of scales that either completely separates or nearly separates the supra- orbital semicircles. In nine of 16 C. antiquus, the supraorbital semicircles are separated by a row of small scales, while in six of 16 specimens, a single pair of scales is in contact, and in one specimen, two scales are in contact. In all other Crotaphytus except C. dickersonae, at least two scales of the su- praorbital semicircles were in contact and this was a relatively rare condition (more than two scales in contact in six of eight C. bicinctores, 26 of 27 C. collaris, four of four C. grismeri, 15 of 20 C. insu- laris, nine of ten C. nebrius, eight of eight C. reti- culatus, and six of seven C. vestigium). Crotaphytus dickersonae is considered most similar with respect to this character to C. antiquus only because one specimen had one pair of scales of the semicircles in narrow contact and three of four additional spec- imens had two scales in contact. Thus, the prevalent condition of C. antiquus (supraorbital semicircles completely separated by a row of scales) was not observed in any other species of Crotaphytus, al- though the condition may very well occur in C. dickersonae given a larger sample size. Thus, C. an- tiquus and other Crotaphytus overlap but little with respect to this feature. Variation (n = 19). — Rostral approximately four times wider than high, usually rectangular in shape. Rostral bordered by three to six postrostrals. Re- maining snout scales irregularly arranged, an en- larged middorsal series may be present. Nasals sep- arated by five to six internasals. Frontonasals oc- casionally enlarged. Canthals three; five to seven scales separate canthals of left and right sides. Su- praorbital semicircles present with 12 to 13 scales per semicircle, median scales never fuse to form azygous frontals, a series of small scales may sep- arate the right and left supraorbital semicircles or one, or rarely two, of the scales of the semicircles may be in contact. Supraoculars flat or convex, smooth, becoming progressively larger medially such that medial scales are 1.5 to two times larger than lateral ones. Circumorbitals present, not well dif- ferentiated from supraoculars. Superciliaries eight to 1 1, extremely elongate medial scale usually pres- ent. Palpebrals ovoid, slightly convex, interspersed with numerous interstitial granules. Preoculars, su- boculars, and postoculars form an arc of seven to 1 1 rectangular scales, second, third, or fourth scale not elongate. Supralabials 12 to 16, usually slightly longer than high. Lorilabials in two to three rows, ovoid to rectangular, juxtaposed, separating su- pralabials from suboculars and nasals. Aperture of external auditory meatus rectangular or ovoid, often constricted at or above the midpoint, approximately two to four times higher than wide, with small, strongly convex, somewhat conical auricular scales lining anterior margin. Mental pentagonal, one to 1.5 times wider than high, bordered laterally by an- terior infralabials and posteriorly by a pair of large postmentals. Postmentals sometimes separated from infralabials by sublabials; mental occasionally con- tacted by one or two sublabials. Chinshields weakly differentiated or undifferentiated. Infralabials 13 to 16, square or wider than high, inferior border con- vex. Gulars granular, strongly convex and beadlike, each scale separated from adjacent scales by nu- merous asymmetrically arranged interstitial gran- ules. Dorsal scales in approximately 128 to 161 rows 1996 McGUIRE— SYSTEM ATICS OF CROTAPHYTID LIZARDS 71 midway between forelimb and hindlimb insertions. Tail long, cylindrical in both sexes and all age groups. Paired, median row of subcaudals may or may not be larger than adjacent subcaudals and lateral cau- dals. Enlarged or slightly enlarged postanal scales present in males. Deep postfemoral dermal mite pocket present at hindlimb insertion. Femoral pores 16 to 20, femoral pores do not extend beyond angle of knee, separated medially by 19 to 25 granular scales. Subdigital la- mellae on fourth toe 18 to 22. Coloration in Life.— Males of this species are characterized by a dorsal color pattern consisting of a thick white reticulum on a dark brown field. The reticulations differ from those of C. reticulatus in that they are thicker, and all, or nearly all, of the dorsal body reticulations enclose black pigments. A few of the forelimb and hindlimb reticulations may also enclose black pigments. As in C. reticulatus, the reticulum is present on nearly the entire dorsal sur- face including the body, the anterior half of the tail, all four limbs, the lateral surface of the head, and the superficial mandibular area. The anterior and posterior collar markings are better developed than those of C. reticulatus and the anterior collar is com- plete ventrally. Black pigmentation is present in the central gular region, as in all other adult male Cro- taphytus except C. collar is and C. nebrius. The dor- sal surface of the head is patternless, but it is not of paler coloration than the remaining dorsal surfaces, as is usually the case with C. dickersonae, C. bi- cinctores, C. grismeri, C. vestigium, and C. insularis. Small inguinal patches largely confined to the prox- imal ventral surface of the thigh are present in all adult males. The femoral pores are jet black. The coloration of females is less vibrant than that of males. The dorsal base color is grayish brown, the white reticulum is not as bright, the dorsal re- ticulum encloses dark gray pigments rather than black, the femoral pore exudate is gray, and the melanic inguinal patches and black pigments of the gular fold and central gular region are absent. Fe- males develop orange gravid coloration during the reproductive period. The one subadult female that I have examined in life had a bright yellow tail and hindlimbs. Distribution (Fig. 43). — Known to occur in the Sierras de San Lorenzo, Texas, and Solis of extreme southwestern Coahuila, Mexico. Fossil Record. —None. Natural History.— The following natural history observations were made on 23 and 25 June 1994. As are all Crotaphytus except C. reticulatus, C. an- Fig. 43. — Geographic distribution of Crotaphytus antiquus. The asterisk indicates the location of the Sierras de San Lorenzo, Texas, and Solis in southwestern Coahuila, Mexico. tiquus is strongly saxicolous and usually is observed basking on large limestone rocks and outcrops. When alarmed, they generally take refuge beneath a nearby rock or under the rock upon which they were perched. The habitat at the type locality is fairly typical Chi- huahuan Desert scrub with the dominant plant spe- cies being Larrea divaricata, Jatropha dioica, Fou- quieria splendens, Agave lechuguilla, Lippia grav- eolens, Opuntia cholla, two unidentified species of Opuntia (one resembling prickly pear, the other sim- ilar in habitus to pencil cholla), and (possibly) Echi- nocactus sp. Additional reptile and amphibian spe- cies observed at the type locality include Cnemi- dophorus inornatus, C. septemvittatus, Coleonyx brevis, Cophosaurus texanus, Phrynosoma modes- tum, Uta stansburiana, Scaphiopus couchii, and an undescribed species of Sceloporus similar to *S. jar- rovii cyanostictus. A third species of Cnemidophorus (possibly C. marmoratus ) is also present. Crotaphytus antiquus are abundant and I ob- served more than 25 individuals in an area of about 1.5 km in length and roughly 200 m in width. This species usually runs quadrupedally, but was ob- served to use bipedal locomotion on occasion. They 72 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 are able to take off bipedally from a standing start, as are all other Crotaphytus species. This species appears to be territorial, which is the case for all other Crotaphytus that have been studied (Fitch, 1956; Moehn, 1976; Montanucci, 1971; Sanborn and Loomis, 1979; Yedlin and Ferguson, 1973; plus numerous additional references). On 25 June 1994, I witnessed apparent territorial behavior when an adult male chased another adult male over approx- imately 10 m after the first male ventured into the area occupied by the second male. During the in- teraction, the pursuing male appeared to have its gular pouch fully depressed, a behavior that appears to be associated with aggression in all Crotaphytus species (Fitch, 1956; Sanborn and Loomis, 1979; personal observation). Very little is known about the reproductive be- havior of this species. However, since all but one of the females observed displayed orange gravid col- oration in various stages of intensity, it is clear that the reproductive cycle includes late June. One of the females bearing gravid coloration appeared emaci- ated, as if she had just oviposited. No juveniles were observed, suggesting that the year’s early clutches had not yet hatched. Some individuals (TNHC 53154, 53159) contained yolked ovarian follicles together with corpora lutea and distended, vascu- larized oviducts, suggesting that this species can pro- duce at least two clutches in a single reproductive season. One large female (SVL = 89 mm) contained four shelled eggs, another (SVL = 89 mm) contained three shelled eggs, and four additional females con- tained between one and four yolked ovarian folli- cles, suggesting that the species has a relatively small clutch size. The only observation made regarding feeding habits is that one adult male that was prepared as a skeleton contained the remains of an unidentified coleopteran insect. Illustrations. — Black-and-white photographs of males and females were provided by Axtell and Webb (1995). Crotaphytus bicinctores Smith and Tanner (Fig. 32A) Crotaphytus collaris bicinctores Smith and Tanner, 1972:27; fig. 1 , 2. Type locality: “Mercury Pass, Nevada Test Site, Nye Co., Nevada” (holotype: BYU 23883). Crotaphytus insularis bicinctores— Axtell, 1972:721; fig. 2, 5b-c, 6. Crotaphytus bicinctores— Sanborn and Loomis, 1979:105. Etymology. — From the Latin bi, two, and cinct, banded or girdled, in reference to “the divided banding (presumably of the collar) in the Great Basin populations” (fide Tanner, personal communication, 1993). Diagnosis. — Crotaphytus bicinctores can be dis- tinguished from C. reticulatus. C. collaris, C. ne- brius, and C. dicker sonae by the absence of black oral melanin. It can be further distinguished from C. reticulatus, C. collaris, and C. nebrius by the pres- ence in adult males of a strongly laterally com- pressed tail with a pale white dorsal caudal stripe, enlarged dark brown or black inguinal patches that extend between one-third and two-thirds of the dis- tance between the hindlimb and forelimb insertions, and a pale tan or off-white patternless region on the dorsal surface of the head. It may be further distin- guished from C. reticulatus as well as C. antiquus by a dorsal body pattern of white spots and dashes on a brown field rather than white reticulations on a gold, tan, or brown field. It may be further distin- guished from C. nebrius by its brown dorsal col- oration rather than pale tan. It may be further dis- tinguished from C. collaris by the presence of dark brown or black pigmentation in the gular fold (= ventrally complete anterior collar). It may be dis- tinguished from C. grismeri, C. vestigium, and C. insularis by the presence of broad tan or buff trans- verse dorsal body bands. It may be further distin- guished from C. grismeri by the absence of a green- ish tint to the white bar that separates the collars, by a pattern of white reticulations on a brown field on the forelimbs and hindlimbs rather than a pattern of yellow forelimbs with minute brown spotting on the proximal dorsal surface of the brachium and a hindlimb coloration that is nearly patternless yellow with scattered minute brown spots from the distal thigh to the distal terminus of the limb, by the ab- sence of a pale orange tail coloration in subadult females, and by the absence of a well-defined pale tan dorsal caudal stripe in juveniles of both sexes. It may be further distinguished from C. insularis and C. vestigium by the presence of a dorsally com- plete or narrowly separated posterior collar rather than a posterior collar that is broadly separated dor- sally or completely absent. It can be further distin- guished from C. insularis by the presence of a rel- atively broad nasal process of the premaxilla, the absence of olive green ventrolateral coloration in adult males, the presence of a pattern of small white spots and dashes (occasionally transverse bands), rather than a pattern of thicker, elongate white dash- 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 73 es, and the absence of extravomerine bones. It can be further distinguished from C. vestigium by the absence of olive green or burnt orange ventrolateral coloration. Variation (n = 20). — Rostral approximately four times wider than high, usually rectangular in shape. Rostral bordered by three to five postrostrals. Re- maining snout scales irregularly arranged, an en- larged middorsal series may be present. Nasals sep- arated by five to six intemasals. Frontonasals oc- casionally enlarged. Canthals three; five to eight scales separate canthals of left and right sides. Su- praorbital semicircles present with ten to 14 scales per semicircle, median scales do not fuse to form azygous frontals. Supraoculars flat or convex, smooth, becoming progressively larger medially such that medial scales are two to four times larger than lateral ones. Circumorbitals present, not well dif- ferentiated from supraoculars. Superciliaries six to 15, extremely elongate medial scale occasionally present. Palpebrals ovoid, slightly convex, inter- spersed with numerous interstitial granules. Preo- culars, suboculars, and postoculars form an arc of five to 1 3 rectangular scales, second, third, or fourth scale only rarely elongate. Supralabials 13 to 17, usually slightly longer than high except anteriormost scale, which is square or pentagonal. Lorilabials in one to three rows, ovoid to rectangular, juxtaposed, separating supralabials from suboculars and nasals. Aperture of external auditory meatus rectangular or ovoid, often constricted at or above the midpoint, approximately two to four times higher than wide, with small, strongly convex, somewhat conical au- ricular scales lining anterior margin. Mental pen- tagonal, one to 1.5 times wider than high, bordered laterally by anterior infralabials and posteriorly by a pair of large postmentals. Postmentals may or may not be separated from infralabials by one to three sublabials. Chinshields weakly differentiated or un- differentiated. Infralabials 12 to 18, square or wider than high, inferior border convex. Gulars granular, strongly convex and beadlike, each scale separated from adjacent scales by numerous asymmetrically arranged interstitial granules. Dorsal scales in approximately 144 to 200 rows midway between forelimb and hindlimb insertions. Tail long, cylindrical to oval in females and juve- niles over entire length, anterior one-half strongly compressed laterally in adult males. Paired, median row of subcaudals larger than adjacent subcaudals and lateral caudals. Enlarged postanal scales in males present. Deep postfemoral dermal mite pocket present at hindlimb insertion. Femoral pores 1 6 to 2 1 , femoral pores do not extend beyond angle of knee, separated medially by 16 to 26 granular scales. Subdigital la- mellae on fourth toe 17 to 23. Coloration in Life.— Dorsal body coloration in adult males is brown, with pale orange or peach- colored body bands. The white component of the dorsal pattern is composed of white spots and dashes on the body, and a reticulum on the tail, hindlimbs, and forelimbs. The reticulate pattern of the fore- limbs may occasionally be broken into spots. Trans- verse body bars are absent. Reticulations are always present on the superficial mandibular and temporal regions. A broad white or off-white caudal vertebral stripe is present. The dorsal surface of the head is pale-colored, and is conspicuously patternless. Ol- ive green or burnt orange ventrolateral coloration is lacking, although fine ventrolateral reticulations are present. The gular coloration in adult males is generally slate gray or gun-barrel blue, with a black central gular component. The peripheral gular pat- tern is the standard reticulate form. Anterior and posterior collar markings are always present and the posterior markings often contact middorsally. The anterior collars are complete ventrally in adult males as black pigments are present within the gular fold. A pair of black nuchal spots are not present mid- dorsally between the anterior collar markings. En- larged melanic axillary patches immediately pos- terior to the forelimb insertion are variably present. Large melanic inguinal patches are always present. The femoral pores are generally off-white to gray in color. Paired, melanic keels are variably present on the ventral surface of the caudal extremity. Females are less vividly marked than males. The dorsal coloration is grayish brown and they lack the white dorsal caudal stripe, black pigments of the gular fold, and melanic inguinal patches, axillary' patches, and gular spot. Gravid females develop vivid orange or reddish lateral bars during the gravid period. The tail is not vividly colored in adult or subadult females of this species. Size. — This species exhibits strong sexual dimor- phism with males reaching larger adult size (maxi- mum observed SVL =111 mm) than females (max- imum observed SVL = 98 mm). Distribution (Fig. 44). — Occurs in xeric rocky hab- itats in southeastern and extreme northeastern Cal- 74 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Fig. 44.— Geographic distribution of Crotaphytus bicinctores. The “?” near Flagstaff denotes a questionable record for the species at Williams, Arizona. The “?” in central Utah represents two records from Emery County that could not be precisely located: Nine miles W of Hanksville Highway at Nixon Uranium Mine and the Mamie Stover Incline. ifomia, western and northern Arizona, southeastern Oregon, western Idaho, western and central Utah, and much of Nevada. In Idaho, the species occurs primarily in association with the Snake River drain- age. Two additional localities in Idaho (approxi- mately 24 km NNE of Atomic City, Butte County, and Montpelier, Bear Lake County) are not indi- cated on the Crotaphytus bicinctores distribution map (Fig. 44) but may represent relict populations. There is a series of three specimens in the Museum of Vertebrate Zoology (MVZ 43415-17) listed as col- lected at Cheney, Spokane County, Washington. This disjunct locality should be considered questionable until verified by additional field work. In southwestern Arizona, the species occurs throughout the volcanic mountain ranges north of the Gila River, while C. nebrius occupies most of the mountain systems south of the Gila River. How- ever, C. bicinctores occurs south of the Gila River near the town of Sentinel, a locality that is not oc- cupied by C. nebrius. In at least two localities, C. bicinctores and C. nebrius are only narrowly sepa- rated by the Gila River. Crotaphytus bicinctores oc- curs in the Laguna Mountains which lie on the north side of the Gila River, while C. nebrius occurs in the Gila Mountains on the south side of the Gila River. Also, C. bicinctores occurs in the Gila Bend Mountains on the west shore of the Gila River, while C. nebrius occurs in the Buckeye Hills on the ad- jacent east shore. I observed a subadult C. bicinc- tores at Black Gap, Maricopa County, Arizona, a narrow pass on the western periphery of the Sauceda Mountains through which Arizona State Highway 85 passes. This observation was extremely surpris- ing given that this area is apparently well isolated from known C. bicinctores populations north of the Gila Bend River and on the Sentinel Plain. If C. bicinctores has an established population at this lo- cality, it is likely that C. nebrius and C. bicinctores contact somewhere in the Sauceda or Maricopa mountains. Several later attempts to find C. bicinc- tores or C. nebrius at this locality were unsuccessful. In northern Arizona, C. bicinctores occurs within and north of the Colorado River drainage (Grand Canyon) and follows the Little Colorado River drainage as well. Over much of this area, the species occurs in close geographic proximity to C. collaris. Two hybrid zones between these species have been documented based on morphological and electro- phoretic evidence (Axtell, 1972; Montanucci, 1983), although it seems likely that additional contact zones exist. The symbol “?” west of Flagstaff on Figure 44 represents a series of specimens (SDSNH 19474- 80) that includes both C. bicinctores and C. collaris. It seems likely that the locality data for the C. bi- cinctores in this series is incorrect. In Utah, Crotaphytus bicinctores occupies most of the desert mountain ranges west of the Wasatch Range and also appears to occupy the arid regions to the east of the Wasatch Range. The symbol “?” on the C. bicinctores map (Fig. 44) represents two localities in Emery County (9 mi W of Hanksville Highway (Hwy 24) near the Nixon Uranium Mine (BYU 16496) and the Mamie Stover Incline [BYU 20089-90]) that are represented by specimens, but for which I could not find the specific localities on topographical maps. Fossil Record. — Pleistocene fossils collected from Rampart Cave, Arizona (Van Devender et al., 1977), 1996 McGUIRE— SYSTEM ATICS OF CROTAPHYTID LIZARDS 75 Gypsum Cave, Clark County, Nevada (Brattstrom, 1954), and Smith Creek Cave, White Pine County, Nevada (Mead et al., 1982) were identified as Cro- taphytus collaris. All fall within the current distri- butional range of C. bicinctores and therefore, on distributional grounds, may be more appropriately referred to this taxon. Natural History. — Many anecdotal reports re- garding the natural history of Crotaphytus bicinc- tores have appeared, although no general treatment of the ecology of the species has been published. The species occurs in some of the most inhospitable regions of North America including the rugged, vol- canic basin and range mountains of the Sonoran, Mojave, and Great Basin deserts. It is generally re- stricted to rocky habitats with scant vegetation, such as alluvia, lava flows, mountain sides, canyons, and rocky plains, but occasionally may be found in pe- ripheral areas with only limited rocky cover. I have observed individuals more than a mile away from the nearest extensive rocky habitat in association with rolling gravely hills with only occasional rocks. Their ability to inhabit such areas may allow this species to disperse across the suboptimal habitats that separate isolated desert mountain ranges, as they are known to inhabit numerous isolated moun- tain systems. These are diurnal lizards often seen perched atop dark volcanic rocks at temperatures over 37°C. When disturbed, they may take refuge beneath a nearby stone or bound bipedally from one rock to the next before taking refuge under a stone or in a nearby rodent hole. Although primarily sax- icolous, this species occasionally may ascend small shrubs (Banta, 1967), possibly to avoid high sub- strate temperatures or in search of food. The diet of this species appears to consist pri- marily of arthropods, including orthopterans, co- leopterans, hemipterans, homopterans, hymenop- terans, lepidopterans, and arachnids, as well as small vertebrates (Camp, 1916; Knowlton and Thomas, 1936; Snyder, 1972; Nussbaum et al., 1983; per- sonal observation). Uta stansburiana is probably the most commonly consumed vertebrate species (Sny- der, 1972; personal observation), although other re- corded taxa include Sce/oporus, Cnemidophorus, Phrynosoma, and Xantusia vigi/is (Banta, 1960; Nussbaum et al., 1983). As do other crotaphytids, C. bicinctores occasionally includes plant matter in its diet (Banta, 1960). Snyder (1972) found that adult Crotaphytus bi- cinctores in northwestern Nevada may become ac- tive as early as April 17 and large numbers may be observed in early May. I have observed adults active as early as March 19 in southwestern Arizona. In southeastern California, I have observed juveniles (probably hatched the previous season), gravid fe- males, and adult males on May 2, indicating that mating activities probably commenced in April. Ne- onates have been observed in August in eastern Or- egon (Brooking, 1934). Axtell (1972) hatched eggs in the laboratory on September 19. Andre and MacMahon (1980) studied the reproductive biology of C. bicinctores in Tule Valley, Millard County, Utah. They discovered that females reached repro- ductive maturity at 85 mm SVL. All females sur- veyed in the first week of June contained oviducal eggs and by the end of June no females contained yolked follicles or oviducal eggs. Mean clutch size was reported as 5.38 with a range of three to seven. Larger females were found to produce larger clutches of eggs. Moehn (1976) showed that exposure to sunlight stimulates aggressive activity and despotism in cap- tives of this species. Sanborn and Loomis (1979) discussed male display patterns. Smith (1974) noted that C. bicinctores may elicit a high-pitched squeal when under duress. Snyder (1972) discussed home range size and territoriality in populations adjacent to Pyramid Lake, Storey County, Nevada. Illustrations.— A detailed black-and-white illus- tration of the entire animal was provided in Stebbins (1954); line drawings of the head squamation were included in Burt ( 1 928Z>:fig. 8) and Axtell (1972). Line drawings of the dorsal and ventral color pat- terns were given in Smith and Tanner (1 974); black- and-white photographs were provided in Axtell (1972), Pickwell (1972), Smith and Tanner (1972), and Nussbaum et al. (1983); color photographs were given by Behler and King (1979) and Sprackland (1990, 1993). Crotaphytus collaris Say Agama collaris Say, 1823:252. Type locality: not given; Arkansas Territory (now Oklahoma) near the Verdigris River implied (holotype: Academy of Natural Sciences of Philadelphia, now lost). Restricted type locality (Stejneger, 1890): “the Verdigris River, near its junction with the Neosho River, Creek Nation, Indian Territory”; (Stejneger and Barbour, 1917): “Verdigris River near its union with the Arkansas River, Oklahoma”; (Webb, 1970): “near Colonel Hugh Glenn’s Trading Post on the east bank of the Verdigris River, about two miles above its confluence with the Arkansas River”; (Axtell, 1989a): Ver- digris River near its union with the Arkansas River, Oklahoma. Crotaphytus collaris— Holbrook, 1842:79; pi. 10. Leiosaurus collaris— Dumeril, 1856:532. Crotaphytus bai/eyi Stejneger (syn. fide Cope, 1900), 1890:103; fig. 1, 2. Type locality: “Painted Desert, Little Colorado River, Arizona” (holotype: USNM 15821). 76 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Crotaphytes collaris baileyi -Stone and Rehn, 1903:30. [Crotaphytes collaris collaris]- Stone and Rehn, 1903:30. Crotaphytus collaris auriceps Fitch and Tanner (syn. fide Mon- tanucci, Axtell, and Dessauer 1975), 1951:553. Type locality: “3 1/2 mi. NNE Dewey, west side of the Colorado River, Grand County, Utah” (holotype: KU 29934). Crotaphytus ( Crotaphytus ) baileyi— Weiner and Smith, 1965:187. Crotaphytus ( Crotaphytus ) collaris- Weiner and Smith, 1965: 174; fig. 1-6. Crotaphytus collaris fuscus Ingram and Tanner, 1971:23; fig. 1. Type locality; “6.5 mi. N. and 1.5 mi. W. of Chihuahua City, Chihuahua, Mexico” (holotype: BYU 16970). Crotaphytus collaris melanomaculatus Axtell and Webb, 1995: 6; fig. 1, 2. Type locality: “25°14T0"N-103°47'W or 3.8 km S-l .7 km E Graseros on the highway to Presa Francisco Zarca, el 1 250 ± m, Durango, Mexico” (holotype: UTEP 15915). Etymology. — From the Latin collaris, in reference to the paired black collars on the lateral and dorsal surfaces of the neck. Diagnosis. — Crotaphytus collaris may be distin- guished from all other species of Crotaphytus by the absence of dark brown or black pigmentation in the gular fold (= ventrally complete anterior collar) of adult males. It may be further distinguished from C. reticulatus and C. antiquus by the absence of a reticulate dorsal pattern in adults of both sexes and from C. reticulatus by the absence of jet black fem- oral pores in males. It may be further distinguished from C. dickersonae, C. grismeri, C. bicinctores, C. vestigium, and C. insularis by the absence in adult males of enlarged dark brown or black inguinal patches, a laterally compressed tail, a white or pale tan dorsal caudal stripe, and a pale tan or off-white patternless region on the dorsal surface of the head. It may be further distinguished from C. grismeri, C. bicinctores, C. vestigium, and C. insularis by the presence of black oral melanin. Variation (n = 30). — Rostral approximately four times wider than high, usually rectangular in shape. Rostral bordered by four to six postrostrals. Re- maining snout scales irregularly arranged, an en- larged middorsal series may be present. Nasals sep- arated by four to six internasals. Frontonasals oc- casionally enlarged. Canthals three; five to seven scales separate canthals of left and right sides. Su- praorbital semicircles present with eight to 1 3 scales per semicircle, median scales may fuse to form azy- gous frontals, especially in eastern part of range. Supraoculars flat or convex, smooth, becoming pro- gressively larger medially such that medial scales are two to four times larger than lateral ones. Cir- cumorbitals present, not well differentiated from su- praoculars. Superciliaries six to 13, extremely elon- gate medial scale occasionally present. Palpebrals ovoid, slightly convex, interspersed with numerous interstitial granules. Preoculars, suboculars, and postoculars form an arc of four to ten rectangular scales, second, third, or fourth scale not elongate. Supralabials 11 to 17, usually slightly longer than high except anteriormost scale, which is square or pentagonal. Lorilabials in one to four rows, ovoid to rectangular, juxtaposed, separating supralabials from suboculars and nasals. Aperture of external auditory meatus rectangular or ovoid, often con- stricted at or above the midpoint, approximately two to four times higher than wide, with small, strongly convex, somewhat conical auricular scales lining anterior margin. Mental pentagonal, one to 1.5 times wider than high, bordered laterally by an- terior infralabials and posteriorly by a pair of large postmentals. Postmentals usually not separated from infralabials by sublabials; mental occasionally con- tacted by one or two sublabials. Chinshields weakly differentiated or undifferentiated. Infralabials 1 1 to 15, square or wider than high, inferior border con- vex. Gulars granular, strongly convex and beadlike, each scale separated from adjacent scales by nu- merous asymmetrically arranged interstitial gran- ules. Dorsal scales in approximately 136 to 186 rows midway between forelimb and hindlimb insertions. Tail long, cylindrical in both sexes and all age groups. Paired, median row of subcaudals larger than ad- jacent subcaudals and lateral caudals. Enlarged post- anal scales present in males. Deep postfemoral dermal mite pocket usually present at hindlimb insertion. Femoral pores 15 to 24, femoral pores do not extend beyond angle of knee, separated medially by 1 4 to 24 granular scales. Subdigital lamellae on fourth toe 15 to 22. Coloration in Life.— The color pattern of Crota- phytus collaris is extremely variable and it is prob- ably not possible to give a complete description of the various color phases that characterize different populations of this wide-ranging species, especially given that the often vibrant coloration displayed by these lizards is quickly lost in preservative. For this reason, the following description of coloration in C. collaris is limited in some respects to those color morphs that I have examined firsthand. Dorsal body coloration of adult males is extreme- ly variable with some populations characterized by a green dorsal coloration, others by a turquoise to pale green body with a yellow head and feet, others by a pale or dark brown coloration, and still others by a gray or combination of gray and olive green. In those populations characterized by a yellow head, the intensity of the yellow pigments may range from 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 77 pale to fluorescent. The white component of the dorsal pattern is retained well in preservative and is easily characterized as nearly all populations have white spots on the body with spots or reticulations present on the tail and hindlimbs. Some populations from Coahuila, Durango, Nuevo Leon, San Luis Potosi, and Zacatecas may have a dorsal pattern consisting at least in part of black spots that may or may not be surrounded by white, a pattern that is reminiscent of that of C. antiquus and C. reticulatus and potentially the result of introgression from the latter species (Montanucci, 1 974). The forelimbs are generally patternless or only obscurely patterned, but may occasionally bear pale reticulations or spots. Transverse body bars are absent. Reticulations gen- erally are confined to the superficial mandibular and temporal regions, as well as the hindlimbs and tail. A broad white or off-white caudal vertebral stripe is lacking. The dorsal surface of the head is not pale- colored, and generally is covered with spots that range in color from rust to chocolate brown. Olive green or orange ventrolateral coloration is lacking. Most of the variation in gular pattern coloration observed within Crotaphytus is restricted to C. col- laris. The gular coloration observed in living adult males examined over the course of this study range between olive green, dark blue, turquoise blue, slate gray, yellow, or orange. However, a black central component is not found in this species. As stated above, the peripheral gular pattern is always com- posed of a white reticulated pattern. Anterior and posterior collar markings are always present and the posterior markings occasionally may contact mid- dorsally. The anterior collars are not complete ven- trally as black pigments are absent from the gular fold. A pair of black spots may be present middor- sally between the anterior collar markings. A pair of enlarged melanic axillary patches are variably present immediately posterior to the forelimb in- sertion, although they are restricted to populations from the western portion of the species’ range (Ar- izona). Small melanic inguinal patches are also vari- ably present in adult males from this portion of the range. The femoral pores are generally off-white to gray in color. Paired, melanic keels may or may not be present on the ventral surface of the caudal ex- tremity. Female Crotaphytus collaris are much less con- spicuously marked than males, particularly in those populations characterized by green dorsal colora- tion. While females may retain a green component in their pattern, it is always of a much duller hue. As in other Crotaphytus, the gular pattern of females is less developed. Inguinal patches, which are vari- ably present in adult males, are lacking in females. Females develop vivid orange or reddish lateral bars during the gravid period. The tail is not vividly colored in either adult or subadult females. Size. — This species exhibits strong sexual dimor- phism with males reaching larger adult size (maxi- mum observed SVL =131 mm) than females (max- imum observed SVL = 106 mm). Distribution (Fig. 45). — Crotaphytus collaris has an extensive distribution in the western and south- central United States and northern Mexico extend- ing from northwestern Arizona, eastern Utah, and western Colorado eastward across the southern Great Plains into Missouri, northern Arkansas, and pos- sibly extreme northwestern Louisiana; and south- ward into extreme northern Sonora and northcen- tral mainland Mexico. Numerous isolated popula- tions occur on the eastern periphery of its range in Missouri and Arkansas. In Texas, the eastern dis- tributional extent of C. collaris is limited by the Balcones Escarpment as suitable rocky habitat does not extend east of this point. For this reason, a num- ber of localities that lie east of the escarpment are considered questionable (FMNH 1171 1 6— 18 — “Ce- dar Creek, Bastrop Co.”; USNM 12762 — “Tehu- acana, Limestone Co.,” 145 18 — “Gainesville, Cooke Co.”; UTA 892-“ 10 mi. S Dallas, Dallas Co.”; see Axtell [1989a] for a more complete assessment of potentially erroneous localities for Texas speci- mens). In Mexico, C. collaris extends as far east as the eastern slopes of the Sierra Madre Oriental, while C. reticulatus occupies the flatland Tamaulipan thomscrub habitats to the immediate east. These two species approach one another closely in the vi- cinity of Allende, Coahuila, Mexico. In western and northern Arizona, the distributions of C. collaris and C. bicinctores abut one another and at least two hybrid zones occur (see description of the distri- bution of C. bicinctores). The questionable (“?”) Colorado locality on the dot distribution map (Fig. 45) refers to a specimen (USNM 58603) from Ar- chuleta County, Colorado, for which no specific lo- cality data were given. The questionable (“?”) lo- cality from near the border between Tamaulipas and San Luis Potosi, Mexico, represents a locality given for C. reticulatus (AMNH 104448 — “rte. 101, 12 mi. SW jet. with side rd. to Tula, 13 mi. NE San Luis Potosi state line”). This locality is dubious for C. reticulatus, but would not be unexpected for C. collaris. An extremely detailed dot distribution map for C. collaris in Texas was provided by Axtell (1989a). 78 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Fig. 45. — Geographic distribution of Crotaphytus collaris and C. reticulatus. The “?” in southern Colorado denotes a specimen without precise locality data from Archuleta County. The “?“s along the eastern periphery of C. collaris’ range in Texas represent dubious localities that lie east of the Balcones Escarpment. The “?” locality from near the border between Tamaulipas and San Luis Potosi, Mexico, represents a locality given for C. reticulatus that is dubious for this species, but would not be unexpected for C. collaris. Dot distribution maps for the states of Colorado (Hammerson, 1986), Kansas (Collins, 1982), Mis- souri (Johnson, 1987), and Oklahoma (Webb, 1970) have also been published. Dundee and Rossman ( 1 989) questioned whether C. collaris occurs naturally in the state of Louisiana. Two specimens are known, one of which may have been accidentally introduced (Frierson, 1 927), while the other was collected by D. Leslie at Boone’s Land- ing on the Toledo Bend Reservoir southwest of Ne- greet, Sabine Parish (cited as a personal commu- nication in Dundee and Rossman, 1989). Fossil Record. —Numerous Pleistocene fossils from several western states have been referred to this taxon, including a number of fossils more rea- sonably referred to other species (see C. bicinctores and C. nebrius accounts). All of the fossils, with the above exceptions, fall within the current distribu- tional limits of C. collaris (Estes, 1983). Natural History. — More has been written about 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 79 the natural history of this species than any other Crotaphytus, with the major ecological study being Fitch (1956). Numerous unpublished master’s the- ses and Ph.D. dissertations have dealt with ecology of C. collaris including (but not necessarily limited to) Mosley (1963), Trauth (1974), Hipp (1977), Bon- trager (1980), McAllister (1980), Rostker (1983), Malaret (1985), Rand (1986), and Uzee (1990). More specific published works have dealt with feeding (Burt, 1928a; Blair and Blair, 1941; McAllister and Trauth, 1982), growth (Sexton et al., 1992), endo- parasites (McAllister, 1985), reproduction (Green- berg, 1945; Clark, 1946; Robison and Tanner, 1962; Cooper and Ferguson, 1972, 1973; Parker, 1973; Ferguson, 1976; Trauth, 1978, 1979; Montanucci, 1983; Ballinger and Hipp, 1985), territoriality and aggression (Greenberg, 1945; Yedlin and Ferguson, 1973; Fox and Baird, 1992), hibernation (Legler and Fitch, 1957), aquatic behavior (McAllister, 1983), and thermoregulatory behavior (Dawson and Tem- pleton, 1963; Cothran and Hutchison, 1979) to highlight just a small sample of the vast amount of literature pertaining to this species. Illustrations. —Numerous illustrations and pho- tographs have appeared in publications and this list, by necessity, is not intended to be complete. Pub- lished figures include black-and-white illustrations of the entire animal (Harlan, 1835; Holbrook, 1842; Baird, 1859), head squamation (Baird, 1859;Stejne- ger, 1890; Cope, 1900; Burt, 19287*; Stebbins, 1954, Ingram and Tanner, 1971), dorsal pattern (Ingram and Tanner, 1971; Smith and Tanner, 1974), limb and preanal squamation (Cope, 1900), and skull, pectoral girdle, and pelvic girdle (Weiner and Smith, 1965). Black-and-white photographs are found in Ditmars (1920) and Van Denburgh (1922); color plates in Ditmars (1920), Webb (1970), Stebbins (1985), Dundee and Rossman (1989), and Conant and Collins (1991); color photographs in Cochran and Goin (1970), Leviton (1971), Behler and King (1979), Collins (1982), Hammerson (1986), Garrett and Barker (1987), Johnson (1987), and Sprackland (1990, 1993). Color photos showing greater road- runners ( Geococcyx californianus ) capturing and consuming C. collaris were presented by Meinzer (1993). Taxonomic Remarks.— As discussed in the Ma- terials and Methods section, all of the subspecies of C. collaris except C. nebrius (C. c. auriceps, C. c. baileyi, C. c.fuscus, and C. c. melanomaculatus) are here synonymized with C. collaris because no evi- dence has ever been presented, nor has any been discovered here, that these taxa represent indepen- dent lineages. For example, Ingram and Tanner (1971) showed the intergrade zone between C. c. auriceps and C. c. baileyi to be larger than the range of C. c. auriceps itself. The only characters that have been presented that are thought to separate C. c. baileyi from C. c. collaris are the following C. c. collaris features: supraorbital semicircles fused me- dially to form one or more azygous frontal scales, gular pouch yellow-orange, a shorter broader head, and larger supraocular scales. Of these, the first two are usually considered to be the principle diagnostic features (Brown, 1903; Meek, 1905; Ruthven, 1907; Strecker, 1909; Burt, 1928 7?; plus numerous other references) and both intergrade extensively. The condition of the supraorbital semicircles varies con- siderably in Colorado, New Mexico, and Texas pop- ulations (Burt, 19287?; personal observation), which prompted Burt (19287?) to synonymize C. c. baileyi with C. c. collaris. The yellow-orange gular pattern of C. c. collaris occurs at least as far south as Fred- ericksburg, Gillespie County, in southern Texas. In- dividuals from northeastern Mexico near the south end of Don Martin Dam and the vicinity of Allende, Coahuila, and 3.2 km NW of Mina, Nuevo Leon, have a gular coloration of yellow-orange surrounded by olive green. Individuals to the south and west (for example, 30 km SSW of Cuatrocienegas) have the standard olive green gular coloration. Thus, it appears that gular coloration grades smoothly from yellow-orange to olive green in northeastern Mex- ico. Fitch and Tanner (1951) were the last to com- ment extensively on the taxonomic status of C. c. collaris and C. c. baileyi. They clearly recognized the two as pattern classes and on these grounds ac- corded them the rank of subspecies. With respect to C. c. fuscus, diagnostic characters were not pre- sented in the type description, which was described on the basis of a distinctive discriminant function (Ingram and Tanner, 1971). Furthermore, Axtell (1989a) suggested that C. c.fuscus, C. c. collaris, and C. c. baileyi show three-way intergradation in western Texas, again implying that all three are pat- tern classes. An additional problem with the current alpha tax- onomy of Crotaphyt us collaris is that the paucity or lack of adequate character support for the subspecies makes it necessary to rely on color pattern differ- ences as a means of identification. Thus, although it was not mentioned in the original description, many herpetologists tend to think of C. c. baileyi as a green collared lizard with a yellow head and C. c. fuscus as a brown or grayish lizard (e.g., Stebbins, 1985; Conant and Collin s, 1 9 9 1 ). U nfortunately , the 80 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 ranges of these subspecies, as they are currently con- strued, are not consistent with these color pattern concepts. Crotaphytus c. baileyi, whose recognized range has been fragmented over the years by the descriptions of C. c. auriceps, C. c.fuscus, and C. c. melanomaculatus, is thought to extend from west- ern Arizona, eastward through central New Mexico, and southward through the panhandle of Texas into northcentral Mexico. The currently recognized dis- tribution of C. c. baileyi makes little sense when one considers that individuals from the Big Bend region (C. c. baileyi) may appear phenotypically identical to those from the Organ Mountains of New Mexico (C. c. fuscus). Thus, the subspecies of C. collaris do not appear to be on separate phylogenetic trajec- tories and do not even seem to represent useful pat- tern classes. Crotaphytus dickersonae Schmidt (Fig. 3 IB, C) Crotaphytus dickersonae Schmidt, 1922:638; fig. 2. Type locality: Isla Tiburon, Gulf of California, Mexico (holotype: USNM 64451). Crotaphytus collaris dickersonae— Allen, 1933:7. Crotaphytus ( Crotaphytus ) collaris dickersonae— Weiner and Smith, 1965:187. Etymology. — Named in honor of Mary C. Dickerson, former curator of herpetology at the American Museum of Natural His- tory, who studied the insular herpetofauna of the Gulf of Cali- fornia, Mexico. Diagnosis. — Crotaphytus dickersonae can be dis- tinguished from Crotaphytus bicinctores, C. gris- meri, C. insularis, and C. vestigium by the presence of black oral melanin, a blue or turquoise dorsal coloration, and the absence of enlarged postanal scales in males. It may be distinguished from C. reticulatus, C. collaris, and C. nebrius by the pres- ence in adult males of a strongly laterally com- pressed tail with a white or pale stripe extending vertebrally and enlarged dark brown or black in- guinal patches extending between one-half and one- third of the distance between the hindlimb and fore- limb insertions. It may be further distinguished from C. reticulatus and C. antiquus by the presence of a dorsal pattern of white spots on a blue or turquoise field rather than white reticulations on a gold, tan, or brown field. It may be further distinguished from C. collaris by the presence of dark brown or black pigmentation in the gular fold (= ventrally complete anterior collar) and the absence of enlarged postanal scales in males. It may be further distinguished from C. nebrius by the presence of a blue or turquoise dorsal coloration rather than tan and the absence of enlarged postanal scales in males. Variation (n = 20). — Rostral approximately two times wider than high, usually rectangular in shape. Rostral bordered by two to four postrostrals. Re- maining snout scales irregularly arranged, an en- larged middorsal series may be present. Nasals sep- arated by four to six internasals. Frontonasals oc- casionally enlarged. Canthals three; four to seven scales separate canthals of left and right sides. Su- praorbital semicircles present with 11 to 15 scales per semicircle, median scales do not fuse to form azygous frontals. Supraoculars flat or convex, smooth, becoming progressively larger medially such that medial scales are two to four times larger than lateral ones. Circumorbitals present, not well dif- ferentiated from supraoculars. Superciliaries eight to 12, extremely elongate medial scale absent. Pal- pebrals ovoid, slightly convex, interspersed with nu- merous interstitial granules. Preoculars, suboculars, and postoculars form an arc of six to nine rectan- gular scales, second, third, or fourth scale not elon- gate. Supralabials 13 to 17, usually slightly longer than high except anteriormost scale, which is square or pentagonal. Lorilabials in two to four rows, ovoid to rectangular, juxtaposed, separating supralabials from suboculars and nasals. Aperture of external auditory meatus rectangular or ovoid, often con- stricted at or above the midpoint, approximately two to four times higher than wide, with small, strongly convex, somewhat conical auricular scales lining anterior margin. Mental pentagonal, one to 1.5 times wider than high, bordered laterally by an- terior infralabials and posteriorly by a pair of large postmentals. Postmentals usually separated from infralabials by a pair of sublabials; sublabials oc- casionally absent on one or both sides. Chinshields weakly differentiated or undifferentiated. Infrala- bials ten to 16, square or wider than high, inferior border convex. Gulars granular, strongly convex and beadlike, each scale separated from adjacent scales by numerous asymmetrically arranged interstitial granules. Dorsal scales in approximately 154 to 186 rows midway between forelimb and hindlimb insertions. Tail long, cylindrical to oval in females and juve- niles of both sexes over entire length, anterior one- half strongly laterally compressed in adult males. Paired, median row of subcaudals larger than ad- 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 81 jacent subcaudals and lateral caudals. Enlarged post- anal scales absent in males. Deep postfemoral dermal mite pocket present at hindlimb insertion. Femoral pores 1 6 to 2 1 , femoral pores do not extend beyond angle of knee, separated medially by 17 to 25 granular scales. Subdigital la- mellae on fourth toe 17 to 21. Coloration in Life. — Dorsal body coloration in adult males is vibrant aquamarine to cobalt blue over the entire dorsal surface of the body except the distal half of the tail. There is no trace of yellow as seen in Crotaphytus collaris. The white component of the dorsal pattern is composed of large white spots and dashes on the body, a reticulated tail and hin- dlimbs, and forelimbs that are generally spotted or mottled. Transverse body bars are absent. Reticu- lations are always present on the superficial man- dibular and temporal regions. A broad white or off- white caudal vertebral stripe is present. The dorsal surface of the head is pale-colored, and is conspic- uously patternless. Olive green or burnt orange ven- trolateral coloration is lacking. The gular coloration is generally slate gray with a black central gular com- ponent. The peripheral gular pattern is the standard reticulate form. Anterior and posterior collar mark- ings are always present and the posterior markings often contact middorsally. The anterior collars are complete ventrally by way of black pigments present within the gular fold. A pair of black nuchal spots are not present middorsally between the anterior collar markings. Enlarged melanic axillary patches immediately posterior to the forelimb insertion are lacking. Large melanic inguinal patches are always present. The femoral pores are generally off-white to gray in color. Paired, melanic keels are always or nearly always present on the ventral surface of the caudal extremity. The coloration of females is much more subdued than that of males. The dorsal coloration is gray or brownish gray, rather than vivid blue, and females lack the melanic inguinal patches, black pigments in the gular fold, black central gular blotch, and white dorsal caudal stripe. Gravid females develop vivid orange or reddish lateral bars. The tail of re- productive females is bright lemon yellow. Size. —This species exhibits strong sexual dimor- phism with males reaching larger adult size (maxi- mum observed SVL =116 mm) than females (max- imum observed SVL = 97 mm). Distribution (Fig. 46). — Isla Tiburon in the Gulf of California, Mexico, and the desert mountains of the adjacent Sonoran coastline (Sierra Bacha and Sierra Seri) between Punta Cirio (1 1 .6 km S Puerto Libertad) and Bahia Kino, Mexico. Fossil Record. — None. Natural History. — No natural history data con- cerning this species have been published to date. Crotaphytus dickersonae apparently does not devi- ate significantly from other saxicolous Crotaphytus species with respect to basic aspects of its ecology and behavior. The species is common on south and east facing slopes with sparse vegetation and scat- tered granitic rocks of various sizes, with lizards generally observed basking on smaller rocks on these slopes. In coastal Sonora, C. dickersonae were ob- served on hillsides characterized by the following plant species: Bursera microphylla, Encelia fannosa, Jatropha cuneata, Pachycereus pring/ei, Stenocereus thurberi, Lycium sp., and Harfordia macroptera. The lizards Uta stansburiana, Cnemidophorus tigris, and Callisaurus draconoides are common on these hill- sides and very likely comprise a large component of the diet of C. dickersonae, a species that appears to prey heavily on lizards (based on gut content observations). This species tends to occur in similar habitats on Isla Tiburon, although juveniles ob- served on the island were concentrated around rocky outcroppings at the summits of the low hills rather than on the scattered rocks along the lower slopes of the hills. However, this observation should not be taken to represent a general phenomenon as very little time (two days) was actually spent on the is- land. Adults of both sexes were observed on 22 March 1991 in coastal Sonora and adults and juveniles were active on Isla Tiburon on 24 March 1991. Adult females did not bear gravid coloration, in- dicating that mating had not yet commenced. How- ever, an adult female observed on 14 April 1992 had striking orange gravid coloration indicating that mating takes place early in the spring in this species. Bright blue Crotaphytus dickersonae males stand out boldly on the pale rocks while basking and one might expect this species to be nervous and difficult to approach. This is not the case, however. Indeed, a Red-tailed Hawk ( Buteo jamaicensis) was ob- served to pass directly over a basking adult male C. dickersonae at a height no greater than 1 0 m without eliciting any observable reaction from the lizard. Illustrations- A black-and-white illustration of the lateral and dorsal head squamation of the holotype specimen is given in Schmidt (1922). Color pho- 82 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Fig. 46.— Geographic distribution of Crotaphytus dickersonae. The map depicts a small section of Sonoran coastline along the eastern margin of the Gulf of California. 1996 McGUIRE— SYSTEM ATICS OF CROTAPHYTID LIZARDS 83 tographs were provided in Avila (1995) and Sprack- land (1993). Crotaphytus grismeri McGuire (Fig. 32B) Crotaphytus grismeri McGuire, 1994:439; fig. 1. Type locality: “Canon David, a low pass that separates the contiguous Sierra de Los Cucapas and Sierra El Mayor, approximately 2 km W Mex. Hwy. 5 on the dirt road to the sulfur mine (turnoff at km 49 S. Mexicali), Baja California, Mexico” (holotype: CES 067- 629). Etymology. — Named in honor of L. Lee Grismer, noted au- thority on the herpetology of Baja California. Diagnosis.— Crotaphytus grismeri differs from all other Crotaphytus in the presence of a dull orange colored tail and hind limbs in subadult females, green pigmentation within the pale gray or white bar that separates the anterior and posterior black col- lars, a well-defined pale tan dorsal caudal stripe in juveniles of both sexes, a hindlimb pattern wherein the region between the middle of the thigh and its distal extremity is yellow and unmarked except for scattered minute brown spots, and in its small adult size (maximum male SVL = 99 mm, n = 7; x = 93.3). The presence in subadult females ( n = 6, in- cluding photographs of living individuals) of three large, lateral black spots with bold white borders may represent another diagnostic feature. Crota- phytus grismeri is further distinguished from C. re- ticulatus, C. antiquus, and C. collaris by the pres- ence, in adult males, of large black or dark brown inguinal patches, a strongly laterally compressed tail, and a bold white dorsal caudal stripe. It differs from these species and from C. dickersonae in that it lacks (in both sexes) black pigmentation of the oral mu- cosa and in the dark brown dorsal ground color of adult males. It differs from the remaining Crota- phytus (C. bicinctores, C. insu/aris, and C. vestigium ) in that the dorsal surface of the forelimb is yellow and almost without pattern, except for a small patch of minute brown spots near the forelimb insertion. It differs further from C. insularis and C. vestigium in that the posterior collar is only narrowly incom- plete middorsally rather than broadly incomplete and in having a dorsal pattern of subequal white spots without transversely oriented white bars. Variation ( n = 10). — Rostral approximately four times wider than high, usually rectangular in shape. Rostral bordered by three to four postrostrals. Re- maining snout scales irregularly arranged, an en- larged middorsal series may be present. Nasals sep- arated by five to six intemasals. Frontonasals oc- casionally enlarged. Canthals three; five to seven scales separate canthals of left and right sides. Su- praorbital semicircles present with ten to 15 scales per semicircle, median scales sometimes fuse to form an azygous frontal. Supraoculars fiat or convex, smooth, becoming progressively larger medially such that medial scales are two to four times larger than lateral ones. Circumorbitals present, not well dif- ferentiated from supraoculars. Superciliaries nine to 13, elongate medial scale occasionally present. Pal- pebrals ovoid, slightly convex, interspersed with nu- merous interstitial granules. Preoculars, suboculars, and postoculars form an arc of seven to 1 3 rectan- gular scales, the second, third, or fourth scale only rarely elongate. Supralabials 14 to 17, usually slight- ly longer than high except anteriormost scale, which is square or pentagonal. Lorilabials in two or three rows, ovoid to rectangular, juxtaposed, separating supralabials from suboculars and nasals. Aperture of external auditory meatus rectangular or ovoid, often constricted at or above the midpoint, approx- imately four times higher than wide, with small, strongly convex, somewhat conical auricular scales lining anterior margin. Mental pentagonal, one to 1.5 times wider than high, bordered laterally by an- terior infralabials and posteriorly by a pair of large postmentals. Postmentals may or may not be sep- arated from infralabials by a pair of sublabials. Chinshields weakly differentiated or undifferentiat- ed. Infralabials 13 to 18, square or wider than high, inferior border convex. Gulars granular, strongly convex and beadlike, each scale separated from ad- jacent scales by numerous asymmetrically arranged interstitial granules. Dorsal scales in approximately 164 to 190 rows midway between forelimb and hindlimb insertions. Tail long, cylindrical to oval in females and juve- niles over entire length, anterior one-half strongly compressed laterally in adult males. Paired, median row of subcaudals larger than adjacent subcaudals and lateral caudals. Enlarged postanal scales present in males. Deep postfemoral dermal mite pocket present at hindlimb insertion. Femoral pores 19 to 23, femoral pores do not extend beyond angle of knee, separated medially by 20 to 25 granular scales. Subdigital la- mellae on fourth toe 18 to 20. Coloration in Life. — Dorsal body coloration in adult males is brown, without pale orange or peach colored body bands. The white component of the dorsal pattern is composed of white spots and oc- casional dashes on the body, as well as the proximal 84 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 portions of the tail and hindlimbs. Transverse body bars are absent. The forelimbs are tan with yellow blotching above and lack the white reticulations or spotting found on other Crotaphytus. The hindlimb is brown with white spots proximally, grading abruptly at about midthigh into yellow-tan with small light brown spots. The minute brown spots terminate proximal to the pes, which is uniform yellow-tan. The lateral surfaces of the proximal half of the tail are brown with white spots, the white component gradually expands distally such that the distal half of the tail becomes uniform pale gray. A broad white or off-white caudal vertebral stripe is present in adult males. The dorsal surface of the head is pale golden tan, and is conspicuously pat- ternless. Reticulations are always present on the su- perficial mandibular and temporal regions. Olive green or burnt orange ventrolateral coloration is lacking. The gular coloration in adult males is dark blue-gray with a black central gular component. The peripheral gular pattern is the standard reticulate form. Anterior and posterior collar markings are always present. The anterior collars are complete ventrally, with black pigments extending through the gular fold. A pair of black nuchal spots are not present middorsally between the anterior collar markings. Enlarged melanic axillary patches im- mediately posterior to the forelimb insertion are absent. Large melanic inguinal patches are always present in adult males. The femoral pores are gen- erally off-white to gray in color. Paired, melanic keels are present on the ventral surface of the caudal extremity. Females are less vividly marked than males. The limbs are not as distinctly yellow as in males, the head and gular markings are duller, the white dorsal caudal stripe is either absent or much less devel- oped, and the melanic inguinal patches, ventrally complete anterior collar marking, and central gular spot are absent. Gravid females develop vivid or- ange or reddish lateral bars. The tail of subadult females is burnt orange in coloration. Size. — This species exhibits strong sexual dimor- phism with males reaching larger adult size (maxi- mum observed SVL = 99 mm) than females (max- imum observed SVL = 83 mm). Distribution (Fig. 47, 48). — Crotaphytus grismeri is known only from the type locality and a sight record in Canada La Palma, approximately 6 km W of El Faro. It is presumed to be restricted to the Sierra de Los Cucapas and the contiguous Sierra El Mayor, an isolated granitic mountain range in ex- treme northeastern Baja California, Mexico. This 80 km-long, 10 km-wide mountain range is isolated from the Sierra de Juarez of the peninsular ranges (inhabited by C. vestigium) to the west by Laguna Salada, a 15 km-wide flood plain that occasionally is inundated by waters from the Gulf of California. The substrate within Laguna Salada is hardpan with scattered aeolian sand. The rocky substratum re- quired by the saxicolous C. grismeri is entirely ab- sent, thus isolating this species to this mountain range. Fossil Record. — None Natural History. — Crotaphytus grismeri is saxic- olous and all lizards observed at the type locality were basking on small- to medium-sized granitic rocks on rock-strewn hillsides. Lizards were ob- served at all levels on the hillsides, from the rocky rubble at the bases of the hills to the tops of the hillsides 100 to 200 m above (McGuire, 1994). The activity season for the species extends at least from early March to early November. An adult male (98 mm SVL) was observed on 6 March 1993 and a juvenile male was observed on 7 November 1992. The latest date on which an adult has been observed was 1 2 September 1 992. However, this was a gravid female and it is certain that the activity period ex- tends at least for a few more weeks. Several gravid females were observed on 2 May and 16 May 1992 and this, together with the presence of a gravid fe- male in early September, suggests that second clutches may be produced. Several neonates ranging in SVL between 50 and 63 mm were observed on 1 2 September along with the gravid female, which further supports the contention that second clutches may occur (McGuire, 1994). Illustrations.— Color photographs of adult male, a gravid female, and a subadult female, as well as a black-and-white photo of the ventral pattern of adult males appeared in McGuire (1994). Crotaphytus insularis Van Denburgh and Slevin (Fig. 32D) Crotaphytus insularis Van Denburgh and Slevin, 1921:96. Type locality: “East coast of Angel de la Guardia Island seven miles north of Pond Island, Gulf of California, Mexico” (holotype: CAS 49151). Crotaphytus ( Crotaphytus ) insularis— Weiner and Smith, 1965: 187. Crotaphytus collaris insularis— Soule and Sloan, 1966:140. Crotaphytus insularis insularis— Smith and Tanner, 1972:27. Etymology. — From the Latin insula, island, and aris, pertain- ing to. In reference to the insular distribution of this species. 1996 McGUIRE— SYSTEM ATICS OF CROTAPHYTID LIZARDS 85 Fig. 47.— Geographic distribution of Crotaphytus grismeri. The wavy pattern indicates the ephemeral playa Laguna Salada. The hand- drawn hatched lines represent the borders of mountain ranges. f a 86 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Fig. 48. — Geographic distribution of Crotaphytus vestigium, C. grismeri, and C. insularis. 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 87 Diagnosis.— Crotaphytus insularis can be distin- guished from all other Crotaphytus by the slender and elongate nasal process of the premaxilla and its dorsal pattern of elongate white dashes, some of which may form thick, wavy transverse lines. It can be distinguished from all Crotaphytus except female C. reticulatus and occasional C. vestigium by the extreme reduction of the posterior collar in both sexes such that it is nearly always absent, and when present, it is extremely reduced. It can be distin- guished from all but C. vestigium by the presence of extravomerine bones. It can be distinguished from all but some C. vestigium (those from north of Bahia de Los Angeles, Baja California) and some C. col- laris by the presence in adult males of olive green ventrolateral coloration. It can be distinguished from C. reticulatus, C. collaris, C. nebrius, and C. dick- ersonae by the absence of black oral melanin. It can be further distinguished from C. reticulatus, C. col- laris, and C. nebrius by the presence in adult males of a strongly laterally compressed tail, a white or off-white dorsal caudal stripe, a pale tan or white patternless region on the dorsal surface of the head, and enlarged dark brown or black inguinal patches (rather than the small inguinal patches of C. nebrius and some C. collaris). It can be further distinguished from C. collaris by the presence in adult males of dark brown or black pigmentation in the gular fold (= ventrally complete anterior collar). It can be fur- ther distinguished from C. grismeri by its forelimb and hindlimb patterns consisting of white reticu- lations on a brown field and the absence of a greenish tint in the white bar that separates the anterior and posterior collars. It can be further distinguished from C. reticulatus and from C. antiquus by the absence of the white dorsal reticulum characteristic of these species. Variation ( n = 14). — Rostral approximately four times wider than high, usually rectangular in shape. Rostral bordered by four to six postrostrals. Re- maining snout scales irregularly arranged, an en- larged middorsal series may be present. Nasals sep- arated by five to six intemasals. Frontonasals oc- casionally enlarged. Canthals three; six to eight scales separate canthals of left and right sides. Supraorbital semicircles present with ten to 14 scales per semi- circle, median scales do not fuse to form azygous frontals. Supraoculars flat or convex, smooth, be- coming progressively larger medially such that me- dial scales are two to four times larger than lateral ones. Circumorbitals present, not well differentiated from supraoculars. Superciliaries eight to 13, ex- tremely elongate medial scale occasionally present. Palpebrals ovoid, slightly convex, interspersed with numerous interstitial granules. Preoculars, subocu- lars, and postoculars form an arc of six to 1 1 rect- angular scales, second, third, or fourth scale not elongate. Supralabials 13 to 18, usually slightly lon- ger than high except anteriormost scale, which is square or pentagonal. Lorilabials in two to three rows, ovoid to rectangular, juxtaposed, separating supralabials from suboculars and nasals. Aperture of external auditory meatus rectangular or ovoid, often constricted at or above the midpoint, approx- imately two to four times higher than wide, with small, strongly convex, somewhat conical auricular scales lining anterior margin. Mental pentagonal, one to 1.5 times wider than high, bordered laterally by anterior infralabials and posteriorly by a pair of large postmentals. Postmentals usually separated from infralabials by a pair of sublabials, occasionally only one sublabial or no sublabials present. Chin- shields weakly differentiated or undifferentiated. In- fralabials 11 to 17, square or wider than high, in- ferior border convex. Gulars granular, strongly con- vex and beadlike, each scale separated from adjacent scales by numerous asymmetrically arranged inter- stitial granules. Dorsal scales in approximately 166 to 206 rows midway between forelimb and hindlimb insertions. Tail long, cylindrical to oval in females and juve- niles over entire length, anterior one-half strongly compressed laterally in adult males. Paired, median row of subcaudals larger than adjacent subcaudals and lateral caudals. Enlarged postanal scales in males present. Deep postfemoral dermal mite pocket present at hindlimb insertion. Femoral pores 19 to 23, femoral pores do not extend beyond angle of knee, separated medially by 19 to 24 granular scales. Subdigital la- mellae on fourth toe 19 to 24. Coloration in Life. — Dorsal body coloration in adult males is brown. The white component of the dorsal pattern is composed of elongate white spots and dashes on the body, with the tail, hindlimbs, and forelimbs reticulated. Transverse body bars are absent. Reticulations are always present on the su- perficial mandibular and temporal regions. A broad white or off-white caudal vertebral stripe is present. The dorsal surface of the head is pale-colored, and is conspicuously patternless. Olive green ventrolat- eral coloration is present in adult males. The gular coloration in adult males is generally slate gray with an olive green tinge. A black central gular compo- 88 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 nent is present. The peripheral gular pattern is the standard reticulate form. Anterior and posterior collar markings are only variably present with both sexes usually lacking pos- terior collar markings and females often lacking both the posterior and anterior collar components. When present, the posterior collar markings are reduced and do not approach one another middorsally. The anterior collars are complete ventrally in adult males, with black pigments extending through the gular fold. A pair of black nuchal spots are not present middorsally between the anterior collar markings. Enlarged melanic axillary patches immediately pos- terior to the forelimb insertion are variably present. Large melanic inguinal patches are always present in adult males. The femoral pores are generally off- white to gray in color. Paired, melanic keels are absent from the ventral surface of the caudal ex- tremity. Females are less vividly marked than males. The head and gular markings are less vibrantly marked and they lack male color pattern characteristics such as the white dorsal caudal stripe and melanic in- guinal patches, axillary patches, central gular patch, and ventrally complete anterior collar marking. Fe- males develop vivid orange or reddish lateral bars during the gravid period. The tail is not vividly colored in adult or subadult females of this species. Size. — This species exhibits strong sexual dimor- phism with males reaching larger adult size (maxi- mum observed S VL = 1 20 mm) than females (max- imum observed SVL - 104 mm). Distribution (Fig. 48). — Restricted to Isla Angel de La Guarda in the Gulf of California, Mexico. Fossil Record.— None. Natural History. — No published accounts are available regarding the natural history of Crotaphy- tus insularis. However, this species does not appear to differ markedly with respect to its behavior and ecology from its sister taxon, C. vestigium. Adults were observed basking on isolated volcanic rocks and a juvenile was basking on a talus slope com- prised of smaller white stones. Individuals are wide- ly spaced, which may be the result of extremely xeric conditions with very scant vegetation. Adults of both sexes and juveniles were active on 28 and 29 June 1991 and one female was observed with gravid col- oration. Illustrations. — A color photograph was provided by Sprackland (1993). Crotaphytus nebrius Axtell and Montanucci, new combination (Fig. 31 A) Crotaphytus collaris nebrius Axtell and Montanucci, 1977:1; fig. 1. Type locality: “28°30'30''N-1 1 1°02'30"W” (14 Km by road N. Rancho Cieneguita), Sonora, Mexico” (holotype: LACM 126617). Etymology. — From the Greek nebrias, meaning spotted, like a fawn. Named in reference to the fawn-like dorsal pattern of large white spots on a pale tan field. Diagnosis. — Crotaphytus nebrius can be distin- guished from C. dickersonae, C. grismeri, C. bi- cinctores, C. vestigium, and C. insularis by the ab- sence in adult males of a laterally compressed tail, enlarged dark brown or black inguinal patches that extend between one-third and one-half the distance between the hindlimb and forelimb insertions, and a pale white dorsal caudal stripe. It can be further distinguished from C. grismeri, C. bicinctores, C. vestigium, and C. insularis by the presence of black oral melanin. It can be distinguished from C. reti- culatus and C. antiquus by its dorsal color pattern of white spots on a pale tan field, rather than white reticulations on a pale tan or brown field and the absence of jet black femoral pores in males. It can be further distinguished from C. reticulatus by the presence in adult males of small dark brown or black inguinal patches. It can be distinguished from C. collaris by the presence in adult males of dark brown or black pigmentation in the gular fold (= ventrally complete anterior collar) and by the presence of burnt orange ventrolateral abdominal coloration in breed- ing males. Variation (n = 20). — Rostral approximately four times wider than high, usually rectangular in shape. Rostral bordered by three to six postrostrals. Re- maining snout scales irregularly arranged, an en- larged middorsal series may be present. Nasals sep- arated by four to six internasals. Frontonasals oc- casionally enlarged. Canthals three; five to eight scales separate canthals of left and right sides. Su- praorbital semicircles present with ten to 1 5 scales per semicircle, median scales do not fuse to form azygous frontals. Supraoculars flat or convex, smooth, becoming progressively larger medially such that medial scales are two to four times larger than lateral ones. Circumorbitals present, not well dif- ferentiated from supraoculars. Superciliaries eight to 1 3, extremely elongate medial scale occasionally present. Palpebrals ovoid, slightly convex, inter- 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 89 spersed with numerous interstitial granules. Preo- culars, suboculars, and postoculars form an arc of five to ten rectangular scales, second, third, or fourth scale only rarely elongate. Supralabials 11 to 17, usually slightly longer than high except anteriormost scale, which is square or pentagonal. Lorilabials in one to three rows, ovoid to rectangular, juxtaposed, separating supralabials from suboculars and nasals. Aperture of external auditory meatus rectangular or ovoid, often constricted at or above the midpoint, approximately two to four times higher than wide, with small, strongly convex, somewhat conical au- ricular scales lining anterior margin. Mental pen- tagonal, one to 1.5 times wider than high, bordered laterally by anterior infralabials and posteriorly by a pair of large postmentals. Postmentals usually not separated from mfralabials by sublabials; mental oc- casionally contacted by one or two sublabials. Chin- shields weakly differentiated or undifferentiated. In- fralabials 13 to 17, square or wider than high, in- ferior border convex. Gulars granular, strongly con- vex and beadlike, each scale separated from adjacent scales by numerous asymmetrically arranged inter- stitial granules. Dorsal scales in approximately 142 to 188 rows midway between forelimb and hindlimb insertions. Tail long, cylindrical or slightly laterally compressed (oval) in both sexes and all age groups. Paired, me- dian row of subcaudals larger than adjacent sub- caudals and lateral caudals. Enlarged postanal scales present in males. Deep postfemoral dermal mite pocket present at hindlimb insertion. Femoral pores 1 7 to 22, femoral pores do not extend beyond angle of knee, separated medially by 1 7 to 24 granular scales. Subdigital la- mellae on fourth toe 17 to 20. Coloration in Life. — Dorsal body coloration is generally straw yellow, although this is subject to some intraspecific variation with some individuals dull tan in color. Contrary to Stebbins (1985), the anterior portion of the head may bear yellow pig- ments similar to those present in some populations of C. collaris. The white component of the dorsal pattern is composed of white spots on the body that are often roughly three times larger middorsally than they are laterally. Spots or a broken reticulum may be present on the tail and hindlimbs, while the fore- limbs are generally spotted or mottled. Transverse body bars are absent. Reticulations may be absent entirely, confined to the superficial mandibular and temporal regions, or present on these regions as well as the hindlimbs below the knee. A broad white or off-white caudal vertebral stripe is lacking. The dor- sal surface of the head is usually pale-colored, and is conspicuously patternless. Burnt orange ventro- lateral coloration may be present in males, partic- ularly those from the western portion of the species’ distribution, and may be a form of breeding col- oration. The gular coloration in males is generally slate gray or dark brown, but may be overlain with a yellow tint. A black central gular component is not present. The peripheral gular pattern is highly variable in this species, with the Tucson Mountains population characterized by the standard reticulated pattern, western populations characterized by obliquely oriented, radiating white stripes, and the remaining eastern and southern populations char- acterized by white spots on a sky blue background. Anterior and posterior collar markings are always present and the posterior markings may contact middorsally. The anterior collars are complete ven- trally in adult males, with black pigments extending through the gular fold. A pair of black nuchal spots may be present middorsally between the anterior collar markings. A pair of enlarged melanic axillary patches are variably present immediately posterior to the forelimb insertion. Small melanic inguinal patches are always present in adult males. The fem- oral pores are generally off-white to gray in color. Paired, melanic keels are always present on the ven- tral surface of the caudal extremity, except in the Tucson Mountains populations where they are lack- ing in two of the three specimens examined. Females are less vividly marked than males. The dorsal coloration is often browner than that of males. The head and gular markings are less vibrantly marked and they lack male color pattern character- istics such as the melanic inguinal patches, axillary patches, and ventrally complete anterior collar marking. Females develop vivid orange or reddish lateral bars during the gravid period. The tail is not vividly colored in adult or subadult females of this species. Size. — This species exhibits strong sexual dimor- phism with males reaching larger adult size (maxi- mum observed SVL =112 mm) than females (max- imum observed SVL = 98 mm). Distribution (Fig. 49). — Crotaphytus nebrius oc- curs in lowland desert and arid-tropical thornscrub mountain ranges of the Sonoran Desert where it appears to be allopatrically distributed with respect to all other Crotaphytus. In southwestern Arizona, 90 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Fig. 49. — Geographic distribution of Crotaphytus nebrius. The cross-hatched area represents the distribution of C. dickersonae. C. nebrius occurs throughout the north-south trend- ing mountain ranges, with specimens known from the Gila, Mohawk, Little Ajo, Ajo, Pozo Redondo, Puerto Blanco, Sikort Chuapo, and Estrella moun- tains, as well as the Buckeye Hills. They are also known from a few mountain ranges further to the east including the Quijotoa, Silverbell, and Tucson mountains (C. collaris occurs on the opposite side 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 91 of the Tucson Valley in the Santa Catalina Moun- tains). It is very likely that they occur in the re- maining mountain ranges south of the Gila River, although the Baboquivari Mountains may be in- habited by C. collaris (Axtell and Montanucci, 1977). Crotaphytus bicinctores generally skirts the northern border of C. nebrius’ range on the north side of the Gila River, but crosses the river at the Sentinel Plain, a region uninhabited by C. nebrius, as no Sonoran mountain ranges project northward into this area. In Sonora, Crotaphytus nebrius occurs in the transversely oriented foothills that follow the Unit- ed States-Mexico border along Mexican Highway 2 (the Pinacate Region). The north-south trending ranges of southwestern Arizona project northward from these foothills and probably provide the cor- ridor through which C. nebrius entered these moun- tains. They have been collected from several moun- tain ranges to the south and east in northern Sonora including the Sierra Cubabi, Sierra La Gloria, Sierra El Alamo, and Sierra El Rajon. One specimen is known from either the Sierra Cibuta or Sierra El Pinto (AMNH 73758, 25.6 km S Nogales), a more eastern locality in the northern foothills of the Sierra Madre Occidental. There is a relatively large gap in the known distribution of the species between the Caborca region (Sierra El Rajon) and the Hermosillo region. However, a series of specimens are known from the foothills between Hermosillo and the Guaymas region. Finally, the remaining specimens have been taken from the foothills of the Sierra Madre Occidental, in a series of north-south trend- ing valleys separated by presumably uninhabitable densely vegetated mountain ranges. It is likely that C. nebrius reached these localities by way of major river drainages entering from the south, such as the Rio Sonora and Rio Yaqui, as suitable open habitat appears to be restricted to these drainage systems. Populations of Crotaphytus nebrius are only nar- rowly separated from those of C. bicinctores at two localities and in both cases the barrier that prevents contact is the Gila River. Crotaphytus nebrius occurs on the northern edge of the Gila Mountains and is separated from a population of C. bicinctores in the Laguna Mountains approximately 0.4 km to the north on the opposite side of the Gila River. Sim- ilarly, C. nebrius occurs on the western margin of the Buckeye Hills, while C. bicinctores occurs on the extreme eastern margin of the Gila Bend Mountains only a few hundred meters to the west on the op- posite shore of the Gila River. Thus, C. nebrius may be observed on the east side of the Gillespie Bridge and C. bicinctores can be observed moments later on the west side. Several questions remain regarding the distribu- tion of C. nebrius. First, C. nebrius occurs as far north as 1 1.7 km N Huasabas and 19.5 km N Ba- cadehuachi in the Sierra Madre Occidental, while C. collaris is known from as far south as the Bavispe Region, approximately 60 km to the north. It is unknown whether this gap is real or an artifact of collecting. The habitat in the Huasabas and Baca- dehuachi regions appears to be marginal and the presence of higher elevation mountains between this area and the Bavispe region strongly suggests that a contact zone does not exist here. However, this re- mains to be substantiated with additional field stud- ies. Second, a specimen of C. bicinctores was ob- served by the author at Black Gap, Maricopa Coun- ty, Arizona, a narrow pass on the western periphery of the Sauceda Mountains through which Arizona State Highway 85 passes. This observation was ex- tremely surprising given that this area is apparently well isolated from known C. bicinctores populations north of the Gila Bend River and on the Sentinel Plain. If C. bicinctores has an established population at this locality, it is likely that C. nebrius and C. bicinctores contact somewhere in the Sauceda or Maricopa mountains. Several later attempts to find C. bicinctores or C. nebrius at this locality were un- successful. Fossil Record.— V an Devender and Mead (1978) referred a maxilla and dentary from late Pleistocene deposits in the Tucson Mountains and Wolcott Peak, Pima County, Arizona, to Crotaphytus collaris. Van Devender et al. (1991) referred dentary, maxillae, and tooth crown material from late Pleistocene de- posits in Organ Pipe Cactus National Monument to either C. collaris or C. insularis. Because the Tucson Mountains and Organ Pipe Cactus National Mon- ument are currently inhabited by C. nebrius, this material probably should be referred to C. nebrius on distributional grounds. Natural History. — Nothing has been published re- garding the natural history of this species but I have made the following observations. Crotaphytus ne- brius occurs in a diversity of habitats, although al- ways in association with rocks. In the northern por- tion of its range it may be found in extremely xeric habitats characterized by granitic outcroppings or volcanic flows. In the southern portion of its range, it is occurs in rocky areas often with relatively dense arid-tropical thornscrub vegetation. In these areas, C. nebrius may be concentrated in arroyo bottoms 92 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 and less vegetated stream and river valleys. In the northwestern portion of its range, the species is often found perched on granitic rocks that lay in sandy washes at the bases of rocky hillsides. No observations have been made with respect to the feeding habits of this species although it is likely that arthropods and small lizards make up the bulk of the diet as in other Crotaphytus species. The activity period for the species may extend between March and at least late September. Adult and subadult males were observed on 1 9 March in the Buckeye Hills, Maricopa County, Arizona, and juveniles were observed north of Guaymas, Sonora, Mexico, on 27 March. Between 15 and 19 April 1992, juveniles and subadults that apparently had just emerged from hibernation (they were still en- crusted with dirt) were observed in the western foot- hills of the Sierra Madre Occidental and at Quijotoa, Pima County, Arizona. On the same day that the Quijotoa subadults were observed, adult males were observed just north of Ajo (Pima County) and at Mohawk, Yuma County, Arizona. Adults are active at least as late as 1 1 August and recently hatched neonates have been observed as late as 19 Septem- ber. It seems likely that adults extend their activities at least into September and juveniles into October or November. Reproductive behavior appears to be typical of the genus. On 14 June 1991 mating was observed in the Gila Mountains, Yuma County, Arizona. The male was observed to grasp the female by a fold of skin of the neck during coitus. The female offered no resistance and thus appeared to be fully receptive. Interestingly, the female bore fully developed gravid coloration, which is consistent with observations made by Montanucci ( 1 965) that this coloration may not deter copulation in Gambelia silus, at least with females that do not display rejection behavior. It therefore seems likely that mating takes place pri- marily in May or June. Recently emergent neonates have been observed on 1 1 August in the Silverbell Mountains, Pima County, Arizona, and on 19 Sep- tember in the Gila Mountains. Neonates collected in the Silverbell Mountains were as small as 44 mm SVL and the individual collected in the Gila Moun- tains was 42 mm SVL and still retained a small portion of the umbilicus. Thus, neonates appear to hatch out between July and/or August and Septem- ber, at least in the northern portion of the range. Illustrations.— A black-and-white photograph ap- pears in Axtell and Montanucci ( 1 977). A color pho- tograph of a gravid female was provided in Sprack- land (1993). Crotaphytus oligocenicus f Holman Crotaphytus oligocenicus Holman, 1972:1613. Type locality: “From early Oligocene, Cypress Hills Formation, north branch of Calf Creek, in L. S. 4, Sec. 8, twp. 8, range 22, W. 3rd mer., elevation 3600 ft (1 100 m)” (holotype: Saskatchewan Museum of Natural History number 1444). Etymology’. —Named in reference to the time period during which these lizards lived. Distribution. — Known only from the type locality. Remarks. — Crotaphytus oligocenicus f is an ex- tinct species of Oligocene age known only from six dentaries collected at the type locality. Because of the fragmentary nature of the type material, it can- not be determined whether this species shares any of the crotaphytid synapomorphies presented here. Thus, I agree with Estes (1983) in questioning whether this species is in fact a crotaphytid. How- ever, given that no data were discovered in this analysis either supporting or rejecting the placement of this species within Crotaphytidae, no taxonomic rearrangements are herein suggested. A black-and- white illustration of the holotype material (a right dentary) is given in Holman (1972). Crotaphytus reticulatus Baird ’ (Fig. 30C) Crotaphytus reticulatus Baird, 1858:253. Type locality: Laredo and Ringgold Barracks, Starr County, Texas— (Smith and Tay- lor, 1950): “Laredo”; (Cochran, 1961) “Ringgold Barracks, Montague County, Texas”; (Montanucci, 1976): “Fort Ring- gold Military Reservation (= Ringgold Barracks), Starr County, Texas” (lectotype Montanucci, 1976: USNM 2692A). Crotaphytus ( Crotaphytus ) reticulatus— Weiner and Smith, 1965: 187. Etymology. — From the Latin reticulatus, made like a net. In reference to the net-like dorsal and gular pattern of white retic- ulations present in this species. Diagnosis.— Crotaphytus reticulatus can be dis- tinguished from all other species of Crotaphytus ex- cept C. antiquus by the presence of an adult color pattern consisting of white reticulations, some of which enclose black pigmentation, and the presence of jet black femoral pores in males. It can be dis- tinguished from C. antiquus by the dorsal coloration of golden tan rather than dark brown and by the presence of black pigments in only a subset of the dorsal body reticulations rather than in all or nearly all of them. It can be further distinguished from C. collaris by the presence of dark brown or black pig- mentation in the gular fold (= ventrally complete anterior collar) in adult males. It can be further dis- tinguished from C. antiquus, C. nebrius, C. dicker- sonae, C. grismeri, C. bicinctores, C. insularis, and 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 93 C. vestigium by the absence in adult males of small or large dark brown or black inguinal patches. It may be further distinguished from C. dickersonae, C. grismeri, C. bid net ores, C. insular is, and C. ves- tigium by the absence in adult males of a strongly laterally compressed tail, a white or off-white dorsal caudal stripe, and a pale tan or white patternless region on the dorsal surface of the head. It may be further distinguished from C. grismeri, C. bicinc- tores, C. insularis, and C. vestigium by the presence of black oral melanin. Variation (n — 17). — Rostral approximately four times wider than high, usually rectangular in shape. Rostral bordered by three to six postrostrals. Re- maining snout scales irregularly arranged, an en- larged middorsal series may be present. Nasals sep- arated by five to seven internasals. Frontonasals oc- casionally enlarged. Canthals three; five to eight scales separate canthals of left and right sides. Su- praorbital semicircles present with ten to 1 5 scales per semicircle, median scales do not fuse to form azygous frontals. Supraoculars flat or convex, smooth, becoming progressively larger medially such that medial scales are two to four times larger than lateral ones. Circumorbitals present, not well dif- ferentiated from supraoculars. Superciliaries seven to 13, extremely elongate medial scale occasionally present. Palpebrals ovoid, slightly convex, inter- spersed with numerous interstitial granules. Preo- culars, suboculars, and postoculars form an arc of seven to 1 2 rectangular scales, second, third, or fourth scale not elongate. Supralabials 11 to 15, usually slightly longer than high except anteriormost scale, which is square or pentagonal. Lorilabials in two to three rows, ovoid to rectangular, juxtaposed, sepa- rating supralabials from suboculars and nasals. Ap- erture of external auditory meatus rectangular or ovoid, often constricted at or above the midpoint, approximately two to four times higher than wide, with small, strongly convex, somewhat conical au- ricular scales lining anterior margin. Mental pen- tagonal, one to 1.5 times wider than high, bordered laterally by anterior infralabials and posteriorly by a pair of large postmentals. Postmentals may or may not be separated from infralabials by one or two sublabials. Chinshields weakly differentiated or un- differentiated. Infralabials ten to 1 5, square or wider than high, inferior border convex. Gulars granular, strongly convex and beadlike, each scale separated from adjacent scales by numerous asymmetrically arranged interstitial granules. Dorsal scales in approximately 156 to 192 rows midway between forelimb and hindlimb insertions. Tail long, cylindrical to oval, sometimes more strongly laterally compressed in adult males. Paired, median row of subcaudals larger than adjacent sub- caudals and lateral caudals. Enlarged postanal scales absent in males. Deep postfemoral dermal mite pocket absent. Femoral pores 1 5 to 18, femoral pores do not extend beyond angle of knee, separated medially by 14 to 20 granular scales. Subdigital lamellae on fourth toe 18 to 22. Coloration in Life. — Dorsal body coloration in adult males and females is golden tan. The white component of the dorsal pattern is composed of a white reticulum found over nearly the entire dorsal surface of the animal, including the body, the tail, all four limbs, and the superficial mandibular and temporal regions. Many of the white reticulations of the body (and occasionally the limbs) enclose black pigments and these black-filled hexagons are present in seven or eight transversely arranged rows. Transverse body bars are absent. A broad white or off-white caudal vertebral stripe is not present in adult males. The dorsal surface of the head is not pale colored, and may bear a mottled pattern. Olive green or burnt orange ventrolateral coloration is lacking. The gular coloration in adult males is gen- erally slate gray or olive green and may be heavily tinged with yellow when the male breeding colora- tion is present. A black central gular component is present in males. Anterior and posterior collar markings are usually present in males, while only the posterior collar markings (in the form of a trans- verse series of black-filled reticulations) are often present in females. In both sexes, the collar markings appear to be more rudimentary than those of other Crotaphytus and appear to represent modified rows of transversely arranged black-filled hexagons from which black pigments have escaped and run togeth- er. When present, the posterior markings do not contact middorsally. The anterior collar markings are complete ventrally in adult males, with black pigments extending through the gular fold. A pair of black nuchal spots are generally present middor- sally between the anterior collar markings. Enlarged melanic axillary patches immediately posterior to the forelimb insertion are lacking. Large melanic inguinal patches are never present in adult males. The femoral pore exudate of males is jet black. Paired, melanic keels are absent from the ventral surface of the caudal extremity. Females develop vivid orange or reddish lateral bars during the gravid period. The tail is not vividly colored in adult or subadult females of this species. 94 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Size. — This species exhibits sexual dimorphism with males reaching larger adult size (maximum ob- served SVL = 122 mm) than females (maximum observed SVL =118 mm). Montanucci (1971) in- dicated that Crotaphytus reticulatus reaches a SVL of 137 mm, a much larger size than was observed in any of the material examined for this study. Distribution (Fig. 45). — Crotaphytus reticulatus occurs in the Tamaulipan Biotic Province of the lower Rio Grande valley of southern Texas and ad- jacent Mexico (Montanucci, 1971, 1976). Montan- ucci (1971) provided a dot distribution map for the species as well as a verbal description of its distri- butional limits. Axtell (1989Z?) provided a detailed dot distribution map for the species within the con- fines of Texas. Montanucci (1971) stated that the western limit of the species occurred at Muzquiz, Coahuila, Mexico, which would suggest that the dis- tributions of C. reticulatus and C. collaris overlap over an extensive area. However, Axtell (198 1) found that the locality data associated with the Muzquiz specimen were erroneous. The questionable (“?”) locality shown in Figure 45 from near the border between Tamaulipas and San Luis Potosi, Mexico, represents a locality given for C. reticulatus (AMNH 104448 — “rte. 101, 12 mi. SW jet. with side rd. to Tula, 13 mi. NE San Luis Potosi state line”)- This locality is dubious for C. reticulatus, but would not be unexpected for C. collaris. Fossil Record. — None. Natural History. — Before Montanucci’s (1971) study, very little was known about the natural his- tory of this species and his publication stands as the major contribution to this topic. Crotaphytus reti- culatus differs in many respects from other Crota- phytus, particularly in that it is much less reliant on rocky habitats. Indeed, while this species will utilize rocky habitats within its range, it is often found on mesquite flats far from the nearest rocky habitat. Montanucci (1971) noted that it is not found on rocky outcroppings along the margins of bluffs (hab- itat that one would expect other species of Crota- phytus to inhabit), but that these outcroppings were occupied by Sceloporus cyanogenys. Montanucci (1971) refers to the preferred habitat of this species as thombrush desert characterized by the following plant taxa: mesquite ( Prosopis glandulosa ), several species of Acacia, Mimosa, paloverde ( Cercidium macrum), white brush ( Aloysia lycioides ), cenizo ( Leucophyllum frutescens), and prickly pear (Opun- tia lindheimeri). Like other Crotaphytus, this species prefers to bask above the surrounding substrate and this is accomplished in rockless areas by perching on fence posts (personal observation) or in the branches of mesquite trees (Montanucci, personal communication). The natural history of Crotaphytus reticulatus bears a number of similarities to that of Gambelia. The utilization of flatland habitats with or without the presence of rocks is one notable similarity. An- other is associated with their escape behavior. When alarmed, they often will run to the base of a nearby bush where they flatten themselves to the ground and remain motionless (Montanucci, 1971; personal observation), a behavior that often is observed in G. silus (Montanucci, 1965), G. wislizenii, and G. copei. As in the latter three species, C. reticulatus often will allow one to approach within one or two meters without attempting escape. The diet of Crotaphytus reticulatus is similar to that of other Crotaphytus with arthropods (primar- ily orthopterans and coleopterans) making up the bulk of the diet, but with lizards ( Cnemidophorus gularis, Eumeces), snakes ( Salvadora grahamiae ), and rodents ( Peromyscus ?) occasionally taken (Klein, 1951; Montanucci, 1971). As has been observed in a number of other Crotaphytus and Gambelia spe- cies, plant matter (in particular Lycium berries) may be consumed. Montanucci (1971) discussed several additional aspects of Crotaphytus reticulatus biology including territoriality, reproduction, diel activity, seasonal activity, hatching and growth, predators, parasites, and injury. Illustrations.— Line drawings of Crotaphytus re- ticulatus were given in Cope (1900) and Burt (1935). Black-and-white photographs were presented in Smith (1946) and Montanucci (1971, 1974). Color illustrations appear in Conant (1975) and Conant and Collins (1991). Color photographs are found in Behler and King (1979), Garrett and Barker (1987), and Sprackland (1993). Crotaphytus vestigium Smith and Tanner (Fig. 32C) Crotaphytus fasciatus Mocquard, 1899:303; pi. 13, fig. 1. Type locality: “Cerro de las Palmas,” Baja California, Mexico (type: none designated). Crotaphytus fasciolatus— Mocquard (substitute name for Cro- taphytus fasciatus Mocquard, 1899), 1903:209. Crotaphytus insularis vestigium Smith and Tanner, 1972:29; fig. 1, 2. Type locality: “Guadelupe Canyon, Juarez Mountains, Baja California” (holotype: BYU 23338). Crotaphytus vestigium— Collins, 1991:43. 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 95 Etymology’. — From the Latin vestigium, a footprint, a track, a trace. In reference to the reduced collars of this species (Tanner, personal communication, 1993). Diagnosis. — Crotaphytus vestigium can be distin- guished from all other Crotaphytus except C. insu- laris and C. reticulatus by the presence of widely separated posterior collars. It can be distinguished from all other species of Crotaphytus by the presence of slender, white transverse dorsal body bars. It can be further distinguished from C. reticulatus, C. col- laris, C. nebrius, and C. dickersonae by the absence of black oral melanin. It can be further distinguished from C. reticulatus, C. collaris, and C. nebrius by the presence in adult males of a strongly laterally compressed tail, a white or off-white dorsal caudal stripe, a pale tan or white patternless region on the dorsal surface of the head, and enlarged dark brown or black inguinal patches (rather than the small in- guinal patches of C. nebrius and some C. collaris ). It can be distinguished from C. antiquus and further distinguished from C. reticulatus in the absence of a dorsal pattern composed of a white reticulum with some or all of the reticulations enclosing black pig- mentation. It can be further distinguished from C. grismeri by the absence of a greenish tint to the white bar that separates the anterior and posterior collars, by the hindlimb pattern consisting of white reticu- lations or spots on a brown field (field occasionally yellowish distal to the knee), by the presence of olive green or burnt orange ventrolateral coloration, and by its much larger maximum adult SVL. It can be distinguished from C. insularis by its broader nasal process of the premaxilla and its more strongly de- veloped posterior collar. Variation (n — 28). — Rostral approximately four times wider than high, usually rectangular in shape. Rostral bordered by two to five postrostrals. Re- maining snout scales irregularly arranged, an en- larged middorsal series may be present. Nasals sep- arated by three to five intemasals. Frontonasals oc- casionally enlarged. Canthals three; five to seven scales separate canthals of left and right sides. Su- praorbital semicircles present, median scales rarely fuse to form an azygous frontal. Supraoculars flat or convex, smooth, becoming progressively larger me- dially such that medial scales are two to four times larger than lateral ones. Circumorbitals present, not well differentiated from supraoculars. Superciliaries nine to 12, extremely elongate medial scale occa- sionally present. Palpebrals ovoid, slightly convex, interspersed with numerous interstitial granules. Preoculars, suboculars, and postoculars form an arc of six to 1 1 rectangular scales, second, third, or fourth scale not elongate. Supralabials ten to 18, usually slightly longer than high except anteriormost scale, which is square or pentagonal. Lorilabials in two to three rows, ovoid to rectangular, juxtaposed, sepa- rating supralabials from suboculars and nasals. Ap- erture of external auditory meatus rectangular or ovoid, often constricted at or above the midpoint, approximately two to four times higher than wide, with small, strongly convex, somewhat conical au- ricular scales lining anterior margin. Mental pen- tagonal, one to 1.5 times wider than high, bordered laterally by anterior mfralabials and posteriorly by a pair of large postmentals. Postmentals may or may not be separated from mfralabials by one or two sublabials. Chinshields weakly differentiated or un- differentiated. Infralabials 1 1 to 17, square or wider than high, inferior border convex. Gulars granular, strongly convex and beadlike, each scale separated from adjacent scales by numerous asymmetrically arranged interstitial granules. Dorsal scales in approximately 156 to 212 rows midway between forelimb and hindlimb insertions. Tail long, cylindrical to oval in females and juve- niles over entire length, anterior one-half strongly compressed laterally in adult males. Paired, median row of subcaudals larger than adjacent subcaudals and lateral caudals. Enlarged postanal scales in males present. Deep postfemoral dermal mite pocket present at hindlimb insertion. Femoral pores 1 5 to 25, femoral pores do not extend beyond angle of knee, separated medially by 1 7 to 24 granular scales. Subdigital la- mellae on fourth toe 15 to 25. Coloration in Life. — Dorsal body coloration in adult males is brown. The white component of the dorsal pattern is composed of white spots and dashes on the body, a reticulated tail and hindlimbs, and forelimbs that are either reticulated, spotted, or nearly patternless. Slender, transverse body bars are present in both sexes. Reticulations are always pres- ent on the superficial mandibular and temporal regions. A broad white or off-white caudal vertebral stripe is present. The dorsal surface of the head is pale-colored, and is conspicuously patternless. Ei- ther olive green or golden orange ventrolateral col- oration is present in adult males, with the former color present in individuals north of Bahia de San Luis Gonzaga, Baja California, Mexico, and the lat- ter color present in individuals from Bahia de Los Angeles southward. The ventrolateral coloration of 96 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 individuals occurring between Bahia de San Luis Gonzaga and Bahia de Los Angeles is not known. The gular coloration in adult males is generally slate gray or gun-barrel blue, with a black central gular component. The peripheral gular pattern is the stan- dard reticulate form. Anterior collar markings are always present and posterior collar marks are only rarely lacking. The posterior markings are widely separated middorsally. The anterior collar markings are complete ventrally in adult males, with black pigments extending through the gular fold. A pair of black nuchal spots are not present middorsally between the anterior collar markings. Enlarged me- lanic axillary patches immediately posterior to the forelimb insertion are variably present. Large me- lanic inguinal patches are always present. The fem- oral pores are generally off-white to gray in color. Paired, melanic keels are always present on the ven- tral surface of the caudal extremity. Females are less vividly marked than males. The dorsal coloration is usually gray or greenish gray. The head and gular markings are less developed and male color pattern characteristics such as the white dorsal caudal stripe, ventrally complete anterior col- lar markings, and melanic inguinal patches, axillary patches, and central gular spot are lacking. Gravid females develop vivid orange or reddish lateral bars. The tail is not brightly colored in adult or subadult females of this species. Size. — This species exhibits strong sexual dimor- phism with males reaching larger adult size (maxi- mum observed SVL = 125 mm) than females (max- imum observed SVL = 98 mm). Distribution (Fig. 48 ). — Crotaphytus vestigium in- habits the peninsular ranges and adjacent rocky hab- itats from the northern slope of the San Jacinto Mountains in southern California to the southern margin of the volcanic Magdalena Plain in Baja Cal- ifornia Sur. In southern California and northern Baja California, C. vestigium is limited to the eastern face of the peninsular ranges. There is a gap in the pen- insular ranges between the southern edge of the Si- erra San Pedro Martir and the northern edge of the Sierra La Asamblea and C. vestigium occurs on ei- ther side of the peninsular ranges from this point southward. Furthermore, its range extends north- ward along the western side of the peninsular ranges from this gap to a point at least as far north as the vicinity of Rancho San Jose (Meling’s Ranch) and even approaches the Pacific Coast at Mesa San Car- los (Bostic, 1971). The known southern distribu- tional limit of C. vestigium is 27.7 km (by road) S of San Jose de Comondu (McGuire, 1991). It is likely that the actual distributional limit is bounded by the volcanic mesas that terminate near this lo- cality. Crotaphytus vestigium apparently does not inhabit the isolated Sierra Santa Clara and Sierra Vizcaino on the Vizcaino Peninsula (Grismer et al., 1994). Fossil Record. — None. Natural History. — Ye ry little has been written re- garding the natural history of Crotaphytus vestigium. Sanborn and Loomis (1979) discussed the display patterns for this species and noted that it inhabits rocky outcroppings on the more rugged portions of the alluvial fans and mountain slopes at their San Jacinto Mountains study site. Common plant spe- cies at this locality included Larrea tridentata, En- celia farinosa, and Ambrosia dumosa. Welsh (1988) collected two individuals, one of which was found on a rocky volcanic slope in central desert scrub and the other on a granitic outcrop in coastal sage scrub. Bostic (197 1) observed two individuals on Mesa San Carlos, a table-topped mountain overlooking the Pacific coast of Baja California approximately 350 km south of the United States-Mexico border. One of these individuals was foraging among large ba- saltic rocks along the edge of the mesa while the other was seen basking on a large basaltic outcrop- ping on top of the mesa proper. Crotaphytus vestigium is a denizen of desert hill- sides, alluvial fans, canyons, and lava flows, always in association with rocks. They occur in some of the most xeric habitats of North America such as the eastern bases of the Sierra de Juarez and Sierra San Pedro Martir where they may be observed basking during the heat of the day. The rocky habitats in which they occur generally are characterized by scant vegetation. Common plant taxa with which C. ve’.s- tigium is often associated include Fouquieria splen- dens, F. digueti, Opuntia, Larrea tridentata, Pachy- cormus discolor, Bursera, Ferocactus, Pachycereus pringlei, Prosopis, and numerous additional xero- philic species. When alarmed, this species can move with great speed over complex rocky terrain by bounding bipedally from one stone to the next, often taking refuge beneath a larger rock. The activity season for adult Crotaphytus vestig- ium probably commences in March. Adults have been observed as early as 1 April 1992 at the foot of the Sierra La Asamblea, Baja California, and adult males, gravid females, and subadults have been 1996 McGUIRE— SYSTEM ATICS OF CROTAPHYTID LIZARDS 97 found as early as 1 1 April 1992 on the lava flows just south of Puertocitos, Baja California. A sub- adult male observed in this area also had conspic- uous orange bars similar to those of gravid females. On 9 April 1993 adults of both sexes as well as juveniles were observed at San Ignacio, Baja Cali- fornia Sur. At this time, large males already bore intense breeding coloration, while a large adult fe- male appeared to have recently emerged from hi- bernation as dried mud was still adhering to the flanks and limbs. Little is known about the predators of Crotaphy- tus vestigium, although it is likely that coachwhip snakes ( Masticophis flagellum ), raptors. Loggerhead Shrikes ( Lanius ludovicianus), and Greater Road- runners ( Geococcyx californianus), all of which are common throughout the range of C. vestigium, probably prey on this species. An American Kestrel ( Falco sparverius ) was observed near Rosarito, Baja California, with a limp C. vestigium in its talons and, thus, represents at least one known predator on the species. Illustrations. — Black-and-white photographs were presented in Smith and T anner (1972), Axtell (1972), and Jones (1993). A black-and-white illustration was given in Mocquard (1899). Color photographs were provided by Sprackland (1990, 1993) and McGuire (1994). Taxonomic Remarks. — In 1899, Mocquard de- scribed Crotaphytus fasciatus from Cerro Las Pal- mas, Baja California. It is clear from his description, and from the accompanying figure, that this is a juvenile Crotaphytus vestigium, and, as the name fasciatus predates that of vestigium by 73 years, the former name has priority. However, at the time of Mocquard’s description, the name fasciatus was al- ready in use as Hallowell (1852) had applied this name to a specimen of G. wislizenii from the sand hills at the lower end of Jornada del Muerte, New Mexico. Apparently realizing his error, Mocquard (1903) provided a substitute name for the Baja Cal- ifornia species, giving it the name C.fasciolatus, but by the time Mocquard had corrected his mistake, C. fasciatus Hallowell had already been synony- mized with C. wislizenii by Cope ( 1 900). Thus, C. fasciatus Mocquard again became the senior syn- onym for the Baja California species of collared liz- ard. The name C. fasciatus has not since been ap- plied to the Baja California population of Crota- phytus (sensu stricto), largely because later workers thought that Mocquard had described another syn- onym of C. wislizenii. Thus, Van Denburgh (1922) erroneously synonymized C. fasciatus Mocquard and C. fasciolatus Mocquard with C. wislizenii. Only Schmidt (1922) and Burt (19286) recognized that Mocquard’s specimen was indeed a Crotaphytus (sensu stricto). Over the following 50 years, the name C. collaris continued to be applied to this population and by the time it was recognized that the Baja California population is a distinct form, the name fasciatus Mocquard had long since been forgotten. Because the name fasciatus has not been used for more than 50 years and because the name vestigium has become firmly entrenched in the herpetological literature, an appeal should be made to the Inter- national Code of Zoological Nomenclature to use its plenary power to suppress the name C. fasciatus in order to maintain taxonomic stability. Gambelia Baird Crotaphytus— Baird and Girard, 1852:69. Leiosaurus, part— Dumeril, 1856:533. Crotaphytus (Gambelia)— Baird, 1858:253. Type species (by monotypy): Crotaphytus wislizenii Baird and Girard, 1852a. Gambelia— Smith, 1946:158. Definition. — Gambelia is defined as a node-based name for the clade stemming from the most recent common ancestor of Gambelia wislizenii and all species that are more closely related to that species than to Crotaphytus. Etymology. —Named in honor of William Gambel, ornithol- ogist and pioneer naturalist of western North America in the mid- 1800s. Coloration in Life.— There is much variation in the color pattern of Gambelia, although much of this is geographic variation within the wide-ranging species G. wislizenii. Nevertheless, several compo- nents of the color pattern are found in all Gambelia, at least during some portion of ontogeny. For ex- ample, the color patterns of neonates are very sim- ilar in all three extant species. They are character- ized by a series of transversely arranged, blood-red dorsal spots that begin on the head and continue onto the base of the tail. Each row of enlarged spots is generally comprised of four spots. Enlarged blood- red spots may extend onto the hindlimbs as well. Each transverse series of enlarged spots is separated by a pale or cream-colored transverse bar. The spots and bars continue onto the tail where the spots pro- gressively coalesce distally, forming dark bars. The dark bars alternate with the pale bars giving the tail a banded appearance, a pattern that remains 98 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 throughout ontogeny. Early in ontogeny, the blood- red color of the dorsal spotting begins to fade to a brown hue that is maintained into adulthood. Another component of the juvenile pattern that is consistent among extant Gambelia are the obliquely oriented, radiating melanic bars present on the head. These have a visually disruptive effect and may play a role in camouflage (McCoy, 1967). These head markings are lost early in ontogeny. The gular pattern of Gambelia is relatively con- sistent, with longitudinally arranged black streaks present in both sexes throughout ontogeny. Gam- belia silus differs slightly from G. copei and G. wis- lizenii in that the streaks are usually fragmented leaving spots or rhombs. Most Gambelia are characterized by the presence of brown dorsal spots that correspond to the blood- red dorsal spots of juveniles. The spots vary in size, density, and position within various populations and species. The spots may be fragmented or may be surrounded by ornamentation in the form of minute white spots. Many Gambelia also retain the juvenile crossbanding into adulthood and these crossbands are in many cases offset paravertebrally. The dorsal base color of Gambelia is generally a pale shade of white, cream, or gray but may be dark brown. The ventral coloration is generally white, off-white, or a pale shade of gray or yellow. All Gambelia lack sexual dichromatism, except in the case of male breeding coloration (present only in G. silus ) and vivid orange or red female “gravid coloration,” which is present in all Gambelia. The “gravid coloration” may be deposited in patches on the sides of the head and on the thighs, in a single or double row of spots along the flanks, and along the ventral surface of the tail. Size. — Gambelia silus exhibits sexual dimor- phism with males larger than females, while females attain much larger sizes than males in G. copei and G. wislizenii. Distribution.— Gambelia is found in the western United States from central Idaho and eastern Ore- gon southward in the Great Basin through western Colorado and western Texas in the east, and through the San Joaquin Valley and eastern deserts of Cal- ifornia in the west; southward into Mexico to west- ern Coahuila, northern Zacatecas, eastern and cen- tral Chihuahua, central Sonora, and into the cape region of Baja California. Fossil Record. — Numerous Pleistocene fossils have been referred to Gambelia, all of which were considered to be G. wislizenii. At least one fossil was found within the current distributional limits of G. silus and may therefore represent this species (Brattstrom, 1953). Gambelia copei Yarrow (Fig. 30B) Crotaphytus copeii Yarrow, 1882:441. Type locality: “La Paz, Cal.” (holotype: USNM 12663). Crotaphytus copii— Garman, 1884:16. Crotaphytus copei— C ope, 1887:34. Crotaphytus wislizenii— Cope, 1900:255. Crotaphytus wislizeni copei— Leviton and Banta, 1964:153. Crotaphytus wislizeni neseotes Banta and Tanner (syn. fide Mon- tanucci, 1978), 1968:186; fig. 1-5. Type locality: “Cedros Is- land, west coast of Baja California Norte, Mexico” (holotype: CAS 79872). Etymology^. — Named in honor of Edward Drinker Cope, noted American herpetologist and paleontologist. Diagnosis. — Gambelia copei is diagnosed from G. corona f by the absence of a broad, transversely con- cave frontal bone, the presence of a frontoparietal suture posterior to the posterior extent of the orbits, and an elongate and slender nasal process of the premaxilla. It is diagnosable from G. silus in its absence of male breeding coloration, absence of sex- ual dimorphism wherein males are larger than fe- males (the reverse condition is present), absence of notched zygosphenes and zygantra, and in the pos- session of an elongate (rather than truncated) and slender nasal process of the premaxilla. Gambelia copei is not easily diagnosed from G. wislizenii, as the primary character that supports the recognition of separate species is their narrowly overlapping dis- tributions (see comments below). Additional differ- ences include the absence of spotting on the head in all but one of 38 G. copei examined (SDSNH 18118, Bahia de San Francisquito, Baja California) and its darker dorsal coloration. Adjacent popula- tions of G. wislizenii are easily diagnosed from G. copei as they are characterized by a pale dorsal col- oration with numerous small punctations that are asymmetrically arranged, extend well onto the head, and continue well down onto the flanks. Variation ( n = 21). — Rostral approximately four times wider than high, usually rectangular in shape. Rostral bordered by four to seven postrostrals. Re- maining snout scales irregularly arranged, an en- larged middorsal series may be present. Nasals sep- arated by six to seven internasals. Frontonasals oc- casionally enlarged. Canthals four; posterior one or two wider than high; seven to nine scales separate canthals of left and right sides. Supraorbital semi- 1996 McGUIRE— SYSTEM ATICS OF CROTAPHYTID LIZARDS 99 circles absent. Supraoculars small, flat or convex, smooth, becoming progressively larger medially such that medial scales are two to four times larger than lateral ones. Circumorbitals absent. Superciliaries eight to 12, extremely elongate medial scale present. Palpebrals ovoid, slightly convex, may be inter- spersed with numerous interstitial granules. Preo- culars, suboculars, and postoculars form an arc of four to seven rectangular scales, second, third, or fourth scale elongate. Supralabials 13 to 17, usually slightly longer than high except anteriormost scale, which is square. Lorilabials in two to four rows, ovoid to rectangular, juxtaposed, separating su- pralabials from suboculars and nasals. Aperture of external auditory meatus rectangular or ovoid, often constricted at or above the midpoint, approximately two to four times higher than wide, with small, strongly convex, somewhat conical auricular scales lining anterior margin. Mental pentagonal, one to 1.5 times wider than high, bordered laterally by an- terior infralabials and posteriorly by a pair of post- mentals that may be enlarged. Postmentals almost always separated from infralabials by sublabials on at least one side. Chinshields weakly differentiated or undifferentiated. Infralabials 12 to 17, square or wider than high, inferior border convex. Gulars usu- ally flat, but occasionally convex and beadlike; each scale may be separated from adjacent scales by nu- merous asymmetrically arranged interstitial gran- ules. Dorsal scales in approximately 160 to 200 rows midway between forelimb and hindlimb insertions. Tail long, cylindrical in both sexes and all age groups. Paired, median row of subcaudals not larger than adjacent subcaudals and lateral caudals. Enlarged postanal scales present in males. Deep postfemoral dermal mite pocket present at hindlimb insertion. Femoral pores 20 to 3 1 , femoral pores extend beyond angle of knee, separated me- dially by ten to 18 granular scales. Subdigital la- mellae on fourth toe 20 to 24. Coloration in Life. — Individuals from southern San Diego County, the Sierra de Juarez and Sierra San Pedro Martir, and cismontane northwestern Baja California generally are dark brown in coloration with a pair of large paravertebral spots that are sep- arated by broad, cream-colored transverse bars. There is much lateral flecking; however, lateral spots are lacking. Spots are nearly always absent from the head. In southern populations, such as those in the Vizcaino Desert, the base color of the dorsum is a paler golden tan, the dorsal spots are fragmented, and lateral spots may be present. In some southern individuals, the dorsal spotting may be nearly in- distinguishable, with the dorsum peppered with fine pale speckling. This pattern may be more cryptic on the fine aeolian sand characteristic of the Vizcaino Desert (Grismer et al., 1994). The speckled pattern of the southern individuals appears to be an onto- genetic fragmentation of the color pattern charac- teristic of northern individuals as subadults have been examined with dorsal patterns very similar to those from the northern portion of the peninsula. Gravid coloration in G. copei is similar to that of G. wislizenii, with orange or red spots often present on the head and/or neck, in two rows of spots on each flank, and on the ventral surface of the tail. Red or orange pigments may be present on the thighs as well. Males lack any form of breeding coloration. A description of the dorsal pattern of G. copei (pattern Cl) was provided in Montanucci (1978). Size. —This species exhibits strong sexual dimor- phism with females reaching larger adult size (max- imum observed SVL = 126 mm) than males (max- imum observed SVL = 120 mm). Distribution (Fig. 50 ). — Gambelia copei occurs in extreme southcentral California in the vicinities of Cameron Comers (Mahrdt, 1973), Campo, and Po- trero Grade southward through all but the San Fe- lipe Desert region of northeastern Baja California to the northern portion of the cape region, Baja California Sur. The species is also found on the Pa- cific islands of Isla de Cedros, Isla Magdelena, and Isla Santa Margarita off of the west coast of the peninsula. Gambelia copei occurs in the lower Col- orado Desert region between El Huerfanito and Ba- hia de San Luis Gonzaga and is also known from the gulf coast desert region in the vicinities of Bahia de Los Angeles and Punta San Francisquito. How- ever, G. copei apparently does not inhabit the Gulf Coast desert region between Santa Rosalia and the vicinity of Loreto and may be excluded from this region by the intervening Sierra San Pedro and Si- erra de La Giganta. This species occurs in high den- sities on the sandy plains of the Vizcaino Peninsula and its distribution appears to be limited to the western side of the peninsular ranges from this re- gion to a point at least as far south as the southern terminus of the Sierra de La Giganta. It is known from as far south as 1 km N Rancho Tres Hermanos (N ofTodos Santos) in the cape region of Baja Cal- ifornia Sur. The only published distribution map specific to Gambelia copei (Banta and Tanner, 1968) is flawed 100 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 in a number of respects. First, the lower Colorado Desert region between San Felipe and Puertocitos is inhabited by G. wislizenii rather than G. copei. Second, G. copei is not known from the eastern half of the peninsula between Bahia de San Francisquito and La Paz. Finally, G. copei is not known to range throughout the cape region. Gambelia copei is narrowly syntopic with G. wis- lizenii over a zone of approximately 1 .6 km in Paseo de San Matias, Baja California (denoted by a rect- angular mark on Figure 50). The two species are separated by a broad, transverse volcanic field that extends from the Sierra San Felipe to the gulf coast between Puertocitos and El Huerfanito, Baja Cali- fornia. This rugged volcanic flow, which is 31.5 km in width (by road), appears to act as an effective dispersal barrier for Gambelia along the gulf coast. On the provided dot distribution map (Fig. 50), the question marks represent localities in Baja Cal- ifornia Sur that are questionable because of impre- cise locality data (CAS 18823 — San Andreas [San Jorge]; MVZ 37260 — Medano Blanco, 37262 — sand dunes 12 mi SE Venancio). Fossil Record. — None. Natural History.— Very little has been written re- garding the natural history of Gambelia copei, al- though it seems likely that it is similar to G. wisli- zenii in most aspects of its biology. This species is particularly common on the sparsely vegetated ae- olian flats of the Vizcaino Peninsula, Baja California Sur, where it is often observed basking on roadside rocks, on the berms adjacent to graded dirt roads, or moving about in open spaces between clumps of vegetation. Gambelia copei is also found in more xeric creosote scrub habitats in the general vicinities of Paseo de San Matias, Bahia de San Luis Gonzaga, and Bahia de Los Angeles, Baja California, and in coastal sage scrub and oak woodland habitats on the western slopes of the Sierra San Pedro Martir and Sierra de Juarez (Welsh, 1988; personal observa- tion). Where G. copei extends its range into extreme southcentral California, it apparently occurs in rel- atively densely vegetated chaparral. Gambelia copei shares a number of behavioral similarities with G. mslizenii and G. silus. For ex- ample, all three share a habit of basking on small stones and roadsides berms. Gambelia copei also displays the familiar “freeze” behavior such that when they are threatened, they run to the base of a bush or thicket, flatten themselves to the ground, and remain motionless (Tevis, 1944; personal ob- servation). Like G. mslizenii, this species appears to be a lizard predation specialist as evidenced by the presence of Uta stansburiana and Callisaurus draconoides in the stomach contents of museum specimens. Banta and Tanner (1968) observed a Uta stansburiana and a grasshopper in the stomach of an adult female (CAS 8843) from Isla de Cedros. Like its sister taxon G. mslizenii, females attain larg- er sizes than males. Although rigorous ecological data are lacking, G. copei appears to be nonterritorial as in G. wislizenii (personal observation). Although little is known of the predators of Gam- belia copei, they are likely to include the coachwhip snake {Masticophis flagellum) and patch-nosed snake ( Salvadora hexalepis) as well as other saurophagous snake species, raptors, the Loggerhead Shrike ( Lan - ius ludovicianus), the Greater Roadrunner ( Geococ- cyx calif or nianus), as well as a number carnivorous mammals such as the coyote ( Canis latrans ). Only one predation event has been observed by the au- thor, in which a Loggerhead Shrike was observed carrying a nearly full-grown G. copei. The shrike could only fly short distances with the relatively large lizard and, when pursued, was forced to pin the dead lizard on the spine of a mesquite ( Prosopis ). Adult Gambelia copei have been observed as early as 1 April 1993 in the vicinity of Catavina, on 9 April 1993 on the Vizcaino Peninsula, and on 10 April 1992 in Paseo de San Matias, indicating that this species emerges from hibernation at a date sim- ilar to that of G. wislizenii from southern California (Miller and Stebbins, 1964; Tollestrup, 1979; Mitchell, 1984). Although adult females were ob- served in April, none were gravid, suggesting that reproductive activity had not yet commenced. The earliest that gravid females have been observed by the author is 3 May 1 993 at the Paseo de San Matias locality. Gravid females have also been seen on 27 June 1991 in the Sierra San Borja and 4 July 1991 in the Sierra Santa Clara. An emaciated female that appeared to have recently deposited eggs also was observed on 4 July in the Sierra Santa Clara, indi- cating that mating probably took place in mid to late June. Fitch (1970) examined 90 leopard lizards from Baja California (but did not list localities, so it is possible that some of the specimens were G. wislizenii) and found two of two females collected in March to be gravid, as well as six of nine collected in June, and three of six collected in July. Thus, the reproductive season is more extensive than my ob- servations would indicate. 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 101 114 110 Fig. 50.— Geographic distribution of Gambelia copei. The small rectangular block in northern Baja California denotes the narrow zone where G. copei and G. wislizenii occur together. The asterisks represent sight records by the author for G. copei near Bahia de San Luis Gonzaga. The question marks along the Pacific coast of Baja California Sur indicate localities that must be considered questionable because of imprecise locality data. 102 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Illustrations. — Black-and-white photographs of adult lizards were provided in Banta and Tanner (1968) and Montanucci (1978). Taxonomic Remarks. -Gambelia copei and G. wislizenii are easily distinguished on the basis of their coloration (see diagnosis above). However, geographic variation in the dorsal coloration of Gambelia wislizenii is extensive and this alone is not particularly compelling evidence for the recog- nition of copei as a distinct species. The primary motivation for this taxonomic rearrangement is the occurrence of both forms in syntopy along a narrow zone within Paseo de San Matias in northeastern Baja California. Within Paseo de San Matias, in- dividuals that are easily identified to species occur together in the same microhabitat over a zone of 1.6 km without showing any obvious evidence of intergradation. Aside from this narrow contact zone, the distributions of G. copei and G. wislizenii are widely separated. Paseo de San Matias is a low elevation dispersal corridor that connects the lower Colorado Desert with the coastal region of northwestern Baja Cali- fornia. Several desert species extend their ranges westward toward the Pacific coast by way of this corridor and some coastal species nearly reach the desert by extending eastward (Welsh and Bury, 1984). It may appear as though G. copei and G. wislizenii are geographic variants and that the pat- tern change is the result of in situ selection where the habitat changes from extremely xeric creosote desert to more mesic mountainous terrain. How- ever, typical G. copei occur in the lower Colorado Desert region in the vicinity of Bahia de San Luis Gonzaga, documenting that the distinctive color pattern of G. copei is not another G. wislizenii pat- tern type that appears only in mesic habitats. Gam- belia copei in the Bahia de San Luis Gonzaga region are approached by G. wislizenii in the vicinity of Puertocitos, where they are separated by a trans- verse volcanic field that is 31.5 road km in width. This lava field extends from the peninsular ranges to the edge of the Gulf of California and appears to be a dispersal barrier for Gambelia. Because the color pattern differences noted above are main- tained in these populations, which occur in essen- tially identical habitats that are separated only by the lava field, the notion that the G. copei and G. wislizenii color pattern differences are the result of in situ selection is unlikely. Nevertheless, because this taxonomic decision is based only on differences in coloration that are relatively subtle, on a single osteological character that differs in frequency (the presence of a well-developed tubercle on the an- terolateral margin of the postorbital was present in all G. copei examined [n = 8], whereas in G. wisli- zenii, the tubercle usually is absent [present in four of 49 specimens]), and on presumed reproductive isolation in this region, the recognition of G. copei as a full species is considered tentative. Electropho- retic analyses of the Paseo de San Matias popula- tions are planned in order to determine if fixed al- lelic differences can be detected that are consistent with the dorsal color pattern data. Montanucci (1978) considered the populations of Gambelia on Isla Tiburon and coastal Sonora be- tween Puerto Libertad and Bahia Kino to be con(sub)specific with copei. Although there are no- table similarities between certain individuals from the coastal Sonoran region and those from Baja Cal- ifornia (particularly in CAS 17050 from the south- eastern end of Isla Tiburon), they differ in that the Sonoran lizards have spots that continue onto the dorsal surface of the head, whereas G. copei nearly always lack this spotting. While some individuals from coastal Sonora clearly resemble those of Baja California, the majority examined here were char- acteristic of those of the remaining portions of So- nora. Gambelia coronal Norell Gambelia corona Norell, 1989:1 1; fig. 10. Type locality: LACM locality 7058, Vallecito Badlands, Anza-Borrego Desert State Park (holotype: LACM 42880). Etymology. — From the Latin corona, a crown, in reference to the distinctive characteristics of the frontal and frontoparietal suture. Diagnosis. — Gambelia corona t is distinguished from other Gambelia by the presence of the fron- toparietal suture anterior to the posterior extent of the orbits. It is further distinguished from Gambelia copei and G. wislizenii by the presence of a trans- versely concave frontal bone. Distribution. — Known only from the type locality. Remarks. — Gambelia corona f is an extinct spe- cies known only from a fossilized skull and man- dibles. Black-and-white photographs of dorsal and lateral views of the skull were provided by Norell (1989). Gambelia silus Stejneger Crotaphytus silus Stejneger, 1890:105. Type locality: “Fresno, Cal.” (holotype: USNM 11790A). Crotaphytus wislizenii— Cope, 1900:255. 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 103 Gambelia wislizenii silus— Smith, 1946:164; pi. 31. Crotaphytus ( Gambelia ) wislizeni silus— Weiner and Smith, 1965: 187. Gambelia silus— Montanucci, Axtell, and Dessauer, 1975:339. Gambelia sila— Jennings, 1987:1 1. Etymology. — From the Latin silus, snub-nosed, in reference to the blunt snout of this species. Diagnosis. — Gambelia silus is diagnosed from G. corona\ by the presence of a frontoparietal suture that is posterior to the posterior border of the orbits. It is diagnosed from G. wislizenii and G. copei by the presence of territoriality, male breeding color- ation, vertebrae with notched zygosphenes and zyg- antra, sexual dimorphism wherein males are larger than females, and in its truncated snout. Variation (n = 15). — Rostral approximately four times wider than high, usually rectangular in shape. Rostral bordered by six to eight postrostrals. Re- maining snout scales irregularly arranged, an en- larged middorsal series may be present. Nasals sep- arated by six to nine internasals. Frontonasals oc- casionally enlarged. Canthals four; posterior one or two wider than high; six to ten scales separate can- thals of left and right sides. Supraorbital semicircles absent, although slightly enlarged scales correspond- ing to the supraorbital series occasionally evident. Supraoculars small, flat or convex, smooth, becom- ing progressively larger medially such that medial scales are two to four times larger than lateral ones. Circumorbitals absent. Superciliaries eight to 13, extremely elongate medial scale present. Palpebrals ovoid, slightly convex, may be interspersed with numerous interstitial granules. Preoculars, subocu- lars, and postoculars form an arc of five to eight rectangular scales, second, third, or fourth scale elongate. Supralabials 13 to 16, usually slightly lon- ger than high except anteriormost scale, which is square. Lorilabials in one to four rows, ovoid to rectangular, juxtaposed, separating supralabials from suboculars and nasals. Aperture of external auditory meatus rectangular or ovoid, often constricted at or above the midpoint, approximately three to four times higher than wide, with small, strongly convex, somewhat conical auricular scales lining anterior margin. Mental pentagonal, one to 1.5 times wider than high, bordered laterally by anterior infralabials and posteriorly by a pair of enlarged postmentals. Postmentals separated from infralabials by subla- bials on at least one side. Chinshields weakly dif- ferentiated or undifferentiated. Infralabials 12 to 16, square or wider than high, inferior border convex. Gulars convex and beadlike; each scale separated from adjacent scales by numerous asymmetrically arranged interstitial granules. Dorsal scales in approximately 156 to 182 rows midway between forelimb and hindlimb insertions. Tail long, cylindrical in both sexes and all age groups. Paired, median row of subcaudals not larger than adjacent subcaudals and lateral caudals. Enlarged postanal scales present in males. Deep postfemoral dermal mite pocket present at hindlimb insertion. Femoral pores 1 5 to 20, femoral pores do not extend beyond angle of knee, separated medially by 17 to 25 granular scales. Subdigital la- mellae on fourth toe 16 to 20. Coloration in Life. — The dorsal base color ranges from pale tan, light or dark gray, or brown and the ventrum is white or yellowish. The dorsum is marked with seven to ten broad, pale transverse bars that may or may not be offsetting. Dark spots are often present between the pale crossbars and generally extend onto the temporal region of the head. The crossbars occasionally may be fragmented into light spots and a vertebral stripe may be present (Mon- tanucci, 1965). Spots and crossbars similar to those of the back are generally present on the limbs and tail, although the crossbars may be absent from the forelimbs. The tail becomes banded distally as de- scribed in the generic account. The posterior of the thigh and the underside of the tail in juveniles is suffused with yellow pigments. Males in certain parts of the range (particularly the foothills surrounding the San Joaquin valley) de- velop a breeding color composed of either a bright rusty red suffusion of the abdomen and the ventral and dorsal surfaces of the hindlimbs and tail or a bright salmon color that extends over the entire ven- tral surface of the body and limbs, sometimes in- cluding the gular region as well (Montanucci, 1965). Occasionally, individuals may develop this color- ation only laterally (Montanucci, 1970). Gravid col- oration in this species is similar to that of G. copei and G. wislizenii in that the orange or red pigments are deposited on the lateral surfaces of the head and flanks, on the under surface of the tail, and occa- sionally on the thighs. However, this pattern differs from that of G. copei and G. wislizenii in that the pigments are generally deposited in a single row along each flank, rather than in two rows (Montan- ucci, 1970). The dorsal pattern of Gambelia silus was de- scribed more fully in Van Denburgh (1922), Smith (1946), and Montanucci (1965, 1970). Size. — This species exhibits strong sexual dimor- 104 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 phism with males reaching larger adult size (maxi- mum observed SVL - 120 mm) than females (max- imum observed SVL =111 mm; Tollestrup, 1979, 1982). Distribution (Fig. 51 ). — Gambelia silus is restrict- ed to the San Joaquin valley of California and its surrounding foothills. They range between “the old town of Carnegie in Corral Hollow,” San Joaquin County, in the north to the Cuyama Valley and base of the Tehachapi Mountains in the south. The spe- cies apparently does not contact G. wislizenii pres- ently, although Montanucci (1970) identified an iso- lated population of putative hybrid origin between the two species in the Cuyama River drainage sys- tem southwest of the southern end of the San Joa- quin valley. Although G. silus and G. wislizenii are isolated from one another, Gambelia wislizenii ap- proaches G. silus in the Cuyama valley drainage where G. wislizenii occurs above 1 1 00 m and G. silus occurs below 790 m (Montanucci, 1970). Fossil Record.— No fossil specimens have been referred to this species, although Brattstrom (1953) considered measurements of two maxillae taken from McKittrick, Kern County, California, a local- ity within the current distributional confines of Gambelia silus, to conform more closely to extant G. wislizenii than to G. silus. However, examination of his figures renders this observation suspect as neither fossil has a complete nasal process. On dis- tributional grounds, it would appear more likely that these specimens represent G. silus. Because the ma- terial has not been reexamined here, the reference to G. wislizenii is considered questionable. Natural History. — Montanucci (1965, 1967, 1970) and Tollestrup (1979, 1982, 1 983) studied the ecol- ogy of Gambelia silus and all of the comments pro- vided here are taken from these references unless otherwise noted. According to Montanucci (1965), the species inhabits sparsely vegetated plains, alkali flats, low foothills, canyon floors, large washes, and arroyos. They prefer open habitat and are absent or rare in areas with dense vegetation or tall grass. As is the case with G. wislizenii, the species appears to be most common in areas with abundant rodent burrows. Common vegetational associates include grasses ( Stipa ), saltbush ( Atrip/ex ), and iodinebush ( Al/enrolfea occidentals). In contrast with Gambelia wislizenii [and presum- ably G. copei ), G. silus is highly territorial and males from many, but not all, populations develop rusty red coloration during the breeding season (Montan- ucci, 1965; Tollestrup, 1979, 1982). The activity season commences in late March or early April and extends through late September, although some ju- veniles may remain active into October given fa- vorable weather conditions (Montanucci, 1965; Tollestrup, 1979). The mating season occurs pri- marily in late April and May, although Germano and Williams ( 1 992) observed gravid females as late as mid-July, and young hatch in late July or early August (Montanucci, 1965; Tollestrup, 1979, 1983). Clutch size is smaller than that of G. wislizenii, with a range of two to six and a mean of 2.90 (Tollestrup, 1979, 1982) to 3.30 (Montanucci, 1970). Germano and Williams (1992) documented that as many as four clutches may be deposited per reproductive season. Gambelia silus shares a number of behavioral similarities with G. copei and G. wislizenii. All three are often observed basking on small roadside rocks and the berms along the edges of graded dirt roads. “Freeze” behavior (Montanucci, 1965), wherein threatened individuals run to the base of a nearby bush, flatten themselves to the ground, and remain motionless (presumably as a means of avoiding de- tection) is also a shared behavior. Montanucci (1965) indicated that Gambelia silus feeds primarily upon locusts (Orthoptera), cicadas (Homoptera), and small lizards, including Uta stansburiana, Phrynosoma coronatum, Cnemidoph- orus tigris, and Sce/oporus magister. Germano and Williams (1994) observed that G. silus eat young conspecifics, as well. Tollestrup (1979) found no ev- idence of lizard predation at her southern San Joa- quin valley study sites and noted the following ar- thropod prey items: orthopterans, coleopterans, hy- menopterans, dipterans, homopterans, lepidopter- ans, and spiders. Regional or seasonal variation may explain the discrepancies in food preferences found in these studies. Montanucci (1965) noted predation on Gambelia silus by several avian species including Loggerhead Shrikes ( Lanius ludovicianus), American Kestrels {Falco sparverius ), Burrowing Owls {Athene cuni- cularia ), and Greater Roadrunners {Geococcyx cal- ifornianus). Prarie Falcons {Falco mexicanus ) are also known to capture this species (Germano and Carter, 1995). Montanucci (1965) also observed predation by the coachwhip snake {Masticophis fla- gellum) and the gopher snake ( Pituophis melano- leucus). Other potential predators include the spot- ted skunk {Spi/ogale putorius) and the ground squir- rel {Spermophilus beecheyi), both of which con- sumed G. silus when captured together in barrel traps, as well as the coyote {Canis latrans ), badger {Taxidea taxus), glossy snake {Arizona elegans), long- 1996 McGUIRE — SYSTEM ATICS OF CROTAPH YT1D LIZARDS 105 nosed snake ( Rhinocheilus lecontei ), and common kingsnake ( Lampropeltis getula). Remarks.— Gambelia silus is now extinct over much of it historical range due primarily to habitat degradation associated with agricultural develop- ment of the San Joaquin valley. As of 1990, only seven percent of the San Joaquin valley had not been altered by agricultural and urban development 106 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 (Steinhart, 1990). As a result, this species is a fed- erally and state listed endangered species. Illustrations. — Montanucci (1965) provided a line drawing of the snout squamation; black-and-white photographs were provided by Van Denburgh (1922), Smith (1946), and Pickwell (1972); color illustrations were given in Smith and Brodie (1982) and Stebbins (1985); color photographs were pro- vided in Behler and King (1979) and Steinhart (1990). Gambelia wislizenii Baird and Girard (Fig. 30 A) Crotaphytus wislizenii Baird and Girard, 1852:69. Type locality: “near Santa Fe,” New Mexico (holotype: USNM, now lost or destroyed); invalid holotype (Yarrow 1882a): USNM 2770; invalid lectotype (Tanner and Banta 1963): USNM 2685. Crotaphytus gambelii Baird and Girard (syn. fide Cope, 1 900), 1852:126. Type locality: “Not precisely known. . .California” (holotype: USNM 2722). Cro/ap/ty/M5/a5c/a/a5Hallowell(syn. fide Cope, 1900), 1852:207. Type locality: “Sand hills, at the lower end of the Jornada del Muerte, New Mexico” (holotype: USNM 2736). Leios. [aurus] fasciatus—'Dnmenl, 1856:533. L. [eiosaurus] hallowellii Dumeril (substitute name for L.fascia- tus Hallowell, 1852), 1856:533. Crotaphytus {Gambelia) wislizenii— Baird, 1858:253. Gambelia wislizenii— Smith, 1946:158; fig. 57, 68; pi. 30. Crotaphytus ( Gambelia ) wislizeni punctatus Tanner and Banta, 1963: 1 38; fig. 1-5. Type locality: “Yellow Cat Mining District approximately 10 miles south of U.S. Highway 50-6, Grand County, Utah” (holotype: BYU 20928). Crotaphytus {Gambelia) wislizeni— Weiner and Smith, 1965: 186; fig. 1-6. Crotaphytus wislizenii neseotes— Banta and Tanner (syn. fide Montanucci, 1978), 1968:186; fig. 1-5. Type locality: “Cedros Island; west coast of Baja California Norte, Mexico” (holotype: CAS 79872). Gambelia wislizeni— Montanucci, Axtell, and Dessauer, 1975: 339. Crotaphytus wislizeni maculosus— Tanner and Banta, 1977:230; fig. 2-4. Type locality: “approximately 200m W of the lookout point along Nevada Highway 33, west side of Pyramid Lake, Washoe County, Nevada” (holotype: BYU 32685). Etymology. —Named in honor of Dr. Frederick Adolphus Wis- lizenus, an army surgeon, who collected the original type speci- men. Diagnosis. — Gambelia wislizenii is distinguished from G. corona^ by the absence of a broad, trans- versely concave frontal bone, the presence of a fron- toparietal suture posterior to the posterior extent of the orbits, and an elongate and slender nasal process of the premaxilla. It is diagnosable from G. silus in its absence of male breeding coloration, absence of sexual dimorphism wherein males are larger than females (the reverse condition is present), absence of notched zygosphenes and zygantra, and in the possession of an elongate (rather than truncated) and slender nasal process of the premaxilla. For a di- agnosis distinguishing G. wislizenii and G. copei, see discussion under the G. copei taxonomic account. Variation ( n = 20). — Rostral approximately four times wider than high, usually rectangular in shape. Rostral bordered by five to eight postrostrals. Re- maining snout scales irregularly arranged, an en- larged middorsal series may be present. Nasals sep- arated by six to nine internasals. Frontonasals oc- casionally enlarged. Canthals four; posterior one or two wider than high; seven to nine scales separate canthals of left and right sides. Supraorbital semi- circles absent. Supraoculars small, flat or convex, smooth, becoming progressively larger medially such that medial scales are two to four times larger than lateral ones. Circumorbitals absent. Superciliaries seven to 13, extremely elongate medial scale pres- ent. Palpebrals ovoid, slightly convex, may be in- terspersed with numerous interstitial granules. Preo- culars, suboculars, and postoculars form an arc of four to seven rectangular scales, second, third, or fourth scale elongate. Supralabials 12 to 17, usually slightly longer than high except anteriormost scale, which is square. Lorilabials in one to four rows, ovoid to rectangular, juxtaposed, separating su- pralabials from suboculars and nasals. Aperture of external auditory meatus rectangular or ovoid, often constricted at or above the midpoint, approximately two to four times higher than wide, with small, strongly convex, somewhat conical auricular scales lining anterior margin. Mental pentagonal, one to 1.5 times wider than high, bordered laterally by an- terior infralabials and posteriorly by a pair of post- mentals that may be enlarged. Postmentals almost always separated from infralabials by sublabials on at least one side. Chinshields weakly differentiated or undifferentiated. Infralabials 12 to 17, square or wider than high, inferior border convex. Gulars usu- ally flat, but occasionally convex and beadlike (es- pecially in southern portion of range); each scale may be separated from adjacent scales by numerous asymmetrically arranged interstitial granules. Dorsal scales in approximately 158 to 224 rows midway between forelimb and hindlimb insertions. Tail long, cylindrical in both sexes and all age groups. Paired, median row of subcaudals not larger than adjacent subcaudals and lateral caudals. Enlarged postanal scales present in males. Deep postfemoral dermal mite pocket present at hindlimb insertion. Femoral pores 1 5 to 25, femoral 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 107 pores extend beyond angle of knee, separated me- dially by 14 to 25 granular scales. Subdigital la- mellae on fourth toe 18 to 25. Coloration in Life. — The dorsal pattern of G. wis- lizenii is extremely variable, with several color morphs that are more or less confined to specific geographic regions. These pattern classes conform to the subspecies wislizenii, punctatus, and macu- losus (Montanucci, 1978; although he recognized an additional unnamed pattern class as well) that are not recognized here. They differ most notably in the size of the dorsal spots (large in the maculosus pat- tern class, intermediate in wislizenii, and small in punctatus), as well as in the character of the dorsal transverse bars. The dorsal spots are often scattered over the dorsum irregularly, extend well down onto the flanks, and continue onto the dorsal and lateral surfaces of the head. The dorsal base color for most individuals is white, cream, or gray, although some individuals apparently may approach the brown col- oration of northern G. copei (Montanucci, 1978). Gravid coloration in G. wislizenii is similar to that of G. copei with orange or red spots often pres- ent on the head and/or neck, in two rows of spots on each flank, and on the ventral surface of the tail. The red or orange pigments occasionally may extend onto the thighs. Males lack any form of breeding coloration. A more detailed description of geographic vari- ation in the dorsal pattern of Gambelia wislizenii is provided in Montanucci (1978). Size. — This species exhibits strong sexual dimor- phism with females reaching larger adult size (max- imum observed SVL = 144 mm) than males (max- imum observed SVL =119 mm; Tollestrup, 1979, 1982). Distribution (Fig. 52). — Gambelia wislizenii oc- curs in the western United States and northern Mex- ico, ranging from eastern Oregon and southern Ida- ho in the north, at least as far south as central Sonora in the west, and southern Coahuila or northern Za- catecas in the east. This species extends westward well beyond the limits of the lower Colorado Desert in southern California where it has been collected at Temecula, near Vail Lake, and at Arlington in Riverside County. However, a specimen purport- edly collected at Arcadia, Los Angeles County (FMNH 203919), seems suspect. The species ap- pears to be absent from the high elevation moun- tains of eastern Arizona and adjacent western New Mexico. Its distribution also appears to be limited in Texas, with a number of specimens known from the sandy northern portion of the Texas panhandle and from the Chihuahuan Desert habitats between Big Bend National Park and El Paso. It is unclear whether G. wislizenii is continuously distributed in the western portion of Texas between Reeves, Ward, and Crane counties and the southern portions of Brewster and Presidio counties. Specimens are rel- atively few from most of northern Mexico, but it appears that G. wislizenii is completely excluded from the higher portions of the Sierra Madre Oc- cidental of eastern Sonora and western Chihuahua. Gambelia wislizenii and G. copei occur together in a narrow zone of syntopy in northern Baja California which is denoted in Figure 52 by an oblong oval marking (for a more extensive discussion of this zone of syntopy, see the G. copei account above). The two northern Oregon localities shown on Figure 52 are old records from The Dalles, Wasco County, and Hat Rock, Umatillo County. The symbol “?” shown on Figure 52 represents a record from Che- ney, Spokane County, Washington. The northern Oregon and Washington records should be consid- ered questionable until verified by additional field work. Fossil Record. —Numerous Pleistocene fossils have been referred to this species (Estes, 1983) in- cluding a pair of maxillae that may be more properly referred to Gambelia silus (see G. si/us account for comments). Natural History. — There is extensive literature as- sociated with the natural history and ecology of Gambelia wislizenii. The reader is referred to the following papers for a more detailed discussion of this topic: McCoy, 1967; Montanucci, 1967, 1970, 1978; Turner et al., 1969; Tanner and Krogh, 1974a, 1974/?; Essghaier and Johnson, 1975; Parker and Pianka, 1976; Tollestrup, 1979, 1982, 1983; and Mitchell, 1984. This widespread species occurs in a number of habitat types, although it is found pri- marily on desert flats and lower foothills character- ized by sparse vegetation. Throughout much of its range in the Sonoran, Mojave, Great Basin, and Chihuahuan deserts it is found in flatlands in as- sociation with creosote bush ( Larrea tridentata ) as well as other xerophilic plants. In the Pyramid Lake region of northwestern Nevada, it is found in as- sociation with filaree storksbill (Erodium cicutar- ium ), mormon tea ( Ephedra nevadensis), four-wing saltbush ( Atriplex canescens ), and Grayia spinosa (Snyder, 1 972). Tollestrup (1979, 1982, 1983) stud- ied G. wislizenii near California City, California, where the dominant shrub was creosote bush {Ear- 108 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Fig. 52. — Geographic distribution of Gambelia wislizenii. The elongate ovoid marking in northern Baja California represents the narrow zone in which G. wislizenii and G. copei occur together. The “?” denotes a questionable locality record from Cheney, Spokane County, Washington. 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 109 rea tridentata ), and other major components of the vegetation included saltbush (Atrip/ex), Haplopap- pus, Lycium andersonii, and Dalea. McCoy (1967) discussed the ecology of this species in the Colorado River valley, Mesa County, Colorado, where it was found in association with greasewood ( Sarcobatus venniculatus) and big sage (Artemisia tridentata). In southeastern Arizona, the species was found on sand dunes with sand sage (Artemisia filifolia) and indigo bush (Dalea sp.) and on bajadas characterized by cat-claw acacia (Acacia constricta), jimmyweed (Haplopappus taenuisecta), Opuntia, and Agave (Mitchell, 1984). Near the northeastern limits of its range (1 1.5 mi S Monahans, Ward County, Texas), Tinkle (1959) found them on sandy flatlands in as- sociation with mesquite (Prosopis), creosote bush (Larrea tridentata). Acacia, and dwarf shin oak (Quercus havardii). Montanucci (1970) found G. wislizenii restricted to the high elevation (above 3600 ft) pinyon-juniper woodland habitats of the Cuyama and Lockwood valleys, southern California, near the hybrid zone between this species and G. silus. How- ever, the pinyon-juniper zone is thought to be sub- optimal habitat for G. wislizenii and they are often absent from such areas (Tanner and Jorgenson, 1 963; McCoy, 1967). Gambelia wislizenii appears to be most common on sparsely vegetated flatlands with large numbers of rodent burrows (Tanner and Banta, 1963; McCoy, 1967; Nussbaum et al., 1983). Unlike Crotaphytus and Gambelia silus, G. wis- lizenii lacks territoriality (McCoy, 1967; Montan- ucci, 1970; Tollestrup, 1979, 1982, 1983) and there is often much overlap in home ranges (Tollestrup, 1979, 1983). Females may even nest communally (Parker and Pianka, 1976). Females attain much larger size than males and appear to consume a high- er proportion of vertebrate prey (Parker and Pianka, 1976; Tollestrup, 1979, 1982, 1983). Southern pop- ulations reach larger adult sizes than more northern populations which Parker and Pianka (1976) again linked to an increased emphasis on vertebrate prey. Gambelia wislizenii are ambush predators, often resting in the shadows at the base of a bush before dashing out to capture passing prey items (Tolles- trup, 1979, 1983). They are able to move with great speed and have been observed to leap as high as 0.6 m to capture flying insects (Franklin, 1914). Known prey items include arthropods, especially orthop- terans, as well as coleopterans, lepidopterans, hy- menopterans, hemipterans, homopterans, dipter- ans, isopterans, neuropterans, and arachnids (Knowlton and Thomas, 1936; McCoy, 1967; Sny- der, 1972; Tanner and Krogh, 1974a, 19746; Essgh- aier and Johnson, 1975; Parker and Pianka, 1976; Tollestrup, 1979; Mitchell, 1984). Vertebrate prey include the lizards Callisaurus draconoides, Cne- midophorus tessellatus, C. tigris, Uta stansburiana, Phrynosoma platyrhinos, Sceloporus graciosus, S. undulatus, smaller G. wislizenii, and small snakes, as well as the pocket mouse Perognathus longimem- bris (Taylor, 1912; Richardson, 1915; Camp, 1916; VanDenburgh, 1922; Knowlton and Thomas, 1936; Banta, 1967; McCoy, 1967; Snyder, 1972; Tanner and Krogh, 1974a, 19746; Parker and Pianka, 1976; Tollestrup, 1979, 1983; Pietruszka et al., 1981; Crowley and Pietruszka, 1983). As has been re- ported for several Crotaphytus species (i.e., C. bi- cinctores, C. vestigium), Lycium berries are often consumed and may even represent a preferred food item during parts of June and July (Tanner and Krogh, 1974a). Turner et al. (1969) observed in- dividuals climbing into Lycium bushes to eat the berries, indicating that this plant material is not consumed inadvertently. Jorgensen and Orton (1962) collected two G. wislizenii in traps baited with oatmeal and found oatmeal in the stomach contents of both. Gambelia wislizenii shares a number of behav- ioral similarities with G. copei and G. silus. All three are often observed basking on small roadside rocks and the berms along the edges of graded dirt roads. “Freeze” behavior (Brooking, 1934; McCoy, 1967) wherein threatened individuals run to the base of a nearby bush, flatten themselves to the ground, and remain motionless (presumably as a means of avoiding detection) is also a shared behavior. A be- havior present in G. wislizenii but not yet noted in other Gambelia is vocalization (Taylor, 1912; Jor- gensen et al., 1 963; Wever et al., 1 966; Crowley and Pietruszka, 1 983). Wever et al. ( 1966) described the sound emitted as “vocal cries of a wailing or moan- ing character.” The ability to vocalize, although ex- tremely unusual within iguanians, has also been not- ed in C. bicinctores (Smith, 1974) suggesting that all crotaphytids may possess this ability. Accounts of predation on Gambelia wislizenii are rare in the literature. Tollestrup (1979) observed a failed predation attempt on an adult female by a Prairie Falcon (Falco mexicanus). Tollestrup (1979) considered the following species to be potential predators at the California City study site: the coach- whip snake (Masticophis flagellum), sidewinder 110 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 ( Crotalus cerastes ), Mojave rattlesnake ( Crotalus scutulatus). Loggerhead Shrike ( Lanius ludovici- anus ), raptors. Burrowing Owl ( Speotyto cunicular- ia), badger ( Taxidea taxus), coyote (Canis latrans ), and kit fox ( Vulpes macrotis). To this list can be added G. wislizenii, which commonly preys on smaller individuals of its own species and a number of saurophagous snakes that occur within its range, such as the patch-nosed snake ( Salvadora sp.), the common kingsnake ( Lampropeltus getula), the go- pher snake ( Pituophis melanoleucus), the glossy snake {Arizona eiegans), and the long-nosed snake ( Rhin - ocheilus lecontei). The length of the activity season of Gambelia wis- lizenii varies latitudinally. Northern and northeast- ern populations (western Colorado, Utah, north- western Nevada, and Ward County, Texas) may not emerge from hibernation until early or even late May (Tinkle, 1959; McCoy, 1967; Snyder, 1972; Parker and Pianka, 1976). Adults enter hibernation in early August and, thus, may have activity seasons less than three months in length (McCoy, 1967). Individuals from southern populations emerge from hibernation in late March or early April (south- eastern Arizona, vicinity of California City, Joshua Tree National Monument) and enter hibernation in late August to late October (Miller and Stebbins, 1964; Tollestrup, 1979; Mitchell, 1984). Reproduc- tion appears to be concentrated in late May and early June in the California City and southeastern Ari- zona populations and after these dates gravid fe- males were not observed (Tollestrup, 1979, 1982; Mitchell, 1984). In Utah and western Colorado, gravid females were observed between early June and early July, indicating that the reproductive sea- son is pushed back by a few weeks in more northern populations (McCoy, 1967; Parker and Pianka, 1976). Clutch size also varies from population to population, with mean clutch sizes ranging between 5.15 (Robison and Tanner, 1962) and 7.3 (McCoy, 1967; Mitchell, 1984). Most studies have found no evidence of multiple clutch production (McCoy, 1967; Tanner and Krogh, 1974a; Parker and Pian- ka, 1976; Tollestrup, 1979, 1982; Mitchell, 1984), although Turner et al. (1969) observed second clutches in a southern Nevada population. Gambelia wislizenii develop vibrant orange or reddish gravid coloration shortly before ovulation (as do all crotaphytid species). This coloration is maintained throughout the gravid period and is lost soon after parturition. The fecal matter of females that are losing their gravid coloration may be heavi- ly saturated with similar orange pigments and this may provide a clue to the yet-to-be-identified pig- ment type responsible for this coloration. Illustrations. —Numerous photographs and illus- trations have been published. Detailed black-and- white illustrations of the entire animal were pro- vided by Baird and Girard (1852c), Hallowell (1852), Baird (1859), and Stebbins (1954); ventral head squamation (Stebbins, 1954); head, limb, and pre- anal squamation by Cope (1900); skull, pelvic and pectoral girdles by Weiner and Smith (1965); an- terior body and head musculature by Robison and Tanner (1962); black-and-white photos were pre- sented by Van Denburgh (1922), Tanner and Banta (1963, 1977), Pickwell (1972), Montanucci (1978), and Nussbaum et al. (1983); color illustrations by Stebbins (1985) and Conant and Collins (1991); co- lorized photo by Ditmars (1920); color photographs were provided by Leviton (1971), Behler and King (1979), Hammerson (1986), and Garrett and Barker (1987). Taxonomic Remarks. —The subspecies Gambelia wislizenii punctatus and G. w. maculosus often are considered to be synonyms of G. w. wislizenii and in their descriptions, broad intergrade zones were identified (Tanner and Banta, 1963, 1977). Fur- thermore, Montanucci (1978) showed that the G. w. maculosus, G. w. punctatus, and G. w. wislizenii dor- sal pattern classes occur sporadically throughout the range of G. wislizenii. Based on these data, G. w. maculosus and G. w. punctatus are here considered to be pattern classes and are synonymized with G. wislizenii. No official holotype specimen of Crotaphytus wis- lizenii was designated by Baird and Girard (1852a) and this created some confusion when later workers attempted to rectify the situation. Tanner and Banta (1963) designated a lectotype (which they referred to as a holotype) for C. wislizeni after recognizing that Yarrow (1882a) had incorrectly designated USNM 2770 as the type specimen, and that the original specimen figured by Baird and Girard (1852c) from near Santa Fe, New Mexico, had been lost or destroyed. The specimen of Crotaphytus wis- lizenii (USNM 2770) designated by Yarrow (1882a) was collected by H. Baldwin Mollhausen in Colo- rado probably in 1853-1854 after C. wislizenii had already been described (Tanner and Banta, 1963) and therefore could not have represented the orig- inal type specimen described by Baird and Girard (1852a). The designation of a lectotype requires that the original description of the species was based on 1996 McGUIRE— SYSTEM ATICS OF CROTAPHYTID LIZARDS a series of syntypes, rather than a single holotype. In their description, Baird and Girard (1852a) stat- ed: “Head proportionally narrow and elongated; ce- phalic plates and scales on the back very small: yel- lowish brown, spotted all over with small patches of deeper brown or black. Caught near Santa Fe, by Dr. Wislizenius (sic); specimens of the same species sent in by Lieut. Col. J. D. Graham, collected be- tween San Antonio and El Paso del Norte.” Thus, the description of the species appears to be based on the specimen from near Santa Fe, whereas the other specimens were referred to the species after- ward. Therefore, according to the International Commission of Zoological Nomenclature (1985), it 1 1 1 is inappropriate to designate a lectotype for the spe- cies. Furthermore, except under “exceptional cir- cumstances,” a neotype is not to be designated ei- ther. “Exceptional circumstances,” such as when a neotype is necessary in the interests of stability of nomenclature, clearly are not evident at present. Therefore, USNM 2685 (collected between San An- tonio and El Paso del Norte), which was designated as the lectotype by Tanner and Banta ( 1963), should not be recognized as either a lectotype or a neotype. If the designation of a neotype should become nec- essary, it should be collected in the immediate vi- cinity of Santa Fe, New Mexico. KEY TO THE SPECIES OF CROTAPHYTUS AND GAMBELIA A key to the species of Crotaphytus is not difficult to produce for adult males because most species are easily distinguished on the basis of conspicuous col- or pattern characteristics. A key for adult females and juveniles of both sexes is more difficult because many of the characteristics that distinguish species are present only in adult males. Adult male Cro- taphytus are easily distinguished from females by the presence of conspicuous gular coloration, larger femoral pores with a greater quantity of exudate, and often by the presence of enlarged postanal scales. A number of additional sexually dichromatic fea- tures may also be employed depending on the spe- cies in question (see taxonomic accounts). With re- gard to Gambelia, the formulation of a key is dif- ficult for both sexes and all age classes due to vari- ation in the coloration of G. wislizenii and G. silus and the absence of distinctive features of squama- tion. With a few notable exceptions, all species of Crotaphytus and Gambelia are allopatrically dis- tributed with respect to their congeners. Thus, ge- ography is usually a reliable means for determining species identifications when morphology fails. For- tunately, where geographic overlap occurs, the spe- cies in question are easily distinguished. Note: The key to juvenile Crotaphytus does not include C. an- tiquus for which no specimens are available. Key to the Species of Crotaphytus (Adult Males) 1. Dorsal pattern comprised of white or pale gray net-like reticulations on a golden tan or brown dorsal base color, femoral pores jet black (Fig. 30C, D) 2 la. Dorsal pattern composed of white spots and/or dashes, with or without white transverse bars, on a brown, blue. green, tan, or straw yellow dorsal base color; femoral pores off-white or gray (Fig. 31, 32) 3 2. Dorsal coloration golden tan, groin patches absent (Fig. 30C, 33) reticulatus 2a. Dorsal coloration brown, groin patches present (Fig. 30D) antiquus 3. Anterior collar markings incomplete ventrally (do not pass through the gular fold (Fig. 33) collaris 3a. Anterior collar markings complete ventrally (Fig. 34, 35) 4 4. Tail round or nearly so in cross section without an off- white vertebral stripe; small melanic inguinal patches are present but confined to immediate vicinity of groin (Fig. 31 A, 34) nebrius 4a. Tail strongly compressed laterally with a white or off- white vertebral stripe; large melanic inguinal patches extend half way to the forelimb insertion or more (Fig. 3 1 B, 32A-D, 35) 5 5. Dorsal coloration aquamarine to cobalt blue; black oral melanin present (Fig. 3 1 B) dickersonae 5a. Dorsal coloration dark brown; black oral melanin absent (Fig. 32A-D) 6 6. Posterior collar markings absent or extremely reduced; white component of dorsal pattern composed of irreg- ularly arranged, elongate, wavy white lines (Fig. 32D) insular is 6a. Posterior collar markings present; white component of dorsal pattern composed of white spots and/or dashes, with or without regularly arranged white transverse bars (Fig. 31A-C; 32A, B) 7 7. White dorsal transverse bars present (Fig. 32C); olive green or yellow-orange ventrolateral breeding coloration present; posterior collar markings widely separated dor- sa*'y vestigium 7a. White dorsal transverse bars absent, olive green or yel- low-orange ventrolateral breeding coloration absent; posterior collar markings in contact or narrowly sepa- rated dorsally g 8. Granular ventrolateral reticulations present; pale orange or peach-colored transverse bands incorporated into brown dorsal base coloration; white bar that separates 1 12 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 anterior and posterior collar markings lacks green tint; hindlimb brown with a white or off-white reticulate pat- tern over entire dorsal surface bicinctores 8a. Granular ventrolateral reticulations absent; white bar that separates the anterior and posterior collar markings with a pale green tint; hindlimb spotted proximally but patternless from just above the knee to the distal ter- minus grismeri Key to the Species of Crotaphytus (Adult Females) 1. Dorsal pattern comprised of white or pale gray net-like reticulations, at least some of which surround gray or black pigments, on a golden tan or brown dorsal base color 2 la. Dorsal pattern composed of white spots and/or dashes, with or without white transverse bars, on a pale brown, bluish, greenish, tan, or straw yellow dorsal base color (Fig. 3 1C) 3 2. Postfemoral mite pockets absent; at least three of the scales of the right and left supraorbital semicircles in contact reticulatus 2a. Postfemoral mite pockets present; supraorbital semi- circles either separated by a continuous row of scales or with one or rarely two scales of the supraorbital semi- circles in contact antiquus 3. Black oral melanin present 4 3a. Black oral melanin absent 5 4. Antehumeral mite fold lacking; tail bright lemon yellow (Fig. 3 1C; note: this feature may prove to be variable) dickersonae 4a. Antehumeral mite pocket present (Fig. 28); tail not bright lemon yellow collaris or nebrius 5. Posterior collar markings absent; anterior collar mark- ings usually absent insularis 5a. Anterior and posterior collar markings present 6 6. White dorsal transverse bars present vestigium 6a. White dorsal transverse bars absent 7 7. Subadult females with orange tail; subadult and adult females with three melanic spots outlined in white along the lateral trunk surface grismeri 7a. Subadult females without orange tail; melanic spots out- lined in white usually absent from lateral trunk surface bicinctores Key to the Species of Crotaphytus (Juveniles) 1. Postfemoral mite pockets absent reticulatus la. Postfemoral mite pockets present (Fig. 29) 2 2. Antehumeral mite pocket absent dickersonae 2a. Antehumeral mite pocket present (Fig. 28) 3 3. Black oral melanin present collaris or nebrius 3a. Black oral melanin absent 4 4. A thin, pale tan dorsal caudal stripe is present and ex- tends anteriorly onto the dorsal pelvic region . . . grismeri 4a. A pale tan dorsal caudal stripe is lacking 5 5. Paired melanic keels on ventral surface of caudal ex- tremity . . vestigium or bicinctores (variable in bicinctores ) 5a. Paired melanic keels on ventral surface of caudal ex- tremity lacking insularis or bicinctores (variable in bicinctores ) Key to the Species of Gambelia (Adults of Both Sexes) 1. Reddish male breeding coloration present; snout trun- cated; gular pattern in both sexes consists of grayish or black linearly arranged spots silus la. Male breeding coloration absent; snout elongate; gular pattern in both sexes consists of longitudinally oriented black streaks 2 2. Dorsal spotting extends onto the temporal region of the head and often to the terminus of the snout; dorsal base coloration off-white or pale tan (Fig. 30A) wislizenii 2a. Dorsal spotting does not extend onto the dorsal surface of the head; dorsal base coloration dark brown or golden tan (Fig. 30B) copei Acknowledgments I would like to thank the following individuals and institutions for allowing me to examine specimens under their care: Philip Damiani, Darrel Frost, and Charles Meyers, American Museum of Natural History (AMNH); Jack Sites, Monte L. Bean Life Science Museum, Brigham Young University (BYU); Jacques Gauthier and Jens Vindum, California Academy of Sciences (CAS); Ellen Censky and the late C. J. McCoy, Carnegie Museum of Natural History (CM); William Duellman, Adrian Nieto, and John Simmons, The University of Kansas Museum of Natural History (KU); Robert Bezy and John Wright, Natural History Museum of Los Angeles County (LACM); Harry Greene, Uni- versity of California Museum of Vertebrate Zoology (MVZ); Os- car Flores-Villela, Museo de Zoologia “Alfonso L. Herrera,” Universidad Nacional Autonoma de Mexico (MZFC); Gregory Pregill, San Diego Natural History Museum (SDSNH); Richard Etheridge, San Diego State University (SDSU); David Canna- tella, Texas Memorial Museum (TNHC); Charles Lowe, Uni- versity of Arizona Department of Zoology (UAZ); Deborah Bak- ken and Steven Sroka, University of Illinois Museum of Natural History (UIMNH); Arnold Kluge and Greg Schneider, The Uni- versity of Michigan Museum of Zoology (UMMZ); Ronald Crombie, Kevin de Queiroz, Ronald Heyer, Addison Wynn, and George Zug, National Museum of Natural History (USNM); and Carl Lieb and Robert Webb, Laboratory for Environmental Bi- ology, The University of Texas at El Paso (UTEP). In addition, Richard Etheridge (REE), L. Lee Grismer (LLG), Bradford Hol- lingsworth (BDH), Ernest Liner (EL), and Jay Savage (JMS) al- lowed me to examine specimens from their personal collections. For assistance in the field I would like to thank Alfonso Del- gadillo-Espinoza, Richard Etheridge, Jerry Feldner, Marty Feld- ner, Erik Gergus, Jesse Grismer, L. Lee Grismer, Bradford Hol- lingsworth, Mario Mancilla-Moreno, Fernando Mendoza-Qui- jano, Sharon Messenger, Richard Montanucci, David Orange, Walter Schmidt-Ballardo, Eric Snow, and John Wiens. Kevin de Queiroz, Richard Etheridge, L. Lee Grismer, Paula Mabee, Shar- on Messenger, Steve Poe, Gregory Pregill, John Wiens, and an 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 113 anonymous reviewer provided valuable criticism of the manu- script. I thank Richard Montanucci for helpful discussions as well as the use of his extensive slide collection. John Huelsenbeck graciously provided the program that allowed me to recompute values specific to my data set. Financial support critical to the completion of this project was provided by the San Diego State University Department of Bi- ology, the San Diego State Univeristy Mabel Myers scholarship fund, the Society of Sigma Xi, the Theodore Roosevelt Memorial Fund of the American Museum ofNatural History, the San Diego Herpetological Society, and the California Academy of Sciences. Scientific collecting permits were provided by the states of Arizona and Texas. 1 am greatly indebted to Oscar Flores-Villela, Arturo Gonzales-Alonso, Erik Mellink, and Fernando Mendoza- Quijano, for obtaining permits (numbers 01303 and A00-700- (2) 01480) that allowed for collecting in Mexico. Literature Cited Allen, M. J. 1933. Report on a collection of amphibians and reptiles from Sonora, Mexico, with the description of a new lizard. Occasional Papers of the Museum of Zoology, Uni- versity of Michigan, 259:1-15. Andre, J. B., and J. A. MacMahon. 1980. Reproduction in three sympatric lizard species from west-central Utah. Great Basin Naturalist, 40:68-72. Arnold, E. N. 1986. Mite pockets of lizards, a possible means of reducing damage by ectoparasites. Biological Journal of the Linnean Society, 29:1-21. Avila, V. 1995. Biology: Investigating Life on Earth. Jones and Bartlett Publishers, Boston, Massachussetts, xxvii + 979 pp. Axtell, R. W. 1972. Hybridization between western collared lizards with a proposed taxonomic rearrangement. Copeia, 1972:707-727. . 1981. A peripheral Mexican record for Crotaphytus reticulatus found erroneous. Herpetological Review, 12:66. . 1989a. Interpretive Atlas of Texas Lizards. Crotaphytus collaris. Self published, East Alton, Illinois, 8:1-38. . 19896. Interpretive Atlas of Texas Lizards. Crotaphytus reticulatus. Self published, East Alton, Illinois, 9:1-8. Axtell, R. W., and R. R. Montanucci. 1977. Crotaphytus collaris from the eastern Sonoran Desert: Description of a previously unrecognized geographic race. Chicago Academy of Sciences Natural History Miscellenea, 201:1-8. Axtell, R. W„ and R. G. Webb. 1995. Two new Crotaphytus from southern Coahuila and the adjacent states of east-cen- tral Mexico. Bulletin of the Chicago Academy of Sciences, 16:1-15. Baird, S. F. 1858. Descriptions of new genera and species of North American lizards in the museum of the Smithsonian Institution. Proceedings of the Academy ofNatural Sciences of Philadelphia, 10:253-256. . 1859. Reptiles of the Boundary. United States and Mexican Boundary Survey Under the Order of the Lieut. Col. W. H. Emory. Washington, D.C., 35 pp. Baird, S. F., and C. Girard. 1852a. Characteristics of some new reptiles in the museum of the Smithsonian Institution. Proceedings of the Academy of Natural Sciences of Phila- delphia, 6:68-70. . 18526. Characteristics of some new reptiles in the mu- seum of the Smithsonian Institution. Proceedings of the Academy ofNatural Sciences of Philadelphia, 6:125-129. . 1852c. Appendix C. Reptiles. Pp. 336-364, in Explo- ration and Survey of the Valley of the Great Salt Lake of Utah, Including a Reconnaissance of a New Route Through the Rocky Mountains (H. Stansbury). Lippincott, Grambo and Company, Philadelphia, Pennsylvania, xiii + 421 pp. Ballinger, R. E„ and T. G. Hipp. 1985. Reproduction in the collared lizard, Crotaphytus collaris, in west central Texas. Copeia, 1985:976-80. Banta, B. H. 1 960. Notes on the feeding of the western collared lizard, Crotaphytus collaris baileyi Stejneger. Wasmann Journal of Biology, 1 8:309-3 1 1 . . 1967. Some miscellaneous remarks on Recent Nevada lizards. Occasional Papers of the Biological Society of Ne- vada, 16:1-5. Banta, B. H., and W. W. Tanner. 1968. The systematics of Crotaphytus wislizenii, the leopard lizards (Sauria: Iguani- dae). Part II. A review of the status of the Baja California peninsular populations and a description of a new subspecies from Cedros Island. Great Basin Naturalist, 28:183-194. Behler, J. L., and F. W. King. 1979. The Audobon Society Field Guide to North American Reptiles and Amphibians. Alfred A. Knopf, New York, New York, 719 pp. Blair, W. F., and A. P. Blair. 1941. Food habits of the collared lizards in northeastern Oklahoma. American Midland Nat- uralist, 26:230-232. Blanc, C. P., and C. C. Carpenter. 1969. Studies on the Iguanidae of Madagascar, III: Social and reproductive be- havior of Chalaradon madagascariensis. Journal of Herpe- tology, 3:125-134. Blanc, C. P., F. Blanc, and J. Rouault. 1983. The interre- lationships of Malagasy iguanids. Journal of Herpetology. 17:129-136. Bontrager, S. K. 1980. The autecology of Crotaphytus collaris. Unpublished M.S. thesis, Oklahoma State University, Still- water, 151 pp. Bostic, D. L. 1971. Herpetofauna of the Pacific coast of north central Baja California, Mexico, with a description of a new subspecies of Phyllodactylus xanti. Transactions of the San Diego Society ofNatural History, 16:237-264. Brattstrom, B. H. 1953. Records of Pleistocene reptiles from California. Copeia, 1953:174-179. . 1954. Amphibians and reptiles from Gypsum Cave, Nevada. Bulletin of the Southern California Academy of Sciences, 53:8-12. Brooking, W. J. 1934. Some reptiles and amphibians from Malheur County, in eastern Oregon. Copeia, 1934:93-95. Brown, A. E. 1903. Texas reptiles and their faunal relations. Proceedings of the Academy of Natural Sciences of Phila- delphia, 55:542-558. Burt, C. E. 1 928a. Insect food of Kansas lizards with notes on feeding habits. Journal of the Kansas Entomological Society, 1:58-68. . 19286. The synonymy, variation, and distribution of the collared lizard, Crotaphytus collaris (Say). Occasional BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 1 14 Papers of the Museum of Zoology, University of Michigan, 196:417-421. . 1929. The sexual dimorphism of the collared lizard, Crotaphytus collaris. Papers of the Michigan Academy of Science, Arts, and Letters, 10:417-421. . 1935. Further records of the ecology and distribution of amphibians and reptiles of the middle West. American Midland Naturalist, 16:31 1-336. Cadle, J. E. 1991. Systematics of the genus Stenocercus (Igu- ania: Tropiduridae) from northern Peru: New species and comments on relationships and distribution patterns. Pro- ceedings of the Academy of Natural Sciences of Philadelphia, 143:1-96. Camp, C. L. 1916. Notes on the local distribution and habits of the amphibians and reptiles of southeastern California in the vicinity of the Turtle Mountains. University of California Publications in Zoology, 12:503-544. Cannatella, D. C., and K. de Queiroz. 1989. Phylogenetic systematics of the anoles: Is a new taxonomy warranted? Systematic Zoology, 38:57-69. Carothers, J. H. 1984. Sexual selection and sexual dimor- phism in some herbivorous lizards. American Naturalist, 124:244-254. Carpenter, C. C. 1967. Aggression and social structure in igua- nid lizards. Pp. 87-105, in Lizard Ecology: A Symposium (W. W. Milstead, ed.), University of Missouri Press, Colum- bia, ix + 300 pp. Clark, H. 1946. Incubation and respiration of eggs of Crota- phytus c. collaris (Say). Herpetologica, 3:136-139. Clarke, R. F. 1965. An ethological study of the iguanid lizard genera Callisaurus, Cophosaurus, and Holbrookia. The Em- poria State Research Studies, 13:1-66. Cochran, D. M. 1961. Type specimens of reptiles and am- phibians in the U. S. National Museum. Bulletin of the U. S. National Museum, 220:104-105. Cochran, D. M., and C. J. Goin. 1970. The New Field Book of Reptiles and Amphibians. G. P. Putnam’s Sons, New York, New York, xxii + 359 pp. Cogger, H. G. 1992. Reptiles and Amphibians of Australia. Cornell University Press, Ithaca, New York, 775 pp. Collins, J. T. 1982. Amphibians and Reptiles in Kansas. Uni- versity of Kansas, Lawrence, xii + 356 pp. . 1991. Viewpoint: A new taxonomic arrangement for some North American amphibians and reptiles. Herpeto- logical Review, 22:42-43. Conant, R. 1975. Reptiles and Amphibians of Eastern and Central North America. Houghton Mifflin Company, Bos- ton, Massachusetts, xviii + 429 pp. Conant, R., and J. T. Collins. 1991. A Field Guide to Reptiles and Amphibians of Eastern and Central North America. Houghton Mifflin Company, Boston, Massachusetts, xviii + 450 pp. Cooper, W. E., Jr. 1984. Female secondary sexual coloration and sex recognition in the keeled earless lizard, Holbrookia propinqua. Animal Behaviour, 32:1 142-1 150. . 1988. Aggressive behavior and courtship rejection in brightly and plainly colored female keeled earless lizards (Holbrookia propinqua). Ethology, 77:265-278. Cooper, W. E., Jr., and D. Crews. 1988. Sexual coloration, plasma concentrations of sex steroid hormones, and re- sponses to courtship in the female keeled earless lizard ( Hol- brookia propinqua). Hormones and Behavior, 22:12-25. Cooper, W. E., Jr., and G. W. Ferguson. 1972. Steroids and color change during gravidity in the lizard Crotaphytus col- laris. General and Comparative Endocrinology, 18:69-72. . 1973. Estrogenic priming of color change induced by progesterone in the collared lizard, Crotaphytus collaris. Her- petologica, 29:107-1 10. Cooper, W. E., Jr., and N. Greenberg. 1992. Reptilian col- oration and behavior. Pp. 298 — 422, in Biology of the Rep- tilia. Volume 18 (C. Gans and D. Crews, eds.), Alan R. Liss, New York, New York, xiv + 564 pp. Cope, E. D. 1887. Catalogue of batrachians and reptiles of Central America and Mexico. Bulletin of the U. S. National Museum, 32:1-98. . 1900. The crocodilians, lizards, and snakes of North America. Annual Report of the U. S. National Museum (1898), part 2:151-1270. Cothran, M. L., and V. H. Hutchison. 1979. Effect of me- latonin on thermal selection by Crotaphytus collaris (Squa- mata: Iguanidae). Comparative Biochemistry and Physiol- ogy, 63A:46 1 — 466. Crowley, S. R., and R. D. Pietruszka. 1983. Aggressiveness and vocalization in the leopard lizard ( Gambelia wislizenii): The influence of temperature. Animal Behaviour, 31:1055- 1060. Dawson, W. R., and J. R. Templeton. 1963. Physiological responses to temperature in the lizard Crotaphytus collaris. Physiological Zoology, 36:219-236. de Queiroz, K. 1982. The scleral ossicles of sceloporine ig- uanids: A reexamination with comments on their phyloge- netic significance. Herpetologica, 38:302-31 1. . 1985. The ontogenetic method for determining char- acter polarity and its relevance to phylogenetic systematics. Systematic Zoology, 34:280-299. . 1987. Phylogenetic systematics of iguanine lizards: A comparative osteological study. University of California Publications in Zoology, 1 18:1-203. . 1989. Morphological and biochemical evolution in the sand lizards. Unpublished Ph.D. dissert., University of Cal- ifornia, Berkeley, 491 pp. . 1 992. Phylogenetic relationships and rates of allozyme evolution among the lineages of sceloporine sand lizards. Biological Journal of the Linnean Society, 45:333-362. de Queiroz, K., and J. Gauthier. 1992. Phylogenetic tax- onomy. Annual Review of Ecology and Systematics, 23:449- 480. Dingerkus, G., and L. D. Uhler. 1977. Enzyme clearing of alcian blue stained whole small vertebrates for demonstra- tion of cartilage. New Stain Technology, 52:229-232. Ditmars, R. L. 1920. The Reptile Book. A Comprehensive, Popularized Work on the Structure and Habits of Turtles, Tortoises, Crocodilians, Lizards and Snakes Which Inhabit the United States and Northern Mexico. Doubleday Page and Company, New York, New York, xxxii -I- 472 pp. Dixon, J. R., and J. W. Wright. 1975. A review of the lizards of the iguanid genus Tropidurus in Peru. Natural History Museum of Los Angeles County Contributions in Science, 271:1-39. Donoghue, M. J., R. G. Olmstead, J. F. Smith, and J. D. Palmer. 1 992. Phylogenetic relationships of dipscales based on rbcL sequences. Annals of the Missouri Botanical Garden, 79:333-345. Duellman, W. E. 1978. The biology of an equatorial herpe- 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 115 tofauna in Amazonian Ecuador. Miscellaneous Publications of the University of Kansas Museum of Natural History, 65: 1-352. Dumeril, A. 1856. Description des reptiles nouveaux ou im- parfaitement connus de la collection du Museum d’Histoire Naturelle et remarques sur la classification et les caracteres des reptiles. Archives du Museum d’Histoire Naturelle, Par- is, 8:56-588. Dumeril, A. M. C., and G. Bibron. 1837. Erpetologie generate ou histoire naturalle complete des reptiles. Paris, 4:1-572. Dundee, H. A., and D. A. Rossman. 1989. The Amphibians and Reptiles of Louisiana. Louisiana State University Press, Baton Rouge, xi + 300 pp. Essghaier, M. F., and D. R. Johnson. 1975. Aspects of the bioenergetics of Great Basin lizards. Journal of Herpetology, 9:191-195. Estes, R. 1 983. Sauria terrestria, Amphisbaenia. Handbuch der Palaoherpetologie, Teil 10 A. Gustav Fischer, Stuttgart, xxii + 249 pp. Etheridge, R. 1959. The relationships of the anoles (Reptilia: Sauria: Iguanidae): An interpretation based on skeletal mor- phology. Unpublished Ph.D. dissert.. University of Michi- gan, Ann Arbor, 236 pp. . 1960. Additional notes on the lizards of the Cragin Quarry fauna. Papers of the Michigan Academy of Science, Arts, and Letters, 45:113-117. . 1964. The skeletal morphology and systematic rela- tionships of sceloporine lizards. Copeia, 1964:610-631. . 1965. The abdominal skeleton of lizards in the family Iguanidae. Herpetologica, 21:161-168. . 1967. Lizard caudal vertebrae. Copeia, 1967:699-721. . 1970. A review of the South American iguanid genus Plica. Bulletin of the British Museum of Natural History (Zoology), 19:237-256. . 1992. A new psammophilus lizard of the genus Lio- laemus (Squamata: Tropiduridae) from northwestern Ar- gentina. Estratto dal Bollettinodel Museo Regionaledi Scienze Naturali— Torino, 10(1): 1-1 9. . 1994. Lizards of the Liolaemus darwinii complex (Squamata: Iguania: Tropiduridae) in northern Argentina. Estratto dal Bollettino del Museo Regionale di Scienze Na- turali—Torino, 1 1 ( 1 ): 1 4 1—204. . 1995. Redescription of Ctenoblepharys adspersaTschu- di 1845, and the taxonomy of Liolaeminae (Reptilia: Squa- mata: Tropiduridae). American Museum Novitates, 3142: 1-34. Etheridge, R., and K. de Queiroz. 1988. A phytogeny of Iguanidae. Pp. 283-367, in Phylogenetic Relationships of the Lizard Families: Essays Commemorating Charles L. Camp (R. Estes and G. Pregill, eds.), Stanford University Press, Stanford, California, xv + 631 pp. Etheridge, R., and E. E. Williams. 1985. Notes on Pristi- dactylus (Squamata: Iguanidae). Breviora, 483:1-18. . 1991. A review of the South American lizard genera Urostrophus and Anisolepis (Squamata: Iguania: Polychri- dae). Bulletin of the Museum of Comparative Zoology, 152: 317-361. Felsenstein, J. 1985. Confidence limits on phylogenies: An approach using the bootstrap. Evolution, 39:783-791. Ferguson, G. W. 1976. Color change and reproductive cycling in female collared lizards ( Crotaphytus collaris). Copeia, 1 976: 491-494. Fitch, H. S. 1956. An ecological study of the collared lizard. University of Kansas Publications of the Museum of Natural History, 8:213-274. . 1970. Reproductive cycles in lizards and snakes. Mis- cellaneous Publications of the University of Kansas Museum of Natural History, 52:1-247. . 1981. Sexual size differences in reptiles. Miscellaneous Publications of the University of Kansas Museum of Natural History, 70:1-72. Fitch, H. S., and W. W. Tanner. 1951. Remarks concerning the systematics of the collared lizard, ( Crotaphytus collaris), with a description of a new subspecies. Transactions of the Kansas Academy of Sciences, 54:548-559. Fox, S. F., and T. A. Baird. 1992. The dear enemy phenom- enon in the collared lizard, Crotaphytus collaris, with a cau- tionary note on experimental methodology. Animal Behav- iour, 44:780-782. Franklin, D. 1914. Notes on leopard lizards. Copeia, 1:1-2. Frierson, L. S., Jr. 1 927. Crotaphytus collaris collaris at Taylor Town, Louisiana. Copeia, 165:1 13-1 14. Fritts, T. H. 1974. A multivariate evolutionary analysis of the Andean iguanid lizards of the genus Stenocercus. Memoirs of the San Diego Society of Natural History, 7:1-89. Frost, D. R. 1992. Phylogenetic analysis and taxonomy of the Tropidurus group of lizards (Iguania: Tropiduridae). Amer- ican Museum Novitates, 3033:1-68. Frost, D. R., and R. Etheridge. 1 989. A phylogenetic analysis and taxonomy of iguanian lizards (Reptilia: Squamata). Uni- versity of Kansas Miscellaneous Publications, 81:1-65. Frost, D. R., and D. M. Hillis. 1990. Species in concept and practice: Herpetological applications. Herpetologica, 48:87- 104. Frost, D. R., A. G. Kluge, and D. M. Hillis. 1992. Species in contemporary herpetology: Comments on phylogenetic inference and taxonomy. Herpetological Review, 23:46-54. Garman, S. 1884. The North American reptiles and batrachi- ans. A list of the species occurring north of the Isthmus of Tehuantepec, with references. Bulletin of the Essex Institute, 16:1-46. Garrett, J. M., and D. G. Barker. 1987. A Field Guide to Reptiles and Amphibians of Texas. Texas Monthly Press, Austin, Texas, xi + 225 pp. Germano, D. J., and C. R. Carter. 1995. Natural history notes— Gambelia sila. Herpetological Review, 26:100. Germano, D. J., and D. F. Williams. 1 992. Life history notes— Gambelia sila. Herpetological Review, 23:1 17-1 18. . 1994. Natural history notes— Gambelia sila. Herpe- tological Review, 25:26-27. Gibbons, J. R. H. 1981. The biogeography of Brachvlophus (Iguanidae) including the description of a new species, B. vitiensis, from Fiji. Journal of Herpetology, 15:255-273. Greenberg, B. 1945. Notes on the social behavior of the col- lared lizard. Copeia, 1945:225-230. Grismer, L. L., J. A. McGuire, and B. D. Hollingsworth. 1994. A report on the herpetofauna of the Vizcaino Pen- insula, Baja California, Mexico, with a discussion of its bio- geographic and taxonomic implications. Bulletin of the Southern California Academy of Sciences, 93:45-80. Guyer, C., and J. M. Savage. 1986. Cladistic relationships among anoles (Sauria: Iguanidae). Systematic Zoology, 35: 509-531. 116 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 . 1992. Anole systematics revisited. Systematic Biology, 41:89-107. Hallowell, E. 1852. On a new genus and three new species inhabiting North America. Proceedings of the Academy of Natural Sciences of Philadelphia, 6:206-209. . 1854. Descriptions of new reptiles from California. Proceedings of the Academy of Natural Sciences of Phila- delphia, 7:91-97. Hammerson, G. A. 1986. Amphibians and Reptiles in Colo- rado. Colorado Division of Wildlife, Denver, vii + 131 pp. Hardy, J. D. 1958. Tail prehension and related behavior in a New World lizard. Herpetologica, 14:205-206. Harlan, R. 1835. Medical and Physical Researches; or Original Memoirs in Medicine, Surgery, Physiology, Geology, Zool- ogy, and Comparative Anatomy. Lydia R. Bailey, Philadel- phia, Pennsylvania, xxxiv + 653 pp. Hendy, M. D., and D. Penny. 1982. Branch and bound al- gorithms to determine minimal evolutionary trees. Mathe- matical Biosciences, 59:277-290. Hennig, W. 1966. Phylogenetic Systematics. University of Il- linois Press, Urbana, 263 pp. Hillenius, D. 1986. The relationship of Brookesia, Rhampho- leon and Chamaeleo (Chamaeleonidae, Reptilia). Bijdragen tot de Dierkunde, 56:29-38. . 1988. The skull of Chamaeleo nasutus adds more in- formation to the relationship of Chamaeleo with Rhampho- leon and Brookesia (Chamaeleonidae, Reptilia). Bijdragen tot de Dierkunde, 58:7-1 1. Hillis, D. M. 1991. Discriminating between phylogenetic signal and random noise in DNA sequences. Pp. 278-294, in Phy- logenetic Analysis of DNA Sequences (M. Miyamoto and J. Cracraft, eds.), Oxford University Press, New York, xvi + 588 pp. Hillis, D. M., and J. J. Bull. 1993. An empirical test of boot- strapping as a method for assessing confidence in phyloge- netic analysis. Systematic Biology, 42:182-192. Hillis, D. M., and J. P. Huelsenbeck. 1 992. Signal, noise, and reliability in molecular phylogenetic analyses. Journal of He- redity, 83: 198-195. Hipp, T. G. 1977. Reproductive cycle and correlated haemo- tological characteristics in Crotaphytus collaris in west cen- tral Texas. Unpublished M.S. thesis, Angelo State Univer- sity, San Angelo, Texas, 79 pp. Hoffstetter, R., and J. P. Gasc. 1969. Vertebrae and ribs of modem reptiles. Pp. 201-310, in Biology of the Reptilia, Volume 1, Morphology A (C. Gans, A. d’A. Bellairs, and T. S. Parsons, eds.), Academic Press, New York, New York, xv + 373 pp. Holbrook, J. E. 1842. North American Herpetology; or, a Description of the Reptiles Inhabiting the United States. J. Dobson, Philadelphia, Pennsylvania, 2:1-142. Holman, J. A. 1972. Herpetofauna of the Calf Creek local fauna (lower Oligocene: Cypress Hills Formation) of Saskatche- wan. Canadian Journal of Earth Sciences, 9:1612-1631. Huelsenbeck, J. P. 1991. Tree-length distribution skewness: An indicator of phylogenetic information. Systematic Zo- ology, 40:257-270. Ingram, W„ and W. W. Tanner. 1971. A taxonomic study of Crotaphytus collaris between the Rio Grande and Colo- rado rivers. Brigham Young University Science Bulletin, 8:1- 29. International Commission on Zoological Nomenclature. 1985. International Code of Zoological Nomenclature, Third Edition, Adopted by the XX General Assembly of the In- ternational Union of Biological Sciences. International Trust for Zoological Nomenclature, London, xx + 338 pp. James, E. 1823. Account of an Expedition from Pittsburgh to the Rocky Mountains, Performed in the Years 1819, ’20, by Order of the Hon. J. C. Calhoun, Sec’y of War; Under the Command of Major Stephen H. Long. H. C. Carey and I. Lea, Philadelphia, Pennsylvania, xcviii + 442 pp. Jenkins, F. A., and G. E. Goslow, Jr. 1983. The functional anatomy of the shoulder of the savannah monitor lizard (Varanus exant hematicus). Journal of Morphology, 175:195- 216. Jennings, M. R. 1987. Annotated check list of the amphibians and reptiles of California. Second, revised edition. South- western Herpetological Society Special Publication, 3:1^48. Johnson, T. R. 1987. The Amphibians and Reptiles of Mis- souri. Missouri Department of Conservation, Jefferson City, xi + 368 pp. Jollie, M. T. 1960. The head skeleton of the lizard. Acta Zool- ogica, 1960:1-64. Jones, T. A. 1993. Collared lizards (genus Crotaphytus). San Diego Herpetological Society Newsletter, 15:1-2. Jorgensen, C. D., and A. M. Orton. 1962. Note of lizards feeding on oatmeal bait. Herpetologica, 17:278. Jorgensen, C. D., A. M. Orton, and W. W. Tanner. 1963. Voice of the leopard lizard Crotaphytus wislizenii Baird and Girard. Proceedings of the Utah Academy of Science, Arts, and Letters, 40:1 15-116. Klaver, C. 1981. Lung-morphology in the Chamaeleonidae (Sauria) and its bearing upon phylogeny, systematics and zoogeography. Zeitschrift fur Zoologische Systematik Evo- lutionsforschung, 19:36-58. Klaver, C., and W. Bohme. 1986. Phylogeny and classification of the Chamaeleonidae (Sauria) with special reference to hemipenis morphology. Bonner Zoologische Monographien, 22:1-64. Klein, T. 1951. Notes on the feeding habits of Crotaphytus reticulatus. Herpetologica, 7:200. Kluge, A. G., and J. S. Farris. 1969. Quantitative phyletics and the evolution of anurans. Systematic Zoology, 18: 1-32. Knowlton, G. F., and W. L. Thomas. 1936. Food habits of Skull Valley lizards. Copeia, 1936:64-66. Lang, M. A. 1989. Phylogenetic and biogeographic patterns of basiliscine iguanians (Reptilia: Squamata: Iguanidae). Bon- ner Zoologische Monographien, 28:1-172. Lazell, J. D. 1969. The genus Phenacosaurus (Sauria: Igua- nidae). Breviora, 325:1-24. Legler, J. M., and H. S. Fitch. 1957. Observations on hiber- nation and nests of the collared lizard, Crotaphytus collaris. Copeia, 1957:305-307. Leviton, A. E. 1971. Reptiles and Amphibians of North Amer- ica. Doubleday and Company, Inc., New York, New York, 250 pp. Leviton, A. E., and B. H. Banta. 1964. Midwinter recon- naissance of the cape region of Baja California, Mexico. Pro- ceedings of the California Academy of Sciences, series 4, 30: 127-156. Leviton, A. E., R. H. Gibbs, Jr., E. Heal, and C. E. Dawson. 1985. Standards in herpetology and ichthyology: Part I. Standard symbolic codes for institutional resource collec- tions in herpetology and ichthyology. Copeia, 1 985:802-832. Lynch, J. D. 1982. Relationships of the frogs of the genus Ceratophrys (Leptodactylidae) and their bearing on hypoth- 1996 McGUIRE— SYSTEM ATICS OF CROTAPHYTID LIZARDS 117 eses of Pleistocene forest refugia in South America and punc- tuated equilibria. Systematic Zoology, 31:166-179. Lynn, W. G., and C. Grant. 1940. The herpetology of Ja- maica. Bulletin of the Institute of Jamaica, Science Series, 1:1-148. Mabee, P. M. 1989. An empirical rejection of the ontogenetic polarity criterion. Cladistics, 5:409-416. . 1993. Phylogenetic interpretation of ontogenetic change: Sorting out the actual and artefactual in an empirical case study of centrarchid fishes. Zoological Journal of the Linnean Society, 107:175-291. Mabee, P. M., and J. Humphries. 1993. Coding polymorphic data: Examples from allozymes and ontogeny. Systematic Biology, 42:166-181. Maddison, W. P., M. J. Donoghue, and D. R. Maddison. 1 984. Outgroup analysis and parsimony. Systematic Zoology, 33: 83-103. Mahrdt, C. R. 1973. Geographic distribution, Sauria, Cro- taphytus wislizenii copei (leopard lizard). Hiss News Journal, 1:98. Malaret, L. 1985. Geographic and temporal variation in the life history of Crotaphytus collaris (Sauria, Iguanidae) in Kansas and Mexico. Unpublished Ph.D. dissert., University of Kansas, Lawrence, 173 pp. Marx, B. H. 1950. Dentition morphology in some North American families of lizards. Unpublished M.S. thesis, Uni- versity of Illinois, Urbana, 109 pp. McAllister, C. T. 1980. Ecological observations of the eastern collared lizard, Crotaphytus collaris collaris (Say), in north- central Arkansas. Unpublished M.S. thesis, Arkansas State University, Jonesboro, 124 pp. . 1983. Aquatic behaviors of collared lizards, Crotaphy- tus c. collaris, from Arkansas. Herpetological Review, 14: 11. . 1985. Endoparasites of Crotaphytus collaris collaris (Sauria: Iguanidae) from Arkansas. Southwestern Naturalist, 30:363-370. McAllister, C. T., and S. E. Trauth. 1982. The instance of the eastern collared lizard, Crotaphytus collaris collaris (Sau- ria: Iguanidae) feeding on Sigmodon hispidus (Rodentia: Cri- cetidae). Southwestern Naturalist, 27:358-359. McCoy, C. J. 1967. Natural history notes on Crotaphytus wis- lizeni (Reptilia: Iguanidae) in Colorado. American Midland Naturalist, 77:138-146. McGuire, J. A. 1991. Geographic distribution— Crotaphytus insularis vestigium. Herpetological Review, 22:135. . 1994. A new species of collared lizard (Iguania: Cro- taphytidae) from northeastern Baja California, Mexico. Her- petologica, 50:438-450. Mead, J.. R. Thompson, and T. Van Devender. 1982. Late Wisconsinan and Holocene fauna from Smith Creek Can- yon, Snake Range, Nevada. Transactions of the San Diego Society of Natural History, 20:1-26. Meek, S. E. 1905. An annotated list of a collection of reptiles from southern California and northern Lower California. Field Columbian Museum, 104, 7:1-19. Meinzer, W. 1993. The Roadrunner. Texas Tech University Press, Lubbock, 104 pp. Miller, A. H., and R. C. Stebbins. 1964. The Lives of Desert Animals in Joshua Tree National Monument. University of California Press, Berkeley, vi + 452 pp. Mitchell, J. C. 1984. Observations on the ecology and repro- duction of the leopard lizard, Gambelia wislizenii (Iguani- dae), in southeastern Arizona. Southwestern Naturalist, 29: 509-511. Mittleman, M. B. 1942. A summary of the iguanid genus Urosaurus. Bulletin of the Museum of Comparative Zoology, 91:105-181. Mocquard, M. F. 1899. Contribution a la faune herpetologique de la Basse-Califomie. Nouvelles Archives du Museum d’Histoire Naturelle, Paris, serie 4, 1:297-343, pis. 11-13. . 1903. Notes herpetologiques. Bulletin du Museum d’Histoire Naturelle, Paris, 5:209-220. Moehn, L. D. 1976. The effects of sunlight on a despotism in the desert collared lizard, Crotaphytus insularis (Reptilia, Lacertilia, Iguanidae). Journal of Herpetology, 10:259-261. Montanucci, R. R. 1965. Observations on the San Joaquin leopard lizard, Crotaphytus wislizenii si/us Stejneger. Her- petologica, 21:270-283. . 1967. Further studies on leopard lizards, Crotaphytus wislizenii. Herpetologica, 23:1 19-126. . 1969. Remarks upon the Crotaphytus-Gambelia con- troversy (Sauria: Iguanidae). Herpetologica, 25:308-314. . 1970. Analysis of hybridization between Crotaphytus wislizenii and Crotaphytus silus (Sauria: Iguanidae) in Cal- ifornia. Copeia, 1970:104-123. . 1971. Ecological and distributional data on Crotaphytus reticulatus (Sauria: Iguanidae). Herpetologica, 27:183-197. . 1974. Convergence, polymorphism, or mtrogressive hybridization? An analysis of interaction between Crota- phytus collaris and Crotaphytus reticulatus (Sauria: Iguani- dae). Copeia, 1974:87-101. . 1976. Crotaphytus reticulatus Baird — Reticulate col- lared lizard. Catalogue of American Amphibians and Rep- tiles, 185:185.1-185.2. . 1978. Dorsal pattern polymorphism and adaptation in Gambelia wislizenii (Reptilia, Lacertilia, Iguanidae). Journal of Herpetology, 12:73-81. . 1983. Natural hybridization between two species of collared lizards ( Crotaphytus ). Copeia, 1983:1-1 1. . 1987. A phylogenetic study of the homed lizards, genus Phrynosoma, based on skeletal and external morphology. Los Angeles County Museum of Natural History Contri- butions in Science, 390:1-36. Montanucci, R. R., R. W. Axtell, and H. C. Dessauer. 1975. Evolutionary divergence among collared lizards ( Crotaphy- tusj, with comments on the status of Gambelia. Herpeto- logica, 31:336-347. Moody, S. M., 1980. Phylogenetic and historical biogeograph- ical relationships of the genera in the family Agamidae (Rep- tilia: Lacertilia). Unpublished Ph.D. dissert.. University of Michigan, Ann Arbor, 373 pp. . 1987. A preliminary cladistic study of the lizard genus Uromastyx (Agamidae, sensu lato), with a checklist and di- agnostic key to the species. Pp. 285-288, in Proceedings of the Fourth Ordinary General Meeting of the Societas Eu- ropaea Herpetologica (J. J. Van Gelder, H. Strijbosch, and P. J. M. Bergers, eds.), Faculty of Sciences, Nijmegen, The Netherlands, 473 pp. Mosley, K. T. 1963. Behavior patterns of the collared lizard 0 Crotaphytus collaris collaris). Unpublished M.S. thesis, University of Oklahoma, Norman, 40 pp. Murphy, R. W. 1983. Paleobiogeography and genetic differ- entiation of the Baja California herpetofauna. Occasional Papers of the California Academy of Sciences, 137:1-48. Norell, M.A. 1989. Late Cenozoic lizards oftheAnza Borrego 118 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Desert, California. Natural History Museum of Los Angeles County Contributions in Science, 414:1-31. Norell, M. A., and K. de Queiroz. 1991. The earliest iguanine lizard (Reptilia: Squamata) and its bearing on iguanine phy- logeny. American Museum Novitates, 2997:1-16. Nussbaum, R. A., E. D. Brodie, Jr., and R. M. Storm. 1983. Amphibians and Reptiles of the Pacific Northwest. The Uni- versity Press of Idaho, Moscow, 332 pp. Oelrich, T. M. 1956. The anatomy of the head of Ctenosaura pectinata (Iguanidae). Miscellaneous Publications of the Mu- seum of Zoology, University of Michigan, 94:1-122. Parcher, S. R. 1974. Observations on the natural histories of six Malagasy Chamaeleontidae. Zeitschrift fur Tierpsychol- ogie, 34:500-523. Parker, W. S. 1973. Notes on the reproduction of some lizards from Arizona, New Mexico, Texas, and Utah. Herpetolo- gica, 29:258-264. Parker, W. S., and E. R. Planka. 1976. Ecological observa- tions on the leopard lizard ( Crotaphytus wislizeni) in different parts of its range. Herpetologica, 32:95-1 14. Pickwell, G. 1972. Amphibians and Reptiles of the Pacific States. Dover Publications, New York, New York, xviii + 234 pp. Pietruszka, R. D., J. A. Wiens, and C. J. Pietruszka. 1981. Leopard lizard predation on Perognathus. Journal of Her- petology, 15:249-250. Pregill, G. K. 1992. Systematics of the West Indian lizard Leiocephalus( Squamata: Iguania: Tropiduridae). University of Kansas Miscellaneous Publications, 84:1-69. Presch, W. 1969. Evolutionary osteology and relationships of the homed lizard genus Phrynosoma (family Iguanidae). Co- peia, 1969:250-275. Rand, M. S. 1986. Histological, hormonal and chromatic cor- relates of sexual maturation in the male lizard, Crotaphytus collaris. Unpublished M.S. thesis, Wichita State University, Wichita, Kansas, 104 pp. Richardson, C. H. 1915. Reptiles of northwestern Nevada and adjacent territory. Proceedings of the U. S. National Mu- seum, 48(2078):403-435. Rieppel, O. 1987. The phylogenetic relationships within the Chamaeleonidae, with comments on some aspects of cla- distic analysis. Zoological Journal of the Linnean Society, 89:41-62. Robinson, P. L. 1975. The function of the hooked fifth meta- tarsal in Lepidosaunan reptiles. Colloque International C.N.R.S., 218:461-483. Robison, W. G., and W. W. Tanner. 1962. A comparative study of the species of the genus Crotaphytus Holbrook (Igua- nidae). Brigham Young University Science Bulletin, Biolog- ical Series, 21:1-31, pis. 1-12. Rostker, M. 1983. An experimental study of collared lizards: Effects of habitat and male quality on fitness. Unpublished Ph.D. dissert., Oklahoma State University, Stillwater. Ruibal, R. 1964. An annotated checklist and key to the anoline lizards of Cuba. Bulletin of the Museum of Comparative Zoology, 130:475-520. Ruibal, R., and E. E. Williams. 1961. The taxonomy of the Anolis homolechis complex of Cuba. Bulletin of the Museum of Comparative Zoology, 125:21 1-246. Ruthven, A. G. 1 907. A collection of reptiles and amphibians from southern New Mexico and Arizona. Bulletin of the American Museum of Natural History, 23:483-603. Sanborn, S. R., and R. B. Loomis. 1979. Systematics and behavior of collared lizards ( Crotaphytus , Iguanidae) in southern California. Herpetologica, 35:101-106. Sanderson, M. J., and M. J. Donoghue. 1989. Patterns of variation in levels ofhomoplasy. Evolution, 43:1781-1795. Savage, J. M. 1958. The iguanid lizard genera Urosaurus and Uta, with remarks on related groups. Zoologica, 43:41-54. . 1960. Evolution of a peninsular herpetofauna. System- atic Zoology, 9:184-212. Schmidt, K. P. 1922. The amphibians and reptiles of Lower California and the neighboring islands. Bulletin of the Amer- ican Museum of Natural History, 46:607-707. Schwartz, A., and R. W. Henderson. 1991. Amphibians and Reptiles of the West Indies: Descriptions, Distributions, and Natural History. University of Florida Press, Gainesville, xvi + 720 pp. Sexton, O. J., R. M. Andrews, and J. E. Bramble. 1992. Size and growth rate characteristics of a peripheral population of Crotaphytus collaris (Sauria: Crotaphytidae). Copeia, 1992: 968-980. Smith, H. M. 1939. The Mexican and Central American lizards of the genus Sceloporus. Zoological Series of the Field Mu- seum of Natural History, 26:1-397. . 1946. Handbook of Lizards. Lizards of the United States and of Canada. Comstock Publishing Company, Ith- aca, New York, xi + 557 pp. Smith, H. M., and E. D. Brodie, Jr. 1982. A Guide to Field Identification — Reptiles of North America. Golden Press, New York, New York, 240 pp. Smith, H. M., and E. H. Taylor. 1950. An annotated checklist and key to the reptiles of Mexico exclusive of the snakes. Bulletin of the U. S. National Museum, 199:1-253. Smith, N. M. 1974. Observation of voice in the western collared lizard Crotaphytus collaris bicinctores. Great Basin Natural- ist, 34:276. Smith, N. M., and W. W. Tanner. 1972. Two new subspecies of Crotaphytus (Sauria: Iguanidae). Great Basin Naturalist, 32:25-34. . 1974. A taxonomic study of the western collared lizards, Crotaphytus collaris and Crotaphytus insularis. Brigham Young University Science Bulletin, Biological Series, 19:1 — 29. Snyder, J. D. 1972. An ecological investigation of sympatric populations of the lizards Crotaphytus collaris and C. wis- lizenii. Unpublished M.S. thesis, San Francisco State Col- lege, San Francisco, California, 88 pp. Snyder, R. C. 1 949. Bipedal locomotion of the lizard Basiliscus basiliscus. Copeia, 1949:129-137. . 1952. Quadrupedal and bipedal locomotion of lizards. Copeia, 1952:64-70. . 1954. The anatomy and function of the pelvic girdle and hindlimb in lizard locomotion. American Journal of Anatomy, 95:1-46. . 1962. Adaptations for bipedal locomotion of lizards. American Zoologist, 2:191-203. Soule, M., and A. J. Sloan. 1966. Biogeography and distri- bution of the reptiles and amphibians on islands in the Gulf of California, Mexico. Transactions of the San Diego Society of Natural History, 14:137-156. Sprackland, R. G. 1990. Collared lizards. Tropical Fish Hob- byist, 39:104-19. . 1993. Husbandry and breeding of collared lizards. Vi- varium, 4:23-26. Stamps, J. A. 1977. Social behavior and spacing patterns in 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 119 lizards. Pp. 256-334, in The Biology of the Reptilia. Volume 7 (C. Gans and D. W. Tinkle, eds.), Academic Press, New York, New York, xvi + 720 pp. Stansbury, H. 1852. Exploration and Survey of the Valley of the Great Salt Lake of Utah, Including a Reconnaissance of a New Route Through the Rocky Mountains. Lippincott, Grambo and Company, Philadelphia, Pennsylvania, xiii + 421 pp. Stebbins, R. C. 1954. Amphibians and Reptiles of Western North America. McGraw-Hill Book Company, New York, New York, xiv + 536 pp. . 1985. A Field Guide to Western Reptiles and Am- phibians. Houghton Mifflin Company, Boston, Massachu- setts, xvi + 336 pp. Steinhart, P. 1990. California’s Wild Heritage— Threatened and Endangered Animals in the Golden State. Craftsman Press, Seattle, Washington, iv + 108 pp. Stejneger, L. 1890. Annotated list of reptiles and batrachians collected by Dr. C. Hart Merriam and Vernon Bailey on the San Francisco Mountain plateau and desert of the Little Colorado, Arizona, with descriptions of new species. North American Fauna, 3:103-118. . 1893. Annotated list of the reptiles and batrachians collected by the Death Valley expedition in 1891, with de- scriptions of new species. North American Fauna, 7:159- 228. Stejneger, L., and T. Barbour. 1917. A Check List of North American Amphibians and Reptiles. Harvard University Press, Cambridge, Massachusetts, iv + 125 pp. Stone, W., and J. A. G. Rehn. 1903. On the terrestrial ver- tebrates of portions of southern New Mexico and western Texas. Proceedings of the Academy of Natural Sciences, Philadelphia, 55:16-34. Strecker, J. K., Jr. 1909. Reptiles and amphibians collected in Brewster County, Texas. Baylor University Bulletin, 12: 11-15. Swofford, D. L. 1995. Phylogenetic Analysis Using Parsi- mony, Version 4.0.0d26. Sinauer Associates, Inc., Sunder- land, Massachusetts. Tanner, W. W. 1978. Zoogeography ofreptiles and amphibians of the intermountain region. Great Basin Naturalist Mem- oirs, 2:43-55. Tanner, W. W., and B. H. Banta. 1963. The systematics of Crotaphytus wislizenii, the leopard lizards. Part I. A rede- scription of Crotaphytus wislizenii wislizenii Baird and Gi- rard, and a description of a new subspecies from the upper Colorado River basin. Great Basin Naturalist, 23:129-148. . 1977. The systematics of Crotaphytus wislizenii, the leopard lizards. Part III. The leopard lizards of the Great Basin and adjoining areas, with a description of a new sub- species from the Lahonton Basin. Great Basin Naturalist, 37:225-240. Tanner, W. W., and C. D. Jorgensen. 1963. Reptiles of the Nevada Test Site. Brigham Young University Science Bul- letin, Biological Series, 3:1-31. Tanner, W. W., and J. E. Krogh. 1974a. Ecology of the leop- ard lizard, Crotaphytus wislizeni at the Nevada Test Site, Nye County, Nevada. Herpetologica, 30:63-72. . 19746. Variations in activity as seen in four sympatric lizard species in southern Nevada. Herpetologica, 30:303- 308. Taylor, W. P. 1912. Field notes on amphibians, reptiles and birds of northern Humboldt County, Nevada, with a dis- cussion of some of the faunal features of the region. Uni- versity of California Publications in Zoology, 7:319—436. Tevis, L., Jr. 1944. Herpetological notes from Lower Califor- nia. Copeia, 1944:6-18. Tinkle, D. W. 1959. Observations on the lizards Cnemidoph- orus tigris, Cnemidophorus tessellatus and Crotaphytus wis- lizenii. Southwestern Naturalist, 4:195-200. Tollestrup, K. 1979. The ecology, social structure, and for- aging behavior of two closely related species of leopard liz- ards, Gambelia silus and Gambelia wislizenii. Unpublished Ph.D. dissert., University of California, Berkeley, 146 pp. . 1982. Growth and reproduction in two closely related species of leopard lizards, Gambelia silus and Gambelia wis- lizenii. American Midland Naturalist, 108:1-20. . 1983. The social behavior of two species of closely related leopard lizards, Gambelia silus and Gambelia wis- lizenii. Zeitschrift fur Tierpsychologie, 62:307-320. Trauth,S. E. 1974. Demography and reproduction of the east- ern collared lizard, Crotaphytus collaris collaris (Say), from northern Arkansas. Unpublished M.S. thesis. University of Arkansas, Fayetteville, 109 pp. . 1978. Ovarian cycle of Crotaphytus collaris (Reptilia, Lacertilia, Iguanidae) from Arkansas with emphasis on cor- pora albicantia, follicular atresia, and reproductive potential. Journal of Herpetology, 12:461-470. . 1979. Testicular cycle and timing of reproduction in the collared lizard ( Crotaphytus collaris) in Arkansas. Her- petologica, 35:184-192. Turner, F. B., J. R. Lannom, P. A. Medica, and G. A. Hoddenbach. 1969. Density and composition of fenced populations of leopard lizards ( Crotaphytus wislizenii) in southern Nevada. Herpetologica, 25:247-257. Underwood, G. 1970. The eye. Pp. 1-97, in Biology of the Reptilia, Volume 2, Morphology B (C. Gans and T. S. Par- sons, eds.). Academic Press, New York, New York, xiii + 374 pp. Uzee, E. M. 1990. The effects of thermal constraints on the daily activity of Crotaphytus collaris. Unpublished M.S. the- sis, Oklahoma State University, Stillwater. Van Denburgh, J. 1922. Reptiles of western North America. Volume 1. Lizards. Occasional Papers of the California Academy of Sciences, 1 0: 1-6 1 1 . Van Denburgh, J., and J. R. Slevin. 1921. Preliminary' di- agnoses of new species of reptiles from islands in the Gulf of California, Mexico. Proceedings of the California Acad- emy of Sciences, 1 1:95-98. Van Devender, R. W. 1982. Comparative demography of the lizard Basiliscus basiliscus. Herpetologica, 38:189-208. Van Devender, T. R., and J. I. Mead. 1978. Early Holocene and late Pleistocene amphibians and reptiles in Sonoran Des- ert packrat middens. Copeia, 1978:464-475. Van Devender, T. R., A. M. Phillips, and J. I. Mead. 1977. Late Pleistocene reptiles and small mammals from the lower Grand Canyon of Arizona. Southwestern Naturalist, 22:49- 66. Van Devender, T. R., A. M. Rea, and W. E. Hall. 1991. Faunal analysis of late Quaternary vertebrates from Organ Pipe Cactus National Monument, southwestern Arizona. Southwestern Naturalist, 36:94-106. Watrous, L. E., and Q. D. Wheeler. 1981. The outgroup comparison method of character analysis. Systematic Zo- ology, 30:1-1 1. 120 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Webb. R. G. 1970. Reptiles of Oklahoma. University Okla- homa Press, Norman, xi + 370 pp. Weiner, N. J., and H. M. Smith. 1965. Comparative osteology and classification of the crotaphytiform lizards. American Midland Naturalist, 73:170-187. Welsh, H. H., Jr. 1988. An ecogeographic analysis of the her- petofauna of the Sierra San Pedro Martir, Baja California, with a contribution to the biogeography of the Baja Cali- fornia herpetofauna. Proceedings of the California Academy of Sciences, 46:1-72. Welsh, H. H., Jr., and R. B. Bury. 1984. Additions to the herpetofauna of the south Colorado Desert, Baja California, with comments on the relationships of Lichanura trivirgata. Herpetological Review, 15:53-56. Werner, D. I. 1978. On the biology of Tropidurus delanonis, Baur (Iguanidae). Zeitschrift fur Tierpsychologie, 47:337- 395. Wever, E. G., M. C. Hepp-Reymond, and J. A. Vernon. 1966. Vocalization and hearing in the leopard lizard. Proceedings of the National Academy of Sciences, Philadelphia, 55:98- 106. Wiens, J. J. 1993a. Phylogenetic relationships of phrynoso- matid lizards and monophyly of the Sceloporus group. Co- peia, 1993:287-299. . 1 993 b. Phylogenetic systematics of the tree lizards (ge- nus Urosaurus). Herpetologica, 49:399-420. . 1995. Polymorphic characters in phylogenetic system- atics. Systematic Biology, 44:482-500. Wiley, E. O. 1981. Phylogenetics. The Theory and Practice of Phylogenetic Systematics. John Wiley and Sons, New York, New York, xv + 439 pp. Wishart, D. 1968. A Fortran II Programme (CLUSTAN) for Numerical Classification. Computing Laboratory, St. An- drews, Fife, Scotland, 50 pp. Wyles, J. S. 1980. Phylogenetic studies of iguanid lizards (Lac- ertilia, Iguanidae). Unpublished Ph.D. dissert.. University of California, Los Angeles, 174 pp. Yarrow, H. C. 1882a. Checklist of North American Reptilia and Batrachia, with catalogue of specimens in the U. S. National Museum. Bulletin of the U. S. National Museum, 24:1-249. . 18827). Descriptions of new species of reptiles and am- phibians in the United States National Museum. Proceedings of the U. S. National Museum, 6:438-443. Yedlin, I. N., and G. W. Ferguson. 1973. Variations in ag- gressiveness of free-living male and female collared lizards, Crotaphytus collaris. Herpetologica, 29:268-275. Appendix 1 Specimens Examined Museum acronyms follow Leviton et al. (1985) except for the following nonstandard abbreviations: BDH (collection of Bradford D. Hollingsworth), CES (Centro Ecologio de Sonora, Hermosillo, Mexico), EL (collection of Ernest Liner), JAM (collection of Jimmy A. McGuire), JMS (collection of Jay M. Sav- age), LLG (collection of L. Lee Grismer), MZFC (Museo de Zoologia “Alfonso L. Herrera,” Univ- ersidad Nacional Autonoma de Mexico), REE (col- lection of Richard Etheridge), and UABC (Univ- ersidad Nacional Autonoma de Baja California, En- senada, Mexico). The following abbreviations de- note the form of preparation for each specimen: D (complete dry skeleton), S (skull only), P (preserved specimen), ARAB (alizarin red, alcian blue stained specimen), and H (hemipenis prepared by wax in- jection). Locality data are presented for all ingroup taxa examined but not for outgroup taxa. Crotaphytidae C. antiquus. — MEXICO: COAHUILA: CM 140199-140200; TNHC 53152-53154, 53157, 53160-53161 (P), 53155-53156, 53158-53159 (D), MZFC 6750-6756 (P)-Sierra de San Lor- enzo, approximately 0.25 mi. W of the pueblo of Santa Eulalia. Crotaphytus bicinctores. —UNITED STATES: ARIZONA: Coconino County: SDSNH 19479 (P)- Williams, 33053 (P)-5 mi. W Kane Ranch, 358 1 2 (P)-Coconino, Lee’s Ferry. Mojave County: USNM 1 15677 (D)-Rampart Cave. Maricopa County: REE 292 1 (D)— 6.3 mi. N Sentinel, 2922 (D)-2.3 mi. N Sentinel, 2923 (D), SDSNH 68624 (P)-7.0 mi. N Sentinel, 2924 (D) — 3.2 mi. N Sentinel; SDSNH 68623 (P)— 5.8 mi. N Sentinel, 68637- 39 (P) — Extreme E slope Gila Bend Mtns. on W shore Gila River at jet. Old U.S. 80 and Gila River (W side Gillespie Bridge). Yuma County: LLG 1397-99 (P) — Trigo Mtns., 12 mi. W Palo Verde. REE 2931 (D)— Nr. Yuma Proving Grounds; SDSNH 16731 (P) — Castle Dome, 17602 (P)— Kofa Mtns., Wilbank Ranch, 22351 (P)— Sentinel, 26911 (P)— Plamosa Mtns., 33301 (P) — Dublin, 68625 (P)— Dome Valley Solid Waste Transfer Site, Co. 7th St., 1.0 mi. E Ave. 20E, NW of Wellton, 68626, SDSU 1723 (P)-S slope Laguna Mtns., 0.3 mi. NW Hwy 95 on rd. to N.R. Adair Park (and shooting range). CALIFORNIA: Imperial County: REE 2928-30 (D), SDSU 1 721-22 (P)-S end Chocolate Mtns., jet. Ogilby Rd. and Hwy 86, REE 2925-27 (D)— jet. Palo Verde Mtns. and Hwy 78, 2933 (D)— Black Mtn., 2.8 mi. SSE Hwy 78 on Black Mountain Rd., 2934 (D), SDSNH 68627-28 (P, H), 68629-36 (P) — Chocolate Mtns., Black Mountain, Black Mountain Rd. Inyo County: AMNH 108970-71 (D) — 9 mi. NE Big Pine, ca 6000’; SDSNH 15878-79 (P)— Death Valley, Fur- nace Creek, 15880-81 (P)— Death Valley, Stovepipe Wells, 15988 (P) — Argus Mtns., 3 mi. E Junction Ranch, 19475-77, 22218- 19 (P)— Ballarat, 22220 (P) — Emigrant Pass, 34113 (P)— 8 mi. SE Keeler, 341 14 (P) — 8 mi. W Panamint Spring, 341 15 (P)— Wildrose Station, 34305 (P) — Independence, Mazurka Canyon, 38255-56 (P)— Panamint Mtns., Wildrose Station. Kern County: JMS 832 (S)— Twin Buttes nr. Mojave; REE 1570 (D) — 2 mi. S Castle Butte. Riverside County: SDSNH 39751 (P) — 3 mi. E Shaver’s Summit, 39752 (P)— foot of Fanhill Canyon, 40139 (P)_4 mi. NE Whitewater. San Bernardino County: REE 2932 (D)-Ord Mtns; SDSNH 2459-61 (P)— Victorville, 4407 (P)- US Rt. 1 1 nr. California-Nevada line, 5874 (P)— Kramer Hills, 1 1087 (P) — N Fort Lytle Creek, 29091 (P)— Mountain Pass, 29229 (P)_7 mj. NE Cronise, 29664 (P)— Providence, Mtns., Cedar Canyon, 38703 (P) — 20-50 mi. W Needles, 39874 (P)— Pipe Can- 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 121 yon. IDAHO: Owyhee County: SDSNH 1 444 — 45 (P) — ID, Owy- hee Co., Hot Springs S Bruneau. NEVADA: Clark County: JMS 203 (S) — Nr. Las Vegas. Lincoln County: USNM 18324 (D) — 25 mi. E Panaca, nr. Utah line. Storey County: LLG 1400-01 (P)— Carson City. Washoe County: JMS 197 (S) — White Mt., Truckee Meadows, 198 (S)— E side Truckee Meadows. UTAH: Millard County: SDSNH 661-62 (P)-20 mi. NW Delta, 26704 (P) — 10 mi. S Kanosh. Washington County: SDSNH 24982-83 (P)— Zion Nat. Park, 24984 (P)-St. George, 25506-07 (P)-7 mi. NW Santa Clara, 25508-09 (P) — Beaver Dam Mtns., 25644- 46 (P)— Toquerville, 25647 (P)- Leeds, 25648-52 (P)-Rock- ville; UIMNH 27723 (S)- Springdale, 93994-95 (S)-Nr. St. George (nr. A Z border). No data: SDSNH 12244-45 (P). C collaris. — MEXICO: CHIHUAHUA: CM 59531 (P)-40 mi. E Julimes, 59532 (P)— 1 8 mi. NE Aldama; KU 49628 (D)— Vado de Fusiles, 157873 (D), SDSNH 49753 (P)— 22 km S Es- tacion Moctezuma, KU 157874 (D), SDSNH 49755 (P)-30.6 mi. S jet Mex. Hwys 45 and 49; REE 1213 (D)— 20 mi. SW Chihuahua, 1214 (D)-6 mi. W Camargo; SDSNH 47932 (P)— Moctezuma, 47937^12 (P)— 1 1-20 mi. N Chihuahua, 49754 (P)— 1 1.6 mi. N jet. Mex. Hwys 45 and 49, 49756 (P)— 12.7 mi. N jet. Mex. Hwys 45 and 49; UIMNH 48295 (S)— 27.7 mi. SCiudad Delicias. COAHUILA: CM 42936 (P)— 8.6 mi. SW Cuatroci- enegas de Carranza, N side San Marcos Mtn., 42938 (P)— 4 mi. N Cuatrocienegas de Carranza, Rio Canon, 42939 (P) — 5 mi. N Cuatrocienegas de Carranza, W slope Sierra del Muerto, 42940 (P) — 5 mi. N Cuatrocienegas de Carranza, N end Rio Canon, 42941 (P) — E edge Cuatrocienegas Basin, along Hwy. 13.5 mi. E Cuatrocienegas de Carranza, 42942 (P)— Rancho San Fernando, 80 km SW Cuatrocienegas de Carranza; KU 147299 (D)— Mo- tacorona; REE 2944 (D)— 1 7.3 mi. E Viesca; SDSNH 49744 (P)— 15.4 mi. S and 1.3 mi. W Sabinas, 49745 (P) — 15.4 mi. S and 0.6 mi W Sabinas, 49746 (P) — 0.9 mi. SE Motacorona, 49747 (P) — 7.4 mi. N Ahuichila, 49748 (P) — 5.8 mi. N Ahuichila, 49749 (P) — 21.5 mi. SW Viesca, 49751 (P)— 1 5.4 mi. S Sabinas, 6 mi. W, 49752 (P)— 15.4 mi. S Sabinas, 3 mi. W, 56752 (P)— 15.4 mi. S and 4.8 mi. W Sabinas; UIMNH 43224-25 (S)— 15.6 mi. E Cuatro Cienagas; SDSU 2061 (P, H)— 4.3 mi. N Bahia de Ahuichila, 2062 (P) — 22.6 mi. S Viesca, 2063 (P)— 4.8 mi. N Bahia de Ahuichila, 2064 (P) — 1.3 mi. N Bahia de Ahuichila, 2065 (P)— 15.1 mi. E Viesca. NUEVO LEON: CM 42943 (P)- 2.7 mi. S Villa Garcia; SDSNH 56750 (P)— 27.9 mi. N Mina. ZACATECAS: SDSNH 5675 1 (P>— 1 . 1 mi. W Tecolotes. UNIT- ED STATES: ARIZONA: Cochise County: AMNH 735 1 8, 74752, 75657 (D) — Portal. Coconino County: AMNH 82297 (D)— 11 mi. NNW Cedar Ridge; SDSNH 2087 (P)— 1 6 mi. N Flagstaff, 9010, 29231 (P)— Canyon Diablo, 25503-05, 25639-40, 29131 (P) — Meteor Crater, 29645 (P)— Two Guns, 32529, 32658, 34466 (P), JMS 200 (S), UIMNH 34337 (S)- Wupatki Nat. Monument, JMS 202 (S)— The Citadel, Wupatki Nat. Monument; SDSNH 40958-59 (P)-1000 yds from Meteor Crater; UIMNH 27727 (D)— Wupatki Nat. Monument, nr. Citadel (4 mi. from Hwy 89). Gila County: SDSNH 27751 (P)— Sierra Ancha Mtns. Pima County: USNM 220214 (D)— Continental. Yavapai County: AMNH 84489, 85381, 85625 (D) — vicinity of Stanton. AR- KANSAS: Brown County: SDSNH 40963-66 (P)-7 mi. N Har- rison. County undetermined: USNM 220216 (D) — Red River. COLORADO: Delta County: CM 39257-39258 (P)— 14 mi. NW Delta, Escalante Canyon. Fremont County: SDSNH 62106-13 (P) — Wet Mountain Project. Garfield County: JAM 3 1 5 (ARAB), REE 2871, 2874, 2879, (D) SDSU 1735, 2108 (P)— 17.9 mi. N Hwy 70 via Hwy 139, REE 2878 (D)— 18.5 mi. N jet. Hwys 139 and 70 via Hwy 139. Mesa County: CM 42932 (P)— Stovepipe Canyon, 2 mi. W, 1 7 mi. N Fruita, 42933 (P) — Colorado National Monument, mouth E Monument Canyon, 44747 (P)— Colorado National Monument. Montezuma County: CM 67094-67097 (P)— Bridge Canyon. San Miguel County: CM 42931 (P) — Dis- appointment Gap Spring (= Gypsum Gap). KANSAS: Cowley County: SDSNH 10982 (P)-7 mi. NE Winfield, 2 1 859-63 (P)- 2 mi. NE Winfield. Douglas County (?): REE 1797, 1823-24 (D)— nr. Lawrence. County undetermined: REE 1836, 1857 (D)— Kansas. NEW MEXICO: Colfax County: JMS 189 (S)— 1 .5 mi. N Chico post office, 7200'. Dona Ana County: REE 2945 (D) — Organ Mtns., 1 .7 mi. S Hwy 82/70 on Baylor Canyon Dr., 2946- 48 (D)— Organ Mtns., 5.1 mi. S Hwy 82/70 on Baylor Canyon Dr., 2949 (D)— Organ Mtns., 5.0 mi. S Hwy 82/70 on Baylor Canyon Dr.; SDSU 2059 (P)-Organ Mtns., 4.1 mi. S Hwy 82/ 70 on Baylor Canyon Dr. Eddy County: SDSU 2067 (P) — Carls- bad. Graham County: JMS 20 1 (S) — 9. 1 mi. ENE San Jose along new paved road to Clifton (1951). Hidalgo County: CM 75544- 75551 (P)— ST 9, 3-5 mi. W Animas. Rio Arriba County: SDSNH 9007-08 (P)— Dixon, 57854 (P) — El Cobre Canyon. San Juan County: AMNH 108314(D), SDSNH 20044 (P)-Chaco Canyon Nat. Monument. San Miguel County: JMS 190 (S) — 10 mi. E Sanchez, 192 (S) — 3.9 mi. NE Trementina. Torrance County: JMS 191 (S) — Manzano. OKLAHOMA: Cherokee County: SDSNH 52752-57 (P)— Tenkiller Ferry Reservoir. Jackson County: JAM 556 (P), REE 2951-52 (D)-Altus. TEXAS: Brew- ster County: SDSU 2058, 2066 (P) — 2.3 mi. W Study Butte via Hwy 170. USNM 217271 (D)— specific locality unknown. Palo Pinto County: JMS 40, 195-96 (S)— Palo Pinto. Pecos County: JMS 194 (S)— Ft. Stockton. Presidio County: SDSU 2060 (P)— the River Road at the Teepees (W of Study butte). Reeves County (?): REE 2950 (D)— Pecos region. Shackelford County: Fort Grif- fin. Travis County: SDSU 2068-7 1 (P)— Milton Reimer’s Fishing Ranch, 0.9 mi. from FM 3238 off Hwy 7 1 . UTAH: Grand Coun- ty: REE 2869, 2877 (D), SDSU 1734 (P, H)-33 mi. N jet Hwy 191 (NE Moab) via Hwy 128, REE 2870, 2875-76 (D), SDSU 2105-07 (P)— 32.2 mi. Njct Hwy 191 (NE Moab) via Hwy 128; SDSU 2109 (P) — 29 mi. NE Moab on Utah Hwy 128. C. dickersonae. — MEXICO: SONORA: AMNH 78949 (PI- SE side Tiburon Island between Monument Pt. and red Bluff; BYU 2425, 39995 (P)— 23 mi. N Kino Bay near coast, 2426, 2433, 3164, 3166-69, 3172 (P) — Punta Perla, NE end Tiburon Island; CAS 14008-12 (P)-Isla Tiburon, SE end. 53265 (P), JMS 208 (S)— Tiburon Island; REE 2774-77, 2787-88, 2904-05 (D), SDSU 1720, 23 1 8 (P), 2319 (P, H)— 1 .2 mi. N Bahia Kino Nuevo via rd. to Punta Chueca, REE 2777-86 (D), SDSU 1718 (P)— Isla Tiburon, El Corralito, Appx. 3 km N of S end of island; SDSNH 47936 (P)— 10 mi. N Bahia Kino; SDSU 1719 (P) — Isla Tiburon, Appx. 5 km N El Corralito (S end of island); UAZ 704- OS, 30226 (P)— 6.5 mi. by rd. NW Desemboque, 9625-26 (P)— Isla Tiburon, Ensenada Blanca, 16578 (P)— Bahia Kino, Mtn. NW Caverna Seri, 20144 (P)— Punta Cirio, 7.0 mi. by rd. S Puerto Libertad, 42569 (P) — Punta Cirio, Sierra Bacha, SE Lib- ertad; USNM 238243-46 (P) — 11.5 mi. N Punta Chueca, 238247- 48 (P), 238249 (P, H)— 4 mi. N Bahia Kino, 248142-43 (P)— Isla Tiburon, S end, 248 1 74-80 (P) — 3 mi. N Bahia Kino (Nuevo). C. grismeri. —MEXICO: BAJA CALIFORNIA: Sierra de Los Cucapas: CES 067-624 (P), 067-627-29 (P), 067-25 (P, H); MZFC 6647-51 (D); UABC 1 15-19 (P) — Canon David, appx. 2 km W Mex. Hwy 2 on the rd. to the Sulfur mine (turnoff at km 49 S Mexicali); USNM 37625 (P) — Volcano Lake. C. insularis. — MEXICO: BAJA CALIFORNIA: CAS 14002 (P) — Isla Angel de La Guarda, SE end, 2 1 948^49 (P)— Isla Angel de La Guarda, nr. small bay opposite Bay of Los Angeles (appx. 122 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 29°6 N, 113°12'W), 22712 (P) — Isla Angel de La Guarda: nr. small bay on SW shore, opp. Bay of Los Angeles (appx. 29°6'N, 1 13°12'W), 50873-79, 86754-55, 86783-84, 148650—52; REE 2793-97 (D), SDSU 1732 (P. H), 1733 (P)— Isla Angel de La Guarda, N end; SDSNH 19971, 19773-75, 50664, 53064 (P)- Isla Angel de La Guarda. C. nebrius. — MEXICO: SONORA; AMNH 73715 (S)-Guay- mas, 73758 (P) — 1 6 mi. (via road) S of Nogales, 75682-83 (P)— Punta San Carlos, 7 mi. N Guaymas, 80852 (P)— 2.3 mi. (road) NEGuaymas; BYU 40930-31 (P)— 8 mi. N Guaymas; CAS 12774 (P) — 31 mi. SW Moctezuma; KU 1 52639-42 (P)- 139.4 km NW Caborca on Mex. Rt. 2, 176402 (P)— 6.4 km S Guaymas; JMS 205 (S)— 10 mi. S Hermosillo; LACM 8798 (P)— 59.5 mi. E San Luis, 8799 (P)-84.1 mi. E San Luis; MVZ 10163-65 (P)-Sierra Alamo, 30 mi. W Caborca, 136687-88 (P) — 1 .9 rd. mi. N Ba- cadehuachi, 136689-90 (P)— ca 4.1 rd. mi. NW Nacori Chico; REE 370-71 (D)— nr. Pitiquito, 404-07 (D), 40 mi. W Sonoita, 2771-73 (D) — appx. 5 mi. N San Carlos Nuevo; SDSNH 49008 (P)— N bay at Guaymas; SDSU 2072 (P) — 66.6 mi. W Sonoita, 2073 (P)— 3.5 mi ENE Huasabas, 2074 (P)— 5.2 mi. ENE Huas- abas. UNITED STATES: ARIZONA: Maricopa County: SDSNH 68657-58 (P) — Buckeye Hills Recreation Area, above picnic area offBuckeye Hills Dr., 68659-61 (P)— Extreme W slope of Buck- eye Hills on E side Gila River at jet. Old U.S. 80 and Gila River (E side Gillespie Bridge). Pima County: CAS 81420 (P)— 20 mi. S Ajo, Alamo Canyon, Organ Pipe Nat. Monument; KU 121460 (P) — Gates Pass, Tucson Mtns. W Tucson; MVZ 76641 (P), UIMNH 5898 (S)— Ajo Mtns., Alamo Canyon, Organ Pipe Cac- tus Nat. Monument; REE 2937-38 (D)— SE edge Tucson Mtns., nr. end Sarasota Dr., 2939-41 (D)— Little Ajo Mtns., 2.9 mi. W Hwy 85 on entrance rd. to Ajo Air Force Station; SDSNH 68640- 41 (P)— Ajo Mtns., ca 1.5 mi. S of Why, 68642-44 (P) — Quijotoa, 68645 (P), 68646 (P, H)-0.9 mi. S Why, 68647 (P) — 4. 1 mi. N Hwy 86 on rd. to Hickiwan, 68648—49 (P)— Silverbell Mtns., 20.4 mi. (by rd.) W Tucson Mtns. by way of Avra Valley Rd. Pinal County (?): KU 14860 (P)-20 mi. SW Phoenix. Yuma County: REE 2925 (D)—W face Gila Mtns., on Hwy 8, 4 mi. E Foothills Dr.; SDSNH 68650, 68652-54, SDSU 1724-25 (P)— W slope of Gila Mtns., ca 2 mi. N Hwy 8, SDSNH 68651 (P)— N slope Gila Mtns., 0.25 mi. S of RR tracks on S side Hwy 95 (at mile marker 39), 68655-56 (P)— Mohawk Mountains, N side Hwy 8. C. reticulatus. — MEXICO: COAHUILA: SDSNH 56753 (P)- 11.1 mi. S Villa Union. NUEVO LEON: EL 4138 (P)— 5.9 mi. SSW 0. 9-2.0 mi. NW Cerralvo along Rancho los Robles rd. to Picacho Mtns, 4816 (P)— 6.2 mi. SW Cerralvo at Rancho Los Montemayores; JMS 211 (S)— between General Teran and El Carbendo; UIMNH 3983 (S), 3984 (P)— 14 mi. E Cadereyta, Reynosa-Monterrey rd. TAMAULIPAS: 4 1 30 (P) — Tamaulipas, 9.9 mi. SW Mier. UNITED STATES: TEXAS: County undeter- mined: KU 128993 (P)— 7 mi. S Chacon Creek on Hwy 83 and 8 mi. NE on road to La Gloria Ranch. Dimmit County: KU 1 26948—52 (P) — 26 mi. SCarrizo Springs on Hwy 186, San Pedro ranch. Maverick County: EL 3250. 1-50.2 (P) — 1 mi. E Eagle Pass on Manges Ranch; KU 481 (P) — Eagle Pass, 143567-69 (P)— 1 mi. E Eagle Pass off U.S. Rt. 277, 147257 (P), 147266-76 (S), 1 47277-78, 1 57875-76 (D)- 1 mi. E Eagle Pass, Manges Ranch; SDSNH 46884-86 (P) — 2 mi. E Eagle Pass, 56754-55 (P) — 1 mi. E Eagle Pass. McMullen County: CM 64677 (P) — 4. 1 mi. W jet. St. Hwy. 16 and FM 624 (ca 22 mi. SSW Tilden). Starr County: KU 9092 (P) — Arroyo El Tigre, ca Rio Grande City, 1 3202 (P)— Rio Grande City, 15388 (P)-23 mi. NW Rio Grande City; UIMNH 20336 (S)— Arroyo Los Alamos, 3 mi. SE Rio Grande City. Webb County: CM 52334-35 (P)— 40 mi. WNW Laredo on FM 1472; EL 4748 (P) — 21.8 mi. W Mirando City on Texas Rt. 359; KU 61 447-49 (P)-40 mi. NW Laredo, 121487, 121489, 121491 (P) — 5.2 mi. E jet. Hwys 44 and 83, 121488 (P) — 10 mi. S Laredo, 126940-47, 126953-56, 126958 (P)-40 mi. WNW Laredo on FM 1472, Trevino Ranch, 128990 (P) — 2 1 mi. NW 1-35, ca Laredo on FM 1472, 128992 (P)— 23 mi. NW 1-35, ca Laredo on FM 1472, 7 mi. NE on El Chapote Rd.; REE 2906 (D)— 37.0 mi. NNW Laredo on FM 1472, 2907 (D)-41.8 mi. NNW Laredo on FM 1472, 2908 (Dj-25.3 mi. NNW Laredo on FM 1472, 2909 (D)-34.4 mi. NNW Laredo on FM 1472, 2910 (D) — 36.0 mi. NNW Laredo on FM 1472, 2911 (D)- 19.6 mi. NNW Laredo on FM 1 472, 29 1 2 (D)-35.7 mi. NNW Laredo on FM 1472, 2913 (D)-22.6 mi. NNW Laredo on FM 1472; SDSNH 41333 (P) — about 10 mi. S Laredo. Zapata County: KU 13203 (P)— San Ignacio. C. vestigium. — MEXICO: BAJA CALIFORNIA: CAS 14000- 01 (P) — Vicinity Bahia de Los Angeles, 154267 (P) — 7 km W (by rd.) of Bahia de Los Angeles; JMS 207 (S) — El Marmol, 210 (S), SDSNH 43226 (P)— Sierra de Juarez, Camillas Canyon; REE 2806 (D)— 1.5 mi. N Bahia de San Luis Gonzaga, 2807-08 (D)— Sierra Las Pintas, 2810 (D)— 10 km W Bahia de Los Angeles, 2814 (D)— 1 km W Bahia de Los Angeles, 2815 (D)— 1.5 km S of hwy to Bahia de Los Angeles at km marker 56, 2822 (D) — 46 km W int. Mex. Hwys 3 and 5 on Hwy 3, 2823 (D)— Sierra San Felipe, Campo La Roca, 18.5 mi. S San Felipe, 2824 (D)— 20 mi. NW San Felipe, int. powerlines and Sierra San Felipe, 2936 (D)— 28.5 mi. N Bahia de Los Angeles; SDSNH 17052 (P)-S base of Sierra de Juarez, 17667 (P) — San Borja, 24391-92 (P) — San Jose, 26754 (P)— E side Sierra San Pedro Martir, Canon del Cardones, 37815 (P)— 1 mi. NW San Felipe, 41612 (P)— appx. 2.5 mi. W Bahia de Los Angeles, 45978 (P) — SE Mesa de San Carlos, 52950-5 1 (P) — Bahia de Los Angeles, 1 .8 mi. S of V.S.E. Field Station; SDSU 1726-27 (P)— 5 km E El Parador on rd. to Bahia de Los Angeles, 1728 (P) — W base Sierra La Asamblea, appx. 20 mi. N El Parador. BAJA CALIFORNIA SUR: CAS 18822 (P)-BCS, 9 mi. W San Ignacio, 146684 (P)-Santa Ague- da, 147683 (P)— 29.1 mi. S (by rd.) Mulege on Mex. Hwy 1, 154268-70, 154272 (P), 154271 (P, H)-Santa Agueda; REE 2809 (D)— 9.7 mi ESan Isidro, 281 1 (D)— 16 km S Mulege, 2812 (D)— 2 km E San Jose de Magdelena, 2813 (D)— 1 km E San Jose de Magdelena, 2816-17 (D) — km marker 76 N Loreto, 2818 (D)-Km 28 E Mex. Hwy 1 on rd to San Francisco de La Sierra, 2819 (D)— 7.6 mi. E San Isidro, 2820 (D)— 5.4 mi. E San Isidro, 2821 (D) — 3.0 mi. E San Isidro, 2825 (D)— 17.2 mi. S by rd. of San Jose de Comondu, 2826 (D)— 1 3.2 mi. S by rd. of San Jose de Comondu; SDSU 1729 (P)— 10.6 mi. S San Jose de Comondu, 1 730 (P)— Rd. to San Francisco de La Sierra. UNITED STATES: CALIFORNIA: Imperial County: JAM 41 (ARAB)-Inkopah Trail, 2 mi. E Jacumba; REE 2935 (D) — Mountain Springs, N side westbound Hwy 8. Riverside County: BYU 2422, 2430, 2432, 2435, 2438 (P) — Chino Canyon, W Palm Springs. San Diego County: CAS 7930 (P)— Palm Canyon, 62794-95 (P) — San Felipe Creek, 62875 (P)- Mason Valley; JAM 617-18 (P)- VallecitosCo. Park, McCain Pit; JMS 199 (S)-Sentenac Canyon, 204 (S) — 3.5 mi. E Jacumba, 209 (S)— Borrego Mtn. No data: SDSNH 19788-92 (P). Gambelia copei. — MEXICO: BAJA CALIFORNIA: MVZ 31794-95 (P)— 3 mi. W Canyon de Llanos, ca 10 mi SW Alaska (= Rumarosa), 31839 (P) — 6 mi. W Alaska (= Rumarosa), 140759 (P) — Sierra San Pedro Martir, 2 mi. SW Paseo de San Matias; 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 123 REE 2800 (D)-1.0 km N San Borja; SDSNH 4071 (P)-San Jose, San Telrno River, 4143 (P) — 3 mi. E. San Telmo, 5078- 80, 26752-53 (P)-San Jose, 5264, 15969 (P)-Isla de Cedros, S. end, 7249, 17411, 24340-42 (P)-Isla de Cedros, 18118 (P)— Bahia de San Francisquito, 18945-46 (P)— Valle de Trinidad, 19787 (P)— 40 mi. W Bahia de Los Angeles, 27693-94 (P)— Isla de Cedros, canyon N middle canyon, 27965 (P) — Isla de Cedros, Middle Canyon, 41613 (P)— Mouth of Arroyo de Rosario, 42622 (P) — 2 mi. N. San Simon, 43007 (P) — 8 mi, E. El Rosario, 45916 (P)-NE Mesa de San Carlos. BAJA CALIFORNIA SUR: CAS 56105 (P)— Isla Santa Margarita, 147738-39 (P)-1.3 mi. NE Punta Abreojos, 147750 (P)— SE Sierra Santa Clara; REE 2798 (D)— km 24.5 on rd. to Punta Abreojos, 2799 (D) — Sierra Santa Clara, 2801 (D)— jet. rds. to Puerto Nuevo and Bahia Tortuga, 2802 (D)— km 57 W of Mex. Hwy 1 on rd. to Punta Abreojos, 2803 (D)— Sierra Santa Clara, Rancho Santa Clara, 2804 (D)— Sierra Santa Clara, 1.0 mi. S Rancho San Ramon, 2805 (D)— km 30.25 on rd. to Punta Abreojos; SDSNH 1 7470 (P) — El Arco, Miraflores Rancho. UNITED STATES: CALIFORNIA: San Di- ego County: CAS 40302, 57865 (P)— Campo, 62964 (P) — Potrero Grade; SDSNH 55251 (P)— 1.5 mi. NE Cameron Comers. G. coronaf. —UNITED STATES: CALIFORNIA: San Diego County: LACM 7058/42880 (S)— Anza Borrego Desert State Park. G. silus. — CALIFORNIA: Fresno County: CAS 22713 (D), 227 14 (D, H)— in dry Panoche Creek bed, mouth of canyon on W side San Joaquin Valley (about 120°39'W, 36°38'N), 20 mi. by rd. WSW Mendota, 23250 (H)— nr. foothills, 3 mi. SE mouth of Panoche Canyon and 16.5 mi. SW Mendota, next to pole line rd., “Staggeredrock trap station,” 141318-19 (D)— 20.2 mi. S jet. Cal Hwys 33 and 1 80, and 1 .7 mi. W on dirt rd. (nr. three Rocks); KU 121493 (P), 121751 (S)-Bundgard Ranch, 10 mi. ESE Men- dota, 121 500 (P) — 8 mi. ESE Mendota ca Double C Ranch, 121504 (P), 121752-53 (S) — 9 mi. ESE Mendota at Double C Ranch, 121507 (P)— 121520, 121526-27 (P)-2 mi. SWjct. Interstate 5 and Shields Ave. on Panoche Plain, 121524 (P), 121510 (S)— Mouth of Little Panoche Creek on Levy-Zentner Ranch, 121754 (S) — 2 mi. S jet. Shields Ave. and Little Panoche Rd., 12175 5— 56 (S) — 2 mi. SSE jet. Shields Ave. and Little Panoche Rd., 1 2 1 758-60, 121762 (S)- 1 mi. NW Three Rocks, 1 5 mi. S Men- dota, 121764 (S) — Levy-Zentner Ranch, 1 mi. E mouth Little Panoche Creek, 121765 (S) — 3 mi. N Mercey Hot Springs along Little Panoche Creek, 121766 (S) — 2 mi. S jet. Interstate 5 and Shields Ave, 121767-68 (S)— Levy-Zentner Ranch, Little Pa- noche Wash. Fresno County (?), 2 mi. S turnoff to Little Panoche Ranch on Little Panoche Rd., 121511 (P), 121757 (S) — Little Panoche Ranch turnoff on Little Panoche Rd. Kern County: KU 121769-75 (S) — Blackwell’s Comer, 30 mi. W Wasco, jet Hwys 33 and 46; SDSNH 16055-59 (P)-3 mi. N McKittrick, 42434- 35 (P) — W end Greenhorn Mtns, 46339 (P) — Bakersfield. Kings County: JMS 206 (S) — 2 mi. S Kettleman City; SDSNH 31697 (P)— Wheeler Ridge Post Office. Madera County: KU 121605, 121610, 121615-16 (P)— 8 mi. E Firebaugh, 121623 (P)-4 mi. E Firebaugh on Rd. 9, 121748, 121750 (S) — 8 mi. E Firebaugh off Ave. 7 1/2, 121749 (S) — 12 mi. E Firebaugh; SDSNH 46888- 89, 49758-59 (P) — 5.9 mi. E Firebaugh. Merced County: KU 121631-34, 121636-38 (P)— 4.5 mi. NWjct. Hwy. I 52 and Hwy. 59, Red Top, 121 644 — 46, 121648 (P) — 10 mi. SW Los Banos on Arburua Rd., 1 2 1 647 (P) — 9 mi. SW Los Banos on Arburua Rd., 121649 (P)— Wjct. Arburua and Langdon Rds., 121650-5 1 (P)— 8.7 mi. S Los Banos off Mercy Springs Rd., 121652 (P) — 1 0 mi. SW Los Banos, W jet. Arburua and Langdon Rds. San Benito County: CAS 22724-25 (D)— in and about the dry wash, SE end Panoche Valley; KU 121537 (P)-2 mi. N jet. Little and Big Panoche Rds. San Luis Obispo County: CAS 23 1 95 (H) — Carrizo Plain, dry creek at N end of false valley between Panorama Hills and the Temblor Mtns., 13 mi. at 235 degrees from Simmler; KU 121657 (P)— 0.3 mi. W and 0.2 mi. S jet. CA Rts. 33 and 166, 121658 (P) — 1.7 mi. E jet. CA Rts. 33 and 166, 121659 (P) — Cuyama Valley at jet. CA Rts. 33 and 166, 121662 (P)— 7.8 mi. N jet. Soda Lake Rd. and CA Rt. 33, 121664 (P) — 7.5 mi. N jet. Soda Lake Rd. and CA Rt. 33, 121671 (P) — 7 mi. W Maricopa. G. wislizenii. — MEXICO: BAJA CALIFORNIA: BYU 23336 (P) — 5 mi. N. San Felipe, 34513 (P) — 3 mi. S. San Felipe (by rd. to Puertocitos), 34514 (P)— 2.9 mi. S. San Felipe (by rd. to Air- port), 34515 (P) — 4 mi. W. San Felipe (at trash dump); CAS 90256 (P)— San Felipe-Ensenada Rd. (Mex. Hwy. 3), 6.8 mi. W. of San Felipe-Mexicali Hwy (Mex. Hwy 5), 119100 (P)— Mouth of Guadalupe Canyon; LACM 94813 (P) — Arroyo Matomi, 132230 (P)— N. end Laguna Salada, 132231 (P) — 3 mi. N. Pozo Penara, Laguna Salada; MVZ 9589 (P)— E. base San Pedro Martir Mountains, El Cajon Canyon, 50017 (P)— Punta San Felipe, 182117 (P) — 5.8 mi. N. San Felipe (via Mex. Hwy. 5). CHI- HUAHUA: UIMNH 6672-73 (S) — 28.7 mi. S Samalayuca, 40408 (S)— sand dunes 35 mi. S Juarez, 43373-75 (S) — 6.8 mi. S Sa- malayuca, 43383 (S) — 0.3 mi. E Carillo. COAHUILA: EL 3129 (P) — 5.2 mi. S. Cuatrocienegas de Carranza along Rio Mesquites; UIMNH 43378 (S) — 7 mi. E Matamoros. DURANGO: UIMNH 43379 (S) — 13.5 mi. S Tlahualilo. SONORA: CAS 15347 (P)- 1.5 mi. W Altar, 15356 (P)— 4.7 mi. SSE La Playa, 17049-50 (P) — Isla Tiburon, SE end of island, 104451 (P)— Isla Tiburon, SW end; REE 2789-91 (D)— Isla Tiburon, appx. 2 mi. N El Corralito (S end of island); SDSNH 38251 (P)-3 mi. NE Punta Penasco, 38252 (P)— 16 mi. NE El Papalote, 38253 (P) — 18.5 mi. NE El Papalote, 38254 (P)— 1 2 mi. NE Punta Penasco, 38605 (P)-El Papalote, 38606 (P)- 1 mi. NE El Papalote, 38888 (P)- 12 mi. NE Punta Penasco, 40601 (P)— 24 mi. N Punta Penasco, 49009 (P)— 36 mi. E. San Luis. UNITED STATES: ARIZONA: Coconino County: SDSNH 6030-32 (P)— Grand Falls of Little Colorado River, 32560 (P)— Nr. Jacob Lake, 35813-14 (P) — 3 mi. SW Navajo Bridge. Maricopa County: JMS 187 (S)— Wick- enburg. Yuma County: REE 810 (D)— Yuma County. Undeter- mined: USNM 220224 (D) — Arizona. CALIFORNIA: Imperial County: REE 1029, 1172 (D)-Glamis, 2915 (D)-Salton City; SDSNH 1879 (P) — 5 mi. E. Holtville, 7143 (P) — 4 mi. N. Kane Spring, 7847 (P)— Kane Spring, 10937 (P) — Mountain Spring, 11346 (P)— Coyote Wells, 13352 (P) — 4 mi. N. Bard, 13911, 20967 (P) — Gray’s Well, 1 8596 (P) — 6 mi. N. Truckhaven, 28762 (P)-Niland. 36541 (P)-Ocotillo, 39735 (P)-5 mi. E San Di- ego-Imperial Co. line, 49002 (P)— 17.5 mi. W. Calexico, 49003 (P) — 3 mi. E. Coyote Wells. Kem County: USNM 18298 (D)— Kernville. San Bernardino County: CAS 190054 (H) — Kelbaker Rd., 2.6 mi. SE Baker; JMS 41 (S) — 29 Palms; REE 1571 (D)- 17 mi. ESE Lucerne Valley, 2916-17 (D), SDSNH 68662, 68664 (P)— Hinkley, Hinkley Rd, REE 2918-20 (D), SDSNH 68663, SDSU 2282 (P)— Johnson Valley, Camp Rock Rd; SDSNH 23636 (P)— Stoddard Well. San Diego County: JMS 21, 186 (S)— Clark Dry Lake, 1 84 (S)— Borrego Valley, 185-1,2 (S)-Split Mountain, 188 (S) — 14 mi. E Benson’s Dry Lake. County undetermined: REE 496, 550 (D) California. IDAHO: Ada County: SDSNH 1450-51 (P) — Ada Co., foothills N. of Boise. Elmore County: SDSNH 1452 (P)-S. of Cleft. NEVADA: Nye County: UIMNH 93992-93 (S) — 1 5 mi. N Mercury. Storey County: REE 2914 (D)— Carson City. Washoe County: UIMNH 3166 (S)— 20 mi. 124 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 N Reno. 3167-68 (S) — 5 mi. S Sutcliffe, Pyramid Lake, USNM 220226 (D) — Sutcliffe, Pyramid Lake. County undetermined: Nevada, camp 1 2. NEW MEXICO: Luna County: EL 5 1 76 (P)- 4.2 mi. S., 0.6 mi. E. Deming on Rock Hound State Park road. UTAH: Javier County: JMS 1 83 (S) — 2 mi. W Monroe. San Juan County: CAS 141349 (H)— 10.8 rd. mi. N Montezuma Creek. No locality: JMS 700 (D). Chamaeleonidae Brookesia kersteni: REE 532 (D). B. stumpffi: REE 1911 (D). Hydrosaunis amboiensis: REE 2068 (D), 2080 (D); SDSNH 47009 (D). H. pustulatus: CAS 11000-01 (P), 28171 (P), 62377 (P), 85642 (P). Leiolepis belliana: REE 1680 (D), 1906 (D), 1908 (D), 1993 (D), 2505 (D); SDSU 2587-90 (P). Physignathus cocincinus: SDSNH 67845 (D), 68062 (D). P. lesueurii: KU 69303 (S, P), 69304 (P); REE 1364 (D), 1722 (D), 1849 (D). Uromastyx acan- thinurus: CAS 135162 (P), 1 35 166-67 (P); KU 94507 (S, P); REE 318 (D), 450 (D); SDSNH 62665 (D). U. aegyptius: SDSU 2584 (P). U. asmussi: CAS 154357 (P). U. benti: SDSNH 68121 (D). U. geyrii: CAS 135006-16 (P). U. hardwickii: REE 1339 (D), 1 840 (D); SDSU 2573-78 (P). U. loricatus: CAS 86379 (P), 86463 (P), 120480 (P). U. macfadyeni: SDSU 2580 (P). U. microlepis (synonymized with U. aegyptius by Moody, 1987): CAS 97834- 35 (P); SDSNH 55288 (D); SDSU 2585-86 (P). U. ocellatus: SDSU 2582-83 (P). U. philbyi: C AS 139537 (P), 141997-98 (P); SDSU 2579 (P). U. thomasi: CAS 190887 (P); SDSU 2581 (P). Corytophanidae Basiliscus basiliscus: KU 84956 (D), 93452-54 (D); REE 2015 (D). B. plumifrons: KU 25660 (P), 91784 (P), 96637 (P), 180368 (P); REE 427 (D), 2014 (D); SDSNH 57098 (D), 57100 (D); SDSU 2093 (P). B. vittatus: REE 49 (D), 555 (D), 637 (D), 1601 (D), 1729 (D), 1757 (D), 1759 (D); SDSU 2095-96 (P). Cory- tophanes cristatus: KU 59602 (P); SDSNH 62345 (D), 67849- 50 (D); SDSU 2098-2100 (P). C. hernandezi: KU 24068 (P), 24070-71 (P), 24073 (P); REE 1 176 (D), 1800 (D); SDSNH 68090 (D). C. percarinatus: KU 93456 (S), 184183-84 (P), 187149-50 (P), 190773 (D). Laemanclus longipes: KU 27529 (P), 59608 (P), 187739 (P); SDSNH 64542 (D), 67835 (D), 68086 (D). L. ser- ratus: KU 70226 (P), 70267 (P), 74910 (D), 75532 (P); REE 619 (D); SDSU 2095 (P). Hoplocercidae Enyalioides laticeps: KU 125967 (D), 147929-34 (P), 147937 (P), 147939-42 (P), 152497-98 (P); REE 76 (D); SDSU 2116- 17. E. oshaughnessyi: KU 122116 (P), 147183 (P); REE 1957 (D). E. praestabilis: KU 122117 (P), 140394 (P), 147184 (P), 169854 (P). Iguanidae Brachylophus fasciatus: REE 1019 (D), 1866 (D), 1888 (D); SDSNH 55601 (D), 55603 (D); SDSU 2591-93 (P). Dipsosaurus dorsalis: JAM 287 (D), 345-51 (D); SDSU 2594-600 (P). Opluridae Chalaradon madagascariensis: KU 187757 (P), 187762-63 (P), 187765 (P), 187756 (S); REE 455 (D), 457 (D), 547 (D); SDSU 2123-29. Oplurus cuvieri: JAM 281 (D); KU 1 87666-68 (P); REE 558 (D), 620 (D), 1835 (D). O. cyclurus: CAS 86739 (P). O. fierinensis: KU 187769 (P), 187770 (S,P), 187771—72 (P). O. quadrimaculatus: REE 658 (D); SDSU 2120-22 (P). O. saxicola: CAS 13958 (P), 14439 (P), 86724 (P); SDSU 2119 (P). Phrynosomatidae Callisaurus draconoides: JAM 88 (D), 184 (D), 202 (D), 361 (D). Petrosaurus mearnsi: JAM 285 (D), 288-90 (D), 295 (D); REE 35 1 (D), 557 (D); SDSU 2253 (P). P. repens: SDSNH 1 7484 (P), 45985 (P). P. thalassinus: REE 575 (D), 765 (D); SDSNH 17484 (P), 32922 (P), 44516 (P), 45985 (P). Phrynosoma asio: REE 1489 (D), 1580 (D), 1676 (D); SDSU 2308-09 (P). P. co- ronatum: REE 310 (D), 390 (D), 527 (D), 609 (D), 1438-39 (D), 1786 (D), 1999 (D); SDSNH 1 6042^43 (P); SDSU 2305-07 (P). P. ditmarsi: SDSU 2278 (P). P. douglassi: REE 1109-11 (D), 1 118(D), 1372 (D); SDSU 2283-84 (P). P. orbicular e: REE 1104 (D), 1181 (D), 1725 (D), 1920 (D), 1931 (D). Uma exsul: REE 2880-81 (D); SDSU 2274-77 (P). U. inornata: KU 90961 (D), 95849 (D); REE 263-64 (D), 602 (D),1538 (D); SDSNH 2754 (P), 48486 (D). U. notata: JAM 172 (D), 235-37 (D), 239-41 (D); SDSU 2558-63 (P). U. scoparia: BDH 117 (D); CAS 42135 (S); REE 509 (D), 551 (D), 2867 (D); SDSNH 7556 (P), 7658 (P), 38419 (P). Urosaurus auriculatus: SDSNH 34853 (P), 34859 (P), 34861 (D), 34866 (P). U. bicarinatus: SDSNH 7371 (P), 10154 (P), 28513 (P). U. clarionensis: SDSNH 22514 (P), 22529 (P), 28507 (P). Uta nolascensis: CAS 14244 (P), 14247-48 (P). U. palmeri: SDSNH 46492-94 (P), 46496 (P). U. squamata: CAS 52343 (P), 52351 (P), 52359 (P). U. stansburiana: JAM 265 (D), 284 (D), 301 (D), 366 (D); REE 274-75 (D), 1877-78 (D); SDSNH 3374 (P), 60800-1 10 (P), 60800-187 (P), 60800-418 (P); SDSU 2525-30 (P). Polychrotidae Anisolepis grilli: REE 1952 (D); SDSU 2130-31 (P). Chamae- leolis chamaeleonides: CAS 1 46 1 0 (P); KU 245644 (P). C. porcus: KU 245645. Enyalius bibronii: MCZ 163783 (P). E. bilineatus: MCZ 5567, 84034, 144556, 163776, 163777 (P); REE 1678 (D), 1958 (D). E. boulengeri: MCZ 163780 (P), 163781 (D). E. bras- iliensis: MCZ 3317, 3322, 4251, 163778-79 (P); REE 1960 (D). E. catenatus: CAS 16101 (P); MCZ 163782 (P); REE 1961 (D). E. iheringii: MCZ 6315, 163786-87 (P); REE 1959 (D); SDSU 2222-23 (P). E. perditus: MCZ 163788 (D), 163789 (P). E. pictus: MCZ 82873 (P), 163784 (D), 163785 (P); SDSU 222 1 (P). Phen- acosaurus heterodennis: SDSU 2224-25 (P). P. richteri: SDSU 2226-27, 2240 (P). Polychrus acutirostris: KU 73436-38 (P); MZUSP 568 (D), 4412 (D), 4448 (D), 4543 (D); SDSU 2236-37 (P). P. femoralis: KU 142682 (P), 218381 (P). P. guttarosus: KU 25170 (P), 76074 (P), 113495 (P); SDSU 2235 (P). P. liogaster: KU 133872-73 (P). P. marmoratus: JMS 1 16-1 17; REE 346 (D), 2283 (D), 2496 (D), 2498 (D), 2863 (S); SDSU 2231-34 (P). Pristidactylus casuhatiensis: MCZ 162924 (D). P. torquatus: CAS 85234 (D); MCZ 33586 (D); REE 2766-68 (D); SDSU 2249-51 (P). Urostrophus vautieri: CAS 13883 (P); REE 2507 (D); SDSU 2522 (P). Tropiduridae Ctenoblepharys adspersus: LACM 49147 (D); MVZ 85415-16 (P); REE 25 1 3 (D). Leiocephalus carinatus: REE 1469 (S), 1 805- 06 (D), 1816 (D); SDSNH 67957-58 (P); SDSU 1996-97 (P). L. greenwayi: REE 1814 (D). L. inaguae: KU 242855 (P), 242859 (P), 242865 (P), 242868 (P). L. macropus: REE 1819 (S); SDSNH 65959-60 (P), 65989 (D), 66002 (P), 66004-05 (D), 66012 (P). L. melanochlorus: KU 243460 (P), 243463 (P), 243470 (P), 243474 (P); REE 1802 (D). L. pratensis: KU 244861-62 (P), 244864 (P), 246145 (P). L. psammodromus: KU 244836 (P), 244838-39 (P), 244843 (P); REE 1813 (D). L. schreibersi: KU 245006-08 (P); REE 1808 (D); SDSNH 64665 (D), 64668-69 (D), 64672 (P), 1996 McGUIRE— SYSTEM ATICS OF CROTAPHYTID LIZARDS 125 64675 (P), 66967 (P), 66970 (P); SDSU 1998 (P). Microlophus duncanensis: CAS 1 2202 (D). M. grayi: CAS 1 1 620 (D). M. stolz- manni: KU 134701 (P), 134712 (P), 134743-44 (P). M. there- sioides: KU 162010-1 1 (P), 162015-16 (P). M. tigris: KU 163750- 52 (P), 163757 (P). Phymaturus palluma: REE 2306 (D), 2309 (D), 2311 (D), 2313 (D), 2326 (D); SDSU 1946-51 (P). P. pa- tagonicus patagonicus: REE 247 1-72 (D); SDSU 1980 (P). P. p. payuniae: REE 233 1-33 (D), 2336 (D), 2339 (D), 2360 (D); SDSU 1981-84 (P). P. p. somuncurensis: REE 2433-36 (D), 2439 (D); SDSU 1780-84 (P). P. p. zapalensis: REE 2451-53 (D); SDSU 1986-90 (P). P. punae: REE 2356-7 (D), 2383-85 (D); 1978-79 (P). P. sp.: SDSU 1991-95 (P). Plesiomicrolophus koepckeorum: KU 163604 (P), 163606-07 (P), 212665 (P). Stenocercus guenth- eri: SDSNH 49472 (P). Uranoscodon superciliosus: KU 128214 (P), 128215 (D), 128216 (P), 128218 (P), 130218 (P), 135269 (D); REE 2589 (D); SDSNH 65497 (D); SDSU 2110 (P). 126 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Appendix 2 Data Matrix The symbols a-y represent frequency ranges within which the derived character state was observed in any particular terminal taxon (see Table 1). “?” = missing or unknown. 1 2 3 4 5 l 6 7 8 9 0 tiiii 1 2 3 4 5 11112 6789 0 2 2 2 2 2 1 2 3 4 5 2 2 2 2 3 6 7 8 9 0 3 3 3 3 3 1 2 345 3 3 3 3 4 6789 0 4 4 4 4 4 1 2 3 4 5 Ancestor aaaa? aa??a aa?aa a??a? ?a??a aaOaa ?aa? ? a?aaa cl 3. EL 3, cL G. copei yyayy yyaay ypayy ayaaa yaayy yaOyy Cwaaa yaaah ydyyy G. corona t a?a?y ?aaa? ?y ? ? ? a?aaa ?y??y ya0?y Dya? ? ?a??? 99999 G. situs fyaya ydaac ayaym axaay ya?w? yaOyy Anaaa yauaa ukyyy G. wislizenii yyayy ycaax yyayy awaaa yaayy yaOyy Bvaaa Q3.8.3.6 xfyyc C. bicinctores babya yayyy aayya aaayy ay?aa bylya Lkyyy ayayy byyde C. antiquus agmya yayys aayya aaayy ayyaa aylya Fsysy gyaay ayyym C. collaris abaya yayyy aayya avayy axyba cylya Gkyyy ayaay asyaa C. dickersonae aayya yayyy aayya aayyy ayyaa dylya Isyyy ayayy ayyai C. grismeri faaya yayyy aayya akayy ay?aa aylya Jkyyy ayayy auyfp C. insularis yaaya yayyy aayya yaayy ayyaa yyiya Kayyy ayayy ayyfa C. nebrius aaaya yayyy aayya aaayy ayyda bylya Hsyyy ayaay axyaa C. reticulatus aaaya yayyy aayya aeayy ayyca cy2ya Eqyyy ayaay awqay C. vestigium baaya yayyy aayya jaayy ayyba gyiya Mkyyy ayayy axydc 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 127 Appendix 2— Extended 4 4 44 5 6 7 8 9 0 5 5 5 5 5 1 2 34 5 5 5 5 5 6 6789 0 6 6 6 6 6 1 2 3 4 5 6 6 667 67 89 0 7 7 7 7 7 1 2 3 4 5 77778 6789 0 8 8 8 8 12 3 4 8 5 8 888 9 67 89 0 9 9 9 9 9 1 2 34 5 999 678 yaaaa aaaa? ?aa?a aaaaa aa??a aa??? aa?a? ??a0 ? ?aa?? 99999 ??? aaaay ayyay aayay ayaaa aaOaa ya??? aa?a? ?yaO (04) a?a?? 99999 ??? 9 9 9 7 9 99999 99999 99999 99999 99999 99999 ???? ? 99999 99999 ??? yaaaa ayaay aayaa ayaya aaOaa ya,? ? ? aa?a? ?yaO 0 aaa?? 99999 ??? yaaaa ayyay aayay ayaaa aaOaa ya??? aa?a? ?yaO 0 ayall 11111 Ill ayayy vaaya yyyya yaaay yy2yL ayay2 ayyay aasO 4 yay44 44444 444 ayayy yaaya yyyya ??aya yyiyy yyay ( 012 ) yyayy aaaO 4 ya??? 99999 ??? gtayy eaaya yyyya yaaaa ya2yh yyaa( 01 ) thaaa aac ( 03 ) (345) yay66 66666 666 ayyyy yaaya yyyya yayay ya2yy yyayi ayyay aaa3 3 yay22 22222 222 ayayy yaaya yyyya yaaay ya2yy ayay2 ayyay aaaO 4 yay?? 99999 ??? apayy yaaya yyyya yaaay ya4ya ayyy3 ayyay aapl 4 yay?? 99999 ??? awayy yaaya yyyya yaaaa ya2yy yyayO gyaaa xas2 2 yay55 55555 555 avayy aaaya yyaya yaaaa yalaa yyyy(Oi) ta?yy aaaO 1 aay77 77777 777 ayayy yaaya yyyya yaaay ya3yy ayyy3 ayyay aau( 12 ) 4 yay33 33333 333 128 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Appendix 3 Outgroup Data Matrix Species that exhibited more than one character state were assigned state V (variable) in this data matrix. “?” = missing or unknown. 1 2 3 4 5 1 67 89 0 11111 1 2 34 5 11112 6789 0 2 2 2 2 2 1 2 34 5 2 2 2 2 3 67 89 0 3 3 3 3 3 1 2 3 4 5 3 3 3 3 4 6 7 89 0 Brookesia stumpffi 0?10? 000?? ?0100 0110? 01??1 00000 ?0??? ????! Chamaeleo kersteni 0?00? 001?? ?0000 0110? 0???1 00000 ?0??? ????! Hydrosaurus amboiensis 0000? 10110 00000 0010? ?01?0 00200 ?0?01 ??001 H. pustulatus 99999 99999 99999 99999 99999 99999 99999 ????1 Leiolepis belliana 0000? 00110 00011 0110? 011?0 10000 ?0?00 V?001 Physignathus cocincinus 0000? 10110 ?0000 0110? 001?0 00000 ?0??? ??001 P. lesueurii 0000? 10110 00010 0111? 001?0 00000 ?0?01 0?001 Uromastyx acanthinurus 0000? voovo 00000 0111? 101?1 01000 ?0??? ??oov U. aegyptius 99999 99999 99999 99999 99999 99999 99999 99999 U. asmussi 99999 99999 99999 99999 99999 99999 99999 ????1 U. benti 0000? 00010 00000 0?11? 1?1?1 00000 ?0??? ??001 U. gevrii 99999 99999 99999 99999 99999 99999 99999 ????! U. hardwickii 0?00? ?0000 00000 Oil?? 001?1 00000 ?0?0? ??001 U. loricatus 99999 99999 99999 99999 99999 99999 99999 ????1 U. mafadyeni 99999 99999 99999 99999 99999 99999 99999 ????1 U. microlepis 0000? 00010 00000 0111? 101?1 01000 ?0??? ??001 U. ocellatus 99999 99999 99999 99999 99999 99999 99999 ????1 U. philbyi 99999 99999 99999 99999 99999 99999 9 9 9 9 9 ????! U. thomasi 99999 99999 99999 99999 99999 99999 99999 ????1 Chamaeleonidae 0000? 00110 00000 0110? 001?? 00000 ?o?o? 0?001 Basiliscus basiliscus 0000? 10110 00000 0000? 10100 00000 ?001? 1?100 B. plumifrons 0000? 10110 00000 0010? 10100 00000 ?001? 1?100 B. vittatus 0000? 101?0 00000 0010? 101?0 00000 ?001? 1?100 Corytophanes cristatus 0000? 101V0 ?0000 0010? ?0100 00000 ?0V1? 1?101 C. hernandezi ovoo? 10110 ?ovoo 0V10? 000?0 00000 ?00?? 1?101 C. percarinatus 0000? 10110 ?ovoo 0110? 0?100 00000 ?01?? ??101 Laemanctus longipes 0000? 101V0 00000 0010? vovoo 00000 ?001? 1?101 L. serratus 0000? 10110 00000 0100? 00100 00000 ?0V1? ??101 Corytophanidae 0000? 101?0 00000 0010? ?0100 00000 ?001? 1?10? Enyaliodes laticeps 0000? V0110 oooov 0000? V0100 00000 ?00?1 0?000 E. oshaughnessyi 0000? 10110 00000 0000? 001?0 00000 ?00?1 0?000 E. praeslabilis 99999 99999 99999 99999 99999 99999 99999 99999 Hoplocercidae 0000? 10110 00000 0000? 00100 00000 ?00?1 0?000 Brachvlophus fasciatus 00001 00110 00000 0010? voooo 00000 ?0111 V?101 Dipsosaurus dorsalis V0001 ooovo voooo 0000? ?0000 00000 ?0001 0?000 Iguanidae 00001 00?10 00000 00?0? ?0000 00000 ?0??1 o??oo Chalaradon madagascariensis 0000? ooovo 1V100 ovvo? V0010 00000 ?0001 0?000 Opiums cuvier i 0000? 00110 10100 0000? 100?0 10010 ?0001 0?000 O. cyclurus 99999 99999 99999 99999 99999 99999 99999 ????0 0. fierinensis 0000? 00010 10000 0000? 00110 01100 ?0?1? 1??00 O. quadrimaculatus 0000? 00??0 10100 00??? ?00?0 00000 ?10?? ??000 0. saxicola 99999 99999 99999 99999 99999 99999 99999 ????0 Opluridae 0000? 00?10 10?00 0000? ?0?10 ???00 ??0?1 ??ooo Callisaurus draconoides ????? 99999 ????0 99999 99999 99999 ???01 0?000 Petrosaurus mearnsi 1000? 00000 01V01 ovoo? 10111 00001 ?0?10 0?000 P. repens P. thalassinus 9 9 9 9 9 99999 99999 99999 99999 99999 99999 99999 0000? 00010 oovoo ovoo? 1101? 0000? ?0?11 0?000 Phrynosoma asio P. coronatum 0001? ?00?0 10100 0000? 00010 V0100 ?0?00 1?001 0000? 00010 10100 0000? 10011 010?0 ?0?0? ??001 P. ditmarsi P. douglassi P. orbicu/are 99999 99999 99999 99999 99999 99999 99999 ???01 0000? ?0000 00?00 0100? 10011 ovvoo ?0??? ??001 0000? 00010 10100 0100? 1001? oov?o ?0??? ??001 Uma exsul 0001? 00000 00101 0000? 10010 00000 ?0?11 v??oo U. inornata 0001? 00000 0010V ovoo? 10010 ovooo ?0?11 0?000 1996 McGUIRE-SYSTEMATICS OF CROTAPHYTID LIZARDS 129 Appendix 3— Extended 44444 44445 55555 55556 66666 66667 77777 77778 88888 88889 99999 999 12345 67890 12345 67890 12345 67890 12345 67890 12345 67890 12345 678 ??0?? ????? ????? ????? ????? ????? ????? ????? ????? 0???? ????? ??? ??0?? ????? ????? ????? ????? ????? ????? ????? ????? 0???? ????? ??? 100?0 00??? ????? ????? ????? ????? ????? ????? ????? 0???? ????? ??? 1?0?? ??101 00?00 ??0?? ????0 00??0 00??? 00?0? ????? 0???? ????? ??? ?00?V 00000 00??1 000?0 0?000 00??0 00??? 00?0? ??0?? 0???? ????? ??? 1?0?1 00??? ????? ????0 ???0? ????? ?0??? 0???? ??0?? 0?0?? ????? ??? 100?V V011? 00??0 100?? ??ooo 00??0 00??? oo?o? ??0?? 0?0?? ????? ??? ?01?0 10V11 00?01 VOO?? ????0 00??0 00??? 00?0? ??0?? ??0?? ????? ??? ??1?? ??011 00?01 000?? ????0 00??0 00??? 00?0? ??0?? ??0?? ????? ??? ??1?? ??111 00?01 000?? ????0 00??0 00??? 0???? ??0?? ??0?? ????? ??? ??1?0 ?0??? ????? ????? ????? ????? ????? ????? ????? ??0?? ????? ??? ??1?? ??1?1 00001 ?00?? ????0 00??0 00??? 00?0? ?00?? ??0?? ????? ??? ?01?0 10111 00?01 000?? ????o 00??0 00??? 01?0? ??o?? ??o?? ????? ??? ??1?? ??101 00001 000?? ????0 00??0 00??? 0?00? ??0?? ??0?? ????? ??? ??1?? ? ? 1 1 1 00?01 000?? ????0 00??0 ?0??? 00?0? ??0?? ??0?? ????? ??? ??1?0 10011 00?01 000?? ????o 00??0 00??? oo?o? ??o?? ??0?? ????? ??? ??1?? ??101 00001 000?? ????0 00??0 ?0??? 00?0? ??0?? ??0?? ????? ??? ??1?? ??111 00?01 100?? ????0 00??0 ?0??? 00?0? ????? ??0?? ????? ??? ??1?? ??V11 00?01 000?? ????0 00??0 ?0??? 00?0? ??0?? ??0?? ????? ??? 100?0 00??1 0000? ?oo?o o?ooo 00??0 00??? oo?o? ??0?? 000?? ????? ??? 000?1 00??? ????? ????0 ??0?? ????? ?0??? 0???? ????? 000?? ????? ??? 000?1 00111 0???? 000?? ?oooo 00??0 00??? 00??? ??0?? 000?? ????? ??? 000?1 00011 0???0 000?0 ?0000 00??0 ?0??? 00??? ??0?? 000?? ????? ??? 0?0?0 00101 o???v ooo?i ??ooo 00??0 00??? 00??? ??o?? o?o?? ????? ??? 010?V 00011 0???? ?00?? ??000 00??0 ?0??? 00??? ??0?? 0?0?? ????? ??? 0?0?? 00V01 0???1 ?00?? ??000 00??0 ?0??? 00??? ??0?? 0?0?? ????? ??? 0?0?V 00001 o???o ?oo?? ??ooo 00??0 00??? 00??? ??o?? 0?0?? ????? ??? 000?1 00011 0???0 000?? ??000 00??0 ?0??? 00??? ??0?? 0?0?? ????? ??? 000?1 00?11 o???o 000?? ?oooo 00??0 00??? 00??? ??o?? 000?? ????? ??? 00000 00101 voiov ?oo?i ????o oo??o 00??? oo?o? ??o?? 0???? ????? ??? 00000 00111 oo?oi ?oo?? ????o oo??o ?o??? oo?o? ??o?? ????? ????? ??? 0?0?? ??111 00?01 ?00?? ????0 00??0 ?0??? 00?0? ??0?? ????? ????? ??? 00000 001?1 0010? ?00?1 ????o 00??0 00??? oo?o? ??o?? ????? ????? ??? 00000 0000? 0000? ?oo?o o?ooo oo??o ?o??? o??o? ??0?? 000?? ????? ??? 00000 V0110 00001 ?01?0 00000 00?00 00??? oo?o? ??o?? 000?? ????? ??? 00000 00??0 0000? ?o??o 00000 00??0 00??? oo?o? ??0?? 000?? ????? ??? 00000 0001V o??oo ooi?o ????o 00??0 00??? 00??? ??0?? ?0??? ????? ??? 00000 100V0 0??01 001?? ??o?o 00??0 ?1??? 00??? ??o?? ????? ????? ??? 0?0?? ??011 0??00 ?01?? ??0?0 00??0 ?1??? 00??? ??0?? ????? ????? ??? 0?0?? ??01V 0??0V 001?? ????0 10??0 ?0??? 00??? ??0?? ????? ????? ??? 000?0 ?0010 0??0? 00??? ??0?0 00??0 ?0??? 00??? ??0?? ????? ????? ??? 0?0?? ??010 0??00 ?01?? ????0 00??? 00??? 00??? ??0?? ????? ????? ??? 00000 ?001? 0??0? 001?0 ??0?0 ?0??0 0???? 00??? ??0?? ?0??? ????? ??? ?0000 01??? ????? ????0 00000 00?10 00??? 00?0? ??0?? 000?? ????? ??? 0?000 10?10 10000 000?0 00000 00?00 00??? 00?0? ??0?? 100?? ????? ??? ????? ??010 10000 ?00?0 ?0000 00?10 ?1??? 01?0? ??0?? 1?0?? ????? ??? 0?000 10010 10000 000?0 ?00?0 00?10 01??? 01?0? ?00?? 1?0?? ????? ??? ?oooo 00?01 00?00 ?00?? ??0?0 00??0 00??? oo?o? ??0?? 0???? ????? ??? ??000 00101 00?00 100?1 00000 00?00 00??? oo?o? ??0?? 0?0?? ????? ??? ??0?? ???01 00?00 000?? 00000 00??0 00??? 00?0? ??0?? 1???? ????? ??? ?0000 00?01 00000 000?1 00000 00?00 00??? oo?o? ??0?? 0?0?? ????? ??? ?0010 00??? ????? ??0?1 ??0?0 00??? 00??? 00?0? ??0?? ????? ????? ??? ooooo moo 00000 000?0 ??0?0 00?00 00??? 01?0? ??0?? 000?? ????? ??? 0000V 11110 0000V 000?0 00000 00?00 00??? 00?0? ??0?? 000?? ????? ??? 130 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Appendix 3 — Continued 1 2 3 4 5 1 6 7 8 9 0 11111 1 2 34 5 11112 67 8 9 0 2 2 2 2 2 1 2 34 5 2 2 2 2 3 6789 0 3 3 3 3 3 1 2 34 5 3 3 3 3 4 67 89 0 U. notata 0001? 00010 00101 0000? 10010 00000 ?0?V1 0??00 U. scoparia 0001? 00010 ?010V ovoo? 10010 00000 ?0?11 0?000 Urosaurus auriculatus 000?? 00010 10000 o?oo? 1?01? 11000 ?0??? ??000 U. bicarinatus 99999 99999 99999 99999 99999 99999 99999 ???00 U. clarionensis 99999 99999 99999 99999 99999 99999 99999 ???00 Uta nolascensis 99999 99999 99999 99999 99999 99999 99999 ???00 U. palmeri 99999 99999 99999 99999 99999 99999 99999 ???oo U. squamata 99999 99999 99999 99999 99999 99999 99999 ???oo U. stansburiana vooo? 00000 00101 OVOO? V1010 00000 ?o?oo 1?000 Phrynosomatidae 0000? 00000 0010? o?oo? 1?010 00000 ?0??0 ??000 Anisolepis grill i 0000? 00000 ?0000 0000? 0110? 01000 ?00?? ??001 Chamaeleolis chamaeleonides 99999 99999 99999 99999 99999 99999 99999 ????1 C. porcus 99999 99999 99999 99999 99999 99999 99999 ????! Enyalius bilineatus 0000? 00010 00000 0010? 0010? 00000 ?V0?? 1?101 E. boulengeri 0000? 00000 00100 0000? 0010? 00000 ?00?? ??100 E. brasiliensis 0000? 00010 01100 0110? 1000? 00000 ?1??? 0?000 C. catenatus 0000? ?0010 ?0100 0?10? 0010? 00000 ?10?? ??ooo E. iheringii 0000? 10010 00100 0010? 1010? 00000 ?001? ??001 E. perditus 0000? 00010 ?0100 0000? 0010? 00000 ?10?? ??ooo E. pictus 0000? ?0??0 ?1100 0???? ?010? 00?00 ?10?? ??000 Phenacosaurus heterodermis 99999 99999 99999 99999 99999 99999 99999 ????1 P. richteri 99999 99999 99999 99999 99999 99999 99999 ????1 Polychrus acutirostris 0001? 00110 ?0000 0100? VV01? 010V0 ?00?? ??001 P. femoralis 99999 99999 99999 99999 ???1? 99999 99999 ???01 P. guttarosus 99999 99999 99999 99999 ???1? 99999 99999 ???01 P. liogaster 99999 99999 99999 99999 ???1? 99999 99999 ???01 P. marmoratus 000V? 00110 00000 0100? V001? 01000 ?0001 0?101 Pristidactylus casuhatiensis 0000? ?0011 00010 00?1? 1110? 01000 ?0??? ???00 P. torquatus ooov? 00110 0V010 ??11? 1110? 01000 ?V001 0?000 Urostrophus vautieri 0000? 00?10 01000 0010? 0100? 01000 ?1??? ???01 Polychrotidae 000?? 00?10 00000 0??0? ?1??? 01000 ?10?1 0?00? Ctenoblepharys adspersus 0010? 00010 00?00 ovoo? 1100? 00000 ?0??? ??ooo Leiocephalus carinatus 0010? 00100 ov?oo ovoo? 11010 00000 ?oo?? ??000 L. green way i 0000? 00000 00?00 0100? 0?010 00000 ?0??? ??0?0 L. inaguae 99999 99999 99999 99999 99999 99999 99999 ????0 L. macropus 001V? ?0000 ?0?00 ?000? 0V010 ovooo ?0?01 o?o?o L. melanochlorus 0010? 00100 01?00 0100? 00010 00000 ?0??? ??0?0 L. pralensis 99999 99999 99999 99999 99999 99999 99999 ???00 L. psammodromus 0000? 00000 00?00 0000? 0?010 00000 ?oo?? ??0?0 L. schreibersi ovvo? 00100 00?00 ovoo? 00010 00000 ?0?11 o???o Microlophus duncanensis ?000? 00000 ?0100 0?0?? 00010 00000 ?0??? ????0 M. grayi 0111? oo??o 00?0? 010?? ?101? 10000 ?0??? ??o?o M. stolzmanni 99999 99999 99999 99999 99999 99999 99999 ???00 M. theresioides ????? 99999 99999 99999 99999 99999 99999 ???oo M. tigris ????? 99999 99999 99999 99999 99999 99999 ???oo Phvmaturus palluma vooo? 00010 ov?oo OVOO? 00000 00000 ?0000 1?000 P. patagonicus patagonicus oovo? ooovo ?0?00 0000? voooo 10000 ?0?10 1?000 P. p. payuniae oovv? oooov 00?00 0000? ovooo 10000 ?0?00 1?000 P. p. somuncurensis oovo? 000?0 00?00 ovoo? voooo 10000 ?0?00 1?000 P. p. zapalensis ovoo? 00000 00?00 ovoo? vvooo 10000 ?o??o 1?000 P. punae vooo? 00010 01?00 0100? ?0000 voooo ?00?0 1?000 P. sp. 99999 99999 99999 99999 99999 99999 99999 ???00 P/esiomicroloph us koepckeoru m ????? 99999 99999 99999 99999 99999 99999 ????0 Stenocercus guentheri ????? 99999 99999 99999 99999 99999 99999 ????0 Uranoscodon superci/iosus ooov? V0010 00000 OVOO? 00V1V 00000 ?0101 o?voo Tropiduridae 00?0? 000?0 00?00 0?00? ??0?0 00000 ?oo?? ??000 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 131 Appendix 3 — Extended— Continued 4 4 4 4 4 1 2 3 4 5 4 4 4 4 5 6789 0 5 5 5 5 5 1 2 3 4 5 5 5 5 5 6 6 7 8 9 0 6666 6 1 2 3 4 5 6 6 667 6 7 8 9 0 77777 1 2 34 5 77778 6789 0 88888 1 2 34 5 8 8 8 8 9 67 89 0 9 9 9 9 9 1 2 34 5 999 678 00000 Vl?10 00000 000?0 00000 00?00 00??? 00?0? ??0?? 000?? 99999 ??? 01000 11110 oooov 000?0 00000 00?00 00??? 01?0? ?10?? 000?? 99999 ??? ??0?0 10010 00000 voo?? ??0?0 00??? ?0??? 00?0? ??0?? 99999 99999 ??? ??0?? ??V10 00000 voo?? ????0 00??0 ?0??? 00?0? ??0?? 0???? 99999 ??? ? ?0 ? ? ??010 00000 000?? ??0?0 00?1? ?1??? 00?0? ??0?? 1???? 99999 ??? ? ?0? ? ? ? ? 10 ?0000 000?0 ?0000 00?10 00??? 01?0? ??1?? 0?0?? 99999 ??? ? ?0 ? ? ??V10 voooo V01?0 00000 00??0 00??? 00?0? ??1?? voo?? 99999 ??? ??0?? ??V10 voooo 101?0 00000 00?10 00??? 00?0? ??1?? 000?? 99999 ??? ?0??? 10010 voooo voo?o 00000 00?10 00??? 00?0? ??1?? voo?? 99999 ??? 00000 ?0??0 ?0000 000?0 00000 00??0 00??? 00?0? ??0?? ?00?? 99999 ??? 001?0 10010 0??00 ?00?1 ??0?0 00??0 10??? 00??? ??0?? 0???? 99999 ??? 0?0?? ??1V1 0???? 000?0 ??0?0 00??0 10??? 00??? ??0?? 0???? 99999 ??? 0?0?? ??001 0???? ?00?1 ??0?0 00??0 10??? 00??? ??0?? 0???? 99999 ??? 001?0 10010 0???0 000?1 ????0 00??0 00??? 00??? ??0?? 0???? 99999 ??? 001?0 10011 0???0 ?00?1 ????0 oo??o 00??? 0???? 99999 0???? 99999 ??? 001?1 00011 o??oo 000?? ????0 00??0 00??? 00??? ??0?? 0???? 99999 ??? 0?1?0 ?0011 0??00 ?00?? ????0 00??0 ?0??? 0???? 99999 0???? 99999 ??? 001?0 10011 0???0 ?00?? ????0 00??0 00??? 01??? ??1?? 0???? 99999 ??? 001?0 ?0011 0???0 ?00?? ????0 oo??o 00??? 0???? 99999 0???? 99999 ??? 001?0 00011 0???? ?00?? ????0 00??0 00??? 00??? ??0?? 0???? 99999 ??? 0?0?? ??01? 0???? ?oo?o ??0?0 oo??o 10??? 00??? ??0?? 0???? 99999 ??? 0?0?? ??01? 0???? ?00?0 ????0 00??0 ?0??? 00??? ??0?? 0???? 99999 ??? 000?0 V0101 00000 000?? ????0 00??0 10??? 00?0? ??0?? 0???? 99999 ??? 0?0?? ??111 000?0 ?00?? ????0 00?10 ?0??? 00?0? ??0?? 0???? 99999 ??? 0?0?? ??011 0000? ?oo?? ????0 oo??o 00??? 00?0? ??0?? 0???? 99999 ??? 0?0?? ??011 000?? ?00?? ????0 00??0 00??? 00?0? ??0?? 0???? 99999 ??? 000?0 V0011 0000? 000?1 ????0 00??0 10??? oo?o? ??0?? 0???? 99999 ??? 001?0 10??? 99999 99999 99999 99999 99999 99999 99999 99999 99999 ??? 001?0 00110 0??00 000?0 ??0?0 00?10 00??? 00??? ??0?? 0?0?? 99999 ??? 001?0 00011 o??oo 000?1 ??0?0 00??0 10??? 00??? ??0?? 0???? 99999 ??? 00??0 ?0??? 00000 000?? ??o?o 00??0 99999 00?0? ??0?? 00??? 99999 ??? 000?1 00001 0??0? 000?0 ??o?o 00?10 00??? 00??? ??0?? 0???? 99999 ??? 00000 10?10 0??00 100?0 ??0?0 00??0 00??? 00??? ??0?? 00??? 99999 ??? o?o?? 99999 99999 99999 ??0?? ?0??? 99999 0???? 99999 0???? 99999 ??? o?o?? ??010 0??0? 000?0 ??0?0 00??0 ?0??? 00??? ?10?? 0???? 99999 ??? 0?000 10V10 0??0? ?00?0 ??0?0 00?10 ?0??? 00??? ??1?? 0???? 99999 ??? 0?000 10010 0??0? voo?o ??0?0 oo??o ?0??? 00??? ?10?? 0???? 99999 ??? o?o?? ??V10 0??0? ?0??0 ????0 00??0 ?0??? 00??? ??0?? 0???? 99999 ??? 0?0?? ?0010 0??0? voo?o ??o?o 00??0 ?0??? 00??? ??0?? 0???? 99999 ??? 00000 00010 0??00 100?0 ??0?0 00?00 00??? 00??? ?10?? 0???? 99999 ??? 001?1 00??? 99999 ????0 ????0 00??0 ?0??? 00??? ??0?? 99999 99999 ??? 001?1 ?0??? 99999 ????1 ????0 00??0 ?0??? 00??? ??0?? 99999 99999 ??? 0?1?? ??010 v??ov ?00?0 ??0?0 00?00 ?0??? 00??? ?oo?? 0???? 99999 ??? 0?1?? ??010 V??01 000?? ????0 00?10 ?0??? 00??? ?10?? 99999 99999 ??? 0?1?? ??01V V??01 000?? ??o?o 00?00 ?0??? 00??? ??0?? 1???? 99999 ??? 10010 10101 0??0? 000?? ??0?0 00??0 00??? 00??? ??0?? 1?0?? 99999 ??? 10010 100V0 0??01 000?? ??0?0 oo??o ?0??? 00??? ??0?? 1?0?? 99999 ??? 10010 100V0 0??01 000?0 ??o?o 00??0 00??? 00??? ??0?? 1?0?? 99999 ??? 10010 100V0 0??01 000?? ??0?0 00??0 00??? 00??? ??0?? 1?0?? 99999 ??? 10010 100V0 0??01 000?? ??0?0 00??0 00??? 00??? ??0?? 1?0?? 99999 ??? 10010 10101 0? ?0? 000?? ??o?o 00??0 00??? 00??? ??0?? 1?0?? 99999 ??? 1?0?? ??ovo 0???1 000?? ????0 00??0 ?0??? 00??? ??0?? 1???? 99999 ??? 0?1?? ??010 v??o? ?00?0 ??o?o 00?00 ?0??? 00??? ?10?? 1???? 99999 ??? 0?1?? ??010 0??0? 001?? ????0 00??0 ?0??? 0???? 99999 99999 99999 ??? 00100 V0001 0??00 ?00?? ??o?o 00??0 00??? 00??? ??0?? 0???? 99999 ??? 000?0 ?00?? o??o? 000?0 ??o?o 00??0 00??? 00??? ?10?? 000?? 99999 ??? 132 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Appendix 4 Step Matrices for Manhattan Distance Frequency Approach Given below are the step matrices employed in the coding of character 31 (number of premaxillary teeth) and in the reanalysis of the Montanucci et al. (1975) allozyme data set using the Manhattan distance frequency approach (Wiens, 1995). Each step matrix is labeled by enzyme and given in the same sequence as presented in Montanucci et al. (1975). Only ten of the 27 original allozyme loci held informative character state changes. The ten included loci are coded as characters 89-98 in the data matrix given in Appendix B. The matrix presented at the bottom of the appendix gives the character “states” that were incorporated into the actual data matrix (Appendix B). 31. Number of Premaxillary Teeth: 1 2 3 4 5 6 7 s 9 10 li 12 13 0 22 35 35 48 65 36 39 14 69 69 38 51 22 0 18 18 70 87 54 60 25 82 87 60 69 35 18 0 0 83 100 72 73 43 100 100 73 82 35 18 0 0 83 100 72 73 43 100 100 73 82 48 70 83 83 0 17 16 15 62 39 39 12 22 65 87 100 100 17 0 33 26 79 25 25 27 25 36 54 72 72 16 33 0 15 46 43 48 16 30 39 60 73 73 15 26 15 0 53 33 33 1 15 14 25 43 43 62 79 46 53 0 72 79 52 61 69 82 100 100 39 25 43 33 72 0 20 32 20 69 87 100 100 39 25 48 33 79 20 0 32 18 38 60 73 73 12 27 16 1 52 32 32 0 14 51 69 82 82 22 25 30 15 61 20 18 14 0 (Note: 1: Gambelia silus, 2: G. wislizenii, 3: G. copei, , 4: G. corona, 5: Crotaphytus reticulatus, 6: C. antiquus, 7: C. col laris, 8: C. nebrius, 9: C. dickersonae, 10: C. grismeri, 1 1 : C. insularis, 12: C. bicinctores, 13: C. vestigium) l 2 3 4 5 6 7 89. H-LDH 0 100 100 100 100 100 100 100 0 100 100 0 0 100 100 100 0 25 100 100 25 100 100 25 0 100 100 0 100 0 100 100 0 0 100 100 0 100 100 0 0 100 100 100 25 0 100 100 0 90. a-GPD 0 0 0 0 0 8 17 0 0 0 0 0 8 17 0 0 0 0 0 8 17 0 0 0 0 0 8 17 0 0 0 0 0 8 17 8 8 8 8 8 0 8 17 17 17 17 17 8 0 91. 6-GPD 0 12 0 12 31 5 12 12 0 12 0 31 7 0 0 12 0 12 31 5 12 12 0 12 0 31 7 0 31 31 31 31 0 31 31 92. ICDs 5 12 0 100 0 0 27 29 0 0 15 0 0 7 0 100 0 100 100 100 100 100 15 0 15 15 5 12 0 100 0 0 27 29 0 0 15 0 0 7 0 0 100 0 0 27 29 0 0 15 0 0 31 31 27 100 27 27 0 17 27 1 1 15 1 1 11 0 7 29 100 29 29 17 0 29 19 11 19 19 7 0 0 100 0 0 27 29 0 0 15 0 0 93. ICDm 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 133 Appendix A — Continued l 2 3 4 5 6 7 11 15 1 1 1 1 0 13 1 1 19 11 19 19 13 0 19 0 15 0 0 11 19 0 94. GOTs 0 69 100 100 69 69 69 69 0 100 100 0 0 0 100 100 0 0 100 100 100 100 100 0 0 100 100 100 69 0 100 100 0 0 0 69 0 100 100 0 0 0 69 0 100 100 0 0 0 95. Pro 0 56 25 4 15 62 58 56 0 81 53 41 6 2 25 81 0 29 40 88 83 4 53 29 0 11 59 55 15 41 40 11 0 48 43 62 6 88 59 48 0 4 58 2 83 55 43 4 0 96. EST1 0 100 0 80 30 71 100 100 0 100 40 70 43 33 0 100 0 80 30 71 100 80 40 80 0 50 9 20 30 70 30 50 0 41 70 71 43 71 9 41 0 29 100 33 100 20 70 29 0 97. Hbpf 0 7 0 8 0 0 50 7 0 7 2 7 7 43 0 7 0 8 0 0 50 8 2 8 0 8 8 42 0 7 0 8 0 0 50 0 7 0 8 0 0 50 50 43 50 42 50 50 0 98. Tr 0 100 100 100 100 100 100 100 0 61 100 61 39 61 100 61 0 100 0 100 0 100 100 100 0 100 100 100 100 61 0 100 0 100 0 100 39 100 100 100 0 100 100 61 0 100 0 100 0 G. wislizenii 1111111111 C. dickersonae 2222222222 C. vestigium 3333333333 C. bicinctores 4444444444 C. nebrius 5555555555 C. collaris 6666666666 C. reticulatus 7777777777 134 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Appendix 5 Character Transformations for Each Stem of the Single Most Parsimonious Tree Characters 1-27, 29-30, 32-67, 69-74, 76-83, 86-88 with a maximum of 24 steps; characters 28, 68, 75, and 84-85 with a maximum of one step; characters 3 1 and 89-98 with a maximum of 100 steps. PAUP does not calculate consistency indices for characters coded using step matrices. Therefore, “n/a” appears in the Cl column for characters 89-98 (allozyme characters cod- ing using Manhattan distances in step matrices). Arrows with double lines indicate unambiguous changes, i.e., those occurring in all optimizations. Arrows with single lines indicate changes that do not occur in all optimizations. (ACCTRAN optimization): Branch Char- acter Steps CI Change HYPANC -*• node A 2 1 0.774 a «-» b 4 24 1.000 a <=> y 6 24 1.000 a <=* y 10 23 0.462 a <=> x 14 24 1.000 a <=> y 26 2 0.453 a <=> c 29 24 1.000 a «=> y 32 16 0.429 a <=> q 40 4 0.774 a <-> e 42 10 0.585 a <=* k 43 24 0.750 a <=> y 45 24 0.247 a <=> y 58 24 0.500 a - y 71 24 0.500 a ** y node A node B 1 5 0.421 a — * f 2 23 0.774 b =5 y 7 2 0.889 a — * c 12 24 0.727 a=?y 15 12 1.000 a =» m 17 19 0.462 e — * x 21 24 1.000 a y 24 20 0.857 C — * w 25 24 1.000 a -* y 26 22 0.462 c =1 y 30 24 1.000 a=£y 36 24 0.632 a=3y 41 20 0.960 a =3 u 44 24 0.436 a=J y 46 24 0.800 a-^y 52 24 1.000 a =? y 55 24 1.000 a -* y 62 24 1.000 a =3 y 82 24 1.000 a -* y node B node C 5 24 1.000 a =5 y 11 24 1.000 a -* y 15 12 1.000 m — 1 ► y 20 24 1.000 y =? a 24 2 0.857 w — » y 31 35 n/a A=£C 32 6 0.444 q =» w 41 3 0.960 u — » X 42 5 0.585 k-*f 53 24 1.000 a -* y 60 24 1.000 a -*■ y 87 24 1.000 a — y Appendix 5 — Continued Branch Char- acter Steps Cl Change node C =» node D 1 19 0.421 f=£y node D =» G. cocci' 7 22 0.889 c =? y 10 1 0.462 x =3 y 12 9 0.727 y p 17 1 0.462 x^y 40 3 0.774 e ^ h 41 1 0.960 x y 42 2 0.585 f=» d node D =» G. wislizenii 17 1 0.462 x — » w 31 18 n/a C=t B 32 1 0.444 w — * V 36 8 0.632 y — * q 45 22 0.247 y=?c node C => G. coronaf 1 5 0.421 f —*■ a 7 2 0.889 c — » a 22 24 0.490 a=G 32 2 0.444 w^ly node B => G. silus 7 1 0.889 c =» d 10 21 0.462 X — * c 32 3 0.444 q -* n 38 20 1.000 a u 40 4 0.774 e — » a 64 24 0.500 a y node A =» node E 8 24 1.000 a -*• y 9 24 1.000 a -* y 10 1 0.462 x =i y 13 24 1.000 a — y 19 24 1.000 a =5 y 22 24 0.490 a =» y 23 24 1.000 a ~ * y 27 24 1.000 a =? y 28 1 1.000 0 — 1 31 38 n/a A — > L 33 24 1.000 a y 34 24 0.800 a ~ * y 35 24 1.000 a -*> y 37 24 1.000 a -► y 40 20 0.774 e=? y 42 12 0.585 k =» w 47 21 0.649 a ^ v 49 24 1.000 a ~ * y 50 24 1.000 a — y 54 24 1.000 a y 56 24 1.000 a ~ * y 57 24 1.000 a=J y 59 24 1.000 a -*• y 61 24 1.000 a =5 y 66 24 1.000 a =! y 68 1 0.800 0^ 1 72 24 1.000 a=J y 76 19 0.558 a =» t 85 1 1.250 0^ 1 88 24 1.000 a=J y 89 100 n/a 1 2 90 8 n/a 1 — ► 6 91 12 n/a 1 -> 2 94 69 n/a 1 -> 2 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 135 Appendix 5 — Continued Appendix 5 — Continued Branch Char- acter Steps CI Change Branch Char- acter Steps CI Change 95 58 n/a 1 -»■ 7 75 1 1.000 2 =8 3 96 71 n/a 1 -» 6 84 1 1.000 0=8 1 98 100 n/a 1 — 3 91 12 n/a 2 — 1 node E = 8 node F 24 1 0.857 c =8 b 95 29 n/a 4 — »• 3 31 1 n/a LUH 96 71 n/a 6 -*• 1 45 16 0.247 y -*■ i node L =8 C. insularis 1 23 0.421 b=8y 51 2 0.889 a =5 e 16 15 1.000 j y 68 1 0.800 1 — » 2 26 18 0.462 g=J y 69 24 1.000 a =8 y 31 18 n/a M =8 K 70 7 0.393 a =5 h 32 10 0.444 k =8 a 77 7 1.000 a =8 h 44 2 0.436 d =8 f 85 1 1.250 1 -> 4 45 2 0.247 c =8 a 86 24 1.000 a=Jy 47 9 0.649 y=£ P node F =8 node G 17 4 0.462 e =8 a 68 1 0.800 3 — ► 4 24 1 0.857 b — >• a 70 11 0.393 1 =8 a 26 1 0.462 c =5 b 83 3 0.465 s — p 32 2 0.444 q =8 s node L =8 C. vestigium 24 1 0.857 a =8 b 42 2 0.585 w =| y 42 1 0.585 y-^x 47 3 0.649 v=8y 70 13 0.393 l — *• y 51 20 0.889 e=J y 83 2 0.465 s =8 u 70 17 0.393 h =8y node J =8 C. grismeri 1 4 0.421 b =8 f 77 17 1.000 h=?y 17 10 0.462 a =8 k 90 8 n/a 6 — 1 26 1 0.462 b ^ a 95 43 n/a 7 -»« 5 31 32 n/a L=8 J node G =3 node FI 2 1 0.774 b — » a 42 4 0.585 y =8 u 76 13 0.558 t=8g 44 2 0.436 d =8 f node H =8 node I 31 1 n/a H =8 L 45 7 0.247 i =8 p 39 24 1.000 a=8y node I =8 C. dickersonae 3 24 0.649 a =8 y 65 24 1.000 a=8y 18 24 1.000 a =8 y 75 1 1.000 0— 1 26 2 0.462 b =8 d 76 6 0.558 g=2a 31 52 n/a L =8 I 78 24 1.000 a=8y 48 24 1.000 a =8 y node I =8 node J 1 1 0.421 a =5 b 63 24 1.000 a ^ y 32 8 0.444 s =8 k 84 1 1.000 0=83 44 3 0.436 a =5 d 85 1 1.250 4 =8 3 71 24 0.500 y=s a 92 100 n/a 1 =8 2 75 1 1.000 1 -*• 2 93 15 n/a 1 =82 89 100 n/a 2^3 95 41 n/a 5 — » 2 94 100 n/a 2 — ► 3 96 43 n/a 6=82 95 11 n/a 5 — ► 4 97 7 n/a 1 =8 2 node J =8 node K 45 4 0.247 i =5 e 98 61 n/a 3 =8 2 70 13 0.393 y — *■ 1 node H =8 C. nebrius 24 3 0.857 a =8 d 83 18 0.465 a =J s 42 1 0.585 y — > x node K =8 C. bicinctores 3 1 0.649 a =5 b 45 8 0.247 i =8 a 41 1 0.960 a =3 b 47 2 0.649 y — » w 51 3 0.889 y=£ v 80 24 0.500 y =8 a 67 24 0.500 a =5 y 81 23 1.000 a =8 x 89 25 n/a 3 — » 4 83 18 0.465 a =8 s 96 9 n/a 6 =84 84 1 1.000 0=82 97 8 n/a 1 =8 4 85 1 1.250 4=8 2 98 100 n/a 3 =84 91 31 n/a 2=85 node K =8 node L 16 9 1.000 a=8j 92 27 n/a 1 =8 5 26 5 0.462 b=8g 93 1 1 n/a 1 =8 5 31 14 n/a L=8M 96 41 n/a 6 =8 5 45 2 0.247 e =8 c node H =8 C. collaris 17 17 0.462 e — » v 68 1 0.800 2 — > 3 22 1 0.490 y — *■ x 73 24 0.500 a=8y 31 15 n/a H =8 G 136 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Appendix 5 — Continued Appendix 5 — Continued Branch Char- acter Steps CI Change Branch Char- acter Steps CI Change 32 6 0.444 q =3 k 26 22 0.462 c=3y 42 4 0.585 w — * s 30 24 1.000 a =£ y 45 8 0.247 i => a 36 16 0.632 a =3 q 46 6 0.800 a ~ g 41 20 0.960 a ^ u 47 2 0.649 v — ► t 43 8 0.750 q ->y 74 24 1.000 y =i a 44 24 0.436 a=J y 80 24 0.500 y =3a 46 24 0.800 a -* y 83 2 0.465 a c 52 24 1.000 a y 91 7 n/a 2 — ► 6 55 24 1.000 a -> y 92 29 n/a 1 =£ 6 58 24 0.500 a -*■ y 93 19 n/a 1 =3 6 62 24 1.000 a =£ y 95 4 n/a 7=J 6 82 24 1.000 a ~ y 98 100 n/a 3 6 node B =» node C 5 24 1.000 a =3 y node F =» C. antiquus 2 5 0.774 b=J g 20 24 1.000 y =?a 3 12 0.649 a^m 25 24 1.000 a — y 10 6 0.462 y s 31 35 n/a A=U 26 1 0.462 b =* a 32 8 0.444 n =5 v 31 26 n/a H =£ F node C =» node D 1 24 0.421 a =! y 34 6 0.800 y =s s 7 2 0.889 a — » c 36 6 0.632 a=£g 10 21 0.462 c * X 44 24 0.436 a =S y 11 24 1.000 a->y 45 4 0.247 i — * m 15 12 1.000 m — 5 64 24 0.500 a =5 y 24 2 0.857 w — ► y 67 24 0.500 a y 40 4 0.774 a — » e 68 1 0.800 2 — » 1 41 3 0.960 u * X 76 5 0.558 t=*y 53 24 1.000 a -► y 79 24 0.500 a=£y 60 24 1.000 a — y node E =£ C. reticulatus 2 1 0.774 b — » a node D => <7. cope/ 7 22 0.889 c=iy 28 1 1.000 1 -» 2 10 1 0.462 x =5 y 31 12 n/a L=?E 12 9 0.727 y =5 p 43 8 0.750 y — Q 17 2 0.462 W=J> 58 24 0.500 y — * a 32 1 0.444 V — » w 73 24 0.500 a — y 36 8 0.632 q -* y 79 24 0.500 a =5 y 40 3 0.774 e =» h 89 100 n/a 2 — > 7 41 1 0.960 x =? y 90 8 n/a 6 7 42 2 0.585 f=J d 96 29 n/a 6 =? 7 45 16 0.247 i -»■ y 97 50 n/a 1 =J 7 node D :=* G. wislizenii 31 18 n/a C => I (DELTRAN optimization): 45 6 0.247 i c Char- 87 24 1.000 a — » y Branch acter Steps CI Change node C =3 G. coronal 22 24 0.490 a =s y HYPANC =i node A 4 24 1.000 a ~ y 32 3 0.444 v =£ y 6 24 1.000 a <=> y node B =» G. silus 1 5 0.421 a — » f 10 2 0.462 a *-* c 7 3 0.889 a ^ d 14 24 1.000 a => y 17 1 0.462 w — » > 26 2 0.462 a - c 36 8 0.632 q y 29 24 1.000 a <=> y 38 20 1.000 a ^ u 32 13 0.444 a <=> n 42 5 0.585 f-» k 42 5 0.585 a <=> f 45 16 0.247 i -► y 43 16 0.750 a - q 64 24 0.500 a =5 y 45 8 0.247 a «-» i node A =* node E 8 24 1.000 a -* y 71 24 0.500 a *=> y 9 24 1.000 a -*• y node A node B 2 24 0.774 a =3 y 10 22 0.462 c^y 12 24 0.727 a =5 y 13 24 1.000 a — y 15 12 1.000 a=*m 19 24 1.000 a=J y 17 18 0.462 e — * w 22 23 0.490 a =» x 21 24 1.000 a - > y 23 24 1.000 a — y 24 20 0.857 c — * w 27 24 1.000 a=J y 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 137 Appendix 5 — Continued Appendix 5 — Continued Branch Char- acter Steps CI Change 31 38 n/a A — » L 33 24 1.000 a=£y 34 24 0.800 a -*y 35 24 1.000 a -* y 37 24 1.000 a — y 40 24 0.774 a =3 y 42 13 0.585 f =3 s 47 19 0.649 a =5 t 49 24 1.000 a -> y 50 24 1.000 a ~ y 54 24 1.000 a =3y 56 24 1.000 a -> y 57 24 1.000 a=3y 59 24 1.000 a -*> y 61 24 1.000 a =3 y 66 24 1.000 a =3 y 72 24 1.000 a=3y 76 19 0.558 a =3 1 88 24 1.000 a =3 y 90 8 n/a 1 -» 6 91 5 n/a 1 — 6 94 69 n/a 1 -* 2 95 56 n/a 1 -> 2 96 71 n/a 1 -» 6 98 100 n/a 1 -► 3 node E =3 node F 24 1 0.857 c=3b 28 1 1.000 0^ 1 31 1 n/a L =3 H 43 8 0.750 q — y 51 4 0.889 a ^ e 58 24 0.500 a -» y 68 1 0.800 0 — ► 2 69 24 1.000 a=£y 70 7 0.393 a =3 h 77 7 1.000 a =3 h 85 1 1.250 0 — * 4 86 24 1.000 a =3 y 89 100 n/a 1 — 2 node F =3 node G 17 4 0.462 e4a 22 1 0.490 x-*y 26 1 0.462 c =3 b 32 5 0.444 n =3 s 42 5 0.585 S =3 X 47 3 0.649 t =3 w 51 20 0.889 e =S y 70 17 0.393 h y 77 17 1.000 h =J y node G =3 node H 76 13 0.558 t=5g 90 8 n/a 6 — 1 node H =3 node I 24 1 0.857 b — » a 31 1 n/a H =3 L 39 24 1.000 a =3 y 47 2 0.649 w — > y 65 24 1.000 a =5 y 76 6 0.558 g=J a 78 24 1.000 a=£y node I =3 node J 1 1 0.421 a =3 b 32 8 0.444 s =3 k Branch Char- acter Steps CI Change 44 3 0.436 a =3 d 71 24 0.500 y =? a 75 1 1.000 0 — » 2 95 41 n/a 2 -> 5 node J node K 45 4 0.247 i =3 e 83 15 0.465 a =3 p 89 100 n/a 2 — »• 3 94 100 n/a 2 — > 3 95 1 1 n/a 5 — *■ 4 node K =3 C. bicinctores 3 1 0.649 a =3 b 41 1 0.960 a =3 b 42 1 0.585 x * y 51 3 0.889 y v 67 24 0.500 a =3 y 70 13 0.393 y-i 83 3 0.465 p -» s 89 25 n/a 3 — ► 4 91 7 n/a 6 — » 4 96 9 n/a 6=34 97 8 n/a 1 =3 4 98 100 n/a 3 =3 4 node K =3 node L 16 9 1.000 a =5 j 26 5 0.462 b=J g 31 14 n/a L =5 M 45 2 0.247 e =3 c 73 24 0.500 a =3 y 75 1 1.000 2=5 3 84 1 1.000 0=5 1 node L4C. insularis 1 23 0.421 b=5y 16 15 1.000 J =5y 26 18 0.462 g=f y 31 18 n/a M =3 K 32 10 0.444 k =3 a 42 1 0.585 x * y 44 2 0.436 d =3 f 45 2 0.247 c =3 a 47 9 0.649 y=3p 68 1 0.800 2 — » 4 70 24 0.393 y=5a node L =3 C. vestigium 24 1 0.857 a =3 b 68 1 0.800 2 — ► 3 83 5 0.465 P =5 u 91 5 n/a 6 — > 3 95 29 n/a 4 — ► 3 96 71 n/a 6 — ► 3 node J =3 C. grismeri 1 4 0.421 b =5 f 17 10 0.462 a =5 k 26 1 0.462 b =3 a 31 32 n/a L=3 J 42 3 0.585 x =5 u 44 2 0.436 d=5f 45 7 0.247 i =3 p node I =3 C. dickersonae 3 24 0.649 a=3y 18 24 1.000 a=5y 26 2 0.462 b=3d 31 52 n/a L=5 1 42 1 0.585 x * y 48 24 1.000 a=5y 138 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Appendix 5 — Continued Appendix 5 — Continued Char- Branch acter Steps CI Change 63 24 1.000 a =5 y 75 1 1.000 0^ 1 84 1 1.000 0=5 3 85 1 1.250 4 =S 3 91 7 n/a 6 — » 2 92 100 n/a 1 =1 2 93 15 n/a 1 =5 2 96 43 n/a 6 =5 2 97 7 n/a 1 =5 2 98 61 n/a 3 =5 2 node H ^ C. nebrius 24 2 0.857 b=5 d 45 8 0.247 i =5 a 80 24 0.500 y =^a 81 23 1.000 a ^ x 83 18 0.465 a =5 s 84 1 1.000 0=5 2 85 1 1.250 4 =5 2 91 31 n/a 6 =5 5 92 27 n/a 1 =5 5 93 11 n/a 1 =5 5 95 41 n/a 2 — ► 5 96 41 n/a 6=5 5 node H =5 C. collaris 2 1 0.774 a — » b 17 17 0.462 e — »• v 31 15 n/a H =5 G 32 3 0.444 n ^ k 45 8 0.247 i =5 a 46 6 0.800 a^g 74 24 1.000 y =* a 80 24 0.500 y =5 a 83 2 0.465 a =5 c 92 29 n/a 1 =5 6 93 19 n/a 1 =5 6 95 6 n/a 2=56 98 100 n/a 3 =5 6 node F =5 C. antiquus 2 6 0.774 a =5 g 3 12 0.649 a =5 m 10 6 0.462 y=5s 24 1 0.857 b — ► a 26 1 0.462 b =5 a 31 26 n/a H =5 F 34 6 0.800 y =5 s 36 6 0.632 a =5 g 42 1 0.585 x — y 44 24 0.436 a =5 y 45 4 0.247 i — * m 47 2 0.649 w — * y 64 24 0.500 a =5 y 67 24 0.500 a =5 y 68 1 0.800 2 -> 1 76 5 0.558 t =5 y 79 24 0.500 a =5 y node E =5 C. reticulatus 22 1 0.490 x — y 28 1 1.000 0 — >• 2 31 12 n/a L =5 E 32 3 0.444 n — q 42 4 0.585 s — » w 45 16 0.247 i — y Char- Branch acter Steps CI Change 47 2 0.649 t -> V 68 1 0.800 0^ 1 73 24 0.500 a-^y 79 24 0.500 a =5 y 85 1 1.250 0^ 1 89 100 n/a 1 -»• 7 90 8 n/a 6=5 7 91 7 n/a 6 — ► 7 95 2 n/a 2 — ► 7 96 29 n/a 6 =5 7 97 50 n/a 1 =5 7 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 139 Appendix 6 List of Character State Changes by Character Characters 1-27, 29-30, 32-67, 69-74, 76-83, 86-88 with a maximum of 24 steps; characters 28, 68, 75, and 84-85 with a maximum of one step; characters 3 1 and 89-98 with a maximum of 100 steps. PAUP does not calculate consistency indices for characters coded using step matrices. Therefore, “n/a” appears in the Cl column for characters 89-98 (allozyme characters cod- ing using Manhattan distances in step matrices). Arrows with double lines indicate unambiguous changes, i.e., those occurring in all optimizations. Arrows with single lines indicate changes that do not occur in all optimizations. (ACCTRAN optimization): Character change lists: Character CI Steps Changes 1 0.421 5 node Aa^f node B 19 node C f =4 y node D 5 node C f — * a G. corona\ 1 node I a =4 b node J 23 node L b =5 y C. insularis 4 node J b =4 f C. grismeri 2 0.774 1 node A b - a HYPANC 23 node Ab4y node B 1 node Gb->a node H 5 node G b g C. antiquus 1 node E b — » a C. reticulatus 3 0.649 1 node K a =4 b C. bicinctores 24 node I a 4 y C. dickersonae 12 node G a =4 m C. antiquus 4 1.000 24 node Ay»a HYPANC 5 1.000 24 node B a =4 y node C 6 1.000 24 node Ay«a HYPANC 7 0.889 2 node A a — » c node B 22 node D c =4 y G. copei 2 node C c — * a G. coronal 1 node B c =4 d G. silus 8 1.000 24 node Aa-^y node E 9 1.000 24 node Aa^y node E 10 0.462 23 node Ax a HYPANC 1 node D x 4 y G. copei 21 node B x — » c G. silus 1 node Ax4y node E 6 node G y =5 s C. antiquus 1 1 1.000 24 node Ba-»y node C 12 0.727 24 node A a =5 y node B 9 node Dy4p G. copei 13 1.000 24 node A a — » y node E 14 1.000 24 node A y <=> a HYPANC 15 1.000 12 node AaTm node B 12 node B m — » y node C 16 1.000 9 node K a =4 j node L 15 node L j =4 y C. insularis 17 0.462 19 node Ae^x node B 1 node D x 4 y G. copei 1 node D x -* w G. wislizenii 4 node F e =4 a node G 10 node J a =5 k C. grismeri 17 node F e — ► v C. collaris 18 1.000 24 node I a =» y C. dickersonae 19 1.000 24 node A a =4 y node E 20 1.000 24 node By^a node C Appendix 6 — Continued Character change lists: Character CI Steps Changes 21 1.000 24 node Aa-*y node B 22 0.490 24 node Ca=*yG. coronaf 24 node A a =4 y node E 1 node F y — ♦ x C. collaris 23 1.000 24 node A a — *■ y node E 24 0.857 20 node Ac-*w node B 2 node Bw-»y node C 1 node Ec=»b node F 1 node Fb-*a node G 1 node L a =4 b C. vestigium 3 node H a =4 d C. nebrius 25 1.000 24 node Aa-^y node B 26 0.462 2 node Ac « a HYPANC 22 node Ac4y node B 1 node Fc=»b node G 5 node Kb4g node L 18 node L g =4 y C. insularis 1 node J b => a C. grismeri 2 node I b =4 d C. dickersonae 1 node G b =4 a C. antiquus 27 1.000 24 node A a =5 y node E 28 1.000 1 node A 0 — » 1 node E 1 node E 1 — » 2 C. reticulatus 29 1.000 24 node Ay « a HYPANC 30 1.000 24 node A a => y node B 31 n/a 35 node B A =4 C node C 18 node D C =4 B G. wislizenii 38 node A A — * L node E 1 node E L =5 H node F 1 node H H =+ L node I 14 node K L => M node L 18 node L M => K C insularis 32 node J L =* J C. grismeri 52 node 1 L => I C. dickersonae 26 node G H =* F C. antiquus 15 node F H =» G C. collaris 12 node E L 4 E C. reticulatus 32 0.444 16 node A q <=> a HYPANC 6 node Bq4w node C 1 node D w — » v G. wislizenii 2 node Cw4y G. coronal 3 node B q — » n G. silus 2 node F q =4 s node G 8 node I s =4 k node J 10 node L k 4 a C. insularis 6 node F q =4 k C. collaris 33 1.000 24 node Aa4y node E 34 0.800 24 node A a — * y node E 6 node G y =4 s C. antiquus 35 1.000 24 node A a — » y node E 36 0.632 24 node Aa4y node B 8 node D y — » q G. wislizenii 6 node G a 4 g C. antiquus 37 1.000 24 node Aa-»y node E 38 1.000 20 node B a 4 u G. Silus 39 1.000 24 node H a =4 y node I 40 0.774 4 node A e - a HYPANC 3 node D e =4 h G. copei 140 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Appendix 6 — Continued Appendix 6 — Continued Character change lists: Character Cl Steps Changes Character change lists: Character Cl Steps Changes 4 node Be^aG. si/us 64 0.500 24 node Ba=»yG silus 20 node Ae4y node E 24 node G a =» y C. antiquus 41 0.960 20 node A a =» u node B 65 1.000 24 node H a => y node I 3 node B u — » x node C 66 1.000 24 node A a =1 y node E 1 node D x =» y G. copei 67 0.500 24 node K a =5 y C. bicinctores 1 node K a =1 b C. bici net ores 24 node G a 4 y C. antiquus 42 0.585 10 node Ak«a HYPANC 68 0.800 1 node A 0 — * 1 node E 5 node B k — » f node C 1 node E 1 — ► 2 node F 2 node D f 4 d G. copei 1 node K2-»3 node L 12 node A k =£ w node E 1 node L 3 — » 4 C. insularis 2 node Fw4y node G 1 node G 2 —* 1 C. antiquus 1 node L y — » x C. vestigium 69 1.000 24 node E a =1 y node F 4 node J y =1 u C. grismeri 70 0.393 7 node Ea4h node F 1 node H y — * x C. nebrius 17 node F h =» y node G 4 node F w — > s C. collaris 13 node J y — *• 1 node K 43 0.750 24 node Ay«a HYPANC 11 node L 1 =» a C. insularis 8 node E y — ► q C. reticulatus 13 node L 1 — » y C. vestigium 44 0.436 24 node A a =5 y node B 71 0.500 24 node Ay»a HYPANC 3 node I a ^ d node J 24 node I y => a node J 2 node L d =» f C. insularis 72 1.000 24 node Aa=>y node E 2 node J d =4 f C. grismeri 73 0.500 24 node Ka4y node L 24 node G a =» y C. antiquus 24 node E a — » y C. reticulatus 45 0.247 24 node Ay w a HYPANC 74 1.000 24 node F y => a C. collaris 22 node D y => c G. wislizenii 75 1.000 1 node H 0 — » 1 node I 16 node E y — * i node F 1 node I 1 — *• 2 node J 4 node J i =5 e node K 1 node K2=>3 node L 2 node Ke4c node L 76 0.558 19 node Aa=»t node E 2 node L c 4 a C. insularis 13 node G t=»g node H 7 node J i =» p C. grismeri 6 node H g^a node I 8 node H i => a C. nebrius 5 node G t =5 y C. antiquus 4 node G i — * m C. antiquus 77 1.000 7 node Ea=»h node F 8 node F i =» a C. collaris 17 node F h ^4 y node G 46 0.800 24 node Aa-*y node B 78 1.000 24 node Ha=>y node I 6 node F a —* g C. collaris 79 0.500 24 node G a 4 y C. antiquus 47 0.649 21 node Aa^v node E 24 node E a =5 y C. reticulatus 3 node F v =» y node G 80 0.500 24 node H y =» a C. nebrius 9 node L y =+ p C. insularis 24 node F y =» a C. collaris 2 node Hy^wC. nebrius 81 1.000 23 node H a =» x C. nebrius 2 node F v — » t C. collaris 82 1.000 24 node Aa->y node B 48 1.000 24 node I a => y C. dickersonae 83 0.465 18 node J a=>s node K 49 1.000 24 node Aa^y node E 3 node L s — » p C. insularis 50 1.000 24 node Aa->y node E 2 node L s => u C. vestigium 51 0.889 4 node Ea4e node F 18 node H a =5 s C. nebrius 20 node Fe4y node G 2 node F a =» c C. collaris 3 node K y => v C. bicinctores 84 1.000 1 node K 0 =» 1 node L 52 1.000 24 node A a =4 y node B 1 node 1 0 3 C. dickersonae 53 1.000 24 node Ba-^y node C 1 node H 0 =5 2 C. nebrius 54 1.000 24 node A a =4 y node E 85 1.250 1 node A 0 — * 1 node E 55 1.000 24 node Aa-*y node B 1 node E 1 — » 4 node F 56 1.000 24 node A a — » y node E 1 node 1 4 =4 3 C. dickersonae 57 1.000 24 node A a =4 y node E 1 node H 4 2 C. nebrius 58 0.500 24 node A y -» a HYPANC 86 1.000 24 node E a => y node F 24 node E y — * a C. reticulatus 87 1.000 24 node Ba^y node C 59 1.000 24 node A a — *■ y node E 88 1.000 24 node A a =5 y node E 60 1.000 24 node Ba->y node C 89 n/a 100 node A 1 — » 2 node E 61 1.000 24 node Aa4y node E 100 node I 2 — » 3 node J 62 1.000 24 node Aa4y node B 25 node K 3 — » 4 C. bicinctores 63 1.000 24 node I a 4 y C. dickersonae 100 node E 2 — » 7 C. reticulatus 1996 McGUIRE— SYSTEM ATICS OF CROTAPHYTID LIZARDS 141 Appendix 6 — Continued Appendix 6 — Continued Character change lists: Character Cl Steps Changes Character change lists: Character Cl Steps Changes 90 n/a 8 node A 1 — » 6 node E 1 node D x =4 y G. copei 8 node F 6 — » 1 node G 22 node A c 4 y node E 8 node E 6 ^ 7 C. reticualtus 6 node G y =4 s C. antiquus 91 n/a 12 node A 1 — » 2 node E 1 1 1.000 24 node C a — * y node D 12 node K 2 — ► 1 node L 12 0.727 24 node Aa4y node B 31 node H 2 4 5 C. nebrius 9 node D y 4 p G. copei 7 node F 2 — » 6 C. collaris 13 1.000 24 node A a — * y node E 92 n/a 100 node I 1 =» 2 C. dickersonae 14 1.000 24 node Ay«a HYPANC 27 node H 1 =4 5 C. nebrius 15 1.000 12 node Aa4m node B 29 node F 1 =» 6 C. collaris 12 node Cm-»y node D 93 n/a 15 node I 1 4 2 C. dickersonae 16 1.000 9 node Ka4j node L 11 node H 1 4 5 C. nebrius 15 node L j =4 y C. insularis 19 node F 1 =» 6 C. collaris 17 0.462 18 node Ae-»w node B 94 n/a 69 node A 1 — » 2 node E 2 node Dw4yC. copei 100 node I 2 — ► 3 node J 1 node Bw-^x & silus 95 n/a 58 node A 1 — ► 7 node E 4 node F e ^ a node G 43 node F 7 — » 5 node G 10 node J a 4 k C. grismeri 1 1 node I 5 — » 4 node J 17 node F e — * v C. collaris 29 node K4^3 node L 18 1.000 24 node I a 4 y C. dickersonae 41 node I 5 — » 2 C. dickersonae 19 1.000 24 node A a =4 y node E 4 node F 7 =4 6 C. collaris 20 1.000 24 node By4a node C 96 n/a 71 node A 1 — » 6 node E 21 1.000 24 node Aa-*y node B 9 node K 6 4 4 C. bicinctores 22 0.490 24 node C a 4 y G. coronaf 71 node K 6 — » 1 node L 23 node A a =» x node E 43 node I 6 =» 2 C. dickersonae 1 node Fx-»y node G 41 node H 6 4 5 C. nebrius 1 node E x — » y C. reticulatus 29 node E 6 =1 7 C. reticulatus 23 1.000 24 node Aa-*y node E 97 n/a 8 node K 1 4 4 C. bicinctores 24 0.857 20 node Ac^w node B 7 node I 1 42 C. dickersonae 2 node C w — * y node D 50 node E 1 =4 7 C. reticulatus 1 node E c =4 b node F 98 n/a 100 node A 1 — » 3 node E 1 node Hb->a node I 100 node K 3 4 4 C. bicinctores 1 node L a 4 b C. vestigium 61 node I 3 =» 2 C. dickersonae 2 node H b 4 d C. nebrius 100 node F 3 4 6 C. collaris 1 node Gb^aC. antiquus (DELTRAN optimization): 25 1.000 24 node B a — >■ y node C Character Cl Steps Changes 26 0.462 2 node Ac»a HYPANC 1 0.421 24 node C a =4 y node D 22 node Ac4y node B 5 node B a — » f G. silus 1 node Fc4b node G 1 node I a =» b node J 5 node K b =4 g node L 23 node L b 4 y C. insularis 18 node L g =4 y C. insularis 4 node J b =4 f C. grismeri 1 node J b 4 a C. grismeri 2 0.774 24 node Aa4y node B 2 node I b 4 d C. dickersonae 6 node G a 4 g C. antiquus 1 node G b =4 a C. antiquus 1 node F a — * b C. collaris 27 1.000 24 node A a =4 y node E 3 0.649 1 node K a 4 b C. bicinctores 28 1.000 1 node E 0 — » 1 node F 24 node I a =4 y C. dickersonae 1 node E 0 —* 2 C. reticulatus 12 node G a =4 m C. antiquus 29 1.000 24 node Ay « a HYPANC 4 1.000 24 node A y - a HYPANC 30 1.000 24 node Aa4y node B 5 1.000 24 node Ba4y node C 31 n/a 35 node B A =4 C node C 6 1.000 24 node Ay « a HYPANC 18 node D C 4 B G. wislizenii 7 0.889 2 node Ca-*c node D 38 node A A — * L node E 22 node D c 4 y G. copei 1 node E L ^4 H node F 3 node B a 4 d G. silus 1 node H H 4 L node I 8 1.000 24 node Aa-»y node E 14 node K L 4 M node L 9 1.000 24 node Aa->y node E 18 node L M 4 K C. insularis 10 0.462 2 node Ac«a HYPANC 32 node J L 4 J C. grismeri 21 node C c — * x node D 52 node I L =4 I C. dickersonae 26 node G H 4 F C. antiquus 142 BULLETIN CARNEGIE MUSEUM OF NATURAL HISTORY NO. 32 Appendix 6 — Continued Appendix 6 — Continued Character change lists: Character Cl Steps Changes 15 node FH^GC. collaris 12 node E L =1 E C. reticulatus 32 0.444 13 node An«a HYPANC 8 node Bn4v node C 1 node D v — > w G. copei 3 node Cv=*y6. corona^ 5 node F n =» s node G 8 node I s k node J 10 node L k =5 a C. insularis 3 node F n ^ k C. collaris 3 node E n — » q C. reticulatus 33 1.000 24 node A a =* y node E 34 0.800 24 node Aa-^y node E 6 node G y ^ s C. antiquus 35 1.000 24 node Aa->y node E 36 0.632 16 node Aa=»q node B 8 node D q -» y G. copei 8 node B q — *• y G. silus 6 node G a 4 g C. antiquus 37 1.000 24 node A a — » y node E 38 1.000 20 node B a =£ u G. silus 39 1.000 24 node H a =5 y node I 40 0.774 4 node Ca-*e node D 3 node D e h G. copei 24 node Aa4 y node E 41 0.960 20 node Aa^u node B 3 node Cu-*x node D 1 node D x ^ y 6. copei 1 node K a ^ b C. bicinctores 42 0.585 5 node Af«a HYPANC 2 node D f =» d G. copei 5 node B f — * k G. silus 13 node Af=»s node E 5 node Fs4x node G 1 node K x — * y C. bicinctores 1 node L x — » y C. insularis 3 node J x =» u C. grismeri 1 node I x — * y C. dickersonae 1 node G x — ► y C. antiquus 4 node E s — w C. reticulatus 43 0.750 16 node A q - a HYPANC 8 node Aq->y node B 8 node Eq->y node F 44 0.436 24 node A a =5 y node B 3 node I a =1 d node J 2 node L d =» f C. insularis 2 node J d =» f C. grismeri 24 node G a ^ y C. antiquus 45 0.247 8 node A i " a HYPANC 16 node D i — *■ y G. copei 6 noe D i =5 c G. wislizenii 16 node B i — > y G. silus 4 node J i =3 e node K 2 node K e =1 c node L 2 node L c =5 a C. insularis 7 node J i =2 p C. grismeri 8 noe H i =* a C. nebrius 4 node G i — » m C. antiquus Character change lists: Character Cl Steps Changes 8 node F i => a C. collaris 16 node E i — » y C. reticulatus 46 0.800 24 node Aa-*y node B 6 node F a — » g C. collaris 47 0.649 19 node A a =» t node E 3 node Ft=»w node G 2 node H w — » y node I 9 node L y => p C. insularis 2 node Gw^yC. antiquus 2 node E t — * v C. reticulatus 48 1.000 24 node I a 4 y C. dickersonae 49 1.000 24 node Aa->y node E 50 1.000 24 node Aa->y node E 51 0.889 4 node Ea=>e node F 20 node F e =5 y node G 3 node Ky=>vC. bicinctores 52 1.000 24 node A a =* y node B 53 1.000 24 node C a — » y node D 54 1.000 24 node Aa=>y node E 55 1.000 24 node A a — > y node B 56 1.000 24 node Aa-»y node E 57 1.000 24 node Aa=>y node E 58 0.500 24 node Aa^y node B 24 node Ea-*y node F 59 1.000 24 node Aa-*y node E 60 1.000 24 node C a — » y node D 61 1.000 24 node Aa=*y node E 62 1.000 24 node A a =» y node B 63 1.000 24 node I a => y C. dickersonae 64 0.500 24 node B a =» y G. silus 24 node G a =» y C. antiquus 65 1.000 24 node H a =5 y node I 66 1.000 24 node A a => y node E 67 0.500 24 node K a y C. bicinctores 24 node G a ^ y C. antiquus 68 0.800 1 node E 0 — * 2 node F 1 node L 2 — * 4 C. insularis 1 node L 2 — » 3 C. vestigium 1 node G 2 — *T C. antiquus 1 node E 0 — » 1 C. reticulatus 69 1.000 24 node Ea=»y node F 70 0.393 7 node Ea=>h node F 17 node Fh=>y node G 13 node K y — ► 1 C. bicinctores 24 node L y => a C. insularis 71 0.500 24 node Ay ^ a HYPANC 24 node I y =» a node J 72 1.000 24 node Aa=»y node E 73 0.500 24 node Ka^»y node L 24 node E a — » y C. reticulatus 74 1.000 24 node F y =* a C. collaris 75 1.000 1 node I 0 — *• 2 node J 1 node K2=»3 node L 1 node I 0 — > 1 C. dickersonae 76 0.558 19 node A a => t node E 13 node Gt=»g node H 6 node H g =» a node I 5 node G t =* y C. antiquus 1996 McGUIRE — SYSTEM ATICS OF CROTAPHYTID LIZARDS 143 Appendix 6 — Continued Appendix 6 — Continued Character change lists: Character Cl Steps Changes Character change lists: Character Cl Steps Changes 77 1.000 7 node Ea4h node F 29 node L 4 — » 3 C. vestigium 17 node Fh=>y node G 41 node H 2 -> 5 C. nebrius 78 1.000 24 node Ha4y node I 6 node F 2 4 6 C. collaris 79 0.500 24 node G a 4 y C. antiquus 2 node E 2 — » 7 C. reticulatus 24 node E a 4 y C. reticulatus 96 n/a 71 node A 1 — * 6 node E 80 0.500 24 node H y 4 a C. nebrius 9 node K 6 4 4 C. bicinctores 24 node F y 4 a C. collaris 71 node L 6 — ► 3 C. vestigium 81 1.000 23 node H a 4 x C. nebrius 43 node I 6 =4 2 C. dickersonae 82 1.000 24 node Aa-*y node B 41 node H 6 4 5 C. nebrius 83 0.465 15 node J a =5 p node K 29 node E 6 =4 7 C. reticulatus 3 node K p — » s C. bici net ores 97 n/a 8 node K 1 4 4 C. bicinctores 5 node L p 4 u C. vestigium 7 node 114 2 C. dickersonae 18 node H a 4 s C. nebrius 50 node E 1 4 7 C. reticulatus 2 node F a 4 c C. collaris 98 n/a 100 node A 1 — * 3 node E 84 1.000 1 node K04 1 node L 100 node K 3 4 4 C. bicinctores 1 node I 0 4 3 C. dickersonae 61 node I 3 4 2 C. dickersonae 1 node H 0 4 2 C. nebrius 100 node F 3 4 6 C. collaris 85 1.250 1 node E 0 — » 4 node F 1 node I 4 4 3 C. dickersonae 1 node FI 4 4 2 C. nebrius 1 node E 0 — * 1 C. reticulatus 86 1.000 24 node E a =4 y node F 87 1.000 24 node D a — » y G. wislizenii Appendix 7 88 1.000 24 node Aa4y node E Scleral Ossicle Data 89 n/a 100 node E 1 — > 2 node F Scleral ossicle numbers and patterns of overlap were assessed 100 node J 2 — »• 3 node K in the listed specimens. All crotaphytids examined match the 25 node K 3 — » 4 C. bicinctores apparently plesiomorphic iguanian condition in which ossicles 100 node E 1 -* 7 C. reticulatus 1 , 6, and 8 are positive and 4, 7, and 1 0 are negative (Underwood, 90 n/a 8 node A 1 — » 6 node E 1970; de Queiroz, 1982). Only one set of scleral ossicles (one eye) 8 node G 6 — *• 1 node FI was examined in the specimens followed by asterisks. 8 node E 6 4 7 C. reticulatus 91 n/a 5 node A 1 — » 6 node E Crotaphytus : 7 node K 6 — > 4 C. bicinctores bicinctores REE 2931, 2932, 2934 5 node L 6 — » 3 C. vestigium antiquus TNHC 53155*, 53156, 53159 7 node I 6 — * 2 C. dickersonae collaris REE 2875, 2944, 2952* 31 node H 6 =4 5 C. nebrius dickersonae REE 2777, 2904, 2905 7 node E 6 —* 7 C. reticulatus grismeri MZFC 6648, 6649, 6650* 92 n/a 100 node I 1 =4 2 C. dickersonae insularis REE 2794- ■2796 27 node H 1 4 5 C. nebrius nebrius REE 2937, 2941, 2942, 2943 29 node F 1 =4 6 C. collaris reticulatus REE 2910, 2911, 2913* 93 n/a 15 node I 1 =4 2 C. dickersonae vestigium REE 2820, 2825, 2826 11 node H 1 =4 5 C. nebrius Gambelia: 19 node F 1 4 6 C. collaris copei REE 2798, 2802. 2804 94 n/a 69 node A 1 — » 2 node E silus CAS 22713 1, 22742*, 141328* 100 node J 2 — *■ 3 node K wislizenii REE 2789* , 2790. 2791. 2792. 2916'. 2917. 95 n/a 56 node A 1 — * 2 node E 29182, 2919. 2920 41 node 12 — *5 node 1 Ossicles 1 and 14 of the right scleral ring are partially overlapping. 1 1 node J 5 — *• 4 node K 2 Ossicles 1 and 1 4 of the right scleral ring and 1 3 and 14 of the'left ring are partially overlapping. ISSN 0145-9058 BULLETIN OF CARNEGIE MUSEUM OF NATURAL HISTORY TAXONOMY AND EVOLUTION OF LATE CRETACEOUS LIZARDS (REPTILIA: SQUAMATA) FROM WESTERN CANADA GAO KEQIN and RICHARD C. FOX DUMBER 33 PITTSBURGH, 1996