Brigham Young University Science Bulletin, Biological Series

Volume 19 Number 4 Article 1

4-1974

A taxonomic study of the Western Collared , collaris and

Nathan M. Smith Graduate Department of Library and Information Sciences, Brigham Young University, Provo, Utah 84602

Wilmer W. Tanner Department of Zoology, Brigham Young University, Provo, Utah 84602

Follow this and additional works at: https://scholarsarchive.byu.edu/byuscib

Part of the Anatomy Commons, Botany Commons, Physiology Commons, and the Zoology Commons

Recommended Citation Smith, Nathan M. and Tanner, Wilmer W. (1974) "A taxonomic study of the Western Collared Lizards, Crotaphytus collaris and Crotaphytus insularis," Brigham Young University Science Bulletin, Biological Series: Vol. 19 : No. 4 , Article 1. Available at: https://scholarsarchive.byu.edu/byuscib/vol19/iss4/1

This Article is brought to you for free and open access by the Western North American Naturalist Publications at BYU ScholarsArchive. It has been accepted for inclusion in Brigham Young University Science Bulletin, Biological Series by an authorized editor of BYU ScholarsArchive. For more information, please contact [email protected], [email protected]. Brigham Young University Science Bulletin

MUS. CCMP. ZOOL LIBRARY

'"^'ATaxonomic study of the •^n^t^stern collared lizards, crotaphytus collaris and cromphytus insularis

by

Nathan M. Smith and Wilmer W. Tanner

BIOLOGICAL SERIES ~ VOLUME XIX, NUMBER 4 APRIL 1974/ISSN 0068-1024 INFORMATION FOR CONTRIBUTORS

Contributions to the Science Bulletin should Illustrations (line drawings or photographs) be primarily monographic in nature. For the should be numbered consecutively throughout most part only manuscripts of approximately the paper, and the approximate place of inser- forty or more typewritten pages will be accepted. tion should be indicated in the margins of the Papers will be published approximately in manuscript pages. Captions for illustrations the order that they are received, pending avail- should be assembled on a separate sheet, and ability of funds. Authors must arrange for fi- each plate of figures must have its correspond- nancing their publications, except that the Uni- ing figure number pencilled lightly on the back. versity Press and the library of Brigham Young Illustrations and cuts will be destroyed unless University will share the publication cost pro- their return is requested when proof is returned portionately to their needs for copies of the to the editor. article printed. A table title should be a short, concise state- In the preparation of manuscripts, authors are ment of what the table purports to show, and requested to follow the Style Manual for Bio- should not include information necessary to the logical Journals, American Institute of Biological interpretation of the table. Every column in the Sciences, 2000 P Street, N.W., Washington, D.C. table should carry a head identifying the data 20006. Manuscripts must be typed on one side in that column; the measure in which the data of the paper only, double-spaced with ample are given should be indicated at the head of margins. Footnotes should be avoided. each column. Tables should not be used when To facilitate review by referees, send to the the same information can be given in a few lines editor the original manuscript and one carbon of text and should not duplicate information in copy, together with the illustrations. Copies of text, graphs, or charts. Symbols (asterisk, dag- the original illustrations may be submitted, but ger, etc. ) should be used to indicate footnotes to should be of quality equal to the originals. tables, with footnotes on the same page as the Illustrations should be referred to as figures table. except for materials requiring inserts of special An abstract of less than four percent of the paper, which may be called plates. Illustrations length of the paper should be prepared. This should be so designed as to fit when reduced summary should be understandable without into a one-column or full-page width. Special reference to the body of the manuscript. The care must be taken to allow for proper reduction abstract must be on sheets separate from the in lettering (i.e., a 50% reduction of the figure manuscript. means also a 50% reduction in the lettering). Proof should be corrected immediately on Photographs should be of a glossy finish, un- receipt and returned to the editor. Authors blurred, and showing sharp contrast. Line draw- should leave forwarding addresses if they move ings should be made with black ink on heavy from the address sent with the manuscript. white drawing paper, blue tracing cloth, or Reprints should be ordered when the proof blue-ruled coordinate paper. Use the same is returned. abbreviations on line drawings as in text. Line Address all manuscripts to Stanley L. Welsh, drawings must be equivalent to a professional Department of Botany, Brigham Young Univer- draftsman's work. Original drawings are pref- sity, Provo, Utah 84602.

erable to photographs, even if they are large. Brigham Young University Science Bulletin

A TAXONOMIC STUDY OF THE WESTERN COLLARED LIZARDS, CROTAPHYTUS COLLARIS AND CROMPHrrus insularis

by

Nathan M. Smith and Wilmer W. Tanner

BIOLOGICAL SERIES — VOLUME XIX, NUMBER 4 APRIL 1974/ISSN 0068-1024 TABLE OF CONTENTS

INTRODUCTION 1

REVIEW OF LITERATURE 2

MATERIALS AND METHODS 3 External Morphology 3 Cranial Morphology 7

RESULTS 8 External Morphology 8 Cranial Morphology 9 DISCUSSION 9 External Morphology 12 Cranial Morphology 19 Population Characteristics and Systematics 20 Phyletie Relationships 24

SUMMARY AND CONCLUSIONS 27

ACKNOWLEDGMENTS 27

APPENDIX I , 28

APPENDIX II 28

LITERATURE CITED 28 A Taxonomic Study of the Western Collared Lizards, Crotaphytus collaris and Crotaphytus insularis

by

Nathan M. Smith' and Wilmer W. Tanner-

INTRODUCTION

The range of the Western Collared habitat for the collared lizard since the Wis- Crotaphytus collaris baileiji Stejneger (1890) consinian. as formerly construed included northern Mexico, Systematists recognizing the potential isolat- the Baja California Peninsula, the Islands Angel ing factors within the range originally attributed de la Guarda and Tiburon in the Gulf of Cali- to C. c. haileiji have described C. c. auriceps fornia, and parts of the seven western states of Fitch and Tanner ( 1951 ) from the Upper Colo- Arizona, California, Colorado, Idaho, Nevada, rado River Basin, and C. c. fusciis Ingram and

New Mexico, and Utah ( Stejneger and Barbour, Tanner ( 1971 ) from the Chihuahuan Desert. 1917). Within this vast area the geophysical Also, two insular forms, C. iiistihiris Van Den- features vary considerably and may serve as burgh and Slevin ( 1921 ) and C. dickersonae population barriers. Schmidt ( 1922 ) , although given rank, Although geophysical variation is broad have long been recognized as closely related to throughout the range, niche requirements are C. collaris Sav ( Burt, 1928; Fitch and Tanner, narrow. The collared Uzards require boulders 1951; Ingran/and Tanner, 1971). Allen (193.3) for basking, lookouts, and shelter, and occupy referred to the collared lizards from Tiburon mountain slopes, canyons, rocky gullies and Island as C. c. dickersonae. boulder-strewn alluvial fans where vegetation Collecting trips were made to determine the is sparse (Stebbins, 1966). The many mountain geographic limits of the several populations ranges oriented on north-south axes have pro- and to gather live specimens for electrophoretic vided favorable conditions for extensive distri- analysis of blood proteins. As a result, two new bution of the collared lizard in the Great Basin, subspecies of Crotapht/tus were named (Smith Arizona-Sonoran Desert, and Baja California and Tanner, 1972). The present paper brings Peninsula. Narrow niche requirements cause together the basic techniques and data used collared lizards to form demes with valleys be- with an analysis of the characteristics of each tween the mountain ranges acting as isolating population of C. collaris and C. insularis. barriers. The Imperial Valley, Gran Desierto In this comparative ta.xonomic study multi- of southern California and northwestern Mexico, variate analysis techniques are utilized (Ingram is an important potential barrier between popu- and Tanner, 1971) to study general external lations of the Great Basin and the Baja Cali- morphology and comparative cranial morpholo- fornia Peninsula. The island populations in the gy. Disc gel electrophoresis, which was used Gulf of California were probably members of to compare populations of C. c. bailey, C. c. the mainland population in late Miocene or auriceps, and C. c. bicinctores is not included early Phocene (Anderson, 1950), with Isla in this report because of the limited scope of the Tiburon still part of the mainland as recently as material available, but the data are available 10,000 years ago (Lawlor, 1971). This possible from the authors upon request. An analysis is geophysical arrangement has been discussed made of the westem-collaris complex and its by Tanner ( 1966 ) to e.xplain the distribution relationship to C. c. auriceps, C. c. bailey, and of the night snakes in and around the Gulf of C. c. fiiscus. The Great Plains population, C. c. California. Portions of the Great Basin, which, collaris, is not included, however; a cursory according to Wells and Jorgensen ( 1964 ) were analysis indicates that it also may be a hetero- significantly cooler (and perhaps moister) 8,000 genous group, and a study of its relationship to 40,000 years ago, have only become suitable is planned.

^Graduate Department of Ijbrary and Information Sciences, Brigham Young Universit>', Provo, Utah 84602. -Department of Zoology, Brigham Young University, Provo, Utah 84602. )

Bricham Young UNrvEHSiTY Science Bulletin

REVIEW OF LITERATURE

An adequate review of the literature up to between C. c. bicinctores and C. c. baileyi, 1971 which is pertinent to this study was pre- chose to designate bicinctores as C. insularis sented by Ingram and Taruier (1971:1-2) and bicinctores. Axtell's choice of designation is will not be repeated here. predicated on the basis his evidence indicates In 1971, Ingram and Tanner reaffirmed C. little genetic introgression. He lists similarities for bicinctores c. insularis as a complete c. auriceps and defined the following distin- and dark ventral collar, similar gular markings, guishing characteristics of C. c. baileiji: (1) com- yellow, if present on pressed tail, extensive dark abdomino-groin pig- the body is dark green; ( 2 ) the head, does not extend posteriorly beyond mentation, small dorsal granules, and no green- the supraorbital semicircles or ventrally to the ish coloration. A brief ta.xonomic history of C. c. baileyi throat; (3) it has fewer fused interorbitals and a greater number of supralabials and loreal- and the insular collared lizard populations from la islands are lorilabials series than C. c. coUaris. Ingram and Angel de Guarda and Tiburon listed their as follows: Tanner then defined C. c. fuscus as the popu- with synonymies lation found in the Chihuahuan Desert. They indicated C. c. fuscus most closely resembles Crotaphytus coUaris baileyi Stejneger C. c. baileyi and is distinguished from it by a Crotaphytus baileyi Stejneger, 1890, N. Amer. combination of external morphological char- Fauna, 3:103 (Type locality: Painted Des- distinc- acters, the sum of which gives good ert, Desert of the Little Colorado R., Ari- tion. In addition, C. c. fuscus has a brown body zona; U.S. Nat. Mus.). dorsum with no trace of green and the head Crotaphijtus coUaris baileyi: Stone and Rehn, is a light to cream color, with no trace of yellow. 1903, Proc. Acad. Nat. Sci. Phila., 55:30. Smith and Tanner (1972) designated the Great Basin collared lizard ;is C. c. bicinctores. Crotaphytus coUaris dickersonae Schmidt It can be distinguished from C. c. baileiji and Schmidt, 1922, Bull. other members of the coUaris complex by ( 1 Amer. Mus. Nat. Hist., 46:638 (Type lo- fewer scales and fewer spots in the dorsal sep- cality: Tiburon Island, Sonora, Mexico; U.S. aration of the first collar, (2) greater number Nat. Mus.); Smith and Taylor, 1950, Bull. of enlarged median internasals, (3) second col- U.S. Nat. Mus., 199:93. lar not extending onto the forearm, (4) greater Crotaphytus coUaris dickersonae: Allen, 1933, number of scales from interparietal eye to the Occ. Pap. Mus. Zool. Univ. Mich., 259:7; anterior edge of first collar, and (5) more ex- tensive black pigmentation of the ventral groin. Smith and Tanner, 1972, Great Basin Nat., 32:26. C. c. bicinctores is easily distinguished from Crotaphytus coUaris baileyi: Stone and Rehn, the Baja California population (C. i. vestigium) 1903, Proc. Acad. Nat. Sci. Phila., 55:30. by its collars and dorsal pattern. It is distin- guished from C. c. dickersonae by its brown body and dorsal pattern (Fig. 1). Crotaphytus coUaris auriceps Fitch and Tanner In addition. Tanner indi- Smith and (1972) Crotaphytus coUaris auriceps Fitch and Tanner, cated a close relationship for the populations 195i, Trans. Kans. Acad. Sci., 54(4) :553 from Angel de la Guarda Island, Gulf of Cali- (Type locality: three and one-half miles fornia, Mexico, and the Baja California Penin- north-northeast of Dewey Bridge, Grand sula. Angel Island population was desig- The Co., Utah; Kans. Mus. Nat. Hist). nated C. /. insularis, and the population inhab- Crotaphytus coUaris baileyi: Stone and Rehn, iting the area north and west of Palm Springs, 1903, Proc. Acad. Nat. Sci. Phila., 55:30. Riverside County, California, and south from there along the eastern slopes of the mountains Crotaphytus coUaris fuscus Ingram and Tanner into Baja California Sur, Mexico, was desig- nated as C. i. vestigium. Characteristics of C Crotaphytus coUaris fuscus Ingram and Tanner, i. vestigium which distinguish it from C. i. 1971, Brigham Young Univ. Sci. Bull. Biol. insularis are fewer subdigital toe lamalae on the Scr., 13(2):23-24 (Type locality: sis and second and fourth toes, a larger second collar/ one-half miles north and one and one-half svl index, and a smaller dorsal separation of miles west of Chihuahua City, Chihauliua, both the first and second collars. Mexico; Brigham Young Univ. Herpet. Axtell (1972), after indicating intergradation Mus.). Biological Series, Vol. 19, No. 4 Western Collared Lizards

Crotapht/tus coUaris haileiji: Stone and Rehn, 1921, Proc. Calif. Acad. Sci., ser. 4, vol. la Guarda 1903, Proc. Acad. Nat. Sci. Phila., 55:30. 11:96 (Type locality: Angel de Island, Baja California, Mexico; Calif. Acad. Crotaphi/tus coUaris hicinctores Sci.). Smith and Tanner Crotaphytus insularis insularis: Smith and Tan- ner, 1972, Great Basin Nat., 32:27. Crotaphi/tus coUaris hicinctores Smith and Tan- Crotaphytus coUaris haileyi: Stone and Rehn, ner," 1972, Great Basin Nat., 32:27-29 (Type Proc. Acad. Nat. Sci. Phila., 55:30. locality: Mercury Pass, Nevada Test Site, 1903, Nye Co., Nevada; Brigham Young Univ. insularis vestigium Hcrpet. Mus. ). Crotaphytus Crotaphtjtus coUaris haileiji: Stone and Rehn, Smith and Tanner 1903, Proc. Acad. Nat. Sci. Phila., 55:30. Crotaphytus insularis vestigium Smith and Tan- Crotaphytus insiilaris hicinctores: Axtell, 1972, ner, 1972, Great Basin Nat., 32:27 (Type Copeia, No. 4:721. locality: Guadalupe Canyon, Ju;u-ez Moun- tains, Baja California, Mexico; Brigham Crotaphytus imukiris insularis Young Univ. Herpet. Mus.). Slevin Van Denburgh and Crotapht/tus colUiris haileyi: Stone and Rehn, Crotapht/tus insularis Van Denburgh and Slevin, 1903, Proc. Acad. Nat. Sci. Phila., 55:30.

MATERIALS AND METHODS

External Morphology the east and west sides of the Colorado River in the area of Moab, Utah. Selection and Gathering Material of The fourth and final collecting trip was to Materials examined are from four principle Tiburon Island, Gulf of California, Mexico, dur- sources: (1) the collection at the Brigham ing the first week in April 1971. A number of Young University Herpetological Museum; (2) collared lizards (14) were caught and photo- materials examined at the California Academy graphed on the island and three were caught of Sciences, Los Angeles County Museum, San on the mainland opposite the island. All speci- Diego Society of Natural History, California mens collected during these four trips are de- State College at Long Beach, University of posited in the Brigham Young University Her- California at Berkeley, and the University of petological Museum. A series from the Nevada Arizona; (3) materials loaned from the insti- Test Site (Research Grant Number AT(ll-l) tutions listed in the acknowledgments; and (4) 1496, between the U.S. Atomic Energy Commis- live specimens collected and color photographed sion and Brigham Young University) and from in the field with 35mm Agfachrome. Pyramid Lake, Nevada, were obtained by the Several collecting trips, supported by an junior author. Also, two trips were made by NDEA Educational Allowance, were made be- him to the Palm Springs area, and specimens tween 1969-1971. The first trip, in June 1969, were taken in Chino Canyon and east of was an extensive one, co\'ering two weeks and Thousand Palms, Riverside County, California. 4,000 miles. During this period, specimens were Following the procedure used by Ingram and collected from southwestern Colorado, western Tanner (1971), collared lizards with a snout- New Mexico to the Mexican border, most of vent length less than 80 mm or that were dam- Arizona, southeastern Cahfomia, and south- aged were not included in the computer analy- western Utah. During the last three weeks in sis portion of this study. May 1970, an extensive collecting trip was made The collared lizards examined in this study to southern Nevada, southern California to the are as follows: Mexican border, and most of Arizona. Speci- ARIZONA: Apache Co., BYU 497; LACM mens were taken from the Oatman-Kingman, 16895; UIMNH 7524; USNM 38056. Coconino Arizona, area, which is near an area of con- Co., CU 30081; SDSNH 12772-76; UIMNH firmed intergradation for C. c. hicinctores and 62445, 74789-92, 6543; USNM 45025-26, 60111, C. c. fca/feyi (Axtell, 1972). In July 1970, speci- 60113, 60115, 60117-20, 86942; UU 217, 3006-08. mens were collected through central and south- Gila Co., UIMNH 74797-98. Graham Co., eastern Utah. Specimens were taken from both UIMNH 82348-50. Mohave Co., BYU 32104; Brigham Young University Science Bulletin Biological Series, Vol. 19, No. 4 Western Collared Lizards

O

-a s

a o

a

o

.§1

w

A 6 Brigham Young UNrvERsm- Science Bulletin

UIMNH 62446, 74785; USNM 86941, 115675-76; 51, 52999; UCMVZ 9590, 51140, 73568. Chihua- SDSNH 5291, 17302, 31769, 31937, LMK 31768. hua, BYU 1.3383-86, 1.3410-11, 13736, 14211, Navajo Co., UU 1603. Pima Co., CAS 81420; 15184, 16969-76, 17010, 17014; UCMVZ 70704. LACM 3983, 5899; UAZ 636, 666, 669-670, 672, Sonora, BYU 2425, 39993, .39995, 409.30-31; CAS 1509, 1511, 19986, 28625, 34364, 35539, 35715; 12768-70, 12770, 12772-77, 104467; CU 44201- Santa Cruz Co., UAZ 11241; Yavapai Co., BYU 03; CSCLB 2730, 2752-55, 2757, 2759-61, 2763- 2423; USNM 11860, 15689-90, 15992; UIMNH 64, 29.59; LACM 52882, 52886, 88798; SDSNH 74776. Yuma Co., BYU 2437, 32097-98, 32100-03; CRM117, .35917-19, 43254, 44063; UAZ 694, CAS-SU 15873, 18299; LACM 16892-93; 697-710, 1513-14, 4634, 9625-26, 10248, 10597, SDSNH 16731, 17602, 33301; UAZ 623-27, 12114, 16578, 20144, 20255, 28072, 28662, 30226, 643-44, 680-81, 35673, 35676, 35714; UIMNH 31391, 31482, .3.3781; UCMVZ 10163; USNM 74793, 74847. 26W. Isla del Tiburon, BYU 2426, 2433, 2443, CALIFORNIA: Imperial Co., BYU 33.321; 3163-65, 3167-69, 3172; CAS 14002-12, 5.3263-64; CAS-SU 13843; LACM 16898, 37783; SDSNH SDSNH 46003-06, 50665. Zacatecas, CAS 95961. 11791, 12244-45, 1.3908-09, 29697-98, 38310. Inyo NEVADA: Clark Co., BYU 461, 40932-33; Co., BYU 409.34-35; CAS 22153-54, 65216, CAS-SU 20083-85. Churchill Co., LACM 268.39- 65221-22, 65224, 65481-82, 65515-17, 65581, 43. Lincoln Co., CAS 37025-.36, .37029-31, 370.33, 65585, 65636, 65662, 65681, 84241, 89621, 89626- 37035-36, 37038-39, 37041-42, 37044, 37046, 30, 89633-34; CAS-SU 22072-73; LACM 26824, 37058-59. Lyon Co., CAS 20576-80, 20583-85, 36667, 36670; SDSNH 15878-79. Kern Co., BYU 22660, 22662-64, 22690, 22705-06; LACM 26838. 31948; CAS 111006. Los Angeles Co., CSCLB Nye Co., BYU 2418, 2424, 17275, 18815-16, 2747-51; LACM 16870, 2681L 26814, 63187-90; 22189-96, 23629-30, 2.3882-83, 30088, 30587; CAS SDSNH 2834-36, 19481-82. Mono Co., CAS 37514, 37693-95, 37697, 37699-710, 37716, .37721; 84962. Riverside Co., BYU 2422, 2429-30, 2432, CAS-SU 20086-9.3, 20096. Washoe Co., BYU 2435, 2438, 31983, 32099; CAS-SU 22711; CSCLB 2401, 2442, 3115; CAS 21487, 21489-90, 21492-94, 600611-1, 2744, RDK 600430-1, RDK RBL 21496, 21498-500, 21.505-06; SDSNH .38311, 600327-2, JAN 700510-2, VAP 660513-4; LACM 38676-77. 16873, 16875-76, 16885-89, 26821-22, 52889-90; NEW MEXICO: Chaves Co., LACM 3974- SDSNH 20698-99, 39751; UCMVZ 1, 210-11, 231, 75. Dona Ana Co., USNM 22268, 25423. Lincoln 243, 74609-11, 76737, 41089, 51850-51, 71723. San Co., LACM 16990. Luna Co., BYU 31940, 31944, Bernardino Co., BYU 3.3.320; CAS 5245, 36268, .32121. Quay Co., USNM 44940. Sierra Co., 42185, 42785-91; CSCLB 2720-21, 2735-36, LACM 16992. 600529-5, 700412-12; 2738-39, RBL RCS OKLAHOMA: Carter Co., BYU 500, 1574. LACM 63179-80, 63182-83, 63188; SDSNH 2490. OREGON: Malheur Co., UAZ 21124. 5874, 11087, 20888, 29664. 38703, 39874; LMK TEXAS: El Paso Co., USNM .59.351; UTEP UCMVZ 5.368, 316.34, 52486. San Diego Co.. .52, 55, 57. Garza Co., CU .32277. Jeff Davis Co., BYU 3994; CAS 62875, 64368; CAS-SU 7930; UAZ .35145. Randall Co., CU 1.3554-56. Roberts SDSNH 1687, 4701, 11088, 11958, 12551, 13250, Co., USNM 32866. Stcphans Co., BYU 13117. 18591-92, 22.327. UTAH: Beaver Co., BYU 12715. Emery COLORADO: Los Animas Co., CU 1292, Co., BYU 16496, 16774, 20089-90. Garfield Co., 2939. Mesa Co., BYU 1134. San Miguel Co., BYU 11740-41, 12699-701, 31882-83; UU 3329. CU 4448, 4451, 4453, 4456, 4458. Grand Co., BYU 2421, 24,34, 3114, 12854-55, 64173. IDAHO: Ada Co., BYU 507; CAS Kane Co., BYU 2732, 11331, 11.384, 11.386b, Butte Co., BYU 30772. Owyhee Co., BYU 2834; 14660, 14894-96, 14898-902, 18921, 21259-60, CAS 55247-48, 55253, 64163-64; NNC 7, 28; 21262-63, 40936-.39; UU 1449. Millard Co., BYU 1444-45. SDSNH 445, 447-48, 8753, 8755, 8883, 12946, 21000-02, KANSAS: Montgomery Co., BYU 22167. 21703; UU 2802. San Juan Co., BYU 2419, Wilson Co., KU 41, 48-50, 54. 18.335, 31945-46, 31951, 31981, 32088. Tooele MEXICO: Isla Angel de la Guarda, CAS Co., BYU 4.305-06, 14818, 14820-21. Utah Co., 49151, 50873-76, 50878-79, 86754-55, 86783-84; BYU 552, 1455, 16.30-31, 2844, 3116, 1.3041, CAS-SU 21948, 22712; LACM 4001-02, 9854-55; 14689. 16602, 22129, .30382; UU 20.38. Washing- SDSNH 46001; UCMVZ 96635-37. Baja Califor- ton Co., BYU 446, 8754, 11.324, 11.386a, 12190, nia, CAS 11545-46, 14000-01, 17048, 18822, 12875. 19124-26; CSCLB LWR 670619-1, LWR670619- SPECIMENS SEEN-NOT USED IN THE 2, KAW 700324-6; CU 45855-61; LACM 16994, COMPUTER ANALYSIS: CAS 14002-03, 21608, 16996-99, 63167-73, 63175-78; SDSNH 17052, 39114-15, 64966; CAS-SU 19125; LACM 8798- 19788-92, 24392, 30107-11, 37815, 41612, 52950- 99, 16993, 36666, 52885-86, 63184; SDSNH .

Biological Series, Vol. 19, No. 4 Westebn Collared Lizabds

40074, 40139; UIMNH 74801; UU 122, 124, Cranial Morphology 33597. Stejneger (1890:103) distinguished C. baileyi from C. coUaris stating that in Statistical Methods the former the "head [is] narrower, and the snout [is] longer." After an initial examination of the museum In 1921, \'an Denburgh and Slevin established specimens available at Brigham Young Univer- C. insularis on the basis of a narrower head and sity and a review of the literature with its de- longer snout than baileyi. Unfortunately, neither scriptions of the collared lizards from Tiburon Stejneger nor Van Denburgh and Slevin listed and Angel de la Guardia Islands, it was hy- the data upon which these statements were pothesized that there are eight distinct collared based. If their data contained measurements lizard populations in the Great Basin, the Baja from juveniles, then it is possible these char- California Peninsula, Tiburon and Angel islands, acters are not significant. To test this possi- the Upper Colorado River Basin, Central Ari- bility, a section on cranial morphology is in- zona, Chihuahua, Mexico, and the Great Plains. cluded in this study. The methods of analyses employed in this study, with minor exceptions, are those of In- Selection of Material In step 2, we were gram and Tanner (1971:4). A total of 126 specimens was selected from respectively, limited to 17 and 13 specimens, the seven areas used: the Great Basin group Tiburon islands which made a for Angel and (N^20) from western Arizona, eastern Cali- of 150 individuals. In step 4, eight total sample fornia, Idaho, Ne\ada, and western Utah; the clusters were fonned. Also we omitted step 5 Baja group (N=20) from Baja and southern and the color and pattern analysis. California; C. /. insularis (N=16) from Angel Tlie terminologv for taxonomic characters is Island; C. c. dickersonae (N=20) from Tiburon that used by Ingram and Tanner (1971:5). Island; C. c. auriceps (N = 10) from the Upper The characters used in the evaluation, with Colorado River Basin; C. c. baileyi (N = 20) their numbers, are as follows: (1) tail length/ from central Arizona; and C. c. fuscus (N = 20) hind-leg length; (2) tail length/snout-vent from Chihuahua, Mexico. Only specimens known length; (3) snout- vent length; (4) intemasal to represent these populations were used and all scales; (5) number of fused interorbital scales; specimens had a snout-vent length greater than (6) fronto-parietal scales; (7) scales from the 80 mm. Approximately equal numbers of males union of the posterior canthal and subocular to and females were included in each sample. the supralabial; (8) supralabial scales; (9) in- The following collared lizards were exam- fralabial contacts postmental; (10) gular scale ined in this portion of the study: rows; (11) number of enlarged internasals; (12) ARIZONA: Coconino Co., BYU 32116; scales from rostral to interparietal; (13) dorsal LACM 28895; UIMNH 6453, 7525, 74786. Mo- scales from interparietal to anterior edge of first have Co., UIMNH 7524, 74781, 74787, 82349. collar; (14) dorsal scales from the anterior edge Yavapai Co., UIMNH 74767-74, 74776-77. 82354. of first collar to posterior edge of second coUar; CALIFORNIA: Inyo Co., LACM 36666, (15) total dorsal scales; (16) total ventral 36670. Los Angeles Co., LACM 63187. Riverside scales; (17) dorsal separation of the first collar; Co., LACM 16873, 16875. San Bernardino Co., (18) dorsal separation of the second collar; LACM 63179. (19) number of spots within the dorsal separa- MEXICO: Isla Angel de la Guarda, CAS-SU tion of first collar; (20) second collar length/ 21948, 22712; CAS 50873-76, 50878-79, 86755, snout-vent length; (21) subdigital lamellae of 86783-84; LACM 4001-02, 9854-55; SDSNH right hind foot, second toe; (22) fourth toe sub- 46001. Baja California, BYU 23.337; CU 45855-58, digital lamellae; (23) fifth toe subdigital lamel- 45860-65; LACM 16993-94, 16996-99, 63171. Chi- lae; (24) femoral pores, right side; (25) second huahua, BYU 13736, 14211, 15184, 15186-87, collar extends onto the upper ami; (26) dorsal 16989, 16971-77, 17010, 17014; USNM 2725, light elongate spots present; (27) the pregroin 14242, 14307, 14307a. Isla del Tiburon, BYU brown black; (28) the coloration of the pre- 3163, 3167-68; CAS 14003-07, 14009, 14011-12, groin fonns spot on the thigh; (29) dorsal 53263-64; CU 26679-80, 35168; SDSNH 44063, ground color is blue green; (30) small spots 46005-06, 50665. on side of head; (31) large spots on side of NEVADA: Clark Co., BYU 23629, 23883. head; and (32) reticulation on side of head Churchill Co., LACM 26843. Lincoln Co., CAS forms patternal bars. Characters 25-32 are 37033, 37035. Lvon Co., CAS-SU 20576, 20585, binary (0 ^= characteristic is absent, 1 ^ it is 22690, 22705; LACM 26838. Washoe Co., CAS present ) 21499. 8 Brigham Young University Science Bulletin

TEXAS: El Paso Co, UU 493. length; (7) head width at eves/total skoill length;

UTAH: Garfield Co, BYU 12700-01. Grand (8) head width at nostrils/total skull length; (9) Co., BYU 551, 1625-26, 11737, 12854-55, 31949- head depth at eyes/total skull length; (10) head 50, 31981, 32111. Millard Co., BYU 21000-02. width at widest point/eye-to-snout length; (11) head width at eyes/ eye-to-snout length; (12) Characters Used head width at nostrils /eve-to-snout length; (13) A total of 16 characters were analyzed from head depth at eyes/eye-to-snout length; (14) taken from the intact, un- six measurements head depth at eyes/head width at widest point; skinned head. All measm^ements were made to (15) head depth at eyes/head width at eyes; nearest 0.1 using a Vernier Caliper. The the mm, and (16) head depth at eyes/head width at skull measurements were as follows: (1) total nostrils. length, distance from posterior of quadrate to Statistical Analysis anterior of rostral; (2) eve to snout, distance from the anterior corner of the e^e to anterior Dixon's BMD07M Computer Program (1968), of rostral; (3) head width at widest point, a the same stepwise discriminant analysis used in point posterior to the eye, in the parietal region; the external moqDhology portion of this study, (4) head width at eyes, distance from jaw to was used at this point because the statistical jaw at a point medial to each eye; (5) head verification of the seven populations was pre- width at nostrils, distance from each jaw at a viously demonstrated.

point medial to each nostril; and (6) head depth The discriminant program calculates the at eves, depth of head from a point medial to mean, standard de\iation, and the within-group supraoculars. correlation coefficient for each character. The The six measurements made for each skull U- and F- statistics which test the null hypothe- were modified to form the following 16 char- sis of no difference among populations are acters which are listed with their character num- listed at each step. The number of correctly ber: (1) total skull length; (2) eye-to-snout identified specimens is displayed at each step length; (3) head width at widest point; (4) head and the a posteriori probability for an individ- width at eyes; (5) eye-to-snout length/total skull ual's membership in each of the seven groups is

length; (6) head width at \videst point/total skull given.

RESULTS

External Morphology eastern edge of the Coachella Valley, Riverside Cluster Analysis County—the Great Basin group (C. c. bicinc- tores); Baja and southern Cahfomia, west of Cluster analysis was used to form eight Coachella Valley—the Baja group (C. /. vestigi- groups of closest morphological similarity (Fig. um); Angel Island-C. /. insularis; Tiburon 2). To test for the independence of the eight Island—C. c. dickersonae; Upper Colorado Riv- groups formed by clustering from the eight er— C. c. duriceps; Central Arizona— C. c. baileyi; groups postulated by geographical locality, a Chihuahuan Desert— C. c. juscus; and other contingency table was formed (Table 1). The specimens from Colorado, New Mexico, Texas, null hypothesis was tested by chi-square with Oklahoma—C. c. collaris. forty-nine degrees of freedom. The test statistic is significant at the 0.001 level. Discriminant Analysis m Discriminant analysis, which predicts mem- X=(l- ,49) =^==1 (0„-Ei-)VEij bership within a group on the basis of a set of i=l j = l continuously scaled attributes for the individual X=(l-oc,49) = 485.7 (Cocley and Lohnes, 1971), was used to define X=( 0.999,49) = 85.4 more accurately the relationships seen after cluster analysis. The BMD07M Computer Pro- Therefore, X^ (l-oc,49) > X= (0.999,49) and gram of Dixon (1968) was used, with the out- the null hypothesis is rejected. put consisting of (1) character mean and stan- Specimens from the various locations were dard deviation for each group (Table 2); (2) labeled as follows: the Great Basin region of the coefficient of the discriminant function for Idaho, western Utah, parts of western Arizona, each group (Table 3); (3) the F-statistic to Nevada, and California south and west to the show the difference between each group (Table )

Biological Series, Vol. 19, No. 4 Western Collared Lizards

4); (4) the number of case.s classified in each 94 percent of the sample was identified correct- group after each new character enters the ly, which is well within the 75 percent taxo- analysis (Table 5); (5) the U-statistic which nomic rule of Mayr (1969). tests the difference between groups (Table 6); and (6) the probability of each specimen's Cranial Morphology membership within a group (Table 7). Discriminant Analysis The output of the BMD07M lists both the The discriminant analysis of Dixon ( 1968 U-statistic and appro.ximate F-statistic. Each was used to test its ability to recognize, on the hypothesis of no difference be- tests the null basis of cranial morphology, the seven previously is 0.00006 with tween groups, the U-statistic postulated populations. The mciin and standard freedom. The approximate 24,6,164 degrees of deviation was calculated for each character F-statistic was used because of the availability (Table 9). Again the approximate F-value was appro.ximate F-value is 24.90764 of tables. The used to test the null hypothesis of no difference degrees of freedom. The tabuhu- with 144,831.93 between groups instead of the U-statistic be- 1.76 (Ostle, 1963). F-value is F(. 999,120,120) = cause of the availability of F-tables. The ap- Therefore, F(l-oc, 144,831.93) > F(. 999,120, proximate F-value is F(l-oc ,54,570.59) =4.7506. 120) and the null hypothesis of no difference The tabular F-value is F( 0.999,40, 120) = 1.84 between groups is rejected. (Ostle, 1963). Therefore, F (1- a: ,54,570.59 ) > The F-statistics to test between each pair of F( 0.999,40,120) and the null hypothesis of no groups are listed in Tabic 4. The tabular F- difference is rejected at the .001 level of signifi- value is F(.999,24, 120) = 2.40 (Ostle, 1963); cance. therefore, the null hypothesis of no difference An F-matrix of 9,111 degrees of freedom was is rejected for each of tlie groups compared at calculated to determine if all groups differ sig- the .001 level of significance. nificantly (Table 10). The tabular F-value is Each postulated group contained lizards F( 0.999,9,60) = 3.69 (Ostle, 1963). Tliercfore, known to be from a gi\ en population. Individual not all groups are distinguishable at the .001 lizards were compared to the discriminant func- level of significance. tions ( Table 3 ) and assigned to one of the seven Table 11 lists the number of individuals possible populations each time a new character classified into each group after nine characters entered the program (Table 6 and 7). To test have entered the analysis. From this informa- the program's ability to classif)' lizards from tion, the number of correctly identified speci- known populations, the percentage of correctly mens was calculated to see how many popula- identified were divided by total number in the tions fit the 75 percent rule (Table 12). Only geographic sample (Ncorr Ntot). This was the Great Basin population was distinguishable summed over all groups to get the total sample \\ith 75 percent accuracy using cranial mor- correctly identified (Table 8). Approximately phology alone.

DISCUSSION

Ingram and Tanner (1972:6) indicated that The objective of this study has been to dis- the collared lizards of western Utah, Idaho, Ne- tinguish the collared lizard populations of these vada, California, and Baja California are ver\' \\ estern states from those which are contiguous different from those east of the Colorado River. to the east (C. c. haileiji, C. c. auriceps, and

Table 1. A contingency table to test the independence of Ward's clustering method and the proposed groups. 10 Brigham Young University Science Bulletin

-2 "Sb

! -S i; Biological Series, Vol. 19, No. 4 Western Collabed Lizards 11

. = = i ; :^Jz < „iisl

' ^ I J ^ 5 J .- its 5 li a 1 1 1 J J I J J J 3 3 I i J J J I J e 1 1 e 1 1 1 e I e i J J .55 I * J I J I 3 5 1 1 ll n S £ i a Jj I e e =S J i a e e 12 Brigham Young University Science Bulletin

Table 2. A listing of the means and standard deviations for the 32 characters measured on the individuals used 13 Biological Series, Vol. 19, No. 4 Western Collared Lizards in forming the discriminant functions.

Groups 14 Brigham Young University Science Bulletin

Table 3. A listing of the coefficients of the discriminant functions after 24 characters have been entered into the discriminant analysis. Biological Series, Vol. 19, No. 4 Western Collared Lizards 15

Table 6. Summarv table of the sequential entry of characters into the BMD07M discriminant analysis program with their F-value to enter or remove along vv-ith the U-statistic at each step. 16 Bricham Young University Science Bulletin

Table 9. A listing of the means and standard deviations of the 16 cranial morphology characters. Biological Series, Vol. 19, No. 4 Western Collared Lizards 17

Groups 18 Brigham Young University Science Bulletin

Great Basin

( Idaho, Nevada, western Utah, southwestern Arizona, eastern Cahfomia)

Percent of

Sa m p le

100-

80-

Sonora, Mexico (West of Highway 15, Organ Pipe National Monument area and southern end Tinajas Altas Mountains exclusive of Tiburon Island) Percent of Sample

100-

80-

Fig. 4. Continued on next page. —— ) .

19 Biological Series, Vol. 19, No. 4 Western Collared Lizards

Baja California, southern California south from the San Gorgonio Pass

Percent of Sample

100-

B BXI 1 G GXB U

Baja Fig. 4. A comparison of collared lizard populations from the Great Basin, northwest Sonora, Mexico, and California-southern California.

A key to the abbreviations used in Fig. 4. A- auriceps population from the Upper Colorado River Basin. B vestigium from the Baja Cdifomia, southern California population. C baileyi population from Central Arizona. F fuscus population from Chihuahuan Desert. G hicinctores population from Idaho, Nevada, western Utah and Arizona, and eastern California.

I ii^sttlaris population from Angel Island. T dickersonae population from Tiburon Island. GXT— intergrade population of hicinctores and dickersonae. CXG— intergrade population of bailctji and hicinctores. A.XC— intergrade population of auriceps and hicinctores. FXG— intergrade population of fuscus and hicinctores. CXF— intergrade population of baileyi and fuscu.^. BXI— intergrade population of vestigium and insularis. G.XB— intergrade population of vestigium and hicinctores. U— Individuals not assignable to any of the above groups.

are Great Ba.sin and Tibviron, witli a few mem- fact that our statistical tests supported these bers of the Tiburon group shifting toward the eastern populations ( Table 4 )

coUaris complex. This is e.xpected, since we con- sider the Tiburon-Sonoran populations to have Cranial Morphology derived from those populations evolving been The skull has been shown to be significant to the cast and now designated as C. c. fuscus. in the definition of subspecific taxa (Avery The Baja California group is more closely re- and Tanner, 1964, 1971; Presch, 1969; Robison to Tiburon group in the characters lated the and Tanner, 1962). Van Denburgh and Slevin studied than it is to the Great Basin popula- ( 1921 ) declared C. imularis to have a longer, tion. This is in agreement with the results of narrower head than C. c. baileyi. In comparing the cluster analysis. There are two valid reasons the cranial morphology of the seven popula- members of the collaris complex did not sep- tions of this study, only the Great Basin group arate as well in this plot as the one shown by reached 75 percent distinction (Table 12). The and Tanner (1971:23): a more re- Ingram (1) character which best separates the western and

scale was used to plot this graph, and ( duced 2 collaris complexes is head width at widest point/ not all the same characters used by Ingram and total skull length (Fig. 6). Within these com- Tanner were used in this study. These two plexes, the distinctions, except for the Great points should be kept in mind along with the Basin group, are not significant. 20 Bricham Young Univebsity Science Bulletin 21 BioLOCiCAi. Series, Vol. 19, No. 4 Western Collared Lizards

leading from Palm Springs to the aerial tram- way in an area which is occupied by vestigium. This specimen, when reconsidered, may repre- sent the results of human intervention in collared lizard distribution. Dr. Ernest Tinkum and James Davis of Indio, Riverside County, Cali- fornia (pers. comm.), indicate they have not ob- served hicinctores from this area. Until a thorough ecological study can be completed within the vicinity of the San Gorgonio Pass, we now prefer to believe hicinctores and ves- tigium represent allopatric species (Fig. 7). Ad- ditional specimens of interest from this area are Trr LACM 16SS9, a female hicinctores from Caba- zon, Riverside County, California; and UCMVZ standard deviation tor head Fig. 6. The mean and 1, a female vestigium from Hallgrade, near width at widest point/total .skull length; (K) C. c. Cabazon. The latter specimen was taken by fuscu?, (C) C. c. haileyi, (A) C. c. auriceps, (G) W. P. Taylor in 1908. Stephen B. Ruth, from C. c. hicinctores, (T) C. c. dickersonae, (B) C. i. field notes and oesfigiuni, (1) C. i. insiilaris. the USMN'Z, has examined the reports that Taylor's base camp was I'A miles from the interparietal to the anterior border of south of Cabazon and that on May 9 Taylor the first collar, and (4) the .second collar usu- walked up to Hurley Flat (which was south of ally extending onto the upper arm. Additional- his camp) via the Hallgrade. In this case, the ly, AxtcU (1972) suggests a rounded tail and specimen was seemingly taken on the low foot- dark oral pigmentation as characteristic of the hills at the base of die San Jacinto Mountains. coUaris complex. Axtell (1972) has taken the position that Crotaphijtus Tlie four populations of the western complex hicinctores should be placed in tlie His arrangement is are believed to represent two species of Crota- insularis species complex. similarities be- phtjtus. The Great Basin and Tiburon Island- based on some morphological Western Sonoran populations are considered tween these species, and we readily agree that is to ac- C. coUaris. The southern California, Baja Cali- there are similarities. However, if one hicinctores is specifically fornia, and Angel Island populations are con- cept his h\pothesis that introgression sidered C. insularis. The Great Basin popula- distinct from C. coUaris because in the Cerbat Mountains where tion, which we consider C. c. hicinctores (al- is very narrow intergradation, then by that though similar to C. insularis with its brownish he has pinpointed information dorsal coloration, ventral collar, gular, and ab- reasoning—and on the basis of the suggest that hi- doniino-lateral pigmentation), is distinctly dif- presented above—we would distinct from in- ferent in dorsal pattern and collar characteristics. cinctores is also specifically information to follow, we will Upon visual inspection, there is no confusion be- sularis. In the indications of additional intergradation be- tween a C. c. hicinctores and C. insularis; this give hicinctores and other coUaris suhspecies is true even between females ( Smith and Tanner, tween Mexico. These data seem to 1972: Fig. 1). Such is not the case when com- in northern Sonora, paring females of two closely related subspecies relate hicinctores more closely to haUeiji, fuscus, we, therefore, prefer to such as C. c. fuscus and C. c. hailet/i. Cluster and dickersonae, and group hicinctores with coUaris at present. analysis ( Fig. 2 ) and canonical analysis ( Fig. 5) sharply distinguish C. c. hicinctores from The possibility that hicinctores, which has C. insularis; however, the distinction between been found as far north in the upper Colorado Baja California and Angel Island populations, River Basin as the Potash Road (BYU 3114), two populations we consider to be subspecifical- Grand County, Utah, intergrades with C. c.

ly related, is not so apparent. auriceps has not not yet been demonstrated. In our earlier paper (Smith and Tanner, However, Axtell (1972) has demonstrated that intergrade in Arizona with C. 1972), we indicated an area of sympatry for C. c. hicinctores does c. haileyi, which is a close relative of auriceps. hicinctores and C. /. vestigium in the vicinity of the San Gorgonio Pass, Riverside County, Cali- In the southern portion of its range, hi- fornia. This assumption was based upon a fe- cinctores has not been shown to contact or inter- male hicinctores, CSCLB VAP 660513-4, said grade directly with C. c. dickersonae; but it to have been taken from the side of the road does intergrade with C. c. fuscus, and C. c. 22 Brigham Young University Science Bulletin

i. veitigium, Fig. 7. Range map for Crotaphtjtus collaris and Crotaphtjtus imularis: (1) C. c. bicinctores. (2) C. baileyi, C. c. /iwcus. (3) C. i. insularis, (4) C. c. dickersonae, (5) C. c. aunceps, (6) C. c. (7) )

Biological Series, Vol. 19, No. 4 Western Collared Lizards 23 fuscus intcrgrades with C. c. dickersonae (See ascertained. However, present data indicates the following discussion of population character- considerable intergradation. istics for dickersonae). The members of C. c. dickersonae are similar

Specimen UAZ 12114 from an area midway to C. c. bicinctores in many respects, including between MacDougal and S\kes Crater in north- collar pattern, but are distinguished from them western Sonora, Mexico, is considered by us to by their blue or green dorsal ground color and be an intergrade between C. c. bicinctores and dorsal light elongate spots. These two popula- C. c. fusctis. Its appearance is very much like tions are so closely related that many of their bicinctores. The a posteriori probability was morphological characteristics, although different, 1.000 for bicinctores. It has, however, jtiscus show considerable overlap (Table 2). The dor- characteristics such as its second collar extend- sal light elongate spots of dickersonae are sug- ing on to the forearm, the first collar having gestive of those found in C. insularis. The ven- an anterior extension that could be interpreted tral pigmentation of male dickersonae is not as as the incomplete formation of spots, a rounded pronounced as that seen in bicinctores or in- tail (a coUaris character suggested by Axtell, sularis. The range of dickersonae can now be patch, minor 1972 ) , a ver\- reduced ventral groin extended north from near Kino Bay to Puerto postmentals con- pigmentation of the dewlap, Libertad ( UAZ 20144 ) and to El Desomboque, tacting the infralabials, and .34 subdigital lamel- Sonora, Mexico (UAZ 704, 705, and 30226). Its lae on the fourth toe. range south from Kino Bav is not completely Similarly, UAZ 10248 from 36.7 miles west known, but two specimens (BYU 409.30-31) of Sonoita has a posteriori probabilities of 0.687 from the hills north of Guaymas show inter- for C. c. aiiriceps and 0..3()7 for C. c. fuscus. grading characters between dickersonae and Visual inspection of the color and dorsal pat- fuscus. Additional tern reminds one of bicinctores. Both of the above specimens have a bicinctores characteristics include: 3 frontopa- posteriori probabilities of 1.000 for dickersonae. rietals, 3 scutae in dorsal separation of first Msual inspection reveals the male ( BYU 40931 dorsal separation of second collar, no scutae in to have dorsal light elongate spots, a blue collar, subdigital lamellae on the fifth and 21 green cast, 3 frontoparietals, no fused interorbi- toe. It is seemingly an intergrade of fuscus tals, 9 loreal-lorilabials and 2 enlarged inter- bicinctores. and nasals— all of which characterize dickersonae. Specimens which we have observed from Conversely, the second collar extends boldly Organ Pipe National Monument, Pima County, onto the upper forearm, the postmentals contact Arizona (UAZ 636, 1511; LACM 3983, 5899), the infralabials, there are 5 scutae within the have appeared to be fuscus: but visual inspec- dorsal separation of the first collar, and the tail tion of the first collar pattern suggests it is not is round—all of which characterize fuscits. In nearly so characteristic as those from the type addition, the first collar unites ventrally and locality in Chihuahua, Mexico. When the dorsal there is bold abdomino-lateral pigmentation, collar pattern is masked and only the pattern on which is not characteristic of either population. the back is \isible, it is often difficult to dis- Axtell (1972) noted this exaggerated effect on tinguish these fuscus specimens from bicinctores. the variables found in his intergrades of bicinc- Several specimens from the area of Caborca tores and l)ailetii. Such intergrade variables and San Luis, Sonora, are especiallv interesting cause one to wonder if bicinctores has not arisen because of the a posteriori probabilit\' listed for as a result of hybridization between several of them by the BMD07M program. On the basis the subspecies and established its range in the of this information, one would suspect this to lo\\'cr Colorado River Basin with its major ex- be an area of intergradation or at least an area tension into the Great Basin. The female of this which represents the remnants of genes repre- pair (BYU 40930) has dickersonae characteris- resenting bicinctores, dickersonae, fuscus, and tics of 3 frontoparietals, 8 loreal-lorilabials, 7 other members of the eastern complex, which scutae in the dorsal separation of the first collar, may nou' be geographically or ecologicallv iso- and dorsal light elongate spots. Her fuscus lated. The specimens, with their localities and characteristics include 1 fused interorbital, 21 probabilities, are UCMVZ 10163, d", 30 miles W scutae from the interparietal to anterior edge of Caborca, .406 fuscus, .353 bicinctores, .241 first collar, 5 scutae in the dorsal separation of dickersonae; CAS 104467, 9 , 9 miles E Caborca, the second collar, and the extension of the sec- .529 bicinctores, .435 baileiji, .024 fuscus. Addi- ond collar onto the arm. We consider both of tional field studies in this area are needed before these specimens to be intergrades between fus- complete relationships and distributions can be cus and dickersonae. )

24 Brigham Young University Science Bulletin

The following additional specimens from the islands in the Gulf of California (Anderson, Guaymas area are listed with their museum 1950). Female collared lizards from Angel number and their BMD07M a posteriori proba- Island, according to Banta (1961), bear a simi- bilities. The probabilities are not consistently in larity to the leopard lizard Crotaphytus tois- favor of any given population indicating a lizenii silus. It is probable that the collared gene pool that has traits for clickersonae, ftiscus, lizards from Angel Island display some of the bicinctores, baileyi, and auriceps. We have not genetic variance which gave rise to C. wislizenii been able to collect collared lizards in the area and are a relict population which was dcri\'ed near Guayn as and suggest that this will be es- from a more widespread population, the rem- sential to understand better the complete rela- nants of which are now found on the island

tionship of (Uckersonae and fuscus. The speci- and Baja California Peninsula. Banta ( 1961 mens, with their identification and a posteriori believes C. collaris is an older population than probabilities, are: LACM 52886, d, 9 miles the leopard lizards. This we have not been able NW Guaymas. .005 dickersonae, .004 fuscus, to demonstrate from our data; however, pre\ious that reticulatus .991 bicinctores; LACM 52882, 9 , 9 miles NW studies have shown Crotaphytus Guaymas, 1.000 fuscus; UAZ 10597, 9, 29.3 of eastern Mexico and southcentral Texas, a miles S Los Pocitos, .795 fuscus, .205 bailetji; fonn which lacks the collar pattern of C. col- SDSNH 35918, d, between Guaymas and Em- laris, is probably an intennediate between C palme, .669 auriceps, .268 fuscus, .058 baileyi, collaris and C. ivislizenii (Robison and Tanner, .006 bicinctores; SDSNH .35919, d, between 1962; Montanucci, 1969). l>e "^wo species in the Guaymas and Empalme, .404 bailetji, ..339 which may be considered the more primi- fuscus, .214 bicinctores, .043 auriceps; and tive members of the collared lizard complex are SDSNH CRM 117, d, Playa Miramar SW of found on the extreme eastern edge (reticulatus) Guaymas, 1.000 fuscus. and the extreme western edge (insidaris) of Specimen GAS-SU 14002, reported collected their distribution. from Tiburon Island, is a perfect example of C. The historical distribution of the collared

i. insularis and presents an anomalistic situation. lizard in the Gulf of California region is a The specimen was collected 7 May 1952. CAS- matter of speculation at present. Those found SU 14003-14012 taken by the same party two on Angel Island and the Baja California Penin- days earlier, 5 May 1952, from Tiburon Island, sula are more closeh' related to I'ach other

are typical dickersonae. To us it seems very than they are to C. c. dickersonae from Tiburon possible the GAS-SU 14002 has been mislabeled. Island or bicinctores in the Great Basin ( Fig. The Baja Califoniia-southem Galifoniia pop- 5), indicating a long period of isolation in the western ulation, C. insularis vesti'^ium, represents a dis- portion of the Gulf. Tlic Baja Cali- fornia Peninsula is thought to have formed tinct taxon. It differs from C. c. baileyi in those some four to ten million years in late Plio- characters which distinguish it from C. c. bicinc- ago cene by crustal movement ( Moore and Buffing- tores. Crotaphytus i. vestigium can be distin- ton, 1968). A peculiar feature in this h\pothesis guished from the Great Basin population by its is the fomiation of of reduced first and second collar and the distinct the northern portion the dorsal transverse bands which separate light Gulf of California before the existence of a elongate spots (Smith and Tanner, 1972: Fig. southern entrance (Larson, Menard, and Smith, 1). 1968). It is possible, therefore, to hypothesize at least three modes of distribution of the col- Crotaphytus i. vestigium can be distin- lared lizard from mainland Mexico to the Baja guished from C. i. insularis by fewer lamellae California on the second and fourth toe, a larger second Peninsula: (1) the peninsular popu- lations collar/s-v ratio, and a smaller dorsal separation were isolated by the peninsular drift which formed the Gulf of California; the of both the first and second collars (Table 2). (2) collared lizard extended its range to the Baja California Peninsula a migration south of the Phyletic Relationships by newly forming Gulf of California, later to be It is noteworthy that cluster analysis separ- isolated as the peninsula was formed; (3) the ated collared lizards into eastern and western collared lizard migrated aroimd the northern groups, producing four clusters for the western end of tlie Gulf of California and later liad .its complexes and four for the eastern. These, we range restricted to the peninsula, isolating it believe, can be used as indicators of phyletic from the more recent populations to the north relationships (Fig. 2). Within the western and east. group, the most distinct population is C. i. The origin of the Angel Island population insularis from Angel Island, one of the older from the Baja California population is difficult Biological Series, Vol. 19, No. 4 Western Collared Lizards 25 to assess because of the absence of a good fossil 7500 feet elevation in the Zuni Mountains of record. Angel Island is a deep-water island northwestern New Mexico (Gehlbach, 1965b). formed during the Pliocene (Anderson, 1950). The preference of a more mesic environment by A possible land bridge between the island and members of the coUaris complex might be ex- the peninsula might have pemiittcd unrestricted plained by natural selection for their green col- distribution of the collared lizard. Later, when oration; conversely, the brown coloration of the land bridge was inundated, the insular pop- the more xeric western fomi seemingly provides ulation became a relict. It is also possible that them an adaptive advantage in their arid en- the collared lizard was rafted to the island be- vironment. fore the channel became extensive. Tiburon-Sonoran and Great Basin popula- Tiburon, a shallow-water island, was un- tions evidently had their genesis in northcentral doubtedlv separated from Sonora, Mexico, when Mexico, probably from the line of C. c. fiiscus, the level of the seas rose during the recession of spreading west to the Gulf Coast and the Colo- the last glacial period of 10,000 to 15,000 years rado River and later invading and occupying the ago (Lawlor, 1971). E\cn today, the channel west side of the river as well as the extremely between the island and the mainland is very dry areas of western Arizona. In Utah, the Colo- narrow (ca. 5 miles) and shallow at low tide. rado River has served as an effective barrier,

Crotaphi/fus c. dickersonac was described from limiting the range of C. c. hicinctores to the Tiburon Island (Schmidt, 1922), but this study west side of the riv(^r. A proposed phylogenic has shown that it is not restricted to the island relationship is shown in Fig. 8. Distribution and occurs at least along the nearby Sonoran and relationship is based primarily upon mor- coast. Moiphologv indicates that dickersonac phological characteristics. has not di\erged to any great extent from the The phylogeny of the genus Crotaphytus ap- original stock which gave rise to it and hicinc- parently had its beginnings in southcentral tores of the Great Basin. The canonical analysis United States and northcentral Mexico before

(Fig. 5) indicates that dickersonac diverged the Pleistocene ( Robinson and Van Devender, from the coUaris complex from whicli all mem- 1973). Indications are that the genus, or its bers of the \\c'Stern compk^x of collaris probably immediate ancestral stock, was well established arose at a much later date than the insularis and gave rise to a radiation, which resulted stock. in a division of the genus into the primitive The Great Basin, formed in late Tertiary collared lizard and the leopard lizard branches, (Banta, 1961), did not present a favorable en- perhaps as early as late Pliocene. Although the vironment for the modern collared lizard until fossil record does not support a Pliocene radi- late Pliocene or early Pleistocene, when many ation for the crotaphxtines, if Aciprion Cope is a north-south oriented mountains were formed member of this iguanid line, its origin may then (Axlerod, 1948). The nortlu'rn range of the be extended back into middle Tertiary. We collared lizard population in the Great Basin surmise that the original stock of the coUaris ebbed and enlarged with changing Pleistocene group may not have exhibited the collar so typi- climatic conditions ( Banta, 1961 ) until the arid cal of the group as we recognize it today. We, conditions beginning about 11,000 vears ago therefore, suspect that C reticidatiis is a rem- (Antevs, 1955; Broecker, 1957) permitted its nant of a much older and originally more wide- present northern extension. The modern collared spread population than members of the present- lizard apparently did not invade the Great ly extant collared lizard group. We further Basin or the Great Plains before Pleistocene suspect that this primitive stock continued to (Brattstrom, 1954: Harris and Findlex , 1964; evolve and gave rise to the coUaris group which and Gehlbach, 1965a), although Gilmore (1941) has been so successful in its invasion of the great described the genus Aciprion Cope from the plains, the Rio Grande Valley, the plateaus Oligocene deposits of Wyoming as similar to and mountains of Sonora, Arizona, New Mexico, Crofaphytus. Utah, and the Great Basin since the Pleistocene. The Great Basin collared lizard, C. c. bicinc- We are not certain as to whether the insularis tores, is possibly a more xeric form than mem- group represents a primitive coUaris stock which bers of the eastern coUaris complex. Miller and has recently been replaced in most of the west- Stebbins (1964) report that collared lizards do em distributional areas or if it represents a not occur at elevations above 5700 feet in the modification of the western coUaris group which moimtains of Joshua Tree National Monument, has, through isolation, become modified from the San Bernardino County, California; whereas ancestral stock which also gave rise to the So- members of the coUaris complex are found up to noran and Great Basin populations. The re- 26 Bricham Young Univehsity Science Bulletin

C . c. bic i nctores

C. i in su la r is

ANCESTRAL STOCK

Fig. 8. Proposed phylogenetic relationship of Crotaphytus.

duced collar in the insularis populations is so is additional indication that they are more recent distinct from the collar pattern found in the arrivals than insularis. These circumstances other western subspecies of coUaris that we are again suggest a relict distributional pattern. As- inchned to believe the instilaris group is more suming this to be the general trend in the past primitive than other collared lizards. The in- development of this branch of the genus Crota- sularis group perliaps preceded the coUaris phytus, we could readih' assume that the ebbing groups into the western deserts and at present of the Ice Age allowed the then established represents a reHct which arose shortly after the coUaris stock to extend its distribution north- coUaris ancestral stock was established in the ward and to the west. It occupied extensive more central and southern part of their present areas in central and western United States and distribution. at the same time was undergoing adaptive The ancestral stock of the coUaris complex changes which led to the distinct subspecies appears to have been derived from a stock simi- populations now extant. lar to C. c. fiiscus. Tlicre seems to be little Perhaps the most extensive subspecific group reason to doubt its close relationship to those is that of hicinctores. Even though this group forms now occurring from the Great Plains west- extends from southern California and southwest- ward to the Colorado River. Although these em Arizona northward into Nevada, Utah, and populations have been separated into a series of southern Idaho, its distribution presumably is subspecies, they are all similar in many basic so recent or its genetical makeup so stable that characteristics. Their distinction from the west- subspeciation has not occurred. This is in con- em coUaris subspecies is undoubtedly a series trast to the subspecies cline extending northward of adaptive modifications resulting from the from Chihuahua into southeastem Utah. The changing environment and perhaps isolation. Great Plains populations must still be care'fully The extensive development of the collars of examined, and we expect that they may also hicinctores and dickersonae, plus the fact that show a similar evolutionary development to that they do not overlap the distribution of insularis. observed in the baileyi branch. Biological Series, Vol. 19, No. 4 Western Collared Lizards 27 SUMMARY AND CONCLUSIONS

The heterogeneous nature of the collared with C. c. hicinctores, is considered Crotaphij- lizards west of the Colorado River prompted tus inmdaris vestigium, and the population from this study. Samples from known populations Isla Angel de la Guarda is designated Crota-

of C. c. auriceps, C. c. haileiji, and C. c. juscus phijtus insularis insularis.

were used as a base for defining the western The range of C. c. dickersonae is shown to populations. External moqihology and cranial extend from Isla Tiburon onto the Sonoran morphology were compared by cluster, canonical, coast opposite the island and from Bahia Kino and discriminant analyses. north to El Desemboque. Crotaphijtus c. hicinc- Results of these analyses show a western tores extends from the Yuma, Arizona, area complex of four populations which are distinct north through eastern California, western Ari- from the three base populations of the coUaris zona, Nevada, central Utah (west of the Colo- complex. Members within the western complex rado River), southern Idaho, and southeastern are distinguishable with at least 90 percent re- Oregon. The western range of C. c. haileyi is liabilit\' b\' discriminant anahsis. Because of restricted to central Arizona. The range of

intergradation patterns seen in western Sonora, C. i. insularis is restricted to Isla Angel de la

Mexico, Crotaplii/tus dickersonae is considered Guarda and C. i. vestigium to a narrow strip to be Crotaphijtus coUaris dickersonae. The extending along the eastern mountain slopes

Great Basin population is retained as Crotaphij- from central Baja California, Mexico, to the tus coUaris hicinctores. The population from foothills south of the San Gorgonio Pass, near southern California and Baja California, which Palm Springs, California.

is distinct and has no demonstrated intergrades

ACKNOWLEDGMENTS

We express our gratitude to Drs. A. Lester (CSCLB); Mr. Allan J. Sloan of the San Diego Allen, Arthur O. Chapman, and Stephen L. Society of Natural History (SDSNH); and Drs. Wood for their helpful suggestions in their re- Charles H. Lowe and Michael D. Robinson,

\'iew of this stud\-. Appreciation is expressed University of Arizona (UAZ). They were very to Mr. William Ingram III for statistical sug- helpful during our visits to their respective in- gestions and use of his computer program stitutions and in providing specimens from the decks, and to Dr. Wayne A. Larson, Dr. Melvin museum collections. We also wish to thank Dr. Carter, Dr. Clive D. Jorgensen, Mr. Keith Blake, T. Paul Maslin, University of Colorado (CU); Mr. John Taylor, and Mr. Jerr\- Hintze for statis- Dr. William E. Deullman, University of Kansas tical and computer assistance; to Mr. Donald (KU); Mrs. D. M. Smith, University of Illinois Karr, Dr. Kenneth Larsen, Mr. Jeffrey Smith, Natural History Museum (UIMNH); Drs. James Mr. William Ingram III, and Mr. Robert Mc- A. Peters and George R. Zug, U. S. National Morris, who have accompanied us on extensive Museum (USNM); Dr. Walter R. Quanstrom, collecting trips during the three years from Northwest Nazarene College (NNC); Dr. 1969 to 1971, we express a heartfelt thanks. Robert G. Webb, University of Texas at El Materials were supplied bv the following in- Paso (UTEP); and Dr. John Legler, University stitutions and indixiduals: Drs. Alan E. Leviton of LTtah (UU), and Brigham Young LTniversity and Steven C. Anderson of the California (BYU). Mr. James Davis, Indio, California, Academy of Sciences (CAS and CAS-SU); Dr. provided us live and preserved specimens from David B. Wake, University of California at the foothills on both sides of the Coachella Berkeley (UCMVZ); Dr. John W. Wright and Valley; for this we extend our appreciation. A Mr. Arden Brame of the Los Angeles County National Defense Education Act Title IV Fel- Museum (LACM); Dr. Richard B. Loomis, lowship provided financial support for the California State College at Long Beach senior author from 1968 to 1970. 28 Brigham Young University Science Bulletin

APPENDIX I

Key to the species of Crotaphytus 3a. Two collars present on dorsum of neck, la. Collars on neck absent dorsally and ven- first collar often incomplete dorsally and traUy 2 with spotting in the dorsum between the lb. Collar or collars present on neck dor- two lateral arms of collar collaris sally and/or ventrally 3 3b. Collars present or absent; when present 2a. Supraorbital semicircles absent; body pat- reduced, usually incomplete dorsally and

tern with spots but not reticulated — . ivislizenii without dorsal spots in the first collar; 2b. Supraorbital semicircles present; occasion- body pattern with distinct light cross ally with a vertical black bar on shoulder; bands, often with light elongate spots body pattern boldly reticulated reticulatus between the cross bands iasuZaris

APPENDIX II

Key to subspecies of Crotaphytus collaris and Crotaphy- 4b. Head yellow with yellow extending under tus insularis the chin and caudally to or just past the la. One or more spots in dorsum of the first second collar C. c. auriceps collar; adult males lack a dark pigmented 5a. Body dorsum blue, green or gray, with gular and groin; first collar not united large light dorsal elongate spots; second ventrally; second collar usually extends collar length/snout-vent averages from 0.13 onto upper arm (Fig. 3G-H) 2 to 0.19; number scales from interparietal

lb. Spots absent in dorsum of first collar; to anterior border of first collar 37-43 -.,. C. c. adult males have a dark pigmented gular dickcrsonae and groin and first collar unites ventrally 5b. Body dorsum brown or tan, witli trans- oc nearly so; second collar not extended verse bands of yellow, orange, or white; onto upper arm 5 second collar lengtli/snout-vent length usu- 2a. Interorbitals form two rows, usually with- ally less than 0.14 6 out a fusion of the scales in these two 6a. Second collar meets, or nearly meets dor- rows; supralabials usually 11 or more; sally, usually never more than 10 scales gular region of adult males boldly reticu- separate the two dorsal arms of the second lated 3 collar; first collar separated by 1-9 scales 2b. Interorbitals have one to four scales fused (Fig. 3B) C. c. hicinctores in the interorbital series; supralabials usu- 61). Second collar reduced or absent, usually ally 10 or less; gular region of adult males with distinct white transverse dorsal not boldly reticulated C c. collaris bands; first collar separated by 16 to 40 3a. Body dorsum brown with no trace of plus scales (Fig. 3A) 7 green, head cream colored with no trace 7a. Second collar very reduced or absent;

of yellow C. c fuscus fourth toe subdigital lamellae 39-4.5 C. i. 3b. Body dorsum green or blue; head yel- insularis low or light 4 7b. Second collar less reduced and usually a

4a. Head yellow or white; if \ellow, not portion of it present; fourth toe sub-

extending under the chin or caudally past digital lamellae 33-37 C. i. vestigium supraorbital semicircles C. c. haileyi

LITERATURE CITED

Allen, M. J. 1933. Report on a collection of am- ( Sauria, Iguanidae ) as determined by osteological phibians and from Sonora, Mexico, with and mvological characters. Brigham Young Univ. the description of a new lizard. Occ. Pap. Mus. Sci. Bull., Biol. Sen, 12(3):l-79.

Zool. Univ. Mich., 259:1-15. A.XELROD, D. I. 1948. Climate and evolution in west- Anderson, C. A. 1950. 1940 E. W. Scripps cruise to ern North America during middle Pliocene time. the Gulf of California. Part 1. Geology of the Evolution, 2:127-44. islands and neighboring land areas. Geol. Soc. AxTELL, R. W. 1972. Hybridization between western Amer. Memoir, 43:l-.53. collared lizards with a proposed taxonomic rear- Antevs, E. 1955. Geologic-climatic dating in the rangement. Copeia 4:707-727. West. Amer. Antiquity, 20:317-35. Banta, B. H. 1960. Notes on the feeding of the Avery, D. F., and W. W. Tanner. 1964. The oste- western collared lizard, Crotaphi/tus collaris haileyi ology and myology of the head thora.x Stejneger, Biol., and regions Wasmann J. 18:309-311,

of the obcius group of the genus Sauromalus . 1961. The variation and zoogeography of Dumeril (Iguanidae). Brigham Young Univ. Sci. the lizards of tlie Great Basin. Unpublished Ph.D. Bull., Biol. Ser., 5(3): 1-30. dissertation. Department of Biological Sciences, . 1971. Evolution of the Iguanine lizards Stanford University, Palo Alto, California. 278 pp. Biological Series, Vol. 19, No. 4 Westehn Collared Lizards 29

Trans- 15HATTSTROM, B. H. 1954. Amphihians and reptiles Moore, D. G., and E. C. Buffincton. 1968. from Gvpsiim Cave, Nevada. Bull. S. Calif. Acad. form faulting and growth of the Gulf of California Science, 161:1238-41. Sci., 53:8-12. since late Pliocene. ed. Iowa Broeckeh, \V. S. 1957. Evidence for a major climatic Ostle, B. 1963. Statistics in research. 2nd change close to 11,000 years B.P. Geol. Soc. Amer. State Univ. Press, Ames, Iowa. p. 129, 512-564. and rela- Bull.,' 68:1703-04. Presch, W. 1969. Evolutionary osteology Burt, C. E. 1928. The synonymy, variation, and dis- tionships of the homed lizard genus Phrijnosoma tribution of the collared lizard, Cmtaphijtus collaris (Family Iguanidae). Copeia, (2):250-75. (Say). Occ. Pap. Mus. Zool. Univ. Mich., 196:1-19. Rao, C. R. 1952. Advanced statistical methods in CoNANT, R. 1958. A field guide to reptiles and am- biometric research. John Wiley and Sons, Inc. phibians of the United States and Canada east of New York. p. 166-172. the 100th meridian. Houghton Mifflin Co., Boston. Robinson, M. D., and T. R. VanDevender. 1973. lizards from Wyoming and Nebraska. 366 pp. Miocene CoOLEY, \V. W., AND p. R. Lohnes. 1971. Multivariate Copeia, (4):698-704. data anaUsis. John W'ilev and Sons Inc., New RoBisoN, W. G., and W. W. Tanner. 1962. A com- species of the genus Crotaphijtus York. 364 pp. p;iratiye study of Cope, E. D. 1900. The crocodilians, lizards, and Holbrook (Iguanidae). Brigham Young Univ. Sci. snakes of North America. Ann. Rep. U.S. Nat. Bull., Biol. Ser.. 2{1):1-31. Mus., 1898:153-1294. St;nMiDT, K. p. 1922. The amphibians and reptiles of Di.xoN, 1968. Biomedical computer programs. lower California and the neighboring islands. Bull. W. J. Univ. Calif. Pubs. Automatic Computation, no. Amer. Mus. Nat. Hist. 46:607-707. 2:214a-214t. Smith, H. M. 1946. Handbook of lizards: Lizards of Fitch, H. S., and W. W. Tanner. 1951. Remarks the United States and of Canada. Comstock Pub. concerning the systematics of the collared lizard Co., Ithaca, New York. p. 166-172. (Crotaplu/fiis colla

spreading and transform faulting. Science, 161:- Van Denuurch, J. 1922. The reptiles of western 781-84. North America. Vol. 1, "Lizards," Occ. Pap. Cahf. Lawlor, T. E. 1971. Evolution of Pcromi/scus on Acad. Sci., 10:9.5-98.

, and R. Slevin. 1921. Preliminary diagnosis northern islands in the Gulf of California, Mexico. J. islands in the Gulf San Diego Soc. Nat. Hist. Trans., 16:91-124. of new species of reptiles from of California, Mexico. Proc. Cahf. Acad. Sci., Ser. Mayr, E. 1969. Principles of systematic zoology. Mc- 4, 11:9.5-98. Graw-Hill Book Co., New York. 428 pp. Wells, P. V., and C. D. Jorgensen. 1964. Pleisto- Miller, A. H., R. and C. Sterdens. 1964. The cene wood rat middens and climatic change in lives of desert in Joshua Tree National Mohave Desert: a record of juniper woodlands. Monument. Lhiiv. Calif. Press, Berkeley. 452 pp. Science, 143:1171-74. Montanucci, R. R. 1969. Remarks upon the Crota- Wishart, D. 1968. A Fortran II programme (CLUS- plu/tus- controversv (Sauria:lguanidae). TAN) for numerical classification. Computing Labo- Herpetologica, 25:308-314. ratory, St. Andrew, Fife, Scotland, p. 1-50,