IHESlS This is to certify that the thesis entitled THE FOSSIL StxAKE FAL’NA OF PIT 91, RANCHO LA BREA, LOS ANGELES COUNTY, CALIFORNIA presented by Thomas Cranston LaDuke has been accepted towards fulfillment of the requirements for M. s . degree in Zoology J. Alan Holman Date August 3, 1983 0.7639 MS U is an Affirmative Action/Equal Opportunity Institution 'uoni' - 0,- 1V1ESI_] RETURNING MATERIALS: Place in book drop to LJBRARJES remove this checkout from l—c—s—L your record. FINES will be charged if book is returned after the date stamped below. \ -_.._._———-——’ THE FOSSIL SNAKE FAUNA OF PIT 91, RANCHO LA BREA, LOS ANGELES COUNTY, CALIFORNIA BY Thomas Cranston LaDuke A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Zoology 1983 rig—arr- ABSTRACT THE FOSSIL SNAKE FAUNA OF PIT 91, RANCHO LA BREA, LOS ANGELES COUNTY, CALIFORNIA BY Thomas Cranston LaDuke The snake fauna of the Rancho La Brea asphalt deposits in Los Angeles, California, was previously poorly documented. The present study results from a current, carefully controlled excavation of a single Rancho La Brea site and increases the Rancho La Brea snake list from four to twelve species. Two of these species are new to the fossil record. The snake fauna is very similar to that of the Los Angeles region today and indicates an only slightly moister climate. It is suggested that the stability of the North American herpetofauna compared to the mammalian fauna may be related to the ability of small ectotherms to maintain viable populations in isolated refugia which will not support populations of large endotherms. ACKNOWLEDGMENTS I would like to thank my major professor Dr. J. Alan Holman for providing the materials and inspiration to perform this study and for guidance and patience during its under- taking. Thanks also go to Dr. Marvin M. Hensley and Dr. Richard J. Seltin who also served on my committee and reviewed the final draft and Leslie P. Fay and Karen Mudar for useful comments on rough drafts. I thank Dr. W. A. Akersten for the loan of fossil material from Pit 91 and Dr. B. H. Brattstrom, Dr. R. A. Nussbaum, Dr. D. A. Rossman, Dr. J. M. Savage, and Dr. J. W. Wright for the loan of comparative specimens. Finally, special thanks go to Rosemarie Attilio for her skillful rendition of the figures. ii LIST OF TABLES . . LIST OF FIGURES . INTRODUCTION . . . SITE DESCRIPTION . METHODS O O O O . SPECIES ACCOUNTS . DISCUSSION . . . . SUMMARY . . . . . APPENDIX . . . . . LIST OF REFERENCES TABLE OF CONTENTS iii Page iv 42 49 50 53 Table 1. LIST OF TABLES Page Numbers of elements of Rancho La Brea snake species in decreasing order of abundance. Hyphens indicate undetermined juvenile material. . . . . . . . . . . . . . . . . . . . 44 Habitat preferences of the Pit 91 snakes, based on the relative abundance of modern species in that habitat. H--highest abundance, I--interme- diate, L--lowest abundance, -not expected . . . 45 iv LIST OF FIGURES Figure Page 1. Partial left maxilla of Diadophis punctatus (LACM-VP R47445); from Pit 91, Rancho La Brea (ventral View, 20x) . . . . . . . . . . . . . . . 8 2. Comparison of vertebrae of Diadophis punctatus (LACM-VP R5196l) (A) dorsal View; and Tantilla sp. (LACM-VP R44374) (B) dorsal view; from Pit 91, Rancho La Brea (20x) . . . . . . . . . . . . lO 3. Vertebra of Coluber mormon (LACM—VP R51950—D) (A) dorsal view, (B) posterior view; from Pit 91, Rancho La Brea (10x) . . . . . . . . . . . . . . l7 4. Vertebra of Masticgphis lateralis (LACM-VP R16976) (A) dorsal view, (B) ventral View, (C) anterior view, (D) posterior view, (B) lateral View; from Pit 91, Rancho La Brea (10x) . . . . . . . . . . l9 5. Comparison of Pituophis melanoleucus (LACM-VP R21401) (A) ventral view, (B) lateral view; and Arizona elegans (LACM-VP R51932) (C) ventral View, (D) lateral View; from Pit 91, Rancho La Brea (10x) . . . . . . . . . . . . . . . . . . . . . . 25 Figure Page 6. Comparison of Lampropeltis getulus (LACM-VP R40284) (A) dorsal View, (B) lateral View, (C) ventral View; and Rhinocheilus lecontei (LACM-VP R51950-C) (D) lateral view, (B) ventral view; from Pit 91, Rancho La Brea (10x) . . . . . 31 Vertebra of Thamnophis of. T. couchi (LACM-VP R25869) (A) dorsal view, (B) ventral View, (C) anterior view, (D) posterior view, (B) lateral view; from Pit 91, Rancho La Brea (10x) . . . . . 40 vi INTRODUCTION The most famous Late Pleistocene fossil vertebrate locality in North America and probably the most important in terms of the information it has and continues to yield about that time period, is Rancho La Brea of Los Angeles, California. This site has produced thousands of fossils in a superb state of preservation providing one of the most detailed accounts of a Late Pleistocene biological community known. Moreover, its fauna is the type material for the Rancholabrean Provincial Land Mammal Age. Early studies of Rancho La Brea concentrated on the spectacular large mammal and avian faunas produced (see Stock, 1956 for a review and bibliography) but largely neglected microvertebrates, invertebrates, microbotanical remains, taphonomy, and details of stratigraphy. In the 1940's, D. W. Pierce developed a technique for recovering minute fossils from the asphaltic matrix (Stock, 1956). Pierce's studies were mostly concerned with insect remains, but microvertebrates were also recovered (Brattstrom, 1953b). The geology and stratigraphy of Rancho La Brea have recently been detailed by Woodard and Marcus (1973). Numerous accounts of dates and dating techniques for Rancho La Brea material have been published (Howard, 1960; Berger and 1 Libby, 1966, 1968; Ho and Marcus, 1969; and McMenamin et. al., 1982). Brattstrom (1953b) published the first comprehensive treatment of amphibians and reptiles from Rancho La Brea, based partly on Pierce's collections. Snakes were poorly represented in this collection and Brattstrom anticipated that a greater diversity would be revealed by further study. The present study is based on the snake remains from a recent excavation and documents that greater diversity. SITE DESCRIPTION Rancho La Brea is located in the Los Angeles Basin within the city of Los Angeles. It is Late Pleistocene in age and is the type section of the Rancholabrean Land Mammal "Age" (Savage, 1951). The fossil deposits lie within the Palos Verdes Sand which is here impregnated with asphalt. Radiocarbon dates for Rancho La Brea vary from about 4,000 to over 40,000 years B.P. (McMenamin et. al., 1982). The reader is referred to Stock (1956), Woodard and Marcus (1973), and Shaw (1982) for detailed discussions of location, age, geology, stratigraphy, and history of excavation. Early workers postulated that the tar deposits repre- sent previously fluid, viscous pools in which animals were trapped and entombed. These pools were conceived of as dynamic fluid bodies that underwent convective overturn which disarticulated skeletons, wore bone surfaces, and obliterated stratigraphy (Stock, 1956). Recent studies (Woodard and Marcus, 1973) have shown that a gross strati- graphy exists and that much of the fossil material is entombed in stream deposits that were inundated by asphalt. In 1969, a study was initiated at Rancho La Brea to improve understanding of aspects of the site that were not well documented by previous investigators. A single pit (LACM 6909, Pit 91 of early workers), originally set aside for display purposes, is being reexcavated with careful attention to taphonomy and the recovery of microfossils (Shaw, 1982). This operation, known as the Rancho La Brea Project (RLP), will substantially increase knowledge of the Pleistocene environment at Rancho La Brea. Pit 91 is roughly 28 feet square (approximately 2.6 sq. m.) and is divided by a grid of one yard squares (approximately 0.83 sq. m.). Each six inch depth interval (approximately 15 cm) of a square is designated by a locality number. Disturbed matrix is disregarded. Portions of the southeast corner of lower levels (below 7'6" below datum (approximately 2.3 m.)) are composed of stream drift (Cox and Jefferson, in litt,, 1983), indicating that this deposit is partly composed of natural stream sediments that should have a reasonably sound stratigraphy. Radiocarbon dates for Pit 91 range from 25,100 to over 40,000 B.P. on wood and bone collagen (Cox and Jefferson, ibid.) revealing that Pit 91 is relatively old among Rancho La Brea deposits. METHODS Identification of snake skeletal elements to genus or species depends on recognition of ranges of variability of subtle characters and proportions and requires the use of extensive comparative collections. Holman (1981, Appendix) presents a detailed guideline for identifying North American Pleistocene snakes. Auffenberg (1963), Brattstrom (1967), and Meylan (1982) also treat this topic. Although many of the diagnostic characters presented herein have been borrowed from the literature, others are original. As in previous studies of fossil snakes, identification was restricted to vertebrae from the middle section of the precaudal vertebral column and to skull elements. Compara- tive specimens are listed in Appendix I. Anatomical termi- nology of vertebrae is derived largely from Auffenberg (1963) and Hecht (1982); that of skull elements from Cundall (1981). Ages and locations of fossil localities discussed in the text are listed below by state. Arizona Curtis Ranch, Irvingtonian (Brattstrom, 1955a). Emery Falls Canyon, Rampart Cave, and Vulture Canyon; Rancholabrean (Van Devender, Phillips, and Mead, 1977). New Water Mountains, Tucson Mountains, Wellton Hills, and Wolcott Peak; Rancholabrean (Van Devender and Mead, 1978). California McKittrick Asphalt, Hawver Cave; Rancholabrean (Brattstrom, 1953a). Rancho La Brea; Rancholabrean (Brattstrom, 1953b, 1958a). Carpinteria Asphalt; Rancholabrean (Brattstrom, 1955b). Mescal Cave; Rancholabrean (Brattstrom, 1958a). Costeau Pit, Newport Beach Mesa; Rancholabrean (Hudson and Brattstrom, 1977). Redtail Peaks, Tunnel Ridge, Whipple Mountains; Rancholabrean (Van Devender and Mead, 1978). Nevada Smith Creek Cave; Rancholabrean (Brattstrom, 1976). Gypsum Cave; Rancholabrean (Brattstrom, 1954, 1958b). New Mexico Howell's Ridge Cave; Rancholabrean (Van Devender and Worthington, 1977). Conkling Cave, Shelter Cave; Rancholabrean (Brattstrom, 1964a). SPECIES ACCOUNTS A checklist of the snake fauna of Pit 91 is presented below followed by an annotated list. The systematic arrangement of higher taxa follows that of Dowling and Duellman (1978). Modern distributions are from Stebbins (1966). Family Colubridae Subfamily Xenodontinae Diadophis ppnctatus (Linnaeus) Hypsiglena torquata Gunther Tantilla sp. Subfamily Colubrinae Arizona elegans Kennicott Coluber mormon Baird and Girard Masticophis lateralis (Hallowell) Pituophis melanoleucus Daudin Lampropeltis getulus (Linnaeus) Rhinocheilus lecontei Baird and Girard Subfamily Natricinae Thamnophis of. T. couchi (Kennicott) Thamnophis sirtalis Linnaeus Family Viperidae Subfamily Crotalinae Crotalus viridis Rafinesque Family Colubridae Subfamily Xenodontinae DiadoPhis Baird and Girard Maxilla: The maxilla of Diadophis has a posterior diastema followed by two compressed, blade—like teeth. The shaft of the maxilla is flexed laterally near the anterior end of the ectopterygoid process (Figure 1). Sonora and Chionactis have shorter maxillary diastemata and weaker flexures. Other small southwestern colubrids lack a similar combination of diastema and flexure. Vertebrae: The vertebrae of Diadophis have been described by Auffenberg (1963). They are relatively long and depressed with a cunate or oblong hemal keel. The neural spine is low and long, but not obsolete posteriorly. Auffenberg (1963), Holman (1967, 1977b), Hill (1971), and Meylan (1982) discuss differences between the vertebrae of Diadophis and Carphophis, Rhadinaea, Stilosoma, and/or Tantilla. Carphophis may generally be distinguished from Diadophis by its lower neural spine which may be obsolete posteriorly. In Rhadinaea and Stilosoma the accessory processes are directed approximately 90 degrees from the long axis of the centrum, while in Diadophis, they are directed obliquely anterior. Although there is some overlap, Tantilla may often be distinguished from Diadophis by its lower neural spine which may be obsolete posteriorly, Figure 1. Partial left maxilla of Diadophis punctatus (LACM-VP R47445); from Pit 91, Rancho La Brea (ventral view, 20x). postzygapophyses that are longer than wide (only slightly longer than wide in Diadophis), and the broadly rounded posterior notch of its neural arch (acute notch in Diadophis, Figure 2). Diadophis punctatus (Linnaeus) Referred material: LACM-VP R47445, partial left maxilla; R154l9-C (2), R15554, R15694, R18293, R20278, R21436-B, R21526, R24307, R44507, R51578, R519l9-C, R51921-B (8), R51928, R51932-D, R51956-B (5), R51961 (14), 41 pre- caudal vertebrae. Diagnosis: This material does not differ noticeably from modern Diadophis punctatus. It differs from Q. elinorae Auffenberg in its longer vertebrae with more depressed neural canals, weaker subcentral ridges, lower neural spines, and broader hemal keels. The neural spine of D. elinorae lacks an anterior overhang, a variable character in Q. punctatus. Comments: This snake prefers relatively moist habitat ranging from forest and woodland through Chaparral and prairie. It moves into arid habitat only along riparian corridors (Stebbins, 1966; Van Devender and Worthington, 1977). Q. punctatus occurs in the Los Angeles area today (Stebbins, 1966). This is the first record of Diadophis from the California Pleistocene. The only other south- western fossil record of D. punctatus is from Howell's Ridge Cave, New Mexico (Van Devender and Worthington, 1977). 10 Figure 2. Comparison of vertebrae of Diadophis punctatus (LACM—VP R51961) (A) dorsal View; and Tantilla sp. (LACM-VP R44374) (B) dorsal View; from Pit 91, Rancho La Brea (20x). 11 Hypsiglena Cope Vertebrae: Hypsiglena torquata vertebrae were described by Van Devender and Mead (1978). They are relatively short, wide, and depressed. The neural spine is medium to low in height, and relatively short. The hemal keel is moderately narrow and strong. Hypsiglena vertebrae are much shorter than those of Diadophis, its neural spine is generally less elongate, and the hemal keel much less flattened and widened. Hypsiglena may be distinguished from Phyllorhynchus and Rhinocheilus by its more depressed neural arch. The zygapophyses of Hypsiglena are larger and are produced laterally more than those of Sonora, Chionactis, and Ficimia. Hypsiglena torquata Gunther Referred material: LACM-VP R14794, R15419—D, R20135, R31273, R35943, R50143, R61921-C (3), R51934-C, R51956-C (3), R51962 (5), 12 precaudal vertebrae. Diagnosis: These vertebrae are assigned to H. torquata based on zoogeographic grounds (Stebbins, 1966). Comments: The habitat of this snake ranges from desert to woodland including Chaparral, creosote, sagebrush, grass- lands, and grass-sedge meadows (Stebbins, 1966; Conant, 1975; 12 Wright and Wright, 1957). H. torquata is found in the Los Angeles area today (Stebbins, 1966). Records of H. torquata from the Southwest include Redtail Peaks and Whipple Mountains in California (Van Devender and Mead, 1978); Wolcott Peak, New Water Mountains (Van Devender and Mead, 1978), and Vulture Canyon (Van Devender, Phillips and Mead, 1977) in Arizona; Smith Creek Cave (Brattstrom, 1976) in Nevada; and Howell's Ridge Cave (Van Devender and Worthington, 1977) in New Mexico. Tantilla Baird and Girard Vertebrae: These minute bones are elongate and highly depressed. Their neural spines are low and often obsolete posteriorly. They are most similar to Diadophis vertebrae, but may be distinguished by characters discussed under that genus (Figure 2). Tantilla sp. Referred material: LACM-VP R43748, R44374, R44617, R51919-D, R51965-A, five precaudal vertebrae. These speci- mens are not assigned to species due to the high degree of morphological similarity in this diverse genus, and to a lack of comparative material. Comments: Tantilla planiceps is the only species found in the vicinity of Rancho La Brea today (Stebbins, 13 1966). It is known from woodland, grassland, Chaparral, and desert habitats. This is the first record of Tantilla from the California Pleistocene. Van Devender and Worthington (1977) reported Tantilla sp. from Howell's Ridge Cave, New Mexico. Subfamily Colubrinae Arizona Kennicott Vertebrae: The vertebrae of this monotypic genus have previously been described by Van Devender and Mead (1978). They are relatively short and broad with a moderately high neural arch and a relatively wide neural canal. The neural spine is medium in height. The condyle is moderately wide and the subcentral ridges are moderately developed. The hemal keel is distinct, low, and spatulate. Arizona vertebrae most closely resemble those of Pituophis. The criteria for distinguishing these two genera will be discussed under Pituophis. Holman (1963) discussed the distinction between A. elegans and several species of Lampropeltis. Arizona elegans Kennicott Referred material: LACM-VP R42025, R51932-A (2), R51951, four precaudal vertebrae. 14 Comments: A. elegans generally prefers open areas with Chaparral, sagebrush, or grassy vegetation, or barren deserts. It is found in the Los Angeles area today (Stebbins, 1966). This species has not previously been reported from the Pleistocene of California. However, it is known from New Water Mountains and Wellton Hills in Arizona (Van Devender and Mead, 1978) and Howell's Ridge Cave (Van Devender and Worthington, 1977) in New Mexico. Coluber Linnaeus Maxilla: Coluber (n=36, 20 individuals, two species) has 13—17 maxillary alveoli and lacks a diastema. It may be distinguished from Masticophis by its shorter palatine process which spans roughly two alveoli. Masticophis (n=30, 16 individuals, five species) has 16-20 maxillary alveoli and a palatine process that generally spans two and one-half to three alveoli. Coluber may be distinguished from Pituophis and Arizona as described under Pituophis. Rhinocheilus, Diadophis and Salvadora all possess diastemata. Compound Bone: Coluber has prearticular and surangular crests of roughly equal height. This character is shared with Masticophis, Rhinocheilus, and some Lampropeltis getulus. In Rhinocheilus the crests are lower than in Coluber. The ventral border of the compound bone of Masticophis becomes very narrow below the quadrate 15 articulation but is wider here in Coluber. Lampropeltis getulus has a narrower mandibular fossa than Coluber. Vertebrae: The vertebrae of Coluber have been discussed by Holman (1962), Auffenberg (1963), and Meylan (1982). Coluber vertebrae are elongate with long, high neural spines and prominent epizygapophyseal spines. Coluber and Masticophis vertebrae may be separated from most other North American colubrid genera by their relatively greater lengths. Opheodrys and Salvadora also have relatively elongate vertebrae, but they may be dis- tinguished by characters given by Holman (1962, 1970, 1976). Coluber and Masticophis vertebrae are often confusingly similar, hence, separation at the generic level is impracti- cal. A consideration of identifiable subgroups is presented below. North American Coluber and Masticgphis may usually be divided into two groups on the basis of vertebral morphology. One group includes such forms as Q. constrictor foxi (n=3), g. g. priapus (n=8), and M. flagellum ssp. (n=11). These forms tend to have moderately high, robust vertebrae, high neural spines, and very prominent epizygapophyseal spines and accessory processes. The second group, including 9. g. flaviventris (n=10), Q. mormon (n=1), M. lateralis (n=4), and M. taeniatus (n=4) has vertebrae that are more depressed, with lower neural spines, less prominent epizygapophyseal spines, and long, yet thinner, accessory processes. 16 Coluber from the second group is distinguishable from M. taeniatus and M. lateralis by zygapophyses that are not produced laterally and have small facets and, frequently, by their elevated postzygapophyseal facets (compare Figures 3 and 4). Coluber mormon Baird and Girard Referred material: LACM-VP R41901, left maxilla; Rl4208, right compound bone; R10066, Rl4304, R154l9-E (7), R15992, Rl7701, R19474, R2224l, R25869, R39737, R47324, R51923-C (2), R51932—C, R51934-B, R51938-B (6), R51950—D, 27 precaudal vertebrae. Diagnosis: Q. mormon vertebrae may usually be distin- guished from Q. constrictor flaviventris vertebrae by their slightly larger zygapophyses, less prominent epizygapophyseal spines, and blunter accessory processes. These remains are referred to Q. mormon on these bases, as well as their relatively smaller size and geographic location. Comments: Fitch, Brown, and Parker (1981) have recently restored Q. mormon to species status. Stebbins (1966) indicates that the western forms of Coluber prefer Open areas including semiarid to moist habitat, but avoid Open desert. g. mormon is found in the Los Angeles area today. Brattstrom (1953b; 376) identified "one partly damaged lower jaw... from Rancho La Brea as "Coluber sp. (sensu 17 Figure 3. Vertebra of Coluber mormon (LACM—VP R51950-D) (A) dorsal view, (B) posterior View; from Pit 91, Rancho La Brea (10x). 18 Figure 4. Vertebra of Masticophis lateralis (LACM-VP R16976) (A) dorsal view, (B) ventral View, (C) anterior View, (D) posterior view, (B) lateral view; from Pit 91, Rancho La Brea (7.7x) . 19 I- 1.. . .335». .‘u . . 9m.“ “UM-"fl ‘.I| .'\ N"! . a; . fl . , . O ' ’I‘.‘l‘.‘l"~ . y ' . .5":- Figure 4. 20 latu)." Vertebrae from Rancho La Brea were later assigned to Coluber constrictor (Brattstrom, 1958a). The only other record of Coluber from the California Pleistocene is a single vertebra from the Carpinteria Asphalt deposits (Brattstrom, 1955b). Southwestern Pleistocene locality records of Coluber also include Curtis Ranch in Arizona (Brattstrom, 1955a), Smith Creek Cave in Nevada (Brattstrom, 1976), and Shelter and Conkling Caves in New Mexico (Brattstrom, 1964a). It is possible that the Carpinteria and Smith Creek Cave records represent Q. mormon since they occur within or near its present range. The other records are extralimital for modern Coluber and may represent either species. Masticophis Baird and Girard Identification of Coluber and Masticophis is discussed under Coluber. Masticophis lateralis (Hallowell) Referred material: LACM-VP R16976, R21016, two precaudal vertebrae. Diagnosis: M. lateralis may be distinguished from M. taeniatus by its narrower neural canal, slightly smaller epizygapophyseal spines, and thicker neural arch laminae in posterior view. Comments: Stebbins (1966) describes this species as 21 an inhabitant of the foothills and, occasionally, forested regions and mountains. It prefers Chaparral associations but also frequents grassy, brushy, or rocky areas, especially areas near freshwater (Stebbins, 1966; Wright and Wright, 1957). M. lateralis occurs in the Los Angeles area today (Stebbins, 1966). This is the first report of Masticophis lateralis from the fossil record. A representative fossil vertebra is illustrated in Figure 4. Coluber or Masticophis Referred material: LACM-VP R12166, R15508, R15887, R16126, R22040, R22565, R29l48, R37999, R4124l, R41386, R48181, R48935, R51917, R519l9-E (2), R51921—D (2), R51932-G (3), R5194l (2), R51948 (3), R51954 (3), R51965-D, R51966 (6), 35 precaudal vertebrae. Diagnosis: These vertebrae are either too fragmentary to assign to species or fall within the range of overlap of these two genera. They may, however, be assigned to this category and probably represent Q. mormon, M. lateralis, or both. 22 Pituophis Holbrook Maxilla: The maxilla of Pituophis has 17 (rarely l6) alveoli, a palatine process which is placed relatively posteriorly and a long ectopterygoid process. The only other genera with similar alveolar counts are Masticophis and Coluber with 17-19 and 14-16, respectively. Pituophis may be separated from these by the length of its ectopterygoid process which spans four alveoli. In Masticophis and Coluber, it spans only two or two and one- half. The maxilla of Pituophis is also distinctive in that the palatine process occurs at a point about two-fifths down the length of the bone. This character is shared with Arizona, but here the process lies opposite the sixth alveolus, while in Pituophis it lies opposite the eighth or part of the seventh and eighth alveoli. Pterygoid: Pituophis and Arizona are the only medium to large colubrid genera observed in which the pterygoid tooth row extends less than half the length of the bone. The two differ in that the dorsal crest is much better develOped in Arizona. Compound Bone: In Pituophis, the prearticular crest is more highly elevated than the surangular crest. They are about equal in height in Coluber, Masticophis, Rhinocheilus, and some specimens of Lampropeltis getulus. Thamnophis, Salvadora, and Arizona have an elevated prearticular crest as in Pituophis, but it is more highly arched. In some 23 specimens of LamprOpeltis getulus, the prearticular crest is elevated, but in all specimens observed, the dorsal opening of the mandibular fossa is narrower than in other large colubrids. Vertebrae: Criteria for identification of Pituophis vertebrae have been discussed by Auffenberg (1963), Holman (1965), Van Devender and Mead (1978), and Meylan (1982). The vertebrae are typically slightly elongate with weak subcentral ridges, spatulate hemal keels, and relatively high neural canals. Epizygapophyseal spines are obsolete. Among western colubrids, Pituophis vertebrae are most readily confused with Arizona, Elaphe, or Lampropeltis getulus. Pituophis vertebrae are generally more elongate than those of Arizona (Figure 5). However, this character varies with the age of the individual such that very young specimens overlap strongly with mature Arizona. The weak subcentral ridges will distinguish most Pituophis from Lampropeltis getulus and some species of Elaphe. E. getulus also has distinct fossae lateral to the hemal keel. This character is only present in the last few precaudal vertebrae of Pituophis. The hemal keel of Pituophis is moderately developed and spatulate. In E- getulus, this keel is much stronger, very thin, and spatulate to gladiate. Pituophis has a relatively narrower centrum in ventral View than do Elaphe or Arizona. Finally, the condyles of small to medium-sized Pituophis are much larger than those of similar 24 Figure 5. Comparison of Pituophis melanoleucus (LACM-VP R21401) (A) ventral View, (B) lateral View; and Arizona elegans (LACM—VP R51932) (C) ventral View, (D) lateral View; from Pit 91, Rancho La Brea (10x). 25 .m musmflm ’nu‘fifl-CIH- o..- ..........u.. 35.x . "O o Ian I. loo-cu...- .. l .0... \ 000-0.. on. of. OJ '0’. . A... 0'00... . . ... CON.- oils o\- \ I oo- o. a o I .r a . n‘ . '00.. u... on s. I... on ‘4. .h! ‘I‘ I . , . . 0 .II .. .x . ... l... a...- c.». o .u.. . .. \7 a. u C. onto-“‘96. n . I 63.5.5: . n . {JG «v u. .0 l. '0‘ m i C.‘ .0. Q. .- .‘fi‘LI‘ ‘0-(I- .‘.”‘I» o‘coOl: .- 0.... .u. on. c o. O o o I 26 sized Lampropeltis, Elaphe, or Arizona. This character appears to change ontogenetically as large specimens do not share this exaggerated condyle. Pitquhis melanoleucus Daudin Referred material: LACM-VP R4264l, R51945, R51958, three left maxillae; R51922, partial right pterygoid; R2446l, right compound bone; R14266, Rl4923, Rl4971, R15041, R154l9-A (8), R16100, R16307, R16405, Rl6584, Rl6852, Rl7607, R17704, R18047, R18786, Rl8865, R18907, Rl9178, R20048 (2), R20301, R21208, R21401, R21436—A, R21825, R22018, R22109, R23008, R23l60, R23505, R23652, R23847, R23848, R23935, R2447l, R24645, R25652, R25957, R25978, R26982, R27718, R30214, R30447, R30930, R32393, R33306, R34743, R37225-A, R377l6, R39618, R42621, R47713, R48877, R50543, R519l9-A, R51921-A (2), R51923-A (3), R51926, R5193l (7), R51933, R51934-A (2), R51939, R51940 (2), R51950-A, R51954 (5). R51956-A, R51964 (2), 90 precaudal vertebrae. Diagnosis: The fossil material is indistinguishable from P. melanoleucus. Comments: 2. melanoleucus is known from most habitat types and to relatively high altitudes throughout its wide range. It is found in the Los Angeles area today (Stebbins, 1966). Brattstrom (1953b) reported Pituophis melanoleucus from Rancho La Brea. Other Late Pleistocene localities in 27 California where P. melanoleucus has been reported include the McKittrick Asphalt (Brattstrom, 1953a) and Costeau Pit (Hudson and Brattstrom, 1977). Other southwestern locality records include Emery Falls Canyon and Rampart Cave (Van Devender, Phillips, and Mead, 1977) and Wolcott Peak (Van Devender and Mead, 1978) in Arizona; Gypsum Cave (Brattstrom, 1954, 1958b) and Smith Creek Cave (Brattstrom, 1976) in Nevada; and Shelter and Conkling Caves (Brattstrom, 1964a) anui Howell's Ridge Cave (Van Devender and Worthington, 1977) in New Mexico. Lampropeltis Fitzinger Two groups within the genus Lampropeltis may be dis- tinguished osteologically. E. getulus and E. calligaster form one group, E. pyromelana, E. triangulum, and E. zonata the other (M. mexicana not available). Maxilla: Maxillae of the E. getulus "group" are short, robust, generally contain 13-15 (n=9) alveoli and lack a diastema. Their palatine process is long, narrow, directed posteriorly, and located near the anterior end of the maxilla. Anteriorly, the maxilla is high and narrow. Pituophis and Arizona have longer, thinner maxillae with palatine processes located more posteriorly. Masticophis and Coluber also have longer maxillae, with wider palatine processes that are directed medially. The anterior portion 28 of the maxilla of Rhinocheilus and the g. triangulum "group" is lower than in the L. getulus "group." Vertebrae: E. getulus vertebrae have been discussed by Auffenberg (1963), Holman (1965), Van Devender and Mead (1978), and Meylan (1982). They are of moderate length with well-developed subcentral ridges; strong, thin hemal keels; well-developed fossae lateral to the hemal keels; zygosphenes that are flat anteriorly; wide neural arches; neural spines that are thickened dorsally, with anterior and posterior overhangs; and accessory processes that are often thickened distally. Vertebrae of the E. getulus "group" may be distinguished from those of Pitquhis as described under that section. Arizona and Elaphe are readily distinguished from Lampropeltis by their shorter vertebrae which lack distinct fossae lateral to the hemal keel. Vertebrae of the g. triangulum "group" have weaker hemal keels that lack fossae lateral to them. They also lack the thickened neural spines and accessory processes. The vertebrae of Rhinocheilus lecontei are similar to the E. getulus "group" but are distinguished by being antero-posteriorly shorter, having wider hemal keels with shallower lateral fossae, smaller zygapophyses, and shorter neural spines (Figure 6). 29 Lampropeltis getulus (Linnaeus) Referred material: LACM-VP R51925, partial left maxilla; Rl4948, R15419-B (3), R19994, R20151, R28575, R37225-B, R50284 (3), R519l9-B, R51923-B, R51932-B (2), R51938-A (2), R51950-B (2), R51960, 20 precaudal vertebrae. Diagnosis: The posterior end of the maxilla is broken off. It is small but closely matches a modern juvenile E. getulus (MSU 2476). The vertebrae differ from E. calligaster in having stronger hemal keels with deeper lateral fossae. Comments: E. getulus is a very widely ranging species throughout the southern half of the United States. It occupies most terrestrial habitats within its range (Wright and Wright, 1957). It is found in the Los Angeles area today (Stebbins, 1966). Brattstrom (1953b, 1958a) reported E. getulus from Rancho La Brea. It has also been reported from the McKittrick Asphalt (Brattstrom, 1953a), Newport Beach Mesa (Hudson and Brattstrom, 1977), and Redtail Peaks and Tunnel Ridge (Van Devender and Mead, 1978) in California. Other sites that have yielded E. getulus include Rampart Cave (Van Devender, Phillips and Mead, 1977) in Arizona, Gypsum Cave (Brattstrom, 1954, 1958b) and Smith Creek Cave (Brattstrom, 1976) in Nevada, and Shelter Cave (Brattstrom, 1964a) and Howell's Ridge Cave (Van Devender and Worthington, 1977) in New Mexico. 30 Figure 6. Comparison of LaMprOpeltis getulus (LACM-VP R40284) (A) dorsal View, (B) lateral View, (C) ventral View; and Rhinocheilus lecontei (LACM-VP R51950-C) (D) lateral view, (E)ventral View; from Pit 91, Rancho La Brea (7.7x). 31 32 Rhinocheilus Baird and Girard Vertebrae: Vertebrae of this monotypic genus have been described by Hill (1971) and Van Devender and Mead (1978). Rhinocheilus vertebrae are short; their zygapophyses are small and moderately produced laterally; the hemal keels are fairly strong, with shallow fossae lateral to them; the neural spines are short, medium-height, thickened dorsally, and have strong anterior and posterior overhangs. This genus most closely resembles the E. getulus "group," but may be diStinguished from it as described under that account. Hypsiglena vertebrae are more depressed and have relatively larger, more laterally produced zygap0physes. Rhinocheilus lecontei Baird and Girard Referred material: LACM-VP R51950-C (3), three precaudal vertebrae. Comments: Stebbins (1966) lists deserts, prairies, and brushland as habitats for this nocturnal species. It is found in the Los Angeles area today (Stebbins, 1966). B. lecontei was reported from Tunnel Ridge in California (Van Devender and Mead, 1978). Other Southwestern records include Rampart Cave (Van Devender, Phillips, and Mead, 1977) and Tucson Mountains, Wolcott Peak and New Water Mountains (Van Devender and Mead, 1978) in Arizona; and 33 Howell's Ridge Cave (Van Devender and Worthington, 1977) in New Mexico. Subfamily Natricinae Thamnophis Linnaeus Maxilla: Tooth-bearing bones of members of this genus generally have more alveoli per unit length than similar bones of colubrines. Thamnophis maxillae have medium-sized ectopterygoid and palatine processes and have 21-25 teeth. Anteriorly, the shaft of the bone curves medially. Thamnophis maxillae lack diastemata. Thamnophis is the only natricine genus presently found in the extreme southwestern United States (Stebbins, 1966). Its maxillae may be distinguished from colubrine genera in this region as described under other generic accounts (see Coluber, LamproPeltis, and Pituophis). Palatine: The palatine of Thamnophis is elongate and has 14-16 alveoli. The choanal process is relatively broad and strongly recurved. The maxillary process is long, narrow, and rounded in cross section. The pterygoid process is usually strongly forked. Lampropeltis and Rhinocheilus palatines have fewer teeth than in Thamnophis. Arizona, Coluber, Masticophis, and Pitquhis have choanal processes that are only weakly recurved, or not at all. The maxillary process is shorter in Arizona and Masticophis and flatter in Coluber. 34 Pterygoid: The pterygoid of Thamnophis has 20-34 teeth (n=35, 10 species represented). The tooth row extends nearly to the posterior end of the bone. The quadrate process is relatively short. LamprOpeltis, Masticophis, Rhinocheilus, and Salvadora have longer, narrower quadrate processes. The pterygoid of Coluber is more strongly constricted posterior to its ectopterygoid articulation. Dentary: The dentary of Thamnophis has 27-31 alveoli. Its Meckel's groove usually closes opposite the sixth alveolus. Thamnophis dentaries are distinguishable from most southwestern colubrines by the higher alveolar counts of the former. Compound Bone: In Thamnophis, the prearticular crest is more elevated than the surangular crest and is relatively strongly arched. The compound of Pituophis differs as described under that genus. In Coluber, Masticophis, and Rhinocheilus the surangular and prearticular crests are roughly equal in height. Lampropeltis getulus has a narrower mandibular fossa than Thamnophis. Salvadora has a distinctly shorter compound bone with a relatively higher surangular crest than in Thamn0phis. Arizona has a relatively short mandibular fossa and a shorter retroarticular process than Thamnophis. Vertebrae: Vertebrae of this genus have been described by Auffenberg (1963). Identification of various genera of 35 natricines has been discussed by several authors: Auffenberg (1963), Brattstrom (1967), Holman (1962, 1977a), and Meylan (1982). Thamnophis vertebrae are generally elongate with short, posteriorly directed hypapophyses and neural spines slightly longer than high. Seminatrix, Storeria, Tropidoclonion, and Virginia are elongate, but have low to obsolete neural spines (Auffenberg, 1963). Nerodia and related genera usually have shorter vertebrae with higher, shorter neural spines, but some overlap with Thamnophis exists. Thamnophis sp. Referred material: LACM-VP R41286 fragmentary left maxilla; R47727, R51936, R51959, three fragmentary right maxillae; R41212 left palatine; Rl7455, right pterygoid; R51947, distal right dentary; R154l9, right compound bone; R10074, R12542, R12783, R13475, R13799, R14154 (2), R14477, Rl4629, R14794 (2), Rl4997 (2), R15036, R15233, R154l9-G (45), R15453, R1546l, R15475, R15508 (2), R15600, R15662, R15680, R15760, R15771, R15807, R15810, R15946, R15984, R15992 (2), Rl6002, Rl6058, Rl6080, R16087, Rl6089, R16104, R16113, R16124 (2), Rl6132, Rl6152 (3), R16153, Rl6301, Rl6362, Rl6367, R16437, Rl6521, Rl6584 (2), R16966 (3). Rl7246, R17318, Rl7363, Rl7453, Rl7476, Rl7638, R17702, Rl7722, R17862, R18115, R18l75, R18228, R18299, R1833l, R18345, R18798, R18993, Rl9067, R19142, R19682, R20361, 36 R204l9, R20893, R20989, R21208 (12), R21311, R21373, R21436-C, R21562, R21701, R21888, R22018, R22095, R22193, R22492, R22598, R23665, R23692, R23863, R23927, R24009, R24465, R24471, R24579 (3), R24621, R25300, R25487, R26007 (3), R26024 (2), R26910 (2), R27090, R27293, R27352, R27637, R27718, R28240, R28251, R29l48, R29827, R3056l (2), R30856, R30922 (2), R3lll9, R31273 (3), R32114, R32475, R32720, R32885 (2), R33049, R33259, R33306, R33766, R35897, R35943, R3599l, R36618, R36882, R37296, R38000, R39671, R40889, R41625 (2), R43ll6, R43420 (2), R43707 (3), R43748, R44056 (2), R446l7, R44684, R44766, R44804, R45535, R45654, R45761, R46043, R46l75, R46205, R46326, R4704l, R48093, R48191, R49l32, R49349, R49557, R50143, R50343, R50706, R50918. R51060, R51205, R51381, R51434, R51460, R51563, R517l7, R51903, R519l9-F (16), R51921-F (9), R51923-D (24), R51926 (2), R51928 (2), R51932-F (26), R51933, R51934 (2), R51938-C (l8), R51943-B (9), R51949 (8), R51953 (27), R51965-C (184), 571 precaudal vertebrae. These remains are either too fragmentary to assign to species or overlap several species in one or more characters. Thamnophis cf. T. couchi Referred material: LACM-VP R14154, R15038, R15419—F, R15975, R16046, R16584-B, R17242, R18859, R20065, R23075, R23133, R23505-B, R23848-B, R25869 (2), R30100, R46925, 37 R51932-E, R51934, R51944 (4), R51952 (3), R51965—B (14), 40 precaudal vertebrae. Diagnosis: A large degree of overlap in features makes specific identification of Thamnophis vertebrae difficult. Holman (1962) used characters of the neural spine to identify species groups within the genus. Thamnophis couchi and T. sirtalis from the Los Angeles area today have neural spines that are somewhat longer than high and usually lack strong anterior overhangs. Anterior overhangs are stronger in the western species T. elegans, T. ggggg, and T. marcianus. T. cyrtopsis has a much lower, longer neural spine. The diverse subspecies of T. couchi display a wider range of vertebral morphology than do those of T. sirtalis, surpassing even most interspecific comparisons. Most of the Pit 91 fossils fall within the range of overlap between the two species and are assigned to ThamnOphis sp. (above). To distinguish between T. couchi and T. sirtalis, ratios of centrum length (CL) and neural arch width (NAW) were compared (see Auffenberg, 1963, pp. 153-155; and Meylan, 1982, pp. 5-6). Five mid-body vertebrae from nine individuals of each species had CL/NAW ranges of 1.15-1.70 (i=1.39) for g. couchi, and 1.55-2.06 (i=1.82) for T. sirtalis. Vertebrae assigned to Thamnophis of. T. couchi are mid-body vertebrae with CL/NAW Ratios of less than 1.50. Those assigned to T. sirtalis have CL/NAW ratios of greater 38 than 1.70. Although the fossils assigned to Thamnophis cf. T. couchi fall comfortably within the wide range of variability displayed by the few modern populations sampled (four, including Arizona, Baja California, and two California localities), they are not identical to any of these. Comments: This species favors aquatic habitats, including marshes and streams. It is found in the Los Angeles area today (Stebbins, 1954, 1966). Thamnophis cf. T. couchi has not previously been reported from the fossil record. A representative fossil vertebra is illustrated in Figure 7. Thamnophis sirtalis (Linnaeus) Referred material: LACM-VP R16584, R20485, R51943, three precaudal vertebrae. Diagnosis: The basis for the identification of this material is described above. It is indistinguishable from comparative material observed. Comments: This widespread species is usually found near aquatic or moist environments, ranging into fields and woodlands. Dixon (1967) reports only two records of this species in Los Angeles County today. This is the southern- most extent of its present range in California (Stebbins, 1966). T. sirtalis has not previously been reported from the Pleistocene of the Southwest. 39 Figure 7. Vertebra of Thamnophis cf. T. couchi (LACM-VP R25869) (A) dorsal View, (B) ventral view, (C) anterior View, (D) posterior view, (B) lateral View; from Pit 91, Rancho La Brea (10x). 40 Figure 7. 41 Family Viperidae Subfamily Crotalinae Crotalus Linnaeus Vertebrae: Crotalus vertebrae may be identified by characters given by Holman (1965), Auffenberg (1963), and Brattstrom (1964b). Short, wide centra with long, thick hypapophyses throughout the vertebral column distinguish Crotalus from colubrid species. Agkistrodon has larger fossae lateral to the cotyle than Crotalus. These fossae have a single, large foramen in Agkistrodon, while in Crotalus the foramina are smaller and may be multiple. Sistrurus has longer centra and may have a small spine on the zygosphene anterior to the neural spine. Crotalus viridis (Rafinesque) Referred material: LACM-VP R14729, Rl4951, Rl4970, Rl4974, Rl4975, R15014, R15996, Rl6033, Rl7798, Rl7997, Rl8049, R20167, R22565 (2), R22603, R22776, R27648, R29466, R29963, R29964, R30992, R31086, R34702, R37476, R41014, R42121, R43061, R47848, R48486, R49556, R51906, R51921-E, R51942, R51955, R51963, R51967, R51968, 36 precaudal vertebrae. Diagnosis: These are assigned to g. viridis on the basis of their centra which are longer than most other 42 Crotalus, and their low neural spines. 9. atrox, Q. molossus, and Q. ruber have shorter centra and higher neural Spines; Q. mitchelli is shorter and has lower neural spines; Q. scutulatus has higher neural spines; while 9. cerastes has a wide, dorsally convex zygosphene. Comments: This species is known from a very wide variety of habitats including coastal sand dunes, wooded regions, near water courses, and prairies, but is not known from desert regions. 9. viridis occurs in the area today (Stebbins, 1966). Q. viridis was reported from Rancho La Brea by Brattstrom (1953b, 1958a). Other Rancholabrean records of this species from California include the McKittrick Asphalt and Hawver Cave (Brattstrom, 1953a), Mescal Cave (Brattstrom, 1958a), and Costeau Pit and Newport Beach Mesa (Hudson and Brattstrom, 1977). It is also reported from the Early Holocene of Nevada at Gypsum Cave (Brattstrom, 1954). DISCUSSION Eight modern snake taxa are added to the herpetofauna of Rancho La Brea including two that are new to the fossil record. The composition of the snake fauna of Pit 91 suggests the presence of a proximal aquatic and a distal semiarid habitat. Differential survivorship between large endothermic species and small ectothermic species at this 43 site may be due to the lower resource requirements of the latter. These tOpics are elaborated below. The present study increases the number of snake species known from Rancho La Brea from four to twelve. Species new to this site include Diadophis punctatus, Hypsiglena torguata, Tantilla sp., Arizona elegans, Masticophis lateralis, Rhinocheilus lecontei, Thamnophis cf. T. couchi, and T. sirtalis. Of these, Thamnophis cf. T. couchi and M. lateralis are new to the fossil record. The habitats of the area surrounding the site of deposition of Pit 91 during the Late Pleistocene as indicated by the snake fauna was probably semiarid Chaparral, with patches of Open grassland, and with a stream and its associated riparian woodland. These conclusions are based on the relative abundance of the species at the fossil site (Table l) and a comparison of present habitat preferences of the species identified (Table 2). The relative abundance of species identified is indicated in Table l by numerically comparing vertebrae of adults, juvenile vertebrae, and skull elements. The number of elements of each species compared to the others is assumed to be representative of the importance of that species in the local ecology at the time of deposition. Abundance is indicated by counting the total identifiable elements because of the difficulties inherent in determining the "minimum number of individuals" in snake faunas (Meylan, 1982). Abundance of elements of fossil animals is considered 44 mmo. moo. oav. mom. Hence no A Hm ha II 0 II 0 II o II o II o N mmo. H II mHH. m II o m mmo. H mm «mm. m Hm cow. m mounwuum> Hmuoe mucmEmHm mflacm>ss Co A Hasxm mo HOQEOZ mo Hwnfisz Nmm AdBOB moo. m mwamuouma manQOOHummz moo. m floucooma msaflmnoocflnm woo. v mammwam chNfln< moo. m .mm mmmmmmmw mao. ma mumsvuou mamamwmmmm mmo. om msasuom mHuHOQOHmqu Hmo. hm GOEHOE HOQDHOO moo. om mfloflufl> msamuouu woo. ow msumuocsm mfinmoomao moH. om moosoaocmama mflnmosuflm was. was Isaac marmonsmse Hobos ownnmuwo> m0 w pasc4 mo wonssz .Hoflumume OHHcm>sn Omcflanmpmoss mumOHUsfl moonmmm .mocmpcsnm mo “mono mcflmmwuowp aw mmflommm oxmcm comm on osocmm mo mucmEmHO mo mumnfisz .H manna 45 H m H H H I mHHmHmumH .m m m m I H I mwmmmmwm .m m m m H H I mmmmmwm .m m m H I q I .ma mmwmmmmm m m H I H I .mmmmmmmm .m A H m H H I mmmmmm .w H m m H H I mmmmmmw .m H m m H H I mmmmmmm .m H m m H H I msosmHosmHmE .M I H H H m I msumuocsm .m I H H H m H mmmmmmmm .M I I I I H m 3.9. OHHox Hmuummmco ccmHmmoHO ummuom OCOHUOOS OHDMOOH coHHMOHm Hmum3£mmuh Owuommxm DOOI .mocmpcsnm umm3oHIIH .mumeOEHODCHIIH .mocmocsnm umwzmHsII: .umanm: umsu CH mmHommm CHOUOE mo mocmpcsnm m>HumHmH may no comma .mmxmcm Hm uHm may mo mwocwummwum umanmm .m mHnt 46 to be the result of two major factors: (1) the species' density in the depositional environment, and (2) the proxi- mity of the once living population to the depositional environment (Shotwell, 1955). A comparison of the abundance of the remains of each taxon with the preferred habitats of those taxa (Table 2) indicates the habitats present and their proximity to the site of deposition during the Pleistocene. Moreover, because of their lightness and greater probability Of transport, skull and juvenile elements are thought to repre- sent a population near the site of deposition. Thus, Thamnophis, which makes up more than 71 percent of the identifiable fauna, indicates that there was probably an aquatic habitat near the site. The presence of Thamngphis cf. T. couchi suggests that this was a permanent body of water. T. sirtalis suggests marshy lowland areas and meadows. Diadophis punctatus indicates relatively moist fields and woodlands. Since there is geological evidence of stream drift at the site (Cox and Jefferson, TE.TT£E., 1983), it may be inferred that this site was located at or near a stream in riparian woodland. The four other abundant species Pituophis melanoleucas, Crotalus viridis, Coluber mormon, and LamprOQeltis getulus, may be found in a wide variety of habitats, but are most characteristic of Chaparral and grassland associations. The remaining species, accounting for only about three percent of the identifiable snake fauna, are characteristic of semiarid or arid habitat 47 ranging from deserttflnxnmflichaparral and prairies. Their low abundance and worn appearance suggest allochthonous origins. The habitat in the region of Pit 91, then, appears to have consisted of two components: a moist riparian site near the site of deposition based on the abundance of ThamnOphis remains and the presence of stream drift in the strata; and a distal semiarid Chaparral and grassland mixture. Warter (1976) gives a more detailed account of the flora of Pit 91. She did not identify any grassland species, but her habitat interpretations tend to support those presented here. It is noteworthy that the semiaquatic species had the highest numbers Of fossils, and also show the lowest diversity, including but two species, Thamnophis cf. T. couchi and T. sirtalis. Conversely, the semiarid species are more diverse, including all of the remaining species except Diadophis punctatus. This suggests that the semiarid habitat is more extensive and probably most characteristic of the region and that the proximal moist riparian woodland habitat is somewhat isolated, supporting large populations of relatively few species. This interpretation mainly reflects the modern conditions in the Los Angeles region, but the presence of a permanent water source indicates slightly moister conditions as suggested by Brattstrom (1953b). Very few extinctions appear to have occurred among North American herpetile species during the Rancholabrean 48 (Holman, pers. comm. 1982), whereas large bird (Welty, 1975) and mammal (Kurten and Anderson, 1981) taxa suffered large numbers of extinctions. Rancho La Brea is no excep- tion to this pattern. All herpetological species reported from the site by Brattstrom (1953b, 1958) and in this report represent extant taxa. This contrasts strongly with the avian and mammalian faunas. Howard (1962) indicates that about 14 percent of the bird species known from Rancho La Brea are extinct. Based on Stock (1956) and Akersten et. al. (1979), roughly 40 percent of the mammalian species are extinct. The stability of the North American herpetofauna during the Late Pleistocene may be the logical result of the ability of small ectotherms to maintain viable pOpulations in relatively small refugia. Guilday (1967) argues con- vincingly that the species of a megafauna will be the logical first victims of a rapidly changing climate in which critical habitat is rapidly reduced to small refugia because of the extremely large quantities of resources necessary to main- tain a viable population. If this is the case, small animals are not as likely to be extirpated by diminished habitats, provided small refugia are available, and small terrestrial ectotherms will be the last species to expire because of their low resource requirements. Johnson (1977) suggests that coastal Southern California represented such a refugium during the Late Pleistocene. Thus, that the herpetofauna of the Rancho La Brea deposits is essentially equivalent to the 49 modern one, in contrast to the mammalian and avian faunas, is probably the result of the ability of herpetiles to weather unfavorable conditions as small populations in refugia. S UMMARY This study adds eight modern species of snake to the herpetofauna of Rancho La Brea and introduces Thamngphis cf. T. couchi and Masticophis lateralis to the fossil record. Modern habitat preferences of species identified suggest that Pit 91 represents a riparian woodland habitat within a more extensive Chaparral and grassland habitat. Finally, it is suggested that populations of small ectotherms survive in isolated refugia more readily than large endotherms because of their lower resource requirements. APPENDIX APPENDIX Comparative material examined. Charina bottae (2) Lichanura trivirgata (1) Contia tennis (2) Diadophis punctatus (13) 9. p. amabilis (l) p. gggyi (l) p. edwardsi (4) 2. p. modestus (1) E . punctatus (6) Hypsiglena torquata (3) Tantilla coronata (l) T. gracilis (8) T. nigriceps (2) T. planiceps (1) Trimorphodon lambda (2) T. vandenberghi (2) Phyllorhynchus browni (l) T. decurtatus (4) Arizona elegans (3) Coluber constrictor (20) Q. g. flaviventris (8) 50 Q. g. foxi (3) Q. g. priapus (9) Q. mormon (l) Elaphe guttata (8) T. g. emoryi (3) E. g. guttata (5) T. obsoleta (6) T. g. obsoleta (1) T. g. quadrivittata (5) T. subocularis (l) T. vulpina (7) Masticgphis bilineatus (l) M. flagellum (11) M. T. flagellum (5) . f. lineatulus (1) If?- . T. testaceous (5) l3 M. lateralis (4) M. mentovarius (l) M. taeniatus (3) l3 . t. ornatus (2) g. t. schotti (l) 51 Pituophis melanoleucus (25) B. M. mugitus (5) P. M. sayi (20) Salvadora grahamiae (l) Salvadora hexalepis (l) LamprOpeltis calligaster (5) T. getulus (12) T. g. getulus (5) T. g. holbrooki (3) T. g. niger (2) T. g. splendida (2) T. pyromelana (2) L. triangulum (10) T. zonata (1) Rhinocheilus lecontei (8) Chionactis occipitalis (8) Ficimia olivacea (l) Sonora EBTSCOpa (5) S. semiannulata (l) Nerodia cyclopion (10) M. erythrogaster (7) N. fasciata (l9) . rhombifera (5) l2 M. sipedon (18) Regina grahami (4) Storeria dekayT (l7) §. occipitomaculata (7) Thamngphis couchi (9) T. gyrtopsis (l) I'd . elegans (l7) egues (2) macrostemma (6) I8 Ila marcianus (6) I'd proximus (7) radix (13) ll-3 l6 II-3 rufipunctatus (l) Ira . sirtalis (24) Trgpidoclonion lineatum (l) Virginia striatula (24) T. valeriae (2) Micruroides euryxanthus (2) Micrurus fulvius (9) Agkistrodon bilineatus (1) A. contortrix (l) A. piscivorus (14) Crotalus adamanteus (7) T. atrox (10) Q. cerastes (6) IO horridus (9) l0 lepidus (1) mitchelli (4) l0 IO molossus (2) IO pricei (l) ruber (2) IO <_:_. IO scutulatus (4) triseriatus (l) viridis (14) 52 LIST OF REFERENCES LIST OF REFERENCES Akersten, W. A., R. L. Reynolds, A. E. Tejada-Flores. 1979. New Mammalian records from the Late Pleistocene of Rancho La Brea. Bull. So. California Acad. Sci. 78:141-143. Auffenberg, W. 1963. The Fossil Snakes of Florida. Tulane Stud. Zool. 10:131-216. Berger, R., and W. F. Libby. 1966. UCLA Radiocarbon Dates V. Radiocarbon 8:491-493. . 1968. UCLA Radiocarbon Dates VIII. Radiocarbon 10:402-416. Brattstrom, B. H. 1953a. Records of Pleistocene Reptiles from California. Copeia:l74-l79. . 1953b. The Amphibians and Reptiles from Rancho La Brea. Trans. San Diego Soc. Nat. Hist. 11:365-392. . 1954. Amphibians and Reptiles from Gypsum Cave, Nevada. Bull. So. California Acad. Sci. 53:8-12. . 1955a. Pliocene and Pleistocene Amphibians and Reptiles from Southeastern Arizona. Jour. Paleontol. 29:150-154. . 1955b. Small Herpetofauna from the Pleistocene of Carpinteria, California. Copeia:l38-139. . 1958a. New Records of Cenozoic Amphibians and Reptiles from California. Bull. So. California Acad. Sci. 57:5-12. . 1958b. Additions to the Pleistocene Herpetofauna of Nevada. Herpetologica 14:36. . 1964a. Amphibians and Reptiles from Cave Deposits in Southcentral New Mexico. Bull. SO. California Acad. Sci. 63:93-103. . 1964b. Evolution of the Pit Vipers. Trans. San Diego Soc. Nat. Hist. 13:185-268. _ 53 54 Brattstrom, B. H. 1967. A succession of Pliocene and Pleistocene snake faunas from the high plains of the United States. Copeia:l88-202. . 1976. A Pleistocene herpetofauna from Smith Creek Cave, Nevada. Bull. So. California Acad. Sci. 75: 283-284. Conant, R. 1975. A field guide to the reptiles and amphibians of the eastern United States. Second Edition. Houghton Mifflin, Boston. 429 pages, 48 plates. Cundall, D. 1981. Cranial osteology of the colubrid snake genus Opheodrys. Copeia:353-37l. Dixon, J. R. 1967. Amphibians and reptiles of Los Angeles County, California. Los Angeles Co. Museum of Nat. Hist. Sci. Ser. 23, Zoology No. 10:1-64. Dowling, H. G., and W. E. Duellman. 1978. Systematic Herpetology: A Synopsis Of Families and Higher Categories. HISS Publ. in Herpetology No. 7. HISS Publications, New York. Fitch, H. S., W. S. Brown, and W. S. Parker. 1981. Coluber mormon, a Species Distinct from Q. constrictor. Trans. Kansas Acad. Sci. 84:196-203. Guilday, J. E. 1967. Differential Extinction during Late- Pleistocene and Recent Times. Pleistocene Extinction, Martin and Wright Eds:121-124. Hecht, M. K. 1982. The Vertebral Morphology of the Cretaceous Snake Dinilysia patagonica Woodward. N. Jb. Geol. Palaont. Mh. H.9:523-532. Hill, W. H. 1971. Pleistocene Snakes from a Cave in Kendall County, Texas. Texas Jour. Sci. 22:209-216. Ho, T. Y., L. F. Marcus, and R. Berger. 1969. Radiocarbon Dating of Petroleum-impregnated Bone from Tar Pits at Rancho La Brea, California. Science 164:1051-1052. Holman, J. A. 1962. A Texas Pleistocene Herpetofauna. Copeia:255-261. . 1963. Late Pleistocene Amphibians and Reptiles of the Clear Creek and Ben Franklin Local Faunas of Texas. Jour. Grad. Res. Center, So. Methodist Univ. 31:152-167. 55 Holman, J. A. 1965. A Late Pleistocene Herpetofauna from Missouri. Trans. Illinois Acad. Sci. 58:190-194. . 1967. A Pleistocene Herpetofauna from Ladds, Georgia. Bul. Georgia Acad. Sci. XXV:154-l66. . 1970. A Pleistocene Herpetofauna from Eddy County, New Mexico. Texas Jour. Sci. 22:29-39. 1976. Snakes of the Split Rock Formation (Middle Miocene), Central Wyoming. Herpetologica 32:419-426. . 1977a. Upper Miocene Snakes (Reptilia, Serpentes) from Southeastern Nebraska. Jour. Herpetology 11:323- 335. . 1977b. The Pleistocene (Kansan) Herpetofauna of Cumberland Cave, Maryland. Annals Carnegie Mus. 46:157-172. . 1981. A Review of North American Pleistocene Snakes. Publ. Mus. Michigan State Univ. Paleontol. Ser. 1:263- 306. Howard, H. 1960. Significance of Carbon-l4 Dates for Rancho La Brea. Science, 131:712-714. . 1962. A Comparison of Avian Assemblages from Individual Pits at Rancho La Brea, California. Los Angeles Co. Mus. Contrib. Sci. 58:1-24. Hudson, D. M., and B. H. Brattstrom. 1977. A Small Herpetofauna from the Late Pleistocene Of Newport Beach Mesa, Orange County, California. Bul. So. California Acad. Sci. 76:16-20. Johnson, D. L. 1977. The Late Quaternary Climate of Coastal California: Evidence for an Ice Age Refugium. Quaternary Research 8:154-179. Kurten, B., and B. Anderson. 1980. Pleistocene Mammals of North America. Columbia University Press, New York. 442 pages. McMenamin, M. A. S., D. J. Blunt, K. A. Kvenvolden, S. E. Miller, L. F. Marcus, and R. R. Pardi. 1982. Amino Acid Geochemistry of Fossil Bones from the Rancho La Brea Asphalt Deposit, California. Quaternary Research 18:174-183. 56 Meylan, P. A. 1982. The Squamate Reptiles of the Inglis IA Fauna (Irvingtonian: Citris County, Florida). Bull. Florida State Mus. Biol. Sci. 27:1-85. Savage, D. E. 1951. Late Cenozoic vertebrates of the San Francisco Bay Region. Univ. California Publ. Bull. Dept. Geol. Sci. 28:215-314. Shaw, C. A. 1982. Techniques Used in Excavation, Prepara- tion, and Curation of Fossils from Rancho La Brea. Curator 25:63-77. Shotwell, J. A. 1955. An Approach to the Paleoecology of Mammals. Ecology: 36:327-337. Stebbins, R. C. 1954. Amphibians and Reptiles of Western North America, McGraw-Hill, New York. 536 pages. . 1966. A Field Guide to Western Reptiles and Amphibians. Houghton Mifflin Co. Boston. 279 pages. Stock, C. 1956. Rancho La Brea: A Record of Pleistocene Life in California. Nat. Hist. Mus. Los Angeles County, Sci. Ser. 20. 81 pages. Van Devender, T. R., and J. I. Mead. 1978. Early Holocene and Late Pleistocene Amphibians and Reptiles in Sonoran DeSert Packrat Middens. Copeia: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 Nat. 22: 49-66. Van Devender, R. R., and R. D. Worthington. 1977. The Herpetofauna of Howell's Ridge Cave and the Paleoecology of the Northwestern Chihuahuan Desert. In: Trans. Symp. on the Biological Resources of the Chihuahuan Desert Region, U.S. and Mexico. R. H. Wauer and D. H. Riskind (eds.). Nat. Park Ser., Wash. D.C. Warter, J. K. 1976. Late Pleistocene Plant Communities-- Evidence from the Rancho La Brea Tar Pits. pp. 32-39 in Latting, J., ed. Plant communities of Southern California. California Native Plant Society Special Publication No. 2, Berkeley. Welty, J. C. 1975. The Life of Birds. 2nd Edition. W. B. Saunders Company, Philadelphia, London, Toronto. 623 pages. 57 Woodard, G. D., Iand 1L. F. Marcus. 1973. Rancho La Brea Fossil Deposits: A Re-evaluation from Stratigraphic and Geological Evidence. Jour. Paleont. 47:54-69. Wright, A. H., and A. A. Wright. 1957. Handbook of Snakes of the United States and Canada. Comstock Publishing Associates, Ithaca, New York. 2 Volumes.