PaleoBios, Volume 15, Number 3, Pages 47-62, May 24, 1993 Harlan's Ground Sloth (Glossotherium harlani) and a Columbian Mammoth (Mammuthus columbi) from Stevenson Bridge, Yolo County, California Robert G. Dundas1 and Laura M. Cunningham2 1 Museum of Paleontology and Department of Integrative Biology, University of California, Berkeley, CA 94720; 2225 Columbia Avenue, Kensington, CA 94708 ABSTRACT Skeletal remains of a subadult and adult Glossotherium harlani and an adult Mammuthus columbi were recovered from fluvial sediments of Putab Creek near Davis, California in July 1975. U/Th radiometric age determination of two mammoth bones suggests a possible late Sangamonian interglacial age (180 isotope stage 5a) for the site. The mammoth material as well as an adult ground sloth femur show evidence of substantial green bone fracturing and bone surface scratches indicative of trampling prior to final deposition. The subadult ground sloth partial skeleton lacks the extensive fracturing and other bone modification noted for the mammoth and adult ground sloth. Additionally close associations of the skull and first three cervical vertebrae, a partial rib cage, and left scapula and left humerus epiphysis suggest that the subadult ground sloth was deposited with some soft connective tissue remaining on the skeleton. These and other taphonomic inferences were made possible by utilization of appropriate data collection methods at the time of excavation. This emphasizes the significance of how good techniques in the collection of field data and laboratory preparation may permit important taphonomic inferences long after a site has been excavated by other researchers. INTRODUCTION Fossil remains of Mammuthus and Glossotherium are common in the Pleistocene of California, particularly in the San Francisco Bay region (Jefferson, 1991). The Stevenson Bridge locality, possibly late Sangamonian (late lsO isotope stage 5a) in age, is one of several late Pleistocene sites along Putah Creek that have yielded ground sloth and mammoth remains. Few specimens are as well preserved as those recovered from the Stevenson Bridge locality; partial skeletons of a subadult and adult Glossotherium harlani and an adult Mammuthus columbi are represented. The Stevenson Bridge site, located in Yolo County approximately 9 km west of Davis, California, is situated on private land of Mrs. J. Jacob about 100-125 m upstream from Stevenson Bridge along the north bank of Putah Creek at an elevation of 16 m (Figure 1): 38°32'13"N, 121*51 WW, Merritt, California USGS 7.5' series topographic quadrangle. HISTORY OF INVESTIGATIONS In mid-July, 1975 three Dixon, California residents, Mark and Mike Little, and Dennis Danielson discovered a mammoth tusk fragment on the north bank of Putah Creek. Subsequently in New Additions to the Pleistocene Vertebrate Record of California (R. G. Dundas and D. J. Long, eds,), PaleoBios v.15. \ * * \ ----------Road * UCMPV76199 >--------¦ 0.5 Km V.CM 7 I I / / / ¦L.-.J :-$«%* I ¦ ¦ / h Figure 1. Generalized map noting the location of the Stevenson Bridge locality (V76199) west of Davis, California. the University of California Museum of Paleontology (UCMP) and Leonard Williams, museum scientist, Department of Anthropology at the University of California (UC) at Davis were contacted about the find. Page 48 Stevenson Bridge Mammoth and Ground Sloth Dundas and Cunningham Following examination of the site Leonard Williams, then graduate students James West and Dwight Simons, and a small field crew from UC Davis began an approximately three week excavation of the locality on July 24,1975. Photos of the excavation are on file in the supplementary locality records of the UCMP. After excavation specimens were taken to UC Davis for preparation and exhibition. The specimens were later transferred to the UCMP where they were accessioned. The UCMP inadvertently catalogued the site under the name Steven's Creek Bridge, UCMP locality V76199, see Jefferson (1991, p.103). The name has been corrected to read Stevenson Bridge. GEOLOGY AND TAPHONOMY The Stevenson Bridge locality and Putah Creek are situated in a 10 m deep by 30 m wide gulley incised by the stream (Figure 2). The gulley bottom sediments are fluvial deposits of Putah Creek. Geologic outcrops are sparse because dense vegetation covers the gulley bottom. Reconnaissance of the creek bank exposures indicates that the typical stream deposits range from clay to pebble size clasts, with sediments larger than gravel a rarity. Presently, Putah Creek, a meandering stream, deposits mostly silt, however this may partially reflect a disruption of natural sedimentation processes because water flow is controlled and partially diverted upstream by Monticello Dam and the Lake Solano Diversion Dam. Detailed sedimentologic and stratigraphic data are unavailable for this site. Fossils were recovered from a 0.5-0.75 m thick greenish-grey silty clay layer capped by a 0.5-1.0 m thick sequence of brownish sediments of silt, sand and !'¦ Unit of brownish silt, •V sand and gravel, 0.5-1.0 m thick. Fossil bearing greenish- grey "silty clay" layer. Dominated by clay at the base but grades into silty clay with sand and pebbles. 0.5-0.75 m thick. Partially indurated gravel. Indurated sandstone. V76199 Putah Creek Figure 2. Section through the Putah Creek stream gulley, showing Putah Creek, Stevenson Bridge and the Stevenson Bridge locality (V76199). Figure 3. Generalized stratigraphic section at the Stevenson Bridge locality (V76199). gravel. Underlying the silty clay layer is a partially indurated gravel underlain by an indurated sandstone (Figure 3). Examination of the residual matrix adhering to several of the mammoth bones revealed that sand and pebbles may be common within this "silty clay" layer; the matrix varies from grey to light brown in color. Although the sediments are fluvial channel deposits of Putah Creek, the lack of adequate sedimentologic, lithologic and stratigraphic data precludes an accurate interpretation of the depositional environment. Further inferences would be mere speculation because both channel fill and point bar deposits share some similarities with the general lithologies represented in the above stratigraphic sequence (Allen, 1965; Reineck and Singh, 1980; Behrensmeyer, 1988). Misinterpretation of channel deposits is possible when lacking a three dimensional section with adequate lateral and vertical exposures (Berhensmeyer, 1988). Paleocurrent direction was assessed for the fossil bearing stratum by examination of long bone orientations. Although dip angles of skeletal elements were not recorded it appears from the excavation photographs that the bones were lying in or near horizontal position within the same stratigraphic horizon. Therefore, long axes of bones were measured from the excavation map and plotted on a 20° sector rose diagram (Figure 4). A preferred orientation is indicated Dundas and Cunningham Stevenson Bridge Mammoth and Ground Sloth Page 49 Figure 4. Sector rose diagram showing long axes orientations of Stevenson Bridge locality skeletal elements reflecting paleocurrent direction. Sample size plotted is 17 elements in 20 degree sectors over 180 degrees; 13 of the elements are ground sloth ribs. particularly with respect to the ground sloth ribs. Long bones may orient parallel or perpendicular to water current direction. Orientation of the long axis parallel to current is characteristic of bones that move by sliding, where one end of the bone is significantly heavier than the other, while bones that roll during transport tend to orient perpendicular to current. Orientation of the long axis parallel to current is typical of most long bones (Voorhies, 1969). In fluvial transport experiments with modern elephant bones, ribs oriented parallel to current with the heavy end pointing upstream (Todd and Frison, 1986), provided that the ribs were unobstructed by other bones, particularly large elements like the pelvis which act as "bone traps" for smaller skeletal elements. It is reasonable to infer that Glossotherium ribs also orient parallel to current which would indicate that water flowed roughly along a west-east trend when the ribs were aligned (Figure 4). Considering the confinement of the stream to the gulley through which it presently flows, it is likely that Putah Creek flowed in the same direction during the late Pleistocene; i.e. west to east. Additionally, the position of the mammoth mandible implies a west to east current direction because it was found upside down with the symphysis pointing to the west (Figure 5). This is a stable position of elephant mandibles observed in fluvial transport experiments with the symphysis pointing upstream; subsequent movement from this position requires large water volume and velocity (Todd and Frison, 1986). Taphonomic inferences based on incomplete data are difficult and considerable caution must be exercised particularly when analyzing data collected by other researchers. Yet, sufficient data recorded during excavation of this locality permits some interpretations of the taphonomic history of the fossil assemblage. The bones composing the partial skeletons were found in association but not articulated (Figure 5). Detailed examination of bone surface features revealed the following. The mammoth material is discussed first. The mammoth specimens exhibit little or no abrasion although some show minor surface weathering in the form of shallow surface cracks, probably indicating brief subaerial exposure prior to deposition; the most advanced weathering is comparable to Stage 1 of Behrensmeyer (1978), although most elements are Stage 0. There are no surface marks that can be confidently attributed to carnivore activity. Nearly all of the mammoth specimens are broken and are more widely dispersed than the ground sloth material. Most fractures are spiral in nature with sharp edges; most breakage occurred prior to final deposition as indicated by matrix filling of fractures and the absence of the broken fragments in the fossil assemblage. The mammoth bone breakage cannot be explained as resulting from fluvial processes. Currently Putah Creek, and likely its antecedants, lacks the energy necessary, even at flood stage, to break the mammoth bones. Moreover, experimental evidence indicates that bone is rarely broken by fluvial transport, rather most breakage occurs prior to transport and deposition (Behrensmeyer, 1991). Although large carnivores are capable of fracturing bones of large ungulates and leaving little evidence behind (Haynes, 1980) it is unlikely that the extensive breakage of the mammoth bones is attributable to large carnivore activity particularly since there are no characteristic carnivore gnawing or bite marks present on any of the elements. Furthermore, considering the robustness of some mammoth skeletal elements it is doubtful that even the largest North American Pleistocene predators could cause the fragmentation exhibited by this partial skeleton (Haynes, 1983). This leaves the possibility of trampling as the main agent of bone fracturing. Trampling is a significant cause of large mammal bone breakage (Haynes, 1983). Although weathered bone is relatively brittle and thus more easily fractured when stepped on by large animals, green bone is difficult to break Page 50 Stevenson Bridge Mammoth and Ground Sloth Dundas and Cunningham by trampling (Haynes, 1983). The breakage of large, robust, relatively unweathered mammoth bones is probably the result of trampling by a large herbivore, possibly another mammoth; living African elephants trample and fracture skeletal elements of dead elephants (Haynes, 1991). In addition to bone fracturing other evidence of bone surface modification indicates trampling. Although excavation and preparation techniques caused some of the surface marks evident on the mammoth skeletal elements, other surface striations resemble trample marks made by large herbivores. The marks are identical to shallow sub-parallel and individual scratch marks observed on subaerially trampled bone (Fiorillo 1984, 1989, 1991; Behrensmeyer et al., 1986). Although similar striations may be caused by human butchering techniques (Fiorillo 1987, 1989, 1991; Behrensmeyer et al., 1986) this possibility can be dismissed because this site significantly predates (see age below) the generally accepted first appearance of humans in North America. Trampling striations are caused by scraping the bone surface with a hard object such as sand grains on the bottom of the foot or pressing the bone against a hard substrate when stepping on it; hooves alone, which are softer than bone, are not sufficient to produce scratch marks (Fiorillo 1987, 1989, 1991). Kicking of elephant bones by other elephants, without trampling them, also produces similar striations (Haynes, 1991). The number of scratch marks on bone depends on the extent of trampling and type of substrate; the coarser (sandier) the substrate, the greater potential frequency of scratch marks (Fiorillo 1987, 1991). With the exception of the mandible all of the other mammoth bones exhibit trample marks. The presence of scratch marks are also related to the amount of bone weathering. Scratch marks are more easily preserved on unweathered bone probably because weathered bone surfaces are more likely to fracture and crumble when stepped upon (Fiorillo 1989,1991). Since the mammoth bones are relatively unweathered, trampling marks are well preserved. It is impossible to infer what animal caused the striations. Scratch marks produced by large herbivores, such as cattle, have been documented but it is unknown how small an animal may be and still produce trample marks (Fiorillo, 1991). Although other physical processes (e.g. fluvial transport) may cause surface marks (Olsen and Shipman, 1988) the lack of both abrasion and abundant coarse clasts in the sediments suggests that the surface scratches are not produced by fluvial processes. Moreover, fluvial processes would more seemingly produce individual marks rather than the multiple sets of parallel striations observed on these specimens. There is also a distinct difference in the amount of trampling marks on the various mammoth bones; the ilium, for example, exhibits numerous scratches, but mostly on one side. In addition to the striations, the scapula displays a 4.5 cm by 6.5 cm depression fracture on the infraspinous fossa which was probably caused by a large animal stepping on it. In summary, breakage of the robust mammoth skeletal elements coupled with the surface striations indicates trampling prior to burial. In considering all skeletal material preserved in the fossil assemblage there is an evident bias towards large skeletal elements which suggests the possibility of dispersion by carnivore activity, trampling or hydraulic sorting. The lack of evidence of large carnivore activity suggests that carnivores were an unlikely agent for moving the mammoth skeletal elements, although large carnivores are known to move large elephant bones (Haynes, 1988) and leave little evidence of activity (Haynes, 1983). Trampling is a viable agent for the observed dispersion because trampling often moves skeletal elements (Haynes, 1991). Modern African elephants will scatter skeletal parts of dead elephants (Saunders, 1977; Haynes, 1988), as will other herbivores (Haynes, 1991). With respect to the potential of fluvial transport, the large mammoth bones were not moved far because their large size equates with a low transport potential (Todd and Frison, 1986); limited fluvial transport of these specimens probably occurred during a flood event. If smaller skeletal elements were present, they may have been transported from the area. Analysis of skeletal sorting behavior indicates that lighter elements transport first (Voorhies, 1969; Behrensmeyer, 1990). However, transport potential of smaller mammoth elements such as foot bones and vertebrae are not well studied although transport distance is high relative to the large bones found in this assemblage (Todd and Frison, 1986). Many of the mammoth bones present in the Stevenson Bridge assemblage are bones with the lowest transport potentials of the mammoth skeleton (Todd and Frison, 1986). A similar argument for limited fluvial transport is made for the subadult I to 3 a ft c 3 5. 5" TO 3" H 3 I I 2 B 3 o to 3 a O -i e c 3 a M Figure 5. Excavation map of the Stevenson Bridge locality (V76199) showing the positions of the skeletal elements. Glossotherium harlani: Gl - cranium; G2 - mandible; G3 - atlas; G4 - axis; G5 - 3rd cervical vertebra; G6 - ribs; G7 - left scapula; G8 - right femur; G9 - left innominate; GIO - right femur. Mammuthus columbi: Ml - cranial fragments; M2 - tusk fragments; M3 - right M^; M4 - left M^; M5 - mandible; M6 - left scapula; M7 - left ulna; M8 - ilium. Modified from field map by James West. -8 ¦0 TO III Page 52 Stevenson Bridge Mammoth and Ground Sloth Dundas and Cunningham ground sloth. The elements of the subadult ground sloth partial skeleton are closely associated (Figure 5). Other than minor weathering (Stage 1) on one side of some of the ribs, there is no evidence of carnivore activity, surface weathering or abrasion. A few of the ribs also exhibit surface sanations but not as abundant as on the mammoth material. The specimens are nearly complete with most of the breakage attributable to postdepositional alteration or excavation procedures because either the breaks are fresh and lack matrix in the bone fracture surfaces or many broken fragments were found in proximity to the broken elements. As noted above the ribs exhibit a preferential orientation. Although disarticulated, the spatial proximity of various skeletal elements (partial rib cage, skull and first three cervical vertebrae, left scapula and left humerus epiphysis) (Figure 5) is indicative of soft-anatomy decomposition followed by water currents strong enough to align the ribs but too weak to move the bones very far (Behrensmeyer, 1991). Saunders (1977) cited the presence of abundant dermal ossicles as an indication of in situ decomposition of skin and underlying soft tissue. The apparent absence of dermal ossicles at this site may be explained in several ways: 1) little or no skin remained by the time of deposition although some connective tissue was present; 2) subsequent water currents strong enough to align the ribs could transport the pebble-sized ossicles from the site; 3) some dermal ossicles may have been present but not recognized during excavation, as is often the case if workers are unfamiliar with dermal ossicles since they can resemble lithic pebbles. Some sediment was screened for micromaterial but neither micromaterial nor dermal ossicles were noted (G. J. West, pers. comm., 1993). The adult ground sloth femur exhibits a spiral breakage pattern reminiscent of trampling although carnivore activity cannot be ruled out. A couple of surface marks could questionably be interpreted as carnivore tooth marks; no other evidence of carnivore activity is noted. The presence of abundant shallow sets of parallel and individual striations identical to those on the mammoth material are inferred as trample marks. Also, damage to the distal part of the diaphysis consists of a 10 cm by 11 cm piece of bone missing from the anterior side above the intercondylar articular surface. This fracture likely resulted from a large animal stepping on the distal diaphysis. The element is very lightly abraded and unweathered. The following scenario is suggested for the accumulation of this fossil assemblage. Following death and decomposition of soft tissues, the mammoth skeleton underwent minor subaerial exposure prior to final deposition. The large skeletal elements suggest limited fluvial transport. Trampling resulting in extensive breakage of the skeletal elements occurred; timing of trampling with respect to deposition is uncertain, although lack of associated large broken fragments implies breakage before final deposition. Breakage of the mammoth bones occurred before deposition of the subadult ground sloth; otherwise the subadult ground sloth would likely exhibit considerable breakage also. Deposition of the subadult ground sloth material occurred at about the same time as the mammoth or shortly thereafter. The ground sloth material was deposited at the site with some soft connective tissue remaining. Subsequent decay of the soft tissue occurred, the ribs may have been subaerially exposed for a short time and lightly trampled then aligned by water current. The adult ground sloth femur was trampled and fractured but the exact timing of deposition in relation to the other partial skeletons is uncertain. Burial of all skeletal material was rapid, preventing further bone modification. AGE The fauna indicates a late Pleistocene age. Although Glossotherium harlani occurs throughout the Pleistocene in North America (Kurten and Anderson, 1980), Mammuthus columbi first appears in the late Pleistocene (Maglio, 1973) and both species became extinct between 12,000-10,000 yr. B. P. (Kurten and Anderson, 1980). In addition to the biostratigraphic age two bone samples were submitted for U/Th age determination to the radiometric analysis lab at the University of California at Davis. A mammoth bone dated in 1980 yielded a Tj234/Th230 age of 7766i + 404 yr. B. p. j^q accuracy of this date is suspect, however. In the age analysis report to J. H. Hutchison of the UCMP, Dennis Garber (UC Davis) states that "The confidence in this age is reduced since the differential between the parent U^38 an£j daughter Tj234 was large, indicating a degree of potential leaching. If leaching did occur, then a Dundas and Cunningham Stevenson Bridge Mammoth and Ground Sloth Page 53 younger age would be indicated." The following year a mammoth rib was submitted for analysis and yielded a U/Th age of 81,181 yr. B. P. (Garber, pers. comm. to D. E. Savage at the UCMP, 1981). If these radiometric dates are accurate a late Sangamonian interglacial age (180 isotope stage 5a) is indicated. SYSTEMATIC PALEONTOLOGY Class Mammalia Linnaeus, 1758 Order Edentata Cuvier, 1798 Family Mylodontidae Ameghino, 1889 Genus Gbssotherium Owen, 1840 Glossotherium harlani (Owen), 1840 Figure 6 Referred specimens: UCMP 116084, partial skeleton including: cranium, mandible, atlas, axis, third cervical vertebra, partial rib cage, left scapula, left humerus epiphysis, right femur, left innominate; UCMP 139027, right femur. Description: The following pertains to UCMP 116084. The skull is mostly complete. The cranial dentition is intact except for missing the left fifth tooth. The left pterygoid is partially broken. The premaxillae, much of the left squamosal and most of the nasals are broken. Both jugals are detached from the cranium; the right jugal is slightly damaged but only the anterior half of the left jugal is present. The mandible is complete with only minor breakage of the left first and third teeth. Only the first three cervical vertebrae are represented. The atlas has sustained minor damage to the lateral borders of each wing. The axis is missing the right half of the neural arch and neural spine. The posterior epiphysis of the centrum and parts of both transverse processes are missing. The third cervical vertebra exhibits minor breakage of the neural spine and the ends of the transverse processes. The body of the centrum is slightly damaged and lacks the posterior epiphysis; the anterior epiphysis is present but unfused. A partial rib cage is represented by thirteen costal ribs which exhibit some breakage of the distal ends. The left scapula is complete with minor damage to the vertebral border epiphysis. The right femur lacks the epiphyses forming the head of the femur and the medial condyle/intercondylar articular surface of the distal end; the epiphyses are unfused. The left innominate exhibits moderate breakage of the pubis. UCMP 139027 represents the distal two- thirds of a right femur with damage consisting of a 10 cm by 11 cm section of bone missing from the anterior of the shaft above the intercondylar articular surface to the anterior portion of the medial condyle, which is broken. The condylar epiphyses are fused. Discussion: Possessing the characteristic lobate teeth of mylodont ground sloths the Stevenson Bridge specimens are morphologically indistinguishable and within the range of variation observed by Stock (1925) for specimens of Glossotherium harlani from Rancho La Brea (Appendix 1). With the exception of UCMP 139027, a right femur, all other material appears to represent one individual (UCMP 116084) based on depositional setting, size and condition of the skeletal elements. The presence of several elements with unfused epiphyses indicates that UCMP 116084 was a subadult. The overall size of UCMP 116084 approaches the mean of Rancho La Brea specimens; the cranium is slightly longer and the postcrania slightly smaller than the Rancho La Brea mean. UCMP 139027, a right femur with fused epiphyses, represents a large adult. The femur is of larger size than the right femur of UCMP 116084. UCMP 139027 is larger than the mean of Stock's (1925) observations (Appendix 1). Order Proboscidea Illiger, 1811 Family Elephantidae Gray, 1821 Genus Mammuthus Burnett, 1830 Mammuthus columbi (Falconer), 1857 Figure 7 Referred specimens: UCMP 116085, partial skeleton including: partial cranium, mandible, left scapula, left ulna, left radius diaphysis, partial ilium, right trapezium, and miscellaneous unidentified fragments, mostly cranial material and portions of limb bone diaphyses. Description: About 20-25% of the shattered cranium is represented. Part of the occipital region remains intact but most fragments average only 10-20 cm in size. The upper third molars are complete. The mandible with both third molars is mostly complete. The left coronoid process is Page 54 Stevenson Bridge Mammoth and Ground Sloth Dundas and Cunningham B Figure 6. UCMP 116084, Glossotherium harlani cranium (A) and mandible (B) from the Stevenson Bridge locality, V76199. Bar scale = 10 cm. partially broken while most of the posterior portion of the right dentary is missing. Nearly all of the supraspinous fossa and about one-third of the infraspinous fossa are missing on the scapula. The scapular tuberosity, coracoid process, acromion process and mid-spinous process are broken. The left ulna is missing the distal articular end and the olecranon process is broken. The right trapezium is complete with minor damage to the articular surfaces for the trapezoid and scaphoid. Discussion: Confusion exists regarding North American mammoth taxonomy (Maglio, 1973; Kurten and Anderson, 1980; Graham, 1986; Agenbroad and Barton, 1991) and until a comprehensive revision is undertaken it is particularly important to cite the taxonomy being used. Here the taxonomy follows Maglio (1973) since it is the most often cited. Dental measurements on the M3s are presented in Table 1 using procedures outlined by Maglio (1973). Measurements on both teeth were the same. The Dundas and Cunningham Stevenson Bridge Mammoth and Ground Sloth Page 55 Figure 7. UCMP 116085, Mammuthus columbi upper third molars (A) and mandible (B) from the Stevenson Bridge locality, V76199. Bar scale = 10 cm. Page 56 Stevenson Bridge Mammoth and Ground Sloth Dundas and Cunningham Table 1. Measurements of Mammuthus left and right M3 from Stevenson Bridge compared with characteristics of late Pleistocene North American mammoth taxa compiled from Maglio (1973). Characteristics UCMP 116085 M. columbi (typical form) M. columbi (derived form) M. primigenius lamellar frequency (lophs per 100 mm) 6 5-7 7-9 7-12 enamel thickness ZOmm 2.0-3.0 mm 15-2.0 mm 1.0-2.0 mm plates per tooth 21 20-24 24-30 20-27 Table 2. Localities in the vicinity of the Stevenson Bridge site (V76199). A voucher specimen is listed for each taxon. UCMP Locality Number Locality Name Taxa County V5430 Putah Creek 1 Smilodon UCMP 44932 Yolo V6911 Putah Creek Nursery Mammuthus UCMP 82994 Yolo V69182 Putah Creek 2 Mammuthus UCMP 139288 Solano V69183 Putah Creek 3 Mammuthus UCMP 139289 Glossotherium UCMP 139290 Solano V69184 Putah Creek 4 Glossotherium UCMP 139291 Solano Dundas and Cunningham Stevenson Bridge Mammoth and Ground Sloth Page 57 lamellar frequency, enamel thickness and plate number support the assignment of these specimens to Mammuthus columbi (Table 1), in particular the typical southern population of Columbian mammoths as classified by Maglio (1973). All specimens are from an adult, probably one individual. Measurements were not taken on specimens other than teeth because of the incomplete nature of most elements resulting in the lack of useful measurements. AREA LOCALITIES The Stevenson Bridge locality is one of several fossil sites in the area along Putah Creek. Nearby localities contain ground sloth and mammoth remains with one locality featuring a Smilodon canine (UCMP 44932) (Table 2). Two of these sites, V69182 and V69183, probably represent part of the same bone assemblage as Stevenson Bridge because both contain mammoth remains of similar preservation as Stevenson Bridge and are located directly across the stream on the south bank of Putah Creek. The Yolo/Solano county line runs along Putah Creek in this area, thus localities directly across the stream from one another are in different counties. Errors in Jefferson (1991, p.102) with regard to the location and taxa present at these sites should be amended according to the descriptions in Table 2. CONCLUSIONS Although initiated as a project to document well preserved ground sloth and mammoth specimens, this paper quickly evolved into a taphonomic study utilizing data collected nearly 20 years ago by other researchers, underscoring the importance of good field and laboratory techniques, especially in data collection. As often noted, paleoecological inferences, and resulting evolutionary studies, can only be accurately assessed if taphonomic biases are known (Voorhies, 1969; Behrensmeyer, 1988; Fiorillo, 1988) and taphonomic interpretations depend on accurate and precise field data as well as the application of appropriate excavation and preparation techniques. As Fiorillo (1988) points out, many taphonomic processes (i.e. weathering, trampling, carnivore activity, fluvial transport, etc.) are recognizable through close examination of bone surfaces. Coupled with data on the depositional environment, this provides the necessary information for making paleoecological inferences. Although a taphonomic study was not the original goal during excavation of this locality, the recording of some data permitted later taphonomic inferences to be made. This serves as a prime example of the significance of how field planning and utilization of appropriate scientific methods can provide important information for future studies even if the data collected during field work will not be used as a part of the current scope of a project. ACKNOWLEDGEMENTS We express our appreciation to Mrs. J. Jacob for allowing access to examine the locality on her land; J. McCoy provided field assistance. Thanks to G. J. West, J. H. Hutchison, and A. D. Barnosky, for their reviews of the manuscript and for discussions about the project. Also thanks to D. Garber, A. R. Fiorillo, L. G. Nelms, C. B. Hanson, and C. J. Bell for information, advice and discussions which improved the study. This paper is contribution number 1608 of the University of California Museum of Paleontology. Funds for collection and preparation of the fossil specimens were provided by the Department of Anthropology at UC Davis. REFERENCES Agenbroad, L. D. and B. R. Barton. 1991. North American Mammoths An Anotated Bibliography 1940-1990. The Mammoth Site of Hot Springs, South Dakota, Inc., Scientific Papers, volume 2, Hot Springs, South Dakota, 118p. Allen, J. R. L. 1965. A Review of the Origin and Characteristics of Recent Alluvial Sediments. Sedimentology, 5(2)^9-191. Behrensmeyer, A. K. 1978. Taphonomic and Ecologic Information from Bone Weathering. Paleobiology, 4(2):150-162. Behrensmeyer, A. K. 1988. Vertebrate Preservation in Fluvial Channels. Palaeogeography, Palaeoclimatology, Palaeoecology 63:183-199. Behrensmeyer, A. K. 1990. Ch.3, Taphonomy, Transport Hydrodynamics, Bones. Pages 232- 235 in Paleobiology, A Synthesis (D. E. G. Briggs and P. R. Crowther, eds.), Blackwell Scientific Publications, Oxford, England. Page 58 Stevenson Bridge Mammoth and Ground Sloth Dundas and Cunningham Behrensmeyer, A. K. 1991. Ch. 6 Terrestrial Vertebrate Accumulations. Pages 291-335 in Taphonomy: Releasing the Data Locked in the Fossil Record, Volume 9 of Topics in Geobiology, (P.A. Allison and D.E. Briggs, eds.), Plenum Press, New York. Behrensmeyer, A. K., K. D. Gordon and G. T. Yanagi. 1986. Trampling as a cause of bone surface damage and pseudo-cutmarks. Nature, 319:768-771. Fiorillo, A. R. 1984. An Introduction to the Identification of Trample Marks. Current Research in the Pleistocene, 1:47-48. Fiorillo, A. R. 1987. Trample Marks: Caution from the Cretaceous. Current Research in the Pleistocene, 4:73-75. Fiorillo, A. R. 1988. Taphonomy of Hazard Homestead Quarry (Ogallala Group), Hitchcock County, Nebraska. Contributions to Geology, University of Wyoming, 26(2):57- 97. Fiorillo, A. R. 1989. An Experimental Study of Trampling: Implications for the Fossil Record. Pages 61-72 in Bone Modification (R. Bonnichsen and M. H. Sorg, eds.), Center for the Study of the First Americans, University of Maine, Orono. Fiorillo, A. R. 1991. Pattern and Process in Bone Modification. Anthropologic 29(3):157-161. Graham, R. 1986. Appendix 2, Part 1, Taxonomy of North American Mammoths. Pages 165- 190 in The Colby Mammoth Site, Taphonomy and Archaeology of a Qovis Kill in Northern Wyoming (G. C. Frison and L. C. Todd, eds.). University of New Mexico Press, Albuquerque. Haynes, G. 1980. Evidence of carnivore gnawing on Pleistocene and Recent mammalian bones. Paleobiology, 6(3):341-351. Haynes, G. 1983. Frequencies of spiral and green- bone fractures on ungulate limb bones in modern surface assemblages. American Antiquity, 48(1):102-114. Haynes, G. 1988. Longitudinal Studies of African Elephant Death and Bone Deposits. Journal of Archaeological Science, 15:131-157. Haynes, G. 1991. Mammoths, Mastodonts, and Elephants, Biology, behavior, and the fossil record. Cambridge University Press, Cambridge, England, 413p. Jefferson, G. T. 1991. A Catalogue of Late Quaternary Vertebrates from California: Part Two, Mammals. Technical Reports n.7, Natural History Museum of Los Angeles County, 129p. Kurten, B. and E. Anderson. 1980. Pleistocene Mammals of North America. Columbia University Press. 442p. Maglio, V. J. 1973. Origin and Evolution of the Elephantidae. Transactions of the American Philosophical Society, N.S., 63(3):1-149. Olsen, S. L. and P. Shipman. 1988. Surface Modification on Bone: Trampling Versus Butchery. Journal of Archaeological Science, 15:535-553. Reineck, H.-E. and I. B. Singh. 1980. Depositional Sedimentary Environments, 2nd edition. Springer-Verlag, Berlin. 551p. Saunders, J. J. 1977. Late Pleistocene Vertebrates of the Western Ozark Highlands, Missouri. Illinois State Museum, Reports of Investigations, n.33,119p. Stock, C. 1925. Cenozoic Gravigrade Edentates of Western North America with special reference to the Pleistocene Megalonychinae and Mylodontidae of Rancho La Brea. Carnegie Institution of Washington Publication 331:1-206. Todd, L. C. and G. C. Prison. 1986. Ch. 2 Taphonomic Study of the Colby Site Mammoth Bones. Pages 27-90 in The Colby Mammoth Site, Taphonomy and Archaeology of a Qovis Kill in Northern Wyoming (G. C. Frison and L. C. Todd, eds.), University of New Mexico Press, Albuquerque. Voorhies, M. R. 1969. Taphonomy and Population Dynamics of an Early Pliocene Vertebrate Fauna, Knox County, Nebraska. Contributions to Geology, University of Wyoming, Special Paper 1, 69p. Dundas and Cunningham Stevenson Bridge Mammoth and Ground Sloth Page 59 Appendix 1. Skeletal measurements of Glossotherium harlani from Stevenson Bridge (V76199) compared with specimens from Rancho La Brea. The Rancho La Brea specimen data is taken from Stock (1925); approximate values were not used. All measurements are in millimeters. Measurements UCMP 116084 Rancho La Brea Stock (1925) N Range Mean Cranium Length from anterior margin of maxillaries to posterior end of occipital condyles. 502 25 461.8-527.6 496.8 Length of palate from anterior end of maxillaries to postpalatine notch. 229 21 176.5-237.8 2145 Greatest width measured across ventral surface anterior to first teeth. 136 26 118.4-156.0 138.1 Width of palate measured between middle of inner sides of second upper teeth. 67 27 61.8-90.5 73.7 Width of palate measured between inner sides of anterior lobes of fifth upper teeth. 41 27 41.7-62.5 51.7 Least width posterior to fifth upper teeth. 63 28 57.8-77.3 67.2 Greatest width across occipital condyles. 129 25 120.0-143.9 132.0 Transverse diameter of foramen magnum. 52 26 35.8-52.9 44.3 Dorso-ventral diameter of foramen magnum. 43 26 32.9-42.0 37.4 Least width behind postorbital processes. 103 29 96.4-134.0 113.0 Height measured from plane of basioccipital to dorsal plane. 144 27 119.7-157.0 138.3 Page 60 Stevenson Bridge Mammoth and Ground Sloth Dundas and Cunningham Appendix 1 continued. Measurements UCMP 116084 Rancho La Brea Stock (1925) N Range Mean Mandible Length from anterior end of symphysis to posterior end of condyle. 387 25 353.1-409.8 379.8 Greatest length of symphysis. 124 21 90.0-116.4 102.8 Greatest pre-dental width. 133 19 102.7-145.4 132.9 Depth of ramus between third and fourth teeth, measured normal to inferior margin. 85 32 69-97.1 85.4 Upper teeth Length of tooth row, alveolar measurement. 148 20 111.3-155.5 1325 1st antero-posterior diameter. 18 8 14.9-22 17.9 1st transverse diameter. 15 7 13.2-20 15.2 2nd antero-posterior diameter. 29 21 28.8-41.3 33.4 2nd transverse diameter. 16 18 14.8-20.7 17.4 3rd antero-posterior diameter. 26 20 22.2-32.1 27.1 3rd transverse diameter. 23 17 17.8-26.1 22.8 4th antero-posterior diameter. 25 16 20.1-27.8 23.1 4th transverse diameter. 23 18 19.1-28.1 23.9 5th antero-posterior diameter. 22 18 20.9-30.4 24.5 5th transverse diameter. 18 16 16.5-22.8 19.2 Lower teeth Length of tooth row, alveolar measurement. 143 18 126.7-151.4 138.9 1st antero-posterior diameter. 25 18 13.2-26.1 21.6 1st transverse diameter. 14 18 12.5-20.2 16.0 2nd antero-posterior diameter. 24 16 24.0-31.3 27.9 2nd transverse diameter. 23 14 19.2-26.5 22.2 3rd greatest diameter across occlusal surface. 30 15 24.6-33.8 28.8 3rd diameter of occlusal surface normal to greatest diameter. 19 12 14.5-20.5 17.8 4th antero-posterior diameter. 54 24 44.4-59.9 51.9 4th greatest diameter of anterior lobe. 30 16 18.9-32.3 23.9 4th greatest diameter of posterior lobe. 23 18 16.1-24.0 20.5 Dundas and Cunningham Appendix 1 continued. Stevenson Bridge Mammoth and Ground Sloth Page 61 Measurements UCMP 116084 Rancho La Brea Stock (1925) N Range Mean Atlas Greatest transverse width across lateral wings. 220.3 minimum 8 211-237 221.7 Greatest antero-posterior diameter. 94 8 77.7-96.8 87.6 Antero-posterior diameter of ventral wall of neural canal along median line. 39 7 33.2-43.9 38.3 Greatest transverse distance between posterior borders of facets for axis. 100 7 89.4-102.3 96.4 Least transverse distance between anterior borders of facets for axis. 46 8 39-55.5 46.0 Greatest antero-posterior diameter of lateral process. 79 8 77.7-97.8 84.5 Axis Greatest length along median line of ventral surface. 81 minimum 6 82-92.3 86.9 Least width behind articulating surfaces for atlas. 89 6 76.0-88.0 84.1 Dorso-ventral diameter of centrum across posterior surface. 42 6 45-49.9 47.7 Greatest width of centrum. 57 6 53-62.7 58.7 Greatest transverse diameter of neural canal at anterior end. 55 6 45-61.4 50.0 Least distance from anterior border of neural canal to border of notch below posterior zygapophyses. 22 6 24.3-29.2 26.3 Greatest width across outer ends of lateral facets for atlas. 97 6 85.8-102 98.1 3rd cervical vertebra Length of centrum measured over ventral surface. 30 minimum — - 38 Width across centrum measured over anterior face and between inner borders of vertebrarterial canals. 73 — 75.3 Depth of centrum measured over anterior face and normal to dorsal surface. 45 — 50.3 Width across outer sides of anterior zygapophyses. 113 — 114 Greatest width across posterior zygapophyses. 97 — 91.7 Page 62 Stevenson Bridge Mammoth and Ground Sloth Appendix 1 continued. Dundas and Cunningham Measurements UCMP 116084 Stock (1925) Mean Scapula Length, measured from outer border of glenoid cavity to supra-scapular border and along the base of the spine. 312 327.7 Greatest length, measured from end of clavicular facet to supra-scapular border and parallel to length of spine. 418 435 Greatest width of blade, measured between ends of supra-scapular border. 400 estimate 422.1 Width measured below base of spine. 226 227.3 Height, measured from inner border of glenoid cavity to point directly above on surface of acromial process. 149 162.5 Greatest antero-posterior extent of glenoid cavity. 113 117.5 Greatest transverse width of glenoid cavity. 74 72.3 Greatest width of coraco-acromial arch. 59 68.2 Greatest diameter of supra-scapular aperture inclosed by coraco-acromial arch. 169 168.4 Measurements UCMP 116084 UCMP 139027 Stock (1925) N Mean Femora Length, measured from great trochanter to inferior surface of outer condyle. 492 26 525.2 Width, measured from inner surface of head to outer surface of greater trochanter. 262 minimum 26 282.9 Least width of shaft. 153 176 26 164.6 Greatest width across distal tuberosities (above condyles). 227 257 26 234.8 Width across condyles. 196 26 1885 Width of intercondyloid space. 47 59 26 47 Greatest width of inner condyle. 81 26 88.7 Vertical extent of inner condyle. 123 26 120 •