The Kyle Mammoth: A Late Pleistocene Columbian mammoth from southern Saskatchewan, Canada

Quaternary International 443 (2017) 79e87

Contents lists available at ScienceDirect

Quaternary International journal homepage: www.elsevier.com/locate/quaint

The Kyle Mammoth: A Late Pleistocene Columbian mammoth from southern Saskatchewan, Canada C.R. Harington Canadian Museum of Nature (Paleobiology), Ottawa, Ontario, K1P 6P4, Canada

a r t i c l e i n f o

a b s t r a c t

Article history: Received 7 June 2016 Received in revised form 30 January 2017 Accepted 15 February 2017 Available online 24 February 2017

Mammoth remains, best referred to the Columbian mammoth (Mammuthus columbi), were uncovered during roadwork near Kyle, Saskatchewan in 1964. Excavation yielded bones that gave a radiocarbon age of 12,200 ± 200 BP (radiocarbon years before present e taken as 1950) suggesting that Columbian mammoths had followed continental glacial ice as it had retreated northward from the South Saskatchewan River valley toward the close of the last glaciation. The Kyle Mammoth evidently died a natural death, perhaps becoming mired in sticky pond deposits. Bones recovered consist of several parts of the skeleton e notably a lower jaw with RM5 and RM6, several vertebrae, a scapula, a femur, parts of a left radius and ulna, and foot bones. Although the bones were scattered they represent the most complete mammoth from the Canadian Prairies (Alberta, Saskatchewan, Manitoba). Paleoenvironmental evidence suggests that the area was partly forested perhaps by a mixed-deciduous forest transition. Maximum summer temperatures could have approached 28
C. © 2017 Elsevier Ltd and INQUA. All rights reserved.

Keywords: Mammuthus columbi Late Pleistocene Kyle Saskatchewan Mammoth skeleton Paleoenvironment Radiocarbon age

1. Introduction It is a pleasure to recognize the contributions to the study of mammoths by my friend and colleague Larry Agenbroad. I got to know of his work at the Hot Springs Mammoth Site about 1978, shortly after I had the opportunity to study remains of the Kyle Mammoth at the Saskatchewan Museum of Natural History (now the Royal Saskatchewan Museum) in Regina in November 1977. We had a continuing correspondence about Canadian mammoths, and I soon realized that he viewed this group in the broadest perspective [e.g. Agenbroad (1984), Agenbroad and Barton (1991)]. Also, he kindly permitted me to measure rare bones of the Hot Springs mammoth skeletons while they were in place on public display. Because of previous work on the Babine Lake Mammoth (Mammuthus cf. M. columbi) from British Columbia (Fig. 1; Harington et al., 1974), I realized that Columbian mammoths likely had spread relatively far north (55
N) in Western Canada. So it was not unusual to find the species at 50
510 N in Saskatchewan, although the prevailing view based on woolly mammoth remains from the Edmonton area of Alberta was that the Columbian mammoth was not common (but see Harington et al., 1974, Table 4)

E-mail address: [email protected]. http://dx.doi.org/10.1016/j.quaint.2017.02.015 1040-6182/© 2017 Elsevier Ltd and INQUA. All rights reserved.

on the Canadian Prairies in the Late Pleistocene. My measurements of the teeth of the Kyle Mammoth led me to conclude that it belonged to a Columbian rather than a woolly mammoth (Table 1). The most authoritative work “Origins and Evolution of the Elephantidae” (Maglio, 1973) showed that these measurements are closest to the M5 and M6 of “Mammuthus armeniacus (Falconer) 1857” (Maglio, 1973, Table 31) e a synonym for the Trogontherian mammoth (Mammuthus trogontherii) and transitional between Mammuthus meridionalis and Mammuthus primigenius (Maglio, 1973, p. 60), which would have its North American equivalent in the Columbian mammoth (Maglio, 1973, p. 62). Limitations of the present account result from: (1) the poor preservation of many of the bones; (2) the fact that many of the bones cannot be specifically located on the original excavation diagram (Fig. 2); (3) a lack of specific catalogue numbers for most of the bones, forcing me to designate many bones with block alphabetical designations; and (4) a lapse of time between my initial research and preparation of this manuscript, when photographs of the bones and samples of the matrix (to help establish the paleoenvironment) were lost. Sketches help to overcome the missing photographs, and paleoenvironmental evidence based on a study of ostracodes and a shell (see Paleoenvironment section below) help to remedy that problem.

80

C.R. Harington / Quaternary International 443 (2017) 79e87

Fig. 1. Map showing locations of: 1. Kyle Mammoth, Saskatchewan. 2. Lindsay Mammoth (Mammuthus columbi), eastern Montana. 3. Columbian mammoth (Mammuthus cf. M. columbi), Babine Lake, British Columbia.

Despite these limitations, the purpose of this paper is to describe and identify the Kyle Mammoth, describe its geographical (Fig. 1) and geological setting, discuss the radiocarbon and individual age of the skeleton, provide measurements and sketches of the best specimens present, and include a rough diagram of the distribution of the bones (Fig. 2), as well as photographs of the excavation in October 1964 (Fig. 3a, b, c, d). 2. The find, its geographic and stratigraphic setting Information regarding the setting is mainly derived from Kehoe (1964). On the morning of October 19, T. Kehoe and a group from the Saskatchewan Museum of Natural History (now RSM) in Regina travelled to Kyle to excavate what turned out to be the most complete mammoth skeleton (about 70 associated specimens) from the province and the Prairies (Alberta, Saskatchewan, Manitoba). An estimated 20,000 people visited the site during excavation along with newspaper, radio and television reporters. The site is about 60 km NNW of Swift Current, and was in the south road ditch in NE¼ S5, T21, R16, W3 (50
510 N, 108
070 W). The matrix enclosing the bone was oxidized, clayey, contorted fossiliferous sand lying regionally below 6 inches (15.2 cm) of soil and 4 feet (1.3 m) of massive lacustrine clay (Fig. 3a). Since the mammoth

remains occurred in a rigorous environment experiencing wetting and drying, freezing and thawing, as well as leaching and staining, they were generally fragile and poorly preserved (letter of E.A. Christiansen to E.J. McCallum of November 2, 1964). At first there was not much to see. On top of a knoll beside the municipal road a small area was roped off (Fig. 3b, c, d). Inside the rope a few workers carefully removed dust from crumbling bones, stopping to apply a mixture of shellac and alcohol every few minutes. Then the bones were wrapped in layers of burlap soaked in plaster of Paris. The plaster bundles were left in a shed to harden and finally trucked to the museum in Regina. A few bones escaped this treatment. The principal value of the Kyle Mammoth lies in the scientifically controlled excavation which allowed removal of bones for radiocarbon dating, potentially yielding a reliable age for the presence of mammoths in the area. The original excavators had hoped that the Kyle Mammoth would provide evidence of having been killed by human hunters. However the mammoth evidently died of natural causes, its body becoming mired in a pond left by melting glacial ice. Drying of the pond, followed by erosion and then the gathering of glacial runoff into a lake, shifted some of the bones, resulting in many of them being covered by a protective layer of lake-deposited clay. During the last few thousand years, the disappearance of the

C.R. Harington / Quaternary International 443 (2017) 79e87

81

Table 1 Summary of comparative measurements (mm) of M6s and M5s of the Kyle Mammoth, Columbian grade mammoths (taken here as near Mammuthus armeniacus) and woolly mammoths (Mammuthus primigenius) e see Maglio 1973, Tables 31, 32.a

Kyle Mammoth RM6 M. armeniacus M6

M. primigenius M6

Kyle Mammoth RM5 M. armeniacus M5

M. primigenius M5

P

L

W

H

LF

ET

M

20

307.0

83.4

130.6

7.0

2.0

M OR N

18.3 15e21 15

298.2 236.0e340.0 14

87.6 70.0e113.0 21

139.5 96.0e160.0 21

6.3 5.0e7.2 22

2.3 1.8e3.0 20

M OR N

21.8 20e25 5

267.4 207.0e320.2 5

87.6 65.0e100.0 8

137.8 123.0e184.1 8

8.5 6.5e10.2 8

1.5 1.3e2.0 8

M

8

e

86.0

e

6.0

2.0

M OR N

11.5 10e14 8

200.5 187.0e220.0 8

79.4 70.0e91.0 8

124.3 101.0e176.2 7

6.4 5.5e7.9 8

2.0 1.5e3.0 8

M OR N

15.3 15e16 7

174.3 147.0e185.0 5

67.0 43.0e85.0 10

121.9 100.0e136.0 7

9.2 7.6e11.4 9

1.3 1.0e2.0 8

a Abbreviations: R e right, M6 e lower sixth molar tooth, M5 e lower fifth molar tooth, M e mean, OR e observed range, N e number of specimens in sample. Note that M6 replaces M3 and M5 replaces M2 according to the nomenclature used here.

glacial lake allowed erosion to again work on the landscape. But the buried mammoth remains were not disturbed until William McEvoy of Lacadena, Saskatchewan, scraped away the clay while constructing a road west of Kyle. McEvoy realized the unusual nature of the bones he exposed and notified authorities who arranged for scientific investigation at the Saskatchewan Museum of Natural History (now the Royal Saskatchewan Museum). The success of the project was largely due to the work of Saskatchewan Archaeology Society members who volunteered to aid the Museum crew (T. Kehoe, Gene Gryba, G. Watson, and Albert Swanston) (Kehoe, 1964; MacEwan, 1964).

Laboratory. This is not surprising considering that the bone fraction (apatite) commonly gives erroneous results (Haynes et al., 1971, Harington, 2003). I note that Agenbroad (1984) incorrectly cited this date as 8680 ± 400 BP. In August 2016 with help from T. Tokaryk (Royal Saskatchewan Museum), I submitted a 20 g sample of bone from the base of the scapula for dating at the Keck Radiocarbon Laboratory, California. The sample yielded no collagen because it lacked cortical bone so no radiocarbon age could be derived (personal communication: J. Southon August 29, 2016). 4. Paleoenvironment

3. Radiocarbon age Two samples of bone were submitted for radiocarbon dating. The first, based on collagen analysis at the Saskatchewan Research Council, Saskatoon yielded an age of 12,000 ± 200 BP (S-246), and is the date I accept here. The dated sample was from a mammoth vertebra (the deepest bone at the site) taken 6 inches (15.2 cm) below the distal end of the scapula in oxidized clayey, fossiliferous sand. This sample, collected by T. Kehoe in 1964, was picked up on a clean trowel, wrapped directly in aluminum foil and submitted by Earl Christiansen. He stated “scattered skeletal remains and cemented pond deposits in which the mammoth was buried indicate that the remains were disturbed during melting of stagnant ice after the mammoth died 12,000 yr ago” (McCallum and Wittenberg 1968; Harington, 2003) when the ice front stood at ice frontal position 4a (which would have been about 64 km NE of Kyle according to Dyke et al., 2003). The second date, based on acid-soluble organic matter from a vertebra about 2 m below the surface at the site was collected by an amateur in 1964 and submitted by Kehoe to the Arizona Radiocarbon Laboratory. The second sample yielded an age of 8650 ± 400 BP (A-619), which was significantly younger than that obtained by the University of Saskatchewan Radiocarbon

No samples of the matrix surrounding the bones were collected for pollen analysis, so little can be said about the specific vegetation covering the area where the Kyle Mammoth died. However, indirect evidence is available. Kehoe (1964, field notes p. 2) mentioned that sediment surrounding the mammoth bones contained gastropod and pelecypod shells. In April 1989, Jane Topping (a malacologist with the Canadian Museum of Nature) identified Stagnicola in matrix I found adhering to the scapula of the Kyle Mammoth. Based on present distribution, ubiquity and type of substrate, it most likely represents the Common Stagnicola [Stagnicola (Stagnicola) elodes] occurring throughout Canada below the tree line and south in the United States to about 38
N. It is found in all kinds of aquatic habitats, especially in thick vegetation and on muddy substrates (Clarke, 1981, p. 142e143). L.D. Delorme (1983, p. 3e5) noted that the Kyle Mammoth died in a permanent pond which was strongly influenced by a permanent stream. Based on his analysis of ostracodes (a very powerful paleoenvironmental indicator) in the matrix, the pond could have been between 0.2 and 1.1 m deep at a distance from shore up to 12 m. The pond's substrate would have been very soft and supported abundant aquatic macrophytes. The water was calcium-bicarbonate in type. The interpreted summer

82

C.R. Harington / Quaternary International 443 (2017) 79e87

Fig. 2. Bones recovered by Saskatchewan Museum of Natural History (now Royal Saskatchewan Museum) at the Kyle Mammoth site. After T.F. Kehoe (note list is incomplete).

surface water temperature of 19
C and summer bottom water temperature of 18
C support the claim that the depth of water was less than 1.1 m. The mean annual temperature of the air was about 2.4
C, however it could have been as low as 0.3
C or as high as

6.5
C. At the site of the modern analogue, temperature is 1.7
C. The maximum summer temperature could have approached 27.8
C. This climatic scenario could have supported a mixeddeciduous forest transition with broad-leafed trees being

C.R. Harington / Quaternary International 443 (2017) 79e87

83

Fig. 3. (a) Initial cutting in the road bank at the Kyle Mammoth site showing the oxidized, clayey contorted fossiliferous matrix; (b) Shows general flatness of the terrain at the Kyle Mammoth site (left side) near the road; (c) Excavation of the Kyle Mammoth site. T.F. Kehoe, leader of the excavation, is seen at left in white shirt; (d) View of the broader area of the Kyle Mammoth site with excavators (center) and onlookers to right. Photographs by Earl Christiansen.

predominant. The general picture indicates a 74% probability of forest existing, 26% for grassland, and no probability of tundra. From a comparative paleoenvironmental viewpoint, it is worth noting that remains of the nearly complete Lindsay Mammoth (Mammuthus columbi) radiocarbon dated to between 12,330 and 11,500 yr BP were found south of the Saskatchewan border near Lindsay, eastern Montana (Davis and Wilson, 1985; Huber and Hill, 2003, Fig. 1; Hill and Davis, 2014). Of 11 measurements obtained between 1973 and 2013, those with the lowest standard deviations range from 12,220 ± 35 [(Stafford Research Laboratories) (SR-8254) (Center for Accelerator Mass Spectrometry, Lawrence Livermore Laboratories, California CAMS-127309) to 12,300 ± 35 yr BP (SR8253) CAMS-127308] … very close to the date I accept for the Kyle Mammoth. Pollen samples were relatively high in non-arboreal pollen (mainly sedge, grass and fern). The presence of spruce (Picea) may indicate cool, moist climatic conditions. Stable isotope measurements of collagen from the mammoth bone are within the range associated with a C3 photosynthetic pathway, indicating a vegetated landscape potentially associated with cool, late-glacial climates.

5. Descriptive paleontology Order Proboscidea Illiger, 1811. Family Elephantidae Gray, 1821. Genus Mammuthus Burnett, 1830. Mammuthus columbi (Falconer, 1857). Since RM5 shows only seven, heavily worn enamel plates, being replaced by RM6 (approximately 20 enamel plates, the anterior five of which are worn), an individual of middle age (perhaps 36 years using African elephant ages as an approximate standard) is represented (Laws, 1966; Shoshani, 1996, Fig. 2.5).

5.1. Specimens Tusk e A fragment (RSM 8810) approximately 0.3 m long was recovered but not well preserved. All other bone measurements are in mm. Mandible e A fragmentary right mandible (RSM 8810) with the anterior part of the horizontal ramus broken near the symphysis containing a heavily worn RM5 and a partly worn RM6 (it was originally separated from the partial mandible containing RM5 and lay about 0.6 m east of it e see diagram Fig. 2). RM5 TL 136 (est.), MW 86, PF 6 plates/100 mm [Skeels (1962) gives 7e10 plates/ 100 mm for Mammuthus jeffersonii from Michigan and the PF of 6 clearly fits the range between about 6 and 8 of M. columbi rather than 10 to 11 in M. primigenius and the ET of 2.0 is closer to the range of 1.7e2.3 for M. columbi rather than 1.2 to 1.3 for M. primigenius (Agenbroad et al., 1994, Figs. 102e103)], N 8 (all worn), ET 2.0. RM6 TL 307.0, MW 83.4, H 170.6 PF 7 plates/100 mm, N 20, ET 2.0. Minimum depth of the mandible just posterior to the symphysis is 237.0, and the height at the last worn plate of RM6 is 231.0. A sketch of the reassembled mandible is shown in Fig. 4. Vertebrae e Cervical vertebrae (RSM 8810 A, B). A seventh measures: MTDC 194, VDC 157, APDC 64, MW 440 (est.), MWN 125, VDNC 48. A sixth (?) with broken spine measures: MTDC 179, VDC 162, APDC 54, MD as restored 325 (est.), MWNC 91, VDNC 92. e Thoracic vertebrae RSM 8810 C, D (neural spine), E.F.G. A second or third having anterior and posterior rib facets: MTDC 220, VDC 147, APDC 67, HS 320, MDV 364, MWNC 108, VDNC 107. Fourth? (with the longest neural spine preserved that flares at top): LS 530. Seventh? Has a long spine sloped strongly back: MTDC 145, VDC 143, APDC 68, HS 440, MDVTP 240, MWNC 61, VDNC 67. Fifth? (with well-marked rib facets): MTDC 132, VDC 129, APDC 83, AHS 298, MDVTP 235, MWNC 50, VDNC 70. Near fourth?

84

C.R. Harington / Quaternary International 443 (2017) 79e87

Fig. 4. Medial view of fragmentary right mandible of the Kyle Mammoth as restored with RM6 behind RM5. Sketch by author.

(with facets for ribs present, but left transverse process is damaged and epiphyseal plates are missing from centrum): MTDC 124, VDC 122, APDC 78þ, AHS 173þ, MWNC 56, VDNC 67. e Lumbar vertebrae RSM 8810 H, I, J.K. H (anterior?) has broken spine and the posterior epiphyseal plate is missing: MTDC 126, VDC 134, APDC 83, MDV 240, MWNC 64, VDNC 57. I (anterior to middle) lacking posterior epiphyseal plate: MTDC 145, VDC 119,

APDC 82þ, AHS 234, MDV 241, MWNC 55, VDNC 52. J (posterior?) has a thick centrum and stubby spine swept well back: MTDC 141. VDC 139, APDC 101, AHS 191þ, MDV 247, MWNC 57, VDNC 56. K (poor condition lacking epiphyseal plates): MTDC 130, VDC 134, APDC 78þ, spine broken near base, therefore AHS e, MDV e (broken), MWNC 83, VDNC 61.

Fig. 5. Lateral view of right scapula of the Kyle Mammoth. Note broken anterior margin. Sketch by author.

C.R. Harington / Quaternary International 443 (2017) 79e87

85

Fig. 6. Anterior view of partial left forefoot [including left metacarpal II(H), III(G) and IV(F)] of the Kyle Mammoth. Sketch by author.

e Rib (presumably #3 in diagram Fig. 2) e VDHR 73, APDHR 55, MWR 81, MAPDR 34, LICR 1170. e Right scapula (Fig. 5, virtually complete except for proximal part of anterior margin – partly restored, #22 in diagram Fig. 2) e LGC 233, WGC 155 (approx.), MVH 950, MAPL 670 (est.), MMLD 255, APLN 303. e Right (?) humerus e part of distal articular surface e incomplete therefore no measurements. e Left radius (distal end broken near fusion with shaft): WAS 131. e Left ulna (part of distal end): MWDAS 134, MLDAS 172. e Left? trapezium fragment: no useful measurements. e Partial left forefoot (Fig. 6, restored e probably #2 on diagram Fig. 2). e Carpals (proximal row) e A (slightly worn on upper surface): MW 126, MAPD 66, MH 47þ. B: MW 116, MAPD 71, MH 68.

e (distal row) C: MW 87, MAPD 146, MH 141. D: MW 105, MAPD 146, MH 96 (est.), E: MW 57, MAPD 115, MH 147. e Metacarpals F: TL 200, PW 92, PD 121, MW 86, MD 68, DW 95, DD 102. G: TL 213, PW (est.) 92, PD 123, MW 77, MD 65, DW 88, DD 92, H: TL 202, PW 92, PD 121, MW 73, MD 66, DW 82 (est.), DD 85. e Phalanges (first) I: TL 97, PW 75, PP 77, MW 57, MD 55, DW 64, DD 50. J: TL 99, PW 82, PD 77, MW 67, MD 54, DW 72, DD 46þ (est.). K: TL 100, PW 73, PD 66, MW 61, MD 55, DW 66, DD 48. e Phalanges (second) L: TL 57, PW 49, PD 43, MW 62, MD 40, DW 52, DD 32. M: TL 57, PW 50, PD 36, MW 61, MD 35, DW 50, DD 35. N: TL 37, PW 55, PD 36, MW 67, MD 39, DW e, DD e e Phalanges (third) O: L 21, W 76, H 20. P: L 23, W 69, H 23. e Sesamoids: P, R, S, T, U, V (not measured).

86

C.R. Harington / Quaternary International 443 (2017) 79e87

e Right femur (shaft only #12 in Fig. 1): MSW 185, MSD 115, MSC 482. 6. Discussion The best current perspective on mammoth evolution is provided by Lister and Sher (2015). These authors consider that the earliest North American mammoths resembled the advanced Eurasian M. trogontherii that covered the Bering Isthmus about 1.5 million years ago, giving rise to M. columbi, and that the woolly mammoths (M. primigenius) later evolved in Beringia and spread into Europe and North America leading to a diversity of morphologies as they encountered endemic M. trogontherii in Europe and M. columbi in North America. That subsequent dispersal led to intermediates (M. jeffersonii, here designated M. columbi jeffersonii), suggesting introgression of M. primigenius with M. columbi. A more recent contribution on this subject (Enk et al., 2016) strongly suggested that various nominal mammoth species interbred, perhaps extensively. The authors hypothesized that at least two distinct stages of interbreeding between conventional paleontological species are likely responsible for this pattern e one between Siberian woolly mammoths and resident North American populations that introduced woolly mammoth phenotypes to the continent, and another between ecomorphologically distinct populations of woolly and Columbian mammoths in North America south of the continental ice sheets. Perhaps Columbian mammoths south of the continental ice in western North America spread northward from eastern Montana and southern Saskatchewan toward Kyle about 12,000 B.P. prior to their regional extinction. 7. Conclusions 1. The Kyle Mammoth was discovered in 1964 and although overdue for study 52 years later, remains the most complete mammoth known from the Canadian Prairies. 2. The Kyle Mammoth represents a middle-aged individual (perhaps 36 African elephant years) that was excavated from glacial lake clay. Evidently it became bogged down in sticky clay and died. There are no signs on the bone of alterations by carnivores, scavengers or humans, so it seems to have died a natural death. Some of the bones were disturbed and eroded during the course of freeze-thaw and wetting-drying cycles. 3. The currently accepted radiocarbon age of 12, 200 ± 200 yr BP indicates that Columbian mammoths had advanced northward from eastern Montana (the nearly complete Lindsay Mammoth of similar radiocarbon age) and the South Saskatchewan River valley toward the close of the last glaciation following the melting southern margin of the Laurentide Ice Sheet; however, woolly mammoths also lived in the western Prairies during the Late Pleistocene. 4. Future work should involve pollen analysis of matrix surrounding the Kyle Mammoth bones in order to better determine paleoenvironment. Indirect evidence of paleoenvironment has come from analyses of ostracodes and mollusc shells. The data suggest that the area occupied by the mammoth was partly forested e perhaps by a mixed-deciduous forest transition and that maximum summer temperatures approached 28
C. 5. Recent perspectives published during 2015 and 2016 note that the earliest North American mammoths resembled the advanced Eurasian M. trogontherii (see M. armeniacus; Table 1) of the Bering Isthmus about 1.5 million years ago, giving rise to M. columbi and to M. primigenius, and that introgression of the two species led to intermediates such as M. jeffersonii, which I consider to be a junior synonym of the Columbian mammoth.

Acknowledgements I thank T.F. Kehoe for his many contributions to this study, particularly for leading the work on the Kyle Mammoth and for providing some handwritten field notes and a diagram showing the distribution and burial of the bones. My friend and colleague Earl Christiansen supplied several photographs of the excavation and the sediments exposed, as well as submitting bone samples for radiocarbon dating, and providing copies of correspondence on the Kyle Mammoth. L.D. Delorme kindly shared his expert view of the paleoenvironment of the Kyle Mammoth site based on a study of ostracodes. Louis Robertson of the Geological Survey of Canada provided the approximate distance from Kyle to the nearest Laurentide Ice about 12,000 BP. Tim Tokaryk (Royal Saskatchewan Museum) kindly provided two radiocarbon samples from the Kyle Mammoth; Alice Telka (Paleotec Services) and John Southon (Keck Radiocarbon Laboratory, California) assisted by speeding up the radiocarbon dating. Comments by Chris Jass (Royal Alberta Museum, Edmonton), Min-Te Chen (Quaternary International) and two anonymous reviewers helped me to improve this paper. I am grateful to Gail Harington for word processing and help with the figures. References Agenbroad, L.D., 1984. New world mammoth distribution. In: Martin, P.S., Klein, R.G. (Eds.), Quaternary Extinctions. University of Arizona Press, Tucson, Arizona, pp. 90e108. Agenbroad, L.D., Barton, B.R., 1991. North American Mammoths. An Annotated Bibliography 1940e1990. Quaternary Studies Program. Northern Arizona University, Flagstaff, Arizona, pp. 1e118. Agenbroad, L.D., Lister, A.M., Mol, D., Roth, V.L., 1994. Mammuthus primigenius remains from the mammoth site of Hot Springs, South Dakota. In: Agenbroad, L.D., Mead, J.I. (Eds.), The Hot Springs Mammoth Site: a Decade of Field and Laboratory Research in Paleontology, Geology and Paleoecology. The Mammoth Site of Hot Springs, South Dakota, Inc., pp. 269e281 see p. 275. Burnett, G.T., 1830. Illustrations of the Quadrupeda, or Quadrupeds, Being the Arrangement of the True Four-footed Beasts Indicated in Outline. Quarterly Journal of Science, London, pp. 336e353. December 1829. Clarke, A.H., 1981. The Freshwater Molluscs of Canada. National Museum of Natural Sciences. National Museums of Canada, Ottawa. Davis, L.B., Wilson, M.C., 1985. The Late Pleistocene Lindsay mammoth (24DW501), eastern Montana. Curr. Res. Pleistocene 2, 97e98. Delorme, L.D., 1983. Paleoclimatic Interpretations. Environment Canada, National Water Research Institute, Burlington, Ontario, pp. 1e6. Dyke, A.S., Moore, A., Robertson, L., 2003. Deglaciation of North America. Geological Survey of Canada. Open File 1574 (Sheet 1, 1214C ka BP map). Enk, I., Dusault, A., Widga, C., 11 other coauthors, 2016. Mammuthus population dynamics in Late Pleistocene north America: divergence, phylogeography and introgression. Front. Ecol. Evol. 4 (42), 11e13. Falconer, H., 1857. On the species of mastodon and elephant occurring in the fossil state in Great Britain. Part I. Mastodon. Q. J. Geol. Soc. Lond. 13, 307e360. Gray, J.E., 1821. On the natural arrangement of vertebrate animals. Lond. Med. Repos. 15, 296e310. Harington, C.R. (Ed.), 2003. Annotated Bibliography of Quaternary Vertebrates of Northern North America e with Radiocarbon Dates. University of Toronto Press, Toronto. Harington, C.R., Tipper, H.W., Mott, R.W., 1974. Mammoth from Babine Lake, British Columbia. Can. J. Earth Sci. 11, 285e303. Haynes Jr., C.V., Grey, D.C., Lang, A., 1971. Arizona radiocarbon dates VIII. Radiocarbon 13 (1), 1e18. Hill, C.L., Davis, L.B., 2014. Multi-scalar geological and paleoenvironmental analysis of the Lindsay mammoth, Yellowstone basin, Montana. In: Poster Paper Presented at the 23rd Biennial Meeting of the American Quaternary Association. Abstracts, Seattle, Washington, pp. 64e65. Huber, J.K., Hill, C.L., 2003. Paleoecological inference based on pollen and stable isotopes for mammoth-bearing deposits of the Oahe Formation (Aggie Brown Member), eastern Montana. Curr. Res. Pleistocene 20, 95e97. Illiger, C., 1811. Prodromus systematis mammalium et avians, p. 302. Salfeld. Kehoe, T., 1964. The Kyle Mammoth. Sask. Archaeol. Newsl. 8, 1e4. Laws, R.M., 1966. Age criteria for the African elephant, Loxodonta a. africana. East Afr. Wildl. J. 4, 1e37. Lister, A.M., Sher, A.V., 2015. Evolution and dispersal of mammoths across the northern hemisphere. Science 350, 805e809. MacEwan, G., 1964. Monsters in these parts. In: Our Natural History. Calgary Herald, after October 19. Maglio, V.J., 1973. Origin and evolution of the Elephantidae. Trans. Am. Philos. Soc.

C.R. Harington / Quaternary International 443 (2017) 79e87 63, 1e149. McCallum, K.J., Wittenberg, J., 1968. University of Saskachewan radiocarbon dates V. Radiocarbon 10 (2), 365e378. Shoshani, J., 1996. Skeletal and other basic anatomical features of elephants. In:

87

Shoshani, J., Tassy, P. (Eds.), The Proboscidea. Evolution and Palaeoecology of Elephants and Their Relatives, pp. 9e20. Skeels, M., 1962. The Mastodons and Mammoths of Michigan, 47. Papers of the Michigan Academy of Science, Arts and Letters, pp. 101e133.