Re-evaluation of mastodon material from Oregon and Washington, USA, Alberta, Canada, and Hidalgo and Jalisco, Mexico

Oregon

Dooley et al. (2019) considered Pleistocene mammutid material from Oregon in their study as Mammut americanum. All of the teeth they examined were either non-diagnostic for distinguishing between M. americanum and M. pacificus (e.g., M2), or came from tooth positions with very small sample sizes (e.g., premolars). Even so, Oregon specimens of tooth positions with small sample sizes, such as P3, still had L:W ratios that were more similar to M. pacificus than to M. americanum. Dooley et al. (2019) noted that these remains were biogeographically and anatomically anomalous if assigned to M. americanum and hypothesized that they may in fact represent M. pacificus.

One specimen not examined by Dooley et al. (2019) was the Tualatin mastodon (F-30282), recovered in 1962 and currently on display in the Tualatin Public Library. Nearly the entire left side of the animal is preserved, including the left tusk and a portion of the maxilla with the preserved M2 and M3 (Fig. 1A). Based on photos of the excavation and rudimentary notes, the remains were situated approximately 3.5–5 feet (1.067–1.52 m) below the surface in a marsh. The Tualatin mastodon dates to the Late Pleistocene (Gilmour et al., 2015), post-dating the Missoula floods and human colonization of the area (Davis et al., 2019; O’Connor et al., 2020). The M3 has a L:W ratio of 1.91, which is just inside the range of M3s of M. americanum in our dataset (1.59–1.95), but well within the range of M. pacificus (1.69–2.33) (Table 1, Fig. 2). Additionally, the left femur (Fig. 1B) is complete and has a maximum length of 807 mm and mid-shaft width of 130 mm, placing it close to small M. pacificus specimens from California, (Dooley et al., 2019: Figure 25; Fig. 3), and well apart from M. americanum. These measurements indicate that the Tualatin mastodon is M. pacificus, suggesting that other Oregon material previously reported as M. americanum may be M. pacificus as well.

Another noteworthy Oregon specimen is USNM 4911, an isolated left M2 described as Mammut oregonense by Hay (1926). Dooley et al. (2019) showed that M2 does not differ in any consistent way between M. pacificus and M. americanum, even in L:W ratio. As M. oregonense is only represented by an M2 and no other specimens have ever been referred to this taxon, we concur with assessment of Dooley et al. (2019) that M. oregonense should be considered a nomen dubium, and its use restricted to the holotype.

Washington

Dooley et al. (2019) assigned three specimens from Washington to M. americanum, two mandibles that included m3s and an isolated M3, all from different localities. The isolated right M3 (UWBM 83312) from Jefferson County is a tetralophodont tooth missing large areas of enamel (Fig. 4F). There is little to no wear on lophs 4 and 5, but damage on the first three lophs make it impossible to assess the wear in these areas. The L:W ratio is 1.82, closer to the average of M. americanum (1.76) that to M. pacificus (1.98), but within the known range of values for both taxa (Fig. 2).

UWBM 88099 is a mandible from Lewis County that includes the left m2 and m3, and the right m3 (Figs. 4C–4E). The anterior tip of the mandible is damaged, as are both ascending mandibular rami, which are missing the condyles. While the anterior tip of the mandible is imperfectly preserved, there is no indication of alveoli for mandibular tusks. The L:W ratio of the left m3 is 1.94. This value is just outside of 2σ for M. pacificus (the lowest known value for M. pacificus is 1.95, average 2.26), and close to the mean value for M. americanum (1.89) (Fig. 4).

UWBM 14491 is a complete mandible from Clallam County that includes both left and right m1, m2, and m3 (Figs. 4G–4I). The m1s show heavy wear, the m2s are in moderate wear, and the m3s are not yet fully erupted and show only slight wear on the first lophids. This is most equivalent to Laws (1966) Group XVII or XVIII, indicating and age of 28–30 ± 2 African elephant equivalent years. There are no alveoli for mandibular tusks. The right m3 has a L:W ratio of 2.11. This is well within the known range for M. pacificus and is greater than all but one M. americanum in our dataset (n = 134), but still with the 2σ range of M. americanum (Fig. 5).

The Manis mastodon was not included in Dooley et al. (2019). This specimen was discovered during excavation of a holding pond near Sequim, Clallam County, Washington in 1977, and became well known because of a reported bone projectile point embedded on one of the ribs (Gustafson, Gilbow & Daugherty, 1979; Waters et al., 2011), although debate continues over the interpretation of this specimen (Haynes & Huckell, 2016; Waters et al., 2023). Carbon dates from bone collagen yielded an age of approximately 13,800 ybp (Waters et al., 2011). The Manis mastodon has never been fully described or figured, but Gustafson, Gilbow & Daugherty (1979) mention tusk segments up to 2 m in length, suggesting that the individual may have been a male. Field sketches reproduced in Gustafson, Gilbow & Daugherty (1979) indicate the presence of numerous ribs, and at least portions of a forelimb including the scapula, humerus, and ulna. They also figure a heavily worn m2, and mention numerous skull fragments.

A number of elements from the Manis mastodon are on exhibit at the Sequim Museum and Arts in Sequim, including a complete right dentary with an in situ m3 (Figs. 4A, 4B). The mandibular symphysis does not have alveoli for mandibular tusks. The m3 is pentaloph, with wear on all five lophids and heavy wear on the first two. This level of wear is consistent with Laws Group XXII or XXIII, yielding an age of 39–43 ± 2 AEY. The L:W ratio of this tooth is 2.09, within the normal range for M. pacificus and much narrower than typical M. americanum m3s (only two M. americanum specimens out of 134 in our dataset are narrower), but still within the 2σ range of M. americanum (Fig. 5).

The Washington State specimens present an interesting dilemma for taxon identification. All the known specimens lack mandibular tusks, but this absence of tusks is not diagnostic in its own right. One specimen (UWBM 88099) has an m3 that falls just outside the 2σ value of M. pacificus and close to the mean of M. americanum; however, due to the absence of lower tusks and the geographic location of this specimen, we hesitate to definitively refer it to either species. Other Washington specimens have M3/m3 L/W ratios that fall within the 2σ range of both taxa, although generally closer to the averages for M. pacifcus than M. americanum, but lack any definitive characters. Therefore, we refer all Washington specimens to Mammut sp. until such time as unequivocal diagnostic material is reported.

Hidalgo

Multiple elements of a single mastodon are known from Rancholabrean deposits at Ventoquipa, Hidalgo, Mexico (UAHMP-311; Bravo-Cuevas, Morales-García & Cabral-Perdomo, 2015) (Fig. 6). Dooley et al. (2019) included this specimen as M. americanum in their dataset, even though the L:W ratio of the M3 of UAHMP-311 is 1.93, close to the mean for M. pacificus (1.98) and close to the maximum value known for M. americanum (1.95) (Table 1, Fig. 2). Measurements of the m3 of this specimen are now available; it has a L:W ratio of 2.29. This is higher than any specimen of M. americanum in our dataset (maximum = 2.23, n = 134), and is greater than the mean for M. pacificus (2.25) (Table 2, Fig. 5). As both M3 and m3 of UAHMP-311 fall within the known range of M. pacificus, and m3 falls outside the 2σ range of M. americanum, we refer this specimen to M. pacificus.

Jalisco

A left M3 (LACM 1854) (Fig. 7) was recovered in 1955 from Lago de Chapala, near San Luis Soyatlan, Jalisco, Mexico. A rich fauna from this site includes remains from both Mammuthus and Cuvieronius (Lucas, 2008), but LACM 1854 is the only mammutid element identified thus far. The age of the Lago de Chapala specimens has been problematic, potentially ranging from Rancholabrean to Blancan, but the material from San Luis Soyatlan appears to be Rancholabrean (Lucas, 2008; Rufolo, 1998).

LACM 1854 is a tetralophodont left M3. Large portions of the tooth, including the entire root area, are encrusted with what appears to be a carbonate or other evaporitic mineral. While there is some damage to the pretrite side of the first loph, the other lophs are undamaged. The lophs are simple, lacking the additional conelets that commonly fill the troughs between lophs in gomphotheriids. There is light to moderate wear on the pretrite side of the first three lophs, with the fourth loph showing only very slight wear. There is a distinct cingulum on the anterior margin, but this does not appear to extend to other portions of the tooth.

The length:width ratio of this tooth is 2.07, well within the known range of M. pacificus (1.69–2.33; mean = 1.98) (Table 1, Fig. 2). No M. americanum M3 in our dataset has such a high L:W value (maximum = 1.95; mean = 1.77), and this value falls outside the 2σ range of M. americanum, justifying the referral of LACM 1854 to M. pacificus.

Alberta

A limited number of mammutid remains are known from Alberta, and were reviewed by Jass & Barron-Ortiz (2017). Nearly all of those records represent isolated specimens recovered from gravel pits in the Edmonton area. The most complete specimen is a partial mandible, RAM P94.16.1 (Fig. 8) from the Apex Galloway Pit, located near Edmonton. The m3 in this specimen has a L:W ratio of 1.90, well outside the 2σ range of M. pacificus and within the known range of M. americanum. Moreover, RAM P94.16.1 has well-developed alveoli for lower tusks, which are unknown in M. pacificus but occur in about 30% of M. americanum mandibles (Green, 2006), regardless of age or sex.

A second specimen, RAM P97.7.1 (Fig. 8), also comes from an Edmonton-area gravel pit (Pit 46). Although damaged, this specimen is identified as a partial left M3 based on the right angle formed by the loph axis and the long axis of the tooth (note: this specimen was reported as an m3 in Jass & Barron-Ortiz (2017)). The tooth includes five lophs, but the first two lophs are damaged, making direct measurement of the maximum tooth width impossible, as in M3 the widest part of the tooth is always at either the first or second loph.

In order to estimate the likely maximum width of RAM P97.7.1, we calculated individual loph widths as a percentage of maximum loph width for 34 Mammut M3s, including 17 specimens each of M. pacificus and M. americanum (Table 3). This enabled us to calculate a range of likely values of the maximum width of a tooth for a given width of the third loph.

RAM P97.7.1 is 183.69 mm long, while the third loph has a width of 81.27 mm. Using the values of Mammut specimens from Table 3 as a guide yields a range of likely widths for the widest loph, from 82.6–94.1 mm. These values result in a LW ratio between 1.95–2.26 (with an average of 2.12). Reconstructions of RAM P97.7.1 using the estimated maximum and minimum loph widths are shown in Fig. 8. As almost this entire range of values falls outside the 2σ values of M. americanum (1.56–1.96), we refer this specimen to M. pacificus.