Genus Cynomys

Prairie dogs (Cynomys) are large, derived ground squirrels endemic to North America throughout their history (Goodwin, 1995b). Morphological evidence suggests a close relationship with derived ground squirrels (subgenus Spermophilus; Bryant, 1945; Goodwin, 1995b); some late Blancan-early Irvingtonian taxa are morphologically intermediate between these taxa, complicating taxonomic assignment (e.g., ?S. cragini; Goodwin and Hayes, 1994).

Two subgenera and five extant species currently are recognized (Pizzimenti, 1975). The subgenus C. (Cynomys), referred to as black-tailed prairie dogs, includes C. mexicanus, today restricted to a small area in northeastern Mexico, and C. lu-dovicianus, which ranges widely across the Great Plains. The subgenus C. (Leucocrossuromys), referred to as white-tailed prairie dogs, includes three species. All three occur on high-elevation basins or plateaus associated with the Rocky Mountains (C. parvidens, southwestern Utah; C. leucurus, central Rocky Mountains and Wyoming Basin; C. gunnisoni, southern Rocky Mountains).

A recent systematic review of fossil Cynomys recognized seven or possibly eight species in pre-Holocene contexts (Goodwin, 1995b), with the lineage known as early as the late Pliocene. A striking feature of the Pleistocene record for the genus was the common and apparently persistent co-occurrence of both subgenera at sites on the central Great Plains and far southern Rocky Mountains. White-tails only occurred to the north of (and at higher elevations than) this zone of overlap; black-tails only were found to the south and east (Goodwin, 1995a). This pattern contrasts with the limited subgeneric overlap evident today. Previous work on Cynomys from Porcupine Cave has suggested the presence of at least two species representing both subgenera (Barnosky et al., 1996; Rouse, 1997). The present study recognizes three prairie dogs or prairie dog-like sciurids from the cave, describes one new species, and reinterprets subgeneric identifications.


HOLOTYPE DMNH 28320, left dentary with m1-m3 (figure 17.4B).

REFERRED MATERIAL See appendix 17.1. NISP = 82, MNI = 23.

TYPE LOCALITY AND AGE Generator Dome locality, Porcupine Cave, Park County, Colorado. Irvingtonian.

DIAGNOSIS Distinguished from Spermophilus (Spermo-philus) by complete protolophid on p4; from ?Spermophilus cragini by smaller size, distally expanded M3, and prominent metaloph on M3; and from all extant and adequately known fossil Cynomys by smaller size, lesser development of the met-alophid on m3, rounded P3, and lack of buccodistal accessory lophule of protoloph on M3.

ETYMOLOGY Named in honor of Elaine Anderson, whose enthusiasm for Pleistocene mammals in general, and for the study of Porcupine Cave specimens in particular, has contributed greatly to knowledge of Pleistocene mammals. Her encouragement was instrumental in getting the author involved in this study.

DESCRIPTION The maxilla bears a triangular infraorbital foramen, but not so strongly as is typical of Cynomys. P3 is sub-equal in length and width (mean [SD] minimum-maximum [n]: L, 2.22 mm [0.04] 2.20-2.27 mm [3]; W, 2.26 mm [0.11] 2.15-2.37 mm [3]), not nearly as wide as is characteristic of ?S. cragini (Goodwin and Hayes, 1994) or extant Cynomys. The mesiobuccal face of the tooth is not flattened, as is typical of Cynomys. The anterior and posterior cingula resemble those of ?S. cragini in most respects (Goodwin and Hayes, 1994), but

2 mm

FIGURE 1 7.4 (A-C) Specimens of ? Cynomys andersoni sp. nov. (A) L p4 (DMNH 14303, Badger Room); (B) holotype, occlusal view of m1-m3 (DMNH 28320, Generator Dome); (C) R M2 and M3 (reversed) (DMNH 28317, Generator Dome).

the anterior cingulum does not extend as far apically at its buccal juncture with the prominent oblique loph of P3.

P4 is poorly represented in the Porcupine Cave sample. Two adequately preserved specimens bear a somewhat compressed protocone, giving a subtriangular appearance. The anterior cingulum, mesiodistally expanded near its buccal end, tapers lingually to its terminus along the mesial surface of the protocone.

M1 and M2 (M2 shown in figure 17.4C) bear a progressively more robust protocone than does P4. The anterior cingulum does not taper as strongly lingually as on P4. In several specimens, the parastyle of M1 (but not M2) extends buccally beyond the margin of the paracone, a feature also seen on the type of ?S. cragini (Goodwin and Hayes, 1994). None of these teeth bears a mesostyle. The buccal half of the protoloph on P4 to M2 usually lacks the prominent distal expansion, bounded distolingually by a prominent notch, that is characteristic of Cynomys. However, one specimen (M1 or M2; UCMP 180186) exhibits this attribute incipiently, and several specimens display a notch on the distal surface of the pro-toloph but lack distal expansion of the protoloph buccal to the notch (figure 17.4C).

M3 (figure 17.4C) averages slightly longer (mean [SD] minimum-maximum [n]: 3.64 mm [0.17] 3.38-3.98 mm [10]) than wide (3.57 mm [0.22] 3.23-3.93 mm [8]); relative M3 length is greater than that exhibited by ?S. cragini (Goodwin and Hayes, 1994). Tooth shape is characteristically sub-triangular with somewhat flattened mesial, buccal, and dis-tolingual surfaces joined at rounded corners. The distolingual surface often bears a distinct notch at the distal terminus of the protocone. The anterior cingulum is shaped much as that of M1 or M2. The protoloph is well developed but never bears the prominent, buccodistal accessory lophule characteristic of Cynomys. A prominent metaloph, usually present, extends mesiobuccally from the distobuccal margin of the protocone, then turns buccally to parallel the protoloph. The metaloph

FIGURE 1 7.4 (A-C) Specimens of ? Cynomys andersoni sp. nov. (A) L p4 (DMNH 14303, Badger Room); (B) holotype, occlusal view of m1-m3 (DMNH 28320, Generator Dome); (C) R M2 and M3 (reversed) (DMNH 28317, Generator Dome).

table 17.3

Comparison of ?Cynomys andersoni with Other Primitive Prairie Dogs and Prairie Dog-Like Ground Squirrels Based on Lower Dental Variables

?Cynomys andersoni ?Spermophilus cragini









(Sample Size)


(Sample Size)



Lower alveolar row

12.40 (0.24)

12.24-12.75 (4)


Length p4

2.70 (0.23)

2.07-2.90 (11)

2.83 (0.11)

2.69-2.95 (9)


Trigonid width p4

3.15 (0.15)

2.89-3.33 (11)

3.45 (0.17)

3.10-3.67 (9)


Talonid width p4

3.28 (0.22)

2.72-3.55 (11)

Not measured


Length m1

2.56 (0.10)

2.43-2.65 (5)

2.82 (0.17)b

2.59-3.06 (12)b



Width m1

3.49 (0.16)

3.25-3.70 (5)

3.81 (0.21)b

3.53-4.14 (12)b



Length m2

2.78 (0.08)

2.65-2.90 (10)



Width m2

3.62 (0.18)

3.33-3.88 (10)



Length m3

3.76 (0.14)

3.55-4.13 (16)

3.93 (0.17)

3.69-4.32 (12)



Width m3

3.75 (0.11)

3.51-3.93 (16)

4.05 (0.11)

3.84-4.17 (12)



sources: Data for?S. cragini are from Goodwin and Hayes (1994) and represent data for left teeth (right teeth with similar statistics). Data for C. hibbardi and C. sappaensis are from Goodwin (1995b). an = 1.

bIncludes both ml and m2 (undifferentiated sample of isolated teeth).

sources: Data for?S. cragini are from Goodwin and Hayes (1994) and represent data for left teeth (right teeth with similar statistics). Data for C. hibbardi and C. sappaensis are from Goodwin (1995b). an = 1.

bIncludes both ml and m2 (undifferentiated sample of isolated teeth).

is absent from one referred specimen from Mark's Sink (DMNH 36336).

?C. andersoni is distinctly smaller than other primitive prairie dogs (C. hibbardi, C. sappaensis) or prairie dog-like ground squirrels (?Spermophilus cragini) on lower dental variables (table 17.3). The dentary is not well preserved in the Porcupine Cave assemblage but exhibits an abrupt drop of the dorsal surface anterior to the p4. On p4, width across the trigonid is subequal to or less than width across the talonid (figure 17.2A), resembling black-tailed (C. [Cynomys]) but not white-tailed (C. [Leucocrossuromys]) prairie dogs (Goodwin, 1995b). This results from prominent buccal expansion of the hypocone (figure 17.4A). A prominent protolophid and met-alophid, defining a distinct trigonid basin, unite the proto-conid and metaconid. As in Cynomys but not Spermophilus, the protolophid is typically complete and is not separated from the metaconid by a distinct notch; a rudimentary notch is present on a few specimens. The talonid basin bears little if any rugosity and is deepest along its buccal margin, typically exhibiting a trench lingual to the ectolophid. The ectolophid bears a slight swelling midway along its length on a few specimens. The lingual margin of the talonid is somewhat flattened in some specimens, then curves buccally at the entoconid, merging into the gently curving posterolophid. In other specimens (figure 17.4A), the lingual and distal margins form a nearly continuous curve. The entoconid is delimited mesially by a small but clear notch; this notch is absent or indistinct in ?S. cragini.

The m1 and m2 (figure 17.4B) take the shape of a rough parallelogram; the talonid is distinctly offset buccally relative to the trigonid. The protoconid and metaconid are not strongly expanded basally, giving the trigonid a somewhat gracile appearance. The metalophid is characteristically complete, bounding a distinct trigonid pit, but is prominently V-shaped when observed in distal view. On specimens with associated m1 and m2, the metalophid is higher and exhibits a more prominent lingual arm on m1. The talonid basin is slightly rugose in some specimens; one specimen (DMNH 23335) exhibits a low, oblique "mesolophid" on m2. The talonid is deepest along the ectolophid, forming an indistinct trench that characteristically deepens and expands as a pit at its mesial terminus, adjacent to the trigonid (figure 17.4B). The outline of the talonid varies in shape. Some specimens have flattened lingual and distal margins, connected at a rounded entoconid; other specimens exhibit a more continuous arc from metaconid to hypoconid. The entoconid is commonly not as distinct as on p4, and the hypoconid is not strongly expanded basally. It invariably projects buccomesially, often constricting the buccal entrance to the hypoflexid.

The m3 (figure 17.4B) is relatively elongate and usually sub-triangular. Bases of the protoconid and metaconid are not strongly expanded. The metalophid is variably developed. In most specimens (figure 17.4B) it extends ventrolingually from the protoconid, terminating along the distal wall of the trigo-nid, and lacks the prominent lingual arm characteristic of modern Cynomys. In others, a rudimentary lingual arm is evident, completely bounding a shallow but distinct trigonid pit. When present, this pit is placed less apically than it is in other Cynomys. The talonid basin is raised, forming a platform with characteristic rugosity. A prominent crescent-shaped trench separates this platform from the hypoconid, ecto-lophid, and protoconid and sometimes continues lingually along the distal wall of the trigonid (figure 17.4B). It usually is unobstructed or only slightly constricted along its length, as in C. (Cynomys), but in one specimen (DMNH 22829B) it is blocked by a bridge, as is characteristic of C. (Leuco-crossuromys). The trench commonly expands and deepens at a point distolingual to the protoconid (figure 17.4B). The ecto-lophid is typically expanded slightly near its midpoint (figure 17.4B), and in several specimens the swelling forms a small conulid. The hypoconid projects strongly mesially in all specimens, as in C. (Cynomys), constricting the buccal opening of the hypoflexid.

DISCUSSION Several sciurids, possibly representing multiple lineages, approached the prairie dog morphotype during the late Blancan and early Irvingtonian. C. hibbardi (Eshel-man, 1975) and C. sappaensis (Goodwin, 1995b) appear to represent the Cynomys lineage. Other taxa (e.g., ?C. vetus, ?S. cragini, and ?C. andersoni) are morphologically mosaic (Goodwin and Hayes, 1994; Goodwin, 1995b; this chapter): they combine derived features otherwise known only in Cynomys (e.g., the complete protolophid on p4 in ?S. cragini and ?C. andersoni) with primitive features otherwise known in Spermophilus but not Cynomys (e.g., the incomplete metalo-phid on m3). The taxonomic position of these taxa remains uncertain. I tentatively ally ?C. andersoni with prairie dogs.

?C. andersoni is clearly distinguished from other prairie dog-like taxa based on smaller size (table 17.3) and morphological details. It is distinctly more primitive, especially in the development of the metalophid on m3, than C. hibbardi and C. sappaensis, and it differs from ?S. cragini in possessing a more elongate M3 with well-developed metaloph. ?C. vetus is known only from a partial skull with extremely worn teeth, but it is clearly larger than ?C. andersoni with a less elongate M3. It is more difficult to establish phylogenetic relationships among these forms. A possible relationship among ?S. cragini, S. cochisei (a smaller taxon known from Arizona; Gazin, 1942), and ?C. vetus has been suggested (Goodwin and Hayes, 1994). In contrast to these three species, ?C. andersoni has a more elongate M3 with a better-developed metaloph and typically lacks strong buccal expansion of the parastyle on M3.


REFERRED MATERIAL Mark's Sink: L dentary with I, p4-m1 (DMNH 36933).

IDENTIFICATION AND DISCUSSION A single specimen from Mark's Sink falls at or near the upper size limit of S. cf. S. elegans in length (Lp4 = 2.21 mm; compare with table 17.2) but differs from that taxon in the much greater width of the trigonid and talonid (figure 17.2A). In addition, p4 bears a complete protolophid with only rudimentary constriction at its contact with the metaconid, a morphology I have not observed in S. elegans or related species in the subgenus S. (Sper-mophilus) but one that resembles Cynomys. Both p4 and m1

are extremely hypsodont and exhibit a distinct notch mesial to the entoconid. This combination of size and morphology is not present in any fossil or Recent species of which I am aware. I tentatively interpret the fossil to represent the smallest prairie dog yet known, but forego formal description in hopes of finding additional material in the fragmentary component of the Mark's Sink sample. Regardless, the great trigo-nid and talonid width of p4 resembles that of C. (Leuco-crossuromys) more closely than that of C. (Cynomys).

0 0

Post a comment