Subgenus Leucocrossuromys Hollister

CYNOMYS CF. C. LEUCURUS MERRIAM, 1890; FIGURE 17.1H

REFERRED MATERIAL See appendix 17.1. NISP=427, MNI = 83.

IDENTIFICATION AND DESCRIPTION Several Porcupine Cave sites (Pit, VR-DMNH, VR-CM, MS, and WH) include fossils of undoubted prairie dogs (Cynomys). Characters supporting this assignment include great hypsodonty (for sciurids), large size, the presence of an accessory lophule along the buc-codistal margin of the protoloph on M3, a complete protolophid on p4, a complete metalophid on m3 (figure 17.1H), and the presence of deep trigonid trenches lingual to the ec-tolophid on m1-m3 (especially m3). All diagnostic specimens from these sites represent white-tailed prairie dogs, C. (Leuco-crossuromys), based on the following characters (Goodwin, 1995b): anterior cingular ridge on P3 terminates along the mesiobuccal face of the oblique loph in a distinct notch; presence of a prominent bridge connecting the ectolophid and talonid platform on most specimens of m3 (figure 17.1H); and hypoconid of m3 deflected anteriorly less strongly than is typical in black-tailed prairie dogs (figure 17.1H). Other typical "white-tailed" characters—notably the less buccal expansion of the hypoconid on p4 and the greater trigonid width of p4-m2 than in black-tails—are more variably developed in the Porcupine Cave assemblage.

Site samples of white-tailed prairie dogs vary morpho-metrically (e.g., figure 17.2A) and qualitatively. In contrast to fossils from relatively old sites (Pit, MS, WH), the VR-DMNH prairie dogs (mostly from horizon B) show less extensive and less frequent anterior deflection of the hypoconid on m3 (figure 17.5A), exhibit a significantly shorter and narrower p4 (across the talonid), and display wider m1-m2 (table 17.4). In degree of anterior deflection of the hypoconid on m3, the two Porcupine Cave samples form a series linking C. gunnisoni (displaying the primitive morphotype for white-tailed prairie dogs; Pizzimenti, 1975; Goodwin, 1995b) with C. leucurus (a member of the crown group of the subgenus). Both samples more closely resemble C. leucurus than small C. gunnisoni on most univariate comparisons (table 17.4). A multivariate discriminant function model, built to discriminate C. leucurus from C. gunnisoni on measures of p4-m2, likewise classifies most fossils in both samples with C. leucurus (figure 17.5B). This model correctly classified 90% of modern specimens.

TABLE 17.4

Comparison of Four Samples of Cynomys on Lower Dental Variables

Velvet Room (DMNH) "Old" Porcupine Cave

C. gunnisoni C. leucurus C. (Leucocrossuromys) C. (Leucocrossuromys)

Minimum-

Minimum-

Minimum-

Minimum-

Mean

Maximum

Mean

Maximum

Mean

Maximum

Mean

Maximum

Variable

(SD)

(Sample Size)

(SD)

(Sample Size)

(SD)

(Sample Size)

(SD)

(Sample Size)

Lp4

2.80 (0.10)

2.63-2.95 (26)

3.04 (0.12)

2.85-3.18 (15)«

2.92 (0.15)

2.55-3.10 (16)

3.13 (0.12)

2.93-3.43 (19)'

WTrp4

3.87 (0.18)

3.50-4.15 (26)

4.08 (0.13)

3.88-4.33 (15)'

4.06 (0.18)

3.65-4.35 (14)'

4.11 (0.23)

3.80-4.63 (19)'

WTlp4

3.58 (0.22)

3.13-3.98 (26)'

3.87 (0.16)

3.58-4.18 (15)*t

3.69 (0.22)

3.38-4.16 (14)'*

3.96 (0.28)

3.33-4.36 (19)t

Lml

2.72(0.13)

2.53-3.00 (26)'

2.80 (0.10)

2.68-2.95 (15)'*

2.85 (0.12)

2.65-3.05 (23)*t

2.97 (0.17)

2.80-3.28 (8)t

Wml

4.27 (0.15)

3.93-4.56 (26)

4.51 (0.16)

4.13-4.78 (15)'

4.78 (0.20)

4.28-5.16 (22)

4.57 (0.14)

4.26-4.68 (7)'

Lm2

2.94(0.11)

2.75-3.15 (26)'

3.03 (0.10)

2.88-3.21 (15)'*

3.02 (0.12)

2.80-3.20 (25)'*

3.12(0.19)

2.90-3.53 (10)*

Wm2

4.48(0.17)

4.15-4.78 (26)

4.84(0.23)

4.30-5.18 (15)'*

4.93 (0.23)

4.43-5.36(25)'

4.71 (0.13)

4.55-4.98 (10)*

Lm3

4.54(0.21)

4.21-5.03 (26)'

4.70 (0.15)

4.35-5.01 (15)'*

4.64 (0.19)

4.18-4.88 (21)'*

4.76 (0.24)

4.20-4.76 (26)*

Wm3

4.31 (0.18)

3.93-4.68 (26)

4.83 (0.20)

4.26-5.08 (15)*

4.67 (0.20)

4.33-5.06 (21)'

4.64 (0.21)

4.10-4.93 (26)'*

NOTES: Significant differences were found for all variables (ANOVA). Cells sharing a symbol (V or '■') are not significantly different (Tukey's post-hoc test). Variable abbreviations indicate length (L) of p4, length (L) and width (W) of ml-m3, and trigonid (WTr) and talonid (WT1) widths of p4.

NOTES: Significant differences were found for all variables (ANOVA). Cells sharing a symbol (V or '■') are not significantly different (Tukey's post-hoc test). Variable abbreviations indicate length (L) of p4, length (L) and width (W) of ml-m3, and trigonid (WTr) and talonid (WT1) widths of p4.

FIGURE 1 7.5 (A) Relative frequency of three classes of anterior deflection of hypoconid on m3 among four samples of prairie dogs: modern Cynomys gunnisoni, "Old" Cynomys from Porcupine Cave (Pit, WH, MS specimens), "Young" Cynomys from Porcupine Cave (VR-DMNH sample), and extant C. leucurus. (B) Discrimination of modern C. gunnisoni and C. leucurus based on discriminant function model using variables of p4-m2. Specimens left of the vertical line are classified by the model as C. leucurus; those right of the line are classified as C. gunnisoni. Fossils of "Old" and "Young" Porcupine Cave Cynomys are depicted as dots and are classified by the discriminant model based on their position relative to the vertical line.

FIGURE 1 7.5 (A) Relative frequency of three classes of anterior deflection of hypoconid on m3 among four samples of prairie dogs: modern Cynomys gunnisoni, "Old" Cynomys from Porcupine Cave (Pit, WH, MS specimens), "Young" Cynomys from Porcupine Cave (VR-DMNH sample), and extant C. leucurus. (B) Discrimination of modern C. gunnisoni and C. leucurus based on discriminant function model using variables of p4-m2. Specimens left of the vertical line are classified by the model as C. leucurus; those right of the line are classified as C. gunnisoni. Fossils of "Old" and "Young" Porcupine Cave Cynomys are depicted as dots and are classified by the discriminant model based on their position relative to the vertical line.

Fossils also cluster with derived white-tails in typically displaying a bridge connecting the ectolophid with the talonid platform of m3, a feature less consistently developed in C. gunnisoni. A few fossil specimens lack this bridge but are otherwise indistinguishable from typical specimens.

These patterns of variation indicate the presence of an evolving lineage probably related to extant Cynomys leucurus, with dental morphology becoming more "modern" up through the sequence. Work is ongoing to determine if variation in the fossil sequence warrants subdivision at the species level.

DISCUSSION The systematic interpretation offered in the present study does not recognize black-tailed prairie dogs in the Pit sequence, as opposed to previous studies (Barnosky et al., 1996; Rouse, 1997). Paleoclimatic inferences hinging on co-occurrence of both subgenera may require revision (Barnosky et al., 1996; Rouse, 1997).

Occurrence of C. cf. C. leucurus in the Porcupine Cave assemblage suggests the presence of that lineage since the Irving-tonian. This conclusion falsifies a phylogenetic hypothesis that treats C. niobrarius as ancestral to C. leucurus and C. par-videns, with the latter species pair originating during range restriction and fragmentation at the end of the Pleistocene

(Goodwin, 1995b). C. niobrarius was significantly larger in size than any other white-tail (Goodwin, 1995b). It was abundant and widespread across the northern and central Plains at least from the late Irvingtonian, extending west onto the Snake River Plain at times during this interval (Goodwin, 1995a, 1995b). The Porcupine Cave record indicates that C. leucurus was probably likewise present from the middle Irvingtonian, living in the high-elevation habitats of the Rocky Mountains.

C. leucurus is not present today in the vicinity of Porcupine Cave but occurs some 80 km to the northwest (Armstrong, 1972). C. gunnisoni is regionally present today, but was evidently absent within the sampling range of the cave during fossil accumulation. As reviewed by Goodwin (1995b), C. gunnisoni was identified from the Hansen's Bluff local fauna, a middle Irvingtonian site located in south central Colorado (Rogers et al., 1985).

0 0

Post a comment