Figure 1.
Landmarks scored on crania and mandibles for analyses of intraspecific cranial morphological shape diversity.
Shown here is Smilodon fatalis LACMHC2001-2 and LACMHC2002-2(L2), associated skull and mandible from pit 67; scale bar equals 5 cm. Landmarks on the cranium are: 1, top of cranium at the junction of sagittal and nuchal crests; 2, top of occipital condyle; 3, dorsal extent of the mastoid musculature; 4, apex of paroccipital process; 5, apex of mastoid process; 6, centre of external auditory meatus; 7, posterior base of zygomatic arch; 8, ventral junction of jugal-squamosal suture; 9, centre of mandibular cotyle; 10, base of postorbital process (jugal portion); 11, apex of postorbital process (frontal portion); 12, centre of orbital aperture; 13, junction of jugal-maxilla suture; 14, posterior, and 15, anterior edge of P4; 15, posterior, and 16, anterior edge of P3; 17, posterior, and 18, anterior edge of C1; 19, anterior edge of premaxilla at incisor alveolus; 20, ventral edge of external narial aperture; 21, apex of nasal; 22, dorsal, and 23, ventral edge of infraorbital foramen; 24, dorsal edge of maxilla-frontal suture; 25, dorsal profile at beginning of temporal fossa. Landmarks on the mandible are: 1, apex of mandibular condyle; 2, posterior and, 3, anterior base of coronoid process; 4, apex of coronoid process; 5, posterior, and 6, anterior extent of retroarticular process; 7, anterior extent of mandibular (for m. temporalis) fossa; 8, posterior, and 10, anterior edge of M1; 10, posterior, and 12, anterior edge of P4; 14, transition of horizontal ramus to ascending portion towards symphysis; 15, posterior, and 16, anterior edge of C1 at the alveolar border; 17, ventral edge of mandibular symphysis; and the depth of the horizontal mandibular ramus posterior to M1 (8, 9), at the M1/P4 junction (10, 11); and anterior to P4 (12, 13).
Figure 2.
Size distribution of condylobasal skull lengths (CBL) in Smilodon fatalis and Panthera spp. along with the average CBL±SD and the coefficient of variation (v) for each sample.
A, Smilodon fatalis (pit 61 only); B. S. fatalis (pit 61+67); C, Panthera leo; D, P. onca; E, P. pardus; F, P. tigris; G, P. uncia. All Panthera are strongly sexually size-dimorphic and in all species the average male CBL is highly significantly larger than the average female CBL: P. leo (F = 643.485, p<0.0001); P. onca (F = 66.168, p<0.0001); P. pardus (F = 162.098, p<0.0001); P. tigris (F = 296.848, p<0.0001); and P. uncia (F = 47.124, p<0.0001). The sexual dimorphism quotient (S) and intersexual coefficient of variation for the samples are P. leo (S = 18.92; v = 9.97); P. onca (S = 13.04; v = 8.64); P. pardus (S = 17.15; v = 10.12); P. tigris (S = 15.24; v = 9.09); and P. uncia (S = 7.77; v = 4.82). In comparison the coefficient of variation for the S. fatalis samples are: pit 61 (v = 5.03); pit 67 (v = 5.03); and pit 61+67 (v = 5.21).
Figure 3.
Size distribution of mandible lengths (ML) in Smilodon fatalis and Panthera spp. along with the average ML±SD and the coefficient of variation (v) for each sample.
A, Smilodon fatalis; B, Panthera leo; C, P. onca; D, P. pardus; E, P. tigris; E, P. uncia. All Panthera are strongly sexually size-dimorphic and in all species the average male ML is highly significantly larger than the average female ML: P. leo (F = 483.091, p<0.0001); P. onca (F = 48.483, p<0.0001); P. pardus (F = 374.975, p<0.0001); P. tigris (F = 256.669, p<0.0001); and P. uncia (F = 37.904, p<0.0001). The sexual dimorphism quotient (S) and intersexual coefficient of variation for the samples are P. leo (S = 21.23; v = 11.19); P. onca (S = 13.51; v = 9.56); P. pardus (S = 19.81; v = 11.04); P. tigris (S = 16.72; v = 9.91); and P. uncia (S = 9.41; v = 6.05). In comparison the coefficient of variation for the S. fatalis samples are: pit 61 (v = 5.02); and pit 61+67 (v = 5.36).
Table 1.
Values of the moment (g) statistics in analysis of normal distribution for size-distribution of condylobasal length (CBL) and mandibular length (ML) in samples of extant Panthera and Smilodon fatalis.
Figure 4.
Plots of the first two canonical axes based on Discriminant Analysis of the Ppartial Warp scores from a Thin Plate Splines analysis on cranial shape in Panthera spp. and Smilodon fatalis.
The first canonical variable explains 66.7% of sample variation and the second canonical variable explains 18.7% of sample variation.
Table 2.
Mahalanobis distance F-matrix of species and sexes based on the Partial Warp scores from a Thin Plate Splines analysis of cranial shape in Panthera spp. and Smilodon fatalis.
Table 3.
Mahalanobis distance F-matrix of species and sexes based on the Partial Warp scores from a Thin Plate Splines analysis of mandible shape in Panthera spp. and Smilodon fatalis.
Table 4.
Post hoc classification matrix of species and sexes based on the Partial Warp scores from a Thin Plate Splines analysis of cranial shape in Panthera spp. and Smilodon fatalis.
Table 5.
Post hoc classification matrix of species and sexes based on the Partial Warp scores from a Thin Plate Splines analysis of mandible shape in Panthera spp. and Smilodon fatalis.
Figure 5.
Box-plots of statistically significant differences in cranial proportions in male (blue) and female (red) extant Panthera spp.
Values are expressed as percentages of CBL, along with the sample averages±SD, coefficients of variation (v) and the sexual dimorphism coefficient (S). The length of each box indicates the central 50% range of the values, and the box hinges denote the first and third quantiles. The whiskers indicate the range of values that fall within the inner fences, and values between the inner and outer fences are indicated with an asterisk. A, length of sagittal crest (differences between male and female samples: P. leo: F = 60.791, p<0.001; P. onca: F = 32.949, p<0.001; P. pardus: F = 175.843, p<0.001; P. tigris: F = 80.551, p<0.001); B, length of premaxilla+maxilla along alveolar series (P. leo: F = 69.761, p<0.001; P. onca: F = 8.246, p<0.001; P. pardus: F = 13.941, p<0.001; P. tigris: F = 27.230, p<0.001); C, width across postorbital constriction (P. leo: F = 58.315, p<0.001; P. onca: F = 19.669, p<0.001; P. pardus: F = 100.793, p<0.001; P. tigris: F = 89.657, p<0.001); D, palatal width across carnassial notch of P4 (P. leo: F = 92.652, p<0.001; P. onca: F = 13.554, p<0.001; P. pardus: F = 28.241, p<0.001; P. tigris: F = 11.177, p<0.001); E, width across occipital condyles (P. leo: F = 104.219, p<0.001; P. onca: F = 14.421, p<0.001; P. pardus: F = 89.402, p<0.001; P. tigris: F = 30.128, p<0.001); F, length of P4 (P. leo: F = 162.110, p<0.001; P. onca: F = 10.482, p<0.001; P. pardus: F = 46.672, p<0.001; P. tigris: F = 72.386, p<0.001).
Figure 6.
Box-plots of statistically significant differences in cranial proportions between large (black; CBL≥300.0 mm) and small (grey; CBL≤285.0 mm) Smilodon fatalis specimens from pits 61+67.
Values are expressed as percentages of CBL, along with the sample averages±SD and coefficients of variation (v). The length of each box indicates the central 50% range of the values, and the box hinges denote the first and third quantiles. The whiskers indicate the range of values that fall within the inner fences, and values between the inner and outer fences are indicated with an asterisk. A, distance from anterior rim of preglenoid process to posterior edge of occipital condyle (large-small specimen samples: combined n = 47; F = 9.742, p = 0.003); B, mastoid height (n = 48; F = 24.806, p<0.001); C, distance from anterior edge of infraorbital fenestra to tip of premaxilla (n = 48; F = 31.977, p<0.001); D, width across the incisor arcade (n = 48; F = 21.535, p<0.001); E, palatal width across carnassial notch of P4 (n = 48; F = 20.093, p<0.001); F, P3 crown length (n = 48; F = 33.740, p<0.001); G, P4 crown length (n = 48; F = 66.530, p<0.001); H, postorbital constriction width (n = 43; F = 28.322, p<0.001).
Figure 7.
Box-plots of P4 and M1 crown lengths in male (blue) and female (red) extant Panthera spp. and large (black) and small (grey) specimens of Smilodon fatalis.
Values are expressed as percentages of ML, along with the sample averages±SD, coefficients of variation (v) and the sexual dimorphism coefficient (S) for Panthera spp. The length of each box indicates the central 50% range of the values, and the box hinges denote the first and third quantiles. The whiskers indicate the range of values that fall within the inner fences, and values between the inner and outer fences are indicated with an asterisk. A, length of P4 crown (differences between male and female samples: P. leo: F = 19.162, p<0.001; P. onca: F = 8.484, p = 0.002; P. pardus: F = 33.237, p<0.001; P. tigris: F = 26.071, p<0.001); B, length of M1 crown (P. leo: F = 8.820, p<0.001; P. onca: F = 12.703, p<0.001; P. pardus: F = 9.462, p<0.001; P. tigris: F = 14.946, p<0.001). In Smilodon fatalis, the ratio variables are also highly significantly different between large vs. small specimens (P4: F = 35.363, p<0.001; M1: F = 20.327, p<0.001).
Figure 8.
Comparative morphology of some Smilodon fatalis skulls from Pits 61 and 67, all to scale.
A–D are inferred to potentially be males and E–H are inferred to potentially be females; note that D and H have complete upper canines, which have been cropped in this image. A, LACMHC2001-261 (Pit 61); B, LACMHC2001-151 (Pit 67); C, LACMHC2001-215 (Pit 61); D, LACMHC2001-2 (Pit 67); E, LACMHC2001-401 (Pit 61); F, LACMHC2001-408 (Pit 61); G, LACMHC2001-434 (Pit 61); H, LACMHC2001-231 (Pit 67). Scale bar equals 10 cm.