Home

Article Search

TABLE 1. List and definitions of landmarks (1-8, 22-23, 47) and semilandmarks (9-21, 24-46) used in the landmark-based geometric morphometric analysis. See also Figure 2B.

1 Labial end of preparacrista (point of contact between preparacrista and ectocingulum)
2 Apex of paracone (center of metacone in all taxa in which this cusp is conical)
3 Midpoint between paracone and metacone (measured as the location of the centrocrista where these cusps meet)
4 Apex of metacone (center of metacone in all taxa for which this cusp is conical)
5 Anterior end of postmetacrista (defined by point of contact between base of metacone and posterolingual edge of tooth)
6 Apex of protocone
7 Labialmost point of talon (measured at between paracone, metacone, and talon)
8 Posteriormost point of talon (measured at point where talon protrudes lingually from trigon)
9-21 Shape of talon from posteriormost to anteriormost part of the talon (semilandmark curve between points 8 and 22)
22 Anteriormost point of talon (measured at point where talon protrudes lingually from trigon)
23 Distal end of postmetacrista
24-46 Shape of stylar shelf (including shapes of para- and metastylar lobes and dimensions of ectoflexus)
47 Anterior end of stylar shelf (typically anterior end of parastylar corner or where tooth contacts preceding tooth)

 

 

TABLE 2. Mahalanobis distances for each discriminant analysis. For all permutation analyses p-values were < 0.001.

  Trigon & Talon Trigon Only
M2 M3 M2 M3
All Taxa M1 2.06 3.55 1.59 3.03
M2 - 2.26 - 1.79
M3 - - - -
No Tiupampa taxa, borhyaenids, or thylacosmilines M1 2.6 4.47 2.5 4.18
M2 - 2.38 - 2.25
M3 - - - -
No borhyaenids or thylacosmilines M1 2.41 4.34 2.27 4.19
M2 - 2.45 - 2.32
M3 - - - -
No borhyaenids M1 2.31 4.32 2.08 4.09
M2 - 2.53 - 2.29
M3 - - - -
Allometric Signal Removed M1 1.87 2.96 1.44 2.75
M2 - 1.98 - 1.54
M3 - - - -

 

TABLE 3. Accuracy (% correct classification) of all CVAs evaluated using cross-validation (“leave one out”) analyses of the training dataset.

Analysis # PCs M1 M2 M3 Total
Trigon + Talon All Taxa 20 66.67 76.92 85.71 77.19
No Tiupampa taxa, borhyaenids, or thylacosmilines 13 80.00 79.31 84.35 81.39
No borhyaenids or thylacosmilines 17 78.57 75.00 85.29 79.78
No borhyaenids 17 70.00 82.35 94.44 83.00
All taxa, allometry controlled 19 66.67 61.54 78.57 69.30
Trigon Only All Taxa 12 73.52 69.05 74.42 72.27
No Tiupampa taxa, borhyaenids, or thylacosmilines 11 76.00 74.19 87.50 79.54
No borhyaenids or thylacosmilines 14 75.00 67.65 85.29 76.04
No borhyaenids 17 70.00 75.00 83.33 76.47
All taxa, allometry controlled 6 75.53 59.52 74.42 68.91

 

 

TABLE 4. Results of the CVA applied to isolated teeth of unknown locus, with predicted locus and posterior probability (PP) for each specimen in each analysis.

    AMNH 28413 AMNH 28415 AMNH 28416 AMNH 28439 MPEF-PV 4345
Analysis Locus PP Locus PP Locus PP Locus PP Locus PP
Trigon + Talon All Taxa - - M1 0.902 - - M3 0.915 - -
No Tiupampa taxa, borhyaenids, or thylacosmilines - - M1 0.996 - - M3 0.962 - -
No borhyaenids or thylacosmilines - - M1 0.982 - - M3 0.971 - -
No borhyaenids - - M1 0.995 - - M3 0.981 - -
All taxa, allometry controlled - - M1 0.582 - - M3 0.870 - -
Trigon Only All Taxa M2 0.836 M1 0.786 M2 0.746 M3 0.974 M3 0.650
No Tiupampa taxa, borhyaenids, or thylacosmilines M2 0.901 M1 0.996 M2 0.908 M3 0.997 M2 0.822
No borhyaenids or thylacosmilines M2 0.971 M1 0.992 M2 0.953 M3 0.982 M2 0.883
No borhyaenids M2 0.877 M1 0.981 M2 0.926 M3 0.988 M2 0.814
All taxa, allometry controlled M2 0.634 M1 0.517 M3 0.738 M3 0.954 M2 0.828

logo smallPalaeontologia Electronica
Webmaster
1998–2022
25 years of electronic palaeontology

PE is archived by Internet Archive.