Fig 1.
Trabecular and bilateral loops.
a) Front and b) oblique view of the dorsal valve of Terebratalia transversa (SBMNH 616990). c) Front and d) oblique view of dorsal valve of Laqueus vancouveriensis (USNM PAL 716055).
Fig 2.
For more detailed information see S1 Table.
Table 1.
List of species analyzed with information on loop type, geographic distribution, and number is individuals analyzed.
Fig 3.
Landmark scheme based on different loop types.
a) Bilateral loop with 20 landmarks and b) three curves based on semilandmarks (I-III). c) Trabecular loop with 15 landmarks and d) three curves based on semilandmarks (I-III). Landmarks in pink (1–15) represent shared landmarks between bilateral and trabecular loops; landmarks in green (16–20) represent unique landmarks found in the bilateral loop.
Fig 4.
Sampling curve from landmark sampling evaluation test (LaSEC) [74] on Laqueus, Terebratalia, and Dallinella dataset (15 landmarks and 69 semilandmarks).
A fit value of 1 indicates that no morphometric information is added when landmark number increases, providing an indication of the minimum meaningful number of landmarks. Gray lines represent iterations of subsampling; black line shows median fit corresponding to every landmark; vertical dashed black lines correspond to number of landmarks at fit values of 0.90, 0.95 and 0.99.
Fig 5.
Principal component analysis (PCA) and canonical variate analysis (CVA) of Laqueus species with associated shape changes in loop morphology.
a) PC1-PC2; b) PC1-PC3; c) CVA; d) specimens with the most negative to most positive PC score values. PC1: L. rubellus USNM PAL 7160785 (PC1-) and L. vancouveriensis USNM PAL 716055 (PC1+). PC2: L. quadratus USNM PAL 716076 (PC2-) and L. erythraeus DAV:SJCLab 0008 (PC2+). PC3: L. vancouveriensis USNM 716058 (PC3-) and L. erythraeus DAV:SJCLab 0007 (PC3+).
Table 2.
Mahalanobis distances between species means in Laqueus (bold) and cross-validated classification results in percentages.
Overall classification accuracy of 100%.
Fig 6.
Principal component analysis (PCA) and canonical variate analysis (CVA) of Dallinella and Terebratalia species with associated shape changes in loop morphology.
a) PC1-PC2; b) PC1-PC3; c) CVA; d) specimens with most negative to most positive PC score values. PC1: D. occidentalis SBMNH 467448 (PC1-) and T. transversa SBMNH 616986 (PC1+). PC2: T. coreanica USNM 716054 (PC2-) and T. transversa DAV:SJCLab 0011 (PC2+). PC3: T. transversa SBMNH 616990 (PC3-) and T. coreanica USNM 716052 (PC3+).
Table 3.
Mahalanobis distances between species means in Dallinella and Terebratalia (bold) and cross-validated classification results in percentages.
Overall classification accuracy of 72.22%.
Fig 7.
Principal component analysis (PCA) and canonical variate analyses (CVA) of a combined dataset (Laqueus, Terebratalia, and Dallinella).
a) PC1-PC2; b) PC1-PC3; c) CVA of species; d) CVA of genera.
Table 4.
Mahalanobis distances among species means for Dallinella, Laqueus, and Terebratalia (bold) and cross-validated classification results in percentages.
Overall classification accuracy of 100%.
Table 5.
Mahalanobis distances among genera means for Dallinella, Laqueus, and Terebratalia (bold) and cross-validated classification results in percentages.
Overall classification accuracy of 100%.