Figure 4 in Mosaic patterns of homoplasy accompany the parallel evolution of suspensory adaptations in the forelimb of tree sloths (Folivora: Xenarthra)

Figure 4. Geometric morphometric principal components analysis of the ulna, humerus and scapula of xenarthrans. A, the first two axes of the ulna GM analysis. No other principal components (PC) axis accounts for more than 5% of variation. B, the first two axes of the humerus GM analysis. The small amount of variation accounted for by PC2 is likely due to the small sample sizes for the taxa it differentiates, specifically armadillos, the two giant ground sloths, and to a lesser extent Hapalops. PC3 accounts for 5.6% of variation and differentiates Cyclopes from Paramylodon and Glossotherium. No other axes account for more than 5% of variation. C, the first two PCs of the scapula GM analysis. PCs 3 and 4 account for 8.6% and 6.3% of variation, respectively. PC 3 separates Cyclopes from other taxa, and PC 4 separates Cyclopes and Dasypus from Choloepus. No other PC accounts for more than 5% of variation. Sloth scapula specimens identified with a thick rimmed circle and black dot indicate the specimens shown in 4D. D, Choloepus (centre) has a relatively conserved gross scapular morphology (compare with Paramylodon on right), especially when compared with Bradypus (left), but it has mapped functional traits such as an angled scapular spine onto that conserved bauplan. Squares indicate armadillos, rounded squares indicate anteaters, and circles indicate sloths.