Virtual Geometric Reconstruction of Australopithecine Femora: A Methodological Evaluation

Abstract

The fragmentary nature of fossils, and the need to extract as much information from them as possible, mandates that broken portions be reconstructed. Virtual reconstruction using geometric morphometric techniques based on the thin plate spline (TPS) function can be used to locate landmarks that represent missing portions of fossil specimens. Here we evaluate the accuracy of TPS‐based reconstruction in artificially damaged australopithecine distal femora.Surface models of modern human (n=35) and chimpanzee (n=37) distal femora, as well as those of two australopiths (casts of A.L. 129‐1a and A.L. 333‐4), were created using a structured light scanner. Articular surfaces were trimmed from the rest of the surface model. 1007 sliding semilandmarks were distributed across the articular surface, and semilandmarks were then slid to minimize the bending energy of the TPS function relative to an updated Procrustes average. Fossils were then virtually destroyed by removing semilandmarks to mimic the type of destruction observed in other fossil specimens. Removed portions increased from a small, central hole in the middle of each condyle to removal of the entire condyle. Average chimpanzee and human femora were used as references to create two separate reconstructions using geometric reconstruction based on the TPS function. Procrustes distances (PD) between each reconstruction and the original fossil were calculated to evaluate reconstruction accuracy.Reconstructions based on both human and chimpanzee averages deviated further from the original fossils as the level of artificial damage increased. When the entire medial condyle was reconstructed, the original morphology of both australopithecine fossils was intermediate between the human‐based and chimpanzee‐based reconstructions. For these reconstructions, Procrustes distances indicate that both A.L. 129‐1a and A.L. 333‐4 are most similar to the human‐based reconstruction (A.L. 129‐1a: PD = 0.001 vs. 0.002; A.L. 333‐4: PD = 0.002 vs. 0.003). When the entire lateral condyle was reconstructed, the fossils fell outside the possible morphologies bounded by the human‐based and chimpanzee‐based reconstructions. Procrustes distances indicate that both fossils are more similar to the human‐based reconstructions than those based on the average chimpanzee (A.L. 129‐1a: PD = 0.004 vs. 0.005; A.L. 333‐4: PD = 0.005 vs. 0.009). These results suggest that while the australopithecine medial femoral condyle may be considered an interpolation between chimpanzee and human morphology, the lateral condyle does not follow the same trend. For australopithecine distal femur reconstructions to be accurate, medial and lateral condyles should be considered separately.This abstract is from the Experimental Biology 2019 Meeting. There is no full text article associated with this abstract published in The FASEB Journal.