Comparing methods for estimating cranial capacity in incomplete human fossils using the Jingchuan 1 partial cranium as an example

Abstract

Cranial capacity is one of the most important features used in hominin taxonomic and morphological analyses. For complete or nearly complete modern human crania, the traditional methods of estimating cranial capacity include filling the vault with seeds, the water displacement method, and the use of regression formulae based on craniometrics. For incomplete human fossils, cranial capacities are estimated by reconstructing endocasts manually or virtually or by using existing modern human skull regression formulae; however, the accuracies of these methods are usually dubious. To find a more accurate way of estimating cranial capacity of partial skulls, seven different estimation methods are compared, including the manual reconstruction of the endocast, models built on skulls and models built on endocasts. We then estimated the cranial capacity of a fragmentary Late Pleistocene cranium, Jingchuan 1. The models are tested on 30 modern human skulls, three Homo erectus fossils and one Late Pleistocene Homo sapiens fossil. In terms of estimating the cranial capacity of the fossil humans, our results indicate that the cranial capacity estimates based on endocasts are more precise than those from exterior skull dimensions, that multivariate models are better than univariate ones, and that the new models using PCR and PLSR have the smallest errors (<50 ml). From the seven methods, the cranial capacity of Jingchuan 1 is estimated to be 1630 ml, 1505 ml, 1533 ml, 1468 ml, 1512 ml, 1470 ml, and 1457 ml, respectively. The most reliable results for the Jingchuan 1 cranial capacity are between 1470 and 1457 ml, and the average is 1464 ml. This study has direct applications to future studies of cranial capacity variation and brain evolution in fossil and modern humans.