Statistical shape modelling as a novel reconstruction tool in palaeoanthropology: A case study on fossil pelves

Reconstructing incomplete anatomical regions of extinct taxa is critical for understanding their ecological and evolutionary context, including their migration patterns, reproductive strategies and environmental adaptations. In hominins, the pelvis can be particularly challenging to reconstruct because crushing and fragmentation during fossilization affects the delicate pubis and the sacrum. Yet the pelvis provides key information about how adaptations to different ecological niches affected functional morphology and reproduction. Here, we introduce a novel method using statistical shape modelling (SSM) to predict missing skeletal morphology from partially preserved hominin pelvic remains. We developed three statistical shape models from different phylogenetic groupings of a dataset of 290 right hipbones comprising 100 Homo sapiens and 190 extant great apes (90 Pan troglodytes , 10 P. paniscus , 57 Gorilla gorilla , 33 G. beringei ) and cross‐validated each in extinct and extant hominins ( Australopithecus afarensis , A. africanus , A. sediba , Neanderthals and Pleistocene humans). We found that models based on more diverse reference samples sometimes produced more accurate predictions. Pooling extant humans, chimpanzees and gorillas (SSM3) improved predictions for two of the fossil specimens compared to models with more restricted taxonomic composition. The human–chimpanzee model (SSM2) best predicted Ohalo 2, MH2 and Sts 14. All three models yielded high prediction error for Kebara 2, failing to predict the length and anterior orientation of its pubis. Our findings establish SSM as a robust tool for reconstructing missing morphologies in extinct hominin pelves, while highlighting the limitations of this method when the quality of the predictor landmarks is insufficient. We provide a methodological framework for addressing incomplete skeletal morphology in evolutionary anthropology and related disciplines, contributing to the broader understanding of pelvic adaptation in hominin evolution and of the interplay between morphology and ecological adaptation.