Abstract
Reconstructing weaning practices in ancient populations is crucial for understanding nutritional strategies and infant care. However, the precise identification of the transition from breastfeeding to solid foods remains a methodological challenge. This study compares two different approaches applied to odontoskeletal remains from the Picene necropolises of Matelica (8th-6th century BCE, central Italy): bulk δ¹⁵N isotopic analysis of 44 individuals elaborated through the WARN model (Weaning Age Reconstruction with Nitrogen isotopes) and histology driven geochemistry of dental enamel by LA-ICP-MS (Laser Ablation Inductively Coupled Plasma Mass Spectrometry). The WARN analysis, performed on 44 non-adult individuals, was based on an approximate Bayesian computational model proposed in 2013 by Tsutaya & Yoneda. The results indicate an onset of weaning at approximately 3.6 months (t1 = 0.3 years) and a complete transition to solid foods at 24 months (t2 = 2.0 years). The nitrogen enrichment value (2.9‰) and the weaning diet δ¹⁵N (8.5‰) suggest a post-weaning diet rich in animal proteins or other nitrogen-enriched foods. In parallel, five thin sections of deciduous molars from five different individuals were analysed for Trace Elements (TE) concentration variation with particular attention to the key markers of weaning (Sr and Ba). An increase in Ba/Ca and Sr/Ca ratios between 100 and 150 days (3-5 months) suggests an early introduction of solid foods. The comparison between WARN model and dental geochemistry shows good agreement in the timing of weaning initiation, while highlighting potential variability in the full transition to solid foods. The integration of these two approaches provides a more comprehensive understanding of infant feeding strategies in past societies.