GoKawiilAn expanded methods section that contains site and sample details, data generation and analysis is provided in the Supplementary Information. We sampled a total of 27 canid specimens (bone or tooth) from 6 archaeological sites across western Eurasia: Gough’s Cave (n = 7), Grotta Continenza (n = 1), Padina (n = 5), Pınarbaşı (n = 6), Vlasac (n = 4) and Wezmeh Cave (n = 4). Sample size was not statistically predetermined, and analysis was conducted without blinding or randomization.
Radiocarbon dating
We generated eight new radiocarbon dates for canids from Gough’s Cave (n = 2), Pınarbaşı (n = 4) and Wezmeh Cave (n = 2) (Fig. 2b, Supplementary Fig. 3 and Supplementary Table 2). Collagen extraction and accelerator mass spectrometry dating was conducted at either the Oxford Radiocarbon Accelerator Unit (Oxford, UK) or the Higham laboratory at the Faculty of Life Sciences, University of Vienna (Vienna, Austria). Radiocarbon age estimates were calibrated using the International Calibration curve IntCal20 (ref. 69) in OxCal v.4.4 (ref. 70) and used to calculate mean age and 95% highest posterior density intervals.
Stable isotope analysis
Bulk and compound-specific isotope analyses (δ13C and δ15N) were performed on both human (n = 4) and canid (n = 8) remains from Gough’s Cave and Pınarbaşı in the University of York’s BioArCh Facilities (Supplementary Table 6). Collagen extraction followed a modified72 protocol, with lyophilized collagen samples analysed in duplicate by elemental analysis-isotope ratio mass spectrometry (IRMS). Stable carbon (δ13C) and nitrogen (δ15N) isotope ratios were calibrated relative to the internationally defined standards (Vienna Peedee belemnite scale, AIR), with standard reference materials (Sigma fish gel, IsoAnalytical Alanine, IsoAnalytical Soy) used to monitor uncertainty. δ15N and δ13C values represent the mean of three standard-corrected measurements. For compound-specific isotope analysis, collagen for each sample was hydrolysed and filtered, before amino acids were derivatized to form N-acetyl-i-propyl esters following ref. 73 (alongside international reference standards and standards). Gas chromatography-carbon-IRMS measurements of amino acids were conducted using a Delta V Plus IRMS linked to a Trace Ultra 1310 gas chromatograph, and all data authenticated using observed versus expected bulk isotope values and observed hydroxyproline:proline δ15N and δ13C amino acid ratios (Supplementary Fig. 9).
Ancient DNA data generation
Ancient DNA laboratory work was performed across dedicated clean-laboratory facilities in London, Munich (Ludwig-Maximilians-Universität; LMU) and Oxford.
Ancient DNA laboratory, Natural History Museum, London
DNA was extracted from canid specimens from Gough’s Cave (n = 7) using a modified protocol version after ref. 74, with different preprocessing steps for teeth and bone. Double-stranded, dual-indexed DNA libraries were prepared from extracts following a modified75 protocol. Amplified indexed libraries were pooled in equimolar concentration for screening on an Illumina NovaSeq X Plus platform with paired-end (2× 150 base pairs (bp)) sequencing chemistry at Novogene. For extracts with higher endogenous DNA content and complexity, we generated further USER-treated libraries and deep sequenced these across several lanes of Illumina NovaSeq X Plus (2× 150 bp paired-end) at Novogene.
Faculty of Veterinary Medicine, LMU, Munich
... continue reading