Contents lists available at ScienceDirect Forensic Science International journal homepage: www.elsevier.com/locate/forsciint A large-scale evaluation of intraperson isotopic variation within human bone collagen and bioapatite ☆ Gregory E. Berg a, ⁎ , Lesley A. Chesson a , Jang Yuryang b , Shin Youngsoon b , Eric J. Bartelink c a Defense POW/MIA Accounting Agency (DPAA) Laboratory, Building 4077, 590 Moffett St., Joint Base Pearl Harbor-Hickam, HI 96853, United States b Ministry of National Defense Agency for KIA Recovery and Identification (MAKRI), 250, Hyeonchung-ro, Dongjak-gu, Seoul, Republic of Korea c Department of Anthropology, California State University, Chico, 400W. First Street, Chico, CA 95929-0400, United States article info Article history: Received 1 February 2022 Received in revised form 17 April 2022 Accepted 25 April 2022 Available online 28 April 2022 Keywords: Forensic anthropology Stable isotopes Intraskeletal variability Bone collagen Bone bioapatite Commingled human remains abstract This study investigated intraperson skeletal (herein referred to as either “intraperson” or “intraskeletal”) variation in stable isotope ratios for collagen (C and N) and bioapatite (C and O) extracted from five to six long bones from 27 modern individuals. The maximum intraperson variation observed for collagen was 0.78‰ for δ 13 C coll values and 1.12‰ for δ 15 N coll values, with a mean variation ( ± SD) of 0.33 ± 0.18‰ and 0.45 ± 0.27‰, respectively. For bioapatite, the maximum intraperson variation was 1.63‰ for δ 13 C ap values and 4.80‰ for δ 18 O ap values, with a mean variation ( ± SD) of 0.81 ± 0.32‰ and 1.00 ± 1.03‰, respectively. These results generally agree with previously reported data on intraskeletal isotopic variation. Using a two- and three-standard-deviations-from-the-mean model with analytical quality control data included, it is proposed that two bones with differing collagen δ 13 C coll values greater than 0.75‰ are probably from dif- ferent individuals, and those that have differing values greater than 0.95‰ are from different individuals. Likewise, differing collagen δ 15 N coll values greater than 1.05‰ are probably different, and greater than 1.35‰ are different. For bioapatite, the proposed values change to 1.55‰ and 1.90‰ for δ 13 C ap , respectively; for δ 18 O ap values no limits were set due to the unexpectedly large variation found in the study population. We highly encourage researchers to use extreme caution when interpreting δ 18 O values from bone apatite. We also note that these parameters were evaluated on modern samples and therefore may not reflect in- traperson variation in past societies. Finally, we demonstrate application of these interpretative limits to sort commingled human remains cases. Published by Elsevier B.V. 1. Introduction Isotopes in human tissues, such as bone or teeth, are tied to diet and drinking water sources [1], thus allowing investigators to use stable isotope ratio analysis to broadly predict the geographic origin of an unknown individual. In 2014, the Defense POW/MIA Ac- counting Agency (DPAA) began using isotope testing and inter- pretation within the forensic identification process for unknown human remains [2]. Broad, geographic interpretations on origin of unknown human remains via stable isotope ratio analysis from collagen can relieve pressure on other, more expensive, and intensive testing modalities. Additional positive down-stream ef- fects, such as potentially segregating commingled skeletal remains and assigning place of origin within narrower geographic regions, are theoretically possible. Addressing commingling via isotope testing is particularly im- portant as a mitochondrial DNA (mtDNA) sequence can be shared between multiple individuals. In the U.S. White population, ap- proximately 7% of individuals share the same mtDNA sequence, and the common frequency sequences steadily decline to nearly 0% thereafter [3,4]. Extrapolating this problem, it can be expected that large, commingled assemblages could easily have multiple in- dividuals that share common mtDNA types. For example, if a skeletal assemblage has a skeletal minimum number of individuals of 23, we could realistically expect that two individuals would share the most common mtDNA sequence, and another two would share the second most common, etc. These duplicated sequences in the assemblage make the problem of sorting individuals difficult and costly, ne- cessitating Y-STR DNA (yDNA) or autosomal DNA (auDNA) testing to https://doi.org/10.1016/j.forsciint.2022.111319 0379-0738/Published by Elsevier B.V. ☆ This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. ]] ]] ]]]]]] ⁎ Corresponding author. E-mail addresses: gregory.e.berg2.civ@mail.mil (G.E. Berg), lesley.a.chesson.civ@mail.mil (L.A. Chesson), ajangcho@naver.com (J. Yuryang), pragmatist_s@naver.com (S. Youngsoon), ebartelink@csuchico.edu (E.J. Bartelink). Forensic Science International 336 (2022) 111319