TY - JOUR
T1 - Towards understanding isotope variability in elephant ivory to establish isotopic profiling and source-area determination
AU - Ziegler, Stefan
AU - Merker, Stefan
AU - Streit, Bruno
AU - Boner, Markus
AU - Jacob, Dorrit E.
PY - 2016/5/1
Y1 - 2016/5/1
N2 - We present here new isotopic data (δ13C, δ15N, δ18O, δ2H, and δ34S) from pulverised ivory powder, measured by continuous flow isotope ratio mass spectrometry from an unprecedented large dataset of 507 ivory samples, derived from twenty-eight African and six Asian elephant range states. The aim of this study is to assess the accuracy of isotopic fingerprinting and to evaluate its forensic potential and limitations to predict the provenance of ivory of unknown origin. We constructed a nominal assignment framework for the African reference samples, consisting of 208 different sites and applied the weighted k-Nearest Neighbor Classifier with reference site as classifier and inferred the accuracy of the assignments of samples from the African elephant species to their correct provenance. Our results show that isotopic profiling of African elephant ivory works on regional scales and we were able to assign 50% of all samples within 381 km, and the majority of the remaining samples within 1154 km. Source area determination is hampered by the fact that within-site and within-individual variation in ivory is immense because elephants as ecological generalists use a wide diversity of plant resources. We propose that forest elephant diets differ more between individuals (i.e. dietary niche partitioning is more significant) than in savanna elephants where individual diets overlap more. Increasing sampling effort in order to decrease median distance of the nominal assignment framework and to better understand within-site variance of the studied isotopic systems are imperative to establish isotopic profiling in the context of law enforcement and wildlife forensics.
AB - We present here new isotopic data (δ13C, δ15N, δ18O, δ2H, and δ34S) from pulverised ivory powder, measured by continuous flow isotope ratio mass spectrometry from an unprecedented large dataset of 507 ivory samples, derived from twenty-eight African and six Asian elephant range states. The aim of this study is to assess the accuracy of isotopic fingerprinting and to evaluate its forensic potential and limitations to predict the provenance of ivory of unknown origin. We constructed a nominal assignment framework for the African reference samples, consisting of 208 different sites and applied the weighted k-Nearest Neighbor Classifier with reference site as classifier and inferred the accuracy of the assignments of samples from the African elephant species to their correct provenance. Our results show that isotopic profiling of African elephant ivory works on regional scales and we were able to assign 50% of all samples within 381 km, and the majority of the remaining samples within 1154 km. Source area determination is hampered by the fact that within-site and within-individual variation in ivory is immense because elephants as ecological generalists use a wide diversity of plant resources. We propose that forest elephant diets differ more between individuals (i.e. dietary niche partitioning is more significant) than in savanna elephants where individual diets overlap more. Increasing sampling effort in order to decrease median distance of the nominal assignment framework and to better understand within-site variance of the studied isotopic systems are imperative to establish isotopic profiling in the context of law enforcement and wildlife forensics.
KW - Elephant ivory
KW - Stable isotopes
KW - Wildlife forensics
KW - Assignment simulation
UR - http://www.scopus.com/inward/record.url?scp=84960952045&partnerID=8YFLogxK
U2 - 10.1016/j.biocon.2016.03.008
DO - 10.1016/j.biocon.2016.03.008
M3 - Article
AN - SCOPUS:84960952045
SN - 0006-3207
VL - 197
SP - 154
EP - 163
JO - Biological Conservation
JF - Biological Conservation
ER -