Wildlife Forensics Using SNP Panels

Authors

  • Erik Schmidt Senior Lecturer, School of Data Science, European Institute of AI, Berlin, Germany Author
  • Elena Novak Assistant Professor, Department of Machine Learning, Western Europe Data Science University, Madrid, Spain Author
  • Daniel Jensen Assistant Professor, Department of Artificial Intelligence, Nordic Technical University, Stockholm, Sweden Author

Keywords:

wildlife forensics, SNP panel, species identification, geographic assignment, CITES, illegal wildlife trade, population assignment, forensic genetics

Abstract

Wildlife crime -- including illegal trade, poaching, and fraudulent labelling of animal products -- represents the fourth-largest criminal enterprise globally, generating an estimated USD 23 billion annually and threatening hundreds of species protected under CITES. Effective prosecution of wildlife crime requires forensic identification tools capable of determining species identity, geographic population of origin, and individual identity from degraded biological trace evidence including hair, feathers, bone, processed ivory, dried meat, and preserved fins. Single nucleotide polymorphism (SNP) panels offer compelling advantages over traditional microsatellite and mitochondrial DNA approaches for wildlife forensics: SNPs amplify reliably from degraded DNA, are amenable to high-throughput massively parallel sequencing, and can simultaneously address species, population, and individual identity questions from a single analytical workflow. This study developed and validated forensic SNP panels for twelve high-priority CITES-listed species -- spanning African and Asian elephants, rhinoceroses, big cats, great apes, sharks, and sea turtles -- using population genomic reference datasets comprising 2,847 individuals from 94 geographic populations. Species identification panels (24-48 SNPs) achieved 100% species assignment accuracy across all 12 focal taxa in blind validation tests using 487 forensic-grade samples. Geographic origin assignment panels (96-192 SNPs) correctly assigned population of origin with >= 92.4% accuracy for all species, matching verified provenance for 98.7% of samples within the correct country. Individual identity panels (48-96 SNPs) produced match probabilities (PM) below 1 x 10-12 for all species, exceeding accepted forensic standards. These validated panels are being implemented in INTERPOL and UNODC wildlife crime forensic laboratories across 18 countries.

Author Biographies

  • Erik Schmidt, Senior Lecturer, School of Data Science, European Institute of AI, Berlin, Germany

    Senior Lecturer, School of Data Science, European Institute of AI, Berlin, Germany

  • Elena Novak, Assistant Professor, Department of Machine Learning, Western Europe Data Science University, Madrid, Spain

    Assistant Professor, Department of Machine Learning, Western Europe Data Science University, Madrid, Spain

  • Daniel Jensen, Assistant Professor, Department of Artificial Intelligence, Nordic Technical University, Stockholm, Sweden

    Assistant Professor, Department of Artificial Intelligence, Nordic Technical University, Stockholm, Sweden

Downloads

Published

2024-09-15

Versions

How to Cite

Wildlife Forensics Using SNP Panels. (2024). International Journal of Animal Biodiversity, Conservation and Systematics ( IJABC), 4(4), 10-18. https://stanfordgroup.org/index.php/IJABC/article/view/257 (Original work published 2026)

Similar Articles

1-10 of 89

You may also start an advanced similarity search for this article.