Pangolin Scale Forensics

All eight pangolin species have been listed on CITES Appendix I since 2016, prohibiting commercial international trade. The four Asian species (Sunda pangolin, Manis javanica; Chinese pangolin, M. pentadactyla; Indian pangolin, M. crassicaudata; Philippine pangolin, M. culionensis) have been commercially exploited far longer and have smaller wild populations than their African counterparts. The four African species (white-bellied, Phataginus tricuspis; black-bellied, P. tetradactyla; giant ground, Smutsia gigantea; Temminck's ground, S. temminckii) are now the dominant source in transnational shipments.

Species identity matters forensically for two reasons. First, sentencing guidelines and conservation impact assessments depend on which populations were affected; a container of critically endangered Chinese pangolin scales is treated differently from the same weight of white-bellied scales in some jurisdictions. Second, geographic origin assignments from DNA can link a specific seizure to a known poaching hotspot, connecting traffickers to source-region criminal networks.

Pangolin scales are built from cornified keratinocytes layered into a dense, overlapping structure. On the outer (dorsal) surface, longitudinal ridges run toward the scale tip. The number and spacing of these ridges, the shape of the tip, and the cross-sectional profile from base to apex all vary in ways that correlate with species. Light microscopy of scale cross-sections shows differences in the thickness ratio between the dense outer cortex and the softer inner spongy core, another character that differs between African and Asian species groups.

Scanning electron microscopy (SEM) adds surface texture at higher resolution, revealing micro-ridge density and the pattern of scale-surface pores. Published reference atlases (notably work from TRAFFIC and from the Wildlife Forensics Network) document these features per species, allowing comparison with unknown samples. Morphology alone produces correct species assignments at high rates for species with distinctive scale profiles, but the Sunda pangolin and Philippine pangolin are closely similar and frequently require molecular confirmation.

Keratin is one of the most durable biological materials. It survives heat, UV exposure, and years of storage in conditions that degrade DNA completely. Liquid chromatography-mass spectrometry exploits this durability. The workflow extracts total protein from a small amount of scale material (typically a few milligrams of powder), performs enzymatic digestion with trypsin to produce a reproducible set of peptide fragments, and then runs the peptide mixture through an LC-MS instrument.

1. Sample preparation
   Scale material is ground, then keratin is extracted using a reducing buffer that breaks disulfide bonds. The extract is cleaned and concentrated.
2. Tryptic digestion
   Trypsin cleaves the protein at predictable sites (lysine and arginine residues), producing a characteristic set of peptide fragments whose masses can be predicted from the gene sequence.
3. LC-MS/MS analysis
   Peptides are separated by liquid chromatography and fragmented in the mass spectrometer. Fragment masses are compared against species-specific keratin peptide databases to assign species identity.
4. Species assignment
   Diagnostic peptides present in one species but absent in others are used as markers. A match above a threshold score supports species identification; a mixed-species result indicates more than one species in the sample.

Studies published by researchers including those affiliated with CITES-supported programs have validated LC-MS against known-species reference scales and achieved species-level discrimination across all eight pangolin species. The method is particularly valuable for old stockpile material, where DNA is typically non-amplifiable but keratin protein peptides remain intact enough to generate confident spectral matches.

Where DNA can be extracted from scales (typically from the scale base where residual dermal tissue persists, or from the spongy core), mitochondrial markers such as cytochrome b and the D-loop provide reliable species identification. For common species with well-populated reference databases such as M. javanica and P. tricuspis, BLAST queries against GenBank or the Barcode of Life Data System (BOLD) return clear top-hit matches. For rarer species with sparse database coverage, phylogenetic tree placement alongside reference sequences is more rigorous than a BLAST top-hit alone.

Population-level assignment requires a denser dataset. Microsatellite loci and mitochondrial haplotype networks have been used to separate West African and Central African populations of white-bellied pangolins, allowing investigators to ask whether a particular seizure came from a specific national poaching corridor. This geographic forensics work is still developing but has contributed evidence in prosecutions by showing that scales claimed to be from a legal stockpile in one country share haplotypes exclusively with animals from a different country's wild population.

Courts and conservation organizations both want to know how many pangolins died to produce a given seizure. Because scales from many individuals are mixed together, you cannot simply count intact carcasses. The standard approach uses the minimum number of individuals, a concept borrowed from zooarchaeology.

The calculation requires two inputs: the total number of scales in the seizure (or, where scales are uncountable, an estimate from weight and average scale mass) and the average number of scales per living individual for the identified species. For M. javanica the figure is approximately 900-1,000 scales per adult; for P. tricuspis it is in the range of 400-600. Dividing total count by species mean gives MNI. The result is a floor, not a true count: where small juvenile scales are absent, the seizure likely included smaller animals whose scales were not tallied, and MNI understates true mortality.

The TRAFFIC pangolin seizure database compiles MNI estimates from hundreds of recorded seizures, providing trend data on annual kill rates and shifts in source species as Asian populations decline. This database is used by researchers to model population-level impacts and by law enforcement to connect shipments across seizure events by comparing species composition, haplotypes, and scale morphology.

A recurring legal argument in pangolin cases is that seized scales come from lawful stockpiles accumulated before the CITES Appendix I uplisting in 2016, or in some jurisdictions before earlier national trade bans. Distinguishing recently harvested material from genuinely old stockpile scales matters enormously for the outcome of a prosecution.

• Radiocarbon dating (AMS): accelerator mass spectrometry of the organic keratin matrix can bracket the decade of harvest using the post-bomb radiocarbon curve. Material from the 2000s onward has a distinctive radiocarbon signature that differs from pre-1960 material. Precision is roughly plus or minus five years depending on calibration curve position.
• DNA degradation profile: DNA yield and fragment-size distribution degrade with age and temperature exposure. Very old material typically yields no amplifiable DNA at all; fresh material yields high-molecular-weight fragments. This is a relative indicator, not an absolute date.
• UV-induced discoloration and cracking: scale surfaces yellow and develop micro-fractures after extended UV and air exposure. Scales stored in controlled, dark conditions age more slowly, complicating this visual criterion, but heavily yellowed, brittle scales are unlikely to be recent.
• Isotopic provenance: stable isotope ratios (carbon, nitrogen, strontium) in scale keratin reflect the animal's diet and local geology, providing geographic origin signals that can conflict with claimed provenance documentation.