|
|
|
|
| |
| Location: |
Africa |
| Subject: |
Management |
| Species: |
African Rhino Species |
 |
|
In the last edition of Pachyderm, I introduced readers to the major WWF-funded project, which the AfRSG is co-ordinating, to develop a forensic test to enable law enforcement staff to source confiscated rhino horn. I outlined progress made with obtaining samples throughout the continent, and discussed the analytical techniques to be used to determine the chemical composition of horn. The need for the development of horn fingerprinting was again highlighted by a member of the Endangered Species Protection Unit of the South African Police Service (ESPU) who indicated the Unit did not know the source of many of the horns it had recovered in illegal busts.
I am pleased to report that initial pilot statistical analyses of the raw chemistry data by the AfRSG's Scientific Officer have been very promising.
The AfRSG's decision to use Laser-Ablation-Inductively-Coupled-Plasma-Mass-Spectrometry (LA-ICP-MS) to quantify the chemical composition of horn (in this case the abundance of heavier isotopes) has been vindicated. Plot analyses show this technique is producing data which better discriminate between areas than any other analytical technique which has been used before. On a pilot sample of 67 black rhino horns, 92.5% of these samples were correctly sourced using only LA-ICP-MS data. The data produced excellent separation between most areas.
The use of Inductively-Coupled-Plasma-Optical-Emission-Spectrometry (ICP-OES) to quantify trace elements was also supported by the pilot analyses. Although not as good as LA-ICP-MS, ICP-OES data still had significant discriminatory power, with the source of 76.1 % of the pilot sample being correctly predicted using only ICP-OES data.
The analysis of horn samples for lighter carbon and nitrogen at the University of Cape Town also produced four variables with significant discriminatory power. The results of the pioneering work by Dr Julia Lee-Thorp and her colleagues were corroborated by the data which confirmed that a stable carbon isotope ratio provides a cast iron diagnostic tech- nique to discriminate between white and black rhino horn. The pilot discriminant function analyses also showed the carbon and nitrogen data had some discriminatory ability, correctly classifying the source of 50% of the pilot samples.
Although the results show LA-ICP-MS is the best single technique, pilot analyses indicated the best results will be obtained by building a statistical model using data from all three methods. By combining LA-ICP-MS and ICP-OES data, 98.5% of the pilot sample was correctly sourced with the only misclassification occurring when a sample was allocated to an adjacent area within the same Park. By combining all three techniques together, the resultant model correctly classified all pilot black rhino horn samples used to build the models. While the real challenge will be the ability of the technique to correctly classify independent samples not used to build the statistical models, these preliminary results are still very encouraging and much better than were originally expected was possible.
|
|
|
African Rhino Specialist Group (AfRSG)
File available
details
Original text on this topic: