Fig 1.
Eight-step compliance procedure devised by DEFF for implementing the legal lion bone export quota from South Africa in 2017 and 2018 (Source: DEFF, pers. comm., 2019).
Sequential steps from permit application submission by source farms and traders to physical export from O.R. Tambo (ORT) airport are provided, indicating the three mandatory compliance inspections (red).
Fig 2.
Spot-checked skeletons (12.5% of every consignment) during a 3rd compliance inspection at ORT airport in 2018 (Step 7).
All checked bags were reweighed, inspected, photographed and had a sample taken for DNA species assignment and pairwise-comparative sample matching. The lion skeleton in (A) and (B) was the heaviest (38.0 kg) of all the skeletons weighed at their source farm (Step 5) in 2018, but only the third heaviest (20.5 kg) on export (Step 7). (C) Collecting a sample following the forensic protocol for genetic testing (farm = 14.0 kg; spot-check = 10.35 kg). (D) A completed inspection (19.2 kg; 14.9 kg).
Fig 3.
Spot-checked skulls were of (A) lions (skeleton in Fig 2A and 2B), except for one specimen in 2018, which resembled that of a (B) tiger (skeleton in Fig 4E). These are distinguished using the alignment of the posterior projections of the nasal-frontal and maxilla-frontal sutures: the sutures align in lions but not tigers. Also, unlike lions, tiger mandibles do not rock when placed on a flat surface [9].
Fig 4.
(A–D) Export-ready lion skeletons, processed and packaged in various ways, as reweighed at O.R. Tambo (ORT) airport during the 3rd compliance inspection (Step 7). (E) A suspected tiger skeleton confiscated at the port. Following DNA-based species assignment of P. tigris on a skeleton with the same tag number at the source farm (Step 5), this skeleton was confirmed as a tiger (by tag and skull morphology).
Fig 5.
Lion skeleton weights by trader, location, consignment and year.
(A) The difference in the mean weight (C.I. 95%) of skeletons recorded [9] at the airport before 2014 (red) compared to 2018 (green). The difference between mean farm (orange) and port skeleton weights packaged (dark blue) and unpackaged (light blue) for matched pairs in 2017. The difference in mean weight of matched-pair spot-checked skeletons weighed at source farms (grey) versus at O.R. Tambo airport (black) in 2018 (Table 1). (B) The differences in mean farm (grey) and airport (black) weights of spot-checked skeletons (12.5%) exported by seven traders in 2018. (C) The linear relationships between airport consignment weight and the number of skeletons in the consignment up to 2014 (red) and for 2016 (purple), as well as for 2017 and 2018 combined (green), where shading about the regression indicates the standard deviation. Note: unless specified otherwise, all port weights are for packaged skeletons.
Table 1.
Mean mass (kg) of captive-bred lion skeletons exported in 2017 and 2018 based on the weights recorded during the compliance inspections at source farms (all skeletons) and ORT airport.
Fig 6.
Species assignment and pairwise-comparative sample matching by relatedness scores between source farm and export consignments tags.
Species is confirmed through (A) an allelic discrimination plot for 12S, where axes correspond to the relative fluorescence of labelled primers. Squares indicate positive (red, homozygous allele 1, lion; blue, homozygous allele 2, tiger) and negative controls (undetermined), where circles and colour represent sample assignments. Genetic differentiation between species based on STRUCTURE analysis (performed with K = 2) is also visualized using DISTRUCT (inset), indicating 1) tiger reference controls, 2) lion reference controls, and 3) the suspected tiger sample (in duplicate). Pairwise-comparative sample matching for the (B) 2017 and (C) 2018 quota indicate farm-port tag matches as significant (P < 0.05) deviations from the normal pairwise relatedness distribution (black whiskers) within comparisons (green) and those non-significant as unmatched and warranting further investigation (yellow).
Table 2.
Number and proportion of identified sampling anomalies (species assignment and pairwise-comparative sample matching) in the DNA profiles by sampling location and year.