Materials and methods
We examined 498 lion skulls kept in museum collections in Europe and southern Africa, originating from as many locations as possible across the species’ natural distribution, including recently extinct populations such as North African Barbary lion, Panthera leo leo Linnaeus, 1758. In addition, previously published information concerning the occurrence of DIF for 23 Indian lion skulls (Smee, 1835; Todd, 1965) was included in the analysis. To investigate the occurrence of DIF in the nearest internally-coherent taxon to the lion, we examined skulls of the extinct Pleistocene European cave lion, P. l. spelaea (Goldfuss, 1810), which is considered to have shared a common ancestor with the modern lion as recently as at c. 500-600 thousand years ago (Burger et al., 2004; Yamaguchi et al., 2004; Barnett et al., 2009). All eight examined cave lion skulls originate from Europe. We also examined 304 tiger, P. tigris Linnaeus, 1758, skulls kept in Europe and the former Soviet Union – from as many locations as possible across its natural distribution, representing all customarily recognised subspecies (Luo et al., 2004), including three extinct ones. The occurrence of DIF was assessed as ‘present’ or ‘absent’ regardless at which side it occurred. The broken specimen recorded by Todd (1965), ̔Mumbai 1261̓ in Appendix 1, was excluded from the analysis because it was not known whether DIF occurred in the skull. The two incomplete divisions recorded at National Museums Scotland (Edinburgh (Z.2001.174.2) and Edinburgh (Z.2008.004) see Appendix 1) were classified as DIF present (Fig 1b). All statistical analyses were carried out using SPSS (SPSS Inc., Chicago, USA), and statistically significant differences were detected using the likelihood ratio chi-square test, which is also referred to as the G-test.
Fig 1a. Undivided infraorbital foramina of a Sub-Saharan African lion (Amathole Museum, King William’s Town, South Africa, ID number: 18319, origin: Etosha Pan, Namibia) (photo: Nobuyuki Yamaguchi).