Contributions to Zoology, 76 (2) 87-94 (2007) Using ancient DNA techniques to identify the origin of unprovenanced museum specimens, as illustrated by the identi cation of a 19th century lion from Amsterdam Ross Barnett1, Nobuyuki Yamaguchi2, Beth Shapiro1 and Vincent Nijman3 1 Henry Wellcome Ancient Biomolecules Centre, University of Oxford, South Parks Road, Oxford, OX1 3PS UK; 2 Wildlife Conservation Research Unit, University of Oxford, Tubney House, Abingdon Road, Tubney, Abingdon, OX13 5QL UK; 3 Zoological Museum, University of Amsterdam, P.O. Box 94766, 1090 GT Amsterdam, The Nether- lands, [email protected] Key words: India, mitochondrial DNA, Panthera leo, skull morphology, South Africa Abstract Introduction In natural history ColleCtions throughout Europe, there are many The lion Panthera leo (Linnaeus, 1758) was onCe wide- old lion speCimens of unknown origin. If these speCimens Can spread throughout AfriCa and southwestern Eurasia, be shown to have originated from now-extinCt populations their ranging from South AfriCa to GreeCe, and from Mo- value would signi Cantly inCrease, as would the value of the ColleCtions. ReCently, a 200-year old mounted skeleton in the roCCo to India (Nowell and JaCkson, 1996). Most ZoologiCal Museum Amsterdam has been identi ed as the extinCt likely due to the expansion of human populations, lions Cape lion Panthera leo melanochaita (Smith, 1842), based disappeared from GreeCe by C. 2000 years ago, and primarily on morphologiCal information inferred from a painting from Palestine around the time of the Crusades C. 11th of this speCimen while it was still alive. To test this hypothesis, - 12th century. In the past 200 years, lions have disap- we used anCient DNA (aDNA) teChniques to extraCt and sequenCe peared from many parts of their former range, inClud- mitoChondrial DNA (mtDNA) from this speCimen, and Compared ing the southern part of South AfriCa (by 1870), Turkey the genetiC results with previously published lion mtDNA se- quenCes. Our results show that the speCimen is not a Cape lion, (1870), Tunisia (1891), Iraq (1918), Iran (1942), Mo- but that it instead possesses the mtDNA haplotype of the Asi- roCCo (1942) and Algeria (by 1960) (YamaguChi and atiC lions P. l. persica (Meyer, 1826) from India. This Indian Haddane 2002; Patterson, 2004; K. Difalla, personal origin hypothesis is further supported by an investigation of its CommuniCation). Among the now extinCt populations Cranial morphology. As the amount of genetiC information avail- two are partiCularly famous: the North AfriCan Bar- able for lions inCreases, in partiCular data from aCross their bary lion P. l. leo (Linnaeus, 1758) and the South Af- historiC distribution, the potential for aDNA teChniques to iden- tify the origins of previously unassigned museum speCimens riCan Cape lion P. l. melanochaita (Smith, 1842) Continues to grow. (Mazák 1975; Nowell and JaCkson 1996). This range ContraCtion has left fragmented populations in sub- Saharan AfriCa with an estimated total population of C. Contents 18,000 - 47,000, and a single small population (C. 300) in India (Nowell and JaCkson 1996; Bauer and Van der Introduction ...................................................................................... 87 Merwe 2002; Chardonnet 2002). Material and methods ..................................................................... 87 Prior to their extinction, lions from North Africa, Skull measurements ................................................................. 87 the Cape Province of South Africa, Iraq, and several DNA-extraction and analysis ................................................. 87 other sites were imported into Europe, where it was Data authenticity ....................................................................... 89 common for lions to be exhibited in menageries and Results ............................................................................................... 90 Discussion ........................................................................................ 91 zoos (Guggisberg 1963; Hemmer 1978). After death, Acknowledgements ........................................................................ 93 these specimens were deposited in natural history mu- References ........................................................................................ 93 seums or private collections, where they remain until 88 R. Barnett et al. – Origin of lion specimen today. Unfortunately, many of these speCimens are as- it may be possible to assign the origins of these speCi- soCiated with little or no provenanCe information and mens with reasonable con denCe. this is partiCularly true for speCimens dating to the 19th In this paper, we re-examine the origin of the Am- Century or earlier. As the reCent range ContraCtion has sterdam Cape lion using both aDNA teChniques and resulted in the loss of a substantial part of the lion’s skull morphology analysis. We also disCuss the feasibil- genetiC diversity, these European speCimens represent a ity of using these teChniques to identify previously signi Cant, as yet untapped resourCe for the investiga- unprovenanced museum specimens. tion of the evolution and historiCal distribution of lions. The reCently “disCovered” 200-year old lion speCimen at the ZoologiCal Museum of the University of Amster- Material and methods dam, The Netherlands, is an important Case in point (Van Bree, 1998). Previous analysis of this speCimen Skull measurements has suggested that it is a Cape lion (Van Bree, 1998). This identi Cation relied heavily on a painting made of The interorbital breadth and the postorbital ConstriCtion this speCimen while it was still alive by the DutCh artist were measured on the Amsterdam “Cape” lion (ID P.G. van Os (1776-1839) (Van Bree and Welman, 1996; number: ZMA710), and Compared to those of lion Van Bree, 1998; Reynaerts et al. 2006). However, os- skulls originated from India, and the southern part of teologiCal evidenCe presented in Van Bree (1998) Con- the former Cape ProvinCe, South AfriCa (Table 1). tradiCts this Classi Cation, as the measurements point Classi Cation was based on museum labels, and we out that this speCimen possesses a large interorbital also inCluded into the analysis a few old speCimens breadth (the smallest distanCe between the orbits) rela- labelled as Cape lion without any speCi C loCations tive to the postorbital ConstriCtion (the smallest distanCe within South AfriCa (see Table 1). We did not follow aCross the postorbital ConstriCtion), whiCh is a typiCal the Classi Cation of the Cape lion suggested by Mazák CharaCteristiC of the North AfriCan – Indian lion skulls, (1975) whiCh was reCently ContradiCted (Barnett et al., rather than sub-Saharan lions (Hemmer 1974). 2006a, 2006b). The ability to extraCt and analyse DNA sequenCes A skull was classi ed as subadult if cemento-enam- from old speCimens provides a meChanism for resolv- el junctions of all canines were already visible above ing this debate. OnCe an organism dies, enzymatiC the alveoli of the Cleaned skull and yet the basioCCipital- proCesses begin to break down its DNA, however re- basisphenoid suture, and/or frontal suture, was still Cent advanCes in moleCular biologiCal teChniques have open. If those sutures were Closed, a skull was Classi ed made it possible to amplify short fragments of anCient adult. Measurements were Carried out using a set of DNA (aDNA), normally mitoChondrial DNA (mtD- metal calipers to the nearest 0.02mm. The coef cient NA), from speCimens up to 100,000 years old (Lin- of variation for the interorbital breadth was 0.06% (n = dahl 1993). While lions were among the rst felids 5 skulls with 3 repeats eaCh), and that for the postorbital whose phylogeny was investigated using moleCular ConstriCtion was 0.09% (n = 5 skulls with 3 repeats methods (O’Brien et al. 1987), genetiC information each). Following Yamaguchi et al. (2004) we arbitrar- has been laCking for this speCies from many parts of ily accept a less than 2% coef cient of variation as a its natural range. Indeed, until very reCently, published cut-off line for reliability and consistency in measure- lion mt DNA sequenCes originated from only seven ments, which both met. Countries, making up less than half of its natural range: In addition to sexual size dimorphism that is Com- Botswana, India, Kenya, Namibia, South AfriCa, Tan- mon in the Felidae, it has been suggested that the skull zania, and Uganda (JanCzewski et al. 1995; Burger et morphologiCal CharaCteristiCs of Captive lions differ al. 2004; DubaCh et al. 2005). Under suCh CirCum- from those of wild animals (Hollister, 1917). While stanCes, even if mtDNA were suCCessfully obtained both the painting and morphometriC data suggest that from an unprovenanCed speCimen, it would be dif - the Amsterdam Cape lion is a male, the sex of the Cult to determine its origin. However, Barnett et al. speCimen was not expliCitly given in the original (2006a, 2006b) ampli ed the HVR1 of mtDNA Con- reCord. Additionally, the speCimen was known to have trol region from 32 modern lion specimens represent- spent at least some time in Captivity (Van Bree, 1998). ing most of the extant and historiCal natural range of We therefore inCluded both male and female and both lions. By Comparing sequenCes obtained from unprov- Captive and wild individuals in our Comparative data enanCed museum speCimens to the inCreased data set, set. Contributions to Zoology, 76 (2) – 2007 89 Table 1. SpeCimens of Cape lions and AsiatiC lions that were analysed in