SEARCH

SEARCH BY CITATION

Keywords:

  • Acinonyx jubatus;
  • conservation genetics;
  • divergence time;
  • phylogeography;
  • population genetics;
  • subspecies

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

The cheetah (Acinonyx jubatus) has been described as a species with low levels of genetic variation. This has been suggested to be the consequence of a demographic bottleneck 10 000–12 000 years ago (ya) and also led to the assumption that only small genetic differences exist between the described subspecies. However, analysing mitochondrial DNA and microsatellites in cheetah samples from most of the historic range of the species we found relatively deep phylogeographic breaks between some of the investigated populations, and most of the methods assessed divergence time estimates predating the postulated bottleneck. Mitochondrial DNA monophyly and overall levels of genetic differentiation support the distinctiveness of Northern-East African cheetahs (Acinonyx jubatus soemmeringii). Moreover, combining archaeozoological and contemporary samples, we show that Asiatic cheetahs (Acinonyx jubatus venaticus) are unambiguously separated from African subspecies. Divergence time estimates from mitochondrial and nuclear data place the split between Asiatic and Southern African cheetahs (Acinonyx jubatus jubatus) at 32 000–67 000 ya using an average mammalian microsatellite mutation rate and at 4700–44 000 ya employing human microsatellite mutation rates. Cheetahs are vulnerable to extinction globally and critically endangered in their Asiatic range, where the last 70–110 individuals survive only in Iran. We demonstrate that these extant Iranian cheetahs are an autochthonous monophyletic population and the last representatives of the Asiatic subspecies A. j. venaticus. We advocate that conservation strategies should consider the uncovered independent evolutionary histories of Asiatic and African cheetahs, as well as among some African subspecies. This would facilitate the dual conservation priorities of maintaining locally adapted ecotypes and genetic diversity.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

At the end of the nineteenth century, cheetahs were widespread across Africa and much of Asia, ranging from the Indian peninsula to Kazakhstan and Southwest Asia (Pocock 1939; Durant et al. 2008). Today only fragmented populations remain on both continents (Durant et al. 2008) and are traditionally classified in four African and one Asiatic subspecies (Fig. 1a) (Krausman & Morales 2005). Despite its vast geographical distribution over two continents, the cheetah is regarded as a genetically depauperate species (O’Brien et al. 1987; May 1995; Menotti-Raymond & O’Brien 1995; O’Brien & Johnson 2005). This low genetic variability is considered to be the result of a bottleneck at the end of the Pleistocene [10 000–12 000 years ago (ya); O’Brien et al. 1987; Menotti-Raymond & O’Brien 1993; O’Brien & Johnson 2005] and has been offered as a possible explanation for the population decline. However, there is little evidence of inbreeding depression in wild cheetahs (Caro & Laurenson 1994). In fact, anthropogenic habitat modification, replacement of wild prey with livestock and concomitant persecution by people (Laurenson 1994; Durant et al. 2008; Marker et al. 2008) account for the dramatic decline in historical range and numbers (Caro & Laurenson 1994). While it is unclear if Asiatic populations ever reached the density of their African counterparts, historical records report large numbers of cheetahs in Asia until the nineteenth century. During the Middle Ages and early Modern Times, Mughal emperors, in particular Akbar the Great (1556–1605), were known to keep thousands of cheetahs as hunting aids (Pocock 1939; Nowell & Jackson 1996; Allsen 2006; Divyabhanusinh 2007). This practice spread to Europe (Masseti 2009) and Southwest Asia (Brehm 1879; Allsen 2006) until cheetahs became rare, which led to regular imports of individuals from East Africa (Pocock 1939; Divyabhanusinh 2007) into India during the European colonial era. Until now, only sub-Saharan populations (Menotti-Raymond & O’Brien 1995; Freeman et al. 2001; Burger et al. 2004; Kotze et al. 2008; Marker et al. 2008) and a few Algerian individuals (Busby et al. 2009) have been investigated using genetic markers. Accordingly, comprehensive data regarding the relationships among all African subspecies and between African and Asiatic cheetah populations are still lacking.

image

Figure 1.  Median-joining (MJ) networks showing phylogeographic structure in African and Asiatic cheetahs. (a) Geographical distribution of the cheetah subspecies and sample repartition. Solid and dashed lines represent the historical distributions of the African and Asiatic cheetah subspecies, respectively (Nowell & Jackson 1996; Krausman & Morales 2005). Hatched fields correspond to current cheetah populations (Durant et al. 2008). The different colour shades refer to the screened cheetah subspecies, Acinonyx jubatus jubatus (red), A. j. raineyi (yellow), A. j. soemmeringii (purple), A. j. venaticus (green), and to the North African cheetah population (blue). Stars indicate the archaeological sites of Bastam and Takht-e Suleyman, Iran. The dotted line represents the southern boundary of the Sahara. The background map was retrieved from http://www.planiglobe.com (accessed 14 January 2010). (b) MJ-network based on the 139-bp concatenated mitochondrial sequence alignment of 94 samples. (c) MJ network based on the 915-bp concatenated mitochondrial fragment obtained from 62 modern and 16 historical cheetah samples. The consensus networks of all the shortest trees are shown. The specimens included are colour-coded according to their geographical origins (country codes following ISO 3166-Alpha 2). Small black squares represent median vectors, which correspond to either homoplasies or missing haplotypes. Red numbers above lines refer to nucleotide mutations separating the haplotypes [numbering according to GenBank (Accession no. GI:38349475.1)]. Positions 12665–12667 correspond to a 3-bp indel in MT-ND5, which we parsimoniously considered as a single evolutionary event. Exact positions of the concatenated mitochondrial fragments are given in Table S1.

Download figure to PowerPoint

In this study, we investigated the phylogeography, genetic structure and evolutionary history of cheetahs from most extant and recently extinct populations in Africa and Asia. We give particular attention to the Asiatic cheetah, because it is critically endangered and restricted to a small remnant population in Iran and possibly a few individuals in Pakistan (Farhadinia 2004) and Afghanistan (Manati & Nogge 2008). Asiatic cheetahs are known to occur in 13 sites in central and northern Iran where the total population is estimated at 70–110 (Farhadinia 2004; CACP 2008). Widespread poaching of the cheetah’s prey base and persecution by local livestock herders are the main causes of the cheetah’s recent decline and, together with road accidents, are likely the limiting factors to their recovery today (Hunter et al. 2007; CACP 2008). Historical records of extinction in the Arabian Peninsula indicate that this population became progressively and ultimately isolated from any potential link to Africa between approximately 1950 and 1980 (Hunter & Hamman 2003). However, it was unclear if demographic and genetic exchange between African and Asiatic cheetahs occurred prior to this recent anthropogenic isolation. To investigate these questions, we apply palaeogenetic analyses to compare extinct and extant Asiatic cheetahs with the major African populations. By demonstrating that all Northern-East African individuals, as well as all Asiatic cheetahs group within independent clusters, clearly distinct from other genotypes and monophyletic for mitochondrial DNA (mtDNA), we identify these two populations as long-term geographic isolates. The identification of taxonomic and populations units, and understanding their evolutionary relationships, is essential for the conservation of biological diversity (Allendorf & Luikart 2007). Within species, preservation of genetically distinct local populations maintains evolutionary processes and potential and minimizes extinction risks (Frankham et al. 2009).

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

Sample collection

Details on the origin and sample type of the 94 cheetahs included in this study are provided in Table 1. Modern samples were either collected non-invasively, during routine veterinary treatment, or post mortem. The osseous remains of cheetahs analysed in this study were collected from the archaeological sites of Bastam and Tahkt-e Suleyman in the Province of West Azerbaijan, Northwest Iran. Bastam is situated ∼50 km north of the city of Khvoy, close to the Turkish border at an altitude of 1300 m. First inhabited in Urartian times and destroyed by a fire c. 650 bce (Before Common Era), Bastam was reoccupied during the Median and Persian period (550–330 bce) and finally in mediaeval times. Archaeological excavations in the 1970s produced a large faunal assemblage (= 26 987) (Boessneck & Krauß 1973; Krauß 1975). From a mediaeval context (9th–15th century ce), a complete Os metatarsale IV was recovered. Originally described as a wolf metatarsal (Boessneck & Krauß 1973), the specimen was re-identified after a thorough comparison with fourth metatarsal bones of modern Acinonyx jubatus. The site of Tahkt-e Suleyman (lit. ‘Throne of Salomon’) lies midway between Urmia and Hamadan, about 30 km north of the town of Takab at an altitude of 2000 m. Archaeological excavations carried out in the 1960s revealed Achaemenid/Persian, Parthian, Sassanid and mediaeval inhabitation, and produced a faunal assemblage of 4000 bones. Kolb (1972) assigned a mandible, a cervical vertebra, and left and right coxal bones to A. jubatus, based on morphological criteria. The presence of postcranial fragments is strongly indicative of a local origin of the animal(s) and not of an individual whose pelt had been traded into the site. The mandible, vertebra and one coxa from the site of Tahkt-e Suleyman, as well as the metatarsus from Bastam, were subjected to palaeogenetic analysis. All cheetah bones date to the 9th and/or 10th centuries ce.

Table 1.   Details on the samples used in this study
No.IDTypeOrigin (time*)Collection place
  1. *Date of collection of historical samples, if available.

  2. #: registration number in the international cheetah studbook (Marker 2009); maxillot. bone: maxilloturbinate bone; museum tissue: dried tissue remaining on the skull; IR: Ariz & Bafq Protected Area, eastern Yazd province, Naybandan Wildlife Refuge south of Tabas, IR; Sarjah: Breeding Centre for Endangered Arabian Wildlife, Sharjah, AE; HZM: Harrison Institute, Sevenoaks, GB; NHM: Natural History Museum; SAP: State Collection of Anthropology and Palaeoanatomy, Munich, DE; SMN: Museum of Natural Sciences, Stuttgart, DE; SMN: Naturmuseum Senckenberg, Frankfurt, DE; DECAN: DECAN rescue centre, Djibouti, DJ; La Palmyre: Parc zoologique de La Palmyre, FR; Zoo Lunaret: Parc Zoologique Henri de Lunaret, Montpellier, FR; NZG: National Zoological Gardens, Pretoria, ZA; NMSZ: National museum of Scotland, Edinburgh, GB; RMCA: Royal Museum of Central Africa, Tervuren, BE. Country codes following ISO 3166-Alpha 2.

 1AJI 01FaecesIran (IR)IR
 2AJI 02FaecesIRIR
 3AJI 03FaecesIRIR
 4AJI M1AFaecesIRIR
 5AJI M2AFaecesIRIR
 6AJI TFaecesIRIR
 7AJI 08FaecesIRIR
 8AJI 11FaecesIRIR
 9AJI 04Museum tissueIRSharjah, AE
10HZM 2.26502HideOman (OM; 1977)HZM, GB
11ZMB 56122HideJordan (JO)NHM Berlin, DE
12BMNH ZD 1939.536HideIRNHM London, GB
13SAPM-Gepard-1a-F6-BaMetatarsal boneIR, Bastam (800–900 ce)SAP Munich, DE
14SAPM-Gepard-1b-TeS1Mandible + vertebraIR, Tahkt-e Suleyman (800–900 ce)SAP Munich, DE
15BMNH ZD 1943.56Museum tissueIraq (IQ; 1928)NHM London, GB
16BMNH 32.4.7.1HideIndia (IN; 1925)NHM London, GB
17SMNS 18941Maxillot. boneEgypt (EG; T. v.Heuglin; 1850s)SMN Stuttgart, DE
18SMF 58993HideEG, Libyan Desert (1974)SM Frankfurt, DE
19NMW 12071HideLibya (LY)NHM Vienna, AT
20NMW 12070HideLYNHM Vienna, AT
21BMNH ZD 1957.312HideLY (1955)NHM London, GB
22BMNH ZD 1939.1685Museum tissueAlgeria (DZ)NHM London, GB
23ZMB 56277Maxillot. boneWestern Sahara (EH)NHM Berlin, DE
24ZMB 42242HideEH, Rio d’OroNHM Berlin, DE
25ADJ 2HairEthiopia (ET), custom’s seizureDECAN, DJ
26ADJ 3HairET, custom’s seizureDECAN, DJ
27ADJ 4HairET, custom’s seizureDECAN, DJ
28ADJ 5HairET, custom’s seizureDECAN, DJ
29ADJ 6HairET, custom’s seizureDECAN, DJ
30ADJ 8HairDjibouti (DJ)DECAN, DJ
31#4421SkinSomalia (SO)Sharjah, AE
32#4499SkinSOSharjah, AE
33#4500SkinSOSharjah, AE
34#4203SkinSOSharjah, AE
35#4208SkinSOSharjah, AE
36#4202SkinSOSharjah, AE
37#4223SkinSOSharjah, AE
38#4205SkinSOSharjah, AE
39#4229SkinSOSharjah, AE
40#4228SkinSOSharjah, AE
41#4206SkinSOSharjah, AE
42#4201SkinSOSharjah, AE
43#4418SkinSOSharjah, AE
44#4222BloodSOSharjah, AE
45#4216SkinSDSharjah, AE
46LP4304HairNorthern-east Africa (N-E.A)La Palmyre, FR
47#4415SkinN-E.ASharjah, AE
48SMNS 38432HideDJ, custom’s seizure (<1985)SMN Stuttgart, DE
49ADJ 1HairET, custom’s seizureDECAN, DJ
50ADJ 7HairET, custom’s seizureDECAN, DJ
51ClaudiaFaecesKenya (KE)DECAN, DJ
52ZMB34306Maxillot. boneTanzania (TZ)MHN Berlin, DE
53ZMB56287Maxillot. boneTZMHN Berlin, DE
54ZMB56302Maxillot. boneTZMHN Berlin, DE
55ZMB56306Maxillot. boneTZMHN Berlin, DE
56ZMB56309Maxillot. boneTZMHN Berlin, DE
57TiggerFaecesKEDECAN, DJ
58ZMB56128Maxillot. boneTZMHN Berlin, DE
59ZMB56289Maxillot. boneTZMHN Berlin, DE
60ZMB56293Maxillot. boneTZMHN Berlin, DE
61ZMB56299Maxillot. boneTZMHN Berlin, DE
62GACH 18/08BloodSouth Africa (ZA)Pretoria NZG, ZA
63GACH 23/06BloodZAPretoria NZG, ZA
64GACH 26/08BloodZAPretoria NZG, ZA
65GACH 33/08BloodZAPretoria NZG, ZA
66GACH 35/08BloodZAPretoria NZG, ZA
67GACH 38/08BloodZAPretoria NZG, ZA
68GACH 42/08BloodZAPretoria NZG, ZA
69GACH 44/08BloodZAPretoria NZG, ZA
70GACH 45/08BloodZAPretoria NZG, ZA
71#1463LungNamibia (NA) descendantLa Palmyre, FR
72#1557/NMSZ 2001.37MuscleNANMSZ, GB
73S1571MuscleZA descendantPrivate owner, DE
741921MuscleNA descendantZoo Salzburg, AT
75#3155/NMSZ 2000.151.2MuscleNA descendantNMSZ, GB
76#3240FaecesZA descendantZoo Vienna, AT
77GACH 34/08BloodZAPretoria NZG, ZA
78#3779/NMSZ 1999.221MuscleZA descendantNMSZ, GB
79DoumaMuscleZAZoo Lunaret, FR
80GACH 25/08BloodZAPretoria NZG, ZA
81GACH33BloodZAPretoria NZG, ZA
82GACH 01/08BloodBotswana (BW)Pretoria NZG, ZA
83GACH 02/08BloodBWPretoria NZG, ZA
84GACH 11/08BloodBWPretoria NZG, ZA
85GACH 12/08BloodBWPretoria NZG, ZA
86GACH 15/08BloodBWPretoria NZG, ZA
87GACH 16/08BloodBWPretoria NZG, ZA
88#4268SkinNASharjah, AE
89#2486/ZFMK 2005.357HideZA, king cheetahMHN Bonn, DE
90RMCA 454Maxillot. boneD.R.Congo (CD)RMCA, BE
91RMCA 1236Maxillot. boneCDRMCA, BE
92RMCA 19237Maxillot. boneCDRMCA, BE
93RMCA 22347Maxillot. boneCDRMCA, BE
94RMCA 22390Maxillot. boneCDRMCA, BE
95Puma concolorBloodUnknownZoo Salzburg, AT

DNA extraction

Ancient DNA extractions from mediaeval cheetah specimens were performed in highly contained laboratories of the palaeogenetic core facility at the Institute Jacques Monod, Paris (see Supporting Information). The superficial layer of the small bone fragments was removed and samples were ground to a fine powder in a freezer mill (Freezer Mill-6750; Spex Certiprep). The bone powder (180 mg) was incubated (37 °C; 48 h) in extraction buffer (0.5 m EDTA pH 8.0; 0.25 m sodium phosphate buffer pH 8.0; 1 mmß-Mercapto-ethanol). The extract was purified according to an improved protocol using the QIAquick gel extraction kit (Qiagen). Museum skin pieces were incubated twice (24 h) in TE buffer to remove potential enzyme inhibitors (Johnson et al. 2004). Complete enzyme digestion was carried out in an improved lysis buffer [100 mm Tris–HCl pH 8.0; 100 mm NaCl; 3 mm CaCl2; 2%N-lauroyl-sarcosyl (NLS); 40 mm DTT; 5 mm PTB (N-phenacyl-thiazolium-bromide (Vasan et al. 1996) in 10 mm phosphate buffer); 340 μg proteinase K] (Pfeiffer et al. 2004). After 24 h one-eighths of an Inhibitex pill (Qiagen) was added to the samples, which had extensively undergone a tanning process. Maxilloturbinate bone shreds (Wisely et al. 2004) were ground and the bone powder was incubated (56 °C; 48 h) in lysis buffer (0.5 m EDTA pH 8.0; 0.25 m sodium phosphate buffer pH 8.0; 1 mmß-Mercapto-ethanol; 2% NLS; 340 μg proteinase K). DNA extraction was performed with commercial kits (Qiagen) in the presence of negative controls. Genomic DNA of modern samples was isolated from blood and tissue with the NucleoSpin®-Tissue Kit (Macherey-Nagel). Faeces were processed, following a two-step storage protocol (Nsubuga et al. 2004), with the QIAamp DNA Stool Mini Kit (Qiagen). Hair samples were digested with a lysis buffer (Pfeiffer et al. 2004) and DNA was extracted with the NucleoSpin®-Tissue Kit (Macherey-Nagel). Two independent extractions were carried out for the mediaeval bones, as well as for the other samples where sufficient material (i.e. museum specimen) was available.

Quantitative real-time polymerase chain reaction and sequencing of mediaeval cheetah specimens

Fragments of 139 base pairs (bp), including 14 informative polymorphisms in NADH-dehydrogenase subunit 5 (MT-ND5) and control region (MT-CR) (Tables 2 and S1), were amplified from two mediaeval and 14 museum specimens by UNG-coupled quantitative real-time polymerase chain reaction (UQPCR) (Pruvost et al. 2005) with the LightCycler® FastStart DNA MasterPLUS SYBR Green I mix (Roche Diagnostics GmbH). Quantification of initial target molecules was performed using a titration curve established with a homologous reference (DNA from a Namibian specimen of A. j. jubatus) according to Pruvost et al. (2005, 2007). The inhibition of the polymerase by the aDNA extracts was quantified (Pruvost & Geigl 2004) and the applied amount of target DNA was adjusted accordingly. The characterization of the PCR products was performed by analysis of the fusion temperature (Tm) using the Lightcycler® and via electrophoresis in a 10% polyacrylamide gel. Multiple PCR amplifications were performed on each independent extract. PCR products were purified using the QIAquick PCR purification kit (Qiagen) and sequenced in both directions by Eurofins MWG GmbH.

Table 2.   Informative sites screened in the 139-bp mtDNA concatenated fragment
AmpliconsntA. j. venaticusA. j. jubatusA. j. soemmeringiiNorth Africa populationA. j. raineyi
  1. nt: nucleotide position (GenBank Accession no. GI:38349475.1); MT-ND5: mitochondrial NADH-dehydrogenase subunit 5; MT-CR: mitochondrial control region. Diagnostic nucleotide polymorphisms are highlighted in bold. A. j. venaticus (Acinonyx jubatus venaticus) refers to the Southwest Asian cheetah population.

MT-ND512665–12667ATCATCATCATC
12679T/CCCCC
12698CCACC
12707AAAGA
MT-CR116448TCCCC
16454TTTTC/T
16473TTCTT
16474AGGAG
MT-CR316817TTTCC/T
16818AAG/AAA
16831AAAAG/A
16854AAAAG/A

DNA sequencing and genotyping

A total of 62 modern and 16 historical cheetah PCR products were sequenced for parts of MT-ND5 [nt 12657–13087; numbering according to GenBank (Accession no. GI:38349475.1)], cytochrome b (MT-CB; nt 15940–16173) and MT-CR without repetitive sequences (Lopez et al. 1996) (nt 16333–16487 and nt 16811–16876), resulting in a 915-bp concatenated mitochondrial fragment. The last 29 bp of the tRNA Leucine (tRNA-Leu; nt 12628–12656) were analysed to ensure the correct amplification of a 3-bp indel mutation at the third and fourth codons of MT-ND5. Sequencing was performed in both directions using a MegaBACE 500 sequencer (GE Healthcare). Genotypes were obtained from 60 modern and seven historical samples at 20 microsatellite loci (FCA005, FCA008, FCA014, FCA026, FCA069, FCA078, FCA085, FCA096, FCA097, FCA105, FCA126, FCA133, FCA171, FCA212, FCA214, FCA220, FCA224, FCA230, FCA247, FCA310) developed in Felis catus (Menotti-Raymond et al. 1999) and tested on cheetahs (Driscoll et al. 2002; Kotze et al. 2008; Marker et al. 2008). Southern African sample analyses were performed at the Centre for Conservation Science of the National Zoological Gardens in Pretoria using an ABI 3130 sequencer (Applied Biosystems Inc.). The genotypes of all other samples were determined with a MegaBACE 500 at the University of Veterinary Medicine, Vienna. At least three independent genotype results were produced for each locus and each individual. Two defined standard individuals were run in each genotyping series. Electropherograms were evaluated using the softwares GeneMapper v3.1 (Applied Biosystems) and MegaBACE Genetic Profiler v2.2 (GE Healthcare), respectively. Results from the loci FCA078 and FCA096 were removed because of the insufficient quality of the electropherograms despite multiple reiterations.

Mitochondrial and nuclear DNA data analysis

Mitochondrial sequences were deposited in GenBank (Accession nos puma, GU984641; cheetah, GU984642GU984735). Sequences were aligned with Codon Code Aligner (version 3.0.2; Codon Code Corporation). A new polymorphism was considered as authentic when it was displayed in at least three independent sequences. The 3-bp deletion (nt 12665–12667) was considered as a single indel event in all subsequent analysis. Mitochondrial haplotype diversity (Hd) and nucleotide diversity (π) (Tajima 1983; Nei 1987) were calculated using Arlequin 3.5 (Excoffier & Lischer 2010). The genetic structure of cheetah populations was analysed using a Bayesian approach implemented in baps 5.2 (Corander & Marttinen 2006; Corander et al. 2008). In this method, the number of populations is treated as unknown parameter and is directly inferred from the data set without defining a prior estimate. For inferring population structure in the mitochondrial (139 bp, = 94; 915 bp, = 60) and nuclear (18 loci, = 60) data sets, we assigned individuals to distinct clusters using the models ‘clustering of linked loci’ (Corander & Tang 2007) and ‘clustering of individuals’, respectively. We specified prior upper bound values for the number of clusters in the data (i.e. 5–10) and performed 10 independent runs for each value. In all independent runs, the assignments of individuals resulted in the same clusters. We performed the admixture analysis based on the results of the mixture clustering of the nuclear data using 500 iterations and a number of 1000 reference individuals per population, each with 10 reiterations. For additional population structure analysis, we used the three-dimensional factorial correspondence analysis (FCA) in genetix 4.05 (Belkir et al. 1999), which portrays the relationship between individuals or populations based on the detection of the best linear combination of allele frequencies. By comparing the clustering solutions of the different methods, we defined the cheetah populations for subsequent population genetic analyses. Average polymorphisms, allele frequencies, expected heterozygosities (HE), pairwise FST, genetic distance (δμ)2 (Goldstein et al. 1995) and proportion of shared alleles between individuals (DPS; Bowcock et al. 1994) were estimated from the microsatellite data set with MSAnalyzer 4.05 (Dieringer & Schlötterer 2003). The stepwise weighted genetic distance (DSW; Shriver et al. 1995) was calculated with Populations 1.2.30 (Langella 1999). Statistical significance for mean HE was tested with the Wilcoxon rank-sum test using r version 2.10.1 (R Development Core Team 2009). An analysis of molecular variance (amova) was performed to determine the proportion of genetic variance explained by the differences within and between modern populations as determined by BAPS and FCA. amova calculations were performed in Arlequin 3.5 and significance levels were obtained with 10 000 permutations. Neighbor-joining (NJ) trees were generated based on the proportion of shared alleles between individuals with the software phylip 3.69 (Felsenstein 1989), visualized in FigTree v1.3.1. (Rambaut 2009) and edited in Adobe® Illustrator® CS4 14.0.0 (Adobe System Inc.). We also tested the populations defined by BAPS and FCA for evidence of a decline in their effective population sizes using the program Bottleneck 1.2.0.2 (Piry et al. 1999). We performed the evaluation using the stepwise mutation (SMM) and two-phase (TPM) models of microsatellite evolution. The significance of the tests was assessed using Wilcoxon sign-rank test, which is the most appropriate test when fewer than 20 microsatellite loci are used (Piry et al. 1999). Median-joining networks were constructed with Network 4.5 (Bandelt et al. 1999) with adapted settings following the software instructions. We applied a transition/transversion ratio of 14 (MJN 139 bp) and 10 (MJN 915 bp) estimated with ModelTest (Posada & Crandall 1998) using the best fitting model [Hasegawa–Kishino–Yano (HKY); Hasegawa et al. 1985] according to the Akaike Information Criterion (AIC; Akaike 1974). An exact test of population differentiation (Raymond & Rousset 1995) was performed, and the population pairwise FST values were calculated based on the 139- and 915-bp mitochondrial fragment using the distance method of Tamura & Nei (1993) implemented in Arlequin 3.5.

Divergence time estimations

In all divergence time estimations, we used the most complete sample set (= 67) for which both mitochondrial (915 bp) and nuclear (18 loci) data could be retrieved. The 3-bp deletion was considered as one evolutionary event. A puma sample (Puma concolor) was sequenced and included as outgroup to build a maximum-likelihood (ML) tree using the HKY model in Tree-Puzzle 5.2 (Schmidt et al. 2003). We tested whether the assumption of a molecular clock was valid by performing a likelihood ratio test between the simpler clock model vs. the more complex model without clock. The log-likelihood of the more complex model was not significantly increased with respect to the simpler model (> 0.05), supporting the assumption of a molecular clock (Felsenstein 1985). Given an estimated cheetah–puma divergence at 4.92 Ma (95% CI = 3.86–6.92) (Johnson et al. 2006), the substitution rate was inferred using the formula dxy = 2μT, where T is the time to the most recent common ancestor, μ is the mutation rate per year and dxy is the genetic distance between species corrected for ancestral polymorphism (Nei 1987). For the computation of dxy, we used the software mega 4.0 (Tamura et al. 2007), which allows for rate heterogeneity among lineages, and the Tamura–Nei substitution model (Tamura & Nei 1993) with Γ = 0.118 (parameter selected by the AIC with correction for small sample size; AICc) as selected by Treefinder (Jobb et al. 2004). The estimated dxy was 0.567 (SD ± 0.175), which translates into a substitution rate of 5.76 × 10−8 substitutions per site per year (95% CI = 1.57 × 10−8–1.19 × 10−7). The divergence times between Asiatic (A. j. venaticus) and Southern African (A. j. jubatus) cheetahs and between Northern-East (A. j. soemmeringii) and Southern African cheetahs were estimated using the equation DA = 2μT in which μ is the average substitution rate per nucleotide, T is the divergence time, and DA is the net number of nucleotide differences between populations (Nei & Li 1979). Divergence times were also calculated following the coalescent method described by Gaggiotti & Excoffier (2000). This method aims to remove the effect of bottlenecks and unequal sizes of the derived populations, which can lead to the overestimation of divergence times from genetic distances. Additionally, these divergence times were estimated using IMa (Hey & Nielsen 2007). This program implements a coalescent-based isolation with migration model that can be applied to genetic data drawn from a pair of closely related populations or species (Nielsen & Wakeley 2001) to infer six demographic parameters [population sizes of the extant as well as the ancestor population, migration rates (m1, m2) per gene in both directions, and time (t) since divergence]. After preliminary runs to optimize settings, four replicate simulations were run. Estimates were generated under the HKY model. Simulations used 10 Markov chains, with 45 chain swap attempts per step, and were run for 20 million steps discarding the first 1 million steps as ‘burn-in’. Genealogies were sampled every 100 steps. Convergence of the simulations was assessed by comparison of their marginal parameter distributions across independent replicate runs. Saved genealogies were used to estimate the joint marginal distribution of t (the estimator of population divergence time) from an evenly spaced sample of 200 000 trees. To convert coalescent times to years before present, we used the substitution rate estimated above and a generation time of 6 years (Marker & O’Brien 1989). ‘Nested models’ (Hey & Nielsen 2007) were also examined and compared to the full six-parameter model using log-likelihood ratio tests. For comparison with previous divergence time estimates between cheetah subspecies based on microsatellite data (Driscoll et al. 2002), we estimated the timing of the splits between Asiatic and African subspecies (A. j. venaticus and A. j. jubatus) and among African cheetahs (A. j. soemmeringii and A. j. jubatus) using the (δμ)2 genetic distance and the equation (δμ)2 = 2μG (μ = mutation rate; G = generations) (Goldstein & Pollock 1997). We applied two estimates of mutation rate for microsatellite loci in humans (5.6 × 10−4 and 2.05 × 10−3), which were previously used by Driscoll et al. (2002), and an additional estimate for the average microsatellite mutation rate in mammals (2.05 × 10−4; Rooney et al. 1999) that has been employed in several studies on other felid species (Spong et al. 2000; Anderson et al. 2004; Ruiz-Garcia et al. 2006). Furthermore, to estimate the divergence times among African populations and between Asiatic and African cheetahs, we used the stepwise-weighted genetic distance (DSW; Shriver et al. 1995) and the equation from Calabrese et al. (2001): DSW = √(2/π) × √(2βτ + 4βNe) − 4βNe/√(8βNe + 1), where π is a constant; β, mutation rate; τ, time in generations; Ne, effective population size calculated under the SMM and inferred from the expected heterozygosity (Ohta & Kimura 1973).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

Genetic variation and population structure analysis

We screened 94 Asiatic and African cheetahs by combining data from 62 modern, 30 historical and 2 zooarchaeological specimens (Table 1). Three mtDNA regions (MT-ND5, MT-CB, MT-CR; partial sequences), corresponding to a total of 915 bp were sequenced from all modern and 16 well-preserved historical specimens. We identified 29 polymorphic sites and one 3-bp deletion resulting in 18 haplotypes (Hd = 0.909, SD = 0.013; π = 0.00659, SD = 0.00351). The highest numbers of polymorphic sites (= 7) were detected within cheetahs originating from Southern Africa and East Africa, respectively, whereas Northern-East African and Asiatic cheetahs showed lower amounts of mitochondrial polymorphism (= 3 and = 2, respectively). A similar pattern was observed for haplotype (Hd) and nucleotide diversities (π) (Table 3). For the palaeogenetic analyses, we selected three diagnostic regions (139 bp) containing 14 informative sites (Table 2). These sites faithfully recovered the partitioning into haplogroups observed in the 915-bp data set. We successfully amplified these regions in mediaeval A. jubatus specimens from two archaeological sites in Iran, Bastam (metatarsal bone) and Tahkt-e Suleyman (vertebra, mandible). These samples represent the few cheetah bones hitherto discovered in archaeological excavations in Southwest Asia. In addition, we amplified these diagnostic regions in 14 highly degraded DNA samples originating from countries where cheetahs are now extinct (e.g. India) or close to extinction. All replicates of the sequences retrieved from the independent extracts were identical.

Table 3.   Genetic variation in cheetahs inferred from mitochondrial DNA (mtDNA) and nuclear DNA (μsat) data
PopulationmtDNA (915 bp)
No. cheetahs (mtDNA/μsat*)No. haplotypesNo. variable sitesHaplotype diversity (SE)π (SE)
  1. *Nuclear genetic variation was assessed only among the extant populations.

Total78/601829 + 1 indel0.909 (0.013)0.00659 (0.00352)
S-West Asia11/8320.345 (0.172)0.00040 (0.00047)
N-East Africa26/25330.551 (0.048)0.00073 (0.00064)
Southern Africa29/27870.828 (0.046)0.00197 (0.00130)
East Africa11/0370.636 (0.090)0.00381 (0.00237)
North Africa1/010
 μsat (18 loci)
% Polymorphic lociTotal no. allelesAverage no. allele/locus (SE)% Private allelesHE
 1001458.06 (1.39)31.720.766
 88.9422.33 (1.03)7.140.397
 1001075.94 (1.70)14.950.674
 1001116.17 (1.38)24.320.698
 
 

We used haplotype network analysis and BAPS to infer the relationships between the different mtDNA haplotypes. In both median-joining networks (Fig. 1b: 139 bp and Fig. 1c: 915 bp), we observed a star-shaped radiation stemming from a Southern African haplogroup corresponding to the subspecies A. j. jubatus. The East African cheetahs, described as A. j. raineyi, emerged in two different branches from the central haplotype. Although none of the East African cheetahs shared a common haplotype with the Southern African individuals, one haplotype comprising Tanzanian and Kenyan cheetahs (defined by nt 16817; Fig. 1b) clustered together with Southern African cheetahs in the BAPS analysis [posterior probability (PP) = 1; Fig. 2a]. Another sub-Saharan cheetah haplogroup was defined (Fig 2a) corresponding to the Northern-East African subspecies A. j. soemmeringii. We observed a monophyletic clustering for this haplogroup in the ML tree (915 bp) with a bootstrap support of 99% (1000 iterations; Fig. S1). Two more haplogroups were recovered (Fig. 1b) in samples from different parts of a range (North Africa and Southwest Asia) that had previously been considered to harbour the same subspecies A. j. venaticus (Krausman & Morales 2005; Belbachir 2007). The partitioning of these cheetahs into two distinct clusters was also confirmed by BAPS (Fig. 2a). One of these clusters encompassed animals from Western Sahara, Algeria, Libya and western Egypt (Libyan Desert; Osborn & Helmy 1980). The other cluster contained three Asiatic haplotypes represented by the current, historic and mediaeval Iranian cheetah samples as well as by museum specimens from India, Oman, Iraq and Jordan (Fig. 1b). In addition, one sample collected by Theodor von Heuglin in eastern Egypt (Table 1) clustered with this Asiatic haplogroup (Figs 1b and 2a). An exact test of population differentiation based on haplotype frequencies resulted in significant differences (< 0.05) between all African and Asiatic clusters as defined by the Bayesian structure analysis (Fig. 2a). The population pairwise genetic distances (FST) among these clusters ranged from 0.724 to 0.930 (within Africa) and 0.818–0.958 (Southwest Asia vs. African populations; Table 4). Comparing the FST values among the African clusters with those calculated between Asiatic and African populations, no significant differences were detected (= 0.246; Wilcoxon rank-sum test corrected for multiple testing).

image

Figure 2.  Bayesian analysis of population structure (BAPS) of African and Asiatic cheetahs. (a) Clustering based on a 139-bp mitochondrial concatenated fragment of 94 cheetahs. Individuals (represented by single bars) are assigned to five distinct clusters (posterior probability, PP = 1). (b) Clustering based on a 915-bp mitochondrial fragment and 18 microsatellite loci using 60 modern cheetahs. Extant cheetahs are assigned to three (PP = 0.999) and five (admixture analysis; PP = 0.999) clusters using mitochondrial (mtDNA) and nuclear DNA (nDNA), respectively. SW-ASIA, Southwest Asia; N-AFRICA, North Africa; EGY, Egypt; NE-AFRICA, Northern-East Africa; E-AFRICA, East Africa; S-AFRICA, Southern Africa.

Download figure to PowerPoint

Table 4.   Population pairwise distances
 S-West AsiaN-East AfricaSouthern AfricaEast AfricaNorth Africa
  1. Population pairwise distances based on the concatenated mitochondrial sequence (below the diagonal: FST; 139 bp; = 94/915 bp; = 78) and 18 polymorphic microsatellite loci (above the diagonal: FST; = 60). All FSTP-values are significant (< 0.0001). na, not applicable. Populations were defined according BAPS and FCA clustering.

S-West Asia0.2950.305nana
N-East Africa0.930/0.9470.170nana
Southern Africa0.818/0.6890.772/0.806nana
East Africa0.951/0.9510.901/0.9390.724/0.613na
North Africa0.958/na0.930/na0.796/na0.972/na

In addition to the mitochondrial sequences, we analysed 18 polymorphic microsatellite loci. Using solely modern samples, we assessed the genetic variation among the extant populations according to the clustering solutions with BAPS [mtDNA 915 bp, PP = 1; nuclear DNA (nDNA), PP = 1; Fig. 2b]. The three clusters obtained with mtDNA reflected the geographical distributions of the described subspecies A. j. jubatus, A. j. soemmerringii and A. j. venaticus. At the nuclear level we could define two additional clusters, which represent substructuring within the Southern and Northern-East African subspecies, respectively. We obtained similar clustering results when we visualized the phylogenetic relationship of the individual genotypes in a three-dimensional FCA (Fig. 3). The results from the admixture analysis based on 500 simulations from posterior allele frequencies revealed no admixture (all P-values = 1; Fig. 2b) and therefore no evidence of past or present gene flow. The Iranian cheetahs (HE = 0.397) were significantly less variable than the Northern-East (HE = 0.674) and Southern African (HE = 0.698) populations (P < 0.001; Wilcoxon rank-sum test corrected for multiple testing). We detected a significant number of loci with heterozygosity excess under the SMM and TPM models, which is consistent with a recent effective population size decline in the Iranian cheetahs. By contrast, no significant signature of a bottleneck was visible in the Southern and Northern-East African populations (Table 5). The population pairwise FST/RST values showed significant differentiation between the three populations (< 0.0001; Tables 4 and S3) and the amova results indicated that 22.7% of the total variation occurred among the different populations/subspecies. In a NJ tree (Fig. 4a), the bootstrap support for the branch assembling all modern Asiatic cheetahs was 100% (100 iterations). As all specimen of the East African subspecies A. j. raineyi were collected from museum or non-invasively (Table 1), their DNA qualities were not sufficient to retrieve consistent and reliable information over all loci. However, we could obtain nuclear data for seven historical samples (#9, 10, 11, 19, 48, 89, 90; Table 1). When these samples were added to the NJ tree analysis (Fig. 4b) the branch leading to all Southwest Asian samples, which cluster separately from the African individuals (including Libya), had a bootstrap support of 72%.

image

Figure 3.  Three-dimensional factorial correspondence analyses (FCA) of African and Asiatic cheetahs based on 18 microsatellite loci. (a) The population structuring of 60 individuals in three clusters corresponding to their geographical origin is shown. The axes 1–4 explain 27.4% of the variation among the populations. (b, c) FCA graphs considering independently the Southern (= 27) and Northern-East African (= 25) cheetah populations. The subclustering within each population reflects the clusters defined with BAPS (Fig. 2b).

Download figure to PowerPoint

Table 5.   Significance of tests for heterozygosity excess assessed using a Wilcoxon sign-rank test under the SMM and TPM model implemented in Bottleneck 1.2.0.2
 P-value (SMM)P-value (TPM)
S-West Asia0.01330.0107
N-East Africa0.87690.1733
Southern Africa0.97000.2475
image

Figure 4.  Neighbor-joining (NJ) trees displaying African and Asiatic cheetahs in independent branches. The NJ trees are based on the proportion of shared alleles (DPS) between individuals using 18 microsatellite loci amplified (a) in 60 modern (b) plus additional seven historical cheetah samples. Modern (solid lines) and historical (dashed lines) samples are colour-coded according their geographical origin. Only bootstrap values (100 reiterations) above 70% are displayed.

Download figure to PowerPoint

Estimation of divergence time

We estimated the divergence time, first between Asiatic and Southern African cheetahs, which correspond to the central haplogroup in the mtDNA network, and within Africa, between the best-sampled subspecies A. j. soemmeringii and A. j. jubatus. Based on the mitochondrial 915-bp fragment, the DA (4.412) between A. j. venaticus and A. j. jubatus was translated into a population split at 41 900 ya (95% CI = 20 300–153 800). Following Gaggiotti & Excoffier (2000), the divergence between these two populations was estimated at 32 170 ya (95% CI = 15 570–118 020). The divergence time between the African subspecies A. j. soemmeringii and A. j. jubatus was calculated at 66 500 ya (95% CI = 32 200–244 000) and 55 085 ya (95% CI = 26 660–202 100) using DA (6.996) and the method of Gaggiotti & Excoffier (2000), respectively. The demographic modelling with IMa suggested a split between A. j. venaticus and A. j. jubatus at 44 403 ya (90% HPD = 27 420–379 222) and between A. j. soemmeringii and A. j. jubatus at 72 296 ya (90% HPD = 43 928–379 317). The upper bound for the credibility interval is not informative, as it critically depends on the assumed prior for the maximum value of t when the curve slowly decreases to zero after the mode of t. The log-likelihood ratio tests did not reject models with m1 = m2 = 0, which are appropriate for studying the divergence of populations under allopatry (Won et al. 2003). Hence, by setting migration rates to zero, we also applied a conventional isolation model (Wakeley & Hey 1997). The splits A. j. venaticus/A. j. jubatus and A. j. jubatus/A. j. soemmeringii were then estimated at 42 120 ya (90% HPD = 16 295–83 677) and 66 698 ya (90% HPD = 24 067–117 615), respectively.

Using the microsatellite genetic distance (δμ)2 and two human microsatellite mutation rates (2.05 × 10−3 and 5.6 × 10−4) employed by Driscoll et al. (2002), we estimated the split between A. j. venaticus and A. j. jubatus at 6700 and 24 700 ya, respectively. Using an average mammalian microsatellite mutation rate (2.05 × 10−4; Rooney et al. 1999), we estimated this divergence at 67 400 ya. The divergence time between A. j. soemmeringii and A. j. jubatus, using the fastest (human) microsatellite mutation rate, was 3200 ya but rose to 32 400 ya when applying the average mammalian mutation rate. To compare these estimates with another distance method, we used the stepwise-weighted genetic distance DSW, which is based on the allele frequency differences among populations (Shriver et al. 1995). Applying again the two human microsatellite mutation rates and the average mammalian mutation rate, and following Calabrese et al. (2001), we calculated the divergence time between Iranian and Southern African cheetahs at 4700, 17 300 and 47 200 ya, respectively. The fastest mutation rate translated into a divergence time estimate among the African subspecies A. j. soemmeringii and A. j. jubatus of 1600 ya whereas it reached the value of 15 600 ya using the average mammalian mutation rate.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

We investigated the genetic diversity and divergence within and between African and Asiatic cheetahs based on mtDNA and microsatellite data of 94 samples including two mediaeval cheetah bones (9th–10th century ce). In general, our data show that there is a higher genetic variation in the current global cheetah population than previously described (O’Brien et al. 1983; Menotti-Raymond & O’Brien 1993; O’Brien & Johnson 2005). This is due mainly to the fact that we included populations that had never been investigated before. The overall amount of mtDNA nucleotide diversity (π) of 0.66% in cheetahs was higher than observed in tigers (0.18%; Luo et al. 2004) and pumas (0.32%; Culver et al. 2000), similar to jaguars (0.77%; Eizirik et al. 2001), and lower than in leopards (1.21%; Uphyrkina et al. 2001). The total nuclear (microsatellite) diversity (HE = 0.766) is comparable with that of other outbred felid species (Culver et al. 2000; Eizirik et al. 2001; Uphyrkina et al. 2001; Driscoll et al. 2002). This might be explained by our use of highly polymorphic markers (all loci were polymorphic), therefore we also compared a set of 15 nuclear microsatellite loci applied in surveys of cheetahs (this study and Driscoll et al. 2002), domestic cats, pumas and lions (supplemental table 3 in Driscoll et al. 2002), and we found similar genetic diversities (HE values ranging from 0.681 to 0.777; Table S2) in the four felid species. The HE levels in the investigated Southern and Northern-East African cheetah populations (Table 3) were similar to Namibian cheetahs (0.640–0.708) and higher than in the Serengeti population (0.599) (Marker et al. 2008). The lower HE (0.397) observed in the Iranian cheetahs might be the consequence of ancestral population divergence or a recent bottleneck, as we found significant evidence for a recent effective population size reduction in this population (Table 5).

Within the African samples, we recovered the previously described relationship between the sub-Saharan cheetah subspecies A. j. raineyi and A. j. jubatus (Menotti-Raymond & O’Brien 1995; Driscoll et al. 2002). The clustering of some Tanzanian and Kenyan animals together with Southern African cheetahs in the mtDNA BAPS analysis (Fig. 2a) suggests a population in East Africa that might be derived from relatively recent re-colonization events as observed in lions (Antunes et al. 2008). This should be investigated combining mtDNA and nDNA data obtained from additional samples of East African cheetahs. At the nuclear level, we observed substructuring in the Northern-East African and Southern African subspecies (Figs 2b and 3) into two subpopulations, which did not correlate with the geographical origin of the individuals. Weak population structure within the South African cheetahs (Kotze et al. 2008) and a panmictic Namibian population (Marker et al. 2008) have been reported previously. Remarkably, the Northern-East African cheetahs were highly differentiated (nuclear FST = 0.170) from the Southern African individuals and clustered independently (Figs 1 and 2) and monophyletic (Fig. S1). Between the African and Asiatic subspecies, we also discovered great differentiation at both nuclear and mitochondrial levels (Table 4). Similar levels of population/subspecies differentiation were described in leopards (Uphyrkina et al. 2001), pumas (Culver et al. 2000) and lions (Antunes et al. 2008). We could not detect significantly higher differentiation (mitochondrial FST) between African and Asiatic cheetahs than among the African subspecies. This indicates deep phylogeographic structure not only between African and Asiatic cheetahs but also among the African cheetah populations.

It is well documented that imports of tamed hunting cheetahs from Northern-East (Pocock 1939) and East Africa (Divyabhanusinh 2007) into India and the Arabian Peninsula were a regular occurrence during the European colonial era. Given the possibility of interbreeding with African escapees, Asiatic cheetahs were not expected to form a genetically distinct unit. However, hunting cheetahs were highly valued, and there are no known records of individuals (accidentally or intentionally) released into the wild (Divyabhanusinh 2007). Moreover, the species is notoriously difficult to breed in captivity. Except for a single litter born in Akbar’s collection of many thousands of cheetahs the first documented captive birth was at the Philadelphia Zoo in 1956 (Marker & O’Brien 1989; Divyabhanusinh 2007). Therefore, the possibility of a captive, hybrid Asiatic-African population as a source of escapees or releases is very low. In our study, we found no evidence of recent gene flow between these populations (Fig. 2b). In all analyses, including the ones with mediaeval Iranian samples, the Asiatic cheetahs constituted a unique cluster (Figs 1–4) suggesting an apparent monophyly of the Asiatic lineage (Fig. S1). Historical museum specimens from Iran, Iraq and India (#12, 15, 16; Table 1) that could not be distinguished by morphological characteristics from African specimens (Divyabhanusinh 2007) were genetically confirmed as Asiatic individuals.

The clustering results of the Asiatic cheetahs were of particular interest because this population has been proposed to form a single subspecies, A. j. venaticus, with North African cheetahs (Ellerman & Morrison-Scott 1951; Krausman & Morales 2005). Notably, Egypt harboured two genetically distinct populations in the past, as we observed clustering of the two Egyptian samples in different haplogroups (Figs 1b and 2a). One now extinct population in eastern Egypt (Northern Sinai; Saleh et al. 2001; Hoath 2003) could be represented by the historical sample collected by Theodor von Heuglin in the early 1850s, which clustered with samples from Southwest Asian countries (Jordan, Iran, Iraq and Oman). The other specimen collected in western Egypt (Libyan Desert; Osborn & Helmy 1980) shared the same haplotype with cheetahs from North Africa and represented a population that might still exist today in the Libyan Desert (Saleh et al. 2001; Hoath 2003). We could not detect hybridization of Asiatic and African cheetahs in our study (Fig. 2a). This suggests that palaeoclimatic constraints, ecological barriers and/or geographical features prevented past gene flow between the two putative populations of this part of Africa. Genetic separation is also supported by the nuclear NJ tree, as the branch leading to the Jordanian (and all Asiatic) samples, which cluster separately from the Libyan (and all other African) individuals, had a bootstrap support of 72% (Fig. 4b). The classification and taxonomy of North African cheetahs are still debated (Krausman & Morales 2005; Belbachir 2007). This population might be genetically contiguous with cheetahs from West Africa (Senegal to Niger), thus it is critical to further investigate current Egyptian and West African populations as suggested by Belbachir (2007). In the light of our results, the previous proposal of a single subspecies, A. j. venaticus, encompassing the Iranian cheetah and its North African congeners (Ellerman & Morrison-Scott 1951; Krausman & Morales 2005) is not supported. In summary, our data based on palaeogenetic analyses demonstrate that the isolation between Asiatic and African cheetahs has existed for millennia.

To date, divergence time between cheetahs has only been estimated among the (closely related; Figs 1 and 2) African subspecies A. j. jubatus and A. j. raineyi. Using mtDNA (mtRFLP) and microsatellite distance data [(δμ)2], the divergence time had been estimated at 28 000–36 000 ya (Menotti-Raymond & O’Brien 1993) and 4253 ya (Driscoll et al. 2002), respectively. The latter was inferred with mutation rates estimated from human microsatellite data (Driscoll et al. 2002). In general, time estimations based on microsatellite distance data can be challenging, particularly if no taxon-specific microsatellite evolution rates are available, as it is the case in felids. Also, it is important to take into account potential homoplasy of microsatellites (Paetkau et al. 1997; Zhivotovsky 2001; Estoup et al. 2002) in the estimation of divergence time between ancient isolates, as some cheetah subspecies are suggested to be by our mtDNA divergence time estimates. In this study, we included two newly investigated subspecies (A. j. soemmeringii and A. j. venaticus) and an average mammalian microsatellite mutation rate previously applied in other felid species (Spong et al. 2000; Anderson et al. 2004; Ruiz-Garcia et al. 2006) to calculate the divergence times within and between African and Asiatic cheetahs. Depending on the genetic marker (mtDNA or nDNA), the nuclear genetic distance [(δμ)2 or DSW] and the choice of the microsatellite mutation rate (human or average mammalian), we retrieved results differing by more than one order of magnitude. Large differences between mitochondrial and nuclear time estimates using human microsatellite rates have been previously observed in wild felids (Menotti-Raymond & O’Brien 1993; Antunes et al. 2008). This could be explained by the fact that the genetic distances (δμ)2 and DSW are considered to underestimate divergence (Paetkau et al. 1997; Calabrese et al. 2001; Zhivotovsky 2001). In our data, we found higher genetic differentiation at microsatellite loci, as measured by FST values and genetic distances [(δμ)2 or DSW], between A. j. venaticus and A. j. jubatus than between A. j. soemmeringii and A. j. jubatus. However, this might be due to a possible stochastic increase in divergence associated with a recent population bottleneck (Chakraborty & Nei 1977; Hedrick 1999), which signature we could apparently detect in our data in A. j. venaticus. Choosing the average mammalian mutation rate and considering the mtDNA estimate, we can place the split between Asiatic and African cheetahs at 32 000–67 000 ya and within Africa, between A. j. soemmeringii and A. j. jubatus, at 16 000–72 000 ya. However, considering the substantial variation in divergence time estimates we acknowledge that decisions for the conservation of this endangered species should not be based on time estimates alone.

Implications for conservation

In this study, we re-visited the currently recognized cheetah subspecies (Krausman & Morales 2005; Durant et al. 2008) in light of our results retrieved from geographically defined populations. We verified the veracity of A. j. venaticus and A. j. soemmeringii based on recognizable phylogenetic partitioning (mitochondrial monophyly and significant divergence at nuclear loci; Moritz 1994) and absence of gene flow (O’Brien & Mayr 1991). Our study suggests a close relationship of A. j. raineyi with A. j. jubatus; however, minisatellite (Menotti-Raymond & O’Brien 1993) and microsatellite variation (Driscoll et al. 2002) support the separation of these two sub-Saharan subspecies. We also clarified the western range limit of the critically endangered A. j. venaticus observing a historical range in contrast to recent accounts, which included North Africa (Krausman & Morales 2005).

The identification of a subspecies recognizes biological distinctiveness and should be sufficient as first-order systematic hypothesis when the aim of conservation is to preserve biological diversity (Green 2005). As large-scale genomic information becomes available also for nonmodel species, adaptive genetic markers might be used to estimate diversification and adaptive genetic variation in subspecies/populations, in combination with supposedly neutrally evolving loci. It might also help to understand how populations can survive despite a low neutral genetic variation (Fraser & Bernatchez 2001; Gebremedhin et al. 2009).

Although there is little evidence that inbreeding depression affects African cheetahs (Caro & Laurenson 1994) and current threats to the species are primarily anthropogenic (Nowell & Jackson 1996; Hunter et al. 2007; CACP 2008; Marker et al. 2008), the lower genetic diversity in the Iranian population is cause for concern in light of their critically low numbers (Hunter et al. 2007; Durant et al. 2008). Any further decline in Iranian cheetah numbers would require increasingly extreme conservation measures, including the consideration of supplemental introductions from Africa, similar to that required for the demographic rescue of Florida panthers (Creel 2006; Pimm et al. 2006; Johnson et al. 2010). However, contrary to the close geographic and genetic distances described in Texas and Florida panthers (Johnson et al. 2010), we did not detect historical gene flow between African and Asiatic cheetahs (Fig. 2c). In addition to the formidable logistical and financial obstacles arguing against introductions, our results emphasize the importance of preserving the genetic distinctiveness of the critically endangered Iranian cheetahs. That will also entail a stronger understanding of possible substructuring in the Iranian population. At least 10 cheetahs moving between known population centres have been killed on roads since 2004 including most recently a female with two cubs in August 2010 (A. Jourabchian, unpublished data). Ongoing, rapid infrastructural development around some cheetah subpopulations in Iran will increase the likelihood of demographic and genetic fragmentation. These processes are currently poorly understood but are the focus of an ongoing multinational research effort led by Iranian biologists which has deployed GPS collars on cheetahs and is undertaking further analysis of genetic differences between cheetah subpopulations in Iran (Hunter et al. 2007). Most significantly, the government of Iran recently renewed its commitment to a major conservation effort of the species (Breitenmoser et al. 2010).

Our results have particular implications for proposed reintroductions in the cheetah’s former range in Asia, especially in India (Ranjitsinh & Jhala 2010). Such endeavours face massive challenges of habitat and prey availability but assuming these are overcome, the question of the founder’s origin remains. The genetic distinctiveness of Asiatic cheetahs would argue that reintroduction efforts should attempt to use cheetahs from Iran. However, this population is critically endangered and cannot sustain removals. Both Southern and Northern-East African cheetahs would have sufficient genetic variability to be considered as independent source populations. Thus, the choice of the most promising source population should be based on ecological, behavioural and viability criteria with minimum taxonomic swamping. In any case, the current Iranian cheetahs will probably remain the only representatives of the subspecies A. j. venaticus in Asia for the foreseeable future.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

Current conservation and management strategies are usually based on the recognition of the subspecies taxonomy. In the case of the cheetah this has rarely been considered, as cheetahs were found to have little genetic variation (Menotti-Raymond & O’Brien 1995; O’Brien & Johnson 2005). Our data suggest this viewpoint to be valid for the two sub-Saharan subspecies A. j. jubatus and A. j. raineyi, which could not be entirely separated in the mitochondrial population structure analysis. However, we show that Northern-East African, Southern African and Asiatic cheetahs are long-term geographic isolates with independent evolutionary histories. Moreover, we demonstrate that the critically endangered Iranian cheetahs are an autochthonous, monophyletic population and the last representatives of the Asiatic cheetah. Our data also support the view of an independent subspecies status for the cheetahs in North Africa. This population may be genetically contiguous with those from West Africa (Senegal–Niger), historically classified as A. j. hecki, but additional sampling is required to resolve this issue (Belbachir 2007). Also, it will be important to survey adaptive genetic variation in the cheetah subspecies to better understand evolutionary differentiation caused by ecological adaptation. Based on our results, we conclude that unique diversity remains in the cheetahs of Africa and Asia and that conservation of these populations, especially of the critically low numbering Iranian individuals, should rank high among felid conservation priorities.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

We thank the curators of the Natural History Museums of London, Vienna, Berlin, Stuttgart and Senckenberg-Frankfurt, the Alexander Koenig Research Museum, the Harrison Institute, the Royal Museum of Scotland and the Royal Museum for Central Africa (Tervuren). We appreciate the support from the veterinarians and keepers of DECAN Rescue Centre Djibouti, Vienna Zoo, Salzburg Zoo, La Palmyre Zoo and the Zoological Garden of Montpellier. We are most grateful to M. Kamyab (UNDP-Tehran), M. Müller (Institute of Animal Breeding and Genetics-vetmeduni Vienna), F. Claro and N. Vidal (MNHN-Paris), J. Corander for comments on BAPS, A. McGregor for language corrections, and Y. Moodley for comments on the manuscript. P.C. and P.B. acknowledge support from the Austrian Science Foundation (FWF; P1084-B17 to P.B.) and C.F. from the Portuguese Science Foundation (FCT-MCTES; contract C2007-UL-342-CBA1). The work was supported by the Felidae Conservation Fund, Panthera and UNDP-GEF (fieldwork in Iran), the French National Research Centre (palaeogenetic analysis at IJM, Paris) and the National Research Foundation of South Africa.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information
  • Akaike H (1974) A new look at the statistical model identification. IEEE Transactions on Automatic Control, 19, 716723.
  • Allendorf FW, Luikart G (2007) Conservation and the Genetics of Populations. Blackwell Publishing Ltd, Oxford, 642 pp.
  • Allsen TT (2006) Natural history and cultural history—the circulation of hunting leopards in Eurasia, seventh-seventeenth centuries. In: Contact and Exchange in the Ancient World (ed. MairVT), 310 pp. University of Hawai’i Press, Honolulu.
  • Anderson CR, Lindzey FG, McDonald DB (2004) Genetic structure of cougar populations across the Wyoming basin: metapopulation or megapopulation. Journal of Mammalogy, 85, 12071214.
  • Antunes A, Troyer JL, Roelke ME et al. (2008) The evolutionary dynamics of the lion Panthera leo revealed by host and viral population genomics. PLoS Genetics, 4, e1000251.
  • Bandelt H, Forster P, Rohl A (1999) Median-joining networks for inferring intraspecific phylogenies. Molecular Biology and Evolution, 16, 3748.
  • Belbachir F (2007) Les grands questions relative a la conservation des grands felins d’Algerie: cas du guepard et du leopard. In: Compte-rendu de la deuxième réunion de l’observatoire du guépard en régions d’Afrique du nord—Tamanrasset (ed. OGRAN), 42 pp. Société Zoologique de Paris, Paris.
  • Belkir K, Borsa P, Goudet J, Chikhi L, Bonhomme F (1999) Genetix, logiciel sous Windows TM pour la ge′ne′tique des populations. Laboratoire Ge′nome et Populations, CNRS UPR 9060, Universite′ de Montpellier II, Montpellier, France.
  • Boessneck J, Krauß R (1973) Die Tierwelt um Bas-tam/Nordwest-Azerbaidjan. Archäologische Mitteilungen aus Iran NF, 6, 113133.
  • Bowcock AM, Ruiz-Linares A, Tomfohrde J, Minch E, Kidd JR, Cavalli-Sforza LL (1994) High resolution human evolutionary trees with polymorphic microsatellites. Nature, 368, 455457.
  • Brehm AE (1879) Brehms Thierleben. Allgemeine Kunde des Thierreichs. Verlag des Bibliographischen Instituts, Leipzig.
  • Breitenmoser U, Breitenmoser-Würsten C, von Arx M (2010) Workshop on the conservation of the Asiatic cheetah. Cat News, 52, 17.
  • Burger P, Steinborn R, Walzer C, Petit T, Mueller M, Schwarzenberger F (2004) Analysis of the mitochondrial genome of cheetahs (Acinonyx jubatus) with neurodegenerative disease. Gene, 338, 111119.
  • Busby GBJ, Gottelli D, Wacher T et al. (2009) Genetic analysis of scat reveals leopard Panthera pardus and cheetah Acinonyx jubatus in southern Algeria. Oryx, 43, 412415.
  • CACP (2008) Conservation of the Asiatic Cheetah, Its Natural Habitats and Associated Biota in the I.R. of Iran. Project Number IRA/00/G35. Final Report. CACP, Tehran, Iran, 44 pp.
  • Calabrese PP, Durrett RT, Aquadro CF (2001) Dynamics of microsatellite divergence under stepwise mutation and proportional slippage/point mutation models. Genetics, 159, 839852.
  • Caro T, Laurenson M (1994) Ecological and genetic factors in conservation: a cautionary tale. Science, 263, 485486.
  • Chakraborty R, Nei M (1977) Bottleneck effects on average heterozygosity and genetic distance with the stepwise mutation model. Evolution, 31, 347356.
  • Corander J, Marttinen P (2006) Bayesian identification of admixture events using multilocus molecular markers. Molecular Ecology, 15, 28332843.
  • Corander J, Tang J (2007) Bayesian analysis of population structure based on linked molecular information. Mathematical Biosciences, 205, 1931.
  • Corander J, Marttinen P, Siren J, Tang J (2008) Enhanced Bayesian modelling in BAPS software for learning genetic structures of populations. BMC Bioinformatics, 9, 539.
  • Creel S (2006) Recovery of the Florida panther—genetic rescue, demographic rescue, or both? Animal Conservation, 9, 125126.
  • Culver M, Johnson WE, Pecon-Slattery J, O’Brien SJ (2000) Genomic ancestry of the American puma (Puma concolor). Journal of Heredity, 91, 186197.
  • Dieringer D, Schlötterer C (2003) Microsatellite analyser (MSA): a platform independent analysis tool for large microsatellite data sets. Molecular Ecology Notes, 3, 167169.
  • Divyabhanusinh C (2007) The End of a Trail, 3rd edn. Oxford University Press, Oxford, 307 pp.
  • Driscoll CA, Menotti-Raymond MA, Nelson G, Goldstein D, O’Brien SJ (2002) Genomic microsatellites as evolutionary chronometers: a test in wild cats. Genome Research, 12, 414423.
  • Durant S, Marker L, Purchase N et al. (2008) In: IUCN Red List of Threatened Species: Acinonyx jubatus. Version 2010.1. Available from http://www.iucnredlist.org (accessed 22 March 2010).
  • Eizirik E, Kim JH, Menotti-Raymond M, Crawshaw PG Jr, O’Brien SJ, Johnson WE (2001) Phylogeography, population history and conservation genetics of jaguars (Panthera onca, Mammalia, Felidae). Molecular Ecology, 10, 6579.
  • Ellerman J, Morrison-Scott T (1951) Checklist of Paleartic and Indian Mammals, 1758–1946, 1st edn. British Museum of Natural History, London, 810 pp.
  • Estoup A, Jarne P, Cornuet JM (2002) Homoplasy and mutation model at microsatellite loci and their consequences for population genetics analysis. Molecular Ecology, 11, 15911604.
  • Excoffier L, Lischer H (2010) Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows. Molecular Ecology Resources, 10, 564567.
  • Farhadinia M (2004) The last stronghold: cheetah in Iran. Cat News, 40, 1114.
  • Felsenstein J (1985) Confidence-limits on phylogenies with a molecular clock. Systematic Zoology, 34, 152161.
  • Felsenstein J (1989) PHYLIP-phylogeny inference package (version 3.2). Cladistics, 5, 164166.
  • Frankham R, Ballou JD, Briscoe DA (2009) Introduction to Conservation Genetics, 2nd edn. Cambridge University Press, Cambridge, UK, 704 pp.
  • Fraser D, Bernatchez L (2001) Adaptive evolutionary conservation: towards a unified concept for defining conservation units. Molecular Ecology, 10, 27412752.
  • Freeman A, Machugh D, Mckeown S, Walzer C, Mcconnell D, Bradley DG (2001) Sequence variation in the mitochondrial DNA control region of wild African cheetahs (Acinonyx jubatus). Heredity, 86, 355362.
  • Gaggiotti O, Excoffier L (2000) A simple method of removing the effect of a bottleneck and unequal population sizes on pairwise genetic distances. Proceedings of the Royal Society B: Biological Sciences, 267, 8187.
  • Gebremedhin B, Ficetola GF, Naderi S et al. (2009) Frontiers in identifying conservation units: from neutral markers to adaptive genetic variation. Animal Conservation, 12, 107109.
  • Goldstein DB, Pollock DD (1997) Launching microsatellites: a review of mutation processes and methods of phylogenetic interference. Journal of Heredity, 88, 335342.
  • Goldstein DB, Linares AR, Cavalli-Sforza LL, Feldman MW (1995) An evaluation of genetic distances for use with microsatellite loci. Genetics, 139, 463471.
  • Green DM (2005) Designatable units for status assessment of endangered species. Conservation Biology, 19, 18131820.
  • Hasegawa M, Kishino H, Yano T-a (1985) Dating of the human-ape splitting by molecular clock of mitochondrial DNA. Journal of Molecular Evolution, 22, 160174.
  • Hedrick PW (1999) Perspective: highly variable loci and their interpretation in evolution and conservation. Evolution, 53, 313318.
  • Hey J, Nielsen R (2007) Integration within the Felsenstein equation for improved Markov chain Monte Carlo methods in population genetics. Proceedings of the National Academy of Sciences, USA, 104, 27852790.
  • Hoath R, ed. (2003) Cheetah. In: A Field Guide to the Mammals of Egypt, pp. 104105. The American University in Cairo Press, Cairo.
  • Hunter L, Hamman D (2003) Cheetah. Struik-New Holland, Cape Town, 144 pp.
  • Hunter L, Jowkar H, Ziai H et al. (2007) Conserving the Asiatic Cheetah in Iran: launching the First Radio-Telemetry Study. Cat News, 46, 811.
  • Jobb G, von Haeseler A, Strimmer K (2004) Treefinder: a powerful graphical analysis environment for molecular phylogenetics. BMC Evolutionary Biology, 4, 18.
  • Johnson WE, Godoy J, Palomares F et al. (2004) Phylogenetic and phylogeographic analysis of Iberian lynx populations. Journal of Heredity, 95, 1928.
  • Johnson WE, Eizirik E, Pecon-Slattery J et al. (2006) The Late Miocene radiation of modern Felidae: a genetic assessment. Science, 311, 7377.
  • Johnson WE, Onorato DP, Roelke ME et al. (2010) Genetic restoration of the Florida panther. Science, 329, 16411645.
  • Kolb R (1972) Die Tierknochenfunde vom Takht-i Suleiman in der iranischen Provinz Aserbeidschan (Fundmaterial der Grabung 1969). Doctorate Thesis, Ludwig-Maximilian University, Munich.
  • Kotze A, Ehlers K, Cilliers D, Grobler J (2008) The power of resolution of microsatellite markers and assignment tests to determine the geographic origin of cheetah (Acinonyx jubatus) in Southern Africa. Mammalian Biology, 73, 457462.
  • Krausman P, Morales S (2005) Acinonyx jubatus. Mammalian Species, 771, 16.
  • Krauß R (1975) Tierknochenfunde aus Bastam in Nordwest-Azerbaidjan/Iran. Doctorate Thesis, Ludwig-Maximilian University, Munich.
  • Langella M (1999) Populations 1.2.30: Population genetic software (individuals or population distances, phylogenetic trees). Available from http://bioinformatics.org/~tryphon/populations/ (accessed 15 July 2010).
  • Laurenson M (1994) High juvenile mortality in cheetahs (Acinonyx jubatus) and its consequences for maternal care. Journal of Zoology (London), 234, 387408.
  • Lopez J, Cevario S, O’Brien SJ (1996) Complete nucleotide sequence of the domestic cat (Felis catus) mitochondrial genome and a transposed mtDNA tandem repeat (Numt) in the nuclear genome. Genomics, 33, 229246.
  • Luo SJ, Kim JH, Johnson WE et al. (2004) Phylogeography and genetic ancestry of tigers (Panthera tigris). PLoS Biology, 2, 22752293.
  • Manati AR, Nogge G (2008) Cheetahs in Afghanistan. Cat News, 49, 18.
  • Marker L (2009) International Cheetah (Acinonyx jubatus) Studbook 2007. Cheetah Conservation Fund, Otjiwarongo, Namibia.
  • Marker L, O’Brien SJ (1989) Captive breeding of the cheetah (Acinonyx jubatus) in north-American zoos (1871–1986). Zoo Biology, 8, 316.
  • Marker L, Wilkerson A, Sarno R et al. (2008) Molecular genetic insights on cheetah (Acinonyx jubatus) ecology and conservation in Namibia. Journal of Heredity, 99, 213.
  • Masseti M (2009) Pictorial evidence from medieval Italy of cheetahs and caracals, and their use in hunting. Archives of Natural History, 36, 3747.
  • May RM (1995) Population genetics. The cheetah controversy. Nature, 374, 309310.
  • Menotti-Raymond MA, O’Brien SJ (1993) Dating the genetic bottleneck of the African cheetah. Proceedings of the National Academy of Sciences, USA, 90, 31723176.
  • Menotti-Raymond MA, O’Brien SJ (1995) Evolutionary conservation of ten microsatellite loci in four species of Felidae. Journal of Heredity, 86, 319322.
  • Menotti-Raymond MA, David VA, Lyons LA et al. (1999) A genetic linkage map of microsatellites in the domestic cat (Felis catus). Genomics, 57, 923.
  • Moritz C (1994) Defining ‘evolutionary significant units’ for conservation. Trends in Ecology and Evolution, 9, 373375.
  • Nei M (1987) Molecular Evolutionary Genetics. Columbia University Press, New York, 512 pp.
  • Nei M, Li W (1979) Mathematical model for studying genetic variation in terms of restriction endonucleases. Proceedings of the National Academy of Sciences, USA, 76, 52695273.
  • Nielsen R, Wakeley J (2001) Distinguishing migration from isolation: a Markov chain Monte Carlo approach. Genetics, 158, 885896.
  • Nowell K, Jackson P (1996) Status Survey and Conservation Action Plan—Wild Cats. IUCN, Gland, 384 pp.
  • Nsubuga AM, Robbins MM, Roeder AD, Morin PA, Boesch C, Vigilant L (2004) Factors affecting the amount of genomic DNA extracted from ape faeces and the identification of an improved sample storage method. Molecular Ecology, 13, 20892094.
  • O’Brien SJ, Johnson WE (2005) Big cat genomics. Annual Reviews of Genomics and Human Genetics, 6, 407429.
  • O’Brien SJ, Mayr E (1991) Bureaucratic mischief: recognizing endangered species and subspecies. Science, 251, 11871188.
  • O’Brien SJ, Wildt D, Goldman D, Merril C, Bush M (1983) The cheetah is depauperate in genetic variation. Science, 221, 459462.
  • O’Brien SJ, Wildt D, Bush M et al. (1987) East African cheetahs: evidence for two population bottlenecks? Proceedings of the National Academy of Sciences, USA, 84, 508511.
  • Ohta T, Kimura M (1973) A model of mutation appropriate to estimate the number of electrophoretically detectable alleles in a genetic population. Genetical Research, 22, 201204.
  • Osborn DJ, Helmy I (1980) The Contemporary Land Mammals of Egypt (Including Sinai). Field Museum of Natural History, Chicago, IL, 579 pp.
  • Paetkau D, Waits LP, Clarkson PL, Craighead L, Strobeck C (1997) An empirical evaluation of genetic distance satistics using microsatellite data from bear (Ursidae) populations. Genetics, 147, 19431957.
  • Pfeiffer I, Volkel I, Taubert H, Brenig B (2004) Forensic DNA-typing of dog hair: DNA-extraction and PCR amplification. Forensic Science International, 141, 149151.
  • Pimm SL, Dollar L, Bass OL Jr (2006) The genetic rescue of the Florida panther. Animal Conservation, 9, 115122.
  • Piry S, Luikart G, Cornuet J-M (1999) BOTTLENCK: a computer program for detecting recent reductions in the effective population size using allele frequency data. Journal of Heredity, 90, 502503.
  • Pocock RI (1939) The Fauna of British India, Including Ceylon and Burma. Taylor and Francis, London, 549 pp.
  • Posada D, Crandall KA (1998) Modeltest: testing the model of DNA substitution. Bioinformatics, 14, 817818.
  • Pruvost M, Geigl E-M (2004) Real-time quantitative PCR to assess the authenticity of ancient DNA amplification. Journal of Archaeological Science, 31, 11911197.
  • Pruvost M, Grange T, Geigl E-M (2005) Minimizing DNA contamination by using UNG-coupled quantitative real-time PCR on degraded DNA samples: application to ancient DNA studies. BioTechniques, 38, 569575.
  • Pruvost M, Schwarz R, Correia VB et al. (2007) Freshly excavated fossil bones are best for amplification of ancient DNA. Proceedings of the National Academy of Sciences, USA, 104, 739744.
  • R Development Core Team (2009) R: A Language and Environment for Statistical Computing. R.2.10.1. R Foundation for statistical computing, Vienna.
  • Rambaut A (2009) FigTree v1.3.1. Institute of Evolutionary Biology, University of Edinburgh, Edinburgh.
  • Ranjitsinh MK, Jhala YV (2010) Assessing the Potential for Reintroducing the Cheetah in India. Wildlife Trust of India, Noida, & Wildlife Institute of India TR2010/001, Dehradun, India, 177 pp.
  • Raymond M, Rousset F (1995) An exact test for population differentiation. Evolution, 49, 12801283.
  • Rooney AP, Honeycutt RL, Davis SK, Derr JN (1999) Evaluating a putative bottleneck in a population of bowhead whales from patterns of microsatellite diversity and genetic disequilibria. Journal of Molecular Evolution, 49, 682690.
  • Ruiz-Garcia M, Payán E, Murillo A, Alvarez D (2006) DNA microsatellite characterization of the jaguar (Panthera onca) in Colombia. Genes & Genetic System, 81, 1527.
  • Saleh MA, Helmy I, Giegengack R (2001) The cheetah, Acinonyx jubatus (Schreber, 1776) in Egypt (Felidae, Acinonychinae). Mammalia, 65, 177194.
  • Schmidt HA, Petzold E, Vingrom M, von Haeseler A (2003) Molecular phylogenetics: parallelized parameter estimation and quartet puzzling. Journal of Parallel and Distributed Computing, 63, 719727.
  • Shriver MD, Jin L, Boerwinkle E, Deka R, Chakraborty R (1995) A novel measure of genetic distance for highly polymorphic tandem repeat loci. Molecular Biology and Evolution, 12, 914920.
  • Spong G, Johansson M, Björklund M (2000) High genetic variation in leopards indicates large and long-term stable effective population size. Molecular Ecology, 9, 17731782.
  • Tajima F (1983) Evolutionary relationship of DNA sequences in finite populations. Genetics, 105, 437460.
  • Tamura K, Nei M (1993) Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees. Molecular Biology and Evolution, 10, 512526.
  • Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4: molecular evolutionary genetics analysis (MEGA) software version 4.0. Molecular Biology and Evolution, 24, 15961599.
  • Uphyrkina O, Johnson WE, Quigley H et al. (2001) Phylogenetics, genome diversity and origin of modern leopard, Panthera pardus. Molecular Ecology, 10, 26172633.
  • Vasan S, Zhang X, Kapurniotu A et al. (1996) An agent cleaving glucose-derived protein crosslinks in vitro and in vivo. Nature, 382, 275278.
  • Wakeley J, Hey J (1997) Estimating ancestral population parameters. Genetics, 145, 847855.
  • Wisely S, Maldonado J, Fleischer R (2004) A technique for sampling ancient DNA that minimizes damage to museum specimens. Conservation Genetics, 5, 105107.
  • Won Y, Young CR, Lutz RA, Vrijenhoek RC (2003) Dispersal barriers and isolation among deep-sea mussel populations (Mytilidae: Bathymodiolus) from eastern Pacific hydrothermal vents. Molecular Ecology, 12, 169184.
  • Zhivotovsky LA (2001) Estimating divergence time with the use of microsatellite genetic distances: impacts of population growth and gene flow. Molecular Biology and Evolution, 18, 700709.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

Table S1 Mitochondrial fragments amplified for aDNA analysis

Table S2 Comparisons of average numbers of alleles per locus and expected heterozygosities calculated from 15 microsatellite loci overlapping in four different felid species.

Table S3 Population pairwise distances (RST) based on 18 polymorphic microsatellite loci.

Fig. S1 Phylogenetic relationship based on a ML tree inferred from the 915-bp concatenated mtDNA fragment.

FilenameFormatSizeDescription
MEC_4986_sm_SupportingInformation.doc98KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.