Early Pleistocene origin and extensive intra-species diversity of the extinct cave lion

www.nature.com/scientificreports OPEN Early Pleistocene origin and extensive intra‑species diversity of the extinct cave lion David W. G. Stanton1,2, Federica Alberti3,4, Valery Plotnikov5, Semyon Androsov6, Semyon Grigoriev7,16, Sergey Fedorov7, Pavel Kosintsev8, Doris Nagel9, Sergey Vartanyan10, Ian Barnes11, Ross Barnett12, Erik Ersmark1,2, Doris Döppes4, Mietje Germonpré13, Michael Hofreiter3, Wilfried Rosendahl4, Pontus Skoglund14 & Love Dalén1,2,15* The cave lion is an extinct felid that was widespread across the Holarctic throughout the Late Pleistocene. Its closest extant relative is the lion (Panthera leo), but the timing of the divergence between these two taxa, as well as their taxonomic ranking are contentious. In this study we analyse 31 mitochondrial genome sequences from cave lion individuals that, through a combination of 14C and genetic tip dating, are estimated to be from dates extending well into the mid-Pleistocene. We identified two deeply diverged and well-supported reciprocally monophyletic mitogenome clades in the cave lion, and an additional third distinct lineage represented by a single individual. One of these clades was restricted to Beringia while the other was prevalent across western Eurasia. These observed clade distributions are in line with previous observations that Beringian and European cave lions were morphologically distinct. The divergence dates for these lineages are estimated to be far older than those between extant lions subspecies. By combining our radiocarbon tip-dates with a split time prior that takes into account the most up-to-date fossil stem calibrations, we estimated the mitochondrial DNA divergence between cave lions and lions to be 1.85 Million ya (95% 0.52– 2.91 Mya). Taken together, these results support previous hypotheses that cave lions existed as at least two subspecies during the Pleistocene, and that lions and cave lions were distinct species. The cave lion (Panthera spelaea) was an apex predator across the Holarctic1,2 until their extinction at the end of the Pleistocene3 (last occurrence in the fossil record 14,219 ± 112 cal B P4). Cave lions were larger than extant lions5, and Pleistocene cave art suggests that they did not have manes. However they may have shared several behavioural traits with their modern counterparts, such as group living and courtship rituals6. Cave lion taxonomy has been contentious, being variously considered a subspecies of Panthera leo7,8, a sister species to extant lions (Panthera spelaea)5,9, or even being more closely related to the tiger (Panthera tigris spelaea)10. In particular, the molecular estimate of the timing of the split between cave lions and extant lions has varied considerably between studies (~ 600 kya8; 1.23–2.93 mya11). Ersmark et al.12 identified two major cave lion mitochondrial DNA haplogroups (based on ~ 348 bp of ATP8 and control region sequences) and showed that 1 Centre for Palaeogenetics, Svante Arrhenius väg 20C, 106 91 Stockholm, Sweden. 2Department of Bioinformatics and Genetics, Swedish Museum of Natural History, Stockholm, Sweden. 3Institute for Biochemistry and Biology, University of Potsdam, Karl‑Liebknecht‑Str. 24‑25, 14476 Potsdam, Germany. 4Reiss-Engelhorn-Museen, Zeughaus C5, 68159 Mannheim, Germany. 5Academy of Sciences of Sakha Republic, Lenin Avenue 33, Yakutsk, Sakha Republic (Yakutia), Russia. 6Museum “Severnyi Mir”, Yakutsk, Sakha Republic (Yakutia), Russia. 7Mammoth Museum of Institute of Applied Ecology of the North, North-Eastern Federal University, Yakutsk, Sakha Republic (Yakutia), Russia. 8Institute of Plant and Animal Ecology, Russian Academy of Sciences, 202 Marta 8 St., Ekaterinburg, Russia 620144. 9Department of Paleontology, University of Vienna, Althanstrasse 14, 1090 Vienna, Austria. 10North-East Interdisciplinary Scientific Research Institute n.a. N.A. Shilo FEB RAS (NEISRI FEB RAS), Portovaya Str., 16, Magadan, Russia 685000. 11Department of Earth Sciences, Natural History Museum, London, UK. 12Natural History Museum of Denmark, University of Copenhagen, Copenhagen, Denmark. 13OD Earth and History of Life, Royal Belgian Institute of Natural Sciences, Vautierstraat 29, 1000 Brussel, Belgium. 14The Francis Crick Institute, 1 Midland Road, London NW1 1AT, UK. 15Department of Zoology, Stockholm University, Stockholm, Sweden. 16Semyon Grigoriev is deceased. *email: Scientific Reports | (2020) 10:12621 | https://doi.org/10.1038/s41598-020-69474-1 1 Vol.:(0123456789) www.nature.com/scientificreports/ there was an association between the age of the specimen and its haplogroup, with one of the two haplogroups disappearing ~ 41 kya. Morphological analysis of skulls and mandibles has shown that cave lions from Yakutia, Alaska and the Yukon Territory are smaller than those from Europe, and led to the conclusion that Beringian lions should be recognised as a distinct subspecies “Panthera spelaea vereshchagini n.subsp"13. However, because all previous genetic studies of cave lions have either used only a small mitochondrial fragment or relied on limited sample number, phylogenetic structure within cave lions, and between cave lions and extant lions has remained largely unresolved. In this study, we investigated mitochondrial genome diversity in 31 cave lions from across their entire prehistoric range, and from an even temporal spread between the last occurrence in the fossil record to beyond the limit of radiocarbon dating. We also generated multiple new radiocarbon dates, allowing us to (1) Use genetic data to estimate the age of specimens with ambiguous or infinite radiocarbon dates, (2) Estimate the date of the split between P. leo and P. spelaea, and (3) Investigate intra-species mitochondrial diversity across the entire historical distribution of cave lions. Results/discussion Mitochondrial genome sequences show that cave lions and modern lions fall into two well-supported reciprocally monophyletic clades (Fig. 1; posterior = 1.00). New radiocarbon dates for the cave lions range between 28.0 kya (thousand radiocarbon years before present; ± 110 years) to beyond the radiocarbon limit (Table S1). By combining these 14C dates as tip priors, alongside a TMRCA prior that takes into account fossil calibration, we estimate this divergence between the cave lions and lion clades to be 1.85 mya (million radiocarbon years before present; 95% credibility interval: 0.52–2.91 mya). This date is in agreement with the previous estimate by Barnett et al.11 (1.23–2.93 million years) that incorporate fossil calibrations, rather than only molecular estimates (1 kb of the Cytochrome b gene from two cave lion individuals, ~ 600 kya8; whole genome data from modern lions and two cave lion individuals, ~ 500 k ya14). Using only 14C tip dates to inform the analysis leads to a divergence estimate that is closer to the younger molecular estimates, at 550 kya (0.17–3.96 mya; Figure S2). It is expected that tip dates will often give younger diver (...truncated)