Figure 1
Geographic distribution of the Pleistocene cave lion. Origins of successfully amplified samples from this study (black circles), and from the study by Barnett et al. (2009) (white circles). The three geographic regions are highlighted; Europe (blue), Siberia (red) and Alaska plus Yukon (yellow).
Table 1
Population summary statistics. Nucleotide diversity, π; gene diversity, Hd; number of polymorphic sites, S; and Tajima’s D, Td. Standard deviations of nucleotide diversity and gene diversity estimates are provided between parentheses. CR: Control Region. Time Period is given in thousands (K) of calendar years before present. *: sample EE26 was excluded for estimating the different summary statistics as the sequence recovered was only partial. NA: Not determined as the number of sequences available was not sufficient.
| Gene partition | Population | Time Period | N | Length (bp) | π | Hd | Td | S |
|---|---|---|---|---|---|---|---|---|
| CR | Europe | 52–16K | 8 | 205 | 0.01167 (0.00218) | 0.929 (0.084) | -0.5409 | 7 |
| Siberia | 33–14K | 10 | 206 | 0.00367 (0.00157) | 0.551 (0.164) | -1.0345 | 3 | |
| Siberia | 62–48K* | 13 | 206 | 0.00884 (0.00190) | 0.872 (0.067) | -1.138 | 8 | |
| Alaska | 22–14K | 9 | 204 | 0.00191 (0.00081) | 0.389 (0.164) | 0.1565 | 1 | |
| P. atrox | 33–13K | 3 | 195 | 0.01368 (0.00645) | 0.667 (0.314) | NA | 4 | |
| P. leo | Modern | 16 | 201 | 0.00829 (0.00160) | 0.833 (0.072) | -0.7398 | 7 | |
| ATP8 | Europe | 52–16K | 3 | 143 | 0.01399 (0.00491) | 1.000 (0.272) | NA | 3 |
| Siberia | 33–14K | 9 | 143 | 0.01049 (0.00478) | 0.583 (0.183) | -1.3984 | 6 | |
| Siberia | 62–41K | 14 | 143 | 0.00200 (0.00166) | 0.143 (0.119) | -1.4807 | 2 | |
| Alaska | 22–14K | 3 | 143 | 0 | 0 | NA | 0 | |
| P. leo | Modern | 3 | 143 | 0.01399 (0.00491) | 1.000 (0.272) | NA | 3 | |
| CR + ATP8 | Europe | 52–16K | 3 | 349 | 0.01146 (0.00403) | 1.000 (0.272) | NA | 6 |
| Siberia | 33–14K | 9 | 349 | 0.00669 (0.00233) | 0.806 (0.120) | -1.3594 | 9 | |
| Siberia | 62–41K | 11 | 349 | 0.00636 (0.00173) | 0.873 (0.089) | -1.4912 | 10 | |
| Alaska | 22–14K | 3 | 348 | 0.00192 (0.00090) | 0.667 (0.314) | NA | 1 | |
| P. leo | Modern | 3 | 350 | 0.01143 (0.00381) | 1.000 (0.272) | NA | 6 |
Figure 2
Population models tested using serial-coalescent simulations. Letters indicate the parameters sampled from priors distributions (other than mutation rates): T, M, I, and Ne, all sampled from uniform distributions. Dashed line indicates the timing (t) of important events: sudden change in population size (priors set to 11.2-25,000 cal a BP in the most recent event of model 3, and 25-60,000 cal a BP for the oldest event in model 3 and events in models 2, 7 and 8), stop in migration (coinciding with times of demographic change), and geographic split (60-75,000 cal a BP in models 5 to 8). Arrows indicate migration (M) set with priors ranging from 0 to 1% of migrants per generation while I is the intensity of the demographic change set with a prior between 1 to 10 times the size of the previous or posterior population change. Support values are listed below each model, for the two calibration rates used. Maximal marginal likelihood is reported with an asterisk.
Figure 3
Posterior distributions of the parameters in model 3. The posterior distributions were estimated using an approximate Bayesian computation procedure and a mutation rate resulting from a tip and split calibration with BEAST v1.6.1. Prior distributions are shown as bars and dotted lines indicate 95 % credibility intervals. a) Female effective size (Nfe) of the population following the demographic bottleneck (number of haploid gene copies). b) Magnitude of the demographic change of each event as a ratio of the larger population size to the smaller one; decline (right curve) and recovery (left curve). c) Timing of the demographic decline (right curve) and recovery (left curve).
Figure 4
Haplotypes in network and temporal graph. a) Median-joining network analysis based on the control region sequence, including samples from Barnett et al. (2009). Letters represent haplotypes given in table 1 and colors correspond to geographic region. Small crossing bars represent number of mutations between haplotypes. Since haplotype Y5 is based on an incomplete sequence it is connected with a dashed line. b) Temporal arrangement of samples (including those from the study of Barnett et al. (2009)) and their respective haplotypes. The estimated bottleneck in Beringia is shaded and delimited by the mode values. Dotted lines show 95% credibility intervals. Extended and faded timelines have been added for samples lacking a finite date.
Table 2
Support for different population demographic models using BEAST and tip calibration. The marginal log likelihood of each model, as well as Bayes Factors (logBF) between alternative models and best models are shown. Different sets of sequences originating from restricted or large geographic regions were considered in order to test for population structure as a possible confounding factor.
| Populations | Model | log Likelihood | logBF |
|---|---|---|---|
| Siberia | Constant | -375.5 (0.144) | -0.355 |
| Siberia | Skyline | -376.0 (0.142) | -0.553 |
| Siberia | Skyride | -374.7 (0.121) | * |
| Siberia + Alaska | Constant | -385.3 (0.145) | -0.258 |
| Siberia + Alaska | Skyline | -384.7 (0.133) | * |
| Siberia + Alaska | Skyride | -385.0 (0.135) | -0.099 |
| Siberia + Alaska + Europe | Constant | -434.4 (0.183) | -0.433 |
| Siberia + Alaska + Europe | Skyline | -433.4 (0.177) | * |
| Siberia + Alaska + Europe | Skyride | -433.7 (0.189) | -0.128 |
Table 3
Parameter estimations of model 3 using ABC. Estimations of: effective population size after the bottleneck (Ne1), magnitude of expansion (Ne1/Ne2), magnitude of decline (Ne3/Ne2), time (cal a BP) to the end of the decline (T1), and time to the beginning of the decline (T2).
| Mode | Median | 95% C.I. | ||
|---|---|---|---|---|
| Ne1 | 1000 | 1476 | 129.64 | 8658.4 |
| Ne1/Ne2 | 0.15 | 0.2796 | 0.0432 | 0.9933 |
| Ne3/Ne2 | 0.05 | 0.0633 | 0.0147 | 0.4392 |
| T1 | 18000 | 17481 | 12252 | 24507 |
| T2 | 46500 | 41603 | 28482 | 54754 |