Highlights
-
•
-
First nuclear paleogenome of the giant panda
-
•
-
Represents a population that forms the sister group to all extant populations
-
•
-
Differential admixture between extinct and extant panda populations
-
•
-
Part of the extinct, ancient population survives in the modern gene pool
Summary
Historically, the giant panda was widely distributed from northern China to southwestern Asia [1]. As a result of range contraction and fragmentation, extant individuals are currently restricted to fragmented mountain ranges on the eastern margin of the Qinghai-Tibet plateau, where they are distributed among three major population clusters [2]. However, little is known about the genetic consequences of this dramatic range contraction. For example, were regions where giant pandas previously existed occupied by ancestors of present-day populations, or were these regions occupied by genetically distinct populations that are now extinct? If so, is there any contribution of these extinct populations to the genomes of giant pandas living today? To investigate these questions, we sequenced the nuclear genome of an ∼5,000-year-old giant panda from Jiangdongshan, Tengchong County in Yunnan Province, China. We find that this individual represents a genetically distinct population that diverged prior to the diversification of modern giant panda populations. We find evidence of differential admixture with this ancient population among modern individuals originating from different populations as well as within the same population. We also find evidence for directional gene flow, which transferred alleles from the ancient population into the modern giant panda lineages. A variable proportion of the genomes of extant individuals is therefore likely derived from the ancient population represented by our sequenced individual. Although extant giant panda populations retain reasonable genetic diversity, our results suggest that this represents only part of the genetic diversity this species harbored prior to its recent range contractions.
Figure 2. Relationship of the Ancient Giant Panda to Modern Giant Panda Genomes
(A) Ordination of individuals along the first and second components of a PCA based on 409,165 variable transversion sites. Axis labels indicate the percentage of variance explained by each component. Symbols for each population are indicated in the key at the bottom left. Singleton positions were excluded from this analysis, which conservatively reduces the overall separation of the ancient from the modern individuals, but the ancient population is still clearly distinct.
(B) Neighbor-joining phylogeny based on 403,235 transversion sites, rooted using the polar bear as outgroup (not shown). Note that the QXL population is not recovered as monophyletic, which likely reflects the substantial divergence of LS within this population cluster (A) [2].
(C) Topology tests for the position of the ancient panda as basal to all modern pandas based on the excess of derived alleles that a modern giant panda shares with the ancient giant panda and not with another modern giant panda, compared to the excess of derived alleles that a modern giant panda shares with another modern giant panda and not with the ancient giant panda. These derived allele proportions (x axis) are expressed as D statistics (black points) calculated for all combinations of individuals for the topologies indicated on the y axis. All Z scores for ((ancient,modern),modern) were >3, whereas many Z scores for ((modern,modern),ancient) were <3. Consistently lower D values for the topology ((modern,modern),ancient) supports the ancient panda as basal to all modern pandas. The giant panda was used as mapping reference for all these analyses.
DOI:10.1016/j.cub.2019.04.021