• biodiversity;
  • Cercidiphyllum ;
  • evolution;
  • genetic markers;
  • phylogeny

Some plant groups are surviving representatives of what were once more diverse or abundant lineages (Sakai, 1971). Among these so-called evolutionary relicts, ‘living fossils’ arouse noteworthy interest because of their static morphology and often surprising shrinkage of distributional ranges (Manchester et al., 2009). Although some relict genera, such as cycads, are still relatively species-rich today, a few are monotypic or contain only two to several species, for example, the Ginkgo and Cercidiphyllum trees. Two important questions connected to the evolution of these groups are: (1) Are the current species of these relict genera true ‘relicts’, which originated tens or even hundreds of millions of years ago (Ma), or did they diversify more recently? (2) Have these relicts experienced further evolutionary changes during shifts in distributional range in response to climatic and geological changes? Although fossil records provide an outline of the general shrinkage of the distributional ranges of such relict genera, they do not provide detailed answers to the above two questions. For several relict genera, fossil-calibrated molecular phylogenies have suggested a burst of recent species diversification from an ancient ‘stem’ lineage (Nagalingum et al., 2011; Renner, 2011; Zhang et al., 2011). However, these species-level phylogenies cannot illustrate the range dynamics of populations and their evolutionary history in response to climatic or geological changes. In this issue of New Phytologist Qi et al. (pp. 617–630) addressed these two questions in a study of the relict genus Cercidiphyllum by using genetic information from both chloroplast (cp) and nuclear DNAs, and assessing molecular structure and diversity in relation to past (Last Glacial Maximum (LGM); c. 21 000 yr before present (bp)) and present geographical distributions based on ecological niche modelling (ENM).

‘… the study illustrates the evolutionary dynamics ofan ancient relict genus, and in doing so deepens our understanding of the possible impacts of modern climate change on biodiversity.’

Cercidiphyllum is assigned to a presently monotypic family Cercidiphyllaceae. Cercidiphyllum-like fossils first appeared in the late mid-Cretaceous (Turonian) in Israel (Fig. 1), and later became widespread in North America, Europe and Asia during the Paleocene (Krassilov, 2010). As with other relict tree genera, the distribution of this genus has become greatly reduced since the Miocene and is presently restricted to low latitudes in Asia (Manchester et al., 2009). Currently, only two species are acknowledged within this dioecious, wind-pollinated and wind-dispersed genus: C. japonicum occurs mainly in low- to mid-elevation warm-temperate deciduous forests in central-southern China and Japan, while C. magnificum is restricted to the cool-temperate/sub-alpine forests of Central Honshu, Japan (Krassilov, 2010) (Fig. 1). Consistent with the fossil record, Qi et al. estimated that Cercidiphyllum split from its sister lineage during the late Cretaceous based on cpDNA evidence. However, they found that the two major cpDNA lineages identified within the genus diverged relatively recently, that is at the Miocene–Pliocene boundary (c. 5 Ma). Moreover, it is intriguing that these two cpDNA lineages are not species-specific: North Japanese populations of C. japonicum share the same cpDNA haplotypes that are otherwise only found in C. magnificum. By contrast, nuclear ribosomal internal transcribed spacer (ITS) sequences and microsatellite (nSSR) loci identified the two species as reciprocally monophyletic units. Factors that frequently cause such cytoplasmic–nuclear discordance include imperfect taxonomy, incomplete lineage sorting of ancestral polymorphisms and introgression (e.g. Neiva et al., 2010). Here, the pronounced inter-specific divergence levels at all surveyed nuclear loci, together with the regional sharing of sequences at independently evolving, maternally inherited cpDNA markers, was interpreted to indicate the regional (North Japanese) introgression of C. magnificum cpDNAs into the otherwise well-differentiated C. japonicum gene pool.


Figure 1. (a) A Cercidiphyllum-like leaf fossil from the Late Cretaceous in Israel (courtesy of V. Krassilov from Krassilov, 2010) and (b) two leaves of one of the two current species (C. japonicum).

Download figure to PowerPoint

Although we basically agree with this scenario, we do not fully concur with the authors’ still widely held view that time to monophyly (in a dioecious species) is usually around four times longer for nuclear than organelle (e.g. cpDNA) genes because of the proportionately larger effective population size (Ne) of nuclear genes (4Ne generations; e.g. Palumbi et al., 2001). In fact, there is accumulating evidence, at least for ITS, suggesting the opposite trend, namely that supposedly neutral mutations at this multi-copy nuclear region become fixed much faster during speciation and following interspecific introgression in angiosperms than mutations in the chloroplast genome (Li et al., 2011; Wang et al., 2011). Moreover, under an incomplete lineage scenario, it appears highly unlikely that the two cpDNA lineages found in Cercidiphyllum japonicum could have been maintained over several (c. 5) millions of years, especially in a tree species that probably has never been a dominant component of East Asia's Plio-Pleistocene flora (Yu et al., 2000). It is possible that the initial divergence between C. japonicum and C. magnificum may have occurred earlier than the estimated times for these two cpDNA lineages, but still within the Late Tertiary. These caveats notwithstanding, regional interspecific introgression, that is, chloroplast capture resulting from hybridization with C. magnificum, indeed provides the most parsimonious explanation for the presence of two deep cpDNA lineages within North Japanese C. japonicum.

Such an interspecific introgression event would suggest that the distributional ranges of the two extant species shifted greatly in their history. Qi et al. further used ENM approaches to compare the species’ potential present and likely past (LGM) distributions. Because of the limited distribution of Cercidiphyllum magnificum, the authors failed to find pronounced range changes in this species. However, when combined with the molecular data, they did infer that during Pleistocene glacial periods, the range of C. japonicum experienced massive reductions in North-central China/North Japan, but increases elsewhere (i.e. in southwest/southeast China, East China Sea landbridge, and South Japan). In China, the Sichuan Basin and/or the middle-Yangtze were likely source-areas of postglacial northward re-colonization. Perhaps most interestingly, in Japan, secondary contact and introgressive hybridization with the cool-temperate congener, C. magnificum, may have increased potentially adaptive genetic diversity in C. japonicum, facilitating the species’ postglacial expansion to the far north of Japan.

This study has several implications. First, the relatively recent origin of Cercidiphyllum japonicum and C. magnificum provides a convincing example to support the recent suggestion that some current plant ‘evolutionary relicts’ or ‘living fossils’ may not have originated as early as thought based on morphological similarity to recorded dated fossils (Nagalingum et al., 2011; Renner, 2011; Zhang et al., 2011). Hence, the diversification history of each putative relict genus comprising more than one species warrants further and closer assessment. Molecular phylogenetic studies addressing this issue are still few, but indicate that unexpectedly recent diversifications of alleged relict plant genera might have been facilitated by climatic change associated with, for example, increased aridity, coldness and seasonality since the Miocene in various parts of the world (Nagalingum et al., 2011; Renner, 2011; Zhang et al., 2011). Second, the distributional ranges of these relict genera may not have continuously contracted as previously thought (Sakai, 1971). Rather, at certain times, their distributions may have also expanded greatly. Such detailed insights could not have been inferred from fossil records alone due to their limitations in both taxonomic resolution and spatial coverage. Finally, the results of Qi et al.'s study have implications for a better understanding of the genetic-demographic context of introgressive hybridization and species delimitation. Thus, their data demonstrate that distributional dynamics triggered by Pleistocene climatic changes might have frequently created opportunities for interspecific introgression within glacial refugia and/or at secondary contact zones following interglacial/postglacial range expansions. Furthermore, the present findings concur with theoretical and empirical evidence, suggesting that seed-transmitted organellar markers (here cpDNA) with low rates of gene flow might be favored to introgress more readily into the genetic background of an expanding species than markers dispersed by both pollen and seed (Petit & Excoffier, 2009; Zhou et al., 2010; Du et al., 2011). The concomitant observation that lineage sorting of nuclear ITS has gone to completion in the face of introgression adds further support to the proposal that this nuclear DNA fragment is a good ‘barcode marker’ for species delimitation (Li et al., 2011; Wang et al., 2011).

In conclusion, the paper by Qi et al. not only draws a very detailed picture of the evolutionary history of Cercidiphyllum, but most importantly also shows that the Late Quaternary range history of C. japonicum and its preferred habitat (warm-temperate conditions) was highly dynamic; thus, questioning a general assumption of relative climatic and environmental stability in East Asia during this period. In addition, the study illustrates the evolutionary dynamics of an ancient relict genus, and in doing so deepens our understanding of the possible impacts of modern climate change on biodiversity. For example, the results raise the possibility that some monotypic relicts or ‘living fossils’, such as the Ginkgo tree, could possibly diverge into more than one species in future. However, ongoing climate change will undoubtedly drive species extinctions and range shrinkages of some species-rich genera in the near future. This should be seriously considered when assessing impacts of modern climate change on biodiversity.


  1. Top of page
  2. References