Abstract
Constituting one of Earth’s major biomes, steppes are characterised by naturally treeless extra-tropical vegetation. The formation of the Eurasian steppe belt, the largest steppe region in the world, began in Central Asia during the Neogene. In the glacial stages of the Pleistocene, steppe displaced forest vegetation, which in turn recolonised the area during the warmer interglacial periods, thus affecting the distribution of plants adapted to these habitats. Krascheninnikovia ceratoides (Chenopodiaceae) is a plant characteristic of dry steppe and semi-desert formations. Earlier studies showed that the ancestor of this autochthonous steppe element originated in Central Asia during the Miocene/Pliocene, i.e., in the same region and at the same time as the first appearance of steppe vegetation. However, as the extant lineages of Krascheninnikovia ceratoides diversified only 2.2 ± 0.9 Mya, it may represent a modern element of current dry steppe and semi-desert formations, rather than a component of the first steppe precursors of the Miocene. As such, it may have capitalised on the climatic conditions of the cold stages of the Quaternary to expand its range and colonise suitable habitats outside of its area of origin. To test this hypothesis, phylogeographic methods were applied to high-resolution genotyping-by-sequencing data. Our results indicate that Krascheninnikovia originated in western Central Asia and the Russian Altai, then spread to Europe in the West, and reached North America in the East. The populations of eastern Central Asia and North America belong to the same clade and are genetically clearly distinct from the Euro-Siberian populations. Among the populations west of the Altai Mountains, the European populations are genetically distinct from all others, which could be the result of the separation of populations east and west of the Urals caused by the Pleistocene transgressions of the Caspian Sea.
Similar content being viewed by others
Introduction
Steppes are characterised as naturally treeless extra-tropical vegetation dominated by drought-resistant xeromorphic grasses, perennial herbs, and low shrubs, geophytes and therophytes1,2,3. The Eurasian steppe belt is the largest steppe region in the world, stretching from the Hungarian basin in the West to the Amur River in the East. Its formation presumably started in Central Asia in the Neogene1 [and literature cited therein]. Significant Northern Hemispheric glaciation started at the beginning of the Pleistocene. In the last 2.6 Million years (My), numerous cold-warm cycles have occurred, as indicated by oxygen isotope studies of marine sediments4. During the cool glacial periods, steppes spread at the expense of forests, while forests recolonised the same areas in the warmer interglacial periods5. The distribution range of plants growing in steppe and semi-desert formations has expanded and contracted accordingly6. Some important landscape structures that may have affected the ability of steppe plants to migrate within their appropriate habitat during the ice ages were permanent, such as the Altai Mountains in Russia and Mongolia, or the Khangai Mountains of Central Mongolia7. On the other hand, temporarily large bodies of water in the area south of the Ural Mountains, caused by transgressions of the Caspian Sea during the ice ages, may have inhibited the migration of steppe plants for only a limited time1, 8,9,10.
In Central Europe, the remains of the more or less continuous Pleistocene Artemisia-Chenopodiaceae cold steppe (tundra-steppe) belt1, 5 can be found nowadays only in azonal steppe islands and in artificial, man-made steppe-like habitats11, 12. In recent years, steppe plants and their shared history with the Eurasian steppe belt have come under the focus of much research13,14,15,16,17,18,19,20,21.
We still know little about where and when the constituent taxa of the various steppe and semi-desert formations originated, and how they have spread. Due to its distribution and age, the xeromorphic shrub species Krascheninnikovia ceratoides (L.) Gueldenst. (Chenopodiaceae) is a particularly suitable subject for reconstructing the evolutionary history of dry steppe and semi-desert formations. It is widespread throughout the Northern Hemisphere, growing on shortgrass steppes, desert steppes, and in semi-deserts22, 23. Although its high morphological variation has led to the recognition of a variety of taxonomical names, Krascheninnikovia is now thought to constitute a single species with two subspecies, the Eurasian K. ceratoides subsp. ceratoides and the North American K. ceratoides subsp. lanata23. The distribution of K. ceratoides subsp. ceratoides is broad and continuous, extending from Mongolia and northern China in the East to the southern Volga and Don regions (European Russia) in the West (Fig. 1). To the South, it reaches the Himalayas and Iran. Large parts of its distribution overlap the Irano-Turanian floristic region24, 25. In Europe, small and isolated populations occur in xerothermic grasslands and on the dry steppes of Austria, Romania, Spain, and Ukraine. Krascheninnikovia ceratoides subsp. lanata is present on the shortgrass prairies, dry scrubland, and in semi-desert regions of central and western Canada, the USA, and Mexico22.
Population sampling of Krascheninnikovia ceratoides. Up to five individuals per population were used, as indicated by dot size. (A) The colour coding indicates the assigned geographical region of the population following Brummit (2001) (CEur Central Europe, EEur eastern Europe, WMAs west Middle Asia, CMAs central Middle Asia, EMAs east Middle Asia, SMAs south Middle Asia, RusAlt Russian Altai, SCAs south Central Asia, WCAs western Central Asia, ECAs east Central Asia, NAm North America). The shaded area indicates the distribution area of K. ceratoides. (B) The ploidy is indicated by colour: diploid (grey) and tetraploid (black). The map was generated in R using the package ‘rworldmap’ (http://journal.r-project.org/archive/2011-1/RJournal_2011-1_South.pdf) and edited in Inkscape v1.0.2 (https://inkscape.org).
Krascheninnikovia ceratoides is a long-lived shrub growing 0.1 to 2 m tall, depending on the habitat26. It can be characterised as an eurythermal species1 that can withstand a wide range of hot and cold temperatures, and grows from sea level (e.g. in the lowlands of Kazakhstan) to high elevations of 4500–4800 m in the southern part of its range, such as in the Pamir and Ladakh mountain ranges26, 27. In Asia, the plant is used for firewood and fodder, with its overuse leading to shrinking population sizes and concomitant possible land degradation as a consequence of increased erosion28. This phenomenon is called the Teresken syndrome for Teresken, the common name for K. ceratoides in Russian28, 29. In North America it is used as fodder for livestock in winter (‘winterfat’)30, 31. Krascheninnikovia ceratoides plants can reach an age of 300 years32, and first reproduction only occurs in plants older than 25 years33. The species is believed to be outcrossing34. The bractlets surrounding its fruit and utricle are densely covered with long hairs, which may facilitate dispersal by wind.
Various ploidy levels exist in K. ceratoides throughout its distribution range. Diploids with 2n = 18 chromosomes have been reported from the Ural Mountains of western Russia, across Kazakhstan and Kyrgyzstan to Mongolia and China35,36,37,38,39,40. Tetra- and/or hexaploids (2n = 36 and 54) have been reported from Europe to the Russian Altai Mountains35, 41,42,43,44,45,46, and south of this range in Armenia47, China36, 38, 39, Iran48, Kyrgyzstan37, and Tajikistan40. A mixed population of di-, tri-, tetra-, penta-, and hexaploids (2n = 18, 27, 36, 45 and 54) has been reported from Kyrgyzstan37. Diploid and tetraploid populations occur in North America35, 49.
According to previous studies, the stem group of Krascheninnikovia originated c. 5 to 23 Million years ago (Mya) (~ 21.2 Mya50; 12.8 Mya (5.3‒23.1 Mya)51; 17.3 Mya (12.0‒23.5 Mya)35), while the crown group diversified 2.2 ± 0.9 Mya35. With a minimum age of more than 5 My, K. ceratoides has persisted throughout several global climatic changes including the glacial and interglacial phases of the Quaternary. Its diversification into extant lineages took place during the Gelasian (Early Pleistocene, 2.6‒1.8 Mya), and seemingly correlates with the appearance of modern types of desert and semi-desert. These are, together with semi-arid temperate grasslands such as prairie, steppe, and pampas, a Plio-/Pleistocene phenomenon52. Therefore, K. ceratoides could have been among the first plants that inhabited modern-type dry steppe areas.
Seidl et al.35 reconstructed the biogeographical history of Krascheninnikovia using a few molecular markers from its nuclear and chloroplast genomes (ITS, ETS, atpB-rbcL, rpl32-trnL, trnL-trnF). Genetic analyses relying on single loci may reflect only the history of these particular loci, rather than the history of the species studied. Therefore, a whole-genome approach is favourable, where single loci have less weight and cannot influence the result as strongly. To investigate the spatio-temporal diversification of K. ceratoides in more detail, we sequenced populations of the species from throughout most of its distribution range employing genotyping-by-sequencing (GBS). We then inferred phylogenetic trees and networks from the GBS data, conducted ancestral range reconstructions, and measured genome size. We specifically addressed the following questions: (1) Where did Krascheninnikovia originate? (2) Can GBS provide the appropriate resolution to retrace its colonising history and migration routes? (3) Can the retrieved phylogeographic and population genetic patterns be related to palaeogeographical and palaeoenvironmental events during the evolutionary history of the Eurasian steppe? With this, our overall aim was to contribute to the knowledge of the florogenesis of the Eurasian steppe belt.
Materials and methods
Plant material
Krascheninnikovia ceratoides leaves were collected in the field and dried with silica gel or taken from herbarium material. Leaves from up to five individuals per population were collected. The five individual specimens from Kyrgyzstan combined into population P025 came from two different localities about 80 km apart. In total, 183 accessions from 43 different populations of K. ceratoides were used for GBS analyses (Fig. 1; Supplementary Table S1). Most of the samples were the same as those used in a previous phylogenetic study of ITS, ETS, and chloroplast DNA sequences35. Three North American populations determined as K. ceratoides subsp. lanata (P039, P040, P041) were included, as well as two eastern Middle Asian populations of K. ceratoides subsp. ceratoides, which were originally determined as K. ewersmanniana (P069, P070). Axyris hybrida and Ceratocarpus arenarius were used as outgroups. Further details to the collected samples can be found in Supplementary Table S1.
The present study complies with relevant institutional, national, and international guidelines and legislation.
Genome size measurements
We determined the genome size and ploidy level of 99 samples of K. ceratoides leaf tissue by flow cytometry, using an internal standard (either Pisum sativum L. ‘Kleine Rheinländerin’ (2C = 8.8 pg53) or Petroselinum crispum (Mill.) Fuss (2C = 4.5 pg54)) and the CyStain PI Absolute P kit (Sysmex, Görlitz, Germany) according to the manufacturer’s instructions. The relative fluorescence of at least 5000 particles per sample was recorded with a CyFlow Space (532 nm diode laser; Sysmex).
DNA extraction and GBS
The accessions used in the study by Seidl et al.35 were complemented with additional individuals from both the same and new populations (P076, P079, P102, P207, P241, P269; Supplementary Table S1). Genomic DNA was extracted from about 20 mg dried leaf material using the DNeasy Plant Mini Kit (Qiagen, Hilden, Germany), the NucleoSpin Plant II kit (Macherey–Nagel, Düren, Germany), or the innuPREP Plant DNA Kit with the SLS CTAB-containing lysis solution (Analytic Jena, Jena, Germany) according to the manufacturers’ instructions. Five DNA samples were extracted from herbarium leaf material (P207‒P218). DNA concentrations were quantified with a DS-11 FX Fluorometer (DeNovix, Wilmington, Delaware, USA), using the DeNovix dsDNA Broad Range Kit (DeNovix) according to the manufacturer’s manual.
Genotyping-by-sequencing library preparation followed Poland et al.55, using the restriction enzymes MspI and PstI. Samples were sequenced in two runs on a HiSeq 2500 (Illumina, California, USA) (single-end sequencing of 100 bp fragments). To achieve the same genomic coverage for individuals with different ploidies, twice the amount of DNA was used for tetraploids as for diploids. Two individuals (one from population P029 and one from population P072) were sequenced twice to assess the reproducibility of the method.
Data analysis
We used STACKS v.2.5556 for the sorting, filtering and de novo assembly of raw read data. To find the optimal parameter settings, we ran STACKS ‘denovo_map’ on a reduced data set (Supplementary Methods S2). We varied one of the three following parameters while keeping the others at their default value: the minimal number of identical reads required per stack, m, was tested in a range from three to six; the differences between loci within an individual, M, was tested for one to four base pair differences; the differences allowed between catalog loci, n, was tested for one to eight base pair differences. The parameter value that maximized the number of variant loci present in at least 80% of all individuals were used for the complete data set: m = 3, M = 3 and n = 8. We set the number of stacks per locus to five to account for possibly present allelic variation in tetraploids. Diploid and tetraploid samples were treated identically. The STACKS ‘population’ program using the ‘write random SNPs’ flag was used to produce unlinked SNP data sets, with and without the Axyris and Ceratocarpus outgroups.
To estimate ploidy levels from GBS data, we used nQuire57 and ploidyNGS58, which recognize bam-files as input. These were built using BWA (Burrows-Wheeler Alignment tool)59 with the concatenated consensus sequence of the respective individuals serving as a reference. BWA settings were the following: maximum gap length of twelve, maximum of one gap per read, and a mismatch parameter of 0.0160. These bam-files were used as input for nQuire using the denoising option and for ploidyNGS using the “guess ploidy” option.
To infer the population genetic structure of Krascheninnikovia, an analysis was performed in R61 using the LEA62 package, using the data set of unlinked SNPs in Variant Call Format (VCF) of ingroup samples only. The ploidy was set to tetraploid. One to fifteen hypothetical ancestral populations were tested with 100 repetitions each. The number of clusters was chosen based on minimal cross-entropy. To show the result per population, the affiliation to each hypothetical ancestral population was calculated per population as the mean over its individuals and shown as pie charts on a map. To validate the results using a different approach, both Discriminant Analysis of Principal Components (DAPC) and Principle Coordinate Analysis (PCoA) were performed using the R packages63, 64 adegenet and dartR, respectively, based on the ingroup data set of unlinked SNPs.
Unlinked SNPs of 6,029 loci were retained in the PHYLIP file of Krascheninnikovia as consensus sequences per population, after removing all uninformative sites65. Phylogenetic analyses were conducted by applying the maximum likelihood (ML) criterion using66 IQ-TREE v.2.0.367 with the GTR+G model and correction for ascertainment bias68. SH-like approximate likelihood ratio test69 and ultrafast bootstrapping (UFBoot)70 were performed with 1,000 repetitions each to obtain bootstrap support (aLRT/UFBoot BS) for nodes. The outgroups Axyris and Ceratocarpus are genetically too distant to assess the correct rooting, so midpoint rooting was used, which resulted in a tree with basal positions for populations of the Russian Altai. The basal population of the IQ-TREE analysis in Seidl et al.35, P242, originated in the Russian Altai as well. The splits data was visualised in SplitsTree v4.14.871 and the consensus tree in FigTree v1.4.372.
To calculate private allelic richness per region, a rarefaction analysis was performed in HP-Rare73. Populations were assigned to biogeographic regions according to their origin based on Brummit74: Central Europe, eastern Europe, western Middle Asia, central Middle Asia, eastern Middle Asia, southern Middle Asia, the Russian Altai Mountains, western Central Asia, southern Central Asia, eastern Central Asia, and North America. Only loci with called bases present in all populations were used. Rarefaction sample sizes for each ‘SNP locus’ were two populations from each region, and two alleles (‘genes’ in HP-Rare) from each population. Mean rarefaction and standard deviation were then calculated per region.
Ancestral range reconstruction
An ultrametric tree was built based on the ML tree using PATHd875 as input tree for the BioGeoBears v. 1.1.176,77,78 R package for R ver. 3.5.076,77,78. The assignment of populations to regions was described above. The analysis was performed based on two different migration hypotheses: free migration between all areas, and restricted migration due to the distance between areas (0.8 for regions with a common land border and 0.2 for areas without a common land border). Three different biogeographic models were tested: DEC, DIVALIKE, and BAYAREALIKE, all both with and without the jumping parameter “j”, which determines the probability of jump dispersals. Based on AICc values, DIVALIKE+J with restricted migration was chosen as the most likely model. DIVALIKE is a likelihood interpretation of parsimony-based DIVA and allows the same processes79 such as dispersal, extinction, narrow sympatry, narrow and widespread vicariance, but not, e.g., subset sympatry.
Results
Ploidy
The mean 2C genome sizes (± standard deviation) of diploid (N = 38) and tetraploid (N = 61) individuals were 2.8 ± 0.3 pg and 5.8 ± 0.2 pg, respectively. The populations measured were uniformly di- or tetraploid. The populations from Central Europe, western Middle Asia, and the Russian Altai Mountains were all tetraploid, but in most regions both diploid and tetraploid populations were found (Fig. 1; Supplementary Table S1). The genome sizes of 85 samples could not be measured due to the old age of the material, their storage conditions (e.g., populations from eastern Europe: P215; central Middle Asia: P034; western Central Asia: P207, P209, P218; eastern Central Asia: P210; Supplementary Table S1), or a lack of material. Only one individual each of the populations P034 (central Middle Asia) and P025 (Kyrgyzstan) could be measured, both turning out to be diploid.
Ploidy level estimations based on GBS data using nQuire and ploidyNGS confirmed most of these measurements. Only the North American sample P041-I0372 was inferred to be tetraploid but appeared diploid according to flow cytometry measurements. Of those not measured using flow cytometry, the populations of eastern Europe (P215), western Central Asia (P207, P209, P218), and eastern Central Asia (P210) were found to be tetraploid; of P025, all individuals were found to be tetraploid except P025-I0223, which was diploid in both analyses, by GBS and flow-cytometry data. Consequently, we consider it very likely that P025-I0223 came from one of the two Kyrgyz localities, and the other four P025 individuals from the other locality, which is ~ 80 km away. P025-I0223 (hereinafter P025_D) and the other four individuals (hereinafter P025_T) were therefore treated as two different populations in all analyses. Of population P034, one individual (P034-I0317) was tetraploid, while all others were diploid. In all populations except for P025 and P034, the GBS-estimated ploidy of unmeasured samples corresponded to the ploidy of measured individuals from the same population.
Raw data analysis
Of 24,249 loci, composed of 2,625,739 sites, 21,254 variant and unlinked sites remained in the STACKS VCF file of Krascheninnikovia. Of those loci, 6029 were kept after removing all uninformative sites. In the PHYLIP file including the outgroups, only 500 SNPs of Axyris and 525 SNPs of Ceratocarpus were retained, while the populations of Krascheninnikovia comprised on average 3731 (± 883) SNPs. Due to this huge difference of the amount of recovered data, the analyses including the outgroups may not be meaningful and were dismissed.
Genetic clusters
LEA assigned all individuals to hypothetical ancestral populations based on their genetic composition. After 100 repetitions, eight hypothetical ancestral populations (K) were chosen based on the cross-entropy criterion (Supplementary Figure S3). The red cluster (Fig. 2) is mainly represented by populations from Central Europe, eastern Europe, and western Middle Asia (all west of the Ural Mountains, mostly tetraploids except for populations P057 and P062 from eastern Europe). The dark green cluster mainly contains tetraploid populations from western, central, eastern, and southern Middle Asia, and southern Central Asia. The light green cluster is mainly represented by diploid populations from eastern and southern Middle Asia and from southern Central Asia and is present in the diploids of eastern Europe and central Middle Asia. The black cluster is represented by three individuals of P081 of eastern Middle Asia, the light blue cluster by P238 of southern Central Asia. The dark blue cluster is mainly comprised of populations from the Russian Altai Mountains and western Central Asia. Populations from eastern Central Asia form the orange cluster. The populations from North America belong to the lilac cluster. Some admixture is observed, especially between the light green, dark green, and red clusters (Fig. 2).
Genetic cluster membership as inferred using LEA is shown as barplot. Vertical bars denote individuals. Eight clusters were retained, which are indicated by different colours. The samples are ordered by their assigned geographical region from west to east. Ploidy is indicated by grey (diploid) and black (tetraploid) bars. The figure was generated in R using the package ‘ggplot2’ (https://ggplot2.tidyverse.org) and edited in Inkscape v1.0.2 (https://inkscape.org).
To illustrate the consecutive splits between the hypothetical ancestral populations, the cluster membership was depicted as pie charts on a map for all K values from two to eight (Fig. 3). The deepest split, with two hypothetical ancestral populations (K = 2), was inferred between the populations east of the Khangai Mountains in Central Mongolia (North America and eastern Mongolia), and the populations west of the Khangai Mountains. The second deepest split (K = 3) is in the area of the Ural Mountains, dividing the populations west of the Khangai Mountains. With K = 4, the Mongolian populations east of the Khangai Mountains and the North American populations each form a group of their own. The populations between the Ural and Khangai Mountains split in two groups according to their ploidy with K = 5: diploid populations are assigned to the light green cluster, while tetraploid populations are assigned to the dark green cluster. With K = 6, the populations of the Russian Altai and western Central Asia form their own group (dark blue). With K = 7, population P238 from southern Central Asia forms its own cluster (light blue), as does P081 with K=8 (black).
Genetic clustering of samples as inferred from LEA is shown as pie charts for K=2 to K=8 clusters. The first split occurred in Central Asia, dividing the populations west and east of the Khangai Mountains (K=2), followed by a split in the area of the Ural Mountains (K=3). At K=4, the American populations separate; at K=5, the diploids and tetraploids of Middle Asia and adjacent areas each form their own group. At K=6, the populations of the Altai Mountains form their own cluster. At K=7 and K=8, P238 and P081 form their own cluster. The map was generated in R using the packages ‘mapplots’ (https://CRAN.R-project.org/package=mapplots) and ‘rworldmap’ (http://journal.r-project.org/archive/2011-1/RJournal_2011-1_South.pdf) and was edited in Inkscape v1.0.2 (https://inkscape.org).
DAPC
Similarly to LEA, DAPC analysis clusters individuals according to their genetic composition (Fig. 4). Four clusters were retrieved: (1) populations from Europe (Ural Mountains to Central Europe, including western Middle Asia), (2) the populations from Middle Asia and southern Central Asia, (3) the populations of the Russian Altai and western Central Asia, and (4) the populations of eastern Central Asia and North America, mirroring the first splits of the analysis performed using the LEA package.
Genetic clustering of samples as calculated by dapc package of R. Four clusters were retained. One cluster comprises the European populations, one the Middle Asian and southern Central Asian populations, one the populations of western Central Asia and the Russian Altai, and another the eastern Central Asian and North American populations. The figure was generated in R and edited in Inkscape v1.0.2 (https://inkscape.org).
PCoA
Samples from Europe, Middle Asia, southern Central Asia, the Russian Altai and western Central Asia clustered together in the Principal Coordinate Analysis (Fig. 5). Apart from this group, two other clusters were found: samples from eastern Central Asia and samples from North America. The populations of western Central Asia and the Russian Altai are genetically different to other populations of the majority group. Among the samples from Europe, Middle Asia and southern Central Asia, diploids and tetraploids segregated along the second PCoA axis.
PCoA scatterplot. Symbols representing samples of the same geographical area were coloured according to Fig. 1. Circles indicate diploids, triangle tetraploids. Upper graph: Distinct from most of the samples, the populations of eastern Central Asia and North America each form their own clusters. Lower graph: Within the remaining populations, those from the western Central Asia and from the Russian Altai are genetically distinct. Within the densely scattered other populations, there is a trend visible, which separates diploids and tetraploids from each other. The figure was generated in R using the packages ‘MASS’ (http://www.stats.ox.ac.uk/pub/MASS4/), ‘car’ (https://socialsciences.mcmaster.ca/jfox/Books/Companion/) and ‘graphics’ (https://www.rdocumentation.org/packages/graphics) and was edited in Inkscape v1.0.2 (https://inkscape.org).
Phylogenetic analysis
According to the ML tree, K. ceratoides comprises two fully supported major genetic groups: populations from the Russian Altai, eastern Central Asia, and North America form one clade, which is a sister to a clade comprising the remaining accessions from the Russian Altai and from southern Central Asia to Central Europe (Fig. 6, Supplementary Figure S4). Within the first group, two well-supported clades of populations from eastern Central Asia and North America, respectively, derive from populations of western Central Asia and one population of the Russian Altai (99.9 aLRT/100 UFBoot BS). In the second group, populations from the Russian Altai Mountains constitute the basally branching lineages, forming a grade from which the southern Central Asian, Middle Asian and European populations are derived (Supplementary Figure S4). The populations of central Middle Asia form a well-supported clade (98.8/99), as do the populations west of the Ural Mountains (63.2/94, excluding P102: 93.6/98). In this clade, populations P101 and P102 from western Middle Asia are paraphyletic sister groups to the European clade. The replicates of the two individuals for which GBS analysis was performed twice (P029-I0272, P072-I0637), occur next to each other in the IQ-TREE analysis (100/100), when using data per individual, confirming the reproducibility of the method (data not shown).
Phylogenetic network based on a ML method (IQ-TREE). The origin of the populations is indicated by colour (compare Fig. 1). Tetraploids are shown in bold. The network was visualized in SplitsTree and edited in Inkscape v1.0.2 (https://inkscape.org).
Biogeographic analysis with BioGeoBears
Based on our BioGeoBEARS analysis, the ancestral range of K. ceratoides lies within the area of western Central Asia (Mongolian Altai Mountains, yellow) and the Russian Altai Mountains (dark blue) (Fig. 7). From western Central Asia, K. ceratoides migrated east to eastern Central Asia (orange) and from there to North America (lilac). From the Russian Altai Mountains, it migrated to southern Central Asia and to eastern Middle Asia (light green). From eastern Middle Asia, it then migrated again to southern Central Asia (light blue) and to southern Middle Asia (brown). It migrated further west to central Middle Asia (pink) and to western Middle Asia (dark green), and on to eastern Europe (red) and Central Europe (blue).
Phylogeographic analysis using the DIVALIKE model of the R package BioGeoBEARS. Colours indicate the origin of the samples, compare Fig. 1. Branches were collapsed when several members of the same populations clustered together. Krascheninnikovia spread from the Altai Mountains to the East, reaching eastern Central Asia and North America, and to the West, reaching Middle Asia and Europe. The figure was generated in R and edited in Inkscape v1.0.2 (https://inkscape.org).
Private allelic richness
The highest private allelic richness was found in western Central Asia (0.014 ± 0.002), followed by eastern Central Asia (0.013 ± 0.004) and the Russian Altai Mountains (0.011 ± 0.001; Supplementary Figure S5). The lowest private allelic richness was found in North America (0.007 ± 0.003).
Discussion
In this study we reconstructed the phylogeographic history of K. ceratoides, a widespread and representative steppe inhabitant, to gain insights into the general evolutionary history of the Eurasian steppe belt. The retrieved molecular patterns fit best to a scenario of a western Central Asian/Russian Altai origin of Krascheninnikovia and its subsequent spread both towards the West, through stepwise migration until reaching Central Europe, and towards the East, finally colonizing western North America.
Steppe precursors originated in Central Asia during the Miocene at the latest1, 80, 81. From there, the steppe started to spread, with many latitudinal range shifts due to changing climatic conditions6, 82,83,84,85. A continuous Eurasian steppe belt was first present during the late Miocene/early Pliocene1. At the end of the Pliocene, true steppes were present in Central Asia, while forest-steppe assemblies were present in south-eastern Europe and on the West Siberian Plain5. Cold steppe vegetation spread during the glacial periods of the Pleistocene1. Eurythermal species could have survived the last glacial maximum in situ1. Krascheninnikovia ceratoides is an autochthonous steppe element, which originated in the steppe belt similarly to some species of the genus Allium86. In contrast, species of the genus Camelina migrated into the Eurasian steppe secondarily out of their presumed area of origin in the western part of the Irano-Turanian floristic region19.
The Altai Mountains are known as a refuge, e.g. for cold-adapted species during warm humid phases87, 88, and could have served as a refugium for steppe and semi-desert plants such as K. ceratoides during the interglacial phases of the Quaternary. As a consequence, the Altai Mountains could have been a source for recolonisation of the spreading steppe area during glacial periods. Because of the numerous large populations of K. ceratoides in Central Asia with high morphological variation in leaf size and shape, hair density, and plant height, this area was assumed to be the cradle of the species35, 46, 49, 89,90,91,92,93,94,95,96. Our findings suggest western Central Asia and the Altai Mountains as the area of origin for Krascheninnikovia (Fig. 7) and the starting point of its range expansion throughout the Northern Hemisphere. As indicated by their high private allelic richness, the populations of eastern and western Central Asia harbour a diverse gene pool (Supplementary Figure S5). Thus, these may have served as a good source for dry steppe plants for recolonising the surrounding areas after times of unsuitable, more humid conditions97 during the interglacial periods.
Although the stem lineage of Krascheninnikovia may have existed since the Pliocene34, 49, 50, it has only diversified in the last ~ 2.2 My35, coincidental with the beginning of the Pleistocene glacial-interglacial macro cycles. It seems that the cold, dry climate during the glacial stages facilitated the migration of Krascheninnikovia, as opposed to the more humid and less continental climatic conditions earlier during the Neogene1.
Krascheninnikovia ceratoides shows an apparent isolation-by-distance pattern: populations that are geographically close are usually genetically similar (Figs. 2, 5, 6, 7). The populations of Central Asia, which belong to two distinct genetic lineages, are exceptions. While geographically close to the populations of western Central Asia and separated only by the Khangai Mountains, the eastern populations are genetically more similar to the populations of North America. The Khangai Mountains (covered by alpine vegetation, taiga, mountain steppe, and forest-steppe98) have divided the populations east and west of the mountains and prevented gene flow for a long time. This split occurred in the early stages of diversification, after Krascheninnikovia migrated from the Altai Mountains to eastern Central Asia (Fig. 7).
A migration from eastern Central Asia to North America as hypothesised by Heklau and Röser23 and Seidl et al.35 is here confirmed. The dispersal to North America may have occurred across Beringia, which repeatedly connected the Asian and American continents during several arid glacial phases in the Pleistocene99, 100.
The sister group relationship of populations in North America (USA) on the one hand and eastern Central Asia (East Mongolia) on the other was found in all phylogenetic inferences (Figs. 3, 4) as well as in Seidl et al.35 The two groups form strongly supported clades (Figs. 2, 3, 4, 6, Supplementary Figure S4), which hints at low or absence of gene flow.
Starting from its likely area of origin in western Central Asia and the adjacent Altai mountains, Krascheninnikovia migrated west to Middle Asia.
During the Middle Pleistocene, glaciers dammed the Siberian Ob-Irtysh-Tobol river system, leading to backwaters that reached the foothills of the Kazakhstan Highlands and the Altai Mountains. The resulting landscape of lakes and swampy areas was not a suitable habitat for steppe plants1, 83, 85. According to our biogeographic analysis, this probably also affected the phylogeographic patterns of K. ceratoides, which was probably present on the Kazakhstan plain and even further west at this time. This damming would have caused range splits and contractions, followed by isolation and subsequent expansion. Repeated secondary contact may be the explanation for the absence of monophyly for individuals from eastern Middle Asia in our IQ-TREE tree (Supplementary Figure S4). It has been shown that the histories of other steppe taxa such as Clausia aprica (Stephan) Korn.-Trotzky12, Goniolimon speciosum (L.) Boiss.18, Capsella Medik.13, Allium cretaceum N.Friesen & Seregin, and A. montanostepposum N.Friesen & Seregin14 were affected by these backwater events as well.
Our results suggest that Krascheninnikovia migrated stepwise from Central Asia to Middle Asia and to Europe. Then, the European populations were separated from populations east of the Urals (Figs. 2, 4, 7). This genetic split may be traced back to a migration barrier imposed by a transgression of the Caspian Sea like the Apsheron transgression ~ 1 Mya1. This suggests that the separated populations found refugia in the Ural mountains17, 18 and west of the Caspian Sea18, 101, 102 to survive the interglacial phases. The same spatio-temporal pattern was observed for the steppe plant Camelina microcarpa Andrz. ex DC., whose two major groups split about 1.2 ± 1 Mya19.
According to our data, Central Europe was colonised through stepping-stone-like migration from the Altai Mountains to Central Europe, rather than by long-distance dispersal. Stepping-stone migration to Central Europe would indicate that trees present during glacial periods may potentially have grown in patches, as opposed to a dense tree cover103, allowing the steppe and semi-desert plant Krascheninnikovia to migrate step by step.
Diploid and tetraploid populations of Krascheninnikovia coexist in most regions48. In Eurasia, both ploidy levels are present in eastern Europe, Middle Asia, southern Central Asia, and eastern Central Asia (Fig. 1, Supplementary Table S1)34, 47. Most diploid samples from eastern Middle Asia (P030, P070, P072, P269) and southern Central Asia (P237, P238) belong to the light green cluster in our LEA analysis, suggesting a common ancestry. The same cluster membership is shared by the diploid individuals from central Middle Asia (P034) and eastern Europe (P057, P062), which are admixed. The westward migration of K. ceratoides was possibly accomplished by diploid individuals, followed by (auto-)polyploidization in the newly colonised areas, a process which occurred several times independently as a response to climatic fluctuations in the Pleistocene104.
There is an ongoing discussion about potential subspecies and species in the genus Krascheninnikovia. While some argue that Krascheninnikovia is monospecific with only two subspecies ceratoides and lanata22, others distinguish several entities at the specific level, such as K. arborescens, K. ewersmanniana, and K. compacta from Middle and Central Asia48, 87,88,89,90. Our data did not include samples representing these potential taxa, except for two populations identified as K. ewersmanniana (P069 and P070). Therefore, this study is not suitable to conclude on this topic. The results of our analyses support however the recognition of subspecies lanata as defined by Heklau and Röser23, as the North American individuals form a clearly distinguishable genetic lineage. Our data did not support the samples identified as K. ewersmanniana to belong to a distinct clade, as they are nested at different positions among morphologically typical subsp. ceratoides specimens from eastern Middle Asia. However, to confirm this hypothesis thorough morphological studies are necessary. Thus, we concur with the conclusions of Heklau and Röser23, recognizing (at the moment) two subspecific entities, i.e. subsp. ceratoides and subsp. lanata within monotypic Krascheninnikovia.
Our analyses provide insights into the history of Krascheninnikovia and its connection to the history of the Eurasian steppe belt, confirming and refining previous findings using GBS. The observed phylogeographic patterns reveal the common history of Krascheninnikovia and the steppe belt: The Krascheninnikovia lineage represents an autochthonous steppe element that evolved in situ during the Miocene when steppe precursors made their first appearance in Central Asia. Additionally, the Caspian Sea transgressions and the ice-dammed backwaters of the Ob-Irtysh-Tobol river system during the Pleistocene left clear imprints on the population genetic structure that persist until today. Nevertheless, there is still much to be investigated: The inclusion of further samples from the southernmost part of the range in Eurasia and from North America in future GBS analyses of Krascheninnikovia could allow the narrowing down of the area of origin and conclusions to be drawn about the number of migration events to America that have taken place.
Data availability
The raw data is available at the European Nucleotide Archive (accessions ERS5794424 to ERS5794612).
References
Hurka, H. et al. The Eurasian steppe belt: Status quo, origin and evolutionary history. Turczaninowia 22, 5–71 (2019).
Walter, H. Die Vegetation Osteuropas (Gustav Fischer Verlag, 1974).
Walter, H. Die Vegetation der Erde in öko-physiologischer Betrachtung , Band II : Die gemäßigten und arktischen Zonen, in ökologischer Betrachtung (Gustav Fischer Verlag, 1968).
Cohen, K. M. & Gibbard, P. L. Global chronostratigraphical correlation table for the last 2.7 million years, version 2019 QI-500. Quat. Int. 500, 20–31 (2019).
Frenzel, B. Grundzüge der Pleistozänen Vegetationsgeschichte Nord-Euroasiens. Geogr. J. 136, 291 (1968).
Tarasov, P. E. et al. Last glacial maximum biomes reconstructed from pollen and plant macrofossil data from northern Eurasia. J. Biogeogr. 27, 609–620 (2000).
Caves Rugenstein, J., Sjostrom, D., Mix, H., Winnick, M. & Chamberlain, C. Aridification of Central Asia and uplift of the Altai and Hangay Mountains, Mongolia: Stable isotope evidence. Am. J. Sci. 314, 1171–1201 (2014).
Yanina, T., Sorokin, V., Bezrodnykh, Y. & Romanyuk, B. Late Pleistocene climatic events reflected in the Caspian Sea geological history (based on drilling data). Quat. Int. 465, 130–141 (2018).
Dolukhanov, P. M., Chepalyga, A. L., Shkatova, V. K. & Lavrentiev, N. V. Late Quaternary Caspian: Sea-levels, environments and human settlement. Open Geogr. J. 2, 1–15 (2009).
Tudryn, A. et al. Late Quaternary Caspian Sea environment: Late Khazarian and Early Khvalynian transgressions from the lower reaches of the Volga River. Quat. Int. 292, 193–204 (2013).
Dengler, J., Janišová, M., Török, P. & Wellstein, C. Biodiversity of Palaearctic grasslands: A synthesis. Agric. Ecosyst. Environ. 182, 1–14 (2014).
Hejcman, M., Hejcmanová, P., Pavlů, V. & Beneš, J. Origin and history of grasslands in Central Europe—a review. Grass Forage Sci. 68, 345–363 (2013).
Franzke, A. et al. Molecular signals for Late Tertiary/Early Quaternary range splits of an Eurasian steppe plant: Clausia aprica (Brassicaceae). Mol. Ecol. 13, 2789–2795 (2004).
Hurka, H., Friesen, N., German, D. A., Franzke, A. & Neuffer, B. ‘Missing link’ species Capsella orientalis and Capsella thracicaelucidate evolution of model plant genus Capsella (Brassicaceae). Mol. Ecol. 21, 1223–1238 (2012).
Seregin, A. P., Anačkov, G. & Friesen, N. Molecular and morphological revision of the Allium saxatile group (Amaryllidaceae): Geographical isolation as the driving force of underestimated speciation. Bot. J. Linn. Soc. 178, 67–101 (2015).
Friesen, N. et al. Dated phylogenies and historical biogeography of Dontostemon and Clausia (Brassicaceae) mirror the palaeogeographical history of the Eurasian steppe. J. Biogeogr. 43, 738–749 (2015).
Friesen, N. et al. Allium species of section Rhizomatosa, early members of the Central Asian steppe vegetation. Flora 263, 151536 (2020).
Friesen, N. et al. Evolutionary history of the Eurasian steppe plant Schivereckia podolica (Brassicaceae) and its close relatives. Flora 268, 151602 (2020).
Volkova, P. A., Herden, T. & Friesen, N. Genetic variation in Goniolimon speciosum (Plumbaginaceae) reveals a complex history of steppe vegetation. Bot. J. Linn. Soc. 184, 113–121 (2017).
Žerdoner Čalasan, A., Seregin, A. P., Hurka, H., Hofford, N. P. & Neuffer, B. The Eurasian steppe belt in time and space: Phylogeny and historical biogeography of the false flax (Camelina Crantz, Camelineae, Brassicaceae). Flora 260, 151477 (2019).
Kirschner, P. et al. Long-term isolation of European steppe outposts boosts the biome’s conservation value. Nat. Commun. 11, 1968 (2020).
Heklau, H. & von Wehrden, H. Wood anatomy reflects the distribution of Krascheninnikovia ceratoides (Chenopodiaceae). Flora Morphol. Distrib. Funct. Ecol. Plants 206, 300–309 (2011).
Heklau, H. & Röser, M. Delineation, taxonomy and phylogenetic relationships of the genus Krascheninnikovia (Amaranthaceae subtribe Axyridinae). Taxon 57, 563–576 (2008).
Takhtajan, A. Floristic Regions of the World (University of California Press, 1986).
Manafzadeh, S., Staedler, Y. M. & Conti, E. Visions of the past and dreams of the future in the Orient: The Irano-Turanian region from classical botany to evolutionary studies. Biol. Rev. Camb. Philos. Soc. 92, 1365–1388 (2017).
Walter, H. & Breckle, S.-W. Ecological systems of the geobiosphere. 2 Tropical and subtropical zonobiomes (Springer, 1986). https://doi.org/10.1007/978-3-662-06812-0.
Hartmann, H. Zur Flora und Vegetation der Halbwüsten, Steppen und Rasengesellschaften im südöstlichen Ladakh (Indien). in Jahrbuch des Vereins zum Schutz der Bergwelt 129–188 (1997).
Kraudzun, T., Vanselow, K. A. & Samimi, C. Realities and myths of the Teresken syndrome—An evaluation of the exploitation of dwarf shrub resources in the Eastern Pamirs of Tajikistan. J. Environ. Manag. 132, 49–59 (2014).
Vanselow, K. & Samimi, C. Predictive mapping of dwarf shrub vegetation in an arid high mountain ecosystem using remote sensing and random forests. Remote Sens. 6, 6709–6726 (2014).
Smoliak, S. & Bezeau, L. M. Chemical composition and in vitro digestibility of range forage plants of the Stipa-Bouteloua prairie. Can. J. Plant Sci. 47, 161–167 (1967).
Waldron, B. L., Eun, J.-S., ZoBell, D. R. & Olson, K. C. Forage kochia (Kochia prostrata) for fall and winter grazing. Small Rumin. Res. 91, 47–55 (2010).
Steshenko, A. P. Formation of the semi-shrub structure in the high mountains of Pamir. Trans Akad Nauk Tadzhik SSR 50, 2 (1956).
Zalenski, O. V. & Steshenko, A. P. On the special features of the main species of the vegetation of the Pamir mountains. Proc. Bot. Soc. 7, 9–12 (1957).
Barnes, M. The Effect of Plant Source Location on Restoration Success: A Reciprocal Transplant Experiment with Winterfat (Krascheninnikovia lanata) (University of New Mexico, 2009).
Seidl, A. et al. Phylogeny and biogeography of the Pleistocene Holarctic steppe and semi-desert goosefoot plant Krascheninnikovia ceratoides. Flora 262, 151504 (2020).
Yang, J. Y., Fu, X. Q., Yan, G. X. & Zhang, S. Z. Analysis of three species of the genus Ceratoides. Grassl. China 1, 67–71 (1996).
Rubtsov, M., Sagimbaev, R., Shakhanov, E., Tiran, T. & Balyan, G. Natural polyploids of prostrate summer cypress and winterfat as initial material for breeding. Sov. Agric. Sci. 4, 20–24 (1989).
Yan, G., Zhang, S., Yan, J., Fu, X. & Wang, L. Chromosome numbers and geographical distribution of 68 species of forage plants. Grassl. China 4, 53–60 (1989).
Kurban, N. Karyotype analysis of three species of Ceratoides (Chenopodiaceae). J. Syst. Evol. 22, 466–468 (1984).
Zakharjeva, O. I. & Soskov, Y. D. Chromosome numbers in desert herbage plants. Bulleten VNII Rastenievod. Im. N.I. Vavilova 108, 57–60 (1981).
Domínguez, F. et al. Krascheninnikovia ceratoides (L.) Gueldenst (Chenopodiaceae) en Aragón (España): Algunos resultados para su conservación. Bol. R. Soc. Esp. Hist. Nat. (Sec. Biol.) 96, 15–26 (2001).
Zakirova, R. Chromosome numbers of some Alliaceae, Salicaceae, Polygonaceae, and Chenopodiaceae of the South Balkhash territory. Citologija 41, 1064 (1999).
Dobes, C. H., Hahn, B. & Morawetz, W. Chromosomenzahlen zur Gefässpflanzenflora Österreichs. Linzer Biol. Beitr 29, 5–43 (1997).
Sainz Ollero, H., Múgica, F. & Arias Torcal, J. Estrategias para la conservación de la flora amenazada de Aragón (Consejo de Protección de la Naturaleza de Aragón, 1996).
Lomonosova, M. N. & Krasnikov, A. A. Chromosome numbers in some members of the Chenopodiaceae. Bot. Zurn. (Moscow Leningrad) 78, 158–159 (1993).
Castroviejo, S. & Soriano, C. Krascheninnikovia ceratoides Gueldenst (Publicaciones del CSIC, 1990).
Takhtajan, A. Numeri chromosomatum magnoliophytorum florae URSS. Aceraceae–Menyanthaceae. (Academis Scientiarum Rossica, Institutum Botanicum nomine VL Komarovii;" Nauka", 1990).
Ghaffari, S. M., Balaei, Z., Chatrenoor, T. & Akhani, H. Cytology of SW Asian Chenopodiaceae: New data from Iran and a review of previous records and correlations with life forms and C4 photosynthesis. Plant Syst. Evol. 301, 501–521 (2014).
eFloras. Published on the Internet http://www.efloras.org. (2008).
Kadereit, G., Mavrodiev, E. V., Zacharias, E. H. & Sukhorukov, A. P. Molecular phylogeny of Atripliceae (Chenopodioideae, Chenopodiaceae): Implications for systematics, biogeography, flower and fruit evolution, and the origin of C4 photosynthesis. Am. J. Bot. 97, 1664–1687 (2010).
Di Vincenzo, V. et al. Evolutionary diversification of the African achyranthoid clade (Amaranthaceae) in the context of sterile flower evolution and epizoochory. Ann. Bot. 122, 69–85 (2018).
Janis, C. M. Tertiary mammal evolution in the context of changing climates, vegetation, and tectonic events. Annu. Rev. Ecol. Syst. 24, 467–500 (1993).
Doležel, J. & Greilhuber, J. Nuclear genome size: Are we getting closer?. Cytom. Part A 77, 635–642 (2010).
Yokoya, K., Roberts, A. V., Mottley, J., Lewis, R. & Brandham, P. E. Nuclear DNA amounts in roses. Ann. Bot. 85, 557–561 (2000).
Poland, J. A., Brown, P. J., Sorrells, M. E. & Jannink, J.-L. Development of high-density genetic maps for barley and wheat using a novel two-enzyme genotyping-by-sequencing approach. PLoS ONE 7, e32253 (2012).
Catchen, J., Hohenlohe, P. A., Bassham, S., Amores, A. & Cresko, W. A. Stacks: An analysis tool set for population genomics. Mol. Ecol. 22, 3124–3140 (2013).
Weiß, C. L., Pais, M., Cano, L. M., Kamoun, S. & Burbano, H. A. nQuire: A statistical framework for ploidy estimation using next generation sequencing. BMC Bioinform. 19, 122 (2018).
Corrêa, A., dos Santos, R., Goldman, G. H. & Riaño-Pachón, D. M. ploidyNGS: Visually exploring ploidy with next generation sequencing data. Bioinformatics 33, 2575–2576 (2017).
Li, H. & Durbin, R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25, 1754–1760 (2009).
Li, H. A statistical framework for SNP calling, mutation discovery, association mapping and population genetical parameter estimation from sequencing data. Bioinformatics 27, 2987–2993 (2011).
R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing (2013).
Frichot, E. & François, O. LEA: An R package for landscape and ecological association studies. Methods Ecol. Evol. 6, 925–929 (2015).
Jombart, T. adegenet: A R package for the multivariate analysis of genetic markers. Bioinformatics 24, 1403–1405 (2008).
Gruber, B., Unmack, P. J., Berry, O. F. & Georges, A. dartr: An R package to facilitate analysis of SNP data generated from reduced representation genome sequencing. Mol. Ecol. Resour. 18, 691–699 (2018).
Bradley, M. raxml_ascbias. GitHub https://github.com/btmartin721/raxml_ascbias (2018).
Stamatakis, A. RAxML version 8: A tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics 30, 1312–1313 (2014).
Minh, B. Q. et al. IQ-TREE 2: New models and efficient methods for phylogenetic inference in the genomic era. Mol. Biol. Evol. 37, 1530–1534 (2020).
Lewis, P. O. A likelihood approach to estimating phylogeny from discrete morphological character data. Syst. Biol. 50, 913–925 (2001).
Guindon, S. et al. New algorithms and methods to estimate maximum-likelihood phylogenies: Assessing the performance of PhyML 3.0. Syst. Biol. 59, 307–321 (2010).
Minh, B. Q., Nguyen, M. A. T. & von Haeseler, A. Ultrafast approximation for phylogenetic bootstrap. Mol. Biol. Evol. 30, 1188–1195 (2013).
Huson, D. H. & Bryant, D. Application of phylogenetic networks in evolutionary studies. Mol. Biol. Evol. 23, 254–267 (2005).
Rambaut, A. FigTree v1.3.1. (2010).
Kalinowski, S. T. hp-rare 1.0: A computer program for performing rarefaction on measures of allelic richness. Mol. Ecol. Notes 5, 187–189 (2005).
Brummitt, R. World geographical scheme for recording plant distributions. (2001).
Britton, T., Anderson, C. L., Jacquet, D., Lundqvist, S. & Bremer, K. Estimating divergence times in large phylogenetic trees. Syst. Biol. 56, 741–752 (2007).
Matzke, N. J. BioGeoBEARS: BioGeography with Bayesian (and likelihood) evolutionary analysis with R scripts. Version 1.1. 1, published on GitHub on 6 November 2018. (2018).
Matzke, N. J. Model selection in historical biogeography reveals that founder-event speciation is a crucial process in island clades. Syst. Biol. 63, 951–970 (2014).
Matzke, N. J. Probabilistic historical biogeography: New models for founder-event speciation, imperfect detection, and fossils allow improved accuracy and model-testing. Front. Biogeogr. 5, 2 (2013).
Ronquist, F. Dispersal-vicariance analysis: A new approach to the quantification of historical biogeography. Syst. Biol. 46, 195–203 (1997).
Strömberg, C. A. E. Evolution of grasses and grassland ecosystems. Annu. Rev. Earth Planet. Sci. 39, 517–544 (2011).
Linder, H. P., Lehmann, C. E. R., Archibald, S., Osborne, C. P. & Richardson, D. M. Global grass (Poaceae) success underpinned by traits facilitating colonization, persistence and habitat transformation. Biol. Rev. 93, 1125–1144 (2017).
Devyatkin, E. V. Meridional distribution of Pleistocene ecosystems in Asia: Basic problems. Stratigr. Geol. Correl. 1, 77–83 (1993).
Arkhipov, S. A. & Volkova, V. S. Geological history of Pleistocene landscapes and climate in West Siberia. (1994).
Akhmetyev, M. A. et al. Chapter 8: Kazakhstan and Central Asia (plains and foothills). In Cenozoic Climatic and Environmental Changes in Russia (Geological Society of America, 2005). https://doi.org/10.1130/0-8137-2382-5.139.
Arkhipov, S. A. et al. Chapter 4: West Siberia. In Cenozoic Climatic and Environmental Changes in Russia (Geological Society of America, 2005). https://doi.org/10.1130/0-8137-2382-5.67.
Li, Q. Q. et al. Phylogeny and biogeography of Allium (Amaryllidaceae: Allieae) based on nuclear ribosomal internal transcribed spacer and chloroplast rps16 sequences, focusing on the inclusion of species endemic to China. Ann. Bot. 106, 709–733 (2010).
Hais, M., Komprdová, K., Ermakov, N. & Chytrý, M. Modelling the last glacial maximum environments for a refugium of Pleistocene biota in the Russian Altai mountains Siberia. Palaeogeogr. Palaeoclimatol. Palaeoecol. 438, 135–145 (2015).
Fedeneva, I. N. & Dergacheva, M. I. Paleosols as the basis of environmental reconstruction in Altai mountainous areas. Quat. Int. 106–107, 89–101 (2003).
Braun-Blanquet, J. & Bolòs i Capdevila, O. de. Les groupements végétaux du bassin moyen de l’Ebre et leur dynamisme. An. la Estac. Exp. Aula Dei 5, 1–266 (1957).
Tutin, T., Webb, D., Heywood, V., Walters, S. & Moore, D. Flora Europaea (Cambridge University Press, 1993).
Heklau, H. Proposal to conserve the name Krascheninnikovia against Ceratoides (Chenopodiaceae. Taxon 55, 1044–1045 (2006).
Davis, P. H. Flora of Turkey and the east Aegean islands (Edinburgh University Press, 1988).
Welsh, S., Atwood, N., Higgins, L. & Goodrich, S. A Utah Flora. Gt. Basin Nat. 9, 123 (1987).
Täckholm, V. Students’ Flora of Egypt (Cairo University Publishing, 1974).
Komarov, V. Flora of the U.R.S.S (Academiae Sciencitarum U.R.S.S, 1964).
Rechinger, K. Flora Iranica (Akademische Druck- und Verlagsanstalt, 1963).
Crawford, K. M. & Whitney, K. D. Population genetic diversity influences colonization success. Mol. Ecol. 19, 1253–1263 (2010).
Hilbig, W. Vegetation of Mongolia (SPB Academic Pubishing, 1995).
Briggs, J. C. Chapter 7 Neogene. In Global Biogeography Vol. 14 (ed. Briggs, J. C.) 147–189 (Elsevier, Amsterdam, 1995).
Yurtsev, B. A. The Pleistocene ‘Tundra-steppe’ and the productivity paradox: The landscape approach. Quat. Sci. Rev. 20, 165–174 (2001).
Stewart, J. R., Lister, A. M., Barnes, I. & Dalén, L. Refugia revisited: Individualistic responses of species in space and time. Proc. Biol. Sci. 277, 661–671 (2010).
Varga, Z. Extra-Mediterranean refugia, post-glacial vegetation history and area dynamics in eastern Central Europe. Relict Species https://doi.org/10.1007/978-3-540-92160-8_3 (2009).
Willis, K. J. & Vanandel, T. Trees or no trees? The environments of central and eastern Europe during the Last Glaciation. Quat. Sci. Rev. 23, 2369–2387 (2004).
Tremetsberger, K. et al. Pleistocene refugia and polytopic replacement of diploids by tetraploids in the Patagonian and Subantarctic plant Hypochaeris incana (Asteraceae, Cichorieae). Mol. Ecol. 18, 3668–3682 (2009).
Acknowledgements
We thank Stephanie Köttl, Leonid Rasran and Nora Stoeckl (Universität für Bodenkultur Wien, Austria), Phil Leitner (California State University, USA), and Peter Schönswetter (Universität Innsbruck, Austria) for providing us with samples of Krascheninnikovia ceratoides, and several others for collecting with us. This research was funded in whole by the Austrian Science Fund (FWF)[project I 3002-B25]. For the purpose of open access, the author has applied a CC BY public copyright licence to any Author Accepted Manuscript version arising from this submission.
Author information
Authors and Affiliations
Contributions
The project was designed by K.-G.Be., H.H., N.F., B.N. and F.Bl.; N.F., A.Sh., P.V. and O.Ba. organised field trips, during which all of the authors collected leaf material; lab work was done by A.Se. and S.P.; A.Se., K.T., A.Z.C. and S.P. analysed the data statistically; A.Se. and K.T. discussed the results. All authors contributed to the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Seidl, A., Tremetsberger, K., Pfanzelt, S. et al. The phylogeographic history of Krascheninnikovia reflects the development of dry steppes and semi-deserts in Eurasia. Sci Rep 11, 6645 (2021). https://doi.org/10.1038/s41598-021-85735-z
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-021-85735-z
This article is cited by
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.
