FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4205867985> ?p ?o ?g. }

• W4205867985 abstract "Article Figures and data Abstract Editor's evaluation Introduction Results Discussion Materials and methods Appendix 1 Appendix 2 Appendix 3 Data availability References Decision letter Author response Article and author information Metrics Abstract With accelerating global warming, understanding the evolutionary dynamics of plant adaptation to environmental change is increasingly urgent. Here, we reveal the enigmatic history of the genus Cochlearia (Brassicaceae), a Pleistocene relic that originated from a drought-adapted Mediterranean sister genus during the Miocene. Cochlearia rapidly diversified and adapted to circum-Arctic regions and other cold-characterized habitat types during the Pleistocene. This sudden change in ecological preferences was accompanied by a highly complex, reticulate polyploid evolution, which was apparently triggered by the impact of repeated Pleistocene glaciation cycles. Our results illustrate that two early diversified Arctic-alpine diploid gene pools contributed differently to the evolution of this young polyploid genus now captured in a cold-adapted niche. Metabolomics revealed central carbon metabolism responses to cold in diverse species and ecotypes, likely due to continuous connections to cold habitats that may have facilitated widespread adaptation to alpine and subalpine habitats, and which we speculate were coopted from existing drought adaptations. Given the growing scientific interest in the adaptive evolution of temperature-related traits, our results provide much-needed taxonomic and phylogenomic resolution of a model system as well as first insights into the origins of its adaptation to cold. Editor's evaluation This work has the potential to be of broad interest to scientists seeking to understand the evolutionary dynamics of plants during past periods of rapid climate change. Specifically, within the target genus of Cochlearia, the results indicate increased rates of speciation and diversification in response to pronounced glacial cycles. Future work to establish more direct mechanistic links between the results and conclusions will improve our understanding of adaptation and speciation. https://doi.org/10.7554/eLife.71572.sa0 Decision letter Reviews on Sciety eLife's review process Introduction Vast spatiotemporal variation across natural environments subjects all organisms to abiotic stressors (Gienapp et al., 2008). Dynamic shifts in these stressors lead to migration, adaptation, or extinction (Aitken et al., 2008). Thus, the current acceleration of global warming and climate volatility demands a better understanding of evolutionary dynamics resulting from climate change (Root et al., 2003; Thomas et al., 2004; Jump and Peñuelas, 2005; Visser, 2008; Franks et al., 2014). Further, there is a strong economic rationale for understanding the consequences of environmental change on plants, which typically lack the option of rapidly migrating away from changing conditions (Xoconostle-Cazares et al., 2010; Olsen and Wendel, 2013). An especially powerful natural laboratory for the study of climate change adaptation is represented by the recurrent cycles of glaciation and deglaciation during the Pleistocene. Thus, looking backwards in time by investigating the evolutionary footprints of this epoch can provide valuable insight for our understanding of adaptive evolution. The genus Cochlearia L. represents a promising study system for the evolutionary genomics of adaptation not only because of proximity to Arabidopsis and other Brassicaceae models, but also because of distinctive ecotypic traits which evolved within a short time span (Koch et al., 1996; Koch et al., 1998; Koch et al., 1999; Koch, 2012). Among these are adaptations to extreme bedrock types (dolomite versus siliceous), heavy metal-rich soils, diverse salt habitats, high alpine regions, and life cycle variation. This diversity is accompanied by a remarkably dynamic cytogenetic evolution within the genus. Two base chromosome numbers exist (n=6 and n=7) and out of the 20 accepted taxa, two-thirds are neopolyploids, ranging from tetraploids to octoploids (previous phylogenetic hypotheses are given in Appendix 1—figure 1; see Supplementary file 1 and Appendix 1 for details). The connecting element between the various cytotypes and ecotypes is the cold character of the diverse habitat types standing in sharp contrast to the preferences of the sole outgroup sister genus Ionopsidium, which occurs only in arid Mediterranean habitats (Koch, 2012). These two genera constitute the monophyletic tribe Cochlearieae with a stem group age of approx. 18.9 million years ago (Walden et al., 2020) and which forms with various other tribes from Brassicaceae the rapidly emerging evolutionary lineage II with highest net diversification rates 16–23 million years ago (Walden et al., 2020). In total, the genus Cochlearia comprises 16 accepted species and 4 subspecies (Kiefer et al., 2014, Supplementary file 1). While on species-level it has been shown in Arabidopsis thaliana that drought- and temperature-adaptive genetic variants are shared among Mediterranean and Nordic regions (Exposito-Alonso et al., 2017), the separation of Cochlearia from Ionopsidium is much deeper, dating to the mid-Miocene (Koch, 2012). However, the formation of the genus Cochlearia as we see it today first started much more recently, during the middle (0.77–0.13 mya) and late (0.13–0.012 mya) Pleistocene (Koch, 2012). This long lag between the divergence of the sole outgroup Ionopsidium and Cochlearia raises the hypothesis of a long-lasting footprint of drought adaptation in Cochlearia. The strong association with cold habitats shown by almost all Cochlearia species may therefore be interpreted as a cold preference that was acquired rapidly in adaptation to the intense climatic fluctuations which characterized this epoch. There is a growing interest in the genus Cochlearia from diverse fields (Reeves, 2019; Brock et al., 2006; Dauvergne et al., 2006; Brandrud et al., 2017; Mandáková et al., 2017; Nawaz et al., 2017; Bray et al., 2020), but the evolutionary history of the genus has been highly recalcitrant. Thus it is still unknown how the genus managed the rapid transition from Mediterranean to circum-Arctic or high-alpine habitat types in combination with a highly dynamic cytogenetic evolution. Here, we overcome the first obstacle, presenting the first genus-wide picture, using comprehensive cytogenetic data and highly resolving phylogenomic analyses, complemented by insights into the Cochlearia metabolome response to cold. Herein the metabolome is primarily used as a complex phenotype, which might characterize potentially different bioclimatically defined biomes or species distribution ranges. However, since it has been shown for A. thaliana that the cold metabolome can reflect continental-scale biogeographically defined clines along temperature gradients (Pritchard et al., 2000; Weiszmann et al., 2020), it can be assumed that on evolutionary scales past footprints of diversification might be detectable, and that in principle genomic analyses (e.g., GWAS) may allow to identify candidate genes involved in pathway regulation and metabolic reaction plasticity. Our phylogenetic and cytogenetic analysis uncovers a recurrent boosting of speciation by glaciation cycles in this cytotypically very diverse genus and indicates that, despite clear challenges brought by global warming, the genus survives evolutionarily while, we speculate, rescuing its species diversity with reticulate and polyploid evolution. Results Cytogenetic analyses show geographic structuring and parallel evolutionary trends toward shrinking haploid genomes in higher polyploids In order to first resolve its cytogenetic evolution, we generated a comprehensive survey of 575 georeferenced chromosome counts and/or genome sizes across the Cochlearia genus (Supplementary file 2) based on our novel cytogenetic data (Supplementary file 3) and a review of published literature over the last century (Supplementary file 4). This survey revealed a clear continental-scale geographical partitioning of diploid cytotypes (2n=12 and 2n=14; Figure 1a). Figure 1 Download asset Open asset Distribution and cytogenomic flexibility of Cochlearia. (a) Geographic distribution of chromosome counts for diploid Cochlearia accessions (n=169; Supplementary file 1 and Figure 1—source data 1), showing a clear separation of 2n=12 (European, green) and 2n=14 (Arctic, pink). (b) Geographic distribution of aneuploidies (orange, n=138) and euploidies (black, n=376) in diploid and polyploid Cochlearia (n=514; Figure 1—source data 2). (c) Measured DNA content per chromosome (given in picograms; Figure 1—source data 3) relative to respective total chromosome numbers (red [21 counts]: 2n=2x [non-Arctic], yellow [nine counts]: 2n=2x [Arctic], green [30 counts]: 2n=4x, blue [six counts]: 2n=6x [excluding C. danica], purple [10 counts]: 2n=6x [C. danica], dark grey [two counts]: 2n=8x) showing a significant decline of genome size per chromosome with increasing total chromosome numbers as revealed by Kendall and Spearman rank correlation analyses (78 individuals from 38 accessions representing 14 taxa analyzed in total) (data are not distributed normally and linear regression has not been performed). (d) Measured DNA content per chromosome (given in picograms; Figure 1—source data 4) relative to respective total chromosome numbers excluding Arctic diploids (yellow) and C. danica (purple) as putative outliers (59 individuals, 29 accessions, 11 taxa analyzed in total), showing a significant decline with increasing chromosome numbers via both rank correlation analyses and linear regression analysis (data are normally distributed and linear regression is significant with p=7.11e−10; R²=0.48; QQ-plot given with Appendix 1—figure 3). (e) Images of three Cochlearia species (left: C. pyrenaica [2n=2x=12], top right: C. tatrae [2n=6x=42], bottom right: C. anglica [2n=8x=48]) (data are normally distributed and linear regression is significant with p=7.11e−10; R²=0.48). Figure 1—source data 1 Coordinates of diploid Cochlearia records in the survey of cytogenetic evolution. https://cdn.elifesciences.org/articles/71572/elife-71572-fig1-data1-v2.txt Download elife-71572-fig1-data1-v2.txt Figure 1—source data 2 Coordinates of Cochlearia accessions with documented euploidies and aneuploidies in the survey of cytogenetic evolution. https://cdn.elifesciences.org/articles/71572/elife-71572-fig1-data2-v2.txt Download elife-71572-fig1-data2-v2.txt Figure 1—source data 3 Measured DNA content per chromosome (given in picograms; full data set). https://cdn.elifesciences.org/articles/71572/elife-71572-fig1-data3-v2.txt Download elife-71572-fig1-data3-v2.txt Figure 1—source data 4 Measured DNA content per chromosome (given in picograms; excluding Arctic diploids and C. danica as putative outliers). https://cdn.elifesciences.org/articles/71572/elife-71572-fig1-data4-v2.txt Download elife-71572-fig1-data4-v2.txt We observed highly dynamic genome compositions throughout the genus, with widespread aneuploidies (aberrations from typical species-specific chromosome numbers) and DNA content variation. Aneuploidies are frequently found in polyploid Cochlearia taxa, especially along the coasts (Figure 1b), but they are only rarely spotted in diploids and therefore are nearly absent from Arctic regions, where only diploids are observed (Appendix 1—figure 2). In order to further investigate the cytogenetic dynamics within Cochlearia, we analyzed the relationships of (1) chromosome number versus genome size and (2) chromosome number versus DNA content per chromosome via rank correlation tests (Supplementary file 5) and, if normality of data was given, via linear regression analyses (Figure 1d, Appendix 1—figure 3). Both analyses revealed that (1) the high frequency of polyploidization events is accompanied by increasing genome sizes (Appendix 1—figure 4), while there is (2) a slight but significant negative correlation of chromosome size with increasing chromosome numbers (Figure 1c and d). This trend was independent of inclusion or exclusion of the annual species C. danica and the short-lived Arctic diploids. These taxa were treated as putative outliers because a relationship between lower genome size and annuality was shown as a significant trend for the Brassicaceae as a whole (Hohmann et al., 2015). Organellar phylogenies provide evidence of recurrent glacial speciation boosting We next assessed spatiotemporal patterns of genetic variation first using cytoplasmic and maternally inherited genomes. Extensive past hybridization and reticulate evolution of polyploid taxa results in complex evolutionary scenarios, which are often not well resolved by strict phylogenetic reconstruction using nuclear data. Therefore, conclusions herein are restricted mostly to diploid taxa and respective gene pools. To provide a highly resolved organellar phylogeny, we generated genome sequence data for 65 Cochlearia accessions, representing all accepted Cochlearia species, as well as three species from the sister genus Ionopsidium (Appendix 1—figure 5 and Supplementary file 6). Using these data, complete plastid genomes were assembled de novo for all samples. A maximum likelihood (ML) analysis using RAxML based on our whole chloroplast genome alignment (122,798 bp, excluding one copy of the inverted repeat) covering a total of 5292 SNPs (1003 within Cochlearia) revealed six well-supported major lineages within the genus (Appendix 1—figure 6, congruent to lineages as illustrated in BEAST chronogram, Figure 2). A radiation of plastome diversity was indicated by the existence of several polytomies, despite the generally high resolution of the ML tree. To test if the phylogenetic scenario as revealed by plastid genome analyses was also supported by the mitochondrial genome, we generated a mitochondrial ML phylogeny based on a combination of de novo assembly and referenced-based mapping (Appendix 1—figure 7). The two maternal phylogenies are largely consistent, and after collapsing all branches below a bootstrap support of 95%, no incongruences remained (illustrated by a tanglegram in Appendix 1—figure 8). Figure 2 Download asset Open asset The maternal footprint of recurrent glacial speciation boosting. BEAST chronogram (Figure 2—source data 1) based on complete plastid genome sequence data supplemented by a geographical distribution pattern of the six main phylogenetic lineages (displayed as colored bars next to the tree; dots in the map are colored accordingly; Figure 2—source data 2). The Ionopsidium outgroup lineage is collapsed and condensed. The tree topology is congruent to the topology as revealed from maximum likelihood (ML) analysis. The full BEAST chronogram and the ML tree (incl. bootstrap support values) are given in Appendix 1—figures 6, 9 and 10. Individuals with a base chromosome number of n=7 (shown for all ploidy levels and diploids only) and accessions with an alpine or subalpine habitat type are marked with black dots next to respective tips. Letters (a)–(e) as displayed on the timeline indicate high glacial periods: (a) 640 kya, end of Günz glacial; (b) 450 kya, beginning of Mindel glacial; (c) 250–300 kya, Mindel-Riss inter-glacial; (d) 150–200 kya, Riss glacial; (e) 30–80 kya, Würm glacial. The black dashed line indicates the extent of the Last Glacial Maximum (LGM ~21 kya; based on Ehlers and Gibbard, 2007). The red dashed rectangle highlights a region with evidence for ice-free areas during the Penultimate Glacial Period (~140 kya; based on Colleoni et al., 2016). Figure 2—source data 1 BEAST chronogram in NEXUS format. https://cdn.elifesciences.org/articles/71572/elife-71572-fig2-data1-v2.txt Download elife-71572-fig2-data1-v2.txt Figure 2—source data 2 Geographical distribution of phylogenetic lineages (plastid genome). https://cdn.elifesciences.org/articles/71572/elife-71572-fig2-data2-v2.txt Download elife-71572-fig2-data2-v2.txt Divergence time estimates generated with BEAST based on our whole plastid genome alignment revealed diversification bursts that closely coincide with high glacial periods. We used two secondary calibration points (Ionopsidium/Cochlearia split: 10.81 mya; Cochlearia crown age: 0.71 mya) taken from a large-scale age estimation analysis performed by Hohmann et al., 2015 which included five Cochlearia samples and one Ionopsidium sample that are also included in the present study. The revealed tree topology (Figure 2; full tree given in Appendix 1—figure 9 and Appendix 1—figure 10) is congruent with the topology of the ML tree. In accordance with Hohmann et al., 2015, our BEAST analysis shows a diversification of the entire genus within the last ~660 kya, after a long period of evolutionary stasis and zero net diversification following a deep split from the genus Ionopsidium ~9.25 mya (see also Koch, 2012), and in concert with the beginning of the Pleistocene’s major climatic fluctuations, which are dated to 700 kya (Webb and Bartlein, 1992; Comes and Kadereit, 1998). Thus, diversification times in the six major chloroplast lineages as revealed via both ML and BEAST analysis closely coincide with high glacial periods (Petit et al., 1999; see timeline in Figure 2; Augustin et al., 2004). The most basal Cochlearia chloroplast and mitochondrial haplotypes (yellow lineage – Arctic I) were found in Eastern Canadian C. tridactylites, a species of unknown ploidy (Appendix 2), in a region where ice-free areas putatively occurred during the Penultimate Glacial Period (~140 kya; Colleoni et al., 2016). The earliest diverged organellar genomes of known diploids were found in the Arctic species C. groenlandica and C. sessilifolia (collected in British Columbia, Canada, and Kodiak Island, Alaska, respectively; pink lineage (Arctic II) in Figure 2, see Appendix 1—figure 6 for details), with distribution ranges covering areas such as Beringia that were thought to have served as ice-free Pleistocene refugia (Abbott and Brochmann, 2003). European diploids are found in the green (Arctic-European) and purple (western European) lineages only. Some of the European polyploids, however, harbor early diverged haplotypes from the otherwise pink (Arctic II) lineage. Thus, except for the eastern European C. borzaeana with 2n=8x=48, all taxa from the pink lineage, as well as several taxa from the early diverged blue (Coastal Western European) and orange (Eastern European) lineages, have a base chromosome number of n=7 (see Figure 2). Genomic data and demographic modeling of diploid gene pools indicate glacial expansion In order to analyze the nuclear fraction of our data, we mapped reads of each sample to our C. pyrenaica transcriptome reference (total length: 58,236,171 bp; Lopez et al., 2017). For 63 samples with sufficient nuclear sequence data quality (62 Cochlearia samples and Ionopsidium megalospermum), we generated a phylogenetic network using SplitsTree (Huson, 1998; Huson and Bryant, 2006) based on 447,919 biallelic SNPs. Concordant with our cytogenetic results, the network shows a clear separation of Arctic and European diploid taxa (Figure 3a and b; see Appendix 1—figure 11 for detailed SplitsTree output). Close associations of both C. tridactylites and C. danica with Ionopsidium support the picture as revealed from organellar phylogenies. Further support for the early divergence of these two species came from an ML analysis based on 298,978 variant sites (same set of samples) performed with RAxML using an ascertainment bias correction and a general time-reversible substitution model assuming gamma distribution (Appendix 1—figure 12). Referring to single polyploid taxa, organellar and nuclear phylogenies are not congruent (Appendix 1—figure 6 and Appendix 1—figure 12), which is best explained by the often allopolyploid origin of the tetraploids (e.g., C. bavarica, Appendix 1—figure 13 and Appendix 1—figure 14), which may be even complicated by multiple and polytopic origin and further reticulation. Therefore, we do not discuss here the individual polyploid taxa (but see Appendix 1), instead focusing on the ancestral diploid gene pools. Figure 3 Download asset Open asset Demographic structure and history of the Cochlearia genus based on nuclear genome sequence data. (a) SplitsTree analysis of 62 Cochlearia samples and Ionopsidium (outgroup) using the NeighborNet algorithm based on uncorrected p-distances (Figure 3—source data 1; network with tip labels is given with Appendix 1—figure 11). (b) Geographic distribution of 62 Cochlearia samples. Chart colors correspond to STRUCTURE results (62 Cochlearia samples; 400,071 variants) at K=3 (Figure 3—source data 2); green: Arctic gene pool, red: European gene pool, purple: C. danica-specific cluster (STRUCTURE result with tip labels given with Appendix 1—figure 15). (c) Coalescent models for diploid populations explored with Approximate Bayesian Computation (ABC). SI=strict isolation, CM=continuous migration from the population split to the present, OSC=ongoing secondary contact with gene flow starting after population split and continuing to the present, PSC=past secondary contact with gene flow starting after population split and stopping before the present. (d) Most likely demographic history of diploid EUR (2n=12) and ARC (2n=14) Cochlearia populations from coalescent modeling (based on 22 EUR individuals and 12 ARC individuals; 2140 SNPs at fourfold degenerate sites) and ABC of models without gene flow and upper bound of the population size (N) prior as 400,000 (Supplementary file 10). N is in number of diploid individuals, and time in number of generations ago. Figure 3—source data 1 SplitsTree network in NEXUS format. https://cdn.elifesciences.org/articles/71572/elife-71572-fig3-data1-v2.txt Download elife-71572-fig3-data1-v2.txt Figure 3—source data 2 STRUCTURE result at K=3 (georeferenced) and Delta K result (Evanno Method). https://cdn.elifesciences.org/articles/71572/elife-71572-fig3-data2-v2.txt Download elife-71572-fig3-data2-v2.txt A STRUCTURE analysis of Cochlearia samples only (same variant calling, 400,071 variants after excluding Ionopsidium) with K=3 (optimal K following Evanno Method; Evanno et al., 2005) revealed a pattern very similar to that obtained via SplitsTree, showing the two diploid clusters and a third cluster comprising C. danica samples (Figure 3b and Appendix 1—figure 15). We note potential signatures of admixture between the diploid clusters, especially in Icelandic 2n=12 and 2n=14 diploids and in several polyploid samples such as C. bavarica (discussed in detail in Appendix 1). Interestingly, C. tridactylites, the earliest diverged lineage according to the plastome analysis, was modeled as a mix of the Arctic gene pool and C. danica. In a separate TreeMix (Pickrell and Pritchard, 2012) analysis of all Cochlearia accessions and Ionopsidium (447,919 variants; up to 10 migration events), the bootstrapped graph for m=1 (optimal number of migration events according to Evanno Method; Appendix 1—figure 16) likewise indicates that C. tridactylites is admixed, with a majority ancestry from near the base of the European (excepting hexaploid C. danica) and Arctic groups, and a minority ancestry from the 2n=14 group of Arctic diploids (see Appendix 1—figures 17–19). In order to elucidate the early stages of the Cochlearia species complex formation that might have facilitated the general cold association of the genus, we tested hypotheses regarding the evolutionary history of the diploid lineages from the Arctic (ARC) and European (EUR) distribution ranges by modeling possible histories using a coalescent framework with Approximate Bayesian Computation (ABC; Tavaré et al., 1997; Beaumont et al., 2002). For a data set of 22 European (2n=12) and 12 Arctic (2n=14) individuals (Supplementary file 7 for sampling), we analyzed 2140 fourfold degenerate SNPs (Materials and methods). The sampling of Arctic accessions covers different taxa with deep evolutionary splits as exemplified by plastome analyses and the entire Arctic range is covered, therefore unequal sampling size is expected to have a minor effect if any. Overall, the EUR metapopulation exhibits much higher genetic diversity than the ARC metapopulation (Supplementary file 8), which is in accordance with population-based analyses (Koch et al., 1998). In both populations, 2n=12 and 2n=14, we found an excess of rare alleles (negative Tajima’s D, Supplementary file 8), indicating that they are not in mutation-drift equilibrium. Differentiation and divergence between ARC and EUR were overall very low (Fst ~0.098, dxy ~0.0036, Supplementary file 8). Given the dynamic nature of their ice-age-associated speciation histories, we hypothesized that after the EUR and ARC metapopulations separated, they underwent dramatic changes in effective population sizes (Ne) over time, and that they experienced gene flow. We first tested four different hypotheses regarding the occurrence and relative timing of gene flow between ARC and EUR, because failure to account for gene flow can confound the inference of population size changes. Our gene flow hypotheses were formulated as different model categories (Figure 3c), and random forest ABC (ABC-RF; Marin and Pudlo, 2015) was used to test under which model the observed data was most probable to have arisen. However, discriminating between the four gene flow models was ambiguous (see Supplementary file 9), as the most probable model depended on the priors, particularly on the upper bound of the Ne priors. When allowing Ne up to 400,000, a model of ongoing secondary contact (OSC) prevailed over a model without any gene flow (model SI; posterior probability >0.75, i.e., Bayes Factor >3), but OSC was not clearly better than a model with continuous gene flow (CM) or a past secondary contact (PSC). Yet, when choosing a less informative Ne prior with a greater upper bound of three million, all four models were similarly in agreement with the observed data. In the absence of strong prior information for Ne, we could not establish the occurrence of gene flow between ARC and EUR metapopulations with confidence (see Appendix 3 for further information on ABC model choice). To estimate changes in Ne through time, we fitted parameters of a model without any gene flow (SI) and a model of OSC, considering both high and low Ne upper prior bounds, amounting to a total of four model fits (Figure 3c; see Supplementary file 10). The general pattern of changes in Ne through time were always modeled such that each of ARC and EUR had an older phase of constant Ne followed by a recent phase of exponential expansion or decline. The model without gene flow (SI) contained the fewest parameters, and this model with a smaller Ne prior bound of maximal 400,000 provided the best fit to the data (smallest Euclidean distances between observed and predicted values from posterior predictive checks; see Supplementary file 11). This model fit (Figure 3d) is consistent with the remaining three model fits. Importantly, all model fits agreed about the relative Ne of EUR and ARC: they evolved drastically differently, with the EUR metapopulation having risen to 4–12-fold the Ne of their common ancestral population followed by a moderate decline (0.4–0.8-fold in three out of four models), or constant size up to the present (OSC with smaller Ne upper bound). In contrast, the ARC metapopulation experienced a bottleneck after splitting from the common ancestor (0.2–0.5-fold), followed by a dramatic expansion of 9–52-fold (Supplementary file 10). Estimated Ne for the ancestral population and for ARC during the ancient phase were robust to the choice of model and priors, but other Ne parameters, in particular the ancient phase of EUR and the recent phase of ARC (i.e., the phases in which their Ne were largest), increased when the prior’s upper bound was increased. However, the relative Ne trends through time were robust to these uncertainties, as mentioned above. If EUR and ARC did not experience gene flow (SI model fit), they must have separated only about 65,000–73,000 generations ago, corresponding to 0.2–3 Ne units. This estimate was robust to the choice of priors and may coincide with the last interglacial period (considering a 2-year average generation time; Abs, 1999). If gene flow is assumed (OSC), this split could have occurred earlier (119,000–227,000 generations ago; Supplementary file 10). Further parameters were poorly estimable as indicated by large prediction error, and little deviation between prior and posterior. These include the timing of the transitions from ancient to recent phases of N, the timing of migration, and the migration rates. Considering our BEAST analyses from plastome data, a split-time of 65,000–73,000 generations ago is more likely (e.g., dating of splits within the green evolutionary lineage). An important implication of this is that polyploid inland taxa, such as alpine hexaploid C. tatrae from the High Tatra mountains, showing footprints of both diploid gene pools cannot have evolved earlier than during the Last Interglacial. The example of hexaploid C. bavarica showed a postglacial origin and footprint of both diploid gene pools dated with approximately 12–15,000 years ago predating rapid Holocenic temperature increase. Genus-wide cold response characterized by metabolic profiling indicates an ancient origin of cold temperature tolerance We hypothesized that a very early evolved tolerance to cold facilitated the observed widespread adaptation to alpine and subalpine habitats across the Cochlearia genus. To test this h" @default.
• W4205867985 created "2022-01-26" @default.
• W4205867985 creator A5013600945 @default.
• W4205867985 creator A5069578294 @default.
• W4205867985 date "2021-08-11" @default.
• W4205867985 modified "2023-09-28" @default.
• W4205867985 title "Decision letter: Evolutionary footprints of a cold relic in a rapidly warming world" @default.
• W4205867985 doi "https://doi.org/10.7554/elife.71572.sa1" @default.
• W4205867985 hasPublicationYear "2021" @default.
• W4205867985 type Work @default.
• W4205867985 citedByCount "0" @default.
• W4205867985 crossrefType "peer-review" @default.
• W4205867985 hasAuthorship W4205867985A5013600945 @default.
• W4205867985 hasAuthorship W4205867985A5069578294 @default.
• W4205867985 hasBestOaLocation W42058679851 @default.
• W4205867985 hasConcept C2778226939 @default.
• W4205867985 hasConcept C39432304 @default.
• W4205867985 hasConcept C42407357 @default.
• W4205867985 hasConcept C86803240 @default.
• W4205867985 hasConceptScore W4205867985C2778226939 @default.
• W4205867985 hasConceptScore W4205867985C39432304 @default.
• W4205867985 hasConceptScore W4205867985C42407357 @default.
• W4205867985 hasConceptScore W4205867985C86803240 @default.
• W4205867985 hasLocation W42058679851 @default.
• W4205867985 hasOpenAccess W4205867985 @default.
• W4205867985 hasPrimaryLocation W42058679851 @default.
• W4205867985 hasRelatedWork W2006523268 @default.
• W4205867985 hasRelatedWork W2089655346 @default.
• W4205867985 hasRelatedWork W2349709498 @default.
• W4205867985 hasRelatedWork W2352677790 @default.
• W4205867985 hasRelatedWork W2360106954 @default.
• W4205867985 hasRelatedWork W2366882156 @default.
• W4205867985 hasRelatedWork W2384598669 @default.
• W4205867985 hasRelatedWork W2899084033 @default.
• W4205867985 hasRelatedWork W2998126594 @default.
• W4205867985 hasRelatedWork W2588135888 @default.
• W4205867985 isParatext "false" @default.
• W4205867985 isRetracted "false" @default.
• W4205867985 workType "peer-review" @default.