Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2020 May 1;11(1):2179.
doi: 10.1038/s41467-020-15641-x.

Recent Hybrids Recapitulate Ancient Hybrid Outcomes

Affiliations
Free PMC article

Recent Hybrids Recapitulate Ancient Hybrid Outcomes

Samridhi Chaturvedi et al. Nat Commun. .
Free PMC article

Abstract

Genomic outcomes of hybridization depend on selection and recombination in hybrids. Whether these processes have similar effects on hybrid genome composition in contemporary hybrid zones versus ancient hybrid lineages is unknown. Here we show that patterns of introgression in a contemporary hybrid zone in Lycaeides butterflies predict patterns of ancestry in geographically adjacent, older hybrid populations. We find a particularly striking lack of ancestry from one of the hybridizing taxa, Lycaeides melissa, on the Z chromosome in both the old and contemporary hybrids. The same pattern of reduced L. melissa ancestry on the Z chromosome is seen in two other ancient hybrid lineages. More generally, we find that patterns of ancestry in old or ancient hybrids are remarkably predictable from contemporary hybrids, which suggests selection and recombination affect hybrid genomes in a similar way across disparate time scales and during distinct stages of speciation and species breakdown.

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1
Fig. 1. Ancient versus contemporary hybrids.
Conceptual overview and a comparative summary of genomic patterns expected in ancient hybrids (a) versus contemporary hybrids (b). Histograms show narrow (ancient) versus wide (contemporary) distributions of hybrid indexes. In ancient hybrids, ancestry blocks (gray vs. blue segments) have been broken up by recombination and some have stabilized, that is, fixed within the hybrid lineage. In contemporary hybrids, larger ancestry blocks are expected and these vary more among individuals. Plots show the expected effect of selection on ancestry frequencies and patterns of introgression in ancient and contemporary hybrids, respectively. In ancient hybrids, selection (pink arrows) shifts ancestry frequencies. In contemporary hybrids, selection shifts genomic clines for individual loci relative to the genome-wide average (dashed line). c Diagram represents the hypothesized history of hybridization in Lycaeides. Our results suggest that Jackson Hole Lycaeides are ancient hybrids, with ancestry blocks from L. idas and L. melissa (akin to panel a), whereas Dubois are contemporary hybrids with ancestry blocks from Jackson Hole Lycaeides and L. melissa (akin to panel b with colors denoting Jackson Hole vs. L. melissa ancestry).
Fig. 2
Fig. 2. Summary of population genetic patterns.
a Map of sample locations with points of shapes based on nominal taxa and colored for different populations within taxa (Supplementary Table 1). b Ordination of genetic variation via principal component analysis (PCA). Points denote individuals (a few low-coverage individuals were removed for visualization). c Boxplots of admixture proportion estimates from entropy with k = 2 source populations for all populations included in the study (n = 835 butterflies from 23 populations). Boxes denote the 1st and 3rd quartile with the median given by the midline; whiskers extend to the minimum and maximum value or 1.5x the interquartile range with points for more extreme values. Tick marks below the plots identify populations based on the population abbreviations in Supplementary Table 1. d Boxplots of admixture proportion estimates from entropy with k = 2 source populations for all populations except L. idas (N = 659 butterflies from 18 populations), and boxes defined as in panel (c). Source data are provided as a Source Data file.
Fig. 3
Fig. 3. Patterns of ancestry in the ancient hybrids.
Boxplots show the distribution of L. idas ancestry across ancestry-informative SNPs (AIMs) for each linkage group in two representative Jackson Hole Lycaeides populations—Bald Mountain, WY (BLD, n = 74 butterflies) (a) and Pinnacle, WY (PIN, n = 20 butterflies) (b). See Supplementary Fig. 16 for additional populations. The Z-sex chromosome is shown in red. Boxes denote the 1st and 3rd quartile with the median given by the midline; whiskers extend to the minum and maximum value or 1.5x the interquartile range with points for more extreme values. Panels (c) and (d) show maps with pie charts reflecting the proportion of L. idas and L. melissa ancestry (mean) for the Z chromosome (c) and autosomes (d) for each of the nine populations (see Supplementary Table 1 for population IDs). Source data are provided as a Source Data file.
Fig. 4
Fig. 4. Summary of the genomic cline analysis.
a The histogram depicts the distribution of hybrid indexes in the Dubois hybrid zone. b This plot shows estimated genomic clines for a subset of ancestry-informative SNPs (AIMs). Each solid line gives the estimated probability of Jackson Hole (JH) ancestry for an AIM. Green lines denote cases of credible directional introgression (95% CIs for α that exclude zero) and purple lines denote credible cases of restricted introgression (95% CIs for β > 0) (gray lines denote clines not credibly different from the genome average). The dashed line gives the null expectation based on genome-wide admixture. Boxplots show the distribution of cline parameters α (c) and β (d) across loci for each linkage group (based on n = 115 butterflies). Boxes denote the 1st and 3rd quartile with the median given by the midline; whiskers extend to the minum and maximum value or 1.5× the interquartile range with points for more extreme values. Source data are provided as a Source Data file.
Fig. 5
Fig. 5. Expected and observed numbers of SNPs with exceptional patterns of introgression in the Dubois hybrid zone and extreme ancestry frequencies in Jackson Hole Lycaeides.
Panels af show results when considering the top 10% of AIMs in each category. Histograms give null expectations from randomization tests, and vertical solid lines show the observed number of AIMs exhibiting a given pattern. Comparisons shown are directional introgression of Jackson Hole alleles (high α) and high L. idas ancestry (in Jackson Hole) (a), directional introgression of L. melissa alleles (low α) and high L. melissa ancestry (b), directional introgression of Jackson Hole alleles (high α) and high L. melissa ancestry (c), directional introgression of L. melissa alleles (low α) and high L. idas ancestry (d), restricted introgression (high β) and high L. idas ancestry (e), and restricted introgression (high β) and high L. melissa ancestry (f). Panels gi show how these results are affected by considering different levels of stringency (i.e., by examining the most extreme 10% to the top 1% of AIMs with each pattern), and when considering only the Z chromosome (g) or only the autosomes (h). Here, circles denote the ratio of the observed to expected overlap from the null, and the circles are filled (P ≤ 0.05) or not (P  > 0.05) to denote whether the overlap is greater than expected by chance from a one-sided randomization test. Source data are provided as a Source Data file.
Fig. 6
Fig. 6. Whole-genome phylogenetic analyses.
Panel (a) shows a map of the sampling localities in the western United States; Dubois is included for reference but was not sampled for whole-genome phylogenetics. The Jackson Hole ancestry cline (i.e., the range of the ancient Jackson Hole hybrids) is shown for reference as well (red zone). Panel b shows the midpoint-rooted maximum likelihood phylogram inferred from the whole-genome SNP set (2,013,201 autosomal SNPs). Panels (ce) show three common unrooted tree topologies inferred from 1000 SNP windows. The tree in panel (c) is the most common topology and matches the whole-genome tree in b. Trees in panels d and e differ by grouping the Warner Mts. population with L. melissa or the Sierra Nevada population, respectively. Panels (fh) show the proportion of 1000 SNP topologies matching the trees in ce for autosomal windows, Z-chromosome windows, and the 1000 SNP windows that contain the candidate barrier loci (see main text). There is a significant deficit of the topology in d on the Z chromosome and among the barrier loci (x-fold = 0.10, one-sided P < 0.001, and x-fold = 0.43, one-sided P = 0.047, respectively), and a significant excess of topology (e) on the Z chromosome and among the barrier loci (x-fold = 5.36, P < 0.001, and x-fold = 3.62, P = 0.008). See Supplementary Fig. 25 for the full set of tree topologies recovered. Source data are provided as a Source Data file.

Similar articles

See all similar articles

References

    1. Arnold, M. L. Natural Hybridization and Evolution (Oxford University Press, 1997).
    1. Mallet J. Hybridization as an invasion of the genome. Trends Ecol. Evolution. 2005;20:229–237. doi: 10.1016/j.tree.2005.02.010. - DOI - PubMed
    1. Prüfer K, et al. The complete genome sequence of a Neanderthal from the Altai Mountains. Nature. 2014;505:43. doi: 10.1038/nature12886. - DOI - PMC - PubMed
    1. De Manuel M, et al. Chimpanzee genomic diversity reveals ancient admixture with bonobos. Science. 2016;354:477–481. doi: 10.1126/science.aag2602. - DOI - PMC - PubMed
    1. Taylor SA, Larson EL. Insights from genomes into the evolutionary importance and prevalence of hybridization in nature. Nat. Ecol. Evolution. 2019;3:170. doi: 10.1038/s41559-018-0777-y. - DOI - PubMed
Feedback