The controversy about the role of hybridization in evolution began almost 300 years ago when the Swedish botanist Carl Linnaeus posited that new forms of plants could arise by hybridization (2<\/em><\/a>). This suggestion was criticized by members of the clergy because it challenged the immutability of species. Darwin (3<\/em><\/a>) also expressed skepticism about the importance of hybridization, albeit for scientific reasons. In the 20th century, the prevailing view among evolutionary biologists aligned closely with Darwin’s, holding that hybridization was rare, and when it did occur, it was likely to be maladaptive and transitory, leading either to the completion of speciation or to the merging of hybridizing entities (4<\/em><\/a>,\u00a05<\/em><\/a>). The alternative view held that hybridization was more common than generally appreciated, but even when rare, it could provide useful genetic variation that could be exploited through natural selection (6<\/em><\/a>).<\/p>\n <\/p>\n Variation in wing patterns in\u00a0Heliconius<\/em>\u00a0butterflies defines different species, which often have admixed genomes.<\/p>\n <\/p>\n <\/p>\n With fully sequenced genomes now available for many organisms, as well as methods for detecting the genomic signatures of past hybridization events (7<\/em><\/a>), it is now possible to assess both the presence and genomic extent of admixture (the presence of DNA in an individual that derives from distantly related lineages, often due to hybridization). Admixture has been detected in numerous genomes across the domains of life, including more than 70 cases where it is known to be adaptive (enhance evolutionary fitness) (8<\/em><\/a>). The extent of genetic exchange is such that some biologists have argued that the use of trees for depictions of the history of life should be abandoned in favor of networks, which can depict both branching and reticulate evolutionary histories (9<\/em><\/a>).<\/p>\n Heliconius<\/em>\u00a0butterflies are endemic to tropical America and comprise \u223c40 species, which are known for the spectacular diversity (and beauty) of their wing patterns (see the photo). The colorful wing patterns not only warn predators of a butterfly’s distastefulness, they also offer cues for mate recognition, which contributes to speciation by promoting breeding within species. Hybridization underlies the origin and exchange of wing patterns, aiding the establishment of new hybrid species (10<\/em><\/a>), as well as the evolution of M\u00fcllerian mimics (11<\/em><\/a>), in which different unpalatable species share similar warning wing patterns. However, these inferences have been questioned (12<\/em><\/a>), partly because of technical difficulties in distinguishing between admixture (hybridization) and the stochastic sorting of ancestral polymorphisms, called incomplete lineage sorting (ILS). The latter occurs when there is not enough time for genetic drift to resolve polymorphic loci (that is, the extinction of all but one allele at a locus) prior to the next speciation event. Consequently, genealogies frequently differ from the species phylogeny, imitating phylogenomic patterns caused by hybridization.<\/p>\n To estimate the importance of hybridization and introgression (interspecific genetic exchange by hybridization) in\u00a0Heliconius<\/em>, Edelman\u00a0et al.<\/em>\u00a0generated 20 new genome assemblies for\u00a0Heliconius<\/em>\u00a0species and related genera. The authors then determined the genealogies of loci from across the\u00a0Heliconius<\/em>\u00a0genome. Most genealogies were incongruent: They differed from each other and from a consensus \u201cspecies\u201d phylogeny for the group, suggesting that ILS and\/or admixture were common.<\/p>\n To distinguish between these possibilities, Edelman\u00a0et al.<\/em>\u00a0developed a new method called \u201cquantifying introgression via branch lengths\u201d (QuIBL). The method takes advantage of the observation that the internal branch lengths in a genealogy are expected to be longer on average if incongruence is due to introgression rather than ILS. Extensive simulations over a wide range of demographic conditions indicate that this method is more accurate than previous approaches. When applied to small genomic regions across 13\u00a0Heliconius<\/em>\u00a0genomes, \u223c20% of genomic windows were estimated to have a history of introgression, and \u016f70% of incongruent genealogies were due to introgression compared with 30% from ILS. This implies that even when speciation events occur close together in time, such as in rapidly diversifying groups, hybridization rather than ILS is the dominant cause of genealogical incongruence.<\/p>\n So, is the observed extensive introgression of evolutionary importance? Analyses of the genomic distribution of introgressions indicate that they are associated with high recombination rates. This distribution implies that most introgressions probably are maladaptive (13<\/em><\/a>). Recombination can remove harmful alleles from introgressions (leaving neutral fragments behind) before they are eliminated from the genome by natural selection. The authors also report a new chromosomal inversion that likely represents an example of adaptive introgression. Inversions typically suppress recombination in heterozygotes and can aid adaptive divergence and speciation by preventing combinations of adaptive alleles from being broken up by recombination. This particular inversion includes the\u00a0cortex<\/em>\u00a0locus, which underlies the evolution of wing coloration patterns and mimicry in another\u00a0Heliconius<\/em>\u00a0species (14<\/em><\/a>). Thus, introgression likely has had both harmful and beneficial consequences in\u00a0Heliconius.<\/em><\/p>\n The study of Edelman\u00a0et al.<\/em>\u00a0adds to a growing literature indicating that the footprints of hybridization are a common feature of plant and animal genomes and that hybridization can provide the fuel for adaptive diversification. However, this does not necessarily mean that hybridization is common at the population level. For example, a single successful hybridization event early in the divergence of a lineage could leave traces in all descendants. Thus, an important next step is to estimate the number and diversity of hybridization events required to account for observed patterns of genomic admixture.<\/p>\n <\/p>\n <\/p>\n![\"Embedded](\"https:\/\/science.sciencemag.org\/sites\/default\/files\/highwire\/sci\/366\/6465\/570\/embed\/graphic-1.gif\")
<\/span><\/div>\nPHOTO: MARCUS KRONFORST\/HARVARD UNIVERSITY<\/q><\/div>\n<\/div>\n