Triplicate parallel life cycle divergence despite gene flow in periodical cicadas

ARTICLE DOI: 10.1038/s42003-018-0025-7 OPEN 1234567890():,; Triplicate parallel life cycle divergence despite gene flow in periodical cicadas Tomochika Fujisawa 1, Takuya Koyama1, Satoshi Kakishima2,3, John R. Cooley4,5, Chris Simon5, Jin Yoshimura2,6,7 & Teiji Sota 1 Periodical cicadas comprise three species groups containing three pairs of 13- and 17-year life cycle species showing parallel divergence, along with a more anciently diverged 13-year species (Magicicda tredecim). The mechanism and genetic basis of this parallel divergence is unknown. Here we use orthologous transcriptome sequences to explore the demographic processes and genomic evolution associated with parallel life cycle divergence. The three 13and 17-year species pairs have similar demographic histories, and the two life cycles diverged 200,000–100,000 years ago. Interestingly, these life cycle differences have been maintained despite substantial gene flow between 13- and 17-year species within species groups, which is possible during co-emergences. Sequence divergence between 13- and 17-year species in each species group (excluding M. tredecim) is minimal, and we find no shared divergent single-nucleotide polymorphisms (SNPs) or loci associated with all instances of life cycle divergence. The two life cycles may be controlled by highly limited genomic differences. 1 Department of Zoology, Graduate School of Science, Kyoto University, Sakyo, Kyoto 606-8502, Japan. 2 Graduate School of Science and Technology, Shizuoka University, Hamamatsu 432-8561, Japan. 3 Department of Botany, National Museum of Nature and Science, Tsukuba 305-0005, Japan. 4 College of Integrative Sciences, Wesleyan University, Middletown, CT 06459, USA. 5 Department of Ecology and Evolutionary Biology, University of Connecticut, Storrs, CT 06268-3043, USA. 6 Department of Environmental and Forest Biology, State University of New York College of Environmental Science and Forestry, Syracuse, NY 13210, USA. 7 Marine Biosystems Research Center, Chiba University, Uchiura, Kamogawa, Chiba 299-5502, Japan. Correspondence and requests for materials should be addressed to T.F. (email: ) or to T.S. (email: ) COMMUNICATIONS BIOLOGY | (2018)1:26 | DOI: 10.1038/s42003-018-0025-7 | www.nature.com/commsbio 1 ARTICLE COMMUNICATIONS BIOLOGY | DOI: 10.1038/s42003-018-0025-7 L ife history diversity is a remarkable feature of living organisms and underlies fundamental evolutionary questions1. Periodical cicadas of the genus Magicicada are found only in the eastern United States and are well known for their unusual life history patterns, characterised by prolonged juvenile periods of 13 or 17 years, followed by synchronised mass emergence of adults within local populations2. Only one cohort, or ‘brood’, of periodical cicadas emerges every 13 or 17 years in any given location. There are three co-occurring species groups of periodical cicadas, Decim, Decula and Cassini. Each has one species with a 17-year life cycle and one or two species with a 13-year cycle, and there are seven described species (four 13-year and three 17-year) in total3–5 (Fig. 1). Although the species groups clearly differ in morphology, male songs and female song preferences, the 13-year and 17-year species within each species group are extremely similar or indistinguishable in these characters4,5; thus, the difference in life cycle length is one of the only diagnostic characters for their identification. The three species groups are estimated to have diverged 3.9–2.5 million years ago (mya), and subsequent divergence of the present 13-year (mostly southern) and 17-year (mostly northern) life cycles has occurred in parallel in the three species groups during the Quaternary, except for the first split of the 13-year species, M. tredecim, in the Decim group (0.5 mya) (Fig. 1)6. The synchronisation of prolonged life cycles among species groups is thought to have evolved for a predation-avoidance strategy7, an ecological problem shared among co-occurring species. The divergence of 13-year and 17-year life cycles may have been related to adaptation to climatic changes across glacial cycles; the 4-year extension of juvenile stages may have been advantageous for surviving in cooler northern environments8,9. The genetic basis of life cycle length has not been studied because the long life cycles complicate genetic crosses. An early explanation for life cycle control in periodical cicadas proposed a one-locus, two-allele system in which either the 13- or the 17-year cycle is dominant10,11. Differences between the a two life cycle lengths may be attributable to differences in juvenile developmental rate12,13, which may be regulated by one locus or a small number of loci. However, life cycle regulation in periodical cicadas may not always be strict, because 4-year acceleration and/or deceleration of emergences have been observed in both groups of cicadas, events unlikely to have resulted from fortuitous mass mutation14. These observations have led to the hypothesis that all periodical cicadas possess monomorphic developmental plasticity14 and that this common plasticity underlies the switching of life cycle lengths triggered by environmental cues (e.g., a drastic change in temperature during juvenile development), followed by a genetic change in a life-cycle control locus (genetic accommodation15), which enables a permanent life cycle shift4. In general, parallelism in adaptive character divergence among closely related species results from parallel mutation or selection, ancestral polymorphism with balancing selection, or adaptive introgression16,17. In periodical cicadas, an ancestral polymorphism in life cycle length followed by collateral genetic evolution16 is considered the most parsimonious explanation for the parallel divergence and the formation of synchronous broods among three species groups, because multiple independent acquisitions of identical life cycles are unlikely6. In addition, a hypothesis of life cycle switching via introgressive hybridisation of the putative 13-year allele from 13- to 17-year cicadas has been proposed10,11,18. This hypothesis was used to explain the existence of two 13-year species in the Decim group18, proposing that introgressive hybridisation from the preexisting 13-year species M. tredecim to the 17-year M. septendecim produced the new 13year species M. neotredecim. However, the hybrid origin hypothesis of M. neotredecim was rejected based on population genetic studies5,19,20. The hybrid origin hypothesis of 13-year species is unlikely to be applicable to the Cassini and Decula groups, which have no early diverged 13-year species (unless hybridisation between species groups drove life-cycle switching from 17- to 13-year cycles). b 3.9 mya 45°N 2.5 mya 17-y 35°N 13-y Species group Decim Cassini 25°N Decula 100°W 0.5 mya 90°W 80°W 70°W 13-year 3 broods: M. tredecim M. neotredecim M. tredecassini M. tredecula XIX, XXII, XX (...truncated)