Genomics Reveals Abundant Speciation in the Coral Reef Building Alga Porolithon Onkodes (Corallinales, Rhodophyta)1

L ETTER

GENOMICS REVEALS ABUNDANT SPECIATION IN THE CORAL REEF BUILDING ALGA POROLITHON ONKODES (CORALLINALES, RHODOPHYTA)1

Paul W. Gabrielson , Jeffery R. Hughey and Guillermo Diaz-Pulido

An essential suite of coral reef ecosystem engineers is coralline red algae. Among these, the smooth, encrusting Porolithon onkodes has historically been considered the most important and common reef building species worldwide. We assess P. onkodes biodiversity by performing a genomic analysis of the lectotype specimen collected in 1892 from the Tami Islands, Gulf of Huon, east of New Guinea. Comparisons of DNA sequences from the lectotype specimen to those deposited in GenBank and to newly generated sequences from both ﬁeld-collected and historical specimens demonstrate that at least 20 distinct species are passing under P. onkodes. We hypothesize that there were multiple evolutionary drivers including ecophysiology, hydrodynamic regimes, and biotic interactions as well as historical biogeography, which resulted in this high diversity of smooth, encrusting Porolithon species throughout the tropics. Our results emphasize the need to document the biodiversity, ecophysiology, and habitats of these tropical, reef-building algae in light of climate change and ocean acidiﬁcation.

Porolithon onkodes is currently 2014). Both of these Porolithon his new genus, Porolithon (Foslie recognized worldwide as the species are distinguished by their 1909). Following earlier successes major coral reef cementing and smooth, encrusting habitat in con- characterizing organellar genomes armoring coralline red alga, pro- trast to other fore reef corallines of red algal type specimens viding structural integrity to fore that form densely branched more (Hughey et al. 2014, 2017, Boo reef habitats (Maneveldt and or less hemispherical mounds. et al. 2016), we analyzed the lec- Keats 2014). The importance of Beginning in the late 1960s and totype of L. onkodes and compared nongeniculate coralline red algae continuing to the present, the its DNA to recent field-collected to coral reef formation was first ecological niches thought to be specimens. The lectotype DNA recognized by Charles Darwin occupied by P. onkodes have was extracted from a 2 9 2mm2 (1890), “Nothing can be more sin- expanded to include back reef fragment following Hernandez- gular than the appearance at low rubble-filled moats as well as slope Kantun et al. (2016) and imple- tide of this ‘flat’ of naked stone, habitats to 5–10 m depth menting the guidelines proposed especially where it is externally (Womersley and Bailey 1970, Lee by Hughey and Gabrielson bounded the smooth convex 1978, Adey et al. 1982). (2012). Illumina 76 bp paired-end mound of Nulliporae appearing Seven coralline species were library and sequencing was car- like a breakwater built to resist described based on specimens ried out by myGenomics LLC using the waves.” Ever since, throughout from the Tami Islands, Gulf of an in-house low-input protocol the tropical Pacific and Indian Huon, collected by native divers (http://www.mygenomics.com). Oceans, P. onkodes has been cited in 1–3 m of water in March 1892 For all other samples, the DNA as the most important species pro- and communicated by the was extracted, amplified, and San- viding major cover on windward Lutheran missionary Rev. Georg ger sequenced following Gabriel- fore reefs from the intertidal reef Bamler, including the coral tine son et al. (2011) and Adey et al. ridge and continuing seaward encrusting Lithothamnion onkodes (2015). Genomic data were assem- through the reef crest and rim (Heydrich 1897), basionym of bled with the default de novo set- (Howe 1912, Setchell 1926, Taylor P. onkodes (Fig. S1 in the Support- tings in CLC Genomics â 1950, Doty and Morrison 1954, ing Information). Much of this Workbench 9.5.2 ( 2016 CLC bio, Marsh 1970, Littler and Doty collection was destroyed in the a QIAGEN Company, German- 1975, Dean et al. 2015). Its eco- bombing of the Berlin Museum town, MD, USA), Velvet 1.2.08 logical counterpart in the Carib- during World War II, but before with a kmer of 59 (Zerbino and bean Sea was known as Porolithon 1909, Heydrich sent a specimen Birney 2008), and default settings pachydermum (Adey 1967), until of L. onkodes to his coralline com- in IDBA-UD (Peng et al. 2012). placed into synonymy under petitor, Mikhail Foslie, in Trond- The three assemblers indepen- P. onkodes (Maneveldt and Keats heim, Norway, who placed it in dently yielded the same results.

429 430 PAUL W. GABRIELSON ET AL.

Genome gaps were closed by PCR and contains 228 genes dermum are the same species, with and sequencing and by iterative (Table S2). It has a high level of P. antillarum having nomenclatu- mapping with Geneious 8.1.5 gene synteny with the two previ- ral priority. Just as there are multi- (BioMatters, Auckland, New Zeal- ously published genomes of Cal- ple species passing under and). Genomes were annotated liarthron tuberculosum and P. onkodes in the Indo-Pacific, using NCBI ORFfinder and align- Sporolithon durum (Janouskovec there are at least two other species ments obtained via BLASTX et al. 2013, Lee et al. 2016). The passing under P. antillarum in the against the reference sequences of mitogenome of Porolithon onkodes tropical western Atlantic (Fig. 1, Calliarthron tuberculosum (GenBank is 27,882 bp and contains 46 represented by MF979946 and accessions KC153978 and genes (Table S3). The mitogen- MF979960). There are four KR005619). tRNAs were identified ome is similar in gene content to unnamed specimens in GenBank using tRNAscan-SE 1.21 web ser- other corallines (C. tuberculosum, (LCB0701, LCB9658, LCB0712, ver (Schattner et al. 2005) and Corallina officinalis, and S. durum); LCB0678), all from Fiji and all rRNAs using RNAmmer 1.2 (Lage- however, it differs in organization with vague habitat data (“isolated ETTER sen et al. 2007). Locally Collinear (Fig. S2 in the Supporting Infor- by scuba diving or on the shore at L Block (LCB) alignments were gen- mation). Most noteworthy, riboso- low tide”), whose psbA erated using ProgressiveMauve mal proteins rps12 and rpl20, the (GQ917931, GQ917955, GQ917 with a seed of 21 with the “Use small subunit rRNA, and nad4L of 964, GQ917969) and mitochon- seed families” option selected the NR segment (Yang et al. drial-encoded COI (GQ917588, (Darling et al. 2010). Sequences 2015) are situated with the CY GQ917571, GQ917592, GQ917 of the plastid genome (GenBank segment, and the tRNAs (R, Y, V, 276) gene sequences indicate that KY212106), mitogenome (Gen- N) are flanking orf172 following they are P. onkodes (for psbA <0.4% Bank KY212107), and nuclear the large subunit rRNA. diverged, Hind et al. 2016, for ribosomal cistron (GenBank Maximum Likelihood (ML) COI <4.5% diverged, Saunders KY212108) were deposited in Gen- phylogenetic analyses of the plas- 2005). None of the other Bank. DNA sequences of the rbcL tid-encoded rbcL gene (Fig. 1) and P. onkodes GenBank sequences of gene and multiple genes (SSU, multigene trees (Fig. S3 in the any marker, including those listed LSU, psbA, COX1, 23S, rbcL) rep- Supporting Information), based as Hydrolithon onkodes, are resenting published and unpub- on newly generated sequences or P. onkodes, but all belong in Poroli- lished Porolithon, and unidentified sequences from GenBank thon except KM369154 from New coralline algae were downloaded (Table S3), showed that over 20 Zealand (Fig. 1). The distributions from GenBank (Table S1 in the species are passing under the of the Porolithon species for which Supporting Information) and name Porolithon onkodes. rbcL we have rbcL sequences are shown aligned with MAFFT (Katoh and sequences from the 30 end of the in Figure 2. Species level status is Standley 2013). Maximum likeli- gene (rbcL-3P) were also obtained based on rbcL pair-wise sequence hood analyses were executed with from the type specimens of divergences >1% (Table S4 in the T-REX (Boc et al. 2012) and the P. sandvicense (basionym: Lithophyl- Supporting Information), a con- GTR + gamma model with 1,000 lum dentatum f. sandvicensis, type servative value given that 0.5%– fast bootstraps. The Bayesian anal- locality Sandwich Islands, now 0.8% has been used to distinguish ysis (Ronquist and Huelsenbeck Hawaiian Islands), P. oligocarpum coralline species in other genera 2003) was performed using the (basionym: Lithophyllum oligo- of Corallinales (Gabrielson et al. same model with default settings carpum, type locality: Puerto Oro- 2011, Hind et al. 2016). in Geneious. The rbcL tree was tava [Puerto de la Cruz], Tenerife, It is clear from the DNA rooted with Harveylithon sp. (Gen- Canary Islands), P. pachydermum sequence data based on multiple Bank MF979962) and the multi- (basionym: Lithophyllum onkodes f. genetic markers that the evolution gene tree with Harveylithon rupestre pachydermum, type locality: Vestin- of tropical and subtropical (SSU, GenBank KM073303; psbA, dien = Danish West Indies, now smooth, encrusting Porolithon spe- KM407535). Phylogenetic trees U.S. Virgin Islands), and P. antil- cies is not marked by morpho-ana- were visualized with TreeDyn larum (basionym: Lithophyllum tomical discontinuities and that 198.3 at Phylogeny.fr (Dereeper antillarum type locality: Culebra none of the morpho-anatomical et al. 2008). Island, Puerto Rico). All of these studies of the past 100+ years has Analysis of the lectotype of species were recently proposed as enabled us to recognize the Lithothamnion onkodes resulted in synonyms of P. onkodes based on numerous species passing under the complete plastid (Table S2 in morpho-anatomical characters P. onkodes. Sequenced specimens, the Supporting Information) and (Maneveldt and Keats 2014, Gal- for which we also have good habi- mitochondrial (Table S3 in the lardo et al. 2016). Not one of tat data, indicate that species evo- Supporting Information) genomes them is P. onkodes (Fig. 1). Poroli- lution has been complex with as well as the ribosomal cistron. thon sandvicense and P. oligocarpum multiple, independent, evolution- The complete plastid genome of are each distinct species, whereas ary drivers yielding the current P. onkodes is 170,468 bp in length Porolithon antillarum and P. pachy- geographic and ecological POROLITHON ONKODES SYSTEMATICS 431 L ETTER

FIG. 1. Maximum likelihood phylogram of Porolithon “onkodes.” Phylogenetic analysis of rbcL sequence data showing evolutionary relationship of P. onkodes to other sequences attributed to P. “onkodes.” Numbers above nodes are bootstrap supports based on 1,000 replicates (<75% not shown) followed by Bayesian posterior probabilities (<0.90 not shown). Asterisks indicate full support in both analyses and dashes no statistical support; A and B label clades discussed in paper; vertical bars and numbers indicate species; corresponding numbers on Figure 2 show distributions.

FIG. 2. Distribution of 21 species called Porolithon onkodes based on rbcL sequences. Numbers correspond to specimens in Figure 2 called P. onkodes; 12 is type specimen of P. onkodes. KM369154 from New Zealand was omitted as that specimen called P. onkodes belongs in another genus. Numbers with letters indicate species present at different localities. 4a = Yonge Reef; 4b = Heron Island; 4c = Cassini Island and the nearby Long Reef; 9a = Playa Munecos, Vera Cruz, Mexio; 9b = Carrie Bow Cay, Belize; 9c = Culebra Island, Puerto Rico; 9d = U.S. Virgin Islands. distributions. The evidence for Pacific Oceans, rather than all pattern. This hypothesis can be this complexity is revealed in the western Atlantic Ocean species tested by sequencing many more phylogenetic tree, from observa- forming a clade and all Pacific specimens with faster mutating tions of field-collected specimens Ocean species forming a separate markers, such as COI (Saunders and from the literature. clade. We hypothesize that the 2005), to more precisely track the Two well-supported clades ancestors of these clades already geographic patterns of these spe- (Fig. 1, A and B) contain smooth, were widely distributed in tropical cies. Rosler€ et al. (2017) proposed encrusting Porolithon species from seas and that their descendants the Indo-Australian archipelago as both the western Atlantic and reflect that historical distribution the center of origin of coral reef 432 PAUL W. GABRIELSON ET AL.

corallines. Our data support that comparable to the most exposed explain the co-occurrence and it also was the center of diversifi- lagoon specimens, were distinct in diversification of smooth, encrust- cation of these smooth, encrust- both the total amount of pigment, ing Porolithon species in reef envi- ing Porolithon species. Thus far, more than twice any of the lagoon ronments. there is no evidence for a single, specimens, and had the highest In light of global climate pantropical, smooth, encrusting concentrations of carotenoids. change and the sensitivity of P. “onkodes.” Likely the reef crest and lagoon Porolithon “onkodes” to ocean acidi- With respect to ecophysiology specimens represented two or fication and warming (Diaz-Pulido being a significant evolutionary more distinct species of Porolithon, et al. 2012, 2014, Johnson and driver, the results are mixed. although alternative hypotheses, Carpenter 2012) and based on the Some species appear to have nar- such as high physiological plastic- evidence presented herein, it is row habitat requirements, whereas ity or local adaptation, cannot be essential that future studies exam- others are broadly adapted to a ruled out. Variable hydrodynamic ining the physiology, minerology, range of PAR, hydrodynamic regimes rather than differences in and response of P. “onkodes”to ETTER regimes, and desiccation. rbcL PAR may influence species physi- ocean acidification be conducted L sequences MF979945, MF979946, ology and be important evolution- on specimens from the same habi- and MF979947, which are identi- ary drivers for some tat with respect to PAR, hydrody- cal, are all from intertidal speci- Porolithon. “onkodes” taxa. namic regime and desiccation, mens on the reef crest at different Adding to this complex evolu- and that these specimens be Caribbean Sea localities (St. Croix tionary picture is the presence of sequenced to establish their iden- Island and Belize). Likewise, morphologically indistinguishable, tity. Only then can we begin to sequences MF979931, MF979932, but distinct species (MF979953, understand evolution, biogeogra- and MF979933 of specimens from MF979938, MF979951, MF979958) phy, ecology, and physiology of the Great Barrier Reef (GBR) that all occur in the same habitat this complex of species function- occur intertidally on either the on exposed fore reefs at 5 m ing as critical ecosystem engineers fore or back reefs, and these are depth, but occupy different on tropical coral reefs. the same species. These two Car- clades throughout the tree ibbean and GBR intertidal species (Fig. 1). The strength of biotic We thank Walter Adey, Luz Elena occur in habitats with high levels interactions, including herbivory Mateo-Cid, D. Wilson Freshwater, Gavin of PAR and are subjected to desic- and interspecific competition, has Maneveldt, Tom Schils, Alexandra cation at low tide. In contrast, been hypothesized to limit Ordonez,~ and Roberta Townsend for sequences MF979937, MF979938, macroalgal species diversity in providing modern material for the MF979934, MF979935, and tropical regions (Keith et al. genetic analyses. We are most grateful MF979936 represent what may be 2014). Indeed, herbivory has to Kristian Hassel (TRH) and Barbara a single, although genetically vari- been tightly linked to coralline Thiers (NY) for loaning the type speci- – mens for examination and analysis, able species (0.8 0.9 rbcL abundance and radiation in trop- without which we would be reduced to sequence divergence) that ranges ical reef habitats where the thick, guessing the application of names to from the intertidal reef crest to high-density calcium carbonate species. Wilson Freshwater, DNA Analy- 10 m depth at exposed sites and construction, as found in Poroli- sis Core Facility, University of North from different localities either on thon species, was hypothesized to Carolina, Wilmington provided sequenc- the GBR or in Northwest Aus- allow them to thrive in environ- ing support and Todd Vision provided research space and equipment to PWG. tralia. This species appears ments of intensive herbivory (Ste- Susan Whitfield (UNC) assisted with broadly adapted to different PAR, neck 1985). Intensive herbivory the figures. The manuscript was wave exposure, and desiccation may also have contributed to the improved by the comments from two regimes. However, the two sample diversification of these smooth, anonymous reviewers. This work was sites on the GBR are 1,300 km encrusting Porolithon species funded by a family trust to P.W.G. and apart and even further from sam- through variable allelopathic supported by an Australian Research Council Grant (DP160103071) awarded ple sites on the northwest coast of interactions, as suggested for the to G.D.P. Australia. Additional collections tropical macroalgal genus Lobo- are needed to determine if this is phora (Vieira et al. 2017). Fur- PAUL W. GABRIELSON*, JEFFERY R. indeed one species. thermore, coral reef † HUGHEY and GUILLERMO In contrast, Payri et al. (2001) environments have been sug- ‡,§ examined photoacclimation in gested as centers of species diver- DIAZ-PULIDO “Porolithon onkodes” (as Hydrolithon sification as they create complex onkodes) based on specimens col- mosaics of habitats (both in size *Herbarium and Biology lected from both the reef crest and variety) and act as a refuge Department, University of North and from the lagoon, the latter from high species extinction for Carolina - Chapel Hill, Coker from three different PAR expo- a variety of reef organisms (Cow- Hall, CB 3280, Chapel Hill, North Carolina 27599-3280, USA sures. The reef crest specimens, man and Bellwood 2011). It is † despite experiencing PAR levels likely that similar processes may Division of Mathematics, Science, POROLITHON ONKODES SYSTEMATICS 433 and Engineering, Hartnell non-specialist. Nucleic Acids Res. 36: Hughey, J. R., Hommersand, M. H., – College, 411 Central Ave., Salinas W465 9. Gabrielson, P. W., Miller, K. A. & Diaz-Pulido, G., Anthony, K. R. N., Kline, Fuller, T. 2017. Analysis of the com- California, 93901, USA ‡ D. I., Dove, S. & Hoegh-Guldberg, O. plete plastomes of three species of School of Environment and 2012. Interactions between ocean Membranoptera (Ceramiales, Rhodo- Science, Australian Rivers Institute acidification and warming on the mor- phyta) from Pacific North America. J. – —Coast and Estuaries, Griffith tality and dissolution of coralline Phycol. 53:32 43. algae. J. Phycol. 48:32–9. Janouskovec, J., Liu, S. L., Martone, P. T., University, Nathan Campus, 170 Diaz-Pulido, G., Nash, M. C., Anthony, K. Carre, W., Leblanc, C., Collen, J. & Kessels Road, Brisbane, R. N., Bender, D., Opdyke, B. N., Keeling, P. J. 2013. Evolution of red Queensland, 4111, Australia Reyes-Vivia, M. & Trotzsch, U. 2014. algal plastid genomes: ancient archi- § Greenhouse conditions induce miner- tectures, introns, horizontal gene Australian Research Council alogical changes and dolomite accu- transfer, and taxonomic utility of plas- Centre of Excellence for Coral mulation in coralline algae on tid markers. PLoS ONE 8:e59001. L Reef Studies, Townsville, tropical reefs. Nat. Commun. 5:3310. Johnson, M. D. & Carpenter, R. C. 2012. ETTER Queensland 4811, Australia Doty, M. S. & Morrison, J. P. E. 1954. Ocean acidification and warming Interrelationships of the organisms on decrease calcification in the crustose Raroria aside from man. Atoll Res. coralline alga Hydrolithon onkodes and 1Received 13 April 2018. Accepted 18 May Bull. 35:1–61. increase susceptibility to grazing. J. 2018. Foslie, M. 1909. Algologiske notiser VI. Exp. Mar. Bio. Ecol. 434–435:94–101. Kong. Norske Vidensk. Selsk. Skr. 1909:1– Katoh, K. & Standley, D. M. 2013. MAFFT 63. Multiple Sequence Alignment Soft- Adey, W. H. 1967. The algal ridges and Gabrielson, P. W., Miller, K. A. & Mar- ware Version 7: improvements in per- coral reefs of St. Croix their structure tone, P. T. 2011. Morphometric and formance and usability. Mol. Biol. Evol. – and Holocene development. Atoll Res. molecular analyses confirm two spe- 30:772 80. Bull. 187:1–67. cies of Calliarthron (Corallinales, Rho- Keith, S. A., Kerswell, A. P. & Connolly, S. Adey, W. H., Hernandez-Kantun, J. J., dophyta), a genus endemic to the R. 2014. Global diversity of marine – Johnson, G. & Gabrielson, P. W. 2015. northeast Pacific. Phycologia 50:298 macroalgae: environmental conditions DNA sequencing, anatomy and calcifi- 316. explain less variation in the tropics. – cation patterns support a mono- Gallardo, T., Barbara, I., Afonso-Carrillo, Global Ecol. Biogeogr. 23:517 29. phyletic, subarctic, carbonate reef- J., Bermejo, R., Altamirano, M., Lagesen, K., Hallin, P., Rødland, E. A., forming Clathromorphum (Hapalidi- Gomez Garreta, A., Barcelo Martı, M. Staerfeldt, H. H., Rognes, T. & Ussery, aceae, Corallinales. Rhodophyta. J. Phy- C., Rull LLuch, J., Ballesteros, E. & D. W. 2007. RNAmmer: consistent col. 51:189–203. De la Rosa, J. 2016. Nueva lista crıtica and rapid annotation of ribosomal – Adey, W. H., Townsend, R. A. & Boykins, de las algas bentonicas marinas de RNA genes. Nucleic Acids Res. 35:3100 ~ W. T. 1982. The crustose coralline Espana. A new checklist of benthic 8. algae (Rhodophyta: Corallinaceae) of marine algae of Spain. Algas, Boletın Lee, R. K. S. 1978. Taxonomy and distribu- ~ the Hawaiian islands. Smithson. Con- Informativo de la Sociedad Espanola de tion of the melobesioid algae on Ron- – tribs. Mar. Sci. 15:1–74. Ficologıa 51:7 52. gelap Atoll, Marshall Islands. Can. J. – Boc, A., Diallo, A. B. & Makarenkov, V. Hernandez-Kantun, J. J., Gabrielson, P. W., Bot. 45:985 1001. 2012. T-REX: a web server for infer- Hughey, J. R., Pezzolesi, L., Rindi, F., Lee, J., Kim, K. M., Yang, E. C., Miller, K. ~ ring, validating and visualizing phylo- Robinson, N. M., Pena, V., Riosmena- A., Boo, S. M., Bhattacharya, D. & genetic trees and networks. Nucleic Rodriguez, R. Le., Gall, L. & Adey, W. Yoon, H. S. 2016. Reconstructing the Acids Res. 40:W573–9. H. 2016. Reassessment of branched complex evolutionary history of Boo, G. H., Hughey, J. R., Miller, K. A. & Lithophyllum spp. (Corallinales, Rhodo- mobile plasmids in red algal genomes. Boo, S. M. 2016. Mitogenomes from phyta) in the Caribbean Sea with glo- Sci. Rep. 6:23744. – type specimens, a genotyping tool for bal implications. Phycologia 55:609 35. Littler, M. M. & Doty, M. S. 1975. Ecologi- morphologically simple species: ten Heydrich, F. 1897. Neue Kalkalgen von cal components of structuring the sea- genomes of agar-producing red algae. Deutsch-Neu-Guinea (Kaiser Wil- ward edges of tropical Pacific reefs. J. – – Sci. Rep. 6:35337. helms-Land). Bibliotheca Bot. 7:1 11. Ecol. 63:117 29. Cowman, P. F. & Bellwood, D. R. 2011. Hind, K. R., Gabrielson, P. W., Jensen, C. P. Maneveldt, G. W. & Keats, D. W. 2014. Coral reefs as drivers of cladogenesis: & Martone, P. T. 2016. Crusticorallina Taxonomic review based on new data expanding coral reefs, cryptic extinc- gen. nov., a non-geniculate genus in the of the reef-building alga Porolithon tion events, and the development of subfamily Corallinoideae (Corallinales, onkodes (Corallinaceae, Corallinales, – biodiversity hotspots. J. Evol. Biol. Rhodophyta). J. Phycol. 52:929 41. Rhodophyta) along with other taxa 24:2543–62. Howe, M. A. 1912. The building of “coral” found to be conspecific. Phytotaxa – – Darling, A. E., Mau, B. & Perna, N. T. 2010. reefs. Science 35:837 42. 190:216 49. ProgressiveMauve: multiple genomic Hughey, J. R. & Gabrielson, P. W. 2012. Marsh, J. A. 1970. Primary productivity of alignment with gene gain, loss, and Acquiring DNA sequence data from the reef-building calcareous red algae. – rearrangement. PLoS ONE 5:e11147. dried archival red algae (Florideo- Ecology 51:255 63. Darwin, C. 1890. On the Structure and Distribu- phyceae) for the purpose of applying Payri, C. E., Maritorena, S., Bizeau, C. & tion of Coral Reefs. In Bettany, G. T. [Ed.]. available names to contemporary Rodiere, M. 2001. Photoacclimation in Ward, Lock and Co., London, 208 pp. genetic species: a critical assessment. the tropical coralline alga Hydrolithon – Dean, A. J., Steneck, R. S., Tager, D. & Botany 90:1191 4. onkodes (Rhodophyta, Corallinaceae Pandolfi, J. M. 2015. Distribution, Hughey, J. R., Gabrielson, P. W., Rohmer, from a French Polynesian reef. J. Phy- – abundance and diversity of crustose L., Tortolani, J., Silva, M., Miller, K. col. 37:223 34. coralline algae on the Great Barrier A., Young, J. D., Martell, C. & Ruedi- Peng, Y., Leung, H. C., Yiu, S. M. & Chin, Reef. Coral Reefs 34:581–94. ger, E. 2014. Minimally destructive F. Y. 2012. IDBA-UD: a de novo assem- Dereeper, A., Guignon, V., Blanc, G., sampling of type specimens of Pyropia bler for single-cell and metagenomic Audic, S., Buffet, S., Chevent, F., Dufa- (Bangiales, Rhodophyta) recovers sequencing data with highly uneven – yard, J. F. et al. 2008. Phylogeny.fr: complete plastid and mitochondrial depth. Bioinformatics 28:1420 8. robust phylogenetic analysis for the genomes. Sci. Rep. 4:5113. Ronquist, F. & Huelsenbeck, J. P. 2003. MRBAYES 3: Bayesian phylogenetic 434 PAUL W. GABRIELSON ET AL.

inference under mixed models. Bioin- Zerbino, D. R. & Birney, E. 2008. Velvet: formatics 19:1572–4. algorithms for de novo short read figure was drawn using Mauve Rosler,€ A., Perfectti, F., Pena,~ V., Aguirre, assembly using de brujin graphs. Gen- 2.3.1. J. & Braga, J. C. 2017. Timing of the ome Res. 18:821–9. evolutionary history of Corallinaceae Figure S3. Maximum likeli- (Corallinales, Rhodophyta). J. Phycol. hood phylogram of Porolithon – 53:567 76. onkodes based on multigene Saunders, G. W. 2005. Applying DNA bar- Supporting Information coding to red macroalgae: a prelimi- sequence data, showing evolu- nary appraisal holds promise for Additional Supporting Infor- tionary relationship of authen- future applications. Phil. Trans. R. Soc. mation may be found in the tic P. onkodes to other B 360:1879–88. sequences attributed to Schattner, P., Brooks, A. N. & Lowe, T. M. online version of this article at 2005. The tRNAscan-SE, snoscan and the publisher’s web site: P. onkodes. Numbers along snoGPS web servers for the detection branches are bootstrap supports of tRNAs and snoRNAs. Nucleic Acids Figure S1. Lectotype speci- based on 1,000 replications Res. 33:686–9. men of Porolithon onkodes (TRH (<75% support not shown). ETTER Setchell, W. A. 1926. Nullipore versus coral A26-1494). Encrusting habit on L in reef formation. Proc. Am. Phil. Soc. coral tine and box lid with col- Table S1. List of taxa with – 65:136 40. lection data. collection data and GenBank Steneck, R. S. 1985. Adaptations of crus- numbers for sequences used in tose coralline algae to herbivory: pat- Figure S2. Locally collinear this paper arranged from east terns in space and time. In Toomey, blocks (LCBs) analysis of red D. F. & Nitecki, M. H. [Eds.] Paleoalgo- to west under each name. logy. Springer-Verlag, Berlin, pp. 352– algal mitogenomes. The figure Sequences included in concate- 66. depicts linearized alignments nated analyses shown in italics. Taylor, W. R. 1950. Plants of Bikini and identifying conserved gene Other Northern Marshall Islands. Univer- regions for Florideophyceae. Table S2. Plastid genome sity of Michigan Press, Ann Arbor, gene content of the lectotype Michigan, 860 pp. Each mitogenome is oriented Vieira, C., Camacho, O., Sun, Z., Freder- horizontally and homologous of Porolithon onkodes. icq, S., Leliaert, F., Payri, C. & De blocks are shown as identically Table S3. Mitogenome gene Clerck, O. 2017. Historical biogeogra- colored regions linked across content of the lectotype of Poro- phy of the highly diverse brown sea- genomes. Regions inverted rela- weed Lobophora (Dictyotales, lithon onkodes. Phaeophyceae). Mol. Phylogenet. Evol. tive to Porolithon onkodes are 110:81–92. shifted below the genome’s Table S4. Pair-wise sequence Womersley, H. M. S. & Bailey, A. 1970. center axis. Sequence similari- divergences (%) for rbcL Marine algae of the Solomon Islands. ties within an LCB are propor- sequences reported in this Phil. Trans. R. Soc. B 259:257–352. study. Yang, E. C., Kim, K. M., Kim, S. Y., Lee, J., tional to the heights of interior- Boo, G. H., Lee, J. H., Nelson, W. A. colored bars. Large sections of et al. 2015. Highly conserved mito- white within blocks and gaps chondrial genomes among multicellu- between blocks indicate lin- lar red algae of the Florideophyceae. eage-specific sequences. The Genome Biol. Evol. 7:2394–406.