J. Phycol. 43, 366–377 (2007) r 2007 by the Phycological Society of America DOI: 10.1111/j.1529-8817.2007.00319.x CHARACTER EVOLUTION IN POLYKRIKOID DINOFLAGELLATES1 Mona Hoppenrath2 and Brian S. Leander Departments of Botany and Zoology, University of British Columbia, 6270 University Boulevard, Vancouver, BC, Canada V6T 1Z4 A synthesis of available data on the morphological Ti:Tv, transition/transversion ratio; WNJ, weighted diversity of polykrikoid dinoflagellates allowed neighbor joining us to formulate a hypothesis of relationships that help explain character evolution within the group. Phylogenetic analyses of new SSU rDNA sequences from Pheopolykrikos beauchampii Chatton, Polykrikos kofoidii Chatton, and Polykrikos lebourae Herdman The dinoflagellate genus Polykrikos was erected by helped refine this hypothetical framework. Our re- Bu¨tschli (1873), with the type species P. schwartzii sults demonstrated that ‘‘pseudocolonies’’ in dino- Bu¨tschli. The most distinctive feature of this athecate flagellates evolved convergently at least three times genus is the formation of multinucleated pseudocolo- independently from different Gymnodinium-like nies comprised of an even number of zooids that are ancestors: once in haplozoans; once in Ph. beau- otherwise similar in morphology to individual dino- champii; and at least once within a lineage contain- flagellates in external view. However, despite every ing Ph. hartmannii, P.kofoidii,andP. lebourae.The zooid having its own cingulum and pair of flagella, the Gymnodiniales sensu stricto was strongly supported zooid sulci are fused together. A pseudocolony often by the data, and the type species for the genus, name- has half the number of nuclei because it has zooids. ly Gymnodinium fuscum (Ehrenb.)F.Stein,formedthe Trichocysts, nematocysts, taeniocysts, mucocysts, and nearest sister lineage to a well-supported Polykrikos plastids have all been reported from different mem- clade. The best synapomorphy for the Polykrikos bers within the group. The genus currently comprises clade was the presence of two nuclei irrespective of four species: P. schwartzii, P. kofoidii, P. lebourae, and zooid number. Two unidentified Gymnodinium spe- P. grassei Lecal. The first three species are relatively cies formed the nearest sister clade to Ph. beau- well described and distinguishable from one another champii, which has four nuclei and four zooids per (Chatton 1914, Kofoid and Swezy 1921, Herdman pseudocolony. The chain-forming dinoflagellate G. 1923, Kofoid 1931, Balech 1956, Dragesco 1965, Hop- catenatum L. W. Graham branched closely to the penrath 2000, Matsuoka et al. 2000, Nagai et al. 2002); clade containing all members of Polykrikos and Pheo- however, the taxonomic separation between P.schwartz- polykrikos, suggesting that an ancestral capacity to- ii and P. kofoidii is difficult because characters like size, ward chain formation existed before the evolution of zooid number, and cingular displacement are overlap- pseudocolonies in this group. Our results also clari- ping in these two species (Kofoid and Swezy 1921, fied the phylogenetic significance of nematocysts, Nagai et al. 2002, Throndsen et al. 2003). The most ocelloids, and photosynthesis in reconstructing the reliable distinguishing character between them is the evolution of polykrikoids and warnowiids. The mo- presence of striated ribs on the posterior-most zooid or lecular phylogenies exposed taxonomic problems ‘‘hyposome’’ of the pseudocolony in P. kofoidii (Nagai associated with Polykrikos, Pheopolykrikos,andGym- et al. 2002). To the best of our knowledge, P.grassei has nodinium, and suggested that a revision for some of only been observed and recorded once (Lecal 1972), these genera is warranted. and doubts about the identity of P. grassei have been expressed in the literature (Greuet and Hovasse 1977, Key index words: character evolution; dinoflagel- Sournia 1986). The reexamination of this species is late; Dinophyceae; Gymnodinium; Pheopolykrikos; critical from a phylogenetic point of view because it is phylogeny; Polykrikos; small subunit ribosomal described to possess an ocelloid (Lecal 1972). These RNA complex organelles otherwise exist only in warnowiid Abbreviations: AU, approximately unbiased; GTR dinoflagellates (Greuet 1987). model, general-time-reversible model; HKY model, The second genus of polykrikoid dinoflagellates is Hasegawa–Kishino–Yano model; MCMC, Monte- Pheopolykrikos, which was first described by Chatton Carlo–Markov chains; ML, maximum likelihood; (1933), with the type species Ph. beauchampii, and sub- sequently emended by Matsuoka and Fukuyo (1986). Pheopolykrikos is different from Polykrikos in having the same number of nuclei as zooids and being able to disassociate into single cells (Chatton 1933, 1952). 1Received 28 January 2006. Accepted 1 November 2006. Pheopolykrikos beauchampii is photosynthetic and ap- 2Author for correspondence: e-mail [email protected]. pears to lack the ability to phagocytize (Chatton 366 CHARACTER EVOLUTION IN POLYKRIKOID DINOFLAGELLATES 367 1933). When emending the genus Pheopolykrikos, hypothesis of relationships that helps explain character Matsuoka and Fukuyo (1986) transferred P.hartmannii evolution within the group. M. H. Zimm. (see also Hulburt 1957) into Pheopolykri- kos for several reasons: the number of nuclei and zo- MATERIALS AND METHODS oids is the same, there is a single-cell life-cycle stage, and the cells are photosynthetic. Furthermore, these Organisms and light microscopy. Near-surface plankton sam- ples were collected in the morning hours with a small net authors emphasized the different cyst morphologies (mesh size 20 mm) at the docks of the Bamfield Marine Sci- present in the two genera and the possibility of differ- ences Center, Vancouver Island (BC, Canada), in June and ent excystment conditions. It is noteworthy that no cyst August of 2005. Immediately after sampling, single cells of stage is known for the type species of Pheopolykrikos, P. kofoidii and Ph. beauchampii were identified at magnifica- namely Ph. beauchampii, and therefore, the value of the tions of Â40 to Â250 (Fig. 1) and isolated from the mixed cyst morphology to separate the two genera remains to plankton sample by micropipetting. Sand samples containing P. l e b o u r a e were collected with a spoon during low tide at be demonstrated. Centennial Beach, Boundary Bay (BC, Canada), in October Polykrikos barnegatensis G. W. Martin is described as a 2005. The sand samples were transported directly to the lab- two-zooid pseudocolony with one large nucleus, with oratory, and the flagellates were separated from the sand by plastids, and without nematocysts (Martin 1929). It has extraction through a fine filter (mesh size 45 mm) using melt- subsequently been synonymized with P. hartmannii ing seawater ice (Uhlig 1964). The flagellates accumulated in (Chatton 1952), but as Hulburt (1957) pointed out, a petri dish beneath the filter and were then identified at magnifications of Â40 to Â250 (Fig. 1). Cells were isolated the ‘‘species’’ differ in the number of nuclei (one in the by micropipetting for the preparations described below. former, and two in the latter). Because Martin (1929) Cells were observed directly and micromanipulated with a based the description of the new species on the obser- Leica DMIL inverted microscope (Wetzlar, Germany) connected vation of only one living cell, P.barnegatensis is in need to a PixeLink Megapixel color digital camera (PL-A662-KIT, of reinvestigation. This uncertainty led us to disregard Ottawa, ON, Canada). For DIC LM, micropipetted cells were this species in the following discussion. placed on a glass specimen slide and covered with a coverslip. There are also different views about the generic Images were produced directly with either the PixeLink Meg- apixel color digital camera or a Zeiss Axioplan 2 imaging micro- classification of polykrikoid dinoflagellates. Loeblich scope (Carl–Zeiss, Oberkochen, Germany) connected to a Leica (1980) transferred Ph. beauchampii into the genus DC500 color digital camera (Wetzlar, Germany). Polykrikos; Dodge (1982) and Sournia (1986) recog- Molecular phylogenetic analysis. Individually isolated pseu- nized only the genus Polykrikos and treated Pheopolykri- docolonies (five for P. k o f o i d i i ,nineforPh. beauchampii,and kos as a junior synonym. As mentioned previously, four for P. lebourae) were individually washed four times in Matsuoka and Fukuyo (1986) retained Pheopolykrikos filtered (eukaryote free) seawater. All pseudocolonies of one species were deposited into a 1.5 mL Eppendorf tube (Dia- as a separate genus from Polykrikos. Fensome et al. Med Lab Supplies Inc., Mississauga, ON, Canada), resulting (1993) were of the opinion that Pheopolykrikos and Poly- in one multispecimen sample for each of the three species. krikos were distinctive enough to be classified into dif- Genomic DNA was extracted using a standard hexadecyltri- ferent families—Pheopolykrikos in the Gymnodiniaceae, methylammonium bromide (CTAB) extraction protocol (Zo- and Polykrikos in the Polykrikaceae. Steidinger and lan and Pukkila 1986) or by placing washed pseudocolonies Tangen (1997) distinguished the two genera but clas- in distilled water that was directly used for PCR. The PCR was carried out using puReTaq Ready-To-Go PCR Beads sified them both into the family Polykrikaceae. Table 1 (Amersham Biosciences, Piscataway, NJ, USA). The PCR am- summarizes the diversity of morphological character- plification protocol using universal eukaryotic primers con-