J. Phycol. 42, 669–678 (2006) r 2006 by the Phycological Society of America DOI: 10.1111/j.1529-8817.2006.00230.x PHYLOGENETIC ANALYSES OF THE BRYOPSIDALES (ULVOPHYCEAE, CHLOROPHYTA) BASED ON RUBISCO LARGE SUBUNIT GENE SEQUENCES1 Daryl W. Lam and Frederick W. Zechman2 Department of Biology, California State University, Fresno 2555 East San Ramon Avenue, Fresno, California 93740, USA Current taxonomy of the Bryopsidales recogniz- morphological diversity (from the simple uniaxial si- es eight families; most of which are further cate- phonous construction found in Bryopsis, Derbesia,and gorized into two suborders, the Bryopsidineae and Caulerpa to the complex interwoven multiaxial siphon Halimedineae. This concept was supported by early patterns found in Codium, Halimeda, and Penicillus). molecular phylogenetic analyses based on rRNA Some genera are heavily calcified as in the genus Udo- sequence data, but subsequent cladistic analyses of tea, while the family Caulerpaceae and the majority of morphological characters inferred monophyly in the suborder Bryopsidineae exhibit no calcification. only the Halimedineae. These conflicting results Some bryopsidalean taxa are invasive and ecologically prompted the current analysis of 32 taxa from this problematic and are known to flourish in temperate diverse group of green algae based on plastid-en- marine waters (e.g. Caulerpa taxifolia and Codium fragile coded RUBISCO large subunit (rbcL) gene sequenc- subsp. tomentosoides; Bouk and Morgan 1957, Trow- es. Results of these analyses suggested that the bridge 1995, Jousson et al. 1998). Halimedineae and Bryopsidineae are distinct mon- Smith (1955) separated the Bryopsidales (his Si- ophyletic lineages. The families Bryopsidaceae, phonales) into the families Halicystidaceae (including Caulerpaceae, Codiaceae, Derbesiaceae, and the algae currently classified under the genus Derbesia), Halimediaceae were inferred as monophyletic, Bryopsidaceae (Bryopsis), Caulerpaceae (Caulerpa), however the Udoteaceae was inferred as non-mon- Codiaceae (Codium) and Dichotomosiphonaceae (Di- ophyletic. The phylogenetic position of two taxa chotomosiphon and Boodleopsis). Hillis-Colinvaux (1984) with uncertain subordinal affinity, Dichotomosiphon subdivided the order into two suborders on the basis of tuberosus Lawson and Pseudocodium floridanum thallus morphology, reproduction, plastid types and Dawes & Mathieson, were also inferred. Pseudocodi- geographic distributions. She defined the Bryopsidi- um was consistently placed within the halimedinean neae (including Bryopsis, Codium, and Derbesia) based clade suggesting its inclusion into this suborder, on non-holocarpic reproduction and the Hali- however familial affinity was not resolved. D. tu- medineae (e.g. Caulerpa, Halimeda, and Udotea)onho- berosus was the inferred sister taxon of the Ha- locarpic reproduction. Furthermore, Hillis-Colinvaux limedineae based on analyses of rbcL sequence (1984) noted that the Halimedineae exhibit heteropla- data and thus a possible member of this suborder. sty (containing both chloroplasts and amyloplasts) Key index words: Bryopsidales; Caulerpales; while amyloplasts are absent in bryopsidinean taxa. Chlorophyta; Phlyogeny; rbcL; RUBISCO; taxo- Global distribution patterns vary as well. That is, nomy; Ulvophyceae bryopsidinean taxa generally inhabit temperate, trop- ical, and subtropical marine waters, while ha- Abbreviations: BI, Bayesian inference; ML, maxi- limedinean taxa are generally restricted to tropical mum likelihood; MP, maximum parsimony; MPT, and subtropical habitats. However, exceptions to these most parsimonious tree; TLD, tree length distri- general patterns occur, including the occurrence of bution. some Caulerpa species in temperate waters (e.g. Caul- erpa taxifolia, Jousson et al. 1998, 2000) and non-holo- carpy in Caulerpella (Prud’homme van Reine and The order Bryopsidales (also referred to as the Lokhorst 1992). Caulerpales, Codiales, and Siphonales) is comprised Pseudocodium and Dichotomosiphon are two bryopsida- of green, mostly macroscopic, siphonous algae with lean genera with uncertain affinity at the subordinal multicellularity arising only in some taxa during sexual level. Hillis-Colinvaux (1984) tentatively placed the ge- reproduction (Silva 1982). The Bryopsidales exhibit a nus Pseudocodium into the Bryopsidineae because of a cosmopolitan distribution; however, some groups are reported common mannan cell wall component with restricted to tropical marine environments. One ge- the genus Codium (Dawes and Mathieson 1972). Like- nus, Dichotomosiphon, is found in freshwater habitats. wise, cladistic analyses of morphological traits grouped Similarly, the Bryopsidales exhibit extremely broad Pseudocodium with Codium (Vroom et al. 1998). How- ever, morphological features, including heteroplasty 1Received 5 October 2005. Accepted 8 March 2006. (Feldman 1946), suggest that the alga has more in 2Author for correspondence: e-mail: [email protected]. common with the Halimedineae. In addition, Weber 669 670 DARYL W. LAM AND FREDERICK W. ZECHMAN van Bosse (1896), Levring (1938), and Womersley (pH 8.0), 20 mM Tris, 2.5 mM polyvinylpyrrolidone, and (1955) noted that the formation of utricles in Pseudo- 2 mL/ Á mL 2-mercaptoethanol) DNA extractions were also codium is closer in affinity to Halimeda than to Codium. performed using the Dneasy Plant Mini Kit (Qiagen Inc. Va- The freshwater alga, Dichotomosiphon tuberosus, possess- lencia, CA, USA) following the protocols found therein. The rbcL gene was amplified in two overlapping fragments es plastid features of the Halimedineae (i.e. heteroplas- from total genomic DNA with oligonucleotide primers ty, Moestrup and Hoffman 1973), however, sexual (Table 2). Each 50 mL PCR reaction consisted of 10 mL diluted reproduction is non-holocarpic (Ernst 1902, Smith DNA, 5 mL10Â Buffer with 1.25 U of Taq polymerase (Fisher 1955). Previous results (Vroom et al. 1998, Hanyuda Scientific Okasis), 2.0 mM MgCl2, 0.2 mM each deoxynucleo- et al. 2000) inferred a sister relationship between Di- tide triphosphate, 0.2 mM each primer, and 0.024% non-acety- chotomosiphon and a clade containing halimedinean lated BSA (Sigma Chemicals, St. Louis, MO, USA). Thermocycling was accomplished on a GeneAmp 2700 PCR taxa. system (Applied Biosystems, Foster City, CA, USA). PCR pa- Previous morphology-based phylogenetic analyses rameters included an initial denaturation at 941 Cfor5min of the order (Vroom et al. 1998) suggested monophyly followed by 40 repeated cycles of 941 C for 45 s, 41.51 C for 45 s, of the Halimedineae. However, bryopsidinean taxa and 721 C for 90 s. A final incubation at 721 C for 10 min was formed a non-monophyletic assemblage. In contrast, used to insure complete polymerization of DNA strands. Re- molecular phylogenies (Zechman et al. 1990, Hanyuda sulting PCR products were electrophoresed on 1% Tris Borate EDTA agarose gels. Products were purified for sequencing us- et al. 2000) supported the monophyly of both subor- ing Qiaquick Spin Columns (Qiagen Inc.) based on protocols ders; however these data sets contained limited num- contained therein. The purified PCR product was quantified bers of bryopsidalean taxa. Recent molecular for cycle sequencing reactions by comparison of band intensities phylogenetic analyses of bryopsidalean taxa (Hillis against known concentrations of unmethylated l virus genomic et al. 1998, Woolcott et al. 2000, Fama et al. 2002, Ko- DNA (Fisher Scientific) on 1% TBE agarose gels. oistra 2002, Kooistra et al. 2002) were aimed at resolv- DNA sequencing. Approximately 40–50 ng of purified PCR ing more specific family, genus and species level product were used as a template for cycle sequencing reac- tions. Sequencing reactions were performed with Big Dye relationships. The current study seeks to elucidate Terminator version 3.1 (Applied Biosystems), purified with the phylogenetic position of major evolutionary line- G-50 fine Sephadex (Amersham Biosciences, Uppsala, Swe- ages within the Bryopsidales. This was accomplished den) columns, dried and dissolved in template suppression by phylogenetic analyses of the RUBISCO large sub- reagent (Applied Biosystems). Automated sequencing was unit (rbcL) sequences for 32 bryopsidalean taxa. Se- performed on an ABI Prism 310 Genetic Analyzer (Applied quencing and amplification of additional taxa was Biosystems). Two opposing strands were sequenced for each fragment using the aforementioned PCR primers. The RUB- attempted but not successful with the methods de- ISCO large subunit sequence fragments were edited and scribed. To the extent possible with current taxon and assembled into contigs using Sequencher version 3.1.1 character sampling, phylogenetic analyses sought to: (Gencodes Corp., Ann Arbor, MI, USA). (1) infer phylogenetic relationships among the Bryop- Sequence alignment and molecular phylogenetic analyses. Se- sidales at the subordinal and familial levels; (2) com- quences were aligned with ClustalX (Thompson et al. 1997). pare rbcL sequence-based phylogenies to results ob- Introns were removed from the data set and the data set was tained from previous data sets; and (3) determine the imported into MacClade version 4.03 (Maddison and Maddi- son 2001). Unweighted MP analyses performed with PAUP* phylogenetic affinity of the genera Pseudocodium and 4.0b10 (Swofford 2002) utilized the heuristic search option Dichotomosiphon. with 10 replicates of random taxon addition and tree-bisec- tion-reconnection (TBR) branch swapping. Characters state changes were