Origin and Evolution of the Canal Raphe System in Diatoms

Protist, Vol. 162, 723–737, November 2011 http://www.elsevier.de/protis Published online date 25 March 2011

ORIGINAL PAPER

Origin and Evolution of the Canal Raphe System in Diatoms Elizabeth C. Rucka,1 , and Edward C. Theriota,b aSection

of Integrative Biology, The University of Texas at Austin, 1 University Station (A6700), Austin, Texas 78712 USA bTexas Natural Science Center, 2400 Trinity Street, The University of Texas at Austin, Austin, Texas 78619 USA Submitted October 13, 2010; Accepted February 7, 2011 Monitoring Editor: Marina Montresor

One lineage of pennate diatoms has a slit through the siliceous cell wall, called a “raphe,” that functions in motility. Raphid pennate diatoms number in the perhaps tens of thousands of species, with the diversity of raphe forms potentially matching this number. Three lineages—the Bacillariales, Rhopalodiales, and Surirellales—possess a complex and presumably highly derived raphe that is physically separated from the cell interior, most often by a set of siliceous braces. Because the relationship among these three lineages is unclear, the number of origins of the canal raphe system and the homology of it and its constitutive parts among these lineages, is equally unclear. We reconstructed the phylogeny of raphid pennate diatoms and included, for the first time, members of all three canal raphid diatom lineages, and used the phylogeny to test specific hypotheses about the origin of the canal raphe. The canal raphe appears to have evolved twice, once in the common ancestor of Bacillariales and once in the common ancestor of Rhopalodiales and Surirellales, which form a monophyletic group in our analyses. These results recommend careful follow-up morphogenesis studies of the canal raphe in these two lineages to determine the underlying developmental basis for this remarkable case of parallel evolution. © 2011 Elsevier GmbH. All rights reserved. Key words: canal raphe; diatom; fibulae; parallel evolution

Introduction One lineage of pennate diatoms has a slit through the siliceous cell wall, called a “raphe,” that functions in motility. Various characters associated with the raphe system have long been used to differentiate taxonomic groups of raphid pennate diatoms. Several genera possess a so-called “canal raphe” system, which among the different raphe types is considered highly derived (Krammer 1989; Mann 2000; Round et al. 1990; Sims and Paddock 1982). In a canal raphe system, the inner side of the 1 Corresponding author: Fax: +1 512 232 3402. e-mail [email protected] (E.C. Ruck).

© 2011 Elsevier GmbH. All rights reserved. doi:10.1016/j.protis.2011.02.003

raphe is separated from the cell interior (Ross et al. 1979), most often by a set of siliceous braces called “fibulae” (arrowheads, Figs 1C, D, I, 2C, F). These fibulae bridge the raphe internally creating a canal beneath them (double arrow, Fig. 1D). Contact between the interior of the raphe canal and the interior of the cell is through a series of portulae (arrows, Figs 1C, I, 2C, F). Finally, canal raphe systems are most often found in marginal valve positions and can be raised onto wings or keels (Figs 1A, B, 2A, D). A total of 27 diatom genera, spanning three diverse taxonomic orders (Bacillariales, Rhopalodiales, and Surirellales), have canal raphe systems (Table 1; Figs 1, 2; Round et al. 1990). Their shared

724 E.C. Ruck and E.C. Theriot

Figure 1. Scanning electron micrographs of exemplar taxa from the Bacillariales and Rhopalodiales. A—D. Hantzschia amphioxys, strain A4. A. External whole valve view. B. Internal whole valve view. C. Close up of internal valve, fibula (arrowhead) and portula (arrow). D. Close up of broken valve, internal view, fibula (arrowhead) and canal raphe (double arrow). E—F. Epithemia turgida, strain Brackenridge. E. External whole valve view. F. Internal whole valve view. G—I. Rhopalodia contorta, strain L1299. G. External whole valve view. H. Internal whole valve view. I. Close up of internal valve, fibula (arrowhead) and portula (arrow), Scale bars A, B,E,F = 10 ␮m, C, G,H = 2 ␮m, D, I = 1 ␮m.

Origin and Evolution of the Canal Raphe System in Diatoms 725

Figure 2. Scanning electron micrographs of exemplar taxa from the Surirellales. A—C. Entomoneis paludosa, strain L65. A. External whole valve view. B. Internal whole valve view. C. Close up of internal valve, fibula (arrowhead) and portula (arrow). D—F. Surirella ovalis, strain M1. D. External whole valve view. E. Internal whole valve view. F. Close up of internal valve, fibula (arrowhead) and portula (arrow). Scale bars A, B = 5 ␮m, D, E = 10 ␮m, C, F = 2 ␮m.

726 E.C. Ruck and E.C. Theriot Table 1. Genera included in the orders Bacillariales, Rhopalodiales and Surirellales within the class Bacillariophyceae, subclass Bacillariophycidae, according to the classification of Round et al. 1990. BACILLARIALES Hendey emend. Mann 1990 Bacillariaceae Ehrenberg 1831 Bacillaria Gmelin 1791 Cylindrotheca Rabenhorst 1859 Cymatonitzschia Simonsen 1974 Cymbellonitzschia Hustedt 1924 Denticula Kützing 1844 Denticulopsis Simonsen 1979 Fragilariopsis Hustedt 1913 Gomphonitzschia Grunow 1868

Gomphotheca Hendey 1982 Hantzschia Grunow 1877 Nitzschia Hassall 1845 Perrya Kitton 1874 Psammodictyon Mann 1990 Simonsenia Lange-Bertalot 1979 Tryblionella Smith 1853

RHOPALODIALES Mann 1990 Rhopalodiaceae (Karsten) Topachevs’kyj & Oksiyuk 1960 Epithemia Kützing 1844 Protokeelia Reimer & Lee 1984 Rhopalodia Müller 1895 SURIRELLALES Mann 1990 Auriculaceae Hendey, 1964 Auricula Castracane 1873 Surirellaceae Kützing 1844 Campylodiscus Ehrenberg ex Kützing 1844 Cymatopleura Smith 1851 Hydrosilicon Brun 1891 Petrodictyon Mann 1990

possession of a canal raphe system has long been considered strong support for a close phylogenetic relationship of these three orders (Hustedt 1928, 1929; Peragallo and Peragallo 1897-1908; Schrader 1973; Simonsen 1979). Rather than a single, uniform character that can be considered homologous in its entirety, the canal raphe might instead represent a composite of many, more-orless independently evolving characters (Paddock and Sims 1977). A detailed and phylogenetically minded study of the constitutive parts of the canal raphe across raphe-bearing taxa is necessary to determine whether the canal raphe is grossly homologous across canal raphe-bearing diatoms, or whether “canal raphe” is a generic term describing a suite of independently evolving characters on the diatom cell wall. In a phylogenetic context, monophyly of all canal raphid diatoms would support the former hypothesis, whereas nonmonophyly of the canal raphids would support the latter. Molecular phylogenetic analyses have, in fact, shown evidence for at least two distinct lineages of canal-raphe-bearing diatoms, the Bacillariales and Surirellales (Medlin et al. 2000; Medlin and Kaczmarska 2004; Sims et al. 2006; Sorhannus

Entomoneidaceae Reimer in Patrick & Reimer 1975 Entomoneis Ehrenberg 1845 Plagiodiscus Grunow & Eulenstein 1867 Stenopterobia Brébisson ex Van Heurck 1896 Surirella Turpin 1828

2004; Sorhannus 2004, 2007). These analyses were based exclusively on small subunit rDNA (SSU or 18S) sequences, with just three genera from the Surirellales and no Rhopalodiales sampled, thus leaving questions about the evolution of the canal raphe outstanding. Beyond the incomplete biological picture, there is a large and increasingly decisive body of literature on the critical importance of both character and taxonomic sampling density on phylogenetic accuracy (Heath et al. 2008; Hillis 1998; Hillis et al. 2003; Zwickl and Hillis 2002), including some examples from the diatoms (Alverson and Theriot 2005). As a result, the relationships among the presently skewed taxonomic sample of canal raphid diatoms are far from certain. The goal of this study is to resolve the phylogenetic relationships among canal raphid diatoms in the context of raphid diatom evolution. In addition to greatly increasing sampling within the Bacillariales and Surirellales, we include, for the first time, members of the third canal-raphe-bearing lineage, the Rhopalodiales. We have also increased character sampling over previous studies by sequencing several genes from different genomic compartments.

Origin and Evolution of the Canal Raphe System in Diatoms 727

These analyses shed valuable light on the relationships among canal-raphe-bearing diatoms and therefore the nature and origins of the canal raphe system. We also resolve the phylogenetic positions of several enigmatic or previously unstudied raphid diatoms.

Results The ML trees resulting from the nuclear SSU rDNA and combined chloroplast datasets are congruent with respect to the phylogenetic positions of the three canal- raphe-bearing orders (Fig. 3A, B). Both the nuclear and CPL datasets resolved the Bacillariales as monophyletic with weak support in the SSU rDNA dataset and strong support in the CPL dataset (BPP/BS 100/97) (Fig. 3A, B). Both datasets also resolve Surirellales and Rhopalodiales as monophyletic with strong statistical support (91–100) under both optimality criteria (SSU 100/91; CPL 99/100). The Rhopalodiales are nested within the Surirellales in both the SSU and CPL trees. The concatenated three-gene dataset gave results consistent with those of the SSU and CPL datasets. That is, Bacillariales are a strongly supported monophyletic group and sister to all other raphid pennates (Fig. 4). At least five strongly supported splits separate the two canal raphid lineages, with the intervening lineages comprised mainly of various “naviculoid” diatoms. The Surirellales and Rhopalodiales again form a strongly supported clade (Fig. 4). Rhopalodiales being monophyletic with high BPP and BS support and nested within the Surirellales (Fig. 4). Hypothesis testing of Surirellales+Rhopalodiales proved unnecessary as the optimal tree from each of the three datasets grouped the two orders into a single lineage. The SSU rDNA dataset alone rejected monophyly of Bacillariales+Rhopalodiales (H2) and Bacillariales+Surirellales (H3) but could not reject the monophyly of a single lineage of the canal raphid diatoms (H1) (p = 0.071). However, all other tests rejected all other null hypotheses.

Discussion Origin and Evolution of the Canal Raphe System Shared possession of canal raphe systems with fibulae has led to speculation of a close phylogenetic relationship between Bacillariales, Rhopalo-

diales, and Surirellales (Hustedt 1928, 1929). Previous SSU rDNA phylogenies that included limited numbers of taxa from the Bacillariales and Surirellales showed that these two orders were distantly related, thereby suggesting that their canal raphe systems were not homologous (Bruder and Medlin 2008b; Medlin and Kaczmarska 2004; Medlin et al. 2000; Sorhannus 2004, 2007). However, none of these studies included representatives of the third canal-raphe-bearing lineage, the Rhopalodiales. It was therefore unclear whether addition of the Rhopalodiales might, (1) provide the necessary phylogenetic signal to “pull” the Bacillariales and Surirellales together into a single clade of canal raphid diatoms, (2) identify the Rhopalodiales as a third, distinct lineage of canal raphid diatoms, or (3) place Rhopalodiales with one of the other two distinct canal raphid lineages. Thus, several important questions about the origin of the canal raphe have remained unanswered. We included, for the first time, members of the third canal-raphebearing lineage, the Rhopalodiales and found strong support for monophyly of the Surirellales and Rhopalodiales and thus a single origin of the canal raphe in this lineage. In addition, each of our three datasets placed this lineage at least five nodes away from the Bacillariales, so it came as no surprise that all hypothesis testing methods rejected monophyly of Bacillariales+Surirellales (Table 5). That is, the canal raphe appears to have independently evolved once in the Bacillariales and again in the Rhopalodiales+Surirellales lineage. A single origin of the canal raphe in the common ancestor of raphid pennates, followed by as many as seven losses in the lineages separating Bacillariales and Rhopalodiales+Surirellales, is a considerably less parsimonious and seemingly implausible alternative scenario. The clear implication of these phylogenetic results is that the canal raphe system has evolved in parallel in Bacillariales and Rhopalodiales+Surirellales and is therefore not homologous between the two lineages. As first suggested by Paddock and Sims (1977), these results suggest that the canal raphe is more a product of a complex set of more-or-less independently evolving constitutive parts than it is a single, coarsely defined character. One promising direction of follow-up research will be to conduct morphogenesis studies focused on the development of the canal raphe in the two distinct canal raphid lineages. These data could provide important insights into the underlying developmental basis for this remarkable case of parallel evolution.

728 E.C. Ruck and E.C. Theriot

99/83/-

97/74

96/-

*/97 82/*/* 89/-

*/*

70/72/*/93 */* */* 98/80 */89 */* 96/-

*/*

*/91 98/99/-

*/84 */83

A

71/85 */92 96/93

Eunotia curvata Eunotia pectinalis Eunotia glacialis Bacillaria paxillifer Nitzschia filiformis Cylindrotheca closterium Denticula kuetzingii Hantzschia amphioxys Tryblionella apiculata Gyrosigma acuminatum Navicula cryptocephala Mastogloia sp. Scoliopleura peisonis Neidium bisulcatum Neidium productum Neidium affine Amphora pediculus Diploneis subovalis Phaeodactylum tricornutum Berkeleya rutilans Climaconeis riddleae Lemnicola hungarica Cocconeis placentula Placoneis elginensis Gomphonema parvulum Gomphonema affine Fallacia monoculata Fallacia pygmea Caloneis lewisii Pinnularia termitina Pinnularia brebissonii Halamphora coffeaeformis Craticula cuspidata Stauroneis acuta Entomoneis sp. Entomoneis ornata Surirella sp. (Fastuosae) Campylodiscus clypeus Surirella ovata Campylodiscus sp. Cymatopleura elliptica Surirella splendida Stenopterobia curvula Rhopalodia contorta Rhopalodia gibba Rhopalodia sp. Epithemia turgida Epithemia sorex Epithemia argus

Bacillariales

*/97

93/82 */*/-

*/*

*/* */* 91/*/98 */*

*/87

*/98 91/- 99/96 */*

*/98

91/89/*/* */* */-

86/79/-

73/83/88/-

99/87

*/91

*/* 99/* */91 97/73 */83 */94 */* */* */98

B Surirellales

*/* */93

Eunotia curvata Eunotia glacialis Eunotia pectinalis Bacillaria paxillifer Hantzschia amphioxys Cylindrotheca closterium Denticula kuetzingii Nitzschia filiformis Tryblionella apiculata Gyrosigma acuminatum Navicula cryptocephala Lemnicola hungarica Placoneis elginensis Gomphonema affine Gomphonema parvulum Caloneis lewisii Pinnularia termitina Pinnularia brebissonii Fallacia pygmea Fallacia monoculata Diploneis subovalis Mastogloia sp. Phaeodactylum tricornutum Scoliopleura peisonis Neidium productum Neidium bisulcatum Neidium affine Climaconeis riddleae Cocconeis placentula Berkeleya rutilans Stauroneis acuta Craticula cuspidata Halamphora coffeaeformis Amphora pediculus Entomoneis sp. Entomoneis ornata Campylodiscus sp. Surirella ovata Cymatopleura elliptica Surirella sp. (Fastuosae) Campylodiscus clypeus Surirella splendida Stenopterobia curvula Rhopalodia sp. Rhopalodia gibba Rhopalodia contorta Epithemia turgida Epithemia argus Epithemia sorex

Rhopalodiales

Figure 3. Maximum likelihood cladograms inferred from the A) SSU rDNA and B) CPL datasets. Branch support is summarized above branches as Bayesian posterior probabilities (BPP)/ML bootstrap (BS) values that are ≥70%. BPP and ML BS values of 100% are shown as an asterisk (*). A dash (-) denotes a BPP or BS value lower than 70%. Canal raphe bearing groups are indicated on diagram as shaded branches.

Origin and Evolution of the Canal Raphe System in Diatoms 729 Eunotia curvata Eunotia glacialis Eunotia pectinalis */*

Bacillaria paxillifer

*/99

*/*

Hantzschia amphioxys Tryblionella apiculata Nitzschia filiformis

*/87 96/84 */97/-

Denticula kuetzingii Cylindrotheca closterium Gyrosigma acuminatum Navicula cryptocephala

*/*

Diploneis subovalis Scoliopleura peisonis Neidium productum

*/* */*

*/85

Neidium affine Neidium bisulcatum 98/70 Phaeodactylum tricornutum Mastogloia sp.

*/98

Climaconeis inflexa Berkeleya rutilans Cocconeis placentula

70/-

Lemnicola hungarica Placoneis elginensis

*/95 */*

Gomphonema parvulum Gomphonema affine

*/*

Stauroneis acuta

*/*

Craticula cuspidata Fallacia monoculata Fallacia pygmea

*/* */93

Caloneis lewisii

*/*

Pinnularia brebissonii Pinnularia termitina Amphora pediculus

*/91

Halamphora coffeaeformis 85/-

Entomoneis ornata

*/98

Entomoneis sp. Rhopalodia sp.

95/71 */*

*/*

*/98

Rhopalodia contorta

*/*

Rhopalodia gibba Epithemia turgida

*/*

Epithemia argus */97

Bacillariales

*/*

Surirellales Rhopalodiales 0.01 substitutions/site

Epithemia sorex Campylodiscus sp. Stenopterobia curvula Surirella splendida

-/75

*/99

71/*/94

Cymatopleura elliptica Surirella ovata Campylodiscus clypeus Surirella sp. (Fastuosae)

Figure 4. Maximum likelihood phylogram inferred from a concatenated dataset of the three markers: SSU rDNA, rbcL and psbC. Branch support is summarized above branches as Bayesian posterior probabilities (BPP)/ML bootstrap (BS) values ≥70%. BPP and ML BS values of 100% are shown as an asterisk (*). A dash (-) denotes a BPP or BS value lower than 70%. Canal raphe bearing groups are indicated on diagram as shaded branches.

730 E.C. Ruck and E.C. Theriot

A handful of morphogenetic studies have shown a consistent developmental pattern of the raphe whereby one raphe-sternum is laid down before the other, resulting in the formation of the actual raphe fissure between them (Round et al. 1990). Considerably more developmental variation has been observed in the formation of fibulae, which are a structurally integral component of the canal raphe system (see Introduction). In Surirella peisonis, the fibulae are formed by valve and mantle interstriae that are drawn together, eventually merging to form “praefibulae” that are further thickened into mature fibulae over the course of valve development (Schmid 1979). Two patterns of fibulae development have been observed in Bacillariales, neither of which resembles that of Surirella peisonis described above. Fibulae formation in Hantzschia amphioxys consistently starts from the valve face and proceeds toward the girdle side in both daughter cells (Pickett-Heaps 1983; Pickett-Heaps and Kowalski 1981). In contrast, in Tryblionella, Nitzschia sigmoidea, and Nitzschia sigma, the fibulae development proceeds from valve face toward the girdle side in one daughter cell and vice versa in the other daughter cell (Pickett-Heaps 1983). In short, there appears to be more than one way to make fibulae. Future studies will show whether the same can be said of other components of the canal raphe and whether taken together, consistent patterns emerge about the developmental evolution of these characters in the two canal raphid diatom lineages identified in this study.

Taxonomic Implications of the Raphid Diatom Phylogeny As emphasized above, the strongly supported relationship between the Rhopalodiales and Surirellales is a key finding of this study. The Rhopalodiales and Surirellales share a few gross morphological and cytological similarities besides a canal raphe system. For example, they both possess a single, lobed, and plate-like plastid (Round et al. 1990). The Rhopalodiales and several Surirellales (e.g. Auricula and Plagiodiscus) also share a wedge-shaped transapical frustule symmetry, the result of both a higher valve mantle and a widening of the girdle bands on the dorsal side. This kind of “cuneate” frustule morphology is also found within some genera in the Thalassiophysales, a few of which (Halamphora and Amphora) were included in this study. The cuneate valve shape might therefore be synapomorphic for the Thalassiophysales, certain members of the polyphyletic genus Surirella, and Rhopalodi-

ales. Formal coding of these shape characters and denser sampling of Surirellales, Thalassiophysales, and other cuneate diatoms will provide a rigorous test of this hypothesis. The phylogenetic instability of Amphora and Halamphora suggests that future phylogenetic studies should include more sequence data as well. The genus Undatella contains just three species and is currently classified within the Thalassiophysales. This genus is an example of the challenge Paddock and Sims (1977) posed with respect to interpretation of the canal and fibulate raphe systems. Within this genus, the occurrence of fibulae ranges from totally absent in Undatella magnifica, occurring only at the raphe ends in U. quadrata, and evenly spaced along the entire raphe in U. lineata (Paddock and Sims 1980). The raphe is also in a canal in the fibulate species, though insofar as we are aware, U. magnifica has not been studied in the SEM. Previous analyses using a single SSU rDNA sequence from an unidentified Undatella species placed it between the Eunotiales and Bacillariales (Bruder and Medlin 2008a, b; Theriot et al. 2009) or within the Bacillariales (Medlin et al. 2000). We included Undatella quadrata in a separate analysis using SSU rDNA data and found it to fall within the Bacillariales (not shown). A transfer of Undatella from the Thalassiophysales to the Bacillariales appears warranted, assuming additional analyses with the remaining Undatella species result in monophyly. If this were so, U. magnifica would represent the first known loss of fibulae within the nitzschoid lineage. Our phylogenetic results support monophyly of the Rhopalodiales, which is currently comprised of, Rhopalodia, Epithemia and Protokeelia (Round et al. 1990). Our analyses included two of three genera, Rhopalodia and Epithemia, which share similar, putatively synapomorphic chloroplast structure and valve symmetry. In addition, some Rhopalodia and all Epithemia species possess intracellular, nitrogen-fixing cyanobacterial endosymbionts (DeYoe et al. 1992; Drum and Pankratz 1965; Floener and Bothe 1980; Prechtl et al. 2004). Our study also seems to have clarified the position of taxonomically enigmatic genus, Denticula, which was originally classified within Nitzschiae (∼ =Bacillariaceae) but was transferred into Epithemiaceae sensu Hustedt (included Rhopalodia and Epithemia) (Hustedt 1930; Krammer and Lange-Bertalot 1988; LangeBertalot and Krammer 1987). Re-examination of the life history, valve symmetry, and raphe morphology led several researchers (Geitler 1977;

Origin and Evolution of the Canal Raphe System in Diatoms 731

Lange-Bertalot and Krammer 1993; Simonsen 1979) to conclude that Hustedt’s classification was incorrect and so they restored the taxonomic affinity of Denticula and Nitzschia (Simonsen 1979), a transfer that was retained by Round et al. (1990). This is the first study to analyze the phylogenetic position of Denticula, and our data place Denticula firmly within the Bacillariales and so support its current taxonomic placement. The SSU rDNA tree did not support monophyly of the Entomoneidaceae and Surirellaceae (Fig. 3A). However, both the CPL and three-gene datasets found the two taxa from the genus Entomoneis to be sister and strongly supported monophyly of the family Entomoneidaceae as well as monophyly of the Surirellaceae although statistical support is low in the CPL and moderate in the three-gene trees (Figs 3B, 4). Nonmonophyly of the genera Surirella and Campylodiscus and the sister relationship between Stenopterobia and the Surirella robustae group supports the morphological analysis of Ruck and Kociolek (2004). Also the close relationship between Cymatopleura elliptica and the Surirella pinnatae group is in agreement with the morphological analysis as it placed them in the same clade (Ruck and Kociolek 2004).

Concluding Remarks Historically, the possession of a canal raphe was considered support for a close evolutionary relationship among its bearers. Focused morphological studies of the canal raphe among the diverse canal raphe bearing taxa suggested a more complex evolutionary history (Paddock and Sims 1977). Previous studies have shown that two canal raphid lineages, Bacillariales and Surirellales, were not monophyletic. However, several key taxa were missing from these analyses, so the number of origins of the canal raphe has been unclear until now. By sampling all of the relevant lineages, we showed that the canal raphe system evolved twice—once in the common ancestor of Bacillariales and once in the common ancestor of Rhopalodiales and Surirellales, which our analyses identified as a monophyletic group.

Methods Taxon sampling and cell cultures: In the course of 100 years of diatom taxonomy, grouping and ranking of many taxa has been altered. Taxa that are relevant to this study have been referenced slightly differently by different authors over time. For example, what Hustedt (1930) called the Epithemiaceae

included Rhopalodia, Epithemia and Denticula Kützing. The latter genus is now in what is today the Bacillariaceae. Rhopalodia and Epithemia are now in what is called the Rhopalodiales along with Protokeelia Reimer & Lee. When referencing the historical literature, we will use the taxon name used by that author and identify the modern approximate equivalent (e.g., Epithemiaceae ∼ = Rhopalodiales). One nuclear (SSU rDNA) and two chloroplast (rbcL and psbC = CPL) DNA regions were amplified for 49 diatoms representing 9 of the 11 orders within the pennate lineage (Bacillariophyceae) (Table 2). Missing data for two orders, Lyrellales and Dictyoneidales, should not affect our results because both genera within the Lyrellales (Lyrella and Petroneis) are closely related to the Cymbellales (Jones et al. 2005), which is well-sampled in this study. Morphologically, the monogeneric Dictyoneidales is highly derived and autapomorphic and might be allied with Mastogloia (Round et al. 1990). Three species of Eunotia were used as outgroup taxa due to the consistent recovery of the Eunotiales as sister to all other raphid pennates (Alverson et al. 2006; Bruder and Medlin 2008a, b; Sims et al. 2006; Theriot et al., 2010). Diatoms were grown in batch culture at temperatures ranging from 14–27 ◦ C on a 12:12 light-dark cycle. Marine species were grown in f/2 medium (Guillard 1975; Guillard and Ryther 1962), and freshwater species were grown in either COMBO medium (Kilham et al. 1998) or biphasic soil medium (Czarnecki 1987). DNA extraction, PCR and sequencing: DNA was extracted according to manufacturer’s instructions with either the DNeasy® Plant Mini Kit (Qiagen, Hilden, Germany) or, if grown in biphasic media, the PowerSoil kit (MO BIO Laboratories, Carlsbad, CA, USA). We were unable to establish cultures for three taxa, so total genomic DNA was extracted via a modified Chelex 100® method from single cell isolations. Molecular markers were amplified by PCR using primer sequences provided in Table 3. PCR reactions consisted of 1.0-5.0 ␮L DNA extract; 6.5 ␮L of Failsafe Buffer E (Epicentre Technologies); 0.5 ␮L of each primer (20 ␮M stocks); 0.5 units Taq polymerase; and ddH2 O to a final volume of 25 ␮L. When necessary, a nested PCR strategy was used that involves two sets of primers in two successive runs of PCR, whereby the product from the initial PCR was used as template for the second reaction. PCR reactions were carried out in a PTC-200TM Peltier Thermal Cycler (MJ research, Watertown, MA, USA) or a DNA Engine (Bio-Rad, Hercules, CA, USA). The following PCR program was used for all reactions: 94 ◦ C for 3:30, 36 cycles of (94 ◦ C for 50 s, 52 ◦ C for 50 s, 72 ◦ C for 80 s), and final extension of 72 ◦ C for 15m. PCR products were purified by treatment with an Exonuclease I (Exo) and Shrimp Alkaline Phosphatase (SAP) protocol that included addition of 1.75 ␮l ddH2 O, 0.25 ␮L Exo, and 1.0 ␮L SAP per 25 ␮L reaction followed by heating for 30 minutes at 37 ◦ C and a 15 minute termination step at 80 ◦ C. Forward and reverse strands were cycle sequenced with BigDye (Applied Biosystems, Foster City, CA, USA) using a combination of primers (Table 3). Sequences were resolved with an ABI 3130 or 3730 DNA Analyzer (Applied Biosystems, Foster City, CA, USA) and sequence chromatograms were edited and assembled into contigs with Sequencher ver. 4.5 (Gene Codes Corporation, Ann Arbor, MI, USA). Sequence alignments and phylogenetic analyses: Terminal regions of each gene were trimmed to minimize the percentage of missing data, and one hypervariable region of the SSU rDNA alignment was excluded because of uncertainty in the alignment. Sequences for the protein coding genes rbcL and psbC were aligned manually with MacClade ver. 4.06 (Maddison and Maddison 2005). The nuclear SSU rDNA sequences were aligned using a template of 136 diatom SSU rDNA sequences, previously aligned

Taxon

Culture ID

Order

SSU

rbcL

psbC

Cocconeis placentula var. lineata (Ehr.) Van Heurck Lemnicola hungarica (Grun.) Round & Basson Denticula kuetzingii Grunow Tryblionella apiculata Greg. Hantzschia amphioxys var. maior Grun. in Van Heurck Cylindrotheca closterium (Ehr.) Reimann et Lewin Nitzschia filiformis (W. Sm.) Van Heurck Bacillaria paxillifer (O. Müll.) Hendey Gomphonema affine var. insigne (Greg.) Andrews Gomphonema parvulum (Kütz.) Kützing Placoneis elginensis (Greg.) Cox Eunotia curvata (Kütz.) Lagerstedt Eunotia glacialis Meister Eunotia pectinalis var. minor (Kütz.) Rabenhorst Mastogloia sp. Caloneis lewisii Patrick Stauroneis acuta W. Smith Climaconeis riddleae A.K.S.K. Prasad Berkeleya rutilans (Trentep.) Grunow Craticula cuspidata (Kütz.) D. G. Mann Diploneis subovalis Cleve Fallacia monoculata (Hust.) D. G. Mann Fallacia pygmea (Kütz.) A. J. Stickle & D. G. Mann Gyrosigma acuminatum (Kütz.) Rabenhorst Navicula cryptocephala Kützing Phaeodactylum tricornutum Bohlin Pinnularia brebissonii (Kütz.) Rabenhorst Pinnularia termitina (Ehr.) Patrick Neidium productum (W. Sm.) Cleve Neidium affine var. longiceps (Greg.) Cleve Neidium bisulcatum var. subampliatum Krammer Scoliopleura peisonis Grunow Rhopalodia sp. Rhopalodia contorta Hustedt Rhopalodia gibba (Ehr.) O. Müller Epithemia argus (Ehr.) Kützinga Epithemia sorex Kützinga

FD23 FD456 FD135 FD465 A4 CCMP1855 FD267 FD468 FD173 FD241 FD416 FD412 FD46 NIES461 29X07.6B FD54 FD51 17VI08.1C BRUT3616 FD35 FD282 FD254 FD294 FD317 FD109 CCMP2561 FD274 FD484 FD116 FD127 FD417 FD13 9vi08.1F.2 L1299 CH155 CH211 CH148

Achnanthales Achnanthales Bacillariales Bacillariales Bacillariales Bacillariales Bacillariales Bacillariales Cymbellales Cymbellales Cymbellales Eunotiales Eunotiales Eunotiales Mastogloiales Naviculales Naviculales Naviculales Naviculales Naviculales Naviculales Naviculales Naviculales Naviculales Naviculales Naviculales Naviculales Naviculales Naviculales Naviculales Naviculales Naviculales Rhopalodiales Rhopalodiales Rhopalodiales Rhopalodiales Rhopalodiales

HQ912592 HQ912626 HQ912610 HQ912600 HQ912404 HQ912645 HQ912589 HQ912627 HQ912608 HQ912595 HQ912607 HQ912599 HQ912586 HQ912636 HQ912632 HQ912580 HQ912579 HQ912644 HQ912637 HQ912581 HQ912597 HQ912596 HQ912605 HQ912598 HQ912603 HQ912556 HQ912604 HQ912601 HQ912582 HQ912583 HQ912591 HQ912609 HQ912405 HQ912406 HQ912407 HQ912408 HQ912409

HQ912456 HQ912490 HQ912474 HQ912464 HQ912390 HQ912509 HQ912453 HQ912491 HQ912472 HQ912459 HQ912471 HQ912463 HQ912450 HQ912500 HQ912496 HQ912444 HQ912443 HQ912508 HQ912501 HQ912445 HQ912461 HQ912460 HQ912469 HQ912462 HQ912467 HQ912420 HQ912468 HQ912465 HQ912446 HQ912447 HQ912455 HQ912473 HQ912391 HQ912392 HQ912393 HQ912394 HQ912395

HQ912285 HQ912319 HQ912303 HQ912293 HQ912376 HQ912338 HQ912282 HQ912320 HQ912301 HQ912288 HQ912300 HQ912292 HQ912279 HQ912329 HQ912325 HQ912273 HQ912272 HQ912337 HQ912330 HQ912274 HQ912290 HQ912289 HQ912298 HQ912291 HQ912296 HQ912250 HQ912297 HQ912294 HQ912275 HQ912276 HQ912284 HQ912302 HQ912377 HQ912378 HQ912379 HQ912380 HQ912381

732 E.C. Ruck and E.C. Theriot

Table 2. List of taxa included in this study, those taxon names and Genbank accession numbers in bold were newly generated. Classification system follows Round et al. 1990.

DNA isolations with Chelex 100® protocol. a Single-cell

Epithemia turgida (Ehr.) Kützinga Entomoneis ornata (Bailey) Reimer in Patrick & Reimer Entomoneis sp. Campylodiscus clypeus Ehrenberg Campylodiscus sp. Cymatopleura elliptica (Bréb.) W. Smith Surirella ovata Kützing Surirella sp. (Fastuosae group) Surirella splendida Ehrenberg Stenopterobia curvula (W. Sm.) Krammer Halamphora coffeaeformis (Cleve) Levkov Amphora pediculus (Kütz.) Grunow

CH154 14A CS782 L951 3613.8 L1333 L1241 DA1 19C L541 FD75 L1030

Rhopalodiales Surirellales Surirellales Surirellales Surirellales Surirellales Surirellales Surirellales Surirellales Surirellales Thalassiophysales Thalassiophysales

HQ912410 HQ912411 HQ912631 HQ912412 HQ912413 HQ912659 HQ912658 HQ912414 HQ912415 HQ912416 HQ912602 HQ912417

HQ912396 HQ912397 HQ912495 HQ912398 HQ912399 HQ912523 HQ912522 HQ912400 HQ912401 HQ912402 HQ912466 HQ912403

HQ912382 HQ912383 HQ912324 HQ912384 HQ912385 HQ912352 HQ912351 HQ912386 HQ912387 HQ912388 HQ912295 HQ912389

Origin and Evolution of the Canal Raphe System in Diatoms 733 against the SSU secondary structure, as a guide (Theriot et al. 2010). All new sequences were submitted to GenBank (Table 2), and alignments were deposited in TreeBase (http://purl.org/phylo/treebase/phylows/study/TB2:S11192). Three datasets were analyzed: the nuclear SSU rDNA alone, a combined chloroplast dataset of rbcL and psbC (CPL), and a concatenated alignment that included all three genes. Models of nucleotide substitution were selected for each marker using the Akaike Information Criterion (AIC) implemented in Modeltest 3.7 (Posada and Buckley 2004; Posada and Crandall 1998). Maximum likelihood (ML) analyses were conducted using RAxML ver. 7.0.4 (Stamatakis 2006) on each dataset (see Table 4 for dataset specifics). Each ML search was carried out 25 times with 500 bootstrap replicates each using the GTRGAMMA model and the run with the best log likelihood score was chosen as our maximum likelihood estimate (MLE). Bayesian inference (BI) analyses were performed using MrBayes v.3.1.2 (Huelsenbeck and Ronquist 2001) on all three datasets, each using default priors and the GTR+G+I model. Posterior probabilities were assessed in two runs using four MCMC chains with trees and parameters sampled every 1000 generations. See Table 4 for number of generations and burn-in information for each dataset. Stationarity was confirmed using the programs AWTY (Nylander et al. 2008) and Tracer ver. 1.5 (Rambaut and Drummond 2007). Hypothesis testing: Three hypotheses were tested regarding the evolutionary history of canal raphe bearing diatoms. The first constrained the orders Bacillariales, Rhopalodiales and Surirellales to monophyly (H0 =B+R+S), this hypothesis supports the canal raphe as a synapomorphic character that arose once in the diatoms. The second hypothesis constrained monophyly of Bacillariales + Rhopalodiales (H0 = B+R), this tests the ideas of Hustedt (1928, 1929, 1930) who grouped Denticula with members of the Rhopalodiaceae. The final hypothesis constrained monophyly of Bacillariales + Surirellales (H0 = R+S), another hypothesis proposed by Hustedt (1929) but one not supported by molecular studies to date. We used three different methods to test these hypotheses. First, using maximum likelihood, we performed parametric bootstrapping (the SOWH test), with the three-gene dataset as the observed data (HA ) (Goldman et al. 2000; Hillis et al. 1996; Huelsenbeck et al. 1996). The SOWH test was performed by simulating 500 matrices according to the three null tree topologies and the GTR+G+I model, with parameters derived from the combined three-gene dataset. Simulations were conducted using Seq-Gen 1.3.2 (Rambaut and Grassly 1997) with the SG Runner 1.5.3 front-end (Wilcox 2004). All simulated matrices were analyzed using the ML criterion as implemented in GARLI 0.96 (Zwickl 2006). Each simulated matrix was analyzed twice, once to calculate the best unconstrained tree and once to calculate the best tree constrained to reflect the null hypothesis. We calculated the difference in likelihood score between these two trees for our observed dataset and compared it to the null distribution generated from the simulated datasets to calculate a test-statistic. Second, we used CONSEL (Shimodaira and Hasegawa 2001) to implement the Approximately Unbiased (AU) test (Shimodaira 2002) with all three datasets. The AU test uses site wise log likelihoods and the multi-scale bootstrap technique to generate a confidence set of trees. If the best tree fitting the null hypothesis falls outside this confidence set of trees, then the hypothesis can be rejected. Third, we calculated the probability of trees consistent with the null hypotheses from the Bayesian post-stationarity distribution of trees for all three datasets (Hedtke et al. 2008; Verbruggen and Theriot 2008). Bayesian analyses result in topologies being sampled in proportion to their posterior probability once stationarity is

734 E.C. Ruck and E.C. Theriot Table 3. Primers used to amplify and sequence SSU rDNA, rbcL and psbC fragments from study taxa. Primer Name SSU1a SSU11+b SSU301+ SSU850+ SSU1004+ SSU1451+ SSU568SSU870SSU1147SSU1672-b ITS1DRa rbcL66+a rbcL40+b rbcL404+ rbcL587rbcL1255rbcL1444-b nd6+ dp7-a psbC+a psbC22+b psbC221+ psbC857psbC1154-b psbC-a a Forward b Forward

Primer Sequence (5 -3 )

Reference

SSU rDNA AAC CTG GTT GAT CCT GCC AGT TGA TCC TGC CAG TAG TCA TAC GCT ATC ATT CAA GTT TCT GCC C GGG ACA GTT GGG GGT ATT CGT A CGA AGA TGA TTA GAT ACC ATC G TGT GAT GCC CTT AGA TGT CCT GG CAG ACT TGC CCT CCA ATT GA TAC GAA TAC CCC CAA CTG TCC C AGT TTC AGC CTT GCG ACC ATA C TAG GTG CGA CGG GCG GTG T CCT TGT TAC GAC TTC ACC TTC C rbcL TTA AGG AGA AAT AAA TGT CTC AAT CTG GGA CTC GAA TYA AAA GTG ACC G GCT TTA CGT TTA GAA GAT ATG GTC TAA ACC ACC TTT TAA MCC TTC TTG GTG CAT TTG ACC ACA GT GCG AAA TCA GCT GTA TCT GTW G GTA AAT GGA TGC GTA TG AAA SHD CCT TGT GTW AGT YTC psbC ACA GGM TTY GCT TGG TGG AGT GG CGT GGT GAT ACA TAG TTA ACG CAT TGT TTC ACC ACC CTT TGG TTA TGA CTG GCG TG GCD CAY GCT GGY TTA ATG G CAC GAC CWG AAT GCC ACC AAT G

Medlin et al. 1988 Alverson et al. 2007 Alverson et al. 2007 This study Alverson et al. 2007 Alverson et al. 2007 Alverson et al. 2007 This study Alverson et al. 2007 This study Edgar and Theriot 2004 Alverson et al. 2007 This study This study Alverson et al. 2007 Alverson et al. 2007 This study Daugbjerg and Andersen 1997 Daugbjerg and Andersen 1997 Alverson et al. 2007 This study Alverson et al. 2007 Alverson et al. 2007 This study Alverson et al. 2007

and reverse primers for initial amplification reaction. and reverse primers for second amplification reaction when nested PCR was performed.

Table 4. Sequence data, evolutionary models and Log-likelihood values (-ln L) from ML estimations. Parameter

SSU

rbcL

psbC

Combined CPL*

Combined 3-gene

Number of sequences Initial aligned length (bp) Positions excluded

49 1797 1–15 643–690 1667–1797 1603 498 358 2(2) 15M 1.5M GTR+G+I 11087.406

49 1571 1–19 1487–1571

49 1231 1–28 1172–1231

49

49

1467 518 391

1143 427 362

GTR+G+I

GTR+G+I

2610 945 753 2(2) 40M 8M GTR+G+I 26280.159

4213 1443 1111 2(2) 60M 6M GTR+G+I 38052.487

Final aligned length (bp) Variable chars. Parsimony inform. chars. Bayesian runs (used) Bayesian generations Bayesian burn-in ML/BI model (AIC) MLE -ln L * CPL

refers to the combined chloroplast markers rbcL + psbC.

Origin and Evolution of the Canal Raphe System in Diatoms 735

Table 5. Results of hypothesis testing by parametric bootstrapping, Approximately Unbiased (AU) test, and estimation of Bayesian posterior probabilities. Hypothesis are as follows: H1 = monophyly of Bacillariales + Rhopalodiales + Surirellales, H2 = monophyly of Bacillariales + Rhopalodiales, and H3 = monophyly of Bacillariales + Surirellales. Significance <0.001 denoted “*”. Dataset

Hypothesis

SSU

H1 H2 H3 H1 H2 H3 H1 H2 H3

CPL 3-gene

␦OBS a

100.53 288.86 288.86

db

0-4.47 0-4.59 0-3.37

PPB c

PAU d

PBPP e

<0.001* <0.001* <0.001*

0.071 <0.001* <0.001* <0.001* <0.001* <0.001* <0.001* <0.001* <0.001*

0.000* 0.000* 0.000* 0.000* 0.000* 0.000* 0.000* 0.000* 0.000*

a Observed test statistic for parametric bootstrapping = ML tree score for observed data with constraint enforce – unconstrained ML tree score for observed data. b Range for ␦’s expected distribution in parametric bootstrapping with 500 simulated datasets. c Parametric bootstrapping p-values. d AU-test p-values. e Bayesian posterior probability calculated by filtering topologies consistent with null hypothesis from distribution of post stationarity trees divided by the total number of post stationarity trees.

reached. Therefore, the posterior probabilities of null hypotheses can be determined by filtering the post burn-in samples using the desired constraint topologies. Post burn-in trees were loaded into PAUP* (ver. 4b10, Swofford 2002) and filtered to find those trees consistent with our three null hypotheses. The probability of a particular constraint topology is the proportion of trees matching this constraint divided by the total number of post burn-in trees.

Acknowledgements We thank David Cannatella, Beryl Simpson, Robert Jansen, John La Claire II, and Andrew Alverson for critical comments and suggestions. We also thank Teofil Nakov, Matt Ashworth and Mariska Brady for helpful discussions. E.C. Ruck was supported by NSF PEET grant DEB 0118883 and NSF AToL grant AToL0629564 awarded to E.C. Theriot.

References

Bruder K, Medlin LK (2008a) Morphological and molecular investigations of naviculoid diatoms. II. Selected genera and families. Diatom Res 23:283–329 Bruder K, Medlin LK (2008b) Morphological and molecular investigations of naviculoid diatoms. III. Hippodonta and Navicula S.S. Diatom Res 23:331–347 Czarnecki DB (1987) The freshwater diatom culture collection at Loras College, Dubuque, Iowa. Not Nat 465:1–16 Daugbjerg N, Andersen RA (1997) A molecular phylogeny of the heterokont algae based on analyses of chloroplast-encoded rbcL sequence data. J Phycol 33:1031–1041 DeYoe H, Lowe RL, Marks JC (1992) Effects on nitrogen and phosphorus on the endosymbiont load of Rhopalodia gibba and Epithemia turgida (Bacillariophyceae). J Phycol 28:773–777 Drum RW, Pankratz S (1965) Fine structure of an unusual cytoplasmic inclusion in the diatom genus Rhopalodia. Protoplasma 60:141–149 Edgar SM, Theriot EC (2004) Phylogeny of Aulacoseira (Bacillariophyta) based on molecules and morphology. J Phycol 40:772–788

Alverson AJ, Theriot EC (2005) Comments on the recent progress toward reconstructing the diatom phylogeny. J Nanosci Nanotechnol 5:57–62

Floener L, Bothe H (1980) Nitrogen Fixation in Rhopalodia gibba, a Diatom Containing Blue-greenish Inclusions Symbiotically. In Schwemmler W, Schenk HEA (eds) Endocytobiology, Endosymbiosis and Cell Biology. Walter de Gruyter & Co, Berlin, pp 541–552

Alverson AJ, Jansen RK, Theriot EC (2007) Bridging the Rubicon: Phylogenetic analysis reveals repeated colonization’s of marine and fresh waters by thalassiosiroid diatoms. Mol Phylogenet Evol 45:193–210

Geitler L (1977) Zur Entwicklungsgeschichte der Epithemiaceen Epithemia, Rhopalodia und Denticula (Diatomophyceae) und ihre vermutlich symbiontischen Sphäroidkörper. Plant Syst Evol 128:259–275

Alverson AJ, Cannone JJ, Gutell RR, Theriot EC (2006) The evolution of elongate shape in diatoms. J Phycol 42: 655–668

Goldman N, Anderson JP, Rodrigo AG (2000) Likelihoodbased tests of topologies in phylogenetics. Syst Biol 49:652–670

736 E.C. Ruck and E.C. Theriot Guillard RRL (1975) Culture of Phytoplankton for Feeding Marine Invertebrates. In Smith WL, Chanley MH (eds) Culture of Marine Invertebrate Animals. Plenum Press, New York, pp 26–60 Guillard RRL, Ryther JH (1962) Studies of marine planktonic diatoms. I. Cyclotella nana Hustedt and Detonula confervacea (Cleve) Gran. Can J Microbiol 8:229–239 Heath TA, Hedtke SM, Hillis DM (2008) Taxon sampling and the accuracy of phylogenetic analyses. J Syst Evol 46: 239–257 Hedtke SM, Stanger-Hall K, Baker RJ, Hillis DM (2008) Allmale asexuality: Origin and maintenance of androgenesis in the Asian clam Corbicula. Evolution 62:1119–1136 Hendey NI (1964) An Introductory Account of the Smaller Algae of British Coastal Waters, Part V: Bacillariophyceae (Diatoms). Fisheries Investigation Series IV: HMSO, London, pp 1–317. Hillis DM (1998) Taxonomic sampling, phylogenetic accuracy, and investigator bias. Syst Biol 47:3–8

Band 2/2. Gustav Fischer Verlag, Stuttgart, New York, pp 1– 596 Lange-Bertalot H, Krammer K (1987) Bacillariaceae, Epithemiaceae. Surirellaceae. Biblioth Diatomol 15:1–289 Lange-Bertalot H, Krammer K (1993) Observations on Simonsenia and some small species of Denticula and Nitzschia. Nova Hedwigia 106:121–131 Mann DG (2000) Auxospore formation and neoteny in Surirella angusta (Bacillariophyta) and a modified terminology for cells of Surirellaceae. Nova Hedwigia 71:165–183 Maddison WP, Maddison DR (2005) MacClade 4: Analysis of Phylogeny and Character Evolution. Sinauer Associates, Sunderland, Massachusetts Medlin LK, Kaczmarska I (2004) Evolution of the diatoms V: Morphological and cytological support for the major clades and a taxonomic revision. Phycologia 43:245–270

Hillis DM, Mable BK, Moritz C (1996) Applications of Molecular Systematics: The State of the Field and a Look into the Future. In Hillis DM, Moritz C, Mable BK (eds) Molecular Systematics. Sinauer Associates, Sunderland, Massachusetts, USA, pp 1–655

Medlin LK, Kooistra WCHF, Schmid A-MM (2000) A Review of the Evolution of the Diatoms – a Total Approach Using Molecules, Morphology and Geology. In Witkowski A, Sieminska J (eds) The Origin and Early Evolution of the Diatoms: Fossil, Molecular and Biogeographical Approaches. W. Szafer Institute of Botany, Polish Academy of Sciences, Cracow, pp 13–35

Hillis DM, Pollock DD, McGuire JA, Zwickl DJ (2003) Is sparse taxon sampling a problem for phylogenetic inference? Syst Biol 52:124–126

Medlin LK, Elwood HJ, Stickel S, Sogin ML (1988) The characterization of enzymatically amplified eukaryotic 16S-like rRNA-coding regions. Gene 71:491–499

Huelsenbeck JP, Hillis DM, Jones R (1996) Parametric Bootstrapping in Molecular Phylogenetics: Applications and Performance. In Ferraris JD, Palumbi SR (eds) Molecular Zoology: Advances, Strategies and Protocols. Wiley-Liss, New York, pp 19–45

Nylander JAA, Wilgenbusch JC, Warren DL, Swofford DL (2008) AWTY (are we there yet?): a system for graphical exploration of MCMC convergence in Bayesian phylogenetic inference. Bioinformatics 24:581–583

Huelsenbeck JP, Ronquist F (2001) MRBAYES: Bayesian inference of phylogenetic trees. Bioinformatics 17:754–755 Hustedt F (1928) Untersuchungen über den Bau der Diatomeen. IV. Zur Morphologie und Systematik der Gattungen Denticula und Epithemia. Ber Deutsch Bot Ges 46:148–157 Hustedt F (1929) Untersuchungen über den Bau der Diatomeen. VII-VIII. VII. Weitere Untersuchungen über die Kanalraphe der Nitzschioideae. VIII. Untersuchungen über die Kanalraphe der Gattung Surirella. Ber Deutsch Bot Ges 47:101–111 Hustedt F (1930) Bacillariophyta (Diatomae). In Pascher A (ed) Die Süsswasser-Flora Mitteleuropas. Gustav Fischer Verlag, Jena, pp 1–466 Jones HM, Simpson GE, Stickle AJ, Mann DG (2005) Life history and systematics of Petroneis (Bacillariophyta), with special reference to British waters. Eur J Phycol 40:61–87 Kilham SS, Kreeger DA, Goulden CE, Lynn S (1998) COMBO: A defined freshwater culture medium for algae and zooplankton. Hydrobiologia 377:147–159 Krammer K (1989) Valve morphology and taxonomy in the genus Stenopterobia (Bacillariophyceae). Br Phycol J 24:237–243 Krammer K, Lange-Bertalot H (1988) Bacillariophyceae. 2. Teil. Bacillariaceae, Epithemiaceae, Surirellaceae. In Ettl H, Gerloff J, Heynig H (eds) Süsswasserflora von Mitteleuropa.

Paddock TBB, Sims PA (1977) A preliminary survey of the raphe structure of some advanced groups of diatoms (Epithemiaceae-Surirellaceae). Nova Hedwigia Beih 54:291–322 Paddock TBB, Sims PA (1980) Observations on the marine diatom genus Auricula and two new genera, Undatella and Proboscidea. Bacillaria 3:161–196 Peragallo H, Peragallo M (1897-1908) Diatomées marines de France et des districts maritimes voisins. Micrographe-Editeur, M.J. Tempère, Grez-sur-Loing, pp 1–491 Pickett-Heaps JD (1983) Valve morphogenesis and the microtubule center in three species of the diatom Nitzschia. J Phycol 19:269–281 Pickett-Heaps JD, Kowalski SE (1981) Valve morphogenesis and the microtubule center of the diatom Hantzschia amphioxys. Eur J Cell Biol 25:150–170 Posada D, Buckley TR (2004) Model selection and model averaging in phylogenetic: advantages of Akaike information criterion and Bayesian approaches over likelihood ratio tests. Syst Biol 53:793–808 Posada D, Crandall KA (1998) Modeltest: testing the model of DNA substitution. Bioinformatics 14:817–818 Prechtl J, Kneip C, Lockhart P, Wenderoth K, Maier U-G (2004) Intracellular spheroid bodies of Rhopalodia gibba have nitrogen-fixing apparatus of cyanobacterial origin. Mol Biol Evol 21:1477–1481

Origin and Evolution of the Canal Raphe System in Diatoms 737 Rambaut A, Drummond AJ (2007) Tracer v1.4, Available at http://beast.bio.ed.ac.uk/Tracer (last accessed 6 October 2010).

Sorhannus U (2004) Diatom phylogenetics inferred based on direct optimization of nuclear-encoded SSU rDNA sequences. Cladistics 20:487–497

Rambaut A, Grassly NC (1997) Seq-Gen: An application for the Monte Carlo simulation of DNA sequence evolution along phylogenetic trees. Comput Appl Biosci 13:235–238

Sorhannus U (2007) A nuclear-encoded small-subunit ribosomal RNA timescale for diatom evolution. Mar Micropaleontol 65:1–12

Ross R, Cox EJ, Karayeva NI, Mann DG, Paddock TBB, Simonsen R, Sims PA (1979) An amended terminology for the siliceous components of the diatom cell. Nova Hedwigia 64:513–533

Stamatakis A (2006) RaxML-VI-HPC; Maximum likelihoodbased phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics 22:2688–2690

Round FE, Crawford RM, Mann DG (1990) The Diatoms: Biology & Morphology of the Genera. Cambridge University Press, Cambridge, pp 1–747 Ruck EC, Kociolek JP (2004) Preliminary phylogeny of the family Surirellaceae (Bacillariophyta). Biblioth Diatomol 50:1–236 Schmid A-MM (1979) The development of structure in the shells of diatoms. Nova Hedwigia Beih 64:219–236 Schrader H-J (1973) Types of raphe structures in the diatoms. Nova Hedwigia 45:195–230 Shimodaira H (2002) An approximately unbiased test of phylogenetic tree selection. Syst Biol 51:492–508 Shimodaira H, Hasegawa M (2001) CONSEL: for assessing the confidence of phylogenetic tree selection. Bioinformatics 17:1246–1247 Simonsen R (1979) The diatom system: Ideas on phylogeny. Bacillaria 2:9–71 Sims PA, Paddock TBB (1982) The fenestral fibula: A new structure in the diatoms. Bacillaria 5:7–42 Sims PA, Mann DG, Medlin LK (2006) Evolution of the diatoms: insights from fossil, biological and molecular data. Phycologia 45:361–402

Swofford DL (2002) PAUP*: Phylogenetic Analysis Using Parsimony (* and other methods). Sinauer Associates, Sunderland, Massachusetts Theriot EC, Cannone JJ, Gutell RR, Alverson AJ (2009) The limits of nuclear-encoded SSU rDNA for resolving the diatom phylogeny. Eur J Phycol 44:277–290 Theriot EC, Ashworth M, Ruck EC, Nakov T, Jansen RK (2010) A preliminary multigene phylogeny of the diatoms (Bacillariophyta): Challenges for future research. Plant Ecol Evol 143:278–296 Verbruggen H, Theriot EC (2008) Building trees of algae: some advances in phylogenetic and evolutionary analysis. Eur J Phycol 43:229–252 Wilcox T (2004) SG Runner vers. 1.5.3: a MAC OSX shellwrapper for Seq-Gen. Published by the author, available at http://homepage.mac.com/tpwilcox/ (last accessed 6 October 2010). Zwickl DJ (2006) Genetic Algorithm Approaches for the Phylogenetic Analysis of Large Biological Sequence Datasets Under the Maximum Likelihood Criterion. Ph.D. Dissertation, The University of Texas at Austin, pp 1–115. Zwickl DJ, Hillis DM (2002) Increased taxon sampling greatly reduces phylogenetic error. Syst Biol 51:588–598