Bru2006b

Bru2006b
I
Dissertation
zur Erlangung des Akademischen Grades
eines Doktors der Naturwissenschaften
-Dr. rer. nat.im Fachbereich 2 (Biologie/Chemie)
der Universität Bremen
vorgelegt von
Katrin Bruder
Bremen, Mai 2006
Erster Gutachter: Prof. Dr. G.O. Kirst
Zweiter Gutachter: Prof. Dr. Ulrich Bathmann
Tag und Ort des öffentlichen Kolloquiums:
11. September 2006, Universität Bremen
II
Hiermit erkläre ich, dass ich die vorliegende Dissertation selbständig verfasst und
keine anderen als die angegebenen Quellen und Hilfsmittel verwendet habe. Die
entnommenen Stellen aus benutzten Werken wurden wörtlich oder inhaltlich als solche
kenntlich gemacht
Katrin Bruder
Contents
III
Table of Contents
Summary ................................................................................................................................... 1
Zusammenfassung.................................................................................................................... 3
1. Introduction .......................................................................................................................... 5
1.1 Diatom systematics .......................................................................................................... 5
1.2 Some problematic genera ................................................................................................. 6
1.2.1 Navicula .................................................................................................................... 6
1.2.2 Pinnularia and Caloneis............................................................................................ 7
1.3 Molecular phylogenetics .................................................................................................. 8
1.3.1 Nuclear-encoded rRNA genes................................................................................... 8
1.3.1.1 Small subunit rRNA gene ................................................................................ 10
1.3.1.2 Large subunit rRNA gene ................................................................................ 10
1.3.2 Plastid-encoded protein-coding genes..................................................................... 10
1.3.2.1 rbcL gene.......................................................................................................... 11
1.3.3 Gene combination ................................................................................................... 11
1.3.4 Molecular phylogenies of diatoms .......................................................................... 12
1.4 Aims of this study .......................................................................................................... 13
2. Materials and Methods ...................................................................................................... 15
2.1. Cultures ......................................................................................................................... 15
2.2. DNA Methods ............................................................................................................... 21
2.2.1 DNA isolation ......................................................................................................... 21
2.2.2. PCR ........................................................................................................................ 21
2.2.3. Sequencing ............................................................................................................. 22
2.3. Sequence Analysis......................................................................................................... 22
2.4. Microscopy.................................................................................................................... 26
2.4.1 Purification of the frustules ..................................................................................... 26
2.4.2. Slide preparation .................................................................................................... 26
3. Results ................................................................................................................................. 27
3.1 Molecular data................................................................................................................ 27
3.1.1 SSU rRNA gene ...................................................................................................... 27
Contents
IV
3.1.2 LSU rRNA gene ...................................................................................................... 36
3.1.3 rbcL gene................................................................................................................. 43
3.1.4 Gene combination ................................................................................................... 51
3.2 Morphological support for molecular data..................................................................... 55
3.2.1 Navicula sensu stricto ............................................................................................. 55
3.2.2 Amphora .................................................................................................................. 56
3.2.3 Pinnularia and Caloneis.......................................................................................... 65
3.2.4 Stauroneis, Craticula and Navicula integra............................................................ 72
3.2.5 Gomphonema .......................................................................................................... 75
3.2.6 Placoneis and Navicula hambergii ......................................................................... 82
3.2.7 Cymbella.................................................................................................................. 82
3.2.8 Navicula brockmannii ............................................................................................. 84
3.2.9 Varieties of Mayamaea atomus............................................................................... 84
4. Discussion............................................................................................................................ 92
4.1 Comparison of the gene trees ......................................................................................... 92
4.1.1 Phylogenies based on the AlgaTerra cultures ......................................................... 93
4.1.1.1 Phylogenies based on SSU rDNA sequences................................................... 93
4.1.1.2 Phylogenies based on LSU rDNA sequences .................................................. 94
4.1.1.3 Phylogenies based on rbcL gene sequences..................................................... 95
4.1.1.4 Phylogenies based on the combined dataset .................................................... 96
4.1.1.5 General results of the analyses of the AlgaTerra cultures................................ 97
4.1.2 Phylogenies based on enlarged datasets................................................................ 100
4.1.2.1 Phylogenies based on SSU rDNA sequences................................................. 100
4.1.2.2 Phylogenies based on LSU rDNA sequences ................................................ 102
4.1.2.3 Phylogenies based on rbcL gene sequences................................................... 103
4.1.3 General relationships of the genera....................................................................... 103
4.2 Relationships within the genera ................................................................................... 105
4.2.1 Navicula sensu stricto ........................................................................................... 105
4.2.2 Amphora ................................................................................................................ 105
4.2.3 Caloneis and Pinnularia........................................................................................ 107
4.2.4 Navicula integra.................................................................................................... 109
4.2.5 Gomphonema ........................................................................................................ 110
4.2.6 Placoneis and Navicula hambergii ....................................................................... 111
Contents
V
4.2.7 Cymbella................................................................................................................ 112
4.2.8 Navicula brockmannii ........................................................................................... 112
4.2.9 Varieties of Mayamaea atomus............................................................................. 113
References ............................................................................................................................. 114
Appendix ............................................................................................................................... 126
Acknowledgements............................................................................................................... 141
Summary
1
Summary
The recent taxonomy of diatoms is basically based on investigations of valve morphology,
cell components and life cycle (e.g., Round et al., 1990). But the development of PCR has
facilitated the use of DNA sequences for inferring phylogenies. Based on morphology, the
taxonomy of the family Naviculaceae (sensu Krammer & Lange-Bertalot, 1986) has been
highly changed (e.g., Round et al., 1990), but little work has been carried out with molecular
data for this large and ecologically important group of diatoms.
My thesis was aimed at the investigation of evolutionary relationships within the naviculoid
pennates using molecular and morphological data. Ninety-one cultures containing 72 species
of 22 genera were isolated and their morphology examined. Sixty-two of these species belong
to the Naviculaceae. Phylogenies based on sequences of the nuclear-encoded SSU rRNA
gene, the LSU rRNA gene, the chloroplast rbcL gene and a combined dataset were compared.
The SSU rRNA gene is the most widely used gene for inferring phylogenetic relationships.
The combination of conserved and variable regions in this gene allows studies of most
phylogenetic relationships. Because the D1/D2-region of the LSU rRNA gene comprises
more highly variable areas than the SSU rRNA gene, a stronger phylogenetic signal for
closely related species was estimated. Also the rbcL gene was used in this study to obtain
clearer information of evolution at lower (order to genus) levels of taxonomic hierarchy in
diatoms. But in this study the trees based on the LSU rDNA and rbcL gene sequences do not
provide stronger supported results for closely related species. The analyses of the combined
dataset resulted in trees with higher bootstrap support than the analyses of the single genes,
although partition homogeneity test resulted in a very low p-value. The results of the partition
homogeneity test should not be used to determine whether or not to combine data sets for
phylogenetic analysis.
This study confirms the assumption that the genus Navicula sensu lato is a very heterogenous
group and my results support the monophyly of Navicula sensu stricto. The separation of
Craticula, Eolimna, Hippodonta, Luticola, Mayamaea and Placoneis from the genus could be
confirmed. “Navicula” species, which do not belong to the section Lineolatae could be
recombined: Navicula integra is the type species of a newly described genus Prestauroneis
Bruder, gen. nov. (Type species: Prestauroneis integra (W. Smith) Bruder comb. nov.);
Navicula brockmannii is transferred to the genus Adlafia (A. brockmannii (Hustedt) Bruder
comb. nov.) and Navicula hambergii is placed within Placoneis (Placoneis hambergii
(Hustedt) Bruder comb. nov.). The differences of their sequences indicates that M. atomus var.
Summary
2
atomus and M. atomus var. permitis were not just two varieties of the same species but two
different species.
The monophyly of the genera Cocconeis, Craticula, Cymbella, Encyonema, Eunotia,
Gomphonema, Lyrella, Mayamaea, Placoneis, Pleurosigma and Sellaphora is supported by
the recent study. But the actual differentiation of the genera Caloneis and Pinnularia is
rejected. The molecular results support groups defined by Krammer & Lange-Bertalot (1985),
based on the morphology of the internal openings of the alveoli. A genus that should be
subdivided is Amphora. Molecular and morphological data strongly support a separation of
the subgenus Halamphora. But further investigations on the subgenus Halamphora is needed
because the results of the analysis from SSU rDNA sequences indicate that this is still an
artificial group. This study does neither support nor refuse a separation of Cymbella, because
of the different results in the molecular phylogenies.
This study also resolve several relationships between different genera. Hippodonta is shown
to be sister to Navicula sensu stricto. The family Stauroneidaceae could be recovered and the
addition of the newly describes genus Prestauroneis to this family is proposed. The results
also support to include the genus Mayamaea into the suborder Sellaphorineae, which could be
recovered in most phylogenies. The marine and freshwater monoraphid genera are clearly
separated. The marine genera form the sister clade to the Bacillariales, whereas the freshwater
monoraphid genera diverge within the naviculoid pennates. The relationship between the
freshwater monoraphid genera and the naviculoid pennates could not be resolved
unambiguously but they might be close relatives of Adlafia brockmannii and the Cymbellales.
The monophyly of the order Cymbellales is strongly supported, but the results contradict the
arrangement of the families Cymbellaceae and Gomphonemataceae, because in most trees
Gomphonema (Gomphonemataceae) diverge within the Cymbellaceae. The order Naviculales
and the suborder Naviculineae as used in Round et al. (1990) are shown to be heterogenous in
all trees.
Zusammenfassung
3
Zusammenfassung
Die Taxonomie der Diatomeen basiert vor allem auf Untersuchungen der Valvenmorphologie,
der Zellkomponenten und des Zellzyklus (z.B. Round et al., 1990). Die Entwicklung der PCR
hat zusätzlich die Verwendung von DNA-Sequenzen bei der Ermittlung von Stammbäumen
ermöglicht. Die Taxonomie der Familie Naviculaceae (sensu Krammer & Lange-Bertalot,
1986) wurde bereits aufgrund morphologischer Untersuchungen stark verändert (z.B. Round
et al., 1990), aber es gibt nur wenige molekularbiologische Arbeiten für diese große und
ökologisch wichtige Gruppe der Diatomeen.
Ziel meiner Arbeit war die Untersuchung der evolutionären Verhältnisse zwischen
naviculoiden Diatomeen unter Verwendung von molekularen und morphologischen Daten.
Insgesamt wurden 91 Kulturen, die 72 Arten aus 22 Gattungen enthielten, isoliert und ihre
Morphologie untersucht. Zur Familie Naviculaceae gehören 62 dieser Arten. Von allen
Kulturen wurden die Sequenzen der im Zellkern vorliegenden SSU rDNA und LSU rDNA
sowie des im Chloroplastengenom kodierten rbcL Gens bestimmt. Die auf den einzelnen
Genen sowie einem kombinierten Datensatz basierenden Phylogenien wurden verglichen.
Zur Bestimmung phylogenetischer Beziehungen wird meist das SSU rRNA Gen verwendet.
Durch die Kombination konservierter und variabler Regionen eignet es sich für die
Untersuchung der meisten phylogenetischer Beziehungen. Die D1/D2-Region der LSU rDNA
beinhaltet mehr hoch variable Regionen als die SSU rDNA, weshalb ein stärkeres
phylogenetisches Signal bei nah verwandten Arten erwartet wurde. Auch die Verwendung des
rbcL Gens sollte eine bessere Auflösung der Evolution auf einem niedrigeren Level (Ordnung
bis Gattung) erzielen. Die auf der LSU rDNA und dem rbcL Gen basierenden Phylogenien
zeigen in dieser Studie aber keine eindeutigeren Ergebnisse für nah verwandte Arten. Die
Analyse des kombinierten Datensatzes ergab die am besten durch Bootstrap-Werte
unterstützten Phylogenien, obwohl der „partition homogeneity test“ einen sehr niedrigen pWert ergab. Dies unterstützt, dass das Ergebnis dieses Testes nicht entscheiden sollte, ob
mehrere Datensätze kombiniert analysiert werden oder nicht.
Diese Studie bestätigt die Annahme, dass die Gattung Navicula sensu lato eine sehr
heterogene Gruppe ist. Zusätzlich unterstützen meine Ergebnisse die Monophylie Navicula
sensu stricto. Die Abspaltung von Craticula, Eolimna, Hippodonta, Luticola, Mayamaea und
Placoneis von der Gattung konnte bestätigt werden. “Navicula” Arten, die nicht zur Sektion
Lineolatae gehören, konnten neu zugeordnet werden: Navicula integra ist die Typus-Art der
neu beschriebenen Gattung Prestauroneis Bruder, gen. nov. (Typus-Art: Prestauroneis
Zusammenfassung
4
integra (W. Smith) Bruder comb. nov.); Navicula brockmannii ist zur Gattung Adlafia (A.
brockmannii (Hustedt) Bruder comb. nov.) und Navicula hambergii zur Gattung Placoneis
(Placoneis hambergii (Hustedt) Bruder comb. nov.) überführt worden. Die Unterschiede ihrer
Sequenzen lassen vermuten, dass es sich bei M. atomus var. atomus und M. atomus var.
permitis nicht nur um zwei Varietäten sondern um zwei Arten handelt.
Die Monophylie der Gattungen Cocconeis, Craticula, Cymbella, Encyonema, Eunotia,
Gomphonema, Lyrella, Mayamaea, Placoneis, Pleurosigma und Sellaphora konnte im
Rahmen dieser Studie bestätigt werden. Dagegen widerlegen die Ergebnisse die derzeitige
Trennung der Gattungen Caloneis and Pinnularia. Stattdessen werden die von Krammer &
Lange-Bertalot (1985) definierten Gruppen, die sich vor allem durch die Morphologie ihrer
internen Alveolenöffnungen unterscheiden, unterstützt. Die Gattung Amphora sollte weiter
unterteilt werden. Sowohl molekulare als auch morphologische Daten unterstützen eine
Abtrennung der Untergattung Halamphora. Es sind jedoch weitere Untersuchungen der
Untergattung Halamphora notwendig, da die Analyse der SSU rDNA Sequenzen andeutet,
dass es sich bei dieser Untergattung noch immer um eine künstliche Gruppe handelt. Auf der
Basis dieser Studie kann eine Aufteilung der Gattung Cymbella weder widerlegt noch
befürwortet werden, da sich die Beziehungen innerhalb dieser Gattung in den einzelnen
Phylogenien unterscheiden.
Diese Studie klärt auch einige Beziehungen zwischen verschiedenen Gattungen auf. So zeigen
die Ergebnisse, dass Hippodonta die Schwestergattung von Navicula sensu stricto ist. Die
Familie Stauroneidaceae konnte bestätigt und die neu beschriebene Gattung Prestauroneis zu
dieser Familie hinzugefügt werden. Aufgrund dieser Studie sollte die Gattung Mayamaea in
die Unterordnung Sellaphorineae eingegliedert werden. Innerhalb der monoraphiden
Gattungen zeigt sich eine klare Trennung der marinen und der Süßwasser-Arten. Die marinen
Gattungen bilden die Schwestergruppe der Bacillariales, während sich die Süßwasser-Arten
innerhalb der naviculoiden Diatomeen abspalten. Das Verhältnis zwischen den monoraphiden
Süßwasser-Gattungen und den naviculoiden Diatomeen konnte nicht eindeutig geklärt
werden, aber meine Ergebnisse weisen auf eine nahe Verwandtschaft mit Adlafia brockmannii
und der Ordnung Cymbellales hin. Die Ergebnisse bestätigen die Monophylie der Ordnung
Cymbellales, aber sie widersprechen der Einteilung der Familien Cymbellaceae und
Gomphonemataceae, da sich Gomphonema (Gomphonemataceae) in fast allen Phylogenien
innerhalb der Cymbellaceae abspaltet. Die Ordnung Naviculales und die Unterordnung
Naviculineae, wie sie in Round et al. (1990) eingeteilt wurden, haben sich in allen
Phylogenien als heterogen erwiesen.
1. Introduction
5
1. Introduction
“Few objects are more beautiful than the minute siliceous cases of the
diatomaceae: were these created that they might be examined and admired under
the higher powers of the microscope?” (Darwin, 1859)
1.1 Diatom systematics
Diatom valves were one of the favourite subjects for study by the early microscopists and the
first diatom was described at the beginning of the 1700s (Round et al., 1990). The description
of diatom species and their taxonomy has been traditionally based on light-microscopical
studies of valve shape and structure. With the introduction of electron microscope techniques,
more details of valve structure (e.g., the areolae, processes or tubes) were visible. Although
diatom classification depends to a great extent on valve morphology, features of the living
cell (e.g., number and form of chloroplasts and pyrenoids) and ecology have also been taken
into account (Mereschkowsky, 1903, Cox & Williams, 2000).
Based on their valve morphology, Schütt (1896) separated the diatoms into two main groups:
Centric diatoms with a radial symmetry and bilaterally symmetrical pennate diatoms. Later
the pennate group was subdivided into species with a raphe slit in at least one valve and those
species without a raphe (e.g., Hustedt 1961-1966, Round et al., 1990). The raphe slit is
necessary for diatom locomotion. This classification implies that centrics and pennates each
represent natural evolutionary lineages. But in fossil records, centric diatoms have been
recovered from Jurassic and Late Cretaceous (e.g., Rothpeltz, 1896, Strelnikowa &
Martirosjan, 1981, Gersonde & Harwood, 1990, Harwood & Gersonde, 1990), whereas
araphid pennate diatoms appear in the Late Cretaceous (e.g., Moshkovitz et al., 1983). Raphid
pennate diatoms, which today represent the most diverse group, have been recovered from
Tertiary (Strelnikova, 1990). In some phylogenetic analyses the centric diatoms grade into the
pennate diatoms (e.g., Kooistra et al., 2003, Sorhannus, 2004). Other molecular phylogenies
show two different clades (e.g., Medlin et al., 2000, Medlin & Kaczmarska, 2004). These
studies differ in the number of used sequences, in species composition and in the outgroup
used. But none of the phylogenies reflect the traditional groups. The centric and the araphid
pennate diatoms are shown to be paraphyletic. Only the raphid pennate diatoms and the
pennate diatoms are monophyletic in all studies.
1. Introduction
6
Medlin and Kaczmarska (2004) proposed a revised classification based on molecular data,
morphological and cytological features:
Subdivision Coscinodiscophytina Medlin & Kaczmarska
Class Coscinodiscophyceae Round & Crawford, emend. Medlin & Kaczmarska, which
comprises the “radial” centrics;
Subdivision Bacillariophytina Medlin & Kaczmarska
Class Mediophyceae (Jousè & Proshkina-Lavrenko) Medlin & Kaczmarska, which
comprises the “multipolar” centrics plus the radial Thalassiosirales;
Class Bacillariophyceae Haeckel, emend. Medlin & Kaczmarska, which comprises the
pennate diatoms.
Study performed by Guillou et al. (1999) and Daugbjerg & Guillou (2001) based on different
genes have shown the Bolidophyceae to be the sister group to the diatoms.
1.2 Some problematic genera
1.2.1 Navicula
The genus Navicula was described by Bory de Saint-Vincent in 1922 based on Navicula
tripunctata (O. F. Müller) Bory. Within the diatoms, this genus is probably the largest and
most diverse because “Navicula has traditionally been a dump for all bilaterally symmetrical
raphid diatoms lacking particularly distinctive features” (Round et al., 1990, p. 566).
Nevertheless, with electron microscopy and the investigation of living cells, the true
morphological diversity of the genus became apparent. Therefore, taxonomic revisions of this
genus are being made or have been carried out and new genera described or old genera
resurrected. Since the description of the genus, the taxonomic treatment of the naviculoid
diatoms has undergone major changes.
Today most diatomists agree that Navicula (sensu stricto) should be used only for species that
belong to the section Lineolatae (sensu Cleve, 1895 and Hustedt, 1930). Navicula sensu
stricto encompasses approximately 200 species, which predominantly (about 150 species)
inhabit freshwater environment (Witkowski et al., 1998). There are still many species named
Navicula that do not belong to this group, but several older genera have been resurrected
(e.g., Placoneis Mereschkowsky in Cox, 1987) and new genera were described and separated
from Navicula sensu stricto because they differ clearly in valve morphology and/or
1. Introduction
7
chloroplast features, e.g., Eolimna (Schiller & Lange-Bertalot, 1997), Hippodonta (LangeBertalot, Metzeltin & Witkowski, 1996) Luticola (Mann, in Round et. al., 1990) or
Mayamaea (Lange-Bertalot, 1997). But not all new genera have been accepted by all
diatomists. For example the separation of the genus Hippodonta is under discussion. In her
investigation of the variation of valve morphology, Cox (1999) doubted the correctness of this
separation. In her study, she could find examples of all characters used to define the genus
Hippodonta in other species of Navicula, but no cytological or reproductive evidence that
would support their separation. Therefore she proposed that the species allocated to
Hippodonta be recognised as a subgenus of Navicula and to enlarge the generic description of
Navicula to cover this.
1.2.2 Pinnularia and Caloneis
The genus Pinnularia was described by Ehrenberg based on P. viridis (Ehrenberg, 1843). In
1894, Cleve described the genus Caloneis with C. amphisbaena as its type and distinguished
the genus from Pinnularia on the basis of light microscopy. He already noted that “smaller
forms of Caloneis with indistinct longitudinal lines closely resemble small Pinnulariae, and
certain of the panduriform species seem to be closely connected with some marine,
panduriform Pinnulariae” (Cleve, 1894).
Since then, many diatomists investigating the two genera have tried to find morphological
characters to make a clear distinction between the two genera. The separation of the genera
Caloneis Cleve and Pinnularia Ehrenberg is discussed controversial: Some infer from their
results, that there is a distinguishing combination of characters to recognise each genus easily
(e.g., Krammer & Lange-Bertalot, 1985, Krammer, 2000). In addition to this conclusion
Krammer & Lange-Bertalot (1985) mentioned a potential separation in three groups: (1) all
species whose alveoli are internally nearly open, as existing in Pinnularia interrupta; (2)
species with partially closed alveoli, e.g., Caloneis amphisbaena and Pinnularia gibba; (3)
species with nearly closed alveoli, like Caloneis silicula.
Other scientists saw great difficulty in distinguishing Caloneis from Pinnularia and consider
it is no longer possible to make a clear distinction. Based on valve morphology and
chloroplast features, Cox (1988 b) concluded, that “there is as much or as little similarity
between Pinnularia and Caloneis as they presently stand, as between species within each.”
Her investigation of the live structure supported three groups, which are different to those
mentioned by Krammer and Lange-Bertalot (1985): (1) Caloneis silicula, Caloneis bacillum
and Pinnularia isostauron; (2) Caloneis based on C. amphisbaena; (3) Pinnularia based on P.
1. Introduction
8
nobilis. Round et al. (1990, p. 556) “were unable to find a satisfactory basis for the traditional
separation of Pinnularia from Caloneis … and conclude that if Pinnularia is ever split, it will
not be along the traditional boundary between the two genera”.
Mann (2001) also doubted the correctness of the traditional Pinnularia-Caloneis distinction
and comes to the conclusion, that “until we have a clearer idea of relationships within the
Pinnulariaceae, especially from gene sequence data, it may be best to accept the unsatisfactory
classification that we have, rather than attempt to produce a new one that might be worse”
(Mann, 2001, p. 34). But hitherto no extensive phylogenetic analysis based on molecular data
has been made.
1.3 Molecular phylogenetics
It has long been evident, that there is useful information about evolutionary history in gene
sequences. The wide application of this method began with the appearance of the polymerase
chain reaction (PCR) in mid-1980 (Mullis et al., 1986, Mullis and Faloona, 1987, Saiki et al.,
1988). Coupled with the direct didesoxynucleotide sequencing of amplified products, the
technique became a powerful tool in life sciences. Sequences of several genes were used to
reconstruct phylogenies of prokaryotes (e.g., Woese, 1987), single-cell eukaryotes (e.g.,
Medlin et al., 1997) and higher plants (e.g., Soltis et al., 2000) and animals (e.g., Söller et al.,
2000). Interest in phylogeny reconstruction has increased so rapidly that now roughly 4,000
articles that include a phylogenetic tree are published each year (Pagel, 1999).
1.3.1 Nuclear-encoded rRNA genes
Because rRNA genes serve a pivotal role in the protein synthesis machinery they occur
universally in prokaryotic and eukaryotic cells without a change in their function. Because
helical formation occurs in their secondary structure (Fig.1), which cannot change otherwise
the function of the molecule would be lost, different regions evolve at very different rates
(Woese, 1987). This combination of conserved and variable regions allows studies of most
phylogenetic relationships from studies of deep phylogeny (e.g., Cavalier-Smith, 2004) to
microdiversity surveys (e.g., Sáez et al., 2003).
The rRNA genes are combined in multigene families with up to thousands of copies arranged
in tandem arrays. Each individual repeat consists of the small subunit rRNA gene (SSU rRNA
gene, SSU rDNA), the gene encoding the 5.8S rRNA, the large subunit rRNA gene (LSU
rRNA gene, LSU rDNA) and two internal transcribed spacers, known as ITS 1 and ITS 2. The
1. Introduction
9
Fig.1: SSU rRNA secondary structure model of Bacillaria paxillifer
(The European Ribosomal RNA databank, http://rrna.uia.ac.be/)
internal transcribed spacers are located between the regions coding for small subunit rRNA
and 5.8S rRNA, and between the latter and the large subunit rRNA coding region. In addition,
an external transcribed spacer (ETS) occurs upstream to the small subunit rRNA gene. These
transcription units were separated by an intergenic spacer (IGS). (Long & Dawid, 1980)
The multiple copies of this cluster appear to be highly homogenised within an organism and
among different individuals of the same species. The main mechanism for this concerted
evolution seem to be gene conversion between sister chromatids after replication and unequal
crossing-over between homologous chromosomes (Schlötterer & Tautz, 1994). The high
number of homogenized copies avoids the extensive sampling required for most single-copy
genes. But some exceptions of the usual gene homogenization are known. For instance in
some species of the protist Plasmodium, two different types of SSU rDNA exist, whose
1. Introduction
10
expression is linked to different stages of the parasitic life cycle of this organisms (Gunderson
et al., 1987, Waters et al., 1989, Qari et al., 1994).
1.3.1.1 Small subunit rRNA gene
The SSU rRNA gene is the most widely used gene for inferring phylogenetic relationships.
Thousands of partial and complete sequences (approx. 1800 bp in eukaryotes) from
prokaryotes, single-celled and multicellular eukaryotes can be found in internet-available
databases like GenBank (http://www.ncbi.nlm.nih.gov/). In diatoms, the gene has been used
to study their position within the heterokont algae (e.g., Daugbjerg & Andersen, 1997), to
reconstruct the evolution of the major classes (e.g., Medlin & Kaczmarska, 2004) or to assess
the monophyly of diatom orders or genera (e.g., Beszteri et al., 2001).
Kooistra & Medlin (1996) calculated a relatively fast substitution rate (1% per 18 to 26 Ma)
in the SSU rDNA of diatoms. In the same study it was proven, that the evolutionary rate
differs within the diatoms. In particular, the SSU rDNA of pennate taxa evolve more slowly
than in the other diatom orders.
1.3.1.2 Large subunit rRNA gene
The LSU rRNA gene comprises more highly variable areas than the SSU rRNA gene (Van
der Auwera & De Wachter, 1998). This indicates a stronger phylogenetic signal for closely
related species in comparison with the SSU rRNA gene. But it may cause problems for
reconstructing deep phylogenies because of saturation effects, the signal might be indistinct.
Additionally, highly variable sequences are difficult to align. Because of the large size of LSU
rDNA (over 3300 bp) complete sequences of this region are rare. Typically used sequences
are derived from several parts of the gene, for example approximately 600 bp from the 5’ end
of 26S rDNA (D1/D2 region).
1.3.2 Plastid-encoded protein-coding genes
Not all DNA in eukaryotes is stored within the cell nucleus. Organelles, like mitochondria or
chloroplasts, contain their own DNA. Organelle genomes usually consist of a single DNA
molecule and each gene is normally present only once. The chloroplast genome contains
predominantly protein-coding genes. In protein-coding genes the evolution rate diverges
between the different codon positions: The mutation rate at the third position is higher than
the rates at the first or second position, because nucleotide changes at the third position in
most cases are synonymous mutations. Synonymous mutations have no influence on the
1. Introduction
11
amino acid coded and thus it appears that they depend only on the background mutation rate.
But nucleotide changes at the first or second codon position nearly always lead to
nonsynonymous substitutions, which result in a change of the encoded amino acid. Therfore
the third codon position is downweighted or omitted very often, if protein-coding genes are
used for phylogenetic analyses.
1.3.2.1 rbcL gene
The enzyme ribulose-1,5-bisphosphate carboxylase (RUBISCO) is responsible for fixation of
carbon dioxide in the Calvin cycle. The holoenzyme is formed by a 16-mer structure that
includes eight identical chloroplast-encoded large subunit polypeptides and eight small
subunit polypeptides. The rbcL gene encodes the large subunit of RUBISCO and is located in
a single-copy region of the chloroplast genome. It is typically 1428-1434 bp in length and
insertions or deletions are extremely rare (Soltis & Soltis, 1998). Although some chloroplastencoded genes are interrupted by introns, this is not the case for the rbcL gene (Clegg, 1993).
This positional conservation of coding information permits the unambiguous alignment of
rbcL sequences.
The relative rate of evolution of SSU rRNA and rbcL genes varies among groups. The rbcL
gene generally evolves about three times faster than SSU rDNA in angiosperms but is slower
in Orchidaceae (Soltis & Soltis, 1998). Within the phaeophytes, a slightly faster mutation rate
of the rbcL gene has been observed (Draisma & Prud’homme van Reine). Compared to SSU
rDNA, the rbcL gene appears more suited to studies of evolution at lower (order to genus)
levels of taxonomic hierarchy in diatoms (Mann et al., 2001).
1.3.3 Gene combination
A gene phylogeny based on a single gene may not agree with the organismal phylogeny
because of such biological processes as introgression, lineage sorting and gene duplication
(Hillis, 1987, Doyle, 1992, Lutzoni & Vilgalys, 1995). Therefore phylogenetic trees derived
from different data sets may also differ. If the primary interest is the phylogeny of organisms
rather than genes, this problem of differential phylogenetic history among data sets argues for
the use of multiple data sets, often concatenated. Several studies have suggested that data sets
should not be combined if the data partitions are heterogenous (e.g., Bull et al., 1993,
Huelsenbeck et al., 1996). The incongruence length difference (ILD) test (Farris et al., 1994)
or the equivalent partition homogeneity test (Swofford, 1995) have been used to determine
whether or not to combine data sets for phylogenetic analysis (e.g. Johnson & Sorensen, 1998,
1. Introduction
12
Hoot et al., 1999). But other studies have found that P-values < 0,05 should not preclude
dataset combination (e.g., Sullivan, 1996, Davis et al., 1998, Flynn & Nedbal, 1998, Yoder et
al., 2001).
Both simulations (e.g., Hillis, 1996, Graybeal, 1998) and empirical studies (e.g., Soltis et al.,
1998, Soltis et al., 2000) indicate that additional data can improve phylogenetic inferences of
molecular phylogenies. For example, analyses of angiosperm relationships on the basis of
gene sequences for rbcL, atpB and 18S rDNA showed increased resolution and internal
support (as measured by bootstrap values), and faster run times when the data sets for these
genes were combined rather than analysed separately (Soltis et al., 1998, Soltis et al., 2000).
Sorhannus (2001) analysed heterokont phylogeny based on a combined dataset (SSU rDNA,
LSU rDNA, rbcL gene and morphological data) using one exemplar of each major group. But
he did not find greatly increased support among class relationships in his analysis.
1.3.4 Molecular phylogenies of diatoms
Most molecular phylogenies of diatoms have been reconstructed from the nuclear-encoded
small subunit (SSU) and the large subunit (LSU) ribosomal RNA genes (Medlin et al., 1991,
1993, Sorhannus et al., 1995, Kooistra & Medlin, 1996, Medlin et al., 1996 a, b, Van der
Auwera & De Wachter, 1998, Medlin et al., 2000, Beszteri et al., 2001, Lundholm &
Moestrup, 2002, Lundholm et al., 2002 a,b, Kooistra et al., 2003, Behnke et al., 2004, Medlin
& Kaczmarska, 2004, Sorhannus, 2004). In addition the internal transcribed spacer regions in
the nuclear-encoded ribosomal DNA cistron (Zechmann et al., 1994, Behnke et al., 2004), the
mitochondrion-encoded cytochrome c oxidase subunit I (coxA, Ehara et al., 2000), the
chloroplast-encoded elongation factor Tu (tufA, Delwiche et al., 1995, Medlin et al., 1997),
the chloroplast-encoded RNA polymerase alpha subunit (rpoA, Fox & Sorhannus, 2003) and
the chloroplast-encoded ribulose-1,5-bisphosphate carboxylase large subunit (rbcL,
Daugbjerg & Andersen, 1997, Daugbjerg & Guillou, 2001, Mann et al., 2001) have been used
for studying molecular systematics in diatoms or their relationship within the heterokont
algae.
The majority of molecular studies investigating the evolution of diatoms have used species
from all classes (e.g., Medlin et al., 1993, Sorhannus et al., 1995, Medlin et al., 1996 a, b,
2000, Kooistra et al., 2003, Medlin & Kaczmarska, 2004, Sorhannus, 2004). But in these
studies most orders are represented by three or less species. Few molecular studies have been
carried out with focus on some closely related genera (e.g., Zechmann et al., 1994, Beszteri et
al., 2001, Lundholm & Moestrup, 2002, Lundholm et al., 2002 a, b, Behnke et al., 2004).
1. Introduction
13
Only two of these studies (Beszteri et al., Behnke et al., 2004) concentrated on species
belonging to the Naviculaceae sensu Krammer and Lange-Bertalot (1986). Beszteri et al.
(2001) determined SSU rDNA sequences of six naviculoid species. Their results slightly
contradicted the monophyly of the Naviculaceae, because Gomphonema parvulum did not
cluster within this group. Based on their data Beszteri et al. (2001) concluded, that further
SSU rDNA sequences from close relatives of G. parvulum could possibly reinforce or reject
the hypothesis about Naviculaceae being a monophyletic group. In more recent studies based
on a large number of sequences (Kooistra et al., 2003, Medlin & Kaczmarska, 2004,
Sorhannus, 2004) G. parvulum cluster within naviculoid diatoms. But the Naviculaceae did
not form a monophyletic group in these studies, because genera like Surirella (family
Surirellaceae) or Cocconeis (family Achnanthaceae) cluster within the Naviculaceae. The
study of Behnke et al. (2004) concentrated on the genus Sellaphora and interclonal
relationships of several clones of S. pupula. In the SSU rDNA phylogeny shown in this study,
the Naviculaceae form a monophyletic group. But this tree did not include species belonging
to the families Surirellaceae or Achnanthaceae. This was the first dataset containing a
Navicula sensu stricto (N. cryptocephala) and a Navicula sensu lato (N. pelliculosa, section
Minusculae) and the two species were clearly separated in the inferred phylogeny. The
greatest number of naviculoid species was present in the dataset used by Sorhannus (2004).
Even there only four genera were represented by more than one species. In the shown
phylogeny inferred with SSU rDNA sequences only the genus Gomphonema (represented by
two species) formed a monophyletic clade. Amphora (three species), Eolimna (two species)
and Navicula sensu stricto (two species) did not form a monophyletic group.
1.4 Aims of this study
Since the electron microscopy was introduced to diatom research and features of live cells,
ecology and molecular data were taken into account, many changes in diatom taxonomy have
occurred. The taxonomy of the family Naviculaceae (sensu Krammer & Lange-Bertalot,
1986) has been changing greatly. Based on morphology the whole family, as well as many of
its genera, have undergone revisions (e.g., Round et al., 1990). But little work has been
carried out with molecular data for this large and ecologically interesting group of diatoms.
In order to estimate evolutionary relationships within the Naviculaceae (sensu Krammer &
Lange-Bertalot, 1986) and to access the nomenclatural problems I performed phylogenetic
analyses of several freshwater species. But a gene tree based on a single gene does not
necessarily agree with the true species tree, that represents the actual evolutionary pathway of
1. Introduction
14
the species involved. Therefore three different genes were sequenced for each culture and
phylogenies were reconstructed for each gene and a phylogenetic analysis based on a
combined data set of all three genes was conducted. Adittionally the morphology of the
sequenced species was investigated.
2. Materials and Methods
15
2. Materials and Methods
2.1. Cultures
The cultures used in this study were established within the scope of the AlgaTerra project
(http://www.algaterra.net/). The field samples were taken from several terrestrial, freshwater
and brackish habitats in northern Germany (Fig.2). Between November 2001 and September
2003 220 samples from 83 different sites were taken.
Fig.2: Sampling sites (map from Stiefel Verlag GmbH, Lenting)
Cultures were initiated from these samples using a DY-IV medium (Andersen et al., 1997)
mixed 2:1 with filter-sterilized (pore size: 0,1 µm) water from the sampling sites. After one to
four days, clonal cultures were isolated from these initial cultures. Most of these isolates still
contain small flagellates. In order to purge these flagellates from the cultures a small number
of diatom cells was transferred to fresh medium several times and than grown on agar plates
(see recipe below) for one to three weeks. From these plates a small number of diatom cells
were transferred to liquid medium. If necessary the entire procedure was repeated several
times.
2. Materials and Methods
16
Recipe for agar plates:
• 9 g Agar Agar was diluted in ½ litre deionised water and autoclaved
• Double concentrated DY-IV medium was filter-sterilized (pore size: 0,1 µm)
• Both mixtures were temperated to approximately 60°C and mixed 1:1.
All isolates were grown under a 14/10 light/dark cycle with photon flux densities between 30
and 120 µM photons m-2 s-1 at 15°C. The clonal cultures were grown in modified DY IV
medium (Andersen et al., 1997) enriched with 5%-10% soil-extract (see recipe below). For
isolates from alkaline, acid or brackish habitats the media was adjusted by addition of sodium
hydroxide, hydrochloric acid or IMR-media (Eppley et al., 1967).
Recipe for soil-extract:
• One l dry nonfertilized garden soil (J.Arthur Bower's African Violet Compost, William
Sinclair Horticulture Ltd.) was saturated with bidistilled water and infused for several
days at room temperature.
• After autoclaving, the hot water/soil-mixture was filtered through a laboratory paper filter.
• Afterwards, the mixture was filtered several times with with stepwise reduced pore size
(10 µm, 5µm, 3µm and 2µm).
The 91 cultures used for this study contain 72 species belong to 22 genera and were isolated
from 45 different field samples. Eighty-one cultures contain 62 species belonging to the
family Naviculaceae. Because monoraphid species of the family Achnanthaceae cluster within
the Naviculaceae in several studies (Kooistra et al., 2003, Medlin & Kaczmarska, 2004,
Sorhannus, 2004), I additionally used sequences of species belonging to this family. Three
cultures contain Eunotia species. Centric and araphid species were used as outgroup. All
cultures grown for this study and their place of origin are shown in Table 1.
2. Materials and Methods
17
Tab. 1: List of diatom cultures established and sequenced within the scope of this study.
DNApreparation
place of origin
culture
author
GPS
discription
source
1438
AT_196Gel02 Achnanthidium minutissimum
(Kützing) Czarnecki
54°10,97N; 10°37,92E Ukelei See
lake, plankton
1427
AT_212.06
Amphora cf. fogediana
Krammer
54°19,86N; 10°17,72E Dobersdorfer See
lake, benthos
1264
AT_117.10
Amphora libyca
Ehrenberg
53°09,51N; 08°42,57E Lesum, near river mouth
river, plankton
1263
AT_105Gel05 Amphora normannii
Rabenhorst
53°09,90N; 08°45,10E Wümme
river, benthos
1265
AT_117.11
Amphora pediculus
(Kützing) Grunow
53°09,51N; 08°42,57E Lesum, near river mouth
river, plankton
1554
AT_221.04
Amphora sp.
Ehrenberg ex Kützing
53°06,41N; 08°11,23E Hunte, near Hundsmühlen
river, plankton
AT_67.02b
Asterionella formosa
Hassall
53°13,79N; 08°41,06E Geeste, bridge near Bramel
river, plankton
1550
AT_177.07
Caloneis amphisbaena
(Bory) Cleve
53°04,08N; 08°29,04E Hasbruch, near hunting lodge ditch, benthos
1323
AT_220.06
Caloneis budensis
(Grunow) Krammer
53°06,41N; 08°11,23E Hunte, near Hundsmühlen
riverside, soil
1446
AT_160Gel04 Caloneis lauta
Carter & Bailey-Watts
52°57,65N; 08°20,67E Poggenpohls Moor
soil, moss
1415
AT_212.07
Ehrenberg
54°19,86N; 10°17,72E Dobersdorfer See
lake, benthos
1418
AT_212Gel11 Cocconeis placentula
Ehrenberg
54°19,86N; 10°17,72E Dobersdorfer See
lake, benthos
1318
AT_200.05
Craticula cuspidata
(Kützing) D.G. Mann
54°11,69N; 10°36,24E Krumm See
lake, benthos
1320
AT_219.03
Craticula cuspidata
(Kützing) D.G. Mann
53°06,41N; 08°11,23E Hunte, near Hundsmühlen
river, benthos
1283
AT_5Nav02
Craticula halophilioides
(Hustedt) Lange-Bertalot
53°09,65N; 08°43,40E Maschinenfleet
canal, plankton
1308
AT_36klein
Craticula halophilioides
(Hustedt) Lange-Bertalot
53°12,72N; 08°26,85E Weser, near Rekum
river, benthos
1284
AT_70Gel14a Craticula molestiformis
(Hustedt) Lange-Bertalot
53°13,79N; 08°41,06E Geeste, bridge near Bramel
riverside, moss
1256
(1, 3)
1493 (2, 3) AT_L1840
(1)
species
Cocconeis pediculus
Cyclotella choctawatcheeana
Prasad
Geeste, near Bremerhaven
river
1414
AT_204Gel02 Cymbella affinis
Kützing
54°09,09N; 10°27,45E Großer Madebroken See
lake, plankton
1423
AT_213.04
Kützing
54°19,86N; 10°17,72E Dobersdorfer See
lake, periphyton
1421
AT_210Gel07 Cymbella aspera
(Ehrenberg) Cleve
54°09,98N; 10°25,19E Trammer See
lake, periphyton
1431
AT_194Gel07 Cymbella helmckei
Krammer
54°08,53N; 10°39,70E Großer Eutiner See
lake, benthos
1317
AT_177.04
(Auerswald) Cleve
53°04,08N; 08°29,04E Hasbruch, near hunting lodge ditch, benthos
Cymbella affinis
Cymbella naviculiformis
DNA and SSU rDNA sequence provided by I. Jung;
(2)
DNA and SSU rDNA sequence provided by B. Beszteri;
(3)
species used as outgroup
2. Materials and Methods
18
Tab. 1: Continued
DNApreparation
culture
1324
AT_221.02
1422
species
author
discription
source
53°06,41N; 08°11,23E Hunte, near Hundsmühlen
river, plankton
AT_210Gel13 Cymbella proxima
Reimer
54°09,98N; 10°25,19E Trammer See
lake, periphyton
1441
AT_214Gel03 Encyonema caespitosum
Kützing
54°19,86N; 10°17,72E Dobersdorfer See
lake, benthos
1266
AT_137.13
Encyonema minutum
(Hilse) D.G. Mann
53°41,96N; 11°29,15E Schweriner See
lake, plankton
1267
AT_70Gel18
Eolimna minima
(Grunow) Lange-Bertalot
53°13,79N; 08°41,06E Geeste, bridge near Bramel
riverside, moss
1268
AT_111Gel09 Eunotia formica
Ehrenberg
53°11,39N; 08°47,05E Hamme, near sluice
river, plankton
1321
AT_219.07
Eunotia implicata
Nörpel, Lange-Bertalot & Alles
53°06,41N; 08°11,23E Hunte, near Hundsmühlen
river, benthos
1269
AT_73Gel02
Eunotia sp.
Ehrenberg
53°38,11N; 10°44,56E Pinnsee
lake, periphyton
AT_185Gel03 Fragilaria crotonensis
Kitton
54°08,53N; 10°39,70E Großer Eutiner See
river, plankton
1410
AT_124.05b
Lyngbye
53°33,00N; 10°55,16E Schaalsee, Zarrentiner Becken lake, benthos
1445
AT_108Gel03 Frustulia vulgaris
(Thwaites) De Toni
53°10,89N; 08°45,70E Hamme, near bridge
river, benthos
1424
AT_219Gel10 Gomphonema acuminatum
Ehrenberg
53°06,41N; 08°11,23E Hunte, near Hundsmühlen
river, benthos
1439
AT_196Gel03 Gomphonema affine
Kützing
54°10,97N; 10°37,92E Ukelei See
lake, plankton
1322
AT_219Gel06 Gomphonema affine
Kützing
53°06,41N; 08°11,23E Hunte, near Hundsmühlen
river, benthos
1409
AT_109Gel08b Gomphonema cf. angustatum
(Kützing) Rabenhorst
53°10,89N; 08°45,70E Hamme, near bridge
river, plankton
1315
AT_161.15
Gomphonema cf. parvulum
(Kützing) Kützing
puddle, soil
1270
AT_117.09
Gomphonema micropus
Kützing
52°57,65N; 08°20,67E Poggenpohls Moor
53°09,51N; 08°42,57E Lesum, near river mouth
river, plankton
1271
AT_117Gel21 Gomphonema micropus
Kützing
53°09,51N; 08°42,57E Lesum, near river mouth
river, plankton
1313
AT_160Gel27 Gomphonema productum
(Grunow) Lange-Bertalot & Reichardt
52°57,65N; 08°20,67E Poggenpohls Moor
soil, moss
1552
AT_195Gel09 Gomphonema truncatum
1272
AT_124.24
Ehrenberg
54°08,53N; 10°39,70E Großer Eutiner See
lake, periphyton
(Ehrenberg) Lange-Bertalot, Metzeltin &
Witkowski
53°33,00N; 10°55,16E Schaalsee, Zarrentiner Becken lake, benthos
1273
AT_104Gel12a Luticola goeppertiana
1274
AT_115Gel07 Mayamaea atomus var. atomus (Kützing) Lange-Bertalot
(1, 3)
Cymbella naviculiformis
GPS
(Auerswald) Cleve
1254
(1)
place of origin
Fragilaria sp.
Hippodonta capitata
DNA and SSU rDNA sequence provided by I. Jung;
(3)
(Bleisch) D.G. Mann
species used as outgroup
53°09,90N; 08°45,10E Wümme
river, plankton
53°11,79N; 08°48,11E Hamme, near Osterholz
river, benthos
2. Materials and Methods
19
Tab. 1: Continued
DNApreparation
place of origin
culture
species
author
GPS
discription
source
1275
AT_101Gel04 Mayamaea atomus var. permitis (Hustedt) Lange-Bertalot
53°40,20N; 10°50,21E Schwarze Kuhle
lake, periphyton
1425
AT_111Gel10 Navicula brockmannii
Hustedt
53°11,39N; 08°47,05E Hamme, near sluice
river, plankton
1417
AT_212Gel07 Navicula capitatoradiata
Germain
54°19,86N; 10°17,72E Dobersdorfer See
lake, benthos
1310
AT_82.04c
Ehrenberg
53°36,36N; 10°54,02E Küchensee
lake, periphyton
1279
AT_114Gel08c Navicula cryptocephala
Kützing
53°13,63N; 08°53,22E Hamme, near Worpswede
river, periphyton
1316
AT_176Gel05 Navicula cryptocephala
Kützing
53°04,08N; 08°29,04E Hasbruch, near hunting lodge
ditch, plankton
1416
AT_212Gel01 Navicula cryptotenella
Lange-Bertalot
54°19,86N; 10°17,72E Dobersdorfer See
lake, benthos
1420
AT_210Gel05 Navicula cryptotenella
Lange-Bertalot
54°09,98N; 10°25,19E Trammer See
lake, periphyton
1435
AT_202Gel03 Navicula cryptotenella
Lange-Bertalot
lake, benthos
1280
AT_117Gel05 Navicula gregaria
Donkin
54°09,86N; 10°32,81E Dieksee
53°09,51N; 08°42,57E Lesum, near river mouth
1436
AT_160Gel09 Navicula hambergii
Hustedt
52°57,65N; 08°20,67E Poggenpohls Moor
soil, moss
1430
AT_177.13
(W. Smith) Ralfs
53°04,08N; 08°29,04E Hasbruch, near hunting lodge
ditch, benthos
1278
AT_114Gel06 Navicula radiosa
Kützing
53°13,63N; 08°53,22E Hamme, near Worpswede
river, periphyton
1433
AT_200.04
Navicula radiosa
Kützing
54°11,69N; 10°36,24E Krumm See
lake, benthos
1440
AT_205.02b
Navicula radiosa
Kützing
54°09,09N; 10°27,45E Großer Madebroken See
lake, benthos
1282
AT_124.15
Navicula reinhardtii
Grunow
53°33,00N; 10°55,16E Schaalsee, Zarrentiner Becken lake, benthos
1411
AT_145.08
Navicula sp.1
Bory
54°06,55N; 10°48,68E Neustädter Binnenwasser
brackish water,
plankton
1319
AT_201Gel01 Navicula sp.2
Bory
54°11,69N; 10°36,24E Krumm See
lake, benthos
1434
AT_202.01
(O. F. Müller) Bory
54°09,86N; 10°32,81E Dieksee
lake, benthos
1276
AT_108Gel01 Navicula veneta
Kützing
53°10,89N; 08°45,70E Hamme, near bridge
river, benthos
1277
AT_110Gel19 Navicula veneta
Kützing
river, benthos
1281
AT_117Gel20b Navicula veneta
Kützing
53°11,39N; 08°47,05E Hamme, near sluice
53°09,51N; 08°42,57E Lesum, near river mouth
river, plankton
1551
AT_177.12
(Ehrenberg) Pfitzer
53°04,08N; 08°29,04E Hasbruch, near hunting lodge
ditch, benthos
Navicula cari
Navicula integra
Navicula tripunctata
Neidum affine
river, plankton
2. Materials and Methods
20
Tab. 1: Continued
DNApreparation
place of origin
culture
1426
AT_161.03
1286
species
Pinnularia acrosphaeria
author
GPS
discription
source
Rabenhorst
52°57,65N; 08°20,67E Poggenpohls Moor
AT_100Gel01 Pinnularia anglica
Krammer
ditch between Plötscher See
53°40,20N; 10°50,21E and Schwarze Kuhle
ditch, periphyton
1314
AT_160Gel30 Pinnularia mesolepta
(Ehrenberg) W. Smith
52°57,65N; 08°20,67E Poggenpohls Moor
soil, moss
1429
AT_161.05
(Ehrenberg) W. Smith
52°57,65N; 08°20,67E Poggenpohls Moor
puddle, soil
1287
AT_105Gel08 Pinnularia microstauron
(Ehrenberg) Cleve
53°09,90N; 08°45,10E Wümme
river, benthos
1288
AT_112Gel04 Pinnularia microstauron
(Ehrenberg) Cleve
53°11,39N; 08°47,05E Hamme, near sluice
river, periphyton
1289
AT_113Gel11 Pinnularia microstauron
(Ehrenberg) Cleve
53°13,63N; 08°53,22E Hamme, near Worpswede
river, plankton
1290
AT_69.06
(Ehrenberg) Cleve
53°13,79N; 08°41,06E Geeste, bridge near Bramel
river, periphyton
1292
AT_70Gel12b Pinnularia obscura
Krasske
53°13,79N; 08°41,06E Geeste, bridge near Bramel
riverside, moss
1311
AT_160Gel10 Pinnularia rupestris
Hantzsch
52°57,65N; 08°20,67E Poggenpohls Moor
soil, moss
Pinnularia mesolepta
Pinnularia microstauron
puddle, soil
1285
AT_100.01
Pinnularia subcapitata
Gregory
ditch between Plötscher See
53°40,20N; 10°50,21E and Schwarze Kuhle
ditch, periphyton
1442
AT_70.09
Pinnularia substreptoraphe
Krammer
53°13,79N; 08°41,06E Geeste, bridge near Bramel
riverside, moss
1291
AT_70.10
Pinnularia viridiformis
Krammer
53°13,79N; 08°41,06E Geeste, bridge near Bramel
riverside, moss
1428
AT_161.02
Pinnularia viridis
(Nitzsch) Ehrenberg
52°57,65N; 08°20,67E Poggenpohls Moor
puddle, soil
1312
AT_160Gel18 Placoneis elginensis
(Gregory) E. J. Cox
52°57,65N; 08°20,67E Poggenpohls Moor
soil, moss
1419
AT_220.09
Mereschkowsky
53°06,41N; 08°11,23E Hunte, near Hundsmühlen
riverside, soil
1412
AT_160Gel11 Stauroneis anceps
Ehrenberg
soil, moss
1294
AT_117Gel17 Stauroneis gracilior
Reichardt
52°57,65N; 08°20,67E Poggenpohls Moor
53°09,51N; 08°42,57E Lesum, near river mouth
river, plankton
1309
AT_70.12
Stauroneis kriegerii
Patrick
53°13,79N; 08°41,06E Geeste, bridge near Bramel
riverside, moss
1444
AT_101.02
Stauroneis kriegerii
Patrick
lake, periphyton
1293
AT_117.04
Stauroneis phoenicenteron
(Nitzsch) Ehrenberg
53°40,20N; 10°50,21E Schwarze Kuhle
53°09,51N; 08°42,57E Lesum, near river mouth
river, plankton
1437
AT_182.07
Stauroneis phoenicenteron
(Nitzsch) Ehrenberg
53°08,06N; 08°53,87E Wümme, Borgfeld
river, plankton
Placoneis sp.
2. Materials and Methods
21
2.2. DNA Methods
2.2.1 DNA isolation
Culture material was concentrated by filtration and quick-frozen in liquid nitrogen. Nucleic
acids were extracted using the Invisorb Spin Plant Mini Kit (Invitek GmbH, Berlin,
Germany). The given protocol was only modified by a duplication of the two washing steps.
2.2.2. PCR
For each culture, the small subunit rRNA coding gene (SSU rDNA), the D1-D2 region of the
large-subunit rRNA gene (LSU rDNA) and the middle part of the ribulose-1,5-bisphosphate
carboxylase/oxygenase large subunit gene (rbcL) were amplified using the polymerase chain
reaction (PCR; Saiki et al., 1988). In the rbcL gene sequence of Rhizosolenia setigera
(GenBank accession number: AF015568) the sequence of the primers F3 and R3 can be found
at the position 292-314 and 1028-1051, respectively. The primers and conditions used for
PCR are shown in the Tables 2 and 3.
Tab. 2: Primers used for PCR
Gene
Primer
Sequence (5' → 3')
SSU
rRNA
1F
AAC CTG GTT GAT CCT GCC AGT
1528R
DIRF
D2CR
F3
R3
TGA TCC TTC TGC AGG TTC ACC TAC
ACC CGC TGA ATT TAA GCA TA
CCT TGG TCC GTG TTT CAA GA
GCT TAC CGT GTA GAT CCA GTT CC
CCT TCT AAT TTA CCA ACA ACT G
LSU
rRNA
rbcL
Author
Medlin et al. (1988),
without polylinker
Medlin et al. (1988),
without polylinker
Scholin et al. (1994)
Scholin et al. (1994)
Beszteri, unpubl.
Beszteri, unpubl.
Tab. 3: Used PCR programs
Cycle step
SSU an LSU rRNA
Temperature Time
Initial
denaturation
94°C
Denaturation
Annealing
Elongation
Cycle repetitions
Final elongation
Cycle
94°C
54°C
72°C
35
72°C
7 min
2 min
4 min
2 min
7 min
rbcL
Temperature
Time
94°C
10 min
Cycle
94°C
56°C
72°C
31
72°C
1 min
1 min
2 min
10 min
2. Materials and Methods
22
The PCR-products were purified by MinEluteTM PCR Purification Kit (QIAGEN, Germany)
according to the manufacturer’s protocol. PCR products with multiple bands were purified by
excising from a 1% agarose gel.
2.2.3. Sequencing
PCR products were sequenced directly on both strands using Big Dye Terminator v3.1
sequencing chemistry (Applied Biosystems, CA, USA). For the LSU rRNA gene and the
rbcL-gene the sequencing reactions were made using the same primers already used in the
PCR. Because of the length of the SSU rRNA gene, additional internal primers (Table 4) were
used. The conditions used for sequencing reaction are shown in table 5. Sequencing products
were purified by DyeExTM Spin Kit (QIAGEN, Germany) and electrophoresed on an ABI
3100 Avant sequencer (Applied Biosystems, CA, USA).
Tab. 4: Additional primers used in the sequencing reactions of the SSU rDNA
Primer
528F
1055F
536R
1055R
Sequence (5' → 3')
GCG GTA ATT CCA GCT CCA A
GGT GGT GCA TGG CCG TTC TT
AAT TAC CGC GGC KGC TGG CA
ACG GCC ATG CAC CAC CAC CCA T
Author
Elwood et al. (1985)
Elwood et al. (1985)
Elwood et al. (1985)
Elwood et al. (1985)
Tab. 5: Used program for the sequencing reaction
Cycle step
Initial
denaturation
Temperature
Time
96°C
1 min
Cycle
Denaturation
Annealing
Elongation
Cycle repetitions
96°C
50°C
60°C
10 sec
5 sec
4 min
25
2.3. Sequence Analysis
Sequences exported from corrected electropherograms were assembled using SeqMan
(Lasergene package, DnaStar, Madison, WI, USA). For the protein-coding rbcL-gene, the
protein-sequence was checked additionally. The alignment of the SSU rDNA sequences was
done with ARB using the secondary structure. The sequences of the D1-D2 region and the
rbcL Gene were aligned using ClustalX (Thompson et al., 1997) and checked manually using
2. Materials and Methods
23
ProSeq v 2.9 beta (Filatov, 2002). The rRNA genes show hypervariable regions for which it is
difficult to obtain an unambiguous alignment. These highly variable sites were excluded from
the alignment.
To get three gene trees with the same set of species an alignment was computed for each gene
using only the sequences of the cultures established within the scope of this study (Table 1).
For each gene a second alignment was made using additional sequences obtained from
GenBank (Table 6).
Tab. 6: List of species of diatoms obtained from GenBank and their accession numbers of the used
gene sequences
Achnanthes
Achnanthes
Achnanthes
Achnanthes
Achnanthes
Achnanthidium
Amphiprora
Amphiprora
Amphora
Amphora
Amphora
Amphora
Amphora
Anomoeoneis
Bacillaria
Campylodiscus
Cocconeis
Cylindrotheca
Cymatopleura
Cymbella
Diadesmis
Dickieia
Encyonema
Encyonema
Entomoneis
Entomoneis
Eolimna
Eolimna
Eunotia
Eunotia
Eunotia
Eunotia
Eunotia
Eunotia
Eunotia
Species
bongranii
brevipes
minutissima
sp.
sp.2
cf. longipes
alata
paludosa
cf. capitellata
cf. proteus
coffeaeformis
montana
sp.
sphaerophora
paxillifer
ralfsii
cf. molesta
closteriva
elliptica
cymbiformis
gallica
ulvacea
cf. sinicum
triangulatum
cf. alata
sp.
minima
subminuscula
minor
bilunaris
cf. pectinalis f. minor
formica var. smatrana
monodon var. asiatica
pectinalis
sp.
SSU rRNA LSU rRNA
AJ535150
AY485476
AJ866992
AY485496
AJ535151
AY485500
AY485497
AY485468
AJ535158
AJ535147
AY485498 AF417682
AJ243061
AB183590
AJ535153
M87325
AF417678
AJ535162
AJ535148
M87326
AJ867030
AJ535156
AJ867023
AY485462
rbcL
AY571754
AJ535157
AJ535160
AF417683
AJ243063
AJ243064
AY571744
AJ866995
AJ535146
AB085830
AB085831
AB085832
AJ535145
2. Materials and Methods
24
Tab. 6: Continued
Species
Fragilariopsis
cylindrus
Gomphonema
capitatum
Gomphonema
parvulum
Gomphonema
pseudaugur
Gyrosigma
limosum
Haslea
crucigera
Haslea
nipkowii
Haslea
ostrearia
Haslea
pseudostrearia
Lyrella
atlantica
Lyrella
hennedyi
Lyrella
sp.
Lyrella
sp.2
Navicula
atomus var. permitis
Navicula
cf. duerrenbergiana
Navicula
cf. erifuga
Navicula
cryptocephala var. veneta
Navicula
diserta
Navicula
lanceolata
Navicula
pelliculosa
Navicula
phyllepta
Navicula
ramosissima
Navicula
salinicola
Navicula
saprophila
Navicula
sclesviscensis
Navicula
sp.
Navicula
sp.2
Navicula
sp.3
Nitzschia
amphibia
Nitzschia
apiculata
Nitzschia
communis
Nitzschia
frustulum
Nitzschia
sigma
Nitzschia
vitrea
Pauliella
taeniata
Peridinium balticum endosymbiont
Peridinium foliaceum endosymbiont
Petroneis
humerosa
Phaeodactylum
tricornutum
Pinnularia
cf. interrupta
Pinnularia
rupestris
Pinnularia
sp.
Placoneis
cf. paraelginensis
Placoneis
constans
Pleurosigma
intermedium
Pleurosigma
planktonicum
Pleurosigma
sp.
SSU rRNA LSU rRNA
AY672802 AF417657
rbcL
AY571751
AJ243062
AB085833
AY485516
AY485482
AY485488
AY485523
AY485524
AJ544659
AY571747
AY571755
AY571756
AJ535149
AJ867024
AY571749
AF417679
AJ297724
AJ535159
AY485484
AY485454
AY485456
AY485512
AY604699
AJ867025
AY485483
AY485513
AY485502
AY485460
AJ867277
M87334
AJ867278 AF417661
AJ535164 AF417671
AJ867279
AJ867280
AY485528 AF417680
Y10566
Y10567
AY571757
AY485459 AF417681
AJ544658
AJ867027
AJ535154
AY571753
AY571752
AY485489
AY485514
AY485515
2. Materials and Methods
25
Tab. 6: Continued
Species
Pleurosigma
sp.2
Pseudogomphonema
cf. kamschaticum
Pseudogomphonema
sp.
Pseudogomphonema
sp.
Rossia
sp.
Sellaphora
bacillum
Sellaphora
laevissima
Sellaphora
pupula
Sellaphora
pupula var. captitata
Seminavis
cf. robusta
Stauroneis
constricta
Surirella
angusta
Surirella
brebissoni
Surirella
fastuosa var. cuneata
uncultured Eunotia-like diatom
Undatella
sp.
SSU rRNA LSU rRNA
AF525664
rbcL
AY571748
AJ535152
AF525663
AJ535144
AY571745
AJ544655
AJ544649
AJ535155
AY571746
AY571750
AY485521
AJ867028
AJ867029
AJ535161
AY821975
AJ535163
Phylogenetic analyses were performed using PAUP* 4.0b10 (Swofford, 1998). In all analyses
the data set was rooted using one centric (Cyclotella choctawatcheea) and two araphid
diatoms (Fragilaria crotonensis and Asterionella formosa), as the use of several outgroup
taxa improves the analyses (Swofford et al., 1996). For maximum likelihood (ML) and
distance based tree calculations, likelihood scores of different nucleotide substitution models
were compared on a neighbor joining tree using Modeltest 3.0 (Posada & Crandall, 1998).
Based on the Akaike Information Criterion (AIC) the best fit model was detected (Table 7).
This was used for phylogenetic analyses using ML and neighbor joining (NJ) tree inference
with ML distances. Maximum parsimony (MP) and ML trees were obtained in heuristic
searches, with 10 random taxa addition sequences. To assess confidence in clades recovered
bootstrapping of MP and NJ analyses was made with 1000 replicates. If necessary, a time
limit of 15 minutes was set for each replicate. The used PAUP command blocks for all
analyses are shown in the appendix.
Tab. 7: Best fit models to perform ML based tree calculations detected by Modeltest based on AIC
(modelblocks are shown in the appendix)
aligned sequences
own cultures
SSU rRNA
GTR +I +G
own cultures and sequences GTR +I +G
obtained from GenBank
gene
LSU rRNA
rbcL
TrN +I +G
GTR +I +G
GTR +I +G
GTR +I +G
combination
GTR +I +G
2. Materials and Methods
26
For weighting the positions in the dataset of the rbcL gene sequences, the entire dataset was
transferred into MacClade (Maddison and Maddison, 1989). In MacClade the third position
was downweighted and the resulting weight block was added to the dataset. Then the entire
weighted dataset was transferred back to PAUP and the phylogenetic analyses were
performed.
For the combined dataset 100 replicates of the partition homogeneity test, as implemented in
PAUP, were performed.
2.4. Microscopy
For identification and morphological investigations of the cultures, light and electron
microscopy were used. Living cells as well as cleaned frustules were examined and
photographed by bright field microscopy using a ZEISS Axioplan microscope with a
AxioCam MRc digital camera. In addition, electron micrographs of cleaned frustules were
taken at 10kV accelerating voltage on a Quanta FEG 200F, a PHILIPS XL30 ESEM or an
I.S.I. DS-130.
2.4.1 Purification of the frustules
To remove all organic material, the cells were oxidized with KMnO4 for 12-16 hours. Then
HCl was added and the mixture boiled until it turned light yellow. The liquid was discarded
and the frustules were washed 4 times with distilled water. The cleaned frustules were stored
in distilled water.
2.4.2. Slide preparation
To prepare permanent slides several drops of cleaned frustule material was placed on a
coverslip and dried on a heating plate at 60°C. Slides for light microscopy were provided with
a drop of a Naphrax/toluene-mixture and the coverslips were placed on this drop. The toluene
was evaporated on a heating plate at 200°C.
For electron microscopy the coverslips were attached to aluminium specimen stubs by
double-sided adhesive tape. The stubs were platinum-coated with a sputter coater (Emscope
SC 500).
3. Results
27
3. Results
3.1 Molecular data
For 89 of the 91 established cultures the SSU rRNA gene, the D1/D2-region of the LSU
rRNA gene and the rbcL gene were sequenced successfully. From Encyonema minutum
(DNA preparation number 1266) and Frustulia vulgaris (DNA preparation number 1445)
only the D1/D2-region of the LSU rRNA gene and the SSU rRNA gene respectively could be
sequenced successfully. Molecular phylogenies were reconstructed on the base of seven
alignments. Four datasets only consists of sequences of the 89 cultures for which all three
genes could be sequenced: One alignment for each gene and one dataset combining these
alignments. For each gene an additional alignment was made comprising the available
sequences from all cultures and sequences obtained from GenBank.
3.1.1 SSU rRNA gene
The SSU rDNA sequences for the sequenced taxa were approximately 1750 nucleotides in
length excluding amplification primers, with the exception of Luticola goeppertiana (DNA
preparation number 1273), which is longer (1904 nucleotides) because of several insertions.
One highly variable region in the SSU rDNA alignment could not be aligned unambiguously.
This segment of 114 nucleotides was excluded from the analyses. It corresponds with the
nucleotides 676 to 790 in the sequence from Luticola goeppertiana. The final dataset had
1827 positions in total, of which 442 were parsimony-informative and 196 parsimonyuninformative characters.
The maximum-likelihood (ML) tree based on the sequences from the AlgaTerra cultures is
shown in Fig. 3. The condensed regions of this figure are shown in detail in Fig. 4.
The three araphid taxa appeared at the base of the ML tree. Asterionella formosa diverged
first, followed by the Fragilaria species. The three Eunotia species formed a monophyletic
group (bootstrap support (BS) based on neighbour-joining (NJ) and parsimony (MP) analysis:
100/97), which diverged next.
Navicula sensu stricto and Hippodonta capitata were sister groups (clade 1) and formed the
basal clade of the raphid pennates. The monophyly of Navicula sensu stricto was supported
by 96% of both bootstrap analyses. The support for Hippodonta being the sister group was
100% in both analyses. Navicula sensu stricto was subdivided in three groups (Fig. 4a). The
3. Results
28
Navicula veneta (1276, 1277, 1281)
Navicula gregaria (1280)
Navicula cryptotenella (1416, 1420, 1435)
96/66
Navicula reinhardtii (1282)
78/76
Navicula sp.2 (1319)
Navicula cryptocephala (1279, 1316)
Navicula sp.1 (1411)
Navicula radiosa (1278, 1433, 1440)
63/61
97/75
Navicula capitatoradiata (1417)
Navicula cari (1310)
Navicula tripunctata (1434)
Hippodonta capitata (1272)
97/98
Cyclotella choctawatcheeana (1493)
Asterionella formosa (1256)
Fragilaria sp. (1410)
Fragilaria araphid
Fragilaria crotonensis (1254)
Eunotia implicata (1321)
100/97
Eunotia formica (1268)
Eunotia
Eunotia sp. (1269)
100/100
96/96
(a)
diatoms
Navicula s.str.
100/100
96/96
98/96
100/100
78/54
100/99
clade 1
Hippodonta capitata (1272)
52/76
42/100/100
99/100
51/55
88/45
Luticola goeppertiana (1273)
Neidum affine (1551)
Stauroneis phoenicenteron (1293, 1437)
Stauroneis anceps (1412)
Stauroneis gracilior (1294)
-/44
Stauroneis kriegerii (1309, 1444)
Navicula integra (1430)
Craticula halophilioides (1283, 1308)
Craticula molestiformis (1284)
Craticula
96/90
Craticula cuspidata (1318, 1320)
Achnanthidium minutissimum (1438)
40/39
50/78
34/48
45/59
(b)
100/92
clade 6
100/100
76/94
80/75
83/84
clade 5
Encyonema caespitosum (1441)
Navicula brockmannii (1425)
Cocconeis pediculus (1415)
100/100
Cocconeis
Cocconeis placentula (1418)
Mayamaea atomus var. atomus (1274)
100/100
Mayamaea
Mayamaea atomus var. permitis (1275)
76/94
Pinnularia/Caloneis
0.05
Gomphonema affine (1322, 1439)
Gomphonema cf. parvulum (1315)
-/Gomphonema truncatum (1552)
100/100
-/Gomphonema acuminatum (1424)
100/100
Gomphonema cf. angustatum (1409)
Gomphonema productum (1313)
Gomphonema micropus (1270, 1271)
100/100
(c)
clade 4
42/54
Eolimna minima (1267)
Amphora sp. (1554)
94/93
53/73
Amphora libyca (1264)
100/100
Amphora cf. fogediana (1427)
Amphora pediculus (1265)
48/48
Amphora normannii (1263)
Amphora
Pinnularia microstauron (1287, 1288, 1289, 1290)
Pinnularia subcapitata (1285)
Pinnularia mesolepta (1314, 1429)
Pinnularia anglica (1286)
Pinnularia obscura (1292)
Pinnularia acrosphaeria (1426)
Caloneis lauta (1446)
Caloneis budensis (1323)
Pinnularia rupestris (1311)
Caloneis amphisbaena (1550)
Pinnularia cf. substreptoraphe (1442)
Pinnularia viridis (1428)
Pinnularia viridiformis (1291)
100/100
Gomphonema
58/79
42/54
0.005
Cymbella
Placoneis elginensis (1312)
Navicula hambergii (1436)
Placoneis sp. (1419)
26/-
clade 2
clade 3
0.01
(d)
0.02
40/39
Fig. 3: Phylogeny inferred with the maximum-likelihood (ML) analysis using SSU rDNA sequences
from the AlgaTerra cultures. Bootstrap values obtained from 1000 replications based on neighborjoining (NJ) analysis using Jukes-Cantor (JC) model and on parsimony analysis have been plotted at
the nodes. Collapsed clades are shown in detail in Fig. 4.
Cymbella affinis (1414, 1423)
Cymbella helmckei (1431)
Cymbella aspera (1421)
Cymbella naviculiformis (1317, 1324)
Cymbella proxima (1422)
100/100
28/44
67/36
0.02
Fig. 4: Details of the ML tree analysis from SSU rDNA sequences from the AlgaTerra cultures. Bootstrap
values obtained from 1000 replications based on NJ analyses using JC model and on parsimony analyses
have been plotted at the nodes. (a) Navicula sensu stricto, (b) Pinnularia and Caloneis, (c) Gomphonema,
(d) Cymbella
3. Results
29
first consisted of N. veneta and N. gregaria (BS: 97/98). N. crytotenella, N. reinhardtii, N.
cryptocephala and the two unidentified Navicula species formed the second group, which was
supported by 98% and 96% of the bootstrap replicates. The third group contained N. radiosa,
N. capitatoradiata, N. cari and the type species N. tripunctata, supported by bootstrap values
of 100 and 99.
Clade 2 in the ML tree (Fig. 3), which comprised Luticola goeppertiana and Neidum affine,
had relatively low bootstrap support (52/76).
The five Amphora species formed a monophyletic clade (BS: 48/48), which diverged next
(clade 3). In this clade, A. normannii is clearly separated from the other Amphora species by
branch length and maximum BS for the monophyly of the other four Amphora species.
Clade 4 in the ML tree includes Eolimna minima, Mayamaea and all Pinnularia and Caloneis
species (BS: 45/59). Eolimna is at the base of this clade and Mayamaea is monophyletic sister
group (BS: 34/48) of Pinnularia and Caloneis. The monophyly of these two genera had
strong MP bootstrap support (94) and medium NJ BS (76). Within Pinnularia/Caloneis clade
three sub-clades could be distinguished (Fig. 4b). One group contained P. acrosphaeria, P.
obscura, P. anglica, P. mesolepta, P. subcapitata and P. microstauron. C. lauta and C.
budensis formed a second group. The third group consisted of P. rupestris, C amphisbaena,
P. viridis, P. cf. substreptoraphe and P. viridiformis.
At the base of clade 5 of the ML tree (Fig. 3) the monoraphid genera Achnanthidium and
Cocconeis and a sub-clade containing Navicula brockmannii and the Cymbellales diverge
from an unresolved polytomy. In this sub-clade N. brockmannii diverges first, followed by
Encyonema caespitosum. Gomphonema (BS: 42/54), Placoneis/Navicula hambergii (BS:
88/45) and Cymbella (BS: 40/39) were monophyletic groups. Gomphonema diverges first and
Placoneis/N. hambergii and Cymbella were sister groups, but this relationship had no BS.
Within the genus Gomphonema (Fig. 4c), G. micropus is clearly separated by the branch
length from the other Gomphonema species, which form a strong group (BS: 100/100). The
genus Cymbella (Fig. 4d) was split into one group containing C. naviculiformis and C.
proxima (BS: 67/36) and another group consisting of C. aspera, C. helmckei and C. affinis
(BS: 28/44).
The clade 6 in the ML tree (Fig. 3) was supported by maximum bootstrap support (100/100).
Within this clade the monophyly of Craticula was supported by 96% and 90% of NJ and MP
bootstrap replicates, respectively. Stauroneis and Navicula integra cluster together, but this
clade had only weak BS (-/44) and the branching order in this group was not fully resolved.
3. Results
30
The maximum parsimony (MP) analysis based on the SSU rDNA sequences resulted in 58
most parsimonious trees. The majority-rule consensus tree of these trees is depicted in the
Figures 5 and 6.
The three araphid taxa appeared at the base as monophyletic group. Bootstrap values of 96
and 86 from NJ and MP bootstrap analysis support this.
Similar to the ML tree the genus Eunotia formed a strongly supported monophyletic group
(BS: 100/97). Luticola goeppertiana and Neidum affine represented the sister group to
Eunotia, although this relationship had nearly no BS (0/29).
The sister groups Navicula sensu stricto and Hippodonta capitata diverged next. The MP
analysis recovered the same three sub-clades in the Navicula sensu stricto as did the ML
analysis. The branching order within the Navicula sensu stricto (Fig. 6a) is similar to the ML
tree (Fig. 4a). But the relationship of N. cryptotenella, the unidentified Navicula species2 and
N. reinhardtii was not resolved.
Following the divergence of Navicula sensu stricto there was a polytomy of three clades.
Clade 1 was the monophyletic Amphora clade (BS: 48/48), in which A. normannii is
separated from the other four Amphora species by maximum BS in the same branching order
as the ML analysis. The other two groups were more divers and differed from the ML
analysis.
Eolimna, Mayamaea, Pinnularia/Caloneis and a clade containing Stauroneis, Craticula and
Navicula integra form clade 2 in the basal polytomy (BS: 23/35). The monophyletic
Pinnularia/Caloneis clade BS (76/94) further diverges into two groups (Fig. 6b). One group
(BS: 0/45) containing C. lauta, P. acrosphaeria, C. amphisbaena, P. obscura, P. anglica, P.
mesolepta, P. subcapitata and P. microstauron. The second group (BS: 55/44) consisted of C.
budensis, P. rupestris, P. viridis, P. cf. substreptoraphe and P. viridiformis. Maximum
bootstrap values (100/100) support the polytomy of Stauroneis, Navicula integra and the well
supported monophyletic Craticula clade (BS: 99/100). The main difference to the three clades
of the ML analysis is that the middle clade of the ML analysis is lost and forms the base of the
two clades in the MP analysis.
In the remaining clade 3 (BS: 50/78) Cocconeis diverged first. Achnanthidium minutissimum
diverged next, followed by Navicula brockmannii and Encyonema caespitosum. A
monophyletic clade containing the two Placoneis species and Navicula hambergii (BS: 88/45)
is the sister group to a clade containing Gomphonema and Cymbella (BS: 14/27). This clade
futher diverges into four branches, which are G. micropus, a clade (BS: 67/36) containing
3. Results
31
96/86
100/100
100/97
-/29
52/76
96/96
100/100
100/100
50/78
67/36
28/44
100/100
96/60
14/27
100/100
100/100
19/29
88/45
58/79
100/100
55/51
100/100
96/90
99/100
100/100
23/35
76/94
Cyclotella choctawatcheeana (1493)
Asterionella formosa (1256)
Fragilaria crotonensis (1254)
araphid
Fragilaria sp. (1410)
Eunotia implicata (1321)
Eunotia formica (1268)
Eunotia
Eunotia sp. (1269)
Luticola goeppertiana (1273)
Neidum affine (1551)
diatoms
(a)
Navicula s. str.
Hippodonta capitata (1272)
Cocconeis placentula (1418)
Cocconeis
Cocconeis pediculus (1415)
Achnanthidium minutissimum (1438)
Navicula brockmannii (1425)
Cymbella proxima (1422)
Cymbella naviculiformis (1317, 1324)
Cymbella aspera (1421)
Cymbella helmckei (1431)
Cymbella affinis (1414, 1423)
Gomphonema productum (1313)
Gomphonema cf. angustatum (1409)
Gomphonema acuminatum (1424)
Gomphonema truncatum (1552)
Gomphonema cf. parvulum (1315)
Gomphonema affine (1322, 1439)
Gomphonema micropus (1270, 1271)
Placoneis sp. (1419)
Placoneis elginensis (1312)
Navicula hambergii (1436)
Encyonema caespitosum (1441)
Stauroneis kriegerii (1309, 1444)
Stauroneis phoenicenteron (1293, 1437)
Stauroneis anceps (1412)
Stauroneis gracilior (1294)
Navicula integra (1430)
Craticula halophilioides (1283, 1308)
Craticula molestiformis (1284)
Craticula
Craticula cuspidata (1318, 1320)
Mayamaea atomus var. atomus (1274)
Mayamaea
Mayamaea atomus var. permitis (1275)
100/100
48/48
Eolimna minima (1267)
Amphora sp. (1554)
Amphora libyca (1264)
Amphora cf. fogediana (1427)
Amphora pediculus (1265)
Amphora normannii (1263)
96/66
96/96
78/76
98/96
100/100
63/61
97/75
100/99
Navicula gregaria (1280)
Navicula veneta (1276, 1277, 1281)
Navicula cryptotenella (1416, 1420, 1435)
Navicula sp.2 (1319)
Navicula reinhardtii (1282)
Navicula cryptocephala (1279, 1316)
Navicula sp.1 (1411)
Navicula radiosa (1278, 1433, 1440)
Navicula capitatoradiata (1417)
Navicula cari (1310)
Navicula tripunctata (1434)
Hippodonta capitata (1272)
clade 3
(b)
100/100
100/92
-/70
56/78
100/100
-/45
97/94
76/94
55/44
93/79
clade 2
80/75
83/84
Pinnularia/Caloneis
94/93
53/73
97/98
Pinnularia microstauron (1287, 1288, 1289, 1290)
Pinnularia subcapitata (1285)
Pinnularia mesolepta (1314, 1429)
Pinnularia anglica (1286)
Pinnularia obscura (1292)
Caloneis amphisbaena (1550)
Pinnularia acrosphaeria (1426)
Caloneis lauta (1446)
Caloneis budensis (1323)
Pinnularia rupestris (1311)
Pinnularia cf. substreptoraphe (1442)
Pinnularia viridis (1428)
Pinnularia viridiformis (1291)
Fig. 6: Details of the parsimony tree analysis using SSU rDNA sequences from the AlgaTerra cultures.
Bootstrap values obtained from 1000 replications based on NJ analyses using JC model and on parsimony
analyses have been plotted at the nodes. (a) Navicula sensu stricto, (b) Pinnularia and Caloneis
Amphora
clade 1
Fig. 5: Majority-rule consensus tree inferred with the parsimony analysis using SSU rDNA
sequences from the AlgaTerra cultures. Bootstrap values obtained from 1000 replications based on
NJ analyses using JC model and on parsimony analyses have been plotted at the nodes. Condensed
regions are shown in detail in Fig. 6.
3. Results
32
C. proxima and C. naviculiformis, a clade (BS: 28/44) consisting of the remaining three
Cymbella species and a strongly supported clade (BS: 100/100) containing the other five
Gomphonema species.
Most of the SSU rDNA sequences obtained from GenBank were similar in length compared
to those sequenced within the scope of this study. But there are several sequences missing up
to 226 nucleotides at the ends (see Table 9 in the appendix). In this extended alignment the
same highly variable region was excluded from the analyses. A MP analysis using this dataset
could not be conducted because of the large number of taxa.
The base of the ML tree inferred from the expanded SSU rDNA dataset (Figs. 7-10) was
similar to the base of the ML tree based on SSU rDNA sequences from the AlgaTerra cultures
(Fig. 3). After a paraphyletic divergence of araphid taxa, the monophyletic Eunotia clade (BS:
71) diverged.
The next branch was formed by a monophyletic clade of four monoraphid species (BS: 50) at
its base, different Bacillariales and the naviculoid Stauroneis constricta. The position of this
naviculoid diatom within this clade is supported by maximum BS, but is likely a contaminant.
The next branch (naviculoid pennates part 1; Figs. 7, 8) diverges at the base into two subclades. Clade 1 (Fig. 8) consists of the well supported monophyletic groups Haslea (BS: 92)
and Pleurosigma (BS: 99) and Gyrosigma limosum, which is sister to Pleurosigma. Clade 2
includes Hippodonta, Navicula sensu stricto and Pseudogomphonema (BS: 100). Hippodonta
capitata was found at the base of this clade, N. diserta diverged next followed by a polytomy
of three clades. The first consists of N. sclesviscensis, N.cryptocephala var. veneta N. veneta,
N. gregaria and two unidentified Navicula species (BS: 55). N. radiosa, N. capitatoradiata,
N. cari, N. tripunctata, N. ramosissima and N. lanceolata formed the second group, which
was supported by 99% of the bootstrap replicates. The third group (BS: 90) diverges into two
clades, which include Pseudogomphonema on one hand and N. crytotenella, N. reinhardtii, N.
cryptocephala, N. phyllepta and three unidentified Navicula species on the other hand.
Neidum affine and Haslea nipkowii diverge next, although this clade had no BS (Fig. 7).
Detail of the next large clade containing naviculoid pennates part 2 is shown Fig. 9. Only
clades at the tip of the tree show strong bootstrap support. The naviculoid pennates part 2
diverge into two clades.
3. Results
33
Cyclotella choctawatcheeana (1493)
Asterionella formosa (1256)
araphid diatoms
100 Fragilaria sp. (1410)
Fragilaria
Fragilaria crotonensis (1254)
Eunotia sp. (1269)
Eunotia implicata (1321)
AJ535146 Eunotia pectinalis f. minor
71
AJ535145 Eunotia sp. 3
AB085832 Eunotia pectinalis
Eunotia formica (1268)
96
AB085830 Eunotia formica var. smatrana
AJ866995 Eunotia bilunaris
AB085831 Eunotia monodon var. asiatica
AY821975 Eunotia sp.2
AJ535163 Undatella sp.
58
66
naviculoid pennates part 1
AY485488 Haslea nipkowii
Neidum affine (1551)
15
naviculoid pennates part 2
23
naviculoid pennates part 3
AJ535150 Achnanthes bongranii
monoraphid
AJ535151 Achnanthes sp.
diatoms
AY485500 Achnanthidium cf. longipes
100
(marine)
AY485476 Achnanthes brevipes
M87325 Bacillaria paxillifer
AJ867280 Nitzschia vitrea
AJ867277 Nitzschia amphibia
93
100
AY672802 Fragilariopsis cylindrus
AY485521 Stauroneis constricta
M87326 Cylindrotheca closterium
Bacillariales
M87334 Nitzschia apiculata
AJ535164 Nitzschia frustulum
AJ867279 Nitzschia sigma
Y10566 Peridinium balticum endosymbiont
AJ867278 Nitzschia communis
Y10567 Peridinium foliaceum endosymbiont
100
50
31
75
94
0.05
Fig. 7: Phylogeny inferred with the ML analysis using SSU rDNA sequences from GenBank and the
AlgaTerra cultures. Bootstrap values obtained from 1000 replications based on NJ analyses (JCmodel) have been plotted at the nodes. Condensed regions are shown in detail in separate figures.
90
92
66
AY485524 Haslea pseudostrearia
AY485523 Haslea ostrearia
Haslea
AY485482 Haslea crucigera
AY485516 Gyrosigma limosum
92
AF525664 Pleurosigma sp.
71
AY485489 Pleurosigma intermedium
99
Pleurosigma
AY485515 Pleurosigma sp.
100
AY485514 Pleurosigma planktonicum
Hippodonta capitata (1272)
AJ535159 Navicula diserta
100 AY485502 Navicula sp.
100
AY485460 Navicula sp.
81
Navicula gregaria (1280)
Navicula veneta (1276)
55
100
50
AJ297724 Navicula cryptocephala var. veneta
AY485483 Navicula sclesviscensis
Navicula radiosa (1278)
87 Navicula capitatoradiata (1417)
Navicula cari (1310)
99
Navicula tripunctata (1434)
67
56 100 AY485512 Navicula ramosissima
AY485484 Navicula lanceolata
100 AJ535152 Pseudogomphonema sp.
AF525663 Pseudogomphonema sp.
AY485513 Navicula sp.
100
90
Navicula cryptocephala (1279)
49
Navicula sp.2 (1319)
Navicula cryptotenella (1416)
Navicula reinhardtii (1282)
AY485456 Navicula phyllepta
44
Navicula sp.1 (1411)
clade 1
clade 2
0.01
Fig. 8: Naviculoid pennates part 1. Detail of the ML tree analysis from SSU rDNA sequences from
GenBank and the AlgaTerra cultures. Bootstrap values obtained from 1000 replications based on NJ
analyses using JC model have been plotted at the nodes.
3. Results
34
AB085833 Gomphonema pseudaugur
AJ243062 Gomphonema parvulum
76
Gomphonema affine (1439)
100
Gomphonema affine (1322)
100
Gomphonema cf. parvulum (1315)
Gomphonema
Gomphonema truncatum (1552)
100
Gomphonema acuminatum (1424)
44
Gomphonema productum (1313)
95
Gomphonema cf. angustatum (1409)
Gomphonema micropus (1270)
Cymbella proxima (1422)
68
Cymbella naviculiformis (1317)
Cymbella aspera (1421)
Cymbella
Cymbella
helmckei (1431)
100
AJ535156 Cymbella cymbiformis
18
100
Cymbella affinis (1414)
Encyonema caespitosum (1441)
Encyonema
99
AJ535157 Encyonema triangulatum
31
Placoneis sp. (1419)
Placoneis elginensis (1312)
84
Navicula hambergii (1436)
AJ535153 Anomoeoneis sphaerophora
28
AJ535149 Lyrella sp.
100
Lyrella
AJ544659 Lyrella atlantica
AY485462 Dickieia ulvacea
Navicula brockmannii (1425)
AY485528 Pauliella taeniata
AJ866992 Achnanthes minutissima
92
monoraphid
Achnanthidium minutissimum (1438)
diatoms
AJ535189 Planothidium lanceolatum
(freshwater)
AJ535148 Cocconeis cf. molesta
Cocconeis
placentula
(1418)
Cocconeis
84
100
Cocconeis pediculus (1415)
Frustulia vulgaris (1445)
AJ867023 Diadesmis gallica
Luticola goeppertiana (1273)
AY485459 Phaeodactylum tricornutum
AJ535147 Amphora cf. proteus
Amphora pediculus (1265)
100
Amphora group 1
100 Amphora cf. fogediana (1427)
Amphora sp. (1554)
53
96 Amphora libyca (1264)
50
28
50
AJ867029 Surirella brebissonii
AJ867028 Surirella angusta
81
AJ867030 Cymatopleura elliptica
AJ535161 Surirella fastuosa var. cunneata
AJ535160 Entomoneis cf. alata
98
AY485468 Amphiprora paludosa
AY485497 Amphiprora alata
AJ535162 Campylodiscus ralfsii
AJ243061 Amphora montana
AB183590 Amphora sp.
AJ535158 Amphora cf. capitellata
Amphora group 2
90
AY485496 Achnanthes sp.
Amphora normannii (1263)
100 AJ867025 Navicula saprophila
AY485454 Navicula pelliculosa
Craticula molestiformis (1284)
Craticula halophilioides (1283, 1308)
Craticula
100
Craticula cuspidata (1318, 1320)
AJ243064 Eolimna subminuscula
86 Stauroneis kriegerii (1309)
Navicula integra (1430)
Stauroneis anceps (1412)
Stauroneis gracilior (1294)
Stauroneis phoenicenteron (1293)
100 Stauroneis phoenicenteron (1437)
AJ535155 Sellaphora pupula var. capitata
98
100
AJ544649 Sellaphora pupula
Sellaphora
AJ544655 Sellaphora laevissima
Eolimna minima (1267, AJ243063)
AY485498 Amphora coffeaeformis
AJ535144 Rossia sp.
100 AJ867024 Navicula atomus var. permitis
100
Mayamaea atomus var. permitis (1275)
Mayamaea
Mayamaea atomus var. atomus (1274)
Caloneis lauta (1446)
Pinnularia microstauron (1290)
100
Pinnularia subcapitata (1285)
100
Pinnularia obscura (1292)
Pinnularia mesolepta (1314)
Pinnularia anglica (1286)
Pinnularia
AJ544658 Pinnularia cf. interrupta
AJ535154 Pinnularia sp.
&
Pinnularia acrosphaeria (1426)
Caloneis
Caloneis budensis (1323)
Caloneis amphisbaena (1550)
AJ867027 Pinnularia rupestris
Pinnularia rupestris (1311)
Pinnularia cf. substreptoraphe (1442)
70
Pinnularia viridis (1428)
80
Pinnularia viridiformis (1291)
100
61
73
100
96
77
clade 2
64
90
61
clade 1
0.02
Fig. 9: Naviculoid pennates part 2. Detail of the ML tree analysis from SSU rDNA sequences from
GenBank and the AlgaTerra cultures. Bootstrap values obtained from 1000 replications based on NJ
analyses using JC model have been plotted at the nodes.
100
clade 2
clade 1
0.05
Fig. 10: Naviculoid pennates part 3. Detail of the ML tree analysis from SSU rDNA sequences from
GenBank and the AlgaTerra cultures. Bootstrap values obtained from 1000 replications based on NJ
analyses using JC model have been plotted at the nodes.
3. Result
35
Clade 1 (Fig 9) consists of Diadesmis gallica as sister to Luticola goeppertiana and
Phaeodactylum tricormutum as sister to a strongly supported (BS: 100) monophyletic group
containing five Amphora species.
Frustulia vulgaris diverge at the base of the clade 2. The next two diverging clades contained
monoraphid species. The first clade consists of Planothidium lanceolatum and a monophyletic
Cocconeis (BS: 84) and the other clade (BS: 50) contained Pauliella taeniata and
Achnanthidium minutissimum. Navicula brockmannii, Dickieia ulvacea and Anomoeoneis
sphaerophora with its sister group Lyrella diverging between the monoraphids and a clade
containing different Cymbellales, but there is no BS for their positions. Within the
Cymbellales, Placoneis and Navicula hambergii diverges first. Following this divergence is a
very rapid divergences of Gomphonema, followed by Encyonema and Cymbella. A
monophyletic clade containing all Gomphonema species has low bootstrap support (BS: 44).
G. micropus branch off at the base of this clade and was separated from the other
Gomphonema species, which are well supported by maximum BS and by branch length. The
genus Encyonema form a well supported monophyletic clade (BS: 99) but the monophyly of
Cymbella has only low BS (28). Cymbella further diverges into sub-clades, one containing C.
proxima and C. naviculiformis and another (BS: 28/44) consisting of the remaining four
Cymbella species.
Navicoloid pennates part 3 is shown in detail in Fig. 10. It diverges into two major clades.
Clade 1 diverges into one branch containing Rossia, Amphora coffeaeformis, Eolimna
minima and the three Sellaphora species and another branch, which includes the
monophyletic Mayamaea clade and a monophyletic clade containing Pinnularia and Caloneis
species. Within this monophyly, C. lauta diverges first, followed by a clade containing P.
acrosphaeria, P. microstauron, P. subcapitata, P. obscura, P. cf. interrupta, P. anglica, P.
mesolepta and one unidentified Pinnularia species. Then C. budensis diverged and a second
larger clade containing P. rupestris, C. amphisbaena, P. viridis, P. cf. substreptoraphe and P.
viridiformis.
Clade 2 also diverges into two clades. At the base of one clade, which had maximum BS,
Navicula saprophila and N. pelliculosa diverges. The branching order of the Stauroneis
species and Navicula integra was not totally resolved. Eolimna subminuscula appeared as
sister to Craticula. The second clade contains several Surirellales and two Amphiprora
species on one hand and a strongly supported (BS: 100) Amphora group with one Achnanthes
in between on the other hand. This Achnanthes sequence is also likely a contaminant.
3. Result
36
The missing BS for several deeper branches within the clades in Figs. 9 and 10 were caused
by one clade in the NJ tree (clade LB in Fig. 59 in the appendix) containing the five species
with the most nucleotide changes (visible as long branches in the ML tree): Caloneis
amphisbaena, Campylodiscus ralfsii, Luticola goeppertiana, Neidum affine and Pinnularia
acrosphaeria. This might be an artefact of long-branch attraction. Therefore these bootstrap
results should be interpreted with caution.
3.1.2 LSU rRNA gene
The LSU rRNA sequences for all sequenced taxa were approximately 540 nucleotides in
length excluding amplification primers, except for Luticola goeppertiana, which was longer
(927 nucleotides) because of several large insertions. One highly variable region that contains
the largest insertion in the sequence from L. goeppertiana was excluded from the analyses.
This region covered 262 nucleotides from L. goeppertiana and approximately 85 nucleotides
from the other taxa. The final datset contained 715 positions, of which 252 were parsimonyinformative and 61 parsimony-uninformative characters.
The maximum-likelihood (ML) tree based on the sequences from the AlgaTerra cultures is
shown in Fig. 11. The collapsed clades of this figure are shown in detail in Fig. 12.
The phylogeny in Fig. 11 diverges at the base into two large clades. Supported by bootstrap
values of 92 and 87 Amphora formed a monophyletic group at the base of clade 1. Within this
group A. normannii was separated from the other Amphora species by maximum BS and
branch length (Fig. 12a). The genus Eunotia diverges next. Hippodonta capitata and Navicula
sensu stricto, interspersed with Neidum affine and Luticola goeppertiana were pooled in the
next clade. H. capitata diverged at the base of this clade. Within this group only the last nodes
were supported by the bootstrap analyses. The araphid diatoms diverge next (BS: 71/73).
Encyonema caespitosum diverges at the base of the next clade, which further diverges into
three monophyletic groups. The first group is formed by the genus Gomphonema (Fig. 12b)
and diverge into G. micropus on one hand and all other Gomphonema species on the other
hand (BS: 99/100). The second group contains Placoneis and Navicula hambergii (BS: 84/42)
and is the sister group to the genus Cymbella (BS: 31/19). This genus diverges into two
groups, but they were not supported by the bootstrap analyses (Fig. 12c).
The polytomy at the base of clade 2 diverges into three branches. One consists of Eolimna
minima only. The second branch was formed by Pinnularia and Caloneis (BS: 0/22). This
branch further diverges into two clades (Fig. 12d), which includes C. budensis, P. rupestris,
P. viridis, C. amphisbaena, P. cf. substreptoraphe, P. acrosphaeria and P. viridiformis
3. Result
37
Cyclotella choctawatcheeana (1493)
81/61
92/87
99/98
71/73
31/19
-/-
84/42
30/16
49/44
96/96
-/19
99/96
49/49
100/100
92/87
diatoms
0.02
Cymbella
Placoneis elginensis (1312)
Navicula hambergii (1436)
Placoneis sp. (1419)
(b)
Gomphonema affine (1322, 1439)
Gomphonema cf. parvulum (1315)
-/57
Gomphonema cf. angustatum (1409)
Gomphonema productum (1313)
Gomphonema truncatum (1552)
100/100
Gomphonema acuminatum (1424)
Gomphonema micropus (1270, 1271)
100/98
33/-
Gomphonema
Encyonema caespitosum (1441)
Navicula
radiosa (1278, 1433, 1440)
74/70
Navicula capitatoradiata (1417)
48/49
100/99 Navicula cari (1310)
45/44
Navicula tripunctata (1434)
Navicula sp.1 (1411)
Navicula reinhardtii (1282)
Neidum affine (1551)
Navicula cryptocephala (1279, 1316)
Navicula sp.2 (1319)
Navicula cryptotenella (1416)
Navicula cryptotenella (1420)
Navicula cryptotenella (1435)
Luticola goeppertiana (1273)
Navicula gregaria (1280)
Navicula veneta (1276, 1277, 1281)
Hippodonta capitata (1272)
Stauroneis phoenicenteron (1293, 1437)
-/39
Stauroneis
anceps (1412)
-/22
Stauroneis gracilior (1294)
74/45
Navicula integra (1430)
Stauroneis kriegerii (1309, 1444)
95/94
Craticula halophilioides (1283, 1308)
Craticula molestiformis (1284)
Craticula
Craticula cuspidata (1318, 1320)
Achnanthidium minutissimum (1438)
Navicula brockmannii (1425)
Cocconeis pediculus (1415)
100/100
Cocconeis
Cocconeis placentula (1418)
Mayamaea atomus var. atomus (1274)
Mayamaea
Mayamaea atomus var. permitis (1275)
Eolimna minima (1267)
-/22
Amphora sp. (1554)
Amphora libyca (1264)
Amphora cf. fogediana (1427)
Amphora pediculus (1265)
Amphora normannii (1263)
98/94
-/-
Eunotia implicata (1321)
Eunotia formica (1268)
Eunotia
Eunotia sp. (1269)
Asterionella formosa (1256)
araphid
100/100 Fragilaria sp. (1410)
Fragilaria crotonensis (1254)
-/13
-/34
(a)
Amphora
99/100
clade 1
49/44
0.02
Navicula
s. str.
(c)
Cymbella affinis (1414, 1423)
Cymbella helmckei (1431)
Cymbella aspera (1421)
Cymbella proxima (1422)
Cymbella naviculiformis (1317, 1324)
99/100
-/31/19
-/-
0.05
(d)
100/83
100/95
-/33
clade 2
Pinnularia/Caloneis
0.05
Fig. 11: Phylogeny inferred with the ML analysis using LSU rDNA sequences from the AlgaTerra
cultures. Bootstrap values obtained from 1000 replications based on NJ analyses using JC model and on
parsimony analyses have been plotted at the nodes. Condensed regions are shown in detail in Fig. 12.
Pinnularia microstauron (1287, 1288, 1289, 1290)
Pinnularia subcapitata (1285)
Pinnularia mesolepta (1314, 1429)
Pinnularia anglica (1286)
Pinnularia obscura (1292)
Caloneis lauta (1446)
Caloneis budensis (1323)
Pinnularia rupestris (1311)
Pinnularia cf. substreptoraphe (1442)
Caloneis amphisbaena (1550)
Pinnularia viridis (1428)
Pinnularia acrosphaeria (1426)
Pinnularia viridiformis (1291)
100/97
-/22
-/21
0.05
Fig. 12: Details of the ML tree analysis based on LSU rDNA sequences from the AlgaTerra cultures.
Bootstrap values obtained from 1000 replications based on NJ analyses using JC model and on
parsimony analyses have been plotted at the nodes. (a) Amphora, (b).Gomphonema, (c) Cymbella, (d)
Pinnularia and Caloneis
3. Result
38
on one hand and C. lauta, P. obscura, P. anglica, P. mesolepta, P. subcapitata and P.
microstauron on the other hand. The third group consist of several genera. Mayamaea
diverges at the base. The next clade consists of Navicula hambergii and Cocconeis and then
Achnanthidium minutissimum diverges. The monophyly of Craticula was supported of 96%
of both bootstrap analyses. Stauroneis and Navicula integra (BS: 74/45) formed the sister
clade (BS: 95/94) to Craticula.
The maximum parsimony (MP) analysis based on the LSU rDNA data sequenced within the
scope of the study resulted in 98 most parsimonious trees. The majority-rule consensus tree of
these trees is depicted in Fig. 13; condensed clades are shown in detail in Fig. 14.
This consensus tree is poorly resolved with a large polytomy at the base. Clades of this
polytomy, which consist of a single genus, were formed by Eolimna minima, Mayamaea (BS:
99/96), Eunotia (BS: 81/61), Amphora (BS: 92/87) and Cocconeis (maximum BS). In clade 1
bootstrap values of 95 and 94 support that Stauroneis and Navicula integra (BS: 74/45)
formed a sister group of the genus Craticula (BS: 96/96). Clade 2, which contains
Achnanthidium minutissimum and Navicula brockmannii, had no bootstrap support.
Pinnularia and Caloneis formed a monophyletic clade (BS: 0/22), which further diverges into
two groups (Fig. 14a). Each group consist of the same species as described for the ML
phylogeny, but the branching order within the clades differs (Figs. 12d and 14a). Also
Hippodonta capitata and Navicula sensu stricto form a monophyletic clade (clade 3, BS:
99/76). The branching order at the base of this clade is not resolved (Fig. 14b) and only the
group containing N. tripunctata, N. cari, N. capitatoradiata and N. radiosa is supported by
high bootstrap values (100/99). Clade 4, which contains Neidum affine and Luticola
goeppertiana, was supported only by 35% of the MP bootstrap replicates. Bootstrap values of
71 and 73 support the clade containing the araphid diatoms. The most diverse clade 5 consists
of Encyonema caespitosum, all species belonging to Cymbella, Gomphonema and Placoneis
and Navicula hambergii. From the polytomy at the base of this clade only two groups diverge.
One contains C. helmckei and C. affinis (BS: 99/100), the other consists of the genus
Gomphonema, whose monophyly is supported by bootstrap values os 49 and 44.
The LSU rDNA sequences obtained from GenBank were similar in length compared to the
sequences, which were sequenced within the scope of this study. In this extended alignment
the same highly variable region was excluded from the analyses. The calculation of some
3. Result
39
99/96
-/39
-/22
74/45
95/94
96/96
-/34
-/22
Cyclotella choctawatcheeana (1493)
Eolimna minima (1267)
Mayamaea atomus var. permitis (1275)
Mayamaea atomus var. atomus (1274)
Stauroneis phoenicenteron (1293, 1437)
Stauroneis anceps (1412)
Stauroneis gracilior (1294)
Navicula integra (1430)
Stauroneis kriegerii (1309, 1444)
Craticula molestiformis (1284)
Craticula halophilioides (1283, 1308)
Craticula cuspidata (1318, 1320)
Achnanthidium minutissimum (1438)
Navicula brockmannii (1425)
Mayamaea
100/83
100/95
-/33
clade 1
99/98
92/87
100/100
40/62
98/94
100/100
99/76
83/74
-/21
Craticula
Eunotia implicata (1321)
Eunotia formica (1268)
Eunotia
Eunotia sp. (1269)
Amphora normannii (1263)
Amphora cf. fogediana (1427)
Amphora pediculus (1265)
Amphora
Amphora libyca (1264)
Amphora sp. (1554)
Cocconeis placentula (1418)
Cocconeis
Cocconeis pediculus (1415)
Hippodonta capitata (1272)
Navicula s. str.
-/35
71/73
71/73
100/100
100/99
100/100
33/57
-/22
70/34
59/33
clade 2
30/29
76/53
Pinnularia/Caloneis
81/61
81/61
49/44
49/49
100/98
99/100
Pinnularia microstauron (1287, 1288, 1289, 1290)
Pinnularia subcapitata (1285)
Pinnularia mesolepta (1314, 1429)
Pinnularia anglica (1286)
Pinnularia obscura (1292)
Caloneis lauta (1446)
Caloneis amphisbaena (1550)
Pinnularia acrosphaeria (1426)
Caloneis budensis (1323)
Pinnularia cf. substreptoraphe (1442)
Pinnularia rupestris (1311)
Pinnularia viridis (1428)
Pinnularia viridiformis (1291)
100/97
(a)
Luticola goeppertiana (1273)
Neidum affine (1551)
Asterionella formosa (1256)
Fragilaria crotonensis (1254)
araphid diatoms
Fragilaria sp. (1410)
Placoneis sp. (1419)
Navicula hambergii (1436)
Placoneis elginensis (1312)
Gomphonema acuminatum (1424)
Gomphonema truncatum (1552)
Gomphonema productum (1313)
Gomphonema cf. angustatum (1409)
Gomphonema
Gomphonema cf. parvulum (1315)
Gomphonema affine (1322, 1439)
Gomphonema micropus (1270, 1271)
Cymbella naviculiformis (1317, 1324)
Cymbella aspera (1421)
Cymbella proxima (1422)
Cymbella helmckei (1431)
Cymbella affinis (1414, 1423)
Encyonema caespitosum (1441)
(b)
77/55
32/20
58/59
clade 3
99/76
-/21
74/70
clade 4
45/44
100/99
48/49
Navicula gregaria (1280)
Navicula veneta (1276, 1277, 1281)
Navicula cryptotenella (1416, 1420, 1435)
Navicula sp.2 (1319)
Navicula reinhardtii (1282)
Navicula cryptocephala (1279, 1316)
Navicula sp.1 (1411)
Navicula radiosa (1278, 1433, 1440)
Navicula capitatoradiata (1417)
Navicula cari (1310)
Navicula tripunctata (1434)
Hippodonta capitata (1272)
Fig. 14: Details of the parsimony tree analysis based on LSU rDNA sequences from the AlgaTerra
cultures. Bootstrap values obtained from 1000 replications based on NJ analyses using JC model and on
parsimony analyses have been plotted at the nodes. (a) Navicula sensu stricto, (b) Pinnularia and
Caloneis
clade 5
Fig. 13: Majority-rule consensus tree inferred with the parsimony analysis based on LSU rDNA
sequences from the AlgaTerra cultures. Bootstrap values obtained from 1000 replications based on NJ
analyses using JC model and on parsimony analyses have been plotted at the nodes. Condensed regions
are shown in detail in Fig. 14.
3. Result
40
replicates in the MP bootstrap analysis needed plenty of time. Therefore a time limit of 15
minutes was set for each replicate. As a result of this time limit 89 of 1000 replicates were
terminated. For these replicates it is not certain if the best tree was detected.
The tree resulted from the ML analysis based on the extended is shown in Fig. 15. The
collapsed clades are only shown in detail (Fig. 16) if they differ from the equivalent clades
shown in Figs. 11 and 12.
In the ML tree clade 1, which contains the monoraphids species, Navicula brockmannii,
Craticula, Stauroneis and N. integra, diverges first. The branching within this clade was
similar to the equivalent clade in Fig. 11. The additional species, Pauliella taeniata, clusters
with Achnanthidium minutissimum.
Mayamaea diverges next, followed by a polytomy of Eolimna minima, a clade containing
Pinnularia and Caloneis and clade 2. Although there were no additional Pinnularia or
Caloneis species in this dataset, the branching order of P. viridis and P. rupestris had changed
(Fig. 16a) compared to the equivalent clade shown in Fig. 12d.
Clade 2 diverges further into two sub-clades. At the base of clade 2a the genus Amphora was
separated by the Entomoneis species into two groups (Fig. 16b), which consists of A.
normannii and A. coffeaeformis on one hand (BS: 99/98) and the remaining Amphora species
on the other hand (BS: 100/100). Phaeodactylum tricornutum diverges next (Fig. 16). Within
the Cymbellales, Encyonema (BS: 99/92) diverges first. The next group contains Placoneis
and N. hambergii. In this tree, the genera Gomphonema and Cymbella were sister groups.
The genus Eunotia (BS: 84/73) diverges at the base of clade 2b, followed by the araphid
diatoms (BS: 69/77). Within these two clades there were no changes compared to the tree in
Fig. 11. The next divergence is a polytomy, from which Luticola goeppertiana, a clade
containing the Bacillariaceae and a clade consisting of Hippodonta capitata and Navicula
sensu stricto interspersed with Neidum affine.
The majority-rule consensus tree (Fig. 17) from the extended alignment based on 242 most
parsimonious trees. The collapsed clades are only shown in detail (Fig. 18) if they differ from
the equivalent clades shown in Figs. 13 and 14.
3. Result
41
(a)
Cyclotella choctawatcheeana (1493)
100/100
Navicula brockmannii (1425)
AF417680 Pauliella taeniata
Achnanthidium minutissimum (1438)
98/95
66/30
22/10
98/91
Pinnularia microstauron (1287, 1288, 1289, 1290)
Pinnularia subcapitata (1285)
Pinnularia mesolepta (1314, 1429)
Pinnularia anglica (1286)
Pinnularia obscura (1292)
Caloneis lauta (1446)
Caloneis budensis (1323)
Pinnularia cf. substreptoraphe (1442)
Pinnularia rupestris (1311)
Pinnularia acrosphaeria (1426)
Pinnularia viridiformis (1291)
Pinnularia viridis (1428)
Caloneis amphisbaena (1550)
100/96
Cocconeis
100/97
Craticula
-/17
clade 1
Stauroneis kriegerii (1309, 1444)
Navicula
integra (1430)
83/48
Stauroneis anceps (1412)
-/25
Stauroneis gracilior (1294)
Stauroneis phoenicenteron (1293, 1437)
Mayamaea
Eolimna minima (1267)
100/98
-/17
Pinnularia/Caloneis
99/89
47/39
Amphora (group 1)
AF417683 Entomoneis sp.
100/100
-/3
Amphora (group 2)
(b)
AF417681 Phaeodactylum tricornutum
34/32
-/16
57/54
86/36
52/48 46/30
99/92
84/73
69/77
37/44
0,05
99/89
Cymbella
Placoneis elginensis (1312)
Navicula hambergii (1436)
Placoneis sp. (1419)
Encyonema caespitosum (1441)
Encyonema minutum (1266)
47/39
2a
Gomphonema
AF417682 Amphora coffeaeformis
Amphora (group 1)
Amphora normannii (1263)
AF417683 Entomoneis sp.
Amphora pediculus (1265)
100/100
Amphora cf. fogediana (1427)
Amphora (group 2)
Amphora libyca (1264)
-/50
97/95
Amphora sp. (1554)
Encyonema
Eunotia
araphid diatoms
Hippodonta capitata (1272)
Navicula veneta (1276, 1277, 1281)
Navicula gregaria (1280)
Navicula cryptotenella (1435)
Navicula cryptotenella (1420)
Navicula cryptotenella (1416)
Navicula sp.2 (1319)
Navicula
Navicula cryptocephala (1279, 1316)
Neidum affine (1551) s. str.
Navicula reinhardtii (1282)
Navicula sp.1 (1411)
AF417679 Navicula cf. erifuga
40/50
Navicula tripunctata (1434)
99/99
Navicula cari (1310)
Navicula capitatoradiata (1417)
77/72
Navicula radiosa (1278, 1433, 1440)
Luticola goeppertiana (1273)
AF417661 Nitzschia communis
AF417678 Bacillaria paxillifer
68/44
50/AF417671 Nitzschia frustulum
96/86
AF417657 Fragilariopsis cylindrus
0.05
clade 2
2b
0.05
Fig. 15: Phylogeny inferred with the ML analysis using LSU rDNA sequences from GenBank and
AlgaTerra cultures. Bootstrap values obtained from 1000 replications based on NJ analyses using JC
model and on parsimony analyses have been plotted at the nodes. Condensed regions, which differ to
the equivalent clades in Figs. 11 and 12, are shown in detail in Fig. 16.
Fig. 16: Details of the ML tree analysis from LSU rDNA sequences from GenBank and AlgaTerra
cultures. Bootstrap values obtained from 1000 replications based on NJ analyses using JC model and on
parsimony analyses have been plotted at the nodes. (a) Pinnularia and Caloneis, (b) Amphora
3. Result
42
(a)
Cyclotella choctawatcheeana (1493)
Eolimna minima (1267)
100/98
Mayamaea
99/99
Navicula brockmannii (1425)
84/73
2/1
96/86
68/44
-/57
69/77
100/100
5/5
31/20
99/84
53/49
AF417657 Fragilariopsis cylindrus
AF417671 Nitzschia frustulum
AF417678 Bacillaria paxillifer
AF417661 Nitzschia communis
75/52
1a
57/54
49/20
-/16
99/92
52/48
48/50/40
34/32
-/42
-/25
83/48
100/100
98/91
98/95
100/96
99/85
100/97
55/46
81/72
48/20
-/29
74/59
-/33
66/30
59/60
araphid diatoms
Cocconeis
Navicula s.str.
75/45
99/84
Hippodonta capitata (1272)
Amphora
clade 1
AF417681 Phaeodactylum tricornutum
75/45
77/72
40/50
Eunotia
Gomphonema
Placoneis sp. (1419)
Navicula hambergii (1436)
Placoneis elginensis (1312)
Encyonema minutum (1266)
Encyonema
Encyonema caespitosum (1441)
Cymbella proxima (1422)
Cymbella aspera (1421)
Cymbella helmckei (1431)
Cymbella
Cymbella affinis (1414, 1423)
Cymbella naviculiformis (1317, 1324)
Stauroneis phoenicenteron (1293, 1437)
Stauroneis anceps (1412)
Navicula integra (1430)
Stauroneis gracilior (1294)
Stauroneis kriegerii (1309, 1444)
(b)
99/89
53/49
100/100
1b
-/50
97/95
AF417682 Amphora coffeaeformis
Amphora normannii (1263)
Amphora pediculus (1265)
Amphora cf. fogediana (1427)
Amphora libyca (1264)
Amphora sp. (1554)
Fig. 18: Details of the parsimony tree analysis based on LSU rDNA sequences from GenBank and
AlgaTerra cultures. Bootstrap values obtained from 1000 replications based on NJ analyses using JC
model and on parsimony analyses have been plotted at the nodes. (a) Navicula sensu stricto, (b)
Amphora
2a
Craticula
Pinnularia microstauron (1287, 1288, 1289, 1290)
Pinnularia subcapitata (1285)
Pinnularia mesolepta (1314, 1429)
Pinnularia anglica (1286)
Pinnularia obscura (1292)
Caloneis amphisbaena (1550)
Pinnularia acrosphaeria (1426)
Caloneis budensis (1323)
Pinnularia cf. substreptoraphe (1442)
Pinnularia rupestris (1311)
Pinnularia viridiformis (1291)
Pinnularia viridis (1428)
Caloneis lauta (1446)
AF417683 Entomoneis sp.
Neidum affine (1551)
Luticola goeppertiana (1273)
Achnanthidium minutissimum (1438)
AF417680 Pauliella taeniata
Navicula cari (1310)
Navicula tripunctata (1434)
AF417679 Navicula cf erifuga
Navicula radiosa (1278, 1433, 1440)
Navicula capitatoradiata (1417)
Navicula sp.1 (1411)
Navicula cryptotenella (1416, 1420, 1435)
Navicula sp.2 (1319)
Navicula reinhardtii (1282)
Navicula cryptocephala (1279, 1316)
Navicula gregaria (1280)
Navicula veneta (1276, 1277, 1281)
Hippodonta capitata (1272)
clade 2
2b
Fig. 17: Majority-rule consensus tree inferred with the parsimony analysis based on LSU rDNA
sequences from GenBank and AlgaTerra cultures. Bootstrap values obtained from 1000 replications
based on NJ analyses using JC model and on parsimony analyses have been plotted at the nodes.
Condensed clades, which differ to the equivalent clades in Figs. 13 and 14, are shown in detail in Fig.
18.
3. Result
43
Eolimna minima and Mayamaea were found at the base of the consensus tree. At the next
lineage the tree diverges into two main clades. Both of these clades further diverge into two
sub-clades.
Navicula brockmannii diverges at the base of clade 1a. Then Eunotia diverges (BS: 84/73).
The next branch consists of two monophyletic groups, the araphid diatoms on one hand (BS:
69/77) and the Bacillariaceae on the other (BS: 68/44). Cocconeis was the sister group (BS:
31/20) to a strongly supported clade (BS: 99/84) containing Hippodonta capitata and
Navicula sensu stricto (BS: 75/45). Navicula sensu stricto did not diverge into separate
groups (Fig. 18a).
The genus Amphora diverges at the base of clade 1b. Within this genus two groups could be
distinguished (Fig. 18b). One consists of A. normannii and A. coffeaeformis (BS: 99/89) and
the other group contains the remaining Amphora species (BS: 100/100). After the divergence
of Phaeodactylum tricornutum the clade diverges into Gomphonema and Placoneis/N.
hambergii on one hand (BS: 49/20) and Encyonema and Cymbella on the other hand (no BS).
Placoneis/N. hambergii were paraphyletic whereas the other three genera formed
monophyletic groups.
Clade 2 consists of the strongly supported clade 2a (BS: 98/91) containing Craticula (BS:
98/95) and Stauroneis/N. integra (BS: 83/48) and a clade 2b containing a mixture of
monoraphid and raphid taxa. The branching order within the Stauroneis/N. integra is not
totally resolved. In clade 2b the monoraphid Pauliella taeniata and Achnanthidium
minutissimum diverge first. The next clade contains Caloneis lauta, the Entomoneis species,
Neidum affine and Luticola goeppertiana. This is followed by a clade of Pinnularia and
Caloneis species. Therefore Pinnularia and Caloneis were not a monophyletic group in this
tree.
3.1.3 rbcL gene
The rbcL gene sequences for most sequenced taxa are 684 nucleotides in length excluding
amplification primers. 15 sequences missing between 3 and 51 nucleotides because of
sequencing problems in the regions close to the primer (Table 8). In none of the sequences
were insertions or deletions. This permitted an unambiguous alignment. The final dataset had
684 positions in total, of which 210 were parsimony-informative and 49 parsimonyuninformative characters.
3. Result
44
Tab. 8: Number of unknown nucleotides in incompletely sequenced rbcL sequences
Species
Achnanthidium minutissimum (1438)
Amphora sp. (1554)
Caloneis amphisbeana (1550)
Cocconeis pediculus (1415)
Craticula molestiformis (1284)
Cymbella helmckei (1431)
Eunotia sp. (1269)
Gomphonema affine (1439)
Gomphonema productum (1409)
Navicula cryptotenella (1416)
Navicula cryptotenella (1420)
Navicula radiosa (1433)
Pinnularia rupestris (1311)
Stauroneis kriegerii (1444)
Stauroneis phoenicenteron (1293)
unknown nucleotides close to
primer F3
primer R3
3
27
3
9
21
30
3
17
6
27
24
18
9
12
6
19
39
7
15
21
24
21
The genus Eunotia (BS: 99/95) formed the base of the tree inferred with the ML analysis
(Fig.19).
Clade 1 consists of Navicula sensu stricto and Hippodonta capitata (BS: 88/78). The
monophyly of Navicula sensu stricto was supported by only 51% and 54% of the bootstrap
replicates. Navicula sensu stricto was subdivided in three groups (Fig. 20a). The first consists
of N. veneta, N. gregaria and one unidentified Navicula species (BS: 29/0). N. crytotenella,
N. reinhardtii, N. cryptocephala and the other unidentified Navicula species formed the
second group which was supported by 86% and 80% of the bootstrap replicates. The third
group contained N. radiosa, N. capitatoradiata, N. cari and N. tripunctata (BS: 83/89).
The araphid diatoms formed a monophyletic clade (clade 2, BS: 85/61), which diverges next.
Clade 3 contains all Pinnularia and Caloneis species, Eolimna minima and Mayamaea, but
this clade had nearly no bootstrap support. Eolimna and Mayamaea formed a monophyletic
sister group to Pinnularia and Caloneis. C amphisbaena and C. budensis diverged at the base
of the clade formed by Pinnularia and Caloneis (Fig. 20b). The other species were subdivided
into two clades. One group contained P. acrosphaeria, P. subcapitata, P. microstauron, P.
mesolepta, P. anglica and P. obscura. The other group consists of C. lauta, P. rupestris, P.
viridis, P. cf. substreptoraphe and P. viridiformis.
Navicula brockmannii diverges at the base of clade 4. At the next lineage the tree diverges
into a sub-clade, which consists of Encyonema caespitosum and the monoraphid species and a
sub-clade containing the remaining Cymbellales. The genus Cymbella did not form a
3. Result
45
(a)
Cyclotella choctawatcheeana (1493)
Eunotia formica (1268)
Eunotia sp. (1269)
Eunotia
Eunotia implicata (1321)
73/89
99/95
51/54
51/54
Navicula s.str.
clade 1
88/78
Hippodonta capitata (1272)
Asterionella formosa (1256)
Fragilaria crotonensis (1254)
100/100
Fragilaria sp. (1410)
85/61
Navicula gregaria (1280)
Navicula sp.1 (1411)
Navicula veneta (1276, 1277, 1281)
Navicula reinhardtii (1282)
86/80
Navicula cryptocephala (1279, 1316)
Navicula cryptotenella (1416, 1420, 1435)
55/43
100/100
Navicula sp.2 (1319)
Navicula
radiosa
(1278, 1433, 1440)
94/92
Navicula capitatoradiata (1417)
83/89
Navicula cari (1310)
69/55
Navicula tripunctata (1434)
Hippodonta capitata (1272)
39/29/-
88/78
araphid diatoms
clade 2
Pinnularia/Caloneis
clade 3
-/8
Mayamaea atomus var. permitis (1275)
Mayamaea
Mayamaea atomus var. atomus (1274)
Eolimna minima (1267)
Navicula brockmannii (1425)
Cocconeis placentula (1418)
100/100
Cocconeis
Cocconeis pediculus (1415)
-/40
Achnanthidium minutissimum (1438)
Encyonema caespitosum (1441)
Cymbella helmckei (1431)
Cymbella affinis (1414, 1423)
Cymbella aspera (1421)
Cymbella naviculiformis (1317, 1324)
Cymbella proxima (1422)
Navicula hambergii (1436)
Placoneis sp. (1419)
Placoneis elginensis (1312)
0.01
100/100
-/8
98/88
50/39
92/75
25/-
22/11
48/35
40/16
(b)
94/81
60/38
clade 4
58/38
53/43
98/96
Gomphonema
Stauroneis phoenicenteron (1293, 1437)
Stauroneis anceps (1412)
73/14
Stauroneis gracilior (1294)
Stauroneis kriegerii (1309, 1444)
33/23
Navicula integra (1430)
Craticula cuspidata (1318, 1320)
-/12
Craticula molestiformis (1284)
65/59
Craticula halophilioides (1283, 1308)
Amphora normannii (1263)
Luticola goeppertiana (1273)
Neidum affine (1551)
Amphora sp. (1554)
Amphora libyca (1264)
100/100
-/59
Amphora cf. fogediana (1427)
98/96 Amphora pediculus (1265)
Pinnularia acrosphaeria (1426)
Pinnularia subcapitata (1285)
Pinnularia microstauron (1287, 1288, 1289, 1290)
Pinnularia mesolepta (1314, 1429)
87/93
Pinnularia anglica (1286)
Pinnularia obscura (1292)
Caloneis lauta (1446)
Pinnularia rupestris (1311)
Pinnularia cf. substreptoraphe (1442)
Pinnularia viridis (1428)
Pinnularia viridiformis (1291)
Caloneis budensis (1323)
Caloneis amphisbaena (1550)
0.01
97/81
-/37
19/24
9/6
38/31
Stauroneis
(c)
40/16
clade 5
0.02
Fig.19: Phylogeny inferred with the ML analysis using rbcL sequences from the AlgaTerra cultures.
Bootstrap values obtained from 1000 replications based on NJ analyses using JC model and on
parsimony analyses have been plotted at the nodes. Collapsed clades are shown in detail in Fig. 20.
78/63
Gomphonema cf. parvulum (1315)
Gomphonema cf. angustatum (1409)
Gomphonema productum (1313)
Gomphonema micropus (1270, 1271)
Gomphonema affine (1322, 1439)
100/100
Gomphonema acuminatum (1424)
Gomphonema truncatum (1552)
0.005
Fig. 20: Details of the ML tree analysis from rbcL sequences from the AlgaTerra cultures. Bootstrap
values obtained from 1000 replications based on NJ analyses using JC model and on parsimony analyses
have been plotted at the nodes. (a) Navicula sensu stricto, (b) Pinnularia and Caloneis, (c) Gomphonema
3. Result
46
monophyletic group. C. helmckei, C. affinis, C. aspera and C. naviculiformis formed a clade
(BS: 50/39), but C. proxima clusters with Navicula hambergii and Placoneis (BS: 22/11).
Within the monophyletic clade formed by the genus Gomphonema (BS: 40/16), the branching
order calculated by the ML analyses had nearly no bootstrap support (Fig. 20c).
Four Amphora species formed a strongly supported monophyletic clade (BS: 100/100) at the
base of clade 5. Amphora normannii was separated from the other Amphora species by a
branch consisting of Luticola goeppertiana and Neidum affine. At the next divergence the
monophyletic Stauroneis clade (BS: 73/14) was separated from a clade containing Craticula
and Navicula integra (BS: 19/24). Craticula was paraphyletic.
The maximum parsimony (MP) analysis based on the rbcL sequences sequenced within the
scope of the study resulted in 2 most parsimonious trees.
The genus Eunotia (BS: 99/95) diverges at the base of the majority-rule consensus tree (Fig.
21). Then a large polytomy with 34 branches followed. 22 of these branches leaded to single
species. Four Amphora species formed a clade, which had maximum bootstrap support.
Within the genus Gomphonema only G. acuminatum and G. truncatum grouped together
(maximum BS). Cymbella affinis and C. helmckei (BS: 98/88) formed a group as well as C.
aspera and C. naviculiformis (BS: 92/75). The Pinnularia species formed two clades, which
consists of P. acrosphaeria, P. obscura, P. anglica, P. mesolepta, P. subcapitata and P.
microstauron one hand (BS: 94/81) and P. rupestris, P. viridis, P. cf. substreptoraphe and P.
viridiformis on the other hand (BS: 91/54). Other small clades consists of Craticula
halophilioides and C. molestiformis (BS: 65/59), Stauroneis anceps and St. phoenicenteron
(BS: 97/81), Mayamaea atomus var. atomus and M. atomus var. permitis (maximum BS), the
araphid species (BS: 85/61) and Cocconeis placentula and C. pediculus (maximum BS). The
largest clade contains Hippodonta capitata and Navicula sensu stricto. The branching order
within Navicula sensu stricto was not totally resolved, but two groups could be distinguished.
One group contained N. capitatoradiata, N. radiosa, N. cari and N. tripunctata (BS: 83/89)
thither group consists of N. reinhardtii, N. cryptocephala, N. crytotenella and an unidentified
Navicula species (BS: 86/80).
The rbcL gene sequences obtained from GenBank were longer, than those sequenced within
the scope of this study. They were all cut to a length of 684 nucleotides.
3. Result
47
Cyclotella choctawatcheeana (1493)
99/95
73/89
100/100
-/59
98/96
63/78
100/100
98/88
92/75
70/75
94/81
87/93
90/61
92/54
96/98
65/59
97/81
100/100
85/61
100/100
88/78
94/92
83/89
51/54
69/55
86/80
100/100
100/100
Eunotia implicata (1321)
Eunotia sp. (1269)
Eunotia formica (1268)
Gomphonema cf. parvulum (1315)
Placoneis elginensis (1312)
Placoneis sp. (1419)
Navicula hambergii (1436)
Cymbella proxima (1422)
Encyonema caespitosum (1441)
Navicula brockmannii (1425)
Eolimna minima (1267)
Caloneis amphisbaena (1550)
Amphora normannii (1263)
Caloneis lauta (1446)
Caloneis budensis (1323)
Luticola goeppertiana (1273)
Neidum affine (1551)
Stauroneis gracilior (1294)
Navicula integra (1430)
Achnanthidium minutissimum (1438)
Amphora sp. (1554)
Amphora libyca (1264)
Amphora pediculus (1265)
Amphora cf. fogediana (1427)
Gomphonema micropus (1270, 1271)
Gomphonema productum (1313)
Gomphonema cf. parvulum (1409)
Gomphonema affine (1322, 1439)
Gomphonema acuminatum (1424)
Gomphonema truncatum (1552)
Cymbella helmckei (1431)
Cymbella affinis (1414, 1423)
Cymbella aspera (1421)
Cymbella naviculiformis (1317, 1324)
Pinnularia acrosphaeria (1426)
Pinnularia subcapitata (1285)
Pinnularia microstauron (1287, 1288, 1289, 1290)
Pinnularia obscura (1292)
Pinnularia anglica (1286)
Pinnularia mesolepta (1314, 1429)
Pinnularia rupestris (1311)
Pinnularia cf. substreptoraphe (1442)
Pinnularia viridiformis (1291)
Pinnularia viridis (1428)
Craticula halophilioides (1283, 1308)
Craticula molestiformis (1284)
Stauroneis anceps (1412)
Stauroneis phoenicenteron (1293, 1437)
Stauroneis kriegerii (1309, 1444)
Craticula cuspidata (1318,1320)
Mayamaea atomus var. atomus (1274)
Mayamaea atomus var. permitis (1275)
Asterionella formosa (1256)
Fragilaria sp. (1410)
Fragilaria crotonensis (1254)
Hippodonta capitata (1272)
Navicula gregaria (1280)
Navicula sp.1 (1411)
Navicula veneta (1276, 1277, 1281)
Navicula capitatoradiata (1417)
Navicula radiosa (1278, 1433, 1440)
Navicula cari (1310)
Navicula tripunctata (1434)
Navicula reinhardtii (1282)
Navicula cryptocephala (1279, 1316)
Navicula sp.2 (1319)
Navicula cryptotenella (1416, 1420, 1435)
Cocconeis placentula (1418)
Cocconeis pediculus (1415)
Fig. 21: Majority-rule consensus tree inferred with the parsimony analysis based on rbcL
sequences from the AlgaTerra cultures. Bootstrap values obtained from 1000 replications based
on NJ analyses using JC model and on parsimony analyses have been plotted at the nodes.
3. Result
48
The phylogeny inferred with the ML analysis using the extended alignment is shown in figure
22. The genus Eunotia, which forms a well supported monophyletic group (BS: 99/96),
diverges at the base of this tree (Figs. 22 and 23a).
In this tree Hippodonta capitata and Navicula sensu stricto did not form a monophyletic
group. H. capitata and N. salinicola diverges first. Then clade 1 with Pseudogomphonema cf.
kamschaticum and Seminavis cf. robusta at the base and a monophyly of the remaining
Navicula sensu stricto species diverge. Within this Navicula sensu stricto two groups could be
distinguished.
The next two clades consist of Petroneis humerosa and Lyrella (clade 2, BS: 96/93) and the
monoraphid species (clade 3, BS: 0/46).
At the next lineage the tree diverges into two clades. At the base of clade 4 Navicula
brockmannii diverges. The next clade consists of the genus Gomphonema (BS: 40/16), which
further diverges into two groups (Fig. 23b). Encyonema cf. sinicum and E. caespitosum
formed a monophyletic group which diverges next. The genus Cymbella was paraphyletic.
Navicula hambergii and Placoneis formed a monophyletic group (BS: 33/24) which cluster
with C. proxima (BS: 16/9). The remaining Cymbella species formed the sister group of this
clade.
The araphid diatoms diverged at the base of clade 5 (BS: 84/67). Maximum bootstrap values
support the monophyly of four Amphora species (group 1). After the divergence of Luticola
goeppertiana and Neidum affine (BS: 36/30) the tree diverges into two groups. Amphora
normannii was at the base of the first group. Stauroneis forms a monophyletic clade, which
separate Craticula halophilioides and C. molestiformis from C. cuspidata and Navicula
integra. In the second group Mayamaea, Sellaphora and Eolimna minima formed the sister
group to Pinnularia and Caloneis. At the base of this clade C. budensis diverged first (Fig.
23c). After the divergence of C. amphisbaena the other species were subdivided into two
clades.
The majority-rule consensus tree (Fig. 24) from the extended alignment based on 105 most
parsimonious trees.
The genus Eunotia (BS: 99/95) diverges at the base of this tree.
Navicula sensu stricto, Hippodonta capitata, Seminavis cf. robusta and Pseudogomphonema
cf. kamschaticum formed clade 1, which diverges next (BS: 66/45). H. capitata, S. cf. robusta
and P. cf. kamschaticum diverge within Navicula sensu stricto.
3. Result
49
Cyclotella choctawatcheeana (1493)
99/96
Eunotia
41/34
25/40
94/58
59/49/33
Hippodonta capitata (1272)
AY604699 Navicula salinicola
AY571748 Pseudogomphonema cf. kamschaticum
AY571750 Seminavis cf. robusta
Navicula sp.1 (1411)
AY571749 Navicula cf. duerrenbergiana
Navicula tripunctata (1434)
Navicula cari (1310)
Navicula capitatoradiata (1417)
-/60
94/90 Navicula radiosa (1278, 1433, 1440)
Navicula gregaria (1280)
Navicula reinhardtii (1282)
Navicula sp.2 (1319)
100/100
Navicula cryptotenella (1416, 1420, 1435)
Navicula cryptocephala (1279, 1316)
Navicula veneta (1276, 1277, 1281)
AY571747 Lyrella atlantica
100/100
AY571755 Lyrella hennedyi
Lyrella
96/93
98/89
AY571756 Lyrella sp.
AY571757 Petroneis humerosa
100/100
-/46
Cocconeis
Achnanthidium minutissimum (1438)
Navicula brockmannii (1425)
44/21
(a)
Eunotia formica (1268)
Eunotia sp. (1269)
100/100
Eunotia implicata (1321)
AY571744 Eunotia minor
84/87
99/96
0.01
clade 1
(b)
Gomphonema cf. parvulum (1315)
Gomphonema cf. angustatum (1409)
Gomphonema productum (1313)
Gomphonema affine (1322, 1439)
Gomphonema micropus (1270, 1271)
Gomphonema acuminatum (1424)
100/100
Gomphonema truncatum (1552)
99/92
AY571751 Gomphonema capitatum
51/21
44/21
77/60
clade 2
clade 3
0.005
Gomphonema
AY571754 Encyonema cf. sinicum
Encyonema caespitosum (1441)
Cymbella helmckei (1431)
99/92
Cymbella affinis (1414, 1423)
41/15
Cymbella aspera (1421)
93/82
Cymbella naviculiformis (1317, 1324)
Cymbella proxima (1422)
AY571753 Placoneis cf. paraelginensis
16/9
Navicula hambergii (1436)
33/24
AY571752 Placoneis constans
Placoneis sp. (1419)
Placoneis elginensis (1312)
(c)
68/66
26/17
52/46
84/67
100/100
36/30
clade 4
62/40
24/24
-/28
araphid diatoms
Amphora (group 1)
98/97
Neidum affine (1551)
Luticola goeppertiana (1273)
Navicula integra (1430)
Craticula cuspidata (1318, 1320)
Caloneis amphisbaena (1550)
Pinnularia acrosphaeria (1426)
94/79
Pinnularia subcapitata (1285)
Pinnularia microstauron (1287, 1288, 1289, 1290)
Pinnularia mesolepta (1314, 1429)
87/93
Pinnularia anglica (1286)
Pinnularia obscura (1292)
Caloneis lauta (1446)
Pinnularia rupestris (1311)
Pinnularia viridis (1428)
Pinnularia viridiformis (1291)
Pinnularia cf. substreptoraphe (1442)
Caloneis budensis (1323)
0.01
Stauroneis
74/26
66/58
Craticula molestiformis (1284)
Craticula halophilioides (1283, 1308)
Amphora normannii (1263)
24/24
Pinnularia/Caloneis
100/100
27/18
63/52
-/43
clade 5
Mayamaea
AY571746 Sellaphora pupula
AY571745 Sellaphora bacillum
Eolimna minima (1267)
0.05
Fig. 22: Phylogeny inferred with the ML analysis using rbcL sequences from GenBank and the
AlgaTerra cultures. Bootstrap values obtained from 1000 replications based on NJ analyses using JC
model and on parsimony analyses have been plotted at the nodes. Collapsed clades, which differ to
the equivalent clades in Figs. 19 and 20, are shown in detail in Fig. 23.
Fig. 23: Details of the ML tree analysis from rbcL sequences from GenBank and the AlgaTerra cultures.
Bootstrap values obtained from 1000 replications based on NJ analyses using JC model and on
parsimony analyses have been plotted at the nodes. (a) Eunotia, (b) Gomphonema, (c) Pinnularia and
Caloneis
3. Result
50
84/87
99/96
100/100
100/100
98/98
-/2
27/18
100/100
68/77
94/97
87/93
98/97
36/30
66/58
74/83
96/93
100/100
98/89
44/21
51/38
26/17
-/27
42/41
93/82
52/46
41/15
99/92
100/100
68/66
-/46
66/45
-/13
41/34
45/35
-/50
-/30
-/10
Mayamaea
44/21
99/92
clade 3
Pinnularia acrosphaeria (1426)
Pinnularia microstauron (1287, 1288, 1289, 1290)
Pinnularia subcapitata (1285)
Pinnularia mesolepta (1314, 1429)
Pinnularia anglica (1286)
Pinnularia obscura (1292)
Pinnularia viridiformis (1291)
Pinnularia viridis (1428)
Luticola goeppertiana (1273)
Neidum affine (1551)
Craticula halophilioides (1283, 1308)
Craticula molestiformis (1284)
Stauroneis phoenicenteron (1293, 1437)
Stauroneis anceps (1412)
AY571757 Petroneis humerosa
AY571747 Lyrella atlantica
AY571756 Lyrella sp.
Lyrella
AY571755 Lyrella hennedyi
Gomphonema
Cymbella proxima (1422)
AY571753 Placoneis cf. paraelginensis
Placoneis elginensis (1312)
Navicula hambergii (1436)
AY571752 Placoneis constans
Cymbella naviculiformis (1317, 1324)
Cymbella aspera (1421)
Cymbella affinis (1414, 1423)
Cymbella helmckei (1431)
Placoneis sp. (1419)
Encyonema chaespitosa (1441)
Encyonema
AY571754 Encyonema cf. sinicum
Achnanthidium minutissimum (1438)
100/100
-/28
25/40
94/58
-/60
94/90
AY604699 Navicula salinicola
Hippodonta capitata (1272)
Navicula sp.1 (1411)
Navicula veneta (1276, 1277, 1281)
AY571750 Seminavis cf. robusta
Navicula reinhardtii (1282)
Navicula cryptocephala (1279, 1316)
AY571748 Pseudogomphonema cf. kamschaticum
Navicula cryptotenella (1416, 1420, 1435)
Navicula sp.2 (1319)
Navicula gregaria (1280)
AY571749 Navicula cf. duerrenbergiana
Navicula tripunctata (1434)
Navicula cari (1310)
Navicula radiosa (1278, 1433, 1440)
Navicula capitatoradiata (1417)
77/60
100/100
clade 2
Cocconeis
araphid diatoms
84/67
53/47
Cyclotella choctawatcheeana (1493)
Eunotia formica (1268)
Eunotia sp. (1269)
Eunotia
Eunotia implicata (1321)
AY571744 Eunotia minor
Navicula brockmannii (1425)
Caloneis amphisbaena (1550)
Amphora normannii (1263)
Pinnularia rupestris (1311)
Pinnularia cf. substreptoraphe (1442)
Caloneis lauta (1446)
Caloneis budensis (1323)
Stauroneis kriegerii (1309, 1444)
Stauroneis gracilior (1294)
Craticula cuspidata (1318, 1320)
Navicula integra (1430)
Amphora libyca (1264)
Amphora sp. (1554)
Amphora pediculus (1265)
Amphora cf. fogediana (1427)
Eolimna minima (1267)
AY571746 Sellaphora pupula
AY571745 Sellaphora bacillum
clade 1
Fig. 24: Majority-rule consensus tree inferred with the parsimony analysis based on rbcL sequences
from GenBank and the AlgaTerra cultures. Bootstrap values obtained from 1000 replications based
on NJ analyses using JC model and on parsimony analyses have been plotted at the nodes. Collapsed
clades, which differ to the equivalent clades in Fig. 21, are shown in detail in Fig. 25.
Gomphonema cf. parvulum (1315)
Gomphonema cf. angustatum (1409)
Gomphonema productum (1313)
Gomphonema affine (1322, 1439)
Gomphonema micropus (1270, 1271)
Gomphonema acuminatum (1424)
Gomphonema truncatum (1552)
AY571751 Gomphonema capitatum
Fig. 25: Gomphonema clade of the parsimony tree analysis based on rbcL sequences from GenBank and
the AlgaTerra cultures. Bootstrap values obtained from 1000 replications based on NJ analyses using JC
model and on parsimony analyses have been plotted at the nodes.
3. Result
51
The next clade consists of araphid diatoms (BS: 84/67).
At the base of clade 2 the monoraphid diatoms diverge (BS: 0/46). The Gomphonema species
formed a monophyletic clade (Fig. 25), but it had only weak BS. Cymbella and Placoneis/N.
hambergii did not form monophyletic clades, because the unidentified Placoneis species
clusters with Encyonema and C. proxima clusters with remaining Placoneis species. But none
of these clades had bootstrap support.
Navicula brockmannii diverged at the base of clade 3. The other species were pooled in a
large polytomy. Four Amphora species formed a branch, which had maximum bootstrap
support. Another branch contained Eolimna minima, Sellaphora and Mayamaea (BS: 27/18).
Two strongly supported clades consist of Pinnularia species. P. acrosphaeria, P.
microstauron, P. subcapitata, P. mesolepta, P. anglica and P. obscura one hand (BS: 94/97)
and P. cf. substreptoraphe and P. viridiformis on the other hand (BS: 98/97). Other small
clades consists of Luticola goeppertiana and Neidum affine (BS: 36/30), Craticula
halophilioides and C. molestiformis (BS: 66/58), Stauroneis anceps and St. phoenicenteron
(BS: 74/83) and Petroneis humerosa and Lyrella (BS: 96/93).
3.1.4 Gene combination
The combination of the three alignments, which consists of data sequenced within the scope
of this study, resulted in an alignment having 3226 positions in total. 896 of these positions
were parsimony-informative and 297 parsimony-uninformative characters.
Results from 100 partition homogeneity test replicates indicated that SSU rDNA, LSU rDNA
and rbcL gene data were significantly heterogeneous (p=0,01).
The ML phylogeny based on this alignment is shown in Fig. 26. The collapsed clades of this
figure are shown in detail in Fig. 27.
The strongly supported (BS: 100/100) monophyletic clade of Eunotia diverged at the base of
this tree.
Clade 1, which includes Hippodonta capitata and the Navicula sensu stricto, had maximum
bootstrap support. Bootstrap values of 100 and 99 from NJ and MP bootstrap analysis support
the monophyly of Navicula sensu stricto. The Navicula sensu stricto was subdivided in three
groups (Fig. 27a). The first consists of N. veneta and N. gregaria (BS: 100/91). N.
crytotenella, N. reinhardtii, N. cryptocephala and the two unknown Navicula species formed
the second group which was supported by 88% and 84% of the bootstrap
3. Result
52
Cyclotella choctawatcheeana (1493)
Eunotia sp. (1269)
Eunotia formica (1268)
Eunotia
Eunotia implicata (1321)
99/99
100/100
100/99
(a)
Navicula s. str.
clade 1
100/100
Hippodonta capitata (1272)
Fragilaria sp. (1410)
Fragilaria crotonensis (1254)
araphid diatoms
Asterionella formosa (1256)
Cymbella affinis (1414, 1423)
100/100
45/78
Cymbella helmckei (1431)
56/70
Cymbella proxima (1422)
Cymbella
99/75
Cymbella aspera (1421)
Cymbella naviculiformis (1317, 1324)
52/30
Placoneis sp. (1419)
-/Navicula hambergii (1436)
80/58
Placoneis elginensis (1312)
100/100
99/99
100/78
100/95
83/87
100/100
Navicula gregaria (1280)
Navicula veneta (1281)
100/91
Navicula veneta (1277)
Navicula veneta (1276)
Navicula cryptotenella (1416, 1420, 1435)
100/100
-/57
100/99
Navicula sp.2 (1319)
96/98
Navicula cryptocephala (1279, 1316)
88/84
Navicula reinhardtii (1282)
Navicula sp.1 (1411)
Navicula radiosa (1278, 1433, 1440)
73/63
98/96
Navicula capitatoradiata (1417)
100/100
Navicula cari (1310)
-/52
Navicula tripunctata (1434)
Hippodonta capitata (1272)
0.01
clade 2
(b)
Gomphonema affine (1322, 1439)
Gomphonema cf. parvulum (1315)
Gomphonema cf. angustatum (1409)
70/43 91/86
Gomphonema productum (1313)
Gomphonema acuminatum (1424)
100/100
Gomphonema truncatum (1552)
Gomphonema micropus (1270, 1271)
100/100
100/100
Gomphonema
45/77
54/47
38/46
100/100
97/95
51/84
Encyonema caespitosum (1441)
Navicula brockmannii (1425)
Achnanthidium minutissimum (1438)
monoraphid
Cocconeis pediculus (1415)
100/100
diatoms
Cocconeis
Cocconeis placentula (1418)
Luticola goeppertiana (1273)
99/97
Neidum affine (1551)
Stauroneis phoenicenteron (1293, 1437)
52/87
29/60
Stauroneis anceps (1412)
Stauroneis gracilior (1294)
76/66
Stauroneis kriegerii (1309, 1444)
Navicula integra (1430)
Craticula halophilioides (1283, 1308)
Craticula molestiformis (1284)
Craticula
100/100
Craticula cuspidata (1318, 1320)
Mayamaea atomus var. atomus (1274)
100/100
Mayamaea
Mayamaea atomus var. permitis (1275)
83/87
0.01
(c)
97/93
80/69
100/100
clade 3
98/99
98/99
Pinnularia/Caloneis
58/30
100/98
100/100
85/79
83/91
Eolimna minima (1267)
Amphora sp. (1554)
Amphora libyca (1264)
Amphora pediculus (1265)
Amphora cf. fogediana (1427)
Amphora normannii (1263)
Pinnularia microstauron (1287, 1288, 1289, 1290)
Pinnularia subcapitata (1285)
Pinnularia mesolepta (1314, 1429)
Pinnularia anglica (1286)
Pinnularia obscura (1292)
Pinnularia acrosphaeria (1426)
Caloneis budensis (1323)
Pinnularia rupestris (1311)
Pinnularia cf. substreptoraphe (1442)
Pinnularia viridiformis (1291)
Pinnularia viridis (1428)
Caloneis amphisbaena (1550)
Caloneis lauta (1446)
100/100
Amphora
0.05
Fig. 26: Phylogeny inferred with the ML analysis using the combined dataset of SSU rDNA, LSU
rDNA and rbcL sequences from the AlgaTerra cultures. Bootstrap values obtained from 1000
replications based on NJ analyses using JC model and on parsimony analyses have been plotted at the
nodes. Condensed clades are shown in detail in Fig. 27.
0.02
Fig. 27: Details of the ML tree analysis from the combined dataset of SSU rDNA, LSU rDNA and rbcL
sequences from the AlgaTerra cultures. Bootstrap values obtained from 1000 replications based on NJ
analyses using JC model and on parsimony analyses have been plotted at the nodes. (a) Navicula sensu
stricto, (b) Gomphonema, (c) Pinnularia and Caloneis
3. Result
replicates. The third group contained N. radiosa, N. capitatoradiata, N. cari and the type
species N. tripunctata and had maximum bootstrap support.
Although the araphid diatoms were assigned as outgroup they diverge next. They formed a
strongly supported monophyletic group (BS: 99/99).
The remaining taxa diverge into two major clades. At the base of clade 2 the monoraphid
genera formed a monophyletic group (BS: 38/46). Then Navicula brockmannii and
Encyonema caespitosum diverged. The next clade contained Gomphonema species. Within
this monophyletic group (BS: 83/87) G. micropus diverged first and was separated from the
other Gomphonema species (Fig. 27b) by branch length and maximum BS. The next
divergence separates Placoneis/Navicula hambergii (BS: 100/78) from Cymbella (BS: 99/75).
At the base of clade 3 in the ML tree Luticola goeppertiana and Neidum affine form a clade,
which was supported by bootstrap values of 99 and 97.
Within the next sub-clade (BS: 83/91) Amphora normannii was separated from the other
Amphora species by branch length and maximum bootstrap support from both analyses.
Eolimna minima is at the base of the next sub-clade, which includes Mayamaea and a
monophyletic group containing Pinnularia and Caloneis. Within this well supported group
(BS: 98/99) Caloneis lauta diverged first (Fig. 27c). The other species were subdivided into
two clades. One clade consists of C. budensis, P. rupestris, P. viridis, P. cf. substreptoraphe,
P. viridiformis and C. amphisbaena, the other contained P. acrosphaeria, P. obscura, P.
anglica, P. mesolepta, P. subcapitata and P. microstauron.
At the next divergence, the monophyletic clade containing the three Craticula species (BS:
100/100) and a clade containing Navicula integra and four Stauroneis species (BS: 76/66)
were separated.
The maximum parsimony (MP) analysis based on the combined data set resulted in 6 most
parsimonious trees. The majority-rule consensus tree of these trees is shown in Figs 28 and
29.
Neidum affine and Luticola goeppertiana formed a well supported clade (BS: 99/97), which
diverged from the polytomy at the base of the tree.
The Eunotia species formed monophyletic group with maximum bootstrap support and a
bootstrap value of 99 from both bootstrap analyses support the monophyly of the araphid
diatoms. They diverge next, but the phylogenetic relationship between these two groups is not
resolved.
53
3. Result
54
100/91
(a)
100/100
100/99
99/97
99/99
100/100
99/99
100/100
100/99
100/100
100/100
38/46
-/34
100/100
91/86
70/43
100/100
100/100
83/87
54/47
-/27
100/100
80/58
45/78
56/70
99/75
100/95
52/30
45/77
100/78
60/53
-/33
100/100
-/46
51/84
98/99
52/87
29/60
68/60
76/66
58/30
100/100
97/95
100/100
48/39
85/79
100/98
100/100
83/91
Cyclotella choctawatcheeana (1493)
Neidum affine (1551)
Luticola goeppertiana (1273)
Eunotia sp. (1269)
Eunotia formica (1268)
Eunotia
Eunotia implicata (1321)
Asterionella formosa (1256)
Fragilaria crotonensis (1254)
araphid
Fragilaria sp. (1410)
96/98
88/84
100/100
73/63
98/96
100/100
diatoms
Navicula s. str.
Hippodonta capitata (1272)
Cocconeis placentula (1418)
Cocconeis monoraphid
Cocconeis pediculus (1415)
diatoms
Achnanthidium minutissimum (1438)
Gomphonema affine (1322, 1439)
Gomphonema cf. parvulum (1315)
Gomphonema cf. angustatum (1409)
Gomphonema productum (1313)
Gomphonema
Gomphonema truncatum (1552)
Gomphonema acuminatum (1424)
Gomphonema micropus (1270, 1271)
Cymbella affinis (1414, 1423)
Cymbella helmckei (1431)
Cymbella proxima (1422)
Cymbella
Cymbella aspera (1421)
Cymbella naviculiformis (1317, 1324)
Placoneis sp. (1419)
Navicula hambergii (1436)
Placoneis elginensis (1312)
Encyonema caespitosum (1441)
Navicula brockmannii (1425)
Mayamaea atomus var. atomus (1274)
Mayamaea
Mayamaea atomus var. permitis (1275)
Eolimna minima (1267)
clade 1
(b)
100/100
97/93
80/69
83/90
100/100
-/72
96/99
98/99
clade 2
-/66
-/45
93/91
100/99
Pinnularia microstauron (1287, 1288, 1289, 1290)
Pinnularia subcapitata (1285)
Pinnularia mesolepta (1314, 1429)
Pinnularia anglica (1286)
Pinnularia obscura (1292)
Caloneis amphisbaena (1550)
Pinnularia acrosphaeria (1426)
Caloneis lauta (1446)
Caloneis budensis (1323)
Pinnularia rupestris (1311)
Pinnularia cf. substreptoraphe (1442)
Pinnularia viridis (1428)
Pinnularia viridiformis (1291)
Fig. 29: Details of the parsimony tree analysis based on a combined dataset of SSU rDNA, LSU rDNA
and rbcL sequences from the AlgaTerra cultures. Bootstrap values obtained from 1000 replications based
on NJ analyses using JC model and on parsimony analyses have been plotted at the nodes. (a) Navicula
sensu stricto, (b) Pinnularia and Caloneis
Pinnularia/Caloneis
Stauroneis phoenicenteron (1293, 1437)
Stauroneis anceps (1412)
Stauroneis
Stauroneis gracilior (1294)
Stauroneis kriegerii (1309, 1444)
Navicula integra (1430)
Craticula molestiformis (1284)
Craticula halophilioides (1283, 1308)
Craticula
Craticula cuspidata (1318, 1320)
Amphora sp. (1554)
Amphora libyca (1264)
Amphora pediculus (1265)
Amphora
Amphora cf. fogediana (1427)
Amphora normannii (1263)
Navicula gregaria (1280)
Navicula veneta (1276, 1277, 1281)
Navicula cryptotenella (1416, 1420, 1435)
Navicula sp.2 (1319)
Navicula cryptocephala (1279, 1316)
Navicula reinhardtii (1282)
Navicula sp.1 (1411)
Navicula radiosa (1278, 1433, 1440)
Navicula capitatoradiata (1417)
Navicula cari (1310)
Navicula tripunctata (1434)
Hippodonta capitata (1272)
clade 3
Fig. 28: Majority-rule consensus tree inferred with the parsimony analysis based on a combined
dataset of SSU rDNA, LSU rDNA and rbcL sequences from the AlgaTerra cultures. Bootstrap values
obtained from 1000 replications based on NJ analyses using JC model and on parsimony analyses
have been plotted at the nodes. Condensed regions are shown in detail in Fig. 29.
3. Result
55
Navicula sensu stricto and Hippodonta capitata were sister groups (BS: 100/100) and formed
the basal clade 1 of the raphid diatoms. Bootstrap values of 100 and 99 support the
monophyly of Navicula sensu stricto. The Navicula sensu stricto clade further diverges into
three groups Fig. 29a). N. veneta and N. gregaria formed the first group, which was supported
by 100% and 91% of the bootstrap replicates. The second group consisted of N. crytotenella,
N. reinhardtii, N. cryptocephala and the two unidentified Navicula species (BS: 88/84). The
third group contained N. radiosa, N. capitatoradiata, N. cari and N. tripunctata. The
phylogenetic relationship within these groups is not totally resolved.
The other raphid diatoms diverge into two large clades, which were supported by bootstrap
values between 30 and 58.
Clade 2 consists of the monoraphid diatoms, Navicula brockmannii and several Cymbellales.
A monophyletic clade containing the monoraphid diatoms (BS: 38/46) is the basal branch. N.
brockmannii diverges next. Within the Cymbellales (BS: 100/95) Encyonema caespitosum
diverges first. The next clade was formed by Gomphonema species. Within this monophyletic
group (BS: 83/87) G. micropus diverges first. Its sister clade consists of Cymbella, Placoneis
and Navicula hambergii (BS: 52/30). The monophyly of Cymbella is supported by relativly
high bootstrap values (BS: 99/75). The Placoneis species and N. hambergii form a
monophyletic group supported by bootstrap values of 100/78.
The basal divergence of clade 3 consists of Eolimna, Mayamaea, Pinnularia and Caloneis
and supported by bootstrap values of 51 and 84. Pinnularia and Caloneis formed a strongly
supported monophyletic group (BS: 98/99), which further diverges into two clades (Fig. 29b).
C. budensis, P. rupestris, P. viridis, P. cf. substreptoraphe and P. viridiformis formed on
clade, the other contained C. lauta, P. acrosphaeria, C. amphisbaena, P. obscura, P. anglica,
P. mesolepta, P. subcapitata and P. microstauron. The genus Amphora diverges next (BS:
83/91). Maximum bootstrap values support the monophyly of Craticula, which diverges next.
The sister group of Craticula consists of Navicula integra at the base of Stauroneis.
3.2 Morphological support for molecular data
3.2.1 Navicula sensu stricto
With the exception of the tree in Fig. 22, the Navicula sensu stricto species were pooled in a
monophyletic clade with Hippodonta capitata diverging at the base. In one tree (Fig. 8) two
Pseudogomphonema species appeared within this clade, in other trees Neidum affine (Figs.
3. Result
56
11, 15) and Luticola goeppertiana (Fig. 15) diverged within Navicula sensu stricto. In all
trees N. radiosa, N. capitatoradiata, N. cari and N. tripunctata form a monophyletic group
(group 1), to which N. lanceolata and N. ramosissima were additionally added in Fig. 8 and
N. cf. erifuga in Figs. 15 and 18. The other species were paraphyletic (Fig. 8 and 24) or
subdivided in two groups. With the exception of one unidentified species, these groups
consists always of N. veneta and N. gregaria on one hand (group 2) and the remaining species
on the other hand (group 3).
The morphological investigations concentrated on the three groups. But no features could be
found that were typical for one group and absent in the other groups. All species show the
typical features for Navicula sensu stricto, such as two plate-like plastids (Figs. 30 a + e, 31 a
+ d and 33 a, c, e, g) or apically elongated liner poroids (Figs. 30 c, d, g, h; 32 and 34).
Differences in the outline of the valves were greater within the groups than between them
(Figs. 30 - 34)
3.2.2 Amphora
The four species A. libyca, A. pediculus, A. cf. fogediana and the unidentified Amphora
species (1554) formed a monophyletic clade, which had maximum BS in all trees. In the tree
in Fig. 9, the SSU rDNA sequence of A. cf. proteus (obtained from GenBank) diverged at the
base of this group (BS: 100). The fifth species isolated within the scope of this study (A.
normannii) also diverged at the base of this group in some trees (Figs. 3, 5, 12, 13, 26 and
27), but in the ML trees (Figs. 3, 12 and 26) the branch length indicated a separation. In all
phylogenies based on rbcL sequences (Figs. 19, 21, 22 and 24), A. normannii is separated
from the other four species. In Fig. 10 A. normannii formed a strongly supported
monophyletic clade with A. montana, A. cf. capitellata and an unidentified species. A.
coffeaeformis is separated from this group in Fig. 10, but formed a well supported clade with
A. normannii in Fig. 16.
All species show the typical asymmetrical valve morphology (Figs. 35 - 37). In an intact
frustule both raphe systems lie on the same side (Figs. 35 c; 36 c, d, f; 37 b, d, h). Therefore
live individuals normally lie on the ventral side (Figs. 35 a; 36 a). From the five species
cultured within the scope of this study, only A. normannii has numerous girdle bands (Figs.
35 c; 36 c, d, f; 37 b, d, f, h). The girdle of the other four species have not more than two
girdle bands (valvocopula, Fig. 37 b, d, f, h) From the species, of which sequences were
obtained from GenBank, A. coffeaeformis, A. montana and A. cf. capitellata have numerous
girdle bands (Frenguelli, 1938, Krammer & Lange-Bertalot, 1986).
3. Result
57
e
a
f
b
g
c
d
Fig. 30: a - d: Navicula gregaria, e - h: Navicula veneta.
a + e: Light micrograph of live individual. b + f: Light micrograph of cleaned valve.
c + g: SEM, internal view of a valve. d + h: SEM, external view of a valve.
h
3. Result
58
b
a
c
d
Fig. 31: Navicula species
a + b: N. cari, light micrograph of live individual (a) and cleaned valve (b);
c: N. tripunctata light micrograph of a cleaned valve;
d + e: N. radiosa, light micrograph of live individual (d) and a cleaned valve (e).
e
3. Result
59
a
b
c
d
e
f
h
g
Fig. 32: Navicula species, SEM of valve exteriors (a, c, e, g) and interiors (b, d, f,
h), some with enlarged midvalve or apice details.
a + b: N. tripunctata, c + d: N. cari, e + f: N. capitatoradiata, g + h: N. radiosa
3. Result
60
a
b
c
d
e
g
f
h
Fig. 33: Navicula species, light micrographs of live individuals (a, c, e, g) and cleaned
valves (b, d, f, h, i)
a + b: N. reinhardtii, c + d: N. cryptotenella, e + f: N. cryptocephala, g + h: Navicula
sp.2 (1319), i: Navicula sp.1 (1411).
i
3. Result
61
a
b
c
d
f
e
g
h
i
k
Fig. 34: Navicula species (group 3), SEM of valve exteriors (a, c, e, g, i) and interiors (b, d,
f, h, k), some with enlarged midvalve or apice details.
a + b: N. cryptotenella, c + d: N. reinhardtii, e + f: N. cryptocephala, g + h: Navicula sp.2
(1319), i + k: Navicula sp.1 (1411).
3. Result
62
a
d
b
e
c
f
Fig. 35: Amphora normannii.
a: Light micrograph of live individual. b, c: Light micrographs of cleaned valve in valve view (b) and
ventral view (c). d: SEM, internal view. e: SEM, external view. f: SEM, girdle band
3. Result
63
a
c
b
f
d
e
g
Fig. 36: Light micrographs of Amphora species.
a - c: A. libyca, live individual (a) and cleaned valves in valve view (b) and ventral view (c).
d + e: A. cf. fogediana, cleaned valves in valve view (d) and ventral view (e).
f + g: A. pediculus, cleaned valves in valve view (f) and ventral view (g).
3. Result
64
a
b
c
d
f
e
g
h
Fig 37: SEM micrographs of Amphora species.
a - c: A. libyca, frustule exteriors (a: valve view, b: ventral side) and valve interior (c).
d - f: A. pediculus, frustule exteriors (d: ventral side, f: dorsal side) and interior (e).
g + h: A.cf.fogediana, frustule exteriors (g top: valve view, h: ventral side) and valve interior (g below).
3. Result
65
3.2.3 Pinnularia and Caloneis
The position of the three Caloneis species and of P. acrosphaeria differs in the different
phylogenies. But in all trees, P. obscura, P. anglica, P. mesolepta, P. subcapitata and P.
microstauron were members of one clade (group 1), whereas P. rupestris, P. viridis, P. cf.
substreptoraphe and P. viridiformis formed a second clade (group 2). Within group 1 P.
obscura, P. anglica and P. mesolepta formed a well supported clade in all trees (BS: 87 –
100).
The morphological investigations of the two Pinnularia groups showed clear differences.
The species in group 1 have a filiform or slightly lateral raphe system, but even if the external
raphe slit is slightly undulate it never crosses the internal slit (Fig. 38 b, d, f, h, i and l). All
species have a large central area, which is often a fascia extended to the margin on one or both
sides (Figs. 38 b, d, f, h, i and l and 39). Midvalve, the striae are radial or parallel and become
convergent or parallel at the apices (Figs. 38 b, d, f, h, i and l and 39). P. subcapitata and P.
microstauron have two plate-like chloroplasts (Fig. 38 g, k). But P. obscura, P. anglica and
P. mesolepta have a single H-shaped plastid arranged as two large plates along each side of
the girdle and a very fine bridge under one valve face (Fig. 38 a, c, e).
All species in group 2 have a lateral raphe system (Fig. 40 b, d, f, g). In some species the
undulate external raphe fissure crosses over the internal fissure along the raphe (complex
raphe, Fig. 40 f, g). The linear or linear-lanceolate axial area pass into a slightly expanded
central area, which is often asymmetric (Fig. 40 b, d, f, g). The striae are radial or parallel at
the centre of the valve and convergent or parallel at the apices (Fig. 40 b, d, f, g). Under the
light microscope the striae were crossed by a lateral line (Fig. 40 b, d, f, g). As it can be seen
in the SEM pictures (Fig. 41), partially closed alveolae are causal for this line. All species
have two plate-like chloroplasts (Fig. 40 a, c, e)
The lateral raphe system of P. acrosphaeria lies in a wide, linear axial area (Fig. 42 b). The
central area is only slightly wider (Figs. 42 b and 43 a). The texture of these areas is very
special (Figs. 42 b and 43 a ). Midvalve the striae are radial or parallel and become
convergent or parallel at the apices (Fig. 42 b and 43 a) and the alveolae are partially closed
(Fig. 43 b). As shown in Fig. 42 a , this species has two lobed chloroplasts.
The raphe of C. lauta is curved, but not lateral (Fig. 42 d). The axial area is relatively narrow
and the central area form a wide fascia (Figs. 42 d and 43 c). The striae are parallel or slightly
radial (Figs. 42 d and 43 c). On the external valve face a raised line cross the striae (Fig. 43 c),
3. Result
66
a
b
c
d
e
f
g
h
k
i
l
Fig. 38: Pinnularia species (group 1), light micrographs of live individuals (a, c, e, g, k) and cleaned
valves (b, d, f, h, i, l)
a + b: P. obscura, c + d: P., anglica e + f: P., mesolepta g - i: P. microstauron, k + l: P. subcapitata.
3. Result
67
a
b
c
d
e
f
g
i
h
k
Fig. 39: Pinnularia species (group 1), SEM of valve exteriors (a, c, e, g, i) and interiors (b, d, f, h, k).
a + b: P. obscura, c + d: P.anglica, e + f: P. mesolepta, g + h: P. microstauron, i + k: P. subcapitata
3. Result
68
a
b
c
d
e
g
f
Fig. 40: Pinnularia species (group 2), light micrographs of live individuals (a, c, e) and cleaned valves
(b, d, f, g)
a + b: P. rupestris, c + d: P. viridiformis, e + f: P. viridis, g: P. cf. substreptoraphe.
3. Result
69
a
b
c
e
g
d
f
h
Fig. 41: Pinnularia species (group 2), SEM of valve exteriors (a, c, e, g) and interiors (b, d, f, h).
a + b: P. rupestris, c + d: P. viridiformis, e + f: P. viridis, g + h: P. cf. substreptoraphe.
3. Result
70
a
b
c
d
e
f
g
h
Fig. 42: Pinnularia and Caloneis species, light micrographs of live individuals (a, c, e, g) and cleaned
valves (b, d, f, h)
a + b: P. acrosphaeria, c + d: C. lauta, e + f: C. amphisbaena, g + h: C. budensis.
3. Result
71
a
b
c
d
e
f
g
h
Fig. 43: Pinnularia and Caloneis species, SEM of valve exteriors (a, c, e, g) and interiors (b, d, f, h).
a + b: P. acrosphaeria, c + d: C. lauta, e + f: C. amphisbaena, g + h: C. budensis.
3. Result
72
internally the alveolae are partially closed (Fig. 43 d). This species has two plate-like
chloroplast, too (Fig. 42 c).
C. amphisbaena has a slightly lateral raphe system and a narrow and linear axial area close to
the apices, which passes into a large rhombic-lanceolate central area (Fig. 42 f). The striae are
radial at the centre of the valve and become convergent or parallel at the apices (Figs. 42 f and
43 e + f), crossed by a lateral line (Figs. 42 f and 43 e ). The alveolae are partially closed (Fig.
43 f). Two lobed, plate-like chloroplast can be observed in live individuals (Fig. 42 e)
C. budensis has a lateral raphe system and the axial area is lanceolately expanded towards the
wide fascia midvalve (Fig. 42 h). The striae are parallel or slightly radial at the centre of the
valve and convergent at the apices (Figs. 42 h and 43 g + h). They are crossed by a lateral line
(Fig. 42 h), because the alveolae are partially closed (Fig. 43 h). Two lobed, plate-like
chloroplast can be observed in live individuals (Fig. 42 g ).
3.2.4 Stauroneis, Craticula and Navicula integra
With the exception of the trees based on the rbcL sequences, Stauroneis, Craticula and N.
integra formed a strongly supported monophyletic clade (BS: 91 – 100). In the ML trees
based on the rbcL sequences, the monophyletic clades of this group had poor BS (Figs. 19,
22) and in the MP phylogenies based on this gene, this group was merged in a large polytomy
(Figs. 21, 24). Within the monophyletic clades, Craticula and Stauroneis were clearly
separated. The only exception was the tree in Fig. 22, where a monophyletic clade of
Stauroneis divided the Craticula species into two groups. In most trees, N. integra clusters
with Stauroneis. Only in the ML phylogenies based on the rbcL sequences does N. integra
cluster with Craticula.
Also the morphological investigation showed several similarities. All species within this
group has two plate-like chloroplasts, lying one against each side of the girdle (Fig. 44 a, d,
g). All frustules were isopolar and tend to lie in valve view, because they are wider
transapically than pervalvarly (Fig. 44). The external central raphe endings are expanded and
the well developed terminal fissures at the poles curve off to the same side of the valve (Fig.
44 b, c, e, f, h, i). The internal central raphe endings are simple and straight or slightly curved
Fig. 45 a, c, e). The girdle composed of several open, porous bands with one or two rows of
small round poroids (Fig. 45 b, d, f). The uniseriate striae consist of small round or elliptical
poroids which were occluded by hymenes at their internal aperatures (Fig. 45 c and Round et
al. 1990: p. 592 Fig. i and p. 595, Figs. i. j). In cleaned material these hymenes were often
eroded (Fig. 45 c).
3. Result
d
a
e
g
b
c
h
f
73
Fig. 44: Craticula cuspidata, Navicula integra and Stauroneis phoenicenteron
a - c: C. cuspidata, light micrograph of a live individual (a) and of a cleaned valve (b) and SEM of valve exterior (c).
d - f: N. integra, light micrograph of a live individual (d) and of a cleaned valve (e) and SEM of valve exterior (f).
g- i: St. phoenicenteron, light micrograph of a live individual (g) and of a cleaned valve (h) and SEM of valve exterior (i).
i
3. Result
e
a
c
b
d
74
Fig. 45: Craticula cuspidata, Navicula integra and Stauroneis phoenicenteron
a - b: C. cuspidata, SEM of valve interior (a) and girdle bands (b).
c - d: N. integra, SEM of valve interior (d) and girdle bands (e).
e - f: St. phoenicenteron, SEM of valve interior (g) and girdle bands.
f
3. Result
75
The most obvious feature of the genus Stauroneis is its stauros (Figs. 44 e + f, 45 e + f and
46). The stauros extends from the raphe-sternum to the valve margins, with decreasing
thickness.
All species belonging to the genus Craticula had parallel and equidistant striae (Figs. 44 a +
b, 45 a + b and 47). The areolae are aligned longitudinal in straight lines parallel to the raphe
system.
N. integra has a lanceolate or lanceolate-elliptical valve with subrostrate apices and an
additional undulation in the valve margin before the apices (Figs. 44 a – c and 45 c). At the
apices are pseudosepta (Fig. 45 c) The striae are radiate and midvalve more distant (16 – 18
striae/ 10 µm) (Figs. 44 a – c and 45 c). The costae separating the striae are thickened at the
centre of the valve, producing a stauros-like structure (Figs. 44 e and 45 c).
3.2.5 Gomphonema
In most gene trees, the species Gomphonema formed a monophyletic clade. Only in the tree in
Fig. 21 did the genus merged in a large polytomy and in Fig. 5, G. micropus is separated from
the other Gomphonema species by Cymbella. With the exception of the trees based on rbcL
sequences, G. micropus diverges at the base of the group. Whereas the entire group had only
low or moderate BS between 16 and 87, the clade without G. micropus was supported by high
bootstrap values of 99 or 100.
Nearly all morphological features of G. micropus can be found in one or more of the other
Gomphonema species cultured within the scope of this study. All species belonging to this
genus have a single H-shaped chloroplast (Figs. 48 a and 49) and heteropolar valves. All
species cultured have a single stigma, which internally opens in a slit (Figs. 48 d and 50).
Generally the striae are uniseriate, but in some species they can become biseriate close to the
raphe (Figs. 48 c + g and 50). Also the separation of the areolas in the alveolus by a small
strut can be found in several species (Fig. 48 d and 50 b, c, f). But G. micropus is the only
species where the areolae externally open in small round poroids (Fig. 48 e – g). In all other
cultured species, they are C- or kidney-shaped (Fig. 51). In all species, the external central
raphe endings are expanded and internally they deviate (Figs. 48 d and 50). Internally the
raphe slit ends in a helictoglossae at both poles. Externally the raphe fissure is hooked. The
only exception can be found in G. micropus with a smooth curved terminal fissure (Figs. 48 e
+ f and 51).
3. Result
76
a
d
e
b
c
f
Fig. 46: Stauroneis species, light micrograph of cleaned valve (a – c), SEM of internal view of the
valve centre (d – f).
a + d: St. anceps, b + e: St. gracilior, c + f: St. kriegerii.
3. Result
77
e
f
a
b
c
g
d
h
Fig 47: a - d: Craticula halophilioides, e - h: Craticula molestiformis.
a + e: light micrograph of live individuals, b + f: light micrographs of cleaned valves, c + g: SEM of
valve exteriors, d + h: SEM of valve interiors.
3. Result
78
a
b
c
d
e
f
g
h
Fig. 48: Gomphonema micropus.
a + b: light micrograph of live individual (a) and cleaned valve (b) in valve view,
c + d: SEM of valve interior of the whole valve (c) and midvalve detail (d),
e - h: SEM of valve exteroir of base (e) and head pole (f), whole valve (g) and girdle (h).
3. Result
79
a
b
c
d
Fig 49: Gomphonema species, light micrographs of live individuals.
a: G. cf. angustatum, b: G. affine, c: G. cf. parvulum, d: G. acuminatum, e: G. productum.
e
3. Result
80
a
b
d
c
e
f
Fig 50: Gomphonema species, SEM of valve interiors showing midvalve.
a: G. cf. angustatum, b: G. affine, c: G. cf. parvulum, d: G. truncatum, e: G. acuminatum, f: G.
productum.
3. Result
81
b
a
d
f
c
e
g
Fig 51: Gomphonema species, SEM of valve exteriors showing the polar raphe curvature and the
areolae.
a: G. cf. angustatum, head pole; b + c: G. affine, head pole of with corroded areolae (b) and
uncorroded areolae (c); d: G. cf. parvulum, base pole; e: G. truncatum, base pole; f : G. acuminatum,
head pole; g: G. productum, base pole.
3. Result
82
3.2.6 Placoneis and Navicula hambergii
The species belonging to the genus Placoneis and N. hambergii formed a monophyletic clade
in all phylogenies inferred with the ML analysis (Figs. 3, 9, 11, 15, 19, 22 and 26). These
clades were supported by bootstrap values from 24 to 100. The MP analyses of SSU rDNA
sequences (Fig. 5) and the combined dataset (Fig. 28) resulted in a monophyly of these
species, too. In the MP phylogenies based on the sequences of the LSU rDNA (Figs. 13, 17)
and rbcL (Figs. 21, 24) these species were not monophyletic, but still closely related.
Morphological investigations of Navicula hambergii and Placoneis elginensis indicated that
the two species are near relatives. The single chloroplast, with a central bridge from which
lobes project into the four quadrants of the cell (Fig. 52 a, b, h, i), is typically for species
belonging to the genus Placoneis. The striae are radiate (Fig. 52 c and k). At the centre of the
valve the striae are irregularly abbreviated (P. elginensis, Fig. 52 c + g) or alternately longer
and shorter (N. hambergii, Fig. 52 k + o). With SEM it can be seen, that, externally, the striae
consist of small round poroids (Fig. 52 g + o). Internally, the striae poroids are almost square
and closed by vola-like occlusions (Fig. 52 d + l). Both species have a straight raphe with
slightly expanded external central endings and at both poles the hook-like raphe fissures curve
to the same side (Fig. 52 c, g, k, o). The internal central raphe endings of both species are
hooked (Fig. 52 f + n) and the internally helictoglossae at the polar raphe endings are strait
and knob-like (Fig. 52 e + m).
3.2.7 Cymbella
In all trees with the exception of the tree in Fig. 21 Cymbella is most closely related to
Placoneis, Gomphonema and Encyonema. If there is no polytomy in this part of the tree, C.
aspera, C. helmckei and C. affinis always belong to a monophyletic clade. But the position of
C. naviculiformis and C. proxima differs in the different trees. In some gene trees, the branch
with these two species formed a monophyletic clade with the other Cymbella species (Figs. 4,
9, 12,). In Fig. 5 this branch is separated from the other Cymbella species by Gomphonema. In
the trees based on the rbcL sequences C. proxima appeared at the base of Placoneis, whereas
in the trees based on the combined data set this species diverged within Cymbella and C.
naviculiformis formed the base of the clade.
The morphological investigations of this genus concentrated on differences between C.
aspera, C. helmckei and C. affinis on one side and C. naviculiformis or C. proxima on the
other side. All species have the typical features of the genus Cymbella, such as dorsiventral
3. Result
83
b
e
a
d
c
g
f
h
i
m
l
k
o
n
Fig. 52: Placoneis paraelginensis (a – g) and Navicula hambergii (h – o)
Light micrographs of a live cell (a + h: girdle view, b + i: valve view) and cleaned valve (c + k).
SEM showing valve interiors (d + l: detail areolae, e + m: total view, f + n: detail central raphe
endings) and exterior (g + o: total view).
3. Result
84
valves, uniseriate striae and dorsal deflected terminal raphe fissures (Fig. 53). Contrary to the
other four species, stigmata (Fig. 55) and apical pore fields (Fig. 54) are absent in C.
naviculiformis. Internally the raphe ends straight in a helictoglossa (Fig. 56 e) in C.
naviculiformis. In the other species, the internally polar raphe slit is curved (Fig. 56 a – d). C.
proxima did not show obvious differences to C. aspera, C. helmckei and C. affinis.
3.2.8 Navicula brockmannii
In most phylogenies, N. brockmannii was closely related to the monoraphid diatoms and the
Cymbellales. Exceptions were only the trees based on LSU rDNA sequences (Figs. 11, 13). In
Fig. 11 N. brockmannii and the monoraphid species were most closely related to a clade
consisting of Craticula and Stauroneis and within the large polytomy in Fig. 13 only a
relationship of N. brockmannii and Achnanthidium minutissimum was shown. In all
phylogenies, N. brockmannii is always clearly separated by several genera from Navicula
sensu stricto.
In contrast to species belonging to Navicula sensu stricto (see 3.2.1), N. brockmannii had only
a single chloroplast (Fig. 57 a). The valves were linear with parallel or slightly convex
margins and broad rostrate or subcapitate ends (Fig. 57 b + c). The raphe was filiform with
scarcely expanded central pores (Fig. 57 b - d) and laterally strongly deflected terminal
fissures (Fig. 57 b + e). The helictoglossae at the internal polar raphe endings are straight and
knob-like (Fig. 57 f). The axial area was linear and narrow and slightly widened close to the
central area, which was variable in size and form because of irregularly abbreviated striae
(Fig. 57 b + c). The striae were radiate, getting parallel towards the poles (Fig. 57 b + c). At
the centre of the valve the striae were less dense (25 – 27/10µm) than towards the valve ends
(30 – 32/10µm). The striae run continuously from the valve surface down onto the mantle
(Fig. 57 e + g ) and consists of uniseriate rows of round areolae, which where externally
closed by hymenes (Fig. 57 c + d). One or two rows of areolae could be found on the girdle
bands (Fig. 57 g).
3.2.9 Varieties of Mayamaea atomus
The two varieties M. atomus var. atomus and M. atomus var. permitis formed strongly
supported (BS: 96 – 100) monophyletic clades in all phylogenies shown above. The
differences between the sequences, which were visualised by the branch length in the ML
phylogenies (Figs. 3, 10, 11, 15, 19, 22 and 26), were almost as many as between the two well
defined Cocconeis species.
3. Result
85
a
b
c
d
e
Fig. 53: Cymbella species, light micrographs of cleaned valves.
a: C. naviculiformis, b: C. proxima, c: C. affinis, d: C. aspera, e: C. helmkei.
3. Result
86
b
a
c
d
e
Fig. 54: Cymbella species, SEM of external polar raphe endings.
a: C. naviculiformis, b: C. proxima, c: C. affinis, d: C. aspera, e: C. helmkei.
3. Result
87
b
a
c
d
e
Fig. 55: Cymbella species, SEM of midvalve interior.
a: C. naviculiformis, b: C. proxima, c: C. affinis, d: C. aspera, e: C. helmkei.
3. Result
88
a
b
c
d
e
Fig. 56: Cymbella species, SEM of valve interior showing the helictoglossae.
a: C. naviculiformis, b: C. proxima, c: C. affinis, d: C. aspera, e: C. helmkei.
3. Result
89
a
b
d
c
e
f
g
Fig. 57: Navicula brockmannii.
a+b: Light micrographs of live individual (a) and cleaned valve (b), c - e: SEM, external view. f: SEM,
internal view. g: SEM, girdle bands
3. Result
90
Both varieties of M. atomus (Fig. 58) had radiate striae, which consists of uniseriate rows of
round areolae. The filiform raphe slit lies in a heavily silicified median costa. The raphe is
slightly curved and the terminal fissures curved to the same side. The two varieties differ in
size and density of striae and areolae. M. atomus var. atomus (Fig. 58 a - c) had a medium size
of length/width =10 µm/4 µm and 20-24 striae/10 µm with approximately 40 areolae/10 µm.
The variety permitis (Fig. 58 d - f) had 35 striae/10µm with approximately 60 areolae/10 µm
and reached a medium size of length/width =7,5/3 µm.
Micrographs of sequenced species that are not present above are shown in the appendix (Figs.
60 – 70).
3. Result
91
a
d
e
b
f
c
Fig. 58: Mayamaea atomus varieties.
a - c: M. atomus var. atomus, d - f: M. atomus var. permitis
a + c: light micrographs of a cleaned valve, b + e: SEM of valve exterior and c + f: SEM of valve
interior.
4. Discussion
92
4. Discussion
The recent taxonomy of naviculoid pennates is basically based on investigations of valve
morphology, cell components and life cycle (e.g., Round et al., 1990). But the development of
the PCR has facilitated the use of DNA sequences for inferring phylogenies and several
studies dealing with diatoms had been carried out (e.g., Medlin et al., 1996 a, b, Medlin et al.,
2000, Kooistra et al., 2003, Medlin & Kaczmarska, 2004, Sorhannus, 2004). With molecular
phylogenetics, the Bolidophyceae were recovered as the sister group to the diatoms (Guillou
et al., 1999) and a revised classification with new subdivisions and classes was proposed
(Medlin & Kaczmarska, 2004). There are several studies, which concentrate on the
relationships of the diatoms with other heterokonta (e.g., Medlin et al., 1997, Guillou et al.,
1999) or on the relationship of the different diatom classes (e.g., Sorhannus et al., 1995), but
there are only two studies with focus on the naviculoid pennates (Beszteri et al., 2001,
Behnke et al., 2004). But in both studies, each genus is only represented by one or two
species. The recent study concentrates on naviculoid pennates. The 91 isolated and sequensed
cultures covered 22 genera and 72 species. 62 of these species belong to the Naviculaceae
covering 16 genera. With the addition of sequences obtained from GenBank the number of
naviculoid species rises up to 66, 76 and 109 in the dataset of rbcL gene, LSU rDNA and SSU
rDNA sequences, respectively.
4.1 Comparison of the gene trees
A gene tree constructed from DNA sequences does not necessarily agree with the true species
tree that represents the actual evolutionary pathway of the species involved. This is well
known from several simulations (e.g., Pamilo & Nei, 1988, Hillis, 1996, Graybeal, 1998) and
empirical studies (e.g., Soltis et al., 1998, Soltis et al., 2000). Most molecular phylogenies of
diatoms based on the SSU rDNA (e.g., Medlin et al., 1996 a, b, Medlin et al., 2000, Kooistra
et al., 2003, Medlin & Kaczmarska, 2004). Those studies based on other gene sequences (e.g.,
Behnke et al., 2004) deal with a different set of taxa, which make it very difficult to compare
the phylogenies. Because the same set of cultures was used for all genes in this study, the
phylogenies based on sequences of the nuclear SSU rDNA, LSU rDNA and the chloroplast
rbcL gene could be easily compared. For the same reason, the sequences could be additionally
analysed in a combined dataset.
For all datasets molecular phylogenies were inferred with maximum likelihood (ML) and
maximum parsimony (MP) analyses. The only exception is the dataset of SSU rDNA
4. Discussion
93
sequences, which contains sequences obtained from GenBank. For this dataset the maximum
parsimony analysis could not be conducted. The time to conduct the MP analysis extremely
increased, because of the large number of species in this dataset.
4.1.1 Phylogenies based on the AlgaTerra cultures
4.1.1.1 Phylogenies based on SSU rDNA sequences
The resulting phylogenetic trees based on the SSU rDNA dataset of AlgaTerra cultures (Figs.
3 – 6) show only few differences. Most relationships between and within the different genera
in the ML tree could be recovered in the MP tree. The polytomies in the MP phylogeny does
not contradict the branching order in the ML tree. Luticola goeppertiana and Neidum affine
form a clade in both gene trees, but this clade diverges at different positions in the two
phylogenies. The very long branches of these two species in the ML tree show that their
sequences differ very much from all other sequences. Especially in MP analyses these rapidly
evolving lineages are inferred to be closely related, regardless of their true evolutionary
relationships or diverge very early in the tree (e.g., Felsenstein, 2004, Salemi & Vandamme,
2003). This phenomenon in phylogenetic analyses is known as “Long Branch attraction”.
Long Branch attraction is most commonly in maximum parsimony analyses but it is also
known for ML or distance methods (e.g., Felsenstein, 2004, Salemi & Vandamme, 2003). The
problem arises when the DNA of two (or more) lineages evolves rapidly. These rapidly
evolving lineages are inferred to be closely related, regardless of their true evolutionary
relationships or diverge very early in the tree. In bootstrap trees, this misinterpretations will
be supported with high bootstrap values. Therefore the close relationship of the two genera
might also be a result of Long Branch attraction. But although their morphology differ clearly
(e.g number of chloroplasts, absence or existence of a stigma, raphe endings) and they were
placed in different families in Round et al. (1990), the two genera belongs to the same
suborder Neidiineae. Because Neidum and Luticola are the only representatives of this
suborder in this phylogeny, their close relationship in the tree might reflect their true relation.
The second clade that changed its position, consisted of Stauroneis, Craticula and Navicula
integra (clade 5 in the ML tree, clade 3 in the MP tree). In both trees, this clade is closely
related to the same two clades, Mayamaea, Eolimna, Pinnularia and Caloneis on one hand
and the monoraphid genera, Navicula brockmannii and the Cymbellales on the other hand. In
the MP tree, Stauroneis, Craticula and Navicula integra are the sister group of the former
clade, in the ML phylogeny the latter is the sister group. Compared to the classification in
Round et al. (1990) the relationship in the MP tree is more likely, because these species were
4. Discussion
94
placed in the same order (Naviculales). On the other hand the order Naviculales could not be
supported by the molecular phylogenies.
The deep divergences had no or only poor bootstrap support, but at lower level (genus to
species) many clades were well supported. Several genera, which were established on the base
of morphological data, could be recovered as monophyletic groups. Most of these groups
were supported by high bootstrap values (≥90). This is true for Fragilaria, Eunotia, Navicula
sensu stricto, Craticula, Cocconeis and Mayamaea. The monophyly of Amphora, Cymbella,
Gomphonema and Placoneis (with Navicula [Placoneis] hambergii, see 4.2.6) had medium or
low bootstrap support and the controversially discussed genera Pinnularia and Caloneis form
a monophyletic clade, which was supported relatively well (for detailed discussion on these
genera see 4.2.3).
Eunotia diverges first after the outgroup, followed by a clade containing Navicula sensu
stricto and Hippodonta capitata. The close relationship of the two genera correspond with the
discussion whether or not to separate Hippodonta from Navicula sensu stricto. The strongly
supported monophyly of Navicula sensu stricto suggest a separation. All other new described
genera that were segregated from Navicula sensu stricto (Craticula, Eolimna, Luticola,
Mayamaea and Placoneis) and all “Navicula” species, that do not belong to the section
Lineolatae (N. integra, N. hambergii and N. brockmannii, for detailed discussion see 4.2.4,
4.2.6 and 4.2.8, respectively) did not cluster with the Navicula sensu stricto. Craticula is most
closely related to Stauroneis and N. integra. The close relationship of Craticula and
Stauroneis agree with the assumption made by Round et al. (1990) and the results of the
phylogenetic analysis of morphological data conducted by Cox and Williams (2000). The
cymbelloid genera (Cymbella, Placoneis, Encyonema and Gomphonema), N. brockmannii and
the monoraphid genera Cocconeis and Achnanthidium form a clade (clade 4). Eolimna
minima and Mayamaea are most closely related to Pinnularia/Caloneis (clade 3). Amphora
diverges at the base of the whole group (clade 2).
4.1.1.2 Phylogenies based on LSU rDNA sequences
The analyses of LSU rDNA alignments resulted in less supported phylogenies (Figs. 11 – 14)
as compared to those based on the SSU rDNA. The tree inferred with the parsimony analysis
using the sequences of the AlgaTerra cultures (Figs. 13 + 14) had several large unresolved
polytomies, but with the exception of the position of N. affine and L. goeppertiana they do not
contradict the branching order in the ML tree (Figs. 11 + 12). In the MP tree, the two species
form a clade, which diverge from the basal polytomy, whereas in the ML tree they diverge
4. Discussion
95
within the Navicula sensu stricto. The integration in the Navicula sensu stricto had no
bootstrap support, whereas the clade consisting of Hippodonta capitata and Navicula sensu
stricto was well supported (BS: 99/76). Therefore and because of the clear morphological
differences, it is unlikely that Neidum and Luticola belong to the Navicula sensu stricto.
Similar to the results of the analyses of the SSU rDNA sequences their close relationship
might represent their true relationship or might be caused by Long Branch attraction.
Although the branching order of the ML tree differs from those of the trees based on SSU
rDNA sequences, several groups could be recovered. Similar to the SSU rDNA gene tree
Amphora, Craticula, Cocconeis, Cymbella, Eunotia, Fragilaria, Gomphonema, Mayamaea
Placoneis (with Navicula [Placoneis] hambergii, see 4.2.6) and Pinnularia/Caloneis form a
monophyletic clade. But only Craticula, Cocconeis, Fragilaria and Mayamaea were
supported by high bootstrap values (≥90). The four cymbelloid genera form a monophyletic
clade with an identical branching order compared to the SSU rDNA gene tree. Craticula and
Stauroneis with N. integra were sister groups and Hippodonta is sister to Navicula sensu
stricto. The close relationship of Eolimna minima, Mayamaea and Pinnularia/Caloneis could
be recovered, even though the branching order of the genera differs. The most unexpected
difference to the SSU rDNA gene tree is that the araphid taxa diverge within the raphid,
although they were assigned as outgroup (see PAUP commands in the appendix). This is
probably, because only one centric diatom was included as outgroup to pull the araphids out
of the raphid diatoms. The ML tree consists of two large clades. In clade 1, Amphora diverged
first followed by Eunotia. Then Hippodonta and Navicula sensu stricto diverges, followed by
the araphid taxa and finally the Cymbellales. In clade 2 Pinnularia/Caloneis and Eolimna
minima form the base and Mayamaea diverges next. The next sub-clade contains N.
brockmannii
and
Cocconeis.
Achnanthidium
minutissimum
is
sister
to
the
Craticula/Stauroneis/N. integra-clade.
4.1.1.3 Phylogenies based on rbcL gene sequences
For the datasets of rbcL sequences the maximum parsimony analyses resulted in a poorly
resolved phylogenetic tree (Figs. 19 + 20). Only Hippodonta capitata and Navicula sensu
stricto and the two araphid genera form clades, which contain two genera. Eunotia, Cocconeis
and Mayamaea were monophyletic. All other genera were merged in a large polytomy. Most
branches in the ML tree had only low bootstrap support (Fig. 21).
Nevertheless, in the ML tree based on rbcL sequences, several clades from the SSU rDNA
gene tree could be recovered. Cocconeis, Eunotia, Fragilaria, Gomphonema, Mayamaea,
4. Discussion
96
Navicula sensu stricto, Placoneis (with Navicula [Placoneis] hambergii, see 4.2.6) and
Pinnularia/Caloneis
form
monophyletic
clades,
again.
Additionally Stauroneis
is
monophyletic because in this gene tree N. integra diverges within Craticula. Three clades
containing the same species but in different branching order could be recovered. The first
clade consists of Eolimna minima, Mayamaea and Pinnularia/Caloneis (clade 3 in Fig. 19).
Stauroneis, Craticula and N. integra form a second recovered clade (within clade 5 in Fig. 19)
and the third clade contains the monoraphid species, N. brockmannii and the Cymbellales
(clade 4 in Fig. 19). But in this gene tree, the Cymbellales are not a monophyletic group,
because Encyonema form a clade with the monoraphid species. But it is unlikely, that this
clade represents the true relationship of Encyonema, because of the strong morphological
support for the Cymbellales and the monophyly of this order in most phylogenies. Analogue
to the phylogenies based on the SSU rDNA sequences, Hippodonta capitata is sister to
Navicula sensu stricto and they diverge close to the base of the tree after Eunotia. The araphid
diatoms diverge next, followed by E. minima, Mayamaea and Pinnularia/Caloneis. Therefore
this is the second gene tree where they diverge within the naviculoid pennates. Luticola
goeppertiana and Neidum affine form a clade that splits the genus Amphora. The branches of
L. goeppertiana and N. affine in the ML tree are not very long. Therefore this clade could not
be caused by Long Branch attraction. Together with Stauroneis, Craticula and N. integra
these species form the sister clade to the group containing the Cymbellales.
4.1.1.4 Phylogenies based on the combined dataset
The analyses of the combined dataset resulted in the best supported trees (Figs. 26 – 29), but
the deep divergences still has only weak bootstrap support. Most relationships of the ML tree
could be recovered in the MP tree. The different positions of the clade containing Luticola
goeppertiana and Neidum affine could be explained by Long Branch attraction. Additionally
the araphid diatoms and the genus Amphora diverged at different positions. In the ML tree the
araphids diverge after Navicula sensu stricto, but in the MP phylogeny they form a clade with
Eunotia and diverge before Navicula sensu stricto. In both trees, Amphora is most closely
related to Craticula/Stauroneis/N. integra, but in the ML tree Amphora diverges before the
divergence of Eolimna minima, Mayamaea and Pinnularia/Caloneis and in the MP tree after
this group.
The ML tree is very similar to the ML tree based on SSU rDNA sequences. Most differences
are found in the deeper divergences. Like in the phylogeny based on the rbcL gene, the
araphid pennates diverge after Eunotia and Hippodonta and Navicula sensu stricto. Similar to
4. Discussion
97
the separate analyses of the LSU rDNA and rbcL gene sequences, the use of a single centric
species might cause the problems to find the real position of the araphid pennates. The next
clade contained the monoraphid taxa, N. brockmannii and the Cymbellales. Then Luticola
goeppertiana and Neidum affine diverge (again with long branches), followed by Amphora.
Craticula, Stauroneis and N. integra forming the sister clade of Eolimna minima, Mayamaea
and Pinnularia/Caloneis.
4.1.1.5 General results of the analyses of the AlgaTerra cultures
The used D1/D2-region of the LSU rDNA comprises more highly variable areas than the SSU
rRNA gene (Van der Auwera & De Wachter, 1998), which makes it even more difficult to
align. A stronger phylogenetic signal for closely related species in comparison with the SSU
rRNA gene and problems for reconstructing deep phylogenies were estimated. The latter
expectation was proven by the MP phylogeny with its large polytomy at the base of the tree.
But in this study the trees based on the LSU rDNA sequences do not provide stronger
supported results for closely related species. Compared to the SSU rDNA gene trees, the
bootstrap values are lower at all levels. Therefore the use of the D1D2-region does not result
in more detailed information of the relationships between the species used in this study
compared to the SSU rDNA.
A part of the rbcL gene was the second sequence additionally used in this study to obtain
clearer information of evolution at lower (order to genus) levels of taxonomic hierarchy in
diatoms. But in the tree resulted from the analyses of the rbcL dataset the bootstrap supports
at all levels were low compared to the SSU rDNA gene trees. It is known that in proteincoding trees the three codon positions evolve at different rates. Therefore, these dataset set
was additionally analysed with differently weighted positions, but the resulting tree differs
only slightly (see Fig. 71 in the appendix). In this study only 684 bp of the rbcL gene, which
has a total length of 1428 – 1434 pb, were used. This might be the reason, that the results fall
short of the expectations.
With the combination of the sequences in a single dataset the information of all genes was
combined. From several studies it is known that an increased number of nucleotides (e.g.,
Saito & Nei, 1986) and the use of different genes that have evolved independently (e.g.,
Pamilo & Nei, 1988). The analysis of the combined dataset should result in trees with an
increased resolution and internal support (as measured by bootstrap values) because the
number of nucleotides increased and the nuclear-encoded rDNA evolved independently from
the plastid-encoded rbcL gene. From other studies it is known that the analyses of combined
4. Discussion
98
data sets run faster times compared to the separate datasets (Soltis et al., 1998, Soltis et al.,
2000). In the recent study, the analyses of the combined dataset ran faster and resulted in trees
with higher bootstrap support than the analyses of the single genes. Especially the divergences
at genus and species level were supported by increased bootstrap values. But the deep
divergences, where the most differences between the different gene trees appeared, still have
only poor bootstrap support. The partition homogeneity test of the combined dataset resulted
in a very low p-value of 0,01. If the test have been used to determine whether or not to
combine data sets for phylogenetic analysis, this p-value denotes separate analyses. But the
resulting best supported tree of the combined analyses in this study agree with other studies,
that have found that P-values < 0,05 should not preclude dataset combination (e.g., Sullivan,
1996, Davis et al., 1998, Flynn & Nedbal, 1998, Yoder et al., 2001).
Most differences between the trees are located at deeper branches. In all phylogenies, the deep
branches had no or only extremely low bootstrap support. Therefore, based on the molecular
data, these divergences could not be resolved unambiguously. Although the trees based on the
different datasets differ, many relationships could be recovered in the analyses of each
dataset. This is a strong support that these relationships in the phylogenetic trees agree with
the true species tree.
Hippodonta capitata diverges at the base of Navicula sensu stricto in all trees. The well
supported monophyly of Navicula sensu stricto in most trees support a separation of the two
genera as promoted by Witkowski et al. (1998) and Round (2001). Although the two genera
appeared as sister groups, the results refutes the idea of Cox (1999, 2002) enlarging the
generic description of Navicula sensu stricto to cover both genera. The results strongly
support the concept of Navicula sensu stricto (Navicula section Lineolatae), because all other
“Navicula” species, that do not belong to the section Lineolatae (N. brockmannii, N.
hambergii and N. integra, for discussion see 4.2.8, 4.2.6 and 4.2.4, respectively) and all new
described genera, that were segregated from Navicula sensu stricto (Craticula, Eolimna,
Mayamaea and Placoneis), did not cluster with the Navicula sensu stricto. This is also true
for Luticola, with the exception of the ML tree inferred using rbcL gene sequences.
As proposed by Round et al. (1990) on the base of plastid behaviour, sexual reproduction and
some aspects of the valve morphology, Craticula is closely related to Stauroneis in all trees.
This also agree with Cox and Williams (2000), who conducted a phylogenetic analysis of
several naviculoid diatoms with a stauros based on morphological data.
4. Discussion
99
Eolimna minima and Mayamaea are closely related to Pinnularia/Caloneis in all trees, but the
branching order of the three genera differs in the different trees. A similar result was found by
Behnke et al. (2004). In their ML phylogeny of SSU rDNA sequences E. minima and several
Sellaphora species form the sister clade to Pinnularia cf. interrupta and Navicula pelliculosa.
Placoneis diverged within the Cymbellales as presumed by Round et al. (1990) based on
frustule and protoplast characters. All genera Round et al. (1990) summarised in the order
Cymbellales, which were present in this study, form a monophyletic clade in the different
phylogenies. But the relationships within this order differ between the phylogenetic trees
inferred in this study and the classification shown in Round et al. (1990). Encyonema was
describes by Kützing (1833) and later added to Cymbella by Cleve-Euler (1948). Round et al.
(1990) restored the genus Encyonema and placed it together with Cymbella, Placoneis,
Brebissonia and Gomphocymbella in the family Cymbellaceae. But in all phylogenetic trees
Gomphonema, which was placed in the family Gomphonemataceae, diverge within the family
Cymbellaceae. This result advises a revision of the involved families Cymbellaceae and
Gomphonemataceae on the base of a detailed morphological and molecular investigation of
all genera. The close relationship of the cymbelloid lineage and the freshwater monoraphid
taxa, which was shown by Medlin and Kaczmarska (2004), could be recovered, although both
studies used a totally different set of taxa.
Eunotia form a monophyletic clade, which diverges at the base of the naviculoid pennates in
most trees. This contradicts the position of this genus found by Medlin and Kaczmarska
(2004), where Eunotia diverges between two clades containing naviculoid taxa. But the
results of other analyses (Medlin et al., 2000, Sorhannus, 2004) in which the Eunotiales fell at
the base of all raphid diatoms, are supported.
In all trees, the position of N. affine and L. goeppertiana differs. The long branches of this
species in both rDNA ML trees indicate that the rDNA evolves more rapidly in these species.
Especially for the sequences of L. goeppertiana this was obvious in the alignment, because of
the large insertions. That these species belong to Navicula sensu stricto as it is shown in the
LSU rDNA gene tree inferred with ML is refused by the mainly well supported monophyly of
Navicula sensu stricto in all other trees. Additionally the valve morphology of both species
deviates from the generic description of Navicula sensu stricto. Beside other differences in
both species the raphe structure does not fit to the generic description of Navicula sensu
stricto (Round et al., 1990). The two species also form a clade in the rbcL gene trees, where
they do not have long branches and in the trees based on the combined data set this clade is
4. Discussion
100
well supported (BS: 99/97). Because the two genera are the only representatives of the
suborder Neidiineae, the clade reflects this relationship. But it should be expected, that the
relationship of the two genera is more distant, than the trees in this study show.
4.1.2 Phylogenies based on enlarged datasets
4.1.2.1 Phylogenies based on SSU rDNA sequences
The ML tree (Figs. 7 – 10) based on the enlarged dataset with additional sequences obtained
from GenBank shows a similar relationship of the genera compared to the tree based on the
smaller dataset. Similar to this phylogeny the deeper divergences had very low bootstrap
support.
Most diatom sequences available at GenBank are sequences of the SSU rDNA. From the huge
amount of available sequences, I choose all naviculoid pennates, several Bacillariales and
Eunotia species. The nomenclatures of two of these additional sequences were obviously
wrong: AY485521 Stauroneis constricta must be a Fragilariopsis species and AY485496
Achnanthes sp. must be an Amphora species, because both species belongs to monophyletic
clades with maximum bootstrap support.
Eunotia form a monophyletic clade, which diverges at the base of all raphid pennates. The
enlarged dataset contains naviculoid and nitzschioid taxa and the phylogeny contradicts the
position of Eunotia found by Medlin and Kaczmarska (2004) and support the results of
Medlin et al. (2000) and Sorhannus (2004). This result also agree with the classification in
Round et al. (1990), in which the Eunotiaphycidae (contain Eunotia and related species) and
the Bacillariophycidae (contain all ather raphid pennates) were combined in one class.
The Bacillariales form a clade with the marine Achnanthes species, which diverges between
Undatella sp. and the other naviculoid pennates. This separation of monoraphid genera
contradict the order Achnanthales, as mentioned in Round et al. (1990). In contrast to the tree
of Medlin and Kaczmarska (2004) the naviculoid pennates form a monophyletic clade, with
the exception of Undatella sp. This clade was subdivided into four sub-clades, of which only
the first one is supported by bootstrap analysis (BS: 66).
The first sub-clade (naviculoid pennates part 1, Figs. 7 + 8) contains Haslea, Gyrosigma and
Pleurosigma as sister to Hippodonta, Navicula sensu stricto plus Pseudogomphonema.
Equivalent to the phylogenies based on sequences from AlgaTerra cultures, H. capitata
diverges at the base of the Navicula sensu stricto. A close relationship of the two genera
Pseudogomphonema and Navicula sensu stricto was already proposed based on
4. Discussion
101
morphological analyses (e.g. Medlin & Round, 1986). The molecular data suggest that
Pseudogomphonema should not be separated from Navicula sensu stricto, although these two
genera differ in their valve symmetry. It is clear that the asymmetry is a derived character
from within the Navicula sensu stricto. The close relationship of Gyrosigma and Pleurosigma
agree with the placement in one family by Round et al. (1990). All genera in this sub-clade
belong to the suborder Naviculineae sensu Round et al. (1990), but not all genera summerized
in this subgenus by Round et al. (1990) appeared in this sub-clade.
Haslea nipkowii did not cluster with the other Haslea species in the first sub-clade, but form a
clade with Neidum affine. But in Damsté et al. (2004) and Poulin et al. (2004) the genus is
monophyletic and the affiliation of H. nipkowii is also suppoted by morphological and
biochemical data.The result of the recent study might be caused by Long Branch attraction.
Clade 1 of the sub-clade naviculoid pennates part 2 (Fig. 9) contains Amphora subgenus
Amphora as sister to Phaeodactylum tricornutum and Luticola goeppertiana as sister to
Diadesmis gallica. The close relationship of Luticola and Diadesmis agree with the
combination of the two genera in the family Diadesmidaceae by Mann (in Round et al.,
1990). But Amphora and Phaeodactylum were placed in different orders by Round et al.
(1990). All monoraphid genera, with the exception of the marine Achnanthes species, could
be found in clade 2. They form two clades with Cocconeis and Planothidium on one hand and
Pauliella and Achnanthidium on the other hand. The divergence in two groups agree with the
separation of two families in Round et al. (1990), although Planothidium and Pauliella were
not mentioned there. But that the monoraphids diverge within several genera placed in the
Naviculales contradict the seperation of these genera in an other order (Achnanthales). Similar
to the phylogenie based on the SSU rDNA sequences of AlgaTerra cultures, the genera
Cymbella, Encyonema, Gomphonema and Placoneis form a monophyletic clade, only the
relationships between these genera differs in the two phylogenies. Bur again the families
Cymbellaceae and Gomphonemataceae could not be recovered in the molecular phylogeny. In
this phylogeny Anomoeoneis, which also belongs to the order Cymbellales (Round et al.,
1990), and Lyrella, which belongs to the order Lyrellales (Round et al., 1990), form the sister
clade of the other Cymbellales. Anomoeoneis and Lyrella were each represented by only one
species and their clade had only low bootstrap support. Therefore the result of this analysis
does not suffice to suggest any chages in the classification of them. A close relationship of the
two orders was also found in the study conducted by Behnke et al. (2004).
4. Discussion
102
The sub-clade naviculoid pennates part 3 (Fig. 10) contains all Amphora species of the
subgenus Halamphora, but they did not form a monophyletic clade (for detailed discussion of
relationships within the genus Amphora see 4.2.2). This dataset contains one species
(Undatella sp.) that is assumend to be closely related to Amphora by Round et al. (1990). But
none of the Amphora species is closely related to Undatella sp.. Most species in clade 1 of
this sub-clade belong to the suborder Sellaphorineae (sensu Round et al., 1990). Based on this
phylogeny the genus Mayamaea, which was not mentioned in Round et al. (1990), should be
placed to the same suborder. That Mayamaea forms the sister clade to Pinnularia/Caloneis
suggests the addition of Mayamaea to the family Pinnulariaceae. The well supported
monophyly of Eolimna minima and Sellaphora agree with Behnke et al. (2004), who
concluded that E. minima “could be regarded as belogning to the Sellaphoraceae, or even to
Sellaphora itself “ (p. 206). In clade 2 Navicula pelliculosa, N. saprophila, Stauroneis/N.
integra, Eolimna subminuscula and Craticula form a strongly supported monophyletic clade
(BS: 100). N. pelliculosa and N. saprophila does not belong to Navicula sensu stricto and the
results support the separation of these species from the genus. But the position of N.
pelliculosa contradicts the assumption of Round et al. (1990) and Behnke et al. (2004) that
this species belongs to Sellaphora or at least to the suborder Sellaphorineae. Based on the
recent phylogeny N. pelliculosa, N. saprophila and E. subminuscula belong to the family
Stauroneidaceae together with Craticula, Stauroneis and N. [Prestauroneis] integra. The
sister clade to the Stauroneidaceae consists of a highly supported monophyletic Amphora
subgenus Halamphora clade and a second clade containing several Surirellales.
As described above, this phylogeny supports several classifications made in Round et al.
(1990) and several families as well as suborders and orders could be recovered as
monophyletic clades. But the results clearly show, that the order Naviculales as described in
Round et al. (1990) is a heterogenous group.
4.1.2.2 Phylogenies based on LSU rDNA sequences
Although only nine sequences could be obtained from GenBank the trees resulting from the
analyses of the enlarged LSU rDNA dataset differs strongly from those inferred with the
smaller dataset. Some relationships between closely related genera, such as the monophyly of
the Cymbellales, could be recovered. But the branching order of the different groups differs
strongly from those found in all other trees. These deep divergences have no support by
bootstrap values and could not be explained by morphological data. Therefore the use of the
D1/D2-region of the LSU rRNA gene appears problematic.
4. Discussion
103
Similar to the the tree based on the enlarged SSU rDNA dataset, the Baccillariales form a
monophyletic clade. But in the phylogenies based on the LSU rDNA sequences this clade
diverge within the naviculoid diatoms. With the addition of Amphora coffeaeformis, the
seperation of the two subgenera of Amphora becomes more obvious. Similar to its position in
the tree based on the enlarged SSU rDNA dataset, Phaeodactylum diverges close to Amphora.
4.1.2.3 Phylogenies based on rbcL gene sequences
The enlarged dataset of rbcL sequences contains 15 additional sequences obtained from
GenBank. The two phylogenies (Figs. 22 – 25) inferred with this dataset show great
differences in the deep branches, compared with each other and with the trees based on the
rbcL sequences from the AlgaTerra cultures. The resolution of the MP tree based on the
enlarged dataset is better compared to the MP tree based on the rbcL sequences from the
AlgaTerra cultures, but it still contains several unresolved polytomies.
The additional species of the genera Encyonema, Eunotia, Gomphonema and Placoneis form
monophyletic clades with the other species of their genera. Lyrella form the sister clade to
Petroneis (clade 2), which agree with the family Lyrellaceae erected by Mann (in Round et
al., 1990). Similar to the tree based on the enlarged SSU rDNA dataset, they are relativly
close related to the monoraphid species and the Cymbellales in the ML tree. But the diverge
before the entire group. Equivalent to the tree based on the enlarged SSU rDNA dataset,
Sellaphora forms a monophyletic clade with Eolimna minima and is closely related to
Mayamaea and Pinnularia/Caloneis. Navicula sensu stricto is paraphyletic, because of N.
salinicola. Both phylogenies suggest that Pseudogomphonema and Seminavis should not be
separated from Navicula sensu stricto. The three genera also share several morphological
feature, like uniseriate striae containing apically elongate, slit-like poroids or the raphe
structure, with simple, straight internal central raphe endings, expanded external central raphe
endings and internal raphe fissures, that open laterally (e.g., Medlin & Round, 1986, Round et
al., 1990, Danielidis & Mann, 2002). For Seminavis it is additionally known, that apart from
creating an asymmetrical shape of the vegetative cell, almost all characteristics exhibited by
the live cell and auxospores agree with what is found in Navicula sensu stricto (e.g., Mann &
Stickle, 1989, Chepurnov et al., 2002).
4.1.3 General relationships of the genera
The phylogenies based on the different datasets differ. Especially the branching order of the
early divergences differs strongly and could not be resolved in this study. The results show,
4. Discussion
104
that it could be difficult to detect the relationships of genera, which are represented by only a
single species. They diverge within closly related genera (e.g., Pseudogomphonema) or
change their position in the different phylogenies, especially if their DNA evolves rapidly
(e.g., Luticola, Neidum). But several relationships on different levels could be determined
based on all or at least most trees.
The results of this study support the monophyly of the genera Cocconeis, Craticula,
Cymbella, Eunotia, Gomphonema, Mayamaea, Navicula sensu stricto and Placoneis (with N.
hambergii, see 4.2.6), because they form monophyletic clades in all or at least most trees. A
monophyletic group containing both, Caloneis and Pinnularia, is also supported.
Additionally, in all phylogenies based on enlarged datasets containing two or more species of
these genera, Encyonema, Lyrella, Pleurosigma and Sellaphora (with Eolimna minima) are
monophyletic. But the monophyly of the genus Amphora is rejected (see 4.2.2).
Navicula sensu stricto and Hippodonta capitata are sister groups in most phylogenies. The
close relationship of the two genera correspond with the discussion whether or not to separate
them. The strongly supported monophyly of Navicula sensu stricto, which appear in most
phylogenies, approve a separation. In all phylogenies based on enlarged datasets containing
sequences from Pseudogomphonema or Seminavis species, these species diverge within
Navicula sensu stricto. In contrast to Navicula sensu stricto Pseudogomphonema and
Seminavis exhibit asymmetrical valves. On the other hand all three genera share several
morphological features, like apically elongate, slit-like poroids, the raphe structure or the two
plastids, lying along each side of the girdle. This suggests, that different valve symmetry
alone does not approve separating genera. Reichardt (1992) came to the same result while
comparing the morphology of Navicula sensu stricto and Rhoikoneis.
The genera Craticula and Stauroneis, which were summarised in the family Stauroneidaceae
by Mann (in Round et al., 1990) and appear as close relatives in a phylogenetic analysis based
on morphological data (Cox & Williams, 2000), are found to be close relatives in the recent
study, too. The results also suggest to add Navicula [Prestauroneis] integra to this family,
because this species is associated with this genera in all phylogenies.
The close relationship of Pinnularia/Caloneis and Sellaphora/Eolimna minima as proposed
with the suborder Sellaphorineae by Mann (in Round et al., 1990) could be recovered in most
phylogenies. The results also support to include the genus Mayamaea to this suborder.
The recent study support the monophyly of the order Cymbellales erected by Mann (in Round
et al., 1990). But the results contradict the arrangement of the families Cymbellaceae and
4. Discussion
105
Gomphonemataceae, because in most trees Gomphonema (Gomphonemataceae) diverge
within the Cymbellaceae.
The order Naviculales and the suborder Naviculineae as used in Round et al. (1990) are
shown to be heterogenous in all trees.
The monoraphid genera diverge within the naviculoid pennates in all phylogenies. They are
close relatives of Navicula [Adlafia] brockmannii in most trees and and diverge at the base of
the Cymbellales in several phylogenies, but the relationship between the monoraphid genera
and the naviculoid pennates could not be resolved unambiguously.
4.2 Relationships within the genera
4.2.1 Navicula sensu stricto
J. B. M. Bory de Saint-Vincent (1922) described the genus Navicula based on N. tripunctata
(O.F.Müller) Bory. In the beginning all diatoms with a central raphe on both valves that lack
other light microscopic characteristics of the frustule were assigned to this genus. But with
further investigations, the morphological diversity of the genus became apparent. Today, it is
widely accepted that Navicula (sensu stricto) should be used only for species that belong to
the section Lineolatae (sensu Cleve, 1895 and Hustedt, 1930).
This study confirms the assumption that the genus Navicula sensu lato is a very
heterogeneous group and the results support the monophyly of Navicula sensu stricto. All
“Navicula” species, that do not belong to the section Lineolatae (sensu Cleve, 1895 and
Hustedt, 1930) did not cluster with the Navicula sensu stricto (see 4.1.3). In the molecular
phylogenies the Navicula sensu stricto are divided into three sub-clades, but the
morphological investigations shows no obvious differences between these sub-clades.
Therefore a further separation of this genus is not reasonable. But this result does not
contradict Witkowski et al. (1998). Based on morphological investigations of freshwater and
marine Navicula sensu stricto species, they reasoned that Navicula sensu stricto is still a
heterogeneous group and distinguishes six different groups. But the five groups, which were
segregated from Navicula sensu stricto, contain mainly marine and few brackish-water taxa.
Therefore none of these taxa is part of this study.
4.2.2 Amphora
The genus Amphora was described by Ehrenberg in 1844 (in Kützing, 1844). The genus
embraced all species whose raphe systems of both valves lie on the same side of the cell. That
4. Discussion
106
Amphora is an artificial genus has been known for over 100 years and Cleve (1895) subdivide
the genus into six subgenera. Three of these subgenera contain freshwater species: Amphora,
Halamphora and Oxyamphora. But this subdivision did not induce the creation of new genera
from Amphora. Only in 1990, did Mann establish the genus Seminavis (in Round et al., 1990),
which covered several marine species previously assigned to Amphora. The species cultured
within the scope of this study belong to the subgenera Amphora (A. libyca, A. pediculus, A. cf.
fogediana and the unidentified Amphora species) and Halamphora (A. normannii). Most
species whose sequences were obtained from GenBank are assigned to the subgenus
Halamphora and only A. cf. proteus belongs to Amphora subgenus Amphora. The most
obvious morphological difference between the two subgenera is the organisation of the girdle.
The girdle of the species belonging to the subgenus Amphora consists only of the valvocopula
(Fig. 36 c, d, f and Schoeman and Archibald, 1986: Figs. 70 – 86), whereas the girdle of the
species belonging to the subgenus Halamphora contains additionally numerous girdle bands
(Fig. 35 c and Krammer and Lange-Bertalot, 1986: Fig.151: 1 – 6, 18 – 27).
The results of the phylogenetic analyses support a partition of the genus Amphora. In all
phylogenies those species, which belong to the subgenus Amphora, formed a monophyletic
clade with maximum bootstrap support. In most phylogenies based on the sequences of the
AlgaTerra cultures (Figs. 3, 5, 12, 13, 26 and 27) A. normannii diverged first in a
monophyletic clade of all Amphora species. Although this monophyly is supported by
bootstrap values up to 92, the branch length in the ML trees (Figs. 3, 12 and 26) indicated a
separation. In all trees based on rbcL sequences (Figs. 19, 21, 22 and 24), A. normannii and
the subgenus Amphora did not form a monophyletic group. An explicit separation of the two
subgenera occurs with the addition of SSU rDNA and LSU rDNA sequences obtained from
GenBank. In the consensus tree inferred with the parsimony analysis based on LSU rDNA
sequences, the six Amphora species formed a monophyletic clade (Figs. 17, 18). But the
bootstrap support for this clade was relatively low (53/49), whereas the monophyly of each
subgenus was supported by high bootstrap values (>95). In the ML phylogeny (Figs. 15, 16)
of this alignment the two subgenera were separated by Entomoneis. In the ML phylogeny of
the SSU rDNA sequences (Figs. 7 - 10), the two subgenera appeared in two different clades.
The subgenus Amphora was most closely related to Phaeodactylum, Diadesmis and Luticola
(Fig. 9). The subgenus Halamphora did not form a monophyletic group (Fig. 10). A.
coffeaeformis was most closely related to Rossia, Eolimna minima and Sellaphora, but most
species of this subgenus formed a sister group to several Surirellales. In fact, the sequence of
A. coffeaeformis missed 29 bases at the beginning and 23 bases at the end of the sequence, but
4. Discussion
107
these were relatively conserved regions. Therefore this could not be the reason for the clear
separation of this species from the subgenus Amphora. The GenBank sequence AY485496
belongs definitely to an Amphora species and not an Achnanthes species. As mentioned above
(see 4.1.2) there must be a contamination or confusion somewhere.
Molecular and morphological data strongly support a separation of the species belonging to
the subgenus Halamphora from the genus Amphora. But further investigations on the
subgenus Halamphora is needed because the results of the analysis from SSU rDNA
sequences indicate that this is still an artificial group.
4.2.3 Caloneis and Pinnularia
As pointed out in the introduction, the separation of the two genera is controversial because
the morphological distinction of Pinnularia and Caloneis is very problematic. Round et al.
(1990) and Mann (2001) doubted the correctness of the traditional Pinnularia-Caloneis
distinction. Based on her investigation on live material, Cox (1988 b) proposed three new
groups: (1) Caloneis silicula, Caloneis bacillum and Pinnularia isostauron; (2) Caloneis
based on C. amphisbaena and (3) Pinnularia based on P. nobilis. Krammer & Lange-Bertalot
(1985) interpreted a different separation, based on the formation of the internal alveoli
aperture: (1) all species whose alveoli are internally nearly open, as existing in Pinnularia
interrupta; (2) species with partially closed alveoli, e.g., Caloneis amphisbaena and
Pinnularia gibba; (3) species with nearly closed alveoli, like Caloneis silicula. Nevertheless
they preferred the traditional Pinnularia-Caloneis distinction.
This study also rejects this traditional distinction of the two genera, because in none of the
trees did Pinnularia or Caloneis form separated monophyletic groups. But in most
phylogenetic trees, the two genera were merged into a monophyletic clade. In all trees two
groups consisting of several Pinnularia species appeared, which could be characterised by
different valve morphology.
One group contained P. obscura, P. anglica, P. mesolepta, P. subcapitata and P.
microstauron. P. interrupta, whose SSU rDNA sequences was obtained from GenBank,
belongs to this group, too. All species in this group have a filiform or slightly lateral raphe
system, where the external raphe slit never crosses the internal slit. Midvalve, the striae are
radial or parallel and become convergent or parallel at the apices. All species have a large
central area, which often form a fascia extending to the margin on one or both sides. Whereas
P. subcapitata, P. microstauron and P. interrupta have two plate-like chloroplasts, P.
obscura, P. anglica and P. mesolepta have a single H-shaped plastid. Although the three
4. Discussion
108
species exhibit a single chloroplast were very close related, the number of plastids could not
alone be used as distinctive feature.
The second group consisted of P. rupestris, P. viridis, P. cf. viridiformis and P. viridiformis.
All species in this group have a lateral raphe system and in some species the undulate external
raphe slit crosses the internal slit several times. The linear or linear-lanceolate axial area
enlarges into a slightly expanded central area, which is often asymmetric. But in contrast to
the first group, the central area does not form a fascia. The striae are radial or parallel at the
centre of the valve and become convergent or parallel at the apices. Under the light
microscope the striae were crossed by a lateral line, which is caused by partially closed
alveolae. All species have two plate-like chloroplasts.
The position of P. acrosphaeria and of the three Caloneis species differs in the different
phylogenetic trees. Based on the valve morphology, P. acrosphaeria is more closely related to
the second group. With this group, P. acrosphaeria shares partially closed alveoli and the
absence of a fascia. But only the ML phylogenies based on LSU rDNA sequences support this
relationship. In most trees, P. acrosphaeria diverged at the base of the first group. In the MP
phylogenies based on the rDNA sequences or the combined data set P. acrosphaeria and C.
amphisbaena form a clade. In the ML phylogenies based on these data, the two species had
very long branches in contrast to the other species. Therefore this grouping might be caused
by Long Branch attraction. Our data does not clarify the position of P. acrosphaeria. The
species might belong to the second group, which is supported by the morphology and the ML
phylogenies based on LSU rDNA sequences. But it could also diverge at the base of the first
group or be the only representative of a sister group.
In most phylogenetic trees, C. amphisbaena belongs to the second group. This is also
supported by valve morphology because C. amphisbaena shares partially closed alveoli and
the absence of a fascia with this group. Therefore it is possible that C. amphisbaena should be
included in this group.
C. budensis shows morphological features of both groups. Like the species belonging to the
first group, C. budensis has a fascia. But the alveolae are partially closed, which is a typical
feature of the species in the second group. C. lauta shows the same character combination. In
most phylogenies, the two taxa diverge early within the Pinnularia/Caloneis clade. In the ML
tree based on SSU rDNA sequences, the two species form a sister clade to the two other
groups. But in most trees they diverge independently, often at the base of one or the other
4. Discussion
109
groups. The results indicate that the two Caloneis species belong to an additional group,
which might be primary within the Pinnularia/Caloneis clade.
These molecular results support the groups defined by Krammer & Lange-Bertalot (1985).
The first group containing P. obscura, P. anglica, P. mesolepta, P. subcapitata and P.
microstauron. P. interrupta is equivalent to their group 1, which includes all species whose
alveoli are internally nearly open, as existing in Pinnularia interrupta. The second group
consisted of P. rupestris, P. viridis, P. cf. viridiformis, P. viridiformis and C. amphisbaena is
identical with their group 2, which contains species with partially closed alveoli, e.g.,
Caloneis amphisbaena and Pinnularia gibba. C. budensis and C. lauta represent typical
Caloneis species, which are the members of their group 3.
4.2.4 Navicula integra
N. integra is not a member of Navicula sensu stricto and Mereschkowsky (1903) include this
species in the genus Placoneis. But the species was not yet renamed and Cox (1987) doubted
the correctness of this combination because N. integra did not have the kind of chloroplast
typical for this genus.
In all phylogenetic trees shown above the species is clearly separated from Navicula sensu
stricto and Placoneis. With the exception of the two MP phylogenies based on the rbcL
sequences N. integra formed a monophyletic group with Craticula and Stauroneis. In most
trees, this grouping had strong bootstrap support. The position of N. integra within this clade
differs. In most trees, the species appears within or at the base of Stauroneis, but in the
phylogenies inferred with ML analyses using rbcL sequences N. integra diverged within
Craticula. N. integra shares several morphological features with Craticula and Stauroneis,
such as number, form and position of the chloroplasts, the formation of the raphe and the
composition of the girdle. But the morphology of N. integra also prohibits its inclusion into
one of these genera. Parallel and equidistant striae with longitudinal aligned areolae forming
straight lines parallel to the raphe system are typical for the genus Craticula. The striae of N.
integra are radiate and at the centre of the valve more distant with thickened costae separating
them. This produces a stauros-like structure. But the species has no stauros, which is the most
defining feature of the genus Stauroneis. Hustedt (1961-1966) placed N. integra in Navicula
section Microstigmaticae. Other species of this section were transferred to the genera
Parlibellus (Cox, 1988 a) and Proschkinia (Karayeva, 1978). Based on the morphological
investigations an affiliation of N. integra to either of these genera could be refuted. For
4. Discussion
110
instant, Parlibellus and Proschkinia have a wide girdle region with numerous girdle bands but
N. integra is wider transapically than pervalvarly.
It must therefore be describes a new genus, for which the name Prestauroneis has been
chosen.
Prestauroneis Bruder, gen. nov.
Type species: Prestauroneis integra (W. Smith) Bruder comb. nov. (Figs. 44 d-f and 45 c-d)
Basionym: Pinnularia integra W.Smith (1856, p. 96).
Synonym: Navicula integra (W.Smith) Ralfs (in: Pritchard, 1861, p. 895).
The two plate-like chloroplasts lie one against each side of the girdle. The frustules were
isopolar and tend to lie in valve view, because they are wider transapically than pervalvarly.
The valves are lanceolate or lanceolate-elliptical with subrostrate apices and an additional
undulation in the valve margin before the apices. Pseudosepta are at the apices. The striae are
radial at the centre of the valve and become nearly parallel at the apices. They are uniseriate
and consist of small round or elliptical poroids, which were occluded by hymenes at their
internal aperatures. Midvalve the striae are more distant and the costae separating them are
thickened, producing a stauros-like structure.The external central raphe endings are expanded
and the well developed terminal fissures at the poles curve off to the same side of the valve.
The internal central raphe endings are simple and slightly curved. The girdle composed of
several open, porous bands with one or two rows of small round poroids.
4.2.5 Gomphonema
Whereas the phylogenetic trees show clearly that the genus Gomphonema and the genera
Cymbella, Placoneis and Encyonema were near relatives, some relationships within the genus
are ambiguous. G. acuminatum and G. truncatum formed the only constant group in all trees.
The position of the other species within this genus differs between the different phylogenies.
The position of G. micropus changed most in the different phylogenies. With exception of the
tree in Fig. 21, the other species were always within a well supported monophyletic clade. G.
micropus diverged at the base of the genus in most trees. But it was also found in the middle
of the genus in the trees based on rbcL sequences or separated from the genus by Cymbella in
Fig. 6. That G. micropus belongs to the monophyletic group in most phylogenies and the
position of this species in the trees based on rbcL sequences support the monophyly of the
whole genus. On the other hand, the separation of G. micropus and the other Gomphonema
species by Cymbella in one tree indicate a division. The long branches between G. micropus
4. Discussion
111
and the rest of the genus in most ML trees and the low bootstrap values for the whole group
indicate a separation, too. The morphological investigations of the Gomphonema species
result in only one feature, which could be found exclusively in G. micropus. In this species
the external openings of the areolae form small round poroids, whereas they are C- or kidneyshaped in all other Gomphonema species. This could be interpreted as a reason to separate the
genus in two groups. But it could also be the basic form of the feature, which appeared at the
base of the genus and evolved to the characteristic found in the other species.
On the base of these results a separation of the genus could not be proposed. With the
exception of G. micropus the monophyly of this genus could be clearly shown. To resolve the
relationship of G. micropus to the other species further investigations with additional species
are necessary.
4.2.6 Placoneis and Navicula hambergii
Although it was already known that N. hambergii does not belong to Navicula sensu stricto
(e.g., Krammer and Lange-Bertalot, 1986), the species has not yet renamed. Only Metzeltin et
al. (2004, p. 8) noted that “Navicula hambergii belongs very probably to Placoneis”.
The phylogenetic trees generated in the recent study show clearly that N. hambergii does
belong to the genus Placoneis because it diverged at the base of or within the genus in most
trees. Altogether the monophyly of N. hambergii and Placoneis was well supported, although
this was not found in all phylogenies. In the four trees inferred with the parsimony analysis
based on LSU rDNA and rbcL sequences (Figs. 13, 17, 21, 24) the genus Placoneis is not
monophyletic. In two trees of these trees (Figs. 13, 21) N. hambergii and the Placoneis
species diverge from a polytomy. As discussed above this is the result of the relatively few
parsimony informative positions. Mereschkowsky described the genus Placoneis in 1903 and
used P. exigua as a typical species. With this genus he separated a group of species from
Navicula sensu lato, which have a single, asymmetrical chloroplast. Cox (1987) re-erected the
genus and chose P. gastrum as type species, because “delineation and nomenclature of P.
exigua are confused” (Cox, 1987, p. 153). In the same paper and a second investigation (Cox,
2003) she adds several morphological features from SEM investigations to the description of
the genus. One of the most important features of the genus Placoneis is the single chloroplast
with a central bridge and lateral lobes, which lies under the valves. The cells are symmetrical
and parallel or elliptical sided in their central region. The striae are radiate near the centre of
the valve, becoming more parallel at the apices. They are composed of small round poroids,
which were internally closed by volae. The usually straight raphe slits lie in a narrow axial
4. Discussion
112
area. Externally, the central raphe endings are straight and slightly expanded and the polar
raphe endings curve to the same side. The internal central raphe endings are usually deflected
to the same side and at the internal polar end small helictoglossae are present. All these
features were found in N. hambergii.
Because of the results of the molecular and morphological analyses a new combination must
be made:
Placoneis hambergii (Hustedt) Bruder comb. nov. (Fig. 52 h-o)
Basionym: Navicula hambergii Hustedt (1924, p 562, pl. 17: fig. 2).
4.2.7 Cymbella
The molecular phylogenies showed different relationships within the genus Cymbella. In most
trees they form a monophyletic clade, but in the phylogenies based on the rbcL sequences C.
proxima is separated. In some trees, C. naviculiformis and C. proxima form a sub-clade within
the monophyletic clade, but in other trees only C. naviculiformis is separated from the other
species. The morphological investigations show no constant feature which support a
separation of C. proxima, but C. naviculiformis shows obvious differences. This corresponds
with Krammer’s (1982) subgenera Cymbella and Cymbopleura. Because of the different
results in the molecular phylogenies, which had only relatively low bootstrap supports, this
study does neither support nor refuse a separation of the subgenera.
4.2.8 Navicula brockmannii
It was already known that N. brockmannii was not a member of Navicula sensu stricto (e.g.,
Krammer and Lange-Bertalot, 1986), but the species was not yet renamed.
In all phylogenetic trees shown above, the species is clearly separated from Navicula sensu
stricto. But N. brockmannii does not belong to one of the genera present in the tree because it
never diverges within another genus. In most trees the monoraphid genera and the
Cymbellales were the nearest relatives. The only exceptions were the phylogenies based on
LSU rDNA sequences. The morphological investigations show, that N. brockmannii does not
belong to any of the genera present in this study. But the morphology of this species fits well
to the diagnosis of the recently established genus Adlafia (Moser et al., 1998). The species in
under 25 µm long. The valve has a linear outline and broad rostrate or subcapitate ends. The
raphe is filiform with scarcely expanded central pores and the terminal fissures are strongly
deflected laterally. The axial area is linear and narrow and slightly widened close to the
central area, which is variable in size and form but not widening to the margins. The striae are
4. Discussion
113
dense (25 – 32/10µm) radiate, getting parallel towards the poles, but in contrast to the
diagnosis of Adlafia the direction does not change abruptly. They run continuously from the
valve surface down onto the mantle and consists of rows of round areolae, which where
externally closed by hymenes. The girdle bands have a uniseriate or biseriate row of areolae.
Because there is only a minor difference between the morphology of N. brockmannii and the
diagnosis of the genus Adlafia, I transfer N. brockmannii to the genus:
Adlafia brockmannii (Hustedt) Bruder comb. nov. (Fig. 57)
Basionym: Navicula brockmannii Hustedt (1934, p. 382, fig. 11).
4.2.9 Varieties of Mayamaea atomus
Although the two varieties formed a strongly supported monophyletic clade in all
phylogenetic trees generated in the course of this study, the difference between the sequences
of the two varieties is relatively large. This is most obvious in the phylogeny inferred with the
ML analysis using the combined dataset (Figs. 26, 27). The two varieties are more distant to
each other than for instant the well defined species belonging to Amphora subgenus Amphora.
In the morphological investigations differences size and density of striae and areolae were
detected. The smaller M. atomus var. permitis showed a higher density of striae and areolae.
In our cultures these differences were consistent, but Mayama and Kobayasi (1988) found
continuity in the size and striation density for their Japanese populations. Based on these
results and the absence of any ecological differences they reject a separation of the two types.
In contrast to the findings of Mayama and Kobayasi (1988) the comparison of the sequences
indicates that M. atomus var. atomus and M. atomus var. permitis were not just two varieties
of the same species but two different species. Mayama and Kobayasi (1988) did not observe
the density of the areolae. Therefore this might be the feature for the differentiation of the two
forms. To clarify this problem further molecular and morphological investigations including
the Japanese populations are necessary.
References
114
References
Andersen, R.A., Morton, S.L. and Sexton, J.P. (1997). CCMP - Provasoli-Guillard
National Center for Culture of Marine Phytoplankton. Journal of Phycology Supplement
33: 1-75.
Behnke, A., Friedl, T., Chepurnov, V. A., and Mann, D. G. (2004). Reproductive
compatibility and rDNA sequence analyses in the Sellaphora pupula species complex
(Bacillariophyta). Journal of Phycology 40 (1): 193-208.
Beszteri, B., Acs, E., Makk, J., Kovacs, G., Marialigeti, K., and Kiss, K.T. (2001).
Phylogeny of six naviculoid diatoms based on 18S rDNA sequences. International
Journal of Systematics and Evolutionary Microbiology 51 (4): 1581-1586.
Bory de Saint-Vincent, J.B.M. (1922). Dictionaire classique d'Histoire Naturelle.
Bull, J.J., Huelsenbeck, J.P., Cunningham, C.W., Swofford, D.L. and Waddell, P.J.
(1993). Partitioning and combining data in phylogenetic analysis. Systematic Biology 42
(3): 384-397.
Cavalier-Smith, T. (2004). Only six kingdoms of life. Proceedings of the Royal Society of
London, Series B, Biological Sciences 271: 1251-1262.
Chepurnov, V.A., Mann, D.G., Vyverman, W., Sabbe, K. and Danielidis, D.B. (2002).
Sexual reproduction, mating system, and protoplast dynamics of Seminavis
(Bacillariophyceae). Journal of Phycology 38: 1004-1019.
Clegg, M.T. (1993). Chloroplast gene sequences and the study of plant evolution.
Proceedings of the National Academy of Sciences of the United States of America 90:
363-367.
Cleve, P.T. (1894). Synopsis of the naviculoid Diatoms. Part 1. Kongliga Svenska
Vetenskaps-Akademiens Handlingar 26, 1-194.
Cleve, P.T. (1895). Synopsis of the naviculoid Diatoms. Part 2. Kongliga Svenska
Vetenskaps-Akademiens Handlingar 27: 1-219.
Cleve-Euler, A. (1948). Süsswasserdiatomeen aus dem Feuerland. Acta Geographica
(Societas Geographica Fenniae) 10 (1): 1-61.
References
115
Cox, E.J. (1987). Placoneis Mereschkowsky: the re-evaluation of a diatom genus originally
characterized by its chloroplast type. Diatom Research 2: 145-157.
Cox, E.J. (1988 a). Taxonomic studies on the diatom genus Navicula. V. The establishment
of Parlibellus gen. nov. for some members of Navicula sect. Microstigmaticae. Diatom
Research 3: 9-38.
Cox, E.J. (1988 b). Variation within the genus Pinnularia Ehrenb.: Further evidence for the
use of live material in diatom systematics. Proccedings of the 9th International Diatom
Symposium: 437-447.
Cox, E.J. (1999). Studies on the diatom genus Navicula Bory. VIII: Variation in valve
morphology in relation to the generic diagnosis based on Navicula tripunctata (O.F.
Müller) Bory. Diatom Research 14: 207-237.
Cox, E.J. (2002). Generic concepts and intrageneric variation in diatoms with particular
reference to Navicula Bory. Proceedings of the 15th International Diatom Symposium:
289-304.
Cox, E.J. (2003). Placoneis Mereschkowsky (Bacillariophyta) revisited: resolution of several
typification and nomenclatural problems, including the generitype. Botanical Journal of
the Linnean Society 141: 53-83.
Cox, E.J. and Williams, D.M. (2000). Systematics of naviculoid diatoms: the
interrelationships of some taxa with a stauros. European Journal of Phycology 35: 273282.
Damsté, J.S., Muyzer, G., Abbas, B., Rampen, S.W., Masse, G., Allard, W.G., Belt, S.T.,
Robert, J.M., Rowland, S.J., Moldowan, J.M., Barbanti, S.M., Fago, F.J.,
Denisevich, P., Dahl, J., Trindade, L.A. and Schouten, S. (2004). The rise of the
rhizosolenid diatoms. Science 304 (5670): 584-587.
Danielidis, D.B. and Mann, D.G. (2002). The systematics of Seminavis (Bacillariophyta):
the lost identities of Amphora angusta, A. ventricosa and A. macilenta. European
Journal of Phycology37: 429-448.
References
116
Darwin, Ch. (1859). On the Origin of Species by Means of Natural Selection, or the
Preservation of Favoured Races in the Struggle for Life. Chapter 6. John Murray,
London.
Daugbjerg, N. and Andersen, R.A. (1997). A molecular phylogeny of the heterokont
algaebased on analyses of chloroplast-encoded rbcL sequence data. Journal of
Phycology 33: 1031-1041.
Daugbjerg, N. and Guillou, L. (2001). Phylogenetic analyses of Bolidophyceae
(Heterokontophyta) using rbcL gene sequences support their sister group relationship to
diatoms. Phycologia 40: 153-161.
Davis, J.I., Simmons, M.P., Stevenson, D.W. and Wendel, J.F. (1998). Data decisiveness
and data quality in phylogenetic analysis: an example from the monocots using
mitochondrial atpA sequences. Systematic Biology 47: 282-310.
Delwiche, C.F., Kuhsel, M. and Palmer, J.D. (1995). Phylogenetic analysis of tufA
sequences indicates a cyanobacterial origin of all plastids. Molecular Phylogenetics and
Evolution. 4: 110-128.
Doyle, J.J. (1992). Gene trees and species trees: molecular systematics as one-character
taxonomy. Systematic Botany 17: 144-163.
Draisma, S.G.A. and Prud’homme van Reine, W.F. (electronic scource) Phylogeny of the
Phaeophyceae. http://www.nationaalherbarium.nl/taskforcemolecular/old_projects.htm
Ehara, M., Inagaki, Y., Watanabe, K.I. and Ohama, T. (2000). Phylogenetic analysis of
diatom coxI genes and implications of a fluctuating GC content on mitochondrial
genetic code evolution. Current Genetics 37: 29-33.
Ehrenberg, C.G. (1843). Mittheilungen über 2 neue asiatische Lager fossiler InfusorieErden aus dem russischen Trans-Kaukasien (Grusien) und Sibirien. Bericht über die zur
Bekanntmachung geeigneten Verhandlungen der Koniglich Preussischen Akademie der
Wissenschaften zu Berlin, 43-49.
References
117
Elwood, H.J., Olsen, G.J. and Sogin, M.L. (1985). The small subunit ribosomal RNA gene
sequences from the hypotrichous ciliates Oxytricha nova and Stylonychia pustulata.
Molecular Biology and Evolution 2: 399-410.
Eppley, R.W., Holmes, R.W. and Strickland, J.D.H. (1967). Sinking rates of marine
phytoplankton measured with a fluorometer. Journal of Experimental Marine Biology
and Ecology 1: 191-208.
Farris, J.S., Källersjö, M.,. Kluge, A.G and Bult, C. (1994). Testing significance of
incongruence. Cladistics 10: 315-320.
Felsenstein, J. (2004). Inferring phylogenenies. Sinauer Associates, Massachusetts.
Filatov D.A. (2002). ProSeq: A software for preparation and evolutionary analysis of DNA
sequence data sets. Molecular Ecology Notes 2: 621-624.
Flynn, J.J. and Nedbal, M.A. (1998). Phylogeny of the Carnivora (Mammalia): congruence
vs. incompatibility among multiple data sets. Molecular Phylogenetics and Evolution 9
(3): 414-426.
Fox, M.G. and Sorhannus, U.M. (2003). RpoA: a useful gene for phylogenetic analysis in
diatoms. Journal of Eukaryotic Microbiology 50, 471-475.
Frenguelli, J. (1938). Diatomeas dela Bahiade San Blas (Privoncia de Buenos Aires). Revista
del Museo de La Plata (Nueva serie), Botany 1: 303.
Gersonde, R. and Harwood, D.M. (1990). Lower Cretaceous diatoms from ODP Leg 113
Site 693 (Weddell Sea), Part 1. Vegetative cells. In: Barker, P. F. et al. [Eds.]
Proceedings of the Ocean Drilling Program, Science Results 113, College Station, TX
(Ocean Drilling Program): 365-402.
Graybeal, A. (1998). Is it better to add taxa or characters to a difficult phylogenetic problem?
Systematic Biology 47: 9-17.
Guillou, L., Chretiennot-Dinet, M.J., Medlin, L.K., Claustre, H., Goeer, SL.-D. and
Vaulot, D. (1999). Bolidomonas: A new genus with two species belonging to a new
algal class, the Bolidophyceae (Heterokonta). Journal of Phycology 35 (2): 368-381.
References
Gunderson, J.H.,
118
Sogin, M.L.,
Wollett, G.,
Hollingdale, M.,
de La Cruz, V.F.,
Waters, A.P. and McCutchan, T.F. (1987). Structurally distinct, stage-specific
ribosomes occur in Plasmodium. Science 238: 933-937.
Harwood, D.M. and Gersonde, R. (1990). Lower Cretaceous diatoms from ODP Leg 113
Site 693 (Weddell Sea), Part 2. Resting spores, Chrysophycean cysts and endosceletal
dinoflaggelates and notes on the origin of the diatoms. In: Barker, P. F. et al. [Eds.]
Proceedings of the Ocean Drilling Program, Science Results 113, College Station, TX
(Ocean Drilling Program): 365-402.
Hillis, D.M. (1987). Molecular versus morphological approaches to systematics. Annual
Review of Ecology and Systematics 18: 23-42.
Hillis, D.M. (1996). Inferring complex phylogenies. Nature 383: 130-131.
Hoot, S.B., Magallón, S. and Crane, P.R. (1999). Phylogeny of basal eudicots based on
three molecular data sets: atpB, rbcL, and 18S nuclear ribosomal DNA sequences.
Annals of the Missouri Botanical Garden 86: 1-32.
Huelsenbeck, J.P., Bull, J.J. and Cunningham, C.W. (1996). Combining data in
phylogenetic analysis. Trends in Ecology and Evolution 11 (4): 152-158.
Hustedt, F. (1924). Die Bacillariaceen-Vegetation des Sarekgebirges. Naturwissenschaftliche
Untersuchungen des Sarekgebirges in Schwedisch –Lappland, Botanik 3: 525-626.
Hustedt, F. (1930). Bacillariophyta (Diatomeae). In: A. Pascher [Ed.], Die Süsswasser-Flora
Mitteleuropas, Heft 10, Gustav Fischer, Jena.
Hustedt, F. (1934). Die Diatomeenflora von Poggenpohls Moor bei Dötlingen in Oldenburg.
Abhandlungen und Vorträge der Bremer Wissenschaftlichen Gesellschaft 8-9: 362-403.
Hustedt, F. (1961-1966). Die Kieselalgen Deutschlands, Österreichs und der Schweiz. 3.Teil.
In: Dr. L. Rabenhorsts Kryptogamen-Flora von Deutschland, Österreich und der
Schweiz, 7. Akademische Verlagsgesellschaft, Leipzig.
Johnson, K.P. and Sorensen M.D. (1998). Comparing Molecular Evolution in Two
Mitochondrial Protein Coding Genes (Cytochromeband ND2) in the Dabbling Ducks
(Tribe: Anatini). Molecular Phylogenetics and Evolution 10 (1): 82-94.
References
119
Karayeva, N.I. (1978). Novyi podporyadok diatomovykh vodoroslei (A new suborder of
diatoms). Botanischeskii Zhurnal 63: 1747-1750.
Kooistra, W.H.C.F. and Medlin, L.K. (1996). Evolution of the diatoms (Bacillariophyta).
IV. A reconstruction of their age from small subunit rRNA coding regions and the fossil
record. Molecular Phylogenetics and Evolution 6 (3): 391-407.
Kooistra, W.H.C.F., De Stefano, M., Mann, D.G., and Medlin, L.K. (2003). The
phylogeny of the diatoms. Progress in Molecular and Subcellular Biology 33: 59-97.
Krammer, K. (1982). Valve morphology in the genus Cymbella C.A. Aghard. In: Helmcke,
J.-G. and Krammer, K. [Eds.], Micromorphology of Diatom Valves XI.
Krammer, K. (2000). The genus Pinnularia. In: Lange-Bertalot, H. [Ed.], Diatoms of
Europe: diatoms of the European inland waters and comparable habitats. A.R.G.
Gantner, Ruggell.
Krammer, K. and Lange-Bertalot, H. (1985). Naviculaceae. Neue und wenig bekannte
Taxa, neue Kombinationen und Synonyme sowie Bemerkungen zu einigen Gattungen.
Bibliotheca Diatomologica 9.
Krammer, K. and Lange-Bertalot, H. (1986). Bacillariophyceae, 1. Teil: Naviculaceae.
Gustav Fischer Verlag, Stuttgart.
Kützing, F.T. (1833). Linnaea 8: 589.
Kützing, F.T. (1844). Die kieselschaligen Bacillarien oder Diatomeen, Nordhausen.
Lange-Bertalot, H. (1997). Frankophila, Mayamaea and Fistulifera: drei neue Gattungen aus
der Klasse Bacilariophyceae. Archiv für Protistenkunde 148: 65-76.
Lange-Bertalot, H., Metzeltin, D., and Witkowski, A. (1996). Hippodonta gen. nov. Umschreibung und Begründung einer neuen Gattung der Naviculiceae. Iconographia
Diatomologica 4, 247-275.
Long, E.O. and Dawid, I.B. (1980). Repeated genes in eukaryotes. Annual review of
biochemistry 49: 727-764.
References
120
Lundholm, N., Daugbjerg, N., and Moestrup, Ø. (2002 a). Phylogeny of the Bacillariaceae
with emphasis on the genus Pseudo-nitzschia (Bacillariophyceae) based on partial LSU
rDNA . European Journal of Phycology 37 (1): 115-134.
Lundholm, N., Hasle, G.R., Fryxell, G.A. and Hargraves, P.E. (2002 b). Morphology,
phylogeny and taxonomy of species within the Pseudo-nitzschia americana complex
(Bacillariophyceae) with descriptions of two new species, Pseudo-nitzschia brasiliana
and Pseudo-nitzschia linea. Phycologia 41: 480-497.
Lundholm, N. and Moestrup, Ø. (2002). The marine diatom Pseudo-nitzschia galaxiae sp.
nov. (Bacillariophyceae): Morphology and phylogenetic relationships. Phycologia 41
(6): 594-605.
Lutzoni, F. and Vilgalys, R. (1995). Integration of morphological and molecular data sets in
estimating fungal phylogenies. Canadian Journal of Botany 73: 649-659.
Maddison W.P. and Maddison D.R. (1989). Interactive analysis of phylogeny and character
evolution using the computer program MacClade. Folia primatologica 53 (1-4):190202.
Mann, D.G. (2001). A discussion of Caloneis and related genera. Diatom 17: 29-36.
Mann, D.G., Simpson, G.E., Sluiman, H.J. and Möller, M. (2001). RbcL gene tree of
diatoms: a second large data-set for phylogenetic reconstruction. Phycologia 40: 1-2.
Mann, D.G. and Stickle, A.J. (1989). Meiosis, nuclear cyclosis, and auxospore formation in
Navicula sensu stricto (Bacillariophyta). British Phycological Journal 24 (2): 167-181.
Mayama, S. and Kobayasi, H. (1988). Morphological variations in Navicula atomus (Kütz.)
Grun. Proccedings of the 9th International Diatom Symposium: 427-435.
Medlin, L.K., Elwood, H.J., Stickel, S. and Sogin, M.L. (1988). The characterization of
enzymatically amplified eukaryotic 16S-like rRNA-coding regions. Gene 71: 491-9.
Medlin, L.K., Elwood, H.J., Stickel, S., and Sogin, M.L. (1991). Morphological and genetic
variation within the diatom Skeletonema costatum (Bacillariophyceae): Evidence for a
new species, Skeletonema pseudocostatum. Journal of Phycology 27 (4): 514-524.
References
121
Medlin, L.K. and Kaczmarska, I. (2004). Evolution of the diatoms: V. Morphological and
cytological support for the major clades and a taxonomic revision. Phycologia 43: 245270.
Medlin, L.K., Kooistra, W.H.C.F., Gersonde, R. and Schmid, A.M. (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
diatoms: fossil, molecular and biogeographical approaches: 13-35.
Medlin, L.K., Kooistra, W.H.C.F., Gersonde, R. and Wellbrock, U. (1996 a). Evolution of
the diatoms (Bacillariophyta). II. Nuclear-encoded small-subunit rRNA sequence
comparisons confirm a paraphyletic origin for the centric diatoms. Molecular Biology
and Evolution 13: 67-75.
Medlin, L.K., Kooistra, W.H.C.F., Gersonde, R. and Wellbrock, U. (1996 b). Evolution of
the diatoms (Bacillariophyta): III. Molecular evidence for the origin of the
Thalassiosirales. Beiheft zur Nova Hedwigia 112: 221-234.
Medlin, L.K., Kooistra, W.H.C.F., Potter, D., Saunders, G.W. and Andersen, R.A.
(1997). Phylogenetic relationships of the 'golden algae' (haptophytes, heterokont
chromophytes) and their plastids. Plant Systematic and Evolution [Suppl.] 11: 187-219.
Medlin, L.K. and Round, F.E. (1986). Taxonomic studies of marine gomphonemoid
diatoms. Diatom Research 1 (2): 205-225.
Medlin, L.K., Williams, D.M. and Sims, P.A. (1993). The evolution of the diatoms
(Bacillariophyta). I. Origin of the group and assessment of the monophyly of its major
divisions. European Journal of Phycology 28 (4): 261-275.
Mereschkowsky, C. (1903). Über Placoneis, ein neues Diatomeen-Genus. Beihefte zum
Botanischen Centralblatt 15: 1-29.
Metzeltin, D., Lange - Bertalot, H. and Garcia - Rodriguez, F. (2004). Diatoms of
Uruguay. Iconographia Diatomologica, 15.
Moser, G., Lange-Bertalot, H. and Metzeltin, D. (1998). Insel der Endemiten Geobonanisches Phänomen Neukaledonien (Island of endemics New Caledonia - a
geobotanical phenomenon). Bibliotheca Diatomologica 38: 87-89.
References
122
Moshkovitz, S., Ehrlich, A., and Soundry, D. (1983). Siliceous microfossils of the Upper
Cretaceous Mishash Formation, Central Negev, Israel. Cretaceous Research. 4: 73-194.
Mullis, K.B., Faloona, F., Scharf, S., Saiki, R., Horn, G. and Ehrlich, H. (1986). Spezific
enzymatic amplification of DNA in vitro: The polymerase chain reaction. Cold Spring
Harbor symposia on quantitative biology 51: 263.
Mullis, K.B. and Faloona, F.A. (1987). Specific synthesis of DNA in vitro via a polymerasecatalyzed chain reaction. Methods in enzymology 155: 336.
Pagel, M. (1999). Inferring the historical patterns of biological evolution. Nature 401: 877884.
Pamilo, P. and Nei, M. (1988). Relationships between gene trees and species trees.
Molecular Biology and Evolution 5: 568-583.
Posada, D. and Crandall, K.A. (1998). Modeltest: testing the model of DNA substitution.
Bioinformatics 14 (9): 817-818.
Poulin, M., Massé, G., Belt, S.T., Delavault, P., Rousseau, F., Robert, J.-M. and
Rowland, S.J. (2004). Morphological, biochemical and molecular evidence for the
transfer of Gyrosigma nipkowii Meister to the genus Haslea (Bacillariophyta).
European Journal of Phycology 39: 181-195.
Pritchard, A. (1861). A general history of infusoria, ed. 4. Taylor and Francis, London.
Qari, S.H., Goldman, I.F., Pieniazek, N.J., Collins, W.E. and Lal A.A. (1994). Blood and
sporozoite stage-specific small subunit ribosomal RNA-encoding genes of the human
malaria parasite Plusmodium vivux. Gene 150 (1): 43-49.
Reichardt, E. (1992). Navicula jakovljevicii Hust. (Bacillariophyceae) Morphologie und
taxonomische Überlegungen. Diatom Research 7 (2): 293-301.
Rothpeltz, A. (1896). Über die Flysch-Fucoiden und einige andere fossile Algen, sowie über
liasische, Diatomeen führende Hornschwämme. Zeitschrift der Deutschen Geologischen
Gesellschaft 48: 914-954.
Round, F.E., Crawford, R.M. and Mann, D.G. (1990). The diatoms: Biology and
morphology of the genera. Cambridge University Press, Cambridge.
References
123
Round, F.E. (2001). Morphology of Navicula sensu stricto – an ecologist’s perspective. In:
Jahn, R., Kociolek, J.P., Witkowski, A. and Compère, P. (eds): Lange-BertalotFestschrift: Studies on Diatoms: 317-326.
Sáez, A.G., Probert, I., Geisen, M., Quinn, P., Young, J.R., and Medlin, L.K. (2003).
Pseudo-cryptic speciation in coccolithophores. Proceedings of the National Academy of
Sciences of the United States of America 100: 7163-7168.
Saiki, R.K., Gelfand, D.H., Stoffel, S., Scharf, S.J., Higuchi, R., Horn, G.T., Mullis, K.B.
and Erlich, H.A. (1988). Primer-directed enzymatic amplification of DNA with a
thermostable DNA polymerase. Science 239: 487-91.
Saitou, N. and Nei, M. (1986). The number of nucleotides required to determine the
branching order of three species, with special reference to the human-chimpanzeegorilla divergence. Journal of Molecular Evolution 24 (1-2):189-204.
Salemi, M. and Vandamme, A.-M. (2003). The phylogenetic handbook – A practical
approach to DNA and protein phylogeny. University Press, Cambridge.
Schiller, W. and Lange-Bertalot, H. (1997). Eolimna martinii n. gen., n. sp.
(Bacillariophyceae) aus dem Unter-Oligozän von Sieblos/Rhön im Vergleich mit
ähnliche rezenten Taxa. Paläonthologische Zeitschrift 71: 163-172.
Schlötterer, C. and Tautz, D. (1994). Chromosomal homogeneity of Drosophila ribosomal
DNA arrays suggests intrachromosomal exchanges drive concerted evolution. Current
Biology, 4: 777-783.
Schoeman, F.R. and Archibald, R.E.M. (1986). Observations on Amphora species
(Bacillariophyceae) in the British Museum (Natural History). V. Some species from the
subgenus Amphora. South African Journal of Botany 52 (5): 425-437.
Scholin, C.A,, Herzog, M., Sogin, M. and Anderson, D.M. (1994). Identification of Groupand
Strain-Specific
Genetic
Markers
for
Globally
Distributed
Alexandrium
(Dinophyceae). 2. Sequence Analysis of a Fragment of the LSU rRNA Gene. Journal of
Phycology 30 (6): 999-1011.
Schütt, F. (1896). Bacillariales (Diatomeae). In: Engler and Prand [Eds.], Die natürliche
Pflanzenfamilien, 1: 31 - 153. Verlag von Wilhelm Engelmann, Leipzig.
References
124
Smith, W. (1856). A Synopsis of the British Diatomaceae, Volume II. Van Voost, London.
Söller, R., Warnke, K., Saint-Paul, U., and Blohm, D. (2000). Sequence divergence of
mitochondrial DNA indicates cryptic biodiversity in Octopus vulgaris and support the
taxonomic distinctiveness of Octopus mimus (Cephalopoda: Octopodinae). Marine
Biology 136: 29-35.
Soltis, D.E. and Soltis, P.S. (1998). Choosing an approach and an appropriate gene for
phylogenetic analysis. In: Soltis, D.E. et al., Molecular systematics of plants II: DNA
sequencing, pp.1-42, Kluwer Academic Publishers.
Soltis, D.E., Soltis, P.S., Chase, M.W., Mort, M.E., Albach, D.C., Zanis, M., Savolainen,
V., Hahn, W.H., Hoot, S.B., Fay, M.F., Axtell, M., Swensen, S.M., Nixon, K.C. and
Farris, J.S. (2000). Angiosperm phylogeny inferred from a combined data set of
18SrDNA, rbcL, and atpB sequences. Botanical Journal of the Linnean Society 133:
381-461.
Soltis, D.E., Soltis, P.S., Mort, M.E., Chase, M.W., Savolainen, V., Hoot, S.B. and
Morton, C.M. (1998). Inferring complex phylogenies using parsimony: an empirical
approach using three large DNA data sets for angiosperms. Systematic Biology 47: 3242.
Sorhannus, U. (2001). A “total evidence” analysis of the phylogenetic relationships among
the photosynthetic Stramenopiles. Cladistics 17: 227-241.
Sorhannus, U. (2004). Diatom phylogenetics inferred based on direct optimization of
nuclear-encoded SSU rRNA sequences. Cladistics 20: 487-497.
Sorhannus, U., Gasse, F., Perasso, R. and Tourancheau, A.B. (1995). A preliminary
phylogeny of diatoms based on 28S ribosomal RNA sequence data. Phycologia 34: 6573.
Strelnikova, N.I. and Martirosjan, G.N. (1981). Lower Cretaceous diatom algae from
Stavropol. Viestnik LGU, Seria biologiya 3: 52-57.
Sullivan J. (1996). Combining data with different distributions of among-site rate variation.
Systematic Biology 45: 375-380.
References
125
Swofford, D.L. (1995). PAUP*. Phylogenetic analysis using Parsimony (* and other
methods). Sinauer Associates, Massachusetts.
Swofford, D.L. (1998). PAUP*. Phylogenetic analysis using Parsimony (* and other
methods). Version 4.0b10. Sinauer Associates, Massachusetts.
Swofford, D.L., Olsen, G.J., Waddell, P.J. and Hillis, D.M. (1996). Phylogenetic Inference.
In: Hillis, D.M., Moritz, C. and Mable, B.K. [Eds.], Molecular Systematics, pp. 407514. Sinauer Associates, Massachusetts.
Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F. and Higgins, D.G. (1997).
The CLUSTAL_X windows interface: flexible strategies for multiple sequence
alignment aided by quality analysis tools. Nucleic Acids Research 25 (24), 4876-4882.
Van der Auwera, G. and De Wachter, R. (1998). Structure of the Large Subunit rDNA
from a Diatom, and Comparison Between Small and Large Subunit Ribosomal RNA for
Studying Stramenopile Evolution. Journal of Eukaryotic Microbiology 45: 521-527.
Waters, A.P., Syin, C. and McCutchan, T.E. (1989). Developmental regulation of stagespecific ribosome populations in Plasmodium. Nature 342: 438-440.
Witkowski A., Lange-Bertalot H. and Stachura K. (1998). New and confused species in
the genus Navicula (Bacillariophyceae) and the consequences of the restricrive generic
circumscription. Cryptogamie, Algologie 19: 83-108.
Woese, C.R. (1987). Bacterial evolution. Microbiological Reviews 51: 221-271.
Yoder, A.D., Irwin, J.A., and Payseur, B. (2001). Failure of the ILD to determine data
combinability for slow loris phylogeny. Systematic Biology 50: 408-424.
Zechman, F.W., Zimmer, E.A. and Theriot, E.C. (1994). Use of ribosomal DNA internal
transcribed spacers for phylogenetic studies in diatoms. Journal of Phycology 30 (3):
507-512.
Appendix
Appendix
Used PAUP command blocks:
Outgroup in all analyses
Outgroup 1493_CYCLOTELLA_CHOCTAWATCHEEA
1254_FRAGILARIA_CROTONENSIS 1256_ASTERIONELLA_FORMOSA;
Maximum Likelihood analyses
set cri=L;
“MODELBLOCK”;
Hsearch start=NJ Timelimit=144000;
savetrees format=phylip brlens=yes file=NAME_AIC_ML.trees;
Modelblocks:
•
SSU rDNA sequences from AlgaTerra cultures:
Lset Base=(0.2765 0.1690 0.2453) Nst=6 Rmat=(1.0531 3.1951 1.2328 0.7719
5.6023) Rates=gamma Shape=0.5864 Pinvar=0.4892;
•
SSU rDNA sequences from AlgaTerra cultures and GenBank:
Lset Base=(0.2644 0.1619 0.2453) Nst=6 Rmat=(1.3518 3.4650 1.3715 1.2991
6.4325) Rates=gamma Shape=0.4599 Pinvar=0.3493;
•
LSU rDNA sequences from AlgaTerra cultures:
Lset Base=(0.3020 0.1525 0.2403) Nst=6 Rmat=(1.0000 2.5417 1.0000 1.0000
5.8760) Rates=gamma Shape=0.6160 Pinvar=0.2261;
•
LSU rDNA sequences from AlgaTerra cultures and GenBank:
Lset Base=(0.2833 0.1713 0.2460) Nst=6 Rmat=(0.8410 2.5640 1.2117 0.8021
4.9980) Rates=gamma Shape=0.5576 Pinvar=0.2085;
•
rbcL gene sequences from AlgaTerra cultures:
Lset Base=(0.2971 0.1400 0.1490) Nst=6 Rmat=(0.6610 2.7121 1.3598 0.7100
3.8746) Rates=gamma Shape=0.6289 Pinvar=0.5455;
•
rbcL gene sequences from AlgaTerra cultures and GenBank:
Lset Base=(0.3128 0.1285 0.1443) Nst=6 Rmat=(0.5876 2.6700 1.1400 0.6432
3.5812) Rates=gamma Shape=0.6738 Pinvar=0.5429;
126
Appendix
•
127
combined sequences from AlgaTerra cultures:
Lset Base=(0.2847 0.1625 0.2262) Nst=6 Rmat=(0.9032 3.0043 1.4657 0.8179
5.2164) Rates=gamma Shape=0.5134 Pinvar=0.4521;
Parsimony analyses
[consensus parsimony tree]
set criterion=parsimony increase=auto;
hsearch addseq=random;
contree/ majrule=yes LE50=yes treefile=NAME_parcon.trees;
[Parsimony bootstrap tree]
set cri=par increase=auto;
bootstrap nreps=1000 search=heu keepall=yes treefile=NAME_PARbootNEW.trees;
savetrees from=1 to=1 savebootp=nodelabels maxdecimals=0
file=NAME_PARbootBaumNEW.trees;
[Parsimony bootstrap tree for SSU rDNA sequences from AlgaTerra cultures and GenBank]
log file=6PAR_BS15.log;
set cri=par increase=auto;
bootstrap nreps=1000 search=heu keepall=yes
treefile=6_keepall_PARBS_Timelimit_15.trees /Timelimit=900 Dstatus=300;
savetrees from=1 to=1 savebootp=nodelabels maxdecimals=0
file=6_keepall_PARBSBaum_Timelimit_15.trees;
Neighbor joining analyses
[NJ bootstrap tree]
set cri=dis;
dset dis=JC;
bootstrap nreps=1000 search=NJ keepall=yes treefile=NAME_NJJCboot.trees;
savetrees from=1 to=1 savebootp=nodelabels maxdecimals=0
file=NAME_NJJCbootBaum.trees;
Appendix
128
Partition homogeneity test (combined dataset)
set increase=auto;
log file=4_PHT100.log;
charpartition genes=18s:1-1828, rbcL:1829-2513, 28s:2514-3229;
hompart partition=genes nreps=100 seed=123 search=heu;
end;
Weightblock (MacClade output used for the rbcL gene dataset from AlgaTerra cultures)
BEGIN CODONS;
GENCODE UNIVNUC;
ENDBLOCK;
BEGIN ASSUMPTIONS;
OPTIONS DEFTYPE=unord PolyTcount=MINSTEPS ;
WTSET = 1.00: 1 5 10 11 16 17 19 20 22 24 26 27 28 29 33 34 35 37 38 40 41 42 43 47 49
50 51 53 55 56 61 62 64 65 66 67 68 70 71 73 74 76 77 78 79 80 81 82 83 85 86 88 89 91 92
94 95 97 98 106 107 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124
125 126 127 128 130 131 132 133 134 136 137 139 140 142 143 147 148 149 150 151 152
154 155 156 157 158 160 161 163 164 166 167 169 170 172 173 174 176 181 182 184 185
187 188 190 191 193 194 195 196 197 198 199 200 202 203 205 206 208 209 214 215 217
218 219 220 221 222 223 224 226 227 229 230 232 233 234 235 236 237 238 239 241 242
244 245 246 247 248 253 254 256 257 259 260 261 262 263 265 266 268 269 271 272 273
274 275 277 278 280 284 285 286 287 288 289 290 291 292 293 294 295 296 298 299 301
302 303 304 305 307 308 310 311 312 313 314 315 316 317 319 320 322 323 325 326 328
329 331 332 334 335 337 338 339 340 341 343 344 346 347 348 349 350 351 352 353 354
355 356 357 358 359 361 362 363 364 368 369 371 373 374 375 377 379 380 381 382 386
388 389 391 392 394 395-398\3 404 406 407 409 410 412 413 414 418 419 420 421 422 424
425 427 428 432 433 434 439 440 445 449 454 455 461 463 469 470 472 478 485 486 491
493 494 496 497 508 509 510 512 514 515 517 518 519 520 521 526 527 529 530 532 533
538 539 541 542 543 550 551 556 557 558 562 563 564 568 569 570 571 572 577 578 582
583 584 585 586 587 588 589 590 595 596 598 599 601 602 604 605 610 611 613 614 616
617 619 620 621 622 623 625 626 628 629 631 632 634 635 637 638 640 641 642 643 644
645 646 647 649 650 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 673
674 675 676 680 682, 2.91: 2 9 48 52 69 281 309 366 372 387-390\3 416 447 488 505 534
544 545, 10.00: 3 7 14 23 25 31 39 44 45 57 63 96 105 135 141 145 159 204 210 243 249-
Appendix
129
252\3 258 276 297-300\3 318 365 376 393 401 411 423 442 451 452 453 459 464 468 473
479 511 546 560 566 575 580 592 594 603 607 618 639 652 654 671 683, 1.18: 6 192 267
477 487 492 507, 1.34: 12 72 99 211 399 531 553 597 624-627\3, 1.50: 13 21 75 144 178 216
522 579, 1.02: 15, 2.13: 18 103 108 153 212 270 383 403 441 540 648, 3.70: 30 90 129 175
225 255 408 435 456 457-460\3 475 483 498 523 525 535 565 574 576 591 636 670 677,
1.14: 36 228 480 552, 2.44: 54 162 171 186 201 213 231 342 415 417 429 446 476 482 495
528 548 633, 5.28: 58 60 84 180-183\3 189 250 264 279 306 330-333\3 378 384 430 436 458
465 466 537 554 561 600 678, 1.23: 87 100 240 324 336 345 397 474 573, 1.90: 93 490 549
615 672, 1.60: 102 138 321 327 405 444 547 630 681, 1.28: 165 207 612, 1.05: 168 462 499
501, 1.41: 177 370 396 400 402 438 450 504 516 606 651, 1.08: 282 360 555 567, 1.73: 426
502 513 559 609 684;
ENDBLOCK;
Appendix
130
Tab. 9: Number of unknown nucleotides in SSU rDNA sequences obtained from GenBank
GenBank
accession number
AB085830
AB085831
AB085832
AB085833
AJ243061
AJ243062
AJ243063
AJ243064
AJ535144
AJ535145
AJ535149
AJ544649
AJ544655
AJ544659
AJ866992
AJ866995
AJ867023
AJ867024
AJ867025
AJ867027
AJ867028
AJ867029
AJ867030
AY485460
AY485462
AY485468
AY485476
AY485482
AY485483
AY485484
AY485488
AY485489
AY485496
AY485497
AY485498
AY485500
AY485502
AY485512
AY485513
AY485514
AY485515
AY485516
AY485521
AY485524
AY485528
AY672802
AY821975
Species
Eunotia formica var. smatrana
Eunotia monodon var. asiatica
Eunotia pectinalis
Gomphonema pseudaugur
Amphora montana
Gomphonema parvulum
Eolimna minima
Eolimna subminuscula
Rossia sp.
Eunotia sp.
Lyrella sp.
Sellaphora pupula
Sellaphora laevissima
Lyrella atlantica
Achnanthes minutissima
Eunotia bilunaris
Diadesmis gallica
Navicula atomus var. permitis
Navicula saprophila
Pinnularia rupestris
Surirella angusta
Surirella brebissoni
Cymatopleura elliptica
Navicula sp.
Dickieia ulvacea
Amphiprora paludosa
Achnanthes breviceps
Haslea crucigera
Navicula sclesviscensis
Navicula lanceolata
Haslea nipkowii
Pleurosigma intermedium
Achnanthes sp.
Amphiprora alata
Amphora coffeaformis
Achnanthidium cf. longipes
Navicula sp.
Navicula ramonissima
Navicula sp.
Pleurosigma planktonicum
Pleurosigma sp.
Gyrosigma limosum
Stauroneis constricta
Haslea pseudostrearia
Paulielle taeniata
Fragilariopsis cylindrus
uncultured Eunotia-like diatom
unknown nucleotides close to
primer 1F
primer 1528R
66
104
79
18
66
25
6
7
8
6
65
34
34
22
32
34
202
25
202
25
202
25
202
25
202
25
202
25
202
25
202
25
202
25
35
22
35
29
33
54
27
10
59
64
35
27
80
100
101
53
64
61
29
23
97
74
27
35
27
62
58
13
56
72
62
139
14
149
14
96
63
2
146
62
Appendix
131
95
100
71
99
92
66
90
92
50
100
8
84
15
57
13
99
21
44
18
28
31
58
100
100
23
56
14
92
22
50
8
56
90
59
100
24
31
4
98
53
7
37
13
100
100
100
28
6
100
67
86
83
66
12
50
77
64
100
90
61
23
23
Eunotia
Pleurosigma
AY485516 Gyrosigma limosum
AY485524 Haslea pseudostrearia
AY485523 Haslea ostrearia
AY485482 Haslea crucigera
Navicula s. str. and Pseudogomphonema
Hippodonta capitata (1272)
Frustulia vulraris (1445)
AY485488 Haslea nipkowii
Navicula hambergii (1436)
Placoneis elginensis (1312)
Placoneis sp. (1419)
Encyonema
Gomphonema
Cymbella
AJ535153 Anomoeoneis sphaerophora
28
14
Cyclotella choctawatcheeana (1493)
Asterionella formosa (1256)
Fragilaria crotonensis (1254)
Fragilaria sp. (1410)
65
Lyrella
Cocconeis
AJ535189 Planothidium lanceolatum
AJ867023 Diadesmis gallica
AJ866992 Achnanthes minutissima
Achnanthidium minutissimum (1438)
AY485528 Pauliella toeniata
AY485462 Dickieia ulvacea
Navicula brockmannii (1425)
Pinnularia acrosphaeria (1426)
Caloneis amphisbeana (1550)
AJ535162 Campylodiscus ralfsii
Luticola goeppertiana (1273)
Neidum affine (1551)
clade LB
Surirellales
Amphora (group 2)
Amphora (group 1)
AY485459 Phaeodactylum tricornutum
Mayamaea
AJ867025 Navicula saprophila
AY485454 Navicula pelliculosa
Stauroneis phoenicenteron (1293, 1437)
Stauroneis anceps (1412)
Stauroneis gracilior (1294)
Stauroneis kriegerii (1309)
Navicula integra (1430)
Craticula
AJ243064 Eolimna subminuscula
Sellaphora
Eolimna minima (1267, AJ243063)
AY485498 Amphora coffeaeformis
Pinnularia/Caloneis
AJ535144 Rossia sp.
11
20
50
Bacillariales
Achnanthes (group 1)
Fig. 59: Neighbor joining tree based on SSU rDNA sequences from GenBank and AlgaTerra
cultures. Bootstrap values obtained from 1000 replications based on NJ analyses using JC
model have been plotted at the nodes. The marked clade LB is in all probability caused by
Long Branch Attraction.
Appendix
132
Additional microraphs
The figures 60 – 70 show micrographs of sequenced species, which were not present in the
results. The species are shown in alphabetical order.
a
Fig. 60: Achnanthidium minutissimum
SEM, raphid (a) and rapheless (b) valve (deformed valves, old culture).
Fig. 61: Amphora sp. (1554). Light micrograph of live individual.
b
Appendix
133
a
b
c
d
e
f
Fig. 62: Cocconeis species.
a + b: C. pediculus. Light micrograph of cleaned valves (a) and SEM, internal view (b).
c - f: C. placentula. Light micrographs of cleaned valves (c + d) and SEM, external view (e + f).
Appendix
a
134
d
e
b
f
c
Fig. 63: a – c: Encyonema caespitosum, d – f: E. minutum
a + d: Light micrographs of cleaned valve, b + e: SEM, external view, c + f: SEM, internal view.
Appendix
135
b
a
c
d
Fig. 64: Eolimna minima
a + b: Light micrographs of live individual (a) and cleaned valve (b), c + d: SEM, external (c) and
internal (d) view.
Appendix
136
c
a
b
d
e
f
g
Fig. 65: Eunotia species
a – d: E. formica. Light micrographs of cleaned valves in girdle (a) and valve view (b) and live
individual (c), SEM showing the raphe (d).
e + f: E. implicata. Light micrographs of cleaned valves in valve (e) and girdle view (f).
g: Eunotia sp. SEM showing several strongly deformed valves.
Appendix
137
a
b
c
d
e
f
Fig. 66: Frustulia vulgaris
a + b: Light micrographs of live individual (a) and cleaned valve (b),
c - f: SEM, external (c + d) and internal (e + f) view.
Appendix
138
a
b
c
d
Fig. 67: Hippodonta capitata
a + b: Light micrographs of live individual (a) and cleaned valve (b), c + d: SEM, external (c) and
internal (d) view.
b
c
a
Fig. 68: Luticola goeppertiana
a : Light micrograph of cleaned valve, b + c: SEM, external view (b) and detail of the internal stigma
aperture (c).
Appendix
139
a
b
Fig. 69: Neidum affine
Light micrographs of live individual (a) and cleaned valve (b).
Fig. 70: Placoneis sp. (1419)
Light micrograph of cleaned valve.
Appendix
140
Eunotia
formica
1268 EUNO
TIA (1268)
FO RMICA
Eunotia
sp. (1269)
1269 EUNO
TIA S P
Eunotia
implicata
(1321)
1321 EUNO
TIA IMP
LICA TA
Amphora
pediculus
12
65 AMPH
ORA PE(1265)
DICULUS
Amphora
cf. fogediana
(1427)
1427 AMPH
ORA CF FO
GE DIANA
Amphora
libyca
(1264)
12 64 AMPH
ORA
LIBYCA
Amphora
sp.ORA
(1554)
15
54 AMPH
SP
Eolimna
(1267)
1267
EOminima
LIMNA MINIMA
Mayamaea
var. atomus
12 74 MAYAatomus
MA EA ATO
MU S (1274)
Mayamaea
var. permitis
12 75 MAYAMAEA
MAYAatomus
MA EA PERMITIS
ATO
MU S (1275)
1275
1291 PINNULARIA
Pinnularia
viridiformisVIRIDIFO
(1291) RMIS
14
28 PINNULA
CF V IRIDIS
Pinnularia
viridisRIA
(1428)
1442 PINNULA
RIA VIRIDIS (1442)
Pinnularia
cf. substreptoraphe
1311
PINNULARIA
Pinnularia
rupestris RUPRE
(1312) STRIS
1446
CALlauta
ONE(1446)
IS LA UTA
Caloneis
12 90 PINNULA
RIA MICROS
TAURO N
Pinnularia
microstauron
(1290)
1429 PINNULARIA
Pinnularia
mesoleptaMESO
(1429)LEP TA
12 86 PINNULA
RIA
ANG LICA
Pinnularia
anglica
(1286)
Pinnularia
obscura
12 92 PINNULA
RIA(1292)
OB SCURA
1285 PINNULA
RIA SUBCAPITATA
Pinnularia
subcapitata
(1285)
14 26 PINNULA
RIA ACRO
SPA ERIA
Pinnularia
acrosphaeria
(1426)
1550 CALONE
IS AMPHISBAENA
Caloneis
amphisbaena
(1550)
13 23 CALONE
IS CF
B UDENSIS
Caloneis
budensis
(1323)
Amphora
normannii
(1263)NNII
1263 AMPH
ORA NORMA
Craticula
halophilioides
(1308) ILIOIDE
1308 NAV
ICULA C F HALOPH
Craticula
(1284) MO ID
1284
NAVmolestiformis
ICUL A C F DIFFICILO
Craticula
cuspidata
(1320)
1320
CRAT
ICULA CUS
PIDA TA
Stauroneis
phoenicenteron
(1437)
14 37 ST AURONE
IS PHO ENICENTERO
N
Stauroneis
anceps IS
(1412)
14 12 ST AURONE
ANCE PS
Stauroneis
graciliorIS
(1294)
12 94 ST AURONE
GR ACIL IOR
Stauroneis
kriegeriiIS(1444)
14
44 ST AURONE
KRIEG ERII
Navicula
integra (1430)
14 30 NAVICULA
INTEG RA
Luticola
goeppertiana
(1273)
12
73 L UTICOL
A GO EPP
ERTIANA
15 51 NEIDUM
CF AFFINE
Neidum
affine (1551)
12 71 G OMPHO
NEMA MICROP
Gomphonema
micropus
(1271) US
1322 G OMPHO
NEMA
CF AFFIN E
Gomphonema
affine
(1322)
1439
G OMPHOaffine
NEMA
SP
Gomphonema
(1439)
1424 G OMPHO
NEMA ACUMINAT
Gomphonema
acuminatum
(1424) UM
15 52 G OMPHO
NEMA TRUNCATUM
Gomphonema
truncatum
(1552)
1409
G OMPHOcf.
NEMA
PRO DUCTUM
Gomphonema
angustatum
(1409)
13 15 G OMPHO
SP (1315)
Gomphonema
cf.NEMA
parvulum
13 12 PLACO
NEIS E (1312)
LGINENS IS
Placoneis
elginensis
1419 NAVsp.
ICULA
SP
Placoneis
(1419)
1436 NAVhambergii
ICULA H AMBERG
Navicula
(1436) II
13 24 CYMBE
LLA NA VICUL
IFORMIS
Cymbella
naviculiformis
(1324)
14 21 CYMBE
LLA(1421)
A SPERA
Cymbella
aspera
1414 CYMBE
LLA
AFFINIS
Cymbella
affinis
(1414,
1423)
1423 CYMBE LLA SP
1431 CYMBE
LLA HE
LMKEI
Cymbella
helmckei
(1431)
1422 CYMBELLA
P ROXIMA
Cymbella
proxima (1422)
Encyonema
chaespitosum
(1441)
1441 CYMBE
LLA CF CAES
PITO SA
14
38 ACHN ANTHIDIUM
MINUTISSIMU
Achnanthidium
minutissimum
(1438)
Cocconeis
pediculus
(1415)
14 15 CO CCONE
IS PEDICULUS
1416 NAV placentula
ICULA C RYPTO
Cocconeis
(1418)TENEL LA
1425
NAVbrockmannii
ICULA B ROCKMA
Navicula
(1425) NNII
Fragilaria
sp. (1410)
1410
DIA TOM
12 54 FRAG
ILARIA CRO
TONE NSIS
Fragilaria
crotonensis
(1254)
Asterionella
formosaLA
(1256)
12 56 AS TERIONEL
FORMO SA
Navicula
1281
NAVveneta
ICUL A(1281)
VENET A
Navicula
12
80 NAVgregaria
ICUL A G(1280)
REG ARIA
Navicula
sp.1
(1411)
14 11 NAV
ICUL
A SP
Navicula
1440
NAVradiosa
ICUL A(1440)
R ADIO SA
Navicula
1418 COcapitatoradiata
CCONEIS PLA(1417)
CENTUL A
1310
NAVcari
ICULA
C ARI
Navicula
(1310)
14
34 NAVtripunctata
ICUL A T RIPUNCTATA
Navicula
(1434)
1316 NAVcryptocephala
ICUL A C RYPTO
CEPHA LA
Navicula
(1316)
14 17 NAV ICUL A C APIT ATOR ADIA TA
14
20 NAVcryptotenella
ICUL A C RYPTO
TENEL
Navicula
(1416,
1420,LA1435)
14 35 NAV ICUL A S P
13 19 NAV
ICUL
A SP
Navicula
sp.2
(1319)
1282 NAVreinhardtii
ICUL A (1282)
Navicula
12
72 HIP POcapitata
DONTA (1272)
CA PITA TA
Hippodonta
14
93 CYCL
OTE LLA CHO CTAW
ATCHEE A
Cyclotella
choctawatcheeana
(1493)
0.01 substitutions/site
Fig. 71: Phylogeny inferred with the ML analysis using a weight block obtained from MacClade
based on rbcL sequences from AlgaTerra cultures.
Acknowledgements
141
Acknowledgements
Für mein Ömchen († 08. Mai 2006).
Mein ganz besonderer Dank geht an Dr. Linda Medlin für die Betreuung meiner Arbeit und
an meine Gutachter Prof. Dr. Gunter-Otto Kirst und Prof. Dr. Ulrich Bathmann.
Für die tolle Zusammenarbeit im Projekt AlgaTerra mit vielen erfolgreichen und lustigen
(gell, Sabienchen) Exkursionen danke ich Bank Beszteri, Ines Jung, Sabine Strieben und Olaf
Wandschneider.
Dr. Richard M. Crawford danke ich für ein allseits offenes Ohr und sprachliche Korrekturen.
Bei Friedel Hinz möchte ich mich für ihre große Hilfsbereitschaft und besonders für die
kompetente Einführung ins REM und in die LM-Fotographie bedanken.
Georgia Klein danke ich dafür, dass sie mich in Stettin ertragen hat und für Tee, Kekse,
Schokolade und gemeinsames jammern.
Dank auch allen anderen derzeitigen und ehemaligen Mitgliedern der AG Medlin (Andrea
Reents, Christine Gescher, Helga Mehl, Jessica Kegel, Katja Metfies, Kerstin Töbe, Klaus
Valentin, Monica Estanqueiro, Niko Hoch, Rene Groben, Shinya Sato, Sonja Dierks, Steffi
Gäbler, Uwe John) für die tolle Arbeitsatmosphäre mit den vielen fachlichen und privaten
Gesprächen und den großen und kleinen Hilfen.
Kurt Krammer und Gabi Hofmann möchte ich für die Verifizierung der taxonomischen
Identifikation meiner Kulturen danken.
Last but not least, möchte ich mich bei meiner Familie für ihre Unterstützung und ihr
Interesse bedanken und ganz besonders bei Alex für seine nervigen Sticheleien J .
Diese Arbeit entstand in Rahmen des BMBF Projektes AlgaTerra 01LC0026 am AlfredWegener-Institut für Polar- und Meeresforschung in Bremerhaven.
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement