J. Phycol.

49, 1167–1180 (2013)

© 2013 Phycological Society of America
DOI: 10.1111/jpy.12128

MOLECULAR AND MORPHOLOGICAL CRITERIA FOR REVISION OF THE GENUS

MICROCOLEUS (OSCILLATORIALES, CYANOBACTERIA)1

y2
Otakar Struneck

Institute of Botany, Centre for Phycology, The Academy of Sciences of the Czech Republic, Dukelsk
a 135, Trebo
n 379 82, Czech
Republic
Centre for Polar Ecology, Department of Ecosystem Biology, Faculty of Science, University of South Bohemia, Branisovska 31,
Ceske Budejovice 370 05, Czech Republic

r
ı Kom

Ji a rek
Institute of Botany, Centre for Phycology, The Academy of Sciences of the Czech Republic, Dukelsk
a 135, Trebon 379 82, Czech
Republic
Institute of Botany, Faculty of Science, University of South Bohemia, Branisovska 31, Ceske Budejovice 370 05, Czech Republic

Jeffrey Johansen
Biology, John Carroll University, University Heights, Ohio 44118, USA

Alena Lukes ov

a
 e Bud
Institute of Soil Biology, Biology Centre ASCR, v.v.i., Na S
adk
ach 7, Cesk
ejovice 370 05, Czech Republic

and Josef Elster

Institute of Botany, Centre for Phycology, The Academy of Sciences of the Czech Republic, Dukelsk
a 135, Trebo
n 379 82, Czech
Republic
Centre for Polar Ecology, Department of Ecosystem Biology, Faculty of Science, University of South Bohemia, Branisovska 31,
Ceske Budejovice 370 05, Czech Republic

Ninety-two strains of Microcoleus vaginatus characters defining Microcoleus are related to the
(=nomenclatural-type species of the genus Microcoleus cytomorphology of trichomes, including: narrowed
Desmazi eres ex Gomont) and Phormidium autumnale trichome ends, calyptra, cells shorter than wide up to
Trevisan ex Gomont from a wide diversity of more or less isodiametric, and facultative presence of
regions and biotopes were examined using a sheaths. The majority of species are 4–10 lm in
combination of morphological and molecular diameter. The possession of multiple trichomes in a
methods. Phylogenies based on the 16S rDNA and common sheath is present facultatively in many but
16S-23S ITS (partial) demonstrated that the 92 not all species.
strains, together with a number of strains in Key index words: Cyanobacteria; Microcoleus; molecu-
GenBank, were members of a highly supported lar evaluation; morphology; Phormidium; phylogeny;
monophyletic clade of strains (Bayesian posterior taxonomy
probability = 1.0) distant from the species-cluster
containing the generitype of Phormidium. Similarity List of Abbreviations: CCALA, Culture Collection of
of the 16S rRNA gene exceeded 95.5% among all Autotrophic Organisms, Trebo n, Czech Republic;
members of the Microcoleus clade, but was less than ISBAL, Culture Collection of Soil Algae at the Insti-
95% between any Microcoleus strains and species 
tute of Soil Biology, Cesk
e Budejovice, Czech
outside of the clade (e.g., Phormidium sensu stricto). Republic; ITS, internal transcribed spacer; PCC,
These findings, which are in agreement with earlier Pasteur Culture Collection; SAG, Culture Collection
studies on these taxa, necessitate the revision of of Algae at Gottingen University, Germany; UTCC,
Microcoleus to include P. autumnale. Furthermore, the University of Toronto Culture Collection of Algae
cluster of Phormidium species in the P. autumnale and Cyanobacteria
group (known as Group VII) must be moved into
Microcoleus as well, and these nomenclatural transfers
are included in this study. The main diacritical
Both Microcoleus Desmazi
eres ex Gomont and Pho-
1
Received 4 October 2012. Accepted 16 July 2013.
rmidium K€utzing ex Gomont are polyphyletic genera
2
Author for correspondence: e-mail [email protected]. in the Oscillatoriales (phylum Cyanobacteria), as fol-
Editorial Responsibility: A. Post (Associate Editor) lows from recent molecular analyses (Boyer et al.

1167
1168 O T A K A R S T R U N E C K Y
E T A L .

2002, Taton et al. 2006, Palinska and Marquardt filaments in a common sheath. The morphological
2008, Siegesmund et al. 2008, Struneck
y et al. 2010, difference between M. vaginatus and P. autumnale is
2011, and others), as well as from detailed cytomor- only in the form of colonies and organization of
phological investigations (Turicchia et al. 2009). filaments, but the morphology of trichomes is dis-
They are both species-rich genera, with about 170 tinctly similar. A molecular synapomorphy has been
validly described species in Phormidium and 30 repeatedly found and reported as diagnostic for
validly described species in Microcoleus up to now M. vaginatus, an 11 nucleotide insert in the H17
(Kom
arek and Anagnostidis 2005). Both genera helix of the 16S rRNA transcript (Garcia-Pichel
date from the starting point of oscillatorialean gen- et al. 2001, Boyer et al. 2002, Hasler et al. 2012).
era (Gomont 1892). The generitype of Phormidium Microcoleus as currently defined is polyphyletic.
is P. lucidum K€
utzing ex Gomont (Geitler 1942); the The generitype, M. vaginatus has successive cell divi-
generitype of Microcoleus is M. vaginatus (Vaucher) sion and thylakoid structure similar to the family
Gomont (Geitler 1942, Drouet 1968). Oscillatoriaceae as defined in Anagnostidis and
Kom
arek and Anagnostidis (2005) presented a Kom
arek (1988), while most of the species in the
revised taxonomy for the nonheterocytous filamen- genus, as presently defined, have cell division and
tous cyanobacteria based on morphological criteria. thylakoid structure typical of the family Phormidia-
As revised, the genus Phormidium contained those ceae (Anagnostidis and Kom
arek 1988). P. autum-
species with following taxonomic features: trichome nale and other species in Group VII share cell
width ~3–11 lm, sheaths with only one trichome, division and thylakoid characteristics of Microcoleus
cells isodiametric or shorter than wide, apical cells sensu stricto. This problem was noted by Boyer
pointed or rounded with or without calyptra, daugh- et al. (2002) and Kom
arek and Anagnostidis (2005).
ter cells dividing and growing to the size of the Of the 30 species of Microcoleus, only seven species
mother cells before the next division, trichome clearly belong to the M. vaginatus cluster (M. amplus,
disintegration facilitated by necrotic cells, without M. acremannii, M. antarcticus, M. baicalensis, M. rushf-
aerotopes, and principally from aquatic biotopes. orthii, M. subandinus, and M. terrestris). Most remain-
A relatively wide spectrum of different morphospe- ing species have strongly constricted cross walls,
cies previously described in Phormidium, Oscillatoria phormidiacean cell division, no calyptra, and end
and Lyngbya was reorganized into eight distinct mor- cells typically longer than wide, whereas a few spe-
phological groups within Phormidium. The type spe- cies differ from both these major groups (see
cies for Phormidium, P. lucidum, belongs to their Kom
arek and Anagnostidis 2005). Because M. vagin-
group VIII, and was originally described by K€ utzing atus is the generitype, these other species must be
(1843) from an aerial mat growing on the walls of removed from Microcoleus and placed in existing or
thermal springs in Karlovy Vary, Czech Republic. new genera. Indeed, some of this revisionary work
The species cluster containing the generitype has already commenced, with Siegesmund et al.
includes taxa, which have the following features: (2008) separating the M. chthonoplastes group into
short cells (usually less than 1/2 cell width), tric- Coleofasciculus. The trichomes of this genus are
homes ((4)-6–12 lm wide), and sheaths that only slightly constricted at the cross walls, not tapering
form facultatively. Some are reported to taper toward ends, with cells with strictly radial or fascicu-
abruptly at the ends and/or possess calyptra. The lated thylakoid arrangement. Cells of Coleofasciculus
genetic and ecological characterization of group are also typically longer than wide, contrary to cells
VIII (based mainly on P. irriguum, related to P. luci- of typical M. vaginatus/P. autumnale clades in which
dum) was presented recently by Sciuto et al. (2012). whole cells are usually isodiametric to shorter than
We have focused our study on the taxonomic wide.
identity of Phormidium group VII (Kom
arek and After examination of a large number of strains of
Anagnostidis 2005), which is characterized by iso- both M. vaginatus and P. autumnale from many
diametric cells, trichomes shortly attenuated toward regions of the world in various biotopes, it has
the trichome ends, and presence of calyptra. P. au- become very clear that Microcoleus must be revised
tumnale (Agardh) Trevisan ex Gomont belongs to to include several well-known Phormidium species.
the core species of this group. Representatives of The purpose of this study is to report the results of
this morphotype are common worldwide in various this work and make the necessary nomenclatural
biotopes. Due to the high degree of environmen- changes.
tally induced morphological variability in cyanobac-
teria (see Kann and Kom
arek 1970, Kom
arek 1972), MATERIAL AND METHODS
sequencing is essential for the proper taxonomic We examined 92 strains (Table S1 in the Supporting Infor-
evaluation of this species. Phylogenetic analysis sug- mation). The geographic origin of strains was worldwide;
gests that P. autumnale is very close to M. vaginatus however, the majority of strains came from the Europe, the
(Boyer et al. 2002, Siegesmund et al. 2008, Stru- Arctic, and Antarctica, a lesser amount of strains came from
neck
y et al. 2010, 2012a,b, Hasler et al. 2012). North America and Asia, only a few strains were from Africa
M. vaginatus belongs to an easily recognizable clade and South America. Microscopic observation of the morphol-
with specific ecology (soil biotope) and fasciculated ogy of the studied strains (Fig. 1 and Fig. S1 in the Supporting
 REVISION OF THE GENUS MICROCOLEUS 1169

FIG. 1. Morphological diversity within Microcoleus vaginatus sensu lato (Fig. 2, Clade A). Strains are presented in same order as in
Figure 2 (top to bottom of phylogeny) and strain numbers are embedded in the figure. All figures are to the same magnification; scale
represents 20 lm.
1170 O T A K A R S T R U N E C K Y
E T A L .

Information) together with habitat of origin confirmed their and EF667962 and EF654077 (Siegesmund et al. 2008;
classification within M. vaginatus, P. autumnale, P. setchellia- Table S1).
num, or P. favosum. Sequences for phylogenetic comparison were chosen after
Most of the strains were isolated by the authors and repre- extensive evaluation of more than 1,300 sequences of Oscilla-
sent previously unreported material. Samples from polar toriales cyanobacteria available in Genbank (www.ncbi.nlm.
regions were transported frozen to the laboratory. Samples nih.gov) and Ribosomal database project (rdp.cme.msu.edu;
from temperate or tropical localities were placed in petri unpublished). Three outgroups for construction of 16S rRNA
dishes or dried on a piece of paper. Upon returning to the gene phylogenetic tree were used: Escherichia coli ATCC
laboratory, a small quantity of each sample was placed on 11775, two representatives of simple unicellular cyanobacte-
agar solidified (1.5%) BG-11 medium (Bischoff and Bold ria, Prochlorococcus marinus MIT 9303 and Synechococcus elonga-
1963). To obtain unialgal strains, the dilution plate method tus PCC 6301, and Nostocales from various genera. Selected
was used for isolation of cyanobacteria (Elster et al. 1999). sequences were aligned in MAFFT (mafft.cbrc.jp; Katoh and
The petri dishes with strains from polar and temperate Toh 2010) considering secondary structure. Minor changes
regions were placed in an illuminated refrigerator at a were done manually with BioEdit 7.0.1 (Hall 1999). A frag-
temperature between 5°C and 15°C. Tropical strains were ment of 1,109 nt was used for the phylogenetic analysis of
cultured at room temperature. In addition to these new the 16S rRNA gene; a fragment of 1,455 nt was used for the
isolates, other Phormidium strains from CCALA and ISBAL 16S rRNA gene with the 16S-23S ITS (both starting at E. coli
were used in this study. The strains used in this study will be ATCC 11775 16S rDNA residues 302).
available upon request. Cultured strains were identified For the 16S rDNA analysis, strains of recurrent/identical
according to Kom
arek and Anagnostidis (2005). sequence belonging to our set of strains were excluded.
For morphological evaluation, the length and width of at Strains Tychonema bourrellyi CCAP 1459/11B (Suda et al.
least 50 cultivated cells were measured for every strain during 2002), T. tenue SAG 4.82, Oscillatoria sp. PCC 7112, M. rushfor-
exponential growth phase under 1,000 9 magnification using thii UTCC296, M. acremanii UTCC 313, and M. antarcticus
an Olympus (Tokyo, Japan) BX 51 microscope equipped with UTCC474 (Casamatta et al. 2005), strains of M. amoenum
DP 71 camera. (BW, Sab) and P. autumnale (CBV, CBV; Lokmer 2007);
Molecular and phylogenetic analysis. DNA from the nonaxe- M. vaginatus (SNM1-KK1, CSU-U-KK1, UBI-KK2, RS1-KK2,
nic unialgal strains was extracted using the modified method CJI-U-KK2, CSI-U-KK1, SEV1-KK3; Siegesmund et al. 2008);
of Yilmaz et al. (2009). A piece of biomass was suspended in M. vaginatus PCC 9802 (Garcia-Pichel et al. 2001) were added
50 lL of TE buffer at pH 7.4 and 750 lL of XS buffer (1% to the phylogenic comparison based on their sequence simi-
potassium ethyl xanthogenate; 100 mM Tris-HCl, pH 7.4; larity to the Microcoleus cluster and especially to M. vaginatus
20 mM EDTA, pH 8; 800 mM ammonium acetate and 1% SAG 22.11 (Boyer et al. 2002). To determine placement of
SDS) in an Eppendorf tube with glass beads. The trichomes the Microcoleus cluster among other Oscillatoriales sensu lato
were manually crushed by micropestle. Tubes were incubated the maximum likelihood method based on the Tamura
for 2 h at 70°C. After incubation, the tubes were vortexed for 3-parameter model in MEGA 5 (Tamura et al. 2007) was
30 s and frozen at 70°C for 30 min. The sample was thawed employed; topology was validated by Bayesian analysis in
and shaken for 15 min, centrifuged for 20 min at 15,000 g, MrBayes 3.1.2. (Huelsenbeck and Ronquist 2001) and
and the supernatant was transferred to a clean microcentri- unweighted maximum parsimony (MP) was implemented in
fuge tube. The DNA was precipitated overnight in a 2:3 vol- PAUP* (Swofford 2003). For the Bayesian analysis, two runs
ume of 100% ethanol with the addition of 1:20 volume of of four Markov chains over 15,000,000 generations, sampling
sodium acetate (3 M, pH 5.2) followed by centrifugation for every 1,000 generations, were employed. The trees were
30 min at 15,000 g. The supernatant was discarded and the checked to show a standard deviation of split frequencies
pellet was washed with 100 lL of 70% ethanol followed by below 0.05. The initial 25% generations were discarded as
centrifugation for 15 min. After discarding the supernatant, burn-in.
the pellet was dried and dissolved in 100 lL of miliQ water. Phylogenetic analysis of both 16S rRNA gene and associ-
The 16S rRNA gene with the 16S-23S intergenetic segment ated 16S-23S ITS was conducted by MrBayes 3.1 at www.meta-
was amplified using the primers 359F (GGGGAATYTTCCG centrum.cz, the tree topology was validated in Mega 5
CAATGGG) (N€ ubel et al. 1997) and 23S30R (CTTCGCCT (Tamura et al. 2007) using the maximum likelihood method
CTGTGTGCCTAGGT) (Wilmotte et al. 1993) with the follow- based on the Tamura three-parameter model with four cate-
ing settings: a starting denaturalization step (94°C, 5 min); 40 gories of variable sites and gamma 2.7 for 16S rRNA and 1.6
cycles of 30 s at 94°C, 30 s at 53°C, and 3 min at 72°C; final for 16S-23S ITS obtained in Modeltest (Posada and Crandall
extension for 7 min at 72°C and cooling to 4°C. A successful 1998) and unweighted MP was implemented in PAUP*.
PCR was confirmed by running a subsample on a 1.5% aga- Because direct sequencing was employed, and because Microc-
rose gel stained with ethidium bromide. PCR products were oleus contains multiple, different ribosomal operons, only the
purified using a QIAquick PCR Purification Kit. Sequencing conserved beginning of the ITS was used (beginning to the
of the 16S rRNA gene fragment was performed on an ABI end of the D3 region). The final graphic refinement of all
3100 sequencer, using BD3.1 (Applied Biosystems, Foster trees was made in Adobe illustrator CS5.
City, CA, USA) chemistry, with six primers (27F, 23S30R, Ultrastructure. For the ultrastructural studies, cultivated
CYA-810R- GTTATGGTCCAGCAAAGCGCCTTCGCCA, CYA7 biological material of strains was fixed with 6% glutaralde-
83F-TGGGATTAGATACCCCAGTAGTC (Struneck
y et al. hyde and kept in this solution for more than 4 h at room
2010), S17*-GGCTACCTTGTTACGAC, and ILE23F- ATTA temperature. After that samples were washed with 0.05 M
GCTCAGGTGGTTAG (Wilmotte and Herdman 2001) to phosphate buffer (pH 7.2) and postfixed with 2% osmium
obtain complementary sequences. The sequences that were tetrioxide in the same buffer for 2 h at room temperature.
not used previously were deposited into GenBank under After washing with 0.05 M phosphate buffer, the cells were
accession numbers KC633953–KC634001 (Table S1). Seque- dehydrated with a graded isopropanol series and embedded
nces used previously in other studies were deposited into in Spurr’s resin using propylene oxide as an intermediate
GenBank under accession numbers GQ504019–GQ504037 stage (Spurr 1969). Thin sections were stained with uracyl
(Struneck
y et al. 2010); JN230326–JN230347 (Struneck
y acetate and lead citrate and observed in a Jeol (Tokyo,
et al. 2012a); JQ769113–JQ769139 (Struneck
y et al. 2012b); Japan) JEN 1010 transmission electron microscope at 80 kV.
 REVISION OF THE GENUS MICROCOLEUS 1171

RESULTS continents, both from aquatic and terrestrial habi-

tats, and from polar regions to hot deserts. Further-
Phylogeny. The clade containing all strains of more, while individual cultures have very narrow
M. vaginatus and P. autumnale was highly supported size ranges (Table S3 in the Supporting Information),
in all phylogenetic analyses and henceforth is the morphological definition of M. vaginatus is histor-
referred to as Microcoleus sensu stricto (Fig. S2 in ically broad (e.g., 3–7 lm wide trichomes), and we
the Supporting Information). All strains in the recovered considerable morphological diversity in our
Microcoleus ingroup were at least 95.6% similar to all observations (Fig. 1). No ecological, evolutionary, or
other members of the group, with many having phylogenetic species concept would accommodate
much higher similarity to select species (>99% simi- such a garbage basket of forms. It is very clear that this
lar). The sister taxon to this clade contained P. ani- species complex requires taxonomic work, but
male CCALA 761 and Geitlerinema CCALA 138 because of the high similarity in the 16S rRNA gene
(Clade A; Fig. S1), and the similarity of Microcoleus species definition will likely require more genetic
to these strains was consistently below 95% markers.
(Table S2 in the Supporting Information). The The phylogenetic tree of the combined data set
molecular discontinuity between Microcoleus and (16S rRNA + 16S-23S ITS) better resolves the rela-
other filamentous genera is distinct, within all other tionships within Microcoleus sensu stricto (Fig. 2).
representatives of Oscillatoriales the similarity of the This phylogeny has better taxon sampling and
16S rRNA gene in Microcoleus sensu stricto strains was higher support (Bayesian posterior probabilities).
not higher than 93% (Table S2). Furthermore, the identification of clades which may
Another phylogenetically close clade to Microcoleus represent species is clearer. Clade A is the largest
(Clade B, Fig. 2) contains strains from Trichodesmium clade with 41 strains from our set. The strains of
sp. and “Oscillatoria sancta” (SAG 74.79 and PCC this clade belongs generally to strains with interme-
7515, respectively). Other species, traditionally diate widths (4–7 lm), the morphology of strains
classified in the family Phormidiaceae and identified mostly corresponds to the typical definition of
as P. tergestinum CCALA155, P. uncinatum SAG81.79, M. vaginatus. Many of these strains were isolated
P. animale CCALA140, P. pseudopriestleyi ACEV53, from soil or damp sediment adjacent a stream or
P. lumbricale UTCC476, and P. terebriforme AB20020, lake. While some were isolated from continually
are located in different parts of the phylogenetic tree sprayed sediment, it seems reasonable to conclude
(Fig. S2), again confirming the polyphyletic nature that this group contains typical M. vaginatus. It
of the genus if a traditional concept is followed. includes several desert soil crust strains from North
Perhaps most importantly, the clade likely containing America that correspond to modern concepts of the
the species closest to the generitype of Phormidium species (Garcia-Pichel et al. 2001, Boyer et al. 2002,
(P. irriguum, P. tergestinum) is very distant from Flechtner et al. 2008). All members of this clade
Microcoleus, and consequently any Phormidium in the contained the 11 bp insert. Indeed, all the strains
Microcoleus clade cannot remain in Phormidium. originally identified as M. vaginatus belong to this
The M. vaginatus and P. autumnale clade (Fig. 3) clade. However, we saw evidence of considerable
demonstrate that identification of species in the morphological divergence in the group. Strain
group cannot be made simply by morphology or SVV11 produced extensive amounts of extracellular
ecology. Originally, all aquatic members in this tree polysaccharide, and laminated sheaths were formed
were considered to be P. autumnale or some other by M. vaginatus K7-08. M. vaginatus ISBAL-M2 is
Phormidium species. The reported habitat preference intensely brown-violet, whereas strains ISBAL-M25,
of M. vaginatus for soils and preference of P. autum- ISBAL-M20, and SV30 are brownish. Morphological
nale for aquatic habitats were not congruent with heterogeneity within Clade A is remarkable (Fig. 1,
the partial 16S rRNA phylogeny. We had thought A–O) and we suppose that even higher variability
that P. autumnale would lack the 11 nucleotide could be seen in nature. We have labeled members
insert, and all prefer aquatic habitats. This was gen- of this clade M. vaginatus in both phylogenies
erally true, but many of the aquatic strains assigned (Figs. 2 and 3) and the tables (Tables S1 and S3)
to P. autumnale by their isolators actually had the recognizing that Clade A represents a species com-
11 bp insert, and thus appeared to belong to plex, not a single species.
M. vaginatus. The primary work of determining Clade B contains strains with intermediate and
species in this clade is a task for a future study. wide width of blue-green trichomes. The cells of
The phylogenetic analysis had very poor bootstrap strains within this clade are often isodiametric, occa-
support due to high similarity of strains. It was inter- sionally (after division) as short as 1/3 as long as
esting to see that the strains without the insert wide (Fig. S1, A–C). Attenuation toward the ends
formed a monophyletic group (Fig. 3). However, occurs but is inconsistent within this clade. Calyptra
species clusters in this analysis are not easily discern- varies among strains from very small caps to dis-
ible. It is clear from this analysis that M. vaginatus tinctly capitate structures. All members of the clade
is polyphyletic, as it contains M. amoenus and have the 11 bp insert. Strains fit the morphology of
M. acremanii, and includes populations from all both P. favosum Gomont ex Gomont 1892 and
1172 O T A K A R S T R U N E C K Y
E T A L .

FIG. 2. Phylogenetic relationships of Microcoleus strains used in this study, with expanded taxon sampling and longer character set.
Analysis based on a concatenated sequence of the partial 16S rRNA gene and beginning region of 16S-23S ITS conducted with MrBayes
3.1 and validated by maximum likelihood (ML) and unweighted maximum parsimony (MP). Thicker node lines represent nodes with
Bayesian posterior probabilities of 0.80–0.89, thickest node lines represent nodes with Bayesian posterior probabilities of 0.90–0.99. The
bootstrap values for ML/MP are indicated. Thickest node lines with an asterisk (//*) indicate a Bayesian posterior probability of 1.00 as
same as values 100 for ML and MP. A fragment of 1455 nt was used for this phylogenetic analysis.

P. autumnale. Given the cold stenothermal condi- (Fig. S1, D–G). Meristematic zones are a common
tions of these arctic and alpine strains, we have feature of strains from this cluster. The strains
placed them provisionally into M. favosus. belonging to cluster C were generally greenish gray
Clade C generally contains strains of intermediate and mostly polar and alpine, living on wet mud,
width from streams and wet soils with a small calyptra below snowfields, and in alpine streams. None of
 REVISION OF THE GENUS MICROCOLEUS 1173

FIG. 3. Phylogenetic relationships within Microcoleus sensu stricto (from analysis shown in Fig. S2, same analyses and parameters). Conti-
nent of origin is shown by continent map codes in front of each taxon. This tree contained forms with highly similar 16S rRNA sequences,
and consequently was poorly supported along the entire backbone of the phylogeny.
1174 O T A K A R S T R U N E C K Y
E T A L .

these strains had the 11 bp insert. The clade repre- Clade F contains strains of morphology similar to
sents one or three species, but further genetic work Clade C, with evident meristematic zones present.
is required to name the taxon/taxa to which they The former clade contains strains of freshwater and
belong. soil origin, the strains forming the latter clade are
Two strains were in a paraphyletic position to aerial. What is more interesting is that all three
Clades A–C (from Luznice and Kralovice). Although strains grow associated with high nitrogen and phos-
adjacent in the phylogeny, the two strains are not phorus enrichment. Microcoleus KG29 was isolated
that similar. Microcoleus strain Kralovice had wider from whale bone in Antarctica, whereas the other
trichomes and lacked the 11 bp insert (Fig. S1I). two strains were from populations associated with
Microcoleus strain Luznice had intermediate width bird rookeries. This distinctive ecology has been asso-
and contained the 11 bp insert, and corresponded ciated with P. attenuatum, and given the morphologi-
exactly to modern concepts of P. autumnale cal similarity of these strains with that species
(Fig. 1H, see Kom
arek and Anagnostidis 2005). (Fig. S1R), we are fairly confident that members of
These strains were both from periphyton of sluices this clade belong to M. attenuatus nov. comb.
of flowing rivers. We have chosen the Luznice strain Morphology and cytology. The thallus of all natu-
to represent M. autumnalis. ral populations, which belong to the Microcoleus/
Clade D contains strains of intermediate-to-wide P. autumnale clade, is gelatinous, forming thin and
width from streams and lakes (with one soil form: usually compact mats on the substrate or rope/band-
Microcoleus Kamenice). A distinctive calyptra could like structures. Mats from aerophytic habitats like just
be found only exceptionally (Fig. S1J), the majority above the water line of the rivers wetted by waves, in
of strains lack a calyptra (Fig. S1, K–M). The mor- periodically raised water lines or in seasonally flooded
phology and ecology of strains from this clade water-sides, form distinguishable macroscopical mats
are similar to the traditionally defined species of with twisted filament-like colonies similar to popula-
P. aerugineo-caeruleum with greenish through dark tions living in wet soil. Cultured mats form two
blue-green to blackish color of trichomes. Slight general patterns, rope-like structures growing in lines
constriction at the cross walls, vacuolization, and from the point of inoculation, or whirl-like structures.
pigmentation of Svalbard strain CCALA 846, as well Trichomes are motile and isopolar. Sheaths are thin,
as its association with strains E6 and E8 from colorless, firm, or diffluent, containing one, two, or
northern Germany suggests some similarity with many parallely arranged trichomes within one fila-
P. griseo-violaceum (Skuja) (Anagnostidis 2001) from ment. Color of the cells is usually bright, dark. or
Swedish lakes. However, the morphological proper- grayish green, occasionally brownish or yellowish,
ties of the cells can considerably vary within one exceptionally reddish. Old cultures in cultivation
strain or even within one trichome. Olive green turn brown regularly, and the color of the cells is
cells have no constrictions, whereas between grayish sometimes dependent on position in the mats with
cells of higher diameter with apparently thickened grayish/blackish green in the upper part and bright
cell walls the constrictions can be found. Given that green in the bottom, even within one trichome, sug-
these Phormidium species are not in group VII, we gesting the availability of photoprotecting substances
are only referring to members of this cluster as or a change in the phycoerythrin–phycocyanin ratio.
Microcoleus sp. Given the tightness of the clade, it The dark green keritomy (distribution of thylakoids)
may be that with further study this clade will be is usually visible in an optical microscope (Figs. 1 and
given a species epithet. All strains in the clade lack S1). Trichomes are attenuated toward the ends, and
the 11 bp insert. generally possess calyptra if mature. Usually numer-
In Clade E, the cell widths and lengths vary con- ous granules of different types are visible in cells. Ca-
siderably even between strains in which combined lyptrae are usually present in only a few trichomes
16S rRNA + 16S-23S ITS sequence was 100% identi- within the population, particularly if the filaments
cal. Microcoleus ISBAL-M26 was thinner and had a grow rapidly and disperse. Trichomes disintegrate
calyptra similar to that of P. autumnale, whereas the with the aid of necrotic cells.
other two strains had the tapering and rounded The detailed measurements for 38 strains indicate
calyptra characteristic of P. setchellianum (Fig. S1, that the mean width of cells is usually between 3
N–P). CCALA 149 resembles P. fonticolum K€ utzing and 8 lm (Table S3). Maximal and minimal widths
ex Gomont 1892, however, with respective ecology of the cells, where the measure of terminal attenua-
of P. autumnale (moist soils and rivers, respectively), tion was excluded, were within the range of 2 and
whereas CCALA 144 should be classified according 10 lm, respectively. The strains with thin trichomes
to the reddish-brown thin thallus and alpine origin with mean cell width under 4 lm are distributed
as proper P. setchellianum sensu stricto. All strains in throughout the phylogenetic tree and do not form
the clade lack the 11 bp insert. This clade is unu- any cohesive groups. The mean length of the cells
sual because the morphological diversity among the ranges from 2 to 6 lm (Table S3). Maximal and
three strains appears greater than the genetic diver- minimal cell lengths do not surpass 10 and 1 lm,
sity. The species names should be considered to be respectively. Cell width-to-length ratio ranges from
provisional. 0.9 to 2.8 with the mean value for all measured
 REVISION OF THE GENUS MICROCOLEUS 1175

strains being 1.7. However, cells can be shorter than radial arrangement of thylakoids may be present
wide, isodiametric, or exceptionally up to two times (Fig. 4E), however, in the same strain thylakoids
longer than wide, depending on the phase of can form the fasciculated packets of thylakoids
growth. On the basis of work with cultures, we (Fig. 4B). Bundles of thylakoids form a pattern
conclude that the traditional morphological charac- that can be often recognized in optical microscope
ter delimiting Microcoleus (multiple trichomes in a with areas of different color of uneven shape and
common wide sheath) is environmentally induced amount within the cells. (Fig. 4, D and F). The cell
and consequently has only limited value for defining wall envelope consists of a cytoplasmatic membrane
the genus and species within the genus. The diacrit- bounded by relatively thin wall layers of peptidogly-
ical autapomorphic markers defining the genus are can, similar to other cyanobacteria in the Oscilla-
the type of cell division, tapering trichomes and toriales. The junctional pore organelles responsible
calyptrate end cells. for slime secretion can be commonly found next
Ultrastructure as observed thus far is highly simi- to the forming septa. Carboxysomes can be
lar within the genus Microcoleus (Fig. 4). Cyanophy- grouped or solitary. Cell division is successive with
cin granules are located mainly near the cross wall formation proceeding before completion of
walls, but can be found also dispersed in the cyto- the last perpendicular binary fission is complete
plasm. Lipid droplets appear to be restricted to (Fig. 4A).
proximity with the cross walls. Thylakoids usually Taxonomic revision. Our results confirm the previ-
form fascicle-like aggregations arranged irregularly ous conclusion of Boyer et al. (2002) and Sieges-
in the cells (Fig. 4, A–D, F). In some strains a mund et al. (2008) that Microcoleus must be revised.

FIG. 4. Ultrastructure of strains of Microcoleus vaginatus. (A) Microcoleus CCALA726. (B and F) Microcoleus JR5. (C and E) M. vaginatus
CCALA861. (D) M. autumnalis Luznice. Note the agglomeration of cyanophycin and polyphosphate granules near the cross walls (A, B,
and D).
1176 O T A K A R S T R U N E C K Y
E T A L .

Phormidium group VII sensu Kom
arek and Anagnos- There are likely many species within Microcoleus
tidis (2005) is based on P. autumnale, and given the and Phormidium “group VII” that will be described
genetic and morphological closeness of P. autum- when the diversity of strains are more intensively sam-
nale with M. vaginatus, we conclude that this cluster pled and more fully characterized morphologically
of Phormidium species must be transferred into and ecologically. The transfer of P. autumnale group
Microcoleus, even though they do not produce the VII to the genus Microcoleus affects the following taxa:
multiple trichomes in a sheath which has been diag-
nostic for Microcoleus in the past. As Phormidium • Microcoleus amoenus (K€ utzing ex Gomont) Stru-
neck
y, Kom
arek et Johansen comb. nov. – Basi-
group VII does not belong to the same clade as the
onym: P. amoenum [K€ utzing, Phycol. Gener.,
generitype species of Phormidium (P. lucidum) and is
p. 192, 1843] ex Gomont, Ann. Sci. Nat. VII. Bot.
morphologically distinct from Phormidium group
16: 225, 1892. [Ecology: submerse, stagnant,
VIII to which P. lucidum belongs, it would be
unpolluted waters, often on water plants.]
nomenclaturally invalid to move M. vaginatus into
Phormidium. The taxonomic transfers into the genus • Microcoleus autumnalis (Trevisan ex Gomont) Stru-
neck
y, Kom
arek et Johansen, comb. nov. – Basionym:
Microcoleus are proposed in agreement with pre-
P. autumnale [Trevisan, Alghe Tenere Udinese, p. 20,
scriptions about later starting point for oscillatoria-
1844] ex Gomont, Ann. Sci. Nat. VII. Bot. 16: 187,
cean cyanobacteria, i.e., Gomont (1892). We here
1892; non–M. autumnalis [Rabenhorst, Deuts. Kryp-
transfer only those traditional species in group VII
tog.-Fl. 2(2):76, 1847], prestarting point name. [Ecol-
of which we are fairly confident that they belong in
ogy: streaming, mesotrophic to eutrophic waters.]
Microcoleus based on cellular and trichome morphol-
ogy: • Microcoleus attenuatus (Fritsch) Struneck
y,
Kom
arek et Johansen comb. nov. – Basionym: Lyn-
Microcoleaceae Struneck
y, Johansen et Kom
arek,
gbya attenuata Fritsch, Scot. Nat. Antarct. Exped.
fam. nov. [syn. (p.p. sine typo): subfam. Microc-
1902–04, p. 328, 1912, P. attenuatum (Fritsch) An-
oleoideae Hansgirg, Prodromus 2: 56, 1892]. Diag-
agnostidis et Kom
arek, Algol. Stud. 50–53: 407,
nosis: Familia Cyanophycearum filamentosum, sine sporis
1988. [Ecology: ornithogenic soils in Antarctica.]
heterocytisque. Filamenta solitaria, intricata vel in fascicu-
lis irregularis. Trichoma monoseriata, isopolares, plus • Microcoleus beggiatoiformis (Rabenhorst ex Gomont)
Struneck
y, Kom
arek et Johansen comb. nov. – Basi-
minusve cylindrica, sine vaginis vel cum vaginis simplic-
onym: O. beggiatoiformis [Rabenhorst, Fl. Eur. Alg. 2:
ior vel cum trichomatibus multis, fasciculatis, in vagina
99, 1865] ex Gomont, Ann. Sci. Nat. VII. Bot. 16:
una, apice attenuata vel cylindrica; cellulae apicales
235, 1892. [Ecology: mineral, alkaline springs.]
rotundate-attenuatae vel calyptratae. Cellulae plus minus-
ve isodiametricae, vel paucim longior vel brevior quam • Microcoleus calidus (Gomont ex Gomont) Stru-
neck
y, Kom
arek et Johansen comb. nov. – Basi-
latae, thylakoides irregulares vel plus minusve radialiter in
onym: P. calidum [Gomont, J. Bot. 4(20): 355,
cellulis dispositae. Reproductio fragmentatione trichomati-
1890] ex Gomont, Ann. Sci. Nat. VII. Bot. 16:
bus cum cellulis necridiis.
182, 1892. [Ecology: thermal, sulfur springs.]
Description: Family of filamentous cyanophytes
without branching, heterocytes and akinetes. Fila- • Microcoleus caucasicus (Elenkin et Kosinskaja) Stru-
neck
y, Kom
arek et Johansen comb. nov. – Basi-
ments are solitary or aggregated in irregular fasci-
onym: O. terebriformis f. caucasica Elenkin
cles and forming mats. Trichomes uniseriate,
et Kosinskaja in Elenkin, Monogr. Alg. Cyanoph.,
isopolar, more or less cylindrical without sheaths or
Pars Spec. 2: 1345, 1949. [Ecology: mineral
with simple sheaths, sometimes more trichomes
springs, high mountains.]
enveloped by one common sheath, fasciculated with
attenuate or cylindrical ends; apical cells are attenu- • Microcoleus crustaceus (Voronichin) Struneck
y,
Kom
arek et Johansen comb. nov. – Basionym:
ated and rounded, often developing distinct calyp-
P. crustaceum, Voronichin, J. Bot. URSS 17(3):
tra. Cells  isodiametrical or slightly shorter or
314, 1932. [Ecology: subaerophytic, epilithic in
longer as wide, with thylakoids located irregularly or
limestone regions.]
 radially. Reproduction by fragmentation of tric-
homes by help of necridic cells. • Microcoleus favosus (Gomont) Struneck
y, Kom
arek
et Johansen comb. nov. – Basionym: P. favosum
Type genus = Microcoleus Desmazi eres ex Gomont,
Gomont, Ann. Sci. Nat. VII. Bot. 16: 180, 1892.
Ann. Sci. Nat. VII Bot. 16: 249, 1892; Generitype =
[Ecology: oligotrophic springs and water habitats
M. vaginatus Gomont ex Gomont, Ann. Sci. Nat. VII
in mountains.]
Bot. 15: 355-356, 1892. This family comprises more
or less genera from the previous family Phormidia- • Microcoleus fonticola (K€ utzing ex Gomont) Stru-
neck
y, Kom
arek et Johansen, comb. nov. – Basi-
ceae sensu Kom
arek and Anagnostidis 2005 (pp.
onym: P. fonticola [K€ utzing, Phycol. Gener.,
340–544, subfamilies Phormidioideae and Microc-
p. 192, 1843] ex Gomont, Ann. Sci. Nat. VII. Bot.
oleoideae), but without the nomenclatoric-type spe-
16: 191, 1892; non–M. fonticola [Rabenhorst,
cies of the genus Phormidium and without taxa,
Deuts. Kryptog.-Fl., 2(2):78, 1847], prestarting
which have been transferred to other families
point name. [Ecology: cold and clear mountain
according to recent taxonomic studies (Coleofascicu-
streams and ditches.]
lus, Wilmottia, Phormidesmis, etc.).
 REVISION OF THE GENUS MICROCOLEUS 1177

• Microcoleus hiemalis (Jaag) Struneck
y, Kom
arek et strains belonging to M. vaginatus had a very
different phylogenetic placement than the clade
Johansen comb. nov. – Basionym: O. hiemalis Jaag,
Mitt. Naturf. Ges. Schaffhausen 14, 1: 117, 1938, containing the “constricted” species M. steenstrupii,
P. hiemale (Jaag) Anagnostidis, Preslia (Praha) 73: M. paludosus, and M. chthonoplastes. Siegesmund
370, 2001. [Ecology: cold, but polluted streaming et al. (2008) confirmed the polyphyly of Microcoleus,
waters, mountains.] and suggested a relationship between P. autumnale
• Microcoleus setchellianus (Gomont) Struneck
y, and M. vaginatus. Their phylogeny clearly showed
that the clade they called Oscillatoriaceae contained
Kom
arek et Johansen comb. nov. – Basionym:
P. setchellianum Gomont, Ann. Sci. Nat. VII. Bot. M. vaginatus and P. autumnale (fig. 2 in Siegesmund
16: 190, 1892. [Ecology: mountain oligotrophic to et al. 2008), whereas other Microcoleus species fell
mesotrophic springs and streams.] distant from that clade. They used this placement to
• Microcoleus subsalsus (Gomont) Struneck
y, justify the creation of Coleofasciculus chthonoplastes,
but did not resolve the confusion of Microcoleus with
Kom
arek et Johansen comb. nov. – Basionym:
P. subsalsum Gomont, Bull. Soc. Bot. France 46: P. autumnale. A number of researchers indepen-
38, 1899. [Ecology: brackish waters.] dently showed a close relationship between M. vag-
• Microcoleus wallrothii (Wallorth ex Gomont) Stru- inatus and P. autumnale (Willame et al. 2006, Comte
et al. 2007, Marquardt and Palinska 2007, Palinska
neck
y, Kom
arek et Johansen nomen nov. – Synonym:
O. violacea Wallroth ex Gomont, Ann. Sci. Nat. VII. and Marquardt 2008). Most recently, Hasler et al.
Bot. 16: 240, 1892; P. violaceum (Wallroth ex Gomont) (2012) studied an extensive number of P. autumnale
Anagnostidis, Preslia (Praha) 73: 37 2001; non–M. vio- from epipelic environments and found that they
laceus [Fremy, Arch. Bot. M
emoires 3(2):81, 1930]. had the 11 bp insert associated with M. vaginatus.
[Ecology: subaerophytic on various substrates, partly They did find differences between 16S-23S ITS
submersed, in temperate zones.] structures of soil and aquatic representatives, but
• Microcoleus vulgaris (K€ utzing ex Gomont) Stru- were unsure whether or not soil forms and aquatic
forms deserved to be recognized as one cosmopoli-
neck
y, Kom
arek et Johansen comb. nov. – Basi-
onym: P. vulgare [K€ utzing, Phycol. Gener., p. 193, tan species or multiple species within a single
1843] ex Gomont, Ann. Sci. Nat. VII. Bot. 16: genus.
188, 1892. [Ecology: soils] This study intensively sampled the P. autumnale
group and definitively demonstrated that P. autum-
To this group belong possibly also other nale and M. vaginatus (together with several other
morphologically and ecologically distinct species, Phormidium and Microcoleus) fall in a highly supported
e.g., P. paludosum, P. penicillatum, P. pseudopriestleyi, clade separate from the type forms of Phormidium
P. schroeteri, P. yellowstonense, and others. (e.g., P. lucidum, P. irriguum in Phormidium group VIII
This transfer, based on molecular analyses, is compat- sensu Kom
arek and Anagnostidis 2005). Our results
ible with morphology of trichomes and ultrastructural were congruent with all past studies, and with the
patterns. The common cytomorphological characters taxon sampling now available and the high similarity
are the structure of trichomes (with variable, but typical of morphology and ultrastructure among this now
morphology and size of cells in certain limits,  nar- well-defined group, the taxonomic resolution we have
rowed ends of trichomes, and calyptras in well-devel- recommended is certainly justified.
oped trichome ends), and life strategy in mats. On the However, further revision of traditional genera
other hand, the character of one or more trichomes Microcoleus and Phormidium is still required. The
arranged in a single sheath, which was considered to be genus now likely contains the species that constitute a
the diagnostic generic feature of the genus Microcoleus clade with a shared common ancestor with the gener-
as formerly defined, is found to be no longer a neces- itype, but it also includes many species that do not
sary character. It is only a specific marker. share that common ancestor. The Microcoleus species
Also the ultrastructure patterns are characteristic with constricted cross walls and pseudanabaenacean
and unique for all representatives so far studied cell division (Anagnostidis and Kom
arek 1988) must
ultrastructurally within the revised genus Microcoleus. be removed from the genus and placed in other gen-
The irregularly fasciculated thylakoids within the era. This work has commenced with the establish-
cells are characteristic for this clade. Successive cell ment of the Pseudanabaenacean genus Trichocoleus
division is also a feature that appears to be consis- (Anagnostidis 2001) and the Phormidiacean genus
tent in the genus. The 11 bp insert in the 16S rRNA Coleofasciculus (Siegesmund et al. 2008). However,
gene is a sufficient but unnecessary character for the majority of freshwater Phormidiacean Microcoleus
definition of Microcoleus. have still to be revised. This group includes M. glacei,
M. paludosus, M. sociatus, M. steenstrupii, M. subtorulo-
DISCUSSION sus, and M. lacustris. We suspect these species eventu-
ally will be transferred into multiple genera.
Boyer et al. (2002) suggested that Microcoleus in The redefined family Microcoleaceae comprises a
the traditional concept (cf. Geitler 1932, Geitler large cluster of both marine and freshwater benthic,
1942) was in need of revision. They found that the periphytic, pelagic, and terrestrial cyanobacteria
1178 O T A K A R S T R U N E C K Y
E T A L .

with simple, unbranched filaments 3–15 lm wide, Bischoff, H. W. & Bold, H. C. 1963. Some Soil Algae from Enchanted
without heterocytes or akinetes, and with fascicu- Rock and Related Algal Species. Phycological Studies IV. Univer-
sity of Texas Publications, Austin, 95pp.
lated thylakoids in cells. The most proximate clade Boyer, S. L., Johansen, J. R., Flechtner, V. R. & Howard, G. L.
to Microcoleus according to molecular phylogeny (and 2002. Phylogeny and genetic variance in terrestrial Microcole-
therefore in Microcoleaceae) consists of cultured us (Cyanophyceae) species based on sequence analysis of the
strains “Oscillatoria lutea” UTEX 390, P. animale SAG 16s rRNA gene and associated 16s-23s ITS region. J. Phycol.
38:1222–35.
1459-6, Geitlerinema sp. strains CCALA 138 and Casamatta, D. A., Johansen, J. R., Vis, M. L. & Broadwater, S. T.
CCALA 139, and uncultured strain GBII. Strain 2005. Molecular and morphological characterization of
UTEX 390 shares similar morphology with similar ten polar and near-polar strains within the Oscillatoriales
thylakoid arrangement according to accessible (Cyanobacteria). J. Phycol. 41:421–38.

Comte, K., Saback
a, M., Carre-Mlouka, A., Elster, J. & Kom
arek, J.
microphotographs from the UTEX culture collec-
2007. Relationships between the Arctic and the Antarctic
tion. Other neighbors on the phylogenetic tree cyanobacteria; three Phormidium-like strains evaluated by a
include strains from the genus Trichodesmium, which polyphasic approach. FEMS Microbiol. Ecol. 59:366–76.
even possess calyptra formation in some strains. Drouet, F. 1968. Revision of the classification of the Oscillatoria-
The recognition of species-level diversity within ceae. Monogr. Acad. Nat. Sci. Phil. 15:1–370.
Elenkin, A. A. 1949. Monographia algarum cyanophycearum aquidul-
Microcoleus is presently poorly resolved. The species, cium et terrestrium in finibus URSS inventarum. izd. AN SSSR,
which form “Phormidium group VII,” were based Moskva-Leningrad, 1908 pp.
mostly on ecology, and there are transitions in mor- Elster, J., Lukesov
a, A., Svoboda, J., Kopeck
y, J. & Kanda, H.
phological markers in these previously described taxa. 1999. Diversity and abundance of soil algae in the polar des-
Our study revealed a number of clades that likely ert, Sverdrup Pass central Ellesmere Island. Polar Record
35:231–54.
represent species, and some correspond closely to Ercegovic, A. 1932. Ekoloske i socioloske studije o litofitskim cij-
described species, whereas others share ecology but anoficejama sa jugoslovenske obale Jadrana. Rad Jugoslov.
lack species affiliation at present (Fig. 2). Our results Akad. Znan. Umjet. Zagreb 244:129–220.
suggest that some aspects of biogeography are critical. Flechtner, V. R., Johansen, J. R. & Belnap, J. 2008. The biological soil
crusts of the San Nicolas Island: enigmatic algae from a geo-
In particular, we saw that polar strains or desert soil graphically isolated ecosystem. West. N. Am. Nat. 68:405–36.
strains often formed distinct groups. It was clear in Frémy, P. 1934. Cyanophycées des côtes d’Éurope. Mem. Soc. Nat.
this study that species clusters only began to make Sci. Nat. Math Cherbourg 41:1–234.
sense when habitat preference, biogeographical Fritsch, F. E. 1912. Freshwater Algæ collected in the South Ork-
region, and phylogeny were all considered together. neys by Mr. R. N. Rudmose Brown, B.Sc, of the Scottish
National Antarctic Expedition, 1902–04. Bot. J. Linn. Soc.
The infrageneric diversity of the revised Microcoleus 40:293–338.
needs further study, and we expect that a polyphasic Garcia-Pichel, F., Lopez-Cortes, A. & N€ ubel, U. 2001. Phylogenetic
approach using detailed morphology, ecological char- and morphological diversity of cyanobacteria in soil desert
acterization, biogeography, 16S rRNA phylogeny, full crusts from the Colorado plateau. Appl. Environ. Microbiol.
67:1902–10.
characterization of the 16S-23S ITS region, and Geitler, L. 1932. Cyanophyceae. In Rabenhorst, L. [Ed.] Kryptoga-
sequencing of additional genetic markers will permit men-flora von Deutschland, Osterreich und der Schweiz. Vol. 14.
diagnosis of taxa in a combination of existing and Akademishe Verlagsgesellschaft, Leipzip, 1196 pp.
new species. We are hopeful and even confident that Geitler, L. 1942. Schizophyta (Klasse Schizophyceae). In Engler,
a revised taxonomy of monophyletic species within €
A. & Prantl, K. [Eds.] Naturliche Pflanzenfamilien 1b. Duncker
& Humblot, Berlin, pp. 1–232.
monophyletic genera will eventually be realized. Golubic, S. 1960. On cyanophyte vegetation in north Adriatic har-
Ecospecies similar to the genus Microcoleus or Pho- bors. Thalassia Jugoslavica 2:5––36.
rmidium group VII from marine or brackish origin, Gomont, M. 1890. Essai de classification des Nostocac
ees homo-
such as M. codii Fr
emy 1934, P. litorale Golubic 1960, cyst
ees. J. Bot. 4:349–57.
Gomont, M. 1892. Monographie des Oscillari
ees (Nostocac
ees ho-
P. submembranaceum Gomont 1892, P. endolithicum mocyst
ees). Annal. Sci. Nat. Bot. 7th Ser. 16:91–264.
Ercegovic 1932, P. nigro-viride (Thwaites ex Gomont) Gomont, M. 1899. Sur quelques Oscillariees nouvelles. Bull. Soc.
Anagnostidis and Kom
arek (1988), and others were Bot. France 46:25–41.
not included within our set of strains and these spe- Hall, T. A. 1999. BioEdit: a user-friendly biological sequence
cies are not yet available in GenBank. It is possible alignment editor and analysis program for Windows 95//
98//NT. Nucleic Acids Symp. Ser. 41:95–8.
that those ecospecies do not belong to the genus Hansgirg, A. 1892. Prodromus der Algenflora von B€ o hmen, Part II.
Microcoleus and that also require revision and trans- Komissons-Verlag von Fr, Rinn
 ac, Prague, 268 pp.
fer to other genera. Hasler, P., Dvor
ak, P., Johansen, J. R., Kitner, M., Ondrej, V. &
Poul
ıckov
a, A. 2012. Morphological and molecular study of
This study was supported by grants supports INGO LA 341, epipelic filamentous genera Phormidium, Microcoleus and Geitleri-
LM2010009 CzechPolar, KONTAKT ME 934, and AV0Z600 nema (Oscillatoriales, Cyanophyta/Cyanobacteria). Fottea 12:341–
56.
50516, and P506/12/1818. We are indebted particularly to J.
  Huelsenbeck, J. P. & Ronquist, F. 2001. MRBAYES: Bayesian infer-
Snokhousov
a and D. Svehlov

a for technical help.
ence of phylogenetic trees. Bioinformatics 17:754–5.
Jaag, O. 1938. Reisebriefe aus Java. Mitt. Naturf. Ges. Schaffhausen
Anagnostidis, K. 2001. Nomenclatural changes in cyanoprokaryot- 14:173–83.
ic order Oscillatoriales. Preslia 73:359–75. Kann, E. & Kom
arek, J. 1970. Systematisch-€ okologische Bemerk-
Anagnostidis, K. & Kom
arek, J. 1988. Modern approach to the ungen zu den Arten des Formenkreises Phormidium autum-
classification system of cyanophytes 3 - Oscillatoriales. Algol. nale. Schweiz. Z. Hydrol. 32:495–518.
Stud. 50–53:327–478.
 REVISION OF THE GENUS MICROCOLEUS 1179

Katoh, K. & Toh, H. 2010. Parallelization of the MAFFT Turicchia, S., Ventura, S., Kom
arkov
a, J. & Kom
arek, J. 2009. Tax-
multiple sequence alignment program. Bioinformatics onomic evaluation of cyanobacterial microflora from alkaline
26:1899–900. marshes of northern Belize. 2. Diversity of oscillatorialean
Kom
arek, J. 1972. Temperaturbedingte morphologische Variabi- genera. Nova Hedwigia 89:165–200.
lit€at bei drei Phormidium -Arten in Kulturen. Preslia 44: Voronichin, N. N. 1932. K biologii soljanych ozer Sibiri. Trudy
293–307. Bot. Inst. Akad. Nauk SSSR 25:435–48.
Kom
arek, J. & Anagnostidis, K. 2005. Cyanoprokaryota -2. Teil/2nd Willame, R., Boutte, C., Grubisic, S., Wilmotte, A., Komarek, J. &
Part: Oscillatoriales. Elsevier/Spektrum, Heidelberg, 759 pp. Hoffmann, L. 2006. Morphological and molecular characteri-
K€
utzing, F. T. 1843. Phycologia generalis oder Anatomie, Physiologie zation of planktonic cyanobacteria from Belgium and Lux-
und Systemkunde der Tange. F.A. Brockhaus, Leipzig, 458 pp. embourg. J. Phycol. 42:1312–32.
Lokmer, A. 2007. Polyphasic approach to the taxonomy of the Wilmotte, A. & Herdman, M. 2001. Phylogenetic relationships
selected oscillatorian strains (Cyanobacteria). Mgr thesis, among cyanobacteria based on 16S rRNA sequences. In
University of South Bohemia, Česk
e Budějovice, Czech Boone, D. R. & Castenholz, R. W. [Eds.] Bergey’s Manual of
Republic, 61 pp. Systematic Bacteriology. Springer, New York, pp. 487–93.
Marquardt, J. & Palinska, K. A. 2007. Genotypic and phenotypic Wilmotte, A., Van der Auwera, G. & De Wachter, R. 1993. Struc-
diversity of cyanobacteria assigned to the genus Phormidium ture of the 16 S ribosomal RNA of the thermophilic cyano-
(Oscillatoriales) from different habitats and geographical bacterium Chlorogloeopsis HTF (‘Mastigocladus laminosus HTF’)
sites. Arch. Microbiol. 187:397–413. strain PCC7518, and phylogenetic analysis. FEBS Lett. 317:96–
N€ubel, U., Garcia-Pichel, F. & Muyzer, G. 1997. PCR primers to 100.
amplify 16S rRNA genes from cyanobacteria. Appl. Environ. Yilmaz, M., Phlips, E. J. & Tillett, D. 2009. Improved methods for
Microbiol. 63:3327–32. the isolation of cyanobacterial dna from environmental sam-
Palinska, K. & Marquardt, J. 2008. Genotypic and phenotypic ples. J. Phycol. 45:517–21.
analysis of strains assigned to the widespread cyanobacterial
morphospecies Phormidium autumnale (Oscillatoriales). Arch.
Microbiol. 189:325–35.
Posada, D. & Crandall, K. 1998. MODELTEST: testing the model
Supporting Information
of DNA substitution. Bioinformatics 14:817–8.
Rabenhorst, L. 1847. Deutschlands Kryptogamen-Flora II, II. Abtei- Additional Supporting Information may be
lung: Algen. E. Kummer, Leipzig, 216 pp.
Rabenhorst, L. 1865. Flora Europaea Algarum aquae dulcis et submari-
found in the online version of this article at the
nae. Sectio II. Algas phycochromaceas complectens cum figuris Gene- publisher’s web site:
rum omnium xylographice impressis. E. Kummer, Leipzig, 319
pp. Figure S1. Morphological diversity within
Sciuto, K., Andreoli, C., Rascio, N., La Rocca, N. & Moro, I. 2012. Microcoleus Clades B–F (Fig. 2). Strains are pre-
Polyphasic approach and typification of selected Phormidium sented in same order as in Figure 2 (top to bot-
strains (Cyanobacteria). Cladistics 28:357–74. tom of phylogeny) and strain numbers are
Siegesmund, M. A., Johansen, J. R., Karsten, U. & Friedl, T. 2008.
Coleofasciculus gen. nov. (Cyanobacteria): morphological and
embedded in the figure. (A–C) Clade B. (D–G)
molecular criteria for revision of the genus Microcoleus Go- Clade C. (H) M. autumnalis Luznice. (I) Microcole-
mont. J. Phycol. 44:1572–85. us Kralovice. (J–M) Clade D. (N–P) Clade E. (Q
Spurr, A. R. 1969. A low-viscosity epoxy resin embedding medium and R) Clade F. All figures are to the same mag-
for electron microscopy. J. Ultrastruct. Res. 26:31–43. nification, scale represents 20 lm.
Struneck
y, O., Elster, J. & Kom
arek, J. 2010. Phylogenetic rela-
tionships between geographically separate Phormidium cyano- Figure S2. Phylogenetic relationships within the
bacteria: is there a link between north and south polar Oscillatoriales, showing the relationship of Microc-
regions? Polar Biol. 33:1419–28.
Struneck
y, O., Elster, J. & Kom
arek, J. 2011. Taxonomic revision oleus sensu stricto within the order Oscillatoriales.
of the fresh–water cyanobacterium “Phormidium” murrayi = Maximum likelihood method based on the Tam-
Wilmottia murrayi. Fottea 11:57–71. ura three-parameter model in MEGA 5 was
Struneck
y, O., Elster, J. & Kom
arek, J. 2012b. Molecular clock evi- employed. Topology was validated by unweighted
dence for survival of Antarctic cyanobacteria (Oscillatoriales,
Phormidium autumnale) from Paleozoic times. FEMS Microbiol.
maximum parsimony in PAUP*. Numbers indi-
Ecol. 82:482–90. cate bootstrap values for ML and MP (1,000 repli-
Struneck
y, O., Kom
arek, J. & Elster, J. 2012a. Biogeography of cates) and bayesian posterior probabilities.
Phormidium autumnale (Oscillatoriales, Cyanobacteria) in wes- Asterisk (*) indicates a value 100 for ML and MP
tern and central parts of Spitsbergen. Polish Polar Res. or Bayesian posterior probability of 1.00. Values
33:369–82.
Suda, S., Watanabe, M. M., Otsuka, S., Mahakahant, A., Yongman- lower than 50% are not shown. A fragment of
itchai, W., Nopartnaraporn, N., Liu, Y. & Day, J. G. 2002. 1109 nt was used for the phylogenetic analysis of
Taxonomic revision of water-bloom-forming species of oscill- the 16S rRNA gene.
atorioid cyanobacteria. Int. J. Syst. Evol. Microbiol. 52:1577–95.
Swofford, D. 2003. PAUP* v.4beta Phylogenetic analysis using parsi- Table S1. Geographic origin, collection habitat,
mony (*and other methods) Version 4. Sinauer Associates, Sun- and year of collection for all strains sequenced by
derland, Massachusetts. us (taxon name adopted from current work, not
Tamura, K., Dudley, J., Nei, M. & Kumar, S. 2007. MEGA4: Molec-
ular Evolutionary Genetics Analysis (MEGA) software version historical designation at time of isolation). Strains
4.0. Mol. Biol. Evol. 24:1596–9. are arranged in the order they appear in Figure 5
Taton, A., Grubisic, S., Balthasart, P., Hodgson, D. A., Laybourn- (top to bottom).
Parry, J. & Wilmotte, A. 2006. Biogeographical distribution
and ecological ranges of benthic cyanobacteria in East Ant- Table S2. Percent similarity matrix (P-distance)
arctic lakes. FEMS Microbiol. Ecol. 57:272–89. of partial 16S rRNA gene sequence (1,072 bp
Trevisan, V. B. A. 1844. Le Alghe del tenere Udinese denominate e desc- compared) for Microcoleus vaginatus SAG 22.11
ritta. Coi tipi del Seminario, Padua, Italy, 24 pp.
1180 O T A K A R S T R U N E C K Y
E T A L .

and selected oscillatorialean strains outside of the Table S3. Cellular dimensions for a sample of
Microcoleus sensu stricto clade. Only the species Microcoleus species, arranged in same order (top
pairs Phormidium animale CCALA761: Geitlerinema to bottom) as in Figure 5, clades (Cl.) represent
sp. CCALA138 and P. pseudopristleyi ACEV53: putative species (Fig. 5).
P. animale CCALA140 could reliably be placed in
the same genera (and for both pairs, possibly the
same species), demonstrating the error-prone tax-
onomy reported for cyanobacteria in GenBank.