Hidden genetic diversity in the green alga Spirogyra (Zygnematophyceae, Streptophyta)

Background The unbranched filamentous green alga Spirogyra (Streptophyta, Zygnemataceae) is easily recognizable based on its vegetative morphology, which shows one to several spiral chloroplasts. This simple structure falsely points to a low genetic diversity: Spirogyra is commonly excluded from phylogenetic analyses because the genus is known as a long-branch taxon caused by a high evolutionary rate. Results We focused on this genetic diversity and sequenced 130 Spirogyra small subunit nuclear ribosomal DNA (SSU rDNA) strands of different origin. The resulting SSU rDNA sequences were used for phylogenetic analyses using complex evolutionary models (posterior probability, maximum likelihood, neighbor joining, and maximum parsimony methods). The sequences were between 1672 and 1779 nucleotides long. Sequence comparisons revealed 53 individual clones, but our results still support monophyly of the genus. Our data set did not contain a single slow-evolving taxon that would have been placed on a shorter branch compared to the remaining sequences. Out of 130 accessions analyzed, 72 showed a secondary loss of the 1506 group I intron, which formed a long-branched group within the genus. The phylogenetic relationship to the genus Spirotaenia was not resolved satisfactorily. The genetic distance within the genus Spirogyra exceeded the distances measured within any other genus of the remaining Zygnemataceae included in this study. Conclusion Overall, we define eight distinct clades of Spirogyra, one of them including the genus Sirogonium. A large number of non-homoplasious synapomorphies (NHS; 114 NHS in total) was found for Spirogyra (41 NHS) and for each clade (totaling 73 NHS). This emphasizes the high genetic diversity of this genus and the distance to the remaining Zygnematophyceae.


Background
The genus Spirogyra is a member of the Zygnemataceae (Zygnematophyceae, Streptophyta). It comprises unbranched, filamentous green algae that are characterized by spirally coiled chloroplasts and sexual reproduction by means of conjugation. Spirogyra is commonly found in stagnant or slowly flowing freshwater habitats all over the world [1,2]. It is sometimes referred to as an alga of roadside ditches and is frequently used in introductory biology courses [3] because it often occurs in huge abundances and is easy to address at the generic level. Species definition is mainly based on hypnozygote (also known as zygospores) morphology because the simple morphology in its vegetative state does not permit species recognition. In the latest monograph of Spirogyra published by Kadlubowska [4], 386 species are included. They were described using morphological traits, many of them based on a single finding. Ashraf and Godward [5] suggested that the mesospore wall structure analyzed using scanning electron microscopy should be added to the species descriptions because the taxonomy of Spirogyra at the light microscopical level remains confusing due to overlapping morphological traits [6]. The morphological species concept, which is also applied in Spirogyra, is not proven to represent true biological species, nor does it provide any information on the ecological or genetic diversity in a genus. It also does not elucidate the phylogenetic relationships between taxa [7,8]. Accordingly, the diversity of a genus remains unclear when estimates are based a single species concept. The problems arising for Spirogyra from findings without ripe hypnozygotes and the low success rate in inducing conjugation [9][10][11][12][13] call for other ways of addressing the issue of species delimitation and identification.
The Zygnematophyceae (Viridiplantae) represent the most species-rich lineage in the Streptophyta except for the embryophytic land plants [14]. Conjugating green algae including the orders Desmidiales (Desmidiaceae, Peniaceae, Closteriaceae) and Zygnematales (Mesotaeniaceae, Zygnemataceae) form a unique and distinct group. Its taxonomic and phylogenetic separation from other algae is definitive [15][16][17][18], but relationships within this group have undergone numerous rearrangements and still remain unclear. The classification schemes within the Zygnematophyceae have been based on morphological traits such as cell size, wall structure, cellular organization or chloroplast structure [1,14,[19][20][21], approaches that have been criticized in the past [22,23]. The Zygnematales are distinguished from the Desmidiales by a smooth cell wall consisting of a single piece, lacking pores and ornamentations [19,21], whereas Desmidiales have cell walls consisting of more than one piece with pores and ornaments [24,25]. Filamentous forms are grouped in the family Zygnemataceae; the unicellular taxa form the family Mesotaeniaceae [4,14,21]. This classification, however, is artificial because polyphyly of both families has been proven by phylogenetic analyses [20,22].
It remains unclear which growth form is primary and which derived [15]. West [26,27] described the ancestral state as filamentous, evolving towards unicellular forms, but Yamagishi [28] stated the opposite. Since the introduction of molecular markers, efforts have been made to solve this question [20,23,29], but the position of the genus Spirogyra within the Zygnematophyceae is not fully resolved. The evolutionary rate is one possible reason for this problem: it differs considerably among Zygnemataceaen genera [30]. The uncertain position of Spirogyra in phylogenetic analyses is also attributed to the long-branch attraction (LBA) phenomenon [31]. Some genera of the Zygnematophyceae originally defined based on morphology have been revealed as artificial based on molecular markers (e.g., Staurastrum; [32]). Furthermore, small subunit nuclear ribosomal DNA (SSU rDNA) phylogenetic studies have often suffered from limited taxon sampling [20]. When genera are represented by just one taxon, authors are unable to address the monophyly of the phylogenetic groups, either at the generic or at higher taxonomic levels. A low number of species within a genus also hinders proving monophyly [33]. In order to assess the monophyly of the genus Spirogyra and to investigate whether the low diversity of its vegetative morphology is also reflected in molecular data, we sampled 130 strains of different origin. The position within the Zygnematophyceae and its long-branch position were evaluated by calculating phylogenetic trees with complex evolutionary models. Minimizing the Spirogyra LBA problem will also help define phylogenetic relationships among genera. Additionally, we searched for Spirogyra taxa with slower evolutionary rates by including isolates from a broad spatial and ecological range and different vegetative morphology. Sampling locations were chosen to cover different types of water bodies in various areas [2]. Morphologically different Spirogyra filaments were isolated and cultivated for later use to check if the morphological differences are also reflected in phylogenetic groups. We focused on sampling the morphological diversity of the genus. SSU rRNA was chosen over ribulosebisphosphate carboxylase large subunit gene (rbcL) because other studies already demonstrated the poor resolution of this marker in the Zygnematales and Desmidiales [3,15].

Molecular phylogenetic analyses
The 130 Spirogyra sequences formed a monophyletic group. This clade, including Sirogonium, was subdivided into eight individual sub-clades A to H. Molecular phylogenies were inferred from two data sets, one combining Zygnematophyceae and Spirogyra alignment and one alignment comprising only Spirogyra and Sirogonium sequences (Figures 1, 2). In the phylogeny inferred from the combined Zygnematophyceae alignment (Figure 1), very high bootstrap support was given for branches within the Spirogyra clade. Only few branches were without support, one indicating a possible polytomy for clade C and another one indicating the lack of support for phylogenetic resolution between clades B and C, and D to H. Only very closely related taxa received high support by bootstrap values, e.g., the Zygnema clade, Sphaerozosma/Cosmocladium, Gonatozygon/Genicularia and Closterium/ Cosmarium. The remaining branches lacked support of at least one method. The individual clades of Spirotaenia and Spirogyra showed very high bootstrap support from all algorithms. The Desmidiales clade was moderately supported (posterior probabilities/maximum likelihood/ neighbor joining/maximum parsimony (PP/ML/NJ/MP): -/50/100/58); the Zygnematales clade received no support at all. Sirogonium sticticum was placed within the Spirogyra clade C. None of the algal families analyzed here formed exclusive clades.
When using an outgroup (Klebsormidium flaccidum and Coleochaete scutata; Figure 1a), the Spirotaenia clade was relocated basal to the rest of the Zygnematophyceae; the Spirogyra/Sirogonium clade formed a sister clade to the remaining Zygnematales and Desmidiales. The overall length of the branches and the classification of taxa to phylogenetic groups did not change. Within the clades, only few rearrangements could be observed in branches that received moderate or poor bootstrap/ bayesian support in Figure 1 (data not shown).When testing tree topologies for the combined Zygnematophyceae -Spirogyra alignment, the "best tree" derived from ML analysis (same as phylogeny in Figure 1) was not the overall best tree (Additional file 1: Table S1). The original ML tree (tree 1) and the user defined (UD) tree with Spirogyra relocated outside in ancestral position to the clades formed by the Desmidiales and the Zygnematales (tree 2) had the same likelihood and same Bayesian posterior probability values. These two trees were the only ones not rejected by the approximately unbiased test (AU); the tree representing Spirogyra as a sister to a Zygnematales clade (including the Spirotaenia clade; tree 3) was rejected by all tests except Shimodaira-Hasegawa test (SH) and weighted SH (WSH); Spirogyra within the Zygnematales clade (tree 4) Figure 1 Combined Zygnematophyceae and Spirogyra SSU rRNA phylogeny: Molecular phylogeny of Desmidiaceae, Peniaceae, Closteriaceae, Mesotaeniaceae and Zygnemataceae based on SSU alignment. The phylogenetic tree was inferred by maximum likelihood analyses of 1720 aligned positions of 33 taxa using PAUP* 4.0b10. TrN+G+I was chosen as best evolutionary model (base frequencies: A 0.25, C 0.23, G 0.27, T 0.25; rate matrix: A-C 1.0000, A-G 1.8721, A-T 1.0000, C-G 1.0000, C-T 4.5252, G-T 1.0000) with the proportion of invariable sites (I= 0.4608) and gamma distribution parameter (G= 0.6376). Posterior Probabilities (>95%; bold; calculated by MrBayes 3.1.2 using the covariation settings (3 million generations, trees from 4100 -30000)) as well as bootstrap values (>50%) of the maximum likelihood (100 replicates; bold italic), neighbor-joining (1000 replicates; italic), and maximum parsimony (1000 replicates; not bold) are given in the tree (PP/ML/NJ/MP). No outgroup was used. ATT4, 5, 7 and 8 refer to alternative tree topologies tested with conselplease refer to Table 1 for details. Figure 1a: upper right corner: Combined Zygnematophyceae and Spirogyra SSU rRNA phylogeny (same as Figure 1) using 2 taxa as outgroup; numbers in brackets indicate number of taxa included in groups.
was also rejected by all tests except SH. All other trees were significantly worse than the best tree at p ≤ 0.05.
In the Spirogyra phylogenetic tree (Figure 2), three sequences could not be placed within any clade: UTEX 1742, UTEX 1745 and 7075 share only a small portion of the identification patterns in base composition with adjacent clades. The major clades received very high bootstrap support, except for Clade E that was not supported by ML. The support for branches within clade B was poor due to high sequence similarity. Taxa with a secondary loss of the group I IC1 intron (marked in Figure 1 & 2) were clearly separated from the taxa containing the intron (clades A to D with intron; UTEX 1742, 7075, UTEX 1745, clades E to H without intron). The placement of Sirogonium in any other clade yielded significantly worse trees in all cases (tested with consel, Additional file 2: Table S2). Also, the relocation of the sequences previously not included in the clades into an adjoining clade was rejected with only one exception: tree 2 (relocation of UTEX 1742 and 7075 into clade E) was not rejected by SH test.

Sequence similarities in Spirogyra
130 Spirogyra nuclear encoded SSU rRNA sequences of strains from 79 different sites were sequenced for this study and, in total, 53 different SSU rDNA types (clones) Figure 2 Spirogyra SSU rRNA phylogeny: Unrooted molecular phylogeny of Spirogyra based on SSU alignment. Individual clades are highlighted in white boxes; group with secondary loss of group I IC1 intron highlighted by light green-box. The phylogenetic tree was inferred by maximum likelihood analyses of 1645 aligned positions of 55 taxa using PAUP* 4.0b10. GTR+G+I was chosen as best evolutionary model (base frequencies: A 0.23, C 0.24, G 0.29, T 0.24; rate matrix: A-C 1.4341, A-G 2.6641, A-T 1.2357, C-G 1.6993, C-T 5.2526, G-T 1.0000) with the proportion of invariable sites (I= 0.6009) and gamma distribution parameter (G= 0.6856). Posterior Probabilities (>95%; bold; calculated by MrBayes 3.1.2 using the covariation settings (2 million generations, trees from 11070 -20000)) as well as bootstrap values (>50%) of the maximum likelihood (100 replicates; bold italic), neighbor-joining (1000 replicates; italic), and maximum parsimony (1000 replicates; not bold) are given in the tree (PP/ML/ NJ/MP). No outgroup was used, tree was rooted using clade A.
were identified. Thirty sequence types were found once; the remaining 23 were represented by up to 11 accessions. Thirty-eight clones were found only at a single site and 19 more were obtained from up to 6 different sampling sites; 47 sites were represented by just one accession, 32 with up to 10 accessions. Sixty-seven sampling sites exhibited just one clone and 12 had up to 3 different clones.
To describe the genetic variability among the discovered eight lineages (clade A-H; see above), the minimal distance and pair-wise differences were calculated in PAUP. The minimal distance ( Table 1, right top) between two clades was 5.09% (clade B and D), the maximum 16.74% (clade A and H), which is 1.57 times higher than the highest value found in the remaining algal groups included in this study. This means that the within-genus difference in Spirogyra exceeds the differences among the remaining genera. The highest withinclade distance was observed in clade H (5.17%), followed by clade C (4.31%), whereas the other clades had comparatively low values from 0.25% to 1.23%. Clade A had the lowest distance value to the Zygnemataceae and Mesotaeniaceae used in our analyses (Zygn.: 22.47%; Desm.: 21.44%), while clade H showed the highest distances (Zygn.: 25.63%; Desm.: 25.71%). The pair-wise differences followed the same pattern (Table 1, left bottom) the biggest difference within Spirogyra was recognized between clade A and H (272 nucleotides (nt) difference); the biggest difference exhibited was clade H to the remaining Zygnemataceae (417 nt difference).

Evolutionary rates in Zygnematophyceae
To test the evolutionary rates among the Zygnematophyceae, the evolutionary models of different data sets were tested by Modeltest. As shown in Table 2, the data sets revealed major differences in base composition. Compared with other Zygnematalean taxa, the G/C content of Spirogyra is elevated. Additionally, high variability of sequences is indicated by a lower portion of constant nt and a different pattern in base substitution rates ( Table 2). 'Zygnemataceae + selected Spirogyra' and 'Spirogyra' represent the same data sets used for phylogenetic analyses; other data sets were obtained by modifying the previous by exclusion of certain taxa ( Table 2). The biggest difference in G/C content (0.0586 units) occurred between the Zygnematophyceae data set excluding Spirogyra and the Spirogyra data set, pointing out the disparity between the two groups (for respective values see Table 2). Interestingly, the lowest C to T substitution rate was found in the data set used for phylogenetic analyses comprising Zygnematophyceae and selected Spirogyra (4.5); the highest value was calculated for the Zygnematophyceae without Spirogyra and Spirotaenia (7.2). In the Spirogyra data set, this value is less elevated (5.3) compared to the other substitution rates. Evolutionary rates were inferred by pair-wise comparison of unambiguously aligned positions of an rRNA SSU alignment of all sequences used in this study in GRate (Table 3 & Additional file 3: Table S3). The comparison was calculated among the genera (Table 3) and among the individual sequences of the alignment (Additional file 3: Table S3). The genus Spirogyra showed significant differences to all other Zygnematophyceaen genera (Table 3); the separate clades also revealed highly different values compared to the other Zygnematophyceae (data not shown). For the remaining genera, the picture was less clearthe evolutionary rates of some taxa such as Mesotaeniaceae, Desmidiaceae and Peniaceae did not differ from each other, but did differ from Closteriaceae. Analyses of the individual sequences (Additional file 3: Table S3) revealed that the evolutionary rates of all Spirogyra sequences differed significantly from all other Zygnematophyceaen sequences. The one exception was KRA2 from clade A: it differed significantly from all Spirogyra sequences, but only from two of the Zygnematophyceaen sequences; all other differences could not be distinguished statistically. Within genera, insignificant values prevailed. The same holds true for Spirogyra clades with one exception: Spirogyra clade G showed significant differences among its sequences. Most of the disparities among non-Spirogyra sequences were not statistically relevant; differences among the Spirogyra sequences of clades B to D were mostly not significant, whereas the differences to the clades E to H (representing accessions not containing the 1506 group I intron) were mostly significant.

Secondary structure and NHS
To discover the variable positions (compensatory base changes (CBCs), hemi-CBCs (HCBCs), and non- Data sets used: 1. Zygnematophyceae and 12 Spirogyra sequences (same as in combined SSU alignment used in Figure 1). 2. Zygnematophyceae. 3. All Spirogyra sequences (same as Spirogyra alignment used in Figure 2). 4. 12 Spirogyra sequences (same Spirogyra sequences as data set 1.). 5. Zygnematophyceae without Spirotaenia. 6. Spirotaenia.  homoplasious synapomorphies (NHSs)) in the SSU rDNA of Spirogyra, the secondary structures of all strains were compared. No major changes were found in the overall secondary structure of the SSU rDNA (Figure 3). Variable parts (denoted by lower case letters in Figure 3) are predominantly peripheral regions such as E10, E10_1, 17, E23_1, E23_2, E23_4, E23_7, 43, 45, 46, and 49. In an alignment of all Zygnematophyceae and Spirogyra sequences used, 114 NHS were identified for the genus Spirogyra and/or the individual clades (Additional file 4: Table S4); for the genus Spirogyra, 41 NHS were found (blue filled circles in Figure 3)

Discussion
We compared 130 SSU rDNA sequences of Spirogyra and found a high genetic diversity that was unexpected from the phenotypes. Our phylogenetic analyses revealed that Spirogyra splits into eight independent lineages within Zygnematophyceae (clades A-H; Figures 1, 2 and  3). In contrast to low phenotypic and high genetic variability in Spirogyra, the genus Staurastrumone the most species-rich genera within the Desmidiaceae comprising around 700 speciesshowed great variability of morphological characters such as cell shape, size or cell wall ornamentation. In contrast, it yielded no information on phylogenetic relationships or genetic distance, which was interpreted to reflect recent radiation [32]. Nonetheless, even a small number of species and their relatively uniform appearance may not guarantee generic monophyly (e.g., Spondylosium, Desmidium, Hyalotheca, Netrium, Cylindrocystis, Mesotaenium; [33]). Morphology apparently does not reflect genetic diversity in this group. Morphology is even less trustworthy in Spirogyra; the genus seems to be very uniform, but the species exhibit a wide overlap of character ranges [2,6,34,35]. Furthermore, changes in ploidy level may occur, also affecting morphology [34,36,37]. The overall phylogenetic relationship of the Zygnematophyceae included in our analysis confirms other studies separating Desmidiaceae, Peniaceae and Closteriaceae from Zygnemataceae and Mesotaeniaceae, although the latter two are not resolved from each other [29,30,38]. Gontcharov et al. [20] already stated that Zygnematales appear to be a polyphyletic assemblage of independent clades. The families Mesotaeniaceae and Zygnemataceae are not monophyletic, proving that the cell wall traits (unornamented, unsegmented) are plesiomorphic [15].
When linking the results of our analysis to rbcL data presented by McCourt [15], the Desmidiales phylogeny is largely congruent, while the Mesotaeniaceae and Zygnemataceae (sensu Bold & Wynne [18]) show major differences. In McCourt's analyses, the branch comprising Mougeotia/Mesotaenium and Cylindrocystis/Zygnemopsis is placed as an ancestor to other Zygnematophyceae; the branch Zygnema/Zygogonium and Spirogyra/Sirogonium/ Spirotaenia emerges at a position basal to the Desmidiales. Contrarily, in our phylogenetic tree, the Zygnematales clade is a sister to the Desmidiales clade, although little bootstrap support is given and the genera Spirogyra/Sirogonium and Spirotaenia form individual clades within the Zygnematales. Those clades form distinct branches basal to the remaining Zygnematales branch due to different evolutionary rates of the SSU rDNA.
One reason to choose SSU rDNA over rbcL for phylogenetic analyses is that, as in the rbcL analyses of McCourt [15], phylogenetic relationships among the Desmidiales show rather poor resolution and do not always receive bootstrap support. Furthermore, in the phylogeny in Drummond's analyses [3], members of the Zygnemataceae exhibit very long branches, whereas members of the genera Spirogyra and Sirogonium have very short branches indicating only little difference in sequence. This could lead to a misplacement of the genus Spirogyra, underestimate diversity in this genus and still yield the LBA problem [31].
The order Zygnematales pools taxa with differing evolutionary rates. On one hand, it is important to include these taxa into phylogenetic analyses to obtain an overview of major relationships. On the other hand, taxa with accelerated evolutionary rates often disturb the (See figure on previous page.) Figure 3 Putative secondary structure model of the nuclear encoded SSU rRNA of Spirogyra sp.; Strain WRH5 (Acc.No. JQ 239088) from Clade B was chosen as representative sequence; 100% consensus bases for the genus Spirogyra are given in upper case letters, variable positions in lower case letters; NHS for Spirogyra are marked with blue circles, NHS for one of the clades are marked by green circles, the assigned number indicates the number in the table; see Additional file 4: Table S4 for details.
analyses. This might be the major reason for low bootstrap support in some of the clades we analyzed. In general, the genetic diversity in the Zygnematales at the generic level has been underestimated in favor of morphological traits that proved to be uninformative at the phylogenetic level [22]. Some genera defined solely by morphological characters are probably artificial and polyphyletic [20]. Thus, the species concept in this group needs urgent revision, and the generic concept requires scrutiny. Spirogyra is positioned ancestral to the remaining Zygnematales (except for Spirotaenia) in our phylogenetic tree. By testing the UDT against the best tree derived from ML analysis, we conclude that Spirogyra has to be placed outside of the Zygnematales clade; this position is definitely not caused by LBA, because no other position in the phylogenetic tree was accepted by UDT testing. Such a position is also supported by analyses of combined Zygnemataceae and Spirogyra 1506 group I intron alignments [39].Similar to the rbcL results of McCourt [15], our data do not support the hypothesis of monophyly for groups exhibiting a similar cell shape. In accordance to McCourt's findings, but in contrast to previous SSU rDNA analyses [29], chloroplast shape seems to be a diagnostic trait: stellate and laminate chloroplast containing taxa form two sister clades [15,40], yet without clear indication of the ancestral chloroplast type [41].
McCourt [15] stated that derived cell and chloroplast forms of placoderm desmids are better photosynthesizers and have achieved greater evolutionary success. This is difficult to reconcile with the ecological success of some of the so-called "primitive forms" such as Spirogyra, which is among the most widespread and speciesrich conjugating green algae [1]. High evolutionary rates seem to be more common in "primitive forms" such as Spirogyra, Zygnema and Mougeotia, whose evolutionary rates reportedly differ from other Zygnematales [20]. Combined, we conclude that the lack of so-called derived cell and chloroplast forms is compensated for by a high evolutionary rate. This yields a large variety of genotypes and helps cover ecological niches more quickly.
C-U ratios are generally elevated compared to the remaining substitution rates ( Table 2). This is because mutations from C to U or U to C in non-coding RNA are not detrimental, as the change in base pairs from G-U or G-C does not affect secondary structure. The biggest difference within the GTR + I + G model occurs between Spirogyra and the remaining Zygnematophyceae without Spirotaenia: Spirogyra shows a 1.9 units lower rate, which partly explains the different mutation rate compared to other algae. Both Spirogyra and Spirotaenia show a slightly elevated A-G substitution rate (approximately 50% and 16% higher than in the Zygnematophyceae data set, respectively). Spirogyra displays not only a higher evolutionary rate, but also a different pattern of base substitution rates compared to the remaining Zygnematales. Evolutionary rates within Spirogyra are more diverse than previously expected; significant differences among clades and sequences outside the same clade prevail. Finally, clades B to D, comprising sequences with the 1506 group I intron, form a group with fewer differences compared to clades E to H, comprising taxa without the intron.
The genus Spirogyra is clearly monophyletic. No sequences were found that had slower evolutionary rates or that could resolve the long branch reported in previous studies [20]. The individual Spirogyra clades found in both of our data sets are essentially the same and well supported by bootstrap and PP values. Moreover, the phylogenetic relationship among those taxa exhibiting the secondary loss of the 1506 group I intron is identical. The Spirogyra clade branches are longer than the branches of most genera of Desmidiales. Two long branches within the genus Spirogyra are present in both trees; one separates clade A from the rest, and the other separates taxa with the secondary loss of the 1506 group I intron. Spirogyra taxa characterized by the absence of the 1506 group I zygnematalean intron form a distinct clade with no exception in both trees. This clearly indicates a single loss event. This explanation is supported by the accelerated evolutionary rate of the exon region of the SSU rDNA [20]. A large number of differentiating NHS exists both within the genus Spirogyra (see results) and for each clade. This emphasizes the distinctness and genetic variety within the genus. In contrast, Gontcharov and Melkonian [42] found only very few NHS to circumscribe the different clades in Cosmarium.
Earlier hypotheses suggested the unicellular Spirotaenia as the ancestor of the filamentous Spirogyra [28,40]. Although Spirotaenia shares many attributes of the genus Spirogyra such as chloroplast shape, absence of the 1506 group I intron [43] and an elevated evolutionary rate, molecular analyses have not proven or clarified its phylogenetic position. Our analyses revealed two well-supported branches at the base of the Zygnematalean clade, but the ancestral form remains unclear. Spirotaenia's unusual position among the Zygnematales was already investigated by Gontcharov and Melkonian [43]. Their results indicate no affiliation of the two genera [43]. They therefore suggested excluding Spirotaenia from the Zygnematophyceae sensu stricto. Perhaps the same applies to the genus Spirogyra, but this remains to be proven by further genetic analyses. Sequence differences of the unambiguously aligned positions within Spirogyra reached 274 nt difference among the strains (a member of clade A and one of clade H), whereas the biggest difference found within the remaining taxa was 247 nt (Closterium ehrenbergii and Spirotaenia obscura). Also, the average, median and mode of the difference between sequences are higher in Spirogyra than in the remaining taxa considered. Within the clade, the range of pair-wise differences within the examined Desmidiales (Desmidiaceae, Peniaceae and Closteriaceae) ranges from 10 to 166 nt: this group includes 3 families and 7 genera. The relative distances within this group do not exceed 10.18. Within the genus Spirogyra, 14 individual distances between clades exhibit bigger values ( Table 1). The same trend was observed in the remaining Zygnematales (Zygnemataceae and Mesotaeniaceae)the range of pairwise differences resulted in 27 to 173 nt, the respective distance values ranged from 1.65 to 10.63. The genetic differences within the genus Spirogyra, i.e. among the individual clades, exceed the differences of genera in either of the other groups. This calls for discussion and further investigation on whether Spirogyra still should be considered a single genus or rather be subdivided based on the clades we differentiated. This once again underlines the different evolutionary rates and reflects the great genetic variability of the genus Spirogyra.
Drummond et al. [3] found Spirogyra to be monophyletic, but still treated Sirogonium as a separate genus based on rbcL data. They were unable to discover morphological characters useful for a generic distinction, simply because the taxa are largely congruent (e.g., number of more or less loosely coiled chloroplasts, reproduction by conjugation and anisogamy of gametangial cells). The diagnostic characters are mainly based on the mode of conjugation: while Spirogyra develops conjugation tubes, Sirogonium filaments are bent towards each other and form kneeshaped bends, so-called geniculations. Drummond et al. [3] also considered the shape and ornamentation of the chloroplast margin as a diagnostic feature, but our observations showed this character to be variable and highly dependent on filament vitality. Other morphological characters such as chloroplast number or cell width are also known to be highly variable and could be explained by polyploidy [1,[34][35][36][37]. Other authors also found Spirogyra and Sirogonium forming a single lineage based on single gene analyses (rbcL, SSU rDNA) and combined data sets [15,20,22,43]. Gontcharov stated in 2002 that the genus Sirogonium has to be rejected and S. sticticum (S. sticticum is the type species of the genus) has to be considered as a species within Spirogyra [20]. Interestingly, Czurda [10] already suggested including Sirogonium into Spirogyra as one of four subgenera. We also found Spirogyra to be monophyletic and inseparable from Sirogonium. Monophyly of the Spirogyra/Sirogonium clade was determined in all our phylogenetic analyses, placing Sirogonium firmly in clade C. All alternative tree topologies relocating Sirogonium outside Spirogyra were significantly worse than the best tree uniting Spirogyra and Sirogonium. Sirogonium shares NHS signatures with the genus Spirogyra and the clade in which it is located.

Conclusion
Spirogyra is monophyletic, incorporating the former genus Sirogonium. Genetic diversity and genetic distances within Spirogyra exceed the diversity and distances found in other Zygnematophyceaen genera. Our results suggest pursuing the question whether Spirogyra should be split into several genera, one of them incorporating Sirogonium.
In the surveyed 130 sequences, 53 individually different clones were identifiedmore than was expected from the simple vegetative morphology. The genus forms eight well-supported clades that differ considerably in NHS patternranging from 3 to 29 NHS for a clade. The genus Spirogyra itself exhibits 41 NHS (4 CBCs). Characterizing those clades will require additional studies considering phylogenetic studies on ITS2 secondary structure, hypnozygote morphology, vegetative characteristics and ecology.

Origin of organisms
Spirogyra clones used in this study originated from a field survey conducted in 2006 and 2007 [2]. Single filaments were isolated by the author (CC) and incorporated into the Algenkultursammlung Wien (ASW). The non-axenic clones were maintained in 100 ml Erlenmeyer flasks with Desmids medium [44]

DNA extraction
Prior to extraction the cultures were transferred into a defined mineral medium (modified Woods Hole medium; [9]). After 4 to 6 weeks, the algae were harvested with a sterile needle and put into a sterile 2 ml microcentrifuge tube. Samples were frozen at −80°C for at least 4 h and then lyophilized for at least 48 h to improve the DNA yield. Afterwards, the samples were placed in 2 ml Eppendorf tubes containing 5 to 7 glass beads (3 mm diam.) and ground with a homogenizing mill. Total DNA was extracted following a modified CTAB protocol ( [45] modified after [46]).

DNA amplification and sequencing
Primers used in this study are given in Table 4 [47,48]. The PCR reaction mixture was prepared according to the manufacturer's recommendation. For each PCR reaction, a 10 μl mixture was prepared containing 9 μl ABGene Reddy Mix PCR Master Mix, 0.2 μl for each primer at 20 pM.μ l -1 , 0.4 μl dimethyl sulfoxide (Sigma) and 0.2 μl DNA template. When the PCR result was unsatisfactory due to low DNA concentration, up to 0.5 μl DNA template was used; when DNA template volume was increased, dimethyl sulfoxide volume was reduced to maintain the total volume of 10 μl, accepting a slight shift in the overall ratio of ingredients. The PCR reaction conditions were an initial hold at 80°C for 5 min followed by 36 cycles starting with a denaturation step at 95°C for 30 s, an annealing step at 55°C for 30 s and an extension step at 72°C for 2 min. A final extension step at 72°C for 8 min and the final hold at 4°C were performed after the 36 cycles were completed. The amplified DNA was cleaned by incubating at 37°C for 45 min, followed by denaturing at 80°C for 15 min together with the enzymes Exonuclease I and Shrimp Alkaline Phosphatase (both from Fermentas) and then subjected to a cycle sequencing reaction. The cycle sequencing reaction conditions were an initial hold at 96°C for 1 min followed by 35 cycles starting with a denaturation step at 96°C for 10 s followed by an annealing step at 50°C for 5 s and an extension step at 60°C for 4 min. The end of the cycles was followed by a final hold at 4°C. Sequencing was performed on a 16-capillary sequencer (Applied Biosystems 3130xl Genetic Analyzer) following the manufacturer's protocols. The SSU rDNA sequences were used in the phylogenetic analyses; their GenBank accession numbers are given in Additional file 5: Table S5.

Sequence alignment and phylogenetic analysis
Sequences were aligned manually taking into account the secondary structure of the SSU rDNA [49]. The alignment was refined by comparison of the secondary structure of the sequences. Secondary structure was determined via the Rensselaer bioinformatics web server using mfold [50]. Only unambiguously aligned regions of the sequences were used for analyses; gap-rich regions were excluded. Two different data sets were analyzed: (1) the SSU alignment of 33 Zygnematophyceaen taxa (including 12 Spirogyra sequences representing the different clades) and (2) the SSU alignment of 55 Spirogyra sequences (clones with identical sequences were represented by only one sequence). The combined Zygnematophyceaen SSU dataset consisted of 1720 unambiguously aligned bases, the Spirogyra dataset of 1645 such bases.
The phylogenetic trees presented were inferred by ML settings using PAUP* 4.0b10 [51], and the best models were chosen according to the Akaike Information Criterion by Modeltest 3.7 [52,53]. To test for the best evolutionary model for the analyses, the loglikelihood values of 56 models using Modeltest 3.7 were compared. No outgroup was applied and unrooted phylogenies were used [22]. This was done to avoid LBA sensu Philippe [31] caused by unsuitable taxa as outgroups. This approach also follows Gontcharov's argument that monophyly of the Zygnematophyceae is undoubted but that its position within the Streptophyta is unclear and therefore no suitable outgroup can be chosen. The combined Zygnematophyceaen SSU dataset (1) was also analyzed together with an outgroup (Klebsormidium flaccidum and Coleochaete scutata) to check for LBA among clades. For the Zygnematophyceaen alignment analyses, the TrN + I + G model was chosen; for the Spirogyra alignment the GTR + I + G was chosen by Modeltest. Individual Spirogyra clades were labeled with letters A to H in the sequence from basal to derived. The Spirogyra alignment was analyzed unrooted to avoid LBA phenomena due to different evolutionary rates [31]. Only individual sequences were used for analyses to reduce computational effort. Bayesian inference (PP) was calculated using MrBayes 3.1.2. [54,55] using 3 million generations, sampling every 100 generations and MCMC chains = 4. All trees below the burnin value of 0.01 were discarded as burnin, the consensus tree was calculated using PAUP*. The robustness of the trees was assessed by bootstrap support values. These were calculated using the corresponding evolutionary model chosen by Modeltest by ML (100 replicates), distance NJ (1000 replicates), and MP (1000 replicates) methods using the accordant settings/evolutionary model for each dataset. Insignificant values were not included in figures (PP < 0.95, ML, NJ, MP < 50%). Details of the corresponding evolutionary models and Bayesian analyses are given in the legend of the accordant figures and in Table 2.
Additionally, a distance matrix was calculated using PAUP* to evaluate the genetic distances among the Spirogyra clades (Table 1). The value was obtained by dividing the number of differing bases by the number of