Algal cells were isolated and cultured from D. luridum var. luridum showing spherical (4–5 µm) to oval (4–8 µm × 3–5 µm) cell shape with the presence of a parietal plate shaped chloroplast (Figure 1A–C). Algal cultures were obtained from four Austrian collected samples and two Canadian samples (Table 1) that were morphologically identified as Diplosphaera chodatii (e.g., Figure 1A–C). An additional algal culture, which was isolated from D. luridum var. luridum thalluscollected in Canada, revealed two different algal cell types (Figure 1D). The first cell type is comprised of spherical to oblate spheroid (7.5–11.5 µm length) shaped cells with the presence of a parietal chloroplast and many globular bodies throughout the cell. The second type is oval to oblong (6–8 µm length) with a width equal to or slightly greater than the width observed in cells of the D. chodatii cultures. The second cell type presented characteristics similar to the characteristics described for the first cell type.

One of 32 most parsimonious trees (Figure 2) summarizes the phylogenetic placement of 22 Internal Transcribed Spacer (ITS) nuclear ribosomal DNA (rDNA) sequences obtained from this study and 36 NCBI GenBank sequences; two of which were used to root the tree. From a dataset of 451 total characters, 140 were parsimony-informative and the final trees shared a tree length of 376 changes with CI and RI scores of 0.6968 and 0.9132, respectively.

The mycobiont tree may be interpreted by identifying four primary clades (Clades I, II, III, and IV), all of which are well supported by bootstrap and posterior probability support. Clade I is supported by 99% bootstrap and 100% posterior probability, and is comprised of all D. luridum samples including two varieties. D. luridum var. decipiens samples are outside this clade and fall into Clade II, which also includes: Dermatocarpon miniatum var. miniatum, D. miniatum var. complicatum, D. arnoldianum, and D. leptophyllum. With the exception of this intraspecific taxon, D. luridum might be considered a monophyletic group.

Clade I may be divided into four sub-clades: a, b, c and d. There are 12 samples that constitute sub-clade Ia with 85% bootstrap and 100% posterior probability support. Four samples constitute sub-clade Ib with 85% bootstrap and 99% posterior probability support. Both sub-clades Ia and Ib represent the species D. luridum var. luridum and are supported with 99% bootstrap and 85% posterior probability support. One sample in sub-clade Ia (M18-1) is from Austria and all others from North America (Manitoba, Minnesota and North Carolina). Sub-clade Ib represents additional Dermatocapon luridum var. luridum sequences from Europe (Austria and Sweden) and is well supported with 85% bootstrap and 99% posterior probability supports. NCBI GenBank fungal ITS sequences of Dermatocarpon luridum var. luridum representing individuals of the Southern U.S. comprise I I sub-clade Id. These samples are segregated from the D. luridum var. luridum collected from Northern Canada and U.S. (sub-clade Ia) and Austria and Sweden (sub-clade Ib). Strong bootstrap and posterior probability, both at 100%, support the placement and segregation of the Southern U.S. from the rest of the D. luridum var. luridum samples. There are six taxa from NCBI GenBank representing D. luridum var. xerophilum that comprise sub-clade Ic which is supported by 100% bootstrap and posterior probability support.

Clade II is supported by 95% posterior probability but the taxa: Dermatocarpon miniatum var. miniatum, D. miniatum var. complicatum, D. luridum var. decipiens and D. leptophyllum comprising Clade II, with the exception of D. arnoldianum,are not monophyletic. Clade III is represented by two monophyletic species D. multifolium and D. arenosaxi each with 100% support. Clade IV is also monophyletic being comprised of D. rivulorum with 100% support.

The ITS rDNA phylogeny represents sequences from 22 algal samples (Figure 3), 16 sequences from this study, and six NCBI GenBank accessions. One of nine most parsimonious trees (Figure 3) is presented from a data set of 699 characters, of which 178 were parsimony-informative characters. The tree length is 707 changes and CI and RI scores were 0.8925 and 0.8281, respectively. Since the Neighbor Joining (NJ) and Maximum Likelihood (ML) trees were consistent with those of the Maximum Parsimony (MP) and Bayesian analyses, the NJ and ML trees are not presented.

Four highly supported clades are presented in the algal phylogenetic tree (Figure 3). Clade I is supported by 95% bootstrap support and it contains algal sequences representing both Austrian and Canadian (Manitoban) photobionts. This clade is composed of the photobionts associated with D. luridum var. luridum and one NCBI GenBank reference. Clade II, supported by 96% bootstrap support and 95% posterior probability support, contains the algae of three different taxa, D. miniatum, D. luridum var. luridum and D. arnoldianum. Algae from three species of Dermatocarpon, in Clades I and II form a highly supported group with 94% bootstrap support and cluster with the UTEX sample of Diplosphaera chodatii. Clade III contains two NCBI GenBank accessions of Stichococcus with 100% of both bootstrap and posterior probability. Clade IV contains a Stichococcus species and the alga from D. rivulorum, and is supported by 95% bootstrap support and 100% posterior probability support.

Algal clade I are all photobionts of D. luridum var. luridum from Fungal Clade Ia (Figure 3; F_Ia), with the exception of one DV4-1 of Fungal Clade Ib (F_Ib). Algal Clade II contains algae that associate with fungal species that are in three different fungal clades (Figure 3; F_Ia, F_Ib, and F_II).

The insert photo (Figure 3) reveals how the addition of two NCBI GenBank sequences of Stichococcus bacillaris disrupts the clade consisting of Stichococcus mirabilis and the photobiont sequence from D. rivulorum. The photobiont from D. rivulorum is grouped in a Clade with the two S. bacillaris sequences. There is 100% bootstrap support for the pairing of the two S. bacillaris sequences. The generation of Clades I, II, and III was identical with 100% bootstrap support. Clade I and II are supported by 84% and 79%, respectively, which are lower support values than the primary tree. Clade III is supported once again by 100% bootstrap support.

The finding that fungal sequences from Austria clustered with sequences from North American specimens (Figure 2) suggested that gene flow of D. luridum var. luridum occurred between continents and is likely ongoing (Figure 2 and Figure 3). Other studies of lichen phylogenetic differentiation at the global scale have provided mixed results. The crustose lichen Porpidia flavicunda was shown to share identical fungal haplotypes between Northern North America, Greenland and Northern Europe [27] and was hypothesized to have undergone long distance dispersal. Geographic segregation between continents was shown to occur for Lobaria pulmonaria [28] and for Letharia vulpina [29]. Both local variation within populations and variation between populations especially as distance increased suggested isolation by distancefor Lobaria pulmonaria. Our study suggests low levels of variation in global and local populations of Dermatocarpon luridum var. luridum with some sharing between continents, but the small sample size does not support conclusive comments to be made on the extent of gene flow. Since long distance dispersal of lichen diaspores was hypothesized for Cetraria aculeata and Cavernularia hultenii [27,30,31], this may also be true for the dispersal of the mycobiont, Dermatocarpon luridum var. luridum. The sharing of genetic information between photobiont populations associated with D. luridum var. luridum from both continents is implied, but further study is underway to determine the extent of photobiont exchange/movement between continents.

The fungal phylogeny obtained from the ITS sequence data reveals the paraphyletic nature of both Dermatocarpon luridum and Dermatocarpon miniatum (Figure 2). The placement of D. luridum var. decipiens into Clade II makes the species paraphyletic. D. luridum var. decipiens may belong as varietal status within D. miniatum. The taxonomic status of D. arnoldianum was questioned by Heiđmarsson [11], suggesting it be incorporated into the species as a variety of Dermarocarpon miniatum [11]. In his review, Heiđmarsson [11] mentions the prior troubles of taxonomically classifying D. arnoldianum by Degelius [32] based on morphological character states and interpretations of iodine reactions relative to D. linkolae. The placement of D. arnoldianum and D. luridum var. decipiens, nested within Clade II suggests that the species maybe more appropriately considered subspecies of D. miniatum rather than independent species. The use of the ITS rDNA gene to address species delimitation is useful because it contains regions that are faster evolving than the surrounding subunit sequences. The ITS gene is routinely used as a phylogenetic marker for fungal species and genera [33] and it has been proposed for use as a bar coding gene [34]. Limitations of the locus include the use of the ITS rDNA as a single marker where additional markers will provide more data for phylogenetic reconstruction. Also, the potential failure of concerted evolution in the multiple tandem repeating units and paralogous ITS regions may obscure phylogenetic relationships [35]. Further study is also needed to resolve D. miniatum var. complicatum and var. miniatum since all samples in this tree were obtained from NCBI GenBank database.

Although D. rivulorum and D. luridum are similar in morphology [36], D. rivulorum is the most basal of the seven species in the tree (Figure 2). Dermatocarpon rivulorum and D. luridum/D. miniatum may have evolved from an ancestor with similar sub-aquatic habitat requirements and adapted to a terrestrial environment while maintaining the same photobiont. The aquatic nature of the lichen may have been lost multiple times through evolution of these morphologically similar species. However, the limited sample size in this study prevents an assessment of the ancestral state reconstruction of aquatic and terrestrial species.

The microscopic analysis of the six cultured photobiont isolates from Dermatocarpon luridum var. luridum and nucleotide sequencing analyses suggest these algae are Diplosphaera chodatii Bialosuknia. The morphology of the cells cultured in this experiment (Figure 1) are consistent with those shown by Zhang and Wei [37], photobiont culture isolates from Endocarpon pusillum and the UTEX algal collection 1177 of D. chodatii. Algal identity is further supported by the cell morphology and dimension information reported for Diplosphaera chodatii (syn. Stichococcus diplosphaera) by Ahmadjian and Heikkila [23] in the hymenium of Endocarpon pusillum.

Blast search in NCBI GenBank of ITS gene sequences of the photobiont of D. luridum var. luridum obtained from this study provided six significantly similar matches with query coverage greater than 80%, which have been incorporated into the algal phylogeny (Figure 3). The sequences of Diplosphaera chodatii UTEX 1177 and those generated by Zhang and Wei [37] were most similar to our sequences. Three Stichococcus sequences with lower similarity were also identified through the BLAST search. The phylogenetic analysis included our algal ITS nrDNA sequences combined with the six NCBI GenBank accessions that supported the morphological identification. The strongly supported clades support that D. chodatii is the preferred photobiont of Dermatocarpon luridum var. luridum, Dermatocarpon arnoldianum and Dermatocarpon miniatum [18,38].

One of the cultures obtained from D. luridum var. luridum (Ne18-1; Figure 1) contained two types of algal cells (autospores and mature vegetative cells). The morphology does not disagree with that of the green alga Elliptochloris bilobata Tschermak-Woess (Figure 1) even though the ITS sequence was amplified from D. chodatii. Elliptochloris bilobata was earlier shown to be a photobiont of Verrucaria sublobulata [18] and is comprised of the morphological features shown in Figure 1. However, the algal cell types observed in the culture (Figure 1) were not observed within the lichen thallus. The finding of an additional algal species using culture methods may be explained if D. luridum var. luridum can associate with more than one photobiont simultaneously; if Elliptochloris bilobata, was present on the thallus as a free-living alga. Even though the thallus was examined and washed before culturing, small amounts of an additional species could be selected under certain culture conditions. Other studies have shown multiple photobionts in one culture [39] and multiple photobionts or photobiont genotypes present in the same thallus [16,17,40,41,42]. The phylogeny in Thüs et al. [18] shows that Diplosphaera sp. is shared with four lichen forming fungi including D. miniatum, supporting the findings of this study. The re-synthesis experiments of Stocker-Wörgötter and Türk [24] identified the photobiont associated with Dermatocarpon miniatum to be Hyalococcus dermatocarponis H. Warén, which does not disagree with the findings in this study. The low selectivity of D. luridum and D. miniatum provides an explanation for the fungal partner associating with different algal partners in different geographic locations. Hyalococcus dermatocarponis was also isolated from Dermatocarpon luridum (syn. Dermatocarpon fluviatilis) and deposited in the UTEX algal collection (UTEX number 908) by V. Ahmadjian. The finding that six of the seven cultures were morphologically identified as D. chodatii provides support for a one-fungus and one-photobiont hypothesis, but it does not eliminate the possibility that more than one photobiont may associate with D. luridum.

The broad geographic origin of the algal sequences from Dermatocarpon species that fall within Clades I and II (Figure 3) suggests that there may be multiple strains of the alga distributed throughout North America and Europe. This broad distribution may be explained by the low selectivity of D. chodatii by three mycobiont species, and that in sexual reproduction the fungal ascospores must come into contact with a different compatible algal strain at each dispersal event. Lichen asexual propagules such as isidia and soredia have been hypothesized to serve as vectors for photobiont dispersal [16,40,43] to promote algal dispersal and algal switching. Members of Verrucaria, Staurothele and Endocarpon illustrate another means of algal dispersal by releasing the photobiont into the environment together with ascospores [18]. As with all sexually reproducing lichens, ascospores must come in contact with a suitable alga in order to establish a symbiosis. The ability to tolerate lengths of desiccation stress up to and exceeding two months [37] and their free-living presence in soils, on bark, on rock and as epiphytes on lichens and other living and dead plants [44,45,46,47,48,49] increases the likelihood that spores will come in contact with D. chodatii for a symbiosis to establish.

Fungal and algal taxa from both Canada and Austria are placed throughout the trees, suggesting that fungi from both continents share some of the same photobiont species. Additionally, the clustering of algal sequences obtained from different Dermatocarpon species suggests sharing of algal genotypes among Dermatocarpon taxa and low specificity. Dermatocarpon miniatum, a terrestrial species, shares the same photobiont, Diplosphaera chodatii, as Dermatocarpon luridum var. luridum and Dermatocarpon arnoldianum, two sub-aquatic species [18,38]. The sub-aquatic Dermatocarpon luridum var. luridum and terrestrial Dermatocarpon miniatum are typically separated by physical distance from one another, where D. luridum is located in the submerged zone and D. miniatum is in the upper splash zone [50] addressing our original question as to whether they share the same alga in these different habitats.

Both algal and fungal phylogenies place Dermatocarpon rivulorum and its algal partner in a basal position to other taxa (Figure 2). Since the DNA sequences of the algal and fungal partners were obtained from the same lichen thallus extract, it is possible that the sequence in the algal tree may represent an epiphyte or contaminant by a Stichococcoid-like alga. Assuming that the photobiont sequence is from D. rivulorum, the placement of the photobiont ITS sequence from Dermatocarpon rivulorum in the basal position in the tree with Stichococcus spp. suggests that the alga that associates with D. rivulorum may be a Stichococcoid-like alga. The degree of morphological similarity between Stichococcus species and Diploshpaera species greatly exceeds the dissimilarity. One difference is the formation of two-celled or short chain clusters [51]. However, the absence of morphological evidence and the long-branch connecting Stichococcus (Figure 3)might suggest that the photobiont associated with D. rivulorum could be a Diplosphaera-like alga. Long-branch attraction [52] might also account for the placement of Stichococcus and photobiont of D. rivulorum together in the tree. Culture evidence and additional molecular analysis of cultured photobionts and thallus extracted DNA may help to resolve this issue. The photobiont genus associated with Dermatocarpon rivulorum was not reported by Thüs et al. [18]; however, Nascimbene et al. [3] stated the photobiont is Diplosphaera chodatii [38]. Thüs and Schultz [22] also indicated that the photobiont of D. rivulorum is D. chodatii. The divergence of the photobiont associated with D. rivulorum in the algal tree suggests that the species of photobiont is more specific to D. rivulorum thanto D. luridum or D. miniatum and may reflect the limit of algal sharing between fungal taxa. Since both D. rivulorum and D. luridum var. luridum grow in similar habitats and in close proximity to one another, it is reasonable to assume that the two fungi associate with the same photobiont species. Even though the single sequence in this study suggests a distant relationship between photobionts of D. luridum and D. rivulorum, the identity of the D. rivulorum photobiont(s) needs more investigation.

Four lakes in west central Manitoba, Canada, and five locations in Austria were sampled for D. luridum var . luridum and other Dermatocarpon lichens. In Canada, ten specimens were collected (Table 1). Thirteen samples were collected from five locations in Austria (Table 1). All lichen thalli were moistened to aid the scraping of the thalli off rock surfaces and reduce damage to the thalli before they were air dried and processed. All vouchers (Table 1) are maintained in the cryptogamic division of the University of Manitoba herbarium (WIN), Winnipeg, Manitoba, Canada.

Photobiont isolation from Dermatocarpon luridum var. luridum was performed using seven thalli (Table 1) according to the Yamamoto method [55], and isolations were performed using the procedure as described in Stocker-Wörgötter [56] and Stocker-Wörgötter and Hager [57]. A lobe of lichen thallus, approximately 20 mm² was cut into smaller fragments and washed in a 50 mL Erlenmeyer flask containing 25 mL of sterile double distilled water for 15 min on a stir table. One drop of Tween 80 was added to the water and the fragments were washed for an additional 15 minutes on a magnetic stirrer. The lichen thallus was transferred using sterile forceps to a clean 50 mL Erlenmeyer flask and washed in 25 mL of fresh sterile double distilled water for 20 min to remove the Tween 80.

Under sterile conditions, the washed lichen fragments were then ground in 3–4 mL of sterile water using mortar and pestle which was filtered through a sieve with mesh pore size of 500 µm and the fragments that passed through were then filtered for the second time through a mesh with pore size 150 µm. The filtrate consisting of fragments of lichen “tissue” (about 150 µm) were selected under a stereomicroscope using sterile bamboo sticks and transferred to slanted agar media.

Woods Hole MBL [58] was used to culture and isolate axenic colonies of the photobiont. Cultures were grown for 2–3 months at 22 °C under a changing light:dark regime of 14:10 h with a light intensity of 100 µE m⁻²s⁻¹. Sub-culturing of the photobiont to liquid medium was performed, once the algal colony size reached 2–3 mm diameter. An inoculation needle was flamed and the colony was fractioned into tiny drops. A small number of algae were transferred to a sterile 50 mL Erlenmeyer flask containing 25–30 mL of liquid medium. Single alga isolates were not feasible, because of the small size of the Diplopshaera cells.

To investigate the morphological features of the algal cells in culture, cells were streaked into water on a microscope slide and observed using a Univar Light Microscope (Reichert INC., Vienna, Austria) and photographs were taken with a Canon PowerShot G5 camera (Canon INC., Tokyo, Japan). Algal species cultures were sent to A. Beck and temporarily maintained at the Culture Collection of the Botanische Staatssammlung Munich, Germany.

Extraction of DNA was accomplished using a protocol modified from Grube et al. [59]. The ITS rDNA of both the mycobiont and photobiont of 22 Dermatocarpon thalli was amplified by Polymerase Chain Reaction (PCR). The DNA was initially amplified using 20 μL reactions to determine the optimal DNA concentration for PCR and then amplified in larger quantities (four to eight 20 μL PCR reactions) for sequencing. One 20 μL reaction consisted of 5–25 ng of genomic DNA, 1× PCR buffer (50 mM KCl; 100 mM Tris-HCl [pH 8.3]), 2.0 mM MgCl₂, 200 μM of each dNTP, 0.5 μM of each primer, and 2 units of Taq DNA Polymerase (Invitrogen, Burlington, ON, Canada).

The fungal and algal ITS genes were amplified using the fungal specific forward primer ITS1F-5' (5'-CTTGGTCATTTAGAGGAAGTAA-3') [60] and the newly designed algal specific primer STICHO-ITS-F-5' (5'-GGATCATTGAATCTATCAACAAC-3'), respectively, both together with the universal reverse primer ITS4-3' (5'-TCCTCCGCTTATTGATATGC-3') [61]. All amplifications of the ITS genes were performed using a Biometra T-Gradient thermal cycler (Fisher Sci, Nepean, ON, Canada). The PCR cycle used to amplify the fungal ITS consisted of an initial denaturation of the DNA at 94 °C for 3 min; then 30 cycles of denaturation at 94 °C for 1 min, annealing at 61.5 °C for 30 s and extension at 72 °C for 45 s; followed by 6 °C soak. Amplification of the algal ITS was accomplished with an initial denaturation of the DNA at 94 °C for 3 min; then 30 cycles of denaturation at 94 °C for 1 min, annealing at 58.5 °C for 35 s, and extension at 72 °C for 45 s; followed by 6 °C soak.

Precipitation of the combined PCR products was carried out by adding 0.2 volumes of 5 M NaCl and 2.5 volumes of 100% ethanol. The tubes were gently inverted several times, placed in the fridge (4 °C) for 30 min, and centrifuged for 10 min at 13,000 rpm. The supernatant was poured off and the pellet was washed with 300 μL of cold 80% ethanol and the pellets were left to air dry for 30 min. This pellet was dissolved in 20 μL of sterile distilled water and the entire 20 μL PCR volume was electrophoresed on a 1% agarose gel, the bands were excised from the gel, and purified using Promega’s Wizard^® SV Gel and PCR Clean-Up System (Promega Corporation, Madison, WI, USA). The purified DNA was resuspended in 35 μL of sterile distilled water, and quantified by gel electrophoresis. Band intensities were compared with the 1650 base pair band (80 ng/μL) of a 1 kb plus DNA ladder (Invitrogen, Burlington, ON, Canada) and visualized using an AlphaImager 2200 transilluminator (Alpha Innotech, Fisher Scientific, Nepean, ON, Canada).

PCR products were cycle sequenced with BigDye v.3.1 (Applied Biosystems, Foster City, CA, USA) following the manufacturer’s directions. Precipitation and cleanup of the cycle sequenced product to remove excess fluorescent dyes was carried out by the EDTA method following manufacturer’s directions, resuspended in Hi-Di formamide (Applied Biosystems, Foster City, CA, USA) and loaded to a 96-well plate for sequencing with an Applied Biosystems 3130 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA).

Nucleotide sequences obtained from samples collected for this study were edited using Sequencher v. 4.8 (Gene Codes Corporation, Ann Arbor, MI, USA). Sequences retrieved from NCBI GenBank were selected by first performing a nucleotide BLAST [62] search on the entire nucleotide collection using each algal sequence of the photobionts from D. luridum var. luridum as the query sequence. Only those with significant e-values of 10e⁻⁵⁰ [63] and high query coverage (>75%) were selected for phylogenetic analysis. Sequences were aligned manually using Se-Al v. 1.0 [64] and two ambiguous regions were delimited based on visual inspection where substitutions were too numerous for a homologous alignment. The alignment was subjected to parsimony analysis in PAUP* 4.0b10 [65] and Bayesian analysis in MrBayes v3.1.2 [66,67]. Four MP analyses were performed for the algal alignment to assess the effect of the ambiguous regions on the phylogeny: (1) one ambiguous region was removed from the analysis; (2) the other ambiguous region was removed from the analysis; (3) both ambiguous regions were removed from the analysis; and (4) both ambiguous regions were included in the analysis. Since the topology was the same and the bootstrap values fluctuated between 3 and 5%, the final analysis presented in the paper includes both ambiguous regions. The fungal data set that was analyzed consisted of 58 fungal ITS sequences, 22 derived from the current study and 36 from NCBI GenBank (Table 1). Likewise, the algal data set that was analyzed consisted of 22 algal ITS sequences, 16 derived from the current study and 6 from NCBI GenBank (Table 1). The fungal outgroup species were Endocarpon pallidulum and Verrucaria viridula, while the algal outgroup was designated as uncultured Eukaryotic algae. The sequences obtained from this study have been deposited in NCBI GenBank and accession numbers are indicated in Table 1.

Maximum Parsimony settings were set to run a full heuristic search with a tree bisection and reconnection (TBR) branch swapping algorithm, with a stepwise addition and 5,000 replicates (Fungal data set) or 10,000 replicates (Algal data set) with random addition of taxa. Swapping was performed on only the best trees, and only the best trees were kept with the saving of Multrees option off. The most parsimonious trees were analyzed for bootstrap support within PAUP with the following MP settings: 1,000 bootstrap replicates [68] were performed and only scores greater than 70% were presented in the phylogenetic trees. All other settings were the same as the initial heuristic search; all but the random addition of taxa setting, which was set to 10 for the fungal analysis and 100 for the algal analysis. Sequence alignment contained ambiguous regions in the algal data set from base pair positions 76–107 and 415–465 of the unaligned UTEX 1177 (HQ129931) sequence. Each of theses regions were individually excluded from subsequent analyses and then all together excluded to see if tree topology was altered significantly. All ambiguous regions had minimal effect on the topology and clade scores, therefore all character data was used for the phylogenetic analysis presented.

Bayesian analysis was performed following the best evolutionary model estimated with JModeltest 0.1.1 [69,70,71]. Both ITS data sets were analyzed according to a TrN+G (Tamura-Nei plus Gamma model), which was presented as the best model according to hLRTs (Hierarchial Likelihood Ratio Tests) output within JModeltest. This model follows a six parameter with a gamma shaped distribution. The TrN+G analysis parameters were set as follows for the fungal dataset: Lset Base = (0.2398, 0.2504, 0.3044); Nst = 6; Rmat = (1.0000, 4.7694, 1.0000, 1.0000, 1.7829); Rates = Gamma; Shape = 0.4751; and Pinvar = 0. Likewise, the TrN+G analysis parameters were set as follows for the algal dataset: Lset Base = (0.2323, 0.3015, 0.2792); Nst = 6; Rmat = (1.0000, 1.8102, 1.0000, 1.0000, 3.8085); Rates = Gamma; Shape = 0.3942; and Pinvar = 0. The model information for both analyses was incorporated into PAUP for ML Analysis (outcomes not presented). The settings of this model correspond to the GTR+G evolutionary model. Bayesian analyses were performed using 2,000,000 generations and were terminated well after the standard deviation of split frequencies fell below 0.01. The number of trees discarded following the analysis included 500 burn-in trees. Posterior probability scores greater than 90 are reported on the phylogenies.

A phenetic analysis of both data sets using NJ was also performed in PAUP using the default settings and a Kimura 2-parameter model (outcomes not shown). Outcomes of the ML and Neighbour Joining analyses are not presented because the results were largely congruent with the MP and Bayesian analyses.