DNA barcoding is a promising tool to facilitate a rapid and unambiguous identification of sponge species. Demosponges of the order Dictyoceratida are particularly challenging to identify, but are of ecological as well as biochemical importance.
Here we apply DNA barcoding with the standard CO1-barcoding marker on selected Indo-Pacific specimens of two genera, Ircinia and Psammocinia of the family Irciniidae. We show that the CO1 marker identifies several species new to science, reveals separate radiation patterns of deep-sea Ircinia sponges and indicates dispersal patterns of Psammocinia species. However, some species cannot be unambiguously barcoded by solely this marker due to low evolutionary rates.
Citation: Pöppe J, Sutcliffe P, Hooper JNA, Wörheide G, Erpenbeck D (2010) CO I Barcoding Reveals New Clades and Radiation Patterns of Indo-Pacific Sponges of the Family Irciniidae (Demospongiae: Dictyoceratida). PLoS ONE 5(4): e9950. https://doi.org/10.1371/journal.pone.0009950
Editor: John J. Welch, University of Edinburgh, United Kingdom
Received: July 8, 2009; Accepted: September 28, 2009; Published: April 1, 2010
Copyright: © 2010 Pöppe et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This study benefited from funding by the German Research Foundation (DFG, www.dfg.de) to GW. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Sponges (Porifera Grant, 1836) are an important group of Metazoa in which species identification based on morphological characters is particularly difficult. However, as a group they are highly diverse, ecologically important as filter feeders and of commercial importance to the pharmaceutical and biomaterials industry as producers of highly potent secondary metabolites e.g. . Most sponge taxa possess only a depauperate suite of complex characters. The basis of poriferan morphological systematics and species identification is based on the skeletal elements, their size, shape, arrangement and combination. Unfortunately, the evolution of skeletal traits is not fully understood. Furthermore, the diversity of these skeletal elements is frequently small, patterns in arrangement are hardly detectable, and environment-induced morphological variability makes their unambiguous interpretation difficult (see e.g. , ). This often results in homoplasies and erroneous classification . Collectively, these factors make Porifera highly susceptible to cryptic speciation , and the actual species diversity and radiation may be under-estimated , , .
Such problematic species identification applies particularly to “keratose” sponges, which comprise taxa of the orders Dendroceratida and Dictyoceratida (Minchin, 1900). Keratose sponges lack a mineral skeleton, known from most other demosponge groups, but possess an organic skeleton made of spongin fibers instead (see ,  for more details). Such an organic skeleton provides less morphological complexity than its mineral counterparts and makes this group a special challenge even for experienced taxonomists. Additionally, the long-term storage of (type) specimens in desiccating preservatives such as ethanol results in changes to specimen colour and tissue shrinking, which makes morphological comparison difficult. As keratose sponges also produce a particularly wide range of bioactive compounds of particular interest for the pharmaceutical industries , means of unambiguous (i.e. non-morphological) species identification have to be employed.
Among the dictyoceratid sponges, the family Irciniidae Gray, 1867 has autapomorphic features distinguishing this taxon from other demosponge families: its taxa possess fine collagen filaments in the mesohyl, which gives the sponges a rubber-like texture. Family Irciniidae currently consists of three genera with 111 described species  with an assumed worldwide distribution. The genus Ircinia currently comprises over 77 described species  and differs from Sarcotragus Schmidt, 1862 (11 known species) by the nature of the primary fibers. Psammocinia Lendenfeld, 1889 for which 23 species are currently described, is distinguished by a dermis armoured with a thick crustose layer of foreign debris. However, the classification of the species is more difficult, increasing the probability of the existence of cryptic species among the known specimens , , and their detection purely by means of morphology appears unlikely.
A potential solution is provided by DNA taxonomic approaches such as DNA barcoding. DNA barcoding was introduced as a method not only for the identification of known species but also for discovery of cryptic speciation by means of diagnostic DNA sequences , . For sponges, concerted barcoding has been set up only recently (www.spongebarcoding.org, , ), and is performed initially under usage of the 5′ region of CO1, the standard barcoding fragment . However, the suitability of this CO1 fragment for sponge species remains to be evaluated given the reduced substitution rates for mitochondria in Porifera and Cnidaria  (see also ), which could diminish the resolution power at species- or genus level.
In this study we follow two goals: First we aim to estimate whether the resolution power of this standard CO1 barcoding fragment may be sufficient for Irciniidae, or whether the reduced mitochondrial substitution rates may prevent any molecular separation below genus level. Second, we aim to assess radiation pattern and evidence of cryptic speciation in Irciniidae. For this purpose we DNA-barcoded an Irciniidae selection from Australia and from other regions of the Indo-Pacific, which is a hotspot for keratose sponge radiation.
The list of specimens, for which we succeeded in receiving amplifyable DNA and unambiguous sequences is given in Table S1. The specimens analyzed in this study originate from the Porifera collection of the Queensland Museum. The taxon set comprised specimens of the genera Ircinia and Psammocinia. Figure 1 provides an overview on their geographical distribution. The genus Sarcotragus was not included in analysis as its status is viewed as uncertain  and awaits revision. Ircinia and Psammocinia of different morphological groups have been selected, which were partially readily determined to species level, partially awaiting their new species description or final determination. In the following we will refer to these as “species” based on their significant morphological distinctness, which is frequently greater than the morphological differences observed in most well-established (i.e. biological) species (e.g. Halichondria panicea and H. bowerbanki ). In other studies these “species” have been referred to as Operational Taxononomic Units, or OTUs. See File S1 for details on the species.
The final dataset for the haplotype analysis consisted of 66 specimens and 519 characters of CO1. We differentiated a total of 6 different haplotypes of Ircinia and 14 haplotypes of Psammocinia. The phylograms in Figure 2 show haplotypes and phylogenetic relationships as reconstructed by parsimony- and bayesian methods. The phylogenetic trees display a clear distinction between the Psammocinia clade and the Ircinia clade. The Ircinia clade consists of six shallow water (<60 m) species with I. spiculosa from Thailand as a sister-group to the remaining Ircinias. The remaining five shallow water species are from Australia and comprise three haplotypes, one of which is shared by I. ramodigitata Burton, 1934 and species 3173 and 2828. There is no intraspecific variation in any of the species, while interspecific variation comprises 0–0.4% (p-distances, see also Table S2).
A. Maximum-parsimony phylogram with the genetic differences of the specimens and the contents of the haplotypes. Haplotypes were named after their species, or in the case of several species per haplotype, with a letter. Species names are followed by their Queensland Museum species number, and the QM-collection number (Gxxxxxx). Numbers above the branches indicate total differences (in substitutions). B. Bayesian inference phylogram on the haplotype relationships. Haplotype names refer to figure A. The numbers above the branches are posterior probabilities. Black bars indicate amino acid changes.
Figure 2 displays a distinctive split between the Australian shallow water sponges and four species collected from the seamounts of the Norfolk Ridge, south of New Caledonia (Figure 1). These deep-sea (180–470 m) specimens possess two distinct haplotypes, which differ by 2 basepairs (0.4%) and represent species likely to be new to science.
In total the genetic variation inside the Ircinia clade comprises not more than 2.7% (p-distance) difference between the species. Three substitutions resulted in non-synonymous codon changes and affected the protein sequence (see Figure 2B).
The 46 sequences of the genus Psammocinia resulted in 14 distinct haplotypes (Figure 2). Ten haplotypes could be assigned to distinct Psammocinia species that were supported by their morphological species groups, whereas four showed distinct morphological variation within their haplotypes. Only in one case it was found that four specimens of a supposedly morphologically distinctive Psammocinia species were split into two haplotypes (species 3983). Several haplotypes comprise geographically distinct locations (e.g. 1909, 2188). In total the genetic variation among Psammocinia species comprises 0–1.8% difference between the species. Only one substitution resulted in changes of the protein sequence within the Psammocinia clade (see Figure 2B).
Our results show that the standard barcoding marker cf.  is an important tool to detect cryptic speciation and to aid taxonomy. On the supra-specific level, the CO1 fragment provides clear separation between the target genera and facilitates the correct identification of genera, which is a problem in several other poriferan taxa such as the Haplosclerida, in which mitochondrial CO1 haplotypes fail to provide clear distinction between (morphologically) accepted genera or even families . Here, Psammocinia and Ircinia, which morphologically differ only by the absence or presence of an armoured ectosome, are neatly divided into two clades. Even though the generic assignments of keratose taxa is problematic for many taxonomists and can only be achieved with histological sections and long experience, we have demonstrated that the standard CO1 barcoding fragment is suitable at this level, although more genera will need to undergo additional testing.
At the species level, the main target of DNA-barcoding projects, the standard CO1 fragment resolved several Ircinia and Psammocinia species from each other. The low genetic differences between these dictyoceratid haplotypes indicate the insufficient resolution power of CO1 for the species analysed in this data set. The interspecific variation of 2.7 and 1.8% for Ircinia and Psammocinia respectively, is clearly below the variation of 10–20% interspecific distances suggested for molecular distinction of species in other non-bilaterian Metazoa , . A suitable barcoding gap , which distinguishes between the species and the next higher taxon, is therefore difficult to find. However, interspecific distances in a sufficient range have been found among other demosponge genera, such as Scopalina (Order Halichondrida), in which up to 22% sequence divergence (uncorrected) among OTUs has been detected  (see also genus Tethya (Order Hadromerida) ).
Our results suggest that the barcoding fragment may be too conserved to provide unambiguous barcodes for every demosponge species, because some haplotypes in our analysis are shared by several otherwise morphologically divergent species. For example, clade “Psammocinia C” contains P. bulbosa and 5 other so far unnamed species, which are clearly morphologically distinct from P. bulbosa, yet share the same CO1 haplotype.
On this basis, the CO1 haplotypes do not result in species-specific barcodes, because several Ircinia and Psammocinia species share the same haplotype. Therefore it is evident that the CO1 standard barcoding marker is not suitable as the only barcoding marker (at least not for the taxa investigated here). Where several species share a haplotype, additional markers should be used in combination with the standard CO1 fragment to provide better species-level resolution within the CO1 haplotype . Suggestions for an alternative marker include ITS  (but see also ) or an additional 3′ region of the CO1 fragment  which has already been successfully tested in Xestospongia (Haplosclerida) .
A deep split between shallow and deep-sea species from the Norfolk Ridge seamounts was demonstrated here. In the literature, seamounts such as the Norfolk Ridge, are regarded as deep sea “islands” with unique biodiversity and restricted species ranges , and new bioactive compounds were detected from Norfolk Ridge Ircinia . Our current data provides evidence for a radiation among Norfolk Ridge species since the four species included here form a monophyletic group with internal genetic differentiation. This pattern raises evidence for a single separation event from other Ircinia, although not much more can be deduced from the presently limited data set.
Furthermore, an extensive radiation of Psammocinia in Australian waters is indicated by our data. Currently there are three described species from Australia (Psammocinia arenosa (Lendenfeld, 1888) and P. vesiculifera (Poléjaeff, 1884) from New South Wales, and P. halmiformis (Lendenfeld, 1888) from Western Australia – none of which occur in tropical waters as do most of the species investigated here– and one from New Caledonia (P. bulbosa Bergquist, 1995). All other known Psammocinia species are from New Zealand, South Korea and Brazil. Therefore, most of the species investigated here undoubtedly represent new taxa. Clearly, the number of described Psammocinia species of Australian (and other) waters does not reflect the total biodiversity of this genus. Psammocinia sp. 3983 is the only species in our data set with two haplotypes and may be another example of cryptic speciation in sponges , ,  considering the comparatively vast divergence (3 nucleotide substitutions) between both haplotypes.
In addition, some Psammocinia species appear to be geographically restricted. Several Psammocinia haplotypes represent specimens from a narrow geographical range (e.g. 1909, 2188 Figure 2), reflecting species-level differentiation. Other populations of the same species would have been indicated by the same haplotype, because intraspecific variation in demosponge CO1 is low. For example, nucleotide diversities among populations of Crambe crambe (Poecilosclerida) and Astrosclera willeyana (Agelasida) collected from locations several thousand kilometres apart were found as low as π = 0.00049 and π = 0.0006, respectively , . This is in congruence with current views that geographic ranges of sponge species are frequently overestimated  and the number of distinct species is higher than expected , .
In conclusion, at the dawn of sponge barcoding our data indicate the great potential for DNA technologies to assist in resolving the taxonomy of sponges. From the examples of Ircinia and Psammocinia we demonstrate that barcoding facilitates rapid assessment of biodiversity, radiation patterns and the detection of cryptic speciation. However, the CO1 standard barcoding fragment should be used in combination with another DNA marker in order to achieve unambiguous taxonomic identification at species level.
Materials and Methods
All specimens were collected by scuba, trawl, or dredge. A piece of about 3 mm3 was taken and DNA was extracted with the DNeasy Tissue kit by QIAGEN (Hilden, Germany) following the protocol for animal tissue. Among the specimens of the taxon set is the holotype of Psammocinia bulbosa (G304689) and Strepsichordaia lendenfeldi (Z5026, Thorectidae), the latter was used as the outgroup for phylogenetic reconstructions. The CO1 fragments were amplified using a twofold-degenerated version of the universal barcoding primers: dgLCO1490 (GGT CAA CAA ATC ATA AAG AYA TYG G) and dgHCO2198 (TAA ACT TCAG GGT GAC CAA ARA AYC A)  with an annealing temperature of 43°C. The PCR product was purified in a second step with silica based method described in Boyle & Lew, 1995. The sequencing reaction was performed with the BigDye-Terminator Mix v3.1 (ABI) following the manufacturers protocol. The template was sequenced on an ABI 3100 automated sequencer. The poriferan origin of the sequences was checked by a BLAST search  against the NCBI Genbank collection (http://www.ncbi.nlm.nih.gov/). Only sponge sequences were analyzed. Sequences were base-called, clipped and assembled by CodonCode Aligner v 2.0.4. MacClade v.4.06  was used for the sequence management including the estimation of haplotype frequency in the data set. Sequences were aligned in Sea-View  using the Muscle  algorithm. Due to the protein coding nature of the sequence, the alignment has been unambiguous. Sequences are deposited in the European Molecular Biology Laboratory (EMBL) under accession numbers FN552810 - FN552875, together with the photographic documentation in the sponge barcoding database (www.spongebarcoding.org). Evolutionary distances and parsimonious tree reconstructions were performed with PAUP 4b10  using heuristic searches in order to display haplotype diversity and relationships.
Haplotypes were phylogenetically reconstructed with Bayesian inference methods using MrBayes 3.12b  under the HKY+G model as suggested by Modeltest 3.7 . Two runs with four Metropolis-coupled chains each were run until the standard deviation of split frequencies dropped below 0.01. Trees were burned in until the distribution of topology likelihoods reached the plateau phase. The map was drawn with MAKE_MAP (http://www.aquarius.ifm-geomar.de/make_map.html).
Species list and collection details of the samples included in the data set.
(0.06 MB PDF)
(0.02 MB PDF)
We thank Melissa Robinson and Merrick Ekins for subsampling the Queensland Museum collections.
Conceived and designed the experiments: GW. Performed the experiments: JP. Analyzed the data: JP PS DE. Contributed reagents/materials/analysis tools: PS JH. Wrote the paper: JP DE.
- 1. Faulkner DJ (2000) Highlights of marine natural products chemistry (1972-1999). Natural Product Reports 17: 1–6.
- 2. Chombard C, Boury-Esnault N, Tillier A, Vacelet J (1997) Polyphyly of “sclerosponges” (Porifera, Demospongiae) supported by 28S ribosomal sequences. Biological Bulletin 193: 359–367.
- 3. Fromont J, Bergquist PR (1990) Structural characters and their use in sponge taxonomy: When is a sigma not a sigma? In: Rützler K, editor. New perspectives in Sponge Biology. Washington D.C.: Smithonian Institution Press. pp. 273–278.
- 4. Mikkelsen P, Cracraft J (2001) Marine biodiversity and the need for systematic inventory. Bulletin of Marine Science 69: 525–534.
- 5. Knowlton N (1993) Sibling Species in the Sea. Annu Rev Ecol Syst 189–216.
- 6. Solé-Cava AM, Boury-Esnault N (1999) Patterns of intra and interspecific genetic divergence in marine sponges. In: Hooper JNA, editor. Memoirs of the Queensland Museum, Proceedings of the 5th International Sponge Symposium, Brisbane, June–July 1998. pp. 591–602.
- 7. Wörheide G, Hooper JNA (1999) Calcarea from the Great Barriea Reef. 1: Cryptic Calcinea from Heron Island and Wistari Reef (Capricorn-Bunker Group). Memoirs of the Queensland Museum 43: 859–891.
- 8. Blanquer A, Uriz MJ (2007) Cryptic speciation in marine sponges evidenced by mitochondrial and nuclear genes: A phylogenetic approach. Molecular Phylogenetics and Evolution 45: 392–397.
- 9. Cook SdC, Bergquist PR (2002) Order Dictyoceratida Minchin, 1900. In: Hooper JNA, Soest RWM van, editors. Systema Porifera A guide to the classification of sponges. New York, Boston, Dordrecht, London, Moscow: Kluwer Academic/Plenum Publishers. 1021 p.
- 10. Bergquist PR, Cook SdC (2002) Order Dendroceratida Minchin, 1900. In: Hooper JNA, Soest RWM van, editors. Systema Porifera A guide to the classification of sponges. New York, Boston, Dordrecht, London, Moscow: Kluwer Academic/Plenum Publishers. 1067 p.
- 11. Blunt JW, Munro MHG (2003) MarinLit. A database of the literature on marine natural products for use on a macintosh computer prepared and maintained by the Marine Chemistry Group (Department of Chemistry, University of Canterbury: Canterbury, New Zealand).
- 12. Van Soest RWM, Boury-Esnault N, Hooper JNA, Rützler K, de Voogd NJ, et al. (2008) World Porifera database. Available online under http://www.marinespecies.org/porifera.
- 13. Cook SDC, Bergquist PR (1998) Revision of the genus Psammocinia (Porifera: Demospongiae: Dictyoceratida), with six new species from New Zealand. New Zealand Journal of Marine and Freshwater Research 32: 399–426.
- 14. Cook SdC, Bergquist PR (2002) Family Irciniidae Gray, 1867. In: Hooper JNA, Soest RWM van, editors. Systema Porifera A guide to the classification of sponges. New York, Boston, Dordrecht, London, Moscow: Kluwer Academic/Plenum Publishers. pp. 1022–1027.
- 15. Hebert PDN, Cywinska A, Ball SL, deWaard JR (2003) Biological identifications through DNA barcodes. Proceedings of the Royal Society of London (B) 270: 313–332.
- 16. Witt JDS, Threloff DL, Hebert PDN (2006) DNA barcoding reveals extraordinary cryptic diversity in an amphipod genus: implications for desert spring conservation. Mol Ecol 15: 3073–3082.
- 17. Wörheide G, Erpenbeck D, Menke C (2008) The Sponge Barcoding Project - aiding in the identification and description of poriferan taxa. In: Custódio MR, Lôbo-Hajdu G, Hajdu E, Muricy G, editors. Porifera Research: Biodiversity, Innovation & Sustainability. Rio de Janeiro: Museu Nacional de Rio de Janiero Book Series. pp. 123–128.
- 18. Wörheide G, Erpenbeck D (2007) DNA taxonomy of sponges - progress and perspectives. Journal of the Marine Biological Association of the United Kingdom 87: 1629–1633.
- 19. Shearer TL, Van Oppen MJH, Romano SL, Wörheide G (2002) Slow mitochondrial DNA sequence evolution in the Anthozoa (Cnidaria). Molecular Ecology 11: 2475–2487.
- 20. Huang D, Meier R, Todd PA, Chou LM (2008) Slow mitochondrial COI sequence evolution at the base of the metazoan tree and its implications for DNA barcoding. J Mol Evol 66: 167–174.
- 21. Vethaak AD, Chronie RJA, Soest RWM van (1982) Ecology and distribution of two sympatric, closely relates sponge species, Halichondria panicea (Pallas, 1766) and H. bowerbanki Burton, 1930 (Porifera, Demospongiae), with remarks on their speciation. Bijdragen tot de Dierkunde 52: 82–102.
- 22. Raleigh J, Redmond NE, Delahan E, Torpey S, Van Soest RWM, et al. (2007) Mitochondrial Cytochrome oxidase 1 phylogeny supports alternative taxonomic scheme for the marine Haplosclerida. J Mar Biol Assoc Uk 87: 1577–1584.
- 23. Dawson MN, Jacobs DK (2001) Molecular evidence for cryptic species of Aurelia aurita (Cnidaria, Scyphozoa). Biol Bull-Us 92–96.
- 24. Holland BS, Dawson MN, Crow GL, Hofmann DK (2004) Global phylogeography of Cassiopea (Scyphozoa: Rhizostomeae): molecular evidence for cryptic species and multiple invasions of the Hawaiian Islands. Mar Biol 1119–1128.
- 25. Meyer CP, Geller JB, Paulay G (2005) Fine scale endemism on coral reefs: Archipelagic differentiation in turbinid gastropods. Evolution 59: 113–125.
- 26. Heim I, Nickel M, Brümmer F (2007) Phylogeny of the genus Tethya (Tethyidae: Hadromerida: Porifera): molecular and morphological aspects. J Mar Biol Ass 87: 14.
- 27. Park MH, Sim CJ, Baek J, Min GS (2007) Identification of genes suitable for DNA barcoding of morphologically indistinguishable Korean Halichondriidae sponges. Molecules and Cells 23: 220–227.
- 28. Wörheide G, Nichols SA, Goldberg J (2004) Intragenomic variation of the rDNA internal transcribed spacers in sponges (Phylum Porifera): implications for phylogenetic studies. Molecular Phylogenetics and Evolution 33: 816–830.
- 29. Erpenbeck D, Hooper JNA, Wörheide G (2006) CO1 phylogenies in diploblasts and the ‘Barcoding of Life’ - are we sequencing a suboptimal partition? Molecular Ecology Notes 6: 550–553.
- 30. López-Legentil S, Pawlik J (2008) Genetic structure of the Caribbean giant barrel sponge Xestospongia muta using the I3-M11 partition of COI. Coral Reefs 9.
- 31. Schlacher-Hoenlinger MA, Pisera A, Hooper JNA (2005) Deep-sea “lithistid” assemblages from the Norfolk Ridge (New Caledonia), with description of seven new species and a new genus (Porifera, Demospongiae). Zoosystema 27: 649–698.
- 32. Bifulco G, Bruno I, Minale L, Riccio R, Debitus C, et al. (1995) Bioactive prenylhydroquinone sulfates and a novel C-31 furanoterpene alcohol sulfate from the marine sponge, Ircinia sp. J Nat Products 58: 1444–1449.
- 33. Nichols SA, Barnes PAG (2005) A molecular phylogeny and historical biogeography of the marine sponge genus Placospongia (Phylum Porifera) indicate low dispersal capabilities and widespread crypsis. Journal of Experimental Marine Biology and Ecology 323: 1–15.
- 34. Duran S, Pascual M, Turon X (2004) Low levels of genetic variation in mtDNA sequences over the western Mediterranean and Atlantic range of the sponge Crambe crambe (Poecilosclerida). Marine Biology 144: 31–35.
- 35. Wörheide G (2006) Low variation in partial cytochrome oxidase subunit I (COI) mitochondrial sequences in the coralline demosponge Astrosclera willeyana across the Indo-Pacific. Marine Biology 148: 907–912.
- 36. Klautau M, Russo CAM, Lazoski C, Boury-Esnault N, Thorpe JP, et al. (1999) Does cosmopolitanism result from overconservative systematics? A case study using the marine sponge Chondrilla nucula. Evolution 53: 1414–1422.
- 37. Bierne N, Bonhomme F, David P (2003) Habitat preference and the marine-speciation paradox. P Roy Soc B-Biol Sci 1399–1406.
- 38. Palumbi SR, Grabowsky G, Duda T, Geyer L, Tachino N (1997) Speciation and population genetic structure in tropical Pacific Sea urchins. Evolution 51: 1506–1517.
- 39. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990) Basic Local Alignment Search Tool. Journal of Molecular Biology 215: 403–410.
- 40. Maddison WP, Maddison DR (1992) MacClade: Analysis of phylogeny and character evolution. 3 ed. Sunderland, Massachusetts: Sinauer Associates.
- 41. Galtier N, Gouy M, Gautier C (1996) SEAVIEW and PHYLO_WIN: Two graphic tools for sequence alignment and molecular phylogeny. Computer Applications in the Bioscience 12: 543–548.
- 42. Edgar RC (2004) MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Research 32: 1792–1797.
- 43. Swofford DL (2002) PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods). 4 ed. Sunderland, Massachusetts: Sinauer Associates.
- 44. Ronquist F, Huelsenbeck JP (2003) MrBayes 3: Bayesian phylogenetic inference under mixed models. BIOINFORMATICS 19: 1572–1574.
- 45. Posada D, Crandall KA (1998) MODELTEST: testing the model of DNA substitution. Bioinformatics 14: 817–818.