https://orcid.org/0000-0002-2745-1704
Figures
Abstract
Class Demospongiae is the largest in the phylum Porifera (Sponges) and encompasses nearly 8,000 accepted species in three subclasses: Keratosa, Verongimorpha, and Heteroscleromorpha. Subclass Heteroscleromorpha contains ∼90% of demosponge species and is subdivided into 17 orders. The higher level classification of demosponges underwent major revision as the result of nearly three decades of molecular studies. However, because most of the previous molecular work only utilized partial data from a small number of nuclear and mitochondrial (mt) genes, this classification scheme needs to be tested by larger datasets. Here we compiled a mt dataset for 136 demosponge species—including 64 complete or nearly complete and six partial mt-genome sequences determined or assembled for this study—and used it to test phylogenetic relationships among Demospongiae in general and Heteroscleromorpha in particular. We also investigated the phylogenetic position of Myceliospongia araneosa, a highly unusual demosponge without spicules and spongin fibers, currently classified as Demospongiae incertae sedis, for which molecular data were not available. Our results support the previously inferred sister-group relationship between Heteroscleromorpha and Keratosa + Verongimorpha and suggest five main clades within Heteroscleromorpha: Clade C0 composed of order Haplosclerida; Clade C1 composed of Scopalinida, Sphaerocladina, and Spongillida; Clade C2 composed of Axinellida, Biemnida, Bubarida; Clade C3 composed of Tetractinellida; and Clade C4 composed of Agelasida, Clionaida, Desmacellida, Merliida, Suberitida, Poecilosclerida, Polymastiida, and Tethyida. The inferred relationships among these clades were (C0(C1(C2(C3+C4)))). Analysis of molecular data from M. araneosa placed it in the C3 clade as a sister taxon to the highly skeletonized tetractinellids Microscleroderma sp. and Leiodermatium sp. Molecular clock analysis dated divergences among the major clades in Heteroscleromorpha from the Cambrian to the Early Silurian, the origins of most heteroscleromorph orders in the middle Paleozoic, and the most basal splits within these orders around the Paleozoic to Mesozoic transition. Overall, the results of this study are mostly congruent with the accepted classification of Heteroscleromorpha, but add temporal perspective and new resolution to phylogenetic relationships within this subclass.
Citation: Lavrov DV, Diaz MC, Maldonado M, Morrow CC, Perez T, Pomponi SA, et al. (2023) Phylomitogenomics bolsters the high-level classification of Demospongiae (phylum Porifera). PLoS ONE 18(12): e0287281. https://doi.org/10.1371/journal.pone.0287281
Editor: Michael Schubert, Laboratoire de Biologie du Développement de Villefranche-sur-Mer, FRANCE
Received: June 1, 2023; Accepted: November 15, 2023; Published: December 4, 2023
Copyright: © 2023 Lavrov 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.
Data Availability: All newly sequenced and/or assembled mtDNA sequences were deposited to the Genbank under accession numbers OM729606-OM729671, All trees and alignments are available at the website https://lavrovlab.github.io/Demosponge-phylogeny.
Funding: This work was supported by the National Science Foundation’s Assembling the Tree of Life program (DEB No. 0829783 to DVL, DEB No. 0829986 to RWT, as well as DEB awards 0829763, 0829791), grant BFU2008-00227/BMC and PID2019-108627RB-I00 of the Spanish Ministry of Sciences and Innovation to MM, and by internal funds from Iowa State University. In addition, we gratefully acknowledge postdoctoral funding from the Irish Research Council to CCM. 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.
Introduction
The phylum Porifera (sponges) consists of four taxonomic classes: Demospongiae, Homoscleromorpha, Hexactinellida, and Calcarea [1]. Among them, class Demospongiae Sollas 1895 is by far the largest (>82% of accepted species) and morphologically the most diverse [2]. Demosponges are found in both freshwater and marine environments from intertidal zone to abyssal depth and include familiar commercial sponges [3] as well as such oddities as carnivorous sponges [4]. Demosponges fulfill several important roles in benthic ecosystems, being essential players both in the carbon flux [5] and in the silicon cycle [6]. In addition, sponges have the capacity to modify boundary flow as they pump large volumes of seawater into the water column [7]. With the decline of reef-building corals on tropical reefs, sponges are becoming one of the most important structural elements in these ecosystems and provide shelter to a variety of other species [8, 9]. From an evolutionary perspective, sponges –– one of the two main candidates for being the sister group to the rest of the animals [10, 11] –– provide insight into the common ancestor of all animals [12, 13]. This knowledge, in turn, can improve our understanding of the origin of animal multicellularity and evolution of animal body plans (reviewed in [14, 15]). Indeed, several genomic [12, 16–19] and transcriptomic [13, 20, 21] studies of demosponges have been used to infer steps in animal evolution as well as to clarify various aspects of sponge biology.
The relationships among the higher taxa of demosponges have been studied since the second half of the 19th century but are still only partially resolved (reviewed in [22]). The most recent pre-molecular taxonomic treatment of the phylum Porifera, Systema Porifera [23], subdivided demosponges into three subclasses, and 14 orders, but warned that “resolving the higher systematics of sponges is clearly beyond the capabilities of this present book” [24]. The advent of molecular systematics led to the rejection of many higher-level taxa defined based on morphological and embryological data and to the recognition of four major lineages within the class: Keratosa (G1) (Dictyoceratida + Dendroceratida), Verongimorpha (G2) (Chondrosida, Halisarcida, and Verongida), Marine Haplosclerida (G3), and the remaining orders (G4) (at the time, Agelasida, Hadromerida, Halichondrida, Tetractinellida (Astrophorina + Spirophorina), Poecilosclerida, and Spongillina (freshwater Haplosclerida)) [25–28]. In addition, a confluence of ultrastructural, embryological and molecular studies resulted in removal of the former sub-class Homoscleromorpha Lévi, 1973 from Demospongiae to form the fourth class of Porifera [1].
A revised classification of the Demospongiae was proposed in 2015 that united G3 and G4 into the subclass Heteroscleromorpha and subdivided the latter group into 17 orders [29]. However, it left phylogenetic relationships among these orders mostly unresolved. Furthermore, because the new system—now accepted as the framework for the demosponge classification in the World Porifera Database http://www.marinespecies.org/porifera/index.php—was based primarily on partial nuclear rRNA and mitochondrial cox1 data, it needed to be tested with larger datasets.
Mitochondrial genomes (mt-genomes) provide two primary data types for phylogenetic inference: DNA sequence and gene arrangements. Mt-sequence data (often translated and concatenated coding sequences) have been used extensively in molecular phylogenetics [30–34] and are particularly well suited for the analysis of demosponge relationships because of the low rate of evolution and relatively homogeneous composition of mtDNA sequences in this group [26, 35–38]. Mt-gene arrangement data have also been used both for the reconstruction of global animal relationships [26, 35] and for testing specific phylogenetic hypotheses (e.g., [39]). However, this dataset provides fewer characters for phylogenetic inference and its analysis is computationally more challenging, so the latter application is more common.
As part of the Porifera Tree of Life project https://portol.org/, we determined mtDNA sequences from 64 demosponges and assembled six more from available genomic and transcriptomic data. Here we report these data and use them, along with publicly available mt-genomes to reconstruct and test demosponge phylogenetic relationships. Importantly, we included in our dataset Myceliospongia araneosa Vacelet & Pérez, 1998 (currently classified as Demospongiae incertae sedis) and resolved its phylogenetic position. In addition, we tested the phylogenetic affinities of several demosponge species used for genomic projects and conducted a molecular clock analysis of demosponge evolution.
Materials and methods
Data collection
Overview.
For this project, we PCR amplified, sequenced, assembled, and annotated complete or nearly complete mtDNA sequences from 57 species of demosponges including 48 belonging to the subclass Heteroscleromorpha. Two additional mt-genome sequences were assembled from low coverage Illumina DNAseq data generated for this project. Furthermore we assembled and annotated five complete mtDNA sequences from publicly available DNAseq and RNAseq data (Table 1). Combined with 61 previously published mt-genomes, we compiled a dataset of 125 complete and nearly complete mt-genomes https://lavrovlab.github.io/Demosponge-phylogeny/published.html. To this dataset we added partial mtDNA sequences from one species of Merliida and one species of Desmacellida as well as a partial mtDNA sequence from an unknown species most closely related to Plenaster craigi. In addition, we utilized available RNAseq data from Scopalina sp. CDV-2016 and newly generated DNAseq data from Hymerhabdia typica and Mycale escarlatei to assemble most of the mitochondrial coding sequences for these species. Finally, we used mt-coding sequences from five incomplete mt-genomes reported by Plese et al. (2021), resulting in a final dataset of 136 taxa.
Taxon sampling.
Species used in this study were chosen to cover much of the suprafamilial diversity in Heteroscleromorpha (Table 1). Eleven additional species of demosponges outside of this subclass were sampled and included in the analyses as outgroups (Table 2). Tissue subsamples, and/or DNA aliquots were derived from three main sources: (1) PorToL-supported expeditions to the Smithsonian Tropical Research Institute at Bocas del Toro, Panama, in 2009 and 2012, where subsamples of specimens were stored in 3M Guanidinium Chloride solution; (2) the Moorea Biocode project, Moorea, French Polynesia (https://geome-db.org/workbench/project-overview?projectId=75), where subsamples of specimens were stored in 3M Guanidinium Chloride solution, and (3) collection effort by C.C.M with the Ulster Museum, Belfast with samples stored in 94% ethanol. Vouchers for these samples were deposited to the the National Museum of Natural History (Washington, USA) (sample IDs starting with USNM and BMOO) and the Belfast Ulster Museum (Belfast, Northern Ireland) (BELUM MC). Additional specimens were collected by M.M. from Blanes coast (41°40.21’N; 2°48.14’E) (sample IDs starting with BL) or were subsampled from the Florida Atlantic University Harbor Branch Oceanographic Institute collection (sample IDs starting with HBOM). Myceliospongia araneosa was collected by T.P. from the “3PP” cave near La Ciotat (43°09.47’N—05°36.01’E) in the Mediterranean Sea. Thymosia sp. was collected by M.M. at the Chafarinas Islands (35°11.05’N; 2°26.08’E) in the Mediterranean Sea. Further samples were provided by Steven Cook (SDCC-NZ-363), Alexander Ereskovsky (MRS0816 and MRS1163), April Hill (CLOOS), Gisele Lôbo-Hajdu (GLH1203), Michael Nickel (TW), Julie Reveillaud (HPRUV), and Gert Wörheide (GW960 and GW1144). The Research activities at the Bocas del Toro Research Station of the Smithsonian Tropical Research Institute in Panama and export of biological materials were conducted with permission of the Autoridad de los Recursos Acuaticos de Panama. No permits were required to sample other specimens.
DNA extraction, PCR amplification.
Collected sponge samples were preserved in either 95% ethanol or 3M Guanidinium Chloride solution. Total DNA was extracted with a phenol-chloroform method modified from [40]. Porifera-optimized conserved primers developed in our laboratory [26] were used to amplify short (400–1000 nucleotide) fragments of several mitochondrial genes for each species. Two species-specific primers were designed for each of these genes for PCR amplification. Complete mtDNA was amplified in several overlapping fragments using the Long and Accurate (LA) PCR kit from TAKARA.
Sequencing.
Three sequencing technologies were utilized in the project (Tables 1 and 2). MtDNA sequences from 11 species were determined using the Sanger method [41]. For each of these species all LA-PCR fragments were combined in equimolar concentrations, sheared into pieces 1–2 kb in size and cloned using the TOPO Shotgun Subcloning Kit from Invitrogen. Colonies containing inserts were collected, grown overnight in 96-well blocks and submitted to the DNA Sequencing and Synthesis Facility of the ISU Office of Biotechnology for high-throughput plasmid preparation and sequencing on the facility’s Applied Biosystems 3730xl DNA Analyzer. Gaps in the assembly were filled by primer-walking.
MtDNA sequences from three species were determined using 454’s sequencing technology. PCR reactions for each species were combined in equimolar concentration, sheared and barcoded as described in Gazave et al. [42]. Barcoded PCR fragments were combined together and used for the GS FLX Titanium library preparation (454 Life Sciences). Pyrosequencing was carried out on a Genome Sequencer FLX Instrument (454 Life Sciences) at the University of Indiana Center for Genomics and Bioinformatics.
Finally, mtDNA sequences for 46 species were determined by using Illumina technology. For these species, PCR reactions for each species were combined in equimolar concentration, sheared and combined together with or without barcoding. Libraries were prepared using the Illumina TruSeq DNA PCR-Free Library Prep Kits. Sequencing was carried on MiSeq and HiSeq instruments at the DNA Sequencing and Synthesis Facility of the ISU Office of Biotechnology.
Sequence assembly.
Different assemblers were used depending on the type of data collected. The STADEN package v. 1.6.0 [43] with Phred basecaller [44, 45] was used to assemble the Sanger sequences. Abyss [46], Mira [47], PCAP [48], and SPAdes [49] were used to assemble 454 and Illumina sequences. In most cases, several programs were utilized for the assembly and results compared and compiled together. When barcodes were used, sequences were first separated by the barcode. When barcodes were not used, species selection was carried out to exclude closely related species from the same library. In the latter case, assembled sequences were identified using short sequences generated for primer design. PCR and Sanger sequencing was used to resolve any ambiguities.
Sequence annotation.
We used flip v. 2.1.1 (http://megasun.bch.umontreal.ca/ogmp/ogmpid.html) to predict ORFs in assembled sequences; similarity searches in local databases and in GenBank using FASTA [50] and NCBI BLAST network service [51], respectively, to identify them. Protein-coding genes were aligned with their homologues from other species and their 5’ and 3’ ends inspected for alternative start and stop codons. Genes for small and large subunit ribosomal RNAs (rns and rnl, respectively) were identified based on their similarity to homologous genes in other species, and their 5’ and 3’ ends were predicted based on sequence and secondary structure conservation. Transfer RNA genes were identified by the tRNAscan-SE program [52]. RNAweasel [53] was used to search for introns in the coding sequences. The exact positions of introns were adjusted based on alignments of coding sequences that contained them.
Phylogenetic inference
Phylogenetic analysis based on mitochondrial coding sequences.
Two datasets were constructed for phylogenetic analysis based on mitochondrial coding sequences. The first dataset comprised 136 demosponge species for which we had complete/nearly complete mtDNA data or most individual mt-coding sequences (CDS). Inferred amino acid sequences of individual mitochondrial proteins were aligned with Mafft v7.475 [54] using L-INS-i stategy. Conserved blocks within the alignments were selected with Gblocks 0.91b [55] using relaxed parameters (parameters 1 and 2 = 0.5, parameter 3 = 8, parameter 4 = 5, all gap positions in parameter 5). Cleaned alignments were concatenated into a supermatrix containing 3,634 amino acid positions for 136 species. The second dataset comprised the same 136 demosponge species plus nine additional species of Homoscleromorpha and was constructed as the first dataset, except the concatenated alignment was filtered with CD-Hit [56] to remove sequences with >95% identity. The final alignment for the second dataset contained 83 species and 3,633 amino acid positions. Both datasets were analyzed with PhyloBayes MPI 1.9 [57] under the CAT+GTR model (-cat -gtr). The chains were sampled every 10th tree after the first 1000 burn-in cycles to calculate consensus trees.
Gene order analysis.
Mitochondrial gene orders were converted to gene adjacency matrices using the gogo program (https://github.com/dlavrov/bio-geneorder, unpublished). The matrices were further modified as required by TNT and RAxML and used in these programs to infer the Maximum Parsimony (MP) and Maximum Likelihood (ML) trees, accordingly. For the parsimony analysis, we tried both the traditional (i.e., random addition of sequences + TBR branch swapping followed by additional branch swapping of trees in memory) and “new technology” (i.e., with Ratchet, Tree-Drifting, and Tree-Fusing followed by additional branch swapping) strategies implemented in TNT [58]. Run scripts for the TNT analysis are available in the supplementary GitHub repository. For the ML analysis, we used the multistate model in RAxML-NG v. 1.1.0 [59]: “raxml-ng --all --msa 77taxa.phy --model MULTI13_MK.” To check for the effect of more frequent tRNA rearrangements in animal mitochondrial genomes, we created an alternative adjacency matrix based, where position of each gene was recorded relative to the position of the closest major (i.e., protein or rRNA) upstream and downstream genes and repeated the MP and ML analyses.
Molecular clock analysis
PhyloBayes 4.1c [60] was used for the molecular clock analysis with the fixed tree topology inferred from mitochondrial coding sequences. The model of sequence evolution was the same as for unconstrained phylogenetic analysis: a generalized time-reversible (GTR) amino acid substitution matrix (-gtr), a Dirichlet mixture profile (-cat), and a discrete gamma distribution with four categories -dgam [4]. We used the log normal (Brownian) autocorrelated clock [61] (-ln) model for the analysis and ran two chains for >15,000 cycles. Convergence was assessed by estimating discrepancies and effective sizes for continuous variables in the model using tracecomp with 250 generations removed as burn-in (‘tracecomp -x 250’). Three calibration points were utilized for the analysis: A uniform prior between 541 and 515 MA (beginning of Cambrian—crown-group heteroscleromorph fossil [62]) was placed on the root of the demosponges (split between Keratosa+Verongimorpha and Heteroscleromorpha). The split between Spongillida and Vetulina sp. was constrained between 410 and 298 MY (the lower bound is defined by the observation that species diversification in lakes prior to the Devonian was limited by low nutrient loads and high sediment loads [63], the upper bound is defined by the first reported freshwater sponge fossil [64]). The origin of the crown group Baikal sponges (the split between Baikalospongia intermedia profundalis and Lubomirskia baikalensis) between 30 and 6 MY based on Lake Baikal history and fossil record of Lake Baikal sponges [65]. All calibration ranges were specified as soft bounds (-sb option), which allocates 0.025 of the total probability outside the specified bounds. Dates were assessed by running readdiv with 250 generations removed as burn-in and every 10th generation sampled for each analysis (‘readdiv -x 250 10‘). The chain with greater number of points was utilized for each molecular clock method.
Results
Mt-genome organization
Structure and gene content.
Most newly characterized mt-genomes were circular-mapping molecules, each containing a conserved set of 14 protein-coding, two ribosomal RNA (rRNA) and 24 or 25 tRNA genes (Fig 1). However, we found the following exceptions to this typical organization:
- The mitochondrial genome of the poecilosclerid Mycale escarlatei did not assemble into a single circular molecule. Instead, several alternative arrangements have been found for most mitochondrial genes, indicating an unusual and likely multi-chromosomal genome architecture. Preliminary data from several other species in the genus Mycale suggest a similar organization (unpublished).
- The mitochondrial genome of the Scopalina sp. assembled into three contigs with AT-rich sequences at the ends of each of them. The contig containing cox1 had twice the coverage of the other two. It is not clear whether these contigs represent individual chromosomes or are results of a genome duplication/mis-assembly.
- atp9 was not identified in mt-genomes of Lissodendoryx sp., Chondropsidae sp. MO1046 (Poecilosclerida) and Neopetrosia sigmafera (Haplosclerida). Both atp9 and atp8 were missing in those of Niphates digitalis and N. erecta (Haplosclerida).
- No stop codon was identified in nad3 of Lissodendoryx sp. and closely related Tedania ignis. Furthermore, this gene was immediately followed by in-frame nad4L, suggesting that the two genes are fused in these species.
- Loss of multiple mt-tRNA genes was inferred in the two Niphates species. In addition, we observed:
- (a). a loss of trnC(gca) in Crambe crambe;
- (b). a loss of trnK(uuu) in Neopetrosia sigmafera;
- (c). losses of trnD(guc) in Agelas schmidtii and closely related Astroclera willeyana, as well as Clathria curacaoensis;
- (d). putative losses of trnP(ugg) and trnL(uag) in Scopalina spp;
- We also observed a few unusual and/or redundant tRNAs in newly sequenced mt-genomes:
- (a). trnI(aau) instead of the usual trnI(gau) was found in Adreus fascicularis;
- (b). trnR(acg) instead of trnR(ucg) was found in Cliona varians;
- (c). trnL(caa), a third gene for Leucine tRNA was found in Heteroxya beauforti;
- (d). trnY(aua), in addition to trnY(gua), was found in Negombata magnifica;
- (e). unusual trnX(uua) that would be predicted to read the stop codon UAA was found in Stelligera stuposa but had the lowest cove score among all tRNAs in this species;
- (f). trnA(ugc) was present twice in Stylissa carteri as in closely related Axinella corrugata, where it was shown to be recruited from trnT(ugu) [66];
- (g). trnR(ucu) was present twice in Vetulina sp.
Superscript number associated with each species name refers to a unique gene order marked by the same number. Species are grouped mainly by taxonomic orders. In a few cases when taxonomy of a species is inconsistent with the results of phylogenetic analyses, the species is grouped according to the latter results. Because phylogenetic analysis does not place Petromica sp. in any accepted orders of demosponges, its systematic position is marked as “???”. Protein and rRNA genes (larger boxes) are: atp6, 8–9—subunits 6, 8 and 9 of the F0 ATPase, cox1–3—cytochrome c oxidase subunits 1–3, cob—apocytochrome b (cob), nad1–6 and nad4L—NADH dehydrogenase subunits 1–6 and 4L, rns and rnl—small and large subunit rRNAs. tRNA genes (smaller boxes) are abbreviated using the one-letter amino acid code. The two arginine, isoleucine, leucine, and serine tRNA genes are differentiated by numbers with trnR(ucg) marked as R1, trnR(ucu)—as R2, trnI(gau)—as I1, trnI(cau)—as I2, trnL(uag)—as L1, trnL(uaa) as L2, trnS(ucu)—as S1, and trnS(uga)—as S2. All genes are transcribed from left to right. Genes are not drawn to scale and intergenic regions are not shown. Missing sequences are indicated by gray boxes.
Most new demosponge mitochondrial genomes were in 16–22 kbp size range, with a mean size of ∼20.8 kbp. A few mitochondrial genomes were larger in size (>30kpb in Dysidea etheria), mostly due to the expansions of non-coding regions. All analyzed mitochondrial genomes had similar nucleotide composition (A+T content between 56–74%) and, with two exceptions, displayed overall negative AT- and positive GC-skews of the coding strand (the two exceptions were Dysidea etheria with AT-skew = 0.01 and Scopalina sp., with AT-skew = 0.03).
Gene order.
Mt-genomes of newly characterized demosponges shared between 0 to 41 gene boundaries. The largest differences, as expected, occurred among subclasses of demosponges and are exacerbated by the loss of the majority of tRNA genes in Keratosa, but also Chondrosia reniformis and the two Niphates species. Most of the differences in mitochondrial gene orders were caused by transpositions of tRNA genes. However, rearrangements of “major” (protein and rRNA) genes were also present. Furthermore, Aiolochroia crassa and the underscribed Pseudoceratinidae species MO1014 had the same inversion in mtDNA as previously studied Aplysina species with 20 genes transcribed in the opposite direction comparing to rnl and the rest of the genes. No inversions were found in Heteroscleromorpha mtDNA: all genes had the same transcriptional polarity.
Introns in cox1, cox2, and rnl.
Seventeen cox1 introns have been found in ten heteroscleromorph sponges sampled for this study: one in Adreus fascicularis (order Tethyida), one in Acanthella acuta (order Bubarida), two in each Axinella polypoides and Axinella infundibuliformis (order Axinellida), two in Cliona varians (order Clionaida), one in Phakellia ventilabrum (order Bubarida), two in “Svenzea” flava (classified as Scopalinida, but—based on our data—closely related to Acanthella acuta and P. ventilabrum), three in Leiodermatium sp., one in Microscleroderma sp., and two in Myceliospongia araneosa at positions. In addition, two cox1 introns were found in the Verongimorpha Thymosia sp., doubling the number of previously known demosponge orders with mt-introns. Unexpectedly, we also found group II introns in two other mitochondrial genes: cox2 of Acanthella acuta and rnl of A. acuta, Dictyonella marsilii, and another Dictyonellidae species (P0911). The cox2 intron in Acanthella acuta contained an ORF most similar to the group II intron reverse transcriptase/maturase of a microalga Ulva ohnoi. The rnl intron in the two Dictyonellidae species was found in the same position as in the three placozoan species [67] and contained a large region that displayed high sequence similarity to a region within mt-lrRNA gene in those species. The structures and phylogenetic affinities of introns found in reported mt-genomes are being analyzed in a separate study (unpublished).
Phylogenetic analyses based on a supermatrix of inferred amino acid sequences
Bayesian phylogenetic analysis based on concatenated amino-acid sequences derived from mitochondrial protein genes from 136 species of demosponges yielded a well-supported consensus tree of demosponge relationships with the mean posterior probability support of bipartitions of 0.94 and 73% of bipartitions having maximum support Fig 2. Analysis of a subset of these species with six additional homoscleromorph species (outgroup) allowed us to position the root of demosponges on a branch between Keratosa + Verongimorpha and Heteroscleromorpha (Fig 1). Most heterosclermorph orders proposed by Morrow and Cardenas [29] were recovered as monophyletic groups, although, species sampling was limited for most of them. Some cases where a particular species was not placed within its accepted order (e.g., Axinella corrugata, A. damicornis, Svenzea flava, and Topsentia ophiraphidites) were probably because that previous order assignment was a misclassification (see [68] and Discussion). Within Heteroscleromorpha, we found support for five major clades (named here C0–C4).
- C0: Order Haplosclerida
- C1: Orders Spongillida, Scopalinida, and Sphaerocladina.
- C2: Orders Axinellida, Biemnida, Bubarida along with Topsentia ophiraphidites and Petromica sp.
- C3: Order Tetractinellida + Myceliospongia.
- C4: Orders Agelasida, Clionaida, Desmacellida, Merliida, Poecilosclerida, Polymastiida, Suberitida, and Tethyida.
Posterior majority-rule consensus tree was obtained from the analysis of concatenated mitochondrial amino acid sequences (3,634 positions) under the CAT+GTR+Γ model in the PhyloBayes-MPI program. The number at each node represents the Bayesian posterior probability. The branches marked by a broken line symbol are shown half of their actual lengths. Five major clades in Heteroscleromorpha are shown as C0–C4.
The phylogenetic relationship among these clades was reconstructed as (C0(C1(C2(C3,C4)))), although the interrelationship among C2, C3, and C4 was only moderately supported.
Phylogenetic analysis based on gene order data
Comparison of mitochondrial gene orders in heteroscleromorph species without major tRNA gene loss revealed that they shared with each other at least six gene boundaries, with the mean number of shared boundaries between a given species and the rest of heteroscleromorphs varying between ∼10 for Plenaster craigi and ∼29 for Svenzea zeai and Spongillida (freshwater sponges) (mean = 24.3). The gene orders were converted into a gene-based matrix, where the identities and transcriptional orientations of the upstream and downstream neighbors of each gene were recorded. The matrix was utilized for maximum likelihood (ML) analysis in RAxML-NG [59] and maximum parsimony (MP) analysis in TNT [58]. The results of these analyses (Fig 3, S2 Fig) generally support the Heteroscleromorpha relationships reconstructed from sequence data, including its subdivision into major clades and the phylogenetic position of Myceliospongia araneosa. However, four main discrepancies were found:
Gene boundaries were encoded as multistate characters based on identity and orientation of each gene. The matrix was used for Maximum Likelihood analysis under the MULTI13_MK model in RAxML-NG. Numbers above branches show the bootstrap support when >50%.
First, C1 clade (Spongillida, Vetulina sp. and Scopalinida) was not reconstructed as a monophyletic group. Instead Vetulina sp. and Scopalina sp. either grouped with Tetractinellida plus Myceliospongia araneosa or was a part of a polytomy at the base of the tree. The lack of support for C1 is not surprising, given that the gene order in Spongillida was inferred to be ancestral for the Heteroscleromorpha [26].
Second, phylogenetic position of the order Agelasida was unstable, as it either grouped with C3 + Vetulina sp. and Scopalina sp. (ML analysis) or its position was unresolved among the main clades. The rest of the C2 clade formed a sister group to all heteroscleromorphs but Haplosclerida.
Third, P. craigi grouped with Agelas schmidti and Astrosclera willeyana within Agelasida. As noted above, P. craigi has the most derived mitochondrial gene order, reflected also in the longest branch in the ML tree (Fig 3).
To check if any inconsistencies between the sequence-based and gene order-based phylogenies were due to saturation of phylogenetic signal in gene order data caused by frequent movement of tRNA genes, we repeated both ML and MP analyses using the multistate encoding based on closest major (protein or rRNA) genes (see [35] for details). Indeed, P. craigi grouped with C2 species in both of these analyses. However, the position of Agelasida was still unresolved and there was no support for the monophyly of Tetractinellida, although Myceliospongia araneosa still grouped with Microscleroderma sp., and Leiodermatium sp.
Finally, we visually identified several unique rearrangements associated with some large lineages of Heteroscleromorpha. Thus, we found that the sampled representatives of the orders Axinellida, Bubarida, Biemnida, along with Topsentia ophiraphidites and Petromica sp. shared a translocation of trnR-nad4L downstream of rnl. Interestingly, trnR-nad4L have also moved to this location in Haplosclerida, but as a part of a much larger genomic fragment that included three additional protein-coding genes: nad1, nad2, and nad5 and several tRNA genes. We also found that orders Suberitida, Polymastiida, Poecilosclerida, and Clionaida shared a translocation of trnY and trnI2 from a tRNA cluster between nad1 and nad2 to the region downstream of rnl. In addition, all Poecilosclerida species had a translocation of cox2 into the gene junction between nad5 and rns, while all Clionaida species had trnV moved immediately downstream of cox1 from its conserved position between trnG and rnl.
Molecular clock analysis
We used the same dataset as for the sequence-based phylogenetic analysis as well as the phylogenetic tree inferred in that analysis to estimate times of divergences among major lineages of demosponges (Fig 4, S3 Fig). Fixing the root of Demosponge phylogeny (between Heteroscleromorpha and Keratosa+Verongimorpha) between 541 and 515 MYA, we estimated that the divergence between Haplosclerida and the the rest of Heteroscleromorpha occurred between 534 and 488 MYA, followed by the split between C1 and C2+C3+C4 between 512–456 MYA, with the splits among C2–C4 between 494 and 417 MYA. Furthermore, most mean estimates for the divergences among orders of Heteroscleromorpha fall between the Ordovician and the Devonian, while those for the basal splits within the orders (that can be used as a proxy for the crown group) between Early Permian to Late Triassic. Several exceptions to this pattern have been found. First, we noted that the split between Myceliospongia araneosa, Microscleroderma sp., and Leiodermatium sp. with the rest of Tetractinellida occurred between 449 and 359 MYA, corresponding to the time of the origin of most orders. Second, the basal split within Axinellida (e.g., between Raspaillidae + Stelligeridae and Axinellidae) was estimated to occur at a similar time, between 438 and 324 MYA. By contrast, Hamacantha johnsoni (order Merliida) and Desmacella informis (order Desmacellida) were estimated to split only between 257 and 96 MYA and to diverge from Poecilosclerida between 363 and 266 MYA. Finally, we note here that the most basal divergence within Haplosclerida, estimated between 516 and 455 MYA, occurred at about the same time as the most basal divergence within the rest of Heteroscleromorpha (between C1 and C2+C3+C4). We suggest that these discrepancies—if confirmed by other datasets—can be used to re-define several orders in Heteroscleromorpha and to revise the taxonomic status of Haplosclerida (see Discussion).
Only heterosleromorph orders are labeled. Numbers at internal nodes indicate their upper and lower age limits. The four large clades within Heteroscleromorpha are shaded in different colors. The vertical lines (left to right) correspond to the Ordovician-Silurian (444 MYA), the Devonian-Carboniferous (359 MYA), and the Permian-Triassic (252 MYA) boundaries. The full tree is shown in S3 Fig.
Discussion
Assembling the new dataset of mtDNA data for demosponges
Despite rapid progress in sequencing technologies, sequence data remain scarce for many taxa of marine animals, including demosponges. When such data are available, they are often limited to partial sequences of individual genes, commonly those for nuclear rRNA and/or mitochondrial cox1. While even these partial data have been instrumental for re-evaluation of demosponge relationships, larger datasets are needed to test and refine proposed phylogenetic hypotheses. Ultimately, a representative sampling of genomic loci should be a dataset of choice for a phylogenetic analysis. However, such datasets are available for only a few species of sponges. Furthermore, assembling and analyzing genomic datasets present various challenges, especially in sponges, where contamination by foreign DNA is always a factor [69].
Here we compiled and analyzed a large dataset of mt-genomic sequences of demosponges, including 66 mtDNA sequences determined or assembled for this study. Importantly, we sampled all but one (Trachycladida) of the proposed orders within Heteroscleromorpha, the largest group of demosponges. We used this dataset to test phylogenetic relationships and reconstruct the timing of major splits within the group. Our results are largely consistent with and add resolution to phylogenetic studies based on single-gene data, compiled and synthesized by Morrow and Cárdenas [29]. For example, our sequence-based phylogenetic analysis not only groups Spongillida with Sphaerocladina (Vetulina)—as proposed by Redmond et al. [70] and investigated in more details by Schuster et al. [71]—but also provides strong support for their association with Scopalinida (clade C1). Similarly, in addition to grouping orders Agelasida, Clionaida, Poecilosclerida, Polymastiida, Suberitida, and Tethyida (reported in [29]; clade C4 in our analysis), our results support a clade comprised of orders Axinellida, Biemnida, Bubarida along with Topsentia ophiraphidites and Petromica sp. (clade C2).
At the same time, there are a few differences between the results of this and previous molecular studies: First, Tetractinellida (C3) forms a sister group to C4 in our analysis rather than to Biemnida [72–74] or C2 (Axinellida, Biemnida, Bubarida) [29]. Second, within the clade C4, Tethyida groups with Suberitida rather than Clionaida [74] or Clionaida & Poecilosclerida [29]. Third, within C2, Bubarida is a sister group to Biemnida, rather than Axinellida [29, 74]. Fourth, Topsentia ophiraphidites and Petromica sp. (classified as Suberitida and incertae sedis, respectively in [29]) are placed within C2 in both sequence-based and gene-order based analyses and form a sister group to Axinellida in the former analysis (similar results are found in citepankey2022). Because there are no molecular data from the type species of either Topsentia or Petromica, the proper classification of these genera remains undetermined. However, it has been suggested that Topsentia could be a polyphyletic genus, with some species belonging to Axinellida and other to Suberitida [72].
Finally, we note that Svenzea flava (Lehnert & van Soest, 1999) groups with Dictyonellidae (Bubarida) species in both sequence and gene order analyses rather than with Scopalinida, which includes the type species Svenzea zeai. This finding is not too surprising, given that S. flava differs from other members of the genus in its skeletal arrangement and lacks characteristic granular cells and large embryos/larvae found in the type species [75]. Thus, in the future, S. flava will need to be renamed and reassigned to the family Dictyonellidae.
Phylogenetic position of emerging sponge genomic model systems.
Our dataset included nine species of sponges for which high throughput DNA and/or RNA data were available and which, therefore, could be considered as emerging sponge model systems. Although the phylogenetic position of most of these species was as expected, two exceptions were found. First, Stylissa carteri (order Scopalinida) grouped closely with Axinella corrugata (order Axinellida) and both of them were placed as a sister group to Axinella damicornis (order Axinellida) and Hymerhabdia typica (order Agelasida) within Agelasida. Furthermore, the coding sequences between Stylissa carteri and Axinella corrugata were >99% identical, indicating a possible misidentification of the sponge. Second, the recently described abyssal sponge Plenaster craigi, grouped with the species from the orders Axinellida, Biemnida, and Bubarida in sequence-based phylogenetic reconstructions, but was not placed in any of these groups. Thus, it should not be consider a representative of Axinellida [76] and its phylogenetic position should be further investigated. Finally, we note that there was a significant level of cross-contamination between DNA sequence data from Xestospongia testudinaria and Stylissa carteri [16], such that both mitochondrial genomes can be assembled from either dataset.
Are haplosclerids heteroscleromorphs?
The current demosponge taxonomy, which we followed in this article, places haplosclerid sponges within the subclass Heteroscleromorpha as the order Haplosclerida [77]. This fits the definition of the subclass as “Demospongiae with a skeleton composed of siliceous spicules which can be monaxons and/or tetraxons and when they are present, microsleres are highly diversified” [29]. However, the same definition was applied earlier to Heteroscleromorpha minus the order Haplosclerida, with the latter group considered as the fourth subclass of demosponges, Haploscleromorpha [22]. Does it matter if we call haplosclerid sponges an order or a subclass? According to Joe Felsenstein, a self-proclaimed founder of the It-Doesn’t-Matter-Very-Much school of classification, “the delimitation of higher taxa is no longer a major task of systematics, as the availability of estimates of the phylogeny removes the need to use these classifications” [78]. Indeed, the position of the order Haplosclerida as the sister group to the rest of the Heteroscleromorpha—inferred in this and several previous studies—is consistent with both the four-subclasses classification system of Cardenas et al. [22] as well as the three-subclasses classification system of Morrow and Cardenas [29]. Nevertheless, we believe that the accepted status of Haplosclerida does influence our treatment of this taxon in two important ways. First, it changes the extent to which this large and diverse group of sponges is and will be represented in comparative studies, with a higher rank leading to a more thorough sampling. Second, it modulates our attention to several unusual features demonstrated by the group, including atypical patterns of evolution of ribosomal RNA genes [79, 80], “streamlined” nuclear genomes [21], and unusual content of the main skeleton-forming genes [81]. While the choice between the three vs. four subclasses classification of Demospongiae will—to a large extent—remain a subjective choice of the sponge taxonomic community, we note that the time estimate for the most basal divergence within Haplosclerida is comparable to that for the most basal divergence within the rest of Heteroscleromorpha (between C1 and C2+C3+C4) rather than the crown group divergences within other heteroscleromorph orders (see also citepankey2022). From this perspective, the classification of Haplosclerida as the fourth subclass of Demospongiae would be preferable
Timing phylogenetic divergences
In addition to resolving phylogenetic relationships among clades of Heteroscleromorpha, we provide molecular clock estimates for the divergences among them. While several molecular clock studies using demosponge mt-genome data have been conducted [36, 38, 82], they utilized much smaller datasets of mitogenomic sequences. Furthermore, two of these studies [38, 82] used a suboptimal selection of the outgroup species, resulting in a likely erroneous phylogenetic reconstruction, which in turn biased molecular dating results (see below).
One major difficulty in molecular clock analysis of sponges is the scarcity and uncertainty of available calibration points. While the Paleozoic record of sponges is well established and extends to the Lowermost Cambrian (529–541 Ma) [83], the taxonomic affinities of Paleozoic sponges are often controversial. Nevertheless, the recent discovery of well-preserved fossils of crown-group demosponges [62] constrain the upper bound for the basal split in demosponges at 515 MY.
The Precambrian record of sponges is even more contentious. Although there have been numerous reports of Precambrian Porifera fossils, most of them are not substantiated [84]. Instead, the current argument for the Precambrian origin of demosponges is based primarily on the presence of fossil steroids (in particular 24-isopropylcholestanes and recently discovered 26-methylstigmastane) in the geological record before the end of the Marinoan glaciation (∼635 MY ago) [85, 86]. However, the problem with both of these biomarkers is their erratic distribution across modern demosponges, which at best can be interpreted as a result of multiple independent losses [86]. Such rampant loss makes it extremely difficult to estimate the origin of this biosynthetic pathway on a phylogenetic tree, especially given the sparse sampling of non-demosponge taxa [86]. Furthermore, the biosynthesis precursors of “sponge biomarkers” are found among Rhizaria, heterotrophic protists common in the ancient and modern oceans [87]. Thus, for the present study, we used the beginning of the Cambrian (541 MY) as the lower bound for the common ancestor of demosponges.
We used the origin of freshwater sponges (approximated by the split between the order Spongillida and Vetulina sp.) as the second calibration point for our analysis. The upper limit for the origin of freshwater sponges was defined by the oldest reported freshwater sponge fossils [64]. The lower bound was defined by a somewhat generic observation that species diversification in lakes prior to the Devonian was limited by low nutrient loads and high sediment loads [63]. There is substantial uncertainty with both of these estimates. First, the fossils we used to define the upper limit of the split predate most other known freshwater sponge fossils (mostly Jurassic and Cretaceous) by more than 100MY [88]. It is not clear if the lack of fossil freshwater sponges between the end of the Palaeozoic and the Jurassic is an artifact of preservation or if the colonization of the freshwater environment by sponges occurred more than once. Furthermore, although our intent was to estimate the origin of freshwater sponges, it is likely that the split between Spongillida and Vetulina happened in marine environment and thus preceded this event.
Our third calibration point was the origin of the crown group Baikal sponges (defined as the split between Baikalospongia intermedia profundalis and Lubomirskia baikalensis) and placed between 30 and 6 MY. The lower bound for this estimate is based on the Lake Baikal age (estimated between 25–30 MY), and lack of Lubomirskiidae spicules in early Tertiary sediments of the Tunkinskaya land basin (approximately Oligocene or 23–33 MYA) [65]. The upper bound is based on the analysis of Baikalian bottom sediment samples conducted during Baikal Drilling Project, which revealed well-formed spicules of several species belonging to all four genera of the Lubomirskiidae family in deposits corresponding to 6,50—4,75 MYA [65].
Given the substantial uncertainties in the calibration points and a large variance associated with molecular clock analysis, divergence times estimates for sponges need to be treated as only rough approximations. Nevertheless, it is illuminating to see that many proposed orders of demosponges have an ancient (Mid-Paleozoic) origin. By contrast, the origin of crown groups within most of them corresponds to the Late-Paleozoic or Mesozoic. These observations suggest that estimated divergence times could be taken into consideration when demosponge classification is revised and/or when alternative classification schemes are considered.
Phylogenetic position of Myceliospongia araneosa
Myceliospongia araneosa, the only described species in the genus Myceliospongia Vacelet & Perez, 1998, is one of the most unusual demosponges in its anatomy and cytology. The sponge is known from a single cave (the “3PP” cave near La Ciotat (43°09.47’N—05°36.01’E)) in the Mediterranean Sea, where it grows on vertical or overhanging walls in a cool stable environment. M. araneosa has an encrusting ‘body’ of up to 25 cm diameter and 1 mm thick surrounded by a reticulation of filaments from as little as 5 μm in diameter. These filaments form an extensive network around the body that can grow over and under other marine life. The sponge body has a reduced aquiferous system and a low number of small choanocyte chambers. Neither spicules nor spongin fibers are present, and collagen fibrils do not form bundles or thick condensations. M. araneosa shows a remarkable uniformity of cell types, with the mesohyl cells being poorly differentiated [89]. Because of the highly unusual organization, which offered little clues as to its affinities to other sponges, Myceliospongia was placed in Demospongiae, incertae sedis [90].
Phylogenetic analyses based both on sequence and gene order data unequivocally placed Myceliospongia within Tetractinellida, and as the sister group to Microscleroderma sp. + Leiodermatium sp. The split between these taxa and the rest of Tetractinellida has been estimated to occur 456—385 MYA, a time more consistent with inter- rather than intra-order divergence (see above). Interestingly, both Microscleroderma sp. and Leiodermatium sp. are “lithistid” sponges characterized by the presence of articulated choanosomal spicules called desmas. It would be interesting to know whether M. araneosa also evolved from a lithistid ancestor, which would predicate losses of both choanosomal and ectosomal skeletons. Overall, losses of mineral skeleton are rare although not unprecedented in Heteroscleromorpha (e.g., Haplosclerida: Dactylia), but can be regarded as a dominant feature in evolution of subclasses Keratosa and Verongimorpha [91]. Independent losses of mineral skeleton have also occurred in various clades of Homoscleromorpha (both within Oscarellidae and also Plakinidae) [92].
Promise, pitfalls, and limitations of Porifera phylomitogenomics
Mitochondrial DNA (mtDNA) in general and animal mtDNA in particular has been a popular and a widely used molecular marker in phylogenetic studies [93]. This popularity is due to a combination of several features that facilitate the acquisition of mtDNA, including its typically circular organization, large copy number per cell, and a high proportion of coding sequences, with those that simplify analysis, such as absence of paralogs and introns, stable gene content and gene order. Phylogenetic analyses of sponge mt-genomes additionally benefit from relatively low rates of sequence evolution and a more homogeneous nucleotide composition (at least within individual classes of sponges) [94]. Indeed, previous studies by our and other groups have demonstrated the utility of mitogenomic datasets for reconstructing phylogenetic relationships within the phylum Porifera and the overall congruence among the results of phylomitogenomic analyses with those based on 18S, 28S, and transcriptomic datasets [26, 36, 42, 95].
Nevertheless, as with any dataset, phylogenetic analysis based on mitogenomic data can fail to reconstruct the true evolutionary history of organisms. The primary challenges faced by a phylogenetic analysis are erroneous (non-phylogenetic) signal due to stochastic noise and/or systematic bias [96] as well as incongruences between gene trees and the species tree due to non-orthology of sequences, horizontal gene transfer, or incomplete lineage sorting [97]. Although large genomic (including mitogenomic) datasets generally benefit from the increase in the ratio of the true phylogenetic signal (which adds across genes) to random noise (which does not), they do not solve the problem of non-phylogenetic signal caused by systematic biases, which also accumulates [98]. Similarly, while paralogs are rare in mtDNA and duplicated sequences are quickly eliminated, potential problems exist both with horizontal gene transfer (primarily through mtDNA introgression) [99] and incomplete lineage sorting [100].
The two well understood factors that can contribute to systematic biases in phylogenomic analyses are heterogeneous sequence compositions and variable evolutionary rates across taxa [96]. While nucleotide composition of mt-genomes is relatively uniform across demosponges (and also similar in homoscleromorphs) [101], it is markedly different between demosponges, glass sponges, and calcareous sponges [102–104]. Because commonly used models in phylogenetic analysis assume a stationary nucleotide/amino-acid composition across the phylogenetic tree, inter-class comparisons of sponges based on mt-sequences can be error-prone. It is important to note that both glass and calcareous sponges also utilize more derived mitochondrial genetic codes, leading to additional differences in patterns of sequence evolution (reviewed in [94]).
In contrast to sequence composition, most standard phylogenetic analyses place no constraints on rates of sequence evolution among branches of a phylogenetic tree. Nevertheless, variable evolutionary rates can result in “long branch attraction” (LBA)—an artificial grouping of taxa with higher rates of sequence evolution [105]. This phenomenon occurs because models of sequence evolution typically underestimate its complexity and thus undercount the number of changes occurring along long branches [106]. One common example of LBA is grouping of fast-evolving in-group taxa with an outgroup, the latter, by its nature, forming a long branch on a phylogentic tree. Thus, the choice of the model and the choice of outgroups are critical for phylogenomic analysis.
The systematic biases discussed above can explain unusual results obtained in two recent studies of demosponge relationships that utilized mitogenomic datasets [38, 82]. First, in the paper by Schuster et al. [82] a time-calibrated phylogeny of Demosponges shows both Haplosclerida and Spongillida nested deep within Heteroscleromorpha with Tetractinellida forming the most basal divergence with the rest of heteroscleromorphs. These unusual placements of Haplosclerida, Spongillida, and Tetractinellida are incongruent with most previous phylogenetic reconstructions as well as with the standard ML and Bayesian trees obtained in the same study but presented only as supplementary data. It is likely that this result is a combination of two factors: the choice of the outgroup (order Dictyoceratida) and the use of molecular clock analysis to co-infer phylogeny. Dictyoceratida are known to form an extremely long branch on the demosponge phylogenetic tree [26], which likely attracts other longer branches to the base of the tree. The suboptimal choice of the outgroup is likely exacerbated by constraints (a prior distribution) placed by molecular clock analysis on the rates of sequence evolution needed for time estimation.
Second, a more recent paper by Plese et al. [38] reconstructed unconventional relationships among subclasses of demosponges, with Verongimorpha grouping with Heterosleromorpha rather than Keratosa. Again, the choice of the outgroup—class Hexactinellida—was unfortunate. Although the accepted view of sponge relationships places Demospongiae + Hexactinellida and Homoscleromorpha + Calcarea as sister groups (reviewed in [107]), mt-genome evolution is more similar in Demospongiae and Homoscleromorpha, whereas the other two classes accumulated multiple idiosyncrasies in mtDNA [94]. In particular, Hexactinellida display patterns of sequence evolution very different from those in demosponges, utilize a distinct mitochondrial genetic code, and have higher rates of mt-sequence evolution [102]. Because Keratosa are also represented by a long branch in a mitogenomic analysis, its inferred position is likely the result of the LBA between this taxon and Hexactinellida.
While one might argue that any phylogeny is just a hypothesis of evolutionary relationships, and that having diverse hypotheses could stimulate discussion and expedite progress, this is often not the case. Conflicting phylogenies derived from the same or similar datasets can diminish confidence in molecular phylogenetics, particularly among biologists who are not specialists in the field. This lack of agreement can also complicate efforts to combine information from multiple phylogenetic datasets, as attempted by projects like OpenTree [108] and TimeTree [109]. Moreover, when inaccurate phylogenies propagate through the biological literature, they can result in misinterpretations of comparative data. Hence, it is crucial to address known issues in phylogenetic reconstruction, both during dataset construction and the actual inference process [69].
Conclusion
In this study, we acquired 64 complete or nearly complete and six partial mt-genome sequences, more than doubling the number of demosponge species available for phylomitogenomic analysis. We compiled these new data with pre-existing records to build a dataset of 136 demosponge species and to test the high-rank phylogeny in the class Demospongiae, in particular the order composition of the subclasses and the relationships between subclasses. Our phylogenetic reconstruction is consistent with a subdivision of Demospongiae into either three subclasses (Verongimorpha, Keratosa, Heterosclermorpha) or four subclasses (Verongimorpha,Keratosa, Heterosclermorpha, Haploscleromorpha). However, we argue that classifying the current heteroscleromorph order Haplosclerida as a fourth subclass of Demospongiae provides better agreement with estimated times of diversification and observed genomic diversity within this group. We confirmed that Heteroscleromorpha forms the sister group to the rest of the demosponges (Haploscleromorpha (Keratosa, Verongimorpha)) and that a recent result claiming closer relationships between Heteroscleromorpha and Verongimorpha is likely an artifact of long branch attraction between Keratosa and the selected outgroup. Analysis of molecular data from Myceliospongia araneosa, a highly unusual skeleton-lacking demosponge, surprisingly placed it within the heteroscleromorph order Tetractinellida, as a sister taxon to the highly skeletonized Microscleroderma sp. and Leiodermatium sp. In general, we argue that mt-genomes provide an informative dataset for studying demosponge phylogenetic relationships but that one must be careful with the choice of outgroups and models of sequence evolution. Finally, we created a website https://lavrovlab.github.io/Demosponge-phylogeny that contains all the data and the trees presented in this paper and that we plan to update as additional mitogenomic data from demosponges become available.
Supporting information
S1 Fig. Rooted phylogenetic tree of Demospongiae.
Posterior majority-rule tree was obtained from the analysis of concatenated mitochondrial amino acid sequences (3,633 positions) under the CAT+GTR+Γ model in the PhyloBayes-MPI program. Mitochondrial coding sequences from nine species of Homoscleromorpha were added to those of demosponges used for the analysis presented in Fig 2 and the resulting dataset was filtered with CD-Hit to remove sequences with >95% identity. The root was placed between Demospongiae and Homoscleromorpha.
https://doi.org/10.1371/journal.pone.0287281.s001
(PDF)
S2 Fig. Maximum Parsimony reconstruction of Demospongiae relationships based on mitochondrial gene boundaries.
Strict consensus of 12800 most parsimonious trees is shown with support values based on 200 bootstrap replicates. All analyses were conducted in TNT v. 1.5 using the “new technology” search options.
https://doi.org/10.1371/journal.pone.0287281.s002
(PDF)
S3 Fig. Time calibrated phylogeny of Demospongiae based on Bayesian analysis in PhyloBayes.
Numbers at internal nodes indicate their upper and lower age limits.
https://doi.org/10.1371/journal.pone.0287281.s003
(PDF)
Acknowledgments
We are grateful to our colleagues Steven Cook, Alexander Ereskovsky, April and Hill, Gisele Lôbo-Hajdu, Michael Nickel, Julie Reveillaud, and Gert Wörheide for samples contributed to this project; Malcolm Hill, Jenna Moore, Bernard Picton, and John Reed for help with specimen collection; Andrzej Pisera and Sven Zea for help with species identification. We also thank the following former graduate and undergraduate students in the Lavrov Lab for their help with the molecular work: Andrea Bekic, Jenessa Filler, Ehsan Kayal, Philip Lange, Katrina Lutap, Benjamin Sheller, Xiujuan Wang, and Katherine Wilson. We are grateful to Nicole Boury-Esnault and three anonymous reviewers for their careful reading of the previous versions of the ms. and many thoughtful suggestions.
References
- 1.
Gazave E, Lapébie P, Ereskovsky AV, Vacelet J, Renard E, Cárdenas P, et al. In: Maldonado M, Turon X, Becerro M, Jesús Uriz M, editors. No longer Demospongiae: Homoscleromorpha formal nomination as a fourth class of Porifera. Dordrecht: Springer Netherlands; 2012. p. 3–10. Available from: https://doi.org/10.1007/978-94-007-4688-6_2.
- 2.
Hooper JNA, Van Soest RWM. Class Demospongiae Sollas, 1885. In: Hooper JNA, Van Soest RWM, editors. Systema Porifera: a guide to the classification of sponges. New York: Kluwer Academic/Plenum Publishers; 2002. p. 15–18.
- 3.
Moore HF. The commercial sponges and the sponge fisheries. In: Bulletin of the Bureau of fisheries, volume xxviii, 1908. Proceedings Of The Fourth International Fishery Congress, Washington, 1908, Volumes 643-671.; 1910.
- 4. Vacelet J, Boury-Esnault N. Carnivorous sponges. Nature. 1995;373(6512):333–335.
- 5. de Goeij J, van Oevelen D, Vermeij M, Osinga R, Middelburg J, de Goeij A, et al. Surviving in a marine desert: the sponge loop retains resources within coral reefs. Science. 2013;342(6154):108–110. pmid:24092742
- 6.
Maldonado M, Ribes M, van Duyl FC. Nutrient fluxes through sponges: biology, budgets, and ecological implications. In: Advances in Marine Biology. vol. 62. Elsevier; 2012. p. 113–182.
- 7. Pawlik JR, McMurray SE. The emerging ecological and biogeochemical importance of sponges on coral reefs. Annual Review of Marine Science. 2019;. pmid:31226028
- 8. Bell JJ. The functional roles of marine sponges. Estuarine, Coastal and Shelf Science. 2008;79(3):341–353.
- 9. Bell JJ, Bennett HM, Rovellini A, Webster NS. Sponges to be winners under near-future climate scenarios. BioScience. 2018;68(12):955–968.
- 10. Simion P, Philippe H, Baurain D, Jager M, Richter DJ, Di Franco A, et al. A large and consistent phylogenomic dataset supports sponges as the sister group to all other animals. Current Biology. 2017;27(7):958–967. pmid:28318975
- 11. Redmond AK, McLysaght A. Evidence for sponges as sister to all other animals from partitioned phylogenomics with mixture models and recoding. Nature Communications. 2021;12(1). pmid:33741994
- 12. Srivastava M, Simakov O, Chapman J, Fahey B, Gauthier ME, Mitros T, et al. The Amphimedon queenslandica genome and the evolution of animal complexity. Nature. 2010;466(7307):720–726. pmid:20686567
- 13. Musser JM, Schippers KJ, Nickel M, Mizzon G, Kohn AB, Pape C, et al. Profiling cellular diversity in sponges informs animal cell type and nervous system evolution. Science. 2021;374(6568):717–723. pmid:34735222
- 14. Dunn CW, Leys SP, Haddock SH. The hidden biology of sponges and ctenophores. Trends in Ecology and Evolution. 2015;30(5):282–291. pmid:25840473
- 15. Renard E, Leys SP, Wörheide G, Borchiellini C. Understanding animal evolution: the added value of sponge transcriptomics and genomics. Bioessays. 2018;40(9):1700237. pmid:30070368
- 16. Ryu T, Seridi L, Moitinho-Silva L, Oates M, Liew YJ, Mavromatis C, et al. Hologenome analysis of two marine sponges with different microbiomes. BMC Genomics. 2016;17:158. pmid:26926518
- 17. Borisenko I, Adamski M, Ereskovsky A, Adamska M. Surprisingly rich repertoire of Wnt genes in the demosponge Halisarca dujardini. BMC Evolutionary Biology. 2016;16(1):123. pmid:27287511
- 18. Francis WR, Eitel M, Vargas R S, Adamski M, Haddock SHD, Krebs S, et al. The genome of the contractile demosponge Tethya wilhelma and the evolution of metazoan neural signaling pathways. bioRxiv. 2017;.
- 19. Kenny NJ, Francis WR, Rivera-Vicéns RE, Juravel K, de Mendoza A, Díez-Vives C, et al. Tracing animal genomic evolution with the chromosomal-level assembly of the freshwater sponge Ephydatia muelleri. Nature Communications. 2020;11(1). pmid:32719321
- 20. Riesgo A, Farrar N, Windsor PJ, Giribet G, Leys SP. The analysis of eight transcriptomes from all poriferan classes reveals surprising genetic complexity in sponges. Molecular Biology and Evolution. 2014;31(5):1102–1120. pmid:24497032
- 21. Guzman C, Conaco C. Comparative transcriptome analysis reveals insights into the streamlined genomes of haplosclerid demosponges. Scientific Reports. 2016;6:18774. pmid:26738846
- 22.
Cárdenas P, Pérez T, Boury-Esnault N. Sponge systematics facing new challenges; 2012.
- 23.
Hooper JNA, van Soest RWM, editors. Systema Porifera: a guide to the classification of sponges. New York: Kluwer Academic/Plenum Publishers; 2002.
- 24.
Hooper JNA, Van Soest RWM, Debrenne F. Phylum Porifera Grant, 1836. In: Hooper JNA, Van Soest RWM, editors. Systema Porifera: a guide to the classification of sponges. New York: Kluwer Academic/Plenum Publishers; 2002. p. 9–13.
- 25. Borchiellini C, Chombard C, Manuel M, Alivon E, Vacelet J, Boury-Esnault N. Molecular phylogeny of Demospongiae: implications for classification and scenarios of character evolution. Molecular Phylogenetics and Evolution. 2004;32(3):823–837. pmid:15288059
- 26. Lavrov DV, Wang X, Kelly M. Reconstructing ordinal relationships in the Demospongiae using mitochondrial genomic data. Molecular Phylogenetics and Evolution. 2008;49(1):111–124. pmid:18583159
- 27. Sperling EA, Peterson KJ, Pisani D. Phylogenetic-signal dissection of nuclear housekeeping genes supports the paraphyly of sponges and the monophyly of Eumetazoa. Molecular Biology and Evolution. 2009;26(10):2261–2274. pmid:19597161
- 28. Hill MS, Hill AL, Lopez J, Peterson KJ, Pomponi S, Diaz MC, et al. Reconstruction of family-level phylogenetic relationships within Demospongiae (Porifera) using nuclear encoded housekeeping genes. PLoS One. 2013;8(1):e50437. pmid:23372644
- 29. Morrow C, Cárdenas P. Proposal for a revised classification of the Demospongiae (Porifera). Frontiers in Zoology. 2015;12:7. pmid:25901176
- 30. Arnason U, Adegoke JA, Bodin K, Born EW, Esa YB, Gullberg A, et al. Mammalian mitogenomic relationships and the root of the eutherian tree. Proceedings of the National Academy of Sciences. 2002;99(12):8151–8156. pmid:12034869
- 31. Lang BF, O'Kelly C, Nerad T, Gray MW, Burger G. The closest unicellular relatives of animals. Current Biology. 2002;12(20):1773–1778. pmid:12401173
- 32. Nardi F, Spinsanti G, Boore JL, Carapelli A, Dallai R, Frati F. Hexapod origins: monophyletic or paraphyletic? Science. 2003;299(5614):1887–1889. pmid:12649480
- 33. Bullerwell CE, Gray MW. Evolution of the mitochondrial genome: protist connections to animals, fungi and plants. Current Opinion in Microbiology. 2004;7(5):528–534. pmid:15451509
- 34. Liu Y, Cox CJ, Wang W, Goffinet B. Mitochondrial phylogenomics of early land plants: mitigating the effects of saturation, compositional heterogeneity, and codon-usage bias. Systematic Biology. 2014;63(6):862–878. pmid:25070972
- 35. Lavrov DV, Lang BF. Poriferan mtDNA and animal phylogeny based on mitochondrial gene arrangements. Systematic Biology. 2005;54(4):651–659. pmid:16126659
- 36. Ma JY, Yang Q. Early divergence dates of demosponges based on mitogenomics and evaluated fossil calibrations. Palaeoworld. 2016;25(2):292–302.
- 37. Schuster A, Lopez JV, Becking LE, Kelly M, Pomponi SA, Wörheide G, et al. Evolution of group I introns in Porifera: new evidence for intron mobility and implications for DNA barcoding. BMC Evolutionary Biology. 2017;17(1):82. pmid:28320321
- 38. Plese B, Kenny NJ, Rossi ME, Cárdenas P, Schuster A, Taboada S, et al. Mitochondrial evolution in the Demospongiae (Porifera): Phylogeny, divergence time, and genome biology. Molecular Phylogenetics and Evolution. 2021;155:107011. pmid:33217579
- 39. Boore JL, Lavrov DV, Brown WM. Gene translocation links insects and crustaceans. Nature. 1998;392(6677):667–668. pmid:9565028
- 40. Saghai-Maroof MA, Soliman KM, Jorgensen RA, Allard RW. Ribosomal DNA spacer-length polymorphisms in barley: mendelian inheritance, chromosomal location, and population dynamics. Proceedings of the National Academy of Sciences. 1984;81(24):8014–8018. pmid:6096873
- 41. Sanger F, Nicklen S, Coulson AR. DNA sequencing with chain-terminating inhibitors. Proceedings of the National Academy of Sciences. 1977;74:5463–5467. pmid:271968
- 42. Gazave E, Lapebie P, Renard E, Vacelet J, Rocher C, Ereskovsky AV, et al. Molecular phylogeny restores the supra-generic subdivision of homoscleromorph sponges (Porifera, Homoscleromorpha). PLoS One. 2010;5(12):e14290. pmid:21179486
- 43. Staden R. The Staden sequence analysis package. Molecular Biotechnology. 1996;5(3):233–241. pmid:8837029
- 44. Ewing B, Green P. Base-calling of automated sequencer traces using phred. II. Error probabilities. Genome Research. 1998;8(3):186–194. pmid:9521922
- 45. Ewing B, Hillier L, Wendl MC, Green P. Base-calling of automated sequencer traces using phred. I. Accuracy assessment. Genome Research. 1998;8(3):175–185. pmid:9521921
- 46. Simpson JT, Wong K, Jackman SD, Schein JE, Jones SJ, Birol I. ABySS: a parallel assembler for short read sequence data. Genome Research. 2009;19(6):1117–1123. pmid:19251739
- 47.
Chevreux B, Wetter T, Suhai S. Genome sequence assembly using trace signals and additional sequence information. In: German Conference on Bioinformatics; 1999.
- 48. Huang X, Wang J, Aluru S, Yang SP, Hillier L. PCAP: a whole-genome assembly program. Genome Research. 2003;13(9):2164–2170. pmid:12952883
- 49. Bankevich A, Nurk S, Antipov D, Gurevich A, Dvorkin M, Kulikov A, et al. SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing. Journal of Computational Biology. 2012;19(5):455–477. pmid:22506599
- 50. Pearson WR. Using the FASTA program to search protein and DNA sequence databases. Methods in Molecular Biology. 1994;25:365–389. pmid:8004177
- 51. Benson DA, Karsch-Mizrachi I, Lipman DJ, Ostell J, Wheeler DL. GenBank. Nucleic Acids Research. 2008;36(Database issue):D25–30. pmid:18073190
- 52. Lowe TM, Eddy SR. tRNAscan-SE: a program for improved detection of transfer RNA genes in genomic sequence. Nucleic Acids Research. 1997;25(5):955–964. pmid:9023104
- 53. Lang BF, Laforest MJ, Burger G. Mitochondrial introns: a critical view. Trends in Genetics. 2007;23(3):119–125. pmid:17280737
- 54. Katoh K, Standley DM. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Molecular Biology and Evolution. 2013;30(4):772–780. pmid:23329690
- 55. Talavera G, Castresana J. Improvement of phylogenies after removing divergent and ambiguously aligned blocks from protein sequence alignments. Systematic Biology. 2007;56(4):564–577. pmid:17654362
- 56. Li W, Godzik A. Cd-hit: a fast program for clustering and comparing large sets of protein or nucleotide sequences. Bioinformatics. 2006;22(13):1658–1659. pmid:16731699
- 57. Lartillot N, Rodrigue N, Stubbs D, Richer J. PhyloBayes MPI: phylogenetic reconstruction with infinite mixtures of profiles in a parallel environment. Systematic Biology. 2013;62(4):611–615. pmid:23564032
- 58. Goloboff PA, Farris JS, Nixon KC. TNT, a free program for phylogenetic analysis. Cladistics. 2008;24:774–786.
- 59. Kozlov AM, Darriba D, Flouri T, Morel B, Stamatakis A. RAxML-NG: a fast, scalable and user-friendly tool for maximum likelihood phylogenetic inference. Bioinformatics. 2019;35(21):4453–4455. pmid:31070718
- 60. Lartillot N, Lepage T, Blanquart S. PhyloBayes 3: a Bayesian software package for phylogenetic reconstruction and molecular dating. Bioinformatics. 2009;25(17):2286–2288. pmid:19535536
- 61. Thorne JL, Kishino H, Painter IS. Estimating the rate of evolution of the rate of molecular evolution. Molecular Biology and Evolution. 1998;15(12):1647–1657. pmid:9866200
- 62. Botting JP, Cárdenas P, Peel JS. A crown-group demosponge from the early Cambrian Sirius Passet Biota, North Greenland. Palaeontology. 2015;58(1):35–43.
- 63.
Cohen AS. Paleolimnology: the history and evolution of lake systems. Oxford: New York: Oxford University Press; 2003.
- 64. Schindler T, Wuttke M, Poschmann M. Oldest record of freshwater sponges (Porifera: Spongillina)—spiculite finds in the Permo-Carboniferous of Europe. Paläontologische Zeitschrift. 2008;82(4):373–384.
- 65. Veynberg E. Fossil sponge fauna in Lake Baikal region. Progress in Molecular and Subcellular Biology. 2009;47:185–205. pmid:19198778
- 66. Lavrov DV, Lang BF. Transfer RNA gene recruitment in mitochondrial DNA. Trends in Genetics. 2005;21(3):129–133. pmid:15734570
- 67. Burger G, Yan Y, Javadi P, Lang BF. Group I-intron trans-splicing and mRNA editing in the mitochondria of placozoan animals. Trends in Genetics. 2009;25(9):381–386. pmid:19716620
- 68. Gazave E, Carteron S, Chenuil A, Richelle-Maurer E, Boury-Esnault N, Borchiellini C. Polyphyly of the genus Axinella and of the family Axinellidae (Porifera: Demospongiaep). Molecular Phylogenetics and Evolution. 2010;57(1):35–47. pmid:20541021
- 69. Philippe H, Brinkmann H, Lavrov DV, Littlewood DTJ, Manuel M, Wörheide G, et al. Resolving difficult phylogenetic questions: why more sequences are not enough. PLoS Biology. 2011;9(3):e1000602. pmid:21423652
- 70. Redmond NE, Morrow CC, Thacker RW, Diaz MC, Boury-Esnault N, Cárdenas P, et al. Phylogeny and systematics of Demospongiae in light of new small-subunit ribosomal DNA (18S) sequences. Integrative and Comparative Biology. 2013;53(3):388–415. pmid:23793549
- 71. Schuster A, Pisera A, Kelly M, Bell LJ, Pomponi SA, Wörheide G, et al. New species and a molecular dating analysis of Vetulina Schmidt, 1879 (Porifera: Demospongiae: Sphaerocladina) reveal an ancient relict fauna with Tethys origin. Zoological Journal of the Linnean Society. 2018;184(3):585–604.
- 72. Morrow CC, Redmond NE, Picton BE, Thacker RW, Collins AG, Maggs CA, et al. Molecular phylogenies support homoplasy of multiple morphological characters used in the taxonomy of Heteroscleromorpha (Porifera: Demospongiae). Integrative and Comparative Biology. 2013;53(3):428–446. pmid:23753661
- 73. Morrow CC, Picton BE, Erpenbeck D, Boury-Esnault N, Maggs CA, Allcock AL. Congruence between nuclear and mitochondrial genes in Demospongiae: a new hypothesis for relationships within the G4 clade (Porifera: Demospongiae). Molecular Phylogenetics and Evolution. 2012;62(1):174–190. pmid:22001855
- 74. Pankey MS, Plachetzki DC, Macartney KJ, Gastaldi M, Slattery M, Gochfeld DJ, et al. Cophylogeny and convergence shape holobiont evolution in sponge–microbe symbioses. Nature Ecology and Evolution. 2022;6(6):750–762.
- 75. Alvarez B, van Soest RWM, Rützler K. Svenzea, a new genus of Dictyonellidae (Porifera: Demospongiae) from tropical reef environments, with description of two new species. Contributions to Zoology. 2002;71(4):171–176.
- 76. Lim SC, Wiklund H, Glover AG, Dahlgren TG, Tan KS. A new genus and species of abyssal sponge commonly encrusting polymetallic nodules in the Clarion-Clipperton Zone, East Pacific Ocean. Systematics and Biodiversity. 2017;15(6):507–519.
- 77.
de Voogd NJ, Alvarez B, Boury-Esnault N, Carballo JL, Cárdenas P, Díaz MC, et al. World Porifera Database. Accessed at https://www.marinespecies.org/porifera on 2022-01-24.
- 78.
Felsenstein J. Inferring phylogenies. Sunderland, Mass: Sinauer Associates; 2004.
- 79. Erpenbeck D, McCormack GP, Breeuwer JAJ, van Soest RWM. Order level differences in the structure of partial LSU across demosponges (Porifera): new insights into an old taxon. Molecular Phylogenetics and Evolution. 2004;32(1):388–395. pmid:15186823
- 80. Voigt O, Erpenbeck D, Wörheide G. Molecular evolution of rDNA in early diverging Metazoa: First comparative analysis and phylogenetic application of complete SSU rRNA secondary structures in Porifera. BMC Evolutionary Biology. 2008;8(1). pmid:18304338
- 81. Aguilar-Camacho JM, Doonan L, McCormack GP. Evolution of the main skeleton-forming genes in sponges (Phylum Porifera) with special focus on the marine Haplosclerida (class Demospongiae). Molecular Phylogenetics and Evolution. 2019;131:245–253. pmid:30502904
- 82. Schuster A, Vargas S, Knapp IS, Pomponi SA, Toonen RJ, Erpenbeck D, et al. Divergence times in demosponges (Porifera): first insights from new mitogenomes and the inclusion of fossils in a birth-death clock model. BMC Evolutionary Biology. 2018;18(1):114. pmid:30021516
- 83. Chang S, Feng Q, Clausen S, Zhang L. Sponge spicules from the lower Cambrian in the Yanjiahe Formation, South China: The earliest biomineralizing sponge record. Palaeogeography, Palaeoclimatology, Palaeoecology. 2017;474:36–44.
- 84. Antcliffe JB, Callow RH, Brasier MD. Giving the early fossil record of sponges a squeeze. Biological Reviews of the Cambridge Philosophical Society. 2014;89(4):972–1004. pmid:24779547
- 85. Love G, Grosjean E, Stalvies C, Fike D, Grotzinger J, Bradley A, et al. Fossil steroids record the appearance of Demospongiae during the Cryogenian period. Nature. 2009;457(7230):718–721. pmid:19194449
- 86. Zumberge JA, Love GD, Cárdenas P, Sperling EA, Gunasekera S, Rohrssen M, et al. Demosponge steroid biomarker 26-methylstigmastane provides evidence for Neoproterozoic animals. Nat Ecol Evol. 2018;2(11):1709–1714. pmid:30323207
- 87. Nettersheim BJ, Brocks JJ, Schwelm A, Hope JM, Not F, Lomas M, et al. Putative sponge biomarkers in unicellular Rhizaria question an early rise of animals. Nature Ecology and Evolution. 2019;3(4):577–581. pmid:30833757
- 88. Pronzato R, Pisera A, Manconi R. Fossil freshwater sponges: taxonomy, geographic distribution, and critical review. Acta Palaeontologica Polonica. 2017;62:467–495.
- 89. Vacelet J, Perez T. Two new genera and species of sponges (Porifera, Demospongiae) without skeleton from a Mediterranean cave. Zoosystema. 1998;20(1):5–22.
- 90.
Vacelet J, Perez T, Hooper JNA. Demospongiae Incertae sedis: Myceliospongia Vacelet & Perez, 1998. In: Systema Porifera. A guide to the classification of sponges. Kluwer Academic / Plenum Publishers; 2002. p. 1099–1101.
- 91. Erpenbeck D, Sutcliffe P, de C Cook S, Dietzel A, Maldonado M, van Soest RWM, et al. Horny sponges and their affairs: On the phylogenetic relationships of keratose sponges. Molecular Phylogenetics and Evolution. 2012;63(3):809–816. pmid:22406528
- 92. Ruiz C, Muricy G, Lage A, Domingos C, Chenesseau S, Pérez T. Descriptions of new sponge species and genus, including aspiculate Plakinidae, overturn the Homoscleromorpha classification. Zoological Journal of the Linnean Society. 2017;179(4):707–724.
- 93.
Avise JC. Molecular Markers, Natural History, and Evolution. Sunderland, MA: Sinauer Associates; 2004.
- 94. Lavrov DV, Pett W. Animal mitochondrial DNA as we do not know it: mt-genome organization and evolution in nonbilaterian lineages. Genome Biology and Evolution. 2016;8(9):2896–2913. pmid:27557826
- 95. Haen KM, Pett W, Lavrov DV. Eight new mtDNA sequences of glass sponges reveal an extensive usage of +1 frameshifting in mitochondrial translation. Gene. 2014;535(2):336–344. pmid:24177232
- 96. Philippe H, Delsuc F, Brinkmann H, Lartillot N. Phylogenomics. Annual Review of Ecology, Evolution, and Systematics. 2005;36(1):541–562.
- 97. Maddison WP. Gene trees in species trees. Systematic Biology. 1997;46:523–536.
- 98. Jeffroy O, Brinkmann H, Delsuc F, Philippe H. Phylogenomics: the beginning of incongruence? Trends in Genetics. 2006;22(4):225–231. pmid:16490279
- 99. Toews DP, Brelsford A. The biogeography of mitochondrial and nuclear discordance in animals. Molecular Ecology. 2012;21(16):3907–3930. pmid:22738314
- 100. Funk DJ, Omland KE. Species-level paraphyly and polyphyly: frequency, causes, and consequences, with insights from animal mitochondrial DNA. Annual Review of Ecology, Evolution, and Systematics. 2003;34(1):397–423.
- 101. Wang X, Lavrov DV. Seventeen new complete mtDNA sequences reveal extensive mitochondrial genome evolution within the Demospongiae. PLoS ONE. 2008;3(7):e2723. pmid:18628961
- 102. Haen KM, Lang BF, Pomponi SA, Lavrov DV. Glass sponges and bilaterian animals share derived mitochondrial genomic features: a common ancestry or parallel evolution? Molecular Biology and Evolution. 2007;24(7):1518–1527. pmid:17434903
- 103. Lavrov DV, Pett W, Voigt O, Wörheide G, Forget L, Lang BF, et al. Mitochondrial DNA of Clathrina clathrus (Calcarea, Calcinea): six linear chromosomes, fragmented rRNAs, tRNA editing, and a novel genetic code. Molecular Biology and Evolution. 2012;30(4):865–880. pmid:23223758
- 104. Lavrov D, Adamski M, Chevaldonné P, Adamska M. Extensive mitochondrial mRNA editing and unusual mitochondrial genome organization in calcaronean sponges. Current Biology. 2016;26(1):86–92. pmid:26725199
- 105. Felsenstein J. Cases in which parsimony or compatibility methods will be positively misleading. Systematic Biology. 1978;27(4):401–410.
- 106. Lartillot N, Brinkmann H, Philippe H. Suppression of long-branch attraction artefacts in the animal phylogeny using a site-heterogeneous model. BMC Evolutionary Biology. 2007;7(Suppl 1):S4. pmid:17288577
- 107.
Wörheide G, Dohrmann M, Erpenbeck D, Larroux C, Maldonado M, Voigt O, et al. Deep phylogeny and evolution of sponges (Phylum Porifera). In: Advances in Sponge Science: Phylogeny, Systematics, Ecology. Elsevier; 2012. p. 1–78.
- 108. Hinchliff CE, Smith SA, Allman JF, Burleigh JG, Chaudhary R, Coghill LM, et al. Synthesis of phylogeny and taxonomy into a comprehensive tree of life. Proceedings of the National Academy of Sciences. 2015;112(41):12764–12769. pmid:26385966
- 109. Kumar S, Stecher G, Suleski M, Hedges SB. TimeTree: a resource for timelines, timetrees, and divergence times. Molecular Biology and Evolution. 2017;34(7):1812–1819. pmid:28387841