The two major extant groups of siliceous sponges, Demospongiae and Hexactinellida, are generally regarded as sister groups forming the clade Silicea, although the nature of their last common ancestor is uncertain. The fossil record contains a diverse range of basal demosponges that appear to have evolved from hexactine-bearing reticulosan ancestors, although a compelling morphological intermediate has not previously been discovered. Here we describe a new species of fossil sponge, Conciliospongia anjiensis gen. et sp. nov., from the Late Ordovician (~444 Ma) Anji Biota of South China. This species has a reticulate, tufted skeleton of minute monaxon spicules, characteristic of the fossil demosponge family Hazeliidae and modern heteroscleromorphs, with hexactine spicules and a globose body form inherited from reticulosan ancestors. This transitional morphology had previously been hypothesized in palaeontological studies. This morphological intermediate between two extant classes further confirms siliceous sponge monophyly and demosponge–hexactinellid spicule homology, and supports the primitive, stem-silicean interpretation of simpler-structured fossil reticulosans.
Demosponges are the most diverse and abundant group of extant sponges1, 2, and have a long and abundant fossil record3, 4. This record, however, consists primarily of those groups with rigid, lithistid skeletons, or hypercalcified groups4,5,6. The skeleton of other demosponges is either primarily organic1, sometimes with some silica and/or aragonite biomineralization associated with chitinous fibres7, or consists of small siliceous spicules1 that are difficult to preserve articulated. These spicules can be relatively large in various groups of tetractine-bearing demosponges, but this appears to be a derived condition8; in the majority of Heteroscleromorpha2, such as the haplosclerids9, the primary spicule type is minute (sub-millimetre) monaxons, with even smaller microscleres.
The fossil record of non-lithistid, non-hypercalcified demosponges is relatively sparse3, 4. Most of the taxa from Cambrian rocks that have been traditionally described as protomonaxonid demosponges3, 10 are now recognized as belonging to a much earlier-branching group of sponges11, 12. Keratosan sponges have been described only occasionally, with the best-known examples being the family Vauxiidae, primarily from the Cambrian Burgess Shale-type faunas12,13,14,15. A few later Palaeozoic examples of keratosans are also known16,17,18, together with some agglutinating taxa that probably also fall into the Keratosa or Verongida16, 19. Spiculate demosponges have been much harder to identify and interpret accurately, and most early examples are included in the Hazeliidae13, 14. This family is characterized by minute monaxon spicules comprising a normally reticulating array of small overlapping tufts, and have been interpreted as basal crown-group demosponges12. In addition, an early Cambrian disarticulated heteroscleromorph has been described, showing the origin of their typical microscleres within the clade20; microscleres appear to have been absent from hazeliids.
In molecular phylogenetics, there is now a general consensus on the sister-group arrangement of demosponges and hexactinellids, even when sponges have been found to be paraphyletic overall21,22,23, including some acceptance of the likely homology of siliceous spicules between the two classes24. Other topologies, however, are still being occasionally recovered25. The explicit question of homology of spicules between demosponges and hexactinellids is normally avoided, although significant differences in secretion mechanism and structure have been noted26, 27. Molecular analysis8 left open the question of whether demosponge skeletons were primitively spicular or organic, but palaeontological revision of the hazeliids and vauxiids12 implies spiculate ancestors, with spicules also present in basal vauxiids; the presence of silica in vauxiids has since been confirmed15. Furthermore, a study of spicule structure in the Ordovician hexactine-bearing reticulosan Cyathophycus 28 demonstrated hexagonal axial canal symmetry diagnostic of demosponges. This supports a speculative earlier suggestion29 that demosponges may have arisen from a sponge similar to Cyathophycus through loss of the external hexactine layer and exaggeration of the inner layer of fine monaxons. The new Ordovician sponge described herein (Figs 1 and 2) from the recently-discovered Anji Biota30 exhibits a previously unknown combination of characters that was predicted by that model.
Silicea Gray, 1867 (ref. 31)
Class, order and family uncertain
The phylogenetic position of the new sponge can be constrained to the later part of the stem group of Demospongiae. This falls within Silicea, but no other less-inclusive clade, and therefore no further higher taxonomy is currently available.
Genus Conciliospongia nov
From Latin conciliare, to unite or bring together, referring to the demonstration of skeletal homology between reticulosans and demosponges.
Globose, moderately thin-walled sponge with reticulate primary skeletal wall composed of small tufts of fine monaxons, becoming perpendicular to wall in outer part; also with isolated, robust, regular hexactines.
Conciliospongia anjiensis sp. nov.
For Anji County, Zhejiang Province, China; the type locality.
Diagnosis. As for genus.
NIGP 165186, from the latest Ordovician Metabolograptus persculptus Biozone of the Wenchang Formation, locality Zhuwukou, Anji County, Zhejiang, China; partial but well-preserved specimen in lateral view.
NIGP 165187, from the type locality; complete, vertically compressed specimen.
Bowl-shaped to globose sponge, 35 mm diameter in type material, with reduced osculum up to 20 mm in diameter. Height estimated at around 25–30 mm. Both specimens pyritized and partly weathered to black oxide minerals that retain clear impressions of spicules.
Body wall coarsely reticulate (Fig. 2b, c), with majority of gaps 0.8–1.2 mm across; intervening spicule tracts variable in width, but normally in the range 0.5–1.0 mm. In lateral view, outer region columnar, showing wall thickness of 1–2 mm. Reticulation broader and more open in upper part of holotype (Fig. 2c), and less clearly visible in paratype (due largely to different orientation of compression, but perhaps also to more massive pyritization).
Monaxon spicule tracts composed of irregularly oriented tufts of spicules up to 0.2 mm wide (Fig. 2d, g, h). Tufts become more parallel close to compressed lateral margin, especially in columnar region, implying that tufts are not wall-parallel, but irregularly oriented in three dimensions; this is also visible internally (Fig. 2g), with variable orientation of spicules inclined to plane of body wall. Thickness of marginal, columnar-appearing zone (Fig. 2b) shows that tufts are superimposed vertically as well as laterally overlapping, such that several tufts occur through the wall thickness. At surface, tufts project as bundles of outwardly-directed spicules (Fig. 2d). Regions between skeletal tracts presumed to have been composed dominantly of soft tissues with a central choanocyte chamber, but are partly or entirely crossed by occasional long, fine monaxon spicules up to 0.6 mm long (Fig. 2b), oriented mainly tangentially (but sometimes somewhat inclined) to the body wall. At lateral margins, these occasional cross-spicules are visible through virtually full thickness of wall.
Individual monaxons (Fig. 2g) densely packed and very fine, and up to 0.3 mm long; axial canals often visible as internal pyrite moulds, but not characterizable. Hexactines (Fig. 2e, f) present through body wall, particularly in outer part, as rare isolated spicules. More frequent in paratype, especially in upper area close to osculum (but not at the precise margin, or acting as marginalia). No regular orientation discernible, and spicules are widely separated and do not appear to be aligned with each other. Hexactines are robust, and mostly short-rayed, with majority having ray length up to 0.4 mm and basal diameter 0.05–0.10 mm; rays are straight and evenly tapered. Slightly larger spicules (Fig. 2f), with similar diameter but ray length up to 0.7 mm, are present in upper region of paratype.
Oscular margin not distinct in spiculation, but with weakly cuspate margin (Fig. 2h); no distinct marginalia visible, but some body wall monaxon tufts project inwards from margin. Basal region not certainly seen, but may be visible in holotype; no basalia observed. No microscleres visible.
The skeletal architecture of this monaxon-based sponge is effectively identical to that of a Cambrian hazeliid demosponge. The three-dimensional wall is more complex than that seen in Crumillospongia 13, but equivalent to that of more derived hazeliids such as Hazelia lobata 14 or H. delicatula 32, as reconstructed by Botting et al.12. The body form, however, is much more primitive, closely resembling the reticulate, globose body of Crumillospongia 13. In most Hazelia species the body form has evolved to become branching, lobate or palmate, but this is only very rarely seen among reticulosans10. The related Vauxiidae show similar reticulate wall architecture, but have strongly developed organic skeletons, reduced or absent monaxon spicules, and usually have an erect, branching body form12, 14.
The unique feature of Conciliospongia anjiensis gen. et sp. nov. is the presence of hexactines in combination with the monaxon-based skeleton. The hexactine morphology is typical of a wide range of reticulosan sponges, although the thinnest-walled reticulosans tend to have more slender-rayed spicules. Similar moderately robust hexactines to those in the new sponge form the dominant skeletal framework in the globose Cyathophycus loydelli Botting33, an advanced reticulosan with a regular spicule grid and an inner skeletal layer of monaxon spicules29, 33 and with rounded parietal gaps28. A similar secondary layer of obscure spicules in other species of Cyathophycus has been previously discussed34, also with rounded gaps through the secondary wall. The spicule microstructure of C. loydelli was considered to be an indicator of a close relationship to the demosponge stem lineage28, which would then have progressed through loss of hexactines to leave the architecturally flexible inner skeletal layer, which is composed of tracts or tufts of fine monaxons. Until now, however, no intermediates within this sequence of hexactine loss have been described, and the hypothesis has remained speculative.
The new species precisely fills this gap in the theoretical sequence. The globose body form is identical to that of C. loydelli and Crumillospongia, and shows none of the architectural complexity of the tubular, frequently branching vauxiids, or the more diverse morphologies seen in Hazelia 14. The tufted arrangement of fine monaxons is typical of the more advanced hazeliids, but easily derived from the tufted, semi-tractose structure seen in Crumillospongia, which is the last genus in the sequence to show a simple, globose body form. This in turn forms the basis for the skeletal architecture of extant heteroscleromorphs such as the Haplosclerida10, 12. The presence of low numbers of hexactines, which are absent from all fossil and recent demosponges, illustrates incomplete loss of the primary skeletal layer of its reticulosan ancestors. Their positions embedded within the perfectly articulated skeletal wall prove that the spicules have not been fortuitously superimposed, and their consistent morphology, intact condition and low density preclude deliberate incorporation of detrital spicules by an agglutinating sponge. The new species therefore represents a perfect evolutionary intermediate between a Cyathophycus-like reticulosan and a globose proto-demosponge like Crumillospongia (in which hexactines have been entirely lost). Conciliospongia therefore illustrates the stem group of demosponges, at a stage prior to the complete loss of the primary symplesiomorphies (hexactines, globose body form, thin wall) and also before the origination of shared demosponge characters (morphological plasticity, organic framework skeleton).
The fossil record of sponges is notoriously incomplete, and important discoveries can emerge from surprising localities. The Anji Biota30 is relatively late (Late Ordovician) in comparison with previously-described early Cambrian crown-group demosponges12, 20, indicating a significant period of evolution since the separation of this sponge from the lineage leading to the base of the demosponge crown group (Fig. 3). This interval allowed the evolution of the relatively derived, three-dimensional wall structure in comparison with that of the thinner wall of Cyathophycus or Crumillospongia. Nonetheless, the body form remains as conservative as in reticulosans, suggesting that the biological or genetic innovations that allowed demosponge morphological plasticity to develop within the Hazeliidae had not yet occurred at this point in the lineage.
This discovery is of critical importance in demonstrating the evolutionary origin of demosponges, but as a result, has additional implications for interpretation of other reticulosan sponge fossils. The phylogenetic scenario that predicted this intermediate also requires stem-group status for many other sponge groups28. In particular, reticulosans that fall below Cyathophycus in structural complexity, such as Protospongia Salter35, Heminectere Botting33 and Asthenospongia Rigby, King and Gunther36 probably occupied phylogenetic positions below crown-group Silicea; this general evolutionary progression of wall complexity within sponges is widely recognized14, with the more simple globose, thin-walled architecture also shared with basal members of stem-group calcareans37. Confirmation of basal demosponge evolution at least requires a critical reassessment of basal sponge relationships based on these fossils, including palaeontological evidence that even calcareous and siliceous spicules are homologous, in the form of biminerallic spicules among both early calcarean-like37, 38 and protomonaxonid39 sponges. This scenario conflicts with traditional understanding based on the differences in secretion mechanisms of calcareous and siliceous spicules27, 40 and the lack of a model for transforming the spicule composition. The current discovery regarding demosponge origins and the homology of silicean spicules emphasizes the need for further detailed palaeontological studies, in addition to molecular and embryological work, to clarify sponge evolution.
Material and Methods
The studied specimens were recovered as part of a diverse sponge fauna from the Upper Ordovician Wenchang Formation in Zhejiang Province, China30. The Wenchang Formation consists mostly of sandstone with an approximately ten-metre-thick mudstone interval in the middle-upper part, which has yielded articulated sponges at several sites. The new species was discovered at locality Zhuwukou (GPS coordinates 30°30.507′N, 119°22.712′E), a roadside section near the Fushi Reservoir, Anji County. The material can be dated as late Metabolograptus persculptus Biozone (Hirnantian, Late Ordovician, ~444 Ma) on the basis of graptolite biostratigraphy30. Material was photographed using a Zeiss Discovery V.20 stereomicroscope and Axiocam 506 camera system (photomicrographs), or with Nikon D80 with extension tubes and 105 mm macro lens (wider-angle photographs). Specimens are deposited in the Nanjing Institute of Geology and Palaeontology (NIGP).
Hooper, J. N. A. & Van Soest, R. W. M. Class Demospongiae Sollas, 1885. In: Systema Porifera: A Guide to the Classification of Sponges (eds Hooper J. N. A. & Van Soest, R. W. M.), pp. 15–51 (Kluwer Academic/Plenum Publishers, New York, 2002).
Morrow, C. & Cárdenas, P. Proposal for a revised classification of the Demospongiae (Porifera). Front. Zool. 12, 1–27 (2015).
Reitner, J. & Wörheide, G. Non-Lithistid fossil Demospongiae — Origins of their palaeobiodiversity and highlights in history of preservation. In: Systema Porifera: A Guide to the Classification of Sponges (eds Hooper J. N. A. & Van Soest, R. W. M.), pp. 52–68 (Kluwer Academic/Plenum Publishers, New York, 2002).
Pisera, A. Palaeontology of sponges—a review. Can. J. Zool. 84, 242–261 (2006).
Carrera, M. G. & Rigby, J. K. Biogeography of Ordovician sponges. J. Paleontol. 73, 26–37 (1999).
Muir, L. A., Botting, J. P., Carrera, M. & Beresi, M. Chapter 8: Cambrian, Ordovician and Silurian non-stromatoporoid Porifera. In: Early Palaeozoic Palaeobiogeography and Palaeogeography (eds Harper, D. A. T. & Servais, T.), pp. 81–95 (Geological Society of London, Memoir 38, 2013).
Ehrlich, H. et al. Insights into chemistry of biological materials: newly-discovered silica-aragonite-chitin biocomposites in demosponges. Chem. Mater. 22, 1462–1471 (2010).
Borchiellini, C. et al. Molecular phylogeny of Demospongiae: implications for classification and scenarios of character evolution. Mol. Phylogenet. Evol. 32, 823–837 (2004).
Van Soest R. W. M. & Hooper J. N. A. Order Haplosclerida Topsent, 1928. In: Systema Porifera: A Guide to the Classification of Sponges (eds Hooper J. N. A. & Van Soest, R. W. M.), pp. 831–832 (Kluwer Academic/Plenum Publishers, New York, 2002).
Finks, R. M., Reid, R. E. H. & Rigby, J. K. Treatise on Invertebrate Paleontology, Part E: Porifera (revised), volume 3: Porifera (Demospongea, Hexactinellida, Heteractinida, Calcarea), pp. i–xxx, pp. 1–872 (Geological Society of America and University of Kansas, Boulder and Lawrence, 2004).
Reitner, J. & Mehl, D. Early Palaeozoic diversification of sponges: new data and evidences. Geologisch-Paläeontologische Mitteilungen Innsbruck 20, 335–347 (1995).
Botting, J. P., Muir, L. A. & Lin, J.-P. Relationships of the Cambrian Protomonaxonida (Porifera). Palaeontologia Electronica 16(2, 9A), 1–23 (palaeo-electronica.org/content/2013/421-protomonaxonid-relationships, 2013).
Rigby, J. K. Sponges of the Burgess Shale (Middle Cambrian), British Columbia. Palaeontographica Canadiana Monograph 2, 1–105 (1986).
Rigby, J. K. & Collins, D. Sponges of the Middle Cambrian Burgess and Stephen Shale Formations, British Columbia. Royal Ontario Museum Contributions in Science 1, 1–164 (2004).
Yang. X., Zhao, Y., Babcock, L. E. & Peng, J. Siliceous spicules in a vauxiid sponge (Demospongia) from the Kaili Biota (Cambrian Stage 5), Guizhou, South China. Sci. Rep. 7, 1–7 (2017).
Botting, J. P. Exceptionally-preserved Middle Ordovician sponges from the Llandegley Rocks Lagerstätte, central Wales. Palaeontology 48, 577–617 (2005).
Botting, J. P. Algae, receptaculitids and sponges. In: Silurian fossils of the Pentland Hills, Scotland (eds Clarkson, E. N. K., Harper, D. A. T., Taylor, C. M. & Anderson, L. I.), pp. 36–49 (Palaeontological Association Field Guides to Fossils 11, 2007).
Luo, C. & Reitner, J. First report of fossil “keratose” demosponges in Phanerozoic carbonates: preservation and 3-D reconstruction. Naturwissenschaften 101, 467–477 (2014).
Muir, L. A. & Botting, J. P. The distribution and diversity of Porifera in the Ordovician Builth Inlier (Wales, UK). Palaeoworld 24, 176–190 (2015).
Botting, J. P., Cárdenas, P. & Peel, J. S. A crown-group demosponge from the early Cambrian Sirius Passet Biota, North Greenland. Palaeontology 58, 35–43 (2015).
Sperling, E. A., Peterson, K. J. & Pisani, D. Phylogenetic-signal dissection of nuclear housekeeping genes supports the paraphyly of sponges and the monophyly of Eumetazoa. Mol. Biol. Evol. 26, 2261–2274 (2009).
Wörheide, G. et al. Chapter one - deep phylogeny and evolution of sponges (Phylum Porifera). Adv. Mar. Biol 61, 1–78 (2012).
Gazave, E. et al. No longer Demospongiae: Homoscleromorpha formal nomination as a fourth class of Porifera. Hydrobiologia 687, 3–10 (2012).
Sperling, E. A., Robinson, J. M., Pisani, D. & Peterson, K. J. Where’s the glass? Biomarkers, molecular clocks, and microRNAs suggest a 200-Myr missing Precambrian fossil record of siliceous sponge spicules. Geobiology 8, 24–36 (2010).
Riesgo, A. et al. Inferring the ancestral sexuality and reproductive condition in sponges (Porifera). Zool. Scr. 43, 101–117 (2014).
Croce, G. et al. Structural Characterization of Siliceous Spicules from Marine Sponges. Biophys. J. 86, 526–534 (2004).
Uriz, M. Mineral skeletogenesis in sponges. Can. J. Zool. 84, 322–356 (2006).
Botting, J. P. & Muir, L. A. Spicule structure and affinities of the Late Ordovician hexactinellid-like sponge Cyathophycus loydelli (Llanfawr Mudstones Lagerstätte, Wales). Lethaia 46, 454–469 (2013).
Botting, J. P. Cyathophycus and the origin of demosponges. Lethaia 36, 335–344 (2003).
Botting, J. P. et al. Flourishing sponge-based ecosystems after the end-Ordovician mass extinction. Curr. Biol. 27, 556–562 (2017).
Gray, J. E. Notes on the arrangement of sponges, with the description of some new genera. Proceedings of the Zoological Society of London 1867(2), 492–588 (1867).
Walcott, C. D. Middle Cambrian Spongiae: Cambrian Geology and Paleontology IV. Smithsonian Miscellaneous Collections 67, 261–364 (1920).
Botting, J. P. An exceptional Caradoc sponge fauna from the Llanfawr Quarries, central Wales, and phylogenetic implications. J. Syst. Palaeontol. 2, 31–63 (2004).
Rigby, J. K. & Chatterton, B. D. E. New Middle Silurian hexactinellid sponge from the McKenzie Mountains, Northwest Territories, Canada. J. Paleontol. 68, 218–223 (1994).
Salter, J. W. On some new fossils from the Lingula-flags of Wales. Quarterly Journal of the Geological Society 20, 233–241 (1864).
Rigby, J. K., King, J. E. & Gunther, L. F. The new Lower Ordovician protosponge, Asthenospongia, from the Phi Kappa Formation in central Idaho. J. Paleontol. 55, 842–847 (1981).
Botting, J. P. & Butterfield, N. J. Reconstructing early sponge relationships by using the Burgess Shale fossil Eiffelia globosa, Walcott. Proc. Natl. Acad. Sci. USA 102, 1554–1559 (2005).
Bengtson, S. & Collins, D. Chancelloriids of the Cambrian Burgess Shale. Palaeontologia Electronica 18.1.61, 1–67 (2015).
Botting, J. P., Muir, L. A., Xiao, S., Li, X.-F. & Lin, J.-P. Evidence for spicule homology in calcareous and siliceous sponges: biminerallic spicules in Lenica sp. (Porifera;?Protomonaxonida) of early Cambrian age (535–520 Ma) from South China. Lethaia 45, 463–475 (2012).
Bergquist, P. Sponges, pp. 1–268 (Hutchinson, London, 1978).
Xiao, S., Hu, J., Yuan, X., Parsley, R. L. & Cao, R. Articulated sponges from the Lower Cambrian Hetang Formation in southern Anhui, South China: their age and implications for the early evolution of sponges. Palaeogeogr. Palaeoclimatol. Palaeoecol. 220, 89–117 (2005).
This work was supported by the Chinese Academy of Science (grant nos. XDB10010100, XDPB050105), the Natural Science Foundation of China (grant no. 41521061) and the Ministry of Science and Technology of China (grant no. 2013FY111000) to Z.Y.D. J.P.B.’s research is funded by a CAS President’s International Fellowship Initiative (no. 2016VEB006). The assistance on fieldwork of Xiang Fan, Xuan Ma, Junye Ma and Yanyan Song of the Nanjing Institute of Geology and Palaeontology and of Longwu Wang and Jianfang Zhang of the Zhejiang Institute of Geological Survey is gratefully acknowledged. We thank the two anonymous reviewers for their constructive and helpful reviews. This is a contribution to the IGCP 653 project.
The authors declare that they have no competing interests.
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Botting, J.P., Zhang, Y. & Muir, L.A. Discovery of missing link between demosponges and hexactinellids confirms palaeontological model of sponge evolution. Sci Rep 7, 5286 (2017). https://doi.org/10.1038/s41598-017-05604-6
This article is cited by
Nature Ecology & Evolution (2018)