Festschrift zum 60. Geburtstag von Helfried Mostler Geol. Paläont. Mitt. Innsbruck, ISSN 0378-6870, Bd. 20, S. 335-347, 1995 • EARLY PALEOZOIC DlVERSIFICATION OF SPONGES: NEW DATA AND EVIDENCES Jüachim Reitner & Dürte Mehl • • With I figure and 2 plates Abstract: Sponges are primitive metazoan organisms which occur since the late Proterozoic. The oldest sponge remains (hexac- tinellids) are from a shallow marine carbonate facies of the late Sinian Shibantan Member (China). From the Tommo- tian (Sansha, China) protospongiid and rossellimorph hexactinellids were discovered in shallow marine silicified rocks. The middle part of this section consists of silty shales with entirely preserved hexactinellid sponges: Sanshadic- tya, Hyalosinica, Solactinella, Triticispongia, Hunanospongia. The hexactinellids typically are soft bottom dwellers, which sometimes grow on and form in situ spiculites (spicule mats) . The sponges from the Chengjiang deposits of At- dabanian age, which show a rossellimorph spicule organization typical of the conservative hexactinellid groups, which inhabited the deep sea during most of the Phanerozo ic, were also typical soft bottom dwellers. All sponge taxa (Lepto- mitella, Leptomitus, Paraleptomitella etc.) described from this locality we consider to be hexactinellids and not de- mosponges, as originally classified. The sponge communities of carbonate-dominated archaeocyathid mounds are completely different from the above de- scribed sponge mounds. The sponges from the archaeocyath mounds of the Flinders Ranges (S.Australia) are characte- rized by Calcarea and demosponges ($; modern type of pharetronid Calcarea Gravestockia pharetroniensis, and many tetractinellid demosponges, Geodiida). Questionable is the origin of four-rayed demosponge spicules found in some archaeocyaths. The oldest ceractinomorph demosponges are documented from the middle Cambrian (Vauxia - an aspi- cular sponge with keratose affinities, and isolated sigmata microscleres). Sponge communities of the Early Cambrian were highly developed, and already then they were taxonomically strongly dependent on substrates and trophic conditions. Zusammenfassung: Schwämme sind primitive Metazoen, die seit dem späten Proterozoikum vorkommen. Die ältesten Schwammfragmen- te (Hexactinelliden) stammen aus einer flachmarinen Karbonatfazies im späten Sinium des Shibatan Members in China. Aus dem Tommotium (Sansha, China) wurden protospongiide und rossellimorphe Hexactinelliden innerhalb flachmariner, verkieselter Gesteinsformationen entdeckt. Der mittlere Abschnitt dieses Profils besteht aus siltigem Tongestein mit körperlich erhaltenen hexactinelliden Schwämmen: Sanshadictya, Hyalosinica, Solactinella, Triticis- pongia, Hunanospongia. Die Hexactinelliden sind typische Weichbodenbewohner, die manchmal in-situ-Spiculite (Spicula-Matten) bilden und auf diesen wachsen. Die Schwämme der Chengjiang Lagerstätte (Atdabatium), die eine • rossellimorphe Sklerenanordnung, typisch für konservative Hexactinellidengruppen des tieferen Wassers während des Phanerozoikums, zeigen, waren ebenfalls charakteristische Weichbodenbewohner. Alle Spoilgien-Taxa, die von dieser Lokalität beschrieben sind (Leptomitella, Leptomitus, Paraleptomitella, ect.), werden von uns als Hexactinellida ange- sehen und nicht als Demospongiae, wie sie sie ursprünglich klassifiziert wurden. Die Schwamm-Gemeinschaften der karbonatreichen Acrchaeocyathidenmounds unterscheiden sich von den oben be- schriebenen Spongien mounds signifikant. Die Schwämme aus den Archaeocyathenmounds der Flinders Ranges (Süd- australien) sind charakteristischerweise Calcarea und Demospongiae (pharetronide Calcarea von "moderner" Organi- sation wie Gravestockia pharetroniensis und viele tetractinellide Demospongiae der Geodiida). Umstritten bleibt der Ursprung der vierstrahligen demospongiiden Spicula, die in einigen Archaeocyathen gefunden wurden. Die bisher äl- testen ceractinomorphen Demospongien sind aus dem mittleren Kambrium nachgewiesen (Vauxia - ein aspikulärer Schwamm mit Affinität zu den Keratosen, und isolierte Sigmata\ Mikroskleren). Spongien-Gemeinschaften des frühen Kambriums waren hochentwickelt und in ihrer taxonomischen Zusammenset- zung schon damals sehr vom Substrat und Nahrungsangebot abhängig. • 335 · Introduction Sponges are a monophyletic metazoan group characterized by choanoflagellate cell types (choanocytes). Based on studies of MEHL & REISWIG (1991), REITNER (1992), REITNER & MEHL (in press) and MÜLLER et al. (1994) the first sponge-metazoans originated in the Pro- terozoic from a choanoflagellate ancestor. Probably the original sponge was an aggregate of choanoflagellates closely related to various microbial communities. It is evident that all known sponge taxa bear various amounts of often species-specific bacterial associations within their mesohyle. Very important data give the analyses of the metazoan ß-galactose- binding lectins (S-type lectins) in sponges which was hitherto analysed only from verte- brates and the nematode Caenorhabditis ele- gans (PFEIFER et al. 1993, MÜLLER et al., 1994). The development of this sponge lectin may have occurred be fore 800 my (HIRABAYASHI & KASAI, 1993). This biochemical hypothesis that sponges are Proterozoic metazoans is tested by the new findings of indubitable sponge spicu- les from the Shibantan Member (Dengying Formation, late Sinian of the Hubei Province, China, STEINER et al., 1993). RrGBY (talk on the 4th internat. Porifera Congress, Amsterdam 1993) has presented a nearly complete speci- men of a hexactinellid sponge from the Ediaca- ra type locality of South Australia. Remarkable are also biomarker analyses made by MOLDD- WAN et al. (1994) who detected C30 sterane which are characteristic for demosponges in 1,8 mrd.y. old black shales! The diversification of the demosponges and calcareous sponges is linked with the shallow calcareous facies often related to the archaeo- cyaths. Within archaeocyath mounds sponge spicules are very common, and beside hexac- tines, tetractines and modified monaxonic ones do occur. One of the authors (JR) has studied the spicules of the Atdabanian and Botomian Archaeocyath mounds of the South Australian Flinders Ranges intensively (REITNER, 1992; DEBRENNE & REITNER, in press). 336 Hexactinellida According to the data presently available, the Hexactinellida are the oldest metazoan group, which can be definitely attributed to an extant taxon. Spicule remains with dear relics ofaxial canals, mainly monaxons but also induding typi- cally hexactinellid triaxones, probably staurac- tins, were documented from thin sections of , limestones from the Dengying Formation in South China (STEINER et al., 1993). These spicu- les are from the Shibatan Member of Late Pro- terozoic age (stratigraphie equivalent of the Aus- tralian Ediacara Member). Well-preserved sponge fossils have been de- scribed earlier from the famous Burgess Shale (RrGBY, 1986) and from the House Range Moun- tains and Wheeler Shale, Utah (RrGBY, 1978, 1983; a.o.). A hexactinellid sponge fauna of comparably high diversity is known from Lower Cambrian (Tommotian) strata of the Sansha sec- tion, Yangtze Platfonn, South China. Six genera and species of sponges, most of which, or maybe all, are Hexactinellida, including the earliest re- presentatives of several large Paleozoic groups, were found in the Sansha section as well-pre- served natural spicule-assemblages. Sanshadictya microreticulata MEHL & R (in: STEINER et al., 1993) is the earliest sponge with regular hori- zontal-vertical rows of diactins and stauractins. This reticulate spicule organization is the diagnos- tic feature of the Dictyospongiidae HALL, 1884, known from the Ordovician till the end of the Permian. However, the monophyly of Dictyo- spongiidae is not definitelyestablished, and it is uncertain whether or not Sanshadictya really be- , longs to this group, since no Middle or Late Cambrian representatives are known so far. The species Hyalosinica archaica MEHL & RETTNER was decribed on account of isolated root tufts of long monaxial spicules, twisted in a dockwise direction. The attachment of the main sponge body by a long, flexible stalk like a glass rope is a feature of the recent amphidiscophoran group, Hyaloneametidae GRAY, which comprise a large number of species (e.g. SCHULZE, 1877). Hyalo- nema GRAY, 1832, known from Late Cretaceous Geol. Paläont. Mitt. Innsbruck, Bd. 20, 1995 • • , • • • .. • • , • • • • . . . . . • • .. , . . , . • • • • • • • , . . .. '. . . .. .. ,. • • .. ' . • • • • . . , . '. . . . . • . . · .. . \ .'-. . . : . . ' " ".' " . • , . 1 • 5 . , . • A , ' . •• • • ~ ' : . I'" ' . • • • · .' '. ~, . · . , .. " . :, . ' :: ,. ' . ', ~ " '", \ - ,' . . . -;' . -, '. , ' : I . .. . · ' • .- . . ' .... ':,' .. :.- ,' . ; "', . ...... . ' '1 ' . · .. .. • • ,. ' . . . . .. • • • . .. • · .. ," . .. . " . . .. · \ " . .':':: • .. . , ' . . ' .. : '. ' .> '. , . . " '. ' ~ , .. • .. '. - .' . • • • • • • • • • • • • . . . " - . · ' '- . · .. ,. - " • • . " . ". . ' 8 .. 2 12 -' , . . '...' . . ' '. .. , • • , '" . . ,. , . . ' . . '., " ,. " · . • · 3 . . . · . , . • • • • • • . . . .. ' . . . ' • • • • , " .' '" " , '. • ~ • • >- ., ••••• • I . • , . 1. -' • . I, . .' .. ' . ~ . '. . . .' . .' i; :" ,v .1,; . . I .. " . . ' ...... " . ' . , " ..... . , " . " .. . .. . ... . ' . . ., " . ...... ' . ", t. .. , ." . ( ' . ' .' , . " . ... ! .. ...... ~ .. , .. . , ' ... ~ .. '. " \ , ,, .. . ': . . - '" , r ,· · . ,, ' :. - , ' ,\::,.- .,. ' . ' '" -" ." , '.- .' . . , ", , ' .. . .. . ,., . : . .. ,N . ,. . . '~"\ ," , ...... '. ' . ~ ., .. ' ' 1 ' :.';'t . . 1 , ' . '.' . ' .I'.~ ' ~ , .... '. ' • ,{~ Y;' ". : ', : (. . r. . . • ·.·t ·\, " .. 't-,'li t,~": !" . • ;;;.:,: .;.C;.~. :3l+;-:.' :.... ___ _ " , ' ) , , ,.. '.- . · . · . .'. , ~ . ' . " .' . . ',' .. . . ' . . · '. '. " . " " . . .... . ,. , " .... _. . . , -, . · . ',- • • I , '. . .. . ' . ., • · . , . . .- . . .. '. .. . { .. .. .-_ .. J,. , .. , ' .. ' . . " .. . . " '. - , . . . ... .. " , · , . . . . . . . . . , """ .' ' . ' . . · ~., . • • , 9 • '. . 4 : . _,-- • • •• • • . '. . .... •• • • . - ~ . • '. ' . I . ' . . ' " , • • • · ~ '. ", , .. ,.., .. · . ' . . . .~ 7 " ; "-,I ." . .. -, , '", ---:-:-:--.... · ," " ,,," . .. .-_ --:.:. . ..;, . . . . ' . . ... . '. . ', . ... .- - ". 1';-: .... . , . . _ .. - -: " '. ":""" .., .. . , . .. " " . .... " . ' , . . . . . . . " . . ' ,-: . .... : ,." .:, ... ; . . ~ .. ' ," , "": :.', ;,~ . _' ," .. . ... ~ . : " ," ,' . , I . • . • - • I" ': ;'- - , • • , . , • . : " . ~ ",;:, ':.. .' ~ .' ... ' :.' .. ~ ... .. ' ,: .· :,'r~, , . ' '. ,. . ~ I, '/ .1 " "" r ~ ," ' 0 ", 1 . ' . ' " .. ,: • . . , _0":; . .; " . :.' . :. . ; .;: .. . _. ..... -. ',: . .. ,'v .. . .' , " , ",. ' . ', . ... .. ... J. , .. .... ... ', . - .... ,' .. ' , ... . " ','" -. -. ' . .. -. : . ' : . -. . ' . "'- . '. . . ' . . .. , . , ,. . , . - J. 'I ., ~ . :, " . ', .', . ... ~. , ' . :.. ' . , ' . ., .. , ... . - " .. ..... . .'. ." '. " " ,.".' ' .. ~ : .. . '.. ........ .. , I . '.".' ' ,~: . " " .. " , " 1, '- \ '/ . . :. ".'. ' .' . "" "" ,' .. ~ .... ~ . ' ' ., . ' . "~ ' " . . ., . . ~ . " ... ' . '.. \ .' . ,. . ... . . " . .. .. .. . . " ,. ~ .... , . ...... . .... " , • . . - I •. • '. ' ~ .. ;." , . " '. ' ·: L,·j·il • J '. ~ ,,,.~ . . ' , " .' "." '" ' •• , .f •. ~." , . , . ~ ' . . • .. • • fo4 · ;. ', ' .' .' . . i.. • .- · . " . . " . . . - '. . .. . • 10 13, • Fig. 1: Drawings of hexactinellid spicules from Georgina Basin (Australia, Middle-Cambrian). ' I: Follipinules of new hexactinellid (MEHL, 1995: holotype). 2-4: Yarious follipinules attributed to thi s species. 5- 7: Heavily in- f1ated stauractins, diactins, and hexactins which are associated with follipinules , and which may probably be attributed to the same taxon as those. 8. Kometiaster. 9, 13: special type of Cambrian stauractins f1attened in the area of the ax ial cross (axial canal broken up in 13). 10: Stauractin with its paratangential rays all bent to point in one direction. 11 : Tylodisc similar to those described as Na- baviella elegans MOSTLER & MOSLEH-Y AZDl, 1976. 12: Clavule, All scales = 100 11m. (MEHL, 1992) till today is characterized by a twist- ed root tuft of anchoring spicules, which may be more than 1 m long and is colonized by sym- GeoL Paläont, Mitt. Innsbruck, Bd. 20, 1995 biotic soft corals encrusting and thus additionally stabilizing the "glass rope", It was a strategy also of many fossil Hexactinellida, such as many Or- 337 dovician Protospongiidae (DAWSON & HrNoE, 1888, 1889) of Little Metis, Canada, the Early Devonian Retifungus rudens RrETSCHEL, 1970, from the Hunsrückschiefer, Germany, or Hyalo- stelia smithii YOUNG & YOUNG, 1877, from the Carboniferous of Ayrshire, England. These are groups which may or may not belong to the Hya- lonematidae. Triticispongia diagonata MEHL & REITNER (in: STEINER et al., 1993) most probably is aprecursor of Diagoniella RAUFF, 1894, a Middle Cambrian representative of the Early pa- leozoic Protospongiidae SALLER. Sanshapentella dapingi MEHL & ERDTMANN, 1994, exhibits a dermal skeleton very similar to the isolated he- xactins with all their paratangential rays bent in one direction, which gave rise to Hunanospongia delicata QIAN & DING, 1988. Sirnilarly triaxial spicules, stauractins, with all rays curved in one direction are found in sediments of the Georgina Basin (fig. 1: 10). No post-Cambrian Paleozoic representatives of this type of hexactins with their paratangential rays all bent in one direction, but they are found in .the Late Cretaceous (Co- niacian) Amagerkalk from Bornholm, Denmark (MEHL, 1992). Solactiniella plumata MEHL & RElTNER, characterized by diactin spicules radiat- ing from a center in all directions represents a conservative bauplan of the Porifera. A radiating spicule organization is very comrnon within the Demospongiae, e.g. the Middle Cambrian Choia ridley WALCOTT, 1920. Hexactinellids which ex- hibit a sirnilarly conservative skeletal architectu- re were called "rossellimorph", which is not the name of a monophyletic group, but merely the designation of a very persistent rnorphological trend within the Hexactinellida (MEHL, 1995). Quadrolaminiella CHEN et al. , 1990, from the Lower Cambrian Atdabatian of Chengjiang from Hunan, China, was originally attributed to the Demospongiae. However, Quadrolaminiella ex- hibits a spicular architecture with strong affinity to the Tommotian hexactinellid Solactiniella. According to our observations on unpublished material from Chengjiang, comprising also spe- cimens which most probably belong to Quadro- laminiella, all Chengjiang sponges including the species described by CHEN et al. (1990), are He- 338 xactinellida. So far, no definite triaxons have been observed within these sponges. However, their spiculation is very sirnilar to that of other Paleozoic "rossilimorphs", some of which do, others apparently do not possess triaxial spicules. Although the Hexactinellida are characterized by triaxons, basically hexactins, most of the hexac- tins of the so-called "lyssacinosans" are reduced to stauractins, e.g. the Protospongiidae (compare , MEHL, 1991) or diactins. Thus, occasionally fossil or recent hexactinellids are found with exclusively diactin "megascleres". In recent re- presentatives and in very well-preserved fossils, their hexactinellid nature can easily be recog- nized because of the typical soft tissue organiza- tion and the triaxial "rnicroscleres" (hexasters and amphidiscs). However, in fossil sponges nor- mally no "rnicroscleres" are preserved, and thus the classification of sponges with exclusively monaxial "megascleres" is often problematic. Hexactinellid "microscleres" are mainly found as isolated spicules washed out of sediments (MEHL & MOSTLER, 1995). One criterion of di- stinction is the fact that hexactinellids often have much larger and coarser spicules than Demo- sponges. The largest poriferan spicules are found in the recent amphidiscophoran hexactinellid Monorhaphis SCHULZE, 1904, which is rooted in the sediment by a single anchoring spicule, up to 3 m long and 1 cm thick. The root-tuft spicules of Hyalonema may exceed 1 m in length and several lillll in thickness. Of course some hexactinellids have small, only delicate "megascleres", but especi- ally fossil sponges characterized by coarse spicules, several cm in length, in almost every case can be proven by the presence of triaxons to be hexac- , tinellid . . In the case of Quadrolaminiella, with co- arse spiculation and strong affmity to Solactiniella the hexactinellid identity is 'very probable, even though the examination of further material would be necessary to definitely prove this hypothesis. From the Late Atdabatian (Qiongzhusi Stage) of the Shaanxi province, southem China, a high- ly diverse association of isolated ? demospongi- an, caIcarean (heteractinid), and mainly hexac- tinellid spicules has been described (ZHANG & PRATT, 1994). It contains oxypentasters and Geol. Paläont. Mitt. Innsbruck, Bd. 20, 1995 -staurasters with 2 or 3 secondary rays which are often c1ading into Tertiary branches, before they end as sirnply pointed rays. These peculiar spicules, so far known from the Cambrian only, might be precusor fonns of true oxyhexasters (compare MEHL & MosTLER, 1995). The Middle Cambrian phosphorites of the Ge- orgina Basin in South Australia are being studied by one of the authors (DM). They contain micro- fossils in extraordinary, often soft-part preservati- on. Further, a diverse association of sponge spicules has been washed out of dissolved sam- pies of these sediments. Their major part consists of hexactinellid spicules, but also the Demospon- giae and Calcarea (Heteractinida) are represented. Hexactins are found in high variability of shapes. Generally, the hexactins are rather large and com- monly measure up to several mm in total length. Average hexactins measure about 2-500 jlIl1 and show six rays of approximately equal length, which may be a little curved or inflated to various degrees (fig. I: 7). One or more rays can show va- rious degrees of reduction and thus the spicules grade into pentactins, stauractins (fig. 1: 9, 13), or other triaxial derivates. Some of the smaller tria- xons, commonly stauractins, may be inflated like pillows, and the triaxial nature of such heavily in- flated spicules is hardly recognizable. Pinules are tree-like spicules, hexactins or stauractins with their distal rays pinulate and often i nfl ated , which, according to the functional morphology and position in recent hecactinellids, are margi- nal, normally dermal spicules. Classical pinules of the Georgina assemblage are rather large, average size 500 to 1 000 jlIl1. Follipinules (MEHL, 1995) are special pinules, average size only about 400 11m, with their dis- tally extremely inflated rays decorated with lon- gitudinal ridges, ball-shaped, and almost "ab- sorbing" the other reduced rays, so the axial cross sometimes is not visible (fig. 1: 1-4). A new taxon is erected on account of a hypotheti- cal reconstruction of the Cambrian sponges carrying this special type of dermal spicules in MEHL (1995). Inflated pentactin pinules with some similarity to follipinules from the Cambri- an of Texas were interpreted as dermal spicules Geol. Paläont. Mitt. lnnsbruck, Bd. 20, 1995 by RIGBY (1975). MOSTLER & MOSLEH-YAZDI (1976) documented smooth hexactins with one or more swollen rays from Late Cambrian sedi- ments of Iran and based on these spicul~ the taxon Rigbyella ruttneri. Heavily inflated spicul- es like "pillow-stauractins" (fig. 1: 5-6) and fol- lipinules are a widespread feature within Paleo- zoic, especially Cambrian, Hexactinellida. This type of probably denlla,l armouring layers indica- tes a special strategy of Early Paleozoic hexac- tinellida, which was realized beside the fragile skeletal architecture known from e.g. the Proto- spongiidae (compare MEHL, 1995). The armou- ring strategy probably reflects the adaptation to another type of environment as that of the fragile hexactinellids, eigther as stabilizor by higher water energies or as a defence against sponge- feeding predators. Kometiasters (gr. Kametes = long-haired star) (Fig. I: 8). These are triaxons 400-800 jlIl1 in dia- metre with numerous short pointed secondary rays. One or two of these principalia are split up into a large number (ca. 30-50) of secondary rays, which are long (350-500 jlIl1), gently curv- ed, and slightly barbed at their distal ends. From the Late Ordovician ofNew South Wales a highly interesting hexactinellid spicule assemblage has been documented recently by WEBBY & TROTTER (1993). The hexasters ofthe new species Kametia crucifarmis WEBBY & TROTlER, 1993, are sirnilar to those from the Georgina Basin. If they can be regarded as true hexasterophorid spicules, these hexasters, which are called kometiasters (MEHL 1995), are the oldest ones found so far. Thus, the major hexactinellid group Hexasterophora might be traced back as far as tQ the Middle Cambrian. . Tylodiscs: Only one spiculum of this type was found within the collection (fig. 1: 11). It is r 290 jlIl1 long with one end like a c1avule from which at least six spines are curved back about 2/3rd of the entire length of the spiculum. The other end is inflated and forms a small budo This spicule, though much smaller, shows great sirni- larity to Nabaviella elegans described by MosT- LER & MOSLEH-YAZDI (1976) from Late Cambri- an sediments of Iran. Some c1avules also occur (fig. 1: 12). 339 The Early Paleozoic radiation of the Hexac- tinellida The monophylum Hexactinellida comprises the sister groups Amphidiscophora and Hexas- terophora, which can be traced back to the Early Paleozoic (MosTLER, 1986; MEHL, 1991, 1995). Oxyhexasters from the Late Ordovician and am- phidiscs from the Late Silurian were documented by MOSTLER (1986). From Late Cambrian sedi- . ments MOSTLER & MOSLEH-YAZDI (1976) docu- mented large (about 0.6--3.2 mm) monaxon spi- cules with one clavulate and one swollen, turbu- lated end. These spicules, fIrst classifIed as chan- celloriides, were re-interpreted as hernidiscs by MOSTLER (1986). In case the spicules Nabaviella elegans are really hemidiscs, it would imply that they were amphidiscs with partly atrophied rays. Further, this would infer that both sister groups, Amphidiscophora and Hexasterophora, were do- cumented since the Cambrian. However, no true amphidiscs are known older than Silurian, in spite of careful investigation of sediments from many different localities and facies types (MosT- LER, pers. comm.). For this reason, MEHL (1992) chose the name tylodiscs for the Early Paleozoic paraclavule-like spicules with one inflated end. The alternative hypothesis that the amphidisc are phylogenetically derived from hemidiscs (FINKS, 1970), is considered even more unestablished (MEHL, 1991). The oldest true Hemidiscs found so far were reported from the sponges, Uralone- ma karpinskii and Microhemidiscia ortmanni of the Late Carboniferous from Ural and Uruguay, respectively (LmROVITCH, 1929; KLING & REIF, 1969). According to close observations of the type material to those species by one of the au- thors (DM), in St. Petersburg and Tübingen, res- pectively, these true hemidiscs are quite different from the Cambrian tylodiscs. Hernidiscs are thus most probably derived from amphidiscs by ray- reduction, and tylodiscs can be considered as convergently evolved, may be related with the paraclavules (the latter spicules are known from some Dictyospongiidae, e.g. Griphodictya epi- phanes HALL & CLARKE, 1898). The occurrence of highly evolved hexactinellids, such as Hyalo- 340 sinica archaica already in the Early Cambrian (Tommotian) as well as the very rich and diverse assemblages of hexactinellid spicules found in Cambrian sediments (also by BENGTSON, 1986; BENGTSON et aI. , 1990) indicate an Early Cam- brian radiation within the Hexactinellida. • Archaeocyatha The archaeocyaths are probably representati- ves of coralline sponges which secrete a secon- dary calcareous skeleton of high Mg-calcite (REITNER, 1990, 1992; VACELET & DEBRENNE, 1984; WOOD et aI. , 1992; ZHURAVLEV, 1989). Within some irregular archaeocyaths of the Flin- ders Ranges (South Australia) tetractine and mo- naxonic spicules were found (REITNER, 1992). The spicules are often intramurally enriched within endo- or exothecal buds (pI. 2, fIgs. 3-6). The calcifying tissue of the archaeocyaths has overgrown more or less entirely the primary spi- cular skeleton which is presumably related to the buds (pI. 2, fIgs. 3-4). The biomineralization of calcareous buds was a fast process which ex- plains that the spicules are often more or less in their natural position. The special biomineraliza- tion of the buds may explain the scarcity of spi- cules within archaeocyath basal skeletons (pI. 2, fIgs. 5-6). The incorporation of spicules within the basal skeleton is probably a calcifIcation arte- fact, a phenomenon which is also seen within the basal skeletons of Triassic coralline sponge Cas- sianothalamia zardinü REITNER (RErTNER, 1987) and Lower Cretaceous ßcanthochaetetes sp. (REITNER & ENGESER, 1987). The obsetved spic- ule types are modifIed tetractines ("dodeca- actine", triaenes) (pI. 2, fIgs.' 4, 6) and mon actine spicules (pI. 2, fIg.4) with demosponge affini- ties. The spicules exhibit the characteristic dia- genesis which is known from fossil siliceous spicules: The early dissolution molds are cemented by a granular andlor equant calcite often associated with early prismatic marine ce- ment. In contrast, fossil calcareous spicules still exhibit their mono- to paucicrystalline character. Geol. Paläont. MUt. Innsbruck, Bd. 20, 1995 The validity of the relationship of these spic- ules to the host archaeocyaths is under discussi- on. The spicules are doubtless tetractine spicules and therefore not related to hexactinellids. However, the archaeocyath tissue was able to incorporate allochthonous material, as trilobate remains, isolated spicules, further organic re- mains, and overgrow sessile benthic organisms (e.g., Gravestockia pharetroniensis REITNER). To decide whether or not spicules were a constituent character of the archaeocyathid animal always depends on a paleobiological reconstruction of the studied specimen. • Demosponges Only few informations on the early history (Cambrian) of demosponges are available • (BENGSTON et aI. , 1990; RIGBY, 1991; KRUSE, 1983, 1990; VAN KEMPEN, 1990) and are restrict- ed to the Middle Cambrian. The sponges described by CI-IEN et aI. (1990) from the Cheng- jang deposits, as explained above, are most prob- ably no demosponges but hexactinellids. Identi- fiable articulated demosponges from the Lower Cambrian are until yet not known. The spicules of the demosponges in Archae- ocyath mounds are exclusively restricted to the Tetractinellida. First Ceractinomorpha do occur during the Middle Cambrian (KRUSE, 1990; REITNER, 1992; REITNER & MEHL, in press). Be- side tetractine spicules, typicaI modified deI mal spicules (nail-types) (pI. 1, fig. 1-3), monaxonic spicules (large tylostyles) (pI. 1, fig. 3) and large aster microscleres (autoapomorphy of the Tetrac- tinellida) (sterrasters, Geodiidae) (pI. 1, fig.4) were found which proves the high level of tetrac- tinellid evolution in the early Cambrian (GRUBER & REITNER 1991). The tetractinellid spicules of the Lower Cambrian differ in some aspects from modern ones. Most of the observed Cambrian spicules are generally larger (500 /lffi to 3 cm) and thicker (ca. 50/lffi) than the recent ones. The second main tribe of the demosponges, the Ceractinomorpha, occurs first in the Middle Geol. Paläont. Mill. Innsbruck, Bd. 20, 1995 Cambrian. KRUSE (1990) has found sigmata microscleres (autapomorphy of this taxon) in the Middle Cambrian Daly and Georgina Basin. First record of a probable keratose sponge is Vauxia sp., e.g. known from the Lower Cambri- an of Greenland (RIGBY 1987). Desma type me- gascleres (orchoclad lithistids with dendro- clones) occurred Iatest in the Middle Cambrian (anthaspidellid spqnge Rankenella, KRUSE, 1983). Based on phylogenetic reconstruction this spicule type should be an archaic one, but no desmas are know from the Lower Cambrian up to now. • Calcarea The Heteractinida are the most characteristic calcareous sponges in the Paleozoic (RIGBY, 1991; RIETSCHEL, 1968). They extinct at the end of the Permian. Within Lower Cambrian strata these sponges are rare and only few spi- cules are known (Eiffe/ia sp.) (pI. 2, fig. 2). The spicule types of the Heteractinida are ex- tremely variable and completely different from the "modern" type Ca1carea. It is astonishing that the "modern" type of calcitic spicules is common in Lower Cambrian strata. Up to now, regular ca1citic triaene spicules were first known from Mesozoic strata. The modern type of spicules of Ca1carea from the Lower Cam- brian give absolutely new aspects of the phylo- geny of this group. Remarkable is this evidence of ca1careous sponges with "modern" affini- ties. The main prob\em is that typical Paleo- zoic Ca1carea are the Heteractinida with their multi-rowed spicules or characteristic octac- ( tines. Entirely preserved sponges with regular triaene or tetractine ca1citic spicules are not known from the Paleozoic. Beside isolated re- gular calcitic spicules one taxon is known from the archaeocyath mounds of the Flinders Ran- ges with cemented choanosomal spicules ("Pharetronida"), Gravestockia pharetroniensis REITNER, 1992 (pI. 1, fig. 5-6). This pharetro- nid sponge exhibits a rigid skeleton of simple 341 tetractine calcareous spicules and diactine-free dermaiones. These sponges grew preferably on archaeocyaths and are sometimes overgrown by the secondary skeletons of the archaeocyaths. This "intramural" spicule record is not to be confused with the isolated tetractinellid spicu- les with demosponge affinities, as described above. Most of the observed calcarean spicules have affinities to the modem Calcaronea (pI. 1, fig. 1), only few with calcinean affinities are known (pI. 2, fig. 1). Conclusions The Hexactinellida are the oldest definite Po- rifera documented, their record goes back to the Late Proterozoic, and their radiation took place in the Early Cambrian. The Archaeocyatha were most probably de- mosponges with secondary calcareous basal skeletons. Evidence of "intramural" spicules within archaeocyaths may be interpreted as trapped dermal spicules that belong to the budding archaeocyath specimens. - The main radiation of the Pinacophora (De- mospongiae/Calcarea-taxon) took place in the Middle Cambrian. - All main sponge taxa occur latest in the Midd- le Cambrian. New fundamental evolutionary trends were not realized during the following time. Acknowledgements Prof Dr. KJ. MÜLLER (Bonn) and Dr. lH. SHERGOLD (Canberra, Australia) . are acknowl- edged for loan of their rich collection of sponge spicules from the Georgina Basin. The Deutsche Forschungsgemeinschaft (DFG) is gratefully acknowledged for financing the pro- ject Ke 322/5-3 on palaeobiology of siliceous . sponges of the Paleozoic. 342 References BENGTSON, S. ; CONWAY MORRIS, S.; COOPER, B.J.; JELL, P.A. & RUNNEGAR, B.N. (1990): Early Cambrian Fos- sils of South Australia. - Ass. Austr. Palaeont., Mem. , 9, 364 p., Brisbane. BENGTSON, S. (1986): Siliceol:ls microfossils from the Upper Cambrian of Queensland. - Alcheringa, 10, 195-216, Adelaide. • CHEN, 1., Hou, X. & LI, H. (1990): New Lower Cambrian de- mosponges - Quadrolaminiella gen. nov. from Cheng- jiang, Yunnan. - Acta Paeont. Sinica, 28, 1, 401-414. CHEN, J. & ERDTMANN, B.-D. (1991): Lower Cambrian fossil Lagerstätte from Yunnan, China: Insights for reconstructing early metazoan Iife. - In: SIMONETTA, A.M. & CONWAY MORRlS, S. (eds.): The early evolu- tion of Metazoa and the significance of problematic taxa, 57- 76, Cambridge. CHEN, J., Hou, X. & Lu, H. (1989): Lower Cambrian Leptomitids (Demospongea), Chengjiang, Yunnan. - Act Palaeont. Sinica, 28, 17-30. CHEN, J. , Hou, X. & Lu, H. (1990): New Lower Cambri- an Demosponges - Quadrolaminiella Gen. Nov. from Chengjiang, Yunnan. - Act Palaeont. Sinica, 29, 410-413. DAWSON, J.W. & HINDE, G.J. (1888): New species of fossil sponges from Little Metis, Province of Quebec, Canada. - Can Rec. Sci., 3, 49-68. DAWSON, J.W. & HINDE, G.J. (1889): New species of fossil sponges from the Siluro-Cambrian at Little Metis on the Lower St. Lawrence. - Trans. Roy. Soc. Can., 7, 31-55. DEBRENNE, F. & REITNER, J. (submitted): New phyloge- netic and palaeoecological aspects of late Precambri- an and early Cambrian sponges and cnidaria. - Univ. Press Columbia. • DEBRENNE, F. & VACELET, J. (1984): Archaeocyatha: Is the sponge model consistent with their structural orga- nization? - Palaeontographica Americ., 54, 358-369. FINKS, R.M. (1970: The evolution and ecologic history of sponges during Palaeozoic times. - Symp. Zoo!. Soc. London,25, 3-22. GRUBER, G. & RElTNER, J. (1991): Isolierte Mikro- und Megaskleren von Porifera aus dem Untercampan von Höver (Norddeutschland) und Bemerkungen zur Phy- logenie der Geodiidae (Demospongiae). - Berliner Geowiss. Abh. (A): 134,107-117, Berlin. Geol. Paläont. Mitt. lnnsbruck, Bd. 20, 1995 HIRABAYASHI, J. & KASAI, K. (1993): The family of me- tazoan metal-independent ß-galactose-binding lec- tins: structure, function and molecular evolution. - Glycobiol. 3, 297-304, Oxford. KEMPEN VAN, T.M.G. (1990): On the Oldest Tetraxon Megascleres. - In: RÜTZLER, K. (ed.): New Perspec- tives in Sponge Biology, 9-16; Washington (Smith. Inst. Press) KLING, S.A. & REIF, W.E. (1969): The Paleozoic history of amphidisc and hemidisc sponges: New evidence from the Carboniferous of Uruguay. - Jour. Paleont., 43, 6, 1429-1434. KRUSE, P. (1983): Middle Cambrian "Archaeocyathus" from the Georgina Basin is an anthaspidellid sponge. - A1cheringia, 7, 49-58. KRUSE, P. (1990): Carnbrian Palaeontologiy of the Daly Basin. - Dept. Mines Energy North. Terr. Geol. Surv. Rep. 7, 58 p., Darwin. LlBROVITCH, L.S. (1929): Uralonema Karpinskii nov. gen., nov. sp. i drugie kremnevie gubki iz kamennou- golnikh otlozhenii vostochnogo sklona Urala. - Trudy Geol. Kom.-ta., nov. ser., 179, 11-57, Mo- skau. MEHL, D. & ERDTMANN, B.-D. (1994): Sanshapentella dapingi n. gen., n. sp. - a new hexactinellid sponge from the Early Cambrian (Tommotian) of China. - Berliner Geowiss. Abh. Ser. E, 13, 315-319. MEHL, D. & MOSTLER, H. (1995): Phylogenetisch-taxo- nomische Bedeutung, stratigraphische Bedeutung und Terminologie der "Mikroskleren" bei den Hexactinel- lida, Porifera.- Geol. Paläont. Mitt. Innsbruck (in press). MEHL, D. & REISWlG, H.M (1991): The presence offla- gell ar vanes in choanomeres of Porifera and their pos- sible phylogenetic implications. - Z. syst. Zoo!. Evo- lutionsforsch., 29, 312-319; Hamburg. MEHL, D. (1991): Are Protospongiidae the stem group of modern Hexactinellida? - In: REITNER, J. & KEUPP, H. (eds.): Fossil and Recent Sponges, 43-53 (Sprin- ger). MEHL, D. (1992): Die Entwicklung der Hexactinellida seit dem Mesozoikum. Paläobiologie, Phylogenie und Evolutionsökologie. - Berliner geowiss. Abh. (E), 2, 1-164, Berlin. MEHL, D. (1995): Phylogenie und Evolutionsökologie der Hexactinellida (Porifera) im Paläozoikum. - GeoL Paläont. Mitt. Innsbruck (in press). Geol. Paläont. Mitt. Innsbruck, Bd. 20, 1995 MOLDowAN, 1.M., DAHL, J., JABOSON, S.R., HUIZINGA, B.J., MCCAFFREY, M. A. & SUMMONS, R.E. (1994): Molecular fossil evidence for late Proterozoic-Early Paleozoic Environments. - Terra nova Abstracts, 3, supp!. Terra nova, 6, 5, Oxford. MOSTLER, H. & MOSLEH-YAZDI, A. (1976): Neue Porife- . ren aus oberkambrischen Gesteinen der Milaformati- on im Elburzgebirge (Iran). - Geol. Paläont. Mitt. Innsbruck, 5, 1, 1-36. • MOSTLER, H. (1986): Beitrag zur stratigraphischen Verbreitung und phylogenetischen Stellung der Amphidiscophora und Hexasterophora (Hexactinel- lida, Porifera). - Mitt. österr. geo!. Ges., 78, 319-359. MÜLLER, W.E.G., SCHRÖDER, H.C., MÜLLER, I.M. & GA- MULIN, V. (1995): Phylogenetic Relationship of Ubi- • quitin Repeats in the Polyubiquitin Gene from the Ma- rine Sponge Geodia cydonium. - J. Mo!. Evo!., 39, in press. PFEIFER, K., HAASEMANN, M., GAMULIN, V., BRETTING, FAHRENHOLZ, F. & MÜLLER, W.E.G. (1993): S-type lectins occur also in invertebrates: High conservation of the carbohydrate recognition domain in the lectin genes from the marine sponge Geodia cydonium. - Glycobiology, 3, 179-184; Oxford REISWIG, H.M & MEHL, D. (1991): Tissue organisation of Farrea occa (Porifera, Hexactinellida). - Zoomor- phology, 110,301:-311; Hamburg REITNER, J. & MEHL, D. (in press): Monophyly and sy- stematics of the Porifera. - Abh. Naturwiss. Verein Hamburg (NF), Hamburg. REITNER, J. & ENGES ER, T. (1987): Skeletal structures and habitates of recent and fossil acanthochaetetes (subclass Tetractinomorpha, Demospongiae, Pori- fera). - Coral Reefs, 6,151-157; Berlin. REITNER, J. (1987): A new ca1citic sphinctozoan sponge , belonging to the Demospongiae from the Cassian Formation (Lower Carnian; Dolomites, Northern , Italy) and ist phylogenetic relationship. - Geobios, 20, 571-589; Lyon. REITNER, J. (1990): The polyphyletic origin of the "Sphinctozoans". - In: RÜTZLER, K. (ed.): New Per- spectives in sponge Biology, 33--42; Washington (Smith. Inst. Pre~s). REITNER, 1. (1992): "Coralline Spongien" Der Versuch einer phylogenetisch-taxonomischen Analyse. - Ber- liner geowiss. Abh. (E), 1,352 p.; Berlin. 343 RJGBY, J.K. (1975): Some unusual hexactinellid sponge spicules from the Cambrian Wilberns Formation of Texas. - Jour. Paleont. , 49, 2, 416-415. RIGBY, J.K. (1978): Porifera of the Middle Cambrian Wheeler Shale, from the Wheeler Amphitheater, House Range, in Western Utah. - Jour. Paleont. , 52, 6,1325-1345. RIGBY, J.K. (1983): Sponges of the Middle Cambrian Marjum Limestone from the House Range and Drum Mountains of Western Millard Country, Utah. - Jour. Paleont. , 57, (2,: 240-270. RIGBY, J.K. (1986): Sponges from the Burgess Shale (Middle Cambrian), British Columbia. - Palaeontogr. Canad., 1986,2, 1-105. • RJGBY, J. K. (1987): Cambrian and Silurian sponges from northern Greenland. - In: PEEL, JS. (ed.): North Greenland Lower Paleozoic paleontology and strati- graphy: Shorter Contributions. - Grn\. Geo\. Under- s0gelse Rapp. , 132,51-63. RIGBY, J. K. (1991): Evolution of Paleozoic Heteractinid Calcareous Sponges and Demosponges. - In: REIT- NER, J. & KEupp, H. (eds.): Fossil and Recent Spon- ges, 83-101, Berlin (Springer). RJGBY, J. K. (1993): Fossils in Time, the record and its interpretation (talk). - 4th Internat. Porifera Congress, Amsterdam. RI ETSC HEL, S. (1968): Die Octactinellida und ihnen • verwandte paläozoische Kalkschwämme (Pori- fera, Calcarea). - Paläont. Z., 42, 13-32, Stutt- gart. SCHULZE, F.E. (1887): Report on the Hexactinellida. - In: MURRAY, J. (ed.): Report on the sc ientific results of the voyage of H.M.S. CHALLENGER during the years 1873-76,21, 1-513, London. STEINER, M., MEHL., D. , RElTNER, 1. & ERDTMANN, B.-D. (1993): Oldest entirel y preserved sponges and other fossils from the Lowermost Cambrian and a new fa- eies reconstruction of the Yangtze platform (China). - Berliner geowiss. Abh. (E), 9, 293-329; Berlin • WALCOTT, C.D. (1920): Middle Cambrian Spongiae. - Smithsonian Mise. Col!., 67, 261-364. • WEBBY, B.D. & TROTTER, 1. (1993): Ordovician sponge spicules from New South Wales. - Jour. Paleont. , 67, 1,21-48. WOOD, R. , ZHURAVLEV, A.Yu & D EBRENNE, F. (1992): Functional Biology and Ecology of Archaeocyatha. - Palaios, 7, 131- 156, Tulsa. YOUNG, J. & YOUNG, J. (1877): On a Carboniferous Hyalonema and other sponges from Ayrshire. - Ann. Mag. Nat. Hist., ser. 4 , 20, 425-432. ZHANG, XI-GUANG & PRATT, B.R. (1994): New and ex- traordinary Early Cambrian sponge spicule assembla- ge from China. - Geology, 22, 43-46. ZHURAVLEV, A.Yu. (1989): Poriferan aspects of archaeo- cyathan skeletal function. - Ass. Australas. Palaeon- tols Mem. 8, 387-399. • Authors' addresses: Prof Dr. Joachim Reitner, Institut und Museum für Geologie und Paläontologie, Universität Göttingen, Goldschmidtstr. 3, D-37077 Göttingen; Dr. Dorte Mehl, Institut für Paläontologie der Freien Universität Berlin, Malteserstr. 74- 100, D-J2249 Berlin. Manuscript submitted: February 20, 1995 Platel: Atdabanian archaeocyath mound of the Flinders Ranges, Wilkawillina Limestone , Fig. 1: Vertical section of a triaene dermal spicule of a tetractinellid demosponge (a). The spicule exhibits the charac- teristic granular cements of former siliceous scleres. In contrast to the diagenetically altered siliceous spicule, the calcarean spicule (b) exhibits a monocrystalline structure. Scale 500 /-Im. . Fig. 2: Horizontal section of a phylotriaene dermal spicule (Demospongiae). Scale 500 /-Im. Fig. 3: Tylostyle (a) and triaene (8b) demospongian megascleres. Scale 500 /-Im. Fig. 4: Modified kidney-shaped sterraster microscleres with demosponge affinities (Geodiidae). Scale 300 /-Im. Fig. 5: Gravestockia pharetronensis REITNER 1992; Calcaronea with pharetronid affinities. This sponge preferred to settle on archaeocyathids. Scale 500 /-Im. Fig. 6: Detailed magnification of the rigid spicular skeleton of Gravestockia pharetroniensis which shows the secon- , dary calcite cement (arrow) of the pharetronids. Scale 150 /-Im . • • 344 Geol. Paläont. Mitt. Innsbruck, Bd. 20, 1995 • • • • • • • " n • • • -• , ~, " " • ". '" • • •• ';', .. ,... 'Y" . ' • , . .... ' • , • • • • - . ' • , • • . , , • ~ • . "r • • ., , • • • • 0' • . , ,.-;,,1.. . ... ~, •. _" • ' " • 345 Plate 2: Atdabanian archaeocyath mound of the Flinders Ranges, Wilkawillina Limestone Fig. 1: Regular monocrystalline calcitic triactine spicule with Calcinean affinities. Scale 250 /lm. Fig. 2: Sexiradiate spicule with heteractinid affinities (Eiffelia ?). Scale 300 /lm. Fig. 3: Endothecal bud of an archaeocyathid with intramural spicules (arrows). Scale 500 /lm. Fig. 4: Detailed magnification of the basal skeleton with intramural spicules of Fig. 3. Clearly visible are triaenes and cross sections of the long spicule axis (white dots). Scale 100 flill. Fig. 5: Entrapped spicules within the primary layer of the main skeleton of an archaeocyath (white dot and arrow). Scale 1 mm. • Fig. 6: Detailed magnification of a dodecaactinellid shape of spicule, a modified siliceous caltrop spicule (Fig. 5, arrow). Scale 200 flill • • • --' • • • • • • • • • • • , • • • • ,, • • • " • , r • , • ,.. • • • r , • -\: 347