G.D. Edgecombe*, G. Giribet, C.W. Dunn, A. Hejnol, R.M. Kristensen

Higher-level metazoan relationships: recent progress and remaining questions Organisms Diversity & Evolution

G.D. Edgecombe*, G. Giribet, C.W. Dunn, A. Hejnol, R.M. Kristensen,

R.C. Neves, G.W. Rouse, K. Worsaae, M.V. Sørensen

* Department of Palaeontology, The Natural History Museum, London, UK

e-mail:

Electronic supplementary material: Palaeontological calibration data for bilaterian divergence dates

The geological time scale and absolute dates for boundaries follow the 2008 International Stratigraphic Chart (Ogg et al. 2008).

Many bilaterian clades are first found in the early Cambrian Burgess Shale-type assemblages of South China conventionally called the Chengjiang fauna. Dates in the literature range from pre-Tommotian to Botoman in the Siberian regional standard (i.e. Stages 1 to 3 of the global 10-stage division of the Cambrian outlined by Zhu et al. 2006). We follow Steiner et al. (2005) and Zhu et al. (2006) in dating Chengjiang to the Atdabanian (Series 2, Stage 3) or 520 Mya. New age constraints for the early Cambrian in Maloof et al. (2010) do not substantially alter this dating; radiometric dates for the late Tommotian and latest Atdabanian therein are 520.9 and 517.0 Mya, respectively (Maloof et al. 2010, fig. 1B). Zhu et al. (2006, fig. 1) are followed for the datings of other key Cambrian Konservat-Lagerstätten.

Bilateria incertae sedis

Kimberella from the upper part of the Ediacaran in South Australia and the White Sea (the latter occurrence radiometrically dated at 555 Mya by Martin et al. 2000) was argued to be a mollusc-like bilaterian by Fedonkin and Waggoner (1997) and Fedonkin et al. (2007). Caron et al. (2006) tentatively placed Kimberella in the mollusc stem group; Conway Morris and Caron (2007) depicted alternative positions in the mollusc or spiralian stem groups. Its apparently soft sole and dorsal, unmineralised “shell” (Fedonkin et al. 2007), coupled with behaviour inferred from fan-shaped sets of scratchings produced by paired hooks associated with a proboscis (Fedonkin 2003; Gehling et al. 2005) are the most informative arguments for identifying Kimberella as mollusc-allied. Kimberella has not received universal acceptance as a mollusc or even as a definite bilaterian, though skeptics generally acknowledge it to be the most promising candidate for an Ediacaran representative of the Bilateria (Budd and Jensen 2000; Valentine 2004), and the behavioural evidence for bilaterian-like grazing has been strengthened in works published subsequently. We assign Kimberella no more specifically than as a bilaterian (i.e. do not use it to date the protostome-deuterostome split or any protostome node). Even if this identification is disputed, a range of trace fossils dating to at least the same age, i.e. 555 Mya, are less controversially identified as attesting to the presence of bilaterians (e.g. Jensen 2003; Jensen et al. 2005).

The proposed bilaterian characters of the putative Duoshantuo (ca. 580 Mya; Yin et al. 2007) bilaterian Vernanimalcula (Chen et al. 2004a) are as reasonably (or more reasonably) interpreted as diagenetic artefacts (Bengtson and Budd 2004). We do not attribute this material to the Metazoa.

Ecdysozoa

Markuelia has been identified as a stem-group scalidophoran (Dong et al. 2005) or as a stem-group priapulid (Cobbett et al. 2007). It is used to date the split of priapulids and kinorhynchs from other ecdysozoans. The oldest fossils of Markuelia from the Cambrian are from the early Tommotian of Siberia (Bengtson and Zhao 1997; Donoghue et al. 2006), and date that split to ca. 525 Mya (date constrained by boundaries of Cambrian Stage 2 fide Zhu et al. 2006; see also Ogg et al. 2008).

The split of Priapulus from Echinoderes is dated by the earliest priapulids, several taxa occurring in the Chengjiang fauna (520 Mya). These include possible crown group priapulids such as Xiaoheiqingella (Huang et al. 2004b) or at least stem-group priapulids (Dong et al. 2005).

The tree used for dating resolves Tardigrada as sister to Nematoida, this split being constrained by Cambrian stem-group tardigrades from the Kuonamka Formation in Siberia (Müller et al. 1995). Association of the tardigrades with Pentagnostus praecurrens constrains their occurrence to ca. 508 Mya.

Divergence between Nematoda and Nematomorpha is calibrated by fossil nematodes. The assignment of palaeoscolecids to Nematomorpha by Hou and Bergström (1994) is rejected in favour of them being stem-group priapulids (Conway Morris 1997; Dong et al. 2005; Harvey et al. 2010; Wills 1998); this excludes Palaeozoic taxa from Nematomorpha, the fossil record of which is confined to the Eocene. The minimal dating used here for the nematode-nematomorph split is 408 Mya, based on the Early Devonian nematode Palaeonema (Poinar et al. 2008).

The split between Arthropoda and Onychophora is dated by the earliest arthropods. Budd and Jensen (2000) cited Rusophycus trace fossils that extend as far back as the base of Cambrian Stage 2 (Crimes 1992), ca. 528 Mya, as the earliest evidence for Arthropoda. We employ this date but note that it pre-dates arthropod body fossils, which are not widely accepted as pre-dating the early part of Stage 3. Although a few arthropod body fossil occurrences have been dated as Stage 2, they are regarded by other workers as being Stage 3. The controversial Stage 2/'Tommotian' records include the nektaspid Liwia (a trilobitomorph) in Poland (Dzik and Lendzion 1988) and a bivalved arthropod from China assigned to Perspicaris (Steiner et al. 1993).

For the basal split within the Arthropoda, Yicaris (Zhang et al. 2007) from the early Cambrian (Stage 3; same trilobite zone as the Chengjiang fauna) of China is a 'crustacean' that can be placed conservatively on the stem lineage of the Tetraconata, i.e. the clade composed of crustaceans and insects. This dates the split between crustaceans and insects versus the myriapods and chelicerates at minimally 520 Mya. Metanauplii named as Wujicaris have maxillopodan crustacean affinities (Zhang et al. 2010) and constrain the same node or, depending on the resolution of crustacean interrelationships, the split between Daphnia and Drosophila. The latter split is more conservatively dated by the Cambrian Rehbachiella, a stem-group branchiopod fide Møller et al. (2004) and Olesen (2004). It occurs in Orsten of the Agnostus pisiformis Zone, ca. 500 Mya.

The split between Scutigera and Chelicerata s.l. (pycnogonids + chelicerates) is dated by various members of Megacheira (stem-group Chelicerata fide Chen et al. 2004b; Cotton and Braddy 2004) in Cambrian Series 2, Stage 3. Chengjiang taxa (such as Haikoucaris) are as old as any other megacheirans, i.e. ca. 520 My.

The split between Anoplodactylus and the euchelicerates is dated by the Cambrian Cambropycnogon, a stem-group pycnogonid (Waloszek and Dunlop 2002). As for other Orsten fossils used to date nodes, the Agnostus pisiformis Zone from which it is sourced dates to ca. 500 Mya.

The split between Carcinoscorpius and the three arachnids is dated by the earliest xiphosurans, documented by Van Roy et al. (2010) from the Lower Fezouata Formation in Morocco, of Tremadocian age (minimally dated at its top as 478.6 My).

The split between Acanthoscurria and Boophilus is provided by Early Devonian (Pragian) prostigmatan mites in the Rhynie Chert, Scotland (Hirst 1923; Selden 1993), ca. 408 Mya. An alleged Early Ordovician oribatid mite (Bernini et al. 2002) is accepted by some acarologists (Alberti 2005) but requires further evidence for authenticity.

Spiralia

Brachiopods are present by at least the early Tommotian (Cambrian Stage 2), represented by paterinids (linguliforms) from Siberia, assigned a date of ca. 525 Mya. This will date the respective splits between brachiopods and nemerteans, brachiopods and phoronids, and of Kryptrochozoa from other spiralians. Slightly earlier dating of Brachiopoda is dependent upon the identity of such fossils as the camenellans and Eccenthotheca as stem-group brachiopods (Skovsted et al. 2009). The oldest representatives of these groups occur in the Watsonella crosbyi Zone in Newfoundland and pre-date the oldest tommotiids (a grade of stem-group Brachiopoda fide Skovsted et al. 2009) in the basal Tommotian in Siberia or Meischucunian in China. The fossil record of Nemertea may extend as far back as the Carboniferous, the earlier of two occurrences being Namurian (Schram 1979), ca. 325 Mya.

The oldest accepted molluscs are from the lowermost Cambrian, with several taxa, including species of Latouchella, in the middle Purella Zone of the uppermost Nemakit-Daldynian (Fortunian) in Siberia and China. This stage ranges between 542 and 528 Mya on the Ogg et al. (2008) time scale, the conservative date being the latter, more recent one (base of the overlying Stage 2). This dating for Mollusca has been used—very conservatively—by Benton and Donoghue (2007) as a minimal age for the protostome-deuterostome split. Latouchella is a helcionellid, and various workers have argued for this group being related to particular crown-group mollusc clades (e.g. Monoplacophora fide Peel 1991; Gastropoda fide Parkaev 2008), though others have (probably overcautiously) assigned them to the mollusc stem group instead (e.g. Budd 2003). Conservatively they would be used no more precisely than as Mollusca.

Early Cambrian Bivalvia (Fordilla, Bulluniella) are known from the latter half of the Tommotian (Parkaev 2008), thus calibrating the divergence of bivalves (Mytilus, Argopecten and Crassostrea) from gastropods to at least 521 Mya (base of Cambrian Series 2, Stage 3).

The divergence of Polyplacophora (Chaetopleura) from cephalopods (Euprymna) and caudofoveates (Chaetoderma) is constrained by late Cambrian polyplacoporans (Matthevia). The oldest records of Matthevia (and thus Polyplacophora) are from the middle Sunwaptan/Trempealeauan (Pojeta et al. 2005), i.e. ca. 492 Mya. The status of alleged early Cambrian polyplacophorans remains uncertain (Parkaev 2008). The cephalopod-caudofoveate (Euprymna-Chaetoderma) divergence is constrained by the oldest cephalopods, plectronocerids from the Furongian (early late to latest Dolgellian) in China, i.e. 489 Mya.

The split of the sipunculans (Themiste) from other annelids is constrained by the Chengjiang Archaeogolfingia and Cambrosipunculus (Huang et al. 2004a), i.e. ca. 520 Mya. An ingroup position of Sipuncula in the Annelida in the tree used for dating dates the broader annelid clade to 520 Mya. If Sipuncula is resolved as sister to all other annelids, this latter clade dates to ca. 519 Mya (Phragmochaeta from the Sirius Passet Lagerstätte; Conway Morris and Peel 2008).

Divergence of Echiura (Urechis) from Capitella is constrained by the Late Carboniferous (Westphalian D, ca. 307 Mya) echiuran Coprinoscloex (Jones and Thompson 1977).

Deuterostomia

The conservative fossil assignments of Benton and Donoghue (2007) and Swalla and Smith (2008) are followed here. The split of Echinodermata (Strongylocentrotus and Asterina) from Hemichordata is dated by early Cambrian echinoderms such as helicoplacoids (earliest fossils in the Nevadella Zone fide Durham 1993, ca. 519 Mya). We do not accept Arkarua from the late Ediacaran (Gehling 1987) as sufficiently well supported as an echinoderm to use it for dating. The internal split between Strongylocentrotus and Asterina is calibrated by the oldest asteroids, with Petraster from the Early Ordovician (late Tremadoc, ca. 480 Mya) being the oldest well-corroborated asteroid (Shackleton 2005).

Some early Cambrian fossils have been identified as hemichordates (e.g. yunnanozoans by Shu et al. 1996), but this identification has been rejected since, including by its original authors; even if it were accepted the echinoderms are still older, i.e. on the lineage where the enteropneusts split off from echinoderms, the safest date for hemichordates is ca. 510 Mya for dendroid graptolites (Maletz et al. 2005), which are well established as ingroup hemichordates (pterobranchs).

The same date provided by the echinoderms cascades down to the node where Xenoturbella splits, because no fossil data speak to the Xenoturbella question.

The basal split in the Chordata (between tunicate and vertebrates) is constrained by the Chengjiang vertebrates Myllokunmingia, Haikouichthys and Zhongjianichthys at ca. 520 Mya. A few authors have identified early Cambrian fossils as tunicates, but the only one that is well supported is Shankouclava (Chen et al. 2003) from the Chengjiang fauna (ca. 520 Mya). Swalla and Smith (2008) suggest that it is an ingroup tunicate (an ascidian).

The split between Homo and Gallus (i.e. the Synapsida-Sauropsida divergence) is constrained by the Carboniferous Hylonomus (Sauropsida: Diapsida) and receives a date of 312 Mya fide Benton and Donoghue (2007).

References

Alberti, G. (2005). On some fundamental characteristics in acarine morphology. Atti della Accademia Nazionale Italiana di Entomologia, 2005, 315–360.

Bengtson, S., & Budd, G. E. (2004). Comment on “Small bilaterian fossils from 40 to 55 million years before the Cambrian”. Science, 306, 1291a.

Bengtson, S., & Zhao, Y. (1997). Fossilized metazoan embryos from the earliest Cambrian. Science, 277, 1645–1648.

Benton, M. J., & Donoghue, P. C. J. (2007). Paleontological evidence to date the tree of life. Molecular Biology and Evolution, 24, 26–53.

Bernini, F., Carnevale, G., Bagnoli, G., & Stouge, S. (2002). An early Ordovician oribatid mite (Acari: Oribatida) from the island of Öland, Sweden. In F. Bernini, R. Nannelli, G. Nuzzaci, & E. de Lillo (eds.), Acarid phylogeny and evolution: Adaptation in mites and ticks (pp. 45–47). Dordrecht: Kluwer.

Budd, G. E. (2003). The Cambrian fossil record and the origin of phyla. Integrative and Comparative Biology, 43, 157–165.

Budd, G. E., & Jensen, S. R. (2000). A critical reappraisal of the fossil record of the bilaterian animal phyla. Biological Reviews, 75, 253–295.

Caron, J.-B., Scheltema, A., Schander, C., & Rudkin, D. (2006). A soft-bodied mollusc with radula from the Middle Cambrian Burgess Shale. Nature, 442, 159–163.

Chen, J.-Y., Bottjer, D. J., Oliveri, P., Dornbos, S. Q., Gao, F., Ruffins, S., et al. (2004a). Small bilaterian fossils from 40 to 55 million years before the Cambrian. Science, 305, 218.

Chen, J.-Y., Hwang, D.-Y., Peng, Q.-Q., Chi, H.-M., Wang, X-Q., & Feng, M. (2003). The first tunicate from the Early Cambrian of South China. Proceedings of the National Academy of Sciences of the United States of America, 100, 8314–8318.

Chen J.-Y., Waloszek, D., & Maas, A. (2004b). A new ‘great-appendage’ arthropod from the Lower Cambrian of China and homology of chelicerate chelicerae and raptorial antero-ventral appendages. Lethaia, 37, 3–20.

Cobbett, A., Wilkinson, M., & Wills, M. A. (2007). Fossils impact as hard as living taxa in parsimony analyses of morphology. Systematic Biology, 56, 753–766.

Conway Morris, S. (1997). The cuticular structure of the 495-Myr-old type species of the fossil worm Palaeoscolex, P. priscatorum (?Priapulida). Zoological Journal of the Linnean Society, 119, 69–82.

Conway Morris, S., & Caron, J.-B. (2007). Halwaxiids and the early evolution of the lophotrochozoans. Science, 315, 1255–1258.