Phillipsastreidae C.F. Roemer, 1883

Wrzołek, Tomasz, 2007, A revision of the Devonian rugosan phillipsastreid genus Smithicyathus, Acta Palaeontologica Polonica 52 (3), pp. 609-632 : 614-616

Family Phillipsastreidae C.F. Roemer, 1883 . Genus Smithicyathus Różkowska, 1980 .

Type species: Phillipsastrea cincta Smith, 1945 , designed by Różkowska 1980: 18, Upper Frasnian, District of Mackenzie, Redknife River, Canada.

Emended diagnosis.— Smithicyathus Różkowska, 1980 , is a massive to submassive phillipsastreid with (1) L/T index 0.335 or lower, i.e., its major septa do not reach deeper into tabularium than 0.335 of its radius; (2) numerous horseshoe dissepiments ( PP = 0.6 or more, typically 1.0, i.e., continuous series) sometimes obscured by expanded septa.

This combination of features seems rare if not unique within the family Phillipsastreidae ; it can be observed also in some (but not all) colonies of Ph. variabilis ( Sorauf, 1988) which, however, can be distinguished by degradation of peripheral septal blades to isolated spines combined with reduction of intercorallite walls ( Fig. 4A ), and rarely in Ph. progressa ( Różkowska, 1953; see Fig. 4B ), which is predominantly thamnasterioid (versus astreoid or phaceloid Smithicyathus ) and with corallites larger than in the species of Smithicyathus and in Medusaephyllum variabile .

Species assigned.— Pachyphyllum lacunosum Gürich, 1896 (including Phillipsastrea lacunosa mariae Wrzołek, 1993 ), Phillipsastrea cincta Smith, 1945 , Smithicyathus mcleani sp. nov. (included into Smithicyathus cinctus by McLean 1994b), Pachyphyllum smithi Różkowska, 1953 , and Smithicyathus lubliniensis Różkowska, 1980 ; also, Smithicyathus cf. lacunosus ( Gürich, 1896) and Smithicyathus cf. smithi ( Różkowska, 1953) as presented below.

Doubtful species. —Here belong the three species described by Spasskiy and Kravtsov (in Besprozvannykh et al. 1975): Phillipsastrea emendata , Frechastraea russakovi , and Stellatophyllum belkovskiense (but not: S. belkovskiense of Hou 1988). These species are discussed below.

Most probably not to this genus should be assigned the two other colonies, originally identified as representing the species of Smithicyathus , but with high L/T indexes, as listed in Appendix 3 (columns “o” and “p”). They are Pachyphyllum cinctum as illustrated by Soshkina (1952: pl. 20: 74), classified as Chuanbeiphyllum by McLean (1994b), with a narrow tabularium and broad aphroid dissepimentarium, and Pseudoacervularia cf. smithi of Pickett (1967: pl 1: 3–5) having marked lateral septal expansion, and possibly close to Phillipsastrea ranciae Coen−Aubert, 1987 . More material would possibly allow for better justification of taxonomical decisions, both at genus and species level, in case of these two taxa.

Comparative material.—Here belong over 50 colonies listed in Appendix 3, which represent 4 species of Phillipsastrea that are morphologically close to Smithicyathus . Among them are two colonies of Ph. progressa ( Różkowska, 1953) from the Holy Cross Upper Frasnian with atypically short major septa (see column “l” of Appendix 3; one of them is illustrated in Fig. 4B ), whereas a majority of colonies of this species (26 out of 28 studied) have septa that are significantly longer in the tabularium. This suggests only minor morphological overlap between Smithicyathus and Ph. progressa .

The other comparable species is Ph. variabilis ( Sorauf, 1988) from the Upper Frasnian of western North America, with a very broad range of variability (Appendix 3: columns “m” and “n”), especially with respect to septal length in the tabularium, septal development between tabularia (very frequently only occurring as isolated spines in this area), and also the frequency of horseshoe dissepiments. Five out of 12 colonies measured might be “numerically” included into Smithicyathus , among them also the specimen shown in Fig. 4A , but review of all the material allows for interpretation of Ph. variabilis as a distinct species, with significant morphological overlap with Smithicyathus . Occurrence of isolated spines (trabeculae) in outer septal regions may be the most conspicuous diagnostic feature of Ph. variabilis . According to McLean (personal communication, 2004) these two species ( S. cinctus and Ph. variabilis ) can occur together in the Upper Frasnian of western Canada.

Two very similar species, Ph.? dybowskii ( Różkowska, 1953) and Ph.? samsonowiczi ( Różkowska, 1953) , thus far only illustrated by two colonies, are peculiar submassive/ cerioid species: Ph.? dybowskii has longer septa in the tabularium and larger corallites ( Fig. 4C , see also Appendix 3, column “q”) than Ph.? samsonowiczi . Multiserial tabularia with distally convex profile and variable but generally poor development of horseshoe dissepiments testify against including these two species into Smithicyathus . Possibly they are related to astreoid Ph. ananas ( Goldfuss, 1826) sensu Różkowska (1953) .

Discussion.—The genus was introduced by Różkowska (1980: 18), who diagnosed it as containing massive to subphaceloid phillipsastreids, with an epithecate (three−layered) intercorallite wall commonly surrounding more than one corallite. In the Devonian of Poland, Różkowska (1980) recognized two species of the genus: S. lacunosus and S. lubliniensis . Wrzołek (1993) classified the species recognized in the present paper as Smithicyathus into Phillipsastrea , as four subspecies of Ph. lacunosa , i.e., Ph. lacunosa lacunosa, Ph. l. smithi, Ph. l. mariae ssp. nov. and Ph. l. lubliniensis . McLean, in a monograph of Canadian phillipsastreids (1994b: 79–80) followed the diagnosis of Różkowska (1980) and discussed possible synonymy of Smithicyathus with Paramixogonaria Liao and Birenheide 1985 , which remains dubious in his opinion. He also described in detail new Canadian material of the type species and discussed several species of the genus, among them Pseudoacervularia samsonowiczi Różkowska, 1953 from the Holy Cross Devonian. Fedorowski (2003), in a review of rugose corals from the Polish Devonian, re−illustrated and re−classified the Polish material of, among others, Różkowska (1953, 1980) and Wrzołek (1993). In his opinion Smithicyathus is represented in Poland, in addition to S. lacunosus lacunosus and S. lacunosus mariae , by S. dybowskii and S. samsonowiczi , but not by Phillipsastrea smithi .

Review of the intercorallite wall structure (see also Fig. 5 ) allows for revision of the generic concept of Smithicyathus . Both the literature and the author’s material indicate that the walls in the species considered here as belonging to Smithicyathus can be either pseudocerioid or, rarely cerioid or both structures can be seen in the same colony.

(1) Pseudocerioid walls. These can be seen in massive colonies and also in the holotype of S. cinctus ( Smith 1945: pl. 22: 4, re−illustrated here in Figs. 5A , 6A) and in the topotypic material, illustrated by McLean (1994b: pl. 3: 1–4). In the latter case pseudocerioid wall is indicated by local occurrence of vertically non−continuous wall, formed by isolated and short wall segments (seen in longitudinal section in Fig. 5A 2 ). Contrary to the opinion of McLean (1994b), pseudocerioid walls can be seen in massive S. smithi ( Różkowska, 1953) and also in its submassive colonies, with intracolony voids bordered by epithecate walls ( Fig. 5C ). Additionally, review of material of S. lubliniensis Różkowska, 1980 , indicates the presence of pseudocerioid walls in this submassive species.

(2) Cerioid (tripartite) walls. These typically are present in submassive–phaceloid colonies of S. mcleani sp. nov. (see Fig. 6B, C) and of S. lacunosus ( Gürich, 1896; see Fig. 5B ). These species are characterized by generally phaceloid form with isolated epithecate corallites, but with significant massive segments of colonies, with cerioid walls dominating. On the other hand pseudocerioid wall segments can also occur there, especially in newly formed partitions between parent and offset corallites ( Fig. 5B 2 ), but also between some adult corallites, as in Fig. 5B 1 .

Occurrence of multiple corallites within a single wall, considered to be an important characteristic of the genus Smithicyathus ( Różkowska 1980: 18; McLean 1994b: 80) does not occur as commonly as suggested in the literature. It can be clearly seen only in the type material of S. cinctus , from western Canada. In other material it seems that only juvenile multiple offsets are associated with their parent corallite to give the impression of this structure (as in S. mcleani sp. nov. in Fig. 6B).

From the other massive Phillipsastreidae , as reviewed by McLean (1994b, see also McLean 1994a), Trapezophyllum and its possible synonym Stellatophyllum are massive and cerioid, with numerous horseshoe dissepiments. Non−cerioid and commonly with the horseshoe dissepiments are: Phillipsastrea (its possible synonym Pseudoacervularia ), Chuanbeiphyllum , Pachyphyllum , and Smithicyathus . Within these Chuanbeiphyllum can be distinguished by large corallites and septal degeneration between the tabularia, whereas Phillipsastrea and Pachyphyllum are with medium−sized corallites. The former genus is with typically massive habit, the latter is massive and transitory to dendroid habit, with few corallites in a colony, highly elevated above its massive part. Smithicyathus and its probable synonym Paramixogonaria is with small corallites and short septa in tabularium. Frechastraea is with small−sized corallites and without horseshoe dissepiments.

Some representatives of the predominantly dendroid genus Thamnophyllum also develop the phaceloid arrangement of corallites and thus are similar to some species assigned to Smithicyathus by me. For example in western Canada phaceloid and short−septate Th. tructense (McLaren, 1959) is more similar to S. mcleani sp. nov. than dendroid and also short−septate Th. pedderi McLean, 2005 (both species illustrated by McLean 2005).

Figures

Fig. 4. Upper Frasnian tetracorals more or less similar to Smithicyathus. A. Phillipsastrea variablis (Sorauf, 1988), Western Canada, GIUS1531CD 09, a colony with short septa in tabularium, see text for exact location (A1 transverse, A2 longitudinal section). B. Phillipsastrea progressa (Różkowska, 1953), fields to the South of Bolechowice village, GIUS 402JB 01, colony with septa atypically short in the tabularium; the thamnasterioid species, with weak lateral septal expansion and numerous horseshoe dissepiments (B1 transverse, B2 longitudinal section). C. Phillipsastrea dybowskii (Różkowska, 1953), Sobiekurów quarry, GIUS 373SW 10, a typical colony with phaceloid to massive habit, ceriod (tripartite) walls seen in places without secondary silification, with long septa, and variable development of horseshoe dissepiments; (C1 transverse, C2 longitudinal section). Scale bar 4 mm for all items.

Fig. 5. Wall and septal microstructures in the Upper Frasnian Smithicyathus. A. The holotype of massive Smithicyathus cinctus (Smith, 1945), Western Canada, GSC 9244, with non−cerioid, locally discontinuous intercorallite walls; for details of sampling location see McLean 1994b: 80, 81 (A1 transverse, A2 longitudinal section). B. Smithicyathus lacunosus (Gürich, 1896), Panek quarry, GIUS 384P 54, densely phaceloid colony with cerioid (tripartite) intercorallite walls predominating, but also with pseudocerioid segments, mainly between parent and offset corallites close to offsetting area, note also presence of 3rd order septa at some corallite peripheries (B1 transverse, B2 longitudinal section). C. Smithicyathus smithi (Różkowska, 1953), Zgórsko quarry, GIUS 382ZS 14, massive pseudocerioid colony with occasional intercorallite voids, bordered by epithecate walls (C1 transverse, C2 longitudinal section).Scale bar 2 mm for all items.

Fig. 2. Schematical geological map of the Devonian outcrop (shaded) in the Holy Cross Mountains, central Poland (after Dzik 2002: fig. 1A), with the main marine Devonian facies belts marked and with indicated coral−bearing sites mentioned in this paper: G, Grabina; JB, environs of Jaźwica quarry; KD, Kowala road cutting; KK, Kowala quarry; M, Miedzianka; P, Panek; PG, Psie Górki; SW, Sobiekurów; W, Wietrznia; ZS, Zgórsko.

Fig. 1. Geography of the Devonian World, after Scotese and McKerrow (1990: fig. 15) with locations, yielding material for the present study. Locations with Smithicyathus: A, Western Canada; B, Holy Cross Mountains (see Figs. 2, 3); C, Lublin area; D, Moravia.Locations with dubious Smithicyathus: E, New Siberian Islands; F, NE Siberia. Additional locations with comparative material: G, New Mexico; H, Harz Mountains; I, Urals; J, Kuzbass; K, South China.