Rugosa Milne Edwards and Haime, 1850

Fedorowski, Jerzy, Bamber, E. Wayne & Richards, Barry C., 2021, Mississippian colonial tabulate and rugose corals from the Flett Formation, Liard Basin, northwest Canada, Acta Palaeontologica Polonica 66 (3), pp. 679-704 : 689-691

Subclass Rugosa Milne Edwards and Haime, 1850

Order Stauriida Verrill, 1865

Suborder Lithostrotionina Spasskiy and Kachanov, 1971

Family Lithostrotionidae d’Orbigny, 1852

Subfamily Lithostrotioninae d’Orbigny, 1852

Genus Cordilleria nov.

ZooBank LSID: urn:lsid:zoobank.org:act:DA56472A-7C79-4EB9-A8E7-8CC9B68FFE56

Etymology: Named after the occurrence of the type species in the North American Cordillera.

Type species: Diphyphyllum mutabile Kelly, 1942: 358 , pl. 51: 7, 8; “southwest slope of the mountain northwest of the junction of the McLeod River with Whitehorse Creek” ( Kelly 1942: 359), lower Rundle Group, upper Tournaisian. Here designated.

Species included: Diphyphyllum mutabile Kelly, 1942 ; Lithostrotion flexuosum Warren, 1927 ; Lithostrotion (Siphonodendron) warreni Nelson, 1960 ; Lithostrotion (Siphonodendron) oculinum Sando, 1963 .

Diagnosis.—Fasciculate colonial Lithostrotioninae . Major septa amplexoid, elongated on surfaces of tabulae to various extent. Protosepta indistinguishable from remaining major septa. Cardinal fossula absent. Pseudocolumella formed on upper surfaces of tabulae independently from major septa; may be continuous, or underdeveloped to various extent, up to atrophy. Increase lateral with peripheral parts of parents’ septa inherited by offsets (atavosepta). Acolumellate growth of offsets long-lasting, may be permanent up to maturity. Tabularium normal; down sloping of peripheral parts of tabulae differently accentuated, rarely so strong as to form short-lasting columnotheca. Microstructure of septa finely trabecular.

Remarks.—The holotype of Diphyphyllum mutabile Kelly, 1942 , has been the most thoroughly studied and best illustrated of all the North American lithostrotionid specimens and the species itself has been formerly included to three different genera: Diphyphyllum Lonsdale, 1845 , Lithostrotion Fleming, 1828 , and Siphonodendron McCoy, 1849 (e.g., White 1875; Warren 1927; Kelly 1942; Nelson 1960; Bamber 1966; Sando and Bamber 1985; Fedorowski and Bamber 2007). Diphyphyllum mutabile is designated here as the type species for Cordilleria gen. nov. Most of the summary by Fedorowski and Bamber (2007), based mainly on the holotype of D. mutabile , is followed here, allowing a considerable reduction of the remarks that follow. Only the idea of the ancestral position of the North American species included in European genus Dorlodotia Salée, 1910 , by Sando (1969a), Sando and Bamber (1985) and Fedorowski and Bamber (2007), is not followed here. All those North American Dorlodotia- like species are here considered as endemics, belonging to a new, yet unnamed genus, closely related to Cordilleria gen. nov. Fedorowski and Bamber (2007) postulated a position of that unnamed new genus as ancestral for the North American Siphonodendron -like corals. We follow that idea for two reasons: (i) the early growth stages of both “ D. mutabile Kelly, 1942 ”, and “ Dorlodotia sp. ” ( Fedorowski and Bamber 2007: pl. 2: 3 and 7, respectively) are nearly identical; (ii) the occurrence of the oldest “ Dorlodotia sp. ”-like species preceded the appearance of the oldest Cordilleria gen. nov., i.e., “ D. mutabile ”.

The collection from the Flett Formation in the Liard Basin is too geographically and stratigraphically restricted for a more comprehensive discussion on the content and the inter-relationships of the species possibly belonging to Cordilleria gen. nov. A large collection derived from more than 200 localities in western Canada and studied by the authors simultaneously, documents very large morphological variability in all characters of colonies potentially belonging to that genus. Limits of the intraspecific variability and emended diagnoses of these species are impossible to establish using only the Liard Basin colonies. They will be discussed in detail and introduced in the study of the large collection mentioned above. Thus, the total species content of the genus Cordilleria gen. nov. is not established here and descriptions of species that follow are reduced to characters typical for the specimens from the Liard Basin collection. For the reasons mentioned above, all species from the Liard Basin are here temporarily identified as affinis to the nominative species. Such an approach is safer than the application of names when the limits of intraspecific variability of those species have not been established. Brief characteristics of the most important features of the holotypes of species present in the Liard Basin are introduced for comparison. Those characteristics are based on the papers by Kelly (1942), Nelson (1960), Sando (1963), Bamber (1966), Sando and Bamber (1985), and Fedorowski and Bamber (2007). A brief discussion below on some morphological characters of corals identified here as Cordilleria gen. nov. is introduced to illustrate their variability and bearing on species identifications. amplexoid major septa. That characteristic, not named in descriptions published so far, is differently accentuated in particular species. Its bearing on the length of major septa has been recognized by several authors dealing with those corals (e.g., Sutherland 1958; Nelson 1960; Sando 1963; Bamber 1966). To that question Bamber (1966: 8) devoted a good deal of his remarks on “ Lithostrotion (Siphonodendron) sinuo- sum ( Kelly, 1942)”, i.e., Cordilleria flexuosa ( Warren, 1927) in the name accepted here. The amplexoid character of the major septa and the consistent length of the non-amplexoid minor septa, led to wide limits being accepted for the major/ minor septa relationship in all detailed descriptions of the species in the papers listed above. We consider that mutual relationship as diagnostically important, but with the restriction that minor septa are compared to the parts of major septa consistent in their length. Only parts of the amplexoid major septa, i.e., those located between external walls or dissepimentarium on the one side and steep parts of tabulae on the other side, fulfill that criterion. Parts of major septa lengthened to various extent on surfaces of inner parts of tabulae, cannot be used for comparison since their length depends on the advancement of their amplexoid character. They are only slightly elongated when weakly amplexoid as exemplified by C. warreni ( Nelson, 1960) or are elongated along surfaces of tabulae to approach the pseudocolumella when strongly amplexoid as in C. flexuosum . That difference should be omitted when the major/minor septal relationships are compared. Thus, we suggest considering only the non-lengthened sectors of major septa as their referring length.

(ii) The morphology in longitudinal sections and the shape of tabulae in particular, has been accepted by Sando (1963: 1076) as the most important feature distinguishing his new species “ Lithostrotion (Siphonodendron) oculinum ” from the most similar species “ L. (S.) mutabile ” and “ L. (S.) warreni ”. Also, Bamber (1966) considered the morphology in longitudinal section as one of the most important characters allowing the distinction between extremely variable species traditionally included by the North American authors in Lithostrotion (Siphonodendron) . Unfortunately, the strong variability in the development of pseudocolumellae within corallites of particular colonies, varying from almost continuous to non-existent, strongly reflects on the shapes of tabulae, making them similar in different species when the pseudocolumella is absent. Diagenetic alterations, in turn, camouflage the distinction of steep, peripheral parts of tabulae from longitudinal sections of peripheral major septa that can imitate steep parts of tabulae resting upon underlying tabulae. All those taxonomically inconvenient features can be eliminated by thorough investigation of perfectly centric longitudinal sections. Nevertheless, they increase the difficulty in applying the characters of longitudinal sections as diagnostic. It is enough to compare the longitudinal sections of the holotype of “ L. (S.) oculinum ” ( Sando 1963: pl. 145: 4, 5), or the longitudinal sections illustrated by Bamber (1966: pl. 1: 4c–i) to demonstrate those difficulties. The illustrated specimens ( Figs. 9A 2 , A 3 , B 1 , B 2 , 11A 2 , A 3 , B 1 , 13D ) may serve as additional examples. Nevertheless, the morphology in longitudinal section is diagnostically important, but it should be treated in a way less rigid than suggested by Nelson (1960), Sando (1963), and Bamber (1966). This means that the prevailing morphology of several characters rather than a strong single feature should be accepted as diagnostic for a given species.

(iii) The presence or absence of a pseudocolumella is one of the most important characteristics for distinguishing between the European Diphyphyllum Lonsdale, 1845 , and Siphonodendron McCoy, 1849. Following that European distinction, Kelly (1942) followed by Sutherland (1958) and Sando and Bamber (1985) applied the name Diphyphyllum to the lithostrotionids from North America that lack a pseudocolumella or have it extensively reduced. Fedorowski and Bamber (2007) rejected that name by informally using the name Siphonodendron for “ Diphyphyllum mutabile Kelly, 1942 ”. However, the question arises what to do with the North American colonies that totally lack pseudocolumella, but expose all other characters resembling particular species of Cordilleria gen. nov. We now provisionally accept such colonies as belonging to the latter genus.

(iv) The dissepimentarium in most North American lithostrotionids described so far can be characterized as consistent for species. Two main groups can be distinguished: those with a single row of interseptal and mostly regular dissepiments and others developing two or more rows of dissepiments with or without lonsdaleoid dissepiments developed in addition to interseptal ones. However, two or even three rows of dissepiments may for a very short period of growth appear in specimens characterized otherwise by a single, incomplete row of interseptal dissepiments ( Fig. 9A 1 ). We provisionally consider such a multiplied development as a diagnostically unimportant and environmentally controlled event and include its bearers within species overwhelmingly developing only a single complete or incomplete row of interseptal dissepiments.

The name Lithostrotion flexuosum , introduced by Warren (1927) for a species included by him in the genus Lithostrotion Fleming, 1828 , was renamed Lithostrotion sinuousum by Kelly (1942) due to the earlier preoccupation of the former by Lithostrotion flexuosum Trautschold, 1879 . However, that pre-occupation failed when specimens of flexuosum –sinuosum were included here in Cordilleria gen. nov. Thus, the name L. flexuosum Warren, 1927 , is here re-introduced as valid and is applied in this paper instead of L. sinuousum Kelly, 1942 .

Figures

Fig. 9. Rugose corals Cordilleria aff. mutabile (Kelly, 1942), from Flett Formation (Mississipian), Rundle Group, Liard Basin, Northwest Territories, Canada. A. C-52124 = GSC142471, locality 3 (Fig. 2); longitudinal thin sections (A1, A2), transverse thin section (A3). B. C-47930 = GSC-142469, locality 14 (Fig. 2), Flett Formation undivided; longitudinal sections, peels (B1–B3), transverse thin section (B4). Note in B1 steeply arranged tabulae when pseudocolumella present (black arrow), flat tabulae when pseudocolumella absent (white arrow).

Fig. 2. Simplified geological map of study region showing distribution of outcrops of Carboniferous Prophet, Flett, and Mattson formations in southwest corner of Northwest Territories and southeastYukon Territory and lines of cross-sectionA–A’ and B–B’ (see Figs.4, 5 for details). Surface and subsurface sections (corallocalities are in orange):1,Twisted Mountain;2, north end Mattson anticline;3, Jackfish Gap (main coral locality); 4, Tlogotsho Plateau;5, Etanda Lakes; 6, Tika Creek; 7, Pan American Home Signal C.S.P. Celibeta Number 7; 8, Texaco N.F.A. Bovie Lake J-72; 9, Pan American A-1 Pointed Mountain P-53; 10, Canada Southern et al. North Beaver River YT 1-27; 11, Pan American Beaver YT G-01; 12, West Flett anticline; 13, Southern Liard Range; 14, Sheaf Cree.Abbreviations:Fm., Formation (formal unit); fm., formation (informal unit); fms.,formations; S., south; W., west;YT,Yukon Territory;C.S.P., Canada Southern Petroleum. Modified from Harker (1963) and an unpublished map by Andrew Okulitch from 2005 (reproduced with permission).

Fig. 3. A. Partly schematic, palinspastic stratigraphic cross-sectionA–A’ showing stratigraphic relationships of Flett Formation in eastern Cordillera of northern part of project area in southwest corner of Northwest Territories and southeast Yukon. B. Environments of deposition. See Fig. 2 for line of cross-section A–A’.Abbreviations:Fm., Formation (formal unit); fm., formation (informal unit); Mbr., Member (formal unit). Modified from Richards et al. (1989).

Fig. 1. Map showing location of study area, Carboniferous lithofacies assemblages subcropping beneath Permian and Mesozoic deposits in the Western Canada Sedimentary Basin, principal sub-basins, uplifts, and shelves. In the Liard Basin western occurrences of the Rundle and Matson lithofacies assemblages overlie and pass westwards into the Besa River lithofacies assemblage. See Fig. 3 for the formational composition of the Banff, Rundle, and Mattson lithofacies assemblages. Modified from Richards et al. (1994).

Fig. 11. Rugose corals Cordilleria aff. mutabile (Kelly, 1942), from Rundle Group undivided, east-central British Columbia, Canada (A) and Meilleur Member, Flett Formation (Mississipian), Rundle Group, Liard Basin, Northwest Territories, Canada (B–D). A. C-07349 = GSC 142468, 8.05 km NW of Bone Mountain, east-central British Columbia; transverse thin section (A1), longitudinal thin sections (A2,A3), pseudocolumella (A4, black arrows) created on tabula surface (white arrow) and tabulae/pseudocolumella relationships in lower part of picture. B. C-52546 = GSC 142470, locality 3 (Fig. 2); longitudinal thin section (B1), transverse thin section (B2). C. C-47930 = GSC 142469, locality 2 (Fig. 2); transverse thin section; very early growth stage of offset. D. C-52124 = GSC 142471, locality 3 (Fig. 2); thin section of three offsets in different growth stage; one cut longitudinally, two transversally.

Fig. 13. Rugose coral Cordilleria aff. oculinum (Sando, 1963), from Meilleur Member, Flett Formation (Mississipian), Rundle Group, Liard Basin, Northwest Territories, Canada (A, C, F), Tlogotsho Member, Tlett Formation (Mississippian), Rundle Group, Liard Basin, Northwest Territories, Canada B, D), and upper Prophet Formation (Mississippian), Roundle Group, Northwest Territories, Canada (E). A. C-52268 = GSC-142473, locality 2 (Fig. 2); transverse thin section (A1), longitudinal thin section (A2). B. IG-16 = UAM-Tc.Can./5 B-D, locality 3 (Fig. 2); longitudinal section, peel (B1), transverse thin section (B2). C. IG-85 = UAM-Tc.Can./6, locality 3 (Fig. 2); longitudinal section (peel). D. IG-30 = UAM-Tc.Can./7, locality 3 (Fig. 2); longitudinal section, peel (D1), transverse thin section (D2). E. C-52205 = GSC-142472, locality 2 (Fig. 2); transverse thin section. F. C-52361 = GSC-142474, locality 13 (Fig. 2); transverse thin section.