Summary account of the Lower Ordovician (Arenig) graptolite biostratigraphy of the Cow Head Group, western Newfoundland

S. HENRY WILLIAMS AND ROBERT K. STEVENS

U Williams, S. H. and Stevens, R. K.: Summary account of the Lower Ordovician (Arenig) graptolite bio­ stratigraphy of the Cow Head Group, western Newfoundland. Bull. geol. Soc. Denmark, vol. 35, pp. 259- 270, Copenhagen, July 1st, 1987.

Cj The Cow Head Group is an allochthonous sequence of Middle Cambrian to late Arenig sedimentary brec­ cias, limestones and shales deposited in a deep lower slope environment close to a continental margin. Im­ bricate thrusting has resulted in repeated exposure of laterally equivalent "proximal" to "distal" facies which may be correlated using graptolitic control in the interbeds. "Proximal" sections are characterised by massive, coarse breccias with interbedded limestones and green/dark grey shales. More distal ex­ posures have fewer and thinner breccias and limestones, while the green/grey shales are replaced pro­ gressively by red, non-graptolitic ones. Although the succession is by no mean unbroken or complete, it furnishes one of the best and most con­ tinuously graptolitic sections through the Arenig. A new zonal scheme is erected for the Cow Head Group, which could prove suitable as a new North American standard. Furthermore, several limestones and siliceous shales have yielded exquisitely preserved isolated material, permitting integration of fine growth detail with complete flattened specimens. With the exception of the uppermost Arenig U. austrodentatus Zone, Arenig graptoloids possess a pro- sicular origin for thl1. The earliest graptoloids with a metasicular origin for the first theca appear in this zone, including Undulograptus, Cryptograptus and Paraglossograptus. This interval, equivalent to Dal of the Australasian scheme, therefore represents a hitherto unrecorded major evolutionary step in graptolite evolution.

S. Henry Williams and Robert K. Stevens. Dept. of Earth Sciences, Memorial University of Newfoundland, St. John's, Nfld, Canada A1B 3X5, August 29th, 1986.

The Cow Head Group is composed of a series of slide breccias, while Johnson (1941) regarded the coarse sedimentary breccias or conglomerates sequence as a broken but relatively complete and with interbedded limestones and shales, together simple one. Oxley (1953) returned to the ideas of with subordinate silts, cherts and dolostones. It Schuchert and Dunbar, concluding the breccias ranges in age from late Middle Cambrian to late to be "thrust breccias" and that the intervening Arenig. The base is nowhere seen, but the unit is limestone-shale beds were large blocks within overlain by green sandstones of the Lower Head these units. Kindle and Whittington (1958) fi­ Formation (James and Stevens 1986), considered nally recognised the Cow Head Group as having to be of turbidite origin (Stevens 1970). a regular sedimentary origin with an age from Much discussion on the Cow Head Group has Middle Cambrian to Middle Ordovician. They been centred around the origin of the breccias. numbered the beds from 1 to 14, an informal sys­ Logan (1863) considered them to be part of a tem still used with the addition of a Bed 15 not simply bedded sedimentary sequence. Dawson exposed at Cow Head. The breccias are now (1883) and Walcott (1894) inferred that they re­ thought to have been deposited through a series sulted from the action of sea ice on a shallow wa­ of debris flows, bringing blocks of shallow water ter carbonate sequence. Schuchert and Dunbar carbonates into a deep water setting (James and (1934) offered an entirely different view; they Stevens 1986). They also eroded the underlying considered that some breccias were tectonic, deep water sediments to a greater or lesser extent while others were derived from the erosion of ad­ which now form the bulk of incorporated clasts, vancing submarine thrust sheets. Barrington especially in the more distal exposures. Brown (1938) compared them to sedimentary The Cow Head Group is considered allochtho- 260 Williams and Stevens: Summary of the Cow Head Group nous; it was probably transported at least 150 km History of graptolite research from the north-east during the Taconic Orogeny (Rodgers and Neale 1963; Stevens 1970 - see Graptolites were first collected from the Cow James and Stevens 1986 for discussion). It now Head Group, together with other fossils, by Ri­ rests in proximity to stratigraphically equivalent chardson in the 1860's during early work by the autochthonous shallow water carbonates of the Geological Survey of Canada, which were subse­ St. George and Table Head Groups. A second al- quently incorporated by Billings in Logan's "Ge­ lochthonous unit, the Curling Group, is also pres­ ology of Canada" (1863). These specimens were ent; it is considered more distal in origin and is among collections of Canadian material sent to mostly older than the Cow Head Group. The Lapworth at Birmingham University for identi­ Lower Paleozoic sediment pile is overlain fication. Although they were studied the results through thrust contact by Grenville gneiss and do not appear to have been used or published. granite which forms the core of the Long Range Schuchert and Dunbar (1934) made extensive fau­ Mountains (fig. 1). nal collections, but the first critical conclusions The imbricate nature of the Cow Head Group on biostratigraphy were based on graptolites col­ and overlying Lower Head Formation exposures lected by Johnson's field assistant Kindle. Rue- is of uncertain genesis. The sequence has never demann studied these in detail and incorporated been buried deeply; this is demonstrated by un- the results in his "Graptolites of North America" deformed macrofossils and low conodont colour published in 1947. He included extensive faunal alteration index of 2.5, making emplacement as a lists, together with a few illustrations and new subophiolitic accretionary prism extremely un­ taxa based on the Cow Head material. likely. The structure does, however, allow an in­ Kindle and Whittington (1958) were the first to sight into the transition of a sedimentary environ­ give a critical stratigraphical account of the grap­ ment from lower continental slope to toe of slope tolite faunas. They compared the succession with in an approximately west-east direction. Most the Australian standard and proposed a correla­ "proximal" sections, such as Lower Head and tion between the graptolitic and North American Cow Head, are characterised by massive poly- shelly zonal schemes. Our work has involved ex­ mict breccias, abundant limestone-shale rhyth- tensive recollecting through the Arenig at a lim­ mites and green/grey/black shale intervals. Dis- ited number of localities (fig. 1). This has per­ tally the breccias decrease in thickness, number, mitted critical taxonomic revision, incorporating clast size and variety of composition, while lime­ isolated, three-dimensional material, and the stones and green/grey shales become scarcer and erection of a new zonal scheme with eight divi­ red shale predominates. This suggests that the sions through the Arenig. oxygen minimum existed at some distance up- slope, oxygen content increasing with depth. Correlation of the sections is made possible Completeness of the Arenig graptolite through graptolites present in the interbeds, and succession conodonts in limestones. Coeval shallow water shelly faunas occur in limestone clasts in the brec­ The Cow Head Group contains possibly the most cias, permitting integration of several biostrat- complete and continuously graptolitic Arenig se­ igraphical schemes (Kindle and Whittington quence in the world. Both faunal and sedimen­ 1958). Previous faunal studies have concentrated tary breaks occur, but these are clearly recognis­ on the trilobites (e.g. Whittington 1963) and con­ able and well understood due to the number of odonts (e.g. Fåhraeus and Nowlan 1978). Ongo­ sections studied. The most obvious erosion oc­ ing research is centred on the conodonts (C. R. curs at the bases of several major breccia hori­ Barnes et al.) and graptolites (S. H. Williams, zons, notably the bases of Beds 10 and 12 at Cow R. K. Stevens and B.-D. Erdtmann) from the Head and Bed 14 at Lower Head, where the Upper Cambrian and Ordovician. cross-cutting relationship with underlying lime­ stone-shale lithologies is unambiguous. Less no­ ticeable, but equally important breaks occur throughout the interbeds. Some erosive intervals Bulletin of the Geological Society of Denmark, vol. 35 1987 261 are indicated by channelling and phosphorite well to bleaching treatment to reveal growth fu- conglomerate horizons, while others may be in­ sellae. The combination of abundant flattened or ferred through absence of lithological or biostrat- partially flattened non-isolated graptolites with igraphical intervals present at neighbouring lo­ flattened bleached specimens and three-dimen­ calities. Thus the thickness of Bed 11 on the sional unbleached ones has allowed an unique in­ Ledge section at Cow Head is only one half to tegration of overall form, growth structure and two thirds the total thickness of that at Jim's fine fusellar detail from the one stratigraphic Cove only a few hundred metres along strike. unit. Likewise beds representing the T. akzharensis Many lithologies are somewhat atypical for Zone thin from about 5 m at the eastern part of the graptolite preservation; the most fossiliferous lith- Cow Head peninsula to 0.5 m at the Ledge sec­ ology is a dark grey, slightly brown, silty shale, tion, a present distance of little more than one which is commonly interbedded with either dark kilometre. Successions in the down slope sections limestone or pale green shale. Black, very fine­ are probably more complete, but are characterised grained paper shales with a somewhat "soapy" by larger proportions of red and green non-grap- texture occasionally yield fragments but are mostly tolitic shales. non-graptolitic. They perhaps represent very slow The picture is complicated further by ubiquitous background deposition which was insufficiently but subtle tectonic deformation. Much cross-fault­ rapid to bury the organic skeletons prior to de­ ing and parasitic folding is obvious, but a large composition. Other graptolitic lithologies include number of low angle faults substantially alter the parted limestones, where the rhabdosomes occur interbed successions and are visible only through mostly on the limestone-shale boundaries, and in detailed critical investigation. A progressive grainstones, where they are normally preserved change from early low angle through higher angle too poorly to be of much taxonomic value. The faulting is apparent, culminating in late-stage, ex- three-dimensional graptolites from St. Paul's and tensional block faulting. Much of the early step Western Brook Pond occur in pale brown, con­ faulting which almost parallels bedding may have cretionary limestones. These represent early di- been syn-sedimentary, related to downslope agenetic precipitation of carbonate along black movement of the sedimentary pile. This is consist­ graptolitic shale bands prior to compaction. ent with the faulting's subtle and commonly Many such nodular bands clearly grade laterally masked relationship within less competent shale into graptolitic shale yielding flattened speci­ lithologies. It would also explain why this style of mens. It is unclear why such horizons should be deformation is most common in the more prox­ restricted to these two localities when similar imal localities on Cow Head where the slope was limestones abound at other sections, particularly probably steeper. Cow Head. Despite these complexities, the Cow Head Group apparently represents one of the world's most complete and thoroughly documented grap- Arenig biostratigraphy of the Cow tolitic successions through the Tremadoc and Are- Head Group nig. Although graptolite occurrence is not contin­ uous, all zonal intervals are represented by several The following summary is preliminary, while sec­ horizons and the large number of exposed sections tions shown (fig. 1) are composite and schematic. permits a more-or-less complete composite picture Fully detailed measured sections and accurate to be compiled. Additionally, some localities, in range charts with zonal definitions, together with particular St. Paul's Inlet and to a lesser extent extensive taxonomic revision, will be given by Western Brook Pond, yield limestones with Williams and Stevens (in press). A selection of three-dimensional graptolites which have been representative taxa is illustrated in figs. 2-4. The isolated successfully using acetic acid. Siliceous new zonal scheme is based partially on the Aus­ shales at both these localities and at Martin Point tralasian standard (see VandenBerg 1981), es­ (North) have also been digested using hydro­ pecially for the isograptid intervals, but is defined fluoric acid, revealing flattened, but equisitely on the Cow Head sections. It is essentially, there­ preserved isolated material that has responded fore, a local biostratigraphic scheme, but is more 262 Williams and Stevens: Summary of the Cow Head Group

PROXIMAL BRITISH -,„„,..,,. , Bed Number SERIES Graptohts Zone (afler K|nd|e & . Whittington 1958) 7LLANVIRN '<*££$* U. austrodentatus J^'il'i ,K?1 15„

ST PAUL'S INLET Bulletin of the Geological Society of Denmark, vol. 35 1987 263

DISTAL

m£

:r:^i

:TT77?71

53 —

/ WESTERN MARTIN / BROOK POND POINT GREEN 20 POINT

10

L0 Fig. 1. Schematic, composite sections through the Cow Head Group. Lithological columns adapted from James and Stevens (1986). Correlation lines are based solely on graptolitic biostratigraphical data. 264 Williams and Stevens: Summary of the Cow Head Group

Fig. 2. Representative graptolite taxa of the Cow Head Group with biostratigraphical ranges. A. Tetragraplus appwximaius Nicholson (x2.5). GSC 79725. 7". approximatus-T, akzharensis zones. B. Tetragraplus acclinans Keble (X2.5). GSC 79885. T. approximatus-T. akzharensis zones. C. Tetragraplus serra (Brongniart) (X2.5). GSC 79687. T. akzharensis-T v. maximus zones. D. Tetragraplusphyllograptoides Strandmark (X2.5). GSC 81473, T. akzharensis Zone. E. Pendeograptus cf. pen­ dens (Elles) (x5). GSC 79672. T. approximatus-P. fruticosus zones. F. Tetragraplus akzharensis Tjaz (X2.5). GSC 82067. T. akz­ harensis Zone. G. Didymograpius (Expansograptus) constrictus (J. Hall) (X5). GSC 79677. T. akzharensis Zone. H. Didymograplus (Expansograptus) latus T. S. Hall (x5). GSC 79727. T. approximatus-T. akzharensis zones. I. Pendeograptus fruticosus (J. Hall) (X2.5). GSC 79626. P. fruticosus-D. bifidus zones. J. Pseudophyllograptus sp. nov. (X2.5). GSC 79897. T. akzharensis Zone. K. Di- chograptus octobrachiatus (J. Hall) (X2.5). GSC 79655a (latex peel). P. fruticosus Zone. L. Sigmagraprus praecursor Rucdemann (xl). GSC 79582. P. fruticosus-l. v. lunatus zones. Bulletin of the Geological Society of Denmark, vol. 35 1987 265

Fig. 3. Representative graptolite taxa of the Cow Head Group with biostratigraphical ranges. A. Didymograptus (Expansograptus) similis (J. Hall) (X5). GSC 79762. T. akzharensis-D. bifidus zones. B. Xiphograptus formosus svalhardensis (Archer and Fortey) (X10). GSC 81705. P. fruticosus-l. v. maximus zones. C. Didymograptus (Expansograptus) nitidus (J. Hall) (X5). GSC 79608. P. fruticosus zone. D. Tetragraptus bigsbyi (J. Hall) (x5). GSC 79605. P. fruticosus-l. v. lunatus zones. E, Didymograptus (Expansograptus) pennatulus (J. Hall) (X2.5) GSC 79639. P. fruticosus Zone. F. Didymograptus cf. hemicyclus Harris (x5). GSC 81514. D. bifidus-1. v. maximus zones. G. Phyllograptus typus J. Hall (X2.5) GSC 79552a (Jatex peel). P. fruliC0SUS-I. V. victoriae zones. H, 1. Didymograptus (Didymograptellus) bifidus (J. Hall). D. bifidus Zone. H. GSC 81699 (xlO). I. GSC 81752 (x2.5). J. Isograptus cf. primulus Harris (xlO). GSC 81586. D. bifidus Zone. K. Isograptus victoriae victoriae Harris (x5). GSC 81731. /. v. victoriae Zone. L. Goniograptus thureaui (M'Coy). (x2.5). GSC 79628 (latex peel). P. fruticosus-l. v. maximus zones. M. Isograptus victoriae lunatus Harris (x5). GSC 81719. /. v. lunatus Zone. 266 Williams and Stevens: Summary of the Cow Head Group

i«

Fig. 4. Representative graptolite taxa from the Cow Head Group with biostratigraphical ranges. A. Isograptus victoriae maximus Harris (X2.5). GSC 81429. /. v. maximus Zone. B. Isograptus victoriae divergens Harris (x2 5) GSC 81458. /. v. maximus-U. auslrodentatus zones. C. Pseudotrigonograptus ensiformis (J. Hall) (X2.5). GSC 81436. /. v. lunatus-U. au­ strodeniatus zones. D. Isograptus caduceus australis Cooper (X2.5). GSC 81444. /. v. maximus-U. austrodeniatus zones. E. Isograptus cf. forcipiformis Ruedemann (x5). GSC 81624b. /. v. maximus Zone. F. Undulograptus austrodeniatus auslrodentatus Harris and Ke- ble)(x5). GSC 81566. U. auslrodentatus Zone. G. Paraglossograptus tentaculatus (J. Hall)(x5). GSC 81552. U. auslrodentatus Zone. H. Isograptus victoriae maximus Harris (X5). GSC 81532. /. i». maximus Zone. I. Pseudisograptus dumosus Harris (x5). GSC 81630. /. v. maximus-U. auslrodentatus zones. likely to be relevant to other Canadian graptolitic Arenig is divided into eight graptolite zones; lat­ units than the "borrowed" sequence given by est geochronological estimates for this interval Barnes et al. (1981). The Cow Head zonation suggest an order of 15-20 million years (Gale et therefore has great potential as a new Canadian, al, 1980; McKerrow in Kokellar et al. 1984, p. and possibly North American, standard. The 247). The average duration of 2-2.5 million years Bulletin of the Geological Society of Denmark, vol. 35 1987 267

per graptolite zone falls short of the estimated stipes than seen in the typically "droopy" pen­ one million years or less for the late Ordovician dent form of P. fruticosus; it is uncertain whether and Silurian (Rickards 1975), but this is far more these early examples consistently possess four sti­ precise than provided by either shelly macrofos- pes as recorded in the Australasian sequence sils or conodonts and more reliably correlated. (Thomas 1960; VandenBerg 1981). The upper­ Because of the similarity with the Australasian most portion of the T. approximatus Zone at graptolite zonal scheme, accurate correlation is Cow Head is characterised by common "Rhabdi- achieved readily. Comparison with zonal succes­ nopora" - like dendroids in rather pale ribbon sions elsewhere is more tentative, with the partial limestones. This is considered to be an environ­ exception of Spitsbergen (Cooper and Fortey mentally, rather than biostratigraphically con­ 1982), and requires fuller discussion than is pos­ trolled assemblage. sible within the present paper; this will be in­ The first major increase in faunal diversity oc­ cluded by Williams and Stevens (in press). curs at the top of Bed 9 at Cow Head, with the We follow the standard practise adopted by first occurrence of Pseudophyllograptus (prob­ graptolite workers in correlating the latest pre- ably a new species), D. (Expansograptus) con- graptoloid burst of dendroids with the top Trem- strictus (J. Hall), Tetragraptus serra (Brongniart) adoc Series, although such an assumption may and Tetragraptus akzharensis Tzaj, together with apparently not be proved in the standard British a rich nematophorous and rooted dendroid as­ sections. The topmost Tremadoc is characterised semblage. The zone index fossil T. akzharensis by a diverse dendroid fauna, with possibly a few was described from the T. approximatus Zone of true graptoloid elements. This is here provision­ Kazakhstan, U.S.S.R. by Tzaj (1974). The Cow ally referred to an "Adelograptus Zone", corre­ Head material of this species is, however, far sponding to the Adelograptus victoriae Zone of more abundant and complete. The T. akzharen­ Australasia (Lancefieldian Stage 2). Following sis Zone varies from 0.5 m thick at the Ledge sec­ this is a non-graptolitic interval in all sections; at tion of Cow Head to ca. 5 m below the Cow Head the more proximal locality of Cow Head this cemetery some 1 km east along strike. At the in­ level is represented by a ca. 3 m bed of thinly in- tervening locality of Jim's Cove the interval is not terbedded dolomites and very pale green shales represented at all. A similar fauna occurs above before the onset of darker grey, graptolitic lime­ the thick breccia referred to Bed 10 at St. Paul's stones and shales yielding a Tetragraptus approx­ Inlet; the zone is also represented at Western imate Zone fauna. Elsewhere this dolomitic unit Brook Pond and possibly Martin Point. At St. is overlain by red, unfossiliferous mudstone Paul's the assemblage additionally includes a new reaching anywhere up to 50 m thick before the subspecies of T. approximatus and Tetragraptus first T. approximatus Zone faunas are encoun­ phyllograptoides Strandmark, while at Western tered. It is here assumed that the T. approx­ Brook Pond rare D. (Expansograptus) similis imatus Zone is equivalent to the earliest Arenig (J. Hall) and D. (E.) ensjoensis Monsen are pres­ (i.e. Lancefieldian 3 of Australasia); at Cow ent. Head and St. Paul's Inlet it yields T. approx­ This biostratigraphic interval separating the ty­ imatus (Nicholson) and small Pendeograptus cf. pical P. fruticosus "Lone, from a T. approximatus pendens (Elles), together with Clonograptus flex- assemblage is considered sufficiently distinct to ilis (J. Hall). This assemblage remains constant define as a new zone. The T. akzharensis Zone throughout much of the zone; towards the upper probably corresponds at least in part to the Aus­ part the earliest Didymograptus, D. (Expanso- tralian "4 - branched fruticosus Zone" (i.e. Ben- graptus) latus T. S. Hall appears at St. Paul's In­ digonian 2), but is defined more accurately and is let and Western Brook Pond, although it has not probably of greater potential for correlation. been recorded from Cow Head. At the latter lo­ The following Pendograptus fruticosus Zone cality the later T. approximatus Zone yields Te­ has an extremely rich and diverse fauna; in addi­ tragraptus acclinans Keble and larger, more ro­ tion to the zone index fossil the assemblage in­ bust specimens of P. cf. pendens reminiscent of cludes abundant Dichograptus octobrachiatus Pendeograptus fruticosus (J. Hall). These how­ (J. Hall), Tetragraptus serra (Brongniart), Didy­ ever have straighter and more sharply divergent mograptus (Expansograptus) nitidus (J. Hall), D. 268 Williams and Stevens: Summary of the Cow Head Group

(E.) extensus (J. Hall), D. (E.) similis (J. Hall), (Expansograptus) species and gradual replace­ D. (E.) ensjoensis Monsen, D. (E.) pennatulus ment by Xiphograptus, which is present but rare (J. Hall), Phyllograptus typusJ. Hall, Sigmagrap- in earlier levels. Both within this and later zonal tus praecursor Ruedemann, Goniograptus thu- intervals, most representatives of the genus may be reaui (M'Coy) and Holmograptus cf. leptograp- assigned to X. formosus svalbardensis (Archer toides (Monsen). D. (E.) pennatulus and the sig- and Fortey) (see also Cooper and Fortey 1982). magraptids appear to be restricted to the upper The earliest populations consist entirely of hori­ part of the zone. The assemblage is most diverse at zontal rhabdosomes, with an ever increasing pro­ Cow Head; indeed the P. fruticosus Zone is only portion of slightly deflexed forms from the suc­ found elsewhere at Lower Head and Western ceeding /. v. lunatus Zone upwards. P. fruticosus Brook Pond, the interval being represented by (both two and three-stiped forms) is relatively red, non-graptolitic shale at other localities. The common in the lower part of the D. bifidus Zone, P. fruticosus Zone of the Cow Head Group corre­ but disappears by the upper part. Sigmagraptids lates with Bendigonian stages 3 and 4 of Aus­ remain abundant, beautifully preserved examples tralasian. The Newfoundland succession shows having been isolated from limestones at St. Paul's little, if any, transition from early P. fruticosus Inlet and Western Brook Pond. with four stipes to late ones with two, and we con­ Rare specimens of Isograptus cf. primulus Har­ sider such a distinction unsuitable for biostrat- ris occur at Cow Head in strata which we assign to igraphic division. the D. bifidus Zone, while at other localities "Di­ The earliest pendent didymograptids are repre­ dymograptus cf. hemicyclus Harris" of Cooper sented by a few minute, slender forms found to­ (1973,1979) appears at an equivalent horizon. Al­ gether with D. (E.) pennatulus at the Jim's Cove though this interval might be separated as a dis­ section on Cow Head. Didymograptus (Didymo- crete zone, insufficient fauna has been collected to graptellus) bifidus (J. Hall) itself occurs soon after justify such a step at the present time. The base of in a sudden burst, associated with two and three- the /. v. lunatus Zone (equivalent to Castlemai- stiped forms of P. fruticosus. This assemblage is nian 1 of Australasia) is marked by the appear­ present at Cow Head, St. Paul's Inlet, Western ance of Isograptus victoriae lunatus Harris, which Brook Pond and Martin Point, and also charac­ appears at the top of Bed 11 on the Cow Head terises the type locality of D. (D.) bifidus at Levis, peninsula. This zone also contains abundant Xi­ Quebec. It is therefore quite clear that the D. bi­ phograptus formosus svalbardensis (mostly hori­ fidus Zone directly follows that of P. fruticosus, zontal forms) and Dichograptus solidus Harris removing the need for an intervening "D. proto- and Thomas, which occurs only rarely in the pre­ bifidus Zone" as adopted for Canada without ceding D. bifidus Zone. A rich /. v. lunatus Zone comment by Barnes et al. (1981). The taxonomic fauna is present in siliceous shales at Martin Point complexities involving the type material of these (North), below a massive breccia referred to Bed taxa will be discussed fully by. Stevens and Wil­ 12 by James and Stevens (1986). The grapto- liams (in press.). Here we merely point out that the lites are flattened, but respond well to isolation by holotype of D. protobifidus Elles from Outerside, hydrofluoric acid and bleaching with concentrated northern England is probably late Arenig or early nitric acid and potassium chlorate. The assem­ Llanvirn, and repeat that the equation of the Brit­ blage includes abundant /. v. lunatus, Pseudotri- ish Llanvirn "D. bifidus" with the lower Arenig gonograptus ensiformis (J. Hall), Xiphograptus types from Levis by Berry (1960; 1968) is mis­ formosus svalbardensis, Kinnegraptus kinnekul- taken (see Jaanusson 1960; Skevington 1963; lensis Skoglund and Didymograptus (s.l.) mendi- Bergstrom and Cooper 1973; Riva in Fischer cus Keble and Harris. This last species is a pecu­ 1977). The D. bifidus Zone of the Cow Head liar pendent didymograptid with apparently folded Group correlates with the Chewtonian Stage (D. thecal apertures and has an asymmetric, almost protobifidus and T. fruticosus, and D. protobifi­ "artus - like" development which Cooper and For­ dus zones) of Australasia. tey (1982) considered characteristic of Llanvirn The fauna of the D. bifidus Zone is largely simi­ didymograptids. lar to that of the preceding P. fruticosus Zone; the The succeeding /. v. victoriae Zone includes major change is a decreasing abundance of D. abundant examples of the zone index fossil, Pseu- Bulletin of the Geological Society of Denmark, vol. 35 1987 269

disograptus hastatus (Harris) and Pseudotrigono- dumosus (Harris), Pseudotrigonograptus ensifor­ graptus ensiformis (J. Hall). It correlates with mis (J. Hall), Xiphograptus formosus cf. for­ Castlemainian 2 of the Australasian scheme. The mosus (Bulman) and Holmograptus sp. nov. Flat­ /. v. maximus Zone is well represented at almost tened examples of this fauna have been success­ all studied sections. We originally considered the fully isolated and bleached from siliceous shales at large isograptid present to be /. v. maximodi- St. Paul's Inlet; although not particularly well pre­ vergens Harris but R. A. Cooper (pers. comm.) served, they do permit observation of proximal states that most lie within his defined population of development in this distinctive assemblage. /. v. maximus Harris. The associated assemblage 1 is second in abundance and diversity only to the P. Origin of thl and the Arenig-Llanvirn fruticosus Zone. Besides the zone index fossil it boundary includes Pseudisograptus hastatus (Harris), Pseu- disograptus dumosus (Harris), "Isograptus cf. During the present study, many Arenig grapto- forcipiformis (Ruedemann)" (of Cooper 1973; lite taxa have been isolated successfully. Al­ 1979), Pseudotrigonograptus ensiformis (J. Hall), though three-dimensional specimens were unable Tetragraptus serra (Brongniart), Xiphograptus sp. to withstand bleaching without collapse, flat­ nov. (= D. v-deflexus Harris of Cooper 1979), X. tened material from siliceous shales reponded formosus svalbardensis (Archer and Fortey), Go- well to such treatment. This permitted observa­ niograptus sp. and other sigmagraptids. The later tion of three-dimensional form both by reflected part of the zone also contains Isograptus caduceus light and scanning electron microscopy and of fu- australis Cooper and early specimens of /. v. di­ sellar structure in flattened specimens through vergens Harris, while rare specimens of Pseudi- transmitted light examination. Species isolated sograptus(?) tau (Harris) and P(?) aggestus (Har­ include Tetragraptus approximate, T. bigsbyi, T. ris) are present at St. Paul's Inlet and Western serra, Pendograptus cf. pendens, P. fruticosus, Brook Pond. Further work may allow the subdivi­ Didymograptus (Didymograptellus) bifidus,D.(Ex- sion of the /. v. maximus Zone into two, based on pansograptus) similis, Phyllograptus typus, Xi­ presence or absence of these taxa. The lowest part phograptus formosus svalbardensis, Isograptus of the /. v. maximus Zone is equivalent to Cas­ victoriae lunatus, I. v. victoriae, Kinnegraptus tlemainian 3 of the Australasian sequence. The kinnekullensis, Holmograptus cf. leptograptoides, distinctive zone index fossils characteristic of the Sigmagraptus praecursor, Undulograptus austro­ | following Yapeenian Stage are not, however, pres­ dentatus austrodentatus, U. austrodentatus amer­ ent in the Cow Head Group; we consider their ab­ icanus and Paraglossograptus tentaculatus. sence due to geographic, environmental or preser- One observation is clear; except for those from vational control rather than to a stratigraphical the uppermost Arenig U. austrodentatus Zone, non-sequence. By comparison with the remaining all graptoloids in the Arenig have a prosicular or­ faunal assemblage, the upper part of the /. v. max­ igin for the first thecae (with the one possible ex­ imus Zone is probably equivalent to Yapeenian 1, ception of H. cf. leptograptoides). Only with the while the unfossiliferous portion of high Bed 14 incoming of the austrodentatus fauna (Dal equiv­ may correlate with Yapeenian 2. alent) do graptoloids such as Undulograptus, The latest Arenig Undulograptus austrodentatus Cryptograptus and Paraglossograptus have a met- Zone (equivalent to Darriwilian 1 of Australasia) asicular origin for thl1. Further study on isolated is not exposed at Cow Head itself, but is well de­ material from equivalent successions elsewhere is veloped in strata underlying and within the basal now required in order to establish whether this green sandstones of the Lower Head Formation at event is cosmopolitan and synchronous with the Lower Head, St. Paul's and Western Brook Pond. appearance of the diplograptids. It would also be The assemblage includes U. austrodentatus aus­ of value to find whether all genera (such as the trodentatus (Harris and Keble), U. austrodentatus isograptids and didymograptids) which survive cf. americanus (Bulman), Cryptograptus cf. ante- the Arenig-Llanvirn transition evolve a metasic- nnarius (J. Hall), Paraglossograptus tentaculatus ular origin for thl1, or whether they retain their (J. Hall), Isograptus victoriae divergens Harris, /. prosicular origin. At present the Arenig-Llanvirn caduceus australis Cooper, Pseudisograptus boundary has yet to be defined formally and cor- 270 Williams and Stevens: Summary of the Cow Head Group

relation with the graptolite succession is unclear. Cooper, R. A. 1979: Ordovician geology and graptolite faunas Future definition is likely to be based on shelly of the Aorangi Mine area, north-west Nelson, New Zea­ land. Palaeont. Bull. N.Z. geol. Surv. 47, 1-127. faunas, but should graptolites be considered a Cooper, R. A. and Fortey, R. A. 1982: The Ordovician grapto­ viable alternative, the onset of metasicular origin lites of Spitsbergen. Bull Brit. Mus. nat. Hist. (Geol.) 36, 1 157-302. for thl , together with the appearance of the Di- Dawson, J. W. 1883: The Quebec Group. In Harrington, B. J.: plograptidae, might prove a suitably significant The life of Sir W. E. Logan., 404-414. Montreal. basis on which to define the boundary. Fåhraeus, L. E. and Nowlan, G. S. 1978: Franconian (late Cambrian) to early Champlainian (middle Ordovician) conodonts from the Cow Head Group, western Newfound­ Acknowledgments. We are grateful to NSERC, NRC and Me­ land. Jl. Paleont. 52, 444-471. morial University of Newfoundland for financial assistance. Fischer, D. W. 1977: Correlation of the Hadrynian, Cambrian and Ordovician rocks in New York State. Univ. State New York, Map and Chart Series 25. Gale, N. H., Beckinsdale, R. D. and Wadge, A. J. 1980: Dis­ Dansk sammendrag cussion of a paper by McKerrow, Lambert and Cham­ berlain on the Ordovician, Silurian and Devonian time scale. Earth Plan. Sci. Lett. 51, 9-17. Cow Head Group er en allochton sekvens omfattende mellem Jaanusson, V. 1960: On the series of the Ordovician System. kambriske til sent arenig sedimentære breeder, kalksten og ski­ Rep. XXI Int. Geol. Congr. Pt. 7, 70-81. fre, der er aflejrede på de dybere dele af kontinental-skrænten Jackson, D. E. 1962: Graptolite zones of the Skiddaw Group in tæt ved en kontinent grænse. Blok-forkastninger har haft til re­ Cumberland, England. Jl. Paleont. 36, 300-313. sultat, at der findes gentagne blotninger af identiske niveauer, James, N. P. and Stevens, R. K. 1986: Stratigraphy of the Cow men sådan, at de blottede lag stammer fra både proximale og Head Group, western Newfoundland. Bull. Geol. Ass. distale facies (dvs. aflejringer fra tæt ved sedimentkilden og Can. 366, 1-143. fjernere derfra). De forskellige facies kan korreleres gennem Johnson, H. 1941: Paleozoic lowlands ot northwestern New­ deres indhold af graptolitter. De proximale aflejringer består af foundland. Trans. N.Y. Acad. Sci. ser. II 3, 141-145. massive, grove breccier med mellemlejrede kalksten og grønne Kindle, C. H. and Whittington, H. B. 1958: Stratigraphy of the til mørkegrå skifre. De mere distale aflejringer viser færre og Cow Head region of western Newfoundland. Bull. geol. tyndere breccier og kalksten og de grønne til grå skifre bliver Soc. Am. 69, 315-342. afløst af røde graptolit-fri skifre. Kokelaar, B.P., Howells, M.F., Bevin, R. E., Roach, R. A. Selvom lagfølgen ikke er komplet, er den, på trods heraf, en and Dunkley, P. N. 1984: The Ordovician marginal basin af de bedste og mest komplette serier gennem arenig perioden. of Wales. Spec. Pap. geol. Soc. Land. 16, 245-270. Der etableres en zonering af Cow Head Group, som kan vise Logan, W. E. 1863: Geology of Canada; Geological Survey of sig at være brugbar som ny nordamerikansk standard. Der ud Canada, report of progress from its commencement to 1863, over har adskillige kalksten og kiesel-rige skifre givet usædvan­ 983 pp. ligt velbevaret fossil materiale, der har tilladt studier af finere Oxley, P. 1953: Geology of Parsons Pont - St. Paul's area, west vækst-detaljer. coast, Newfoundland. Rep. Nfld. geol. Surv. 5, 1-53. På nær i det aller-øverste arenig (U. austrodentatus Zonen) Rickards, R. B. 1975: Palaeoecology of the Graptolithina, an besidder alle arenig graptoloider en prosicula oprindelse for extinct class of the phylum Hemichordata. Biol. Rev. 50, thi'. De første graptoloider med en metasiculær oprindelse af 397-436. den første theca fremkommer i denne zone og omfatter Undu- Rodgers, J. and Neale, E. R. W. 1963: Possible "Taconic" klip- lograpius, Cryptograptus og Paraglossograptus. pen in western Newfoundland. Am. Jl. Sci. 261, 713-730. Dette interval, som er ekvivalent til Dal i den austral-asiske Ruedemann, R. 1947: Graptolites of North America. Mem. zonering, repræsenterer derfor et hidtil ubeskrevet trin i grap- geol. Soc. Am. 19, 1-652. tolit evolutionen. Schuchert, C. and Dunbar, C. O. 1934. Stratigraphy of western Newfoundland. Mem. geol. Soc. Am. 1, 1-123. Skevington, D. 1963: A correlation of Ordovician graptolite- bearing sequences. Geol. For. Stockh. Forh. 85, 298-319. Stevens, R. K. 1970: Cambro-Ordovician flysch sedimentation References and tectonics in west Newfoundland and their possible bear­ ing on a proto-Atlantic Ocean. Spec. Pap. geol. Ass. Can. Barnes, C. R., Norford, B.S. and Skevington, D. 1981: The 7, 165-177. Ordovician System in Canada. Correlation chart and ex­ Thomas, D. E. 1960: The zonal distribution of Australian grap­ planatory notes. LV.G.S. Pub!8. tolites. Jl. Proc. Roy. Soc. New South Wales 94,1-58. Barrington Brown, C. 1938: On a theory of gravitational sliding Tzaj, D. T. 1974: Graptolites of the Lower Ordovician in Ka­ applied to the Tertiary of Ancon, Ecuador. Q. Jl. geol. zakhstan. Izd. Nauk. Moscow, 1-127. Soc. Loud. 94, 359-370. VandenBerg, A. H. M. 1981: Victorian stages and graptolite Bergstrom, S. M. and Cooper, R. A. 1973: Didymograptus bi- zones. In Webby, B. D. (ed.). The Ordovician System in fidus and the trans-Atlantic correlation of the Lower Or­ Australia, New Zealand and Antarctica. Correlation chart dovician. Lethaia 6, 1625-1660. and explanatory notes. I.U.G.S. Publ. 6, 1-64. Berry, W. B. N. 1960: Graptolite faunas of the Marathon re­ Walcott, C. D. 1894: Paleozoic intra-formational conglomer­ gion, west Texas. Pubis. Bur. econ. Geol, Univ. Texas ates. Bull. geol. Soc. Am. 5, 191-198. 6005, 1-129. Whittington, H. B. 1963: Middle Ordovician trilobites from Berry, W. B. N. 1968: British and North American Lower Or­ Lower Head, western Newfoundland. Bull. Mus, Comp. dovician correlation; reply. Bull. geol. Soc. Am, 79,1265- Zool. 129(1), 1-118. 1272. Williams, S. H. and Stevens, R. K. in press. Arenig graptolites Cooper, R. A. 1973: Taxonomy and evolution of Isograptus of the Cow Head Group, western Newfoundland. Pal- Moberg in Australasia. Palaeontology 16, 45-115. aeontogr. Canad.