Stratigraphy of the Upper Ordovician upper Vaureal and Ellis Bay ...

2 downloads 94 Views 2MB Size Report
Sands present in the upper Vaureal and lower Ellis Bay formations in the east appear to have deterred the ...... Jeffery G Richardson and William I Ausich. 2007.
Stratigraphy of the Upper Ordovician upper Vaureal and Ellis Bay formations, eastern Anticosti Island, Quebec D. G . F. LONGAND PAULCOPPER Department of Geology, Laurentian University, Sirdhuv, Ont.. Canada P3E 2C6

Can. J. Earth Sci. Downloaded from www.nrcresearchpress.com by 99.236.90.37 on 04/10/11 For personal use only.

Received April 17, 1986 Revision accepted August 6, 1986 Marked facies changes occur in Late Ordovician strata, assigned to the uppermost Vaureal and Ellis Bay formations (Ashgill: Rawtheyan-Himantian) on Anticosti Island, Quebec. Western Anticosti features shales and carbonates, whereas outcrops along the eastem coast contain prominent, discontinuous, mixed siliciclastic-carbonate units. Detailed section measurement along the northeast coast allows, for the first time, accurate definition of seven new members within this uninterrupted sequence. Sands present in the upper Vaureal and lower Ellis Bay formations in the east appear to have deterred the growth of muddy-bottom brachiopod communities comparable to those in the western and central regions of Anticosti. Sand units within the upper Vaureal Formation contain 1 m diameter colonies of Paleofavosites; coeval small coral patch reefs are found in the central part of the island, where sands are absent. The uppermost Ellis Bay Formation of northeast Anticosti is marked by a shallow, subtidal, coral-algal oncolite bed or by small (2-4 m across, 1-2 m thick) local coral patch reefs, the tops of which have been used to define the Ordovician - Silurian boundary. No supratidal or intertidal sediments and faunas are evident in the Anticosti succession, suggesting that Late Ordovician sea-level drawdown was insufficient to provide shelfemergent conditions in this region. Des changements remarquables de faciks apparaissent dans les strates de l'ordovicien suptrieur, assignCes aux formations de VaurCal tardi-supCrieur et d'Ellis Bay (Ashgill : Rawtheyen-Himantien) sur l'ile d'Anticosti, QuCbec. Les schistes argileux et les carbonates dominent la partie occidentale de l'ile d'Anticosti tandis que les affleurements qui longent la cbte orientale sont formts principalement d'un melange d'unitts dicontinues de roches carbonatees et silicoclastiques. Une mesure dCtaillCe dans une coupe le long de la cBte nord-est permet, pour la premibre fois, de dCfinir avec pkcision sept nouveaux membres dans cette sCquence continue. Les sables pksents dans les formations de Vaukal supCrieur et d'Ellis Bay infbrieur i l'est semblent avoir nui B la croissance des communautCs de brachiopodes des fonds boueux comparativement 2 celles qui ksidaient dans les kgions de l'ouest et du centre d'Anticosti. Les unitCs arCnacCes au sein de la Formation de VaurCal contiennent des colonies de 1 m de diambtre de Paleofavosites; de petites constructions coralliennes contemporaines sont observCes dans la partie centrale de l'ile, oh les sables sont absents. La Formation d'Ellis Bay tardi-supkrieur du nord-est d'Anticosti est caractkrisCe par une couche d'algues oncoides et de coraux de haut-fond infralittoral ou par des petites constructions coralliennes locales (2 -4 m de large, 1-2 m d'kpaisseur); les toits de ces demibres ont servi i dCfinir la limite entre I'Ordovicien et le Silurien. Dans la sCquence i Anticosti il n'existe pas de faunes ou de sCdiments supralittoraux ou interlittoraux, ce qui indique que l'abaissement du niveau de la mer B l'ordovicien supCrieur Ctait insuffisant pour crCer des conditions d'Cmergence du plateau continental dans cette region. [Traduit par la revue] Can. J. Earth Sci. 24, 1807-1820 (1987)

Introduction The St. Lawrence Platform of eastern Canada contains a thick, continuous sequence of Middle Ordovician through Lower Silurian limestones, mudstones, and minor siliciclastics, which record deposition on the northwest margin of the Paleozoic Iapetus Ocean. These strata are flat lying, virtually undisturbed, and extensively exposed, especially in remote coastal cliff sections on Anticosti Island (Fig. 1). It is nevertheless difficult to correlate individual units of some of these formations along strike because of pronounced facies changes from east to west. The most conspicuous of these facies changes is the development of thick, local sand bodies on the northeast coast of Anticosti. These sands form part of the Upper Ordovician (Ashgill) strata of the uppermost Vaureal Formation and the lower Ellis Bay Formation. On western Anticosti Island the same formations are dominated by calcareous shales and limestones. The Anticosti Island Late Ordovician section is particularly important because it is the only location in eastern North ~ m e r i c awhere a thick, unbroken sequence of post-Richmondian rocks is exposed. Elsewhere, erosion or nondeposition of the latest Ordovician demonstrates a hiatus of perhaps as much as 1 or 2 Ma. This post-Richmondian succession is typified by the introduction of a rich shelly benthos of Silurian affinities, which has led some Pnnted m Canada 1 Imprime au Canada

workers in the past to assign the uppermost Ellis Bay Formation to the Silurian (e.g., Beny and Boucot 1970). The upper Vaureal and Ellis Bay formations were deposited at a time of marked glaciation in North Africa and may thus shed clues to sea-level changes mirroring such events (see Johnson et al. 1985). The objectives of this study are to present a revised stratigraphic framework for the upper Vaureal and Ellis Bay formations, to establish or more precisely define formation boundaries, and to identify lithologic units within the formations. An attempt is made at correlation with described units at the west end of the island, some 100 km away. The stratigraphic revision is based on detailed bed-by-bed measurement of continuous sections exposed between Mill Bay and Fox Point (Pointe du Renard) on the northeast coast of Anticosti Island. A detailed sedimentological study, with new paleogeographic implications, represents part two of this study and is the companion paper in this issue.

Previous work The earliest investigation of the geology of Anticosti Island was that of Richardson (1857). He divided the outcropping sequence into six divisions labelled A to F: these include Ashgill (Ordovician) to late Llandovery (Silurian) rocks. We

1808

CAN. J . EARTH SCI. VOL. 24.1987

I

ELLIS BAY FORMATION

Can. J. Earth Sci. Downloaded from www.nrcresearchpress.com by 99.236.90.37 on 04/10/11 For personal use only.

Schmitt Creek

VAUREAL FM.

VAUREAL FORMATION

FIG. 1 . Measured type sections of the upper Vaureal and lowermost Ellis Bay formation members on the northeast coast of Anticosti Island. The numbered units within each member are those identified in the type sections. Numbers on the map margins identify the National Topographic System (NTS) metric grid system; the northings and eastings are spaced at 1 km, and contour lines are at 25 ft (7.6 m). Symbols: black colour in the sections, shales; stippled, sands; and mounds, favositid colonies.

primarily consider here the upper part of Richardson's division B and lower part of division C , of Ordovician age (Fig. 2). The base of division B was marked by means of a "track bed" (a distinctive zipper-like trace fossil labelled Saerichnites by Billings in 1866; this fossil occurs at several levels). The top of unit B was defined by Richardson as 53 ft (16 m) of sandstones identified from Grindstone Cliff (Cap aux Meules) east to Table Head. Rocks above this sandstone, including shales and limestones, a biohermal unit, and another 139 ft (40 m) of limestone, were assigned by Richardson (1857) to division C. No rock units were named. Logan et al. (1863), using Richardson's data supplemented by paleontological identifications carried out by Billings (1857), coined the name Anticosti Group for "the remainder portion of the island [from Fox Point east

and south] . . . occupied by newer rocks. ' ' The underlying beds on the north coast were assigned to the Hudson River Group. Thus, although Richardson and Logan provided good descriptions of the sections, which are still recognizable, the early subdivisions show considerable stratigraphic overlap in the light of modem evidence. Despite this, Logan et al. were very close to identifying the Ordovician - Silurian boundary, which is near Fox Point. Schuchert and Twenhofel(1910) were the first to use formal stratigraphic formation names for the Ordovician and Silurian rocks of Anticosti. The Ordovician outcrops were divided into the English Head (units A1 -A6), Charleton (B1 -B1 I), and Ellis Bay formations (C 1-C 11) in ascending order (Fig. 2). They were early to recognize the Richmondian nature of the

I

I

LONG AND COPPER

Can. J. Earth Sci. Downloaded from www.nrcresearchpress.com by 99.236.90.37 on 04/10/11 For personal use only.

.

---

m-

PRINSTA VELLEDA

---_

LIMESTONE +SHALE ARGILLACEOUS LST. BIOHERMAL LST.

NODULAR LST. SANDSTONE CONGLOMERATE

FIG. 2. Correlation of the proposed stratigraphic nomenclature with previous schemes developed for the Ellis Bay Formation on Anticosti Island. The Ellis Bay - Becscie formational boundary shown conforms to the initial usage by Schuchert and Twenhofel (1910) at the top of the bioherms; the formational boundaly was later placed above this by Twenhofel (1928) and Bolton (1972).

faunas in their lower two formations and also the first to identify the Ellis Bay Formation as a new post-Richmondian, Ordovician series. They called it the Gamachian (after Gamache, the former name of Ellis Bay). Schuchert and Twenhofel(1910, p. 703) defined the boundary between their units B and C, i.e., their Charleton and Ellis Bay formations, as the base of a "thick sandstone showing in many places deposition by agitated waters," as exposed on the north shore. This contact is not exposed at the west end of the island at Ellis Bay. The resistant-weathering sandstone unit at the base of the Ellis Bay Formation is best exposed along the east side of Mill Bay adjacent to Cape James, at Grindstone Cliff, along the shoreline of Mill Bay, and at the 20 m high Schmitt Creek falls, some 300 m up Schmitt Creek (Fig. 1). The top of the Ellis Bay Formation, as defined originally by Schuchert and Twenhofel (1910, p. 702), was the top of a "coral reef limestone. . . 10 ft" thick. As a result, lithological boundaries were established for both top and bottom of the Ellis Bay

Formation, though at opposite ends of the island. In the most comprehensive work on Anticosti Island, Twenhofel (1928) used a double notation system for the subdivision of formations, using both broad "zones" and numbered, measured rock units. He dropped the tern Charleton Formation and replaced it with Vaureal. He then divided the Vaureal Formation into six "zones." In describing the measured sections along the Vaurkal River and the north coast of the island, he used a lithological numbering scheme (the north shore Vaureal being divided into six rock units numbered from top to base). No direct attempt was made by Twenhofel to correlate the nine lithological units measured on the Vaudal River with the six lithological units on the north shore, except in a very broad way in the zonal scheme. Indeed this would have been, and still is, difficult. Bolton (1961, 1970, 1972) provided new descriptions of the Anticosti sections, extending the Vaureal Formation downward to eliminate the name English Head Formation. In 1972

Can. J. Earth Sci. Downloaded from www.nrcresearchpress.com by 99.236.90.37 on 04/10/11 For personal use only.

1810

CAN. I. EARTH SCI. VOL. 24. 1987

Bolton also introduced a numbering system for the Ellis Bay Formation, splitting it into six members, the uppermost member 6 including "60 to 120 ft [18 - 39 m] . . . of thickto thin-bedded brown, dense, semilithographic limestone, [and] . . . isolated bioherms . . . at the base," which we include here in the Becscie Formation, except for the biohermal and interbiohermal beds at the base. The five western members (and base of member 6) outlined by Bolton, however, are generally not the same lithological or paleontological eastern members described in this paper. The top of the Ellis Bay Formation, as first stated by Schuchert and Twenhofel (1910), was the top of the biohermal unit. Bioherms are only selectively present in the Late Ordovician here, but the lateral extension of the biohermal unit is a widespread and uniform, thick-bedded coral -algal bed, the top of which is a convenient lithological -faunal break used to define the ordovician silurian boundary (petryk 1979, 198la; Cocks and Copper 1981). It should be noted that although the coral -algal bed and the bioherms are richly ~ o s s ~ ~ ~ ~ eno r o ordovicianus, Silurian boundary "golden spike9' based on gnptolite fossils, as the one defined at Linn in Scotland (Cocks 1985), can yet be defined at a specific bedding plane on top of the coral - algal bed or on top of the bioherms. Twenhofel (1921, 1928), Bolton (1972), Copeland (1973), and Petryk (1979, 1981a) primarily divided the Ellis Bay lzormation into zones or numbered members based on the partly exposed Junction Cliff to Cape Henry sections on the west end of Anticosti (Fig. 2). The base of the Ellis Bay Formation does not outcrop here, and there is considerable glacia1 or soil cover of the various members in the coastal bluffs. The tidal flats exposed at low tide yield continuous sampling through much of the Ellis Bay Formation, but like the western coastal bluffs, cannot be measured accurately for thickness (as seen in the anomalous thickness in McCracken and Barnes 1981) and ~ i e l dpoorly reserved megafossils. The contact between the Ellis Bay and Vaured formations at Anse aux Fraises is taken at the base of the northernmost hummocky cross-bedded calcarenites distinctly exposed on the tidal flat: this is substantially ( 10 m) below the boundary used there by ~ e t r y k(198la, 1981b map). Directly above this Cdcarenite are the characteristic Ellis Bay brachiopods Hindella and Eospirigerina. he salmon ~ i v e (Rivibre r au SaUmon) and Vaurkal River sections, used for conodont studies by McCracken and Barnes (1981), also are discontinuous. We have remedied this problem 'measuring a complete, composite section on the northeast coast. These coastal sections with vertical bluffs, accessible only by boat at Present, enabled us for the first time to determine reliable thicknesses for the uppermost Vaureal Y and to attempt a correlation with setand ~ l l i B s ~ formations tions on the west end of the island.

-

Stratigraphic revisions Stratigraphic sections were measured at six locations along

19 km of coastline on the northeast shores of Anticosti, from Mill Bay to past Table Head (Figs. 1, 4). These sections provide a continuous sequence from the upper 45 m of the Vaureal Formation through the Ellis Bay Formation and across the Ordovician- Silurian boundary into the lower 30 m of the Becscie Formation. This covers a time span in the Late Ordovician (Ashgill) from approximately Rawtheyan through Hirnantian to early Rhuddanian (Silurian: Llandovery), or from late Richmondian through Gamachian and into Clinton time, in North American terminology. 'pper vaureal Formation The Formation consists of to of shallowing-upward, thinly bedded carbonate mudstones, limestones, shales, and minor sandstones beneath the central and northem parts of the island, thickening by about 60 m beneath the south, as determined from drill-hole data (Roliff 1968; Riva 1969; INRS-Pktrole 1974). Only about 300 m of the Vaureal Formation is exposed as outcrop on the island, mostly in coastal and river exposures, and only the upper 13 of the Vaureal Formation will be considered in this paper. A conothe taken by NOwlan and Barnes Salmon River "about 10 m below the basal shales of the Ellis Bay Formati0n" indicates that the upper Vaureal has a Conodont Fauna 12 age (i.e.9 Richmondian). The uppermost Vaureal sequence is divided into two distinctive herein the Bay Member and Schmitt Creek Member, which outcrop in sections west from Tunnel Head Joseph Point (Fig. The members are described the base Inembers9

Mill B~~ ~~~b~~ (new) ~h~ w ill B~~ ~~~b~~ is accessible along the eastern side of ill B~~ from west of Schmitt Creek to a small coastal waterfall about 1 km to he east p i g . 1, section 3: NTS 12 E/8E Broom Bay 64-670:69-700). The type section here consists of 9.65 of resistant-weathering, laminated to mediUm-bedded, c o m m o ~ ycross-bedded, medium- to coarse-grained sandstones, interbedded with laminated carbonate siltstones and minor h i n shales ( ~ i 3e~-3f). ~ . ~h~ three upper units within the member, a 2.35 m ,-ross-bedded sandstone at the top platy, micaceous silt-sandstone in the (unit 4), a 1.7 middle (unit 3), and a 1.5 m ,-oa~e-grained sandstone below this (unit 2), show large-scale trough and hummocky crosscross-bedding in the upper and lower sandbedding. stones is beautifully displayed on the adjacent tidal flats at low tide (in unit at schmitt creek mouth, and 1 km west in unit 2). The uppermost 2-35 m unit (4) of sandstone contains large, in situ, 1 m diameter, hemispherical colonies of Paleofavosites (Fig. 3f) and smaller, sporadic, usually broken specimens of the rugose coral Paleophyllum, the tabulate coral Ellisites, and the cylindrical stromatoporoid Aulacera. Shelly fossils are very rare and broken in these sandstones. The lowermost 4.1 m of the Mill Bay Member (unit 1) con-

FIG. 3. Measured and type sections of the upper Vaureal and lower Ellis Bay formations. (a) Section at Prinsta Point showing the lower part of the Velleda Member (vel) at the top of the cliff and Grindstone Member (gs) towards the base (the lowest unit is not exposed here). ( b ) Table Head lighthouse rests on the Grindstone Member of the Ellis Bay Formation; the arrow marks the undulating surface of a large sand wave within the Grindstone Member. ( c ) Section southeast of Prinsta Point showing a prominent channel out at the top of the Grindstone Member (unit 4: top of stick) and the overlying Velleda Member (units 1 and 2). (d) Schmitt Creek falls (drop -20 m) marks the upper part of the Grindstone Member, with the upper ledges approximately defining the soft shales of the Velleda Member. (e) East Mill Bay section, demonstlating the Mill Bay Member (base; units 2-4), the Schmitt Creek Member (sc), and the Grindstone Member (gs) of the Ellis Bay Formation at the top; the figure in the lower left defines the scale. (f) Type section of the Schmitt Creek Member shales (units 1-4 at the top) and the underlying crossbedded sandstones of the Mill Bay Member (unit 4). Note the large in situ colony of Paleofavosites (arrow), which is typically associated with the sandstones.

I

I

1

Can. J. Earth Sci. Downloaded from www.nrcresearchpress.com by 99.236.90.37 on 04/10/11 For personal use only.

LONG AND COPPER

1811

1812

CAN. 1. EARTH SCI. VOL. 24, 1987

Can. J. Earth Sci. Downloaded from www.nrcresearchpress.com by 99.236.90.37 on 04/10/11 For personal use only.

Laframboise Member

ELLIS BAY FORMATION

70-

-

65

-

FIG.4. Map of the northeast coast showing the location of measured type sections of the Ellis Bay Formation. This map may be joined with Fig. 1 , from which it continues eastwards. Contour lines are 25 ft (7.6 m).

sists of medium-bedded, medium-coarse-grained, brown sandstones, interbedded with grey to greenish grey, soft-weathering shaly siltstones or silty shales with the brachiopods Oepikina and Sowerbyella. The unnamed Vaureal member below the Mill Bay Member differs markedly in having finely laminated, dark grey siltstones and calcareous shales, with lenticular limestone lenses or nodules, and broken shell lags containing small brachiopod fragments, ostracodes, bryozoans, and trilobites. The conformable lower contact of the Mill Bay Member is readily visible in cliff sections. Sandstones in this member are bioclastic feldsarenites with a spar cement. Most of the thicker bed sets are characterized by trough cross-stratification in sets up to 80 cm thick and 10 m wide. Sinuous-crested bedforms are well exposed on the rocky tidal platform in eastern Mill Bay, where they have an alongcrest wavelength of 8 - 10 m and a crestal amplitude of about 2 m. Planar cross-stratification is developed locally at the base of the member. Thinner bed sets have plane and wavy bedding and ripple cross-lamination, and some contain abundant mudstone intraclasts. Bioturbation is minor. Hummocky cross-

bedding is locally developed in very fine sand-grade units. The Mill Bay Member corresponds approximately to unit 4 of Twenhofel's (1928, p. 46) north shore section, but Twenhofel must have visited the section at high tide, since he did not mention the well-exposed trough- and hummocky-bedded sandstones visible for some 800 m along the tidal flats, nor did he mention the existence of a fault with 2 m displacement in the shore bluffs east of Schmitt Creek. The Mill Bay is a mappable member visible in the cliffs at Joseph Point to the west and around Cape James to the east. The abundance of very large in situ favositids and the localized overgrowths and intergrowths of these with other corals suggest that this unit may be the lateral equivalent of the small upper Vaureal Formation bioherms exposed in the centre of Anticosti island. However, the sandstones in the Mill Bay Member probably disappear westwards near the Salmon River: they have not been reported on the Vaurkal River in the centre, or on the west coast of the island. In our interpretation, the lowermost "arenaceous limestone" (SO1 to SO4 or as high as S07) of section IlIA along the Salmon River, in McCracken and Barnes (1981), assigned by

Can. J. Earth Sci. Downloaded from www.nrcresearchpress.com by 99.236.90.37 on 04/10/11 For personal use only.

LONG AND COPPER

them to the Ellis Bay Formation, is probably the Mill Bay Member of the Vaureal Formation. One of the prime reasons for this assumption is that there would otherwise be an anomalously thick carbonate section of the Ellis Bay Formation only a few kilometres from our measured sections. A second reason is that the covered section there could mask thin sandstone units as the westward fingering out of the thick sandstone of the lower Ellis Bay. A third reason is the appearance of the conodont Gamachignathus in unit 12 of the Salmon River section of McCracken and Barnes (1981). This identifies the basal Ellis Bay Formation at Ellis Bay and would thereby also make the eastern sections more consistent in their correlation. However, the exact range of Gamachignathus and its status as a genus have recently been questioned. Savage and Bassett (1985) regarded Gamachignathus as a junior synonym of Birksfeldia Orchard 1980, which is of Rawtheyan age. No megafaunas were reported by McCracken and Barnes (1981, appendix) in their lower Ellis Bay Formation at the Vaukal River except for "Beatricea," which ranges through both formations. The Mill Bay Member is part, probably the lower part, of undefined member 5 of Petryk (1981~). Schmitt Creek Member (new) The Schmitt Creek Member, named after Schmitt Creek, which flows into Mill Bay (Fig. I), describes 6.05 m of laminated, nodular, dark grey, calcareous shales or nodular argillaceous limestones with minor thin, discontinuous beds of laminated to very thin bedded, medium to very fine lime sandstones to siltstones, which are locally channeled and crosslaminated. This is defined here as the uppermost member of the Vaureal Formation. The type section includes the first low bluffs exposed west of Schmitt Creek (Fig. 3f), extending towards Grindstone Creek, where there is a small waterfall on the shoreline (NTS 12 El8E Broom Bay 64700-65800:6960070200). The base of the member, only partly exposed at the type section, is defined at the contact with the underlying cross-bedded, bioclastic feldsarenites of the Mill Bay Member outcropping on the tidal flats. East of Schmitt Creek the lower contact is visible and consists of grey, laminated shales. The top of the Schmitt Creek Member is the last silty shale bed below the thin- to medium-bedded flaggy sandstones of the Grindstone Member above. This also marks the top of the Vaureal Formation. The Schmitt Creek sandstones, a very minor component of the member, range from spany, bioclastic subfeldsarenites to biosparites, with only a minor terrigenous component (these occur as thin ledges, which have been used as markers to define local units in the type section; Fig. 3f). They are typically flat to wavy laminated, with local development of ripple cross-lamination in units with erosional bases. Bioturbation is common, including irregular burrows and minor domichnia (Diplocraterion). Mudstone intraclasts are present in some ripple-laminated units. Nodules and discontinuous carbonate beds in the grey, laminated mudrocks that dominate the Schmitt Creek Member appear to represent flasers of carbonate sands that foundered in a muddy substrate. Pseudonodules show surface borings (?Trypanites), which possibly indicate early lithification or cementation. The upper shales and thin limestones of this member contain a relatively modest to sparse fauna of Oepikina fluctuosa, Sowerbyella, rhynchonellids (?Lepidocyclus), crinoid and bryozoan fragments, rare small, spherical heliolitids, rare large Paleofavosites ( 1 m), and small-diameter aulacerids. The beds are bioturbated, especially the silty ledges, with silty and

-

1813

sandy beds increasing upwards. The Schmitt Creek Member usually weathers recessively; this is probably the reason it is not well or is only partly exposed on the west side of the island. At Anse-aux-Fraises, on the west coast, the top of the Schmitt Creek Member is placed by us below the resistantweathering calcarenites due west of the old cemetery on the tidal flats (see Barnes et al. 1981, Fig. 5). Thus, directly north of this on the tidal flats, the soft Schmitt Creek shales give way to an eroded embayment and small delta. Equivalent strata on Vaukal Falls are inaccessible on the falls escarpment, though only thin shales occur there. The member probably corresponds to units 5 and 6 of the north shore section of Twenhofel (1928, p. 46: he indicated more than 40 ft of strata), to the lower 6 m of Vaureal member 5 of Petryk (1981a), and to units S05-SO6 and covered units above these of section IIIA on the Salmon River in McCracken and Barnes (1981). They may also be equivalent to the "upper 15 feet of argillaceous, nearly a nodular limestone" mentioned by Bolton (1961, p. 16) for the uppermost Vaureal of central Anticosti. We have placed this member in the Vaureal Formation because the megafaunas are similar to those observed in unnamed nodular shales and limestones directly below the Mill Bay Member and different from those of similar lithology in the overlying member 1 of the Ellis Bay Formation (sensu Bolton 1972, p. 8). No diagnostic Ellis Bay Formation brachiopods, e. g ., Hindella, Eospirigerina, Leptaena, or Parastrophinella, occur in this unit, despite the fact that the lithology here is similar to the lithology in which Hindella occurs most abundantly in the Ellis Bay Formation. Ellis Bay Formation Along the northeast coast of Anticosti, the Ellis Bay Formation consists of 52-62 m of limestone, mudstone, and sandstone, with minor conglomerate, local thickness variations being attributed to the development of clastic units. It is exposed in cliff, bluff, and river sections between Joseph Point to the west and Fox Point to Lousy Bay to the east, near Table Head (Figs. 1, 4). It is here divided into five distinct units named, from the bottom up, the Grindstone, Velleda, Prinsta, Lousy Cove, and Laframboise members. The Grindstone Member is the lowest member within the Ellis Bay Formation, and thus its base also defines the base of the formation. Since the base is covered at the west end of the island around Ellis Bay and similarly covered or inaccessible elsewhere, the north coast, with continuous outcrop, should be used to define the precise boundary. Schuchert and Twenhofel (1910, p. 699), in describing the north coast section, stated: "It is at the base of this sandstone that for the present the writers would draw the dividing line between B and C." We are interpreting that thick sandstone unit as the Grindstone Member. The "arenaceous shale" below it is therefore part of the Vaureal Formation. Billings (1857, p. 252) was the first to recognize that fossils in Richardson's division C (the Ellis Bay Formation) contained "Upper Silurian species" (i.e., Silurian sensu stricto). This includes distinctive brachiopods such as Hindella and Eospirigerina, which have Silurian descendants but no direct Ordovician counterparts. These fossils have so far only been discovered in the lowermost beds assigned to the Ellis Bay Formation on the west side of Anticosti and do not occur in the clastic unit at the base on the east side, e.g., the Grindstone Member. Grindstone Member (new) The Grindstone Member was first clearly identified by Richardson (1857, pp. 213 -214), who described 53 ft of

Can. J. Earth Sci. Downloaded from www.nrcresearchpress.com by 99.236.90.37 on 04/10/11 For personal use only.

1814

CAN. J. EARTH SCI. VOL. 24, 1987

sandstones as follows: "in the bight of a cove . . . [I] give to a cliff the name of Grindstone Cape," placing the sandstones at the top of his division B. The type section is located at Grindstone Cliff in Mill Bay (NTS 12 El8E Broom Bay 6480070200), with a second reference section exposed not far away at Schmitt Creek falls, one of the highest waterfalls on the island (66300-68800: Fig. 3d). Both sections are dominated by thinly laminated to thin-bedded, very fine sandstones with minor laminated mudstones. In the section at Schmitt Creek falls, the Grindstone Member is 18.25 m thick. At the type section, only the basal 14.1 m is exposed. The base of the member is conformable with the underlying shales of the Vaureal Formation, with no sign of an erosional contact (Fig. 3e). At Schmitt Creek falls the Grindstone Member is, in turn, conformably overlain by recessive-weathering nodular calcareous shales and limestones of the Velleda Member. The base of the Grindstone Member is not exposed at Table Head lighthouse, even at lowest tide, despite the fact that the Grindstone Member makes up most of the high cliff on which the lighthouse is built (Fig. 36). Sandstones of the Grindstone Member are subfeldspathic to feldspathic bioarenites or bioclastic subfeldsarenites or feldsarenites with spar cement. Carbonate detritus (fossils and peloids) is slightly more abundant in the upper part of the member (42 %) than in the lower part (32%). There is a corresponding decrease in the relative abundance of feldspar (17% dropping to 7 % ) and an increase in the quartzlfeldspar ratio from 2.8: 1 in the lower half to 6.8: 1 in the upper. In the type section the sandstones at the base are mainly flat to wavy laminated, with some bioturbation and shale intraclasts. These beds are locally affected by slumping, with consequent development of numerous, conspicuous, isolated load balls from 20 to 100 cm across, confined to channels up to 3 m thick and 50 m across with erosional bases (Fig. 1: unit 1, section 1). The upper part of the lower half of the member is trough and hummocky cross-stratified and capped by a thin veneer of very coarse sandstone. This marker sandstone was not observed in the Schmitt Creek section, but it appears to be equivalent to channel-fill sandstones in exposures at Table Head. The latter section also contains minor medium-grained bioclastic sandstones at a level approximately coincident with the top of the slumped units in Grindstone Cliff. The upper half of the member in the type section at Grindstone Cliff begins with flat to wavy laminated sandstones at the base, giving way to a thick sequence of hummocky and trough cross-bedded sandstones up section in sets up to 60 cm thick. These are overlain by plane-laminated sandstone and mudstone and an upper unit of laminated sandstone. The top of the member is not exposed at Grindstone Cliff; at Schmitt Creek falls it is marked by laminated to very thin, nodular to wavy bedded, fine sandstones with some large Paleofavosites. Strata at an equivalent level at Prinsta Point include pebbly medium and coarse sandstones, which infill 1.5 m deep, 8.2 m wide channels (Fig. 3c). In the Table Head sections the top of the member is marked by a distinctive intraformational boulder lag containing intraformational clasts, corals such as Ellisites, Tollina, Paleofavosites, and the cylindrical stromatoporoid Aulacera, which rest on a broad erosional surface. In one section this can be traced at least 80 m along section, with a relief of at least 70 cm. Maximum extent of this zone probably exceeds 2 krn. In general, fossils are not common in the sandstones. Paleofavosites may have diameters of more than 1 m and are not

overturned (these occur especially at the top), but other corals show evidence of erosional wear or breakage and are much smaller. Aulacera are horizontally oriented and consist of broken fragments averaging 20 cm in diameter and 1 m long: one Aulacera had a diameter of nearly 40 cm. In the channel fill at Prinsta Point, pockets of transported shells (preserved as moulds) of the large bivalve Technophorus(?), gastropods, and rhynchonellids are abundant (Fig. 3c). Trilobite, smooth ostracode, bryozoan, solitary rugosan, and nautiloid fragments are scarcer. Except for the one occurrence of rhynchonellids, the most common brachiopods are broken strophomenid shells. Nearly all the shelly fossils are disarticulated, broken, or worn, and many occur as lag deposits, indicating sorting and transport; fossils are more common in the shaly partings. In the lower sandstones, large Aulacera, large Paleofavosites, and small-diameter Cryptophragmus occur with sowerbyellids. Twenhofel (1928, p. 54), following earlier, more general discussions in Schuchert and Twenhofel (1910, p. 699), established the sandstones exposed at Grindstone Cliff as the basal unit of the Ellis Bay Formation. Units 7 and 8 of Twenhofel's north shore section (1928, p. 54) consisted of "70 ft" (22 m) of quartz sandstones, which are here assigned to the Grindstone Member. An "arenaceous limestone" unit on the Salmon River (units S9 - S 10, section IIIA), sampled for conodonts by McCracken and Barnes (1981), may be the part equivalent of the Grindstone Member. At Vaurkal Falls to the west, Twenhofel (1928, p. 51) defined 22 ft (7 m) of thinbedded limestone of the basal Ellis Bay Formations, "forming the low cliff bordering the river just above Vaureal Falls," as the lowermost Ellis Bay Formation. McCracken and Barnes (1981) defined the Vaureal - Ellis Bay boundary at Vaurkal Falls lower down in the section, just below the lip of the falls, also in limestones. In our interpretation, this boundary may be best placed higher up, below the ripple-laminated limestones occurring above the falls, i.e., at or about 1-2 m below sample V10 (McCracken and Barnes 1981, p. 54), which marks the first appearance of Gamachignathus in the Vaukal section. The Grindstone Member sandstones are missing at Vaurkal River and at the west end, but the shales and limestones of member 1 of Bolton (1972) may be stratigraphically near the top of the Grindstone Member. However, the diagnostic Ellis Bay shelly fossils of Bolton's shaly member 1 on the west end of the island have not yet been discovered in the Grindstone sandstones. This is attributed to shale-facies dependence of the diagnostic shelly fauna. At the west end of the island, we believe that the stratigraphic equivalent of the base of the Grindstone Member is the lowermost (and northernmost), thick, hummocky-bedded calcarenite exposed as a resistant-weathering ridge on the tidal flats due east of the Anse-aux-Fraises cemetery (see Barnes et al. 1981, Fig. 5). We found a good Eospirigerina -Hindella fauna, characteristic of the Ellis Bay Formation, less than 2 m above this contact, thus about 10- 13 m below the top of Petryk's ( 1 9 8 1 ~ )member 5 of the Vaureal Formation and below Petryk's Vaureal - Ellis Bay contact on the west coast. Velleda Member (new) The Velleda Member is named after a small lake, draining into Prinsta Bay and located southwest of Table Head (Fig. 4). The type section of the member is at the bluffs north of Lousy Cove (Anse Mauvaise), about 2 krn south of Table Head (NTS 12 F15w Table Head 80800:65000; Fig. 6d, section 6). Lousy Creek marks the start of the old service road to Table Head

1815

LONG AND COPPER -

Can. J. Earth Sci. Downloaded from www.nrcresearchpress.com by 99.236.90.37 on 04/10/11 For personal use only.

'F:c

1

=---.., , . . -,., -, .

.- ...,zi: - -- ,, , -.-. .. - ..:.= ..:.;.: ..,,,.

.

--.:. ..er,:-.r=

?:-!:,=:k.

..

. ..

...2,, . :