'Bu([ctms of

Begun in 1895

VOLUME 109, NUMBER 349

MARCH 5, 1996

Stromatoporoids from the

Emsian (Lower Devonian) of

Arctic Canada

by

Eric Prosh and Colin W. Steam

Paleontological Research Institution

1259 Trumansburg Road

Ithaca, New York, 14850 U.S.A.

S

r or)

?)

>P ^

PALEONTOLOGICAL RESEARCH INSTITUTION

Officers

President John C. Steinmetz

Vice-President Richard E. Petit

Secretary Henry W. Theisen

Treasurer Pamela Wait

Director Warren D. Allmon

Trustees

R. Tucker Abbott (to 6/30/96) Bruce M. Bell (to 6/30/96) Carlton E. Brett (to 6/30/98) WiLLL\M L. Crepet (to 6/30/97) J. Thomas Dutro. Jr. (to 6/30/96) Shirley K. Egan (to 6/30/98) Anton J. Egner (to 6/30/97) M. G. Harasewych (to 6/30/98) Harry G. Lee (to 6/30/97) Amy R. McCune (to 6/30/97) Samuel T. Pees (to 6/30/98)

Richard E. Petit (to 6/30/96) Edward B. Picou (to 6/30/98) Gary Rosenberg (to 6/30/96) Constance M. Soja (to 6/30/97) James E. Sorauf (to 6/30/97) John C. Steinmetz (to 6/30/97) Susan B. Stephens (to 6/30/96) Henry W. Theisen (to 6/30/98) Mary Kane Trochim (to 6/30/98) Pamela Wait (to 6/30/96) Thomas C. Whiteley (to 6/30/97)

Trustees Emeritus

Harry A. Leffingwell

Robert M. Linsley Raymond Van Houtte William P. S. Ventress

BULLETINS OF AMERICAN PALEONTOLOGY

and

PALAEONTOGRAPHICA AMERICANA

Warren D. Allmon Editor

Reviewers for this issue

Barry D. Webby

Carl W. Stock

A list of titles in both series, and available numbers and volumes may be had on request. Volumes 1-23 of Bulletins ofAnicncaii Paleontology are available from Periodicals Service Company. 1 1 Main St., Germantown, New York 12526 USA. Volume 1 of Palaeontograpliica Americana has been reprinted by Johnson Reprint Corporation, 1 1 1 Fifth Ave., New York. NY 10003 USA.

Subscriptions to Bulletins ofAnicrican Paleontology' may be started at any time, by volume or year. Current price is US $62.50 per volume. Numbers of Palaeontograpliica Americana are priced individually, and are invoiced separately on request.

for additional information, write or call:

Paleontological Research Institution

1259 Trumansburg Road

Ithaca, NY 14850 USA

(607) 273-6623

FAX (607) 273-6620

0 This paper meets the requirements of ANSI/NISO Z39. 48-1992 (Permanence of Paper).

yaUon

Begun in 1895

tolc^

VOLUME 109, NUMBER 349

MARCH 5, 1996

Stromatoporoids from the

Emsian (Lower Devonian) of

Arctic Canada

by Eric Prosh and Colin W. Steam

^ MAR 29 1996

Paleontological Research Institution

1259 Trumansburg Road

Ithaca, New York, 14850 U.S.A.

ISSN 0007-5779

ISBN 0-87710-440-9

Library of Congress Catalog Card Number: 95-071054

Printed in the United States of America

Allen Press, Inc.

Lawrence, KS 66044 U.S.A.

CONTENTS

Page

Abstract 5

Introduction 5

Acknowledgements 5

Stratigraphy

Introduction 6

Eids Formation 6

Blue Fiord Formation 6

Bird Fiord Formation 9

Disappointment Bay Formation 9

Unnamed Formation 9

Conodont Biostratigraphy

Dehiscens and Gronbergi Zones 9

Inversus Zone 10

Serotinus Zone 10

Undifferentiated Patulus Zone 10

Stromatoporoid Biostratigraphy and Paleogeography

Southwestern Ontario and North-Central United States 11

Northern Asia II

Australia and China II

Discussion 13

Systematic Paleontology

Introduction 13

Systematics

Order Actinostromatida

Family Actinostromatidae

Genus Actinostroma 14

Genus Plectostroma 15

Genus Aculatostroma 16

Order Clathrodictyida

Family Clathrodictyidae

Genus Clathrodictyon 17

Genus Gerronostroma 18

Genus Petridiostroma 19

Genus Atelodictyon 20

Family Tienodictyidae

Genus Anostylostroma 21

Genus Pseudoactinodictyon 22

Genus Schistodiclyon 23

Order Stromatoporellida

Family Stictostromatidae

Genus Stictostroma 24

Genus Stromatoporella 26

Genus Clathrocoilona 27

Family Hermatostromatidae

Genus Trupetostroma 30

Order Stromatoporida

Family Stromatoporidae

Genus Stromatopora 31

Genus Ferestromatopora 32

Genus Glyptostromoides 33

Family Syringostomellidae

Genus Syringostromella 34

Genus Salairella 34

Order Syringostromatida

Family Synngostromatidae

Genus Atopostroma 35

Genus Habrostroma 36

Genus Parallelopora 37

Appendix 1 . Collection Localities 38

Appendix 2. Catalogue of Type Numbers and Locations 39

Appendix 3. Collecting locations and stratigraphic positions 40

References cited 42

Plates 46

Index 64

LIST OF ILLUSTRATIONS

Text-figure Page

1 . Position of collecting localities in Canadian arctic islands 7

2. Correlation chart of formations from which stromatoporoids were collected 8

3. Slratigraphic ranges of Lower and lower Middle Devonian stromatoporoids 12

LIST OF TABLES

Table Page

1 . Comparative measurements of Slictostroma gornense Steam 25

STROMATOPOROIDS FROM THE EMSIAN (LOWER DEVONIAN) OF

ARCTIC CANADA

Eric C. Prosh and Colin W. Stearn

Earth and Planetary Sciences, McGill University, Montreal, Canada

ABSTRACT

Early Devonian limestones of EUesmere, Bathurst, and smaller islands between them in the Canadian Arctic Archipelago contain a diverse fauna of stromatoporoid sponges. This fauna provides the best evidence in North America of the early recovery phase of this reef-building group from a diversity low at the Silunan/Devonian boundary, a recovery that lead to its diversity peak in Givetian time. Stromatoporoids from the lower member of the Blue Fiord Formation locally form large reefal masses. Well preserved stromatoporoids also occur less abundantly: 1. in the top of the underlying Eids Formation, 2. in the upper member of the Blue Fiord Formation, 3. in the Disappointment Bay Formation, which is correlative of the upper Blue Fiord, and, 4. in the overlying Bird Fiord Formation, and a correlative unnamed formation, both of which span the Lower/Middle Devonian boundary. The stratigraphic distribution of these stromatoporoids can be accurately determined according to conodont biostratigraphy as spanning the dehiscens to partitus (Emsian to basal Eifelian) conodont zones. Common occurrences of Slro- maloporelta perannulata. Stictostroma gorriense. Hahroslroma proxilammatum. and Parallelopora campbelh in the arctic fauna and southern Ontario and the adjacent United States, indicate that the Detroit River Group is of similar Emsian age, and that the Eastern Americas realm was open to migration from the Arctic. Similarity of species with the Emsian faunas of Russia, Australia and China suggests the cosmopolitan and equatorial distribution of stromatoporoids in Emsian time and opens possibilities for using the group in correlation. The fauna is therefore important in establishing both the evolution of the order and also its geographic distnbution in Early Devonian time.

Twenty-five species (assigned to 22 genera) are described. The species concept used is a broad one and the range of variation in each taxon is documented. New species described are: Genonoslroma septentrionalis. Anostylostroma anfraclum. Pseudoac- tmodictyon conglutmalum. Stictostromal nunavutense. Clathrocoilona vexata, Stromatopora hensoni. The morphologic limits of the following genera are considered in the description of species: Plectostroma. Aculalostroma and Atelodictyon, Clathrocoilona. Salairella and Syringostromella. The range of Trupetostroma is extended downward into Emsian strata.

INTRODUCTION

Rocks of earliest Devonian age cover a relatively small area of the North American platform because this was a time of worldwide regression at the close of the Tippecanoe sequence. Reef faunas of this age are restricted in their distribution and of low diversity. In North America the Kaskaskia transgression appears to have started in the north and, spreading southward in shallow seas across the platform, brought with it a reef- building fauna rich in stromatoporoids. The first reef complexes of regional extent built in this sea were in the area that is now the Canadian Arctic Archipelago and are contained within the Blue Fiord Formation. This study describes the elements of this resurgent reef fauna. In eastern and arctic North America the growth of the reef fauna in subsequent Middle and Late De- vonian time was progressively inhibited by siliciclastic input, but in the Western Canada Sedimentary Basin reefs thrived through the middle part of the period and their growth culminated in giant Frasnian reef com- plexes.

This study expands and refines earlier work by Steam (1983) on the Emsian stromatoporoids of southern EUesmere Island. Steam described a fauna collected

from the lower 100 m (lower Emsian, dehiscens Zone) of the Blue Fiord Formation in the type area between Eids and Sor Fiords (Text-fig. lA). Of the 11 species described by Stearn (1983), eight are recognized in this study: Clathwdictvon ellesmerense Steam, 1983, Ger- ronostroma septentrionalis n. sp., Clathrocoilona vex- ata n. sp., Stromatopora polaris. Steam, 1983, Stro- matopora cf S. hiipschii (Bargatzky, 1881), Glyptos- tromoides simplex (Yang and Dong, 1979), Salairella prima Khromych, 1971, and Atopostroma distans (Ripper, 1937b). The three species that do not occur in the collections studied here, which were made by Gary Smith and Eric Prosh, are all represented in Steam's (1983) collections by single specimens. These include Gerronostroma cf. G. immemoratum Bogo- yavlenskaya, 1977, Amphipora sp., and Labechia sp. No amphiporids were found in these larger collections. The single poorly preserved specimen referred to La- bechia sp. has been tentatively identified as Syringo- dictyon tuberculatum (Nicholson) by St. Jean (1986).

ACKNOWLEDGEMENTS

We are grateful to Gary P. Smith who, in the prep- aration of his doctoral dissertation (Smith, 1984), col-

Bulletin 349

lected most of the specimens on which this study is based. The field work of Smith (in 1978, 1979, and 1980) and of Prosh (m 1983 and 1992) was made pos- sible by the logistic support of the Polar Continental Shelf Project of Natural Resources Canada. Smith's field work was funded by Natural Resources Canada, the Natural Science and Engineering Council, and the McGill Centre for Northern Studies and Research. The research of Eric Prosh and Colin Steam is funded by grants from the Natural Sciences and Engineering Re- search Council, Canada to Steam. For critical com- ments that have improved the manuscript we are grate- ful to Barry Webby, Carl Stock, and Warren Allmon.

STRATIGRAPHY

Introduction

The Emsian and basal Eifelian stromatoporoids de- scribed in this study were collected from the Eids, Blue Fiord, Disappointment Bay, Bird Fiord, and an un- named formation. The great majority of specimens were collected by Gary Smith and Colin Steam from the Blue Fiord Formation (lower to upper Emsian) of Ellesmere Island. Most of the Blue Fiord specimens were collected along a broad outcrop belt extending from Blue and Bird fiords in the west to Sor Fiord in the east (Text-fig. lA). Because this outcrop belt in- cludes the type section of the Blue Fiord Formation, localities within the belt are referred to in the text as within the "type area" or in the "vicinity of the type section". Supplementary collections from the Eids and Bird Fiord formations, which respectively underlie and overlie the Blue Fiord Formation (Texi-fig. 2), were also included in the study. Collections of Smith and Steam from the Blue Fiord Formation northeast of the type area in the vicinty of Vendom Fiord (Text-fig. 1 B) and by Smith from southwestern Ellesmere Island at Muskox and Goose Fiords (Text-fig. IC) were also studied. The location of the collections studied and the stratigraphic sections measured by Smith (1984) are plotted on Text-figure 1 and listed by latitude and lon- gitude in Appendix 1 . The geology of this southem coast of Ellesmere Island has been described recently by Mayr et al. ( 1 994).

The Disappointment Bay Formation of Bathurst, Comwallis, and adjacent islands is a correlative of the middle-to-upper Blue Fiord Formation of Ellesmere Island (Text-fig. 2). Small collections from the Dis- appointment Bay Formation on Truro Island (Text- fig. 1, loc. 34) made by Eric Prosh in 1983 and 1992 and from Lowther Island (Text-fig. 1, loc. 33) in 1983 are also described here.

On Bathurst Island and neighbouring islands, the uppermost Emsian and Eifelian are represented by an

unnamed limestone that has been referred to incor- rectly as the Blue Fiord Formation. Stromatoporoids were collected from this unnamed formation on Bath- urst Island by Smith and on nearby Tmro Island by Prosh.

Eids Formation

The Eids Formation (Text-fig. 2) consists mostly of calcareous siltstone and shale, with less limestone, silt- stone and sandstone. It conformably underlies the Blue Fiord Formation in southem and central Ellesmere Island (Trettin, 1978). West of Sor Fiord it reaches a thickness of 767 m (Uyeno, 1990). With the exception of the uppermost beds transitional to the Blue Fiord Formation, the Eids Formation is poorly fossiliferous. It ranges in age from Lochkovian to earliest Emsian. Silty carbonates on Bathurst and adjacent islands as- signed by Kerr (1974) to the Eids Formation are much younger than the Eids on Ellesmere Island.

Stromatoporoid specimens identified in this study as from the Eids Formation come from the upper 100 m in the Blue and Sor fiords areas of Ellesmere Island. In these areas the upper Eids consists of dark gray, calcareous siltstone and mudstone with interbeds of fossiliferous lime wackestone and isolated bioherms (Smith, 1984).

Blue Fiord Formation

The Blue Fiord Formation is an important and wide- ly distributed cliff-forming unit in the arctic islands. The best exposed sections occur on southem Ellesmere Island and adjacent islands. The type section was des- ignated by McLaren (1963) between Blue Fiord and Eids Fiord (Text-fig. lA). The typical Blue Fiord is predominantly a dark fossiliferous limestone spanning most of the Emsian Stage (Smith, 1984).

The Blue Fiord Formation has been recorded from the arctic islands west of Ellesmere Island (Thorsteins- son and Tozer, 1962) and from the subsurface (Mayr, 1980) in the Bent Horn oilfield (a short distance west of the northwest tip of Bathurst Island and just offText- fig. IC). Much of what has been referred to the Blue Fiord Formation in the westem arctic islands, how- ever, is of Eifelian age, and should ultimately be re- ferred to a new, as yet undescribed formation (see un- named Formation below).

Studies of Blue Fiord macrofossils include those of Brice (1982) and Jones and Boucot ( 1 983) on brachio- pods, Ormiston (1967) on trilobites, and Redder ( 1 982, 1983) on corals.

In the type area of the Blue Fiord Formation, McLaren (1963) recognized two members: a lower limestone and shale member and an upper brown lime- stone member. The lower member is about 700 m thick

Devonian Canadian Stromatoporoids: Prosh and Stearn

Text-figure 1. Position of collecting localities in Canadian arctic islands. Precise positions by longitude and latitude for the numbered localities are given in Appendix 1 . A. Blue Fiord-Sor Fiord area. B. Vendom Fiord area. C. Localities outside A and B. The stippled boxes show the positions of maps A and B. Truro Island 34. Lowther Island— 33.

(689 m; Uyeno, 1 990) and consists of brownish gray to brown nodular limestones and lesser interbedded gray calcareous mudstones and shales (McLaren, 1 963). The lower member is cliff-forming, abundantly fossil- iferous and contains large stromatoporoid-bearing bioherms. Smith (1984), and Smith and Steam (1982, 1987a) recognized a variety of lithologic units within the lower member: interbedded lime mudstone and shale, argillaceous fossiliferous wackestone, skeletal

grainstone to packstone, coral-stromatoporoid bound- stone. The upper brown limestone member is 572 m thick in a section near the type section (Uyeno, 1990) and consists of brown and brownish gray, bioclastic, coarse-grained limestone but contains relatively few fossils.

On the west side of Vendom Fiord (Text-fig. IB) the Blue Fiord Formation consists of about 1200 m of limestone and minor siltstone. Two informal units are

Bulletin 349

Conodont zones

SOUTHERN ELLESMERE ISLAND

BATHURST, TRURO, LOWTHER ISLANDS

<

Bird Fiord Fm.

unnamed formation

W5

"a 1

partitus

patulus

serotinus

^ Blue Fiord y/^ Dolomitic Facies \. (Muskox & Goose / Fiords)

Blue \

Disappointment Bay Fm.

inversus

gronbergi

Fiord Fm.

dehiscens

Eids Fm.

< O

i

Text-figure 2.— Correlation chart of formations from which stromatoporoids were collected.

recognized; 1 ) a lower carbonate member about 900 m thick consisting of gray limestone with dolostone interbeds and dark grayish-brown limestone, and 2) an upper siltstone-carbonate member of greenish silt- stones, calcareous siltstones and limestone interbeds overlain by a resistant grayish-yellow limestone and dolostone (Uyeno, 1990). The upper siltstone-carbon- ate member is the same unit provisionally assigned to the Bird Fiord Formation by Jones (1982).

To the north and east of the type area as far as central EUesmere Island, Blue Fiord lithologies represent pro- gressively more restricted depositional environments (Kerr, 1976; Trettin. 1978). On southernmost Files- mere Island at Goose Fiord and adjacent Muskox Fiord, the formation is mostly dolomitic and deposition be-

gan later than in the type area, in Late Emsian time (Text-fig. 2)(Mayr et al., 1994). This platformal dol- omitic facies also occurs on adjacent parts of Devon Island (Kerr, 1977; Prosh et al.. 1988). Lithologically the dolomitic Blue Fiord consists of vuggy dolostone, vuggy lime mudstone, and dark fossiliferous lime wackestones and dolowackestones (Smith and Steam, 1987b). The dolomitic Blue Fiord records a period of late Emsian transgression and platform inundation; it is approximately correlative with the upper member of the type Blue Fiord. Smith and Steam (1987b) pro- posed that the dolomitic facies of the Blue Fiord be assigned to the Disappointment Bay Formation, but the term Blue Fiord is retained for these beds here (Prosh et al.. 1988).

Devonian Canadian Stromatoporoids: Prosh and Stearn

Bird Fiord Formation

The Bird Fiord Formation overlies the Blue Fiord Formation conformably, and reflects in its lithology a transition to siliciclastic depositional conditions. It is widely distributed on southwestern Ellesmere Island, northwestern Devon Island and the Bathurst Island group (Goodbody, 1989). It ranges in age from latest Emsian to Eifelian. In the type area on Ellesmere Island it is over 800 m thick (Goodbody, 1989).

A few stromatoporoid specimens were studied from the lower Bird Fiord Formation in the type area on the north side of Bird Fiord. These beds (the Norwe- gian Bay Member of Goodbody, 1989) consist of bio- clastic sandy limestone and calcareous siltstone with minor shale and argillaceous siltstone.

Disappointment Bay Formation

This late Emsian formation (Text-fig. 2) has been studied on Comwallis Island, eastern Bathurst Island, and intervening smaller islands (Kerr, 1974; Thor- steinsson, 1980, 1986). It is mostly dolostone but lo- cally on Bathurst, Truro, and Lowther islands (Text- fig. IC), rare limestones (low in the formation) are abundantly fossiliferous.

On Truro Island, the Disappointment Bay Forma- tion consists mostly of light-colored, massive, vuggy, microcrystalline dolostone with subordinate laminated dolostone. It is about 200 m thick (Kerr, 1974). On the northeastern tip near the base of the formation, small reef knolls occur. Kerr ( 1 974) and Thorsteinsson (1986) do not indicate this occurrence on their maps. Stromatoporoid specimens were collected from these reef knolls of dark gray-brown, bituminous, fossilif- erous packstone. The small knolls are lithologically similar to coeval larger knolls in the Disappointment Bay Formation on Lowther Island (Thorsteinsson, 1980; Prosh, 1989).

Unnamed Formation

An unnamed unit consisting of limestone and minor dolostone and shale conformably overlies the Disap- pointment Bay Formation on eastern Bathurst, Com- wallis, and intervening small islands (map unit D-1 of Thorsteinsson, 1986). On Bathurst Island it incorpo- rates most or all of what Kerr (1974) erroneously re- ferred to as the Blue Fiord Formation. Kerr's Bathurst Island "Eids Formation", a distal equivalent of his "Blue Fiord Formation", probably also belongs to the unnamed formation (Thorsteinsson, pers. comm., 1992). The unnamed formation spans the Emsian-Ei- felian boundary and is 20-100 m thick (Thorsteinsson, 1986) (Text-fig. 2).

Stromatoporoids used in this study were collected from the lower 100 m of the formation at the north-

easternmost end of Bathurst Island and at Dyke Ack- land Bay on the southern coast (Text-fig. IC, Iocs. 35, 36). On the western coast of Truro Island (Text-fig. IC, loc. 34) 4-6 m of the unnamed formation contain a coral-stromatoporoid biostrome locally up to 6 m thick. Stromatoporoids from this biostrome are in- cluded in this study.

CONODONT BIOSTRATIGRAPHY

In this study we have been able to relate the strati- graphic ranges of stromatoporoids directly to an es- tablished conodont zonation, in other words to the standard Emsian zonal scale (Text-fig. 2). This is in great measure due to the excellent work of Uyeno ( 1 990) who has documented the conodont zonation of the Devonian of southern Ellesmere Island. Many of the stromatoporoids used in this study come from local- ities identical or closely comparable to those of Uyeno and Klapper ( 1 980) and Uyeno ( 1 990). In the following paragraphs, the biostratigraphic framework is outlined and the stratigraphic intervals bearing stromatopo- roids are placed in the standard conodont zonation.

DEHISCENS and GRONBERGI ZONES

These zones comprise the lower Emsian. In a mea- sured section in the type area of the Blue Fiord For- mation, Uyeno and Klapper (1980) and Uyeno (1990) have recognized the dehiscens Zone in the lower 267 m of the lower member. The gronbergi Zone is rec- ognized by the first appearance of Polygnathus aff. P. perbonus as Polygnathus gronbergi itself does not oc- cur. In this section the gronbergi Zone is recognized in the 267 to 393 m interval of the lower member. Be- cause P. gronbergi is absent and the gronbergi Zone here spans a relatively small stratigraphic interval, we commonly combine the dehiscens and gronbergi zones in dating. Consequently, stromatoporoids occurring exclusively from the lower 250 m or so of the formation of the type area Blue Fiord are assigned to the dehiscens Zone. Stromatoporoids occurring through the lower member up to about 400 m are assigned to a combined dehiscens/ gronbergi Zone.

In the Vendom Fiord area, conodont zones in the lower member of the Blue Fiord Formation cannot be as precisely placed as in the type area (Uyeno, 1990). Conodonts and macrofossils associated with the de- hiscens Zone occur low in the formation. The dehiscens and gronbergi zones probably occur in the lower two thirds of the lower member here: P. inversus first ap- pears high in the lower member. Stromatoporoids from low in the lower member at Vendom Fiord are assigned to the dehiscens Zone; those from low to medial parts of the member to the combined dehiscens/ gronbergi Zone.

10

Bulletin 349

In the Sor Fiord area, Uyeno and KJapper (1980) have identified P. dehiscens from the uppermost beds of the Eids Formation. The full extent of its occurrence in this formation is not known, but presumably the base of the Emsian is near the top of the formation. Stromatoporoids from the highest beds of the Eids Formation are assigned to the dehiscens Zone; those 50 or more meters below the top of the Eids are pos- sibily uppermost Pragian.

INVERSUS Zone

This zone is widely and readily recognized in arctic Devonian strata. In the Blue Fiord type area, the in- versus Zone spans the interval from 393 to 1 104 m, the upper half of the lower member and much of the upper member (Uyeno, 1990). At Vendom Fiord the inversus Zone ranges from the upper part of the lower member through much of the upper member of the Blue Fiord Formation. In both the type and Vendom Fiord areas, in the upper member of the formation, stromatoporoids are relatively rare and are assigned to a combined inversus/serotinus Zone.

On southernmost Ellesmere Island, the dolomitic facies of the Blue Fiord Formation ranges in age from the inversus to serotinus zones (Smith and Steam, 1987b). Stromatoporoids from this dolomitic facies from Muskox Fiord and Goose Fiord areas, are all from low in the formation and assignable to the in- versus Zone.

Although the lower Disappointment Bay Formation is firmly dated as inversus Zone, the upper part is dol- omitic and evaporitic and unfossiliferous. Stratigraph- ic relations suggest that the formation spans the full serotinus Zone as well. Stromatoporoids from the lower Disappointment Bay Formation at Truro and Lowther islands occur within the inversus Zone (Thorsteinsson, 1980).

SEROTINUS Zone

In a section 2.5 km east of the type section of the Blue Fiord Formation the serotinus Zone is identified in the interval of the upper member from 1 104 m to the top of the formation at about 1260 m (Uyeno, 1990). At Vendom Fiord the zone is recognized in the uppermost part of the Blue Fiord upper member. Al- though P. serotinus itself ranges up into the overlying Bird Fiord Formation, the overall faunal and strati- graphic context suggests that the zone ends at the top of the Blue Fiord Formation in the type area (Uyeno, 1990).

Undifferentiated patulus Zone

The Emsian-Eifelian boundary is placed at the boundary between the patulus and partitus zones (Zie-

gler and KJapper, 1985). Because in the arctic islands the nominal conodonts of neither of these zones has been recognized, this interval must be dated by other fossil occurrences and by stratigraphic context. For dating stromatoporoid occurrences, we treat this in- terval as the undifferentiated patulus Zone (i.e.. com- bined patulus and partitus zones). Two formations that contain stromatoporoids span this interval: 1) the low- er Bird Fiord Formation of Ellesmere Island, and 2) the lower unnamed formation of Truro and eastern Bathurst Islands.

1) Uyeno (1990) placed the Lower-Middle Devo- nian boundary at or near the base of the Bird Fiord Formation at its type section. Although diagnostic con- odonts are absent, brachiopods and corals from higher in the formation suggest Eifelian and Dalejan (late Em- sian) ages (Uyeno, 1990). Goodbody (1989) considerd the basal Bird Fiord to be included in the patulus Zone. On this basis we date the stromatoporoids that have been collected from the lower Bird Fiord Formation on Ellesmere Island as coming from the undifferen- tiated patulus Zone.

2) Similar uncertainty in dating applies to the un- named formation of Truro and Bathurst islands. Kerr (1974) considered the lower unnamed formation (Blue Fiord of Kerr) as Eifelian and possibly latest Emsian in age, largely on the basis of its trilobite fauna. A sample of the unnamed formation analyzed for con- odonts by T. T. Uyeno yielded no specimens (Uyeno, pers. comm. 1993; GSC internal report 02-TTU-93).

The age of the unit underlying the unnamed for- mation is important in estimating the age of the for- mation. In the area of Comwallis, Truro and Lowther islands, the Disappointment Bay Formation is dated as inversus Zone (Thorsteinsson, 1980) on the basis of diagnostic conodonts recovered from rare limestones low in the formation. The upper Disappointment Bay Formation is assumed to span most, or all, of the ser- otinus Zone and the transition to limestones of the unnamed formation is assumed to begin at about the patulus Zone. This conclusion is based on correlation of the transition from Blue Fiord to Bird Fiord for- mations on Ellesmere Island with that between the Disappointment Bay and unamed formation in the Comwallis Island area suggested by Thorsteinsson (1986).

Stromatoporoids collected from the unnamed for- mation come from the base at Truro Island and from the lower 100 m of the formation on eastem Bathurst Island. In both areas they are probably of latest Emsian age (patulus Zone). Due to the uncertain dating of this interval, however, they are assigned to the undiflFer- entiated patulus Zone.

The stromatoporoid fauna offers some clue as to the

Devonian Canadian Stromatoporoids: Prosh and Stearn

11

age of the lower unnamed formation (Text-fig. 3). Some faunal turnover is apparent between the the Blue Fiord {dehiscens to serotinus zones) and the unnamed for- mation (undifferentiated /)an</;« Zone) collections, but a significant number of species range through both as- semblages. This suggests that no hiatus separates the Blue Fiord and younger assemblages.

STROMATOPOROID BIOSTRATIGRAPHY AND PALEOGEOGRAPHY

In this study stromatoporoid ranges have been pre- cisely dated, allowing meaningful comparisons of the arctic Emsian fauna to coeval faunas world-wide and the assessment of the biostratigraphic usefulness of stromatoporoids (Text-fig. 3). Of the 25 taxa described in this study, 1 3 have been recorded from other places and supply information on the distribution of stro- matoporoid faunas in Emsian time. These regions are, in decreasing order of similarity and importance: 1) north-central United States and southwestern Ontario, 2) Russian Asia, 3) southeastern Australia, and 4) Chi- na. When localities sharing species with the Arctic Em- sian fauna are plotted on a reconstruction of the con- tinental configurations of Early Devonian time, such as that of Scotese (Stock, 1990), they are distributed throughout the tropical world of that time. On a broad scale the Emsian stromatoporoid fauna appears to have been a cosmopolitan one and stromatoporoids must have propagated along the tropics with little impedi- ment.

Southwestern Ontario and North-Central United States

Four species known only from this mid-continent region, Stromatoporella perannulata Galloway and St. Jean, Stictostroma gorriense Steam {=mamiUiferum Galloway and St. Jean), Habrostroma proxilaminatum (Fagerstrom), and Parallelopora campbelli Galloway and St. Jean, have been identified in the arctic Emsian fauna. In addition the arctic Tnipetostroma sp. and Pseudoactinodictyon conglutinatum n. sp. are closely comparable to species from southwestern Ontario (Fa- gerstrom, 1982). This sharing of species has led Prosh and Steam (1993) to conclude that the Detroit River Group of southem Ontario is entirely of Emsian age. The reasoning that led to the reinterpretation of the correlation of the Detroit River Group is presented elsewhere (Prosh and Steam, 1993). The identity of species indicates a dispersal of stromatoporoids along an open seaway connection from the Arctic to the mid- continent during serotinus Zone time and perhaps as early as during inversus Zone time. That the isolation of the Eastern Americas Realm of Oliver ( 1 976b) based on the distribution of rugose corals (Oliver, 1976a;

Oliver and Pedder, 1989) and on stromatoporoids (Stock, 1990) was not as absolute as proposed, has already been suggested by Steam (1983) and is fully supported by this study. The migration route of stro- matoporoids across the present site of Hudson Bay may have been closed to mgose corals by ecological barriers. Evidence is accumulating that environments favorable to stromatoporoids and tabulate corals may have been different from those that favored mgose corals (Mallamo et al. 1993). In any case, the paleo- geographic reconstructions of Oliver (1976b), Oliver and Pedder (1989), and Witzke ( 1 990) need to be mod- ified to show an open seaway connection between the Arctic Islands and southwestern Ontario.

Northern Asia

Plectostroma salairicum (Yavorsky) is known in the Arctic from the unnamed formation (Bathurst Island) and from the Kuznetsk Basin of central Russia (Ya- vorsky, 1930). The age of both these occurrences is virtually identical, at or near the Emsian-Eifelian boundary. Aculatostroma cf A. kaljanum (Bogoyav- lenskaya) from the lower Emsian part of the Blue Fiord Formation is known (as Coenellostroma kaljanum Bo- goyavlenskaya) from the eastern slope of the northern Urals and from lower Emsian strata (Bogoyavlenskaya, 1977; Khodalevich et al., 1982). Gerronostroma cf G. immemoratum Bogoyavlenskaya, described by Steam (1983) from the lower Blue Fiord, also was originally described from this fauna. Atelodictyon cf A. solidum Khromych (Blue Fiord Formation, inversus Zont) was previously recorded (A. cylindricum solidum Khro- mych) from probable mid-to-late Emsian rocks in northeastem Siberia (Khromych, 1971, 1976). Two other arctic species are also recorded from Russia, but appear to be longer ranging. Syringostromella zintch- enkovi Khalfina occurs in the upper Lochkovian of the Salair, southcentral Russia (Khalfina, 1961) and 5a- lairella prima Khromych from probable Pragian rocks in Severo-Vostok, eastern Siberia (Khromych, 1971, 1 976). Both of these species also occur in Australia (see below).

Australia and China

Affinities with the Early Devonian faunas of Victoria (Webby et al.. 1 993) are indicated by the cosmopolitan species Syringostromella zintchenkovi Khalfina, Stro- matopora polaris Steam, Atopostroma distans (Ripper) and Salairella prima Khromych. The arctic Emsian fauna shares 7 genera in common with the 1 1 taxa described by Webby and Zhen (1993) from the Jesse Limestone of New South Wales. At the specific level Atopostroma distans and Salairella prima occur in both faunas.

12

Bulletin 349

i]pqdwvD vuodopipjDd

tUTUDUJWDJlXOjd DlUOJlSOjqVfi

suvjsip vwojisodoiy

vtuud OlpjlDlDS M0^31I3}U12 VJpiUOJJSOSuUKS

xBidiujs s3piouiojisoidK]{) }A0^mu3dru:>l j jo vjodojtnuouis3J3^

nt{osdm( s P vjodojmuojis

lU0SU3t{ DUOdOlDUiOJJS

suvpd Djodojviuojjs ds DWOJtsoiddrux

VIVXM DUOjtODOUliJVU

asuajriADunu ipiuoJisoiJxis

dSUdUUOS DIUOJISOJDIIS

DivjnuuDjad vjpuodojtmojjs

•ds iuo/Q3ipo}Sii(JS

lunjmimiSuoj uotQjjpouijODopnasj

lunjovjfiiv vwoJisoj/Qsouy

lunpips v P uuKioipopjv

ds Dmojjsoipu}3j

SipU0U)U3ld3S DIUOJISOUOJJ3Q 3SU3J3U4S3JP UOWpOjyjVJJ

wnuv[p:>t V 'P owojisojvpoy

lUnOUWpS DWOJJSOJ031J

V "ds vutouisoutpy

O

3

u

a

Nvnajia

3

CO

a

3

o u <u

3

U

> e

±

u

B O

3 a u

'.3

Mvispva

•g 3

w O

a o ta E o

o -J

C/5 I

Devonian Canadian Stromatoporoids: Prosh and Stearn

13

The species Glyptostromoides simplex (Yang and Dong) is common to the Emsian of the south China continental plate and the arctic Emsian.

Discussion

This synthesis demonstrates that the stromatopo- roids have more biostratigraphic value than has been conventionally attributed to them. A prerequisite for the confident use of these fossils in correlation is that dating of occurrences be precise as in this study of the Emsian of the Arctic. Such dating has generally been lacking in classical studies of stromatoporoid system- atics in part because rigorous, internationally accepted standards of dating postdate many of these studies. Stromatoporoids have been conventionally regarded as poor candidates for correlation in the belief that the species are too long ranging and too provincial. Con- servative taxonomic philosophies, species defined without clear appraisal of range of variation within skeletons or between specimens, and oversplitting of taxa have made recognition of species across national boundaries difficult. The recognition of the world-wide distribution and narrow stratigraphic range of some species in this study demonstrates that the biostrati- graphic and paleogeographic value of this group is po- tentially great. Some genera in which taxonomic prob- lems at the species level persist, such as Stromatopom and Clathrocoilona, may be poor candidates for cor- relative purposes, but other genera, such as Stictostro- ma and Stromatoporella, where species are better de- fined, have great biostratigraphic potential.

The chronologic development of stromatoporoids in Early Devonian time has recently been reviewed by Webby et al. (1993). The diversity and wide distri- bution of the Emsian world fauna shows that the rapid diversification of stromatoporoids to form the better- known reef complexes of Middle and Late Devonian time was well underway in Early Devonian time.

SYSTEMATIC PALEONTOLOGY

Introduction

Stromatoporoids are an extinct class of sponges al- lied to the extant coralline sponges (Steam, 1972, 1975a). Little can be gleaned from the distant stro- matoporoid-sponge relationship, however, to assist in setting limits to extinct stromatoporoid species. Most of the classical systematic studies of stromatoporoids (pre- 1 960s) employed a species concept that by current standards would be considered narrow. Sample suites were often limited in size and geographic scope, and as a result new species were plentiful. Many of these have since been combined in synonymy. In more re- cent years, larger samplings have been collected, and more attention has been paid to morphometric analysis

(e.g., Fagerstrom, 1982), and a much broader concept of the stromatoporoid species has ensued. The species concept we apply here is a relatively broad one (Steam, 1989a).

Many stromatoporoid species demonstrate signifi- cant structural or morphometric variability. In many species a large proportion of the range of variability may be expressed within a single skeleton. In general, a grouping of specimens may be confidently considered a unique species when the range of variability (however broad) is continuously expressed across the full col- lection, with no major gaps or discontinuities. An ob- vious precondition for recognizing the range of vari- ability is a relatively large number of specimens. A good indication of the broad species concept as applied herein is shown by a comparison of total species di- versity with generic diversity. The Arctic Emsian stro- matoporoid fauna comprises 25 species in 22 genera; only two genera are represented by more than a single species.

A consequence of a broad species concept is that, in our opinion, the notion of subspecies is generally in- applicable to stromatoporoids. Because variability within a species is already great, rarely (if ever) will a given population consistently demonstrate character- istics deserving subspecies rank.

Species synonymies listed are meant to be all-inclu- sive, listing all species considered identical to the spe- cies being described. Those species described as confere (cf.) to an earlier description have the compared ref- erences listed in the synonymy (following cf.). Confere (cf.) is the only nomenclatorial qualifier employed here. It is used for described species that closely resemble those described elsewhere. Reservation may be due to minor morphologic differences, too small a sampling to assure that the full range of variability is present, or inadequacies in the original description. Expanded study of such species may ultimately demonstrate that they are the named species (delete cf) or, much less likely, that they are separate but closely related species.

Species-level taxa are distinguished by letters (e.g., Actlnostroma sp. A) when we judge that a distinctive morphology is exhibited that is likely to result in a new species being described when more and better speci- mens have been studied. If the material is sufficient for a generic identification only, the generic name is modified only by sp. (Trupetostroma sp. = a species of Trupetostroma).

In the course of this study, approximately 800 thin- sections were examined. The preservation of identified specimens is graded according to the scale: exception- ally well preserved; well preserved; moderately well preserved; mediocre; poor. These categories are ap- proximate and subjective, but in general imply the

14

Bulletin 349

following. Exceptionally well preserved specimens pre- serve all macrostructure and especially microstructure in very fine detail, uniformly throughout the specimen; specimens of this quality are exceedingly rare. Well preserved specimens show little or no diagenetic al- teration; macrostructural details are fully preserved, and original microstructures are preserved over most of the specimen. Moderately well preserved specimens show some diagenetic alteration; macrostructure is largely unaffected, but microstructural detail may be obscured, although it is generally preserved at least locally. Most identified specimens are either well or moderately well preserved. Mediocre preservation ap- plies to specimens in which macrostructural details are obscured, at least partly, and no original microstruc- tural detail remains; specimens so preserved remain identifiable on macrostructural grounds only. Poorly preserved specimens preserve no microstructural de- tail, and macrostructures are largely obscured. In gen- eral, poorly preserved specimens cannot be confidently assigned (to species) in isolation, but can be identified in the context of the full collection studied.

The value of classifying the preservation of speci- mens (in combination with the number of specimens identified) is that it allows the reader to gauge inde- pendently the confidence of the species identifications/ descriptions. This is particularly true for newly de- scribed species.

Morphological measurements provided are in met- ric units, commonly centimetres (cm), millimetres (mm), or micrometres (/um). Most measurements are derived from thin-section examination. Most fre- quently cited morphological parameters are laminar and pillar spacing, and laminar and pillar thickness. Laminar and pillar spacing are conventionally ex- pressed as number of laminae/pillars occurring per 2 mm distance, commonly cited as an average of n mea- surements with a minimum and maximum range of values per specimen or species.

Morphological terminology follows the established literature on stromatoporoid systematics. Definitions of recently introduced terms may be found in such papers as Steam, ( 1 989b, 1 99 1 , 1 993) and Stock ( 1 989). Introductory glossaries appear in older papers such as Galloway and St. Jean ( 1 957). In this report, we deviate from the normal terminology only for the following morphological feature. Where astrorhizae in tangential section are without walls and defined only by areas devoid of structural elements, the term "canal" or "tube" is inapproporiate; we use the term "astrorhizal path" or "path".

Type specimens designated in this report are housed in the Type Collection at the Geological Survey of Canada, Ottawa, Ontario, and are identified with a six-

digit number with the prefix GSC {e.g., GSC 108863). The locations from which these specimens were col- lected are listed in Appendix 2. The precise location of collections is recorded by latitude and longitude in Appendices 1 and 2 because very few geographic names are available in the High Arctic. Appendix 3 is a catalog of the other specimens identified in this study with their collection localities referenced to Appendix 1 and Text-figure 1 . These specimens are in the general col- lections of the Geological Survey in Ottawa marked with a prefix number 110, 111, 120, or 129 followed by a hyphen and a second number (e.g., 1 10-286). The few types referred to by the prefix ROM are from the collection of the Royal Ontario Museum, Toronto, On- tario, Canada.

Systematics

Class STROMATOPOROIDEA

Nicholson and Murie, 1878

Order ACTINOSTROMATIDA

Bogoyavlenskaya, 1969

Family ACTINOSTROMATIDAE Nicholson, 1886

Genus ACTINOSTROMA Nicholson, 1886

Type species, —.ictinostroma clathratuin Nicholson, 1886.

Actinostroma sp. A

Plate 1, figures 1-3

Description. —Skcltxon large, hemispherical; surface bearing low mamelons closely and regularly spaced; no astrorhizae visible on surface.

Vertical section: Laminae thin, flat or very gently undulate,