Biogeography and biofacies pattems of Middle polymeroid from North Greenland: palaeogeographic and palaeo-oceanographic implications

Loren E. Babcock

Fossils from Middle Cambrian rocks of Peary Land and Nyeboe Land, North Green­ land, show that polymeroid biofacies were strongly segregated near the shelf margin ofLaurentia. Some biofacies include trilobites of Laurentian aspect that lived in warm, relatively shallow water, and other biofacies have trilobites of Baltic aspect that lived in cool, deeper water. Polymeroids in the Glossopleura and Ptychagnostus gibbus zones are from open-shelf lithofacies and are mostly of Laurentian aspect. Polymeroids in the Lejopyge laevigata Zone are from deep, outermost shelf to upper slope lithofa­ cies and are entirely of Baltic aspect. Polymeroids in the Ptychagnostus atavus Zone are from an intermediate, outer-shelf setting and are of mix.ed Laurentian and Baltic aspect. Mixing of polymeroids from different biofacies is inferred to have occurred through downslope movement by localised sediment gravity flows. The presence of segregated trilobite biofacies at the shelf margin of Laurentia is the first evidence for a thermocline in marine waters surrounding Laurentia during the Middle Cambrian, although evidence for a thermocline during the Early and Late Cambrian has been reported previously. The presence of polymeroids of Baltic aspect in Laurentian rocks of North Green­ land has important implications for the analysis of tectonostratigraphic terranes. The presence of cool-water trilobites is not sufficient evidence to conclude that the terrane originated in high palaeolatitudes. Identification of an autochthonous terrane of a continental margin is possible, however, on the basis of assemblages containing mixed biofacies representatives. Provenance identification is possibie because the assemblages are expected to include taxa characteristic of warm-water cratonic areas along which the terrane originated.

L. E. B., Department af Geological Sciences and Byrd Polar Research Center, The Ohio State University, Columbus, Ohio 43210, U.S.A.

Historically, Cambrian tri10bite faunas from the cir­ nation of Sweden's tri10bites (Westergård, 1946, 1953). cum-Atlantic region have been assigned to either an At­ Information about trilobite biogeography is also routinely lantic or Pacific fauna1 province. These provinces have used to supplement pa1aeomagnetic, stratigraphic, struc­ also been called Acado-BaHic and North American, re­ tura1, or other geo1ogic evidence about the configuration spective1y. Several biogeographica1 studies (e.g., Loch­ and history of continents and oceans during the ear1y man-Balk & Wi1son, 1958; Palmer, 1969, 1972, 1979; Pa1aeozoic. Cowie, 1971, 1974; Theokritoff, 1979, 1985; Shergo1d, The biogeography of Cambrian tri10bites has been cen­ 1988) have shown that the Pacific provinee is character­ tral to some arguments conceming the origin of tecto­ ised by po1ymeroid tri10bites that were mostly endernic to nostratigraphic terranes in the Appa1achian Orogen (see the she1f of Laurentia, whereas the Atlantic province is Keppie, 1989 and Horton et al., 1989 for definitions of characterised by po1ymeroids that were first described terranes). The presence of po1ymeroids having At1antic from localities in Scandinavia, southem Great Britain, affinities in rocks of Maritime Canada and New England Maritime Canada, or New England. (e.g., Howell, 1925; Hutchinson, 1952, 1962; Poulsen & Tri10bites of the Atlantic provinee have been important Anderson, 1975; Bergstrom & Levi-Setti, 1978) has been in studies of early Pa1aeozoic biostratigraphy and pa1aeo­ used to infer the exotic origin of these terranes and geography. Zonations deve10ped for Middle Cambrian subsequent accretion to North America during the midd1e tri10bites worldwide are common1y compared to the zo- or 1ate Pa1aeozoic (see Williams & Hatcher, 1982; Secor

Bull. Grønlands geol. Unders. 169, 129-147 (1994) © GGU. Copenlwgen. 1994 130

50 km A Brønlund Fjord & I Tavsens Iskappe Groups

> > > Wandel Valley Fm Amundsen Land > o o O Amundsen Land Group "C il Gp "E "E O ~ es es O ...J ...J f.:c ~ 15" f-- S2- E E (,) I- (,) E! <:: Kap Stanton Cl Ol Kap Stanton Formation IIII~ ~.g ::::l ::::l B Fm i= (,) S - r2. FmKOC~s~opetSSU Holm ~~ ..2- -.c .c Cl Dal Fm E E .§ " Ol Ol .c .c (,) i= ~ Fimbuldal Fm :t~o!1: (,) E E Henson Gletscher Formation Hensan Ol ::;; B Ekspedition Bræ Fm (,) ~~ Gletscher r- ::;; Cl Sydpasset Fm ::;; ~ .c = Fm æ ~ E ~ Henson Gletscher Fm al""j5" Ol Aftenstjernesø Formation Ol ~ (,) (,) E E Aftenstjernesø Fm Aftenstj. Fm Ol ...J Buen Formation ...J (,) B Buen Fm ...J Buen Fm ...J W N Nyeboe Land NW Peary Land E S SW Peary Land NW Peary Ld N c Localities 1 & 2 Locality 3 B Locality 3

[. :::l E Ul IL l!!-, D r!fCl Locality 3 ----?--_.

c: o c: Locality 2 ~~ Lejopyge en E laevigata 0.0 ZOne Locality 1 ~IL IS " * Ss 1::.2 Ptychagnostus tiJ~ punctuosus ~~ 0.0 0.0 ~IL Zone ---- ...... -- ~IL Fig. 1. Derivation of fossiliferous samples. A, map showing the distribution of the Brøn­ Plychagnostus lii .c: * Holm Dal lund Fjord and Tavsens Iskappe Groups in atavus .2" lil " Zone 1ii

Fig. 2. Maps of reconstructed Middle Cambrian paleogeogra­ phy (after Scotese & Denham, 1988). Outlines of present-day geographic areas are inc1uded for reference. The view is from 30 degrees south latitude. I, View emphasising the relationships of Laurentia, Baltica, and Siberia. 2, View emphasising Gondwana and the peri-Gondwanan terranes.

Baltica, which consisted of north-western Europe ex­ microcontinental blocks, and includes an archipelago 10­ clusive of the British Isles and western Norway, was cated in low palaeolatitudes (Cook & Taylor, 1989; Cook probably in moderately high to high southern latitudes et al., 1991). Considerable uncertainty remains concern­ during most of the Cambrian (Scotese et al., 1979; Sco­ ing the pa1aeogeographic positions of tectonostrati­ tese, 1987; Scotese & McKerrow, 1990; Torsvik et al., graphic terranes of China. 1991; but see Smith et al., 1981). Shelf sedimentation A variety of geophysical, stratigraphic, structural, pet­ was dominated by fine-grained siliciclastics, including rologic, and palaeontologic evidence has been used to a1um shales. Carbonate units comprise a minor part ofthe assess plate-tectonic history. Faunal evidence is of partic­ Cambrian section. ular importance in the lower Palaeozoic because other Major components of Gondwana were South America, reliable geologic and geophysica1 evidence can be more Africa, Florida, Arabia, Turkey, Iran, Afghanistan, India, difficult to obtain than for other intervals of the Phanero­ Madagascar, Australia, and parts of Antarctica. Gond­ zoic. Distributions of played a major role in Wil­ wana extended from tropicaI to polar latitudes. England son's (1966) hypothesis of a proto-Atlantic Ocean, as (Wales, southern Ireland, and southern England), Acadia well as in many subsequent studies of the history of the (eastern New England and Maritime Canada except for lapetus Ocean. Distributional data continue to be impor­ eastern Nova Scotia and western Newfoundland), and tant in current arguments about the definition and history Armorica (Bohemia, southern Germany, France, Iberian of terranes in the circum-Atlantic region (e.g., Williams Peninsula, northern Morocco, and eastern Nova Scotia) & Hatcher, 1982; Secor et al., 1983; Samson et al., 1990) were three important peri-Gondwanan terranes. All were and elsewhere. located in polar or subpolar latitudes of the North African Studies of the distribution of Cambrian trilobites sector of Gondwana. worldwide (e.g., Palmer, 1969, 1972, 1979; Jell, 1974; Other important terranes include those of Kazakhstania Burrett & Richardson, 1980; Shergold, 1988; Robison, (including present-day Kazakhstan, Tien Shan, and the 1991) have demonstrated that shelf seas of the major Aral Sea area) and present-day China. Recent evidence cratonic areas of Laurentia, Baltica, Siberia, and Gond­ suggests that Kazakhstania is composed of several fused wana tended to have relatively large numbers of endernic 133 trilobite genera and species. Endemism is strongest often been disregarded when synthesising geological and among polymeroids, most of which were probably ben­ geophysical data for palaeogeographic reconstructions. thic. Most genera and many species of the largely pelagic agnostoids had cosmopolitan distributions in world Biogeographic affinities of North Greenland oceans (e.g., Robison, 1972, 1984, 1988; bpik, 1979). polymeroid trilobites Patterns of distribution of ear1y Pa1aeozoic tri10bites in the circum-Atlantic region have traditionally been used to Middle Cambrian rocks of North Greenland contain define At1antic (or Acado-Baltic) and Pacific (or North polymeroid trilobite assemblages of both Laurentian and American) fauna1 provinces (e.g., Palmer, 1969, 1972; Baltic aspect. Most polymeroids described by Poulsen Poulsen, 1969; Cowie, 1971, 1974). Biotic exchange be­ (1927), Poulsen (1964), Palmer & Peel (1981), and Robi­ tween these regions has been generally thought to have son (1988) have c10sest affinities with trilobites that are been limited by severa1 factors, inc1uding the time avail­ characteristic of open-shelf areas of Laurentia. Polyme­ able for 1arval migration across geographic distances roids identified by Poulsen (1969) from Nyeboe Land, (McKerrow & Cocks, 1976; Burrett & Richardson 1980' inc1uding Eodiscus punctatus and a paradoxidid (possib1y Fortey & Cocks, 1986). In this interpretation, iapetu~ Centropleura) have c10sest affinities with taxa character­ may have been wide during the Cambrian (Conway Mor­ istic of cratonic areas of BaItica. The observed geo­ ris & Rushton, 1988). A trend of increasing similarity graphic distributions of polymeroid taxa identified from between North American and western European faunas the Henson Gletscher and Kap Stanton formations are during the and Si1urian has been inferred to summarised in Tables 1 to 4. result from the progressive c10sure of lapetus and the All genera in new Geological Survey of Greenland elimination of fauna1 barriers (McKerrow & Cocks, (GGU) collections (Babcock, 1994, but see a1so a prelim­ 1976). inary notice by Fletcher et al., 1988) from the Glos­ Studies of distributiona1 patterns in the circum-Atlantic sopleura Zone (Tab1e 1) are widely distributed in open­ region, 1ike most that have used biogeographic data to shelf lithofacies of Laurentia. One genus, , is he1p determine the configuration of plates during the a1so known from shelf lithofacies ofpresent-day Asia and ear1y Pa1aeozoic, have incorporated three major assump­ Australia. Two species, klotzi and Glosso­ tions (Taylor & Forester, 1979). First, fauna1 elements pleura walcotti, have been recognised elsewhere in Lau­ rentia. were assumed to have been largely restricted to shallow­ shelf seas of single continental blocks. Second, the shal­ In addition to its occurrence in North Greenland, O. low-shelf seas were surrounded by deeper oceans, which klotzi is known from the of British inhibited dispersal. Third, the degree of fauna1 dissimi­ Columbia (Rasetti, 1951), the Metalline Formation of larity between areas is assumed to have been a function of Washington (McLaughlin & Enbysk, 1950), the Miller their past geographic separation (also see Sneath & Mountain Formation of Nevada (Nelson, 1963), the Cow McKenzie, 1973). Mounting evidence from the distribu­ Head boulders of western Newfoundland (Young & Lud- tions of tri10bites (Palmer, 1973; Fortey, 1975; Taylor, 1976, 1977; Cook & Taylor, 1975; Taylor & Cook, 1976; Table 1. Distributions ofpolymeroid genera Ludvigsen, 1978; Taylor & Forester, 1979; Babcock & in the Glossopleura Zone of the Robison, 1989; Robison & Babcock, 1990) and other Henson Gletscher Formation organisms (Skevington, 1974; Sweet and Bergstrom, 1974; Bames & Fåhraeus, 1975; Bergstrom & Cames, TAXA AN AM GB AL CR OL 1976; Bergstrom, 1977; Fortey & Cocks, 1986) indicates that these assumptions are not always valid. Factors of Glossopleura walcotti G the marine environment other than distance have prob­ Kootenia nodosa G G GGG G Ogygopsis klotzi ably had an important infiuence on the degree of fauna1 G Syspacephalus sp. 3 GGG G resemblance between geographic regions. Such factors may inc1ude latitudina1 c1imatic gradients (e.g., Ekman, Observed geographic distributions of polymeroid genera (G) or 1953; Valentine, 1973) and differences between shelf genera and species (s) identified in the Glossopleura Zone of the seas and the deep sea (e.g., Hedgpeth, 1957; Bruun, 1957; Henson Gletscher Formation (hyphen indicates absence of a Menzies et al., 1973; Benson, 1975, 1988; Taylor, 1977). genus). Abbreviations of geographic regions are AN, northern The effects ofboth ofthese factors on fauna1 distributions Appalachia (western Newfoundland and Quebec); AM, middle Appalachia (New York and Pennsylvania); GB, Great Basin have been disregarded or accorded minor importance in (Utah, Nevada, and California); AL, east-central Alaska; CR, many studies of early Palaeozoic faunas. Also, they have Canadian Rocky Mountains; and OL, tectonic blocks other than Laurentia. 134 from restricted-she1f lithofacies (e.g., Robison, 1976; Table 2. Distributions ofpolymeroid genera Palmer & HaUey, 1979), but examination of collections in the Ptychagnostus gibbus Zone of the at the University of Kansas and the U.S. National Mu­ Henson Gletscher Formation seum of Natura1 History indicates that some species were more eurytopic than previous1y thought. In the Henson TAXA AN AM GB AL CR OL Gletscher Formation, G. walcotti is associated with such Bathyuriscus sp. GGG G G characteristic open-shelf taxa as Ogygopsis and Syspa­ BolaspidelIa sp. G GGG cephalus. Kootenia, with which it is also associated, is sp. G GGG G eurytopic (Robison, lJ)76). Syspacephalus sp. 2 G GGG Genera of po1ymeroids in GGU collection 298970 Zacanthoides sp. G G GGG from the Ptychagnostus gibbus Zone have strongest af­ Observed geographic distributions of polymeroid genera (G) finities with genera from open-she1f environments of identified in the Ptychagnostus gibbus Zone of the Henson Laurentia (Tab1e 2). Syspacephalus and Zacanthoides are Gletscher Formation (hyphen indicates absence of a genus). known only from Laurentia. Bolaspidella is known with Abbreviations of geographic regions are AN, northern Appalachia certainty from Laurentia (Robison, 1988) and the Precor­ (western Newfoundland and Quebec); AM, middle Appalachia dillera terrane of Mendoza, Argentina (Poulsen, 1960). (New York and Vermont); GB, Great Basin (Utab, Nevada, and The Precordillera terrane may have been derived from California); AL, east-central Alaska; CR, Canadian Rocky Mountains; and OL, tectonic blocks other than Laurentia. Laurentia (Rarnos et al., 1986; Robison, 1991). The re­ ported occurrence ofBolaspidella from Kashmir (Shah & Sudan, 1982) is probab1y in error (Robison, 1988). Ex­ vigsen, 1989), and the Kinzers Fonnation of Pennsylva­ cept for Zacanthoides, which is eurytopic, all the genera nia (Campbell, 1971). Each occurrence represents an are most common1y associated with open-shelf litho­ open-shelf setting near the ocean-facing side of a carbon­ fades. Olenoides inhabited open-shelf sites in Laurentia ate platform. Other occurrences of Ogygopsis (Nelson, but was also present in Siberia, Kazakhstania, and Gond­ 1963; Palmer & HaUey, 1979) indicate that the genus is wana. Kootenia is a eurytopic genus that is common in most common1y associated with p1atfonn-edge environ­ Laurentia, but 1ike some other dorypygids, it was rather ments (Palmer & HaUey, 1979). widely distributed in low palaeo1atitudes. Finally, Eo­ G. walcotti was first described from the Cape Wood discus was widespread in open-shelf settings around Fonnation of Ing1efield Land, North-West Greenland. much of the world. Functional analyses of morphology Later, abundant materia1 was reported from the Carrara and wide geographic distributions indicate that most eo­ Fonnation of Nevada and Ca1ifomia (Palmer & Hal1ey, discids were pe1agic or perhaps planktonic (JeU, 1975). 1979). Glossopleura was previous1y identified primari1y

Table 3. Distributions ofpolymeroid genera in the Ptychagnostus atavus Zone of the Henson Gletscher and Kap Stanton formations

TAXA AN AM GB CR AL CM SW EN SP AU

Bathyuriscus concavus G GGGG Corynexochus? sp. G G G Costadiscus minutus Dasometopus groenlandicus G G Eodiscus scanicus Hartshillia inflata G G Olenoides cf. O. convexus GGG G Opsidiscus longispinus G G Parasolenopleura aculeata G Solenopleurella transversa? G? G G? Syspacephalus sp. l G GGG

Observed geographic distributions of polymeroid genera (G) or genera and species (s) identified in the Ptychagnostus atavus Zone of the Henson Gletscher and Kap Stanton formations (hyphen indicates absence of a genus). Abbreviations of geographic regions are AN, northern Appalachia (western Newfoundland and Quebec); AM, middle Appalachia (New York and Pennsylvania); GB, Great Basin (Utab and Nevada); AL, east-central Alaska; CR, southern Canadian Rocky Mountains; CM, maritime Canada except for western Newfoundland; SW, Sweden; EN, England south of the Caledonian suture; SP, Siberian platform; and AU, Australia. 135

Table 4. Distributions ofpolymeroid genera in the Lejopyge laevigata Zone of the Kap Stanton Formation

TAXA AM GB SW NO EN CM BO BI SP AU

Anomocarina excavata s Centropleura angelini GG GGG GG G Centropleura loveni GG G G G G G Elyx trapezoidalis G G G G G Solenopleura bucculenta G? GG G

Observed geographic distributions ofpolymeroid genera (G) or genera and species (s) identified in the Lejopyge laevigata Zone of the Kap Stanton Formation (hyphen indicates absence of a genus). Abbreviations of geographic regions are AM, middle Appalachia (New York and Vermont); GB, Great Basin (Nevada); SW, Sweden; NO, Norway; EN, England south ofthe Caledonian suture; CM, maritime Canada except for western Newfoundland; BO, Bohemia; BI, Bennett Island; SP, Siberian platform; and AU, Australia.

This, together with the numerous pelagic agnostoids in coefficients can be tested for statistical significance by GGD 298970 (Robison, 1994), suggests that trilobites in using binomial probabilities (see Archer & Maples, 1987; this collection had unrestricted access to the open ocean. Maples & Archer, 1988). Confidence intervals estab­ Polymeroids in new collections (Babcock, 1994) repre­ lished for dusters on data from the L. laevigata and P. senting the Ptychagnostus atavus Zone are of mixed atavus zones are listed in Table 5. Laurentian and Baltic aspect (Table 3), and those in the To test further the reliability ofdusters, MDS was used Lejopyge laevigata Zone are entirely of Baltic aspect in a manner analogous to R mode. The rationale for using (Table 4). Pattems in the palaeogeographic distributions it is that CA can distort between-group relationships, of polymeroids occurring in these two zones in North although relationships within compact groups are likely Greenland were explored using multivariate methods. to be accurate (Kruskai, 1964a, 1964b; Hazel, 1977). Results, summarised in the following section, reveal a MDS distributes distortion evenly between large and strong distinction between major faunal associations small distances (Hazel, 1977). Lastly, fidelity and con­ (biofacies) and some subtle but important associations. stancy indices (Hazel, 1970, 1977; Hood & Robison, 1988) were calculated for all genera in principal dusters Multivariate analyses of trilobite collections as measures of how well they characterise a particular duster. A fidelity index measures the extent to which a Multivariate methods were used to explore pattems of taxon is confined to samples of a duster. Fidelity (F) is faunal association among polymeroid trilobites from se­ defined by lected localities in North Greenland, North America, Scandinavia, and Great Britain. Faunas analysed are from the Ptychagnostus atavus and Lejopyge laevigata zones. Binary (presence-absence) data were analysed using where Ni is the number of collections in each duster in duster analysis (CA) and multidimensional scaling (MDS). Analyses were performed using version 3 of the SYSTAT statisticai package (SYSTAT, Inc.) on an Apple Table 5. Confidence intervals for clusters of Macintosh II computer. Cluster analysis was used to trilobite presence-absence data separate data sets into discrete groups (see Hazel, 1970, 1977). In Q-mode analyses, objects (collections) were ZONE N P

Fig. 3. Cluster analysis of data on co c:i polymeroid collections from the Lejo­ pyge laevigata Zone of North Green­ land and selected localities in North A America, Scandinavia, and Great Bri­ tain. Data are plotted in order formed

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

Modocia r----- [ :::~:elJa Olønoldøs L..- [ EJrathla Bonnslørrlna KJngsfDnla II [ Hemlrhodon ------~ MsIøo18Spls ------[ Oe/18Cf1phalus Trlcrøplcephalus Corynexochus ...-______AcrocøphaHtøs [ Dorypyge Opsldlscus Dollchometopus • Bsillaspls • L- ..,-__ [ ==:'ldes Oasometopus Gronwall/a L- [ ~:~~es

Solønopløu18 L- [ ::mocarina

Cenfropleu18

Baltica, Gondwana, and peri-Gondwanan terranes. Ano­ low palaeolatitudes. Modocia is a common and diverse mocarina, Anomocarioides, and Dasometopus are known genus in Laurentia, and the only verified occurrence other from sites in Baltica and Siberia. Dasometopusjllld Ano­ than Laurentia is a single specimen from England (Rush­ mocarina are reported by Babcock (1994) from North ton, 1978). Greenland. Elyx has now been observed in Baltica, Sib­ Results using MDS (Fig. 4) are in close agreement eria, Acadia, and North Greenland. From published re­ with results of the CA. Genera that formed clusters in cords, Centropleura, Corynexochus, Opsidiscus, and 80­ R-mode analysis form discrete, rather compact groups in lenopleura all have relatively wide distributions, al­ two-dimensional MDS. The first dimension represents a though not necessarily during the L. laevigata Biochron. contrast between the 16 genera ofBaltic aspect and the Il Cluster II contains eleven genera which are restricted genera of Laurentian aspect. The second dimension pri­ to Laurentia except Olenoides and Modocia. Olenoides is marily represents a contrast between more widespread diverse in Laurentia but was also present in shelf seas of genera and less widespread ones. Siberia, Kazakhstania, China, and the Precordillera ter­ Fidelity and constancy indices for clusters (Fig. 3) are rane of Argentina. It, like some other genera assigned to listed in Table 6. Cluster I contains 16 genera that define the (Poulsen in Harrington et al., 1959: cluster A and have been collected from localities in Scan­ 0217-0219), seems to have been widely distributed in dinavia, Great Britain, or Peary Land, North Greenland 138 remarkable in view of the fact that section 3, where trilobites ofBaltic aspect were found, is only about 40 km 0.5 4 from the type section of the Holm Dal Formation in Peary 00 N 3 Land (Fig. 1). The Holm Dal Formation contains trilo­ 30 o C O - _2 o O bites primarily of Laurentian aspect (see Robison, 1988). 'u; 3- -3 02 C 20 Major tectonic dislocation has not been observed in this Q) -0,5 o o E region (Fletcher et al., 1988; Higgins et al., 1991). Rocks (5 of the Holm Dal Formation represent an inferred low -1 energy, open-shelf environment (Ineson, 1988); water depths were probably significantly less than those in -1.5 -1.5 -1 -0.5 O 0.5 1.5 2 2.5 which the sediments of the lower Kap Stanton Formation, as seen along J. P. Koch Fjord, were deposited. Dimension l Collections analysed from the P. atavus Zone include Fig. 4. Scattergram of coordinates of genera analysed by multi­ two from North Greenland (GGU 298969 and 319790; dimensional scaling. Taxa analysed are from the Lejopyge laevi­ Babcock, 1994), one from western North America gata Zone. Open circles correspond to polymeroid genera of (House Range, Utah; Robison, 1964 and unpublished Baltie aspect as defined in Q-mode cluster analysis, and closed data), two from eastem Canada (Cow Head Group, Que­ circles correspond to genera of Laurentian aspect. A numeral bec, Kindle, 1982; and south-eastem Newfoundland, next to a circle indicates the number of genera that are stacked vertically at that point. Table 6. Fidelity and constancy indices for (GGU 301313). All have high fidelity indices. Genera polymeroids from the Lejopyge laevigata Zone that also have high constancy indices are Agraulos, Ano­ mocarina, Bailiaspis, Centropleura, Elyx, Gronwallia, CLUSTER A CLUSTER B F C NF C N Paradoxides, and Solenopleura. Some genera appear to have been largely restricted to Baltica and England but, Modocia 10 10 3 as indicated above, this is an artifact of the localities that BolaspideIla 10 10 3 were chosen for analysis. Cluster II includes genera that Cedaria 10 10 3 define cluster B and that are most common in open-shelf Olenoides 10 7 2 lithofacies of North America. All have high fidelity and Elrathia 10 7 2 Bonneterrina 10 7 2 constancy indices. Except for Olenoides and Modocia, all Kingstonia 10 7 2 are restricted to Laurentia. Extra-Laurentian samples con­ Hemirhodon 10 7 2 taining these genera, however, were not included in the Meteoraspis 10 7 2 analysis. Deiracephalus 10 7 2 Trilobite samples from the L. laevigata Zone of the Tricrepicephalus 10 7 2 Kap Stanton Formation are interpreted to represent au­ Corynexochus 10 3 2 2 tochthonous associations of trilobites in deep, outermost Acrocephalites 10 3 Dorypyge 10 3 2 shelf or possibly upper slope lithofacies. Specimens are Opsidiscus 10 3 2 commonly articulated and fragile fossils preserved paral­ Dolichometopus 10 5 3 lel to bedding in dark, organic-rich lime mudstone. Dis­ Bailiaspis 10 7 4 articulated sclerites are also common, but are generally Anomocare 10 5 3 unbroken. Some specimens represent undisturbed Anomocarioides 10 5 3 moulted exoskeletons. Specimens range from small me­ Dasometopus 10 5 3 Gronwallia 10 7 4 raspides to large holaspides. Little post-mortem or post­ Agraulos 10 8 5 moulting transportation of sclerites is indicated. Fl)rther­ Paradoxides 10 8 5 more, the presence of undisturbed laminae indicates little Solenopleura 10 10 6 or no faunal mixing by bioturbation. All four genera and Elyx 10 8 5 four of five species present in these collections (Table 4) Anomocarina 10 8 5 are also present in coeval rocks of Scandinavia. Some Centropleura 10 8 5 genera or species are found in open-shelf lithofacies of Fidelity (F) and constancy (C) indices for poiymeroids in two southern Great Britain, Siberia, Kazakhstan, Australia, or Q-mode dusters shown in Fig. 3 from the Lejopyge laevigata Zone. China. None of the genera is present in coeval rocks of The number of collections containing a genus is iisted under N. A the Holm Dal Formation (see Robison, 1988). This is hyphen indicates the absence of a genus from a duster. 139 Hutchinson, 1962), three from Scandinavia (Norway, tains collections from the House Range, Dtah, and the Vlistergotland, and Bornholm; Westergård, 1946, 1953), Cow Head Group, western Newfoundland. and one from southern Great Britain (Thomas et al., GGD collections 298969 and 319790 from North 1984). The GGD collections include specimens in both Greenland and collections from boulders of the Cow mudstone and packstone. Genera preserved in both lithol­ Head Group of Newfoundland all consist of mixed as­ ogies were grouped together in this analysis because all semblages of polymeroids having Laurentian and Baltic that are present in mudstone are also present in pack­ aspect. All polymeroids in the GGD collections preserved stone. in mudstone are of Baltic aspect and probably represent CA in Q mode produced two primary clusters (Fig. 5, trilobites indigenous to the deep, outermost shelf litho­ A, B) that show major faunal differences between collec­ fades. Most polymeroids preserved in packstone are also tions containing polymeroids of Laurentian and Baltic of Baltic aspect. One of 5 genera (20 percent) in GGD aspect. The Euclidean distance coefficient between clus­ 298969 and 2 of 6 genera (33 percent) in GGD 319790, ters A and B is 0.76, which is considered to be of however, are of Laurentian aspect. Most polymeroids borderline statisticai significance at the 0.1 level (see from the Cow Head collections are of Laurentian aspect Table 5). Cluster A contains collections from Great Bri­ but 4 of 10 genera (40 percent) are ofBaltic aspect. A few tain, south-eastern Newfoundland, Scandinavia (Vlister­ trilobites of Laurentian or Baltic aspect may have coex­ gotland, Bornholm, Scania, and Norway), and North isted in some places, but the differences between taxa Greenland (GGD 298969 and 319790). Cluster B con- preserved in mudstone and packstone in North Greenland

«l ci ..... A ci

-q o B Q-mode clusters ..'"ci ci

""d

N ci

o R-mode clusters ci

Euclidean distance

Fig. 5. Cluster analysis of data on 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 polymeroid collections from the .--- [ O/enoidss Ptychagnostus atavus Zone of North Greenland and selected 10­ II .------1 [~~ calities in North America, Scandi­ _ navia, and Great Britain. Data are ___r--jL------' SyspacephaJusBolaspidelIa plotted in order formed by Q­ .--- [ Solenopleurella ,....--I Corynexochus I mode clustering of collections CosJadlscus and R-mode clustering of genera. Hartshlllla Anopolenus Primary clusters are labelled with ClarelIa Roman numerals and uppercase [ BaJllella Holocephallna letters. Cluster A contains genera .--- [ BalIlaspIs of Baltic aspect (except Oie­ MenevIella noides, Bathyuriscus, and Syspa­ CBnlropleura I III Eodlscus I cephaius), and cluster B contains I Parasolenopleura genera of Laurentian aspect. Paradoxidss 140 319790, and Cow Head Group) in the analysis. The

2 distance coefficient between dusters I + H and III is 0.82, o 4 Q.5 0 which is considered to be of borderline statisticai signif­ C\I o o icance at the 0.1 level (see Table 5). Cluster III contains C o Q 4. O O three genera that were evidently widespread. The duster 'en 2 C Q) has a high levelof dissimilarity from the other dusters -Q.5 • E 2 • o probably because the contained genera are widely distrib­ O o uted among localities that contain polymeroids of Baltic -1 o aspect.

-1.5 Results of MDS (Fig. 6) illustrate the relatively lower -1.5 -1 -Q.5 Q Q.5 1.5 2 2.5 distance coefficients between dusters containing poly­ meroids of mostly Baltic aspect (Fig. 5, A) and Lauren­ Dimension l tian aspect (Fig. 5, B), compared to samples analysed Fig. 6. Scattergram of coordinates of genera analysed by multi­ from the L. laevigata Zone. Genera that comprise dusters dimensional scaling. Taxa analysed are from the Ptychagnostus A and B form loose dusters. Most ofthe taxa in duster B atavus Zone. Open eircles eorrespond to polymeroid genera of have coordinates near those in duster A. Eodiscus, Para­ Baltie aspeet as defined in Q-mode cluster analysis, and closed doxides, and Parasolenopleura have coordinates that are circles eorrespond to genera of Laurentian aspeet. A numeral most distant from taxa in duster A. The first axis has next to a circle indicates the number of genera that are stacked been fitted primarily to the contrast between polymeroids vertically at that point. of Laurentian and Baltic aspect. The second axis is fitted primarily to the contrast between more widespread and suggest that biofacies of polymeroids were strongly dif­ less widespread genera. ferentiated. Fidelity and constancy indices for polymeroids from Stratigraphic and sedimentologic evidence suggest that trilobites in GGU collections 298969 and 319790 and collections from the Cow Head Group have been mixed Table 7. Fidelity and constancy indices for by sedimentary processes. All ofthese collections contain polymeroids from the Ptychagnostus atavus Zone disarticulated and fragmented trilobite sderites and occur in deeper-water successions that accumulated basinward CLUSTER A CLUSTER B of steep-rimmed, shallow-water carbonate platforms. The F C N F C N specimens from North Greenland that are preserved in Olenoides 7 10 3 packstone are crudely size-sorted, and are interpreted to Bathyuriscus 7 10 3 have been redeposited by sediment gravity flows (see Modocia 10 10 2 Ineson, 1980 and discussion in Ineson et al., 1994). The Alokistocare 10 10 2 Cow Head material was collected from limestone breccia Elrathia 10 10 2 that was inferred to have been deposited by gravity flow BolaspidelIa 10 10 2 (Hubert et al., 1977; Hiscott & James, 1985; James & Syspacephalus 5 5 2 Solenopleurella 10 3 2 Stevens, 1986), Boulders in the Cow Head Group repre­ Corynexochus 10 3 2 sent several depositionallithofacies (Kindle, 1982; James Costadiscus 10 3 2 & Stevens, 1986). Hartshillia 10 4 3 R-mode CA produced three primary dusters (Fig. 5, Anopolenus 10 3 2 l-III). Cluster I contains all oLthe genera of Baltic aspect ClarelIa 10 3 2 except Eodiscus, Parasolenopleura, and Paradoxides, Bailiella 10 3 2 Holocephalina 10 3 2 which form duster HI. Genera in dusters I and III are Bailiaspis 7 3 3 responsibIe for duster A produced in Q mode. Cluster H Meneviella 7 3 3 contains seven genera characteristic of open-shelf sites in Centropleura 5 1 2 Laurentia. These taxa are responsibie for duster B pro­ Eodiscus 9 9 8 duced in Q mode. The average Eudidean distance coeffi­ Parasolenopleura 10 8 6 cient between dusters I and H is 0.63. The null hypothe­ Paradoxides 10 8 6 sis that there is no difference between dusters I and H Fidelity (F) and constancy (C) indices for polymeroids in two cannot be rejected (P> 0.1; see Table 5). The rather low Q-mode clusters shown in Fig. 5 from the Ptychagnostus atavus distance coefficient is probably due mostly to the pres­ Zone. The number of collections containing a genus is listed under ence of three mixed collections (GGU 298969, GGU N. A hyphen indicates the absence of a genus from a cluster. 141 the P. a/al'lIS Zone are listed in Tahle 7. Except for Environmental relationships of trilobites Cel1/rop/eura, all 14 genera thaI produccd duster A have high fidelity indices. Those that also have high constancy Middle Cambrian strata af lhe Lo /aeviga/a and P. indices :Jre EodisClIS, Paradoxides, and Paraso/el1o­ atams zones ol' ~orth Greenland inc!ude biofacies dom­ pleum. All genera in duster B except Syspacephalus inated by polymeroids hLwing either Laurentian ar Baltic have high fidelity and con lancy values. Syspaceplwllls is aspect. Trilobiles ol' Baltie aspeel were evidently indige­ restricted to Laurentia, but di articulaled 'c1erites have nous to lhe deep outenno t heil' ar upper ba. in slope, b en found in ane mixed assemblage, GGU 298969, in whereas tho'e uf Laurenlian a peet inhabited open- heil' which mo t polymeroid are af Ballic a pecl. lilhofneies. Resedimentation ol' u'ilobite se!criles has rc- In summary, collections from packstone beds of lhe P. uhed in the mixing af trilobites af differelll aspeets in atal'lIS Zone ol' North Greenland (e.g., GGU 298969 and same bed af the P. CIIavus Zone (Fig. 7). 319790) eontain polymeroids ofboth Laurentian and Bal­ In recent years, f:Junnl differences in Cambrian rocks tic aspects. Most cxamined specimens are c1isartieularecl have been eommonly attributed to environmental gra­ and fragmenled, Polymeroicls in lhe inlervening mucl­ di nt· in whieh temperature played a major role (Cook & slones are entirely af Ballie a pecl, and tend to be found Taylor, 1975, 1977; Taylor, 1976, 1977; Taylor & Cook artieulated. Trilobites in mudstone are inlerpreted LO have 1976; Taylor & Forester, 1979; Theokritoff, 1979; Con­ heen indigenous to the deep outennost shelf. Those in way Morris & Ru hton, 19l5S; Shergold, 1988; Babcoek packstone, however, have prohahly been mixed by sedi­ & Robison, J989; Robi an & Babcock, 1990). Stratigra­ ment gravity-Jlow proces es inlroducing skeletal material phie and hiogeographie evidcncc has bcen prcscnled to typical ol' outer-shelf environmel1l' into the dceper-water show that same ocean were tJlermally stratified during elling. thc Early Cambrian (TheokritolT, 1979) and the Late Cam brian (e.g., T:Jylor & Forcstcr, 1979), just as they are during the Holocenc (c.g., Benson, 1975. 1988). The rcsults prcsclltcd in this paper suggc 'llhal ucean waters

TR/LOBITES OF LAURENT/AN ASPECT •••••••••••••• Autochthonous or Parautochthonous Remains Allochthonous Sclerites in Debris Beds

TR/LOBITES OF BALT/C ASPECT Autochthonous or Parautochthonous Remains Fig. 7. Modelof the Jnnuitian margin of Laurenlia orr present-day orth Greenland during the Middle Call1brian illustrating the relationships of physiography. sedimenlation pallerns, and lrilobite biofaeies. PoJymeroid trilobiles of Baltic aspect (represcnted by Cen/rap/eum. right) are infcrred to have lived beluw lhe permanent therlllocline. Fos. ils af Ballie trilobites are presefl-cd in lime mudSlone and are eommonly articulated. Polymeroids of Laurentian aspec.:l (represenled by Bmlzyuriscus. left) are inferred IO have lived in wann shallower water above the therllloeline but Iheir disanic.:ulaled remains (represellled by isolated .eJerites of fjmlryuris LIS) were dispersed basinward hy gravily displaccmenl of sediment and redeposited in debris beds. 142 surrounding Laurentia were also thermally stratified dur­ open-shelf to deep, outermost shelf (or possibly basin­ ing the Middle Cambrian (Fig. 7). Other postulated ex­ slope) environments along the Innuitian margin of Lau­ planations for biogeographic barriers, including a deep­ rentia. water basin (see Wilson, 1966; Cowie, 1971) and a land Second, indigenous deeper-water polymeroid faunas of barrier (Ulrich & Schuchert, 1902), were reviewed by North Greenland show strong dissimilarities with coeval Theokritoff (1979). Theokritoff rejected the idea that a faunas from the adjacent warm-water Laurentian shelf. deep-water basin was alone sufficient to act as a bioge­ Polymeroid faunas of the L. laevigata Zone that were ographic barrier on both uniformitarian arguments and indigenous to putative deep, outermost shelf environ­ evidence about the distribution of some Early Cambrian ments do not share any genera or species with a fauna trilobites. The notion of a land barrier was rejected on the reported from the coeval Holm Dal Formation (Robison, basis of biogeographic and stratigraphic evidence. 1988), which was deposited in an inferred low energy, The terms thermosphere and psychrosphere (Bruun, open-shelf setting (Ineson, 1988). Trilobites described 1957) describe the divisions of the world oceans that are from the type section of the Holm Dal Formation were respectively above and below the permanent thermocline. collected approximately 40 km south of locality 3 (Fig. In low and intermediate latitudes, warmer, less dense l), indicating that faunal changes between the shelf and waters of the thermosphere, which are characterised by slope were abrupt. temperatures higher than 10°C, float above cooler, more Shallower water faunas from the P. atavus Zone of dense waters of the psychrosphere, which are character­ North Greenland were not available for comparison. Nev­ ised by temperatures lower than 10°C (Bruun, 1957; ertheless, comparison of trilobites in mudstone and pack­ Benson, 1975). In present-day, high-latitude shelf areas, stone from GGU collections 298969, 319789, and the thermosphere is not present. 319790 indicate a strong faunal dissimilarity between Based on an analysis of modem isopod biofacies, Tay­ putative open-shelf-dwelling and deep, outermost shelf­ lor & Forester (1979) found that faunas indigenous to dwelling polymeroids. Some polymeroids found as rese­ warm shelves in low latitudes show strong similarities to dimented partieles in GGU collections are interpreted to other warm-water faunas and strong dissimilarities to have been derived from open-shelf settings where poly­ faunas in cool marine waters. Those from cool waters meroids of Laurentian aspect predominated. Two species, show a high degree of similarity whether they are in high Olenoides cf. O. convexus and Solenopleurella trans­ latitudes or in low latitudes below the permanent thermo­ versa, have been previously described from open-shelf eline. It is likely that thermal stratification of water sites elsewhere in Laurentia. Except for the eodiscid masses during parts of the early Palaeozoic similarly Costadiscus, which is known only from North Greenland, affected the biogeographic distributions of trilobites (see polymeroids in undisturbed mudstone are exelusively of Taylor & Forester, 1979). The mostly endernic shelf fau­ Baltic aspect. nas of Laurentia are inferred to have been adapted to Finally, polymeroids indigenous to putative deep, out­ warm water, whereas faunas described from shelf areas ermost shelf or upper slope environments of the L. laevi­ of Baltica and southem Gondwana or from peri-Gond­ gata and P. atavus zones in North Greenland are taxo­ wanan terranes located in south-polar regions are inferred nornically most sirnilar to faunas from shallow-shelf en­ to have been adapted to cool water. vironments of cool, high-latitude areas of Baltica, Taylor & Forester (1979) listed three criteria that they Acadia, Armorica, and England. Some taxa, including considered to be necessary to demonstrate the presence of Centropleura, Eodiscus, and solenopleurids, are also a two-layered thermally stratified Palaeozoic ocean. They known from deep-water Laurentian sites in north-westem are (1) evidence of deep-water environments, (2) strong Vermont (Howell, 1937; Shaw, 1966), south-eastem New biofacies differences between indigenous faunas of deep­ York (Rasetti, 1967) and central Nevada (Stewart & water environments and those of adjacent warm shelves, Palmer, 1967; Palmer & Stewart, 1968; Babcock, 1990). and (3) widely distributed deep-water faunas that show Indigenous deep-water polymeroids of the Henson greatest taxonornic resemblance to faunas in other deep­ Gletscher and Kap Stanton formations evidently did not water environments and shallow-water environments of also inhabit shallower water environments of the Lauren­ high latitudes. tian shelf. Where trilobites of mixed Laurentian and Bal­ Rocks and faunas of the L. laevigata and P. atavus tie aspect have been found in the same beds, evidence zones ofNorth Greenland meet Taylor & Forester's crite­ suggests that they were brought together by taphonomic ria. First, deep-water shelf environments, basinward of processes. the carbonate platform are represented by the Henson Polymeroids of the P. atavus Zone in Nyeboe Land Gletscher and Kap Stanton formations. Rocks from those support stratigraphic and structural evidence (Fletcher et units are interpreted (Ineson et al., 1994) to represent al., 1988; Higgins et al., 1991; Surlyk, 1991) that these 143 rocks are in situ and were not accreted by plate collision. commonly used as evidence of an origin in high latitudes. Minimal fragmentation and lack of size sorting suggest Some cool-water biofacies, however, are similar to each that specimens preserved in mudstone probably are au­ other regardless of latitude or depth and a suspect terrane tochthonous or parautochthonous to environments at con­ containing such trilobites could in fact have an origin tinental margins, including deep, outermost shelf and either in high latitudes or in deep water of low latitudes. perhaps slope areas. Specimens in packstone, which are Cool waters evidently were avenues for the dispersal of more fragmented and size sorted, are interpreted to have some organisms (Taylor & Forester, 1979). On this ac­ been redeposited. Many were probably carried basinward count, the degree of taxonomic dissimilarity among cool­ from proximal shelf sites by gravity flows. Analogous water faunas of different regions is not necessarily a deposits containing resedimented trilobites that were near function of the palaeogeographic distance between those the Laurentian continental margin have been described areas. Thus, conclusions about the latitudinal separation from the Upper Cambrian of central Nevada (Cook & of tectonic blocks during the geological past that are Taylor, 1975, 1977; Taylor, 1976; Taylor & Cook, 1976). based on differences between warm-water and cool-water There, redeposited warm-water trilobites of Laurentian biofacies are suspect. Conclusions about palaeolatitude aspect are interbedded with indigenous trilobites having based on faunas and lithofacies from shelf successions an Asian aspect. Other possibie continental slope deposits are more reliable, but palaeomagnetic evidence remains that contain reseaimented trilobites of mixed provincial the best indicator of the original latitudinal positions of affinities (Taylor & Forester, 1979) include some in the terranes. Taconic sequence of New York (Bird & Theokritoff, The presence of mixed assemblages of warm-water 1968; Bird & Rasetti, 1968), the Cow Head Group of and cool-water biofacies in the P. atavus Zone of North western Newfoundland (Babcock, 1991; see Kindle, Greenland illustrates a method by which autochthonous 1982), and the Collier Shale of Arkansas (Hart et al., terranes of the outer continental shelf or slope can be 1987; Hohensee & Stitt, 1989). identified (Babcock, 1991). Taphonomic, sedimentologic, and stratigraphic evidence needs to be examined before assemblages of mixed aspect can be determined to repre­ Geological implications sent organisms that either lived together or were redepo­ During the Middle Cambrian, separate polymeroid bio­ sited as bioclastic debris from ecologically separate bio­ facies developed in warm waters of the Laurentian open­ facies. In addition to improving interpretations of the shelf and in cool waters of deeper, outermost shelf to tectonic history of suspect terranes, study of mixed as­ uppermost basin or slope? areas. Most genera and almost semblages in continental shelf and slope environments all species found in open-shelf lithofacies were endemic may help to improve the resolution of biostratigraphic to Laurentia. Important exceptions are genera of the Do­ correlations between separate continents. rypygidae (especially Dorypyge, Kootenia and Ole­ noides) which are widespread in shelf facies of some Acknowledgements. Grateful acknowledgement to the many people tectonic plates that were in low palaeolatitudes. Most and institutions who made this study possibIe is given in the accom­ genera and species of polymeroids from deeper-water panying paper an polymeroid trilobites (Babcock, 1994) in this habitats of North Greenland seem to be the same as those volume. from cool, shallow-shelf sites of Baltica or peri-Gond­ wanan terranes. References The presence of polymeroids of Baltic aspect in Lau­ Archer, A. W. & C. G. Maples. 1987: Monte Carlo simulation of rentian rocks of Nyeboe Land and Peary Land, North selected binomial similarity coefficients (I): effect of number Greenland, has important implications for reconstructing af variables. Palaios 2, 609-617. the history of tectonic blocks during the early Palaeozoic. Babcock, L. E. 1990: Biogeography, phylogenetics, and system­ It is evident that the presence of trilobites of Baltic aspect atics af same Middle Cambrian trilobites from open-shelf to in a suspect terrane attached to a tectonic plate that was in basinallithofacies af North Greenland and Nevada. Unpub­ low palaeolatitudes is not sufficient evidence in itself to lished Ph.D. dissert., Univ. Kansas, 222 pp. conclude that the terrane is exotic. An exotic terrane is Babcock, L. E. 1991: Debris beds containing mixed faunal representatives from shelf-margin biofacies: a tool for use in one that has moved from a distant source and subse­ terrane analysis. Geo!. Soc. Am. Abstr. Prog. 23(3), 2. quently accreted to another tectonic plate. A native ter­ Babcock, L. E. 1994: Polymeroid trilobites from the Henson rane, by comparison, can include deep-shelf, slope or Gletscher and Kap Stanton formations (Middle Cambrian), basinal areas adjacent to or on the tectonic plate with North Greenland. Bul/. Grønlands geol. Unders. 169 (this which it is now associated. The presence of cool-water volume). trilobites of Baltic aspect in a suspect terrane has been Babcock, L. E. & Robison, R. A. 1989: Paleobiogeographic 144 significance of Middle Cambrian trilobites from North Green­ em Kazakhstan, Soviet Union. In Cooper, J. D. & Stevens, land. Geol. Soc. Am. Abstr. Prog. 21(6), A340. C. H. (ed.) Paleozoic paleogeography of the Western United Barnes, C. R. & Fåhraeus, L. E. 1975: Provinces, communities, States - II. Pacific section, SOC. Econ. Paleont. Mineral., and the proposed nektobenthic habit of Ordovician conodon­ 847-872. tophorids. Lethaia 8, 133-149. Cowie, J. W. 1971: Lower Cambrian faunal provinces. In Mid­ Beebe, M. A. 1989: Unusual preservation of Cambrian trilobites dlemiss, F. A., Rawson, P. F. & Newall, G. (ed.) Faunal prov­ in the Weeks Formation of Utah. Geol. Soc. Am. Abstr. Prag. inces in space and time. Spec. Publ. geol. SOC. Land. 4, 21(6), A71. 31-46. Benson, R. H. 1975: The origin of the psychrosphere as re­ Cowie, J. W. 1974: The Cambrian of Spitsbergen and Scotland. corded in changes of deep-sea ostracode assemblages. Leth­ In Holland, C. H. (ed.) Lawer Palaeozoic rocks ofthe world, aia 8, 69-83. Vol. 2, Cambrian ofthe British Isles, Norden and Spitsbergen, Benson, R. H. 1988: Ostracods and palaeoceanography. In De 125-155. London: J. Wiley & Sons. Decker, P., Colin, J.-P. & Peypouquet, J.-P. (ed.) Ostracoda in Ekman, S. 1953: Zoogeography of the sea, 417 pp. London: the earth sciences, 1-26. Amsterdam: Elsevier. Sidgwick and Jackson. Bergstrom, J. & Levi-Setti, R. 1978. Phenotypic variation in the Fletcher, T. P., Higgins, A. K. & Peel, J. S. 1988: A Balto­ Middle Cambrian trilobite Paradoxides davidis Salter at Scandian Middle Cambrian fauna from Peary Land, North Manuels, SE Newfoundland. Geolog. Palaeont. 12, 1-40. Greenland. Rapp. Grønlands geol. Unders. 137, 118 only. Bergstrom, S. M. 1977: Early Paleozoic conodont biostrati­ Fortey, R. A. 1975: Early Ordovician trilobite communities. graphy in the Atlantic borderlands. In Swain, F. M. (ed.) Fossils & Strata 4, 331-352. Stratigraphic micropaleontology of Atlantic basin and bor­ Fortey, R. A. & Cocks, L. R. M. 1986: Marginal faunal belts and derlands, 85-110. New York: Elsevier. their structural implications, with examples from the lower Bergstrom, S. M. & Carnes, J. B. 1976: Conodont biostrati­ Paleozoic. Jl geol. Soc. Land. 143, 151-160. graphy and paleoecology of the Holston Formation (Middle Gromet, L. P. 1989: Avalonian terranes and late Paleozoic tecto­ Ordovician) and associated strata in eastern Tennessee. In nism in southeastern New England; constraints and problems. Barnes, C. R. (ed.) Conodont paleoecology. Spec. Pap. geol. In Dallmeyer, R. D. (ed.) Terranes in the circum-Atlantic Assoc. Canada 15, 27-57. orogens. Spec. Pap. geol. Soc. Am. 230, 193-211. Bird, J. M. & Rasetti, F. 1968: Lower, Middle, and Upper Harrington, H. J. et al. 1959: Systematic descriptions. In Moore, Cambrian faunas in the Taconic sequence of eastern New R. C. (ed.) Treatise on invertebrate paleontology, part O, York: stratigraphic and biogeographic significance. Spec. Arthrapoda 1,0170-0540. Lawrence: Geol. Soc. Am. and Pap. geol. Soc. Am. 113, 66 pp. Kansas Univ. Press. Bird, J. M. & Theokritoff, G. 1968: Mode of occurrence of Hart, W. D., Stitt, J. H., Hohensee, S. R. & Ethington, R. L. fossils in the Taconic allochthon. Spec. Pap. geol. Soc. Am. 1987: Geological implications of Late Cambrian trHobites 101, 248 only. from the Collier Shale, Jessieville area, Arkansas. Geology Bruun, A. F. 1957: Deep sea and abyssal depths. In Hedgpeth, 15,447-450. J. W. (ed.) Treatise on marine ecology and paleoecology, Vol. Hazel, J. E. 1970: Binary coefficients and clustering in biostrati­ 1, Ecology. Mern. geol. Soc. Am. 67, 641-672. graphy. Bull. geol. Soc. Am. 81, 3237-3252. Burrett, C. & Richardson, R. 1980: Trilobite biogeography and Hazel, J. E. 1977: Use of certain multivariate and other tech­ Cambrian tectonic models. Tectonophys. 63, 155-192. niques in assemblage zonal biostratigraphy: examples using Campbell, L. D. 1971: Occurrence of 'Ogygopsis shale' fauna in Cambrian, , and Tertiary benthie invertebrates. In southeastern Pennsylvania. J. Paleont. 45, 437-440. Kauffman E. G. & Hazel, J. E. (ed.) Concepts and methods of Conway Moms, S. & Rushton, A. W. A. 1988: Precambrian to biostratigraphy, 187-212. Stroudsburg: Dowden, Hutchinson Tremadoc biotas in the Caledonides. In Harris, A. L. & & Ross. Fettes, D. J. (ed.) The Caledonian-Appalachian Orogen. Hedgpeth, J. W. 1957: Marine biogeography. In Hedgpeth, J. W. Spec. Publ. geol. Soc. Land. 38, 93-109. (ed.) Treatise on marine ecology and paleoecology, Vol. 1, Cook, H. E. & Taylor, M. E. 1975: Early Paleozoic continental Ecology. Mern. geol. SOC. Am. 67, 359-382. margin sedimentation, trilobite biofacies, and the thermo­ Higgins, A. K., Ineson, J. R., Peel, J. S., Surlyk, F. & Sønder­ eline, western United States. Geology 3, 559-562. ho1m, M. 1991: Lower Palaeozoic Franklinian Basin of North Cook, H. E. & Taylor, M. E. 1977: Comparison of continental Greenland. Bull. Grønlands geol. Unders. 160, 71-139. slope and shelf environments in the Upper Cambrian and Hiscott, R. N. & James, N. P. 1985: Carbonate debris flows, lowest Ordovician of Nevada. In Cook, H. E. & Enos, P. (ed.) Cow Head Group, western Newfoundland. J. sed. Petrol. 55, Deep-water carbonate environments. Spec. Publ. Soc. Econ. 735-745. Paleont. Mineral. 25,51-81. Hohensee, S. R. & Stitt, J. H. 1989: Redeposited Elvinia Zone Cook, H. E. & Taylor, M. E. 1989: Evolution of early Paleozoic (Upper Cambrian) trilobites from the Collier Shale, Ouachita carbonate seamount, Malyi Karatau Range, southern Kazakh­ Mountains, west-central Arkansas. J. Paleont. 63, 857-879. stan, USSR: newevidence for early history of Kazakhstania. Hood, K. C. & Robison, R. A. 1988: Trilobite and lithofacies Abstr. 28 int. geol. Congr., Washington D.C. 1, 322 only. relationships in the Holm Dal Formation (late Middle Cam­ Cook, H. E. et al. 1991: Comparison of two early Paleozoic brian), central North Greenland. Meddr Grønland Geosci. 20, carbonate submarine fans - western United States and south- 105-112. 145 Horton, J. w., Jr., Drake, A. A., Jr. & Rankin, D. W. 1989: Lochman-Balk, C. & Wilson, J. L. 1958: Cambrian biostrati­ Tectonostratigraphic terranes and their Paleozoic boundaries graphy in North America. J. Paleont. 32, 312-350. in the central and southern Appalachians. In Dallmeyer, R. D. Ludvigsen, R. 1978: Middle Ordovician trilobite biofacies, (ed.) Terranes in the circum-Atlantic orogens. Spec. Pap. southern Mackenzie Mountains. Spec. Pap. geol. Assoc. Can. geol. Soc. Am. 230, 213-245. 18, 1-37. Howell, B. F. 1925: The faunas of the Cambrian Paradoxides Maples, C. G. & Archer, A. W. 1988: Monte Carlo simulation of beds at Manuels, Newfoundland. Bulls Am. Paleont. 11(43), selected binomial similarity coefficients (II): effect of sparse 140 pp. data. Palaios 3, 95-103. Howell, B. F. 1937: Cambrian Centropleura vermontensis fauna McKerrow, W. S. & Cocks, L. R. M. 1976: Progressive faunal of northwestern Vermont. Bull. geol. Soc. Am. 48, 1147­ migration across the lapetus Ocean. Nature 263, 304-306. 1210. McLaughlin, K. P. & Enbysk, B. B. 1950: Middle Cambrian Hubert, J. F., Suchecki, R. K. & Callahan, R. K. M. 1977: The trilobites from Pend Oreille County, Washington. J. Paleont. Cow Head Breccia: sedimentology of the Cambro-Ordovician 24, 466-471. continental margin, Newfoundland. In Cook, H. E. & Enos, P. Menzies, R. J., George, R. Y. & Rowe, G. T. 1973: Abyssal (ed.) Deep-water carbonate environments. Spec. Publ. Soc. environment and ecology af the world oceans, 488 pp. New econ. Paleont. Mineral. 25, 125-154. York: Wiley-Interscience. Hutchinson, R. D. 1952: Middle Cambrian ofthe Atlantic realm Nelson, C. A. 1963: Stratigraphic range of Ogygopsis. J. Pale­ in eastern Gaspe. Am. J. Sci. 250, 275-280. ont. 37, 244-248. Hutchirison, R. D. 1962: Cambrian stratigraphy and trilobite bpik, A. A. 1979: Middle Cambrian agnostids: systematics and faunas of southeastern Newfoundland. Bul!. geol. Surv. Cana­ biostratigraphy. Bul!. Bur. Miner. Resour. Austral. 172; vol. da 88, 156 pp. 1, 188 pp.; vol 2, 67 pis. Ineson, J. R. 1980: Carbonate debris flows in the Cambrian of Palmer, A. R. 1969: Cambrian tri10bite distributions in North south-west Peary Land, eastern North Greenland. Rapp. America and their bearing on Cambrian paleogeography in Grønlands geol. Unders. 99, 43-49. Newfoundland. In Kay, M. (ed.) North Atlantic - geology and Ineson, J. R. 1988: Lithostratigraphy and depositional setting of continental drift. Mern. Am. Assoc. petrol. Geol. 12, 139-148. the Holm Dal Formation (Middle Cambrian), central North Palmer, A. R. 1972: Problems of Cambrian biogeography. Proc. Greenland. Meddr Grønland Geosci. 20, 9-21. 24 int. geol. Congr., Montreal 7,310-315. Ineson, J. R., Surlyk, F., Higgins, A. K. & Peel, J. S. 1994: Slope Palmer, A. R. 1973: Cambrian trilobites. In Hallam, A. (ed.) apron and deep shelf sediments of the Brønlund Fjord and Atlas ofpalaeogeography, 3-11. Amsterdam: Elsevier. Tavsens Iskappe Groups (Lower Cambrian - Lower Ordo­ Palmer, A. R. 1979: Cambrian. In Robison, R. A. & Teichert, C. vician) in North Greenland: stratigraphy, facies and deposi­ (ed.) Treatise on invertebrate paleontology, part A, biogeog­ tional setting. Bull. GrØnlands geol. Unders. 169 (this vol­ raphy and biostratigraphy, AI19-A135. Lawrence: Geo!. urne). Soc. Am. & Kansas Univ. Press. James, N. P. & Stevens, R. K. 1986: Stratigraphy and correlation Palmer, A. R. & Halley, R. B. 1979: Physical stratigraphy and of the Cambro-Ordovician Cow Head Group, western New­ trilobite biostratigraphy of the Carrara Formation (Lower and foundland. Bull. geol. Surv. Canada 366, 143 pp. Middle Cambrian) in the southern Great Basin. Prof Pap. Jell, P. A. 1974: Faunal provinces and possibie planetary recon­ u.S. geol. Surv. 1047, 1-131. struction of the middle Cambrian. J. Geol. 82,319-350. Palmer, A. R. & Peel, J. S. 1981: Dresbachian trilobites and Jell, P. A. 1975: Australian Middle Cambrian eodiscoids with a stratigraphy of the Cass Fjord Formation, western North review of the superfamily. Palaeontographica A 150, 1-97. Greenland. Bull. Grønlands geol. Unders. 141, 46 pp. Kaesler, R. L. 1966: Quantitative re-evaluation of ecology and Palmer, A. R. & Stewart, J. S. 1968: A paradoxidid trilobite distribution of recent Foraminifera and Ostracoda of Todos from Nevada. J. Paleont. 42, 177-179. Santos Bay, Baja California, Mexico. Paleont. Contr. Univ. Peel, J. S. (ed.) 1988: Stratigraphy and palaeontology of the Kansas Pap. 10, l-50. Holm Dal Formation (Iate Middle Cambrian), central North Keppie, J. D. 1989: Northern Appalachian terranes and their Greenland. Meddr Grønland Geosci. 20, 168 pp. accretionary history. In Dallmeyer, R. D. (ed.) Terranes in the Poulsen, C. 1927: The Cambrian, Ozarkian and Canadian faunas circum-Atlantic orogens. Spec. Pap. geol. soc. Am. 230, 159­ of northwest Greenland. Meddr Grønland 70, 233-343. 192. Poulsen, C. 1960: Fossils from the late Middle Cambrian Bo­ Kindle, C. H. 1982: The C. H. Kindle collection: Middle Cam­ laspidella Zone of Mendoza, Argentina. Mat.-Fys. Medd. brian to Lower Ordovician trilobites from the Cow Head Danske Vid. Selsk. 32, 42 pp. Group, western Newfoundland. Pap. geol. Surv. Can. 82-1C, Poulsen, V. 1964: Contribution to the Lower and Middle Cam­ 1-17. brian paleontology and stratigraphy of northwest Greenland. Kruskai, 1. B. 1964a: Muitidimensional scaling by optimizing Meddr Grønland 164(6), 105 pp. goodness of fit to a nonmetric hypothesis. Psychometrika 29, Poulsen, V. 1969: An Atlantic Middle Cambrian fauna from 1-27. North Greenland. Lethaia 2, 1-14. Kruskal, J. B. 1964b: Nonmetric multidimensional scaling: a Poulsen, V. & Anderson, M. M. 1975: The Middle-Upper Cam­ numerical method. Psychometrica 29, 28-42. brian transition in southeastern Newfoundland, Canada. Can. J. Earth Sci. 12,2065-2079. 146

Ramos, V. A., Jordan, T. E., Allmendinger, R. W, Mpodozis, C., brian of Kashmir and its stratigraphic significance. J. geol. Kay, S. M., Cortes, J. M. & Palma, A. 1986: Paleozoic ter­ Soc. India 23, 236--245. ranes of the central Argentine-Chilean Andes. Tectonics 5, Shaw, A. B. 1966: Paleontology of northwestern Yermont. XI. 855-880. Fossils from the Middle Cambrian St. Albans Shale. J. Pale­ Rasetti, F. 1951: Middle Cambrian stratigraphy and faunas of ont. 40, 843-858. the Canadian Rocky Mountains. Smithson. misc. Collns 116 Shergold, J. H. 1988: Review of trilobite biofacies distributions (5),277 pp. at the Cambrian-Ordovician boundary. Geol. Mag. 125,363­ Rasetti, F. 1967: Lower and Middle Cambrian trilobite faunas 380. from the Taconic sequence of New York. Smithson. misc. Skevington, D. 1974: Controls influencing the composition and Collns 152(4), 111 pp. distribution of Ordovician graptolite faunal provinces. Spec. Robison, R. A. 1964: Late Middle Cambrian faunas from west­ Pap. Palaeontology 13, 59-73. ern Utah. J. Paleont. 38, 510-566. Smith, A. G., Hurley, A. M. & Briden, J. C. 1981: Phanerozoic Robison, R. A. 1972: Mode of life of agnostid trilobites. Proc. paleocontinental world maps, 102 pp. New York: Cambridge 24 int. geol. Congr., Montreal 7, 33--40. Univ. Press. Robison, R. A. 1976: Middle Cambrian trilobite biostratigraphy Sneath, P. H. A. & McKenzie, K. G. 1973: StatisticaI methods of the Great Basin. Brigham Young Univ. geol. Stud. 23(2), for the study of biogeography. Spec. Pap. Palaeontology 12, 93-109. 1--42. Robison, R. A. 1984: Cambrian of North America Staff, G. M., Stanton, R. J., Jr., Powell, E. N. & Cummins, H. and Greenland, Part I, Ptychagnostidae. Paleontol. Contr. 1986: Time-averaging, taphonomy, and their impact on pale­ Univ. Kansas Pap. 109, 59 pp. ocommunity reconstruction: death assemblages in Texas Robison, R. A. 1988: Trilobites ofthe Holm Dal Formation (late bays. Bull. geol. Soc. Am. 97, 428--443. Middle Cambrian), central North Greenland. Meddr Grøn­ Stewart, J. H. & Palmer, A. R. 1967: Callaghan window - a land Geosci. 20,23-103. newly discovered part of the Roberts thrust, Toiyabe Range, Robison, R. A. 1991: Cambrian period. In Goetz, P. W (ed.) The Lander County, Nevada. Prof Pap. U.S. geol. Surv. 575D, new Encyclopaedia Brittanica, 785-792. (15th edit.). Yol. 19. D56--D63. Macropaedia. Chicago. Surlyk, F. 1991: Tectonostratigraphy of North Greenland. Bul!. Robison, R. A. 1994: Agnostoid trilobites from the Henson Grønlands geol. Unders. 160, 25--47. Gletscher and Kap Stanton formations (Middle Cambrian), Sweet, W C. & Bergstrom, S. M. 1974: Provincialism exhibited North Greenland. Bull. Grønlands geol. Unders. 169 (this by Ordovician conodont faunas. In Ross, C. A. (ed.) Paleoge­ volume). ographic provinces and provinciality. Soc. econ. Paleontol. Robison, R. A. & Babcock, L. E. 1990: Biogeography ofMiddle Mineral. Spec. Publ. 21, 189-202. Cambrian trilobites from North Greenland: implications for SYSTAT, Inc. 1986: SYSTAT: the system for statistics. terrane analysis. 3 int. Symp. Cambrian System, Novosibirsk, Taylor, M. E. 1976: Indigenous and redeposited trilobites from USSR, Abstr. 151. Late Cambrian basinal environments of central Nevada. J. Rushton, A. W. A. 1978: Fossils from the Middle-Upper Cam­ Paleont. 50, 668-700. brian transition beds in the Nuneaton district. Palaeontology Taylor, M. E. 1977: Late Cambrian of western North America: 21, 245-283. trilobite biofacies, environmental significance, and biostra­ Samson, S., Palmer, A. R., Robison, R. A. & Secor, D. T., Ir. tigraphic implications. In Kauffman, E. G. & Hazel, J. E. 1990: Biogeographical significance of Cambrian trilobites (ed.) Concepts and methods of biostratigraphy, 397--425. from the Carolina slate belt. Bull. geol. Soc. Am. 102, 1459­ Stroudsburg: Dowden, Hutchinson & Ross, Inc. 1470. Taylor, M. E. & Cook, H. E. 1976: Continental shelf and slope Scotese, C. R. 1987: Phanerozoic plate tectonic reconstructions. facies in the Upper Cambrian and lowest Ordovician of Ne­ Paleoceanogr. map. Proj., Univ. Texas, Inst. Geophys., Tech. vada. Brigham Young Univ. geol. Stud. 23(2), 181-214. Rept 40, 54 pp. Taylor, M. E. & Forester, R. M. 1979: Distributional model for Scotese, C. R., Bambach, R. K., Barton, c., Yan der Yoo, R. & marine isopod crustaceans and its bearing on early Paleozoic Ziegler, A. M. 1979: Paleozoic base maps. J. Geol. 87, 217­ paleozoogeography and continental drift. Bull. geol. SOG. Am. 277. 90, 405--413. Scotese, C. R. & Denham, C. R. 1988: User's manual for Terra Theokritoff, G. 1979: Early Cambrian provincialism and bioge­ Mobilis: plate tectonics for the Macintosh, 43 pp. Houston: ographic boundaries in the North Atlantic region. Lethaia 12, Earth in Motion Technologies. 281-295. Scotese, C. R. & McKerrow, W S. 1990: Revised world maps Theokritoff, G. 1985: Early Cambrian biogeography in the and introduction. In McKerrow, W. S. & Scotese, C. R. (ed.) North Atlantic region. Lethaia 18, 283-293. Palaeozoic palaeogeography and biogeography. Mem. geol. Thomas, A. T., Owens, R. M. & Rushton, A. W A. 1984: SOC. Land. 12, 1-21. Trilobites in British stratigraphy. Spec. Rept geol. SOC. Lond. Secor, D. T., Jr., Samson, S. L., Snoke, A. W & Palmer, A. R. 16,78 pp. 1983: Confirmation of the Carolina slate belt as an exotic Torsvik, T. H., Ryan, P. D., Trench, A. & Harper, D. A. T. 1991: terrane. Science 221, 649-650. Cambrian-Ordovician paleogeography of Baltica. Geology Shah, S. K. & Sudan, C. S. 1982: Bolaspidella from the Cam- 19, 7-10. 147

Ulrich, E. O. & Schuchert, C. 1902: Pa1eozoic seas and barriers accretionary history' of the Appa1achian orogen. Geology 10, in eastem North America. Bul!. N. Y. State Mus. 52, 633-663. 530-536. Valentine, J. W. 1973: Evolutionary paleoecology oIthe marine Wi1son, J. T. 1966: Did the Atlantic dose and then reopen? biosphere, 511 pp. New Jersey: Prentice-Ha11. Nature 211, 676-681. Westergård, A. H. 1946: Agnostidea of the Midd1e Cambrian of Young, G. A. & Ludvigsen, R. 1989: Mid-Cambrian tri10bites Sweden. Sver. geo!. Unders. C 477, 141 pp. from the lowest part of the Cow Head Group, western New­ Westergård, A. H. 1953: Non-agnostidean tri10bites of the Mid­ foundland. Bul!. geo!. Surv. Can. 392, 49 pp. dIe Cambrian of Sweden III. Sver. geo!. Unders. C 526, 59 Young, G. C. 1986: Cladistic methods in Pa1eozoic reconstruct­ pp. ion. J. Geol. 94, 523-537. Williams, H. & Hatcher, R. D., Ir. 1982: Suspect terranes and