Canada's diverse continental landmass has long been recognized as a collage of distinct orogenic provinces (e.g. Price and Douglas 1972). Work over the past twenty years during the tenure of the LITHOPROBE program has revealed intriguing similarities and differences in orogens formed during the four billion year record of geological time. Armed with common geological, geophysical, geochemical and geochronological observations from each orogen, Earth scientists can now focus on key orogenic parameters to explore how crust- and mantle-forming processes may have changed over time. At a recent PanLITHOPROBE workshop, representatives working on the major orogenic provinces of Canada investigated by LITHOPROBE (Cordillera, Appalachians, Grenville, Paleoproterozoic, Superior, Slave orogens; Fig. 1) met to distill and synthesize information from their respective areas of expertise, and to compare orogenic evolution over large tracts of ground and a wide swath of geological time.

Participants were provided with a list of parameters considered to be significant in describing orogenic structure and history, and were asked to compile characteristics (in bold below) of their orogen of interest within this common template. Beginning with physical parameters, the list also included important aspects of orogenic and postorogenic evolution. Crustal thickness and its variation describe the end result of major crust-forming events, as well as evidence of "scars" produced during the process. Crustal structure, composition, seismic velocity and reflectivity define many first-order characteristics of the orogenic crust. Mantle lithospheric structure and composition incorporate information from geophysical and xenolith studies constraining the nature of the shallow orogenic mantle. The nature of orogen margins (e.g. continents) is important in assessing whether orogens had "rigid backstops" and conversely, the degree of orogenic reworking. The age and style of rifting describe the nature of the continental breakup phase preceding orogeny. The nature and age of accreted terranes provide insight into processes and paleogeography preceding terrane accretion. The nature and age of continental magmatic arcs give information on the character and longevity of subduction preceding collision. The location and nature of paleo-suture zones point to zones where oceanic crust has been consumed. The age and style of accretion , including collisional geometry explore the three-dimensional form (thin-skinned vs thick-skinned), kinematics and duration of collisional processes. Current or paleotopography and evolution of P-T conditions through time describe vertical crustal movements during orogenesis as well as the movement of heat through the crust. The age and style of post-collisional processes provide insight into the final processes that have shaped orogenic crust. Variation in the development and/or preservation of these parameters explains many of the differences between Canada's orogenic provinces.

Each working group compiled background material in advance of the meeting, then spent one to two days preparing a short synthesis (Intra-orogen correlations) presented in a plenary session. The six presentations and accompanying discussions occupied most of a day and led to a list of about 50 discussion topics. These were assembled and digested by a panel (Chris Beaumont, Jeremy Hall, Karl Karlstrom, Ray Price), who then led a discussion on comparative orogenic anatomy on the final day of the workshop.

F.A. Cook (on behalf of the Cordillera working group: S. Carr, R. Clowes,D. Francis, L. Hollister, R. Hyndman, A. Jones, R. Price, G. Ross, D. Symons, D. Thorkelson)

The Canadian portion of the Cordillera in western North America is one of the classic orogens of the world. It is bounded on the east by a stable craton that was extended during Proterozoic and Paleozoic rifting events and on the west by the actively deforming convergent/transform margin of the eastern Pacific Ocean (Gabrielse and Yorath, 1991). The Cordillera is thus the youngest orogen in which LITHOPROBE has worked and portions of two major transects have been devoted to studying it (Southern Canadian Cordillera Transect, 1984– 1995; Cook, 1995, and Slave NORthern Cordillera Lithospheric Evolution, or SNORCLE, transect, 1995– 2003; Fig. 1), thus providing a basis for comparisons along strike. Geomorphologic belts are used for simplicity in referencing the descriptions (Fig. 2).

In the southern Cordillera(Price, 1981), the crustal thickness varies from >40 km in the Foreland Belt to 33 to 35 km beneath the Omineca Belt. The transition occurs in the vicinity of the southern Rocky Mountain Trench (Fig. 2), west of which the thickness remains uniform to the Coast Belt, where it thickens again by a few kilometres. The northern Cordillera exhibits the same gross crustal structure, except that the thickest crust occurs ~100 to 150 km east of the deformation front. The 33 to 35 km crustal thickness of the interior southern Cordillera is commonly explained as a consequence of Paleocene– Eocene extension and associated uplift of mid-crustal rocks. The explanation for the thin crust in the northern Cordillera is more problematic as regional late extension has not been recognized there and the regional metamorphic grades of surface rocks do not indicate similar large uplifts. Estimates of lithospheric thickness range from 50 to 60 km in the Intermontane Belt, where high heat flow is interpreted to represent high temperatures at depth, to >100 km in the Foreland Belt and east of the Cordillera.

The oldest rocks in the Cordillera (~2.1 Ga) are present in uplifts that were exposed as a result of post-contractional extension and provide a temporal link with the sub-Phanerozoic rocks of the western Canadian Shield. These are generally interpreted as remnants of the cratonic crust that was rifted during Mesoproterozoic extension, the eastern limit of which may be reflected in a prominent series of west-facing crustal-scale homoclines beneath the Foreland Belt. The nature and style of the earliest rifting is preserved in thick strata of the Wernecke Supergroup (ca. 1.8– 1.7 Ga) in the north and the Belt-Purcell Supergroup (ca. 1.5– 1.4 Ga) in the south, and may have been diachronous from north to south. Whether these rocks were deposited in deep, intracratonic basins, or whether they were open to large oceans is still debated. There is evidence for periods of pre-Cordilleran contractional deformation within these and other Precambrian strata throughout the Cordillera, but the regional extent and paleogeographic significance of such deformation are uncertain. Neoproterozoic– Cambrian rifting initiated the development of the westward-thickening Cordilleran miogeocline.

Continental basement and cover rocks are traced in the crust beneath much of the Cordillera. In the north, the SNORCLE reflection data have been interpreted as having images of the thick supracrustal strata and their basement projecting as far west as the British Columbia– Alaska border (Cook et al., unpublished work). In the south, some basement rocks that were metamorphosed to amphibolite grade are now exposed as a result of post-contractional extension (e.g., Parrish et al., 1988). Seismic reflection data from the southern Cordillera have been interpreted as indicating that North American basement tapers westward as far as the Fraser fault (Cook, 1995).

For the most part, therefore, accreted terranes appear to be thin and confined to the crust. In some areas of the northern Cordillera, the surface geology and seismic profiles indicate terrane flakes that are only a few kilometres thick; other terranes may comprise much of the crust. In contrast to many other orogens, however (e.g., Newfoundland Appalachians), ophiolites are either nonexistent or only fragmentary throughout the Cordillera. Terranes are thus defined by distinct paleontological, stratigraphic, structural, metamorphic, and igneous histories, but suture zones are commonly manifested solely as faults. In some regions, the stratigraphic relationships between the North American miogeoclinal rocks (Neoproterozoic– Early Mesozoic) and adjacent accreted terranes are ambiguous. If strata are correlative from one to the other, such relationships may indicate there was no major terrane; if they are not correlative, the accretion boundaries are cryptic.

Magmatic activity has punctuated the development of the Cordillera since the Paleoproterozoic. Evidence of early magmatic episodes is a key tool for regional correlations of Proterozoic stratigraphic sequences because both the igneous rocks and their detritus provide age constraints. Paleozoic magmatism was sparse along the North American margin, but many of the accreted terranes have distinct magmatic signatures. Regionally extensive magmatic activity accompanied two major accretionary episodes, in the Jurassic– Early Cretaceous and the Late Cretaceous– Paleocene, during which widespread arc magmas and granites intruded through both accreted rocks and underlying North American rocks (e.g., Guichon Creek, Cassiar, etc.). Post-contraction regional extension in the south during the Eocene was accompanied by the last phase of orogen-scale magmatism.

Distinct differences in geological evolution between the northern Cordillera and the southern Cordillera date back to the Mesoproterozoic record of diachronous extension. They are, however, emphasized by post-collisional features such as variations in structural evolution, magmatism, level of exposure, and neotectonics. For example, strike-slip faults, such as the Tintina-Northern Rocky Mountain Trench, were associated with transcurrent displacement following accretion in the north. This motion was probably transformed southward into coeval contractional structures of the Foreland Belt. In contrast, regional lithospheric extension (Paleocene– Eocene) in the south was partly coeval with, and/or followed by, limited-strike slip faulting. Throughout the Cordillera, therefore, the magnitude of orogen-parallel terrane displacement (e.g., "Baja BC") is at present unresolved, large estimates (1000– 3000 km) having been determined from paleomagnetic measurements, and considerably more modest estimates determined from visible structures.

Post-collisional Late Tertiary to Holocene mafic magmatism occurs in several localities throughout the Canadian Cordillera. The largest centres, which may be associated with back-arc extension, are located in northern British Columbia and the Yukon. The widespread small centres are more likely related to fractures. Significantly, mantle xenoliths throughout the Cordillera have Proterozoic Rb/Sr and Re/Os ages, indicating that whatever has happened to the crust during accretion and contraction, the underlying old mantle is still present.

Along-strike differences between north and south are also apparent in some geophysical signatures such as the electrical conductivity of the crust (more resistive along the SNORCLE transect profiles than along the southern Cordillera profiles), heat flow (generally higher in the north), and Neogene tectonism (more earthquakes in the north). Some of these differences may be due to variations in the initial state of the Proterozoic margin; some may be due to variations in convergence during orogenesis, and some may be due to variations in convergence today. A challenge for continuing analyses of the LITHOPROBE and other data sets will be to determine the relative importance of each.

C. R. van Staal (on behalf of the Appalachian working Group:C. Beaumont, S. Castonguay, J. Hall,R. Jamieson, G. Jenner, S. Lin,B. Murphy and A. van der Velden)

The Canadian Appalachians (Maritimes and Newfoundland) are an example of a long-lived (ca. 250 Ma) Paleozoic accretionary orogen, punctuated by accretion of small oceanic terranes and four major oblique collisional events: Taconic, Salinic, Acadian and Alleghenian. The first three orogenic events are mainly the result of the successive arrival of four micro-continents (Dashwoods, Ganderia, Avalonia and Meguma) at the Laurentian margin (Fig. 3), which was thus expanding oceanwards with time (van Staal et al., 1998; Waldron and van Staal, 2001 and references therein), whereas the Alleghenian orogeny was due to continent collision during final closure of the Rheic Ocean. Alleghenian deformation in Newfoundland is mainly concentrated along major strike-slip faults (e.g. Cabot Fault) and to a first approximation can be ignored elsewhere on the island.

The Canadian Appalachians (Fig. 3) are imaged with deep seismic reflection lines acquired along several marine transects and two major on-land profiles across Newfoundland (Fig. 1), the results of which were earlier discussed by Marillier et al. (1989), Quinlan et al. (1992) and Hall et al. (1998). The two on-land transects(Fig. 4) were reprocessed by van der Velden and Cook (2002a).

The Canadian Appalachians are bounded to the west by the Laurentian craton , where little- or undeformed Paleozoic sediments overlie a dominantly Grenville-age basement. The Humber Zone (Williams, 1979) represents the peripheral part of the Laurentian craton that was involved in Appalachian orogenesis.

Crustal thickness in the Canadian Appalachians varies from 45 km beneath the Laurentian craton to 30 to 35 km beneath the Central Mobile Belt (Dunnage and Gander zones), to 40 km beneath the Avalon Zone. Thus the crust is the thinnest beneath the CMB, which was most affected by deformation, metamorphism and magmatism. Since evidence for syn- or post-accretion extension in the CMB is sparse, this feature may be due to the pre-accretion, thin nature of the crust of the oceanic terranes and micro-continents amalgamated in the CMB (Ellis et al., 1998). The Meguma Zone, absent in Newfoundland, has a thickness of about 30 to 35 km in Nova Scotia. The Moho beneath the Newfoundland CMB displays both across- and along-strike relief. Beneath the Gander zone the Moho shallows and reflective upper mantle layers appear to thin along strike towards the northeast, coinciding with an offshore Carboniferous basin in Notre Dame Bay (Chian et al., 1998). In addition, there is a marked lateral increase in seismic velocity in the lower crust along the same section, implying the presence of mafic rocks below this basin. These structures suggest a relationship between localized Carboniferous extension and mafic underplating. Hence, if correct, the seismic Moho beneath the CMB is largely a deformed, pre-Carboniferous structure. Other high-velocity layers possibly related to underplating occur at the edge of the Laurentian craton and beneath the Magdalen Basin southwest of Newfoundland. Prominent, west-dipping reflections of Gander Zone lower crust continue from the southeastern end of the transect at least as far west as the Red Indian line (Fig. 3) and project into the mantle (Fig. 4). The mantle reflections have been interpreted as subduction scars formed during Siluro-Devonian, west-directed underthrusting of Ganderia and Avalonia beneath Laurentia (van der Velden and Cook, 2002a). This is consistent with geometrical and relative age relationships between the reflection fabrics and known Siluro-Devonian faults. Interpretations of the origin and age of the prominent lower crustal reflectivity fabric beneath Ganderia range from largely due to syn-collisional shearing (e.g. Ellis et al., 1998) to an inherited, pre-collisional layering (van der Velden and Cook, 2002a).

Evidence of a major rifting and opening of an ocean is well preserved in the Humber Zone. Rift-related clastic sediments thicken toward the east, and associated rift-related magmatism took place intermittently between 615 and 550 Ma. A transgressive sequence overlying the rift clastics indicates a rift-drift transition near the base of the Cambrian (ca. 540 Ma). The rift-drift event relates to the departure of the peri-Laurentian Dashwoods micro-continent rather than creation of the Iapetus Ocean, which is purported to have opened earlier (Waldron and van Staal, 2001). Evidence related to the age and style of rifting of Ganderia, Avalonia and Meguma from Gondwana is poorly preserved in the sedimentary record, although an independent drift across the ocean(s) is suggested by geological, fossil and paleomagnetic data.

The nature and age of accreted oceanic terranes . The Dunnage Zone of the Northern Appalachians, and Newfoundland in particular, contains the remnants of several unrelated, thin, allochthonous, Iapetan oceanic terranes. Paleomagnetic and fossil data suggest they can be separated into peri-Laurentian (Notre Dame Subzone) and peri-Gondwanan (Exploits Subzone) terranes, which formed on opposite sides of Iapetus and are separated by the Red Indian Line. The 507 to 489 Ma Baie Verte oceanic tract and the 480 to 460 Ma Annieopsquotch accretionary tract (AAT) (van Staal etal., 1998) comprise oceanic terranes preserved in the Notre Dame Subzone. The peri-Gondwanan oceanic terranes of the Exploits Subzone have a similar age range (513– 460 Ma). The accreted oceanic terranes invariably consist of suprasubduction zone rocks (i.e. forearc, arc and back-arc). True MORB-type oceanic crust is not preserved anywhere. Accretion of the oceanic terranes to continental crust was mainly characterized by mélanges and localized deformation. Accretion of the Dashwoods microcontinent to Laurentia produced the Taconic Orogeny (ca. 475– 465 Ma), which caused significant shortening and metamorphism of the magmatic rocks of the Notre Dame arc built on the Dashwoods. Accretion of the Dashwoods microcontinent was followed rapidly along its eastern margin by 470 to 460 Ma accretion of the AAT, and the 455 to 450 Ma arrival of the Victoria-Exploits arc, which was built on the leading edge of Ganderia. This latter collision produced the Red Indian Line suture zone , characterized by duplex-style thrust complexes, oblique reverse ductile shear zones, local folding and mélanges, and continued deformation and metamorphism of the Dashwoods and Humber zones. After accretion of the Victoria-Exploits arc, deformation related to closure of the Exploits back-arc basin to the east juxtaposed both sides of the basin along the Dog Bay Line (Fig. 3, 4) during the ca. 430 Ma Salinic Orogeny. This collision produced both high-level structures (e.g. mélanges) and metamorphic tectonites, and coincides with termination of the last phase of arc magmatism (ca. 440– 435 Ma) in the Notre Dame arc to the west (Fig. 4). Arc magmatism was followed by ca. 430 to 425 Ma, bimodal, within-plate magmatism in the Notre Dame and Dashwoods subzones (Fig. 3).

While the western part of Ganderia was colliding with Laurentia, oceanic crust was subducting beneath it from the Avalon side. This produced Early to Late Silurian arc - back-arc magmatism along its southern and eastern edge, cryptic evidence of which is preserved in Newfoundland (e.g. La Poile Basin and Burgeo granite, Kerr et al., 1994), but much better in the Maritimes (e.g. Barr et al., 2002). Termination of the Silurian arc, inversion of the La Poile Basin (O'Brien et al., 1991), pervasive Early Devonian deformation and low- to high-grade metamorphism, and voluminous, largely granitoid magmatism throughout the Gander Zone and immediately adjacent Exploits Subzone, is probably due to the docking of Avalonia (the onset of the Acadian Orogeny). Acadian structures near the Gander-Avalon boundary (Fig. 3, 4) generally have a southeasterly vergence (e.g. Colman-Sadd, 1980). However, a large west-directed hot nappe (Fig. 4, Meelpaeg allochthon) exposed in southwestern and central Newfoundland (Fig. 3) may form partof a retro-wedge.

The predominance of low-grade metamorphism and preservation of large tracts of Siluro-Devonian marine and terrestrial sedimentary and volcanic rocks throughout the CMB, locally in nearly pristine condition, indicate that the core of the Canadian Appalachians, with the exception of some restricted areas with exposed high-grade metamorphic rocks near the external boundaries of the CMB, generally experienced moderate to low amounts of denudation. Furthermore, since evidence for extension is rare or absent in the CMB, this suggests that thickening of the Appalachian core generally was limited and the paleotopography low. The Megumazone, the furthest outboard terrane in the Northern Appalachians, is absent on the island of Newfoundland, but its accretion to composite Laurentia was probably responsible for Middle to Late Devonian orogenesis (Neoacadian) and voluminous granitoid magmatism in Nova Scotia.

Post-collisional processes mainly involved localized Middle to Late Devonian granitoid magmatism in the CMB and orogen-wide Late Devonian– Carboniferous strike-slip faulting associated with transtension and/or transpression, and basin formation.

Toby Rivers (on behalf of the Grenville working group: Sharon Carr, Louise Corriveau, Nicholas Culshaw, Tony Davidson, Michael Easton, Charles Gower, Andrew Hynes, Aphrodite Indares, Rebecca Jamieson, Jacques Martignole)

The Grenville Province comprises the exposed part of the late Mesoproterozoic Grenville orogen, which extends in the subsurface and in isolated outliers southwest as far as Texas (Fig. 5). The Grenvillian orogeny, which took place between ~1190 and 990 Ma, is interpreted to be the result of collision between Laurentia and another continent, possibly South America, to the southeast. The Himalayan-scale collision was preceded by about 300 Ma of northwest-dipping subduction that produced an Andean arc – back-arc system (1500– 1200 Ma) on the southeastern margin of Laurentia (e.g., Hanmer et al., 2000; Rivers and Corrigan, 2000).

Present crustal thicknes ranges between 42 and 47 km under much of the Grenville Province, on average a few kilometres thicker than under the adjacent Archean and Paleoproterozoic orogens. The orogenic crustal thickness may have been up to 70 km locally, based on metamorphic pressure estimates. Relief on the Moho appears low in most LITHOPROBE soundings, as imaged in the GLIMPCE, Abitibi-Grenville and ECSOOT transects, compatible with a hot lower crust and thermal relaxation following orogenic thickening. However, crust-penetrating normal faults offset the Moho by a few kilometres in the western Quebec transect (C, Fig. 6, 7), and in Labrador, the Grenville Front gravity low is caused by crust up to 5 km thicker than elsewhere. Seismic velocities in Grenville crust are generally slightly greater than those in adjacent orogens at similar depths, compatible with the higher grades of metamorphism. Reflectivity is weak to absent in the upper lithospheric mantle , the thickness of which is poorly constrained.

At the scale of the orogen, the Grenville's tectonic architecture is doubly vergent, although most of the exposed part verges to the northwest. A marginal parautochthonous belt in the northwest can be readily linked with the adjacent foreland, despite the effects of Grenvillian metamorphism and polyphase deformation (Fig. 6; Rivers et al., 1989, 2002). The Parautochthonous belt is structurally overlain by the High Pressure belt, characterized by eclogiteand/or granulite-facies rocks, which is in turn overlain by the Low Pressure belt in the orogenic hinterland. Throughout much of the province, the southeast-dipping crustal architecture has been exceptionally imaged in LITHOPROBE seismic profiles, which show gently to moderately dipping reflectors flattening into the mid to lower crust (Fig. 7; Ludden and Hynes, 2000a). Major reflectors can be correlated with both thrust- and normal-sense shear zones that mark contractional and extensional structures, respectively (e.g., Martignole et al., 2000). A wedge of Superior Province crust can be traced up to 350 km beneath the central Grenville Province (transects C and D, Fig. 7), where it appears to have behaved as a relatively competent ramp above which ductily deformed Proterozoic crust was imbricated.

Most of the Grenville Province is formed from reworked Laurentian crust of Archean to Mesoproterozoic age. Rift deposits have not been identified and the earliest Mesoproterozoic activity relevant to the ensuing Grenvillian orogenesis was the construction of a continental magmatic arc between ca. 1500 and 1200 Ma. The arc, which comprises a suite of calc-alkaline and A-type volcanic and plutonic rocks, along with coeval back-arc mafic dyke swarms, anorthosite complexes and clastic sedimentary deposits, extends along the Laurentian margin for over 5000 km of strike length.

Juvenile terranes that were accreted to the Laurentian margin immediately prior to the Grenvillian orogeny include the Composite Arc belt (ca. 1300-1250 Ma) and the metasedimentary Frontenac-Adirondack Lowland terrane of the southwestern Grenville Province, as well as the Coal Creek domain (1325-1275 Ma) of the Llano Uplift in Texas (Fig. 5; Mosher, 1988; Carr et al., 2000). Depositional environments of the Composite Arc belt and Coal Creek domain include arc and rifted arc oceanic settings; MORB-like compositions are rare. Where mapped, sutures separating the accreted terranes from the reworked Laurentian margin are ductile shear zones up to several kilometres wide. No suture with the putative southeastern continent has been identified, implying that with the exception of the accreted terranes, all parts of the orogen in North America have Laurentian affinities.

Grenvillian metamorphic grade in the Grenville Province is generally high, in the upper amphibolite and granulite facies. Metamorphic events also vary spatially and temporally, and have been dated at ~1190 to 1160 Ma, ~1080 to 1020 Ma and ~1010 to 990 Ma in different parts of the province (e.g., Rivers, 1997). Grenvillian P-T-t paths display wide variations depending on crustal level. The deep crustal rocks in the High Pressure belt evolved along very steep P-T-t paths, compatible with rapid exhumation by tectonic extrusion, whereas coeval high-T, medium-P rocks in the Low Pressure belt show evidence for more gradual exhumation (Indareset al., 2000; Rivers et al., 2002). The Parautochthonous belt followed intermediate P-T-t trajectories. Peak temperatures in excess of 800°C in the lower and middle crust, which led to anatexis and widespread emplacementof granitoid plutons during and immediately following the metamorphic peak, are unlikely to have been achieved by crustal thickening alone. It has been surmised that magmatism associated with the ca. 1.1 Ga mid-continental Keweenawan Rift system (Fig. 5) may have contributed ‘ excess' heat in the western Grenville Province (Culshaw et al., 1997). Elsewhere, evidence for widespread synorogenic mafic magmatism in the form of mafic dykes and gabbronorite-anorthosite intrusions has been interpreted as being due to extensional thinning of the lithospheric mantle following collisional thickening (Corrigan and Hanmer, 1997). Several isolated blocks lack evidence of penetrative Grenvillian metamorphism, implying development at shallow crustal levels prior to downfaulting during late-stage orogenic collapse.

In addition to late extensional structures, late- to post-collisional features include two widespread suitesof late-tectonic and post-tectonic granitoid plutons in the eastern Grenville Province (Gower and Krogh, 2002). Their geographic distribution suggests a unique lower crustal and/or mantle evolution in this region, which is not well understood. Throughout the Grenville, hornblende and muscovite ⁴⁰Ar/³⁹ Ar cooling ages of ~950 to 905 Ma indicate locally divergent uplift and cooling paths for adjacent structural domains, implying that brittle fault activity continued during uplift and exhumation.

In summary, the Grenville Province was part of a hot, wide orogen that developed on the site of a long-lived Mesoproterozoic continental arc on the southeastern Laurentian margin, in response to collision of a putative outboard continent. Some aspects of the Grenville orogen, such as its scale, the presence of eclogites, and the inference of tectonic extrusion of high-grade rocks, invite comparisons with the Himalayan and Variscan belts. Coeval high-temperature, medium-pressure metamorphic terranes in the hanging wall above the High Pressure belt may have developed under an uplifted plateau region. Numerical modelling of tectonic extrusion suggests that hot wide orogens are the sites of massive crustal-scale horizontal displacements and shortening along gently dipping shear zones, compatible with the LITHOPROBE images of the Grenville orogen.

G.M. Ross (on behalf of the participants of the PanLITHOPROBE IV meeting working group: Kevin Ansdell, Zoli Hajnal, Dick Wardle, Jeremy Hall, David Eaton, Ray Price, Carolyn Relf, Bill Davis, David Snyder, Ken Ashton, John Ketchum, Nick Culshaw)

Paleoproterozoic orogenic belts of Canada comprise nearly 6000 kilometres of strike-length orogenic material that record the formation of the core of the Laurentian craton between 2.0 and 1.8 Ga (Fig. 8; Hoffman, 1989). Older orogens (ca. 1.98– 1.90 Ga) are restricted to northwest Canada from central Alberta to the Arctic coast (Thelon-Taltson and Wopmay orogens). Widespread Paleoproterozoic collisional orogenesis occurred between ca. 1860 and 1800 Ma, including the circum-Superior Trans-Hudson orogen (THO) and collisional zones in central Alberta (East Alberta orogen)(Wardle and Hall, 2002; Ross, 2002) and western Northwest Territories (western Wopmay orogen). Paleoproterozoic orogenic belts include both accretionary and collisional orogens. The former involve outward-younging belts of material added to the growing orogen (e.g. Makkovik, Wopmay). In contrast, collisional orogens involved significant pre-existing crustal material (Archean and older Proterozoic in Paleoproterozoic orogens). Accretionary orogens may evolve to collisional orogens with time (Trans-Hudson orogen, for example) or this change may occur as an along-strike variation reflecting the nature (size, age and strength) of material arriving at the consuming plate margin.

The crustal thickness of Paleoproterozoic orogens ranges from about 35 km (East Alberta) to over 50 km (Torngat orogen). The crust- mantle boundary (Moho) invariably displays relief, which is inferred to date back to the time of orogeny. The origins of the Moho relief may be composite and can reflect simple crustal thickening, subcrustal accretion (Sask craton; THO), magmatic underplating, and late-stage fault juxtaposition of crustal elements of contrasting thickness. Virtually all LITHOPROBE lines across Paleoproterozoic orogens (SNORCLE, Alberta basement, THO and ECSOOT transects; Fig. 1) have revealed substantial seismic reflectivity throughout the entire crust. As a result of the excellent observed reflectivity, a number of major reflection styles and geometries can be discerned, and compare favorably with orogenic geometries seen in Phanerozoic orogens (Beaumont and Quinlan, 1994). Structure in the mantle lithosphere has been imaged on both reflection and refraction surveys within Paleoproterozoic orogens and includes the "frozen subduction" of the SNORCLE program (Cook et al., 1999), the Deep Probe/SAREX refraction profile (Gorman et al., 2002), and local short segments of dipping reflectivity in the shallow subcrustal mantle, which are interpreted as subduction scars (Eaton and Cassidy, 1996; White et al., 2002).

Crustal levels preserved in Paleoproterozoic orogens range from shallow levels of exhumation, typified by the Flin-Flon assemblage of THO (300– 400 MPa; 10– 12 km), to the deepest levels, which are implied by 1.0 to1.2 GPa rocks of the core zone of Torngat Orogen. An intriguing observation is that ca. 1.9 Ga orogens of northwest Canada have well-preserved foreland basins whereas ca. 1.8 Ga orogens lack coeval forelands. One of the key findings of the intraorogenic comparison approach is that evidence for synorogenic (or even late orogenic) extension, which normally assists exhumation, is lacking in Paleoproterozoic orogens. This includes seismic evidence as well as field relationships and kinematic analysis of shear zones. Further work defining P-T-t loops for Paleoproterozoic orogens will help constrain exhumation paths.

Paleoproterozoic orogens involved the assembly of pre-existing crustal fragments and microcontinents of both Archean and older Proterozoic age, and thus the shape and strength of the collided elements played a key role in orogenic style and evolution. The age and style of rifting sets the collisional template, defining the shape of the rifted margin (salient-reentrant geometries) and the thermal age (strength) of the margin, similar to younger orogens. The time interval between the age of rifting and the onset of convergence exerts a first-order control on the thermal state and elastic thickness of the lithosphere loaded during contraction and may explain the variable development and preservation of foreland sequences between Proterozoic orogens.

Paleoproterozoic orogens have accreted oceanic material comparable in diversity to Phanerozoic rocks although the ophiolite components are less prevalent. Circum-Superior orogens have a relatively young (ca. 1.88 Ga) magmatic-oceanic component related to rifting that seems to be concentrated along the continental margin of the Superior craton and may have been related to a plume impinging on the underside of the rifted Superior lithosphere. Continental magmatic arcs form the magmatic glue that welded constituent cratonic elements during consumption of oceanic lithosphere, ocean closure and eventual collision. A range of different erosional levels within these arcs is preserved, from deep structural levels in the Thelon-Taltson, to shallow levels in Great Bear Magmatic Zone. The duration of arc magmatismis comparable to that in Phanerozoic arcs, ranging from about 70 to 10 Ma, although the volume of some arcs, such as the Wathaman batholith, is remarkable given its short duration. One lesson learned from reflection profiles in Paleoproterozoic orogens (and elsewhere in LITHOPROBE surveys) is that the ascent of magmatic material and emplacement in the mid to upper crust has done little to modify highly reflective crust, through which the magmas must have traversed during emplacement. This observation suggests that magmas rose along transient conduits that may subsequently have closed. However, post-magmatic deformation and metamorphism during terminal phases of continental collision and assembly may also have overprinted the magmatic conduit systems.

With the exception of the Torngat orogen, which underlies the highest peaks in eastern North America, Paleoproterozoic orogens in Canada have little contemporary topographic relief. Reconstructing paleotopography is a difficult proposition, especially if the synorogenic foreland basin is not preserved. In Paleoproterozoic orogens of western Canada, sedimentary basins that are demonstrably post-orogenic, such as the cratonic sheet sandstones of the Athabasca, Thelon, and Hornby Bay basins (all approximately 1700 Ma), should provide a measure of remnant orogenic topography. On the basis of facies patterns and paleocurrents, none of these basins seem to show an influence by orogenic topography, suggesting that it was essentially removed by the time of deposition of these basins. On average that rate would correspond to 100 to 200 Ma for removal of orogenic topography. This contrasts with Phanerozoic orogens such as the Appalachians, which still have strong modern-day topographic expression more than 300 Ma after cessation of orogeny.

Paleoproterozoic orogens were variably affected by post-collisional magmatism , i.e. magmatism that is less than 50 Ma younger than the age of collision, as distinct from magmatism that is separated from orogenic activity by more than 200 Ma (e.g. transcontinental magmatism 1490– 1350 Ma that affected New Quebec and Makkovik orogens more than 300 Ma after orogenesis). In Alberta, post-collisional magmatism is inferred from seismic images, which are interpreted as showing the presence of the intrusive equivalents of large igneous provinces (LIPs) in the form of underplated material and midcrustal sills, the latter covering a region of at least 120,000 km² . Post-collisional granitic magmatism is common in some Paleoproterozoic orogens (e.g. Rae/ Hearne hinterland), whereas it appears to be absent in others (e.g. Torngat, New Quebec). This may be due in part to the state of the lower crust, considering that dehydrated, refractory, Archean lower crust, a common basement to Paleoproterozoic orogens, would be less fertile than accreted Proterozoic crust, which may not have undergone previous melting.

J.A. Percival and PanLITHOPROBE IV working group: K. Bethune, A. Calvert,T. Corkery, J. Craven, D. Davis,H. Helmstaedt, A. Leclair, S. Lin,J. Parks, M. Sanborn-Barrie, T. Skulski,G. Stott, P. Thurston, K. Tomlinson,G. West, D. White.

Although not traditionally viewed as an orogenic belt, the Superior Province has many features interpreted to record mountain-building processes on a 2000-km length scale, including arc magmatic rocks, orogenic flysch deposits, regionally extensive strain fabrics and medium to high metamorphic grades. Recent LITHOPROBE reflection profiles, from the Western Superior, Kapuskasing and Abitibi-Grenville transects (Fig. 1), interpreted in light of new geological mapping and a robust geochronological framework, have revealed a crustal architecture that invites comparisons to modern collisional orogenic belts (Williams et al., 1992; Calvert and Ludden, 1999; Ludden and Hynes, 2000b; Fig. 9).

Where measured, the crustal thickness of the Superior Province ranges between 38 and 45 km, apparently peneplaned from the maximum 44 to 70 km thickness at the orogenic climax inferred from metamorphic assemblages. Anomalies associated with Proterozoic structures include thicker crust (53 km) beneath the Kapuskasing uplift, a Paleoproterozoic intracratonic thrust (e.g., Percival, 1994), and thinner crust (35 km) beneath the Nipigon plate, a Mesoproterozoic aulacogen developed on Superior Province basement. Estimates of lithospheric thickness range up to 300 km, and shear wave analyses indicate east-west seismic anisotropy (Musacchio et al., in press). Shallow mantle sampled in refraction experiments also displays marked anisotropy (8.3– 8.8 km/s), suggestive of depleted harzburgitic compositions. Magnetotelluric experiments indicate strong east-west anisotropy at mantle depths below the ca. 2.7 Ga southwestern Superior Province, in contrast to a distinctly isotropic mantle domain beneath the ca. 3.0 Ga North Caribou terrane of the northwestern Superior Province (Craven et al., 2001).

Domains of old (>2.8 Ga) continental crust within the Superior Province (Fig. 10) are separated by belts of younger oceanic material. In the north, the Northern Superior superterrane (Skulski et al., 2000)("superterrane" implies a collage of terranes assembled prior to incorporation into the Superior Province) records an ancestry back to ca. 3.82 Ga. The North Caribou-La Grande-Goudalie superterrane (Stott, 1997), dominated by juvenile 3.0 Ga crust, extends through northern Manitoba, Ontario and Quebec and appears to reflect a craton-scale swing in structural grain (e.g. Percival et al., 1994), although tectonic discordance between older NW trends in the north and younger east-west trends in the south may also explain this pattern. Smaller domains of old continental crust to the south include the Winnipeg River (up to 3.4 Ga), Marmion (3.0 Ga) and Opatica (2.81 Ga) terranes. At the southern margin of the Superior, the Minnesota River Valley terrane(3.6 Ga)(Fig. 8) may be part of a bounding craton, although the continental fragments were variably reworked by Neoarchean (2.75– 2.65 Ga) magmatic and metamorphic activity, such that the extent of orogen-bounding continents is uncertain.

Scant evidence is available on the nature, style and timing of early rifting within the Superior Province. Thin rift sequences are locally associated with old cratonic fragments, along with distal platformal equivalents (Thurston and Chivers, 1990). These sequences generally comprise lower clastic (fuchsitic quartz arenite, grit, conglomerate) and chemical (carbonate, iron formation) sedimentary layers deposited in intertidal, shallow-marine, platformal and submarine-fan environments, overlain by volcanic (komatiite, basalt, rhyolite) sequences. Age constraints for two clastic sequences near the southern edge of the North Caribou superterrane suggest deposition prior to 2.93 Ga at the margin of an ocean that closed after2.72 Ga. Generally less than 500 m thick, these apparent rift sequences resemble those at volcanic passive margins such as in East Greenland.

Isotopically juvenile rocks (εNd^{2.7 Ga} =+5 to +2) occur in greenstone belts separating continental domains. Dominated by tholeiitic and komatiitic magmatism derived from depleted mantle sources, these juvenile belts are interpreted to have formed in oceanic settings, possibly analogous to oceanic plateaus. Relatively rare but widespread (Fig. 10) associated basalts characterized by Th and LREE depletion bear resemblance to midocean ridge basalt (N-MORB). Calc-alkaline volcanic rocks akin to islandarc edifices typically overlie lower mafic sequences in many juvenile greenstone belts. Rare adakite, Nb-enriched basalt and komatiite point to a considerable range of arc/plume processes, particularly well illustrated in the Abitibi subprovince. Subvolcanic intrusions of dominantly tonalitic composition make up large volumes of juvenile crust and were generated by recycling of oceanic crust through melting and fractionation processes. Oceanic rocks formed between 2.77 and 2.70 Ga are widespread throughout the southern Superior Province, and ca. 2.84 Ga examples are present in the Oxford-Stull Lake terrane, northwestern Superior Province.

Continental arcs are well developed in the northeastern Superior Province, where hydrous to charnockitic granodiorite and tonalite crystallized as massive vertical sheets between 2.78 and 2.70 Ga (Percival et al., 2001). Erosional levels (avg. 15 km) are slightly deeper than in the northwestern Superior where magmatic rocks of similar composition and age(2.74– 2.71 Ga) form horizontal sheets at ca. 12 km emplacement levels (Stone, 2000). Pre-arc continental basement ages are reflected in enriched isotopic signatures and inherited zircon populations.

Sutures are inferred at several interfaces between continental and oceanic domains (Fig. 10), although flysch deposits and younger structures invariably obscure their character. Subduction polarity is deduced from the presence of calc-alkaline arc magmatic belts that were active up to a few million years prior to collision, as well as from subduction scars: dipping upper mantle reflections presumed to image ancient subducted slabs (e.g. Calvert and Ludden, 1999). Suture zones are characterized by regionally extensive belts of greywacke flysch deposited early during the collisional process and subsequently deformed and metamorphosed during continued convergence. Their depositional ages constrain the time of docking events across the Superior Province (Davis, 1998; Fig. 11): 1) Northern Superior– North Caribou collision ca. 2.72 Ga, forming the initial northern portion of the composite Superior superterrane (CSS) during the Northern Superior orogeny; 2) accretion of the Winnipeg River– Wabigoon to CSS ca. 2.71– 2.70 Ga (Uchian orogeny; Stott, 1997);3) accretion of the Abitibi– Wawa to CSS ca. 2.69 Ga (Shebandowanian orogeny); and 4) terminal collision of the Minnesota River Valley to CSS ca. 2.68 Ga (Minnesotan orogeny).

Pressure-temperature-time (PTt) paths are best documented in flysch sequences deposited just prior to burial and metamorphism. For example, the English River flysch sequence, deposited in the North Caribou-Winnipeg River suture zone between 2.704 and2.698 Ga, reached peak P-T conditions (650 MPa; 800^° C) by 2.69 Ga. Sustained high temperatures and thick crust are indicated by continued metamorphism and emplacement of crust-derived granites until 2.65 Ga. Peak Shebandowanian conditions (600 MPa; 800^° C) were attained in the <2.698 Ga Quetico flysch sequence at ca. 2.66 Ga, and declined southward in the Abitibi belt (220 MPa; 260°C)

Post-collisional features include development of dextral transpressive zones, linked temporally and spatially to emplacement of mantle-derived magmas such as sanukitoid-suite rocks (high-Mg granodiorite, monzodiorite), rare carbonatite and associated alkaline rocks. These events occurred 1 to 5 Ma after the inferred time of accretion, with progressively younger ages toward the south (2.71– 2.68 Ga). Molasse deposits ("Timiskaming" conglomerates) show a similar age distribution. Post-tectonic granites and pegmatites crystallized between 2.68 and 2.64 Ga throughout the Superior Province; their emplacement may relate to fluid circulation responsible for growth of late monazite and titanite(2.69– 2.61 Ga). Deep crustal extension and high-grade metamorphism between 2.66 and 2.62 Ga are further indications of persistent high temperatures and ductile conditions following assembly of the province. Together, these phenomena point to a process of mantle magma generation, followed by deep-crustal melting and associated thermotectonic activity. The ultimate cause may relate to slab breakoff or more widespread lithospheric delamination.

W. Bleeker (on behalf of the Slave craton working group: B. Davis, R. Ernst, B. Griffin, A. Jones, J. Ketchum, C. Relf, D. Snyder, A. van der Velden)

The Slave craton, one of approximately 35 principal Archean crustal fragments scattered around the globe (Bleeker, 2003), is a small to moderate size craton (about 500 km EW by >700 km N-S) in the northwestern Canadian Shield (Fig. 12). Its key features include a high proportion of ancient crust, a craton-scale stratigraphy, high-amplitude mantled granitoid-gneiss domes, voluminous late granites, and widespread high-T, low-P metamorphism (Bleeker and Davis, 1999; Bleeker, 2002). It has been imaged in the eastern SNORCLE transect (Cook et al., 1999; Fig. 1).

The present topography of the Slave is dominated by a broad dome or arch rising to 550 m above sea level. The present crustal thickness averages 38 to 40 km, increasing to approximately 45 km near Paleoproterozoic bounding structures, and thinning to about 35 km beneath in the north-central part of the craton below parts of the Paleoproterozoic Kilihigok basin (KB, Fig. 12) (Bank et al., 2000). A crude estimate of average orogenic crustal thickness of 52 km at ca. 2.58 Ga is suggested by widespread andalusite-grade metamorphism (<360 MPa), indicating an average of ca. 10-12 km of syn- and post-orogenic exhumation. This exhumation was essentially complete by ca. 2.0 Ga, the age of onlapping cover sequences.

The nature of the subcontinental lithospheric mantle (SCLM) of the Slave is particularly well defined through geophysical and xenolith data. Slave lithosphere extends to a depth of about 180 to 220 km, with local maxima to 250 km (Bostock, 1998; Cook et al., 1999). Ancient crust of the Central Slave Basement Complex (CSBC, Fig. 12; Bleeker et al., 1999a,b) is underlain by seismically fast mantle lithosphere(e.g., Bank et al., 2000), xenoliths of which have Re-Os ages reflecting crustal ages (Aulbach et al., 2001; Irvine et al., 2001). Mantle xenoliths and xenocrysts, brought to the surface by diamondiferous kimberlite pipes, provide a complex picture of distinct and stratified mantle domains (Kopylova et al., 1998; Griffin et al., 1999; Grütteret al., 1999; Kopylova and Russel,2000; Carbno and Canil, 2002). A key feature of the central Slave SCLM, including that of the Lac de Gras kimberlite field, is an approximately 150 km thick, conductive (Jones et al., 2001), ultra-depleted harzburgite layer above more fertile lherzolite (Griffin et al., 1999). A relative abundance of diamondiferous eclogite xenoliths in some pipes may hint at a role for subduction and subcretion of oceanic crust at the base of the Slave lithosphere (Kopylova et al., 1998; Heaman et al., 2002; see also Bostock, 1998).

The crustal architecture of the Slave craton is dominated by the CSBC and its thin, widespread Central Slave Cover Group (2850-2800 Ma; Bleeker et al., 1999a,b; Sircombe et al., 2001; Fig. 13). This is overlain by voluminous tholeiitic basalt and minor komatiite (2740-2700 Ma), probably indicating rifting and/or plume activity, at a scale comparable to that of modern large igneous provinces. Neither geochemical nor field data have yet provided indications of this ancient crust and its cover stratigraphy in the eastern Slave, with the possible exception of the Hope Bay Block in the northeast (HB,Fig. 12), which remains insufficiently tested.

The east-west dichotomy has been explained in terms of collision of an exotic eastern arc terrane with an ancient crustal block (e.g., Kusky, 1989). However, because the <2690 Ma volcanic and sedimentary rocks of the eastern Slave are continuous or can be correlated with the upper part of the western Slave stratigraphy, any terrane collision would have to have occurred between ca. 2700 and 2690 Ma, a time interval considered too short for a full Wilson cycle. Progressive rifting, accompanied by voluminous volcanism and plutonism (back-arc setting?) may be an alternative to terrane collision to explain the lack of an ancient basement signature in the eastern Slave.

After 2690 Ma, the structural grain changed from N-S to NE-SW, on the basis of: 1) facies boundaries (i.e. distribution of banded iron formation (BIF) and gold deposits) in ca. 2680 to 2660 Ma Burwash Formation turbidites; 2) a chain of ca. 2661 Ma felsic volcanic centers ( continental volcanic arc? ); 3) the NE-SW trend ofca. 2642 to 2635 Ma F₁ folds in Burwash turbidites; 4) a magmatic arc of ca. 2635-2620 Ma Defeat Suite plutons; and 5) speculative boundaries between contrasting mantle lithosphere domains (Grütter et al., 1999; Davis et al., 2003). Collectively, these features suggest a supra-subduction zone setting (arc and back-arc?), possibly with a trench to the southeast of the present Slave. Significantly, all post-Burwash volcanic rocks occur in the northwestern Slave (Fig. 12), suggesting that after D₁ collision (folding of the Burwash Basin) and Defeat magmatism, a subduction flip may have occurred, moving the active margin to the northwest of the present Slave craton. These features were overprinted by F₂ folds and subsequent strike-slip faults along dominantly NNW-SSE to N-S trends between 2600 and 2590 Ma, leading to interference and partial masking of earlier trends (e.g., Davis and Bleeker, 1999).

Peak-metamorphic conditions generally were attained syn- to post-D₂ , near the time of voluminous granitoid magmatism, including a craton-wide "granite bloom" at about 2590 to 2580 Ma (Davis and Bleeker, 1999). This intracrustal depletion event extracted aqueous fluids, minimum melt fractions and most of the heat-producing elements from the lower crust. Large-scale advection of granitic melts into the upper crust is thought to have contributed to the crustal-scale amplitude (10– 15 km) of composite granite-gneiss domes, many of which are cored by young granites (Bleeker, 2002). This irreversible transfer of heat-producing elements to upper crustal levels allowed the lower crust to cool and stiffen, setting the stage for "cratonization" of the Slave crust. Although not easily ruled out, there is no direct evidence for mafic magmatic input from the mantle during the granite bloom. In rare sections of Slave crust exposing 700 to 900 MPa metamorphic levels, these younger events are manifested by a subhorizontal migmatitic layering and intercalated granite sheets, which may be the principal cause for strong lower crustal reflectivity. SNORCLE images of offsets and imbrication of this reflectivity and the Moho (Fig. 13) may be subduction scars (Cook et al., 1999; van der Velden and Cook, 2002b), or younger (e.g., Wopmay orogen) structures.

Following the 2590 to 2580 Ma thermal peak, the upper crust underwent rapid cooling (e.g., Bethune et al., 1999). However, zircon growth events in lower crustal xenoliths record a number of younger events at 2560 and 2520 Ma (Davis et al., 2003), indicating that metamorphic activity in the lower crust continued for up to 60 Ma after the metamorphic peak, similar to the pattern observed in the Kapuskasing uplift of the Superior craton (Moser and Heaman, 1997). Post-Archean thermal events recorded in accessory minerals of some xenoliths (zircon, titanite, rutile) indicate that transient heating of the lower crust, and possibly magmatic addition, occurred repeatedly, accompanying intrusion of Proterozoic dyke swarms at 2.2 and 1.27 Ga (Davis, 1997). Rutile ages from lower crustal xenoliths provide a minimum age for establishment of a cratonic geotherm at ca. 1.8 Ga.

Summary by J.A. Percival (from discussion organized by C. Beaumont, J. Hall, K. Karlstrom and R. Price, facilitated by notes from A. Calvert and K. Tomlinson)

An early discussion point considered whether Precambrian terranes, particularly those of Archean age, should be considered as orogens. The original definition of orogeny relates to the construction of mountain belts, although other characteristics (collisions, presence of sutures, metamorphic zonation, fold-thrust belts, etc.) have become engrained in our common understanding of the term. It was argued for the Superior Province that belts of 800– 900 MPa metamorphic rocks with Archean cooling ages and normal crustal thickness must have had major topographic expression during the orogenic climax, and therefore qualify as bona fide orogenic belts.

A theme that pervaded the discussion concerned secular variation of orogenic processes. At the extremes, are Archean orogens fundamentally different from Cenozoic ones? In order to answer this question, individual parameters need to be isolated, considered and compared. For example, low-P, high-T metamorphism is not a uniquely Archean feature; thermal and tectonic controls must be defined before appealing to generally hotter Archean conditions. Owing to higher concentrations of heat-producing elements, the Earth must have generated more heat during the Archean, but many believe that once stabilized, continental keels were not subjected to thermal erosion. Consequently, proportionally more heat must have been dissipated through the oceans.