Geolimnology of Freefight Lake: an unusual hypersaline lake in the northern Great Plains of western Canada*

Freefight Lake, Canada's deepest salt lake, is a meromictic, hypersaline lake located in the most arid part of the northern Great Plains. The lake has a distinctive basin morphology, with a large expanse of seasonally flooded mud flats and sand flats adjacent to a deep, flat bottomed perennial waterbody. The mixolimnion, dominated by magnesium, sodium and sulphate ions, has an average salinity of 110 ppt and overlies a monimolimnion of 180 ppt total dissolved salts. The entire water column is strongly supersaturated with respect to a variety of calcium and magnesium carbonate minerals; the lower water mass is also saturated or supersaturated with respect to a number of very soluble sodium, magnesium and sodium + magnesium salts. The modern sedimentary processes operating in Freefight Lake give rise to six main sedimentary facies: (i) colluvium, (ii) mud flats and sand flats, (iii) algal flats, (iv) delta, (v) slope and debris apron, and (vi) deep basin. The colluvium, mud flats and sand flats, and delta facies are dominated by physical processes and consist mainly of detrital siliciclastic sediment. The algal flats, slope and debris apron, and deep basin facies are dominated by endogenic and authigenic sediments derived mainly by physicochemical and biologically mediated carbonate and evaporite mineral precipitation. As one of very few deep water lakes in the world in which soluble evaporite minerals are forming and being preserved, Freefight Lake occupies an important position within the realm of lacustrine sedimentology. Although many of the sedimentological and geochemical processes taking place in the basin today are unusual, the delineation and evaluation of these processes is essential in order to decipher properly the stratigraphic records of Quaternary lake sediments in this large area of North America, as well as lacustrine sequences from other arid and semi-arid regions of the world.     

Climatic and hydrologic variability during the past millennium in the eastern Rocky Mountains and northern Great Plains of western Canada

Modelling of tree-ring δ¹³C and δ¹⁸O data from the Columbia Icefield area in the eastern Rocky Mountains of western Canada provides fuller understanding of climatic and hydrologic variability over the past 1000 yr in this region, based on reconstruction of changes in growth season atmospheric relative humidity (RH_grs), winter temperature (T_win) and the precipitation δ¹⁸O-T_win relation. The Little Ice Age (~ AD 1530s-1890s) is marked by low RH_grs and T_win and a δ¹⁸O-T_win relation offset from that of the present, reflecting enhanced meridional circulation and persistent influence of Arctic air masses. Independent proxy hydrologic evidence suggests that snowmelt sustained relatively abundant streamflow at this time in rivers draining the eastern Rockies. In contrast, the early millennium was marked by higher RH_grs and T_win and a δ¹⁸O-T_win relation like that of the 20th century, consistent with pervasive influence of Pacific air masses because of strong zonal circulation. Especially mild conditions prevailed during the “Medieval Climate Anomaly” ~ AD 1100-1250, corresponding with evidence for reduced discharge in rivers draining the eastern Rockies and extensive hydrological drought in neighbouring western USA.     

Sedimentology and evaporite genesis in a Holocene continental-sabkha playa basin—Bristol Dry Lake, California

Bristol Dry Lake, a 155 km² continental-sabkha playa basin in the Mojave Desert of south eastern California, is filled with at least 300 m of interbedded terrigenous clastics, gypsum, anhydrite, and halite. Evaporite facies conform approximately to a bull's eye pattern with gypsum and anhydrite surrounding a basin centre accumulation of halite. Transects through Bristol Dry Lake, from the alluvial fan to the centre of the playa, reveal: (1) crudely-bedded, alluvial fan clastics interfingering with (2) playa-margin sand flat and wadi sand and silt, followed by (3) gypsum, anhydrite, chaotic mud halite, and clay of the saline mud flat, and (4) salt-pan halite beds. Terrigenous clastics were deposited in Bristol Dry Lake by sheetflow and by suspension settling from ponded floodwater. Some sediment has been reworked by aeolian processes to form barchan dunes around the playa margin. Thin nodular-like beds of anhydrite and several types of gypsum occur across most of the playa. Giant hopper-shaped halite cubes are suspended in saline mud flat facies, suggesting that they grew displacively in brine soaked sediment just below the surface. Thick beds (4 m) of halite, in the playa centre, may have formed through a complex alternating history of subaqueous and intrasedimentary precipitation under the influence of periodic floods, intense evaporation and brine-level lowering, and capillary discharge of brines. The stratigraphy in the playa centre is cyclic. An ideal cycle consists of: (1) chaotic mud halite at the base overlain by (2) green to red clay with abundant, giant hoppers, and at the top (3) red clay, gypsum, and anhydrite with flaser- to wavy-bedded sand and silt. This type of cycle probably records a gradual progradation of mud-flat facies over salt pans. Bristol Dry Lake sediments are nearly identical to some of the Permian evaporites of the Permian Basin region, U.S.A. and they can serve as modern analogues for ancient-sabkha facies analysis.     

加拿大Athabasca地区下白垩统McMurray组沉积特征分析

在前人研究基础上,通过钻井岩心描述和测井曲线对比,根据岩石学特征、沉积相标志、测井相标志等对加拿大Athabasca地区下白垩统McMurray组进行了沉积特征分析。McMurray组主要发育了河流相、三角洲相和滨海沉积,总体呈海平面相对上升变化。纵向上,中下部层序以河流相为主,沉积物多为来自盆地周边的近物源,以西部物源为主。短河流汇入盆地后可沿盆地长轴南北发育,成为北部区块的长物源。上部层序受西北方向海侵的影响,自北向南发育滨海-河口湾-三角洲相沉积。     

Loess record of the Pleistocene–Holocene transition on the northern and central Great Plains, USA

Various lines of evidence support conflicting interpretations of the timing, abruptness, and nature of climate change in the Great Plains during the Pleistocene–Holocene transition. Loess deposits and paleosols on both the central and northern Great Plains provide a valuable record that can help address these issues. A synthesis of new and previously reported optical and radiocarbon ages indicates that the Brady Soil, which marks the boundary between late Pleistocene Peoria Loess and Holocene Bignell Loess, began forming after a reduction in the rate of Peoria Loess accumulation that most likely occurred between 13.5 and 15 cal ka. Brady Soil formation spanned all or part of the Bølling-Allerød episode (approximately 14.7–12.9 cal ka) and all of the Younger Dryas episode (12.9–11.5 cal ka) and extended at least 1000 years beyond the end of the Younger Dryas. The Brady Soil was buried by Bignell Loess sedimentation beginning around 10.5–9 cal ka, and continuing episodically through the Holocene. Evidence for a brief increase in loess influx during the Younger Dryas is noteworthy but very limited. Most late Quaternary loess accumulation in the central Great Plains was nonglacigenic and was under relatively direct climatic control. Thus, Brady Soil formation records climatic conditions that minimized eolian activity and allowed effective pedogenesis, probably through relatively high effective moisture.Optical dating of loess in North Dakota supports correlation of the Leonard Paleosol on the northern Great Plains with the Brady Soil. Thick loess in North Dakota was primarily derived from the Missouri River floodplain; thus, its stratigraphy may in part reflect glacial influence on the Missouri River. Nonetheless, the persistence of minimal loess accumulation and soil formation until 10 cal ka at our North Dakota study site is best explained by a prolonged interval of high effective moisture correlative with the conditions that favored Brady Soil formation. Burial of both the Brady Soil and the Leonard Paleosol by renewed loess influx probably represents eolian system response that occurred when gradual change toward a drier climate eventually crossed the threshold for eolian activity. Overall, the loess–paleosol sequences of the central and northern Great Plains record a broad peak of high effective moisture across the late Pleistocene to Holocene boundary, rather than well-defined climatic episodes corresponding to the Bølling-Allerød and Younger Dryas episodes in the North Atlantic region.     

Sedimentology of cherts in the Early Proterozoic Wishart Formation, Quebec-Newfoundland, Canada

The siliciclastic Wishart Formation of the Early Proterozoic Labrador trough is a high-energy shelf deposit. Wishart sandstones contain both interstitial chert with textures of void-filling cement and thin chert intercalations contaminated with siliciclastic mud. Although volumetrically minor, these cherts occur in several thin, areally extensive stratigraphic intervals. The Wishart contains intraclasts of both the chert-cemented sandstone and the impure chert layers (as well as several other types of chert sand and gravel). This suggests the cherts formed penecontemporaneously, which is consistent with the absence of any signs of replacement in all but one of the chert types and the clear-cut distinctions between chert types, even where they are side by side in a single thin section. The origin which appears to be most compatible with available evidence is that the cherts represent silica precipitated from thermal waters that rose through the sediments of the Wishart shelf and discharged into suprajacent seawater. A biogenic origin is unlikely in view of the lack of appropriate organisms during the Early Proterozoic and the rapidity with which the cements formed. A volcanogenic origin is unlikely because volcaniclastic textures are plentiful in associated formations but absent from the Wishart. Precipitation induced by evaporative concentration is unlikely in view of the widespread evidence of tidal currents and the lack of evidence of desiccation in the Wishart. Finally, the cherts are not restricted to the lowest-energy facies, and therefore they presumably did not accumulate as a background sediment. Deposition of silica above the sediment/water interface was probably made possible by ambient concentrations of silica that were significantly higher than those of Phanerozoic seawater. Cherts with similar textures occur in other Early Proterozoic sediments, most notably arenitic or granular iron-formations.     

Sedimentology of the Ekwan Shoal,Akimiski Strait,James Bay,Canada

One of the steepest depositional coasts of western James Bay is found along the west shores of Akimiski Strait, north of the mouth of the Ekwan River. This shore receives considerable amounts of sediment during the spring break-up of the rivers. The sediments are stored on the steep narrow tidal flats and marshes, and in thinner (up to 80 cm) drapes on till-cored shoals that parallel and protect the coast. The low areas between the shoals and the mainland are swept and reworked by relatively powerful (2 m s^?1) reversing currents due to flooding and ebbing of tides into the strait.A series of distinct environments and sedimentary facies develop on this western coast and its antecedent longshore shoal. The outer part of the shoal is characterized by tidal bedding, Macoma balthica burrows and considerable ice scour. The inner part of the shoal has winnowed sand, the greatest abundance of Macoma, and well-developed flaser bedding. The longshore tidal channel separating the shoal from the mainland has coarse sand lags in the shallower parts and silty sand in deeper protected areas. The steep tidal flats develop laminated silty sands locally saturated and slumping toward the channel. The high saturation of the sediments inhibits colonization of the flats by Macoma. The narrow marshes have characteristic vegetation zonation, with Puccinellia phryganodes colonizing the lower marsh. The sedimentary sequence of the marsh displays irregular, bioturbated laminated sequences of silt, silty sand and organic matter.     

A high-resolution record of climate variability and landscape response from Kettle Lake,northern Great Plains,North America

A decadal-scale multiproxy record of minerals, pollen, and charcoal from Kettle Lake, North Dakota provides a high-resolution record of climate and vegetation change spanning the entire Holocene from the northern Great Plains (NGP) in North America. The chronology is established by over 50 AMS radiocarbon dates. This record exhibits millennial-scale trends evident in other lower-resolution studies, but with much more detail on short-term climate variability and on the rapidity and timing of major climatic shifts. As a proxy for precipitation, we utilize the rate of endogenic carbonate sedimentation, which depends on groundwater inflow, which in turn depends on precipitation. Independent cluster analyses of mineral and pollen data reveal major Holocene mode shifts at 10.73 ka (ka = cal yr BP), 9.25 ka, and 4.44 ka.The early Holocene, 11.7–9.25 ka, was generally wet, with perhaps a trend to higher evaporation associated with warming temperatures. A switch from calcite to aragonite deposition associated with a severe, but brief drought occurred at 10.73 ka. From 10.73 ka to 9.25 ka, climate was generally humid but punctuated at 100–300 yr intervals by brief droughts, including the most severe drought of the entire Holocene at 9.25 ka. This event was coeval with the 9.3–9.2 ka event in the Greenland ice cores and observed at a number of sites worldwide. In contrast, the prominent 8.2 ka event in Greenland is not remarkable at Kettle Lake. The prominence of the 9.25 event locally in the NGP may be due to a major drawdown and northward retreat of Lake Agassiz at this time, reducing its mesoclimatic effect on the NGP and thrusting the region into an insolation controlled regime.The mid-Holocene, 9.25–4.44 ka, was characterized by great variability in moisture on a multi-decadal scale, with severe droughts alternating with more humid periods. The high abundance of the weedy but drought intolerant Ambrosia generally during the mid-Holocene and specifically during the multi-decadal drought periods is seemingly paradoxical, but can be explained by high interannual variability of moisture overlaid on the multi-decadal variability.The late Holocene, 4.44 ka–present, was also characterized by multi-decadal variability in moisture, but was generally wetter than the mid-Holocene and the magnitude of variability was less. The trends in wet-dry mineral, pollen, and charcoal proxies were similar to the mid-Holocene, but late Holocene mineral-pollen assemblages are distinct from mid-Holocene. The shift to wetter climate in the late Holocene was more gradual than the abrupt shift to arid conditions 9.25 ka, which may explain the asymmetric retreat and readvance of forest along the eastern margin of the NGP.Precipitation variations in the NGP have been linked with Pacific and Atlantic sea-surface temperatures, and mid-Holocene drought in the NGP has been linked with sustained La Niña-like conditions in the Pacific. These linkages may explain the decadal- and millennial-scale variations seen in the NGP, but cause of the prominent century-scale variations remains elusive.     

Sedimentology and history of a Late Wisconsinan glacial lake, Grande Prairie, Alberta, Canada

The Grande Prairie region of northwestern Alberta was partially covered by glacial Lake Peace, which was dammed against the retreating Laurentide ice sheet. Two levels of glacial Lake Peace are identified in the study are by closely spaced groups of strandlines and minor deltas lying at 805 to 840 m a. s. l., and 655 to 710 m a.s.l. Sedimentation associated with the upper of the two lake levels is marked by rhythmites of silt and clay deposited by turbid underflow, interbedded with diamicton deposited by debris flow. Dropstones and dump structures indicate common ice-rafting. Thick sequences are only found on the axes of major valleys, where sediment gravity flows were concertrated. Thin sequences of ice proximal glaciolacustrine sediments reflect topographic setting and do not indicate a short-lived lake. Retreat of the ice front resulted in a decrease in ice-rafted material and diamicton in sediments. The fall in lake level to the second stage resulted in deposition of sequences of vaguely laminated silt and clay close to the modern Beaverlodge River. These sediments were deposited by suspension settling from interflow or overflow of the Beaverlodge River as it entered the lake. Lake sedimentation was dominated by inflow from unglaciated areas, rather than the ice front.     

Wind erosion of saline playa sediments and its ecological effects in Ebinur Lake,Xinjiang, China

In many arid and semiarid areas, dry lake beds (saline playa) represent a tremendous source of unconsolidated salt-rich sediments that are available for aeolian transport. Severe salt-dust storms caused by the erosion of such landforms have become very harmful natural phenomena. In this study, sample analysis and field erosion monitoring of Ebinur Lake was conducted to investigate the salt content, chemical composition, and wind erosion intensity of surface salt-rich sediments. The effects of salt-dust rising from the playa on the growth and physiological health of plants were also evaluated in this study through a leaf dustfall test. The results indicate that water-soluble salts assemble densely on the dry lake bed surface. At a depth of 0–2 cm, the highest salt contents can exceed 40%, with sulfate and chloride being the main anions present and Na⁺, Ca²⁺, and Mg²⁺ being the primary cations. The annual wind erosion rate ranged from 0.48 to 5.6 cm in the northwest portion of the lake and from 0.24 to 0.96 cm in the southeast portion. Salt-dust storms caused by wind erosion of saline playa sediments seriously influenced the normal absorption of minerals by plant leaves. Under the influence of salt-dust storms, plant leaves absorb more Na⁺, but far less K⁺.     

Mid-Wisconsinan environments on the eastern Great Plains

Few sites on the eastern Great Plains contain paleobotanical records for the mid-Wisconsin. We report on four sites, two stream cutbanks and two quarry exposures, ranging in age from >50 to ～23.4 ka. The oldest site at >50 ka contains a suite of macrofossils from prairie and disturbed ground habitats, with no representation of trees, indicating an open prairie. By ～38 ka the assemblages include aquatic, wetland, mudflat, and prairie elements with rare specimens of Populus, Betula cf. papyrifera, Salix and at the most northerly site, Picea. This assemblage suggests a prairie/parkland with interspersed marshes, cooler temperatures and increased moisture. Populus and Salix continued to be represented from ～36 to ～29 ka, but the only other taxon was Carex. A hiatus may be present at some time during this interval. After ～29 ka, Picea became dominant on the uplands and it was joined by sedges in local wetlands. At sites near riverine loess sources, loess accumulation began to fill in the wetlands and organic deposition ceased some time after 29 ka.     

Sedimentology and stratigraphy of a transgressive, muddy gravel beach: Waterside Beach, Bay of Fundy, Canada

Sediments exposed at low tide on the transgressive, hypertidal (>6 m tidal range) Waterside Beach, New Brunswick, Canada permit the scrutiny of sedimentary structures and textures that develop at water depths equivalent to the upper and lower shoreface. Waterside Beach sediments are grouped into eleven sedimentologically distinct deposits that represent three depositional environments: (1) sandy foreshore and shoreface; (2) tidal‐creek braid‐plain and delta; and, (3) wave‐formed gravel and sand bars, and associated deposits. The sandy foreshore and shoreface depositional environment encompasses the backshore; moderately dipping beachface; and a shallowly seaward‐dipping terrace of sandy middle and lower intertidal, and muddy sub‐tidal sediments. Intertidal sediments reworked and deposited by tidal creeks comprise the tidal‐creek braid plain and delta. Wave‐formed sand and gravel bars and associated deposits include: sediment sourced from low‐amplitude, unstable sand bars; gravel deposited from large (up to 5·5 m high, 800 m long), landward‐migrating gravel bars; and zones of mud deposition developed on the landward side of the gravel bars. The relationship between the gravel bars and mud deposits, and between mud‐laden sea water and beach gravels provides mechanisms for the deposition of mud beds, and muddy clast‐ and matrix‐supported conglomerates in ancient conglomeratic successions. Idealized sections are presented as analogues for ancient conglomerates deposited in transgressive systems. Where tidal creeks do not influence sedimentation on the beach, the preserved sequence consists of a gravel lag overlain by increasingly finer‐grained shoreface sediments. Conversely, where tidal creeks debouch onto the beach, erosion of the underlying salt marsh results in deposition of a thicker, more complex beach succession. The thickness of this package is controlled by tidal range, sedimentation rate, and rate of transgression. The tidal‐creek influenced succession comprises repeated sequences of: a thin mud bed overlain by muddy conglomerate, sandy conglomerate, a coarse lag, and capped by trough cross‐bedded sand and gravel.     

Coalbed methane resources and reservoir characteristics from the Alberta Plains,Canada

Alberta has vast coal resources that may be a potential source of coalbed methane (CBM). Exploration and research are currently underway in the province to quantify gas potential, identify key geological factors that maximize CBM potential, and identify the ‘most favourable’ areas for CBM production potential.There are four main coal zones within the Plains and Foothills of Alberta. The Ardley Coal Zone of the Plains is undergoing CBM exploration and production piloting. Much of this effort is within the west-central Pembina area. Horseshoe Canyon Formation coals of south-central Alberta were initially thought to have gas concentrations too low to be economic CBM producers. It is these coals, however, that host Alberta's first commercial CBM production project. Although similar in both geographic distribution and coal quality to Horseshoe Canyon coals, not much is known regarding the gas potential of underlying coals of the Belly River Group. The deeper Mannville coals have some of the highest gas concentrations of Alberta coals; however, they are also relatively deep and have lower permeability than the overlying Belly River, Horseshoe Canyon and Ardley coals.Maximum gas-in-place for the Plains and Foothills has been estimated to be greater than 500 trillion cubic feet (1.42 * 10¹³ m³). Although this number is very large, little is known as to how much of this huge resource is actually producible. A key challenge to producibility in Alberta has been the generally low permeability of coals with the highest gas concentrations (Mannville coals), and the moderate to low gas concentrations of higher permeability coals (Horseshoe Canyon and Ardley coals).Regionally, coal distribution and average gas-in-place concentrations are calculated for the Alberta Plains. Identifying and explaining local areas with favourable CBM production characteristics within the regional setting is necessary to establish economic CBM plays within Alberta. There are currently several pilots and numerous exploration efforts underway in the province to demonstrate production potential.This study integrates existing data with new data collected from key areas that show favourable CBM potential. There is great opportunity for CBM development in Alberta. Local areas have enhanced characteristics favourable to production. Ongoing geological investigations are needed to identify characteristics that will serve as an exploration tool for future CBM discoveries.     

Sedimentology and paleohydrology of Late Quaternary lake deposits in the northern Namib Sand Sea, Namibia

The Namib Sand Sea is the largest active desert dunefield in southern Africa, and is comprised mainly of large north-south linear dunes. In the interdune areas of the northern Sand Sea eleven small areas of calcareous lacustrine sediment have been studied. These beds are typically less than a metre thick and are dominantly comprised of calcareous sandstones to mudstones and sandy limestones. The carbonates are mainly magnesian calcites (1–14% MgCO₃) with some protodolomite and aragonite. Calcified reed casts and fresh to brackish water gastropods, diatoms, and ostracods are present in some beds. δ¹⁸O values indicate a hot and dry climate. A number of enriched δ¹³C values may reflect high salinity, low organic populations, or carbonate recrystallization.These carbonate-rich lacustrine deposits are indicative of increased periods of moisture availability in this normally hyperarid region during the Late Quaternary. The origin of the water responsible for depositing these sediments may be: (1) ponding at the end point of the Tsondab River, which at one time extended farther west into the Sand Sea; (2) flooding into interdune corridors when water levels rose in rivers such as the Kuiseb; (3) groundwater seepage into depressions either through dunes that border rivers or from the underlying Tsondab Sandstone; and (4) increased rainfall. We do not believe that there is evidence to support a major increase in precipitation over the region. However, even a small increase in precipitation in the headwaters of valleys that drain toward the Sand Sea might: (1) generate enough additional runoff to extend the terminal point of rivers such as the Tsondab farther into the dunes; (2) cause lateral flooding from major valleys into interdune corridors; and (3) recharge aquifers. The sedimentary records at Narabeb, Ancient Tracks, and West Pan, which lie along the old course of the Tsondab River, favor a ponded river origin for them, whereas groundwater seepage is favored at other sites. The chronology of deposition, based on radiocarbon dates, suggests that ponding and recharge occurred earlier in the lower, western part of the area, and later in the east. This is in harmony with the view that the end point of the Tsondab River progressively retreated eastward between about 30 and 14 ka BP, as dunes blocked its route.     

Sedimentology and tectonic implications of Cretaceous fan-delta conglomerates, Queen Charlotte Islands, Canada

The Honna Formation, of Coniacian age, consists of several hundred metres of polymictic clast-supported conglomerate associated with sandstone and mudstone. Five conglomerate facies are recognized: ungraded beds; inverse graded beds; normal graded beds; inverse-to-normal graded beds; and parallel-stratified beds. These facies are interpreted as the deposits of subaqueous cohesionless debris flows and/or high-density turbidity currents. The depositional environment was a deep-water, gravelly fan that draped a fault-controlled, basin-margin slope. The fan is inferred to have passed upslope directly into an alluvial fan (unpreserved); hence, the name fan delta can be applied to the overall depositional system. This type of fan delta, of which the Brae oilfield in the North Sea is an example, is defined here as a deep-water fan delta. The lack of a shelf is in marked contrast to other types of fan delta. Three facies associations are recognized in the Honna Formation: subaqueous proximal-fan conglomerates, distal-fan turbiditic sandstones, and pro-fan/interfan mudstones with thin sandy turbidites. The proximal fan is envisaged as an unchannelled gravel belt with a downslope length of at least 20 km; such a long subaqueous gravel belt lacks a known modern analogue. The distal fan was an unchannelled sandy extension of the proximal gravel belt. It is postulated that the Honna Formation accumulated in a foreland basin which migrated westwards from the Coast Mountains where the Wrangellia-Alexander terrane was colliding with North America. In this model, the Honna fan delta was sourced by a (west-verging) thrust sheet whose sole-thrust was the Sandspit Fault immediately to the east. Deep-water fan deltas appear to develop preferentially when eustatic sea-level is relatively high, so that the‘feeder’ alluvial fan is small, and gravelly throughout. In petroleum exploration and field development, care should be taken to distinguish deep-water fan deltas from base-of-slope (canyon-fed) submarine fans, because the two systems differ significantly in terms of coarse-sediment distribution.     

Sedimentology of amphitrite bank,a commercially exploited groundfish habitat—Southwestern continental shelf,British Columbia,Canada

Among trawling grounds for Pacific cod (Gadus macrocephalus) off western Canada, Amphitrite Bank is small (∼220 km²) but important both because it is a major spawning area and because it sustains a relatively high yield At the south and east central part of the 40- to 60-m deep bank, sediments consist mainly of densely packed, poorly sorted, very coarse, immobile gravels Boulders which are commonly 3–5 m wide and 1–2 m high offer the most severe hazards on the bank to trawling gear. On the west central part of the bank, areas containing generally finer, but still poorly sorted gravels alternate with areas having coarse rippling (wavelength=4–5m, height=0.4–0.5m as measured from sonograms) and size-segregated gravels. Along the outer bank, areas having coarse and fine rippling alternate with relatively smooth patches where large boulders are generally absent. Local wave regime is adequate to mobilize the coarsest (mean size=−4ϕ) relatively well-sorted (σ=≤1.9ϕ) sediment and generate the largest observed oscillatory ripples in materials of this size. Incubating eggs were not observed directly on the bank surface, but Canadian trawl captains have reported catching actively spawning cod exclusively over areas of the bank which we have identified as heavily rippled. This substrate may be suitable for Pacific cod, whose eggs are demersal and temporarily adhesive, possibly because it is moved sufficiently to maintain gravel particle surfaces free of deposits which may inhibit adhesion, but not frequently enough during the spawning and incubation period to damage eggs     

Petrology of the Great Abitibi Dyke, Superior Province, Canada

The Great Abitibi Dyke (GAD) which can be traced northeast,for >> 700 km, across the Abitibi Belt in the southeasternSuperior Province of the Canadian Shield, is composed of olivinegabbro to monzodiorite, weakly saturated to undersaturated insilica. All rocks of the GAD can be derived by mainly plagioclase andolivine fractionation from a parental magma corresponding incomposition to chilled margin samples. Two units can be distinguished,a marginal unit (Unit 1) representing 0–50% crystallizationand a central unit (Unit 2), found over about half of the dykelength, representing 50–70% crystallization. Modelling,using Pearce elemental ratio analysis, quantifies the fractionationhistory and allows mass–balance calculations over thepresent exposure level of the dyke. The approximate balancebetween the amounts of cumulate and fractionated rocks suggeststhat Unit 1 differentiated essentially in situ as a closed system.In contrast, Unit 2 rocks were formed by loss of substantialplagioclase and olivine from the parental magma. This fractionatemust have either been lost to depth or left behind in an externalchamber. Unit 2 rocks show depletion in plagioclase and enrichmentin mafic minerals along strike towards the southwest (deeperexposure level?), a trend explained by density stratification.Regional variation in Unit 1 chemistry is interpreted in termsof lateral magma injection towards the northeast from the locusof Keweenawan rift volcanism. Feldspar, olivine, and augitecompositions are linearly correlated with equilibrium temperatureand extent of magma evolution. The parent magma had a trace element chemistry correspondingto a ‘within–plate’ setting and was probablyderived from an incompatible–element enriched mantle similarto the source for ocean island basalts (OIBs).