Paradoxical mid-crustal displacements and stratigraphic continuity: structural evolution of the northeastern Paleoproterozoic Amer belt, Nunavut, Canada

Mapping and ancillary studies by several authors (Heywood 1977; Blackwell 1978; Young 1979; Knox 1980; Patterson 1986; Rainbird et al. 2010; Pehrsson et al. 2013; Jefferson et al. 2015) have established the geology of the Paleoproterozoic Amer belt, Nunavut (Fig. 1) at a range of scale and breadth of disciplinary interest. The Amer group is one of several Paleoproterozoic sequences deposited on the Archean Rae craton, each of which is assigned to a distinct belt or basin (e.g., Amer, Ketyet River and Piling groups). The currently applied tectono-stratigraphic framework of the Amer group (Fig. 2) comprises four Paleoproterozoic sequences—Ps1, Ps2, Ps3, Ps4—that were largely established by Young (1979); sequence boundaries were defined by some combination of unconformities, major lithological change, or fine- to coarse-grained siliciclastic transitions. The stratigraphy and sedimentology of these units is described in detail by Jefferson et al. (2023).

Detailed studies over the past decade (MacIsaac 2011; McEwan 2012; Tschirhart et al. 2012; Calhoun et al. 2014; Calhoun 2016) have generated new particulars for the region that sometimes conflict with existing views. The latter studies have concentrated on acquisition of detailed structural data in concert with the specific stratigraphic relationships. Notwithstanding the overall contiguity and ordering (Rainbird et al. 2010) of the Ps1–Ps4 sequences throughout the Amer belt (Fig. 2), the intensity of deformation throughout can be shown to be greatly understated. This, in turn, poses an apparent paradox wherein a regionally extensive stratigraphic sequence is retained in its apparent primary order despite the observation of intense deformation.

The Amer belt (Fig. 3) trends east northeast–west southwest as a series of synclinoria of variable areal width having overall shallow to moderate west southwest plunges. The portion of the belt examined herein is morphologically divided into two domains (Fig. 3)—the Main Synform and the Oval Synform that are separated by a regional linear feature, the Medial Zone. The exposed extent of the belt is defined by a basal quartz arenite (quartzite) that overlies Neoarchean gneissic basement (Fig. 4a). The four Proterozoic lithostratigraphic sequences (Fig. 2; Jefferson et al. 2023) that form the basis for geometric analysis of the belt are

• Ps1: basal micaceous schist ± conglomerate, quartz arenite (Ayagaq Lake formation)

• Ps2: black graphitic phyllite feldspathic sandstone ((Resort Lake formation), siliceous dolostone/dolomitic marble (Aluminum River formation)

• Ps3: tholeitic basalt (Five Mile Lake formation), phyllitic siltstone (Three Lakes formation), feldspathic sandstone (Oora Lake formation), phyllitic siltstone/fine-grained arkose (Showing Lake formation)

• Ps4: feldspathic sandstone (Itza Lake formation)

Exposition of the structural geology comprises three sections. Detailed macro- and mesoscopic structural/deformation features of the northeastern Amer belt are first described with reference to the lithostratigraphic units (Fig. 2) that act as the stratigraphic paradigm for the belt (Young 1979, Jefferson et al. 2023); the latter is advisably read as complementary to this contribution. This first section comprises field descriptions of observed structures that establish the data on which eventual interpretations are based; these features also document the extent and intensity of deformation absent from some prior studies. Next, a systematic analysis of orientation data and structures (stereonet analysis) establishes concrete relationships among the different structural elements. Lastly, the structures within the Amer belt, their sequential development, and the issue of how such a regionally extensive stratigraphic ordering can retain its fundamental character in the faces of significant tectonic disruption are addressed. A resolution of regional tectonics is not the primary objective, but any such interpretations must be consistent with the detailed deformation attributes and history described herein.

An important component of the data acquisition has been extensive review, evaluation, and incorporation of primary source legacy data (see Calhoun 2016), particularly from archived material held by the Geological Survey of Canada. The sources examined included field notes, field samples, and original thin sections from the latter. Direct comparison of these legacy data with remapped outcrops enabled correlations among different workers to be established, particular as related to map-scale interpretations of the geology. The legacy data dramatically increase the effective area of detailed examination without a complete remapping. The incorporation of both detailed gravity (Tschirhart et al. 2012) and aeromagmetic data (Jefferson et al. 2015) contributed significantly to deciphering deformation geometries, especially in areas of poor or limited outcrop. Correlations of geophysical signatures with mesoscopic geometries, deformation styles, and finite fabrics were used to extrapolate into the subsurface to a greater extent than otherwise possible.

The detailed and, to some degree, repetitive description of the mesoscopic structures from individual sedimentary sequences serves as the primary database that records the intense deformation of the northeastern Amer belt, the nature of which is significantly underreported. Structures are purposely described by order of the individual lithostratigraphic sequences (Fig. 2) so as to (1) avoid any unconscious bias toward describing a common behaviour throughout the belt, (2) establish, in detail, structural elements that have been overlooked during regional mapping, and (3) highlight the contrasting rheological responses among specific lithologies that are potentially masked by improper combination of observational (deformation style) and measured (orientation) data across the belt.

To aid reading of the structural descriptions, a brief summary of the structural development is given here; this will allow for easier adaptation to notation used throughout the text and provide a scheme for considering that data. It is emphasized that the summary reflects synthesis of the evidentiary observations and did not pre-empt their objective acquisition. Upon completion of mapping and orientation analysis, structural elements have been separated into two major deformation phases, D₁ and D₂. D₁ comprises internal deformation of the Proterozoic units Ps1 through Ps3, while D2 produces the current regional synclinoria after unconformable deposition of Ps4. Within D₁, three deformation generations are recognized through overprinting relationships, i.e., S_1a, S_1b, S_1c, using standard subscript notation, in this case for foliations. When generations within D₁ cannot be discriminated, a single subscript, such as S₁, is assigned; the use of single subscript for D₁ features implies that overprint relationships or other information are not available to separate generations. The style of otherwise co-generational cleavages/foliations varies with lithology between individual sequences. As well as their style, structural elements vary in orientation as a function of accumulated strain plus their location on overprinting structures. All specific orientations are given in vector format as inclination → azimuth, that is, dip → dip azimuth for planar features and plunge → plunge azimuth for linear features.

The <2.3 Ga Ayagaq Lake formation quartz arenite (vitreous quartzite in parts) defines the Proterozoic base of the Amer belt overlying Archean basement gneiss (Fig. 4a); the quartzite is typically exposed as resistant ridges along the northern and southern belt boundaries (Fig. 3). The Amer “basin” is outlined by exposures of this unit and the high preservation index of this lithology is such that, if it exists, it is expected to outcrop. The gneiss–quartzite contact is markedly dislocated either along the contact or at different levels within the gneiss, by thrust faults forming imbricate structures. Imbrication at the structural base of the belt is most obvious between Ayagaq Lake quartzite and Archean basement along the northern and northeastern edges of the belt (Fig. 3). Thrusts currently dip moderately west to south–southwest but have been modified by later folding (to be discussed later).

Thin conglomerates (Fig. 2) at the base of the quartzite (Jefferson et al. 2023)) establish the initial unconformity between Ps1 and Archean gneiss. A thin schistose unit is sometimes preserved between the arenite and the basement gneiss (Fig. 4b) and presents varyingly as an illitic/sericitic phyllite of distinctive composition and texture (Figs. 4b and 4c) and micaeous, quartzo-feldspathic schist. The schistose unit is consistent with the conglomerate/paleosol protolith expected initially at the major unconformity. Disturbance of the primary texture is seen as deformation microstructures in thin section (Fig. 4b) within the basal phyllite/schist and include both dynamic recrystallization (quartz subgrain formation and rotation, as well as serrated grain boundaries defined by new grain interfaces) and stress-induced diffusional mass transfer (pressure solution), respectively, indicative of stresses sufficient to induce the intracrystalline deformation and fluid flux along the Archean–Proterozoic boundary.

Direct contacts between quartzite and gneiss along the southern edge of the belt both units exhibit a contact–parallel cleavage or foliation (Figs. 5a and 5b) that is the earliest recognized Proterozoic fabric denoted S_1a. In thin section, S_1a at the basement contact (Fig. 5c) is observed to comprise an S–C type fabric (Berthé et al. 1979). Here, muscovite lenses define the primary slip surfaces that present as the mesoscopic cleavage (Figs. 5a and 5b), while a grain shape fabric formed of elongate quartz and feldspar grains, plus some additional micas, characterizes the S-surface.

S_1a varies in absolute orientation with folding of the contact but maintains a fixed orientation relative to both the basement contact and gross layering (Fig. 6) within the quartzite. The northern edge of the belt exhibits distinct quartz ridges that define the boundary between gneissic basement and overlying units. Here, S₁ surfaces can contain a strong lineation (Figs. 6a and 6b) formed by the intersection of isoclinally folded quartzite layering that is presumptive, relict bedding (Fig. 6c). The intersection lineations lie within the isocline axial planes that are effectively defined by the S_1a layering; that is, S_1a and the isocline axial planes lie at a small angle to each other, both lying subparallel to fold hinges and demonstrating that the dominant layering is itself not “unaffected” bedding. Throughout the quartzite, folds of different generations can be observed in different stages of accumulated strain (Figs. 6c–6e); the end morphology of most is as recumbent isoclinal folds with axial surfaces parallel to the regional lithological contacts (Figs. 6e and 6f). In combination, the observed deformation geometries demonstrate a sequence comprising D_1a (near) bedding-parallel displacement mechanically controlled by the basement–cover contact and bedding, both of which would be initially subhorizontal (Fig. 5b) followed by two additional isoclinal and recumbent (Figs. 6d–6f) fold generations (F_1b, F_1c) of quartzite layers and older fold axial surfaces. Major F_1b folds can have hinges effectively orthogonal to each other (Figs. 6e and 6f) implying strong reorientation during evolution of the isoclinal folds. The latter deformation is inferred to have produced an overall subhorizontal array of axial surfaces with the apparent thickness of the Ayagaq Lake formation being increased by D₁ folding.

Kilometre-scale D₁ isoclinal folds of quartzite (Fig. 3) occur along the southeastern edge of the belt, within the northeastern imbricate thrust sequence and along the northwestern edge of the belt (see Discussion). The differing orientations of the latter reflect later folding. Notably, many of the regional quartzite isoclines form in combination with imbricate basement thrusts. Overall, the Ayagaq Lake formation can be described as a package of thrusts and isoclinal folds with a common end-state D₁ axial surface.

The regional basin-like disposition of the Ayagaq Lake quartzite arises from the shallowly plunging, reclined synforms imposed by D₂. The dominant regional cleavage/foliation, S₂, serves to discriminate D₁ and D₂ structures (Fig. 6g). Whereas D₁ foliations are effectively parallel to lithologic layering throughout the belt, S₂ clearly transects the latter with belt-parallel north−northeast trends and dips to the west–northwest >45°. D₂ controls the current orientation of D₁ structures, which can place structural elements of both tectonic phases parallel to each other; the latter can cause difficulties in discriminating foliation generation in cases of poor exposure.

The Resort Lake formation, a graphitic, pyritic unit that stands in abrupt contrast to the underlying Ayagaq Lake formation most commonly forms the base of Ps2 (Fig. 2) and is itself overlain by thick siliceous dolomitic limestones, the Aluminum River formation, that provides a marker within the Amer belt second only to the basal quartzites.

Deformation features within the Resort Lake formation are routinely concealed by its recessive nature that sees it occurring as highly weathered detritus along the top of the Ayagaq Lake quartzite or base of the Aluminum River carbonate. Less graphitic, coarser grained components of the Resort Lake formation are more resistant and preserve deformation fabrics in situ. The latter most commonly exhibit multiple, anastomosing slatey cleavages intersecting at shallow angles that collectively produce a composite S₁ foliation (Fig. 7a) with isolated isoclinal fold closures. Relatively lower strain exposures (albeit still intensively deformed) exhibit successive S₁ cleavage overprinting older generations (Fig. 7b). A rare occurrence of multigenerational structures is recorded in Fig. 7c. Here, very fine-grained graphitic phyllite, now eroded, lies structurally below a coarser grained, quartzo-feldspathic layer. The ensuing competency contrast accentuates the tightness of hinges in the phyllitic later, such that the coarser grained unit exhibits mullion-like fold hinges (Wilson 1953). The youngest foliation is assigned to S_1c, axial planar to the largest folds and defines the surface parallel to the lithostratigraphic boundary with overlying the Aluminum River formation. The dominant layering is then assigned to S_1b which is folded to produce L_1c lineations parallel to the fold axes. S_1b (actually composite S_1a/b) carries mineral and intersection lineations, L_1b, folded about the F_1c fold hinge. The persistence of this deformation sequence throughout the Showing Lake formation is demonstrated by comparable thin sections fabrics within phyllitic layers (Fig. 7d). Sequential D₁ cleavages are expressed by multiple crenulation foliations. The oldest observed foliation, S_1a/b, is itself composite slatey cleavage, justifying assigning it to S_1b. The axial planar S_1c foliation is preserved in an early stage nearly orthogonal to S_1b. S₂ lies in the steep plane bounding the fine-grain–coarse grain contact and is poorly expressed or weakly developed at the site in Fig. 7c. Incipient S₂ occurs in thin section as a crenulation of S_1c.

The Aluminum River formation, in the main, forms tabular outcrops resembling bedded carbonate (Fig. 8a); however, the dominant layering is instead a pervasive compositional layering comprising variably distorted, metamorphosed dolomite–calcite–quartz (chert) bands (Fig. 8b) confined by gross compositional layering that is presumed to have had an origin as primary bedding. The most intense deformation is confined within layers of relatively lower competence bounded by more competent layers with the rheological contrast characterized by relative proportions of carbonate (more competent) versus calc-silicate minerals and quartz (less competent, finer layering). Within the bulk lithological layering (Fig. 8c), attenuated, disrupted, and isoclinally folded S₀/S_1a defines composite foliations with F₁ axial surfaces parallel to the main layering. Extreme layer elongation is typical of both F_1a and F_1b isoclines (Figs. 8d and 8e) that routinely exhibit “rootless” folds and isolated hinges characteristic of transposition (Sander 1911 in Turner and Weiss 1963; Sander 1970; Williams 1983). D₁ fold interference style varies between Type 2 and Type 3 (Ramsay 1967) but the significance of such observations is moot given the general rotation and transposition to a single axial surface.

D₂ regional folding produces a moderately to steeply dipping S₂ axial plane foliation (78° → 152°) that overprints the lithological layering and associated structures (Figs. 8d and 9). D₁/D₂ relationships are particularly well demonstrated in the hinge zones of F₂ folds. Within the body of the carbonate unit, compositional layering is the cumulative effect of D₁ folds and transposition, producing a composite S₁ foliation overprinted at a high angle by spaced D₂ cleavage (Fig. 9a). In sections viewed perpendicular to plunge (essentially the F₂ profile plane), F₂ fold closures exhibit layering parallel to F₁ axial surfaces defining Type 3 folds that are folded around the F₂ hinge, with vertical S₂ cleavage (Figs. 9b and 9c). Figure 9b shows compositional layering comprising Type 3 interference with the F₁ axial plane that is in turn folded about fold axis parallel to layering a high-angle D₂ axial plane, producing a Type 2 interference fold. The combined metamorphism and deformation of the Aluminum River sequences locally produce exceptionally fine S₁ fabrics (Fig. 9d) that record steep, axial planar S₂ cleavage at high angles to the compositional layering. The upright to inclined axial surfaces and shallow plunge of F₂ folds complicates the S₁/S₂ relationships in peneplained surfaces typical of the study area. Outside F₂ hinge zones, steepened S₁ layering on F₂ fold limbs is essentially parallel to the S₂ cleavage on horizontal outcrop surfaces (Fig. 8e). This could contribute to the underrepresentation of D₁ observations in previous studies.

The siliciclastics of the Ps3 sequence form the core of the Amer belt synforms (Fig. 3). Two units can mark the base of Ps3—the Five Mile Lake formation basalts and the Three Lakes formation phyllitic silt/sandstones. The Five Mile Lake basalts are restricted to the southern edge of the Amer belt where they overlie either Resort Lake phyllites or underlie Aluminum River formation carbonates. The basalts and underlying units along the southern edge of the belt are thrust and folded into the previously noted major D₁ isocline (F1^**, Fig. 3) that parallels the strike of the belt. With the exception of the Five Mile Lake-cored syncline, the basal unit of Ps3 is the Three Lakes formation comprising phyllitic siltstones and sandstones.

D₁ structures occur as tight to isoclinal folds with variably orientated fold axial surfaces depending on their location on later D₂ folds (Figs. 10a–10c). Fold asymmetry is rare given the tightness of folds but where present indicates northeasterly vergence (Fig. 10a). Extremely long-limbed F₁ folds (Fig. 10c) preserve what is probably the clearest folding of bedding. D₂ reorients the aforesaid folds into the steeply dipping regional trends (Fig. 10c) on F₂ limbs such that S₁ fabrics now parallel S₂ cleavage. In the absence of exposed fold closures, outcrops of these “bedded” units present as extensive arrays of vertical to steep bedding. The consistency of the latter deformation sequence is preserved in phyllitic units (Fig. 10d) which largely comprise a composite S₁ foliation from which the fabric evolution is obliterated by transposition. The full sequence of F_1a to F_1c folding and ensuing generation of the generic S₁ composite foliation can be fortuitously preserved in thin section (Fig. 10d). Incipient S₂ foliation can be seen forming as spaced crenulation cleavage that overprints all the earlier deformation fabrics.

Succeeding units assigned to sequence Ps3 comprise the Oora Lake (salmon pink arkose with carbonates at top) and the overlying Showing Lake (carbonate-bearing silitstones and mudstones with distinct magnetic horizons) formations. Differentiation between these siliciclastic units can be difficult. Along the southern edge of the northeast extent of the Main Synform (Fig. 3), near its contact with the Showing Lake formation, the Oora Lake formation exhibits well-preserved, very early D₁ features (Fig. 11a). Arkose beds intercalated with thin silt layers show incipient bedding-plane thrusts cutting up-section and defining S_1a in conjunction with asymmetric folds. In highly strained arkose, spaced cleavage (Fig. 11b) defines cleavage lithons that are asymmetrically folded in concert with continued layer–parallel displacement. The uppermost Oora Lake formation is a carbonate unit (dolarenite) with a composite S₀/S₁ foliation parallel to the lithological contacts (Fig. 11c) overprinted by steeply dipping S₂ cleavage, as are all Ps3 units.

The Showing Lake formation can resemble the Oora Lake formation but is discriminated by having a larger fraction of silt and mudstones; importantly, the Showing Lake formation contains distinctive magnetite-rich horizons of considerable lateral extent (Fig. 2). Also, the contrasting initial bed compositions provide excellent markers for observing structures. Isoclinally folded arkose/siltstone layers (Fig. 12a) establish the composite S₁ layering seen throughout the unit, while in finer grained, mudstone units, S₁ is expressed as slatey cleavage (Fig. 12b). In open hinge zones of F₂ synforms (Fig. 12c), early stages of S₂ foliation can be observed developing by transposition of the S₁ compositional layers (Fig. 12c). In tighter F₂ hinge zones, S₂ presents as steep spaced crenulation cleavage.

Within the map area, sequence Ps4 is limited to isolated exposures (Fig. 3) overlying Ps1–Ps3 above a significant unconformity (Fig. 2). Molasse units of the Itza Lake formation contain extensive fluvial sedimentary features (cross-lamination, graded bedding) that are only mildly deformed (warped) and contain no D₁ deformation fabrics described in the underlying sequences, consistent with exhumation and erosion to produce the unconformity. S₂ cleavage is weak and sparsely developed.

The pre-emptive summary provided at the beginning of this section can now be justified. The individual Proterozoic sequences exhibit wide variations in deformation intensity and style. Notwithstanding the latter, a consistent sequence comprising sequential generations of isoclinal and transposed structures is identified in sequences Ps1–Ps3.

Three generations of D₁ thrusting, folding, and layer transposition can be systematically identified through overprint relationships (Figs. 6e, 7b–7d, 8e, 10d, and 12d). Although presumably related originally to primary sedimentary layering, the current compositional layering (S₀/S₁) reflects three cycles of layer parallel deformation and fabric transposition. Whereas D₁ foliations can be of three generations, S_1a, S_1b, or S_1C, they commonly occur as an undifferentiable composite S₁ that mesoscopically is the end-state of the various transposed foliations. At high strain typical of most exposures, the individual generations of structures, e.g., S_1a, S_1b, or S_1C, are subsumed into a single denoted S₁, exemplified by a ubiquitous composite compositional layering or intense slatey cleavage depending on the specific lithology. D₁, thus, produces a lithostratigraphic sequence with major contacts parallel to internally transposed compositional layers that are themselves approximately parallel to an array of fold axial surfaces and similarly orientated early thrust faults.

Although the sequence of D₁ folding occurs throughout sequences Ps1–Ps3, the subjective intensity of strain is greatest within phyllitic (e.g., Resort Lake, Three Lakes formations) and other relatively low competency units such as the Aluminum River formation. The more competent units, such as the Ayagaq River and Oora Lake formations exhibit the same sequence of deformation, but internal transposition of the units is more constrained. Additionally, it is the juxtaposition of relatively strong against relatively weak mechanical layers that can induce layer-parallel focusing of displacement. Detachments at stratigraphic horizons are evident in Fig. 3 as discontinuities between Ayagaq Lake (quartzite) and Archean gneiss Ayagaq Lake (quartzite) and Resort Lake (graphitic) formations, repetition of Ayagaq Lake on Resort Lake, Oora Lake on Showing Lake, among other examples of disruptions to the stratigraphy.

The current disposition of the Amer belt owes much to its preservation within D₂ synformal keels. The steepening of arguably near-horizontal sequences to north–northwest-overturned synform limbs has a consequential link to the Medial Zone separating the Main and Oval Synforms. The primary structural attribute of the Medial Zone is the abrupt steepening of D₁ fabrics along its length (Fig. 7c), although it has a varied and complex character. The zone appears to have a long-lived depositional influence in that there are significant facies changes across it, in addition to being a structural break. Faulting and dykes parallel to the zone also contribute to its definition. Post-D₂ northwest-trending vertical to steep faults clearly dissect the Main Synform (Fig. 3), but do not pass through the Medial Zone in any meaningful amount; displacements change from kilometres to decametres from the north to south side of the zone.

In addition to the aforementioned structures, isolated folds (three or four in number) that appear to deform S₂, sometimes as warping, once as a crenulation. They do not fall within existing nor do they provide new systematics for the folding history. Except in one instance, new fabrics (e.g., cleavage) have not been observed. Although not playing a significant role in the current geometry, their occurrence is noted here and their interpreted significance will be discussed in subsequent sections.

Given the complex deformation styles and overprint relationships ascertained by field inspection, determination of the relationships among fabric elements requires a detailed analysis of measured data. The structural element measurements were plotted and analyzed as lower hemisphere, equal area stereonet projections using Stereonet v. 11.4.0 copyright© by R.W. Allmendinger, 2011–2020 (https://www.rickallmendinger.net/stereonet). Raw orientation data are plotted as poles to surfaces for planar elements (π-plots; Turner and Weiss 1963) and are plotted directly as lines for linear elements. First-order analyses of orientation data included examination of the spatial distribution of poles and linear features for nonuniform concentrations by contouring on the hemispherical projection; for the latter, the Kamb method of sampling area is used with contour intervals of 2 standard deviations from a uniform distribution. Likewise, π-plot distributions were fit with great circles (Turner and Weiss 1963) to test for the absence or presence of cylindrical folding. The modelled best-fit great circles include the calculated ideal RP for the distribution.

Conventions for the stereonet plots are as follows. Lithological layering observed in a given location is designated as S₀ and can include definitive bedding as well as gross compositional layering now parallel to composite S₀/S₁ foliations. All generations of D₁ foliations have been conflated into S₁ in recognition of the progressive nature of the folding and the fact that individual D₁ generations cannot be discriminated at many outcrops. B-type lineations that form in association parallel with fold axes of a given generation (Turner and Weiss 1963) have been grouped together and include fold hinges, intersection lineations, mineral lineations, and stretching lineations that meet this criterion; for example, stretching lineations commonly initiate as intersection lineations. These aggregations of lineations are denoted as L₁ and L₂. Poles to S₂ cleavages are plotted separately. As already noted, all specific orientations are given in vector format as inclination → azimuth, that is, dip → dip azimuth for planar features and plunge → plunge azimuth for linear features.

Synoptic orientation plots (Fig. 13) were generated by combining data from throughout the study area regardless of the lithological sequence from which they were acquired. Notwithstanding the ubiquity of multigenerational mesoscopic deformation observed for D₁, orientation patterns belie this complexity, that is, the multiplicity of deformation during D1 is not immediately evident from orientation plots alone.

The distributions in Fig. 13 set the basic patterns for structural elements. The orientation patterns of both S₀ and S₁ poles-to-planes (Figs. 13a and 13b) exhibit strong single maxima lying within well-constrained great circle distributions. The calculated best-fit great circles of S₀ and S₁ poles are represented by their respective minimum eigenvectors to the distributions (RP) which are 04° → 266 for S₀ (Fig. 13a) and 20° → 260° for S₁ (Fig. 13b). The maximum pole concentrations were used to determine the mean orientation of S₀ planes (Fig. 13a) which dips 28° → 188° and S₁ (Fig. 13b) dipping 30° → 210°. The π-plot of S₂ cleavages (Fig. 13c) has a strong single maximum defining mean cleavage dipping 80° → 169°. Dispersion of the S₂ poles is weaker than for S₀ and S₁, but significant number of poles to S₂ remain spread along a weak girdle with best-fit great circle having an RP plunging 20° → 255°. D₁ linear elements in Fig. 13d (F₁ fold axes and intersection lineations) are widely distributed but have a mean orientation plunging 17° → 255°. The plot of L₂ (Fig. 13d) has a very strong concentration of orientations compared to L₁ with the maximum concentration plunging 24° → 242°. Notably, the mean vectors (fold axes) for both L₁ and L₂ are close in orientation, implying near collinearity of the D₁ and D₂ fold axes.

A fundamental attribute of the S₀ and S₁ poles is their distribution along well-defined great circles for which the RPs are close to being parallel. The occurrence of such patterns is indicative of re-orientation of planes by cylindrical folding for which there are specific initial constraints: (1) the layers being folded must be parallel (at least in a similar orientation) and planar prior to the relevant folding and (2) layering must be rotated about a common axis lying within the layering (Turner and Weiss 1963). It follows that given the similar orientations of calculated RPs and mean planes (Fig. 13), both the gross compositional layering, S₀, and most importantly, the multigenerational S₁ cleavage/foliation must have comprised, at the time of the cylindrical folding, internally parallel layers with a common absolute spatial orientation. Additionally, the fold axis must lie within or close to the plane of both S₀ and S₁.

Lineations indicative of D₁ fold axes (Fig. 13d) are roughly distributed along a great circle parallel to the mean S₂ with a mean vector plunging 17° → 255°. Other D₁ lineations plunge shallowly at high azimuthal angles to the mean vector. D₂ lineations are concentrated with southwest trends and a mean orientation of 24° → 242°. The striking aspect of all the linear structural elements is their similar orientation. In effect, the calculated RPs for S₀, S₁, and S₂ are collinear, while the mean vectors for L₁ and L₂ are proximal to the latter. Physically, this suggests that a single fold axis is responsible for the current distribution of planes, in this case, F₂. The dispersion of S₂ (Fig. 13c) still suggests additional effects.

Notwithstanding the general, regional simplicity of D₂ structures, there is a notable degree of dispersion of D₂ fold axis-related lineations and axial plane cleavage. Possible scenarios include (1) misidentification of L₁ and S₁ as L₂ and S₂, (2) dispersion reflects influence of D1 fabric orientations on subsequent D2, (3) an additional collinear generation F₃ or D_2b, or (4) formation of cleavage in progressive manner throughout D₂ folding.

The current near-coplanarity of mean S₀ and S₁ is fundamentally significant because it exists in concert with intense internal deformation. The similar orientation maxima of overprinted S₀ and later D₁ foliations is explained by the transposition of bedding and early tectonic fabrics to produce a single dominant composite foliation, S₁; that is, the lithological sequence is not the simple preservation of primary stratigraphy. The dispersion of D₁ lineations reflects both the noncylindrical nature of many D₁ folds, the reorientation of said folds, especially rotation of initial fold hinges from northwest–southeast trends to southwest–northeast, parallel to the F₂ fold axis.

The key element of the stereonet analysis is that at the termination of D₁ the Amer belt lithostratigraphic sequences retained a high degree of parallelism with much bedding transposed into a near-coplanar D₁ foliation while accumulating significant strain during discrete episodes of folding and transposition. The end-state is a highly deformed sequence dipping moderately to the southwest with intense linear structure plunging with a similar trend. Subsequent cylindrical folding during D₂ is effectively controlled by this pre-existing fabric whereby F₂ fold lies in the plane of S₁ and approximately parallel to L₁.

Being the synoptic analysis provides an overall description of the structural evolution within the northeastern Amer belt, there remain stark contrasts in deformation styles between the two principal morphological domains (Fig. 3): (1) the Main Synform largely north of the medial line and (2) the Oval Synform south of the Medial Zone. Although both areas exhibit similar mesoscopic structural styles, they remain distinct at the map scale. Each of these domains has been analyzed separately in the order of the lithostratigraphic sequences in an attempt to extract any nuanced differences that may have been masked by conflating all data into the synoptic plots.

The Main Synform extends from the exposed southeast-plunging closure in the northeast of the belt to the limit of data acquisition approximately 100 km to the southwest (Fig. 3) and is bounded on the southeast by the Medial Zone and, where parallel, the southern exposure of the Ayagaq Lake formation. In the Ayagaq Lake quartzite, both bedding and S₁ foliations exhibit tightly clustered pole maxima with mean planes dipping 29° → 196° (Fig. 14a) and 29° → 224° (Fig. 14b), respectively. The similar orientation between bedding and the generationally undifferentiated D₁ foliations (Fig. 14c) is consistent with the outcrop-scale transposition to generate the current southwesterly dipping panel of rocks. The low dispersion of planes from the mean orientations argues for the orientation maxima contains the immediate end-D₁ orientations.

S₂ cleavage planes (Fig. 14c) are distributed along the plane perpendicular to the F₂ fold axis. Such variations in cleavage orientation could result from misidentification of rotated S₁, as described earlier, or this is an indication that S₂ cleavage development is progressive throughout D₂; likewise, the distribution of F₂ lineations (Fig. 14c) can be explained by formation of folds on nonparallel D₁ although the fold axis remains common to all. In the latter case, they should like on the D₂ axial plane, which they do not.

Of sequence Ps2, the Resort Lake formation provides a sparse data set (Fig. 15a) reflecting its poor exposure. Mean S₀ and S₁ have effectively the same orientation, with S₀ dipping 30° → 225°, and S₁, 33° → 225°. The high degree of parallelism between S₀ and S₁ is considered a qualitative indicator of strain intensity whereby the fabric is evolving toward an end-orientation in the field. The Aluminum River formation (Fig. 15b) is typified by shallowly dipping mean bedding (16° → 240°) and D₁ foliations (10° → 218°) that are also effectively parallel. The mean S₂ dips steeply southeasterly at 78° → 152°, although the two clusters of cleavage poles suggest two dominant, discrete orientations. Rotation axes (RP) were calculated for all planar distributions cluster together (Fig. 15b) lending support to the existing of a single D₂ fold axis.

D₁ and D₂ fold axis-parallel lineations are concentrated on the mean D₂ fold axis (Fig. 15c), as well as the clustered RPs in Fig. 15b. The D₁ lineations are distributed along the mean S₂ plane, as well as at high angles to S₂. Whereas L₁ lineations occupy the same orientation fields as L₂, the converse does not hold. Given the relative timing of the lineations, the L₁ distribution reflects rotations during different generations of D₁ as well reorientation during D₂ folding. The southwesterly plunging concentration of L₁ is interpreted as the end-orientation of D₁ lineations formed during folding and transposition. This inference is supported by the observation of sheath fold axes (Fig. 15c, solid blue diamond) in this orientation only, that is, rotations and extension of early D₁ fold axes produces the observed loci of L₁. The northeasterly plunging D₂ lineations (Fig. 15c) are consistent with formation of fold hinges on D1 structures, and need not indicate an additional deformation event.

Sequence Ps3 of the Main Synform has several distinct orientation signatures. As with the Resort Lake formation, S₀ is virtually absent from Three Lakes formation while S₁ exhibits a strong orientation maximum (Fig. 16a). The calculated mean orientations for both S₀ (32° → 210°) and S₁ (33° → 218°) are effectively parallel, while D₂ lineations plunge shallowly west–southwest, a distinctly westerly trend. The Oora Lake formation stands out in having mean bedding (61° → 172°) and S₁ foliations (58° → 178°) dipping much steeper and to the south (Fig. 16b) than any other units. S2 is near-vertical dipping 87° → 349° (Fig. 16c).

The Showing Lake formation has south–southwest dipping mean bedding (24° → 201°) and foliation (36° → 196°) both of which are broadly dispersed within great circles (Fig. 16b). Overall, mean D₁ and D₂ fold axes are approximately collinear with the calculated best-fit RPs. Overall lineation orientations vary little between different units and are largely collinear with the calculated RPs.

Orientation patterns of the structural elements within the Oval Synform (Fig. 17) have general similarities, but some notable differences, compared with those of the Main Synform. Ps1 Ayagaq Lake quartzite exhibits a strong, elongated π-girdle maximum (Fig. 17a). The calculated mean bedding orientation (13° → 037°) and S₁ have virtually the same orientation (Fig. 17a). In contrast to the Main Synform, F₁ and F₂ fold hinges (Fig. 17a) plunge easterly at <10° as do the calculated rotation axes (RP) for best-fit great circles. The concentration of bedding poles giving the mean north–northeast dip is biased by the preponderance of exposures at the western end of the synform (Fig. 3). The southern edge of the synform instead comprises vertical to overturned beds as a result of D₂ folding. S₂ cleavage (53° → 173°; Fig. 17a) defines the reclined nature of the current synform.

The Ps2 Resort Lake unit (Fig. 17b) presents as a highly cleaved phyllite dominated by S₁ with few other features. D₂ fold axes plunge shallowly west–southwest, the opposite trend of those in the underlying Ayagaq formation. These opposing plunges are reflected in the doubly plunging nature of the overall Oval Synform. The mean S₁ (61° → 173°) has an orientation similar to that of S₂ in overlying (Fig. 17a) and underlying (Fig. 17c) units suggesting that D₂ folding has reorientated many of the sampled S₁ foliations into parallelism with S₂.

Albeit based on scant data, the Oora Lake formation records F₂ hinges with opposing plunges and disperse layering about these lineations. Like the Ayagaq Lake formation, mean Oora Lake bedding (Fig. 17d) dips distinctly towards the north, as opposed to the mean west–southwest dips of the Main Synform. The Showing Lake formation (Fig. 17e) layering is distinctive in forming strong maxima indicative of southerly dipping mean planes (S₀ 53° → 175°; S₁ 45° → 184°).

The current geometry of the Amer belt results from a few basic influences: (1) sequences Ps1–Ps3 of the Amer belt formed as an initially (sub)horizontal array of mechanically contrasting lithologies on Archean basement; (2) at the onset of tectonism, layer-parallel displacements of the strata occurred in conjunction with thrusts, rheological detachments, and recumbent, isoclinal folds; this deformation, D₁, involved several generations of ductile overprinting, and had the effect of producing a shallowly dipping, imbricated, and transposed tectonic architecture of subparallel primary and tectonic fabrics; (3) subsequent exhumation, erosion, and deposition of sequence Ps4 were followed by upright to inclined folding of the D₁ architecture by F₂ folds to produce steep fold limbs. Early (in the development sequence) F₁ fold axes trend approximately 90° to those of F_2. The latter observations, as well as those of Calhoun et al. (2014) verify and extend the structural paradigm established by Pehrsson et al. (2013)wherein sequences Ps1–Ps3 are distinguished from sequence Ps4 in the central Rae Craton. Since the lower three sequences are penetratively deformed during D₁ and assigned to the ca. 1.9–1.865 Snowbird orogeny, sequence Ps4 was unaffected by D₁, and deformed solely by D₂ during the ca. 1.87–1.81 Ga Hudsonian orogeny. Variations in the intensity of S₂ in Ps1–Ps3 sequences versus Ps4 and the deviations’ orientations of S₂ across F₂ synforms suggests that the D₂ fabrics may record concomitant deformation, exhumation, and erosion, that is, S₂ is somewhat deformed (rotated) by D₂ without the need to invoke a separate deformation.

The relative simplicity of D₂ folding is central to unravelling aspects of the deformation, in conjunction with the assumption that the Archean basement/Proterozoic contact was initially close to horizontal. The high degree of cylindricity of D₂ folding defining the π-plots of S₀ enables rotation about the F₂ fold axis to produce a subhorizontal basement contact; concomitantly, subparallel bedding and D₁ foliations are also rotated into shallowly dipping orientations. For example, opposing limbs of F₂ folds can be rotated into similar subhorizontal orientations. A principal difficulty in interpreting older structures is now removed in that D₁ features steepened by D₂ can now be clearly discriminated. These are imperfect reconstructions but constrain the belt-scale behaviour.

Early, subhorizontal displacements are recognized through bedding-parallel movement, imbricated thrusts, and recumbent, isoclinal folds. The earliest tectonic fabric in the Proterozoic units, denoted S_1a, occurs at lithological contacts in concert with evidence of displacement. S/C fabrics at the Archean/Proterozoic contact (Figs. 5b and 5c) indicate shear at the base of the sequence. Bedding-parallel displacement, imbricate thrusts, and incipient folds in the Oora Lake formation (Fig. 11a) provide specific evidence of the kinematics. By assuming an initial horizontality to the beds, S_1a can be reoriented, and gives an initial orientation of 20° → 196°, that is, displacement is towards 015°. The latter orientation of the S_1a cleavage correlates closely to the mean orientations of the aggregate S₁ foliations (Figs. 14b; 15a, 15b; 16a, 16c) within the Main Synform. Such correlations have important implications: (1) displacement, both brittle and ductile, was directed toward the north–northeast to east–northeast along planes dipping generally to the southwest at 10°–30° and (2) where a sufficient number of measurements exist (Fig. 15c), D₁ lineations are concentrated in a similar southwesterly trend. This correlates with the rotation of intersection lineations and fold axes from initially at a high angle to the displacement direction (Fig. 15c), toward parallelism with the displacement (extension) direction during ductile flow, i.e., the end-orientation at high strain.

Belt-scale thrust imbrication is most obvious between Ayagaq Lake quartzite and Archean basement along the northern and northeastern edge of the belt (Fig. 3). The nature of thrusting can be illustrated by reconstructing geometries of the Ayagaq Lake formation prior to D₂ folding and northwest-trending faulting. Current exposures from the northeastern end of the belt have been projected into a common horizontal plane (Fig. 18) after removal of the D₂ fold rotation to produce a shallowly dipping S₀/S₁. Again, this is feasible because of the relatively straightforward near-cylindrical folding by D₂ identified in the stereonet analyses. The reconstructed imbricate slices (Fig. 18) verge to the northeast, consistent with in-sequence thrust propagation and displacement in the same direction. An interesting consequence of the reconstruction is the effect on the Medial Zone thrust which appears to be geometrically consistent with other imbricate thrusts. This is noteworthy because of the enigmatic nature of the Medial Zone as a now dominantly steep zone.

The basement–quartzite thrusts often occur somewhat below the contact. In this way imbricate slices can isolate the basement/cover contact so as to preserve the original quartzite–gneiss contact locally, while in other cases, thrusts move along the contact. Therefore, two flavours of contact can coexist, one with primary unconformable relationships, and the other as a tectonic contact. The imbrication of the basement is believed to be more extensive than existing data can confirm and there is map-scale evidence for D₁ Archean gneiss–quartzite folds (Fig. 3).

In conjunction with thrusts, recumbent isoclinal folds dominate the D₁ structures (Figs. 6, 8, 10, and 12) from thin section to map-scale. In reconstructing Fig. 18, a kilometre-scale isoclinal fold is identified with its axial trace parallel to the thrusts, comparable to the mesoscopic example in Fig. 11a. The multigenerational isoclinal folding and transposition in turn produces the closely correlated orientations of bedding and S₁ foliation (e.g., Figs. 14, 15, and 16). F₁ folds and related thrust imbricates are illustrated in a pre-D₂ configuration in Fig. 19. Folds are reclined to recumbent, are intimately associated with basement thrusts, and verge to the northeast in the same sense as the thrusts. This is the same sense of vergence recognized by McEwan (2012) within the Ketyet River belt to the south. The S₀/S₁ intersection at this scale correspond to the mesoscopic lineations in Fig. 6a. Large D₁ folds reiterate the degree to which D₁ affects the overall architecture of the belt.

The southern edge of the belt differs dramatically in outcrop pattern from that of the north (Fig. 3) in that equivalent D₁ features appear absent. This ambiguity is at least partially erased when the overturned Archean–quartzite contact (southern limb of D₂ fold) is returned to horizontal. The current distribution of imbricate slices reflects the effect of D₂ which steepens and overturns the southern edge of the belt such that initially shallow-dipping faults are masked within the now steep lithological layering and/or are reactivated as late steep faults (TF in Fig. 3). Likewise, the tight F₁ syncline extending along the southern edge of the belt (Fig. 3), cored by the Five Lakes formation, converts from a steep D₂-like feature to a shallow D₁ thrust–fold complex (see Fig. 21).

Thrusts controlled by bedding planes are less easily identified in Ps2 and Ps3 sequences. Thrusts are commonly localized at a “weak” interface (phyllite, marble), especially in the case of more competent layers such as Ayagaq Lake and Oora Lake formations. It appears that discrete displacements observed in Ps1 are accommodated within less competent units at higher stratigraphic levels without explicit faulting. There are two principal styles of D₁ deformation with Ps2 and Ps3 sequences: (1) intensely cleaved and foliated units such as the Resort Lake formation and (2) units in which isoclinal folds and transposition are ubiquitous, such as the Aluminum River formation.

The restriction of high shear intensity parallel to the gross lithological layering (e.g., Figs. 8a and 8b) enables individual units to remain isolated from each other by detachments (zones of high strain) during D₁. Within D₁, noncoaxial fabrics range S/C foliations in the basal micaceous schist through to the extreme extension and transposition evidenced within the Aluminum River carbonates. Considered as a well-bounded subhorizontal zone of high strain during D₁, the Aluminum River formation has the character of a layer-parallel shear zone or at least a zone accumulating extensive noncoaxial strain. Except in cases of imbricate stacking or on the overturned limbs of large recumbent F₁ folds, the normal stratigraphic sequence is maintained in the right-way-up orientation (Fig. 19). Identifying the recumbent nature of large F₁ folds can be particularly problematic in Ps3 given the difficulty in discriminating between Oora Lake and Showing Lake formations and the paucity of way-up features in these units. The latter was first recognized during mapping by Blackwell (1978) and confirmed in legacy drill core during this study.

The absolute resolution of some of the aforementioned issues can only be achieved by 3D analysis. The inset for Fig. 20 indicated in Fig. 3 is part of the Main Synform for which more detailed mapping was possible given the relatively large number of exposures (compared to, for example, the Oval Synform), the abundance of mesoscopic structures preserved in compositionally layered Showing Lake formation, and the presence of a distinctive magnetite-rich horizon in the Showing Lake unit that provides an aeromagnetic signature giving a useful guide to structure. Although the different generations of fold axial traces can be observed in Fig. 3, the detail in Fig. 20a gives a clearer demonstration of the geometries. The area contains a major F₂ fold closure segmented along its axial trace by steep northwest-trending faults into three subareas, of which Fig. 20a is the central block. Both F_1b and F_1c fold axial traces can be identified (Fig. 20a). The polyphase nature of the folding is demonstrated by the vergence of (F_1b) minor folds on the limbs of the F₂ structure that are antisymmetric indicative of a simple D₂ syncline. F₂ fold axial traces have orientations that vary within lithostratigraphic sequence. The difference in F₂ axial trace orientations between Ps1 and Ps3 (Fig. 20a) could be caused by relative displacement within Ps2 during even D₂.

As the best-constrained area, a formal orthographic block diagram (Fig. 20b) was constructed to represent the geometry of the Main Synform at the conclusion of D₂ but before truncation by the northwest-trending faults. Movements on northwest-trending faults was removed by calculating the vertical offset needed to match the map plane displacement of inclined markers (e.g., bedding contacts). Although subject to some simplifying assumptions about fault displacement, the fit was judged to be acceptable. The contact between the Oora Lake and Showing Lake formations was chosen as the projection surface. Projection of surface data produced the cross-hatched surface shown in Fig. 20b. The reclined D₂ structure can now be seen to comprise closed, folded surfaces defined by the stratigraphic contact. The folded surfaces are D₁ structures that developed as early, possibly F_1b, large noncylindrical folds. Even with simplistic rotation of the steepened F₂ limbs, the D₁ structures appear to be Type 2 folds (sheath folds) that are recumbent nappe-like structures, with axes trending northeast–southwest. The latter trend duplicates that of most F₁ fold axes and related lineations, as well as the axes of the two mesoscopic sheath folds identified observed within the Aluminum River formation (Fig. 15c). The high angle between F_1b and F_1c fold axis traces (Fig. 20a), is sequentially consistent with the early stage of isoclinal folds having axes near perpendicular to the displacement direction, with increasing strain subsequently rotating axes toward that direction, that is, F_1c tells the early (geometric) history of F_1b.

The Oval Synform (Fig. 3) has been historically problematic in that its morphology stands in contrast to the Main Synform, it is a doubly plunging structure and its relationship to the Medial Zone lineament remains in question. The combined nature of the Ps2 and Ps3 sequences and the paucity of outcrop confound obvious interpretations. A distinctive aspect of the Oval Synform is the asymmetric repetition(s) of strata; these right-way-up, but incomplete sequences (Fig. 3) must relate to detachments (thrusts) parallel to lithological boundaries and subsequent truncation of units. However, there is minimal evidence of such truncation against adjacent units.

Tschirhart et al. (2012) and Calhoun et al. (2014) presented vertical cross-sections of the Oval Synform along the Line 6NW–6SE in Fig. 3. The cross-sections were originally created by incorporating surface geology and field acquired magnetic susceptibility and gravity field measurements to produce a self-consistent model. These have been adapted in this contribution to create an end-D₁ cross-section (Fig. 21) along within the same plane, but with the effect of D₂ folds removed. The primary assumption, as elsewhere, is that the basement/cover contact has a pre-D₂ orientation near-horizontal. With the removal of D₂ folding, basically by rotating about a horizontal axis in the case of the Oval Synform, an interpretation of the current outcrop pattern in a pre-D₂ configuration can be made. As the cross-section is roughly perpendicular to the synform, the view is effectively parallel to the F₂ fold axis. The magnetic horizons within the Showing Lake and Three Lakes formations were essential to the 3D interpretation.

Certain attributes stand out. As mentioned previously, the upright Five Mile Lake-cored syncline (F1^**, Fig. 3) is now seen as a recumbent D₁ structure. Reorientation of units on either side of current Medial Zone indicates no early structure of great significance other than shallowly inclined layering and detachments. Discontinuities in the lithostratigraphic sequence are the loci for detachments for which some truncations (imbrications) can be inferred. Such truncations can be hidden within the coeval folding of low competence units (Fig. 21). From outcrop-scale observations, it is argued that these discontinuities are common within the stratigraphic sequence and play the equivalent role within less competent units of discrete imbricates within the quartzite. There is no unambiguous sense of vergence within Ps2 and Ps3 sequences. In addition to the detachments and reordering of strata required by the current stratigraphic sequence, the bulk of the Oval Synform now presents as large recumbent noncylindrical folds (Ramsay Type 2). In this case, the view is along the axis of presumed sheath-like folds. As with the Main Synform, the structures developed here are the minimum to be consistent with observations; it is anticipated that we remain unaware of many other aspects of the geometric details given data limitations.

The current geometry of the Oval Synform does show thrusting to the north over the Main Synform along the belt-parallel Median Zone (Fig. 3). The existing steep dip of the Medial Zone and its long-lived existence as such a structure is at odds with reconstructions such as Fig. 21 that do not provide clear evidence for its future significance. The current disposition of this linear zone argues for some important association with D₂ to generate the steep dips along it. Its longevity as a focus of crustal activity exists in the evidence for strike-slip faulting and dyke injection. A most enigmatic feature is the abrupt change across this lineament in fault displacement along the northwest-trending faults. Such a rapid change in behaviour implies much more extension northwest of the Medial Zone than to the southeast, that is, the Medial Zone could have acted as an extensional detachment. The scattered, apparently nonsystematic folds observed deformed S₂ might be related to such an extensional episode, but there is a general absence of supporting mesoscopic data at this time.

Maintenance of the gross stratigraphic order throughout ample portions of the belt results from a combination of fortuitous geometry and mechanical decoupling parallel to the original bedding with the loci of deformation along these zones. To a first order, the overturned nature of D₁ folds will produce large portions of the belt that remain upward-facing. Subsequently, D₂ folding and erosion has produced a largely fortuitous relationship with F₁ folds whereby large areas of exposure comprise lower limbs of F₁ folds with the right-way-up stratigraphic sequence; in this case, F₂ folds produce what appear outwardly as synclines, but which are in fact more complex synformal structures.

In the western portion of the Main Synform, recumbent folds in the quartzite have axial surface traces trending northwest–southeast parallel to thrust faults between Archean basement and the quartzite (Fig. 3), both indicative of shallowly inclined displacement. The apparent failure of these thrusts to pass into the overlying Ps2 and Ps3 sequences argues for detachment along the Resort Lake horizon and are much harder to identify along the steepened, southern limb of the belt. Asymmetric repetition of Archean basement, Ayagaq Lake, and Ps2 units in the northeast region indicates thrust imbrication (Fig. 18), whereas Ps3 units show no similar repetitions, again suggesting decoupling from the quartzite within Ps2 and/or a major change in thrust propagation so as to concentrate within less competent units.

In addition to the displacements seen as basement/cover imbricate thrusting, large displacement within and along lithologic units is consistent with the documented progressive sequence of D₁ isoclinal folding and high strains. The degree of decoupling between contrasting units is central to maintaining versus transposing an initial layering. The effect of a strong mechanical stratigraphic that nevertheless accommodates intra-unit transposition is to sustain stratigraphic order; the latter would reflect both the competency contrast and overall kinematics. This would suggest a large component of noncoaxial flow during deformation whereby interaction across lithological boundaries is minimized. Although shear zones sensu stricto are not required for the latter kinematics, the structures developed are effectively the same. Preservation of stratigraphy in areas of high strain of this type has been observed in ocean floor sequences with thinning from kilometre- down to decimetre-scale (Molli et al. 2006; Malavielle et al. 2011). In other environments (temperature, pressure, kinematics), compositionally distinct layers may act more as passive markers, such transposition will lead to large-scale mixing and loss of any stratigraphic contiguity. This is evident “within” the Aluminum River formation, where carbonate and silica bedding have been completely transformed from their initial relationships (e.g., Fig. 8d).

It is important to make the point that it is not the case that the stratigraphic order is perfectly preserved. Mixing, or rather juxtaposition of units exists throughout D₁ thrusts and folds involving more than one lithological unit, such as along the southern edge of the belt where Five Mile Lake basalts core a tight F₁ syncline (F1^**, Fig. 3). It has been noted that there can be ambiguity in differentiating the Oora Lake and Showing Lake formations. To a large extent, separation of the units is based on the relative proportion of “salmon pink sandstone” to siltstone and mudstone units; mapped Showing Lake formation could be in part tectonically transposed mixtures of contrasting units. Notwithstanding the sedimentary evidence of lithological transitions, transposition at contacts within “right-way-up” sequences cannot be precluded wherein apparent lithological transitions reflect a significant component of tectonic textural changes (Fig. 21); for example, deformation induced reduction in grain size and intermixing of mud and sandstone units by transposition.

Of major importance is the internal repetition of units by isoclinal folding. This effectively precludes stratigraphic analysis, particular determinations of unit thicknesses, and orientations of sedimentological flow features. The initial occurrence of the Ayagaq Lake quartzite as a thinner unit that has subsequently been repeated in structural sections by isoclinal folds of a multilayer is one explanation for how easily it appears to fold (Figs. 6 and 19). Similarly, the reported metamorphism throughout the Amer belt is both scant and contradictory, potentially reflecting discontinuities in the current architecture. Metamorphic conditions from anchizone (<200 °C) to kyanite amphibolite (∼500 °C) are reported. The deformation of the quartzite is indicative of minimum temperatures of 300 °C. The geometries recorded in this study could increase the ambiguity of the spatial variation in metamorphic conditions (e.g., Patterson 1980, 1986). Mineral assemblages would need to be interpreted as variations in structural depth (temperature and pressure); for amphibolite grade conditions, tectonic thickening on the scale of 15–20 km, of which 3–5 km thickness is preserved. Significant extensional exhumation could, for example, explain higher observed metamorphic grades to the northwest, than south of the belt. There are currently insufficient correlated data to test such ideas.

The principal conclusion of this study is that the level of strain within the Amer belt is substantially greater than indicated in most previous studies. The bulk of the strain has been accumulated during at least three generations of isoclinal thrusting, folding, and related transposition during D₁ (2.05–1.865 Ga Snowbird Orogeny). A sequence of D₁ development can be summarized as

• 1) Bedding parallel displacements and imbricate thrusting of sequences to the northeast involving basement gneiss and basal quartzite (Ps1)

• 2) Large-scale D₁ recumbent folding verging north–northeast associated with imbricate thrusting within Ps1–Ps3.

• 3) Three generations of isoclinal folds and transposition are coeval with formation of large recumbent folds. High strains develop within most units, with some acting as discrete detachment zones, notably the Resort Lake and Three Lakes formation. Folds are distinctly noncylindrical and form sheath-type geometries, particularly in Ps3 units.

• 4) (i) The combination of transposition and rotation of primary and D₁ layering toward a common orientation with rotation of D₁ fold hinges toward displacement direction during noncylindrical folding produces a strong L–S fabric that controls subsequent deformation. (ii) A major unconformity developed between deformed Ps1–Ps3 and overlying Ps4. (iii) D₂ folding produced upright to reclined regional folds and cleavage that defines the current regional synformal geometry (1.90–1.81 Ga Hudsonian Orogeny). Sequence Ps4 was not affected by D₁ and is only mildly deformed by D₂. Faulting of variable character reactivated along the Median Zone, with segmentation of the northwest portions of the belt by northwest-trending extensional faults. The latter have kilometre-scale offsets northwest of the Medial Zone, but upon crossing it, offset decrease to a few tens of metres. Such behaviour carries the implication of extensional accommodation localized along the Medial Zone.

Field work was supported through the Geological Survey of Canada GEM NE Thelon project. Additional field support and laboratory studies were supported through an NSERC Discovery Grant to JCW. Graham Halcrow provided assistance in collating the structural measurements for stereonet analysis. Stephen Delahunty and Calvin Nash of the UNB Earth Sciences workshop provided their usual excellent thin sections. K. Bethune and E. Thiessen provided detailed and beneficial reviews that improved the final paper. Finally, Grant Young is fondly remembered for the intellectual honesty he displayed throughout his many years of mentoring.

Data generated or analyzed during this study are available from the corresponding author upon reasonable request. The principal data repository is Calhoun (2016), with additional data held by Geological Survey of Canada, C.J. Jefferson, and J.C. White.

Conceptualization: LC

Data curation: JCW, LC, CWJ

Formal analysis: JCW, LC, CWJ

Funding acquisition: JCW, CWJ

Investigation: JCW, LC, CWJ

Methodology: JCW, LC, CWJ

Project administration: CWJ

Resources: JCW, CWJ

Software: JCW, CWJ

Supervision: JCW, CWJ

Visualization: JCW

Validation: LC, CWJ

Writing – original draft: JCW

Principal funding was provided through NRCan by the Geological Survey Geomapping for Energy and Minerals and Secure Canadian Energy Supply programs. Additional funding was provided through an NSERC Discovery Grant to JCW.