Constraining the tectonic evolution of rifted continental margins by U–Pb calcite dating

We employ U–Pb calcite dating of structurally-controlled fracture fills within crystalline Caledonian basement in western Norway to reveal subtle large-scale tectonic events that affected this rifted continental margin. The ages (15 in total) fall into four distinct groups with ages mainly ranging from latest Cretaceous to Pleistocene. (1) The three oldest (Triassic-Jurassic) ages refine the complex faulting history of a reactivated fault strand originated from the Caledonian collapse and broadly correlate with known rifting events offshore. (2) Two ages of ca. 90–80 Ma relate to lithospheric stretching and normal fault reactivation of a major ENE-WSW trending late Caledonian shear zone. (3) We correlate five ages between ca. 70 and 60 Ma with far-field effects and dynamic uplift related to the proto-Iceland mantle plume, the effect and extent of which is highly debated. (4) The five youngest ages (< 50 Ma) from distinct NE–SW trending faults are interpreted to represent several episodes of post-breakup fracture dilation, indicating a long-lived Cenozoic deformation history. Our new U–Pb data combined with structural and isotopic data show that much larger tracts of the uplifted continental margin of western Norway have been affected by far-field tectonic stresses than previously anticipated, with deformation continuing into the late Cenozoic.


Results
We collected 35 calcite samples from fractures and obtained 15 meaningful U-Pb ages from 14 different sites. The remaining 21 samples did not contain enough uranium (U) to extract reliable age information. All successfully dated calcites were collected from fracture planes lacking slickensides. Three different types of calcites have been dated (Fig. 2): type 1 represents calcite clasts or void filling incorporated in cataclasites within major brittle fault zones formed related to extensional faulting; type 2 represents calcite veins formed along distinct fault slip surfaces during extensional faulting, and type 3 represents simple dilational fractures not connected to brittle fault zones. We assume that the normal faults (type 1 and 2 calcites) formed at 30° to σ1, whereas the dilational fractures (type 3) formed parallel to σ1 and perpendicular to σ3. For detailed sample descriptions and U-Pb data, see Supplementary Material. Based on orientation, U-Pb age, calcite type and stable isotope composition, we assigned our results to four distinct groups, which in the following are described from the oldest to the youngest.
Triassic to Cretaceous ages from the Dalsfjord fault. The three oldest obtained U-Pb calcite ages are from the Dalsfjord fault, separating the Western Gneiss Complex from Caledonian crystalline nappes 14 , through fault activity in the Late Permian-Early Triassic, Jurassic and Cretaceous 12,15,16 (Fig. 1). The Dalsfjord fault contains distinct zones of green and red cataclasite as well as a prominent layer of fault gouge 12,15,17 (see Supplementary Material 1). The two oldest calcite U-Pb ages come from a sub-rounded calcite clast or void filling within red cataclasite of the hanging wall (208.0 ± 25.0 Ma-VAH_286_8; Type 1; Fig. 3a, Table 1) and from a fracture plane cutting red cataclasite in the footwall (206.4 ± 6.2 Ma-VAH_286_2; Type 3; Fig. 3a (Fig. 3b), while the δ 13 C value is − 5.8‰ for both samples (Fig. 3c). The younger sample shows a δ 18 O value of − 12.5‰ (Fig. 3b) and a δ 13 C value of − 9.3‰ (Fig. 3c).

Discussion
Our new 15 U-Pb calcite ages are the first radiometric constraints of calcite-filled fractures from the onshore rifted margin of western Norway. They span across ~ 200 million years of tectonic evolution and correlate to some extent with well-known tectonic events but reveal also more subtle tectonic processes. The two oldest calcite ages yield two possible interpretations for the age of the red cataclasite, both being older than the previously suggested age of c. 150 Ma 15 . In our preferred textural interpretation, the oldest calcite of 208 ± 25 Ma represents a calcite clast deformed within the cataclasite (Type 1) and the younger calcite of 206 ± Ma comes from a fracture cutting the cataclasite (Type 2). In this case, the two ages bracket the age of red cataclasite formation, which then must have occurred in the latest Triassic, representing a faulting phase not previously been detected (Fig. 3). Alternatively, if the age of 208 ± 25 Ma comes from a void-filling calcite postdating cataclasite formation, then both ages represent minimum ages for red cataclasite formation, which then is constrained to have occurred in the Triassic between the 260-248 Ma green cataclasite and the 208-205 Ma calcite ages. The youngest calcite age142 ± 16 overlaps with the paleomagnetic age of ca. 150 Ma and represents renewed fault activity in the latest Jurassic-earliest Cretaceous, prior to gouge formation. The older ages of this study coincide with the observed increase in fault activity in western Norway prior to the Late Jurassic rift phase 12,18 (Fig. 3d), whereas the younger age coincides with the prominent phase of rifting offshore in the Late Jurassic-Early Cretaceous 19 (Fig. 3a,f).  Fig. 1), temporally coinciding with the calcite U-Pb ages (Fig. 3d). The NNW-SSE trending σ 3 inferred from the fracture orientations is compatible with the Cretaceous regional stress field affecting the mid-Norwegian passive margin at this time 18,21 . The timing of fracture formation corresponds to the renewed onset of rifting between Greenland and Norway at around 80 Ma, after a ~ 40 Ma period of quiescence 22 . This lithospheric extension left the crust flexed and weakened, and erosion induced unloadingloading along the margin has been interpreted to have caused Cretaceous top-NW-down normal reactivation of the Møre-Trøndelag fault system 5 (Fig. 1). We interpret the 90-80 Ma calcite-filled fractures to be the product of the same process (Fig. 3f). The ages also coincide with pulses of deep-water turbidite deposition offshore 23 .

70-60 Ma: large-scale doming prior to North Atlantic break-up? Five samples with ages between
70 and 60 Ma represent a regional period of fracture opening and calcite precipitation, pre-dating North Atlantic break-up. This time frame broadly coincides with the emplacement of the proto-Icelandic plume. It resulted in the onset of widespread volcanic activity from 63 to 62 Ma (North Atlantic Igneous Province or NAIP), the formation of regional unconformities offshore and dynamic uplift of off-and onshore regions during the Paleocene along the NW European margin [24][25][26][27][28] . The NAIP was active until Early Eocene times, ending c. 54 Ma ago 24,29 (Fig. 3a). Different models have shown that the arrival of the proto-Icelandic plume could have caused uplift starting at c. 70 Ma and reaching a maximum of 0.1-0.6 km around 56-55 Ma 28,30 . Uplift of western Norway is supported by the presence of an up to 2 km thick clastic sedimentary wedge on the Måløy Slope offshore of the study area, which has been related to hinterland uplift 31,32 , and which overlaps in time with our fracture calcite ages (Fig. 1, 3e). We therefore interpret the 70-60 Ma calcite-filled fractures to represent onshore dynamic uplift affecting the NW Europe margins during the arrival of the proto-Icelandic plume (Fig. 3f). The variable orientation of the fracture surfaces in this group may indicate that fracture formation was related to large-scale domal uplift rather than the result of a regional unidirectional stress field (Fig. 1). < 50 Ma: unravelling Eocene, Miocene, Pliocene and Pleistocene tectonic pulses. The five ages that are < 50 Ma were all obtained from NE-SW striking fractures (Fig. 1), and document continued tectonic activity throughout the entire Cenozoic (Fig. 3d). The offshore stratigraphy from the northern North Sea (Fig. 3e) shows an Oligocene phase of increased sediment input around 33-27 Ma, as well as a Mid Miocene unconformity from c. 25 to 8 Ma 33 . Different mechanisms have been suggested to explain these observations, such as long-term tectonic uplift due to far-field compression, episodic normal fault reactivation related to lithospheric flexure, and/or climatic variations including Quaternary glaciations, all potentially enhanced by unloading-loading mechanisms [34][35][36][37][38] . As a first order assumption, the fracture orientation is compatible with a NW-SE trending σ 3 stress, indicating that these fractures might have formed in a NW-SE extensional or transtensional stress field from 50 Ma, similar to the inferred stress regime offshore from the Eocene to present 39 .  (Figs. 3, 4) (Fig. 4). The general δ 18 O trend of getting heavier with younger ages (Fig. 3b), might indicate precipitation of calcite from fluids of a similar source under decreasing temperatures 40 , possibly indicating decreasing depth of crystallisation. The variably negative δ 13 C values point to the groundwater being influenced by dissolved HCO 3 from the oxidation of overlying organic material 43 , whereas the slightly positive δ 13 C values of the coastal samples VAH_192_B and VAH_96 (Fig. 1), point to a lack of overlying organic material and a possible influence of sea water.

Conclusions
The first U-Pb calcite ages from fracture fills in basement rocks of western Norway demonstrate the potential of U-Pb calcite geochronology to reveal both large-scale regional and more subtle tectonic events along deeply eroded and uplifted continental margins. In particular, our data (1) reveal the potential influence of the proto-Icelandic plume in the region, leading to extensive fracturing with variable strike orientation, and (2) highlight the importance of ongoing post-break up tectonic activity throughout the entire Cenozoic. Our new U-Pb data combined with structural and isotopic data provide evidence that much larger crustal tracts of continental crust have been affected by far-field tectonic stresses than previously anticipated. These data are of prime importance for the understanding of rifted continental margins comprised of old basement, where the lack of younger sedimentary cover otherwise hampers constraining the regional tectonic events.

Methods
Laser ablation ICP-MS U-Pb calcite dating. Polished rock slabs in 25 mm diameter epoxy mounts were analysed for characteristic major and trace elements and for U and Pb isotopes using either a mapping approach or a spot ablation sampling strategy. Analyses were performed at the Department of Geology at Trinity College, Dublin using a Photon Machines Analyte Excite 193 nm ArF excimer laser ablation system coupled to an Agilent 7900 quadrupole ICP-MS.
We followed the general analytical and data processing routine for image-based U-Pb geochronology 44 and specific details are given below in Supplementary Material 1 Table S1. In brief, laser sampling employed ablation of successive linear rasters that were compiled into element, elemental ratio and isotope ratio maps whereby one pixel represents one time-slice of the time-resolved signal. Characteristic major, minor and trace elements were measured along with U and Pb isotopes. Filtering of the data associated with the pixels in the maps was undertaken by applying a combination of specific geochemical criteria and/or manually drawn regions of interest to separate pixels from chemically and texturally different domains. The selected pixels were then pooled into 'pseudo-analyses' by using an empirical cumulative distribution function (ECDF) of a suitable channel ( 238 U/ 208 Pb or 207 Pb/ 235 U) and plotted on isochron diagrams. Details on the selection criteria and regions of interest are provided with the sample portraits of Supplementary material 1.
Spot analysis experiments were conducted in the conventional way. Downhole fractionation was negligible due to the large spot diameter and associated large width to depth ratio. Specific operating conditions and details on data processing are given in Supplementary Material 1 Table S2.
We use recommended uncertainty propagation 45 with modifications 46 . The first uncertainty quoted in the tables (Supplementary Material 1) is a session-wide estimate including the data point uncertainty, uncertainty on weighted means of primary reference material ratios and their excess scatter. The second uncertainty additionally includes systematic uncertainties such as the uncertainty on the reference age of WC-1, uncertainty on the 238 U decay constant and a laboratory-specific long-term reproducibility based on the results of the QC material (2%).
Ages were calculated using Isoplot 4.15. All U-Pb ages are from unanchored model 1 regressions in Tera Wasserburg (TW) plots (except VAH_130: unanchored regression in 86-TW space 47 ). Uncertainties quoted in the main text and figures are 2σ and include systematic uncertainties.  Fig. 1d) and compared to data from the literature from similar geological settings.