Fracture network heterogeneity in continental lithosphere: An example from the Lewisian Gneiss Complex, NW Scotland

Pre-existing ductile and brittle structures in continental lithosphere are thought to influence the orientation and spatial characteristics of subsequent deformation sets. By using quantitative techniques to describe fracture attributes it should be possible to determine whether these pre-existing weaknesses affect fracture network development. To test this hypothesis, the mainland Lewisian Gneiss Complex (LGC), in NW Scotland is being used as an example. The Late Archaean - Early Proterozoic LGC comprises TTG gneisses, mafic and ultramafic dykes and meta-volcanic and meta-sedimentary sequences that were accreted as a series of terranes during the Precambrian. Three regional fracture trends are recognised (from oldest to youngest); (1) steeply-dipping NW-SE Paleoproterozoic faults (mainly sinistral oblique) that are preferentially developed as foliation-parallel structures in pre-existing ductile shear zones (2) N-S to ENE-WSW trending, hematite stained normal fault ‘ladder fractures’ associated with the deposition of the overlying Neoproterozoic (1.2 Ga.) Stoer Group rift-related sediments. (3) NE-SW trending younger (likely Mesozoic) faults. Each fault set is associated with characteristic fault rock and mineral assemblages. The present project focuses on characterising these fracture sets on the regional to outcrop scale, using a variety of remote and fieldwork analysis techniques. Regional data comprises 2D lineament maps created from high resolution NEXTMap® digital elevation models. Outcrop data consists of 1D sample lines and 2D photo-mosaics which have allowed fracture attribute characterisations to be made. This project centres on the Assynt and Rhiconich terranes to assess the heterogeneity in fracture networks due to variations in lithology and metamorphic grade. Results from statistical analysis of outcrop and regional orientation data show that the tonalitic and granulite-facies Assynt terrane shows a correlation between intense foliation and preferential development of fracture sets. Fieldwork confirms these results. For example, the Canisp Shear Zone with NW-SE trending foliation exhibits NW-SE trending faults with a regularity of 10-20cm along the foliation planes. Outside this shear zone similarly trending faults have a spacing of 10’s to 100’s of metres. ’Ladder fractures in the Stoer Group preferentially occur within these shear zones although they cross-cut the foliation. In contrast, there is no correlation in the granodioritic, amphibolite-facies Rhiconich terrane, where, fractures do not reactivate pre-existing fabrics and instead cut across the foliation, even within shear zones. In conclusion, the heterogeneity of the fracture sets seen within the mainland LGC can be attributed to variations in lithology and changes in pre-existing ductile and brittle basement structure, but only the NW-SE fracture set, seen most commonly in the Assynt terrane, can be directly linked to reactivation of pre-existing ductile fabrics.