Provenance from the geochemical composition of terrestrial clastic deposits — A review with case study from the intracontinental Permo-Triassic of European Pangea

How can inorganic geochemical data from terrestrial deposits be used as a provenance tool? We demonstrate a new approach as earlier studies have focused on marine deposits. We present a review of the factors that affect the whole-rock geochemical composition of terrestrial deposits and a case study to test how geochemical provenance differentiation can be made for such deposits. We confirm that both K₂O/Na₂O and immobile trace-element ratios are suitable provenance indicators for terrestrial deposits. We stress that previously developed discrimination schemes constructed for marine deposits of different depositional facies, specific grain size and texture (commonly greywacke) for identification of tectonic settings are unsuitable for terrestrial deposits as they often are related to different tectonic frames, depositional facies, and sorting degrees. Different terrestrial tectonic situations will influence the detrital composition such that rift, foreland, and sag basins likely have different chemical compositions. Humid climatic conditions may reduce such differences. The case study is based on the intracontinental Permian Rotliegend Group and Early Triassic Buntsandstein Group and equivalents in central Europe, mainly Germany and Denmark. Both units had mainly felsic sources. We present results from >1500 sedimentary samples from both published and unpublished data, mostly from sandstone. K₂O/Na₂O efficiently differentiates sandstone from different regions. The compositional variation is in line with reported variations in feldspar compositions and indicates local catchments. Thus, the feldspar has been preserved in the arid climate due to limited weathering. Y/Co, in combination with other indices (La/Sc, La/Lu, Ti/Nb, Th/Sc), is one of the most successful mafic–felsic trace-element indices that we use. We postulate that the trace-element variation is governed by mineral inclusions mainly in feldspar in addition to individual detrital grains. V/Zr and Zr/Sc mostly correlate with Y/Co, which indicates different zircon affinity in the source rocks rather than sorting. More dominance of felsic compositions and higher Zr/Sc for Triassic than Permian strata are in line with more sedimentary recycling during the Triassic than Permian. Thus, the results reveal that source-rock differentiation can be made for different geographical areas and stratigraphic units based on K₂O/Na₂O due to variations in feldspar compositions in the source areas, provided little chemical weathering. Based on the case study, the mafic–felsic trace-element ratios La/Sc, Th/Sc, Y/Co, and Ti/Nb seem most promising for provenance discrimination, in combination with La/Lu, and Zr/Sc and V/Zr as tracers of heavy minerals. We conclude that the combination of arid climatic conditions and terrestrial depositional environments makes it possible to use both major oxides and trace element ratios to reveal source-rock differences. For robust provenance interpretations, we recommend to focus on immobile elements that are little affected by grain-size variations, to combine several element ratios for interpretation, to avoid comparison of data produced by different methods (for instance x-ray fluorescence and mass spectrometry), and to complement geochemical investigations with data from other methods.