Petrography and geochemistry of feldspathic and mafic sediments of the northeastern Pacific margin

ABSTRACT This study attempts to ascertain whether the differing provenance sources and plate tectonic settings of deposition of clastic sediments and rocks can be identified by chemical means, thus opening the increased use of these rocks and their metamorphosed derivatives in plate tectonic modelling. Mineralogically immature feldspathic and mafic sands, muds, sandstones and shales from Mesozoic, Cenozoic and Holocene clastic deposits in California, Oregon, Alaska and Colorado have been both modally and chemically analysed providing a valuable data base (217 samples; 216 samples chemically analysed, many for 28 elements). There is significant upward chemical variation in the Great Valley sequence of California which mimics previously described petrographic variation and in turn reflects provenance changes with igneous episodes and erosional stripping of the Sierra Nevada in late Mesozoic time. Differing sandstone petrofacies result in varying chemical signatures and while the provenances of monomict sediments are easiest to identify, polymict sources involving granitoid or ophiolitic material can often be identified by potassium feldspar or Cr contents. The distinction of K which is derived from detrital potassium feldspar from K derived from detrital illite, micas or other sheet minerals, is best made by a Niggli al–alk plot against k . Mafic sandstones derived from mafic volcanic or plutonic rocks preserve essentially mafic igneous chemistry and could be possible parent sources of some amphibolites which grade into metasediments. The lithified erosion products of the Sierra Nevada calc-alkaline igneous rocks have higher Niggli al–alk and higher average Niggli si at any given mg value than the fresh igneous rocks enabling meta-arkoses to be distinguished from meta-igneous rocks. Applying previously published chemical criteria gives the actual plate tectonic setting of deposition of most of the sandstones studied. This suggests that the chemical composition of sandstones can yield much more information about the provenance and plate tectonic setting than hitherto recognised.