The San Gabriel fault, a deeply eroded late Oligocene to middle Pliocene precursor to the San Andreas, was chosen for petrologic study to provide information regarding intrafault material representative of deeper crustal levels. Cataclastic rocks exposed along the present trace of the San Andreas in this area are exclusively a variety of fault gouge that is essentially a rock flour with a quartz, feldspar, biotite, chlorite, amphibole, epidote, and Fe-Ti oxide mineralogy representing the milled-down equivalent of the original rock (Anderson and Osborne, 1979; Anderson et al., 1980). Likewise, fault gouge and associated breccia are common along the San Gabriel fault, but only where the zone of cataclasis is several tens of meters wide. At several localities, the zone is extremely narrow (several centimeters), and the cataclastic rock type is cataclasite, a dark, aphanitic, and highly comminuted and indurated rock. The cataclastic rocks along the San Gabriel fault exhibit more comminution than that observed for gouge along the San Andreas. The average grain diameter for the San Andreas gouge ranges from 0.01 to 0.06 mm. For the San Gabriel cataclastic rocks, it ranges from 0.0001 to 0.007 mm. Whereas the San Andreas gouge remains particulate to the smallest grain-size, the ultra-fine grain matrix of the San Gabriel cataclasite is composed of a mosaic of equidimensional, interlocking grains. The cataclastic rocks along the San Gabriel fault also show more mineralogiec changes compared to gouge from the San Andreas fault. At the expense of biotite, amphibole, and feldspar, there is some growth of new albite, chlorite, sericite, laumontite, analcime, mordenite (?), and calcite. The highest grade of metamorphism is laumontite-chlorite zone (zeolite facies). Mineral assemblages and constrained uplift rates allow temperature and depth estimates of 200 ± 30° C and 2–5 km, thus suggesting an approximate geothermal gradient of ~50°C/km. Such elevated temperatures imply a moderate to high stress regime for the San Andreas, which is consistent with experimental rock failure studies. Moreover, these results suggest that the previously observed lack of heat flow coaxial with the fault zone may be the result of dissipation rather than low stress. Much of the mineralogy of the cataclastic rocks is still relict from the earlier igneous or metamorphic history of the protolith; porphyroclasts, even in the most deformed rocks, consist of relict plagioclase (oligoclase to andesine), alkali feldspar, quartz, biotite, amphibole, epidote, allanite, and Fe-Ti oxides (ilmenite and magnetite). We have found no significant development of any clay minerals (illite, kaolinite, or montmorillonite). For many sites, the compositions of these minerals directly correspond to the mineral compositions in rock types on one or both sides of the fault. Whole rock major and trace element chemistry coupled with mineral compositions show that mixing within the zone of cataclasis is not uniform, and that originally micaceous foliated, or physically more heterogeneous rock units may contribute a disproportionally large amount to the resultant intrafault material. As previously found for the gouge along the San Andreas, chemical mobility is not a major factor in the formation of cataclastic rocks of the San Gabriel fault. We see only minor changes for Si and alkalies; however, there is a marked mobility of Li, which is a probable result of the alteration and formation of new mica minerals. The gouge of the San Andreas and San Gabriel faults probably formed by cataclastic flow. There is some indication, presently not well constrained, that the fine-grained matrix of the cataclasite of from the San Gabriel fault formed in response to superplastic flow.
Read full abstract