Dilution and propagation of provenance trends in sand and mud: Geochemistry and detrital zircon geochronology of modern sediment from central California (U.S.A.)

Abstract

Integrated, multi-method provenance studies of siliciclastic sedimentary deposits are increasingly used to reconstruct the history of source-to-sink transport, paleogeography, and tectonics. Invariably, analysis of large-scale depositional systems must confront issues regarding how to best sample the system and adequately cope with the details of sediment mixing. Potential biases including variations in grain size, sediment flux, and mineral concentration may cause provenance tracking tools to misrepresent the contributions of source-areas that contribute to large drainage networks. We have acquired U-Pb detrital zircon data from modern sand and whole-rock geochemistry from mud sampled from the Sacramento-San Joaquin drainage of central California to elucidate conditions that can skew provenance trends along the course of a major river system. This drainage network is fed by headwaters that tap the Mesozoic pluton-dominated southern Sierra Nevada, the Paleozoic-Mesozoic wallrock and volcanic-dominated northern Sierra Nevada, the ultramafic-dominated eastern Klamath Mountains, and the intermediate to mafic Cascades volcanic arc. Analysis of the results indicates that detrital zircon provenance trends effectively record source variations for the southern, granite-dominated portion of the drainage network where contrasts in lithology and inferred zircon fertility are relatively minor. In these circumstances, mixture modeling of U-Pb detrital zircon data calibrated with a measure of zircon fertility approximates relative sediment flux contributed by individual drainages. Alternatively, in the northern parts of the system, source regions underlain by ultramafic and/or volcanic rocks are poorly represented, or entirely missing, in downstream detrital zircon records. In some cases, mud geochemistry data more faithfully represent sediment provenance trends. Sampling performed at the confluence of the Sacramento, American, Mokelumne, and San Joaquin Rivers within the Sacramento-San Joaquin Delta region yields a detrital zircon age distribution that is indistinguishable from that of an independently established database of Sierra Nevada batholith crystallization ages. However, when the combined river flows along a recently established passage to the Pacific through the San Francisco Bay region, dredged sediment is found to be significantly contaminated by locally eroded material from the Franciscan Complex and other rocks that crop out within the Coast Ranges. Large variation of Zr concentrations measured throughout the Bay area document that significant hydrodynamic fractionation impacts sediment delivery through this segment of the system. The more Sierra Nevada-like detrital zircon age distribution yielded by a piston-core sample from the continental slope may be explained by either early-stage unroofing of the Coast Ranges or more efficient sand delivery from the delta to the Pacific by a free-flowing river driven by a low stand in sea level.