Rapid subsidence and stacked Gilbert-type fan deltas, Pliocene Loreto basin, Baja California Sur, Mexico

Abstract

Pliocene nonmarine to marine sedimentary rocks exposed in the Loreto basin, Baja California Sur, provide a record of syntectonic subsidence and sedimentation in a transform-rift basin that developed along the western margin of the Gulf of California. A thick sequence of twelve Gilbert-type fan deltas, having a total measured thickness of about 615 m, accumulated near the fault-bounded southwestern margin of this basin. Based on stratal geometries and lithofacies associations, sedimentary rocks are divided into Gilbert-delta topset, foreset and bottomset strata, shell beds and background shallow-marine shelf deposits. Topset strata of each Gilbert-type delta cycle are capped by laterally persistent molluscan shell beds containing diverse assemblages of bivalves, pectens, oysters, gastropods and echinoids. These shell beds are interpreted to be condensed intervals that record sediment starvation during abandonment of the fan-delta plain. Delta abandonment may have been caused by large episodic faulting events, which submerged each pre-existing fan-delta plain, substantially slowed detrital input by drowning of alluvial feeder channels, and created new accommodation space for each new Gilbert-type fan delta. Alternatively, it is possible that delta-plain abandonment was caused by upstream avulsions and autocyclic lateral switching of fan-delta lobes during relatively uniform rates of slip along the basin-bounding fault.Two contrasting, plausible basin models are proposed for the Loreto basin: (1) asymmetric subsidence along a high-angle oblique-slip normal fault, producing a classic half-graben basin geometry with vertically stacked Gilbert-type fan deltas; or (2) lateral stacking and horizontal displacement of strata away from a relatively fixed depocenter due to fault movement in the releasing bend of a listric strike-slip fault. We favor the first model because field relations and simple geometric constraints suggest that most of the total measured section represents a true vertical stratigraphic profile. Assuming vertical sediment accumulation and using ages of interbedded tuffs obtained from high-precision 40Ar/ 39Ar dating of plagioclase and biotite, quantitative decompaction and geohistory analysis was carried out for the Loreto basin sequence. Tuff ages range from 2.61 ± 0.01 Ma in the lower part of the basinal sequence to 1.97 ± 0.02 Ma near the top, with two intermediate tuffs dated at 2.46 ± 0.06 and 2.36 ± 0.02 Ma that are separated by 782 m of measured section. Basin subsidence initially took place at moderate rates of 0.43 ± 0.17 mm/yr and accelerated dramatically at 2.46 Ma to 8.1 ± 5.1 mm/yr. This phase of extremely rapid subsidence lasted for only about 100 ka, and it produced much of the total accomodation space and sedimentary thickness in the basin. Accumulation of Gilbert-type fan deltas took place only during the short pulse of very rapid subsidence, between 2.46 and 2.36 Ma. Prior to this time interval, alluvial-fan and shelf-type fan-delta depositional systems prevailed; afterwards no fan deltas of any kind were deposited, and the basin evolved to a slowly subsiding low-energy carbonate shelf setting. This suggests that very rapid subsidence, combined with rapid sediment input, may be required to maintain steep basin-margin slopes and continually create new accommodation space, conditions that seem necessary for the development of thick sequences of stacked Gilbert-type fan deltas.