Large-scale stratigraphic architecture and sequence analysis of an early Pleistocene submarine canyon fill, Monte Ascensione succession (Peri-Adriatic basin, eastern central Italy)

The Monte Ascensione succession (c. 2.65-2.1 Ma) is a well-exposed example of an exhumed submarine canyon fill embedded within slope hemipelagic mudstones. This gorge represented a long-lasting pathway for sediment transport and deposition and during the Gelasian delivered Apennine-derived clastic sediment to the adjacent Peri-Adriatic basin. A total of six principal lithofacies types, representing both canyon-confining hemipelagic deposits and canyon-filling turbidity current and mass-transport deposits, can be delineated in the studied sedimentary succession. The canyon-fill deposits display a marked cyclic character and the component lithofacies succeed one another to form at least fifteen fining-upward stratal units, which are interpreted to represent high-frequency, unconformity-bounded depositional sequences. Variability in the vertical repetition of constituent lithofacies allows the identification of three basic styles of sequence architecture that can be interpreted in terms of differing positions along a conceptual down-canyon depositional profile. An integrated chronology, based on biostratigraphic data and on palaeomagnetic polarity measurements, strongly supports a one-to-one correlation between the sequence-bounding surfaces and oxygen isotope stages G2-78, suggesting that the most feasible sequence-engendering mechanism is that of orbitally dictated glacio-eustatic changes in sea level, which regulated timing of sediment storage on the shelf and its redistribution beyond the shelf edge. One of the most significant aspects of this study is the demonstration that processes occurring within upper slope canyons can be expected to be strongly influenced by variations in sea level; that is, the erosional and depositional features evident in these deposits can be strongly controlled by allocyclic processes rather than autocyclic or random processes.