The concept of turbidite has evolved so much since its original definition by Kuenen and Migliorini in 1950 – i.e. the deposit of turbidity currents exemplified by the sandy flysch successions of the Northern Apennines – that it is now used to define a variety of deposits, some of which have little in common with sandy flysch formations in terms of facies, geometry and geological significance. The extension of the concept to other geodynamic settings and deposits of non-siliciclastic composition is considered only briefly in the concluding sections.
With the diffusion of the concept of turbidity current, in the 1950s and early 1960s, an entirely new branch of sedimentology came into being, concerned with the inventory of sedimentary structures, palaeocurrent measurements and bedding patterns. The most representative expression of this branch came from the ‘Dutch school’ of Philip H. Kuenen and his students. Between the late 1960s and the mid-1970s, there was a new development: facies analysis, in terms of modern environments and depositional systems. This development led to the introduction and discussion of ‘fan models’ that became an increasingly thorny issue with the accumulation of data from modern deep-marine settings. In particular, most researchers emphasized the importance of channel and lobe elements and their mutual relationships in space and time. These models may differ in terms of specific features, e.g. canyon-fed versus delta-fed ramp settings and terminology, but the basic distinction between channels (sediment pathways), lobes and basin plains (sheet-like depositional features) was and still is widely retained – a model that simply refers to a system where a distributary channel passes downstream to a depositional zone, like in most fluvio-deltaic systems. Great caution should, however, be exercised when comparing modern and ancient fans – a problem discussed at length in the Committee on Submarine Fans I convened by A.H. Bouma and held in Pittsburgh in 1982. Different data sets and geological contexts, scaling problems and terminology still cast doubt over how meaningful such a comparison may be. Despite the many problems encountered, the elemental approach provides an easy, essentially descriptive tool to significantly compare recent with ancient, recent with recent, and ancient with ancient systems.
Beginning in the 1970s, process-oriented facies analysis led to increasingly complex facies classification schemes, which showed substantial departures from the classic Bouma sequence and introduced many new concepts: proximal versus distal sedimentation, sediment bypass and flow efficiency, in addition to deflection, reflection and ponding of turbidity currents in confined basins. During the last two decades, there has been an increased interest in attempting to interpret the incredibly detailed submarine landscapes obtained through advances in marine geology, technology and high-resolution three-dimensional seismic data provided by the oil industry. Outcrop ‘analogues’ derived from orogenic belts are used commonly to improve the interpretation of seismic-reflection facies, although their actual value may be questioned in many cases.
Seismic–stratigraphic concepts are used routinely to describe and interpret turbidite systems of continental margin basins where cyclic sea-level variations are thought to be essentially controlled by eustasy. These concepts are difficult to apply to flysch basins, where the tectonic control on the development of cycles of relative sea-level variations appears to be dominant. In particular, the huge volumes of sediment involved in the infill of flysch basins imply amounts of uplift of the source areas and subsidence of the receiving basins that clearly outstrip those of divergent continental margins controlled by eustasy and thermal subsidence. Cycles of tectonic uplift and denudation (Davisian-type cycles in the sense of Mutti et al., 1996) apparently play a major role here.
Most recent attempts to understand turbidite deposition are related to the increased economic importance of turbidite sandbodies as hydrocarbon reservoirs in many offshore basins (e.g. Gulf of Mexico, West Africa, Brazil, the North Sea). The many problems inherent to this situation have been reviewed extensively in a workshop held in Parma in 2002; only some of these problems are reconsidered briefly in this paper. Sandy turbidite systems can be generated by the resedimentation of deltaic deposits through submarine slides or be derived directly from flood-generated hyperpycnal flows; in the latter case, climatic variations must have played a fundamental role in controlling flood frequency and magnitude with time. Recognizing these two different types of system is not always easy and requires a good understanding of the geological context of the basin under consideration and particularly of the role of marginal fluvio-deltaic systems from which turbidites are ultimately derived. Unfortunately, this kind of integrated analysis is still in its infancy. There are other types of turbidite deposits, such as the calcareous flysch of the Western Alps and the Northern Apennines, whose origin still remains a matter of debate in terms of sediment source and triggering mechanisms of large-volume turbidity currents essentially loaded with fine-grained biogenic sediment. Some authors have referred to these sediments either as ‘megaturbidites’ or ‘seismoturbidites’.
The importance of tectonic control and geodynamic setting is stressed for turbidite systems of orogenic belt basins, which is justified both by historical reasons (turbidites were from their recognition included in the definition of flysch) and recent studies of thrust belts. The time is now ripe for reconsidering these sediments within a broader framework that takes into account the enormous quantity of data and concepts that have been developed in the last 50 years; this in itself raises a problem, and no small one: the accuracy and quality of data collected in the field and the training of young scientists. How many field geologists are being produced in these times of increasingly computerized geology; and how good are they?