Abstract
Sequence stratigraphy arose as a paradigm in stratigraphy after the integration of the descriptive seismic method, introduced by Exxon researchers in the 1970s, with genetic concepts linking seismic attributes to sedimentary dynamics. Since then, the sequence stratigraphy model underwent significant modification owing to the increasing scenarios of application, each with its own practical requirements. This led to the fragmentation of the original model into a plethora of sub-methods and the flourishing of redundant notions and terminology. Reviewers striving to preserve the unity and fitness of sequence stratigraphy systematically weakened the central assumption by which it stood as a novelty and a paradigm: the relevance of relative sea-level cycles in shaping strata ‘sequentially’. In contrast to this attitude, the value of a model explicitly relying on the upholding control of sea-level is herein reconsidered, based on the marine record of Quaternary, climate-driven, sea-level cycles. Traditionally conceived as exceptionally short-lived and extraordinary events in the history of the Earth, these cycles are documented worldwide on modern continental margins, providing convincing evidence of how sea-level fluctuations actually shape sequences.
Stratigraphy has played a key role in major conceptual revolutions, proving highly influential in determining the growth of geological thought. In the 1970s, stratigraphy underwent its own revolution with the emergence of sequence stratigraphy, a paradigm made possible by the advent of seismic stratigraphic methods (Payton 1977). The sequence stratigraphy paradigm relied on the principle that relative sea-level exerted the main control on the architecture of continental margins, thus forming the stratal patterns observed at the scale of seismic profiles. More specifically, the sequence stratigraphy paradigm stated that continental margins displayed a general motif made of basic units called depositional sequences, as the result of cyclical sea-level changes (Mitchum et al. 1977). The cyclical pattern of sea-level was reflected by the existence of specific units (systems tracts) at the sub-sequence level, each with distinctive stratal features recording a particular trend (lowstand, rising, highstand and falling) of the sea-level cycle (Vail et al. 1977a; Vail 1987). Following the rapid diffusion of sequence stratigraphy, a synthesis was proposed, focusing on the application of the model in different contexts and by different investigation methods and data (Wilgus et al. 1988). This revised model included conceptual schemes linking stratal patterns to rates of relative sea-level variation and sediment supply (Fig. 1).
Basic sequence model modified from the original of Vail (1987) and later reviews (Wilgus et al. 1988). Estimates of sea-level amplitude (i.e. <100 m) and cycle duration (i.e. >1 – 2 myr) were not part of the original scheme. Units with distinct geometric pattern (irrespective of depositional environment and sedimentological facies) are systems tracts genetically related to specific phases of the sea-level curve (1 – 5). The following features should be noted: shelf bypassing and base of slope deposition during sea-level fall (1); margin construction during lowstand (2) and highstand (4); reduced lowstand progradation (5) and enhanced highstand preservation during lower amplitude sea-level falls (4).
The interpretative key of sequence stratigraphy, however, soon revealed inconsistencies with the current view of field sedimentologists, based on which environmental factors and tectonics, rather than global sea-level, controlled sedimentary dynamics. Moreover, tectonics had been long accepted as the governing force of regional cyclical patterns (i.e. ‘cyclothems’) in the stratigraphic record (Sloss 1963; Fischer 1964). Conflicts between these opposite views entailed both the supposed originality and the greater efficacy of sequence stratigraphy compared with methods already in use (Dott 1996). These fundamentally opposite views condensed in the controversy between ‘allocyclic’ and ‘autocyclic’ factors as having the foremost control on strata formation (Schlager 1993). As criticism heightened, eustasy and relative sea-level were progressively replaced by concepts of less specific connotation, such as ‘base level’ and ‘accommodation space’; this approach set the tendency towards depriving sea-level of any primary importance in controlling the origin and architecture of sequences (Van Wagoner 1995). As a result of this process, the latest versions of sequence stratigraphy exclude explicit reference to sea-level as a prime agent in shaping sequences (Catuneanu et al. 2009; Neal & Abreu 2009).
It is herein argued that the development of sequence stratigraphy as a novel, important approach relied on its standing as a new paradigm (Miall & Miall 2001; Catuneanu 2006), centred on the assumption that sea-level cycles have exerted a first-hand control on the sequence architecture of continental margins. By rejecting relative sea-level as a general and primary mechanism, sequence stratigraphy no longer stands as a breakthrough or paradigm in stratigraphy. In contrast to this approach, a sequence stratigraphy model relying on the prominent control of relative sea-level can be refined based on the extensive marine record of Quaternary sequences. Although largely ignored during major reviews of sequence stratigraphy (Van Wagoner 1995; Catuneanu et al. 2009), Quaternary sequences distinctively reveal the ‘sequential’ architecture imparted by reiterated sea-level cycles of known shape, duration and amplitude (Lobo & Ridente 2014). Quaternary sequences should not therefore be regarded as ‘special cases’, but as exemplars of the dominant role of relative sea-level change in determining the form of stratigraphic architectures.
Background and overview
The inborn link with sea-level: a methodological bias
Despite claims to the contrary (Van Wagoner 1995), the sequence stratigraphy model, in its original form and early assessments, was essentially rooted on a twofold postulation dealing with sea-level: (1) the origin of depositional sequences was thought to be controlled primarily by sea-level cycles (Mitchum et al. 1977); (2) sea-level cycles were claimed to be eustatically driven, so that sequences from different continental margins would reflect the same global sea-level curve (Vail et al. 1977b). Severe criticism was received on the second proposal (Miall 1986, 1991, 1992), which implied that sequences could be correlated globally on the basis of one global sea-level curve extending throughout the Meso-Cenozoic (Haq et al. 1987). However, this controversial argument did not detract from the interrelated assumption that, in a context of cyclical sea-level change, deposits would stack as predicted by the sequence stratigraphy model (Carter et al. 1991, 1998).
The genetic bond of sequence stratigraphy with sea-level, which is critical to its status as a general model, also has profound implications on its applicability by different approaches and in different contexts. By its birth as a seismic method in marine settings, sequence stratigraphy provided a powerful model for defining basin-scale stratigraphy rather than processes at the scale of single depositional systems (Brown & Fischer 1977). Retrospectively, it can be argued that the conceptual and methodological imprint from seismic stratigraphy has biased sequence stratigraphy as a model of patterns of depositional architecture in clastic sediments, accumulating under the control of changes in accommodation modulated by relative sea-level cycles (Posamentier et al. 1988). Therefore, the efficacy of sequence stratigraphy as a model may not be irrespective of scale and settings, as well as of the methods employed to exploit them. This methodological bias has been at the heart of the early controversy between field sedimentologists and the pioneers of sequence stratigraphy. Questions arose on the limitations of seismic/sequence stratigraphy in resolving details at the scale of facies analysis, and the focus of the debate long remained on whether sequence stratigraphy was conceptually wrong compared with traditional methods, rather than complementary and inclined towards larger scale stratigraphic relationships.
The new debate and the ultimate sequence stratigraphy: evolution or involution?
In recent time, the debate on sequence stratigraphy has changed its skin, being no longer a matter between ‘sequence stratigraphers’ and their opponents, but rather drawing from the many contrasting interpretations emerged from its wide application. According to Catuneanu et al. (2009), the blooming of different ‘schools’ within sequence stratigraphy reflects methodological constraints rather than conceptual inconsistencies; hence, many of the controversies can be resolved by emphasizing the complementary aspects among the different working approaches. This is the premise for a new synthesis of sequence stratigraphy and the standardization of a general scheme defined as ‘model independent’ (Catuneanu et al. 2009). ‘Model independent’ implies that concepts underpinning this advanced version of sequence stratigraphy need not be rooted in any assumption-based paradigm (i.e. model); but rather they should provide a suite of practical guidelines for defining stratigraphic units in the field or on the basis of geophysical data.
The model-independent approach is well synthesized in the definition of a sequence as ‘a succession of strata deposited during a full cycle of change in accommodation or sediment supply’ (Catuneanu et al. 2009, p. 19). This definition, substantially different from the original (Mitchum et al. 1977), lays emphasis on the fact that sequence stratigraphy should be embedded in a concept of cyclicity unbiased by the mechanism of the cycle, and mostly relying on changes in accommodation and sedimentation rates. Equating any product of changes in accommodation or supply to a ‘sequence’ eliminates conflicts of all sorts in defining, recognizing and delimiting sequences; however, it also implies the dismissal of any general explanation as to what, ultimately, causes depositional cycles. Instead, because of the multiple sedimentary processes that may control depositional cycles and stratal patterns (Burgess & Prince 2015), a general model should provide criteria for reducing ambiguity when discriminating among diverse processes and controlling factors.
It has been long acknowledged that processes at the scale of depositional systems may generate cyclical patterns, a phenomenon known as autocyclic control (Beerbower 1964). Products of autocyclicity frequently result from a ‘change in accommodation or sediment supply’ (Muto & Steel 2002; Muto et al. 2007), therefore conforming to the definition of ‘sequence’ proposed by Catuneanu et al. (2009). However, the early model of sequence stratigraphy, and the related concept of ‘sequence’, envisages basin-wide exposure and flooding of shelf to upper slope settings (Fig. 1), with regional-scale stratigraphic effects that are in contrast to local effects typical of autocyclic processes (Cecil 2013). Basin-scale events of the type envisaged in the early model of sequence stratigraphy are consistent with allocyclicity in the form of relative sea-level cycles resulting from the compound effect of eustasy, tectonic uplift and subsidence (Vail et al. 1977a; Jervey 1988). In this view, most autocyclic changes in accommodation and supply probably produce units that are limited in lateral extent compared with those reflecting allocyclic control.
The supra-local influence of allocycles provides valuable constraints for defining sequences based on their spatial scale (i.e. areal extent), independently of thickness or temporal attributes. However, it also poses the problem of the identification and lateral correlation of key stratigraphic surfaces and units, to discern auto- and allocycles before defining sequences and applying sequence stratigraphy concepts.
A new perspective for an old assumption
Quaternary cycles and ‘sequential architecture’
According to the founders of sequence stratigraphy, the ‘sequential’ arrangement of progradational–aggradational–retrogradational patterns unveiled by sequences was governed by relative sea-level, with a secondary role from supply dynamics (Kendall & Lerche 1988; Posamentier & Vail 1988; Posamentier et al. 1988). To restore the nexus between this original assumption and any revised version of sequence stratigraphy, evidence should be found in the stratigraphic record of patterns of ‘sequential architecture’ that can be explained on the basis of wide-ranging, relative sea-level cycles (i.e. allocycles). Quaternary stratigraphies from marine settings provide worldwide evidence of such basin-scale architectural patterns controlled by sea-level cycles and environmental processes (Lobo & Ridente 2014).
Quaternary sea-level cycles reflect Milankovitch climate cycles with periods of c. 20, 40 and 100 kyr (Hays et al. 1976; Schwarzacher 2000). These climate cycles are uneven, with abrupt warming pulses and ice cap melting, followed by gradual cooling promoting slow ice cap growth (Ruddiman 2003). Sea-level cycles are thus asymmetric, with high-amplitude (>100 m in the case of 100 kyr cycles) rapid sea-level rises and slow sea-level falls (Lobo & Ridente 2014). Depositional sequences reflect this asymmetry in their overall shape and internal architecture, which is essentially formed by progradational shelf deposits (tens of metres thick) of the highstand (HST), falling (FST) and lowstand (LST) phase (Fig. 2); the rapid sea-level rise (TST) is generally recorded by thinner, aggradational distal drapes (Fig. 2b and c) and reworked deposits forming patchy clinoforms or channel fill deposits (Fig. 3). A schematic synthesis of this general sequence-stratigraphic pattern, with possible variants, is simplified in Figure 4.
Seismic profiles (interpretation on the right) from different settings highlighting the stratigraphic architecture displayed by 100 kyr Quaternary sequences (modified after Lobo & Ridente 2014): (a) Gulf of Lions (original data from Jouet et al. 2008); (b) Bengal Shelf (original data from Hübscher & Spieß 2005); (c) Adriatic margin (original data from Ridente et al. 2008). Despite the very different and distant settings, a general pattern is observed, consisting of stacked ‘regressive sequences’ separated by prominent shelf-wide erosional unconformities (i.e. sequence boundaries). The following features should be noted: continuous deposition and preservation of deposits from highstand (HST) to lowstand (LST); dominance of overall self-similar regressive units of the falling stage (FST); variable pattern of lowstand deposition (LST), with reduced slope progradation in (a) and (b) compared with (c); overall draping pattern (b, c) or patchy preservation (a) of transgressive deposits (TST); sharp-based contact between regressive units (c) evidenced by very high-resolution seismic data, which marks the HST–FST and FST–LST transition (HST and FST units have been characterized based on borehole chronostratigraphic data discussed by Ridente et al. 2008, 2009).
Examples of coarse-grained, high-angle progradational units (TST) deposited during the rapid sea-level rise that typically drowns the shelf at the turn-around point between successive cycles: (a) Northern Tyrrhenian margin (modified after Ridente et al. 2012); (b) East China Sea (modified after Berné et al. 2002). These deposits lay directly on the main unconformity and are capped by an equally marked erosional (ravinement) surface. They often display a linear or patchy distribution, depending on the laterally variable availability of coarse sediment. SB1–SB6 and D115–D140 are sequence boundaries; U110–U125 are sequences.
(a) Basic model representing 100 kyr Quaternary sequences, synthesized based on examples from modern continental margins (Lobo & Ridente 2014). Noteworthy features are the dominance of progradational clinoforms and the preservation of HST (4) limited to its distal foresets, closely resembling those of FST units (5). (b) Ideal Quaternary margin recording multiple cycles each exemplifying variants of the basic model in (a). Sequences sa–sc stack laterally, indicating that no additional accommodation is created on the shelf during each cycle and deposits are preserved seaward of the previous shelf margin: this pattern is typical of some pre-Middle Pleistocene narrow shelf settings, or of uplifting margins. Sequences sd–sg stack vertically as on subsiding margins, where additional accommodation enhances preservation on a shelf that has attained significant extent throughout successive depositional cycles. The variability of LSTs should be noted (grey units): a–c, slope prograding lowstand; d, thin shelf-margin lowstand; e, shelf-perched lowstand delta; f, delta front prograding lowstand (similar to a and b); g, slope-confined ‘draping’ lowstand (similar to c).
In many examples from Quaternary margins, and in contrast to the model in Figure 1, sediment flux to the shelf does not decrease appreciably during the slow sea-level fall (Fig. 2); instead, there is evidence that sediment bypassing of the shelf actually occurs at the end of the sea-level fall (Ridente et al. 2012), determining reduced progradation during lowstand intervals (e.g. Fig. 5; compare with Fig. 2a and b). This determines the observed stratigraphic continuity and overall similarity between HST, FST and LST progradational units (e.g. Fig. 2; compare with Fig. 4). However, tectonic subsidence or uplift and sediment supply may vary significantly on Quaternary margins, resulting in a changeable shape of the relative sea-level curve and dissimilar exploitation of the accommodation, respectively. As a consequence, Quaternary sequences show considerable variability of their stratigraphic architecture, despite the overall dominance of the sea-level component (Fig. 4b); this variability mostly affects lowstand deposition, during which a complex balance exists between overall sediment flux, sedimentation rates on the shelf margin and bypassing of the shelf (Fig. 6; compare with Figs 2, 4b and 5).
Seismic and core stratigraphy from the Northern Tyrrhenian margin. Sparker seismic profile shows sequence architecture; chirp-sonar high-resolution profile along the same line shows detail of the cored interval (modified after Ridente et al. 2012). Progradational clinoforms within sequences DS1 and DS2 pass distally into upslope-thinning units (tu). Core Z-145 (151 m water depth) penetrated 6 m through the uppermost of these units, recovering a 4.11 m thick continuous interval. Seismic reflectors a–c mark the transition between sediment intervals A–D resulting from the alternation of mud (A and C) and sandy mud (B and D). Interval C largely corresponds to the uppermost upslope-thinning unit. Bio- and chronostratigraphic analysis, and also 14C age dating, indicate that the transition from lithofacies C to D (reflector c, corresponding to the distal conformity of erosion surface SB1) is dated between 21.3 and 22.5 ka BP. Consequently, the upslope-thinning unit (C) records an interval encompassing part of the Last Glacial Maximum lowstand and early post-glacial sea-level rise. It should be noted that lower sedimentation rates characterize the lowstand and the early sea-level rise (between c. 24 and 11 ka), whereas higher rates occur during the Late Holocene.
Stratigraphic variability of lowstand progradational deposits within 100 ka sequences (modified after Lobo & Ridente 2014: (a) low-angle shelf margin progradational unit thinning downslope (delta-influenced setting of the Gulf of Mexico; original data from Anderson et al. 2004); (b) evolution of lowstand deposition from shelf-margin high-angle progradational shoreface units (LSTsf) to slope-prograding prodeltaic unit (LSTpd) (Niger submarine delta; original data from Riboulot et al. 2012).
In the vicinity of a fluvio-deltaic source, sediment flux may be high enough to construct shelf-perched lowstand units (Fig. 4b, LST e; compare with Fig. 2a and LSTsf in Fig. 6b) typically consisting of coarse-grained deltaic deposits (e.g. Lobo et al. 2005; Rabineau et al. 2005; Jouet et al. 2008; Riboulot et al. 2012); or to promote large-scale shelf margin and slope progradation (Porębski & Steel 2003; Ridente et al. 2008; Fig. 4, LST f; compare with Fig. 2c and LSTpd in Fig. 6b), as predicted by the classical lowstand scheme of the Exxon model (Fig. 1). With decreasing influence of deltaic sources, slope progradation is hampered (Fig. 4b, LST g; compare with Fig. 5) and overall low-angle progradational LSTs form, in some cases closely resembling FSTs in terms of overall thickness, sedimentological composition and depositional geometries (Steckler et al. 2007; Ridente et al. 2009; Fig. 4b, LST d; compare with Fig. 6a).
Finally, in spite of theoretical models in which composite sequences stack in a ‘matrioska-like’ style (Mitchum & Van Wagoner 1991), the Quaternary record of composite cyclicity provides evidence of scale constraints (essentially duration versus amplitude of the sea-level excursion) that affect the preservation of sequence boundaries and sequences formed during cycles of different order (Lobo & Ridente 2014). In particular, because of the predominance of 100 kyr cycles during the past c. 800 kyr, 100 kyr sequences have largely obliterated the stratigraphic expression of 20 and 40 kyr cycles, which are instead typically represented during the Pliocene and Early Pleistocene (Naish & Kamp 1997; Carter et al. 1998; Massari et al. 1999; Kitamura et al. 2000; Pomar & Tropeano 2001).
The sequence boundary and other key surfaces
Despite redundant terminology and conceptualization of key stratigraphic surfaces in sequence stratigraphy (for a detailed review see Catuneanu et al. 2009), basically no more than three stratigraphic surfaces are required for delimiting sequences and systems tracts; these were all defined in early times (Plint 1988; Van Wagoner et al. 1988; Fig. 7a). In many Quaternary examples, key surfaces most difficult to identify (sedimentologically and/or seismically) are those marking the onset and the end of sea-level fall (e.g. Porębski & Steel 2003), which separate HST–FST and FST–LST, respectively (e.g. ‘sharp basal contact’ in Fig. 2). These surfaces, which develop progressively as wave base erosion shifts seawards during sea-level fall (Plint & Walker 1987; Plint 1988), attained a first formal definition as ‘regressive surface of marine erosion’ (Nummedal et al. 1993). An equivalent wave-base controlled marine erosion process occurs during sea-level rise and produces an erosion surface both on top of (ravinement surface) and below (transgressive surface) deposits of the TST (e.g. Fig. 3; compare with Fig. 7).
Principal bounding surfaces. (a) Original model (from Fig. 1): Eu, erosional unconformity; Ts, transgressive surface; Mfs, maximum flooding surface. The dashed rectangle outlines the Eu segment not subject to subaerial erosion (regressive surface of marine erosion, Rm). Rm bridges the true Eu with the correlative conformity (Cc) of the sequence boundary. Incised valley systems, if present, may contain sediment fill of the FST, LST or TST; in the last case the erosional surface atop is a ravinement surface (Rs), and the Ts correlates with the Eu at the base of the TST fill. (b) Example of problems posed by Quaternary sequences in identifying the LST from the late FST units and tracing the distal segment of the sequence boundary (from Fig. 4a). At least three interpretations are possible, which influence the choice of the distal correlative of the unconformity (Rm1, Rm2, Rm3), of the related LST (LST1, LST2, LST3) and of the correlative conformity (Cc1, Cc2, Cc3).
The identification of the regressive surface of marine erosion at the base of the LST is crucial for extending the sequence boundary seaward of the erosional unconformity, into the correlative conformity (Fig. 7). Although itself referred to as ‘correlative conformity’ (Catuneanu 2006; Catuneanu et al. 2009), the regressive surface of marine erosion at the base of LSTs retains erosional and/or geometric attributes for which the term ‘conformity’ is inappropriate. Hence, a distinction is proposed between the ‘distal correlative’ of the unconformity, which is part of the sequence boundary as a marine erosional surface (i.e. the regressive surface of marine erosion), and the ‘correlative conformity’ of the sequence boundary, which is its deep-water, non-erosional counterpart (Fig. 7).
In many instances the correlative conformity lacks any unambiguous physical continuity with the sequence boundary; indeed, on modern continental margins, shelf-perched LSTs display basal regressive surfaces limited to the shelf edge or abruptly terminating on the upper slope (Fig. 4b, LST d–e). As an exception, large river delta-front progradation may occur at the scale of the entire continental slope (e.g. Porębski & Steel 2003; Fig. 4b, LST f), in which case the base of the LST extends from the shelf edge to the basal slope, as schematically shown in the original sequence stratigraphy model (Cc in Fig. 1).
From a practical perspective (especially in seismic analysis, and in most Quaternary examples), there are cases where the unconformity may be more confidently correlated with the transgressive surface (e.g. Ts in Fig. 4) at the top of the LST (e.g. Ridente & Trincardi 2002), resulting in the so-called ‘T–R cycle stratigraphy’ of Embry & Johannessen (1992). Analogously, ‘genetic stratigraphy’ (Galloway 1989) relies on maximum flooding surfaces as alternative sequence bounding surfaces at least where these are more readily detectable by means of available investigation methods (e.g. facies analysis). Although different, there are no conceptual inconsistencies between these alternatives, which are easily translatable into one another and do not represent different models as to how the sequential architecture was formed. Nevertheless, it should be admitted that the above stratigraphic relationships are more evident in seismic data, and are more difficult to define through analysis of outcrops.
Discussion
Back to sea-level
To preserve the original centrality of sea-level, a model of ‘sequential stratigraphy’ should lay emphasis on the allocyclic component of changes in accommodation controlling stratigraphic architecture at the scale of a sedimentary basin. More specifically, such a model implies the following: (1) relative sea-level in the form of allocycles is a key mechanism by which marine sediments of the shelf environment are shaped into depositional sequences; (2) allocycles determine regional-scale, shelf-wide unconformities that can be regarded as sequence boundaries; (3) maximum flooding surfaces, also the product of allocycles, may compete with the erosional unconformity as suitable sequence boundaries; (4) possibly (but not at all certainly), the shelf unconformity can be correlated seaward with a conformable surface and the sequence boundary extended to the deeper basin; (5) because in deep water accommodation is not significantly affected by sea-level change, sediments forming there retain no architectural (i.e. sequential) relationship with the driving sea-level cycle; nonetheless, sediment flux to the sink is sensitive to shoreline migration across the shelf (Posamentier & Vail 1988; Posamentier et al. 1988; Porębski & Steel 2003), although timing and patterns of deep-water sediment accumulation may be different from those predicted by the sequence stratigraphy model (Burgess & Hovius 1998).
Based on the above, the architecture of depositional sequences consists of building blocks (i.e. systems tracts) reflecting a cyclical pattern in which shelf deposits of the lowstand, transgressive, highstand and falling stage display specific stratigraphic relationships. Stratigraphic relationships among and within sequences are essentially reflected by geometric attributes of strata; these geometric attributes are affected, although not directly reflected, by the variable nature of sediments as expressed by facies assemblages from different depositional environments. The dominance of regional/geometric versus environmental/facies attributes within the criteria for defining sea-level-controlled sequences is a bias of the original sequence stratigraphy model being derived from seismic stratigraphy; this implies that the efficacy of the model will eventually vary along with the very diverse working contexts and methods of exploitation.
Ancient and Quaternary sequences: which model?
With respect to the applicability of sequence stratigraphy, ‘standard’ and ‘non-standard’ cases can be distinguished based on the degree to which they reflect the central assumption of the model; therefore, sea-level-dominated settings provide better standard cases than supply-dominated settings. The Exxon model, however, descends from case studies where defining the role of sea-level is problematic, thus it cannot unambiguously represent the standard cases. Whereas several researchers have pointed out that Quaternary settings, dominated by glacio-eustatic sea-level cycles of known periodicity and amplitude, represent a natural laboratory for testing the sequence stratigraphy model, little attention was given to the implications of such a claim: the central assumption of the sequence stratigraphy paradigm is assuredly reliable in the case of Quaternary sequences, hence they provide the reference frame not only for testing but also for establishing a sea-level-driven model of sequence stratigraphy.
The overall architecture of Quaternary sequences, with its possible variants (Fig. 4b), may result in being considerably different from that predicted by the original sequence stratigraphy model (Fig. 1), largely derived from Meso-Cenozoic greenhouse cycles of 1 – 3 myr duration (Williams 1988; Miller et al. 2005). In particular, the stratigraphic continuity of HST, FST and LST regressive deposits within Quaternary sequences is not accounted for in the Exxon model, in which sea-level fall is conceived only as an erosional phase and sedimentation reaches its acme around the lowstand phase and during the late highstand (Fig. 1). Such differences affect the prediction potential of the model relative to sedimentation rates during falling, lowstand and highstand intervals; however, they can be accounted for by modulating the ratio between cycle duration and amplitude of the sea-level excursion within a cycle: different scaling of these parameters affects the amount of sediment flux and the filling of the available accommodation during a full cycle, thus controlling the overall thickness and shape of systems tracts and sequences (e.g. Lobo & Ridente 2014). On this basis, Quaternary and older sequences are reconcilable as representing distant members within a model where, at increasing duration and/or decreasing amplitude of relative sea-level change, the basic architecture departs from that of Quaternary sequences and progressively meets that of older sequences (Fig. 8).
Qualitative scheme of the competing influence exerted by main factors controlling depositional cycles and sequence architecture. Each vertex of the diagram represents the maximum effect of one of three parameters, which decreases towards the other two vertices. Competing of sea-level amplitude and sedimentation flux results in different sequence architectures that can be represented by the ‘Quaternary’ (e.g. Fig. 4) and Exxon (e.g. Fig. 1) models, respectively. It is assumed that increasing duration of the cycle is also important and that it probably enhances the sedimentary response, thus favouring exploitation of accommodation as in high-supply settings; on longer time scales, however, average sedimentation rates are expected to be lower than in typical supply-dominated settings. Sea-level-dominated settings (SLd) produce the fourth- and fifth-order-like Quaternary sequences; supply-dominated settings (SFd) produce the third-order-like pre-Quaternary sequences; long period cycles produce the first- and second-order-like pre-Quaternary sequences. The stratigraphic architecture of sequences forming under conditions that fall close to the three vertices is well documented; however, possible variants and transitional contexts between the three main fields need to be exploited and eventually used to test and calibrate the model.
Therefore, the schemes in Figures 1 and 4 do not represent fundamentally different models but rather a different balance, within the same general model, between the main parameters of cycle duration, sea-level amplitude and sedimentation rates (Fig. 8). By modulating this balance (i.e. by accounting for the variable interplay of sea-level and supply, within a given time interval) sequence architecture shifts from one end member (i.e. highly dominated by sea-level) to the other (highly dominated by local supply dynamics and autocyclic mechanisms). Many examples of well-constrained architectures and related controlling parameters (i.e. the overall balance) are documented in the literature, and can be used to explore and define the different contexts and possible scenarios that can be represented by a model in which the control of sea-level, compared with other depositional factors, can vary significantly (Burgess et al. 2006).
Summary and conclusions
It is herein argued that the central importance of sea-level, as one genetic mechanism above all others, was plainly stated in the original presentation of seismic stratigraphy. In fact, the descriptive method in AAPG Memoir 26 relied on the assumption that eustasy controlled the origin of sequence boundaries and depositional sequences worldwide. Significant pitfalls inherent to such a generalized conception have been readily stressed by early workers (i.e. Wilgus et al. 1988); however, recommendations to avoid straightforward application of simplified schemes have not always been pursued by subsequent practitioners, and also critics of sequence stratigraphy have somewhat overlooked the repeated warnings and suggestions towards a critical application of the model (i.e. Kendall & Lerche 1988; Posamentier & Vail 1988; Posamentier et al. 1988).
The commitment to the importance of sea-level discloses conceptual constraints that may limit the application of sequence stratigraphy in cases where sea-level control is neither overwhelming nor unambiguously recognizable in the stratigraphic record. These practical limitations, however, do not affect the conceptual consistency of sequence stratigraphy; consistency, instead, is seriously undermined when loose interpretation of basic concepts is permitted in order to endorse application of sequence stratigraphy to the most diverse settings.
Whereas in the most recent versions of sequence stratigraphy emphasis is given to how sequences can be delimited, in the original model the focus was on what controlled the sequential architecture of continental margins, as unveiled in the (seismic) stratigraphic record. However, although claiming to represent basically eustatically driven depositional cycles, the Exxon model suffered the bias of being derived from contexts where sea-level oscillations (of unknown origin and amplitude) were probably less overwhelming compared with other environmental factors (particularly sediment supply regimes).
Quaternary examples indicate that the forcing control of sea-level over sedimentation rates depends strongly on the duration of the full cycle and of single phases, scaled to the amplitude of the sea-level oscillation: the modulation of these parameters (notwithstanding the great variability of depositional settings and environments) largely influences the competing effects of environmental factors in controlling sedimentary dynamics and shaping sequences. In this view, Quaternary glacio-eustatic cycles of known periodicity and amplitude afford the optimal scenario for defining a general model of how sediments are shaped under the dominant (although not unique) control of relative sea-level change.
Disparities between Quaternary sequences and the architecture predicted by the Exxon model have been traditionally accommodated by interpreting the former as special cases, long estranged from the refining of the latter. In contrast, Quaternary sequences are herein conceived as representing a standard case in terms of strong sea-level impact, whereas the Exxon model probably represents contexts with subdued sea-level impact. By bridging the two, with the background of the variable interplay of relative sea-level cycles and sedimentary dynamics, the construction of a comprehensive model of sequential stratigraphy can be addressed.
Acknowledgements and Funding
I am grateful to P. M. Burgess and P. A. Allen for greatly improving the paper with their review and for their support and assistance as Editors. I also thank the colleagues and friends who read early versions, sharing discussion and providing insightful comments.
- © 2016 The Author(s)
