Chapter 3
Basin Models

PAUL B. WIGNALL & JIM BEST

3.1 Introduction

The outcrops of the Shannon region provide some of the best Carboniferous sections to be seen in north-west Europe. They are also amongst the most debated. Basin style is an especially contentious issue and several conflicting, and to some extent irreconcilable, models have been proposed. Each model differs in the implications it has for depositional conditions, the orientation of depositional systems and sediment provenance. All of these debates can be pondered when examining the outcrops detailed in this field guide, which provides a chance to see some of the key evidence and, we hope, allow the visitor to come to their own conclusions based on what they have seen. This chapter highlights the key attributes of the various competing depositional models that have been proposed for the Shannon Basin.

The Carboniferous basin has acquired several names in various studies over the years: the Shannon Trough (Sevastopulo, 1981a; Sleeman & Pracht, 1999; Tanner et al., 2011), the Shannon Basin (Strogen, 1988; Sevastopulo, 2001; Martinsen et al., 2003; Pyles, 2007, 2008), the Clare Basin (Deeny, 1982; Croker, 1995), the West Clare Basin (Gill, 1979) and the Western Irish Namurian Basin, usually abbreviated to WINB (Collinson et al., 1991). All are valid (and the Shannon Basin is used here), although Martinsen (et al., 2003) recommended the WINB name be discarded because basin development began prior to the Namurian.

The Shannon Basin is generally recognized to have had an elongate east-north-east to west-south-west orientation aligned along the trend of the modern Shannon Estuary. The basin stretches from central Ireland and passes westward into the Atlantic Ocean where it is sharply truncated by the north-south trending Porcupine Basin; a much younger structure developed around 150 km west of the County Clare coastline (Strogen, 1988). Overall, the Shannon Basin was approximately 250 to 300 km wide along its roughly east-west axis and at least half that value in a north-south orientation. However, neither its northern or southern terminations are well defined due to a combination of erosion and/or poor inland outcrops. For example, the Namurian stratigraphy of north-west County Cork is nearly identical to that seen in County Clare (Morton, 1965), suggesting that the County Cork strata should be included within the Basin even though this area is usually considered to be south of it. Despite its considerable extent, most studies (and this field guide) concerning the Shannon Basin have focused on the best-exposed sections around the Shannon Estuary and the County Clare coastline, an area that only measures 80 km by 80 km – a small part of the whole Basin.

3.2 Dinantian Basin History

Broad regional subsidence began at the start of the Early Carboniferous (Tournaisian Series; see Figure. 1.7.2) over a broad area, with the result that shallow-water heterolithic deposition became widespread (Strogen, 1988; Sevastopulo & Wyse Jackson, 2009). At this point, the Shannon Basin was not a clearly distinguishable topographic entity but instead was merely part of an extensive shelf sea extending across central southern Ireland. These facies pass southwards into a thick succession of clastic sediments (the Culm Facies) that accumulated in the South Munster Basin, a north-west to south-east trending trough (Price & Todd, 1988).

A topographically distinct Shannon Basin became established in the later Tournaisian as subsidence rates accelerated along the axis of the Shannon Estuary. This coincided, presumably coincidentally, with the onset of highly distinctive, but rather enigmatic, Waulsortian mudbank carbonate deposition. Sometimes forming coalescing, reef-like mounds, these Waulsortian carbonates lack constructing organisms like normal reefs and instead have a high proportion of cement that infills large-stromatactis-like cavities and fissures (Lees, 1961; Lees & Miller, 1995). The Waulsortian facies of southern Ireland are the most extensive strata of this unusual limestone, and the huge thickness in the Shannon Basin, where it approaches 1200 m thick, is by far the thickest development in the world (Shephard-Thorn, 1963; Sleeman & Pracht, 1999). The Waulsortian limestones can be seen in the excursions described by Somerville in Chapter 4.

The Shannon Basin Waulsortian limestones accumulated in the final ~ 6 Myr of the Tournaisian (Fig. 1.7.2), indicating that accumulation rates were of the order of 200 mm kyr⁻¹. However, the absence within these limestones of any wave reworking or organisms that lived in the photic zone (e.g. calcareous algae) suggests that water depths were more than 200 m deep, despite the impressive aggradation rates (Lees & Miller, 1995). Clearly subsidence was extremely high in the Shannon Estuary area at this time.

The rapid formation of the Shannon Trough was also associated with volcanic activity. Around Limerick, on the eastern edge of the super-thick Waulsortian mud mound development, the limestones are overlain by a thick succession of earliest Viséan (Chadian Stage) alkali basalts and trachytes with tuffs. Further volcanics were also developed south of Limerick (Shelford, 1967; Strogen, 1988). The later Viséan strata in the Shannon Estuary area record an overall deepening-upwards trend, with the development of fine-grained, thin-bedded, limestones of the Parsonage Formation. The presence of some rudaceous limestones (the Inishtubrid Beds) found in the Corgrig Lodge Formation suggests steep slopes were developed, at least locally, probably associated with small reefs (Sleeman & Pracht, 1999). However, overall a ramp was developed in the Shannon Trough, deepening to the west and north-west from Limerick (Somerville & Strogen, 1992; Sleeman & Pracht, 1999). Ultimately, by the Brigantian Stage, calciturbidites developed at Ballybunnion in the western-most outcrops of the Shannon Basin. In contrast, contemporaneous limestones in northern County Clare remained at shallow depths throughout the Viséan and several emergence horizons with palaeosols are seen (Sleeman & Pracht, 1999; Somerville, Chapter 5). Thus, a northern carbonate platform region appears to have developed adjacent to a deep Shannon trough. After the Waulsortian Limestone development and its exceptional local thickness development in the Shannon Estuary area, lateral variations in thickness of Viséan strata become much more subdued (Fig. 1.7.3).

So, what type of basin was the Shannon Basin in the Dinantian? Several ideas have been suggested. Most agree that the site of greatest subsidence lay along the Shannon Estuary and probably reflects the presence of an underlying Iapetus Suture (e.g. Sevastopulo & Wyse Jackson, 2009; Warr, 2012; Graham, Chapter 2). However, despite this consensus, there is disagreement concerning basin style. Several studies have suggested that the Shannon Basin was the product of extensional rifting in a similar manner to that seen in contemporaneous basins in northern Britain (Deeny, 1982; Haszeldine, 1984; Collinson et al., 1991; Martinsen et al., 2003). However, the sudden but short-lived accelerated pulse of subsidence in the late Tournaisian and the development of volcanoes on the eastern margin of the Shannon Basin are both features atypical of extensional basins. Volcanism produced by extension and decompression melting should see eruptions occurring in the centre of a basin. Instead, these observations better fit a transtensional origin for the Shannon Basin (Haszeldine, 1984; Strogen, 1988; Warr, 2012). A third alternative is that the onset of foreland flexure, as the Gondwanan continent approached the Laurasian continent, may have caused subsidence along the Iapetus Suture (Tanner et al., 2011). In this scenario, the South Munster Basin to the south of the Shannon Estuary became an under-filled foreland basin in the Dinantian, whilst the Shannon Basin was the product of block faulting in the fore-bulge region (Higgs, 2004). A problem for this third model is, once again, the volcanism present at Limerick, which is not anticipated in foreland settings.

3.3 Namurian Basin History

The Late Carboniferous witnessed a fundamental and dramatic change in depositional style. Limestones disappeared and anoxic black shale facies of the Clare Shales Formation developed in the Shannon region in the early Namurian (Hodson, 1954a; Braithwaite, 1993). Identical black shale facies developed over much of Ireland and the British Isles at the same time (Davies et al., 2012). Despite the carbonate-to-clastic transition, the same north-south thickness trends persisted from the Dinantian through to the Namurian. Thus, the Clare Shales are extremely thick, peaking at 232 m at Inishcorker in the centre of Shannon Estuary region (this is the same area of peak thickness of the Waulsortian limestones), and they are much thinner (< 20 m) in northern County Clare. The succeeding deep-water clastic sediments of the Shannon Group also show the same thickness trends, although the thickness variation is much reduced in the overlying Central Clare Group.

The similar thickness trends between the Dinantian and Namurian sediments have led some to suggest that the Namurian Shannon Basin inherited the same bathymetry and subsidence regime as seen in the Dinantian (Haszeldine, 1984; Collinson et al., 1991; Sevastopulo & Wyse Jackson, 2009). However, continuous deposition between Viséan and Namurian strata is only seen within the Shannon Estuary area. Elsewhere in the Shannon Basin, and across much of central and southern Ireland generally, there is a major hiatus at the boundary during which uplift, folding and erosion occurred (Hodson, 1959). Thus, in the Doonbeg No. 1 borehole in central County Clare, the Dinantian limestones dip at a different angle to the unconformably overlying Namurian clastics (Croker, 1995). The significance attributed to this tectonic episode varies considerably from model to model but, before discussing the merits of the various alternatives, we first need to present them.

3.4 The CoMa Model

In a highly influential paper, Collinson et al. (1991) presented an inclusive model for the Shannon Basin that built on the earlier observations of Hodson & Lewarne (1961). The model was subsequently challenged (see below) and vigorously defended (e.g. Martinsen & Collinson, 2002; Martinsen et al., 2003). Here we term it the CoMa model after its two principal proponents: Collinson and Martinsen.

In essence, the CoMa model envisages business as usual between the Dinantian and Namurian phases of basin development. Thus, continued extensional rifting along the Iapetus Suture/Shannon Estuary trend caused a narrow, confined basin “elongated along an ENE-WSW line” (Martinsen et al., 2003, p. 791) to persist, with shallower conditions lying to the north-east in northern County Clare. After the initial phase of fill by thick, black shales, the sandstones of the Ross Sandstone Formation turbidite system arrived from a source area to the north-west and, having been deflected along the axis of the basin, aligned along the Shannon Estuary, then prograded along the basin floor. The majority of palaeocurrents in the Ross Sandstone Formation consequently record “flow to the north-east” (Fig. 3.4.1A; Collinson et al., 1991, p. 235).

Fig. 3.4.1. Reconstruction of depositional conditions during Namurian infill of the Shannon Basin, based on the CoMa model (after Collinson et al., 1991). Contours, labelled 1 to 3 from shallow to deep, show slope orientations, whilst the arrows show dominant sediment dispersal directions. A) the Ross Sandstone Formation turbidite system; B) the Gull Island Formation slope system, with slumps and slides; C) the Tullig Cyclothem deltaic system.

The biostratigraphic framework established by Hodson & Lewarne (1961) shows that the Ross Sandstone Formation passes north-eastward into a thin development of the Clare Shales. The black shales in this region are considered to have accumulated on a basin margin in a “shallow, marginal area [where] agitation kept fines in suspension” (Collinson et al., 1991, p. 237). Slumps are abundant in the upper Ross Sandstone Formation and they become an even more common style of soft-sediment deformation in the overlying siltstone-dominated Gull Island Formation (Strachan & Pyles, Chapter 8). The down-slope movement direction of the majority of the slumps is interpreted to be to the south-east (Martinsen & Bakken, 1990) and thus is at 90° degrees to the progradation direction of the turbidites. In the CoMa model, this is interpreted to represent collapse of a slope prograding to the south-east whilst the turbidites are flowing along the foot of the slope to the north-east (Fig. 3.4.1B). The slope system of the Gull Island Formation was fed by shelf-edge deltas and this unit is thus overlain by a series of deltaic units, beginning with the Tullig Cyclothem (Fig. 3.4.1C). Like the Gull Island Formation, the down-slope direction, in this case palaeocurrent indicators in channel sandstones, are said to suggest “flow dominantly to the south-east” (Collinson et al., 1991, p. 236), in accord with Pulham’s (1989) study of the cyclothem sediments of the Central Clare Group.

In the original manifestation of the CoMa model, there was some discrepancy in the suggested orientation of the deep-water trough in the Shannon Basin (Wignall & Best, 2000). Although Collinson and colleagues considered the basin to be developed above the Iapetus Suture, which runs east to west along the Shannon Estuary, the Ross Sandstone Formation turbidites were stated to flow along this axis to the north-east in the direction of the northern basin margin in the CoMa model.

In a re-validation of the CoMa model, Martinsen et al. (2003) modified it somewhat. They suggested that the early phase of fill, up to a level in the mid Gull Island Formation, was primarily an aggradational infill of a deep, narrow, trough centred at Loop Head in south-western most County Clare. At this point, the topographic low had been infilled, thus allowing the turbidity currents in the upper Gull Island Formation to spill-over and prograde northwards. Thus, the “basin high [in northern County Clare] received turbiditic sediments once the central depression of the basin had been filled and margins onlapped” (Martinsen et al., 2003, p. 802). Nonetheless, the principal progradation of the Gull Island Formation slope remained to the south-east in this modified version of the CoMa model. Intriguingly, this modification has turbidity currents flowing along a slope up onto a basin high. The healed slope accommodation is then downlapped by the delta slope of the Tullig depositional system prograding south-east (Martinsen et al., 2003).

3.5 The Clockwise Minibasin Model

Pyles (2007, 2008) has proposed a similar basinal model to the modified CoMa model in which the Ross Sandstone Formation is once again envisaged to be a ponded submarine fan (Fig. 3.5.1; Pyles, 2007, 2008; Pyles & Jeanette, 2009). However, the model of Pyles differs in significant details. The key difference is that the Shannon Basin is no longer considered to be an elongate trough but rather a minibasin of circular outline analogous to “structurally confined, salt-withdrawal basins [such as those found today] on the northern Gulf of Mexico continental slope” (Pyles, 2008, p. 568). Palaeocurrent data from the Ross Sandstone Formation show a dominant north-eastwards flow but with considerable spread from north-west to south-east (Pyles, 2007; Pyles & Jeanette, 2009). This is interpreted to record a radially-dispersive, lobe-dominated, aggradational turbidite system (Fig. 3.5.1A). The Minibasin model also differs in its interpretation of the Clare Shales/Ross Sandstone Formation contact in the Shannon Estuary. In the CoMa model this is seen to be a downlap relationship, whereas in the Minibasin model the Ross Sandstone Formation turbidites onlap a lateral slope of Clare Shale (Pyles & Jeanette, 2009). It should be noted though, that this purported slope of organic-rich muds lacks any evidence for slope failure.

Fig. 3.5.1. The Clockwise Minibasin Model of Pyles (2007, 2008). Contours, labelled 1 to 3 from shallow to deep, show slope orientations, whilst the large grey arrows show dominant sediment dispersal directions. A) a ponded submarine fan model for the Ross Sandstone Formation; B) the Gull Island Formation slope system, with slumps and slides. C) the Tullig Cyclothem deltaic system.

The Minibasin model also differs from the CoMa model in the inferred provenance of the sediment supply. The CoMa model considers the hinterland to be a Caledonian Province to the north-west of the Shannon Basin whereas Pyles (2008) suggests initial sediment supply was from the south-south-west. This source was then joined by westerly-supplied sediment during deposition of the upper Ross Sandstone Formation. Later, in the Gull Island Formation, only the western source was active and finally the sediment source was proposed to change again in the Central Clare Group to a north to north-west provenance. Thus, there is “a clockwise change in the sediment transport direction through time associated with the filling of the basin” (Pyles & Jeanette, 2009, p. 1975–6; Pyles & Strachan, Chapter 7) – hence a clockwise minibasin model (Fig. 3.5.1).

3.6 The WiBe Model

A fundamentally different basin model was proposed by Wignall & Best (2000, 2002, 2004) for the Namurian infill of the Shannon Basin, here abbreviated to the WiBe model, in which basin style is envisaged to change fundamentally between the Dinantian and the Namurian. In this alternative model, Namurian clastic sediment was proposed to be derived from the south and south-west with the basin being infilled by depositional systems that were building, down-dip, to the north-east (Fig. 3.6.1). The Ross Sandstone Formation turbidite system is viewed as unconfined and showing north-eastward downlap onto the deeper water black shales of the Clare Shale Formation (Fig. 3.6.1A). A plot of the extent of turbidite deposition accords with this north-eastward progradation (Wignall & Best, 2000, fig. 7; Pyles & Strachan, Chapter 7). In contrast, the turbidites onlap a distal basin margin in the modified CoMa model (Martinsen et al., 2003). The Gull Island Formation remains a slope system in the WiBe model but with the primary down-dip directions, based on evidence from growth faults and slumps, to the north-east (Fig. 3.6.1B). The deltaic systems of the Tullig Cyclothem continue the theme of north-easterly progradation (Fig. 3.6.1C).

Fig. 3.6.1. Reconstruction of depositional conditions during Namurian infill of the Shannon Basin, based on the WiBe model (after Wignall & Best, 2000, 2002). Contours, labelled 1 to 3 from shallow to deep, show slope orientations, whilst the large grey arrows show dominant sediment dispersal directions. A) the Ross Sandstone Formation turbidite system; B) the Gull Island Formation slope system, with slumps and slides; C) the Tullig Cyclothem deltaic system.

This WiBe model interpretation of Tullig Cyclothem migration is in accord with the earlier work of Rider (1974) who also suggested that the Tullig deltas show progradation to the north-east. This is at variance to the proposed south-eastward direction proposed by Pulham (1989). However, Pulham’s own data clearly shows that nearly all the palaeocurrents in the Tullig Cyclothem record flow to the north-east (Pulham, 1989, fig. 14), but he suggested that this was because the measurements come from a distributary channel developed at a high angle (~ 90°) to the main progradation direction (Pulham, 1989, fig. 19).

The CoMa and Clockwise Minibasin models both consider the area of thickest Shannon Group sediments to record the deepest water parts of an extensional basin where infill was predominantly aggradational. In contrast, the WiBe model views the infill as essentially progradational, albeit with thickening in the Shannon Estuary region due to rapid subsidence along the Iapetus Suture. This has been attributed to flexural loading (Higgs, 2004) and suggests that the Namurian phase of the Shannon Basin saw a foreland basin develop. Thus the change in structural style between the Viséan and Namurian coincides with a regional phase of uplift and moderate folding, probably of a transient fore-bulge, especially to the south of the Shannon Basin (Strogen, 1988). The tectonic readjustment from a transtensional to compressional regime explains the change of basin configuration from one that deepened from east to west in the Viséan (Sleeman & Pracht, 1999) to one that deepened from south-west to north-east in the Namurian.

The development of a foreland basin in the Namurian also provides an explanation for the shallow- to deep- water transition seen in northern County Clare in the WiBe model (Wignall & Best, 2002). In this region, the Brigantian platform carbonates are succeeded by Clare Shales but are separated from them by a thin development of phosphatic pebbles (the St. Brendan’s Well Phosphate Bed – see Chapter 6; Braithwaite, 1993). These phosphates rest on a truncation surface that records a hiatus spanning the late Brigantian to Chokerian stages – an interval of ~ 10 Myr (Wignall & Best, 2000; Barham et al., 2014). The region was probably emergent during some of this time, as evidenced by rare, reworked limestone pebbles that occur in the Phosphate Bed; but the renewed onset of subsidence was not compensated by sediment infill until the onset of black shale accumulation. Even if subsidence only spanned half the duration of the hiatus and with modest rates of 0.5 mm kyr⁻¹, water depths of ~ 200 m could have been achieved in northern County Clare prior to black shale deposition. Subsidence rates were undoubtedly higher in southern County Clare, but here subsidence was compensated for by the accumulation of hundreds of metres of black shales and turbidites.

Support for a foreland basin model has also come from the kinematic retro-modelling of fold and thrust structures seen in the Namurian sections of the County Clare coastline. Previously considered to be a product of Variscan compression in the latest Carboniferous, Tanner et al. (2011) have shown syn-sedimentary compression and folding had begun by at least the base of the Tullig Cyclothem. Tanner et al. (2011) therefore favour a foreland basin model, but they note that subsidence rates in the Shannon Estuary region were exceptionally high for such a basin style. This was presumably due to the ready subsidence on the weak Iapetus Suture.

3.7 Debate on Basin Development

All three alternative models outlined above raise key questions that can be pondered whilst visiting the Namurian sections detailed in this field guide. Pertinent areas for debate include:

1. What was the nature of the northern County Clare sections? In particular, are the Namurian strata in this region deeper or shallower water than the equivalent strata exposed around the Shannon Estuary? The interpretation of the Clare Shale in northern County Clare is especially contentious. In the WiBe model, the black shales are deep water, low energy deposits accumulated in a distal basinal setting, whilst the other models suggest high energy, winnowed accumulation on a basin margin.
2. What direction was downslope? There are an unrivalled variety and abundance of slope collapse indicators to be seen in the Shannon Basin. They include spectacular slumps, debris flows and numerous growth faults that provide the opportunity to assess their movement direction (Rider, 1974; Gill, 1979; Martinsen et al., 2003; Wignall & Best, 2000, 2002, 2004). The CoMa model envisages slope progradation to the south-east and its proposers have supported this assertion with several studies of large slumps in the Gull Island Formation (Martinsen & Bakken, 1990). However, detailed re-evaluation of several large examples has re-interpreted movement directions to show a narrow range of movement directions spanning east-north-east to north-east (Strachan & Alsop, 2006), in accord with the WiBe model. Whilst difficult to interpret, some major slope collapse features, such as the slumps and growth faults seen at the Point of Relief (see Strachan & Pyles, Chapter 7), unequivocally record down slope collapse to the west – a direction that is impossible to reconcile with the CoMa model.
3. What was the sediment provenance? Whilst difficult to judge in the field, when visiting the Namurian outcrops in County Clare it is worth considering whether there is any lithological change to be seen amongst the various sandstone units. The Clockwise Minibasin model in particular, invokes highly variable provenance at different horizons, whereas the WiBe model proposes a consistent source to the south-west and the CoMa model has sediments with a north-westerly provenance.

   In the only provenance study on the basin, Pointon et al. (2012) have dated detrital zircons from throughout the Namurian infill and found a range of age groupings. Of these, the most diagnostic are a 500 to 700 Ma group that indicates a source area that was sampling “Cadomian-Avalonian orogenic activity within terranes to the south” (Pointon et al., 2012, p. 77). As there is no available source for these zircons north of the Shannon Basin, their presence unequivocally indicates that Gondwanan sediment was reaching County Clare with the Ross Sandstone Formation turbidity currents – clear and unequivocal support for foreland basin development early in the Namurian.

4. What was the predominant progradation direction of the deltaic systems of the Central Clare Group? There is a clear difference between the CoMa and WiBe models, a conflict that is also mirrored in the different views of Pulham (1989) and Rider (1974). Abundant palaeocurrent evidence is available to test these two models and fit these in with reconstructions of the depositional systems, including the changing size of fluvial-deltaic channels in a proximal-distal traverse and the orientation of any palaeovalleys within these sediments. Furthermore, there is good evidence that the progradation direction of the deltaic systems changed between the Tullig Cyclothem and the overlying Kilkee and Doonlicky Cyclothems (Rider, 1974; Gill, 1979; Pulham, 1989). Thus, the Tullig palaeocurrents generally record flow to the north and north east, whereas flow directions swing clockwise in the younger cyclothems and the Doonlicky flow directions are to the south east. Despite this trend, provenance data suggest there is little change. The study of Pointon et al. (2012) showed that all cyclothem sediments possess a major component of southerly-derived sediment. It is worthy of consideration as to why the predominant flow direction may have changed, how this influenced the resulting depositional geometry and if the source of sediment also changed or whether sediment reworking was a major contributor to the upper cyclothems.
5. What are the various scales of autocyclic and allocyclic control on sedimentation within these Namurian sediments and how are base-level changes recorded in the rock record within these sediments? As one example, although the distinct marine bands show periods of marine incursion, other evidence of marine influence, in the form of fully bioturbated sediments, is highly variable especially in the Central Clare Group. How are base-level changes manifested within both the deeper water sediments of the Clare Shales, Ross Sandstone Formation and Gull Island Formation and what role do autocyclic and allocyclic scour play in the marked erosion surfaces that can be seen within the cyclothems? For instance, the distinct erosion surfaces that are present at the base of some of the fluvial channel sandbodies have been proposed to represent type-1 sequence boundaries (Davies & Elliott, 1996; Hampson et al., 1997) as well as showing the influence of autocyclic scour (Wignall & Best, 2000). It is clear that the nature of the facies above and below these boundaries, and their lateral variation, hold the key to this debate.