2. Stratigraphy

The oldest proven rocks on the peninsula are the Tremadoc to Arenig shales, sandstones and mélange of the Annascaul Formation (Fig. 3; Todd et al. Reference Todd, Connery, Higgs and Murphy2000). These Ordovician deep-water sediments have been correlated with the Ribband Group of SE Ireland and, as such, are assigned a Leinster terrane or Avalonian affinity (Todd et al. Reference Todd, Connery, Higgs and Murphy2000). The Annascaul Formation is confined to two inliers in the east of the peninsula, the most extensive being the Annascaul inlier (Fig. 4). The Annascaul Formation is overlain unconformably by the mudstones, tuffs and limestones of the Wenlock Ballynane Formation (Dunquin Group, Fig. 3; Parkin, Reference Parkin1976; Todd et al. Reference Todd, Connery, Higgs and Murphy2000). The Ballynane Formation is overlain, apparently conformably, by the mudstones/siltstones of the Caherconree Formation and the sandstones/mudstones of the Derrymore Glen Formation (Dunquin Group, Fig. 3). Both these formations are proven to be Ludlow in age through graptolite and shelly faunas (Parkin, Reference Parkin1976).

Figure 3. Chronostratigraphy of the Palaeozoic rocks of the Dingle Peninsula (updated from Todd Reference Todd, Friend and Williams2000, fig. 3). Compiled from various sources by area. 1. Todd, Williams & Hancock (Reference Todd, Williams and Hancock1988), Pracht (Reference Pracht1996), Richmond & Williams (Reference Richmond, Williams, Friend and Williams2000) and Higgs & Williams (Reference Higgs and Williams2011); 2. Todd, Boyd & Dodd (Reference Todd, Boyd, Dodd, McMillan, Embry and Glass1988), Todd, Williams & Hancock (Reference Todd, Boyd, Dodd, McMillan, Embry and Glass1988) and Richmond & Williams (Reference Richmond, Williams, Friend and Williams2000); 3. Holland (Reference Holland1988), Todd, Boyd & Dodd (Reference Todd, Boyd, Dodd, McMillan, Embry and Glass1988) and Todd, Williams & Hancock (Reference Todd, Williams and Hancock1988); 4. Holland (Reference Holland1987, Reference Holland1988), Todd, Boyd & Dodd (Reference Todd, Boyd, Dodd, McMillan, Embry and Glass1988), Todd, Williams & Hancock (Reference Todd, Williams and Hancock1988), Todd (Reference Todd, Arthurton, Gutteridge and Nolan1989, Reference Todd and Went1991) and Higgs (Reference Higgs1999); 5. Parkin (Reference Parkin1976), Todd, Boyd & Dodd (Reference Todd, Boyd, Dodd, McMillan, Embry and Glass1988) and Todd et al. (Reference Todd, Connery, Higgs and Murphy2000); 6. Horne (Reference Horne1974), Parkin (Reference Parkin1976), Higgs & Russell (Reference Higgs and Russell1981) and Pracht (Reference Pracht1996); 7. Higgs & Russell (Reference Higgs and Russell1981), Pracht (Reference Pracht1996) and Williams et al. (Reference Williams, Sergeev, Stössel and Ford1997, Reference Williams, Friend and Williams2000); 8. Russell (Reference Russell1978), Higgs & Russell (Reference Higgs and Russell1981) and Williams et al. (Reference Williams, Sergeev, Stössel and Ford1997, Reference Williams, Friend and Williams2000); 9. Higgs & Russell (Reference Higgs and Russell1981) and Williams et al. (Reference Williams, Sergeev, Stössel, Ford, Higgs, Friend and Williams2000). AN – Annascaul Formation; BF – Ballyferriter Formation; BH – Bull's Head Formation; BM – Ballymore Formation; BR – Ballyroe Group; BS – Ballinskelligs Sandstone Formation; CD – Carrigduff Group; CG – Coosglass Formation; CO – Coumeenoole Formation; CP – Cappagh Sandstone Formation; CS – Chloritic Sandstone Formation; DQ – Dunquin Group; EK – Eask Formation; GB – Glashabeg Conglomerate Formation; IC – Inch Conglomerate Formation; KM – Kilmurry Sandstone Formation; LK – Lack Formation; LLS – Lower Limestone Shales; LS – Lough Slat Conglomerate Formation; PG – Pointagare Group; SF – St Finan's Sandstone Formation; SH – Slea Head Formation; SM – Slieve Mish Group; SW – Smerwick Group; TB – Trabeg Conglomerate Formation; VS – Valentia Slate Formation. The time scale is from Gradstein, Ogg & Schmitz (Reference Gradstein, Ogg and Schmitz2012). The Welsh-derived stage names commonly used in the UK for the Ordovician are provided as well as the international standard.

Figure 4. Simplified geological map of the Dingle Peninsula showing the main structural elements and the locations of the maps and cross-sections shown in later figures. The lower-hemisphere equal-area projections provide orientations and dimensions of strain data collected during field mapping of the peninsula. BH – Bull's Head inlier; BIF – Ballyickeen Fault; BN – Ballynane; BSS – Ballysitteragh Syncline; CC – Caheracruttia; DG – Derrymore Glen inlier; DOF/FA – Dún an Oír Fault/Feohanagh Anticline; DQF – Dunquin Fault; FC – Farranacarriga; FS – Fahan Syncline; SMA – Slieve Mish Anticline.

This Silurian succession in the eastern Bull's Head, Annascaul and Derrymore Glen inliers is quite different in facies from the main Llandovery–Ludlow succession in the Dunquin inlier, some 20–40 km to the west (Fig. 4; Holland, Reference Holland and Kay1969; Sloan & Bennett, Reference Sloan and Bennett1990; Sloan & Williams, Reference Sloan, Williams and MacDonald1991; Boyd & Sloan, Reference Boyd, Sloan, Friend and Williams2000; Higgs & Williams, Reference Higgs and Williams2011). The oldest rocks in the west are the slates of the Coosglass Formation, now proven to be of Llandovery – early Wenlock age through palynomorphs (Higgs & Williams, Reference Higgs and Williams2011). These rocks are separated by a fault from the rest of the Wenlock–Ludlow Dunquin Group, which comprises non-marine to shallow-marine mudstones, siltstones, sandstones, various volcaniclastic deposits and lavas (Sloan & Bennett, Reference Sloan and Bennett1990; Sloan & Williams, Reference Sloan, Williams and MacDonald1991; Boyd & Sloan, Reference Boyd, Sloan, Friend and Williams2000).

The Dingle Group, a succession of mudstones, sandstones and conglomerates, was deposited in a non-marine basin (Todd, Boyd & Dodd, Reference Todd, Boyd, Dodd, McMillan, Embry and Glass1988). A conformable transition from the uppermost shallow-marine sediments of the Dunquin Group is inferred in two localities. On the mainland, a partially exposed sequence from the upper Ludlovian Croaghmarhin Formation (Dunquin Group) to the Bull's Head Formation (Dingle Group) is interpreted to be the result of the progradation of a wave-dominated barrier shoreline (Boyd & Sloan, Reference Boyd, Sloan, Friend and Williams2000). A better-exposed section crops out on Inishabro, one of the Blasket Islands off the western tip of the Dingle Peninsula. This section also appears to be transitional from the Silurian Inishnabro Formation to the Bull's Head Formation (Todd, Reference Todd1991). In other places, the contact between the Dunquin and Dingle groups is clearly faulted or unconformable (Todd, Williams & Hancock, Reference Todd, Williams and Hancock1988). The variation between unconformable and apparently conformable relationships is interpreted to be the consequence of contrasting subsidence styles between the basin depocentre, intra-basin highs and the margins of the basin (Fig. 3; Todd, Boyd & Dodd, Reference Todd, Boyd, Dodd, McMillan, Embry and Glass1988; Boyd & Sloan, Reference Boyd, Sloan, Friend and Williams2000).

The age of the Dingle Group is constrained by three fossil finds. K. Higgs and colleagues (pers. comm. 2014) have discovered fish fragments and spores from the Trabane Member in the upper part of the Bull's Head Formation. These fossils not only support a lacustrine sand flat margin origin for these deposits, but indicate a middle–late Lochkovian age for the upper part of the Bull's Head Formation (lower Dingle Group; Fig. 3). A palynological assemblage obtained from grey siltstones in the alluvial apron deposits of the Eask Formation is ascribed an early late Pragian age (Higgs, Reference Higgs1999; Fig. 3). A further assemblage from a medial to upper position in the Dingle Group, within the axial fluvial deposits of the Slea Head Formation which is the lateral equivalent of the alluvial fan/apron facies of the Trabeg and Glashabeg formations (Todd, Boyd & Dodd, Reference Todd, Boyd, Dodd, McMillan, Embry and Glass1988; Todd, Reference Todd, Arthurton, Gutteridge and Nolan1989, Reference Todd, Friend and Williams2000), has been interpreted to be early–?middle Emsian in age (Higgs, Reference Higgs1999; Fig. 3). Biostratigraphic evidence therefore shows that most of the Dingle Group is Early Devonian in age. However, the age of the base of the Dingle Group still remains uncertain as no biostratigraphical data have yet been found in the basal Heterolithic Member of the Bull's Head Formation. The Silurian–Devonian boundary most likely occurs in the condensed lacustrine sequence in the lower part of the Bull's Head Formation (Higgs, Reference Higgs1999, p. 195).

The north limb of the Feohanagh Anticline, along the north coast of the peninsula, contains a distinct succession of unconformity-bounded ORS groups (Figs 3, 4). The Dingle Group is represented only by a variable thickness of mudstones and conglomerates of the Glashabeg Formation (Todd, Williams & Hancock, Reference Todd, Williams and Hancock1988). The Smerwick Group overlies the Dingle Group, either in faulted or unconformable contact (Todd, Williams & Hancock, Reference Todd, Williams and Hancock1988). Richmond & Williams (Reference Richmond, Williams, Friend and Williams2000) proposed that the Smerwick Group is of Lower Devonian age like the Dingle Group, and was translated into place as an allochthonous terrane. In the absence of definitive age data and the compelling evidence of a major terrane-bounding fault, the simpler interpretation is that the Smerwick Group is younger than the Dingle Group with the boundary as an angular unconformity that has been faulted in some places (Todd, Williams & Hancock, Reference Todd, Williams and Hancock1988). Four successive unconformity-bounded groups follow the Smerwick Group (Todd, Boyd & Dodd, Reference Todd, Boyd, Dodd, McMillan, Embry and Glass1988; Richmond & Williams, Reference Richmond, Williams, Friend and Williams2000). Three of these are confined to the northern Dingle Peninsula; the youngest, the Slieve Mish Group, oversteps the Dingle Group and older rocks. The Carrigduff Group has yielded spores in a solitary and rather sparse assemblage evaluated by K. T. Higgs as early–middle Frasnian in age (L. K. Richmond, unpub. thesis, 1998; Richmond & Williams, Reference Richmond, Williams, Friend and Williams2000). The Slieve Mish Group, cropping out across the Dingle Peninsula, is the youngest ORS group and passes conformably into marine sandstones and mudrocks with Carboniferous fauna and flora (Todd, Boyd & Dodd, Reference Todd, Boyd, Dodd, McMillan, Embry and Glass1988; Higgs, Clayton & Keegan, Reference Higgs, Clayton and Keegan1988).

On the south side of the Dingle Peninsula the ORS Caherbla Group contains breccio-conglomerates of the Inch Conglomerate Formation and sandstones of the Kilmurry Sandstone Formation. This group is not directly dated, but is generally believed to be of Middle Devonian age (Fig. 3; Todd, Boyd & Dodd, Reference Todd, Boyd, Dodd, McMillan, Embry and Glass1988; Williams et al. Reference Williams, Sergeev, Stössel and Ford1997; Todd, Reference Todd, Friend and Williams2000; Morrissey et al. Reference Morrissey, Braddy, Dodd, Higgs and Williams2011).

3. Variscan deformation

The expression and amount of end-Carboniferous Variscan strain in the Dingle Peninsula is best gauged by structural assessment of the Slieve Mish Group and other upper Palaeozoic formations which clearly post-date Caledonian and Acadian events. Major reverse faulting on the offshore Dingle Bay and North Kerry lineaments is inferred from structural geometries and stratigraphic separations recorded in onshore outcrop in the Dingle and Iveragh peninsulas (Price & Todd, Reference Price and Todd1988; Todd, Reference Todd, Arthurton, Gutteridge and Nolan1989). In the Dingle Peninsula, shortening has produced open to closed, upright folds in the Slieve Mish Group and overlying Carboniferous shales and limestones (Shackleton, Reference Shackleton1940; Capewell, Reference Capewell1951, Reference Capewell1965). These folds have an east–west trend in the east of the peninsula but take on a more NE–SW orientation in the west and have wavelengths of up to several kilometres (Figs 4, 5). They developed by buckling, facilitated by interlayer flexural slip as indicated by bed-parallel quartz slickenfibres with lineations oriented normal to fold axes. Rudimentary line length measurement of unbalanced cross-sections indicates that the buckling achieved shortening of about 14% (Fig. 5).

Figure 5. Structural cross-section through the Variscan structures of the Slieve Mish Group. (a) Slieve Mish Anticline (after Capewell Reference Capewell1951). (b) Structural cross-section through the area of Brandon Bay. For locations of the cross-sections see Figure 4.

Throughout the Irish Variscides there is a single cleavage that is typically axial planar to Variscan folds. In the quartz-rich sandstones and conglomerates of the Slieve Mish Group in the Dingle Peninsula, the cleavage is formed as smooth, disjunctive, variably spaced (a few centimetres) and anastomosing pressure solution seams (Fig. 6). In mudrocks, the cleavage is more tightly spaced (less than 1 cm). The cleavage forms strongly convergent fans about anticlines in the Slieve Mish Group; these fans are symmetrical about the axial plane of the folds (Fig. 7) and are therefore axial planar, sensu lato, with no transection in plan view. Cleavage fans like this are believed to have formed early in the deformation, prior to buckling (e.g. Cooper et al. Reference Cooper, Collins, Ford, Murphy, Trayner, Hutton and Sanderson1984).

Figure 6. Photographs of the Variscan pressure solution cleavage in sandstones of the Slieve Mish and Dingle groups. (a) Curvi-planar anastomosing cleavage on a sandstone bedding plane (e.g. arrows) at Pointnarianna, west of Bull's Head (Fig. 4; V 491 979). The pen for scale is 14 cm long. (b) Spaced planar cleavage (marked by arrow in left of picture) cut by joints and quartz veins (marked by arrow in right) in a sandstone bed at Caher Point, Brandon Bay (notebook for scale is 20 cm long). (c) Spaced planar cleavage cutting quartz veins in a sandstone bed of the Cappagh Formation at Caher Point, Brandon Bay (closed compass clinometer is 10 cm long). Note the ptygmatic folding and offsets of the quartz vein by the cleavage folia, evidence of substantial volume loss. (d) Cliff face at Pointnarianna showing classic unconformity between gently dipping Slieve Mish Group and underlying steeply dipping and overturned Bull's Head Formation (Dingle Group). The cliff is c. 40 m high. Note the channels in the unconformity picked out by the block arrows and the cleavage in the sandstones of the Slieve Mish Group and bedding/cleavage fabric in the Bull's Head Formation picked out by double-headed line arrows. (e) Unconformable contact between base of Slieve Mish Group and Bull's Head Formation at Pointnarianna, viewing underside of bedding plane of sandstone bed at base of Slieve Mish Group. The field of view is c. 5 m wide. Note the trace of closely spaced pressure solution cleavage on the base of the Slieve Mish Group, parallel to the intersection of the combined cleavage/bedding fabric in the Bull's Head Formation. (f) Cross-sectional view of same basal unit (c. 1.8 m thick) of Slieve Mish Group at Pointnarianna as in (e), with cleavage trace on vertical joint/erosion surfaces.

Figure 7. Geological map of the Brandon Bay area to illustrate the Variscan structure of the area. Lower-hemisphere equal-area projections show the variation of bedding and cleavage across macroscopic folds with cleavage forming fans around axial planes, maintaining a high angle to bedding irrespective of position in the fold. For location see Figure 4.

Comparative strain data, collected on a spot basis during field mapping in this research, are presented in Figure 4. A roughly NNW-oriented Z direction with the XY plane vertical is usually indicated by markers such as reduction spots, diagenetic nodules, mudcracks and burrows. A small number of measurements in mudrocks of the Slieve Mish Group suggest that ductile deformation associated with cleavage development has accomplished a strain ratio in the yz plane of the strain ellipsoid, R _s yz, of c. 2.5 in the NNW direction (Fig. 4). This appears to have been accompanied by some vertical stretching so R _s xz ~ 3.75. Bulk strain of Slieve Mish Group sandstones appears to have been more heterogeneous with some shortening achieved by the largely unquantifiable effects of volume loss during pressure solution (but see Fig. 6c for an unusual case where volume loss can be observed). An R _f/Ø analysis (a graphical approach for estimating finite strain of deformed elliptical objects) of clasts in sandstones and conglomerates in the Slieve Mish Group at Bull's Head gave R _s yz values in the range of 1.1–1.3 (Meere & Mulchrone, Reference Meere and Mulchrone2006). This approach tends to provide a minimum estimate of bulk strain, as the process is not homogenous as theoretically assumed (Meere & Mulchrone, Reference Meere and Mulchrone2006). Taenidium barretti burrow tops exposed on the gently SE-dipping Caherbla Group sandstones in the SE of the peninsula have an average long:short axis ratio of 1:0.7 and are always parallel to structural strike, discordant with palaeocurrent directions. They are therefore considered to have been originally subcircular and deformed by tectonic strain (Morrissey et al. Reference Morrissey, Braddy, Dodd, Higgs and Williams2011). A minimum (assuming X is vertical) R _s yz strain for the burrows in these post-Acadian rocks is c. 1.42.

There therefore appears to be a dichotomy between bulk strain between mudrocks and sandstones/conglomerates in the Slieve Mish Group (and other irrefutable post-Acadian groups) which appears to be best explained by assuming that the R _s yz strain in the mudrocks of c. 2.5 is homogeneous and is matched in the sandstones/conglomerates through a heterogeneous combination of strain (c. 1.1–1.4) and volume loss on cleavage folia (Fig. 6c).

Bulk strain (including a cleavage) and layer buckling therefore account for the Variscan deformation of the Slieve Mish Group and other post-Acadian ORS groups. The former apparently accounts for at least twice as much total strain as the latter, which is similar to findings in the main Variscan Belt to the south (Cooper et al. Reference Cooper, Collins, Ford, Murphy, Trayner, Hutton and Sanderson1984, Reference Cooper, Collins, Ford, Murphy, Trayner and O’Sullivan1986; Ford, Reference Ford1987). One implication is that considerable Variscan bulk shortening (probably more than 40%) must have overprinted Acadian and older structures in the older strata. This includes the Variscan cleavage, present in even the most competent, quartz-rich lithologies in the Slieve Mish Group, passing downwards into older pre-Late Devonian rocks (Fig. 6d, e).

4. Late Devonian extension

The effects of Late Devonian extension of the Dingle area are manifest in faults and fault-block rotations that occurred prior to deposition of the latest Devonian Slieve Mish Group (Horne, Reference Horne1977; Todd, Boyd & Dodd, 1988; Todd, Williams & Hancock, Reference Todd, Boyd, Dodd, McMillan, Embry and Glass1988; Todd, Reference Todd, Arthurton, Gutteridge and Nolan1989). For example, several ENE-trending normal faults cut the Caherbla Group, but are overstepped by the Slieve Mish Group. A series of such faults form the northern boundary of the Caherbla Group outcrop and the southern boundary of the Annascaul inlier (Fig. 4). For example, the Caherconree Fault displaces the Caherbla Group by several tens of metres but fails to affect the Slieve Mish Group (Horne, Reference Horne1977). The Minard Head Fault Zone, at the opposite SE end of the outcrop strip, is a Caledonian/Acadian fault zone that was reactivated in a normal down-to-the-south motion (Todd, Reference Todd, Arthurton, Gutteridge and Nolan1989; see Section 8.a). In the northern footwall of the fault, the Caherbla Group was removed by erosion prior to overstep by the Slieve Mish Group. In the hanging wall, the Caherbla Group was tilted northwards before unconformable deposition of the Slieve Mish Group; it now dips less steeply southwards than the Slieve Mish Group on the southern limb of the Slieve Mish anticline. The product of erosion of the Caherbla and older groups of the Dingle Peninsula may have been deposited as three distinct conglomerate units observed in the otherwise finer-grained Late Devonian alluvium of NW Iveragh (Capewell, Reference Capewell1975; Todd, Boyd & Dodd, Reference Todd, Boyd, Dodd, McMillan, Embry and Glass1988; Williams et al. Reference Williams, Bamford, Cooper, Edwards, Ford, Grant, Maccarthy, Mcafee, O'sullivan, Arthurton, Gutteridge and Nolan1989, Reference Williams, Sergeev, Stössel and Ford1997, Reference Williams, Friend and Williams2000; Williams, Reference Williams, Friend and Williams2000).

Similar extensional faulting and fault-block rotations (cf. Jackson & McKenzie, Reference Jackson and McKenzie1983; Barr, Reference Barr, Coward, Dewey and Hancock1987) are believed to have affected the geometry of the post-Dingle Group unconformity-bounded ORS groups in the Feohanagh Anticline (Figs 3, 4, 5; Todd, Boyd & Dodd, Reference Todd, Boyd, Dodd, McMillan, Embry and Glass1988; Todd, Williams & Hancock, Reference Todd, Williams and Hancock1988, Todd, Reference Todd, Arthurton, Gutteridge and Nolan1989). The Smerwick Group crops out only on the northern limb of the anticline and is inverted, dipping steeply towards the SE (Fig. 8). Successive groups lie unconformably on the previous groups, dipping north at progressively lower angles until the Slieve Mish Group oversteps all and extends across the peninsula. Todd, Williams & Hancock (Reference Todd, Williams and Hancock1988) interpreted these geometries as the result of the Feohanagh anticline being developed as a Late Devonian rollover anticline into a NE–SW-trending extensional fault off the current north coast (the North Kerry Lineament; Todd, Reference Todd, Arthurton, Gutteridge and Nolan1989). This rollover or tilted fault-block lay in the regional footwall of the Munster Basin separated by the Dingle Bay Lineament, across which there was more than 5 km displacement during Late Devonian time (Price & Todd, Reference Price and Todd1988; Todd, Reference Todd, Arthurton, Gutteridge and Nolan1989). Successive pulses of extension are believed to have sequentially caused rotation of the pre-existing strata in the Feohanagh Anticline, uplift and erosion of the Dingle regional footwall to the Munster Basin and then subsidence and accommodation of the next group. Note that in Figure 3 the three conglomerate units in Late Devonian NW Iveragh, thought to be products of recycling of the Dingle footwall during uplift, are tentatively time-correlated with the rotational unconformities on the northwest limb of the Feohanagh Anticline (Fig. 8a).

Figure 8. Acadian/Variscan structural style of the Dingle and Dunquin groups. For locations of the cross-sections see Figure 4. (a) Structural cross-section through the Acadian/Variscan structure of the Dingle Group. (b) Structural cross-section through the west coast of the Dingle Peninsula, illustrating the dominantly Acadian structure of the Dunquin Group. (c) Field sketch of the view (looking east) of the east cliff of Coosmore that exposes a cross-section through the faulted core of the Ballyferriter Syncline including chevron folds in the Drom Point Formation (hanging wall) and regularly dipping conglomerates of Glashabeg Formation (footwall). The cliff is c. 30 m high. (d) Lower-hemisphere equal-area stereographic projection depicting structural data for the folds in the Drom Point Formation. Note that the single Variscan cleavage strongly transects the axial planes of these Acadian folds.

5. Acadian deformation of the Dingle Group

The angular unconformity between the Dingle and Slieve Mish groups, particularly at the classic locality of Pointnarianna (V 491 979; Irish National Grid references are provided for key localities not specifically located in figures) west of Bull's Head, has often been cited as the unconformity marking a late Caledonian deformation event associated with the Iapetus closure; this is now interpreted as a distinct event, the Acadian orogeny, that occurred in the Early Devonian some time after the closure of Iapetus (Fig. 6d, e; Shackleton, Reference Shackleton1940; Soper, Webb & Woodcock, Reference Soper, Webb and Woodcock1987; Todd, Reference Todd, Arthurton, Gutteridge and Nolan1989, Reference Todd, Friend and Williams2000). At this location the gently dipping conglomerates and sandstones of the Slieve Mish Group lie unconformably on the Bull's Head Formation (lowermost Dingle Group), inverted and dipping steeply NNW on the southern limb of the Fahan Syncline (Figs 6, 8, 9). The Fahan Syncline, like other macroscopic folds in the Dingle Group, is much tighter than any folds in the Slieve Mish Group (Fig. 8a; compare with Fig. 5). The single cleavage in the Dingle Group is steeper than the axial plane of the Fahan syncline and other folds, and often transects the fold in plan view in an anticlockwise direction (Fig. 10) in contrast to the symmetrical cleavage in folds in the Slieve Mish Group described in Section 3 (see also Figs 5, 7). This observation was first used by Shackleton (Reference Shackleton1940) to argue that the folds in the Dingle Group, including the Fahan syncline, were initiated in the ‘Caledonian orogeny’ (now distinguished as Acadian) and overprinted by Variscan compression. This interpretation is upheld here (but see Meere & Mulchrone, Reference Meere and Mulchrone2006).

Figure 9. Geological map of the western Dingle Peninsula, illustrating the macroscopic Acadian and Variscan folds and faults.

Figure 10. Geological map of the Ballymore area to illustrate the Acadian/Variscan style of the faulted core of the Fahan Syncline. Lower-hemisphere equal-area stereographic projections depict (a) poles to bedding and cleavage and (b) the pole to the axial plane of the Fahan Syncline. Note how the Variscan cleavage has a steeper dip and strikes anticlockwise of the Acadian-initiated Fahan Syncline, and is much more consistent in orientation compared to its development in purely Variscan folds (Fig. 7). For location see Figure 4.

The macroscopic folds in the Dingle Group trend ENE–WSW and have axial planes that dip SSE. The folds verge NNW and short limbs are overturned in that direction (Figs 8a & 9). From south to north across the peninsula, there is a decrease in fold wavelength and amplitude (Fig. 8a). The folds are cut by a number of strike faults, both normal and reverse. Folding and faulting together account for c. 40% shortening in the Dingle Group, estimated by rudimentary line length measurement of unbalanced cross-sections.

The Dingle Group is cut by a single, spaced pressure solution cleavage, which is closely spaced to slaty in lithic arenites and mudrocks. Considerable flattening accompanied the development of the cleavage (Fig. 4; Meere & Mulchrone, Reference Meere and Mulchrone2006). Strain observations in mudrocks such as reduction spots suggest an R _s yz strain of 2.5–3.5 in a NNW–SSE direction (Fig. 4; Meere & Mulchrone, Reference Meere and Mulchrone2006). Similarly to the Slieve Mish mudrocks, there was vertical stretching with an R _s xz strain of c. 3.8 (Meere & Mulchrone, Reference Meere and Mulchrone2006). Strain in sandstones and conglomerates is more heterogeneous, involving a cleavage partitioned to the matrix with volume loss difficult to quantify and vertical stretching evidenced by mica beards on sandstone clasts and ladder veins (Meere & Mulchrone, Reference Meere and Mulchrone2006, fig. 4). The comparison of strain above and below the Acadian unconformity is fraught with uncertainties, including: debate on the level of the Acadian unconformity in the north of the peninsula (see above); paucity of mudrocks in the irrefutable post-Acadian ORS in comparison with mudrocks in the Dingle Group; contrasting sandstone/conglomerate lithologies ranging from lithic wacke sandstones and petromict conglomerates (with labile clasts) in the Dingle Group to quartz arenite sandstones and oligomictic conglomerates (with indurate clasts) in the Slieve Mish Group; and the effect of different bedding orientations and structural locations. A compilation of the available quantitative data indicates that the Dingle Group was considerably shortened by folding and faulting during the Acadian event and these folds were probably tightened further by Variscan bulk shortening (Table 1). Bulk strain appears to be mostly Variscan in origin with, leaving aside lithological differences and volume loss estimates, R _s yz strain differences between pre-Acadian and post-Acadian strata ranging from 12% in mudrocks to 25% in sandstones/conglomerates. The Variscan cleavage is therefore axial planar (in fans) to folds in the Slieve Mish Group, forming early in the Variscan deformation prior to buckling, but is non-axial planar and superimposed on Dingle Group folds to transect axial planes in both plan and cross-section.

Table 1. Summary of strain data illustrating shortening of the Dingle Group (pre-Acadian) and Caherbla, Pointagare and Slieve Mish groups (post-Acadian) in a NNW–SSE direction, approximately parallel to the YZ plane.

6. Acadian deformation of the Dunquin Group

There is also evidence that some Acadian structures began to develop through deposition of the Dingle Group. For example, in the northwest of the peninsula at coastal exposures in Coosmore, Coosgorrib and Coosglass, the Dunquin and Dingle Group succession is attenuated compared to thicker sequences to the south (Fig. 9; Todd, Williams & Hancock, Reference Todd, Williams and Hancock1988). The thinned Dingle Group sequence is composed of conglomerates of the Glashabeg Formation and unconformably oversteps onto older Dunquin Group strata in a north and west direction (Fig. 9; Todd, Williams & Hancock, Reference Todd, Williams and Hancock1988). This same area is bounded to the east by the largely inferred NNE-trending Ballyicken Fault and to the south by the Dunquin Fault (Figs 4, 9; Todd, Williams & Hancock, Reference Todd, Williams and Hancock1988). The macroscopic folds of this domain have an orientation and morphology that contrasts with the folds in the Dingle Group to the east (Figs 4, 8b, 9). The folds here have a WNW orientation, are tighter than folds in the Dingle Group and are markedly discordant to the single ubiquitous cleavage (Fig. 8b–d). From the angular unconformity and the anomalous folds in the Dunquin Group, it is inferred that this area was being deformed and uplifted in early Dingle Group times; the ancestral Ballyickeen and Dunquin Faults form the structural boundaries, dividing this area from the main depocentres of the Dingle Basin to the south and east where the much thicker Dingle Group follows conformably from the Dunquin Group. These basin margin effects are interpreted to have caused the coarser influxes of sand and paraconglomerate into the lacustrine Bull's Head Formation and recycling of Dunquin Group lithotypes into the Dingle Group (Todd, Boyd & Dodd, Reference Todd, Boyd, Dodd, McMillan, Embry and Glass1988; Boyd & Sloan, Reference Boyd, Sloan, Friend and Williams2000; Todd, Reference Todd, Friend and Williams2000). The Dingle Group conglomerate formations contain sedimentary evidence of differential subsidence, episodic fluvial plain tilting and the strike-lengthening of a major fan lithosome by syn-sedimentary strike-slip displacement of the feeder system (Todd, Boyd & Dodd, Reference Todd, Boyd, Dodd, McMillan, Embry and Glass1988; Todd, Reference Todd, Arthurton, Gutteridge and Nolan1989; Todd & Went, Reference Todd and Went1991).

A picture of ongoing late Silurian – early Devonian deformation, subsidence and finally inversion of the Dingle Basin to produce the Acadian structure emerges from the combination of stratigraphic and structural observations. The contemporaneous development of compressional structures adjacent to areas of rapid subsidence suggests a transtensional/transpressive regime (Todd, Boyd & Dodd, Reference Todd, Boyd, Dodd, McMillan, Embry and Glass1988; Todd, Reference Todd, Arthurton, Gutteridge and Nolan1989, Reference Todd, Friend and Williams2000; Boyd & Sloan, Reference Boyd, Sloan, Friend and Williams2000).

7. Caledonian/Acadian deformation of the Annascaul Formation

The early Ordovician Annascaul Formation provides evidence of Caledonian deformation that is pre-Wenlock in age, as well as bearing the overprints of Acadian deformation, Devonian extension and Variscan compression. There are at least three cleavages observed together in the Annascaul Formation, and the presence of a fourth is inferred from collation of structural styles and local overprinting relationships (Figs 11, 12).

Figure 11. (a) Geological map of the Caherconree area. Lower-hemisphere equal-area projections depicting structural attributes of the (b) Annascaul Formation and (c) Dunquin Group.

Figure 12. Geological map of Minard Bay, with insets of lower-hemisphere equal-area projections depicting orientations of the structures in the Annascaul Formation and Bull's Head Formation. Middle and Upper ORS refers to the Caherbla and Slieve Mish groups, respectively.

Annascaul Formation mudrocks contain a bedding-parallel slaty cleavage labelled S* (after the definition of Powell & Rickard, Reference Powell and Rickard1985), deformed by at least two, possibly three, later cleavages. S* is developed by the preferred dimensional orientation (PDO; Elliott & Williams, Reference Elliott and Williams1988) of illite and chlorite grains. The fabric post-dates, but mimetically overgrows, a depositional fabric of illite and chlorite grains. The fabric is very typically parallel to bedding, occasionally divergent (<15°). S* augens competent clasts in the mélange units, but passes through shale clasts. Early slump folds (Parkin, Reference Parkin1976) are also cut by S*. S* is interpreted to be the product of burial metamorphism and is also observed in the corollary parts of the Ordovician Ribband Group in east and southeast Ireland (Shannon, Reference Shannon, Harris, Holland and Leake1979; Murphy, Reference Murphy1987).

F₁ folds are distinguished in Minard Bay, the best exposure of the Annascaul Formation structure, as tight to isoclinal Class 1C folds that have upright axial planes and axes that plunge SW or NE (Fig. 12). One macroscopic F₁ fold is mapped in Foilaniska where an upright synclinal core is exposed (Fig. 12). South of the fold hinge, later superimposed F₂ folds are downward facing. The F₁ fold is overturned to the NW and has a wavelength greater than 1.6 km. S* is crenulated in the hinges of F₁ folds where it is cut by S₁, which is axial planar to F₁ and forms a disjunctive cleavage with some crenulation of S*. On the limbs of F₁ folds a very strong phyllitic, papery fabric parallels bedding with S₁ parallel to and enhancing S*, forming disjunctive pressure solution folia.

Annascaul D₂ structures are much more widespread and easy to identify than D₁. F₂ folds are steeply inclined to upright, steeply to moderately plunging and tight to isoclinal. In pelites, the folds have a Class 1C profile; isolated sandstone units form Class 1B folds with rounded hinges; packages of interbedded greywacke and shale form chevron folds. These folds are parasitic to larger macroscopic F₂ folds and have wavelengths of 1–6 m. The macroscopic antiforms have wavelengths of 20–300 m (Fig. 12).

An S₂ cleavage, approximately axial planar to F₂ folds, is generally restricted to (or at least discernible in) the hinge zones of F₂ folds or the intensely deformed D₂ shear zones described in more detail in the following section. S₂ tends to be parallel to the axial surfaces of F₂ folds, but slightly transects the fold hinge in an anticlockwise sense in some places.

In the Kilteenban–Caherconree area (Fig. 11), S* and S₂ are generally not parallel and S* has an unusual orientation dipping NE. F₂ fold axes plunge east or west rather than northeast or southwest as in Minard Bay. S₂ also has an atypical morphology and orientation. In some places S₂ appears as a conjugate crenulation cleavage and is overprinted by a second crenulation cleavage, S₃.

8. Structure of fault zones

The structure of faults provides enhanced evidence of the nature and timing of deformation. The effects of Acadian faulting are most prominent in the southeast of the Dingle Peninsula and die out to the northwest.

8.a. Minard Head Fault Zone

One of the best-exposed fault zones in the area occurs on the northern side of Minard Head as an outcrop strip c. 40 m wide (Fig. 12). This fault zone was referred to as the Caherconree Fault Zone by Todd (Reference Todd, Arthurton, Gutteridge and Nolan1989), following Horne (Reference Horne1970, Reference Horne1974, Reference Horne1977). Subsequent mapping has shown the Caherconree Fault at Caherconree is however a separate structure to that at Minard Head; the latter was therefore renamed the Minard Head Fault Zone (MHFZ). Rocks of the Annascaul and Bull's Head Formations are both deformed in the zone. To the southeast, on the southern side of Minard Head, the Inch Conglomerate Formation (Caherbla Group) is largely unaffected by the fault zone except for some simple normal faulting (Fig. 12).

Most discrete faults and fault-rock fabrics within the zone dip steeply SE. In the footwall, the Annascaul Formation is a mélange of black shale matrix and clasts of greywacke, shale, vein quartz, dolomite and brecciated lava.

Three zones with irregular and gradational boundaries can be recognized in the deformed footwall rocks of the Annascaul Formation (Fig. 13). In Zone 1, the main fault fabric is an intensification of the S₂ cleavage in the Annascaul Formation outside the fault zone. Within the mélange, S₂ crenulates S* which is present in both the matrix and clasts, particularly shale clasts; more competent blocks such as greywacke clasts are augened by S*. In some places, blocks containing S* are cut or augened by S₂, indicating some cataclasis and formation of clasts post-S* but pre-S₂.

Figure 13. Fault rocks in the Minard Head Fault Zone and at Farranacarriga. (a) MHFZ Zone 1 rocks: plan view of mélange/cataclasite in Annascaul Formation with penetrative S₂ cleavage (the compass clinometer is 10 cm long). (b) MHFZ Zone 3 rocks: plan view of foliated cataclasite in Annascaul Formation protolith adjacent to the principal fault plane marked by quartz-fibre sheets at the top of the photograph (compass-clinometer is 18 cm long). (c) Slabbed, oriented specimen of Zone 1 fault rocks of the MHFZ cut parallel to the main fault fabric. Note the equidimensional shapes of the clasts. (d) Same sample as (c), but cut perpendicular to fault fabric. (e) Photomicrograph of the Zone 1 rocks in the MHFZ. The section is cut in the horizontal plane, perpendicular to the main fabric (S₂) picked out by haematite-rich cleavage seams. A large porphyroclast in the centre of the field of view displays asymmetric pressure-shadow tails of comminuted grains, the asymmetry indicative of sinistral shear as marked by the half arrows. (f) MHFZ Zone 1 cataclasite foliated by S₂ cut by a later S₃ cleavage. Both (e) and (f) photographs taken in plane-polarized light. (g) Plan view photomicrograph of the Farranacarriga Fault rocks developed in the Annascaul Formation showing boudins if phyllonite within the principal fault fabric are bounded by small shear planes. The anticlockwise rotation of the boudins indicates sinistral shear. (h) Plan view photomicrograph of the Farranacarriga Fault rocks developed in the Annascaul Formation showing a quartz ribbon mylonite in which the light areas are quartz with a little feldspar and the dark areas are graphitic phyllonite. In both (g) and (h), the photographs were taken in plane-polarized light and the scale bar is 0.1 mm long.

The rocks are cut by numerous discrete faults, some of which are mineralized with quartz and sulphide minerals. In plan most steep faults trend anticlockwise of the S₂ cleavage, which is deflected into the faults; some faults lie clockwise of the cleavage (Fig. 13). The dominance of the anticlockwise relationship between cleavage and shear planes indicates a predominant component of sinistral shear in the zone. Although the majority of the individual strike-slip faults in the MHFZ are sinistral, some are dextral. Sinistral faults are aligned mainly about 060°, whereas dextral faults are aligned about 080° (Fig. 13; Todd, Reference Todd, Arthurton, Gutteridge and Nolan1989). Similar preferred orientations are observed elsewhere in the Annascaul inlier, outside the MHFZ (Fig. 14; Todd, Reference Todd, Arthurton, Gutteridge and Nolan1989).

Figure 14. Histogram of the frequency of occurrence of mesoscopic fault strikes in the Annascaul inlier. Vertical axis in each histogram is the strike angle in degrees.

Other mesoscopic indicators of shear sense in Zone 1 include porphyroclasts that range in diameter from a few millimetres to 10 m, but are generally less than 50 mm long. These clasts, and the clasts of sedimentary origin, are highly flattened parallel to subparallel to the S₂ fault fabric (Fig. 13). The porphyroclasts commonly have long axes that are oriented at a slight, oblique angle to the S₂ fault fabric (Fig. 13). Some of the augened clasts possess asymmetric pressure-shadow tails which in plan are mostly clockwise of the fabric, indicating back-rotation of the clast in sinistral transpressive shear (e.g. Lister & Williams, Reference Lister and Williams1979; Lister & Snoke, Reference Lister and Snoke1984). In cross-sections perpendicular to the fault fabric, similar asymmetric tails indicate down-to-the-north shear. Clasts in the mélange are flattened and in agreement with strain indicators outside the fault zone, indicating an approximately NE–SW-oriented Z direction with Y horizontal and X vertical. From these indicators it is concluded that shortening across the zone was more important than simple shear (Sanderson & Marchini, Reference Sanderson and Marchini1984).

Field observations of Zone 1 fault rocks are supported by microscopic analysis. S₂ forms a pervasive crenulation cleavage that in Zone 1 achieves transposition of S*. Some of the phyllitic clasts still contain S* which has been typically rotated clockwise of the S₂ foliation, again indicative of sinistral shear (Fig. 13e, f). Other cataclasized arenite clasts are rounded to sub-angular and often contain tails of comminuted material; some of the tails are obliquely oriented to the fault fabric. Thin veins (c. 100 μm thick) of recrystallized mosaic quartz occur in some quartz and arenite clasts. These veins, plus the undulose extinction of quartz with the clasts and comminuted grains (cf. Watts & Williams, Reference Watts and Williams1979; Anderson & Oliver, Reference Anderson and Oliver1986) indicate limited quartz deformation followed by grain recovery by recrystallization (Sibson, Reference Sibson1977; Wise et al. Reference Wise, Dunn, Engelder, Geiser, Hatcher, Kish, Odom and Schamel1984). Nevertheless, the dominance of brittle cataclastic processes over plastic deformation suggests that these rocks should be termed foliated cataclasites (Chester, Friedman & Logan, Reference Chester, Friedman and Logan1985).

In many places in Zone 1 of the MHFZ S₂ is cut by later cleavages. A gently dipping crenulation cleavage is apparently of minor significance. Another spaced pressure solution cleavage occurs more widely, and is correlated on the basis of geometry and form with S₃ observed in the Annascaul Formation outside the fault zone and with the main cleavage observed in the Bull's Head Formation in the hanging wall of the MHFZ (Figs 12, 13f).

Zone 2 fault rocks form elongate lenses a few metres wide. Although they are not necessarily more strained than Zone 1, Zone 2 fault rocks are distinctive in being finer grained and possessing abundant quartz veins. They are possibly the product of deformation of fine-grained black shales without abundant clasts. The quartz veins display a variety of geometries. Simple lenticular extension veins lie at a variety of angles to the S₂ cleavage. The veins may be oriented clockwise, suggesting dextral shear, or anticlockwise to the fabric, suggesting sinistral with the latter predominating. Some quartz veins are folded into tight to isoclinal folds with variable vergence. Others are pulled apart or boudined to form isolated vein quartz clasts. These different types of quartz vein may be the products of the same progressive simple transpressive shear.

Zone 3 fault rocks of the MHFZ are also derived from the Annascaul Formation and occupy a band 3–4 m wide adjacent to the principal fault plane (Fig. 13b). The yellow phyllitic rocks are intensely foliated and contain numerous shear planes. The foliation is domainal and anastomosing. Some resistant competent blocks of sandstones and conglomerate persist, but generally the grain size is small.

The principal fault plane of the MHFZ, separating Zone 3 rocks developed from the Annascaul Formation from Zone 4 rocks developed from the Bull's Head Formation (Dingle Group), is marked by a composite series of quartz-fibre sheets (Fig. 13b). These sheets cut all the previously described fault fabrics on the Annascaul Formation. The sheets generally dip SE and bear lineations indicative of oblique slip. The risers of the accretionary steps on the lineated quartz slickenfibre surfaces face downwards on the footwall indicating normal, down-to-the-south displacement combined with a subordinate component of dextral strike-slip. The quartz-fibre sheets are laterally extensive and largely undeformed, although in some places are cut and displaced by steep reverse faults that dip more steeply SE, are buckled into gentle inclined folds or are boudined.

Two zones may be defined in the Bull's Head Formation deformed within the MHFZ. Zone 4 is c. 1–2 m wide and lies parallel to the principal fault plane. It is characterized by an intensification of the brecciation, shearing and cleavage that occurs less commonly in the less-deformed rocks of Zone 5 to the east. Thinly bedded sandstones of the Bull's Head Formation, their usual purple hue altered to a bright green colour, have been sheared and boudined. The product is a cataclasite containing phacoids of resistant sandstone in a finely brecciated, foliated matrix.

A few metres southeast of the principal fault plane, cataclasis is reduced and Zone 5 is characterized by thinly bedded purple sandstones and siltstones of the Bull's Head Formation deformed by mesoscopic folds and faults (Fig. 12). Only limited cataclasis has occurred on some of the fault planes. The folds have either tight chevron-shaped profiles with angular hinges, or more open Class 1B profiles with rounded hinges. The folds have wavelengths of c. 5–15 m. They are variably oriented; tighter folds tend to be more steeply plunging and moderately to steeply inclined while open folds are generally more upright. The variability of the folds is taken as evidence of two phases of folding (Fig. 12; Todd, Reference Todd, Arthurton, Gutteridge and Nolan1989).

A ubiquitous cleavage cross-cuts most of the inclined tight folds and is axial planar only to the open upright folds (Fig. 12) and is interpreted to be Variscan in age (see further discussion in Section 9.a). In one place on the north side of the Minard Head sandstones of the Bull's Head Formation contain two cleavages: (1) a closely spaced cleavage trending NE–SW and dipping moderately SE; and (2) a spaced cleavage cuts the NE–SW cleavage, trends ENE and dips steeply SSE. This is one of the few places where the Dingle Group contains more than one cleavage.

Numerous faults cut the folds. Some of them are accompanied by brecciation, but most are sharp, unmineralized planar and curvi-planar shear planes. A silt matrix supports angular arenite clasts up to 5 cm long in the breccia bands. A cleavage is commonly developed in the matrix of the breccias and, in some places, this fabric is deflected asymptotically into the walls of the fault-breccia band. In several places, particularly in the region of the southwest promontory of Minard Head, faults bound packages of little-deformed sandstone beds. Bedding within these packages or duplexes (Woodcock & Fischer, Reference Woodcock and Fischer1986) dips steeply and is often deflected asymptotically from the centre of the package towards the bounding fault (see fig. 10 of Todd, Reference Todd, Arthurton, Gutteridge and Nolan1989). The faults trend approximately NE–SW and sinistral offset is most common where offset is demonstrable or where kinematic indicators such as deflected cleavage or bedding are present (Fig. 13). The sense of shear on the faults is concomitant with shearing of the short limbs of the folds, and the shearing and folding appears to have been contemporaneous. In some places fold hinges are completely cut out; abrupt changes in younging of the beds therefore occurs across the faults that bound these duplexes.

The east–west trend of the axial planes of the earlier folds in the Bull's Head Formation represents north–south shortening which is aligned anticlockwise of the NE-trending faults. This arrangement would be expected from the development of the folds and faults contemporaneously in a regime of sinistral transpression (Sanderson & Marchini, Reference Sanderson and Marchini1984; Todd, Reference Todd, Arthurton, Gutteridge and Nolan1989).

8.f. Caheracruttia

A very small inlier of the Annascaul Formation crops out within the Inch Conglomerate Formation of the Caherbla Group and is exposed in the stream that flows through the townland of Caheracruttia (Fig. 4; Q 726 067). The inlier forms a strike-parallel strip enveloped by breccio-conglomerates of the Inch Conglomerate Formation. The Annascaul Formation is here represented by green, glossy phyllonites with a well-developed penetrative slaty cleavage. Locally, patches of cataclasites with clasts of phyllite are developed. There are abundant isolated quartz pods interspersed in a matrix of comminuted clasts of phyllite (Fig. 15a). In plan view the long axes of quartz clasts trend at an acute angle, anticlockwise of the cleavage. The cleavage is cut and deflected by shear planes that also indicate sinistral shear. In vertical cross-section the shear fabrics indicate down-to-the-north displacement. In common with the vast majority of the fault rocks of the southeast Dingle Peninsula, these phyllonites developed in sinistral transpression.

Figure 15. Fault rocks associated with the Inch Conglomerate Formation. (a) Slabbed specimen of cataclasite developed in the Annascaul Formation exposed in the stream bed at Caheracruttia. (b) Slabbed specimen of foliated breccio-conglomerate of the Inch Conglomerate Formation (lithosome 2 of Todd Reference Todd, Friend and Williams2000, fig. 15) that directly overlies the cataclasites illustrated in (a). Photomicrographs, taken in cross-polarized light, of two clasts of mylonite from lithosome 1 (Todd, Reference Todd, Friend and Williams2000) showing the spectrum of mylonitization of schist/gneiss protoliths including (c) quartz ribbon mylonite and (d) ultramylonite. The slabs in (a, b) are c. 15 cm long and the scale bars in (c, d) are 0.1 mm long.

A distance 10 m north of the exposure of the Annascaul Formation, blocks protrude from the banks of the Caheracruttia stream that are formed of foliated breccia of the Inch Conglomerate Formation. The breccia is composed of clasts of phyllite and quartz set in red sandy matrix and the foliation is interpreted to be sedimentary (Fig. 15b). The clasts are identical to the lithologies that occur in the cataclasites below. The breccia may be interpreted as a talus deposit, reworked directly from the underlying fault rocks (Capewell, Reference Capewell1951).

9. Nature and timing of deformation episodes

Table 2 summarizes the timing and effects of the deformation episodes that affected the Palaeozoic succession of the Dingle Peninsula. Four aspects of this deduced tectonic sequence merit additional discussion.

Table 2. Structural history of the Palaeozoic rocks of the Dingle Peninsula.

9.c. Partitioned Acadian transpression/transtension

The broad zonation of the Acadian structure of the Dingle Peninsula invites comparison with that of a flower structure (Harding, Reference Harding1985) or a strain-partitioned transpression zone (Fig. 16; Dewey, Holdsworth & Strachan, Reference Dewey, Holdsworth, Strachan, Holdsworth, Strachan and Dewey1998; Dewey & Strachan, Reference Dewey and Strachan2003). During Dingle Basin subsidence structural and stratigraphic evidence indicates coeval strike-slip along the major bounding lineaments and intra-basinal effects such as localized folding, tilting and faulting of the Dunquin Group and older rocks. Similarly, the sinistral transpressive inversion of the Dingle Basin appears to have caused folding and faulting of the basin fill with strain apparently intensifying closer to the Dingle Bay Lineament, manifest by fault rocks in the Annascaul inlier. Although not currently exposed, the Dingle Bay Lineament is considered to represent the main fault segment of the zone. The model predicts simple shear strain increasing towards, or distinctly partitioned in, the main fault segment (Dewey, Holdsworth & Strachan, Reference Dewey, Holdsworth, Strachan, Holdsworth, Strachan and Dewey1998). Dingle Group and older rocks southeast of the southern limb of the Fahan Syncline are therefore typically subvertical, often overturned and involved in sinistral transpressive fault zones such as the MHFZ. North of the Fahan Syncline the fold belt which shallows northwards, affecting both Dingle and Dunquin groups, is interpreted as the ‘petals’ of the flower; it is predicted to accommodate more compression and reverse slip across thrusts, perhaps with a subordinate amount of strike-slip.

Figure 16. Early Devonian transtensional then transpressive development of the Acadian structure of the Dingle Pensinsula. (a) Schematic plan showing structural and sedimentation patterns in the Dingle Basin during early Emsian time. While the patterns observed could be due to very oblique sinistral transpression, a slightly oblique divergence in sinistral transtension is favoured. (b) Schematic plan showing structural pattern of the Dingle Basin in late Emsian time after inversion by sinistral transpression partitioned across the basin with folding and faulting in the prior basin area and sinistral shear and flattening in the steep zone towards the Dingle Bay Lineament.

9.d. Late Devonian syn-rift

The observed pattern of Late Devonian normal faulting affecting Caherbla Group and older rocks, then reversed and overprinted by Variscan shortening through folding, faulting and bulk strain, is consistent with the models of Price & Todd (Reference Price and Todd1988) and Todd (Reference Todd, Arthurton, Gutteridge and Nolan1989) invoking the reuse of Caledonian (and Acadian) structures in the development of the Munster Basin framework (Fig. 17). Pre-existing faults such as the Dingle Bay Lineament and synthetic faults such as the MHFZ were reactivated as largely normal faults to form the Late Devonian Munster Basin, perhaps with a small amount of dextral strike-slip as is observed on the MHFZ (Todd, Reference Todd, Arthurton, Gutteridge and Nolan1989).

Figure 17. Regional chronostratigraphy of the Iapetus suture zone in Ireland showing the Caledonian and Acadian deformation history and linkages between the main terranes or blocks. Updated from Todd (Reference Todd, Friend and Williams2000) with correlation with the timing data compiled by Soper & Woodcock (Reference Soper and Woodcock2003) and modification of the time scale to that of Gradstein, Ogg & Schmitz (Reference Gradstein, Ogg and Schmitz2012).

Williams (Reference Williams, Friend and Williams2000) presented a detailed modelling analysis of the subsidence history of the Munster Basin. He concluded that extension was accommodated through displacement on steep crustal faults, such as the Dingle Bay Lineament. Deformation of the regional footwall of the Dingle area, north of the Dingle Bay Lineament, was modelled to include the formation of a ‘rim-type’ basin in which the Late Devonian Carrigduff Group was preserved. Williams (Reference Williams, Friend and Williams2000) only included the Carrigduff Group, preferring the interpretation of Richmond & Williams (Reference Richmond, Williams, Friend and Williams2000) that the Smerwick and Pointagare Groups are Early–Middle Devonian in age and ‘pre-rift’ to the Munster Basin. The alternative interpretation that all the post-Dingle Group succession of the Feohanagh Anticline is age-equivalent to the Munster Basin is upheld here as at least equally permissible, if not more consistent, with the available evidence (Price & Todd, Reference Price and Todd1988; Todd, Boyd & Dodd, Reference Todd, Boyd, Dodd, McMillan, Embry and Glass1988; Todd, Williams & Hancock, Reference Todd, Williams and Hancock1988; Todd, Reference Todd, Arthurton, Gutteridge and Nolan1989). This model sees the deposition and subsequent tilting of successive unconformity-bounded syn-rift groups in a punctuated fashion through progressive rotations of a half-graben perched on the regional footwall of the Munster Basin (Fig. 17).

10. Tectonic evolution of Dingle Peninsula in a regional context

The oldest rocks exposed in Dingle, the Early Ordovician Annascaul Formation, have been correlated with the Ribband Group in SE Ireland and the Skiddaw Group in the Lake District (Figs 2, 17; Todd et al. Reference Todd, Connery, Higgs and Murphy2000). As such, a palaeogeographic position on the southern margin of the Iapetus Ocean seems very likely as part of Eastern Avalonia (see Prigmore, Butler & Woodcock, Reference Prigmore, Butler and Woodcock1997). It is generally agreed that the Iapetus itself was in existence during Cambrian time, although there is some debate on when Avalonia separated from Gondwana. Some favour a late Cambrian or Tremadocian separation, influenced by the Tremadocian onset of arc volcanism. Others, influenced more by palaeomagnetic and faunal data, infer a Llanvirn or early Caradocian separation. In either case, the deep-water facies of the Annascaul Formation, including evidence of slope instability to produce slumps and mélange deposits with contemporaneous volcanism, is consistent with both a fore-arc and back-arc ocean margin.

D₁ in the Annascaul Formation, deforming the bedding parallel S* cleavage believed to be associated with deep burial metamorphism, is difficult to date. Correlation with similar early structures in the Ribband Group of east and southeast Ireland (Shannon, Reference Shannon, Harris, Holland and Leake1979; Murphy, Reference Murphy1987) suggests a pre-Caradocian age. If this is correct, then perhaps the stress associated with D₁ is related to subduction either in a fore-arc setting (accretionary processes?) or through inversion of a prior back-arc as subduction trajectories changed. Late Ordovician rocks have not been preserved in Dingle, but in SE Ireland the period is dominated by the Duncannon Group volcanic rocks representing arc volcanism through continued southwards subduction of Iapetus under Avalonia.

It is now generally accepted that Avalonia had accreted to Laurentia by middle Silurian time and that the Silurian succession in the Lake District and central Ireland represents infilling of the vestigial seaway (e.g. Woodcock, Soper & Strachan, Reference Woodcock, Soper and Strachan2007). The development of Wenlock subduction-related volcanic rocks (Dunquin Group) in western Dingle has been related to a final burst of Iapetus-related volcanism with a vestigial re-entrant of Iapetus oceanic crust being the preferred explanation for these arc-signature volcanic rocks (Sloan & Bennett, Reference Sloan and Bennett1990; Boyd & Sloan, Reference Boyd, Sloan, Friend and Williams2000). The collision is generally considered to be ‘soft’ within the Iapetus suture zone of Britain and Ireland, with continuous late Silurian subsidence recorded in many places (Soper & Woodcock, Reference Soper and Woodcock2003). There is evidence of Silurian deformation in South Mayo (Dewey & Strachan, Reference Dewey and Strachan2003), in Anglesey (Treagus, Treagus & Woodcock, Reference Treagus, Treagus and Woodcock2011) and in Rosslare (Murphy, Reference Murphy, D’Lemos, Srtachan and Topley1990). Such an episode might also be recorded in the main metamorphism of the schists and gneisses preserved as clasts in the Inch Conglomerate (Fig. 17; Todd, Reference Todd, Friend and Williams2000).

The end of volcanicity in Dingle was followed a transition from marine to non-marine sedimentation and increasingly apparent evidence of syn-depositional tectonics attributed a strike-slip regime (Todd, Boyd & Dodd, Reference Todd, Boyd, Dodd, McMillan, Embry and Glass1988; Boyd & Sloan, Reference Boyd, Sloan, Friend and Williams2000). The patterns of conformable-unconformable boundary variation, stratigraphic thickness change, sedimentary facies and dispersal patterns of the Dingle Group are strongly evidential of a sinistral transtensional/transpressional regime (Fig. 16a; Todd, Reference Todd, Arthurton, Gutteridge and Nolan1989). The Dingle Basin was filled by lakes, alluvial fans and rivers sourced from rising hinterlands to the north and south with geological signatures similar to the terranes now north and south of the Iapetus Suture (Fig. 17; Todd, Boyd & Dodd, Reference Todd, Boyd, Dodd, McMillan, Embry and Glass1988; Todd, Reference Todd, Friend and Williams2000). The Dingle Group was then itself deformed through folds, faults and some penetrative strain. This period of inversion correlates very well with the late Emsian cleavage-forming event in the Acadian slate belts of SE Ireland and Lake District, which was synchronous with granites dated at c. 400 Ma (Fig. 17; Soper & Woodcock, Reference Soper and Woodcock2003). A single Rb–Sr mica-whole-rock age of 398±5 Ma from a clast from the Inch Conglomerate Formation might similarly reflect this event in the basement of the Dingle Bay Lineament (Fig. 17; Todd, Reference Todd, Friend and Williams2000).

It is therefore possible to envisage ongoing transtensional/transpressive strain in the region of the Dingle Bay Lineament throughout Dingle Group times (Todd, Reference Todd, Arthurton, Gutteridge and Nolan1989). The fault zones described above are a window into that major fault zone and display a variety of shortening structures (folds and cleavage), forming together but at different angles with sinistral and dextral strike-slip faults and both normal and reverse faults. It is possible that basin subsidence through pull-apart and/or flexural loading continued synchronously with shortening and uplift of neighbouring blocks, partitioned into or bounded by zones of pre-existing weakness such as the Dingle Bay Lineament (Fig. 16a). Given the local evidence, there is little need to invoke a significant change in plate convergence to accomplish the observed tectonic sequence. For example, the change from a regime of differential subsidence and uplift described for Dingle Group times (Fig. 16a) to general shortening and uplift (Fig. 16b) could be accomplished by a modest rotation of convergence across the 060°-oriented zone from very oblique to less oblique convergence, or by the strike movement of the Dingle block along the Dingle Bay Lineament from releasing to restraining bends (see also Dewey, Reference Dewey2002).

However, in addition to the Acadian orogenic episode being well established as a late Emsian event involving transpression across the slate belts south of the Iapetus Suture, there is now growing evidence of a preceding transtensional period in the latest Silurian – Early Devonian (Dewey & Strachan, Reference Dewey and Strachan2003; Soper & Woodcock, Reference Soper and Woodcock2003). This regional transtension reflects large strike-slip displacements inferred for many of the major faults (e.g. Great Glen Fault; Dewey & Strachan, Reference Dewey and Strachan2003), provides a driver for subsidence and accommodation space for ORS sedimentation (Dewey & Strachan, Reference Dewey and Strachan2003; Soper & Woodcock, Reference Soper and Woodcock2003) and enables the intrusion of the lamprophyre suites and ultimately the newer granite suite across the Iapetus suture zone (Brown et al. Reference Brown, Ryan, Soper and Woodcock2008). It is therefore simplest to believe that the Dingle Basin change from differential subsidence/uplift during late Silurian – early Emsian time to general shortening in late Emsian time also reflects the same regional tectonic climate change from largely transtensional to largely transpressional. Woodcock, Soper & Strachan (Reference Woodcock, Soper and Strachan2007) suggest either a change of subduction trajectory or that the collision of a continental fragment in the Rheic subduction zone induced the stress to generate the slate belts in the fossil Iapetus suture zone (Fig. 17).

A similarly profound change in tectonic style must have occurred to cause the onset of Munster Basin extension that probably began during Givetian time or perhaps earlier during Eifelian time. Leeder (Reference Leeder1982) suggested that slab-pull from the Rheic subduction zone caused extension in many upper Palaeozoic basins in Britain. A geotectonic picture therefore emerges for the fossil Iapetus suture zone, in which changes from to transpression to extension are generated by changes in Rheic subduction styles to the south, culminating in the closure of the Rheic with collision of Gondwana and the development of the Irish Variscides.