ABSTRACT: The abundance of volcanic rocks found in theOrdovician period of Wales suggests widespread volcanism. A variation ingeochemical compositions, mainly basalt-andesite-rhyodacite typical of modern volcanic arcsfound in the Rhobell Volcanic Complex, and ocean floor basalts found in theFishguard Volcanic Complex of Wales suggest two distinct forms of volcanismgenerally agreed upon in the literature. The causes of these two distinct typesof volcanism, arc and marginal basin volcanism, have been the subject ofongoing debates.
Arc volcanism is thought to have occurred in late Tremadoc-early Arenig times with a southerly subductionof the Iapetus Ocean beneath Avalonian margins. Paleomagnetic data backed up byfaunal evidence place the rifting of Avalonia from Gondwana and its northwardmigration in Arenig-Llanvirn times, which raises the question ofwhether rifting was involved in the onset of arc volcanism. Predominantlymarine stratigraphic samples from Arenig times suggest a period of volcanicquiescence with possible marine transgression. Geochemical data from magmasamples of mid to late Ordovician in the Fishguard Volcanic Complex point outto a back-arc spreading, but the location of this back-arc in Wales still remains unclear. To conclude, rifting of Avalonia-Gondwana might have happened in two-stages with different time intervals and theproposed location of the back-arc near Lake District is tectonically notpossible and must have been closer to Wales. 1.INTRODUCTIONThegeological evolution of Britain during the Ordovician is characterised byperiods of intense volcanic and tectonic activities. In Wales, the rocksparticularly record the evolution of arc volcanism and a subsequent transitionto marginal basin volcanism.
Paleogeographic reconstructions (Fig. 2) place the position of Avalonia,Gondwana as well as other continental fragments at Southern Hemispherelatitudes during the early Ordovician (Pickeringand Smith, 1995; Torsvik et al.,1996).
Throughout the Ordovician (Fig 1) Wales was part of Avalonia, amicrocontinent which was located on the northern margin of supercontinent Gondwana.However,Avalonia did not remain connected to Gondwana for much longer and instead mayhave migrated northward after rifting. A time range for this separation at theAvalonia-Gondwanan margin has been devised through the use of faunal contrastsbetween continents by Cocks and Fortey.(1990) and paleomagnetic data by the likes of Torsvik et al.
(1996) andChannell et al. (1992). Although this rifting and northward migration hasgenerally been agreed upon in the literature, certain details are still beingdebated. Aperiod of arc volcanism dominated the early Ordovician of Wales as proposed by Kokelaar (1979) and Koklaar et al. (1982), during theclosure and southerly subduction of an Iapetus ocean below the northern marginof Avalonia. This arc volcanic episode was succeeded by marine transgressionand the development of marginal basin volcanism in Wales. Thispaper aims to review the existing evidence in the literature pointing out tothe events that might have resulted in the development of arc and marginalbasin-type volcanism in Wales during the Ordovician. In addition, the controversialrelationship between rifting and the onset of arc volcanism based on timing, aswell as the proposed location of a back-arc during the developmentof a marginal basin in Wales are examined and alternative solutions proposed.
Figure 1: Chronostratigraphic chartshowing global series and stages in correlation with chronostratigraphic unitsrecognised in rocks of the UK. Source: Fortey et al. (2000). Figure 2: Paleogeographicreconstruction showing the position of Avalonia and Gondwana (South) as well asother continental fragments during the Early Ordovician (Arenig).
Source: Scotese and McKerrow (1990). 2.DATA & METHODOLOGY FROM WALES 2.1 PALEOMAGNETIC DATANumerous attempts at palaeomagnetic datacompilation for the lower Paleozoic (Ordovician to Permian) of Southern andNorthern Britain have been made, notably by Torsvik et al. (1990a) are shown in Fig 3 and earlier by Bridenet al. (1984, 1988). Southern Britain, of which Wales was part, belonged tothe Avalonian continental fragment while Northern Britain belonged to theLaurentian continent. Fig 3a comprisesthree groups of poles corresponding to the Ordovician (squares),Silurian-Devonian (diamonds) and Carboniferous-Permian (circles) time periodsand Fig 3b displays the revisedApparent Polar Wander Paths (APWPs) indicated as stars on the Equal area plot.
Figure 3: (a) Palaeomagnetic polesrepresenting the Ordovician-Permian of Southern Britain (b) Revised model with new amended paleomagnetic poles (SH), (CA) and(TV). Sources: Torsvik et al. (1990a), Trench & Torsvik (1990) Three Ordovician poles appear to be off bycomparison of Fig 3a and Fig3b, namely (SH) obtained fromintermediate hypabyssal or volcanic intrusions of the Shelve Ordovician inlierin mid-wales, (CA) obtained from the Carrock Fell Gabbro and (TV) from theTramore Volcanics of SE Ireland. A correct paleogeographic reconstruction (Fig 4) shows the position of Avaloniancontinental fragments relative to Gondwana during late Tremadoc-early Arenig(a) and Llanvirn (b). Figure 4a: Paleogeographicreconstruction of late Tremadoc-early Arenig.Sources: Western and Eastern Avalonia Bullard et al. fit (1965), Gondwana Vander Voo (1988) (pole: 34″N, 007″E, African co-ordinates),Laurentia and Northern Britain (Torsvik et al.
19906, pole: 13″S,29″E, European co-ordinates in aBullard et al. fit (1965)), Baltica(Torsvik er al. 19906, pole: 31″,086″E), Armorica (A) (Torsvik etal.
199oa, table 7,490 Ma pole: 30″N,334″E), Siberia with a mean pole of 30″N, 330″E (Torsvik et al. 1990a) Figure 4b: Paleogeographicreconstruction of the Late Llanvirn (Llandeilo). Sources: Avalonia from Trench & Torsvik (1991, table 2, 470 Ma pole: 12″N, 23″E), Baltica (Torsvik er al. 19906,combined path X and Y, 470 Ma pole: 21″N, 32″E), Armorica (Torsvik eral. 1990a, table 7,470 Ma pole: 33″N, 345″E), Laurentia and Northern Britain(Torsvik er al. 19906, pole: 223, 19″E, European co-ordinates). 2.2 SUBSIDENCE ANALYSIS Stratigraphic sections have been sampled by Cowie et al.
(1972) and Williams et al. (1972) in areas ofWales to obtain well dated data for the Cambrian-Ordovician stratigraphicprofile of Eastern Avalonia (Fig 5),the continental fragment of Avalonia that contained Wales. Figure 5: Cambrian-Ordovicianstratigraphic sections of eastern Avalonia. Sources: Cowie et al., (1972); Williams et al., (1972).
The sequences are mainly composed sedimentarydeposits of mudstones, sandstones, conglomerates, limestones and turbidites,with volcanic deposits observed in the Tremadocian to Caradocian time interval.Before obtaining a stratigraphic profile, the stratigraphic sections were back-stripped,a process which removes the additional subsidence resulting from sediment loading and isolates thesubsidence due to tectonic forces (or decompacted) as first described by Steckler and Watts (1978). Thedepositional thicknesses were then substituted by a same water thickness and usedto obtain a stratigraphic profile as shown in Fig 6. Thestratigraphic profile is a graphical representation of vertical movement whichcan help us reconstruct the subsidence history of a basin and thus rifting (Van Hinte, 1978). Figure 6: Cambrian-Ordoviciansubsidence curves from localities shown in Figure 5. Main sources: Cowie et al.(1972) and Williams et al. (1972).
2.3 FAUNAL EVIDENCE The trilobite and brachiopod faunas found inSouth Wales and elsewhere in Avalonia belong to the calymenacean-dalmanitaceanprovince typical of western Gondwana (Cocks& Fortey 1982, 1988) as shown in Fig7 below. Figure 7: Early Ordovician (Arenig-Llanvirn)paleogeographic reconstruction based on trilobite fauna, showing thepalaeolatitudinal relationship between Gondwana and Avalonia (South) as well asother continental fragments.
Avalonian trilobite fauna belong to theCalymenacean-Dalmanitacean province. Source: Cocks and Fortey (1988).Aremnant of the late tremadoc volcanic episode and ensialic destructive platemargin volcanism in Wales, mainly inferred from basalt-andesite-rhyodacite, which survivedthrough Arenig time to the present is the Rhobell Volcanic Complex (RVC) assuggested by Kokelaar (1977,1979,1986);Kokelaar et al. (1982). The RVC is a series of N-S trending dykes andintrusions which cut across uplifted and folded, broadly N-S trending areaslike the Harlech dome in Wales. The tectonic relations of the RVC and Harlechdome is somewhat complex but can be summarised as shown in Fig 8. Figure 8: Geological map summarising thetectono-volcanic relationship between the Harlech dome and the Rhobell VolcanicComplex (RVC) in Wales.
All the dykes at the RVC site are shown and sedimentsare presumably mainly composed of sediments derived from the Harlech dome.Sources: Unpublished work of C. A.Matley (BGS archive); Matley & Wilson (1946); Kokelaar (1977); Institute ofGeological Sciences (1982); Allen & Jackson (1985); M. Smith pers. comm. (1987). Arenig(non-volcanic) sediments have been sampled in part by Kokelaar et al.
(1985)and Craig (1985) in South Wales as shown in Fig 9, and their facies give the stratigraphic panel shown in Fig 10. These facies confirmpredominantly marine conditions in Wales during the Arenig and have beenproposed by Fortey (1984) to beassociated with an eustatic sea-level change that may have occurred worldwideat the time. Figure 9: Map showing locations forthe six areas studied (1- St David’s, 2- Llanferran, 3- Treffgarne Gorge, 4-Whitland, 5- Llangynog inlier, 6- Carmarthen). L= Basement lineaments, RF= RamseyFault. Fennian, Whitlandian and Moridunian are all subdivisions of the Arenigtime interval. Sources: Kokelaar et al. (1985); Craig (1985). Figure 10: Stratigraphic panelshowing correlation of logs from areas studied.
The difference members and formationsare numbered from 1 to 27 and a key is provided to help identify and correlatefacies. Sources: Kokelaar er al. (1985); Fortey and Owens (1987). Othersamples collected by Bevins (1982) are volcanic deposits of basic to acidcomposition, located to the northwest and northeast of the Fishguard, in theFishguard Volcanic Complex of Wales. To the northwest, the volcanic depositsoccur as basaltic pillow lavas and sheet flows as described by Kokelaar et al. (1984b) thus providingevidence for a subaqueous setting.
Tothe northeast however, the rocks are predominantly thinner rhyolites andrhyolitic tuffs that were in instances formed and deposited in a subaqueousenvironment (Lowman and Bloxam, 1981).The discriminant diagram in Fig 11 andclassification scheme in Fig 12below help classify the provenance of magma in the Fishguard Volcanic Complexand therefore can be used as evidence for a subaqueous environment in Wales andperhaps marginal basin transition from early Llanvirn to later Ordoviciantimes. An explanation for the derivation of this discriminant diagram is alsoprovided in Fig 13. Figure 11: Discriminant diagram for Titanium (Ti),Yttrium(Y) and Zircon (Zr) of the Fishguard Volcanic Complex in Wales. LKT =low-K tholeiite, CAB = calc-alkaline basalt, OFB = ocean floor basalt, WPB =within-plate basalt. Source: Derived from Pearce and Cann (1973). Figure 12: Proposed classification of magma provenance based on tectonicplate motion. Source: Pearce and Cann (1973).
Figure 13: Diagrams explaining how discriminantdiagrams are derived. Source:Pearce and Cann (1973). 3.DISCUSSION3.1 INTERPRETATIONSThe existence of a proto-Atlantic (Iapetus) ocean is unquestionable based on evidence of afaulted junction separating two faunal realms, ‘Pacific’ and ‘Atlantic, ofScotland and Wales as proposed by J.Tuzo Wilson’s (1966). The distribution of lower Ordovician graptolitefaunal provinces shows differences between wales in the South-eastern marginwith an Atlantic graptolite fauna (dominantly pendent Didymograptus species)and southern Scotland in the North-western margin with a Pacific graptolitefauna (no pendent Didymograptus species) (SkevingtonI974). These differences progressively disappeared during the Ordovician,thus proving the closure of an Iapetus ocean.
Arcvolcanism arises during the subduction of a denser plate beneath a plate oflower density. The addition of fluids due to high temperatures and pressuresduring subduction, lowers the solidus and causes partial melting of the crustinto less dense magma which rises up to the surface to form volcanoes. The palaeomagneticdata showing palaeogeographic reconstructions (Figs 3, 4a and 4b), the subsidence analysis (Fig 6) showing a major subsidence eventin South Pembrokeshire on the Welsh basin margin which is backed up by theincrease in marine facies in the Arenig of Wales (Fig 10) and the faunal data (Fig7), all point towards an Avalonia-Gondwanarifting during Arenig-Llanvirn time and asubsequent northward migration of Avalonia towards Laurentia. The closure ofIapetus would therefore result in its subduction beneath Avalonia and theformation of volcanic arcs which are survived today in North Wales as theRhobell Volcanic Complex (Kokelaar 1980)as shown in Fig 8.
Asopposed to the Tremadoc, Arenig and later volcanic rocks found in Wales appearto have mainly rhyolitic composition, typical of marginal basin volcanism (Kokelaar et al. 1984; Bevins et al. 1984).In addition, the discriminant diagram (Fig11) shows that magma samples from the Fishguard volcanic complex of Walesconcentrate in the 2+4 region corresponding toOFB (Ocean Floor Basalt), further proving a marine type volcanism, and pointingtowards a back-arc spreading as themechanism for marginal basin formation. 3.2 CONTROVERSIESAccordingto Kokelaar (1980), arc volcanismoccurred in Wales during Late Tremadoc time. On one hand in theliterature, a late Cambrian-Earliest Ordovician time for the rifting ofAvalonia-Gondwana is proposed by Cocksand Torsvik (2002).
On the other hand, all the evidence gathered in thisreport suggests a later time for the rifting. This brings out two main controversies,(1) whether arc volcanism was actually triggered by this rifting, and (2)whether the timing for this rifting between Avalonia-Gondwana is correct. Regarding the first controversy, thepaleogeographic reconstructions based on paleomagnetic poles obtained bysources listed in (Figs 4a and 4b)show that Gondwana and Laurentia remained roughly in place as Avalonia wasmigrating northward.
The subduction of the Iapetus ocean beneath Avalonia musthave therefore been caused by the rifting between Gondwana and Avalonia. Paleomagneticevidence (Fig 4) and faunal links (Fig 7) support a late Llanvirn time forthe initiation of rifting. In addition, any faunal contrast observed on Avaloniaearlier than Llanvirn times can potentiallybe attributed to Avalonia being on a more northern latitude than Gondwana. Eventhe stratigraphic evidence in Fig 6may be interpreted as having a rapid period of subsidence betweenArenig-Llanvirn time but also during Late Cambrian-Early Tremadoc, andtherefore does not really constitute a solid evidence from which timing for therifting of Avalonia and Gondwana can be derived.
It is therefore more likelythat the Arenig-Llanvirn timing for thiscomplete rifting is correct. One possible solution to the issue of latetremadoc arc volcanism preceding the timing of rifting, is partial rifting.Eastern Avalonia which contained Wales may have rifted-off Gondwana earlier andthus may have observed destructive plate margin activity before Western Avalonia.Anothercontroversy is the proposed back-arc marginal basin model, first put forward byKokelaar et al.(1984b) and which mayhave formed behind the Lake District-Leinster terrane. It is important toremember first here that the subduction of the Iapetus ocean occurred along andbelow the Northern margins of Avalonia where any movement of an arc in Waleswould have therefore been seaward. Thus, the major problem with this model is thatthe formation of a back-arc basin in Lake District during a southerlysubduction of Iapetus beneath Wales and the Lake district in their presentposition and distance relative to each other is unlikely, mainly because thisimplies the northward and so seaward migration of an original forearc fromWales to the Lake District, a process known as trench roll-back, over presentdistances.
This is not supported by current tectonic models, see Dewey (1980). A distance is clearlystated, over which trench roll-back may occur at current subduction rates, andis lower than the one being considered here. A possible alternative might beforearc migration from Wales over a shorter distance northward, but no evidencesupports this. 4.CONCLUSION The chronological causes for arc volcanism andmarginal basin transition in Wales can be assembled. The first proposed, is therifting between Gondwana and Avalonia during either the earliest ormid-Ordovician. Given that Avalonia (comprising Wales) was located on thesouth-eastern margin of the Iapetus ocean and that graptolite fauna showprogressive similarities between the south-eastern and north-eastern margins ofIapetus, we can safely infer a southerly subduction of Iapetus below thenorthern margin of Avalonia during closure.
Thissubduction of the Iapetus ocean was followed by periods of uplift and magmaticintrusions in Wales, subsequently leading to arc volcanism during late Tremadoctimes which is survived today by the Rhobell Volcanic Complex (RVC) in Wales (Fig 8). Marine transgression followed inWales during the Arenig, and is supported by the predominantly marine facies inFig 10. Spreading of a back arc inthe Lake District-Leinster terrane north of Wales followed this marinetransgression, in agreement with Kokelaaret al.(1984b).
Evidence in support of this transition is also given by theFishguard volcanic complex in Wales (Fig11).