Chapter 1 Introduction to the Old Red Sandstone of Great Britain

W.J. Barclay

Introduction

The Old Red Sandstone is one of the two major 'red-bed' sequences of sedimentary rock in Great Britain, the other being the younger Permo-Triassic rocks that were formerly termed the 'New Red Sandstone' (see the companion GCR volume by Benton et al., 2002) to distinguish them from the Old Red Sandstone, rocks that are about 150 million years (Ma) older. In the early days of geological research in the 1830s, the Old Red Sandstone was included in the Carboniferous System, but soon after was given separate status and accorded a Devonian age, in recognition of its equivalence to the marine Devonian rocks of south Devon and Cornwall.

The GCR sites described in this volume are representative of the continental Old Red Sandstone facies in Great Britain. The rocks are mainly of what is now formally defined as Devonian age (about 418 to 362 million years (Ma) old), but according to modern definitions extend back into the Silurian Period, perhaps locally into the Wenlock Series (424 Ma). They also extend upwards into what is now defined as the early Carboniferous at less than 362 Ma (see (Figure 1.4), 'Stratigraphical framework for the Old Red Sandstone', this chapter).

The Old Red Sandstone crops out principally in five areas in Great Britain (Figure 1.1), which broadly reflect the original sedimentary basins in which they were deposited. These are:

• the Orkney and Shetland islands and northeast Scotland (the Orcadian Basin);
• the Midland Valley of Scotland (in an amalgamation of several basins of which the largest was the Strathmore Basin);
• the Scottish Borders and Northumberland (the Scottish Border Basin);
• the southern Lake District (the Mell Fell Trough); and
• south Wales, the Welsh Borderland and Bristol (the Anglo-Welsh Basin).

(Figure 1.2) shows the stratigraphical distribution of the main Old Red Sandstone sequences.

Traditionally, the base of the Old Red Sandstone in the Anglo-Welsh Basin was placed at the base of the Ludlow Bone Bed, a thin, lenticular, phosphatized 'lag deposit' marking the top of the Silurian Ludlow Series. However, the international agreement at the Montreal Devonian Symposium in 1972 to define the base of the Devonian System in the fully marine, graptolite-bearing succession exposed at Klonck in the Czech Republic, at the base of the Monograptus ultimus Biozone (e.g. House, 1977) now places the basal parts of the Old Red Sandstone in the modern Silurian System. The strata from the Ludlow Bone Bed up to the base of the modern Devonian System, which is as yet poorly defined in the Old Red Sandstone, belong to the Přídolí Series, the fourth, uppermost series of the Silurian System (White and Lawson, 1989). The age intervals (or stages) of the Devonian Period, also defined in the marine rocks of Europe, are applied to the terrestrial Old Red Sandstone succession with some difficulty because of its absence of marine fossils.

GCR site selection

The selection of Geological Conservation Review (GCR) sites described in this volume was carried out in the 1980s and 1990s, following the criteria set out in Ellis et al. (1996). The main reasons for qualification of a site for a particular GCR site selection category are:

• international importance — for example, the site may be important because it is a type locality for a geological time period, rock unit or fossil species, or is of historical importance in the development of geological science;
• possession of unique or exceptional geological features;
• national importance because a site is representative of a feature, event, process or rock body that is fundamental to the understanding of the geological history of Great Britain.

Scope

The GCR sites were selected according to thematic GCR 'Blocks', the present volume describing the 'Non-marine Devonian' GCR Block, which consists of 64 ratified GCR sites, together with a small number of potential GCR sites. The site descriptions are arranged geographically, from north to south, in areas that correspond to the original depositional basins. The sites are listed in (Table 1.1), together with the principal criteria for their selection. Many of the sites have features that satisfy several selection criteria. Furthermore, there are numerous Old Red Sandstone sites that have been independently selected for other GCR palaeontological 'Blocks'. These sites are described in the companion GCR volumes on fossil fishes ((Table 1.2); Dineley and Metcalf, 1999) and Palaeozoic palaeobotany ((Table 1.3); Cleal and Thomas, 1995).

(Table 1.1) GCR Old Red Sandstone sites and proposed sites, with main criteria for their selection. Continued on page 7.

(Table 1.2) GCR sites in the Old Red Sandstone described in the fossil fishes GCR volume. After Dineley and Metcalf (1999). Continued on page 9.

(Table 1.3) GCR sites in the Old Red Sandstone described in the Palaeozoic palaeobotany GCR volume. After Cleal and Thomas (1995).

The GCR sites provide representative localities for the entire stratigraphical range of the Old Red Sandstone. The initial selection of GCR sites for the 'Non-marine Devonian' GCR Block included sites in the Anglo-Welsh Basin in strata that extended down from the base of the Devonian System to the Ludlow Bone Bed. These strata, comprising the Downton Group (the former Downtonian Stage), are of Silurian (Přídolí Series) age and the sites ((Table 1.4)) are described in the GCR volume on Silurian stratigraphy (Aldridge et al., 2000). The GCR volume on Caledonian igneous rocks (Stephenson et al., 1999) includes sites ((Table 1.5)) in which Old Red Sandstone strata are present in addition to the contemporaneous igneous rocks for which they are cited. All of the Old Red Sandstone sites described in the other GCR volumes are listed in (Table 1.2), (Table 1.3), (Table 1.4) and (Table 1.5), along with the level of detail in which they are described in the present volume. Only some of these 'overlapping' sites are given full descriptions in the present volume, which emphasizes the sedi-mentological and lithostratigraphical features. The Freshwater West potential Old Red Sandstone GCR site, and part of the Freshwater East–Skrinkle Haven GCR site are also Variscan to Alpine structures GCR sites.

(Table 1.4) GCR sites in the Old Red Sandstone described in the Silurian stratigraphy GCR volume. After Aldridge et al. (2000).

Site selection is inevitably subjective, but the aim of the GCR is to identify the minimum number and area of sites needed to demonstrate the current understanding of the diversity and range of features within each GCR 'Block'. The preferred sites are generally those that are least vulnerable to the potential threat of destruction, are more accessible and are not duplicated elsewhere (Ellis et al., 1996). The original selection of sites was made over 20 years ago, and all of these sites are included in this volume. In addition, a small number of sites were identified during the course of the compilation of the volume as representing stratigraphical units or unique features not included in the original GCR site selection, and are described as 'potential' GCR sites.

History of research

The name 'Old Red Sandstone' appears to have been first applied to the red rocks below the Mountain (Carboniferous) Limestone in the mistaken belief that they were the equivalents of the Permian Rotliegendes of Germany (Jameson, 1821; Simpson, 1959). It was initially mapped and named in southern Wales and the Welsh Borderland (Phillips, 1818; Conybeare and Phillips, 1822) and included as the lowermost part of the Carboniferous System. Murchison was the first to champion the Old Red Sandstone as a separate geological entity. According to Miller's account (1841), a visiting foreign geologist advised Murchison that 'you must inevitably give up the Old Red Sandstone: it is a mere local deposit, a doubtful accumulation huddled up in a corner, and has no type or representative abroad.'

(Table 1.5) GCR sites with Old Red Sandstone sedimentary rocks described in the Caledonian igneous rocks volume. After Stephenson et al. (1999).

'I would willingly give it up if Nature would,' replied Murchison, 'but it assuredly exists, and I cannot'. Compared to the richly fossiliferous rocks of the Silurian System below and the Carboniferous System above, the Old Red Sandstone seemed relatively barren to the early Victorian workers, but as the remains of early fishes were discovered, first in Scotland by the young Swiss naturalist Louis Agassiz, and later in south Wales and the Welsh Borderland, interest gradually increased.

The Devonian System was established by Sedgwick and Murchison (1839) for the pre-Carboniferous marine rocks of Devon. These rocks were readily correlated with the Rhenish nearshore rocks and the Bohemian deep-water rocks of mainland Europe (House, 1977). With the recognition of large tracts of Old Red Sandstone in North America, Norway, Siberia, Poland and Russia (the last containing many of the same fish species as Great Britain), the strata assumed a new importance (Geikie, 1879). At the same time, Murchison (1839), impressed by the great thickness of Old Red Sandstone strata in the Welsh Borderland, and the difference between them and the overlying Carboniferous rocks, with which they had hitherto been merged, applied the status of 'system' to the Old Red Sandstone. This situation held, in the UK at least, for over 130 years, with the term used in a quasi-chronostratigraphical sense for rocks of continental facies and Devonian age. However, with the advent of more precise stratigraphical procedures and classification, and, in 1972, the new definition of the base of the Devonian System at a higher level, equivalent to a horizon within the Old Red Sandstone (see below), the term is now no longer used in a quasi-chronostratigraphical sense. Biostratigraphers tend not to use the term at all; the glossary in the companion GCR volume on Silurian stratigraphy, for example (Aldridge et al., 2000), defines the Old Red Sandstone as 'a classic term still applied to the terrestrial, largely clastic facies of the late Silurian to earliest Carboniferous in Britain'. Sedimentologists retain the name as a facies (or magnafacies) term for all the terrestrial red beds and lacustrine deposits of Silurian to early Carboniferous (but predominantly Devonian) age (e.g. Friend and Williams, 2000). The term 'Old Red Sandstone' is also applied in an informal lithostratigraphical sense. The three subdivisions of the Old Red Sandstone recognized by Murchison — Lower, Middle and Upper — are similarly retained as informal, but long-established lithostratigraphical terms onshore in the United Kingdom and as formal groups offshore in the North Sea.

Old Red Sandstone palaeogeography

The Old Red Sandstone represents a period when ocean closure and continental collisions resulted in a world geography vastly different to that of much of early Palaeozoic times. The drift of the early Palaeozoic continents and their relative positions can be estimated from the correlation of geological successions and their faunas, with palaeomagnetism providing data on palaeolatitudes. The Iapetus Ocean, which separated the northern (Laurentian) and southern (Gondwanan) continents, closed throughout the Ordovician and Silurian periods as the smaller continent of Avalonia fragmented from Gondwana and drifted northwards (Figure 1.3). As the Iapetus Ocean closed north of Avalonia, the Rheic Ocean opened behind it. To the east, the continent of Baltica also drifted northwards and eastwards and the Tornquist Sea, an arm of the Iapetus Ocean, slowly closed. The timing and nature of the convergence of the three components that were to make up the Old Red Sandstone continent remain matters of debate. Trench and Torsvik (1992) considered that Baltica and the eastern part of the Avalonia microcontinent collided first, in late Ordovician times, moving northwards together to make first contact with Laurentia by late Silurian time at about 420 Ma (Torsvik et al., 1996; see Dewey and Strachan, 2003, fig. 1). However, Dewey and Strachan (2003) interpret the Scandian Orogeny as the result of collision, by sinistral transpres-sion, of Baltica and Laurentia from about 435 Ma to 425 Ma, with a soft collision between Avalonia and Laurentia/Baltica (Laurussia) at about 425 Ma (Soper and Woodcock, 2003). By late Silurian (Ludlow) time, the continents had fully docked, with the Iapetus Ocean dosed along the line of subduction (the Iapetus Suture) under the Southern Uplands. Thus, the Caledonian-Appalachian Orogen (or North Atlantic Caledonides) and the newly amalgamated Old Red Sandstone continent (Laurussia or Euramerica) were formed. Continuing compression and shortening of the continental crust resulted in the filling to sea level of the Silurian marine basins, their inversion to upland areas and the establishment of terrestrial conditions in newly developing basins.

Palaeogeographical reconstructions (e.g. Scotese, 2001) suggest that the continent lay in tropical to sub-tropical latitudes from the equator to about 30°S, with the Anglo-Welsh Basin lying approximately 5°S to 15°S. Palaeomagnetic data from the Lower Old Red Sandstone in southern Wales suggest a latitude of 17 ± 5°S (Channel et al., 1992). Sedimentological studies of the Old Red Sandstone, and particularly of its fossil carbonate soils (calcretes) confirm, by analogy with modern calcretes, a warm to hot, semi-arid tropical to sub-tropical setting (e.g. Allen, 1986) with rainfall confined to wet seasons (e.g. Marriott and Wright, 1993). Uplift of the orogen may have caused broad variations in the rainfall pattern, producing periods of wetter and drier climate.

Woodcock (2000a), Friend et al. (2000), Dewey and Strachan (2003) and Soper and Woodcock (2003) presented recent overviews of the tectonics and kinematics of — Old Red Sandstone basin formation. Superimposed on the broadly compressive stresses associated with convergence of the Avalonian and Laurentian continental margins, the oblique angle of closure produced strike-slip transpressive and transtensional movements. The nature and extent of these movements remain the matter of debate, largely centred on whether there was a major, orogen-wide sinistral megashear or whether basins were controlled by strike-slip movements of different sense and at different times during the Caledonian orogenic cycle. Another debate concerns the possible role of gravitational collapse of the uplifted, granite-buoyed Caledonian Orogen in the formation of some at least of the internal basins (e.g. Woodcock, 2000a). Dewey and Strachan (2003) conclude that the diachronous closure of Iapetus, and subsequent deformation and basin formation were controlled by sinistrally dominated relative movement between the Laurentian and Avalonia–Baltica plates. The Old Red Sandstone basins probably formed as a result of sinistral transtension, with an estimated 1200 km of strike-slip movement between Laurentia and Baltica. Rheic convergence in the Emsian Age (late Early Devonian) from 400 Ma to 390 Ma resulted in the Acadian Orogeny, which affected the basins south of the Highland Boundary Fault (Soper and Woodcock, 2003).

The Old Red Sandstone basins were formerly divided into two main groups on the basis of their positions relative to the Caledonian Orogen (e.g. Allen, 1977; Woodcock, 2000a). Those within it (internal or intramontane basins) include the Orcadian Basin, the basins of the Midland Valley of Scotland (but see below) and the Scottish Border Basin. The Anglo-Welsh Basin was regarded as an external, or extra-montane basin, open to the sea to the south. However, the recent models, invoking major orogen-parallel, sinistral movement and three separate, temporally discrete collision events (Grampian, Scandian and Acadian) collectively making up the Caledonides have revised the former view of a continuously 'prograding Caledonian mountain front.

The recent models (Dewey and Strachan, 2003; Soper and Woodcock, 2003) envisage that the highly oblique, sinistral closure of the Iapetus Ocean resulted in, sequentially, transpression, strike-slip and transtension. The area of maximum uplift in the Scandian Orogen was to the north of Britain, in Scandinavia, in an orogen of Himalayan proportions (Dewey and Strachan, 2003). The compression in the Laurentian crust, of which the Scottish Highlands were part, caused thrusting along major NE-trending faults, granitic intrusion, andesitic volcanicity and low-grade metamorphism in northern Britain (Stephenson et al., 1999). The volcanic rocks were probably extensive, their eroded remnants being seen at Ben Nevis, Glen Coe, Lorn and just north of the Highland Boundary Fault. Volcanic rocks also occur extensively within the Midland Valley of Scotland, at Montrose, in the Sidlaw, Ochil and Pentland hills, and in Ayrshire. They also occur more locally in the Southern Uplands, where granitic intrusions such as the Cheviot were emplaced.

During the transcurrent and transtensional phases, much of the orogen-parallel, sinistral movement appears to have been taken up by the Great Glen Fault and its north-east continuation, with at least 700 km of displacement (Dewey and Strachan, 2003). The formation of the Late Silurian–Early Devonian Old Red Sandstone basins is also attributed to sinistral transtension (Dewey and Strachan, 2003; Soper and Woodcock, 2003). The Acadian Orogeny ended this phase of basin formation and caused transpressive shortening of the early Palaeozoic basins flanking the Midland Microcraton, as well as the inversion and erosion of the Old Red Sandstone rocks not underlain by the micro-craton (Soper and Woodcock, 2003). The cause of the Acadian event was probably the collision of a Gondwana-derived continental fragment (Soper and Woodcock, 2003) with the Midland Microcraton segment of the amalgamated Laurussian continent. The evidence for the terrane boundary in the vicinity of the Bristol Channel is now confined to the Lizard mafic–ultramafic complex, interpreted as an ophiolite and a fragment of the Rheic suture (e.g. Soper and Woodcock, 2003).

The Orcadian Basin was a large Mid-Devonian intramontane lake basin, totally unconnected to the open sea, apart perhaps from a brief period. Its formation was probably due to a combination of both gravitational extension, and transtensional movements on basin-margin faults. The Midland Valley of Scotland was not a single discrete basin in the Devonian Period. Weakened by a long history of igneous activity, internal, transtensional fault movements opened pull-apart basins and transpressive movements subsequently inverted them, resulting in the recycling of the basin-fills and providing weak points for continuing volcanic extrusion (e.g. Bluck, 2000). The preserved sequences thus represent the deposits of separate pull-apart basins, formed and brought together in a strike-slip faulted collage. The Stonehaven Basin in the north-east is the earliest, its sedimentary fill dating perhaps from the Wenlock Epoch (Marshall, 1991). It and its larger successor basins, the Crawton and Strathmore basins, formed by sinistral strike-slip along the Highland Boundary Fault. The southerly Lanark Basin formed along the Southern Upland Fault. Large volumes of arc-related volcanic rocks were extruded along the central axis of the Midland Valley, on lines weakened by the transtensional stresses (e.g. Bluck, 2000). The late Devonian Scottish Border Basin formed after Acadian inversion in Mid-Devonian times and extended into the Midland Valley, Northumberland and the Solway Firth.

The Anglo-Welsh Basin was formerly interpreted primarily as the product of load-generated flexural subsidence of the Caledonian foreland (James, 1987; King, 1994; Friend et al., 2000). Dewey and Strachan (2003) and Soper and Woodcock (2003) prefer a transtensional mechanism for its formation. Transtensional movement on faults produced variations in the basin-fill in Pembrokeshire (e.g. Marshall, 2000a,b) and introduced coarse, clastic, detritus farther north (Tunbridge, 1980a). The isolated succession in Anglesey was probably deposited contiguously with the Přídolí–Pragian sequences to the south, with which there are marked similarities, although the initial coarse conglomerates are unique and of local derivation, and lacustrine deposits suggest internal or impeded drainage.

Stratigraphical framework for the Old Red Sandstone

Stratigraphical classification of the rocks in the geological record has traditionally fallen into two broad categories — lithostratigraphical and chronostratigraphical. Lithostratigraphical classification is based on the physical characteristics of a rock body, such as colour, rock type (lithofacies) and mode of formation. Chronostratigraphical classification is based on the relative age of a rock body, determined by its fossil content (biostratigraphy) as correlated with standard, defined and internationally agreed geological marker horizons (the 'golden spikes'), and in the case of igneous rocks, by radiometric age dating. Biostratigraphical classification is achieved by the study of component fossil and microfossll groups, with subdivisions based on marker species or assemblages of species. Thus, in the Devonian System, there are biostratigraphical zonal schemes for graptolites, ammonoids, brachiopods, fish, conodonts, microvertebrates and miospores.

The chronostratigraphical subdivisions of the Upper Silurian and Devonian (Figure 1.4) are internationally agreed and defined in fossili-ferous marine strata in continental Europe. The Upper Silurian Přídolí Series (not yet divided into stages) and Lower Devonian Lochkovian and Pragian stages are defined in the deep-water, graptolite-bearing succession of the Prague Basin in the Czech Republic. The last two replace the previously used, but not completely equivalent Gedinnian and Siegenian stages defined in the nearshore succession of the German Rhenish Basin. The highest Lower Devonian stage is the Emsian, defined in Belgium. The Middle Devonian stages (Eifelian and Givetian) are defined in Germany, the Upper Devonian stages (Frasnian and Famennian) are named from the carbonate-bearing marine succession of southern Belgium.

The problems of classification and correlation of the Old Red Sandstone of Great Britain are inherent in its terrestrial origins and the patchy preservation of its non-marine fossils. The fossils that are present indicate that the Devonian Period was a time of profound changes in the evolutionary record, with the first significant colonization of terrestrial habitats by vascular plants (Edwards, 1979a), the rapid expansion of the first aquatic vertebrates, and their emergence onto land. However, no direct correlations can be made with the European marine successions and the internationally agreed stages. Because of this, a series of loosely defined local stages (Downtonian, Dittonian, Breconian and Farlovian) were erected for the Anglo-Welsh Basin, but now have been largely subsumed into the international stages as a result of increasing refinement in correlation. (Figure 1.4) (based on House et al., 1977; and Marshall and House, 2000) shows the stages and their correlation.

The principal macrofossils are fish fragments. A biozonal scheme was erected for the Old Red Sandstone in the Anglo-Welsh Basin (see (Figure 5.3), Chapter 5) and was extended to continental Europe. Refinement of the scheme continues (Blieck and Janvier, 1989; Blieck and Cloutier, 2000), but the occurrence of fish remains is patchy and of limited use in high-resolution correlation. Miospore classifications (e.g. Richardson et al., 2000; Streel et al., 2000) and microvertebrate classification (Vergoossen, 2000) also aid correlation and stratigraphical resolution. However, the problem of detailed correlation of the terrestrial Old Red Sandstone succession with the Bohemian, German and Belgian marine stages, in which miospores are rare, remains. Progress is, however, being made by chains of correlation involving miospores that are common to the Old Red Sandstone and the Rhenish marine succession, the latter then being correlated with the Bohemian stages. For example, the recognition of the Breconensis-zavallatus Zone in the Ardennes allows correlation of the Anglo-Welsh and Rhenish Gedinnian-Siegenian successions (Richardson et al., 2000) ((Figure 5.3), Chapter 5). A widespread volcanic ash deposit (the Townsend Tuff Bed) and a basin-wide calcrete (the Psammosteus Limestone) are valuable marker horizons in the Anglo-Welsh Basin, providing lithostratigraphical correlation of the succession.

References