GEOLOGICAL JOURNAL Geol. J. 41: 1–22 (2006) Published online 10 November 2005 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/gj.1029
A late Ordovician (Hirnantian) karstic surface in a submarine channel, recording glacio-eustatic sea-level changes: Meifod, central Wales PATRICK J. BRENCHLEY1, JIM D. MARSHALL1*, DAVID A.T. HARPER2, CAROLINE J. BUTTLER3 and CHARLIE J. UNDERWOOD4 1
Department of Earth and Ocean Sciences, University of Liverpool, Liverpool, UK 2 Geological Museum, University of Copenhagen, Denmark 3 Department of Geology, National Museums and Galleries of Wales, Cardiff, UK 4 School of Earth Sciences, Birkbeck College, London, UK
The growth and decay of the end-Ordovician Gondwanan glaciation is globally reflected by facies changes in sedimentary sequences, which record a major eustatic fall and subsequent rise in the Hirnantian Stage at the end of the Ordovician. However, there are different reported estimates of the magnitude and pattern of sea-level change. Particularly good evidence for end-Ordovician sea-level change comes from a sequence at Meifod in central Wales, which has a karstified limestone unit within a channel incised into marine shelf sediments. Pre-glacial (Rawtheyan) mudstones have a diverse fauna suggesting a mid-to-deep-shelf water depth of c. 60 m. The channel, 20 m deep, was incised into these mudstones and partially filled with a mixture of fine sand and detrital carbonate. The taphonomy of bioclasts and intraclasts indicates that many had a long residence time on the sea floor or suffered diagenesis after shallow burial before being resedimented into the channel. The presence of carbonates on the Welsh shelf is atypical and they are interpreted as having accumulated as patches during a minor regression prior to the main glacio-eustatic fall. Comparison of the carbon stable-isotopic values of the bioclast material with the global isotopic record confirms that most of the material is of Rawtheyan age, but that some is Hirnantian. The resedimented carbonates lithified rapidly and formed a limestone, several metres thick, in the deepest parts of the channel. As sea-level fell, this limestone was exposed and eroded into karstic domes and pillars with a relief of over 2 m. The overall, glacioeustatic, sea-level fall is estimated to be in excess of 80 m. A succeeding sea-level rise estimated to be 40–50 m is recorded in the laminated crust that mantles the karstic domes and pillars. The crust is formed of encrusting bryozoans, associated cystoids, crinoid holdfasts and clusters of the brachiopod Paromalomena, which is normally associated with mid-shelf environments. Fine sands buried the karst topography and accumulated to fill the channel. In the sandstones at the base of the channel there is a Hirnantia fauna, while in the sandstones high in the channel-sequence there is cross-stratification characteristic of mid-shoreface environments. This would indicate a fall of sea-level of c. 30 m. The subsequent major transgression marking the end of the glaciation is not recorded at the Meifod locality, but nearby exposures of mudstones suggest a return to mid-to-deep-shelf environments, similar to those that prevailed before the Hirnantian regression. The Meifod sequence provides strong evidence for the magnitude of the Hirnantian sea-level changes and by implication confirm larger estimates for the size of the ice sheets. Smaller oscillations in relative sea-level seen at Meifod may be local phenomena or may reflect eustatic changes that have not been widely reported elsewhere. Copyright # 2005 John Wiley & Sons, Ltd. Received 14 June 2004; revised version received 29 April 2005; accepted 3 May 2005 KEY WORDS sea-level; facies; karst; Hirnantian; Ordovician; Wales
* Correspondence to: J. D. Marshall, Department of Earth and Ocean Sciences, University of Liverpool, Liverpool, L69 3GP, UK. E-mail:
[email protected] Contract/grant sponsor: British Geological Survey; contract/grant number: GA/95E/23.
Copyright # 2005 John Wiley & Sons, Ltd.
2
p. j. brenchley
ET AL.
1. INTRODUCTION The late Ordovician Gondwanan glaciation, which prevailed briefly within a prolonged greenhouse regime, induced a major sea-level fall at the start of the Hirnantian and a complementary sea-level rise in the midHirnantian when the ice caps melted (Figure 1). Oxygen stable isotopes reflect changes in ice volume and temperature of ocean waters and so record the growth and decay of the Gondwanan ice caps (Figure 1). Coeval effects on the carbon cycle in the oceans are reflected in the carbon isotope record (Figure 1). The timings of the sea-level and isotopic changes are particularly significant, because they are coeval with the main phases of the late Ordovician mass extinction, and are relevant to the timing and magnitude of the anomalous late Ordovician glaciation (Brenchley et al. 1994, 1995). Hirnantian facies-sequences seen at many localities around the world reflect the eustatic changes (Brenchley 1988). Estimates of the amount of sea-level change range from 50–100 m (Brenchley and Newall 1980) to 30–36 m (Long 1993) or ‘possibly less than 20 m’ (Chen 1984). The fall in sea-level has been interpreted as occurring in two phases with an intervening rise (Brenchley and Newall 1980; Brenchley and Sˇtorch 1989), or as a series of fluctuations (Long 1993; Armstrong and Coe 1996). Hirnantian glacial sequences in North Africa show two cycles of ice sheet advance, consistent with two phases of lowered sea-level (Sutcliffe et al. 2000) within the full glacial interval. The shelf sequence on the eastern margin of the Welsh Basin records the Hirnantian sea-level fall with a widespread erosion surface including channels deeply incised into shelf mudstones (Brenchley and Newall 1980, 1984). Above the erosion surface are sandstones, possibly deposited during a regressive phase within the subsequent transgression. In one of the channels, at Meifod in Powys, central Wales, there is a more complex fill, which includes a unit of karstified limestone below the main sandstone body. The importance of buried erosional topography in establishing the magnitude of sea-level change has been emphasized by Johnson et al. (1998) with reference to Silurian sequences. The implications of the karst in the Meifod channel-fill sequence are discussed in this paper. 2. PALAEOGEOGRAPHIC CONTEXT Prior to the Hirnantian, throughout the Rawtheyan, deep neritic mudstones with a mixed brachiopod–trilobite fauna prevailed along the margin of the Welsh Basin across North Wales and along the eastern shelf. From there, a west-facing palaeoslope descended to deeper basin mudstones (Figure 2a), with a fauna mainly of trilobites (Price and Magor 1984; Brenchley 1992, p. 32). The Hirnantian glacio-eustatic sea-level changes radically altered this palaeogeography. The initial regression created a widespread erosion surface across the shelf and incised channels that apparently fed mainly fine sediment onto the slope and created mud-dominated fans at the base of the slope (Figure 2b). During the subsequent overall transgression the channels were filled with sand, which also extended beyond the channel to blanket much of the rest of the erosion surface. 3. STRATIGRAPHY The upper Ordovician and lower Silurian stratigraphy of the Meifod area was first mapped and described by King (1928), who identified the Ordovician–Silurian boundary as being at the base of the sandstone unit. The erosion surface below the sandstone was described and figured, and the sandstone referred to as the ‘Basal Sandstone of Graig-wen’ in the erroneous belief that it was earliest Silurian in age, rather than latest Ordovician. The discontinuous limestones at the base of the sandstones were first recorded by Sedgwick (1843, cited in King 1928) and were regarded as large concretions, as they were much later by King (1928). Fearnsides, in a discussion of King’s (1928) paper, erroneously referred to great blocks of limestone, which had been derived from a more continuous lithified sheet, and later Brenchley (1993) made a rather similar interpretation. 3.1. Lithostratigraphy A summary of the lithostratigraphy is shown in Figure 3.4. Mudstones form the lowest part of the relevant stratigraphy in the Meifod area. The upper part of the mudstone sequence is characterized by the presence Copyright # 2005 John Wiley & Sons, Ltd.
Geol. J. 41: 1–22 (2006)
3
late ordovician karstic surface, central wales
RHUDDANIAN
HIRNANTIAN RAWTHEYAN
Ordovician
Silurian
Stage Graptolite Zone
Stable Isotopes
Benthic Faunas Glaciation Sea-level
δ 18O δ 13C
Survival/Recovery fauna
ascensus
Survival fauna
major ice caps
persculptus
Environmental Changes
Global warming, change in carbon cycling,sea-level rise, widespread anoxia.
Survival/Recovery fauna extraordinarius
Start of global cooling, change in carbon cycling, start of sea-level fall. pacificus
Pre-extinction fauna -6 -4 -2 0 2 4 6
0
40 80 m
Extinctions
Figure 1. A summary diagram showing late Ordovician events, biotic changes and stratigraphy (based on data as summarized in Brenchley et al. 2003).
of thin (< 3 cm) laminated siltstones and dispersed phosphate nodules up to about 1.5 cm in diameter, containing a diverse fauna of small fossils. The upper surface of the mudstones is erosional, with deep broad channels and shallower and irregular incisions in many parts of the surface. The overlying limestone fills the depressions in the topography, where it locally has an estimated maximum thickness of 15 m, but thins to zero at the channel margins. The limestone is composed of bioclasts, intraclasts and fine-grained sand that, in parts, constitute 50% of the rock. The upper surface of the limestone is in the form of domes and pillars with an estimated topographic relief of up to 3 m. Mantling the domes and pillars is a crust, formed predominantly of encrusting bryozoans. The overlying Graig-wen Sandstone fills the topographic relief of the limestones and accreted upwards to a form a thickness of about 25 m. The sandstones are massive or thick bedded, with parallel- or cross-lamination: they are composed of fine-grained, well-sorted, angular quartz sand. The overlying bioturbated shales with thin sandstones are not exposed above the sandstones, but are seen in small exposures a few kilometres away (see below). 3.2. Biostratigraphy In general the mudstones are sparsely fossiliferous, but the dispersed phosphate nodules have many fossils of small size constituting a diverse and undeformed fauna including the trilobites Primaspis (Primaspis) cf. evoluta, Cybeloides sp., Stenopareia sp., Flexicalymene? sp., Remopleurides spp. and Tretaspis sortita, indicating a Zone 7, youngest Rawtheyan age. Other elements of the fauna from the nodules are brachiopods, Sulevorthis and Chonetoidea, Plectambonites (now Eoplectodonta) simulans, recorded by King (1928), an indeterminate non-articulated brachiopod and a torynelasmatid. In addition, there are molluscs including Concavodonta, Sinuites, conulariids, indeterminate small gastropods and orthoconic nautiloids; indeterminate rare ostracodes, crinoid columnals and bryozoans are also present. The limestones have yielded a small brachiopod fauna with Mirorthis, Sowerbyella?, Eoplectodonta? that are typical of the Rawtheyan. The brachiopods occur as uncommon, dispersed single clasts; their association with resedimented bioclasts and intraclasts suggests they too might be resedimented and do not necessarily indicate the depositional age of the limestone (see discussion below). Within the bryozoan crust on the limestone there Copyright # 2005 John Wiley & Sons, Ltd.
Geol. J. 41: 1–22 (2006)
4
p. j. brenchley
ET AL.
a. Rawtheyan Palaeogeography of Wales Anglesey
Shelf mudstones /siltstones
m
te
Bioclastic carbonates
s Sy lt
u
y le
Fa
n
d
Meifod
r fo
s
te
n Po
Li
n tto re m t S te ch s ur t Sy h l C au F
Basinal Mudstones
Land
Llandovery
50km
Figure 2. Palaeogeography of the Welsh Basin and adjacent shelf during (a) late Rawtheyan when sea-level was high and (b) Hirnantian when sea-level was low.
are locally common articulated and disarticulated valves of Paromalomena polonica, a species which appears at the start of the Hirnantian and persists into the Rhuddanian, but is most common in the Hirnantian. The Graig-wen Sandstone has a well-preserved fauna comprising articulated and disarticulated brachiopods dominated by Hindella crassa incipiens together with Dalmanella testudinaria, Hirnantia sagittifera, Kinnella kielanae and Plectothyrella crassicostis (see Figure 12). The assemblage is typical of the ‘Hirnantia fauna’, which indicates an Hirnantian age. The immediately overlying bioturbated shales have yielded no benthic fossils, but King (1928) recorded two graptolites from a higher level. One of the graptolites was from greenish-grey mudstones, which yielded a small specimen of Normalograptus (formerly Climacograptus) normalis, indicating a probable Hirnantian age. The second graptolite, from a higher level, is Lagarograptus (formerly Monograptus) acinaces, which is likely to be of Llandovery age. 3.3. Stratigraphic relationships The channel and its fill, described in this paper, are exposed in a line of small quarries adjacent to the north side of the road (A495, Ordnance Survey Sheet 125, grid references SJ102093 to SJ104095) (Figure 3.3). The relevant sequence comprises upper Rawtheyan mudstones, overlain by discontinuous sandy limestones covered by Hirnantian sandstones (the Graig-wen Sandstone) (Figure 3.3). The sandy limestones at their thickest are estimated to be 15 m thick, but probably do not extend laterally for more than a few hundreds of metres at most. They are Copyright # 2005 John Wiley & Sons, Ltd.
Geol. J. 41: 1–22 (2006)
late ordovician karstic surface, central wales
5
Figure 2. Continued
demonstrably absent along strike in the Graig-wen Quarry (locality D, Figure 3.3) and are not recorded elsewhere in the region, even where the Graig-wen Sandstone is present (King 1928). The upper surface of the limestones is deeply karstified to form domes and pillars at least 2 m high. These karstic limestones only occupy the deeper parts of the channel (Figure 4), so that the overlying sandstone not only mantled the limestone, but came to rest on an erosion surface on the mudstones, in places where the limestone has been removed by erosion or where the limestone had never been deposited (Figure 4). The complex relationships between the component units of the stratigraphy are commonly difficult to determine because the rocks dip at c. 85 towards the observer, who consequently has a view nearly perpendicular to the plane of bedding. The quarrying was for the sandstone, so it ceased when it encountered the stratigraphically older limestones, now seen as the karstic domes and pillars, emerging through the sandstone cover. Locally, quarrying also ceased where the limestone was missing and it encountered the underlying shales. The sandstones are
