Dinantian conodont biostratigraphy of the Northumberland Trough

Conodont faunas from Dinantian strata of the Northumberland Trough and Tweed Basin are described. Five zones are recognised. These zones are correlated with the standard British Stages and where possible with European and American sections. Chadian and Arundian faunas are dominated by Taphrognathus, Cloghergnathus and Cavusgnathus which typically occur in shallow intertidal and shallow subtidal environments. Holkerian and Asbian strata are barren of conodonts. In the Brigantian deeper-water genera including Gnathodus are represented. The shallow water nature of the lower faunas causes difficulty in correlation with goniatite-bearing sequences in the Craven Basin but reasonable correlations can be made to Ravenstonedale and the Avon Gorge.


INTRODUCTION
During the Carboniferous the Northumberland Trough developed as a half graben between the tectonically high Southern Uplands to the north and the fault-bounded Alston and Askrigg blocks to the south. Lower Carboniferous strata were deposited upon a surface of marked relief and lie either conformably on Old Red Sandstone "facies" (which may in part be Carboniferous in age) or unconformably on Lower Palaeozoic basement (Johnson, 1984). In the west lowermost Carboniferous sediments lie above the Birrenswark Lavas and traditionally the base of the Carboniferous system has been taken at the base of these (Robson, 1980). The stratigraphy of the Northumberland Trough is of interest as it provides a link between the well known Midland Valley sequences to the north and the area of Ravenstonedale in Cumbria, Garwood's type (1913, p. 451) for the Carboniferous succession of northern England. In addition, the trough continues to be of interest in hydrocarbon exploration and a good biostratigraphy is required to produce basin development models. Here we present a conodont biozonation for the Dinantian rocks of the trough, the first comprehensive micropalaeontological zonation.

STRATIGRAPHY
Early lithological classification (Tate, 1867;Lebour, 1875;Miller, 1887) established the Cementstone, Fell Sandstone, Scremerston Coal 'and the Lower, Middle and Upper Limestone groups. This scheme was not applicable to the sequence in the west of the trough and Lumsden et al., (1967) and  revised this lithostratigraphy introducing the terms Lower, Middle and Upper Border groups with the Liddesdale Group forming the top of the succession (Fig. 1).
The top of the Dinantian Subsystem has been placed on goniatite evidence at the base of the Great Limestone. Ramsbottom (1973. p. 573) used patchily dolomitised, commonly algal, calcite mudstones, to mark the regressive upper boundaries of his mesothem cycles, documenting six for the Dinantian Subsystem. Each was supported by palaeontological evidence except for the basal Asbian event. More recent work (George, 1978) suggests these cycles may be suspect, the lack of definitive internal criteria often making their identification uncertain or wrong (Austin & Davies, 1984, for a review). This causes some difficulties in using the Dinantian stages of George et al. (1976).

Lower Border Group (LBG)
These are the oldest Carboniferous sediments in the west of the trough. The succession in the Bewcastle area ( Fig. 2) was designated by Garwood (1931) as the type for the Group. Here it comprises 960m of alternating limestones. shales and sandstones deposited in a variety o f environments associated with periodic delta progradation into a shallow marine gulf (Leeder, 1974a).
Until modern biostratigraphical methods were applied, much of the LBG was considered to be Courceyan in age.  in his resurvey of the area noted the generally impoverished nature of the marine fauna, most of the molluscs, brachiopods, algae and serpulids common throughout the group are both long ranging and of limited diversity. Only the Cambeck Beds were reliably dated as C2 (i.e. ChadianiArundianage) using the macrofauna (Day, op. cit.). Ostracods from the Lower Border Group including Beyrichiopsis c.f. carinuta and Lichvinia sp. nov. were described as being of "essentially Tournaisian type" (Robinson, in Day, 1070. p. 166) though these persisted into the Vist-an Series.
Miospore studies on the Bewcastle Formation (Butterworth & Spinner, 1967) and the Main Algal Forma-tion (Gueinn, in Neves et al., 1972) provided little Ramsbottom (1973) interpreted the Main Algal further refinement. The miospores recovered are char-Formation as the regressive phase at the top of acteristic of the Lycospora pusilla or Pu Biozone Dinantian Mesothem 1, which he correlated with the (Neves et al., 1972) then thought to range from the Tournaisian-Visean boundary. George et al., (1976) Courceyan to Visean in age. Clayton (1984) revised the followed Ramsbottom (op. cit.) and equated the miospore zonation of the Dinantian Subsystem making Courceyan-Chadian boundary of their new stages with the Pu zone lower Visean in age.
the base of the Cambeck Beds. Leeder (1974a, b; RAVENSTONEDALE BEWCASTLE BRAMPTON BEADNELL (Higgins and Varker. 1982 197Sa, b;) also considered the Lynebank, Bewcastle and Main Algal formations to be Tout-naisian in age. Rnmsbottom (1977) reported the discovery in the lowermost Lynebank Formation of Chadian foraminifcra m d the conodont species Mestognathus beckmunni and Polygnathus hischoffi, elsewhere found together in strata of Chadian age. Subsequently the Lynebank, Bewcastle and Main Algal formations were considered to be Chadinn in age and the Cambeck Formation Arundian in age (Robson, 1980, fig. 2).

Middle Border Group (MBG)
Most of the MBG has been referred to the Arundian and Holkerian Stages Robinson, in Day, 1970;George et al., 1976;Robson, 1980). There are few diagnostic fossils in the sequence which is dominated by the Fell Sandstone and coeval Scremerston Coal Group in north Northumberland. The Fell sandstone has traditionally been correlated with the Ashfell Sandstone in the Stainmore Trough to the south (George et al., 1976, fig. 11) the latter sandstone being confined to the upper Arundian Stage.

Upper Border Group (UBG)
The base of the Asbian Stage can be placed in the Archerbeck borehole using foraminifera, within the Glencartholm Volcanics at an horizon correlated with the Clattering Band of Bewcastle (Day, 1970, pl. V). This unit is defined as the base of the UBG and contains a macrofauna of Lithostrotion martin, L. portlocki and Semiplanus sp., a fauna which can be traced eastwards from Bewcastle into the North Tyne area (Fowler, 1966, p. 76;Westoll et al., 1955, p. 81). In northern Northumberland George et al., (1976) placed the base of the Asbian Stage some 600m beneath the Redesdale Ironstone, consequently much of the Scremerston Coal Group became Asbian rather than Holkerian in age as traditionally conceived (Taylor et al., 1971, pl. V).
Robinson (in . p. 334) examined ostracods from the Spadeadam No. 3 (Camp) Bore and from the Clattering Band. Many of the species recorded are also known from the Middle Border Group. Neves & Williams (in Day. , p. 1972. X1) reported miospore assemblages from eight samples of coal and shale collected from the UBG of the River Irthing. They found a well preserved though low diversity microflora with few species of diagnostic age. They considered this reflected the poor state of knowledge of Visean assemblages. Comparing their microfloras with the more recent range charts compiled by Clayton (1984) it appears they are typical of the upper Holkerian Stage (Tc Biozone) to Asbian Stage (Nm Biozone).

Liddesdale Group
The base of thc Brigantian Stage has been placed at the base of the Peghorn Limestone, correlated by Trotter 24 Hollingworth (1932, p. 52) with the LOW Tipalt Limestone of the Greenhead-Bewcastle area. This limestone lies midway through the group (Fig. 1).
The lower part of the Liddesdale Group contains many fossils typical of the Asbian Stage, including a G. bilineatus Biozone fauna in the Denton Mill Limestone. Higher in the Group (Penton Limestone and above) typical Brigantian fossils are found, such as Lonsdaleia floriformis, Gigantoproductus striatosulcatus and G. Group. Dots represent present but not counted; ? remesents auestionable assignment.
giganteus. Saccamminopsis jloriformis is recorded from the Buccleugh Limestone. The equivalence of part of the P I zone of the Brigantian Stage has been confirmed by t h e presence of Posidonia becherii a n d Beyrichoceratoides truncatus in the shales between the Gastropod and Tombstone limestones (Ramsbottom, in Day, 1970, p. 175). Above in the higher limestones of the group macrofaunas and a L. monodosa Biozone conodont fauna are indicative of the P2 zone, this cannot be confirmed on goniatite evidence.

CONODONT FAUNAS
Generally conodonts throughout the sequence are low in abundance and diversity (Figs. 2, 3, 4, 5 ) , reflecting the marginal marine, shallow-water nature of the sequence. Preservation is generally good with thermal maturity values only exceeding conodont colour alteration index (CAI) 2 in proximity to the Whin Sill and related late Carboniferous-Permian intrusions (Armstrong & Strens, 1987). It is noticeable particularly in the Lower Border Group sequence that many specimens are exceedingly small, particularly Vogelgnathus and Taphrognathus species. This may be a result of the euryhaline nature of the environment, evidenced in the upper part of the group where algal laminites and gypsum pseudomorphs are found (Leeder, 1975a, b). A sequence of five conodont biozones is recognisable in the Northumberland sequence (      \ CAI reflect more marine conditions and where a more complete and continuous sequence is exposed. The lower three zones can be traced to Ravenstonedale and the upper two zones correlate well with schemes proposed by Higgins (1975;not shown in Fig. 8) and Metcalfe (1981) for the Central Pennines and the Craven Basin respectively (Fig. 8).
Currently, no single conodont zonation for Lower Carboniferous strata is applicable throughout Britain (refer to Varker & Sevastopulo, 1984 for a review) though attempts have been made at a conflated scheme (Varker & Sevastopulo, op. cit.;Higgins, 1984). This problem is caused by the strong environmental controls on Lower Carboniferous conodonts, particularly the distinction between shallow and deeper-water environments. Direct correlations with the Stage stratotypes of the Dinantian Subsystem would be preferable but only limited conodont data is currently available for these I I I I I I + I + I I I + I I I 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 : (see Austin & Davies, 1984 for a review). Whilst the Northumberland Trough represented a distinct depositional basin during Dinantian time, the conodont faunas recovered are similar to those reported from the Ravenstonedale shelf sequence. There are several advantages in correlating the Bewcastle succession with that of Ravenstonedale. The latter is the stratotype for the Lower Carboniferous succession of northern England and is demonstrably more complete than coeval shelf sections elsewhere in Britain. It is geographically close to Bewcastle and despite minor differences the depositional environments for both areas appear to be similar (Higgins & Varker, 1982;Leeder, 1975a, b).

Taphrognathus Biozone
T. varians (Branson & Mehl), Cloghergnuthus carinatus Higgins & Varker and 'Apatognathus' cuspidatus Varker, characterise this partial range zone. In Ravenstonedale, the appearance of ,C, carinatus and the disappearance of T. variuns mark the lower and upper boundaries respectively. Hindeodus scitulus (Hinde) was recorded in Ravenstonedale but is absent from the Northumberland sequence. T. varians and C. carinatus occur in the Lynebank Formation. The former disappears in the middle of the Bogside Limestone where we have a high sample density. Consequently we are confident in placing the top of this zone a few metres above the base of the Bewcastle Formation. Tuphrognathus varians with affinities to Cloghergnuthus con-tinues above this zone into the Kitty Beck Limestone and may in future allow recognition of a local subbiozone within the Cloghergnuthus Biozone. Within this biozone the fauna in Northumberland is much more diverse than that of Ravenstonedale which suggests a more favourable environment for conodonts, perhaps slightly deeper and less prone to salinity fluctuations. Of particular interest is the occurrence of Vogefgnathus sp. nov., which is considerably older than any other species reported in this genus (Norby & Rexroad, 1985).
T. varians has a wide distribution being recorded from VisCan strata in the Upper Mississipi Valley (Collinson, Rexroad & Thompson, 1971) and from Australia (Jenkins, 1974). In Britain T. varians was reported from the Main Algal Limestone of Roxburghshire (Rhodes et al., 1969) where it is early Visean in age. In Ravenstonedale it is associated with foraminifera in the Stone Gill Beds which are clearly referable to the Chadian Stage (Ramsbottom, 1977). In Europe this Taphrognathus fauna is uncommon, it occurs in Ireland (Austin & Mitchell, 1975) with the characteristic early Visean conodonts Mestognathus beckmanni and Cloghergnuthus. Taphrognathus is known from the Windsor Group of Novia Scotia (von Bitter, 1976) where it occurs with Cavusgnathus species and is likely to be Holkerian in age (Higgins & Varker, 1982, p. 154;Von Bitter & Austin, 1984, p. 100).

Cloghergnathus Biozone
This assemblage zone ranges from the middle of the Bogside Limestone to the lower-Main Algal Formation. These beds yield an impoverished fauna and of the species recorded from similar levels in Ravenstonedale only C. carinatus is present in the Bewcastle area. Higgins & Varker (1982, p. 154) considered this zone to be an "interregnum between the disappearance of Taphrognathus and the appearance of Cavusgnathus". This appears to be the case in Bewcastle. M . beckmanni, C. carinatus and Cloghergnathus sp. nov. disappear synchronously before the appearance of Cavusgnathus. M . beckmanni reappears in the Asbian Dun Limestone of the Tweed Basin. The Cloghergnathus Biozone is poorly defined elsewhere (Higgins & Varker, 1982). It appears to correlate with the T.

varians-Apatognathus interval in the Upper Mississippi
Valley and corresponds to a broad Cavusgnathusbearing zone in Ireland, between the top of the Lower Carboniferous Shale and the Upper Calp Limestone (Austin & Mitchell, 1975). In Ravenstonedale this zone occupies the Scandal Beck Limestone to mid Ashfell sandstone interval. George et af., (1976) dated the former as Chadian and the latter as Arundian in age.

Cavusgnathus Biozone
In Ravenstonedale the lower limit of this zone is marked by the first appearance of Cavusgnathus regularis and C. unicornis, the upper limit cannot be defined (Higgins & Varker, 1982). Correlation of the lower boundary in the Bewcastle area is problematical as the earliest Cavusgnathus are fragments and a single specimen of C. cf. navicufus from the Birky Cleugh Limestone in the Main Algal Formation.
Barren samples between the last Cloghergnuthus Biozone Fauna in the Ashy Cleugh Limestone (Bewcastle Formation) and the Cavusgnathus Biozone Fauna in the Birky Cleugh Limestone leave approximately 77m of section in which Cavusgnathus could make its first appearance. C. charactus has been recorded from the Liddel Formation of Harden Burn. Roxburgshire (Rhodes et al., 1969), ten miles to the north of Bewcastle. The Liddel Formation has been traditionally correlated with the Main Algal Formation on lithostratigraphic, macrofaunal and miospore      grounds (Garwood, 1931;Gray, 1971;Leeder, 1974a). Gueinn (discussion in Leeder, 1974a, p. 177) questioned this on the basis of miospore data. The presence of C. charactus in the Liddel Formation lends support to the traditional correlation and suggests the base of the Cavusgnathus Biozone should fall within the lower Main Algal Formation. In Ravenstonedale this zone includes the upper beds of the Ashfell Sandstone and at least the lowest beds of the Ashfell Limestone and therefore according to George et al. (1976) spans the Arundian-Holkerian boundary.

Gnathodus girtyi-Gnathodus bilineatus Biozone
Samples from the upper Cambeck Formation to the lower Liddesdale Group lack conodonts. The Denton Mill Limestone yields a low diversity fauna including G. girtyi and G. bilineatus. In Britain and Ireland the base of the G. bilineatus Biozone is marked by the first appearance of these two species, although G. girtyi girtyi is known to range down into the upper part of the Lochriea commutata Biozone (Varker & Sevastopulo, 1984). The base of this zone lies within the Asbian Stage (Metcalfe, 1981;Varker & Sevastopulo, 1984) and correlates well with the combined Mestognathus beckmanni-G. bilineatus and L. mononodosa zones of the Avon Gorge (Austin, 1973). In Britain the top of the zone is defined by the first appearance of L. mononodosa in the mid Brigantian Stage. In the Northumberland trough this lies within the Harelawhill Limestone and in the Tweed Basin between the Eelwell and Acre Limestones. Here the fauna includes M .

Lochriea mononodosa Biozone
As defined by Varker & Sevastopulo (1984) the base of this zone is marked by the first appearance of L. mononodosa (Rhodes et al.) and the top by the first appearance of G. girtyi collinsoni Rhodes et al. The latter subspecies has not been recovered from Northumberland and the L. mononodosa Biozone is taken to the top of the Brigantian Stage. The base of the synonomous G. nodus Zone (Metcalfe, 1981) occurs within PZb of the Brigantian Stage and Higgins (1975) defined his G. girtyi collinsoni Zone as ranging through the P2 subdivision of the Stage.
The appearance of a L. mononodosa Biozone fauna in the Acre and Sandbanks limestones of the Tweed Basin allows interbasinal correlation and our data therefore supports the correlations made by Holliday et al., (1975), that the Harelawhill and Eelwell limestone are in part equivalent. Austin (1976) distinguished a shelf biofacies comprising Cavusgnathus, Mestognathus and Taphrognathus from a deeper-water biofacies dominated by Gnathodus. This work was later substantiated in northern England (Higgins & Varker, 1982) and to some degree refined (Varker & Sevastopulo, 1984;Austin & Davies, 1984). Without exception the Chadian and Arundian conodont faunas of Northumberland are of shallow-water type. Von Bitter (1976) placed Taphrog     showed a high degree of facies control in Ravenstonedale, occurring typically in littoral and lagoonal facies but ranging into offshore infratidal facies. The Holkerian and Asbian rocks of Northumberland have failed to yield conodontsthe Middle and Upper Border Groups represent a sequence laid down over a flat lower delta plain that allowed frequent inundation by marine conditions. Cyclothems were predominant with the marine argillaceous limestones containing faunas dominated by brachiopods. The reason for the absence of conodonts remains unclear.

Facies Control of Carboniferous conodonts
The incoming of the Liddesdale Group over the Northumberland Basin marks a change in sedimentation and the conodont faunas. Thicker fully marine limestones can be traced across the whole area, the Denton Mill (or Dun Limestone) being the first to cross the Cheviot Block into the Tweed Basin. This marked the end of the separate identities of the basin and blocks and triggered the appearance of the deeper water Gnathodus-dominated fauna of the mid Brigantian Stage. Shallower water must have persisted in the Tweed Basin during the Denton Mill sedimentary cycle as Mestognathus reappears in the Dun Limestone. This distinction between the main trough and the Tweed Basin disappears as one passes up the sequence where thick Yoredale cycles continue to the top of the Dinantian succession. Limestones are generally thicker and contain a marine coral-brachiopod fauna. Conodonts are typically more abundant and represented by the genera Gnathodus, Lochriea, Hindeodus and 'Apa-strata of the Bewcastle area. tognuthus'.

CONCLUSIONS
It has been shown that, although environmentally controlled, conodonts are of biostratigraphic use within the shallow-water depositional environments of the Northumberland Trough. Five biozones have been established: the Taphrognathus, Cloghergnathus, Cavusgnathus, C . bilineatus and L. mononodosa biozones. The lower three allow good correlations to be made with the sequences in the Stainmore trough. The upper two zones can be traced throughout the trough into the Tweed Basin, and correlate with schemes proposed for the Central Pennines and Craven Basin. The Lower Border group is recorrelated in terms of the Visean Stages, the Lynebank and lower part of the Bewcastle Formation are Chadian in age, the upper part of the Bewcastle and Main Algal Formation are Arundian in age and the Cambeck Formation is in most part Holkerian in age. A Brigantian age is confirmed for the upper part of the Liddesdale Group. Our correlations support the lithostratigraphic correlations made by Holliday et al. (1975) between the Liddesdale Group and Upper Limestone Group of the Tweed Bas@, in that the Harelawhill and Eelwell limestones are in part equivalent.
Conodont age data help to constrain the timing of tectono-stratigraphic events within the development history of the trough. Rapid sedimentation, associated with a rifted origin for the trough occurred until the mid Brigantian age. Depositional conditions on either side of the Alston Block varied through time. Two comparisons can be made. Firstly the base of the Cloghergnathus Biozone lies within algal laminities in Ravenstonedale and within the shallow marine Bogside   The Northumberland Trough sequence reflects the continuation of shallow-marine sedimentation throughout Dinantian time and is thus ideal for the study of the palaeontology and palaeoecology of the somewhat unusual and restricted Dinantian conodont faunas which could obviously tolerate fluctuating salinities.