Integration of Darriwilian (Middle Ordovician) δ13Corg chemostratigraphy with graptolite biostratigraphy in the classical Röstånga area in northwestern Scania (southern Sweden)

The largely covered Middle Ordovician succession in the classic geological Röstånga area in northwestern Scania has not been studied for some 80 years. A new drill core through a succession ranging from the lower–middle Darriwilian to the lower Sandbian has provided a unique opportunity to investigate the graptolite biostratigraphy and the δCorg chemostratigraphy, and clarify their stratigraphic relations, through this ~90 m thick interval, which is developed within a black shale facies. The lithology, biostratigraphy and chemostratigraphy are closely similar to those of the coeval strata in the Fågelsång area, south­central Scania, including the presence of the Fågelsång Phosphorite, which was previously unrecorded in the Röstånga area. The new data are particularly important in providing evidence of the relations between graptolite biostratigraphy and δCorg chemostratigraphy. The Fågelsång­3 and Röstånga­2 drill core successions are currently the only Darriwilian sequences in the world where these relations have been well established.


INTRODUCTION
An important area for the study of Darriwilian graptolite biostratigraphy not only in Scania (Skåne), southern Sweden, but in the entire Baltoscandia is at the village of Röstånga, which is located approximately 35 km north of the city of Lund (Fig. 1). Although the Middle Ordovician outcrops in that region are of insignificant size, shallow diggings through unexposed intervals, such as those carried out by Hadding (1913) and Ekström (1937), have provided important information on some Middle Ordovician stratigraphic intervals. The pioneer work on the Middle Ordovician of the Röstånga area was carried out a very long time ago (e.g. Tullberg 1880Tullberg , 1882Tullberg , 1883Moberg 1910;Hadding 1913;Ekström 1937) and virtually no new studies of the Lower and Middle Ordovician of this area have been published for decades. The principal reason for this scientific inactivity may have been that the existing small outcrops were judged unlikely to yield significant new information. Hence, in order to establish a useful standard succession for the Röstånga area, there was an urgent need for a new drill core through the Darriwilian succession. The usefulness of such a new drill core was illustrated by a recent drilling (Röstånga1) through the Upper Ordovician and lower Silurian that has yielded a wealth of new information (e.g. Bergström et al. 1999Bergström et al. , 2014Bergström et al. , 2016Pålsson 2002;Maletz et al. 2014).
A new core drilling, here referred to as Röstånga2, through the Darriwilian succession, was carried out in 2016. The purpose of the present report, which is preliminary in some respects, is to summarize the principal stratigraphical results of this drilling, especially the relations between the graptolite biostratigraphy and the δ 13 C org chemostratigraphy through the Darriwilian succession. It is of special interest that comparable stratigraphic data have recently become available from a similar drilling (Fågelsång3) that penetrated the Darriwilian succession in the Fågelsång area in southcentral Scania (Bergström et al. 2018;. The drilling at Röstånga is important in providing an opportunity to assess the validity of the relations between graptolite biostratigraphy and δ 13 C org chemostratigraphy recognized by Bergström et al. (2018Bergström et al. ( , 2020 in the Fågelsång succession. Although different in some respects, these two key successions exhibit notable similarities, which suggests that they are representative of the outer shelf-upper slope deposition across this part of Baltoscandia during Darriwilian time. Both these succes sions are located in the Scanian Confacies Belt of Jaanusson (1976Jaanusson ( , 1995 in which the dominant lithology is graptolite bearing fineclastic sediments that were deposited in deeper water than the coeval coldwater carbonates (see Lindskog & Eriksson 2017 and references therein) with diverse shelly faunas that were laid down in the central parts of the Baltoscandic platform.

PREVIOUS WORK ON THE DARRIWILIAN SUCCESSION AT RÖSTÅNGA
Early studies of this part of the Ordovician in the Röstånga region were carried out during the second half of the nineteenth century by, among others, Linnarsson (1875Linnarsson ( , 1879 and Tullberg (1880Tullberg ( , 1882Tullberg ( , 1883. In an informative summary of previous work together with much new information, Moberg (1910) provided a still useful guide to the Lower Palaeozoic geology of the area. Subsequently, two of Moberg's students, namely Hadding (1913) and Ekström (1937) Ekström's (1937) studies, which were based on field work from 1913 to about 1920, very few new data have been published on the Darriwilian succession at Röstånga. This is in contrast to the significant amount of work that in recent years has been devoted to the Katian and Hirnantian part of the succession (cf. Pålsson 1996Pålsson , 2002Bergström et al. 1997Bergström et al. , 1999Bergström et al. , 2010Bergström et al. , 2011Bergström et al. , 2018Badawy et al. 2014;Maletz et al. 2014). A major reason for this is the lack of outcrops of especially the Lower Ordovician that still remains essentially unknown in this area.  Hadding (1913) and Ekström (1937) made their excavations, are along the west side of the brook (stippled oval) a short distance northwest of the Röstånga2 drill site. Modified after   fig. 1).

LITHOLOGY AND LITHOSTRATIGRAPHIC CLASSIFICATION OF THE RÖSTÅNGA2 DRILL CORE SUCCESSION
The Röstånga2 core drilling, which took place on a field 30 m east of the Church Brook and 370 m southeast of Röstånga Church (Fig. 1), resulted in a 116.11 m long core succession, reaching a depth of 116.91 m (Fig. 2). The drilling was carried out in September 2016 by the Engineering Geology Group of the Department of Measurement Technology and Industrial Electrical Engineering at Lund University using an Atlas Copco CT20 drill rig (Riksriggen). The core diameter is 82 mm down to 12.10 m and 62 mm between 12.10 and 116.91 m.
The cored sequence has a dip of ~34-43 degrees and core loss was minimal. Using an average dip of ~38 degrees, the calculated stratigraphical thickness of the penetrated succession may be approximately 90 m. This may not represent the true thickness of the study succession as a few bedding plane surfaces may be interpreted as possibly representing fault planes. However, the completeness of the graptolite zone succession suggests that there are no major gaps in the succession, but it cannot be excluded that there are some faultgenerated minor duplications or deletions of one or several intervals. Nevertheless, it is difficult to avoid the conclusion that the Darriwilian sequence at Röstånga is significantly thicker than that at Fågelsång, possibly by as much as 50%.
Virtually the entire core succession is a rather mono tonous black to grey shale with a few thin beds of finecrystalline impure limestone mixed with siliciclastic material (Fig. 2). The core interval from 111.10-116.91 m, that is, roughly the stratigraphically lowermost 5 m of the study succession, is light to mediumgrey mudstone. Both graptolite biostratigraphy and chemostratigraphy suggest that this interval may represent a transition to the early Darriwilian carbonate unit known as the Komstad Limestone that is rather widely distributed in Scania (see Bergström et al. 2018 for a comparison with the cor responding interval in the Fågelsång area). Moberg (1910) recorded this unit in boulders from a locality at Röstånga, but later authors (e.g. Ekström 1937, p. 23) have expressed doubts about its in situ occurrence in this area (see also Nielsen 1995, p. 13). An occurrence of the Komstad Limestone at Röstånga appears not unlikely in view of the fact that it has a thickness of 2.75 m in the Albjära drill core (cf. Maletz 2005), which was drilled at a site approximately 8 km southwest of Röstånga (Fig. 1A). It seems that deepening the drill hole of Röstånga2 a few metres would clarify if the Komstad Limestone is present at this site.
The drill core succession up to a depth of 19.7 m is classified as the Almelund Shale (Fig. 2). Most interestingly, its top is marked by a few centimetres thick phosphorite bed (Fig. 3) that appears to have the same stratigraphic position as the prominent Fågelsång Phosphorite in the Fågelsång succession (cf. Bergström et al. 2000). This partly con glomeratic bed has not been identified with certainty in northwestern Scania, but Tullberg (1883) recorded a lithologically similar bed in a drill core from Stabbarp, which is approximately halfway between Röstånga and Fågelsång (Fig. 1A). Hadding (1913, p. 82) noted that phosphorite lenses and beds occur at several levels in the Fågelsång succession (cf. Bergström et al. 2000) and he felt that in the absence of biostratigraphy, it was premature to identify the Stabbarp bed with what is now referred to as the Fågelsång Phosphorite. At any rate, this prominent bed apparently represents a period of very slow net deposition, erosional reworking and/or sea floor phos phogenesis during the earliest part of the Upper Ordovician Sandbian Stage (cf. Finney & Bergström 1986). As is the case in the Fågelsång succession, the Almelund Shale succession in the new drill core contains a small number of thin Kbentonite beds.
The topmost 19.7 m of the drill core, that is, the interval above the Fågelsång Phosphorite, includes dark shale with a couple of thin Kbentonite beds. Following the litho stratigraphy of the Fågelsång succession (Bergström et al. 2002), it is herein classified as the Sularp Shale.

GRAPTOLITE BIOSTRATIGRAPHY OF THE RÖSTÅNGA2 SUCCESSION
The graptolite biostratigraphy of the Röstånga2 core was recently described by , who showed that the cored succession ranges from the middle part of the Darriwilian Stage into the lowermost Sandbian Stage (Fig. 2). It is conceivable that also the basal part of the Darriwilian Stage is present in the Röstånga area, but in the absence of outcrops of this interval, the only way to verify this idea is by a new drilling, or by deepening the Röstånga2 drill hole.
Assuming that the Röstånga phosphorite bed, identified as the Fågelsång Phosphorite, is coeval with that bed at Fågelsång (Bergström et al. 2000), the top of the Darriwilian Stage in the Röstånga succession would be somewhat below the phosphorite bed. According to Bergström et al. (2000) and Bergström & Ahlberg (2004), the base of the Nemagraptus gracilis Zone should be found in the uppermost part of the Almelund Shale, as it was defined in the GSSP section at Fågelsång.
The graptolite faunas of the Almelund Shale in the Röstånga2 drill core show the presence of the Darriwilian Holmograptus lentus, Nicholsonograptus fasciculatus, Pterograptus elegans, Pseudamplexograptus distichus and Jiangxigraptus (formerly Dicellograptus) vagus zones , that is, the same graptolite zone succession as at Fågelsång (Fig. 2). The index fossil of the  basal Darriwilian Levisograptus austrodentatus Zone has not yet been found in Scania, but the index species of two of its subzones have been recorded just below the Komstad Limestone at Fågelsång (Bergström et al. 2018;. The corresponding lowermost Darriwilian strata are not yet known from Röstånga.
The lowermost portion of this drill core, which includes the drill core depth from 89.1 m to 116.25 m, is referred to the Holmograptus lentus Zone. Using an average dip of 38 degrees, the stratigraphical thickness of this interval is calculated to be ~21.5 m. Graptolites are generally not very well preserved in this interval, but the fauna includes, apart from the zone index, pendent didymograptids, phyllo graptids identified as Pseudophyllograptus sp. and Aulograptus climacograptoides (see .
The base of the overlying Nicholsonograptus fasciculatus Zone is placed at the level of the first occurrence of the zone index, which is at a core depth of 89.1 m. The top of this zone is taken at the level of the first occurrence of Pterograptus elegans, which is at a core depth of 74.88 m. Using 38 degrees as the average dip of these strata, the total stratigraphical thickness of this zone is ~11 m. Graptolites recorded from this zone include, among others, Cryptograptus schaeferi, Proclimacograptus angustatus and specimens of Haddingograptus and Archiclimacograptus .
As just noted, the base of the overlying Pterograptus elegans Zone is taken to be at the core level of 74.88 m. In this drill core, this zone ranges up to a core level of 26.02 m and has a calculated total stratigraphical thickness of ~39 m. Most of this zone along the Church Brook was excavated by Ekström (1937), who estimated its thickness there to be ~15 m, which includes a severalmetrethick unexposed interval. The graptolite fauna of this zone includes slender didymograptids, such as Didymograptus artus, and more robust specimens, such as Didymograptus murchisoni.
The base of the overlying Pseudamplexograptus distichus Zone is placed at the level of the first occurrence of the zone index species at a core depth of 26.02 m. The top of the zone is marked by the first appearance of the zone index Jiangxigraptus vagus at a core depth of 20.7 m. The stratigraphical thickness of this zone is ~4.4 m. In the past, this interval has frequently been referred to as the Glyptograptus teretiusculus Zone, but the extended vertical range of the zone index species into both older and younger strata, along with the unclear morphology of the type specimens of this species (cf. Jaanusson 1960, p. 322), have made it appropriate to abandon this species as a zonal index. Herein, we follow Maletz et al. (2007), Chen et al. (2016),  and  in using Jiangxigraptus vagus as a zone designation. In the Röstånga2 drill core, the graptolite fauna of this zone is relatively diverse and includes, among others, Dicranograptus irregularis and Gymnograptus ejuncidus (see . The top of the Jiangxigraptus vagus Zone is currently not well defined in the Röstånga2 drill core succession because the stratigraphically oldest specimens of the index of the next higher graptolite zone, the Nemagraptus gracilis Zone, have been found well above the base of the Sularp Shale at the 13.22 m drill core level . Taking the zone base at that level would be in conflict with the conditions at Fågelsång where the base of this zone is 1.4 m below the base of the Fågelsång Phosphorite (Bergström et al. 2000). As is the case at Fågelsång, the graptolite fauna of the uppermost part of the Almelund Shale is poor and the species found in the new drill core are not diagnostic of a specific zone. However, judging from the δ 13 C chemostratigraphy and a comparison with the carbon isotope curve from Fågelsång (Bergström et al. 2018(Bergström et al. , 2020, it would appear that the base of the Nemagraptus gracilis Zone corresponds to a 125 S. M. Bergström et al.: Middle Ordovician stratigraphy in Scania, Sweden   Fig. 3. Photo of the Fågelsång Phosphorite from just above the 20m level in the Röstånga2 drill core. This is the first record of this lithologically characteristic bed in the Röstånga area. The arrow is placed just above the top surface of the bed. Note the conglomeratic nature of especially the lower part of this bed and the apparent disconformity surface in its upper part (marked by arrowheads). level well below the first occurrence of the zone index in also the Röstånga2 drill core.
It should be noted that the Darriwilian graptolite faunas from Röstånga have much in common with the coeval ones from Fågelsång (cf.  and there is also a close similarity to the graptolite faunas of the Albjära drill core faunas recorded by Maletz (1995). The latter core was estimated to include about 75 m of the Almelund Shale (identified as the Elnes Formation by Maletz 1995). Based on Ekström's (1937) records, the middle-upper Darriwilian interval has yielded 29 graptolite species at Röstånga and 33 species at Fågelsång among which 27 (93%) are in common. Although these figures may change somewhat after a modern revision of the faunas, there is no doubt that these faunas have essentially the same composition. This is not surprising in view of the relatively close distance between these areas and their closely similar depositional environments.
In view of the great similarity between the Darriwilian graptolite faunas from Röstånga and Fågelsång, it is obviously of interest to compare the thickness of the graptolite zones in the two regions (Fig. 4). Obviously, without consideration of the dip, such a thickness comparison is meaningless. As is shown in the literature (e.g. Hede 1951), the dip varies through the Fågelsång Darriwilian succession but we have adopted a figure of 10 degrees as an average in our attempt to calculate the stratigraphical thickness of the graptolite zones in our study successions in that area. As shown in Fig. 4, the stratigraphical thickness of the Pseudamplexograptus distichus and Nicholsonograptus fasciculatus zones is about 4-7 m and 10 m, respectively in the two successions, whereas the Pterograptus elegans Zone has a stratigraphical thickness of about 8 m in the Fågelsång3 drill core, but close to 40 m in the Röstånga2 succession. We consider the latter figure to be greatly inflated and probably an artifact of thickness duplication by faulting, although this is not obvious from the isotope curve shown in Fig. 5. It should be noted that in Ekström's (1937) excavations along the Church Brook, the total estimated thickness of this graptolite zone would be no more than ~15 m, including an estimated 5 m thick unexposed interval. This estimated thickness of the Pterograptus elegans Zone is not much different from that in the Fågelsång3 drill core indicated above.

Samples and laboratory analysis
Among 233 samples collected from the drill core interval 1.49-116.90 m, 230 yielded useful carbon isotope data (Table 1). Assuming an average dip of 38 degrees, the sampled interval would have a stratigraphical thickness of approximately 90 m. In most cases sampling took place at 50 cm drill core intervals which corresponds to a stratigraphical thickness of 40 cm. The samples were subjected to the same laboratory treatment as those from the Fågelsång3 drill core as described by Bergström et al. (2018). Carbon isotope analyses of organic carbon were performed using a Flash EA 2000 elemental analyser connected online to a ThermoFisher Delta V Plus mass spectrometer at the Geocentre of Northern Bavaria,

Results
The stratigraphically closely controlled δ 13 C org curve is illustrated and compared with the isotope record from the Fågelsång3 drill core in Fig. 5. Although at first look appearing rather smooth and without prominent isotopic excursions, the δ 13 C org curve exhibits some notable features. Also, it is of significant interest to compare this isotope curve with that of the approximately same stratigraphic interval from the Fågelsång3 drilling (see Bergström et al. 2018 Albanesi et al. 2013;Wu et al. 2017;Lindskog et al. 2019). In both the Fågelsång succession and Röstånga2 drill core (Fig. 6), this excursion interval is followed by several metres with declining isotopic values just above the Fågelsång Phosphorite. In a recent study, Bergström et al. (2020) identified this negative excursion as equivalent to what Kaljo et al. (2007) in Estonia referred to as the 'Upper Kukruse low', which occurs in the Baltoniodus variabilis Subzone of the Amorphognathus tvaerensis conodont Zone and in the lower part of the Nemagraptus gracilis graptolite Zone. In their study of the chemostratigraphy of the Tartu453 drill core from southern Estonia, Bauert et al. (2014) renamed this isotopic perturbation the Lower Sandbian Negative Isotope Carbon Excursion (LSNICE). This excursion has potential as a useful stratigraphic marker for the lower Sandbian. As shown in Fig. 6, the onset of the LSNICE is displayed in the uppermost part of the Röstånga2 drill core where it shows isotopic values down to at least -30.5‰, which is of nearly the same magnitude as the isotopic values of the corresponding interval at Fågelsång. Although it has not been linked to any known global biogeochemical event, it may be noted that the LSNICE appears to be geographically widespread. For instance, in the Yichang area, China, it is displayed in the Nemagraptus gracilis Zone of the Zhenjin and Jieling sections (Ma et al. 2015).
In summary, the isotope chemostratigraphy is in agreement with the interpretation that there is a close similarity between the Röstånga and Fågelsång successions in the interval around the base of the Sandbian Stage, and the carbon isotope record supports the idea that the base of the Nemagraptus gracilis Zone corresponds to a level well below the appearance of the zone index in the Röstånga2 drill core.

CONCLUDING REMARKS
Although the widely cited study by Underwood et al. (1997) on the δ 13 C org chemostratigraphy of the Ordovician-Silurian boundary interval at Dob's Linn, South Scotland, was published more than 20 years ago, and has been followed by several δ 13 C org investigations of Upper Ordovician (e.g. Bergström et al. 2014 DE"CB# D="BB# D="CB# D?"BB# D?"CB# AB"BB# AB">?# @D?"=># @D?"FE# @D?"EA# @D?"==# @D?"=?# @D?"E?# @D?"?=# CE"BB# CE"CB# C="BB# C="CB# C?"BB# C?"CB# # @D?"?F# @D?"?<# @D?"EF# @D?"=?# @D?"E<# @D?"=?# # # =C"??# =F"CB# =F"??# =E"C<# =="BB# =="C<#GHI# @AB">C# @AB"<C# @AB"BC# @AB"BF# @AB"A># @AB"<B# # <<>">E# <<>"?E# <<C"CB# <<F"BB# <<F">?# <<F"?B# # # @D?"EB# @D="?A# @D="D<# @D?"<E# @D?"<># @D?"AC# # =="C<#GHI# # Darriwilian carbonate successions (e.g. Young et al. 2016), such studies of Ordovician graptolitebearing dark shale successions are still far less common than those based on the δ 13 C carb . This is especially the case in the Darriwilian interval where, as far as we are aware, Zhang et al. (2010) and Bergström et al. (2018) Fig. 6. Comparison of the uppermost Darriwilian-lowermost Sandbian interval (from the top of the Pseudamplexograptus distichus Zone to the lower Nemagraptus gracilis Zone) in the Röstånga2 drill core and at Fågelsång. The latter is a composite of part of the Fågelsång3 drill core (Bergström et al. 2018, fig. 5) and the E14b outcrop (Bergström et al. 2020, fig. 5), this outcrop section being the GSSP of the Global Sandbian Stage. Note the general similarity between the isotope curves. Three apparently anomalous data points just above the Fågelsång Phosphorite in the Röstånga2 drill core (see Table 1) are omitted in this figure. The isotope curves suggest that in terms of δ 13 C chemostratigraphy, the base of the Sandbian Stage occupies a position between the positive curve perturbation referred to as MDICE and the negative one known as LSNICE. P. d., Pseudamplexograptus distichus. the first time, the distinctive Fågelsång Phosphorite is identified at Röstånga, where it appears to occupy the same stratigraphic position just above the base of the Sandbian Stage as at Fågelsång. This suggests that this characteristic bed may have a much wider geographic distribution than has been known previously. 3. Based on 230 samples the δ 13 C org chemostratigraphy has been established through the entire Röstånga2 drill core. Two excursions are recognized, namely the positive MDICE in the middle-upper Darriwilian and the negative LSNICE in the lower Sandbian. As a whole the chemostratigraphy is closely comparable to that of the coeval interval at Fågelsång. 4. The Röstånga2 chemostratigraphy has been tied to the graptolite biostratigraphy. The relations between these stratigraphies are basically the same as at Fågelsång. 5. Both the graptolite biostratigraphy and the chemostrati graphy indicate that the end of the Röstånga2 drill core is somewhat above the stratigraphic interval corresponding to the lower Darriwilian Komstad Limestone. Hence, it remains uncertain if the latter calcareous unit is present at Röstånga. 6. To date, the Röstånga2 and Fågelsång3 drill cores represent the only stratigraphically reasonably complete Darriwilian shale successions in the world in which the δ 13 C org chemostratigraphy is directly tied to the graptolite biostratigraphy.