(2006), Pp. 85–90. Article

GFF volume 128 (2006), pp. 85–90. Article

13 The Late Wenlock Mulde positive carbon isotope (δ Ccarb) excursion in North America

BRADLEY D. CRAMER1, MARK A. KLEFFNER2 and MATTHEW R. SALTZMAN1

Cramer, B.D., Kleffner, M.A. & Saltzman, M.R., 2006: The Late Wenlock Mulde positive carbon isotope excursion in North America. GFF, Vol. 128 (Pt. 2, June), pp. 85–90. Stockholm. ISSN 1103-5897. Abstract: Marine carbonates from two well-studied areas of the Silurian of North America were analyzed 13 for stable carbon isotope (δ Ccarb) stratigraphy. A graptolite-bearing sequence from the eastern margin of the Panthallasic Ocean (Nevada) and a conodont-bearing sequence from the mid-continent epeiric 13 seaway (Tennessee) were sampled for δ Ccarb stratigraphy in order to improve the correlation between these areas and the Swedish island of Gotland, which has become the global standard for Wenlock conodont and carbon isotope stratigraphy. The Homerian (Late Wenlock) Mulde positive carbon isotope excursion serves as a useful chronostratigraphic marker for Homerian sequences, especially in regions such as the two included in this investigation, where zonal fossils are absent or poorly represented. In Nevada, using presently available biostratigraphic data, a detailed modern graptolite zonation cannot be applied due to a lack of several key species. Likewise, the zonally important conodont species are poorly 13 represented in Tennessee. Our recognition of the dual-peaked Mulde δ Ccarb excursion in North America allows improved correlation between these sequences and any other locality where the Mulde Excursion has been recorded in sufﬁcient detail. Keywords: Wenlock, Ludlow, Silurian, carbon isotope, stratigraphic correlation, conodont, graptolite, extinction.

1 Department of Geological Sciences, The Ohio State University, 125 S. Oval Mall, Columbus, Ohio 43210, USA; [email protected]; [email protected] 2 Department of Geological Sciences, The Ohio State University at Lima, 4240 Campus Drive, Lima, Ohio 45804, USA; [email protected] Manuscript received 15 August 2005. Revised manuscript accepted 3 May 2006.

Introduction 13 The well-studied Swedish island of Gotland has become the The two protracted positive carbon isotope (δ Ccarb) excursions global standard for conodont biostratigraphy and carbon isotope (Ireviken and Mulde Excursions) can serve as reliable chronos- stratigraphy of the Wenlock (Jeppsson 1997; Bickert et al. 1997; tratigraphic markers for Wenlock successions. The Sheinwoodian Calner & Jeppsson 2003; Jeppsson & Calner 2003; Munnecke (Early Wenlock) Ireviken Excursion reaches values as high as et al. 2003; Calner et al. 2004; Calner et al. 2005; Calner et al. +5.00‰ (VPDB) and spans the lower half of the Sheinwoodian this volume). The numerous carbon isotope investigations of (Bickert et al. 1997; Munnecke et al. 2003; Cramer & Saltzman Wenlock sequences have promoted global climate research sur- 2005; Calner et al. 2006; Kaljo & Martma this volume). The rounding the two major positive carbon isotope excursions that Homerian (Late Wenlock) Mulde Excursion is a dual-peaked took place during the Wenlock (see Calner et al. 2004) however excursion with values approaching +4.00‰ (VPDB) in some the inability to precisely correlate sections outside of Baltica localities (Corﬁeld et al. 1992; Por´bska et al. 2004; Calner et with the Gotland sequence has hindered accurate global climate al. 2006) and is associated with the ʻBig Crisisʼ graptolite ex- reconstructions of this unstable phase of the Silurian. In many tinction (“der Grossen Krise” Jaeger 1959; 1991; Quinby-Hunt sections world-wide, available conodont data is often insufﬁ- & Berry 1991). Here, we sampled marine carbonates from a cient to use the detailed zones recognized on Gotland (Calner & graptolite-bearing sequence in central Nevada (Simpson Park Jeppsson 2003). Likewise, the classical data available for many I), and a conodont-bearing sequence in north-central Tennessee 13 graptolitic sections do not contain all of the now recognized (McCrory Lane) at regular stratigraphic intervals for δ Ccarb 13 zonal species (e.g. Por´bska et al. 2004). In instances such as stratigraphy. By locating the Mulde δ Ccarb excursion in the this, where biostratigraphic correlation between geographically study areas, we can constrain more precisely the timing of global distant sections that often lack key species leaves a measure of climate changes recorded in strata deposited during this interval uncertainty, another means of correlation is desirable. of the Silurian. 13 86 Cramer et al.: The Late Wenlock Mulde positive carbon isotope (δ Ccarb) excursion in North America GFF 128 (2006)

by assigning a δ13C value of +1.95‰ and a δ18O value of –2.20‰ to NBS19. Reproducibility was checked by replicate analysis of laboratory standards and is better than ±0.05‰. Samples were analyzed at the Erlangen University stable isotope laboratory in Germany.

Results and Discussion Simpson Park I, Nevada The well-known graptolite-bearing succession of the Roberts Mountain Formation in the Simpson Park Range, Nevada, was studied in detail for graptolite biostratigraphy by Berry & Murphy (1975). The Mulde Excursion is well developed in the Fig. 1. Wenlock paleogeography (after Woodcock 2000; Cocks 2001; Roberts Mountain Formation with baseline values of +0.50‰ Johnson et al. 2001). Grey areas denote land. Capital letters refer to and highs approaching +2.50‰ during both peaks of the excur- major emergent landmasses: L – Laurentia; B – Baltica; S – Siberia; G sion (Fig. 2; Table 1). The graptolite data of Berry & Murphy – Gondwana. Arrows point to locations of the two sampled sections in Nevada and Tennessee. (1975) allow the recognition of the Monograptus testis and M. flemingii-P. dubius Zones of Por´bska et al. (2004), but the rest of the zonally useful species for higher in the sequence were not reported from Nevada with the exception of G. nassa and Pristiograptus ludensis (Berry & Murphy 1975). The carbon isotope curve follows the graptolite correlation of Por´bska et al. (2004) in the lower portions of the sequence Geological background where zonally useful species are abundant. As predicted by the The two sampled sections were located on the paleocontinent of correlation of Por´bska et al. (2004), the most rapid positive rate 13 Laurentia, with Nevada (Simpson Park I) located on the western of change during the first peak in δ Ccarb took place shortly after margin facing the Panthallasic Ocean, and Tennessee (McCrory the disappearance of the M. flemingii, and the first peak occurred Lane) located towards the southern end of the mid-continent near the FAD of G. nassa. Graptolites present higher in the se- epeiric sea (Fig. 1). Both sections were located between 15° and quence are problematic however, when compared with the carbon 35° S latitude (Witzke & Scotese 1990; Woodcock 2000). isotope data. From the first peak and the first positive identifica- The Mulde Excursion has been well-correlated to both grap- tion of G. nassa, the excursion extends more than 10 meters up tolite and conodont biostratigraphy (Jeppsson & Calner 2003; section before the end of the second peak. The problem arises Calner & Jeppsson 2003; Por´bska et al. 2004; Calner et al. with the occurrence of Neodiversograptus nilssoni, which marks 2006). In graptolite-bearing sequences, the fastest rate of posi- the base of the Ludlow Series, only 2.4 meters above G. nassa 13 tive change in δ Ccarb occurs shortly after the disappearance of and well within the center of the Mulde Excursion as recorded in Monograptus flemingii with the first peak occurring near the first this section. Previously, the Mulde Excursion was considered to appearance datum (FAD) of Gothograptus nassa. The low-point have occurred entirely within the Wenlock (Corfield et al. 1992; between the peaks occurs near the boundary between the Pristi- Por´bska et al. 2004; Calner et al. 2004). Our data therefore in- ograptus dubius-G. nassa Zone and the overlying Colonograp- dicate one of several possibilities; (1) that the Mulde Excursion tus praedeubeli Zone. The second peak appears to be restricted actually crossed the Wenlock-Ludlow boundary and should be primarily to the C. praedeubeli Zone (Por´bska et al. 2004). In re-correlated with respect to Silurian time and that all other sec- conodont-bearing sequences, the first peak occurs in the Ozarko- tions where the Mulde Excursion has been recorded need to be dina bohemica longa Zone with the low-point coming near the likewise adjusted; (2) that N. nilssoni was misidentified by Berry boundary with the overlying Kockelella ortus absidata Zone. & Murphy (1975); (3) that N. nilssoni first occurred in Nevada The second peak is restricted to the Ctenognathodus murchisoni substantially earlier than elsewhere in the world; or (4) that a Zone (Calner et al. 2005). significant break in deposition occurred just below the first occurrence of N. nilssoni cutting out the second peak of the Mulde Excursion and serendipitously replacing it with another isotopic Methods excursion of similar magnitude in the Early Ludlow. Marine carbonates have been shown to faithfully preserve the None of the four possible explanations suggested for the prob- 13 13 primary marine carbon isotopic δ Ccarb signature (compare lematic second peak of the δ Ccarb curve in the Simpson Park I Munnecke et al. 1997 with Bickert et al. 1997), and the use of section clearly best explains the distribution of N. nilssoni rela- fine-grained carbonates for carbon isotope stratigraphy has been tive to it in that section: (1) Based on the work in the classical demonstrated in every system of the Paleozoic (Saltzman 2005). British sections by Corfield et al. (1992), and the fact that nei- Analyzed powders were micro-drilled from micritic matrix ther N. nilssoni, nor any other graptolite indicative of the Early whenever possible, but some samples were too coarse grained Ludlow, was found during the Mulde Excursion in Poland by to guarantee that no skeletal material was sampled. Carbonate Por´bska et al. (2004), the first option seems unlikely; (2) The powders were reacted with 100% phosphoric acid (density >1.9; fact that Bohemograptus bohemicus, Saetograptus colonus, and Wachter & Hayes 1985) at 75°C using a Kiel III online carbonate Monograptus uncinatus, all of which are indicative of the Early preparation line connected to a ThermoFinnigan 252 mass spec- Ludlow, share an FAD with N. nilssoni in the Simpson Park I trometer. All values are reported in ‰ notation relative to V-PDB section makes the misidentification possibility unlikely; (3) 13 GFF 128 (2006) Cramer et al.: The Late Wenlock Mulde positive carbon isotope (δ Ccarb) excursion in North America 87

Fig. 2. Stable carbon isotope data from the Simpson Park I section, Coal Canyon, Nevada. Graptolite information from Berry & Murphy (1975), shown with the feet above the base of section they reported. Our sampling datum is the end of the bulldozer road that runs from ~170 m to 300 m in their section. LADs of M. testis and M. flemingii flemingii shown to the right of the stratigraphic column. The sole occurrences of G. nassa and P. ludensis are also illustrated. B. bohemicus, N. nilssoni, S. colonus, and P. dubius frequens occurrences shown are all FADs. Chronostratigraphy shown is based on the first occurrence of N. nilssoni in the section, see text for further discussion. M. testis and M. flemingii Zones shown sensu Por´bska et al. (2004). The first peak of the excursion matches well, biostratigraphically, what was expected based on Por´bska et al. (2004), but the occurrence of N. nilssoni before the second peak raises concern regarding the Mulde Excursion in Nevada.

Table 1. Stable C and O isotope data. Simpson Park 1, Coal Canyon, Nevada. Concerning extending the range of N. nilssoni into the Wenlock, meter* δ13C δ18O Formation the three aforementioned species which share an FAD with N. carb –10 0.62 –6.90 Roberts Mountain nilssoni would also have their ranges extended if this were the –17 0.57 –6.83 Roberts Mountain case, which seems unlikely; (4) A break in deposition was not –22 0.92 –8.40 Roberts Mountain noted by Berry & Murphy (1975), nor was one observed during –24 0.50 –6.32 Roberts Mountain –25 –0.41 –6.91 Roberts Mountain our field work. The fact that all four Ludlow graptolite species –28 –0.59 –9.63 Roberts Mountain first occur at the same datum does lend support to the possibility –29 1.85 –6.86 Roberts Mountain –30 1.96 –7.76 Roberts Mountain of a break in deposition immediately below their FAD. If that –31 1.76 –9.41 Roberts Mountain is the case however, the carbon isotope curve requires special –32 2.30 –7.42 Roberts Mountain pleading. –33 2.33 –6.97 Roberts Mountain –34 2.58 –7.77 Roberts Mountain The Mulde Excursion is a dual-peaked excursion, which is –35 2.45 –12.69 Roberts Mountain something rarely seen in the Paleozoic. Only during the Mulde –36 0.94 –11.96 Roberts Mountain Excursion and the Frasnian–Famennian boundary excursion in –37 0.65 –9.67 Roberts Mountain –38 0.52 –7.69 Roberts Mountain the Devonian (e.g. Joachimski & Buggisch 1993) is there an ini- –39 0.89 –6.92 Roberts Mountain tial peak which decays nearly to pre-excursion baseline values –40 0.62 –8.10 Roberts Mountain δ13 –41 2.14 –12.99 Roberts Mountain creating a significant trough in the Ccarb curve, before reaching –44 2.28 –14.06 Roberts Mountain a second peak of similar magnitude. Our data show what appears –45 1.46 –14.72 Roberts Mountain to be this type of dual-peaked excursion. Although it does seem –48 0.82 –6.53 Roberts Mountain –49 0.38 –7.49 Roberts Mountain unlikely that the second peak of our data is not that of the Mulde –50 0.57 –7.71 Roberts Mountain Excursion, we cannot rule out the possibility that the second peak –51 0.45 –9.14 Roberts Mountain 13 –52 0.82 –7.72 Roberts Mountain in our data is the Ludlow age ʻLindeʼ δ Ccarb excursion, the next 13 –53 0.73 –9.11 Roberts Mountain stratigraphically higher positive δ Ccarb excursion (Early Lud- –54 0.74 –8.45 Roberts Mountain low, Middle to Late Gorstian; Munnecke et al. 2003; Calner et al. –55 0.58 –7.26 Roberts Mountain 2004; Calner et al. 2005), which has never been documented in –56 1.14 –7.09 Roberts Mountain 13 –57 0.57 –6.75 Roberts Mountain δ Ccarb from fine-grained carbonate samples and is relatively un- –58 0.57 –6.34 Roberts Mountain known by comparison to other Silurian excursions. Furthermore, –59 0.34 –6.39 Roberts Mountain the second peak may represent some previously unrecognized –60 0.36 –7.00 Roberts Mountain δ13 * Sampling datum is the end of the bulldozer road along wich the measured global Ccarb excursion; or an unexpected basin-scale isotopic section of Berry and Murphy (1975) runs from approx. 170 m to 300 m. The feature of Ludlow age that occurs in precisely the position we end of the road occurs at c. 170 m in their measured section. 13 88 Cramer et al.: The Late Wenlock Mulde positive carbon isotope (δ Ccarb) excursion in North America GFF 128 (2006)

Fig. 3. Stable carbon isotope data from the McCrory Lane section, Nash- ville, Tennessee. Conodont occurrence from Barrick (1983). Sampling datum is the base of the middle carbonate unit within the Waldron Mem- ber (Waldron [b]; Fig. 4). Note the close agreement between the two separate data-sets included in the figure (grey triangles, Table 1; black circles, Table 2). We chose to show these as two separate data-sets to 13 Fig. 4. Photograph of the McCrory Lane section to demonstrate our help demonstrate the reproducibility of δ Ccarb stratigraphy as a reliable chronostratigraphic marker during times of major δ13C excursions. informal terminology of the Waldron. We have identified two shaley carb intervals that bracket a clean carbonate interval within the Waldron. The first peak of the Mulde Excursion occurs in the basal shaley unit (Wal- dron [a]), with the trough during the carbonate (Waldron [b]), and the second peak within the upper shaley interval (Waldron [c]). Note that in some regions of the mid-continent of North America the units Waldron [b] and Waldron [c] discussed herein, have been variably included in overlying formations, often due to poor development of the second shale interval (Waldron [c]). expected to see the second peak of the Mulde Excursion due to a break in deposition during the Late Homerian (and Earliest Gorstian?). Another possibility we must investigate is that the second peak of our data represents the Middle-Late Ludlow ʻLauʼ Excursion 13 however biostratigraphic and δ Ccarb data (Pete Hanson Creek II, Saltzman 2001) from the nearby Roberts Mountains in Nevada middle portion of the Saltzman (2001) curve, we feel that we can 13 13 prohibit this interpretation. The Lau δ Ccarb excursion is associ- demonstrate the second peak of our data is not the Lau δ Ccarb ated with a minor extinction event (Calner 2005a) marked by the excursion. disappearance of the conodont Polygnathoides siluricus. In the The problematic distribution of early Ludlow graptolites within Roberts Mountains Pete Hanson Creek II section, P. siluricus, the apparent second peak of the Mulde Excursion in the Simpson 13 which went extinct during the rising limb of the Lau δ Ccarb Park I section requires an integrated, modern, high-resolution excursion (Saltzman 2001; Calner 2005b), was recovered above study of graptolites and carbon isotope stratigraphy from parallel the first occurrence of the graptolite Saetograptus chimaera samples in that section. (Berry & Murphy 1975). As shown in Saltzman (2001), the Lau Excursion occurred within the range of S. chimaera in Nevada. Although P. siluricus has not been recovered from the Simp- McCrory Lane, Tennessee son Park I section, S. chimaera has a first occurrence at 497ft The McCrory Lane section was sampled for conodont bios- above the base of the section, which is roughly 10 m above the tratigraphy by Barrick (1983). The comparatively low-resolu- stratigraphically highest sample in this investigation. Therefore, tion North American conodont zonation assigned the upper due to the biostratigraphy and the similarity of our data with the Maddox and the Waldron Members of the Wayne Formation to 13 GFF 128 (2006) Cramer et al.: The Late Wenlock Mulde positive carbon isotope (δ Ccarb) excursion in North America 89

13 Fig. 5. δ Ccarb based isotopic correlation of lithostratigraphic units between Tennessee and Gotland shown against the McCrory Lane 13 δ Ccarb curve and integrated graptolite and conodont biostratigraphy. Isotopic correlation of Gotland stratigraphy after the Hunninge-1 Drillcore (Fig. 1 in Calner et al. this volume). Placement of graptolite biostratigraphy against the isotope curve accomplished via Por´bska et al. (2004) and Calner et al. (2006). Boxes in graptolite zonation labeled 1 and 2 refer to extinction levels (Calner & Jeppsson 2003; Calner et al. 2006). Conodont biostratigraphy from Gotland after Calner & Jeppsson (2003), with minor changes. Placement of the base of the Klinteberg Formation with respect to the base of the Lego Formation remains unclear due to limited isotope data from higher in the section on Gotland and limited conodont data from Nashville. It may be that the base of the Klinteberg Fm. is coincident with the base of the Lego Fm. making the Djupvik Mbr. and expanded section roughly equivalent to the Waldron [c]. the Kockelella amsdeni Zone with the boundary to the overlying conodonts Ozarkodina bohemica longa and Kockelella ortus Kockelella stauros Zone tentatively placed low in the overlying absidata occur in the Lego, while the Early Ludlow species Lego Formation. None of the zonally useful species utilized in Kockelella variabilis occurs near the top of the Lego (J. Barrick, the biostratigraphic scheme erected for Gotland were reported pers. comm. 2005). from the section sampled in this investigation, however some The Mulde Excursion is well represented in this section with of them have been documented elsewhere in Tennessee. The baseline values of +1.00‰ rising gently through the upper conodont Pseudooneotodus linguicornis, which is indicative of the Early Homerian Ozarkodina sagitta sagitta Zone (Calner & Jeppsson 2003), occurs in the upper Maddox of western Tennes- see (J. Barrick, pers. comm. 2005). In western Tennessee, the Table 3. Stable C and O isotope data. McCrory Lane, Nashville, Tennessee. 13 18 meter* δ Ccarb δ O Formation 2.5 1.76 –4.42 Lego 2.3 2.07 –3.80 Lego 2.1 2.31 –3.38 Lego Table 2. Stable C and O isotope data. McCrory Lane. Nashville. Tennessee. 2 2.43 –3.10 Waldron [c] meter* δ13C δ18O Formation 1.7 2.17 –3.33 Waldron [c] carb 1.6 2.02 –3.14 Waldron [c] 7.5 0.83 –4.01 Lego 1.5 2.41 –2.97 Waldron [c] 7.25 0.99 –4.19 Lego 1.4 2.16 –3.32 Waldron [b] 7 0.77 –4.25 Lego 1.2 1.86 –3.71 Waldron [b] 6.75 0.77 –4.42 Lego 1.1 1.59 –3.49 Waldron [b] 6.5 0.92 –4.16 Lego 0.9 1.39 –4.23 Waldron [b] 6.25 0.86 –4.16 Lego 0.7 1.61 –3.76 Waldron [b] 6 0.93 –3.79 Lego 0.5 1.63 –3.52 Waldron [b] 5.75 0.97 –3.89 Lego 0.3 1.82 –3.89 Waldron [b] 5.5 0.83 –4.30 Lego 0.1 2.18 –3.21 Waldron [b] 5.25 0.83 –4.27 Lego 0 2.30 –2.98 Waldron [b] 5 1.02 –3.92 Lego –0.05 2.31 –3.27 Waldron [a] 4.75 1.02 –4.35 Lego –0.1 2.60 –2.73 Waldron [a] 4.5 0.98 –5.80 Lego –0.3 2.10 –2.87 Waldron [a] 4.25 1.19 –4.05 Lego –0.6 2.63 –2.96 Waldron [a] 4 1.21 –4.57 Lego –0.8 2.59 –2.93 Waldron [a] 3.75 1.36 –4.36 Lego –0.9 2.82 –3.55 Waldron [a] 3.5 1.36 –4.23 Lego –1.25 2.42 –2.84 Maddox 3.25 1.17 –4.48 Lego –1.35 1.68 –4.05 Maddox 3 1.55 –3.95 Lego –1.45 1.21 –3.93 Maddox 2.75 1.93 –3.34 Lego –1.55 1.03 –4.47 Maddox 2.5 1.63 –3.98 Lego –1.65 1.44 –4.41 Maddox 2 2.12 –3.26 Waldron [c] –1.85 1.28 –3.55 Maddox 1.5 2.00 –3.21 Waldron [c] –2.05 1.23 –3.67 Maddox 1.25 2.37 –3.17 Waldron [b] –2.25 1.25 –3.73 Maddox 1 1.70 –3.51 Waldron [b] –2.45 1.32 –3.90 Maddox 0.75 1.62 –3.93 Waldron [b] –2.65 1.09 –4.38 Maddox 0.5 1.71 –3.63 Waldron [b] –2.85 1.02 –3.92 Maddox 0.25 1.94 –3.53 Waldron [b] –3.05 1.00 –4.70 Maddox 0 2.18 –3.19 Waldron [b] –3.35 0.94 –4.01 Maddox * Sampling datum is the base of the unit Waldron [b]. Members listed are * Sampling datum is the base of the unit Waldron [b]. Members listed are within the Wayne Formation. within the Wayne Formation. 13 90 Cramer et al.: The Late Wenlock Mulde positive carbon isotope (δ Ccarb) excursion in North America GFF 128 (2006)

Maddox before the maximum rate of positive change occurs in References the uppermost Maddox samples (Fig. 3). The first peak occurs Barrick, J.E., 1983: Wenlockian (Silurian) conodont biostratigraphy, biofacies, within the lower shaley unit of the Waldron (Waldron [a]; Fig. and carbonate lithofacies, Wayne formation, Central Tennessee. Journal of 4) with values approaching +3.00‰. The low-point between the Paleontology 57, 208–239. Berry, W.B.N. & Murphy, M.A., 1975: Silurian and Devonian graptolites of two peaks occurs within the central limestone of the Waldron central Nevada. University of California Publications in Geological Sciences (Waldron [b]) before the second peak occurs within the upper 110, 109 pp. shaley unit (Waldron [c]). δ13C values steadily decline in the Bickert, T., Pätzold, J., Samtleben, C. & Munnecke, A., 1997: Paleoenvironmen- carb tal changes in the Silurian indicated by stable isotopes in brachiopod shells lower Lego returning to a baseline near +1.00‰. The two data from Gotland, Sweden. Geochimica et Cosmochimica Acta 61, 2717–2730. sets shown in Fig. 3 represent two separate field trips and two Calner, M., 2005a: A Late Silurian extinction event and anachronistic period. Geology 33, 305–308. separate sample sets run through the lab. The stratigraphically Calner, M., 2005b: Silurian carbonate platforms and extinction events – ecosys- higher data (diamonds in Fig. 3; Table 2) were sampled and ana- tem changes exemplified from Gotland, Sweden. Facies 51, 603–610. lyzed in 2003 in a preliminary search for the Mulde Excursion Calner, M. & Jeppsson, L., 2003: Carbonate platform evolution and conodont stratigraphy during the middle Silurian Mulde Event, Gotland, Sweden. Geo- in Tennessee strata. The stratigraphically lower data (circles in logical Magazine 140, 173–203. Fig. 3; Table 3) were sampled and analyzed in 2005 in order to Calner, M., Jeppsson, L. & Munnecke, A., 2004: The Silurian of Gotland - Part pinpoint the onset of the Mulde Excursion. The close agreement 1: Review of the stratigraphic framework, event stratigraphy, and stable car- 13 bon and oxygen isotope development. Erlanger Geologische Abhandlungen in δ Ccarb values between the two sample sets where they over- Sonderband 5, 113–131. lap demonstrates the reproducibility of high-resolution carbon Calner, M., Jeppsson, L. & Eriksson, M.E., 2005: The Baltic basin and the Silurian strata of Gotland, Sweden. In M.E. Eriksson & M. Calner (eds.): The Dynamic isotope stratigraphy. Silurian Earth – Subcommission on Silurian Stratigraphy Field Meeting 2005. Sveriges geologiska undersökning, Rapporter och Meddelanden 121, 6–20. Calner, M., Koz∏owska, A., Masiak, M. & Schmitz, B., 2006: A shoreline to Summary deep basin correlation chart for the middle Silurian coupled extinction-stable isotopic event. GFF 128, 79–84. 13 The recognition of the Mulde positive δ Ccarb excursion in the Cocks, L.R.M., 2001: Ordovician and Silurian global geography. Journal of the study areas allows precise correlation of the studied sections Geological Society of London 158, 197–210. Corfield, R.M., Siveter, D.J., Cartlidge, J.E. & McKerrow, W.S., 1992: Carbon with the well-studied sequence on Gotland. There are no major isotope excursion near the Wenlock-Ludlow (Silurian) boundary in the Anglo- divisions of the Roberts Mountain Formation during this interval Welsh area. Geology 20, 371–374. Cramer, B.D. & Saltzman, M.R., 2005: Sequestration of 12C in the deep ocean of the Silurian, but the formal names given to strata in Tennessee during the early Wenlock (Silurian) positive carbon isotope excursion. Palaeo- can now be precisely correlated to those on Gotland (Fig. 5). geography, Palaeoclimatology, Palaeoecology 219, 333–349. In the case of Nevada, recognition of the Mulde Excursion has Jaeger, H., 1959: Graptolithen und stratigraphie des jungsten Thuringer Silurs. Abhandlungen Akademie Wissenschaften Berlin, Klasse Chemie, Geologie brought to light a concern regarding either the placement of the und Biologie 2, 197 pp. excursion within Silurian time or the possibility of a previously Jaeger, H., 1991: New standard graptolite zonal sequence after the “big crisis” at unrecognized carbon isotopic event. Recognition of the Mulde the Wenlockian/Ludlovian boundary (Silurian). Neues Jahrbuch für Geologie und Paläontologie, Abhandlungen 182, 303–354. Excursion in Nevada may also be serving as an independent test Jeppsson, L., 1997: A new latest Telychian, Sheinwoodian, and Early Homerian on the timing of the first occurrence of a species globally. In (Early Silurian) Standard Conodont Zonation. Transactions of the Royal Soci- ety of Edinburgh 88, 91–114. Tennessee, the recognition of the Mulde Excursion significantly Jeppsson, L. & Calner, M., 2003: The Silurian Mulde Event and a scenario for improves our ability to correlate stratigraphic sections across secundo-secundo events. Transactions of the Royal Society of Edinburgh 93, continents, and allows meaningful comparison of records of 135–154. Joachimski, M.M. & Buggisch, W., 1993: Anoxic events in the late Frasnian global climate change developed from the two geographically – Causes of the Frasnian-Famennian faunal crisis? Geology 21, 675–678. separated cratonic sequences of Gotland and Tennessee. For in- Johnson, M.E., Rong, J., Wang, C. & Wang, P., 2001: Continental island from the stance, the interpreted eustatic sea level rise beginning at the base Upper Silurian (Ludfordian Stage) of Inner Mongolia; implications for eustasy and paleogeography. Geology 29, 955–958. of the Bara Oolite Member of the Halla Formation on Gotland Kaljo, D. & Martma, T., 2006: Application of carbon isotope stratigraphy to dat- (Calner & Jeppsson 2003, and references therein), which has ing the Baltic Silurian rocks. GFF 128, 123–129. δ13 Munnecke, A., Westphal, H., Reijmer, J.J.G. & Samtleben, C., 1997: Microspar been tied to the highest rate of positive change in Ccarb values development during early marine burial diagenesis: a comparison of Pliocene during the first peak on Gotland (Calner et al. 2006), appears to carbonates from the Bahamas with Silurian limestones from Gotland (Swe- be supported by the deepening represented by deposition of the den). Sedimentology 44, 977–990. Munnecke, A., Samtleben, C. & Bickert, T., 2003: The Ireviken Event in the lower Waldron shale on top of crinoidal grainstones of the uppermost Silurian of Gotland, Sweden – relation to similar Palaeozoic and Proterozoic Maddox during the maximum rate of positive change during the events. Palaeogeography, Palaeoclimatology, Palaeoecology 195, 99–124. first peak of the Mulde Excursion reported herein. Por´bska, E., Koz∏owska-Dawidziuk, A. & Masiak, M., 2004: The lundgreni event in the Silurian of the East European Platform, Poland. Palaeogeography, Our investigation demonstrates the utility of high-resolution Palaeoclimatology, Palaeoecology 213, 271–294. 13 carbon isotope δ Ccarb stratigraphy for stratigraphic correlation Quinby-Hunt, M.S. & Berry, W.B.N., 1991: Late Wenlock (Middle Silurian) global bioevents: possible chemical cause for mass graptolite mortalities. in regions where biostratigraphic detail is insufficient, and also Historical Biology 5, 171–181. 13 13 the potential of δ Ccarb stratigraphy during major global carbon Saltzman, M.R., 2001: Silurian δ C stratigraphy: A view from North America. isotope excursions to serve as an independent test of the syn- Geology 29, 671-674. Saltzman, M.R., 2005: Phosphorus, nitrogen, and the redox evolution of the chronicity of FADs of Zone fossils in geographically disparate Paleozoic oceans. Geology 33, 573–576. regions. Wachter, E. & Hayes, J.M., 1985: Exchange of oxygen isotopes in carbon-diox- ide - phosphoric acid systems. Chemical Geology 52, 365–374. Acknowledgements. – We would like to thank T. Bickert and M. Calner for detailed and help- Witzke, B.J. & Scotese, C.R., 1990: Palaeoclimatic constraints for Palaeozoic ful reviews that made this a stronger manuscript. K. Tierney provided invaluable assistance in the field at both localities. M. Calner, J. Barrick, C. Brett, and P. McLaughlin are thanked palaeolatitudes of Laurentia and Euramerica. In C.R. Scotese (ed.): Palaeozoic for access to unpublished data and many enlightening conversations. We would also like to Palaeogeography and Biogeography. Geological Society of London Memoir thank M. Joachimski at the Stable Isotope Laboratory of the University of Erlangen. This 12, 57–73. investigation was partially funded by a Student Research Grant from the Geological Soci- Woodcock, N.H., 2000: Introduction to the Silurian. In R.J. Aldridge, D.J. Siveter, ety of America to B.D. Cramer, and National Science Foundation Grant number 0517929 D.J. Siveter, P.D. Lane, D. Palmer & N.H. Woodcock (eds.): British Silurian awarded to M.A. Kleffner. Stratigraphy, Geologic Conservation Review Series 19, 3–22.