Two new Emsian (Early Devonian) U-Pb zircon ages from volcanic rocks of the Rhenish Massif (Germany): implications for the Devonian time scale

Journal of the Geological Society, Mar 2005 by Kaufmann, B, Trapp, E, Mezger, K, Weddige, K

Abstract:

So far, only few biostratigraphically well-constrained isotopic ages constitute the geochronological framework of the Devonian period. To expand this sparse dataset, two new U-Pb isotope dilution-thermal ionization mass spectrometry (ID-TIMS) zircon ages were obtained from volcanic rocks of the Rhenish Massif (Germany). These rocks are interbedded in Emsian (Lower Devonian) marine sedimentary successions that are well documented by index fossils. Given that the U-Pb zircon data represent primary crystallization and eruption ages, they provide an important contribution to the Devonian time scale. The older sample, a stratiform volcaniclastic layer (Hans-Platte) is collected from Bundenbach (Hunsrück), the type locality of the Lower Emsian Hunsrück Slate. A cluster of five coherent single-zircon analyses obtained from this layer yielded a weighted mean ^sup 206^Pb/^sup 238^U age of 407.7 ± 0.7 Ma, which is regarded as the time of eruption of the pyroclastic rock. Biostratigraphically, the layer can be assigned to the upper part of the Polygnathus excavatus conodont zone. The younger U-Pb age was obtained from the Global Stratotype Section and Stratigraphic Point section at Wetteldorf (Eifel), where the K-bentonite 'Hercules I' is intercalated 13 m below the Lower-Middle Devonian boundary in the uppermost part of the Polygnathus costatus patulus conodont zone. Some of the 19 single-zircon analyses obtained from this bentonite appear to document significant influences of inheritance. The 13 youngest points are, however, concordant and form an elongated cluster along concordia with ^sup 206^Pb/^sup 238^U ages between 392.2 and 407.7 Ma. Given the (risky) assumption that this age scattering is exclusively caused by varying amounts of inheritance, the youngest point of this cluster represents the analysis with the least or no inheritance. In this case, its ^sup 206^Pb/^sup 238^U age of 392.2 ± 1.5 Ma should approach the eruptional age of the bentonite. The ages presented here in combination with U-Pb ID-TIMS zircon and monazite ages from other workers allow a recalibration of the early Emsian to early Eifelian interval of the Devonian time scale. The result is a duration of the Emsian stage (excluding the basal Emsian P. kitabicus conodont zone) of more than 17.2 Ma and an Emsian-Eifelian (= Early-Middle Devonian) boundary age of 391.8 ± 0.4 Ma.

Keywords: Devonian, U/Pb, absolute age, zircon, time scales.

The most important precondition for the construction of accurate time scales is a sufficient dataset of reliable and biostratigraphically well-bracketed isotopic ages. Not so long ago, only a few were available for the Devonian period. These ages were acquired from different isotopic decay schemes, they were imprecise and mostly did not fit the requirement of a close biostratigraphic control. Accordingly, time-scale constructions based on these ages diverged considerably. In the 1990s, there was substantial progress in Devonian chronology based on new, highly precise U-Pb zircon and monazite ages (Roden et al. 1990; Tucker el al. 1998). The U-Pb method has turned out, because of high initial parent/daughter clement ratios and a unique dual decay scheme, to be significantly superior to other isotopic systems in precision and reliability. However, even within the U-Pb system, different analytical methods (isotope dilution-thermal ionization mass spectrometry (ID-TIMS) v. high resolution-secondary ionization mass spectrometry (HR-SlMS)), applied to identical zircon material, have formerly yielded different ages (see detailed discussions by Tucker & McKerrow 1995; Compston 2000a, b; Williams et al. 2000). HR-SIMS ages 1-2% younger than ID-TIMS ages have turned out to be caused by a recently detected heterogeneity in the SL 13 zircon reference material and by the erroneous inclusion of youngest data points (which were later assumed to have lost Pb) in the weighted mean ages (Compston 200Oa, h). These problems have now been overcome. However, HR-SlMS ages are typically less precise (0.5-1.4%) than ID-TIMS ages (0.2-0.5%) and there are merely four reliable Devonian HR-SIMS ages available (Jagodzinski & Black 1999; Nesbitt et al. 1999; Compston 2004), of which some are not well controlled by biostratigraphy. In comparison, seven U-Pb ID-TIMS ages have been obtained from biostratigraphically well-constrained volcanic ash layers (Roden et al. 1990; Tucker et al. 1998; Richards et al. 2002). This study is part of a project that deals with the expansion of this dataset by the acquisition of new U-Pb ID-TIMS single-zircon ages from volcanic rocks (mainly K-bentonites) of the Rhenish Massif (Germany) (Kaufmann et al. 2004; Trapp et al. 2004). Here, two new ages obtained from Emsian (Lower Devonian) strata are presented. They are intercalated in biostratigraphically well-defined sedimentary successions and are therefore suitable for a numerical recalibration of the Emsian stage. Estimates for the duration as well as ages for the beginning and end, respectively, of the Emsian range from 4.4 Ma (390.4-386.0 Ma) to 15.5 Ma (409.5-394.0 Ma) (Harland et al. 1990; Fordham 1992; Young 1995; Gradstein & Ogg 1996, 2004; Sandberg & Ziegler 1996; Tucker et al. 1998; Compston 20006) (Fig. 1). This discrepancy is due to the fact that some workers have ignored the U-Pb monazite age (390.0 ± 0.5 Ma) of the lower Eifelian Tioga Ash (Pennsylvania) published by Roden et al. (1990), which represents a minimum age constraint for the Emsian-Eifelian boundary. Tucker et al. (1998) confirmed this result with a U-Pb ID-TIMS zircon age of 391.4 ± 1.8 Ma from the same ash zone. In addition, the latter study presented another U-Pb zircon age of 408.3 ±1.9 Ma for the Lower Emsian Esopus Formation (New York). The resulting time-scale calibration constrained the duration of the Emsian stage to 15.5 Ma from 409.5 to 394.0 Ma (Tucker et al. 1998) (Figs 1 and 2). To date, this is the best estimate for the duration of the Emsian stage because it was interpolated between two methodically consistent and biostratigraphically well-bracketed isotopic ages.