Seawater Sr isotopes in belemnites from the Upper Jurassic of southern Germany

The Late Jurassic seawater 87 Sr/ 86 Sr composition is constrained by relatively few data of well preserved and dated fossils, mainly from the Boreal realm, and data from Tethyan shelves are very limited. The previous seawater curve shows a continuous and nearly linear rise from a minimum in the Oxfordian through the Kimmeridgian and Tithonian. In order to improve the resolution of the 87 Sr/ 86 Sr curve with data from the Tethyan realm, 33 new 87 Sr/ 86 Sr data for 22 Kimmeridgian to Tithonian belemnites from the Franconian Jurassic, Germany, are supported by 52 δ 18 O and δ 13 C analyses of both belemnites, and the limestone matrix in which they are preserved. The stable isotope results of belemnite constituents are compatible with existing data, and all diagenetically altered parts of belemnites (apical area, microfractured, recrystallized and marginal parts of the rostrum) show isotope shifts towards matrix values. The originally aragonitic septa show very negative δ 18 O (−9.1 to −11.8 ‰) and δ 13 C (−2.6 to −5.3 ‰), probably due to disequilibrium fractionation, i.e., vital effects. The 87 Sr/ 86 Sr data of belemnite rostra, composed of unaltered low-Mg calcite (0.706888 to 0.707072) by and large overlap the previously inferred seawater 87 Sr/ 86 Sr curve, but provide more details, showing an apparent more stepwise increase during the Late Jurassic, with rapid increases followed by invariant ratios. Jurassic seawater 87 Sr/ 86 Sr was most likely uniform, but data from well preserved and dated belemnites of the same age show a small but significant variation in 87 Sr/ 86 Sr of up to 0.000018. This is most likely an artifact caused by unintentional sample contamination with macroscopically pristine but microfractured areas, attached matrix, or the apical zone. The stepwise increase of Late Jurassic seawater 87 Sr/ 86 Sr could also be an artifact caused by inaccuracies in the absolute ages assumed for Late Jurassic ammonite zone boundaries and imprecise relative ages within ammonite zones. Alternatively, changes in Jurassic seawater 87 Sr/ 86 Sr occurred far more rapidly (∼0.000200 per Ma) than the residence time of Sr would suggest, similar to increases previously reported for Phanerozoic anoxic events and related to the change in the flux of Sr from the continents.