SULFUR ISOTOPES IN ANHYDRITES FROM THE UPPER DEVONIAN PRYPIAC' AND DNIPRO- DONETS BASINS (BELARUS AND UKRAINE) I Tadeusz Marek Peryt, 2 Anatoliy A. Makhnach, 3 Stanislaw Halas, 4 Oleh Y. Petrychenko, 2 Leonid F. Gulis, and 2 Svietlana M. Abravets 'Panstwowy Instytut Geologiczny, Rakowiecka 4,00-975 Warszawa, Poland 2Institute of Geochemistry and Geophysics, National Academy of Sciences of Belarus, Kuprevich 7, Minsk 220141, Belarus 3 Mass Spectrometry Laboratory, Institute of Physics, Maria Curie-Sklodowska University, 20-03 I Lublin, Poland 'Institute of Geology and Geochemistry of Combustible Minerals, National Academy of Sciences of Ukraine, Naukova 3A, 79053 Lviv, Ukraine ABSTRACT: Upper Devonian evaporite rocks worldwide record a marked sulfur isotope excursion to heavier 8 34 8 values. New 8 34 8 data from anhydrite laminae in halite, anhydrite in water-insoluble residue in halite and sulfate from massive sulfate units in evaporite sequencesof the Prypiac' and Dnipro-Donets Basins of Belarus and Ukraine are presented. The mean is 27.6 %0 ± 3.0 %0 (n = 60) with a range from 21.9 %0 to 35.6 %0. The mean values for Frasnian and Famennian evaporites are similar: 27.5%0 ± 3.8 %0 (n = 30) and 27.7 %0 ± 1.9 %0 (n =. 30), respectively. The rocks formed in a restricted basin but its parent waters had a marine source. Bacterial sulfate reduction was the dominant process controlling 8 348 of Upper Devonian anhydrite and pyrite although also disproportionation processes and secular variation played an important role. True secular isotope variation is indicated between stratigraphic units, and a range of c. 24-29 %0 is assumed to be representativefor oceanic sulfate of this time. New data from the Lower Eifelian suggest that the value of c. 15 %0 was representative for oceanic sulfate of Eifelian time, and this indicates the increase of 8 348 values of oceanic sulfate by 12 %0 during ca. II Ma. Subsequently, 8 348 values show high plateau of27 %0 during ca. 12 Ma (Early Frasnian to Late Famennian) with some oscillation reflected in the mean values for particular evaporite units. INTRODUCTION The 8 34S of evaporite sulfates of different ages are used to construct the approximate seawater secular 8 34S curve (Claypool et al. 1980; Strauss 1997). The question of homogeneity and representativeness of sulfur isotopic data for ancient evaporite deposits was discussed by Nielsen (1989) and Strauss (1997). They concluded that several samples from any given evaporite deposit and several evaporite deposits of any given time from a number of different places worldwide should be studied prior to an average isotope value for a given time is assigned. Strauss (1997, p. 104) assumed that this approach should narrow a spread of isotope values. On the other hand, redox reactions between sulfur species during and/or before sulfate precipitation could results in significant variations of 8 34S (Lu et al. 200 I). Claypool et al. (1980: fig. 7) recognized a broad minimum at 8 34S = +17 %0 in the Lower Devonian, followed by a sharp maximum of 8 34S = 24 %0 in the Frasnian. The validity of the recorded curve for the Devonian has been conformed by data presented by other authors (e.g. Pierre and Rouchy 1986; Richardson and Hansen 1991; see also Claypool et al. 1980, p. 214). Claypool et al. (1980, p. 214) noted that a large number of their analyses in the Upper Devonian have higher 8 34S values (up to 34 %0) but as these analyses are from one basin in western Canada, they have not been given weight in drawing the "best estimate" curve. The evolution of sulfur isotopic composition of seawater sulfate is fairly discontinuous as far as time resolution is concerned (Strauss 1997, p. liS) and thus the Upper Devonian Prypiac' and Dnipro-Donets Basins of Belarus Carbonates and Evaporites, v. 22, no. 1,2007, p. 43-54. and Ukraine, with their multiple episodes of evaporite deposition and reasonable constrained age assignments, offer a possibility to construct the secular 8 34S curve for the major part of the Late Devonian times. In addition, despite of the overall temporal trend based on samples from the Prypiac' and Dnipro-Donets Basins, they can indicate shorter-termfluctuations that could not have been previously resolved (cf. Strauss 1999). Previous informationpublished on the sulfur isotopes in sulfates from the Upper Devonian Prypiac' and Dnipro-Donets Basins of Belarus and Ukraine (Eremenko and Pankina 1971; Ivanov et al. 1971; Lein et al. 1974; Pankina et al. 1975; Makhnach et al. 2000) also fits this picture, however, the stratigraphic position of the previously-studied samples was very general. A total of sixty new samples taken from the precisely-constrained evaporites, in terms of their stratigraphic position, were analyzed for 8 34S aiming to provide a detailed sulfur isotope record for the Upper Devonian of those basins. GEOLOGICAL SETTING Widespread evaporites occurred in Middle and Late Devonian (Zharkov 1984; Vysotskiy et al. 1988), including the Dnipro-Donets and Prypiac' evaporite basins related to the Donbas-Prypiac' rift, one of the best-studied Paleozoic structures of its type (Figs. 1-3) (Ulmishek et al. 1994). Devonian and younger rocks (2 km thick in the western part of the Prypiac' Basin and up to 15 km in the SE part of the Dnipro-Doncts Basin) unconformably overlie the Proterozoic or crystalline basement. The Devonian and younger sedimentary rocks may be divided into prerift platform, synrift, and postrift sag sequences, and the overlying platform sequence of post-Early Permian age (Ulrhishek et al. 1994; Konishchev et al. 2001). The