Palaeogeography, Palaeoclimatology, Palaeoecology 211 (2004) 157–169 www.elsevier.com/locate/palaeo

Palaeoenvironmental reconstructions by stable isotopes of Middle Miocene gastropods of the Central Paratethys

Christine Latala,*, Werner E. Pillera, Mathias Harzhauserb

aInstitute for Geology and Palaeontology, University of Graz, Heinrichstrage 26, A-8010 Graz, Austria bMuseum of Natural History Vienna, Geological-Palaeontological Department, Burgring 7, A-1014 Vienna, Austria Received 11 April 2003; received in revised form 9 April 2004; accepted 5 May 2004

Abstract

The stable isotopes of carbon and oxygen were used to provide information of mixing of marine and freshwater and subsequently used to determine palaeosalinities. In the Middle to Late Miocene St. Margarethen bZollhausQ section (Upper Sarmatian bMactra ZoneQ/Pannonian), a succession of three gastropod assemblages indicates changing environmental conditions from a limnic-fluviatile to littoral and a sublittoral marine environment. Several shells of the main representatives of the three assemblages [ hartbergensis (Hilber), Granulolabium bicinctum (Brocchi), Potamides disjunctus (Sowerby)], and additionally shells of Lymnaea sp., a typical freshwater gastropod, and Pomatias conicus,a terrestrial gastropod, were taken for stable isotope analyses. Oxygen and carbon isotope data of the freshwater endmember are distinctively lower (d18O~À4.5x/d13C~À9x) than of the marine endmembers (d18O~À2.2x/d13C~2.3x). In contrast to the expected intermediate isotopic composition between the freshwater and marine environment, the shells of supposed brackish environments have higher d18O values than the shells of the marine environment. These higher d18O values are explained by evaporation, which affected the protected areas of riverine to lacustrine environments and small pools on , formerly considered to be the habitats of the brackish assemblages. Thus, the oxygen and carbon isotope data extend the already existing palaeoenvironmental information. D 2004 Elsevier B.V. All rights reserved.

Keywords: Gastropods; Stable isotopes; Palaeoenvironment; Middle Miocene; Central Paratethys

1. Introduction Though the Paratethys underwent distinct changes in environmental conditions forced by palaeogeograph- Investigations of stable isotope ratios of carbon and ical changes, the use of stable isotopes on Miocene oxygen on carbonate shells are a powerful tool for the fossils of the Paratethys has been rather scarce (Geary reconstruction of palaeoenvironmental parameters. et al., 1989; Sutovska and Kantor, 1992; Matyas et al., 1996; Durakiewicz et al., 1997; Hladilova et al., 1998; * Corresponding author. Fax: +43 316 380 9871. Gonera et al., 2000; Bojar et al., 2004). The Badenian/ E-mail address: [email protected] (C. Latal). Sarmatian (Middle Miocene regional stages) boun-

0031-0182/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.palaeo.2004.05.003 158 C. Latal et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 211 (2004) 157–169 dary marks a distinct change in faunal communities littoral to a marine-littoral environment. These three where nearly all stenohaline groups such as corals, assemblages were named after their main represen- echinoids and planktic foraminifera, became extinct. tatives as bPotamides hartbergensis assemblageQ, The same scenario is observed in the molluscan fauna, bGranulolabium bicinctum assemblageQ and bPo- which is devoid of pectinids, strombids or other tamides disjunctus assemblageQ. The assumed palae- stenohaline taxa. The closure of the marine connection oenvironmental changes are connected with changes of the Paratethys is the main factor, which caused an in palaeosalinities. If differences in the environmental isolation of the Paratethys fauna from the Mediterra- parameters, like temperature and salinity, existed nean bioprovince. This geodynamic process coincided between the three assemblages, these should be with a drastic change in water chemistry within the reflected in oxygen and carbon isotope signatures of Paratethys (Ro¨gl, 2001). Traditionally, a decrease in carbonate shells of representatives of the three salinity is advocated as cause for the biotic changes assemblages. Lowest d18O and d13C values would (Papp, 1954). Recently, however, an increase in be expected for the limnic-fluvial environment, as alkalinity of the water body as trigger for the faunistic freshwater typically has lighter isotope values than collapse has been put forward (Pisera, 1995, 1996). marine water. If oxygen and carbon isotopes behaved Especially during the Middle Sarmatian, occurrence conservatively in the seawater–freshwater mixing of ooliths and small biohermal carbonate bodies zone, shells from the brackish-littoral setting should indicate normal marine conditions and contradict the yield intermediate values between the freshwater and brackish water scenario (Harzhauser and Piller, 2001). marine endmember. The use of oxygen and carbon The use of d18O and d13C values to provide isotope analyses should therefore provide a test of the information on mixing of marine and freshwater, and proposed palaeoenvironmental interpretation based on therefore inferring palaeosalinities, has long been a taxonomic uniformitarian approach. tested (Keith and Weber, 1964; Keith et al., 1964; Keith and Parker, 1965; Stanton and Dodd, 1970; Mook, 1971; Dodd and Stanton, 1975; Anderson and 2. Geological setting Arthur, 1983). Commonly freshwater has lower d18O and d13C values than marine water, whereas brackish The Late Sarmatian/Pannonian section at St. water systems show intermediate isotopic ratios. But Margarethen bZollhausQ (Burgenland, Austria) is an the relationship between salinity and isotopic compo- exploited gravel pit located in the Neogene Eisen- sition is often more complicated: oxygen isotope stadt-Sopron Basin. This small basin of about 20Â20 values of carbonates are controlled by temperature km size is interpreted to represent a satellite basin of and d18O of the ambient water. Evaporation may the Vienna Basin (Steininger, 1991)(Fig. 1). enrich water in 18O, if it has a long residence time in Recently a 30-m-thick section has been exposed hydrologically restricted settings (Swart et al., 1989; comprising two sedimentary units of the Sarmatian Hudson et al., 1995; Hendry and Kalin, 1997), and (Middle Miocene) and Pannonian (Late Miocene). therefore lead to higher d18O values than typically Based on the occurrence of Sarmatian Mactra vital- expected for a specific setting. As freshwater dilution iana (d’Orbigny) and the characteristic composition as well as evaporation can lead to depletion in d13Cin of the mollusc fauna, a dating to the Upper Sarmatian surface waters, the interpretation of d13C values of ecostratigraphic bMactra ZoneQ (zonation sensu Papp, carbonate samples for inferring paleosalinities may 1954)(Fig. 2) is allowed. It is represented by gravels, also be equivocal (Patterson and Walter, 1994; Swart marly sands and scattered detritic limestones, whereas et al., 2001). the Pannonian consists of sand and silt. The Sarmatian The molluscan fauna of the Upper Sarmatian part of the section yields only deposits which can be section of St. Margarethen bZollhausQ has been correlated with the Mactra Zone of the Vienna Basin recently investigated by Harzhauser and Kowalke ecostratigraphic mollusc zonation. Short lived out- (2002). They interpreted a succession of three crops in the base of the current outcrop proved the gastropod assemblages as an indication of changing existence of sands, which already represent the environments from a limnic-fluvial to a brackish- preceding Ervilia Zone. Due to the lack of drilling C. Latal et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 211 (2004) 157–169 159

data, the exact thickness of the Sarmatian strata in this part of the Eisenstadt-Sopron Basin is unknown. The entire section is described in detail by Harzhauser and Kowalke (2002) and Harzhauser et al. (2002). In this work, however, special focus is given to a layer, which shows a distinct succession of gastropod assemblages (Fig. 3).

3. Gastropod assemblages and palaeoecological constraints

The analysed shells derive from a short sequence of about 40 cm thickness (Fig. 3). Sedimentologi- cally distinct layers yield three ecologically defined successive assemblages, represented by a lacustrine stage, a marine mudflat stage, and a Sarmatian- marine littoral stage. They were named by Harz- Fig. 1. Location of the St. Margarethen/Zollhaus section in the hauser and Kowalke (2002) after the predominating Eisenstadt-Sopron Basin (Schmid et al., 2001). potamidids the Potamides hartbergensis, the Gran-

Fig. 2. Stratigraphic correlation chart of the Central and Western Paratethys with the Mediterranean. 160 C. Latal et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 211 (2004) 157–169 C. Latal et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 211 (2004) 157–169 161 ulolabium bicinctum and the Potamides disjunctus transition from the littoral to very shallow sublittoral, assemblages. linked with a shift in lithofacies from silts and clays The lowest marly silt layer was settled by the towards marly silts and marly sands with scattered bPotamides hartbergensis assemblageQ consisting pebbles. The sediments are interpreted as representing mainly of P. hartbergensis (Hilber), and abundant more or less normal marine conditions, based on the freshwater gastropods (lymnaeids and planorbids) as occurrence of various archaeogastropods, columbell- well as terrestrial gastropods (carychiids). This mixed ids and opisthobranch gastropods, which shun brack- assemblage was therefore interpreted by Harzhauser ish water environments. The freshwater influx had and Kowalke (2002) to have settled in a rather obviously ceased. protected riverine to lacustrine environment which Thus, Granulolabium bicinctum occurs in two developed within the deltaic complex bordering the different assemblages. Nevertheless, its shell mor- Sarmatian Sea. P. hartbergensis, however, cannot be phology seems to be strongly affected by the environ- interpreted as inhabitant of pure freshwater environ- ment. Specimens from the mudflats are small, glossy, ments, and point to local or temporary development of thin-shelled and poorly sculptured. In contrast, those ponds within the delta plain, suitable for this potamidid. from the bmarineQ Potamides disjunctus assemblage The succeeding bGranulolabium bicinctum assem- are robust, thick shelled and develop a conspicuous blageQ is characterised by a nearly monospecific mass ornamentation. This observation is an important occurrence of G. bicinctum (Brocchi) within a marly evidence that the specimens deriving from the marly silt or clay. It is interpreted as having been formed sand overlying the mudflat are elements of the P. during a minor transgressive phase during the disjunctus assemblage rather than reworked shells expansion of the sea within the Mactra Zone. This from the G. bicinctum assemblage. relative sea level rise allowed the establishment of Potamidid and batillariid gastropods are perennial vast mudflats on the flat delta plain in the Eisenstadt- but exact data on the life span of recent Sopron Basin. The fossil mudflat containing the G. representatives of these littoral animals are poor. By bicinctum assemblage can be traced throughout the comparing with modern relatives (Yamada, 1982; gravel pit and displays an extension of approximately Houbrick, 1991; Tojo and Ohno, 1999; Nishihira, 500 m2. The basal layer of this mudflat deposit yields 1983) life ages for the sampled gastropods may be masses of Hydrobia frauenfeldi and Hydrobia cf. deduced. stagnalis. Soon after, the hydrobiids are replaced by monospecific G. bicinctum assemblages yielding excellently preserved shells. This internal succession 4. Materials and methods of hydrobiids to Granulolabium assemblages seems to reflect the gradual shift from the upper intertidal Well preserved gastropods were used for stable zone to the lower intertidal zone. isotope investigations. For detecting diagenesis, sam- The uppermost gastropod assemblage (bPotamides ples were analysed for their shell mineralogy and disjunctus assemblageQ) is characterised by P. dis- microstructures with X-ray diffraction analyses and junctus (Sowerby), Acteocina lajonkaireana (Bas- scanning electron microscopy (SEM). At least three terot), Gibbula buchi (Dubois), Gibbula podolica shells of Potamides hartbergensis, Granulolabium (Dubois), Jujubinus turricula (Eichwald), Granulo- bicinctum from the G. bicinctum assemblage as well labium bicinctum (Brocchi) and Acmaea soceni as from the Potamides disjunctus assemblage and P. (Jekelius). This assemblage is bound to a depositional disjunctus were chosen for X-ray diffraction and SEM

Fig. 3. St. Margarethen/Zollhaus section: enlarged part shows the layers from which the gastropod shells for stable isotope measurements were taken with their specific gastropod assemblages; (1) typical representatives of the Potamides hartbergensis assemblage: (a) P. hartbergensis (Hilber), (b) Gyraulus vermicularis (Stoliczka), (c) Valvata pseudoadeorbis Sinzov, (d) Stagnicola cf. bouilleti (Michaud), (e) Carychium pachychilus Sandberger, (f) Pomatias conicus (Klein); (2) typical representatives of the Granulolabium bicinctum assemblage: (a) G. bicinctum (Brocchi), (b) Hydrobia frauenfeldi (Ho¨rnes); (3) typical representatives of the Potamides disjunctus assemblage: (a) G. bicinctum (Brocchi), (b) P. disjunctus (Sowerby), (c) Acmaea soceni Jekelius, (d) Acteocina lajonkaireana (Basterot), (e) Mitrella agenta (Harzhauser and Kowalke), (f) Gibbula buchi (Dubois) (after Harzhauser and Kowalke, 2002). 162 C. Latal et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 211 (2004) 157–169 analyses. Mineralogical composition of the shells was calcite might be present, comparison of isotopic results analysed with a Bruker D8 Discover (CuKa1)with of the pure aragonitic shells with shells containing General Area Detector Diffraction solutions (GADDS). traces of calcite indicated no differences in the isotopic It is possible to investigate single points on a sample composition. SEM investigations performed on the with this X-ray diffractometer. Therefore, three to five samples also confirmed the aragonitic composition. As analyses on each shell on different whorls were made aragonite is a relatively unstable phase and easily without destroying the shells. Multiple analyses on converts to calcite during diagenesis, the results of each shell should provide information about variations these two methods show that all samples are nearly in shell mineralogy within one shell. For SEM analyses unaffected by diagenesis and provide strong evidence shell fragments from the aperture were used. The same that stable isotope values are primary and can be used shells which were investigated for diagenesis were for palaeoenvironmental analyses. chosen for stable isotope measurements: five shells of G. bicinctum of the G. bicinctum assemblage, five P. 5.2. Oxygen isotopes hartbergensis of the P.hartbergensis assemblage, three P. disjunctus and five G. bicinctum of the P. disjunctus Intrashell variability in d18O is distinct between assemblage. In addition, two specimens of Lymnaea the different assemblages (Fig. 4): within the Pota- sp. as representative of typical freshwater gastropods, mides hartbergensis assemblage gastropods show three specimens of Hydrobia sp. from the basal layer of low variability of at most 1x. In the Granulolabium the mudflat and two specimens of Pomatias conicus as bicinctum assemblage, intrashell changes are higher a terrestrial snail were analyzed. For the G. bicinctum, with values around 1.6–2.5x. In the Potamides P. hartbergensis and P. disjunctus shells, at least six disjunctus assemblage, no differences in the intrashell samples from each shell from the aperture to the apex variability between the two species are discernible: were drilled with a 0.8-mm drill. Multiple samples from G. bicinctum and P. disjunctus show fluctuations of the shells were taken in order to estimate the range of about 0.5–1.4x. isotopic values within one shell. For Hydrobia sp., In addition to the mean values of d18O and d13Cof Lymnaea sp. and P. conicus, whole specimens were each measured shell, bulk data of the shells of crushed, and the powder used for isotope analyses. Lymnaea sp., Hydrobia sp. and Pomatias conicus Whole shells were used because of the small size of are given in Table 1. each single shell. To compare these different types of In the Potamides hartbergensis assemblage, d18O samples, also mean values of the multiple samples from values range from À1.3x to 2.0x, in the Gran- the other shells were used for interpretation. Samples ulolabium bicinctum assemblage from À3.2x to were reacted with 100% phosphoric acid at 70 8Cina 0.5x. The Hydrobia sp., which settled the mudflats Finnigan Kiel II automated reaction system and first, show values from À2.4x to À1.2x. In the measured with a Finnigan Delta Plus isotope-ratio Potamides disjunctus assemblage, d18O values vary mass spectrometer at the Institute of Geology and from À3.1x to 0.4x. The freshwater gastropods Palaeontology, University of Graz. Repeated measure- Lymnaea sp. exhibit values of À5.1x and À4.2x, ments of NBS-19 and an internal laboratory standard and shells of the terrestrial snails yielded data from yield a standard deviation of 0.1x for both d18O and À3.0x to À1.9x. d13C. Values are given against VPDB. 5.3. Carbon isotopes

5. Results Intrashell variability in d13C is approximately the same in nearly all specimens, ranging from 0.5% to 5.1. Shell structure and mineralogy 2.0x(Fig. 4). No distinct difference between the assemblages is observable, but two of the measured The shell mineralogy was determined using X-ray specimens show extreme high variability. One shell diffraction techniques. In most of the samples only from Granulolabium bicinctum of the G. bicinctum aragonite was found. Although in a few samples assemblage and one shell of G. bicinctum of the C. Latal et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 211 (2004) 157–169 163

Fig. 4. Stable isotope profiles from the apex to the aperture of the shells of the different gastropod assemblages. 164 C. Latal et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 211 (2004) 157–169

Table 1 Potamides disjunctus assemblage have intrashell 18 13 Mean d O and d C values in x and standard deviations of the variabilities of 3.2% and 3.4x, respectively. gastropod shells In the Potamides hartbergensis assemblage, d13C 18 13 Assemblage Species Number of d O d C values range from 0.3x to 2.0x, in the Granulo- samples/shell (x) (x) labium bicinctum assemblage from À1.1x to 3.5x, Pomatias conicus bulk À3.0 À6.4 and in the Potamides disjunctus assemblage from Pomatias conicus bulk 1.9 5.6 À À x x Lymnaea sp. bulk À5.1 À7.6 1.6 to 6.8 . The three shells of Hydrobia sp. Lymnaea sp. bulk À4.2 À10.0 showed values between À0.8x and 0.7x.Bulk analyses of two shells of Lymnaea sp. exhibit values Potamides hartbergensis of À7.6x and À10.0x, and the two shells of Potamides n=6 À0.4 1.0 Pomatias conicus of À5.6x and À6.4x (Table 1). hartbergensis (F0.4) (F0.4) Potamides n=6 À1.1 1.2 hartbergensis (F0.1) (F0.4) Potamides n=6 1.7 1.0 6. Discussion hartbergensis (F0.3) (F0.3) Potamides n=6 À0.7 1.3 The excellent preservation of metastable aragonite hartbergensis (F0.4) (F0.4) Potamides n=6 1.2 1.3 in shells has been used to argue that the isotopic hartbergensis (F0.3) (F0.4) composition is also primary. Additionally, molluscs are considered to precipitate their shells in isotopic Granulolabium bicinctum equilibrium (Grossman and Ku, 1986; Cornu et al., Granulolabium n=8 À1.1 À0.7 1993). So it may be reasonably assumed that the bicinctum (F0.6) (F0.2) Granulolabium n=8 À1.5 À0.3 isotopic composition of the analyzed aragonitic bicinctum (F0.7) (F0.6) gastropod shells reflects the interplay of the general Granulolabium n=8 À2.0 1.4 conditions (e.g. global climate), influencing the bicinctum (F0.7) (F1.1) isotopic composition of the ambient water and local Granulolabium n=8 À1.3 À0.5 palaeoenvironmental parameters (e.g. temperature and bicinctum (F0.5) (F0.5) Granulolabium n=7 À1.1 À0.3 salinity). 18 bicinctum (F0.7) (F0.3) The main controlling factors of d O values in Hydrobia sp. bulk À1.2 À0.3 gastropod shells are temperature and the d18O of the Hydrobia sp. bulk À2.4 0.7 ambient water. Freshwater d18O values are primarily Hydrobia sp. bulk À1.9 À0.8 controlled by the isotopic composition of the rain- 13 Potamides disjunctus water. In contrast, d C values of marine and fresh- Granulolabium n=6 À0.2 5.8 water gastropods are influenced by a multitude of bicinctum (F0.4) (F1.2) factors, e.g., upwelling, seasonal productivity, diet, Granulolabium n=6 À2.9 2.8 living mode, growth rate, reproductive status (Geary F F bicinctum ( 0.2) ( 0.6) et al., 1992 cum lit.; Bonadonna et al., 1999). Granulolabium n=6 À2.5 2.5 bicinctum (F0.4) (F0.4) Eocene marine molluscs from the Paris Basin and Granulolabium n=6 À2.4 1.5 England show high intra-annual or intrashell variations bicinctum (F0.5) (F0.3) in d18O between winter and summer of about 2.5x Granulolabium n=6 À2.5 1.4 and 3.0x, while d13C values vary by about 2.5x in F F bicinctum ( 0.2) ( 0.1) the Paris Basin and up to 4.0x in England (Andreas- Potamides n=10 À2.3 2.6 disjunctus (F0.5) (F0.6) son and Schmitz, 1996; Purton and Brasier, 1997). Potamides n=10 À2.7 2.6 Samples from this work, as shown in Fig. 4, exhibit disjunctus (F0.2) (F0.4) similar intrashell variabilities, probably reflecting Potamides n=10 À2.4 2.8 seasonal temperature changes during their lifetime. F F disjunctus ( 0.2) ( 0.3) In palaeoenvironmental isotope studies, it can be difficult to distinguish between the effects of local temperature and salinity on account of the d18O signal C. Latal et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 211 (2004) 157–169 165 in gastropod shells. General palaeoenvironmental and marine water and a less saline, generally isotopically climatic conditions for Sarmatian environments in the depleted freshwater, used for proxy estimations of Central Paratethys can be inferred from new results by palaeosalinities, may be masked by other factors (e.g. Piller and Harzhauser (2002). Tropical to subtropical evaporation) influencing the stable isotope composi- conditions are assumed, based on the occurrence of tion of the mixing components (Swart et al., 1989, small carbonate platforms with thick oolite sequences 2001; Patterson and Walter, 1994; Hudson et al., and limestones with abundant larger miliolid fora- 1995; Hendry and Kalin, 1997; Holmden et al., 1997). minifera. Bryozoan-foraminiferan build-ups formed The isotope data of shells of the freshwater bioconstructions of up to 20m width in the Eisenstadt- molluscs from St. Margarethen can be used to Sopron Basin during the Mactra Zone. On land, a estimate the d18O composition of the freshwater semi-arid climate is indicated by the occurrence of endmember. Present members of the Lymnaea calcrete formations (Piller and Harzhauser, 2002). are typical freshwater gastropods which tolerate These evidences also point to rather uniform warm salinities lower than 0.3–0.7 ppt (Schmitz and temperature conditions in the marginal areas. Hence, it Andreasson, 2001). The Lymnaea sp. from St. can be safely assumed that changes in salinity and Margarethen show the lowest isotope values in d18O water composition are more likely the cause of the as well as in d13C. A mean d18O value of À4.7x and variations in the d18O values between the assemblages a mean d13C value of À8.8x are consistent with a than temperature. freshwater environment (Fig. 5). The possibility to differentiate marine from fresh- The Potamides disjunctus assemblage was consid- water environments by means of investigations of ered to be a marine assemblage in a shallow littoral oxygen and carbon isotopes is based on the generally setting. Normal marine conditions can be inferred contrasting d18O values of seawater (~0x SMOW) from the gastropods in this assemblage, which do not and freshwaters (bÀ5x), and differences in d13Cin occur in brackish environments. Isotopic composition dissolved inorganic carbon (DIC) between seawater of P. disjunctus and Granulolabium bicinctum shells (~0x PDB) and freshwaters (generally lower d13C from this assemblage show great similarities. The values). But the simple mixing between a saline mean d18O value of the three P. disjunctus shells is

Fig. 5. Plot of the mean values of the stable isotope data of each measured gastropod shell. Open symbols are mean values of single shells, full symbols are mean values of each genus in one assemblage. Different gastropod assemblages can be clearly distinguished and their inferred environmental conditions are shown. 166 C. Latal et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 211 (2004) 157–169

À2.1x and 2.8x for d13C, respectively, while the G. cene, instead of having recorded the real water bicinctum shells yielded a mean d18O value of À2.6x chemistry of the lake. and a mean d13C value of 2.1x, excluding one shell Sarmatian shells of Cardium and Cerastoderma with anomalous high isotope values. As this shell from the Pannonian Lake show a wide range of shows no obvious differences in its mineralogy and isotope ratios (d18O from À2.6x to À0.2x/d13C shell structure, these anomalous positive oxygen and from 1.0x to 1.9x), but these data were obtained carbon isotope values have to remain unexplained by without differentiating between littoral and sublittoral now. elements (Matyas et al., 1996). Due to the then current In general, the mean d18O value of À2.6x for assumption that the Sarmatian sea was a brackish the Potamides disjunctus assemblage seems a little environment, these authors interpreted the data as bit low for a marine environment. For example, pointing to brackish conditions. But also some isotope data of three marine Conus shells of the Badenian shells deriving from an undoubted fully Miocene of northern Florida yielded an average marine environment showed similar isotopic compo- d18OofÀ1.0x(Kobashi et al., 2001). But it is sitions, especially d18O values are in the same range highly questionable if the Eisenstadt-Sopron Basin (À2.5x to À1.2x), whereas d13C values of the is comparable to an open oceanic environment. Badenian shells are generally lower (À2.9x to 2.3x). Though major ice sheet expansions around 14 Ma These inconsistencies could only be explained by raised d18O of the seawater (Lear et al., 2000; considerations of a salinity and oxygen isotope mass Zachos et al., 2001), d18O estimates of seawater for balance for a closed or nearly closed lake (Matyas et theLateMiddleMiocenearelowerthan0x al., 1996). SMOW, which is today’s worldwide oceanic aver- The mean d18O value of the Potamides disjunctus age (Epstein and Mayeda, 1953). Local environ- assemblage is in the range of the data from Geary et mental conditions may have influenced the d18Oof al. (1989) and Matyas et al. (1996), while the mean the seawater of this small satellite basin of the d13C value is a little bit higher. As the carbon isotopic Vienna Basin, leading to even more reduced d18O composition of mollusc shells is influenced by several values. Therefore comparison of the results with factors, and vital effects for d13C in gastropod shells isotopic data from open oceanic localities might be cannot be excluded (Romanek and Grossman, 1989; inadequate. Unfortunately, only few stable isotope Wefer and Berger, 1991), d13C signals are more data of Sarmatian molluscs from the Central Para- complicated to interpret. New data of the Late tethys are available for comparison (Geary et al., Sarmatian reject the slowly freshening brackish basin 1989; Matyas et al., 1996). Carbon and oxygen and propose, at least for the Late Sarmatian, normal isotopic ratios of melanopsids from the Pannonian marine, even partly hypersaline conditions (Piller and Basin system seem to indicate a shift in salinity of Harzhauser, 2002). the Pannonian Lake from brackish towards fresh- In spite of the relatively low d18O data, which can water environments during the Late Miocene be explained by slightly reduced d18O values of (Geary et al., 1989). However, the selected gastro- seawater in the Eisenstadt-Sopron Basin, the Pota- pod genus inhabits mainly freshwater mides disjunctus assemblage can be considered as environments of riverine and deltaic ecosystems, as marine because of evidences from palaeofaunal can be inferred from data provided by Van Damme considerations. (1984), Glaubrecht (1993), Heller and Sivan (2000), Differentiation between the freshwater endmem- Heller et al. (1999) and Mienis (1983). Even if the ber, represented by the Lymnaea sp., and the marine endemic species of Lake Pannon might have settled Potamides disjunctus assemblage, is evident in d18O also coastal areas adjacent to riverine systems, a as well as in d13C(Fig. 5). Intermediate values considerable influence of the water chemistry would have been expected for brackish water fossils. should be expected. Hence, one may suppose that But as shown in Fig. 5, the Potamides hartbergensis Melanopsis is a critical candidate for isotopic and the Granulolabium bicinctum assemblages do studies as it might have settled rather similar not align along this simple mixing line. The G. freshwater influenced habitats throughout the Mio- bicinctum shells show similar values as the Hydrobia C. Latal et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 211 (2004) 157–169 167 sp. The mean d18O values are À1.8x for Hydrobia slightly modified in light of oxygen and carbon sp. and À1.4x for G. bicinctum. These values are isotope data. slightly higher than the mean values for this species Based on the palaeofaunal evidence in combination of the P. disjunctus assemblage. The shells of P. with the isotopic results the Potamides disjunctus hartbergensis exhibited the most extraordinary val- assemblage is considered as the marine endmember. ues. In addition to the fact that the mean d18O value The P. disjunctus assemblage shows distinctively of the P. hartbergensis assemblage is the highest of higher d18O and d13C values as compared to the the measured groups with 0.2x, the mean values of freshwater endmember, represented by Lymnae sp. each shell span a range from À1.1x to 1.7x, which The Potamides hartbergensis assemblage, originally is quite a large variation within a group. But the interpreted as representing a protected riverine to range of d13C values for the five P. hartbergensis lacustrine environment, shows the highest d18O shells is small (1.0–1.3x). Therefore, this environ- values. In addition, the shells of the Granulolabium ment must have provided special conditions. Such a bicinctum assemblage, and the shells of Hydrobia sp., wide range of d18O and an enrichment against the both settling on the mudflat, also show higher d18O freshwater–marine mixing environment can be values than the marine G. bicinctum and P. disjunctus caused by high evaporation rates. Similar conditions, shells, but lower values than the P. hartbergensis where high evaporation rates lead to an enrichment shells. These data seem at first to contradict the of d18O in brackish bay water, are documented from original interpretation, where it was considered that the Florida Bay (Lloyd, 1964; Swart et al., 1989) freshwater influx decreased from the environment and the Lower Miocene of the Mainz Basin (Rothe represented by the P. hartbergensis assemblage to the et al., 1974). mudflat until fully marine conditions were reached. The originally supposed conditions for the Pota- However, evaporation may have led to an enrichment mides hartbergensis assemblage were a protected in 18O which affected especially the small protected riverine to lacustrine environment which easily could areas of riverine to lacustrine environments of the P. experience higher evaporation rates within small hartbergensis assemblage or small pools on the ponds or pools leading to the high d18O values. A mudflats, settled by Hydrobia sp. and the G. similar situation can be assumed for the Granulo- bicinctum assemblage, which were only flooded labium bicinctum assemblage, which is interpreted to sporadically. Thus, the oxygen and carbon isotopic have settled on mudflats. A high evaporation level data provide a palaeoenvironmental insight not readily inferred from the isotope data of the P. hartbergensis apparent from the study of the gastropod assemblages assemblage fits well to the frequent calcrete for- alone. mation, as observed at the section St. Margarethen At least during the Upper Sarmatian Mactra (see Fig. 3), as well as in other Upper Sarmatian Zone, gastropods from a freshwater, a brackish and outcrops of the Vienna Basin and the Styrian Basin. a shallow sublittoral environment yielded distinc- At least short phases of hypersaline conditions have tively different isotope signatures. The classical been predicted by Ja´mbor (1978) and Halmai (1998), brackish water scenario for the Sarmatian (Papp, based on evaporite intercalations in the Sarmatian 1954) is not supported by these data. Furthermore deposits of the Hungarian Zsambek Basin and the data of the Potamides hartbergensis assemblage Zagyva Valley and the Transylvanian Basin in and from the mudflat assemblage point to consid- Rumania. erable evaporation during the Late Sarmatian.

7. Conclusion Acknowledgements

Ecological interpretation of molluscan assemblages We thank P. Swart and A. Longinelli for useful in the St. Margarethen bZollhausQ section predicted a critical reviews. Comments of E.L. Grossman are change from a limnic-fluvial to a brackish-littoral to a much appreciated. Financially, the study was sup- marine-sublittoral environment. This interpretation is ported by the Austrian Science Foundation (FWF, 168 C. Latal et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 211 (2004) 157–169 project bStable Isotopes and changing Miocene palae- 13C) of Middle Miocene (Badenian) foraminifers in the Central oenvironments in the East Alpine regionQ (P-14366 Paratethys. Terra Nova 12, 231–238. b Grossman, E.L., Ku, T.-L., 1986. Oxygen and carbon isotope BIO) and project Evolution Versus Migration: fractionation in biogenic aragonite: temperature effects. Chem- Changes in Austrian Marine Miocene Molluscan ical Geology 59, 59–74. PalaeocommunitiesQ (P-13745 BIO). Halmai, J., 1998. Lower and Middle Miocene Formations in Hungary. In: Cicha, I., Ro¨gl, F., Rupp, C., Ctyroka, J. (Eds.), Oligocene–Miocene Foraminifera of the Central Paratethys. Abhandlungen der Senckenbergischen Naturforschenden References Gesellschaft, vol. 549, pp. 54–56. Harzhauser, M., Kowalke, T., 2002. Late Middle Miocene Anderson, T.F., Arthur, M.A., 1983. Stable isotopes of oxygen and (Sarmatian) Potamidid-Dominated Gastropod assemblages of carbon and their application to sedimentologic and paleo- the Central Paratethys as tool for facies interpretation. Facies 46, environmental problems. In: Arthur, M.A. (Organizer), Stable 57–82. Isotopes in Sedimentary Geology. Society of Economic Paleon- Harzhauser, M., Piller, W.E., 2001. The Sarmatian of the Central tologists and Mineralogists Short Course, vol. 10, pp. 1-1–1-151. Paratethys—an urgent need of a reevaluation of models. The 3rd Andreasson, F.P., Schmitz, B., 1996. Winter and summer temper- European Paleontological Congress 2001, Leiden, The Nether- atures of the Early Middle Eocene of France from lands, pp. 20–21. d18O profiles. Geology 24/12, 1067–1070. Harzhauser, M., Kowalke, T., Mandic, O., 2002. Late Miocene Bojar, A.-V., Hiden, H., Fenninger, A., Neubauer, F., 2004. Middle (Pannonian) Gastropods of Lake Pannon with Special Emphasis Miocene seasonal temperature changes in the Styrian basin, on Early Ontogenetic Development. Annalen des Naturhistor- Austria, as recorded by the isotopic composition of pectinid and ischen Museums Wien 103A, 75–141. brachiopod shells. Palaeogeography, Palaeoclimatology, Palaeo- Heller, J., Sivan, M., 2000. A new species of Melanopsis from the ecology 203, 95–105. Golan heights, Southern Levant (: ). Bonadonna, F.P., Leone, G., Zanchetta, G., 1999. Stable isotope Journal of Conchology 37/1, 1–5. analyses on the last 30 ka molluscan fauna from Pampa Heller, J., Sivan, M., Motro, U., 1999. Systematics, distribution and grassland, Bonaerense region, Argentina. Palaeogeography, hybridization of Melanopsis from the Jordan Valley (Gastro- Palaeoclimatology, Palaeoecology 153, 289–308. poda: Prosobranchia). Journal of Conchology 36/5, 49–81. Cornu, S., P7tzold, J., Bard, E., Meco, J., Cuerda-Barcelo, J., 1993. Hendry, J.P., Kalin, R.M., 1997. Are oxygen and carbon Palaeotemperature of the last interglacial period based on d18O isotopes of mollusc shells reliable paleosalinity indicators of Strombus bubonius from the western Mediterranean Sea. in marginal marine environments? A case study from the Palaeogeography, Palaeoclimatology, Palaeoecology 103, 1–20. Middle of England. Journal of the Geological Society Dodd, J.R., Stanton, R.J., 1975. Palaeosalinities within a Pliocene 154, 321–333. Bay, Kettleman Hills, California: a study of the resolving power Hladilova, S., Hladikova, J., Kovac, M., 1998. Stable Isotope record of isotopic and faunal techniques. Geological Society of in Miocene Fossils and Sediments from Rohoznik (Vienna America Bulletin 86, 51–64. Basin, Slovakia). Slovak Geological Magazine 4/2, 87–94. Durakiewicz, T., Gonera, M., Peryt, T.M., 1997. Oxygen and carbon Holmden, C., Creaser, R.A., Muehlenbachs, K., 1997. Paleosalin- isotopic changes in the Middle Miocene (Badenian) forami- ities in ancient brackish water systems determined by 87Sr/86Sr nifera of the Gliwice are (SW Poland). Bulletin of the Polish ratios in carbonate fossils: a case study from the Western Canada Academy of Sciences. Earth Sciences 45, 145–156. Sedimentary Basin. Geochimica et Cosmochimica Acta 61/10, Epstein, S., Mayeda, T.K., 1953. Variations of the 180/160 ratio in 2105–2118. natural waters. Geochimica et Cosmochimica Acta 4, 213–224. Houbrick, R.S., 1991. Systematic review and functional morphol- Geary, D.H., Rich, J., Valley, J.W., Baker, K., 1989. Stable isotopic ogy of the snails and evidence of salinity change: influence on the evolution of (; Prosobranchia). Malacologia 33, 289–338. melanopsid gastropods in the Late Miocene Pannonian basin. Hudson, J.D., Riding, J.B., Wakefield, M.I., Walton, W., 1995. Geology 17/11, 981–985. Jurassic paleosalinities and brackish-water communities—a case Geary, D.H., Brieske, T.A., Bemis, B.E., 1992. The influence and study. Palaios 10, 392–407. interaction of temperature, salinity, and upwelling on the Ja´mbor, A` ., 1978. U´ j elke´pzele´sek a magyarorsza´gi Neoge´n stable isotopic profiles of strombid gastropod shells. Palaios 7, tengerek vize´nek so´tartalom-va´ltoza´sio´l (New ideas about the 77–85. changing salinity of the Neogene seas pre-existing in the Glaubrecht, M., 1993. Mapping the diversity: geographical distri- present-day Hungarian Territory). A Magyar A´ llami Fo¨ldtani bution of the Melanopsis (Gastropoda:? Inte´zet E´ vi Jelente´se Budapest, 1976, pp. 260–265. In : Melanopsis) with focus on its systematics in the Hungarian with English abstract. Mediterranean Basin. Mitteilungen aus dem Hamburgischen Keith, M.L., Parker, R.H., 1965. Local variation of 13C and 18O Zoologischen Museum und Institute 90, 41–97. content of mollusk shells and the relatively minor temperature Gonera, M., Peryt, T.M., Durakiewicz, T., 2000. Biostratigraphical effect in marginal marine environments. Marine Geology 3, and palaeoenvironmental implications of isotopic studies (18O, 115–129. C. Latal et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 211 (2004) 157–169 169

Keith, M.L., Weber, J.N., 1964. Carbon and oxygen isotopic Ecosystem Dynamics of the Eurasian Neogene (EEDEN)- composition of selected limestones and fossils. Geochimica et Stratigraphy and Paleogeography, Workshop Graz 2001. Cosmochimica Acta 28, 1787–1816. Berichte des Institutes fqr Geologie und Pal7ontologie der Keith, M.L., Anderson, G.M., Eichler, R., 1964. Carbon and oxygen Karl-Franzens-Universit7t Graz, O¨ sterreich, vol. 4, pp. 49–59. isotopic composition of mollusk shells from marine and Romanek, C.S., Grossman, E.L., 1989. Stable isotope profiles of freshwater environments. Geochimica et Cosmochimica Acta Tridacna maxima as environmental indicators. Palaios 4, 28, 1757–1786. 402–413. Kobashi, T., Grossman, E.L., Yancey, T.E., Dockery III, D.T., 2001. Rothe, P., Hoefs, J., Sonne, V., 1974. The isotopic composition of Reevaluation of conflicting Eocene tropical temperature esti- Tertiary carbonates from the Mainz Basin: an example of mates: molluskan oxygen isotope evidence for warm latitudes. isotopic fractionations in bclosed basinsQ. Sedimentology 21, Geology 29/11, 983–986. 373–395. Lear, C.H., Elderfield, H., Wilson, P.A., 2000. Cenozoic Deep-Sea Schmid, H.P., Harzhauser, M., Kroh, A., 2001. Hypoxic Events on a Temperatures and Global Ice Volumes from Mg/Ca in Benthic Middle Miocene Carbonate Platform of the Central Paratethys Foraminferal Calcite. Science 287/5451, 269–272. (Austria, Badenian, 14 Ma). Annalen des Naturhistorischen Lloyd, R.M., 1964. Variations of the oxygen and carbon isotope Museums in Wien 102 A, 1–50. ratios of Florida Bay mollusks and their environmental Schmitz, B., Andreasson, F.P., 2001. Air humidity and lake significance. Journal of Geology 72, 84–111. d18O during the Paleocene–earliest Eocene in France from Matyas, J., Burns, S.J., Mqller, P., Magyar, I., 1996. What can stable recent and fossil fresh-water and marine gastropod d18O, isotopes say about salinity? An example from the Late Miocene d13C, and 87Sr/86Sr. Geological Society of American Bulletin Pannonian Lake. Palaios 11, 31–39. 113/6, 774–789. Mienis, H.K., 1983. A preliminary checklist of the freshwater Stanton, R.J., Dodd, J.R., 1970. Paleoecologic techniques— molluscs of Israel and the administered areas. Levantina 47, comparison of faunal and geochemical analyses of Pliocene 543–550. paleoenvironments, Kettleman Hills, California. Journal of Mook, W.G., 1971. Paleotemperatures and chlorinities from stable Paleontology 44, 1092–1121. carbon and oxygen isotopes in shell carbonate. Palaeogeogra- Steininger, F.F., 1991. Das Terti7rinO¨ sterreich-ein U¨ berblick. In: phy, Palaeoclimatology, Palaeoecology 9, 245–263. Roetzel, R., Nagl, D. (Eds.), Exkursionen im Terti7rO¨ sterreichs. Nishihira, M., 1983. Ecological distribution, population structure O¨ sterreichische Pal7ontologische Gesellschaft, pp. 1–11. and feeding habit of (Linne´) in the Sutovska, K., Kantor, J., 1992. Oxygen and carbon isotopic analysis mangrove swamp of Funaura in Iriomote Island, Okinawa. of Karpatian foraminifera from LKS borehole (Southern Survey Report for the Fishing-ground Development in Irimote Slovakian Basin). Mineralia Slovaca 24, 209–218. Island. Okinawa Agency, Okinawa, pp. 28–36. Swart, P.K., Sternberg, L., Steinen, R., Harrison, S.A., 1989. Papp, A., 1954. Die Molluskenfauna im Sarmat des Wiener Controls on the oxygen and hydrogen isotopic composition of Beckens. Mitteilungen der Geologischen Gesellschaft Wien waters from Florida Bay. Chemical Geology. Isotope Geo- 45, 1–112. science Section 79, 113–123. Patterson, W.P., Walter, L.M., 1994. Depletion of 13C in seawater Swart, P.K., Price, R.M., Greer, L., 2001. The relationship between

ACO2 on modern carbonate platforms: significance for the stable isotopic variations (O, H and C) and salinity in waters and carbon isotopic record of carbonates. Geology 22, 885–888. corals from environments in South Florida; implications for Piller, W.E., Harzhauser, M., 2002. Reinterpretation of Sarmatian reading the paleoenvironmental record. Bulletins of American Environments and Climatic Constraints in the Central Para- Paleontology 361, 17–29. tethys: 16. Internationale Senckenberg Konferenz Forschung- Tojo, B., Ohno, T., 1999. Continuous growth-line sequences in sinstitut und Naturmuseum Senckenberg, bThe Middle Miocene gastropod shells. Palaeogeography, Palaeoclimatology, Palae- CrisisQ, EEDEN-Programm, pp. 91–92. oecology 145, 183–191. Pisera, A., 1995. The role of skeletal and non-skeletal components Van Damme, D., 1984. The Freshwater of Northern in the Sarmatian (Miocene) reefs of Poland. Publications du Africa, Distribution, Biogeography and Paleoecology. Dr. W. Service Geologique du Luxembourg 29, 81–86. Junk Publishers, Dordrecht. Pisera, A., 1996. Miocene Reefs of the Paratethys: a review. In: Wefer, G., Berger, W.H., 1991. Isotope paleontology: growth and Franzen, E., Esteban, M., Ward, B., Rouchy, J.M. (Eds.), composition of extant calcareous species. Marine Geology 100, Models for Carbonate Stratigraphy from Miocene Reef Com- 207–248. plexes of Mediterranean Regions. Special Publication of the Yamada, S.B., 1982. Growth and longevity of the mud snail Society for Sedimentary Geology, vol. 5, pp. 97–104. atramentaria. Marine Biology 67, 187–192. Purton, L., Brasier, M., 1997. Gastropod carbonate d18O and d13C Zachos, J., Pagani, M., Sloan, L., Thomas, E., Billups, K., 2001. values record strong productivity and stratification shifts during Trends, rythyms, and aberrations in global climate 65 Ma to the Late Eocene in England. Geology 25/10, 871–874. present. Science 292/5517, 686–693. Ro¨gl, F., 2001. Mid-Miocene Circum-Mediterranean paleogeogra- phy. In: Latal, C., Piller, W.E. (Eds.), Environments and