Middle Miocene (Serravallian; Upper Badenian–Lower Sarmatian) Dinoflagellate Cysts from Bad Deutsch-Altenburg, Vienna Basin, Austria

ABSTRACT Middle Miocene (Serravallian; upper Badenian–lower Sarmatian) strata recovered in 10 cored boreholes (83 samples) from Bad Deutsch-Altenburg, Vienna Basin, Austria, were analysed palynologically for the first time. The strata belong to the Rabensburg Formation of the Baden Group. The lateral distribution of the boreholes in reference to a Mesozoic ridge makes this area interesting for studying various aspects such as distribution of deposits, stratigraphy and palaeoenvironmental reconstruction. A diverse and well-preserved in situ dinoflagellate cyst association has been identified. Middle Miocene age-diagnostic species including Cannosphaeropsis passio, Cerebrocysta poulsenii, Habibacysta tectata, Labyrinthodinium truncatum, Operculodinium? borgerholtense and Unipontidinium aquaeductum are recorded. Their occurrences allow correlation with dinoflagellate cyst biozonations on a regional scale. Based on the identified dinoflagellate cysts and by correlation with calcareous nannoplankton and ostracods, from the same set of samples, a Serravallian age – corresponding to a late Badenian and early Sarmatian age (regional Paratethys stages) – is confirmed. Reworked (Cretaceous and Paleogene) dinoflagellate cyst assemblages, also well preserved, were recorded abundantly from boreholes HA 521 and HA 573 (south-west of the Mesozoic ridge). In accordance with ostracods, the Badenian–Sarmatian boundary can be traced in the upper part of borehole HA 66 based on an abrupt change in the dinoflagellate cyst assemblages. The upper Badenian strata were deposited in a coastal to inner shelf environment with terrigenous (siliceous sand and clay) and carbonate sediments. The recorded dinoflagellate cysts reflect marine, tropical to warm-temperate climatic conditions. The composition of the recorded dinoflagellate cyst assemblages is very close to that of the Middle Miocene assemblages of the Mediterranean, indicating water exchanges between the Central Paratethys and the Mediterranean during the late Badenian (early Serravallian).


Introduction
The Paratethys Sea came into existence when the Western Tethys was split by the rising Alpine chains into the Proto-Mediterranean Sea in the south and the Paratethys Sea in the north. This separation occurred around the Eocene/ Oligocene boundary and the connections of both realms changed frequently over time, with temporarily active gateways between the Paratethys and the Proto-Mediterranean and also with the North Sea and the Indian Ocean, but they suffered from palaeogeographical isolation during intervals in between. The Paratethys itself is subdivided into a Western/Central part extending from the Gulf of Lyon in the west to the Pannonian Basin in the east, and the Eastern Paratethys covering the area from the Pannonian Basin in the west and far east to the extant Lake Aral. The Central Paratethys was predominantly marine during the Oligocene to Middle Miocene but changed to the long-lived brackish/ freshwater Lake Pannon during the early Late Miocene (e.g. Magyar et al. 1999).
The connections/separations of the Paratethys to/from other oceans and seas resulted in a specific palaeogeographical and palaeoenvironmental development which necessitated the establishment of a regional chronostratigraphical scheme (Figure 1), based mostly on endemic biota, that cannot always easily be correlated to the global chronostratigraphy. Correlation outwith the Paratethys is based on planktonic foraminifera and calcareous nannoplankton and, to a minor extent, on magnetostratigraphy and sequence stratigraphy. Most of the biota, e.g. foraminifera, ostracods, molluscs and echinoids, are well studied in the Central Paratethys. Dinoflagellate cysts have been described but a continuous record is missing. From the Oligocene, there is only poor documentation of the North Alpine Foreland Basin (NAFB) (e.g. Hochuli 1978;Sachsenhofer et al. 2010;Soliman 2012) covering the Kiscellian regional stage. A better record exists for the Early and Middle Miocene; Lower Miocene successions, mostly Ottnangian, were studied in the NAFB (e.g. Hochuli 1978;Jim enez-Moreno et al. 2006;Grunert et al. 2010;Palzer-Khomenko et al. 2018a, 2018b, and the Lower/ Middle Miocene (Karpatian/Badenian) boundary was studied by Soliman and Piller (2007) Harzhauser et al. 2022). (B) Miocene geochronology, geomagnetic polarity chrons, biozonations of planktonic foraminifera and calcareous nannoplankton, sequence stratigraphy, third-order sea-level curve, and oxygen isotope stratigraphy partly recalibrated and correlated to the regional chronostratigraphy of the Central Paratethys. The light blue strip represents the studied interval (after Piller et al. 2007 and references therein). Soliman and Riding (2017) and Soliman and Lucas-Clark (2018) in the Vienna and Styrian basins.
An important gap in knowledge of the VB exists in the late Badenian (early Serravallian) from which no dinoflagellate cysts have been described so far. In this study, we describe for the first time late Badenian dinoflagellate cyst associations, and discuss their stratigraphical importance, taphonomic interpretation and palaeoenvironmental indications in an already well-known setting.

Geological background of the Vienna Basin
The VB is one of the largest hydrocarbon reservoirs in Europe and also one of the best known pull-apart basins in the world (Hamilton et al. 2000;Arzm€ uller et al. 2006;Boote et al. 2018;Rupprecht et al. 2019). However, the basin did not originate from simple pull-apart kinematics, as supposed by earlier studies (e.g. Royden 1985Royden , 1988; its evolution is much more complex, starting with a series of piggy-back  Table 1 (modified from Gross 2002Gross , 2006. (D) Simplified geological map with the position of the Vienna Basin related to the Alpine-Carpathian Mountains and the Pannonian Basin System (modified after Wessely 1998). RDFU ¼ Rhenodanubian Flysch Unit; NCA ¼ Northern Calcareous Alps; CA ¼ Central Alps. The green colour marks the boreholes in the north-east of the Mesozoic ridge, the brown colour indicates the boreholes on the crest of the ridge and yellow indicates those boreholes in the south-west.
basins at the front of the north-to north-west-propagating thrust belt of the Eastern Alps. This initial phase is also called the (proto-)VB, developed during the Early Miocene (Fodor 1995;Decker 1996;Lee and Wagreich 2017). In the Middle to Late Miocene, the basin was shaped by extensional tectonics due to the lateral extrusion of the Eastern Alps (Fodor 1995;Decker 1996;H€ olzel et al. 2010;Lee and Wagreich 2017). Consequently, distinct fault systems developed, subdividing the VB into a complex horst and graben system and creating several sub-basins (Kr€ oll and Wessely 1993;Hamilton et al. 2000;Wessely 2006;H€ olzel et al. 2010;Siedl et al. 2020). The basin extends c. 200 km in a north-south direction and c. 55 km west-east; it lies mostly in Austria but also reaches into the Czech and the Slovak republics. Sedimentation started in the Early Miocene (Ottnangian) and continued to the early Late Miocene (Pannonian). The maximum sediment thickness reaches about 5500 m in the Schwechat deep. The complex tectonics with fault-bounded uplifted blocks and depressions make correlation within the VB difficult. This pattern also produced a complex facies distribution in the basin interrelated with synsedimentary tectonics (e.g. Harzhauser et al. 2017Harzhauser et al. , 2019Siedl et al. 2020). The interplay of complex regional tectonics and global/ regional sea-level changes (Strauss et al. 2006;Piller et al. 2007) produced a complex pattern of lithostratigraphical units which have been recently revised (Harzhauser et al. 2019. Due to many deep drillings in the basin and intra-basinal seismic lines, correlation has been carried out by wireline logs supported by biostratigraphy which represents a poorly constrained ecobiostratigraphy . Correlation with open oceanic equivalents is mostly based on planktonic foraminifera and calcareous nannoplankton, but also on cyclostratigraphy and astronomical tuning (Hohenegger et al. 2009;Lirer et al. 2009;Hohenegger and Wagreich 2012).
The Badenian age (Langhian-early Serravallian) is an interval of effective water exchange with the Mediterranean Sea via the Slovenian 'Trans-Tethyan Trench Corridor' (Bistricic and Jenk 1985) ( Figure 1A) and shows the highest biodiversity in the Miocene of the Central Paratethys . Also, in the VB this open connection is well expressed by the recognition of three third-order sequences subdividing the Badenian deposits into lower, middle and upper substages ( Figure 1B). The base of the Badenian is characterised by a hiatus, which, in addition to the global sea level lowstand Bur5/Lan1, is caused by the so-called Styrian Tectonic Phase (Stille 1924) and represents the lower Badenian Ba1 sequence (Strauss et al. 2006), which can be correlated to the third-order sequence TB 2.3 of Hardenbol et al. (1998), ending again with a hiatus. The middle Badenian is reflected by the hiatus-bounded sequence Ba2 (TB 2.4) and the upper Badenian by sequence Ba3 (TB 2.5), starting again with a hiatus representing the lowstand Ser2. The Badenian-Sarmatian boundary is also characterised by a major sea-level drop and by a subsequent restriction from the Mediterranean area resulting in reduced biodiversity .
Palaeoenvironmental conditions changed substantially through the Badenian (and Sarmatian), guided by strong tectonics and palaeoclimate Kranner et al. 2021aKranner et al. , 2021b. Besides the application of planktonic foraminifera and calcareous nannoplankton for correlation to the Mediterranean Miocene within the VB, an ecobiostratigraphical zonation was developed (Grill 1941) and repeatedly modified, but more recently abandoned (Kranner et al. 2021a(Kranner et al. , 2021b. The zonation mostly used consists of the Lower Lagenid and Upper Lagenid zones, the Spirorutilus Zone and the Bulimina-Bolivina Zone. The Lagenid zones cover the lower Badenian and the lower part of the middle Badenian, the Spirorutilus Zone covers the middle Badenian, and the Bulimina-Bolivina Zone covers the upper Badenian. Both the definition and the stratigraphical extent of these foraminiferal zones are poorly constrained .

Study area
The studied area is located in the Hainburg Mountains (Lower Austria), which mark the eastern margin of the VB and also the boundary with the Danube Basin in the east ( Figure 2). The latter is part of the Pannonian Basins System and exhibits a diverging origin from the VB. The Hainburg Mountains represent the southern end of the Little Carpathians and belong geologically to the Tatrid Unit. The Hainburg Mountains are made up of granites, gneiss and mica schists and Permo-Triassic quartzites overlain by Middle Triassic limestone and dolomite (Wessely 1961;Kristan-Tollmann and Spendlingwimmer 1975). These rocks are occasionally covered by Badenian and Sarmatian strata. A detailed description of the geology of the Hainburg Mountains is provided by Wessely (1961), who also studied the Miocene sedimentary rocks overlying the crystalline basement and Mesozoic cover. He described a transgression of Badenian strata onto the basement along a steep shoreface with coralline algal limestones and marls in more distal positions. Biostratigraphically, he classified these deposits as part of the foraminiferal Spirorutilus and Bulimina-Bolivina zones.
In the reconnaissance phase of the Danube powerplant project at Hainburg, possible impacts on the mineral springs  at Bad Deutsch-Altenburg were also studied in great detail. For this sensitive topic, a dense grid of 78 shallow, fully cored boreholes were drilled in the spa area and its surroundings as well as in the Danube river bed and on the left river bank, mostly located in the municipality of Bad Deutsch-Altenburg (Gangl 1988(Gangl , 1990) ( Figure 2C). These drillings penetrated Quaternary-Miocene strata, and some boreholes also reached the Mesozoic basement. The specific geological situation necessitated this detailed study with a Mesozoic dolomite ridge which crops out at the Kirchenberg and Pfaffenberg in Bad Deutsch-Altenburg but dips to the north-west into the deeper parts of the VB ( Figure 2C). This ridge is onlapped and overlain by Badenian strata, which show small-scale lateral and vertical facies variations. Above a transgressive conglomerate, the Triassic dolomite is overlain by coralline algal limestones referred to in the literature as 'Leitha Limestone' or 'Leitha Limestone Facies' (Wessely 1961(Wessely , 2006Gangl 1988Gangl , 1990. Overall, the Badenian sedimentary rocks in the south-west of the Triassic spur are dominated by sandy and marly strata, while in the north-east of the spur coralline algal limestones prevail. This unusual situation makes this location interesting for studying various aspects, such as reworking of the Mesozoic dinoflagellate cysts. So far, the focus has been placed mainly on the study of ostracods, with the first study and new descriptions of taxa by Danielopol et al. (1991). A comprehensive examination of ostracods was later carried out by Gross (2002Gross ( , 2006 based on 66 samples from 10 boreholes. In accordance with the data from Wessely (1961), most of the samples (59) were assigned to the upper Badenian; only a few (seven) from the tops of five boreholes were placed in the lower Sarmatian (HA 66, HA 519, HA 540, HA 541, HA 573). The Badenian samples all indicate fully marine conditions; the ostracod assemblages indicate freshwater input in the sandier samples south-west of the spur and an epi-to mesoneritic environment in the north-east of the spur (Gross 2002).

Material and methods
A total of 83 samples from 10 shallow boreholes (HA 66, HA 511, HA 512, HA 514, HA 521, HA 529, HA 540, HA 542, HA 549, HA 573) have been studied. Three cores originate from the crest of the Mesozoic ridge (HA 514, HA 540, HA 549), two from south-west of this ridge (HA 521, HA 573) and five from the north-east (HA 66, HA 511, HA 521, HA 529, HA 542) (Figure 2). Six boreholes reached the Mesozoic Basement (for details see Table 1). Core description and sampling were carried out directly after drilling. Samples are stored at the Institute of Earth Sciences, University of Graz (Austria).
All samples have been palynologically processed (Soliman 2012). Approximately 10-20 g of dry sediment, whenever possible, was processed from each sample. To remove the carbonate, up to 30-50 ml of hydrochloric acid (HCl, 38%) was added, left for 24 hours and agitated frequently. After that, the acid was decanted, and the residues were washed three times. To remove silicates, up to 40 ml of hydrofluoric acid (HF, 48%) was added and left for 48 hours, agitated frequently, and then washed with distilled water. The remaining residue was put in an ultrasound bath for c. 10-20 seconds, sieved at 20 lm, and stained with 'Safaranin O'. No oxidation was applied. Two microscope slides (76 Â 26 mm with a 24 Â 50 mm cover slip) were made from each sample using glycerine jelly as a mounting medium and sealed with nail varnish. Additional slides were made from some samples to aid in determining the presence/absence of the marker taxa. Mounts for Scanning Electron Microscope (SEM) studies were made by air drying water-suspended residues (sieved at 30 mm) on circular glass cover slips (12 mm) mounted on aluminium stubs with thin double-sided sticky tape. Stubs were coated with platinum. All slides were scanned at Â200 and Â400 magnification. The microscopical investigation was carried out, and the transmitted light photomicrographs were taken, using a bright field Carl Zeiss (Axioplan 2) microscope fitted with a Leica DFC320 Digital Camera and a Leica DM 750 microscope fitted with Leica MC170HD digital camera. SEM observations and photomicrographs were made using a LEO DSM 982 Gemini SEM, operating at a working voltage of 10 kV. These studies and documentation were conducted at the Institute of Earth Sciences, University of Graz (Austria). Fluorescence microscopy used an Optika IM-3F microscope housed in the Faculty of Pharmacy, Tanta University (Egypt). All microscope slides and residues are housed in the Geology Department, Tanta University.
Up to 200 organic-walled dinoflagellate cysts (in situ and reworked) were counted in most samples. The remainder of the slides was then scanned for rare taxa and well-preserved specimens. Acritarchs, green algae, foraminiferal test linings and terrestrial palynomorphs were also recorded during the counting.
The nomenclature of dinoflagellate cysts and acritarchs follows Fensome et al. (2019, and references therein) and Head et al. (2020). Because of their limited stratigraphical importance during the Neogene, all species of the genera Spiniferites and Achomosphaera are grouped together as Spiniferites/Achomosphaera spp. The full names of the identified taxa are listed in Supplementary material: Appendices A and B and the marker taxa are illustrated in Plates 1-7.

Palynology
Although the palynomorphs of 10 boreholes have been studied, we give more detailed information only for five boreholes, which represent the three different geographical positions/facies (north-east of the ridge: HA 66, HA 511; top of the ridge: HA 540; and south-west of the ridge: HA 521, HA 573) and the majority of the samples (64). The data from the other five boreholes (HA 512, HA 514, HA 529, HA 542, HA 549) support the data from the five boreholes discussed in detail herein. The full dataset is provided in the Supplementary data.
Dinoflagellate cysts (both in situ and reworked) and other palynomorphs (acritarchs, prasinophytes, pollen and spores, foraminiferal test linings) show fair to good preservation in most of the investigated samples. None of the studied samples was completely barren of palynomorphs. Overall, a considerable variation in Miocene dinoflagellate cyst diversity has been recognised throughout the studied interval (Figures 3-7; Supplementary data). The reworked dinoflagellate cyst taxa originate from the Upper Cretaceous and Paleogene.
In general, the gonyaulacoid taxa Spiniferites, Achomosphaera, Lingulodinium, Operculodinium, Hystrichokolpoma, Melitasphaeridium, Reticulatosphaera and Cleistosphaeridium are common in most samples. Peridinioids occur only sporadically or are completely absent. Homotryblium is abundant in samples where the assemblage is characterised by high numbers of reworked dinoflagellate cysts. The highest occurrence (HO) and highest common occurrence of Homotryblium in the Mediterranean and north-west Europe are assigned to the upper Oligocene-Lower Miocene (e.g. Powell 1986;El-Mehdawi and El Beialy 2008;Dybkjaer et al. 2021). In the studied material, we treated the Homotryblium specimens as reworked (see the section titled Reworked dinoflagellate cysts).
Pollen and spores were not identified to genus or species level during routine counting. The pollen assemblages are dominated by the gymnosperms Pinus, Cathaya, Abies, Tsuga and Picea and the angiosperms Liguliflorae and Chenopodiaceae. Spores are mostly represented by pteridophytes (e.g. Schizaeaceae, Pteridaceae).

Borehole HA 66
Twenty-six samples were investigated. Marl is the dominant lithofacies, with some coralline algal limestones, while sand and sandstone intercalations are present in the lower and upper parts. The Middle Miocene deposits unconformably overlie Mesozoic and underlie Quaternary deposits (Figure 3). The marine/non-marine ratio oscillates around 50%, but a notable decline is recorded in sample HA 66/6 ( Figure 3). The dinoflagellate cyst diversity is, overall, moderate. This borehole is considered as a reference section for the study area.
The samples are productive, with fair to good preservation of the dinoflagellate cysts. The taxa Spiniferites/ Achomosphaera spp., Operculodinium spp., Hystrichokolpoma spp., Lingulodinium machaerophorum, Melitasphaeridium choanophorum, Reticulatosphaera actinocoronata and Cleistosphaeridium placacanthum are abundant. Common to frequent occurrences of Labyrinthodinium truncatum, Batiacasphaera sphaerica, Hystrichosphaeropsis obscura, Tectatodinium pellitum, Nematosphaeropsis labyrinthus, Pentadinium laticinctum, Operculodinium? borgerholtense, Unipontidinium aquaeductum, Operculodinium piaseckii and Polysphaeridium zoharyi were also encountered ( Figure 3). Additional well-known Miocene taxa that occurred rarely include Minisphaeridium latirictum, Habibacysta tectata and Cannosphaeropsis passio. In general, peridinioid dinoflagellate cysts such as Lejeunecysta and Selenopemphix are rare, with an average of 3%. Other aquatic palynomorphs were also recorded and are represented by Cyclopsiella spp., Cymatiosphaera spp., Pterospermella spp., Paralecaniella indentata and Nannobarbophora gedlii. Based on the distribution of the encountered dinoflagellate cysts, three intervals can be distinguished: (i) the lower interval includes samples HA 66/30 to HA 66/24 and is characterised by sandy facies intercalations. The dinoflagellate cyst assemblages recovered in this interval are characterised by a low diversity of in situ taxa, with notable occurrences of reworked taxa in samples HA 66/25 and HA 66/24 ( Figure 3). (ii) The middle interval includes samples HA 66/22 to HA 66/ 7 and represents the peak abundance and high diversity of in situ Miocene taxa, associated with the common occurrence of open-marine index taxa, e.g. Nematosphaeropsis labyrinthus and Impagidinium spp. Reworked taxa are absent from this interval. This interval may represent the system tracts TST3 and HST3 of Strauss et al. (2006). (iii) The upper interval is represented by the samples HA 66/6 to HA 66/1. It is characterised by a dominance of reworked taxa, with low percentages of in situ Miocene taxa. This may indicate a shallowing trend continuing into the ?Sarmatian. Interestingly, sample HA 66/6 shows a low recovery of dinoflagellate cysts, and many of the Middle Miocene taxa have not been recorded in or above this sample. Thus, the interval between samples HA 66/7 and HA 66/5 may contain a hiatus between the Badenian and Sarmatian.

Borehole HA 511
This borehole is located north-east of the Mesozoic ridge. Seven samples were selected and investigated for their dinoflagellate cyst content. The lithology comprises marl, sandy marl, coralline limestone and two intercalations of sandstone ( Figure 4) Homotryblium in samples HA 511/4 and HA 511/5 were noticed, but no Cretaceous/Paleogene reworked taxa were recorded during the routine counting. The occurrences of Labyrinthodinium truncatum, Unipontidinium aquaeductum, Cerebrocysta poulsenii and Operculodinium? borgerholtense strongly indicate a Middle Miocene age (Serravallian; late Badenian). The Uaq and Cpo dinoflagellate cyst biozones of Jim enez- Moreno et al. (2006) are represented in this borehole (see the Dinoflagellate cyst biostratigraphy section).

Borehole HA 521
This borehole is located south-west of the Mesozoic ridge ( Figure 2C). Seven samples were selected for palynomorph analysis. The lower part of the borehole is dominated by marls, while the upper part is dominated by sandstones with some marl intercalations ( Figure 6). The marine proportion is low, close to 10% in all studied samples except sample HA 521/1 in which it reaches 30%. Also the diversity of in situ dinoflagellate cysts is lowgenerally around 10 species per sample ( Figure 6). Reworked Cretaceous and Paleogene dinoflagellate cyst taxa (e.g. Deflandrea spp., Axiodinium augustum, Dracodinium spp., Enneadocysta spp., Lanternosphaeridium spp., Rhombodinium spp., Subtilisphaera spp.) are the main constituents of the assemblages in all samples ( Figure 6). Also, Homotryblium spp. are abundant in all samples, but most pronounced in the upper part. In addition, occurrences of some long-ranging taxa including Reticulatosphaera actinocoronata, Polysphaeridium zoharyi, Lingulodinium machaerophorum, Melitasphaeridium choanophorum, Spiniferites spp. and Operculodinium spp. were encountered. Single occurrences of the Middle Miocene species Labyrinthodinium truncatum (HA 521/1) and Operculodinium? borgerholtense (HA 521/5) were recorded. Notable occurrences (19-22%) of the protoperidinioid taxa Lejeunecysta and Selenopemphix in the lower part of the section indicate nutrient-rich water. Evidence of strong terrigenous influx from the same set of samples in this borehole has been detected by ostracods (Gross 2002). The absence of Middle Miocene marker taxa, such as Unipontidinium aquaeductum hampers the precise age assignment of these samples and no dinoflagellate cyst biozones are recognised. However, a late Badenian age is suggested based on ostracods (Gross 2002).

Borehole HA 573
Nineteen samples from marl, sandy marl, sand and sandstone were selected for palynological analysis (Figure 7). HA 573 is located south-west of the Mesozoic ridge. The marine ratio does not exceed 20%, except in sample HA 573/4 where it reaches 80% and the diversity of the in situ dinoflagellate cysts is low (generally fewer than 10 species per sample) (Figure 7). The assemblage consists mostly of reworked taxa (e.g. Areosphaeridium diktyoplokum, Phthanoperidinium filigranum, Membraniphorum sp., Rhadinodinium politum, Cerodinium sp., Wetzeliella symmetrica, Cerodinium striatum, Circulodinium sp. and Glaphyrocysta wilsonii) from the Cretaceous and Paleogene. Homotryblium spp. are also abundant. The remaining dinoflagellate cysts are regarded as in situ Miocene taxa; however, some of them are long-ranging taxa, such as Reticulatosphaera actinocoronata, Lingulodinium machaerophorum, Melitasphaeridium choanophorum, Nematosphaeropsis labyrinthus, Batiacasphaera sphaerica, Polysphaeridium zoharyi, Hystrichosphaeropsis obscura and Tuberculodinium vancampoae, in addition to Spiniferites spp. and Achomosphaera spp.  The core is characterised by high abundances of terrigenous freshwater material transported from the hinterland based on the ostracod data (Gross 2002). This is supported by the abundance of miospores and, especially, bisaccate pollen (Figure 7). Due to the absence of Middle Miocene marker taxa from nearly all samples, no assignment to dinoflagellate cyst biozones has been made. Gross (2002) placed the Badenian-Sarmatian boundary between samples HA 573/ 3 and HA 573/2 based on ostracods, but there is no clear evidence from the dinoflagellate cysts.

Integrated stratigraphy
The studied deposits have been assigned to the regional foraminiferal Bulimina-Bolivina Zone and therefore to the upper Badenian (Wessely 1961). Further support for this assignment comes from the study of uvigerinid foraminifera (Haunold 1995) and ostracods (Danielopol et al. 1991;Gross 2002Gross , 2006. This biostratigraphy is well established in the VB, but it is an ecobiostratigraphy in a rather restricted marginal sea and therefore of questionable reliability. To provide a better framework for this regional stratigraphy, the calcareous nannoplankton from borehole HA 66 has been studied. Samples HA 66/30 and HA 66/28 are barren of nannoplankton, but HA 66/19 shows a diverse and moderately preserved flora, containing Discoaster exilis, Coccolithus pelagicus, Reticulofenestra pseudoumbilicus (> 7 lm), common Calcidiscus premacintyrei and Helicosphaera walbersdorfensis, and rare Cyclicargolithus floridanus and Coronocyclus nitescens. This assemblage indicates the nannoplankton zones (lower) NN6 (Martini 1971), MNN6b (Fornaciari et al. 1996) and CNM8 (Backman et al. 2012). Nannofossils in samples HA 66/8 and HA 66/7 were moderately to well preserved, with common occurrences of Helicosphaera walbersdorfensis, together with rare Helicosphaera stalis and Helicosphaera vederi. Based on their abundant Coccolithus pelagicus and Reticulofenestra pseudoumbilicus, the occasional presence of Cyclicargolithus floridanus and the notable absence of Calcidiscus premacintyrei, these samples were assigned to the zones (upper) NN6 corresponding to (lower) MNN7 and lower CNM9 (Martini 1971;Fornaciari et al. 1996;Backman et al. 2012). Reworked Cretaceous and Paleogene taxa are very common. Sample HA 66/5 is nearly barren of stratigraphically indicative in situ taxa, although common Coccolithus pelagicus, as well as rare Reticulofenestra minuta, Reticulofenestra haqii and Reticulofenestra floridanus occur; reworked Cretaceous and Paleogene taxa are common. The NN6 Zone gives a very clear indication of Serravallian, with CNM8 and CNM9 indicating lower and upper Serravallian, respectively. For the Central Paratethys, the CNM8 Zone correlates with the upper Badenian, and CNM 9 would indicate Sarmatian. The latter, however, is primarily based on the absence of Calcidiscus premacintyrei and therefore not very strongly supported.
The calcareous nannoplankton assemblages and also the dinoflagellate cysts clearly match the earlier correlation. The late Badenian age also matches the established sequence stratigraphy, which clearly places the Badenian deposits of the studied boreholes in the regional sequence Ba3 (Strauss et al. 2006;Siedl et al. 2020) ( Figure 1B). Ba3 can be correlated with the third-order sequence TB 2.5 of Hardenbol et al. (1998) which is associated with the lowstands Ser2 and Ser3. This sequence stratigraphical approach has been verified both in the southern (Strauss et al. 2006) and the northern VB ). The studied cores from Bad Deutsch-Altenburg begin with a transgression which can be correlated to the TST of TB 2.5 and the respective HST. The boundary between the upper Badenian and the lower Sarmatian is recognised with ostracods in the uppermost parts of several boreholes, but a hiatus has not been detected. This can be explained by the widely uniform marls at the top of the Badenian and at the base of the Sarmatian.
This clear biostratigraphical and sequence stratigraphical assignment allows the Badenian deposits in the Hainburg Mountains to be classified lithostratigraphically within the Rabensburg Formation, which was defined and described by Harzhauser et al. (2020). This unit belongs to the Baden Group, is well known from many deep wells in the VB and reaches up to 1000 m in thickness. It is laterally highly variable due to erosion at the Badenian-Sarmatian boundary, representing the Ser3 lowstand. The lithological composition is highly variable, with a dominance of marls but also with corallinacean limestones, fluvial gravel and even lignite seams . The corallinacean limestones resemble the recently defined Leitha Formation which is, however, restricted to the middle Badenian   (Figures 3-8). The marker taxa are directly compared to the global compilation of Williams et al. (2004) and other Neogene dinoflagellate cyst biozonations (Munsterman and Brinkhuis 2004;Jim enez-Moreno et al. 2006;Dybkjaer and Piasecki 2010;Schreck et al. 2012;Dybkjaer et al. 2021 King (2016) and Raffi et al. (2020), its lowest occurrence (LO) and HO, respectively mark the base and top of the dinoflagellate cyst Zone DM5 ranging from upper Langhian to middle Serravallian. Habibacysta tectata occurs in samples HA 66/19 and HA 66/21 and has its LO in the lowermost Serravallian. The two taxa together indicate a late Badenian age. Cannosphaeropsis passio, recorded in samples HA 66/17 and HA 66/19, is considered as a marker for the dinoflagellate cyst Zone DM6b, since its LO and HO delimit the base and top of this zone, respectively. Following the DM zonation, this species indicates a late Serravallian age, which accords with the Sarmatian of the Central Paratethys. However, Cannosphaeropsis passio and Unipontidinium aquaeductum co-occur in sample HA 66/19. The co-occurrence of Cannosphaeropsis passio and Unipontidinium aquaeductum contradicts the zonation of King (2016), according to which the ranges of these two taxa are separated by a gap representing subzone DM6a. The co-occurrence of the two species in this material may be related to a local geological issue. The first of these is the Unipontidinium aquaeductus Interval Biozone (Uaq), whose base is defined by the top of the Cribroperidinium tenuitabulatum Assemblage Biozone (Cte) and the top by the HO of Unipontidinium aquaeductus. The authors assigned this zone to the lower Serravallian. The Uaq Biozone is followed by the Cerebrocysta poulsenii Assemblage Biozone (Cpo) with its top correlated to the late Serravallian. The uppermost zone in the biozonation of Jim enez- Moreno et al. (2006) is the Cleistosphaeridium placacanthum Assemblage Biozone (Cpl), which is assigned to the lower Sarmatian (upper Serravallian) based on other fossils, because the dinoflagellate cysts have an endemic character (Jim enez- Moreno et al. 2006, p. 134).
In our samples, Cerebrocysta poulsenii is rather rare and occurs already at the base of some cores, immediately above the Triassic dolomite (e.g. Ha 573/19), but its HO is usually above the HO of Unipontidinium aquaeductum as in the HA 66 and HA 511 boreholes. This bioevent was also documented by Jim enez-Moreno et al. (2006, fig. 3). Cleistosphaeridium placacanthum occurs consistently in all the investigated samples. Its abundance (ca. 8-13%) in samples HA 66/10 and HA 66/9 is contemporaneous with the HO of Labyrinthodinium truncatum and Operculodinium? borgerholtense. These bioevents define the boundary between the dinoflagellate cyst zones Cpo and Cpl of Jim enez-Moreno et al. (2006, p. 133-134).
In summary, the studied sections can be correlated to the dinoflagellate cyst Zone (upper) DM5 of King (2016) based on the co-occurrence of Habibacysta tectata and Unipontidinium aquaeductum (Figure 8). This zone correlates with the Uaq Zone of Jim enez- Moreno et al. (2006) and indicates a late Badenian (early Serravallian) age for the studied succession. Cannosphaeropsis passio, although occurring rarely (two single specimens in HA 66/17 and HA 66/19, respectively), points to the DM6b Zone of King (2016) which would allow assignment to the Sarmatian (upper Serravallian).
The recorded dinoflagellate cyst markers Unipontidinium aquaeductum, Operculodinium piaseckii, Habibacysta tectata and Cannosphaeropsis passio thus support a late Badenian (Serravallian) age as indicated by calcareous nannoplankton. Furthermore, the occurrence of Cleistosphaeridium placacanthum and Cerebrocysta sp. A of Powell (1986) may indicate a Sarmatian age (late Serravallian) for the uppermost part of the studied succession. Gross (2002) identified the Badenian-Sarmatian boundary in the upper part of borehole HA 66, between samples Ha 66/7 and HA 66/6 ( Figure 3) and between samples HA 573/3 and HA 573/2 of HA 573 (Figure 7) based on ostracods. An abrupt decline in both total counts and in diversity of dinoflagellate cysts is seen from samples Ha 66/7 to HA 66/6, followed by an increase of reworked taxa and Homotryblium spp. in HA 66/5 (Figure 3). Interestingly, sample HA 66/6 is characterised by abundant bisaccate pollen grains and other unidentified miospores. Jim enez-Moreno et al. (2006, fig. 3) delineated the Badenian-Sarmatian boundary at the gap between the dinoflagellate cyst zones Cpo and Cpl. Following these authors, the Badenian-Sarmatian boundary in borehole HA 66 is located between samples HA 66/7 and HA 66/5. The boundary cannot be defined in borehole HA 573.

Reworked dinoflagellate cysts
Reworked well-preserved dinoflagellate cysts of Cretaceous and Paleogene age occur in high numbers and even dominate some samples (Figures 6 and 7; Supplementary material: Appendix B; Plates 6-7). The state of palynomorph preservation (Muller 1959), species colour (Stanley 1966) and chronostratigraphical range (Supplementary material: Appendix C) are conventionally used to distinguish in situ from reworked palynomorphs.
Several source areas for these cysts can be identified. Deposits of the Gosau Group on top of the Northern Calcareous Alps and of the Rhenodanubian Flysch Unit are restricted to this time interval with sedimentation ending in the Eocene ( Figure 2D). Both geotectonic units became uplifted in the course of the Alpine orogeny during the Oligocene to Early Miocene, becoming exhumed and exposed to erosion (Trautwein et al. 2001). Both units crop out at the western margin of the VB and the eroded material could easily have been transported across the basin to its eastern margin. A similar geotectonic history is reported from the Western Carpathians (Kov a c et al. 1994), which are very close and could also have acted as a source area. An additional source for Paleogene reworked biota is the occurrence of the upper Eocene deposits of Wimpassing an der Leitha (Burgenland), which represents an erosional relict of a former widespread Eocene cover (Fuchs 1980;Pahr and Herrmann 2000). These deposits are currently very poorly exposed and consist of limestones and sandstones with coralline algae, larger foraminifera (Nummulites), bryozoans, molluscs, etc. (Zorn 2000). The close vicinity of the boreholes at Bad Deutsch-Altenburg may explain the good preservation and the high abundance of reworked dinoflagellate cysts in some intervals of the studied boreholes.
The high abundance of reworked taxa in all samples of boreholes HA 573 and HA 521 reflects a very localised sedimentation regime. The Mesozoic ridge acted as a barrier with respect to sediment distribution. In the south-west and/ or west of the ridge, where these two boreholes are located, a higher amount of coarser siliciclastics occurs, whereas to the north-east of the ridge, where HA 511, HA 512 and HA 542 are located, limestones and marly limestones dominate. HA 66 is in an intermediate position with siliciclastics at the base, overlain by carbonates in the lower part of the section. The barrier thus seems to have prevented the transportation of coarser terrigenous input containing reworked taxa to the north-east.
Distinguishing between in situ and reworked taxa is easy when their stratigraphical distribution is clearly different (as it is in the above-mentioned cases; see Supplementary material: Appendix C). It may be more difficult, however, with long-ranging taxa. One such example is the genus Homotrybliumnot differentiated here at the species levelwhich is abundantly recorded in some intervals. The genus Homotryblium ranges from the late Paleocene to the Early Pliocene (Dybkjaer 2004) and raises the question whether specimens assigned to this genus are in situ or reworked.
A solution to this problem can be provided by fluorescence microscopy, a well-known technique applied in palynology to differentiate reworked from in situ palynomorphs (van Gijzel 1967;Strother et al. 2017;Hoyle et al. 2018). Van Gijzel (1971) and Tyson (1995) suggested that in situ palynomorphs fluoresce more strongly than reworked ones. To test these results and detect reworked palynomorphs, the palynological slides from boreholes HA 66 and HA 573 that are known to contain both in situ and reworked palynomorphs were studied to observe palynomorph fluorescence (Plate 8). Taxa known to be reworked (based on their stratigraphical ranges) show low fluorescence (Plate 8, figures 2, 4, 6) and the same low degree of fluorescence was seen in Homotryblium cysts (Plate 8,figures 4,8). The dinoflagellate cysts and miospores considered in situ clearly show a higher degree of fluorescence (Plate 8, figures 8, 10). These results indicate that the abundant Homotryblium specimens in the upper part of HA 66 and HA 573 are reworked.

Palaeoenvironment and palaeoclimate
The geological development in the studied area at Bad Deutsch-Altenburg reflects a marine transgression. Due to variable positions of the transgression horizons in the boreholes, the transgression can be traced from lower to higher altitude levels. The lowermost level is documented with a transgression conglomerate in borehole HA 66 which occurs at around 50 m above the actual sea level (asl). In borehole HA 540 the transgression horizon is at around 100 m above the asl (Figure 2; Table 1). When at the position of HA 540, the Mesozoic basement was flooded, there was an altitude difference of about 50 m to HA 66. However, the transgressive contact of the late Badenian is much higher in altitude farther south (e.g. at the Kirchenberg and the Pfaffenberg in Bad Deutsch-Altenburg). This much wider range can only be explained by local/regional tectonic movements as suggested  previously for other areas in the VB (Kranner et al. 2021a). Wessely (1961) studied various tectonic structures in the Hainburg Mountains and Gangl (1988Gangl ( , 1990 found good evidence for faults in the area for the planned Hainburg powerplant project. Altogether, the upper Badenian deposits in the studied boreholes were deposited in a coastal to inner shelf position, probably not deeper than 50 m (epi-to mesoneritic, sensu Gross 2002). This estimate is supported by the occurrence of larger benthic foraminifera (LBF: Borelis, Amphistegina, Planostegina) that are restricted to shallow water (Hohenegger et al. 1999;Hohenegger 2004).
The occurrence of coralline algal-incrusted breccias/conglomerates at the transgression horizon and larger foraminifera clearly indicate fully marine conditions already from the beginning of the transgression. The foraminiferal fauna and the corallinaceans occur throughout the drilled cores and point to a continuous marine environment. Only in HA 66 can a shallowing be observed in the topmost samples (HA 66/6 through HA 66/1) which may belong already to Sarmatian strata. The Sarmatian date is also proven in the samples from the top of HA 573 by ostracods (Gross 2002).
The sediment differentiation along the Mesozoic spur shows a clear dominance of sand-rich sedimentary rocks in the south-west of the spur. This sand can be considered of fluvial origin because the ostracod assemblages in these samples show a mixture of fully marine and freshwater taxa (upper part of HA 514: Danielopol et al. 1991;HA 521, HA 573: Gross 2002). Freshwater ostracods are also associated with characean gyrogonites (green algae) which clearly indicate freshwater influx (Danielopol et al. 1991). A major part of the terrigenous input is made of the clay fraction, which is transported in suspension and does not necessarily affect the in situ dinoflagellate cyst occurrences.
Lingulodinium machaerophorum, the cyst of the motile Lingulodinium polyedrum, can survive in a salinity range of 8.5 to 42 psu (Mertens et al. 2009;Zonneveld et al. 2013). Mertens et al. (2009) concluded that cysts with relatively long and stout processes are an indication of increased salinity and temperature. Such morphotypes are common (ca. 10%) in HA 66 (samples HA 66/26 through HA 66/7; Plate 2, figure 3). Nematosphaeropsis labyrinthus (Plate 5, figure 11) which indicates a salinity range of 25.8-39.4 psu (Zonneveld et al. 2013) is frequently recorded within the same interval, except the lowermost samples representing deposition at shallow water depths. In samples HA 66/27 to HA 66/16, Polysphaeridium zoharyi (Plate 5, figure 15) which indicates 28.4-39.4 psu (Zonneveld et al. 2013) is also recorded (0.5% to 2.9%). This fits well with the calculated salinity (average of 35 psu and a range between 31 and 40 psu) for the late Badenian of the VB based on foraminiferal assemblages (Kranner at al. 2021b).
Dinoflagellate taxa such as   (Dale 1996;Head 1997;Zonneveld et al. 2013;Quaijtaal et al. 2014;Sangiorgi et al. 2021). This group of taxa has been consistently recorded through the dinoflagellate cyst zones Uaq and Cpol (Figures 3-5; Jim enez- Moreno et al. 2006). In HA 66, for example, these taxa reach up to 23% (HA 66/11). Thus, a tropical to warm-temperate environment is firmly suggested for the late Badenian. This is also supported by the continuous and consistent occurrences of Labyrinthodinium truncatum (Plate 1, figures 13-14; Plate 4, figure 6), associated with Nematosphaeropsis labyrinthus (Plate 2, figure 4; Plate 5, figure 11), which indicate relatively warm conditions (Zonneveld et al. 2013;Quaijtaal et al. 2014;Schreck et al. 2017;Guler et al. 2021). Furthermore, these warm climatic conditions can also be detected from the thermophilic and mesothermic terrestrial flora, based on the occurrence of pollen of Bombax sp. (Bombacaceae) and Tsuga, and pollen of Carya, Pterocarya (Juglandaceae) and Tilia (Tiliaceae) (Plate 3) . They are indicated also by the occurrence of tropical/subtropical ostracod taxa, e.g. Heliocythere vejhonensis, Phlyctenophora arcuata and Bairdoppilata subdeltoidea (Gross 2002). It is worth mentioning that the thermophilic taxon ratio reaches about 1% in the upper part of the HA 66 borehole, for example (Cpl interval zone; late Serravallian; Sarmatian), which may indicate a shift towards a Mediterranean temperate climate . A clear climatic assignment to tropical to warmtemperate conditions is supported by the occurrence of LBF such as Borelis, Amphistegina and Planostegina. This is fully in agreement with the data of Kranner et al. (2021b), who calculated an average water temperature of ca. 18 C for the late Badenian based on foraminifera, for both surface and bottom waters. Although the late Badenian can be attributed to the Miocene Climate Transitiona cooler time interval following the warm Mid-Miocene Climatic Optimumthe overall relatively high temperatures may be explained by shallowing of the VB and a minimal depth gradient (Kranner et al. 2021a).
In addition, we noted a consistent occurrence of thermophilic taxa (e.g. Polysphaeridium zoharyi, Selenopemphix nephroides, Tectatodinium pellitum and Tuberculodinium vancampoae) in the studied boreholes, reaching 27% (average 11%) in HA 66 for example (Figures 3-5). The occurrences of taxa that are common in low latitudes such as Operculodinium israelianum (Plate 4, figure 12) and Operculodinium longispinigerum (Plate 4, figure 11) (the two species recorded within Operculodinium spp. in the range chart) (Zonneveld et al. 2013;de Vernal et al. 2020) possibly support the connection between the Central Paratethys and the Mediterranean during the Middle Miocene ( Figure 1A).
In general, the upper Badenian dinoflagellate cyst assemblages recorded in Bad Deutsch-Altenburg (BAD) and those documented by Jim enez- Moreno et al. (2006;Tengelic-2 core) are very close to the Serravallian assemblage of the Mediterranean (northern Italy) (Powell 1986;Zevenboom 1995). This consistency indicates an exchange of water masses between the Central Paratethys and the Mediterranean during the early Serravallian (late Badenian).

Conclusions
Dinoflagellate cysts have been studied, for the first time, in 83 samples from 10 cored boreholes of the Rabensburg Formation (Baden Group), Bad Deutsch-Altenburg, VB, Austria. A diversified assemblage of in situ dinoflagellate cyst taxa has been recorded. Middle Miocene key taxa including Unipontidinium aquaeductum, Labyrinthodinium truncatum, Cerebrocysta poulsenii, Habibacysta tectata, Cannosphaeropsis passio and Operculodinium? borgerholtense are recorded. Their occurrences allow correlation with the Central Paratethys dinoflagellate cyst biozonation of Jim enez- Moreno et al. (2006). Their biozones Uaq (upper part), Cpo and Cpl are defined in the wider area of this study. The dinoflagellate cysts studied herein strongly indicate a Serravallian (late Badenian-early Sarmatian) age. The suggested age is supported by calcareous nannoplankton and ostracods from the same set of samples. Additionally, the studied succession can be correlated with the DM5 (in part) and DM6 (in part) dinoflagellate cyst zones of King (2016).
Well-preserved dinoflagellate cysts reworked from the Cretaceous and Paleogene occur in high numbers or even dominate in some boreholes (HA 521, HA 573) or samples (the upper part of HA 66). These dinoflagellate cysts may have been transported to the study area due to erosion from the Gosau Group at the top of the Northern Calcareous Alps and/or the Rhenodanubian Flysch Unit and/or the Little Carpathians in Slovakia. Specimens of the genus Homotryblium, generally associated with reworked taxa ranging from the Paleocene to the Middle Miocene and showing a low degree of fluorescence, are also considered reworked.  (2006). The global and regional Miocene chronostratigraphical and biostratigraphical framework is after Gradstein et al. (2020) and Harzhauser et al. (2020).
Based on the identified dinoflagellate cyst assemblage and the associated miospores, the upper Badenian deposits in Bad Deutsch-Altenburg were deposited in a coastal to inner shelf position under tropical to warm-temperate climatic conditions with a normal salinity. The composition of the dinoflagellate cyst association is similar to those of the Serravallian association found in the Mediterranean region, which supports the interpretation of water exchanges between the Central Paratethys and the Mediterranean during the late Badenian.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
The laboratory work for this study was supported by the Commission for Stratigraphical and Palaeontological Research of the Austrian Academy of Science. The SEM investigation and photography were conducted at the Institute of Earth Sciences, University of Graz. Paul Dodsworth and Martin Head kindly provided many constructive comments on the manuscript.