Pleistocene loess deposits and mollusc assemblages in the Eastern PreAlps

In comparison to other areas in low mountain regions, the widespread occurrence and thickness of loess is impressive in the northern Vienna Forest. Due to differences in grain size, it is obvious that the loess deposits of the Hagenbach Valley deviate from those of other locations. In comparison to the results of Krems and Stillfried, the loess of the Hagenbach Valley has a pronounced maximum in the sand fraction reflecting an essential influence of the Flysch sandstone and a proximity to the source sarea. The loess of the Hagenbach Valley is specified as sediment with significant local impact due to a remarkable influence of short distance transport. Partly, the loess is of alluvial origin as it contains small pebbles and therefore it reflects cool and wet paleoenvironmental conditions. The malacological evidences coincide with the geomorphodynamic conditions. Redeposition processes cause a generally high degree of fragmentation. The malacological analyses proved 28 species of terrestrial gastropoda, with a total number of 3,283 specimens. The results indicate very humid and cool climate and a weakly expressed, slightly more favorable period is visible in one of the horizons. (Pleistozäne Lösssedimente und Molluskengesellschaften in der östlichen Voralpenzone) Kurzfassung: Die Lössablagerungen im nördlichen Wienerwald sind im Vergleich mit anderen Mittelgebirgsregionen aufgrund ihrer Mächtigkeit sehr eindrucksvoll. Charakterista in der Korngrößenverteilung zeigen deutlich, dass die Lösse im Hagenbachtal sich von denen anderer Lösssgebiete unterscheiden. Ein Vergleich mit Lössprofilen in Krems und Stillfried hat ergeben, dass der Löss im Hagenbachtal einen erhöhten Sandanteil aufweist und damit den Einfluss der Flysch-Sandsteine widerspiegelt. Das spricht für einen lokalen Sedimenteintrag und kurze äolische Transportstrecken. Zudem wurde der Löss unter kühl-humiden Paläoklimabedingungen zum Teil als Schwemmlöss abgelagert. Die malakologischen Ergebnisse stimmen mit den geomorphodynamischen Bedingungen überein. Die Umlagerungsprozesse haben zu einer intensiven Fragmentierung der Schalenreste geführt. Die malakologischen Untersuchungen belegen insgesamt 28 unterschiedliche Arten von terrestrischen Gastropoden mit 3283 Individuen. Die paläoökologische Auswertung spricht für sehr humide, kühle Klimabedingungen mit einer schwach ausgeprägten, klimatisch etwas günstigeren Phase.


Introduction
In Lower Austria, eolian sediments are widespread (Fig. 1) and form thick and well developed loess sequences.These deposits of Lower Austria have been studied for several decades.Detailed descriptions and sedimentological data exist for many loess/paleosol sequences situated in the basins of Vienna and the Alpine foreland (Fink 1976;Kohl 1976;Haesaerts et al. 1996, Terhorst 2007, Peticzka et al. 2010).Geochronological framework for these deposits are most commonly based on radiocarbon (e.g.Haesearts et al. 1996) and luminescence dating (Wallner et al. 1990;Noll et al. 1994, Zöller et al. 1994;Thiel et al. 2011aThiel et al. , 2011b, c), c).Both dating methods are suitable, at least for the time range of up to 40 ka for radiocarbon (Hajdas 2008) and to 300 ka for luminescence (Thiel et al. 2011a), for the sediments under question and have been proven to be reliable approaches.
While detailed descriptions, sedimentological and chronological data exist for many loess/paleosol sequences in the Vienna basin and the Alpine foreland (e.g.Fink 1956Fink , 1976Fink , 1978;;Kukla 1975, Döppes & Rabeder 1997, Terhorst 2007), loess sediments in the low mountain areas have not been studied up to present.The map of loess distribution for Lower Austria clearly shows that loess covers the northern margins of the Vienna Forest (Fig. 1).There, loess deposits are frequent and occur in altitudes above 300 m a.s.l. and reach remarkable thicknesses of 4 m in steep slope positions (Damm & Terhorst 2010).Inside the Hagenbach Valley, loess sequences are exposed by landslide processes.
In general, the malacological analyses of a complete pedosedimentary sequence of the Hagenbach Valley are promising for insights into the faunal and paleoenvironmental development during the last Middle to Late Glacial period.
Two characteristic sequences and first dating results combining both radiocarbon and luminescence dating are presented with a focus on mollusc analyses in this study.The results are compared to characteristic loess sequences of Lower Austria.

Study area
The Hagenbach Valley (Hagenbachklamm) is situated approximately 16 km to the NW of the city of Vienna (Fig. 2).Climate data for the period between 1971 and 2000 show mean annual temperature of 9.2°C and mean annual precipitation of 741.5 mm (Zamg 2002).The valley is part of the northern margin of the Vienna Forest that represents a rolling landscape of the Central European low mountain regions.The flat-topped mountains of the Vienna Forest with altitudes between 300 and 500 m a.s.l. are deeply incised by valleys.
The study area belongs to the Rhenodanubian Flysch Zone, which is W-E oriented and consists of various layers of (calcareous) sandstones, marly shales, calcareous marls  (Terhorst et al. 2009).
and clay schists.The formations inside the Hagenbach Valley belong to the Altlengbach and the Greifenstein beds.They are dominated by calcareous quartzitic sandstones, which are subdivided by stratified marls and clays (Wessely 2006).
The slopes of the Hagenbach Valley are characterized by various types of Quaternary sediments.The bedrock is covered by loess and Pleistocene periglacial cover beds.
First results concerning Quaternary slope deposits have been published by Damm & Terhorst (2010) and Terhorst et al. (2009).However, the chronology of loess deposits and periglacial layers is not known up to present in the Vienna Forest.

Field survey and sedimentological analyses
The description and determination of soil horizons were conducted according to the German Field Book for Soil Survey (AD-HOC-Arbeitsgruppe Boden 2005).The field description was adapted to the World Reference Base for soil resources (IUSS Working Group WRB, 2006).During the field survey information on texture, bedding, content of carbonate, and colour of soils and sediments was gathered.The differentiation of periglacial cover beds was based on the German Field Book for Soil Survey (AD-HOC-Arbeitsgruppe Boden 2005).The main criteria were given by sedimentological features such as the occurrence of specific silt contents and rock fragments, which are essential indicators for the differentiation of periglacial cover beds (Semmel & Terhorst 2010).Samples were taken by horizons.Two sequences (A and B) are presented in detail (Fig. 3).
Laboratory analyses included grain size determination according to the combined sieve-and pipette-analysis according to Köhn (Schlichting et al. 1995).The calcium carbonate contents (CaCO 3 ) were measured gasvolumetrically with the Scheibler instrument.

Radiocarbon and luminescence dating
For age determination both radiocarbon and luminescence dating have been applied.Whereas radiocarbon dating estimates the time elapsed since the death of an organism (e.g.Hajdas 2008), luminescence dating techniques determine the time passed since the last exposure of mineral grains to sunlight, and thus enable to constrain the time of deposition (Aitken 1998).
The radiocarbon age (Hv25871) was obtained from charcoal using gas proportional counters.The sample was taken in Sequence B at a depth of 3.9 m in loessic sediments.The calculation of the radiocarbon age is based on the radioactive half-life after Libby, i.e. 5568 years.The age was fitted to the international radiocarbon timescale by NBS oxalic acid standard; the data is δ 13 C corrected.All uncertainties of the chemical and technical treatment of the sample are included in the expressed age error (2-sigma standard deviation).
Two luminescence samples were taken in the loess of Sequence A (Fig. 3) by hammering metal tubes into the freshly cleaned profile; sample 1217 was taken at a depth of 2.4 m and sample 1218 at 3.1 m, respectively.Luminescence measurements were made with an automated Risø TL/OSL read-ers (DA-15;Bøtter-Jensen et al. 2003) equipped with a calibrated 90 Sr/ 90 Y beta source.The feldspar signal of the polymineral samples was stimulated with infra-red (IR) light diodes emitting at 870 nm, and the luminescence was detected in the blue-violet region through a Schott BG39/Corning 7-59 filter combination.
For equivalent dose (D e ) determination, a single-aliquot regenerative (SAR; Murray & Wintle 2000) protocol with a preheat of 250°C (60 s) and IR stimulation at 50°C for 100 s was applied.An IR illumination at 280°C (100 s) was inserted at the end of each SAR cycle (Murray & Wintle 2003).The laboratory fading rate was measured as the IRSL signal decrease over time using artificially irradiated ali-quots (Lamothe et al. 2003); this is expressed in terms of the percentage decrease of signal intensity per decade (the g-value; Aitken 1985, Appendix F).
Material for dosimetry measurements (as part of the luminescence dating) were taken from immediately around the luminescence samples.After drying and homogenizing, 50 g of each sample were packed in N-type beakers, which were stored at least for one month to ensure equilibrium between radon and its daughter nuclides.The concentrations of U, Th and K were determined by high-resolution gamma-ray spectrometry.The dose rates were derived using the conversion factors of Adamiec & Aitken (1998).Calculation of the cosmic dose rate is based on Prescott & Hutton (1994).A water content of 20 ± 5% was estimated, allowing for changes in water content throughout time, and a mean a-value of 0.08 ± 0.01 was used.

Mollusc analyses
Soaked samples (1000g) were treated by wet sieving to 2 mm-630 µm and 200 µm mesh, dried and selected.The 200 µm fraction contained thousands of minute fragments, so only the significant ones (apices, apertures) were picked out for calculating the individual numbers (as proposed by Ložek 1964).

Field survey and sedimentological analyses
Two sequences at a distance of 300 m were studied.The altitude of both profiles ranges between 260 m and 310 m a.s.l.. Sequence A is situated on the western slope of the Hagenbach Valley, sequence B on the eastern slope.Both loess deposits are situated on top of basal periglacial cover bed (Flysch debris) superimposed on Flysch.In general, a Luvisol is developed on top of the loess.

Sequence A
The sequence starts with a Luvisol consisting of a thin A horizon, an E horizon, partly followed by a transitional EBt horizon, Bt horizon and, as parent material, loess that forms the C horizon (Fig. 3).Soil horizons are free of calcium carbonate and show a loose structure interspersed with pores of different sizes.
Below the 1 m thick Luvisol, a loess deposit with a thick-ness of 1.4 m is recorded (Fig. 4).The loess shows a pale yellow colour, involves partly small rounded sandstone fragments as well as secondary carbonates.On average, the silt content in the loess deposits is higher than in related soil horizons.Inside the silt fraction, coarse silt constitutes with values from 43.1% to 48.0% makes up the most important fraction (Table 1: 6-4 to 6-6), whereas the fine and middle silt contents are comparatively small.The loess has low clay contents with a maximum value of 7.8%.The carbonate content of the unweathered loess deposits ranges from 34.9% to 36.2% (Table 1).The basal part of the loess (Fig. 3) is characterized by the occurrence of hydromorphic features such as thin brownish and grayish Fe-bands.Below the eolian deposits, a periglacial cover bed is present, which consists of clays, marls, and debris (Fig. 3).The clay content amounts to 31.8%, the silt content to 53.8% and the sand fraction is lower than in the previously described layers.The carbonate content is 31.9%.The consistence and distribution inside the silt classes deviates from those of the loess.In Tab.1: Grain size and calcium carbonate content of sequences A and B, modified after Terhorst et al. (2009).
Tab. 1: Ergebnisse der Korngrößen-und Kalziumkarbonatanalysen der Profile A und B, geändert nach Terhorst et al. (2009).Samples that originate from decomposed sandstone are certainly dominated by the sand fraction, which varies be-tween 61.2% and 79.7% (Table 1: 8,10,11).It is two to three times higher than in the loess and marls.The maximum contents are in the middle and fine sand fraction, whereas the coarse fraction is of minor importance.Silt contents range between 17.4% and 31.7% with maxima in coarse silt.

Sequence B
The second example of a characteristic pedosedimentary sequence (Fig. 3, B) is situated on the E slope of the Hagenbach Valley, in about 300 m linear distance from sequence A. The upper part of the sequence is formed by a Bt horizon of an eroded Luvisol, disturbed by anthropogenic impact and redeposition of colluvial material.The eroded Luvisol has formed in loess, which has a remarkable thickness of 3.5 m.The upper part of the pale yellow loess shows low clay con- tents between 6.7% and 10.2% (Table 1: 17-3 to 17-6).The silt content dominates the grain size distribution and reaches values up to 75.0%.The sand content ranges from 14.8% to 24.1%.The CaCO 3 contents are on average about 23%, and only the parts of the sequence, which have secondary carbonates, contain more CaCO 3 (33.9%,Table 1: 17-3).In the basal parts of the loess hydromorphic features are present (Fig. 3).The horizon C2 is interspersed with small mottles and bands of rusty iron alternating with grayish mottles of reduced iron.The lowermost horizon, C3 (17-12) consists of sandstone debris and sand, which is exclusively made of Flysch sediments.This debris layer corresponds to the Pleistocene basal periglacial cover bed of sequence A. It consists of local bedrock fragments with stone fragments orientated related to the slope inclination.

Comparison of loess sites in Lower Austria
The characteristics of the loess deposits of the Hagenbach Valley differ significantly from other loess deposits in Lower Austria (Table 2).The loess sediments of the Vienna Forest show a higher percentage in the sand fraction and lower clay content.However, the silt fraction is comparable with those of the loess sequence in Stillfried (Table 2: 20-1 and 20-2).There is a significant analogy in the coarse silt fraction for all study sites.On the contrary, the loess of Krems contains, with values up to 83.5%, significantly more silt in total as well as in the coarse silt fraction (Table 2: 59.1% to 62.3%).In general, the portion of middle silt is higher in the loess regions and reduced in the Vienna Forest study sites.

Dating
The charcoal sampled in Sequence B was dated to 31,330-27,860 years cal.BC (Table 3; calibration after Reimer et al. 2009).
For the luminescence samples the dose recovery tests for all aliquots were within 10% of unity, showing the applicability of the protocol to our samples.Recycling ratios, which show the ability of a measurement protocol to reproducibly measure the response to a laboratory dose given after repeated heating of the sample, are close to unity (1.01 ± 0.01; n=30), and are hence very satisfactory.Furthermore, recuperation for all aliquots was measured to be <5%.The dose rates of 2.4 ± 0.2 Gy/ka and 2.7 ± 0.2 Gy/ka, respectively, are in the range known from Lower Austrian loess deposits (e.g.Thiel et al. 2011a, b, c;Zöller et al. 1994) (Table 3).
The fading uncorrected age for sample 1217 is 20.7 ± 3.7 ka, and sample 1218 was dated to 22.2 ± 1.2 ka.However, for both samples a fading rate of >2%/decade was measured, implying that these uncorrected ages are underestimating the depositional age.Fading correction had to be applied; this is based on Equation 4 of Huntley & Lamothe (2001).The fading corrected ages of 25.4 ± 4.8 ka (sample 1217) and 26.9 ± 2.4 ka (sample 1218) are considered more reliable.

Mollusc analyses / Coenological analysis
Mollusc samples (Table 4 and Table 5) originate from the lower part of the loess sequence B (Fig. 3).They have been taken in a vertical order from the sample position S1 (top) to S3 (base).In sample S2 there is a remarkable decrease in total number of individuals (S1: 1344, S2: 843).However, in sample S3 the number of species is declining (S2: 25, S3: 21), whereas the number of individuals increases noticeably (1106) (Fig. 5 and Fig. 6).
Singular shell fragments of highly demanding woodland species appear only in S1 and S2: Ena montana and Ruthenica filograna in both of them, cf.Cochlodina laminata only in S1, Aegopinella cf.nitens only in S2.Semilimax kotulae has been found all over the profile.This species occurs also in thanatocoenoses (=mollusc assemblages) pointing to moderate or cold climate.The first-mentioned species are allochthoneous in samples S1 and S2.
Tab. 3: Radiocarbon age and IRSL ages (fading uncorrected and fading corrected) and its corresponding information (dose rates, equivalent doses, fading rates).
living in present day malacocoenoses in the neighbourhood of the excavation area, does not occur in the studied samples.This large form seems to be restricted to some regions of the Vienna Forest, at colline to low-mountain altitudes with favourable climate.
-Contaminations and redeposition of material is expressed in the thanatocoenoses of horizons S1 and S2, at a depth of 2.8-3.7m,where single minute, corroded shell fragments of highly demanding woodland species (Ena montana, Cochlodina laminata, Ruthenica filograna, Aegopinella nitens) appear.These species certainly occurred in the study area during the middle Holocene climatic optimum but also live there today.

Discussion
In comparison to central European low mountain regions, the widespread occurrence and thickness of loess is impressive in the northern Vienna Forest.It is clearly distinguishable from other sediments such as periglacial cover beds and decomposed Flysch bedrock.Furthermore, loess deposits of the characteristic loess regions of Lower Austria can be clearly distinguished.Due to differences in grain size, it is obvious that the loess deposit of the Hagenbach Valley deviates from those of the other locations.The loess profiles of Krems and Stillfried show the characteristic grain size distribution of periglacial eolian sediments with an enhanced input of grains originating from a far distance transport (Pécsi & Richter 1996).In comparison to the results of Krems and Stillfried, the loess of the Hagenbach Valley has a pronounced maximum in the sand fraction, thus reflecting an essential influence of the Flysch sandstone.However, short distance eolian transport is significantly higher in the area under study than in the main loess regions of Lower Austria.Therefore, the loess of the Hagenbach Valley might be specified as sediment with a considerable local impact.Partly, the loess has been redeposited as it contains small pebbles and therefore it reflects cool and wet paleoenvironmental conditions.The results of the datings might be interpreted in this way as well.
The malacological evidences coincide with geomorphodynamic processes like erosion and sedimentation as assumed by Damm & Terhorst (2007, 2010), Damm et al. (2008), andTerhorst et al. (2009).Furthermore, redeposition of loess is suggested by the generally high degree of fragmentation of molluscs shells, especially in sample S1.The periglacial origin of the sediments is reflected by the structure of the thanatocoenoses, which is distinctly contradictory to the rich and varied molluscan assemblage of today.
Malacological evidence points to comparatively least favorable paleoenvironmental factors during the deposition of the uppermost part of the basal loess (Fig. 3, S1).The conditions reflected in the lowermost sample S3 are equally infavorable, but of mitigated character, whereas the sedimentation period of S2 appears slightly moderated.
Following this, the sedimentation period is marked by very humid and cool climate, which does not coincide with the occurrence of very manifold vegetation: -mainly open to semi-open, probably tundra-like areas -predominantly herbaceous vegetation and undemanding trees or brushwood -not very extensive damp habitats -rocks overgrown with mosses and lichens, shady or exposed Although the three thanatocoenoses reflect rather cold climate, extreme conditions are nowhere recognizable, indicated by the low percentage rates of the typical loess species Pupilla, Vallonia tenuilabris and Columella columella, the absence of cold-tolerant Vertigo's, the obviously not very differentiated structure of the thanatocoenoses, the mass development of the Trochulus species, and the presence of the Neostyriaca corynodes subspecies austroloessica.According to Klemm (1969), the latter is associated with environmental conditions similar to those in the Eastern Alps from about 1,400 m upwards to about 2,300 m asl.Klemm (1969) has listed the occurrence in different loess deposits in the Danube Valley of Upper and Lower Austria (c.f.Frank 2006).Furthermore, the absence of large Orcula dolium specimens is worth mentioning in this context, since the study area is situated within the "infima zone" sensu Zimmermann (1932) and Klemm (1974).Local populations of such "fattened" forms would rather point to optimum living conditions.Therefore, one may conclude that this large subspecies that occur nowadays in the southern and eastern Vienna Forest have developed subsequently during the Postglacial warming.
In general, there is good agreement between malacological results and the obtained dating results.The slightly older age estimate for the charcoal compared to the luminescence age could be either due to problems in fading correction as observed by others (e.g.Wallinga et al. 2007), i.e. even after fading correction the luminescence ages are underestimates, or due to reworking and correspondingly later incorporation of the charcoal into the hydromorphic unit C2.However, within two standard deviations, there is good agreement of the obtained ages.
Both, 14 C-and IRSL-dating indicate deposition of loess and loess-like sediments prior to the Last Glacial Maximum.This is consistent with formerly presented luminescence ages of ~25-30 ka (Thiel et al., 2011a, b, c) for loess deposits in the Kremser Feld.From these studies it seems evident that the phase of loess deposition during the Last Glacial Maximum is not well documented in Lower Austria, whereas the period between ~25-30 ka can be found in several outcrops, for instance in Stillfried (cf. Frank 1997), in the archaeological sites Willendorf (Haesaerts et al. 1996), andKrems-Wachtberg (Neugebauer-Maresch 2008).Hence, the observations made for the Hagenbach Valley fit to results obtained from other loess sequences of Lower Austria.However, there is still a lack of high resolution dating in relevant profiles, also for the Hagenbach Valley.Moreover, investigations have focussed on the relevant archaeological horizons of a sequence (Haesaerts et al. 1996), and younger loess might not have been observed due to sampling and dating strategies.It is evident that loess/paleosol sequences of the LGM are ubiquitous in Upper Austria (Terhorst et al. 2002;Starnberger et al. 2009).
case of the periglacial cover bed the contents of fine, middle, and coarse silt are more or less homogeneously distributed.The stratum is densely bedded, has an intensively undulated boundary and the debris is composed of Flysch bedrock.The latter meets the diagnostic pre-requisites for the classification as a Pleistocene basal periglacial cover bed.Samples were taken in the surface formations of the Flysch sandstones to compare loess deposits with the local bedrock (Table 1: 8, 10, 11).Weathered bedrock varies in carbonate content between 2.8% and 6.0%, whereas the carbonate content of less weathered Flysch bedrock can reach values up to 41.7% and reflects the potential carbonate level of the bedrock (Table 1: 11). the Tab. 2: Grain size and calcium carbonate content of sequences A and B in comparison to analyses of loess profiles ofKrems and Stillfried in Lower Austria ). -The indication of humidity due to the predominance of Trochulus hispidus and Trochulus suberectus (together: S1: 61.7%, S2: 47.9%, S3: 55.0%).-Elements inhabiting areas of medium humidity, with herbaceous vegetation, hedges or clumps of trees are more distinctly represented in S2 (12%) and S3 (12.2%) than in S1 (3.9%).-The percentage rates of Pupilla species (mostly: sterrii, fur- Sandberger 1875), 2 most of them: muscorum, alpicola densegyrata, 3 most of them: alpicola densegyrata, loessica 4 muscorum, alpicola densegyrata, loessica, 5 fragments of body whorl, 6 atypical fragment, 7 cf.fragment of aperture, 8 cf.fragment of body whorl, 9 atypical fragment, 10 cf.two of them: fragments of body whorl, 11 hispidus + suberectus, 12 like in S1, 13 like in S1 and S2