New evidence for persistent impact-generated hydrothermal activity in the Miocene Ries impact structure, Germany



The extent of impact-generated hydrothermal activity in the 24 km sized Ries impact structure has been controversially discussed. To date, mineralogical and isotopic investigations point to a restriction of hydrothermal activity to the impact-melt bearing breccias, specifically the crater-fill suevite. Here, we present new petrographic, geochemical, and isotopic data of postimpact carbonate deposits, which indicate a hydrothermal activity more extended than previously assumed. Specifically, carbonates of the Erbisberg, a spring mound located upon the inner crystalline ring of the crater, show travertine facies types not seen in any of the previously investigated sublacustrine soda lake spring mounds of the Ries basin. In particular, the streamer carbonates, which result from the encrustation of microbial filaments in subaerial spring effluents between 60 and 70 °C, are characteristic of a hydrothermal origin. While much of the primary geochemical and isotopic signatures in the mound carbonates have been obliterated by diagenesis, a postimpact calcite vein from brecciated gneiss of the subsurface crater floor revealed a flat rare earth element pattern with a clear positive Eu anomaly, indicating a hydrothermal fluid convection in the crater basement. Finally, the strontium isotope stratigraphic correlation of the travertine mound with the crater basin succession suggests a hydrothermal activity for about 250,000 yr after the impact, which would be much longer than previously assumed.


Impact-generated hydrothermal activity has been reported from a number of terrestrial impact craters (Newsom 1980; Naumov 2002; Pirajno 2009a; Osinski et al. 2013; and references therein), from large impact structures 200–250 km in diameter, such as the Sudbury structure (e.g., Ames et al. 1998, 2006), to smaller craters of only 1.8 km in diameter (e.g., Lonar crater; Newsom and Hagerty 2003). The thermal energy driving these hydrothermal systems is derived from (1) the remnant heat of impact melts or melt-bearing rocks and (2) the hot sections of the central uplift with rocks from initially greater depths (Pirajno 2009a). So far, no evidence has been provided that impact-generated magmatism could create hydrothermal fluid convection on (the post-Hadean) Earth independent from impact melts. Indeed, impact-generated hydrothermal systems differ from magmatism-related systems by the fact that a single high-temperature pulse is followed by a longer duration tail of the cooling system (e.g., Nelson et al. 2012).

The evidence of hydrothermal activity is commonly derived from subsurface mineral veins, altered impactites, and altered postimpact sediments (e.g., Newsom et al. 1986; Osinski et al. 2001, 2013; Hecht et al. 2004; Zürcher and Kring 2004; Escobar-Sanchez and Urrutia-Fucugauchi 2010). In addition, mineral deposits of subaerial thermal springs have been reported in few cases, such as the Haughton impact structure, Canada (Osinski et al. 2001). Such hydrothermal precipitates potentially fossilize microbial communities and are of special interest as analog sites for the search of life on other planetary bodies (e.g., Farmer and Des Marais 1999; Pirajno 2009b; Osinski et al. 2013).

With respect to the 24 km sized Ries impact crater, one of the best investigated impact structures on Earth (Pohl et al. 1977; von Engelhardt 1990, 1997; Stöffler et al. 2013), the intensity of hydrothermal activity has been controversial to date: Up-to-date, mineralogical investigations point to a restriction of hydrothermal activity to the impact-melt bearing breccia, specifically the crater-fill suevite (Osinski 2005). In addition, fluid channels and degassing pipes have been reported from the outer (“fallout”) suevite (Newsom et al. 1986). On the other hand, the formation of hydrous clay minerals in the outer (“fallout”) suevite, previously assigned to hydrothermal alteration (Newsom et al. 1986), could be explained alternatively by late meteoric diagenesis (Muttik et al. 2008, 2011). No hydrothermal spring deposits or veins from surface outcrops have been reported so far from the Ries, whereas cool-water (i.e., waters with ambient temperatures) spring mounds are well known from the Ries basin (Bolten 1977; Arp 1995; Pache et al. 2001).

Here, we present the discovery of a small spring mound with travertines that differs in fabric and stratigraphic position from the previously described mound carbonates (Bolten 1977; Arp 1995; Pache et al. 2001). Petrographic and geochemical data are provided to discuss their hydrothermal origin and the extent of hydrothermal fluid convection outside the crater-fill suevite. Furthermore, we discuss the potential duration of the hydrothermal system and its relation to impact melts and deep fracturing in the Ries basin, based on these spring-related surface carbonate deposits (travertine mound) and subsurface calcite veins of the crater basement.

Geological Setting and Location

The impact crater Nördlinger Ries is located in southern Germany, 110 km NW of Munich. Its circular morphological depression separates the limestone karst plateaus of the Franconian Alb in the east and the Swabian Alb in the west. The crater formed by the impact of an approximately 1 km sized asteroid of unknown composition (Shoemaker and Chao 1961; Schmidt and Pernicka 1994), which penetrated through the 580 m thick sedimentary cover into the crystalline rocks of the Variscian basement (for review, see Pohl et al. 1977; von Engelhardt 1990, 1997). 40Ar-39Ar data indicate an age of 15.1 ± 0.1 Ma for the impact event (Staudacher et al. 1982), which has recently been confirmed by new 40Ar-39Ar data from suevite impact glasses from Öttingen (14.89 ± 0.1 Ma; Abdul Aziz et al. 2008) and by dating of volcanic ashes underlaying (15.00 ± 0.02 Ma) and overlaying (14.93 ± 0.01 Ma) a distal ejecta bed in the North Alpine Molasse basin (Rocholl et al. 2011). The recalculation of standards used in other 40Ar-39Ar datings of moldavites and suevites, and the revision of the 40K decay constant, partly explains the discrepancy with published younger ages of the Ries impact (Buchner et al. 2013).

After the formation of a transient cavity of 2.0–2.8 km depth (Stöffler 1977; von Engelhardt et al. 1995), gravitational collapse resulted in a complex basin 24 km in diameter and 600 m deep, including a central crater, an inner ring, and a megablock zone (Reich and Horrix 1955; Pohl et al. 1977; Stöffler 1977). There is no central peak evident, but geoelectric data point to the existence of a second peak ring of basement rocks within the central crater (Pohl et al. 1977).

Ejecta layers comprise (1) the Bunte Breccia consisting of a mélange of sedimentary and crystalline rocks and (2) the “outer suevite,” an impact-melt bearing crystalline rock breccia with minor sedimentary components. According to Pohl et al. (1977), Stöffler et al. (1977, 2013), and Artemieva et al. (2013), the collapse of the vapor plume and ground surging led to the formation of the “crater-fill suevite” within the crater. Possibly due to the high volatile content of the target rocks, no coherent impact melt sheet formed and impact melt agglomerates are restricted to small patchy areas (e.g., von Engelhardt 1997; Pohl et al. 2010; Reimold et al. 2011; Artemieva et al. 2013; Stöffler et al. 2013). As an alternative model, Osinski et al. (2008) considers the groundmass of the suevite, i.e., calcite, silicate glass, and clay minerals, as a series of impact-generated melts, which crystallized upon cooling. Hence, the crater suevite may have been emplaced as a melt sheet and the outer suevite could reflect an outward flow of impact-melted material during uplift movements in the central crater (Osinski et al. 2004, 2008).

Crystalline rocks of the inner ring are exposed in a number of hills in a discontinuous, almost circular, arrangement. Most of these hills are capped by algal reefs or cool-water spring mounds, such as the Wallerstein (Bolten 1977; Bolten and Gall 1978; Pache et al. 2001). In the southern Ries, between Nördlingen and Enkingen, where fine-grained siliciclastic crater lake sediments conceal the structural units below, a small isolated occurrence of spring mound carbonates indicates the position of a submerged section of the inner ring. This carbonate mound, the Erbisberg (Fig. 1), surpasses the surrounding plain of the crater by only 5–10 m. The Erbisberg limestones and their peculiar fabrics were first mentioned by Nathan (1926) and Reis (1926), however, without recognizing their spring-related nature.

Figure 1.

Geological overview of the Ries impact crater and location of the Erbisberg spring mound (hill top 436.9 m asl, topographic map 1: 25000 sheet 7129 Deiningen R: 43 90 855, H: 54 11 475).

In June 2006, the construction of a biogas plant was associated with the excavation of a pipeline trench (Fig. 2). This trench exposed marginal parts of the Erbisberg spring mound and recovered carbonates with subhorizontal fibrous fabrics resembling those of present-day hydrothermal travertines. This facies type was previously unknown from the Ries basin and gave rise to more detailed investigations.

Figure 2.

Overlay of field image and sketch of the Erbisberg pipeline trench showing the macroscopic layering of the travertines in a tangential section of the mound slope. Numbers 1–7 indicate position of samples. Filament tube framestones (sample 8) were found 10 m farther left (north) to this mosaic photograph.

Material and Methods

Petrographic investigations were carried out on a Zeiss Axioplan microscope (Carl Zeiss MicroImaging GmbH, Jena, Germany), using 13 large thin sections (7 × 10 cm, 80 μm thick). One half of each thin section was stained with alizarin red S to distinguish dolomite from calcite and aragonite (Friedman 1959).

The mineralogical composition of the samples was analyzed by powder X-ray diffraction using a Philips X'Pert MPD PW 3040 diffractometer, equipped with a PW 3050/10 goniometer operating at 40 kV and 30 mA with CuKα radiation. The range 20–59.5°2θ was scanned with a step width of 0.02°2θ. The counting time was 20 s per step. Mineral identification was carried out using the X'Pert HighScore software (PANalytical).

The elemental composition of the samples was determined by X-ray fluorescence analysis (PANalytical Axios Advanced XRF spectrometer fitted with a 4 kW Rh anode SSTmAx X-ray tube) on glass fusion discs. Samples with high S and Corg contents were previously heated at 650 °C for 3 h in a furnace. For the analyses of major elements, the analytical precision is better than 1–2%. For trace element determinations, the standard deviations of consecutive analyses are in the range of 2–5% relative at the level of 20–30 ppm. Detection limits vary from 5 to 0.3 ppm for the measured elements.

Contents of organic (Corg) and carbonate carbon (Ccarb) were determined using an Euro EA CNS-analyzer (HEKAtech GmbH, Germany). Organic carbon (Corg) was measured after decarbonization with 2 N HCl. Carbonate carbon (Ccarb) was calculated as the difference of directly determined total carbon (Ctot) and Corg.

Samples for carbon and oxygen stable isotope measurements have been obtained under a binocular from cut core slabs and hand specimen using a steel needle to sample separate textures. Carbonate powders were reacted with 100% phosphoric acid (density 1.95) at 70 °C using a Thermo Kiel IV carbonate preparation line connected to a Finnigan Delta plus mass spectrometer. All values are reported in per mil relative to V-PDB by assigning a δ13C value of +1.95‰ and a δ18O value of −2.20‰ to NBS 19. Oxygen isotopic compositions of dolomite were corrected using the fractionation factor given by Rosenbaum and Sheppard (1986), i.e., 1.00993 at 70 °C. Reproducibility was checked by replicate analysis of laboratory standards and is better than ±0.05‰ (1σ).

Twenty-five carbonate fractions of carbonate and calcareous siliciclastic rock samples were analyzed for 87Sr/86Sr isotopic composition. All analyses were carried out on a Thermo-Finnigan Triton© thermal ionization mass spectrometer (TIMS). Prior to digestion, all samples were mixed with a tracer solution enriched in 87Rb-84Sr. Concentrations were calculated using the isotope dilution technique. Rb and Sr were separated from one single rock digest using standard cation-exchange procedures. Reproducibilities for NBS SRM 987 (n = 5) are 0.71025 ± 8 and 0.05650 ± 4 for 87Sr/86Sr and 84Sr/86Sr. The analytical mass bias was corrected with 88Sr/86Sr of 0.1194 using an exponential law. Analytical mass bias correction for Rb measurements was achieved via repeated analyses of NBS SRM 984 yielding a 85Rb/87Rbraw of 2.6013 ± 2 (n = 5) resulting in an exponential mass bias of 2.45‰ amu−1. All sample measurements were performed under the same conditions and corrected with an exponential mass bias derived from the standard measurements.

The rare earth element (REE) compositions of whole rock powders (approximately 100 mg, <0.63 μm grain size) were measured by inductively coupled plasma mass spectrometry (ICP-MS) model DRC II of Perkin Elmer© (Sciex, Canada). Rock powders were dissolved in a mixture of HNO3-HClO4-HF and dissolved in 2% HNO3 by a dilution of 1:1000 according to the procedure of PicoTrace© (Bovenden, Germany). Internal standardization by addition of constant concentration levels of Ge, Rh, In, and Re to any sample and standard was performed to minimize standard deviation as a measure of precision (<2% RSD [2σ]). Accuracy was better than 10% RSD (2σ) relative to standard reference material (GJA-2 and KK).

Laser ablation (193 nm Excimer laser, COMPEX 110 by Lambda Physik©, Göttingen, Germany) was adapted to the ICP-MS to analyze small portions (120 μm diameter, 300 μm depth) of sample material in situ for the REE composition. As internal standard 43Ca and for calibration of the measured intensities the reference material NBS 610 (NIST, USA) was used.


Facies Types, Their Mineralogical and Elemental Composition

A partial section of the Erbisberg mound has been temporarily exposed by a 21 m long and 1 m deep pipe trench. The trench provided a tangential section of the mound slope at its SE foothill showing layered travertine facies types (Fig. 2). In addition, a small abandoned quarry 1 m deep at the hill top exposes similar facies types plus travertines with subhorizontal fibrous structure and calcified gas bubbles (“Lochblasenkalk”; Reis 1926). On the basis of thin sections, five facies types could be identified.

Filament Tube Framestone

This facies type is composed of a highly porous framework of subhorizontal, slightly undulating carbonate tubes of 80–100 μm diameter. In cross section, the tubes are ovoid and vertically compressed to irregular (Figs. 3A and 3B). No branching has been observed. The tube walls are calcitic, multilayered, and brownish, enclosed by inclusion-rich fibrous calcite, which locally grades into fan-like to dendroid arrays of calcite crystals with sweeping extinction (“calcite shrubs showing undulous extinction”: Rodríguez-Berriguete et al. 2012; “non-crystallographic calcite dendrites”: Jones and Renaut 1995). At the center of the dendroid calcites, patches of chalcedony occur locally (Fig. 3C). Later cement phases are skalenoedric calcite spar and inclusion-rich fibrous calcite shrubs (Fig. 3C). No fecal pellets and no other microfossils, such as ostracodes or hydrobiid gastropods, have been observed in this facies type. XRD and XRF analyses (Table 1) indicate that this facies type is largely composed of calcite, with traces of quartz, dolomite, and ankerite. Mg and Sr concentrations are low. Notable are Si concentrations of 4 wt% and increased Mn contents.

Table 1. XRD, XRF, and Corg-Ccarb data of carbonates of the Erbisberg spring mound
SampleLithfacies typeMineralogySiO2TiO2Al2O3MnOMgOCaONa2OK2OP2O5Fe2O3tSSrCaCO3CaMg(CO3)2COrgCCarbDolomiteCalcite
(XRD)|←   (wt%)   →|(ppm)(ppm)(wt%)a(wt%)a(wt%)(wt%)(mole%)(mole%)
  1. n.d. = not detected; n.a. = not analyzed; XRD = X-ray diffraction; XRF = X-ray fluorescence.

  2. a

    Calculated on basis of CaO and MgO values.

Erb 8Filament tube framestonecc, (qz),(dol)4.07n.d.1.10.1380.550.
Erb AFilament tube framestonecc, (qz), (dol), (ank)0.08n.d.0.170.0171.0755.
Erb 1Crystalline crust travertinecc, (dol)1.52n.d.0.870.023.8550.
Erb 2Crystalline crust travertinecc, dol0.60n.d.0.80.0334.
Erb 7Vesicular-clotted stromatolitecc, dol0.07n.d.0.580.0388.
Erb 5Vesicular-clotted stromatolitecc, dol0.44n.d.0.690.0238.744.
Erb 3Vesicular-clotted stromatolitecc, doln.d.n.d.0.420.0418.944.
Erb 4Vesicular-clotted stromatolitedol, ccn.d.n.d.0.430.09813.537.
Erb 6Nonskeletal stromatolite with Artemia fecal pelletsdol, ccn.d.n.d.0.410.03415.
Erb CCoated bubble travertineccn.d.n.d.
Figure 3.

Travertine facies types of the Erbisberg. A) Filament tube framestone showing cross sections of tubes with fibrous calcite encrustation. Sample Erb 8. B) Filament tube framestone in longitudinal section. Sample Erb 8. C) Filament tube with fibrous calcite encrustation. Note patchy chalcedony in center of dendroid calcite (arrow). Later cements comprise skalenoedric calcite spar and inclusion-rich fibrous calcite shrubs. Sample Erb B. D) Crystalline crust travertine with angular crystal traces causing a serrated lamination. Sample Erb 1. E) Vesicular-clotted stromatolite with coarse lamination. Sample Erb 4. F) Fecal pellet-filled pockets in vesicular-clotted stromatolite. Sample Erb 6. G) Coated bubble travertine. Sample Erb C.

Crystalline Crust Travertine

Guo and Riding (1998, p. 166): This type of stromatolite is formed by an alternation of microspar-replaced crystal layers and microcrystalline calcite or dolomite laminae (Fig. 3D). Crystal layers are composed of ghost structures of angular crystals causing a serrated top of the layers. The crystal layer and laminae thicknesses vary between 250 and 1 mm. There is no fenestral fabric. No fecal pellets, ostracodes, or gastropods have been observed. The bulk mineralogical composition is about 80 wt% calcite and 20 wt% dolomite, with the dolomite concentrated in microcrystalline laminae. Mg and Sr concentrations are considerably higher if compared with the previously described filament tube framestones (Table 1).

Micropeloid-Microsparite Stromatolite

The highly porous, irregular framework of this stromatolite type is composed of dolomitic micropeloids, dolomite threads and subhorizontal-wavy lamellae, and subangular dolomite rhombs rich in inclusions. Between the dolomitic constituents, an irregular mosaic of nonuniform microspar to spar crystals is observed. Rare former open voids between stromatolitic protuberances, 1–6 mm in size, are lined by finely laminated calcite-dolomite coatings (Fig. 3E). Dolomite and microspar-dominated layers macroscopically alternate to result in a stromatolitic appearance. Dolomite and calcite constitute approximately half of the rock (40–60 wt%). Sr concentrations are high and vary between 2500 and 4200 ppm (Table 1).

Micropeloid-Microsparite Stromatolite With Fecal Pellets

This stromatolitic facies type is similar to that previously described, but shows larger quantities of micropeloids. In addition, pocket-like accumulations of dolomitic peloids are present between the stromatolitic protuberances. Between the peloids, dolomitic fecal pellets of 80–100 μm diameter and 300–400 μm length occur (Fig. 3F). These fecal pellets are similar to that of the present-day brine shrimp Artemia (see e.g., Eardley 1938). Major parts of the rock are dolomitic (70 wt%), with 30 wt% calcite forming parts of the microcrystalline laminae, peloids, and fecal pellets. A late calcite spar cements the remaining pore space. This facies type exhibits the highest Sr concentrations of all samples analyzed (6200 ppm) (Table 1).

Coated Bubble Travertine

Guo and Riding (1998, p. 168): This facies type is characterized by former gas bubbles 250 μm to 5 mm in diameter, fossilized by microcrystalline to fibrous calcite veneers of 10-20 μm thickness (“Lochblasenkalk”; Reis 1926). Upon these faint membranes, which occasionally show features of collapse, numerous phases of fan-like to dendroid arrays of calcite crystals with sweeping extinction coalesce to form a rigid framework (Fig. 3G). The remaining pore space is partially or fully cemented by late blocky spar. The coated bubble travertine of the Erbisberg consists of 100 wt% calcite. Sr concentrations are similar to those of crystal stromatolites, i.e., about 1300 ppm (Table 1).

In addition to the mound carbonates of the Erbisberg, postimpact calcite veins of the fractured crystalline basement below the crater floor have been investigated. These veins, obtained by the research drilling Nördlingen 1973, have already been mentioned by Bauberger et al. (1974: enclosure 1). For the present study, veins from fractured orthogneisses at 805.7 m, 871.7 m, and 881.4 m depth have been investigated. The host rock is gneiss composed of albite, quartz, biotite, and accessory anatase. The calcite veins are 2–5 mm thick, and show spar crystals (100–500 μm in size) rich in inclusions (Fig. 4). The veins are devoid of deformation or shock features and cement impact–generated fractures and thus are postimpact in origin.

Figure 4.

Postimpact calcite vein of the crater basement, drilling Nördlingen 1973 at 871.7 m depth. Sample Nö 73-3.

Stable Carbon and Oxygen Isotopes

Dendroid calcite cements of filament tube framestones (“streamer carbonate”) are characterized by slightly negative δ13C values around −1.2‰, and δ18O values at 0.0‰ (Table 2). Microcrystalline calcite patches in the same samples show similar values (Fig. 5).

Table 2. Stable carbon and oxygen isotope data of carbonates of the Erbisberg and reference samples
LocationSampleFaciesDescription (mineralogy)δ13CStd. dev.δ18OStd. dev.
(‰ PDB)(‰ PDB)
Erbisberg, western foothillErb A-1Filament tube framestoneTube carbonate (calcite)0.300.0070.900.016
Erbisberg, western foothillErb A-2Filament tube framestoneDendroid cement (calcite)0.130.0060.360.008
Erbisberg, top of hillErb BFilament tube framestoneTube carbonate with minor dendroid cement (calcite)0.710.0101.660.016
Erbisberg, top of hillErb CCoated bubble travertineMicrocrystalline bubble wall with fibrous cements (calcite)−0.280.0120.920.017
Erbisberg, trench at SE foothillErb 8Q-1Filament tube framestoneDendroid cement (calcite)−1.250.0070.100.018
Erbisberg, trench at SE foothillErb 8Q-2Filament tube framestoneDendroid cement (calcite)−1.300.0070.320.015
Erbisberg, trench at SE foothillErb 8Q-3Filament tube framestoneMicrocrystalline carbonate (calcite)−1.090.004−0.290.011
Erbisberg, trench at SE foothillErb 8Q-4Filament tube framestoneMicrocrystalline carbonate (calcite)−1.240.004−0.020.014
Erbisberg, trench at SE foothillErb 8Q-5Filament tube framestoneMicrocrystalline carbonate (calcite)−0.790.004−0.160.014
Erbisberg, trench at SE foothillErb 8Q-6Filament tube framestoneMicrocrystalline carbonate (calcite)−1.280.006−0.020.016
Erbisberg, trench at SE foothillErb 1-1Crystalline crust travertineMicrocrystalline lamina (calcite)−0.660.0040.920.017
Erbisberg, trench at SE foothillErb 1-1Crystalline crust travertineMicrocrystalline lamina (calcite)−1.390.0100.100.028
Erbisberg, trench at SE foothillErb 1-2Crystalline crust travertineSpar cement (calcite)−6.090.008−7.320.013
Erbisberg, trench at SE foothillErb 1-3Crystalline crust travertineMicrospar replacing crystal traces (calcite)−6.630.007−5.720.016
Erbisberg, trench at SE foothillErb 2Crystalline crust travertineBulk sample (powder) calcite with traces of dolomite−1.570.006−1.050.011
Erbisberg, trench at SE foothillErb 2-1Crystalline crust travertineMicrocrystalline lamina (calcite)−1.490.009−0.600.017
Erbisberg, trench at SE foothillErb 2-1Crystalline crust travertineLight microcrystalline lamina (dolomite)0.400.0120.420.022
Erbisberg, trench at SE foothillErb 2-1Crystalline crust travertineDark microcrystalline lamina (calcite)−2.010.004−1.180.022
Erbisberg, trench at SE foothillErb 2-1Crystalline crust travertineLight microcrystalline lamina (dolomite)1.490.0082.110.024
Erbisberg, trench at SE foothillErb 2-1Crystalline crust travertineDark microcrystalline lamina (calcite)−1.310.008−0.100.021
Erbisberg, trench at SE foothillErb 2-1Crystalline crust travertineLight microcrystalline lamina (dolomite)1.220.0041.630.013
Erbisberg, trench at SE foothillErb 2-1Crystalline crust travertineDark microcrystalline lamina (calcite)−0.330.0111.140.020
Erbisberg, trench at SE foothillErb 2-2Crystalline crust travertineMicrospar replacing crystal traces (calcite)−6.990.006−4.820.010
Erbisberg, trench at SE foothillErb 2-2Crystalline crust travertineMicrospar replacing crystal traces (calcite)−7.560.009−5.620.016
Erbisberg, trench at SE foothillErb 2-3Crystalline crust travertineSpar cement (calcite)−6.640.010−5.920.014
Erbisberg, trench at SE foothillErb 3Micropeloid-microsparite stromatoliteBulk sample (powder) calcite with traces of dolomite−2.220.017−0.610.021
Erbisberg, trench at SE foothillErb 3-1Micropeloid-microsparite stromatoliteLight microcrystalline lamina (dolomite)1.260.0092.160.016
Erbisberg, trench at SE foothillErb 3-2Micropeloid-microsparite stromatoliteDark microcrystalline lamina (calcite)0.290.0052.450.009
Erbisberg, trench at SE foothillErb 4-1Micropeloid-microsparite stromatoliteLight microcrystalline lamina (dolomite)1.060.0052.540.028
Erbisberg, trench at SE foothillErb 4-1Micropeloid-microsparite stromatoliteFibrous vadose cement (calcite)−7.770.009−4.310.017
Erbisberg, trench at SE foothillErb 4-2Micropeloid-microsparite stromatoliteDark microcrystalline lamina (calcite)1.190.0094.040.008
Erbisberg, trench at SE foothillErb 4-3Micropeloid-microsparite stromatoliteSpar cement (calcite)−8.210.010−4.260.018
Erbisberg, trench at SE foothillErb 4-4Micropeloid-microsparite stromatoliteLight microcrystalline lamina (dolomite)2.870.0083.880.014
Erbisberg, trench at SE foothillErb 5-1Micropeloid-microsparite stromatoliteLight microcrystalline lamina (dolomite)−0.680.0091.010.017
Erbisberg, trench at SE foothillErb 5-2Micropeloid-microsparite stromatoliteSpar cement (calcite)−8.580.014−2.720.020
Erbisberg, trench at SE foothillErb 6Micropeloid-microsparite stromatolite with pelletsBulk sample (powder) dolomite with traces of calcite0.730.0081.300.023
Erbisberg, trench at SE foothillErb 6-1Micropeloid-microsparite stromatolite with pelletsMicrobial peloids (dolomite, calcite) with microspar−0.890.0070.650.023
Erbisberg, trench at SE foothillErb 6-2Micropeloid-microsparite stromatolite with pelletsMicrobial peloids (dolomite), Artemia-pellets (calcite), microspar0.740.0131.130.031
Erbisberg, trench at SE foothillErb 7-1Micropeloid-microsparite stromatoliteLight microcrystalline lamina (dolomite)0.210.0060.800.010
Drilling Nördlingen 1973 at 871.7 m depthNoe 3Basement postimpact calcite veinInclusion-rich spar cement (calcite)−10.690.012−14.180.028
Drilling Nördlingen 1973 at 805.7 m depthNoe 4Basement postimpact calcite veinInclusion-rich spar cement (calcite)−10.280.019−12.550.016
Drilling Nördlingen 1973 at 881.4 m depthNoe 5Basement postimpact calcite veinInclusion-rich spar cement (calcite)−8.660.020−14.890.047
400 m W' Ursheim; 508 m ü.NNUrs 1Arenaceous ooid packstoneShell of Lymnaea dilatata (Noulet) (aragonite)−3.170.034−4.110.033
400 m W' Ursheim; 508 m ü.NNUrs 2Arenaceous ooid packstoneShell of Lymnaea dilatata (Noulet) (aragonite)−2.780.022−3.920.041
400 m W' Ursheim; 508 m ü.NNUrs 3Arenaceous ooid packstoneShell of Lymnaea dilatata (Noulet) (aragonite)−2.960.018−3.510.041
400 m W' Ursheim; 508 m ü.NNUrs 4Arenaceous ooid packstoneShell of Planorbarius mantelli (Dunker) (aragonite)−0.900.0150.250.042
Figure 5.

Cross-plot of stable carbon and oxygen isotopes of travertines of the Erbisberg and crater floor postimpact calcite veins of the drilling Nördlingen 1973. For comparison, stable isotope data of several modern travertines (Guo et al. 1996; Kele et al. 2008, 2011; Rodríguez-Berriguete et al. 2012) are shown, too.

The carbonate subsamples obtained from crystal crust travertines exhibit a clear covariation of δ13C and δ18O values (Fig. 5), from slightly negative values for calcite laminae (δ13C = −2.01‰, δ18O = −1.18‰) to moderately positive values for dolomicrite laminae (δ13C = 1.49‰, δ18O = 3.90‰) (Table 2). However, a contamination of dolomicrite samples by minor amounts of calcite from adjacent laminae, and vice versa, cannot be excluded.

Calcite spar cements and neomorphic calcite microspar are characterized by moderately negative values for both δ13C and δ18O (Table 2), i.e., clearly distinct from all other values of the Erbisberg travertine components. A single value obtained from late calcite vadose cement falls in the same range of δ13C and δ18O values.

Calcite veins of the fractured gneiss below the crater floor (research drilling Nördlingen 1973) show the lowest δ13C and δ18O values obtained from the samples investigated (Table 2). Their δ13C values are 1–3‰ lower, their δ18O values are 2–3‰ lower than the corresponding values of meteoric sparite in the travertines (Fig. 5).

Strontium Isotopes

Eight bulk rock samples from the Erbisberg travertine have been analyzed with respect to 87Sr/86Sr ratios, with values ranging from 0.71148 to 0.71171 (Table 3). The values show a covariation with dolomite contents (r = 0.91; Fig. 6A).

Table 3. Rb and Sr concentrations and 87Sr/86Sr isotope ratios of carbonates of the Erbisberg
LocalitySampleFaciesCarbonate mineralogy (mole%)RbSr  
ErbisbergErb 1Crystalline crust travertine8713615250.711660.00002
ErbisbergErb 2Crystalline crust travertine802028520.711630.00003
ErbisbergErb 3Micropeloid-microsparite stromatolite57430.339400.711580.00003
ErbisbergErb 4Micropeloid-microsparite stromatolite33670.328140.711480.00003
ErbisbergErb 5Micropeloid-microsparite stromatolite57430.956160.711640.00002
ErbisbergErb 6Micropeloid-microsparite stromatolite with fecal pellets27730.125680.711540.00003
ErbisbergErb 7Micropeloid-microsparite stromatolite60400.416050.711610.00002
ErbisbergErb 8Filament tube framestone9730.1449170.711710.00002
Figure 6.

Covariation of 87Sr/86Sr values with A) dolomite, and B) stable carbon and oxygen isotope ratios of carbonates of the Erbisberg.

The filament tube framestone sample, which is almost devoid of dolomite, exhibits highest 87Sr/86Sr ratio, i.e., 0.71171. Crystalline crust travertines, with dolomite contents between 13 and 20 mole%, have 87Sr/86Sr ratios of 0.71166 and 0.71163, respectively. Micropeloid-microsparite stromatolites, with dolomite contents between 40 and 67 mole%, show a broad range of 87Sr/86Sr ratios between 0.71148 and 0.71164, which is, however, in general lower than ratios of cement crust stromatolites. Indeed, the sample Erb 4, with 67 mole% dolomite, shows the lowest 87Sr/86Sr value, i.e., 0.71148, of the samples analyzed. Finally, the sample of micropeloid-microsparite stromatolite with fecal pellets, with 73 mole% dolomite, shows the second lowest 87Sr/86Sr ratio (0.71154).

Based on the calculated covariation, pure dolomite and calcite at the Erbisberg are expected to have an 87Sr/86Sr of 0.71140 and 0.71173, respectively (Fig. 6A). Likewise, 87Sr/86Sr isotope ratios correlate with stable carbon and oxygen isotope ratios, although with less high correlation coefficients (Fig. 6B).

Rare Earth Element Pattern

Eleven travertine samples from the Erbisberg and one calcite vein from the crystalline basement of the research drilling Nördlingen 1973 have been analyzed for the REE pattern. For comparison, further 14 carbonate sediment samples from the Ries basin and adjacent areas were investigated (Table 4). The highest ∑REE values have been detected in the calcite vein from the crystalline basement and dolomites of central lake basin sediments (Table 4). The obtained PAAS-normalized REE values show distinct pattern (Fig. 7). A true, negative Ce anomaly, as defined by the Ce/Ce*SN and Pr/Pr*SN ratios (e.g., Murray et al. 1991; Bau and Dulski 1996), has only been found in one of the reference samples, a Jurassic marine limestone.

  1. REE patterns enriched in HREE are typical for filament tube framestones (“streamer carbonates”), coated bubble travertines, and microsparite stromatolites. In particular, the filament tube framestones exhibit steep pattern highly depleted in LREE. A single analysis of a filament tube framestone deviates from this observations by a flat pattern showing a positive Eu anomaly (Fig. 7A).
  2. Crystalline crust travertines of the Erbisberg revealed flat REE patterns (Fig. 7B). A crystalline crust travertine from Böttingen (Miocene, Swabian Alb) exhibited an uneven, more or less flat REE pattern with relatively high La and Lu contents (Fig. 7D). However, the uneven pattern may reflect the extremely low ∑REE value.
  3. Flat REE patterns with a positive Eu anomaly (EuSN/Eu*SN = 1.68–2.11) characterize the calcite vein of the fractured crystalline basement (Fig. 7A). The vein margin, analyzed by LA-ICP-MS, exhibited very high REE concentrations and the most pronounced positive Eu anomaly (Table 4).
  4. In contrast, lacustrine carbonates of the Ries basin show flat to upward convex, roof-shaped REE patterns (Fig. 7C). While the stratigraphically oldest sample, an argillaceous Hydrobia-limestone, exhibits a flat REE pattern, younger lake basin dolomites, as well as dolomitized marginal lacustrine algal reefs, are characterized by roof-shaped REE pattern. Similarly, the stratigraphically youngest limestones containing freshwater mollusks exhibit a pronounced roof-shaped REE pattern (Fig. 7C).
  5. A flat REE pattern with negative Ca anomaly has only been obtained from a marine Jurassic limestone of the Bunte Breccia (Fig. 7D).
Table 4. Rare earth element (REE) concentrations and shale-normalized (SN) REE values of carbonates of the Erbisberg and reference samples
LocationSampleDescriptionBaLaCePrNdSmEuGdTbDyHoErTmYbLu∑ REE∑ LREE∑ HREE
ErbisbergErb BStreamer carbonate (hill top)82500781592313351168923274984408771712723165501766
ErbisbergErb CCoated bubble travertine (hill top)438008119334185642291211854615624170251298671627
ErbisbergErb 1Crystalline crust travertine21800168030503681820320752973924346138191342082497610639
ErbisbergErb 2Crystalline crust travertine34300124019302501060221532414121847155241742656804995685
ErbisbergErb 3Micropeloid-microsparite stromatolite76900628188025915405371637431349211885557954184825257502502
ErbisbergErb 4Micropeloid-microsparite stromatolite8650068318102721550524154707128928206703112818124871957003019
ErbisbergErb 5Micropeloid-microsparite stromatolite499001150250032416004341265481007051484446140658860466821922
ErbisbergErb 6Micropeloid-microsparite stromatolite with fecal pellets7970039995713162219459277605031254406948272439026391751
ErbisbergErb 7Micropeloid-microsparite stromatolite377007381370163782168421913424861214352503943353454881
ErbisbergErb 8Streamer carbonate (trench)6210053791052167329822172128614545249296
Research drilling Nö 73Noe 871Calcite vein of crystalline basement (871 m depth)2300016200270003180134002854983265040024004671400190129019372607662676340
Research drilling Nö 73Noe 871Calcite vein 871 m depth (LA-ICP-MS)13500007000001420000128000393000773003080060800726041400750020700312018900316029119402809900102040
NähermemmingenNaeh 38Peloid-wacke/packstone with ostracods71800269058107063340737178788114711147442583885616165142491916
NähermemmingenNaeh 41Cladophorites-bafflestone8640022904600596269057513960090581119336462804012982114901492
NähermemmingenNaeh 44Ostracod-wackestone68500402079509994420103325211201731130236705976239322851197943057
Research drilling EnkingenEnk 12.01Dolomite lamina in bituminous shale1660001630036200470021100490012004860710424877522502741670230994178926010157
Research drilling EnkingenEnk 21.10Argillaceous Hydrobia-limestone6120013700212002510102002070410180026817003431060149101014556565518904675
Wallerstein spring moundWa 98/8Nonskeletal stromatolite8050048111301365901454116624153339715981631252689436
Goldberg spring moundGold 1Coated bubble travertine24800135214351924516641290249115114181064700364
Ehingen algal biohermEhMu 55Cladophorites-framestone134000500013000222011600267067723603371850312824935056941517375273990
Utzwingen algal biohermUtz 1Skeletal stromatolite170000215054208224070106027810901831160233690925979017935148903045
Hasenberg algal biohermHas 1Skeletal stromatolite143000208043505222440589165632106721153465604006012743107781965
UrsheimUrs 1Arenaceous ooid packstone with Limnaea87600278055906923100688169671100625115340422593615207136901517
BreitenloheBrei 1Charophyte wackestone with Gyraulus116000542012400157072201550395156023914502838151076769733782301153667
BöttingenBoetCrystalline crust travertine95002941871864102122102616261658729
Aumühle QuarryAu 1aUpper Jurassic micrite with Tubiphytes960031001390343152029775379614411063324528740841671041312
  1. Ce/Ce*SN = CeSN/(0.5PrSN + 0.5LaSN); Pr/Pr*SN = PrSN/(0.5CeSN + 0.5NdSN); Eu/Eu*SN = EuSN/(0.5SmSN + 0.5GdSN).

Research drilling Nö 73Noe 871Calcite vein of crystalline basement (871 m depth)0.4240.3390.3600.3950.5140.9100.5690.5170.5130.4710.4910.4690.4570.4460.950.870.981.68
Research drilling Nö 73Noe 871Calcite vein LA-ICP-MS18.32517.83914.49611.59313.92828.51913.0479.3808.8467.5687.2637.7046.7027.2982.511.090.992.11
Erbisberg travertine moundErb BStreamer carbonate (hill top)0.0020.0020.0030.0040.0090.0150.0190.0300.0590.0990.1540.2150.2540.2930.010.860.881.07
Erbisberg travertine moundErb 8Streamer carbonate (trench)0.0010.0010.0010.0020.0030.0070.0070.0110.0180.0210.0250.0300.0300.0320.040.790.901.37
Erbisberg travertine moundErb CCoated bubble travertine (hill top)0.0020.0020.0040.0050.0120.0210.0200.0270.0400.0460.0550.0590.0600.0580.040.810.981.33
Goldberg spring moundGold 1Coated bubble travertine0.0040.0030.0040.0060.0080.0140.0140.0150.0190.0240.0320.0370.0400.0420.090.720.951.32
Erbisberg travertine moundErb 1Crystalline crust travertine0.0440.0380.0420.0540.0580.0700.0640.0500.0520.0460.0480.0470.0480.0460.950.890.911.15
Erbisberg travertine moundErb 2Crystalline crust travertine0.0320.0240.0280.0310.0400.0490.0520.0530.0470.0470.0540.0590.0620.0600.540.801.021.08
Wallerstein spring moundWa 98/8Nonskeletal stromatolite0.0130.0140.0150.0170.0260.0380.0360.0310.0330.0330.0340.0370.0350.0370.341.010.971.24
Erbisberg travertine moundErb 3Micropeloid-microsparite stromatolite0.0160.0240.0290.0450.0970.1510.1590.1730.1970.1900.1950.1950.1920.1940.081.030.851.18
Erbisberg travertine moundErb 4Micropeloid-microsparite stromatolite0.0180.0230.0310.0460.0940.1430.1520.1650.1980.2080.2470.2770.2900.2860.060.930.901.16
Erbisberg travertine moundErb 5Micropeloid-microsparite stromatolite0.0300.0310.0370.0470.0780.1170.1180.1290.1510.1490.1560.1510.1440.1340.220.940.931.19
Erbisberg travertine moundErb 6Micropeloid-microsparite stromatolite with fecal pellets0.0100.0120.0150.0180.0350.0550.0590.0780.1070.1260.1540.1700.1710.1660.060.950.981.16
Erbisberg travertine moundErb 7Micropeloid-microsparite stromatolite0.0190.0170.0180.0230.0300.0390.0410.0440.0530.0620.0750.0860.0890.0900.210.910.921.10
Research drilling EnkingenEnk 12.01Dolomite lamina in bituminous shale0.4270.4550.5320.6220.8831.1111.0430.9170.9080.7820.7890.6770.5920.5310.800.950.991.15
Research drilling EnkingenEnk 21.10Argillaceous Hydrobia-limestone0.3590.2660.2840.3010.3730.3800.3860.3460.3630.3460.3720.3680.3580.3351.070.831.001.00
NähermemmingenNaeh 38Peloid-wacke/packstone with ostracods0.0700.0730.0800.0990.1330.1650.1690.1470.1520.1480.1550.1430.1380.1290.540.970.931.09
NähermemmingenNaeh 41Cladophorites-Bafflestone0.0600.0580.0670.0790.1040.1290.1290.1160.1240.1200.1180.1140.0990.0920.650.910.981.11
NähermemmingenNaeh 44Ostracod-wackestone0.1050.1000.1130.1300.1860.2330.2400.2240.2410.2380.2470.2400.2210.2150.490.910.981.09
Ehingen algal biohermEhMu 55Cladophorites-framestone0.1310.1630.2510.3420.4810.6270.5060.4350.3950.3150.2890.2300.1790.1590.820.850.991.27
Utzwingen algal biohermUtz 1Skeletal stromatolite0.0560.0680.0930.1200.1910.2570.2340.2360.2480.2350.2420.2270.2120.2080.270.910.991.21
Hasenberg algal biohermHas 1Skeletal stromatolite0.0540.0550.0590.0720.1060.1530.1360.1370.1540.1540.1630.1480.1420.1390.390.960.931.26
UrsheimUrs 1Ooid packstone with Radix0.0730.0700.0780.0910.1240.1560.1440.1290.1340.1160.1190.1040.0920.0830.880.930.971.17
BreitenloheBrei 1Charophyte wackestone with Gyraulus0.1420.1560.1780.2130.2790.3660.3350.3090.3100.2860.2860.2640.2400.2240.630.970.961.19
BöttingenBoetCrystalline crust travertine0.0080.0020.0020.0020.0020.0020.0030.0030.0020.0020.0020.0020.0020.0061.390.480.960.75
Aumühle QuarryAu 1aJurassic micrite with Tubiphytes0.0810.0170.0390.0450.0540.0700.0810.0790.0940.1070.1160.1110.1020.0920.880.291.251.03
Figure 7.

Shale-normalized rare earth element patterns of carbonates of the Erbisberg and reference samples. A) Erbisberg filament tube framestones, coated bubble travertines, and postimpact calcite vein of the crater basement. B) Crystalline crust travertines and micropeloid-microsparite stromatolites of the Erbisberg. C) Lacustrine carbonates of the Ries basin. D) Travertine and speleothem reference samples (Möller et al. 2004, 2008; Zhou et al. 2012).


Depositional Setting

After the discovery that the Ries basin is not of volcanic (e.g., Gümbel 1870), but of impact, origin (Shoemaker and Chao 1961), the “travertine mounds” of the Ries have been considered cool-water artesian spring deposits (Bolten 1977; Bolten and Gall 1978) and, later, cool-water mixing-zone deposits (which have temporarily been subject to subaerial exposure) of subaquatic freshwater seeps of a soda lake (Arp 1995; Pache et al. 2001). This certainly applies for most of the spring mounds previously investigated, which are dominated by so-called sickle-cell limestones, thrombolites, nonskeletal stromatolites with Artemia fecal pellets, and various types of speleothems (Reis 1926; Bolten 1977; Arp 1995; Pache et al. 2001) known from present-day analogs (Russel 1889; Scholl 1960; Arp et al. 1998). However, two facies types not known from the Ries basin before are now recognized at the Erbisberg mound:

  1. The filament tube framestone is closely reminiscent of “streamer carbonates” (Fig. 8), which are well known from a number of extant and fossil subaerial hot springs (e.g., Walter et al. 1996; Mammoth hot springs; Farmer 2000; Fouke et al. 2000; Fouke 2011). This facies type forms in the “apron and channel facies” at 60–70 °C (Fouke 2011) of gently sloped channels where dense bunches of fine stringy microbial filaments cover the floor. Lower temperature streamers (e.g., 24 °C; Bonny and Jones 2008) apparently do not mineralize intensively to form tubes and corresponding streamer carbonates. Likewise, streamer-like fabrics have not yet been reported from arctic, cold-spring deposits, which otherwise show travertine-like macroscopic morphologies (Pollard et al. 1999). The subparallel, horizontally arranged tubes (also termed “Capellini fabric” [Fouke 2011]) result from mineral encrustation of flow-oriented filamentous micro-organisms (“streamers”) (e.g., Walter et al. 1996; p. 513; fig. 18). In present-day examples, sulfur-oxidizing members of the Aquificales form major parts of the streamers (Fouke et al. 2003). The calcifying streamer walls are composed of aggregates of aragonite needle botryoids (Fouke et al. 2000; Fig. 8c). Indeed, tube dimensions, their slightly undulating, subparallel arrangement, and the loop-like tube cross sections are almost identical to present-day analogs (Fig. 8). However, aragonite not preserved at the Erbisberg and the fabric now consists of fibrous calcitic crusts and dendroid calcite cements veneering the streamer tubes (Figs. 3A and 3B). While fibrous calcitic crusts and dendroid calcite cements may represent recrystallized former fibrous aragonite dendroid shrubs, later skalenoedric calcites indicate secondary cementation under vadose conditions.
    Figure 8.

    Filament tube framestone (“streamer carbonate”) of A) the Erbisberg mound, Ries impact structure, and B) modern analog from a hydrothermal spring (65 °C) near Viterbo, Italy. View on bedding planes.

    Figure 9.

    Strontium isotope stratigraphic position of the Erbisberg mound in the Ries basin. Magnetostratigraphy of the research drilling Nördlingen 1973 according to Pohl (1977), absolute age of Ries impact according to Rocholl et al. (2011), and absolute ages of magnetochrons according to Cande and Kent (1995).

  2. Crystalline crust travertines also have not been reported from other, sublacustrine, spring mounds of the Ries. The conspicuous serrated microfabric (Fig. 3D) of this travertine facies type demonstrates the former presence of mm-sized angular crystals and crystal arrays, which are now diagenetically converted into calcite. Indeed, these crystal layers point to rapid, rather inorganic precipitation on biofilm-poor surfaces, known from thermal spring orifices (“vent facies” 73–66 °C; Fouke 2011) to proximal slopes (47–60 °C; Kele et al. 2008: fig. 6c). Terracette-like morphologies, however, were not observed at the Erbisberg.

A further peculiarity of the Erbisberg mound, and difference from other, sublacustrine spring mounds in the Ries basin, is its complete deficiency in lacustrine fossils. In turn, Hydrobia shells and ostracode valves are present, although rare, in marginal parts of other, sublacustrine spring mounds. Here, nonskeletal stromatolites and pellet-tube-boundstones show abundant fecal pellets of the brine shrimp Artemia, locally associated with insect larval tubes (Arp 1995; Pache et al. 2001). At the Erbisberg, as opposed to that, fecal pellets have been detected only very rarely in one sample of micropeloid-microsparite stromatolites (Fig. 3F), indicating a limited, episodic influence of lake water for this facies type.

Coated bubble travertines (Fig. 3G) have previously been described from various hydrothermal springs (e.g., Allen and Day 1935; Kitano 1963; Schreiber et al. 1981; Chafetz et al. 1991; Koban and Schweigert 1993; Guo and Riding 1998; Jones and Renaut 2010), but also from the Goldberg and Wallerstein spring mound of the Ries basin (Reis 1926; Bolten 1977; Pache et al. 2001). Coated bubble travertines form in shallow pools on slopes or distal settings such as valley floors (e.g., Guo and Riding 1999), and apparently are not restricted to high temperatures. Indeed, coated bubbles occur in shallow-water microbial mats of cool-water sublacustrine soda lake springs too (Arp et al. 1998).

Therefore, based on petrography, a subaerial setting near the lake shore is indicated for the Erbisberg mound, with filament tube framestones (“streamer carbonates”) in channels, crystalline crust travertines on proximal slopes (45–70 °C), coated bubble travertines in pools on the slope, and micropeloid-microsparite stromatolites on distal slopes (<45 °C).

Discharge of Groundwater

Strontium isotopes form a valuable tool to trace ground waters, spring waters, and diagenetic fluids (Faure 1977; Banner and Hanson 1990; Barrat et al. 2000). Geochemical evidence of ancient groundwater discharge at the Erbisberg can be derived from 87Sr/86Sr values of mound carbonates in comparison with those of lacustrine carbonates, rocks of the Bunte Breccia, and crystalline basement rocks. Taken as a whole, 87Sr/86Sr values at the Erbisberg show a wide range between 0.71148 and 0.71171 (Table 3). They are lower than average ratios of the Variscan crystalline basement and much higher than ratios of the Mesozoic sedimentary cover (Schnetzler et al. 1969; Horn et al. 1985; Pache et al. 2001).

When compared with lacustrine 87Sr/86Sr data of the Ries basin, the Erbisberg values are much higher than values of green algal bioherms slightly younger in age (Nähermemmingen: 445 m a.s.l.). This means that the Erbisberg 87Sr/86Sr ratios are shifted toward ratios of the crystalline basement, with increasing values from lacustrine-affected facies types to proximal slope facies types. Therefore, a groundwater discharge from the crystalline ring below the mound appears well substantiated.

However, the Erbisberg 87Sr/86Sr values show a clear covariation with dolomite contents (Fig. 6A). Because Mg concentrations should have been much higher in the lake water of the closed basin (Wolff and Füchtbauer 1976) if compared with spring waters of the Ries basin (Winkler 1972), facies types with high dolomite content might have been affected by lake water. This is certainly true for micropeloid-microsparite stromatolites with fecal pellets of the brine shrimp Artemia. On this basis, the lake water 87Sr/86Sr value of 0.71140 can be calculated for 100% dolomite content (Fig. 6A). On the other hand, facies types with only very low dolomite content, i.e., crystalline crust travertines and filament tube framestones, approach an 87Sr/86Sr ratio of 0.71173 for 100% calcite, which might have been the spring water value (Fig. 6A). This is certainly not a “pure” crystalline basement signal, and may indicate (1) mixing with lake water at the mound surface, and/or (2) mixing with lake-derived groundwater, and/or (3) mixing of Sr from basement and marine rocks in the groundwater pathway. While scenario (1) applies to sublacustrine spring mounds, scenario (2) also appears unlikely in the case of the Erbisberg because the proximal slope facies types are very low in dolomite, which is inconsistent with lake-derived brines. Only scenario (3) appears likely: If meteoric waters contributed to the spring waters of the Erbisberg, their pathway not only went through crystalline rocks but also through fractured Mesozoic sedimentary rocks of the megablock zone and Bunte Breccia.

Similarly, the 87Sr/86Sr versus δ13C and δ18O mixing lines (Fig. 6B) are not necessarily reflecting a primary mixing of waters, but could alternatively be explained by mixing of two separate phases (i.e., calcite and dolomite) during sampling for isotope analysis. Nonetheless, both phases may contain memory signals of primary δ13C and δ18O values, partially masked by the mixed sampling.

In summary, the high 87Sr/86Sr ratios of cement crust stromatolites and streamer carbonates indicate groundwater seepage from the crystalline basement rocks of the inner ring, which forms the basis of the mound. Indeed, a recent drilling recovered crystalline basement rocks below the Erbisberg travertine mound (Jung et al. 2011).

Diagenetic Alterations

With the exception of late calcite spar cements, calcitic and dolomitic components of the Erbisberg carbonates form a mixing line between slightly negative and moderately positive δ13C and δ18O ratios (Fig. 5), which could at least partly reflect primary isotopic signatures. This covariation of δ13C and δ18O at the Erbisberg is similar to that observed in present-day hydrothermal springs (e.g., Guo et al. 1996; Kele et al. 2008, 2011; Rodríguez-Berriguete et al. 2012; Fig. 5). There, the decrease in temperature, evaporation, and Rayleigh fractionation during CO2 degassing lead to successive higher values for δ18O and δ13C along the outflow of thermal springs (Kitano 1963; Gonfiantini et al. 1968; Friedman 1970; Chafetz and Lawrence 1994; Fouke et al. 2000).

However, petrography (Fig. 3) and REE pattern (Fig. 7) suggest that Erbisberg carbonates suffered diagenetic alteration, by early subaerial exposure and later meteoric-phreatic conditions. In particular, this applies to the filament tube framestone. Their presumably primary aragonitic tube walls are not preserved and now consist of inclusion-rich fibrous calcite cements and fan-like to dendroid arrays of calcite crystals, which possibly represent neomorphic, former aragonite shrubs. Alternatively, the dendroid calcites may have formed at lower temperatures (<40 °C; e.g., Fouke 2011), dissolved, and reprecipitated from an aragonitic precursor. Indeed, Sr contents are low, and steep REE patterns highly depleted in LREE (Fig. 7A) suggest a substantial dissolution-reprecipitation effect. In any case, low-temperature diagenetic overprints are to be expected in dying hydrothermal systems, when latest spring waters drop to ambient temperatures.

The microcrystalline calcite laminae in crystalline crust travertines may represent the least diagenetically altered phase of proximal types. Indeed, bulk rock REE patterns show less distinct depletion in LREE to more or less flat pattern (Fig. 7B). Also, microcrystalline calcite laminae show δ13C and δ18O values intermediate between dolomitic components of lacustrine-influenced micropeloid-microsparite stromatolite and the strongly diagenetic altered filament tube framestones. However, clear evidence for in situ recrystallized former aragonite cannot be derived from the present data and the associated calcite spar cements show unequivocal meteoric isotopic signatures (Fig. 5).

Likewise, micropeloid-microsparite stromatolites show mixed stable isotope signatures of purely dolomitic components and diagenetic calcite (Fig. 5). Again, clearly meteoric values were found for the calcite spar cements for this facies type.

In conclusion, no unequivocal pristine signals are evident for δ13C and δ18O of calcites at the Erbisberg. Fibrous, inclusion-rich calcites and calcite shrubs possibly developed from aragonitic precursors, while microcrystalline calcites may contain partly primary precipitates of the spring. In turn, the isotopic composition of dolomitic components reflects temporary flushing by lake water at marginal mound parts.

Despite these diagenetic overprints on stable isotopes, alteration of primary 87Sr/86Sr ratios should have been negligible, as meteoric dissolution of travertine carbonate and reprecipitation further below were without external Sr influx.

Hydrothermal Waters from Basement: Eu Anomaly versus Reprecipitation

  • 1.High ∑REE and REE pattern with clear positive Eu anomaly have been obtained from a calcite vein of crater floor crystalline basement (Table 4; Fig. 7A). This vein cross-cuts shocked and fractures gneiss, thus postdates impact (Fig. 4). Such clear, positive Eu anomalies are commonly explained by the interaction of hydrothermal fluids with plagioclase, where Ca2+ is substituted by Eu2+ (Sverjensky 1984; Bau 1991). Eu2+ leached from plagioclase is then precipitated with calcite in veins or surface deposits (e.g., McLennan 1989). High ∑REE, which could reflect low pH conditions, higher concentrations of complexing ligands, and/or higher temperatures (e.g., Cantrell and Byrne 1987; Michard 1989; Lottermoser 1992), are consistent with this interpretation.

Indeed, the calcite veins have lowest δ18O values (Fig. 5), i.e., much lower values than cool meteoric waters of the Ries basin, indicating increased temperatures. Hence, the hydrothermal origin of these veins appears conclusive. Indeed, Stettner (1974: p. 45) already mentioned the possibility that the calcite veins, associated with zeolites, result from hydrothermal activity, with Ca2+ derived from decomposed plagioclase.

  • 2.Travertine mound carbonates of the Erbisberg, however, exhibit less high ∑REE and only insignificantly increased Eu2+ (Table 4; Fig. 7). Instead, depletion in LREE and relative enrichment in HREE (i.e., a low La/Lu ratio) are the norm in these facies types. This unusual pattern either reflects a primary signal of the spring water or a diagenetic effect.

However, such LREE-depleted and HREE-enriched pattern are not known from present-day spring waters. Instead, they have been reported from highly alkaline soda lake waters (Johannesson et al. 1994) and corresponding “tufa” deposits (Wilcox 2010), where they reflect the preferential complexation of HREE to carbonate ions (Johannesson and Lyons 1994). Consequently, one may speculate about highly alkaline, HCO3/CO32−-rich spring waters for the Ries basin, possibly similar to the situation at Lake Bogoria thermal springs (Renaut et al. 2013). Indeed, present-day groundwaters of the crystalline ring show increased alkalinities (5–10 meq L−1; Winkler 1972), which is, nonetheless, much lower than typical soda lake waters (22–777 meq L−1; see e.g., Council and Bennett 1993; Kempe and Kazmierczak 1997; Arp et al. 1998, 1999).

The alternative explanation, i.e., diagenesis, therefore appears more likely: Assuming a primary aragonitic mineralogy of the Erbisberg filament tube framestones, the conversion to calcite may have affected the REE pattern. While the neomorphic conversion of aragonite to calcite is reported to cause only a minor depletion in LREE (Webb et al. 2009), patterns almost identical to that of the Erbisberg filament tube framestone have been described from certain laminae of a stalagmite in China, although at much lower ∑REE (Zhou et al. 2012; Fig. 7D). While the overall REE pattern of these speleothems appears poorly affected by host rock dissolution and reprecipitation, a set of laminae formed under cool and dry conditions exhibit LREE depleted/HREE enriched pattern, explained by longer travel time and increased ionic strength of karst groundwater (Zhou et al. 2012). We therefore suggest that the Erbisberg filament tube framestone REE patterns reflect subaerial exposure and slow karstification under dry, semiarid conditions.

This scenario is in accordance with the presence of meteoric-vadose sinter crusts and spar cements in these facies types, and that the likely primary precipitate aragonite is not preserved. Apparently less affected by dissolution and reprecipitation are cement crust travertines, showing flat REE pattern compatible with neomorphism (cf. Webb et al. 2009). A significant diagenetic alteration of mound carbonates therefore is evident, which obliterated possible primary REE signatures of hydrothermal activity. Indeed, this obliteration of primary signatures is in accordance with the general cooling history of impact-generated hydrothermal systems, where early high-temperature phases are commonly overprinted or erased by the later following cooler conditions (e.g., Nelson et al. 2012).

  • 3.Likewise, a negative Ce anomaly, which is commonly observed in marine carbonates (e.g., Piper 1974; Hu et al. 1988), such as those of the Upper Jurassic carbonate of the Weißjura-Group (Fig. 7D), is absent in the Erbisberg carbonates. Negative Ce anomaly results from high pε, i.e., oxidative conditions, when Ce (III) is microbially transformed to Ce(IV)O2, and preferentially removed from solution (De Baar et al. 1985; Elderfield 1988; Moffett 1990). In turn, Ce (III) is kept in solution under reducing conditions. The absence of a negative Ce anomaly therefore would be consistent with reducing conditions during formation of primary mound carbonates.
  • 4.Open-lake carbonates are distinct from the travertines by their roof-shaped REE pattern. A negative Ca anomaly is absent, too. Such roof-shaped REE patterns, known from a variety of different waters (Hoyle et al. 1984; Michard 1989) and hydrothermal deposits (Bau and Möller 1992; Hecht et al. 1999), are considered to reflect a remobilization of REE from pelitic metasediments, probably under acid conditions (Lüders et al. 1993; Hecht et al. 1999). The REE pattern of the Ries open lake carbonates therefore may reflect fluids derived from metapelitic basement rocks and pelitic sediments of the Mesozoic cover. With respect to the high ∑REE in the dolomicrites intercalated between bituminous shales, a concentration via organic matter-bound REE solution complexes (Tang and Johannesson 2003) into the anoxic monimolimnion is an explanation.

In summary, subsurface calcite veins show unequivocal hydrothermal REE pattern, while diagenesis has largely obliterated primary signals in the surface travertine mound carbonates.

Discussion: Extent and Duration of Postimpact Hydrothermal Activity

While the principal existence of a post-impact hydrothermal system is undisputed for the Ries basin, its extent and duration appear less well established. Previous investigations on hydrothermal activity in the Ries basin have focused on alterations of the suevite, which is considered the major source of heat for fluid convection (von Engelhardt 1972; Newsom et al. 1986; Osinski et al. 2004).

Initial interpretations suggesting a postimpact hydrothermal activity in the outer suevite were based on vertical chimney-like structures (von Engelhardt 1972; Chao et al. 1978) with Fe-rich oxide or hydroxide coating, which were interpreted as degassing pipes (von Engelhardt 1972; Newsom et al. 1986). Moreover, the high contents of Fe-rich montmorillonite in the outer suevite were considered to reflect hydrothermal alteration at temperatures below 100–130 °C (Newsom et al. 1986).

Hydrothermal alterations at higher temperatures were detected in the 270 m thick crater-fill suevite in the research drilling Nördlingen. Here, albitization, chloritization, and zeolitization led to a significantly altered matrix, shocked rock fragments, and glassy components (Stähle and Ottemann 1977; Osinski 2005). Later investigations suggested that the alteration phases in outer suevite could be explained by meteoric weathering alone, leaving the crater-fill suevite as the only pervasively hydrothermally altered rock type (Muttik et al. 2008, 2010).

However, a specific feature of impact-generated hydrothermal systems is that an initial pulse of high temperatures is followed by a longer tail of cooling. During this temperature decline, early high-temperature precipitates and their geochemical signals are commonly altered or completely converted, so that their detection could be impeded (e.g., Zürcher and Kring 2004; Zürcher et al. 2005; Nelson et al. 2012). To date, a hydrothermal fluid convection system has, therefore, only been shown for the crater suevite, although craters of similar size exhibit much more extended systems, e.g., the Haughton impact structure (Osinski et al. 2001, 2005), where temperatures (about 210 °C) lasted for approximately 5 kyr (Parnell et al. 2005; Lindgren et al. 2009).

Time estimates are derived from calculations of suevite cooling assuming that the hydrothermal activity is largely linked with impact melts: Model calculations by Daubar and Kring (2001) suggest periods of 104–105 yr in 100 km sized craters and up to 106 yr in 180 km sized craters for a cooling to 20% of the initial temperature. For a 25 km sized crater, a cooling of a 200 m thick melt sheet from 800 °C to 160 °C within a period of 3 × 103 yr can be derived (Daubar and Kring 2001; Fig. 2). With respect to the Ries impact structure, (Pohl 1977: p. 345) calculated that the cooling of a 200 m thick melt-rich suevite layer from initially 800 °C to 100 °C took several thousands of years. Apart from these model calculations, as in many other impact craters (see, e.g., Daubar and Kring 2001), the duration of the hydrothermal activity in the Ries impact structure remained not well constrained.

For the Erbisberg mound, the 87Sr/86Sr lacustrine estimate provides a tentative stratigraphic correlation with sediments encountered in the central crater (Fig. 9). Carbonate sediments of the Ries basin exhibit a unidirectional 87Sr/86Sr trend from higher to lower ratios, due to a change from weathering crystalline basement rocks and suevite to Jurassic marine limestones in the catchment area as major source of Sr2+ influx to the lake water (Arp et al. 2013).

The 87Sr/86Sr lacustrine estimate of the Erbisberg (0.71140) falls in range of values of carbonate intercalations in bituminous shales encountered by the drilling Enkingen (Arp et al. 2013). These shales are considered equivalent to youngest parts of the laminite member of central basin succession (Jankowski 1981), thereby narrowing the relative age of the Erbisberg to this stratigraphic interval (Fig. 9).

While the whole currently preserved lacustrine sediments in the basin center may cover a time period up to 1.2 Ma (Jankowski 1981: p. 206), the 145 m thick laminite member alone could have formed within 100,000–300,000 a, based on laminae couplet thicknesses of 0.5–1.5 mm. This is in accordance with magnetostratigraphic data (Pohl 1977), which suggest about 276,000 a for the laminite interval (polarity chrons C5ADr and C5Bn.1n; Cande and Kent 1995). Neglecting a much shorter time period for the coarse-siliciclastic basal member and assuming that the strontium isotope stratigraphic correlation is correct, Erbisberg mound was active about 250,000 yr after the impact event. This implies that the impact-generated hydrothermal system was active much longer than previously assumed, although this time estimate refers to a late, low-temperature hydrothermal phase, as there are no indications of temperatures higher that 70 °C at the Erbisberg. Indeed, in a recent study, Schmieder and Jourdan (2013) argue, on the basis of a 1.1 ± 0.5 Ma offset of ages between the primary impact melt rocks and younger, slowly cooled K-feldspar melt particles, that postimpact hydrothermal activity in the 23 km sized Lappajärvi impact structure could last between approximately 600 ka and approximately 1.6 Ma.

It has to be noted that, besides impact melts, uplifted warm basement rocks form a further significant source of heat in impact structures (e.g., McCarville and Crossey 1996; Pirajno 2009a). This may indeed apply to the Ries basin, where a substantial part of the initial impact melt sheet may have been lost by a phreatomagmatic-like “fuel-coolant interaction” (Artemieva et al. 2013; Stöffler et al. 2013). Then, the limited hydrothermal activity concentrates at inner ring, where uplifted warm basement rocks gave rise to the Erbisberg travertine formation. Moreover, further potential indicators of an extended postimpact hydrothermal activity exist in the Ries basin: Localized hydrothermal sulfidic mineralization may explain that some suevite occurrences are characterized by high pyrite or pyrrhotite contents (Pohl, personal communication). In addition, high-temperature hydrothermal activity (vapor-dominated), earlier than the travertine mound formation, may be represented by silicified crystalline rock breccias, which have been reported as very localized occurrences in the Ries basin (N of Amerbach: Weber 1941; Dressler et al. 1969; NW of Schmähingen; Nathan 1926: pp. 66–67). Unfortunately, these rare breccias are poorly investigated, to date, and their potential impact-related genesis remains to be shown.


The Erbisberg mound, located on the inner crystalline ring of the Ries impact structure, shows two distinct facies types previously not described from the Ries basin: streamer carbonates and crystalline crust travertines. These facies types indicate a subaerial, hydrothermal origin of the mound, in contrast to previously described sublacustrine spring mounds of the Ries basin. Indeed, the fabric of the streamer carbonates, composed of subhorizontal slightly undulating tubes encrusted by dendroid arrays of calcite crystals, is almost identical to present-day streamer carbonates of hydrothermal springs.

While 87Sr/86Sr values of the carbonates indicate a discharge of groundwater from crystalline basement rocks of the Inner Ring below the mound (with admixture of waters from Bunte Breccia), diagenesis obliterated much of the primary geochemical signatures of the travertine. A REE pattern with a clear positive Eu anomaly has only been obtained from a calcite vein of the crater floor (drilling Nördlingen 1973), indicating hydrothermal fluid circulation for the crystalline basement.

87Sr/86Sr of lacustrine-influenced facies types of the Erbisberg suggest a correlation of the mound with upper part of the laminite member of the central basin succession; based on that, it is estimated that the hydrothermal activity of the mound persisted about 250000 years after the impact. In summary, our new data indicate that the postimpact hydrothermal activity in the Ries basin was not restricted to the crater-fill suevite, but more extended than previously assumed.


We thank Volker Dietze, Nördlingen, for informing us on the Nähermemmingen road construction. Thilo Bechstädt, Heidelberg, supported a field trip to the Erbisberg in 2006. We are grateful to Ingrid Reuber for stable isotope analysis, Brigitte Dietrich for strontium isotope analysis, Andreas Reimer for Corg-Ccarb analysis, Gerald Hartmann for XRF analysis, and Axel Hackmann for preparing thin sections. The manuscript benefited from discussions with Jean Pohl, München; Uwe Reimold; and Dieter Stöffler, Berlin. Very helpful and valuable comments, suggestions, and corrections by Horton Newsom and Gordon Osinski further improved the manuscript.

Editorial Handling

Dr. Christian Koeberl