Post-Miocene landscape rejuvenation at the eastern end of the Alps

The recent topographic evolution of the European Alps and its relationship to climate and tectonics remain strongly debated (e.g., Cederbom et al., 2004, 2011; Herman et al., 2013; Baran et al., 2014). In this debate, no consensus has emerged about the relative tectonic versus climatic significance of the driving forces responsible for the last stages of uplift, in particular those in post-Miocene times (Dunkl et al., 2005; Wagner et al., 2010; Hergarten et al., 2010; Willett, 2010; Delunel et al., 2010; Wittmann et al., 2007; Kuhlemann et al., 2002; Champagnac et al., 2007; Legrain et al., 2014).

Some studies have argued that a significant part of the post-Miocene rock uplift in the Swiss Alps can be attributed to climatic drivers: For example, Wittmann et al. (2007), Champagnac et al. (2007), Sternai et al. (2012), and Norton et al. (2010a) suggested that Pliocene–Pleistocene glacial carving resulted in erosion-driven uplift, and that this uplift is an effect of long-term landscape transience due to the repeated glacial-interglacial cycles. At the eastern end of the Alps, glacial carving was minimal, but a Pliocene–Pleistocene uplift event is recognized nevertheless (e.g., Wagner et al., 2010). This finding suggests that for the eastern Alps, a tectonic cause for the recent uplift pattern should not be excluded. This argument is consistent with most recent suggestions by Baran et al. (2014) or Cederbom et al. (2011) and a series of other recent studies (e.g., Cloetingh et al., 2006).

In order to place better constraints on this debate, the magnitude and the regional distribution of recent uplift and erosion in the Alps need to be documented in regions where glacial erosion can largely be excluded. The Koralpe range at the eastern end of the Alps is such a region, because it remained unglaciated except for isolated small cirque glaciers around the highest peaks (Fig. 1). More than 90% of the Koralpe range, including summits above 2000 m elevation, features a fluvial landscape that has been interpreted as an incised relict landscape of Miocene age (Robl et al., 2008; Legrain et al., 2014). The region therefore represents a unique opportunity to investigate young uplift processes and place them into the framework of the post-Miocene erosion and rock uplift increase in the Alps.

Here, we present ¹⁰Be-derived denudation rates (in this paper, these are also referred to as “erosion rates”) from different parts of the Koralpe landscape in order quantify landscape-forming processes in a range that features transient incision into a relict landscape. In particular, we determined erosion rates from “old” and “young” parts of the landscape in order to test and quantify our earlier mapping results (Legrain et al., 2014). Then, we used the ¹⁰Be-derived erosion rates for the two parts of the landscape and the known amount of incision into the relict landscape (Legrain et al., 2014) to calculate an estimate for the age of onset of incision and the amount of total relative base-level fall since onset of incision. We compared these results with different data sets from surrounding regions to explore if tectonic or climatic drivers may better explain the observed post-Miocene rock uplift increase in this part of the Alps. Finally, we inferred the topographic evolution of the Koralpe range since the late Miocene and compared it with the recent topographic evolution of the rest of the Alps.

The Koralpe range is located at the eastern end of the Alps, at the transition to the Pannonian Basin (Fig. 1). It is bordered to the west by the Lavanttal fault, one of the major strike-slip faults that had its peak of dextral strike-slip activity in the early Miocene, during the lateral extrusion of the eastern Alps (Ratschbacher et al., 1991; Frisch et al., 1998). The fault is a conjugate couple to the sinistral Mur-Mürz fault. Together, the two faults delineate the Styrian block to their east—a region inferred to have acted as a single coherent unit during the last few million years of tectonic evolution of the region (Wagner et al., 2011). The Styrian block includes both parts of the westernmost Pannonian Basin and its bordering ranges like the Koralpe. These bordering ranges also include the Styrian karst, which is part of the so-called “Paleozoic of Graz.” In the Styrian karst, burial-age dating evidence from caves has been used to date some 600 m of uplift within the last 4 m.y. (Wagner et al., 2010), with the uplift rate being rapid at first and decreasing in the last 2 m.y. In the Pannonian Basin area of the Styrian block east of the Koralpe range (the Styrian Basin), a kilometer-thick sequence of Miocene sediments forms a hilly landscape around 300 m surface elevation. The Koralpe range itself consists of high-grade metamorphic rocks that experienced peak metamorphism at ∼80 km depth around 90 Ma (Tenczer and Stüwe, 2003). Apatite fission-track ages document that the range was exhumed to near surface around 40–50 Ma (Hejl, 1997). (U-Th)/He thermochronology from the Koralpe range suggests a relatively steady exhumation from the Eocene to the Miocene of ∼100 m/m.y. (Legrain et al., 2014; Hejl, 1997), but no more detailed resolution of the changes in uplift rate within the last few million years is known. During the early Miocene, the Koralpe range is interpreted to have been tilted eastward, with the tilting having occurred during the main activity of the Lavanttal fault—between 18 Ma and 16 Ma (Neubauer and Genser, 1990; Kurz et al., 2011; Legrain et al., 2014). The Lavanttal fault may have been active later (see Scharf et al., 2013), but the lack of tilting in geomorphological features indicates that this later activity involved no vertical motion (Legrain et al., 2014). Lithologically, the Koralpe range is composed of paragneisses and mylonitic gneiss, and it includes the type locality for the rock type eclogite (Fig. 2). There are also small marble lenses, but most rocks are very quartz rich and allow for ¹⁰Be-derived erosion rates to be determined in all locations of the range.

Morphologically, the Koralpe region is characterized by two different areas: (1) a smooth, low-gradient relict landscape above roughly 1000 m elevation and reaching 2140 m on the highest point of the range (Speikkogel summit) and (2) a steeper, incised landscape below ∼1000 m elevation, involving deeply incised valleys that successively dissect the range from the east and west sides (Legrain et al., 2014; Winkler-Hermaden, 1957; Frisch et al., 1998). Interpretation of channel profiles shows that the upland relict landscape is clearly separated from the incised landscape by knickpoints and can be mapped around the range (white line on Fig. 2; Legrain et al., 2014). However, note that the relict landscape itself contained substantial local relief even prior to the young uplift and incision event discussed here, so that local channels within the relict landscape may also contain steep sections. It was also shown that the incision into the relict landscape resulted from some 400 m of uplift and incision during the post-Miocene geological history of the region (Legrain et al., 2014). However, the timing of this uplift period remains unknown. During the glaciation periods of the last million years, and in particular during the Last Glacial Maximum (LGM), the Koralpe range was never pervasively glaciated (van Husen, 1997). Cirque glaciers around the highest peaks were small and isolated (Fig. 3B; van Husen, 1997), but they significantly disrupted the smooth surface of the relict landscape and created isolated glacial cirques with steep faces of up to 200 m height (Figs. 3C and 4). The local influence of these cirque glaciers on the Koralpe morphology is readily visible in terms of morphological features and moraine deposits. These areas were excluded from the interpretation presented here in order to avoid confusion with fluvially sculpted landscapes and the interpretation of their erosion rates. Today, the mean annual precipitation of Koralpe scales with elevation and ranges from 900 mm/yr at the lowest elevations to 1600 mm/yr in the summit area (http://www.zamg.ac.at).

We determined catchment-wide erosion rates across the Koralpe range using the in situ–produced cosmogenic nuclide ¹⁰Be. Nuclide concentrations in quartz integrate erosion rates over a 10³ to 10⁵ yr time scale (e.g., Granger et al., 1996). We sampled 21 rivers for cosmogenic ¹⁰Be analysis, collecting sand from channel bottoms and active channel bars at several points along a 20 m reach at each stream location. Quartz was extracted from the 250–800 μm size fraction using magnetic separation and standard chemical leaching methods. We digested ∼40 g of clean quartz in a 5:1 concentrated hydrofluoric acid:nitric acid mixture, along with 215 μg of a ⁹Be carrier derived from a phenakite mineral. Beryllium was extracted from digested quartz using methods outlined in von Blanckenburg et al. (2004). We measured ¹⁰Be/⁹Be ratios on BeO targets with accelerator mass spectrometry (AMS) at ETH Zürich in Switzerland in June 2010 and 2011. Initial AMS results were normalized to AMS standard S2007N, with an isotope ratio of 2.81 × 10^–11. All results were renormalized to the 07KNSTD standardization. The ¹⁰Be concentrations were blank corrected (average ¹⁰Be/⁹Be ratio of five chemical processing blanks = 3.05 ± 2.31 × 10^–15) by subtraction. Analytical results are presented in Table 1. The ¹⁰Be concentrations were used to derive catchment-wide erosion rates following scaling factors from Dunai (2000) and absorption laws for nucleonic and muonic interactions from Schaller et al. (2002). Basin-averaged production rates were determined using 10 m gridded elevation data and using a sea-level, high-latitude production rate of 4.5 atoms/g quartz/yr. Corrections for skyline shielding were made following Norton and Vanacker (2009). In the absence of better information, snow shielding was calculated following Norton et al. (2011) using elevation–snow depth relationships determined in the Swiss Alps by Auer (2003). We set production rates equal to zero in parts of drainage basins where non-quartz-bearing lithologies were present (e.g., limestones). The ¹⁰Be derived denudation rates and analytical data are presented in Table 1.

We also defined another classification of catchments based on the drainage area of the sampled catchments. Previous analysis of catchment morphology and log-slope, log-area relationships identified the critical drainage area for each catchment that separates portions of the landscape dominated by hillslope and fluvial processes (Legrain et al., 2014). We assumed a conservative critical area of 7 km² (in accordance with Legrain et al., 2014) for the Koralpe region. Based on these results from digital elevation model (DEM) analysis, we defined three categories based on drainage area of the catchments: bedrock sample (A = 0), small catchments (A < 7 km²), and large catchments (A > 7 km²). Combining both classifications, our ¹⁰Be-derived erosion rate samples can be divided into five categories (Fig. 2; Tables 1 and 2): Samples from (1) relict small catchments (6 samples), (2) mixed small catchments (7 samples), (3) mixed large catchments (2 samples), (4) incised small catchments (2 samples), and (5) relict bedrock (1 sample) locations plus three samples from the Styrian Basin. There are no samples from incised large catchments (A > 7 km²), as all larger catchments in the Koralpe contain at least some fraction from the relict landscape. However, we did calculate theoretical erosion rates for large incised catchments using the two largest samples from the “mixed” catchments and Equation 1.

The ¹⁰Be-derived erosion rates from the Koralpe range from 36 mm/k.y. to 149 mm/k.y. (Fig. 5; Table 1). They are consistent with the long-term exhumation rate of ∼100 m/m.y. since the Eocene as inferred from fission-track and (U-Th)/He ages (Hejl, 1997; Legrain et al., 2014) for the Koralpe range. Erosion rates show a clear difference between relict and incised landscape (Fig. 5B). The mean erosion rate of the incised catchments is more than 2.5 times higher than the mean erosion rate of the relict landscape (137 ± 15 mm/k.y. and 49 ± 8 mm/k.y., respectively; Table 1). Erosion rates from catchments within the Koralpe relict landscape range from 36 ± 3 mm/k.y. to 59 ± 4 mm/k.y. (Table 1). The erosion rates of the two entirely incised catchments are 111 ± 9 mm/k.y. (Pro 6) and 149 ± 14 mm/k.y. (Pro 2). Erosion rates of mixed catchments comprising both incised and relict landscape range from 56 ± 3 mm/k.y. to 123 ± 9 mm/k.y. and average 81 ± 24 mm/k.y.

Erosion rates from the Styrian Basin east of the Koralpe range are 33 ± 2 mm/k.y. to 123 ± 9 mm/k.y. (Fig. 6; Table 1). Two larger catchment (Sti-1 and Pic-1) exhibit similar erosion rates, while the small catchment sampled in the Styrian Basin (Sau-1) shows a low rate of 33 ± 2 mm/k.y. However, in general, the erosion rates from the Styrian Basin appear to be relatively high across our sample region and average 88 ± 40 mm/k.y (Table 1).

Erosion rate data exhibit poor correlation with mean catchment slope (Fig. 5A), and plotted together, they do not support correlation models previously shown in other mountainous landscapes (e.g., Montgomery and Brandon, 2002; DiBiase et al., 2010; Carretier et al., 2013; Abbühl et al., 2011). However, these relationships become more clear when we divide our sampled catchments into the previously identified classifications based on size and degree of incision. For example, the correlation of erosion rate with mean catchment slope is notably more scattered for data from mixed small catchments than for incised small or relict small catchments (Fig. 5A). Overall, there is a positive linear correlation of erosion rate with the fraction of incised landscape within the catchment (r² = 0.49; Fig. 5B). The correlation coefficient omitting the mixed small catchments is significantly better (r² = 0.90; n = 10) than the correlation coefficient for all the data (r² = 0.49; n = 17).

The erosion rate data suggest that the relationships between landscape morphology and erosion rates are modulated by catchment size. Incised and relict portions of catchments each reflect distinct geomorphic processes. In large mixed catchments, the erosional expression of these processes is averaged out across the basin. However, we suggest that the heterogeneity of processes eroding small mixed catchments is not similarly averaged out due to the catchment size. Instead, erosion rates within these catchments capture the true spatial and temporal variability of erosion, as caused, for example, by rock type, jointing, local discharge, weathering, or the local degree of transient incision and hillslope response. A similar dependence of catchment size on erosion rate variability was previously demonstrated for the central Alpine Maggia Valley (Wittmann et al., 2007; Figs. 5 and 6).

In order to improve on the understanding of the correlation between erosion rate and proportion of the incised landscape in a catchment, we calculated theoretical erosion rates (E*) for mixed catchments based on a linear fit between the two end-member erosion settings (Fig. 5B). We defined these two end-member settings by the mean erosion rate of all measured catchments that are located either in 100% incised or 100% relict landscapes, respectively (incised catchments; n = 2; E* = 130 mm/k.y.; relict catchments: n = 6; E* = 49 mm/k.y.). The erosion rate E* is given by: E* = 81f + 49, where f is the fraction of incised landscape within the catchment (Fig. 5B; Table 1). We see that all mixed catchments feature erosion rates between these two end members, a finding similar to that of Van den Berg et al.,(2012) in the central Alps. The fact that mixed small catchments scatter significantly around the predicted value is consistent with the idea of small catchments expressing their local conditions and local erosion response. Interestingly, end-member catchments (100% relict small or 100% incised small catchments) that are smaller than this critical catchment size do not feature this variability—possibly because these portions of the landscape have attained a geomorphic equilibrium, such that catchments within each setting erode at the same rate.

Using Equation 1 to calculate erosion rates for the incised parts of large mixed catchments (Bis-1* = 146 ± 7 mm/k.y. and Las-3* = 143 ± 8 mm/k.y.), we find that the modeled erosion rates are consistent with measured rates for the small catchments entirely located in the incised landscape (Table 1; Pro-6 = 111 ± 9 mm/k.y. and Pro-2 = 149 ± 14 mm/k.y.). It is important to note that these two types of erosion rates are measured independently from each other. Therefore, the average of these four erosion rates (137 ± 15 mm/k.y.) should represent the erosion rate of the incised landscape of Koralpe (Fig. 6).

Erosion rates of two large catchments (Pic-1 and Sti-1) in the Styrian Basin average 117 mm/k.y. This value is comparable to the mean erosion rate of the Koralpe incised landscape (137 ± 15 mm/k.y; Fig. 6) and also to the long-term incision rate determined for this region from cosmogenic dating of cave sediments in Styrian karst (Wagner et al., 2010). This agreement suggests that the Styrian Basin is currently responding to a relative base-level fall that is transmitted to the Koralpe range but has not yet reached the relict landscape. This is logical if the Styrian Basin and the Koralpe range respond to a common rock uplift process as parts of the Styrian block.

The processes of successive adjustment of a catchment to a new base level can be explored in more detail by investigating a large mixed catchment that crosses both the relict and the incised landscape. The Prössingbach catchment northwest of the Koralpe summit is the most incised of the large Koralpe catchments (Fig. 2). It is therefore a good area to investigate how channel, hillslopes, and erosion rates adjust to increased uplift rates and associated base-level drop (Fig. 7).

Hillslope and channel morphology of the Prössingbach River show a clear separation between the upper relict landscape and the lower incised landscape. The river network in the channel shows a prominent knickpoint at this transition (Fig. 7). Interestingly, the mean slope of small tributary catchments connected to the Prössingbach River increase sharply 1.5 km downstream of the knickpoint (Figs. 7A and 7D). The spatial lag between the knickpoint and the increase in slope of adjacent hillslopes is inferred to reflect a temporal lag by the time needed for the hillslopes to propagate the incision signal upstream and increase the mean slope of the catchment.

The erosion rates of tributary catchments correlate well with the mean slope of the catchments, and the transition from incised to relict portions of the landscape. Erosion rates in the two relict catchments (36 mm/k.y.) are several factors lower than those of two incised catchments (average of 130 mm/k.y.; Fig. 7E). As above, the mixed hillslope catchments Pro-8 and Pro-7 are interpreted as scattering values between the two erosion rate signals.

The time response of hillslopes can be estimated by the distance between the knickpoint and the downstream location where the hillslopes within a tributary catchment are entirely within the incised portion of the landscape. These hillslopes have propagated the entire erosion signal up to the crest. The assumption is that the shorter is the distance, the faster is the time response (Hurst et al., 2012). For the Prössingbach catchment, one would therefore infer that the hillslope response time is different between the north- and south-facing hillslopes. Indeed, in downstream tributaries such as Pro-8, which is located 9 km downstream of the knickpoint, the incision wave has not yet propagated up to the crest, while hillslopes located south of the Prössingbach River are entirely incised only ∼3 km downstream of the knickpoint (Fig. 7A). However, these differences between north-facing and south-facing catchments might also be explained by the geometry of the fluvial network of this catchment, since southern tributaries joining the Prössingbach River are larger than those on the north side (Figs. 2 and 7A).

The data presented here provide several important constraints on the uplift of the Koralpe region. In particular, the difference in erosion rates between the two parts of the Koralpe landscape clearly shows that the landscape is currently adjusting to a wave of incision. We infer that rock uplift is largely equal to erosion for the Styrian Basin (assuming almost no surface uplift), so that the basin’s elevation remains approximately constant at the level of the regional base level, while the Koralpe is a transient landscape currently responding to recent uplift. The causes of the uplift and incision of the entire region and timing are discussed next.

Based on these relationships and the known amount of incision (Δz = 350 ± 90 m; from Legrain et al., 2014), we used a simple two-stage uplift model for the Koralpe region to define the different vertical motions (rock uplift, surface uplift, erosion) across the study area (Fig. 8). We also compared the incision of the Koralpe relict landscape with three other data sets: (1) the incision record of the Mur River in the Styrian karst as measured from burial age dating of cave sediments by Wagner et al. (2010), (2) subsidence analysis of the Styrian Basin (Ebner and Sachsenhofer, 1995; Sachsenhofer et al., 1998), and (3) subsidence analysis of the northern Molasse Basin (Genser et al., 2007). For the Koralpe, Δz represents the amount of incision since the rock uplift increase and ΔB_k is the total amount of base-level fall since the onset of rock uplift, also equivalent to the total amount of rock uplift since the uplift increase (Fig. 8). Subsidence analyses in the Styrian Basin by Ebner and Sachsenhofer (1995) have constrained some of these vertical motions in the Styrian block, and we use their results as a proxy for uplift.

Using these model and variables, we calculated the base-level fall and its timing for the Koralpe range using Equations 2 and 3. For the values of U_i and U_f, we used the average erosion rate of the catchment from the relict landscape (49 ± 8 mm/k.y.) and the average erosion rate from the incised landscape catchments (137 ± 15 mm/k.y.), respectively (Tables 1 and 2). Calculations yield an incision time Δt_k = 4.0 ± 1.0 Ma and a total relative base-level fall ΔB_k = 540 ± 140 m since the beginning of incision (Table 2). This timing of incision can be interpreted as the minimum age for the onset of incision because only the vertical incision is taken into account with Equation 3. Lateral knickpoint migration is neglected, and the calculated time of incision could be shorter than the real onset of incision.

Our estimates of Δt_k = 4.0 ± 1.0 Ma and ΔB_k = 540 ± 140 m are based on the assumption that erosion rates have been constant across the averaging time scale of ¹⁰Be-derived erosion rates (4–18 ka for our catchments). This assumption appears problematic because climate has changed repeatedly in the last million years, so that changes within the time-averaging of the method are expected. However, support for this assumption is provided by a comparison between short-term erosion from the Koralpe incised landscape and the long-term incision rate of the Mur River in the Styrian karst averaged over the last 4 m.y. (Wagner et al., 2010). Wagner et al. (2010) interpreted a detailed incision history of the Mur River for the last 4 m.y. and derived a mean incision rate of 130 ± 9 m/m.y. based on the oldest cave sediments (Fig. 3A; e.g., sample DH4: 4.05 ± 0.28 Ma). The close match between the ¹⁰Be-derived erosion rate of the Koralpe incised landscape (137 ± 15 mm/k.y.) and the long-term incision rate of the Mur River (130 ± 9 m/m.y. of Wagner et al., 2010) strongly supports the assumption that the ¹⁰Be derived-erosion rates of Koralpe are representative of the long-term erosion rates.

Our results and interpretation regarding the magnitude and timing of uplift can also be compared to the subsidence record of wells from both the Styrian Basin (Ebner and Sachsenhofer, 1995; Sachsenhofer et al., 1998) and the northern Molasse Basin (Genser et al., 2007). For the Koralpe region, we calculated a total relative base-level fall of ∼540 ± 140 m. This change is similar to the calculated base-level fall in the Styrian karst (ΔB_c = 543 m). Additionally, the base-level fall in the Styrian Basin (ΔB_b = 440 ± 100) is comparable within error to that in the northern Molasse Basin north of the region investigated here (ΔB_k = 540 ± 170 m; Fig. 8; see Genser et al., 2007). The similarity between these four independent data sets suggests that the entire region may have responded at approximately the same time (between 6 Ma and 4 Ma) to an increase in rock uplift rate, with a relative base-level fall of ∼500 m following uplift (Fig. 8). Figure 9 shows the spatial distribution of the amount of rock uplift for the different wells as derived from subsidence analyses of Ebner and Sachsenhofer (1995), Sachsenhofer et al. (1998), and Genser et al. (2007) near the eastern end of the Alps.

While climate has been invoked as a driver of uplift in other portions of the Alps (e.g., Wittmann et al., 2007), a climate-related post-Miocene uplift driver seems unlikely at the eastern end of the Alps. This region was well-removed from the main LGM ice body (Fig. 1), and post-LGM rebound due to ice unloading is limited to ∼100 m (Sternai et al., 2012). Also, our calculated timing for the increase in rock uplift at 4 ± 1 Ma is significantly older than the onset of periodic glaciations (see also Cederbom et al., 2011). Thus, the recent increase in uplift of the Koralpe range documented here neither fits with the timing nor with the spatial occurrence of the last periodic glaciations in the Alps. However, the uplift and incision seem closely related to the inversion of the vertical motions of the Styrian and northern Molasse Basins from subsidence throughout the Miocene. We therefore suggest that the causes for the uplift event discussed here may be closely related to the causes for the inversion of the Pannonian Basin.

In this context, it is useful to use our calculation of Δt_k and the amount of incision Δz to provide a rough estimate of the topographic evolution of the Koralpe range since the late Miocene. The present-day relief of Koralpe (R_f), between the summit and the elevation of the Styrian Basin, is ∼1800 m. By removing the amount of incision (Δz) from the present-day relief, we calculate an initial relief (R_i) of ∼1450 m (Fig. 8). This simple calculation suggests that ∼80% of Koralpe relief was present before the increase in rock uplift rate, and ∼20% was created by incision after the rock uplift rate increased. If we assume a constant regional base level, the elevation of the Koralpe summit was around 1800 m elevation compared to its present-day elevation of 2140 m. Although this suggests that most of the topography of Koralpe was present before 4 Ma, the increase of relief of 20% since that time is significant.

A simple explanation for this postorogenic tectonic-related increase in rock uplift could be the ongoing slow crustal convergence between Adria and stable Europe measured from global positioning system measurements (Bus et al., 2009). However, in this context, foreland basins like the northern Molasse Basin should subside due to plate loading and not be uplifted, as pointed out by Genser et al. (2007). We can therefore exclude a direct link between the convergence between Adria and Europe and the post-Miocene uplift of the eastern end of the Alps and surrounding sedimentary basins. If uplift related to climate and to active convergence can both be excluded, a deep-seated process seems to be a likely possibility to explain the observed increase in rock uplift rate since the late Miocene. Such an explanation is supported by the large wavelength of the emerging uplift pattern of several hundreds of kilometers, as shown in Figure 9. The spatial correlation of the emerging uplift pattern correlates to some extent with the location of the Adriatic slab at depth (Luth et al., 2013) and suggests mantle-derived causes for the uplift. Lippitsch et al. (2003), Kissling et al. (2006), and Genser et al. (2007) provided various models on the topology of the mantle lithosphere beneath the eastern Alps, and we refrain from discussing mantle processes further in view of the geomorphic focus of this paper (Fig. 9B).

The current erosion rates of the Koralpe are one order of magnitude lower than ¹⁰Be-derived erosion rates from the western, central, and eastern Alps, where they are typically of the order of a few millimeters per thousand years (Delunel et al., 2010; Wittmann et al., 2007; Norton et al., 2010a, 2010b, 2011). Such difference could be explained by a greater extent of glacial erosion to the west, which may precondition topography (e.g., Norton et al., 2010a) and/or further drive uplift and postglacial erosion (e.g., Wittmann et al., 2007).

At the scale of the Alps, the amount of young surface uplift documented here (350 m) seems very modest compared to the present-day topography of the Alps, where many summits are well above 3500 m in elevation. Additional uplift of 350 m would only provide an increase of summit elevation of 10%. However, we speculate that it is possible that this 10% of tectonically driven additional uplift (as documented here) may have been critical to trigger the creation of the Alpine ice cap in the glacial periods by contributing to a positive feedback mechanism between the small-amplitude tectonic-related uplift documented here and the well-constrained climate-related uplift (Norton et al., 2010b; Sternai et al., 2012).

We summarize our results and interpretation as follows:

• (1) The Koralpe and surrounding mountain ranges are the only peaks in the Alps that have summits above 2000 m surface elevation but were never pervasively ice covered during the glacial periods of the last 2 m.y. As such, they form one of the few locations within the range where glacial erosion was absent, and the geomorphology can be interpreted in terms of pre-Pleistocene processes.

• (2) ¹⁰Be-derived erosion rates from the Koralpe range are approximately one order of magnitude lower (36–149 m/m.y.) than erosion rates from previously glaciated regions of the Alps at similar elevation. However, there is a clear separation of relatively high erosion rates in what is interpreted as young, incised parts of the landscape (120–149 mm/k.y.) at low elevations and significantly lower erosion rates in the old relict landscape at high elevations (36–55 mm/k.y.). This confirms the interpretation that the Koralpe is a transient landscape that is currently responding to an increase in uplift rate that commenced only some few million years ago.

• (3) The data allow us to calculate an onset of incision for the Koralpe range of 4 ± 1 Ma and amount of total relative base-level fall of 540 ± 140 m. These results fit with data sets from surrounding regions in the Styrian karst, the Styrian Basin, and the northern Molasse Basin and suggest a common rock uplift increase for the whole area at the end of the Miocene. Such a post-Miocene uplift event has been documented for the central Alps (e.g., Baran et al., 2014), but it is new for the eastern Alps.

• (4) At the eastern end of the Alps, the post-Miocene uplift driver is unlikely to be directly due to active convergence between Adria and Europe or the result of a climatic change. We suggest that the increase in rock uplift rate may be due to a deep-seated process in the lithosphere, for example, thinning of the mantle part of the lithosphere by slab breakoff or delamination.

This work was funded by the Topo-Alps project (Austrian Science Fund project I152-N19). We thank Hella Wittmann for laboratory support. Fritz Schlunegger, Paul Bierman, and three further anonymous reviewers are thanked for their thoughtful, positive, and constructive comments.