A rare piedmont glaciation in the Mediterranean: Insights from cosmogenic 36Cl dating of Davraz hummocky moraine field (SW Türkiye)

Piedmont glaciers (lobes), typically found in high latitudes and large mountainous regions, extend from ice sheets and ice caps to lower altitudes. However, they can also occur, although less commonly, on mid‐latitude mountains. When these fan‐like glaciers retreat, they leave behind hummocky moraines scattered in a chaotic pattern. In this study, we have mapped one of these mid‐latitude sites and established a Terrestrial cosmogenic nuclide (TCN) glacial chronology on Mount Davraz, namely Davraz hummocky moraine field (37°46′00″N, 30°43′15″E). Our findings indicate that the glaciers in this area started receding from the early local Last Glacial Maximum (LGM) period (21.8 ± 2.4 ka) to the early Late‐glacial period (17.7 ± 2.2 ka), and eventually disappearing. The deglaciation of the Mt. Davraz palaeoglacier matches nearby mountains, supported by southerly winds as significant for regional glaciation. Our discoveries reveal a robust connection between southerly winds and nearby glaciation, contributing to our understanding of how climate influences glaciers. Likewise, the glacial timelines of the neighbouring mountains align with the glacial history of Mt. Davraz.


| INTRODUCTION
In various regions worldwide, particularly since the global Last Glacial Maximum (LGM; 26.5 to 19 ka, Clark et al., 2009), the growth of specific glaciers has been limited by the topography of the mountainous areas in which they originated.However, some glaciers surpass these boundaries and extend onto relatively flat terrain, spreading out like to a colluvial fans.This transition marks their shift to a distinct spatial existence, leading to their classification as piedmont glaciers (or lobes).Valley glaciers that flow through bed-rock troughs and terminate in low-lying regions can give rise to these types of glaciers (Benn & Evans, 2010).They may either have a cirque glacier origin (e.g., Smith et al., 1997), or they may originate from ice caps or ice fields (e.g., Evans et al., 2016;Ivy-Ochs, 2015), like the famous Malaspina Glacier in Alaska, USA (Post, 1972;Sharp, 1958); the largest piedmont glacier on the earth, covering an area of 3,209 km 2 (Kargel et al., 2014).Piedmont glaciers also exist on different planets than Earth, such as the Late Amazonian piedmont glacier located northwest of Olympus Mons on Mars (Milkovich et al., 2006).Several piedmont glaciers may merge and thicken to form small ice fields, which in turn merge to form larger ones.They thicken and grow due to their low gradients and ice velocities, which prevent them from discharging the ice from fast-moving valley glaciers that originate from steep mountainous areas behind (Bennett & Glasser, 2009).
After spreading out onto a flat area, piedmont lobes form frontal moraines within preexisting end moraine systems that were established during earlier ice advances (Ivy-Ochs et al., 2022).They provide great examples of large hummocky moraine fields, such as the ones in the Northern Alpine Foreland (Salzach Glacier) (Salcher et al., 2010), on Geyikda g, Taurus Mountains, Türkiye (Çiner et al., 2015) and on the Rhine and Reuss glacier systems (Kamleitner et al., 2023).
Hummocky moraines are primarily characterized as uncontrolled ice-disintegration features that are predominantly irregular and exhibit no apparent elongation (Gravenor & Kupsch, 1959).They are formed by a seemingly irregular collection of morainic mounds and ridges, giving the appearance of a chaotic or disorganized assemblage of glacial deposits (Balfour, 2014).This definition may include a diverse range of landforms, which may have various origins such as deposition linked to active or stagnant (i.e.dead) ice, and deformation due to glaciotectonic activity (Benn & Evans, 2010).Hambrey (1994) described hummocky moraine origin as a down-wastage of a glacier, rather than its recession, particularly during a state of glacial inactivity.
Correspondingly, the term 'hummocky moraine' has been used to refer to glacial sediments deposited during the down-wasting of debris-mantled glaciers.
Several dating studies aimed to reconstruct and analyse the behaviour of palaeoglaciers in cirques and glacial valleys located in the Turkish mountains.Thanks to the TCN dating method, especially in the last two decades, a comprehensive glacial chronology for Türkiye has been established.Accordingly, the earliest glacial retreat in Türkiye is traced back to late Marine Isotope Stage 5 (MIS 5; 78.5 ± 17.6 ka) (Altınay et al., 2022).Nevertheless, the most recent ages, Late Holocene glacial deposits (1.2 ± 0.3 ka, 2.5 ± 0.3 ka, and 3.8 ± 0.4 ka), linked to the Neoglacial expansion, have also been identified (Sarıkaya et al., 2009).The local LGM (21.5 to 18.5 ka) in Quaternary glaciation areas in Türkiye is correlated with the global LGM (26.5 to 19 ka) (Sarıkaya & Çiner, 2015).Active glaciers in Türkiye are relatively scarce compared to the Quaternary glaciations.The total area of Turkish glaciers amounts to 12.29 km 2 , with Mt.Ararat (5,137 m a.s.l.; the highest mountain of Türkiye) being the predominant contributor, encompassing over 60% of the region (Azzoni et al., 2020).
While the hummocky moraines created by piedmont glaciers represent uncommon phenomena in the Mediterranean, it is not surprising that there is a limited body of research on this subject.A comprehensive study (Çiner et al., 2015)

| Study area
Mt. Davraz (37 45 0 16 00 N, 30 43 0 30 00 E; 2637 m a.s.l., above sea level) is located in the southwest of Türkiye (Figure 1a), between the city of Isparta and Lake E girdir (916 m a.s.l.).It is a part of the western section of the Taurus Mountain Range and appears like a stairway structure from north to south, cascaded by SW-NE-oriented active normal faults (Emre et al., 2011) (Figure 1b).Old thrust faults such as Antalya complex thrust (Palaeocene) on the northeast and Lycian complex thrust (Langhian) with Aksu thrust (Late Miocene-Pliocene) on the southeast set the boundaries of Mt.Davraz (Poisson et al., 2003).It has also been argued that Mt.Davraz represents a slice of the Beyda gları Massif (a shallow-marine carbonate platform) by Poisson et al. (2003).The lithology of Mt.Davraz is mostly dominated by Triassic and Jurassic-Cretaceous neritic limestones.
The climate of Türkiye is regulated by seasonal changes between mid-latitude and sub-tropical/tropical pressure and wind systems, resulting in a subtropical Mediterranean macroclimate with hot and dry summers (Lolis & Türkeş, 2016).The distribution of Mediterranean moisture across Anatolia is significantly influenced by the Taurus Mountains.This mountain range acts as a natural barrier between the coastal and interior regions due to their high altitudes (>3 km, a.s.l) (Kuzucuoglu & Roberts, 1997;Sarıkaya et al., 2009).The study area is located on a transition between coastal areas and interior regions.It is characterized by a sub-humid climate (Akman & Keteno glu, 1986;_ Iyigün et al., 2013).The average summer and winter temperatures in the study area are 16.6 C and À0.9 C, respectively, while the mean annual temperature is 7.8 C, based on data measured from Isparta Davraz Ski Center weather station (37 46 0 13.2 00 N, 30 44 0 32.9 00 E; 1951 m a.s.l., however, only 5 years of observation, since 2017) (Figure 1c).The same data shows that the average total annual precipitation is 1,184 mm, with 67% of it occurring during winter and spring seasons.

| Previous studies
The first glacial study on Mt.Davraz was carried out by Louis (1944) within the context of Pleistocene glaciations in Anatolia.He identified several cirque glaciations on the mountain.The most detailed research was done by Ardos (1977a), who focused on the glacial geomorphology of the mountain.He mapped the development of cirques, moraines and rock glaciers on the northern slopes of Mt.Davraz.Ardos (1977a) also remarked the equilibrium line altitude (ELA), based on the cirque floor altitudes, as in between 2400 and 2450 m a.s.l.during Pleistocene.A recent regional study was done by Evans et al. (2021) about the palaeoclimatic evaluations from morphometric features of the cirques of all Western Taurus Mountains of Türkiye.Ten cirques of Mt.Davraz were part of this study.Based on their results, glacier mass balance at times of cirque development is mainly controlled azimuthally by solar radiation contrasts between shaded and exposed slopes (Derbyshire & Evans, 1976;Evans, 1977).However, this conclusion is not supported by the data from Mt. Davraz.Because the vector means aspect value of the cirques on Mt.Davraz (347 ± 17 ) does not agree with the overall vector mean aspect (015 ± 12.5confidence limits: 002 and 027 ) of the cirques in the Western Taurus.
They also calculated the LGM ELA of Mt.Davraz as 2233 m a.s.l, using moraine-based glacier reconstructions, independent of cirque floors' altitude.

| GIS and map producing
Mapping was carried out utilizing 1:25.000scale topographic maps provided by the General Directorate of Mapping (HGM), highresolution satellite imagery from sources such as Bing Maps, Google Earth, and HGM Küre (https://kure.harita.gov.tr/), as well as extensive fieldwork conducted during the summer of 2019 and autumn of 2020.During the field studies, oblique images were captured using a UAV in the study area.These images were subsequently utilized for interpreting the geomorphological features of the region.Digital elevation/surface models (DEMs/DSMs) with pixel resolutions of 30 m (SRTM 1 arc-second global, obtained from https://earthexplorer.usgs.gov/) and 5 m (generated from HGM stereo aerial photos) were utilized in the generated maps, depending on the scales of the maps.DSMs (with a 5-m pixel resolution) were utilized to generate crosssections illustrating structural features, such as faults and depressions.
To enhance the visualization of the topography, both a Hillshade and F I G U R E 1 Location maps of the study area: (a) location of the study area within Türkiye, (b) showing the study area and its immediate surroundings, along with their structural features, on the digital elevation model (DEM) and (c) a red relief image map (RRIM, Chiba et al., 2008) showing the location of the study area, Davraz hummocky moraine field, northern slopes of Mt.Davraz (C#: cirques names; NC: nivation cirque).
[Color figure can be viewed at wileyonlinelibrary.com] a Red relief image map (RRIM, Chiba et al., 2008) were generated from the DEM.We calculated the LGM ELA using a GIS toolbox developed by Pellitero et al. (2015).The Area-Altitude Balance Ratio (AABR) method was employed for the calculation of the LGM ELA on Mt.Davraz.The coordinate reference system chosen for the study area was WGS 84/UTM zone 36 N EPSG:32636.
T A B L E 1 Geographic sample locations, boulder dimensions, sample thicknesses, topographic shielding factors of the cosmogenic 36 Cl samples.Attributes and local corrections to production rates for sampled Davraz hummocky moraine boulders.

| Terrestrial cosmogenic nuclide (TCN) dating
The cosmic ray exposure ages of moraines in Mt.Davraz were determined by using in situ-produced cosmogenic 36 Cl.We collected six carbonate rock samples from the boulders of the hummocky moraine field (Table 1).While four of the samples were extracted from boulders located on crests near the external boundaries, the two last samples were collected on boulders from the inner part of the area (Figures 1c and 2a-f).The coordinates of sampled boulders were recorded using a handheld GPS device.In order to ensure optimal exposure to cosmic rays, preference was given to flat or gently sloping moraine crests.Horizon angles to be used in topographic shielding calculations were measured at each sampling site using a compass and a clinometer at 45 azimuthal intervals.A hammer and chisel were used to collect rock samples from boulder tops on the crests of the moraines (Figure 3).Our selection of boulders was determined by their size, position and post-depositional preservation.
The chemical preparations were followed by the procedures described by Sarıkaya (2009).Preparation of the samples was done at ITU-Kozmo-Lab (Istanbul Technical University, Türkiye) (https:// kozmo-lab.itu.edu.tr/en).The carbonate rock samples were crushed and ground to a fractional size of 0.25 mm to 1 mm before being analysed.Samples were leached with milli-Q and dilute nitric acid (10%) to remove meteoric chlorine and biogenic material.Then, natural chlorine in the rock matrix was liberated by dissolving leached rocks in HDPE bottles within the solution of 400 ml 10% HNO 3 .
Before complete dissolving, the materials were mixed with natural NaCl (Merc Emsure) and 35 Cl (99.7%) enhanced carrier (Aldrich).By adding AgNO 3 , chlorine was precipitated.By precipitating the samples twice as BaSO 4 , 36 S, and an isobar of 36 Cl, the rate in the detector can be kept low enough not to influence the identification of 36 Cl (Mechernich et al., 2019).The Australian Nuclear Science and Technology Organization (ANSTO, Sydney-Australia), analysed chlorine isotopic ratios of the samples, using its 6 MV SIRIUS Tandem Accelerator (Wilcken et al., 2017(Wilcken et al., , 2019)).
Average CaO concentrations are 54.95%,although K 2 O concentrations are very low and typically close to the detection limits (0.01%), except in Sample #2 (0.07%, still very low) (starting from this point, we will use "#" instead of the last two digits for sample identification; e.g., sample #1 instead of DVRZ19-01).The natural Cl concentrations are also very low, between 3.6 and 24.1 ppm.Thus, the main production mechanism for 36 Cl is from the spallation of 40 Ca.For all samples, it was assumed that the rock density was 2.6 g/cm 3 .The sample ages were determined using the CRONUS Web Calculator version 2.1 (https://cronus.cosmogenicnuclides.rocks/2.1/)(Marrero, Phillips, Borchers, et al., 2016).Corrections for thickness and topographic shielding were done for all production rate calculations.We also used a snow correction ($0.97) on spallation responses with 50 cm snow thickness on top of moraine boulders, assuming a snow cover period from December to April.We utilized the corrected 36 Cl production rates reported by Marrero et al. (2021) [55.9 ± 4.5 atoms 36 Cl (g Ca) À1 a À1 for Ca spallation, 153 ± 12 atoms 36 Cl (g K) À1 a À1 K spallation and 696 ± 196 fast neutrons (g air) À1 a À1 ].The Fe and Ti spallation production rates of 36 Cl are 1.9 atoms (g Fe) À1 a À1 and 3.8 atoms (g Ti) À1 a À1 , respectively (Marrero, Phillips, Caffee, & Gosse, 2016).The time-dependent Lifton-Sato-Dunai method (also called 'LSD' or 'SF' scaling) (Lifton et al., 2014), was used to scale 36 Cl production rates.
We calculated the ages with and without accounting for erosion.
We employed the erosion rates on carbonate rock rates of Mt.Landform ages were calculated using a weighted mean.The analytical and production rate uncertainties were included in the age calculations with a 1-sigma uncertainty for all ages.

| Glacial geomorphology
There are no active glaciers on Mt.Davraz today.However, the geomorphology of the mountain provides evidence of glaciation in the past.On the northern slopes of the mountain, there exists a relatively extensive moraine area in comparison to the overall coverage of Mt.Davraz, accompanied by successive cirques developed on the upper slopes.It can be asserted that the magnitude of these glaciations is greater on the northern slope than on the southern and eastern slopes of the mountain.The Davraz hummocky moraine field on the north and cirque-dominated slopes on the northern slopes of the mountain are the two main parts of glacial geomorphology in the study area (Figure 4).We calculated the ELA during the LGM period as 2,161 m a.s.l.For the Western Taurus region, including Mt. Davraz, the modern ELA depression has been indicated by Akçar (2022) to be between 1000 and 1400 m a.s.l.
Mid-latitude glaciated mountains often show signs of previous glaciation as cirque development.These sites were characterized by a particular topography that controlled the accumulation and movement of ice, with ice flowing from the back walls of cirques towards the margins of glaciers over short distances.Ardos (1977a) mentioned three large cirques and a few minor nivation hollows on Mt.Davraz.
However, he mentioned neither the formation of the hummocky T A B L E 2 Chemical compositions of the bulk rock samples after leaching with dilute nitric acid (10%).Analysis performed at the AcmeLab, Canada (Bureau Veritas Commodities Canada Ltd.) by ICP-ES (major elements), ICP-MS (trace element).Chemical compositions of the bulk rock samples after leaching with dilute nitric acid (10%  1c and 4).
Topographical restrictions prevented the cirques from forming proper glacial valleys.However, palaeoglaciers developed in these cirques have been able to form piedmont glaciers despite the absence of well-developed glacial valleys.Accordingly, it is possible to say that cirque glaciers directly started to accumulate on the slopes and then slow into the flat-lying area located in front of the northern slope of Mt.Davraz (Figure 1c).Hereby, we observed evidence of palaeopiedmont glaciers developed by these cirque glaciers during the LGM (c.26.5 to 19 ka).As the late LGM marked the onset of deglaciation, piedmont glaciers began to diminish and retreat.Throughout the retreat, spanning from the late LGM to the Late-glacial (c.19 to 11.7 ka; cf.Penck & Brückner, 1901/1909;Brazier et al., 1998;Böhlert et al., 2011;Shrestha & Aryal, 2011;Hagg, 2022) period, a substantial quantity of hummocky moraines was deposited on these relatively flat surfaces.As mentioned above, piedmont glaciers may also originated from ice caps extending on mountain tops, instead of being fed by cirque glaciers.Nonetheless, during our field surveys, we could not find evidence of ice cap formation on the top of Mt.Davraz.
Besides the northern Davraz hummocky moraine field, there is another hummocky moraine field located on the eastern side of the mountain, which we called the Northeast hummocky moraine field (Figure 1c).Presumably, it was formed by a single cirque glacier.
There, a part of this hummocky moraine field proximal to the slope was reworked by a rock glacier.The rock glacier was probably formed by frost-shattering processes on a well-developed talus slope.However, we preferred to exclude these two areas from this study and focused on the largest 'Davraz' hummocky moraine field to ensure the integrity of the research.Kıryayla sections (Figure 1c).In making this division, we have used some of the notable differences between these two sections.For instance, Kıryayla has a larger area than the Kundurpınarı section.Concerning that, the hummocky moraine field shows a larger area from the west (Kundurpınarı) to the east (Kıryayla).It also loses elevation, roughly 60 m, along the overall length of the field.At the eastern terminus part, hummocky moraines intertwined with karst morphology.The moraine boulder of both sections usually consists of poorly sorted limestone clasts in a fine-grained matrix with boulder-to-debris size (Figure 5).
We obtained three landform ages based on two scenarios: (1) a uniform glacier retreat with a single mean landform age (Figure 6a and Table 3); and (2) a gradual glacier retreat with two mean landform ages (Figure 6b,c and Table 3).We adopted the methodology of Griffin et al. (2022) to estimate the age of landforms, despite the larger uncertainty in the approach.However, we consider that larger uncertainties better represent the landform age, especially if there is a limited age dataset.

| Denudation/Erosion rates on carbonate rocks
Rates of denudation in karst landscapes, encompassing both chemical denudation and mechanical erosion processes (Ford and Williams, 2007), can exhibit substantial variations.The preservation of glacial deposits in glaciokarst landscapes is facilitated by efficient rainfall percolation, limited fluvial reworking, and the cementation of landforms (Woodward et al., 2008;Žebre and Stepišnik, 2015;Çiner et al., 2015).In a recent study, Allard et al. (2020) investigated the scattered ages of moraines within the Tymphi massif in the Pindus Mountains, northwest Greece.They correlated these findings with moraine degradation, attributing it to the erosion of limestone surfaces, which is more prevalent in glacio-karst landscapes compared to granitic or siliceous lithologies.This phenomenon arises from the presence of karst depressions and the solubility of limestone, as highlighted by Putkonen and Swanson (2003).Despite the limited number of samples collected from our study area, we believe that we have obtained well-clustered age results.Consequently, we do not anticipate any issues arising from moraine degradation.
In their erosion corrections applied to ages calculated using the 36 Cl isotope on carbonate rocks, Žebre et al. ( 2019) assumed an erosion rate of 40 mm ka À1 (with the upper limit calculated even at 60 mm ka À1 ).This choice was influenced by the fact that the study area is one of the regions (The Dinaric Mountains) with the highest rainfall in Europe.On the other hand, cosmogenic dating investigations focused on carbonate lithology in the Taurus Mountains, researchers have adopted an erosion rate of either 5 mm ka À1 (Sarıkaya et al., 2014;Çiner et al., 2015;Çiner and Sarıkaya, 2017;Köse et al., 2022) or 10 mm ka À1 (Sarıkaya et al., 2017;Çiner et al., 2017;Köse et al., 2019;Altınay et al., 2022).In a recent study by Hashemi et al. ( 2023), the long-term denudation rate of carbonate rocks in the Taurus Mountains was determined range from 1.92 ± 0.31 to 45.77 ± 3.91 mm ka À1 using the cosmogenic 36 Cl isotope.
In certain investigations, due regard was given to the depth of the solution grooves (cf., rinnenkarren; Bögli, 1980) when ascertaining the erosion rate.In certain investigations, careful consideration was given to the depth of the solution grooves when determining the erosion rate.
For instance, in studies conducted in the Taurus Mountains, solution grooves ranging from a few centimetres to 10-15 centimetres (Sarıkaya et al., 2017;Çiner et al., 2017) were inferred as 10 mm ka À1 erosion.On the other hand, in areas with solution grooves measuring several centimetres, the erosion rate was determined to be 40 mm ka À1 for Velež and Crvanj mountains and Blindje polje in Bosnia and Herzegovina ( Žebre et al., 2019; Çiner et al., 2019).In our study area, the maximum depth of solution grooves on moraine boulder surfaces is only a few centimetres, contrasting with other studies that assume a 10 mm ka À1 erosion rate.In karst terrains, extreme cold climatic conditions lead to a scarcity of liquid water, limiting dissolution and allowing other geomorphological processes to dominate morphological evolution (Ford and Williams, 2007).Given the interior location (i.e., transition to continental climate) of Mt.Davraz, the extreme cold may explain why we could not identify relatively deeper solution grooves on the sampled moraine boulders.Chert veins can also serve as indicators to estimate differential weathering rates on limestones.Allard et al. (2020) used the thicknesses of elevated chert veins in limestone to assess the varying erosion rates of glacial boulders in Greece.While they proposed cosmogenic 36 Cl ages corrected for erosion as low as 0.6 mm ka À1 , they acknowledge the possibility of higher rates (greater than 10 mm ka À1 ) within their study area in Greece.In another study conducted by Köse et al. (2022) in the eastern part of the Central Taurus Mountains, the erosion rate was determined as 3 mm ka À1 based on the chert vein thickness.This rate was evaluated alongside the erosion rate of 7 mm ka À1 obtained from a depth profile of an alluvial fan (Sarıkaya et al., 2015) near the region, and the erosion rate on the surface of the moraine blocks was accepted as 5 mm ka À1 .However, we were unable to locate any chert veins on moraine boulders in our study area.
Considering the previous cosmogenic 36 Cl dating studies and the published denudation rates by Hashemi et al., (2023) for the Taurus Mountains, we have decided to apply an erosion rate of 10 mm ka À1 for the ages obtained from the moraine blocks in the case of Mt.
Davraz.Consequently, the cosmogenic ages we obtained are consistent with the local LGM (21.5 to 18.5 ka), indicating that the erosion rate of 10 mm ka À1 is appropriate.

| Evolution and age interpretation of the hummocky moraine field
Based on our morphological observations, we are assuming that the hummocky moraine field preceding the glacier occupation was characterized by relatively flat terrain, resembling the plateau topography observed at the summit of Mt.Davraz (Figure 1c).Accordingly, the Kıryayla section appears to exhibit lower relief compared to the Kundurpınarı section after the disappearance of the Davraz palaeoglacier (Figure 7).We inferred that the more widely expanded palaeoglacier in the Kıryayla section is the direct cause of this difference.Because, as the glacier expands, its thickness tends to decrease, especially in the ablation area (Flint & Demorest, 1942).Correspondingly, thinner glaciers will have less debris to carry vertically, so the morphology of the hummocky moraines formed by ice wastage may be relatively low.On the other hand, the Kundurpınarı section has a higher relief.We believe that this can be attributed to the longitudinal structure of this section.As the palaeoglacier did not advance significantly in this area, it might have expanded without losing much of its maximum thickness.The debris carried by the glacier vertically might have formed higher moraines here because it was deposited in a relatively narrow region.
A multitude of factors, including the quantity of debris transported by ice, the location of the debris on, within or beneath the ice, the quantity of meltwater, and the resulting erosion and deposition, can lead to the formation of diverse landforms (such as hummocky moraines) as a consequence of disintegration (Gravenor & Kupsch, 1959).Hummocks have a chaotic appearance due to their random distribution and varying dimensions.Therefore, it may be challenging to understand deglaciation patterns in hummocky moraine fields.However, the morphology and TCN dating ages of the Davraz hummocky moraine field enable us to deliberate on the pattern of glacier retreat in this area.At this point, we believe that the relatively slight variations in the clustered ages are also connected to moraine stabilization resulting from the melting of dead-ice caused by ice disintegration.The distribution and morphology of moraines, combined with our cosmogenic ages, suggest that the deglaciation of the Davraz palaeoglacier began in the Kıryayla section and terminated in the middle of the Kundurpınarı section (Figure 8).We believe that the deglaciation in the area was led by the formation of four primary palaeoglacier patches due to ice stagnation-dominated deglaciation.
The palaeoglacier patches centred with samples #1, 2, 3 and 4  (Figure 8).However, the moraine morphology shows that the slopeproximal part of the glacier retreated into the slope and disappeared in the Kıryayla section (Figure 8b-d).In this respect, we do not consider there was an ice-stagnation-controlled melt-out occurred in this segment of the palaeoglacier.
In the first deglaciation scenario, we considered all ages to ascertain the optimal representation of the geomorphological age.After excluding the outlier, Sample #6 (53.0 ± 8.9 ka), we calculated the landform age of the Davraz hummocky moraine field as 19.3 ± 3.0 ka (Figure 6a).In a recent study by Tielidze et al. (2022) in the Southern Alps, similar results were obtained for the ages represented in these two scenarios on a terminal moraine unit with a hummocky character.
The landform age obtained from 11 boulders from the northern part of the above-mentioned moraine is 20.1 ± 0.4 ka, while on the We have two ages in the Kıryayla section as Sample #3 (22.7 ± 2.4 ka; 2060 m a.s.l.) and #8 (20.8 ± 1.9 ka; 2038 m a.s.l.).Sample #3, the oldest age, is located close to the terminus of the glacier.By examining the location of this sample and its surroundings, one can observe the presence of relatively high moraines (Figure 7).This area was among the first parts to separate from the palaeoglacier and commenced melting, following the onset of the stagnant ice period after the deglaciation of the Davraz palaeoglacier (Figure 8b,c).During the retreat and stagnation of the Davraz palaeoglacier, the southwest boundary of the area is most likely controlled by the fault that marks this boundary (Figure 1b).This fault, the so-called Holocene fault, extends from the southwest to the northeast direction in the slopeproximal part of the Davraz hummocky moraine field (Figures 4 and   9).The presence of a fault zone can boost karstification by creating pathways for groundwater seepage, thus expediting the dissolution process.In the slope-proximal part of the Kıryayla section (Figure 1a), the depression along with the Holocene fault could be linked to accelerated dissolution in this region (Figure 9).The dissolution can be attributed to the meltwaters (Bennett & Glasser, 2009;Maire, 1990;Smart, 1983Smart, , 2004) from the retreating paleoglacier on the slopes of Mt.Davraz.Cross-sections (Figure 9) reveal that the depression along the fault is likely associated with postglacial karstification.We posit that this depression evolved through a process analogous to the formation of solution dolines (cf.Ford & Williams, 2007).Furthermore, based on the topographic features observed, it is inferred that the area in question (southeast part of the Kıryayla section) served as the discharge point for the meltwater (Figure 7).This process might also affect the relief of the adjacent hummocky moraine fields (e.g., Salcher et al., 2010).As a consequence, the depression in this ± 1.8 ka (Sample #2) (Figure 8e).
As mentioned before, the Kundurpınarı section represents a different area of cover compared with the Kıryayla section.Glaciers covered with debris are composed of a core of glacier ice that is covered by a layer of surface debris.The layer of surface debris, which can be several meters thick, typically results from a combination of rockfall onto the glacier ice and/or sublimation of ice-containing debris (Milkovich et al., 2006).Accordingly, given the formation of the piedmont glaciers, they can be covered by thick debris and thus develop high-relief hummocky moraine topographies during the down-wasting of glacier ice.Based on the longitudinal extent and high hummocky moraine relief of the Kundurpınarı section, we are considering that the palaeoglacier was thicker in here (Figure 7).Besides, the dissolution along the Holocene fault seems not intense in the Kundurpınarı section as observed in the Kıryayla section.Here, the hummocky moraines have quite distinct morphology along with the Holocene fault.These moraines exhibit smaller dimensions and a more compact arrangement compared to other moraines.The distribution of the moraines observed in the area may reasonably be attributed to the deglaciation pattern (i.e., as the palaeoglacier retreats towards the slope) of the palaeoglacier.
Our findings may indicate that the ages obtained in the Kundurpınarı section are relatively young.We assume that deglaciation occurred in two phases here.Sample #1 (18.7 ± 1.8 ka; 2102 m a.s.l.), as a second palaeoglacier patch, is located on probably the first surface exposed in this section (Figure 8d).This area underwent disintegration from both the southwest and northeast, as well as from Samples #4 and #2, respectively.We believe that the intense glacial deposition in this area is due to its proximity to Cirque #1 (C1), the largest cirque on Mt.Davraz.The morphology of the moraines observed in Sample #2 and its immediate vicinity (Figure 7) suggests that this area corresponds to the third patch of the palaeoglacier.This patch, which retreated westward from the Kıryayla section, also appears to have retreated northeastward from the high-relief area where Sample #1 is situated (Figure 8d,e).Finally, the fourth palaeoglacier patch which consists of Sample #4 (17.1 ± 2.6 ka; 2083 m a.s.l.), probably melt-out synchronously with the third palaeoglacier patch (Figure 8e).This area also has the clearest boundaries of the hummocky moraine field with two terminal moraine-like loops that include Sample #4.On the other hand, Sample #6, considered an outlier, is situated approximately 250 m northwest of Sample #4 on an adjacent moraine loop.The age of Sample #6 is attributed to its potential detachment from the bed-rock through frost-shattering in its original position.Subsequently, it may have rolled onto the glacier, being transported supraglacially to its current location during the flow.Throughout this period, it could have been exposed to cosmic rays, leading to the potential accumulation of cosmogenic 36 Cl over tens of thousands of years.Ultimately, we consider that the average initiation timing of deglaciation in the Kundurpınarı section is approximately 17.7 ± 2.2 ka (weighted means of Samples #1, 2, and 4), occurring in the early Late-glacial period (Figure 8f).
In conclusion, based on morphological evidence and the obtained TCN ages, the palaeoglacier over the Davraz hummocky moraine field appears to have disappeared approximately from the early Late-glacial period (c.17 ka).However, some morphostratigraphic evidence suggests the presence of more recent glacial advances in this area.Terminal moraines mapped in front of Cirque 1 and 2 (Figure 4) can be considered as evidence for these advances.

| Wind-induced snow redistribution effect on glaciation
Temperature and precipitation represent the principal climatic parameters essential for the formation of glaciers.Nevertheless, glacier growth tends to occur on the leeward side of prevailing winds.(Sanders et al., 2013).Snow that is blown by the wind accumulates on the leeward side of crests without any depressions and in sheltered basins (Seppälä, 2004).It is also possible to form glacierets on the leeward slopes by wind-blown snow as well as snow avalanches (Benn & Evans, 2010).These observations highlight the phenomenon where windward slopes are frequently cleared of accumulated snow, and the turbulent conditions on these slopes intensify heat transfer and facilitate ablation processes (Evans, 1977).The aspect of the cir- However, significant differences can be observed between these two mountains.Mt.Dedegöl stands out with its considerably higher relief presenting an intense glaciation compared to Mt. Barla.On the other hand, the glacial geomorphology of Mt.Barla appears to be more marginal (Altınay et al., 2022;Ardos, 1977b).However, when compared to Mt. Davraz, these differences are even more pronounced.Topographic factors on Mt.Davraz have hindered the development of glacial valleys.
Instead, the glaciers that formed within the cirques accumulated and spread at the mountainside, giving rise to a piedmont-type glaciation.
Therefore, Mt.Davraz exhibited a predominance of glacial deposition characterized by hummocky moraines.Köse et al. (2019) reported that the primary source of precipitation in the current area on Mt.Dedegöl is the prevailing easterly winds.Additionally, they noted that these easterly winds played a significant role in the formation of cirque and valley glaciers during glacial periods.It can be regarded as one of the most heavily glaciated mountains in the Western Taurus region.The first dating study conducted in this area was by Zahno et al. (2009).They found evidence that the glaciers formed in three distinct periods in the W-E-oriented Muslu Valley: during the global LGM (24.3 ± 1.8 ka), late LGM (21.5 ± 1.5 ka), and in the Late-glacial period (15.2 ± 1.1 ka).This study also holds a place in the literature as the first and only cosmogenic 10 Be and 26 Al isotopes surface exposure dating in the Taurus Mountains.Despite the variations, the glacial histories of these three mountains largely overlap (Figure 11).For instance, summer insolation (Figure 11a) between $25 and 13 ka reveals a potential commonality among these three mountains.In this specific range, which fluctuates roughly from 460 to 530 w/m 2 , especially above 1800 m a.s.l., a monotonic relationship appears to exist between the insolation line and glacier retreat concerning elevation of the cosmogenic samples.
However, upon examining the NGRIP Ice core data (Figure 11b), the monotonic relationship terminates just before the Younger Dryas (at the onset of the Bølling-Allerød interstadial; 14.6-12.9ka, Rasmussen et al., 2014)  LGM-aged samples on Mt.Dedegöl suggests a more extensive glaciation in this region (Figure 11).In contrast, the mean elevation of the Late-glacial samples on Mt.Barla (Altınay et al., 2022) is 273 m below the LGM ELA, while the mean elevation of the samples on Mt.Davraz is 168 m below (96 m below the LGM ELA calculated with AABR method) the LGM ELA of the mountain.Therefore, it can be concluded that there was marginal glaciation in these two mountains.
The key advantage of Mt.Davraz among these three mountains is its positioning along the path of humid air masses originating from the west and south (Figure 10b).In contrast to Evans et al. (2021), we hold the view that the impact of moisture-bearing southerly winds on Mt.Barla is rather limited due to the location of Mt.Davraz.We believe that Mt.Davraz receives the majority of the precipitation by intercepting the air masses arriving from the south towards Mt.Barla.
Accordingly, we hold the belief that the primary factor contributing to the formation of glaciers on Mt.Barla is the influence of westerly winds.It is known that the palaeoglaciers on Mt.Dedegöl developed under the control of easterly winds (Evans et al., 2021;Köse et al., 2019).In this regard, Mt.Davraz holds a more advantageous position for glacier development, despite its relatively smaller scale, as it benefits from the convergence of humid air masses originating from both the south and west.

| Piedmont glaciations and hummocky moraines in the Mediterranean region
Piedmont glaciations are infrequent in the Mediterranean region during Quaternary glaciations (e.g., Çiner et al., 2015;Fernandes et al., 2017).Additionally, while hummocky moraines are observable on valley floors, the chances of encountering extensive areas of hummocky moraines, particularly those originating from ice sheets, are low.
Prior to this investigation, Geyikda g stood as the singular mountain in Türkiye known for hosting piedmont glaciers during Quaternary glaciations.It is situated 165 km northwest of Mt.Davraz.The palaeo-piedmont glaciers here ($25 km 2 ) were five times larger than those on Mt.Davraz ($4 km 2 ).Smaller glaciers, known for their higher sensitivity (Bahr et al., 1998;Huss & Fischer, 2016;Oerlemans & Fortuin, 1992;Zekollari et al., 2020), suggest a rapid glacial retreat on Mt.Davraz.We propose that the Davraz hummocky moraine field has likely been ice-free after the early Late-glacial (17.7 ± 2.2 ka).A thorough investigation conducted by Smith et al. (1997) Woodward et al. (2004) argue that the chronology in this study is not reliable and lacks internal consistency, thereby limiting its value as a source of palaeoclimatic information.On the other hand, relying on the ages derived from two moraine boulders (11.97 ± 0.9 ka and 12.06 ± 0.78 ka) in the hummocky moraine area within the Megala Kazania cirque on Mt.Olympus, Styllas et al. (2018) believe that these ages signify the conclusion of the second Late-glacial phase ($12.5 ka) of glacial activity on Mt.Olympus.They also identified hummocky moraine areas in the Pindus Mountains in Greece (Hughes et al., 2006).The U-series dating results obtained from cement formed in tills around hummocky moraines indicate that the minimum ages of the tills fall between MIS 5a interstadial and MIS 6.
Nevertheless, the formation of hummocky moraines in Greece is not related to either piedmont glaciations or ice-sheets, as seen in

CONFLICT OF INTEREST STATEMENT
No potential conflict of interest was reported by the author(s).
conducted on Geyikda g (2,877 m a.s.l., SW Türkiye), an important palaeo-piedmont glaciation site and vast hummocky moraine areas were dated by cosmogenic 36 Cl.They discovered that the ages of hummocky moraines originating from piedmont glaciers range from 16.0 ± 3.1 ka to 19.1 ± 3.4 ka based on 10 moraine boulders.The weighted average of all samples (excluding outliers) indicates that the earliest occurrence of glaciation took place approximately 18.0 ± 1.1 ka on Geyikda g.Within this study, we present a modest yet consistent glacial chronology of Mt.Davraz, while also providing an exploration of its comprehensive geomorphological features.Our analysis involved detailed mapping and topographical analyses of Mt.Davraz.We employed the terrestrial cosmogenic 36 Cl dating method to unveil the timing of the last deglaciation on Mt.Davraz.As our discussion unfolds, we delve into various dimensions, including assessing erosion rates on carbonate rock/erosion rates on TCN dating of carbonate rocks, interpreting deglaciation-age relationship and uncovering the development of the Davraz hummocky moraine field.We also examined the potential influence of wind-induced snow redistribution on glaciation.Finally, comparisons are drawn with glacial records in the nearby region, and intriguing insights are gained into piedmont glaciations and hummocky moraines within the broader Mediterranean context.

F
I G U R E 3 Moraine blocks sampled in the Davraz hummocky moraine field.Sampled surfaces are indicated by white arrows.[Color figure can be viewed at wileyonlinelibrary.com] The glacial geomorphology of Mt.Davraz underwent changes with the onset of postglacial modification.The relatively thick talus slopes that developed on the walls of the cirques are the most important of these changes in the geomorphology of Mt.Davraz.These slopes were developed by the combined impact of paraglacial and periglacial processes that dominated after the deglaciation.Talus slopes are generally seen on the steep slopes of the mountain apart from the cirque headwalls.Especially in the areas where these cirque headwalls are located, rock falls due to paraglacial and periglacial processes are common.Finally, limited fluvial processes have continued to be effective on the mountain.They caused the formation of gully channels and debris flows.Colluvial fans and deposits were formed in certain areas due to these processes.The Davraz hummocky moraine field covers a larger part of the northern slope of Mt.Davraz, as introduced above.It extends approximately 4 km from southwest to northeast, covering an area of about 4 km 2 .On the western margin of this area, moraines show slightly terminal moraine morphology (Figure1c).These moraines can be considered as the terminus part of the Davraz palaeoglaciers.Contrary to the western part, moraines on the eastern margin of the area are difficult to distinguish, especially from the karst morphology due to postglacial karstification and fluvial incision.Thus, we divided the area into two sections to a better definition of the hummocky field as Kundurpınarı and

F
I G U R E 4 The geomorphological map of the northern slopes of Mt.Davraz (C#: cirques names; NC: nivation cirque).[Color figure can be viewed at wileyonlinelibrary.com]Geomorphological mapping and oblique images (derived from UAV and Google Earth) show that the moraine crests vary in distribution and morphometry.While proximal moraine lobes have relatively small scales, distal moraines in the Kundurpınarı section are larger and well-developed.Towards the east, the Kundurpınarı hummocky moraine field becomes narrower, and then the Kıryayla section starts.Here, the slope-proximal area has a lower relief compared with the whole hummocky field.The northern and eastern boundary of this area is partly removed by fluvial processes.Based on debris flow developments, the northern slopes of the moraines intertwined with debris flow valley heads.Therefore, it has become difficult to determine the boundaries of the moraines in this area.

F
I G U R E 6 Diagrams depicting the distribution of the individual cosmogenic 36 Cl surface exposure ages (n = 5, outliers excluded).(a) the mean surface age was obtained by combining all ages calculated according to the first glacier retreat scenario; surface ages were calculated separately for the Kundurpınarı (b) and Kıryayla (c) sections, following the second glacier retreat scenario.[Color figure can be viewed at wileyonlinelibrary.com]F I G U R E 7 The Davraz hummocky moraine field is shown on the DSM with field sections, cosmogenic ages and cirques (top), while the graph (bottom) displays the southwest-northeast oriented profile across this field.[Color figure can be viewed at wileyonlinelibrary.com]

F
I G U R E 8 Speculative representation of the melt-out/retreat of the Davraz palaeoglacier, progressing from the east (Kıryayla section) to the west (Kundurpınarı section) with cosmogenic ages.[Color figure can be viewed at wileyonlinelibrary.com] southern part it is calculated as 19.5 ± 0.4 ka.However, since this landform is considered as a single continuous moraine, its formation is interpreted as occurring 19.8 ± 0.3 ka.In this regard, the Davraz hummocky moraine bears similarity to the first scenario in terms of the deglaciation process.Alternatively, our results suggest a two-phased deglaciation at 21.8 ± 2.4 ka and a more recent phase at 17.7 ± 2.2 ka (Figure6b,c) in the second scenario.The shared outcome of the scenarios is that the glaciers began their retreat during the global LGM (26.5 to 19 ka) and concluded it throughout the Late-glacial (c.19 to 11.7 ka).However, considering the morphological evidence and the obtained ages, we posit a two-phase deglaciation, with the second scenario appearing more consistent.Hence, we find it appropriate to conduct the discussion based on the second scenario.
area has become deeper.Currently, seasonal debris flows are still witnessed along this relevant route of flow.Sample #8 is indicative of the second phase of the Davraz palaeoglacier deglaciation in the Kıryayla section.It appears that the glacier retreated in two directions in this area -towards the west, in the direction of Sample #2 (17.2 ± 1.8 ka; 2101 m a.s.l.), and southsoutheastwards towards the slope.Consequently, this led to bidirectional deglaciation: (i) retreat in the direction of the slope and (ii) down-wastage due to ice disintegration at the central part of the Davraz palaeoglacier (Figure 8c,d).This bidirectional deglaciation also may have resulted in a high amount of meltwater discharge into the area, accelerating surface lowering through dissolution.Based on the average of samples #3 and 8, we can infer that deglaciation in this area began around by 21.8 ± 2.4 ka.Eventually, the palaeoglacier patches in the Kıryayla section presumably disappeared before 17.2

F
I G U R E 9 The graph (left) and digital surface model (DSM) (right) display the profiles taken nearly perpendicular to the Holocene fault that traverses between the base of the north slope of Mt.Davraz and the hummocky moraine field.The Holocene fault is indicated by the bold orange line on the DSM.[Color figure can be viewed at wileyonlinelibrary.com]

F
I G U R E 1 0 Wind influence on snow accumulation over palaeoglaciers of Mt.Davraz: (a) a schematic illustration depicting the influence of wind on the glacial formation of Mt.Davraz; (b) the spatial pattern of simulated precipitation (mm) in Türkiye during winter (DJF) based on a 10-year average (1990-2000).Additionally, displayed are the 10-year seasonal average of 850 hPa wind vectors (m/s) from NCEP/NCAR reanalysis on the left and the control simulation on the right (Bozkurt and Şen, 2011); (c) on the left, a wind rose diagram illustrating the yearly frequency and direction of wind blowing at the Davraz Ski Center weather station.On the right, pie charts indicate the percentage of wind direction frequency during winter (DJF) and march; (d) the columns represent the monthly average wind speed recorded by the Davraz ski center weather station.Below the graph, the average number of days with strong winds (10.8-13.8m/s, Beaufort 6) per month is shown in blue.Notably, the average wind speed and the average number of days with strong winds are presented for the winter (DJF) and March when SW-SE winds are prevalent.[Color figure can be viewed at wileyonlinelibrary.com] ques, which refers to the direction of cirque glacier development, provides valuable insights into the correlation between wind direction and glacier development at this point.When considering the influence of local wind patterns on Mt.Davraz, it is noteworthy that winds might be affected by the glacier developments on Mt.Davraz (Figure 10a).The geomorphology of Mt.Davraz indicates a notable influence of wind patterns on the formation and evolution of glaciers in the area.According to Evans et al. (2021), the presence of westerly winds (Figure 10b) and moisture-bearing southerly winds (Figure 10c) likely exerted an influence on the snowfall patterns experienced by Mt.Davraz during the Quaternary period.The cirques on Mt.Davraz collectively developed in the NNW-N direction.Consequently, it can be inferred that the development of these cirques is influenced by local winds predominantly originating from the SSE-S direction.By examining the current wind directions and frequencies (Figure 10c), as well as wind speeds (Figure 10d) between December and March on Mt.Davraz, we may gather insights into the role of wind for the Quaternary glaciations.Certainly, to assert such a F I G U R E 1 1 The cosmogenic ages of Mt.Davraz are shown together with the cosmogenic ages of the nearby Dedegöl and Barla mountains.(a) the mean daily insolation at 65 N on the summer solstice data from Laskar et al. (2011) and (b) the oxygen isotopic profile from the NGRIP ice core (North Greenland Ice Core Project members, 2004).The map at the top left displays the locations of Mt.Davraz, Mt.Dedegöl and Mt.Barla with their cosmogenic age locations.[Color figure can be viewed at wileyonlinelibrary.com]proposition, one must initially acknowledge the assumption that the palaeo-wind parameters are either identical or at least comparable to the present-day conditions.However, based on the relationship between the aspect of the cirques and the current wind directions, it is reasonable to conclude that the current and palaeo-wind directions are likely similar.In conclusion, we believe that winds exert a significant influence on the relatively small-scale yet intense palaeopiedmont glaciations on Mt.Davraz.4.4 | Glacial records in the nearby regionMt.Dedegöl (2,992 m a.s.l.) and Mt.Barla (2,798 m a.s.l.) are two mountains in close proximity to Mt. Davraz (Figure11).Mt.Dedegöl is located $50 km east of Mt.Davraz, while Mt.Barla, situated $30 km north of Mt.Davraz, stands as the northernmost peak in this Western Taurus mountains.Both mountains exhibit cirque-valley-type glaciation, which has played a significant role in shaping their overall geomorphology.
Çiner et al. (2015)  identified a Late Pleistocene ice cap on Geyikda g, attributing piedmont glaciers and hummocky moraine formation to it.In contrast, our study suggests that the piedmont glaciers on Mt.Davraz originated from cirque glaciers, given the absence of glacial evidence on the summit plateau.Satellite images resembling a hummocky moraine region contradicted our field observations, revealing a doline-uvala complex formed by karstification.Insufficient snow accumulation due to the windward position is believed to hinder glacier formation in this area (Figure10).

Mt
palaeo)piedmont glaciations and moraine formation are also documented in the Dinaric Mountains, situated in the central part of the Mediterranean region along the Adriatic Sea.According toÇiner et al. (2019), on Mt.Čvrsnica (2,226 m a.s.l.), piedmont glaciers advanced towards Blindje Polje, where they created hummocky, lateral, and terminal moraines.Ages derived from five hummocky moraine boulders from Svinjača area indicate that the glacial masses reached their maximum approximately 22.7 ± 3.8 ka years ago.This conclusion was drawn using the age of the oldest among the five moraine boulders.They also asserted that this result is spatiotemporally consistent with the lateral (13.2 ± 1.8 ka) and terminal (13.5 ± 1.8 ka) moraine ages within the polje.The onset of deglaciation on the Svinjača hummocky moraine area correlates with the deglaciation of the Davraz palaeoglacier.5 | CONCLUSIONThe identification of palaeo-piedmont glaciers, which give rise to the hummocky moraine field on the northern slope of the mountain, represents a significant contribution to the literature, particularly in the realm of Mediterranean glaciations.Such features are relatively uncommon in mid-latitudes, making their discovery on Mt.Davraz all the more noteworthy.Based on the ages derived from six cosmogenic samples taken from the Davraz hummocky moraine field, it was determined that the region experienced two distinct phases of glacial retreat.The first retreat of the Davraz palaeoglacier commenced approximately 21.8 ± 2.4 ka (late LGM) years ago in the eastern part (Kıryayla section).Subsequently, the western (Kundurpınarı section) part of the palaeoglacier retreat began approximately 17.7 ± 2.2 ka (early Lateglacial) years ago.Nevertheless, the terminal moraines in front of some cirques indicate morphostratigraphically younger glacial advances.Despite its relatively small scale, topographical effects around Mt. Davraz are considered to be decisive factors in creating favourable conditions for the development of piedmont glaciers.During this period, the influence of moist air masses originating from western and southern winds played a significant role in the glaciation of Mt.Davraz.Another important impact of the southerly winds was their ability to transport snow from the windward side to the leeward side of the slopes.We believe these winds directly contributed to the formation and development of cirque glaciers on the northern slopes.Mt.Barla and Mt.Dedegöl, as nearby mountains, do not share any similarities with Mt.Davraz in terms of glacial geomorphology.Nevertheless, their glacial chronologies are correlated during the global LGM (26.5 to 19 ka) and Late-glacial (c.19 to 11.7 ka).Mt.Davraz exhibits significant geomorphological and geochronological similarities to Geyikda g in Türkiye.However, these resemblances do not extend to piedmont glaciations and extensive hummocky moraine fields in the Mediterranean region.The remaining piedmont glaciations in the region are either old (e.g., pre-LGM) or haven not demonstrated the potential to form extensive moraine areas.This is likely attributed to the topographies of these areas which do not support the formation of an ice sheet by piedmont glaciers, as seen in Mt.Davraz.Overall, despite the challenge of dating hummocky moraines with carbonate lithology, our limited data evidence a consistent chronology both with global and local LGM.With a limited number of samples, general conclusions about the deglaciation of the study site could not made exclusively.The future studies in the region may focus on particularly in understanding karst processes of carbonate rocks on glaciated regions with a larger set of samples.However, we believe that the established chronology also enlightens and emphasizes how much the topography and climate can be effective on glaciations, even in small-scaled mountains.The rarity of the relationship between piedmont glaciations and hummocky moraine fields in a mid-latitude region like the Mediterranean serves as evidence for this assertion.acknowledge the financial support from the Australian Nuclear Science and Technology Organisation (#AP11366) through the National Collaborative Research Infrastructure Strategy (NCRIS).We extend our sincere appreciation to Editor Stuart Lane for his dedicated efforts throughout the review process.We are grateful to the first reviewer, Jamie Woodward, for providing insightful comments and suggestions.Additionally, we would like to express our gratitude to the anonymous reviewers for their thorough evaluation of our manuscript and their invaluable insights.Special thanks are also owed to Attila Çiner, Cengiz Yıldırım, Manja Žebre, Uroš Stepišnik, Aydo gan Avcıo glu, Khadijeh Hashemi, O guzhan Köse and Şeyma Yıldız Köse for their assistance during fieldwork.
based on the grouped ages.From this juncture onward, according to the existing chronology, neither Mt.Davraz nor Mt.Barla can be associated with any glacial retreat.