Evolution of karst and landscape of the Lessini Veronesi Mountains (Venetian Prealps, north-east Italy) from the Paleogene to the Present

The stratigraphic setting

The main Mesozoic and Cenozoic stratigraphic formations and units in the Veronese area (Fig. 2) are briefly described in stratigraphic succession with references to karst (Mietto & Sauro, 1989):

Fig. 2 A stratigraphic column from the Triassic to the Quaternary eras shows rock formations, lithology layers, a vertical volcanic intrusion, and a legend.

Dolomia Principale (upper Carnian-Rhaetian). This formation constitutes the rocky backbone of the Veronese reliefs. It consists of whitish and pinkish dolomite, sometimes yellowish, often in banks, in which bituminous limestone is interbedded. It has an outcrop thickness of approximately 500-600 m and presents a cyclic stratification. Morphologically, it gives rise to overhanging walls and is characterised by deep karst.

Calcari Grigi Group (Hettangian-Pliensbachian). It is approximately 450 m thick and is characterised by predominantly deep karstification. The unit is divided into four formations (Clari, 1975; 1981):

- “Monte Zugna formation” (previously called the “Lower Member”): made up of massive micritic limestones formed in subtidal environments, with layers of laminated limestones and green clays. Its total thickness is around 150 m;

- “Loppio ooolitic limestone” (formerly “Middle Member”): consists of stratified oolitic limestone with a thickness of about 60 m in the Val d’Adige that reduces eastwards to about 8-10 m in the upper Val d’Illasi;

- “Rotzo formation” (formerly “Upper Member”): this is the most representative unit of the group. The lithology is very variable and consists of limestones, marls and clays with a thickness of over 200 m. The upper part is easily recognisable due to the presence of “Lithiotis”;

- “Massone Oolitic limestone” (formerly “Upper Oolite”): this formation has the same characteristics as the Oolite di Loppio.

Oolite di San Vigilio (Toarcian p.p.-Aalenian). The formation (Beneke, 1886) consists mainly of oolitic limestones with abundant remains of echinoderms and crinoids, as well as ammonites. The thickness of the formation varies from about 100 m to a few metres. Karst features are frequent both at the surface and in the deep.

Rosso Ammonitico Veronese (lower Bajocian-Tithonian). The formation consists of very fine-grained nodular limestones, predominantly red in colour, and rich in ammonites. They are approximately 25-30 m thick. Morphologically, it is easily recognisable as it forms overhanging rocky ledges. Surface and deep karst phenomena are important.

Maiolica (upper Tithonian-Aptian p.p.). It is a thickly stratified, ivory-white limestone, rich in nodules and flint strips. Its thickness varies from just over 30 m to about 140 m. Morphologically, it gives rise to undulating meadows with surface karstification (dolines).

Scaglia Variegata Alpina (Aptian p.p.-Cenomanian). It consists of alternating layers of limestone as well as layers of thickly stratified marls of variable colour. It is not affected by karstification.

Scaglia Rossa (Turonian-Maastrichtian). It consists mainly of pink, reddish, and whitish limestones and marly limestones. It has a thickness of about 70-75 m within which there is a horizon of about 6.5 and 8.0 m (Prun stone or “lastame”) of nodular limestone affected by karstification.

“Nummulitic limestones” (lower and middle Eocene). They are approximately 150 m thick. These limestones are affected by karstification and are rich in foraminifera (nummulites, assilines, alveolins), corals, molluscs, etc.

Priabona Marls (upper Eocene). It occurs on the summit of hill ridges to the north of Verona. The unit is mostly greenish-grey or yellowish-grey marls rich in discoclines. Outcrop thickness does not exceed 80 m. Locally, the formation is covered, in discordance, by Miocene rocks. There are surface karst landforms.

“Incaffi limestone” (upper Oligocene). It is rich in foraminifera (Luciani, 1989) and has an outcrop thickness of approximately 90 m. It is not affected by karstification.

“Miocene limestones” (middle Miocene p.p.). The stratigraphic succession outcropping in the Verona hills is mainly composed of calcarenites and bioclastic calcarenites rich in pectinids and large melobesites (Fabiani, 1915). They are not affected by karstification.

In general, most Cenozoic “limestones” are skeletal grainstone-rudstone facies.

Over time, the platform was affected by various magmatic phases involving deep parts of the lithosphere. The volcanic eruptions of the eastern portion of the LVM are documented by approximately 400 m of volcanic rocks (Barbieri et al., 1991) with inclusions of peridotites (Cannatelli, 2012).

The succession of various sedimentary and volcanic rocks that make up the Veronese plateau has a thickness of almost 2,000 m.

The tectonic setting

Two main structural sectors can be distinguished: a northern and a southern tableland connected by an escarpment with SSW dip and inclination of approximately 40°, whose height is progressively reduced towards the east from 400 m at the Corno d’Aquilio to zero close to Scandole (Sauro & Ferrarese, 2016). The LVM forms a gently dipping monocline about 5° inclined southward. The current tableland situation is controlled by structure and must be considered a form inherited from the early stages of the uplift of the LVM (Castiglioni & Stegagno, 1935; Fuchs, 1969; Castiglioni et al., 1988).

The present tectonic-structural setting of the LVM results from a polyphase deformation of the southern Trento platform that started in the Early Jurassic with the opening of the Alpine neo-Tethys between Adria and Europe. The synsedimentary tectonics produced differential subsidence of platform sectors (Castellarin, 1972; Masetti et al., 2012). In the Late Cretaceous, the movement of tectonic plates changed direction, causing the area to shift from an extensional to a compressional setting before the collision. However, in the Lessini area, extension continued until the Paleogene and was linked to mafic and ultramafic magmatism (Piccoli, 1965, 1966; Zampieri, 1995). At that time, the Lessini area was the peripheral bulge of the uplifting Dinaric chain (Doglioni & Bosellini, 1987). In the eastern southern Alps, the shortening began in the Neogene, reactivating previous normal faults as strike-slip faults (Panizza et al., 1981; Zampieri, 1995, 2000). The western portion of the Lessini Mountains, and even more so the Mount Baldo chain, owes its origin to thrusts from the north-west (transpressional activity of Giudicarie faults), particularly active in the Late Miocene. From the Middle Miocene to the Quaternary, the Lessini were also incorporated in the foreland of the Apennines chain. The strong regional flexuring towards the Apennines chain produced up to 7000-8000 m thick foredeep depocenters close to the Apennines and the opposite uplift of the Lessini Mountains close to the southalpine chain (Fantoni & Franciosi, 2010; Pola et al., 2014).

The most significant tectonic deformations effects mainly the western margin (Mount Pastello ridge) and eastern margin (east of the Castelvero fault) of the LVM. Within the Veneto Region, this area is identified as the “Lessini-Schio District” and is affected by mainly NW-SE strike-slip faults (Zampieri et al., 2021). The historical seismic activity of the area includes two maior events: the distructive Verona earthquake of January 3, 1117 (Mw = about 6.5) and the Brescia event of December 25, 1222 (Mw = 6.05) (Sugan & Peruzza, 2011).

The morphological setting

During the Neogene, important events occurred that led to the emergence, at different times and in different ways, of large portions of the Venetian Prealps (Panizza et al., 1981), including the future LVM (Fig. 3B, 3C, 3D).

Fig. 3 Five tectonic evolution diagrams A to E from Oligocene to Quaternary showing faults and tilting strata, with a topographic map of Monti Lessini.

To the west of the investigation area, the youngest Miocene carbonates outcrop at Rocca di Garda and Mt. Moscal. In the Verona area, moving from west to east, several outcrops of Miocene limestones can be observed. At the Torricelle di Verona (Fig. 4E), the Miocene deposits rest on the underlying Priabona Marls along an erosional surface, which reveals a significant stratigraphic hiatus. This hiatus encompassed part of the upper Eocene, the entire Oligocene, and the lowermost Miocene (Sommaruga et al., 2020). Many authors have dealt with this unconformity. According to Fabiani (1915), the Verona hills had already emerged during the Oligocene, preventing sediments deposition of that age. Conversely, Castellarin & Farabegoli (1974) suggested that sedimentation occurred in the early to middle Oligocene, followed by erosion during late Oligocene-Early Miocene uplift events.

Fig. 4 A topographic map of the Lessini Veronesi mountain with seven detailed geological inset maps labeled A to G and a color-coded numerical legend.

Castiglioni et al. (1988) and Bartolini et al. (2004), identified relict flattened surfaces in the Venetian Prealps, including the LVM. They distinguished three types of relict landforms: a) flattened surfaces strongly controlled by geological structure; b) hilly landscape resulting from the dissection of hypothesised flattening surfaces; c) levelled surfaces with minimal structural influence.

In the case of the LVM, it can be hypothesised that the initial hydrographic network developed over the newly emerged land was “controlled” by tectonic activity, bedding orientation, lithology, and the distribution of lithostratigraphic units. Beginning in the Late Miocene, the drainage system evolved differently from the modern one, reflecting the impact of the local tectonics persisting until at least the end of the Neogene. A key geological event occurred in the Late Miocene, around 6 Myr ago, when the connection between the Atlantic Ocean and the Mediterranean Sea was interrupted leading to the formation of the “Lago-Mare” phase and, subsequently, the Messinian salinity crisis (Garcia-Castellanos & Villaseñor, 2011). By the end of the Messinian crisis, the main Alpine valleys were already deeply incised, while the LVM remained a broad limestone plateau with shallowly entrenched valleys. The plateau surface, underwent differential erosion depending on bedrock dip and litology. If the erosion processes had not been activated, the plateau would be several hundred metres higher. Gabert (1965) calculates the erosion of the Alps over the last 5 million years at around 950 m, while Pasa (1960) and Sauro (2010) estimated that the summit of Cima Carega (currently 2,259 m a.s.l.) would have exceeded 3,000 m. Again, according to Sauro (2010), Cima Trappola (currently 1,865 m a.s.l.) would have reached 2,500-2,600 m (Fig. 3C).

The ubstantial erosion is mainly attributed to glacial and periglacial exaration processes, enhanced by Messinian, Villafranchian, and Middle Pleistocene neotectonic phases (Sauro, 1982).

The Pleistocene landscape of the Lessini Mountains: glacialism and gravitational deposits

In our view, the processes responsible for repeated surface flattening in the LVM over the past 2.5 Myr can be primarily attributed to glacial and periglacial erosion, as well as widespread karst activity. According to Corrà (1993, 1997), flattening and other glacial morphologies are the result of the dismantling of glacial masses coming from the north in a condition of Alpine inlandsis. Castiglioni et al. (1988), on the other hand, report the recurrence of extensive glaciation during the Pleistocene. Finally, Busoni & Bondesan (2019), in describing Alpine glacialism, acknowledge the possibility of a large ice sheet (icefield) reaching the high plains and giving rise to piedmont glaciers. We fully support this hypothesis. These processes operate over broad areas, in contrast to fluvial actions, which tend to act in more localised and linear patterns. Consequently, it remains plausible that the icefield, which extended across the entire Alpine arc during its earliest phases, may also have reached the Venetian Pre-Alps and much of the LVM. According to estimates by Pasa (1960) and Sauro (1982, 2010), the elevation of the northern plateau has decreased by approximately 750 m over the last 5 Myr. Furthermore, Sauro (1982), discussing the relationship between karst morphogenesis and glacial-periglacial morphogenesis, states that in the areas repeatedly occupied by glaciers during the Pleistocene, glacial erosion and deposition processes were prevalent over karst processes. From the calculations reported by Sauro (2010), it appears that the superficial karst action on the Veronese relief accounts for approximately 25% of the average denudation of the mountain surface. Therefore, karst acts rather marginally. On the other hand, glacial action, with the contribution from fluvial processes, likely dominated both in direct erosion of the substrate and in total sediment volumes removed.

Glacial evidence in the Lessini Veronesi Mountains

Compared to those produced by the nearby Garda and Adige foothill glaciers, the glacial morphologies of the LVM, although smaller in size, are no less important for understanding the recent history of this portion of the territory. Hitherto, scholars had mostly limited themselves to attributing them to the last great glaciation, in the belief that traces of more ancient climatic crises had been erased during the last one. Only Pasa (1960) had sketched a written synthesis of MIS 12. In the same work, the author represented, on a special table, the glaciated areas of the Verona and Vicenza Lessini Mountains at the peak of MIS 2. In Fig. 9, we have reported the portion of the table from Pasa (1960) relating to the LVM, overlain the glacial areas hypothesised by Sauro (1973). Note the greater extent of ice shown by Pasa and the altitudes below 1200 m that it reached.

Fig. 9 A topographic geological map showing M. Castelberto and M. Tomba peaks with contour lines, purple and yellow formations, and a kilometer scale.

For decades, the only known morphologies for Pleistocene glaciation in the LVM were those related to the Mt. Tomba canopy, the Gasparine-Scortigara glacier (Figs. 4B-4C) and the alpine-type glacier that descended along the Vallon del Malera (Pasa, 1940; Pasa, 1960; Corrà, 1970; Sauro, 1973). It must be remembered that for the Würm glacial period, Sauro (1973) had deduced an orographic snow limit of 1,570 m a.s.l. on the south-facing slopes of the plateau and 1,450 m on those facing north. He observed, however, that a substantial ice supply, protection provided by cryoclastic deposits, and factors such as slope orientation could allow glacial fronts to extend below these altitudinal limits. In particular, he documented the descent of the Vallon del Malera glacier down to 1,270 m in Val Squaranto (or 1,200 m a.s.l., according to Pasa, 1960).

The studies of these two Authors also indicate that, during the last glaciation, two distinct types of glaciers were present on the plateau: an alpine-type glacier in the Vallon del Malera area, likely shaped over multiple glaciations and an ice cap that extended between Mt. Castelberto (1,765 m a.s.l.) and Mt. Tomba (1,777 m a.s.l.).

As far as the upper Valle d’Illasi is concerned, Pasa (1940) hypothesised that a glacial tongue about 5 km long descended from Mt. Carega to the Acqua Fresca fountain (1,046 m a.s.l.) and reached Selva di Progno (about 570 m a.s.l.) in the penultimate glacial stage. According to Sauro (1973), in the last glaciation, the glacial front stopped just upstream of Giazza, while reconstructions updated by other Authors (Bassetti & Borsato, 2007) for wider contexts attest to the ice further downstream of the village. Recent observations by M.S. and R.Z. confirm the latter interpretation at least as far as the foot of the present-day cemetery, where a glacial trough in the Dolomia Principale is observed.

The numerous hanging valleys in LVM, characterised by flat or rounded bottoms, deserve careful study, as they are evidence of more ancient and extensive glaciation. Particularly notable is the valley opening on the right-hand side at the top of the slope connecting the Revolto and Illasi valleys (Fig. 4C); it has a sub-rectangular layout with a major axis N-S oriented and extends between 1,225 and 1,526 m a.s.l. near Mt. Potteghe (Fig. 10).

Fig. 10 A mountain range view showing dense autumn-colored forested lower slopes, prominent rocky cliffs, and snow-dusted peaks against a blue sky.

Further south, still along the right side of the Illasi, at the foot of the Purga and Stolz mountains (Fig. 4A), lies a similar valley from which, at about 875 m, the Valle del Covolo descends. On the opposite side of the Valle d’Illasi, between the latitudes of Campofontana and Contrada Pernigotti, other valleys overhang from about 1,000 m a.s.l.; they drained a vast area occupied by an ancient and extensive glaciation (Sommaruga & Zorzin, 2020, 2023).

From the summit of the Lessini plateau, moving eastward, several similar landforms descend towards the Valle dei Ronchi. These include the larger depression of Sega d’Ala, located north of Mt. Cornetto; the smaller semicircular basin of Malga Revoltel, approximately 1.5 km in diameter; the cirque of Malga Scortigara di mezzo, which spans a couple of kilometres; and the Vajo delle Casare Gasparine (Fig. 4B), which differs in shape, being primarily elongated. Since they belong to the Valle dei Ronchi, during the last glaciation these basins must have hosted marginal lobes that fed the Ronchi glacial tongue (Fig. 11). At the western end of this alignment, remnants of glacial basins point southward: evidence includes the keeled profile with a NNW-SSE trend at the base of Mt. Cornetto and, more centrally, landforms consistent with ice flow along the wide “flexure” between Corno Mozzo and Busimo (cirques, U-shaped valleys with flat floors, moraine deposits). At the eastern end of this watershed, the upper Val Squaranto, which descends opposite the aforementioned Casare Gasparine valley, bears clearer evidence of an ancient transfluence, as indicated by pebbles of metamorphic rocks.

Fig. 11 A 3D Google Earth satellite view showing alpine terrain, labeled mountain peaks, valleys, a dashed path route, and a 2-kilometer scale bar.

Paleolandslides or seismites?

Morphologies referable to ancient landslides are described by Carraro (1965) and Carraro et al. (1970) in the Lessini Mountains. After analysing the structure, grain size, and bedding of these deposits, we suggest that at least some of them originated from gravitational movements likely linkedto slope decompression during the Lower and Middle Pleistocene glaciations. Such deposits have been identified at the sites of Moie and Le Colombare in Negrar (Corrà, 2006), as well as on Mt. San Giovanni, Gravazzo, Mt. Tesoro, Mt. Nuvola, Mt. Masua di Cerna, Mt. Comune, and Mt. Purga di Velo Veronese (Figs. 4D-4G), primarily along the southern and central ridges of the Mezzane, Squaranto and Valpantena valleys. A comprehensive study covering much of the central-western Lessinia (Sorbini et al., 1993) shows that the paleo-landslides predominantly affect lithologies from the Scaglia Rossa Formation and Nummulitic limestones. They are generally aligned parallel to the ridges and are found on gently inclined plateau-like morphologies at mid to high elevations. De Zanche et al. (1977) argue that, due to their location, these paleo-landslides cannot be attributed to fault tectonics but rather to gravitational sliding events that took place during Pleistocene climatic phases. We agree with this interpretation, associating these movements to slope decompression that occurred as ancient glacial tongues retreated. Cozzaglio (1933) documented breccias several tens of metres thick on the northern slopes of southern Mount Baldo near Caprino, attributing them to tectonic activity and thus defining them as “syntectonic breccias”. More recently, Sauro & Ferrarese (2016) identified extensive, stable breccia plates on the left slopes of the Anguilla, Falconi and Marciora valleys, as well as around Stallavena. Consistent with Cozzaglio’s interpretation, they classified these as seismites and definitively excluded both a “morpho-climatic” origin and a gravitational cause. They further proposed that even the paleo-landslides identified by Carraro et al. (1970) in the Lessini Mountains and in Valpolicella are seismites. Seismites are also invoked in the description of the Middle Palaeolithic site (Peresani et al., 2022) of Vajo Sansone (Valpantena) to describe the loose slope and fill deposits of the local karst shaft. In our opinion, these are not seismites, since this term, coined by Seilacher (1969) and very often misused, should be understood to mean sedimentary deposits and fine-grained structures, generally soft-plastic, deformed during an earthquake (Shanmugam, 2016; Müller et al., 2021). Seismites sensu stricto involving Quaternary deposits are unknown in the Lessini Mountains. Authors who applied the term improperly were, in fact, referring to earthquake-triggered landslides.

Sauro (2003, 2010), Sauro & Ferrarese (2016), Magaldi & Sauro (1978) identify a neotectonic genesis for the Orsara (Valpantena) and Scandole (Vajo dell’Anguilla) forms (slope deposits, scarp with top or base bluff, surface ruptures, earthquake trenches, gorge with steps, dolinas, bluish and greenish clay deposits). The particularly “fresh” appearance of the Orsara fault walls (surface faulting) is ascribed to the 1117 earthquake, while the Scandole trenches can be traced back to one or more violent earthquakes (surface ruptures).

The earthquake of 3 January 1117, which was widely reported in Europe in contemporary chronicles, is often “used” to justify the genesis of various morphologies and deposits in Lessinia. The epicentre of this earthquake, defined with a magnitude M=6.5, has more recently been identified to the east of the Verona area. According to Galadini et al. (2001), the seismogenic structure is probably located along the active south-verging thrust systems that delimit the Venetian Plain to the north.

Regarding the morphologies of Scandole (Fig. 4B), the hypothesis that their formation is solely due to gravity-driven movements is also excluded in this case, since the buildings within the affected area remain undamaged. Therefore, so-called “morphological fractures” must have existed prior to the construction of these buildings (Sauro & Zampieri, 2001; Sauro, 2003). According to these authors, if the slope were subject to gravitational movements, the buildings, probably dating back to the 16th century, would be affected by fractures and subsidence. Such an argument would apply only if movement were active, not in the case of a stabilised paleo-landslidelike Scandole. These forms can also be traced back to decompression phenomena on the valley slopes. Such loosening could derive from the local tectonics/fracturing in relation to the morphology/hydrogeology (landslide), but could also be a consequence of the retreat of ancient glacial tongues. Indeed, the strong cementation of debris sealing the trenches and the deep alteration/corrosion of the bordering rock walls do not suggest recent formation.

Ridge splitting, counter slope, trench, closed depression, etc., are forms present on the Orsara slope. Such morphologies (Arpav, 2009) were interpreted as Deep-Seated Gravitational Slope Deformation (DSGSD). These deformations are part of Slope Tectonics (Jabeyodoff et al., 2011). These are phenomena that can also affect extensive and large volumes of rock slopes whose genesis is mainly related to gravity (Carraro et al., 1979; Discenza & Esposito, 2021). The displacements of the rock mass towards the valley occur at a slow and steady rate (a few mm/y), while the temporal progression of these deformations can be influenced by multiple factors, including the geomechanical properties of the lithotypes, the structural configuration, tectonic activity, relief energy, and climatic factors such as glacial masses. The tensional release due to the loss of load as a consequence of the retreat of a glacial tongue seems to be the main cause of the most well-known stresses affecting the Alpine arc. The absence of support by the Pleistocene glacier causes detensioning of the slopes, leading to fracture systems parallel to the slopes.

Similarly, at Orsara and Scandole, within the “Il Castello-Corso” area located at the intersection of the Valpantena and the Vajo dei Falconi (Fig. 4D), directly opposite the Ponte di Veja, we observe trenches several metres wide (surface ruptures) partially filled with cemented detritus, karstic cavities developed along vertical fractures extending tens of metres deep, and evidence of paleo-landslides. In particular, the Corso district lies in a morphological situation similar to that of Orsara. Part of the paleo-landslide on which it lies is made up of voluminous blocks of Rosso Ammonitico Veronese from the above relief of Costamora, some of which have been deeply excavated to make artefacts. Slope deposits, composed of limestone blocks of highly variable size and breccias, have been correlated with tectonic phases from the Pliocene and Pleistocene epochs (Pasa, 1954; Carraro et al., 1969; Magaldi & Sauro, 1978). Magaldi & Sauro (1978), describing the Lughezzano, Lughezzano-Cà di Sotto, Lumini, and Naole sections, point out difficulties in interpreting the bluish and greenish laminated clays that, in their opinion, could refer to glacial barriers or tectonics. Magaldi et al. (1980) attributed loess deposits in Lughezzano, with Palaeolithic artifacts, to MIS 12.

It cannot be ruled out that the DSGSDs of Orsara, Corso, and Scandole derive from complex mechanisms, which include movements referable to “Lateral spread” (Jahn, 1964; Hutchinson, 1988) and “Block Slide”. These movements are triggered by the structural arrangement of the slope, detensioning due to the retreat of a Pleistocene glacial tongue and/or increased linear erosion with rapid deepening of the riverbed, as well as tectonic activity and, in some cases, shaking by strong earthquakes.

Regarding the extent of ice in Lessinia during the last glaciation, Fig. 8 allows, by superimposition, a comparison between the extensive glaciated area hypothesised by Pasa (1960) and the smaller area estimated by Sauro (1973). However, it cannot be excludedthat Pasa mapped the limits of an older glaciation and attributed them to the last one.

A brief return of interest in local glaciation was observed from the 1980s onwards (Ragnolini & Sauro, 1982; Sommaruga, 2009), and concerned some morphologies identified on the high ridge of Cima Lobbia (Fig. 4C).

Finally, in more recent years, as part of a research project promoted by the Natural History Museum of Verona and still in progress, important glacial morphologies were detected in the Upper Chiampo Valley, confirming the forms described by Pasa (Sommaruga & Zorzin, 2018, 2020b, 2023). Within this framework, the Campofontana ridge proved to be the ‘showcase’ of Lessini glaciation. Returning to the northern sector of the plateau, the Malera glacier and its surroundings present important reasons for interest. The first relates to the conspicuous deposits of its moraine apparatus and the marked glacial cirque features. The lack of an evident glacial cirque at its apex has led to the assumption that this glacier was once continuos with that of Cima Posta (Carega Group), but that the regressive erosion of the Valle dei Ronchi interrupted their connections (Corrà, 1970, 1976; Sauro, 2020, 2024). This hypothesis is difficult to support, because the current floor of the glacial cirque near the western threshold by Costa media (1,755 m a.s.l.) is at the same elevation as the present head of the Vallon del Malera, despite being 2.5 km distant. A similar situation occurs with the Campobrun glacial tongue, where the cirque floor at Malga Campobrun (1,650 m a.s.l.) lies at a lower elevation than the current head of Vallon del Malera. We therefore suggest that the source area feeding the Vallon glacier may have been located in the Carega-Lessini dislocation fold.

A dozen years ago, M.S. collected, downstream of Malga Campegno (about 1,460 m a.s.l.), on the hydrographic left of the Val Squaranto, three pebbles of metamorphic rocks, decimetric in size, abraded and partly altered, then lost. In 2022, M.S. and R.Z. collected some well-rounded quartzite pebbles, a few centimetres long, in the same vicinity, but at a lower altitude (Zorzin et al., 2022a). The discovery of these pebbles among the excavation materials of a marmot burrow, a long way from the local road, excludes an anthropogenic origin. Occurrence of quartzite pebbles in Val Squaranto can be attributed to a transfluence of the Athesian glacier and the attribution, at least to the Middle Pleistocene, of the oldest glacier in Vallon del Malera.

Karst deposits as records of Pleistocene glacial events

Several caves in the LVM have preserved fossil remains of animals and/or geological evidence related to surface processes within their sediments. One notable example is the “Buso della Fadanana” cave, which contains numerous exotic pebbles (Sauro & Zorzi, 2017). The cave, subject of archaeological excavations (Sauro et al., 2007), is located about 9 km south of Malga Campegno, at an elevation of 752 m a.s.l., on the right side of Val Squaranto and approximately 150 m above the valley floor. A fossiliferous breccia, supported by a reddish matrix, attributed to the Lower Pleistocene (Upper Biharian), contains numerous small pebbles (5-12 mm in diameter), many of which are exotic (Sauro & Zorzi, 2017). The authors interpreted these pebbles as “ecofacts” transported by birds of prey frequented the cave. However, based on the discovery of Athesian pebbles at the surface near Malga Campegno, the site of the “Buso della Fadanana” deposit - together with the presence of horizontal and vertical karst conduits above it and the lenticular stratification observed within the breccia, we consider these features more consistent with a glacial or fluvioglacial deposit preserved within a karst trap. Moreover, gastroliths are generally found in herbivorous birds rather than in raptors. Most clasts are either local or derived from lithologies absent from the present catchment area. Among these are Upper Cretaceous (Scaglia Rossa) and Cenozoic limestones, while micascists, gneisses, serpentines, and other exotic rocks are characteristic of Athesian provenance. The presence of clasts belonging to the Lessini stratigraphic succession, now missing due to erosion, may reflect transport by morphodynamic agents such as glaciers or watercourses. This is a very old deposit: no longer outcropping series must originally have been at least 200 m thick. Under conditions similar to those of today, karst corrosion acting on the surface of the Lessini Mountains would have lowered the bedrock by about 74 m during the Quaternary (Sauro, 1982, 2010, 2024). It is therefore plausible that roughly two-thirds of the total rock volume removedfrom the plateau over the last 2.5 million years resulted from other erosive agents, glaciers included. Since Pasa (1960) and Sauro (1982, 2010) estimate an average lowering of the northern plateau by approximately 150 m per million years over the last 5 million years, the deposit can reasonably be attributed to the Lower Pleistocene. The age inferred from the oxiferous breccia is compatible with that calculated from surface erosion rates. The “Buso della Fadanana” deposit would therefore represent the oldest glacial event so far documented on the plateau. Pasa (1960) had already reported the descent of a glacial tongue reaching Scardon, a locality a few hundred metres from the “Buso della Fadanana” (Fig. 4F). Near this cave, Salzani & Sauro (1986) described a conspicuous light-grey silt deposit (loess type), probably of glacial origin, inside the Buso del Beco, which opens at 770 m a.s.l. As part of a research project promoted by the Civic Museum of Natural History of Verona and still in progress, excavation campaigns were carried out between 2001 and 2008 (Zorzin & Bona, 2002) at the Grotta Inferiore dei Covoli di Velo (875 m a.s.l.) yelded radiocarbon dating on a phalanx of Ursus spelaeus (Rossi et al., 2018), providing an age of 29,130 ± 90 yr BP. This perfectly coincides with the climatic conditions during the advance of the ALGM (Alpine Last Glacial Maximum), just before a temperate-humid phase. A later sample from the same site, but from the older stratigraphic level, produced an age of 32,230 yr BP (Rossi et al., 2023).

In the LVM, there is also evidence of older glacial morphologies that are difficult to interpret because of alteration, particularly on carbonate and volcanic deposits.