Journal Pre-proof

Geomorphology of conglomerate terrains – Global overview

Piotr Migoń

PII: S0012-8252(20)30348-2
DOI: [Link]
Reference: EARTH 103302

To appear in: Earth-Science Reviews

Received date: 5 May 2020

Accepted date: 22 July 2020

Please cite this article as: P. Migoń, Geomorphology of conglomerate terrains – Global
overview, Earth-Science Reviews (2020), [Link]

This is a PDF file of an article that has undergone enhancements after acceptance, such
as the addition of a cover page and metadata, and formatting for readability, but it is
not yet the definitive version of record. This version will undergo additional copyediting,
typesetting and review before it is published in its final form, but we are providing this
version to give early visibility of the article. Please note that, during the production
process, errors may be discovered which could affect the content, and all legal disclaimers
that apply to the journal pertain.

© 2020 Published by Elsevier.

 Journal Pre-proof

GEOMORPHOLOGY OF CONGLOMERATE TERRAINS – GLOBAL OVERVIEW

Piotr Migoo

Institute of Geography and Regional Development, University of Wrocław, pl. Uniwersytecki 1, 50-137

Wrocław, Poland

Abstract. Conglomerates are widespread clastic rock, often occurring in thick successions and

of
supporting complex geomorphological landscapes. Yet, by contrast to some other rock types, little

ro
systematic work has been done on geomorphology of conglomeratic terrains, major geographical

gaps exist, and no review has been offered up to now. This paper partially fills this gap and provides
-p
an overview of characteristic landforms, landscape types and geologic controls present in areas built
re
of conglomerates. Distinctive landforms developed in conglomerates include bedrock cliffs and
lP

overhangs, domed hills and hilltops, crags, slots and clefts, tafoni, caves and various karst features.

They combine into specific landform assemblages of dome clusters, dissected plateaus and cuestas,
na

stepped hillslopes and conglomeratic badlands if lithification of sediment is poor. Some dissected
ur

plateaus and dome clusters represent spectacular morphology, awarded the status of UNESCO World

Heritage (China Danxia, Meteora, Kata Tjuta). A variety of geological controls is identified and
Jo

presented, including composition of conglomerate packages, their thickness, jointing patterns,

presence of sandy and clayey intercalations, as well as the style and magnitude of regional uplift.

Clast-by-clast breakdown is identified as a principal mechanism of rock disintegration, analogous to

grain-by-grain disintegration pattern of many sandstones. This explains striking similarities between

some sandstone and conglomeratic sceneries and the common presence of rounded shapes. In the

closing part of the paper research avenues for the future are suggested. Both regional studies, aimed

at the complete inventory of conglomeratic landforms, as well as detailed insights into the nature of

bedrock control at the variety of scales are recommended.

1
 Journal Pre-proof

Key words. Rock control in geomorphology; hillslopes; domes; karst; Danxia

1. Introduction

Rock-cut landscapes are fascinating even to people without any background and

understanding of geoscience. This is because of the scenic values these landscapes have. Steep cliffs,

isolated domes, towers and pinnacles, deep canyons, labyrinths of widened fractures, huge

of
overhangs, and curiously shaped rock residuals are common features of many erosional terrains in

ro
granites, limestones, dolomites, sandstones, or quartzites. Following the appreciation of the
-p
aesthetic value, many attempts to understand the origin of and controls behind such sceneries have

been made, but here a striking imbalance emerges. Whilst the morphology associated with some
re
rock types, such as granite or limestone, has been researched very frequently and a considerable
lP

amount of knowledge has accumulated (e.g., Twidale, 1982; Godard et al., 2001; Twidale and Vidal

Romaní, 2005; Migoo, 2006 for granite areas; Jennings, 1984; Gunn, 2004; Ford and Williams, 2007
na

for karst), landscapes developed upon some other lithologies have generated much less interest
ur

although they are no less distinctive. For decades, sandstone geomorphology remained under-

researched and the first comprehensive presentation of the variety of sandstone landforms in English
Jo

appeared as recently as in early 1990s (Young and Young, 1992). An updated and much enlarged

version (Young et al., 2009) remains an authoritative source of information and the research progress

in sandstone karst was recently summarized by Wray and Sauro (2017). Recent reviews by Migoo et

al. (2017) and Duszyoski et al. (2019) are also highly relevant to sandstone geomorphology, so that

much of the balance was restored.

However, the same cannot be said about landforms developed in the coarser members of

the clastic rock family – conglomerates. Here the deficit of information is striking and a big gap to fill

exists in rock geomorphology, especially regarding larger spatial scales. Conglomerates have most

2
 Journal Pre-proof

commonly been investigated for their sedimentological characteristics and a considerable amount of

research exists that explains the origin of coarse sediments and their small- to medium-scale

sedimentary structures (e.g., Koster and Steel, 1984; Blair and McPherson, 1994; Nichols, 2009). As

early as 1987 Ori and Roveri (1987, p. 845) commented that ’Coarse-grained sediments have

received considerable attention and significant progress has been made lately towards their

understanding’. More than 30 years later, there is still a long way to go before a similar statement

can be offered regarding the geomorphology of conglomerate terrains. The paucity of information

and the little attention paid to conglomerates in large-scale rock geomorphology is even more

of
surprising if one realizes that some landform assemblages supported by conglomerates are

ro
immensely scenic and have even gained the status of iconic geomorphological landscapes. The
-p
examples include Montserrat in north-eastern Spain, the towers of Meteora (Greece) – a UNESCO

World Heritage property since 1988, the range of localities in south-east China known under the
re
collective term of Danxia Landform, also inscribed on the UNESCO World Heritage List (since 2009),
lP

or the Kata Tjuta inselberg group in central Australia, similarly a World Heritage property (since

1987). Each of these localities is associated with significant cultural heritage showing that
na

geomorphic scenery in conglomerates was valued by people for many reasons, not only scientific.
ur

Given the above circumstances, this review is presented with the following main aims. First,

it is to highlight specific landforms and landscape types that develop upon conglomeratic
Jo

successions, including poorly lithified conglomerates of late Neogene and Quaternary age. Second,

geological controls (sensu lato) behind conglomeratic landforms and processes involved in their

shaping are discussed. Third, rock-cut landscapes in conglomerates are compared with those in

sandstones, and similarities and differences between the two are examined. Finally, research

avenues for the future are indicated, so that our understanding of conglomeratic landscapes may

increase.

Consequently, the following logic of presentation is adopted in this paper. Conglomerate as a

distinctive type of sedimentary rock is presented first. Then characteristic landforms developed in

3
 Journal Pre-proof

conglomerate are discussed, including slope and hilltop shapes, residual bedrock outcrops (tors,

crags), some concave linear features and finally, caverns and other karst phenomena. After that, the

spatial scale is extended and specific landscapes are given attention, ranging from dissected

highlands through clusters of residual hills to strongly eroded badlands. The next part of the paper

seeks explanation of this diversity of landforms and landscapes and addresses various instances of

geological control, from the role of regional uplift to specific features of lithological composition of

conglomerates. In the last but one part similarities and differences between conglomerate and

sandstone landscapes are discussed. The closing part summarizes the state-of-the-art and indicates

of
research needs.

ro
One note is necessary here. Minor interest in conglomerate geomorphology until now is
-p
reflected in a limited number of authoritative publications on the subject, especially such which go

beyond the mere description of landforms. Balanced geographical coverage is yet to be achieved (Fig.
re
1). Much of potentially important information is only available as field reports, travel logs, notes
lP

offered in passim, not as properly documented, reviewed and referenced papers. This shortage of

sources impacts on the nature of this review which cannot use large volumes of accumulated
na

knowledge. Perhaps unusually, significant use is made of Google Earth images. For many areas they
ur

are the only available background materials to examine basic features of conglomerate landforms,

their dimensions and spatial patterns. Nevertheless and having all these caveats in mind, it is
Jo

believed that this review shows both where we are and how much needs to be done to explain

striking conglomeratic landforms.

2. Conglomerate as a distinctive sedimentary rock

Conglomerate is a term that embraces sedimentary rocks composed dominantly of a mixture

of pre-existing rock fragments or clasts of virtually any rock type that have a diameter greater than 2

mm (gravel). ‘True’ conglomerates display clasts that have been transported significant distance and

4
 Journal Pre-proof

so are generally rounded in form. The term can also include breccias where the individual clasts are

very close to their source and retain a more angular shape.

Folk (1954) and Folk et al. (1970) believed that the use of the term conglomerate was only

warranted when at least 80% of clasts were greater than 2 mm in size (Moncrieff, 1989). If the

proportion of gravel was less than 80%, but more than 30%, terms such as sandy-conglomerate or

muddy-conglomerate could be applied (Folk, 1954). Collinson and Thompson (1982), in contrast,

suggested that at least 30% of a rock by volume needs be composed of large (2 mm or larger) clasts

before it should be termed a conglomerate or breccia. Clasts of different sizes have their specific

of
names such as granules (2–4 mm), pebbles (4–16 mm), cobbles (16–256 mm) and boulders (>256

ro
mm) (Nichols, 2009). A conglomeratic rock may represent a range of clast sizes or be composed of
-p
one dominant size, which means it is well sorted. Sorting is related to the sedimentary processes

involved. Fluvial processes typically result in better sorting than that achieved from alluvial fan
re
deposition by sheetfloods, whereas debris flows produce very poorly sorted deposits. Clasts in
lP

conglomerates may represent different lithologies, both considered as very resistant (quartz,

rhyolite, flint) and rather weak (schist) and/or soluble (limestone fragments), reflecting geology of
na

the source area. Similarly to the sorting property, one lithology may dominate among clasts
ur

(monomictic conglomerates) or clasts of various rocks may be present (polymictic conglomerates).

The clasts of conglomerates are typically held together by a matrix of finer material
Jo

composed of sand, silt, clay, often bound by some mineral cement. These cements may be composed

of carbonate, silica, iron or other minerals and may be macrocrystalline to cryptocrystalline (Bamlett

1990). Matrix and cement type can have important effects on rock properties. For example,

carbonate-cemented conglomerates are prone to karstic dissolution processes, whereas those

cemented by silica or iron may be much less susceptible to chemical weathering. Conglomerate

matrix material may have been deposited at the same time as the larger clasts, in which case the

rock would be likely to be matrix-supported (Bamlett 1990), with the clasts suspended in matrix and

rarely touching one another. If the finer matrix material entered the spaces between clasts after

5
 Journal Pre-proof

deposition, either being washed in or due to precipitation of carbonate, silica or other minerals from

groundwater, the clasts would be in contact with one another and the material would be clast-

supported. These two different forms, matrix- versus clast-support, can impart quite different

geomechanical properties upon the rock.

Typical sedimentary structures in conglomerates are planar or slightly inclined stratifications,

bedding planes, normal or inverse grading, gravel imbrication and the occurrence of lenses of

particularly large gravels. However, many conglomerates are massive and non-graded, with no clear

boundary planes discernible over many metres of sediment thickness (Boggs, 2009).

of
ro
3. Characteristic landforms in conglomerates
-p
3.1. Cliffs and overhangs
re
lP

Very steep to vertical slopes are a common and repetitive landform unit in many

conglomerate terrains. The height of conglomerate cliffs varies from a mere few metres to imposing
na

precipices exceeding 300 m in a single drop to the footslope (Fig. 2). High cliffs may rise directly from
ur

the surrounding plain or gently rolling relief, as in Longhushan, China (Kusky et al., 2010; Peng, 2020),

or at Agüero, Spain (Guerrero et al., 2014), from a high, but non-conglomeratic pedestal, as in
Jo

Jianglangshan, China (Zhu et al., 2009), or be distinctive facets within a slope, as in the Ceahlau

Mountains, Romania, where vertical cliffs 100–150 m high developed in conglomerates (Oprea et al.,

2012).

In detail, cliff morphology can be variable and geomorphic diversity of a cliff face is related to

the complexity of architecture of the respective conglomerate body. It is most diverse where

sandstone and mudstone/claystone intercalations occur within a conglomerate package as reported

from the Danxia landscapes in China (Peng et al., 2015a; Peng, 2020). These are often exploited by

weathering more easily than massive conglomerates and can be seen inside horizontal niches and

6
 Journal Pre-proof

overhangs (see section 3.5). In other instances, as in Belogradchik, Bulgaria, wavy cliff surfaces rather

than distinct shelters form, reflecting alternations between coarser and finer conglomerate units.

However, overhangs do not require finer inset deposits to form (Tronkov and Sinnyovsky, 2012). In

the Meteora, Greece, horizontal slots into the wall follow the bases of individual sedimentary

packages of Gilbert-type deltas. Permeability contrasts between adjacent beds seem the main cause

of preferential weathering and the origin of shelters.

The sheer height of many conglomerate cliffs indicates a considerable compressive and

tensile strength of the material, able to withstand both the load of the overlying strata as well as

of
tension imposed by the lack of lateral support. Nevertheless, in circumstances as described above

ro
one might expect abundant evidence for massive rock slope failures or signs of accumulating strain
-p
leading eventually to rock failures. This seems rarely the case and the high conglomerate cliffs in

Danxia landscapes in China show little surface evidence of instability in the form of opened clefts or
re
outward tilted compartments. However, the example of the November 2009 event in Langshan (Yan
lP

et al., 2016) shows that stresses can be released suddenly and the whole-slope collapses may occur.

Post-event analysis revealed that preferential weathering of a siltstone layer at the base of the
na

collapsed rock pillar changed the state of stress, leading first to two minor detachments and then to
ur

the disintegration of the entire 40 x 50 m block. The runout distance of the resulting rock avalanche

was more than 200 m (Fig. 3). Likewise, evidence of past whole-slope failures is present, for instance
Jo

in Longhushan (Fig. 4). Here, the talus rubble occupies c. 110 x 100 m at the river bank and it is

possible that fluvial undercutting caused the collapse of the shoulder of a conglomeratic dome.

The geomorphic impact of lithological variations across the cliff face does not follow one rule.

In thick conglomerate successions with subordinate sandstones and mudstones, it is the latter which

are the weak units. Hence, shelters and recesses form in rocks finer than conglomerates. However, in

sedimentary successions dominated by sandstones it may be the thin insets of fine-grained sandy

conglomerates which are eroded at a faster rate and slope recesses are associated with these (Fig. 5).

7
 Journal Pre-proof

Reasons apparently reside in the elevated porosity of coarser, poorly lithified units which promotes

weathering by a range of available mechanisms.

3.2. Rounded tops and towers

Although cliffs are so pervasive in some conglomerate terrains, top parts of bare elevations

built of well-lithified conglomerates are often strikingly rounded. They assume the shape of a dome,

with very steep slopes in the lower and middle sections, gradually declining in angle while going

of
upwards, until a convex hilltop is attained (Fig. 6a). Rounded tops stand in evident contrast to

ro
erosional landscapes in Quaternary gravel deposits (conglomeratic badlands), dominated by narrow

fins, pinnacles and hoodoos (Fig. 6b).

-p
Despite the fact that the theme of rounded hilltops in conglomerates is rarely explicitly
re
addressed in geomorphological literature (see section 5.5), photographs in various publications show
lP

that domed outlines of the upper parts of residual hills are fairly common. Examples can be cited

from various geographical and climatic settings, including Mediterranean environments of southern
na

Europe in Spain (Nichols, 2004, 2005; Guerrero et al., 2014) and Greece (Ori and Roveri, 1987), humid
ur

subtropical south-east China (Peng, 2020) and monsoonal Republic of Korea (Kim, 2008), as well as

semi-arid interiors of Australia (Ollier and Tuddenham, 1961; Twidale, 2010). From this limited
Jo

database some specific features of hilltop rounding may be identified. This phenomenon may be

observed at residual hills of various height, from minor pillars to large domes in excess of 300 m high.

In terms of area, minor domes may be c. 100 x 100 m, whereas the largest rounded compartments

are elongated ridges up to 2 km long and 0.5 km wide (Fig. 7). In Jiangslangshan, the erosional

landscape consists of three peaks of 270–370 m relative height, separated by deep and narrow slots

(Zhu et al., 2009). Although their sides are nearly vertical, despite cutting across different facies of

the conglomeratic sequence (Chen and Guo, 2017), their summits are gently rounded, with almost

flat sections on the very top (Fig. 8).

8
 Journal Pre-proof

3.3. Tors and related features

While many landscapes cut in conglomerates show the presence of high continuous cliffs,

including those delineating domed hills, other morphologies with exposed bedrock can form too.

There are many reports about landscapes typified by the presence of tors. ‘Tor’ is the term derived

from ancient Welsh and means a ‘pile of boulders’. Although used informally prior to the 1950s, it

was the paper by Linton (1955) on granite residuals in Dartmoor, SW England, through which tors

of
entered geomorphological literature and vocabulary. Tors have been defined as ‘solid rock outcrops

ro
as big as a house rising abruptly from the smooth and gentle slopes of a rounded summit or broadly
-p
convex ridge’ (Linton, 1955, p. 470) and may be considered synonymous with castle koppies, which is

another term widely used in rock geomorphology, particularly for granite terrains (Twidale, 1981).
re
Following the definition, tors are from a few to about 20 m high and residuals of this size are
lP

common in some conglomerate areas. However, if bedrock outcrops emerge from steep slopes, the

term ‘crag’ is more appropriate.

na

One such area is the Central Iberian Chain, northwest of the city of Teruel, where Triassic
ur

conglomerates crop out (Gutierrez et al., 2005; Lozano et al., 2007). Around the Peracense Castle

tens of conglomerate tors occur, rising from the gentle slopes or the steep front of an escarpment
Jo

(Fig. 9). They vary both in shape and height, from less than 5 m to more than 15 m high towers

isolated from the escarpment face. Conglomerates occur as thick lenses within a variegated clastic

sequence and behave as a more resistant rock type, supporting protruding beds and tor upper

surfaces.

Conglomerate crags are present in the low mountain area of the Sudetes, in Central Europe

(Czech Republic, Poland), where kilometre-thick sedimentary successions of Carboniferous and

Permian age occur. The former was deposited in warm and humid climates, whereas the latter

contains typical red beds from continental deserts and semi-deserts, but each locally supports crags.

9
 Journal Pre-proof

In both cases conglomerates occur alongside finer materials, from sandstones to claystones.

Outcrops are from 3–4 to 20 m high and typically subordinate to valley sides and secondary ridges,

within which they form clusters of steps, spurs and towers (Pilous, 1990). The distribution of crags is

clearly controlled by regional structure. Valley sides following the dip of strata have very few and low

bedrock outcrops, whereas the opposite slopes host the most impressive examples. If the dip is close

to vertical due to subsequent deformation, conglomerate tors take the form of narrow fins along an

axis of the ridge. Pasterskie Skały (Shepherd’s Rocks, Fig. 10a) in SW Poland (Bartuś and Mastej,

2013) and crags of Babí Lom near Brno in SE Czechia (Dvořák et al., 1975) (Fig. 10b) provide

of
examples. Conglomerate tors have been also reported from the Bucegi Mountains in the Romanian

ro
Carpathians, where they rise from the mountain crest line and reach 12 m high (Oprea et al., 2012).
-p
The descriptive definition of tors makes it difficult to agree on the lower and upper size limits

for tors. Transitions to both smaller and larger features occur. The former include low (<2 m)
re
pinnacles built of more cemented conglomerates protruding from gentle slopes (Coronato, pers.
lP

comm.; Fig. 11), whereas the latter are shaped as multiple cones and crags rising above crest lines in

mountainous settings (Fig. 12a) or as >20 m high isolated rocky hills overlooking flat erosional
na

surfaces as in the Cueva del Milodon protected area in southernmost Chile (Fig. 12b). At the latter
ur

site one can also see how rock spurs become fragmented into a dozen or so residuals next to one

another, to form a small example of ruiniform relief (sensu Migoo et al., 2017). Furthermore,
Jo

transitions from isolated towers and pinnacles to more continuous conglomerate cliffs occur (Pilous,

1990).

3.4. Clefts, slot canyons and related features

Deep and narrow linear incisions are common in many conglomerate terrains (Fig. 13).

Although they are quite clearly related to structural discontinuities with the rock mass (fractures,

faults), they are not necessarily all of the same origin. Both weathering and fluvial processes may

10
 Journal Pre-proof

play a role, as well as mass movements leading to joint opening. Here, following Young et al. (2009),

the term ‘slot canyon’ applies to features shaped primarily by erosion of running water and lined by

vertical, locally overhanging rock cliffs. If the incision is wider and resembles the letter ‘V’ in cross-

section, the term ‘ravine’ is used. As in better researched sandstone terrains, fluvial origin of slot

canyons in conglomerates is assumed on the basis of several lines of evidence. Most convincing are

typical fluvial bedrock erosional landforms such as potholes, chutes, scalloped sidewalls and rounded

waterfall steps. They are interspersed with gravel bars and fans in sections of decreased gradient and

wider than upstream. It seems that little systematic research has been done on slot canyons in

of
conglomerates, but scarce information (e.g. Pilous, 1990; Burlando et al., 2011) and images available

ro
in various Internet resources show undoubted fluvial features (e.g.,
-p
[Link] access

date 2020-01-06).
re
‘Clefts’ in turn are linear slots originated due to non-fluvial processes, chiefly weathering
lP

focused on zones of higher fracture density. They may be less than 1 m wide and barely passable (Fig.

14), but more than 10 m wide in other cases, whether in relation to the width of jointed zone or time
na

elapsed since the onset of weathering. Their floors may be horizontal or steeply sloping and little
ur

squares occur at intersections of clefts of different trend, so that small rock cities are formed (Pilous,

1990). The depth of clefts varies too, from a few metres as within conglomerate cliffs in Czechia to
Jo

large-scale features reported from north-eastern Spain (Bergada et al., 1997; Guerrero et al., 2014)

and Danxia landscapes of China (Peng, 2020). Serra da Capivara in NE Brazil offers excellent examples

of co-existence of linear features of various dimensions and origin, from joint-aligned slots barely

incised into the plateau surface, through a range of intermediate landforms to deep flat-floored

valleys (Fig. 15). Likewise, a spectrum of linear incisions of various size has been reported from the

Mallos landscapes of northern Spain (Guerrero et al., 2014).

Together, these linear features may form complex patterns. At Kata Tjuta they separate

individual domes which form the inselberg cluster, mostly along W–E trend, with subordinate N–S

11
 Journal Pre-proof

and SW–NE directions, especially in the eastern part (Fig. 7). They vary in width and whereas the very

narrow ones are due to preferential weathering of more fractured zones, transitions to wider,

fluvially-shaped valleys can be observed, especially along N–S fractures. At a larger scale, the grid-like

pattern of corridors of different width can be observed at several China Danxia localities in southern

China, including Langshan and Taining (Fig. 16). However, not all ravines and canyons necessarily

follow straight courses. Local scale sinuosity, as seen within slot canyons, is related to the

hydrodynamics of the flow (Carter and Anderson, 2006), whereas the large-scale variant likely results

from rounding of corners at intersections of differently oriented reaches of an initial valley.

of
ro
3.5. Caves and other karst features
-p
The occurrence of caves has been reported from many conglomerate terrains, although
re
information is often limited to ground plan description and simple maps, with little explanation
lP

provided about the origin and temporal context of formation. The very term ‘cave’ is used in various

ways, referring to both extensive and complex networks of underground passages as well as minor
na

overhangs within rock slopes, fully exposed to daylight. The latter are particularly common in
ur

erosional landscapes developed from well lithified conglomerates, intercalated with rather thin

mudstone and claystone layers. Pilous (1990) provided several examples from Permian
Jo

conglomerates of northern Czechia. Large overhangs, referred to as bedding caves, were described

from Danxia landscapes of China (Zhu et al., 2010; Peng et al., 2015a). They may be tens of metres

long, up to 10 m high at the opening and penetrate into the rock slopes for more than 10 m (Fig.

17a). Some are wide and high enough to accommodate temples and monasteries, which are quite a

common sight in China. In Fangyan, Zhejiang province, the development of caves from the opposite

sides of a hill led eventually to their connection through the rock mass (Fig. 17b). It is suggested that

the process of differential weathering starts along weak fine-grained layers and as the initial

horizontal slot enlarges, roof collapses in the overlying conglomerate play an increasing role, so that

12
 Journal Pre-proof

the cave appears to be formed entirely within conglomerate. Water emerging at the

conglomerate/mudstone contact enhances weathering at the rear of the cave, helping differential

cave wall retreat. The enlargement process is hypothesized to culminate in the whole rock slope

collapse, being thus an important component of cliff retreat. The geometry of bedding caves is

clearly controlled by the attitude of strata. If rock layering is horizontal, bedding caves are horizontal

too. If a rock mass is tilted, then they follow inclined bedding planes as in Longhushan in China (Ren,

2013; Guo et al., 2018). Similar effects are achieved if major conjugate sets of compressive fractures

occur across rock faces, as is the case in Langshan, China. Then deeply incised overhangs run

of
diagonally across the rock walls. Liu and Li (2018) attempted a classification of caves in clastic

ro
erosional landscapes of China, looking at both morphology and origin. Whereas the former is
-p
problematic and hardly considers scale and the change of form through time, the latter clearly

distinguishes caves formed by weathering, water erosion which may produce both horizontal caves
re
and vertical slots, and gravity-induced collapses. Karstic origin was apparently not taken as another
lP

possibility.

Perhaps some of the biggest caves belonging to the category of simple overhangs are those
na

at Cueva del Milodon site in Chile, known for findings of prehistoric animals from the times of
ur

Pleistocene/Holocene transition. There are three caves within the property, with the Cueva Grande

extending for 200 m into the rock mass, being up to 80 m wide and 30 m high at the opening (Fig.
Jo

18). The floor is covered by huge blocks of conglomerate resulting from ceiling collapses. The origin

of caves is poorly known and seems not to be reported in scientific papers. Popular science

explanation offered at the locality proposes abrasion by waves of a proglacial lake, noticing

altitudinal consistence of the cave opening and the reconstructed strandlines. However, the

presence of stalagmite stumps and flowstones in the nearby Cueva Chica suggests the role of karstic

processes as well, focused on carbonate components in the conglomerate. The size of the caves do

not seem to be readily compatible with abrasion on the shore of a short-lived proglacial lake.

13
 Journal Pre-proof

Proper caves have been reported from various localities, including northern Italy (Ferrarese

and Sauro, 2005), Slovenia (Lipar and Ferk, 2011), Northern Alps in Germany (Goeppert et al., 2011),

and Catalonia in Spain (Bergada et al., 1997). Less systematic are observations from outside Europe,

sometimes restricted to basic descriptions (e.g., Piccini et al., 2009; Dunkley et al., 2017), but they

are nevertheless valuable as enlarging the general picture. The study of Lipar and Ferk (2011) from

the Udin Boršt plateau, underlain by early or mid-Pleistocene conglomerates, originally an outwash

fan, contains detailed inventory of 15 caves, among which the longest one is 1154 m long and 16 m

deep. The majority of them is drained by an active stream, hence termed as ‘linear stream caves’,

of
whereas other varieties include shelter caves from surface weathering, breakdown caves due to

ro
collapses and vadose shafts. Caves of the Udin Boršt plateau are presented as examples of eogenetic
-p
caves, whose formation is concurrent with diagenesis of the host rock. The touristy Gupteshwar Cave

in Pokhara, Nepal, developed in the late Quaternary coarse fill of the Pokhara Basin (Gachok
re
conglomerates) and drained by a large stream (Fig. 19), is another example. However, caves
lP

developed in well-lithified pre-Quaternary conglomerates cannot be considered as eogenetic. The

distinction into caves of horizontal and vertical extension was made by Bergada et al. (1997) in
na

conglomerate landscapes of Catalonia. Some of these caves attain large dimensions, with Cuberes
ur

Cave being nearly 13 km long and 327 m in vertical range. Another important cave in the area is Font

Major cave near the town of l’Esplunga. It consists of largely a single conduit that extends for 3.6 km,
Jo

draining a conglomerate aquifer located at the base of the cliff-bound highland. Castel Sotterra Cave

in the Montello karst in north-eastern Italy is c. 10 km long (Sauro, pers. comm.). Some caves in

conglomerates are highly complex in plan, such as Tham Din Phieng cave in eastern Thailand

(Dunkley et al., 2017). It represents a maze cave, developed in a thin (c. 2 m) conglomerate interbed

between two harder and less permeable beds. In all these cases the conglomerates involved consist

mostly of limestone pebbles and cement is calcareous, so that karstic origin is implied.

Large caves in non-carbonate conglomerates, but still considered to be karstic, have been

rarely reported. Auler (2012) mentioned Lapão Cave which is a 1 km long cave partly cut through

14
 Journal Pre-proof

conglomerate beds within a sedimentary sequence in Chapada Diamantina tableland in Bahia, NE

Brazil.

Surface karst features have also been reported from various localities. Ferrarese and Sauro

(2005) described karst of the Montello Hill in the Alpine foreland of NE Italy and claimed it as

‘classical karst in conglomerate’. The most characteristic features are more than 2,000 dolines

scattered across an area of c. 60 km2, some merging into larger uvalas. Noticing topographic control

on the spatial distribution of dolines they suggested that diffuse porosity is a more important factor

in the development of dolines than fracture network. The presence of dolines, many acting as

of
swallow holes, was reported by Goeppert et al. (2011), who also used the term polje for flat-floored

ro
depressions, acknowledging their hydrological functioning despite their small size (~1 ha). Dolines
-p
were also mentioned from the Udin Boršt plateau (Lipar and Ferk, 2011) and Kalaw conglomerate

karst in Myanmar (Piccini et al., 2009), whereas a comprehensive geophysical and morphometric
re
study of dolines from Pleistocene conglomeratic terraces of different ages in north Slovenia was
lP

offered by Čeru et al. (2017). On the other hand, poor development of surface karst was noted in

conglomerates massifs of Catalonia, with dolines and lapiez notably absent (Bergada et al., 1997). It
na

was suggested that efficient fracture-guided dissection of conglomerate plateaus into isolated
ur

massifs meant loss of potential to create a fully developed karst scenery.

Jo

3.6. Minor weathering landforms

The very nature of thick conglomeratic successions, with alternating layers of different

lithological composition, including sandstone and claystone intercalations, dictates that their

outcrops are subject to differential weathering. Some have analogues in other lithologies, others are

notably missing, whereas certain weathering patterns seem peculiar to conglomerates. However, it

seems that little systematic work on weathering phenomena in coarse clastic rocks has been done.

15
 Journal Pre-proof

Based on observations in the ruiniform relief around Belogradchik in Bulgaria, three surface

weathering ‘morphotypes’ have been distinguished, namely (a) massive smooth columns and walls,

(b) columns with wavy walls, typified by alternating horizontal grooves and bulges, and (c) platy

shapes (Tronkov and Sinnyovsky, 2012). This diversity is attributed to the variable scope for the rock-

controlled selective weathering. Massive, poorly bedded conglomerates tend to weather rather

uniformly and hence, little differentiation of their outer walls takes place. By contrast, lithological

contrasts and the presence of bedding planes focus weathering which exploits less resistant units

and discontinuities. Nevertheless, rock walls are nearly vertical and it was observed that the tops of

of
the residuals are rounded if bedding is poor, or more planar if bedding planes are better developed.

ro
Scientific reports indicating the presence of tafoni in well-lithified conglomerates are few,
-p
although these weathering features seem to be quite common and some may attain huge

dimensions (e.g., Bourne and Twidale, 2003; Oh et al., 2017; Viljoen and Cawthra, 2019). Work from
re
the Korean Peninsula (Shin et al., 2015; Oh et al., 2017) indicates that tafoni are preferentially
lP

associated with southern aspect which enhances the efficacy of mechanical and chemical

weathering. In addition, their horizontal extension follows bedding planes and it is assumed that
na

large forms develop through the coalescence of smaller ones. Tafoni are common in conglomeratic
ur

Danxia landscapes of south-east China and take a variety of shapes, as the summary by Peng (2020)

shows, but systematic research is very scarce. Generally, they seem to follow the same pattern as in
Jo

Korea. Many are horizontal and there is gradation towards larger and more open recesses (bedding

caves), others follow the dip of bedding planes and boundary surfaces. Guo et al. (2018) have

recently offered a more thorough study of tafoni at Longhushan and made three important

contributions. First, they envisage the initiation of tafoni in vertical cliffs by means of fall out of

individual pebbles or their packages. The sediment itself contains loosely packed pebble lags and

these pebbles are poorly supported by one another and easily removed. Then positive feedback

mechanism is implied, as the precursory cavern becomes subject to temperature and humidity

changes on a daily basis. In this way caverns grow and may coalesce to form larger hollows. Second,

16
 Journal Pre-proof

they noted that tafoni, although extended horizontally as singular features, are arranged in vertical

strips if the whole rock wall is considered (Fig. 20). They maintain that this is not accidental but

related to the role of water flowing down the rock walls along certain pathways. The term ‘beaded

tafoni’ was introduced to account for this pattern. Third, whereas the development of tafoni is

usually considered solely in terms of weathering processes, Guo et al. (2018) suggest that the role of

running water is also important. Heavy rainfall results in sufficiently high discharges from the micro-

catchments above the cliffs to initiate waterfalls, channelized along the furrows. At the same time,

wavy rock surfaces disturb the flow and induce turbulence which increases erosional potential within

of
already weakened bedrock. Thus, a three-stage evolution of ‘beaded’ tafoni is proposed, consisting

ro
of: (a) formation of an initial hollow due to selective weathering, (b) deepening by water flow and
-p
origin of tiered ‘beaded’ pattern, (c) lateral growth and coalescence, with sub-horizontal bedding

taking the role of a major control.

re
Other primarily weathering features present in conglomerate terrains include wide karren
lP

(Fig. 21), widened vertical slots which may evolve into rock gates and arches (Fig. 12a), irregular

shallow pans and arcades, ledges and benches (Pilous, 1990; Peng, 2020). Hoodoos (mushroom
na

rocks) are part of ruiniform relief formed on conglomerates, owing their origin to non-uniform
ur

cementation within the succession (Fig. 22). In Bucegi Mountains, however, hoodoos formed within

mixed sandstone – conglomeratic sequences, whereby more porous conglomerates behaved as

Jo

weaker units and they are exposed within the stems (Oprea et al., 2012).

By no means ‘minor’, one of the highest natural rock arches in the world, with an estimated

height of more than 300 m, the Shipton’s Arch in Xinjiang, China, appears to have developed in a

coarse conglomeratic succession ([Link]

accessed 2020-04-11). Likewise, the now collapsed large arch on Legzira Beach near Sidi Ifni in

Morocco was shaped in conglomerates. Here, weathering was certainly assisted by marine abrasion

which eventually led to the major cliff failure and arch disappearance in September 2016. However,

17
 Journal Pre-proof

another arch still exists at this location ([Link]

accessed 2020-04-11).

4. Characteristic conglomerate landscapes

4.1. Dome clusters

In various areas an ever-increasing dissection of an initial highland is hypothesized as a

of
pathway leading to the development of characteristic conglomeratic landscapes within which high

ro
and bare domes, circular or elongated in ground plan, are separated by narrow linear valleys. Such an
-p
evolutionary sequence, from a recently uplifted plateau to the senile stage with scattered residual

hills, was proposed as a conceptual framework to explain the diversity of Danxia landscapes in south-
re
east China, specifically within the UNESCO World Heritage properties (Peng et al., 2015b; Peng,
lP

2020).

Domes are the most characteristic elementary landforms and may occur in a variety of sizes,
na

from as little as a few tens of metres in diameter to impressive monoliths c. 2 km long. Likewise, their
ur

heights vary from 10–20 m to more than 300 m. Broad summit accordance was noted at several

localities in China, lending support to the hypothesis that domes evolved through dissection of an old
Jo

surface of low relief. Dissection is focused on vertical discontinuities which are deepened and

widened by both weathering and fluvial erosional processes. If only the former is involved, the

process is slow and narrow clefts and avenues originate. However, they may attain considerable

depth such as the one at Jianglangshan, which is 308 m long, 298 m high and 3.5–4.3 wide in the

bottom part (Zhu et al., 2009). Fluvial processes, both episodic and perennial, help to enlarge the

initial slots, turning them into wider ravines and then flat floored valleys (Fig. 23). The spatial pattern

of dissection mimics the structural pattern of discontinuities, although as valley widening continues,

precise location of lines of weakness become more and more problematic.

18
 Journal Pre-proof

Within this type of conglomerate terrains, several sub-types may be distinguished, depending

on the distance between individual domed hills. Taining represents limited dissection, whereas in

Langshan fragmentation of the conglomerate belt is more advanced and valleys are wider, locally

with alluvial flat floors (Fig. 16 vs. Fig. 23). Pedras Negras in Angola show similar dissection and

degradation of conglomerate blocks within a 12.5 km x 5 km remnant upland (Fig. 24), which rises c.

150–200 m above the surrounding plain. In Danxiashan, many separate clusters of bedrock

elevations can be observed, with minor basins in between, apparently developed from lateral

enlargement of joint intersections. The erosional landscape of Meteora in Greece illustrates several

of
phases of dissection and individualization of domes, with partial emergence in the north, where

ro
spurs and ravines at the plateau edge may be observed, to almost complete isolation in the south
-p
(Fig. 25). However, even in the south the domes are closely spaced and no outlying hills are present.

Domes may be symmetrical, if bedding is nearly horizontal or asymmetrical, if large-scale tilting

re
occurred. With an increasing dip of bedding planes domes and towers are replaced by asymmetric
lP

ridges of cuesta type (Peng et al., 2015b; Peng, 2020).

Whereas it is tempting to follow this classic sequential scenario of landscape development

na

from an initial uplift to the end-plain (Yan et al., 2015), it may be too simplistic. For example, Chen
ur

and Guo (2017) urged to give more credit to facies diversification within thick conglomeratic

successions and argued that monolithic towers will be present in very coarse conglomerates whereas
Jo

the surrounding areas of low relief may in fact be developed upon finer-grained members of basin

infill.

4.2. Inselbergs

A clear-cut boundary line between dome clusters and inselberg landscapes is impossible to

draw and there will be always transitions in space in time. Nevertheless, in several places evident

inselbergs are built of conglomerates. Kata Tjuta in central Australia is the most cited example (Ollier

19
 Journal Pre-proof

and Tuddenham, 1961; Twidale, 2010), although it is not a singular monolithic dome but represents

complex morphology of about 30 domes of various dimensions, separated by narrow clefts or wider

valleys. The largest compartments are elongated and measure approximately 2000 x 300 m and 1800

x 400 m, and there is a range of intermediate and small size domes, ranging from c. 600 x 400 m to

125 x 125 m. The height of the domes varies from less than 100 m to c. 450 m. The inselberg massif

itself may be divided into sub-areas, with a more compact western part, where narrow slots and

avenues separate the monoliths, and the eastern part, where domes form minor clusters and are

separated by low whaleback outcrops or alluvial-colluvial depositional surfaces (Fig. 7). However, it is

of
also true that Kata Tjuta massif as a whole stands in considerable isolation from other hills. The

ro
nearest ones are nearly 25 km towards ESE (Uluru) or more than 30 km towards NW (Mount Currie).
-p
4.3. Plateaus, cuesta escarpments and dissected backslopes
re
lP

Thick and laterally extensive packages of conglomerates may retain horizontal attitude of

bedding and underlie plateaus or, if they are gently tilted, support cuesta morphology consisting of
na

gentle slopes broadly concordant with the dip of 5–25° (backslopes) and steep frontal escarpments
ur

(cuesta faces). Unless the plateau is strongly dissected (see section 4.2), bedrock outcrops within the

plateau itself are few and generally limited to marginal escarpments and valley shoulders cut into the
Jo

plateau. Serra da Capivara National Park in north-east Brazil provides an example (Fig. 26).

In cuesta landscapes conglomerates usually belong to stronger units within sedimentary

successions and hence, act as caprock on cuesta faces and/or support mid-slope benches. In both

settings, they may support crags and even small rock cities emerging from the steep slope sections

(Pilous, 1990). Whereas backslopes of the conglomerate cuesta described by Pilous (1990) have

rather gentle relief and a few crags only occur within valley sides, a very different picture comes from

the landscape around the town of Belogradchik. The area is famous for distinctive conglomeratic

landforms which developed through the dissection of a backslope of a cuesta. The sediments, laid

20
 Journal Pre-proof

down in an alluvial fan environment, are of Lower Triassic age (Tronkov, 1998; Tronkov and

Sinnyovsky, 2012). The entire outcrop area of conglomerates is c. 20 km long and up to 3 km wide,

but the most scenic part just south of the town extends over approximately 3 x 3 km. The thickness

of the clastic succession is several hundred meters and is dominated by coarse and poorly bedded

conglomerates in the lower part, whereas towards the top the proportion of sandstones increases

and bedding becomes more prominent. The backslope is incised by several south-trending valleys,

with the depth of incision up to 150 m. In some valleys high conglomerate walls line the valley floors

(Fig. 27a), but generally bare bedrock outcrops can be found on interfluves, either as solitary

of
pinnacles and towers (Fig. 27b), or more extensive massifs, akin to mesas and buttes, dissected

ro
themselves along widened joints (Fig. 27c). Locally they turn into large assemblages of ruiniform
-p
relief, as within the old Kaleto fortress and its immediate vicinity (Fig. 27d). However, further west of

Belogradchik the style of dissection is different and narrow parallel fins with no intervening valleys
re
occur (Fig. 27e). They are up to 500 m long and have gaps along its length.
lP

4.4. Stepped and terraced hillslopes

na
ur

In sedimentary successions consisting of alternating hard and soft strata stepped hillslopes

typically develop, with steeper sections supported by the former. They are associated, among others,
Jo

with molasses, where cemented conglomerates are cliff formers. These cliffs may be continuous over

kilometres-long distances or rather localized, depending on the architecture of the sedimentary

package. Likewise, their height is variable, reflecting the thickness of individual conglomerate units.

Lavaux area in western Switzerland, protected as UNESCO World Heritage for its wine cultural

landscape, is an example of this kind of topography which underpinned land use and the

development of wine culture (Reynard and Estoppey, 2020). The succession consists of distal

sandstones and proximal conglomerates, about 1000 m of total thickness, with conglomerates

accounting for approximately half of it, but alternations are multiple. Hence, slope morphology

21
 Journal Pre-proof

consists of gentler sections in sandstones (although still fairly steep, 15–30°) separated by

conglomerate cliffs which are 5 to 15 m high. Sandstone treads were found very suitable for vine

planting, whereas natural conglomerate banks, locally strengthened to improve slope stability, are

the risers. Further north, terraced hillslopes surround concentrically the entire massif of Mount

Pelerin, again distinctly influencing land use, with pastures and meadows on sandstone treads and

forest belts on conglomeratic risers. Less spectacular landscapes of this kind can be found around

cities of Bern and Luzern further east, all associated with the Alpine molasses series (Reynard, pers.

comm.)

of
ro
4.5. Badlands
-p
Poorly lithified and non-lithified conglomerates, generally of Quaternary age, may give rise to
re
badlands. Badlands are defined as deeply dissected erosional landscapes consisting of numerous rills
lP

and gully systems, dominated by overland flow erosion (Harvey, 2004). Although this type of

erosional topography is most common in soft, erodible deposits such as clay, marl or shale,
na

numerous examples can be also found in thick gravel deposits. Furthermore, whereas badlands are
ur

most frequent in climates typified by seasonal drought, conglomerate badlands are present in humid

climates too.
Jo

Conglomerate badland localities include, but are certainly not limited to, several places in

central and southern Taiwan (Lin and Su, 2019), Putangirua Pinnacles east of Wellington (north

Island) and Clay Cliffs near Omarama (South Island) in New Zealand, Djavolja Varoš in southern Serbia

and Têt Valley in the eastern Pyrenees in France. Satellite images of these localities show that in

contrast to more extensive mudstone badlands they do not extend over wide areas, but are

restricted to singular valley systems (Fig. 28). For example, the main part of Putangirua Pinnacles is

600 m long and up to 300 m wide, covering about 14 ha on both sides of one minor valley, with the

most eroded and bare section limited to 6 ha. Local relief is up to 80 m, with singular vertical cliffs up

22
 Journal Pre-proof

to 30 m. Characteristic landforms include side gullies and ravines, fluted conglomeratic walls, isolated

hoodoos capped by remnants of more cemented layers, and gravel fans at the outlet of individual

ravines (Fig. 6b). In this particular case, conglomerates are not Quaternary but date back to the

Neogene, c. 7–8 Ma (Begg and Johnston, 2000).

At Shihba Luohanshan (Eighteen Buddha Arhans) locality in the Maolin Scenic Area in

southern Taiwan one can observe transitional landforms between strongly eroded badland and

clusters of peaking and domed residual hills, with characteristic relief consisting of steep slopes and

rounded tops, typical for morphology of well-lithified conglomerates (see section 4.1) (Fig. 29).

of
Although no detailed studies of this locality are available, it is reasonable to suspect that

ro
cementation of the conglomeratic package plays an important part.
-p
However, the development of badlands is by no means the norm in poorly lithified

conglomerates. For example, regionally extensive and thick conglomerate sheets in South Patagonia
re
of late Neogene to Pleistocene age (Martínez et al., 2009; Martínez and Kutschker, 2011) support
lP

almost featureless plains. Their marginal escarpments are scalloped by numerous valley heads, with

occasional cliffs, but badlands appear absent.

na
ur

5. Geological controls and relief-forming processes

Jo

In geomorphology, two overarching controls on landform evolution are usually invoked. One

is related to geology, covering various aspects, from regional uplift to lithological details and

mechanical properties (Yatsu, 1966; Suzuki, 2004; Young et al., 2009). Another one considers

environmental characteristics, emphasizing climate at a variety of temporal scales, but vegetation

also plays an increasingly recognized role (Tricart and Cailleux, 1972; Goudie, 2004; Viles, 2011).

Unfortunately, such a balanced approach to the geomorphology of conglomeratic terrains is not

possible due to very patchy and uneven geographical coverage in relevant publications. Moreover,

mechanisms of processes involved in landform origin are rarely explored in detail, especially using

23
 Journal Pre-proof

background climatic data. Hence, any generalizations regarding climatic control will be premature

and not sufficiently based on empirical evidence. In this section, therefore, only geological controls

are addressed.

5.1. Lithology and mechanical properties

Among features of the conglomeratic rock mass itself which have significant influence on

landforms and geomorphic processes are thickness, composition, clast-to-matrix proportion, degree

of
of lithification and resulting mechanical properties, particularly strength and porosity. Unfortunately,

ro
very few studies of conglomeratic landforms explore lithological background in more detail.
-p
Conglomerate-dominated successions have variable thickness, from metre-thick beds

sandwiched among clastic deposits of finer grain (sandstones, mudstones) to a few hundred metres
re
and even more than 1 km thick piles of coarse conglomerates deposited in orogenic settings. The
lP

former have limited bearing on morphology and control landforms at the very local, single outcrop

scale. Depending on the properties of matrix and related porosity, such beds may behave as strong
na

rocks, supporting rims and ledges, or as weak rocks, favouring weathering and cliff recession. Thick
ur

packages support distinctive regional landscapes such as mallos towers and rocky uplands in Spain

(Guerrero et al., 2014) or Danxia landscapes in China (Peng, 2020). Relative relief in these terrains is
Jo

obviously related to the thickness of the whole succession.

At the larger spatial scale, it was noted that conglomeratic highlands often terminate in

distinct escarpments and wide transitional belts are missing. This point was addressed in respect to

isolated mallos monoliths in Spain. Rapid transition from coarse alluvial fan facies, including debris

flow deposits, to finer fluvial facies within the sedimentary body architecture, sets the stage for

effective differential erosion (Nichols, 2004, 2005). Similar observations were put forward for some

Danxia landscapes in China (Chen and Guo, 2017). However, significant slope retreat has certainly

24
 Journal Pre-proof

occurred since uplift and dissection of conglomeratic packages and it would be wrong to equate the

outer limit of mallos relief with the original lithological boundary.

Clast and matrix lithology decide about long-term resistance of conglomerates to erosion,

with the nature of matrix being probably of key importance. Therefore, it would be risky to propose

any general relationships between clast lithology and size on one hand, and rock behaviour against

weathering and erosion on the other one. In addition, inherent dynamics of sedimentary

environments favouring deposition of coarse materials results in complicated architecture of

conglomeratic packages, with non-systematic changes in lithology and compactness over metric

of
distances. Guo et al. (2018) showed that coarse pebble insets are often loosely packed and are

ro
preferential places for the development of tafoni. Carbonate matrix in turn favours dissolution and

disintegration of the whole rock mass.

-p
To provide meaningful strength characteristics for conglomerates is difficult, given their
re
bipartite composition and there are not too many studies in this respect (Shafiei and Dusseault,
lP

2008; Akram, 2010). Data compiled by Akram (2010) indicates that the uniaxial compressive strength

(UCS) may be as low as 30 MPa, but examples of UCS in excess of 200 MPa were also reported. One
na

might assume that the latter would be valid for the very high cliffs observed in nature. Accordingly,
ur

the Young Modulus varies from as little as 1–2 GPa to more than 100 GPa. Marinos and Hoek (2000)

decided not to suggest any values of material constant mi (related to the frictional properties of the
Jo

rock) for conglomerates in their recommendations how to use Geological Strength Index in practice.

In fact, they commented that ‘Conglomerates and breccias may present a wide range of mi values

depending on the nature of the cementing material and the degree of cementation, so they may

range from values similar to sandstone, to values used for fine grained sediments’. Solutions to

overcome problems of heterogeneity of rock complexes were offered for flysch and sandy-silty

molasses (Marinos and Hoek, 2000; Hoek et al. 2005), but not for successions dominated by

conglomerates. Duran (2016) compared several methods to determine rock mass strength of

conglomerate and concluded that each gave different results. The term ‘bimrock’ (block-in-matrix)

25
 Journal Pre-proof

was introduced in engineering geology (Medley, 2007) and although primarily intended for chaotic

melanges, conglomerates are also considered to be an example of such ‘bimrocks’ from engineering

perspective (Bonetto et al., 2016). It appears that there is still long way before reliable methods to

evaluate rock mass strength of conglomerates will be available for geomorphological purposes.

5.2. Regional uplift

The role of uplift is twofold and not unique to conglomerates. First, it provides available relief

of
for subsequent landscape development. Thus, if the magnitude of uplift is in the order of hundreds of

ro
metres at least and the conglomeratic succession is of comparable thickness, which is not uncommon
-p
in front of ancient orogenic ranges, conditions for the origin of deeply dissected highlands are met.

The conceptual model used in China to explain the diversity of Danxia landscapes implies the
re
fundamental role of basin inversion and regional uplift, which creates a plateau subject to further
lP

erosion (Peng, 2020). Over time, it turns into dissected plateau and then arrays of residual massifs

until a senile stage of scattered residuals is attained. The presence of high cliffs, a few hundred
na

metres high, is the result of considerable uplift of a poorly differentiated coarse sedimentary
ur

succession.

Second, regional uplift may be associated with large-scale deformation of the former basin
Jo

fill. In effect, conglomeratic packages may retain the horizontal attitude of beds, but they may also

become tilted. Depending on the angle of tilt, respective morphologies are those of cuestas (gentle

dip), monoclinal ridges (moderate dip), up to hogback ridges, with nearly vertical arrangement of

strata. Anticlinal and synclinal structures may also form. The southern part of the Pyrenees in the

vicinity of the town of Sant Llorenç de Morunys (Spain) shows the transition from hogback ridges in

the north through monoclinal ridges to a bowl-shaped shallow syncline in the south, consistent with

an increasing distance from major thrust faults (Fig. 30). Súl’ovské skaly area in Slovakia provides an

example of large-scale relief inversion, where the core of the anticline was eroded and outward

26
 Journal Pre-proof

dipping conglomerate beds form rocky ridges surrounding the central depression (Fig. 31). The

Brecon Beacons range in south Wales, Great Britain, represents a large cuesta with resistant

Devonian clastic deposits acting as caprock, although the morphology of the cuesta face was

considerably remodelled by Quaternary glaciers (Carr, 2020). Babí lom ridge north of Brno, Czechia, is

a fine example of a hogback ridge, crowned by fin-type crags up to 10 m high (Dvořák et al., 1975;

Fig. 10b).

Distribution of medium-size residual landforms such as tors and crags is subordinate to the

gross relief dictated by the regional pattern of tectonic deformation. In tilted conglomeratic

of
successions bedrock outcrops are associated with front scarps and backslope canyons, rarely

ro
occurring on the interfluves (Pilous, 1990). If the dip exceeds 45°, they cluster along water divides, in

the axial parts of hogback ridges.

-p
re
5.3. Major fractures
lP

Major fractures (master joints) are the first-order lines of weakness, focusing dissection and
na

fragmentation of a rock mass. They are exploited by weathering, gravity and water-related transport
ur

processes. The latter include episodic transport of debris down the steep ravines during torrential

rainfall as well as continuous transport in established channels. Altogether, fractures provide a

Jo

geometric framework for landscape evolution and they do so at the variety of scales.

Many Mesozoic conglomeratic successions which support Danxia landscapes in China

developed a fairly regular pattern of major discontinuities during subsequent orogenic phases (Kusky

et al., 2010). In the Cenozoic these fracture systems played a pivotal role in landform evolution and

the contemporary distribution of highlands, domed hills, narrow corridors and wider plains mimics

the grid-like pattern of fractures (Fig. 23). Regular fracture pattern was assumed in the modelling

exercise by Yan et al. (2015). The resultant output topographies, although simplified, showed some

27
 Journal Pre-proof

similarities to real Danxia landscapes at various stages of evolution. The influence of tectonic pre-

design on the landscape of Pedras Negras is also evident (Fig. 24).

Addressing smaller spatial scales, non-uniform density of fractures is of considerable

geomorphic significance. In Montserrat area in Spain one can clearly see domes rising above the

general upland level. They are some 100–150 m high in respect to that level and associated with

more massive bedrock compartments, where major fractures are at least 50 m apart, locally more

than 100 m from one another (Fig. 32). Similar relationships occur in Kata Tjuta, where the highest

domes are associated with very wide spacing of master fractures (Twidale, 2010; Fig. 7). Benito

of
(1993) examined relationships between the density of major fractures and morphology of mallos

ro
domes in Spain, concluding that if spacing is too high, denudation is efficient and high residual domes
-p
hardly form. Conversely, if spacing is too wide, high-elevation platforms occur.

At the local scale, various geomorphic features characteristic for conglomerates develop in
re
relation to joints. Clefts, ravines and slots typically evolve along zones of closely spaced joints, more
lP

prone to weathering and erosion. Morphology of tors and crags is controlled by joint density too,

with projecting spurs in more massive compartments. Dominance of one fracture direction gives rise
na

to narrow fins as in Belogradchik area (Fig. 27e). Details of ground plan of rock cities in
ur

conglomerates are also controlled by jointing pattern.

Jo

5.4. Bedding and interlayering of finer material

Conglomerate packages may be very uniform in vertical profile, which favours the origin of

little differentiated, high cliff faces, but they may also be more diverse. Bedding planes and

intercalations of sandstone, mudstone and claystone are of particular importance for landform

evolution. They influence pathways of groundwater circulation, diverting water flows sideways,

towards the surface. Insets of strata of finer grain size slow down percolation and may even act as

impermeable horizons. Forced lateral movement of groundwater results in seepage from rock faces,

28
 Journal Pre-proof

initiating the process of sapping (Laity and Malin, 1985; Howard and Kochel, 1988). Although seepage

and sapping are best known from sandstone terrains, they occur in conglomerates as well. Their

importance for the evolution of cliffs in Danxia landscapes was emphasized by Young et al. (2009)

and studied in detail by Chinese scholars (Zhu et al., 2010; Peng et al., 2015a; Guo et al., 2018).

Combination of greater water availability and low strength of these fine-grained, porous interlayers

results in enhanced weathering, higher rates of recession, roof undermining, decline of stability and

finally collapse of an overhang.

Conglomerate landforms directly associated with bedding and lithological control of this kind

of
include tafoni, basal rock shelters, bedding caves, wavy microrelief of rock cliffs, and stepped profiles

ro
of hillslopes in micro-scale, with ledges and benches.
-p
5.5. Jointing, bedding and the origin of domes
re
lP

Wide spacing of vertical joints is quite a common feature of conglomeratic packages and is

strikingly associated with the presence of domed residual hills (Fig. 7, 32). This strongly suggests
na

causal relationship. The origin of rounded hillslopes was more often addressed in respect to
ur

sandstones (Ollier and Tuddenham, 1961; Bradley 1963; Young 1986, 1987, Young et al., 2009; Migoo

and Goudie, 2014), but is perhaps even more relevant to conglomerates as other hillslope shapes,
Jo

common in sandstones, are rather infrequent and sharp edges are rare. It is important to note that

these curved shapes are not necessarily related to sheeting joints, which are actually quite rare in

conglomerates. Sheeting joints are convex partings parallel to the slope and usually considered to

result from rock mass unloading due to ongoing denudation. Rather, the origin of curved shapes in

sandstones is attributed to dominant grain-by-grain disintegration and not to detachment of larger

rock pieces delineated by joints (Young, 1986; Howard and Kochel, 1988; Young et al., 2009). In the

sandstone landscape of south Jordan iron-rich intercalations in one sandstone member give rise to

distinct rims, ledges and benched slopes, whereas another one, lacking iron inliers, develops perfect

29
 Journal Pre-proof

convex hillslope shapes (Migoo and Goudie, 2014). Extending this concept to conglomerates, one can

replace grain-by-grain breakdown by clast-by-clast disintegration, as already indicated by Young et al.

(2009). Aprons of thick debris mantles composed of clasts released from the outcrops testifies to this

mechanism, whereas rock fall derived talus is rare. The preference for clast release was observed by

Young (1992) in respect to the conglomeratic Ragged Ranges in north Australia. He emphasized

causal linkages between rock breakdown and form stating that ‘they (domes – P.M.) are rounded

because the large clasts are readily dislodged from the relatively weak arkosic matrix on faces where

tensional or shearing stresses dominate’ (p. 144). The key prerequisite seems to be the paucity of

of
other structural features which may control the progress of cliff retreat or lowering of upper

ro
platforms. In Meteora bedding planes are quite prominent and hence, the towers have vertical cliffs
-p
and planar tops rather than summit cupolas. However, an old report by Arnold (1907) contains

description of a minor dome which was built of well-bedded conglomerate and nevertheless
re
developed surface-parallel fractures. Weathering processes rather than lithostatic stress release
lP

were assumed to explain the presence of cracks.

The height of a hill may also play an important role. Whereas low domes, up to 50 m high or
na

so, have gently inclined slopes all around (e.g. in the eastern part of Kata Tjuta group), the higher
ur

ones have their slopes bipartite, with cliffs in the lower and middle part and rounded tops in the

upper part (Fig. 6A). The former may be affected by rock slope failures and locally overhanging, the
Jo

latter appear stable and undergoing very slow change. This suggests that rounded shapes are long-

term equilibrium slopes for massive, poorly bedded conglomerates (see Selby, 1980), but ongoing

incision and excavation of fractured zones steepens the slopes turning them into cliffs. However, the

rock mass strength of conglomerates in compression is sufficiently high for cliffed faces to survive for

quite a long time, their overall stability being punctuated by probably infrequent rock falls. If this

scenario holds true, rounded hilltops would be inherited landforms, developed prior to the more

recent incision (Fig. 33). Views of Twidale (2010) about the antiquity of Kata Tjuta morphology would

be consistent with this proposal.

30
 Journal Pre-proof

5.6. Presence of carbonates and amorphous silica

Clastic rocks have long been considered non-karstic, with dissolution playing negligible role in

the evolution of landforms. This view is no longer tenable and the evidence of karst processes in

sandstones is now abundant (Jennings, 1983; Young, 1986, 1987; Wray, 1997, Young et al., 2009;

Wray and Sauro, 2017). In respect to conglomerates, the issue seems more complex as karst in

conglomerates has many faces. First, the majority of constituting clasts may be derived from

of
limestone. Second, the cement can be calcareous. Third, both clasts and cement may be

ro
predominantly made of calcium carbonate. Fourth, the role of cement may be played by amorphous
-p
silica which may quite effectively dissolve under specific circumstances (high pH, high temperature,

long timescale).
re
Examples of karst provided earlier in this paper come from calcareous conglomerates,
lP

typically with calcium carbonate cement (Bergada et al., 1997; Ferrarese and Sauro, 2005; Goeppert

et al., 2011; Lipar and Ferk, 2011). They show that karst in such lithological variant can be quite
na

evolved, both at the surfaces, as indicated by the density of dolines (Ferrarese and Sauro, 2005; Čeru
ur

et al., 2017), and underground, as testified by >1 km long cave systems (Bergada et al., 1997). By

contrast, karst landforms in non-carbonatic conglomerates, analogous to those in sandstone and

Jo

comprehensively reviewed by Wray and Sauro (2017), are yet to be properly documented.

6. Conglomerate versus sandstone landforms and landscapes

The distinction between sand and gravel particles is arbitrary set at 2 mm. Likewise, the

distinction between sandstones and conglomerates in terms of grain size is arbitrary, even more so,

since constituent particles within a rock mass may be of both sand and gravel size. This mixed

composition is reflected in names such as sandy conglomerates or gravelly sandstone (Nichols, 2009).

31
 Journal Pre-proof

Consequently, it is reasonable to assume that continuity of landforms and landscapes will exist

between sandstone-dominated and conglomerate-dominated terrains. Indeed, overviews of

sandstone landforms (Härtel et al., 2007; Young et al., 2009; Adamovič et al., 2010) show sceneries

remarkably similar to those present in conglomerates, but they also show landscapes that are

evidently different (Fig. 26).

Comparative studies of sandstone topographies worldwide indicate two major mechanisms

of rock breakdown and slope development (Young and Young, 1992; Young et al., 2009, Migoo and

Goudie, 2014). One is grain-by-grain breakdown which results in smooth slope shapes and curved

of
profiles, typically convex. The consequence is poor development of talus mantling the footslopes,

ro
since sandy aprons are easily eroded and sand can be carried further down by overland flow. There
-p
are several factors facilitating this mechanism such as poor and non-systematic jointing, wide

bedding, low degree of induration, poor development of cement, very low tensile and shear strength
re
(Young and Wray, 2015). Conversely, if cementation is strong and jointing pattern fairly dense and
lP

regular, grain-by-grain disintegration mechanism is suppressed and taken over by the release of

larger blocks. Rounded slopes are then replaced by cliffs, angular towers and pinnacles. Observations
na

from adjacent areas, hence representing the same weathering environment, but underlain by
ur

different types of sandstone, support the reasoning based on rock characteristics (Young et al., 2009,

Migoo and Goudie, 2014; Young and Wray, 2015).

Jo

The above argument can be extended to conglomerates. In fact, Young (1992) noted

contrasts between topographies developed in some conglomeratic successions versus sandstones in

the East Kimberley of northern Australia and emphasized rounded shapes (domes) in conglomerates,

attributing them to preferential clast-by-clast disintegration. Thus, if cement is weak and/or can be

dissolved, but jointing is poorly developed, slope evolution will proceed towards rounded shapes, as

in some sandstones. However, if bonding is strong and dense fracturing is present, more angular

shapes will originate. But dense and regular fracture patterns of the kind comparable to those in

sandstone are not common and therefore, sandstone landscapes dominated by angular shapes (e.g.,

32
 Journal Pre-proof

Zhangjiajie in China, Yang et al., 2012) do not have too many counterparts in conglomerates. Angular

towers of Meteora may be cited here, where prominent bedding controls flattish tops, preventing

advanced rounding to occur. Flattened tops in Serra da Capivara (Fig. 26) have also developed in a

well-bedded succession, with local induration.

Coexistence of rounded hilltops and cliffs below, noted for conglomerates, was also observed

in sandstones. Young and Wray (2015) explained it by emphasizing the disparity between high

compressive and low shear strength of sandstones of Purnululu. In their own words, ‘As the

sandstone consists essentially of a network of grains in point-to-point contact, the interlocking

of
stresses increase with depth due to the increasing weight of overburden; the less-loaded summits

ro
thus are more readily eroded than the flanks, thereby increasing the convexity’ (p. 9). It is possible
-p
that similar explanation can be offered for conglomerates although part of the stress will be borne by

matrix.
re
Many minor landforms are common for conglomerates and sandstones. These include
lP

horizontal shelters (bedding caves) developed along bedding planes and intercalations of strata

which for various reasons are weaker than the host rock. Karren and tafoni can be found on both
na

types of clastic rocks, although those on conglomerates seem to be bigger. However, solid empirical
ur

evidence is yet to be provided. On the other hand, small-scale alveolar (honeycomb) weathering,

with individual pits of a few centimetres long and deep, is generally absent on conglomerates or very
Jo

crudely developed. In sandstones, by contrast, this weathering pattern is very widespread (e.g.,

Turkington and Phillips, 2004; Young et al., 2009; Bruthans et al., 2018). Likewise, regular weathering

pits on planar surfaces are not common. In both cases textural differences seem to hold an

explanation. Conglomerates are far too anisotropic at the local scale to allow these regular, but small

geometric shapes to develop.

7. Concluding remarks

33
 Journal Pre-proof

7.1. Summary of observations

Our knowledge and understanding of conglomeratic landscapes is still very incomplete, both

in terms of geographical coverage as well as, perhaps more importantly, processes and controls

involved in the geomorphic evolution of these sceneries. Given striking appearance of many

conglomerate landscapes, this very limited interest is puzzling, especially if contrasted with the

considerable attention paid to sedimentology of coarse deposits within the geological community.

Nevertheless, critical analysis of available publications coupled with the personal experience

of
of the author allow one to emphasize certain key features of conglomeratic terrains. These include:

ro
 considerable proportion of bare bedrock surfaces in the landscape, especially in semi-humid,

semi-arid and arid climates.

-p
 high degree of erosional dissection, focused along master joints, leading to large-scale
re

labyrinthine relief.
lP

 the presence of steep-sided domes with rounded summits and rock cliffs as characteristic

medium-scale landforms.
na

 predominance of clast-by-clast disintegration, with major rock failures not being very
ur

common.


Jo

important role of selective weathering, guided by lithological heterogeneities and structural

features. Tafoni, rock shelters and bedding caves are present in many areas.

 the presence of karst in conglomerates predominantly made of limestone clasts and with

calcareous cement. Both surface (dolines) and subsurface (caves) karst landforms occur.

 impressive badlands occur in some non-lithified gravel deposits.

 correspondence between the distinctiveness of the scenery and the thickness of

conglomerate successions.

However, it must not be assumed that these features will be present everywhere and may play a

diagnostic role. Some conglomeratic terrains are rather subdued, with scarce bedrock outcrops in the

34
 Journal Pre-proof

form of indistinct and widely dispersed crags. Apart from thickness, geomorphic distinctiveness is

related to the composition and hence, strength and resistance of conglomerates. Well-lithified

varieties, densely packed, with high proportion of resistant clasts and siliceous cement, are much

more likely to support impressive rock-cut scenery.

7.2. Research avenues for the future

The above review indicates that gaps in our knowledge of conglomerate geomorphology are

of
many, applicable to all scales of research, and it will take long time before they are all filled.

ro
However, two main directions of future research are strongly recommended. First, the fundamental
-p
knowledge about the range of relief types on conglomerates is very incomplete. Published papers

often focus on the most impressive sceneries (e.g., Tronkov and Sinnyovsky, 2012; Guerrero et al.,
re
2014; Peng, 2020), but this should not obscure the fact that large tracts of conglomerate terrains lack
lP

comparable grand features. In fact, this review too is focused on distinctive landforms, which may

not be entirely representative. Extending landform inventories present in conglomerates, from as

na

many different climatic settings as possible, will help to achieve a more balanced picture. Seeking
ur

geomorphometric signatures of conglomerate terrains would be a worthwhile exercise, facilitating

future comparative analysis. These may include the distribution of slope angles and shapes, valley
Jo

network density, valley form cross-sections, proportion of bedrock versus regolith-covered surfaces.

Global approach to inventory is now possible due to availability of satellite imagery (e.g., Goudie,

2020). Large tracts of continents can be inspected quite quickly and although not all properties of the

terrain can be extracted from images, perspectives for global overviews are now much better than

anywhere before, at least in respect to conglomerate areas with bedrock widely exposed. Numerous

examples how Google Earth imagery can be used were provided in this study.

Second, it is quite clear that conglomerates are a very heterogeneous family of rocks and this

understanding is yet to be fully realized in geomorphological studies. It is hypothesized that much of

35
 Journal Pre-proof

the terrain variability in conglomerates at the medium and small scale can be explained by

examination of rock characteristics and properties. This kind of research is present in the work on

China Danxia landscapes (e.g., Peng et al., 2015a; Chen and Guo, 2017; Guo et al., 2018) and is very

promising, especially when sedimentology is linked with geomorphology. Distinguishing between

matrix- and clast-supported conglomerates, different types of cement and clast lithology, coupled

with due attention to larger sedimentary structures should underpin geomorphological endeavours.

This research direction should be connected with proper consideration of mechanical properties of

conglomerates, which are unfortunately rarely addressed. The rock mechanics perspective

of
considerably benefitted sandstone geomorphology research (Young et al., 2009; Bruthans et al.,

ro
2014; Young and Wray, 2015). It is reasonable to assume that comparable progress will be made in

understanding conglomerate landscapes.

-p
re
Acknowledgements
lP

This paper is the outcome of an invited lecture presented at the IAG Regional Conference on
na

Geomorphology in Athens in September 2019. My thanks are to the organizers for their kind
ur

invitation to give the plenary presentation, to Achim Beylich for suggesting this review paper to be

prepared, and to Tim Horscroft (Elsevier) for encouragement and guidance. The idea of writing a
Jo

review on conglomerate geomorphology is however not new and was discussed by the present

author with the late Robert Wray nearly 10 years ago. With deteriorating health of Robert, the joint

paper has never really started but I acknowledge Robert’s inspirational role and dedicate this paper

to him. I am also very grateful to people who in various ways helped me to see conglomeratic

landscapes in various places in the world, also sharing information about sites which I was unable to

visit. In particular I wish to mention the late Professor Peng Hua who made me involved in Danxia

research in China, as well as my Chinese friends Ren Fang and Pan Zhixin. Rubson Pinheiro Maia

(Fortaleza, Brazil) gladly responded to my request and organized a long trip to Serra da Capivara

36
 Journal Pre-proof

National Park, where impressive conglomerate and sandstone rock landscapes occur side by side.

Other individuals who deserve mention are Jiun-Chuan Lin (Taiwan), Mike Crozier (New Zealand),

Valentina Nikolova (Bulgaria) and Emmanuel Reynard (Switzerland). Relevant information was also

provided by Augusto Auler (Brazil), Andrea Coronato (Argentina) and Francesco Sauro (Italy). Bettina

Ercolano, Jiun-Chuan Lin, Liuqin Chen, Ren Fang and Pan Zhishin kindly supplied some photographs

used in this paper.

References

of
ro
Adamovič, J., Mikuláš, R., Cílek, V., 2010. Atlas pískovcových skalnych měst České a Slovenské
-p
republiky: Geologie a geomorfologie [Atlas of Sandstone Rock Cities of the Czech and Slovak

Republic: Geology and Geomorphology]. Academia, Praha (460 pp.) (in Czech).
re
Akram, M.S., 2010. Physical and numerical investigation of conglomeratic rocks. Unpubl. PhD Thesis,
lP

University of New South Wales, Sydney.

Arnold, R., 1907. Dome structure in conglomerate. Journal of Geology 15, 560–570.
na

Auler, A., 2012. Quartzite caves of South America. In: White, W.B., Culver, D.C. (eds.), Encyclopedia of
ur

Caves (2nd ed.). Academic Press, Chennai, pp. 635–639.

Bamlett, M., 1990. Conglomerates. Geology Today, 6(4), i–iv.

Jo

Bartuś, T., Mastej, W., 2013. The Shepherds’ Rocks. In: Słomka, T. (ed.), Katalog obiektów

geoturystycznych w obrębie pomników i rezerwatów przyrody nieożywionej. AGH, Kraków, pp. 97–

100.

Begg, J.G., Johnston, M.R. (comp.), 2000. Geology of the Wellington area. Institute of Geological &

Nuclear Sciences, Lower Hutt, New Zealand (64 pp.).

Benito, G., 1993. Genesis and evolution of the Mallo pinnacle landform. In: Second Intensive Course

on Applied Geomorphology, Zaragoza, pp. 179–183.

37
 Journal Pre-proof

Bergadá, M.M., Cervello, J.M., Serrat, D., 1997. Karst in conglomerates in Catalonia (Spain):

morphological forms and sedimentary sequence types recorded on archaeological sites. Quaternaire

8, 267–277.

Blair, T.C., McPherson, J.G., 1994. Alluvial fans and their natural distinction from rivers based on

morphology, hydraulic processes, sedimentary processes and facies assemblages. Journal of

Sedimentary Research A64, 450–589.

Boggs, S. Jr., 2009. Petrology of Sedimentary Rocks (2nd ed.). Cambridge University Press, Cambridge

(600 pp.).

of
Bonetto, S., Comina, C., Colombero, C., Dela Pierre, F., Ferrero, A.M., Giordano, N., Rispoli, A.,

ro
Vagnon, F., 2016. Study of the mechanical properties of a conglomerate. Procedia Engineering 158,

248–253.
-p
Bourne, J.A., Twidale, C.R., 2003. Geomorphological development of the Baxter Hills, a conglomeratic
re
upland near Iron Knob, South Australia. Zeitschrift für Geomorphologie N.F. 47, 351–371.
lP

Bradley, W.C., 1963. Large-scale exfoliation in massive sandstones of the Colorado Plateau.

Geological Society of America Bulletin 74, 519–528.

na

Bruthans, J., Soukup, J., Vaculikova, J., Filippi, M., Schweigstillova J., Mayo, A.L., Masin, D.,
ur

Kletetschka, G., Rihosek, J., 2014. Sandstone landforms shaped by negative feedback between stress

and erosion. Nature Geoscience 7(7), 597–601.

Jo

Bruthans, J., Filippi, M., Slavk, M., Svobodová, E., 2018. Origin of honeycombs: Testing the hydraulic

and case hardening hypotheses. Geomorphology 303, 68–83.

Burlando, M., Firpo, M., Queirolo, C., Rovere, A., Vacchi, M., 2011. From Geoheritage to Sustainable

Development: Strategies and Perspectives in the Beigua Geopark (Italy). Geoheritage 3, 63–72.

Carr, S., 2020. The Brecon Beacons. In: Goudie, A.S., Migoo, P. (eds.), Landscapes and Landforms of

England and Wales. Springer, Switzerland, pp. 557–570.

Carter, C.L., Anderson, R.S., 2006 Fluvial erosion of physically modeled abrasion-dominated slot

canyons. Geomorphology 81, 89–113.

38
 Journal Pre-proof

Čeru, T., Šegina, E., Gosar, A., 2017. Geomorphological dating of Pleistocene conglomerates in

Central Slovenia based on spatial analyses of dolines using LiDAR and Ground Penetrating Radar.

Remote Sensing 9, 1213.

Chen, L., Guo, F., 2017. Upper Cretaceous alluvial fan deposits in the Jianglangshan Geopark of

Southeast China: implications for bedrock control on Danxia landform evolution. Journal of Mountain

Science 14, 926–935.

Collinson, J.D., Thompson, D.B., 1982. Sedimentary Structures. Allen & Unwin, London (194 pp.).

Dunkley, J., Ellis, M., Bolger, T., 2017. Unusual caves and karst-like features in sandstone and

of
conglomerate in Thailand. Helictite 43, 15–31.

ro
Duran, A., 2016. Rock mass assessment — what goes wrong? In: Dight, P.M. (ed.), Proceedings of
-p
APSSIM 2016 Australian Centre for Geomechanics, Perth, pp. 493–506.

Duszyoski, F., Migoo, P., Strzelecki, M.C., 2019. Escarpment retreat in sedimentary tablelands and
re
cuesta landscapes – Landforms, mechanisms and patterns. Earth-Science Reviews 196, 102890.
lP

Dvořák et al., 1975. Karásek, J. Netopil, R. (1975): Soliflukční kamenné proudy na Babím lomu u Brna

[Solifluction block streams at Babí lom near Brno]. Časopis pro Mineralogie a Geologie 20, 303–306
na

(in Czech only).

ur

Ferrarese, F., Sauro, U., 2005. The Montello hill: The “Classical Karst” of the conglomerate rocks. Acta

Carsologica 34, 439–448.

Jo

Folk, R.L., [Link] distinction between grain size and mineral composition in sedimentary-rock

nomenclature. Journal of Geology, 62, 344–359.

Folk, R.L., Andrews, P.B., Lewis, D.W., 1970. Detrital sedimentary rock classification and

nomenclature for use in New Zealand. New Zealand Journal of Geology and Geophysics, 13, 937–968.

Ford, D.C., Williams, P.W., 2007. Karst Hydrogeology and Geomorphology (2nd ed.). Wiley, Chichester

(576 pp.).

Godard, A., Lagasquie, J.-J., Lageat, Y. (eds.), 2001. Basement Regions. Springer, Heidelberg (306 pp.).

39
 Journal Pre-proof

Goeppert, N., Goldscheider, N., Scholz H., 2011. Karst geomorphology of carbonatic conglomerates in

the Folded Molasse zone of the Northern Alps (Austria/Germany). Geomorphology 130, 289–298.

Goudie, A.S., 2004. Climatic geomorphology. In: Goudie, A.S. (ed.), Encyclopedia of Geomorphology.

Routledge, London, pp. 162–164.

Goudie, A.S., 2020. Global barchans: A distributional analysis. Aeolian Research 44, 100591.

Guerrero, J., Gutierrez, F., Gutierrez, M., 2014. Conglomerate monoliths and karst in the Ebro

Cenozoic Basin, NE Spain. In: Gutierrez, F., Gutierrez, M. (eds.), Geomorphological Landscapes of

Spain. Springer, Heidelberg – Dordrecht, pp. 81–89.

of
Gunn, J. (ed.), 2004. Encyclopedia of Karst and Caves. Taylor and Francis, London (902 pp.).

ro
Guo, F., Chen, L., Xu, H., Liu, X., 2018. Origin of beaded tafoni in cliffs of Danxia landscapes,
-p
Longhushan Global Geopark, South China. Journal of Mountain Science 15, 2398–2408.

Gutiérrez, F., Gutiérrez, M., Gracia, F.J., 2005. Karst, neotectonics and periglacial features in the
re
Iberian Ranges. Field Trip C-5. Sixth International Conference on Geomorphology, Zaragoza (58 pp.).
lP

Härtel, H., Cílek, V., Jackson, A., Herben, T., Williams, R. (eds.), 2007. Sandstone Landscapes.

Academia, Praha (493 pp.).

na

Harvey, A., 2004. Badland. In: Goudie, A.S. (ed.), Encyclopedia of Geomorphology. Routledge,
ur

London, pp. 45–48.

Hoek, E., Marinos, P., Marinos, V., 2005. Characterisation and engineering properties of tectonically
Jo

undisturbed but lithologically varied sedimentary rock masses. International Journal of Rock

Mechanics and Mining Sciences 42, 277–285.

Howard, A.D., Kochel R.C., 1988. Introduction to cuesta landforms and sapping processes on the

Colorado Plateau. In: Howard, A.D., Kochel, R.C., Holt, M.E. (eds.), Sapping features of the Colorado

Plateau. A comparative planetary geology field guide. NASA Scientific and Technical Information

Division 1988, Washington DC, pp. 6–56.

40
 Journal Pre-proof

Jennings, J.N., 1983. Sandstone pseudokarst or karst. In: Young, R.W., Nanson, G.C. (eds.), Aspects of

Australian Sandstone Landscapes. Australian and New Zealand Geomorphology Group Special

Publication 1, pp. 21-30.

Jennings, J., 1985. Karst Geomorphology, 2nd ed. Blackwell, Oxford (293 pp.).

Kim, Y.B., 2008. Geoheritage of Korea. In: Leman, M.S., Reedman, A., Pei, C.S. (eds.), Geoheritage of

East and Southeast Asia. Lestari, Malaysia, pp. 115–147.

Koster, E.H., Steel, R.J. (eds.), 1984. Sedimentology of Gravels and Conglomerates. Memoir Canadian

Society of Petroleum Geology, Calgary 10 (441 pp.).

of
Kusky, T.M., Ye, M., Wang, J., Wang, L., 2010. Geological evolution of Longhushan World Geopark in

ro
relation to global tectonics. Journal of Earth Science 21, 1–18.
-p
Laity, J., Malin, M.C., 1985. Sapping processes and the development of theater headed valley

networks on the Colorado Plateau. Bulletin of the Geological Society of America 96, 203–217.
re
Lin, J.-C., Su, S.-J., 2019. Geoparks of Taiwan. Their Development and Prospects for a Sustainable
lP

Future. Cham, Springer (178 pp.).

Linton, D.L., 1955. The problem of tors. Geographical Journal 121, 451–470.
na

Lipar, M., Ferk, M., 2011. Eogenetic caves in conglomerate: an example from Udin Boršt, Slovenia.
ur

International Journal of Speleology 40, 53–64.

Liu, F., Li, X., 2018. A brief analysis of cave types of Danxia Landform in Xinjiang Basin. IOP Conference
Jo

Series: Earth and Environmental Science 170, 022129.

Lozano, M.V., Fabregat, C., López, S., González, J.M., 2007, Albarracín Rodeno landscape (Spain). In:

Härtel, H., Cílek, V., Jackson, A., Herben, T., Williams, R. (eds.), Sandstone Landscapes. Academia,

Praha, pp. 368–371.

Marinos, P., Hoek, E., 2000. GSI: A geologically friendly tool for rock mass strength estimation. In:

GeoEng2000: An International Conference on Geotechnical & Geological Engineering, International

Society for Rock Mechanics, Melbourne, pp. 1422–1442.

41
 Journal Pre-proof

Martínez, O.A., Kutschker, A., 2011. The ‘Rodados Patagónicos’ (Patagonian shingle formation) of

eastern Patagonia: environmental conditions of gravel sedimentation. Biological Journal of the

Linnean Society 103, 336–345.

Martínez, O.A., Rabassa, J., Coronato, A., 2009. Charles Darwin and the first scientific observations on

the Patagonian Shingle Formation (rodados patagónicos). Revista de la Asociación Geológica

Argentina 64(1), 90–100.

Medley, E.W., 2007. Bimrocks – Part 1: Introduction. In: Newsletter of the Hellenic Society of Soil

Mechanics and Geotechnical Engineering 7, 17–21.

of
Migoo, P., 2006. Granite Landscapes of the World. Oxford University Press, Oxford (384 pp.).

ro
Migoo, P., Goudie, A., 2014. Sandstone geomorphology of south-west Jordan, Middle East.

Quaestiones Geographicae 33, 119–126.

-p
Migoo, P., Duszyoski, F., Goudie, A., 2017. Rock cities and ruiniform relief: Forms – processes –
re
terminology. Earth-Science Reviews 171, 78–104.
lP

Moncrieff, A.C.M., 1989. Classification of poorly-sorted sedimentary rocks. Sedimentary Geology 65,

191–194
na

Nichols, G.J., 2004. Sedimentation and base level in an endorheic basin: the early Miocene of the
ur

Ebro Basin, Spain. Boletín Geológico y Minero 115, 427–438.

Nichols, G.J., 2005. Tertiary alluvial fans at the northern margin of the Ebro Basin: a review.
Jo

Geological Society, London, Special Publications 251, 187–206.

Nichols, G., 2009. Sedimentology and Stratigraphy (2nd ed.). Wiley-Blackwell, Chichester (419 pp.).

Oh, C., Lee, S., Lee, B., 2017. The forming process of the Maisan and nearby famous mountains and

the related mountain ranges and water systems. Journal of Petrological Society of Korea 26, 201–

219.

Ollier, C.D., Tuddenham, W.G., 1961. Inselbergs of Central Australia. Zeitschrift für Geomorphologie

N.F. 5, 257–276.

42
 Journal Pre-proof

Oprea, R., Nedelea, A., Comănescu, N., 2012. Petrographic relief in the Bucegi (Prahovean Area) and

Ceahlău Mountains (Central Area) – the Romanian Carpathians. Scientific approach vs local legends.

GeoJournal of Tourism and Geosites, 10, 183–192.

Ori, G.G., Roveri, M., 1987. Geometries of Gilbert-type deltas and large channels in the Meteora

Conglomerate, Meso-Hellenic basin (Oligo-Miocene), central Greece. Sedimentology 34, 845–859.

Peng, H., 2020. China Danxia. Higher Education Press and Springer Nature, Singapore (391 pp.).

Peng, H., Pan, Z., Qiu, Z., 2015a. Weathering pattern and development process of bedding-controlled

caves at Mt. Danxiashan, China. Zeitschrift für Geomorphologie N.F. 59, Suppl. 1, 167–184.

of
Peng, H., Ren, F., Pan, Z., 2015b. A review of Danxia landforms in China. Zeitschrift für

ro
Geomorphologie N.F. 59, Suppl. 1, 19–33.
-p
Piccini, L., De Vivo, A., Lo Mastro, F., Straford, T., 2009. Preliminary overview on the karst areas of

Kalaw (Myanmar). In: Proceedings 15th International Congress of Speleology, Kerville, Texas (USA),
re
vol. 3, 1866–1872.
lP

Pilous, V., 1990. Tvary zvětrávání a odnosu Vlčických a Zámeckých skal u Trutnova. Opera Corcontica

27, 5–46.
na

Ren F., 2013. Study of geological controlling factors in the formation of Danxia Landform in
ur

Longhushan Geopark, southeast China. In: Migoo, P., Kasprzak, M. (eds.), Sandstone Landscapes.

Diversity, Ecology and Conservation. Proceedings of the 3rd International Conference on Sandstone
Jo

Landscapes, Kudowa-Zdrój (Poland), 25-28 April 2012. Instytut Geografii i Rozwoju Regionalnego

Uniwersytetu Wrocławskiego, Wrocław, pp. 163–170.

Reynard, E., Estoppey, E., 2020. The Lavaux World Heritage terraced vineyard. In: Reynard, E. (ed.),

Landscapes and Landforms of Switzerland. Springer, Switzerland (in press).

Selby, M.J., 1980. A rock-mass strength classification for geomorphic purposes: with tests from

Antarctica and New Zealand. Zeitschrift für Geomorphologie N.F. 24, 31–51.

43
 Journal Pre-proof

Shafiei, A., Dusseault, M.B., 2008. Geomechanical properties of a conglomerate from Iran. In: 42nd

US Rock Mechanics Symposium and 2nd U.S.-Canada Rock Mechanics Symposium, San Francisco,

June 29-July 2, 2008, 8 pp.

Shin, J.-R., Lee, J.-K., Choo, C.-O., Park, K.-G., 2015. A study on weathering processes of tafoni in Mt.

Cheonsaeng, Gumi, the Republic of Korea: Interpretation of water content data using GIS

interpolation analysis. Journal of Korean Earth Science Society 36, 543–552 (in Korean, with English

summ.).

Suzuki, T., 2004. Rock control in geomorphology. In: Goudie, A.S. (ed.), Encyclopedia of

of
Geomorphology. Routledge, London, pp. 873–876.

ro
Tricart, J., Cailleux, A., 1972. Introduction to Climatic Geomorphology. Longman, London (295 pp.).
-p
Tronkov, D., 1998. Belogradchik rock sculptures — a child of erosional power of nature and of

geological controls. Geologica Balcanica 28(3–4), 153–158

re
Tronkov, D., Sinnyovsky, D., 2012. Belogradchik Rocks, Bulgaria: Geological setting, genesis and
lP

geoconservation value. Geoheritage 4, 153–164.

Turkington, A.V., Phillips, J.D., 2004. Cavernous weathering, dynamical instability and self-
na

organization. Earth Surface Processes and Landforms 29, 665–675.

ur

Twidale, C.R., 1981. Granite inselbergs: domed, block-strewn and castellated. Geographical Journal

147, 54–71.
Jo

Twidale, C.R., 1982. Granite Landforms. Elsevier, Amsterdam (372 pp.).

Twidale, C.R., 2010. Uluru (Ayers Rock) and Kata Tjuta (The Olgas): Inselbergs of central Australia. In:

Migoo, P. (ed.), Geomorphological Landscapes of the World. Springer, Dordrecht, pp. 321–332.

Twidale, C.R., Vidal Romaní, J.R., 2005. Landforms and Geology of Granite Terrains. Balkema,

Rotterdam (351 pp.).

Viljoen, J.H.A., Cawthra, H.C., 2019. Lithostratigraphy of the Buffelskloof Formation (Uitenhage

Group), South Africa. South African Journal of Geology 122, 97–104.

44
 Journal Pre-proof

Viles, H., 2011. Biogeomorphology. In: Gregory, K.J., Goudie, A.S. (eds.), The SAGE Handbook of

Geomorphology. SAGE, London, pp. 246–259.

Wray, R.A.L., 1997. A global review of solutional weathering forms on quartz and sandstone. Earth-

Science Reviews 42, 137–160.

Wray, R.A.L., Sauro, F., 2017. An updated global review of solutional weathering processes and forms

in quartz sandstones and quartzites. Earth-Science Reviews 171, 520–557.

Yan, Z., Du, X., Fan, X., 2015. Numerical simulation of the evolutionary process of Danxia landforms,

Physical Geography 36, 322–336.

of
Yan, L., Peng, H., Hu, Z., Kašanin-Grubin, M., Greif, V., Chen, Z., Simonson, S., 2016. Stone pillar

ro
rockfall in Danxia landform area, Mt. Langshan, Hunan Province, China. Physical Geography 37, 327–

343.
-p
Yang, G., Tian, M., Zhang, X., Chen, Z., Wray, R.A.L., Ge, Z., Ping, Y., Ni, Z., Yang, Z., 2012. Quartz
re
sandstone peak forest landforms of Zhangjiajie Geopark, northwest Hunan Province, China: pattern,
lP

constraints and comparison. Environmental Earth Sciences 65, 1877–1894.

Yatsu, Y., 1966. Rock Control in Geomorphology. Sozosha, Tokyo (135 pp.).
na

Young, R.W., 1986. Tower karst in sandstone: Bungle Bungle massif, northwestern Australia.
ur

Zeitschrift für Geomorphologie N.F. 30, 189–202.

Young, R.W., 1987. Sandstone landforms of the tropical East Kimberley region, northwestern
Jo

Australia. Journal of Geology 95, 205–218.

Young, R.W., 1992. Structural heritage and planation in the evolution of landforms in the East

Kimberley. Australian Journal of Earth Sciences 39, 141–151.

Young, R.W., Young, A.R.M., 1992. Sandstone Landforms. Springer, Heidelberg (163 pp.).

Young, R.W., Wray, R.A.L., 2015. Rock control in sandstone geomorphology: a tribute to Eiju Yatsu

with some Australian examples. Zeitschrift für Geomorphologie N.F. 59, Suppl.-Bd. 1, 3–17.

Young, R.W., Wray, R.A.L., Young, A.R.M., 2009. Sandstone Landforms. Cambridge University Press,

Cambridge (304 pp.).

45
 Journal Pre-proof

Zhu, C., Peng, H., Li, Z., Zhang, G., Li, L., Yu, J., Zhu, G., Ouyang, J., Zhong, Y., Zhu, Q., Zhou, S., Zheng,

C., Zhou, R., Li, D., Zhu, Y., Lv, W., Wu, H., Xu, L., 2009. Age and genesis of the Danxia landform on

Jianglang Mountain, Zhejiang Province. Journal of Geographical Sciences 19, 615–630.

Zhu, C., Peng, H., Ouyang, J., Hu, Z.H., Li, L., 2010. Rock resistance and development of horizontal

grooves on Danxia slopes. Geomorphology 123, 84–96.

of
ro
-p
re
lP
na
ur
Jo

46
 Journal Pre-proof

Captions

Fig. 1. Location map. Localities (sites and regions) mentioned in text: 1 – Montserrat, 2 – Spanish

Pyrenees and their foreland, 3 – Teruel, 4 – Têt Valley, 5 – Lavaux, 6 – Northern Alps (Germany), 7 –

Montello karst, 8 – Slovenia, 9 – Djavolja Varoš, 10 – Belogradchik, 11 – Meteora, 12 – Romanian

Carpathians (Bucegi Mts., Cehlau Mts.), 13 – Súl’ovské skaly, 14 – Babí lom, 15 – Sudetes, 16 – Brecon

Beacons, 17 – Pokhara, 18 – Kalaw karst, 19 – east Thailand, 20 – south China (Danxiashan, Langshan,

Taining), 21 – south-east China (Longhushan, Jianglangshan, Fangyan), 22 – Shipton’s Arch, 23 –

of
Maisan, 24 – Taiwan, 25 – Ragged Ranges and East Kimberley, 26 – Kata Tjuta, 27 – Putangirua

ro
Pinnacles, 28 – Clay Cliffs, 29 – Legzira Beach, 30 – Pedras Negras, 31 – South Africa/Cape Mountains,
-p
32 – Serra da Capivara, 33 – Chapada Diamantina, 34 – south Patagonia, 35 – Cueva del Milodon.
re
Fig. 2. Examples of conglomerate cliffs: A – low (~10 m high) cliff in Carboniferous conglomerates in
lP

SW Poland; B – high cliffs (>100 m) in a thick Mesozoic conglomeratic sequence in Langshan (China).
na

Fig. 3. Recent collapse of a rock pillar in Langshan (China). A – image dated 9-03-2009; B – image
ur

dated 25-12-2013 (source: Google Earth)

Jo

Fig. 4. Evidence of massive ancient rock slope collapse in Longhushan (China). Photo supplied by Pan

Zhishin.

Fig. 5. Overhang due to more efficient weathering of a thin (~1 m) pebbly layer within a sandstone

succession (West Sudetes, SW Poland).

47
 Journal Pre-proof

Fig. 6. Contrasting morphologies of well-lithified and poorly lithified conglomerates. A – high cliffs

and rounded tops in the Danxia landscape of Fangyan, China; B – pinnacles and fins in Putangirua

badlands, North Island, New Zealand.

Fig. 7. Spatial pattern of conglomeratic domes in Kata Tjuta massif, Australia (source: Google Earth).

Fig. 8. Massive conglomerate towers at Jianglangshan (China) have vertical cliffs more than 300 m

high (A), but their top surfaces are remarkably flat or slightly convex, as highlighted by patches of

of
vegetation (B).

ro
-p
Fig. 9. Tors in Triassic conglomerates near Teruel, central Spain.
re
Fig. 10. Distinctive crags form the top parts of hogback ridges. A – Pasterskie Skały (SW Poland) –
lP

Upper Cretaceous conglomerates; B – Babí lom (SE Czechia) – Devonian conglomerates.

na

Fig. 11. Small residual pinnacle in cemented gravels of Rodados Patagonicos formation in south
ur

Argentina. Photo by Bettina Ercolano.

Jo

Fig. 12. Large crags/tors (ruiniform relief) in conglomerates. A – Súl’ovské skaly (Slovakia), with a

natural arch along a vertical joint, B – partly collapsed tor near Cueva del Milodon (Chile). Reasons for

tor decay are unclear.

Fig. 13. Slot canyon in Taining, China, partly submerged due to reservoir building. Photo by Ren Fang.

Fig. 14. Clefts widened by weathering within a minor rock city of Kamenná hora (Czechia). Notice

their intersection at the right angle in the middle of image.

48
 Journal Pre-proof

Fig. 15. Diversity of linear incisions in Serra da Capivara (Brazil), from very narrow clefts to wide flat-

floored valleys. Note limited penetration of clefts and ravines into the plateau (compare with ground

photo, Fig. 26) (source: Google Earth).

Fig. 16. Thorough dissection of conglomerate relief in Taining, China (source: Google Earth).

Fig. 17. Horizontal caverns (bedding caves) in conglomerate cliffs of Danxia landscapes of China. A –

of
cliff face in Danxiashan, with caverns at different elevations, some apparently following the same

ro
major discontinuity; B – horizontal slot big enough to accommodate a temple (Fangyan). Behind the
-p
back wall is a passable opening to the opposite side of the hill.
re
Fig. 18. Inside Cueva del Milodon (Chile). Note people for scale.
lP

Fig. 19. Gupteshwar Cave in Pokhara, Nepal, was eroded in Quaternary conglomerate infill of the
na

Pokhara Basin.
ur

Fig. 20. Beaded tafoni (locally grading into bedding caves – lower left) in Longhushan. Photo by Liuqin
Jo

Chen.

Fig. 21. Fluted convex upper break of slope in Triassic conglomerates of Belogradchik, Bulgaria.

Fig. 22. Pedestal rocks (‘rock mushrooms’) in Triassic conglomerates of Belogradchik, Bulgaria.

Fig. 23. Influence of joint pattern on the spatial distribution of convex (domes, ridges) and concave

(clefts, valleys) landforms in Langshan, China (source: Google Earth).

49
 Journal Pre-proof

Fig. 24. Relief adjustment to joint pattern in Pedras Negras, Angola. Note differences in upland

fragmentation between the western and eastern part of the area, as well as preferential exploitation

of SW–NE (west) to SSW–NNE (east) trending master joints (source: Google Earth).

Fig. 25. Conglomerate towers of Meteora, Greece, form distinctive rock spurs along the margin of a

plateau. Note more advanced fragmentation of more distant spurs (above the town, lower left)

(source: Google Earth).

of
ro
Fig. 26. Marginal escarpment of Serra da Capivara, Brazil. The red rocks in the foreground are
-p
sandstones, overlain by white conglomerates. Note that the top of the plateau lacks bedrock

outcrops, in notable contrast to the thorough dissection present in Danxia landscapes of China.
re
lP

Fig. 27. Diversity of landforms in the dissected conglomeratic plateau near Belogradchik, Bulgaria. A –

flat-floored valley lined by continuous cliffs; B – isolated pinnacles on the interfluves; C – large
na

residual massifs (buttes); D – ruiniform relief; E – parallel arrangement of rock fins, controlled by
ur

dominant SSW–NNE joints (source: Google Earth).

Jo

Fig. 28. Satellite images of badlands in poorly lithified conglomerates. A – Putangirua pinnacles, New

Zealand; B – Miaoli County, Taiwan. Note that the occurrence of badlands is restricted to singular

valley systems and huge debris fans at the outlet of ravines in Taiwan (source: Google Earth).

Fig. 29. Erosional landscape in Maolin Scenic Area in south Taiwan represents a transition from a

typical badland dominated by pinnacles to residual relief typified by rounded domes. Photo by Jiun-

Chuan Lin.

50
 Journal Pre-proof

Fig. 30. Influence of structure on conglomeratic landforms near Sant Llorenç de Morunys, Pyrenees,

Spain. Note the transition from hogback ridges on vertical and overturned beds (middle left) to

cuesta (middle) and plateau within a shallow syncline (right).

Fig. 31. Large amphitheatre of Súl’ovské skaly in Slovakia, with conglomerate crags facing each other

on the opposite limbs of a dissected anticline.

Fig. 32. Influence of regional jointing pattern on the geomorphic landscape near Montserrat, Spain.

of
Dominant joints strike SSW–NNE (A), but their variable density results in the emergence of massive

ro
domes where joint spacing is wider (B). White rectangle on (A) shows the position of (B) image

(source: Google Earth).

-p
re
Fig. 33. Conceptual model of how rounded towers in conglomerates evolve. A – initial surface; B –
lP

slow surface lowering and the development of strength-equilibrium domes in massive bedrock

compartments; C – rapid surface lowering following uplift in jointed zones, origin of vertical cliffs, but
na

persistence of dome shapes on top of the massive compartments. BC – bedding caves, RSF – rock
ur

slope failures.
Jo

51
 Journal Pre-proof

Declaration of interests

☒ The authors declare that they have no known competing financial interests or personal
relationships that could have appeared to influence the work reported in this paper.

of
ro
-p
re
lP
na
ur
Jo

52
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 10
Figure 11
Figure 12
Figure 13
Figure 14
Figure 15
Figure 16
Figure 17
Figure 18
Figure 19
Figure 20
Figure 21
Figure 22
Figure 23
Figure 24
Figure 25
Figure 26
Figure 27
Figure 28
Figure 29
Figure 30
Figure 31
Figure 32
Figure 33