a r t i c l e i n f o
Article history:
Received 9 August 2013
Received in revised form 20 February 2014
Accepted 22 February 2014
Available online 1 March 2014
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
Landsliding is an important process inﬂuencing mountain range and
landscape evolution (Shroder and Bishop, 1998; Kirchner and Lacina,
2004; Korup et al., 2010). Large landslides can impact catchment
morphology (Korup et al., 2010), river long proﬁles (Hewitt, 1998), or
sediment delivery (Hovius et al., 1997). Landslides may block river
courses and form temporal dams (Costa and Schuster, 1988; Nicoletti
and Parise, 2002; Baroň et al., 2004; Korup, 2004a,b; Pánek et al.,
2007). Landslide dams could be rather short-lived phenomena (Ermini
and Casagli, 2003), while others seem to be stable during the whole
Holocene (Clague and Evans, 1994; Reneau and Dethier, 1996).
The interplay between drainage networks and hillslopes concerning
the role of landslides can be grouped into the slope–ﬂuvial (landslidingchannel
initiation) and ﬂuvial–slope (drainage headward extensionlandsliding)
relationships (Korup, 2005; Ng, 2006). Various causes
of landslide initiation have been documented such as earthquakes,
water-saturation of rock due to intense rainfalls, snowmelt, and, to a
minor extent, stream piracy as was demonstrated for instance from
the SE Spain (Mather et al., 2003). Korup (2004a,b) noticed avulsion
(i.e., rapid abandonment of a river channel and formation of a new
river channel in ﬂat areas) due to landslides, by which pulsed or chronic
supply of landslide debris to valley ﬂoors causes substantial aggradation
and channel instability; he distinguished three types of landslideinduced
channel avulsion: (i) upstream/backwater avulsions, (ii) contact
avulsions, and (iii) downstream/loading avulsions.
Stream piracy is a process of ﬂuvial erosion whereby headward erosion
of one stream captures the upper part of an adjacent stream (Sala,
2004). River piracy and the corresponding incision into a bedrock as a
result of an increased stream power usually induce change of hillslope
geometry and subsequent landslides (Azañón et al., 2005). River piracy
is common in tectonically active regions (e.g., Simoni et al., 2003; Stokes
and Mather, 2003) where an excess of potential energy exists. Stream
capture can be caused by headward erosion (Bishop, 1995), lateral erosion
(Douglass and Schmeeckle, 2007), sapping (Pederson, 2001), and
even regional tectonic movements when minor earth movements
change slope inclination and thus have the potential to alter stream
ﬂow (Twidale, 2004). Stream piracy was also modelled in physical
experiments (Douglass and Schmeeckle, 2007).
Stream piracy induced by landslide activity is rare in recent literature.
In Southern Alps of New Zealand, Korup and Crozier (2002) documented
river piracy of ﬁrst-order stream channels by headward
truncation and subsequent subsidence as a consequence of headward
extension of new scarplets and tensional ﬁssures. Oesleby (2005) mentioned
a landslide aided stream piracy and abandonment of Unaweep
Canyon in western Colorado.
The aim of this paper is to provide an insight into stream piracy in a
typical landslide prone region, the Outer Western Carpathians. The
paper will focus on (i) documenting a rare case of landslide-triggered
river piracy in the Malá Brodská Valley, eastern Czech Republic, (ii) understanding
the geological and geomorphic factors of the piracy, and
(iii) evaluating the consequences of the piracy, i.e., estimate the change
of local erosion rates caused by river caption.
2. Geological and geomorphic settings
The investigated site is situated in the upper part of the Malá
Brodská Valley near Nový Hrozenkov, in the central part of the
Vsetínské vrchy Mts. (Fig. 1). The highest point of the study site is the
Tanečnice Mt. (912 m asl), and the lowest is the valley ﬂoor (559
m asl). The average slope is 19.5°; maxima are located near the elbow
of capture and locally reach up to 48°. Topography of the Vsetínské
Mts. has a character of uplands and highlands up to 1024 m asl, with
major slope inclinations of 10° to 20° (Krejčí et al., 2002). It is of
structural-denudational and erosion-denudational origin; the difference
between competent and incompetent beds, mass movements as
well as ﬂuvial erosion played a basic role in forming the local relief
(Kirchner, 2000). The mountain ranges have a typical strike ENE–
WSW, predisposed by the geological structure.
The study area belongs to the Rača Unit of the Magura Group of
Nappes (Fig. 1B), which is a part of the ﬂysch zone of the Outer Western
Carpathians (Pícha et al., 2006). The ﬂysch zone is a folded, tectonically
imbricated, thin-skinned thrust stack, which was thrust over the North
European platform (Fennosarmatia) during Palaeogene and early
Neogene phases of the Alpine orogeny (Pícha et al., 2006). The study
site is located within the Vsetín Member of the Zlín Formation (Rača
Unit, Magura Group of Nappes). It is mainly composed of alternating
calcareous mudstone, shale, and metre-scale thick turbidite sandstone
layers of Eocene to early Oligocene age. The bedrock is strongly weathered
(Pícha et al., 2006). The ENE–WSW trending thrust faults and numerous
radial faults frequently occur in the area. The study area has
never been glaciated in the Quaternary (Nývlt et al., 2011), and it has
been intensively affected by shallow to deep-seated landslides throughout
the Holocene (Krejčí et al., 2002; Baroň et al., 2004; Pánek et al.,
2013). Numerous examples of stream piracy can be seen in the ﬂysch
belt area (e.g., Vojtko et al., 2012), but the relationship between the
gravity mass movements and stream capture has gained only little
attention.
3. Methods
Landslide-related erosional features and their spatial distribution,
dip direction, and dip angle of outcropping sandstones were mapped
by means of the ﬁeld geological and geomorphologic mapping in
order to reconstruct the local geological structure and leading geomorphic
processes. Bedding patterns and sedimentary structures of the
Quaternary windgap sediments were studied in a 1.1-m-deep trench
in the abandoned channel.
In order to get a precise digital elevation model (DEM), we carried
out a ﬁeld survey by using the Trimble 5503 total station. The resulting
DEM has a 1-m resolution and covers all important terrain features in
the area. With these data we were able to generate both longitudinal
and cross-sectional proﬁles of the area.
In order to investigate subsurface structure of the landslide and the
windgap, we conducted a two-dimensional (2D) electrical resistivity tomography
(ERT) survey using the ARES automatic geoelectrical system
Fig. 1. General settings of the Malá Brodská Valley: (A) location of the site within central European context (location of Fig. 1B is marked as a rectangle, the study area is marked as the
complex star); (B) geological map of the Vsetínské vrchy Mts. superimposed on the slope gradient map from the SRTM data; the brown colours represent ﬂysch rocks of the Rača Unit,
the green colours represent ﬂysch rocks of the Silesian Unit, and the extent of Fig. 3A is deﬁned by the black rectangle.
Source of data: Czech Geological Survey and USGS/NASA.
357
(GF Instruments, Czech Republic). The Wenner–Schlumberger array of
64 electrodes was used with a 3-m electrode spacing. The 237-m total
length of the section was accomplished using the roll-along method
with a 16-electrode (48-m) increment. Stacking of four pulses with a
0.5 s pulse length was used in each measured point. The maximum
depth of the apparent resistivity pseudosection was 33.7 m. Inverse
model resistivity section was produced from the apparent resistivity
pseudosection by least-square inversion method using RES2DINV software
(Loke, 1997).
The age of the landslide was estimated by AMS radiocarbon (14
C)
dating of ~2.2-m-thick ﬂuviolacustrine deposits originated as a result
of the blockage of adjacent valley by right-lateral levee of the landslide.
Sediments at the deepest part of the lake were drilled by the Eijkelkamp
peat sampler with 60 mm diameter. We dated needles of Abies alba situated
at depths of 223 and 88 cm in order to state the onset and range of
the sedimentation behind the landslide barrier. Dating was carried out
at the University of Georgia, Center for Applied Isotope Studies. Radiocarbon
dates were converted into calibrated ages using IntCal 09 calibration
curve (Reimer et al., 2009) in OxCal v 4.1.7 software (Bronk
Ramsey, 2009). Calibrated ages are presented in the form of their 1σ
probability ranges.
To compare the erosional dynamics before and after the river piracy,
we calculated a minimum denudation rate (DRmin) for the captured
catchment and neighbouring noncaptured catchment. Both catchments
were previously blocked by different phases of the described slope
deformation (captured valley in 1997 — still existing lake, the adjacent
valley — lake ﬁlled with sediments, see above). For estimation of the
denudation rate we formulated this equation:
DRmin ¼ VL=tð Þ=AC;
where DRmin (mm.ky−1
) is the minimum denudation rate in the
catchment above the particular landslide-dammed lake, VL (m3
) is
the volume of sediments trapped in the landslide-dammed lake, t
(year) is the duration of sedimentation, and AC (km2
) is the area of
the catchment above the particular landslide-dammed lake. The duration
of sedimentation (t) was acquired as a difference between medians
of the probability density functions of the calibrated radiocarbon dates.
Fig. 2. Airborne oblique photos: (A) of the entire studied area from the NW, (B) of the 1997 landslide from the east, and (C) detail of the depletion zone of the recent landslide from the SW.
Photos by Ivo Baroň.
358
Medians were chosen as the most appropriate central values for calibrated
dates (Telford et al., 2004; Engel et al., 2010).
Bathymetry of the existing lake was measured in 2009 (the 2009
bottom and the original 1997 ground-surface levels) and again in
2012, using probe and Topcon 232 total station. Thickness of the sedimentary
inﬁll of the silted lake was measured in 2010 (using the same
equipment). Volumes of the sediments in both lakes were calculated
from the acquired digital depth models in Surfer 8 (Golden Software).
To get the duration of sedimentation in the lake completely ﬁlled with
ﬂuviolacustrine deposits, we used medians of the calibrated 14
C dates
as a robust central-point estimate of the age (Telford et al., 2004).
Area of both catchments was acquired in ArcGIS 10 (Esri) from the
DEM mentioned above.
4. Results
4.1. Local topography
The studied case of stream piracy has evolved in a catchment for as
large of only about 1 km2
. From the ﬁeld survey it was possible to identify
the following features related to stream piracy: an abandoned
valley, a wind gap, a ridge, old inactive landslides and a landslide
reactivated in 1997, an elbow of capture, deeply incised upper part of
the captured stream, remnants of a strath, and a knickpoint in the abandoned
valley where a small original tributary overtook the role of the
active stream (Figs. 2–4).
The abandoned valley is about 400 m long, trending NNE–SSE, widely
opened, and mild with a dry bottom of the former streambed. The maximum
slope gradient of the abandoned valley is about 7–10°, and the
abandoned valley has a U-shaped cross-section. Looking upward the
abandoned channel suddenly ends in the wind gap, which is situated
12 m above the active streambed and which indicates the minimum
depth of erosion since the river piracy. The ridge is about 130 m long,
N–S trending and it constitutes a local divide between the abandoned
valley and the main active one. The ridge shape is not strictly controlled
by competent sandstone beds. The upper part of the captured stream is
deeply incised in the bedrock; the incision is the most intense at the
junction with the main stream and it disappears upward in 250 m.
The longitudinal channel proﬁle is inﬂuenced by outcrops of competent
sandstone beds, which form series of cascades. At the level of the abandoned
valley ﬂoor upward, the strath terrace occurs at western slopes
along the pirated stream; this is an erosional remnant of the former
streambed. The strath terrace is only minor with a maximum width of
2–3 m.
The abandoned valley is dry only in its 150-m-long upper course.
Two minor tributary streams, which joined the original (now pirated)
main stream a few hundreds of metres downward from the wind gap,
ﬂow in the valley. The resulting stream has much lower erosive capacity
than the main active stream. This is well evident at the place of the junction
of the remnant and the main active streams. Several strath terraces
of about 5 m elevated above the riverbed of the active stream are cut by
the residual stream in the abandoned valley, which erodes the bottom
of the abandoned valley ﬂoor. Headward erosion by the residual stream
continues progressively forming the strath terraces from the smooth
ﬂoor of the abandoned valley. At the place of most intense headward
erosion an ~1.5-m-high knickpoint occurs (Fig. 4).
The eastern slopes of the main active valley are formed by distinct,
deep-seated landslides of different morphology and activity state
(Figs. 2 and 3). East of the site of the capture, a large complex ﬂowlike
landslide is situated; the landslide was activated on 7 July 1997 following
distinct heavy rainfalls and ﬂoods in the eastern Czech Republic
(Rybář and Stemberk, 2000; Krejčí et al., 2002). The landslide is 685 m
Fig. 3. Topography of the study site: (A) DTM of the upper part of the catchment of Malá Brodská River with major studied features; (B) detail view of the active and abandoned valleys,
accumulation zones of the active and dormant landslides, dammed lakes, and proﬁles.
Source of data: ZABAGED.
359
long (total volume of ~0.3 × 106
m3
), and it is situated on the westfacing
slope of the Zadní Kyčera hill (802 m asl). It has a distinct
trough-like morphology and extremely low width-to-length ratio
(maximum width of landslide is ~50 m). The wedge-like morphology
of the detachment zone of the landslide area was controlled by the intersection
of steeply inclined bedding planes and a normal fault striking
ENE–WSW. A basal sliding surface is exposed in the upper part of the
slope, where the landslide mass liqueﬁed and transformed into an
earthﬂow. Several 2–6 m high levees occur along the middle part of
the landslide. The lower part of the earthﬂow completely buried the valley
ﬂoor and caused blockages of two adjacent streams. One of these
blockages emerged during the July 1997 event and still accommodates
a small permanent lake. The lake is recently about 20 × 10 m in size
and 3 m deep. An older lake (paleolake) is situated upward of the
main active stream (Fig. 3B) and it is now completely ﬁlled with
ﬂuviolacustrine deposits. The recent landslide developed within more
extensive, old deep-seated landslides affecting ca. 20 ha of slope
(Figs. 2 and 3). Based on their topography, they are both rotational
and at least two distinct bodies could be distinguished. Their geomorphic
features are, however, very smooth in contrast to the recent
landslide.
Active processes, related to the 1997 landslide, changed some original
features related to stream piracy. In particular, a part of the main active
valley was ﬁlled by the landslide's accumulations and the entrance
to the gorge was ﬁlled with water thus reducing its relative depth
(Fig. 4). More recent processes also include subsequent incision into
the landslide body forming a new course of this stream.
4.2. Geological structure
The local geology of the investigated area consists of steeply dipping,
almost vertical, competent sandstone beds up to 2 m thick and less competent
calcareous shale and claystone beds up to 10 m thick, alternating
here in several similar sequences. The general strike of the ﬂysch beds is
about ENE–WSW, oblique to the river channels. The competent sandstones
tend to build ridges and act as protective dams in a newly incised
Fig. 4. Topographic proﬁles of the abandoned and active valleys with marked major studied features: (A) longitudinal proﬁle, (B and C) cross proﬁles; for their location see Fig. 3B.
Fig. 5. Interpreted ERT proﬁle indicating the local geological structure, presence of the subvertical N-striking fault, old and recent landslide accumulations, sediments of the abandoned
channel, colluvium, and intense weathering; for localization of the ERT proﬁle, see Fig. 3.
360
piracy stream signiﬁcantly reducing erosion. On a broader scale, the area
is affected by south-dipping thrust faults with general strike ENE–WSW
and NNW–SSE trending radial faults.
The ERT survey (Fig. 5) revealed distinct, high resistivity (~280 to
~650 Ω.m) domains, which are interpreted as thick layers of unweathered
sandstones. These domains alternate with low resistivity (~10 to
~70 Ω.m) ones interpreted as shales and calcareous mudstones. Being
steeply inclined to ESE in the WNW–ESE proﬁle, their boundaries correspond
to the regional dips of the ﬂysch beds (toward the south). This
layered pattern is disturbed in the middle part of the ERT proﬁle, at
~100 m distance, by a subvertical zone, which is interpreted as a radial
fault (running approximately N–S, consistent with the geology of the
broader region). The resistivity values are highly variable, ranging
from ~30 to ~250 Ω.m, in a near-surface domain, ~3 to ~5 m thick,
which are interpreted as colluvium or landslides (Fig. 5). The relatively
shallow-seated 1997 landslide as well as the deep-seated old landslide
can be visible in the ERT proﬁle.
A trench, dug out in the upper central part of the abandoned valley
about 10 m below the wind gap (Figs. 3 and 4), revealed about a 20cm-thick
layer of clast-supported gravel composed of imbricated,
semiangular sandstone clasts, generally 10–20 cm in diameter (Fig. 6).
The matrix of the sediment contains (probably secondarily) a signiﬁcant
amount of clay. This layer is interpreted as ﬂuvial sediment (stream
channel residue). The ﬂuvial sediment lies at 70 to 90 cm deep below
the ground surface, overlying ﬂysch mudstone of the Vsetín Member.
Above it, about 60-cm thick, unsorted muddy–stony colluvium (clastrich
muddy diamicton according to Moncrieff, 1989) with angular clasts
up to a few centimetres in diameter is deposited, overlain by about 10cm-thick
soil with O and A horizons.
4.3. Slope-failure history
Morphological evidence shows that the slope transects with the July
1997 failure is particularly prone to recurrent landslides/earthﬂows. A
dammed palaeolake (Fig. 3B), now completely ﬁlled by 2.2-m-thick
sediments, was formed as a result of some earlier phases of landslide
activity. Dating of the basal lacustrine deposits to 1170–1060 cal BP
(Fig. 7) provides a minimum age estimate of a major ancient landslide
that occurred at the site and was likely the immediate trigger responsible
for the river piracy as a consequence of the permanent
landslide-induced shift of the stream against the opposite eastfacing
slope. Besides lacustrine and palustrine deposits, two intercalations
of alluvial gravels (200–212 cm; 69–88 cm) reveal high energy
sedimentary inputs to the reservoir (Fig. 7). The lake is still
subject to ﬂuviolacustrine sedimentation, which is suggested by
almost recent age (290–0 cal BP) of organic particles deposited at
the depth of 88 cm below the surface.
Sediment volumes trapped in both lakes (Fig. 8) enabled a rough
estimation of the denudation rates in the contributing subcatchments
of the pirating stream. Areas of both subcatchments are very similar
(0.356 km2
for the adjacent catchment and 0.352 km2
for the captured
Fig. 6. Schematic sketch of the sedimentary log of the trench with main lithologic characteristics
evidencing the presence of ﬂuvial deposits of the original brook; for localization of
the trench, see Figs. 3 and 4.
Fig. 7. Schematic cross-section of the upper dammed lake and sedimentary log with dated ﬂuviolacustrine deposits: 1 — alluvial channel deposits; 2 — high energy sandy and gravely deposits;
3 — lacustrine silts; 4 — organic-rich silts deposited in swampy environment; the calibrated ages of the deposited organic matter are presented.
361
catchment), whereas the size of the palaeolake is ~3.5 times larger than
the 1997 lake (660 and 190 m2
for the palaeolake and existing lake,
respectively). Calculated denudation rate (DRmin) for the captured
catchment is 18.8 mm.ky−1
. The result was acquired from sediment
volume accumulated between the years 1997 and 2012. Denudation
rate for the adjacent (noncaptured) catchment is 3.6 mm.ky−1
and
it was calculated from sediment volume accumulated between
180 cal BP (median of the probability density function of the calibrated
radiocarbon date from the basal lacustrine deposits) and
the present.
5. Discussion
The presented site in the Malá Brodská Valley in the ﬂysch belt of the
Outer Western Carpathians documents well a rare case of stream piracy
induced by gravitational mass movements. Our reconstruction shows
that the accumulations of deep-seated landslides pushed one of two
subparallel river channels with different erosion activity westward.
This has forced an intensive lateral erosion toward the recently abandoned
valley and caused the piracy (Fig. 9). These geomorphic features
of the piracy here are of local scale, but such a complex topography at a
small area is rarely seen in the ﬂysch Carpathians.
The most critical factor of the river piracy seems to be the occurrence
of two subparallel river channels with different erosion activities; the
eastern (recently active) channel performed much deeper incision,
which was probably enhanced by the estimated N–S striking fault indicated
by its geomorphic appearance on DTMs and detected by the ERT
survey. The contrast between incisions of the eastern (pirating) and
western (pirated) channels was also accentuated by higher ﬂow rates
of the eastern one as a consequence of its larger catchment. As a result,
the elevation gradient between these parallel channels was about 12 m
in a horizontal distance of about 50–80 m.
Accumulations of the two dormant old landslides and of the original
ﬂow-like landslide, which was reactivated in 1997, pushed the stream
westward and forced intensive lateral erosion of the active channel toward
the recently abandoned one. When the lateral erosion reached
the western channel, the less active river changed the course out of
the former valley. Later on, intensive retrograde erosion in the elbow
of capture and also within the upper portion of the western catchment
occurred. The coarse ﬂuvial deposits of the abandoned channel were
subsequently buried with about 60-cm-thick colluvium.
As landslides directly controlled the evolution of the studied case of
river piracy and the paleoﬂuvial deposits of the abandoned channel
were not datable, we tried to achieve information on the age of stream
truncation through dating the landslides. The oldest achieved age of the
Fig. 8. (A, B) Sedimentation in the recent landslide-dammed lake reﬂected in the bathymetry changes between the years 1997 (initial state) and 2012. (C) Depth of sedimentary ﬁll in the
palaeolake.
362
complex ﬂow-like landslide was about 1170–1060 cal BP by indirect
dating of the ancient landslide dam (dating of basal lacustrine sediments,
which evolved after the damming). This landslide was, however,
preceded by two deep-seated (recently inactive) landslides to the south
of it, which could unfortunately not be dated because no datable deposits
were found there. Based on the rather smooth shape of the old
landslides and the analogy with other dated landslides of the same magnitude
in this area (e.g., Kobylská landslide; Baroň et al., 2004), we can
only speculate about the age of the river piracy to be of early Holocene
and/or late Glacial.
Although this case of river piracy induced by landslides is rather
small, it documents the high rate of recent erosion and slope processes
in the area of the mountains of the ﬂysch belt of the Outer West
Carpathians. Topographic gradient between the pirating and pirated
streams enhanced deep incision and relatively high erosion rates of
the western (pirated) stream. Recent high erosional dynamics in the
western catchment is still detectable by comparing the denudation
rates calculated from sediments trapped in the landslide-dammed
lakes (Fig. 8). We consider the acquired denudation rates to be minimum
values because an unknown amount of sediments might be missing
as a result of possible suspended load outﬂow in both lakes and
because of the higher rate of the sediment compaction in the palaeolake.
In the western (captured) catchment (18.8 mm.ky−1
), the value is ﬁve
times higher than in the adjacent catchment (3.6 mm.ky−1
). Though
some levels of uncertainty exist in the calculated values, we think it
clearly shows the higher recent erosional dynamics in the captured
stream. Our estimation of the erosion rate of the captured catchment
is, however, an order of magnitude smaller than the recent denudation
rate within a similar catchment of the Bystřička River also situated in the
Vsetínské Mts. Baroň et al. (2010) estimated it by balancing the artiﬁcial
dam clastic sediments with the area of the catchment above and assumed
it to be between 82.1 and 28.7 mm.ky−1
by not considering
and considering porosity of the dam deposit, respectively. Contrary to
our study site, the Bystřička River catchment is much larger and comprises
large portions of arable land and pastures. For a longer period,
Bíl et al. (2004) calculated the erosion in the ﬂysch belt of the Outer
Fig. 9. Interpretative model of the evolution of river piracy in the Malá Brodská Valley: (A) reconstructed topography of the original state with a trace of the suspected fault that probably
enhanced faster erosion in the eastern valley, (B) the oldest and largest landslide slightly pushed the eastern valley westward, (C) younger (dated) landslides pushed the eastern valley
more to the west and intensiﬁed lateral erosion toward the recently abandoned valley, which (D) caused capturing of the western stream and caused the piracy (present day topography).
363
West Carpathians since the Sarmatian (Neogene) by analysing anisotropy
of magnetic susceptibility, vitrinite reﬂectance, and K/Ar dating of
volcanic rocks to be 1400 m; i.e., the mean erosion rate was 102–
127 mm.ky−1
. These values are higher than the recent erosion rates obtained
from the sediment accumulation in the Bystřička River reservoir
probably as a result of more sparse vegetation cover in the ﬂysch
Carpathians during cold periods of the Pleistocene, which could not
protect the land surface from sheet erosion. Isolated forest vegetation
similar to the present Siberian coniferous taiga surrounded by a mosaic
of steppe communities and tundra patches occurred in the area during
the last glacial period (Jankovská and Pokorný, 2008). The high erosion
rates calculated by Bíl et al. (2004) are probably inﬂuenced by periods of
high tectonic activity of the Western Carpathians in the Neogene
(Golonka and Pícha, 2006).
Although the recent erosion rate of the pirated stream in the Malá
Brodská Valley is much lower than the previously published values, it
is slightly higher than the range of late Holocene denudation rates of
2.5–13.4 mm.ky−1
calculated for several Holocene landslide-dammed
catchments in the Outer Western Carpathians by Pánek et al. (2010).
All of these values are probably low because of the dense forestation
of the studied catchments in the Holocene, which considerably slows
down sheet erosion, whereas the piracy-induced intensiﬁcation of erosion
had only a local occurrence.
6. Conclusions
This paper presents results of the ﬁeld investigations of river piracy
in the Malá Brodská Valley in the ﬂysch belt of the Outer Western
Carpathians controlled by mass movement processes. Based on the
ﬁeld geological, geomorphological, and geophysical data, we could
roughly summarize a scenario of this river piracy case. The landslides'
accumulations pushed the more active river westward and forced
intensive lateral erosion towards the recently abandoned valley.
Except for the landslide processes, the occurrence of two subparallel
river channels with different erosion activity and the presence of a Nstriking
fault, accentuated by higher ﬂow rates of the eastern channel
as a consequence of its larger catchment area, were the most critical
factors of the river piracy.
Later, intensive retrograde erosion in the elbow of capture and also
within the upper portion of the western catchment occurred. Slope
mass-transport processes subsequently buried the coarse ﬂuvial
deposits of the inactive channel with about 60-cm-thick colluvium.
The maximum age of the river piracy is estimated to be of late Glacial
and/or early Holocene. The presented case of river piracy as a result of
slope failures documents recent erosion rates in the area of the mountains
of the ﬂysch belt of the Outer Western Carpathians.
Acknowledgements
The investigations were ﬁnanced by the project ISPROFIN
215124-1 of the Ministry of the Environment, Czech Republic;
‘Slope Failures in the CR’ conducted by the Czech Geological Survey;
by the University of Ostrava Foundation project SGS15/PřF/2013;
and by the project of the Ministry of the Interior of the Czech
Republic Nr. VG20102015057. Preparation of the paper was cofunded
by ﬁnancial subsidy from the RRC /04/2012: ‘Support for
science and research in Moravian-Silesian Region 2012’ programme
provided by the Moravian-Silesian Region. The authors
appreciate the help of David Konečný, Jan Kubeček, Dorota
Rabinská, and Petra Tománková during the ﬁeld surveys and geophysical
investigations. The authors would like to acknowledge
Daniel Nývlt, Richard A. Marston and two anonymous referees for
their fruitful comments and suggestions, which helped improve
the manuscript's readability.
References
Azañón, J.M., Azor, A., Pérez-Peña, J.V., Carrillo, J.M., 2005. Late Quaternary large-scale
rotational slides induced by river incision: the Arroyo de Gor area (Guadix basin, SE
Spain). Geomorphology 69, 152–168.
Baroň, I., Cílek, V., Krejčí, O., Melichar, R., Hubatka, F., 2004. Structure and dynamics of
deep-seated slope failures in the Magura Flysch Nappe, Outer Western Carpathians
(Czech Republic). Nat. Hazards Earth Syst. Sci. 4, 549–562.
Baroň, I., Baldík, V., Fifernová, M., 2010. Preliminary assessment of a recent denudation
rate of the ﬂysch belt of Outer Western Carpathians — case study: Bystřička River
catchment in Vsetínské Hills. Geol. Res. Moravia Silesia Year 2009 17 (1–2), 10–13
(in Czech).
Bíl, M., Krejčí, O., Franců, J., Hrouda, F., Přichystal, A., 2004. Approximation of the missing
eroded sediments in the Bílé Karpaty Unit (Outer Western Carpathians). Stud.
Geomorphol. Carpatho Balc. 38, 58–66.
Bishop, P., 1995. Drainage rearrangement by river capture, beheading and diversion. Prog.
Phys. Geogr. 19 (4), 449–473.
Bronk Ramsey, C., 2009. Bayesian analysis of radiocarbon dates. Radiocarbon 51, 337–360.
Clague, J.J., Evans, S.G., 1994. Formation and failure of natural dams in the Canadian
Cordillera. Geol. Surv. Can. Bull. 464, 35.
Costa, J.E., Schuster, R., 1988. The formation and failure of natural dams. Geol. Soc. Am.
Bull. 100, 1054–1068.
Douglass, J., Schmeeckle, M., 2007. Analogue modeling of transverse drainage mechanisms.
Geomorphology 84, 22–43.
Engel, Z., Nývlt, D., Křížek, M., Treml, V., Jankovská, V., Lisá, L., 2010. Sedimentary evidence
of landscape and climate history since the end of MIS 3 in the Krkonoše Mountains,
Czech Republic. Quat. Sci. Rev. 29, 913–927.
Ermini, L., Casagli, N., 2003. Prediction of the behaviour of landslide dams using a geomorphological
dimensionless index. Earth Surf. Process. Landf. 28, 31–47.
Hewitt, K., 1998. Catastrophic landslides and their effects on the Upper Indus streams,
Karakoram Himalaya, northern Pakistan. Geomorphology 26, 47–80.
Hovius, N., Stark, C.P., Allen, P.A., 1997. Sediment ﬂux from a mountain belt derived from
landslide mapping. Geology 25, 231–234.
Jankovská, V., Pokorný, P., 2008. Forest vegetation of the last full-glacial period in the
Western Carpathians (Slovakia and Czech Republic). Preslia 80, 307–324.
Kirchner, K., 2000. Geomorfologie. In: Pavelka, J., Trezner, J. (Eds.), Příroda Valašska. Český
Svaz Ochránců Přírody, Vsetín (488 pp.).
Kirchner, K., Lacina, J., 2004. Slope movements as a disturbance agent increasing heterogeneity
and biodiversity of landscape in eastern Moravia. Ekologia 23 (1), 94–103.
Korup, O., 2004a. Geomorphometric characteristics of New Zealand landslide dams. Eng.
Geol. 73, 13–35.
Korup, O., 2004b. Landslide-induced river channel avulsions in mountain catchments of
southwest New Zealand. Geomorphology 63, 57–80.
Korup, O., 2005. Geomorphic imprint of landslides on alpine river systems, southwest
New Zealand. Earth Surf. Process. Landf. 30, 783–800.
Korup, O., Crozier, M., 2002. Landslide types and geomorphic impact on river channels,
Southern Alps, New Zealand. In: Rybář, J., Stemberk, J., Wagner, P. (Eds.), Landslides
— Proceedings of the 1st European Conference on Landslides, Prague, Czech
Republic, June 24–26. Swets & Zeitlinger, Lisse, The Netherlands, pp. 233–238.
Korup, O., Densmore, A.L., Schlunegger, F., 2010. The role of landslides in mountain range
evolution. Geomorphology 120, 77–90.
Krejčí, O., Baroň, I., Bíl, M., Jurová, Z., Hubatka, F., Kirchner, K., 2002. Slope movements in
the Flysch Carpathians of eastern Czech Republic triggered by extreme rainfalls in
1997: a case study. Phys. Chem. Earth 27, 1567–1576.
Loke, M.H., 1997. Res2DINV Software User's Manual.
Mather, A.E., Grifﬁths, J.S., Stokes, M., 2003. Anatomy of a ‘fossil’ landslide from the
Pleistocene of SE Spain. Geomorphology 50, 135–149.
Moncrieff, A.C.M., 1989. Classiﬁcation of poorly-sorted sedimentary rocks. Sediment. Geol.
65, 191–194.
Ng, K.Y., 2006. Landslide locations and drainage network development: a case study of
Hong Kong. Geomorphology 76, 229–239.
Nicoletti, P.G., Parise, M., 2002. Seven landslide dams of old seismic origin in southeastern
Sicily (Italy). Geomorphology 46, 203–222.
Nývlt, D., Engel, Z., Tyráček, J., 2011. Pleistocene glaciations of Czechia. In: Ehlers, J.,
Gibbard, P.L., Hughes, P.D. (Eds.), Quaternary Glaciations — Extent and Chronology,
A Closer Look. Developments in Quaternary Science, 15. Elsevier, Amsterdam,
pp. 37–46.
Oesleby, T.W., 2005. Abandonement of Unaweep Canyon, western Colorado; stream
piracy aided by major landslide. Geol. Soc. Am. Abstr. Programs 37 (7), 297.
Pánek, T., Smolková, V., Hradecký, J., Kirchner, K., 2007. Landslide dams in northern part of
Czech Flysch Carpathians: geomorphic evidences and imprints. Stud. Geomorphol.
Carpatho Balc. 41, 77–96.
Pánek, T., Smolková, V., Hradecký, J., 2010. Reconstruction of a Holocene average
catchment denudation from the landslide-dammed lakes in the Outer Western
Carpathians. Geophys. Res. Abstr. 12, EGU2010–EGU9625.
Pánek, T., Smolková, V., Hradecký, J., Baroň, I., Šilhán, K., 2013. Holocene reactivations
of catastrophic complex ﬂow-like landslides in the Flysch Carpathians (Czech
Republic/Slovakia). Quat. Res. 80, 33–46.
Pederson, D.T., 2001. Stream piracy revisited: groundwater-sapping solution. GSA Today
11, 4–10.
Pícha, F.J., Stráník, Z., Krejčí, O., Kirchner, K., 2006. Geology and Hydrocarbon Resources of
the Outer Western Carpathians and their Foreland, Czech Republic. In: Golonka, J.,
Pícha, F.J. (Eds.), The Carpathians and Their Foreland: Geology and hydrocarbon resources.
AAPG Memoir, pp. 49–175.
Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey, C.,
Buck, C.E., Burr, G.S., Edwards, R.L., Friedrich, M., Groothes, P.M., Guilderson, T.P.,
364
Hajdas, I., Heaton, T.J., Hogg, A.G., Hughen, K.A., Kaiser, K.F., Kromer, B., McCormac,
F.G., Manning, S.W., Reimer, R.W., Richards, D.A., Southon, J.R., Talamo, S., Turney,
C.S.M., Van Der Plicht, J., Weyhenmayer, C.E., 2009. IntCal09 and Marine09 radiocarbon
age calibration curves, 0–50,000 years cal BP. Radiocarbon 51, 1111–1150.
Reneau, S.L., Dethier, D.P., 1996. Late Pleistocene landslide-dammed lakes along the Rio
Grande, White Rock Canyon, New Mexico. Geol. Soc. Am. Bull. 108, 1492–1507.
Rybář, J., Stemberk, J., 2000. Avalanche-like occurrences of slope deformations in the
Czech Republic and coping with their consequences. Landslide News 13, 28–33.
Sala, M., 2004. River capture. In: Goudie, A.S. (Ed.), Encyclopedia of Geomorphology.
Routlege, London, pp. 860–861.
Shroder Jr., J.F., Bishop, K.P., 1998. Mass movement in the Himalaya: new insights and research
directions. Geomorphology 26, 13–35.
Simoni, A., Elmi, C., Picotti, V., 2003. Late Quaternary uplift and valley evolution in the
Northern Apennines: Lamone catchment. Quat. Int. 101–102, 253–267.
Stokes, M., Mather, A.E., 2003. Tectonic origin and evolution of a transverse
drainage: the Rio Almanzora, Betic Cordillera, southeast Spain. Geomorphology
50, 59–81.
Telford, R.J., Heegaard, E., Birks, H.J.B., 2004. The intercept is a poor estimate of a calibrated
radiocarbon age. The Holocene 14, 296–298.
Twidale, C.R., 2004. River patterns and their meaning. Earth Sci. Rev. 67, 159–218.
Vojtko, R., Petro, L., Benová, A., Bóna, J., Hók, J., 2012. Neotectonic evolution of the northern
Laborec drainage basin (northeastern part of Slovakia). Geomorphology 138,
276–294.
365