Variations in bar material grain-size and hydraulic conditions of managed
and re-naturalized reaches of the gravel-bed Bečva River
(Czech Republic)
Václav Škarpich a,
⁎, Tomáš Galia a
, Stanislav Ruman a
, Zdeněk Máčka b
a
Department of Physical Geography and Geoecology, Faculty of Science, University of Ostrava, Chittussiho 10, Slezská Ostrava, Ostrava 710 00, Czech Republic
b
Department of Geography, Faculty of Science, Masaryk University, Kotlářská 2, 611 37 Brno, Czech Republic
H I G H L I G H T S
• Channel geometry corresponded well
with modelled hydraulic variables.
• Bar grain size and simulated hydraulics
did not produce any clear relationships.
• Sediment inputs and disconnectivities
gave better explanation of the grain size.
G R A P H I C A L A B S T R A C T
a b s t r a c ta r t i c l e i n f o
Article history:
Received 12 February 2018
Received in revised form 22 August 2018
Accepted 24 August 2018
Available online 26 August 2018
European multi-thread rivers have undergone rapid morphological changes during past centuries due to the extensive
direct and indirect human impacts on ﬂuvial systems. As a consequence, we can identify altered patterns
of bed sediment calibre reﬂecting disturbed sediment connectivity and modiﬁed ﬂow hydraulics. Changes in the
grain-sizes of samples collected on 68 gravel bars in August 2015 were studied along 14.0-km river reach of the
Bečva River (Outer Western Carpathian Mts., Czech Republic). The grain-size characteristics obtained were
confronted with modeled ﬂow hydraulics and the present stage of the channel. The studied channel reach is presently
characterized by several distinctive sections: for a long time (ca. 100 years) regulated single channel sections
with artiﬁcial bank stabilizations incised several meters in the ﬂoodplain and by contrast, multi-thread
channel patterns of two sections, which have witnessed retrograde development after large ﬂoods in 1997 and
2010 with 100- and 50-year recurrence intervals, respectively. The present channel behaviour of managed (regulated)
and re-naturalized (multi-thread) river sections corresponded well with the modeled hydraulic variables
for one-year discharge recurrence interval. Especially, re-naturalized river sections showed lower values of ﬂow
competence which facilitated the deposition of sediment material in the form of gravel bars. The high occurrence
of lateral sediment sources (e.g., tributaries, bank failures) together with sediment disconnectivities (e.g., boulder
ramps) in the longitudinal river reach were observed, and grain-size parameters did not particularly reﬂect the
Keywords:
Multi-thread river
Channelization
Gravel bar grain-size
Channel hydraulics
Sediment connectivity
The Bečva River
Science of the Total Environment 649 (2019) 672–685
⁎ Corresponding author.
E-mail address: vaclav.skarpich@osu.cz (V. Škarpich).
https://doi.org/10.1016/j.scitotenv.2018.08.329
0048-9697/© 2018 Elsevier B.V. All rights reserved.
Contents lists available at ScienceDirect
Science of the Total Environment
journal homepage: www.elsevier.com/locate/scitotenv
hydraulic conditions. Especially tributaries as sediment inputs had signiﬁcant effect on bar grain size and increase
of channel diversity, although, in general results indicate a gradual downstream ﬁning.
© 2018 Elsevier B.V. All rights reserved.
1. Introduction
Multi-thread river channels are integral parts of ﬂuvial systems,
which connect the mountainous landscapes with lowland areas
(Schumm, 1977). In the past few centuries, multi-thread river channels
have undergone rapid changes induced by human activities (Gregory,
2006; Kondolf, 1997). A large number of studies across the globe have
identiﬁed the most serious issues of multi-thread channel pattern transformation
with regards to progressive narrowing (Korpak, 2007;
Liébault and Piégay, 2002; Rinaldi et al., 2005; Surian and Rinaldi,
2003; Škarpich et al., 2013, 2016a, 2016b; Zawiejska and Wyżga,
2010) and incision (Kondolf et al., 2002; Martín-Vide et al., 2010;
Preciso et al., 2012; Rovira et al., 2005; Surian and Rinaldi, 2003;
Uribelarrea et al., 2003; Wyżga, 2001; Zawiejska and Wyżga, 2010).
These alterations resulted in channel ﬂow hydraulics changes and
brought a new set of boundary conditions for erosion, transport and deposition
processes (Knighton, 1999; Madej, 1995), in which amount,
calibre and sorting of sediment as well as its spatial distribution are adjusted
to local variations in ﬂow competence (Powell, 1998). The spatial
distribution and sorting pattern of sediment have important consequences
and implications for the functioning of river ecosystems,
e.g., the potential for ﬁsh spawning or the suitability of the riverbed as
a habitat for benthic macroinvertebrates (Vannote et al., 1980).
Sediment transport within river systems has been described as a cascade
system (Schumm, 1977). It can be thought of as a threedimensional
system in the longitudinal, lateral and vertical directions
(in the sense of Brierley et al., 2006; Fryirs, 2013; Fryirs et al., 2007;
Hooke, 2003). Bed sediment changes along altered or natural river
channels have been relatively well recognized (Rengers and Wohl,
2007; Surian, 2002; Zawiejska et al., 2015). In contrast, much less is
known about sediment distribution and sorting patterns linkages to
speciﬁc values of hydraulic parameters in channels.
A characteristic feature of many alluvial rivers is the gradual decrease
in bed material grain-size in the downstream direction, which is called
‘downstream ﬁning’ (Gomez et al., 2001). Two generic mechanisms
have been invoked to explain this phenomenon: particle abrasion and hydraulic
sorting. The pioneering observations of Sternberg (1875) in a
260 km reach of the Rhine River described an exponential downstream
decrease in grain-size due to abrasion. The derived exponential law
takes the form:
Di ¼ D0e−KL
; ð1Þ
where Di is the particle diameter (m), D0 represents the initial diameter of
the particle (m) transported for distance L (m), and K is the coefﬁcient of
reduction for given rock resistance. Based on this, abrasion depends on a
variety of factors, including the lithology of the material, climate conditions
in the basin, human impact, etc. Hydraulic sorting is promoted by
the differential entrainment, transport, and deposition of particles, as a
function of their size and shape (Gomez et al., 2001; Russell, 1939).
There has been discussion about the extent to which condition of equal
mobility or departures from it dominate (e.g., Andrews, 1983; Gomez,
1995; Gomez et al., 2001; Wilcock, 1992). The consensus is that there is
a tendency for particles of different sizes to be equally mobile when the
boundary shear stress greatly exceeds the threshold for motion and that
at lower shear stresses, sorting is much less pronounced than would be
the case if the size fractions were transported independently
(e.g., Wilcock and McArdell, 1993; Gomez, 1995; Gomez et al., 2001).
However, downstream coarsening in some river reaches was observed
for gravel-bed alluvial rivers (Dawson, 1988; Knighton, 1980;
Powell, 1998). These anomalies are the result of (dis)connectivity
(Schumm, 1977; Fryirs et al., 2007; Galia et al., 2016), which disrupts
lateral, vertical and/or longitudinal linkages in sediment ﬂux and affects
the sediment cascade. The downstream ﬁning disruption of gravel-bed
rivers occurs not only in response to sediment supply from tributaries
and other lateral inputs but also in response to anthropogenic interventions
in the ﬂuvial system (Surian, 2002; Fryirs et al., 2007; Ondráčková
and Máčka, 2018). Additionally, a conceptual framework identifying
the status of river reaches and linkages between them were proposed
by Hooke (2003). Based on this, the types of connectivity including
unconnected, partially and potentially connected, connected and
disconnected reaches were identiﬁed. It has the implications for
understanding the sensitivity to perturbations and channel
development.
A characteristic feature of low gradient and wide alluvial channels
is the accumulation of sediment which is created by reduced ﬂow
conveyance (Powell, 1998). In contrast, some channel reaches accelerate
the downstream movement of sediments (Fryirs et al., 2007).
These boosters are represented by natural predispositions (e.g., by
the presence of gorges or bedrock reaches) or as the consequences
of regulation works. River channelization includes the narrowing
and shortening of the channel together with the steepening of bed
gradient, which lead to the increase of ﬂow velocities and forces acting
on stream bed and banks (Fryirs et al., 2007; Škarpich et al.,
2016a; Wyżga, 1993).
The potential for incision and sediment transport can be assessed
via the unit stream power (ω) or the bed shear stress (τb). Many studies
(Krapesch et al., 2011; Magilligan, 1992; Nardi and Rinaldi, 2015;
Zawiejska et al., 2015) have conceptualized this issue via the
calculated parameters of ω and τb, because obtaining the input
variables is easy. The parameter ω (W m−2
) was introduced by
Bagnold (1966):
ω ¼ ρgQSð Þ=b; ð2Þ
where ρ is the density of water (kg m−3
), g is the acceleration due to
gravity (9.8 m s−1
), Q is the discharge (m3
s−1
), S is the channel slope
(m m−1
), and b is the channel width (m).
The parameter τb (N m−2
), another fundamental variable in
assessing ﬂow conditions, is deﬁned as:
τb ¼ ρgRS; ð3Þ
where R is the hydraulic radius (m). In the past few decades, one- or
two-dimensional models have allowed us to calculate these hydraulic
parameters (Czech et al., 2016; Pasternack et al., 2006).
This study is focused on the grain-size variation of 68 gravel bars in a
14-km long unconﬁned reach of the gravel-bed Bečva River (Western
Carpathians, Czech Republic) and its confrontation with modeled ﬂow
hydraulics and the present state of the channel. In the 19th century,
the Bečva River had multi-thread characteristics over the entire study
reach, which was converted into a uniform single-channel by intensive
channelization works at the turn of 20th century. These changes promoted
sediment conveyance, acting as boosters and causing a gradual
incision of several meters into the ﬂoodplain. The high-magnitude
ﬂoods in 1997 and 2010 partially destroyed artiﬁcial channelization
structures in some sections. This was accompanied by notable channel
673V. Škarpich et al. / Science of the Total Environment 649 (2019) 672–685
widening in these sections and development of gravel bars. This study
aims to:
(i) identify contemporary general patterns of channel
morphology, lateral sediment inputs (connectivities), potential
disconnectivities in sediment ﬂux and erosion/deposition
processes in the channel of the laterally unconﬁned river val-
ley,
(ii) analyse and describe the variations in the grain-size parameters
of gravel bars in the context of general patterns of channel
morphology and sediment ﬂux (dis)connectivities,
(iii) assess the relationships between the modeled hydraulic
variables responsible for the transformation of the gravel
bars' surface layer and the grain-size parameters of these
bars.
2. Study area
2.1. Basic regional settings
The Bečva River is 61.5 km-long gravel-bed stream with a drainage
area of 1620 km2
(data source: Czech Hydrometeorological Institute)
and feather-like or dendritic channel network (Ivan et al., 2000). It is a
left side tributary of the Morava River (Danube River basin). The Bečva
River begins at the junction of two main source streams, the Rožnovská
Bečva River and the Vsetínská Bečva River (see Fig. 1). They drain out
Czech part of Western Carpathians, namely the Moravskoslezské Beskydy
Mts., Hostýnsko-vsetínská hornatina Mts., Javorníky Mts., Vizovická
vrchovina Mts. to the relatively ﬂat piedmont. The locations of the highest
and lowest elevations within the Bečva R. basin are, Čertův Mlýn at
1205 m asl, and the mouth to the Morava R. at 195 m asl, respectively.
Annual precipitation in the basin area ranges from 500 mm at the
lower parts of the basin to 1200 mm in the mountains (data source:
Czech Hydrometeorological Institute). The Bečva River is characterized
by the occurrence of frequent ﬂoods of moderate magnitude due to
snow melting, and large ﬂoods caused by prolonged rainfalls, which
are connected to summer cyclones.
The study was performed in a 14.0-km long river reach of the
Bečva River, which was entirely located in an unconﬁned valley of
the piedmont area (i.e., out of primary sediment sources). The main
tributary in the studied river reach is left-side gravel-bed tributary of
the Juhyně River at 1.1 km downstream the beginning of the studied
longitudinal reach. The mean annual discharge is 15.5 m3
s−1
at the
Teplice nad Bečvou gauging station (0.6 km immediately downstream
the studied river reach where basin area is 1275 km2
, for location
in Fig. 1 see gauging station a). Table 1 shows the mean
annual discharges and recurrence intervals (R.I.) of ﬂood discharges
of additional gauging stations of the Bečva, Vsetínská Bečva and
Rožnovská Bečva Rivers.
The study area belongs to the Outer Carpathian Group of ﬂysch
nappes. The historical occurrence of the multi-thread channel of the
Bečva River was mainly caused by high sediment supply into the channel
network from the ﬂysch lithology of the basin's mountainous areas.
The predisposition of local ﬂysch lithology to high sediment supply in
this area was described by Galia et al. (2016) and Škarpich et al.
(2013). Lithology of the mountainous part of the Bečva River drainage
basin is mainly composed of alternating layers of claystone, sandstone
and conglomerate ﬂysch, from Jurassic to Palaeogene rocks (Chlupáč
et al., 2011; Menčík et al., 1983). The valley ﬂoors are dominated by unconsolidated
Miocene (Badenian) clays and sands, and ﬂuvial
Fig. 1. Geographical position of the studied river reach; legend: 1 — streams, 2 — studied river reach, 3 — the Bečva River basin border, 4 — gauging stations, 5 — mountain peaks; 6 — state
border, 7 — settlements; gauging stations: a — Teplice nad Bečvou, b — Jarcová, c — Vsetín, d — Valašské Meziříčí, e — Rožnov pod Radhoštěm, f — Kelč (data source: U.S. Geological Survey —
SRTM 1 Arc Second Global; T. G. Masaryk Water Research Institute, public research institution).
674 V. Škarpich et al. / Science of the Total Environment 649 (2019) 672–685
Quaternary sediments (Czudek, 2005; Dvořák et al., 2001). The valley
ﬂoor is formed by a wide (up to 3 km) ﬂoodplain fringed by Mid–Late
Pleistocene river terraces (Stacke et al., 2014; Macoun et al., 1965).
The downstream part of the valley was partly affected by Pleistocene
continental glaciation during its maximum advance in the Saalian 1
(Ehlers and Gibbard, 2004).
2.2. History of direct human impact on channel development
At the end of the 19th century, the ﬁrst channelization works in
the study reach were conducted by restricting the ﬂow to a narrow
channel through the stabilization of banks and gravel bars with vegetation.
At the turn of 20th century, even more systematic regulation
was applied, and the Bečva River was completely regulated in the
studied 14.0-km long river reach (Havlík, 1999). The original
multi-thread river channel pattern was converted into a uniform
single-channel. The channel was designed in trapezoidal shape
which was accompanied by levees, and the channel bed was stabilized
by boulder ramps (Havlík, 1999; Krejčí and Krejčí, 2012). The
main reasons for channelization of the Bečva River were ﬂood control
and a demand for arable land. As a consequence of river management,
the channel has typically incised several meters into the
Table 1
Mean annual discharge and ﬂood recurrence intervals of the Bečva, Vsetínská Bečva, Rožnovská Bečva and Juhyně Rivers (data source: Povodí Moravy Basin State Enterprise and Czech
Hydrometeorological Institute).
Gauging station River/Stream Basin area of given gauging station
[km2
]
Start year of
measurement
Mean annual discharge [m3
s−1
]
N-year recurrence interval
discharge (QN) [m3
s−1
]
Q1 Q5 Q10 Q50 Q100
Teplice nad Bečvou Bečva 1275.32 1920 15.50 219 452 555 799 908
Jarcová Vsetínská Bečva 723.87 1939 9.39 151 274 333 479 547
Vsetín Vsetínská Bečva 505.81 1940 6.79 126 234 279 378 420
Valašské Meziříčí Rožnovská Bečva 252.45 1941 3.79 66.5 161 214 364 441
Rožnov pod Radhoštěm Rožnovská Bečva 160.24 1951 2.72 42.8 99.1 134 241 301
Kelč Juhyně 86.12 1957 0.83 9.3 31.5 43.6 79.8 98.8
Fig. 2. View of the Bečva River channelized (managed) single channel sections between (A) 0.0–1.7 km, (C) 3.0–9.0 km and (E) 10.1–14.0 km and retrogradually developed (renaturalized)
multi-thread channel sections between (B) 1.7–3.0 km and (D) 9.0–10.1 km; designations A, B, C, D and E are corresponding with designations of channel sections in Fig. 4.
675V. Škarpich et al. / Science of the Total Environment 649 (2019) 672–685
ﬂoodplain along its entire reach. The sediment supply balance has
also been upset by reforestation of the mountainous part of the
Bečva River basin since the end of 19th century (Krejčí and Krejčí,
2012; Pavelka and Trezner, 2001).
The studied channel is presently characterized by several distinctive
sections. Regulated single channels with artiﬁcial bank stabilizations
are incised several meters into the ﬂoodplain (see Fig. 2A, C and
E). In contrast, the multi-thread channel pattern (see Fig. 2B and D)
has retrogradually developed after the 1997 (Hrádek, 2000) and
2010 (Krejčí and Krejčí, 2012) ﬂoods (maximum discharge
950 m3
s−1
with ~100-year R.I. and 799 m3
s−1
with 50-year R.I., respectively;
data source: Czech Hydrometeorological Institute) in two
short river sections. Before the extreme ﬂood in 1997, the regulated
channel was trapezoidal in cross-section, the width was 33–35 m
and the depth was 2.6–3.7 m (Mihola, 1992). The bank and bed stabilization
structures were systematically repaired by water management
agencies after individual ﬂoods (e.g., in 1950s). This state was
given by the political situation (Communist Era) in the Czech
Republic (1948–1989) which aimed to maximal using of land for agriculture
(Chloupkova, 2002). However after the last signiﬁcant
ﬂoods in 1997 and 2010, these stabilization structures in renaturalized
sections were not repaired again because (i) the arable
land along the river reach was no longer suitable for agriculture,
and (ii) no larger settlements are present along the river which
could be endangered by ﬂoods.
3. Methods
3.1. Geomorphic and (dis)connectivity mapping
The contemporary state of channelized/re-naturalized channel sections,
lateral sediment inputs and potential disconnectivities in sediment
ﬂux were mapped in detail along the entire studied reach. We
used conventional method of geomorphic mapping based on combination
of ﬁeld survey, analysis of aerial photos and inspection of digital elevation
model obtained by LiDAR altimetry airborne laser scanning in
2013. Aerial photos and LiDAR dataset were produced in S-JTSK/
Krovak East North coordinate system by State Administration of Land
Surveying and Cadastre of Czech Republic. Aerial photos were collected
in year 2016 with spatial resolution of 0.2 m. The LiDAR-based digital elevation
model, produced in raster format with spatial resolution 5 × 5 m
and mean altitude accuracy 0.3 m, provided us with information about
the geometry of the ﬂoodplain. We generally distinguished the following
patterns: (i) depositional, relatively wide channel-sections with
the development of gravel bars as in the re-naturalized multi-thread
river pattern, and (ii) transport-balanced channel-sections, which
were typically incised several meters into the ﬂoodplain as the consequence
of river management. Sediment inputs consisted of individual
bank failures, bedrock outcrops and tributaries. Nine boulder ramps stabilizing
the channel bed decreased upstream bed slopes and local ﬂow
competences. Artiﬁcial bank stabilizations (riprap and concrete ﬂex
mats) were representatives of lateral disconnectivities, which also
prevented lateral channel migration.
3.2. Sampling of gravel bars
Investigated gravel bars were sampled during low ﬂow conditions
(1.3–1.6 m3
s−1
) in August 2015. To avoid differences in sample populations
with respect to the position on the gravel bar, all samples were
taken from the middle part (in longitudinal direction of the river
reach) of the bar, close (~1–1.5 m) to the wetted channel. In total, 68
samples were collected in the 14-km long studied reach. The grainsize
characteristics of bar surfaces were obtained by photographic
grain-size analysis. This analysis was performed with the Sedimetrics
Digital Gravelometer software package (Loughborough University Enterprises
Ltd) (Graham et al., 2005a, 2005b). The method enables
rapid image-processing-based procedure for the measurement of exposed
ﬂuvial gravels from digital photographs. The precision achieved
is comparable with conventional pebble count sampling strategy of
Wolman (1954) as noted by Graham et al. (2005a). Rice and Church
(2010) compared values of D50 and D95 obtained from paired pebble
count and photographic samples of gravel bar surfaces. They revealed
that the photographic method has not introduced any signiﬁcant bias,
i.e. the mean-square differences were ±3 and ±9 mm, respectively.
One photo per sampled bar was taken. Photos covered a planar area
of 0.75 m2
which was cleaned of leaves or remains of small branches before
taking the photographs. Reference points (plastic targets) were
placed at each corner of the rectangular wooden sample patch (see
Fig. 3A). After the placing of reference points, rectangular wooden sample
patch was removed (see Fig. 3B). The reference points provided a
scale and they deﬁned the boundary of the patch in the image. The photographs
were taken (see Fig. 3C) using a SAMSUNG ST88 compact camera
(35 mm equivalent focal length = 25 mm, 16.1 Mpx). The grid-by
number method truncated on the lower bound at 8 mm was applied
to obtain D50 and D95 percentiles (mm) and the value of sediment
sorting SI at Φ scale (in the sense of Folk and Ward, 1957) as representative
of the previous episodes which affected bed conditions in the
river channel.
3.3. Hydraulic modeling
HEC-RAS 5.0.3 software (U.S. Army Corps of Engineers - USACE) was
used to calculate hydraulic parameters in the positions of investigated
sediment samples; namely, (i) the water surface slope (m m−1
) — Sws,
(ii) the wetted channel width (m) — Wf, (iii) the ﬂow area (m2
) — Af,
(iv) the bed shear stress (N m−2
) — τb, (v) the unit stream power
(W m−2
) — ω, and (vi) the mean ﬂow depth (m) — d.
The software has been widely used in numerous studies at various
scales in relation to sediment processes (Song et al., 2015), hydrokinetic
assessment (Punys et al., 2015), river restoration (Pregun, 2016) and
ﬂood hydraulics (Czech et al., 2016).
1-year R.I. discharge (219 m3
s−1
in Teplice nad Bečvou gauging station)
was simulated by 1D model running in unsteady mode and hydraulic
parameters were calculated in the positions of investigated
sediment samples. We observed complete submergence and morphological
changes of studied gravel bars on 5/15/2014 (260 m3
s−1
,
~1–2-year R.I.) and 1/11/2015 (136 m3
s−1
, b1-year R.I., data source:
Czech Hydrometeorological Institute). Based on these observations,
we consider 1-year R.I. discharge as the discharge which was sufﬁcient
to rework the surface layer of all studied bars.
Hydraulic modeling consists of several crucial steps. The present digital
elevation model obtained by airborne laser scanning (LiDAR) is not
able to capture terrain under water surfaces, so the bathymetry of the
river channel was added from geodetically measured cross-sections.
These were measured during a low ﬂow stage in February and March
2016 at appropriate positions within the studied river reach, when a
higher density of cross-sections was obtained in irregular channelsections
(i.e., in river bends or in the case of the presence of a multithread
pattern). Every single cross-section was taken with respect to
the bank and bed morphology, usually including 10–20 measurement
points obtained using the total station TOPCON: GTS 212 — 2000.
Mean bed slope between individual cross-sections was taken from the
digital elevation model. In total, 32 cross-sections (i.e., 2.3 crosssections
per km) were collected for the studied river reach of the
Bečva River. An additional ﬁve cross-sections were taken upstream
and downstream of the studied 14-km long reach, which served as section
for reﬁning of simulated hydraulic variables due to absence of explicit
boundary conditions in lower and upper part of the 14-km long
reach. In addition, the parameters of boulder ramps (height, length,
cross-sections) were geodetically measured and added into the model.
To preserve channel sinuosity, parts of the channel between the
surveyed cross-sections were inspissated using the shape of originally
676 V. Škarpich et al. / Science of the Total Environment 649 (2019) 672–685
surveyed cross-sections. The criterion for the selection of original crosssection
shape was the channel width, measured from aerial photos in
ArcGIS 10.2.1 software, while S-JTSK/Krovak East North coordinate system
was applied. Elevation of points in these inspissated cross-sections
was recomputed using the elevation from the ﬁrst point in cross-section
obtained from LiDAR. Channel cross-sections were merged with adjacent
ﬂoodplain (LiDAR) in HEC-RAS software.
The hydrograph of the ﬂood event from 2010 with ~50-year R.I. peak
discharge was used for basic calibration of the model. Discharge data from
the time span between 5/11/2010 (12:00 a.m.) and 06/10/2010 (11:00 p.
m.) at 1 h intervals were provided by the Czech Hydrometeorological Institute.
The model was calibrated by adjusting Manning's n hydraulic
roughness. Simulated discharge was calibrated for the timing of ﬂood
peak (799 m3
s−1
) and for the value of maximum ﬂood peak with the
Manning's n-values and compared with the ﬂood hydrograph observed
at the outlet gauging station Teplice nad Bečvou (see Fig. 1 for location).
Additional calibration of water levels at individual cross-sections for
1-year R.I discharge reﬂected known water level in Teplice nad Bečvou
gauging station. The maximal water level error was 0.07 m, simulated
time of ﬂood peak occurred 0.25 h later and maximal simulated ﬂood
peak was equal to the observed one. Finally, the Manning's n-value
0.035 in the channel was established in accordance with this calibration
process. To preserve gradual increase in discharges from upper to lower
part of studied reach, boundary conditions of modeled river reach were
based on the linear relation between the increasing basin area and discharges
from upstream and downstream gauging stations in the Bečva
River basin. In the lower part of the river reach, the normal depth was
applied as the water level was unknown. Because the normal depth is
the estimation of the measured water depth, the lower part was
prolonged to reﬁne simulated hydraulic variables. Furthermore, Teplice
nad Bečvou gauging station (in Fig. 1, see gauging station a) was a part
of this extended reach. The same extension took place in the upper part,
because the discharge data from two gauging stations (in Fig. 1, see
gauging station b — Jarcová and d — Valašské Meziříčí) were available
here. It implies that the extended river reach for hydraulic modeling
has in total 21 km length. There occurred fourteen tributaries the
contributions of which to discharge were modeled as lateral inﬂow
boundary conditions. The 1-year R.I. discharge was derived from the
contributing area of individual tributaries. The discharge was computed
from direct relationship (Q1 = 1.0457A0.7486
, R2
= 0.98, p b 0.001) between
the 1-year R.I. discharge (Q1) and the basin area (A) of gauging
stations in the Bečva River basin. The similar approach of discharge
derivation was applied in the studies of Buttle et al. (2016) and Galia
and Škarpich (2016). The contributing area was computed by ArcGIS
on the basis of LiDAR digital elevation model and ArcHydro 10.2
extension.
3.4. Data analysis
Calculated values of the transport stage (TS) were used for the assessment
of ﬂow competence. TS is described as a ratio between the
available bed shear stress (τb) and the reference (critical) shear stress
(τr) producing a very small, but measurable bedload transport rate
(Parker, 1990):
TS ¼ τb=τr: ð4Þ
It implies that the threshold for measurable bedload transport rates
assumes the value of TS is N1.
The bed shear stress values were calculated by HEC-RAS model with
Eq. (3) for shallow and wide channels. The reference shear stress for the
Fig. 3. Illustration of the photographic procedure: (A) placing of reference points (plastic targets) at each corner of the rectangular wooden sample patch, (B) removed wooden sample
patch after the placing of reference points and taking approximately vertically photograph by digital camera, (C) scale and deﬁned boundary of the patch in the image provided by
reference points in photograph, the photographed area (shaded) in (B) must include all grains intersecting the patch edge (dashed outline).
677V. Škarpich et al. / Science of the Total Environment 649 (2019) 672–685
representative bed grain-size takes the form (Shields, 1936):
τr ¼ ρs−ρð ÞθgDi; ð5Þ
where ρs is the density of grain, θ is the dimensionless shear stress and
Di is the representative grain diameter (e.g., D50). The assessment of the
θ value is the subject of ongoing discussions in bedload calculations for
gravel-bed rivers (e.g., Bufﬁngton and Montgomery, 1997; Lamb et al.,
2008). For our purposes, we used the slope-dependent equation of
Parker et al. (2011):
θ ¼ 0:19Sws
0:28
; ð6Þ
where Sws is the water surface slope (m m−1
). The Parker et al. (2011)
equation was derived from an extensive dataset of ﬁeld measurements
in gravel-bed rivers covering a wide range of channel slopes. The parameter
of Sws was calculated by HEC-RAS for individual cross-sections
adjacent to investigated gravel bars.
3.5. Statistical methods
Hydraulic (Sws, Wf, Af, τb, ω, d, TS) and grain-size (D50, D95, SI) parameters
were statistically analysed. Because the majority of hydraulic and
grain-size parameters were not normally distributed and they showed
differences in variance between the channelized (managed) and renaturalized
channel sections, the nonparametric Kruskal-Wallis (KW)
test was used. This test was used to examine the differences in the hydraulic
and grain-size variables among three channelized (managed)
and two re-naturalized channel sections. If KW tests indicated a signiﬁcant
difference (p b 0.05) in a particular variable, multiple pairwise
comparison was performed using Dunn's procedure with the Bonferroni
corrected signiﬁcance level to identify mutually different channel
sections.
A principle components analysis (PCA) with varimax rotation was
used to ﬁnd potential relationships between the grain-size of gravel
bars (SI and D50, D95) and hydraulic parameters (Wf, Af, ω, τb, d, Sws,
TS) of examined cross-sections. The PCA was chosen because of its
capability to reduce the number of entering variables (nine in our
case). Resulting components are represented by a linear combination
of variables to summarize the patterns of intercorrelation between the
variables, yet the components are mutually independent.
Varimax rotation is a change of coordinates in PCA by rotating the
axes of two-dimensional plane while keeping the 90-degree angle
between them. This maximizes the sum of the variances of the squared
loadings. Thus, all the coefﬁcients (squared correlation with factors) will
be either large or near zero, with a few intermediate values, which helps
to easier interpretation of calculated results (Kaiser, 1958).
4. Results
4.1. Contemporary patterns in the channel morphology
General patterns of channel morphology varied considerably along
the entire studied reach. The river sections between 0.0 and 1.7 km
(designated A in Fig. 4), 3.0–9.0 km (designated C in Fig. 4) and
10.1–14.0 km (designated E in Fig. 4) exhibited channelized (managed)
sections. They were characterized by a narrow channel with a trapezoidal
cross-section and artiﬁcial bank stabilizations (riprap and locally
concrete ﬂex mats). These structures represented lateral
disconnectivity, which prevented lateral channel migration and sediment
supply from adjacent river terraces. In total, nine artiﬁcial boulder
ramps were observed in the river section A, C and E, which stabilized the
channel bed. In addition, they acted as potential longitudinal forms of
sediment disconnectivity due to the local decrease in bed slope upstream
from these structures. At 10.5 km, an additional boulder ramp
was partially destroyed to the present height of ~0.3 m.
Overall, the river sections A, C and E were characterized as transportbalanced
zones with the presence of small gravel bars. Between 11.7
and 14.0 km (the part of E section), several relatively larger gravel
bars were observed despite the stream channelization.
In contrast, the river sections between 1.7 and 3.0 km (designated B
in Fig. 4) and between 9.0 and 10.1 km (designated D in Fig. 4) exhibited
the multi-thread pattern with the occurrence of larger gravel bars and
the absence of artiﬁcial bank stabilizations. Original artiﬁcial bank stabilization
structures were destroyed by large ﬂoods in 1997 and 2010. The
removal of bank stabilizations enabled lateral channel migration accompanied
by the development of large bank failures in outer river bends
and increased sediment supply to the channel. These sections B and D
were characterized as accumulation zones with active deposition of
Fig. 4. (A) Contemporary state of channelization/re-naturalization, (B) patterns in the channel, (C) lateral sediment inputs and potential disconnectivities in sediment ﬂux.
678 V. Škarpich et al. / Science of the Total Environment 649 (2019) 672–685
sediment material. The occurrence of large woody debris was observed
locally in the re-naturalized river section D. Partially, the effect of vegetation
succession stabilizing gravel bars was observed in the renaturalized
channel sections.
Along the entire studied reach, the channel bed was formed by
gravel-size material. No visible outcrops of bedrock were observed in
the channel except in cross-sections at ~2.5 km and at ~3.0 km, where
bedrock outcrops of unconsolidated Miocene (Badenian) clays and
sands occurred. Except the left-side Juhyně River tributary, other eight
short tributaries were observed in the entire studied reach of the
Bečva River, which operated as sediment inputs of various grain-size
and magnitudes. Based on the ﬁeld observations, left-side Juhyně
River (see tributary number 1 in Figs. 4 and 5) and two tributaries
(see tributaries number 7 and 9 in Figs. 4 and 5) were gravel-carrying
with developed gravel bed surface layer. Other tributaries (see
tributaries number 2–6 and 8 in Figs. 4 and 5) were distinguished as
sand-carrying because developed sand bed surface layer was observed.
4.2. Downstream variations in grain-size parameters
Along the entire studied reach, the D50 demonstrated a nonsigniﬁcant
exponential tendency in downstream ﬁning (see Fig. 5A;
D50 = 37.6766e−0.0085L
, R2
= 0.03, p = 0.17). It implies that the exponential
model produced a very low reduction (ﬁning) coefﬁcient of
0.0085 mm km−1
. The exponential model produced a signiﬁcant trend
in D95 of downstream ﬁning (see Fig. 5B; D95 = 91.9719e−0.0153L
, R2
= 0.07, p = 0.03) and the exponential model gave a reduction (ﬁning)
coefﬁcient equal to 0.0153 mm km−1
. The disruption of downstream
ﬁning in D95 was observed at the 1.0–2.0 km (in section A), at
5.0–6.5 km (in section C) and also at 10.0–10.5 km (in section E).
Fig. 5. The downstream variations in (A) D50, (B) D95 and (C) SI along the Bečva River entire studied reach; solid lines show exponential trends/tendencies and dashed lines show running
average of the grain-size parameters of D50, D95 and SI.
679V. Škarpich et al. / Science of the Total Environment 649 (2019) 672–685
The non-signiﬁcant linear tendency in SI were observed for the entire
studied river reach (see Fig. 5C; SI = 0.9307e−0.0052L
, R2
= 0.05, p
= 0.07). According to common classiﬁcations (in the sense of Folk and
Ward, 1957) of SI, gravel bars (where SI ranges between 0.7 and 1.1Φ)
were moderately or poorly sorted.
4.3. Differences in hydraulic and grain-size variables between managed and
re-naturalized channel-sections
Differences in modeled ﬂow hydraulics during 1-year R.I. discharges
and obtained grain-size parameters of bars between managed sections
(A, C and E) and re-naturalized sections (B and D) were tested (Fig. 6).
The hydraulic parameter of Wf (p b 0.001) reﬂected that the renaturalized
sections B and D had larger wetted widths than channelized
sections C and E. In addition, difference in Wf was observed between the
channelized sections C and E where section E had wider wetted channel
during 1-year R.I. discharge. Additionally, the hydraulic parameter Sws
showed the only signiﬁcant difference between the channelized section
A and the re-naturalized section D where section A had higher values of
Sws than section D (p = 0.016). The hydraulic parameter ω (p b 0.001)
showed the differences, when re-naturalized channel sections B and D
had lower values of the unit stream power than the channelized sections
A and C. Moreover, signiﬁcant differences were reported between
the re-naturalized section D and channelized section E, when section D
had lower ω than section E. Signiﬁcant differences between renaturalized
section D and channelized sections A, C and E occurred in respect
to TS parameter (p = 0.006). No signiﬁcant differences in grainsize
parameters D50 (p = 0.730) and SI (p = 0.355) were found
between the re-naturalized and the channelized sections (see Fig. 6H
and I).
4.4. Potential relationships between the grain-size and hydraulic variables
Principal components PC1 and PC2 accounted only for 40.4% and
22.8% of the total variance, respectively. The added third component
Fig. 6. Boxplots of hydraulic and grain-size parameters of managed channel-sections (A, C and E) and re-naturalized channel-sections (B and D); letters above boxes show signiﬁcantly
different channel-sections obtained by Kruskal-Wallis test and followed by multiple pairwise comparisons (Dunn's procedure with Bonferroni corrected signiﬁcance level); the
centreline shows the median and the edges of the box represent the ﬁrst and third quartile.
Fig. 7. Bar chart of factor loadings of individual PCs rotation and the communalities of
principal components after varimax rotation; signiﬁcant correlations at α = 0.05 are
marked by an asterisk (*); D50 and D95 — 50th and 95th percentile of grain-size
parameters, SI — sorting index, Wf — wetted channel width, Af — ﬂow area, ω — unit
stream power, τb — bed shear stress, d — mean depth, Sws — water surface slope, TS —
transport stage.
680 V. Škarpich et al. / Science of the Total Environment 649 (2019) 672–685
PC3 (19.5%) increased the total communality to 82.7% (Fig. 7). PC1 included
hydraulic parameters Af, ω, τb, Sws, and TS. The grain-size parameters
of D50, D95 and SI were included in PC2. Principal component PC3
clustered hydraulic parameters Wf, Af and d. Wf and Af have an inverse
relationship with d (Fig. 7). It implies that PCA analysis did not demonstrate
any clear relationship between the grain-size parameters (D50,
D95 and SI grouped in PC2) and the hydraulic parameters modeled for
1-year R.I. discharge.
We observed some distinctions among the hydraulic variables after
plotting the evaluated channel sections, where the x-axis and y-axis
are represented by PC1 and PC2 (Fig. 8), respectively. Based on this
analysis, the hydraulic parameters (PC1) were more important to distinguish
between managed and re-naturalized sections than the grainsize
parameters (PC2) while the high downstream variability of the
grain-size parameters was observed in PC2 in relative independence
on channelized/re-naturalized state of the river.
5. Discussion
The dominant downstream ﬁning trend (in the sense of Sternberg,
1875) of many alluvial rivers is punctuated by positive grain-size steps
at coarse-sediment recruitment points such as tributaries, bedrock outcrops
or bank failures (Rengers and Wohl, 2007; Rice, 1999; Rice and
Church, 1996; Surian, 2002), barriers in sediment ﬂux (Knighton,
1999; Surian, 2002) or changes of channel pattern (Constantine et al.,
2003; Zawiejska et al., 2015). Detection of these coarse-sediment recruitment
sources and subsequent identiﬁcation of barriers can help
mapping of connectivity in ﬂuvial system, which is necessary for sustainable
management of dynamic ﬂuvial systems (e.g., gravel-bed rivers)
in cultural landscape (Fryirs, 2013). In the entire studied reach of
the Bečva River, the analysis of gradual downstream decrease in bed
grain-size produced very low reduction coefﬁcients in D50 (non-signiﬁcant
exponential tendency) and D95 (signiﬁcant exponential trend),
where the effect of relatively short evaluated longitudinal distance
(only 14 km) should also be considered. The better goodness-of-ﬁt between
D95 and the stream length than for D50 was also previously observed
by Zawiejska et al. (2015) in a Carpathian gravel-bed river of
similar dimensions to our case. The rates of grain-size reduction were
relatively similar compared with those reported by other studies of
gravel-bed rivers from the Polish and Ukraine Carpathians (see
Table 2) reported by Malarz (2004) or Zawiejska et al. (2015). On the
other hand, higher rates of reduction were observed in headwater
reaches (Galia et al., 2015) and braided reaches (Dawson, 1988; Cowie
and Brierley, 2008; Surian, 2002). In these channel patterns, relatively
higher bed slope accelerates sediment ﬂux connectivity and produces
more intensive abrasion and sorting, which is reﬂected in rapid downstream
ﬁning of bed sediments (Dawson, 1988).
As expected, grain-size parameters in the studied reach did not
show a consistent downstream decrease but a much more complex pattern
occurred (see in Fig. 5 dashed lines showing running average of the
grain-size parameters and in Fig. 6H variation in grain-size parameters
for sections A–E). Similar complex patterns were previously described
in other gravel-bed rivers (Dawson, 1988; Surian, 2002). A high
Fig. 8. Plotting of the ﬁrst two principal components and loadings of individual PCs;
designation A, B, C, D and E is corresponding with designation of channel sections in Fig. 4.
Table 2
Reduction coefﬁcients and basic morphological parameters of (active) channela
width and bed slope from ﬁeld studies in braided, multi-thread and single streams.
River Reduction
coefﬁcient
[mm km−1
]
Base analysis Length of
study
reach [km]
(Active)
channel
width [m]
Bed slope
[m·m−1
]
Channel pattern Reference
Piave River (Italy) 0.027 Mean grain
diameter
50 225 0.008 Braided/multiple-thread Surian (2002)
Ngaruroro River (New
Zealand)
0.052 Mean grain
diameter
19 50–750 0.005 Braided Cowie and Brierley
(2008)
Sunwapta River (Alberta,
Canada)
0.104 Mean grain
diameter
11.5 300–500 0.007 Braided Dawson (1988)
Rio Chagres (Panama) 0.013–0.017 Mean grain
diameter
40 20–100a
0.009 Multiple-thread/single Rengers and Wohl
(2007)
Rhine River (Switzerland) 0.003–0.016 Largest pebbles 42 85–115 0.003 Multiple-thread Mikos (1994)
Cosumnes River (California,
USA)
0.072 Mean grain
diameter
43 10–80 0.001 Multiple-thread Constantine et al.
(2003)
Ringarooma (Tasmania) 0.009–0.015 Mean grain
diameter
60 20–200a
0.009 Multiple-thread/single Knighton (1999)
Czarny Dunajec (Poland) 0.012 Mean grain
diameter
18 20–120 0.007 Multiple-thread/single Zawiejska et al. (2015)
Sola River (Poland) 0.032 Largest pebbles 25 50–300a
0.002 Multiple-thread/single Malarz (2004)
Skawa River (Poland) 0.036 Largest pebbles 21 30–150a
0.002 Multiple-thread/single Malarz (2004)
Prut River (Ukraine) 0.015 Largest pebbles 85 150–300a
0.002 Multiple-thread Malarz (2004)
Cheremosh River (Ukraine) 0.010 Largest pebbles 33 200–300a
0.003 Multiple-thread Malarz (2004)
Bečva River (Czech Republic) 0.009–0.015 Mean grain
diameter
14 34–291 0.006 Multiple-thread/single This study
Carpathian headwaters 0.283–0.465 Largest particles 0.63–1.08 1.2–3.8 0.050–0.280 Single Galia et al. (2015)
a
(Active) channel width measured using Google Maps (Image © 2018 CNES/Airbus).
681V. Škarpich et al. / Science of the Total Environment 649 (2019) 672–685
occurrence of lateral sediment sources (as tributaries, bank failures)
along with sediment disconnectivities in the river reach (as boulder
ramps) disrupted downstream ﬁning (see Figs. 4 and 5).
Both river sections A (managed) and B (re-naturalized) produced a
high variability in grain-sizes (see Fig. 5 and Fig. 6H). The role of
sediment supply from the major left-side Juhyně River (see tributary
number 1 in Fig. 5) (in sections A and B), and partially from bank failures
(in section B) was reﬂected in the ﬂuctuations of measured grain-size
parameters. Decreased values of the bed slope upstream of three
additional highest boulder ramps in section A led to the reduction of
local stream power and bed shear stress upstream of these ramps,
which directly affected the intensity of bedload transport in these
locations. The relationship between the presence of grade-control
structures and decrease of bedload transport rates in a Carpathian
stream via reduced bed slope was documented by Galia et al. (2016),
who reported gradual downstream coarsening of bed sediments. Such
downstream coarsening through the sequence of check-dams was also
previously documented in Mediterranean rivers (Boix-Fayos et al.,
2007).
The entire section C was affected by channelization and higher
values of ω, τb and TS were observed and compared with the renaturalized
section D (Fig. 7). Section C was characterized as a
transport-balanced section, where deposition processes are limited by
higher rate of ﬂow competence. At the uppermost part of this section,
the increased rate of ﬂow competence of the regulated channel and
only one boulder ramp partially reduced the bed slope, potentially decreasing
sediment transport rates. Absence of boulder ramps reduced
the bed slope in the middle part of this section C and affected disruption
of downstream ﬁning where the increased rate of ﬂow competence of
the regulated channel was observed. In the downstream part of the
channelized section C, initial bed slope was reduced by four boulder
ramps (6.7, 7.1, 7.3 and 8.3 km) which decrease the potential in sediment
ﬂux by decreased rate of ﬂow competence. The river section D
with large gravel bars developed by re-naturalization processes by
1997 and 2010 ﬂoods was predisposed to deposition, the general ﬁning
of sediment material and lower values of ω, τb or TS (Fig. 7) compared
with sections A, C and E. The large woody debris observed in the channel
of section D most likely disrupted sediment ﬂux coming from the downstream
part of section C. The abrupt local decline in bed grain-sizes may
be attributed to the presence of large woody debris, which is in accordance
with the works of Beschta (1979) who studied the effect of
large woody debris removal on bed grain size, Bufﬁngton and
Montgomery (1999) who modeled the inﬂuence of bank irregularities,
large woody debris and gravel bars on sediment ﬁning and Brooks
et al. (2004) who monitored bed grain-size changes after the addition
of engineered log jams.
Section E represented a channelized single channel pattern. The leftside
tributary (see tributary number 7 in Fig. 5) and destroyed boulder
ramp (at 10.5 km), where the increased rate of ﬂow competence of
the regulated channel was assumed, produced occurrence of coarser
material in the part downstream from the 10.5 km (see Fig. 5A and B).
The difference in Wf was observed between the channelized sections C
and E, where section E had larger Wf during 1-year R.I. discharge. Reported
hydraulic conditions in section E were more prone to deposition
than the transport-balanced conditions in sections C or A. At the lowermost
part of section E, the deposition was observed in the form of large
gravel bars. The high variation in D50 (see Fig. 5A) at the lowermost part
of section E was linked to larger tributaries supplying sediments to the
main channel.
Mapping of channel behaviour (transport-balanced and depositional
sections) in the studied Bečva River reach corresponded well
with the modeled hydraulic variables for 1-year R.I. discharge (see
Fig. 6A–G or Fig. 8). In particular, the re-naturalized river sections
showed lower values of ﬂow competence (ω, τb) and facilitated deposition
of sediment material in the form of larger gravel bars compared
with the channelized river sections characterized by limited presence
of small gravel bars. A similar situation was documented by Czech et al.
(2016) on the gravel-bed Biała River (Polish Carpathians) who described
that within unmanaged channel reaches the ﬂows can be conveyed
with relatively low shear forces in contrast with channelized
reaches, where the ﬂow velocity and shear forces are substantially
higher.
The wide multi-thread channels with low values of ω and τb and
high channel-form roughness facilitated sediment deposition and
reﬂected relatively ﬁne grades of sediment (Komar and Carling, 1991;
Constantine et al., 2003; Zawiejska et al., 2015; Babej et al., 2016). In
contrast, channelized sections (sometimes referred to as conveyor
belts) where higher rates of ﬂow competence and low channel-form
roughness, facilitated downstream transfer of the bed material
(Komar and Carling, 1991; Fryirs et al., 2007; Zawiejska et al., 2015).
The assessment of potential correlations between the grain-size and hydraulic
variables modeled for 1-year R.I. discharge in the Bečva River did
not demonstrate any clear relationships. The PCA analysis clustered the
hydraulic variables in the ﬁrst and third component and the grain-size
parameters in the second component. Rice and Church (2010) documented
that single bed material samples cannot be representative of
the grain-sizes apparent on a bar or across a channel width because
high local variability of sediment sizes. But in the case of the Bečva
River and in another Carpathian gravel bed river Czarny Dunajec of similar
dimensions (Zawiejska et al., 2015), this local variability is low. Despite
the occurrence of re-naturalized channel sections in the Bečva and
the Czarny Dunajec, their channels are relatively stable, much narrower,
and the elevations between the channel bed and top of the gravel bars
are lower (in the Bečva River channel max. ~2 m) by contrast to the
braided, very dynamic Fraser River where Rice and Church (2010) performed
their research. Additional question arises which parameters of
grain-size distribution (ranging from the median D50 to the 95th percentile
D95) are suitable for the assessment of ﬂow competence. The
use of largest particles on the upper tail of the grain-size curve has potential
uncertainties. It may be difﬁcult to recognize the largest particles
within the ﬂood deposits, and there are questions concerning how representative
one or a few large particles might be of the transported sediments
and therefore of the ﬂood hydraulics. However, the analysis of
Komar and Carling (1991) showed that the trend of increasing sizes of
the largest particles in the bed sediment with increasing ﬂow competence
is consistent within interval D50–D95.
Limits concerning the hydraulic modeling are mainly connected
with the model selection, accuracy of digital elevation model together
with resolution of bathymetry obtained from geodetically surveyed
cross-section, as well as with the boundary conditions and model calibration
(Ferguson and Church, 2009). Also typically, 1D hydrodynamic
models are used as 1D representation of the main channel and
2D representation of the ﬂoodplain (Vojinovic and Tutulic, 2009).
Crowder and Diplas (2000) or Benjankar et al. (2015) stated that 1D
model may be sufﬁcient for a study requiring only a description of
the general ﬂow patterns (e.g. ﬂow depth, water surface elevation)
at a macro-spatial scale similar to our case. Furthermore, based on another
factors such as available input data, model set-up and computational
time results obtained by 1D model HEC-RAS are reliable and
sufﬁciently accurate (Horritt and Bates, 2002; Ferguson and Church,
2009). Jowett and Duncan (2012) and Dimitriadis et al. (2016) applied
1D and 2D model to braided and single channel rivers, respectively.
They compared measured and predicted hydraulic parameters
(e.g., water depths and velocities) and suggest that the differences in
predicted hydraulic variables between 1D and 2D modeling approach
are within units of percent.
Following the results of Casas et al. (2006), combination of
geodetically surveyed cross-sections and LiDAR are cost-effective tools
for developing of a digital elevation model and possible errors in simulated
hydraulic variables are sufﬁciently low. There remain uncertainties
in the calibrated uniform Manning's values in single crosssection.
More detailed calibration data from gauging stations,
682 V. Škarpich et al. / Science of the Total Environment 649 (2019) 672–685
hydrometric measurements, aerial photos of ﬂooded area or wreck
marks elevations measurement would reduce uncertainty inherent in
the calibration procedure and supplied uniform Manning's value, but
at the expanse of high economical and time requirements. The similar
approach of uniform values was applied in the study of Ferguson and
Church (2009) who simulate gravel transport and aggradation along a
highly irregular 38-km reach of lower Fraser River.
The one of the main factors affecting potential relationships between
the hydraulic and grain-size variables was the short length of studied
reach where high occurrence of lateral sediment sources with sediment
disconnectivities in the longitudinal river reach was observed and
grain-size parameters did not reﬂect hydraulic conditions. The frequent
occurrence of lateral sediment inputs with sediment disconnectivities
created no differences in D50 and SI (p = 0.355) between the renaturalized
and the channelized sections (Fig. 6H and I) and high
downstream variability of the grain-size parameters observed in PC2
(see Fig. 8). It implies that sediment inputs and longitudinal
disconnectivities provided better explanation of the grain-size patterns
along the studied Bečva River reach than hydraulic conditions calculated
for 1 to 2-year R.I. discharge. The downstream variations in bed
material through identiﬁcation of disconnectivities and sediment inputs
were comprehensively described by Surian (2002). He explained that a
regular downstream ﬁning pattern was observable only in those
reaches of 120-km examined length of the Piave River, where infrequent
lateral sediment inputs and the barriers had minor effects on
the resulting bed grain sizes. Similarly, Rice and Church (1998) documented
that identiﬁcation of sediment inputs were fundamental for understanding
bed sediment grain-size variations within the studied
length of the gravel-bed Pine and Sukunka Rivers in British Columbia.
In the Bečva River studied reach, some tributaries caused local coarsening
of the bar grain-sizes which interrupted downstream ﬁning trend
(see tributary numbers 1 and 7 in Fig. 5). Benda et al. (2004) studied
the role of tributaries on downstream ﬁning in main-stem river channels.
They found that a higher probability of signiﬁcant effect on changing
grain-size in main-stem channel had tributaries with dendritic
channel network compared to rectangular or trellis channel networks
with lower signiﬁcant effect. These ﬁndings support conclusion that
the tributaries of the Bečva River affected signiﬁcantly change in the
grain-sizes because feather-like or dendritic channel network of these
tributaries were documented here by Ivan et al. (2000). Benda et al.
(2004) and Rice (1998) also discussed the importance of tributaries in
resetting geomorphic and sedimentary characteristics and the effect of
tributaries on morphological heterogeneity. The local coarsening was
also observed in the regulated channel reaches where the increased
rate of ﬂow competence was assumed and the grade control structures
(boulder ramps or weirs) were missing. Additional factors as sediment
inputs from bank failures as well as sediment ﬂux disconnectivities
caused by boulder ramps, weirs and partially large woody debris or vegetation
succession produced less signiﬁcant changes in the bar grain size
in the Bečva River reach. In future, more detailed research will be necessary
to explain and identify effect of vegetation and woody debris on bar
material stabilization, e.g. by the analysis of time-lapse aerial photos or
detailed repeated mapping.
Reversal re-naturalization of gravel-bed channelized reaches to
multi-thread pattern helps to progressive improvement of the condition
for aquatic and riparian communities (Wyżga et al., 2018). In the multithread
channel, lower transport capacity affects deposition and ﬁning of
bed sediments in contrast to channelized reaches with higher transport
capacity, coarsening of bed sediment and incision or lateral erosion processes
(Zawiejska et al., 2015). Multi-thread gravel bed channels are dependent
on supply of larger volumes of coarse sediments which are
deposited in channel in the form of gravel bars (Gurnell et al., 2009).
Therefore, identiﬁcation and mapping of sediment inputs and
disconnectivities are necessary for future possibilities of restoration,
natural spontaneous or reversal re-naturalization of these types of
river channels.
6. Conclusions
In the past few centuries, European multi-thread gravel-bed rivers
have undergone rapid changes induced by human activities. These alterations
resulted in changes to channel morphologies and ﬂow hydraulics.
Similar trends are apparent in the Bečva River, where an original
multi-thread river channel pattern was converted into a uniform
single-channel. As a consequence of river management of multithread
gravel-bed rivers, including our studied case, channels were typically
incised several meters into the ﬂoodplain (Škarpich et al., 2013;
Zawiejska and Wyżga, 2010; Wyżga, 2001). However, within the studied
reach of the Bečva River, the multi-thread channel pattern has
retrogradually developed after two large ﬂoods (1997 and 2010) in
two sections.
The mapping of sediment (dis)connectivity brought important ﬁndings
about the character of sediment ﬂux in the studied reach, which
were confronted with simulated cross-sectional hydraulics. We demonstrated
that in the case of a high occurrence of lateral sediment inputs
(tributaries, bank failures) and longitudinal sediment ﬂux
disconnectivities (weirs or boulder ramps), the assessment of the longitudinal
distance, bar grain size and simulated hydraulics submerging
bars did not produce any clear relationships. Although the sections
with re-naturalized multi-thread patterns showed distinctive hydraulic
variables (i.e., larger wetted width or lower unit stream power), we did
not observe direct relationships with their bar sediment sizes. This implies
that for complex ﬂuvial systems of multi-thread rivers as the transition
reaches connecting mountainous and lowland areas, even those
in unconﬁned valley settings out of the primary sediment sources, additional
factors (i.e., effect of bank failures and especially tributaries as
sediment inputs, weirs or boulder ramps as sediment ﬂux
disconnectivities) beyond local ﬂow hydraulics and distance from the
main sediment sources contribute to better explanation of the downstream
evolution of grain-size patterns. Especially tributaries reset the
sedimentary characteristics of the main-stem Bečva River. It implies
that detailed mapping of sediment (dis)connectivity at basin scale (or
at least at reach scale) including particular elements of sediment ﬂux
is crucial for sustainable management of gravel-bed rivers. The explanations
of the bar-scale variability in relatively narrow gravel bed rivers
(e.g. managed Flysch Carpathian Rivers) will be necessary for the identiﬁcation
of simple predictive models in the context of channel morphology
and hydraulic parameters.
Acknowledgements
The authors sincerely appreciate the comments and suggestions of
the four anonymous reviewers leading to a signiﬁcant improvement of
the manuscript. The study was funded by the Student grant competition
of the University of Ostrava [SGS05/PřF/2017-2018] and Speciﬁc Research
project MUNI/A/1251/2017 of the Masaryk University.
References
Andrews, E.D., 1983. Entrainment of gravel from naturally sorted riverbed material. GSA
Bull. 94 (10), 1225–1231. https://doi.org/10.1130/0016-7606(1983)94b1225:
EOGFNSN2.0.CO;2.
Babej, J., Máčka, Z., Onderka, P., Peterová, P., 2016. Surface grain size variation within
gravel bars: a case study of the River Opava, Czech Republic. Geogr. Fis. Din. Quat.
39 (1), 3–12. https://doi.org/10.4461/GFDQ.2016.39.1.
Bagnold, R.A., 1966. An approach to the sediment transport problem from general physics.
United States Department of the Interior, U.S. Geological Survey, Professional
Paper 422-I.
Benda, L., Andras, K., Miller, D., Bigelow, P., 2004. Conﬂuence effects in rivers: interactions
of basin scale, network geometry, and disturbance regimes. Water Resour. Res. 40,
W05402. https://doi.org/10.1029/2003WR002583.
Benjankar, R., Tonina, D., McKean, J., 2015. One-dimensional and two-dimensional hydrodynamic
modeling derived ﬂow properties: impacts on aquatic habitat quality predictions.
Earth Surf. Process. Landf. 40 (3), 340–356. https://doi.org/10.1002/
esp.3637.
Beschta, R.L., 1979. Debris removal and its effects on sedimentation in an Oregon Coast
Range stream. Northwest Sci. 53 (1), 71–77.
683V. Škarpich et al. / Science of the Total Environment 649 (2019) 672–685
Boix-Fayos, C., Barberá, G.G., López-Bermúdez, F., Castillo, V.M., 2007. Effects of check
dams, reforestation and land-use changes on river channel morphology: case study
of the Rogativa catchment (Murcia, Spain). Geomorphology 91, 103–123. https://
doi.org/10.1016/j.geomorph.2007.02.003.
Brierley, G.J., Fryirs, K., Jain, V., 2006. Landscape connectivity: the geographic basis of geomorphic
applications. Area 38 (2), 165–174. https://doi.org/10.1111/j.1475-
4762.2006.00671.x.
Brooks, A.P., Gehrke, P.C., Jansen, J.D., Abbe, T.B., 2004. Experimental reintroduction of
woody debris on the Williams River, NSW: geomorphic and ecological responses.
River Res. Appl. 20 (5), 513–536. https://doi.org/10.1002/rra.764.
Bufﬁngton, J.M., Montgomery, D.R., 1997. A systematic analysis of eight decades of incipient
motion studies, with special reference to gravel-bedded rivers. Water Resour.
Res. 33 (8), 1993–2029. https://doi.org/10.1029/96WR03190.
Bufﬁngton, J.M., Montgomery, D.R., 1999. Effects of hydraulic roughness on surface textures
of gravel-bed rivers. Water Resour. Res. 35 (11), 3507–3521. https://doi.org/
10.1029/1999WR900138.
Buttle, J.M., Allen, D.M., Caissie, D., Davison, B., Hayashi, M., Peters, D.L., Pomeroy, J.W.,
Simonovic, S., St-Hilaire, A., Whitﬁeld, P.H., 2016. Flood processes in Canada: regional
and special aspects. Can. Water Resour. J. 41 (1–2), 7–30. https://doi.org/10.1080/
07011784.2015.1131629.
Casas, A., Benito, G., Thorndycraft, V.R., Rico, M., 2006. The topographic data source of
digital terrain models as a key element in the accuracy of hydraulic ﬂood modelling.
Earth Surf. Process. Landf. 31 (4), 444–456. https://doi.org/10.1002/esp.1278.
Chloupkova, J., 2002. Czech agricultural sector: organisational structure and its transformation.
Unit of Economics Working Papers 2002/1. Institute of Food and Resource
Economics, University of Copenhagen, Copenhagen, Denmark.
Chlupáč, I., Brzobohatý, R., Kovanda, J., Stráník, Z., 2011. Geologická minulost České
republiky, revisited. second ed. Academia, Praha (in Czech with English summary).
Constantine, C.R., Mount, J.F., Florsheim, J.L., 2003. The effects of longitudinal differences
in gravel mobility on the downstream ﬁning pattern in the Cosumnes River, California.
J. Geol. 111, 233–241. https://doi.org/10.1086/345844.
Cowie, M., Brierley, G., 2008. The inﬂuence of lateral conﬁnement upon the downstream
gradation in grain size of the lower Ngaruroro River, New Zealand. Open Geol. J. 2,
46–63. https://doi.org/10.2174/1874262900802010046.
Crowder, D.W., Diplas, P., 2000. Using two-dimensional hydrodynamic models at scales of
ecological importance. J. Hydrol. 230, 172–191. https://doi.org/10.1016/S0022-1694
(00)00177-3.
Czech, W., Radecki-Pawlik, A., Wyżga, B., Hajdukiewicz, H., 2016. Modelling the ﬂooding capacity
of a Polish Carpathian river: a comparison of constrained and free channel conditions.
Geomorphology 272, 32–42. https://doi.org/10.1016/j.geomorph.2015.09.025.
Czudek, T., 2005. Quaternary Development of Landscape Relief in the Czech Republic. The
Moravian Museum, Brno (in Czech, with English summary).
Dawson, M., 1988. Sediment size variation in a braided reach of the Sunwapta River, Alberta,
Canada. Earth Surf. Process. Landf. 13, 599–618. https://doi.org/10.1002/
esp.3290130705.
Dimitriadis, P., Tegos, A., Oikonomou, A., Pagana, V., Koukouvinos, A., Mamassis, N.,
Koutsoyiannis, D., Efstratiadis, A., 2016. Comparative evaluation of 1D and quasi-2D
hydraulic models based on benchmark and real-world applications for uncertainty
assessment in ﬂood mapping. J. Hydrol. 534, 478–492. https://doi.org/10.1016/j.
jhydrol.2016.01.020.
Dvořák, J., Stráník, Z., Tyráček, J., 2001. Geologická mapa, list 25-14 Valašské Meziříčí.
Czech Geological Survey (in Czech).
Ehlers, J., Gibbard, P.L., 2004. Quaternary Glaciations — Extent and Chronology: Part I:
Europe. Elsevier, Amsterdam.
Ferguson, R., Church, M., 2009. A critical perspective on 1-D modeling of river processes:
gravel load and aggradation in lower Fraser River. Water Resour. Res. 45, W11424.
https://doi.org/10.1029/2009WR007740.
Folk, R.L., Ward, W.C., 1957. Brazos River bar: a study in the signiﬁcance of grain size parameters.
J. Sediment. Petrol. 27, 3–26. https://doi.org/10.1306/74D70646-2B21-
11D7-8648000102C1865D.
Fryirs, K., 2013. (Dis)connectivity in catchment sediment cascades: a fresh look at the
sediment delivery problem. Earth Surf. Process. Landf. 38, 30–46. https://doi.org/
10.1002/esp.3242.
Fryirs, K., Brierley, G.J., Preston, N.J., Kasai, M., 2007. Buffers, barriers and blankets: the
(dis)connectivity of catchment-scale sediment cascades. Catena 70, 49–67. https://
doi.org/10.1016/j.catena.2006.07.007.
Galia, T., Škarpich, V., 2016. Do the coarsest bed fraction and stream power record contemporary
trends in steep headwater channels? Geomorphology 272, 115–126.
https://doi.org/10.1016/j.geomorph.2015.07.047.
Galia, T., Škarpich, V., Hradecký, J., 2015. Connectivity of the coarsest fraction in headwater
channels: imprints of ﬂuvial processes and debris-ﬂow activity. Geogr. Ann. Ser. A
Phys. Geogr. 97, 437–452. https://doi.org/10.1111/geoa.12086.
Galia, T., Škarpich, V., Přibyla, Z., Hradecký, J., 2016. Effect of grade-control structures at
various stages of their destruction on local channel parameters. Geomorphology
253, 305–317. https://doi.org/10.1016/j.geomorph.2015.10.033.
Gomez, B., 1995. Bedload transport and changing grain size distributions. In: Gurnell, A.,
Petts, G. (Eds.), Changing River Channels. John Wiley, New York, pp. 177–199.
Gomez, B., Rosser, B.J., Peacock, D.H., Hicks, D.M., Palmer, J.A., 2001. Downstream ﬁning in
a rapidly aggrading gravel bed river. Water Resour. Res. 37 (6), 1813–1823. https://
doi.org/10.1029/2001WR900007.
Graham, D.J., Rice, S.P., Reid, I., 2005a. A transferable method for the automated grain
sizing of river gravels. Water Resour. Res. 41, W07020. https://doi.org/10.1029/
2004WR003868.
Graham, D.J., Reid, I., Rice, S., P., 2005b. Automated sizing of coarse-grained sediments:
image-processing procedures. Math. Geol. 37 (1), 1–28. https://doi.org/10.1007/
s11004-005-8745-x.
Gregory, K.J., 2006. The human role in changing river channels. Geomorphology 79,
172–191. https://doi.org/10.1016/j.geomorph.2006.06.018.
Gurnell, A., Surian, N., Zanoni, L., 2009. Multi-thread river channels: a perspective on
changing European alpine river systems. Aquat. Sci. 71, 253–265. https://doi.org/
10.1007/s00027-009-9186-2.
Havlík, A., 1999. Pilotní projekt Spojená Bečva — studie. Revital — inženýrská a projekční
kancelář, Praha (in Czech).
Hooke, J.M., 2003. Coarse sediment connectivity in river channel systems: a conceptual
framework and methodology. Geomorphology 56, 79–94. https://doi.org/10.1016/
S0169-555X(03)00047-3.
Horritt, M.S., Bates, P.D., 2002. Evaluation of 1D and 2D numerical models for predicting
river ﬂood inundation. J. Hydrol. 268, 87–99. https://doi.org/10.1016/S0022-1694
(02)00121-X.
Hrádek, M., 2000. Geomorphic effect of the July 1997 ﬂood in the North Moravia and
Silesia (Czech Republic). Geogr. Časopis 52, 303–321 (in Czech with English
summary).
Ivan, A., Kirchner, K., Krejčí, O., 2000. The morphostructural features of the Moravian part
of the Western Carpathians and Pannonian Basin. Geogr. Časopis 52 (3), 221–230 (in
Czech with English summary).
Jowett, I.G., Duncan, M.J., 2012. Effectiveness of 1D and 2D hydraulic models for instream
habitat analysis in a braided river. Ecol. Eng. 48, 92–100. https://doi.org/10.1016/j.
ecoleng.2011.06.036.
Kaiser, H.F., 1958. The varimax criterion for analytic rotation in factor analysis.
Psychometrika 23 (3), 187–200. https://doi.org/10.1007/BF02289233.
Knighton, A.D., 1980. Longitudinal changes in size and sorting of stream-bed material in
four English rivers. Geol. Soc. Am. Bull. 91, 55–62. https://doi.org/10.1130/0016-
7606(1980)91b55:LCISASN2.0.CO;2.
Knighton, A.D., 1999. The gravel–sand transition in a disturbed catchment. Geomorphology
27, 325–341. https://doi.org/10.1016/S0169-555X(98)00078-6.
Komar, P.D., Carling, P.A., 1991. Grain sorting in gravel-bed streams and the choice of particle
sizes for ﬂow-competence evaluations. Sedimentology 38 (3), 489–502. https://
doi.org/10.1111/j.1365-3091.1991.tb00363.x.
Kondolf, G.M., 1997. Hungry water: effects of dams and gravel mining on river channels.
Environ. Manag. 21, 533–551. https://doi.org/10.1007/s002679900048.
Kondolf, G.M., Piégay, H., Landon, N., 2002. Channel response to increased and decreased
bedload supply from land use change: contrast between two catchments. Geomorphology
45, 35–51. https://doi.org/10.1016/S0169-555X(01)00188-X.
Korpak, J., 2007. The inﬂuence of river training on mountain channel changes (Polish
Carpathian Mountains). Geomorphology 92, 166–181. https://doi.org/10.1016/j.
geomorph.2006.07.037.
Krapesch, G., Hauer, C., Habersack, H., 2011. Scale orientated analysis of river width
changes due to extreme ﬂood hazards. Nat. Hazards Earth Syst. Sci. 11, 2137–2147.
https://doi.org/10.5194/nhess-11-2137-2011.
Krejčí, L., Krejčí, M., 2012. Živá Bečva – cesta z regulace. Vodní Hospodářství 62 (12),
358–390 (in Czech with English summary).
Lamb, M., Dietrich, W., Venditti, J., 2008. Is the critical shear stress of incipient sediment
motion dependent on channel-bed slope? J. Geophys. Res. Earth Surf. 113. https://
doi.org/10.1029/2007JF000831.
Liébault, F., Piégay, H., 2002. Causes of 20th century channel narrowing in mountain and
piedmont rivers of southeastern France. Earth Surf. Process. Landf. 27, 425–444.
https://doi.org/10.1002/esp.328.
Macoun, J., Šibrava, V., Tyráček, J., Kneblová-Vodičková, V., 1965. Kvartér Ostravska a
Moravské Brány. Nakladatelství ČSAV, Praha.
Madej, M.A., 1995. Changes in channel-stored sediment, Redwood Creek, northwestern
California. US Geological Survey Professional Paper 1454-P.
Magilligan, F.J., 1992. Thresholds and the spatial variability of ﬂood power during extreme
ﬂoods. Geomorphology 5, 373–390. https://doi.org/10.1016/0169-555X(92)90014-F.
Malarz, R., 2004. The rate of gravel abrasion in the Carpathian rivers. Geogr. Časopis 56,
99–109.
Martín-Vide, J., Ferrer-Boix, C., Ollero, A., 2010. Incision due to gravel mining: modelling a
case study from the Gállego River, Spain. Geomorphology 117, 261–271. https://doi.
org/10.1016/j.geomorph.2009.01.019.
Menčík, E., Adamová, M., Dvořák, J., Dudek, A., Jetel, J., Jurková, A., Hanzlíková, E., Houša,
V., Peslová, H., Rybářová, L., Šmíd, B., Šebesta, J., Tyřáček, J., Vašíček, Z., 1983. Geologie
Moravskoslezských Beskyd a Podbeskydské pahorkatiny (Geology of the
Moravskoslezské Beskydy Mts and Podbeskydská pahorkatina Hilly land). ÚÚV v
nakladatelství ČSAV, Praha (in Czech, with English summary).
Mihola, L., 1992. 100 let úprav řeky Bečvy. Vodní hospodářství 2, 90–93 (in Czech with
English summary).
Mikos, M., 1994. The downstream ﬁning of gravel-bed sediments in the Alpine Rhine
River. In: Ergenzinger, P., Schmidt, K.H. (Eds.), Dynamics and Geomorphology of
Mountains Rivers. Springer, Berlin, pp. 93–108.
Nardi, L., Rinaldi, M., 2015. Spatio-temporal patterns of channel changes in response to a
major ﬂood event: the case of the Magra River (central-northern Italy). Earth Surf.
Process. Landf. 40, 326–339. https://doi.org/10.1002/esp.3636.
Ondráčková, L., Máčka, Z., 2018. Geomorphic (dis)connectivity in a middle-mountain
context: human interventions in the landscape modify catchment-scale sediment
cascades. Area https://doi.org/10.1111/area.12424.
Parker, G., 1990. Surface-based bedload transport relation for gravel rivers. J. Hydraul. Res.
28, 417–436. https://doi.org/10.1080/00221689009499058.
Parker, G., Clifford, N.J., Thorne, C.R., 2011. Understanding the inﬂuence of slope on the
threshold of coarse grain motion: revisiting critical stream power. Geomorphology
126, 51–65. https://doi.org/10.1016/j.geomorph.2010.10.027.
Pasternack, G.B., Gilbert, A.T., Wheaton, J.M., Buckland, E.M., 2006. Error propagation for
velocity and shear stress prediction using 2D models for environmental management.
J. Hydrol. 328, 227–241. https://doi.org/10.1016/j.jhydrol.2005.12.003.
684 V. Škarpich et al. / Science of the Total Environment 649 (2019) 672–685
Pavelka, J., Trezner, J., 2001. Příroda Valašska: okres Vsetín. Český svaz ochránců přírody
ZO 76/06 Orchidea, Vsetín (in Czech).
Powell, D.M., 1998. Patterns and processes of sediment sorting in gravel-bed rivers. Prog.
Phys. Geogr. 22, 1–32. https://doi.org/10.1191/030913398666402127.
Preciso, E., Salemi, E., Billi, P., 2012. Land use changes, torrent control works and sediment
mining: effects on channel morphology and sediment ﬂux, case study of the Reno
River (Northern Italy). Hydrol. Process. 26, 1134–1148. https://doi.org/10.1002/
hyp.8202.
Pregun, C.Z., 2016. Ecohydrological and morphological relationships of a regulated lowland
river; based on ﬁeld studies and hydrological modeling. Ecol. Eng. 94,
608–616. https://doi.org/10.1016/j.ecoleng.2016.06.125.
Punys, P., Adamonyte, I., Kvaraciejus, A., Martinaitis, E., Vyciene, G., Kasiulis, E., 2015. Riverine
hydrokinetic resource assessment — a case study of a lowland river in Lithuania.
Renew. Sust. Energ. Rev. 50, 643–652. https://doi.org/10.1016/j.rser.2015.04.155.
Rengers, F., Wohl, E., 2007. Trends of grain sizes on gravel bars in the Rio Chagres,
Panama. Geomorphology 83, 282–293. https://doi.org/10.1016/j.
geomorph.2006.02.019.
Rice, S., 1998. Which tributaries disrupt downstream ﬁning along gravel-bed rivers? Geomorphology
22, 39–56. https://doi.org/10.1016/S0169-555X(97)00052-4.
Rice, S., 1999. The nature and controls on downstream ﬁning within sedimentary links.
J. Sediment. Res. 69 (1), 32–39. https://doi.org/10.1306/D426895F-2B26-11D7-
8648000102C1865D.
Rice, S., Church, M., 1996. Bed material texture on low order streams in the Queen Charlotte
Islands, British Columbia. Earth Surf. Process. Landf. 21, 1–18. https://doi.org/
10.1002/(SICI)1096-9837(199601)21:1b1::AID-ESP506N3.0.CO;2-F.
Rice, S., Church, M., 1998. Grain-size along two gravel-bed rivers: statistical variation, spatial
pattern and sedimentary links. Earth Surf. Process. Landf. 23, 345–363. https://
doi.org/10.1002/(SICI)1096-9837(199804)23:4b345::AID-ESP850N3.0.CO;2-B.
Rice, S., Church, M., 2010. Grain-size sorting within river bars in relation to downstream
ﬁning along a wandering channel. Sedimentology 57, 232–251. https://doi.org/
10.1111/j.1365-3091.2009.01108.x.
Rinaldi, M., Wyżga, B., Surian, N., 2005. Sediment mining in alluvial channels: physical effects
and management perspectives. River Res. Appl. 21, 805–828. https://doi.org/
10.1002/rra.884.
Rovira, A., Batalla, R.J., Sala, M., 2005. Response of river sediment budget after historical
gravel mining (the lower Tordera, NE Spain). River Res. Appl. 21, 829–847. https://
doi.org/10.1002/rra.885.
Russell, D.R., 1939. Effects of transportation on sedimentary particles. In: Trask, P.D. (Ed.),
Recent Marine Sediments. AAPG, Tulsa, Oklahoma, pp. 33–47.
Schumm, S.A., 1977. The Fluvial System. John Wiley and Sons, New York.
Shields, A., 1936. Anwendung der aÈhnlichkeitsmechanik und der turbulenzforschung auf
die geschiebebewegung. Mitteilung der Preussischen versuchsanstalt fuer Wasserbau
und Schiffbau Heft 26 Berlin (trans. Ott, W.P. and van Uchlen, J.C., Pasadena, CA:
United States Department of Agriculture, Soil Conservation Service, Coop Lab., California
Institute of Technology).
Škarpich, V., Hradecký, J., Dušek, R., 2013. Complex transformation of the geomorphic regime
of channels in the foreﬁeld of the Moravskoslezské Beskydy Mts: case study of
the Morávka River (Czech Republic). Catena 111, 25–40. https://doi.org/10.1016/j.
catena.2013.06.028.
Škarpich, V., Galia, T., Hradecký, J., 2016a. Channel bed adjustment to over bankfull discharge
magnitudes of the ﬂysch gravel-bed stream — case study from the channelized
reach of Olše River (Czech Republic). Z. Geomorphol. 60 (4), 327–341. https://
doi.org/10.1127/zfg/2016/0395.
Škarpich, V., Kašpárek, Z., Galia, T., Hradecký, J., 2016b. Anthropogenic impact and morphology
channel response of Beskydian gravel-bed rivers: a case study of the
Ostravice River, Czechia. Geograﬁe 121, 99–120.
Song, S., Schmalz, B., Fohrer, N., 2015. Simulation, quantiﬁcation and comparison of inchannel
and ﬂoodplain sediment processes in a lowland area — a case study of the
Upper Stör catchment in northern Germany. Ecol. Indic. 57, 118–127. https://doi.
org/10.1016/j.ecolind.2015.03.030.
Stacke, V., Pánek, T., Sedláček, J., 2014. Late Holocene evolution of the Bečva River ﬂoodplain
(Outer Western Carpathians, Czech Republic). Geomorphology 206, 440–451.
https://doi.org/10.1016/j.geomorph.2013.10.015.
Sternberg, H., 1875. Über Längen- und Querproﬁl geschiebeführender Flüsse. Zeitschr.
Bauw. 25, 483–506.
Surian, N., 2002. Downstream variation in grain size along an Alpine river: analysis of controls
and processes. Geomorphology 43, 137–149. https://doi.org/10.1016/S0169-
555X(01)00127-1.
Surian, N., Rinaldi, M., 2003. Morphological response to river engineering and management
in alluvial channels in Italy. Geomorphology 50, 307–326. https://doi.org/
10.1016/S0169-555X(02)00219-2.
Uribelarrea, D., Pérez-González, A., Benito, G., 2003. Channel changes in the Jarama and
Tagus rivers (central Spain) over the past 500 years. Quat. Sci. Rev. 22, 2209–2221.
https://doi.org/10.1016/S0277-3791(03)00153-7.
Vannote, R.L., Minshall, G.W., Cummins, K.W., Sedell, J.R., Cushing, C.E., 1980. The river
continuum concept. Can. J. Fish. Aquat. Sci. 37 (1), 130–137. https://doi.org/
10.1139/f80-017.
Vojinovic, Z., Tutulic, D., 2009. On the use of 1D and coupled 1D-2D modelling approaches
for assessment of ﬂood damage in urban areas. Urban Water J. 6 (3), 183–199.
https://doi.org/10.1080/15730620802566877.
Wilcock, P.R., 1992. Experimental investigation of the effects of mixture properties on
transport dynamics. In: Billi, P., Hey, R.D., Thorne, C.R., Tacconi, P. (Eds.), Dynamics
of Gravel-bed Rivers. John Wiley, New York, pp. 109–131.
Wilcock, P.R., McArdell, B.W., 1993. Surface-based fractional transport rates: mobilization
thresholds and partial transport of a sand-gravel sediment. Water Resour. Res. 29,
1297–1312. https://doi.org/10.1029/92WR02748.
Wolman, M.G., 1954. Method of sampling coarse river-bed material. EOS Trans. Am.
Geophys. Union 35 (6), 951–956. https://doi.org/10.1029/TR035i006p00951.
Wyżga, B., 1993. River response to channel regulation: case study of the Raba River,
Carpathians, Poland. Earth Surf. Process. Landf. 18, 541–556. https://doi.org/
10.1002/esp.3290180607.
Wyżga, B., 2001. Impact of the channelization-induced incision of the Skawa and Wisłoka
Rivers, Southern Poland, on the conditions of overbank deposition. Regul. Rivers Res.
Manag. 17, 85–100. https://doi.org/10.1002/1099-1646(200101/02)17:1b85::AID-
RRR605N3.0.CO;2-U.
Wyżga, B., Zawiejska, J., Gurnell, A., 2018. Effects and persistence of river restoration measures:
ecological, management and research implications. Sci. Total Environ.
628–629, 1098–1100. https://doi.org/10.1016/j.scitotenv.2018.02.071.
Zawiejska, J., Wyżga, B., 2010. Twentieth-century channel change on the Dunajec River,
southern Poland: patterns, causes and controls. Geomorphology 117, 234–246.
https://doi.org/10.1016/j.geomorph.2009.01.014.
Zawiejska, J., Wyżga, B., Radecki-Pawlik, A., 2015. Variation in surface bed material along a
mountain river modiﬁed by gravel extraction and channelization, the Czarny
Dunajec, Polish Carpathians. Geomorphology 231, 353–366. https://doi.org/
10.1016/j.geomorph.2014.12.026.
685V. Škarpich et al. / Science of the Total Environment 649 (2019) 672–685