Stream bed stabilization using boulder check dams that mimic
step-pool morphology features in Northern Italy
Mario Aristide Lenzi
Department of Land and Agroforest Environment, University of Padova, Agripolis, Via Romea 16, 35020 Legnaro (PD), Italy
Received 10 April 2001; received in revised form 2 October 2001; accepted 4 October 2001
Abstract
During the last 10 years, the check dams made of boulders have been fully employed for torrent management, particularly
for high-gradient stream stabilization. The analysis of this typology is shown in relation to three aspects: building features, field
of use and design according to geomorphologic principles. The ``Maso di Spinelle'' torrent located in the Province of Trento,
Italy (North Italian Alps) is an example where a sequence of low-check dams made of boulders have been used for bed
stabilization. The design criteria are taken from the step­pool morphology features and the results are encouraging for further
applications. The artificial step­pool grade-control structures in the Maso di Spinelle torrent have been successfully tested by
floods events with return periods of about 7­10 and 20­25 years. D 2002 Elsevier Science B.V. All rights reserved.
Keywords: Catchment management; Streambed stabilization; Boulder check dams; Step­pool morphology
1. Introduction
Torrent control requires the design of diverse
interventions in the basin. Advanced technology and
superior knowledge enable the construction of inter-
ventions that are increasingly appropriate to their
objectives. These objectives are determined by what
is to be protected, from what danger and, most impor-
tantly, to what extent.
It is undeniable that each intervention in a complex
ecosystem, such as a stream, alters the balance of its
biotic communities (vegetation, fauna, microfauna)
and, therefore, its ecological stability. Equally undeni-
able, however, is the need to protect infrastructure and
habitations that could be damaged by floods, which in
a mountain area are absolutely ``natural'' and destined
to reoccur cyclically (Kettl, 1994; Della Giacoma et
al., 1991).
Current trends for mountain stream control, which
blend in with the ecosystem as much as possible, entail
the combined use of naturalistic engineering techni-
ques and methods of morphologically ``reconstruct-
ing'' the stream. These should be as compatible as
possible with the stream's natural tendency to reach a
stable configuration over a long period (Kondolf,
1996). Such criteria should be applied with care and
be accompanied by traditional structural interventions
(Lenzi et al., 2000). The latter are, in fact, essential both
in preventing the most severe disorders (such as rapid
phases of channel erosion and incision, debris flow and
near-bank landslides) and guaranteeing stability in
situations where the safety of the population must be
guaranteed. The two naturalistic­geomorphological
0169-555X/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved.
PII: S0169-555X(01)00157-X
E-mail address: marioaristide.lenzi@unipd.it (M.A. Lenzi).
www.elsevier.com/locate/geomorph
Geomorphology 45 (2002) 243­260
aspects and the classic structural one must, therefore,
interact, taking on different degrees of importance
within the control work, according to the level of pro-
tection required, the return time of the flood and the
objectives established during planning. Biological­
naturalistic engineering interventions aim above all to
restore the ecological functionality of the stream, es-
sentially through the reintroduction of riparian vegeta-
tion which, once established, can effectively perform a
number of functions (Schweitzer, 1995). Aspects of
restoration or rehabilitation of torrents connected to a
morphological ``reconstruction'' of the stream are, at a
practical level, much less common than biological­
naturalistic or engineering remedial measures (Sear et
al., 1995; Newson, 1995; Knighton, 1998). Neverthe-
less, research intending to propose constructional var-
iations of traditional concrete check dams and to study
the conditions of formation and stability of some
natural morphological units (riffle-pool, step­pool)
has greatly increased within the scientific community
and in some organisations with specific expertise in this
sector (Brooks and Douglas Shields, 1996; Lenzi et al.,
2000).
In particular, traditional check dams arranged in a
cascade, which reduce gradient through a series of
concrete works (Benini, 1990), have recently been
constructed using large boulders. In some cases, old
concrete check dams have even been converted (Fig.
1) into check dams made from boulders or ramps
(Haltiner et al., 1996; Brooks et al., 1996; Lenzi et
al., 2000). Reconstruction is aimed reducing the arti-
ficial effect of the concrete on the channel and, at the
same time, adhere firmly to control principles and
objectives. Such interventions are particularly appro-
priate in areas of tourist and recreational importance,
but more generally, they allow a more natural control
arrangement in many streams where there are already
large boulders on the bed (1­1.5 m diameter). Further
refinement of the planning stage for boulder check
dams may be carried out, taking into account the
similarity between the morphology of ``step'' and
``pool'' physiographical units and traditional stabilisa-
tion works. Indeed, the longitudinal profile of step­
pool reaches resembles a series of check dams (Fig. 2).
The distance between the steps in a sequence, Ls, is
determined by stream slope, S, step height, H, and bed
roughness as well as by the floods that formed them
(Lenzi, 1999; Lenzi and D'Agostino, 1998, 2000). The
channel arrangement found in step­pools appears to
be the outcome of a natural evolutionary process
which tends towards a certain stream stability (Lenzi,
1999, 2001). The end results of complex natural hy-
draulic, sedimentological and geomorphological inter-
actions, which are characteristic aspects of step­pool
morphology, may be taken as a starting point and ap-
plied appropriately to the aimed channel ``reconstruc-
tion'' (Lenzi, 2001).
In this article, construction designs currently used
for boulder check dams will be described. An example
of their planning and design according to geomorpho-
logical and stream bed stabilization criteria will be
given, with reference to an artificial step­pool seq-
uence structure constructed with large boulder in the
Province of Trento, North Italian Alps.
2. Construction designs and fields of use for boulder
check dams
Large boulder stabilisation works include concrete
check dams covered with boulders, structures with
boulders anchored to the bed by micropiles and boul-
der check dams either uncemented or strengthened
with cement. The use of large boulders in torrent con-
trol was directed mainly at longitudinal bank protec-
tion structures (rip-rapping, deflectors, training walls).
During the course of these works, it was possible to
reduce labour requirements considerably, thereby cut-
ting the costs of traditional forms of stabilisation. On
the other hand, these interventions could not and, still
cannot, be applied in more complex situations. For
example, in steep streams where energy excesses with
sudden planimetric and altimetric changes, consider-
able bedload or debris flows have to be dealt with. The
use of these construction designs is further conditioned
by technical limitations. Exceeding certain heights (2­
2.5 m) is problematic even though the structure's
strength is largely assured by the size of the boulders.
Durability is another factor that should not be over-
looked because it appears to be slightly more limited
than that in traditional concrete check dam structures.
To overcome some of the limitations of using unce-
mented or cemented boulder check dams, new con-
struction designs have been researched which
guarantee greater strength against violent floods and/
or debris flows, even for high-drop structures. At the
M.A. Lenzi / Geomorphology 45 (2002) 243­260244
same time, attempts have been made to increase du-
rability, to keep costs within limits comparable to those
of traditional concrete check dams and to better con-
ceal any concrete which may be visible through the
interstices (D'Agostino et al., 1997; Lenzi, 2000).
In an attempt to combine all these objectives, the
Torrent Control Agency of the Autonomous Province
of Trento has been developing a design, widely used in
the east of Trentino, since 1990. This involves works
with a reinforced concrete framework to which very
large, uncut boulders are firmly tied and anchored.
This design has been subject to development, which
aims to save concrete and achieve greater construc-
tional simplicity. In the most recent works, the up-
stream framing has been replaced by a rudimentary
baffle against the scour front. This means there is less
scour, greater safety for workers and reduced con-
struction time. This construction design is now the
Fig. 1. Longitudinal profile showing the reconstruction of the old grade control structure (redesign on the base of a project of the Torrent Control
Agency of the Autonomous Province of Trento) (source: Lenzi et al., 2000).
Fig. 2. Sketch of a natural step­pool sequence (Ls = step wavelength; Lp = pool wavelength; H = step height; B = width of the step riser).
M.A. Lenzi / Geomorphology 45 (2002) 243­260 245
most commonly used in Trento Province and results in
works of considerable strength and durability (Lenzi et
al., 2000). Some such stabilisation works have already
been rigorously tested by floods with return intervals
over 30 years and even by a debris flow.
Cemented boulder check dams are obviously a
simplification of the check dams previously described:
the upstream framing is rarely made, there is no
reinforcement and fastening and the amounts of con-
crete are clearly low as the body of the construction
behind the downstream face is reduced to the mini-
mum necessary to structurally join the boulders and
give the check dam a certain monolithicity. Two to four
square metres of concrete are used per square metre of
downstream face for tied works, but only 0.5­1.5 m3
for works that are just cemented. Even the designs
considered here have their limitations and their con-
struction requires particular care. Their use is recom-
mended in channels characterised by a uniform, coarse
distribution of bed material. Indeed, if no large boulder
is present, the control would have a completely artifi-
cial effect, which may be worse than that obtained
from traditional works. In these circumstances, it
would be necessary to transport blocks from other
locations with a consequent increase in costs. When
choosing between these different boulder check dam
designs, determining factors are protection aims and
the nature of what is to be protected (residential areas,
infrastructure, extensive or intensive farmland, natural
landscape). It is, therefore, logical to hypothesise a
high degree of protection for areas of high value and a
lower degree of protection for farmland and forest
areas, leaving aside the now outmoded concept of a
standard size for a flood with a 100-year return time
(UFEA, 1995). Table 1 proposes an outline of the
fields of use for the main kinds of boulder check dams,
taking into account the degree of protection required,
the kind of solid transport and some basic geomorpho-
logical parameters of the stream. Conditions for the
construction of artificial steps (uncemented boulder
check dams or check dams strengthened with cement)
are as follows: (a) bed slope reach should be typical of
that of step­pool sequences (5­20%); (b) sufficient
availability of large boulders in the stream; (c) possi-
Table 1
Fields of use for boulder check dams
Uncemented boulder check dams
(artificial steps, riffle steps)
Check dams strengthened with cement
(artificial steps, riffle steps)
Check dams with reinforced concrete
framework and tied boulders
Characteristics of reach where control works are to be constructed
Gradient up to 12­14% Gradient up to 18­20% Gradient up to 18­20%
Grain size of surface material
(a) Wide and complete (heterogeneous) (a) Quite graduated (a) Quite uniform
(b) Consistent presence of large boulders (b) Prevalence of gravel and pebbles,
but also a certain number of boulders
(b) Prevalence of sand, gravel and
small pebbles; a few boulders
Morphology
(a) Not changed by previous interventions (a) Slightly changed by previous interventions (a) Already changed by previous
interventions
(b) Presence of step­pool and/or riffle
step­pool sequences
(b) Presence in the stream and/or its tributaries
of short step­pool and/or riffle step­pool sequences
(b) Not very structured channel
Sediment transport
(a) Bedload (a) Bedload (a) Bedload
(b) Hyperconcentrated flow (b) Hyperconcentrated flow
(c) Debris flow
Abundance of large boulders present
in stream
Considerable availability of boulders in the stream Hardly any boulders in the stream
Surrounding environment
Natural, farmland Natural, farmland, slight urbanisation Built up
Degree of safety required
Return time: 20­30 years Return time: 30­50 years Return time: > 50 years
M.A. Lenzi / Geomorphology 45 (2002) 243­260246
bility of basing these works on large boulders already
present in stream; (d) presence of well-defined step­
pool sequences in stream or in some of its tributaries.
Design criteria for these transverse works, which
attempt to artificially reconstruct a natural step­pool
morphology, can make use of research carried out in
fluviomorphological and hydraulic fields on step­
pool sequences (Lenzi, 2001).
To guarantee a wide safety margin, it is advisable to
base the step on at least two of these foundation
boulders whilst a suitable foundation should be laid
for the rest of the structure, with boulders at least 1.5­
2.0 m deep. Building the step requires careful excavator
work so that the boulders anchor each other in, making
the whole step stable. Except in certain cases, the step
should not be higher than 2.5 m above the ground. This
guarantees a more natural effect and also avoids laying
more than three rows of boulders, which would make
the step structurally less stable. Structural stability can
be improved by slightly arching the step towards
upstream or it can be supported on the banks with an
open ``V'' shape. Even the increased stability of an
arched planimetric shape can be put at risk when banks
erode. In these cases, there should be both good lateral
keying (4­5 m), to prevent possible turning of the step,
and a continuous lateral bank protection with rootwads/
boulders (D'Agostino et al., 1997; Lenzi, 2000).
A constructional precaution to increase stability can
consist in creating supplementary anchoring points,
such as micropiles placed upstream of the step and tied
to the foundation boulders with steel cables. Torrent
control works built according to the above criteria
blend in well with their natural surroundings. To fur-
ther reduce the impact of rip-rap on the environment,
and to speed up revegetation, willow cuttings can be
planted between the boulders.
3. Study area
In 1996 and 1997, the Torrent Control Agency of
the Autonomous Province of Trento, with the scientific
support of the University of Padua, carried out hy-
draulic control works on some reaches of a mountain
stream, the Maso di Spinelle, using transverse boulder
works for channel protection against incision. The
Maso di Spinelle forms, along with the Maso di Ca-
lamento, the Maso, a tributary that flows into the River
Brenta, from the left bank, near the village of Castel-
nuovo (Fig. 3a) in Northern Italy. The main physio-
graphic characteristics of its basin are synthesised in
Table 2. The climatic conditions are typical of an
Alpine environment. Precipitation occurs mainly as
snowfall from November to April. Runoff is usually
dominated by snowmelt in May and June but summer
and early autumn rains represent an important contri-
bution to the flow regime. Usually, late autumn, winter
and early spring lack noticeable runoff events.
Soils in general are thin and belong to three main
families: (a) skeletal soils, occurring on steep slopes
with a discontinuous vegetation cover; (b) organic
soils, with more continuous vegetation cover; (c)
brown earth soils. Vegetation cover consist of herba-
ceous associations, including both continuous-cover
mountain grassland (6% of catchment surface) and
sparse grassland (5%). Shrubs are well widespread
(24%), while forest stands made up of spruce and larch
are found in the and lower parts of the catchment and
occupy 53% of the total area; 12% of the catchment
consists of bare land.
Sediment sources in the Maso di Spinelle catchment
were mapped through field surveys in the summer of
1996. Area, elevation, slope and vegetation cover of
sediment sources were measured in the field. Inves-
tigated areas were sketched, photographs were taken
and sediment samples were collected. The resulting file
includes several sediment source areas, covering 7.2%
of total basin area. Active sediment sources mainly
consist of stream bed incisions, eroded banks and
landslides along streams, shallow landslides, bare
slopes, overgrazed areas, low-order channels and deb-
ris flow tracks (Sonda, 1998). The geology of the basin
has contributed consistently to such large alterations of
the stream, which are also documented by aerial photo-
graphs taken after the 1966 flood (Figs. 3b and 4). The
geological formations of the valley result from the
great crystalline (magmatite) presence of the Cima
D'Asta. Vulcanite protrudes in the lower reach (partic-
ularly on the right bank). Metamorphites (phyllites,
cornubianites, paragneiss) are present in the upper
reach. The channel is dominated by degraded deposits
of Quaternary age, particularly in the lower reach
between Pontarso and the confluence with the Rivo
Caserine. Moraine and fluvioglacial deposits are highly
evident close to the banks and are often covered by
talus. The moraine material has led to a high percent-
M.A. Lenzi / Geomorphology 45 (2002) 243­260 247
Fig. 3. (a) Location map of the Maso di Spinelle catchment. (b) Erosion processes on the main stem.
M.A. Lenzi / Geomorphology 45 (2002) 243­260248
age of sandy and sandy­gravelly banks on which there
are large boulders of various shapes. Some of these
boulders, which range up to some hundreds of cubic
metres, caused huge flow diversion during the 1966
and 1993 floods, leading to bank slides (Sonda, 1998;
Cerato, 1999).
The streambed of the Rio Maso di Spinelle con-
sists of five different types of channel reach: they were
identified during summer 1996 as step­pool, cascade­
pool (step­pool­cascade), pool­riffle and riffle step
(plane-bed­step­pool) (Lenzi et al., 2000). Short bed-
rock reaches are also present but they are restricted to
the catchment headwaters and tributaries. Step­pools
reaches of the Maso di Spinelle showed a disorganiz-
ed pattern, associated with steep gradients (10% < S <
15%), and confinement by the valley walls. The cas-
cade­pool reaches are step­pool sequences irregular-
ly punctuated by small heaps of ill-formed gravel bars
of coarse material, while pool­riffle reaches are char-
acterized by well defined lateral and central bars and
are bounded by a small eroded alluvial plain. Cas-
cade­pool reaches are similar to step­pool reaches
but with coarse particle bars deposited upstream of the
boulder steps, or adjacent isolated big boulders, and
largely infilling the upstream end of pools. This latter
Fig. 4. Aerial photograph taken after the 1966 flood showing the location of the different types of protective structure built since 1980
(photograph courtesy of the Torrent Control Agency of the Autonomous Province of Trento).
Table 2
Main physiographic and climatic characteristics of the Maso di
Spinelle catchment
Catchment area (km2
) 45.0
Average elevation (m.a.s.l.) 1860.0
Minimum elevation (m.a.s.l.) 909.7
Maximum elevation (m.a.s.l.) 2561.6
Mean slope gradient (%) 50.3
Length of the main stem (km) 10.0
Mean gradient of the main stem (%) 13.8
Mean annual precipitation (mm); mean annual temperature (°C)
U Pontarso station 1100­7.7
U Costa Brunella station 1300­3.4
M.A. Lenzi / Geomorphology 45 (2002) 243­260 249
channel type generally occurs in the Maso di Spinelle
torrent on step slopes (S >12­14%) and seems to
reflect both the description and the diagnostic features
of the ``cascade channel'' proposed by Montgomery
and Buffington (1997). Riffle step reaches occur in the
Maso di Spinelle torrent at moderate to high slopes
(7% < S < 10%) in relatively straight reaches and un-
confined by valley walls. They are composed of bould-
er grain sizes, but are dominantly gravel to cobble
bedded. Tributaries of the Maso di Spinelle stream:
Rivo di Caserine, Rio Montalon, Rio Val Sorda and
Rio Val Conseria are characterised by prevailing well
organised step­pool reaches sequences.
Channel reaches 10­15 channel widths in length
were surveyed throughout the Maso di Spinelle drain-
age basin. Each reach was classified into one of the
above-defined channel types. Reach slopes were sur-
veyed using either a total station positioning system or
a clinometer, a meter tape, a laser diastimeter and a
stadia rod. Topographic surveys and channel-spanning
pebble counts of 100 grains (Wolman, 1954) were con-
ducted at representative cross-sections. Reach slopes
were determined from topographic maps for some
additional reaches where morphologies were mapped,
but slope and grain-size measurements were not col-
lected (Lenzi and D'Agostino, 1998).
Particular attention was paid to the analysis of the
geometric relationships of the step­pools reaches sur-
veyed on the Maso di Spinelle tributaries, due to their
use in the torrent control designs.
4. Channel stabilization according to
morphological criteria
The Maso di Spinelle's bed, especially the middle-
lower part, was subject to considerable incision and
deepening during the historic flood of 3­4 November
1966 (100 < Tr < 200 years). To give an idea of the
degree of disorder of the channel, the average width of
the bed downstream of the Rivo di Caserine (Figs. 3
and 4) increased from 6­7 to 35­50 m after the flood,
while the bed profile went down by about 10­15 m in
some reaches. Similar results, though not so accentu-
ated, were recorded more recently following the flood
event of 2 October 1993 (50 < Tr < 75 years).
One reach of the Maso di Spinelle, near the con-
fluence with the Maso di Calamento, was subject to a
classic intervention with concrete check dams during
the 1980s (Fig. 4). As a result of the incision caused by
the 1966 flood, these dams were about 7 m high.
Following the 1993 flood event, further intervention
was deemed necessary to stabilise the worst affected
reaches of the channel. Within the framework of the
Maso di Spinelle Watershed Master Plan, proposed
activities and priorities of measures (Action plans)
were elaborated taking into consideration a geomor-
phological approach at catchment scale and the most
important objectives of flood protection and ecology.
The procedure used for linking both the catchment
analysis to planning and the researches at catchment-
reach scale is reported on Fig. 5. A broader catchment-
scale perspective which emphasizes the physical integ-
rity of the drainage basin and the close links between
catchment and channel dynamics was adopted. Catch-
ment response to catastrophic flooding (1982, 1966
and 1993) was reconstructed from historical docu-
ments and field evidence in order to indicate spatial
variations in geomorphological effectiveness. The role
played by events of different magnitude and frequency
and the conditions which can lead to instability of the
main channel were also analysed. Upstream-progress-
ing degradation as response to a fall in base-level at the
confluence with the Brenta River was identified as the
maindominanterosionprocesses.Degradationprocess-
es became more complex because interaction between
a main stream and its tributaries was also deected.
Channel confinement, due to the narrow valley mor-
phology, accelerates the incision and inhibits gradient
adjustment through a change in sinuosity and rough-
ness characteristics.
Maso di Spinelle Master Plan highlights the neces-
sity of preserving stable stream environments which
means preserving the long term stability of factors that
influence stream dynamics and evolution. To enable
this, the Maso di Spinelle torrent was divided into
characteristic reaches to be preserved or rehabilitated
while recognizing that these reaches are parts of a
continuum of flow and sediment transfer (Fig. 5). This
plan adopts a geomorphological approach, which both
complements and supplements that of river engineer-
ing, as underlined by Sear et al. (1995), Thorne et al.
(1997) and by Knighton (1998), as well as applying
some of the principles of ``engineering geomorphol-
ogy'' developed by Bravard et al. (1999). Complete
restoration to a predisturbance state was not considered
M.A. Lenzi / Geomorphology 45 (2002) 243­260250
Fig. 5. Schematic illustration of the Maso di Spinelle Watershed Master Plan.
M.A. Lenzi / Geomorphology 45 (2002) 243­260 251
a viable option, so a partial restoration (rehabilitation)
of the stream was proposed, attempting to improve the
hydraulic, biological and aesthetic status of the torrent.
The abundance of fine and poorly sorted material on
the bed, combined with the disorderly arrangement of
the large boulders present, suggested remedial options
of minor impact which would also enhance the water
course from an aesthetic and environmental aspect.
The first such attempt was made following the 1993
flood in a reach affected by 10­15 m of bed lowering
(Fig. 4). A block stone ramp was built from large boul-
ders tied together by steel cords and laid on the sandy
subsoil. The creation of this long, very rough ramp
prevented the channel from taking on an artificial,
engineered appearance, whilst its fixed bed armouring
ensured almost complete immunity from erosion. The
only drawback was perhaps the large workforce
required to drill the boulders making up the artificially
paved channel.
Since the channel required stabilising in two further
reaches upstream that were affected by headward
cutting, both to avert the risk of new bed incisions
and to strengthen some bank slides overlooking the
channel, it was decided to reproduce natural step­
pool structures using boulder check dams. Conditions
allowing this kind of intervention were: the gradients
(10­14%) typical of step­pool reaches, the high
availability of large boulders near the stream, the
possibility of building these works on large boulders
already present in the streambed and the presence of
well-defined step­pool sequences on some tributaries
of the Maso di Spinelle (Rivo di Caserine, Rio Val
Sorda and Rio Montalon). The building solutions
adopted envisaged the construction of uncemented
boulder check dams for a 500-m reach of the channel
downstream of the Maso's confluence with the Rivo di
Caserine (Figs. 3 and 4) and the installation of boulder
check dams strengthened with cement for the control
of a reach further downstream, situated upstream of
the block stone ramp (Sonda, 1998; Lenzi et al.,
2000).
In accordance with the largest grain-sizes in the
stream (D90, D100), a structure's height above ground
(H) can be calculated, maintaining an average relation
H/D90 = 2 and keeping within the range H/D90 = 1­4
(Lenzi and D'Agostino, 1998; Lenzi, 1999). After
conducting four grain size surveys in the zones under-
going stabilisation works according to the ``grid by
number method'' (Wolman, 1954, modified by Lenzi
1992), the most suitable heights were assigned to the
steps (check dams). According to the largest grain
sizes sampled in the channel (D90 = 0.7­2 m), values
between 1 and 3.5 m (with a more frequent value of
2.5 m) were assigned to the height above the ground
(H) of the structures.
The distance between steps (Ls = L) can be esti-
mated beforehand by applying the maximum flow
resistance criteria as determined by Abrahams et al.
(1995) in the laboratory. The authors, through four
interconnected series of experiments, have observed
that on a step­pool reach, there tends to be a con-
dition of low mean flow velocity (and, therefore, of
maximum global resistance to flow), when there is a
combination of regular spacing between steps and the
geometric condition:
1 < H=L=S < 2 1
where H = the vertical height of the steps in the
sequence and L= distance between steps, estimated
in a parallel direction to the slope S of the stream
(Fig. 2).
A comparable result to Eq. (1) has been obtained
by D'Agostino and Lenzi (1998) from a sample of 49
natural step­pool sequences (Eastern Alps), for which
the following equation has been obtained:
0:5 < H=L=S < 2:1: 2
D'Agostino and Lenzi (1998) have, however,
obtained a mean value of 1.4 for the (H/L)/S parameter
which, therefore, corresponds with value proposed by
Abrahams et al. (1995). In planning, one can attempt to
create artificial steps which fall into the range sug-
gested by Eq. (1) and which tend, where possible,
towards H/(LS) = 1.5. The criteria, however, should be
applied in a flexible way, trying to follow the altimetric
and planimetric course of the stream. One should also
bear in mind local changes in gradient, widening, and
sections which are most suitable for lateral keying
(Lenzi and D'Agostino, 1998). Likewise, in assigning
the wavelength, it is better to move within a certain
range of variation for Ls, rather than attempting to keep
a fixed distance between works. Lenzi and D'Agostino
(1998) have noted average values of Ls between 0.5
and 1.6 times the breadth (B) of the stream in corre-
spondence to the steps. An analogous study by Chin
M.A. Lenzi / Geomorphology 45 (2002) 243­260252
(1999) (Santa Monica Mountains, CA) highlights how
frequency peaks of the variable Ls/B are between 0.5
and 1.5 and how there was a notably low occurrence of
values higher than 5.
After conducting a detailed longitudinal topo-
graphic survey of the restored main stream, the field
of wavelength, L= H/(1.1­1.3S) was chosen with dis-
tances of 10­25 m between the works.
Determination of discharge over the artificial steps
can be carried out according to the estimation of peak
discharge under critical flow conditions. It is neces-
sary that the flow downstream of each step is not
submerged by the water level caused by the next step
downstream. An observation of the flow profile
between the pool and the head of the next step during
floods shows a strong ponding of the water and an
almost horizontal hydraulic grade line with a very
gradual lowering of the free surface towards the step.
Cautiously one can then assume that the level of the
hydraulic grade line does not go beyond the value:
ym  1:5q2
=g0:333
3
where ym = flow depth slightly upstream of the step,
estimated according to the height of the jutting edge
and q = design inflow per unt of transverse width of
the step (m2
s  1
).
The design water discharge for a return interval of
100 years was estimated to be 85 m3
s  1
by Sonda
(1998). Considering the maximum width of the flow
as equal to 23.5 m, is possible to obtain a value of ym
equal to 1.65 m; the flow depth slightly upstream of the
step, estimated according to the height of the jutting
edge and q the design inflow per unit of transverse
width of the step (m2
s  1
). This value suggested that
lateral bank stabilization should be built to a height of
2 m, using rootwads/boulders.
To estimate any possible structure failure, a refer-
ence design hypothesis, especially in the case of
uncemented artificial steps, is represented by the cal-
culation of stability for each boulder in the work. It
should be pointed out that when analysing the extreme
equilibrium condition for a cubic element (Benini,
1990), the limited sliding velocity is lower than that
for rolling. For boulders in an artificial step, for which
the longer axis is laid in the direction of flow, the
limited sliding velocity on the bed or on the boulder
below is, therefore, worse for stability than that of
rolling. It is difficult to describe the general shape of a
boulder with regular geometric terms. Usually they are
pseudo-elliptic shapes which can be categorised
according to their three main axes: in the direction of
flow (Dx), perpendicular to the bottom (Dy) and in a
transversal direction (Dz) to the flow (Fig. 6). If a
generic boulder is placed on a slope S = tgb and struck
by a water flow with speed Vf (Fig. 6), the forces which
destabilise the boulder are the drag force ( FD), the lift
force and the weight of the element in the water in the
direction of movement ( Px).
Fig. 6. Sketch of the forces acting.
M.A. Lenzi / Geomorphology 45 (2002) 243­260 253
Friction with the bed ( FA), which depends on the
immersed weight of the boulder, and actions of lateral
friction and anchoring between surrounding boulders,
contribute to boulder stability.
In developing the condition of stability, it has been
decided to ignore the lateral anchoring and the effect
lift force has in making the boulder lighter. The stabi-
lity condition, therefore, is:
FD  Px  FA: 4
The drag force is a function of; the coefficient of
hydrodynamic resistance of the element (CD) struck
by flow, the density of the liquid (q), the surface
covered in the normal direction of flow (Ax) and of the
flow velocity (Vf). Therefore:
FD  CDqAXVf 2=2: 5
Px can be expressed by the relation:
Px  qs  qg V sin b 6
where V is the boulder volume and qs its density.
Friction force FA is given by the product of the
orthogonal immersed weight force to the bottom for
the friction coefficient ( f ) between boulder and
boulder (v between boulder and bed):
FA  f qs  qg V cos b: 7
The boulder volume can be expressed as:
V  aAxDx 8
where a is a shape parameter with 1 for a prismatic
solid, 2/3 for an ellipsoid and 0.6­0.8 for natural or
broken boulders.
If Eq. (8) is now replaced in Eqs. (6) and (7) and
these are then inserted with Eq. (5) in Eq. (4) one
obtains:
CDqAxV2
f =2  qs  qg a AxDx sin b
 f qs  qg a AxDx cos b 9
from which the near bed velocity under incipient
movement conditions can be obtained:
Vf  M2qs  qg a f cos b  sin bDx=CDq:
10
The drag coefficient CD essentially depends on four
factors: Reynolds number, Froude number, shape of
the boulder and relative submersion. Depending on the
value obtained for the Reynolds number, the boundary
layer on the boulder varies from laminar to turbulent
with a corresponding shift of the separating zone of
flow and, therefore, of the form resistance induced by
the boulder. Bathurst (1993) has observed that this can
be considered a secondary factor for a natural boulder,
since irregularities in form and possible angularity of
the boulder have considerably more influence than the
boundary layer on the position of the separating zone.
In the case of moderate flows, the effects of Froude
number (Fr) and relative submersion (that is the
relation between the flow ( ym) upstream of the boulder
and its size (Dy) ) have a greater influence. The in-
fluence of Froude number is connected to the energy
loss that the boulder's protrusion causes at the free
surface. Even though there is not an exact estimation
for the relation that links the coefficient CD to Fr and to
ym/Dy, a significant indication is given in a study by
Flammer et al. (1970). The authors carried out experi-
ments on a hemisphere with height Dy on a smooth
surface, and showed that, as with relative submer-
gence, the coefficient CD decreases with the increase
of Fr. Flammer et al. (1970) highlight three distinct
behaviour fields: the first is characterised by ym/Dy
values higher than 4 where the dissipating surface
effects are negligible and, therefore, where CD, not
influenced by the Fr value, remains nearly constant
(CD = 0.3­0.5); a second field, included in the interval
ym/Dy = 1.5­4, where the surface effects begin to be
noticeable and the coefficient CD just manages to over-
take the unit for Fr = 0.5 and ym/Dy % 1.7; the third
field has a low relative submergence ( ym/Dy < 1.5)
where surface disturbance is very evident and the in-
fluence of Froude number on CD predominates (Lenzi
et al., 2000).
To propose a simplification of Eq. (10), which con-
servatively accounts for the CD value, one could hypo-
thesise that the boulders on the upstream face of the
crest of an artificial step are not completely silted up
(or even that there has been a partial removal of ma-
terial artificially placed between the boulders). Fol-
lowing this hypothesis, each boulder would be struck
by a flow whose regime is slightly lower than the
critical state (Fr = 0.9) and characterised by a relative
submergence value close to unity. With these values,
M.A. Lenzi / Geomorphology 45 (2002) 243­260254
Flammer et al.'s graph suggests CD = 2. If, in addition,
one enters into Eq. (10) a submerged boulder density
of 1.65, a friction coefficient f = 0.7, an intermediate
value of the form parameter (a = 0.7) and one ignores
the bed gradient (b = 0), the equation is reduced to:
Vf % 3MDx: 11
A similar relation was already indicated by Stern-
berg (1875) with a coefficient of 4 and a reference
diameter corresponding to the b axis rather than to the
long axis Dx as in Eq. (11).
In testing the stability of artificial steps, the velocity
which is obtained from Eq. (11) has to be estimated for
the Dx values of smaller particles in the construction.
Moreover, in low submergence conditions, the flow
speed affecting boulders in the central zone of surface
Ax cannot be estimated by hypothesising a logarithmic
profile of velocity (Nowell and Church, 1979; Bath-
urst, 1993). It is, therefore, advisable to directly
compare the velocity Vf with the mean flow velocity
(V ) (Benini, 1990). A conservative deduction can be
made by applying the hypothesis that the design inflow
is already found in transition conditions slightly
upstream of the boulder. This velocity, therefore, is:
V  Mg ym=1:5 12
where ym is expressed by Eq. (3). The safety coef-
ficient for checking the stability of boulders in a step,
then, will be represented by the relation between Vf
and V. It would have to be higher than the unit for
cemented boulder works and at least equal to 1.2­1.3
for uncemented boulder steps.
Boulders forming the upper part of the steps are
characterised by an axis parallel to the flow direction
(Dx) of at least 1.6­1.7 m. If Eq. (11) is applied, a
value of Vf = 3.8 m s  1
is obtained; this value is the
mean greater than the mean flow velocity calculated
from Eq. (12) (V = 3.3 m s  1
), and so a degree of
safety relating to the stability of boulders is greater
than unity.
5. Geomorphological implications of boulder check
dams
Traditional streambed stabilization through the use
of concrete check dams requires the designer to de-
termine the ultimate slope of the channel upstream of
the dam (Fig. 1). In artificial boulder step sequences
with a reverse slope longitudinal profile (Fig. 2) and
characterized by a condition of maximum flow resist-
ance (steepness parameter c = (H/Ls)/S, between 1.2
and 1.4 in this study), it is not necessary to predict the
equilibrium slope. In step­pool systems, water flows
over groups of large elements that act like weirs and
plunges into pools below these elements. The flow in
step­pool systems is described as tumbling flow, in
which a great part of the flow energy is dissipated by
rollers in the pools. Wohl and Thompson (2000) meas-
ured velocity profiles and velocity fluctuations over
steps and pools throughout the snowmelt hydrograph.
They found that flow became more turbulent as stage
increased, particularly at lower-gradient reaches with
less variable bed roughness. Flow also became more
turbulent as gradient increased, and as bed roughness
increased. The wake-generated turbulence leads to
higher energy dissipation in step­pool reaches relative
to more uniform-gradient reaches (Wohl, 2000).
Morphological implications and effects of boulder
check dams can be synthesized as follow: if we assume
both that the breath (B) of the restored stream and the
distances between step, Ls, can not change even during
less-frequent higher-discharges, the only possibility
for changing the configuration is the variation of the
H parameter. Then, during a series of ``ordinary''
(Lenzi et al., 1999) flood events, we will expected
bed armouring to dominate the sediment transport res-
ponse and a stable bed. However, following an extra-
ordinary flood (Tr > 50 Years) and unlimited supply
conditions, the steepness factor, c, should suddenly
decrease as a result of sediment trapped in the pools,
and a general levelling of reaches should also be ex-
pected. The higher gradient of the artificial step­pools
along with the high overall channel roughness should
be replaced by a morphological configuration with a
flatter profile and reduced channel roughness. Lenzi
(2001) suggested that this transition from a step­pool
profile to a plane bed profile is more attributable to the
downstream transportation of solid materials than
erosion processes in pools, and causes the burial of
transversal structures. After an extraordinary flood,
subsequent ordinary floods can reestablished the mor-
phological features of the step­pools scouring out
loose sediment from pools. In this way, pools become
deeper inducing the formation of a negative bed slope
M.A. Lenzi / Geomorphology 45 (2002) 243­260 255
between steps, and due to fact that bed erosion pro-
cesses and pool scouring will prevail over deposition,
the relationship c = (H/Ls)/S will assumes a value equal
or greater than 1.
Bedload studies in natural step­pool channels dem-
onstrate complex relations between discharge and sedi-
ment transport: transport rates are dependent on
seasonal and stochastic sediment inputs, flow magni-
tude and duration, and antecedent events (Ashida et al.,
1981; Sawada et al., 1983; Whittaker, 1987; Schmidt
and Ergenzinger, 1992; Warburton, 1992; Montgom-
ery and Buffington, 1997; Blizard and Wohl, 1998;
Wohl, 2000; Lenzi, 2001). Warburton (1992) sug-
gested three phases of sediment transport in natural
step­pool channels: a low-flow flushing of fines;
frequent high-flow mobilization of pool-filling gravels;
and less-frequent higher-discharge mobilization of
step-forming grains (Montgomery and Buffington,
1997). Ashida et al. (1981) observed a 10-h lag
between hydrograph peak and onset of bedload trans-
port for step­pool channels scoured of all pool-filling
sediment during a previous storm. Hydrograph peak
and sediment transport were, however, directly corre-
lated during a subsequent storm due to the availability
of sediment deposited in pools. Attempts to predict
sediment transport along step­pool channels using
some of the standard sediment relations developed
for coarse sediments and steep gradient indicated that
these relations tended to over-predict actual measured
bedload transport by more than three orders of magni-
tude (Wohl, 2000). Working on a step­pool channel in
the Rio Cordon (Northeastern Italy), Lenzi (2001)
found that ``the evolution over time of the c parameter
mimics the trend of the measured bedload rate for
different floods events. Before the September 1994
large flash flood the c = (H/Ls)/S parameter close to 1.5
reflects a stable bed and well developed coarse armour.
The bedload rate was, therefore, almost two orders of
magnitude less than the transport capacity computable
from bedload formulae (Lenzi et al., 1999; D'Agostino
and Lenzi, 1999). The decrease of c during the Sep-
tember 1994 flood was accompanied by the break-up
of coarse armour, the ``equal-mobility'' of the trans-
ported sediment grains and bedload rates close to those
predicted by the Schoklitsch (1962) formula. Succes-
sive ordinary events took the bedload rate back towards
the values recorded before 1994. Nevertheless, most
recent values (1997­1998) are slightly increased
owing to the excessive supply of fine sediment present
in the bed''.
The intervention described was, from a hydraulic
point of view, put to a rigorous test during the final
stages of construction. Between 14 and 18 October
1996, there was considerable heavy rainfall (in partic-
ular from 15­17 October when the Cruccolo rainfall
measuring station recorded around 170 mm of rain).
The maximum discharge caused by the event was
around 43 m3
s  1
. This was estimated by the clear
traces left by the flow upstream of the check dams at
the end of the torrent. The exact contribution of the
flood ( % 0.9 m3
s  1
km2
) indicated a return interval of
between 7 and 10 years. On October 7, 1998 another
flood, with a peak discharge of 52 m3
s  1
and a return
interval of 20­25 years occurred. On both these
occasions, the behaviour of both check dams strength-
ened with concrete and uncemented boulder steps was
satisfactory. Neither subsidence or damage to the check
dams nor bank erosion or fragmentation were evident.
From a hydraulic point of view, the investigations
carried out during the flood have also highlighted the
energy dissipating behaviour typical of step­pool
sequences, with a tendency to almost completely sub-
merge the vortex created by the step's sault into the
next pool downstream during slow flow (Fig. 7).
To evaluate both the stability and the hydraulic
behaviour of these artificial step­pool sequences a
set of experiments was performed at HR Wallingford
(UK). The experimental activity was performed using
the ``Tilting Sediment Duct'' facility, a flume which can
be tilted up to 40°. A model of the artificial steps in the
Maso di Spinelle torrent was built using Froude sim-
ilarity; with a length scale of 1/40 (time scale about 1/6
and discharge scale about 1/10,000). Calibration, per-
formed by simulating the flood hydrograph occurred in
1998, showed both the effectiveness of the model and
the importance of a good positioning of the boulders
while building the steps. Another hydrograph with a
peak discharge of 85 m3
s  1
(Tr = 100 years) has been
used to verified the stability of the artificial steps
(Giacometti, 2000). Moreover, the distance between
two subsequent steps has been changed trying to
validate the antidune and the maximum flow resistance
theory. Finally, the analysis of the failure of a step at
high discharge values pointed out the importance of toe
protection to ensure reliability and more durability to
the artificial structure. The results showed that toe
M.A. Lenzi / Geomorphology 45 (2002) 243­260256
failure is the main cause of collapse for a well-shaped
step (Giacometti, 2000). This hypothesis is also sup-
ported by the results of field research on the evolution
of natural step­pool sequences following a large flood,
carried out by Lenzi (2001) on the Rio Cordon, Italy.
Evaluation of the ecological effects of the artificial
boulder step structures was carried out with the RCE-2
model. To understand the differences in the ecological
assessment of the varieties of the protective structures
implemented on the Maso torrent, a comparison of
results was made between concrete check dams, arti-
ficial boulder steps, block stone ramp and a natural
step­pool sequences of the Maso di Calamento stream
(AKL-SASSM, 1999). The RCE-2 model is based on
the ``Riparian Channel and Environmental Inventory''
(RCE-1), a qualitative method of assessment of the
ecological and flood prevention function of water
courses. It was developed by Petersen (1990) and re-
Fig. 8. Total result of the ecological evaluation of the various structures.
Fig. 7. Uncemented boulder step­pool sequence during the October 7, 1998 flood (photograph courtesy of D. Sonda).
M.A. Lenzi / Geomorphology 45 (2002) 243­260 257
vised and adapted to the North Italian mountain envi-
ronment by Siligardi (1997). The survey shows clear
differences in the ecological assessment of the varieties
of protective structures building on the torrent Maso
(Fig. 8). Section 1 (Maso di Spinelle with artificial
boulder step and pool) reached a total of 203 or 207
points and, therefore, is just within class 2 quality;
Section 2 (Maso di Spinelle block stone ramp) reach
176 points and class 3 quality although there is the
possibility that it will improve when the vegetation of
the bank has grown again along the ramp; both the
sections with the concrete check dams (Section 3 of the
Maso di Spinelle and Section 4 of the Maso) also reach
class 3 quality; a more natural sequence of the Maso di
Calamento (Section 5) was investigated for compar-
ison. This natural step­pool sequence reach a score of
290 points and, therefore, reached class 1 quality.
Altogether, the ``boulder step and pool structure''
(Section 1) prove to be the best ecological and environ-
mental variation. Where channel stabilization struc-
tures are required, a stepped series of low boulder
step­pools structures is preferable to a single large
check dam. Multiple low structure designing according
to morphologic principles provide, as reported by
Haltiner et al. (1996), improved habitat for macro-
invertebrates, fish passage and also passage by other
wildlife which often use the channel as transportation
corridors between regional habitat areas.
6. Conclusions
Progress of construction technology in hydraulic
stream control sites has given a considerable impetus to
the creation of boulder check dams. These have been
built according to several design and construction
procedures. This category of works is suitable for
channel management in accordance with the usual
criteria of step­pool protection, offering greater natu-
ralness together with structural intervention. The differ-
ent designs for boulder check dams also allow the
planner to choose the kind of work appropriate to the
degree of protection needed. Nevertheless, geomor-
phologic criteria should also be considered, so that use
of boulder check dams, as an alternative to the use of
more traditional protection works and ones with a
greater impact on the course of water, does not merely
become of aesthetic and landscape benefit. Only by
identifying the physiographical units making up the
stream, analysing channel roughness, bed grain size
distribution and longitudinal profile, taking into con-
sideration geological conditioning, and linking channel
and catchment sediment system can one deduce the
informative elements on which to apply principles of
engineering geomorphology for managing channel
erosion and bedload. Further efforts have to be made
to understand fully the morphological individuality of
streams, even in situations that are not easily classified,
and to better predict channel dynamism. This will lead
to increasingly functional interventions that are more in
keeping with channel stability.
Acknowledgements
The SASSM-Servizio Azienda Speciale di Siste-
mazione Montana of Autonomous Province of Trento,
and particularly M. Cerato, are kindly acknowledged
for implementing on the Rio Maso di Spinelle, torrent
control techniques based on a realistic assessment of
morphologic criterion. Discussions and interactions
with Vincenzo D'Agostino and Diego Sonda were
instrumental in developing and clarifying thinking
regarding both natural step­pools evolution and
channel stabilization using boulder check dams. Part
of the research was carried out in the framework of the
``Damocles'' (Debrisfall Assessment in Mountain
Catchments for Local End-users) project, Contract n.
EVG1-CT-1999-00007, financed by the European
Union, DGXII. Field activities were supported by the
University of Padova, Research Project 2001­2002
``Hydraulic control and restoration of mountain
streams according to morphological and environmental
criteria''. I thank G.E. Grant (USDA Forest Service) for
critical reading of the text and for valuable comments
and suggestions. The valuable comments both of C.R.
Thorne and of an anonymous reviewer are also
acknowledged.
References
AKL (Amt der Kartner Landesregierung, Austria) and SASSM (Ser-
vizio Azienda Speciale di Sistemazione Montana, Provincia Au-
tonoma di Trento, Italia), 1999. Ecologically Orientated Flood
and Erosion Management in Alpine River Basin. Klagenfurt,
Trento, September, 68 pp.
M.A. Lenzi / Geomorphology 45 (2002) 243­260258
Abrahams, A.D., Li, G., Atkinson, J.F., 1995. Step­pool streams:
adjustment to maximum flow resistance. Water Resour. Res. 31,
2593­2602.
Ashida, K., Takahashi, T., Sawada, T., 1981. Processes of sediment
transport in mountain stream channels. IAHS Publ. 132, 166­
178.
Bathurst, J.C., 1993. Flow resistance through the channel network.
In: Beven, K., Kirkby, M.J. (Eds.), Channel Network Hydrol-
ogy. Wiley, Chichester, pp. 69­98.
Benini, G., 1990. Sistemazioni idraulico-forestali. UTET, 283.
Blizard, C.R., Wohl, E.E., 1998. Relationships between hydraulic
variables and bed load transport in a subalpine channel, Colo-
rado Rocky Mountains, USA. Geomorphology 22, 359­371.
Bravard, J.P., Landon, N., Peiry, J.L., Piégay, H., 1999. Principles of
engineering geomorphology for managing channel erosion and
bedload transport, examples from French rivers. Geomorphol-
ogy 31, 291­311.
Brookes, A., Douglas Shields Jr., F., 1996. River Channel Restora-
tion: Guiding Principles for Sustainable Projects. Wiley, Chiches-
ter, p. 433.
Brookes, A., Knight, S.S., Douglas Shields Jr., F., 1996. Habitat
enhancement. In: Brookes, A., Douglas Shields Jr., F. (Eds.),
River Channel Restoration: Guiding Principles for Sustainable
Projects. Wiley, Chichester, pp. 103­126.
Cerato, M., 1999. Contributo per una storia delle sistemazioni idrau-
liche e forestali in Valsugana. P.A.T., Azienda Speciale di Siste-
mazione Montana. Coordinamento Editoriale G. Goffer, 109 pp.
Chin, A., 1999. The morphologic structure of step­pools in moun-
tain streams. Geomorphology 27, 191­204.
D'Agostino, V., Lenzi, M.A., 1998. La massimizzazione della re-
sistenza al flusso nei torrenti con morfologia a step­pool. XXVI
Convegno di Idraulica e Costruzioni Idrauliche, Catania, Sept.
9­12, 1998, Italy, vol. 1, pp. 281­293.
D'Agostino, V., Lenzi, M.A., 1999. Bedload transport in the instru-
mented catchment of the Rio Cordon: Part 2. Analysis of the
bedload rate. Catena 36, 191­204.
D'Agostino, V., Cerato, M., Da Re, F., Lenzi, M.A., 1997. La
sistemazione idraulica dei torrenti con briglie in massi. Dendro-
natura 2, 39­52.
Della Giacoma, F., Florineth, F., Zolin, G., 1991. Opere di sistema-
zione e di regolazione dei corsi d'acqua, valutazione di impatto
ambientale. Prov. Auton. Trento, 1­23.
Flammer, G.H., Tullis, J.P., Mason, E.S., 1970. Free surface, veloc-
ity gradient flow past hemisphere. Proc. Am. Soc. Civ. Eng., J.
Hydraul. Div. 96, 1485­1502.
Giacometti, G., 2000. Local scour downstream bed sills and check
dams in high gradient streams. PhD Thesis, University of Pa-
dova, 190 pp. (in Italian).
Haltiner, J.P., Kondolf, G.M., Williams, F.B., 1996. Restoration
approaches in California. In: Brookes, A., Douglas Shields Jr.,
F. (Eds.), River Channel Restoration: Guiding Principles for
Sustainable Projects. Wiley, Chichester, pp. 291­329.
Kettl, W., 1994. Techniken der Wasser-und Gechiebebe-Wirstshaf-
tung als Beitrag zur Stabilisierung von Okosystemn. Wlidbach-
verbauung and Okologi 2, 58, Jahrgang, Heft, 126, pp. 41­45.
Knighton, D.A., 1998. Fluvial Forms and Processes. Edward Ar-
nold, London, 383 pp.
Kondolf, M., 1996. A cross section of stream channel restoration. J.
Soil Water Conserv. 51, 119­125.
Lenzi, M.A., 1992. Campionamento ed analisi di materiale d'alveo
con componenti grossolane. Il bacino attrezzato del Rio Cordon,
Regione Veneto, Quad. Ricerca, vol. 13, pp. 159­178.
Lenzi, M.A., 1999. Morfología y estabilidad de las secuencias en
escalones en los torrentes alpinos de elevada pendiente. Ing. Agua
6, 151­162.
Lenzi, A. (Ed.), 2000. Dynamics of Water and Sediments in Moun-
tain Basins. Quaderni di Idronomia Montana, vol. 20. Editoriale
Bios, Cosenza, Italy, 266 pp. (Special Issue).
Lenzi, A., 2001. Step­pool evolution in the Rio Cordon, North
Eastern Italy. Earth Surf. Processes Landforms 26, 991­1008.
Lenzi, M.A., D'Agostino, V., 1998. Dinamica dei torrenti con mor-
fologia a gradinata e interventi di sistemazione dell'alveo. Quad.
Idron. Montana 17, 37­56.
Lenzi, M.A., D'Agostino, V., 2000. Step­pool evolution in an
alpine torrent. In: Maione, U., Majione Lehto, B., Monti, R.
(Eds.), New Trends in Water and Environmental Engineering
for Safety and Life. A.A. Balkema, Rotterdam CD-Rom, 12 pp.
Lenzi, M.A., D'Agostino, V., Billi, P., 1999. Bedload transport in
the instrumented catchment of the Rio Cordon: Part 1. Analysis
of bedload records, conditions and threshold of bedload entrain-
ment. Catena 36, 171­190.
Lenzi, M.A., D'Agostino, V., Sonda, D., 2000. Ricostruzione mor-
fologica e recupero ambientale dei torrenti: criteri metodologici
ed esecutivi. Bios, Cosenza, Italy, 208 pp.
Montgomery, D.R., Buffington, J.M., 1997. Channel-reach mor-
phology in mountain drainage basins. Geol. Soc. Am. Bull.
109, 596­611.
Newson, M.D., 1995. Fluvial geomorphology and environmental
design. In: Gurnel, A.M., Petts, G.E. (Eds.), Changing River
Channels. Wiley, Chichester, pp. 413­432.
Nowell, A.R.M., Church, M., 1979. Turbulent flow in a depth-
limited boundary layer. J. Geophys. Res. 84, 4816­4824.
Petersen, R.C., 1990. The RCE: A Riparian, Channel, and Environ-
mental inventory for small streams in agricultural landscape.
Freshwater Biol., pp. 71­83.
Sawada, T., Assida, K., Takahashi, T., 1983. Relationship between
channel pattern and sediment transport in a step gravel bed river.
Z. Geomorphol., Suppl. 46, 55­66.
Schmidt, K.H., Ergenzinger, P., 1992. Bedload entrainment, travel
lengths, step lengths, rest periods--studied with passive (iron,
magnetic) and active (radio) tracer techniques. Earth Surf. Pro-
cesses Landforms 17, 147­165.
Schocklitsch, A., 1962. Handbuch des Wasserbaues, 3rd edn.
Springer, Viena.
Schweitzer, R., 1995. Metodi di ingegneria naturalistica lungo i
corsi d'acqua. In: Maione, U., Brath, A. (Eds.), Moderni Criteri
Di Sistemazione Degli Alvei Fluviali. Bios, Cosenza, pp. 319­
326.
Sear, D.A., Newson, M.D., Brookes, A., 1995. Sediment-related
river maintenance: the role of fluvial morphology. Earth Surf.
Processes Landforms 20, 629­647.
Siligardi, M., 1997. Ecologia del paesaggio e sistemi fluviali. Eco-
logia del Paesaggio, a cura di V. Ingegnoli. Citta` Studi, Milano,
187 pp.
M.A. Lenzi / Geomorphology 45 (2002) 243­260 259
Sonda, D., 1998. Applicazione di criteri morfologici nella sistema-
zione dei torrenti: il caso della Val Campelle (TN). Tesi di Laurea
in Scienze Forestali ed Ambientali Universita` di Padova, 406 pp.
Sternberg, H., 1875. Untersunchungen uber das langen und Quer-
profil geschiebefuhrender flusse. Zeitung Bauvesen, 25 pp.
Thorne, C.R., Newson, M.D., Brookes, A., 1997. Applied Fluvial
Morphology for River Engineering and Management. Wiley,
Chichester, 376 pp.
Ufficio Federale dell'Economia delle Acque, UFEA, 1995. Esi-
genze in materia di protezione contro le piene'95. Berna, 6 pp.
Warburton, J., 1992. Observations of bed load transport and channel
bed changes in a proglacial mountain stream. Arctic Alpine Res.
24, 195­203.
Whittaker, J.G., 1987. Modelling bedload transport in step mountain
stream. IAHS Publ. 165, 319­332.
Wohl, E.E., 2000. Mountain rivers. Water Resour. Monogr. 14, 320
Washington, DC, USA.
Wohl, E.E., Thompson, D.M., 2000. Velocity characteristics along a
small step­pool channel. Earth Surf. Processes Landforms 25,
353­367.
Wolman, M.G., 1954. A method of sampling coarse river-bed ma-
terial. Am. Geophys. Union Trans. 35, 951­956.
M.A. Lenzi / Geomorphology 45 (2002) 243­260260