You are on page 1of 25

Rivers of dreams: on the gulf between

theoretical and practical aspects of an


upland river restoration
Blackwell Publishing, Ltd.

Adrian McDonald*, Stuart N Lane, Nick E Haycock


and E A Chalk
This paper is concerned with development and application of a conceptual model of
the river restoration process and, through a specific case study, a demonstration of the
challenges that arise when combining ecological and geomorphological principles with
social and economic issues. Thus, it demonstrates the difference between the ideal
goals of river restoration and the practical determinants of the nature of a river
restoration that derive from individual and community goals. It begins by exploring
theoretically what a river restoration project might involve. It then introduces the
options developed for restoration of a particular case study river and the associated
decisionmaking process. As a result of strong community involvement, institutional
and financial regulations, the location of the restoration site in a National Park, and
the influence of governmental and non-governmental organizations, the actual
decisionmaking process resulted in a practical river restoration that departed
significantly from the idealized goals of a restoration defined in purely scientific terms.
It is argued that these practical limitations are likely to remain the dominant influence
upon the nature and scope of river restoration projects.
key words

river restoration

rehabilitation

gravel-bed rivers

decisionmaking

*School of Geography, University of Leeds, Leeds LS2 9JT


Department of Geography, University of Durham, Science Laboratories, South Road, Durham DH1 3LE
email: s.n.lane@durham.ac.uk
Haycock Associates Ltd, St Albans, Hertfordshire AL4 0DE
Environment Agency, North-East Region, Dales Area, Clifton Moor, York YO30 4GZ

revised manuscript received 4 June 2004

Introduction
The last 20 years have seen a fundamental shift in
the purpose and approach to river channel management. First, the purpose of management has shifted
from simple anthropocentric, utilitarian needs associated with river channel engineering for flood and
erosion/sedimentation control towards incorporation of a range of goals traditionally optional, but
now required (e.g. ecological concerns). As this
purpose has changed, so the number of interest
groups associated with the management process
has increased (Sear et al. 2000). Second, there has
been recognition that effective river channel management must be holistic at the catchment scale

(e.g. Newson 1997; Clark 2002; Everard and Powell


2002). For instance, in England and Wales, the
Environment Agency of England and Wales (the
Environment Agency) has introduced Catchment
Abstraction Management Strategies that explicitly
recognize that the river has multiple users and that
the needs of all users must be addressed to identify
optimal solutions for river management. This will
be greatly impacted upon by the European Water
Framework Directive (WFD) (European Commission 2000) requirement to produce river basin management plans for all major rivers in all member
and candidate countries at least nine years after the
date the treaty comes into force. This policy change
recognizes that one of the clearly defined user

Trans Inst Br Geogr NS 29 257281 2004


ISSN 0020 -2754 Royal Geographical Society (with The Institute of British Geographers) 2004

258

groups is the rivers fauna and flora, introducing a


strongly environmental requirement. This is also
increasingly legislated [e.g. under the requirement
of the European Union Habitats Directive (European Commission 1992) and the UKs Conservation
Regulations (Conservation (Natural Habitats, &c.)
Regulations 1994) to protect sites of nature
conservation interest].
It is hardly surprising, given these changes, that
there has been progressive development of river
restoration projects. Whilst river restoration is one
example of a broader restoration movement, it
has received much less attention as a concept as
compared with other ecosystems (e.g. Hobbs and
Norton 1996). The major exceptions to this fall into
three groups. The first is a biophysical view that
emphasizes the need to think critically about our
conception of how rivers work (e.g. Sear 1994;
Stanford et al. 1996; Ward et al. 2001). The second
recognizes that restoration is not just about the science of the biophysical environment, but the perspectives of non-scientists (e.g. Rhoads and
Herricks 1996). The third follows on from these
two (e.g. Sear 1994; Eden et al. 1999 2000; Rhoads et
al. 1999; Tunstall et al. 1999 2000) and shows that
interaction between scientific definitions of restoration, and the goals of society as a whole and individuals involved in the restoration process, creates
a very different form of river restoration to that
which might be expected from a biophysical perspective alone. For example, Eden et al. have demonstrated how UK river restoration projects
represent hybrids: unique combinations of nature
and culture, that are contingently situated and
with very little resemblance to either aspects of
nature or of culture when the latter are taken in
isolation. Rhoads et al. (1999) argue that this type of
restoration is better viewed as naturalization: it
generally results in improvement of the biophysical environment, when defined from a narrow scientific perspective; but other processes mean that
restoration does not and cannot proceed along biophysical principles alone.
Hobbs and Harris (2001) note that restoration
has been practised widely without reference to any
background conceptual framework and that effective restoration requires ongoing dialogue between
the general concepts associated with system behaviour and the specific practical restoration activities
of immediate concern at a particular location. Van
Diggelen et al. (2001) argue that this arises because
a rapid response is required (e.g. due to opportunistic

Adrian McDonald et al.

funding, a local river management crisis) and leads


to: (a) restoration becoming more of an art than
a science; (b) a strong emphasis upon templates
for restoration, as per the Manual of river restoration
techniques (River Restoration Centre 2002) and
Stream corridor restoration (Federal Interagency Stream
Restoration Working Group 1998), and reflected
in UK (http://www.rrc.org.uk) and European
(http://www.minvenw.nl/rws/riza/home/ecrr/)
river restoration sites; and (c) a reliance upon
experienced practitioners to guide the restoration process. From both scientific and cultural
(Pfadenhauer 2001) perspectives, there may be
good reasons why restoration projects are site specific and not always seen as effective when judged
in conventional terms (see below). Indeed, many of
these issues are reflected in the common distinction
(e.g. Eden et al. 1999 2000; van Diggelen et al. 2001)
between restoration, a more radical attempt to recreate a prior system, and rehabilitation, the recreation of certain ecosystem functions as a result of
natural and social and economic limits upon a pure
restoration process, often connected with present
catchment conditions, as defined in the broadest
sense (i.e. geomorphological, ecological, climatic,
social, economic etc.).
In this paper we seek to develop some of the
conceptual ideas emerging in the restoration literature to develop a conceptual model for approaching river restoration. This is more developed than
the leitbild or guiding image for river restoration
(Kern 1992) as it explicitly recognizes the influence
of social and economic processes in influencing the
restoration process. We believe that, whilst it is a
somewhat mechanical view of what restoration
involves, it provides the starting point from which
to understand how the restoration process works
in practice: there have to be guidelines for doing
restoration, and it is better that these are recognized critically and explicitly, including social and
economic influences, rather than left as implicit to
the management process. This allows us to demonstrate the strong contextual influence that comes
from socio-economic and cultural processes, and
which results in the creation of projects that are
best described as rehabilitation, due to the complex
interactions between the river corridor and the
communities within them. Rehabilitation of a reach
of the River Wharfe, North Yorkshire, is then used
to show how this conceptual model operates in
practice. This allows us to achieve a second aim: a
demonstration of the difficulties associated with

Rivers of dreams

259

achieving river restoration in the framework of


integrated catchment management.

Conceptualizing river restoration


The ecological literature has witnessed considerable emphasis upon conceptualization of the
restoration process (e.g. Hobbs and Norton 1996;
Higgs 1997; Hobbs and Harris 2001). Much of this
has focused around the need to identify restoration
goals (e.g. Higgs 1997; Hobbs and Harris 2001)
based upon an appropriate ecosystem response
model (Hobbs and Harris 2001). Figure 1 shows a
generalized model for a restoration project, modified from Hobbs and Harris (2001). It has a number
of core aspects: (1) a system response model; and
(2) an associated knowledge base. These interact
with: (3) restoration goals to produce; (4) restoration options, implemented though policy instruments. Monitoring and evaluation, in relation to
project goals, are used to adapt the system
response model as further understanding of the
system emerges, but also to feed back into individual and community goals in recognition that
experience of an implemented project will change
these goals. This is an ideal model, and we reflect
upon its application to the River Wharfe later in

Figure 1

the paper. It follows directly from the leitbild


model (Kern 1992) for an ideal river restoration
project which recognizes: (a) natural stream properties; (b) the influence that comes from irreversible
changes in abiotic and biotic factors; and (c) the
need to recognize cultural and historical influences
upon restoration options. Kern advocates an holistic
approach to restoration, based upon a building
block, the leitbild. We now develop this in relation
to natural stream properties and, more importantly,
bring social, economic and political processes explicitly into the restoration process.

The system response model


A system response model (e.g. Stanford et al. 1996;
Ward et al. 2001) is a key aspect of river restoration:
a clear statement of what system behaviour might
be expected in relation to a particular study site.
Hobbs and Harris (2001) suggest it might be simple
or complex, quantitative or qualitative, but that, at
its core, it captures the essence of the system being
studied. In the case of river restoration this would
include the dynamics of the specific river corridor
under consideration and wider catchment-scale
processes that influence the dynamics of that reach.
Thus, the catchment baseline surveys, geomorphological audits and dynamic assessments suggested

Generalized model for a restoration project modified from Hobbs and Harris (2001)

260
Table I

Adrian McDonald et al.


Four aspects of a river response model

Structure (or Form)


Cross-section and planform morphology
Bedforms
Perimeter sedimentology
Habitat
Diversity
River channel pattern
Within-reach cross-sectional variability
Sediment sorting
Downstream fining

by the geomorphological community as part of


river management (e.g. Downs and Thorne 1996;
Thorne et al. 1996; Environment Agency 1998)
provide a required context prior to a restoration
activity.
The Society for Ecological Restoration (1990,
cited in Higgs 1997) defines the dimensions of possible behaviour as the: structure, function, diversity and dynamics of the specified (eco)system.
Similar definitions come from: (a) Hobbs and Norton (1996) composition, structure, function, heterogeneity and resilience; and (b) Higgs (1997)
fidelity, based upon structural/compositional
replication, functional success and durability.
Rather than defining and evaluating these terms,
we use the Society for Ecological Restoration (1990)
dimensions and Table I illustrates them in relation
to river channels. It recognizes the importance of
interaction between form and process (structure
and function; Ehrenfeld 2001; Ward et al. 2001) in
creating both spatial variation in river channel
structure (diversity) and changes in river channel
structure through time (dynamics), which in turn
control formprocess interactions. All systems
should be spatially and temporally variable as a
result of formprocess interactions: both diversity
(e.g. Pickett and White 1985) and dynamics (e.g.
Resh et al. 1988; Sparks et al. 1990; Hildrew and
Giller 1994; Kondolf 1998) should matter. Indeed,
the ecological literature has noted that much restoration tends to draw too strongly upon outdated
idealizations of ecosystem stability and the ability
of an ecosystem to return to a particular equilibrium state following disturbance (Adams 1997;
Hobbs and Harris 2001). This may also hold for
river systems (Newson and Sear 1997). For
instance, Ward et al. (2001) argue that the failure to
recognize the diversity and dynamism of natural
river corridors has hindered conceptual advances

Function (or Process)


Flow acceleration and deceleration
Secondary circulation
Turbulence
Sediment entrainment, transport and deposition
Dynamics
Bedform migration
Bank erosion
Sedimentation
Channel migration and channel change

in river ecology and hence river restoration. Similarly, geomorphological diversity provides the
footprint within which ecological diversity is
found. Recognition of the importance of diversity
and dynamics causes us to introduce notions of
durability and resilience, and hence the importance
of considering sustainability (e.g. Eden et al. 1999;
van Diggelen et al. 2001; Whalen et al. 2002) within
the restoration process.
In summary, a system response model must be
both analytical and normative: it needs to combine
an appreciation of the way a river is working with
consideration of how it could work, in order to
identify geomorphologically and ecologically sustainable management restoration options.

Knowledge
Figure 1 identifies both traditional scientific sources
of understanding and local community knowledge
as drivers of a system response model. The
involvement of experts in the river restoration
process emphasizes why this fusion of knowledge
is important. Restoration is commonly a locationspecific activity over a pre-defined time horizon.
The involvement of river scientists (fluvial geomorphologists, aquatic ecologists, engineers,
hydrologists) is clearly necessary as a result of the
basic generic knowledge that has been accrued in
river science over the last 50 years. It is hard to
develop the system response model without a basic
understanding of how different aspects of the river
system work in practice (Wissmar and Beschta
1998). However, this is rarely sufficient. From a
scientific viewpoint, an ideal approach to river
restoration (Wissmar and Beschta 1998) would
involve a long-term investigation of how the target
river reach works, including catchment-scale and
local-scale audit (e.g. Downs and Thorne 1996;
Thorne et al. 1996) to establish basic context, plus

Rivers of dreams

more detailed measurement of formprocess


interactions through time. Time is commonly not
available, and so other methods (scaled laboratory
analysis, numerical modelling or study of
neighbouring river reaches that have not been
engineered) might be required to provide the
necessary understanding. However, the number of
restoration projects being undertaken, coupled to
resource limitation, means that this sort of in-depth
knowledge base is rarely available (e.g. Scholz et al.
2002). Similarly, and certainly in a UK context, we
have very little history of long-term measurement
of river behaviour and funding constraints mean
that this is likely to remain the case. For this
reason, generic scientific knowledge must be fused
with local knowledge (accounts of past events,
descriptions of previous river location, estimates of
bank erosion rates) to develop a reasonable
understanding of the place-specific behaviour of
the system being studied (Rhoads et al. 1999). The
local community possesses a vitally important
knowledge base (van Diggelen et al. 2001; Scholz
et al. 2002) which should not be subsumed to
some privileged scientific knowledge (Rhoads et al.
1999). Similarly, local institutions may add value to
scientific and community knowledge bases.

Links between a system response model and


restoration options
A set of restoration options follows from a system
response model. Following Palmer et al. (1997),
there are three ways in which we may approach
river restoration in relation to structure, function,
diversity and dynamics. Palmer et al. (1997) label
the first as field of dreams, drawing heavily upon
the traditional engineering approach to river
management: form (or structure) is engineered to
look like what you want it to look like in the hope
that this will create the processes necessary to maintain it. It is characterized by creation of particular
natural morphological attributes (e.g. meanders, a
riffle-pool sequence, a mid-channel bar or island)
and detailed morphological controls (e.g. bendway
weirs to reduce outer bank erosion, buried weirs to
encourage riffle formation, deflectors to break up
secondary circulation). It assumes a characteristic
form, which will produce processes (e.g. secondary
circulation). To maintain the restored form, some
of these processes will need to be managed (e.g.
breaking up secondary circulation to prevent overdeepening of the bed and potential bank erosion).
Traditionally, it is well-supported from notions of

261

equilibrium and regime analysis in river systems


(e.g. Hey and Thorne 1986) and the assumption
that a river system adjusts its morphology to the
water and sediment delivered to it. By identifying
the correct regime (e.g. width, depth, slope, channel
pattern, bed material, cross-section area etc.) in
relation to the dominant processes, reinstating that
regime will result in a sustainable river restoration.
It assumes an historically identifiable reference
state to which the system can be restored, and that
appropriate river engineering can be used to
recreate this state.
The field of dreams approach has a number of
problems. Hobbs and Harris (2001) argue that it is
backward looking, seeking to recreate ecosystems
with properties that were characteristic of the system at some time in the past. Indeed, four problems arise from using an historically defined
endpoint. The first is technical (e.g. Hobbs and
Harris 2001). We have to assume that a single state
can be identified from a complex landform history.
Unfortunately, our understanding of the historical
dimension of a river is often inadequate for the
purpose of reconstruction as there are problems of:
(1) separating seasonal, inter-annual and periodic
variation from a given depositional record; (2)
understanding spatial context and the role of
upstream and downstream influences on a given
river reach; and (3) differences in preservation
potential so that some fluvial features are more
well-preserved than others. Thus, there can be considerable uncertainty as to exactly what the historical structure or form should be (van Diggelen et al.
2001; Ward et al. 2001). These technical difficulties
are compounded by a need to decide which point
in history is the endpoint that is to be restored
given long-term changes in river morphology in
relation to climatic controls (e.g. Passmore and
Macklin 2000) and the fact that a river may be out
of equilibrium in the sense that it is adjusting to
drivers that are not locally or currently active
(Newson 2002). It is quite possible that an endpoint
could be the river associated with present or even
projected catchment characteristics, although if we
assume that most river systems are still adjusting
to past and present climatic changes, difficulties
emerge in predicting exactly the endpoint that
might be expected.
Second, events since the historical end point may
prevent restoration to the a priori system (Stanford
et al. 1996; van Diggelen et al. 2001). In a UK context, environmental variability, plus the long history

262

of cultural modification of the landscape (Sear


et al. 2000) make this likely to be the norm. Third,
especially in the ecological literature (e.g. Ehrenfeld 2001), the extent to which unique structure
function assemblages exist has been questioned.
Much of this relates to the level of complexity of
ecosystems. In a river channel system, this complexity will be reduced (there will be fewer structures or forms), but even then there may be
different river channel states for a given process
combination. Graf (1988) describes a cusp catastrophe model where, for a given combination of
stream power and stream resistance, the river may
be either braided or meandering (i.e. have very different values of sinuosity) according to whether or
not the control variables (power and resistance) are
increasing or decreasing through time. This
increases the probability that a given form or structure is not sustainable because the precise combination of process sequencing required to maintain
it is not present. Finally, it assumes that the system
has some sort of equilibrium that is the target of
the restoration activity. Whilst research in fluvial
geomorphology traditionally has accepted this
assumption (e.g. Blench 1969), it can only hold if
the system is closed and its entropy can only
increase. In an open system, external influences
mean that there can be no equilibrium state to
which a system evolves, and a range of system
states is possible: there is no stable structure to be
concerned about; and disturbance is the rule rather
than the exception. Indeed, disturbance may be
necessary if a system is to be self-sustaining. This is
now well-recognized in ecology more generally
(e.g. Pickett and White 1985; Pahl-Wostl 1995;
Adams 1997). The irony of some restoration
projects is that once a given river channel state has
been created, additional channel modifications (e.g.
outer bank deflectors) have to be introduced to
eliminate the very processes (e.g. secondary circulation) that the channel will create and that will
lead to channel change. This is well-illustrated by
Shields: The risk of a restored channel being
damaged or destroyed by erosion or deposition
can be reduced if economic considerations permit
installation of control measures (1996, 65).
The second approach is labelled the system
function approach (e.g. Palmer et al. 1997), where
the initial conditions required to achieve a projects
overall goals are identified and ameliorated. It
seeks to address the criticism of fields of dreams
approaches (e.g. Boon et al. 1992; Kondolf 1998;

Adrian McDonald et al.

Ward et al. 2001; Everard and Powell 2002) as summarized by Christensen where goals are often cast
in terms of goods and services (e.g. natural
resources, recreation, water quality etc.) rather
than the (eco)system processes necessary to deliver
those goods and services (1997, 168). Thus, it is a
process-oriented approach and recognizes that
restoration strategies can sometimes fail, either
because more fundamental root causes have not
been addressed (e.g. Kondolf 1998) or because of
misunderstandings of the nature of system
response (e.g. Ehrenfeld 2001; Hobbs and Harris
2001; van Diggelen et al. 2001). Three problems
emerge: (1) where individual or community goals
are at odds with the structure (i.e. river channel
form) that comes from restoring system function;
(2) where cultural or environmental limitations to
process amelioration exist (e.g. where there are
strong upstream land use influences, such as sediment delivery or water quality, that cannot be
managed for the purposes of restoration; van
Diggelen et al. 2001); and (3) there can be considerable uncertainty in (van Diggelen et al. 2001) and
time required before the system responds.
The third approach (Palmer et al. 1997) is called
the keystone method. The keystone method seeks
to recognize that restoration must identify and
incorporate crucial components of both form and
function, but that there must be an open-ended
view of what the final restoration project achieves:
both diversity and dynamics must be created by
the interaction between form and process. The
keystone approach requires identification of key
aspects of form and process within river systems
(e.g. Brookes and Sear 1996) and their initial restoration. For instance, the riffle-pool sequence has
been shown to be a fundamental aspect of river
channels with mobile beds (e.g. Newson and Sear
1997) and this can provide a basic keystone around
which a river restoration might be developed
(Frissell et al. 1986). Similarly, there are approximate
rules that define expected channel pattern and
morphology given channel width, discharge,
perimeter sedimentology and slope (e.g. Leopold
and Wolman 1957; Schumm 1960 1968; Carlston
1965; Ackers and Charlton 1970). All of these relationships recognize the importance of both current
catchment state (as reflected in terms of discharge
parameters) and reach location (as reflected in local
slope and sedimentology). These might need to be
extended (e.g. Brierley et al. 2002) to account for
river channel variability, both current (natural or

Rivers of dreams

human-induced) and in the future (due to land use


or climate change), although the latter may have
substantial uncertainty. The keystones might also
include what Brierley et al. (2002) call biophysical
linkages: i.e. key ecosystem attributes that also
need to be restored in order to achieve an effective
long-term restoration. Stock exclusion to encourage
the development of natural vegetation or willow
planting are both examples of active intervention
that has at least two purposes: (1) assistance with
creation of an expected formprocess interaction in
geomorphological terms; and (2) recognition of the
need to include keystone members that are ecological as well as geomorphological. These are not general requirements of a restoration project but rather
attributes that are associated with a specific restoration context.
Following from the above, the keystone approach
may appear to be just like another field of dreams
approach based upon the expected river regime.
Indeed, too rigid an adoption of certain keystones
(e.g. riffle pool sequences with a set spacing) may
result in a restoration that becomes a field of
dreams approach. This emphasizes that these
three approaches are overlapping. However, they
have some distinctiveness: the keystone (and hence
the restoration) is assumed to provide the template
from which the river creates its own formprocess
interaction through space and time ( Ward et al. 2001):
Ward et al. describe how the aim of river restoration in situations where islands might be expected
to form should not be to recreate an island but
rather: (1) the reconstruction of the ecological and
geomorphological processes that enable the river to
construct its own islands (e.g. through sufficient
sediment supply in the correct size range); and
(2) identification of the necessary morphological
changes to allow the river to develop islands (in
this case an unconstrained channel). The restoration can be either finite (where the river is left to its
own devices) or adaptive (e.g. Kondolf 1998): if
types of diversity or dynamics emerge that are not
wanted, further intervention may be needed (e.g.
setting back of flood defences as a river bend
migrates; allowing a river to migrate within a zone
of stock exclusion or riparian set aside). However,
a crucial issue in a keystone approach relates to the
spatial scale of restoration. The sustainability of a
keystone cannot be seen as independent from
externally imposed process drivers that might
undermine survival of the keystone. For instance,
catchment-scale changes in fine sediment delivery

263

may mean that a riffle-pool sequence is rapidly infilled (Sear 1994). Similarly, if unacceptable levels
of river migration emerge, the approach may
appear to be overtly idealistic. At its core, it is
accepting the future uncertainty and unpredictability of river behaviour, however well the river is
known, such as by giving the river space (a river
corridor) to adapt to future climatic and catchmentscale changes. The keystone approach helps to achieve
a set of formprocess interactions rather than simply
to restore a particular form or forms.

Individual and community goals


Incorporation of individual and community goals
into the decisionmaking process (Figure 1) is
crucial to the restoration process (Boon 1998).
Pfadenhauer (2001) identifies four reasons why
broader community goals have an important
influence: (1) goals are strongly influenced by
public opinion and open to dispute through the
decisionmaking processes that result in a particular
project being adopted (Swart et al. 2001); (2)
emotional and aesthetic factors may be especially
important, such that identifying with a
component of a restoration project is necessary;
(3) the popular acceptance of goals is required for
a project to be socially relevant ( Wissmar and
Beschta 1998) and for public involvement
(Goodwin 1998); and (4) the value of a unit of the
landscape will be referenced to changing values
such that even highly degraded systems (e.g. a
canalized river in an urban environment) can
acquire value (e.g. Adams 1997). Thus, restoration
plays an important social role in mediating our
relationship with valued places (e.g. Jordan 1994;
Adams 1996; Higgs 1997; Eden et al. 2000). This
represents a move on from a purificationist view
(Eden et al. 2000) of river restoration to one that is
more community-centred. Restoration goals are not
only grounded in a system response model and
how nature works but also what the restoration
means to the community that defines and redefines
it (Eden et al. 2000) and the need to see river
restoration as sustainable in more than just a
physical or biological sense is noted (e.g. Eden et al.
2000). This is necessary. First, restoration is partly
an economic process (Holl and Howarth 2000) as a
result of the substantial costs that might be
involved, the spatial extent over which costs may
be felt if landscape level changes are required (e.g.
van Diggelen 2001; Ehrenfeld 2001; Pfadenhauer
2001), and the potentially long period of time over

264

which activities may need to be sustained (van


Diggelen 2001). Second, research suggests that provided appropriate mechanisms for community
involvement can be identified, people can develop
enthusiasm for restoration (Tunstall et al. 2000).
The River Skerne demonstration project in the UK
has been recognized for its effective engagement with
the community and this seems to be associated
with use of an effective community liaison officer.
The idea that community desires matter is wellrecognized by the restoration ecology community
(e.g. Higgs 1997; Hobbs and Harris 2001; Pfadenhauer 2001; Clark 2002) and it has even been
argued (e.g. Higgs 1997) that what the restoration
achieves in relation to the ecosystem being considered matters less than the process by which restoration is undertaken in relation to those people
with relevant interests in the site being restored
(Wyant et al. 1995; Hobbs and Norton 1996; Pfadenhauer 2001): the worth of restoration is adjudicated on historical, social, political, cultural,
aesthetic, and moral contexts (Higgs 1997, 339).

Links between individual and community goals


and restoration options
Given that the process of restoration can be just as
important as the restoration itself, thought needs to
be given as to how to identify and to evaluate
restoration options. There is a considerable body of
literature that relates to environmental decisionmaking processes (see Clark 2002). Much of this
stems from the business world where there is an
agreed primary objective, say income, which has to
be maximized or optimized to a maximum relative to
risk. Such a simple scenario on which to base decisionmaking is unlikely to be found in the management
of river systems. A more complex decision context
exists here (Clark 2002), because there are likely to
be many stakeholders (e.g. ramblers, anglers, farmers,
boaters, conservationists) all with valid, but often
contradictory, objectives or expectations: a compromise has to be sought that satisfies most stakeholders and achieves a lowest common denominator
of acceptance by all. Although commonly used
(e.g. National Rivers Authority 1995), traditional
approaches using costbenefit analysis and risk
analysis have considerable drawbacks in the river
management context, especially in relation to longterm sustainability (Clark 2002). In costbenefit
analysis, benefits to one group of stakeholders may
not be given the same value, may be a real or
perceived disbenefit, or may even be a perceived or

Adrian McDonald et al.

real cost to another stakeholder group. Overall


benefit is very hard to categorize, far less quantify
(Bennett 2002). Even if benefits could be agreed,
quantifying them is contentious and complex
(Bennett 2002). Risk analysis suffers some of the
same problems and it is recognized (e.g. OBrien
2000) that risk assessment stresses the negative.
OBrien also argues that risk analysis, costbenefit
analysis and the various hybrids of benefitrisk
assessment are extremely costly and so managers
tend to commission analysis of only some preferred
options from the technical providers. In effect then,
a part of the decisionmaking process has been
removed from stakeholders or managers. Some
(e.g. OBrien 2000) have identified the advantages
of options assessment, in which all available
options are presented, including those that are
in disfavour by some stakeholders. Incorporation
of these sorts of approaches into definition of
restoration options is crucial. However, it almost
certainly means that there will be contradictions
emerging in relation to system response models,
regardless of the level of community involvement
in defining the associated knowledge base.

Monitoring and evaluation


Post project appraisal is now an established component of most river management activities (Boon
1998), even if it is not always adopted (Brookes
1996; Kondolf 1998). This is formalized into both
geomorphological activity, but also socio-economic
assessments such as costbenefit analysis (e.g.
National Rivers Authority 1995). A river restoration project could be evaluated with respect to:
(1) whether or not the river resembles (and stays
resembling) what was intended (a field of dreams
evaluation); (2) whether or not the river is behaving in the intended way (a system function
evaluation); or (3) whether or not the essence of the
restoration is leading to development of sustainable formprocess interactions (a keystone evaluation). This is the sense in which establishing
the restoration goals is crucial, as it will shape
expectations in relation to monitoring. However,
involvement of individual and community goals
in the restoration process means that there may be
significant departure from what science says is the
best means of restoration, and this allows for an
adopted restoration option to become unsustainable
in relation to a chosen system response model if
it is defined in a narrowly scientific way. Thus,
monitoring and evaluation is not simply a technical

Rivers of dreams

activity (Whalen et al. 2002): the views of the local


community, stakeholders etc. should be incorporated into the evaluation process (e.g. Eden et al.
1999 2000; Tunstall et al. 1999 2000), including both
the scheme itself and the process by which the
scheme was conceived and implemented. This ties
directly with the social role and economic implications of a restoration project (cf. Higgs 1997),
including ongoing funding streams and community acceptance of the project.

Adaptation
There is a tendency to view a management activity
as a singular act. In practice, river restoration has
to involve adaptation (Kondolf 1998; Clark 2002).
Eden et al. (2000) call this translation, occurring in
response to observed river response as well as
economic changes. There are good reasons for
building adaptation into the process. First, it has
considerable support in the presence of uncertainties (e.g. Lindblom 1959 1979; Clark 2002). The
option adopted is information-based, in that part of
the strategy is to develop the information and
understanding required to support further decisions. This recognizes that not all uncertainty is
determinate in river basin science as a result of
unpredictability in the socio-economic environment that partly influences catchment-scale hydrological processes (Clark 2002). Even if a rivers
response to restoration could be predicted accurately and precisely, the factors that might influence
river response may not be so. Second, the need to
adapt follows from the way in which community
and individual goals change as a result of the
experience that comes from adopting a particular
restoration option, and the associated monitoring and
evaluation. Third, adaptation may also be the only
means of achieving a more radical river restoration:
restoration of an unconstrained, wandering gravelbed river, in a region of high agricultural or
landscape value may not be acceptable; allowing
a river to migrate in a managed way, along with
other components that make up the landscape (e.g.
footpaths, river crossings etc.), may achieve the
same end but through a more acceptable means.

Identification and implementation of


rehabilitation options for the River
Wharfe
The above description identified conceptually the
range of activities that might comprise an effective

265

river restoration, based upon the combined input


from a system response model, which embodies an
understanding of the way the river works, and
institutional and community goals. This section
illustrates these themes and issues in relation to a
specific project.

Background to the restoration site


The River Wharfe is an upland gravel-bed river
that flows through a formerly glaciated valley in
the Yorkshire Dales National Park, North Yorkshire, UK (Figure 2a). It is sinuous, with riffle-pool
sequences in both meandering and relatively
straight reaches. From Hubberholme to as far downstream as Grassington, the river is depositional,
although observations suggest spatial variation in
deposition rates (ERM 1983). It has a long-term
history of both sedimentation and incision in
response to changing catchment hydrology and
sediment delivery (Howard et al. 2000), reflected in
evidence of channel migration across the floodplain (Figure 2b). There is a marked change in river
behaviour downstream from Hubberholme because
of a sudden reduction in channel gradient as the
river flows from out of a confined valley into a
wider valley. Sediment that is relatively rapidly
transferred through the limestone bedrock reach
upstream begins to move more slowly and increasingly in the form of self-regulating sediment
waves, leading to alternating phases of erosion and
deposition (cf. Passmore and Macklin 2000).
Locations of deposition affect adjacent riparian
land in two ways. First, they lead to bank erosion:
the formation of alternate, point and mid-channel
gravel bars leads to bank erosion on the opposite
(alternate/point bars) or both (mid-channel bars)
sides of the river. This is a natural characteristic of
a gravelly river in the presence of high rates of sediment delivery. Second, if delivery rates are higher
and the net result is reach-scale bed aggradation,
the magnitude and frequency of local floodplain
inundation increases, with greater fine sediment
transfers to the floodplain and the probability of
coarse sediment transfer during extreme flow events.
This results in an active process of floodplain construction and possible avulsion, as is reflected in
the records of sedimentation and incision available
for the valley (Howard et al. 2000) and the palaeochannels on the floodplain (Figure 2b).
The last few hundred years have seen active
management of this system, with the river being
constrained to the same channel, and the channel

266

Adrian McDonald et al.

Figure 2 The location of the rehabilitation site (2a)


and a Digital Elevation Model showing floodplain
topography (2b)

Rivers of dreams

to the same position. This has involved stone walling the banks of the river with local material,
removing accumulated gravels and managing the
trees along the banks. When this river training
started is not known exactly, but similar river
works in the Lake District are associated with
fourteenth- and fifteenth-century farming practices
and the conversion of floodplains to meadows to
support growing sheep numbers. The last 50 years
have seen three major types of changes. First, there
have been profound changes in the institutional
responsibility for planning and preservation of the
landscape associated with designation of the area
as a National Park. Many of the features of the
river that were introduced traditionally to manage
the environment (e.g. river training and stabilization activities using stone walling) are now part of
the cultural heritage of this area. Second, the traditional method for gravel management, based upon
localized extraction for construction, may have
ended. Third, it has been shown (e.g. Hey and
Winterbottom 1990; Longfield and Macklin 1999)
that catchment hydrological change (increasing frequency of high magnitude flow events) is occurring as a result of the interacting effects of land use
change and climate change, including upland
drainage, possible afforestation impacts and, as yet
only hypothesized, a marked increase in stocking
densities (Sansom 1996).
These changes have been accompanied by continued accumulation of gravels within the river,
exacerbating flood risk significantly (Hey and
Winterbottom 1990). The recent manifestation of this
historical river training practice was the Buckden
Scheme in the mid-1980s: river gravels were removed;
new flood banks and river-bank walling (hard
revetment) were installed; and floodplain drainage
was improved. Part of the Buckden Scheme was the
creation of the Buckden Gravel Trap (Figure 2a),
intended to retain gravels in a fixed place, to allow
their ready removal when the trap filled up, and to
ensure that, downstream, the Wharfe maintained
its engineered riverbed level. The trap filled at a
faster rate than expected with the result that, by the
mid-1990s, the gravel trap was considered ineffective. Gravel accumulation has continued and large,
migrating gravel shoals are evident. The latter
force high velocity water to flow against the river
flood banks and, in some cases, lead to erosion of
the banks and leves. Similarly, there has been progressive aggradation of the river bed, reducing the
effectiveness of land drainage and increasing flood

267

risk. In this paper, we focus upon the process by


which the gravel trap was eventually removed and
the river reach redesigned.

The system response model


On the basis of the above description, we can
identify a highly simplified system response model
for the river (Figure 3). In the case of the Wharfe, this
was aided by a relatively knowledge-rich situation.
Maps and design sections for the original Buckden
scheme were acquired. The political sensitivity of
the scheme, coupled to a local Environment Agency
initiative (the Upper Wharfedale Best Practice Project,
see below), resulted in a series of commissioned
reports in the 1990s (Heritage and Newson 1998a
1998b; ARUP-RKL 1999; Jeremy Benn and Associates
2000). The model was developed from the above
reports, plus additional field investigations and
provided the sort of simplified model that can be
shown to and appreciated by a local community.
The model identified the root causes of behaviour in the system, in this case associated with
upstream gravel delivery linked to catchment-scale
hydrological processes. High rates of gravel delivery, coupled to a reduction in valley slope and an
increase in valley width (i.e. catchment-scale geomorphology) result in the reach downstream of
Hubberholme, covered by the 1980s Buckden
Scheme, being associated with a change in the
nature of sediment transfer. Under current delivery
rates, this leads to a tendency towards net sediment deposition, but with distinct erosion and
deposition zones. As gravel deposition normally
occurs in the form of bar forms, aggradation of this
type encourages bank erosion, as bar deposition
creates the flow divergence and convergence and
consequent secondary circulation that leads to
channel change. General bed aggradation also
increases the magnitude and frequency of flood
inundation and hence the transfer of fine and
coarse sediment to the floodplain as part of the natural process of floodplain construction. Bed aggradation and channel change in turn determine the
spatial patterns of future erosion and deposition.
Thus the river does not behave as a sediment conveyor (cf. ARUP-RKL 1999): rather it shunts sediment down through the river system, leading to
localized erosion and aggradation (i.e. a morphological signature) as it does so. On the basis of
these interactions, we can identify the expected
structure, function, diversity and dynamics expected
in an upland river of this type (Figure 3).

268

Adrian McDonald et al.

Figure 3

System response model for the Upper Wharfe

The Wharfe runs out of gravel somewhere around


Tadcaster, just before it joins the Yorkshire Ouse,
downstream of York, approximately 80 kilometres
downstream of the restoration site. Given the size
of particles delivered to the restoration site (maximum 0.320 metres) and typical rates of abrasion, it
is unlikely that abrasion can be the sole explanation of the downstream fining: i.e. the gravels must
be going into storage. Thus, the Wharfe as a whole
is almost certainly a gravel storage river. This is

important as any measures (e.g. channel realignment) that serve to increase local transport capacity
and reduce floodplain construction may result in
downstream impacts (i.e. increased coarse sedimentation) that in turn may need to be managed.

Individual and community goals


The legal responsibility for the maintenance of
river beds and banks and the provision and
maintenance of flood defences rests primarily with

Rivers of dreams

the riparian owners. However, during the last half


of the twentieth century, legislation gave powers to
the relevant Authority to carry out river works for
the purpose of flood defence and land drainage.
The Environment Agency (and its predecessor
organizations, the National Rivers Authority and
Yorkshire Water Authority) has permissive powers
under the Water Resources Act 1991 to carry out
certain works in connection with the main river.
These include construction and maintenance of
both flood defences and the water course. Up until
the early 1990s, nationally, river management had
a strong focus upon land drainage and flood
defence. Only in the last decade have environmental considerations emerged. However, active
management of the Wharfe extends back into the
twelfth and thirteenth centuries, if not earlier,
associated with the maintenance of high-value
floodplain meadow land. Records from the
nineteenth and early twentieth centuries show
active removal of river gravels by local landowners
to reduce river-bed aggradation and hence the
magnitude and frequency of both overbank
flooding and bank erosion.
The growing environmental requirement has
combined with designation of the Yorkshire Dales
National Park, the development of national-scale
landowner interests (notably the National Trust)
and increasing influence from Non-Governmental
Organizations (e.g. The National Trust) to complicate responsibility and to increase the range of stakeholders with legitimate and/or perceived interests
(e.g. Clark 2002). In practical terms this is a river
that has, through precedent, become the focus of
multiple management, with a less clear understanding of who is responsible for what and no
clear identification of management goals (or who
should define them). However, in 1997, the Upper
Wharfe became the site for an important pilot study
on best practice for the sustainable management of
the land and water (Upper Wharfedale Best Practice Project, UWBPP). This is a partnership, managed by the Environment Agency, with a Steering
Group. Membership of the latter included: (1) local
landowners and farmers, both individuals and
groups like the National Trust; (2) the Yorkshire
Dales National Park; (3) the Forestry Commission;
(4) the Department of the Environment, Farming
and Rural Affairs; (4) Angling Associations;
(5) English Nature; (6) environmental managers,
including Tilhill Economic Forestry; (7) representatives of local decisionmakers, notably parish councils

269

and farmers; and (8) academics. The partnership


received a grant from the European Commissions
European Agricultural Guidance and Guarantee
Fund, Objective 5b programme for the Northern
Uplands and the Millennium Commission through
Yorkshire Dales Millennium Trust. The aim of the
UWBPP was to determine the principles, techniques and benefits of an integrated way of achieving good land and water management in an upland
environment. It is based on an ecological approach,
protecting habitats and water quality in the catchment, whilst encouraging a move towards more
sustainable hill farming, and taking into account
social and economic considerations. The extent of
the UWBPP provided a fundamental focus for both
the identification of individual and community
goals and a process by which those goals could be
identified and evaluated.

Process for identifying restoration options


Figure 1 does not help in the operationalization of
a restoration project: it explains what needs to be
done, but not how to achieve it. For instance,
Goodwin (1998) notes that achieving the participation of local people is desirable, but the basis by
which this might be achieved rarely has been
addressed. In river restoration terms, there are
good examples of public participation (e.g.
Tunstall et al. 2000; Carroll et al. 2002; Scholz et al.
2002), but less written on how exactly this might be
achieved, with the exception of Clark (2002) and
Clark and Richards (2002). In the case of the upper
Wharfe, the major steps in the process are shown
in Figure 4. This process came about from 1999
onwards as a result of dissatisfaction over river
management within the UWBPP linked to initial
attempts (19972000) to do something about the
gravel trap problem. This dissatisfaction reached
its nadir in spring 2000, at which point the
designed solution for the gravel trap reach
recommended a series of management activities
including stabilization of the river and maintaining
a continuous transfer of sediment through the
system as the core goals (ARUP-RKL 1999). This
plan predominantly involved conventional hard
engineering (i.e. bank protection measures) within
the river. At least some of the stakeholder
representatives felt that: (1) this was not reflecting
their perceptions and observations of how the river
worked (it was naturally dynamic with sediment
moving discontinuously between aggrading and
degrading zones, in turn changing the propensity

270

Figure 4

Adrian McDonald et al.

The agreed process adopted for operationalizing the interface between data, a system response model
and individual and community goals for the case of the Upper Wharfe

to aggradation and degradation); and (2) that the


current gravel trap problem was a result of an
over-reliance upon attempts to train or to stabilize
the river. Thus, whilst the dissatisfaction was
manifest from a perceived approach to river
management that was unacceptable, its root driver
was local knowledge of river system behaviour.
This is reflected in Figure 4. The process, similar
to that identified by Clark and Richards (2002), was

agreed by the UWBPP, with the stakeholder representatives making sure that this process related to
the needs, desires and aspirations of the wider constituencies that they represented. From an early
stage the UWBPP wanted a Delphi group process
operating at two levels in order to identify restoration options, similar to that adopted by Scholz et al.
(2002) and described by Clark and Richards (2002).
The first group was all interested stakeholders, as

Rivers of dreams

defined by the UWBPP, and making direct links to


the desires of the local community. The second
group was a sub-group of experts, which the
UWBPP found necessary for a number of reasons.
First, there was a need to obtain the required
knowledge for identifying a system response
model, using experience that went beyond the geographical and discipline experience of the UWBPP.
This was not so much in relation to templates for
restoration (cf. Eden et al. 2000), but much more
generally in relation to the dynamics of an upland
gravel-bed river.
Second, the technical group provided a crucial
educational role in helping both the UWBPP and
the local community to understand how their river
worked. This involved tying the observations and
experience of local people in relation to their river
(which had driven the move to the process shown
in Figure 4) to the state of upland gravel bed river
science. This was interesting in that it emphasized
the very grey role of the expert in this process. On
the one hand, following Wynne (1993) and Goodwin (1998), there was a lack of trust in the expertise
available. Continued problems of river management following the 1980s Buckden Scheme as well
as the aborted 2000 management plan were cited as
reasons to be cynical about whether or not yet
another set of management activities might have a
beneficial impact. However, the issue of trust was
more subtle than this. The local community have
since acknowledged that one of the most crucial and
successful components of the restoration activity
was the way in which the technical group helped
them to understand what their riverfloodplain
system was doing. For instance, a farmer who had
moved into the Dale ten years previously had
always wondered what the long linear depressions
on his land were. To a river scientist, these were
clearly palaeochannels, indicative of an historical
river course and hence of natural processes of erosion and deposition. The farmer welcomed the
associated understanding, even if it did not change
his desire for a stabilized river. In general, using
community members observations of what their
environment was and what it did allowed the technical group to bring to the fore the contradiction
between river channel management activities that
seek to stabilize the river and the natural dynamics
of a wandering upland gravel bed river in the
absence of human activity. Understanding this contradiction allowed a much more critical community
involvement in the range of restoration options

271

presented, even if the eventual option chosen was


less satisfactory from a river science perspective
alone, something that the community, and the
project as a whole, eventually defined as quite
satisfactory.
Third, the local community wanted to be presented with a range of options, with positive and
negative benefits for them to evaluate, rather than
them being asked to identify the options themselves.
There was a genuine feeling that they didnt understand what the river was doing, and that experts of
some sort helped them to make sense of it. In all of
this, the technical group had to be careful not to
fall foul of the problems identified by Rhoads et al.
(1999) where we: (1) failed to distinguish between
our knowledge and our values, especially given
uncertainties in the knowledge base; (2) privileged
our knowledge over other forms of knowledge;
and (3) presented plans and strategies as recommendations that could run counter to what Rhoads
et al. (1999) label the sociocultural lifeworlds of
nonscientists. Indeed, in this process, we found it
hard to separate analytical statements from normative statements, and it was clear that combining lay
understanding with traditional scientific expertise
was a very new and very challenging enterprise. In
this instance, the social and historical context of the
Wharfe landscape proved crucial. The local community (both landowners and non-landowners)
tied river management to maintaining high quality
hay meadow by stopping bank erosion; and preserving a landscape that is perceived as stable
as part of our cultural heritage, both because of
its economic value (tourism) and because of its
cultural value (especially to those who had retired
into the Dale).
Figure 4 shows that a preliminary site meeting
(May 2000) between the main landowners and
stakeholders, with some technical input, agreed an
option appraisal approach. This required technical
input through which the technical group, agreed by
the stakeholders (May 2000), explored and clarified
the options determined by the main group. The
technical group reported directly to a site meeting
of the Project Steering Group (July 2000) and to
the local community (September 2000) through the
auspices of the main community group within
the Upper Wharfe, a Parish Council. A large on-site
meeting in October 2000 presented the options to
the Project Steering Group and stakeholders. At
this stage, the options were broad in scope and
spanned the full interventionnon-intervention

272

range, including a set of intermediate options. The


Project Steering Group and stakeholders then
rated each option in terms of acceptability to them.
These options were then taken to an open meeting
of the Parish Council (November 2000), attended
by about 40 community members and stakeholders,
including an open discussion of how their river
worked. Thus, a system response model and restoration options were presented simultaneously. The
system response model was refined through a
mixture of technical presentation and community
discussion and debate. The preferred stakeholder
and Project Steering Group options were presented and the community preferred a combination
of two of the presented options. The Project Steering Group subsequently accepted the community
recommendation and proceeded to detailed design
with both stakeholder and technical input (January
2001).

The management options


The management options were developed with the
aim of seeking a range of solutions that lie between
two extremes. On the one hand, maintaining the
current channelized river was identified as a
solution, based upon removing gravels frequently,
fixing or re-walling the channel and felling trees
that have the potential to damage walls or obstruct
the channel. This was identified as having a high
need for on-going maintenance. It would have
fewer ecological benefits and the frequency of
works needed may reduce the visual appeal of the
general valley landscape. At the other extreme was
the option of letting the river define its own
channel. This would mean letting gravels
accumulate, banks move by erosion, new channels
form in time, and generally allowing natural
processes to create the form and position of the
channel continually. In the full range of solutions,
attention had to be given to possible downstream
impacts of any decisions made. Evidence from
cross-section records showed that there had been
a long-term geomorphological response to gravel
trap installation, with initial degradation until
the trap was filled, followed by slow aggradation.
Any decision to manage the gravel trap reach in
a different way could have major downstream
impacts.
Between these two extremes, a number of
options were identified (Table II). A number of
important points emerge. First, and following from
Figure 1, Table II is neither an exact statement of

Adrian McDonald et al.

the technical context of the project nor of the goals


and aspirations of either society or individuals, but
the combined influence from a technical appraisal
and stakeholders views. Second, and following
from this, the positive and negative aspects identified are both objective and, importantly, subjective:
they reflect community views of each option.
Third, throughout the restoration options, there is
recognition of the problem of root causes, associated with sediment delivery in combination with
catchment-scale geomorphology, which would be
difficult to remove. Community views, coupled to
the difficulty in addressing root causes, meant that
the restoration options included cultural and economic factors in the positive and negative aspects
and medium- and long-term issues identified.
Finally, the options identified differed in their definition of naturalness in river behaviour and, more
importantly, in their recognition of downstream
impacts. By increasing sediment throughput (e.g.
as per Option 1), there will be reduced sediment
retention in the restoration reach and increased
sediment delivery downstream. This represented
the view that the natural function of a river was to
transfer sediment and is at odds with the evidence
that suggests that the natural function of this river
reach (gravel doesnt reach the sea, lateral channel
migration, palaeochannels) was to construct floodplain through sediment storage. Sediment transfer
and erosion, deposition and floodplain construction are two sets of natural functions that will coexist, but the discontinuous nature of sediment
transfer means that it wont be independent of
floodplain construction and erosion. Options that
prioritize sediment transfer are likely to have
downstream impacts that may be serious.
The vehicle by which community views were
recognized and incorporated involved a two-stage
process. First, options were reduced by the technical group, which community members and other
stakeholder representatives were members of. This
gave each option a community perspective, prior to
presentation to the full community. Second, all
options and stakeholders views were presented to
a full meeting of the community, defined geographically by the catchment divides and extending about 13 kilometres downstream from the
restoration site. This was facilitated by the spatial
arrangement of rural parishes, which meant that
Buckden (restoration site) and Kettlewell (immediately downstream) parish councils were able to
invite any interested community members to

Rivers of dreams
Table II

273

Summary of restoration options

Option

Brief summary

1. An engineered
structure to
consolidate eroding
bank, protect the
footpath and to
protect against
flooding at that site

Enhance current bank protection works on the true right bank by providing an engineered structure
capable of consolidating the bank thus enabling the long-distance footpath (the Dales Way) to
remain in its current position. Acknowledges that the Wharfe channel, local to Cray Beck, has
narrowed because of the northern (left) bank walling, gravel accumulations from Cray Beck, plus
a large gravel shoal opposite the confluence that has accumulated in the last 130 years. The
accumulated right bank gravel shoal opposite Cray Beck would be removed allowing free passage
of water and further reducing the erosion stress on the new river bank walling upstream. This
section of channel seems to be developing into a tight meander; maps produced over the last 130
years illustrate this process. The option therefore seeks to reverse this trend in order to protect the
right banks and current footpath. The planform would be designed so as to enhance sediment
transport through the reach.
Combines elements of bank protection on the right bank with elements of removing bank protection
on the left bank. Acknowledges that the meander within this reach has become tighter, and allows
this process to continue on the left bank (currently heavily protected), local to Cray Beck, although
it may reduce erosion pressure upstream on the right bank. By allowing some river migration,
there is the possibility of floodplain construction, and this begins to address the issue of
downstream impacts. Option 2 seeks to use more vegetation to protect the banks. This would
extend upstream and downstream of the meander. The islands associated with the Cray Beck
confluence prior to the Buckden Scheme would be restored.
This option tries to overcome one of the key concerns with Option 2, namely that the area for sediment
accumulation is marginally smaller. This aspect is of concern since ideally we are trying to protect
downstream river sections through the management of the gravel trap reach. The proposal is
therefore to establish a by-pass channel on the northern floodplain (left bank) that allows some
water and sediment to by-pass the right bank defences thereby reducing erosion pressure. The
area of woodland increases to protect the new channel through soft bank protection methods.
This proposal is to realign the River Wharfe to the lowest point in the floodplain. Centuries of river
training, gravel use and levee construction has resulted in levee levels well above the flood plain elevation
and riverbed levels that are approaching the general floodplain level. If continued and in the light
of the cessation of gravel exploitation, the riverbed level may well be significantly above the
floodplain as has occurred in some Lake District rivers. Following realignment, management
would allow normal flooding and sediment to spread laterally over the floodplain. This would
involve the restoration of natural processes and enable the river to define its own channel. It
would also restore the crucial process of floodplain construction associated with sediment storage.
This is the do nothing option now required to be evaluated in much US environmental legislation.
However, the do nothing option can, in this case, take two forms. The first, do nothing, would
be to make no change to the planned current management regime. The Buckden Scheme would
be maintained as intended with the gravel trap emptied at regular intervals. This is treated as
Option 5 here. The second do nothing option is the hands-off option. In effect then this
maintains the de facto status quo of the last decade, namely neither maintain the gravel trap, nor
any of the gravel shoals. But, and this is a major change, it also requires non-intervention with
current bank erosion and allows the river to adjust its form and position from its current position
in the valley. This is treated as Option 6 below.
No intervention or maintenance. Allow the river to erode and eventually change course and form
with a return to natural processes of sediment erosion, deposition and storage.
This proposal takes elements of Option 6, namely do nothing and allow the river to develop its own
form, with elements of Options 1 and 2. The main elements are to remove the Buckden Scheme
bank and riverbed protection measures. Allow Cray Beck island to be reformed and the natural
sinuosity of this river reach to develop, both on the left and right bank. The flood risk to fields
south of the meander would be protected by strengthening the current levee. In the medium to
long-term, and with no gravel clearance within this section, the meander should be expected to
tighten on the basis of current changes. The levee will be reformed behind the eroding bank at
intervals to reduce the risk of the river breaching this part of the bank and flowing southwards
over the lower lying floodplain (blister repair). Farm access to the field and the public footpath
would follow the newly formed levee. The proposal therefore seeks to allow the river to define is
bed form, allow gravels to accumulate and banks to erode until a quasi-stable river form is
reached where erosion rates become lower. Bank protection works will retreat with the eroding
bank or advance with the revegetation and soil development of older gravel areas. This option
seeks to avoid an uncontrolled breach of the levee while promoting some natural processes,
including sediment storage, with beneficial downstream impacts.

2. Reduce left bank


protection and
encourage movement
on the north side of
the river

3. Bypass channel

4. Realignment and
planned retreat

5. Gravel extraction
from gravel trap and
manage the gravel
deposition

6. No intervention.
Unplanned retreat
7. Working with the
river

274

attend, along with the technical advisors and all


other stakeholders. Involvement of Kettlewell was
important as the downstream community that
would clearly be impacted by any upstream activities, but also as the local focus of tourist activity. At
this meeting, the final recommendations were
made: (a) to adopt a modified version of Option 2,
combined with removal of the gravel trap (Figure 5);
but (b) to agree that a longer term management
plan would be required to manage the gravels
(under Option 5). This meeting was not uncritical
of these solutions (i.e. their limitations were recognized). However, during the meeting, the community strongly aligned themselves with local
landowners whose land was most prone to river
erosion problems. After a delay due to the foot and
mouth outbreak that began in February 2001, planning permission for the scheme was obtained in
November 2001. This was not a straightforward
process, and the decision was taken after careful
review by the National Park Planning Committee,
which took into account written and verbal comments from all stakeholders. It was at this point
that some stakeholders, notably those with a substantive (e.g. enhancement of river ecology) rather
than a place-specific or local perspective, voiced
concerns over the restoration project and notably
the use of bank erosion protection measures. For
a short period, divisions amongst stakeholders,
aligned with different lay understandings, dominated the debate. However, the project was
approved along the lines agreed at the community
meeting and care was taken to include design features that would help to soften hard design elements and provide some ecological benefit. The
project was implemented in summer 2002. Before
implementation, a detailed topographical survey
was undertaken throughout the reach and extending ten kilometres downstream. This was repeated
immediately after the survey and will be repeated
in a mixed stratified-systematic manner: surveys
every six months, interspersed with additional surveys after sediment transporting events. Additionally, a series of meetings of the UWBPP and of the
community are planned to allow the evaluation
and monitoring to be extended to include community and institutional goals.

Evaluation: lessons learnt


There are two different ways in which a scheme
such as this one may be evaluated. The first is in

Adrian McDonald et al.

terms of the restoration actually adopted. The


second is in terms of the process by which the
restoration project was identified under the sort of
general conceptual model shown in Figure 1 and
taking into account the specific approach adopted
in this case (Figure 3).

The rehabilitation project


From the view of the knowledge base and associated system response model, the restoration option
adopted was not ideal: it was clearly a rehabilitation approach. The decision to remove the gravel
trap was taken: despite it being full, it appeared to
be acting as a sediment transfer discontinuity
associated with upstream sediment accumulation.
This was unsustainable and the decision to return
the overall reach to a more continuous meander
had the capacity to increase gravel throughput,
with attendant downstream impacts in terms of
increased sediment delivery. This is the classical
dilemma of all local river management activities.
To break out of this problem, it was necessary to
restore the natural process of sediment storage.
However, Option 2 largely focused upon the
stabilization of the true right bank to protect high
value pasture on that side of the river. A number of
issues arose. First, it passed over the root cause of
the problem as indicated by the system response
model, the delivery of gravel to the reach, by
advocating that a gravel management plan should
be developed. The nature of such a plan, and its
feasibility, were not considered at the point at
which Option 2 was adopted, although it was
acknowledged by certain stakeholders and the
technical advisors that such a plan would be
required if the scheme were to be sustainable in the
longer term.
Second, in terms of Figure 3, it was a restoration
project that was largely about preventing natural
system dynamics. Without a control on gravel
delivery, expected channel response was erosion of
the outer bank of the meander in response to bed
aggradation. To prevent this, once the works were
completed, it was decided that deflectors would be
required on the outer bank to help reduce erosion
due to processes that would be expected in any
curved river channel. The restoration very much
followed the field of dreams view. The restoration
process was dominated by community concerns to
stop bank erosion and not general concerns to
allow the river to behave in a natural way. Indeed,
albeit implicitly, questions of naturalness were

Figure 5

The final option adopted for restoration of the Buckden Gravel Trap Scheme based upon Option 2 (Table II)

Rivers of dreams
275

276

continually negotiated. The original Buckden scheme


was introduced in the 1980s to reduce the high frequency of flooding associated with high rates of
bed sediment aggradation and under the assumption that a range of engineering activities could
increase the throughput of sediment. Reintroducing natural channel dynamics was not acceptable
to the local community for a range of socioeconomic reasons (affects farming activities, closes
footpaths). Simultaneously, the high rate of gravel
delivery and the natural tendency for the river to
migrate meant that some form of hard engineering
was inevitable in order to satisfy local community
concerns. Thus, high rates of sediment delivery,
channel migration and flooding were all both good
(natural processes) and bad (from the perspective
of the community), and the exact definition of
which view holds was the consequence of negotiation. This is an example of how contingent natural
(the nature of the river) and social processes can
exert an important effect upon the way in which
restoration evolves. In this case, as has been
observed in other river restoration studies (e.g.
Rhoads et al. 1999), there was a strong communitybased ethic, case-centred on stabilizing the river.
Even though the expert group spent considerable
time describing and explaining why channel
migration was a fundamental aspect river behaviour, the stabilizing ethic remained engrained in all
generations of the community. Many of the more
radical options (e.g. Option 7) were simply unacceptable because they not only allowed bank erosion to occur, but they would have resulted in
profound changes to a landscape that has acquired
significant historical and cultural value. The landscape was already a considerable distance from
being natural as a result of the duration of human
occupation and intensity of management activity
(Brookes and Shields 1996). However dissatisfying
this may be from a technical view point, it is almost
certainly the case with any river restoration project
in an upland environment that has distinct economic and cultural value. Notions of history, as
held by river scientists armed with plenty of supporting evidence (such as palaeochannels, stratigraphic exposures in eroding river banks), were
markedly at odds with the histories defined by
local communities, some of whom have lived for
many years within the river ecosystem and who
see in their landscape a strong local distinctiveness
(Adams 1997). These points emphasize that
phrases like natural have very little meaning in

Adrian McDonald et al.

landscapes that have acquired value in a range of


cultural, historical and economic senses and most
restoration projects will be distanced from the
associated system response model. This emphasizes the need for a fundamental change in the
nature of post project appraisal to move away from
narrowly technical assessment to one that incorporates much broader goals.
From the perspective of individual and community goals, once some attempt to stabilize the river
had been agreed, the local community became supportive. However, the history of intervention to
manage coarse sediment as part of a flood control
strategy had not been marked by success and this
led to scepticism on the part of both the stakeholders in general and the community in particular.
Community goals were also blurred by the political
position of the project. For an upland site in the
north of England, there were a large number of
stakeholders, several of whom were powerful
agencies. The decision would certainly set a precedent because the work arose from a Europeanfunded study of best practice; it was therefore
integral to the study that the process and outcome
would set at least a well-documented example of
the decision process. This aside, the decisionmaking process recognized that because public
resources are used and because the community is
impacted, the community must be able to influence
the decisionmaking process. Indeed, expert knowledge was effectively subservient to local consultation. The crucial role (from a system response
model viewpoint) of dynamics (bank erosion, channel migration and so on), as identified by the
experts, was accepted by stakeholders and at community meetings, but not to the extent to which the
desire for bank stabilization was undermined.
As Rhoads et al. (1999) and Eden et al. (2000) consider for restoration in general, the Wharfe project
became neither concerned with purely a system
response model nor the definition of a set of individual and community goals. Rather, it is a hybrid
combination (Eden et al. 2000) of both nature and
culture that is very different to what either nature
or culture might suggest in isolation. The model in
Figure 1, and the way in which the options in Table
II were developed, reflects this combination. From
a scientific point of view, it is difficult to defend
any of the options suggested from purely the perspective of how the system works, and this demonstrates the way in which scientific involvement
in the restoration process differs considerably from

Rivers of dreams

traditional models for how science is linearly incorporated into policy (Clark 2002). Similarly, Figure 1
encapsulates the idea that restoration becomes a
hybrid process (Eden et al. 2000) because of the
way in which community and individual goals, as
well as the system response model, must both
evolve in relation to the experience of active intervention in the system.

The restoration process


The above section started to illustrate the challenges of defining individual and community goals.
Whilst theoretically appealing, securing adequate,
effective and fair community involvement in the
restoration process is not straightforward because
of the effects of spatial scale. Whilst the physical
and ecological dimension of restoration might be
readily based around natural catchment and landscape units, defining the community with interests
in a particular river restoration is much more
difficult. There are a number of reasons for this.
The first is antithetical. If river restoration is to be
predicated upon removal of the root causes of river
degradation (see above), and those causes are geographically distant from the restoration site, then
this requires the involvement of a community that
may also be geographically distant (Pfadenhauer 2001).
This is especially the case in river restoration,
where the river provides a strong physical connection
between upstream catchment-scale processes and
the restoration site that is under consideration,
which may be some way downstream. Thus, in the
case of the Wharfe, the root cause was identified as
the problem of high rates of gravel delivery from
upstream. The extent to which this was related to
upstream land use management practices and
environmental changes was uncertain. The Wharfe
was associated with many instances of enlightened
land use management (e.g. active involvement of
land owners in the restoration of blanket bog in
areas of artificially drained moorland). It follows
that a crucial component in addressing catchmentscale root causes is a land management process
that incorporates the range of downstream impacts
associated with particular land management
activities through appropriate mechanisms for
land-use decisionmaking (van Diggelen et al. 2001).
In this case, it was aided by funding mechanisms
for alternative land use management under both
national and EU legislation.
Second, the Wharfe is a river that has value for
a very wide range of users that are not readily

277

defined in geographical terms. These include anglers,


ramblers and nature conservationists. The Best
Practice Project was founded upon a broad definition of community membership, with representation from angling associations and others, and this
was factored into the decisionmaking process.
The process identified in Figure 4 gave a strong
local community input to the decisionmaking process. However, the remit of the Planning Committee
who eventually approved the scheme extended
well beyond the concerns of the immediate, local
community and this gave non-local stakeholders
the opportunity to influence the final decision
directly. The result was the re-emergence of conflict divided largely along the lines of locals (particularly landowners) and those concerned with
issues (anglers, NGOs concerned with nature conservation). As Rhoads et al. (1999) show, conflict is
an inevitable component of the restoration process
and this leaves a major challenge for the restoration process: who should have a legitimate interest
in a given restoration project and what weights
should be given to different individuals and groups?
What constitutes the community? Although the
range of options presented resulted in the articulation of competing perspectives and views of the
desired nature and landscape (Swart et al. 2001),
there was no simple guidance as to the weight that
should be given to each perspective or view. When
the final decision was made, and following
Pfadenhauer (2001), the restoration was negotiated,
but around terms largely defined by a restricted
number of stakeholders dominated by the local
community. This was accompanied by feelings
amongst the community that, at the Planning Committee, a series of issue-based NGOs who were
remote (Goodwin 1998) to the project had tried
to interfere in their river despite them having little
understanding of the local context within which
the river was situated. This comment overlooks
significant within-NGO variation in their interaction
with the local community, and within-community
variation in the sympathies aligned to each NGO.
For instance, the National Trust was actively combining national-level policy objectives for the
conservation of soil and water with strong local
involvement with tenants and National Trust
Wardens. Thus, not all NGOs were remote,
although they varied strongly in how they were
perceived within local communities.
These issues need to be linked back into the concept of integrated catchment management (ICM).

278

ICM is commonly cited as a key prerequisite of an


effective river restoration (e.g. Boon 1998; Harper
et al. 1999; Wissmar and Beschta 1998; Ward et al.
2001; Everard and Powell 2002). However, many
river restoration projects, as with this case study,
tend to be small-scale and/or ad hoc in nature
(Harper et al. 1999). As noted in the Introduction,
there is an increasing emphasis, certainly in Europe
(e.g. through the Water Framework Directive) and
the UK (through Catchment Flood Management
Plans), upon the catchment as a natural management unit. However, the case study described here
starts to raise important questions about integrated
catchment management as a practical concept.
Whilst catchments might be readily identified in
biophysical terms, their social definition is much
more difficult, and the political and administrative
structure of our landscape is rarely structured
around a catchment. As the spatial scale of a restoration project increases, so do the political and
social constraints (Sear 1994; Brookes and Sear
1996), such that there may have to be growing
compromise between both biophysical and community perspectives as well as within different
community perspectives. At present, we have very
little idea as to how to achieve such compromise.
Whilst integrated catchment management has
much theoretical appeal, and is well supported by
biophysical processes, it provides little guidance as
to how it might be effective in practice. The common argument from a biophysical perspective that
a catchment-scale treatment of root causes is
required overlooks important research that suggests that it is often only the local catchment scale
that can engage stakeholders and communities
such that neighbourhood catchments (Carroll et al.
2002) may be the only socially meaningful scale
of restoration. Further, in a democratic society, if
political units are defined as spatial units that do
not match catchment boundaries, following a
catchment-based approach will raise questions of
democratic accountability. This effectively explains
why large-scale or catchment-scale experiments
have become less favoured (Darby and Thorne
2000), despite an increasingly large-scale legislative
framework (e.g. the Water Framework Directive).
Accepting this requires direct links to the arguments made by Eden et al. (2000): the outcome of
the restoration should not be seen in terms of
whether or not it has successfully restored the natural, as this is a label that only has meaning in a
narrowly defined sense. Rather, the outcome needs

Adrian McDonald et al.

to be seen in terms of how the restoration has


changed the understanding of and relationships
between stakeholders, technical specialists and
community participants, in relation to those who
work and walk along the local river bank. This
may lead to odd restoration projects, such as those
that focus upon the reintroduction of a single species, to the detriment of other potentially valued
aspects of the environment, and may fail to address
the necessary conditions for the sustainability of a
reintroduced species. This emphasizes the fact that
restoration is a complex process, at the heart of the
relationship between the natural and the social.
Simple biophysical conceptions that deny this connectivity may be theoretically ideal but practically
and democratically unacceptable.

Conclusion
In this paper, we have sought to develop a basic, if
mechanistic, conceptualization of the restoration
process as the basis from which we have
demonstrated some of the challenges facing river
restoration. These were explored through a case
study. Whilst mechanistic, the conceptual model
forced the explicit combination of a system
response model (and hence reflections upon how
rivers work) with consideration of community and
individual goals, in order to identify a set of
restoration options. In the case of the restoration
described here, it resulted in a set of restoration
options that were technically informed but,
crucially, were evaluated from a much broader
range of perspectives. Thus, the actual option
chosen appeared unsound when evaluated from a
technical perspective alone. However, such an
evaluation would miss the fundamental point of a
restoration project in which process of restoring a
system is just as important as the restoration itself.
Thus, the case-study reach of river was not
restored, but managed in a manner that reflected
strong community feelings as articulated through
certain stakeholders and supported by a wider
local community.
What makes this acceptable? First, a crucial
aspect is that the decisions made in this case are
not irreversible. The current proposal is limited
and could be recovered if the other commitments
(e.g. to develop a sustainable management plan to
address the root causes of the problem, gravel
delivery) were not secured. Second, the adopted
solution has the characteristics of an incremental

Rivers of dreams

change in river management policy away from


hard engineering. Such incremental change finds
more favour where there are extreme stances in the
stakeholder group, as in this case. Further, as there
was a high degree of uncertainty, incremental
change permitted adaptation and flexibility (Clark
2002). In addition, the cost of errors, defined in the
broadest sense, is reduced. Third, strong cultural
and historical influences largely prevented the restoration options that might seem ideal in biophysical terms. The case study here might be viewed as
an extreme one, given the landscapes cultural heritage and complex and stringent planning environment. However, the reported findings concord
with the work of others, notably of Rhoads et al.
(1999) and Eden et al. (2000), in demonstrating that
river restoration is not just a biophysical process,
but also a social, political and economic process. It
might lead to improvements in the biophysical
environment, but the process needs to be evaluated
in broader terms than ecosystem values alone.
Looking beyond the biophysical environment
raises difficult issues that those involved with river
management must address. The real issue is that,
however well a river system response model is
developed, defining community and individual
goals is more difficult. This is especially the case in
landscapes of significant historical and cultural
value, or where the structure of the landscape, in
terms of land management and land ownership,
prevents the ready achievement of integrated
catchment management. In some senses, the fluvial
geomorphologist has the straightforward responsibility: physical and ecological analyses of catchments as drainage basins are straightforward as
they are geographically well-defined. Incorporating lay understanding, defining what community
and individual goals are and should be, how they
could and should be used, and who defines them,
are problems that make the concept of integrated
catchment management appear to have real academic appeal but many more practical difficulties.

Acknowledgements
This research is supported by NERC Grant NER/
D/S/2000/01269 awarded to SNL, AM and Professor M. Kirkby, by Environment Agency award E1108 to AM and SNL, and by the National Trust.
The views expressed are solely those of the
authors. The second author was funded by the
Department of Geography, University of Illinois at

279

Urbana-Champaign to present a seminar on this paper


and this allowed discussion with Professor Bruce
Rhoads which significantly sharpened the ideas in
the manuscript. Dr Paul Waley, Professor Neil Ward
and Dr Geraldene Wharton provided constructive
comments on an earlier version of this manuscript.

References
Ackers P and Charlton F G 1970 Meander geometry arising from varying flows Journal of Hydrology 11 23052
Adams W M 1996 Future nature: a vision for conservation
Earthscan, London
Adams W M 1997 Rationalization and conservation: ecology and the management of nature in the United Kingdom
Transactions, Institute of British Geographers 22 27791
ARUP-RKL 1999 Dynamic assessment of unstable assessment
of reaches of the Upper Wharfe Report to Environment
Agency, Dales Area of the North-East Region Coverdale
House, York
Bennett J 2002 Investing in river health Water Science and
Technology 45 8590
Blench T 1969 Mobile-bed fluviology University of Alberta
Press, Edmonton
Boon P J 1998 River restoration in five dimensions Aquatic
Conservation: Marine and Freshwater Resources 8 25764
Boon P J, Calow P and Petts G E eds 1992 River conservation and management Wiley, Chichester
Brierley G, Fryirs K, Outhet D and Massey C 2002
Application of the River Styles framework as a basis for
river management in New South Wales, Australia
Applied Geography 22 91122
Brookes A 1996 River restoration experience in Northern
Europe in Brookes A and Shields F D eds River channel
restoration Wiley, Chichester 23367
Brookes A and Sear D A 1996 Geomorphological principles
for restoring channels in Brookes A and Shields F D
eds River channel restoration Wiley, Chichester 75101
Brookes A and Shields F D Jr eds 1996 River channel restoration Wiley, Chichester
Carlston C W 1965 The relation of free meander geometry
to stream discharge and its geomorphic implications
American Journal of Science 263 8645
Carroll C, Rohde K, Millar G, Dougall C, Stevens S,
Ritchie R and Lewis S 2002 Neighbourhood catchments: a new approach for achieving ownership and
change in catchment stream management Water Science
and Technology 45, 18591
Christensen N L 1997 Managing for heterogeneity and
complexity in dynamic landscapes in Pickett S T A,
Ostfield R S, Shachak M and Likens G eds The ecological basis for conservation: heterogeneity, ecosystems and biodiversity Chapman and Hall, New York 16786
Clark M J 2002 Dealing with uncertainty: adaptive approaches to sustainable river management Aquatic Conservation: Marine and Freshwater Ecosystems 12 34763

280
Clark M J and Richards K J 2002 Supporting complex
decisions for sustainable river management in England
and Wales Aquatic Conservation: Marine and Freshwater
Ecosystems 12 47183
Conservation (Natural Habitats, &c) Regulations 1994
Statutory instrument of the United Kingdom parliament number 2716, October 1994, ISBN 0110457161
Darby S E and Thorne C R 2000 A river runs through it:
morphological and landowner sensitivities along the
Upper Missouri River, Montana, USA Transactions,
Institute of British Geographers 25 91107
Downs P W and Thorne C R 1996 A geomorphological
justification of river channel reconnaissance surveys Transactions of the Institute of British Geographers 21 45568
Eden S, Tunstall S M and Tapsell S M 1999 Environmental restoration: environmental management or environmental threat? Area 31 1519
Eden S, Tunstall S M and Tapsell S M 2000 Translating
nature: river restoration as nature-culture Environment
and Planning D 18 25773
Ehrenfeld J G 2001 Defining the limits of restoration: the
need for realistic goals Restoration Ecology 8 29
Environment Agency 1998 River geomorphology: a practical
guide Environment Agency guidance note 18 National
Centre for Risk Analysis and Options Appraisal, London
ERM 1983 River Wharfe Buckden scheme: assessment report
Environmental Resource Management, Norwich
European Commission 1992 Directive 92/43/EEC of the
European Parliament and of the Council of 21st May
1992: On the conservation of habitats and of wildlife
and fauna Official Journal of the European Communities
L206 17
European Commission 2000 Directive 2000/60/EC of the
European Parliament and of the Council of 23rd October
2000: Establishing a framework for Community action
in the field of water policy Official Journal of the European
Communities L327 172
Everard M and Powell A 2002 Rivers as living systems
Aquatic Conservation: Marine and Freshwater Ecosystems
12 32937
Federal Interagency Stream Restoration Working Group
1998 Stream corridor restoration: principles, processes and
practices National Technical Information Service,
Springfield VA
Frissell C A, Liss W J, Warren C E and Hurley M D 1986
A hierarchical framework for stream classification:
viewing streams in a watershed context Environmental
Management 10 199214
Goodwin P 1998 Hired hands or local voice: understandings and experience of local participation in conservation Transactions, Institute of British Geographers 23
48199
Graf W L 1988 Applications of catastrophe theory in
fluvial geomorphology in Anderson M G ed Modelling
geomorphological systems Wiley, Chichester 3347
Harper D M, Ebrahimnezhad M, Taylor E, Dickinson E,
Decamp O, Verniers G and Balb T 1999 A catchmentscale approach to the physical restoration of lowland

Adrian McDonald et al.


UK rivers Aquatic Conservation: Marine and Freshwater
Ecosystems 9 14157
Heritage G J and Newson M D 1998a Geomorphological
audit of the Upper Wharfe Report to Environment
Agency, Dales Area of the North-East Region, Coverdale House, York
Heritage G and Newson M D 1998b Dynamic assessment
of the gravel trap on the River Wharfe, upstream of Buckden
Report to Environment Agency, Dales Area of the
North-East Region, Coverdale House, York
Hey R D and Thorne C R 1986 Stable channels with
mobile gravel beds Journal of Hydraulic Engineering,
American Society of Civil Engineers 112 67189
Hey R D and Winterbottom A N 1990 River engineering
in National Parks: the case of the River Wharfe, UK
Regulated Rivers: Research and Management 5 3544
Higgs E S 1997 What is good ecological restoration? Conservation Biology 11 33848
Hildrew A G and Giller P S 1994 Patchiness, species
interactions and disturbance in the stream benthos in
Giller P S, Hildrew A G and Raffaelli D S eds Aquatic
ecology, pattern, scale and process Blackwell, Oxford 2162
Hobbs R J and Harris J A 2001 Restoration ecology:
repairing the earths ecosystems in the new millennium
Restoration Ecology 9 23946
Hobbs R J and Norton D A 1996 Towards a conceptual
framework for restoration ecology Restoration Ecology 4
93110
Holl K D and Howarth R B 2000 Paying for restoration
Ecological Restoration 8 2607
Howard A J, Macklin M G, Black S and HudsonEdwards K A 2000 Holocene river development and
environmental change in Upper Wharfedale, Yorkshire
Dales, England Journal of Quaternary Science 15 23952
Jeremy Benn and Associates 2000 Upper Wharfedale best
practice project: hydraulic model final report to the
Environment Agency 2 volumes
Jordan W R 1994 Sunflower Forest: ecological restoration as the basis for a new environmental paradigm in
Baldwin A D, de Luce J and Pletsch C eds Beyond preservation: restoring and inventing landscapes University of
Minnesota Press, Minneapolis 1734
Kern K 1992 Restoration of lowland rivers: the German
experience in Carling P A and Petts G E eds Lowland
floodplain rivers: geomorphological perspectives Wiley,
Chichester 27997
Kondolf G M 1998 Lessons learned from river restoration
projects in California Aquatic Conservation: Marine and
Freshwater Ecosystems 8 3952
Leopold L B and Wolman M G 1957 River channel
patterns braided, meandering and straight Professional
Paper of the US Geological Survey 282B
Lindblom C 1959 The science of muddling through Public
Administration Review 19 7988
Lindblom C 1979 Still muddling, not yet through Public
Administration Review 39 51726
Longfield S and Macklin M G 1999 The influence of
recent environmental change on flooding and sediment

Rivers of dreams
fluxes in the Yorkshire Ouse basin Hydrological Processes
13 105166
National Rivers Authority 1995 Post project performance
evaluation for River Wharfe Buckden Scheme Report prepared by Sir William Halcrow and Partners, Swindon,
February 1995
Newson M D 1997 Land, water and development: sustainable
management of river basin systems 2nd edn Routledge,
New York
Newson M D 2002 Geomorphological concepts and tools
for sustainable river ecosystem management Aquatic Conservation: Marine and Freshwater Ecosystems 12 36579
Newson M D and Sear D A 1997 The role of geomorphology in monitoring and managing river sediment
systems Journal of the Institute of Water and Environmental Management 11 26470
OBrien M 2000 Making better environmental decisions MIT
Press, Cambridge MA
Pahl-Wostl C 1995 The dynamic nature of ecosystems: chaos
and order intertwined Wiley, Chichester
Palmer M A, Ambrose R F and Poff N L 1997 Ecological
theory and community restoration ecology Restoration
Ecology 5 291300
Passmore D G and Macklin M G 2000 Late Holocene
channel and floodplain development in a wandering
gravel-bed river: the River South Tyne at Lambley,
Northern England Earth Surface Processes and Landforms
25 123756
Pfadenhauer J 2001 Some remarks on the socio-cultural
background of restoration ecology Restoration Ecology 9
2209
Pickett S T A and White P S eds 1985 The ecology of natural
disturbance and patch dynamics Academic Press, New York
Resh V H, Brown A V, Covich A P, Gurtz M E, Li H W,
Minshall G W, Recce S R, Sheldon A L, Wallace J B
and Wissmar R C 1988 The role of disturbance in
stream ecology Journal of the North American Benthological Society 7 43355
Rhoads B L and Herricks E E 1996 Naturalization of
headwater agricultural streams in Illinois: challenges
and possibilities in Brookes A and Shields F D eds
River channel restoration: guiding principles for sustainable
development Wiley, Chichester 33169
Rhoads B L, Wilson S, Ruban M and Herricks E E 1999
Interaction between scientists and nonscientists in
community-based watershed management: emergence of
the concept of stream naturalization Environmental Management 24 297308
River Restoration Centre 2002 Manual of river restoration
techniques 2nd edn River Restoration Centre
Sansom A 1996 Floods and sheep is there a link? Circulation 49 14
Scholz G, VanLaarhoven J, Phipps L, Favier D and
Rixon S 2002 Managing for river health integrating
watercourse management, environmental water requirements and community participation Water Science and
Technology 45 20913

281
Schumm S A 1960 The shape of alluvial channels in relation to sediment type Professional Paper of the US Geological Survey 352B 1730
Schumm S A 1968 River adjustment to altered hydrological regime Murrumbidgee River and palaeochannels,
Australia Professional Paper of the US Geological Survey 598
Sear D A 1994 River restoration and geomorphology
Aquatic Conservation 4 16977
Sear D A, Wilcock D, Robinson M R and Fisher K 2000
Channel modifications and impacts in Acreman M ed
The changing hydrology of the UK Routledge, London 5581
Shields F D 1996 Hydraulic and hydrologic stability in
Brookes A and Shields F D eds River channel restoration: guiding principles for sustainable projects Wiley,
Chichester 2374
Sparks R E, Bayley P B, Kohler S L and Osborne L L
1990 Disturbance and recovery of large floodplain rivers Environmental Management 14 699709
Stanford J A, Ward J V, Liss W J, Frissell C A, Williams
R N, Lichatowich J A and Coutant C C 1996 A general
protocol for restoration of regulated rivers Regulated
Rivers: Research and Management 12 391413
Swart J A A, van der Windt H J and Keulartz J 2001 Valuation of nature in conservation and restoration Restoration Ecology 9 2308
Thorne C R, Allen R G and Simon A 1996 Geomorphological river channel reconnaissance for river analysis,
engineering and management Transactions of the Institute of British Geographers 21 46983
Tunstall S M, Penning-Rowsell E C, Tapsell S M and
Eden S E 2000 River restoration: public attitudes and
expectations Water and Environmental Management 14
36370
Tunstall S M, Tapsell S M and Eden S 1999 How stable
are public responses to changing local environments?
A before and after case study of river restoration
Journal of Environmental Planning and Management 42
52747
van Diggelen R, Grootjans Ab P and Harris J A 2001
Ecological restoration: state of the art or state of the science? Restoration Ecology 9 1158
Ward J V, Tockner K, Uehlinger U and Malard F 2001
Understanding natural patterns and processes in river
corridors as the basis for effective river restoration Regulated Rivers: Research and Management 17 31123
Whalen P J, Toth L A, Koebel J W and Strayer P K 2002
Kissimmee River restoration: a case-study Water Science
and Technology 45 5562
Wissmar R C and Beschta R L 1998 Restoration and management of riparian ecosystems: a catchment perspective Freshwater Biology 40 57185
Wyant J G, Meganck R A and Ham S H 1995 A planning
and decision-making framework for ecological restoration Environmental Management 19 78996
Wynne B 1993 Public uptake of science: a case for
institutional reflexivity Public Understanding of Science 2
3217

You might also like