23
AQUATIC CONSERVATION: MARINE AND FRESHWATER ECOSYSTEMS Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998) Quality assessment using Ri7er Habitat Sur7ey data P.J. RAVEN a, *, N.T.H. HOLMES b , F.H. DAWSON c and M. EVERARD d a En6ironment Agency, Rio House, Waterside Dri6e, Aztec West, Almondsbury, Bristol BS32 4UD, UK b Alconbury En6ironmental Consultants, The Almonds, 57 Ramsey Road, Warboys, Huntingdon, Cambridgeshire PE17 2RW, UK c NERC -Institute of Freshwater Ecology, Ri6er Laboratory, East Stoke, Wareham, Dorset BH20 6BB, UK d The Natural Step, Thornbury House, 18 High Street, Cheltenham, Gloucestershire GL50 1DZ, UK ABSTRACT 1. A representative sample of habitat features from rivers in the UK and Isle of Man has been generated by River Habitat Survey (RHS) during 1994 – 1997. This baseline reference provides a sound basis for describing the physical character and assessing the habitat quality of 500 m lengths of river shown on 1:250000 scale maps and classified for water quality purposes. 2. The use of a standard field method, with associated accreditation controls, stratified random sampling strategy and computer database all provide a robust foundation for habitat quality assessment. The outputs have a sound statistical basis and satisfy the practical needs of river management as well as providing policy-makers with relevant information. 3. Within the UK, sites of national, regional, or local importance for wildlife habitat quality can be determined, using criteria based on the presence of features of known conservation interest. The occurrence, individually, of rare features, or rare combinations of representative features, can also be used to identify sites of particular regional or local importance for their river habitat quality. 4. A habitat quality assessment (HQA) scoring system, based on features considered to be of wildlife importance can be used to compare sites surveyed by the RHS method. However, for meaningful results this comparison must involve the same river type. 5. For completeness, the HQA scores for a given river type can be calibrated using known top quality sites. For individual sites, HQA scores should be used in conjunction with a measure of artificial modification to the channel. A simple Habitat Modification Score (HMS) system can be applied to RHS data for this purpose. 6. RHS can be further developed to define and predict the likely distribution of sites which satisfy the known habitat requirements of certain aquatic and riparian species. It can also provide the basis for better understanding of the spatial relationships between geomorphological processes and habitat type. © 1998 John Wiley & Sons, Ltd. KEY WORDS: RHS; conservation assessment; river habitats, river quality; wildlife habitat INTRODUCTION The basis for the sustainable management of rivers is (i) the availability of good quality information and (ii) professional judgement based on sound science. Organizations such as the Environment Agency, * Correspondence to: Environment Agency, Rio House, Waterside Drive, Aztec West, Almondsbury, Bristol BS32 4UD, UK. CCC 1052–7613/98/040477 – 23$17.50 © 1998 John Wiley & Sons, Ltd. Received 20 August 1997 Accepted 10 April 1998

Raven et al

Embed Size (px)

Citation preview

Page 1: Raven et al

AQUATIC CONSERVATION: MARINE AND FRESHWATER ECOSYSTEMS

Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Quality assessment using Ri7er Habitat Sur7ey data

P.J. RAVENa,*, N.T.H. HOLMESb, F.H. DAWSONc and M. EVERARDd

a En6ironment Agency, Rio House, Waterside Dri6e, Aztec West, Almondsbury, Bristol BS32 4UD, UKb Alconbury En6ironmental Consultants, ‘The Almonds’, 57 Ramsey Road, Warboys, Huntingdon,

Cambridgeshire PE17 2RW, UKc NERC-Institute of Freshwater Ecology, Ri6er Laboratory, East Stoke, Wareham, Dorset BH20 6BB, UK

d The Natural Step, Thornbury House, 18 High Street, Cheltenham, Gloucestershire GL50 1DZ, UK

ABSTRACT

1. A representative sample of habitat features from rivers in the UK and Isle of Man has beengenerated by River Habitat Survey (RHS) during 1994–1997. This baseline reference provides asound basis for describing the physical character and assessing the habitat quality of 500 m lengthsof river shown on 1:250000 scale maps and classified for water quality purposes.

2. The use of a standard field method, with associated accreditation controls, stratified randomsampling strategy and computer database all provide a robust foundation for habitat qualityassessment. The outputs have a sound statistical basis and satisfy the practical needs of rivermanagement as well as providing policy-makers with relevant information.

3. Within the UK, sites of national, regional, or local importance for wildlife habitat quality canbe determined, using criteria based on the presence of features of known conservation interest. Theoccurrence, individually, of rare features, or rare combinations of representative features, can alsobe used to identify sites of particular regional or local importance for their river habitat quality.

4. A habitat quality assessment (HQA) scoring system, based on features considered to be ofwildlife importance can be used to compare sites surveyed by the RHS method. However, formeaningful results this comparison must involve the same river type.

5. For completeness, the HQA scores for a given river type can be calibrated using known topquality sites. For individual sites, HQA scores should be used in conjunction with a measure ofartificial modification to the channel. A simple Habitat Modification Score (HMS) system can beapplied to RHS data for this purpose.

6. RHS can be further developed to define and predict the likely distribution of sites whichsatisfy the known habitat requirements of certain aquatic and riparian species. It can also providethe basis for better understanding of the spatial relationships between geomorphological processesand habitat type.© 1998 John Wiley & Sons, Ltd.

KEY WORDS: RHS; conservation assessment; river habitats, river quality; wildlife habitat

INTRODUCTION

The basis for the sustainable management of rivers is (i) the availability of good quality information and(ii) professional judgement based on sound science. Organizations such as the Environment Agency,

* Correspondence to: Environment Agency, Rio House, Waterside Drive, Aztec West, Almondsbury, Bristol BS32 4UD, UK.

CCC 1052–7613/98/040477–23$17.50© 1998 John Wiley & Sons, Ltd.

Received 20 August 1997Accepted 10 April 1998

Page 2: Raven et al

P.J. RAVEN ET AL.478

which are involved in the protection and management of water, need to operate in a coherent waythrough integrated river basin management. One of the stumbling blocks to full realization of thisapproach has been the lack of a system to characterize and classify the physical structure of rivers, neededto complement those already developed for reporting on water quality, based on chemical and biologicalsampling (Environment Agency, 1997a). Without this capability, it is difficult to set targets for habitatquality, or to measure the impact (both negative and positive) of river channel management.

In the UK, recent development of the General Quality Assessment (GQA) scheme allows for reportingon other aspects of the quality of rivers, including invertebrate biology, nutrients and aesthetic quality(Nixon et al., 1996). Since physical structure is one of the primary factors which determines the type ofaquatic biological communities present in a river, an assessment and reporting mechanism for habitats haslong been overdue.

Some form of monitoring changes in habitat quality is soon likely to be a statutory requirement. Forinstance, the draft European Framework Directive on Water Policy has a reporting requirement for thephysical as well as chemical and biological conditions of inland waters, to determine whether they areachieving ‘good ecological status’ (Commission for the European Communities, 1997). The Directive willrequire that national reporting systems adopted by EU Member States must be notified to the scientificcommunity and that the details are published. Under the Habitats Directive (92/43/EEC), EU MemberStates are required to identify and designate Special Areas of Conservation (SACs), and to preventdeterioration in their conservation status. In so doing, there is a monitoring requirement with a view tomaintaining, or where necessary restoring, ‘favourable conservation status’. The ecological quality ofrivers within SACs needs to be reported as part of this process.

In the near future, local planning authorities in the UK will need to judge whether or not developmentproposals will have a significant effect on the environment. This is a requirement of the EU Environmen-tal Impact Assessment Directive (85/337/EEC), as amended by EU Directive 97/11/EC. Inevitably, somedecisions will involve proposals that alter the physical structure of river channels. The basis fordetermining ‘significant effect’ therefore needs to take full account of available information and qualitymeasures. The UK Biodiversity Action Plan places significant emphasis on the need to monitor and reporton biodiversity, and a number of the key species identified as requiring priority action depend on rivers(Department of the Environment, 1994); using physical structure as a surrogate indicator of biodiversityis likely to feature as part of the reporting on key habitats such as chalk rivers (Department of theEnvironment, 1995). The physical structure of rivers could also be used as a measure of sustainability inthe context of maintaining or restoring conservation interest and reversing the effects of habitatfragmentation in the wider countryside.

In the past, different river management needs have generated the development of separate assessmentmethods associated with aquatic biology (Wright et al., 1994, 1998), flow requirements of fish and otheraquatic life (Bovee, 1982), and habitat requirements of salmonid fish populations (Milner et al., 1993,1998). More recently, information from these and other sources have been used to assess conservationvalue at a catchment or sub-catchment scale, using SERCON (Boon et al., 1997, 1998).

This paper considers how River Habitat Survey (RHS) data can be used to assess the physical characterand habitat quality of rivers in the UK and identifies some practical applications of the system. Thepotential synergies between RHS and other methods are discussed elsewhere (Raven et al., 1998a).

Definitions

For the purposes of this paper, the following definitions are used: ‘site’ is a 500 m length of river surveyedby the standard RHS method; ‘RHS reference sites’ are those specifically surveyed to establish a baselineRHS reference network; ‘river type’ is a descriptive term for rivers of similar physical character; ‘feature’is a distinctive, readily recognized physical object or form recorded during an RHS survey (Environment

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 3: Raven et al

QUALITY ASSESSMENT USING RIVER HABITAT SURVEY DATA 479

Agency, 1997a). A semi-natural channel is defined by the absence of any artificial physical modificationto the river bed and banks in at least nine out of 10 spot-checks surveyed (i.e. 90%) within a site and aHabitat Modification Score of 2 or less (Appendix B). Land use and biological factors such as alienspecies are not accounted for in this definition of semi-natural. This means that a semi-natural channelcould support non-native plant species and be bordered by exotic coniferous plantation.

BACKGROUND

River Habitat Survey

RHS is a system for assessing the character and habitat quality of rivers based on their physical structure(Raven et al., 1997, 1998b). It has four distinct components: (i) a standard field survey method; (ii) acomputer database, for entering results from survey sites and comparing them with information fromother sites throughout the UK and Isle of Man; (iii) a suite of methods for assessing habitat quality; and(iv) a system for describing the extent of artificial channel modification.

Habitat quality is determined by the occurrence and diversity of habitat features of known value forwildlife, and is derived by comparing observed features at a site with those recorded at other sites fromrivers of similar character. High quality is determined by habitat features occurring at sites in apredominantly unmodified physical state.

The RHS field method is a systematic collection of data associated with the physical structure ofwatercourses. Data collection is based on a standard 500 m length of river channel. Map information iscollected for each site and includes grid reference, altitude, slope, geology, height of source and distancefrom source. During the field survey, features of the channel (both in-stream and banks) and adjacentriver corridor are recorded (Table 1). In all, more than 200 compulsory data entries are made at each site,in the form of the presence, absence and (in some case) extent of specific features, collectively building acomprehensive picture of habitat diversity and character (Environment Agency, 1997b). A full descriptionand rationale for the survey method can be found in Fox, et al. (1998). Both the map-derived and fielddata are computerized, thus allowing easy access to a database, and rapid analysis of the informationcollected.

To test for consistency, 38 RHS sites were visited by both experienced and novice surveyors.Comparing the results from individual data entries established the variation in the recording of features

Table 1. The main features recorded during an RHS survey

In sweep-upAt 10 spot-checksFeatures recorded

Predominant valley form ãPredominant channel substrate ãPredominant bank material ãFlow type(s) and associated features ã ã

ããChannel and bank modificationsBankface and banktop vegetation structure ã

ããChannel vegetation typesãBank profile (unmodified and modified)ãBankside trees and associated features

Channel habitat features ã ããArtificial features ããFeatures of special interest

Land use ã ã

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 4: Raven et al

P.J. RAVEN ET AL.480

by individual surveyors and this information was used to improve the survey forms, training methods andalso to determine the RHS accreditation test (Fox et al., 1998).

Channel substrate, habitat features, aquatic vegetation types, the complexity of bank vegetationstructure and the type of artificial modification to the channel and banks are all recorded at each of 10‘spot-checks’ located at 50 m intervals. The recording format is simple, and a two-letter abbreviation foreach feature is used. These abbreviations are included both on the form and a laminated spot-check key,acting as a prompt for the surveyor. A ‘sweep-up’ checklist is also completed to ensure that features andmodifications not occurring at the spot-checks are recorded (Table 1). Cross-section measurements ofwater and bankfull width, bank height and water depth are made at one representative location to provideinformation about geomorphological processes acting on the channel. The number of riffles, pools andpoint bars found in the site is also recorded.

Establishing a baseline of reference sites

An early requirement for RHS was to establish, for England and Wales in the first instance, arepresentative baseline sample of river habitat features, collected in a consistent and repeatable fashion.This was achieved by surveying a network of reference sites based on a stratified random sample of thoserivers classified for water quality purposes. The RHS reference sites were selected independently ofexisting chemical and biological sampling points, because the latter are located, for purely practicalreasons in a non-random manner (Nixon et al., 1996). Data from the RHS reference sites now provide,for the first time, a geographically representative sample of habitat features from rivers and streamsclassified for water quality throughout the UK and Isle of Man. As a result, any 500 m length of riversurveyed in the UK using RHS can be categorized and its habitat quality assessed, by comparing it withother sites of similar physical character.

As with other UK-based surveys, the Ordnance Survey 10×10 km grid squares were used as asampling framework (Gibbons et al., 1993). For convenience, however, all coastal squares with less than50% of land area above high water mark were omitted from the baseline reference sample. The originalfocus for RHS development was England and Wales, and a 3-year sampling period was planned. Threeindividual RHS sites in each of the 1523 qualifying 10 km squares were sampled during 1994–96, one ineach square in successive years, giving 4569 sites in all. Rivers indicated on 1:250000 scale topographicalmaps qualified for inclusion in the network, but tidal reaches and canals were specifically excluded. Siteswere located on the basis of random selection of tetrads (2×2 km) within each qualifying 10 km square.The main qualifying criterion was that the watercourse had been classified for water quality, as indicatedby the 1985 River Quality Map based on the National Water Council classification (National WaterCouncil, 1981). Where no such classified watercourses existed within a 10 km square, any watercoursequalified. In the three cases in England where a 10 km square did not contain a watercourse shown onthe 1:250000 scale map, the 1:50000 scale Ordnance Survey map was used to determine site location.

In 1995–96, Scotland and Northern Ireland sites were added. In Scotland, one site in each of 779qualifying squares was sampled over a 2 year period in 1995 and 1996. Because some 10 km squares wereinaccessible by road, site selection had the added practical requirement of being within 2 km of a vehicletrack. In Northern Ireland, one site in each of 133 qualifying squares was sampled in both 1995 and 1996,giving 266 sites in all. To extend the picture, three RHS sites in each of six qualifying squares on the Isleof Man were sampled in 1997. The total length of river classified for water quality in the UK, andtherefore the population sampled, is approximately 85000 km.

Habitats as features

The main driving force behind the development of river channel features is stream energy (Frissell et al.,1986). Key aspects related to this are site elevation, channel slope, height of site compared with height of

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 5: Raven et al

QUALITY ASSESSMENT USING RIVER HABITAT SURVEY DATA 481

source, distance from source, valley form, and erosive force as measured by ‘flashiness’ or, inversely, flowstability as indicated by a high groundwater to surface run-off ratio. The physical character of anunmodified channel will principally be determined by: (i) erosion/transportation forces (energy); and (ii)erodibility (or resistance to erosion) of bed and bank material (Newson, 1997). Together, energy andresistance to erosion determine channel shape, planform, and the frequency and spatial occurrence oferosional and depositional features. The frequency of riffles, pools, point bars and eroding banks are ofprime importance in determining, in a scale-dependent fashion, the type, distribution and abundance ofaquatic wildlife communities, as do features such as in-stream and bankside vegetation, including trees(Harper and Everard, 1998). The attributes recorded by RHS capture structural variation relevant toorganisms across a range of scales from diatoms and aquatic macroinvertebrates to birds and mammals(Ormerod et al., 1997). Riparian wetland features indicate that the hydrological links between the channeland river corridor are still intact. Not all sites will have the potential for riparian wetland features, andsteep-gradient and gorge reaches may lack such obvious evidence of connectivity. High habitat value inthese areas will be represented by such features as flushes, wet heath, blanket bog, or ancient nativewoodland (Ratcliffe, 1977).

THE BASIS FOR QUALITY ASSESSMENT

The RHS system is not a conservation classification per se, but a method which allows the physicalcharacteristics of sites to be compared, and habitat quality to be assessed using criteria derived either fromknown conservation interest or the occurrence of specific features recorded within the baseline referencenetwork. It is independent of the process used to select river Sites of Special Scientific Interest in the UK(Boon, 1991).

Habitat quality assessment using RHS data can be achieved by four broad approaches: (i) astraightforward rule-based separation to identify the very best (outstanding) sites; using habitat featureswhich (ii) singly, or (iii) in combination, are rare; and (iv) a scoring system.

The basis for assessing habitat quality is as follows:

� evaluation is determined at site (500 m) level;� quality is based on the presence of channel and river corridor features which are known to be of value

to wildlife;� the two main factors which determine habitat quality are the diversity and ‘naturalness’ of physical

structure;� the system is calibrated, wherever possible, using known top quality sites surveyed specifically for this

purpose.

RHS quality assessment does not take account of naturalness in the strictly biological sense. Habitatfeatures are recorded as physical structures, and vegetation categories are recorded irrespective of whetherthey contain native or non-native species.

Outstanding sites—separating the best from the rest

Given the priority afforded to naturalness as the primary criterion for wildlife conservation interest, thereare two simple rules for identifying an outstanding RHS site: one is that the river channel is pristine, thatis, totally free from artificial modification; the second is that land use in the river corridor is entirelysemi-natural. Examples of semi-natural land use in the UK include broadleaf woodland, native pinewood,peatbog, fen or whatever represents the climax vegetation (Ratcliffe, 1977). Since there are significantregional differences in semi-natural vegetation, outstanding sites need to be confirmed on advice from

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 6: Raven et al

P.J. RAVEN ET AL.482

Table 2. Some selection criteria for deriving river type

Primary attributes Secondary attributes

Solid geology Above or below tree-lineDrift geology Planform of channelAltitude Distance to on-line lakeChannel slope (gradient) Valley formDistance from source Size (water width, mean discharge)Height of source Baseflow index

relevant conservation experts. The ability to determine outstanding sites on this basis, regardless of rivertype, provides a valuable tool for the planning process, enabling the best sites in whatever context to beidentified and appropriate protection measures taken.

River type as a basis for comparing habitat quality

During the initial development of RHS, attempts were made to devise a single, fixed national classifica-tion of river types for assessing habitat quality. Eleven preliminary river segment types proved unwork-able for this specific purpose, but nevertheless provide a useful basis for describing overall river character(National Rivers Authority, 1996). Given the difficulties in trying to apply a fixed classification of rivertypes to the continuum of natural variation in the physical character of channels, a more practical andflexible approach is required. This means interrogating the RHS database using a set of rules, or selectioncriteria, to identify those reference sites with the most similar attributes. The criteria chosen will vary,depending on the purpose of the quality assessment exercise, but may be based on map-derived variablessuch as altitude, geology, and height of source. If necessary, field generated measurements such as channelwidth can also be used (Table 2). The resulting group of sites represents a sample of the ‘river type’determined by the selection rules or criteria used. Some descriptive examples appear in Table 3. The moregeneral the selection rules, the less precisely defined the river type will be and, because of the naturalvariability involved, the lower the confidence for comparing either general characteristics or measures ofhabitat quality.

Table 3. Some descriptive examples of river types in the UK and their occurrence in the RHSreference network

River type Present in RHS reference network

Winterbournes Very fewYesChalk rivers

Groundwater-fed sandstone rivers YesVery fewPeatbog streamsYesSteep streams

Upland plateau rivers YesLimestone rivers Very fewGorge rivers YesCoastal streams Very few

YesMountain valley riversSmall, lowland riffle-dominated rivers YesClay rivers Yes

YesInter-lake (lochan) riversLarge lowland rivers Yes

YesDrains and dykes

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 7: Raven et al

QUALITY ASSESSMENT USING RIVER HABITAT SURVEY DATA 483

Table 4. Natural habitat features with an occurrence of 5% or less in upland and lowland RHS reference sites. Insufficient uplandsites sur6eyed in Northern Ireland for meaningful comparison

Feature UK and Isle of Man England and Wales Scotland Northern Ireland

Upland Lowland Upland Lowland Upland Lowland Lowland

Gorge ã ã ã ã ã ãBraided/side channels ã ã ãWaterfalls \5 m high ã ã ã ãExtensive exposed ãã ã ã

bankside tree rootsExtensive underwater ã ã ã ã ã ã

tree rootsExtensive coarse woody ãã ã ã ã ã ã

debrisExtensive fallen trees ã ã ã ã ã ã ã

Carr ã ã ã ã ã ã ãExtensive broadleaf ã

woodland on bothbanks

Extensive wetland on ã ã ã ã ã ãboth banks

Upland defined as more than 200 m altitude to the north and west of a line joining Start Point and Flamborough Head.

Rarity as a quality measure

Sites of high habitat quality can be determined using rarity as the primary criterion. This can be appliedto the occurrence of one or more rare features of known wildlife value, or a rare combination ofrepresentative features which individually may not be scarce. Regardless of the approach, rarity needs tobe based on the occurrence of a feature or a combination of features in the baseline RHS referencenetwork. Given the significant regional variations in landscape character across the UK, it is essential thatthe geographical context for rarity is explicit. A rare feature in the Midlands of England may be quitecommon in Wales or Scotland, but nevertheless deserves special attention when putting site value into aregional or local context.

Table 4 lists a selection of natural features occurring in 5% or less of upland and lowland RHSreference sites overall, and individually for England and Wales, Scotland and Northern Ireland. Otherfeatures occur at 5% or less, of reference sites, but cannot be considered rare in a wider context. Similarlists can be derived for any region of the UK. Table 5 illustrates how combinations of representativefeatures can be used to define rarity for a given river type, in this instance, steep streams. This exampleindicates that, whilst 91.4% of all RHS reference sites of this type have exposed boulders, only 11.0% ofsites have the combination of exposed boulders, bedrock, cascades, waterfalls and extensive broadleafwoodland (or native pinewood) on both banks which together are considered special. This approach canbe applied to any river type in an overall, regional, or even a local context and is likely to be the mostuseful day-to-day method for assessing the relative habitat quality of a particular site.

An alternative, but similar, approach could be used, with sites only qualifying if minimum levels forhabitat features are met. For example, continuous or semi-continuous trees along both banks andextensive broadleaf woodland could be used as two tree-related quality criteria for identifying how manysites qualify for a particular habitat requirement.

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 8: Raven et al

P.J. RAVEN ET AL.484

Habitat Quality Assessment (HQA) score

Scoring is a simplistic instrument for describing any ecological system, especially complex and dynamicsystems such as rivers. Whilst it does have its detractors, it provides a useful way for evaluating therelative quality of a site and the potential impact of proposed management. Scoring can be used toquantify (i) improvement or degradation of habitat quality; and (ii) how an anticipated change couldaffect a particular habitat resource at catchment, regional or national level. The habitat quality assessment(HQA) scoring system, described in Appendix A, is basically a broad measure of the diversity and‘naturalness’ of physical (habitat) structure in the channel and river corridor. The HQA score of a site isdetermined by the presence and extent of habitat features of known wildlife interest recorded during thefield survey. Rare features nationally such as waterfalls more than 5 m high and extensive fallen treesattract additional points.

The current HQA system is based on expert opinion and has been developed primarily for use inEngland and Wales. It needs to be tested, refined and adapted to take account of regional variationsacross the UK and beyond. For example, the special quality of pristine rivers and streams occurring innaturally treeless mountain areas above 700 m, blanket bogs, coastal machair and other wind-stressedlandscapes found in parts of Scotland are not adequately accounted for in the current version.

Point scoring for the HQA system is based on a consensus of informed professional judgment. It issubjective, but provides the necessary consistency for comparisons. Features that score are consistent withthose included in the ‘System for Evaluating Rivers for Conservation’ (SERCON), for which a panel ofecological experts identified the attributes of most value to riverine wildlife (Boon et al., 1997). However,given the limitation of scoring, the system may not always identify immediately those sites which have themost extensive or very best examples of habitat features. The HQA score should therefore not be usedalone to determine management action. Although independent of river type, comparison of individualHQA scores has to involve sites of similar river type, determined in turn by specific selection rules andhence the purpose of the exercise. Comparison of HQA scores across different river types is notmeaningful.

Calibration using top quality sites

During 1994, the first survey year, it became clear that very few RHS reference sites in England and Waleshad a combination of totally unmodified channels and an extensive semi-natural landscape. Anticipatingthat some of the very best sections of river would not be included in the full set of reference sites, it wasdecided to establish a ‘top quality’ series of benchmark sites by undertaking special RHS surveys of

Table 5. An example of determining the quality of steep streams, using combinations of habitat features

Features Percentage of steep stream RHS reference sites withrequired attributes (n=336)

91.4Exposed bouldersExposed boulders+bedrock 67.9Exposed boulders+bedrock+cascades 64.3Exposed boulders+bedrock+cascades+waterfall(s) 32.7Exposed boulders+bedrock+cascades+waterfalls 11.9

+extensive moorland/heath on both banksExposed boulders+bedrock+cascades+waterfalls 11.0

+extensive broadleaf woodland on both banks

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 9: Raven et al

QUALITY ASSESSMENT USING RIVER HABITAT SURVEY DATA 485

additional rivers of known high wildlife importance. Most of the benchmark sites were selected for surveybecause existing information confirmed their high nature conservation value in terms of plant or animalcommunities, although this did not always mean that the river habitat was of high value. Otherbenchmark sites, particularly in lowland England, needed to be traced using map-based information as apreliminary guide. In addition to an RHS survey, each benchmark site had a full macrophyte surveycompleted and water chemistry analysed. These data and HQA scores are included as a special benchmarksection in the RHS database and provide a quality marker for certain river types. Full calibration of HQAscores is only possible for those river types which include one or more benchmark sites by their selectionrules.

It was remarkably difficult to find 500 m lengths of river in the lowlands which were suitable forbenchmarking in the strictest sense. Such is the extent of human influence in the UK, that there are hardlyany pristine 500 m lengths of large lowland river flowing through a semi-natural landscape. Somebenchmarks are therefore some way short of natural, but they nevertheless represent high habitat qualityin the context of the structural degradation of particular river types.

Classifying habitat quality

Many management decisions require that the quality of sites is classified. The links between RHS site,river type, HQA scores and other quality assessment outputs are illustrated in Figure 1. River habitatquality can be classified in response to specific questions, such as:

� is the site outstanding?� is the site important within a specified geographical context due to one or more rare features, or a rare

combination of features?how do features in the site, such as the number of riffles or extent of trees,compare with other sites of the same river type?

� how does the site HQA score compare with others of similar type, and with top quality benchmarkexamples?

Classification of habitat quality using RHS is the grouping of sites with similar attributes or scores.Great care is needed when categorizing sites as ‘excellent’, ‘good’, ‘fair’ or ‘bad’ and this approach, usedfor describing water quality, is helpful only for reporting purposes and informing broad managementpolicies. As a general rule, habitat quality assessment, however generated, should be used in conjunctionwith other available information such as species data if the wider conservation value of a site needs to bedetermined. This is the basis for using RHS data to help evaluate the overall conservation interest ofrivers through the SERCON system (Boon et al., 1997).

Using RHS to assess artificial channel modification

River habitat quality is strongly influenced by the type and extent of artificial modification. By applyinga simple set of rules to RHS data, artificial modification to the physical structure of the channel can beexpressed as a Habitat Modification Score (HMS). Like the HQA score, the system is an objectiveapplication of a set of subjective rules, and is necessary for consistent comparison between sites. It canalso be used at a site level to audit predicted or actual impacts resulting from channel works. Pointsscored are based on the relative impact of modification on habitat features. By assigning a score of 1 toeach spot-check entry for resectioning, 2 for reinforcement and allowing for other types of modification,such as weirs, to be accounted for as well, a cumulative HMS score can be used to summarize the severityand extent of structural alteration to the channel (Appendix B).

Using the HMS system, sites with a pristine channel, having no artificial modification, score zero.Semi-natural channels score a maximum of 2, while the most heavily and extensively modified channelsscore 45 or more (Table 6). The HMS score is independent of river type, so it can be used to describe

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 10: Raven et al

P.J. RAVEN ET AL.486

Figure 1. The links between site quality assessment and river type, HQA and HMS scores.

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 11: Raven et al

QUALITY ASSESSMENT USING RIVER HABITAT SURVEY DATA 487

artificial modification to channel structure across the board. However, biological factors such as thepresence of non-native plant species are not included in the scoring system. In describing individual sites,both the HQA and HMS scores should be used in conjunction, because together they give a broadindication of how overall habitat quality and structural modification to the channel might be linked.

PRACTICAL APPLICATIONS OF RHS

Reporting

Although development of a UK-wide database has been a major achievement in itself, the ultimate testfor the RHS system is in its practical applications. A national overview of the physical state of rivers inthe UK, based on the RHS baseline reference network has now been published (Raven et al., 1998b). Byway of example, habitat quality and channel modification associated with chalk rivers, using HQA andHMS scores are shown in Figure 2. The broadly bell-shaped distribution frequency of HQA scores for anumber of river types confirms that the current system provides a good basis for classifying sites on thebasis of habitat quality (Raven et al., 1998b).

Catchment management plans

Catchment management plans provide an important framework for delivering integrated river basinmanagement (Werritty, 1997). RHS has already started to add another dimension to these by providing:

� a descriptive framework for describing the physical character, quality and modification of rivers;� a basis for setting habitat-related targets and measuring the performance of river management;� a link with other measures of quality so that river management can take full account of all users’

requirements.

The geographical coverage of RHS baseline reference sites provides a useful basis for the generaldescription of river character at the regional level or above, but extra survey work may be necessary forindividual catchments, depending on the level of detail required. An optimum catchment-based samplingstrategy, using RHS surveys every 2 km (a frequency of one survey every fourth 500 m), is recom-mended—a conclusion drawn from analysing a complete survey of the River Wyre in Lancashire(Sansbury, 1994). Entire lengths of other mainstream rivers in the UK have also been surveyed, providinga complete RHS assessment of physical character, quality and modification which has been used inconjunction with other information for catchment management purposes.

Table 6. Habitat modification score (HMS) categories for describing the physical state of the riverchannel at RHS sites

HMS score Descriptive category of channel modification

Pristine00–2 Semi-natural3–8 Predominantly unmodified9–20 Obviously modified

Significantly modified21–44Severely modifiedMore than 45

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 12: Raven et al

P.J. RAVEN ET AL.488

Figure 2. The frequency distribution of HQA and HMS scores for chalk rivers in the RHS reference network.

Environmental impact assessment

A key requirement of environmental impact assessment is determining the current value of a particularsite and predicting the ecological changes likely to be caused by proposed works. This is particularlyimportant in deciding whether a proposal is likely to have a ‘significant effect’ on the environment, inaccordance with the EU Environmental Impact Assessment Directive (85/337/EEC). RHS can help toestablish the relative importance of a site in habitat quality terms. Decisions on whether a fullEnvironmental Impact Assessment is needed, together with broad policies regarding protection, mitiga-tion and enhancement can then be related to this and other quality assessments.

The three main types of physical modification affecting river channels are: reinforcement (revetment inthe form of concrete, steel piling, gabion, rip-rap, etc.); resectioning (reprofiling through dredging of the

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 13: Raven et al

QUALITY ASSESSMENT USING RIVER HABITAT SURVEY DATA 489

bed and banks); and regulation of flow by impounding structures. The impact of these modifications maywell influence the occurrence of habitat features for a considerable distance downstream or, in the case ofmajor impoundments, upstream as well.

Both HQA and HMS scores can be used to assist in ‘before and after’ appraisals of physical alterationsto the river channel and adjacent land. Existing HQA scores, reflecting features such as trees, underwatertree roots, marginal deadwater, or riparian wetland, provide a simple numerical baseline against whichlosses can be anticipated and real changes measured. This provides both a management tool forimproving decision-making and a means of auditing decisions and the resulting consequences. It isimportant to note that RHS cannot replace other specific assessment methods, but can provide aconsistent framework for deciding what, if any, other more detailed studies (such as geomorphological orbotanical survey work) are required. The RHS database allows the relationship between features andmodifying factors at both site and individual spot-check level to be analysed. Analyses can be carried outon a whole range of features, comparing their occurrence in impacted and semi-natural channels of thesame river type. The following examples illustrate how selected habitat features in small, lowland,riffle-dominated rivers in the UK can be affected by artificial modification. Geomorphological studiessuggest that riffles in this type of river develop at a modal spacing equivalent to 5–7 bankfull channelwidths (Ferguson, 1981). This is reflected by RHS reference sites with a semi-natural channel structure(Figure 3); however, modification affecting at least a third of channel length leads to increased rifflespacing and, by implication, an associated reduction of in-stream habitat diversity.

Individual types of modification also reduce habitat structure. Bank resectioning is mainly associatedwith land drainage and flood defence works, and involves the mechanical reprofiling of the river toproduce a larger, more uniform cross-section to allow faster passage of flood flows. There are clearadverse impacts on the diversity of both in-stream and bank habitats when resectioning is extensive.Features such as riffles, point bars, and tree cover are all affected (Figure 4). Extensively resectioned sitesalso have a greater preponderance of silt and less gravel/pebble size substrate than semi-natural channels.The impact on point bars, a habitat important for certain invertebrates, is significant since these featuresoccur on the inside of meander bends. Resectioning is often associated with channel straightening, which

Figure 3. Riffle-spacing, expressed as bankfull widths, in semi-natural and extensively modified channels of small, lowland,riffle-dominated rivers. (See Raven et al., 1998b for derivation.)

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 14: Raven et al

P.J. RAVEN ET AL.490

Figure 4. The impact of extensive bank resectioning, bank reinforcement and channel impoundment on five selected features ofsmall, lowland, riffle-dominated rivers: percentage variation between semi-natural and extensively modified sites.

reduces the number of meanders and consequently the occurrence of point bars. The implications for rivermanagement are clear, and confirm the need for techniques such as working from one bank to minimizeenvironmental impact (RSPB et al., 1994).

Reinforcement is used to protect all or part of a bank from erosion. Various materials can be used,depending on the level of protection needed. Concrete, sheet piling, bricks, stone, rip-rap and rock-filledgabion baskets are used for ‘hard’ reinforcement, and these have a clear adverse impact on point bars andbankside trees compared with semi-natural channels (Figure 4). Channel impoundment occurs for a variety

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 15: Raven et al

QUALITY ASSESSMENT USING RIVER HABITAT SURVEY DATA 491

of purposes, including to manage water levels of surrounding land, control flooding, abstract water andto diversify the in-stream habitat of featureless reaches. The impact depends both on the height of theimpounding structure and the channel gradient. In sites with one or more weirs, there is a reduction inthe occurrence of riffles and point bars. The occurrence of silt in the channel is noticeably greater (Figure4).

Habitat enhancement

Management techniques to enhance the structural diversity of degraded river channels is now well-estab-lished (RSPB et al., 1994; Hansen et al., 1998). The RHS database can help to assess likely changes infeatures and habitat quality associated with enhancement and rehabilitation measures. In some instances,well-intentioned enhancement works, such as riffle reinstatement or pool excavation, can be inappropriatefor a particular location. Gravel imported to form ‘riffles’ can be left high and dry, or, in other instances,swept away by major spates, if located in the wrong place. The risk of rehabilitation failure can bereduced by providing information on channel character and behaviour. Decisions on the reintroductionof features such as riffles can be made more confidently using the RHS database, because the pattern oftheir occurrence in other sites of the same river type can be determined. In this way, sustainable andeffective solutions can be more readily identified.

Further analysis of RHS data should be able to quantify better which management practices arebeneficial for wildlife habitats, either through minimizing the loss of natural features or improvingdegraded reaches. However, RHS alone cannot produce the answers. In all cases, RHS informationshould be used with other available data to guide decisions. Geomorphological principles also need to beapplied, taking full account of catchment characteristics, historical influences and those sedimentaryprocesses which are operating at the reach level (Newson, 1997).

DISCUSSION

There have been numerous attempts worldwide to classify rivers with a view to conservation management,but a recurrent problem is one of scale, with different processes influencing features from the catchmentto microhabitat level (e.g. Frissell et al., 1986; Mitchell, 1990; Kershner and Snider, 1992; Naiman et al.,1992; Hawkins et al., 1993; Rosgen, 1994). The pitfalls and limitations of classification schemes imposingarbitrary divisions in what is essentially a continuum such as a river system, are comprehensively exploredby Kondolf (1995). Some attempts to classify rivers according to their environmental or ecological qualityhave incorporated biological, physical and chemical parameters to derive a single score (e.g. Mitchell,1990; Bauer, 1992; Verniers, 1992; Schneiders et al., 1993). This approach can fail because of the inherentcomplexities involved, particularly when different parameters give conflicting quality measures.

Until now, no other method has generated the necessary baseline reference information required for anobjective assessment of river habitat quality on a national scale. The RHS method has produced apractical but robust system by concentrating on physical structure, using a standard sample unit length,and generating a representative sample of habitat features for the population of rivers in the UK classifiedfor water quality purposes. The level and type of information generated by RHS map and field dataallows for comparison of the occurrence of specific individual features, or combinations of featuresbetween sites of similar physical character. RHS data can also be used for ordination analysis to establishrelationships between map-derived variables and habitat features recorded in the field (Jeffers, 1998).

Comparison between RHS sites can be rule-based, using selection criteria, or for broad managementpurposes, score-based. The use of site-based HQA and HMS scores provides an insight into the linkbetween the ‘diversity’ of physical structure and the extent of artificial constraining factors. This will vary

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 16: Raven et al

P.J. RAVEN ET AL.492

depending on the type of river involved and its response to differing intensities of management.Comparison of actual and potential HQA scores can also provide an insight into the enhancementpotential of physically degraded rivers and help to set restoration objectives. It also provides the basis formeasuring achievement of those objectives.

Requirements for wildlife conservation vary according to the species or community concerned.Interrogation of the RHS database can reveal, at site level, the occurrence of key habitats which representhigh quality for certain taxa: for example, exposed riverine sediments, coarse woody debris, andunderwater tree roots are important for various aquatic invertebrate taxa: eroding river-cliffs are essentialfor nesting sand martins Riparia riparia L. and kingfishers Alcedo atthis L.; and certain fish and aquaticinvertebrates require particular flow conditions (Harper and Everard, 1998). Provided that habitatrequirements and other needs (such as water chemistry) are well established, RHS-based criteria can beused to help predict the occurrence of particular species. An example of this approach using nativecrayfish Austropotamobius pallipes (Lereboullet) is described by Naura and Robinson (1998), whilstdetailed studies of the habitat preferences of upland river birds using RHS data have been explored byBuckton and Ormerod (1997).

By determining the necessary hydrological and ecological criteria, a classification of RHS sites can bemade on the basis of habitat requirements for particular species. This approach would parallel that ofusing combinations of representative features to assess high quality (Figure 1). In effect, the permutationssuch as ‘good kingfisher habitat’ are almost endless, although prediction will always be limited by theknowledge of species-habitat relationships, and other factors influencing distribution. Bearing this inmind, more specific microhabitat-scale information may need to be gathered in addition to RHS data.

Quality assessment using RHS data does not take into account the full spectrum of conservationcriteria, so the presence of alien plant species, for instance, does not ‘downgrade’ the HQA score. Inconservation terms, this is an important distinction, particularly since more than 20% of RHS referencesites with semi-natural channel structure in England and Wales contain Himalayan balsam Impatiensglandulifera Royle (Raven et al. 1998b). Likewise, the presence of a rare plant or animal species does notupgrade the HQA score. These factors can be taken into account by using SERCON (Boon et al., 1997)provided the relevant data have been collected in addition to carrying out the RHS survey (Wilkinson etal., 1998).

Reporting on the type and extent of artificial channel modification is theoretically far simpler, and canbe done irrespective of river type. However, the information depends on the confidence with whichsurveyors can distinguish the sometimes subtle effects of physical modification caused by both past andpresent river management. RHS training courses, surveyor accreditation tests an illustrated guidancematerial have been introduced to minimize inconsistency in what is essentially and observational methodof recording (Fox et al., 1998). There is also an inherent bias in the baseline reference network. Longculverted stretches of river do not appear on maps and are therefore excluded from the sample. In largeurban areas, such as London, where extensive culverting has occurred, this is an important historicalfactor to consider (Barton, 1992).

Quality assessment extends beyond wildlife conservation (Boon and Howell, 1997; Tunstall et al., 1997).For example, anglers, canoeists, and drainage engineers will all have their own ideas about the relativeimportance of river-related features. In each instance, selection criteria can be established to determinequality according to the particular interest group. Within the limits imposed by the level of detail offeatures recorded, the RHS database can be used to help identify those sites which can be considered ofgood quality for a particular purpose. Features representing high quality for wildlife conservation willinvariably differ, to a lesser or greater extent, from those desired by other interest groups. Anglers, forexample, require sufficient gaps in tree cover to allow for unimpaired casting, whilst low overhangingboughs, extensive coarse woody debris and debris dams will impede canoeists. Walkers need good accessalong the bank, whilst river engineers require a uniform smooth channel form where efficient, unimpededflood conveyance is critical.

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 17: Raven et al

QUALITY ASSESSMENT USING RIVER HABITAT SURVEY DATA 493

Table 7. Some current and future uses and users of RHS

How RHS has already been usedEnvironmental Impact AssessmentDescriptive statistics and habitat targets for catchment plansMonitoring and post-project appraisalNational and regional reportingLinking with SERCONPredicting the likely distribution of some aquatic species based on their habitat requirements

Future usesCalibration of sites sampled for water quality purposesTargeting and monitoring waterside agri-environment schemesBiodiversity Action Plan target-setting and monitoringEducational interactive CD-ROM

Future development and applications

The RHS system is now operational with an initial set of applications (Table 7). The emphasis now passesto the systematic development of further planning, options appraisal and operational tasks. Close linksbetween RHS and other systems, such as SERCON, have already been made, enabling a better overallassessment of rivers to be carried out. A simple and effective link is that with the chemical and biologicalassessments of river water quality. Figure 5 illustrates this link, comparing habitat modification categoriesat RHS reference sites in England and Wales with those in Scotland with good river water quality. Givenappropriate survey data, a site in the UK can now be classified according to river plant community type(Holmes et al., 1998a), chemical and biological water quality (Environment Agency, 1997a; Wright et al.,1998) HQA and HMS score (Appendix A and B), fisheries status (Mainstone et al., 1994) and othersystems currently under development, such as mean trophic ranking (Holmes et al., 1998b).

Figure 5. The proportion of habitat modification categories at RHS reference sites with good river water quality: a comparisonbetween England and Wales, and Scotland.

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 18: Raven et al

P.J. RAVEN ET AL.494

During the development phase, RHS targeted applications focused on habitat quality assessment, witha strong link to biodiversity issues. The application phase will provide a new dimension by adding similaremphasis on channel behaviour. This will mean exploring further those geomorphological aspects of riverchannel stability and sensitivity, and the role of water quantity (both high and low flows), which togetherare major considerations for sustainable river management (Ferguson, 1981).

So far, the RHS database has provided site-based outputs. There is considerable potential for usingRHS data for categorizing, on a catchment basis, the nature of direct and indirect impacts on riverchannels. It is clearly possible to look at sequences of site data within an individual catchment, or atcomparative locations between catchments. This approach is needed to investigate the relationshipbetween the ‘driving’ variables of gradient, flow and substrate and the influence of channel management,land use and other features, such as lakes and wetlands.

RHS can also be used to describe the habitat character of chemical and biological sampling points usedfor assessing water quality. Indeed, because habitat conditions influence the type of aquatic invertebratecommunities present in rivers, RHS can provide a calibration method for assessing the impact of channelmodification at biological sampling sites.

The RHS reference sites provide a baseline for monitoring changes in the physical character of riverchannels in the UK. Resurvey of a subset of the RHS reference sites could be used to help assesslong-term changes of habitat features in different river types. Targeted monitoring could also help toreveal how changes in catchment and riparian land use affect habitat quality. In the context of climatechange, alterations to river habitats might result from reduced baseflows in groundwater-fed streams onthe one hand, and increased flood peaks on the other, caused by the increased frequency of droughts andstorms, respectively.

Although RHS has been designed primarily for conservation purposes, its geomorphological contexthas considerable significance in its own right (Padmore, 1997). Analysis of overall and regional patterns,and the establishment of geomorphological benchmarks, are just two examples of further developmentpotential in this field of study. The RHS field method has been developed specifically for the scale andmanagement of rivers found in the British Isles. The precise method is not suitable for very large riversor multi-thread channels. However, the underlying principles can be applied to larger-scale continentalrivers and to channels in areas of higher, steeper relief, even if individual components such as unit samplelength, distance between spot-checks, or vegetation categories, have to be adapted to suit local conditions.

ACKNOWLEDGEMENTS

Constructive discussions and comments have been made throughout by fellow members of the River Habitat Surveyproject team. Particular acknowledgements are extended to Peter Fox, Professor John Jeffers, Professor MalcolmNewson, Professor Mike Clark, Dr Phil Boon, Professor Ron Edwards and Ian Fozzard. Professor Stephen Ormerodprovided excellent redrafting advice in his role as a referee. This paper includes material from Ri6er Habitat Quality:the physical character of ri6ers and streams in the UK and Isle of Man (Raven et al., 1998b) which the EnvironmentAgency has kindly allowed to be reproduced in this paper.

APPENDIX A. HABITAT QUALITY ASSESSMENT (HQA) SCORING SYSTEM: VERSION 1.21

The HQA score for a site is the total of all the component scores in the categories listed below.

1 HQA scores should only be used when comparing sites of similar river type or character. For instance, sites in naturally treelessexposed or mountain areas should not be compared with those in lowland wooded valleys.

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 19: Raven et al

QUALITY ASSESSMENT USING RIVER HABITAT SURVEY DATA 495

Flow types

Each predominant flow-type recorded scores 1; if it occurs at two to three spot-checks, it scores 2; if itoccurs at four or more spot-checks, it scores 3. If only one type occurs at all 10 spot-checks, the score willbe 3. Dry channel scores 0.

If recorded in the sweep-up, score 1 for each of the following channel features provided that an equivalentflow-type has not been recorded in any spot-check: waterfall(s), if free fall flow absent; cascade(s), if chuteflow absent; rapid(s), if broken standing wa6e absent; riffle(s), if unbroken standing wa6e absent; run(s), ifrippled flow absent; boil(s), if upwelling absent; glide(s), if smooth flow absent; pool(s), if no perceptibleflow absent. Score 1 for marginal deadwater recorded as present or extensive in the sweep-up.

Channel substrates

Each predominant natural substrate type (i.e. bedrock, boulder, cobble, gravel/pebble, sand, silt, clay,peat) recorded scores 1; if it occurs at two to three spot-checks it scores 2; if it occurs at four or morespot-checks, it scores 3.

If only one predominant type is recorded at all 10 spot-checks, the score will be 3.Extra substrate(s) recorded (on the 1997 form) do not count.‘Not visible’ does not score, unless recorded at six or more spot-checks, when it scores 1.

Channel features

Each ‘natural’ channel feature (i.e. exposed bedrock/boulders, unvegetated mid-channel bar, vegetatedmid-channel bar, mature island) recorded scores 1; if it occurs at two to three spot-checks, it scores 2; ifit occurs at four or more spot-checks, it scores 3. [N.B.: more than one feature can occur at a singlespot-check.]

If any of these features are not recorded in the spot-checks, but occur as present or extensive in thesweep-up, then they will score 1 each.

Bank features

Each bank is scored separately.Each natural feature (i.e. eroding earth cliff, stable earth cliff, unvegetated point bar, vegetated point

bar, unvegetated side-bar, vegetated side-bar) recorded scores 1; if it occurs at two to three spot-checks,it scores 2; if it occurs at four or more spot-checks, it scores 3. [N.B.: more than one feature can berecorded at a single spot-check.]

If any of unvegetated point bar, vegetated point bar, unvegetated side-bar or vegetated side-bar are notrecorded in the spot-checks, but appear in the sweep-up, then they will score 1 each. [N.B.: 6ertical/under-cut cliff profile recorded in the sweep-up does not equate to eroding or stable earth cliff.]

Bank vegetation structure

Only simple and complex vegetation structure score. Both score equally.Each bank is scored separately.Bankface and banktop are scored separately.

Bankface

If simple or complex is recorded at one spot-check it scores 1; if simple and/or complex recorded at twoto three spot-checks, score 2; if simple and/or complex occur at four or more spot-checks, the score willbe 3.

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 20: Raven et al

P.J. RAVEN ET AL.496

Banktop

If simple or complex is recorded at one spot-check it scores 1; if simple and/or complex recorded at twoto three spot-checks, score 2; if simple and/or complex occur at four or more spot-checks, the score willbe 3.

Point bars

Add together the total number of unvegetated and vegetated point bars recorded.Score 1 if the total is three to eight; score 2 for nine or more.

In-stream channel vegetation

In-stream channel vegetation types are grouped into six categories for scoring purposes: (i) liverworts andmosses; (ii) emergent broad-leaved herbs; (iii) emergent reeds/rushes/sedges; (iv) floating-leaved, free-float-ing and amphibious; (v) submerged broad-leaved; and (vi) submerged linear and fine-leaved.

Score 1 for each category recorded within the site, and 2 for those categories recorded either as presentor extensive at four or more spot-checks.

Filamentous algae do not score.

Land-use within 50 m

Each bank is scored separately.Only the sweep-up information is used.Only broadleaf woodland (or native pinewood), moorland/heath, and wetland score.Broadleaf woodland, moorland/heath and wetland each score 1 if present, and score 2 if extensive.If broadleaf woodland (or native pinewood) or wetland, alone or together are the only land-use

categories recorded, then score 7 for that bank. For naturally treeless sites, moorland/heath or equivalentqualifies.

Trees and associated features

Trees

Each bank is scored separately.Score 1 if trees are isolated/scattered; score 2 if regularly-spaced or occasional clumps; score 3 if

semi-continuous or continuous.

Associated features

Overhanging boughs, exposed bankside roots, underwater tree roots, coarse woody debris and fallentrees each score 1 if present.

Extensive exposed bankside roots and underwater tree roots each score 2.Extensive coarse woody debris scores 3.Extensive fallen trees score 5.

Special features

Score 5 if any of the following have been recorded: waterfall more than 5 m high, braided or side channel,debris dams, natural open water, fen, carr, flush, bog. [Score 5 regardless of number of special featurespresent.]

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 21: Raven et al

QUALITY ASSESSMENT USING RIVER HABITAT SURVEY DATA 497

APPENDIX B. HABITAT MODIFICATION SCORE (HMS) RULES: VERSION 1.1

The HMS score for a site is the total of all the component scores in the categories listed below.

A. Modifications at spot-checks Score per spot-checkReinforcement to banks 2Reinforcement to bed 2Resectioned bank or bed 1Two-stage bank modification 1Embankment 1Culvert 8Dam, weir, ford 2Bank poached by livestock 0, if less than three

spot-checks1, if three to five

spot-checks2, if six or more

spot-checks

B. Modification present but not recorded at Both banksOne bank (or channel)spot-checksArtificial bed material 1 —Reinforced whole bank 2 3Reinforced top or bottom of bank 1 2Resectioned bank 1 2Embankment 1 1Set-back embankment 1 1Two-stage channel 1 3Weed-cutting 1 —Bank-mowing 1 1Culvert 8 for eachDam, weir, ford 2 for each

C. Scores for features in site as a whole One Two or moreFootbridge 0 0Roadbridge 1 2Enhancements, such as groynes 1 2Site partly affected by flow control 1Site extensively* affected by flow control 2Partly realigned channel** 5Extensively* or wholly realigned channel** 10

* Extensive means at least a third of channel length.** Information from map.

REFERENCES

Barton, N.J. 1992. The Lost Ri6ers of London—A Study of Their Effects Upon London and Landowners, and theEffects of London and Landowners Upon Them, second edition, Historical Publications, London.

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 22: Raven et al

P.J. RAVEN ET AL.498

Bauer, H.J. 1992. ‘Methode d’evaluation multicritere de la qualitie ecologique des cours d’eau’, in Impact ofWatercourse Impro6ements, Assessment, Methodology, Management, Assistance, Proceedings of an InternationalSymposium, Wepion, Belgium, 177–191.

Boon, P.J. 1991. ‘The role of Sites of Special Scientific Interest (SSSIs) in the conservation of British rivers’,Freshwater Forum, 195–108.

Boon, P.J. and Howell, D.L. (Eds) 1997. Freshwater Quality: Defining the Indefinable?, The Stationery Office,Edinburgh.

Boon P.J., Holmes N.T.H., Maitland P.S., Rowell T.A. and Davies J. 1997. ‘A system for evaluating rivers forconservation (SERCON): development, structure and function’, in Boon, P.J. and Howell, D.L. (Eds), FreshwaterQuality: Defining the Indefinable?, The Stationery Office, Edinburgh, 299–326.

Boon, P.J., Wilkinson, J. and Martin, J. 1998. ‘The application of SERCON (System for Evaluating Rivers forConservation) to a selection of rivers in Britain’, Aquatic Conser6ation: Marine and Freshwater Ecosystems, 8,597–616.

Bovee, K. D. 1982. ‘A Guide to Stream Habitat Analysis Using the Instream Flow Incremental Methodologies’, USFish and Wildlife Ser6ice Instream Flow Information Paper No. 12, Report No FWS/OBS-82/26, USFWS InstreamFlow Group, Fort Collins, USA.

Buckton S. T., and Ormerod S. T. 1997. ‘Use of a new standardised habitat survey for assessing the habitatpreferences and distribution of upland river birds’, Bird Study, 44, 327–337.

Commission for the European Communities, 1997. ‘Proposal for a Council Directive establishing a framework forCommunity action in the field of water policy’, CEC COM(97)49, Official Journal, C184, 17 June 1997.

Department of the Environment, 1994. Biodi6ersity: the UK Action Plan, Command 2428, HMSO, London.Department of the Environment, 1995. Biodi6ersity: the UK Steering Group Report, Two volumes, HMSO, London.Environment Agency, 1997a. The Quality of Ri6ers and Canals in England and Wales 1995, Bristol.Environment Agency, 1997b. Ri6er Habitat Sur6ey, 1997 Field Sur6ey Guidance Manual, Bristol.Ferguson, R.I. 1981. ‘Channel form and channel changes’, in Lewin, J. (Ed), British Ri6ers, George Allen and Unwin,

London, 90–125.Fox, P.J.A, Naura, M. and Scarlett, P. 1998. ‘An account of the derivation and testing of a standard field method:

an example using River Habitat Survey’, Aquatic Conser6ation: Marine and Freshwater Ecosystems, 8, 455–475.Frissell, C.A., Liss, W.J., Warren, C.E. and Hurley, M.D. 1986. ‘A hierarchical approach to classifying stream habitat

features: viewing streams in a watershed context’, En6ironmental Management, 10, 199–214.Gibbons, D. W., Reid, J. B. and Chapman, R. A. 1993. The New Atlas of Breeding Birds in Britain and Ireland

1988–1991, T & A.D. Poyser, London.Hansen, H.O., Boon, P.J., Madsen, B.L. and Iversen, T.M. (Eds) 1998. ‘River restoration: the physical dimension’,

Aquatic Conser6ation: Marine and Freshwater Ecosystems, 8, 1–264.Harper, D.M. and Everard, M. 1998. ‘Why should the habitat-level approach underpin holistic river survey and

management?’, Aquatic Conser6ation: Marine and Freshwater Ecosystems, 8, 395–413.Hawkins, C.P., Kershner, J.L., Bisson, P.A., Bryant, M.D., Decker, L.M., Gregory, S.B., McCullogh, D.A., Overton,

C.K., Reeves, G.H., Steedman, R.J. and Young, M.K. 1993. ‘A hierarchical approach to classifying stream habitatfeatures’, Fisheries, 18, 3–12.

Holmes, N.T.H., Boon, P.J. and Rowell, T.A. 1998a. ‘A revised classification system for British rivers based on theiraquatic plant communities, Aquatic Conser6ation: Marine and Freshwater Ecosystems, 8, 555–578.

Holmes, N.T.H., Newman, J.R., Dawson, F.H., Chadd, S., Rouen, K.J. and Sharp, L. 1998b. ‘Mean trophic rank:a user’s manual’, Environment Agency R & D Technical Report E38, Bristol.

Jeffers, J.N.R. 1998. ‘Characterization of river habitats and prediction of habitat features using ordination tech-niques’, Aquatic Conser6ation: Marine and Freshwater Ecosystems, 8, 529–540.

Kershner, J.L. and Snider, W.M. 1992. ‘Importance of a habitat level classification system to design instream flowstudies’, in Boon, P.J., Calow, P. and Petts, G.E. (Eds), Ri6er Conser6ation and Management, John Wiley,Chichester, 179–192.

Kondolf, G.M. 1995. ‘Geomorphological stream channel classification in aquatic habitat restoration: uses andlimitations’, Aquatic Conser6ation: Marine and Freshwater Ecosystems, 5, 127–141.

Mainstone, C.P., Barnard, S. and Wyatt, R. J. 1994. ‘The NRA National Fisheries Classification Scheme, a guide forusers’, R&D Note 206, National Rivers Authority, Bristol.

Milner, N., Wyatt, R.J. and Scott, M.D. 1993. ‘Variability in the distribution and abundance of stream salmonids,and the associated use of habitat models’, Journal of Fish Biology, 43(Supplement A), 103–109.

Milner, N.J., Wyatt, R.J. and Broad, K. 1998. ‘HABSCORE—applications and future development of related habitatmodels’, Aquatic Conser6ation: Marine and Freshwater Ecosystems, 8, 633–644.

Mitchell, P. 1990. The En6ironmental Condition of Victorian Streams, Internal Report to Department of WaterResources, Victoria, Australia.

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)

Page 23: Raven et al

QUALITY ASSESSMENT USING RIVER HABITAT SURVEY DATA 499

Naiman, R.J., Lonzarich, D.G., Beechie, T.J. and Ralph, S. C. 1992. ‘General principles of classification and theassessment of conservation potential in rivers’, in Boon, P.J., Calow, P. and Petts, G.E. (Eds), Ri6er Conser6ationand Management, John Wiley, Chichester, 93–124.

National Water Council, 1981. Ri6er Quality— the 1980 Sur6ey and Future Outlook, National Water Council,London.

National Rivers Authority, 1996. Ri6er Habitats in England and Wales—A National O6er6iew, RHS Report 1, Bristol.Naura, M. and Robinson, M. 1998. ‘Principles of using River Habitat Survey to predict the distribution of aquatic

species: an example applied to the native white-clawed crayfish Austropotamobius pallipes ’, Aquatic Conser6ation:Marine and Freshwater Ecosystems, 8, 515–527.

Newson, M.D. 1997. Land, Water and De6elopment, second edition, Routledge, London.Nixon, S.C., Clarke, S.J., Dobbs, A.J. and Everard M. 1996. De6elopment and Testing of General Quality Assessment

Schemes, NRA R&D Report 27, HMSO, London.Ormerod, S.J., Baral, H.S., Brewin, P.A., Buckton, S.T., Juttner, I., Rothfritz, H. and Suren, A.M. 1997. ‘River

Habitat Survey and biodiversity in the Nepal Himalaya’, in Boon, P.J. and Howell, D.L. (Eds), Freshwater Quality:Defining the Indefinable?, The Stationery Office, Edinburgh, 241–250.

Padmore, C.L. 1997. ‘Biotopes and their hydraulics: a method for defining the physical component of freshwaterquality’, in Boon, P.J. and Howell, D.L. (Eds), Freshwater Quality: Defining the Indefinable?’ The Stationery Office,Edinburgh, 251–257.

Ratcliffe, D.A. (Ed.) 1977. ‘A Nature Conservation Review’, Cambridge University Press, Cambridge.Raven, P.J., Fox, P., Everard, M., Holmes, N.T.H. and Dawson, F.H. 1997. ‘River Habitat Survey: a new system for

classifying rivers according to their habitat quality’, in Boon, P.J. and Howell, D.L. (Eds), Freshwater Quality:Defining the Indefinable?, The Stationery Office, Edinburgh, 215–234.

Raven, P.J., Boon, P.J., Dawson, F.H. and Ferguson, A.J.D. 1998a. ‘Towards an integrated approach to classifyingand evaluating rivers in the UK’, Aquatic Conser6ation: Marine and Freshwater Ecosystems, 8, 383–393.

Raven, P.J., Holmes, N.T.H., Fox, P.J.A., Dawson, F.H., Everard, M., Fozzard, I.R. and Rouen, K.J. 1998b. Ri6erHabitat Quality: The Physical Character of Ri6ers and Streams in the UK and the Isle of Man, Environment Agency,Bristol.

Rosgen, D.L. 1994. ‘A classification of natural rivers’, Catena, 22, 169–199.RSPB, NRA and RSNC, 1994. The New Ri6ers and Wildlife Handbook, Royal Society for the Protection of Birds,

Sandy.Sansbury, J. 1994. ‘The applicability of River Habitat Survey for catchment management planning’, unpublished

report to the National Rivers Authority, North West Region, Warrington.Schneiders, A., Verhaert, E., De Blust, G., Wils, C., Bervoets, L. and Verheyen, R.F. 1993. ‘Towards an ecological

assessment of watercourses’, Journal of Aquatic Ecosystem Health, 2, 29–38.Tunstall, S., Fordham, M., Green, C. and House, M. 1997. ‘Public perception of freshwater quality with particular

reference to rivers in England and Wales’, in Boon, P.J. and Howell, D.L. (Eds), Freshwater Quality: Defining theIndefinable?, The Stationery Office, Edinburgh, 39–58.

Verniers, G. 1992. ‘Proposition d’une methodologie d’evaluation ecologique des cours d’eau avant amenagement etapplication au cas de l’Ourthe’, in Impact of Watercourse Impro6ements, Assessment, Methodology, Management,Assistance, Proceedings of an International Symposium, Wepion, Belgium, 224–232.

Werritty, A. 1997. ‘Conserving the quality of freshwater resources: the role of integrated catchment management’, inBoon, P.J. and Howell, D.L. (Eds), Freshwater Quality: Defining the Indefinable?, The Stationery Office, Edinburgh,489–505.

Wilkinson, J., Martin, J., Boon, P.J. and Holmes, N.T.H. 1998. ‘Convergence of field survey protocols for SERCON(System for Evaluating Rivers for Conservation), and RHS (River Habitat Survey)’, Aquatic Conser6ation: Marineand Freshwater Ecosystems, 8, 579–596.

Wright, J.F., Furse, M.T. and Armitage, P.D. 1994. ‘Use of macro-invertebrate communities to detect environmentalstress in running waters’, in Sutcliffe, D.W. (Eds), Water Quality and Stress Indicators in Marine and FreshwaterEcosystems: Linking Le6els of Organisation (Indi6iduals, Populations and Communities), Freshwater BiologicalAssociation, Ambleside, 15–24.

Wright, J.F., Furse, M.T. and Moss, D. 1998. ‘River classification using invertebrates: RIVPACS applications’,Aquatic Conser6ation: Marine and Freshwater Ecosystems, 8, 617–631.

© 1998 John Wiley & Sons, Ltd. Aquatic Conser6: Mar. Freshw. Ecosyst. 8: 477–499 (1998)