A review of ecology based classification systems for standing freshwaters
SR(02)03
July 2002

EXECUTIVE SUMMARY

The European Community Water Framework Directive (WFD: 2000/60/EC) requires Member States to employ ecologically-based classification systems ("broad, integrated biologically-based systems which provide an assessment of ecosystem health") as procedures for biological monitoring of surface waters, in order to assess system biointegrity. In many parts of the world new legislative drivers, such as the WFD, demand biologically-based monitoring systems to assess the "health" (biointegrity) of aquatic ecosystems. These systems are based either on predictive modelling or multimetric approaches, depending on individual requirements and legislation.

The main objective of this study was to review and report on ecologically-based classification and monitoring systems, utilising macrophytes, phytoplankton and benthic invertebrates, individually or in combination, for potential application to standing waters in the UK. The study was commissioned by SNIFFER, the Environment Agency and English Nature because currently-available standing water classification systems in use in the UK (e.g. Fozzard et al 1999) do not fully meet the needs of the Directive in this respect. The contractors were asked to:

• review existing survey, monitoring and classification systems used for ecological assessment of temperate freshwater system biointegrity (emphasising standing waters > 0.5km2 surface area) on an international basis, primarily targetting three biotic assemblages: aquatic plants (phytoplankton and macrophytes) and benthic invertebrates.
• determine the relative suitability of such schemes, with regard to scientific basis, stage of development, and degree of utilization of the "reference conditions" approach prescribed by the WFD.
• assess the applicability (or potential for development to suit UK conditions) of candidate individual schemes and approaches, with regard to factors such as scale, biota, biogeoclimatic typologies, and costs of development, implementation and operational use.

The structure of the report covers four main areas: (i) summary review of the relevant scientific literature (including both peer-reviewed and "grey" studies and sources); (ii) compilation of contacts amongst regulators/ agencies and scientists working on practice and procedure in this field, worldwide; (iii) development of an assessment protocol for identification of candidate methods based on criteria of scientific information, resource implications, timescales for development, and health/ safety aspects; (iv) application of the assessment protocol to produce recommendations on biomonitoring methodologies and classification systems appropriate for application to UK standing waters for WFD implementation purposes.

Numerous schemes for classifying and monitoring the biointegrity of lakes, ponds and reservoirs have been developed and are examined to varying extents in the literature. These schemes typically utilise macrophytes, phytoplankton, periphyton, benthic invertebrates and fish, either singly or in combination. Increasingly, biomonitoring schemes utilise the changed-state approach, or incorporate comparison with "pristine" baseline sites (e.g. the river environmental monitoring scheme, RIVPACS): both approaches being variants on the WFD-prescribed reference conditions approach.

Biological monitoring programmes fall within the framework of the three types of monitoring (surveillance, operative and investigative) prescribed by WFD. These have differing aims and intensities of sampling.

With regard to standing freshwaters, the objectives of the monitoring programmes which must be defined and "ready for commencement" by 2007 by Member States are to:

• provide a coherent and comprehensive overview of ecological and chemical status of lakes and other standing waters;
• permit classification of standing waters into five classes of ecological status: high, good, moderate, poor and bad;
• be based on characterisation and impact assessment carried out for each river basin district;
• cover parameters which are indicative of the status of each relevant quality element.

Whilst ecologically-based systems of river classification in use in the UK (e.g. RIVPACS) meet the reference-based requirements of the WFD, no similar comprehensive reference-based classification or monitoring schemes (utilising aquatic biota) are in place for lakes and other standing waters. In general, work on lakes for WFD-based monitoring programmes lags some way behind the work done in rivers. Existing lake classification schemes are more fully implemented in Scotland than in England/ Wales but utilize only chemical data.

Currently uncertainties in monitoring requirements relating to the implementation of WFD include:

• Methods of assessing water bodies at risk from pressures.
• Future flexibility of the surveillance network.
• Need to identify pressures and undertake impact assessments by 2004.
• Derivation of pressure-sensitive indicators (with R&D implications).
• · Balance between programme of measures and environmental objectives, in relation to costs-benefits.

Issues relating to typology requirements and risk based prioritisation protocols for sampling lakes are currently being developed in the UK accordance with the requirements of the WFD.

Considering current relevant research within the UK, the EC-funded ECOFRAME project looks promising as a biomonitoring system, with a tiered approach permitting the method to be customized to the requirements, level of effort, and resource availability of the user. Other useful progress has been made involving monitoring systems based on chironomids, diatoms and macrophytes. Recent projects on lake biomonitoring in Northern Ireland and Wales have potential for future development in this context. There has also been a substantial amount of work, mainly in Europe and the USA, on biomonitoring in rivers and, to a lesser extent, lakes discussed in detail in this report.

An example of an assessment system, classifying lake quality using physico-chemical data, is the Scottish Standing Waters Quality Classification Scheme, developed by the Scottish Environment Protection Agency. This scheme classifies standing waters along three gradients: (i) increase in P concentration (as an indicator of potential for eutrophication) since values hindcast to 1850; (ii) loss of acid neutralising capacity (ANC): an indicator of acidification; and, (iii) degree of contamination with toxic substances. Standing waters are placed into one of 4 quality bands along these gradients. The overall quality class is that of the lowest score on any of these bands. Although there is provision for back-up comparison with biotic data, no biological information directly contributes to the lake classification.

While the aim inherent in the WFD is a laudable one, in that it will require all EU states to come up with more or less comparable standard approaches to water quality assessment, it may be questioned whether one EU-wide system can really cater for widely varying practical situations and requirements in catchment water management needs, at all spatial and temporal scales on a continent-wide basis. It is very important to ensure that local-scale variability in response of aquatic systems is not lost in the application of large-scale standardised schemes. However, the experience of US state and federal agencies, for example, in developing and implementing the successful LRBB scheme provides encouragement in this respect (Gerritsen et al 2000). The development and implementation of a good system typology is shown by US experience to be a vital issue in developing a successful system in this context.

Monitoring schemes based on the use of biological indicators of aquatic ecosystem biointegrity should in general aim to meet six criteria (Norris & Hawkins 2000). Effective methods will:

• quantify and simplify complex ecological phenomena;
• provide easily interpretable outputs;
• respond predictably to damage caused by humans, while being insensitive to natural spatial/ temporal variation;
• relate to an appropriate scale;
• relate to management goals;
• be scientifically defensible.

The use of aquatic macrophytes, phytoplankton and benthic macroinvertebrates as target bioassessment assemblages was addressed in this project. Other organisms were excluded from the terms of reference of the contract. Various advantages of using each of these groups of biota in a number of existing reference-based (or other) schemes have been recognised. While these are often well tested for rivers, schemes for lakes are much less well developed.

Numerous other schemes for freshwater biomonitoring and classification exist or are in varying stages of development, utilising either plants or aquatic invertebrates, or both (sometimes in combination with other biota).

Specific tasks carried out during the project, to meet the terms of reference, included the following:

• A comprehensive international list was drawn up of potential contacts, including people and organisations undertaking research on bioassessment and classification of freshwater systems; regulatory agencies involved in implementing such schemes, and other relevant organisations. Contacts and websites comprise 38 in the UK and Ireland, 96 in the rest of Europe, and 53 in the rest of the world. We succeeded in contacting approximately half of the named contacts.
• A suitability matrix was drawn up for assessing the relative value of individual schemes. This was agreed with SNIFFER/EA as an early deliverable of the project. The assessment protocol included criteria for scientific basis (implementation and scientific quality), resource implications, timescales and health and safety issues. Hierarchical scoring was applied to identify schemes which did not meet minimum SNIFFER/EA requirements, and which could therefore be disregarded. Further scoring of those which did meet minimum requirements allowed relative comparisons to be made between various schemes.
• Brief details are given for a large number of sampling procedures, and minor studies concerning the use of the target biota (aquatic macrophytes, phytoplankton and benthic macroinvertebrates). However, attention was focussed on three large-scale (primary candidate) schemes which are either implemented, or at an advanced stage of development, and which appear to have reasonable potential for development to fit the requirements of SNIFFER/EA:
• the USEPA Bioassessment Protocols (LRBB) for lakes and reservoirs;
• the UK-led European shallow lake assessment scheme currently under development (ECOFRAME); and,
• the Swedish Environmental Quality Criteria (SEQC) scheme for freshwater assessment.

A number of smaller schemes, or schemes which are less apparently suited to UK lake monitoring purposes for other reasons, but which may, in due course, show promise for application for WFD implementation in relation to UK lake biomonitoring requirements were included for comparative purposes and to assess the success of the scoring methodology. These schemes were the UK Trophic Ranking Scheme (TRS), the UK pond and canal metrics-based assessment scheme (PSYM), the Welsh lakes conservation classification schem (CCLW)e, the Canadian BEAST, and the UK chironomid assessment protocol (CHIRON). Each of these schemes utilised two or less of the target biota.

Ranked in order of total score (using the matrix scoring scheme described above) the schemes come out as follows:

 Score > 40:   1. SEQC
                      2. LRBB
 Score 35-40: 3. TRS
                      4. PSYM
                      5. CHIRON
                      6. CCLW
 Score <35:    7. ECOFRAME and BEAST (equivalent scores)

The Swedish scheme scores well in almost all categories, being comprehensive and well presented, and covering landscape, biogeoclimatic and water chemistry typologies similar to those found in the UK. The scheme uses a range of metrics largely appropriate to the UK lake flora and fauna, and is relatively well tested. The US scheme is very comprehensive and well tested, covering a much broader typology range than is necessary for the UK, but has the disadvantage of geoclimatic remoteness from the British Isles. There are also major biological differences which make it impossible to apply directly to the UK. The third large-scale scheme, ECOFRAME, loses points because of its incompleteness in development and testing, and also because it is limited to shallow-water lakes. However, despite its low score on this scale, it undoubtedly has interesting potential once completed.

None of the schemes identified fulfilled all of the requirements laid down by the project terms of reference in the context of the WFD, and all were lacking in one or more of the areas of testing (and availability of reference values), coverage of sufficient biological groups (e.g. both littoral and profundal invertebrates), ease of application at the correct geographical scale, and coverage of required typologies. The integration of some of the more successful minor schemes to the potential improvement of the primary candidate schemes is an option which should be considered by SNIFFER/EA.

• It was not possible to place actual figures on the various cost headings. Instead, and in agreement with the SNIFFER/EA Project Managers, it was decided to make an informed comparison of the likely relative costs of development, implementation and operational use of the three primary schemes. This analysis provides a rough guide to likely costs. It appears that SEQC is unlikely to be any more expensive to implement than the other candidate schemes, despite the need for translation of technical materials. Because of its low level of current development (and absence of coverage of various key elements), ECOFRAME is likely to require higher development and implementation costs (especially in the pilot study phase). The main cost in dealing with LRBB is amending it to fit UK lake typologies and biota, which we consider would be substantial.

We concluded from this study that no single existing scheme, or combination of schemes could reasonably be used unchanged for WFD biomonitoring implementation purposes in lakes of the British Isles, at no extra net cost over current monitoring budgets.

The recommended option is to develop an approach based on the Swedish EQC scheme, as being most likely to permit SNIFFER/EA to have in place a working lake bioassessment approach by the WFD deadline, at acceptable cost. It should be noted that the scheme includes physico-chemical criteria and biotic elements (periphyton, fish) excluded from the brief covered by our study, but a slimmed-down version may offer good potential. The SEQC genarally meets the conditions prescribed by SNIFFER/EA for acceptability of the recommended classification system and associated surveillance methodologies.

However, it may prove useful to integrate a tiered sampling approach (as adopted within the US LRBB and ECOFRAME) to allow implementation of bioassessments based upon an informed choice of what is to be monitored at each tier, and therefore to improve cost-effectiveness. Also, prioritisation requirements may aid the identification of sites at high risk.

KEYWORDS:
Standing Freshwaters, Lakes, European Water Framework Directive, Bioassessment, Biocriteria, Macrophytes, Phytoplankton, Macroinvertebrates.

Copies of this report are available from the Foundation, price £ 35.00 less 20% toFWR members.