You are here:Home//Results//Methods database (results)//WFD River Diatom method or Trophic Diatom Index version 3 Method

back to overview methods

TDI 3 [id:]

Method: WFD River Diatom method or Trophic Diatom Index version 3 Method [WFD River Diatom method or Trophic Diatom Index version 3 Method]

1. General information

1.01 GIG: Central-Baltic, Northern
Relevant intercalibration types: n.a.
1.02 Category: Rivers
1.03 BQE: Benthic Diatoms
1.04 Country: United Kingdom
1.05 Specification: none
1.06 Method name: WFD River Diatom method or Trophic Diatom Index version 3 Method
1.07 Original name: WFD River Diatom method or Trophic Diatom Index version 3 Method
1.08 Status: Method is/will be used in First RBMP (2009), Second RBMP (2015)
1.09 Detected pressure(s):
Eutrophication, General degradation Specification of pressure-impact-relationship:
Data has been presented in EA Science Report SCO301030/SR1 Environment Agency 2007 (Kelly et al, 2007. Use of diatoms for evaluating ecological status in UK freshwaters), and by Kelly et al, 2008, Assessment of ecological status in U.K. rivers using diatoms, Freshwater Biology, 53: 403?422
Yes, with quantitative data (e.g. against range of sites reflecting continuous gradient of pressure).
1.10 Internet reference:
1.11 Pertinent literature of mandatory character:
Water Framework Directive - United Kingdom Advisory Group (WFD-UKTAG), 2008. UKTAG river assessment methods macrophytes and phytobenthos. Phytobenthos- diatom assessment for River ecological status (DARES1).
Kelly, M.G., S. Juggins, H. Bennion, A. Burgess, M. Yallop, H. Hirst, L. King, J. Jamieson, R. Guthrie & B. Rippey, 2007. Use of diatoms for evaluating ecological status in UK freshwaters. Environment Agency Science report SCO301030/SR.
Environment Agency England & Wales use these operational instructions (regularly reviewed):
EA Ref. No. 027_07 Sampling diatoms from rivers and lakes
EA Ref. No. 087_07 Fixing phytoplankton and diatom samples with Lugol's iodine
EA Ref. No. 028_07 Diatom sample digestion and slide preparation
EA Ref. No. 029_07 Diatom slide analysis, recording and archiving
EA Ref. No. 198_07 Quality Assurance Scheme for diatom samples
EA Ref. No. 387_09 Interpreting and reporting freshwater ecology data
1.12 Scientific literature:
Kelly et al., 2008. Assessment of ecological status in UK rivers using diatoms. Freshwater Biology 53: 403-422.
Kelly, M.G., H. Bennion, A. Burgess, J. Ellis, S. Juggins, R. Guthrie, B.J. Jamieson, V. Adriaenssens & M. Yallop, 2009. Uncertainty in ecological status assessments of lakes and rivers using diatoms. Hydrobiologia 63: 5-15.
Kelly, M.G., L. King, G. Clarke, H. Bennion & M. Yallop, 2006. Recommendations for sampling littoral diatoms in lakes for ecological status assessments. Journal of Applied Phycology 18: 15-25.
Kelly, M.G., L. King, R. Jones, P. Barker & B.J. Jamieson, 2008. Validation of diatoms as proxies for phytobenthos when assessing ecological status in lakes. Hydrobiologia 610: 125-129.
Yallop, M., H. Hirst, M. Kelly, S. Juggins, B.J. Jamieson & R. Guthrie, 2009. Validation of ecological status concepts in UK rivers using historic diatom samples. Aquatic Botany 90: 289-295.
1.13 Method developed by: Dr Martyn Kelly
Email of developer:
Institute of developer: Bowburn Consultancy
1.14 Method reported by: Jan Krokowski, Imelda O?Neill, Jane Jamieson
Email of person reporting the method:,,
Email of institute reporting the method:
Scottish Environment Protection Agency (SEPA), Northern Ireland Environment Agency (NIEA), Environment Agency (EA, England and Wales)
1.15 Comments: none

2. Data acquisition

Field sampling/surveying

2.01 Sampling/Survey guidelines:
Kelly, M.G., A. Cazaubon & E. Coring et al., 1998. Recommendations for the routine sampling of diatoms for water quality assessments in Europe. Journal of Applied Phycology 10: 215?224.
EN 13946, 2003. Water Quality ? Guidance Standard for the Routine Sampling and Pretreatment of Benthic Diatoms from Rivers.
EN 14407, 2004. Water Quality ? Guidance Standard for the Identification, Enumeration and Interpretation of Benthic Diatom Samples from Running Waters.
Environment Agency England & Wales also uses these operational instructions (regularly reviewed):
EA Ref. No. 027_07 Sampling diatoms from rivers and lakes
EA Ref. No. 087_07 Fixing phytoplankton and diatom samples with Lugol's iodine
EA Ref. No. 028_07 Diatom sample digestion and slide preparation
EA Ref. No. 029_07 Diatom slide analysis, recording and archiving
EA Ref. No. 198_07 Quality Assurance Scheme for diatom samples
EA Ref. No. 387_09 Interpreting and reporting freshwater ecology data
2.02 Short description:
Cobbles are the recommended substratum because they are stable (allowing diatom communities to develop) and manoeuvrable. Cobbles are available in most river types. Five cobbles/small boulders, free from algae, are collected from mid-stream and placed into a tray with a little stream water and the top surface of each brushed with a clean toothbrush to remove the biofilm. The resulting suspension was collected in a plastic bottle, fixed with Lugol?s iodine and stored prior to analysis.
Step Action
1 From the sampling area, collect at least five cobbles (64 to 256 mm) or small boulders (> 256 mm) that have an obvious diatom film (brown colour and slimy texture).
In standing waters, collect samples from depths where cobbles are permanently submerged and that you can reach wearing thigh waders.
Note: If suitable substrata are very abundant, select each cobble from a separate location within reach or within the sampling area.
2 Gently agitate the cobbles in river or lake water to remove loosely attached surface contamination (this will not dislodge the biofilm).
Surface contamination might include small particles of organic matter or sediment.
3 Place the stones in a tray with about 50 ml of river or lake water.
4 Wash a stiff toothbrush in clean river or lake water and rub it on waders or a similar surface to remove any diatoms from previous samples.
5 Brush the upper surface of the stone vigorously to remove the diatom film, rinsing the toothbrush periodically in the tray water to transfer the diatoms.
If there are filamentous algae or silt deposits on the stone, try to remove diatoms from the stone where it is free of contaminants. they don?t, try brushing them for a sample.
6 Replace the stone in the river or lake and repeat the steps above for other stones.
7 Transfer the tray water (which should be brown and turbid from the diatoms) from the tray to the sample bottle.
8 If samples will be stored for some time, you can concentrate the suspension by:
1. allowing it to settle overnight;
2. decanting the supernatant; transferring the sediment to a smaller (60 - 100 ml) bottle.
2.03 Method to select the sampling/survey site or area: n.a. Other method to select the sampling/survey site or area:
Sites most representative of waterbody (riffles, runs, glides with suitable substrata) and associated with sampling sites used for other biota
2.04 Sampling/survey device: n.a.
Other phytobenthos sampling device: Toothbrush
Any other sampling device: Toothbrush
2.05 Specification:
toothbrush, strong scissors, white plastic tray, wide-mouthed plastic sample bottles with watertight lids, waterproof permanent marker pen or another means of labelling samples, (house bricks with holes in, and polypropylene rope ? only if using introduce
2.06 Sampled/surveyed habitat:
Specification of sampled habitat:
Generally cobbles but other habitats when cobbles are not present. Sample habitat is chosen based on that which is appropriate for optimising the presence of diatoms at a site.
Sampled habitat: All available habitats per site (Multi-habitat)
2.07 Sampled/surveyed zones in areas with tidal influence: Intertidal zone
2.08 Sampling/survey month(s):
SEPA: Spring (mid-April to end of May) and autumn (September to end of November)
EA/NIEA: summer (June to end of August)
2.09 Number of sampling/survey occasions (in time) to classify site or area: 6 samples/survey occasions in a classification period
2.10 Number of spatial replicates per sampling/survey occasion to classify site or area: 5
2.11 Total sampled/surveyed area or volume or total sampling duration to classify site or area:
5 randomly selected cobbles/small boulders free of algae. Environment Agency (England & Wales) and SEPA use different sampling methods for different substrata, in order of preference: 5 randomly selected cobbles/small boulders, free of algae Algae-covered

Sample processing

2.12 Minimum size of organisms sampled and processed: n.a.
2.13 Sample treatment:

Sample is divided (sub-sampling) and organisms of a sub-sample are identified.
2.14 Level of taxonomical identification:
Level: Species/species groups
Specification of level of determination: n.a.
2.15 Record of abundance:
Determination of abundance: Individual counts, Relative abundance
Abundance is related to: n.a.
Unit of the record of abundance: Number of valves
Other record of abundance: sampled 5 cobbles/small boulders
2.16 Quantification of biomass: n.a.
2.17 Other biological data:
Other photosynthetic organisms e.g. filamentous algae (% in 10M reach) Cover of sewage fungus above and below stones, presence and density
2.18 Special cases, exceptions, additions: none
2.19 Comments: none

3. Data evaluation


3.01 List of biological metrics:
The TDI3 is based on the weighted average equation of Zelinka & Marvan (1961).

WMS = Sum of (aj * sj) / Sum of aj
where aj the abundance or proportion of valves of species j in sample; sj, the revised nutrient sensitivity class (1?5) of species j; WMS, the weighted mean score. The second step was performed to present the TDI on a score ranging from 0 (very low nutrients) to 100 (very high nutrients).
3.02 Does the metric selection differ between types of water bodies: No
3.03 Combination rule for multi-metrics: Not relevant
3.04 From which biological data are the metrics calculated:
List of biological metrics: Aggregated data from multiple sampling/survey occasions in time

Reference conditions

3.05 Scope of reference conditions: Surface water type-specific
3.06 Key source(s) to derive reference conditions:
Scope of reference conditions:
Existing near-natural reference sites, Expert knowledge, Historical data, Modelling (extrapolating model results)
3.07 Reference site characterisation:
Number of sites:
169 sites across Scotland, England and Wales and Northern Ireland were used to derive reference conditions for the method.
Geographical coverage: Scotland, England and Wales and Northern Ireland
Location of sites:
Large numbers of reference sites were found in Scotland, Wales and south-west England ? almost none in densely populated areas of the midlands and southern England
Data time period:
Reference sites were identified from the DARES database ? comprising subsets of data from 1970?s through to 2005
The process of identifying reference sites from the DARES was iterative, as data were screened and hypotheses tested. Guidelines from UK studies associated with the Habitats Directive (European Community, 1992) set limits no higher than 30 µg l-1 SRP in rivers without significant anthropogenic influences (Pitt et al., 2002) and this value was used to filter out an initial pool of potential reference sites. A further criterion used in the first iteration was that the invertebrate biology, as evaluated by RIVPACS, had to fall into the top two classes. The precise limits varied between the Environment Agency, SEPA and EHS but all correspond, approximately, to 'good status' or better.
Following this, a further iteration (based on discussions with other experts in the UK) set a threshold of 20 µg l-1 SRP for sites with total alkalinity < 50 mg l-1 CaCO3 and 30 µg l-1 SRP for sites with alkalinity ≥ 50 mg l-1 CaCO3.
As more sites with high resolution SRP data become available, these limits will need to be revisited. The data were also screened to remove sites with high nitrate-N concentrations. A value of 2 mg l-1 nitrate-N was applied to Low Alkalinity sites while a higher value (4 mg l-1 nitrate-N) was applied to sites with total alkalinity ≥ 50 mg l-1 CaCO3 for the same reasons as described above, though this will almost certainly include some slightly impacted sites. Initial analysis of the resulting reference sites showed some to have high TDI values, suggesting that even after screening using chemical criteria the reference groups still contained sites suffering from the impacts of elevated nutrient concentrations. We therefore applied a further screening and removed sites with TDI scores > 50. The above screening identified a subset of 278 reference samples from 169 sites, from the total database of 1051 samples. Figure 4.1 shows the spatial distribution of reference and non-reference samples, and the distribution of reference samples in relation to alkalinity and altitude. Figure 4.2 summarises additional environmental characteristics of the reference samples. Reference sites are distributed primarily around the periphery of Great Britain, with large numbers in Scotland, Wales and north and south-west England. There are almost no reference sites in the densely populated areas of the midlands and southern England. This geographic bias is also reflected in the hydrochemistry: while 661 (63%) samples in the total database are from sites with mean annual alkalinity
3.08 Reference community description:
High relative abundance of Achnanthidium spp.( many sites also contained A. biasolettiana and/or A.microcephalum), attached taxa Gomphonema spp, and loosely-attached Fragilariophyceae (Fragilaria capucina was the most abundant, but Meridion circulare, Hannae arcus and Tabellaria flocculosa were all common at lower alkalinities), but few motile taxa. A few lower alkalinity sites were dominated by Achnanthes oblongella, and Cocconeis placentula was also abundant on some occasions.
3.09 Results expressed as EQR: Yes

Boundary setting

3.10 Setting of ecological status boundaries:
High-good boundary derived from metric variability at near-natural reference sites
Using paired metrics that respond in different ways to the influence of the pressure
3.11 Boundary setting procedure:
The high/good boundary is set at the 75th percentile of EQR values for reference sites within a particular type.
?Crossover? between nutrient-sensitive and nutrient-tolerant species (Pollard and van de Bund, 2005).
Biological metrics tend to show gradual change as the level of nutrient/organic pressure increases, with no distinct discontinuities that could act as criteria for setting class boundaries. An alternative approach ? based on the proportions of nutrient-tolerant, nutrient-sensitive and indifferent taxa within samples ? was used to define status class boundaries in the UK, with the good/moderate boundary set at the point where the proportion of sensitive taxa falls below that of tolerant taxa. In ecological terms, the diatom flora at high and good status is characterised by a number of taxa, often with relatively broad niches (e.g. Achnanthidium minutissimum, Fragilaria capucina) which occur at different phases of a microsucession from colonisation of bare rock up to a mature biofilm (see Biggs et al., 1989). At high status, these are accompanied by other nutrient-sensitive taxa but as nutrient concentrations increase, the most sensitive of these taxa disappear whilst the numbers of nutrient tolerant taxa increases. Therefore ?crossover? is the point at which the taxa which form the ?association? characteristic of a site in the absence of pressure become subordinate to taxa which are favoured by a pressure (nutrients, in this case). The EQR gradient below the good/moderate boundary is then divided into three equally-spaced portions from which the moderate/poor and poor/bad boundaries are derived.
3.12 "Good status" community:
A. minutissimum, F. capucina, F. vaucheriae and N. dissipata were present in a majority of sites at 99.5%, 69.2%, 70.6% and 70.6% respectively, but the maximum relative abundance recorded was lower than in samples at high status (62.5%, 25.7%, 26.0% and 41.6% relative abundance respectively). Other species including G. parvulum, A. pediculus, Planothidium lanceolatum, Reimeria sinuata and motile species including N. gregaria, N. lanceolata, Navicula minima and N. dissipata were present in over 70% of all samples in this status class. Concerning these species the highest maximum of relative abundance was recorded for G. parvulum (61.4%).


3.13 Consideration of uncertainty: Yes
Specification of uncertainty consideration:
Detailed in chapter 6 of the report: Use of diatoms for evaluating ecological status in UK freshwaters. Environment Agency Science report SCO301030/SR1 In this chapter we ask two questions: What is the uncertainty associated with a single sample as an estimate of ecological status on the day that the sample was collected? How well does this sample reflect the long-term average condition of the biology? These questions are addressed separately. The former uses a nested analysis of variance that examines variation in metrics associated with variability on a slide nested within variability at a site. No attempt has been made to separate (natural) spatial variability from variability introduced by the operator but the latter sources of error were minimised by use of standard methods. Errors associated with making slides are relatively small and differences between lakes and rivers are minor. If analysts adhere to protocols, one slide per sample is sufficient to estimate the taxonomic composition and derived indices from a sample. The variance between replicate samples taken at one time from one location in lakes was much smaller than in rivers. There is a large amount of temporal variation at single sampling locations in rivers and reliable indications of status class will need to be based on repeated sampling from the same location. Results suggest that at least six replicates (i.e. two per year for three years or three per year for two years) will be required in order to provide a firm basis for regulation. A sampling intensity greater than this might be at risk of ?pseudo-replication?. The risk of misclassification depends on the proximity of the mean EQR for a site to the status class boundary. When the EQR value is very close to the boundary, the risk of misclassification will be approximately 50%, regardless of the number of samples available.
3.14 Comments: none

back to overview methods

WISER: "Water bodies in Europe: Integrative Systems to assess Ecological status and Recovery"
Online: [date: 2022/08/16]
© 2022 WISER (Contract No. 226273). All rights reserved.