general enquiries on this form should be made...

General Enquiries on the form should be made to:Defra, Procurements and Commercial Function (Evidence Procurement Team)E-mail: [email protected]

Evidence Project Final Report

NoteIn line with the Freedom of Information Act 2000, Defra aims to place the results of its completed research projects in the public domain wherever possible. The Evidence Project Final Report is designed to capture the information on the results and outputs of Defra-funded research in a format that is easily publishable through the Defra websiteAn Evidence Project Final Report must be completed for all projects.

This form is in Word format and the boxes may be expanded, as appropriate.

ACCESS TO INFORMATIONThe information collected on this form will be stored electronically and may be sent to any part of Defra, or to individual researchers or organisations outside Defra for the purposes of reviewing the project. Defra may also disclose the information to any outside organisation acting as an agent authorised by Defra to process final research reports on its behalf. Defra intends to publish this form on its website, unless there are strong reasons not to, which fully comply with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000.Defra may be required to release information, including personal data and commercial information, on request under the Environmental Information Regulations or the Freedom of Information Act 2000. However, Defra will not permit any unwarranted breach of confidentiality or act in contravention of its obligations under the Data Protection Act 1998. Defra or its appointed agents may use the name, address or other details on your form to contact you in connection with occasional customer research aimed at improving the processes through which Defra works with its contractors.

Project identification

1. Defra Project code WT1001

2. Project title

FACTORS AFFECTING THE MICROBIAL QUALITY OF SHELLFISH

3. Contractororganisation(s)

Centre for Environment, Fisheries & Aquaculture Science (Cefas)

54. Total Defra project costs £ 74,831(agreed fixed price)

5. Project: start date................. 1 July 2009

EVID4 Evidence Project Final Report (Rev. 06/11) of 29

mailto:[email protected]

end date.................. 30 June 2011


6. It is Defra’s intention to publish this form. Please confirm your agreement to do so......................................................................................YES X NO (a) When preparing Evidence Project Final Reports contractors should bear in mind that Defra intends that

they be made public. They should be written in a clear and concise manner and represent a full account of the research project which someone not closely associated with the project can follow.Defra recognises that in a small minority of cases there may be information, such as intellectual property or commercially confidential data, used in or generated by the research project, which should not be disclosed. In these cases, such information should be detailed in a separate annex (not to be published) so that the Evidence Project Final Report can be placed in the public domain. Where it is impossible to complete the Final Report without including references to any sensitive or confidential data, the information should be included and section (b) completed. NB: only in exceptional circumstances will Defra expect contractors to give a "No" answer.In all cases, reasons for withholding information must be fully in line with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000.

(b) If you have answered NO, please explain why the Final report should not be released into public domain


Executive Summary7. The executive summary must not exceed 2 sides in total of A4 and should be understandable to the intelligent

non-scientist. It should cover the main objectives, methods and findings of the research, together with any other significant events and options for new work.

The relationship between faecal indicator organism (FIO) concentrations in coastal waters and their content in commercially harvested bivalve shellfish (e.g. clams, oysters, mussels) is driven by the effect of environmental factors and the physiological and ecological characteristics of each species. The relative importance of environmental factors on FIO levels may however change significantly after major sewerage infrastructure improvement schemes.

The overall aim of this project was to refine our understanding of bivalve mollusc microbial uptake and retention kinetics, upon which environmental factors are then superimposed to influence water:flesh relationships for given shellfish species under specific conditions. New evidence on this matter is required to underpin policy on microbiological standard(s) for Shellfish Protected Areas under the Water Framework Directive. Specifically, this project aimed to:

Study site specific factors that influence the take-up of FIOs by shellfish; Appraise the effectiveness of previous water company investment programmes (to improve point

source discharges) in improving the quality of shellfish flesh; and Investigate how strong the link is between shellfish flesh quality and water quality, and why the ratios

differ so widely from one site to another.

Evidence from peer-reviewed literature on the influence of environmental factors on FIO contamination in coastal waters was reviewed. Empirical water quality data, held by Cefas or supplied by the Environment Agency, were also examined. The following conclusions emerged from the review:

Significant differences in FIO accumulation factors between shellfish species have been reported. Usually, mussels and cockles show higher FIO accumulation rates than oysters.

When shellfish are exposed to environmental conditions within the “natural range” of the species, maximum FIO accumulation rates in shellfish may occur within 30min. exposure to the contaminating source.

Shellfish exposed to water temperatures outside the “natural range” of the species show low metabolic activity which could alter the relationship between FIO in shellfish and overlying waters.

Usually, the highest bacterial die-off rates occur when light intensity is high, sewage content is low, mixing conditions in the water column are high and turbidity is low.

The environmental conditions favourable to the survival of FIOs in seawater are low levels of solar radiation, low salinity, elevated levels of nutrients and organic matter and low temperatures.

FIO survival in the marine environment depends heavily on pre-adaptation, energetic state and physiological response of bacteria to stresses in the seawater.

Predatory activity by protozoa could determine up to 1log reduction of FIOs in the water column. Rainfall is the parameter most frequently implicated with peak FIO levels in shellfish waters. The

impact of rainfall on FIO content in shellfish waters can be detected from one to several days after a rainfall event. These lag times are often associated with catchment topography, geology, the complexity of the riverine network and water residence times.

Temporal changes in compliance with the Guideline faecal coliform standard of the Shellfish Waters Directive (SWD) and shellfish hygiene classifications under Regulation (EC) No. 854/2004 were reviewed for all UK shellfish harvesting areas for which regulatory monitoring data exist. As of 1 September 2009, there were 375 beds within 66 classified shellfish production areas in England and Wales (E&W). Of these, five were class A beds, 325 class B beds and 45 class C beds. In addition, there were eight areas from which the collection of bivalves was prohibited. Since the implementation of statutory hygiene controls in 1992, the percentage of beds achieving class B has increased from less than 60% to more than 80%. However, the percentage of class A beds decreased significantly during the period.Trend analyses on annual geometric mean E. coli levels in shellfish flesh from 203 hygiene monitoring points in E&W for the period 1999–2008 were undertaken using the non-parametric Mann-Kendall test. These analyses revealed:

No trend in E. coli levels in shellfish from 79% of the monitoring points;Significant downward trends in E. coli levels in shellfish from 12% of the monitoring points; andSignificant upward trends in the levels of the indicator in shellfish from 9% of the monitoring points.

Compliance with the Guideline faecal coliform standard of the SWD for the period 2000–2006 was also reviewed. There has been an increase in the number of monitored waters meeting the G standard, although there is year on year variation superimposed on this trend in E&W. Furthermore, the number of


monitored waters meeting the standard in 2007 and 2008 declined in relation to that in 2006.

Quantitative sanitary profiles were produced for six shellfish water catchments [Chichester Harbour (Chichester Channel), Poole Harbour West, Fal/Ruan, Yealm, Taw/Torridge (Estuary Mouth) and Ribble] in which significant improvements in sewerage infrastructure, notably the installation of UV disinfection at sewage treatment works (STWs) and improvements in intermittent discharges (IDs) have been made; and the Conwy catchment, which has not received significant improvements. FIO sources were categorised as follows:

Sewerage sources that have been improved: STWs and IDs associated with these; and Other catchment sources: primarily other sewage-related sources (STWs, IDs, septic tanks, etc.) and

agricultural sources.

For sewerage sources where improvements have been made, every effort has been made to obtain flow and faecal coliform (FC) and enterococci (EN) concentration data both pre- and post-improvement. Unfortunately, empirical data for quantifying FIO fluxes from sewerage (and the other catchment) sources within the study catchments are totally inadequate. Where data are not available, generic data from previous CREH studies have been used and various assumptions made. For other catchment sources, regression models developed by CREH/Environment Agency, using predictor variables such as density of residences and livestock, have been used to predict the FC and EN concentrations in rivers, and to make provisional source apportionment estimates. Concentration data were combined with river flow data to quantify fluxes.

The sanitary profiles, which must be regarded as estimates since they are based largely on generic data and statistical models, indicated: Reductions in FC and EN fluxes of 39.83–87.98% and 35.64–93.91%, respectively, following

sewerage infrastructure improvements, with the smallest improvements in the Yealm catchment and largest in Taw/Torridge.

In all six catchments in which improvements have been made to the key STWs, treated effluents from these now make only very minor contributions (≤ 0.61%) to the total fluxes.

In the five catchments where IDs have been improved (i.e. excluding Chichester Channel and Conwy), then, on the assumption of a 90% reduction in the estimated volume of ID flow following improvement, the IDs post-improvement contribute only small proportions of the FC (≤ 4.21%) and EN (≤ 6.82%) fluxes. However, under a worst-case scenario in which estimated ID flow volumes are greater pre-improvement and are reduced by only 50% following improvement, ID contributions increase to > 50%.

Preliminary source-apportionment estimates suggest that sewage- and agriculture-related sources both contribute significantly to present fluxes from all seven catchments. In none of the catchments does one of the sources account for ≥ 90% of the flux, In fact, only in the case of Chichester Channel does one source (sewerage-related) account for more than about 70% of the FC and EN fluxes, and this almost certainly reflects the lack of ID improvements in this catchment.

While these data suggest that investments in both sewerage infrastructure improvements and agricultural best management practices (BMPs) will lead to further reductions in FIO fluxes to shellfish waters, the effectiveness of BMPs in reducing FIO fluxes at the catchment scale has yet to be fully established, and implementation may prove costly in large catchments such as Taw/Torridge (2094 km2). In contrast, investment in further improvements to STWs and IDs, such as those proposed in the EA’s Pollution Reduction Plans (PRPs), is more easily targeted and the benefits more readily evaluated.

Clearly, the current empirical evidence base for underpinning policies to reduce FIO fluxes to shellfish waters is limited. Detailed monitoring is needed to determine the fluxes from individual sources, both before and after intervention, to allow accurate characterisation.

Finally, the relationships between levels of FIOs in seawater and shellfish flesh were studied using linear and logistic regression on paired data where E. coli were enumerated in shellfish samples (native oysters, Pacific oysters and mussels) and nearby waters collected during the period 1991–1994. It was found that:

Levels of E. coli in shellfish increase with those in the seawater. The “pooled species” (mussels and oysters) linear regression model accounted for 35% of the variance observed in the variables.

Logistic regression models were used to determine targets for geometric mean and 90 th percentile of E. coli in seawater compliant with the SWD G standard. A geometric mean of 10 and a 90 th percentile of 55 E. coli per 100ml of water were predicted to be equivalent to the SWD G standard.

Differences in compliance rates between mussels and Pacific oysters emerged from logistic regression models indicating that these relationships are indeed complex and require further investigation.

The E. coli levels in seawater for which 90% are below a threshold of 4,600 E. coli 100g-1 FIL is predicted


to be achieved as follows: 33 E. coli 100ml-1 (mussel samples), 177 E. coli 100ml-1 (native oyster samples) and 4,200 E. coli 100ml-1 (Pacific oyster samples).In terms of correspondence with the class A threshold (≤230 E. coli 100g-1 FIL), none of the logistic models achieved 95% correspondence rates with the range of E. coli levels detected in the water.

Project Report to Defra8. As a guide this report should be no longer than 20 sides of A4. This report is to provide Defra with details of the outputs of the research

project for internal purposes; to meet the terms of the contract; and to allow Defra to publish details of the outputs to meet Environmental Information Regulation or Freedom of Information obligations. This short report to Defra does not preclude contractors from also seeking to publish a full, formal scientific report/paper in an appropriate scientific or other journal/publication. Indeed, Defra actively encourages such publications as part of the contract terms. The report to Defra should include: the objectives as set out in the contract; the extent to which the objectives set out in the contract have been met; details of methods used and the results obtained, including statistical analysis (if appropriate); a discussion of the results and their reliability; the main implications of the findings; possible future work; and any action resulting from the research (e.g. IP, Knowledge Exchange).

1. Introduction

There are two national monitoring programmes of the microbial quality of commercially harvested shellfish undertaken in UK waters. These are related to the requirements of the Shellfish Waters Directive (SWD) (2006/113/EC which replaced 79/923/EC) and European Union (EU) food hygiene legislation [primarily Regulation (EC) No 854/2004, as amended]. Each programme is organised separately for England & Wales (E&W), Scotland and Northern Ireland. The monitoring for the SWD is undertaken by the Environment Agency (EA) in E&W, the Scottish Environment Protection Agency in Scotland and the Northern Ireland Environment Agency in Northern Ireland. The shellfish hygiene monitoring under Regulation (EC) 854/2004 is the responsibility of the Food Standards Agency, with the monitoring for England and Wales being overseen by FSA London and that for Scotland and Northern Ireland by the offices in the respective countries.

The SWD seeks to protect shellfish and encourage favourable conditions of growth for their populations. Under the Directive, EU Member States are required to designate coastal and brackish waters which need protection or improvement to support shellfish. In addition, each Member State is charged with establishing pollution reduction programmes to ensure that designated waters comply with water quality parameters contained in the annex to the Directive. Under article 3 of the SWD, Member States are required to set values for the parameters listed in the Annex to the Directive for the designated shellfish waters (SWs). The standards set must not to be less stringent than the imperative (‘I’) values within the Directive and Member States are to ‘endeavour to observe’ the guideline (‘G’) values in the Directive. The SWD specifies a guideline standard of ≤300 faecal coliforms per 100ml of shellfish flesh and intervalvular fluid (FIL) that should not be exceeded on an annual basis in 75% of samples. The minimum sampling frequency specified in the Directive for this parameter is quarterly.

Over recent decades, UK water companies have made substantial investments to improve point-source discharges from sewerage infrastructure (e.g. installation of ultraviolet (UV) disinfection at sewage treatment works (STWs) and increases in the storage capacity of intermittent discharges (IDs), notably combined sewer overflows (CSOs) and storm tank overflows (STOs)). In particular, the National Environment Programme included in the Asset Management Plan (AMP) 3 was the first water company investment period in which the microbial quality of designated SWs was a driver for sewerage infrastructure improvement. Figure 1 shows that there has been a general increase in the number of monitored SWs meeting the G standard over the period 2000–2006, although there is year on year variation superimposed on this.


39

79

3336

60

44

32

42

76

2

12 1113

20

31

37

32 33

0

10

20

30

40

50

60

70

80

2000 2001 2002 2003 2004 2005 2006 2007 2008

Number of monitored shellfish waters in England and Wales meeting guideline faecal coliform standard, 2000 to 2008

Fail Pass

Figure 1. Number of monitored SWs in England and Wales meeting the Guideline faecal coliform standard of the SWD, 2000 to 2008. Source: Environment Agency. NB. Additional sites were designated in 2004 and for all years only sites meeting the minimum monitoring requirement of four samples per year are included.

The year to year variation is partly due to the fact that only a small number of results contribute to the assessment (four, taken quarterly) and also the variability in environmental factors, principally rainfall which causes the operation of intermittent discharges (such as combined sewer overflows) and land run-off. The number of monitored waters meeting the guideline standard in 2007 and 2008 declined in relation to that in 2006. This may be due to above average summer rainfall in these years.

Chapter A of Annex II of Regulation (EC) No 854/2004 prescribes three classes of levels of Escherichia coli monitored in bivalve mollusc FIL and for which a level of post-harvest treatment is required before marketing of shellfish for human consumption (Table 1). A fourth class (Prohibited) is used for those areas that do not comply with the requirements of the regulation.

Table 1. Microbiological standards for classification of bivalve mollusc production areas under Regulation (EC) No 854/2004.

Class Microbiological standarda Post-harvest treatment required

A Live bivalve molluscs from these areas must not exceed 230 Most probable number (MPN) E. coli per 100g of FILb

None

B Live bivalve molluscs from these areas must not exceed the limits of a five-tube, three dilution MPN test of 4,600 E. coli per 100g of FIL in more than 10% of samples. In the remaining 10 % of samples, live bivalve molluscs must not exceed 46,000 MPN E. coli per 100g of FILc

Purification, relaying or cooking by an approved method

C Live bivalve molluscs from these areas must not exceed the limits of a five-tube, three dilution MPN test of 46,000 E. coli per 100g of FILd

Relaying or cooking by an approved method

Prohibited >46,000 E. coli per 100g of FILe Harvesting not permitteda Reference method is given as ISO 16649-3.b By cross-reference from Regulation (EC) No 854/2004, via Regulation (EC) No 853/2004, to Regulation (EC) No 2073/2005.c From Regulation (EC) No 1021/2008.d From Regulation (EC) No 854/2004. e This class is not specifically given in the Regulation but does not comply with classes A, B or C. The competent authority has the power to prohibit any production and harvesting of bivalve molluscs in areas considered unsuitable for health reasons.

Although the number of shellfish production areas achieving class B under Regulation (EC) No 854/2004 has increased significantly in E&W over the past decade, the number of those achieving class A under the same regulation is, at the time of writing this report, less than 1% of the total number of classified production areas (Figure 2; Table 2).


0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

Perc

enta

ge o

f bed

s in

eac

h cl

ass

Class A Class B Class C Prohibited

Figure 2. Classification of bivalve mollusc production areas in England and Wales under Regulation (EC) No 854/2004 during the period 1992–2009. Source: Cefas.

Table 2. Percentage of UK bivalve mollusc production areas in each classification category (November 2009) under Regulation (EC) No 854/2004.

% United Kingdom England and Wales Scotland Northern Ireland

Class A 14.3 1 42.8 8.5Seasonal A/B 13.4 0 46 0Class B 61.5 82.5 8.9 91.5Seasonal B/C 1.8 2.5 1.1 0Class C 7.5 12 0.6 0Prohibited 1.3 2 0 0NB. Where a bed is represented by more than one monitoring point, the data from each point is considered and an overall assessment made. Should one point show obviously worse contamination than others on the same bed, then the possibility of treating these as separate enforceable entities may be considered. If such a split is not considered justifiable scientifically or is it not practical due to the enforcement problems that it would create, then the overall assessment would be based on the data from the point showing worst results.

In 2000, the European Community adopted Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 establishing a framework for Community action in the field of water policy [hereafter referred EU Water Framework Directive (WFD)]. This legislation is an important operational tool setting the objectives for future water protection within the European Union. The WFD seeks to, among others, streamline existing legislation, expand the scope of water protection to all waters and achieve "good status" for these waters. It incorporates new requirements for water management based on river basins. In 2013, the SWD will be repealed by the WFD in the UK. According to recital 51 and Article 4.9 of the WFD, at least the same level of protection afforded by the old legislation should be achieved with the implementation of the WFD. However, currently the WFD does not incorporate a microbiological standard upon which this objective could be enforced with respect to SWs (which the WFD defines as protected areas).

1.1.Objectives

The purposes of this project were the following:

Study site specific factors that influence the take-up of faecal indicator organisms (FIOs) by shellfish; Appraise the effectiveness of previous water company investment (to improve point source

discharges) in improving the quality of shellfish flesh; and Investigate how strong the link is between shellfish flesh quality and water quality, and why the ratios

apparently differ so widely from one site to another.

These aims would underpin the following policy outputs and resultant evidence base:

Information as to the effectiveness of previous water company investment to inform future investment decisions;

A realistic assessment of what improvements to SWs are achievable and will result in real improvement in the short to medium term;

Specific recommendations as to most cost effective remedial measures to improve targeting of investment which, in the past, has not consistently delivered the expected improvement;


A recommendation on a microbial standard (water column standard versus shellfish flesh standard) for Shellfish Protected Areas under the WFD and a discussion on the possibility of using shellfish flesh data from the FSA shellfish hygiene monitoring programme to inform water quality in SWs; and

Information to enable us understand better the possible causes of and influences on FIO contamination in shellfish flesh.

2. Methodology

2.1.Literature review

Published literature on biological, meteorological and hydrographic factors influencing the relationship between FIO in shellfish flesh and overlying waters was reviewed. Papers were retrieved from Scopus (http://www.scopus.com/home.url), ScienceDirect (http://www.sciencedirect.com/), Ingentaconnect (http://www.ingentaconnect.com/), Scientific Electronic Library Online - SciELO (http://www.scielo.br/) and the Biodiversity Heritage Library (http://www.biodiversitylibrary.org/Default.aspx). The terms searched were “clams”, “enteric bacteria”, “enterococci”, “Escherichia coli”, “faecal coliforms”, “faecal indicator organisms”, “irradiance”, “mussels”, “oysters”, “rainfall”, “river flow”, “salinity”, “seasonality”, “seawater”, “sediment”, “solar radiation”, “temperature”, “tides” and “turbidity”. A number of studies were also retrieved from Cefas reference archives. Abstracts were reviewed and those considered relevant were flagged for full text review. Studies undertaken in shellfish harvesting areas or bathing waters in the proximity of SWs and those reporting results from species commercially harvested in the UK were prioritised. Only studies showing results for faecal coliforms (FC), E. coli and enterococci (EN) were reviewed. A limited number of non-peer reviewed studies (e.g. conference proceedings, doctoral dissertations and project reports) were also reviewed.

2.2.Sanitary profiles of selected shellfish water catchments pre- and post-improvements in sewerage infrastructure

The aim of this study was to assess the effectiveness of recent water company investment in reducing faecal indicator organism (FIO) fluxes to SWs at selected sites by undertaking quantitative sanitary profile investigations of their catchments. The study focuses on two key FIOs: FC (the regulatory microbial parameter for SWs) and EN. Seven sites, covering different levels of sewerage infrastructure improvement and including sites showing both an improvement and deterioration in shellfish flesh quality from 1999–2008, were investigated:

1. Chichester Channel in Chichester Harbour (referred to as ‘Chichester Channel’); 2. Poole Harbour West; 3. Yealm;4. Fal/Ruan; 5. Taw/Torridge Estuary Mouth (referred to as ‘Taw/Torridge’);6. Conwy; and7. Ribble.

Their levels of sewerage infrastructure improvement and trends in shellfish flesh quality are summarised in Table 3 and locations shown in Fig. 3. Six of the sites (i.e. except Conwy) have had significant improvements to their sewerage infrastructure over recent years. Sanitary profiling for these six sites was undertaken both pre and post sewerage infrastructure improvements. For Conwy, a more limited analysis was undertaken to establish the present-day sanitary profile.

Table 3. Shellfish waters selected for sanitary profiling: sewerage infrastructure improvements and trends in E. coli concentrations in shellfish flesh.

Shellfish water Significant improvement in sewerage infrastructure (start year)a

Change in GM E. coli concentrations in shellfish flesh: 1999–2008b

Chichester Channel Yes (2008) No trendPoole Harbour West Yes (2003) No trendYealm Yes (2004) IncreaseFal/Ruan Yes (2002) No trendTaw/Torridge Yes (1997) DecreaseConwy No No trendRibble Yes (1999) No trend

Year when UV disinfection first introduced at one or more STWs within the catchment. Results from trend analysis.


http://www.biodiversitylibrary.org/Default.aspx

http://www.scielo.br/

http://www.ingentaconnect.com/

http://www.sciencedirect.com/

http://www.scopus.com/home.url

Figure 3. Location of shellfish waters investigated.

2.2.1. Overall approach

The various FIO sources within the catchment of each shellfish water (defined as comprising all land draining to the designated shellfish water), were categorised as follows:

Sewerage sources that have been improved: STWs (termed ‘key STWs’) and IDs associated with these.

Other catchment sources: primarily other sewage-related sources (STWs, IDs, septic tanks, etc. where there has been little/no improvement) and agricultural sources.

Improvements are taken to have commenced when UV disinfection was first implemented at one or more STWs. Ideally, data would have been generated for the annual cycle. However, so few empirical or generic effluent FIO and flow data are available outside the summer bathing season (here termed ‘summer’; cf. ‘winter’ for rest of year) (several requests for these data were made to water companies and EA during the period December 2010–March 2011 but, in most cases, no data were received or existing data for a very limited numbers of STWs relates to post-UV installation period only) that this proved impossible. It is argued that the average daily fluxes over the winter and summer periods are likely to be broadly similar, and winter (hence annual) fluxes have been estimated on this basis.

2.2.2 Sewage sources that have been improved

For all sewerage sources where improvements have been made, every effort has been made to obtain data on flow volumes and FC and EN concentrations both pre- and post-improvement, during the summer and winter periods. Where data are not available (which was particularly the case for IDs and for STW effluent flows and FIO fluxes pre-improvement), generic data from previous CREH studies have been used, as follows:

FC and EN concentrations in sewage and treated effluents: Base- and high-flow geometric mean (GM) concentrations in effluents from specific treatment types during the summer months (as published in Kay et al., 2008a).

Treated effluent flow volumes at STWs: Treated effluent flow data from various STWs investigated by CREH reveal a mean total flow of 355 litres PE (human population equivalent)-1 day-1 and mean proportions of base and high flow over the summer period of 0.733 and 0.267, respectively.

Flow volumes of IDs (CSOs, STOs, etc.): Detailed monitoring and/or modelling of ID flows associated with a large STW undertaken by CREH in one catchment during the summer bathing season gave a ratio of ID:STW treated effluent flow of 0.0429. This ratio has been used to estimate


the total volumes of ID flow associated with the STWs at which improvements have been made. Data on IDs that have been improved: Existing empirical ID data for the catchments are inadequate

for estimating FIO fluxes pre- and post-improvement. The following approach has been adopted for profiling the IDs: ID flow volumes associated with the key STWs pre-improvement have been estimated to be 0.0429 x total STW final effluent flow. Improvements have been assumed to have reduced the flow volumes across all IDs (i.e. improved and non-improved) by 90%, which seems reasonable since the more active IDs will be the ones targeted for improvement. In addition, the sensitivity of using these figures has been assessed by calculating likely ‘best- and worst-case scenarios’: Best-case scenario assumes a lower ID flow volume of 0.01 x STW effluent flow and a greater reduction in flow (99%) following improvement; and Worst-case scenario assumes a much greater ID flow volume of 0.873 x STW effluent flow, which is the highest estimated figure from previous CREH studies, and a smaller reduction in flow (50%) following improvement.

It should be noted that the water companies responsible for the sewerage infrastructure in each of the catchments have each been invited by Defra to comment on these assumptions, the overall methodology and resulting sanitary profiles, and have not raised any concerns.

2.2.3. Other catchment sources

Generic models developed by CREH/Environment Agency (EA) have been used to predict the FC and EN concentrations in rivers that are derived from other catchment sources during the summer, using approaches reported in Crowther et al. (2011). These have been combined with river flow data generated by the EA to quantify FC and EN fluxes.

2.2.4. Source apportionment of fluxes from other catchment sources post-improvement

Knowledge of the contribution of different FIO sources (especially sewage vs agriculture) is critical to the development of future investment strategies for the remediation of catchment-derived FIOs. Ideally, such source apportionment would be based on detailed empirical data, but these are lacking for the seven catchments investigated, or be derived using process-based catchment models, but their application to FIOs is prevented by the absence of empirical data with which to parameterise and evaluate these models (Crowther et al., 2011). In the present study, provisional source apportionment estimates have been derived using a modification of the regression modelling approach developed by Kay et al. (2010). Here, the sewage-related component has been assumed to be represented by the ‘residences’ term in the regression equation, which is entered first in each of the models, and the agriculture-related component by the sum of the various livestock and base flow index (BFI) terms. The latter is included since it is a key factor affecting the survival and transport of FIOs derived from livestock inputs to land through direct voiding of faeces, slurry applications, etc.

2.3.Relationships between FIO levels in shellfish flesh and seawater

The study was based on paired data where E. coli were enumerated in shellfish FIL and in nearby water samples by Public Health Laboratory Service (PHLS) laboratories during the period 1991–1994. The data were made available by six local authorities to Cefas1, to support an investigation on the relationship between bacterial levels in seawater and shellfish and help inform bacteriological water quality standards equivalent to those set out in Directive 91/492/EEC2 (EU Scientific Veterinary Committee Working Group on Faecal Coliforms in Shellfish, 1996).

The data cover native oysters (Ostrea edulis), Pacific oysters (Crassostrea gigas) and mussels; it includes but does not distinguish results from the common mussel Mytilus edulis and the Mediterranean mussel M. galloprovincialis. Levels of E. coli in shellfish were quantified using the MPN technique as described by the standard protocol used in the official monitoring programme (MAFF et al., 1992). Levels of E. coli in seawater were quantified using the standard method in use in each PHLS laboratory which was either membrane filtration (MF) or MPN.

The dataset used in this study consisted of 602 paired samples from 40 water and 40 flesh monitoring points within six different production areas (see Appendix 1). At 34 monitoring points (85% of the total number of points), only one species of shellfish was collected. The dataset includes E. coli results from two production areas (Conwy and Yealm) for which fluxes of FIOs impacting on SWs were quantified as part of this project. The dataset does not include information on how close in time or space water samples were collected in relation to shellfish samples. On the basis of the data available, it was not possible to investigate tidal influences on E. coli contamination at these sites.

1 Formerly designated Ministry of Agriculture, Food and Fisheries. Cefas manages the microbiological monitoring programme for bivalve mollusc harvesting areas in England and Wales on behalf of the competent authority (Food Standards Agency). 2 Superseded by Regulation (EC) No 854/2004.


Bacterial counts are conventionally converted to a log scale for analysis, so values described as E. coli may be assumed to refer to log10(MPN). Logarithmic transformation is commonly used to ensure a more symmetrical distribution of the data and is justified by biological considerations (bacteria grow exponentially). Censored data at the limit of detection (LOD) for each method were taken at face value. After log10-transformation, both datasets remained significantly non-normal (skewness/kurtosis test; p=0.007 and 0.005, respectively). However, quantile plots indicate that the datasets have similar distributions as suggested by the good fit over most of the range of E. coli results.

Simple linear regression (also known as ordinary least squares) models were computed to investigate the co-variation between E. coli levels in shellfish flesh and E. coli levels in water. Linear regression is particularly useful when estimating or predicting values of one variable based on the knowledge of another variable, for which more data are available. For the purposes of this analysis, the variable “E. coli levels in shellfish flesh” is considered the response because the mechanism of contamination is assumed to result from the filter-feeding mechanism of shellfish and accumulation of bacteria present in the seawater, i.e. the mechanism of contamination integrates contamination available during seawater flows over the preceding hours of the tidal cycle. It is assumed the E. coli do not multiply within the shellfish, but may be retained or excreted.

Logistic regression is used when the response variable is observed only as a binary characteristic: yes/no, present/absent, or in this case, comply/fail coded as 1/0. To assess the various degrees of association between the SWD faecal coliform G and levels of E. coli in seawater, weighted logistic regression models were computed for:

Compliance with SWD G versus the geometric mean of E. coli in seawater for all species tested (pooled species model);

Compliance with SWD G versus the geometric mean of E. coli in seawater for each species tested;Compliance with SWD G versus the 90th percentile of E. coli in seawater for all species tested (pooled

species model); andCompliance with SWD G versus the 90th percentile of E. coli in seawater for each species tested.

The models test the relationship between the threshold levels used for the purposes of classifying harvesting areas under Regulation (EC) No 854/2004 and the levels of E. coli in seawater. Because no covariates were available for each flesh sample other than the monitoring point and time, the data are grouped at that level and the fitted value is the proportion of samples that come under the threshold for classification. This predicted response is the probability of a sample passing the test at each E. coli level in seawater.

3. Results

3.1 Environmental factors influencing the microbial qualities of SWs and shellfish flesh

The relationship between FIO concentration in the water and their content in the shellfish shows various levels of complexity. Physiological and biological characteristics of each shellfish species often alter the relationship between levels of FIOs in shellfish flesh and overlying waters. Evidence from the literature suggests that:

Fate of FIOs in shellfish waters

Typically, the environmental conditions favourable to the survival of FIOs in seawater are low levels of solar radiation, low salinity, high levels of nutrients and organic matter and low temperatures.

The deleterious effect of solar radiation on FIO survival in seawater is caused by the synergistic interaction of various portions of the electromagnetic spectrum of UV light (UV-A, UV-B and PAR).

Significant (>1log) differences have been reported in FIO contamination in shellfish waters between seasons. Levels of E. coli in surface waters detected in 13 sampling points in the Taw/Torridge (1993–2008) and 11 sampling points in Chichester Harbour (2007–2009) evidenced differences between


winter and summer seasons exceeding 1log in sampling sites closer to the mouths of rivers and pollution sources. For instance, the monthly variation of E. coli at Bideford Bridge (head of the estuary) mirrors the monthly variation in water levels in the River Torridge. In contrast, no differences were detected in the levels of the indicator in Appledore Slipway, a site more exposed to fully saline waters from Bideford Bay. A reliable assessment of the microbial quality of SWs requires the examination of these spatial differences in FIO levels. Where possible, consideration should be given to the processes driving this seasonal variability in microbial concentrations for the purposes of informing sampling regimes.

The attachment of FIO to mobile or semi-mobile solid phases has important implications for their transport in nearshore waters. Particle size determines the movement of attached FIOs, i.e. finer particles provide greater surface area available for bacterial attachment per unit mass and remain in suspension for longer.

FIO survival in seawater depends heavily on pre-adaptation, energetic state and physiological response of bacteria to stresses in the seawater. Enteric bacteria are able to pre-adapt at a sewage works for example, before releasing into coastal waters. Their survival capability could be enhanced if bacterial cells possess certain genes associated with osmoregulation, which require nutrients to be expressed. Bacterial resistance in seawater is also determined by a range of anti-stress responses and dormancy.

Predatory/antagonistic activity by protozoa in the water column could determine up to 1log of FIO die-off. Under certain circumstances selective predation may also promote adaptive responses in bacteria.

Usually, the highest bacterial die-off rates occur when light intensity is high, sewage content in the water is low, mixing conditions in the water column are high and turbidity is low.

Salinity has been used as a proxy for FIO contamination across shellfish harvesting areas as high bacterial loads are often associated with freshwater inputs to coastal waters. Water:flesh ratios may also differ between sites because salinities significantly higher or lower than the optimum for the species will reduce pumping action of the bivalves, with the animals ceasing activity and closing at extremely high or low levels. Levels of faecal coliforms in the Ribble shellfish water were found to be negatively correlated with salinity (Spearman’s rho=-0.767). No significant correlations were found between levels of the microbiological indicator and salinity in Tresillian or Yealm shellfish waters.

Fate of FIOs in shellfish flesh

Bacterial release into the water column via sediment resuspension is an important factor affecting shellfish quality in shallow, low-energy and depositional inshore areas impacted by high magnitude storm events or in high-energy areas where wave bases are large when compared with the overall bathymetric profile.

Rainfall is the environmental parameter most frequently implicated with peak levels of faecal contamination detected in shellfish and seawater. Rainfall effects depend on the magnitude of the event and hydrological characteristics of the catchment. Rainfall-induced peak FIO concentrations could be detected in shellfish as much as 7 days after the rainfall event. The relationships between rainfall and the levels of E. coli in mussels and Pacific oysters from the Yealm Estuary over the period 2000–2009 were studied. Significant positive relationships (Spearman’s rho: 0.3–0.5, p<0.05) were obtained between these parameters at all sampling points. The magnitude of response of E. coli levels to rainfall varied between sampling points, but was consistently higher 1–2 days after the rainfall event. The fact that positive associations were found for rainfall station in Watercombe suggests potentially significant impact from pollution sources in the upper catchment. The estimated source apportionment of fluxes of FIOs to the shellfish water is given below.

Rainfall may not adequately reflect FIO levels in deep, high-energy coastal waters subject to higher dilution and affected by tidal/wind-driven currents. However, FIO contamination associated with rainfall-runoff processes could be the dominant source in some coastal waters with deep bathymetric profile in the vicinity of urbanised areas more impacted by pipeline discharges.

Significant positive association has been found between levels of FIOs in shellfish and spring tides, particularly during the ebb when the impact of freshwater plumes from upstream sources is more pronounced. In some areas, this effect may take less than 1h. However, tidal effects often vary considerable between sampling points within the same shellfish harvesting area.

Settlement of sediments removes FIOs from the water column. Therefore, surface layers of the sediments, which are preferential habitats for many species of shellfish, may contain the concentrations of


bacteria that could be expected in the water column following a recent pollution event. Data from the Shellfish Hygiene monitoring programme in England and Wales showed higher average E. coli accumulation levels in shellfish growing in riverbed (cockles, mussels) than those growing in the water column (Pacific oysters). Whilst physiological characteristics may account for some of the differences observed, the position/method of growth may also be significant. Therefore, sampling undertaken in deeper areas and in the proximity of pollution sources is more likely to return high levels of indicator organisms in stratified shellfish waters.

Significant differences in FIO accumulation factors between shellfish species have been reported in the literature. The most consistent differences are between mussels and oysters. Comparative analysis of FIO accumulation factors between native oysters and Pacific oysters has been inconclusive.

The uptake of FIO in shellfish also increases with the change in temperature towards optimum and the increase of food particles in the water. However, the degree of association between water temperature and FIO contamination in water/shellfish often depends on the proximity to pollution sources. Table 4 shows no substantial differences in average water temperatures between five of the shellfish waters selected for detailed analysis. However, there have been times when water temperatures reach levels below minimum for shellfish growth (Pacific oysters in Conwy and Yealm).

Table 4. Levels of faecal coliforms, temperature and salinity in surface waters at five shellfish waters and summary of physical factors controlling growth and survival of shellfish.

Parameter†

Range (average)Conwy(n=10)

Ribble(n=9)

Truro(n=11)

Tresillian(n=12)

Yealm(n=12)

Commercially harvested species

Mussels Mussels; cockles Mussels Mussels Pacific oysters; mussels

Temperature (°C) 6.5–17 (11.5) 2.6–17 (11.9) 8.3–16.9 (12.9) 8–17.4 (12.9) 7.8–18.5 (12.7)

Salinity (ppt) 32.5–33.5 (32.9)

29–32.9 (30.6) 22.1–34.7 (32.9) 30.2–33.9 (32.3) 28.4–34.8 (32.5)

Range (median)Faecal coliforms (surface waters) (No 100ml-1)

<2–38 (1) <2–952 (136) <2–32 (4) 2–350 (23) <10–277 (18)

Physical factors*Native oysters Pacific oysters Mussels

Temperature (minimum for growth) (°C)

8–9 8–9 -1

Temperature (minimum for survival) (°C)

3–4 5–6 -4

Temperature (maximum) (°C)

26–27 29–30 27–28

Optimum salinity range (psu)

25–35 20–30 20–35

†January 2010–February 2011. * Source: Laing and Spencer (2006).

Different species of shellfish have different requirements with respect to environmental factors. Maximum FIO accumulation rates in shellfish maintained under water temperature and salinity conditions optimal for growth and food supply typical of the growing water may occur within 30min. exposure to the contaminating source. The variations in salinity in shellfish waters selected for detailed analysis are generally within the optimum range for the species produced. Although it is not expected that salinity constitutes a limiting factor for FIO accumulation/elimination in shellfish at these sites, monitoring on a monthly/quarterly basis is systematically biased not to characterise episodic peak salinity levels and therefore there is insufficient data to test this hypothesis.

3.2. Sanitary profiles of selected shellfish water catchments pre- and post-improvements in sewerage infrastructure

3.2.1. Overview

For each catchment, the following data have been generated: Maps of land use and key STWs/IDs at which improvements have been made. Catchment characterisation, including geology, soils and land use. Significant improvements to STWs. Significant improvements to IDs. Areas located upstream of lakes/reservoirs (treated separately in the modelling).


Base flow index (BFI) and land cover. Residences and adjustments for those served by key STWs pre-/post-improvement. Stocking density pre- and post-improvement. Long-term mean flow data during the ‘summer’ and ‘winter’ periods. Predicted GM FC and EN concentrations in drainage waters in summer. Predicted summer fluxes of FC and EN to the SWs from key STWs, IDs associated with these and

other catchment sources, pre- and post-improvement. Estimated summer, winter and annual FC and EN fluxes pre- and post-improvement to SWs from all

sources. Provisional estimates of percentage of fluxes of FC and EN post-improvement that are derived from

sewage- and agriculture-related sources.

3.2.2. Comparison of sanitary profiles pre- and post-improvement

The summer fluxes of FC and EN reported in Table 5 represent the pollution load delivered to each SW pre-improvement. Apart from the Yealm and Ribble, >59% of the fluxes are derived from key STWs and their associated IDs.

Table 5. Estimated overall FC and EN fluxes to the shellfish waters during the summer bathing season PRE-IMPROVEMENT and their sources.

Shellfish water FC ENFlow: Base High Total Base High TotalFlux (cfu)Chichester Channel 4.9E+15 4.3E+15 9.2E+15 4.7E+14 5.8E+14 1.0E+15Poole Harbour West 2.1E+15 2.5E+15 4.5E+15 1.8E+14 3.7E+14 5.5E+14Yealm 2.5E+14 7.8E+14 1.0E+15 3.0E+13 1.5E+14 1.8E+14Fal/Ruan 1.5E+15 2.8E+15 4.3E+15 1.3E+14 3.5E+14 4.8E+14Taw/Torridge 1.3E+17 5.6E+16 1.8E+17 1.6E+16 7.7E+15 2.4E+16ConwyRibble 5.5E+16 1.9E+17 2.5E+17 4.6E+15 1.7E+16 2.2E+16Key STWs (%)Chichester Channel 99.81 54.82 78.88 99.68 41.35 67.34Poole Harbour West 62.38 31.48 45.48 67.09 32.41 44.05Yealm 52.00 9.08 19.40 36.28 4.50 9.86Fal/Ruan 83.47 25.12 45.71 83.61 18.81 36.24Taw/Torridge 98.72 41.06 81.12 99.39 58.47 86.12ConwyRibble 86.73 12.73 29.22 92.73 13.43 30.06IDs associated with key STWs (%)Chichester Channel 0.00 45.13 21.00 0.00 58.58 32.48Poole Harbour West 0.00 25.55 13.97 0.00 30.38 20.18Yealm 0.00 8.17 6.21 0.00 7.54 6.27Fal/Ruan 0.00 22.60 14.63 0.00 31.50 23.03Taw/Torridge 0.00 7.91 2.41 0.00 9.93 3.22ConwyRibble 0.00 10.66 8.28 0.00 20.36 16.09Other catchment sources (%)Chichester Channel 0.19 0.05 0.12 0.32 0.07 0.18Poole Harbour West 37.62 42.96 40.54 32.91 37.21 35.76Yealm 48.00 82.74 74.39 63.72 87.96 83.88Fal/Ruan 16.53 52.28 39.66 16.39 49.69 40.73Taw/Torridge 1.28 51.04 16.47 0.61 31.59 10.66ConwyRibble 13.27 76.62 62.50 7.27 66.22 53.86

Table 6 presents the equivalent post-improvement data. Following UV installation at the key STWs, contributions from these are now negligible.

Table 6. Estimated overall FC and EN fluxes to the shellfish waters during the summer bathing season POST-IMPROVEMENT and their sources.

Shellfish water FC ENFlow: Base High Total Base High TotalFlux (cfu)Chichester Channel 1.1E+13 2.0E+15 2.0E+15 3.0E+12 3.4E+14 3.5E+14Poole Harbour West 8.0E+14 1.2E+15 2.0E+15 6.2E+13 1.5E+14 2.1E+14Yealm 1.1E+14 5.0E+14 6.2E+14 1.9E+13 9.5E+13 1.1E+14Fal/Ruan 2.6E+14 1.4E+15 1.7E+15 2.0E+13 1.6E+14 1.8E+14Taw/Torridge 1.7E+15 2.0E+16 2.2E+16 9.8E+13 1.4E+15 1.5E+15Conwy 1.1E+15 4.4E+16 4.5E+16 1.0E+14 4.1E+15 4.2E+15Ribble 7.0E+15 1.2E+17 1.2E+17 3.6E+14 8.0E+15 8.3E+15


Key STWs (%)Chichester Channel 10.77 0.02 0.08 51.39 0.16 0.61Poole Harbour West 0.09 0.02 0.05 0.63 0.09 0.25Yealm 0.31 0.03 0.08 0.26 0.02 0.06Fal/Ruan 1.38 0.09 0.29 2.82 0.13 0.42Taw/Torridge 0.03 0.00 0.00 0.38 0.01 0.03Conwy 0.00 0.00 0.00 0.00 0.00 0.00Ribble 1.76 0.04 0.14 5.91 0.10 0.35IDs associated with key STWs (%)Chichester Channel 0.00 99.88 99.35 0.00 99.73 98.86Poole Harbour West 0.00 5.09 3.03 0.00 6.89 4.91Yealm 0.00 1.26 1.03 0.00 1.17 0.97Fal/Ruan 0.00 4.97 4.21 0.00 7.65 6.82Taw/Torridge 0.00 1.85 1.71 0.00 4.81 4.49Conwy 0.00 0.00 0.00 0.00 0.00 0.00Ribble 0.00 1.73 1.63 0.00 4.41 4.22Other catchment sources (%)Chichester Channel 89.23 0.10 0.57 48.61 0.11 0.53Poole Harbour West 99.91 94.89 96.91 99.37 93.02 94.84Yealm 99.69 98.71 98.89 99.74 98.81 98.97Fal/Ruan 98.62 94.94 95.50 97.18 92.23 92.77Taw/Torridge 99.97 98.15 98.29 99.62 95.18 95.48Conwy 100.00 100.00 100.00 100.00 100.00 100.00Ribble 98.24 98.23 98.23 94.09 95.50 95.44

Based on an assumed reduction of 90% in ID flow, the IDs are now also minor contributors to the overall fluxes, and other catchment sources are overwhelmingly dominant. The exception is Chichester Channel, where no ID improvements were made.

3.2.3. Reductions in FIO fluxes resulting from sewerage infrastructure improvements

Comparison of the total fluxes of FC and EN pre- and post-improvement reveals reductions over the six sites of 39.83–87.98% and 35.64–93.91%, respectively. The smallest percentage improvements are in the Yealm catchment and the largest in the Taw/Torridge (Figure 4).

Figure 4. Change (%) in overall FC and EN fluxes during summer following improvements to STWs and IDs to the various shellfish waters (no improvements in the Conwy catchment).

3.2.4. Provisional source apportionment of FIO fluxes post-improvement from other catchment sources

Preliminary source-apportionment estimates (Table 7) suggest that sewage- and agriculture-related sources both contribute significantly to present fluxes from all seven catchments. In none of the catchments does one source account for ≥ 90% of the flux, In fact, only in the case of Chichester Channel does one source (sewerage-related) account for more than about 70% of the FC and EN fluxes, and this almost certainly reflects the lack of ID improvements in this catchment.

Table 7. Provisional estimates of percentage of fluxes of FC and EN to the shellfish waters POST-IMPROVEMENT from all catchment sources, except from key STWs and their associated IDs, that are derived from sewage- and agriculture-related sources.

Shellfish water FC ENBase flow

High flow

Base flowa

High flow

Sewage-related sourcesChichester Channel 85.4 79.0 - 69.1Poole Harbour West 69.1 60.3 - 56.4


Yealm 75.1 64.3 - 63.1Fal/Ruan 51.8 44.2 - 45.2Taw/Torridge 50.7 36.9 - 38.1Conwy 61.3 38.4 - 39.5Ribble 60.5 48.7 - 50.3Agriculture-related sourcesChichester Channel 14.6 21.0 - 30.9Poole Harbour West 30.9 39.7 - 43.6Yealm 24.9 35.7 - 36.9Fal/Ruan 48.2 55.8 - 54.8Taw/Torridge 49.3 63.1 - 61.9Conwy 38.7 61.6 - 60.5Ribble 39.5 51.3 - 49.7

a The base-flow regression model for EN was unsuitable for source apportionment.

While these data suggest that investments in both sewerage infrastructure improvements and agricultural best management practices (BMPs) will lead to further reductions in FIO fluxes to shellfish waters, the effectiveness of BMPs in reducing FIO fluxes at the catchment scale has yet to be fully established, and implementation may prove costly in large catchments such as Taw/Torridge (2094 km2). In contrast, investment in further improvements to STWs and IDs, such as those proposed in the EA’s Pollution Reduction Plans, is more easily targeted and the benefits more readily evaluated.

Figure 5. Estimated source apportionment of present total (i.e. base + high flow) FC flux to the shellfish water in summer.

3.2.5. Comparison of E. coli levels in shellfish flesh pre- and post-improvement


Comparison of E. coli levels in shellfish flesh pre- and post-improvement reveals no statistically significant (Mann-Whitney test) reductions over fourteen sites within four of the SWs selected for sanitary profiling. Geometric means of E. coli in shellfish from some of the sites (Fal/Ruan: Tresillian River, Tolverne; Yealm: Thorn, Fox Cove) are indeed higher post-improvement (Table 8).

Table 8. Summary statistics for E. coli levels in shellfish from sites classified under Regulation (EC) No 854/2004 within four shellfish waters.

Shellfish bed name Shellfish water SpeciesDate of first

sampleDate of last

sample n Min Max GMDate of first

sampleDate of last

sample n Min Max GM31 Ribble Walls (North) Ribble Mytilus spp. 18/08/1992 01/12/1998 37 110 35,000 2,508 19/01/2000 17/12/2008 89 90 35,000 1,815Tresillian River Fal/Ruan Mytilus spp. 08/02/1999 09/10/2001 25 160 16,000 1,101 17/02/2003 15/12/2008 72 40 92,000 1,724Grimes Bar Fal/Ruan O. edulis 11/01/1999 18/12/2001 28 110 16,000 763 21/07/2003 15/12/2008 67 40 16,000 1,242Maggotty Bank Fal/Ruan O. edulis 08/02/1999 20/11/2001 70 40 >18,000 528 13/01/2003 15/12/2008 70 40 >18,000 765R. Pontoon/Tregothnan Fal/Ruan Mytilus spp. 11/01/1999 18/12/2001 51 40 >18,000 1,172 13/01/2003 15/12/2008 74 <20 >18,000 1,093Tolverne Fal/Ruan O. edulis 23/02/1999 18/12/2001 33 <20 >18,000 561 13/01/2003 15/12/2008 75 40 >18,000 1,113Coombe Creek Fal/Ruan O. edulis 11/01/1999 18/12/2001 30 20 9,100 419 13/01/2003 15/12/2008 66 20 16,000 702T. Pontoon/South Wood Fal/Ruan Mytilus spp. 11/01/1999 18/12/2001 51 <20 >18,000 741 13/01/2003 15/12/2008 78 20 180,000 841Fox Cove Yealm C. gigas 18/01/1999 24/11/2003 61 50 5,400 717 10/01/2005 16/02/2008 50 110 16,000 910Thorn Yealm Mytilus spp. 18/01/1999 22/12/2003 58 <20 9,100 671 08/02/2005 16/12/2008 54 50 54,000 1,287Fox Cove Yealm Mytilus spp. 07/12/1999 22/12/2003 54 40 24,000 1,005 10/01/2005 16/12/2008 48 110 54,000 2,141Thorn Yealm C. gigas 03/02/1999 22/12/2003 56 40 9,100 549 10/01/2005 16/12/2008 50 90 16,000 765

Barrel 'O' Poole Harbour West Mytilus spp. 01/02/1999 03/12/2002 23 220 >180,000 2,562 20/01/2004 01/12/2008 54 50 54,000 1,293

Rockley Poole Harbour West Mytilus spp. 01/02/1999 05/11/2002 27 70 9,100 799 20/01/2004 01/12/2008 54 <20 54,000 575

Pre-improvement Post-improvement

3.3. Relationship between FIO levels in shellfish flesh and seawater

Table 9 shows summary statistics for arithmetic levels (i.e. not logged) of E. coli in shellfish flesh and water from the six UK production areas. Minimum results correspond to the limit of detection of the enumeration methods. Overall, mussels were found to be more contaminated than oysters, as indicated by statistics representing the central tendency of the data. The higher levels of E. coli in mussels than those in oysters correspond to the general pattern detected in classified production areas in England and Wales (Younger and Reese, 2011).

The coefficients of kurtosis indicate sharper peak and longer tails in the distribution of E. coli results in native oysters than those in mussels and Pacific oysters. The distributions of E. coli data for the three species are however leptokurtic, i.e. have a more acute peak of E. coli results around the mean.

Table 9. Summary statistics for E. coli levels each species tested in six UK production areas. Pacific oysters

(C. gigas)Mussels

(Mytilus spp.)Native oysters

(O. edulis)Shellfish

fleshSeawater Shellfish

fleshSeawater Shellfish

fleshSeawater

N 111 — 313 — 178 —Minimum 20 <2 20 <2 20 <2Median 280 160 2,400 600 500 66Maximum 16,000 5,400 91,000 240,000 160,000 91,000Geometric mean 797 512 4,578 3,102 2,792 940Log10 standard deviation 0.641 0.908 0.780 1,112 0.723 0.799Skewness 0.150 -0.364 -0.744 -0.622 0.229 0.491Kurtosis 2.430 2.250 2.987 2.818 3.536 4.411

Linear regression of log10-transformed E. coli levels in flesh versus log10-transformed E. coli levels in seawater shows that a very significant proportion of E. coli results lie above the line of equality (Figure 6). This is expected as the mechanism of E. coli in shellfish often determines higher levels of E. coli in the shellfish flesh than those found in the seawater. Therefore, the correlation coefficient detected ( r=0.59) is indicative of the level of agreement between variables. A test for non-linearity (Ramsey RESET test) shows no significant curvature in the relationship. Overall, the tendency for E. coli levels in shellfish to increase with E. coli levels in the seawater and the wide spread of values around the regression line are evident.


Regressionflesh=1.88+0.46*water

Line ofequality

0

1

2

3

4

5

log1

0 fle

sh E

col

i /10

0g

0 1 2 3 4 5log10 water E coli / 100ml

Figure 6. Scatterplot of E. coli levels in shellfish flesh and seawater for 602 paired samples from six production areas in the UK (all species). The R2 shows that the regression accounts for 35% of the variance observed in the variables, suggesting that other factors would help explain the variance between variables. This coefficient is higher than most coefficients obtained for environmental studies published in the literature (Table 10). A “moderate” R2 is typical of data obtained under natural environmental conditions, i.e. the relationship between FIOs in shellfish and growing waters is influenced by various factors, including physiological mechanisms influencing bacterial accumulation in shellfish and environmental factors determining FIO survival and transport in the marine environment. Depending on the response of each of the catchments to the pollution event, there may be significant time delays between the event and the arrival of contaminated water at the shellfish water. For instance, it has been shown that maximum levels of E. coli in shellfish from the Dart Estuary usually occur 3–4 days after rainfall events whereas levels in the water are expected to occur with peak water levels in tributaries within 12–24h after the rainfall event (Campos et al. 2011).

Table 10. Simple linear regression models for levels of faecal indicator organisms in shellfish flesh versus water published in the literature.

ShellfishSpecies

Faecal indicator*

Regression equation R2 Reference

Clams, Oysters, Mussels

FC/EC n/a 0.3 Burkhardt III et al. (2009)

Mussels FC/EC n/a 0.2 Plusquellec et al. (1983)Mussels FC Log FC (M. galloprovincialis) 100g-1 = 0.79Log FC

(seawater) + 0.790.6 Šolić et al. (19

9)

ysterFC/EC Log10 FC (seawater) 100ml-1 =

0.7495Log10 EC(shellfish) 100g-1 + 0.52150.2 Ogburn and Wh

te (2009)†Musels

EC Log10 EC (Mytilus spp.) 100g-1 =

0.8087 Log10 EC (water)+1.7312

0.8 Kay et al. (2007)

†Oysters EC Log10 EC (C. gigas) 100g-1 =

0.4651 Log10 E. coli (water)+2.8717

0.9 Kay et al. (2007)

*FC (faecal coliforms); EC (E. coli). †Microcosm studies. n/a-not available.

Figure 7 shows the linear regression fit thematically represented by species. The good spread of E. coli results around the regression line for all three species is evident. The majority of results above the class B


threshold (4,600 E. coli 100g-1 FIL; log10 = 3.663) relates to mussel samples.

Regr ession:log( f lesh) =1. 88+ 0. 46* log( wat er )

Line ofequalit y

1

10

100

1000

4600

10 000

100K

E co

li / 1

00g

flesh

(log

10 s

cale

)

1 10 100 1000 10 000 100KE coli / 100ml water (log10 scale)

edulisgigasmussels

Figure 7. Scatterplot of E. coli levels in shellfish flesh and seawater for 602 paired samples from six production areas in the UK for each species tested.

Table 11 summarises linear regression models for E. coli levels in shellfish flesh versus water for each species. The regression model for mussels shows that E. coli levels in seawater explain a higher proportion of the variation of E. coli levels in that species than that in the model for the three species combined. This is probably due to the fact that mussel samples represent approximately 52% of the total number of samples. In contrast, levels of E. coli in Pacific oysters and native oysters explain relatively less proportion of the variation of E. coli in seawater. Furthermore, the difference between E. coli levels in mussels and in Pacific oysters is highly significant (one-way ANOVA; Scheffé’s test) whereas the difference in E. coli levels between Pacific oysters and native oysters is marginal (p=0.07).

Table 11. Summary of regression models for levels of E. coli in shellfish flesh versus seawater for each species tested.

Species

log10 of E. coli in seawaterGeometric

meanStandard deviation n

Regression equation R2

Pacific oysters (C. gigas) 4.75 2.09 111 log10 shellfish flesh = 1.821+0.299log10 seawater 0.18

Mussels (Mytilus spp.) 5.78 2.56 313 log10 shellfish flesh = 2.027+0.464log10 seawater 0.44

Native oysters (O. edulis)

4.12 1.84 178 log10 shellfish flesh = 1.999+0.389log10 seawater 0.18

All species combined 5.10 2.40 602 log10 shellfish flesh = 0.299+1.653log10 seawater 0.35

Figures 8–9 show logistic regression models of compliance with SWD G for all species tested (hereafter referred to as “pooled species” model) and for each species tested, respectively versus geometric mean of E. coli in seawater. The “pooled species” model is supported by mussel and native oyster E. coli results across the range of geometric mean E. coli values in seawater.


0

.25

.5

.75

.9

1

prop

ortio

n of

sam

ples

com

plia

nt

1 10 100 1000 10000 100Kmean log E coli per 100ml water, log scale

C gigasmuss elsO edul is

logistic curves for pooled species

Symbol s izes proportional to number of samples

Compliance curve: FC300 v water mean

Figure 8. Logistic regression model of compliance with 300 faecal coliforms per 100ml of shellfish flesh in 75% of samples versus the geometric mean of E. coli in seawater for pooled species. Each symbol represents a site average.

When fitted for each species separately, Pacific oysters achieve higher compliance rates at each water quality than mussels and native oysters (Figure 9).

gigasmussels

edul is0

.25

.5

.75

.9

1

prop

ortio

n of

sam

ples

com

plia

nt

1 10 100 1000 10000 100Kmean log E coli per 100ml water, log scale

C gigasmusselsO edul is

logistic curves by species


Compliance curves: FC300 v water mean

Figure 9. Logistic regression model of compliance with 300 faecal coliforms per 100ml of shellfish flesh in 75% of samples versus the geometric mean of E. coli in seawater for each species.

Models individually fitted for each species show a better fit than the “pooled species” model as indicated by the lower values of information coefficients [Akaike’s Information Coefficients (AIC) and Bayesian Information Coefficients (BIC)] (Appendix IV).The “pooled species” model of compliance with SWD G versus 90th percentile of E. coli in seawater are supported by a smaller range of E. coli results in water (Figure 10).


0

.25

.5

.75

.9

1

prop

ortio

n of

sam

ples

com

plia

nt

1 10 100 1000 10000 100K90%ile E coli per 100ml water, log scale

C gigasmusselsO edul is

logistic curves for pooled species


Compliance curve: FC300 v water 90% ile

Figure 10. Logistic regression model of compliance with 300 faecal coliforms per 100ml of shellfish flesh in 75% of samples versus the 90%-ile of E. coli in seawater for all species. Each symbol represents a site average.

Figure 11 shows differences in compliance curves for individual models fitted for each species tested. Pacific oysters achieve higher compliance rates (>90%) than mussels and native oysters when the 90 th

percentile of E. coli in seawater is considered.

gigasmuss els

edul is0

.25

.5

.75

.9

1

prop

ortio

n of

sam

ples

com

plia

nt

1 10 100 1000 10000 100K90%ile E coli per 100ml water, log scale

C gigasmuss elsO edul is

logistic curves by species


Compliance curves: FC300 v water 90 percentile

Figure 11. Logistic regression model of compliance with 300 faecal coliforms per 100ml of shellfish flesh in 75% of samples versus the 90%-ile of E. coli in seawater for each species. Each symbol represents a site average.

The R2 between levels of E. coli in shellfish and water obtained in the “pooled species” model demonstrates a moderate level of agreement between variables. Table 12 indicates that a geometric mean and 90 th

percentile of E. coli of 10 and 55 respectively would be equivalent to the SWD G standard.

Table 12. Estimated geometric means and 90th percentile of E. coli in seawater at 75% compliance with 300 faecal coliforms in shellfish fluid and intravalvular liquid.

Species Geometricmean

of E. coliin seawater

90th percentileof E. coli

in seawater

Samplepairs(n)

Mussels (Mytilus spp.) 8.9 102 313Native oysters (O. edulis) 8.3 64 178Pacific oysters (C. gigas) 41 492 111“Pooled species” model 9.6 55 602

Dataset 1991–1994.

The differences between water E. coli geometric means in mussel/native oysters and those in the “pooled species” model are deemed of no practical effect in terms of microbial quality of shellfish. Furthermore, the


difference between the 90th percentile in mussels and that in the “pooled species” model could be attributed to high standard deviation of E. coli results in mussels, reflecting the fact that some of the areas where mussels grow are/have been affected by high levels of contamination of faecal origin.

The ability of Pacific oysters to be compliant at higher water contamination levels could be explained by:

Differences in physiological mechanisms determining the uptake or retention of microbial

contaminants;

Different growing methods (e.g. oysters commonly grown on trestles and mussels commonly grown on

sea/riverbed); and

Differences in impacting pollution sources.

The logistic regression models computed to assess whether E. coli levels in the seawater would come under the E. coli thresholds for class B and class A are shown in Figures 7–8. The logistic models highlight different responses for each species tested. The E. coli levels in seawater for which 90% of compliance with threshold 4,600 E. coli 100g-1 FIL is predicted to be achieved as follows: 33 E. coli 100ml-1 (mussel samples), 177 E. coli 100ml-1 (native oyster samples) and 4,200 E. coli 100ml-1 (Pacific oyster samples). The high E. coli level tolerated in the water for class B compliance in Pacific oyster samples is due to the fact that only three samples failed the pass/fail test (Figure 12). In terms of compliance with the class A threshold, no observed samples met the 95% probability criteria demanded (Figure 13) and extrapolation of the model is not firmly grounded. The logistic models for mussels and native oysters are not significantly different from each other.

mussels

gigas

edulis

pass

fai l0

.1

.2

.3

.4

.5

.6

.7

.8

.9

1

Pro

babi

lity

of c

ompl

ianc

e

90% compliances 33 177 4200

1 28 164 900 10K 242KWater E coli reading (log scale)

mussels

edulis

gigas

logist ic f it s

n o te : p a s s /fa i l wo u l d p l o t a s 1 /0 fo r l o g i s t i c b u t a re s h i f te d v e rt i c a l l y to s e p a ra te th e s p e c i e s

B-class compliance

Figure 12. Logistic regression models of compliance (pass/fail) of levels of E. coli in seawater with the class B threshold (≤4,600 E. coli 100g-1 FIL) under Regulation (EC) No 854/2004. Each point represents an individual sample.


mussels

gigas

edulis

No line r eaches . 95wit hin dat a r ange

pass

fai l0

.1

.2

.3

.4

.5

.6

.7

.8

.951

Pro

babi

lity

of c

ompl

ianc

e

1 28 164 900 10K 242KWater E coli reading (log scale)

mussels

edulis

gigas

logist ic f it s

n o te : p a s s / fa i l wo u l d p l o t a s 1 /0 fo r l o g i s ti c b u t a re s h i f te d v e rt i c a l l y to s e p a ra te th e s p e c i e s

A-class compliance

Figure 13. Logistic regression models of compliance (pass/fail) of levels of E. coli in seawater with the class A threshold (≤230 E. coli 100g-1 FIL) under Regulation (EC) No 854/2004. Each point represents an individual sample.

3.2.1. Recommendation on a microbial standard for shellfish protected areas under the Water Framework Directive

The UK Technical Advisory Group on the WFD has noted that a revised microbiological standard for the purposes of the WFD may be more stringent (e.g. nearer existing G) or less stringent than the existing standard (UKTAG, 2007). From the results shown above, it becomes evident that the level of stringency required to achieve this objective is determined by the strength of the association between levels of FIOs in shellfish and those in the overlaying waters. As discussed in the interim technical report 5, the complexity of this association arises from the fact that different shellfish species growing in the same area may differ in the degree to which they accumulate FIOs. This is often attributed to biological and/or physiological characteristics of shellfish species (e.g. age, size, maturity, nutritional condition, physiological mechanisms regulating feeding and digestion) and the combined effect of prevailing environmental conditions. Furthermore, the growing methods used for the commercial production of shellfish will determine the time during which shellfish are immersed in the water column and their location within the water column and hence both their temporal and spatial exposure to microbial contaminants. On this point, the EU Scientific Veterinary Committee Working Group on Faecal Coliforms in Shellfish (1996) emphasised that, although in general both sets of legislation placed similar upper threshold contamination limit on shellfish destined for commercial depuration, it is very difficult to quantify any public health implications arising from species specific differences in FIOs accumulation levels. However, to date none of the comparisons of the EU and US systems has included considerations of pathogen monitoring data, and so assessments of differences in margins of safety have not necessarily been made on appropriate grounds.

4. Conclusions and recommendations

4.1. Sanitary profiles of selected shellfish water catchments pre- and post-improvements in sewerage infrastructure

4.1.1. Data limitations

This study highlights the inadequacy of existing data for quantitative sanitary profiling:

Catchment-derived fluxes: none of the catchments have suitable monitoring data for calculating annual, base- and high-flow FC and EN fluxes pre- and post-improvement.

STW treated effluent fluxes: for none of the key STWs are data available pre-UV installation, and


even for the post-improvement period the data available are often very limited, with no base-/high-flow separation.

ID fluxes: the existing flow/FIO database is inadequate, thereby precluding the calculation of empirically-based flux reductions at most of the IDs where improvements have been made, and providing no basis for calculating the overall fluxes derived from all the IDs within the sewerage network of each of the key STWs.

Clearly, the current empirical evidence base for underpinning policy is limited. Detailed monitoring is needed to determine the fluxes from individual sources, both before and after intervention, to allow accurate characterisation.

4.1.2. Assessment of the impact of sewerage improvements upon FC and EN fluxes

Comparison of the total FC and EN fluxes pre- and post-improvement reveals reductions of 39.83–87.98% and 35.64–93.91%, respectively. The smallest percentage improvements are in the Yealm catchment, which largely reflects the relatively small contribution that the key STW and its IDs were making to FC and EN fluxes (25.61 and 16.12%) pre-improvement. The largest improvements were recorded for Taw/Torridge. Here the treated effluents from the key STWs pre-improvement (some of which effected only primary treatment) accounted for higher proportions of the FC (81.12%) and EN (86.12%) fluxes, and therefore the potential for improvement through UV disinfection was greater. In fact, with the establishment of Cornborough STW, a much higher proportion of treated effluent is now discharged to sea outside the shellfish water catchment.

In all six catchments in which improvements have been made to the key STWs, treated effluents from these now make only very minor contributions (≤ 0.61%) to the total fluxes. In the five catchments (i.e. excluding Chichester and Conwy) where IDs have been improved, then the IDs post-improvement contribute only small proportions of the FC (≤ 4.21%) and EN (≤ 6.82%) fluxes. However, these figures are strongly dependent upon the assumptions made. Thus, under the worst-case scenarios for the IDs, the contributions of IDs following improvement in these same five catchments are all > 50%, with figures ranging from 51.47–81.72% for FC and 50.01–88.15% for FC.

4.1.3. Recommendations regarding strategies for further reducing FIO fluxes

Preliminary source-apportionment estimates suggest that both sewage- and agriculture-related sources contribute significantly to present fluxes from all seven catchments. However, in none of the catchments does one of the sources account for ≥ 90% of the flux. It may be assumed therefore that both improvements to sewerage infrastructure and the implementation of agricultural BMPs will be effective in reducing FIO fluxes.

a) Assessment of likely impacts of further improvements to sewerage infrastructureIn the case of STWs, then the most likely area of future investment would be the extension of UV disinfection to other STWs. On the basis of the generic data reported, estimates can be made of the flux reductions that could be achieved. For example, in the case of a STW with secondary treatment, then the effect of installing UV disinfection would be estimated reductions in FC and EN fluxes of 4.9 x 10 11 and 4.3 x 1010 cfu PE-1 yr-1, respectively. To evaluate the impact of improvements to IDs, data would be needed on the present volumes of flow and GM FIO concentrations, and on the reductions in flow volume resulting from improvement. Under AMP5, UV will be installed at the following STWs, with expected completion in 2013: Holton Heath STW (Poole Harbour West) and Blackburn, Croston and Walton-le-Dale STWs (Ribble). These improvements, together with the various others identified in the EA’s Pollution Reduction Plans for the seven SW catchments as being proposed or under investigation will undoubtedly lead to further reductions in FIO fluxes.

b) Assessment of likely impacts of future implementation of BMPs to reduce FIO fluxes from agricultural (i.e. livestock-related) sourcesThere is a reasonably good understanding of the effectiveness of individual BMPs in reducing FIO fluxes to watercourses from steading and field sources (CREH, 2010). By comparison, relatively few empirical data are currently available on the effectiveness of agricultural BMPs, such as streambank fencing and steading measures, in reducing FIO fluxes at the catchment scale. These have proved equivocal and inconsistent (e.g. Kay et al., 2005; Kay et al., 2007), and provide an inadequate basis for evaluating the reductions in FIO fluxes that might be achieved within the present catchments. The need for further catchment-scale investigation is presently being addressed by Defra- and SEPA-funded projects on the effectiveness of farm ponds and streambank fencing in reducing catchment fluxes; the England Catchment Sensitive Farming Delivery Initiative (ECSFDI) and the Bathing Waters and Diffuse Pollution (BWDP) project in Wales. Studies in Scotland suggest that detectable improvements in FIO concentrations may only be achieved where ≥ 30% of stream banks on livestock farms are fenced. BMP implementation sufficient to make a significant


impact in large SW catchments such as the Taw/Torridge (2094 km2) could therefore prove extremely costly. Compared with investment in further improvements to STWs and IDs, BMP implementation is therefore less easily targeted and the resulting benefits less easily evaluated.

4.2. Relationship between FIO levels in shellfish flesh and seawater

4.2.1. Simple linear regression modelling was used to study the relationship between levels of E. coli in shellfish flesh versus levels of the indicator in overlying waters from six UK production areas. The “pooled species” model explained approximately 35% of the variance observed in the variables. This is higher than that reported in the literature for some environmental studies. There was no evidence for a non-linear relationship within the relevant ranges.

4.2.2. Exploratory data analysis by means of convex hull graphical representation showed that each production area has more limited range of E. coli results in seawater and forms a more concentrated spread than the overall range of E. coli results. Therefore, separate regression lines fitted for each production area were not considered to be a valid representation of the range of contaminating levels. However, it was not possible to investigate spatial and temporal sampling bias, as data on time and precise location of sampling were not available.

4.2.3. Logistic regression models were used to determine targets for geometric mean and 90 th percentile of E. coli in seawater compliant with the SWD G standard. A geometric mean of 10 and 90 percentile of 55 E. coli per 100ml water are predicted to be equivalent to the SWD G standard (300 faecal coliforms per 100g FIL). We recommend that these thresholds should be validated on a new dataset and compared to results generated in microcosm studies.

4.2.4. Differences in compliance rates between mussels and Pacific oysters emerged from the logistic regression models indicating that these relationships are indeed complex and require further investigation, namely on the role of environmental factors and physiological mechanisms determining the uptake of FIOs by shellfish.

4.2.5. Compliance with the class B threshold (≤4,600 E. coli 100g-1 FIL) with 90% probability) in mussel, native oyster and Pacific oyster samples was predicted at 33, 177, and 4,200 E. coli levels in seawater, respectively. The high E. coli level in Pacific oyster samples reflects the fact that the vast majority of samples passed the test of class B threshold compliance. However, logistic models indicate that small changes in the proportion of samples compliant may result in relatively large changes in the required E. coli targets in seawater.

4.2.6. In terms of compliance with the class A threshold (≤230 E. coli 100g-1 FIL), none of the logistic models achieved 95% compliance rates within the range of E. coli levels detected in the water.

5. Considerations on the practical application of a microbial standard for shellfish protected areas under the WFD

In England and Wales, compliance with the SWD G has been assessed on the basis of monitoring one species from a unique sampling point deemed to be representative in each SW. The WFD specifies the approach for selecting sampling points as follows:

For bodies at risk from significant point sources, sufficient monitoring points within the body of water in order to assess the magnitude and impact of the point source and where a body is subject to a number of point sources, monitoring points may be selected to assess the magnitude and impact of these sources as a whole;

For bodies at risk from significant diffuse sources, sufficient monitoring points within a selection of the bodies in order to assess the magnitude and impact of the diffuse sources. The selection bodies shall be made such that they are representative of the relative risks of the occurrence of the diffuse source pressures, and the relative risks of the failure to achieve good surface water status;

For bodies at risk from significant hydromorphological pressure, sufficient monitoring points within a selection of the bodies in order to assess the magnitude and impact of the hydromorphological pressures. The selection of bodies shall be indicative of the overall impact of the hydromorphological pressure to which all the bodies are subject.

This categorisation could facilitate harmonisation of practices and better use of resources. For instance, there are other statutory requirements in force in the EU outside the scope of the WFD aiming to assess the potential sources of faecal contamination in coastal waters, such as sanitary


surveys being undertaken for shellfish production areas and bathing water profiles for designated bathing waters which could help inform the level of risk from point/diffuse sources of contamination. The following baseline level of information required to understand site-specific causes of and influences on FIO contamination of shellfish flesh and overlying waters:

Analyses of existing monitoring programme data with respect to spatial and temporal patterns of shellfish flesh and shellfish water quality;

Empirical data on FIO fluxes from the various pollutant sources, encompassing seasonal and base/high flow variations, to allow the construction of accurate quantitative sanitary profiles;

Characterisation of site specific processes driving the transport of microbial pollutants as they pass through the catchment to nearshore waters, namely those driven by rainfall processes (for example, CSO discharges) which determine the highly episodic and seasonal variability in microbial levels in SWs

Any significant differences in underlying levels of FIOs between commercially harvested species should be evaluated if shellfish flesh data from the Shellfish Hygiene monitoring programme is used for the purposes of informing water quality in a given SW.


References to published material9. This section should be used to record links (hypertext links where possible) or references to other

published material generated by, or relating to this project.

Burkhardt III, W., Chirtel, S., Watkins, W.D., 2009. Microbiological comparisons of the U.S. and E.U. Shellfish Safety Programs. U.S. Food and Drug Administration Center for Food Safety and Applied Nutrition (unpublished internal report).Campos, C.J.A., Kershaw, S., Lee, R.J., Morgan, O.C., and Hargin, K. 2011. Rainfall and river flows are predictors for β-glucuronidase positive Escherichia coli accumulation in mussels and Pacific oysters from the Dart Estuary (England). Journal of Water and Health 09.2: 368–381.CREH. 2010. Source strengths and attenuation of faecal indicators derived from livestock farming activities: Literature review. Unpublished Interim Report (Project WQ0203) to Defra. Crowther, J., Hampson, D.I., Bateman, I.J., Kay, D., Posen , P.E., Stapleton, C.M., Wyer, M.D. 2011. Generic modelling of faecal indicator organism concentrations in the UK. Water 3: 682–701.Environment Agency, 2009. Directive (2006/113/EC) on the quality required of shellfish waters – Article 5 programme, Chichester Harbour (Chichester Channel).Environment Agency, 2009. Directive (2006/113/EC) on the quality required of shellfish waters – Article 5 programme, Conwy.Environment Agency, 2009. Directive (2006/113/EC) on the quality required of shellfish waters – Article 5 programme, Fal Estuary.Environment Agency, 2009. Directive (2006/113/EC) on the quality required of shellfish waters – Article 5 programme, Poole Harbour (West).Environment Agency, 2009. Directive (2006/113/EC) on the quality required of shellfish waters – Article 5 programme, Ribble Estuary.Environment Agency, 2009. Directive (2006/113/EC) on the quality required of shellfish waters – Article 5 programme, Ruan Creek.Environment Agency, 2009. Directive (2006/113/EC) on the quality required of shellfish waters – Article 5 programme, Taw Torridge Estuary Mouth.Environment Agency, 2009. Directive (2006/113/EC) on the quality required of shellfish waters – Article 5 programme, Yealm.European Communities, 2000. Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 establishing a framework for Community action in the field of water policy. Official Journal of the European Communities, 22.12.2000, L327: 1–72.European Communities, 2004. Regulation (EC) No 852/2004 of the European Parliament and of the Council of 29 April 2004 on the hygiene of foodstuffs. Official Journal of the European Union, 25.6.2004, L226: 3–21.European Communities, 2004. Regulation (EC) No 853/2004 of the European Parliament and of the Council of 29 April 2004 laying down specific hygiene rules for food of animal origin. Official Journal of the European Union, 25.6.2004, L226: 22–82.European Communities, 2004. Regulation (EC) No 854/2004 of the European Parliament and of the Council of 29 April 2004 laying down specific rules for the organisation of official controls on products of animal origin intended for human consumption. Official Journal of the European Union, 25.6.2004, L226: 83–127.European Communities, 2005. Regulation (EC) No 2073/2005 of 15 November 2005 on microbiological criteria for foodstuffs. Official Journal of the European Union, 22.12.2005, L338: 1–22.European Communities, 2006. Directive 2006/113/EC of the European Parliament and of the Council of 12 December 2006 on the quality required of shellfish waters (codified version). Official Journal of the European Union, 27.12.2006, L376: 14–20. European Communities, 2006. Regulation (EC) No 1666/2006 of 6 November 2006 amending Regulation (EC) No 2076/2005 laying down transitional arrangements for the implementation of Regulations (EC) No 853/2004, (EC) No 854/2004 and (EC) No 882/2004 of the European Parliament and of the Council. Official Journal of the European Union, 18.11.2006, L320: 47–49. European Communities, 2008. Regulation (EC) No 1021/2008 of 17 October 2008 amending Annexes I, II and III to Regulation (EC) No 854/2004 of the European Parliament and of the Council laying down specific rules for the organisation of official controls on products of animal origin intended for human consumption and Regulation (EC) No 2076/2005 as regards live bivalve molluscs, certain fishery products and staff assisting with official controls in slaughterhouses. Official Journal of the European Union, 18.10.2008, L227: 15–17.EU Scientific Veterinary Committee Working Group on Faecal Coliforms in Shellfish, 1996. Report on the Equivalence of EU and US Legislation for the Sanitary Production of Live Bivalve Molluscs for Human Consumption, 23 p.Kay, D., Kershaw, S., Wyer, M.D., Lee, R. 2007. Impact of intermittent discharges on the microbiological quality of shellfish. Defra R&D Project WT0708. UK Water Industry Research Report Reference No. 07/WW/18/1, UKWIR.Kay, D., Wilkinson, J., Crowther, J., Reid, S., Francis, C., Kay, C., Hopkins, M., Watkins, J., Edwards, A.E., McDonald, A.T., Wyer, M.D., Stapleton, C.S. 2005. Monitoring the effectiveness of field and steading measures to reduce diffuse pollution from agriculture to bathing waters in the Ettrick, Cessnock, Nairn and Sandyhills catchments. Unpublished report (ref: ENV/7/4/04) to Scottish Executive Water Division.Kay, D., Aitken, M., Crowther, J., Dickson, I., Edwards, A.C., Francis, C., Hopkins, M., Jeffrey, W., Kay, C., McDonald, A.T., McDonald, D., Stapleton, C.S., Watkins, J., Wilkinson, J., Wyer, M.D. 2007. Reducing fluxes of faecal indicator compliance parameters to bathing waters from diffuse agricultural sources: the Brighouse Bay study, Scotland. Environmental Pollution 147: 138–149.Kay, D., Crowther, J., Stapleton, C.M., Wyer, M.D., Fewtrell, L., Edwards, A., Francis, C.A., McDonald, A.T., Watkins, J. and Wilkinson, J. 2008. Faecal indicator organism concentrations in sewage and treated effluents. Water Research 42: 442–454.Kay, D., Anthony, S., Crowther, J., Chambers, B.J., Nicholson, F.A., Stapleton, C.M., Wyer, M.D. 2010. Microbial water pollution: a screening tool for initial catchment-scale appraisal and source apportionment. Science of the Total Environment 408: 5649–5656.Laing, I., and Spencer, B.E. 2006. Bivalve cultivation: criteria for selecting a site. Cefas Science Series Technical Report 136.MAFF, DoH, PHLS Working Group, 1992. Bacteriological examination of shellfish. PHLS Microbiology Digest 9: 76–82.Ogburn, D.M., White, I., 2009. Evaluation of fecal pollution indicators in an oyster quality assurance program: application of epidemiological methods. Journal of Shellfish Research 28(2): 263–271.Plusquellec, A., Beucher, M., Le Gal, Y., 1983. Enumeration of the bacterial contamination of bivalves in monitoring the marine bacterial pollution. Marine Pollution Bulletin 14(7): 260–263.UKTAG, 2007. Guidance on objectives and standards for Protected Areas. Final Working Paper WP4a of the UK Technical Advisory Group on the Water Framework Directive. Available at: http://www.wfduk.org/tag_guidance/Article_06-07/guidance_protected_areas. Accessed January 2011.Younger, A.D.. Reese, R.A. 2011. E. coli accumulation compared between mollusc species across harvesting sites in England and


http://www.wfduk.org/tag_guidance/Article_06-07/guidance_protected_areas.%20Accessed%20January%202011

http://www.wfduk.org/tag_guidance/Article_06-07/guidance_protected_areas.%20Accessed%20January%202011

Wales. Cefas Report to the Food Standards Agency, January 2011.


general enquiries on this form should be made...

Documents