Biological Effects of Contaminants on Estuarine Organisms

Home , Acanthochondria cornuta, Lernaeocera branchialis, Velvet (fish disease)

BIOLOGICAL EFFECTS OF CONTAMINANTS ON ESTUARINE ORGANISMS

For

The Department for Environment, Food and Rural Affairs (Defra), Marine and Waterways Division.

FINAL REPORT

Reference as: Feist, S. W., Bateman, K, B., Bignell, J., Longshaw, M., Lyons, B. and Stentiford, G. D. (2005) Biological effects of contaminants on estuarine organisms. Final report for Defra, Marine and Waterways Division, (CDEP 84/5/31 ME3016). 141 pp.

Project leader: Dr Stephen W. Feist Cefas Weymouth Laboratory, Barrack Road, The Nothe, Weymouth, Dorset DT4 8UB, United Kingdom

Other staff: Dr Grant Stentiford, Dr Matt Longshaw, Dr Brett Lyons, Ms Kelly Bateman, Mr John Bignell. Cefas Weymouth Laboratory, Barrack Road, The Nothe, Weymouth, Dorset DT4 8UB, United Kingdom C1617/ME3106

Contents Page No

Section 1: Overview 4

1.0 Executive summary 5 1.1 Background to the program 7 1.2 Project objectives 12 1.3 Milestones 13 1.4 General Methodologies 14 1.4.1 Selection of target species 1.4.2 Sampling effort 1.5 The Aquatic Health Database (AHD) 14 1.5.1 Database structure 1.5.2 What data is recorded?

Section 2: Fish 19

2.1 Introduction 20 2.1.1 Contaminants and organ pathology 2.1.2 Parasites and populations 2.1.3 Molecular markers of exposure 2.1.4 Model fish species 2.2 Materials and methods 24 2.2.1 Histopathology and parasitology 2.2.2 Molecular pathology 2.2.3 Statistical methodologies, indices and data analysis 2.3 Results 28 2.3.1 Flounder 2.3.1.1 Histopathology 2.3.1.2 Targeted study on fibrillar inclusions 2.3.1.3 Molecular pathology 2.3.1.4 Parasitology 2.3.2 Viviparous blenny 2.3.4 Sand goby

Section 3: Crustaceans 61

3.1 Introduction 62 3.2 Materials and methods 63 3.3 Results 64 3.3.1 Shore crab 3.3.2 Brown shrimp 3.4 Discussion 67 3.4.1 Shore crab 3.4.2 Brown shrimp

C1617/ME3106

Section 4: Molluscs 81

4.1 Introduction 82 4.2 Materials and methods (Phase 1) 82 4.3 Results and discussion (Phase 1) 84 4.4 Conclusions (Phase 1) 86 4.5 Phase 2: Monthly sampling of mussels 87 4.6 Materials and methods (Phase 2) 88 4.7 Results and discussion (Phase 2) 89 4.8 Conclusions (Phase 2) 91

Section 5: Multivariate statistics 107

5.1 Discrimination using multivariate statistics 108 5.1.1 Parasites 5.1.2 Liver lesions 5.1.2.1 Within site variation over time 5.1.3 Gill lesions 5.1.4 Combined data for all sites (2002)

Section 6: General discussion and conclusions 123

6.1 Fish 124 6.1.1 Recording ‘top level’ markers of health 6.1.2 Specific pathologies and parasites 6.1.3 Synthesis and recommendations 6.2 Invertebrates and biological effects monitoring 129

Section 7: Key findings and forward look 131

7.1 Key findings 132 7.2 Forward look 135 7.3 Acknowledgements 136

Section 8: References 137

C1617/ME3106

Section 1: Overview

Cefas Contract C1617: This report describes the work undertaken by Cefas in fulfilment of the final milestone relating to the DEFRA (Marine and Waterways Division) funded project on biological effects of contaminants on estuarine organisms.

C1617/ME3106

1.0 Executive Summary

The concern over the degradation of estuarine environments is a major driver behind large-scale initiatives such as the Water Framework Directive (WFD). While such programmes are primarily concerned with the presence, absence and abundance of key species and the overall water quality where these exist, at present, it does not consider the health status of resident populations.

This report presents the outcomes of a field programme designed to assess the biological effects of contaminants on estuarine organisms. It has built upon previous work carried out by Cefas and other laboratories on the use of sentinel species (such as flounder) for investigating exposure and effect in the estuarine environment. Furthermore, it has attempted to expand upon this approach to provide baseline data for common invertebrate species found in UK estuarine and coastal environments.

Key findings –

• This study has reinforced the use of pathology for assessment of ‘top–level’ effects in marine organisms. Histopathological changes reflect the combined effects of environmental and biological factors on the study organism.

• Establishment of a relational database (the Aquatic Health database) and innovative use of multivariate statistics for such data has enabled clear discrimination of contaminated and reference sites. The discrimination season-to season and year-to-year has also been demonstrated.

• New evidence is presented, suggesting that a widespread and highly prevalent liver pathology (hepatocellular fibrillar inclusions) in flounder is associated with endocrine disruption. In addition, intersex (ovotestis) condition was detected in flounder from the Clyde and Mersey estuaries and in viviparous blennies from the Tyne estuary.

• Parasitology data has enabled discrimination of flounder populations from relatively contaminated sites with fish from reference sites. In addition, such data offers a key ability in determining estuary residency times for individual fish based on the species of marine and/or freshwater parasites present.

• The discovery of significant differential biological response in mussel populations previously assumed to be solely blue mussel Mytilus edulis. Certain populations were found to consist of M. edulis, M. galloprovincialis and a hybrid of the two species.

• Baseline data on disease levels in crustaceans was obtained. No direct contaminant related pathology detected. However, evidence of depressed immunocompetance suggested by differing prevalence of infections with yeasts, viruses and parasites at different sites was detected.

C1617/ME3106

• Results from the erythrocyte micronucleus assay using flounder appear to suggest that this assay is not sensitive enough for deployment for assessment of biological effects of contaminant exposure in UK estuaries.

• A previously un-reported pathology of oogonial and spermatogonial apoptosis in flounder and viviparous blenny requires further study for potential use as a marker of reproductive health in those species.

C1617/ME3106

1.1 Background to the program

The estuarine environment is a major sink for potentially hazardous chemical pollutants emitted from industrial and domestic sources. Inventory-based chemical monitoring programmes are restricted to identification of a limited range of contaminants and provide no information on their biological significance. Typically, the sublethal effects of exposure to environmental contaminants are assessed by measuring the responses of one, or few, biochemical or physiological systems within an organism. An individual's survival potential, however, is determined by the integrated phenotypic response of all its systems working in concert. As such, to gain a holistic view of the impact of chemical exposure on the organisms living within a particular environment, the simultaneous measurement of numerous biological variables, in a number of sentinel species, may well be the optimum approach. By taking this multiple species approach, it may be possible to determine intra- and inter-specific variability in the type of responses exhibited and thus enable the potential impacts on populations and communities to be better assessed. In addition, according to the type of pollutants impacting a site, a suite of site-specific biomarkers may be deployed.

In recent years there has been concern over the possible degradation of UK estuarine environments from chemical pollution. This has led to increased efforts attempting to evaluate the causal relationships between contaminant exposure and observable biological effects in aquatic organisms (De Flora et al., 1991). Previously, Cefas has studied UK estuarine waterways extensively with respect to both chemical and biochemical indicators of contaminant exposure (Matthiessen et al., 1993, 1998; Kirby et al., 1999; Lyons et al., 1999). These studies have allowed us to identify those UK estuaries where the present level of contamination is posing a threat to the long-term health of resident biota. A number of the contaminants present in these industrialised waterways, in particular the polycyclic aromatic hydrocarbons (PAH’s), are known to have mutagenic and/or carcinogenic properties (IARC, 1983). Significantly, a recent investigation has investigated the use of a linked series of biomarkers and has provided preliminary data on the pathology in three estuarine fish species (flounder, viviparous blenny and sand goby), including histopathology, parasite burdens, genotoxic damage and various metabolite biomarkers (Feist et al. 2001; Stentiford et al., 2003). However, there remains a paucity of information on the health status of aquatic organisms in contaminated UK estuaries and on the prevalence of hepatic lesions in fish species other than flounder and for shellfish (crustacean, molluscan) pathology in general. The identification of genetic damage associated with elevated contamination levels is a significant finding. There is a need for ongoing studies addressing the consequences of such exposure in relation to the induction of cancer and general indicators of aquatic organism health (Lyons et al., 2004a, 2004b) as well as for assessments on the abilities of populations to recover from prior exposure (Leonard et al., 1999; Nicholson and Leonard, 1999).

One of the main goals of an environmental monitoring programme is to provide, where possible, a somewhat simplistic answer to a complex problem.

C1617/ME3106

Coupled with this, the elucidation of drivers responsible for the different responses of populations and individuals to contaminants is a long-term goal of all involved in protection of aquatic organisms and in ensuring the quality and health of ecosystems (Moore et al., 2004). The first step in this process is to determine if there are differential responses by populations of resident or migratory species in contaminated and reference sites. If differences are shown to exist, the next step is to determine what biological characteristics are responsible for these observed differences and then to determine which extraneous factors such as pollutants, seasonality, interannual variation, etc. are driving the observed changes. In order to deal with these complex datasets (that may include biological, physicochemical and contaminant information), multivariate statistical approaches have been developed. The results obtained by using these multivariate approaches have been utilised in the current project to begin to determine which biological endpoints are altered in animals from impacted environments and as such, which ones are likely to be useful as markers for contaminant exposre. Such an approach leads to the development of novel markers that have environmental relevance and that accurately reflect the status of individuals and populations.

Fish diseases and pathology, whether caused by infectious agents, environmental factors or xenobiotics, are increasingly used as indicators of environmental stress at the population level (ICES, 1997). The use of histological and histochemical biomarkers of toxic injury, dysfunction and carcinogenesis is now well established. These approaches provide a powerful toolbox for detection and characterisation of the biological end points of toxicant and carcinogen exposure. The utility of these lesions as indicators for the assessment of health of wild fish populations has been demonstrated in several European and North American studies (Kranz & Dethlefsen, 1990; Myers et al., 1990, 1991, 1992, 1994, 1998; Köhler, 1991; Köhler et al., 1992; Bucke & Feist, 1993; Vethaak & Wester, 1996, Bogovski et al., 1999; Lang et al., 1999; Stentiford et al., 2003). In addition, several laboratory and mesocosm studies have demonstrated causal links between exposure to xenobiotics and the development of toxicopathic hepatic lesions (Malins et al., 1985 a & b; Varanasi et al., 1987; Stein et al., 1990, 1992; Moore & Myers, 1994; Vethaak et al., 1996).

Vitellogenin is commonly employed as a biomarker of effect for exposure to endocrine disrupting chemicals (Allen et al. 1999, Janssen et al. 1997). Both male and female flounder hepatocytes have relatively non-specific cell surface oestrogen receptors. Levels of vitellogenin in blood plasma of male and immature animals, or un-seasonally elevated levels in females are an effective means of measuring environmental impact of these compounds (Kleinkauf et al. 2004). Vitellogenin levels are routinely measured by means of an Enzyme-Linked Immunoabsorbant Assay (ELISA) (Janssen et al. 1997) and long term trend studies have centred on the measurement of this plasma protein in flounder from several UK estuaries (Kirby et al., 2004).

Of the Phase I biotransformation enzymes, CYP1A is the commonly used biomarker of exposure to sub-lethal levels of PAHs (Bogovski et al. 1998). CYP1A levels meanwhile are measured as the O-deethylase activity of liver

C1617/ME3106 microsome preparations against the substrate ethoxyresorufin (Burke et al. 1985). It is easy to measure and statistically more sensitive than other biotransformation enzymes (Van der Oost 1997). As a result the EROD assay is now routinely deployed as part of the UK NMMP and other studies (MPMMG 1998, Kirby et al. 1999).

There are however concerns about the sensitivity of the EROD assay. The EROD enzymic activity of CYP1A has been shown to be susceptible to inhibition by a number of compounds. This can include both direct inhibition by CYP1A inducers as well as other compounds present within the contaminant mixture found in the environment (Petrulis & Bunce 1999, Willet et al. 2001). Activity can also be affected by seasonal ambient water temperatures and sexual cycle of both males and females (Hylland et al. 1998, Kirby et al. 2004).

Parasites are an integral component of any ecosystem and it has been suggested that every fish carries at least one parasite (Lester, 1990). The parasite fauna of a fish or shellfish can accurately reflect the life history of that individual and therefore provide some indication of the health of the population and of the ecosystem from which it is derived (MacKenzie et al., 1985; Williams et al., 1992; MacKenzie & Abaunza, 1998). Parasites of fish can be affected by pollutants in a number of ways, either during a free-living phase or whilst within a host. Indeed, parasites are useful biomarkers that can, in some circumstances, be more sensitive to environmental stressors than their fish host (Landsberg et al., 1998). The number of parasites on or in a fish host can increase as a result of an increased success of the intermediate host in a polluted site, or as a result of impaired immune responsivity in the fish. Alternatively, parasite numbers can decrease due to toxic effects on the free-living stages of the parasite or on the survival of intermediate host, or a change in the physiology of the fish host. The prevalence of external parasites with direct lifecycles often increases in response to a stressor such as pollutant. For example, the number of trichodinid parasites infecting the gills have been shown to increase in fish captured from sites contaminated by crude oil (Khan et al., 1994). Furthermore, Marcogliese et al. (1998) found an increase in the prevalence and abundance of the monogenean parasite Gyrodactylus sp. in fish exposed to sediments contaminated with high levels of PCBs and PAHs. They attributed this to the impaired immunological status of the hosts.

Previous studies of parasite communities of estuarine fish as biological indicators have demonstrated the utility of this approach (Longshaw et al., 2001; Stentiford et al., 2001). These studies demonstrated in general terms that the prevalence of parasites with direct lifecycles increased while those with indirect lifecycles decreased in response to environmental degradation. Additionally, parasite community diversity and infracommunity structure were reduced in estuaries subjected to environmental contamination.

Even though most studies have focussed on the use of vertebrates for environmental hazard assessment, biomarker measurements are equally feasible in invertebrate samples (Fossi et al., 1997). Indeed, there are several

C1617/ME3106

reasons why invertebrates may be preferable for use in this ecotoxicological discipline, not least being their relatively static habit when compared to fish (Depledge & Fossi, 1994), that they are common, and that they tend to bioaccumulate toxicants present in their environment. In addition, their biochemical, physiological and histological characteristics are sufficiently well known to discriminate exposed from non-exposed individuals (Viarengo, 1993). Most studies on biomarkers in invertebrates have concentrated on the role of the mixed function oxidase system in detoxifying xenobiotics (see Snyder, 2000), while a few studies have evaluated the genotoxic effect of contaminants (Mortimer & Hughes, 1991). A considerable literature also exists on the histopathological changes that are manifested in invertebrates following exposure to a range of exogenous aquatic pollutants. In general, the tissues most affected by contaminant challenge are the hepatopancreas and the gills (for review see Gardner, 1993). The presence of external disease has also been used as an indicator of industrial pollution. In particular, shell disease in various crab and shrimp species has been directly linked to exposure to contaminated estuarine and marine waters and sediments (see Sindermann, 1979). Results from numerous controlled laboratory exposures of shellfish (crustaceans and molluscs) to toxicants have shown that histopathological changes occur in the organ and tissue systems of these animals, relatively few field studies have included shellfish histopathology in the suite of monitoring tools employed. Most studies of this type have centred on the use of the common mussel (Mytilus edulis) (Lowe et al., 1981; Moore et al., 1987; Lowe, 1988; Wedderburn et al., 2000) while studies on crustacean histopathology are scarce (Couch, 1978; Sindermann 1979; Overstreet, 1988).

The investigative approach adopted in this proposal aimed to provide a wealth of data on the general health status of ecologically important estuarine fish and shellfish species at a variety of biological levels of complexity, including effects on parasite fauna. The proposed programme of work was designed to continue and extend upon previous studies (Feist et al., 2001) and to work towards extrapolation to the population-level effects of exposure to anthropogenic contaminants. In the current program, the health status of common estuarine crustacean species (Carcinus maenas and Crangon crangon) has been added to the sampling programme. In addition, the inclusion of mussels (Mytilus spp) into the program allows for a more robust and holistic assessment of the impact of estuarine contaminants on resident biota.

Due to its capacity to visualise effects in situ, histopathology remains an essential tool for biological effects studies at the whole organism level, providing an integrative endpoint of contaminant and pathogen exposure. As such, a central theme to this program has been the use of histopathology as a ‘top level’ biomarker that can potentially be used to assess the usefulness of associated biochemical and physiological markers. It may also provide the basis for estimating ‘biological effect’ of exposure to particular contaminants at defined concentrations. Histopathology forms the basis for the newly constructed ‘Aquatic Health Database’ and we see this as central tool for integrating disparate datasets (e.g. chemistry, biomarkers, environment) in

C1617/ME3106

future programs. Such an approach will move us closer to defining ‘ecological quality’ at specific field sites.

The examination of biological effect of contaminant exposure trends can be considered at three important levels. Firstly, site-to-site comparisons allow for ‘hotspot’ and ‘reference’ sites to be identified. This has been the traditional approach to monitoring since it allows for wider spatial coverage and for targeting of effort to the most heavily impacted sites. The second approach, based on year-to-year monitoring allows for degradation or improvement in specific geographical sites to be related to particular intervention measures (e.g. reduced sewage input, siting of a new factory). As such, year-to-year monitoring plays an important role when allied with site-to-site monitoring since it allows environmental managers to track changes at particular sites over time. Finally, recent monitoring efforts using sentinel species (such as flounder and particularly the mussel, as demonstrated in this report) have highlighted the significant biological changes that may occur in these sentinels over a given season. In this way, it may not be feasible to compare data collected during one part of the season with that collected during another. To address this issue, effort during the project established a season-to-season monitoring study on the mussel taken from two geographically distinct sites.

The current study provides more robust data on the health status of UK estuarine environments and establishes a work plan for moving individual- level health assessments to population and ecosystem level assessment. As such, we have carried out an assessment of temporal trends in the health status of flounder and for the first time, shellfish species, using a suite of techniques including macroscopic pathology and multi-organ histopathology and markers of genetic damage. Furthermore, we have attempted to integrate marker data by carrying out multivariate analyses based on the PRIMERTM software. Such an approach can be employed to link potentially disparate datasets to provide higher-level analysis of site, season and year specific discrimination in health status. The application of this software has been dependent on the establishment of the Aquatic Health Database (AHD), also in the current program. The AHD is now established as the centralised data recording system into which biomarker data pertinent to aquatic ecosystem health can be deposited. It will play a key role in future studies of this type.

C1617/ME3106

1.2 Project objectives

1. To extend, within the framework of the UK NMMP, observations on the prevalence of external diseases, parasites and hepatic pathology in selected estuarine fish and shellfish species at ‘contaminated’ and ‘reference’ locations in the UK.

2. To determine the utility of using shellfish species (crustaceans and molluscs) as bioindicators of contaminant effects. Specifically to identify pathological lesions and markers of genotoxic damage that may be used for monitoring purposes.

3. To determine the utility of liver and gill histopathology to detect adverse biological effects of contaminants in non-flatfish species and to identify which species and lesion types may be suitable for routine monitoring purposes.

4. To determine the utility of the application of molecular epizootiology biomarkers in selected estuarine fish and shellfish. Particularly, in linking contaminant exposure, DNA damage and gene expression patterns to tissue-level alterations.

5. To implement the use of parasite fauna of target fish and shellfish species as an ecological quality indicator.

C1617/ME3106

1.3 Milestones

Milestone Task Due Date Assessment of archive shellfish material, literature survey, selection of target species. 1 14 June 2002 Completion and submission of first interim report. Completion of sampling exercise, preparation 14 November 2 of samples for analysis. Completion and 2002 submission of second interim report. Additional miletstone: Special survey and 2a report on fibrillar inclusions in flounder 14 May 2003 hepatocytes Assessment of shellfish material from Oct 3 2002, submission and completion of third 14 May 2003 interim report. Assessment of all material collected in Oct 2002. Completion of sampling exercise, 14 November 4 completion and submission of fourth interim 2003 report Assessment of shellfish material from Oct. 5 2003, submission and completion of fifth 14 May 2004 interim report. Assessment of all material collected in Oct 2003. Completion of sampling exercise, 14 November 6 completion and submission of fourth interim 2004 report. Targeted assessment of material collected in 30 May 2004 7 Oct 2004. Production and submission of final (amended) report.

C1617/ME3106

1.4 General methodologies

1.4.1 Selection of target species

An aim of the current contract was to identify suitable non-flatfish and shellfish (crustacean and molluscan) species for the assessment of external diseases, parasites, organ pathology and gene mutations and to determine a sampling strategy which minimised the effects of confounding factors such as season and age. The current study utilised previous research carried out by CEFAS, which has shown that the flounder, viviparous blenny and sand goby are suitable sentinels for exposure to estuarine contaminants. These species are generally known to be common to most UK estuarine sites and can be captured using the same sampling methodology. Invertebrates, with their abundance, ubiquity and relative sessility, show considerable potential for use as environmental sentinels. Shore crab and brown shrimp are common in all UK estuaries and can be captured using the same methodology as that used for capture of the fish species mentioned above. Molluscan species have also been identified as having considerable potential for use as environmental sentinels. Mussels (Mytilus spp.) were sampled during the current study.

1.4.2 Sampling effort

A number of transitional water sites were selected for sampling during the current project. Selected sites included the Clyde, Thames, Mersey, Alde, Tyne, Forth and Exe estuaries and Southampton Water. In addition, an inshore marine site at Brancaster Statithe was sampled for mussels. Fish and shellfish samples were collected in Autumn 2002 and Autumn 2003, before offshore migration of flounder during the winter months. Mussel samples were collected monthly from September 2004 to May 2005. All motile animals (Fish and crustaceans) were caught using a standard 2m-beam trawl deployed from a range of fishing vessels at the selected estuarine sites. Samples of mussels were collected at the low water mark from respective sites.

1.5 The Aquatic Health Database (AHD)

As with all large-scale survey-based programmes there are specific informatics requirements associated with the need to store and analyse the data collected. The AHD approaches this problem from a multidisciplinary direction and address’s the requirements of not only MWD and Cefas but also any other research body that could utilise the pathological data interface to better understand and interpret their observations.

The AHD will continue to develop beyond the lifetime of this program and will become a central repository for storing laboratory and field data. The AHD is currently designed using a Microsoft Access database. Access was chosen due to its availability across the Cefas network and its potential for easy migration to a more structured database system in the near future as required. Data mined from the AHD can then be analysed using more specialised

C1617/ME3106

statistical packages (such as PRIMERTM) to provide relatively simplistic advice lines using complex multivariate data.

1.5.1 Database structure

Microsoft Access is a competent relational database tool but users require extensive knowledge of the database system to interrogate the data effectively. However, it is versatile and flexible in its handling approaches. The AHD system combines the various benefits of Access (functionality, user friendly Graphical User Interface (GUI), cost and simplicity) with those of specialist downstream statistical software (e.g. PRIMERTM) for analysis of integrated datasets. In this context, Access is used primarily as a data storage and retrieval tool and not as a complete analysis package.

At present the database has limited GUI support (outside of the remit of this contract). The ‘front end’ of the AHD will eventually be designed to ensure user friendliness and reduced chance of data entry errors. The AHD will develop as an environmental data archive and may even be applicable to data entry in the field (e.g. on board research vessels). Hence an integrated menu driven GUI will be developed to enhance user familiarity.

Regarding the data, the database is based on a ‘fluidic’ approach. The entry of data is restricted to a particular format, which must be adhered to for each field. However the formats can be changed to meet changing needs without loss of information. Data from common fields and frequently used search parameters such as specimen location, which require specific details to match to make the data search effective, are linked and fully amendable.

Aside from these obviously functional tables and reports allowing data harmonisation across the species groupings, they allow for a single and unified approach to categorising data, which is a simple and logical. This will aid the future dynamism of the AHD. Outputs from the database will take the form of reports and queries tables, which will then be exported to suitable analyses packages.

1.5.2 What data is recorded?

There is a wide diversity of data recorded in the AHD including the severity grading of specific diseases, parasitological data, weight, sex and origin of individuals. At present the database contains data categorised into 540 fields for each specimen. Not all data is recorded for each specimen (relevant fields are dependent on the species being examined), however the wide variability of the recording parameters gives a versatile and highly functional research tool, which will undergo continual development and improvement.

The data recorded is first divided into species groups. Each ‘species group’ is distinct, appearing as ‘fish‘, ‘mollusc’ or ‘crustacean’ since the data parameters collected for each group are different. Splitting the data in to the species groups and the origin of the material allows the databases to remain compact and rapid to interrogate. The first dataset to populate the AHD has

C1617/ME3106

been the ‘fish’ data. This consists of several species including flounder sampled under previous and current programs. Data relating to the offshore sentinel species dab (Limanda limanda), collected during the UK National Marine Monitoring Programme (NMMP) cruises has also been utilised (data not shown here). Statistical analyses of this integrated data is now providing an insight into differences between sites, seasons and years and will be useful in identifying potential new biomarkers for discrimination of heavily impacted and reference sites. In addition, it allows for a more intelligent approach to identifying areas for research as well as providing a sound basis for the creation of meaningful health indexes.

Finally, the AHD can also be utilised as a specimen tracking system whereby researchers can rapidly locate specimens with a specific attributes (e.g. by location, pathology, species, size, sex etc). We invisage extension of the AHD to incorporate a ‘tissue bank’ whereby individual specimens can be cross- referenced to archived samples (e.g. frozen plasma or tissue, formalin-fixed material etc) useful for further research purposes.

C1617/ME3106

Section 2: Fish

Fish diseases and pathology, whether caused by infectious agents, environmental factors or xenobiotics, are increasingly used as indicators of environmental stress at the population level since they depict the endpoint of a stress cascade that links exposure to effect. Techniques such as histopathology are used to assess toxic injury, carcinogenesis and other diseases and as such provide a powerful tool for monitoring health status of wild fish populations. Furthermore, molecular techniques that allow for sensitive diagnosis of damage at the genomic level provide further evidence for effect. A recent investigation (Feist et al., 2001)) investigated the use of a linked series of biomarkers for assessing health status of three estuarine fish species (flounder, viviparous blenny and sand goby) from UK estuaries. Here, we have extended this study by using histopathological, parasitological and genotoxicological markers for assessing health status. Due to its suitability for monitoring, the flounder has been the central species for investigation. Linking ‘top level’ effects (e.g. presence of liver tumours or intersex gonads) to subcellular biomarkers and contaminant burdens at particular sites remains logistically challenging. However, with an improved understanding of the life history of sentinel species, more efficient methods for collating potentially disparate datasets (e.g. via the Aquatic Health Database) and multivariate approaches to statistical analysis, complex patterns may be deciphered for an improved understanding of processes within estuarine environments.

The results of these studies on fish pathology are published in:

Lyons, B.P., Stentiford, G.D., Green, M., Bignell, J., Bateman, K., Feist, S.W., Goodsir, F., Reynolds, W.J., Thain, J.E. (2004). DNA adduct analysis and histopathological biomarkers in European flounder (Platichthys flesus) sampled from UK estuaries. Mutation Research 552, 177-186

Associated publications relating to this work:

Stentiford, G.D., Longshaw, M., Lyons, B.P., Jones, G., Green, M., Feist, S.W. (2003). Histopathological biomarkers in estuarine fish species for the assessment of biological effects of contaminants. Marine Environmental Research 55, 137-159.

Bateman, K.S., Stentiford, G.D., Feist, S.W. (2004). A ranking system for the evaluation of intersex condition European flounder (Platichthys flesus). Environmental Toxicology and Chemistry 23, 2831-2836.

Lyons, B.P., Bignell, J., Stentiford, G.D., Feist, S.W. (2004). The viviparous blenny (Zoarces vivparus) as a bioindicator of contaminant exposure: application of biomarkers of apoptosis and DNA damage. Marine Environmental Research 58, 757- 761.

C1617/ME3106

2.1 Introduction

In recent years, increasing emphasis has been placed on the evaluation of the causal relationships between contaminant exposure and observable biological effects in aquatic organisms (De Flora, Bagnasco & Zanacchi, 1991). As such, fish diseases and pathologies, with a broad range of aetiologies, are increasingly being used as indicators of environmental stress since they provide a definite biological end-point of historical exposure (Matthiessen, Thain & Law et al., 1993).

2.1.1 Contaminants and organ pathology

Histopathological biomarkers provide powerful tools to detect and characterise the biological end points of toxicant and carcinogen exposure (Hinton & Lauren, 1990; Wester & Canton, 1991; Hinton, Baumen & Gardener et al., 1992; Moore & Simpson, 1992). As such, the utility of histological lesions as sensitive and reliable indicators of the health of wild fish populations has been demonstrated in several European and North American studies (Kranz & Dethlefsen, 1990; Myers, Landahl & Krahn et al., 1991; Myers, Olson & Johnson et al., 1992; Myers, Stehr & Olson et al., 1994; Myers, Johnson & Hom et al., 1998; Köhler, 1991; Köhler, Deisemann & Lauritzen, 1992; Bucke & Feist, 1993; Vethaak & Wester, 1996; Bogovski, Lang & Mellergard, 1999; Lang, Mellergard & Wosniok et al., 1999). Several laboratory and mesocosm studies have also demonstrated causal links between exposure to xenobiotics and the development of toxicopathic hepatic lesions (Malins, Krahn & Brown et al., 1985a; Malins, Krahn & Myers et al., 1985b; Varanasi, Stein & Nishimoto et al., 1987; Stein, Reichert & Nishimoto et al., 1990, Stein, Collier & Reichert et al., 1992; Moore & Myers, 1994; Vethaak & Jol, 1996; Vethaak, Jol & Meijboom et al., 1996). Following studies of this type, it is generally accepted that certain liver lesions in marine flatfish can be induced by environmental contaminants and that these represent an ecologically relevant biological endpoint of exposure to pollution.

2.1.2 Parasites and populations

In addition to direct contaminant effects on tissues, diseases caused by parasites and pathogens are integral components of any population. The parasite fauna of individual fish reflect the life history of that individual and their interaction with other conspecifics in the population. As such, parasite status provides some indication of the health of the population and of the ecosystem from which it is derived (MacKenzie et al., 1985; Williams et al., 1992). The relationship between the presence of a particular parasite infection and contamination may however be complex: Parasites of fish can be affected by pollutants in a number of ways, either during a free-living phase or whilst within a host. Indeed, parasites are useful biomarkers that can, in some circumstances, be more sensitive to environmental stressors than their fish host (Landsberg et al., 1998). The number of parasites on or in a fish host can increase as a result of an increased success of the intermediate host in a

C1617/ME3106

polluted site, or as a result of impaired immune responsivity in the fish. Alternatively, parasite numbers can decrease due to toxic effects on the free- living stages of the parasite or on the survival of intermediate host, or a change in the physiology of the fish host. The prevalence of external parasites with direct lifecycles often increases in response to a stressor such as pollutant. For example, the number of trichodinid parasites infecting the gills have been shown to increase in fish captured from sites contaminated by crude oil (Khan et al., 1994).

Previous studies of parasite communities of estuarine fish as biological indicators have demonstrated the utility of this approach. These studies demonstrated in general terms that the prevalence of parasites with direct lifecycles increased while those with indirect lifecycles decreased in response to environmental degradation. Additionally, parasite community diversity and infracommunity structure were reduced in estuaries subjected to environmental contamination.

2.1.3 Molecular markers of exposure

We have previously used the flounder in field monitoring programmes and have demonstrated that populations of fish collected from industrialized UK estuaries display a range of pathologies, which can be linked to contaminant exposure, including ovotestis and hepatic nuclear/cellular pleomorphism and altered hepatic cell foci (Stentiford et al., 2003). Supporting evidence from biomarker research has suggested that contaminants such as xenoestrogens and PAHs may in part contribute to these observed conditions (Lyons et al., 1999, 2004). Therefore, as part of the current programme of research we have employed a suite of biomarkers including hepatic and ovarian DNA adducts and the erythrocyte micronucleus assay, to further investigated the pressures genotoxic and mutagenic contaminants place on fish living in polluted environments.

The erythrocyte micronucleus assay (MN) has previously been used with varying success as a cheap and rapid in vivo assay for detecting genotoxic damage in a variety of fish species (for review see, Al-Sabti and Metcalfe, 1995). Micronuclei are small secondary nuclei visible during interphase that originate from chromosomal fragments that are not incorporated into the nucleus during mitosis. Although they may occur spontaneously, the induction of MN is commonly used to detect genotoxic damage resulting from exposure clastogenic or aneugenic contaminants in aquatic organisms (Carrasco et al., 1990). However, to date no information is available as to its suitability when used in conjunction with flounder collected from UK estuarine and coastal areas. Therefore, during 2002 we initiated a blood sampling programme using flounder collected from UK estuaries and investigated the potential for the erythrocyte MN assay to be used as a biomarker for mutagenic and genotoxic exposure.

In addition, previous studies have indicated that the levels of genotoxin-DNA complexes (adducts) in selected aquatic bioindicator species could be used as biomarkers of environmental contamination (Harvey et al., 1997; Lyons et

C1617/ME3106

al., 1997, 1999, 2004). The majority of studies to date have focused on the formation of DNA adducts in tissues such as the liver and gills of exposed aquatic organisms. However, our previous research has indicated that contaminant related pathologies are commonly observed in the reproductive tissues of fish from impacted estuaries (Stentiford et al., 2003; Lyons et al., 2004). Therefore, along with the formation of DNA adducts in primary target tissues such as the liver we have investigated the potential for DNA damage to occur in the reproductive tissues of flounder collected from selected UK estuaries.

Vitellogenin is commonly employed as a biomarker of effect for exposure to endocrine disrupting chemicals (Allen et al. 1999, Janssen et al. 1997). Both male and female flounder hepatocytes have relatively non-specific cell surface oestrogen receptors. Levels of vitellogenin in blood plasma of male and immature animals, or un-seasonally elevated levels in females are an affective means of measuring environmental impact of these compounds (Kleinkauf et al. 2004). Of the Phase I biotransformation enzymes CYP1A is the commonly used biomarker of exposure for sub-lethal levels of PAHs (Bogovski et al. 1998).

Vitellogenin levels are routinely measured by means of an Enzyme-Linked Immunoabsorbant Assay (ELISA) (Janssen et al. 1997). CYP1A levels meanwhile are measured as the O-deethylase activity of liver microsome preparations against the substrate ethoxyresorufin (Burke et al. 1985). It is easy to measure and statistically more sensitive than other biotransformation enzymes (Van der Oost 1997). As a result the EROD assay is now routinely deployed as part of the UK NMMP and other studies (MPMMG 1998, Kirby et al. 1999).

There is however concerns about the sensitivity of the EROD assay. The EROD enzymic activity of CYP1A has been shown to be susceptible to inhibition by a number of compounds. This can include both direct inhibition by CYP1A inducers as well as other compounds present within the contaminant mixture found in the environment (Petrulis & Bunce 1999, Willet et al. 2001). Activity can also be affected by seasonal ambient water temperatures and sexual cycle of both males and females (Eggens et al. 1996, Hylland et al. 1998, Kirby et al. 2004).

The measurement of vitellogenin and CYP1A mRNA by means of real-time PCR is less technically demanding than currently used methods, allows earlier detection of exposure induced response and avoids problems of protein inhibition. Real-time analysis has previously been applied to the measure of vitellogenin in rainbow trout (Oncorhynchus mykiss) (Celius et al. 2000), and to CYP1A in the European flounder (Dixon et al. 2003). However both studies employed the use of a non-specific inter-chelating dye (SYBER- Green) to monitor the PCR reaction. The use of TaqMan probes utilizing FRET Chemistries (Fluorescent Resonance Energy Transfer) offers a more specific, robust method for quantitative real-time PCR.

C1617/ME3106

2.1.4 Model fish species

The flounder is one of the commonest of the flatfish species found in UK waters and is the only European flatfish species known to penetrate well into estuaries (even to freshwater). Found in inshore and estuarine locations at depths of below 50 metres, it is particularly common in areas of reduced salinity. Its range in the NE Atlantic extends from Norway and the Baltic to Morocco and the Meditteranean, with subspecies in the Adriatic and Black Seas. Distant migrations (of up to 200 miles) have been recorded, though local migrations of less than 20 miles are thought to be more common (some tagging studies have shown that individual specimens may not move significantly over several years). Migrations are generally associated with the breeding cycle and lead to an inshore habit during the summer and an offshore (or deeper water) habit during the winter months. Offshore migration precedes spawning in the spring (February to June), this occurring in deeper waters offshore. Eggs hatch at temperatures above 12oC; these float to the surface and sink as development proceeds. Metamorphosis to the flatfish form occurs at a length of 15-30 mm, after which they become benthic (inshore or estuarine). Growth depends on feeding rate but generally equates to 8, 14, 19 and 24 cm in the first four years of life. Sexual maturity in males is around 11cm and in females is around 17cm.

Their benthic habit, ubiquity and tendency to express biomarkers of exposure to contaminants have identified the flounder as a useful model for investigating the effects of anthropogenic pollution in estuarine systems. Studies have centred on the measurement of a range of biological endpoints, including gross disease and histopathology (Köhler, 1990; Vethaak et al., 1996; Vethaak & Jol, 1996; Vethaak & Wester, 1996; Bogovski et al., 1999; Lang et al., 1999; Simpson et al., 2000; Stentiford et al., 2003, Bateman et al., 2004; Lyons et al., 2004), and a range of biochemical markers indicative of exposure to particular groups of contaminants (Goksøyer et al., 1996; Matthiessen et al., 1998; Allen et al., 1999a and 1999b; Kirby et al., 1999; Lyons et al., 1999; Kirby et al. 2004; Lyons et al., 2004). More recently, genomic and proteomic markers of contaminant exposure have been investigated (Williams et al., 2003; Sheader et al 2004).

However, as flounder are known to undergo seasonal offshore migrations, the incorporation of species with limited migrational tendencies into biological effects monitoring programmes has recently been recommended (Kirby et al., 2000). The viviparous blenny (Zoarces viviparus) and sand goby (Pomatoschistus microps) are such a species that reside in contaminated and clean estuarine environments around the UK. Previous studies have demonstrated the potential use of these species as sentinels for biological effects monitoring programmes (Celanderet al., 1994; Waring et al., 1996; Stentiford et al., 2003; Lyons et al., 2004).

The increasing emphasis on the assessment and monitoring of estuarine ecosystems has highlighted the need to deploy appropriate biological indices

C1617/ME3106 and measurements for these locations. This study reports data for multi-organ histopathology, genotoxic markers and parasitology in flounder from several UK estuaries with differing contaminant burdens. In addition, histopathology and genotoxic data for viviparous blenny from selected UK sites is presented. A later section in this report uses data presented here to carry out multivariate statistical analyses of site-to-site differences in individual and population status of flounder.

2.2 Materials and methods

A number of transitional water sites were selected for sampling during the current project. Selected sites included the Clyde (Spring and Autumn 2002), Thames (Autumn 2002), Mersey (Autumn 2002), Alde (Autumn 2002), Tyne (Autumn 2002), Inner Forth (Autumn 2002), Outer Forth (Autumn 2002) and Southampton Water (Autumn 2002). Low numbers of fish were also collected from Belfast Lough (data not included). Fish were collected using a standard 2m-beam trawl deployed from a range of fishing vessels at the selected estuarine sites. Samples collected under the DMECS (Development of a National Marine Ecotoxicological Analytical Control Scheme - CDEP 84/5/284) were also utilized in the current programme. All Fish were kept alive in running estuarine water prior to euthanasia and dissection.

2.2.1 Histopathology and parasitology

The gill, liver and gonad were dissected from up to 50 flounder from each site and prepared for histology. In addition, the head and guts were removed and preserved for parasitological assessment. Where present, viviparous blenny and sand goby were fixed whole, and samples taken for histological and parasitology assessment. All fish samples were fixed in NBF for 24 hours before being transferred to 70% industrial methylated spirits (IMS) for transport and storage. Tissues for histopathology were processed using standard protocols; thin sections (3-5μm) were obtained using a standard rotary microtrome and were stained with haematoxylin and eosin (HE). Stained sections were analysed by standard light microscopy and digital images were captured using the Lucia™ Screen Measurement System (Nikon, UK). Where appropriate, samples identified using standard histology were serially re-sectioned and labelled using the terminal deoxynucleotidyl transferase-mediated dUTP-biotin nick end labelling (TUNEL) technique. This technique allows for the detection of 3’-OH groups at the end of apoptotic and necrotic DNA fragments found in dying cells.

2.2.2 Molecular pathology

For the erythrocyte micronucleus assay (MN), blood smears were collected from the caudal vein of flounder and blenny via a heparinized needle. An aliquot of 10μl was smeared onto pre-cleaned microscope slides air-dried, fixed in 90 % ice-cold methanol and stained with 20 % giemsa solution for 5 minutes. Slides were scored for the presence of MN according to the criteria of (Al-Sabti and Metcalfe, 1995).

C1617/ME3106

For detection of DNA adducts, samples of ovary and liver were removed from flounder and snap frozen in liquid nitrogen before transfer to storage at -80°C. Ovarian DNA was prepared from crude nuclei homogenates using a phenol/chloroform extraction method essentially as described by French (1996) with minor modifications. Frozen (-80oC) ovary tissue samples (125- 250 mg) were homogenised in 1.7 ml of 10 mM Tris/100 mM EDTA, pH 7.4 and then centrifuged at 4oC for 10 min at 6000 rpm to pellet the crude nuclei. The resultant pellet was then re-suspended in 850 µl of 1 % SDS/20 mM EDTA (pH 7.4). To remove contaminating glycogen and RNA, the samples were incubated with α-amylase, RNaseA and RnaseT1. Proteins were then digested by the addition of 0.4 mg of proteinase K in 40 µL of 1 M Tris-HCl (pH 7.4). The DNA was purified from contaminating proteins by a series of sequential organic solvent extractions using phenol, phenol/chloroform /isoamylalcohol (25:24:1) and chloroform/isoamylalcohol (24:1), after which it was precipitated by the addition of 0.1 volumes of 5 M NaCl and 100 % ice- cold ethanol. The DNA pellet was then washed in 70 % ethanol before finally re-suspending in 10 mM Tris/1 mM EDTA, pH 7.4 (TE buffer). All dissolved DNA isolates were quantified spectrometrically (absorbance ratio A260/A280 nm) and deemed to be free from RNA and protein contamination. The DNA samples were then adjusted to a final concentration of 1 µg/µl and stored at - 32 80ºC prior to P-postlabelling.

DNA adducts were determined using the standardized nuclease P1 version of 32 the P-postlabelling assay, as described previously (Phillips and Castegnaro, 1999). Briefly, 10 μg samples of DNA were digested to deoxyribonucleoside 3'-monophosphates in a total volume of 9.5μl of digestion mix (6mU/μl calf spleen phosphdiesterase (Calbiochem, UK), 36mU/μl micrococcal nuclease, 100mM sodium succinate, 50mM CaCl2). 2μl of the DNA digest was diluted and held for the labelling of the normal undamaged nucleotides for subsequent quantification. To the remaining 7.5μl of digested DNA 1μl sodium acetate buffer (final concentration 40mM), 1μl ZnCl2 (final concentration 0.2mM) and 1μl nuclease P1 solution (final concentration 0.31 μg/μl) was added to digest normal nucleotides. The reaction mixture was incubated at 37ºC for 30 minutes then stopped by the addition of 1μl Tris solution. Adducted and normal nucleotides were then labelled separately, but simultaneously, for 30 minutes with labelling buffer (200 mM bicine NaOH, pH 9, 100 mM MgCl2, 100 mM dithiothreitol, 10 mM spermidine), 6 units T4 polynucleotide kinase (30U/μl; Amersham) and 50 μCi of [γ-32P] ATP (>7000 Ci/mmol, ICN).

The adducted deoxyribonucleoside-3'-5'-biphosphates present in the sample were then purified and separated from their normal undamaged counterparts using multidimensional anion exchange thin layer chromatography (TLC), on 10 x 10 cm polyethyleneimine (PEI)-cellulose plates (Camlab, Cambridge, UK). The levels of DNA adduct radioactivity were determined using an AMBIS radioanalytical scanning system (LabLogic, Sheffield, UK). Upon the quantification of both the adducted nucleotides and the normal nucleotides, the relative adduct labelling values were calculated and converted to the

C1617/ME3106

number of adducted nucleotides per 108 undamaged nucleotides. Appropriate negative and positive DNA controls were analysed throughout the studies as described by Harvey and Parry (1998). Dr Alan Hewer and Prof. David Phillips (Institute of Cancer Research, Sutton UK) kindly provided the positive control, which consisted of BaP labelled DNA.

All real-time PCR technologies rely upon the detection and quantification of a fluorescent reporter, the signal of which increases in direct proportion to the amount of PCR product in the reaction. All primer-probe sets used for TaqMan real-time PCR were designed using the Primer Express software package (ABI UK), and are given below. The optimum size product size for TaqMan is <80bp. The CYP1A primer pair was designed such that they spanned an intron-exon boundary. Unfortunately such detail is not currently available in the sequence information for either vitellogenin or the 18s gene.

5’ Primer Name Sequence Tm ºC Label 18s Reverse 5’-TGC AGC AAC TTT AAG ATA CGC TAT TG-3’ 60.1 None 18s Probe 5’-TGC CAG CAG CCG CGG TAA TTC-3’ 63.7 JOE CYP1A Forward 5’-GAC GAG AAG ATT GTA GGA ATT GTC AAC-3’ 52 None CYP1A Reverse 5’-CGA CCA TGA CAG GGC AGT AGA-3’ 51 None CYP1A Probe 5’-CCT GTT TGG AGC TGG ATT CGA TAC CG-3’ 66.4 FAM VIT Forward 5’-ATC CGT GCA GTA CGA GTT TGG-3’ 50 None VIT Reverse 5’-GGC ATT GCT GAT CCT CAG AAG-3’ 50 None VIT Probe 5’-AGC GAG CTT CTC CAG ACA CCC G-3’ 65.8 FAM

The availability of multiple reporter dyes for TaqMan analysis enables the analysis of multiple genes in the same tube (multiplexing). It is essential that the reporter dyes are compatible. The chosen reporter dyes were 6-FAM (λmax 518nm) and JOE (λmax 554nm). All probes were labeled with the Tamera quencher dye (λmax 582nm) at the 3’-end. All primers and probes were purchased from MWG.

Total RNA was isolated from flounder liver tissue (5 fish per sex, per site) (~100mg) using the SV Total RNA Isolation system according to the standard protocol (Promega Ltd, Southampton UK). Yield was estimated by spectroscopy at 260nm and was approximately 2μg/μl.

First strand cDNA synthesis was performed using Superscript II RT, purchased from Invitrogen, Paisley UK. RNA was mixed with random primers (300ng) and dNTPs (0.5mM). The reaction was gently mixed by pipetting and incubated at 65ºC for 5 minutes to denature mRNA secondary structures and allow primer annealing. First strand synthesis buffer (5x) was then added and the reaction incubated at 45ºC for 2 minutes. Twenty units of Superscript II RT were then added and the reaction incubated at 45ºC for 1½ hours. The reaction was terminated by addition of EDTA (33mM) and NaOH (66mM) and cleaned using the Qia-quick spin columns purification system (Qiagen, Crawley UK). There are a number of quantification methods using real-time and TaqMan technologies, which result in either absolute or relative quantification of the

C1617/ME3106

message. Relative quantification is applicable to the majority of applications and allows for simplified experimental protocols and data analysis through the comparative CT (ΔΔCT) method (CT= PCR cycle threshold). The first step in the calculation is to designate one group of fish as the calibrator sample to which all other samples gene expression is expressed as a relative value. In this experiment from fish collected from the Alde reference estuary were used as the base line and fish from the Tyne expressed as a relative value.

Step 1: Calculate the ∆CT = CT target gene – CT reference gene (18s). Step 2: Calculate the ∆∆CT = Mean sample ∆CT – Mean calibrator ∆CT. Step 3: Calculate relative expression = 2-∆∆CT

∆CT between target and reference gene in a single sample is independent of starting concentration of PCR template and affectively normalizes for this variable. For further information and mathematical basis behind the above calculation protocol, refer to the ABI 7700 User Bulletin 2.

2.2.3 Statistical methodologies, indices and data analysis

Histological methods and diagnostic criteria follow those developed by ICES and according to the quality assurance requirements required under the Biological Effects Quality Assurance in Monitoring (BEQUALM) programme (Feist et al., 2004). Assessments of the intersex condition in flounder followed the criteria set out by Bateman et al. (2004).

All parasites observed in fish were enumerated and identified to species level where possible. The measures of the level of infection used in this study were the prevalence (defined as the number of fish in a sample infected with a particular parasite species, divided by the number of fish examined in that sample, expressed as a percentage) and abundance (defined as the total number of individuals of a particular parasite species in a sample of hosts divided by the total number of fish examined in that sample). The range of parasite species was defined as the minimum and maximum number of individual parasites found in fish in a particular sample. Intensity levels for Trichodina sp., and Lernaeocera branchialis were graded as present/absent. Sample prevalences for all species of parasite were compared for statistical significance using the χ2 test and analysis of variance.

A number of ecological parameters and indices were used to classify the sample populations. Parasites were classified as allogenic (developmental stages are found in different environments and most frequently have birds as the final host) or autogenic (species complete their whole development cycle in the same environment as their final hosts) parasite species (Esch et al., 1988). Core species were defined as ones in which the prevalence >60%, secondary species with prevalences between 41 to 60%, satellite species with prevalences between 6 and 40% and rare species were defined as those with prevalences <6%. Diversity indices were determined for parasite populations and compared for statistically significant differences in values. Parasite 27

C1617/ME3106

population structures were compared using multidimensional scaling based on presence/absence of parasites or on prevalence of all parasites in order to discriminate host populations. Linear discriminant analysis (LDA) was applied to the data from the previous contract (DETR report CDEP 84/5/287) and proved useful in discriminating flounder populations and thus the effects of contamination on parasite communities and their respective hosts.

Data relating to flounder parasites and pathologies collected under the current MWD and the previous contract investigating 'estuarine pathology biomarkers' (CDEP 84/5/287) were extracted from the aquatic health database (AHD) (see section 3.0 below). Data were imported into PRIMER (ver. 6 ß r6) and variously converted according to data type. The main transformation overall was to treat each pathology or parasite as present/absent within an individual fish. Individual fish data was subsequently converted to prevalence within the population sub-sample. Analyses of the data include non-metric multidimensional scaling (NMDS) and principal component analysis (PCA). Two major analyses were carried out on the data. These were intended to detect (i) changes within a site over time and (ii) differences between sites within a given year (see Section 5 of this report).

2.3 Results

Results for flounder are split in to two discrete sampling periods. The first, in Spring 2002 included only the Clyde estuary while the second, in Autumn of 2002 included the Clyde and all other sites as listed in Section 2.2.

2.3.1 Flounder

2.3.1.1 Histopathology

Clyde Estuary (Spring 2002). No visible cases of external disease were recorded on any of the 30 flounder sampled from the Clyde site. Internally, no visible liver lesions were recorded. It should be noted however that the mean size of flounder captured and sampled at the Clyde site was 17.7 ± 0.4 cm. This is considerably smaller than the recommended length of 20 cm that has been employed during external disease measurements in previous studies.

Multi-organ histopathology revealed a number of lesions, some of which have been reported to be of toxicopathic aetiology. Over sixty-six percent of flounders captured from the Clyde estuary displayed at least one of the liver lesion categories listed under BEQUALM guidelines (Feist et al., 2004). While no lesions of definite toxicopathic aetiology were found, the livers of 56.7 % of flounders contained hepatocytes with proliferated rough endoplasmic reticulum (RER) (Fig. 1). These ‘fibrillar inclusions’ were associated with apparent proteinaceous deposits in most fish (Fig. 2) and with megalocytosis (nuclear and cellular polymorphism) in 20 % of fish.

The most significant pathology recorded in the gonads of flounder captured from the Clyde estuary was the presence of intersex (ovotestis) in one male fish (9.09 % apparent prevalence). The condition manifested as multiple foci

C1617/ME3106

of pre-vitellogenic oocytes amongst otherwise normal fields of immature testicular material (Fig. 3). In this fish, ovotestis was bilaterally distributed and of similar severity in each testicular lobe (Bateman et al. 2004).

A number of parasites were observed in histological sections of tissues and organs from flounder captured in the Clyde estuary. In the gill, a Rickettsia- like infection was seen in one fish (3.73 % apparent prevalence) (Fig. 4) and a digenean infection was seen in two fish (7.41 % apparent prevalence). Furthermore, infestation of the gill by a ciliate of the genus Trichodina was seen in 88.8 % of flounder (Fig. 5). An unidentified microsporidian parasite was recorded in the kidneys of 72.4 % of flounder captured from the Clyde estuary. The parasite is found in discrete xenoma’s in the Bowman’s space of the glomeruli (Fig. 6). Inflammatory reactions, apparently toward these xenoma’s, and often involving the melanomacrophage aggregates of the kidney interstitial tissue are also observed in several cases.

All sites (Autumn 2002). The sites with the largest proportion of flounder containing at least one BEQUALM liver pathology category were the Mersey, Tyne, Clyde and Forth estuaries (all over 80 %). Lower proportions were recorded in flounder from the Thames and Outer Forth sites (below 70 %), while the site with the lowest proportion of flounder with at least one lesion was the Alde estuary (below 25 %) (Fig. 7). In addition to variations in the proportion of fish containing at least one lesion category, flounder captured from the Alde estuary also contained a significantly lower total number of lesion categories in the liver than those captured from all other sites (p<0.001). Once again, the livers of flounder captured from the Mersey, Tyne, Clyde and Inner Forth estuaries displayed the highest number of lesions (Fig. 8).

Non-toxicopathic lesions such as those associated with cell turnover (apoptosis, necrosis, regeneration) and immune-related functions (melanomacrophage aggregates, inflammation) were seen in flounder from all sites. Although it is difficult to associate higher prevalence of these lesion types with specific sites, generally, the lowest prevalence was seen in flounder captured from the Alde estuary, with higher prevalence (particularly of melanomacrophage aggregates, inflammation and necrotic foci) seen in fish from the other sites (Fig. 9). Since at least some of these lesion types are known to be associated with pathogen and parasite infections, care must be taken when associating their presence with site-specific factors (such as contamination).

In addition to pathologies associated with cell turnover and the immune system, two other significant pathologies of unknown aetiology were also seen in flounder captured from several sites. Hepatocellular fibrillar inclusions were seen in flounder captured from all sites apart from the Alde estuary. Particularly noteworthy was the high prevalence of this condition in flounders captured from the Mersey, Clyde and Inner Forth sites (Fig. 10). Hepatocellular fibrillar inclusions were either present as inclusions of isolated cells or as a progressively diffuse lesion involving the majority of hepatocytes. In severe cases, cells containing inclusions appeared degenerate and often

C1617/ME3106 contained eosinophilic inclusions reminiscent of phospholipid droplets. Eventually, foci of cells containing these inclusions became necrotic. Associated with the hepatocellular fibrillar inclusions was a high prevalence of nuclear pleomorphism. (Fig. 11)

All four types of foci of cellular alteration (FCA); basophilic (b), eosinophilic (e), vacuolated (v) and clear cell (cc) were observed in flounders captured during the current study. The prevalence of bFCA was highest in flounder captured from the Mersey, Inner Forth and Thames estuaries (though this lesion was also recorded in low prevalence in fish captured from the Alde). The prevalence of eFCA was highest in the Thames, Tyne and Mersey estuaries (with a slightly lower prevalence at the Outer Forth site). Vacuolated FCA were seen at all sites (apart from Outer Forth) and were even seen at relatively high prevalence (6 %) in flounder from the Alde estuary while ccFCA was only detected in flounder from the Tyne and Mersey estuaries. The Mersey was the only estuary that contained flounder that exhibited all four categories of FCA (Fig. 12).

Benign and malignant neoplasms of the liver were rarely observed in the current study. Basophilic adenoma was diagnosed in one flounder captured from the Thames estuary while another fish, also captured from the Thames, exhibited hepatocellular carcinoma.

Although no cases of intersex were recorded in flounder captured from the various sampling sites included in the current study, a subsequent sampling exercise to the same site on the Mersey estuary in January 2003 revealed an apparent prevalence of 21.4 % (3 fish out of 14 males) of intersex fish.

A number of pathologies were observed in the gills of flounder captured from each site. Most of these pathologies were associated with the presence of pathogens or parasites either attached to the gill lamellae or associated with the epithelial cells. Epitheliocystis, a pathology associated with an infection by a Chlamydia-like organism (CLO) was common in flounder captured from the Outer Forth site (26 %), with lower prevalence observed at the Tyne, Southampton, Forth, Thames and Mersey sites respectively. In many cases, hyperplasia and other gill abnormalities were seen to be associated with infection by larvae of the parasitic copepod Lernaeocera branchialis. As such, almost all of the flounder captured from the Outer Forth sites were infected (94 %), with relatively lower prevalence at the Mersey, Forth, Alde, Tyne and Thames sites respectively. Interestingly, L. branchialis infection was not recorded at the Southampton Water site by histology. Two other parasitic infections were also associated with the gill of flounder captured at several sites. The first, the metacercarial stage of a digenean trematode inhabiting the secondary lamellae was particularly common in flounder from the Alde estuary (38.6 %). Its prevalence was relatively less at the Southampton, Clyde, Inner Forth, Outer Forth, Mersey, Tyne and Thames sites respectively. The second parasite, Trichodina sp. inhabiting the inter-lamellar spaces was present in all flounder captured from the Clyde estuary and at a relatively high prevalence in flounder from the Mersey (91.8 %), Tyne (83.6 %), Thames (60 %), Alde (56.8 %), Forth (40 %) and Southampton (4.65 %) sites. Trichodina

C1617/ME3106

sp. was not observed on the gills of flounder captured from the Forth control site (see Table 1).

2.3.1.2 Targeted study on fibrillar inclusions

A total of fifty flounder were obtained from the Clyde and Mersey estuaries. Samples of liver for histological evaluation at the light microscope level were taken from all fish and liver samples from twenty fish from each site were fixed for subsequent ultrastructural examination

In affected fish, most hepatocytes were seen to contain conspicuous fibrillar arrays (Fig. 13). Under closer examination these arrays could be seen to occupy most of the cellular volume and often displaced the nucleus to a peripheral location close to the cell membrane (Fig. 14). The fibrillar arrays appeared basophilic in haematoxylin and eosin stained sections and were orientated in various directions within the cell. At the light microscope level it was not possible to discern organelle pathology, however it has been noted that in livers exhibiting high levels of hepatocellular fibrillar inclusions, necrotic foci may occur. This suggests that in more extreme cases the pathological changes within hepatocytes are sufficient to result in lysis of the cell.

In some cases affected hepatocytes also contained varying numbers of small spherical eosinophilic droplets (Fig.15). At the light microscope level these tended to be present mainly within the cytoplasm but occasionally small droplets were observed apparently embedded between the fibrils of the fibrillar arrays. Further investigation revealed that these droplets were indistinguishable from those present in the condition referred to in the literature as 'phospholipoidosis'. From examination of material obtained from this contract we have been able for the first time to identify a sequential pathology. From the development of fibrillar inclusions, production of eosinophilic droplets, which increase in number and apparently coalesce, to the more overt condition shown in Figure 16 referred to here as phospholipoidosis. Careful examination of archive material together with samples obtained during this study has confirmed that in many cases of phospholipoidosis the hepatocytes can be seen to contain remnants of fibrillar inclusions. Since there has not been any histochemical evaluation of the eosinophilic droplets it is inappropriate to consider that the material is in fact a phospholipid. Indeed the staining characteristics alone suggest the presence of a protein/lipid complex.

Samples obtained from the Mersey estuary yielded excellent material for study. The data presented here are from fish exhibiting moderate degrees of the condition where not all hepatocytes contained inclusions. From resin embedded semi-thin 1μm sections stained with toluidine blue it was clear that several isolated hepatocytes within the liver parenchyma showed atrophy and increased basophilia in association with the presence of fibrillar arrays. These specimens were selected for further study by electron microscopy.

The main features of affected cells were increased osmiophilia and general condensation of the cytoplasm. The characteristic fibrillar arrays were clearly

C1617/ME3106 visible (Fig. 17). From observations of several affected hepatocytes the most likely origin of these arrays is from pathologically altered endoplasmic reticulum (ER). This can be seen in Fig. 18 where condensed ER is continuous with the anasmotosing network of 'tubules', which in that case are sectioned transversely. Other cytoplasmic structures present include contains numerous small vesicles containing granular substance, larger amorphous lipid inclusions, smaller electron dense droplets and lysosomes containing membranous material consistent with effete organelles. In both examples the mitochondria show marked degenerative changes, including swelling and loss of cristae. It is clear that affected cells are themselves undergoing a degenerative process, this providing supportive evidence for the findings at light microscope level where necrotic foci have been associated with increased levels of hepatocytes with 'fibrillar inclusions'.

Hepatocytes displaying mild pertubations of the ER were also observed (Fig. 19). In these cases the damage to mitochondria were much less severe and the overall osmiophilia of the cytoplasm was reduced to near normal levels. However, proliferation of the ER was still present together with the presence of small electron dense droplets seen in more severely affected hepatocytes. Closer examination of the ER has revealed that the cisternae have become extended to form the tubule-like structures seen in longitudinal section. Numerous free ribosomes were observed in between the cisternae together with the droplets referred to above with none clearly attached to the surface of the ER (Fig. 20). In this case the mitochondria appeared normal.

2.3.1.3 Molecular pathology

Table 2 shows the results for peripheral erythrocyte micronucleus (MN) frequencies in flounder from UK estuaries. This study suggests that the erythrocyte MN assay as applied to European flounder is not of the required sensitivity to be used as a routine biomarker of contaminant exposure within UK estuaries.

The total DNA adduct levels from individual samples of flounder ovarian tissue are displayed in Fig. 22. All of the sites studied contained at least one sample that contained DNA adduct profiles indicative of exposure to genotoxic contaminants. While flounder collected from the contaminated Mersey, Thames and Tyne contained higher levels of ovarian DNA adducts when compared with the control reference site (Alde) they were not significantly elevated (P>0.05). Overall the level of hepatic DNA adducts detected were similar to those reported previously for the liver of flounder collected from UK estuaries (Lyons et al., 1999; CEFAS, 2003) (Fig. 21). At the majority of industrialised sites (Southampton, Thames, Clyde, Tyne and Mersey) the predominant hepatic DNA adduct profile consisted of diagonal radioactive zones (DRZs), which is the characteristic profile obtained following exposure to complex mixtures of aromatic and/or hydrophobic genotoxins, such as those formed by PAHs. In contrast, flounder collected from the Forth, Alde and Belfast lacked DRZs with only background levels of DNA damage being observed

C1617/ME3106

The most striking inter-site differences where observed between the Cyp1A transcript levels where male flounder from the Tyne had a 266-fold increase over male fish sampled from the Alde (Fig. 22a). Similarly, female flounder from the Tyne when compared with female fish from the Alde displayed a 15- fold increase in Cyp1A transcription. The induction of Cyp1A is known to be induced in response to PAH exposure and its elevation in fish collected from the Tyne supports previous work showing that PAH’s present in the Tyne sediment are bioavailable to fish residing there (Woodhead et al., 1999; Lyons et al., 1999; 2004). Results for the induction of Vtg where less conclusive with Tyne females showing an increase in Vtg production (31 fold), while male fish displayed a decrease in transcription (2.3 fold). However, further information on the reproductive state of these fish would be required before any clear conclusions could be drawn from these findings.

The data from this pilot study should be viewed with caution due to the low number of fish examined. However, these initial results demonstrate the suitability of real time PCR technique for detecting gene expression differences between clean and polluted sites using number of important biomarker genes. This will approach will prove particularly powerful when combined with data being generated from the use of DNA microarrays (such as that being generated under the current NERC Post-Genomic and Proteomics Programme: Identifying and defining the bases of individual and population susceptibility and adaptation to environmental pollutants in fish: An integrated transcriptomic, proteomic and metabolomic approach), where it is envisaged that the application of DNA arrays containing over 10,000 candidate marker genes will highlight the need to develop site specific biomarker assays. The use of real time PCR will allow assays to be set up for individual marker genes in the fraction of the time it would take to develop traditional enzymatic or immunological based assays.

2.3.1.4 Parasitology

Clyde Estuary (Spring 2002). Flounder analysed for parasitology from the spring sample revealed a total of 10 parasite species including the acanthocephalan Echinorhynchus gadi. Levels of Trichodina spp. in Clyde flounder were 80 %. Values for the Shannon-Wiener diversity index (H’ = 1.858) and the Brillouin’s diversity index (H = 1.700) were slightly lower than previously found from fish at this site, with an overdominance of Diplostomum spathaeceum. The infracommunity structure of the parasite populations was calculated to include all parasite species found and then, only to include the helminths. When all parasites are included in the analysis, the modal value for infracommunity structure was 2, with all fish parasitised by at least one species, and the majority parasitised with 2 or 3 species, with a maximum of 4 species in any one flounder. However, when only helminths are considered, the modal value shifts to 1, with a maximum value of 2 in any one fish. Additionally, 8 fish in the sample were not infected with any helminths. Only two parasite species were considered as core species with the remaining 8 species considered as either satellite or rare.

C1617/ME3106

All sites (Autumn 2002). Flounder from seven sites were examined for parasitology. A total of 24 parasite species were recorded from the combined sites. These comprised of one Protistan, nine Digenea, eight Nematoda, two Acanthocephala, two Cestoda and two Copepoda. Most of the parasites were located in the gut and gills (See Tables 3, 4 and 5 for details of species found, ecological characteristics, lifecycle patterns, prevalences, range and abundance data). Variations in a number of measures were found between sites including diversity indices, prevalences, abundance, infracommunity structure and parasite component community structure. Additionally, the use of multidimensional scaling (MDS) demonstrated the ability to discriminate three of the heavily contaminated sites from the reference site and three “intermediate” sites (see Section 5).

Most sites had at least two core species, usually represented by Trichodina sp. and L. branchialis (Fig. 22). The exception to this was flounder collected at Southampton water that possessed three core species represented by one nematode, a digenean and a copepod. The core parasite species by prevalence in flounder from the Alde were Trichodina sp. (70%) and L. branchialis larvae (100%). The main species by abundance were D. minutus (mean abundance 3.63) and Diplostomum sp. (mean abundance 1.77). The maximum number of individual parasites found on one fish was 26 (C. heterochrous). The core species by prevalence in flounder from the Mersey were Trichodina sp. (100%), L. branchialis larvae (100%) and C. heterochrous (73%). The main species by abundance was C. heterochrous (mean abundance 18.0) with the maximum number of individual parasites found on two fish: 142 and 105 (C. heterochrous). In the Clyde the core species by prevalence in flounder were Trichodina sp. (100%) and D. minutus (90%). The main species by abundance was also D. minutus (mean abundance 5.43) and Zoogonoides viviparus (mean abundance 4.50). The maximum number of individual parasites found on one fish was 81 (Z. viviparus). The core species by prevalence in flounder from the Tyne were Trichodina sp.(100%) and L. branchialis larvae (100%). The main species by abundance was Z. viviparus (mean abundance 2.20). The maximum number of individual parasites found on one fish was 13 (Z. viviparus). The core species by prevalence in flounder from the Forth were L. branchialis larvae (97%), Trichodina sp.(83 %) and Z. viviparus (60). The main species by abundance were Z. viviparus (mean abundance 5.93) and C. heterochrous (mean abundance 1.77). The maximum number of individual parasites found on two fish were 36 and 33 (Z. viviparus). The core species by prevalence in flounder from Southampton were Z. viviparus (90%), D. minutus (80%) and A. cornuta (60%). The main species by abundance also Z. viviparus (mean abundance 24.93) and D. minutus (mean abundance 9.47). The maximum number of individual parasites found on three fish were 88, 70 and 70 (Z. viviparus). Lastly, from the Thames the core species by prevalence in flounder were L. branchialis larvae (100%), Trichodina sp. (100 %) and Diplostomum sp. (80%). The main species by abundance was Diplostomum sp.(mean abundance 13.17), Acanthocephalan larvae sp. (mean abundance 7.24), Z. viviparus (mean abundance 2.34) and C. heterochrous (mean abundance 1.90). The maximum number of individual parasites found on one fish was 60 Acanthocephalan larvae.

C1617/ME3106

Highly statistically significant (p<0.001) differences in parasite prevalences between sites were noted for eight parasites, including C. heterochrous, D. minutus, Contracaecum sp., Acanthocephalan larvae sp., Z. viviparus, H. communis, Diplostomum sp., and S. baccatum. Furthermore, four parasite species show significant differences (p<0.05) in prevalences between sites (Table 6).

The number of metazoan parasite species varied for each site. The Inner Forth contained 19 species of parasites, followed by the Thames (14), and 11 parasites species were found in flounder form the Alde, Mersey and Tyne. Flounder from Southampton had 9 species and the Clyde had the fewest with only 7 species of parasites (Table 7). The diversity of the metazoan parasites as expressed by H' was highest at the Inner Forth site (H' 1.98), which was the site that also showed to have the highest species richness of (d = 2.94). The lowest diversity index was found at the Mersey site (H' 0.58), with the lowest evenness value of (J' 0.24).

Comparison of values of diversity between individual sites and the Alde indicated that the diversity values were significantly different (p<0.001) between sites (See Table 8). No difference in diversity was noted between flounder caught in the Alde and the Tyne, despite some differences in the types and prevalences of parasites in fish form these sites.

Only six species of parasite were common to all sites; Trichodina sp., L. branchialis, C. heterochrous , Z. viviparus , P. varius and Diplostomum sp. and most parasites isolated were autogenic, only being able to complete their life cycle within the aquatic environment.

The frequency of infra-community classes varied between sites (Fig. 23). At all sites, all fish possessed at least two parasite species per individual fish. The range of parasite numbers per site was 2-11 (on average the range was between 3-7 parasites per fish). For the Alde, flounder had a maximum of 7 parasite species per fish, with 30% having a modal value of 5 parasites species per fish. In the Mersey and Forth the maximum number of parasite species per fish was 11 for each, both sites had the modal value of 5 species per fish at 27 % and 37 % respectively. For the Clyde, the maximum number of species per fish was 5, with 43 % for each of two and three parasite species per fish, the modal value. Southampton had a maximum of 6 species of parasites per fish, whilst 3 parasites per fish was the modal value at 40 % and in the Thames the maximum number of species was 8 and 9 per fish, with 66 % spread between 4, 5 and 6 species per fish.

2.3.2 Viviparous blenny

Clyde Estuary (Spring 2002). The liver from the majority of fish (17/29) showed moderate vacuolation of hepatocytes (Vac II) with 8/29 exhibiting Vac III hepatocytes. Ten of the fish examined showed no abnormalities. Of the remainder, seventeen fish showed hepatocellular nuclear polymorphism (Fig 24 and 25). Individual additional fish exhibited vacuolated focus of cellular

C1617/ME3106

alteration (vFCA) and coagulative necrosis. Gonadal pathology was detected in both male and female fish examined. In females, oocytic atresia was encountered in ten fish and apoptosis was visualised from H&E stained sections (Fig. 26) and by TUNEL staining utilising a fluorescent labelled probe (Fig. 27). Two fish exhibited evidence of muscle necrosis. No abnormalities were detected in the kidney or gill.

All sites (Autumn 2002). Viviparous blennies were only captured in large enough amounts for analysis from the Clyde estuary (n = 25). The liver from the majority of fish exhibited morphology consistent with macrovesicular steatosis (Vac III vacuolation status). Whether such morphology is normal in this species cannot be speculated until sufficient numbers of fish from other sites are examined. In addition, the liver of several fish from the Clyde estuary also contained hepatocytes with pleomorphic nuclei. No fish collected in Autumn 2002 exhibited foci of cellular alteration. Mild gonadal pathology was detected in both male and female fish examined. In females, oocytic atresia was encountered while in males, spermatogonial apoptosis was recorded. Only three parasite species were recorded in these fish from the Clyde, including Diplostomum sp. (32%), nematode larvae sp. (12%) and the digenean P. varius (8%). Due to the insufficient sample size no further analysis was carried out.

2.3.3 Sand Goby

Clyde Estuary (Spring 2002). A total of thirty fish were examined histologically. The vacuolation status of hepatocytes was determined for each fish. Only one fish exhibited hepatocytes with the lowest category of vacuolation (Vac I), 55% with Vac II and 39% with Vac III. Over 26% of fish harboured hepatocytes displaying nuclear pleomorphism (Fig. 28). Other pathologies recorded in gobies from the Clyde were granuloma formation (13.2%) (Fig. 29), hepatocellular necrosis (3.3%) and hepatocellular regeneration (3.3%) (Fig. 30). Two fish exhibited gill aneurysms at low level and oocyte atresia was seen in three fish. No significant changes were observed in the eye, brain, spleen or skin. Parasitic infections were detected in several fish. The most prevalent being Kudoa sp. infections affecting the somatic musculature. A total of 36.6% (11 fish) exhibited this infection with six of those displaying a vigorous host inflammatory response to the parasite (Fig. 31). Nematode infections in the mesentary and gut elicited little host response, nor did digenean parasites resident in the digestive tract. A single fish harboured a myxozoan infection in the renal tubules.

All sites (Autumn 2002). Sand goby were collected from the Clyde, Mersey, Thames and Alde estuaries and from Southampton water. Ten fish were examined for histopathology from each site. While some inflammatory pathologies were recorded in the liver of some fish, it was not possible to ascertain whether significant differences existed between fish captured at different sites. The gonad of one goby captured from the Clyde estuary was tentatively identified as being of the intersex condition, though confirmation is problematic due to the association of inflammatory lesions with the organ. The most significant differences between fish captured at the different sampling

C1617/ME3106 sites was in their pathogen and parasite fauna. A high prevalence of goby from the Thames (33.3 %) and Southampton Water (50 %) sites exhibited gill epitheliocystis, probably caused by a similar CLO to that observed infecting flounder (see Fig. 4). CLO infection was not observed in goby captured from the other sites. Digenean trematode metacercaria were found infecting the gill (Clyde), liver (Alde, Southampton Water) and muscle (Alde, Thames, Southampton Water) from several sites. Interestingly, digenean infections were not observed in any goby captured from the Mersey estuary. Finally, myxozoan infections were recorded in the muscle and kidney of goby captured from Southampton Water (kidney infection, 80 %) and the Clyde estuary (kidney infection, 20 %; muscle infection, 66.7 %). Myxozoan infections were not observed in goby captured from the other sites (Table 9).

Gobies for parasitological examination were obtained from the five sites; Alde, Mersey, Clyde, Southampton and the Thames. Numbers of fish examined for parasitology ranged from 21 to 30. Eleven parasite species were identified. In the Alde four species were identified. The highest prevalence was for P. varius (15%), then Contracaecum sp. (10%), A. simplex and Z. viviparus both had a prevalence of 5%. All the abundances were <0.25. In the Mersey four species were identified. Prevalence data showed that Cestode sp. were the highest at 19%, followed by both Contracaecum sp. and Z. viviparus were 10% and lastly A. simplex (5%). All abundances were <0.25. For the Clyde only two species were identified. The prevalence for P. varius was 59% (mean abundance 2.69) and for Z. viviparus 18%. The gobies from Southampton had 6 species identified. The highest prevalence was for A. demeli (92%). Both the larvae of A. demeli and S. baccatum showed prevalences of 46%. Next followed P. varius (33%) and Cestode sp. (13%). The abundance was highest for A. demeli (mean abundance 11.67). Lastly, in the Thames the gobies contained 8 species of parasites. The highest prevalence was for A. demeli larvae (67%), then Cestode sp. (27%), L. gibbosus (23%), A. demeli (17%), Contracaecum sp. (7%) and lastly H. aduncum and Z. viviparus each having a prevalence of 3%. The highest abundance was obtained for A. demeli larvae (mean abundance 3.6). The maximum number of an individual parasite species found in one fish was 28 (A. demeli larvae). No further analysis was carried out on the parasite data collected from gobies due to the low numbers of representative species isolated and the limited variability in prevalences and abundances noted between sites.

1 2

3 4

5 6

13 14

15 16

17 18

19 20

C1617/ME3106

7 Proportion of flounder with liver exhibiting at least one BEQUALM category lesion 1 0.9 0.8 0.7

n 0.6 0.5

Proportio 0.4 0.3 0.2 0.1 0 Alde Tyne Clyde Forth Mersey Thames

Forth Con. Site Southampton

8 Mean number of BEQUALM category lesions per liver, in flounder 2 1.8 1.6 n 1.4 1.2 1 0.8 0.6 Mean number of lesio 0.4 0.2 0 Alde Tyne Forth Soton Clyde Mersey Thames

Forth Con Site

C1617/ME3106

Proportion of flounder with livers exhibiting lesions associated with cell turnover and immune-related function

0.7 Coagulative necrosis Apoptosis MMA Inflammation Regeneration

0.6

0.5

0.4

0.3 Proportion

0.2

0.1

0 Alde Clyde Tyne Forth Mersey Thames

Forth Con. Forth Sit e Southampton

10 Proportion of flounder with liver exhbiting hepatocellular fibrillar inclusions

0.9 0.8 0.7 0.6 0.5 0.4 Proportion 0.3 0.2 0.1 0 Alde Tyne Clyde Forth Mersey Thames

Forth Con. Site Southampton

C1617/ME3106

11 Proportion of flounder with liver exhibiting hepatocellular nuclear pleomorphism

0.8 0.7 0.6

n 0.5 0.4

Proportio 0.3 0.2 0.1 0 Alde Tyne Clyde Forth Mersey Thames Forth Con.

Southampton Site

12 Proportion of flounder with livers exhibiting categories of foci of cellular alteration (FCA) 0.1 ccFCA vFCA eFCA bFCA 0.09 0.08 0.07 0.06 0.05

Proportion 0.04 0.03 0.02 0.01 0 Alde Clyde Tyne Forth Mersey Thames Forth Con.

Southampton Site

C1617/ME3106

Site No. Fusion CLO Aneurysm Hyperplasia Monogenea Trichodina Digenea Copepoda

Alde 50 2.3 0 9.1 81.8 0 56.8 38.6 59.1

Forth Control 50 8 26 4 100 4.1 0 4 94

Southampton 45 7 4.7 11.6 4.7 4.7 4.7 9.3 0

Thames 50 0 4 4 14 2 60 2 14

Forth 45 0 4.4 6.7 73.3 0 40 4.44 66.7

Clyde 50 0 0 4 6 0 100 6 4

Tyne 50 2 6.1 8.2 57.1 6.1 83.7 2 20.4

Mersey 50 0 2 8.2 75.5 4.1 91.8 2 71.4

Table 1. Prevalence of pathologies in gill of flounder.

Sample location Number of fish sampled Mean ± SE Thames 15 0.66 ± 0.3 Belfast 7 0.43 ± 0.17 Southampton 11 0.41 ± 0.16 Alde 15 0.27 ±0.18 Mersey 9 0.17 ± 0.08

Table 2. Mean MN assay scores for flounder, by site

C1617/ME3106

Southampton Thames Clyde Mersey Tyne (Howden) Forth Alde Belfast Southampton x > 0.05 > 0.05 > 0.05 < 0.01 < 0.01 < 0.01 < 0.01 Thames x x > 0.05 > 0.05 < 0.05 < 0.05 < 0.05 < 0.05 Clyde x x x > 0.05 > 0.05 < 0.05 < 0.05 < 0.01 Mersey x x x x > 0.05 > 0.05 > 0.05 > 0.05 Tyne x x x x x > 0.05 > 0.05 > 0.05 Forth x x x x x x > 0.05> 0.05 Alde x x x x x x x > 0.05 Belfast x x x x x x x x

140.0

120.0

100.0 undamaged

8 80.0

60.0

40.0 nucleotides 20.0

0.0

e t d s DNA 10 per adducts den) Alde lfa Cly Forth w Be Thames Mersey (Ho outhampton e S n Ty Sample location

30.0

25.0

20.0

15.0 undamaged nucleotides 8 10.0

5.0

0.0 Mersey Tyne Thames Alde Belfast Adducts per 10

C1617/ME3106 22a

Real time PCR Cyp1A and Vtg Expression: Tyne v Alde

2.5

1.5

0.5

0 Tyne Male Vtg Tyne female Vtg Tyne Male Cyp1A Tyne female

Fold induction (log) rel. to Alde fish Alde rel.Fold to induction (log) -0.5 Cyp1A

C1617/ME3106

Table 3. Showing the species of parasite found in the survey of flounder caught in the Alde, Mersey, Clyde, Tyne, Forth, Southampton, Thames and Belfast estuaries Included is the site of infection, mode of lifecycle and host-specificity of the parasites recorded (specialist/generalist).

Freshwater (F)/ Specialist (SP) Allogenic (AL) / Parasite species Site of infection Marine (M)/ / Generalist Autogenic (AU) Euryhaline (E) (GE) PROTISTA Trichodina sp. Gills E AU SP DIGENEA Derogenes varicus Gut M AU GE Lecithaster gibbosus Gut M AU GE Hemiuris communis Gut M AU GE Plagioporus varius Gut M AU GE Zoogonoides viviparus Gut M AU GE Stephanostomum baccatum Muscle M AU GE Cryptocotyle lingua Muscle E AL GE Diplostomum sp. Eye E AL SP? Digeanean metacecaria Intestinal wall M? AL ? NEMATODA Dichelyne minutus Gut E AU GE Hysterothylacium aduncum Gut M AL GE Contracaecum sp. Gut M AU GE Cuccullanus heterochrous Gut M AL GE Pseudocapillaria tomentosa Gut F AU GE Nematode larval sp. Gut ? ? ? Raphidascaris acus Gut M/F AU GE Anisakis simplex Mesenteries M AL GE ACANTHOCEPHALA Pomphorhynchus laevis Gut M AU GE Gut / Acanthocephalan larvae M/F AU GE Mesenteries CESTODA Cestode sp. Gut/ viscera M AU SP Tetraphyllidea Gut M AU GE COPEPODA Acanthochondria cornuta Buccal cavity M AU SP Lernaeocera branchialis Gills M AU GE larvae. Lepeophtherius pectoralis Skin M AU GE

C1617/ME3106

Table 4. Lifecycles of parasite species recovered from flounder collected at all stations.

Parasite species First intermediate Second intermediate host Definitive host host Trichodina - - Flounder D. varicus Natica sp. Paracalanus parvus, Pseudocalanus Over 100 species of elongatus, Temora longicornis, Acartia teleosts sp., Centropages hamatus, Calanus finmarchius, Sagitta spp., coelenterates, polychaetes (Tomopteris sp.) L. gibbosus Odostomia Acartia sp. Many marine teleosts eulimoides H. communis Acartia sp. Many marine teleosts P. varius Flounder, plaice, Z. viviparus Buccinum undatum Brittle stars, polychaetes, lamellibranches, Flounder, plaice, dab, prosobranchs long rough dab S. baccatum Natica sp. Gobies Flounder C. lingua Littorina sp., Several teleost Gulls Hydrobia ventrosa Diplostomum sp. Lymnaea sp. Several teleosts Gulls Digenean molluscs Teleosts Fish eating birds, other metacecaria sp. teleosts D. minutus Unknown Unknown Pleuronectids H. aduncum Acartia bifilosa, Polychaeta, Amphipoda, Copepoda, Many teleosts Eurytemora affinis, Chaetognatha other copepoda C. heterochrous Unknown Unknown Teleosts Contracaecum sp. Crustaceans - Several teleost A. simplex Euphasia, Teleosts Cetaceans and Thysanoessa, Pinnipeds Pandulus P. tomentosa Oligochaeta - Many Freshwater Cyprinids, Percidids R. acus Oligochaeta, insect Various aquatic invertebrates Numerous sp. larvae (Diptera) piscivorous fish

Nematode larval sp. Copepod, insect Several teleosts Teleosts nymphs P. laevis Copepods, - Various teleosts ostracods, amphipods, isopods Acanthocephlan Copepods, - Various teleosts larvae ostracods, amphipods, isopods Cestode sp. Copepods, - Various teleosts amphipods, isopods Tetraphyllidea - Elasmobranchs larvae A. cornuta - - Flounder L. branchialis larvae. Flatfish - Gadoids

C1617/ME3106

Table 5. To illustrate the score for the Ecological niche (E N) to show core (C), secondary (2nd), satellite (S) and rare (R) species of parasites, prevalence (%), Abundance (A) and Range (R) of different parasite species in flounder from the sites sampled.

PARASITE Alde Mersey Clyde Tyne Forth SPECIES E N % A R E N % A R E N % A R E N % A R E N % A R Trichodina sp. C 70 - - C 100 - - C 100 - - C 100 - C 83 - - D. varicus R 0 - - S 30 0.50 0-6 R 0 - - S 17 0.50 0-1 S 23 0.47 0-5 Z. viviparus R 3 0.03 0-1 S 7 0.10 0-2 S 20 4.50 0-81 2nd 57 2.20 0-13 C 60 5.93 0-36 L. gibbosus R 3 0.03 0-1 R 0 - - R 0 - - R 0 - - R 3 0.23 0-7 P. varius S 27 0.87 0-8 S 10 0.17 0-3 S 7 0.10 0-2 R 7 0.13 0-2 S 23 0.63 0-9 H.communis R 0 - - R 3 0.03 0-1 R 0 - - R 0 - - S 27 2.80 0-27 Cryptocotyle sp. R 0 - - R 0 - - R 0 - - R 0 - - S 7 0.07 0-1 Diplostomum sp. S 30 1.77 0-15 S 13 0.17 0-2 S 33 1.07 0-16 S 7 0.37 0-10 S 20 0.63 0-13 Digenean larvae R 3 0.10 0-3 R 0 - - R 0 - - R 0 - - R 3 0.17 0-5 S. baccatum R 0 - - R 0 - - R 0 - - R 0 - - R 3 0.03 0-1 H. aduncum R 0 - - S 10 0.10 0-1 R 3 0.03 0-1 S 7 0.07 0-4 S 17 0.33 0-4 C.heterochrous S 27 1.87 0-26 C 73 18.00 0-142 R 3 0.10 0-3 2nd 47 0.87 0-4 2nd 43 1.77 0-13 D. minutus 2nd 53 3.63 0-17 R 0 - - C 90 5.43 0-21 R 3 0.03 0-3 S 13 0.20 0-2 A.simplex S 27 0.83 0-10 S 33 0.33 0-1 R 0 - - S 20 0.43 0-9 S 30 1.40 0-15 Contracaecum sp. R 0 - -- S 30 0.57 0-6 R 0 - - S 13 0.17 0-7 S 10 0.10 0-1 R. acus R 0 - - R 0 - - R 0 - - R 0 - - R 3 0.07 0-2 Nematode larvae sp unknown R 0 - - R 3 0.10 0-3 R 0 - - R 0 - - R 3 0.03 0-2 P. tomentosa R 0 - - R 0 - - R 0 - - R 0 - - R 0 - - Liver nematode (from Histology) S 7 - S 37 - - R 3 - - S 20 - - S 30 - - Acanthocephalan larvae sp. R 3 0.17 0-5 R 0 - - R 3 0.03 0-1 R 3 0.13 0-1 R 0 - P. laevis R 3 0.03 0-1 R 0 - - R 0 - - R 0 - - S 7 0.13 0-3 Tetraphyllidean larvae R 0 - - R 0 - - R 0 - - R 0 - - R 3 0.07 0-2 Cestode sp. R 3 0.03 0-1 S 10 0.20 0-3 R 0 - - R 3 0.03 0-2 S 7 0.07 0-1 L. branchialis C 100 - - C 100 - - S 13 - - C 100 - - C 97 - - A. cornuta 2nd 43 - - 2nd 50 - - R 0 - - R 3 - - S 7 - -

C1617/ME3106

Table 5 (Continued) To illustrate the score for the Ecological niche (E N) to show core (C), secondary (2nd), satellite (S) and rare (R) species of parasites, prevalence (%), Abundance (A) and Range (R) of different parasite species in flounder from the sites sampled.

PARASITE Southampton Thames SPECIES E N % A R E N % A R Trichodina sp. S 7 - - C 100 - - D. varicus S 7 0.10 0-2 R 3 0.03 0-1 Z. viviparus C 90 24.93 0-88 S 24 2.34 0-34 L. gibbosus R 0 - - R 0 - - P. varius S 7 0.17 0-3 S 7 0.52 0-14 H. communis R 0 - - R 3 0.03 0-1 Cryptocotyle sp. R 0 - - R 0 - - Diplostomum sp. S 13 0.37 0-7 C 80 13.17 0-70 Digenean larvae R 0 - - R 0 - - S. baccatum S 17 1.23 0-17 R 0 - - H. aduncum R 0 - - S 10 0.10 0-1 C. heterochrous S 7 0.07 0-1 2nd 48 1.90 0-23 D. minutus C 80 9.47 0-72 S 17 0.34 0-5 A .simplex R 0 - - S 3 0.14 0-4 Contracaecum sp. R 0 - - S 10 0.10 0-1 R. acus R 0 - - R 0 - - Nematode larvae sp unknown R 0 - - R 0 - - P. tomentosa R 0 - - R 3 0.03 0-1 histo Liver nematode R 0 - - R 3.3 - - Acanthocephalan larvae sp. S 10 0.23 0-5 2nd 55 7.24 0-60 P. laevis R 0 - - S 21 0.86 0-12 Tetraphyllidean larvae R 0 - - R 0 - - Cestode sp. R 3 0.03 0-1 S 7 0.34 -0-9 L. branchialis R 3 - - C 100 - - A. cornuta C 60 - - S 23 - -

C1617/ME3106

23 80 70 60 50 Alde 40 % Mersey 30 Clyde 20 Tyne 10 Forth 0 S'oton Thames R

S S'oton Forth Thames

2 Tyne Ecological

C Clyde

niche status Alde

Mersey site

24 45

prevalence % prevalence Alde 25 Mersey 20 Clyde Tyne 15 Forth Soton 10 Thames 0

2 5 4

6 0

number of parasite species / 8 fish 10 Soton Forth Tyne Thames 12 Clyde Alde Mersey sites

C1617/ME3106

Table 6. One-Way ANOVA for each parasite species at sites (original parasite count data used from Metazoan species). *** p<0.001, ** p<0.01, *p<0.05 and NS, Not significant

Parasite species Statistical significance H .aduncum * C. heterochrous *** D.minutus *** A.simplex ** Contracaecum sp. *** Acanthocephalan larvae sp. *** P. laevis ** R. acus NS Nematode larvae sp unknown NS P. tomentosa NS D. varicus * Z. viviparus *** L. gibbosus NS P. varius NS H. communis *** Cryptocotyle sp. NS Diplostomum sp. *** Digenean larvae NS S. baccatum *** Tetraphyllidean larvae NS Cestode sp. NS

C1617/ME3106 Table 7. Parasite component communities showing the number of parasite species found (S), total number of individual species counted (N) plus mean in parenthesis, species richness (d), Pieloiu's evenness values (J’) and Shannon diversity index (H').

Number of Site S N (rank) d (rank) J' (rank) H' (rank) fish Alde 30 11 (3) 281 (9.4) 1.77 (4) 0.68 (2) 1.64 (3) Mersey 30 11 (3) 608 (20.3) 1.56 (5) 0.24 (7) 0.58 (7) Clyde 30 7 (7) 338 (11.3) 1.03 (7) 0.54 (5) 1.06 (5) Tyne 30 11 (3) 148 (4.9) 2.00 (2) 0.73 (1) 1.74 (2) Forth 30 19 (1) 454 (15.1) 2.94 (1) 0.67 (3) 1.98 (1) Southampton 30 9 (6) 1098 (36.6) 1.14 (6) 0.39 (6) 0.86 (6) Thames 29 14 (2) 801 (27.6) 1.94 (3) 0.56 (4) 1.48 (4)

Where: d= Species richness (Margalef) (S-1)/logeN J' = Pieloiu's evenness H'/ logeS H'(loge) = Shannon and Wiener H' = - Σ PI log (PI), where the logs are to the base e

Table 8. Results of tests of statistical significance for Shannon indices identified in Table 7 for pairs of locations compared to the Alde.

SITE Alde Mersey Clyde Tyne Forth Southampton Thames Alde NS *** *** NS *** *** **

Where *** p<0.001, ** p<0.01, *p<0.01 and NS Not significant.

25 26

27 28

Gill Gill Liver Kidney Muscle Muscle CLO Digenea Digenea Myxozoa Digenea Myxozoa

Alde 0 0 11.1 0 90 0

Thames 33.3 0 0 0 10 0

Southampton 50 0 11.1 80 70 20 Water

Clyde 0 16.7 0 0 0 66.7

Mersey 0 0 0 0 0 0

Table 9. Prevalence of pathogens recorded in sand gobies.

29 30

31 32

C1617/ME3106

C1617/ME3106 Section 3: Crustaceans

Invertebrates show considerable potential as sentinel organisms for the monitoring of the health status of aquatic systems. They are generally small, abundant, and relatively sessile and may readily bioaccumulate toxins. Cascade-like stress responses can occur following acute or chronic exposures to contaminated environments and as such, the overall health status of individuals within those environments, both in terms of histopathological lesions and the presence of infecting organisms, may ultimately reflect the general health status of these sites. The current study provides baseline multi-organ histopathological data for two common crustacean species, the shore crab (Carcinus maenas) and the brown shrimp (Crangon crangon) collected from six UK estuarine sites. Changes in the metabolic condition of crustaceans from these sites (measured in terms of connective tissue storage cell status) were interpreted in relation to other health measures (including parasite load and the presence of microbial pathogens). The relative ease at which a holistic assessment of health can be made using histopathology and the suitability of these species as environmental sentinels provide support for the inclusion of crustaceans as indicators of aquatic environmental health. Studies linking disease status to burdens of industrial contamination in these environments are now required.

The results of these studies on crustacean pathology are published in:

Stentiford, G.D., Bateman, K., Feist, S.W. (2004). Pathology and ultrastructure of an intranuclear bacilliform virus (IBV) infecting brown shrimp Crangon crangon (Decapoda: Crangonidae). Diseases of Aquatic Organisms 58, 89-97.

Stentiford, G.D., Feist, S.W. (2005). A histopathological survey of shore crab (Carcinus maenas) and brown shrimp (Crangon crangon) from six estuaries in the United Kingdom. Journal of Invertebrate Pathology 88, 136-146.

C1617/ME3106 3.1 Introduction

Stress has been implicated in the development of disease in aquatic animals, both in aquaculture (Houghton & Matthews 1986) and in natural conditions (Ross et al., 1996). In such cases, induction of the ‘stress cascade’ can allow latent infections to manifest as disease via generalised host immunosuppression induced by the prevailing environmental conditions (Johnson, 1983). Under adverse conditions, the relatively primitive invertebrate immune system (of phagocytosis, encapsulation and initiation of the prophenoloxidase system) may be compromised so that opportunistic pathogens, both from the ambient environment and those living on or within the host are allowed to proliferate, with pathological consequences. When baseline information is gathered on the prevalence of these diseases under a range of scenarios, the data may then be used as a marker to discriminate stressful aquatic environments.

Fish diseases and pathologies, with a broad range of aetiologies, are increasingly being used as indicators of environmental stress since they provide a definite biological end-point of historical exposure (Matthiessen et al., 1993). Benthic invertebrates too show considerable potential as sentinel marker species in ecosystem health monitoring programs since they are small, common, relatively sessile and tend to bioaccumulate toxicants present in their environment. In addition, the biochemical, physiological and histological characteristics of several common species are sufficiently well known to discriminate exposed from non-exposed individuals (Viarengo, 1993). However, whilst studies on finfish have shown that histopathology is a sensitive indicator of individual and population health status and results from numerous controlled laboratory exposures of shellfish (crustaceans and molluscs) to toxicants have shown that histopathological changes also occur in the organ and tissue systems of these animals, relatively few field studies have included shellfish histopathology in the suite of monitoring tools employed. Most studies of this type have centred on the use of the common mussel (Mytilus edulis) (see Lowe et al., 1981, Moore et al., 1987, Wedderburn et al., 2000). Studies on crustacean health status have focussed on the response of individual organ systems to laboratory exposure to a range of contaminants (Couch 1977, 1984, Sarojini et al., 1993, Victor, 1993, 1994, Soegianto et al., 1999a, 1999b). However, whilst studies of this type undoubtedly provide an invaluable insight into the cellular response to pollutants, with an increasing emphasis on the effect of stress at the population-level, more holistic approaches are likely to be required to aggregate the combined effect on multiple organ systems. Relatively few studies have applied such an approach to wild crustacean populations, despite their pivotal role within food chains (Couch, 1978, Sindermann, 1979, Bang, 1980, Overstreet, 1988).

The current study reports baseline prevalence data for a range of parasites and pathologies present in two common crustacean species (Carcinus maenas and Crangon crangon) found in UK estuaries. To the best of our knowledge, this is the first attempt to collect multiple organ histopathology data for these species from indigenous field sites. It has demonstrated the

C1617/ME3106 relative ease at which disease status can be assessed in these species and provides insights into how invertebrate disease may be used as a high-level indicator of ecosystem health.

3.2 Materials and methods

Shore crabs (Carcinus maenas) were captured during the autumn of 2002 by standard beam trawling from the Alde, Mersey, Tyne, Forth and Clyde estuaries and from Southampton Water (Table 1). Approximately 30 crabs from each site were euthanised and fixed by direct injection with Davidson’s seawater fixative (Hopwood, 1996). The carapace of euthanised crabs was carefully removed before placing the remaining carcass into the same solution for 24 h before transfer to 70 % industrial methylated spirit for transport and storage. Following fixation, the hepatopancreas, heart, gonad, gill, midgut and muscle were removed and processed for histological examination using standard protocols. Thin sections (3-5 μm) were obtained using a rotary microtome and were stained with haematoxylin and eosin (HE). Stained sections were analysed by standard light microscopy and digital images were captured using the LuciaG™ Screen Measurement System (Nikon, UK).

An index on the relative abundance of glycogen-containing reserve inclusion (RI) cells in histological sections of the connective tissues of crabs was devised based upon their presence in the connective tissues of the hepatopancreas. The index ranged from Stage 0 (RI cells absent) through Stage 1 (RI cells present but scarce), Stage 2 (RI cells scattered), Stage 3 (RI cells frequent) to Stage 4 (RI cells abundant and constituting the majority of connective tissue volume).

Brown shrimp (Crangon crangon) were captured using the same standard beam trawling methodology as mentioned above. Up to 35 brown shrimp (Crangon crangon) from the Alde, Thames, Mersey and Clyde estuaries were euthanised and fixed in the same way as for C. maenas. Euthanised shrimp were immediately placed into the same solution for 24 h before transfer to 70 % industrial methylated spirit for transport and storage. For processing, whole shrimp were sectioned longitudinally and the cut surface was embedded outermost in the block for ease of sectioning. Whole shrimp sections were stained with HE and selectively re-sectioned for staining with the Farley- Feulgen (FF) and Periodic acid-Schiff (PAS) stains. Stained sections were analysed by standard light microscopy and digital images were captured as above.

An index describing the pathological manifestation of a previously described virus infection, Crangon crangon bacilliform virus (CcBV) (Stentiford et al. 2004) was also constructed and applied to all shrimps sampled during the current study. The details of the index are given in Table 10. For ultrastructural confirmation of CcBV, the hepatopancreas was removed from 20 shrimp captured from the Clyde estuary and prepared for electron microscopy using the method previously described by Stentiford et al. (2004). Thick sections were stained with Toluidine Blue for viewing with a light microscope to identify suitable target areas. Ultrathin sections (70-90 nm) of

C1617/ME3106 these areas, mounted on uncoated copper grids and stained with uranyl acetate and Reynolds’ lead citrate were examined using a JEOL 1210 transmission electron microscope.

Data pertaining to the index of the relative abundance of RI cells in C. maenas and the pathological manifestation of CcBV infection in C. crangon was tested for normality. Normally distributed data was analysed using one way analysis of variance (ANOVA) while the Mann-Whitney test was applied to non- normally distributed data. Significance differences between sites was considered to be at p< 0.05.

3.3 Results

3.3.1 Shore crab (Carcinus maenas)

Sampling for shore crab (Carcinus maenas) occurred during two periods. The first, a preliminary screen of crabs from the Clyde estuary took place in Autumn 2002. The pre-screen was used to identify the range of pathogens and pathologies that should be included in a more comprehensive screen of multiple sites in Autumn 2003.

Preliminary sampling. General pathologies recorded in crabs from the Clyde included the presence of haemocytic aggregates and melanised nodules in the gills, heart and hepatopancreas. Connective tissues (particularly between the tubules of the hepatopancreas) contained reserve inclusion (RI) cells in all crabs, suggesting that metabolic storage products were being produced and stored. Secretory packets were seen within the vacuoles of the Blassenzellen (B)-cells of the hepatopancreas and in a number of cases, vacuoles, or even whole cells, were apparently ejected into the lumen of the hepatopancreatic tubules. Other observations of note in crabs captured from the Clyde included parasitic infection by the digenean trematode Microphallus primas in the hepatopancreas and the fouling of the gills by filamentous bacteria and stalked ciliates (Vorticella spp.). Atresia (degeneration) of the vitellogenic oocytes was also observed. Examples of these pathologies are described below.

Main study. Thirty crabs were sampled from all sites apart from the Mersey (n = 12). Recently moulted crabs and those approaching imminent moult were not sampled. The sex ratio of crabs (determined by histology) differed between sites, with the Mersey (1.2:1), Alde (1.3:1) and Clyde (1.33:1) showing male bias and the Forth (0.36:1) and Southampton water (0.44:1) showing a strong female bias. A number of pathologies were observed in crabs from various estuarine sites. The prevalence of these pathologies in each organ sampled is given in Table 11.

The RI cell index for crabs captured from the various sampling sites is given in Table 1. A representative example of connective tissue RI cells is given in Fig. 33. The mean RI cell score was highest in crabs from the Alde estuary (1.7) and Southampton Water (1.63) and was lowest in crabs from the Mersey (0.75) and Tyne (0.97) estuaries. The mean RI cell score in crabs from the

C1617/ME3106 Mersey, Tyne and Clyde estuaries were significantly lower than in crabs from the Alde estuary (all p < 0.05).

Secretory packets were seen within the vacuoles of the Blassenzellen (B)- cells of the hepatopancreas of crabs from the Tyne, Forth and Clyde estuaries (Table 11). In a number of cases, it appeared that these vacuoles, or even whole cells, were being ejected into the lumen of the hepatopancreatic tubules (Fig. 34). No secretory packets were observed in crabs captured from the Mersey or Alde estuaries or from Southampton Water. General pathologies recorded in crabs from all sites included the presence of haemocytic aggregates and melanised nodules in the gills, heart, hepatopancreas and gonad (Table 11, Fig. 35). The presence of these pathologies showed no clear site bias.

A number of parasites were observed infecting crabs from several sites. Of particular note was the high prevalence of Sacculina carcini infecting crabs from Southampton Water and the Forth estuary. No crabs from the Mersey, Alde and Tyne estuaries were infected, while only a very low prevalence was observed in crabs from the Clyde estuary (Table 11). Rootlets of the parasite could be seen forming a systemic network in infected crabs, often displacing the tubules of the hepatopancreas (Fig. 36). In infected female crabs, invasion of the gonad led to atresia of the oocytes and to an apparent arrest of their development in the pre-vitellogenic phase (Fig. 37). Several crabs appeared to mount an immune response to the invasive rootlets, manifested as haemocytic encapsulation and in some cases, melanisation (Fig. 38). It is noteworthy that the sites with the highest S. carcini prevalence (Southampton Water, Forth estuary) also showed a sex ratio with a strong female bias. The mean RI cell index was also lowest at sites with the highest S. carcini infection prevalence (see Table 11).

Another parasite of note in crabs captured from the Alde, Forth and Clyde estuaries and from Southampton water was the digenean trematode Microphallus primas. Infection prevalence was highest in crabs from Southampton Water and from the Alde estuary, with lower prevalence in the Forth and Clyde estuaries (Table 11). The most common sites of infection were the hepatopancreas (Fig. 39) and the gill (Fig. 40). Infection did not appear to elicit an acute host immune reaction, though apparent ‘pearl’ formation did occur around metacercaria, outside of which, whorls of haemocytes were occasionally observed. Within the gill, encysted metacercaria appeared to form blockage of the secondary gill filaments and in some cases led to haemocyte stasis within the lamellae.

Fouling of the gills by filamentous bacteria (Fig. 41) and by stalked ciliates (Fig. 42) was observed in crabs captured from all sites. Bacterial fouling was most prevalent in crabs from Southampton Water and the Tyne estuary, while the prevalence of ciliates was highest in crabs from Southampton Water and the Tyne and Clyde estuaries. However, since detailed information on moult status (i.e. the particular stage of the intermoult) was not collected, care should be exercised when interpreting external fouling data of this kind.

C1617/ME3106 A pathology similar to that previously associated with infection by the bacilliform viruses was observed in the hepatopancreatic epithelium of crabs from the Clyde (10.7 %) and Tyne (3.3 %) estuaries. The infection did not appear pathogenic in those crabs observed and since material was not collected for electron microscopy, no inferences can be made on the exact aetiology of this condition or of any similarity to the C. crangon virus recorded in shrimp from the Clyde, Mersey, Alde and Thames estuaries (see below) (Fig. 43). Finally, one of the crabs collected from the Southampton Water site harboured an infection putatively identified as Hematodinium perezi. Uninucleate parasite cells with distinct nuclei formed small aggregates in the sinuses of the myocardium (Fig. 44).

3.3.2 Brown shrimp (Crangon crangon)

The most significant pathology noted in C. crangon was that caused by infection by a previously undescribed non-occluded bacilliform virus (hereby named Crangon crangon Bacilliform Virus - CcBV). While there were no obvious signs of external disease among the shrimp collected from the Alde, Thames, Mersey and Clyde estuaries, animals from these sites exhibited high prevalences of infection (almost 100 % in the Clyde and Mersey estuaries).

Histological analysis of whole shrimp sections revealed a characteristic pathology: at low magnification, pathology consisted of degeneration of the hepatopancreatic tubules, causing a loss of histological structure in the organ while at higher magnification, significant changes were seen in the differentiated epithelial cell types lining the hepatopancreatic tubules and of the epithelial cells lining the midgut. Hypertrophied and other aberrant nuclei either contained enlarged nucleoli (Fig. 45) or an amorphous eosinophilic matrix. Nuclei which contained this matrix often had a disrupted chromatin profile with condensation of nuclear material at the inner periphery of the nuclear membrane. Cells containing aberrant nuclei appeared singly or in clusters and the cytoplasm of intact affected cells appeared dense and basophilic, with an apparent loss of lipid storage inclusions. Aberrant nuclei were confined to the storage (R) and possibly fibrillar (F) cells of the hepatopancreatic tubules. In contrast, they were not observed in the regenerating epithelial cells (e-cells) of the distal hepatopancreatic tubules, while mitotic figures were commonly seen within these cells. The percentage of hepatopancreatic epithelial cells containing aberrant nuclei in shrimps displaying this condition ranged from approximately 5 % to 40 %.

The basement membrane of affected epithelial cells in the hepatopancreatic tubules and the midgut was often separated from that of its neighbouring cells. These cells appeared to contain nuclei that were apoptotic and in several cases, appeared to be in the process of expulsion into the lumen of the tubule or the midgut (Fig. 46). Sloughed-off epithelial cells containing aberrant nuclei could be seen within the lumen of degenerated tubules and in the midgut. In severe infections, tubules appeared necrotic, with degeneration of epithelial cells and only remnants of the basement membrane and myoepithelial lining remaining.

C1617/ME3106 Transmission electron microscopy of these aberrant hepatopancreatic epithelial cells revealed the presence of rod-shaped and cylindrical, envelope- bound particles accumulated within aberrant nuclei (Fig. 47). These virions did not form arrays and some appeared to be partially occluded in an amorphous matrix which corresponded to a granular viroplasm. The tri-laminar envelope that surrounded the nucleocapsid of the virion was expanded at one end and appeared to accommodate a tail-like structure (Fig. 48). The morphology and size of the virions and the lack of any apparent occlusion body associated them most closely with the non-occluded bacilliform viruses.

In addition to the CcBV infection in shrimp, 26.7 % of shrimp captured from the Clyde estuary were co-infected by a previously undescribed yeast-like organism. Yeast cells could be observed free in the haemosinuses of infected shrimp and occasionally within the haemocytes and associated with encapsulation responses (Fig. 49, 50). The haemopoeitic centres were disrupted by yeast cells (Fig. 51), while muscle tissue appeared to be undergoing proteolysis due to yeast infection (Fig. 52). The yeast infection was not detected in shrimp from any of the other estuarine sites and was not detected in C. maenas captured from the Clyde estuary during the same sampling trip.

3.4 Discussion

This study has provided baseline data on a range of pathologies and pathogens present in two species of wild crustacean (Carcinus maenas and Crangon crangon) from several estuarine sites in the United Kingdom. The relative ease at which disease markers in these species can be recorded using histopathology is encouraging and may be further applied to dedicated studies of disease in relation to prevailing environmental conditions at such sites (e.g. the presence of contaminants).

3.4.1 Carcinus maenas

A number of pathologies were recorded in the organs and tissues of crabs from all sites. Interestingly, the sex ratio differed considerably between sites, with the Mersey, Alde, and Clyde showing a male bias, the Tyne showing a relatively even ratio and the Forth estuary and Southampton Water showing a strong female bias. In coincidence with this data was the high prevalence of the parasitic Rhizocephalan barnacle Sacculina carcini in crabs from the Forth and Southampton sites (with almost complete absence of infection in crabs from the other sites sampled). Infection prevalence was highest (over 90 %) at the Southampton site, this considerably higher than that reported in previous studies at other sites (Werner, 2001). It has been suggested that infected hosts are more lethargic and are less easily caught by passive fishing methods such as traps, thereby underestimating true prevalence in the field (Crothers, 1968, Werner, 2001). In the current study, active capture by trawling may have provided a more realistic estimate of S. carcini prevalence at the sites where it is endemic.

C1617/ME3106 S. carcini can modify a range of physiological, biochemical, endocrinological and behavioural traits in its host (Werner, 2001). Significantly, the secondary sexual characteristics of male crabs may also be altered (Høeg, 1995) by destruction of the androgenic gland by the parasite (Veillet & Graf, 1958). The manifestation of this pathology leads to feminisation of males, externally visible by a widening of the abdominal pouch and retarded development of other male-oriented features (such as claw size) (Høeg, 1995). Retardation of testicular growth and hermaphrodite sexual gland generation, with ovarian tissue appearing in the testis of previously infected hosts, has also been reported (Reinhard, 1956). It is not inconceivable then that the female biased sex ratios observed at the Southampton and Forth sites may, at least in part, be due to current and previous infection by this parasite. We can make no inference as to the cause of the male biased sex ratios at the Mersey, Alde and Clyde sites though from the current study, it is recommended that any attempts to link external sex ratio or morphometric data to prevailing environmental conditions (e.g. contamination) should be carried out in conjunction with histological assessment of S. carcini prevalence in the population.

In addition to possible effects on the sex ratio of crabs captured from the various sites, S. carcini infection also appeared to alter the metabolic status of hosts in the form of reduced reserve inclusion (RI) cell abundance. The function of RI cells are likely associated with the synthesis and storage of haemocyanin and other products such as glycogen; these reserves being utilized during stressful periods such as during moulting, disease or hibernation and during normal reproduction (Johnson, 1980). Historical studies have stated that RI cells are common in well fed C. maenas while absent in starving crabs (Cuénot, 1893). In light of this information, during the current study, we made an attempt to grade the RI cells status in crabs captured from the different estuarine sites. Interestingly, the RI cell score for the Alde estuary was significantly higher than that seen in crabs from the Mersey, Tyne and Clyde estuaries. Significant differences between Southampton water and the Mersey, Tyne and Clyde estuaries appeared to be precluded by the high prevalence of S. carcini at the Southampton site (which caused a reduction in connective tissue structure and an apparent loss of RI cells). The RI cell score may provide a useful tool for grading overall condition in C. maenas populations in estuaries. However, it should be noted that accurate moult staging and additional histopathological data (e.g. of parasitic infection prevalence) is required in order to provide confidence for interpretation of data of this type.

Previous studies on rhizocephalan infections of crabs have suggested that the internal rootlets of the parasite somehow avoid attack by the host immune system, possibly by generation of a glycoprotein-rich cuticle that prevents recognition of the parasite as non-self (Walker, 2001). However, in the current study, we report that in several cases, S. carcini rootlets were vulnerable to the host immune response and were often encapsulated with aggregates of flattened haemocytes and occasionally, within melanised nodules that corresponded to granuloma (Sparks, 1980). Whether this response is

C1617/ME3106 sufficient to regress the infection and to starve the externa of the parasite is not known.

Another parasite, Microphallus primas, was also found in high prevalence in populations of C. maenas captured from Southampton Water and the Alde estuary. A lower prevalence of infection was also observed in crabs from the Forth and Clyde sites. Adult M. primas are found as parasites of the digestive tract of several species of marine birds (Dawes, 1968, James et al., 1976) while the metacercarial cysts of the parasite infect C. maenas. The remaining larval stages occur within molluscan hosts such as Hydrobia ulvae (Saville & Irwin, 1991). In this study, metacercarial stages of the parasite were most commonly found encysted within the haemal sinuses of the hepatopancreas and gill, though occasionally, cysts were also detected within the muscle and the heart. Upon dissection, heavily infected crabs could be recognised by the presence of small melanised foci throughout the hepatopancreas. At microscopic level, these foci consisted of metacercarial stages of the parasite within an eosinophilic capsule, presumably of host origin (Martorelli & Schuldt, 1990). Infection by Microphallus spp. in other crab species has been reported to lead to structural damage to the hepatopancreas, necrosis of tubules and an alteration in the concentration of physiological storage products (Robaldo et al., 1999). In the current study, necrosis of tubules was not observed, though displacement of tubules was seen to occur in heavy infections. Apart from generation of the capsule, the encysted M. primas did not appear to elicit an acute haemocytic encapsulation response in the host, though on occasion, encysted parasites were surrounded by a thin layer of flattened haemocytes.

It is of note that no M. primas-infected crabs were detected at either the Mersey or Tyne sites and that infection prevalence was low at both the Forth and Clyde sites. Similar patterns of reductions in digenean infection prevalence have also been recorded in fish exposed to industrial pollution (Burn, 1980). In relation to such findings, MacKenzie et al. (1995) suggested that unequivocally linking environmental stressors such as pollution with parasite prevalence in fish is highly complex and thus problematic without consideration of various biotic and abiotic factors and the capability for migration of hosts. However, it is noteworthy that in the current study, the use of relatively sessile benthic invertebrates circumvents at least the issue of host migration. As such, the study of parasite communities as a means of assessing the impact of industrial contamination of aquatic ecosystems may be particularly relevant when applied to invertebrate hosts.

Fouling organisms (filamentous bacteria, ciliates) were detected on the secondary gill lamellae of C. maenas captured at all sites. The prevalence of ciliates was highest in crabs from the Tyne and Clyde estuaries and from Southampton Water, while filamentous bacterial fouling was most common in crabs from the Tyne estuary and from Southampton water. The occurrence of microbial epibionts on marine crustaceans has been documented in numerous studies (for review see Carman & Dobbs, 1997. In fish, it has been suggested that increased burdens of gill ciliates may be due to immunosuppression of the host, coupled with changes in the structure of the gill caused by pollutants (Khan, 1990; Yeomans et al. 1997). While the

C1617/ME3106 ecological significance of similar infestations in crustaceans is not well understood, there is evidence that epibiont load is correlated to intermoult duration, with longer intermoult periods or even terminal moult status expected to accumulate the greatest loads (Carman & Dobbs, 1997). As such, it is likely that epibiont loading such as that described in this study may be used to indicate not only prevailing microbiotic conditions within the aquatic environment but also the existence of normal moulting cycles within the crabs inhabiting these environments.

A low prevalence and intensity of a putative viral infection of the hepatopancreatic epithelial cells was recorded in C. maenas captured from the Tyne and Clyde estuaries. The histopathological manifestation was similar to that described for the intranuclear bacilliform virus described in Crangon crangon from the Clyde estuary in a previous study (Stentiford et al., 2004) and in the Mersey, Alde and Thames estuaries in the current study. Finally, one crab captured from Southampton Water was found to harbour an infection by the dinoflagellate parasite Hematodinium perezi. This is the first report of this parasite in shore crabs from European waters since the original description by Chatton and Poisson (1931). It’s rediscovery from close to the type location heralds’ future opportunity for a full description and genetic sequence analysis of H. perezi in comparison to Hematodinium sp. isolates described from other crustacean hosts.

3.4.2 Crangon crangon

The ubiquity and relative abundance of the brown shrimp (Crangon crangon) affords it considerable potential for use as an environmental sentinel. The most significant pathologies noted in C. crangon during the current study was the intranuclear bacilliform virus (IBV) previously described infecting the hepatopancreatic epithelial cells (Stentiford et al., 2004). In the current study, the infected host range was extended from the Clyde to the Thames, Mersey and Alde estuaries. The lesions observed in the epithelium of the hepatopancreas and midgut of C. crangon were identical to those previously described by Stentiford et al. (2004) and are typical of those caused by the bacilliform viruses infecting other crustacean hosts. The pathological manifestation of CcBV was marked by eventual degeneration of the storage (R) epithelial cells of the hepatopancreatic tubules and of the midgut. In severe cases, large areas of the hepatopancreas appeared necrotic, with loss of tubular structure and organ integrity.

Although in several cases, the hepatopancreas of CcBV-infected shrimp appeared to be in a state of severe degeneration, no haemocytic encapsulation response (as noted by Johnson (1984) for baculovirus infections) was recorded. In addition to the possible effect of apoptosis in prevention of inflammatory reactions in virus-infected shrimp (Stentiford et al., 2004), the lack of host immune response to this pathology may also suggest that the immune system of shrimps captured from the various estuarine sites may be compromised. Previous studies on the defence capability of C. crangon have shown that exposure to contaminated harbour dredge spoils led to a reduced total haemocyte count and blood cell phenoloxidase activity

C1617/ME3106 (Smith et al., 1995). In such a way, it is conceivable that the difference in infection prevalence and infection intensity (derived from the severity index) may be caused at least partly by differences in environmental stressors present between the sites. Such interpretation is reinforced by previous studies that have indicated that environmental stressors can enhance the prevalence and severity of baculovirus infections. Crowding or exposure to sublethal amounts of PCBs increases the prevalence of Baculovirus penaeii in penaeid shrimps (Couch, 1974, Couch & Courtney, 1977), while poor environmental conditions, shell disease, general bacterial infections and carapace fouling have been associated with an increased prevalence of Monodon baculovirus (MBV) in Penaeus monodon (Lightner & Redman, 1981). That stress has been found to exacerbate viral prevalence and disease in marine invertebrates is further reinforced by studies of vertebrate populations where it has been stated that ‘…the most important factor in transforming an infection into a disease is stress.’ (Overstreet, 1978). In such a way, latent infections may transform to patent disease (Sindermann, 1979) and even cause epizootics (Lightner & Redman, 1981). Johnson (1984) even states that studies on the effects of stress on viral disease in marine invertebrates will prove invaluable to studies of the effects of pollutants or other man-made stressors on natural populations and in aquaculture. The CcBV model in Crangon crangon may provide such an example.

A number of studies have reported the occurrence of secondary bacterial infections in invertebrates, particularly for virus-infected crustaceans under aquaculture conditions. In such cases, secondary infections are generally attributed to immunosuppression induced by the primary pathogen (Johnson, 1983). In the current study, a co-infection by a yeast-like organism was discovered in virus-infected C. crangon from the Clyde estuary. It is of note that while haemocytic encapsulation responses were not observed against virus-infected cells (see above), large encapsulation responses, were associated with the yeast-like cells, some of which were also observed within the cytoplasm of free haemocytes. Granuloma-like lesions, similar to those described by Sparks (1980) were also present. Numerous studies have shown that the prophenoloxidase system in crustaceans can be initiated by microbial cell wall components, such as β 1,3-glucans in fungi and lipopolysaccharides and peptidoglycans in Gram-negative and Gram-positive bacteria, respectively (Thörnqvist & Söderhäll, 1997). It appears likely then that the immune system of virally infected C. crangon may still be able to respond to secondary pathogens, such as the yeast-like organisms, that express suitable stimulatory molecules in their cell walls. However, the mere presence of yeast like cells in the haemolymph of shrimp may also suggest that the immune response is not sufficient to contain the proliferating yeast cells. The ubiquity of C. crangon in estuarine and coastal locations and the potential for epizootic viral and yeast infections in this species, may provide an ideal disease model for future studies on the effect of stress and pollution on the prevalence, pathogenesis and severity of disease.

C1617/ME3106

Stage 0 • Hepatopancreatocytes and midgut epithelial cells normal.

No sign of aberrant nuclei or epithelial sloughing into lumen. Absent •

Stage 1 • Few aberrant nuclei (eosinophilic, enlarged, peripheral chromatin). • Most hepatopancreatic tubules not affected. Scattered • No epithelial sloughing into lumen.

Stage 2 • Frequent aberrant nuclei present in numerous hepatopancreatic tubules. • Some separation of infected cells from their neighbours. Frequent • Epithelial sloughing is infrequent

• Majority of hepatopancreatic tubules contain numerous aberrant nuclei. Stage 3 • Separation of large numbers of infected cells from their neighbour cells.

• Some epithelial sloughing of infected cells into lumen. Abundant • Some tubules appear degenerate.

• Majority to all hepatopancreatic tubules contain cells with numerous aberrant nuclei Stage 4 • Separation and apparent apoptosis of large numbers of infected cells.

• Large numbers of epithelial cells are sloughed into lumen. Severe • Tubules appear degenerate, often involving epithelial cells of the midgut

Table 10. Staging of Crangon crangon Bacilliform Virus (CcBV) infection severity.

C1617/ME3106

A: Mean RI Secretory Epithelial Sacculina Microphallus Microsporidian* Hepatopancreas cell score packets virus carcini primas Mersey 0.32 0 0 0 0 0 Southampton 1.63 0 0 93.3 63.3 20 Alde 1.7 0 0 0 63.3 26.7 Tyne 0.97 13.3 3.3 0 0 0 Forth 1.23 10 0 73.3 20 0 Clyde 0.92 7.1 10.7 3.6 17.9 3.6

Haemocyte Melanised Stalked Other fouling Microphallus Sacculina B: Gill aggregates nodules ciliates organisms** primas carcini Mersey 41.7 8.3 25 25 0 0 Southampton 73.3 10 60 93.3 66.7 3.3 Alde 73.3 0 46.7 40 60 0 Tyne 26.7 16.7 76.7 66.7 0 0 Forth 36.7 13.3 16.7 3.3 3.3 3.3 Clyde 36.7 6.7 73.3 6.7 0 0

Haemocyte Melanised Myocardial Sacculina Microphallus Hematodinium C: Heart aggregates nodules necrosis carcini primas perezi*** Mersey 45.5 0 9.1 0 0 0 Southampton 25.9 7.4 0 3.7 0 3.7 Alde 46.4 7.1 0 0 0 0 Tyne 56.7 10 3.3 0 0 0 Forth 28 16 0 0 0 0 Clyde 58.3 12.5 0 0 0 0

Mature Mature Oocyte Haemocyte Melanised Sacculina D: Gonad oocytes (F) spermatozoa (M) atresia (F) aggregates nodules carcini Mersey 80 66.7 0 0 0 0 Southampton 33.3 100 44.4 46.1 15.4 61.5 Alde 100 84.6 10 0 0 0 Tyne 92.9 84.6 35.7 0 0 0 Forth 78.6 80 28.6 5 0 42.1 Clyde 77.8 91.6 33.3 0 0 0

Table 11. Health index parameters recorded from organs of Carcinus maenas. All numbers refer to percentage prevalence of parameter in particular organ of crabs from a given site. Key: *Unidentified microsporidian infection of hepatopancreatic epithelial cells. **Mixed population of fouling epibionts, dominated by filamentous bacteria. ***Putative Hematodinium perezi diagnosis based on original description of parasite in Carcinus maenas (Chatton, 1931). (F) In female crabs, (M) in male crabs.

Fig 33-38. (33) Hepatopancreas of Carcinus maenas. Hepatopancreatic tubules (H) are separated by connective tissues rich in reserve inclusion (RI) cells (asterisk). H&E, Bar = 250μm. (34) Hepatopancreas of Carcinus maenas. Tubule epithelium containing blister-like cells (Tep) and ‘secretory packets’ (asterisk). Tubule lumen (Tlu) (arrow). H&E Bar = 100 μm. (35) Heart of Carcinus maenas. Haemocytic aggregate (short arrow) and melanising granuloma- like lesion (long arrow) in the myocardium (My). H&E Bar = 50 μm. (36) Hepatopancreas of Carcinus maenas. Tubules containing normal epithelial cells (Tep) and lumens (Tlu) surrounded by parasite rootlets of Sacculina carcini (arrow). H&E Bar = 100 μm. (37) Ovary of Carcinus maenas. Pre-vitellogenic oocytes (short arrows) interspersed with rootlets of Sacculina carcini (long arrows). (38) Hepatopancreas of Carcinus maenas. Remnant Sacculina carcini infection showing massive infiltration by host haemocytes and apparent granuloma-like formations (arrows). H&E Bar = 250 μm

33 34

H H * * * Tep * Tlu

35 36

Tep My

Tlu

37 38 Tep

Tep

Fig. 39-44. (39) Hepatopancreas of Carcinus maenas. Metacercaria of Microphallus primas (arrows) displace normal hepatopancreatic tubules. An unidentified microsporidian parasite is also shown forming a large xenoma-like cyst (asterisk). H&E Bar = 100 μm. (40) Gill of Carcinus maenas. Metacercarial cyst of Microphallus primas (arrow) lodged within the secondary lamellae. H&E Bar = 100μm. (41) Gill of Carcinus maenas. Fouling of the secondary lamellae (2o) by filamentous bacteria and other unidentified microbial epibionts (arrow). H&E Bar = 25 μm (42) Gill of Carcinus maenas. Fouling of secondary lamellae by stalked ciliates (short arrows). Ciliates were often present with filamentous bacteria and other fouling microbial epibionts (see Fig. 9). A metacercarial cyst of Microphallus primas is seen in the primary lamellae (long arrow). H&E Bar = 100μm (43) Hepatopancreas of Carcinus maenas. Epithelial cells of a hepatopancreatic tubule containing enlarged nuclei with peripheral chromatin (arrows). H&E Bar = 25μm (44) Hepatopancreas of Carcinus maenas. Putative Hematodinium perezi infection of the haemal sinus. Uninucleate parasites containing characteristic nuclear profiles (arrows). H&E Bar = 25μm

39 40

41 42

43 44

Fig 45-48. (45) Hepatopancreas of Crangon crangon. Unidentified yeast-like organism (arrows) in the haemal Fig 13-16. (46) Hepatopancreas of Crangon crangon. CcBV-infected nucleus (arrow). H&E Bar = 25 μm (14) Hepatopancreas of Crangon crangon. CcBV-infected nucleus (arrow) and adjacent infected nucleus being expelled to tubule lumen (black arrow), H&E Bar = 25 μm (47) TEM of CcBV infected nucleus. Virions (arrows) are embedded in a viral stroma that fills the host nucleus (arrow). (48). TEM of CcBV infected nucleus. High power micrograph of a CcBV virion showing envelope (arrow) and nucleocapsid (double arrow).

45 46

47 48

Fig 49-52. (49) Hepatopancreas of Crangon crangon. Unidentified yeast-like organism (arrows) in the haemal sinus (S). PAS Bar = 25 μm (50) Connective tissues (Ct) of Crangon crangon. Large granuloma-like lesion associated with yeast-like organism. PAS Bar = 25 μm (51) Haemopoeitic tissue of Crangon crangon. Yeast- like organisms (Ye), apparently contained within the bounding membrane of the tissue, replace large regions of the tissue. H&E Bar = 100μm. (52) Skeletal muscle of Crangon crangon. Yeast-like cells are seen between intact muscle fibres (arrows). Remnants of muscle fibres, apparently necrotic are interspersed between (asterisk). H&E Bar = 100μm.

49 50

51 52 *

Ye *

C1617/ME3106

C1617/ME3106 Section 4: Molluscs (Mytilus species)

Mussels (Mytilus spp.) have long been used as sentinel organisms for the monitoring of the health status of aquatic ecosystems. Since the 1970’s where the global “Mussel Watch” programme attempted to monitor the long-term effects of anthropogenic substances on the marine environment, the mussel has since been the subject of numerous studies due to its ubiquity and tendency to show effects of exposure relative to its environment. More recent studies have centred on the use of more sensitive biomarkers (such as induction of the P450 system, lysosomal changes and proteomic profile changes) in an attempt to improve the resolution of exposure-effect. However, many such studies have failed to take in to account the natural variation in populations present at particular sites in particular seasons (and over several years). As such, this programme has attempted to address baseline issues of speciation, natural variation over season and site-specific differences in pathogen burden. The collection of this type of data is paramount if we are to interpret the additional factor of contaminant exposure in these species.

C1617/ME3106

4.1 Introduction

Mussels have shown considerable potential as sentinel organisms for the monitoring of the health status of aquatic systems. They are generally small, abundant, sessile and readily exhibit a range of biological responses when exposed to impacting environments. As such, they are often utilised as an ideal “biomarker” species in studies that monitor the aquatic environment. One such study, the “Mussel Watch” programme (MWP) initiated in 1976, was designed to monitor the long-term effects of anthropogenic substances on the marine environment (Goldberg et al., 1978). The programme generally focused on the quantification of synthetic chlorinated compounds, polychlorinated biphenyls (PCBs), polynuclear aromatic hydrocarbons (PAHs), radionuclides and trace metal concentrations in the tissue of mussels. Goldberg & Bertine (2000) reviewed the results of the MWP and other initiatives. They highlighted that the majority of studies carried out were directed towards the measurement of contaminant concentration, with little emphasis being placed on the actual impact of such substances. Such an approach is now considered vital in gaining a more holistic assessment of the general “health status” of a given environment. As a result, it has been suggested that the analyses of the detoxifying enzymes cytochrome P-450) and metallothionein (in organisms) in addition to estrogenic substances (in waters, sediments and organisms) might provide an indication of the potential effects of pollution on the aquatic environment and on the organisms that reside there, giving rise to a suite of biological effects techniques employed in national environmental monitoring programmes (Goldberg & Bertine, 2000). However, due to its capacity to visualise effects in situ, histopathology still remains the ‘gold-standard’ for effects at the whole organism level providing an integrative endpoint of contaminant and pathogen exposure. As such, its incorporation into these studies is recommended. The overall health status of individuals within such environments, both in terms of histopathological lesions and the presence of infecting organisms is a significant component for assessments aimed at determining ecological quality at various sites.

This current programme can be described under two main phases. Phase one involved a preliminary assessment of mussels (Mytilus spp.) collected from the Mersey, Tees and Tyne estuaries and from sites in Southampton Water and from a marine site at Brancaster during the winter (March) and summer (June) of 2003. This scoping study identified an array of pathologies and pathogens, some of which may be linked to the biological effect of exposure to contaminants. Phase two used life history information collected during Phase one to design a targeted study of two sites: the Exe estuary and Southampton Water. Results presented below are divided into those collected during Phase one and those collected in Phase two.

4.2 Materials and Methods (Phase 1)

Mussels were obtained from the Mersey, Tees and Tyne estuaries and from sites in Southampton Water and from a marine site at Brancaster. Mussels

C1617/ME3106 were kept cool in an insulated sealed container for transport back to the laboratory. For dissection, shells were opened by severance of the adductor muscle followed by removal of the tissue mass. For histology, a 3mm “steak” section was removed along a standard plane allowing for all major organs and tissues (gill, mantle, kidney, digestive gland and gonad) to be incorporated into a single section. Excised samples were placed into histological cassettes and immediately transferred to Davidson's seawater fixative (Hopwood, 1996). Fixation was allowed to proceed for 24 h before transferring to 70% industrial methylated spirit until further processing. Sections were stained with haematoxylin and eosin and analysed using standard light microscopy techniques. Digital images were obtained using the LuciaG image analysis system (Nikon, UK). Diagnostic criteria applied to histological sections are shown in Table 12.

An index of the relative abundance of adipogranular (ADG) cells in mussels was devised based upon their presence in the connective tissue of the mantle; Stage 0 (Absent- No ADG cells apparent in mantle tissue), Stage 1 (Present- ADG cells are scarce), Stage 2 (Scattered- ADG cells appear occasionally throughout mantle tissue), Stage 3 (Frequent- ADG cells present through much of the mantle. Some areas may not show consistency) and Stage 4 (Abundant- ADG cells constitute the majority of connective tissue volume).

The gonadal index as set out by Seed (1976) was used to evaluate the reproductive status of the gonads from all mussels sampled. This is briefly described as:

RESTING • STAGE 0- Inactive/neuter gonad. May also include mussels that have completed spawning.

DEVELOPING • STAGE 1 - Gametogenesis begins. No gametes can yet be seen. • STAGE 2 - Ripe gametes first appear. Gonad is one-third of its final size. • STAGE 3 - Each follicle contains roughly equal proportions of developing and ripe gametes • STAGE 4 - Gonad is two-thirds of its final size. Follicles contain mainly ripe gametes.

RIPE • STAGE 5 - Spawning commences. Male gonad is distended with morphologically ripe sperm. Female gonad is compacted with polygonal ova.

SPAWNING • STAGE 4 - General reduction in sperm density and rounded off ova as pressure is reduced. • STAGE 3 - The gonad is approximately half empty.

C1617/ME3106 • STAGE 2 - Further reduction in gonad size. Follicles are approximately one-third full of ripe gametes. • STAGE 1 - Residual gametes remain.

4.3 Results and discussion (Phase 1)

The mean gonadal stage by site was as follows (spring and summer scores shown respectively); Brancaster (3.4, 2.0), Mersey (3.0, 1.5), Tees (3.5, 0.9) and Southampton Water (2.1, 1.8) (Fig. 53). The results appear to be consistent with the normal reproductive physiology previously described in mussels, bearing in mind that spawning takes place at different times subject to geographical location, temperature and other environmental conditions. Sex ratio data indicated increased bias to male animals in samples from Brancaster, the Tees and Southampton Water with an increased bias towards females in mussels from the Mersey. In all cases this was more pronounced during the summer sampling (Fig. 54).

The application of an ADG cell index (described above) revealed little difference in the mean ADG rate across sites between spring and summer sampling. However, mussels collected from the Mersey during summer exhibited a relatively large reduction in the mean ADG rate (Fig. 55). The assessment of ADG cells present within the mantle tissue, and the reproductive status may be a good physiological state marker for use in monitoring programmes (see Phase 2).

Gametogenesis in mussels is supported by two main storage cell types, the adipogranular (ADG) cells and vesicular connective tissue (VCT) cells. These cells form a matrix of connective tissue and support the gonadal follicles where gametogenesis takes place. ADG cells consist primarily of membrane bound protein granules, lipid and glycogen and are readily seen using standard histological techniques. VCT cells consist of a single large glycogen vesicle appearing microscopically as a relatively large transparent cytoplasmic vacuole (Pipe, 1987). A previous study suggested that exposure to hydrocarbons leads to a reduction in the levels of storage cells. The quantification of these cells can therefore be used to assess the health status/condition of mussels from different environments. However, care must be used when interpreting these results. The levels of storage cells in mussels are inversely correlated to the annual reproductive cycle. During the summer months, mussels are spent and there is a period of low demand for energy once spawning is complete. This is reflected in the low levels of storage cells present. However, there is a greater abundance of food thus feeding and storage of nutrients commences. Consequently, there is an increase in the frequency and size of both ADG and VCT cells reflecting elevating levels of stored glycogen, protein and lipid in mantle tissue towards the end of the summer and beginning of the autumn period. Autumn and winter sees a higher demand for energy when gametogenesis commences. Consequently, during gametogenesis, levels of glycogen fall with concomitant reduction in storage cell numbers. However, Sunila (1986) noted that sudden acute chemical exposure can bring about the spontaneous spawning of mussels outside of the normal spawning period. 84

C1617/ME3106

Granulocytoma prevalence was found to be lowest in mussels collected during the spring sampling period compared to those sampled during early summer (Fig. 56). However, the prevalence differed between sites. Here, figures in parenthesis refer to percentage prevalence in spring and summer respectively: Southampton Water (4, 34), Tees (14, 48), Brancaster (24, 34) and Mersey (32, 60). Granulocytomas were classified by the presence of focal areas consisting of many tightly packed granulocytes and agranulocytes (Fig. 57), both of which have phagocytic properties. Often, discrete granulocytomas were melanised. Granulocytoma prevalence at each site was consistently higher during the summer, presumably associated with increased ambient water temperature, increased metabolic rate and inflammatory response to pathogens present during this period. Furthermore, it is well documented that the formation of granulocytomas occur as a result of chronic pollution and chemical exposure, both in the field and in laboratory studies (Auffret, 1988; Rasmussen et al., 1983; Sunila, 1986; Svärdh, 1999). Granulocyte aggregates are known to appear in various tissues within Mytilus sp. including the vesicular connective tissue, digestive interstitial tissue, mantle, stomach, intestine and the lumen of gonadal follicles. Auffret (1988) demonstrated the presence of such granulocytomas at all sites sampled from a Norwegian Fjord contaminated with metals and organic xenobiotics of industrial origin. Sunila (1984, 1986) also demonstrated the presence of granulocytomas in mussels sampled in close proximity to an outlet from a titanium dioxide plant. Because granulocytomas have also been associated with exposure to domestic waste (Bayne, 1980), they may also reflect a physiological response to this type of contamination, thus providing a general marker for health status in individuals. Additionally, once spawning is completed, granulocytes migrate to the male gonadal follicles where cytolysis of residual gametes takes place. However, granulocyte migration can also occur after exposure to certain chemicals such as copper and cadmium, followed by the cytolisis of even ripe sperm (Sunila, 1986).

Kidney melanisation was classified by the presence of dark brown deposits within the kidney epithelial cells (Fig. 58). During spring, melanised aggregates were only observed in mussels collected from Southampton Water and the Tees estuary at 2% and 31% prevalence respectively. However, summer sampling revealed an increase in prevalence across all sites (Brancaster, 28%; Mersey, 93%; Tees, 92%; Southampton, 43%). Although some melanisation may be associated with normal excretion of waste products, previous chemical and histological analysis has revealed that an elevated level of excretory products in the kidney is representative of the environment to which mussels are exposed (Sunila, 1986).

During the histological assessment of sections through whole mussels we also recorded the presence and absence of lysosomes within the epithelial cells of the digestive diverticula. Here, they predominantly appear as round, brown organelles. On occasion, lysosomes that were eosinophillic in appearance could also be seen (Fig. 59). Generally, lysosome prevalence was highest in mussels across all sites during the summer compared to those sampled during spring. It is well established that lysosomes in marine

C1617/ME3106 molluscs are often affected by xenobiotic-induced pathological change (Moore et al., 1987) with both structure and function becoming altered. Studies of this type have mainly centred on lipofuscin accumulation within such lysosomes (see Lowe et al., 1981; Moore et al., 1987; Lowe and Clarke, 1989). A previous study reports the presence of melanised lysosomes within the epithelial cells of the digestive diverticula (Auffret, 1988), however the frequency of lysosomes was inversely correlated to the total PAH concentration from sampled field sites. This suggests that acute xenobiotic exposure leads to a decrease in lysosomal stability. These findings suggest that recording the mere presence and absence of lysosomes may not be sufficient for use in monitoring programmes. As a result, quantitative rather than qualitative histological methods of lysosomal assessment should be included in future histopathological studies. Well established cytochemical methods measuring lysosomal stability remain the tools of choice for assessing pathological change in these organelles.

Several parasites and pathogens were also observed during Phase 1 of the study. These included Ancistrum mytili (Fig. 60), Mytilicola intestinalis (Fig. 61 and 62), Proctoeces maculates (Fig. 63), Digenean sp. (Fig. 64), Rickettsia- like organisms (Fig. 65) and the intracellular mussel protozoan termed MPX (Fig. 66). We also report the first case of Marteilia maurini from mussels collected in UK waters (Fig 67 and 68). This parasite was observed infecting mussels during the summer sampling with prevalences of 2% from Brancaster and 6% from Southampton Water. Another species, Martelia refringens has been associated with mortalities within oyster populations in mainland Europe and is currently a notifiable disease under the OIE. Consequently, this first report of a Martelia sp. parasite in UK waters is significant. Additionally, we also report the first case of Steinhausia mytilovum from UK waters. This parasite was found in mussels collected during the summer sampling at Brancastor and the Tees estuary at 6% and 2% respectively. Termed ‘Mussel Egg Disease’, this microsporean is believed to cause reduced fecundity of those individuals infected (Bower, 2001). It is characterised by the presence of multiple spores within the cytoplasm of the host egg that is generally associated with a haemocytic infiltrate (Fig. 69). Mussels potentially harbour several different parasites and this feature has been exploited in studies seeking correlations between parasite burden and contaminant exposure (Auffret, 1988; Svärdh et al., 2002; Svärdh, 1999;). Auffret (1988) described infections of a larval trematode within the digestive gland of Mytilus edulis, closely resembling the trematode Proctoeces maculates. Although the parasite has been known to have an adverse effect on gametogenesis during heavy infection, no correlation was found between parasite burden and contaminant exposure. Similarly, no correlations were found during this study. A multivariate statistical analysis of parasite burden at the individual level in relation to other physiological and environmental factors will be possible at the end of Stage 2 of this study (see below).

4.4 Conclusions (Phase 1)

The first phase of this study has demonstrated an array of histopathological features that can be used for both the assessment of health status and

C1617/ME3106 condition of mussels at multiple sites. Such data is essential in order to accurately interpret specific biomarker data (e.g. CYP1A) that may also be collected during monitoring activities. This component of the present study has provided invaluable data on pathological changes present in mussels in UK waters. However, since gonadal maturity and ADG rate can relate to the overall health status of an individual (and presumably its ability to respond to additional stressors such as contamination), a more detailed sampling strategy needs to be initiated so that sound baseline data can be gathered. In addition, the issue of mixed species populations (Mytilus edulis, Mytilus galloprovincialis and hybrids of the two species) that are known to inhabit UK sites and the implications of their use in monitoring programmes needs to be stringently addressed. A monthly sampling programme at two sites designed to investigate interrelationships of these different factors, including environmental parameters is described in Phase 2.

4.5 Phase 2: Monthly sampling of mussels

Phase 1 revealed how histopathology is a valuable tool for assessment of health status in mussel populations. Historically, a large percentage of research carried out incorporating mussels as a sentinel biomarker species have used stocks believed to be composed of pure Mytilus edulis (Earnshaw et al., 1986; Auffret, 1988; Wedderburn et al., 2000). However, previous studies have suggested that certain geographical locations might in fact contain mixed populations of mussels (Wood et al., 2002); with Mytilus edulis being joined by the supposed Mediterranean co-species Mytilus galloprovincialis. Moreover, it is also known that “hybrids” of these two species can occur also occur, with populations at some sites being composed of M. edulis, M. galloprovincialis and hybrids of the two.

In terms of biomonitoring using mussels this poses a series of important questions:

1. Have biomarker studies on supposed M.edulis actually been carried out on mixed populations of pure species and hybrids? 2. Do different Mytilus species show variations in reproductive physiology, maturation and condition at different times of year? 3. Do different Mytilus species show differential response to contaminant exposure and is this response seasonally dependent? 4. Do different Mytilus species show different pathogen burdens? 5. Is it sufficient to assign a dominant species type to a particular site or do different sub-populations of particular species exist? 6. Should speciation tests be included in all monitoring studies using mussels?

Some evidence for 4. (above) does exist. It has also been reported that parasitism affects M. edulis and M. galloprovincialis collected from the same site differently. It has been reported that Pinnotheres pisum, commonly known as the Pea Crab, appears to preferentially infect Mytilus galloprovincialis, over Mytilus edulis (Seed 1969). Additionally, the trematode parasite Prosorhynchus squamatus has been detected at higher prevalence in M.

C1617/ME3106 edulis than in M. galloprovincialis (Coustau et al., 1991). Lobel et al. (1990) discovered higher concentrations of metals present in Mytilus trossulus compared to those seen in M. edulis of the same size range. However, it was discussed that this could be related to the slower growth rate of M. trossulus. Nevertheless, it is clear that this issue requires full investigation. Consequently, this may have serious implications for accurate interpretation of biological data resulting from environmental studies. This is of particular importance to the OSPAR Joint Assessment and Monitoring Programme (JAMP) because of the need and requirement for a quality assurance reference relating to mussels and their use in monitoring programmes.

Phase 2 of the mussel sampling programme from the Exe estuary and Southampton Water involves monthly sampling and results collected are presented below. The primary aim of the sampling programme is to collect mussels from each of the sites in order to assess temporal trends in gonadal status, condition, pathology and parasites. This baseline data from a relatively uncontaminated (Exe) and relatively contaminated waterways (Southampton) provides invaluable baseline reference data for further studies using mussels as a biomonitor. Furthermore, using a molecular test, speciation of individual mussels to determine the extent of mixed populations at the study sites is possible. This data will be used to investigate differences in gonadal status, condition, pathology and parasites between species groups from a single site (to determine effect of speciation on status) and between sites (to determine effect of species proportions on marker status). This is the first time that speciation data has been linked to biological effects data in such a way. Preliminary outcomes from these analyses are provided below.

4.6 Materials and Methods (Phase 2)

Mussels were collected from fixed sampling sites, once a month over a nine- month period between September 2004 and May 2005 and will be continued outwith this contract to produce a 12-month dataset. Sampling sites were Cracknore Hard in Southampton Water and Starcross on the River Exe. For standardisation, 50 mussels were collected from the low water mark on all sampling occasions. Salinity and temperature recordings were also obtained for each site and each month. Mussels were subsequently transferred into approximately 15 L of estuarine water collected from the sampling location and kept cool in an insulated sealed container. For dissection, shells were opened by severance of the adductor muscle followed by removal of the tissue mass. For histology, a 3mm “steak” section was removed along a standard plane allowing for all major organs and tissues (gill, mantle, kidney, digestive gland, gonad and foot) to be incorporated into a single section. A sample of adductor muscle was also obtained. Further samples obtained included mantle and digestive gland (for DNA adduct analysis), mantle and adductor muscle (for metabalomic analysis) in addition to mantle and gill for speciation. All were frozen in individual cryovials prior to analysis. Metabolic analysis will take place at the University of Birmingham under a separate contract.

C1617/ME3106 Speciation. For investigation of species status, molecular speciation was carried out on frozen mussels sampled from Southampton Water during August and September 2004; and mussels sampled from the River Exe during September 2004. Eventually, each individual mussel sampled will be speciated. DNA was extracted from individual samples following sterile maceration and vortexing with 1ml of maintenance media. One hundred μl of tissue supernatant from each vial was added to 1ml DNAzol. Total DNA was extracted using the standard DNAzol extraction protocol with nucleic acid being re-suspended in 100 μl of Rnase/Dnase free water and then stored at - 20°C pre PCR. For PCR, a standard 35 round PCR cycle was used to generate products from the 5’ end of the Glu gene employing the ME15 and ME16 primer (Inoue et al. 1995). The Glu-5’ gene distinguishes alleles specific to M. edulis (180 bp) and M. galloprovincialis (126 bp) by a 54 bp insertion/deletion polymorphism. While we only investigate a single allele marker in this study research has demonstrated that the use of Glu-5’ gene marker and allozyme loci used in previous studies of hybrid mussels are strongly correlated (Hillbish et al., 2002). Each assay contained negative controls and a 100bp DNA ladder as a marker. Post PCR, all products were separated on a 4% agarose gel by electrophoresis and the images recorded using UVP equipment and relevant software.

Histology. Excised samples were placed into histological cassettes and immediately transferred to Davidson's seawater fixative (Hopwood, 1996). Fixation was allowed to proceed for 24 h before transferring into 70% industrial methylated spirit until further processing. Sections were stained with haematoxylin and eosin and analysed using standard light microscopy techniques.

4.7 Results and Discussion (Phase 2)

Speciation. Results for speciation carried out in Southampton Water during August 2004 are shown in Fig. 74 a-c. Three distinct profiles were observed; a form containing a single product of approximately 200bp, a form containing a single product of approximately 150bp, and a form containing both products. The products were confirmed as the 180bp and 126bp products described by Inoue et al. (1995) by sequence analysis. The samples chosen for analysis represented the three different gel profiles (Southampton- August Samples/Lanes 3, 7, 10, 20, 26 and 30). Approximately 40ul of each band was run through a 4% agarose gel and each fragment of interest was then carefully excised into individual corresponding containers. Each tube was then frozen at -20oC and then subjected to the 'Freeze n Squeeze' method of DNA recovery.

A standard (Direct) sequencing PCR protocol then followed for each product band with both strands of the DNA being sequenced using the ABI Prism cycle sequencing system. Sequences were analysed on the ABI 3100 genetic analyser and a consensus sequence was generated using sequences generated from the forward and reverse sequences. The consensus sequences were compared to sequences submitted to GenBank and EMBL databases using the BLAST search program available at the UK HGMP 89

C1617/ME3106 Resource Centre, Hinxton UK. Sequence alignments were performed using the Sequencher software package. Analysis revealed the following result for each band:

- Products for sample numbers 10 and 20 were 180 base pairs long and shared 99% nucleotide identity with the published sequence for M. edulis. - Products for sample numbers 3 and 30 were 126 base pairs long and shared 98% nucleotide identity with the published sequence for M. galloprovincialis - Products generated for samples 7 and 26 were subdivided in to two (7a - 7b & 26a - 26b).- products 7a and 26a were 126 base pairs long and shared 100% nucleotide identity with the published sequence for M. galloprovincialis while products 7b and 26b were 180 base pairs long and shared 99% nucleotide identity with the published sequence for M. edulis.

Results for speciation carried out in the River Exe during October 2004 are shown in Fig. 74 d-e. This revealed a population consisting entirely of M. edulis type alleles. As a result of previous mussel identification from the Southampton August 2004 samples above, we were able to identify the amplified fragments due to their molecular weight/marker size relation (approx 180bp). Although these DNA products have not undergone sequencing identification we are able to suggest with a high degree of probability that the samples of interest are of M.edulis.

Histology. The mean gonadal stage and adipogranular (ADG) rate of mussels from the Exe site appeared to change in a phase-shifted manner. This was demonstrated by a decreasing ADG rate as mussels presumably directed energy reserves into gametogenesis (Fig.76). Mussels sampled from Southampton Water showed a similar pattern (Fig.78), however reproductive effort appeared less intensive with some differences observed in those mussels sampled between September 2004 and November 2005. During September 2004, the mean gonadal stage was c.2, gradually rising to c.3.5 by March 2005. Mussels sampled from the Exe during September 2004 had a mean gonadal stage of c. 0.5. However, the gonadal stage during March was similar to that observed in mussels from Southampton Water (c. 3.5). The ADG rate of mussels from both sites was similar in that they both reached a plateau between September 2004 and November 2004, followed by a decrease in December 2004. During March 2004, mussels from Southampton Water and the Exe both continued to show a decline in the ADG cell rate. Although ADG rate during the first 7 months was similar throughout for both sites, Southampton Water generally had a lower ADG rate compared to the River Exe (Fig. 76 and 78). The differences that can be seen between the two sites in figures 76 and 78 may be related to the prevalence of species at each site. As stated above, initial findings appear to indicate that the River Exe is a pure M. edulis site which may explain the classic gonadal stage/ADG rate plot seen in figure 76. However, the majority of those mussels sampled from the Exe had a gonadal stage of 1 or 2, and an ADG rate of 3 or 4. This is consistent with the theory that M. edulis, M. galloprovincialis and their hybrids have different reproductive strategies since 100% of the mussels at the River Exe appear to be M. edulis.

C1617/ME3106 Analysis of data based upon speciation of mussels collected from each of the sites has revealed some interesting differences. Gonadal stage and ADG rate for mussels sampled from Southampton Water during August 2004 was calculated based upon species. M. edulis displayed a significantly lower mean gonadal stage in comparison to M. galloprovincialis sampled from the same site in the same month Furthermore, hybrid mussels appeared to demonstrate a gonadal stage intermediate between the two species (Fig.80). ADG rate also appeared to be related to species. M. edulis exhibited a higher ADG rate compared to that of M. galloprovincialis sampled from the same site in the same month (Fig. 81). The result would be responsible for differences in the reproductive strategies of these two species (and their hybrids) inhabiting this site. Interestingly, the ADG rate of hybrids was between M. edulis and M. galloprovincialis types.

In additional to analyses based upon species differences, our studies have also revealed differences in the percentage prevalence of certain pathological and parasitological parameters in mussels sampled from the Exe and Southampton Water sites. In addition, the relative significance of these differences appears to change throughout the year. Mussels from the River Exe appear to harbour a larger number of parasites with higher numbers of Digenean sp. and Ancistrum mytili being recorded (Fig. 82 and 88) compared to Southampton Water. Mytilicola intestinalis appeared to have a similar prevalence at both the River Exe and Southampton. The prevalence of inflammation and granulocytomas between sites also varied. At Southampton, prevalence decreased between September and November (Fig. 87). However, those levels observed at the Exe have generally remained constant throughout the sampling period to date (Fig. 89).

4.8 Conclusions (Phase 2)

A number of important findings have emerged from the preliminary analysis of data pertaining to seasonal sampling of mussels from the Exe estuary and Southampton Water. These are:

1. Variation in reproductive and gonadal status by season and differences in seasonal patterns between sites. 2. The presence of numerous pathogens and pathologies that can be easily monitored using histopathology (and therefore incorporated into environmental monitoring programmes using mussels). 3. Variation in pathogens and pathologies by season and differences in seasonal patterns between sites. 4. The presence of mixed populations of M. edulis, M. galloprovincialis and a hybrid of the two species at sites currently used for estuarine monitoring. 5. Different patterns in reproductive and reserve storage allocations between the different species present at a given site. 6. Evidence for differences in pathogens and pathologies between the different species populating a given site.

C1617/ME3106 As further monthly data is collected, multivariate statistical analysis, as shown for fish in Section 4, will be applied. In this way, rigorous baseline data for mussels residing at these sites will be obtained. The major finding of differential response in M. edulis, M. galloprovincialis and the hybrid form is considered a key finding of this study and has implications for future UK and global monitoring efforts using these species. Further work is now required to investigate the extent of mixed species at monitoring sites. .

C1617/ME3106

Table 12. Criteria used during the histological evaluation of mussels.

PHYSIOLOGICAL/PATHOLOGICAL EFFECTS PARASITE FAUNA

CONDITION LOCATION IN HOST PARASITE SPECIES LOCATION IN HOST

Sex Gonad Ancistrum mytili Gill

Reproductive stage Gonad Digenean sp. Vesicular connective tissue / foot Lysosomes Digestive diverticula Marteilia refringens Digestive diverticula / intestine Degeneration Digestive diverticula Marteilia maurini Digestive diverticula / intestine Melanised aggregates Kidney Mytilicola intestinalis Intestine / stomach Haemocytic neoplasia Vesicular connective tissue Steinhausia mytilovum Oocytes Other neoplasia Throughout Proctoeces maculatus Vesicular connective tissue

Vesicular connective tissue / Inflammation Vesicular connective tissue Haplosporidium sp. digestive diverticula

Granulocytomas Vesicular connective tissue Rickettsia-like organisms Digestive diverticula / gill Apoptosis Gonad Chlamydia-like organisms Digestive diverticula / gill Atresia Gonad Mussel protozoan X (MPX) Digestive diverticula Intersex Gonad Pea crab infection Shell cavity

C1617/ME3106

53 Mean gonadal stage by site

2 Gonadal stage

0 ) ) ) ) M) P P P M S S S M) S ( ( ( (S ( R (SP) R S EY E N (S TE EY (SM) S TON TO RS RS TE TEES O O A ASTE S ME S NC ME RA BRANC B Site sam pled

Fig. 53 Mean gonadal stage by site for spring (SP) and summer (SU) sampling.

54 Mussel sex ratio data (M:F)

2.5

1.5

1 Ratio (M:F)

0.5

0 ) P) U U) S SP) (S ( S (SU) ER EY (SP) ES T E TON RS T TEES ( OTON (SP) O S S ME MERSEY (SU) RANCASTER ( BRANCAS B Site sam pled

Fig. 54 Sex ratio (M:F) of mussels by site for spring (SP) and summer (SU) sampling. 94

C1617/ME3106

55 Mean ADG rate by site

3.5 3 2.5 2 1.5

ADG RATE 1 0.5 0

) ) ) ) P U P P S S (SP) (S ( (SU) R (S Y S N E E TE S ON O S RSEY ( TE TEES (SU)OT A E S SOT NC MER M

BRA BRANCASTER (SU) Site Sampled

Fig. 55 Mean adipogranular (ADG) cell rate by site for spring (SP) and summer (SU) sampling.

56 Prevalence of granulocytomas

30 Prevalence (%) 20

0 ) ) P) U P (S (S (SU) R E ER EY T S ST R TEES (SP) TEES (SU) A E SOTON (SP) MERSEY (S SOTON (SU) NC M A R B BRANCAS Site Sam pled

Fig. 56 Prevalence of granulocytomas by site for spring (SP) and summer (SU) sampling. 95

Fig. 57 Granulocytomas (arrow) within the vesicular connective tissue of mussel.

Fig. 58. Melanised aggregates within the epithelial cells of kidney indicative of increased secretion of waste products.

Fig. 59 Lipofucshin-containing lysosomes within the epithelial cells of the digestive diverticula. Note the presence of both brown (white arrow) and pink (black arrow) lysosomes.

Fig. 60 Ancistrum mytili attached to the gill lamellae of mussel. These ciliate parasites are thought to be harmless filter-feeding commensals.

Fig. 61 Copepod, Mytilicola intestinalis present within the stomach lumen of mussel.

Fig. 62 Intestine containing an individual Mytilicola intestinalis parasites. Note the legs of the parasite attached to intestinal wall (arrows).

Fig. 63. Mussel trematode disease (Proctoeces maculates). Sporocysts contain numerous developing cercariae within the vesicular connective tissue of the digestive diverticula.

Fig. 64 Digenean sp. parasites present in the mussel foot (arrow). In this figure, a total of fifteen individuals can be seen

Fig. 65. Rickettsia-like organisms (arrow) within the secondary gill lamellae.

Fig. 66. Mussel protozoan X (MPX) within the digestive epithelial cells. This intracellular ciliate appears to cause no host response even in the presence of large numbers

Fig. 67 Marteilia maurini infection within the epithelial cells of the digestive diverticula.

Fig. 68 Higher magnification of Marteilia maurini infection within epithelial cells of the

digestive diverticula.

Fig. 69. Oocytes containing the microsporidian, Steinhausia mytilovum (mussel egg disease). Note the presence of haemocytic infiltration (*) and multi-nucleate sporocysts within the oocyte cytoplasm.

Fig. 70. Degeneration of the digestive diverticula tubules. Note the presence of sloughed cells in the tubule lumen.

SOUTHAMPTON

RIVER EXE

Fig. 71 Map of the SW England showing sampling sites for the mussel histolopathology temporal trends study (Phase 2).

Fig. 72 Temporal trend sampling location: Cracknore Hard, River Test, Southampton Water, Hampshire.

Fig. 73. Temporal trend sampling location: Starcross, River Exe, Devon.

Fig. 74. PCR product gels for mussel samples from Southampton (August 2003). Samples 1-10 (a), Samples 11-30 (b) and Samples 31-50 (c). Black arrow highlights larger band specific to M. edulis (180 bp), blue arrow highlights smaller band specific to M. galloprovincialis (126 bp) and while arrow shows a hybrid containing both sized bands. The negative controls are in lane 1 and 12 on gel a, lane 1 on gel b and lane 22 on gel c. The 100bp size marker is present in the right hand lane of all gels.

74a 1 12

1 74b

74c

74d 74b

N1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 L

19 20 21 22 23 24 25 N2 26 27 28 29 30 31 32 33 34 35 36 L

74e

37 38 39 40 41 42 43 44 45 46 47 48 49 50 N3 L

Fig. 74. PCR product gels for mussel samples from River Exe (October 2004). Samples 1- 36 (d), Samples 37-50 (e). The negative controls are in lane N1 and N2 on gel d, and lane N3 on gel e. The 100bp size marker is present in the lane L of both gels..

SPECIES SOUTHAMPTON RIVER EXE M. edulis 12 100 M.edulis/gallo 50 0 M. galloprovincialis 38 0

Fig. 75. Table showing percentage prevalence of the three species that were present at Southampton Water (August 2004) and the River Exe (October 2004).

Gonadal stage versus ADG rate (River Exe) Gonadal stage vs ADG rate (River Exe) 76 77 Gonadal stage 5 4 ADG rate y = -0.6143x + 3.9833 R2 = 0.7774 4 3

3 2

2 ADG rate 1

1 0

0 012345 Sept Oct Nov Dec Jan Feb Mar

Month Gonadal stage

Gonadal stage vs ADG rate (Southampton) ADG rate vs Gonadal stage (Southampton) 78 79 Gonadal stage 5 4 ADG rate y = -1.1616x + 4.8135 R2 = 0 .78 12 4 3 3 2

Rate 2 ADG rate 1 1 0 0 Sept Oct Nov Dec Jan Feb Mar 012345 Month Gonadal stage

Species against Gonadal stage Species against ADG rate (Southampton) 80 (Southampton) 81 4 4 3 3 2 2

1 ADG rate 1 0 0 Gonadal Stage M.e M.e/M. g M. g All M.e M.e/M. g M. g All Species Species

A. Mytili prevalence (Exe) Prevalence of M. Intestinalis (Exe) 82 83

100 100 90 90 80 80 70 70 60 60 50 50 40 40 30 30 20 20 Percentage % Percentage % 10 10 0 0 Sept Oct Nov Dec Jan Feb Mar Sept Oct Nov Dec Jan Feb Mar

A. Mytili prevalence (Southampton) Prevalence of M. Intestinalis 84 85 (Southampton).

100 90 100 80 90 70 80 60 70 50 60 50 40 40 30 30 20 Percentage % 20

10 Percentage % 10 0 0 Sept Oct Nov Dec Jan Feb Mar Sept Oct Nov Dec Jan Feb Mar

Prevalence of Digenean sp. Parasite Inflammation and Granulocytoma 86 87 Inflammation (Southampton) prevalence (Southampton). Gr anul ocytoma

100 100 90 90 80 80 70 70 60 60 50 50

40 40 30 30 20 Percentage %

Percentage % 20 10 10 0 0 Sept Oct Nov Dec Jan Feb Mar Sept Oct Nov Dec Jan Feb Mar

Prevalence of Digenean sp. Parasite (Exe) Inflammation and Granulocytoma 88 100 89 prevalence (Exe) Inflammation 90 Gr anul ocytoma 80 10 0 90 70 80 60 70 50 60 40 50 40 30 30 Percentage % 20 20 Percentage % 10 10 0 0 Sept Oct Nov Dec Jan Feb Mar Sept Oct Nov Dec Jan Feb Mar

C1617/ME3106

106

C1617/ME3106

Section 5: Multivariate statistics

Environmental monitoring programmes should aim to collect multi-parameter data at a range of different physico-chemical and ecological levels in order to assign a relatively simple category to a particular site of interest. Once categories have been assigned, managers can decide on where resources should be targeted for remediation or further assessment and monitoring effort. To date, alignment of this complex data has been problematic with coherent outcomes often lacking. The current programme has helped to clarify an approach to this issue by attempting to utilise relevant datasets to explore relationships on three major levels: site-to-site, season-to-season and year-to-year. It has become apparent that in order to achieve this, a baseline must be established against which changes over space and time can be assessed. Only then can complicating factors (such as those associated with contaminant exposure) be estimated. The establishment of the Aquatic Health Database (AHD) and first attempts to apply multivariate statistical approaches to analysis of aquatic animal health during this programme constitute a new approach to biological effects assessment. This offers significant benefits for analysis of environmental and biological effects data for managers to make informed policy decisions regarding the aquatic environment.

107

C1617/ME3106

5.1 Discrimination using multivariate statistics

A major challenge in modern environmental monitoring programmes is to align and interpret complex and potentially disparate datasets into decipherable and coherent advice lines. Traditional approaches to aquatic health monitoring have relied on specific markers for exposure (e.g. biochemical markers, pathology) and their tentative link to contaminant burdens in biota, sediment or water. However, to date, inconsistent sampling protocols and a lack of coherence between chemists and biologists have blurred any clear patterns between exposure and effect. In addition to analysing cause and effect, modern approaches to monitoring also need to utilise all relevant datasets to explore relationships on the three major levels of site-to-site, season-to-season and year-to-year. The importance of establishing this baseline (i.e. the patterns that exist in normal populations over space and time) will be fundamental in deciphering complicating factors associated with contaminant exposure. Handling such datasets requires a structured system (e.g. the Aquatic Health Database: AHD) that can be interrogated by environmental scientists for subsequent downstream statistical analysis (e.g. by multivariate statistical analysis). Only when these systems are in place can best estimate advice be reliably provided to policy makers.

This section of the report describes preliminary multivariate analyses of fish and shellfish health data collected under the current program. Analyses are based upon the use of the PRIMERTM software package developed at the Plymouth Marine Laboratory, UK (PRIMER-E Ltd., UK). Analysis has highlighted some major issues concerning regularity of sampling and the establishment of baseline data for the main sentinel fish and shellfish species employed in UK estuarine monitoring. However, the combination of analysis packages such as PRIMERTM with rigorous data collection and storage systems (e.g. based on the AHD) will increase our ability to provide advice on true biological effects of contaminant exposure and will form a central resource for tracking improvement or decline in site quality over time.

Due to the limitations on sampling during the current program, pathology and parasitology data was analysed to highlight site-to-site differences. Analyses of this type will be usefully for pinpoint ‘problem’ and ‘reference’ areas and as such will be useful for directing monitoring effort in the future. However, as stated above, a significant gap in knowledge exists for changes that may occur in sentinel species over a whole season. These season-to-season changes may help to mask true biological effects of contaminant exposure, this making site-to-site or year-to-year comparison impossible. For this reason, we have collected mussels (Mytilus spp.) from two sites (Exe estuary and Southampton Water) monthly over a 9-month period (12 months when completed) to address normal seasonal changes. This type of data is crucial for successful planning of biological effects monitoring studies using this species (see Section 4). It is suggested that future studies on flounder from key estuarine sites should also take these seasonal changes into account, particularly due to the migratory nature of this species (see recommendations in Section 7 of this report).

108

C1617/ME3106

5.1.1 Parasites

The use of multiple dimensional scaling (MDS) using parasite presence/absence data was able to discriminate the estuarine sites sampled during October 2002 (Fig. 90). The parasite populations from the Alde reference site had a similar structure to those from the Thames, Tyne and Southampton water. Parasite populations from the Forth, Mersey and Clyde were different from the reference site. Interestingly, flounder from the Forth, Mersey and Clyde were also shown to contain the highest number of pathological changes in the liver (see section 2 of this report and data shown below). The alignment of parasite data with observational data pertaining to likely contaminant effects on the liver demonstrates that parasite data could potentially be utilised to estimate contaminant effect. The method utilised here was unable to accurately discriminate flounder from the Tyne from those collected in the Alde. The parasite diversity indices for both sites were not statistically significant and the types of parasites recorded were also similar. However, examination of the differences in parasite prevalences between the two sites clearly shows differences. Higher levels of Z. viviparus, C. heterochrous, Contracaecum sp. and H. aduncum in Tyne flounder broadly reflects the presence of pollution tolerant invertebrate intermediate hosts such as Buccinum and some polychaetes at this site. In comparison the higher levels of P. varius, Diplostomum sp. and D. minutus in flounder from the Alde likely reflects the presence of pollution sensitive invertebrate hosts at this site.

Use of MDS was able to discriminate the reference site (Alde) from the four contaminated sites included in the analysis when only parasite data from flounder collected in autumn 2002 was used (Fig. 91a). The pattern was best explained by the digeneans Hemiuris communis, Diplostomum sp. and digeneans in the gills, the nematode Cucullanus heterochrous and the parasitic copepod Acanthochondria cornuta. A significant correlation coefficient of 0.988 was obtained. This value is equivalent to saying that 98.8% (i.e. 0.988 of 1) of the pattern observed is due to these 4 parasites. The bubble plots associated with this MDS plot (Fig. 91b to f) provide visual confirmation of the reason that these parasites are able to discriminate the sites, with the Mersey and Thames sites generally having intermediate levels of parasitism for any given parasite species compared with the reference site and the Tyne and Forth sites. Whilst the combination of a number of species can be used to discriminate all sites, individual species can be used to discriminate individual sites from each other. For example the digenean H. communis is found at relatively high prevalences at the Forth and Tyne sites, at intermediate levels at the Mersey and Thames and absent from the Alde (Fig. 91b) and thus the Mersey and Thames sites can be discriminated clearly from the other sites. Conversely, digeneans in the gills are found at relatively high levels in the Alde, with a decreased prevalence in the contaminated sites (Fig. 91d).

5.1.2 Liver lesions

One of the clearest MDS plots was generated using data pertaining to liver lesions in flounder collected during the Autumn 2002 sampling period. As with

109

C1617/ME3106

the parasite data, using only liver lesions data it is possible to clearly discriminate between the contaminated sites and the reference site (Alde) (Fig. 92a). The pattern observed was best explained by the increasing prevalence of fibrillar inclusions, hepatocellular and nuclear polymorphisms, coagulative necrosis, melanomacrophage centres and lymphocytic/monocytic infiltration at contaminated sites compared with the Alde (correlation coefficient of 0.987 - i.e. 98.7% of the discriminating pattern observed is due to these 5 pathologies). The most obvious factor that explained the pattern was the prevalence of fibrillar inclusions (Fig. 92b – see Section 2 of this report). The plot shows that the prevalence of fibrillar inclusions decreases from top to bottom and from left to right on the plot going from a high prevalence in the Clyde in the top left quadrant to a low prevalence in the Thames in the lower right corner and a prevalence of 0% in the Alde (top right corner).

5.1.2.1 Within site variation over time

Although as stated above, analysis of temporal differences is restricted due to sampling limitations in the current programme, sufficient numbers of data points were obtained for prevalence of liver lesions in flounder collected in the rivers Alde, Mersey and Tyne over the time course of the program to allow for temporal trends to be assessed. For liver lesions in flounder from the Mersey, the MDS plot generated formed a typical horseshoe shape with a clear change in the profile over time (Fig. 93). This pattern is suggestive of seasonal changes in liver pathology throughout the sampling period.

In contrast to the data collected for liver lesions in flounder from the Mersey, liver lesion data for flounder from the Tyne showed a somewhat more erratic, but nonetheless distinct pattern in the MDS plot (Fig. 94). The patterns observed in liver lesions prevalence/presence were most similar between Autumn 2000 and Autumn 2002. Samples taken in Spring 2000 and Autumn 2002 were very dissimilar.

A similar pattern was observed in the MDS plots of prevalence of liver lesions in flounder caught in the Alde between Spring 2000 and Autumn 2002 (Fig. 95). However, samples taken in Spring 2000 and Autumn 2000 were most similar, whilst the variation between the three autumn samples taken in 2000, 2001 and 2002 was greatest. The similarity in shape of the plots (zigzag) for the Alde and Tyne data would suggest that there is a factor impacting on these sites at a regional level. Subtle site-specific factors explain the individual plots seen and highlight the requirement for site specific baseline information to be collected for sentinel species residing at these sites if they are to be used as environmental sentinels as intended.

When taken at the larger geographical scale and incorporating liver lesion data from all sites over the period 2000 to 2003, patterns are still apparent in the data, as demonstrated by Fig. 96. Sites with supposedly higher contaminant burdens still group together with the exception of a sample from the Tees in 2000 which groups with the Alde sample from 2002. Otherwise, the four reference samples taken in Spring and Autumn 2000, Autumn 2001

110

C1617/ME3106

and Autumn 2002 are occupy the left of the plot and are clearly demarcated from the contaminated sites. Although not immediately clear from the plot, there appears to be a slight cyclical trend in the data that is similar to the trend seen in the Tyne and Alde data with data collected in 2000 grouping on the right side of the plot, 2001 data on the top left quadrant, 2002 towards the bottom of the plot and the 2003 more or less centrally in the plot. The significance of this pattern is not known.

5.1.3 Gill lesions

Data for gill lesions in flounder sampled from 7 estuarine sites around the UK during Autumn 2002 analysed using Principle components analysis (PCA) and MDS, are shown in Fig. 97 (A-G). Sites could be discriminated from each other, with the Alde reference site apparently closely associated with the Forth site. The main reason for the strong association between the Forth and Alde sites is the increased levels of hyperplastic gills and lower prevalence of Trichodina spp. at these two sites as demonstrated by the PCA plot (Fig. 97A). A model that includes the prevalences of aneurysms, hyperplasia of the filaments, Trichodina and Rickettsia, which has a correlation coefficient of 0.990 (i.e. 99% of the pattern seen is due to these pathologies), best explains the patterns observed in the MDS plots. Individual pathologies that are useful to discriminate individual sites include aneurysms that group the Thames and Clyde sites separately from the other 5 sites (Fig. 97 C). Three distinct groups can be identified on the basis of the prevalence of Trichodina spp., namely Southampton; Forth/Alde; and Thames/Clyde/Tyne and Mersey (Fig. 97 F).

5.1.4 Combined data for all sites in 2002

Although the continuity of data presented in this analysis is somewhat disjointed, it may be feasible to eventually combine parameters relating to pathology, parasitology, biomarkers and chemistry to create a overall measure of health status within a site, season or year. By using Bray-Curtis similarity index (BCSI), MDS and PCA we were able to discriminate relatively contaminated sites (Tyne, Thames, Southampton, Forth and Mersey) from the reference site (Alde), and from each other. Whilst each method provided slightly different results, it is clear that the relatively contaminated sites have similar pathologies and pathogens that may contribute to the production of an “index” to allow classification of the health status of individual estuaries. This is the ultimate goal for multivariate analyses of this type.

Use of the Bray-Curtis similarity index (Fig. 98), clustered the data into three main groups, namely Forth and Southampton; Mersey, Thames and Tyne; and the Alde. This grouping is clearly not related to geographical location. Given the apparent discrimination of the relatively contaminated sites from the reference site, and the highly variable environmental conditions at each of these sites, the data may reflect differences in the contamination status of respective sites. Further analysis of data of this type (preferably using continual datasets collected from respective sites) will clarify these issues.

111

C1617/ME3106

Use of MDS (Fig. 99), has also demonstrated that each site can be discriminated from each other, and more importantly that the relatively contaminated sites can be discriminated from the reference site although the patterns more clearly observed in the BCSI are less clear using the MDS approach.

One of the major benefits of PCA (and the bubble plot function using MDS in PRIMERTM) is firstly that it allows sites to be discriminated from one another and secondly, it provides information on which components (parameters) within the combined dataset are most important at generating this discrimination (Fig. 100). In the current study, 100% of the variance was explained by the first 5 principal components. In the example of flounder, these components were inflammation and hyperplasia of the skin, monogeneans and fusion on the gills, liver pathologies and the presence of the parasitic copepod Lepeophtheirus pectoralis on the skin. This type of information will be instrumental in generating relevant ‘biomarker’ lists that are proven discriminators of sites, seasons and years.

112

C1617/ME3106

Fig.90 MDS plot of parasite presence/absence data for flounder collected at all sites in Autumn 2002. Parasite populations in flounder from the Alde, Thames and Tyne group together, compared with those from the Forth, Mersey, Southampton and Clyde, which are more distant from the central group.

113

C1617/ME3106

Fig. 91 Parasite fauna of flounder collected from 5 estuarine sites in Autumn 2002. (A) MDS configuration plot with clusters of similarity overlaid onto the plot. Relatively contaminated sites are clearly discriminated from the reference site (Alde). (B) Same MDS configuration plot with superimposed circles representing prevalence of the digenean Hemiuris communis. (C) Same MDS configuration plot with superimposed circles representing prevalence of the digenean Diplostomum sp. (D) Same MDS configuration plot with superimposed circles representing prevalence of digeneans in the gills. (E) Same MDS configuration plot with superimposed circles representing prevalence of the nematode Cucullanus heterochrous. (F) Same MDS configuration plot with superimposed circles representing prevalence of the copepod Acanthochondria cornuta.

Transform: Presence/absence Transform: Presence/absence Resemblance: S17 Bray Curtis similarity Resemblance: S17 Bray Curtis similarity Stress: 0.01 Stress: 0.01 D-HEM

A Mersey B Mersey 2

Alde TyneForth Alde TyneForth 14

Thames Thames

Transform: Presence/absence Transform: Presence/absence Resemblance: S17 Bray Curtis similarity Resemblance: S17 Bray Curtis similarity Stress: 0.01 D-DIP C D Stress: 0.01 GI-DIG Mersey 5 Mersey 4

20 16

Alde TyneForth 35 Alde TyneForth 28

50 40

Thames Thames

Transform: Presence/absence Transform: Presence/absence Resemblance: S17 Bray Curtis similarity Resemblance: S17 Bray Curtis similarity Stress: 0.01 N-CUC Stress: 0.01 CO-ACA E F Mersey 5 Mersey 3

20 12

Alde TyneForth 35 Alde TyneForth 21

50 30

Thames Thames

114

C1617/ME3106

Fig.92 Liver lesions in flounder collected from 7 estuarine sites in Autumn 2002. (A) MDS configuration plot with clusters of similarity overlaid onto the plot. Relatively contaminated sites are clearly discriminated from the reference site (Alde). (B) Same MDS configuration plot with superimposed circles representing prevalence of fibrillar inclusions (max bubble size 80 %). (C) Same MDS configuration plot with superimposed circles representing prevalence of hepatocellular and nuclear polymorphism (max bubble size 60 %). (D) Same MDS configuration plot with superimposed circles representing prevalence of coagulative necrosis (max bubble size 20 %). (E) Same MDS configuration plot with superimposed circles representing prevalence of melanomacrophage centres (max bubble size 70 %). (F) Same MDS configuration plot with superimposed circles representing prevalence of fibrosis (Max bubble size 1.5%).

Transform: Presence/absence Transform: Presence/absence Resemblance: S17 Bray Curtis similarity Resemblance: S17 Bray Curtis similarity

Stress: 0 Stress: 0 Alde Alde A Clyde B Clyde

MerseyForth MerseyForth

Soton Soton

Tyne Tyne

Thames Thames

Transform: Presence/absence Transform: Presence/absence Resemblance: S17 Bray Curtis similarity Resemblance: S17 Bray Curtis similarity

Stress: 0 Stress: 0 C Alde D Alde Clyde Clyde

Mersey Forth MerseyForth

Soton Soton

Tyne Tyne

Thames Thames

Transform: Presence/absence Transform: Presence/absence Resemblance: S17 Bray Curtis similarity Resemblance: S17 Bray Curtis similarity

Stress: 0 Stress: 0 E Alde F Alde Clyde Clyde

Mersey Forth MerseyForth

Soton Soton

Tyne Tyne

Thames Thames

115

C1617/ME3106

Fig.93 MDS configuration of liver lesions in flounder collected in the Mersey estuary between Spring 2000 and Autumn 2003. MDS plot derived from prevalence of 16 lesion categories found in liver of flounder from the estuary.

Transform: Presence/absence Resemblance: S17 Bray Curtis similarity

Stress: 0 AUTUMN2002 AUTUMN2001

AUTUMN2003

AUTUMN2000

SPRING2000

Fig. 94 MDS configuration of liver lesions in flounder collected in the Tyne estuary between Spring 2000 and Autumn 2002. MDS plot derived from prevalence of 13 lesion categories found in liver of flounder from the estuary.

Transform: Presence/absence Resemblance: S17 Bray Curtis similarity

Stress: 0 AUTUMN2002

AUTUMN2001 AUTUMN2000

SPRING2000

116

C1617/ME3106

Fig. 95 MDS configuration of liver lesions in flounder collected in the Alde estuary between Spring 2000 and Autumn 2002. MDS plot derived from prevalence of 13 lesion categories found in liver of flounder from the estuary.

Transform: Presence/absence Resemblance: S17 Bray Curtis similarity Stress: 0 AUTUMN2002

AUTUMN2001

AUTUMN2000 SPRING2000

Fig. 96. MDS configuration of liver lesions in flounder collected from 8 estuarine sites between 2000 and 2003. Fish from relatively contaminated sites appear to group separately from the reference site (Alde – green triangles).

Transform: Presence/absence Resemblance: S17 Bray Curtis similarity 2001 Stress: 0.11 ORIGIN OF SAMPLE Alde Clyde Forth Mersey Soton Tees 2001 2000 Thames Tyne 2001 2000 2001 2002 2000 20002002 2000 2000 2002 2003 2000 20022002 2002

2002 2000

117

C1617/ME3106

Fig. 97. Gill lesions in flounder collected from 7 estuarine sites in Autumn 2002. (A) PCA plot of gill lesions with variable vectors in bottom left corner. (B) MDS configuration plot with clusters of similarity overlaid onto the plot. (C) Same MDS configuration plot with superimposed circles representing prevalence of aneurysms. (D) Same MDS configuration plot with superimposed circles representing prevalence of hyperplasia. (E) Same MDS configuration plot with superimposed circles representing prevalence of ‘no abnormalities’. (F) Same MDS configuration plot with superimposed circles representing prevalence of Trichodina spp. (G) Same MDS configuration plot with superimposed circles representing prevalence of Rickettsia- like organism (RLO)

Transform: Presence/absence Resemblance: S17 Bray Curtis similarity 60 Stress: 0 A B Clyde Clyde

Thames 40

20 Thames

Soton GI-TRI PC2 Tyne TyneMersey

0 Mersey GI-MONGI-RIC Soton GI-NADGI-FUSGI-ANE GI-DIG -20

GI-HYP Forth Forth Alde Alde -40 -80 -60 -40 -20 0 20 40 60 PC1 Transform: Presence/absence Transform: Presence/absence Resemblance: S17 Bray Curtis similarity Resemblance: S17 Bray Curtis similarity Stress: 0 GI-ANE Stress: 0 GI-HYP C Clyde D Clyde 1 8 Thames Thames

4 32

Soton 7 Soton 56 TyneMersey TyneMersey

10 80

Forth Forth Alde Alde

Transform: Presence/absence Transform: Presence/absence Resemblance: S17 Bray Curtis similarity Resemblance: S17 Bray Curtis similarity Stress: 0 GI-NAD Stress: 0 GI-TRI E Clyde F Clyde 6 10 Thames Thames

24 40

Soton 42 Soton 70 TyneMersey TyneMersey

60 100

Forth Forth Alde Alde

Transform: Presence/absence Resemblance: S17 Bray Curtis similarity Stress: 0 GI-RIC G Clyde 0.6 Thames

2.4

Soton 4.2 TyneMersey

Forth Alde

118

C1617/ME3106

98 60

80 Similarity

100 Alde Tyne Forth Soton Mersey Thames

99 Stress: 0.01 Thames

Mersey Forth Alde Tyne

Soton

119

C1617/ME3106

100

120

C1617/ME3106

121

C1617/ME3106

122

C1617/ME3106

Section 6: Discussion and conclusions

The approach taken in the current study has provided new data on the utility and applicability of selected histopathological, molecular and parasitological “biomarkers” to discriminate contaminated and reference sites. In addition, we have been able to detect variations in these biomarkers that occur between years within some of these sites (for fish) and between seasons within a particular site (fish and molluscs). This approach is the first step in selection of a suite of ‘health markers’ that are likely to best reflect actual biological effect of contaminant exposure in the field and furthermore, informing on the effectiveness of measures aimed at ensuring the continued health of British estuaries. Application of this methodology to other data sets of environmental and biological parameters and other biomarkers will provide important insights for the continued use of established biomarkers in biological effects monitoring or for the need for inclusion of new ones. The current study has discriminated reference sites from contaminated sites, has shown that there are variations within sites between years and that there is a continuum of change between sites.

123

C1617/ME3106

6.1 Fish

6.1.1 Recording ‘top-level’ markers of health

Flounder remain an attractive sentinel species for assessing the biological effect of exposure to contaminated environments. Their benthic habit, ubiquity and tendency to express biomarkers of exposure to contaminants have identified this species as a useful model for investigating the effects of anthropogenic pollution in estuarine systems. Studies have centred on the measurement of a range of biological endpoints, including gross disease and histopathology (Köhler, 1990; Vethaak et al., 1996; Vethaak & Jol, 1996; Vethaak & Wester, 1996; Bogovski et al., 1999; Lang et al., 1999; Simpson et al., 2000; Gill et al., 2002; Stentiford et al., 2003, Bateman et al., 2004; Lyons et al., 2004), and a range of biochemical markers indicative of exposure to particular groups of contaminants (Goksøyer et al., 1996; Matthiessen et al., 1998; Allen et al., 1999a and 1999b; Kirby et al., 1999; Lyons et al., 1999; Kirby et al. 2004; Lyons et al., 2004). More recently, genomic and proteomic markers of exposure have been investigated (George et al., 2004; Sheader et al., 2004).

Flounder is one of the commonest flatfish species in UK waters and is the only one known to penetrate well into estuaries and freshwater environments. Its range in the NE Atlantic extends from Norway and the Baltic to Morocco and the Meditteranean, with subspecies in the Adriatic and Black Seas (Galleguillos and Ward, 1982), making it an accessible species for international monitoring programmes. The migratory tendency in this species has been suggested to impact upon its effectiveness as an in site monitor for estuarine health status. However, we believe that the migrational pattern could actually be of benefit to assessment, particularly when combined with information relating to residence time in the estuary (e.g. time since returning from sea) or to offshore status of animals destined to return to a particular estuarine system. This facet of the flounder life history pattern should be exploited in future studies using this species as a sentinel (see below).

There are a number of issues that have arisen following the use of flounder as a sentinel for environmental monitoring of estuarine health status. Most of these issues relate to the provenance of the samples collected and to the natural history traits of individuals at specific sites at specific times of year. Residence time in the estuary is a major factor in determining the true biological effect of exposure to any chemical (including endocrine disruptors) (i.e. was this flounder at this site yesterday and if so, was it here the day before?). Linked to this is the issue of ‘site loyalty’ (i.e. do flounders always return to the same estuary to feed or are they ‘nomadic’?). Such life history features will preclude the interpretation of biological effects data since the researcher will be unsure that the specimens under investigation are a true ‘exposure group’ (i.e. that they have all been exposed to the same contaminants for the same amount of time). In addition to these factors, flounder residing in different estuaries may show different life history features (e.g. compare migration tendencies from shallow estuaries that cool rapidly in winter with deeper, fjordic estuaries that are relatively more constant with

124

C1617/ME3106

surrounding sea temperatures). In such cases, it may not be suitable to directly compare fish captured at different sites since those from certain sites may show almost year-round residence with only short offshore (or outer estuary) migrations to breed, while those from other sites exhibit shorter estuarine residence times (e.g. summer), with more distant migrations to deeper offshore marine sites for breeding (in winter). Since biological effect (e.g. endocrine disruption) is a function of exposure to effector chemicals, knowledge of these life history traits for each study site is vital, particularly if the population level effect of exposure is to be estimated.

Despite these limitations, the flounder has been the subject of fairly extensive study and due to ubiquity, size, relative ease of capture, known biomarker response and categorised pathologies, it remains a suitable candidate for future application to endocrine disruption and related toxicological effects monitoring in the UK.

Other species, such as the viviparous blenny and sand goby also show potential as sentinel species. Viviparous blennies are known to be sensitive to endocrine disrupting chemicals (EDCs) and show induction of VTG and development of ovotestis. In addition, previous studies have demonstrated that this species is also susceptible to a range of toxicopathic hepatic pathologies (Stentiford et al. 2003). In the current programme, extension of this work has applied specific markers for apoptotic change to the gonad of blennies. We have identified this as a potential marker for reproductive effect of contaminants in this (and other species) though the use of this marker in monitoring work requires further validation. From our studies, we have concluded that whilst viviparous blennies may a useful supplementary marker species for specific sites (they are not found on the west coast of the UK), they are unlikely to be abundant enough for comparative studies at multiple sites. However, since this species is the focus of international monitoring efforts (e.g. it is the main sentinel species for environmental monitoring in Sweden), we should continue to record background data for this species from sites where it exists.

Sand gobies are ubiquitous inhabitants of most estuarine and inshore habitats in the UK. Previous work has demonstrated their potential as a sentinel species for monitoring. However, top-level markers (e.g. health, disease, condition, parasites etc) are not so apparent, making clear site-to-site discrimination problematic. Once again, it is suggested that this species is used as a supplementary marker species at sites where it is abundant.

6.1.2 Specific pathologies and parasites

Fibrillar inclusions and megalocytosis have previously been found in association in the livers of flounder captured from contaminated estuaries (Stentiford et al., 2003). The significance of the fibrillar inclusions as indicators of contaminant exposure is not entirely understood. However, enlarged Golgi fields and proliferation of RER have been interpreted as an adaptive sublethal response indicating successful detoxification of contaminants and may be associated with the active synthesis of metallothionein and other proteins for

125

C1617/ME3106

the sequestration and detoxification of heavy metals (see Köhler, 1990). In the current study, we have expanded our understanding of these inclusions. They were only recorded at sites considered to be relatively contaminated and from electron miscroscopy and histology are most are most likely pathologically altered endoplasmic reticulum. It is therefore likely that the pathology is associated with perturbations in cellular metabolism involving protein synthesis. Evidence obtained during this study suggests that there is progression in severity of cellular pathology from the development of fibrillar inclusions, production of an unidentified protein-like material to severe hepatocellular changes where the cytoplasm is filled with this material (previously referred to as 'phospholipoidosis'). The pathology does not appear to be related to the hepatocellular carcinogenic process (i.e. adenoma/carcinoma formation) that has regularly been recorded in flounder from contaminated estuaries but instead is perhaps associated with endocrine disruption and potentially the over stimulation of hepatocytes by estrogenic compounds. This latter theory is strengthened by recent associated work that has demonstrated the presence of vitellogenin within the inclusions and proteinaceous deposits. Specific laboratory studies will demonstrate the specificity and of this pathology and will strengthen the case for fibrillar inclusions to be included as a marker for endocrine related pathology in the flounder.

Megalocytosis has previously been associated with contaminant-induced liver damage in flatfish (Becker et al., 1987; Köhler, 1990). As cellular and nuclear pleomorphism has been associated with other hepatocellular necrotic conditions such as degenerative necrosis, they have been viewed as the initial toxicopathic lesions resulting from exposure to toxic and carcinogenic chemicals (see Myers et al., 1987). It may perhaps be suggested that due to the slow formation of neoplastic lesions (Myers et al., 1987), these more advanced lesions are unlikely to be found in juvenile and young fish (such as those captured in this study). As such, the histopathological detection of lesions such as cellular and nuclear pleomorphism may offer a useful biomarker for contamination where adult fish are unavailable or in low abundance.

While elevated levels of vitellogenin have been reported in the plasma of male flounder captured from this estuary and the intersex condition has been reported in male viviparous blennies (Zoarces viviparous) from the same site (Matthiessen et al., 2000), here we have documented the first case of intersex in flounder from the Clyde. Reproductive effects of industrial chemicals so far detected in fish range from disturbed gonadal maturation and abnormal levels of vitellogenin in the plasma, to intersex gonads (Gimeno et al., 1996). Natural and synthetic estrogens are known to induce sexual reversal in male fish provided that the treatment takes place during the critical period of sexual differentiation (Gimeno et al., 1996). Due to the offshore migration event in flounders, any sexual reversal or disruption in newly hatched larvae is likely to have taken place either at offshore sites, or depending on local circulatory patterns, in coastal waters (see Allen et al., 1999b). This creates somewhat of a paradox with using intersex as an indicator of contamination within an estuarine environment as the presence of this condition may be due to larval

126

C1617/ME3106

rather than adult exposure. However, numerous studies have demonstrated the artificial induction of, and natural presence of, vitellogenin in adult male flat fish (see Sumpter & Joblin, 1995; Pereira et al., 1992; Lye et al., 1997; Allen et al., 1999a, 1999b). It follows that the intersex condition in male flounders may not only be due to an initial exposure to endocrine disrupting chemicals during the critical sexual differentiation period, but also to subsequent exposure to these chemicals in the estuarine environment during adulthood.

When examining the interannual variation that occurs in liver lesions, it is clear that there is a great deal of change from year to year. At first glance, it would seem that such variation would make interpretation of the data difficult. However, by taking the multivariate analysis approach, a broader picture can be observed in the data. For example, the similar overall pattern in the MDS plots of changes in liver lesions in flounder from the Tyne and Alde would suggest that whilst there are site-specific differences (e.g. similarity/dissimilarity in Spring and Autumn 2000 sample), there is an overriding effect that may well be driven by a factor that operates at a larger regional level, such as temperature, atmospheric pressure, etc). The nature of this factor is however currently unknown. It is also clear that there is a greater difference in the spring and autumn samples taken in 2000 from the Tyne when compared to the same two sampling times in the Alde. The main reason for this difference is likely due to the effect of contamination on the liver of flounder in the Tyne compared with the relatively few minor changes that occur in fish that are resident in the reference site (Alde) over the summer. Clearly such interannual variations in any biomarker needs to be considered in any monitoring programme and highlights the need for a consistent approach when sampling to include using target sites and species to obtain samples and the need to carry out any sampling regime during the same season from year to year. Without a consistent approach, interpretation of any data obtained is problematic.

Although there are differences within selected sites from year to year, it is still possible to compare different sites within the same year. The contaminated sites were readily separated from each other and from the reference site using either liver or gill lesions or parasite prevalences (see below). In addition, it is possible to detect a continuum of change in these estuaries on the basis of a range of prevalences in selected pathologies. An example of this is in the prevalence of fibrillar inclusions, which shows a continuum of change from the Clyde to the Thames in the contaminated sites whilst still discriminating the Alde from all of these sites.

Along with the specific pathologies described above, the parasite fauna on and in specific individual fish and within the population as a whole can be considered as top-level indicators of health status. The presence or absence of a particular parasite species within a system is dependent on a range of factors relating to the environment, the host and the parasite itself. As such, it is possible to record the communities of parasite present within a population and, given knowledge of the life history of those parasites (and their hosts), investigate differences that exist between site, season and year. Although

127

C1617/ME3106

some of these variables are clearly not directly related to exposure of host or parasite to contaminants, the presence or absence of parasites at higher or lower prevalence or abundance than predicted may indicate some deviation from ecosystem stability (e.g. absence of a required intermediate host for the parasite, immunosuppression of hosts within the population etc). These measures can be then be employed as an indirect measure of biological effect since they reflect some deviance from the normal scenario that would be predicted at a given site. Finally, since parasites and pathogens can be directly responsible for mortality or sub-lethal effects on growth and reproduction in affected populations, their status in the population of sentinel species chosen for monitoring may give a direct insight into the likely population effect of disease imposed by living in sub-optimal environments.

6.1.3. Synthesis and recommendations

Whilst individual pathologies and parasites are useful as discriminators between sites, the power of multivariate statistics is in the integration of multiple biomarkers into the production of an overall “model” to determine the best suite of biomarkers to use in biological effects monitoring programmes. In addition, such models also provide useful information on the impact of anthropogenic alterations on the ecosystem, allowing the efficacy of implemented changes to be determined. Ultimately it is anticipated that using a similar approach to that taken here, it will be possible to identify those biomarkers that are most able to describe differences between sites, seasons and years and further to determine which specific contaminants or combination of contaminants are influencing the patterns observed.

A suggested approach to this is as follows:

1. Collect multivariate top-level health data on species of interest from site of interest (e.g. pathology, pathogens, condition factors). 2. Collect supporting data (e.g. contaminant burdens, biomarkers etc) from same species/sites/seasons/years. 3. Generate multivariate statistics for health measures described in 1. (above) 4. Generate groupings or types dependent on PCA or MDS plots. 5. Overlay supporting data on these groupings to assess correlative relationships between health and various markers and contaminants. 6. Select health measures and markers that best depict contaminant profiles and best describe overall condition of individual sites. 7. Assign a site/season/year ‘type’ that can be compared to future monitoring efforts at the same site. 8. Record status of ‘type’ over time. Advise on changes and on any intervention strategies required.

128

C1617/ME3106

6.2 Invertebrates and biological effects monitoring

Via the sampling strategy employed within the current programme we have identified three invertebrate species with potential as sentinels for measuring ‘top-level’ health effects in aquatic ecosystems. The three species (mussel, brown shrimp and shore crab) display an array of pathologies and pathogens that may be utilised to discriminate site, season and year-on-year variations.

The relative ease at which a holistic assessment of health can be made using histopathology and the suitability of these species as environmental sentinels provide support for their inclusion as indicators of aquatic environmental health. However, as for fish, there are a number of considerations that must be taken into account if data pertaining to invertebrate health status is to be interpreted unequivocally. These include:

1. Seasonal changes in health, condition and reproductive status are apparent in all of the invertebrate species assessed. These patterns are best depicted in mussels collected during Phase 2 of the mussel- sampling program described in Section 4 of this report. In addition to changes related to contaminant exposure, seasonal changes may also be due to effects of gonadal maturation or pathogen burden. It is important that baseline data is collected from study sites to ensure that normal life history changes within hosts are considered. 2. Linking multivariate features of health in invertebrates will be paramount for interpreting true effect of exposure to contaminants. In such a way, it is important to understand the effect of other stressors (such as pathogens, reproductive stage etc) on biomarker responses since these are likely to change by season. 3. For mussels, we have demonstrated for the first time that gonadal status, condition, pathogens and pathologies within a population is dependent on the species that comprise that population (i.e. M. edulis vs. M.galloprovincialis vs. hybrids). For future biological effects studies using mussels, speciation of individual animals will be required to prevent subjective interpretation of true biological effect of exposure to contaminants. 4. For crustaceans, we have demonstrated that it is possible to discriminate sites based upon condition, pathologies and pathogens. However, to depict true biological effect of exposure to contaminants, similar studies to those carried out under Phase 2 of the mussel sampling program (Section 4) should occur. In such a way, normal seasonal effects of reproduction, moult status and pathogen burden can be eliminated from analyses.

Finally, the current programme has highlighted a number of host-pathogen model systems that may have considerable potential for assessing the effect of disease at the population level. As such, multigenerational studies could be carried out whereby relevant aspects of growth, fecundity, immunity and condition are assessed in the presence and absence of disease. Such a study may be possible using the brown shrimp and CcBV model. In this case, it may

129

C1617/ME3106 be feasible to establish the effect of virus infection on those life history parameters mentioned. Furthermore, by adding factors such as contamination to the experimental system would allow for the host-pathogen-environment model to be assessed more objectively. Findings from such studies would likely have direct relevance to effects observed within the field environment and would move us closer to understanding the effect of stress and pollution on prevalence, pathogenesis and severity of disease.

130

C1617/ME3106

Section 7: Key findings and Forward Look

131

C1617/ME3106

7.1 Key findings

The key findings from the current programme and their relevance to future biological effects monitoring studies are given below.

1. This study has reinforced the use of pathology and parasitology as an essential tool for the overall assessment of health in fish and invertebrates. By combining evidence relating to the parasitology and pathology of multiple organs and tissues, it provides the first step in disentangling anthropogenic from endogenous biological effect. The ‘top-level’ data relating to fish and shellfish health provides the basis for diagnosing a ‘biological effect’ in a given individual, population or ecosystem and should therefore be considered as an ‘endpoint effect’ marker with which to compare ‘exposure’ biomarkers (e.g. DNA adducts, EROD etc). In addition, contaminant data can be correlated with the top-level and exposure biomarkers to predict causality. 2. A relational database has been created (the Aquatic Health Database – AHD) that allows for the potential for integration of endpoint effects with exposure effects and contaminant burdens. The AHD will be used in future programmes to generate outputs applicable to downstream analysis using multivariate statistics applications such as PRIMERTM. It is envisaged that the AHD can be further expanded to incorporate other data types, including environmental, contaminant and biological as well as possessing the ability to be linked to external databases as appropriate. Information contained in the AHD will become available to other researchers via links with the UK wide ‘Marine Data and Information Partnership’ (MDIP) initiative, which is currently under development. 3. Multivariate statistical assessment using endpoint effects data from flounder was used to discriminate populations from relatively contaminated and relatively uncontaminated sites. As stated above, the ability and relevancy in discriminating sites will be enhanced following incorporation of exposure biomarker and contaminant data into the AHD. In this way, exposure and effect can be correlated. The key role of the AHD and PRIMERTM is to allow for discrimination on a site-to-site, season-to-season and year-to-year basis. 4. Flounder from the Mersey, Tyne, Clyde and Forth exhibited the highest prevalences of liver pathology at levels similar to those reported previously. Lowest levels were recorded at the Alde reference site. Liver pathology has proved to be a sensitive indicator of contaminant effect in this species and we propose this as an excellent example of a top-level effect marker in this species. Quality assurance for the measurement of this marker is provided under the EU BEQUALM programme (lead laboratory Cefas Weymouth). 5. Hepatocellular fibrillar inclusions were observed in flounder and were most prevalent in fish from the Mersey and Clyde estuaries. Fish from the Alde estuary did not exhibit these inclusions. The inclusions are most likely pathologically altered endoplasmic reticulum and/or Golgi apparatus and appear to represent an effect of exposure to contaminated environments. Since fibrillar inclusions have been shown

132

C1617/ME3106

to be associated with the production of the egg yolk protein vitellogenin, they may be a useful indicator of exposure to endocrine disrupting chemicals (EDCs). Further work is now required to test this hypothesis in a controlled laboratory setting. 6. The intersex (ovotestis) condition was recorded in flounder from the Clyde and Mersey estuaries and in viviparous blennies from the Tyne estuary. To the best of our knowledge, this is the first histologically confirmed case of intersex in flounder from the Clyde estuary. 7. A previously un-reported pathology of potentially high reproductive significance was that of oogonial and spermatogonial apoptosis in flounder and viviparous blenny. The discovery of this pathology using histology and its subsequent verification using the TUNEL histochemical assay should be followed up in future biological effects studies as a potential marker for reproductive health in these species. 8. Results from the erythrocyte micronucleus (MN) assay using flounder appear to suggest that this assay is not sensitive enough for deployment in assessment of biological effects of contaminant exposure in UK estuaries. 9. The level of DNA adducts in liver tissue of flounder varied between estuarine sites with increased levels at Southampton Water, Clyde, Thames and Mersey estuaries in particular when compared to the Alde. These data are similar to those reported elsewhere. Adduct levels in gonadal tissues of flounder collected from Southampton Water and the Clyde, Thames and Mersey estuaries were not generally elevated when compared to fish from the Alde site. 10. Assessment of a number of statistical measures based upon data collected for parasitology allowed for discrimination of populations of flounder from relatively contaminated sites compared to those caught from the Alde providing evidence that different ambient conditions at these sites acted upon the flounder host and their parasite communities within these sites. The responses of parasite communities to anthropogenic inputs are likely to vary between sites and reflect the impact of contaminants on the parasites directly, on the fish hosts, and on invertebrates harbouring intermediate parasite stages. 11. At relatively contaminated sites, the absence or reduction in numbers of certain parasite species utilising pollution sensitive intermediate hosts provides evidence of an anthropogenic alteration in the estuarine ecosystem impacting on these invertebrate hosts. No statistically significant differences were found in the diversity index between the contaminated Tyne and the reference site. However, the parasite faunas of flounder from the two sites differ. The absence of parasites utilising pollution sensitive intermediate hosts in flounder from the Tyne and the presence of the marine nematode Hysterothylacium aduncum in Tyne flounder shows that these fish are feeding in the wholly marine environment. The ability of these host to change feeding strategies to a marine diet shows that they are likely to move out of the estuary during the summer months and to therefore, potentially negate the impact of contaminants on the host and their parasites. 12. Multivariate statistics (PCA, MDS) have been utilised for the first time to discriminate flounder populations based upon the presence/absence

133

C1617/ME3106

of parasites. The data clearly shows that the reference site (Alde) and the relatively uncontaminated Thames group together. The marine feeding behaviour of flounder in the Tyne and thus the potential for limited environmental impact of contaminants on hosts (and parasites) from this site may explain the close grouping of fish from this site with those from the reference site. Furthermore, fish from the relatively contaminated Forth, Mersey and Clyde sites are easily discriminated from the reference site by this method. 13. While no definite toxicopathic lesions were identified in the crustacean models utilised in this study, most anomalies appeared to be associated with departures from the normal state of immune competence. By this means, parasites and pathogens (such as the virus and yeast pathogens in brown shrimp, the viral and digenean infections in shore crab and the ciliate infestations in mussels are more likely to persist to form disease. It is recommended that further efforts are directed towards the study of immunocompetence in invertebrates captured from these sites and to associate departures from the optima with laboratory based studies using known contaminant levels and a suitable host-pathogen model. 14. The current study has demonstrated how histopathology can be used to generate multivariate condition and disease status data for the blue mussel (Mytilus spp.). An array of pathogens and pathologies were present in animals collected from a range of estuarine sites in the UK (Phase 1). These parameters were used to design a seasonal monitoring survey of mussels from two of the sites (Exe estuary and Southampton Water) (Phase 2). Preliminary data from the Phase 2 study has suggested that it is possible to discriminate mussels from the two sites based upon pathologies and their pathogen burden. 15. Phase 2 of the mussel sampling program has also led to a fundamental discovery regarding multiple speciation at monitoring sites. Populations at several of these sites can be composed of M. edulis, M. galloprovincialis and a hybrid of the two species. Preliminary analysis of mussels collected from individual sites has shown that reproductive status, overall condition, pathologies and pathogen status can all vary significantly between the different species collected at the same sites. In previous studies by other workers, these mussels are likely to have been grouped under the heading of M. edulis. The discovery of differential response in discrete species is a key consideration for future environmental monitoring work using these species. 16. An overarching emphasis of the current program has been the movement towards definition of population and ecosystem-level effects of contaminant exposure in aquatic environments. Throughout the programme, it has become apparent that in order to achieve this complex goal, a number of key features are now required. These include: 1) A better appreciation of baseline values, seasonal alterations and effect of complicating factors on biomarker (endpoint and exposure) expression in sentinel species; 2). A means of recording, integrating, analysing and interpreting complex datasets; 3). A means of classifying an endpoint effect when one is seen (e.g. how do we define a population effect and what will it look like when we see

134

C1617/ME3106

one?); 4). A better system for selecting and testing biomarkers with regard to their usefulness in describing true biological effect of exposure; 5). A better classification scheme for grading effects and for integrating these measures with evolving aquatic classifications schemes (such as Water Framework Directive); and finally, 6). An improved scheme for informing environmental managers of real changes in observed effects over specified spatial and temporal scales.

7.2 Forward look

The current programme has highlighted a requirement for continuation of work in several key areas. These are listed below.

1. Most of the issues related to interpretation of biological effects monitoring studies are associated with the presence of incomplete or fragmented datasets, lack of coherence in chemical and biological measures or an inability to integrate potentially disparate datasets to provide coherent outcomes to the questions of environmental managers. Furthermore, there is a general lack of understanding of the baseline changes that occur in sentinel species over a given season and more importantly, on the effect of these normal changes on the expression of biomarkers of contaminant exposure. As such, site-to- site comparisons may in reality be comparing normal season-to-season changes in species and not, as interpreted, a difference in response to the particular contaminant burden of a site. With this in mind, we suggest a back to basics approach for the key monitoring species (in particular flounder and mussels) whereby basic life history data is utilised to design studies that truly test the differential effect of contaminant exposure between sites. 2. Population of the Aquatic Health Database (AHD) with multivariate data relating to host, disease and environment is seen as an essential goal in describing ecosystem health status. The workflow described in 1. above would lead to this and would provide a baseline dataset against which to compare relatively impacted and non-impacted sites. Such studies are particularly relevant to the requirements of the Water Framework Directive (WFD) where measures of anthropogenic impacts are likely to become more important. 3. Incorporating biological effects data from sentinel species at different positions within the aquatic food chain is now possible. When such data is integrated with chemical analyses of contaminants at these trophic levels, it will be possible to better predict fate and subsequent effect of exposure to contaminants. 4. From the results of the current study, we recommend that an invertebrate laboratory model should be developed to investigate mechanisms of interaction of contaminant exposure with onset of disease. Brown shrimp would appear to be a suitable challenge model (a small invertebrate, common, amenable to culture, simple immune system, environmentally relevant), with the newly discovered CcBV (virus) or yeast pathogens as suitable candidate disease agents. As

135

C1617/ME3106

such, disease-free stock, retained under different environmental conditions could be exposed to the pathogen; the resulting pathogenesis and host defence parameters being measured in a controlled manner. Such a model would allow us to better understand the complex interaction between anthropogenic contamination, the immune system and disease status. 5. Molecular tools are fast becoming mainstream for the diagnosis of biological effect of exposure to contaminants. Furthermore, they provide a critical link to the observed formation of endpoint effect markers (such as cancer). Molecular tools may well be utilised as an early warning system for exposure (e.g. which array of genes/proteins are expressed in early stage pathologies or in juveniles at a particular site?). It is recommended that data arising from these technologies is incorporated into systems such as the AHD to better investigate cause- effect in sentinel species. 6. Future studies on biological effects of exposure to contaminants will need to take account of genetic structure of the target sentinel species. A programme currently funded by Defra MWD (Development of population genetics markers in dab (Limanda limanda) and European Flounder (Platichthys flesus) to assess population structure in impacted and unimpacted areas. ME3206) will collect data of this type from inshore/estuarine (flounder) and offshore (dab) populations from around the UK coastline. Data will be aligned with disease data relating to the same individual fish to examine the potential for genetic components of disease susceptibility.

7.3 Acknowledgements

The authors of this report acknowledge Defra (MWD) for funding this work and the Environment Agency for use of their inshore vessels for sample collection.

136

C1617/ME3106

Section 8: References

Allen, Y., Scott, A.P., Matthiessen, P., Haworth, S., Thain, J.E. and Feist, S. (1999a). Survey of estrogenic activity in United Kingdom estuarine and coastal waters and its effect on gonadal development of the flounder Platichthys flesus. Environmental Toxicology and Chemistry 18, 1791-1800.

Allen, Y., Matthiessen, P., Scott, A.P., Haworth, S., Feist, S., and Thain, J.E. (1999b). The extent of oestrogenic contamination in the UK marine environment – further surveys of flounder. Science of the Total Environment 233, 5-20.

Al-Sabti, K and Metcalfe, C.D. (1995). Fish micronuclei for assessing genotoxicity in water. Mutation Res., 343, 121-135.

Auffret M (1988) Histopathological changes related to chemical contamination in Mytilus edulis from field and experimental conditions. Marine Ecology – Progress Series 46, 101-107.

Bang FB (1980). Monitoring pathological changes as they occur in estuaries and in the ocean in order to measure pollution (with special reference to invertebrates). Rapp.P.- v.Réun.Cons.Int.Explor.Mer. 179, 118-124.

Bayne BL, Brown DA, Harrison F, and Yevich PP, (1980) Mussel Health. In “The International Mussel Watch,” pp. 163-235. National Academy of Sciences, Washington DC.

Becker, D.S., Ginn, T.C., Landolt, M.L. and Powell, D.B. (1987). Hepatic lesions in English sole (Parophrys vetulus) from Commencement bay, Washington (USA). Marine Environmental Research 23, 153-173.

Bogovski S., Sergeyev,B., Muzyka,V., and Karlova,S. (1998). Cytochrome P450 system and heme synthesis enzymes activity in flounder liver as biomarkers of marine environments pollution. Marine Environmental Research 46, 13-16.

Bogovski, S., Lang, T. and Mellergaard, S. (1999). Short communication: histopathological examinations of liver nodules in flounder (Platichthys flesus L.) from the Baltic Sea. ICES Journal of Marine Science 56, 152-156.

Bower, S.M. (2001): Synopsis of Infectious Diseases and Parasites of Commercially Exploited Shellfish: Steinhausia mytilovum (Mussel Egg Disease).

Bucke, D and Feist, S. W. (1993). Histological changes in the livers of dab, Limanda limanda (L.). Journal of Fish Diseases, 16, 281-296.

Burn, P.R. (1980). Pollution effects on fish parasites. Coastal Ocean Pollution Assessment News 1, 3-4

Carman, K.R., Dobbs, F.C. (1997). Epibiotic microorganisms on copepods and other marine crustaceans. Microscopy Research and Technique 37, 116-135.

Carrasco, K.R., Tilbury, K.L., Myers, M.S. (1990). Assessment of the piscine micronucleus test as an in situ biological indicator of chemical contaminant effects. Can. J. Fish. Aquat. Sci., 47, 2123-2136.

Couch JA (1978). Diseases, parasites, and toxic responses of commercial Penaeid shrimps of the Gulf of Mexico and south Atlantic coasts of north America. Fishery Bulletin 76, 1-44.

137

C1617/ME3106

Couch, J.A. (1974). Free and occluded virus, similar to Baculovirus, in hepatopancreas of pink shrimp. Nature 247, 229-231.

Couch, J.A., Courtney, L. (1977). Interaction of chemical pollutants and virus in a crustacean: a novel bioassay system. Annals of the New York Academy of Science 298, 497-504.

Coustau C, Renaud F, Maillard C, Pasteur N, Delay B (1991). Differential susceptibility to a trematode parasite among genotypes of Mytilus edulis/galloprovincialis complex. Genet Res 57: 207-212.

Crothers, J.H. (1968). The biology of the shore crab Carcinus maenas (L.) 2. The life of the adult crab. Field Studies 2, 579-614.

Cuénot, L. (1893). Etudes physiologiques sur les Crustacés Décapodes. Arch. Zool. Exp. Gén., Ser. 4, 3, 1-16.

Dawes, B. (1968). The Trematoda. Cambridge: Cambridge University Press.

De Flora, S., Bagnasco, M. and Zanacchi, P. (1991). Genotoxic, carcinogenic and teratogenic hazards in the marine environment, with special reference to the Mediterranean Sea. Mutation Research 258, 285-320.

Depledge MH and Fossi MC (1994). The role of biomarkers in environmental assessment. (2). Invertebrates. Ecotoxicology 3, 161-172.

DETR (1999). Final report: The Effects of Polycyclic Aromatic Hydrocarbons on Marine Ecosystems. Report number CW0695.

Dixon,T.J., Taggart,J.B., and George,S.G. (2002). Application of real time PCR determination to assess interanimal variabilities in CYP1A induction in European flounder (Platichthys flesus). Marine Environmental Research 54, 1-4.

Doughtie DG and Ranga Rao K (1984). Histopathological and ultrastructural changes in the antennal gland, midgut, hepatopancreas, and gills of grass shrimp following exposure to hexavalent chromium. Journal of Invertebrate Pathology 43, 89-108.

Earnshaw MJ, Wilson S, Akberali HB, Butler RD and Marriott KRM (1986), The action of heavy metals on the gametes of the Marine Mussel, Mytilus edulis (L.)-III. The effect of applied copper and zinc on sperm motility in relation to ultrastructural damage and intracellular metal localisation. Marine Environmental Research, 20: 261-278.

Eggens,M.L., Opperhuizen,A., and Boon,J.P. (1996). Temporal variations iof CYP1A indicies, PCB and 1-OH pyrene concentration in flounder, Platichthys flesus, from the Dutch Wadden Sea. Chemosphere 33, 1579-1596.

Feist, S. W., Jones, G., Longshaw, M., Lyons, B. and Stentiford, G. D. (2001) Pathology biomarkers in estuarine fish species for the assessment of biological effects of contaminants. DETR Final report, March 2001 CDEP 84/5/287, pp.100.

Feist, S.W., Lang, T., Stentiford, G.D., Koehler, A. (2004). Use of liver pathology of the European flatfish dab (Limanda limanda L.) and flounder (Platichthys flesus L.) for monitoring. ICES Techniques in Marine Environmental Sciences 38. 42pp

Fossi MC, Savelli C, Casini S, Franchi E, Mattei N, and Corsi I (1997). Multi-response biomarker approach in the crab Carcinus aestuarii experimentally exposed to benzo (a) pyrene, polychlorobiphenyls and methyl mercury. Biomarkers 2, 311-319.

Gardner GR (1993). Chemically induced histopathology in aquatic invertebrates. Couch JA and Fournie JW. In: Pathobiology of marine and estuarine organisms. pp. 359-392. CRC Press, Boca Raton, Florida, USA.

138

C1617/ME3106

Gimeno, S., Gerritsen, A., Bowmer, T. and Komen, H. (1996). Feminization of male carp. Nature 384, 221-222.

Goksøyr, A., Beyer, J., Egaas, E., Grøsvik, B.E., Hylland, K., Sandvik, M. and Skaare, J.U. (1996). Biomarker responses in flounder (Platichthys flesus) and their use in pollution monitoring. Marine Pollution Bulletin 33, 36-45.

Goldberg ED, Bertine KK (2000) Beyond the Mussel Watch- new directions for monitoring marine pollution. Science of the Total Environment 247, 165-174.

Goldberg ED, Bowen VT, Farrington JW, Harvey G, Martin JH, Parker PL, Riseborough RW, Roberstson W, Schneider E, Gamble E. (1978). The Mussel Watch. Environmental Conservation Vol 5, 2, 101-125.

Harvey, J.S., Lyons, B.P., Waldock, M. and Parry, J.M. (1997). The application of the 32-P postlabelling assay to aquatic biomonitoring. Mutation Research (Fundamental and Molecular Mechanisms of Mutagenesis) 378, 77-88.

Hinton, D.E., & Lauren, O.J. (1990). Liver structural alterations accompanying chronic toxicity in fishes: potential biomarkers of exposure. In: (McCarthy, J.F., & Shugart, L.R. eds) Biomarkers of environmental contamination. Lewis Publishers, MI. pp.12-68.

Hinton, D.E., Baumen, P.C., Gardener, G.C., Hawkins, W.E., Hendricks, J.D., Murchelano, R.A., & Okhiro, M.S. (1992). Histopathological biomarkers. In: (Huggett, R.J., Kimerle, R.A., Mehrle, P.M., & Bergman, H.L. eds) Biomarkers: biochemical, physiological and histological markers of anthropogenic stress. Lewis Publsihers, MI. pp.155-210

Høeg, J. (1995). The biology and life cycle of the Cirripedia Rhizocephala. Journal of the Marine Biological Association, United Kingdom 75, 517-550/

Hopwood D (1996) Theory and practice of histopathological techniques, 4th edn. In: Bamcroft JD, Stevens A (eds) Fixation and fixatives. Churchill Livingstone, Hong Kong. p 23-46

Houghton, G., & Matthews, R.A. (1986). Immunosuppression of carp (Cyprinus carpio L.) to ichthyophthiriasis using the corticosteroid triamcinolone acetonide. Veterinary Immunology and Immunopathology 12, 413-419

Hylland,K., NissenLie,T., Christensen,P.G., and Sandvik,M. (1998). Natural modulation of hepatic metallothionein and cytochrome P4501A in flounder, Platichthys flesus L. Marine Environmental Research 46, 51-55.

IARC (1983). IARC monographs on the evaluation of carcinogenic risk of chemicals to humans. Certain polycyclic aromatic hydrocarbons and heterocyclic compounds. IARC monographs, 3.

ICES (1997). Report of the special meeting on the use of liver pathology of flatfish for monitoring biological effects of contaminants. Weymouth, UK. October 1996.

James, B.L., Sannia, A., Bowers, E.A. (1976). Parasites of birds and shellfish. In: Problems of a small estuary (ed. Nelson-Smith, A. and Bridges, E.M.). Swansea: Institute of Marine Studies and Quadrant Press Ltd.

Janssen,P.A.H., Lambert,J.G.D., Vethaak,A.D., and Goos,H.J. (1997). Environmental pollution caused elevated concentrations of oestradiol and vitellogenin in the female flounder Platichthys flesus. Aquatic Toxicology 39, 195-214.

Johnson PT (1980) Histology of the blue crab, Callinectes sapidus. A model for the Decapoda. Praeger, New York

139

C1617/ME3106

Johnson, P.T. (1983). The Biology of Crustacea, (Bliss DE ed in charge), Vol 6, Pathobiology. In: Provenzano, Jr. A.J. (ed). Diseases caused by viruses, rickettsiae, bacteria, and fungi. Academic Press, New York. p 1-78

Johnson, P.T. (1984). Viral diseases of marine invertebrates. Helgoländer Meeresunters 37, 65-98

Khan, R.A. (1990). Parasitism in marine fish after chronic exposure to petroleum hydrocarbons in the laboratory and to the Exxon Valdez oil spill. Bulletin of the Environmental Contamination and Toxicology 44, 759-763.

Khan, R.A., Barker, D.E.,Williams-Ryan, K. and Hooper, R.G. (1994). Influence of crude oil and pulp and paper mill effluent on mixed infections of Trichodina cottidarim and T. saintjohnsi (Ciliophora) parasitizing Myoxocephalus octodecemspinosus and M. scorpius. Canadian Journal of Zoology 72, 247-251.

Kirby,M.F., Matthiessen,P., Neall,P., Tylor,T., Allchin,C.R., and Kelly,C.A. (1999). Hepatic EROD Activity in Flounder (Platichthys flesus) as an Indicator of Contaminant Exposure in English Estuaries. Marine Pollution Bulletin 38, 676-686.

Kirby,M.F., Neall,P., Bateman,T.A., and Thain,J.E. (2004). Hepatic ethoxyresorufin O- deethylase (EROD) activity in flounder (Platichthys flesus) from contaminant impacted estuaries of the United Kingdom: Continued monitoring 1999-2001. Marine Pollution Bulletin (In Press).

Kleinkauf,A., Macfarlane,C., Yeates,S., Simpson,M.G., and Leah,R.T. (2004). A biomarker approach to endocrine disruption in flounder - estrogen receptors, hepatocyte proliferation, and sperm mobility. Ecotoxicology and Environmental Safety (In Press).

Köhler, A. (1990). Identification of contaminant-induced cellular and subcellular lesions in the liver of flounder (Platichthys flesus L.) caught at differently polluted estuaries. Aquatic Toxicology 16, 271-294.

Köhler, A. (1991). Lysosomal perturbations in fish liver as indicators for toxic effects of environmental pollution. Comparative Biochemistry and Physiology, 100, 123-127.

Köhler, A., Deisemann, H. and Lauritzen, B. (1992). Histological and cytological indices of toxic injury in the liver of the dab Limanda limanda. Diseases of Aquatic Organisms 91, 141- 153.

Kranz, H and Dethlefsen, V. (1990). Liver anomalies in dab Limanda limanda from the southern North Sea with special consideration given to neoplastic lesions. Diseases of Aquatic Organisms 9, 171-185.

Landsberg, J.H., Blakesley, B.A., Reese, R.O., McRae, G. and Forstchen, P.R. (1998). Parasites of fish as indicators of environmental stress. Environmental Monitoring and Assessment 51, 211-232.

Lang, T., Mellergaard, S., Wosniok, W., Kadakas, V. and Neumann, K. (1999). Spatial distribution of grossly visible diseases and parasites in flounder (Platichthys flesus) from the Baltic Sea: a synoptic study. ICES Journal of Marine Science 56, 138-147.

Larson KG, Roberson BS, and Hetrick FM (1989). Effect of environmental pollutants on the chemiluminescence of hemocytes from the American oyster Crassostrea virginica. Diseases of Aquatic Organisms 6, 131-136.

Leonard, D. R. P., Law, R. J. and Kelly, C. A. (1999) Responding to the Sea Empress oil spill. Proceedings of the International Symposium on Marine Pollution. Monaco, 5-9th October, 1998. pp 177-182.

140

C1617/ME3106

Lightner, D.V., Redman, R.M. (1981). A baculovirus-caused disease of the penaeid shrimp, Penaeus monodon. Journal of Invertebrate Pathology 38, 299-302.

Lowe DM, Clarke KR (1989). Contaminant-induced changes in the structure of the digestive epithelium of Mytilus edulis. Aquatic Toxicology 15, 345-358.

Lowe DM, Moore MN (1978). Cytology and quantitative cytochemistry of a proliferative atypical hemocytic condition in Mytilus edulis (Bivalvia, Mollusca). Journal of the National Cancer Institute 60, 1455-1459.

Lowe DM, Moore MN, and Clarke KR (1981). Effects of oil on digestive cells in mussels: quantitative alterations in cellular and lysosomal structure. Aquatic Toxicology 1, 213-226.

Lowe DM, Pipe RK (1987) Mortality and quantitative aspects of storage cell utilization in mussels, Mytilus edulis, following exposure to diesel oil hydrocarbons. Marine Environmental Research, 22: 243-251.

Lowe DM, Pipe, RK (1986). Hydrocarbon exposure in mussels: a quantitative study of the responses in the reproductive and nutrient storage systems. Aquatic Toxicology 8, 265-272.

Lye, C.M., Frid, C.L.J., Gill, M.E., & McCormick, D. (1997). Abnormalities in the reproductive health of flounder Platichthys flesus exposed to effluent from a sewage treatment works. Marine Pollution Bulletin, 34, 34-41.

Lyons, B.P., Harvey J.S. and Parry J.M. (1997). The initial assessment of the genotoxic impact of the Sea Empress oil spill by the measurement of DNA adduct levels in the intertidal teleost Lipophrys pholis. Mutation Research 390, 263-268.

Lyons, B.P., Bignell, J.P., Stentiford, G.D., Feist, S.W. (2004a). The viviparous blenny (Zoarces viviparus) as a bioindicator of contaminant exposure: application of biomarkers of apoptosis and DNA damage. Marine Environmental Research 58, 757-761

Lyons, B.P., Stentiford, G.D., Green, M., Bignell, J., Bateman, K., Feist, S.W., Goodsir, F., Reynolds, W.J., Thain, J.E. (2004b). DNA adduct analysis and histopathological biomarkers in European flounder (Platichthys flesus) sampled from UK estuaries. Mutation Research 552, 177-186.

Lyons, B.P., Stewart, C., & Kirby, M.F. (1999). The detection of biomarkers of genotoxin exposure in the European flounder (Platichthys flesus) collected from the River Tyne Estuary. Mutation Research, 446, 111-119.

MacKenzie, K. (1985). The use of parasites in population studies of herring, Clupea haengus L., in the North Sea and to the north and west of Scotland. J. Cons. perm. Int. Explor. Mer. 42, 33-64

MacKenzie, K. (1987). Long-term changes in the prevalence of two helminth parasites (Cestoda: Trypanorhyncha) infecting marine fish. Journal of Fish Biology. 31, 83-88

MacKenzie, K. (1999). Parasites as pollution indicators in marine ecosystems: a proposed early warning system. Marine Pollution Bulletin. 38, 955-959

MacKenzie, K., Williams, H.H., Williams, B., McVicar, A.H. and Siddall, R. (1995). Parasites as indicators of water quality and the potential use of helminth transmission in marine pollution studies. Advances in Parasitology 35, 85-144.

Malins, D.C., Krahn, M.M., Brown, D.W., Rhodes, L.D., Myers, M.S., McCain, B.B and Chan, S-L. (1985a). Toxic chemicals in marine sediment and biota from Mukilteo, Washington: Relationships with hepatic neoplasms and other hepatic lesions in English sole (Parophrys vetulus). Journal of the National Cancer Research Institute 74, 487-494.

141

C1617/ME3106

Malins, D.C., Krahn, M.M., Myers, M.S., Rhodes, L.D., Brown, D.W., McCain, B.B and Chan, S-L. (1985b). Toxic chemicals in sediments and biota from a creosote-polluted harbor: relationships with hepatic neoplasms and other hepatic lesions in English sole (Parophrys vetulus). Carcinogenesis 6, 1463-1469.

Martorelli, S.R., Schuldt, M. (1990). Encapsulación de dos metacercarias (Digenea: Microphallidae) en Cyrtograpsus angulatus y Palaemonetes argentinus (Crustacea: Decapoda). Rev Biol Trop 38: 295-304.

Matthiessen, P. and Law, R. Contaminants and their effects on estuarine and coastal organisms in the United Kingdom in the late twentieth century. (in preparation).

Matthiessen, P., Allen, Y., Bignell, J., Craft, J., Feist, S., Jones, G., Katsiadaki, I., Kirby, M., Robertson, F., Scott, A., Stewart, C., & Thain, J. (2000). Studies of endocrine disruption in marine fish – progress with the EDMAR programme. ICES CM 2000/S: 08, International Council for the Exploration of the Sea, Copenhagen, 20 pp.

Matthiessen, P., Bifield, S., Jarrett, F., Kirby, M. F., Law, R. J., McMinn, W. R., Sheahan, D. A., Thain, J. E. and Whale, G. F. (1998a). An assessment of sediment toxicity in the River Tyne estuary, UK by means of bioassays. Marine Environmental Research 45, 1-15.

Moore MN, Livingstone DR, Widdows J, Lowe DM, Pipe RK (1987). Molecular, cellular and physiological effects of oil-derived hydrocarbons on molluscs and their use in impact assessment. Phil. Trans. R. Soc. Lond. B 316, 603-623.

Moore, M.J. and Myers, M.S. (1994). Pathobiology of chemical-associated neoplasia in fish. Aquatic Toxicology 24, 327-386.

Moore, M.N., Simpson, M.G. (1992). Molecular and cellular pathology in environmental impact assessment. Aquatic Toxicology 22, 313-322.

Moore, M. N., Readman, J., Depledge, M. J. and Leonard, D. R. P. (2004) An integrated biomarker-based strategy for ecotoxicological evaluation of risk in environmental management. Mutation Research, 552, 247-268.

Mortimer MR and Hughes JM (1991). Effects of organophosphate pollution on genetic structure in two species of estuarine crabs. Marine Pollution Bulletin 22, 353-359.

Myers, M.S., Johnson, L.L., Hom, T., Collier, T.K., Stein, J.E & Varanasi, U. (1998). Toxicopathic lesions in subadult English sole (Pleuronectes vetulus) from Puget Sound, Washington, USA: Relationships with other biomarkers of contaminant exposure. Marine Environmental Research 45, 47-67.

Myers, M.S., Landahl, J.T., Krahn, M.M and McCain, B.B. (1991). Relationships between hepatic neoplasms and related lesions and exposure to toxic chemicals in marine fish from the U.S. West Coast. Environmental Health Perspectives 90, 17-26.

Myers, M.S., Landahl, J.T., Krahn, M.M., Johnson, L.L and McCain, B.B. (1990). Overview of studies on liver carcinogenesis in English sole from Puget Sound; evidence for a xenobiotic chemical aetiology, 1, Pathology and epizootiology. Science of the Total Environment 94, 33- 50.

Myers, M.S., Olson, O.P., Johnson, L.L., Stehr, C.M., Hom, T. and Varanasi, U. (1992). Hepatic lesions other than neoplasms in subadult flatfish from Puget Sound, WA: Relationships with indices of contaminant exposure. Marine Environmental Research 34, 45- 51.

Myers, M.S., Rhodes, L.D. and McCain, B.B. (1987). Pathologic anatomy and patterns of occurrence of hepatic neoplasms, putative preneoplastic lesions and other idiopathic hepatic

142

C1617/ME3106 lesions in English sole (Parophrys vetulus) from Puget Sound, Washington, USA. Journal of the National cancer Institute, 78(2), 333-363.

Myers, M.S., Stehr, C.M., Olson, O.P., Johnson, L.L., McCain, B.B., Chan, S-L., & Varanasi, U. (1994). Relationships between toxicopathic hepatic lesions and exposure to chemical contaminants in English sole (Pleuronectes vetulus), starry flounder (Platichthys stellatus), and white croaker (Genyonemus lineatus) from selected marine sites on the pacific coast, USA. Environmental Health Perspectives, 102, 200-215.

Nicholson, M. D. and Leonard, D. R. P. (1999) The impact of the Sea Empress oil spill on fish and shellfish. Proceedings of the Sea Empress Conference. 11-13 February, 1998. Cardiff. CIWEM.

Overstreet RM (1988). Aquatic pollution problems, Southeastern U.S. coasts: histopathological indicators. Aquatic Toxicology 11, 213-239.

Periera, J.J., Ziskowski, J., Mercaldo-Allen, R., Kuropat, C., Luedke, D. and Gould, E. (1992). Vitellogenin in winter flounder (Pleuronectes americanus) from Long Island Sound and Boston Harbor. Estuaries 15, 289-297.

Petrulis,J.R. and Bunce,N.J. (1999). Competitive inhibition by inducer as a confounding factor in the use of the ethoxyresorufin-O-deethylase (EROD) assay to estimate exposure to dioxin- like compounds. Toxicology Letters 105, 251-260.

Pipe RK (1987). Ultrastructural and cytochemical study on interactions between nutrient storage cells and gametogenesis in the mussel Mytilus edulis. Marine Biology 96, 519-528.

Rasmussen LPD, Hage E, and Karlog O, (1983) Light and electron microscopic studies of the Acute Toxic and chronic toxic effects of N-Nitroso compounds on the marine mussel, Mytilus edulis (L). I. N-nitrosodimethylamine dimethylamine. Aquatic Toxicology, 3 (1983) 285-299.

Reinhard, E.G. (1956). Parasitic castration of Crustacea. Parasitology 5, 79-107

Robaldo, R.B., Monserrat, J., Cousin, J.C.B., Bianchini, A. (1999). Effects of metacercariae (Digenea: Microphallidae) on the hepatopancreas of Chasmagnathus granulata (Decapoda: Grapsidae). Diseases of Aquatic Organisms 37, 153-157.

Ross, P.S., De Swart, R.L., Van Loveren, H., Osterhaus, A.D.M.E., & Vos, J.G. (1996). The immunotoxicity of environmental contaminants to marine wildlife: a review. Annual Review of Fish Diseases 6, 151-165

Sarojini R, Reddy PS, Nagabhushanam R, and Fingerman M (1993). Napthalene-induced cytotoxicity on the hepatopancreatic cells of the red swamp crayfish, Procambarus clarkii. Bulletin of Environmental Contamination and Toxicology 51, 689-695.

Saville, D.H., Irwin, S.W.B. (1991). In ovo cultivation of Microphallus primas (Trematoda: Microphallidae) metacercariae to ovigerous adults and the establishment of the life-cycle in the laboratory. Parasitology 103, 479-484.

Seed R (1969). The incidence of pea crab, Pinnotheres pisum in two types of Mytilus (Mollusca: Bivalia) from Padstow, south west England. J Zool 158: 413-420.

Seed R (1976) Ecology, In: Marine Mussels, their ecology and physiology (Bayne BL), International Biological Programme, Cambridge University Press, pp 19-31.

Sheader, D.L., Gensberg, K., Lyons, B.P., Chipman, K. (2004). Isolation of differentially expressed genes in contaminant exposed European flounder (Platichthys flesus) by suppressive subtractive hybridisation. Marine Environmental Research 58, 553-558.

143

C1617/ME3106

Simpson, M.G., Parry, M., Kleinkauf, A., Swarbreck, D., Walker, P. and Leah, R.T. (2000). Pathology of the liver, kidney and gonad of flounder (Platichthys flesus) from a UK estuary impacted by endocrine disrupting chemicals. Marine Environmental Research 50, 283-287.

Sindermann CJ (1979). Pollution-associated diseases and abnormalities of fish and shellfish: a review. Fishery Bulletin 76, 717-749.

Smith, V.J., Swindlehurst, R.J., Johnston, P.A., Vethaak, A.D. (1995). Disturbance of host defence capability in the common shrimp, Crangon crangon, by exposure to harbour dredge spoils. Aquatic Toxicology 32, 43-58.

Snyder MJ (2000). Cytochrome p450 enzymes in aquatic invertebrates: recent advances and future directions. Aquatic Toxicology 48, 529-547.

Soegianto, A., Charmantier-Daures, M., Trilles, J.P., Charmantier, G., 1999a. Impact of copper on the structure of gills and epipodites of the shrimp Penaeus japonicus. J. Crustac. Biol. 19, 209-223.

Soegianto, A., Charmantier-Daures, M., Trilles, J.P., Charmantier, G., 1999b. Impact of cadmium on the structure of gills and epipodites of the shrimp Penaeus japonicus (Crustacea: Decapoda). Aquat. Living Res. 12, 57-70.

Sparks AK (1980) Multiple granulomas in the midgut of the Dungeness crab, Cancer magister. J Invertebr Pathol 35: 323-324.

Stein, J.E., Collier, T.K., Reichert, W.L., Casillas, E., Hom, T & Varanasi, U. (1992). Bioindicators of contaminant exposure and sublethal effects – studies with benthic fish in Puget Sound, Washington. Environmental Toxicology and Chemistry 11, 701-714.

Stein, J.E., Reichert, W.L.., Nishimito, M. and Varanasi, U. (1990). Overview of studies on liver carcinogenesis in English sole from Puget Sound; evidence for a xenobiotic chemical etiology, 2, Biochemical studies. Science of the Total environment 94, 51-69.

Stentiford, G.D., Evans, M., Bateman, K., Feist, S.W. (2003). Co-infection by a yeast-like organism in Hematodinium-infected European edible crabs Cancer pagurus and velvet swimming crabs Necora puber from the English Channel. Diseases of Aquatic Organisms 54, 195-202.

Stentiford, GD, Longshaw, M, Lyons BP, Jones G, Green, M, Feist, SW (2002). Histopathological biomarkers in estuarine fish species for the assessment of the biological effects of contaminants. Marine Environmental Research (in press).

Sumpter, J.P. and Joblin, S. (1995). Vitellogenesis as a biomarker for estrogenic contamination of the aquatic environment. Environmental Health Perspectives 103, 173-178.

Sunila I (1984) Copper- and cadmium-induced histological changes in the mantle of Mytilus edulis L. (Bivalvia). Limnologica (Berlin) 15 (2): 523-527.

Sunila I (1986) Histopathological changes in the mussel Mytilus edulis L. at the outlet from a titanium dioxide plant in Northern Baltic. Ann Zool Fennici 23, 61-70.

Svärdh L (1999) Bacteria, Granulocytomas, and Trematode Metacercariae in the digestive gland of Mytilus edulis: Seasonal and Interpopulation Variation. Journal of Invertebrate Pathology 74, 275-280.

Svärdh L, Johannesson K, (2002) Incidence of haemocytes and parasites in coastal populations of blue mussels (Mytilus edulis)- testing correlations with area, season, and distance to industrial plants. Journal of Invertebrate Pathology 80, 22-28.

144

C1617/ME3106

Thörnqvist P-O, Söderhäll K (1997). Diseases in Asian aquaculture III. In: Flegel TW, MacRae IH (eds). Crustacean immune reactions, a short review. Fish Health Section, Asian Fisheries Society, Manila. p 203-218 van der Oost R (1997). Biomonitoring aquatic pollution with feral eel (Anguilla anguilla). III. Statistical analyses of relationships between contaminant exposure and biomarkers. Aquatic Toxicology 39, 45-75.

Varanasi, U, Stein, J.E., Nishimoto, M., Reichert, W.L and Collier, T.K. (1987). Chemical carcinogenesis in feral fish: uptake, activation, and detoxication of organic xenobiotics. Environmental Health Perspectives 71, 155-170.

Veillet, A., and Graf, F. (1958). Dégénérescence de la glande androgéne des Crustacés décapodes parasités par les Rhizocéphales. Bulletin mensuel de la Sociétés des Sciences de Nancy 18, 123-127.

Vethaak, A.D and Wester, P.W. (1996). Diseases of flounder Platichthys flesus in Dutch coastal and estuarine waters, with particular reference to environmental stress factors. II. Liver histopathology. Diseases of Aquatic Organisms 26, 99-116.

Vethaak, A.D. and Jol, J.G. (1996). Diseases of flounder Platichthys flesus in Dutch coastal and estuarine waters, with particular reference to environmental stress factors. I. Epizootiology of gross lesions. Diseases of Aquatic Organisms 26, 81-97.

Vethaak, A.D., Jol, J.G., Meijboom, A., Eggens, M.L., Rheinhalt, T., Wester, P.W., Van de Zande, T., Bergman, A., Dankers, N., Ariese, F., Baan, R.A., Everts, J.M., Opperhuizen, A., & Marquenie, J.M. (1996). Skin and liver diseases induced in flounder (Platichthys flesus) after long-term exposure to contaminated sediments in large-scale mesocosms. Environmental Health Perspectives, 104, 1218-1229.

Viarengo, A. (1993). Mussels as bioindicators in marine monitoring programs. In: Proceedings of the Symposium of the Mediterranean Seas, Santa Margherita Ligure, 23-27 September 1991 (ed. Della Croce, N.F.R.), pp. 161-171

Victor, B., 1993. Responses of hemocytes and gill tissues to sublethal cadmium chloride poisoning in the crab Paratelphusa hydrodromous (Herbst). Arch. Environ. Contam. Toxicol. 24, 432-439.

Victor, B., 1994. Gill tissue pathogenicity and hemocyte behavior in the crab Paratelphusa hydrodromous exposed to lead chloride. J. Environ. Sci. Health A29, 1011-1034.

Walker, G. (2001). Introduction to the Rhizocephala (Crustacea: Cirrepedia). Journal of Morphology 249, 1-8.

Waring, C.P., Stagg, R.M., Fretwell, K., McLay, H.A. and Costello, M.J. (1996). The impact of sewage sludge exposure on the reproduction of the sand goby, Pomatoschistus minutus. Environmental Pollution 93, 17-25.

Wedderburn J, McFadzen I, Sanger RC, Beesley A, Heath C, Hornsby M and Lowe D (2000). The field application of cellular and physiological biomarkers, in the Mussel, Mytilus edulis, in conjunction with early life stage bioassays and adult Histopathology. Marine Pollution Bulletin Vol. 40, No. 3 pp 257-267.

Wedderburn, J., McFadzen, I., Sanger, R.C., Beesley, A., Heath, C., Hornsby, M., Lowe, D. (2000). The field application of cellular and physiological biomarkers, in the mussel Mytilus edulis, in conjunction with early life stage bioassays and adult histopathology. Marine Pollution Bulletin, 40, 257-267.

145

C1617/ME3106

Werner, M. (2001). Prevalence of the parasite Sacculina carcini Thompson 1836 (Crustacea, Rhizocephala) on its host crab Carcinus maenas (L.) on the west coast of Sweden. Ophelia 55, 101-110.

Wester, P.W., & Canton, J.H. (1991). The usefulness of histopathology in aquatic toxicity studies. Comparative Biochemistry and Physiology, 100 C, 115-117.

Willett,K.L., Wassenberg,D., Lienesch,L., Reichert,W., and Di Giulio,R.T. (2001). In vivo and in vitro inhibition of CYP1A-dependent activity in Fundulus heteroclitus by the polynuclear aromatic hydrocarbon fluoranthene. Toxicology and Applied Pharmacology 177, 264-271.

Williams T.D., Gensberg K., Minchin S.D., Chipman J.K. (2003) A DNA expression array to detect toxic stress response in European flounder (Platichthys flesus). Aquatic Toxicol. 65, 141-157.

Williams, H.H., MacKenzie, K. and McCarthy, A.M. (1992). Parasites as biological of the population biology, migrations, diet, and phylogenetics. Reviews in Fish Biology and Fisheries 2, 144-176

Yeomans, W.E., Chubb, J.C. and Sweeting, R.A. (1997). Use of protozoan communities for pollution monitoring. Parasitologia 39, 201-212.

Zander, C.D. (1991). Accumulation of helminth parasites in Zoarces viviparus (L.) Teleostei of SW Baltic. Seevogel 12, 70-73

Zander, C.D. (1993). The biological indication of parasite life-cycles and communities from the Lubeck Bight, SW Baltic Sea. Zeitschrift fur Angewandte Zoologie 79, 377-389

Zander, C.D. (1998). Ecology of host parasite relationships in the Baltic Sea. Naturwissenschaften. 85, 426-4

146