Bioaccumulation surveillance in Milford Haven Waterway

Abstract

Biomonitoring of contaminants (metals, organotins, polyaromatic hydrocarbons (PAHs), PCBs) was undertaken in Milford Haven Waterway (MHW) and a reference site in the Tywi Estuary (St Ishmael/Ferryside) during 2007-2008. Bioindicator species encompassed various uptake routes-Fucus vesiculosus (dissolved contaminants); Littorina littorea (grazer); Mytilus edulis and Cerastoderma edule (suspension feeders); and Hediste (=Nereis) diversicolor (sediments). Differences in feeding and habitat preference have subtle implications for bioaccumulation trends though, with few exceptions, contaminant burdens in MHW were higher than the Tywi reference site, reflecting inputs. Elevated metal concentrations were observed at some MHW sites, whilst As and Se (molluscs and seaweed) were consistently at the higher end of the UK range. However, for most metals, distributions in MH biota were not exceptional. Several metal-species combinations indicated increases in bioavailability upstream, which may reflect the influence of geogenic/land-based sources-perhaps enhanced by lower salinity. TBT levels in MH mussels were below OSPAR toxicity thresholds and in the Tywi were close to zero. Phenyltins were not accumulated appreciably in M. edulis, whereas some H. diversicolor populations appear subjected to localized (historical) sources. PAHs in H. diversicolor were distributed evenly across most of MHW, although acenaphthene, fluoranthene, pyrene, benzo(a)anthracene and chrysene were highest at one site near the mouth; naphthalenes in H. diversicolor were enriched in the mid-upper Haven (a pattern seen in M. edulis for most PAHs). Whilst PAH (and PCB) concentrations in MH mussels were mostly above reference and OSPAR backgrounds, they are unlikely to exceed ecotoxicological thresholds. Bivalve Condition indices (CI) were highest at the Tywi reference site and at the seaward end of MH, decreasing upstream-giving rise to several significant (negative) relationships between CI and body burdens. Despite the possible influence of salinity gradient as a complicating factor, multivariate analysis indicated that a combination of contaminants could influence the pattern in condition (and the biomarkers metallothionein and TOSC). Integrating bioaccumulation data with biological and biochemical endpoints is seen as a useful way to discriminate environmental quality of moderately contaminated areas such as MHW and to prioritise cause and effect investigations.

Full text

Bioaccumulation in Milford Haven Waterway
W J Langstona*, N D Popea , S O’Haraa, M Imamuraa, H Harinob, Kim, A.Wc. and
C H Vanec
aMarine Biological Association, Citadel Hill, Plymouth PL1 2PB
bSchool of Human Sciences, Kobe College, 4-1 Okadayama, Nishinomiya, Hyogo 662-8505, Japan
cBritish Geological Survey, Kingsley Dunham Centre, Keyworth, NG12 5GG
Keywords: Milford Haven, bioaccumulation, metals, PAHs, PCBs, organotins
Abstract: Biomonitoring of contaminants (metals, organotins, PAHs, PCBs) was
carried out along the Milford Haven Waterway (MHW) and at a reference site in the
Tywi Estuary during 2007-2008. The species used as bioindicators encompass a
variety of uptake routes - Fucus vesiculosus (dissolved contaminants); Littorina
littorea (grazer); Mytilus edulis and Cerastoderma edule (suspension feeders); and
Nereis diversicolor (omnivore which often reflects contaminants in sediment).
Differences in feeding strategy and habitat preference have subtle implications for
bioaccumulation trends though, with few exceptions, contaminant body burdens in
Milford Haven (MH) were higher than those at the Tywi reference site, reflecting
inputs.
Elevated concentrations of metals were occasionally observed at individual MH sites,
whilst As and Se (molluscs and seaweed) were, for much of MHW, consistently at the
higher end of the UK range. However, for the majority of metals, distributions in MH
biota were not exceptional by UK standards. Several metal-species combinations
indicated increases in bioavailability at upstream sites, which may reflect the
influence of geogenic or other land-based sources – perhaps enhanced by lower
salinity (greater proportions of more bioavailable forms).
TBT levels in MH mussels were below OSPAR toxicity thresholds and in the Tywi
were close to zero. Phenyltins were not accumulated appreciably in Mytilus, whereas
some Nereis populations may have been subjected to localized (historical) sources.
PAHs in Nereis tended to be evenly distributed across most sites, but with somewhat
higher values at Dale for acenaphthene, fluoranthene, pyrene, benzo(a)anthracene and
chrysene; naphthalenes tended to be enriched further upstream in the mid-upper
Haven (a pattern seen in mussels for most PAHs). Whilst concentrations in MH
mussels were mostly above reference site and OSPAR backgrounds, it is unlikely that
ecotoxicological guidelines would be exceeded.
PCBs in mussels were between upper and lower OSPAR guidelines and were unusual
in their distribution in that highest levels occurred at the mouth of MH.
Condition indices (CI) of bivalves (mussels and cockles) were highest at the Tywi
reference site and at the seaward end of MH, decreasing upstream along the
Waterway. There were a number of significant (negative) relationships between CI
and body burdens and multivariate analysis indicated that a combination of
contaminants could influence the pattern in condition (and sub-lethal responses such
as MT and TOSC) across sites. Cause and effect needs to be tested more rigorously in
future assessments.
*Corresponding author: wjl@mba.ac.uk
1

1 Introduction
The Milford Haven Waterway is a ria-type estuary, unique in Wales, forming part of
the Pembrokeshire Marine Special Area of Conservation (Burton, 2006). The Haven
is fully marine for some 12km from the mouth (to the confluence of the Pennar River)
with a shoreline of >100km. The Daucleddau -the common Estuary of E&W Cleddau
Rivers (Figure 1) - is also marine (mesohaline) for much of its length because of the
small ingress of FW relative to the tidal incursion (Nelson-Smith, 1965).
Despite its important conservation status, the Waterway is subjected to contaminants
from several sources including E & W Cleddau Rivers, industry (e.g. oil refineries),
waste water discharge (WWTW), diffuse inputs associated with landfill leachate,
urban development, agricultural run-off and atmospheric deposition. Maritime
operations and pollution incidents (hydrocarbons and antifouling), dredging and spoil
disposal add to this inventory (Atkins, 2005). The harbour was host to >50M gross
tonnes shipping in 2009, supplying, among other cargoes, 25-30% of the UK's petrol,
diesel and, recently, liquefied natural gas. Contamination by biologically-deleterious
substances has implications for the ‘favourable condition status’ of the site and other
conservation objectives. Potential impact on fauna, at its most extreme, was
highlighted by the ‘Sea Empress’ oil spill in 1996.
Milford Haven has been a major oil terminal since the 1960s and has received chronic
releases of hydrocarbons (<250 tonnes pa, overall) by a variety of routes including
refineries, power stations, shipping, road run-off and small-scale domestic sources.
These inputs are dispersed along the waterway in association with suspended
particulates (Little et al., 1987; Nikitik and Robinson, 2003). The loss of >70000
tonnes of light crude oil and 480 tonnes of heavy fuel oil from the Sea Empress in
1996 represented a significant departure from this trend (though <15000t are
estimated to have reached the shoreline, and only a fraction of this entered the Haven,
2

with very little reaching the Cleddau Estuaries). Nevertheless, hydrocarbon residues
were dispersed along considerable stretches of the MH waterway, with heaviest
contamination of sediments reported in the lower reaches (Nikitik and Robinson,
2003). Prior to this (and subsequently) hydrocarbon levels in fine sediments generally
displayed a gradient decreasing from the upper estuary, seaward, and were more
indicative of pyrogenic and degraded chronic input than crude oil. The higher values
upstream are presumed to reflect the up-estuary transport of fines in this flood-tide
dominated system.
Sediment PAH ‘hotspots’>100,000 µg kg-1 at MH, reported by Woodhead et al.,
(1999), were above the Threshold for Effects Level (TEL) for ∑PAHs of 1684 µg kg-1
(dry weight), and also above the Probable Effects Level (PEL) of 16770 µg kg-1 (dw)
implying ecotoxicological significance. However up until the time of the accident
there was little monitoring of biological-effects baselines, particularly sub-lethal
‘markers’ which may be most relevant in quantifying the ‘health’ status of
populations.
Reports of oil-related biological impacts from the Sea Empress imply varying degrees
of deleterious effects, though in most cases recovery appears to have been complete
within five years. Shellfish including limpets (Patella spp), winkles (Littorina spp),
razor shells (Ensis ensis), trough shells (Mactra corallina) and cockles Cerastoderma
edule were subjected to mass mortalities in oiled areas though mussels Mytilus edulis
were relatively tolerant to oiling and were only acutely affected in areas sprayed with
detergent. C. edule populations were the only commercial shellfish to suffer major
mortalities but recovery appears to have been rapid, albeit fluctuating in scale
(Rostron 1998). Our own observations in 2007/8 indicate fairly abundant populations
of mussels and cockles (and other bioindicator species) at each of the selected sites in
MH waterway, upstream as far as Black Tar (figure 1). A subsequent survey in 2010
3

has shown that distributions of M.edulis and C.edule extend even further upstream - to
the confluence of E. and W. Cleddau. Subjective indications are that cockles may be
less common than in 2008, however, coinciding with reduced numbers described
elsewhere in the region, particularly the Loughor Estuary, S Wales.
Following the spill, biomarkers of genotoxicity revealed responses in some fish
species - though not in mussels (Law et al., 1998; Harvey et al., 1999).
Immunosuppression in mussels was, however, severe immediately after the accident,
but recovery followed a few months later (Dyrynda et al., 2000). Reduction of
phagocytosis and intra-cellular superoxide production (both defence mechanisms for
killing infectious agents) were correlated with mussel body burdens of oil-derived
and, in particular, combustion-derived PAHs; extra-cellular superoxide production
was more strongly related to oil-derived PAHs.
Reduced scope for growth (SFG)- a physiological stress response in mussels - was
recorded in samples from Milford Haven in 1996 and 1997, much of which was
thought attributable to PAHs and, to a lesser extent, organotins, though the authors
suggest that other ‘unknown’ compounds may have also contributed to effects
(Widdows et al., 2002). Of all Irish Sea sites sampled, the concentrations of 2- and 3-
ring PAHs were highest in mussels from MH, sampled 6 months after the Sea
Empress oil spill (>20000 µg kg-1 dw). By the following year (1997) concentrations
had declined by ~ 66%, reflecting recovery after the oil-spill, but were still relatively
high (4000 to 7800 μg kg−1 dry wt.) – more than an order of magnitude above those in
the current survey.
Bioassays using sediments from some sites in MHW induced a reduction in feeding
activity in lugworm Arenicola marina though acutely toxic effects were not
demonstrated in these assays (Law et al., 1998). At the community level, there were
reduced numbers of sensitive amphipods following the spill, accompanied by
4

increases in opportunist polychaetes (e.g. Capitella) in the Middle and Lower Haven.
Diversity has recovered to varying degrees in different parts of the Waterway by
2000, even though abundances were generally still low (Nikitik and Robinson, 2003).
The importance of MH port is likely to increase in coming years, and will see a rise in
contaminant-related pressures: hence the value of establishing a monitoring
programme - to ensure unacceptable deterioration does not occur as a result of
anthropogenic events. Bioaccumulation surveillance will help statutory agencies
(Milford Haven Port Authority, Countryside Council for Wales, the Environment
Agency), and industry, in their environmental protection responsibilities.
Biomonitoring is valuable addition to the tools available to assess the adverse effects
of contamination (water quality analysis, toxicity testing and ecological survey)
because it provides a direct measure of the bioavailability of contaminants.
Bioaccumulation is not only an important component of environmental quality
assessment but also, for commercial species, can have implications for human health.
Shellfish gathering in MHW is currently mainly small-scale - for mussels, winkles,
cockles, clams, oysters and razor fish. Limited seaweed harvesting occurs, primarily
for the making of laver bread.
It may seem most relevant to base the choice of biomonitoring organism on one or
more species consumed by humans. However, there are other considerations which
stem from the fact that different contaminants have their own characteristics and that
organisms accumulate them from a variety of sources, often at different rates,
adopting diverse accumulation strategies. Consequently, there is no universal
indicator organism and the most useful monitoring programmes are likely to include
analyses of several species, preferably of differing ecological types (primary
producer, detritivore, herbivore, filter feeder). Selection of indicators should be
appropriate to the chemistry and form (dissolved, particulate, dietary) of the
5

contaminants of concern. Hence, by integrating results for several different species a
broad appraisal of impact is obtained. The selection of species in the current project
represents an appropriate blend i.e. Fucus vesiculosus (dissolved contaminants);
Littorina littorea (grazer); Mytilus edulis and Cerastoderma edule (suspension feeders
that accumulate from both dissolved phase and suspended particulates); and Nereis
diversicolor (an omnivore which often strongly reflects bioavailable contaminants in
sediment e.g. Bryan et al, 1985; Langston and Spence, 1995).
By combining the information gained from these species a reasonable picture of the
significance of biologically-available contaminants in MHW will be achieved. A
similar rationale was adopted for earlier EA (NRA) bioaccumulation surveys in Wales
(Davies and Ellery, 1995), which, though limited, represent the only long-term
bioaccumulation data for the Haven. Continuation of this strategy, based on similar
species and sites, provides an opportunity to establish whether there has been
improvement or deterioration in contamination levels, as well as providing a
contemporary baseline against which future change can be gauged.
In the context of detecting environmental change, and to make data comparable with
earlier surveys, we have made efforts to minimize the effects of biotic factors such as
seasonality and size (Bryan et al., 1985; Langston and Spence, 1995). For metals and
some organotins, MBA has data for the same species for most estuaries in Wales and
England. Harmonised sampling routines have thus helped to facilitate nationwide
comparisons and to achieve other key objectives of the current study, namely to
establish a rigorous bioaccumulation programme that offers wide coverage within
MHW; to undertake bioaccumulation surveillance on a range of metals and organic
contaminants; and to assess current status and to provide benchmarks for future
comparisons, making use of earlier surveys where possible.
6

2 Materials and methods
2.1 Sampling
Field-survey work in Milford Haven was undertaken on two occasions: September
2007 (N. diversicolor and L. littorea) and March 2008 (M. edulis, C. edule and F.
vesiculosus). Reference samples for each species were also collected in the Tywi
Estuary at these times. Sites are shown in Figure 1; grid references and species
occurrences are summarised in Table 1.
Because of different habitat preferences, sites for sediment dwellers (N.diversicolor,
C.edule) and rocky shore species (L.littorea, M. edulis and F. Vesiculosus) were not
always identical but were as close as practical.
All biota were returned live to the laboratory and immediately submitted to clean-up
procedures (Bryan et al., 1985) in preparation for freeze-drying and analysis. Whole
organism size, weight and tissue wet and dry weight data were recorded. Condition
indices (CI) for bivalves describe the relationship between soft tissue dry weight and
the organism total size. High CI values are considered an integrated signal of better
‘health’ status. The condition index used in this study was CI4 as defined in
Lundebye et al., 1997:
CI= (Soft tissue dry weight (g) x 1000)/(shell length (cm))3
2.2 Analysis
Contaminant analyses were performed on batches of 60 pooled mussels and cockles,
30-100 winkles, 250-600 Nereis and 20 Fucus plants, freeze-dried to constant weight
and homogenised to a fine powder. Metals were analysed in all species; several
classes of organic contaminants (organotins, PAHs, PCBs) were also analysed in M.
edulis and N. diversicolor. 7

Metals
Metals analysed included: Ag, As, Cd, Co, Cr, Cu, Fe, Hg, Mn, Ni, Pb, Se, Sn, and
Zn. Sub-samples (0.5g) of freeze-dried homogenate were digested with 5ml HNO3
(Fisons Primar grade) and 1 ml H2O2 in a Milestone (1200 Mega) microwave
digestion system. For analysis of Ag, a hotplate digestion of a separate subsample was
employed (Langston et al., 1994a). Digests were analysed by Flame Atomic
Absorption, or, where concentrations were low, by Graphite Furnace AA. Hg and Se
were analysed by cold vapour and hydride generation systems, respectively. To
prepare samples for arsenic and total tin analysis, 5 ml ashing slurry (6% magnesium
nitrate, 10% magnesium oxide) were added to sub- samples of the freeze-dried
homogenate; these were ashed in a muffle furnace and dissolved in 10ml HCl prior to
analysis by hydride generation AA. Quality assurance included the use of the
Certified Reference Materials DORM-2 (National Research Council), LUTS-1 and
IAEA-140 (seaweed), ensuring that determinations fell within the confidence intervals
of the assigned values. Detection limits for metals ranged from 0.01(Sn) to 0.1
(Cu,Zn) µg g-1 dw
Biometric and contaminant data were input to Microsoft Excel spreadsheets.
Statistical analyses were performed using Statistica (Statsoft Corp.). Metals data were
interlinked to the larger MBA database (Langston et al 1994b). In order to place MH
results into context, MH data are ranked in comparison to the rest of the UK and
expressed as percentiles of the values present in the database. Summary plots for all
species and maps for M.edulis are shown here. Maps for other species are provided as
electronic supplementary material (Figs A-D). MH data below the lower quartile
value (lowest 25% of UK values) are plotted as green bars, red if above the upper
quartile (highest 25%). Values in the mid range (25-75th percentile) are plotted as grey
bars.
8

Organotins
Compounds analysed included: monobutyl-, dibutyl- and tributyltin (MBT,DBT,
TBT) and monophenyl-, diphenyl- and triphenyltin (MPT, DPT,TPT) based on
methodology of Harino et al. (2005). Tissue samples, including aliquots spiked with
standards, were extracted with HCl and acetone, followed by tropolone-benzene, then
propylated and cleaned on florisil, prior to analysis by GC-FPD. The detection limits
were ~0.004µg g-1 dry wt. Quality assurance was established using certified reference
materials, PACS-1, CRM 462,477.
Polyaromatic hydrocarbons (PAHs)
Seventeen PAHs were assayed in M.edulis and N. diversicolor, ranging from 2 ring
naphthalene and 1-methyl-naphthalene (petrogenic), to six ring PAHs
benzo[ghi]perylene and indeno[1,2,3-cd]pyrene, of probable pyrogenic origin (Law et
al., 1999).
Freeze-dried tissue samples were extracted with acetonitrile and tetrahydrofuran
(THF) aided by sonication. Clarified, filtered extracts were analysed by HPLC
(gradient programming) equipped with scanning fluorescence detector (see Vane et
al., 2007 for details). Quality control was achieved by subjecting a well-characterised,
low-level PAH proficiency-testing marine sediment (Quasimeme – QPH048MS) to
the above procedure. Three QCs and three procedural blanks were conducted at
intervals throughout the analysis of the samples (duplicates). Limits of detection for
PAHs ranged from 0.1 to 0.7 µg kg-1.
Polychlorinated biphenyls (PCBs)
The ICES 7 PCB congeners, 28, 52, 101, 118, 138, 153 and 180 were determined by
gas chromatography mass spectrometry (GCMS) analysis. Freeze dried samples and
CRMs were spiked with internal standards (PCBs: 019, 034, 062, 119, 147, 131, 173
9

at 1.4-2.5 ng/g in hexane) The CRM was BCR-718 and the procedural blank was inert
hydromatrix dispersant.
Samples were extracted by Dionex 200 accelerated solvent extraction (ASE) system.
Stainless steel extraction cells (33ml) were filled with the sample, 2-3g of copper
powder and hydromatrix. Extraction conditions were: solvent acetone:hexane (1:1),
temperature 100°C, pressure 2000 psi, heat-up time 1 min, static time 8 minutes, flush
volume 120%, purge time 60 seconds, total solvent used 37-44 mL per sample.
During clean-up, acetone was removed by mixing the extract with de-ionised water
and disposing of the aqueous layer. The fatty-emulsion in the organic fraction was
broken-down by the drop-wise addition of concentrated sulphuric acid (10mL). The
aqueous / organic phases were allowed to separate and the acidic layer removed. The
organic phase was reduced to 8-10mL and dried with anhydrous sodium sulphate.
Interferences were removed using back-extraction with dimethylsulphoxide (DMSO).
DMSO fractions were mixed with de-ionised water (25 mL) and extracted with n-
hexane (2 x 50mL). The hexane extracts were combined, reduced to approximately
5mL using rotary evaporation, and dried with anhydrous sodium sulphate. A recovery
standard was added (containing PCB029 and PCB157, 0.40ng/g in hexane) and
reduced to approximately 0.2mL using nitrogen.
GCMS analysis was performed using Varian CP-3800 GC fitted with a Factor Four
column (VF-1ms, 60m, 0.32mm, 0.25µm). Oven temperature program was: 60°C (1
minute isothermal) to 200°C (at 5°C / min) to 280°C (at 2.6°C / min) to 320°C (at
20°C / min, 20minutes isothermal at 320°C). Injection (1µL, 250°C) was splitless for
the first 0.7min and 1:50 thereafter. Helium carrier gas was at 1mL/min. Mass
spectrometer (Varian 1200L) operating conditions were: selective ion monitoring (m/z
57-328), 0.5 scans/sec, ionization energy 70 eV, detector voltage fixed at 1700V. Data
10

processing was performed using Varian MS Workstation (version 6.5). The limit of
detection was 0.5ng/g.
Extensive contextual data are not available for organic contaminants and MH data
have been compared with OSPAR Guideline values. For TBT and PCBs, OSPAR
Environmental Assessment Criteria (EAC) in mussels have been used as generic
benchmarks. The OSPAR scheme identifies two types of EAC: (1) “lower”
concentrations, below which it is reasonable to expect that there will be an acceptable
level of protection from chronic effects (presented as green bars in the maps for TBT
and PCBs shown in the text and supplementary data); and (2) EAC “higher” –
concentrations above which it is reasonable to expect acute toxic effects on marine
species (plotted as red bars). The concentrations in between these upper and lower
values indicate sub-lethal effects cannot be ruled out (plotted in grey). Black bars are
used where reference values have not been set for a particular contaminant.
The lower and higher OSPAR environmental assessment criteria (EAC) for TBT
(mussels) are 0.012 and 0.175 mg kg-1 wet weight, respectively (OSPAR, 2004).
The lower and higher OSPAR environmental assessment criteria (EAC) for ∑ICES7
chloro-biphenyls (CBs) in mussels are 0.75 and 7.5 µg kg-1 wet weight, respectively
(OSPAR, 2000; NMMP 2004). These have been converted to dry weight value of 5
and 50 µg kg-1 dw by multiplying by the average wet:dry weight ratio of 6.66.
For a number of PAHs we have compared values in relation to OSPAR (2007)
‘Background Concentrations’ (BC–concentrations expected at undeveloped sites
around the North Atlantic) and ‘Background Assessment Criteria’ (BAC) for mussels
(above BAC concentrations values can be considered ‘above background’- see Table
A in the supplementary electronic material). PAH concentrations at or below the BC
values are mapped in green, those above the BAC in red. Concentrations between
these values are plotted in grey. Black bars are used for those PAHs where reference
11

values have not been set. It should be noted that all such classifications are for
guideline purposes only and are based on generic data.
Biomarkers
To demonstrate a possible way forward for future assessment of MHW, in terms of
better integration of environmental data with biological ‘health’ we have trialled two
biological response ‘biomarkers’ in mussels- metallothionein (MT -a measure of
metal exposure) and TOSC (total oxyradical scavenging capacity- a measure of
oxidative stress). A sub-sample of eight M. edulis in samples collected for
contaminant analysis were dissected into gills and digestive glands. Each tissue
sample was “flash-frozen” with liquid nitrogen in an Eppendorf tube, stored at -70°C
and processed as described by Chesman et al. (2007). Metallothionein was determined
by differential pulse polarography (DPP) using a PARC model 174A analyser, and a
PARC/EG&G model 303 static mercury drop electrode. TOSC was determined in the
same samples by measuring the suppression of ethylene formation by a system
containing the peroxy-radical generator 2,2'-azobis-amidinopropane dihydrochloride
and a labile substrate α-keto-γ-methiolbutyric acid (Chesman et al., 2007).
3. Results
Metals
Maps of metals in M. edulis are shown in Figure 2 and those in C. edule, N.
diversicolor, L. littorea and F. vesiculosus are included as supplementary electronic
material (figures A-D). Species differences in body burdens and spatial trends are due
to physiological and ecological attributes of individual bioindicators and the chemical
properties of different metals. However, it is possible to make some general
12

observations regarding bioavailability. Thus, in all species, Milford Haven body
burdens were generally higher than those in the Tywi Estuary- the latter appeared to
be a suitable regional estuarine reference site for most contaminants. One of the few
exceptions concerns the slightly higher Mn burden in Nereis at the upper Tywi
sediment site: this may reflect local sediment pore water conditions, particularly lower
salinity (i.e. this apparent anomaly may be due to natural factors rather than
contamination).
For the majority of metals and species, concentrations in MH biota were at the lower
or middle part of the UK range (green and grey bars, respectively, in Figure 2 and
Figures A-D in supplementary material; see legends for explanation).
Exceptionally, concentrations of a few elements in certain taxa of MHW were
consistently at the higher end of the UK range (within the upper 25% of values, as
represented by red bars in maps). These included As and Se in molluscs and seaweed.
Also, as indicated in these figures, elevated levels were observed for individual
sites/species: namely, Mn (molluscs, seaweed), Co (mussels, seaweed), Sn (bivalves),
Ni (cockles) and Fe (ragworm).
Increases in bioavailability at upstream sites were evident in several metal-species
combinations, which may reflect the influence of geogenic or other land-based
sources. As with the example of Mn in Nereis, above, this pattern may be enhanced
by the presence of lower salinities upstream (greater proportions of more bioavailable
forms and less competition from chloride complexation). The strongest of gradients
were seen for Cd (bivalves), Co (molluscs, seaweed), Mn (bivalves, seaweed), Ag, Ni
(bivalves, ragworm, seaweed) and Sn (cockles, winkles). There was little indication of
raised levels of bioaccumulation from sources in the lower part of Milford Haven.
It is, perhaps, simplistic to generalise further over the status of metals in biota from
the Haven but, by averaging percentiles across the range of species used, an overview
13

of rankings (for maximum, minimum and mean values in the waterway) can be
derived for each metal, in the context of UK ranges (Figure 3). The major features
from this data treatment are that: (1) The upper concentrations of As and Mn at MH
sites, averaged across all five study species (max percentiles, black bars in Figure 3),
indicate that bioaccumulation is consistently high in UK terms, (>80th percentile).
Upper values for Co, Fe, Ni, Se and Sn were > 50th percentile. (2) Mean levels in MH
biota (grey bars, Figure 3) were > 50th UK percentiles for As, Mn and Se confirming
a widespread degree of contamination with these elements throughout much of the
estuary. For most other metals, mean levels were close to or below the 25th UK
percentile. (3) Minimum concentrations in MH biota - usually at seaward sites –were
within the lowest 10% of UK concentrations in most cases (white bars, Figure 3) and
can be considered close to background. Exceptions were As, Mn (Se, Fe) implying
some enrichment above background for these elements, even in the least contaminated
parts of the Haven.
There is a broad similarity in the spatial patterns of metal bioaccumulation described
here with earlier descriptions of sediment-metal distributions, in that highest sediment
concentrations tend to be found upstream, whilst minimum values were generally
those at the mouth of MH Waterway (Smith and Hobbs, 1994). Metal inputs probably
occur along much of the waterway from sewage and industrial sources, though natural
concentrating mechanisms (eg adsorption) can result in net transport of contaminated
particles towards the head of the estuary. Here they combine with catchment sources,
including a component from weathering of mineralized rocks (Cu, Pb, Zn) and
sulphide-rich coal measures (Se?). Exceptions to this geochemical trend do occur,
however: for example, anomalously high sediment concentrations have occasionally
been observed near the mouth of the Haven (e.g. Cd and Hg), at the mouth of
Cosheston Pill (Hg), and elsewhere, and appear to represent confined ‘hotspots’. A
14

landfill site at the industrial estate bordering Cosheston Pill (near Pembroke Ferry,
Figure 1) may be responsible for this particular localized enrichment (Smith and
Hobbs, 1994).
For physiological reasons however, sediment metal levels are not necessarily reflected
in bioaccumulation. Copper, zinc (and lead) were considered by Smith and Hobbs
(1994) to be substantially enriched in most of the MH sediments (compared with
world-wide standard shale), partly as a consequence of anthropogenic inputs. In
contrast, there were few indications of significant Cu or Zn enrichment in biota during
the current study. This is not as contradictory as may appear, since many organisms
have essential requirements for these elements and may regulate body burdens, thus
underestimating environmental contamination. Lead levels in most biota sampled in
the present survey were at the lower end of the UK range which may reflect declining
levels from anthropogenic sources, notably following removal of alkyl lead from
petrol.
The content of most other metals in MH sediments (Smith and Hobbs, 1994) were,
with the exception of Co, typical of concentrations in UK estuaries (Bryan and
Langston, 1992). Co enrichment in sediment from upstream sites in MHW may be
indicative of catchment sources, probably natural in origin, and was reflected in the
pattern of bioaccumulation in mussels and seaweed in the current survey.
Estimates made 20y ago indicated that the relative inputs of (organic) pollution to
MHW were divided equally between freshwater, industrial and sewage discharges,
though volumes were dominated by freshwater (96%), of which the W and E Cleddau
represented some 80% of total FW flow. Inputs of Cu, Pb and Zn from these two
rivers were estimated at 336, 336, 1000 kg y-1 and 381, 381, 1300 kg y-1, respectively
(Hobbs and Morgan, 1992; Bent, 2000). Atmospheric discharges from Pembroke
Power Station (1993) were estimated to be 4.4 (As), <300 (Cd, Cr, Cu), 6600 (Fe)
15

13600 (Ni) and 65300 (V) kg y-1. Less than 1% of total pollutant load was estimated
to come from road run-off. Contaminated land/ landfill inputs are potential, as yet
unquantified sources. Likewise, budgets for Cu and Zn leaching from antifouling are
unknown. More accurate and updated loadings estimates for contaminants would be
useful in terms of identifying specific sources of bioaccumulation.
Results from the few previously published studies do not indicate exceptional metal
bioaccumulation in MH biota, in agreement with the current data. Mussel samples
from 1996 and 1997 collected at St Ishmaels, near Dale (Widdows et al., 2002) are
most relevant to our 2008 survey. Taking the 1996 results as baselines, recent As and
Hg concentrations were higher, by 50% and 200%, respectively. In contrast, there
have been consistent declines for Cd, Ni, Se (50-100%) and, to a small extent, for Cu
(~20%).
Earlier (1993-5) biomonitoring with seaweeds, as part of the EA (NRA) Welsh
Region Marine Biological Programme (Davies and Ellery, 1995), included analyses of
Cd, Cu and Zn at Dale and Lawrenny (MH), together with Ferryside in the Tywi (used
in the current survey, Figure 1) and comparisons are plotted in Figure 4. During the
intervening period, Cd concentrations throughout the area dropped by ~50-60%
(comparable to trends in mussels). This reduction in Cd bioaccumulation is
presumably in response to controls on Cd inputs in the wider area (including major
sources in the Severn Estuary). Copper concentrations have also dropped, by ~30%, at
MH sites and by > 60% in the Tywi (though the 1995 data appear to have been
exceptionally high). In contrast, changes in Zn concentrations have been relatively
small over the last 15 years (Figure 4).
16

TBT and other organotins
Organotin (OT) burdens in mussels are displayed in Figure 5 (see also table B,
supplementary material). Butyltin (TBT, DBT, MBT) concentrations mostly increased
upstream. The exception was a relatively high value for MBT at Dale. When
compared with OSPAR Environmental Assessment Criteria (EAC), TBT
concentrations were above the lower EAC but below the upper EAC. In contrast, TBT
concentrations (and other BTs) in mussels from the Tywi reference site were close to
zero.
The proportion of ∑BT (TBT+DBT+MBT) present in environmental samples as the
parent compound, TBT, is an indication of the relative importance of recent sources (a
high proportion of TBT indicates fresh inputs). In the current survey, the proportion of
TBT varied from 12% at Dale to almost 50% at Pembroke Ferry (mean 31%). In
contrast, mussels from the reference site in the Tywi Estuary contained <1% TBT
with most of the BT burden present as the breakdown product MBT. This implies that
the influence of recently introduced TBT was greatest in the vicinity of Pembroke
Ferry and lowest at the reference site.
The concentrations of phenyltins in mussels were much lower than butyltins (TPT and
DPT are shown in Figure 5, MPT was below detection). This pattern probably reflects
the low usage of TPT as an antifouling agent, compared with TBT. Nevertheless, the
pattern of TPT distribution in mussels closely resembles that of TBT. The proportion
of total PT present as the parent compound, TPT, varied from 66-100% in MH (mean
81%), compared with 29% at the Tywi reference site. Again this infers more recent
input to the MHW and that phenyl tins, though present at lower concentrations than
butyltins, may persist for longer before degradation.
In their survey of mussel condition around the Irish Sea in 1996/7, Widdows et al.,
(2002) found that the concentrations of OT were generally low at coastal sites but
17

elevated near harbours, with ΣBT concentrations ranging from ‘not detected’ to 0.66
μg g−1 dry wt. The highest value on the UK mainland occurred in MH mussels - from
St. Ishmaels, in the vicinity of the oil terminal (Widdows et al., 2002). The reported
concentrations of TBT (0.44 μg g−1) were above the threshold of 0.2 μg g−1 at which
uncoupling of oxidative phosphorylation, the primary mechanism of toxicity,
commences. The maximum TBT concentration in the current study was 0.12 μg g−1
(Pembroke Ferry). Whilst not threatening the survival of mussels, accumulated body
burdens of this order would represent a challenge to the reproductive success of TBT-
sensitive neogastropods such as dogwhelks Nucella lapillus. Previous surveys have
indicated that TBT body burdens and imposex in N. lapillus were generally highest
near the central and outer part of Milford Haven (where tankers moor), decreasing
(with anomalies) seaward (Harding et al., 1998). Neyland Marina, opposite Pembroke
Ferry, has also previously been identified as a TBT ‘hotspot’, whilst upstream, TBT
levels in the Cleddau were considered typical of UK estuaries (Kitts, 1999; Bent,
2000).
Maximum BT values in MH mussels in 1996/7 (Widdows et al., 2002), were higher
than any of the current values: TBT and DBT concentrations at Dale are now almost
100% lower than those previously recorded at nearby St. Ishmaels. Theoretically, the
global ban on TBT in 2008 should start to produce a further decline in TBT
contamination. The current data set is therefore a timely and suitable platform to test
this hypothesis and to establish the rate of recovery.
Organotin concentrations in the ragworm N. diversicolor (see Table B and figure E,
supplementary information) display somewhat different spatial distributions to those
in mussels for a number of possible reasons, including the fact that they live and feed
in benthic sediments. Bioavailability of TBT in this phase may influence body
burdens to a greater extent than filter feeding mussels whose main source of TBT is
18

probably the overlying water (including suspended organic matter). Secondly, the
sites at which N. diversicolor were collected were slightly different to mussel sites. N.
diversicolor may also be better at metabolizing TBT to DBT and MBT. Thus,
concentrations of TBT in N. diversicolor were on average fourfold lower than in
mussels.
In contrast to mussels, the average triphenyltin (TPT) concentration in worms was
five-fold higher than that in mussels (DPT seventy-fold higher), consistent with
observations that phenyltins partition more strongly towards sediment and might
therefore be more available to sediment-dwelling species. Additionally, the Nereis site
with highest TPT (and MPT) levels, at Pembroke (Waterloo), was within Cosheston
Pill –and perhaps subjected to localized sources, compared with the nearest mussel
site (Pembroke Ferry) in the main waterway. The use of TPT in antifouling has been
less extensive than TBT in the UK; however, phenyltins have in the past also been
used in agriculture as pesticides and fungicides.
PCBs
Results summarizing biomonitoring of PCBs in Mytilus edulis and Nereis diversicolor
based on analysis of the ∑ICES 7 congeners (CBs 28, 52, 101, 118, 138, 153 and 180)
are plotted in Figure 6 (see also table C, supplementary material). Concentrations in
MH mussels showed an unexpected spatial trend in that highest values occurred at the
mouth of the Haven and decreased in an upstream direction: The lowest values here
were comparable to the reference site in the Tywi Estuary. Manufacture of PCBs has
long since ceased and it seems unlikely that there are recent discharges of any
significance which influence distribution. Rather, biotic factors may be involved since
PCB concentrations in mussels follow a pattern which is similar to the condition
index (below). It may be that body burdens of these lipophilic contaminants are a
19

function of lipid reserves which would be anticipated, from condition data, to be
highest in those populations at the seaward end of the waterway.
The lower and higher OSPAR environmental assessment criteria (EAC) for ∑ICES7
PCBs in mussel are 5 and 50 µg kg-1 dw and all values for Milford Haven and Tywi
samples were below the upper threshold (above which effects on marine species
might be expected), but exceeded the lower ‘no-effects’ threshold (OSPAR, 2004;
NMMP, 2004). These EACs are only guidelines to ecotoxicological impact - and may
be set with over-precautionary ‘safety factors’: nevertheless, the results indicate that
supplementary biological effects studies would be appropriate.
Some context is gained by comparison of PCB values in N. diversicolor from MH
with others from the Severn Estuary (table 2). For the lower chlorinated congeners,
concentrations in both systems were relatively low: for the more highly chlorinated
CBs 138, 153 and 180 (hexa- and hepta-chlorobiphenyls), and total PCBs, there was
considerably more bioaccumulation in the Severn (on average almost 20-fold for
∑CBs). This excess in the Severn is not unexpected, since PCBs were previously
manufactured at in the region at Newport, a known hotspot for contamination.
There are few comparative PCB data from earlier surveys. Concentrations of ∑25
PCB congeners (composition not stated), determined in mussels from Milford Haven
(St Ishmaels) in 1996/7, were in the range 0.009-0.013 µg g-1 (Widdows et al., 2002).
These were somewhat lower than values determined in the current study. In contrast,
concentrations in mussels from hotspots in the Irish Sea, such as Liverpool Bay and
the Mersey Estuary, were an order of magnitude higher (Widdows et al., 2002).
PAHs
PAH concentrations in M. edulis sampled during the current project are tabulated in
the supplementary data (table D). Some of the key features are illustrated in Figure 7.
20

For many PAHs, including 1-methyl-naphthalene, acenaphthene, anthracene,
chrysene, perylene, benzo(ghi)perylene, ideno(1,2,3-cd)pyrene,
benzo(k)fluoranthene, benzo(b)fluoranthene, benzo(a)anthracene and benzo(a)pyrene
(latter shown in figure 7), body burdens were elevated in the central and upper parts
of MH and were higher than the Tywi reference site (this applies also to ∑PAHs).
However, for several PAHs (fluorene, phenanthrene, fluoranthene, and pyrene)
concentrations were not substantially different across the sites (example shown in
figure 7 is pyrene). It appears that distributions fall into groupings which may reflect
their similarity in origins and chemistry.
Placed in the context of OSPAR Background Concentrations (BC) and Background
Assessment Criteria (BAC) for mussels, levels above background (BAC) are observed
widely for phenanthrene, fluoranthene, benzo(a)anthracene, benzo(ghi)perylene,
benzo(a)pyrene and pyrene (see examples for latter two PAHs in figure 7; bars
depicted in red are above background (>BAC); green bars represent concentrations
characteristic of undeveloped sites (BC); grey bars lie in between). There were a few
PAH concentrations which would be considered as background and representative of
undeveloped sites (see, for example, anthracene Figure 7). Again it is stressed that
these comparisons are guidelines only, not proof of ecotoxicological threat. Despite
evidence of bioaccumulation of most PAHs in mussels, acute effects are not expected
from the body burdens measured; in reality, however, the true extent of combined
sub-lethal effects from these and other compounds is unknown and requires
assessment.
Distributions of several PAHs in Nereis diversicolor were similar to those in M.edulis
(see supplementary data, table D); subtle differences could be due in part to localized
sampling site differences between the two species and the stronger influence of
sediment contamination on N. diversicolor. Species differences in the activity of
21

enzymes which metabolise PAHs (e.g. cytochrome P450 system) will also contribute
to the variation in bioaccumulation patterns. The main conclusions drawn from PAHs
in Nereis were that Milford Haven body burdens were often higher than at the Tywi
reference site, with enrichment most notable for naphthalene and 1-methyl-
naphthalene, especially upstream. Acenaphthene, fluoranthene, pyrene,
benzo(a)anthracene and chrysene were characterized by elevated levels at Dale. A
third group (fluorene, phenanthrene, anthracene, benzo(b)fluoranthene, indeno (1,2,3-
cd) pyrene) was typified by a more even distribution across all sites, including the
Tywi reference site.
Results for PAHs in N. diversicolor from MH have been compared with data for the
Severn Estuary, with ratios for individual PAHs shown in table 3, indicating that
Severn worms contained higher levels than those in Milford Haven – particularly so
for the higher molecular weight PAHs. Only phenanthrene was higher in Nereis from
MH. ∑PAHs were on average three-fold higher in worms from the Severn Estuary.
To place mussel data in similar perspective, the pattern in PAH bioaccumulation in
M.edulis along the Bristol Channel/Severn Estuary was established in a study of
benzo(a)pyrene (BaP) residues conducted after the Sea Empress grounding in MH in
1996 (CEFAS, 2000; Law et al., 1999). An increase in BaP bioaccumulation in
mussels eastwards from the Mouth of MH, towards Cardiff, was tentatively related to
the trend in urban development along the coastline and to the delivery of PAHs from
the Severn catchment upstream. PAH enrichment in mussels from the Severn, relative
to that in Milford Haven, therefore mirrors the pattern in Nereis.
Sources of PAHs in Severn Estuary have included the south Wales coalfield, oil
bearing shales in Bridgwater Bay, combustion of fossil fuels, land run-off and the
precipitation of airborne particulates. As a result, sediments in the Severn were
considered moderately high in PAHs in a survey of UK coastal sites conducted by
22

Woodhead et al. (1999), though the highest value was obtained from Milford Haven.
Correspondingly, ∑18 PAHs in MH mussels at the time of the Sea Empress oil spill
rose to extremely high levels within a few days at sites such as Angle and Dale
(26189 -100946 µg kg-1 ww ) though within three months these had dropped back by
>90% (to 500 µg kg-1 ww) notably for oil-derived PAHs like naphthalene and
phenanthrene and their alkylated derivatives (Dyrynda et al., 2000). Current ∑ PAH
values at Angle and Dale indicated further reductions and were 30 and 18 µg kg-1 ww,
respectively.
Pyrogenic PAHs exhibited somewhat different temporal behaviours following the
grounding of the tanker (Law et al., 1999; Dyrynda et al., 2000). Bioaccumulation in
mussels peaked 23 days after the February spill (up to ~80 µg kg-1 ww), before
declining close to zero in mid-summer. One year post-spill, further small seasonal
(winter) peaks in combustion-derived (high molecular weight) PAHs, such as benzo-
a-pyrene (BaP), indeno[1,2,3-cd] pyrene and benzo[ghi] perylene† were observed in
mussels from Dale and Angle (upto ~40 µg BaP kg-1 ww at the latter site), before
returning to baseline levels of a few µg kg-1 (< 1 µg BaP kg-1 ww in the current
survey).
Similar seasonal patterns have been observed for pyrogenic PAHs in N Sea mussels
(reaffirming that PAH bioaccumulation is not always related to oil spillage, and may
influenced by biotic as well as physicochemical and climatic factors such as run-off
and precipitation). The return to lower PAH levels in spring may also be a function of
depuration and losses during spawning (gonadal tissues are enriched in PAHs relative
to other tissues), subsequent rapid growth (dilution), and perhaps increased
photodegradation. Though metabolism of hydrocarbons is considered to be relatively
† Other combustion derived PAHs include anthracene, fluoranthene, pyrene, benzanthracenes,
chrysene, benzofluoranthenes. There was also an indication of a peak in immunosupression effects in
these mussels, coinciding with raised PAHs. Dyrynda et al., 2000)
23

limited in molluscs, this ability will tend to rise with seasonally increasing
temperatures (Law et al., 1999; Dyrynda et al., 2000). The samples taken in our
survey of Milford Haven, in March, would coincide with the maxima in pyrogenic
PAHs described previously.
Discussion
Integrating bioaccumulation, environmental and biological response data –the
way forward?
Biometric data for M. edulis and C. edule are expressed in terms of condition indices
(CI) in the current survey, with higher values signifying better condition. Thus the CI
of mussel (a measure of tissue weight relative to shell length) is shown in Figure 8A.
Highest values were those at St Ishmael (SI -the reference site in the Tywi) and at
Angle Bay (A) in the lower Haven. All other samples in MH were significantly lower
than the reference site, SI, with values generally lowest upstream from Pembroke
Ferry (PF). The CI of whole cockles is shown in Figure 8B and resembles strongly the
pattern in mussels. Highest values were those at St Ishmael (the reference site in the
Tywi) and at Dale; CI decreased upstream in the Cleddau Estuary.
Correlation coefficients between CI and metals in mussels indicate several significant
(negative) relationships (Table 4) with CI decreasing according to the sequence
Pb>Se>Cu>Zn>Hg>Ni. (The corresponding sequence for metals in cockles was
Co>Ni>Hg=As>Se>Mn>Cu=Fe).
For mussels it is possible to extend the range of contaminants in these comparisons to
include PAHs, PCBs and organotins. The ranking of contaminants, for those which
are significantly (negatively) correlated with CI, is shown in Table 4. Some of the
more toxic (e.g. Pb, TBT, Cu, benzo(a)pyrene) appear high on this list and it seems
plausible that a combination of contaminants could have an influence on CI.
24

Multivariate analysis of Milford Haven data
Body burdens of a number of contaminants co-vary illustrating possible common
sources. However, such covariance contributes to the difficulty in attributing cause
and effect. There are also other factors which may be influential in the condition of
bivalves; notably feeding rate, food availability and salinity. There is a strong case to
attempt better co-ordination of chemical and biological monitoring in future. To
establish cause and effect more substantially would require additional measurements
of sub-lethal effects and environmental parameters, coupled with multivariate
statistical techniques to try and tease out the relative importance of these parameters.
We have attempted to illustrate the benefits of this approach by including
measurements of sample biomarkers TOSC (total oxyradical scavenging capacity – a
measure of oxidative stress) and MT (metallothionein – a metal binding protein
synthesised to protect organisms from excess intracellular metal) in MH mussels.
These have been analysed alongside other biological (CI) and chemical variables
(body burdens of metals, organotins, ∑PAHs and ∑PCBs), using the PRIMER-E
statistical package to examine the principle components responsible for site
distinctiveness and to demonstrate a possible way forward for future assessment.
This principal component analysis (PCA) approach enables reduction of complex
multivariate data to a much simpler representation of ‘similarity’ i.e. which sites are
most similar to, or different from, each other. Normalisation was performed so that
each variable has equal weighting in the final PCA ordination which reduces the
location of each sample to a best-fitting 2-D or 3-D solution. A sample 2-D PCA plot
for mussels is shown in Figure 9 and represents 58.5% of the variation, while 3-D
ordination captures 72.7%. There is approximately equal (positive) contribution to
PC1 from each of the metals and organotins, together with (negative) contributions
25

from some of the biometric data (weight and condition index). PC2 is dominated by
eigenvectors representing ∑PAHs (positive) and digestive gland TOSC (negative).
Figure 9 appears cluttered due to superimposition of the eigenvectors for each
variable (shown as individual axes), but shows how the different sampling sites relate
to each other within this dimensionally-reduced multivariate space. Clearly the
‘control’ site (St Ishmael, Tywi) on the bottom left of the plot, is distant from most of
the other sites. By considering the eigenvectors, this appears largely to be due to a
combination of low metals, low hydrocarbons and PCBs, high levels of digestive
gland TOSC (high capacity to scavenge oxyradicals) and high levels of biometrics
(size, weight and condition index). Collectively this indicates that mussels from the St
Ishmael reference were generally larger, in good condition, with low levels of
accumulated contaminants and high capacity to cope with pollutant-induced stress.
Outer sites in Milford Haven - Dale and Angle - also separated from inner estuary
sites in Figure 9. Dale is at a similar level to St Ishmael on PC2 (hydrocarbons, PCBs
and DG TOSC) but further to the right on PC1 indicating higher metal burdens and
lower biometrics. In contrast, at Angle, across the waterway there appears to be
slightly less influence of metals, but much higher effect of hydrocarbons and PCBs
together with a reduction in DG TOSC.
Pennar Mouth, the entrance to the Pembroke River, sits between Dale and Angle in
Figure 9, although it shows greatest similarity with Angle, probably reflecting the
influence of hydrocarbons and PCBs at this site.
The remaining inner estuary sites fall on the right-hand side of the PCA ordination,
with an overall tendency for the most upstream sites to lie towards the bottom of the
plot. The location of these sites on the right would appear to result principally from
the combined accumulation of metals and organotins together with metallotheionein
induction in mussels, while the vertical ‘zonation’ results from the components in PC2
26

(principally total hydrocarbons, PCBs and DG TOSC, although the eigenvectors show
that several metals also exert influence in this direction as well).
There is a risk in ‘over-analysing’ PCA plots when determining the factors that best
describe the similarities or differences seen between sites, since PCA is a
simplification of the true multivariate data. Given these caveats, it may be regarded
that overall, in Figure 9, there is an increasing level of contaminant impact and
biological response moving from bottom-left to top right across the PCA plot. Sites
closest to each other (in the PCA plot, but not necessarily geographically) are the most
similar, while, conversely, those furthest apart are most different in terms of the
parameters measured in mussels during this survey.
One consideration for the future is to link biological community data,
bioaccumulation, biomarkers and environmental variable data, which will require that
harmonised sampling is conducted. This linked data may be useful to generate
compatible datasets so that we may better understand the principal drivers acting on
the biota in Milford Haven.
Acknowledgements
The authors are grateful to the members of the Milford Haven Waterways Group
(MHWESG) for their support for this work and in particular to Dr Blaise Bullimore
and Captain Mark Andrews for their guidance and suggestions. Thanks are also due to
Nia Watkins for help with sampling and to Dr Hiroya Harino for contributions to
analysis.
Bibliography
Atkins (2005). Development of an Inputs Budget for Milford Haven Waterway.
Technical report to the MHWESG.
27

Bent, E. (2000). A review of environmental studies in Milford Haven Waterway 1992-
2000. Report to the MHWESG.
Bryan, G.W., Langston, W.J., Hummerstone, L.G., and Burt, G.R.(1985). A guide to
the assessment of heavy-metal contamination in estuaries using biological
indicators, Mar. Biol. Ass. U.K., Occasional Publication No. 4, 92 pp.
Burton, S. (2006). Pembrokeshire Marine Special Area of Conservation: Draft
management scheme. 157pp.
CEFAS (2000) Monitoring and surveillance of non-radioactive contaminants in the
aquatic environment and activities regulating the disposal of wastes at sea.
Aquatic Environment Monitoring Report, 52, 92pp.
Chesman, B.S., O’ Hara, S., Burt G.R. and Langston W.J. (2007). Hepatic
metallothionein and total oxyradical scavenging capacity in Atlantic cod Gadus
morhua caged in open sea contamination gradients. Aquatic Toxicology, 84,
310-320.
Davies, G. and Ellery, S. (1995). Results of the NRA Welsh Region marine
bioaccumulation programme 1991-1995. NRA Report No. SE/EAU/95/9
Dyrynda, E.A., Law, R.J., Dyrynda, P.E.J., Kelly, C.A., Pipe, R.K. and Ratcliffe,
N.A. (2000). Changes in immune parameters of natural mussel Mytilus edulis
populations following a major oil spill ('Sea Empress', Wales, UK). Marine
Ecology Progress Series, 206, 155-170.
Harino, H., O'Hara, S.C.M., Burt, G.R., Chesman, B.S. and Langston, W.J. (2005).
Accumulation of butyltin compounds in benthic biota of the Mersey Estuary.
Marine Pollution Bulletin, 50(2), 223-226.
28

Hobbs, G and Morgan, C I (eds.) (1992). A review of the current state of
environmental knowledge of the Milford Haven Waterway. Report to the
Milford Haven Waterway Environmental Monitoring Steering Group. 140pp.
Langston, W.J. and Spence, S.K. (1995). Biological Factors involved in metal
concentrations observed in aquatic organisms. In: Metal Speciation and
Bioavailability, Tessier A. and Turner D.R. (Eds), John Wiley and Sons Ltd.,
407-478
Langston, W.J., Bryan, G.W., Burt, G.R. and Pope, N.D. (1994a) Effects of sediment
metals on estuarine benthic organisms. Project Record, National Rivers
Authority , 49pp.
Langston, W.J., Bryan, G.W., Burt, G.R.(1994b). Heavy metals in UK estuaries:PML
data and mapping programme,R&D Note 280, National Rivers Authority, 85pp.
Law, R.J., Thain, J.E., Kirby, M.F., Allen, Y.T., Lyons, B.P., Kelly, C.A. et al.
(1998). The impact of the Sea Empress oil spill on fish and shellfish. In: The
Sea Empress oil spill, Edwards, R and Sime, H. (Eds), pp 109-136. Lavenham:
Lavenham Press
Law, R.J., Kelly, C.A. and Nicholson, M.D. (1999). Polycyclic aromatic
hydrocarbons (PAH) in shellfish affected by the Sea Empress oil spill in Wales
in 1996. Polycyclic Aromatic Compounds, 17(1-4), 229-239.
Lundebye, A.K., Langston, W.J. & Depledge. M.H. (1997) stress proteins and
condition index as biomarkers of TBT exposure. Ecotoxicology, 6, 127-136
Nelson-Smith, A. (1965). Marine biology of Milford Haven: The physical
environment. Field Stud., 2 (2): 155-188.
29

Nikitik, C.C.S. and Robinson, A.W. (2003). Patterns in benthic populations in the
Milford Haven Waterway following the Sea Empress oil spill with special
reference to amphipods. Mar. Pollut. Bull. 46, 1125-1141.
NMMP (2004). UK National Marine Monitoring Programme. Second Report (1999-
2001). Marine Environment Monitoring Group, CEFAS, ISBN 0 907545 20 3,
136pp.
OSPAR (2000). Quality Status Report 2000 for the North-East Atlantic. OSPAR
Commission, London, 108+viipp.
OSPAR (2004). OSPAR/ICES Workshop on the evaluation and update of background
reference concentrations (B/RCs) and ecotoxicological assessment criteria
(EACs) and how these assessment tools should be used in assessing
contaminants in water, sediment and biota. Hazardous Substances Series. ISBN
1-904426-52-2, 167pp.
OSPAR (2007). 2006/2007 CEMP Assessment: Trends and concentrations of selected
hazardous substances in the marine environment. Assessment and Monitoring
Series. Publication Number: 330/2007, 63pp.
Rostron, D. (1998). Sea Empress sediment shore impact assessment monitoring:
Infauna of heavily oiled shores in Milford Haven and Carmarthen Bay. A
report to the Countryside Council for Wales from SubSea Survey,
Pembroke. 51pp + appendices.
Smith, J. and Hobbs, G. (1994). Metal concentrations in Milford Haven sea bed
sediments – data storage, analysis and initial interpretation. Field Studies
Research Council Research Centre FSC/RC/12/94 , 8pp+ appendices
Vane, C.H., Harrison, I. and Kim, A. (2007). Assessment of polyaromatic
hydrocarbons (PAHs) and polychlorinated biphenyls (PCBs) in surface
30

sediments of the Inner Clyde Estuary, UK. Marine Pollution Bulletin, 54, 1301-
1306.
Widdows, J., Donkin, P., Staff, F.J., Matthiessen, P., Law, R. J., Allen, Y. T., Thain,
J. E., Allchin, C. R. and Jones, B.R. (2002). Measurement of stress effects
(scope for growth) and contaminant levels in mussels (Mytilus edulis) collected
from the Irish Sea. Marine Environmental Research, 53, 327-356.
Woodhead, R.J., Law R.J. & Matthiessen P. (1999). Polycyclic aromatic
hydrocarbons in surface sediments around England and Wales, and their
possible biological significance. Marine Pollution Bulletin 38, 773-790.
31

FIGURES
Figure 1 Location of sampling sites for biota, Milford Haven and Tywi Estuary
32

75 km.
Ag Musse ls
0
0.05
0.1
0.15
0.2
0.25
Ag µg/g dw
75 km.
As Mussels
0
10
20
30
40
50
As µg/g dw
75 km.
Cd Muss els
0
0.2
0.4
0.6
0.8
1
Cd µg/g dw
75 km.
Co Mussels
0
0.2
0.4
0.6
0.8
1
Co µg/g dw
75 km.
Cr Mus sels
0
0.5
1
1.5
2
2.5
Cr µg/g dw
75 km.
Cu Muss els
0
0.5
1
1.5
2
2.5
Cu µg/g dw
75 km.
Fe Mussels
0
100
200
300
400
500
Fe µg/g dw
75 km.
Hg Muss els
0
0.1
0.2
0.3
0.4
0.5
Hg µg/g dw
Figure 2. Metals in mussels Mytilus edulis, µg g-1 dry weight. Values below the
lower quartile value (lowest 25%) of values in MBA UK data base are plotted as
green bars and red if above the upper quartile (highest 25%). Values in the mid
range (25-75th percentile) are represented as grey bars. (cont.)……
33

75 km.
Mn Musse ls
0
5
10
15
20
25
Mn µg/g dw
75 km.
Ni Mus sels
0
0.5
1
1.5
2
2.5
Ni µg/g dw
75 km.
Pb Musse ls
0
1
2
3
4
5
Pb µg/g dw
75 km.
Se Mussels
0
2
4
6
8
10
Se µg/g dw
75 km.
Sn Musse ls
0
0.2
0.4
0.6
0.8
1
Sn µg/g dw
75 km.
Zn Muss els
0
20
40
60
80
100
Zn µg/g dw
...Figure 2 (cont.). Metals in mussels Mytilus edulis, µg g-1 dw. Values below the
lower quartile value (lowest 25%) of values in MBA UK data base are plotted as
green bars and red if above the upper quartile (highest 25%). Values in the mid-
range (25-75th percentile) are represented as grey bars.
34

Figure 3. Metals in Milford Haven biota 2007/8. Minimum, maximum and mean
values expressed as percentiles of UK ranges (averaged across Fucus vesiculosus,
Littorina littorea, Mytilus edulis, Cerastoderma edule and Nereis diversicolor).
35

Figure 4. Fucus vesiculosus: metal concentrations measured in the current
survey at Lawrenny and Dale (MH) and the Tywi reference site, compared with
concentrations determined in the previous decade by Davies and Ellery, 1995).
36

75 km.
TBT Mussel
based on OSPAR EAC criteria (2004)
0
20
40
60
80
100
TBT µg/kg dw
75 km.
TPT mussels
0
1
2
3
4
5
TPT µg/kg dw
75 km.
DBT Mussels
0
10
20
30
40
50
DBT µg/kg dw
75 km.
DPT Mussels
0
0.2
0.4
0.6
0.8
1
DPT µg/kg dw
75 km.
MBT Muss els
0
150
300
450
600
750
MBT µg/kg dw
Figure 5. Organotin concentrations (µg kg-1 dw) in mussels Mytilus edulis
75 km.
Tot al PCBs Muss el s
usi ng OSPAR EAC cri t er i a
0
15
30
45
60
75
µg/ kg dw
75 km.
Tot al PCB Ner ei s
0
5
10
15
20
25
µg/ kg dw
Figure 6. PCBs (∑ICES 7 congeners) in mussels Mytilus edulis and ragworm
Nereis diversicolor. 37

75 km.
Anthra cene Myt ilus
OSPAR BC and BAC cr iter ia
0
0.2
0.4
0.6
0.8
1
µg/kg
75 km.
Pyre ne Mytilus
OSPAR BC & BAC cr iteri a
0
15
30
45
60
75
µg/kg dw
75 km.
Benzo(a)pyrene Mytilus
OSPAR BC & BAC cr iteri a
0
2
4
6
8
10
µg/kg dw
Figure 7. PAHs in Mytilus edulis in context of Background Concentration
(BC) and Background Assessment Criteria (BAC) for mussels. Bars depicted
in red are above background (BAC); green bars represent concentrations
characteristic of undeveloped sites (BC); grey bars lie in between.
A B
±Std. Dev.
±Std. Err.
Mean
Mussel whole tissue condition i ndex (4) - Milford Haven March 2008
comparison with "reference site" St Ishmael (SI) Tywi Estuary
Asteris ks indicate significant differenc e (* p<0.05, ** p<0.01, *** p<0.001)
site
Condition Index (whole tissue)
0
2
4
6
8
10
SI D A PM PPF FH LBT
***
***
***
***
***
***
***
75 km.
Condition Index whole cockles
March 2008
0
5
10
15
20
25
CI
Figure 8. Condition indices in bivalves: (A) Mytilus edulis (B) Cerastoderma
edule (*Sites significantly different to the Tywi reference site at St Ishmael
(SI). Other labels D-Dale; A-Angle; PM- Pennar Mouth; P-Pennar; PF-
Pembroke Ferry; FH-Ferry Hill; L-Lawrenny; BT- Black Tar).
38

-6 -4 -2 0 2 4
PC1
-4
-2
0
2
4
PC2
Pennar mouth
Dale
Black Ta
Ferry Hill
Lawrenny
St Ishmael
Angle
Pembroke Ferry
Pennar
Mean Size
Mean Wet Wt
Mean Dry Wt
Log(0.1+Ag)
Log(0.1+As)
Log(0.1+Cd)
Log(0.1+Co)
Log(0.1+Cr)
Log(0.1+Cu)
Log(0.1+Fe)
Log(0.1+Hg)
Log(0.1+Mn)
Log(0.1+Ni)
Log(0.1+Pb)
Log(0.1+Se)
Log(0.1+Sn)
Log(0.1+Zn)
Log(0.1+MBT)
DBT
TBT
TPT
Total hydrocarbons
Total of 7 PCBs
Gill MT (dw)
Gill MT (p rot)
Gill TOSC
DG TOSC
whole ST CI (4)
Figure 9. 2-dimensional PCA ordination of Milford Haven mussel data.
Eigenvectors for individual variables (contaminants, biomarkers,
biometrics) are indicated. Locations of sampling sites in the diagram show
how they relate to each other within this dimensionally-reduced
multivariate space.
39

TABLES
Table 1. Summary of sampling sites and species distributions
Site
Map ref (sites sampled)
Fuc
Ner
Lit
Myt
Cer
MILFORD HAVEN WATERWAY
Landshipping
SN011118
(+)
++
(+)
(+)
(+)
Landshipping Quay
SN008108
(++)
++
(+)
(+)
Black Tar
SM999093
++
(+)
++
++
Lawrenny (Cresswell/Carew Mouth)
SN017063
+
++
Lawrenny (Jenkin’s Point)
SN009062
++
++
++
++
Ferry Hill
SN003061
++
(+)
++
++
Pembroke Ferry (Waterloo)
SM982040
++
++
++
Pembroke Ferry (Ferry Inn)
SM974047
++
(+)
++
++
Pembroke River (Pennar)
SM959020
++
++
++
++
++
Pembroke River (Pennar Mouth)
SM943028
++
(+)
++
++
Angle Bayb
SM870027
++
++
++
++
Angle Baya
SM868028
++
Daleb
SM809065
++
++
++
++
Dalea
SM815075
++
TYWI REFERENCE SAMPLES
Tywi (1.2km u/s of Ferryside)
SN370117
++
(+)
Tywi (St. Ishmael)
SN361082
++
++
++
+
Key: Fuc, Fucus vesiculosus; Ner, Nereis diversicolor; Lit, Littorina littorea; Myt, Mytilus
edulis; Cer, Cerastoderma edule. + species present and sampled. ++ species numerous and
sampled. (+) species present, but not sampled (in some cases, specimens too small or sparse).
Table 2. Nereis diversicolor. Comparison of CBs in Milford Haven samples with
those in the Severn Estuary.
CB028 CB052 CB101 CB118 CB138 CB153 CB180 ∑CBs
0.5 1.0 2.5 6.7 12.9 10.9 28.0 18.7
Ratios Severn Estuary: Milford Haven
Table 3. Nereis diversicolor. Comparison of PAHs in Milford Haven samples with
those in the Severn Estuary.
Ratios Severn estuary : Milford Haven
Naphthalene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo[a]-
anthracene
Chrysene
Benzo[a]-
pyrene
Benzo[ghi]-
perylene
Indeno[1,2,3-
cd]pyrene
25
0.85
7.7
2.5
6.4
14
13
24
48
110
40

Table 4. Mytilus edulis. Comparison of (negative) Correlation coefficients
between Condition Index and metal body burdens (all values are significant,
P<0.05; ns not significant; nd not determined)
contaminant
Correlation with Condition Index
Pb
-0.86325
TBT
-0.85835
Se
-0.83187
Benzo(a)pyrene
-0.80803
Cu
-0.80061
Zn
-0.78868
Benzo(ghi)perylene
-0.78032
DBT
-0.75958
Dibenzo(ah)anthracene
-0.74688
Perylene
-0.74128
Hg
-0.69862
Benzo(k)fluoranthene
-0.69206
Ni
-0.68709
41

SUPPLEMENTARY MATERIAL
75 km.
Ag Cockles
0
0.02
0.04
0.06
0.08
0.1
Ag µg/g dw
75 km.
As Cockles
0
5
10
15
20
25
As µg/g dw
75 km.
Cd Cockles
0
0.15
0.3
0.45
0.6
0.75
Cd µg/g dw
75 km.
Co Cockles
0
1
2
3
4
5
Co µg/g dw
75 km.
Cr Cockles
0
0.5
1
1.5
2
2.5
Cr µg/ g dw
75 km.
Cu Cockles
0
2
4
6
8
10
Cu µg/g dw
75 km.
Fe Cockles
0
500
1000
1500
2000
2500
Fe µg/g dw
75 km.
Hg Cockles
0
0.1
0.2
0.3
0.4
0.5
Hg µg/g dw
Figure A. Metals in cockles Cerastoderma edule, µg g-1 dw. Values below the
lower quartile value (lowest 25%) of values in MBA UK data base are
plotted as green bars and red if above the upper quartile (highest 25%).
Values in the mid-range (25-75th percentile) are represented as grey bars
(cont.)….
42

75 km.
Mn Cockles
0
20
40
60
80
100
Mn µg/g dw
75 km.
Ni Cockles
0
20
40
60
80
100
Ni µg/g dw
75 km.
Pb Cockles
0
1.5
3
4.5
6
7.5
Pb µg/g dw
75 km.
Se Cockles
0
2
4
6
8
10
Se µg/g dw
75 km.
Sn Cockles
0
1
2
3
4
5
Sn µg/g dw
75 km.
Zn Cockles
0
20
40
60
80
100
Zn µg/g dw
…Figure A(cont.). Metals in cockles Cerastoderma edule, µg g-1 dw. Values
below the lower quartile value (lowest 25%) of values in MBA UK data base
are plotted as green bars and red if above the upper quartile (highest
25%). Values in the mid range (25-75th percentile) are represented as grey
bars.
43

75 km.
Ag Nereis
0
0.2
0.4
0.6
0.8
1
Ag µg/g dw
75 km.
As Nereis
0
5
10
15
20
25
As µg/g dw
75 km.
Cd Nereis
0
0.02
0.04
0.06
0.08
0.1
Cd µg/g dw
75 km.
Co Nereis
0
1.5
3
4.5
6
7.5
Co µg/g dw
75 km.
Cr Nerei s
0
0.15
0.3
0.45
0.6
0.75
Cr µg/ g dw
75 km.
Cu Nereis
0
10
20
30
40
50
Cu µg/g dw
75 km.
Fe Nereis
0
200
400
600
800
1000
Fe µg/g dw
75 km.
Hg Nereis
0
0.05
0.1
0.15
0.2
0.25
Hg µg/g dw
Figure B. Metals in ragworm Nereis diversicolor, µg g-1 dw. Values below the
lower quartile value (lowest 25%) of values in MBA UK data base are
plotted as green bars and red if above the upper quartile (highest 25%).
Values in the mid range (25-75th percentile) are represented as grey bars.
(cont.)…..
44

75 km.
Mn Ner eis
0
5
10
15
20
25
Mn µg/g dw
75 km.
Ni Nerei s
0
1.5
3
4.5
6
7.5
Ni µg/g dw
75 km.
Pb Nereis
0
1
2
3
4
5
Pb µg/g dw
75 km.
Se Nereis
0
2
4
6
8
10
Se µg/g dw
75 km.
Sn Nereis
0
0.1
0.2
0.3
0.4
0.5
Sn µg/g dw
75 km.
Zn Nereis
0
50
100
150
200
250
Zn µg/g dw
....Figure B (cont.). Metals in ragworm Nereis diversicolor, µg g-1 dw. Values
below the lower quartile value (lowest 25%) of values in MBA UK data base
are plotted as green bars and red if above the upper quartile (highest
25%). Values in the mid range (25-75th percentile) are represented as grey
bars.
45

75 km.
Ag Littorina
0
1
2
3
4
5
Ag µg/g dw
75 km.
As Littorina
0
10
20
30
40
50
As µg/g dw
75 km.
Cd Littorina
0
0.2
0.4
0.6
0.8
1
Cd µg/g dw
75 km.
Co Littorina
0
0.2
0.4
0.6
0.8
1
Co µg/g dw
75 km.
Cr Littorina
0
0.15
0.3
0.45
0.6
0.75
Cr µg/ g dw
75 km.
Cu Littorina
0
20
40
60
80
100
Cu µg/g dw
75 km.
Fe Littorina
0
100
200
300
400
500
Fe µg/g dw
75 km.
Hg Littorina
0
0.05
0.1
0.15
0.2
0.25
Hg µg/g dw
Figure C. Metals in winkles Littorina littorea, µg g-1 dw. Values below the
lower quartile value (lowest 25%) of values in MBA UK data base are
plotted as green bars and red if above the upper quartile (highest 25%).
Values in the mid range (25-75th percentile) are represented as grey bars.
(cont)…..
46

75 km.
Mn Littorina
0
50
100
150
200
250
Mn µg/g dw
75 km.
Ni Littorina
0
1
2
3
4
5
Ni µg/g dw
75 km.
Pb Littorina
0
1
2
3
4
5
Pb µg/g dw
75 km.
Se Littorina
0
0.5
1
1.5
2
2.5
Se µg/g dw
75 km.
Sn Littorina
0
0.02
0.04
0.06
0.08
0.1
Sn µg/g dw
75 km.
Zn Li ttori na
0
20
40
60
80
100
Zn µg/g dw
…Figure C (cont.). Metals in winkles Littorina littorea, µg g-1 dw. Values
below the lower quartile value (lowest 25%) of values in MBA UK data base
are plotted as green bars and red if above the upper quartile (highest
25%). Values in the mid range (25-75th percentile) are represented as grey
bars.
47

75 km.
Ag Fucus
0
0.1
0.2
0.3
0.4
0.5
Ag µg/g dw
75 km.
As Fucus
0
20
40
60
80
100
As µg/g dw
75 km.
Cd Fucus
0
0.2
0.4
0.6
0.8
1
Cd µg/g dw
75 km.
Co Fuc us
0
2
4
6
8
10
Co µg/g dw
75 km.
Cr Fuc us
0
0.5
1
1.5
2
2.5
Cr µg/ g dw
75 km.
Cu Fucus
0
2
4
6
8
10
Cu µg/g dw
75 km.
Fe Fucus
0
150
300
450
600
750
Fe µg/g dw
75 km.
Hg Fucus
0
0.02
0.04
0.06
0.08
0.1
Hg µg/g dw
Figure D. Metals in seaweed Fucus vesiculosus, µg g-1 dw. Values below the
lower quartile value (lowest 25%) of values in MBA UK data base are
plotted as green bars and red if above the upper quartile (highest 25%).
Values in the mid range (25-75th percentile) are represented as grey bars.
(cont.)…..
48

75 km.
Mn Fucus
0
200
400
600
800
1000
Mn µg/g dw
75 km.
Ni Fuc us
0
2
4
6
8
10
Ni µg/g dw
75 km.
Pb Fucus
0
0.5
1
1.5
2
2.5
Pb µg/g dw
75 km.
Se Fuc us
0
0.5
1
1.5
2
2.5
Se µg/g dw
75 km.
Sn Fucus
0
0.2
0.4
0.6
0.8
1
Sn µg/g dw
75 km.
Zn Fucus
0
50
100
150
200
250
Zn µg/g dw
…Figure D (cont.). Metals in seaweed Fucus vesiculosus, µg g-1 dw. Values
below the lower quartile value (lowest 25%) of values in MBA UK data base
are plotted as green bars and red if above the upper quartile (highest
25%). Values in the mid range (25-75th percentile) are represented as grey
bars.
75 km.
TBT Nereis
0
10
20
30
40
50
TBT µg/kg dw
75 km.
TPT Nereis
0
10
20
30
40
50
TPT µg/kg dw
Figure E. TBT and TPT concentrations (µg kg-1 dw) in ragworm Nereis
diversicolor
49

Table A Background Concentrations (BC) and Background Assessment
Criteria (BAC) for PAHs in mussels (2004/5 data; OSPAR, 2007)
(
µ
g kg-1 dry weight)
BC
BAC
Naphthalene
1
81.2
Phenanthrene
4.5
12.6
Anthracene
1
2.7
Fluoranthene
7
11.2
Pyrene
5.5
10.1
Benz[a]anthracene
1.5
3.6
Chrysene
6.5
21.8
Benzo[a]pyrene
1
2.1
Benzo[ghi]perylene
2.5
7.2
Indeno[123-cd]pyrene
2
5.5
50

Table B. Organotins (µg kg-1) in Nereis diversicolor, September 2007 and Mytilus edulis, March 2008.
Organotins µg kg-1 (dry weight)
Site
Map Ref.
Mouth
Date
MBT
DBT
TBT
MPT
DPT
TPT
Nereis diversicolor
Tywia
SN370117
-
11/09/2007
64.03
12.22
18.48
28.72
31.53
6.47
Dalea
SM815075
6.5
13/09/2007
41.76
15.84
30.72
35.81
24.02
5.02
Angle
a
SM868028
7.4
12/09/2007
166.26
104.77
33.72
44.58
30.93
12.04
Pennar
SM959020
16.8
12/09/2007
42.20
15.73
13.85
37.86
13.38
5.13
Pembroke Ferry (Waterloo)
SM982040
19.0
12/09/2007
88.91
35.41
19.24
163.65
115.60
36.58
Lawrenny
b
SN017063
21.8
13/09/2007
60.15
31.28
9.04
38.54
19.51
5.51
Landshipping
SN011118
28.0
13/09/2007
68.12
39.85
15.41
48.73
180.69
7.14
Mytilus edulis
Tywib
SN361082
-
09/03/2008
54
2
<1
<1
0.66
0.27
Dale
b
SM809065
5.5
11/03/2008
454
14
62
<1
<1
1.26
Angleb
SM870027
7.3
10/03/2008
79
13
59
<1
0.9
1.71
Pennar Mouth
SM943028
13.8
10/03/2008
106
15
72
<1
<1
1.48
Pennar
SM959020
16.8
10/03/2008
82
13
78
<1
0.93
2.64
Pembroke Ferry (Ferry Inn)
SM974047
17.0
10/03/2008
104
34
123
<1
<1
1.49
Ferry Hill
SN003061
21.0
11/03/2008
104
19
103
<1
0.83
3.07
Lawrenny
a
SN009062
21.0
11/03/2008
336
29
100
<1
0.73
2.73
Black Tar
SM999093
24.8
11/03/2008
129
17
83
<1
<1
2.95
< in tables denote values below Limit of Detection.
a,b ‘Soft substrate’ sites may be slightly different from ‘rocky shore habitats’ : see Appendix 1 for site details
51

Table C PCBs (µg kg-1) in Nereis diversicolor, September 2007 and Mytilus edulis, March 2008.
a,b ‘Soft substrate’ sites may be slightly different from ‘rocky shore habitats’ : see Appendix 1 for site details
Site
Map Ref.
Mouth
Date
PCB
028
PCB
052
PCB
101
PCB
118
PCB
153
PCB
138
PCB
180
Total of
7 PCBs
Nereis diversicolor
Tywia
SN370117
-
11/09/2007
1.044
1.181
0.000
0.000
3.111
1.708
0.000
7.044
Dalea
SM815075
6.5
13/09/2007
2.022
0.608
0.456
0.000
2.266
1.234
0.000
6.586
Anglea
SM868028
7.4
12/09/2007
2.930
0.578
0.364
0.726
3.211
1.551
1.686
11.046
Pennar
SM959020
16.8
12/09/2007
3.336
0.660
0.434
0.623
1.751
0.943
0.705
8.452
Pembroke Ferry (Waterloo)
SM982040
19.0
12/09/2007
2.900
0.493
0.000
0.915
3.279
1.686
0.966
10.239
Lawrennyb
SN017063
21.8
13/09/2007
2.648
0.568
0.491
0.776
2.167
1.255
0.974
8.878
Landshipping
SN011118
28.0
13/09/2007
1.579
0.816
0.000
0.625
3.662
1.601
1.253
9.535
Mytilus edulis
Tywib
SN361082
-
09/03/2008
0.272
0.558
1.250
1.330
7.406
2.841
0.030
13.688
Daleb
SM809065
5.5
11/03/2008
0.633
0.525
11.452
16.019
7.608
2.873
0.382
39.492
Angleb
SM870027
7.3
10/03/2008
3.013
0.551
2.677
2.867
23.188
7.364
0.322
39.983
Pennar Mouth
SM943028
13.8
10/03/2008
0.000
0.000
2.496
2.620
10.836
4.039
0.312
20.302
Pennar
SM959020
16.8
10/03/2008
-
-
-
-
-
-
-
-
Pembroke Ferry (Ferry Inn)
SM974047
17.0
10/03/2008
0.498
0.735
2.376
2.584
11.557
3.747
0.561
22.057
Ferry Hill
SN003061
21.0
11/03/2008
0.247
0.261
1.628
1.699
7.781
2.395
0.449
14.459
Lawrenny
a
SN009062
21.0
11/03/2008
0.504
0.493
1.441
2.192
8.872
2.523
0.052
16.077
Black Tar
SM999093
24.8
11/03/2008
0.545
0.332
1.018
1.604
4.346
1.631
0.002
9.479
52

Table D. PAHs (µg kg-1) in Nereis diversicolor, September 2007 and Mytilus edulis, March 2008.
Site
Naph.
1-Me-
naph.
2-Me naph
Ace.
Fluor.
Phen.
Anth.
Fanth.
Pyr.
B(a) A
Chrys.
B(b)F
Pery.
B(k)F
B(a)P
DB(ah)
A
B(ghi)
P
Ind(123
cd)P
Total
PAHs
Nereis diversicolor (September 2007)
Tywia
0.97
1.26
NA
0.94
4.54
49.47
1.96
14.91
5.86
1.27
2.48
2.65
0.67
0.24
0.40
< 0.06
0.61
< 0.07
88.3
Dalea
3.51
1.91
NA
2.07
4.22
55.02
3.23
65.09
66.82
3.72
4.33
5.47
< 0.22
0.67
1.48
< 0.18
0.33
< 0.12
218.4
Anglea
5.04
2.19
NA
0.88
4.64
45.39
2.10
11.72
7.20
1.97
2.66
4.12
< 0.25
0.19
0.76
< 0.75
1.80
< 0.13
91.8
Pennar
2.95
< 0.99
NA
0.51
3.51
39.79
2.18
8.12
3.89
1.20
1.46
1.39
< 0.05
< 0.33
0.53
< 0.06
0.37
< 0.13
67.4
Pembroke Ferry (Waterloo)
11.65
2.21
NA
< 0.30
3.92
35.71
1.48
10.20
4.58
1.62
1.62
2.61
< 0.16
< 0.03
0.65
< 0.29
1.12
< 0.15
78.3
Lawrennyb
7.09
4.62
NA
< 0.14
3.74
35.21
1.40
16.65
15.08
1.19
1.96
3.62
< 0.05
1.58
1.60
< 0.06
0.85
< 0.03
94.6
Landshipping
3.35
2.00
NA
0.78
4.26
39.94
2.07
10.16
5.97
1.31
1.41
2.15
< 0.05
< 0.01
0.29
< 0.06
0.30
< 0.06
74.2
Mytilus edulis (March 2008)
Tywi
b
4.26
<0.77
0.86
<0.40
2.00
20.6
<0.00
25.4
24.1
2.80
5.23
9.13
4.39
3.08
1.00
<0.00
5.29
1.01
110.4
Daleb
7.10
0.92
2.10
<0.40
1.39
16.3
0.70
16.1
19.6
9.92
12.6
17.1
6.26
4.70
1.89
<0.00
7.92
2.57
127.7
Angleb
3.62
1.86
1.63
<0.25
2.15
23.1
1.09
27.9
33.6
12.5
26.8
34.2
12.0
8.08
3.74
<0.03
13.5
3.80
209.8
Pennar Mouth
6.34
3.22
2.62
0.84
1.44
18.1
0.31
23.8
38.8
25.5
29.9
43.3
14.3
11.4
5.59
<0.02
14.2
4.30
244.0
Pennar
4.74
2.89
1.83
<0.35
1.35
15.5
0.79
16.9
23.1
17.8
12.6
25.0
12.8
7.01
3.28
<0.08
11.4
1.93
159.3
Pembroke Ferry (Ferry Inn)
4.13
9.29
2.01
0.62
1.75
19.8
1.02
29.2
33.5
30.7
27.4
44.2
16.2
12.2
9.58
1.55
17.6
1.59
262.3
Ferry Hill
7.54
2.52
2.27
0.52
1.80
17.7
0.92
14.5
25.2
20.4
11.3
28.8
11.7
8.46
6.30
<0.82
12.7
4.66
178.1
Lawrennya
3.84
1.28
1.59
0.68
1.54
13.4
1.21
17.5
24.6
7.55
11.2
29.2
14.2
8.22
5.92
1.25
15.4
3.70
162.3
Black Tar
5.46
2.85
1.74
<0.15
1.32
11.2
0.70
10.4
14.5
6.62
6.32
28.7
13.6
9.15
6.54
<0.92
16.9
3.45
140.7
< in tables denote values below Limit of Detection. NA not analysed
a,b ‘Soft substrate’ sitesa for Nereis may be slightly different from rocky shore habitatsb for Mytilus :
53