Assessment of Heavy Metal and PAH Contamination of Urban Streambed Sediments on Macroinvertebrates

ASSESSMENT OF HEAVY METAL AND PAH CONTAMINATION OFURBAN STREAMBED SEDIMENTS ON MACROINVERTEBRATES

GARY BEASLEY and PAULINE E. KNEALE∗School of Geography, University of Leeds, Leeds LS2 9JT, U.K.

(∗ author for correspondence, e-mail: [email protected]; phone: 0113 343 3340;fax: 0113 3433308)

(Received 20 August 2002; accepted 12 April 2003)

Abstract. The results from measuring PAH and metal contamination together with macroinverteb-rate communities at 62 headwater stream sites gives a significant insight into the range and scale ofcontamination. Monitoring streambed sediments at 62 sites from rural to inner city and in industriallocations presented a unique opportunity to distinguish the conditions that enhance pollution runoffat sites that are less obviously ‘at risk’ and to compare these results with sites of expected highcontamination, for example in industrial areas and at motorway junctions. We used pCCA (partialCanonical Correspondence Analysis) to tease out the relationships between individual macroinver-tebrate families and specific metal and PAH contaminants, and showed that it is not always themetals and PAHs with the greatest total concentrations that are doing the damage to the ecology. Niand Zn are the critical metals, while benzo(b)fluoranthene, anthracene and fluoranthene are the mostcontaminating PAHs. The results identify previously unrecognized ‘high risk’ pollution sources, laybyes used for commercial parking, on-street residential parking areas, and the junctions at the bottomof hills with traffic lights, where surface runoff feeds rapidly to the streams. While this study looksat sites across Yorkshire, UK, it clearly has a broader significance for understanding contaminationrisks from diffuse runoff as a prerequisite for effective sustainable urban drainage system (SUDS)agendas and the protection of urban stream ecology.

Keywords: diffuse urban runoff, macroinvertebrates, metal contamination, PAHs, stream bed sedi-ments

1. Introduction

Despite efforts to control water quality through targeting point sources of majorcontaminants, many rivers and streams experience biological quality below thatsuggested by their environmental characteristics. Streambed sediments can be ex-pected to accumulate have metal- and oil-based contamination (polycyclic aromatichydrocarbons, PAHs) in urban areas where runoff is associated with vehicle traf-ficking, pavements, roofs, guttering and industry (Marsalek et al., 1999; van Metreet al., 2000). Metals and oils preferentially attach to fine particles (Estèbe et al.,1997; Lee et al., 1997), which are entrained in surface runoff and deposited as stre-ambed sediments. The sediment contamination is often a magnitude greater thanin the overlying water column (Power and Chapman, 1992). These contaminantsexert a persistent and wide-reaching stress on the freshwater ecosystem leading to

Water, Air, and Soil Pollution: Focus 4: 563–578, 2004.© 2004 Kluwer Academic Publishers. Printed in the Netherlands.

564 G. BEASLY AND P.E. KNEALE

the impairment of tolerant species and the disappearance of the sensitive macroin-vertebrate species and, through their accumulation, damage species in the highertrophic levels (Field and Pitt, 1990; Beyer et al., 2000). The lack of a detailedknowledge of these interrelationships, and therefore the ability to identify diffusecontamination risks, has critical implications for the intelligent implementation ofsustainable urban drainage system (SUDS) agendas (Environment Agency, 2000;D’Arcy and Frost, 2001; Harremoes, 2002).

This study reports the empirical results of monitoring the ecology and sedimentchemistry in a significant number of streams that receive only diffuse urban runofffrom a variety of surfaces. The analysis compares the field results with RIVPACSforecasts for pristine streams (Wright, 2000) and the risks of each heavy metaland PAH to individual macroinvertebrate families are isolated using pCCA (partialCanonical Correspondence Analysis).

2. Methodology

A detailed examination of surface sewer network maps and verification from fieldinvestigations identified 62 sites on 27 first order streams in west and southYorkshire, United Kingdom, for sampling in May and September 1999 (Figure 1;Beasley, 2001; Beasley and Kneale, 2002). In this paper, the September results arediscussed. Sediment and macroinvertebrate samples were collected 25 m above andbelow surface storm water inflows in rural, residential, industrial, and motorwaysubcatchments. In the analysis, sites were subjectively arranged in order of hypo-thesised increasing contaminant risk, rural through to motorway land use. In orderto ensure comparability with data for the Environment Agency (EA) RIVPACSsystem model, site data waere collected on each visit using EA procedures (En-vironment Agency, 1997). At each site, data were obtained for: altitude, distancefrom source, slope, stream width, depth, discharge class, percentage particle coverfrom silt to boulders, dissolved oxygen, electrical conductivity, and pH. Similarly,sediment and macroinvertebrate sampling followed EA protocols.

2.1. STREAM SEDIMENT COLLECTION

Representative triplicate random samples were taken from the upper 0–5 cm layerand composited to minimise heterogeneity in heavy metal concentrations and samp-ling variance (Argyraki et al., 1995). Trials showed that sediment collected fromthe near surface (0–5 cm) with a plastic trowel caused minimum disturbance, lowrisk of cross contamination, and retention of the finest particles. The sediments formetal determinations were placed in airtight, zip-sealed polythene bags, and doublebagged. The same field procedures were implemented for the collection of PAHsediment samples, which were packed tightly into 500 mL glass bottles with glassstoppers and covered with aluminium foil to deter light penetration (Greenberg et

CONTAMINATION IN URBAN STREAMBED SEDIMENTS 565

Figure 1. Site locations, Yorkshire, UK.

al., 1992). Samples were refrigerated at 4 ◦C, not frozen, as PAH determinationswith frozen sediments have been linked to reduction in concentrations of up to 99%(Fox et al., 1991).

2.2. MACROINVERTEBRATE COLLECTION

Standard macroinvertebrate collection involved three 1-min kick samples, eachfrom 1 m2 of streambed. Samples were decanted into 1100 mL polypropylenebottles with just enough water to maintain dampness, which reduced damage and


retarded carnivore activity during transportation. They were preserved using 95%ethanol, sorted for identification to family level, and recorded using EA auditsheets. Examples of each taxon were placed in vials containing ethanol for qualityassurance checking with EA scientists. The abundance of each family was recordedusing EA audit sheets.

2.3. LABORATORY ANALYSIS

2.3.1. PAHsAir-dried PAH samples were ground, passed through a riffle box to ensure homo-genisation and reground to pass a 212 :m aperture sieve. 3 g of sieved sediment wasextracted with 15 mL of dichloromethane in glass vials spiked with d14-Terphenylfor 16 hr on a mechanical shaker. The extract was analysed by gas chromatographymass spectrometry (GCMS) using a Varian 3800 series gas chromatograph, con-nected to a Saturn 2000 ion trap mass spectrometer. The GC oven was programmedfrom 50 ◦C (held for 1 min) to 290 ◦C at a rate of 10 ◦C min−1. Temperature washeld constant at 290 ◦C for 6 min (total run time 31 min). The injector was held at aconstant temperature of 300 ◦C with a split ratio of 50:1. The carrier gas was heliumsupplied at a constant flow rate of 1.5 mL min−1 for 6 min. Duplicate samples weretaken for quality control in conjunction with a CRM (Coal carbonised site soil LGC6138).

2.3.2. Heavy metalsFollowing standard drying processes (Mudroch and MacKnight, 1994) subsam-ples were taken by coning and quartering, and then ground lightly using an agatemortar and pestle. Each subsample was passed through a 2000 microns syntheticpolymer woven screen to minimise contamination (Mudroch and Azcue, 1995),while retaining only the sand and silt-sized particles. Analytical-grade (AnalaR)acids were used for all extraction solutions and cleaning procedures. The metals(Cd, Cr, Cu, Fe, Pb, Ni and Zn) were extracted from 0.5 g of each size fraction in50 mL PTFE vials using a three-step sequential extraction technique (Quevellier,1997). This method identifies the metals from the three geochemical phases; ex-changeable, reducible and oxidisable (Beasley, 2001). The supernatant producedafter extraction was acidified to pH 2 to prevent any adsorption or precipitationof metals. Metal concentrations were determined using a Jarrell-Ash InductivelyCoupled Plasma Optical Emission Spectrometer (IAP-AES). Duplicate samples,blanks and CRM 601 were incorporated in each run for quality assurance.

2.4. ANALYSIS

2.4.1. RIVPACSComparison of observed stream biological quality with that expected assuming nocontamination assists in identifying the severity of contamination and ecologicalrisks between and within land uses. Therefore, the predicted biological quality at


each site was forecast with the EA RIVPACS model (Wright, 2000). The RIVPACSmodel, (Wright, 2000) uses river and catchment environmental characteristic datato forecast the expected Biological Monitoring Working Party (BMWP) biotic in-dex under non-polluted conditions, providing a baseline comparison with measuredresults. The model uses data collected at each site for altitude, distance from source,slope, stream width, depth, discharge class, percentage particle cover from silt toboulders, dissolved oxygen, electrical conductivity, and pH.

2.4.2. pCCAPartial canonical correspondence analysis (pCCA) used here followed the methodof ter Braak (1987, 1994) to determine the relative importance of contaminants(bioavailable metals and PAHs) in explaining the variability in the macroinver-tebrate community composition. In order to do so, the explanatory variables weresubdivided into a set of covariables and a set of variables-of-interest. The covari-ables represent habitat variables are not the prime focus of the research and assuch did not enter the synthetic gradients (ter Braak and Verdonschot, 1995). Thevariables-of-interest used to construct the synthetic gradients are the streambedsediment heavy metal and PAH concentrations. With the covariables representinggradients that are already extracted having been partialled out (ter Braak, 1996),the ordination diagrams display the unimodal relationships (optimum abundance)between macroinvertebrates, heavy metals and PAHs.

The analysis was carried out using the programme CANOCO 4.0 following thedata set conversion into CANOCO 4.0 format using the utility program CanoImp(Beasley, 2001). In Run 1, the first extraction, bioavailable metals were incorpor-ated against macroinvertebrates recorded for September. In Run 2, the influence ofPAHs on the macroinvertebrates was assessed. Essentially, the data are displayedas rankings (from the ‘forward selection’ option in CANOCO; Table I) and ex-plored further using ordination diagrams. Table I shows the top 10 ranked variablesthat explain community composition by metals and PAHs. Interpretation of theordination diagrams facilitates the ranking of families in relation to each elementidentifying those that are tolerant and sensitive (Tables II and III).

On ordination diagrams, the variables with long arrows are more strongly re-lated to the pattern of variation in species composition than those with short arrows.Families whose points are projected close to the arrow tips are largely restrictedto streams with those characteristics and vice versa for families whose endpointsproject onto the lower end of the arrow.

3. Results and Discussion

As has been discussed elsewhere (Beasley, 2001; Beasley and Kneale, 2002) theresults support the general hypothesis that as suburbanisation increases, the con-centrations of metals and PAHs can be expected to rise, agreeing with Andoh,


TABLE I

Ranked pCCA results for two runs using unrestricted Monte Carlo significance test (p < 0.05)

pCCA RUN 1 pCCA RUN 2

Top 10 rankings for weighted bioavail-able metal concentrations and waterchemistry variables

Top 10 rankings for PAHs and waterchemistry variables

Variable F p Variable F p

Ni 2.01 0.020* Benzo(b)fluoranthene 2.93 0.005*

Zn 1.50 0.045* Electrical conductivity 1.85 0.030*

PH 1.31 n.s Anthracene 1.61 0.035*

Pb 1.05 n.s pH 1.61 0.030*

Cu 0.91 n.s Fluoranthene 1.57 n.s

Electrical conductivity 0.92 n.s Napthalene 1.45 n.s

Dissolved oxygen 1.00 n.s Dibenz(a,h)anthracene 1.17 n.s

Cd 1.20 n.s Indeno(1,2,3-cd)pyrene

1.02 n.s

Cr 0.66 n.s Acenaphthene 0.89 n.s

Fe 0.53 n.s Fluorene 0.93 n.s

(1994), Ellis et al. (1997), Marsalek et al. (1999), and Lee and Bang, (2000).Careful analysis of the metals data indicates that motorway runoff is not alwaysthe worst culprit (Figure 2; n.b. the sites were subjectively arranged in order ofhypothesised increasing contaminant risk, rural through to motorway land use).Zn contamination is found at several of the residential sites, most noticeably site16 (and corresponding downstream sites 18 and 19) and sites 20–30. These siteswere all found to have serious traffic-related sources. Zinc is the major metal invehicle tires. It is postulated that the high sediment Zn levels are the result ofrunoff from heavily trafficked, steep ‘A’ roads where deceleration causes tires towear with concurrent increased release of Zn, agreeing with the findings of Kim etal. (1998) and Draper et al. (2000). Sites 59 and 60 have the highest concentrationsoverall because runoff derives from a major junction roundabout and motorwayexit lanes that are hot spots for vehicle braking. Sites 40 and 41 receive drainagefrom an industrial outlet where heavy trafficking by lorries and where tight turningrestrictions induces tire wear, which is thought to account for the enhanced Znlevels, but some building-related contamination may occur here as well.

Site 54 showed very high Cu concentrations in September 1999, although notin the previous May, and further investigation suggests that this was due to spillageor illegal dumping. The persistence of such short-term variations on catchmentsediment chemistry requires further investigation. The patterns between May and


September vary, but not very substantially, the overall patterns being consistent andof more significance than the fluctuations.

The PAH results also showed an increasing concentration gradient from ruralthrough residential to industrial land uses supporting the link with vehicles foundby Maltby et al. (1995) and van Metre et al. (2000). However, results from thisstudy show clearly that this is not always true and depends on the individual ele-ment (Beasley, 2001; Beasley and Kneale, 2002). Napthalene does appear to ad-here to this general pattern; the highest concentrations were recorded in streamsediments below industrial and motorway inflows (Figure 2). However, certainresidential subcatchments, for example sites 46 and 47, posed a greater contam-ination risk than either the industrial or motorway source areas. The reason liesin specific catchment characteristics. These residential areas with roadside parkingaccumulate higher concentrations of total PAHs largely because of high concen-trations of fluoranthene, phenanthrene and pyrene (Figure 3), which originate fromcrankcase oil drippings leaked onto the road surface (Latimer et al., 1990). Sup-porting evidence for leaked oil rather than traffic volume being responsible for highconcentrations of PAHs derives from sites 45 and 59 that drain heavily utilisedlay-byes.

Understanding contamination risk is further complicated by the stream or rivercharacteristics into which the runoff drains. If the stream has a silt-laden stre-ambed due to low flow allowing fine particles to settle, it would be expected thatcontaminant concentrations would be greater than a gravel bed stream receivingidentical contaminant input. Differences in stream characteristics have undoubtedlyinfluenced the levels of contamination in this study and one would presume theecological quality. It is hypothesised that a silt laden, shallow, slow flowing streamwould support a relatively depauperate macroinvertebrate community structure.However, the characteristics of the streams in this investigation are all shown to becapable of supporting a ‘good’ biological quality, as evidenced by in the RIVPACSsmodel forecasts. The RIVPACs forecast incorporates detailed information aboutaltitude, geology, stream bed sediment size, and channel characteristics (Wright,2000). Actual biological quality is inversely related to the observed contaminationlevels, with a general reduction in macroinvertebrate diversity as contaminationincreases. The predicted RIVPACS, scores range from 102–155 and always exceedthe actual scores. Actual BMWP values ranged from 25–95 except for one score of123 in the September samples.

3.1. TOTAL METAL CONTAMINATION

Evaluating the toxicity of the concentrations of the sediment associated metals islimited by the absence of UK or European standards, but we can compare the res-ults with the Ontario Ministry of Environment (OME) total metal sediment toxicityguidelines (Persaud et al., 1989) (Figure 2). For each metal there are some sitesthat exceed the critical thresholds, but the largest number of excedences were for


Figure 2. Total metal sediment toxicity compared with the OME toxicity guidelines.


Figure 3. Streambed sediment concentrations of three PAHs.

Pb and Zn. This classifies the sediments as potentially toxic to macroinvertebratesand fish. The association of Pb and Zn from vehicle emissions and Zn from vehicletires can be linked to the more contaminated and heavily trafficked sites, such asat site 16 and the motorway sites 59, 60 and 62, all of which exceed the toxicityguidelines. Extrapolating Canadian sediment guidelines to Yorkshire streams mustbe done cautiously, but is nonetheless indicative of problems in headwaters thathave been expected to be much cleaner. It would, of course, be helpful to have UKsediment chemistry toxicity guidelines that are based on bioavailable rather thantotal metals. Such guidelines would be more sensitive than those based on totalmetal extractions, although their application would still require sensitivity to thelocal hydrological and catchment circumstances.


TABLE II

Families determined as tolerant or sensitive to metals and water chemistry variables, Run 1

Variable Five most tolerant families(tolerance ranked left to right)

Five most sensitive families(sensitivity ranked left to right)

Ni Phil, Perl, Chlo, Rhya, Hept Plan, Ephemeri, Valv, Hydrob,Dyti

Zn Chlo, Phys, Hali, Hydroph, Rhya Hydrom, Plan, Nemo, Ephemeri,Ephemere

pH Plan, Ephemeri, Valv, Phys, Hali Perl, Phil, Hydrom, Leuc, Hept

Pb Chlo, Phys, Hali, Hydroph, Rhya Hydrom, Nemo, Plan, Ephemeri,Ephemere

Cu Chlo, Phys, Hali, Hydroph, Rhya Hydrom, Plan, Nemo, Ephemeri,Ephemere

Electrical conductivity Chlo, Phys, Hali, Plan, Lymna Perl, Phil, Hydrom, Nemo, Leuc

Dissolved oxygen Plan, Ephemeri, Valv, Hydrob,Dyti

Phil, Perl, Chlo, Rhya Hept

Cadmium Chlo, Phys, Hali, Rhya, Hydroph Hydrom, Plan, Ephemeri, Nemo,Ephemere

Cr Chlo, Phys, Rhya, Hydroph, Hali Hydrom, Plan, Ephemeri, Nemo,Ephemere

Fe Phil, Perl, Chlo, Rhya, Hydroph Plan, Ephemeri, Hydrom, Valv,Limne

Asel, Asellidae; Chlo, Chloroperlidae; Dyti, Dytiscidae, Ephemere, Ephemerellidae;Ephemeri, Ephemeridae; Erpo, Erpobdellidae; Hali, Haliplidae; Hept, Heptageniidae; Hydrob,Hydrobiidae; Hydrom, Hydrometridae; Hydroph, Hydrophilidae; Leptoc, Leptoceridae; Leuc,Leuctridae; Lymna, Lymnaeidae; Nemo, Nemouridae; Odon, Odontoceridae; Perl Perlodidae;Phil, Philopotamidae; Phys, Physidae; Plan, Planorbidae; Rhya, Rhyacophilidae; Seri,Sericostomatidae; Simu, Simulidae; Spha, Sphaeriidae; Valv, Valvatidae

3.2. BIOAVAILABLE METAL CONTAMINATION

In the pCCA analyses, the bioavailable rather than total metal data were used,showing that Zn and Ni were consistently the most significant metals in determ-ining macroinvertebrate community structures (Table II). Although total concen-trations of Ni were much lower than for example Cu or Pb, Ni is more readilybioavailable and it is this understanding of metal speciation that will assist in thelonger term in assigning appropriate catchment mitigation measures. AlthoughCu and Pb are found in higher total concentrations, their bioavailable concen-trations are insufficient to pose a toxic threat in these streams. By targeting thesource areas of those metals with known high bioavailabilities and establishingbioavailability toxicity thresholds, concerted inroads could be made into reducingcontaminant-induced ecological stress.


TABLE III

Families determined as tolerant or sensitive to metals and water chemistry variables, Run 2

Variable Five most tolerant families(tolerance ranked left to right)

Five most sensitive families(sensitivity ranked left to right)

Benzo(b)fluoranthene Asel, Valv, Ephemere, Spha, Hy-drob

Hydrom, Leptoc, Phil, Odon,Hept

Electrical conductivity Hydroph, Ephemere, Asel, Spha,Phys

Leptoc, Hydrom, Phil, Ephemeri,Hali

Anthracene Valv, Asel, Spha, Phys, Hydrob Phil, Hydrom, Odon, Hept,Leptoc

pH Valv, Hydrob, Asel, Spha, Phys Hydrom, Phil, Odon, Hept, Leuc

Fluoranthene Asel, Valv, Ephemere, Spha, Hy-drob

Hydrom, Leptoc, Phil, Odon,Hept

Naphthalene Hydroph, Ephemere, Asel, Perl,Spha

Leptoc, Ephemeri, Hydrom, Phil,Leptop

Dibenz(a,h,)anthracene Hydroph, Ephemere, Asel, Spha,Phys

Leptoc, Leptop, Hali, Ephemeri,Seri

Indeno(1,2,3-cd)pyrene

Valv, Asel, Spha, Phys, Hydrob Hydrom, Phil, Odon, Hept,Leptoc

Acenaphthene Valv, Asel, Spha, Phys, Eph-emere

Hydrom, Phil, Leptoc, Odon,Hept

Fluorene Valv, Asel, Spha, Phys, Eph-emere

Hydrom, Phil, Leptoc, Odon,Hept

Run 2 (Table II) shows that it is not only those PAHs that are found in the highestconcentrations that have the greatest impact on stream ecology. The group repres-ented by fluoranthene was ranked fifth despite being the most prevalent in terms ofconcentration. The most significant PAHs are those represented by benzo(b)fluo-ranthene, which although present in lower concentrations, have carcinogenic im-pacts on macroinvertebrates. The pCCA indicate that it is not total contaminationthat matters in terms of either the metals or the PAHs, the stresses on streamecology are more subtle.

3.3. INDICATOR FAMILIES

The pCCA allows us to tease out the more detailed relationships between thecontaminants and macroinvertebrate families (Figures 4 and 5). The literature sug-gests metal contamination leads to reduction in specific Ephemeroptera (mayfly)families (Kiffney and Clements, 1994; Gower et al., 1994; Clements et al., 2000),and their replacement by chironomids and oligacheates. This effect is confirmedin these Yorkshire streams, particularly by the absence of Leptoceridae and Eph-emerellidae. The widespread distribution of Baetidae and its central position in


Figure 4. Species – Environment biplot based on pCCA Run 1. Families are represented by crosses.

the species-environment ordination diagrams supports the findings of Gower et al.,(1994) in that is has moderate metal tolerance. Similarly, there is variation in thetolerance of Plecoptera (stonefly) families, Leuctridae being slightly more tolerantthan Nemouridae, whereas Chloroperlidae appear here to be particularly tolerantto metals. Tricoptera (caddis fly) families display slightly greater tolerances, espe-cially Polycentropodidae, Rhyacophilidae, and Hydropsychidae, with the latter twofamilies preferring the sites with high dissolved oxygen concentrations. The casedcaddis flies, Limnephilidae and Sericostomatidae are more sensitive. Asellidae areseen here to be moderately tolerant to a number of heavy metals.

The rankings (Table II) generated from the ordination plot (Figure 3) uncoverseveral indicator families that were at risk from bioavailable metal pollution, not-ably Ephemerellidae, Ephemeridae, Nemouridae, and Planorbidae. Looking atthe risk from PAH, Hydrometridae and Ephemeridae were most at risk. Lepto-ceridae, Philopotamida, Ondontoceridae, and Heptageniidae were not as sensitive


Figure 5. Species – Environmental biplot based on pCCA Run 2. Families are represented by crosses.

to metals, but were particularly sensitive to enhanced PAH levels (Table IV). In thePlecoptera (stonefly) families, the Leptoceridae were particularly sensitive, muchmore so than the Nemouridae, Chloroperlidae and Perlodidae. Of the Tricopteralarva (caddis fly), only the caseless Polycentropodidae exhibited moderate toler-ance towards PAHs (Figure 5). The Ephemerellidae, although sensitive to metals,were shown to tolerate PAH pollution.

4. Conclusions

The SUDS (sustainable urban drainage system) agenda indicates that runoff man-agement should be targeted to mitigate contamination of streams so that the eco-logical quality is retained or restored. Identifying streams, or sections of streams,at risk is complicated by local circumstances. Whereas previous studies have con-centrated on a priori ‘worst case’ sites, this study has looked at a suite of sites. The


results from 62 sites are inevitably noisy, but indicative. They provide a broadercontext for management consideration. Surface runoff from lay byes used for com-mercial goods parking and on-street parking in residential areas were shown togenerate more PAH contamination than a motorway slip road. Runoff from a junc-tion at the foot of a hill with traffic lights, where braking is hard and regular, was abigger problem than runoff from roads elsewhere. Overnight on-street parking ledto highly contaminated runoff from the street surfaces.

We have shown that looking at ‘total’ metals or ‘total’ PAH results can be verymisleading. Bioavailable metals are much more useful and the contaminants thathave the greatest abundance do not necessarily cause the most biological impair-ment. The pCCA results pull out the greater importance of bioavalable Ni and Zn,although their total concentrations are less than for Pb, and the importance of thebenzo(b)fluoranthene and anthracene groups.

Catchments are heterogeneous, so catchment-specific characteristics are crucialto our understanding. These results suggest that risk management would benefitfrom: 1) controlling runoff from high hydrocarbon collection areas, such as parkingareas, regardless of their size or traffic flow data; 2) identifying catchment sourceareas for bioavailable metals, such as Ni and Zn, from road junctions, roundabouts,exit lanes, and roads with steep gradients, and 3) developing bioavailable metal andPAH sediment toxicity guidelines and thresholds.

References

Andoh, R. Y. G.: 1994, ‘Urban runoff: Nature, characteristics and control’, J. Inst. Water Eng.Managers 8, 371–378.

Argyraki A., Ramsey M. H. and Thompson M.: 1995, ‘Proficiency testing in sampling: Pilot studyon contaminated land’, Analyst 120, 2799–2803.

Beasley, G. E.: 2001, ‘Investigating the relations between streambed sediment contaminants andmacroinvertebrate assemblages in urban headwater streams in Yorkshire’, Unpublished PhDthesis, University of Leeds.

Beasley, G. and Kneale, P. E.: 2002, ‘Reviewing the impact of metals and PAHs on macroinverteb-rates in urban watercourses’, Prog. Phys. Geogr. 26, 236–270.

Beyer, W. N., Day, D., Melancon, M. J. and Sileo, L.: 2000, ‘Toxicity of Anacostia River, WashingtonDC, USA, sediment fed to mute swans (Cygnus olor)’, Environ. Toxicol. Chem. 19, 731–735.

Clements, W. H., Carlisle, D. M., Lazorchak, J. M. and Johnson, P. C.: 2000, ‘Heavy metals structurebenthic communities in Colorado mountain streams’, Ecol. Appl. 10, 626–638.

D’Arcy, B. and Frost, A.: 2001, ‘The role of best management practices in alleviating water qualityproblems associated with diffuse pollution’, Sci. Total Environ. 265, 359–367.

Draper, D., Tomlinson, R. and Williams, P.: 2000, ‘Pollutant concentrations in road runoff: SoutheastQueensland case study’, J. Environ. Eng. 126, 313–320.

Ellis, J. B., Revitt, D. M. and Llewellyn, N.: 1997, ‘Transport and the environment: Effects of organicpollutants on water quality’, J. Chartered Inst. Water Environ. Manage. 11, 170–177.

Environment Agency: 1997, Assessing Water Quality – General Quality Assessment (GQA) Schemefor Biology, Environment Agency Fact Sheet, Environment Agency, Bristol.

Environment Agency: 2000, Sustainable Urban Drainage Systems, Environment Agency, Bristol.


Estèbe, A., Boudries, H., Mouchel, J.M. and Thevenot, D.R.: 1997, ‘Urban runoff impacts on par-ticulate metal and hydrocarbon concentrations in the river Seine, suspended solid and sedimenttransport’, Water Sci. Technol. 36, 185–195.

Field, R. and Pitt, R. E.: 1990, ‘Urban storm-induced discharge impacts: US EnvironmentalProtection Agency research program review’, Water Sci. Technol. 22, 1–7.

Fox, M. E., Murphy, T. P., Thiessen, L. A. and Khan, R. M.: 1991, ‘Potentially significant under-estimation of PAHs in contaminated sediments’, Proceedings 7th Eastern Region Conference,Quebec, Canadian Association on Water Pollution Research and Control, pp. 80–86.

Gower, A. M., Myers, G., Kent, M. and Foulkes, M. E.: 1994, ‘Relationships between macroinver-tebrate communities and environmental variables in metal-contaminated streams in south-westEngland’, in D. M. Harper and A. J. D. Ferguson (eds), The Ecological Basis for RiverManagement, John Wiley & Sons, New York, New York, USA, pp. 181–191.

Greenberg, A. E., Clesceri, L. S. and Eaton, A. (eds): 1992, Standard Methods for the Examinationof Water and Wastewater, 18th edition, American Public Health Association, Washington, DC,USA.

Harremoes, P.: 2002, ‘Integrated urban drainage, status and perspectives’, Water Sci. Technol. 45,1–10.

Kiffney, P. M. and Clements, W. H.: 1994, ‘Effects of heavy metals on a macroinvertebrate as-semblage from a Rocky Mountainstream in experimental microcosms’, J. N. Am. Benthol. Soc.13, 511–523.

Kim, K. W., Myung, J. H., Ahn, J. S. and Chon, H. T.: 1998, ‘Heavy metal contamination in dustsand stream sediments in the Taejon area, Korea’, J. Geochem. Explor. 64, 409–419.

Latimer, J. S., Hoffman, E. J., Hoffman G., Fasching, J. L. and Quinn, J. G.: 1990, ‘Sources ofpetroleum hydrocarbons in urban runoff’, Water, Air, Soil Pollut. 52, 1–21.

Lee, D. K., Touray, J. C., Bailif, P. and Ildeonse, J. P.: 1997, ‘Heavy metal contamination of settlingparticles in a retention pond along the A71 motorway in Sologne, France’, Sci. Total Environ.201, 1–15.

Lee, J. H. and Bang, K. W.: 2000, ‘Characterisation of urban stormwater runoff’, Water Res. 34,1773–1780.

Maltby, L., Boxall, A. B. A., Forrow, D. M., Calow, P. and Betton, C. I.: 1995, ‘The effects ofmotorway runoff on freshwater ecosystems: 2. Identifying major toxicants’, Environ. Toxicol.Chem. 14, 1093–1101.

Marsalek, J., Rochfort, Q., Brownlee, B., Mayer, T. and Servos, M.: 1999, ‘An explanatory study ofurban runoff toxicity’, Water Sci. Technol. 39, 33–39.

Mudroch, A. and Azcue, J. M.: 1995, Manual of Aquatic Sediment Sampling, Lewis Publishers, AnnArbor, Michigan, USA.

Mudroch, A. and MacKnight, S. D.: 1994, Handbook of Techniques for Aquatic Sediments Sampling,2nd edition, Lewis Pubishers, Ann Arbor, Michigan, USA.

Persaud, D., Jaagumagi, R. and Hayton, A.: 1989, Development of Provincial Sediment QualityGuidelines, Ontario Ministry of the Environment and Energy, Toronto, Ontario, Canada.

Power, E. A. and Chapman, P. M.: 1992, ‘Assessing sediment quality’, in G. A. Burton, Jr. (ed.),Sediment Toxicity Assessment, Lewis Publishers, Ann Arbor, Michigan, USA, pp. 1–18.

Quevellier, P. H., Rauret, G., Lopez-Sanchez, J. F., Rubio, R., Ure, A. and Muntau, H.: 1997, ‘Thecertification of the EDTA-extractable contents (mass fractions) of Cd, Cr, Ni, Pd and Zn insediment following a three-step sequential extraction procedure – CRM 601’, Office for OfficialPublications of the European Communities, Luxembourg.

ter Braak, C. F. J.: 1987, CANOCO – A FORTRAN Program for Canonical Community Ordination by[partial][detrended][canonical] Correspondence Analysis, Principal Components Analysis andRedundancy Analysis (Version 2.1), DLO-Agricultural Mathematics Group, Wageningen, TheNetherlands.


ter Braak, C. F. J.: 1994, ‘Canonical community ordination. Part I: Basic theory and linear methods’,Ecoscience 1, 127–140.

ter Braak, C. F. J.: 1996, Unimodal Models to Relate Species to Environment, AgriculturalMathematics Group, Wageningen, The Netherlands.

ter braak, C. F. J. and Verdonschot, P. F. M.: 1995, ‘Canonical correspondence analysis and relatedmultivariate methods in aquatic ecology,’ Aquat. Sci. 57, 255–289.

van Metre, P. C., Mahler, B. J. and Furlong, E. T.: 2000, ‘Urban sprawl leaves its PAH signature’,Environ. Sci. Technol. 34, 4064–4070.

Wright, J. F.: 2000, ‘An introduction to RIVPACS’, in J. F. Wright, D. W. Sutcliffe and M. T. Furse(eds), Assessing the Biological Quality of Fresh Waters: RIVPACS and other Techniques’, TheFreshwater Biological Association, Ambleside, pp. 1–24.

Documents

Assessment of Heavy Metal and PAH Contamination of Urban Streambed Sediments on Macroinvertebrates