Top Banner
Polycyclic aromatic hydrocarbons (PAHs) in sediments from lake Lille Lungegårdsvannet in Bergen, western Norway; appraising pollution sources from the urban history Malin Andersson , Martin Klug 1 , Ola Ann Eggen 1 , Rolf Tore Ottesen 1 Geological Survey of Norway (NGU), Postboks 6315 Sluppen, 7491 Trondheim, Norway HIGHLIGHTS PAH in urban lake sediments present large concentration variations. The anthropogenic inuence can clearly be seen as higher PAH concentrations. The most prominent PAH sources are urban res, gasworks and trafc. 14 C dating does not provide consistent data in the anthropogenic period. abstract article info Article history: Received 3 September 2013 Received in revised form 24 October 2013 Accepted 24 October 2013 Available online 16 November 2013 Keywords: PAH Lake sediment Urban 14 C Bergen Norway This study aims to determine the temporal character and concentration variability of polycyclic aromatic hydrocarbon (PAH) during the last 5400 years in urban lake sediments through a combination of dating and chemo-stratigraphical correlation. We investigate the chemical history of the city of Bergen and determine the effect of specic point sources, as well as diffuse sources, and also help assess the risk of remediation plans. By using several organic compounds, metals and cyanide, we demonstrate the more accurate timing of sedimentation. The PAH results display very low concentrations in pre-industrial times, followed by a general increase that is punctuated by a few signicant concentration increases. These most probably correspond to urban res, domestic heating, gaswork activity and most recently due to trafc pollution. At the same depth as a signicant rise in concentration from background levels occurred, the high relative occurrence of low-molecular-weight PAH-compounds, such as naphthalene, were replaced by heavier compounds, thus indicating a permanent change in source. The general observation, using ratios, is that the sources have shifted from pre-industrial pure wood and coal combustion towards mixed and petrogenic sources in more recent times. The 14 C dating provides evidence that the sedimentation rate stayed more-or-less constant for 4500 years (from 7200 to 2700 calibrated years before present (cal yr BP)), before isostatic uplift isolated the water body and the sedimentation rate decreased or sediments were eroded. The sediment input increased again when habitation and industrial activities encroached on the lake. The 14 C dating does not provide consistent data in that period, possibly due to the fact that the lake has been used as a waste site throughout the history of Bergen city. Therefore, results from 14 C dating from anthropogenically inuenced sediments should be used with caution. © 2013 Elsevier B.V. All rights reserved. 1. Introduction Polycyclic aromatic hydrocarbons (PAHs) are a group of organic pollutants in our environment, which can be formed naturally during res and volcanic eruptions, and have thus existed long before human activity. However, their heightened status as pollutants originates from the fact that some of the compounds are proven carcinogenic, mutagenic and teratogenic and may have an adverse effect on human health (EU, 2002). With the onset of industrialisation, PAH from the combustion of fossil fuel has surmounted contributions from natural sources. Urban soil studies have documented that the highest anthropogenic- caused concentrations of PAH are found in urban areas with the longest habitation history (Haugland et al., 2008; Jensen et al., 2011; Mielke Science of the Total Environment 470471 (2014) 11601172 Corresponding author. Tel.: +47 73904321; fax: +47 73921620. E-mail address: [email protected] (M. Andersson). 1 Tel.: +47 73904321; fax: +47 73921620. 0048-9697/$ see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.scitotenv.2013.10.086 Contents lists available at ScienceDirect Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv
13

2014-Science of the Total Environment 470–471 (2014) 1160–1172

Dec 23, 2015

Download

Documents

Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: 2014-Science of the Total Environment 470–471 (2014) 1160–1172

Science of the Total Environment 470–471 (2014) 1160–1172

Contents lists available at ScienceDirect

Science of the Total Environment

j ourna l homepage: www.e lsev ie r .com/ locate /sc i totenv

Polycyclic aromatic hydrocarbons (PAHs) in sediments from lake LilleLungegårdsvannet in Bergen, western Norway; appraising pollutionsources from the urban history

Malin Andersson ⁎, Martin Klug 1, Ola Anfin Eggen 1, Rolf Tore Ottesen 1

Geological Survey of Norway (NGU), Postboks 6315 Sluppen, 7491 Trondheim, Norway

H I G H L I G H T S

• PAH in urban lake sediments present large concentration variations.• The anthropogenic influence can clearly be seen as higher PAH concentrations.• The most prominent PAH sources are urban fires, gasworks and traffic.• 14C dating does not provide consistent data in the anthropogenic period.

⁎ Corresponding author. Tel.: +47 73904321; fax: +47E-mail address: [email protected] (M. Anderss

1 Tel.: +47 73904321; fax: +47 73921620.

0048-9697/$ – see front matter © 2013 Elsevier B.V. All rihttp://dx.doi.org/10.1016/j.scitotenv.2013.10.086

a b s t r a c t

a r t i c l e i n f o

Article history:Received 3 September 2013Received in revised form 24 October 2013Accepted 24 October 2013Available online 16 November 2013

Keywords:PAHLake sedimentUrban14CBergenNorway

This study aims to determine the temporal character and concentration variability of polycyclic aromatichydrocarbon (PAH) during the last 5400 years in urban lake sediments through a combination of datingand chemo-stratigraphical correlation. We investigate the chemical history of the city of Bergen and determinethe effect of specific point sources, as well as diffuse sources, and also help assess the risk of remediationplans. By using several organic compounds, metals and cyanide, we demonstrate the more accurate timingof sedimentation.The PAH results display very low concentrations in pre-industrial times, followed by a general increase thatis punctuated by a few significant concentration increases. These most probably correspond to urban fires,domestic heating, gaswork activity and most recently due to traffic pollution. At the same depth as a significantrise in concentration from background levels occurred, the high relative occurrence of low-molecular-weightPAH-compounds, such as naphthalene, were replaced by heavier compounds, thus indicating a permanentchange in source. The general observation, using ratios, is that the sources have shifted from pre-industrialpure wood and coal combustion towards mixed and petrogenic sources in more recent times.The 14C dating provides evidence that the sedimentation rate stayed more-or-less constant for 4500 years(from 7200 to 2700 calibrated years before present (cal yr BP)), before isostatic uplift isolated the waterbody and the sedimentation rate decreased or sediments were eroded. The sediment input increasedagain when habitation and industrial activities encroached on the lake. The 14C dating does not provideconsistent data in that period, possibly due to the fact that the lake has been used as a waste site throughoutthe history of Bergen city. Therefore, results from 14C dating from anthropogenically influenced sedimentsshould be used with caution.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

Polycyclic aromatic hydrocarbons (PAHs) are a group of organicpollutants in our environment, which can be formed naturally duringfires and volcanic eruptions, and have thus existed long before human

73921620.on).

ghts reserved.

activity. However, their heightened status as pollutants originatesfrom the fact that some of the compounds are proven carcinogenic,mutagenic and teratogenic and may have an adverse effect onhuman health (EU, 2002). With the onset of industrialisation, PAHfrom the combustion of fossil fuel has surmounted contributionsfrom natural sources.

Urban soil studies have documented that the highest anthropogenic-caused concentrations of PAH are found in urban areas with the longesthabitation history (Haugland et al., 2008; Jensen et al., 2011; Mielke

Page 2: 2014-Science of the Total Environment 470–471 (2014) 1160–1172

1161M. Andersson et al. / Science of the Total Environment 470–471 (2014) 1160–1172

et al., 2011; Morillo et al., 2007) and soils with high organic mattercontent (Morillo et al., 2007). The “urban pattern” from diffuse sourceslike traffic and domestic heating is superimposed by point sourcessuch as aluminium smelters, gasworks, impregnation facilities and con-taminated sites that contain for example tar, oil or other petroleumproducts (Neff, 1979).

Studies of PAH in urban lake sediments cores have been conductedby Guo et al. (2010), Guo et al. (2011b), Ikenaka et al. (2005), Itohet al. (2010), Jung et al. (2008), Van Metre and Mahler (2000) andYang et al. (2011) and in the Bergen region by Eide et al. (2011), Deter-mination of PAH concentrations in sediment coreswas a common factorfor these studies. These studies also used different methods forsource apportionment through the assessment of possible sourcesto distinguish between petrogenic and pyrogenic PAHs. They re-vealed increasing PAH-concentrations concomitant to population

Fig. 1.Maps presenting: a) The water bodies surrounding Bergen centre, the original drainagecenturies. c) the location of shoreline over the last eight centuries.

and economic growth often combined with an increasing numberof PAH sources. However, all of these studies were limited to thelast century whilst other sediment studies have focussed only onsurface sediment samples.

Due to land reclamation activity and an interest in cultural heritage,several harbour sediment investigations have been conducted in theBergen area (Fig. 1). PAH studies of surface and subsurface sedimentsamples revealed comparatively higher concentrations in subsurfacesediments (Fylkesmannen i Hordaland, 2005; Johnsen et al., 1998).Johnsen and Brekke (2009) presented high PAH concentrations insurface sediments in lake Lille Lungegårdsvannet (LL). The sedi-ments of this lake are the focus of attention in this study.

One of the aims of the EuropeanUnion'sWater Framework Directiveis to ensure that by 2015 extensively modified water bodies, such as LL,will be required to meet good ecological potential criteria (EU, 2008).

area and the smaller, current drainage area. b) The sprawl of habitation over the last eight

Page 3: 2014-Science of the Total Environment 470–471 (2014) 1160–1172

1162 M. Andersson et al. / Science of the Total Environment 470–471 (2014) 1160–1172

In 2003 aNorwegian Council for Contaminated Sedimentswas appointedto identify gaps in scientific knowledge and to make recommendationsconcerning sediment management (Breedveld et al., 2010). Severalconclusions were drawn:

1. Before making decisions on remediation the uncertainty and risksmust be analysed

2. Clear, realistic, understandable and verifiable environmental targetsshould be established for each remediation project.

3. Source control should be documented before remediation is initiated.

This study is a step in the knowledge-building and documentationwork required before the remediation process plan starts. Here wepresent results from a high-resolution study of PAH content from pre-historic to anthropogenic influenced lake sediments. This study offersa unique opportunity to investigate the content of PAHs from pristineto industrial environment and to retrace the chemical history of a cityprogressively growing in size.

2. Study site

2.1. Lake Lille Lungegårdsvannet and the population development of Bergen

The present-day lake LL is located in the city of Bergen on thewest coast of Norway (Fig. 1). Bergen, which was founded in the 11thcentury, has gradually grown in size to over 260 000 inhabitants atthe present day. The climate is maritime with mean winter (DJF) andsummer (JJA) temperatures of 1.7 and 13.9 degrees Celsius respectivelyand a mean annual precipitation of 2250 mm per year (StatisticsNorway, 2012). Central Bergen, which surrounds LL, is situated at sealevel and surrounded by a mountainous area up to 640 m (Fig. 1a)that frequently causes inversion, also called the “cap of Bergen” (Cityof Bergen, 2007). Today the surface area of the lake is about 2.8 haand the maximum water depth is 6 metres in the centre of the lake.

Over the centuries LL has been subjected to various modificationsthat have transformed its primary size, shore line (Fig. 1c), water ex-change rate and usage. Frompreviously being a part of the fjord system,the open access on both sides to the sea was reduced and the waterbody became a fjord inlet. This occurred approximately 2700 yr BP asa result of isostatic uplift (Øystein Lohne, pers. comm.). The inlet subse-quently reduced in size and the open connection to the sea decreasedand finally closed due to land reclamation. From the early 20th century,the margins of the lake have been at their current location.

When Bergen became a settlement in the 11th century, the inletwas surrounded by heathland (Helle, 1995). Subsequent growth ofBergen city and the increasing need for building sites transformed theland usage around the inlet from pasture to an urban area. In the 17thcentury, the northern shoreline of LL was inhabited. Subsequently,habitation spread around the whole inlet (Harris, 1991) (Fig. 1b).In addition, parts of the inlet have gradually been filled and some ofthe natural shorelinewas replacedwith quay structures. The populationand related industrial history of the study area is characterised bysteady urban expansion that gradually incorporated LL and also affectedits catchment area (Harris, 1991).

2.2. Water quality

LL has for centuries been characterised by poor water quality.Previous studies have documented high nutrient values, stratifica-tion and hypoxic conditions within the bottom water column duringthe last 90 years (Johansen and Skarheim, 1968; Larsson, 1994;Johnsen and Brekke, 2009; Johnsen, 2008). At intervals, the lakewater has deteriorated due to algal bloom. The first reported studiesundertaken in the 1920s and 1930s tried to find a cause and solutionto the production of H2S (Johansen and Skarheim, 1968). Even recentstudies of LL indicate leakages of municipal waste water into the lake(Johnsen and Brekke, 2009).

The exchange of water between LL and the sea was gradually re-duced as the open connection was first replaced by a channel andthen by drainage pipes. The pipes extended down to the deepestpart of the lake in an attempt to drain deep saline water and replaceit with fresh water and thus reduce the production of H2S in the stag-nant bottom water column (Larsson, 1994).

Presently urban stormwater is redistributed into the lake to increasethe inflow of fresh water from a 90 040 m2 large drainage area.Although numerous attempts have been made to reduce salt waterin LL and to increase its water quality, the coarse and porous fillingmaterial used for land reclamation still allows saline water tore-circulate into the lake. This can be seen at low and high tideas the water level fluctuates.

2.3. Anthropogenic PAH point sources

In addition to environmental changes, craft and industrial activi-ties with subsequent input of refuse and pollutants during varioustime periods have affected water quality and sediment deposition.This was accompanied by changes in bioproductivity and grain sizedistribution. The concentration of PAH results from a number ofpoint and diffuse sources that have been active within the catchmentarea over time.

During the 18th and 19th century it was customary to place industryand crafts outside the inhabited area to obviate industrial stench andnoise. Thus the lake provided a convenient and local waste disposalplace for industries and households and several major point sourceshave existed by the lake.

Themain pollution source in the 19th century was from a gas pro-duction unit that operated at the quayside of the lake for almost50 years (1856–1908). Process waste such as diluted ammoniacalfluids from the gas cleaning was emitted directly into LL until 1870(Statsarkivet, 2012). From about the mid 19th century new produc-tion methods gave rise to a very sulphur-rich by-product calledspent oxide that may have a total cyanide content of up to 7%(Meehan, 2000). On-site deposition of spent oxide remnants are apotential pollution source of contaminated soil and sediment aroundthe gasworks. In 1908 the gasworks was replaced by a facility byPuddefjorden shore, outside the catchment of LL. The larger, newgasworks was producing gas until 1985, thus continuing the atmo-spheric deposition of PAH and metals in Bergen.

Two additional point sourceswere situated on the opposite quaysidethat utilised coal in electricity production. These were active during thefirst decades of the 20th century (Gjerstad, 2005). The electricity worksshut down as large-scale hydroelectric development was completed inBergen. In addition to the main point sources by the lake, many smallindustries, petrol stations and craft activities have existed by the lake.

2.4. Anthropogenic diffuse PAH sources

Most of the present potential PAH sources in Bergen area are diffuse:urban soil, stormwater, atmospheric deposition and city fires.

During pre-industrial times it was common to burn heath andother plant remains such as old nutrient-poor woody Calluna plantsto prepare land for agriculture and pasture. This burning also fertilisesthe soil and hinders shrub and tree colonisation (Nilsen, 2004).This tradition likely resulted in the spread of PAHs into the environ-ment and formed a diffuse PAH background signal in the pre-industrial sediments.

Generally, the urban soil in Bergen shows PAH concentrations thatexceed those in other Norwegian towns (Jensen et al., 2011), especiallywithin the city centre (Forfang, 2009;Haugland et al., 2008;Ottesen andVolden, 1999). Soil studies conducted in the immediate vicinity of thelake show that the soil is polluted with PAHs with concentrations upto 10 mg/kg and oil to at least 4 to 6 m depth depending on the location(Noteby, 1998; Multiconsult, 2004). Therefore, it can be expected that

Page 4: 2014-Science of the Total Environment 470–471 (2014) 1160–1172

1163M. Andersson et al. / Science of the Total Environment 470–471 (2014) 1160–1172

part of the PAH concentrations in the sediments derive from pollutedsoil that has been eroded and flushed into the lake.

According to national statistics of atmospheric emissions (www.ssb.no), the current main diffuse emission source of PAH is domesticheating (Kocbach et al., 2006). In addition to using wood for domes-tic heating (Finstad et al., 2004), coke was mainly used for heatingwhile co-existing with the gasworks. In more recent times oil-firedheating has been more prevalent. Within the watershed area, manydomestic oil tanks are buried underground. although not all ofthem are in use. These are of varying quality and age, such that oilspillage can be a potential source (Multiconsult, 2004). A higherdegree of deposition of atmospheric particles can be expected inthe Bergen area due to relatively high precipitation levels and thefrequent inversion.

Urban storm water is gaining attention as a major, diffuse pollut-ant source in urban areas (Brown and Peake, 2006; Cornelissen et al.,2008; Davis et al., 2001; Gromaire-Mertz et al., 1999; Jartun et al.,2008). Storm water runoff includes a mixture of anthropogenicPAH sources; combustion of fossil-fuel, vehicular emissions, abrasionof asphalt and car tires, combustion processes of waste incineratorsand domestic heating (Jung et al., 2008). In the present study somesediment samples from stormwater traps within the drainage areawere analysed in order to investigate whether active pollution path-ways exist around the lake that in present time contain PAHs. Thecity of Bergen is currently, redirecting increasing amount of urbanstormwater in order to increase the inflow of fresh water into thelake. This means that a large proportion of suspended particles inthe stormwater that derive from paved surfaces within the drainagearea flush straight into the lake.

Bergen has partially burnt down numerous times and since the16th century, six substantial city fires destroyed large areas of habi-tation in the near vicinity of the lake. These fires took place in 1561,1582, 1623, 1640, 1702, 1855 and 1916 (Byen Brenner, 2002). Thecity fires in 1640 and 1702were particularly extensive and destroyedover half of the city, which was at that time largely built of woodenhouses. Over time, the fires influenced the shaping of the city centreinto what it is today, consisting largely of concrete/stone buildingsseparated by wide streets. The inlet was at all times a convenientplace to dispose of fire refuse and has been extensively used as adumping location, which can be seen in pictures taken from the lake-shores in the early 20th century where the waste was used as mate-rial to fill in the lake (Fossen and Grønlie, 1985). Burnt refuse has alsobeen documented within soil profiles in environmental assessments(Asplan Viak, 2003).

Fig. 2. Coring vessel at Lille Lungegårdsvann (photographer: Atle Nesje).

3. Methods

3.1. Field work

Two cores were collected at separate dates. The sediment coring forthe long corewas performed in 2009 from a raft in the centre of the lake(Fig. 2) by using a piston corer (Nesje, 1992). The extraction of the corewas performed by pushing a 6 m long plastic tube into the sediments.The diameter of the plastic tube was 100 mm. To supplement thelong sediment record with undisturbed surface sediments a secondcore was taken in May 2011 with a Hongve type gravity corer (Boyle,1995) with an inner diameter of 66 mm. The total length of sedimentrecovered was 25 cm for the surface core and 535 cm for the maincore. The long sediment core was lengthwise opened in the laboratoryand sub-sampled at 2 cm intervals. Aliquots for PAH, metals and 14Cdating were extracted. The bulk sample aliquots for PAH analysis werestored in the dark at 4 °C in aluminium foil-sealed glass jars until furtherprocessing and analyses. The surface core was sub-sampled in the fieldat 1 cm intervals and the samples were stored at 4 °C in Rilsan bagsuntil analysed. Not all samples have been analysed from either core,but a collection of samples were selected from regular intervals inorder to give an un-biassed picture of the concentration in the sedimentcore. The bottom 2 m of the long core were not analysed for PAH.

In order to uncover the displacement between the top of the longcore and the actual sediment surface, a visual comparison analysis com-binedwith an analysis of correlation coefficients were used. All subsam-ples from both cores, not just a selection of samples, were additionallyanalysed for inorganic elements and TOC. A selection of ten elementswas used to visualise the amount of displacement between the sedi-ment surface and the first sample of the long core. The displacementamounted to 18 cm, which also could be confirmed by 137Cs-dating ofthe top core (NGU, unpublished data).

Samples of settled sediments within stormwater traps around thelake were extracted in May 2011 using a long-handled aluminiumscoop. The scoop was cleaned in the water within the trap that wassampled to avoid contamination between traps. The samples werestored in a similar manner as the sediment samples from the cores.

3.2. Laboratory work

3.2.1. PAH analysisThe PAH extraction of the bulk samples was performed by weighing

the samples in an Accelerated Solvent Extraction (ASE) cell. Internalstandards that contain 16 deuterium marked compounds were addedas well as adsorbent agent diatomaceous silica. Laboratory blankswere analysed. The samples were extracted using ASE and elutedby toluene with 5% acetic acid. Excess water was removed usingsodium sulphate before the extract was transferred to a Zymarktube and subsequently reduced to 500 μl. The analyses were per-formed using selective ion monitoring on a Agilent 6890/5973 gaschromatography–mass spectrometer (GC-MS) with EI-ionisation.The column was a J&W DB-5MS 30 m × 0,25 mm × 0.25 μm.

All 16 parent PAH compounds prioritised by the United StatesEnvironmental Protection Agency were determined. The recoveriesfor the internal standards are presented in Table 1.

3.2.2. TOC analysisTOC analysis was performed at the laboratory of the Geological

Survey of Norway, Trondheim. In order to remove any CO2 boundto carbonates samples were weighed and treated with 10% HCl atabout 50 °C for 30 minutes. After complete removal of CO2 sampleswere washed 10 times with double distilled water in vacuum anddried in 80 °C for a minimum of two hours. The remaining organiccarbon content was determined using a Leco SC-444 (Michigan,USA). To control the sample preparation and instrumental perfor-mance, duplicate samples (4% of the samples) and project standard

Page 5: 2014-Science of the Total Environment 470–471 (2014) 1160–1172

Table 1Minimum, maximum and median recoveries for the internal standards during analysis.

Compound Minimum recovery(%)

Maximum recovery(%)

Median recovery(%)

Standard deviation

Naphthalene 6 75 54 18.3Acenaphthene 23 100 79 15.1Acenaphtylene 12 63 47 22.2Fluorene 39 112 80 21.2Phenathrene 50 124 101 18.1Anthracene 29 109 79 20.4Pyrene 64 129 109 15.9Fluoranthene 72 172 126 19.5Benzo(a)anthracene 61 159 117 26.7Chrysene 54 125 98 16.2Benzo(k)fluoranthene 60 135 97 17.3Benzo(b)fluoranthene 59 175 127 26.9Benzo(a)pyrene 1 156 90 33.2Dibenzo(a,h)anthracene 44 159 111 26.2Indeno(1,2,3, cd)pyrene 55 180 112 35.5Benzo(g,h,i)perylene 39 143 91 23.9

1164 M. Andersson et al. / Science of the Total Environment 470–471 (2014) 1160–1172

samples (6%) were inserted into the sample series at intervals.The mean TOC value of the inserted standard samples (N = 16) is0.61% and the standard deviation of the measurements 0.10.

3.2.3. Cyanide analysisThe cyanide (total) analysis was performed according to ISO 17380.

The sediment was leached using a NaOH-solution.

3.2.4. Grain size analysisFor grain size analyses, samples were pre-treated with H2O2 to

remove organic matter. Prior to the measurement the samples weremixed with 5–10 ml 5 % sodium diphosphate and 40 ml deionisedwater with a subsequent 5 minutes of ultrasonic treatment. Measure-ments were performed using a Coulter LS200 (California, USA) with ameasuring range of 0.4 μm–2000 mm.

3.2.5. 14C datingFor chronological control, terrestrial plant remains were separated

from predefined sediment depths by sieving using demineralisedwater and dried at room temperature. The core contains generallyvery low concentrations of terrestrial macrofossils. Some samples thatwere planned to be dated did not contain enough material. Samplepreparation for accelerated mass spectrometry was performed withAcid–Alkali–Acid (AAA) method according to internal routine at theScience Museum, Norwegian University of Science and Technology,Trondheim. The samples were treated with diluted sodium hydrox-ide (10 g/100 ml) and further treated with diluted hydrochloricacid (5 ml/100 ml). The samples were then graphitized with a cleans-ing step of the CO2 and further analysed by accelerator mass spectrom-etry (AMS) at the Tandem laboratory, Uppsala University, Sweden.Blank samples and samples of reference material Ox2 were preparedin the same way as the samples and inserted into the analytical batch.

The age-depth model was developed using Bayesian sequencemodelling (P_Sequence deposition model, variable k-value), imple-mented in the OXCAL v4.2 programe (Ramsey, 2008). Uncertaintiesreported are 2-sigma.

4. Results

Most of the sediment appears homogenous, with only minor colourchanges. The grain size within the whole core is silty sand and poorlysorted according to Folk and Ward (1957). According to colour andgrain size characteristics the sediment record can be divided into twomain units. The lower part (Unit 1) from 549 to 185 cmbelow sedimentsurface (bss), is characterised by stiff sediments with a dominantlybrownish colour. Clear indications of bioturbation are found withinthe lowest 10 cm and abundant shell fragments occur in intervals

(at 549–529, 469–459, 441–437, 299–301, 151–153 and 109–111 cmdepth). The grain size within the lowermost 50 cm is dominated bysandy silt with a gradually increasing amount of the finer fraction.According to Rise and Brendryen (2013) the amount of clay should bemultiplied with four as the sediments have been analysed by the aidof laser. This sediment section is slightly better sorted than the rest ofthe core (Fig. 3). From 350 cm bss medium silt dominates towards theupper unit boundary. At 210 cm bss gravel is present. The sedimentcolour of Unit 2 (185–0 cm bss), gradually shifts from brownish todark brownish softer sediments at the top. Some lighter sections occurat (147–127 cm; 87–77 cm and 55–49 cm bss). Grain size is compara-bly finer than in Unit 1 dominated by medium to fine silt, interspersedwith coarse silt sections at 145 to 133 cm bss and more frequent sifttowards the top. Gravel occurs within the sediments at 110 cm bss.Unfortunately the top surface samples did not provide enough materialto perform a grain size analysis.

The PAHsum16 (Table 2) results display very low concentrations inthe bottom of the analysed samples as the median of the bottom 371–197 cm bss (8 samples), all within Unit 1, is 0.15 mg/kg PAHsum16

(Fig. 4/Table 2). A general increase in concentration can be seen overthe whole time period that the core represents. As mentioned before,samples from the bottom core were not analysed as 350 cm depthwould provide a background concentration, not influenced by an-thropogenic activity. More dramatic increases in concentration occurat several intervals within the core. The PAH results present approxi-mately three time periodswhen the PAH concentrations have risen dra-matically, with a subsequent gradual, reduction in concentration. Theconcentrations increase again in recent times to high concentrations(max. 57 mg/kg PAHsum16) to reduce again towards the very topsediment. The correlation between grain size/median) and levels ofPAHsum16 is slight, but not high (R2 = 0.26).

The results for the bottom samples present a high (75–90%) relativeoccurrence of light PAHs (2–3 ring) compared to the heavier fraction(Fig. 5). The high relative occurrence drops dramatically around190 cm bss to around 10%, the depth being approximately where adistinct drop in grain size is observed. To our knowledge, the rela-tively high amount of naphthalene in the bottom samples has notbeen observed in any other lake sediment study. However, not allstudies have included naphthalene in the analytical programme(e.g. Jung et al., 2008). We have been unable to find another studythat has analysed sediments from so far back in time. Therefore, pre-vious studies offer no explanation as to why such a high percentageof low molecular PAHs occurs in the bottom of the core.

The results for TOC should reflect the amount of inflow and thedegree of degradation of organic matter and the degree of dilution byinorganic matter. Several studies have noted a correlation betweenthe PAH-levels and the TOC-levels (e.g. Choudhary and Routh, 2010).

Page 6: 2014-Science of the Total Environment 470–471 (2014) 1160–1172

Fig. 3. Schematic log of the Lille Lunggårdsvann stratigraphy. The grain size distribution core shows the density of grain sizes against depth. The scale 0 to 5 corresponds to volumepercentages. The top samples were not analysed due to lack of sample material. The grey graphs present TOC, phosphorus and sodium content in the sediments against depth.

Table 2Results for the 16 PAH compunds analysed. All results in μg/kg.

Depth Na Acn Ac Fl Phe An Fa Py B(a)A Chr B(b)F B(k)F B(a)p D(ah)A B(ghi)P IP

0–3 cm 270 270 64 97 450 190 1100 940 540 560 870 410 680 180 880 6605–7 cm 250 390 64 91 560 250 1400 1400 740 980 1500 630 1000 290 1400 97010–12 cm 390 690 120 160 1000 680 4000 3900 2200 2900 4400 1700 2500 520 3000 220014–16 cm 700 1200 190 340 2000 1300 7200 7200 4500 5300 8000 2800 4600 770 4700 400018–21 cm 570 250 42 110 960 720 4400 3750 1550 2300 2950 1300 1550 450 1500 150025–27 cm 1100 270 100 140 1700 1100 8100 12000 7000 4600 6600 2850 5200 1300 3300 300031–33 cm 460 49 15 8 140 100 270 350 94 99 120 46 58 26 78 7037–39 cm 430 8 4 360 220 140 560 950 430 380 530 210 680 82 320 30043–45 cm 815 62 55 80 350 200 520 540 220 250 260 100 200 40 180 18049–51 cm 820 9 34 1 120 130 280 250 50 53 53 15 6 7 22 3755–57 cm 860 150 84 440 730 420 1200 1300 650 670 730 290 580 140 530 50061–63 cm 720 100 73 530 620 410 1100 1400 670 760 970 430 1700 190 890 77067–69 cm 700 48 25 75 490 280 750 920 230 340 480 200 300 57 280 34073–75 cm 1300 0.1 150 4000 200 260 120 320 61 81 55 19 23 15 35 2179–81 cm 900 95 77 120 500 310 800 870 390 430 420 150 320 61 280 24085–87 cm 910 130 110 250 800 490 1300 1600 580 630 690 260 560 100 430 38091–93 cm 1100 240 160 360 1500 1200 2700 2500 1500 1500 1600 690 1800 260 1200 110099–101 cm 580 36 22 99 410 210 510 540 120 170 230 81 150 20 140 120105–107 cm 1300 350 210 400 1900 2100 3400 4100 2200 2400 2600 1100 – 420 2800 1900111–113 cm 1200 210 160 260 1100 710 2500 3000 1300 1500 1700 660 2900 310 1500 1300117–119 cm 2500 400 290 460 2000 1200 4500 4200 2500 2600 2800 1100 3100 480 2000 1900123–125 cm 1800 410 240 450 1600 1100 4600 4100 2600 2800 2900 1200 4700 460 2200 1800129–131 cm 850 98 100 200 650 370 1200 1200 520 560 540 250 440 75 330 360135–137 cm 290 0.1 3 2 70 110 140 140 34 37 65 32 1 8 20 31141–143 cm 890 47 44 400 330 230 550 610 280 290 310 120 280 48 200 190147–149 cm 870 81 67 86 540 410 980 1100 370 490 600 280 340 60 410 485155–157 cm 380 35 29 410 230 140 430 380 130 140 200 94 150 24 170 170163–165 cm 860 190 100 490 970 760 1900 1700 670 710 1200 700 770 57 580 580171–173 cm 480 35 61 92 440 250 680 520 170 160 260 140 140 20 160 170179–181 cm 610 100 100 310 610 180 550 480 200 160 230 69 120 30 130 120187–189 cm 110 2 21 480 110 76 250 190 63 62 70 24 33 9 34 36197–199 cm 560 8 140 0.7 47 150 59 38 23 13 10 5 6 8 8 10205–207 cm 370 27 4 6 49 35 24 27 12 6 8 2 3 3 10 9215–217 cm 660 26 110 0.1 5 3 15 14 9 7 10 3 8 3 5 7225–227 cm 240 8 7 300 99 67 38 38 17 14 17 7 6 4 10 8245–247 cm 270 3 7 370 25 10 14 10 8 5 7 3 8 2 8 9269–271 cm 200 0.1 2 6 4 38 20 15 6 2 4 2 6 0.1 0.1 4317–319 cm 330 0.7 7 2 3 0.5 5 3 4 0.4 4 0.6 2 0.2 3 5369–371 cm 170 0.1 4 5 7 4 13 7 7 7 15 7 0.5 0.1 4 9

PAHsum16: Na: naphthalene; Acn: Acenaphthylene; Ac: Acenaphthene; Fl: Fluorene; Phe: Phenathrene; An: Anthracene; Fa: Fluoranthene; Py: Pyrene; B(a)A: Benz(a)anthracene; Chr:Chrysene; B(b)F: Benzo(b)fluoranthene; B(k)F: Benzo(k)fluoranthene;B(a)P: Benzo(a)pyrene; D(ah)A:Dibenz(a, h)anthracene; B(ghi)P: Benzo(ghi)perylene; IP: Indeno(1,2,3-cd)-pyrene.

1165M. Andersson et al. / Science of the Total Environment 470–471 (2014) 1160–1172

Page 7: 2014-Science of the Total Environment 470–471 (2014) 1160–1172

Fig. 4. PAHsum16 and cyanide(total) concentrations within the core.

1166 M. Andersson et al. / Science of the Total Environment 470–471 (2014) 1160–1172

Such a correlation, however, can only be expected if TOC and PAHweredeposited by the same pathways. This study presents no correlationbetween these parameters (R2 = 0.09), which may indicate differentsources. Very low correlation between these parameters is also docu-mented by Gocht et al. (2001) and Yang et al. (2011).

The gasworks are a known, major source of PAHs, where the activeperiod may be pinpointed by using cyanide as a proxy. The cyanide ispresent as a legacy from the gas purification process. The samples dem-onstrate a distinct peak of cyanide concentration that coincides with aPAH peak (Fig. 4). On either side of the cyanide peak, the concentrationsdrop to under the detection limit.

The 14C dates are presented in Fig. 6 and Table 3. The AMS ages dis-play a period (Unit 1) of even supply of sedimentmaterial from 7200 to2700 cal. yr BP. The age model shows a significant change in sedimentdeposition around that time. 2700 yr BP corresponds to approximately200 cmdepth (Fig. 6). Unit 2 also displays an even, but higher sedimen-tation rate than sediments from Unit 1.

Sediment samples from ten stormwater trapswere analysed and thePAHsum16 concentrations ranged from b0.45 to 162 mg/kg (median5.6 mg/kg). Even though the sample number is low, it gives an indica-tion that urban stormwater is a diffuse PAHsource that is real and activetoday. A study byNGU (unpublished data) determines that a substantialamount of suspendedmaterial enters the lake through stormwater. Theresults therefore indicate that stormwater has partly contributed to thehigh PAH concentrations in the top of the sediment.

5. Discussion

This study gives the opportunity to investigate environmentalchanges that take place before and during the growth of a city. This isdone by combining the results from detailed chemical analysis of the

sediment cores, chronological knowledge and the known history ofthe city of Bergen. By analysing samples of sediments that were de-posited before the influence of man, the diffuse background levelsare known and the degree of anthropogenic influence can thereforebe recognised.

The AMS dating results provide a framework for understanding thesediment deposits. The sedimentation model (Fig. 6) includes resultsfrom 137Cs dating (NGU, unpublished data), and a few periods withknown urban activities tied to chemical data.Within Unit 1, the bottomof the core represents the Mesolithic period within the Stone age, a pe-riod of scattered habitation along coastal Norway. An two-fold increasein the concentration of lead, at approximately 320 cm bss (unpublisheddata, NGU), may be supported by dating results that coincide with themain period of Greek and Roman lead mining (Hong et al., 1994). LLlost its connection towards the sea on both sides around 2700 yr BP.This is coincidingwith the large hiatus (Fig. 6) and a thin gravel horizonat 210 cm bss. We envisage that as the connection between LL and thesea was reducing, tidal currents would have a higher intensity, possiblyeroding the fine grained surface sediments, leaving an erosion horizon.How much of the hiatus was caused by erosion and how much byreduced sedimentation rate is not known today. The isolation of LL issupported by the fact that the sodium concentration in the sedimentsdrop significantly at the same sediment depth (Fig. 3), confirming theage when the connection was closed and LL became an inlet with adecreased input of saline water.

The 14C dating results from Unit 2 present a more chaotic picturewhere outliers with a stratigraphically reversed age lie within the unit.The age model has added several marker horizons in addition to theAMS dates within this unit. These horizons include events such as theonset of the gaswork activity, which can be seen in an increased PAHconcentration, and the attempt to improve thewater quality by pouring

Page 8: 2014-Science of the Total Environment 470–471 (2014) 1160–1172

Fig. 5. The relative occurrence of PAH compounds from the light (darker grey) to the heavier (white) of the analysed compounds.

1167M. Andersson et al. / Science of the Total Environment 470–471 (2014) 1160–1172

sand and gravel on top of the sediments, which is evident by anothergravel horizon. The ages of the extreme outliers might represent mate-rial that has been reworked or inserted by human hand into the waterinlet, as the inlet was utilised as a dumping ground for many differentwaste types (Waldmann et al., 2011). Within Unit 2, representing theAnthropocene, the AMS dating results appear older than their positionin the core suggests and may be considered unreliable. The generallyolder ages may possibly be as a result of sample contamination by oldor dead carbon sources such as coal (Björck and Wohlfarth, 2001) thatwas not completely removed during the pretreatment process. Fromthe initial habitation (17th century) of the inlet shores until today, thesedimentation rate increased and has remained high. The inlet has,through its role as a sewage andwaste recipient, accumulated consider-able anthropogenic additions. Sewage has also led to a higher biologicalproduction within the water column, which further increases the sedi-mentation rate. This is also evident from the phosphorus concentration(Fig. 3).

Our results showing the PAH concentration paint a general picturewith low concentrations in the bottom, a general increasewith dramaticconcentration increases at several points in history. Observing back-ground levels relative to the levels that are occurring today, they arein agreement with Wilcke et al. (2005) who observed that the lowestmeasured PAH concentration in temperate soils are approximatelyten times higher than the concentrations assumed to exist prior toindustrialisation. The same ratio has also been documented in sedi-ments (Van Metre et al 2000). The first obvious change occurs asa rise in concentration from background levels, at approximately190 cm bss (Fig. 4), the same depth as when the relative PAH com-pound occurrences show some distinct changes in concentration(Fig. 5). At this depth in the sediments, the high relative occurrenceof low-molecular-weight (LMW) PAH-compounds such as naphthalene

are replaced by heavier compounds. The change results in a higherrelative occurrence of fluoranthene, pyrene, benz(a)anthraceneand chrysene. The higher relative occurrence of these compoundsremains permanently high all the way up in the core, thus indicatinga permanent change in source(s). The shift might also indicate deg-radation of more readily bio-available components, resulting in ap-parent shift to heavier compounds. It has been suggested that LMWPAHs mainly originate from low- to moderate-temperature combus-tion processes, such as biomass burning and domestic wood and coalcombustion (Guo et al., 2011b; Wang et al, 2007; Yang et al, 1998),while high-molecular-weight (HMW) PAHs are thought to be aproduct of high-temperature combustion (Khalili et al., 1995) or coalcombustion (Duval and Friedlander, 1981). A dominance of LMWPAHs (2–3 rings) may also, according to Guo et al. (2011a), representlocal sources of PAH. Source fingerprinting as presented by Khaliliet al. (1995) emphasises the difficulties in allocating such sources,as emissions from very different sources include the same PAH com-pounds. Khalili et al. (1995) documented that wood combustion pro-duces emissions that include about 70% of two- and three-ring PAHs.Naphthalene is regarded as very volatile, therefore it is not expectedto be detected in high amounts in older sediments. But if the substantialamounts of naphthalene that are produced are taken into account, aconsiderable part might still reach the soil and sediments although alarge amount is ultimately volatilised (Morillo et al., 2007). It cannotbe excluded that naphthalene concentrations in the oldest sedimentsmight derive from degradation of heavier compounds (Seo et al., 2009).

Fig. 7 displays the concentration of summed light compounds(including naphthalene) and the sum of the heavier compounds.The bottom of the core not only exhibits higher relative occurrenceof light compounds but the concentrations are also much higher(note log scale). The concentration of the heavy compounds exceeds

Page 9: 2014-Science of the Total Environment 470–471 (2014) 1160–1172

Fig. 6. Age model using all 14C-results, results from 137-Cs-dating and probable depth definition of gasworks and gravel horisons.

1168 M. Andersson et al. / Science of the Total Environment 470–471 (2014) 1160–1172

the light compounds at two distinct periods, 130–100 cm bss and inrecent sediments.

The concentrations of phenathrene, fluoranthene and pyrene arehigh further up in the core, which is in agreement with data in the liter-ature from other urban soil studies. The sum of these three is above 35%of PAHsum16 from 190 cm depth upwards in the core, while the sum ofthe three compounds below 190 cm core depth is generally below

15% and even as low as 10%. According to Morillo et al. (2007) the pre-dominance of these three compounds confirms the pyrogenic origin ofthe PAH as they are considered pyrogenic products from high-temperature condensation of LMW aromatic compounds. The changein relative occurrence from LMW to HMWPAHs therefore strongly sug-gests that the sediments from 190 cm bss upwards represent the pres-ence of industrial processes by the lake as the relative occurrence of

Page 10: 2014-Science of the Total Environment 470–471 (2014) 1160–1172

Fig. 7. The concentration of summed light compounds— 2 and 3 rings (including naphtha-lene) and the sum of the heavier compounds against depth.

Table 3Radiocarbon dates calibrated using Oxcal 4.1.

Laboratory IDa Core depthb

(cm)δ13C(‰)

14C age(year BPc)

Calibrated age range

TRa-2821 64 −24.6 650 ± 50 AD 1275–1404TRa-2820 88 −26.7 65 ± 35 AD 1690–1926TRa-2814 94 −27.2 195 ± 30 AD 1648–1955TRa-2819 98 −26.7 115 ± 35 AD 1679–1940TRa-2818 122 −28.3 365 ± 40 AD 1448–1635TRa-2812 144 −26.8 3780 ± 35 BC 2336–2046TRa-2817 146 −27.5 880 ± 40 AD 1035–1225TRa-2816 170 −25.9 320 ± 35 AD 1476–1647TRa-2815 184 −25.5 285 ± 35 AD 1490–1795TRa-4332 236 −27.5 3295 ± 40 BC 1686–1466TRa-4331 316 −27.7 3915 ± 70 BC 2577–2201TRa-2813 418 −28.7 4675 ± 45 BC 3630–3362Beta-268556 549 −25.1 6260 ± 40 BC 5320–5076

a TRa: Samples prepared at Laboratory for radiological dating, Norwegian University ofScience and Technology, analysed at Tandem laboratory, Uppsala University.

b sample depths are midpoints of 2 cm sample slices.c BP = before present.

1169M. Andersson et al. / Science of the Total Environment 470–471 (2014) 1160–1172

HMW rises as well as the occurrence of compounds that originate fromhigh-temperature combustion processes. The transition into a moreindustrialised period is supported by the fact that metals such as lead,mercury, zinc, copper and cobalt exhibit abrupt concentration increasesat that same core depth (unpublished data, Geological Survey ofNorway).

5.1. Ratios

The use of ratios between different PAH compounds to distinguishbetween natural and anthropogenic sources has been in extensive use(e.g. Yunker et al., 2002). Ratios for PAH have mainly been used to dis-tinguish between compounds from the incomplete combustion of or-ganic matter (pyrogenic sources), from petroleum and its products(petrogenic sources), or from the transformation of biogenic precursorsafter deposition (diagenetic sources). Source apportionment of PAHrests on several assumptions: Firstly, each potential atmospheric sourceis assumed to be associated with relative isomer proportions that areunique (Galarneau, 2008). Secondly, Katsoyiannis et al. (2007) assumedthat paired chemicals are diluted to a similar extent and that the ratiosremain constant en route from source to receptor. The complexity arisesfrom the co-occurrence of point and diffuse sources, where the level ofchanges in isomer ratios is uncertain. However, isomer ratiosmay showsubstantial variability, and should therefore only be seen as indicative.

To investigate whether a change in sources can be seen over time,we utilise PAH ratios to determine whether there are any graduallong-term changes in PAH sources over the thousands of years thatare covered in the core samples. To determine an indication of the rela-tive dominance of pyrogenic versus petrogenic sources, ratios of the iso-mers of parent PAH with the principal mass 178, 202, 228 and 276 (asused by many authors, e.g. Yunker et al., 2002) were calculated. The ra-tios of Fluoranthene and Pyrene and Benzo(a)anthracene and Chryseneare presented (Fig. 8). Due to overlap of characteristic PAH profiles fromsources and selective decay of more labile compounds we utilise a cou-pling of two ratios instead of using a single ratio. This is more helpful inproviding a better estimation of PAH sources (Lehndorff and Schwark,2004).

The ratio between Fluorathene and Pyrene (Fla/(Fla + Pyr)) indi-cates unburned petroleum if it is below 0.4, while a ratio at and above0.5 indicates coal, grass and wood combustion origins. The sedimentshere display a gradual decline in ratio from 0.6 from the bottom of thecore towards 0.4 in recent times. The ratio between Benzo(a)anthra-cene and Chrysene (BaA/(BaA + Chr) at a value of 0.35 distinguishesbetween combustion sources andmixed sources. Fig. 8 displays a time-line from the bottom of the core towards the present-day by using acombination of these two ratios.We observe in general that the sources

have shifted from pure wood and coal combustion sources in the bot-tom of the core towards mixed sources and petrogenic sources inmore recent times. This could indicate a gradual change from woodand grass combustion at the start of the PAH timeline, through thelater addition of coal combustion and towards amixing of these sourceswith petrogenic sources in more recent times.

5.2. PAH versus city history

Themost recent concentration peak (between about 10–30 cm bss)that represents the overall highest concentrations of PAH, stands outwith a high degree (Fig. 5) and concentration (Fig. 7) of HMW PAHs.The build up towards this peak is possibly a reflection of the degree ofurbanisation and the increased production in the local, bigger gasworks.The PAH high amount of heavy compounds strongly indicates high-temperature processes and the peak has thereby also a possible correla-tion with vehicle traffic, which increased dramatically in the 1950s and1960s (Fossen and Grønlie, 1985).

We observe that the PAH concentrations decrease in the most re-cent sediments. In other studies in the Bergen area where PAH hasbeen analysed in the top 50 cm of the sediment column (Eide et al.,2011; Kryvi and Grevskott, 2002), similar concentration patternswere observed. Maximum PAH concentrations are found at depthin the sediment, followed by decreasing values towards the presentday. In addition, the harbour sediments in Oslo, the capital of Norway,shows results that indicate a peak of PAH concentration between1910 and 1950 with levels of PAH, of pyrogenic origin, decreasing

Page 11: 2014-Science of the Total Environment 470–471 (2014) 1160–1172

Fig. 8. Plot presenting ratios between Fluoranthene/Pyrene and Benzo(a)anthracene/Chrysene. The numbers correspond to the depth in the core.

1170 M. Andersson et al. / Science of the Total Environment 470–471 (2014) 1160–1172

over the last decades (Cornelissen et al., 2008). Both studies determinea similar trend as the results from the near-surface samples within thisstudy. The decrease in PAH in sediments is thought to be due to a gen-eral decrease of PAH emissions in Norway as transition from coal gasifi-cation to hydroelectric power occurred during the 1950s (Cornelissenet al., 2008). In addition to a general decrease of emissions also localemissions decreased dramatically in the 1980s as the last gasworks inBergen closed, thereby concluding a period with a major point source.

As indicated earlier, many city fires have taken place in Bergen. Theburnt wood, coal and soot that were dumped into the water body is ex-pected to have spread large amounts of PAH into the lake and subse-quently the sediments and give rise to higher PAH concentrations. Theinput from urban fires is expected to be direct input, but also subse-quent, gradual input as on-land burnt refuse is eroded and transportedinto the lake. A few, large-scale fires may possibly be connected withhigher PAH concentrations in the sediments. The peak concentrationsat approximately 160 cm bss may be related to the extensive urbanfire in 1702 that destroyed around 85% of the habitated area in Bergen(Byen Brenner, 2002). An additional peak PAH value at approximately65 cm bss may be a result of the 1916 fire.

5.3. PAH versus other compounds

The occurrence of cyanide within the sediments may indirectlypinpoint the period of gaswork activity. This is supported by the factthat a higher relative concentration of heavy PAHs is present in thatparticular time period. The cyanide enters the aquatic environmentthrough liquid and solid waste from the gas purifying process. Thespent oxides, which originally were iron oxides that were used to re-move hydrogen sulphide and cyanide from the gas, were most possiblydumped on-site beside the gasworks. The concentration of cyanide inthe core increases from a steady concentration at 100–120 cm depthto a high peak at 92 cm depth (Fig. 4). Cyanide may be produced bycertain algae in the nitrate metabolism process (Kaufman et al., 2011),but algal blooms are not expected to produce the high cyanide concen-trations presented here. The high peak of cyanide is likely derived froma period when the gasworks site was closed and rehabilitated for newusage. It is possible that the solidwastewas deposited in the lake, there-by causing a large input of cyanides into the sediments. Therefore it is

possible that the high concentration peak of cyanide was depositedafter the cessation of gaswork activity. The authors are aware of thefact that the analysis of more closely spaced samples would provide amore accurate picture of the PAH distribution, but the analyses areonly used as indicators of the input processes at that time. The periodof gaswork activity is clearly pinpointed from the concentration oflead, arsenic, iron and mercury (unpublished data, Geological Surveyof Norway), which all display concentration peaks at the same specificinterval. These are all elements that are present in relatively high levelsin coal; therefore a major coal combusting point source as the gasworksshould be evident as higher concentrations of these elements (Reimannand de Caritat, 1998). When studying PAH in an urban environment,simultaneously looking at other compounds such as cyanide andheavy metals increases the amount of information that contributesto a more accurate timing of the sediments and therefore a morecomprehensive pattern of the PAH distribution.

The most recent peak of PAH argues for a significant input that isindependent of the input of TOC, since there is no correlation betweenthese two parameters. When the oxygen levels are low, the organicmatter degradation is very slow and the amount of organic matterremains high, which should lead to higher levels of organic matter inrecent sediments. This hypoxic state, however, is not expected to havebeen always present, since the lake was previously a sea inlet with seawater throughout the water column. Therefore it is possible that thelower TOC levels are not just a result of a longer degradation time, butalso of concentrations of oxygen in the bottom sediments when theinlet was part of the fjord system. The low correlation with grain sizealso suggests that the concentrations are independent of grain size dif-ferences. Thus this study does not follow studies for other compoundswhere changes in grain size contribute to a difference in the concentra-tion of metals, in that a reduction in grain size leads to an increase inmetal concentrations (e.g. Reimann et al., 1998).

The analytical results from the stormwater trap sediments clearlyshow that active PAH sources are present today, even though the con-centrations aremuch higher within the sediments.We consider wheth-er it is possible to determine and control these active sources. It is, as yetunknownwhat the nature of these sources are but they are likely to be acombination of traffic pollution, pollution from domestic heating andsome minor industrial sources. Kocbach et al. (2006) presented the dif-ficulties in distinguishing between these sources. These are all sourcesthat will not reduce over time. The city of Bergen is increasing theamount of inflow of untreated urban stormwater into the lake throughthe expansion of the drainage area. Therefore, a reduction in PAH con-centration at the surface sediment will not be possible as long as thesolids in the storm water are contaminated.

6. Conclusions

Combining results from sediments that span the time period frombefore urban settlement until today, it is possible to determine the de-gree of influence that habitation, urban fires, and major point sourcesof PAH have had on the environment. The PAH results correlate wellwith some historical events. In addition, we identify several definitesources that are apparent from the results. The onset of craft industriesand the usage of the lake as a filling place for burnt refuse can clearly beseen in the sediments aswell as the influence of the gasworks and trafficpollution. At the same time as the effect of the craft industries becameevident, we clearly detect changes in sources are through the replace-ment of LMW PAH compounds by HMW compounds. These data addto the understanding of the development from background values,which is essential when assessing remediation efforts.

The results show that it is possible to aid the 14C dating modellingby using other chemical proxies to pinpoint environmental changes,in other words, to perform a kind of “chemical dating”. For examplethe period of the gaswork activity displays higher PAH, cyanide and

Page 12: 2014-Science of the Total Environment 470–471 (2014) 1160–1172

1171M. Andersson et al. / Science of the Total Environment 470–471 (2014) 1160–1172

metal concentrations. In addition the same time period displays a higherrelative concentration of HMW PAHs.

Our study clearly shows that14C dating alone does not provide reli-able results within the period of anthropogenic influence and therefore,its use in urban environments during the Anthropocene, without theaddition of chemical proxies, should be employed tentatively.

We demonstrate that there are one or more presently active diffusesources of PAH in the urban environment contributing PAH to the lakevia urban stormwater.

Conflict of interest

There is no existing conflict of interest between the contents in themanuscript or the authors and other parties, (including any financial,personal or other relationships with other people or organisations).

References

Asplan Viak. Bybanen i Bergen, delprosjekt 1-reguleringsplaner. Forurenset grunnlangs bybanetraseen. (translated title from Norwegian: Bybanen in Bergen –

part 1 – development plan. Polluted soil along the Bybane route). Asplan Viakproject: 301106; 2003. [http://www3.bergen.kommune.no/bybane/bybanen_gml/10_Forurenset_grunn.pdf].

Björck S, Wohlfarth B. 14C chronostratigraphic techniques in paleolimnology. In: LastWM,Smol JP, editors. Tracking environmental changes using lake sediments; basin analy-sis, coring and chronological techniques. Dordrecht: kluwer Academic Publishers;2001. p. 205–47.

Boyle JF. A simple closure mechanism for a compact, large-diameter gravity corer.J Paleolimnol 1995;13:85–7.

Breedveld BD, Skei J, Hauge A. Contaminants in Norwegian fjord sediments: industrialhistory or future source? J Soils Sediments 2010;10:151–4.

Brown JN, Peake BM. Sources of heavy metals and polycyclic aromatic hydrocarbons inurban stormwater runoff. Sci Total Environ 2006;359:145–55.

Brenner Byen. www.bergenskartet.no http://www.histos.no/bergen/vis.php?visbrann=1,2002. [(03.09.2012)].

Choudhary P, Routh J. Organic geochemical record of increased productivity in LakeNaukuchiyatal, Kumaun Himalayas, India. Environ Earth Sci 2010;60(4):837–43.

City of Bergen. Handlingsplan for bedre luft i Bergen 2007 (translated title fromNorwegian: Action plan for better air in Bergen 2007); 2007 [https://www.bergen.kommune.no/bk/multimedia/archive/00098/Handlingsplan_for_be_98555a.pdf(03.09.2012)].

Cornelissen G, Pettersen A, Nesse E, Eek E, Helland A, Breedveld GD. The contribution ofurban runoff to organic contaminant levels in harbour sediments near twoNorwegiancities. Mar Pollut Bull 2008;56:565–73.

Davis AP, Shokouhian M, Ni S. Loading estimates of lead, copper, cadmium and zinc inurban runoff from specific sources. Chemosphere 2001;44:997–1009.

Duval MM, Friedlander SK. Source resolution of polycyclic aromatic hydrocarbons in theLos Angeles atmospheres: application of a CMBwith first order decay. U.S. EPA ReportEPA-600/2-81-161, Washington, DC; 1981.

Eide I, Berg T, Thorvaldsen B, Christensen GN, Savinov V, Larsen J. Polycyclic AromaticHydrocarbons in dated freshwater and marine sediments along the Norwegiancoast. Water Air Soil Pollut 2011;218:387–98.

EU. Water Note 4. Reservoirs, Canals and Ports: Managing artificial and heavily modifiedwater bodies. http://ec.europa.eu/environment/water/waterframework/pdf/water_note4_reservoirs.pdf, 2008.

EU, European Commission, Scientific Committee on Food. Polycyclic AromaticHydrocarbons — Occurrence in Foods, Dietary Exposure and Health Effects.http://ec.europa.eu/food/fs/sc/scf/out154_en.pdf, 2002.

Finstad A, Flugsrud K, Haakonsen G, Aasestad K. Vedforbruk, fyringsvaner ogsvevestøv. Undersøkelse om vedforbruk og fyringsvaner i Trondheim og Bergen2003 (translated title from Norwegian: Wood consumption, heating habits anddust in suspension. A study in the wood consumption and heating habits inTrondheim and Bergen 2003). Statistics Norway Report 2004/27; 2004.

Folk RL, Ward WC. Brazos river bar: a study in the significance of grain size parameters.J Sediment Petrol 1957;27:3–26.

Forfang I. Jordundersøkelse i barnehager i Bergenhus bydel, Bergen kommune (translatedtitle from Norwegian: Soil study in day-care centres in Bergenhus district, BergenCity). NIVA-report 5744-2009; 2009.

Fossen AB, Grønlie T. Bergen bys historie, bind IV (translated title from Norwegian:The history of Bergen, vol. IV. Bergen: Universitetsforlaget; 1985.

Fylkesmannen i Hordaland. Tiltaksplan for Bergen havn, fase II (translated title fromNorwegian: plan of action for Bergen harbour, phase II). http://hordaland.miljostatus.no/dm_documents/Bergen_havn_sluttrapport12_des_ol8egS53817es_F50uy.pdf, 2005.

Galarneau E. Source specificity and atmospheric processing of airborne PAHs: implica-tions for source apportionment. Atmos Environ 2008;42:8139–49.

Gjerstad J. Fra Kaien til Møllendal — og retur (translated title from Norwegian: FromKaien to Møllendal — and return). Bergen: Bodoni; 2005.

Gocht T, Moldenhauer K-M, Pūttmann W. Historical record of polycyclic aromatic hydro-carbons (PAH) and heavy metals in floodplain sediments from the Rhine River(Hessisches Ried, Germany). Appl Geochem 2001;16:1707–21.

Gromaire-Mertz MC, Garnaud S, Gonzalez A, Chebbo G. Characterisation of urban runoffpollution in Paris. Water Sci Technol 1999;39:1–8.

Guo J, Wu F, Luo X, Liang Z, Liao H, Zhang R, et al. Anthropogenic input of polycyclicaromatic hydrocarbons into five lakes in Western China. Environ Pollut 2010;158:2175–80.

Guo W, Pei Y, Yang Z, Chen H. Historical changes in polycyclic aromatic hydrocarbons(PAHs); input in Lake Baiyangdian related to regional socio-economic development.J Hazard Mater 2011a;187:441–9.

Guo J-Y, Wu F-C, Zhang L, Liao H-Q, Z R-Y, Li W, et al. Screening level of PAH in sedimentcore from lake Hongfeng, Southwest China. Arch Environ Contam Toxicol 2011b;60:590–6.

Harris CJ. Bergen i kart fra 1646 til vårt århundre (translated title from Norwegian: Mapsof Bergen from 1646 to our century). Bergen: Eide; 1991.

Haugland T, Ottesen RT, Volden T. Lead and polycyclic aromatic hydrocarbons (PAHs) insurface soil from day care centres in the city of Bergen, Norway. Environ Pollut2008;153:266–72.

Helle K. Bergen bys historie. Bind 1. Kongssete og kjøpstad fra opphavet til 1536(translated title from Norwegian: The history of Bergen, from Royal Estate andmarket town from the origin until 1536). Bergen: Alma Mater AS; 1995.

Hong SM, Candelone JP, Patterson CC, Boutron CF. Greenland ice evidence of hemispherelead pollution 2-millennia ago by Greek and Roman civilizations. Science 1994;265:1841–3.

Ikenaka Y, Eun H, Watanabe E, Kumon F, Miyabara Y. Estimation of sources and inflow ofdioxins and polycyclic aromatic hydrocarbons from the sediment core of Lake Suwa,Japan. Environ Pollut 2005;138/3:529–37.

Itoh N, Tamamura S, Kumagai M. Distributions of polycyclic aromatic hydrocarbons ina sediment core from the north basin of Lake Biwa, Japan. Org Geochem 2010;41:845–52.

JartunM, Ottesen RT, Steinnes E, Volden T. Runoff of particle bound pollutants from urbanimpervious surfaces studied by analysis of sediments from storm water traps. SciTotal Environ 2008;396:147–63.

Jensen HKB, Eggen OA, Frøland SL, Skårn JS. Polycyclic aromatic hydrocarbons inurban surface soil in Oslo, Bergen, and Trondheim, Norway: PAH16 levels, com-positions and ratios. In: Johnson CC, Demetriades A, Locutura J, Ottesen RT, edi-tors. Mapping the Chemical Environment of Urban Areas. Chichester, UK:Wiley-Blackwell; 2011.

Johansen SS, Skarheim HS. Lille Lungegårdsvann— En vurdering av saneringstiltak basertpå undersøkelser av utskiftningsforhold og vannkvalitet (translated title fromNorwegian: Lille Lungegårdsvann — an assessment of remediation based on studiesof water exchange and quality). Norwegian Institute for Water Research(NIVA)—report O-21/62; 1968.

Johnsen GH. Punktundersøkelse av Lille Lungegårdsvatn, Bergen kommune, 25.oktober2007 (translated title from Norwegian: Point study of Lille Lungegårdsvatn, BergenCity, 25th of October, 2007). Rådgivende biologer AS report 1082; 2008.

Johnsen GH, Brekke E. Overvåking av Lille Lungegårdsvatn, Bergen kommune, 2008.(translated title from Norwegian: Monitoring study of Lille Lungegårdsvatn, 2008).Rådgivende biologer AS report 1178; 2009.

Johnsen TM, BjerkengB,Molvær J, Nygaard E. Miljøvurderinger av utfylling av sprengsteini Store Lungegårdsvann (translated title fromNorwegian: Environmental assessmentfor the filling in of Store Lungegårdsvann with rocks). Norwegian Institute for WaterResearch report 3927-98; 1998.

Jung S, Arnaud F, Bonté P, Chebbo G, Lorgoux C, Wiarski T, et al. Temporal evolution ofurban wet weather pollution: analysis of PCB and PAH in sediment cores from lakeBourget, France. Water Sci Technol 2008;57(10):1503–10.

Katsoyiannis A, Terzi E, Cai QY. On the use of PAH molecular diagnostic ratios insewage sludge for the understanding of the PAH sources. Is this use appropriate?Chemosphere 2007;69:1337–9.

Kaufman MM, Rogers DT, Murray KS. Urban watersheds, geology, contamination andsustainable development. Boca Raton: Taylor and Francis Group; 2011.

Kocbach A, Li Y, Yttri KE, Cassee FR, Schwarze PE, Namork E. Physicochemical character-isation of combustion particles from vehicle exhaust and residential wood smoke.Particle and Fibre Toxicology 2006;3:1–10.

Khalili NR, Sceff PA, Holsen TM. PAH sourcefingerprints for coke ovens, diesel and gasolineengines, highway tunnels and wood combustion emissions. Atmos Environ 1995;4:533–42.

Kryvi H, Grevskott K. Tiltaksplan for Bergen havn (translated title fromNorwegian: Actionplan for Bergen harbor). Report for Fylkesmannen in Hordaland; 2002. [http://hordaland.miljostatus.no/dm_documents/Bergen_havn_sluttrapport_fase_1_62jKz.pdf].

Larsson P. Lille Lungegårdsvann, utvikling gjennom 10 år (translated title fromNorwegian: Lille Lungegårdsvann, development over 10 years). Report of Zoolog-ical institute. Norway: Bergen University; 1994.

Lehndorff E, Schwark L. Biomonitoring of air quality in the Cologne conurbation usingpine needles as a passive sampler—part II: polycyclic aromatic hydrocarbons (PAH).Atmos Environ 2004;38:3793–808.

Meehan S. The fate of cyanide in groundwater at gasworks sites in South-easternAustralia. PhD-thesisSchool of Earth sciences. The University of Melbourne;2000.

Mielke HW, Alexander J, Langedal M, Ottesen RT. Children, soils and health: How dopolluted soils influence children's health? Mapping the chemical environment ofurban areas. Chichester, UK: Wiley-Blackwell; 2011.

Morillo E, Romero AS, Maqueda C, Madrid L, Ajmone-Marsan F, Grcman H, et al. Soilpollution by PAHs in urban soils: a comparison of three European cities. J EnvironMonit 2007;9:1001–8.

Multiconsult. Nybygg Gulating lagmannsrett, miljøteknisk grunnundersøkelse (translatedtitle from Norwegian: New construction by Gulating lagmannsrett, environmentalassessment). Multiconsult report 610263-1; 2004.

Page 13: 2014-Science of the Total Environment 470–471 (2014) 1160–1172

1172 M. Andersson et al. / Science of the Total Environment 470–471 (2014) 1160–1172

Neff JM. Polycyclic aromatic hydrocarbons in the aquatic environment: sources, fates andbiological effects. England: Applied Science Publishers, Essex; 1979.

Nesje A. A piston corer for lacustrine and marine sediments. Arct Alp Res 1992;24:257–9.Nilsen LS. Coastal heath vegetation in central Norway; recent past, present state

and future possibilities. PhD thesisDepartment of Biology, Faculty of NaturalSciences and Technology. Trondheim, Norway: Norwegian University of Scienceand Technology; 2004.

Noteby. Bergen offentlige bibliotek utvidelse. Miljøtekniske grunnundersøkelser(translated title from Norwegian: Extending Bergen public library. Environmentalstudy of ground). Noteby report: 52082-1; 1998.

Ottesen RT, Volden T. Jordforurensning i Bergen (translated title from Norwegian: Soilpollution in Bergen). Report Geological Survey of Norway 99.022; 1999.

Ramsey CB. Deposition models for chronological records. Quat Sci Rev 2008;27:42–60.Reimann C, de Caritat P. Chemical elements in the environment. Factsheets for the

geochemist and the environmental scientist. Berlin, Heidelberg, New York, London,Paris, Tokyo, Hong Kong: Springer Verlag; 1998.

Reimann C, de Caritat P, Niskavaara H, Finne TE, Kashulina G, Pavlov VA. Comparison ofelemental contents in O-and C-horizon soils from the surroundings of Nikel, KolaPeninsula, using different grain size fractions and extractions. Geoderma 1998;84:65–87.

Rise L, Brendryen J. Leirinnhold i jordarter—en sammenlignende studie med vekt påmetodene Coulter Laser 200 og Sedigraph, og forslag til beregning av ekvivalentleirinnhold i prosent (translated title from Norwegian: Clay content in soils—acomparative study of Coulter Laser 200 and Sedigraph, suggesting a recalculation ofclay contents in precentage). Report Geological Survey of Norway 2013.012; 2013.

Seo JSS, Keum YSK, Li QX. Bacterial degradation of aromatic compounds. Int. J. Environ.Res. Public Health 2009;6:278–309.

Statistics Norway. Yearly book of statistics, 2010; 2012 [http://www.ssb.no/].Statsarkivet. http://www.arkivverket.no/arkivverket/Arkivverket/Bergen/Nettartikler/

Tekniske-arkiv#Notar, 2012. [03.09.2012].Van Metre PC, Mahler BJ, Furlong ET. Urban sprawl leaves its PAH signature. Environ Sci

Technol 2000;34:4064–70.Waldmann N, Anselmetti FS, Ariztegui D, Austin Jr JA, Pirouz M, Moyz CM, et al. Holocene

mass-wasting events in Lago Fagnano, Tierra del Fuego (541S): implications forpaleoseismicity of the Magallanes–Fagnano transform fault. Basin Res 2011;23:171–90.

Wang Z, Chen J, Qiao X, Yang P, Tian F, Huang L. Distribution and sources of polycyclicaromatic hydrocarbons from urban to rural soils: A case study in Dalian, China.Chemosphere 2007;68:965–71.

Wilcke W, Krauss M, Safronov G, Fokin AD, Kaupenjohann M. Polycyclic aromatic hydro-carbons (PAHs) in soils of the Moscow region. Concentrations, temporal trends andsmall-scale distribution. J Environ Qual 2005;34:1581–90.

YangH-H, LeeW-J, Chen S-J, Lai S-O. PAH emission fromvarious industrial stacks. J HazardMater 1998;60:159–74.

Yang Z, Tang Z, Shen Z, Niu J,WangH. One-hundred-year sedimentary record of polycyclicaromatic hydrocarbons in urban lake sediments from Wuhan, Central China. WaterAir Soil Pollut 2011;217:577–87.

Yunker MB, MacDonald RW, Vingarzan R, Mitchell RH, Goyette D, Sylvestre S. PAHs in theFraser River basin: A critical appraisal of PAH ratios as indicators of PAH source andcomposition. Org Geochem 2002;33:489–515.