temporal and spatial variabilities of atmospheric polychlorinated biphenyls (pcbs), organochlorine...

www.elsevier.com/locate/scitotenv

Science of the Total Environm

Temporal and spatial variabilities of atmospheric polychlorinated

biphenyls (PCBs), organochlorine (OC) pesticides and polycyclic

aromatic hydrocarbons (PAHs) in the Canadian Arctic:

Results from a decade of monitoring

H. Hunga,T, P. Blancharda, C.J. Halsallb, T.F. Bidlemanc, G.A. Sternd, P. Felline,

D.C.G. Muirf, L.A. Barrieg, L.M. Jantunena, P.A. Helmd, J. Maa, A. Konoplevh

aMeteorological Service of Canada, 4905 Dufferin Street, Downsview, ON, Canada M3H 5T4bEnvironmental Science Department, Lancaster University, Lancaster, Lancashire LA1 4YQ, UKcCentre for Atmospheric Research Experiments, 6248 Eighth Line, Egbert, ON, Canada L0L 1N0

dFreshwater Institute, 501 University Crescent, Winnipeg, MB, Canada R3T 2N6eAir ZOne, 2240 Speakman Drive, Mississauga, ON, Canada L5K 1A9

fNational Water Research Institute, Environment Canada, 867, Lakeshore Road Burlington, ON, Canada L7R 4A6gAtmospheric Research and Environment Programme, World Meteorological Organization, 7 bis, Avenue de la Paix,

BP2300, CH-1211 Geneva 2, SwitzerlandhSPA Typhoon, Lenin Avenue 82, Obninsk, 249038, Russia

Abstract

The Northern Contaminants Program (NCP) baseline monitoring project was established in 1992 to monitor for persistent

organic pollutants (POPs) in Arctic air. Under this project, weekly samples of air were collected at four Canadian and two

Russian arctic sites, namely Alert, Nunavut; Tagish, Yukon; Little Fox Lake, Yukon; Kinngait, Nunavut; Dunai Island, Russia

and Amderma, Russia. Selected POPs, including polycyclic aromatic hydrocarbons (PAHs), polychlorinated biphenyls (PCBs)

and organochlorine (OC) pesticides, were analyzed in both the gas and particulate phases.

This paper summarizes results obtained from this project in the past 5 years. Temporal trends were developed for

atmospheric PCBs and OCs observed at Alert using a digital filtration (DF) technique. It was found that trends developed with 5

years of data (1993–1997) did not differ significantly from those determined with 7 years of data (1993–1999). This implies that

with the DF technique, long-term trends can still be developed with less than 10 years of data. An acceleration in decline of OC

and PCB air concentrations was noted in 1999 for some compounds, although the reason is unknown. Monitoring efforts must

continue to assess the effect of this decline on the long-term trends of POPs in the Canadian Arctic.

Occasional high trans-/cis-chlordane ratios and heptachlor air concentrations measured at Alert between 1995 and 1997

suggests sporadic fresh usage of chlordane-based pesticides. However, significant decreasing trends of chlordanes along with

0048-9697/$ - s

doi:10.1016/j.sc

T Correspondi

E-mail addr

ent 342 (2005) 119–144

ee front matter. Crown Copyright D 2004 Published by Elsevier B.V. All rights reserved.

itotenv.2004.12.058

ng author. Tel.: +1 416 739 5944; fax: +1 416 739 5708.

ess: [email protected] (H. Hung).

H. Hung et al. / Science of the Total Environment 342 (2005) 119–144120

their chemical signatures has provided evidence that emission of old soil residues is replacing new usage as an important source

to the atmosphere.

Measurements of OC air concentrations conducted at Kinngait in 1994–1995 and 2000–2001 indicated faster OC removal at

this location than at Alert. This may be attributed to the proximity of Kinngait to temperate regions where both biotic and

abiotic degradation rates are faster.

The PAH concentrations observed at Alert mimic those at mid-latitudes and are consistent with long-range transport to the

Arctic, particularly for the lighter PAHs. A decline in particulate PAH was observed, similar to atmospheric sulphate aerosol and

can be attributed to the collapse of industrial activity in the former Soviet Union between 1991 and 1995.

Spatial comparisons of OC seasonality at Alert, Tagish, Dunai and Kinngait show elevated air concentrations of some

compounds in spring. However, elevated spring concentrations were observed for different compounds at different sites.

Potential causes are discussed. Further investigation in the atmospheric flow pattern in spring which is responsible for the

transport of POPs into the Arctic is required. OC and PCB air concentrations at Alert were found to be influenced by two

climate variation patterns, the North Atlantic Oscillation (NAO) and the Pacific North American (PNA) pattern. Planetary

atmospheric patterns must be taken into account in the global prediction and modelling of POPs in the future.

Crown Copyright D 2004 Published by Elsevier B.V. All rights reserved.

Keywords: Persistent organic pollutants (POPs); Arctic air monitoring; Spatial variations; Seasonality; Temporal trends; Climate variation

patterns

1. Introduction

Since the 1980s, there has been increasing evi-

dence that the seemingly remote and pristine environ-

ment of the Arctic has been contaminated by

anthropogenic contaminants, such as the persistent

organic pollutants (POPs), mercury and heavy metals.

As most of these chemicals are not used or emitted in

the Arctic region, they must have been transported

over great distances from their sources in the south.

The atmosphere has been regarded as one of the most

important and rapid route of transport for POPs to the

Arctic. Air samples collected in the Canadian and

European Arctic during the 1980s confirmed the

presence of these chemicals in Arctic air (Patton et

al., 1989, 1991; Pacyna and Oehme, 1988). Since

these contaminants tend to bioaccumulate through the

foodchain and has been known to be toxic to humans

and many organisms, a long-term monitoring project

for POPs in Arctic air was initiated in 1992 under the

Canadian-operated Northern Contaminant Program

(NCP). Weekly air samples (particulate and gaseous)

have been taken from stations in the Canadian and

Russian arctic for the determination of POP concen-

trations in the Arctic atmosphere. These atmospheric

pollutants include the polychlorinated biphenyls

(PCBs), polycyclic aromatic hydrocarbons (PAHs)

and various organochlorine pesticides (OCs) and

synthetic industrial compounds. Sampling sites

(sampling period) include: Alert, Nunavut (1992–

ongoing); Tagish, Yukon (December 1992–March

1995); Little Fox Lake, Yukon (July 2002–July

2003); Kinngait, Nunavut (March 1994–February

1996, October 2000–September 2002); Dunai Island,

Russia (April 1993–May 1995) and Amderma, Russia

(1999–2001). As part of the Arctic Monitoring and

Assessment Program (AMAP) Work Plan, data

collected under this project, as well as those collected

at various European Arctic monitoring sites (e.g.

Pallas, Finland; Storhofdi, Iceland; Zeppelin, Norway,

etc.), are presented at the international negotiation

table in order to achieve the appropriate restrictions on

release of pollutants of concern for the Arctic

environment.

Arctic atmospheric data for PCBs, PAHs and OC

pesticides through 1994 collected under this program

were presented for the sites of Alert (Nunavut), Tagish

(Yukon) and Dunai Island in Russia in the first

Canadian Arctic Contaminants Assessment Report

(CACAR) (DIAND, 1997). Results included a varia-

tion in the PCBs homologue pattern according to the

origin of air masses affecting the receptor sites. For

the OCs, the lack of spatial differences indicated

widespread atmospheric contamination while levels at

Tagish reflected relative proximity to North American

and Asian source regions. The particle/gas partitio-

ning of PAHs during the haze season at Alert was

successfully represented using the Junge–Pankow

H. Hung et al. / Science of the Total Environment 342 (2005) 119–144 121

model (Pankow, 1987; Halsall et al., 1997). The PAH

partitioning was also well correlated with vapour

pressure at Alert and Tagish, similar to temperate sites

around the Great Lakes. This demonstrated that

temperature dependent vapour pressure was a good

descriptor of gas/particle distribution of PAHs even at

extremely cold Arctic temperatures.

Since CACAR, several articles have been pub-

lished reporting results from the air monitoring

program under NCP. These are reported in CACAR

II (DIAND, 2003). In this paper, we intend to

summarize, as well as update, the results of this

program for the past few years. These include: (1) the

development of time trends from PCB and OC (1993–

1999) air concentrations obtained at Alert; (2)

evidence that bpast-useQ pesticides, such as chlordane

and heptachlor, might still be in use sporadically, but

general sources of such compounds have shifted from

fresh application to a greater proportion brecycledQfrom soil; (3) spatial variations in seasonality of OC

air concentrations observed at different Arctic sta-

tions; as well as (4) evidence of climate variation

patterns on atmospheric transport of POPs to the

Canadian High Arctic. Based on these results,

recommendations will be provided for future air

monitoring studies in Arctic regions.

2. Experimental procedures and data analysis

2.1. Sample collection and chemical analysis

The locations of the five sampling sites and a

sampling timeline are shown in Fig. 1. At each site,

weekly air samples were collected using a high-

volume air sampler. The sampling and analytical

procedures for this program have been given else-

where (Hung et al., 2002, Halsall et al., 1998, Stern et

al., 1997). In summary, each air sample represents

~13000 m3 of air sampled over 7 days, resulting in

approximately 52 samples per year. For each sample,

air was drawn through a 20 cm glass fibre filter (GFF)

and two polyurethane foam (PUF) plugs (each: 20 cm

diameter�4 cm thick), to collect the respective

particle and vapour fractions. Field blanks were

collected every 4 weeks by handling a PUF and filter

in the same manner as a sample but without airflow.

Collected samples were soxhlet extracted and the

resulting extracts were split. One half of the extract

was archived and the other subjected to analysis for

PAHs, PCBs and OCs.

2.2. Data handling and quality assurance

The system of data handling and quality assurance

has been described in details by Stern et al. (1997).

Quality assurance measures included the calculation of

method detection limits (MDLs) for each sampling

year (average blank value of each sampling medium

(filter or PUFs) plus three standard deviations from the

mean), inclusion of recovery and internal standards in

each sample extract, analysis of EPA standard reference

materials, routine reanalysis of extracts and participa-

tion in interlaboratory comparison studies. Air con-

centration data reported under this program have not

been blank- or recovery-corrected. A statistical data-

base in SAS called the Research Data Management and

Quality Control System (RDMQ) with an extensive

flagging system is used to track and calculate all air

concentration data and quality control information.

Although sampling started at Alert in 1992, OC

and PCB data from the first year have not been

included in analyses in this article due to the exceed-

ence of certain quality control criteria, which were

discussed in Stern et al. (1997). To retain the

maximum amount of information, samples with

concentrations below MDLs are included in all

analysis. Samples that were found to be below

instrument detection limits (IDL) were replaced with

2/3 IDL of the corresponding compound.

2.3. Temporal trends and apparent first order halflives

Atmospheric concentrations for semi-volatile pol-

lutants, including PCBs and OCs, often exhibit

seasonality. If a temporal trend is to be determined, it

is important that the intra-annual variability be

reduced. As part of the data analysis, a statistical time

series analysis technique (digital filtration, DF) was

employed to smooth any seasonality in the air data

(Nakazawa et al., 1997). For each compound, an

approximate long-term trend and an average seasonal

cycle were determined by fitting a smoothing Reinsch-

type cubic spline and Fourier components to the data,

respectively. Long- and short-term variations of the

trend and the seasonal cycle were then extracted using

Fig. 1. Location of sample sites and sampling timeline of program.


iterative digital filters with two cut-off periods: a short

term cut-off period, set to 4 months, and a long term

cut-off period, set to 24 months. The variabilities

longer than 4 months and shorter than 2 years were

extracted to obtain the overall seasonal cycle and the

variabilities longer than 2 years were extracted to

obtain the final long-term trend (Hung et al., 2001,

2002). Outliers, defined as data larger than 3 standard

errors away from the fitted curve, were successively

rejected after each iterative fit.

For compounds that showed a decline in air

concentration, an apparent first order halflife, t1/2,

can be estimated by dividing ln 2 with the negative

value of the linear regression slope of the trendline

between air concentrations, represented as the natural

log of partial pressure, P, and time. Long-term trends


and t1/2 of PCBs and OCs in Alert air (1993–1997)

have been reported previously by Hung et al. (2001,

2002). In this paper, we are updating these results with

2 more years of data, through the end of 1999.

Temporal trends were only developed using the

vapour phase concentrations because the majority of

PCBs and OCs in Arctic air samples were found in the

gas phase.

Results of the trend analyses are presented in

Sections 3.2.1 and 3.2.2 for the PCBs and OCs,

respectively. Also summarized in the OC section

(Section 3.2.2) are results of a study on the correlation

between chlordane isomers and enantiomers, and the

time trends of chlordane isomers in air based on

compiled data from 1984 to 1998 observed at four

Arctic stations, including Alert (Bidleman et al.,

2002).

Alert is the only location in this study with multi-

year air concentration data where long-term trends and

t1/2 can be developed. However, the Kinngait site

located in the Eastern Canadian Arctic has been re-

activated in October 2000 (Kinngait II) to compare

with results obtained in 1994–1996 (Kinngait I). As

this paper goes to press, only data from 2000 week 43

to 2001 week 15 from Kinngait II are available.

Changes in OC concentrations observed during the

same time period of the year between Kinngait I and II

are briefly discussed in Section 3.3.

2.4. Data for spatial comparisons of OCs

Spatial comparisons of OCs at Alert, Tagish and

Dunai between 1993 and 1994 have been reported

previously by Halsall et al. (1998). However, only 1

year of data from Dunai was available at that time.

Also, air concentration data from Kinngait have never

been reported before although there was a one-year

overlap in sampling period (March 1994–March

1995) with the other three sites. Weekly air samples

were taken at this site but only 19 samples were

analysed individually as weekly extracts while the

rest were analysed as four-week composites. An

update on the spatial comparison in the 1993–1995

co-sampling period is provided in Section 3.4,

focusing on the differences in OC seasonality

observed at the four sites.

3. Results and discussions

3.1. Annual mean concentrations

3.1.1. PCBs

The mean and range of APCB concentrations in both particulate and vapour phases for 102 congeners, as well

as a subset of the 10 AMAP congeners are listed in Table 1. As observed in CACAR (DIAND, 1997) for 1993, the

PCB levels for 1995–1996 were generally lower at Alert than those reported for other AMAP Arctic sites. For

example, annual means (1995 and 1996) for the 10 AMAP congeners were 16.9 and 19.4 pg/m3 at Storhofdi,

Iceland and 68.2 and 165 pg/m3 at Spitsbergen, Norway (AMAP, 1999). The mean vapour phase concentrations

observed at Amderma, Russia and Kinngait, Nunavut more recently (1999–2001) were similar to that observed at

Alert. A selected set of 25 samples collected at Amderma between 1999 and 2000 showed a mean vapour phase

concentration of 4.05 pg/m3 (AMAP subset). At Kinngait II, 12 samples collected between 20 November 2000 and

12 March 2001 showed a mean vapour phase concentration of 3.28 pg/m3 (AMAP subset). The mean vapour

phase concentrations of APCBs (102 congeners) in 1994, when the four NCP sites were operating simultaneously,

were 26, 27, 30 and 35 pg/m3 for Tagish, Kinngait I, Alert and Dunai, respectively. In comparison to the temperate

IADN sites of Eagle Harbour (APCB=128F12 pg/m3), Sleeping Bear Dunes (160F10 pg/m3) and Sturgeon Point

(315F20 pg/m3) over 1991 to 1995 (Hillery et al., 1997), levels observed at the four sites were 5- to 11-fold lower.

3.1.2. OCs

The annual average total concentrations (gas+particle) at Alert are shown in Fig. 2. h-HCH, heptachlor, cis-nonachlor, mirex, photomirex, endrin, trifluralin and the DDT group (aside from p,pV-DDE) were detected at very

Table 1

Annual mean and range of APCBs concentrations (pg/m3) of total PCBs (102 congeners) and the AMAP subset (10 congeners) in vapour

(polyurethane foam, PUF) and particulate (glass fibre filter, GFF) phases at Alert

Total PCBsa AMAP subsetb

PUF GFF PUF GFF

Mean (S.D.) Range Mean (S.D.) Range Mean (S.D.) Range Mean (S.D.) Range

Alert

1993 29 (18) 3.2–98 3.7 (1.9) 1.5–15 5.6 (4.6) 0.18–26 0.33 (0.70) 0.089–5.0

1994 30 (20) 8.0–150 3.7 (2.5) 2.6–21 5.8 (3.8) 1.0–25 0.27 (0.23) 0.15–1.6

1995 30 (14) 12–94 3.3 (0.8) 2.6–8.3 4.9 (3.1) 2.0–22 0.25 (0.18) 0.16–0.88

1996 26 (13) 7.5–100 3.4 (0.9) 1.9–8.2 4.2 (2.4) 0.69–17 0.30 (0.17) 0.11–0.97

1997 23 (6.7) 10–42 3.9 (2.2) 2.7–16 3.4 (1.2) 0.99–6.4 0.39 (0.64) 0.15–4.2

1998 25 (6.0) 14–42 3.5 (1.0) 1.6–8.9 3.9 (1.4) 2.5–9.0 0.29 (0.20) 0.093–1.2

1999 21 (43) 7.3–320 3.6 (0.64) 3.0–5.9 5.6 (19) 1.2–140 0.39 (0.30) 0.17–1.9

Kinngait I

1994c 27 (8.5) 14–46 3.1 (1.0) 1.7–6.6 5.4 (1.3) 2.8–8.4 0.18 (0.062) 0.096–0.38

1995d 34 (16) 21–72 2.9 (1.5) 0.82–5.3 5.4 (3.7) 1.3–14 0.40 (0.45) 0.047–1.21

Tagish

1993 18 (6.0) 8.4–37 4.2 (2.1) 2.0–15 3.4 (1.6) 0.45–8.0 0.28 (0.31) 0.11–2.2

1994 26 (42) 9.1–320 5.6 (12) 2.0–92 4.1 (5.5) 0.64–42 0.32 (0.71) 0.12–5.2

Dunai

1993 36 (18) 4.6–98 5.6 (1.7) 3.5–10 7.7 (4.6) 0.27–22 0.44 (0.43) 0.20–2.4

1994 35 (9.3) 22–69 4.9 (1.1) 3.5–9.0 5.9 (2.8) 2.4–18 0.35 (0.23) 0.20–1.4

a Sum of 102 PCB congeners. Refer to Stern et al. (1997) for congener numbers.b Sum of PCB 28, 31, 52, 101, 105, 118, 138, 153, 156 and 180.c Consists of 19 weekly samples and 5 four-weekly composite samples.d Consists of 11 four-weekly composite samples.


low levels (45% to 95% of all samples were below the method detection limit in gas phase). The relative levels

were generally: (AHCHs and HCB)N(Achlordanes and endosulfan I)NdieldrinNADDT.a-, g-hexachlorocyclohexanes (a-, g-HCHs) and hexachlorobenzene (HCB), which are OCs with the highest

vapour pressures, achieved the highest levels in all 7 years and annual average air concentrations observed at Alert

were similar to those found at Spitsbergen, e.g. in 1996, AHCHs and HCB were 63.7 and 63.4 pg/m3 at Alert, and

73.1 and 92.5 pg/m3 at Spitsbergen (AMAP, 1999). For Alert, HCB shows the highest percentage of breakthrough

(on average, 27.3% of all samples); thus, the concentrations might be underestimated. The breakthrough criterion

was defined as (amount of chemical on front PUF/amount on back PUF)N0.333 for each sample. For all the OCs

monitored at Alert, only the following compounds are significantly affected by breakthrough, i.e. greater than 10%

of all samples exceeded the breakthrough criterion: HCB, pentachloroanisole (17.0%) and tetrachloroveratrole

(15.1%). Most of the other OCs showed evidence of breakthrough in less than 5% of all samples.

Vapour phase concentrations of a-HCH from Kinngait II (November 2000–March 2001) (meanFS.D.=25.2F10.0 pg/m3) have shown lower levels than in 1994–1995 (reported as four-week composites)

(meanFS.D.=101F6.8 pg/m3) during the same period of the year. This decrease is expected since the use of

technical HCH (containing 55–80% a-HCH (Li, 1999)) has been banned by its two major users, China and the

former USSR in 1983 and 1990, respectively (Li et al., 1998).

After 1996, Alert endosulfan I concentrations exceeded Achlordanes (Fig. 2). Endosulfan is a current-use

pesticide while the chlordanes have been deregistered in most European and North American countries. Global

endosulfan emissions have increased continuously since it was first applied in the 1950s (Li and Macdonald, this

0.1

1

10

100

1993 1994 1995 1996 1997 1998 1999

Year

CA (

pg

/m3 )

HCHs HCB Chlordanes Endosulfan Dieldrin DDT

Fig. 2. Relative concentrations of selected OCs at Alert (1993–1999).


issue of Sci Total Environ). Li and Macdonald (this issue) have also reviewed the global usage and emission of

DDT. Although DDT has been banned or severely restricted in many countries, it is still being used in others,

especially those in the tropical region, to control insects that carry diseases such as malaria and typhus.

Summertime DDT/DDE ratios at Alert were of the order of 1–1.5 whereas for the Russian Arctic site of Amderma

(1999–2001), the DDT/DDE ratios were about 3. Technical DDT is made up of 80% p,pV-DDT and 20% o,pV-DDE(Ramesh et al., 1989). A larger DDT/DDE ratio has also been observed at Tagish in Western Canada and is linked

to trans-Pacific transport from Asia (Bailey et al., 2000).

Table 2 presents annual average and range of OC concentrations for Kinngait, Tagish and Dunai between 1993

and 1995. As Halsall et al. (1998) have found, concentrations of OCs were similar at all sites indicating a uniformity

in contamination in Arctic air. Spatial differences in seasonality of air concentrations will be discussed in Section 3.3.

3.1.3. PAHs

Fig. 3a and b presents the annual arithmetic means of APAH air concentrations (vapour+particle phase) at Alert

from 1993 to 2000 (APAH=16 compounds; acenaphthylene, acenaphthene, fluorene, phenanthrene, anthracene,

fluoranthene, pyrene, benzo[a]anthracene, chrysene, benzo[b]fluoranthene, benzo[k]fluoranthene, benzo[e]pyr-

ene, benzo[a]pyrene, indeno[1,2,3-c,d]perylene, dibenzo[a,h]anthracene, benzo[ g,h,i]perylene). The APAHvapour phase concentrations ranged from 113 to 516 pg/m3 and the particle phase concentrations ranged from

38 to 392 pg/m3.

Fig. 4 shows the profile of the PAH compounds in 1994 at Alert, Spitsbergen and Kinngait. The concentrations

observed at Spitsbergen were generally much higher than those found at Alert and Kinngait. For the dominant

compounds, including fluorene, phenanthrene, fluoranthene and pyrene, the annual mean concentrations were

higher at Spitsbergen by a factor of 2 to 3.

3.2. Temporal trends

3.2.1. PCBs

Fig. 5 shows the seasonal cycles and trends of PCBs 28, 31, 52, 101,138, 180 at Alert (1993–1999) determined

by the DF technique. Air concentrations were converted to partial pressure, P, using the ideal gas law. Also

Table 2

Arithmetic mean (range) of OCs in Arctic air (filter+PUF)

(pg/m3) Kinngait I Tagish Dunai

1994 1995 1993 1994 1993 1994

HCB 55.4 (15.6–123) 63.4 (28.1–123) 41.8 (10–138) 55.0 (0.34–901) 54.7 (18.6–173) 63.2 (32.0–90.9)

a-HCH 78.4 (19.3–193) 71.2 (31.6–112) 79.3 (24–160) 78.0 (0.25–757) 40.0 (0.63–76.9) 56.6 (11.2–144)

h-HCH 0.59 (0.03–2.39) 0.29 (0.04–0.57) 1.08 (0.09–3.06) 0.47 (0.02–7.99) 1.02 (0.07–4.73) 0.36 (0.06–1.99)

g-HCH 14.0 (2.25–39.3) 9.38 (3.07–26.8) 11.0 (3.33–25.1) 13.2 (0.18–138) 9.83 (3.68–22.7) 12.1 (2.68–35.7)

Heptachlor 0.05 (0.01–0.23) 0.06 (0.01–0.14) 0.06 (0.02–0.73) 0.05 (0.01–0.75) 0.05 (0.02–0.08) 0.05 (0.03–0.26)

Octachlorostyrene 0.97 (0.12–1.50) 0.62 (0.06–1.32) 0.44 (0.08–1.31) 0.89 (0.07–11.9) 0.60 (0.11–1.74) 0.86 (0.14–2.37)

Oxychlordane 1.15 (0.25–3.11) 0.58 (0.26–0.86) 0.83 (0.30–2.12) 1.05 (0.06–11.9) 0.80 (0.04–1.64) 0.83 (0.15–2.53)

t-Chlordane 0.69 (0.10–1.21) 0.50 (0.27–0.93) 0.53 (0.16–1.65) 0.66 (0.02–7.67) 0.40 (0.02–1.27) 0.59 (0.13–1.93)

c-Chlordane 1.49 (0.22–3.25) 0.89 (0.45–1.28) 1.22 (0.44–2.70) 1.30 (0.13–11.9) 0.96 (0.03–2.65) 1.00 (0.07–3.19)

t-Nonachlor 1.29 (0.62–2.95) 0.81 (0.30–1.51) 1.01 (0.33–2.12) 1.12 (0.04–10.0) 0.79 (0.02–2.11) 0.89 (0.04–3.21)

c-Nonachlor 0.16 (0.02–0.44) 0.07 (0.01–0.23) 0.13 (0.02–0.51) 0.12 (0.02–0.92) 0.12 (0.03–0.30) 0.08 (0.04–0.43)

Heptachlor epoxide 1.42 (0.03–2.50) 1.08 (0.18–1.55) 1.12 (0.04–2.19) 1.47 (0.04–16.2) 0.92 (0.03–2.13) 1.15 (0.11–2.98)

Dieldrin 1.63 (0.60–3.96) 1.45 (0.53–2.16) 0.93 (0.37–2.15) 1.28 (0.04–20.6) 1.11 (0.29–2.04) 1.46 (0.10–5.28)

o,pV-DDE 0.13 (0.02–0.47) 0.10 (0.01–0.31) 0.19 (0.04–0.86) 0.23 (0.03–1.16) 0.09 (0.04–0.28) 0.12 (0.04–0.51)

p,pV-DDE 0.42 (0.12–0.92) 0.49 (0.20–1.05) 0.53 (0.03–1.82) 0.71 (0.03–13.6) 0.36 (0.03–1.02) 0.38 (0.10–0.85)

o,pV-DDT 0.56 (0.05–1.40) 0.41 (0.17–1.20) 0.39 (0.04–1.05) 0.65 (0.04–7.19) 0.18 (0.05–0.76) 0.36 (0.06–2.24)

p,pV-DDT 0.37 (0.05–1.05) 0.29 (0.12–0.64) 0.17 (0.04–0.66) 0.22 (0.04–1.40) 0.14 (0.05–0.46) 0.33 (0.06–1.74)

Pentachloroanisole 3.80 (0.34–6.60) 3.13 (1.50–5.62) 2.56 (0.08–15.3) 3.99 (0.04–73.4) 2.92 (0.95–6.92) 3.27 (0.96–9.29)

Endosulfan I 5.81 (2.05–13.4) 5.64 (1.94–11.5) 5.76 (0.08–15.3) 8.33 (0.08–88.6) 2.99 (0.05–7.18) 6.25 (0.06–38.4)

Methoxychlor 0.23 (0.12–0.46) 0.12 (0.06–0.19) 0.31 (0.13–1.12) 0.40 (0.09–6.53) 0.41 (0.22–0.73) 0.37 (0.26–0.63)

Trichloroveratrole 0.16 (0.08–0.32) 0.08 (0.04–0.12) 0.20 (0.09–0.32) 0.30 (0.07–4.52) 0.44 (0.15–2.19) 2.24 (0.19–9.09)

Tetrachloroveratrole 2.06 (0.16–4.74) 2.02 (0.90–4.19) 1.96 (0.17–6.71) 1.74 (0.07–20.6) 1.49 (0.15–3.65) 2.06 (0.20–5.86)

Endrin 0.20 (0.08–0.70) 0.15 (0.05–0.25) 0.22 (0.09–0.44) 0.29 (0.07–4.52) 0.28 (0.15–0.51) 0.27 (0.18–0.59)

Trifluralin 0.10 (0.05–0.21) 0.05 (0.03–0.08) 0.13 (0.06–0.21) 0.18 (0.04–2.92) 0.18 (0.09–0.33) 0.16 (0.11–0.28)


included in Fig. 5a is the weekly average temperature recorded at the sampling site. The seasonal cycles of the

congeners likely resulted from the temperature dependency of atmospheric PCB concentrations as well as

photochemical and radical (OH) reactions during the Arctic summer. This is in contrast to what has been reported

by Stern et al. (1997) who have observed no apparent seasonal cycles in 2 years of Alert data (93–94). However,

this is simply a result of the longer time series now available.

With 2 additional years of data, the findings are similar to that reported by Hung et al. (2001). By comparing the

trends, it was noted that the lower chlorinated congeners, PCB 28, 31 and 52, declined from 1993 to 1998 (Fig. 5).

No clear trend was apparent for the higher chlorinated congeners, except for PCB 180. However, in 1999, the

decline seems to have accelerated for some congeners with no apparent reason. This is especially evident for PCBs

28, 101 and 138.

In spite of the abrupt decline in 1999 for some congeners, apparent first order halflives, t1/2, determined with 2

more years of data are similar to those estimated with data from 1993 to 1997 (Table 3). The strongest correlations

between concentration and time occurred for the lighter congeners (r2 ranged from 0.53 to 0.86), with evidence of

declining trends for these compounds over the 7 years of monitoring. The r2 values for the heavier congeners were

generally lower (b0.2), except for PCB 180, which showed a highly correlated decreasing trend (r2=0.78). No

pattern was apparent between halflives and chlorine numbers, with halflives ranging from 4 to 79 years. The

halflives at Alert were typically longer than those derived from temperate background/rural sites from the early

1990s (Table 3).

According to Hung et al. (2001), the lack of a declining PCB temporal trend for the more chlorinated

congeners may be due to several factors. Lower Arctic temperatures and winter darkness will slow both biotic and

abiotic degradation relative to temperate regions. Global fractionation (Wania and Mackay, 1996), a process

(a) Vapour Phase

0

100

200

300

400

500

600

1992 1993 1994 1995 1996 1997 1998 1999 2000

Year

Year

An

nu

al A

rith

met

ic M

ean

(p

g/m

3 )A

nn

ual

Ari

thm

etic

Mea

n (

pg

/m3 )

(b) Particle Phase

0

50

100

150

200

250

300

350

400

1992 1993 1994 1995 1996 1997 1998 1999 2000

Fig. 3. Annual arithmetic mean of APAHs (16 compounds) in a) vapour phase and b) particulate phases at Alert.


where chemicals may be latitudinally dfractionatedT according to the ambient temperature and their physical–

chemical properties, may help to sustain levels in Arctic air, while concentrations in temperate regions are

declining. The apparent reduction in air concentrations of the lighter congeners over the years is probably the

result of the dominating effect of long range transport, whereby levels in the Arctic echo falling air concentrations

in source regions. This, in combination with their relatively fast OH depletion rates during the arctic summer, may

explain the temporal decline of the lighter congeners. For example, at an average temperature of �17.6 8C at

Alert, the reaction rate constants of PCB 31 (2,4V,5-trichlorobiphenyl) and 110 (2,3,3V,4V,6-pentachlorobiphenyl)would be 8.0�10�13 and 3.4�10�13 cm3/s, respectively (estimated using Arrhenius parameters given by

Anderson and Hites, 1996); combining with an estimated OH radical concentration of 1.7�105 molecules/cm3 in

the Arctic atmosphere in April (Barrie et al., 1988), the lifetime of PCB 31 would be 85 days compared to 202

days for PCB 110. However, the picture is complicated by notable exceptions, for instance PCB 180

(heptachlorinated) shows a clear downward trend and has a calculated halflife of only 4.2 years at Alert. It is

believed that heavier congeners might be subjected to different removal processes than the lighter ones.

Precipitation or scavenging processes, e.g. particulate deposition and snow scavenging, might be the dominant

removal processes for the heavier congeners, while the lighter congeners were mainly removed by degradation via

reaction with OH radicals. The relative removal rates pertaining to the different processes and the physical–

chemical properties of the congeners may be responsible for the differing trends observed.

acen

apht

hyle

ne

acen

apht

hene

fluor

ene

phen

anth

rene

anth

race

ne

anth

ene

pyre

ne

rete

ne

B[a

]A

chry

sene

B[e

]P

B[a

]P

Ind-

P

B[g

hi]P

dibe

nzth

ioph

ene

pery

lene

25

50

75

100

125

150

175

200

225

250

500

525

Alert SpitzbergenKinngait

l

r anth

ene

pyre

ne

rete

ne

B[a

]A

chry

sene

B[e

]P

B[a

]P

Ind-

P

B[g

hi]P

dibe

nzth

ioph

ene

pery

lene

Arit

hmet

ic M

ean

Con

cent

ratio

ns (

pg/m

3 )

0

550

Alert SpitzbergenKinngait

fluor

fluor

Fig. 4. PAH profiles at Alert, Spitsbergen and Kinngait in 1994. Kinngait data are arithmetic mean concentrations from Week 16, 1994 to Week

15, 1995.


3.2.2. OCs

Fig. 6 shows the seasonal cycles and trends of selected OCs. Most of the selected OCs showed declining trends

over the 7 years. Similar to the PCBs, temporal trends and halflives (Table 3) developed with 7 years of OC

concentration data (1993–1999) do not show significant differences from those developed with only 5 years of data

from 1993 to 1997, except for a-HCH. Two years of additional data resulted in a shortening of the t1/2 for a-HCH,

a major component of technical HCH, from 17 to 9 years. This probably reflects the continuous decline in global

emission of technical HCH as many countries have banned its use since the 1970s (Li and Macdonald, this issue).

From Fig. 1 of Li and Macdonald (this issue), it can be seen that global a-HCH emission continued to decline

throughout the 1990s. Another component of technical HCH, h-HCH, has shown rapid decline in Arctic air

concentrations between 1993 and 1995 (median concentrations: 0.45, 0.27 and 0.07 pg/m3 for 1993, 1994 and

1995, respectively. Median concentrations are reported here due to high percentage of non-detects). Its

concentrations then remain steady till the end of 1999 (annual median concentrations range from 0.04 to 0.09 pg/

m3) with greater than 70% of all samples below method detection limits after 1995. Since its concentrations were

non-detectable in many samples, a valid halflife and trend cannot be determined for h-HCH.The accelerations in decline in 1999 are not only noted for the PCBs but are also observed for the OCs in

Arctic air. Among the selected OCs shown in Fig. 6, this phenomenon is especially apparent for trans-chlordane

(TC), trans-nonachlor (TN) and dieldrin, but is also shown in the trends of a- and g-HCH, cis-chlordane (CC) and

endosulfan I. Since the analysis has been performed in the same laboratory using the same equipment and

analytical procedures by the same staff and the phenomenon was observed for both PCBs and OCs, this

acceleration in decline appears to be real.

Dieldrin and endosulfan I, which did not show a decline in trends when data from 1993 to 1995 were used,

showed halflives of 13 and 38 years, respectively, with the 2 additional years of data. For dieldrin, this is evidently

the result of the downturn in air concentrations starting early in 1998 and continued until the end of 1999. Before

then, dieldrin levels in Alert air had been fairly steady between 1994 and 1997. For endosulfan I, the level also

-30

-29

-28

-27

-26

-25

-24

-23

-30

-29

-28

-27

-26

-25

-24

-23

ln P

PCB 31b)

-27

-26.5

-26

-25.5

-25

-24.5

-24

-23.5

-23

ln P

-110-100-90-80-70-60-50-40-30-20-1001020

Tem

per

atu

re (ºC

)

PCB 28a)

ln P

1993 1994 1995 1996 1997 1998 1999

PCB 52c)

-31

-30

-29

-28

-27

-26

-25

-24

-23

ln P

1993 1994 1995 1996 1997 1998 1999

PCB 180f)

-31

-30

-29

-28

-27

-26

-25

-24

-23

-22

ln P

ln P

PCB 138e)-30

-29.5

-29

-28.5

-28

-27.5

-27

-26.5

-26

-25.5

-25

PCB 101d)

Seasonal Cycle Trend Measured Temperature

Fig. 5. Trends and seasonal cycles of PCBs.


declined from mid-1998 to end of 1999 but the extent was not as significant as that observed for dieldrin. The

concentration of endosulfan I in the Arctic has been relatively steady throughout the sampling period. This may be

because endosulfan is a current-use pesticide, but dieldrin and aldrin, which converts quickly to dieldrin in the

environment, have been banned in most western industrialized countries.

The halflives of the OCs are again found to be similar to those observed at the three Great Lakes sites (Table 3).

Since most OCs were still being used within the last decade, changes in atmospheric levels of the OCs in the Arctic

appear to mirror those in temperate source regions.

Only slight differences were found between t1/2 derived for both PCBs and OCs in the current study and those

determined previously with 5 years of data (Table 3), despite the sudden acceleration in decline in 1999. This

implies that with the DF technique perhaps 5 to 7 years of data are adequate for the derivation of long-term time

trends at remote background locations, like those in the Arctic, where concentrations are not affected by high

concentration episodes from sources. However, with the data that are currently available to us, it is not possible to

assess whether the acceleration in declining rates observed for most OCs and PCBs in 1999 will continue and what

Table 3

Halflives of various POPs at Alert compared to those at temperate locations

PCBs Alert (93–99) Alert (93–97) PCBs Sleeping Bear Dunesa Sturgeon Pointa

t1/2(years)b

r2 t1/2(years)b

r2 t1/2(years)b

rel SE

(%)ct1/2(years)b

rel SE

(%)c

28 10 0.55 12 0.58 28+31 2.7 22 3.0 27

31 6.9 0.86 6 0.87 – – – – –

52 4.0 0.86 3 0.86 52 4.3 30 7.5 56

101 7.3 0.53 11 0.73 101 2.9 24 4.0 32

105 11 0.21 ND 0.021 118 2.1 19 2.1 24

118 ND 0.03 ND 0.093 105+132+153 1.8 22 1.8 17

138 ND 0.01 ND b0.01 138+163 2.2 23 2.8 28

153 79 0.01 17 0.18 – – – – –

180 4.2 0.76 4 0.78 180 1.7 18 3.7 46

OC Alert (93–99) Alert (93–97) Eagle Harbord Sleeping Bear Dunesd Sturgeon Pointd

t1/2(years)b

r2 t1/2(years)b

r2 t1/2(years)b

rel SE

(%)ct1/2(years)b

rel SE

(%)ct1/2(years)b

rel SE

(%)c

a-HCH 9.1 0.80 17 0.79 4.4 4.5 3.5 5.7 3.3 6.1

g-HCH 5.7 0.86 4.9 0.88 9.1 14 4.6 8.7 5.4 7.4

Oxychlordane 4.1 0.90 3.5 0.84 – – – – – –

t-Chlordane 5.5 0.67 8.3 0.48 6.9 42 5.2 33 3.2 18

c-Chlordane 6.1 0.70 4.1 0.71 23 110 9.7 50 5.9 25

t-Nonachlor 6.2 0.73 6.2 0.71 33 160 6.0 35 5.0 30

c-Nonachlor 6.6 0.55 7.9 0.29 – – – – – –

Heptachlor epoxide 8.3 0.70 6.3 0.72 – – – – – –

Dieldrin 13 0.21 ND b0.10 3.7 18 2.4 17 2.9 21

Endosulfan I 38 0.06 ND 0.25 – – – – – –

a Simcik et al. (1999) Supporting Information. t1/2 calculated using multiple linear regression method. Only t1/2 calculated with airmass

originating over land is presented.b t1/2=halflife, calculated as ln 2/slope. Compounds that do not show a consistent declining trend (i.e. greatly fluctuating or steady or slightly

increasing trends) are indicated as ND=not determinable. In such cases, r2 is the linear regression coefficient of the slope of ln P versus time.c rel SE (%)=Relative Standard Error (%).d Cortes and Hites (2000), except for a- and g-HCH. The study periods were November 1990 to December 1997 for Eagle Harbor, December

1991 to December 1997 for Sleeping Bear Dunes and Sturgeon Point. Halflives that are statistically significant with 95% confidence are

indicated in normal fonts, those that are not significant are in italics. For a- and g-HCH, halflives derived with data up to the end of 2001 were

obtained from Buehler et al. (2004).


effect this might have on the long-term trends. Therefore, atmospheric monitoring efforts should continue in the

Canadian Arctic in order to achieve this purpose.

Hung et al. (2002) have noted an increase in the ratio of the trans- and cis-isomers of the deregistered pesticide

of chlordane, TC/CC, between 1993 and 1997. Several episodes of TC/CC exceeded 1 were observed during that

period. The ratio of TC/CC in technical chlordane mixtures used in the US is 1.16 (Jantunen et al., 2000; Mattina et

al., 1999). The average ratios in the Arctic atmosphere have been reported as b1 in recent years (Bidleman et al.,

2002; Oehme et al., 1996). This indicates a more weathered chlordane profile in general to be affecting the Arctic

compared to previous decades since the trans-isomer is relatively less stable and more susceptible to degradation

by microorganisms in soil (Beeman and Matsumura, 1981). Episodes of TC/CC N1 measured in Arctic air imply

sporadic fresh application of chlordane-based pesticides in the south. Fig. 7 shows a plot of TC/CC versus time

with the seasonal cycle and trend developed by DF.

Superimposed on this plot is the concentration of heptachlor (gas and particle phases) measured at Alert.

Heptachlor concentrations measured at Alert were usually very low and mostly below detection limits prior to

-24

-23

-22

-21

-20

-19

-18

ln P

ln P

ln P

ln P

ln P

ln P

ln P

-110

-90

-70

-50

-30

-10

10

30

Tem

per

atu

re (

C)

-28

-27

-26

-25

-24

-23

-22

-21

-20

-19

γ-HCH

-30

-29

-28

-27

-26

-25

-24

-23

-22

t-chlordane

-30

-29

-28

-27

-26

-25

-24

c-chlordane

b)

c)

d)

a) α-HCH

-30

-29

-28

-27

-26

-25

-24

ln P

t-nonachlor

-31

-30

-29

-28

-27

-26

-25

-24

p,p’-DDE

-31

-30

-29

-28

-27

-26

-25

-24

Dieldrin

-30

-29

-28

-27

-26

-25

-24

-23

-22

Endosulfan I

h)

g)

f)

e)

1993 1994 1995 1996 1997 1998 1999 1993 1994 1995 1996 1997 1998 1999

Seasonal Cycle Trend Measured Temperature

Fig. 6. Trends and seasonal cycles of (a) a-HCH, (b) g-HCH, (c) t-chlordane, (d) c-chlordane, (e) t-nonachlor, (f) p,pV-DDE, (g) dieldrin and (h)endosulfan I at Alert (1993–1999).


1995, but increased to relatively higher levels (0.2–0.3 pg/m3) during the winter months of 1995/1996 and 1996/

1997 (Fig. 7). If heptachlor were truly a bpast-usedQ pesticide, the residue should mainly be degraded to

heptachlor epoxide and there should not be fresh peaks of heptachlor. Karlsson et al. (2000) have observed cyclic

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

TC

/CC

0

0.1

0.2

0.3

0.4

0.5

0.6

Hep

tach

lor

con

c. (

gas

+ p

arti

cle)

(p

g/m

3 )

Seasonal Cycle Trend Measured Heptachlor

1993 1994 1995 1996 1997 1998 1999

Fig. 7. TC/CC and heptachlor concentration at Alert (1993–1999).


temporal patterns of heptachlor at Senga Bay, South Africa, between 1997 and 1998 which indicates that there

has been repeated use of heptachlor rather than technical chlordane. Therefore, fresh inputs into the Arctic

atmosphere are not surprising. Nonetheless, technical chlordane mixtures contain some heptachlor (~7%) (Gooch

et al., 1990) and technical heptachlor is contaminated with chlordane isomers, especially TC (Bidleman et al.,

2002). When the TC/CC peaked together with heptachlor, for instance 5 to 12 February 1996, it may suggest

fresh usage of either heptachlor or chlordane mixtures. Five-day back trajectories calculated by Hung et al. (2002)

at 925 hPa during the periods when TC/CC N1 indicated that Alert was subjected to stronger than normal

influence from the Eurasian sector but the actual source regions of the fresh input episodes cannot be determined.

In the last 2 years of 1998 and 1999, the TC/CC ratio apparently declined at Alert (Fig. 7). After 1997, there

was only one episode that TC/CC exceeded 1. Heptachlor concentrations also remained low for most of 1999.

Velsicol, the sole producer of technical chlordane and heptachlor, has stopped its production of both pesticides in

May 1997 (PANUPS, 1997). Therefore, it can be expected that the TC/CC would generally remain less than 1 in

Arctic air as the technical chlordane stockpile depletes.

Bidleman et al. (2002) have compiled TC, CC and TN measurements in Arctic air from 1984 to 1998 at the

monitoring stations of Alert (1993–1997), Tagish (1993–1994), Dunai (1993–1994) and the AMAP station of

Pallas, Finland (68858V N, 24 807V E) (1996–1998). (Data from Pallas was kindly supplied by Eva Brorstrfm-

Lunden (Swedish Environmental Research Institute) and Sirkka Juntto (Finnish Meteorological Organisation).)

The data set contains over 500 measurements, 70% from 1993 to 1998 and 30% from earlier. Significant declines

in TC, CC and TN concentrations over the 14-year period in both summer–fall (May–October) and winter–spring

(November–April) were observed ( pb0.001). Apparent values of halflives, t1/2, obtained by regressing the natural

logarithm of concentration versus year (first order plot) were (summer–fall, winter–spring) 4.9, 6.4 years (TC), 6.7,

9.7 years (CC) and 7.4, 6.3 years (TN). They are similar to the time trends observed between 1993 and 99 at Alert,

which were 5.5 years for TC, 6.1 years for CC and 6.2 years for TN (Table 3).

As seen in Fig. 7 and in other studies over many years (Bidleman et al., 2002; Halsall et al., 1998; Oehme et

al., 1996), the ratio of TC/CC undergoes seasonal cycles in Arctic air, higher in winter–spring and lower in

summer–fall. Between 1984 and 1997, the fraction of TC=FTC=TC/(TC+CC) declined significantly in both

winter–spring ( pb0.001) and summer–fall ( p=0.014) seasons. This again suggests that sources of chlordane in

Arctic air have changed over the last 14 years and are now more depleted in TC. The FTC in technical chlordane


is 0.54, and when weighed for the slightly higher vapour pressure of TC the expected atmospheric FTC in

equilibrium with technical chlordane is 0.62 (Jantunen et al., 2000). The FTC in soils is about the same or lower

than in technical chlordane (Aigner et al., 1998; Jantunen et al., 2000). Confounding the interpretation of FTC

values is the fact that technical heptachlor was also used in agriculture and as a termiticide. As mentioned above,

technical heptachlor is contaminated with chlordane isomers, especially TC, so higher FTC in soils or air could be

associated with this source.

In two studies, Bidleman et al. (2002, 2004) have used chlordane enantiomers in air samples from the arctic

stations at Alert (1993–1996 and summer of 1999); Dunai (mostly 1994–1995, with a few from 1993 and 1996);

Pallas (1998, 2001); and a European Monitoring and Assessment Program (EMEP) station at Rfrvik (1998, 2001)

on the Swedish west coast (57825VN, 11856VE) as an indicator of historical residue emissions, since non-racemic

chlordanes have been found in soils (Aigner et al., 1998; Jantunen et al., 2000; Meijer et al., 2001; Wiberg et al.,

2000) and in ambient air from the Great Lakes region (Bidleman et al., 1998; Ulrich and Hites, 1998). Depletion of

the (+)TC and (�)CC enantiomers was found for most of the samples. The average enantiomer fraction, EF=(+)/

[(+)+(�)], of TC was different from racemic (racemic EF=0.500) at a high level of significance ( pb0.001) at Alert

(1993–1996: 0.471F0.011; 1999 summer: 0.470F0.0090), Pallas (1998: 0.477F0.0061; 2001: 0.471F0.012),

Dunai (1994–1995: 0.471F0.0087) and the temperate coastal station Rfrvik (1998: 0.463F0.013; 2001:

0.454F0.026). The average EFs of CC were also significantly non-racemic at Rfrvik (1998: 0.511F0.0054,

pb0.001; 2001: 0.514F0.0071, pb0.001) and Pallas (1998: 0.507F0.0050, pb0.01; 2001: 0.510F0.0042,

pb0.001) but not at Dunai (1994–1995: 0.503F0.0046, pN0.2) and Alert (1993–1996: 0.504F0.0048, pN0.1; 1999

summer: 0.507F0.0043, pN0.2). At Alert (1993–1996), the EF of TC was positively correlated with FTC (r2=0.41,

p=0.0032), but not at Dunai (r2=0.0024, pN0.9) nor Pallas (r2=0.029, pN0.6). The mean EFs at each station in

1999/2001 (Bidleman et al., 2004) were not significantly different ( pN0.05 to N0.2) from those obtained during the

previous campaign (Bidleman et al., 2002). However, there appeared to be a slight drift toward more nonracemic

signatures in 2001.

Have chlordanes in Arctic air always been non-racemic, even in the mid-1980s and earlier? To examine this

question, the EFs of chlordanes were determined in archived samples of atmospheric deposition that were collected

in Sweden, Iceland and Slovakia during 1971–1973 (Bidleman et al., 2004). The average EFs of TC and CC in the

historical samples were 0.496F0.0025 and 0.500F0.0027, racemic for CC and nearly so for TC. Sediment layers

from Lake DV-09 on Devon Island in the Canadian Archipelago also showed a trend in the EFs of TC, from more

nearly racemic in the 1950s and 1960s to less racemic in recent times (Stern et al., this issue; Bidleman et al.,

2004).

Heptachlor exo-epoxide (HEPX) was significantly ( pb0.001) non-racemic in the 1994–1998 air samples from

all stations and showed an excess of the (+) enantiomer. Sources of HEPX include metabolism of heptachlor in

soils and photochemical oxidation of heptachlor. The former process is thought to yield HEPX enriched in the (+)

enantiomer, as occurs when heptachlor is metabolised by rat liver microsomes, while the second yields racemic

HEPX (Buser and Muller, 1994). HEPX in ambient air samples from temperate regions is also non-racemic,

suggesting that emission from soils is the primary source (Bidleman et al., 1998).

Although occasional spikes of high TC/CC and heptachlor air concentrations were seen at Alert between 1995

and 1997 suggesting that these pesticides might still be in use sporadically, from the study of enantiomer fractions,

it appears that general sources of chlordane have shifted over time, from atmospheric transport of the freshly

applied pesticide to a greater proportion of chlordane brecycledQ from soils. The non-racemic HEPX in Arctic air

exemplifies the volatilisation of soil residues on a large scale and influence on atmospheric levels in remote regions.

3.2.3. PAHs

Attempts were made to extract trends using the DF technique from the entire PAH dataset and from segregated

winter/spring data, but the data were extremely scattered and reliable trends could not be derived. This is mainly due

to the extreme seasonality of PAH concentrations at Alert (Halsall et al., 1997), more years of data are required to

obtain reliable temporal trends. Nonetheless annual means (Fig. 3a and b) can be used to discuss general tendencies.


The decrease observed at Alert for gas phase PAH (Fig. 3a) between 1993 and 1996 follows measurements

around the Great Lakes (Cortes and Hites, 2000) and at urban sites in the UK (Halsall et al., 1994, Coleman et al.,

1997). The UK decreasing trends have been attributed to a substantial decline in domestic consumption of fossil

fuels for space heating and improvement in domestic and industrial combustion devices. Given the importance of

long range transport of PAHs to the Arctic (Halsall et al., 2001), it is not surprising that the levels observed at Alert

mimic the declining trends at mid-latitudes. The decline in particulate phase PAHs from 1993 to 1996 (Fig. 3b) is

consistent with the general declining trend of atmospheric sulphate aerosol at Alert after 1991 reported by Sirois

and Barrie (1999). Particulate-sulphate accounts for the majority of the aerosol loading in the Arctic (Halsall et al.,

1997) and the size distribution of sulphate (V1 Am) corresponds to the range where most PAHs have been observed

(Allen et al., 1996). Sirois and Barrie (1999) have shown that between 1985 and 1990, there was a lack of decline

in sulphate concentration at Alert and Barrow reflecting emission data from the former Soviet Union. Between

1991 and 1995, they have attributed the decline in sulphate concentrations to the collapse of industry during the

early years of the new Eurasian republics. It is hypothesized that the decline in PAH air concentrations observed at

Alert could also be due to such a collapse in industrial activities or, alternatively, it could be caused by a change in

emission source and/or transport pathway which influence the site. As a result of global climate change,

contaminant transport pathways may change accordingly. The potential influence of climate change on the

transport of contaminants to the Arctic is discussed in detail in Macdonald et al. (this issue). Longer air

concentration time series of PAHs are required in order to test these hypotheses and to determine the cause behind

their decreasing concentrations observed in Arctic air.

The observed profiles at the three Arctic stations of Alert, Spitzbergen and Kinngait (Fig. 4), dominated by

lighter compounds such as fluorene, phenanthrene, fluoranthene and pyrene, confirm the modelling result

presented by Halsall et al. (2001). They have found that lighter compounds, such as fluorene and phenanthene

would be able to reach the Arctic while most of the heavier compounds, such as benzo[a]pyrene (B[a]P), would

have been removed from the atmosphere before reaching the Arctic, either by reaction with OH radicals or

deposition along the transport pathway.

3.3. Kinngait—changes in OC air concentrations (2000–2001 vs. 1994–1995)

Although air monitoring at the site of Kinngait has not been continuous, it is still possible to compare the relative

changes in OC air concentrations at this location and those observed at Alert through the ratios of air concentrations

measured recently and those observed in 1994–1995. These bNow-to-ThenQ ratios, N:T, are defined as:

N : T ¼ Now Concentration ðKinngait : 2000� 2001; Alert : 1998� 1999ÞThen Concentration ð1994� 1995Þ ð1Þ

Four-week composite concentrations at Kinngait were calculated from the 2000–2001 weekly data for

comparison with monthly composite concentrations measured during the same weeks in 1994–1995. N:T ratios of

2000–2001 to 1994–1995 composite concentrations of the HCHs and chlordanes are presented in Fig. 8. Also

shown in Fig. 8 are the ratios of the 1998–1999 to 1994–1995 composite concentrations for Alert. Alert 2000–

2001 air concentrations are still under analysis and are not available at this time. The N:T ratios at both sites were

mostly less than 1, indicating lower atmospheric concentrations in recent years compared to 1994–1995.

The N:T ratios at Kinngait were generally much lower than those at Alert, indicating faster OC removal at

Kinngait. Despite the 2-year difference between the bnowQ data used to calculate the ratios for the two sites, the

atmospheric decline rates of OCs, expressed as halflives at Alert (Table 3), cannot explain the difference between

the N:T ratios of the two locations. To confirm this, air concentrations at Alert in 1998–1999 and 2000–2001 were

estimated using measured concentrations in 1994–1995 and halflives determined from OC trends at Alert between

1993 and 1999 (Table 3). The corresponding halflife-estimated ratios are presented in Fig. 8 for comparisons (grey

[1998–1999] and white [2000–2001] bars). It is believed that OC air concentrations seem to decline faster at

Kinngait because it is closer to temperate regions, where both biotic and abiotic degradation rates are usually faster.

(a) α-HCH

0

0.2

0.4

0.6

0.8

1

1.2

1.4

Composite 4(wk 43-46)


Composite 6(wk 5-8)


Composite 8 Composite 4 Composite 5 Composite 6 Composite 7 Composite 8

(wk 13-16)



Composite 6(wk 5-8)



N:T

(b) γ-HCH

0

0.2

0.4

0.6

0.8

1

1.2

(wk 43-46) (wk 47-49) (wk 5-8) (wk 9-12) (wk 13-16)

Composite 4 Composite 5 Composite 6 Composite 7 Composite 8(wk 43-46) (wk 47-49) (wk 5-8) (wk 9-12) (wk 13-16)

N:T

(c) t-chlordane

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

N:T

(d) c-chlordane

0

0.5

1

1.5

2

2.5

3

N:T

Kinngait Alert Estimated Alert 98-99 Estimated Alert 00-01

Fig. 8. bNow-to-ThenQ ratios, N:T, of (a) a-HCH, (b) g-HCH, (c) t-chlordane and (d) c-chlordane N:T ratios estimated for Alert 98–99 and Alert

00–01 were determined using 94–95 measured Alert air concentrations and 93–99 halflives determined using DF.


3.4. Spatial variations in OC seasonality

It can be clearly seen that at Alert (Fig. 6) elevated OC concentrations were generally observed before and after

the warmest time of the year. This phenomenon was noted not only for the current-used pesticides, such as lindane

(g-HCH) and endosulfan, but also for banned pesticides, such as the chlordanes. The question is whether this

occurs at other Arctic locations or is this a unique phenomenon at Alert. To answer this question, gas phase air

concentrations from Alert, Tagish, Dunai and Kinngait measured during the 1994–1995 co-sampling period are

plotted against months of the year for comparisons. As an example, Figs. 9–11 show the box-and-whisker plots of

a-HCH, g-HCH and TC at Alert, Tagish and Dunai. Since there are very few data for Kinngait during this co-

sampling period as most samples were analyzed as four-week composites, scatter plots of all data are given in Figs.

9d, 10d and 11d. For Kinngait, Tagish and Dunai, all available data during the 2 years of sampling are used in Figs.

9–11 to retain the most information although part of the sampling periods do not overlap (see timeline in Fig. 1).

For Alert, data from 8 March 1993 to 24 April 1995 (same as sampling period at Dunai) are used in the plots. Also

shown in Fig. 11 are the box-and-whisker plots of the weekly averaged temperatures at the sites. For Kinngait, the

temperatures are weekly averaged when 7-day samples were taken and four-weekly averaged when four-week

composites were analyzed.

For the three compounds shown in Figs. 9–11, it is apparent that generally elevated air concentrations were

observed in April and May at Alert, Tagish and Dunai. At Kinngait, this bSpring Maximum EventQ usually

Dunai α-HCH

1 2 3 4 5 6 7 8 9 10 11 12

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12

1 2 3 4 5 6 7 8 9 10 11 12

02468

10121416182022242628303234363840

Month

Alert α-HCH

gas

con

cent

ratio

n (p

g/m

3 )ga

s c

once

ntra

tion

(pg

/m3 )

gas

con

cent

ratio

n (p

g/m

3 )ga

s c

once

ntra

tion

(pg

/m3 )

0

50

100

150

200

250

Month

Kinngait α-HCH

0

10

20

30

40

50

60

70

80

90

100

110

120

130

140

Month

Tagish α-HCH

0102030405060708090

100110120130140150160170500600700800

Month

(a) (b)

(c) (d)

Fig. 9. a-HCH seasonality at (a) Alert, (b) Tagish, (c) Dunai and (d) Kinngait. For the box-and-whisker plots, the centre box is bounded by the

25th and 75th percentile with the horizontal line representing the median. Outliers (z1.5�interquartile range) are represented by a ..


occurred slightly earlier, beginning in March and ending in May. Other than these three compounds, other OCs also

showed elevated concentrations in spring at one or more of the four sites. For instance, spring peaks occurred for

CC and endosulfan I at Alert, Kinngait and Dunai but not at Tagish; TN at Kinngait and Dunai only; o,pV-DDE at

Alert only; o,pV- and p,pV-DDT at Alert and Dunai only; octachlorostyrene at Dunai only; and tetrachloroveratrole

at Alert and Tagish only. Most transformation products/metabolites of parent pesticides, i.e. oxychlordane,

heptachlor epoxide and dieldrin, generally do not show spring peaks. Rather, they followed the seasonal

temperature cycles and reached highest air concentrations in Arctic air during the warmest time of the year.

There are many potential factors which may be responsible for bSpring Maximum EventsQ observed. For

instance, it has been hypothesized that the spring elevation in Arctic air concentrations of the HCHs was due to

fresh applications, either as lindane (99% g-HCH) or technical HCH (55–80% a-HCH, 8–15% g-HCH; Li, 1999),

in temperate regions during spring, followed by long-range transport to the Arctic (Halsall et al., 1998; Hung et al.,

2002). However, these events are observed not only for current-use pesticides, but are also seen for banned

pesticides, e.g. chlordanes, and industrial by-products, e.g. octachlorostyrene and tetrachloroveratrole. Even for the

(a) (b)

(c) (d)

gas

conc

entr

atio

n (p

g/m

3 )

Month

Alert γ-HCH

0

5

10

15

20

25

30

35

40

45

50

55

60

Tagish γ-HCH

02468

10121416182022242628303234

110120130140

gas

conc

entr

atio

n (p

g/m

3 )

Month

Kinngait γ-HCH

0

5

10

15

20

25

30

35

40

45

50ga

s co

ncen

trat

ion

(pg/

m3 )

Month

gas

conc

entr

atio

n (p

g/m

3 )

Month

Dunai γ-HCH

02468

10121416182022242628303234363840

1 2 3 4 5 6 7 8 9 10 11 12

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12

1 2 3 4 5 6 7 8 9 10 11 12

Fig. 10. g-HCH seasonality at (a) Alert, (b) Tagish, (c) Dunai and (d) Kinngait. For the box-and-whisker plots, the centre box is bounded by the

25th and 75th percentile with the horizontal line representing the median. Outliers (z1.5�interquartile range) are represented by a ..


current-use pesticides such as lindane and endosulfan, they are probably being used at different times of the year at

different locations. While lindane was applied in Canada around May, mainly as a canola seed treatment,

endosulfan is used to treat crops such as apples, cherries and potatoes in the Great Lake basin and is mainly used

during the summer rather than in spring (Buehler et al., 2004). When freshly applied pesticides reached the Arctic,

it is unlikely for their concentrations to peak around the same time of the year.

Another possible reason is the so-called dspring pulseT. Gouin et al. (2002) have observed high spring air

concentrations of polybrominated diphenyl ethers (PBDEs) in Southern Ontario. They have attributed this dspringpulseT to chemicals deposited on the snowpack over winter and were released back to the air with snowmelt in

spring, resulting in elevated air concentrations. However, it can be seen from the box-and-whisker plots of

temperature in Fig. 11 that local ice/snowmelt (TN0 8C with T=0 8C indicated by the solid line on each plots) did

not occur until May in Kinngait and June in Alert and Dunai, which did not coincide with the spring OC

concentration peaks. The only exception is at Tagish where the temperature reaches 0 8C in March. Although local

dspring pulseT is unlikely the cause for spring elevated air concentrations at these sites, ice/snowmelt which occurs

earlier at southern latitudes might have been. Chemicals released from snowpacks in the south during the spring

(b)

(a)

(c)

(d)

Alert t-chlordane

1 2 3 4 5 6 7 8 9 10 11 120.00.2

0.40.60.8

1.01.21.41.61.8

2.02.22.4

gas

con

cent

ratio

n (p

g/m

3 )ga

s c

once

ntra

tion

(pg

/m3 )

gas

con

cent

ratio

n (p

g/m

3ga

s c

once

ntra

tion

(pg

/m3

))

Month

Month

T (

°C)

Alert Temperature

1 2 3 4 5 6 7 8 9 10 11 12-40

-35

-30

-25

-20

-15

-10

-5

0

5

10

Month

T (

°C)

Month

Tagish Temperature

1 2 3 4 5 6 7 8 9 10 11 12-30

-25

-20

-15

-10

-5

0

5

10

15

20

25

Month

T (

°C)

Dunai Temperature

1 2 3 4 5 6 7 8 9 10 11 12-4 0

-3 0

-2 0

-1 0

0

10

T (

°C)

Month

Kinngait I Temperature

1 2 3 4 5 6 7 8 9 10 11 12-3 0

-2 5

-2 0

-1 5

-1 0

-5

0

5

10

15

Tagish t-chlordane

1 2 3 4 5 6 7 8 9 10 11 120.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.86.0

8.0

Month

Dunai t-chlordane

1 2 3 4 5 6 7 8 9 10 11 120.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

Month

Kinngait t-chlordane

1 2 3 4 5 6 7 8 9 10 11 120.00.10.20.30.40.50.60.70.80.91.01.11.21.3

Fig. 11. t-Chlordane seasonality and temperature at (a) Alert, (b) Tagish, (c) Dunai and (d) Kinngait. For the box-and-whisker plots, the centre

box is bounded by the 25th and 75th percentile with the horizontal line representing the median. Outliers (z1.5�interquartile range) are

represented by a ..



snowmelt might be transported to the Arctic by advection resulting in higher Arctic air concentrations at this time

of the year. On the other hand, it is impossible to explain why only some OCs showed elevated concentrations in

spring but not others, especially the transformation products/ metabolites, and only at some sites but not others. If

chemicals trapped and re-released from snowpack in spring was the reason for the dspring maximum eventsT in the

Arctic, both parent pesticides and metabolites should show high concentrations in air during snowmelt, although

some compounds might be more subjected to snow scavenging, revolatilization and long-range transport than

others due to their physical–chemical properties.

The delevatedT concentrations observed in the Arctic in spring could be an artifact resulted from increased

atmospheric removal in late spring/summer. After polar sunrise around April, removal rates of OCs by

photodegradation and depletion reactions with OH radicals would increase, resulting in a decrease in air

concentrations during the summertime. Yet, if this was the only reason for the seemingly high air concentrations in

spring, the dspring maximum eventT should occur at all sites, not just some sites for some compounds.

Another process that might reduce atmospheric concentrations in late spring/summer is scavenging by forests

along the transport pathway to the Arctic during the growing season (Wania and McLachlan, 2001). At leaf burst in

late spring, the waxy cuticle on the new leaves would adsorb semivolatile organic compounds in air in the vicinity

of forests. This process would reduce the amount of chemicals in air available for transport to the Arctic, resulting

in a decrease in Arctic air concentrations in late spring/summer.

Why certain compounds show a spring peak in Arctic air at some locations is probably the result of a

combination of the above factors. Spring is a sensitive time of the year when various activities happen in the

Arctic as well as at southern latitudes. In the Arctic, polar sunrise occurs at this time and removal by

photodegradation and reaction with OH radicals increase over the summer. In the south, agricultural activities

start to revive hence the application of pesticides; the snow and ice starts to melt hence re-releasing of OCs

trapped in snowpacks; temperature starts to rise resulting in enhanced revolatilization from secondary sources;

and leaves start to burst in late spring/summer scavenging OCs from air thus reducing the amount of chemicals

available for transport to the Arctic. Yet, the atmospheric circulation pattern between the south and the Arctic in

spring, which controls the movement of contaminants into, through and out of the Arctic, has not been well

characterized. Barrie et al. (1992) have summarized air flow patterns in both winter and summer. In the winter,

air is driven by the strong Siberian anticyclone and flows mainly from Eurasia into the Arctic, then out over

North America, enhances transport of pollutants into the Arctic region. In the summer, the Siberian high

dissipates and the south-to-north transport from Eurasia is replaced by weak north-to-south transport. In order to

understand the dspring maximum eventT, one must understand how the atmospheric circulation pattern changed

between winter and summer, and the relationship between this spring flow pattern and the locations of potential

sources. For example, the air concentrations of lindane may be high in the Canadian Prairies during the spring

planting season but if the air flow pattern does not encourage transport to Alert, air concentrations may not have

peaked in spring at this specific site. To determine the influence of each of the abovementioned factors on the

dspring maximum eventT of different OCs at different locations, further investigation in the atmospheric flow

pattern in spring is required.

3.5. Influence of atmospheric variation patterns on Arctic air concentrations of OCs

Another factor which may affect transport of pollutants to the Arctic, which is closely related to atmospheric

circulation patterns, is climate fluctuations. Ma et al. (2003) have found strong association between air

concentrations of hexachlorobenzene (HCB) measured in the Great Lakes region and the El Nino–Southern

Oscillation (ENSO) during the 1990s. They have attributed this correlation to enhanced revolatilization from soil

contaminated due to past usage as a result of surface air temperature increase in North America associated with the

El Nino events. Their findings implied the potential influence of climate variation on the global distribution of

POPs. Climate variations not only affect the revolatilization of POPs from secondary sources but also the

atmospheric circulation patterns which are responsible for their distribution around the globe. In this section, some


evidence of the influence of two climate variation patterns, the North Atlantic Oscillation (NAO) and the Pacific

North American (PNA) pattern, on the atmospheric concentrations of OCs at Alert will be presented.

The NAO refers to a meridional shift in atmospheric mass between subpolar and subtropical latitudes. It is an

indication of the strength of westerly winds blowing across the North Atlantic Ocean between 40 8N and 60 8N.Such strength influences the surface air temperature and balance of precipitation and evaporation across the

Atlantic Ocean and the adjoining landmasses (Stenseth et al., 2002). In spring, sea level pressure anomaly patterns

shift between the Arctic and the northern Atlantic during the positive and negative phases of the NAO, influencing

surface air temperatures. During the positive NAO phase, western and central North America experiences warmer

than normal spring. Roughly reversed conditions occur during the negative NAO phases when eastern and central

North America experiences colder springs. The NAO is correlated with the Arctic Oscillation (AO) pattern which

is a surface manifestation of the strength of the polar vortex (Macdonald et al., this issue). The PNA is

characterized by atmospheric flow in which the west coast of North America is out of phase with the Eastern

Pacific and Southeast United States. During its positive phase, wavy flow occurs over the continent with increased

temperatures and decreased storminess in the Northwest and cold temperatures in the Southeast (NOAA-CIRES

Climate Diagnostic Center, http://www.cdc.noaa.gov/Teleconnections/pna.html).

The NAO and the PNA are represented by the NAO and PNA indices, respectively. These indices can be

obtained from the Climate Diagnostics Center of the National Oceanic and Atmospheric Administration (NOAA)

at bhttp://www.cdc.noaa.gov/Teleconnections/pna.htmlN.

The Spearman rank-order correlation is used to determine the association of HCB, HCHs and PCB air

concentrations observed at Alert and the 5 Great Lakes sites with the NAO index. The Spearman rank-order

correlation is a distribution-free, nonparametric measure of association based on the rank of the data values which

is exact for small sample sizes and weakens the effect of outliers (Sachs, 1984). Relationships are assessed using

seasonal averages of all parameters in winter (December–February) and spring (March–May). During these two

seasons, the influence of these climatic patterns is most prominent.

Table 4 shows the Spearman correlation coefficients between air concentrations and NAO. HCB, a- and g-HCH

and some lower PCB congeners (e.g. PCB 8, 31, 44 and 52) correlate significantly ( pb0.05) with NAO at various

Great Lake sites in spring. All Spearman correlation coefficients, rS, were found to be greater than zero, indicating

that air concentrations of these compounds in the Great Lakes region tend to increase towards the positive phase of

Table 4

Spearman rank-order correlation coefficients, rS, between NAO indices and OC air concentrations

Station OC rS p Na

Spring (March–May)

Alert PCB 31 0.71 0.05 8

PCB 44 0.74 0.04 8

PCB 138 0.76 0.03 8

Point Petre PCB 8 0.72 0.03 9

PCB 31 0.82 0.02 7

PCB 52 0.73 0.02 9

Burnt Island PCB 44 0.83 0.04 6

Eagle Harbor HCB 0.86 0.01 7

a-HCH 0.76 0.03 8

Sleeping Bear Dunes a-HCH 0.82 0.02 7

Sturgeon Point g-HCH 0.86 0.01 7

Winter (December–February)

Burnt Island PCB 105 0.83 0.04 6

Eagle Harbor PCB 101 0.71 0.05 8

Sturgeon Point g-HCH 0.75 0.05 7

a N=number of seasonally averaged data used in the correlation for each site.

http://www.cdc.noaa.gov/Teleconnections/pna.html





NAO, when higher-than-usual temperatures were observed in western and central North America. This could be

the result of enhanced re-volatilization from sources in these regions where such compounds have been or are still

being used.

For Alert, the lighter compounds of HCB and HCHs do not show any correlation with NAO. This is probably

because air concentrations of these compounds in the arctic are less dependent on local temperatures. Correlations

between PCB air concentrations and NAO may simply reflect their relationships at lower latitudes.

For all sites, few compounds show significant correlations in winter. It is believed that the overall low

temperature in winter makes revolatilization of POPs difficult.

Fig. 12 shows plots of the mean gas phase concentrations of PCBs 31, 101 and 153 and g-HCH measured at

Alert in spring between 1993 and 1999 against the PNA index. All four compounds showed statistically significant

correlation with the PNA index ( pb0.1).

Despite the relatively small datasets (b10 years) of POP air concentrations used, good correlations observed

between concentrations measured at Alert and indices of climate variation patterns, such as the NAO and PNA,

have highlighted the potential influence of climate variation on the distribution of POPs, both spatially and

temporally. Temperature anomalies associated with climate variations not only may enhance revolatilization of

chemicals from previously accumulated sinks, they may also affect the apportionment of these compounds

between the gas and particle phases in air which, in turn, influence their transport to remote locations like the

Arctic. Besides, climate variation patterns, such as the NAO and AO, also carry with them altered moisture

fluxes (Macdonald et al., this issue). Such changes will affect upstream scavenging of POPs which will influence

0

0.1

0.2

0.3

0.4

0.5

0.6

-1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8

PNA Index

0.8

2.8

4.8

6.8

8.8

10.8

12.8

14.8

16.8

18.8

20.8

-1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8

PNA Index

PCB 153γ-HCHrs= 0.86p = 0.01

rs= 0.82p = 0.02

PCB 31rs= 0.79p = 0.04

0

0.1

0.2

0.3

0.4

0.5

0.6

-1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8

PNA Index

0.3

0.35

0.4

0.45

0.5

0.55

0.6

0.65

0.7

0.75

0.8

-1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8

PNA Index

PCB 101rs = 0.68p = 0.09

Mea

n ga

s ph

ase

conc

. in

sprin

g (p

g/m

3 )M

ean

gas

phas

e co

nc. i

n sp

ring

(pg/

m3 )

Mea

n ga

s ph

ase

conc

. in

sprin

g (p

g/m

3 )M

ean

gas

phas

e co

nc. i

n sp

ring

(pg/

m3 )

Fig. 12. Mean gas phase concentrations of OCs in spring at Alert and PNA index.


the transport potential of these compounds. The air flow, precipitation and surface air temperatures associated

with each climatic pattern must be investigated carefully in order to explain how the transport of POPs to the

Arctic would change in response to such pattern. Since climate variations occur on a decadal time scale in the

Arctic (Nakamura and Oort, 1988), many more years of POP air concentration data are required to reveal the

underlying mechanisms.

4. Conclusions and recommendations

Long-term trends of PCB and OC air concentra-

tions measured at Alert were developed using a digital

filtration technique with data from 1993 to 1999. It

was found that most banned pesticides showed

declining trends while current-use pesticide, endo-

sulfan I, exhibits a slow decline with t1/2 of 38 years.

Trends developed with 5 years of data are similar to

those developed with 7 years of data. Although this

indicates that less than 10 years of data may be

adequate for deriving trends for remote background

locations, atmospheric measurement must be contin-

ued. The notable acceleration in decline of air

concentrations for some OCs and PCBs in 1999

may affect the long-term trend and the 5 to 10 years of

atmospheric data currently available barely provide

enough information to develop statistically significant

trends for forecasting purposes.

Both bcurrent-useQ and banned pesticides showed

spring maxima events. However, only some com-

pounds show such peaks at some sites but not others.

Relationships between potential sources and atmos-

pheric circulation pattern in spring must be inves-

tigated. The correlation between air concentrations of

POPs measured at Alert and climate variation patterns

indicates that planetary atmospheric patterns must be

taken into account in the global prediction and

modelling of POPs in the future. It is important to

bear in mind that temperature is not the only

parameter of climate variation that would affect the

transport processes of such compounds. For instance,

changes in other climate parameters, e.g. quantity,

quality and spatial variation of rain and snow

associated with various climate patterns would also

affect scavenging and deposition of organic pollu-

tants. Therefore, it is essential to continue long-term

atmospheric monitoring of POPs in the Arctic as well

as in temperate North America to further establish

relationships between POP concentrations and the

fluctuations of different climate parameters. As it

stands, the POP air concentration dataset is too small

to fully characterize the effect of climate variations on

global transport of POPs.

Acknowledgement

Financial support for this work was provided by

the Arctic Environmental Strategy Northern Contam-

inants Program, Department of Indian Affairs and

Northern Development (DIAND). We would like to

thank D.C.G. Muir for his effort in developing this

project and continual support of our work.

References

Aigner EJ, Leone AD, Falconer RL. Concentrations and enantio-

meric ratios of organochlorine pesticides in soils from the US

Cornbelt. Environ Sci Technol 1998;32:1162–8.

Allen JO, Dookeran NM, Smith KA, Sarofim AF, Taghizadeh K,

Lafleur AL. Measurement of polycyclic aromatic hydrocarbons

associated with size-segregated atmospheric aerosols in Massa-

chusetts. Environ Sci Technol 1996;30:1023–31.

AMAP. AMAP data report: arctic pollution issues. Arctic Monitor-

ing and Assessment Programme (AMAP), Oslo, Norway. xii +

859 pp. 1999.

Anderson PN, Hites RA. OH radical reactions: the major removal

pathway for polychlorinated biphenyls from the atmosphere.

Environ Sci Technol 1996;30:1756–63.

Bailey R, Barrie LA, Halsall CJ, Fellin P, Muir DCG. Atmospheric

organochlorine pesticides in the western Canadian arctic:

evidence of transpacific transport. J Geophys Res 2000;

105(D9):11805–11.

Barrie LA, Bottenheim JW, Schnell RC, Crutzen PJ, Rasmussen

RA. Ozone destruction and photochemical reactions at polar

sunrise in the lower arctic atmosphere. Nature 1988;334:

138–41.

Barrie LA, Gregor D, Hargrave B, Lake R, Muir D, Shearer R, et al.

Arctic contaminants: sources, occurrence and pathways. Sci

Total Environ 1992;122:1–74.

Beeman RW, Matsumura F. Metabolism of cis- and trans-chlordane

by a soil microorganism. J Agric Food Chem 1981;29:84–9.

Bidleman TF, Jantunen LMM, Wiberg K, Harner T, Brice K, Su K,

et al. Soil as a source of atmospheric heptachlor epoxide.



Bidleman TF, Jantunen LMM, Helm PA, Brorstrfm-Lunden E,

Juntto S. Chlordane enantiomers and temporal trends of

chlordane isomers in Arctic air. Environ Sci Technol 2002;

36:539–44.

Bidleman TF, Wong F, Backe C, Sfdergren A, Brorstrfm-Lunden E,

Helm PA, et al. Chiral signatures of chlordanes indicate changing

source to the atmosphere over past 30 years. Atmos Environ

2004;38:5963–70.

Buehler SS, Basu I, Hites RA. Causes of variability in pesticide and

PCB concentrations in air near the great lakes. Environ Sci

Technol 2004;38:414–22.

Buser H-R, Mqller MD. Identification of the (+) and (�)

enantiomers of chiral chlordane compounds using chiral high

performance liquid chromatography–chiroptic detection and

chiral high resolution gas chromatography–mass spectrometry.

Anal Chem 1994;66:2155–62.

Coleman PJ, Lee RGM, Alcock RE, Jones KC. Observations on

PAH, PCB, and PCDD/F trends in UK urban air, 1991–1995.


Cortes DR, Hites RA. Detection of statistically significant trends in

atmospheric concentrations of semivolatile compounds. Environ

Sci Technol 2000;34:2826–9.

DIAND (Department of Indian Affairs and Northern Development).

Canadian Arctic Contaminants Assessment Report (CACAR),

Department of Indian Affairs and Northern Development,

Ottawa, Ontario, Canada. 1997, 459 pp.

DIAND (Department of Indian Affairs and Northern Development).

Canadian Arctic Contaminants Assessment Report II (CACAR

II). Sources, occurrence, trends and pathways in the physical

environment. Department of Indian Affairs and Northern

Development, Ottawa, Ontario, Canada. 2003, 332 pp.

Gooch JW, Matsumura F, Zabik MJ. Chlordane residues in great

lakes lake trout: acute toxicity and interaction at the Gaba

receptor of rat and lake trout brain. Chemosphere 1990;21:

393–406.

Gouin T, ThomasGO, Cousins I, Barber J,MackayD, Jones KC. Air-

surface exchange of polybrominated diphenyl ethers and poly-

chlorinated biphenyls. Environ Sci Technol 2002;36:1426–34.

Halsall CJ, Coleman PJ, Davis BJ, Burnett V, Waterhouse KS,

Harding-Jones P, et al. PAHs in UK urban air. Environ Sci

Technol 1994;28:2380–6.

Halsall CJ, Barrie LA, Fellin P, Muir DCG, Billeck BN, Lockhart L,

et al. Spatial and temporal variation of polycyclic aromatic

hydrocarbons in the Arctic atmosphere. Environ Sci Technol

1997;31:3593–9.

Halsall CJ, Bailey R, Stern GA, Barrie LA, Fellin P, Muir DCG, et

al. Atmospheric organochlorine pesticides in the western

Canadian Arctic: evidence of transpacific transport. Environ

Pollut 1998;102:51–62.

Halsall CJ, Sweetman AJ, Barrie LA, Jones KC. Modelling the

behaviour of PAHs during atmospheric transport from the UK to

the Arctic. Atmos Environ 2001;35:255–67.

Hillery BR, Basu I, Sweet CW, Hites RA. Temporal and spatial

trends in a long-term study of gas-phase PCB concentrations

near the Great Lakes. Environ Sci Technol 1997;31:1811–6.

Hung H, Halsall CJ, Blanchard P, Li HH, Fellin P, Stern G, et al.

Are PCBs in the Canadian Arctic atmosphere declining?

Evidence from 5 years of monitoring. Environ Sci Technol

2001;35:1303–11.

Hung H, Halsall CJ, Blanchard P, Li HH, Fellin P, Stern G, et al.

Temporal trends of organochlorine pesticides in the Canadian

arctic atmosphere. Environ Sci Technol 2002;36:862–8.

Jantunen LMM, Bidleman TF, Harner T, Parkhurst WJ. Toxaphene

and other organochlorine pesticides in Alabama air. Environ Sci

Technol 2000;34:5097–105.

Karlsson H, Muir DCG, Teixiera CF, Burniston DA, Strachan WMJ,

Hecky RE, et al. Persistent chlorinated pesticides in air, water,

and precipitation from the Lake Malawi area, Southern Africa.


Li YF. Global technical hexachlorocyclohexane usage and its

contamination consequences in the environment: from 1948 to

1997. Sci Total Environ 1999;232:121–58.

Li YF, Macdonald RW. Sources and pathways of selected organo-

chlorine pesticides to the Arctic and the effect of pathway

divergency on HCH trends in biota: a review. Sci Total Environ

2004;342:87–106 [this issue].

Li YF, Bidleman TF, Barrie LA, McConnell LL. Global

hexachlorocyclohexane use trends and their impact on the

arctic atmospheric environment. Geophys Res Lett 1998;25:

39–41.

Ma J, Venkatesh S, Jantunen LM. Evidence of the impact of

ENSO events on temporal trends of hexachlorobenzene air

concentrations over the Great lakes. Sci Total Environ 2003;

313:177–84.

Macdonald RW, Harner T, Fyfe J. Recent climate change in the

Arctic and its impact on contaminant pathways and interpreta-

tion of temporal trend data. Sci Total Environ 2004;342:5–86

[this issue].

Mattina MJI, Lannucci-Berger W, Dykas L, Pardus J. Impact of

long-term weathering, mobility and land use on chlordane

residues in soils. Environ Sci Technol 1999;33:2425–31.

Meijer SN, Halsall CJ, Harner T, Peters AJ, Ockenden WA, Jones

KC. Organochlorine pesticide residues in archived UK soil.


Nakamura N, Oort AH. Atmospheric heat budgets of the polar

regions. J Geophys Res 1988;93(D8):9510–24.

Nakazawa T, Ishizawa M, Higuchi K, Trivett NBA. Two curve

fitting methods applied to CO2 flask data. Environmetrics

1997;8:197–218.

Oehme M, Haugen J-E, Schlabach M. Seasonal changes and

relations between levels of organochlorines in Arctic ambient

air: first results of an all-year-round monitoring program at Ny-

2lesund, Svalbard, Norway. Environ Sci Technol 1996;30:

2294–304.

Pacyna JM, Oehme M. Long-range transport of some organic

compounds in the Norwegian arctic. Atmos Environ 1988;

22:243–57.

Pankow JF. Review and comparative analysis of the theories on

partitioning between the gas and aerosol particulate phases in

the atmosphere. Atmospheric Environment Atmospheric Envi-

ronment 1987;21:2275–83.

PANUPS (Pesticide Action Network North America). http://

www.igc.org/panna/resources/_ pestis/PESTIS.1997.43.html

(May 23, 1997).

http://www.igc.org/panna/resources/_pestis/PESTIS.1997.43.html

http://www.igc.org/panna/resources/_pestis/PESTIS.1997.43.html


Patton GW, Hinckley DA, Walla MD, Bidleman TF, Hargrave BT.

Airborne organochlorines in the Canadian high arctic. Tellus

1989;41B:243–55.

Patton GW, Walla MD, Bidleman TF, Barrie LA. Polycyclic

aromatic and organochlorine compounds in the atmosphere of

northern Ellesmere Island, Canada. J Geophys Res 1991;

96:10867–77.

Ramesh A, Tanabe S, Tatsukawa R, Subramanian AN, Palanichamy

D, Mohan D, et al. Seasonal variations of organochlorine

insecticide residues in air from Porto Novo, South India.

Environ Pollut 1989;62:213–22.

Sachs L. Applied statistics. A handbook of techniques. New York

(U.S.A.)7 Springer-Verlag New York Inc.; 1984. p. 395.

Simcik M, Basu I, Sweet CW, Hites RA. Temperature dependence

and temporal trends of polychlorinated biphenyl congeners in

the Great Lakes atmosphere. Environ Sci Technol 1999;33:

1991–5.

Sirois A, Barrie LA. Arctic lower tropospheric aerosol trends and

composition at Alert, Canada: 1980–1995. J Geophys Res

1999;104(D9):11599–618.

Stenseth NC, Mysterud A, Ottersen G, Hurrell JW, Chan K-S, Lima

M. Ecological effects of climate fluctuations. Science 2002;297:

1292–96.

Stern GA, Halsa CJ, Barrie LA, Muir DCG, Fellin P, Rosenberg B,

et al. Polychlorinated biphenyls in Arctic air: 1. Temporal and

spatial trends: 1992–1994. Environ Sci Technol 1997;31:

3619–28.

Stern GA, Braekevelt E, Helm PA, Bidleman TF, Outridge PM,

Lockhart WL, et al. Modern and historical fluxes of halogenated

organic contaminants to a lake in the Canadian arctic, as

determined from annually laminated sediment cores. Sci Total

Environ 2005;342:223–43 [this issue].

Ulrich EM, Hites RA. Enantiomer ratios of chlordane-related

compounds in air near the Great Lakes. Environ Sci Technol

1998;32:1870–4.

Wania F, Mackay D. Tracking the distribution of persistent organic

pollutants. Environ Sci Technol 1996;30:A390–6.

Wania F, McLachlan M. Estimating the influence of forests on the

overall fate of semivolatile organic compounds using a multi-

media fate model. Environ Sci Technol 2001;35:582–90.

Wiberg K, Letcher RJ, Sandau CD, Norstrom RJ, Tysklind M,

Bidleman TF. The enantioselective bioaccumulation of chiral

chlordane and a-HCH contaminants in the polar bear food

chain. Environ Sci Technol 2000;34:2668–74.

temporal and spatial variabilities of atmospheric polychlorinated biphenyls (pcbs), organochlorine...

Documents