- Email: [email protected]
Fs in a mountain forest by means of active and passive sampling
ARTICLE IN PRESS
Environmental Research 105 (2007) 300–306 www.elsevier.com/locate/envres
Monitoring of PCDD/Fs in a mountain forest by means of active and passive sampling W. Levya,, B. Henkelmanna, G. Pfistera, M. Kirchnera, G. Jakobia, A. Niklausa, J. Kotalika, S. Bernho¨fta, N. Fischera, K.-W. Schramma,b a
GSF-Forschungszentrum fu¨r Umwelt und Gesundheit, Institut fu¨r O¨kologische Chemie, Ingolsta¨dter Landstr. 1, D-85764 Neuherberg, Germany b TUM, Wissenschaftszentrum Weihenstephan fu¨r Erna¨hrung und Landnutzung, Department fu¨r Biowissenschaften, Weihenstephaner Steig 23, 85350 Freising, Germany Received 30 October 2006; received in revised form 9 March 2007; accepted 3 May 2007 Available online 13 August 2007
Abstract Polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/F) are sampled and investigated in a forested area in Middle-Europe. The campaigns, consisting in active and passive samplings, were conducted in the Bavarian and Bohemian Forest at four sites chosen for their similar soil and forest stand characteristics. Passive sampling was conducted using both semi-permeable membrane devices (SPMDs) and needles of well-exposed dominant spruce trees. Active air sampling was also performed at one site with a low volume air sampler. Correlations were performed to identify relationships and trends of PCDD/F. Lower chlorinated PCDD/F are accumulated in SPMDs, needles collected all compounds among homologues and their PCDD/F pattern is close to that of active sampling. Results of the analysed compounds obtained with the different sampling methods served as a basis for the establishment of advantages and disadvantages of the sampling tools applied and their possible optimisation. r 2007 Elsevier Inc. All rights reserved. Keywords: PCDD/F; SPMDs; Active sampling; Spruce needles; Passive sampling
1. Introduction The role of mountain regions in the behaviour and global distribution of persistent organic pollutants (POPs) is crucial. In particular, forests play an important role as regulators of the deposition of these pollutants due to their enhanced surface roughness and the high capacity to retain them in the lipophilic foliages acting as a sink. Semivolatile organic compounds (SOCs) are transferred from the atmosphere to the soil compartment by dry and wet deposition where the vegetation is the intermediate compartment between them (Smith and Jones, 2000). Within the scope of studying the deposition and accumulation of SOCs, different sampling methods were chosen for monitoring persistent compounds to generate a complete overview of the temporal and spatial distribution in the Corresponding author. Fax: +49 89 3187 3371.
E-mail address: [email protected] (W. Levy). 0013-9351/$ - see front matter r 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.envres.2007.05.018
forest system. The accumulation of lipophilic compounds in needles allows them being used as a passive sampler where this accumulation is strongly related to the species (Bo¨hme et al., 1999) and generally increases with the needle age until senescence starts (Hellstro¨m et al., 2004). Coniferous trees have good collection efficiency for both gases and particles and a high interception capacity of rain, fog, and snow precipitation and these factors make this species suitable as a passive sampler (Umlauf and McLachlan, 1994). For these reasons, needles of spruce trees (Picea abies (L.) Karst) are adequate to use as natural passive samplers. Whereas spruce needles consider not only the gaseous uptake but also the uptake of particulate matter, semi-permeable membrane devices (SPMDs) are based on diffusion processes of the compounds mainly related to the gaseous phase (Ockenden et al., 1998). Additionally, SPMDs as integrative passive samplers are of importance because they mimic the accumulation of organic hydrophilic compounds in biological systems in a
ARTICLE IN PRESS W. Levy et al. / Environmental Research 105 (2007) 300–306
301
simplified way. Active sampling with a low volume sampler allows us to determine the concentration of organic compounds in the air. The analytical determinations were focused on dibenzop-dioxins and dibenzofurans (PCDD/F) due to their persistence, bioaccumulation, and toxicity. Correlations among compounds were calculated using results obtained from the active and passive sampler measurements. Comparing the results obtained under the different sampling methods, similarities and differences between them could be investigated. Therewith an additional aim of this study was to determine strength and weakness of each sampler method applied in the monitoring campaign of PCDD/F in mountainous areas.
triolein (1,2,3-tris[cis-9-octadecenoylglycerol]). SPMDs were transported in hermetic clean glass material to and from the place of deployment to prevent possible contaminations caused by transportation. Quadruplicates of these membranes were placed in parallel on square frames. These frames were put into deployment devices of untreated wood, which were exposed at 3.0 m above the ground in small forest clearings. These deployment structures allow a baffle flow through them. The devices protect the SPMDs from direct sunlight and meteorological conditions as precipitation, hindering the wet deposition on the SPMD. After exposure, SPMDs were stored without cleaning their surface at 20 1C until analysis. SPMDs blanks were also kept under identical storage conditions. Needles, SPMDs and cartridges from the active sampler were sampled simultaneously. Thus, the last 3-month exposure coincides for the active and passive samplings in October 2003. The sampling scheme is shown in a chronological timetable (Fig. 1).
2. Material and methods
The procedures for the sample preparation, clean up and quantification of PCDD/F in needles are described in Niu et al. (2003). In brief, SPMDs were cut to slices, spiked with 13C-labeled PCDD/F standard mixtures (Cambridge Isotope Laboratories, USA) and were extracted for 24 h with 100 ml cyclohexane each by use of a rotating shaking machine. After extraction, water was removed with anhydrous sodium sulphate. Afterwards, a clean-up procedure with a sandwich (silica, H2SO4-treated silica), an alumina and a florisil column was conducted. The extract was then concentrated to 10 ml under a stream of nitrogen before quantification. The compounds were determined by highresolution gas chromatography (HRGC) on a Rtx-2330 column for PCDD/F (Restek, Germany). The GC is coupled with a high-resolution mass spectrometer MAT95s (Thermo Electron GmbH, Germany) operated in single ion monitoring mode. Results for PCDD/F were given in air concentration (fg m3) for the active measurements, in mass per dry weight (ng kg1) for the spruce needles and total amount of each compound per tube (ng tube1) for SPMDs. Quadruplicates of SPMDs blanks were analysed for the first sampling period (12 year) and duplicates for the extended period (112 years). Duplicates of the clean-up blanks were also conducted as an internal quality control. The standard recoveries for PCDD/Fs were between 80% and 100% for SPMDs and needles.
2.1. Sampling The sampling was conducted in Eastern-Bavaria at four sites (Mitterfels 1030 m a.s.l., Ruckowitzschachten 1130 m a.s.l., Haidel 1160 m a.s.l., and Boubin 1300 m a.s.l.) chosen for their comparably cold temperature regime and their similar soil and forest stand characteristics. Since the four sites are situated at different distances from the central mountainous barrier of the Bavarian Forest and Sumava (to the southeast: Mitterfels approximately 30 km, Ruckowitzschachten 1 km, and Haidel 8 km; to the north-east, Boubin 20 km), both the precipitation regime and the distance to possible sources in Bavaria and Czech Republic are different. Stem density varies between the four sites increasing from Mitterfels with 40, Haidel with 46, Boubin with 54, and Ruckowitzschachten with 60 m2 ha1 (measurement error73 m2 ha1). Further information about the sampling sites and active and passive sampling methods is available in Kirchner et al. (2006). The low volume active sampler (Digitel blower, DPA96) was deployed at Haidel and comprises the particulate as well as the gas phase in the air sampling. Four campaigns corresponding to the periods August–October 2003, October 2003–January 2004, January 2004–March 2004, and September–October 2004 were conducted in order to establish possible seasonal trends. 1 1 1 2, 12, and 22 years old spruce needles (Picea abies (L.) Karst) were sampled in October 2003 and October 2004, respectively. Needles were collected from the seventh branch whorl of dominant trees of closed spruce stands. Foliar biomass of the sampling sites was also determined in order to detect possible influences of this factor (Jaward et al., 2005). The four spruce forests growing on typical soils of granite and gneiss bedrock and belonging to the same altitudinal zone have been selected to reduce differences attributable to stand characteristics. SPMDs were deployed for 12 and 112 years at the sampling sites and collected in October 2003 and 2004, respectively. SPMD tubes (23 cm 2.5 cm, membrane thickness 67.4 mm) were filled with 0.7 ml
2.2. Analysis
3. Results and discussion 3.1. PCDD/F: needles and active sampler The homologue distribution for dioxins and furans in spruce needles is shown in Fig. 2. The distribution at the four sites showed a PCDD percentage increase towards the higher chlorinated congeners in needles. The trend for PCDF was the opposite one; lower percentage was found for higher chlorinated congeners. This pattern was observed in all needles where less percentage differences
Fig. 1. Chronological table of SPMDs deployment, spruce needles collecting and cartridges sampling.
ARTICLE IN PRESS 302
W. Levy et al. / Environmental Research 105 (2007) 300–306
Fig. 2. PCDD/F homologues distribution (%) in 12 year old spruce needles and SPMDs exposed for M: Mitterfels, and B: Boubin. Active sampling, campaign A at Haidel. Sampling October 2003.
1 2
year at H: Haidel, R: Ruckowitzschachten,
Fig. 3. PCDD/F homologues distribution (%) in 12, 112, and 212 years old spruce needles and active sampling, campaign A at Haidel. Sampling October 2003.
P between OCDD and TetraCDF were found in aged needles (Fig. 3). Therefore, the homologue distribution of PCDD/F is characteristic in all the collected needles with P an increase of TetraCDF related to OCDD in aged needles. Comparing the four sampling sites, a higher P percentage of HeptaCDD and OCDD was detected at Haidel. Atmospheric bulk deposition studied by Horstmann et al. (1997) in a spruce forest near Bayreuth (NorthEastern Bavaria) exhibited the same homologue pattern where the maximal concentration values were also determined for OCDD. Active sampling in Southern Bavaria close to an urban area (Augsburg) performed by Hippelein et al. (1996) reported on average a comparable homologue profile. PCDD/F homologue profiles determined in spruce needles by Rappolder et al. (2007) at locations (Eastern Bavaria and Sumava) close to the monitored one in the present study showed a similar PCDD/F trend where the main percentage P of homologue compounds is shifted from OCDD to TetraCDF. PCDD air concentrations increased from lower to higher chlorinated homologues nonetheless PCDF air concentrations tend to decrease at
higher chlorine substitutions (Fig. 2). Concentrations of PCDD homologues measured in spruce needles and active sampling trials taken from the same location showed a strong correlation (Fig. 4A). A similar correlation was also achieved for older spruce needles and the active sampling (r2 ¼ 0.96 and 0.98 for 112 and 212 years old needles, respectively) pointing out the similar pattern of PCDD found in both sampling methods. PCDD congeners (Fig. 5A) mark the same pattern found in PCDD homologues (Fig. 4A). The congeners in this graphic were chosen considering the compounds that were always detectable or presented low numbers of non-detectable values in the analysed samples. Needles with older ages tend to accumulate compounds (Piccardo et al., 2005; Herceg and Krauthacker, 2006) showing seasonal differences during the year (Kylin and Sjo¨din, 2003). In our study, the bioaccumulation of PCDD/F is observable in older needles in both samplings of October 2003 and 2004. The needles were collected in the earlier autumn season after the summer period where compounds partially bound to particles tend to show a higher accumulation than in the
ARTICLE IN PRESS W. Levy et al. / Environmental Research 105 (2007) 300–306
303
for PCDF in needles is changing with time, whereas this change was not observed for PCDD. This difference obtained for older needles for PCDF could be attributed to the lower chlorinated compounds, TetraCDF and PentaCDF can be accumulated more easily into the leaf compartment than higher chlorinated PCDF homologues due to higher air concentrations and higher availability in the gaseous phase. The factor that during the sampling period the absorption of higher chlorinated compounds is favoured cannot counteract these two above mentioned effects. Another influencing factor is the compound stability. Recent studies indicated a possible increase of lower chlorinated congeners due to photolytic dechlorination (Niu et al., 2003) where PCDF tend to be more reactive than the PCDD. Thus, the higher amounts of PentaCDF and TetraCDF may also be related to an increase of photochemical degradation products accumulated in the needle at longer sunlight exposure (Fig. 5B). Another variable to take into account is the mobilisation of the compounds into the leaf compartments. The work performed by Kaupp et al. (2000) suggested that from the PCDD/F accumulated in the cuticle and classified as particle bounded fraction, PCDD are more immobile than PCDF in maize leaves. If this could also be applied to the needle compartment this may be another factor that implies the major stability in the pattern obtained for PCDD in comparison to PCDF. 3.2. PCDD/F: SPMDS and needles as abiotic and biotic passive samplers
Fig. 4. (A) PCDD homologue concentrations of spruce needles and active sampler (campaign A, 3 months exposure) at Haidel. Diamonds, squares and circles represent 12, 112, and 212 needle ages, respectively. Sampling October 2003. Correlation coefficient between active sampling and 12 year old needles. (B) PCDF homologue concentrations of spruce needles and active sampler (campaign A, 3 months exposure) at Haidel. Diamonds, squares and circles represent 12, 112, and 212 needle ages, respectively. Sampling October 2003. Correlation coefficient between active sampling and 12 year old needles.
rest of the annual seasons (Hellstro¨m et al., 2004). The factor that higher chlorinated PCDD are at higher air concentrations than lower chlorinated PCDD determines the pattern obtained in needles but the season of the sampling also influences the obtained results for PCDD. For PCDF at older needle ages, the correlation between air and needle concentrations decreased clearly (Fig. 4B) due to the different shift of PCDF homologues towards higher concentrations so we cannot infer a direct relationship between the active sampling and needle concentrations for aged needles. Despite this, a good agreement between both methods for all homologues was found for 12 year old needles. Regarding PCDF congeners, only a correlation between the younger needles and the active sampler could be found. This indicates that the concentration distribution
The PCDD/F homologue concentrations obtained from SPMD samples were different from those related to spruce needles. The PCDD/F homologue distribution (in % of the total amount of PCDD/F) is shown in Fig. 2. OCDF was not detectable in the majority of the samples. Taking into account that the SPMD is able to sample mainly the gaseous phase and not the contaminants attached to particulate aerosol the different homologue distribution obtained is not surprising. Homologues with comparable air concentrations but different partitioning behaviour had different distributions in these two passive samplers. This is the case for HeptaCDD and TetraCDF homologues where their air concentrations are similar (and the highest after OCDD air concentrations) but their distribution in both passive samplers is clearly different. HeptaCDD tends to be mainly bound to particles during all seasons meanwhile TCDF is mainly in the gaseous phase (Hippelein et al., 1996). This causes that HeptaCDD homologues able to achieve similar percentage amounts as the TetraCDF homologues in needles are significantly different in the SPMD pattern. Experimental studies in a semi-rural field station showed that 2,3,7,8-TetraCDD/Fs present a higher proportion in the gaseous phase in comparison to the Penta to OCDD/F (Harner et al., 2000). Therefore, higher concentrations of the tetra congeners in the gas phase reflect the pronounced uptake of these compounds into
ARTICLE IN PRESS W. Levy et al. / Environmental Research 105 (2007) 300–306
304
40 Concentration of congeners in active sampler ( fg m-3 air)
OCDD
30 R2 = 0,99 20
10
1,2,3,4,6,7,8HpCDD HxCDD 2 HxCDD 1 HxCDD 3
0 1,2,3,7,8-PeCDD 0.0
2.5
5.0
7.5
Concentration of congeners in needles (ng
kg-1
dry wt)
Concentration of congeners in active sampler (fg m-3 air)
9 1,2,3,4,6,7,8HpCDF
R2 = 0.92
6
HxCDF 3 3 HxCDF 2
PeCDF 1
HxCDF 1
2,3,7,8-TCDF PeCDF 2
0 0.00
0.25
0.50
0.75
Concentration of congeners in needles (ng kg-1 dry wt) Fig. 5. (A) PCDD congener concentrations in spruce needles and active sampler 3 months exposure (campaign A) at Haidel. Diamonds, squares, and circles represent 12, 112, and 212 needle ages, respectively. HxDD 1: 1,2,3,6,7,8-HxCDD, HxDD 2: 1,2,3,7,8,9-HxCDD, HxDD 3: 1,2,3,4,7,8-HxCDD. Sampling October 2003. Correlation coefficient between active sampling and 12 year old needles. (B) PCDF congener concentrations in spruce needles and active sampler 3 months exposure (sampling A) at Haidel. Diamonds, squares and circles represent 12, 112, and 212 needle ages, respectively. PeCDF 1: 1,2,3,7,8/1,2,3,4,8-PeCDF, PeCDF 2: 2,3,4,7,8-PeCDF, HxCDF1: 1,2,3,4,7,8/1,2,3,4,7,9-HxCDF, HxCDF 2: 2,3,4,6,7,8-HxCDF, HxCDF 3: 1,2,3,6,7,8HxCDF. Sampling October 2003. Correlation coefficient between active sampling and 12 year old needles.
SPMDs in comparison to the spruce needles. The tetra homologues peaked in the distribution of all sampling sites and it is also observed an increase for the Penta and HexaCDD/F in comparison to the HeptaCDD/Fs and OCDD. Fig. 5 points out these observations. In the study of Harner et al. (2000), the division of PCDD/Fs in two groups was suggested according to their chlorine substitution pattern where the group 2 (any Tetra to HexaCDD/F with three or four chlorines in the 2,3,7,8 substitution) is characterised for uniform charge distribution in comparison to the group 1 (the rest of PCDD/F). The lower polarity exhibited in group 2 means higher octanol solubility, thus a higher KOA (partition coefficient octanol-air) for this PCDD/F group. This implies a better uptake into the SPMD device for the group 2 in
comparison to HeptaCDD/Fs and OCDD/F that belong to the group 1. On the other hand, SPMDs exposed for 112 years do not show linear increase of PCDD/F concentrations compared to the 12 year exposed devices but a relative P increasing of Tetra P to HexaCDF compounds (Fig. 6). TetraCDD and PentaCDF increase signifiP cantly, even the HexaCDF, close to the limit of detection after 12 year exposure is well determined at longer exposure periods. This is concordant with the study of Schro¨der et al. (1997) where it was determined that of the whole PCDD/Fs homologues dry gaseous deposition occurs for TetraCDD and Tetra to HexaCDFs. The importance of this process in the uptake of compounds in SPMDs is clear when we compare the homologue patterns of SPMDs and needles in Fig. 2. It has to be considered that longer periods
ARTICLE IN PRESS W. Levy et al. / Environmental Research 105 (2007) 300–306
Fig. 6. PCDD/F homologue concentrations in SPMDs
1 2
305
and 112 years exposure at three sites. Samplings October 2003 and 2004.
allow the mass transfer from the particles attached to the membrane and the membrane itself into the SPMD. The homologue distribution of SPMDs presented a similar pattern after 112 years exposure at all sampling sites, which was not so clearly observed for SPMDs exposed 12 year (Fig. 6). Extended exposure periods can cause non-linear uptake so such prolonged periods are not advised for environmental kinetic sampling with SPMDs. PIn conclusion, theP lower chlorinated compounds, TetraCDD/Fs and PentaCDD/Fs, presented a better uptake in SPMDs devices whereas higher chlorinated homologues, OCDF in particular, are hardly detectable or not detected in these devices after 6-month exposure. Needles in the same exposure period collected compounds from all the studied homologue groups following similar distribution at all sampling sites. In summary, SPMDs appear to be a very feasible tool for lower chlorinated PCDD/Fs, in particular TetraCDD/Fs due to the ability to uptake compounds from the gaseous phase. On the other hand, needles are less ‘selective’ than SPMDs but being able to uptake compounds from both, the particle bound and gaseous phase and can embrace all the homologue families in spite of the less pronounced uptake of lower chlorinated PCDD/Fs. 4. Funding The funding sources that made possible this work were facilitated from the Bavarian State Ministry of the Environment, Health and Consumer Protection under the project number 76a-8731.2—2000/1. Acknowledgments The authors wish to thank the Bavarian State Ministry of the Environment, Health and Consumer Protection (Project no. 76a-8731.2—2000/1) for supporting the study, the Bavarian State Institute for Forestry for the valuable input to the work and the Deutsche Telekom for electrical power support at Haidel.
References Bo¨hme, F., Welsch-Pausch, K., McLachlan, M.S., 1999. Uptake of airborne semivolatile organic compounds in agricultural plants: field measurements of interspecies variability. Environ. Sci. Technol. 33, 1805–1813. Harner, T., Green, N.J.L., Jones, K.C., 2000. Measurements of octanol–air partition coefficients for PCDD/Fs: a tool in assessing air–soil equilibrium status. Environ. Sci. Technol. 34, 3109–3114. Hellstro¨m, A., Kylin, H., Strachan, W.M.J., Jensen, S., 2004. Distribution of some organochlorine compounds in pine needles from Central and Northern Europe. Environ. Pollut. 128 (1–2), 29–48. Herceg Romanic´, S., Krauthacker, B., 2006. Distribution of persistent organochlorine compounds in one-year and two-year-old pine needles bull. Environ. Contam. Toxicol. 77, 143–148. Hippelein, M., Kaupp, H., Do¨rr, G., McLachlan, M., Hutzinger, O., 1996. Baseline contamination assessment for a new resource recovery facility in Germany part II: atmospheric concentrations of PCDD/F. Chemosphere 32, 1605–1616. Horstmann, M., Bopp, U., McLachlan, M.S., 1997. Comparison of the bulk deposition of PCDD/F in a spruce forest and an adjacent clearing. Chemosphere 34, 1245–1254. Jaward, F.M., Di Guardo, A., Nizzetto, L., Cassani, C., Raffaele, F., Ferretti, R., Jones, K.C., 2005. PCBs and selected organochlorine compounds in Italian mountain air: the influence of altitude and forest ecosystem type. Environ. Sci. Technol. 39, 3455–3463. Kaupp, H., Blumenstock, M., McLachlan, M.S., 2000. Retention and mobility of atmospheric particle-associated organic pollutant PCDD/ Fs and PAHs in maize leaves. New Phytol. 148, 473–480. Kirchner, M., Henkelmann, B., Gert, J., Kotalik, J., Fischer, N., Oxynos, K., Schramm, K.-W., 2006. Concentration measurements of PCDD/F in air and spruce needles in the Bavarian Forest and Bohemian Forest (Sumava): first results. Ecotoxicol. Environ. Saf. 63, 68–74. Kylin, H., Sjo¨din, A., 2003. Accumulation of airborne hexachlorocyclohexanes and DDT in Pine Needles. Environ. Sci. Technol. 37, 2350–2355. Niu, J., Chen, J., Henkelmann, B., Quan, X., Yang, F., Kettrup, A., Schramm, K.-W., 2003. Photodegradation of PCDD/Fs adsorbed on spruce (Picea abies (L.) Karst) needles under sunlight irradiation. Chemosphere 50, 1217–1225. Ockenden, W.A., Prest, H.F., Thomas, G.O., Sweetman, A., Jones, K.C., 1998. Passive air sampling of PCBs: field calculation of atmospheric sampling rates by triolein-containing semipermeable membrane devices. Environ. Sci. Technol. 32, 1538–1543. Piccardo, M.T., Pala, M., Bonaccurso, B., Stella, A., Redaelli, A., Gaudenzio, P., Valerio, F., 2005. Pinus nigra and Pinus pinaster needles as passive samplers of polycyclic aromatic hydrocarbons. Environ. Pollut. 133, 293–301.
ARTICLE IN PRESS 306
W. Levy et al. / Environmental Research 105 (2007) 300–306
Rappolder, M., Schro¨ter-Kermani, C., Scha¨del, S., Waller, U., Ko¨rner, W., 2007. Temporal trends and spatial distribution of PCDD, PCDF and PCB in pine and spruce shoots. Chemosphere 67 (9), 1887–1896. Schro¨der, J., Welsch-Pausch, K., McLachlan, M.S., 1997. Measurement of atmospheric deposition of polychlorinated dibenzo-p-dioxins (PCDDs) and dibenzofurans (PCDFs) to a soil. Atmos. Environ. 31, 2983–2989.
Smith, K.E.C., Jones, K.C., 2000. Particles and vegetation: implications for the transfer of particle-bound organic contaminants to vegetation. Sci. Total Environ. 246, 207–236. Umlauf, G., McLachlan, M., 1994. Deposition of semivolatile organic compounds to spruce needles. Environ. Sci. Pollut. Res. 1 (3), 146–150.
Environmental Research 105 (2007) 300–306 www.elsevier.com/locate/envres
Monitoring of PCDD/Fs in a mountain forest by means of active and passive sampling W. Levya,, B. Henkelmanna, G. Pfistera, M. Kirchnera, G. Jakobia, A. Niklausa, J. Kotalika, S. Bernho¨fta, N. Fischera, K.-W. Schramma,b a
GSF-Forschungszentrum fu¨r Umwelt und Gesundheit, Institut fu¨r O¨kologische Chemie, Ingolsta¨dter Landstr. 1, D-85764 Neuherberg, Germany b TUM, Wissenschaftszentrum Weihenstephan fu¨r Erna¨hrung und Landnutzung, Department fu¨r Biowissenschaften, Weihenstephaner Steig 23, 85350 Freising, Germany Received 30 October 2006; received in revised form 9 March 2007; accepted 3 May 2007 Available online 13 August 2007
Abstract Polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/F) are sampled and investigated in a forested area in Middle-Europe. The campaigns, consisting in active and passive samplings, were conducted in the Bavarian and Bohemian Forest at four sites chosen for their similar soil and forest stand characteristics. Passive sampling was conducted using both semi-permeable membrane devices (SPMDs) and needles of well-exposed dominant spruce trees. Active air sampling was also performed at one site with a low volume air sampler. Correlations were performed to identify relationships and trends of PCDD/F. Lower chlorinated PCDD/F are accumulated in SPMDs, needles collected all compounds among homologues and their PCDD/F pattern is close to that of active sampling. Results of the analysed compounds obtained with the different sampling methods served as a basis for the establishment of advantages and disadvantages of the sampling tools applied and their possible optimisation. r 2007 Elsevier Inc. All rights reserved. Keywords: PCDD/F; SPMDs; Active sampling; Spruce needles; Passive sampling
1. Introduction The role of mountain regions in the behaviour and global distribution of persistent organic pollutants (POPs) is crucial. In particular, forests play an important role as regulators of the deposition of these pollutants due to their enhanced surface roughness and the high capacity to retain them in the lipophilic foliages acting as a sink. Semivolatile organic compounds (SOCs) are transferred from the atmosphere to the soil compartment by dry and wet deposition where the vegetation is the intermediate compartment between them (Smith and Jones, 2000). Within the scope of studying the deposition and accumulation of SOCs, different sampling methods were chosen for monitoring persistent compounds to generate a complete overview of the temporal and spatial distribution in the Corresponding author. Fax: +49 89 3187 3371.
E-mail address: [email protected] (W. Levy). 0013-9351/$ - see front matter r 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.envres.2007.05.018
forest system. The accumulation of lipophilic compounds in needles allows them being used as a passive sampler where this accumulation is strongly related to the species (Bo¨hme et al., 1999) and generally increases with the needle age until senescence starts (Hellstro¨m et al., 2004). Coniferous trees have good collection efficiency for both gases and particles and a high interception capacity of rain, fog, and snow precipitation and these factors make this species suitable as a passive sampler (Umlauf and McLachlan, 1994). For these reasons, needles of spruce trees (Picea abies (L.) Karst) are adequate to use as natural passive samplers. Whereas spruce needles consider not only the gaseous uptake but also the uptake of particulate matter, semi-permeable membrane devices (SPMDs) are based on diffusion processes of the compounds mainly related to the gaseous phase (Ockenden et al., 1998). Additionally, SPMDs as integrative passive samplers are of importance because they mimic the accumulation of organic hydrophilic compounds in biological systems in a
ARTICLE IN PRESS W. Levy et al. / Environmental Research 105 (2007) 300–306
301
simplified way. Active sampling with a low volume sampler allows us to determine the concentration of organic compounds in the air. The analytical determinations were focused on dibenzop-dioxins and dibenzofurans (PCDD/F) due to their persistence, bioaccumulation, and toxicity. Correlations among compounds were calculated using results obtained from the active and passive sampler measurements. Comparing the results obtained under the different sampling methods, similarities and differences between them could be investigated. Therewith an additional aim of this study was to determine strength and weakness of each sampler method applied in the monitoring campaign of PCDD/F in mountainous areas.
triolein (1,2,3-tris[cis-9-octadecenoylglycerol]). SPMDs were transported in hermetic clean glass material to and from the place of deployment to prevent possible contaminations caused by transportation. Quadruplicates of these membranes were placed in parallel on square frames. These frames were put into deployment devices of untreated wood, which were exposed at 3.0 m above the ground in small forest clearings. These deployment structures allow a baffle flow through them. The devices protect the SPMDs from direct sunlight and meteorological conditions as precipitation, hindering the wet deposition on the SPMD. After exposure, SPMDs were stored without cleaning their surface at 20 1C until analysis. SPMDs blanks were also kept under identical storage conditions. Needles, SPMDs and cartridges from the active sampler were sampled simultaneously. Thus, the last 3-month exposure coincides for the active and passive samplings in October 2003. The sampling scheme is shown in a chronological timetable (Fig. 1).
2. Material and methods
The procedures for the sample preparation, clean up and quantification of PCDD/F in needles are described in Niu et al. (2003). In brief, SPMDs were cut to slices, spiked with 13C-labeled PCDD/F standard mixtures (Cambridge Isotope Laboratories, USA) and were extracted for 24 h with 100 ml cyclohexane each by use of a rotating shaking machine. After extraction, water was removed with anhydrous sodium sulphate. Afterwards, a clean-up procedure with a sandwich (silica, H2SO4-treated silica), an alumina and a florisil column was conducted. The extract was then concentrated to 10 ml under a stream of nitrogen before quantification. The compounds were determined by highresolution gas chromatography (HRGC) on a Rtx-2330 column for PCDD/F (Restek, Germany). The GC is coupled with a high-resolution mass spectrometer MAT95s (Thermo Electron GmbH, Germany) operated in single ion monitoring mode. Results for PCDD/F were given in air concentration (fg m3) for the active measurements, in mass per dry weight (ng kg1) for the spruce needles and total amount of each compound per tube (ng tube1) for SPMDs. Quadruplicates of SPMDs blanks were analysed for the first sampling period (12 year) and duplicates for the extended period (112 years). Duplicates of the clean-up blanks were also conducted as an internal quality control. The standard recoveries for PCDD/Fs were between 80% and 100% for SPMDs and needles.
2.1. Sampling The sampling was conducted in Eastern-Bavaria at four sites (Mitterfels 1030 m a.s.l., Ruckowitzschachten 1130 m a.s.l., Haidel 1160 m a.s.l., and Boubin 1300 m a.s.l.) chosen for their comparably cold temperature regime and their similar soil and forest stand characteristics. Since the four sites are situated at different distances from the central mountainous barrier of the Bavarian Forest and Sumava (to the southeast: Mitterfels approximately 30 km, Ruckowitzschachten 1 km, and Haidel 8 km; to the north-east, Boubin 20 km), both the precipitation regime and the distance to possible sources in Bavaria and Czech Republic are different. Stem density varies between the four sites increasing from Mitterfels with 40, Haidel with 46, Boubin with 54, and Ruckowitzschachten with 60 m2 ha1 (measurement error73 m2 ha1). Further information about the sampling sites and active and passive sampling methods is available in Kirchner et al. (2006). The low volume active sampler (Digitel blower, DPA96) was deployed at Haidel and comprises the particulate as well as the gas phase in the air sampling. Four campaigns corresponding to the periods August–October 2003, October 2003–January 2004, January 2004–March 2004, and September–October 2004 were conducted in order to establish possible seasonal trends. 1 1 1 2, 12, and 22 years old spruce needles (Picea abies (L.) Karst) were sampled in October 2003 and October 2004, respectively. Needles were collected from the seventh branch whorl of dominant trees of closed spruce stands. Foliar biomass of the sampling sites was also determined in order to detect possible influences of this factor (Jaward et al., 2005). The four spruce forests growing on typical soils of granite and gneiss bedrock and belonging to the same altitudinal zone have been selected to reduce differences attributable to stand characteristics. SPMDs were deployed for 12 and 112 years at the sampling sites and collected in October 2003 and 2004, respectively. SPMD tubes (23 cm 2.5 cm, membrane thickness 67.4 mm) were filled with 0.7 ml
2.2. Analysis
3. Results and discussion 3.1. PCDD/F: needles and active sampler The homologue distribution for dioxins and furans in spruce needles is shown in Fig. 2. The distribution at the four sites showed a PCDD percentage increase towards the higher chlorinated congeners in needles. The trend for PCDF was the opposite one; lower percentage was found for higher chlorinated congeners. This pattern was observed in all needles where less percentage differences
Fig. 1. Chronological table of SPMDs deployment, spruce needles collecting and cartridges sampling.
ARTICLE IN PRESS 302
W. Levy et al. / Environmental Research 105 (2007) 300–306
Fig. 2. PCDD/F homologues distribution (%) in 12 year old spruce needles and SPMDs exposed for M: Mitterfels, and B: Boubin. Active sampling, campaign A at Haidel. Sampling October 2003.
1 2
year at H: Haidel, R: Ruckowitzschachten,
Fig. 3. PCDD/F homologues distribution (%) in 12, 112, and 212 years old spruce needles and active sampling, campaign A at Haidel. Sampling October 2003.
P between OCDD and TetraCDF were found in aged needles (Fig. 3). Therefore, the homologue distribution of PCDD/F is characteristic in all the collected needles with P an increase of TetraCDF related to OCDD in aged needles. Comparing the four sampling sites, a higher P percentage of HeptaCDD and OCDD was detected at Haidel. Atmospheric bulk deposition studied by Horstmann et al. (1997) in a spruce forest near Bayreuth (NorthEastern Bavaria) exhibited the same homologue pattern where the maximal concentration values were also determined for OCDD. Active sampling in Southern Bavaria close to an urban area (Augsburg) performed by Hippelein et al. (1996) reported on average a comparable homologue profile. PCDD/F homologue profiles determined in spruce needles by Rappolder et al. (2007) at locations (Eastern Bavaria and Sumava) close to the monitored one in the present study showed a similar PCDD/F trend where the main percentage P of homologue compounds is shifted from OCDD to TetraCDF. PCDD air concentrations increased from lower to higher chlorinated homologues nonetheless PCDF air concentrations tend to decrease at
higher chlorine substitutions (Fig. 2). Concentrations of PCDD homologues measured in spruce needles and active sampling trials taken from the same location showed a strong correlation (Fig. 4A). A similar correlation was also achieved for older spruce needles and the active sampling (r2 ¼ 0.96 and 0.98 for 112 and 212 years old needles, respectively) pointing out the similar pattern of PCDD found in both sampling methods. PCDD congeners (Fig. 5A) mark the same pattern found in PCDD homologues (Fig. 4A). The congeners in this graphic were chosen considering the compounds that were always detectable or presented low numbers of non-detectable values in the analysed samples. Needles with older ages tend to accumulate compounds (Piccardo et al., 2005; Herceg and Krauthacker, 2006) showing seasonal differences during the year (Kylin and Sjo¨din, 2003). In our study, the bioaccumulation of PCDD/F is observable in older needles in both samplings of October 2003 and 2004. The needles were collected in the earlier autumn season after the summer period where compounds partially bound to particles tend to show a higher accumulation than in the
ARTICLE IN PRESS W. Levy et al. / Environmental Research 105 (2007) 300–306
303
for PCDF in needles is changing with time, whereas this change was not observed for PCDD. This difference obtained for older needles for PCDF could be attributed to the lower chlorinated compounds, TetraCDF and PentaCDF can be accumulated more easily into the leaf compartment than higher chlorinated PCDF homologues due to higher air concentrations and higher availability in the gaseous phase. The factor that during the sampling period the absorption of higher chlorinated compounds is favoured cannot counteract these two above mentioned effects. Another influencing factor is the compound stability. Recent studies indicated a possible increase of lower chlorinated congeners due to photolytic dechlorination (Niu et al., 2003) where PCDF tend to be more reactive than the PCDD. Thus, the higher amounts of PentaCDF and TetraCDF may also be related to an increase of photochemical degradation products accumulated in the needle at longer sunlight exposure (Fig. 5B). Another variable to take into account is the mobilisation of the compounds into the leaf compartments. The work performed by Kaupp et al. (2000) suggested that from the PCDD/F accumulated in the cuticle and classified as particle bounded fraction, PCDD are more immobile than PCDF in maize leaves. If this could also be applied to the needle compartment this may be another factor that implies the major stability in the pattern obtained for PCDD in comparison to PCDF. 3.2. PCDD/F: SPMDS and needles as abiotic and biotic passive samplers
Fig. 4. (A) PCDD homologue concentrations of spruce needles and active sampler (campaign A, 3 months exposure) at Haidel. Diamonds, squares and circles represent 12, 112, and 212 needle ages, respectively. Sampling October 2003. Correlation coefficient between active sampling and 12 year old needles. (B) PCDF homologue concentrations of spruce needles and active sampler (campaign A, 3 months exposure) at Haidel. Diamonds, squares and circles represent 12, 112, and 212 needle ages, respectively. Sampling October 2003. Correlation coefficient between active sampling and 12 year old needles.
rest of the annual seasons (Hellstro¨m et al., 2004). The factor that higher chlorinated PCDD are at higher air concentrations than lower chlorinated PCDD determines the pattern obtained in needles but the season of the sampling also influences the obtained results for PCDD. For PCDF at older needle ages, the correlation between air and needle concentrations decreased clearly (Fig. 4B) due to the different shift of PCDF homologues towards higher concentrations so we cannot infer a direct relationship between the active sampling and needle concentrations for aged needles. Despite this, a good agreement between both methods for all homologues was found for 12 year old needles. Regarding PCDF congeners, only a correlation between the younger needles and the active sampler could be found. This indicates that the concentration distribution
The PCDD/F homologue concentrations obtained from SPMD samples were different from those related to spruce needles. The PCDD/F homologue distribution (in % of the total amount of PCDD/F) is shown in Fig. 2. OCDF was not detectable in the majority of the samples. Taking into account that the SPMD is able to sample mainly the gaseous phase and not the contaminants attached to particulate aerosol the different homologue distribution obtained is not surprising. Homologues with comparable air concentrations but different partitioning behaviour had different distributions in these two passive samplers. This is the case for HeptaCDD and TetraCDF homologues where their air concentrations are similar (and the highest after OCDD air concentrations) but their distribution in both passive samplers is clearly different. HeptaCDD tends to be mainly bound to particles during all seasons meanwhile TCDF is mainly in the gaseous phase (Hippelein et al., 1996). This causes that HeptaCDD homologues able to achieve similar percentage amounts as the TetraCDF homologues in needles are significantly different in the SPMD pattern. Experimental studies in a semi-rural field station showed that 2,3,7,8-TetraCDD/Fs present a higher proportion in the gaseous phase in comparison to the Penta to OCDD/F (Harner et al., 2000). Therefore, higher concentrations of the tetra congeners in the gas phase reflect the pronounced uptake of these compounds into
ARTICLE IN PRESS W. Levy et al. / Environmental Research 105 (2007) 300–306
304
40 Concentration of congeners in active sampler ( fg m-3 air)
OCDD
30 R2 = 0,99 20
10
1,2,3,4,6,7,8HpCDD HxCDD 2 HxCDD 1 HxCDD 3
0 1,2,3,7,8-PeCDD 0.0
2.5
5.0
7.5
Concentration of congeners in needles (ng
kg-1
dry wt)
Concentration of congeners in active sampler (fg m-3 air)
9 1,2,3,4,6,7,8HpCDF
R2 = 0.92
6
HxCDF 3 3 HxCDF 2
PeCDF 1
HxCDF 1
2,3,7,8-TCDF PeCDF 2
0 0.00
0.25
0.50
0.75
Concentration of congeners in needles (ng kg-1 dry wt) Fig. 5. (A) PCDD congener concentrations in spruce needles and active sampler 3 months exposure (campaign A) at Haidel. Diamonds, squares, and circles represent 12, 112, and 212 needle ages, respectively. HxDD 1: 1,2,3,6,7,8-HxCDD, HxDD 2: 1,2,3,7,8,9-HxCDD, HxDD 3: 1,2,3,4,7,8-HxCDD. Sampling October 2003. Correlation coefficient between active sampling and 12 year old needles. (B) PCDF congener concentrations in spruce needles and active sampler 3 months exposure (sampling A) at Haidel. Diamonds, squares and circles represent 12, 112, and 212 needle ages, respectively. PeCDF 1: 1,2,3,7,8/1,2,3,4,8-PeCDF, PeCDF 2: 2,3,4,7,8-PeCDF, HxCDF1: 1,2,3,4,7,8/1,2,3,4,7,9-HxCDF, HxCDF 2: 2,3,4,6,7,8-HxCDF, HxCDF 3: 1,2,3,6,7,8HxCDF. Sampling October 2003. Correlation coefficient between active sampling and 12 year old needles.
SPMDs in comparison to the spruce needles. The tetra homologues peaked in the distribution of all sampling sites and it is also observed an increase for the Penta and HexaCDD/F in comparison to the HeptaCDD/Fs and OCDD. Fig. 5 points out these observations. In the study of Harner et al. (2000), the division of PCDD/Fs in two groups was suggested according to their chlorine substitution pattern where the group 2 (any Tetra to HexaCDD/F with three or four chlorines in the 2,3,7,8 substitution) is characterised for uniform charge distribution in comparison to the group 1 (the rest of PCDD/F). The lower polarity exhibited in group 2 means higher octanol solubility, thus a higher KOA (partition coefficient octanol-air) for this PCDD/F group. This implies a better uptake into the SPMD device for the group 2 in
comparison to HeptaCDD/Fs and OCDD/F that belong to the group 1. On the other hand, SPMDs exposed for 112 years do not show linear increase of PCDD/F concentrations compared to the 12 year exposed devices but a relative P increasing of Tetra P to HexaCDF compounds (Fig. 6). TetraCDD and PentaCDF increase signifiP cantly, even the HexaCDF, close to the limit of detection after 12 year exposure is well determined at longer exposure periods. This is concordant with the study of Schro¨der et al. (1997) where it was determined that of the whole PCDD/Fs homologues dry gaseous deposition occurs for TetraCDD and Tetra to HexaCDFs. The importance of this process in the uptake of compounds in SPMDs is clear when we compare the homologue patterns of SPMDs and needles in Fig. 2. It has to be considered that longer periods
ARTICLE IN PRESS W. Levy et al. / Environmental Research 105 (2007) 300–306
Fig. 6. PCDD/F homologue concentrations in SPMDs
1 2
305
and 112 years exposure at three sites. Samplings October 2003 and 2004.
allow the mass transfer from the particles attached to the membrane and the membrane itself into the SPMD. The homologue distribution of SPMDs presented a similar pattern after 112 years exposure at all sampling sites, which was not so clearly observed for SPMDs exposed 12 year (Fig. 6). Extended exposure periods can cause non-linear uptake so such prolonged periods are not advised for environmental kinetic sampling with SPMDs. PIn conclusion, theP lower chlorinated compounds, TetraCDD/Fs and PentaCDD/Fs, presented a better uptake in SPMDs devices whereas higher chlorinated homologues, OCDF in particular, are hardly detectable or not detected in these devices after 6-month exposure. Needles in the same exposure period collected compounds from all the studied homologue groups following similar distribution at all sampling sites. In summary, SPMDs appear to be a very feasible tool for lower chlorinated PCDD/Fs, in particular TetraCDD/Fs due to the ability to uptake compounds from the gaseous phase. On the other hand, needles are less ‘selective’ than SPMDs but being able to uptake compounds from both, the particle bound and gaseous phase and can embrace all the homologue families in spite of the less pronounced uptake of lower chlorinated PCDD/Fs. 4. Funding The funding sources that made possible this work were facilitated from the Bavarian State Ministry of the Environment, Health and Consumer Protection under the project number 76a-8731.2—2000/1. Acknowledgments The authors wish to thank the Bavarian State Ministry of the Environment, Health and Consumer Protection (Project no. 76a-8731.2—2000/1) for supporting the study, the Bavarian State Institute for Forestry for the valuable input to the work and the Deutsche Telekom for electrical power support at Haidel.
References Bo¨hme, F., Welsch-Pausch, K., McLachlan, M.S., 1999. Uptake of airborne semivolatile organic compounds in agricultural plants: field measurements of interspecies variability. Environ. Sci. Technol. 33, 1805–1813. Harner, T., Green, N.J.L., Jones, K.C., 2000. Measurements of octanol–air partition coefficients for PCDD/Fs: a tool in assessing air–soil equilibrium status. Environ. Sci. Technol. 34, 3109–3114. Hellstro¨m, A., Kylin, H., Strachan, W.M.J., Jensen, S., 2004. Distribution of some organochlorine compounds in pine needles from Central and Northern Europe. Environ. Pollut. 128 (1–2), 29–48. Herceg Romanic´, S., Krauthacker, B., 2006. Distribution of persistent organochlorine compounds in one-year and two-year-old pine needles bull. Environ. Contam. Toxicol. 77, 143–148. Hippelein, M., Kaupp, H., Do¨rr, G., McLachlan, M., Hutzinger, O., 1996. Baseline contamination assessment for a new resource recovery facility in Germany part II: atmospheric concentrations of PCDD/F. Chemosphere 32, 1605–1616. Horstmann, M., Bopp, U., McLachlan, M.S., 1997. Comparison of the bulk deposition of PCDD/F in a spruce forest and an adjacent clearing. Chemosphere 34, 1245–1254. Jaward, F.M., Di Guardo, A., Nizzetto, L., Cassani, C., Raffaele, F., Ferretti, R., Jones, K.C., 2005. PCBs and selected organochlorine compounds in Italian mountain air: the influence of altitude and forest ecosystem type. Environ. Sci. Technol. 39, 3455–3463. Kaupp, H., Blumenstock, M., McLachlan, M.S., 2000. Retention and mobility of atmospheric particle-associated organic pollutant PCDD/ Fs and PAHs in maize leaves. New Phytol. 148, 473–480. Kirchner, M., Henkelmann, B., Gert, J., Kotalik, J., Fischer, N., Oxynos, K., Schramm, K.-W., 2006. Concentration measurements of PCDD/F in air and spruce needles in the Bavarian Forest and Bohemian Forest (Sumava): first results. Ecotoxicol. Environ. Saf. 63, 68–74. Kylin, H., Sjo¨din, A., 2003. Accumulation of airborne hexachlorocyclohexanes and DDT in Pine Needles. Environ. Sci. Technol. 37, 2350–2355. Niu, J., Chen, J., Henkelmann, B., Quan, X., Yang, F., Kettrup, A., Schramm, K.-W., 2003. Photodegradation of PCDD/Fs adsorbed on spruce (Picea abies (L.) Karst) needles under sunlight irradiation. Chemosphere 50, 1217–1225. Ockenden, W.A., Prest, H.F., Thomas, G.O., Sweetman, A., Jones, K.C., 1998. Passive air sampling of PCBs: field calculation of atmospheric sampling rates by triolein-containing semipermeable membrane devices. Environ. Sci. Technol. 32, 1538–1543. Piccardo, M.T., Pala, M., Bonaccurso, B., Stella, A., Redaelli, A., Gaudenzio, P., Valerio, F., 2005. Pinus nigra and Pinus pinaster needles as passive samplers of polycyclic aromatic hydrocarbons. Environ. Pollut. 133, 293–301.
ARTICLE IN PRESS 306
W. Levy et al. / Environmental Research 105 (2007) 300–306
Rappolder, M., Schro¨ter-Kermani, C., Scha¨del, S., Waller, U., Ko¨rner, W., 2007. Temporal trends and spatial distribution of PCDD, PCDF and PCB in pine and spruce shoots. Chemosphere 67 (9), 1887–1896. Schro¨der, J., Welsch-Pausch, K., McLachlan, M.S., 1997. Measurement of atmospheric deposition of polychlorinated dibenzo-p-dioxins (PCDDs) and dibenzofurans (PCDFs) to a soil. Atmos. Environ. 31, 2983–2989.
Smith, K.E.C., Jones, K.C., 2000. Particles and vegetation: implications for the transfer of particle-bound organic contaminants to vegetation. Sci. Total Environ. 246, 207–236. Umlauf, G., McLachlan, M., 1994. Deposition of semivolatile organic compounds to spruce needles. Environ. Sci. Pollut. Res. 1 (3), 146–150.