Assessment of Chlorinated Dibenzo-p-dioxin and Dibenzofuran

7137 Wallace Drive, Brentwood Bay, British Columbia, Canada V8M 1G9, .... assessment of crab consumers from the Newark Bay Complex, New Jersey, USA...
0 downloads 0 Views 175KB Size
Research Assessment of Chlorinated Dibenzo-p-dioxin and Dibenzofuran Trends in Sediment and Crab Hepatopancreas from Pulp Mill and Harbor Sites Using Multivariate- and Index-Based Approaches M A R K B . Y U N K E R , * ,† WALTER J. CRETNEY,‡ AND MICHAEL G. IKONOMOU‡ 7137 Wallace Drive, Brentwood Bay, British Columbia, Canada V8M 1G9, and Contaminants Science Section, Institute of Ocean Sciences, Box 6000, Sidney, British Columbia, Canada V8L 4B2

Mill process changes and source controls instituted in the late 1980s at pulp and paper mills along the British Columbia (BC) coast produced dramatic reductions in both the concentrations of polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs) and the TEQ (2,3,7,8-TCDD toxic equivalents). Nevertheless, questions remained as to whether the sediments were acting as contaminant sources or sinks, whether crab uptake pathways have stayed the same since the mills ceased producing PCDD/ Fs, and whether improvement was faster at some sites than others. To examine the processes governing the changes in PCDD/F composition, PCDD/F sources and compositional trends for the coastal BC sites were interpreted using sediment and crab PCA (principal components analysis) models and two indices based on the three main groups of PCDD/Fs that covary in all PCA models. One index measured the relative inputs of chlorine bleaching PCDFs and PCP (pentachlorophenol)-related PCDDs at each mill and harbor site whereas the other measured the relative inputs of PCP wood preservatives and pulp mill effluents. Results indicated that PCDFs produced during chlorine bleaching accumulate from sediments to crabs more readily than the PCDDs but that the chlorine bleaching TCDFs have generally decreased in importance at all mill sites. In the years just before 1995, little change was observed in either the proportion of toxic 2,3,7,8-chlorinated congeners or the PCDD/F concentrations, and the improvements seemed to have largely ceased by 1995. The lack of change in PCA composition for the harbors provided a further, disturbing indication that ecosystem recovery may have stalled. Results also implied that the main route of PCDD/ Fs into crabs shifted from uptake via a pelagic food web incorporating effluent-borne PCDD/Fs associated

* Corresponding author e-mail: [email protected]; telephone: (250)363-6805; fax: (250)363-6807. † 7137 Wallace Drive. ‡ Institute of Ocean Sciences. 10.1021/es0112893 CCC: $22.00 Published on Web 03/22/2002

 2002 American Chemical Society

with suspended particulate matter to uptake via a benthic food web.

Introduction In coastal British Columbia, Canada (BC; Figure 1), the proximity of forest-covered mountains to sheltered ports with abundant freshwater led to the establishment of marine pulp and paper mills in settings with diverse and abundant aquatic life and some of the world’s largest salmon runs (1). Accordingly, the discovery in 1987 of polychlorinated dibenzop-dioxins and dibenzofurans (PCDD/Fs) in ecologically significant concentrations in sediments, shellfish, and groundfish near BC mills and in BC harbors (2-6) quickly led to coast-wide monitoring programs and extensive fisheries closures (7). The major processes that have produced PCDD/Fs in pulp mill effluent are the chlorination of dibenzofuran and dibenzo-p-dioxin during pulp bleaching and the condensation of polychlorinated phenoxyphenols during pulp digestion (8-10). Hence, the elimination of dibenzofuran and dibenzo-p-dioxin from defoamer products, the exclusion of chlorophenol-contaminated wood chips, and the introduction of chlorine dioxide bleaching have led to a dramatic reduction in the production of mill-related PCDD/Fs (7, 1113). The result has been a decline in PCDD/F concentrations in biota, and in recent years a number of previously closed fisheries areas have been re-opened (7, 13). Mill-related PCDD/Fs from both chlorine bleaching and the condensation of polychlorinated phenoxyphenols are present in the marine ecosystem at the nine BC mill sites that use a bleached-kraft process (Figure 1) (6, 13). Chlorinebleaching PCDD/Fs are low or absent at Kitimat and Port Alice where the mills use, respectively, an unbleached-kraft process without chlorine and sulfite bleaching with mild chlorination (6). PCDD/Fs from pentachlorophenol (PCP) wood preservatives (8, 9, 14) also are present at locations with sawmills or wood treatment facilities, at storage sites in the BC harbors, and near docks at pulp mills where chlorophenol-contaminated wood chips have been unloaded (6). Finally, PCDD/Fs from various combustion processes and the photolysis of PCP in atmospheric condensed water (primarily OCDD with lesser amounts of HpCDDs) are ubiquitous in sediments (15-19). Along the BC coast, male Dungeness crab (Cancer magister) was chosen as the principal sentinel organism. Crustaceans have been particularly useful as biomonitors because they have limited ability to metabolize chlorinated contaminants and the PCDD and PCDF contaminants in crab tissues often mirror surface sediments for the lesschlorinated congeners (e.g., refs 13, 20, and 21). Previously, we have demonstrated that oceanographic factors at the mill sites affected the rate of decrease of PCDD/F contamination in crabs between 1987 and 1993 (6, 13), while the bioavailability of PCDD/Fs remained independent of environmental factors (e.g., oceanic environment, depositional regime, sediment organic carbon) and compositional differences for the sedimentary PCDD/Fs (22). However, questions remained as to whether the sediments were acting as contaminant sources or sinks, whether crab uptake pathways have remained the same since the mills ceased producing PCDD/ Fs, and whether improvement was still faster at some sites than others. VOL. 36, NO. 9, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1869

FIGURE 1. Map of the British Columbia coast showing the pulp mills and principal harbors. Accordingly, the present study adds analyses of 1987PCDD/F analyses by Axys Analytical Ltd. of 89% of the 1995 sediment and 1994-1995 crab hepatopancreas samples sediment and 74% of the crab PCA samples used the general to the 1987-1993 crab data set. It compares PCDD/F procedure described in ref 25. Samples were spiked with 13C-labeled surrogate standards for 2,3,7,8-TCDD, 1,2,3,7,8distributions between sediment and crab and uses principal components analysis (PCA) to interpret PCDD/F sources and PnCDD, 1,2,3,6,7,8-HxCDD, 1,2,3,4,6,7,8-HpCDD, OCDD, trends for the coastal BC sites. The PCA model results also and 2,3,7,8-TCDF; as a minimum 13C-labeled 1,2,3,4-TCDD are used to define a pair of indices that encapsulate the most was added as a recovery standard. For samples analyzed by important PCDD/F compositional information and to fahigh-resolution GC/MS (i.e., 1991 onward), 13C-labeled cilitate the interpretation of changes and trends in PCDD/F 1,2,3,7,8-PnCDF, 1,2,3,4,7,8-HxCDF, and 1,2,3,4,6,7,8-HpCDF composition and concentration at the BC mill and harbor surrogates and 1,2,3,7,8,9-HxCDD and 1,2,3,4,6,7,8-HpCDD sites. recovery standards were also added. Minor method differences occurred with the other laboratories (listed in the Methods Supporting Information), but all analyses employed 13Clabeled standards (26, 27). A detailed examination of the PCDD/F concentration data for coastal BC from 1987 to 1995 data set has found no inter-laboratory bias (this work and were available for 310 surficial sediment and 686 male ref 6). Dungeness crab hepatopancreas samples. Yunker and Concentration data were available for all 2,3,7,8-substiCretney (22) list the monitoring program reports contuted PCDD/F congeners and the non-2,3,7,8-homologue taining all mill site data. Vancouver Harbour sediment data totals (Table 1). 1,2,3,4,7,8-HxCDD, 1,2,3,4,7,8,9-HpCDF, and for 1991 and 1992 were from ref 23 and Goyette (unpublished). the 2,3,7,8-substituted HxCDFs were undetectable in most Data have been archived in the Canadian National Consamples (Table 1) and were excluded from PCA models. OCDF taminants Information System and can be obtained by writing was undetectable in almost all crab samples and was only to the Division Information Manager at the same address as used for sediment PCA. OCDD and the HpCDFs were W.J.C. undetectable in ca. 30% of the crab samples and were used Samples with fewer than nine detectable congeners (ca. only after it was verified that their presence did not skew the one-half of the congener total) were removed from the data PCA results. Undetectable values (average 14% for sediment set to produce a 1990-1995 sediment PCA model with 280 and 8.8% for crab) for all PCA congeners and homologue samples and a 1987-1995 male Dungeness crab hepatototals were replaced by the limit of detection (typically 0.1pancreas model with 671 samples (the sediment and 19870.4 pg/g; 2-10 pg/g for pre-1990 low-resolution MS data) 1995 crab PCA models, respectively). An additional PCA model (22). Data were midrange normalized (28) and autoscaled with 361 samples was also constructed that contained just (scaled to the mean and unit variance) before PCA, and then recent (1993-1995) crab samples from the mill sites plus a Varimax rotation (29) was applied to the first three principal selected harbor and reference site samples (the 1993-1995 components (PCs) to simplify the physical interpretation of crab PCA model). Twenty-four other 1987 and 1988 sediment the PCA projections (22). The potential density functions for samples (2, 24) had either incomplete or missing data for the 3D contour plots were generated with routines POTJACK, 2,3,7,8-substituted congeners or unsuitable detection limits, POTFUN, and POTLINE from Q-Parvus 3.0 (Italian School and these samples were used only for concentration comof Chemometrics, Genoa, Italy) and plotted using Surfer 7.0 parisons (Figure 2 and Supporting Information Table 1). 1870

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 9, 2002

FIGURE 2. Year-to-year changes in ∑PCDD/F for sediment and male Dungeness crab hepatopancreas (ng/g dry and wet weight, respectively) and in the crab hepatopancreas TEQ (2,3,7,8-TCDD toxic equivalents referenced to fish in ng/g wet weight) at representative locations close to each pulp mill. Mill sites from fjord/restricted circulation areas and areas well-flushed by tidal currents are shown separately. (Golden Software Inc.). Fixed Gaussian potential functions and the equivalent determinant method were used to estimate density.

Results and Discussion PCDD/F Concentration Trends. At most mill sites, the PCDD/F concentration total (∑PCDD/F) and TEQ (2,3,7,8TCDD toxic equivalents referenced to fish; Table 1) (30) for Dungeness crab hepatopancreas has declined, while ∑PCDD/F for sediment generally has exhibited less change over time (Figure 2; Supporting Information Table 1) (6). ∑PCDD/Fs for both sediment and crab hepatopancreas are typically 10-100 times higher at the nine mill sites using a kraft process with chlorine bleaching than at Kitimat or (based on very limited data) Port Alice (Figure 1; Supporting Information Table 1). For Port Alice, the few data available suggest that ∑PCDD/F in crab hepatopancreas is very low (