As part of the work to examine the impact of development on ecosystem health and integrity on the Peace and Athabasca river basins in Alberta, the Northern River Basin Study (NRBS) was required to determine “the contents and nature of the contaminants entering the system ... particular reference to water, sediments and biota" and to determine “... the current concentration of contaminants in water and edible fish tissue and how are these levels changing through time and by location". The Reach Specific Study (RSS) was designed to measure spatial and temporal trends of contaminants including polychlorinated dibenzo-p-dioxins and -dibenzofurans (PCDD/Fs) in sediment, water and and biota (fish and invertebrate) samples collected at six locations on the upper Athabasca River downstream of Hinton (AB) in spring 1992, fall 1992 and spring 1993. The General Fish Collection (spring 1992), the Long nose sucker and Northern pike liver study (fall 1994) and the Special Burbot Collection (fall 1992 and 1994), and the Ft. Chipewyan winter fishery study (1994/95) were also conducted to examine levels in fish tissues within the Athabasca, Peace and Slave River basins. The purpose of this report is to summarize the levels of PCDDs and PCDFs in fish from these various studies and to assess temporal trends of 2,3,7,8-TCDD and 2,3,7,8-TCDF by comparison with previously published data. A second objective was to reexamine pathways of accumulation of 2,3,7,8-TCDD and 2,3,7,8-TCDF from water and suspended sediment to fish, first measured in the upper Athabasca River in 1992 (Pastershank and Muir 1995).
The major PCDD/F congeners in muscle (skinless fillet) of mountain whitefish and northern pike samples collected in the upper Athabasca River downstream of Hinton in fall 1992 and spring 1993 were 2,3,7,8-TCDD and -TCDF. Mean concentrations of 2,3,7,8-TCDD in mountain whitefish ranged 0.6 to 7.7 pg-g'1wet wt and from 1.7 to 9.8 pg-g"1for 2,3,7,8-TCDF. Concentrations of other 2,3,7,8- substituted penta- to octachloro- PCDD/F congeners were generally much lower or non-detectable in both species. Two lower chlorinated congeners, 2,7/2,8-dichlorodibenzodioxin and 2,3,8- trichlorodibenzofuran were detected in most samples ofmountain whitefish from fall 1992 at low pg-g'1 concentrations. TCDF was the most frequently detected PCDD/F congener in longnose sucker and northern pike livers collected from the Wapiti/Smoky and Peace Rivers in fall 1994. TCDF concentrations in liver were in the low pg-g'1range similar to levels in muscle of these species. Highest concentrations of TCDF in livers of longnose sucker (9.2 ± 17.8 pg-g"1) were found at a site on the Smoky River (SRI) downstream of the pulp mill effluent near Grande Prairie.
Temporal trends in 2,3,7,8-TCDD and -TCDF in mountain whitefish were examined over a four year period by combining the three sampling times in the upper Athabasca River with data from previous studies (DFO National Dioxin Program 1989). There was a definite decline in 2,3,7,8-TCDD and - TCDF concentrations in mountain whitefish downstream of the Hinton but most of the decrease took place in the period 1989 to 1992. The extent of the decline depends to a large extent on which results for spring 1993 are used. If samples from the near-field sites of Weldwood and Obed (mean concentrations of 1.1 and 2.6 pg-g"1wet, for TCDD and TCDF respectively) are used the decline is about five-fold for both TCDD and TCDF over four years. But if the fish from Emerson Lake (48 km downstream) are included (mean concentrations are 3.6 and 7.1 pg-g"1wet, for TCDD and TCDF, respectively) the decline is about 3-fold.
In general, concentrations of PCDD/Fs were higher in burbot liver than in muscle or liver of mountain whitefish or northern pike and a greater number of congeners were detected. TCDF was detected
(mean concentrations, 0.30 to 65 pg-g'1) in 86% o f all 203 burbot liver samples analysed, while 2,3,7,8- TCDD was detected in 35% of samples (mean concentrations, <0.3 to 8.5 pg-g'1). Two other 2,3,7,8- substituted- PCDD/F congeners, 1,2,3,6,7,8-HxCDD and the heptachlorodioxin, 1,2,3,4,6,7,8-HpCDD were detected in 37% o f burbot liver samples. OCDD was also detected relatively frequently (17%) while OCDF was found in only 3 of 203 samples. Di and trichloro-CDDs and CDFs were detected infrequently in burbot liver and at low levels relative to tetra- to octachloro congeners. Significantly higher levels (ANCOVA; Tukey’s or least squares means test) of TCDD and TCDF were found in burbot liver downstream of the Hinton BKM than at all other sites.
Levels of 2,3,7,8-TCDD and -TCDF in burbot liver were lower in the. fall 1994 collection than in fall 1992 at four sites; downstream of the Grande Prairie pulp mill outlet, PR2 on the Peace River near the mouth ofthe Notikewin River (674 km from confluence ofthe Peace/Slave), and PR3 upstream ofFort Vermillion (396 km). Comparison of concentrations in burbot liver near the BKM at Grande Prairie was problematic because sampling sites were not in the same locations each year. Nevertheless, the results show a decline of 4 to 17-times in the case of 2,3,7,8-TCDF at three sites. No significant decline of TCDD or TCDF concentrations was found in burbot livers from PR2. The burbot liver results, expressed as TCDD TEQ’s, also agreed well with those of Swansonet ak (1995) who found a 5-fold decline in TEQs downstream of the Grande Prairie BKM between summer 1991 and spring
Concentrations of all 2,3,7,8-substituted PCDD/F congeners in composite samples of fish muscle from the Ft. Chipewyan domestic winter fishery in the Peace-Athabasca delta were at or near detection limits (<0.1 to <0.8 pg-g'1). Only 2,3,7,8-TCDF was detectable in most samples (<0.1 to 0.5 pg-g'1). Burbot liver samples from the three sites in the Peace-Athabasca delta had higher levels o f 2,3,7,8-TCDF than burbot muscle (1.7 to 2.9 pg-g'1). These levels were similar to those at other far-field and reference sites located far from BKMs.
The bioavailability of TCDD and TCDF to mountain whitefish and northern pike was assessed using biota-sediment (or suspended sediment) accumulation factors (BSAF/BSSAFs). BSAFs for 2,3,7,8- TCDD ranged from 1.1 to 2.0 and for TCDF from 0.19 to 1.63 in mountain whitefish in spring 1992. A similar range of BSAFs was found in 1993. BSSAFs for both 2,3,7,8-TCDD and TCDF were generally lower and showed greater consistency than BSAFs with distance from the BKM. The results suggest that TCDD/TCDF levels in fish can be estimated with an average, site specific, BSAF or BSSAF using concentrations of TCDD/F in bed sediment or suspended sediments. Application of the Thomann and Connolly food chain model (steady-state version) to predict levels of TCDF in the food web downstream of Hinton showed that good agreement between predicted and observed results could be obtained for benthic feeding organisms (and longnose suckers and pike) which were close to equilibrium with sediments or biofilm. The model overpredicted concentrations in filter-feeding invertebrates and mountain whitefish; these organisms are not in equilibrium with TCDF in the water and suspended solids in the river due to the dynamic nature ofthe system.
All mean concentrations of TCDD TEQs in fish muscle or liver were below the limit of 20 pg-g'1(wet wt) set by Health Canada for commercial sale and export of fish. A few individual samples, mainly burbot liver from the Athabasca River downstream of Hinton, exceeded the 20 pg-g'1guideline. Assuming TCDD TEQs of 8.3 pg-g'1in mountain whitefish downstream of Hinton a 60 kg individual would have to consume 72 g of mountain whitefish muscle per day to exceed the Health Canada