Environ. Sci. Technol. 2005, 39, 984-990
Storm Disturbance of Sediment Contaminants at a Hot-Spot in the Baltic Sea Assessed by 234Th Radionuclide Tracer Profiles M I C H A E L K E R S T E N , * ,† THOMAS LEIPE,‡ AND FRANZ TAUBER‡ Geoscience Institute, Gutenberg-University, Becherweg 21, D-55099 Mainz, Germany, and Baltic Sea Research Institute, Seestrasse 15, D-18119 Rostock-Warnemu ¨ nde, Germany
Fly ash sludges from an abandoned metal smelter were dumped into the shallow inner part of the Mecklenburg Bay until 1971, representing the most severe heavy metal contamination hot-spot along the German coast. Half of the dumped Zn (455 t) and Pb (173 t) inventory was found to be spread from the originally 0.5 km2 hot-spot site to a now 360 km2 affected adjacent area. Wave-driven resuspension during gale events produced large pulses of contaminated sediments from this hot-spot due to the only 23 m water depth. Instantaneous sediment mixing down to 10 cm occurs during such a wave event as evidenced by activity profiles of the short-lived radionuclide 234Th in sediment cores. According to these estimated sediment exchange fluxes in the transport bottom area, each wave event may have mobilized Zn and Pb pulses on the order of several hundreds of kilograms from the dump site, sufficient to build up a plume in sediments of the outer bay area. With each centimeter (∼5 yr) of additional natural sediment capping, however, the amount of metal remobilization would decrease by about 50%.
Introduction Fly ash sludges from the blast furnace of the now abandoned copper smelter “Metallhu ¨ttenwerke Lu ¨beck AG” were dumped together with other unspecified industrial waste into the inner part of the Mecklenburg Bay until 1971. This represents the most severe heavy metal and organic contamination hotspot along the whole German coast. Monitoring data show redistribution and transport of metals from this relatively shallow (23 ( 2 m) spot into the outer bay area (1, 2). Owing to difficulties in obtaining measurements and to vagaries of the weather at sea, few field data exist pertaining to sediment mobilization and transport in storm conditions (3). Estimations based on geochemical proxys indicate that 50-80% of what is finally settling out in the deep basins of the Baltic derives from storm-induced erosion processes (4). This does not necessarily mean, however, that there is no net sedimentation in the dynamic Baltic coastal areas. Many of these areas may rather be characterized by transport bottoms, with merely episodic deposition of fine particles, where periods of accumulation are interrupted by periods of erosionresuspension-transportation. To quantitatively assess the * Corresponding author phone: +49 6131 392 4366; fax: +49 6131 392 3070; e-mail:
[email protected]. † Gutenberg-University. ‡ Baltic Sea Research Institute. 984
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 4, 2005
storm-induced sediment turnover and concomitant contaminant redistribution from the hot-spot area, sediment profiles of the naturally occurring radiogenic isotope 234Th have been measured at this and a reference site. This radionuclide is produced in seawater from the decay of 238U, and is particularly useful in assessments of suspended particulate matter (SPM) transport in coastal environments due to its relatively rapid decay (5-8). The parent nuclide 238 U is dissolved in seawater under oxidizing conditions as the anionic uranyl carbonate complex, and therefore behaves conservatively with salinity. The daughter nuclide 234Th, on the other hand, is present in the form of the particle-reactive hydrolysis product Th(OH)n(4-n)+ (5). Measured distribution coefficients of the latter are on the order of Kd ≈ 106 L kg-1, while 238U has only a maximum Kd ≈ 102 L kg-1. 234Th is therefore scavenged rapidly by SPM accumulating in excess of the background fraction supported by the particulate 238U load. The disequilibrium which is induced between 234Th and its parent 238U has for many years been recognized as a radiotracer of particle residence times in the marine environment (9). In coastal systems with no significant river input as in the present case study, scavenging particles originate mainly from biological processes and sediment resuspension. Both these processes vary with the season, but not necessarily in phase, and may control the total concentrations and residence times of 234Th (10). In a previous paper, SPM residence times were determined for the Mecklenburg Bay on the basis of the short-term variability in the partitioning of 234Th in the water column (11). Once scavenged by SPM and deposited to the bottom, particulate 234Th decays too rapidly to accumulate in sediment (half-life t1/2 ) 24.1 days). Its distribution in surficial layers is therefore purely controlled by sediment mixing (12). In the present study, the 234Th inventory in sediment cores near the dump site was therefore measured to detect short-term wave-induced sediment turnover events, and to balance the contaminant redistribution during such events within the bay area. Ecotoxicological effects of this sediment turnover were assessed in another study (2).
Methodology and Sampling Site Characteristics Study Area. The present-day morphology of the western Baltic seafloor was formed by repeated advances and retreats of ice during the Pleistocene and the Holocene Transgressions. The advancing ice masses dug basins and troughs in the continental shoreline. Mecklenburg Bay is such a trough with a maximum depth of 28 m, i.e., well below the fair weather wave base. Up to 7 m of mud is accumulated below the 20 ( 2 m isobath, which includes over half of the bay (1400 out of 2500 km2, Figure 1), the remainder being fineto coarse-grained sand. Net sediment accumulation rates of 30-65 mg cm-2 yr-1 (dry weight) were reported for mud sampling sites in Mecklenburg Bay on the basis of 210Pb radionuclide dating of (albeit severely disturbed) sediment cores (13). During the course of a year, the SPM composition is controlled by enhanced biological production in spring and summer, and wave-driven resuspension in the autumn and winter periods. Tidal currents in the semienclosed Baltic are negligible, and the hydrographic structure is mostly related to atmospheric forcing. There is a large seasonal and a minor annual wind variation in the area. During summer, stratification results from heating of the upper layer, and leads to anoxia in bottom waters. Winter cooling and winter storms remove much of the upper layer stratification. Since the Baltic Sea is fetch limited, the dominant wind direction is important 10.1021/es049391y CCC: $30.25
2005 American Chemical Society Published on Web 01/11/2005
FIGURE 1. Schematic maps showing for Mecklenburg Bay the bathymetry and spatial monitoring sampling site grids: (a) sideways (orange) triangle, 1983 campaign; upright (red) triangle, 1985 campaign; upside down (green) triangle, 1997 campaign; (red) square, location of sampling site nos. 12 and 23a, the latter near the dump site; (b) surface (top 2 cm) sediment concentrations for Zn determined for the 1983/1985 monitoring campaigns (data from ref 1); (c) the same for the 1997 campaign (data from ref 2); (d) differential map showing the change in surface sediment Zn concentrations between the 1983/1985 and 1997 campaigns. for the maximum wave heights. The brackish salinity conditions also change both vertically and seasonally due to admixture of less saline waters from the east and more saline waters from the northwest. Winter storms can introduce sufficient quantities of saltwater from the North Sea into the bay, adding to the pronounced summer stratification through the buoyancy input. A comparison of the near-bottom hydrodynamic conditions by bottom tripod and currentmeter mooring and experimentally derived critical shear stress velocities suggests that particle transport is controlled by storm events in winter, whereas under calm summer season conditions shear stress velocities do not exceed the critical values (14). Bottom current scour leads in the former case to intense resuspension of the surface mud layer to a yet unknown depth, and to elevated near-bottom SPM concentrations in the range of 1-10 mg L-1 (11). Light scattering measurements have shown that in the winter season the SPM-rich nepheloid layer may extend up to 2-5 m above the bottom, and be transported by the bottom
currents counterclockwise through the bay to the northeast (15). Sediment Sampling. Two sampling sites in the central and southern parts of the Mecklenburg Bay were studied which are part of the long-term pollution monitoring network of the Baltic Sea Research Institute (Figure 1). Sampling site no. 12 at 54°18.90′ N and 11°33.00′ E in the outer part of the bay (water depth of 25 m) is the same as the International Baltic Monitoring Program reference sampling site BMPM2, while sampling site no. 23a (54°02.85′ N, 11°03.50′ E) is situated in the inner part of the bay less than 5 nautical miles east of the hot spot. Both sites are covered by mud (>90% silt sized), but the clay mineral content was higher at station no. 23a (4.3 ( 0.7%) than at station no. 12 (3.1 ( 0.6%). Sediment cores for thorium radiotracer analysis were taken on two cruises November 8, 1994, and September 12, 1995. The water column during the November 1994 sampling campaign was virtually completely mixed as evidenced by a low vertical salinity and temperature gradient between the VOL. 39, NO. 4, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
985
surface and bottom water. In September 1995, a salinity rise between 9 in surface water to 24 in bottom water indicated a clear-cut halocline, typical throughout the calm summer season. All sediment cores were collected with a multicorer of 10 cm i.d. Perspex tubes. The sediment cores were retrieved by pressing a piston from the bottom up through the multicorer tubes, and cut by 1 cm increments into clean plastic dishes, where the outer 1 cm rim was immediately trimmed off each slice to reduce cross contamination from adjacent slices during coring (i.e., the Petri dishes had an 8 cm i.d.). All sediment subsamples were kept frozen until processing in the laboratory. The porosity (i.e., volumetric water content) and dry bulk density of each sediment slice of known volume were determined by weight loss upon freeze-drying. The resulting dry density of the sediment throughout the whole core is 2.0 ( 0.2 g cm-3 at the hot-spot. This is lower than usual because of the high amount of anthropogenic organic matter in the waste layer (25 ( 1% loss on ignition) and natural organic matter in unaffected horizons (17 ( 2% loss on ignition, but only
