Contaminant variability in a sedimentation area of the river Rhine = Variabiliteit van verontreinigingen in een sedimentatiegebied van de Rijn

Aquatic sediments in sedimentation zones of major rivers are in general sinks for pollutants. The sedimentation zone Ketelmeer/IJsselmeer is an important sink for contaminants of the river Rhine (i.e. river IJssel). Recent and historical pollution interact here. Redistribution of suspended solids and erosion of deposited sediment in the shallow Dutch lakes (due to wave action) are likely to change contamination levels of sediments in these lakes, which is the subject of this thesis. The aim of this research was to study and explain the variability of contaminants in the sedimentation area Ketelmeer/IJsselmeer in order to predict the fate of the contaminants in the future. For this purpose a number of methodologies and models were developed and/or adapted.<em>Chapter</em> 2 describes the collection and analysis of sediment cores, top-layer sediments and geologically different layers in Lake Ketelmeer. Sediment cores were sectioned into thin slices and the year of deposition of each layer was determined using radio-chemical analyses. The contaminant concentrations were plotted versus the year of deposition of each sediment layer to (re-)construct the history of contamination. Similar vertical changes in contaminant concentrations were found as in a number of sediment cores sampled in sandpits in Lake Ketelmeer. Further, differences in concentration between the top-layer sediments and the degree of contamination in the entire recent IJsselmeer deposits (IJm-deposits) of Lake Ketelmeer were found. The older Zuiderzee deposits (Zu-deposits) underlying the IJm-deposits have low background values for heavy metals, PAHs and PCBs. This indicates that downward transport of these contaminants with infiltrating water is negligible in this lake. The concentrations of metals and PAHs in the sediment cores reflect, without any serious alterations, the historical input of the past five decades. The pollution history is characterized by, in the early 1940s, low concentrations of metals and already elevated levels of PAHs; a possible reduction of these contaminants during the Second World War and attaining, their highest levels between 1955 and 1970. Rather low levels occur in recently deposited sediments, some of which are the lowest ever observed over the last five decades (Pb, As, and all studied PAHs). Almost all chlorinated compounds showed a certain decline in concentration in anaerobic sediments as compared to samples of the top-layer collected in 1972 and stored in the laboratory, which still reflect the original pollution input. For several PCBs this decline proved to be significant; it may have been caused by microbial dechlorination reactions in the anaerobic sediment. Consequently, the concentration profiles of the chlorinated compounds do not reflect the original pollution history directly. Despite the attenuation of concentration, peaks in PCB concentration profiles were still observed. The following trends in concentrations of PCBs can be currently observed in Lake Ketelmeer sediment:- Almost all PCBs studied had rather low concentrations in the early 1940s.- The highest levels of PCBs occurred between 1960 and 1975.- Recently deposited sediments also have elevated levels of PCBs as compared to the levels in layers from the early 1940s.Overall, recently deposited material is far less-polluted than sediment deposited in the 1960s and 1970s. These findings prove that in this lake older, highly polluted sediments are buried under a younger, less- polluted layer. However, at some locations, such as in dredged parts or erosive zones, the highly polluted layers may remain uncovered, so aquatic organisms may still be exposed to highly polluted sediments from the 1960s and 1970s through the benthic food chain.In <em>Chapter 3</em> attention is focused on the distribution and geochronology of the sediments of Lakes Ketelmeer and IJsselmeer The concentrations of metals, PCBs, PAHs and various sediment characteristics were determined in 77 samples of the surface sediments and one 3 in core of both lakes. Absolute concentrations of these pollutants were normalized for sediment composition (e.g. clay fraction and organic matter contents). In Lake IJsselmeer the youngest geological layer (IJm-deposit) is mainly found in deep sedimentation areas (25%). This deposit is severely polluted in Lake Ketelmeer (Chapter 2). Concentrations of all polluting compounds in the IJm-deposit of Lake Ketelmeer proved to be 1.6 - 9 times higher than in Lake IJsselmeer Concentrations in the same deposit in Lake IJsselmeer were 2 - 4 times higher than those in the older sandy sediments of the lake. Concentrations of heavy metals, As and PCBs initially increase with depth, but then decrease to lower or even background levels. This corresponds with the inputs of the river IJssel). during the past five decades. As the distance from the river mouth (i.e. Lake Ketelmeer) to Lake IJsselmeer increases, there is a decrease in the degree of pollution in this (IJm-deposit) The hypothesis is developed that primary production (with related calcite formation) and mixing with eroded sediment from elswhere in Lake IJsselmeer are together responsible for this dilution.<em>Chapter 4</em> describes the core sampling and analysis for two similar sedimentation zones of two major river deltas. Uniformly soft anoxic sediments in the Volga and Danube deltas were collected, using satellite images, which reflect the concentration of suspended solids. Cesium-isotope dating and measurement of the concentration profiles of heavy metals and PAHs, which reflect (without serious alterations) the historic pollution input into these rivers, were used in the comparison. The contents of the 7 PCBs investigated and of cadmium were below the detection limits for all sediment samples in the Volga and Danube deltas. Low, more or less constant concentrations of arsenic, copper, zinc and all studied PAHs were observed in sediments of the last five decades in the Volga river. Nickel concentrations in Volga delta sediments were rather high, and recently deposited sediments seemed to show slightly increasing levels for zinc, chromium and arsenic. The pollution history of the Danube is characterized by low concentrations of metals but elevated PAH levels in the early 1940s; increasing levels of metals and PAHs between 1950 and 1987; and decreasing levels in more recently deposited sediments. When comparing the concentrations of heavy metals, PAHs and PCBs in the aquatic sediments of the rivers Rhine, Danube and Volga deltas for the past five decades it is evident that the Volga delta was, and still is, the cleanest of the three. A combination of natural (background) inputs, industrial inputs and man made technical changes in the river systems (like the building of storage lakes) can explain most differences in the historical contaminant profiles of the three deltas. Nowadays the concentrations of heavy metals (except copper and nickel), PAHs and PCBs in sediments of the river Rhine are still higher than in the other two rivers, but the sediment loading rate for heavy metals (except cadmium and zinc) of the Danube is higher than for the other two rivers.In <em>Chapter 5</em> the geostatistical sampling approach chosen for Lake Ketelmeer is explained. When monitoring contaminants and related sediment characteristics in an aquatic environment, their spatial variability needs to be taken into account. The sampling strategy covered short-distance variability (65 m) and long-distance variability (500 m) of the investigated variables. In Lake Ketelmeer we chose three sub-areas. The distances between sampling points takes into account the size of each sub-area. With this approach the number of sampling points needed to monitor trends of contaminants in sediments can be minimized, taking into account the necessary accuracy. The choice of sampling strategy for monitoring sub-areas, characterised by either water depth, sedimentation/erosion behaviour or sediment type, will result in different sampling spacings. For example, in Lake Ketelmeer the optimal sampling distance for monitoring Benzo(A)pyrene (BAP) in the central part of the lake was larger than near the harbour and shore, where gradiënts in water depth are steeper. Thus, when designing a dredging programme to remove seriously contaminated sediments, the identification of sub-areas is essential to ensure the adequate dredging of the sediments. If spatial variability is not taken into account for dredging contaminated layers, seriously contaminated spots may be overlooked or rather clean sediments may be dredged needless. Thorough (although expensive) spatial investigations of the contaminated layer before dredging starts, identifying critical sub-areas, is therefore recommended. Practical, cost-effective, geostatistical methods allow an efficient use of limited financial resources for monitoring aquatic sediments.Another important process affecting sediment concentration profiles is consolidation. <em>Chapter 6</em> deals with this physical process of settling of suspended solids and the loss of water after deposition. Consolidation in principle can be described by mathematical models, but because of local circumstances in the Lake IJsselmeer area an empirical approach seemed more reliable. Five representative cores of the (IJm- deposit) were taken from deep zones of the lakes. Periodic water depth surveys at these locations over the last sixty years provided information on the net sedimentation rate and total thickness of the (IJm-deposit) at known time intervals. To calculate a time-equivalent of the depth scale, correction factors for sediment consolidation were needed. These factors were based on a simplification of the various stages of compression (i.e. 0%, 30% and 45%). A factor <em>n</em> , which represents changes of water content of the sediment as a dependent variable of clay content, was derived for each layer, making it possible to determine the initial, uncompressed thickness of each layer by an inverse calculation procedure. Hence, a fairly reliable time-scale for depth could be reconstructed. This time-scale was compared with radio-isotope-dated layers and the results showed close consistency.Annual variability of contaminants in the IJsselmeer area is described in <em>Chapter 7.</em> Measurements of the concentrations of six heavy metals in suspended solids, discharged by the river IJssel). and settling solids at two locations in Lake IJsselmeer showed a typical spatial gradient. The heavy metals concentrations decreased with increasing distance from the river IJssel). inlet. This spatial gradient corresponds with gradients observed in the bottom sediment (Chapter 3). Measurements in sediment cores from Lake Ketelmeer, i.e. the river mouth, and the central part of Lake IJsselmeer showed that the heavy metals concentration in sediments, deposited during the same periods, is 2 to 3 times higher in Lake Ketelmeer than in Lake IJsselmeer The concentration gradient in the settling solids is still significant when changes in the clay and organic matter content are accounted for by using normalized metal concentrations. A rough sediment mass balance for heavy metals, based on river input data and observed sedimentation fluxes, indicates that the total internal sedimentation fluxes of heavy metals in Lake IJsselmeer far exceed the external aereal loading by the river. Due to the complexity of the relationships between the measured variables, the heavy metal concentrations and variables related to primary production and erosion, single correlation analysis did not reveal clear relations and processes that could explain this dilution. Principal component analysis and stepwise multiple regression however showed that the variation in the heavy metals concentration in settling solids is related to wind velocity and clay content, both of which are related to resuspension/erosion of sediments; or alternatively, to pH, chlorophyll and CaCO _{<font size="-2">3</font>} , which are in turn related to algal growth in the lake. Resuspension/erosion-related variables are the dominant factors explaining the variation in heavy metals concentration in the southern part of Lake IJsselmeer whereas algal-growth-related variables explain most of the variation in the metal concentrations in settling solids in the central part of the lake. In this central part, where algal concentrations are high, the negative relation between the concentration of most of the heavy metals in settling solids and the chlorophyll and organic matter concentration in the water compartment justifies the conclusion that dilution of contaminated suspended solids by primary production is active there. In the southern part of the lake, the heavy metals concentration is positively related to wind velocity and clay content. This indicates that resuspension of recent deposits contributes to the metal concentrations of the water compartment. The gradual increase in the δ <sup><font size="-2">13</font></SUP>C value with distance from the river IJssel). mouth (passing from the southern to the central part of Lake IJsselmeer shows that the freshwater sediments are mixed with eroded Zu deposits from within the lake, which were originally saline. It can be concluded, therefore, that the decrease in the heavy metals concentrations in suspended and bottom sediments in Lake IJsselmeer is the result of dilution of contaminated sediments due to erosion of old sediments and of primary production related to algal growth.In <em>Chapter</em> 8 the specific knowledge reported in all previous chapters on sediment contamination, dilution, transport and processes like consolidation and primary production occurring in this area, is integrated in a modeling framework. The sediment distribution processes and the related contaminant dilution of previous decades has been reconstructed. This exercise was the basis for prediction of the future sediment composition and contamination resulting from the expected changes in the contaminant load of the river IJssel). The adapted model STRESS-2d gives a reconstruction of resuspension, erosion, sedimentation and horizontal sediment transport processes, resulting in a good simulation of total-suspended-solids concentration and of the sediment fluxes. This model was nested into the more aggregated model DIASPORA, which simulates changes in morphology (net sedimentation fluxes) quite well. It also accounted for the effects of biogenic calcite production and the consolidation of sediment layers. The decrease of the lead concentration in sedimentation areas in Lake IJsselmeer reflects the reduced suspended input from the river IJssel). but this is dominated by the effects of internal redistribution of sediments in the area. The modelling results support the above mentioned conclusion of Chapter 7. Simulation of sediment fluxes and future contamination by lead are simulated for a constant sediment input and a 50% reduction of lead input from the river IJssel). Modelling results show a steady decrease of the lead concentration in sedimentation zones of Lake IJsselmeer and a purification effect in these zones due to the dredging of sediments in Lake Ketelmeer.