Organic matter (OM) and dissolved inorganic nitrogen (DIN: nitrite, nitrate and ammonium) in the sediments as well as in the water column of two temperate estuaries, the Scheldt Estuary in Belgium and the Netherlands, and the Fraser Estuary in Canada, were investigated. Three representative stations, differing in salinity and representing areas of fast sedimentation, were selected in each estuary. Samples were taken during periods of high and low river discharge. The results show, in both estuaries, that the vertical distributions of OM and DIN in a sediment layer are affected by the instability, caused by episodic resuspension and re-deposition, of the uppermost sediment layer. The findings of this study suggest a hypothesis, next to biogeochemical processes, that the OM and DIN distributions in upper sediment layers are influenced by sedimentary processes in the estuarine environment. The same sedimentary processes even in different estuaries affect OM and DIN distributions in an equivalent way. Correspondingly, the similarity or difference in OM and DIN distribution to a certain extent reflects the sedimentary dynamics. River runoff and sediment resuspension and sedimentation have important impacts on sediment behaviour and thus regulate OM and DIN distributions and shape their vertical profiles in the sediments. As a reflection, the coupling of sediment resuspension followed by redeposition can be deduced from the vertical profile of DIN in the bottom sediments which, in turn, can provide a time-integrated periodic record of the most recent sedimentary history.
Water-sediment interactions are of significant ecological importance in an estuary (e.g., Herman et al., 2001). However, the nature and significance of these interactions with respect to physical forces has barely received attention. As nutrients pass along an estuary, they are subject to biogeochemical processes that are driven by river discharge, tidal exchange, wave activity, and resuspension-sedimentation events. The combination of these processes governs the fate of organic matter and nutrients in an estuary. Estuarine sediments in particular play a significant role in the transformations of nitrogen compounds in coastal systems (Fenchel and Blackburn, 1979; Sloth et al., 1995), and are often viewed as a black box which behaves as a sink for organic material and as a source and a sink for dissolved nutrients (Billen, 1982). Moreover, estuarine sediments are often more important than the water column in the transformations of biologically active elements, and exceed the water column as sites for the processing of nitrogen during organic matter (OM) degradation in estuaries (e.g., Nedwell et al., 1999). Fluxes of nitrate, nitrite and ammonium (dissolved inorganic nitrogen, DIN) across the water-sediment interface are dynamic processes (Billen et al., 1989), and DIN distributions in sediments govern the diffusion of DIN across the water-sediment interface in an estuary. Parallel field investigations were carried out in representative estuarine areas (Table 1) showing fast sediment accumulation (detailed below) to observe the vertical distributions of sedimentary OM and DIN in muddy deposits in two dimensionally very different estuaries, the Scheldt Estuary in Belgium and the Netherlands, and the Fraser Estuary in Canada. The study was designed in hopes that investigations of these parameters would bring up information whether or not they were dependent on estuarine type and geographic location. The objective of this paper is to document the observed OM and DIN distributions from the water-sediment interface down to the uppermost 10 cm in the sediments. Sedimentary OM and DIN dynamics are largely discussed by many authors (e.g., Blackburn and Henriksen, 1983; Jenkins and Kemp, 1984; Nedwell et al., 1999), namely, OM content of bottom sediments are expected to decrease with time; ammonium concentrations increase gradually with depth in the sediments; and, nitrate and nitrite concentrations usually are high just below the water-sediment interface and decline strongly with depth as a result of oxygen depletion. However, in this study the observed OM and DIN vertical profiles in the sediments of both estuaries did not fit conventional or regular sedimentary OM and DIN dynamics, as briefed above, but showed stratifications and truncations. Therefore an alternative hypothesis, considering estuarine physical processes like sedimentation, resuspension and redeposition, that can explain the observations is proposed here, although actual sedimentation-resuspension and redeposition physical processes at the sampling time of this study were not performed.
The Scheldt Estuary
The Scheldt River basin covers an area of nearly 22,000 km2. The 355 km long river rises in the north-east of France, flows through the west of Belgium and the south-west of the Netherlands (Figure 1A). The well-mixed coastal plain estuary extends about 160 km and opens to the North Sea. The river discharge is governed by rainfall, and has an annual average of 100 m3 s−1 and a maximum around 600 m3 s−1 (Belmans, 1991; Baeyens et al., 1998; Chen et al., 2005). The annual sediment load near the Belgian-Dutch border is of the order of 106 tonnes (Wartel, 1977; Terwindt, 1977). The majority of this load is locally resuspended sediment (Chen et al., 2005).
The Fraser Estuary
The Fraser River, with its source in the Rocky Mountains (Canada), is the largest river draining the Pacific margin of Canada. It extends 1,378 km in length and drains an area of 230,000 km2 (Figure 1B). The Fraser Estuary is a stratified estuarine delta with a length of 90 km and extends from the straight of Georgia to Mission. The source of fresh water entering the Fraser Estuary is snowmelt, glacier melt and rainwater. The river discharge varies from 400 m3 s−1 to as much as 15,000 m3 s−1 with an average of 3,500 m3 s−1 (Environment Canada, 1992). The annual sediment load at the head of the estuary is approximately 17 × 106 tonnes (Barrie and Currie, 2000).
The temperate estuaries of the Scheldt and of the Fraser rivers are of a very different type, i.e., estuarine coastal plain versus estuarine delta, well mixed versus stratified, and have divergent water discharge regimes. With these sceneries in view, the study sites are selected in such a way that they show analogous morphological settings, are exposed to similar hydrodynamic processes, and sedimentation from a suspension characterised by fine sediment compositions. In the light of these conditions, three stations, varying in salinity and situated in semi-enclosed areas with one end open to the main channel, were selected in each estuary and covered essentially the complete estuarine reach under the tidal influence: ranging from freshwater above the estuarine turbidity maximum (ETM) zone, through fresh-to-brackish water within the ETM zone, to brackish-to-saline water near the river mouth below the ETM zone (Figures 1A and 1B; Table 1). Samples were collected from four campaigns during 1998–1999 on both the Scheldt and the Fraser at the time around the maximum flood current and at periods of high and low river runoff respectively. At the time of sampling, water discharge reached 200 and 50 m3 s−1 in the Scheldt, and 3,500 and 425 m3 s−1 in the Fraser for high and low runoff respectively. In the Scheldt, samples were taken in the access channels to locks. The sedimentation rate in these areas is very high, reaching up to 10 cm per month, and hence sediments are regularly removed in order to sustain navigation (Taverniers, 1999). In the Fraser, sampling sites were chosen in backwater areas and sloughs. McLaren and Ren (1995) identified fine sediment (mud) accumulation zones in these areas. No exact sedimentation rates are available for the Fraser sampling sites, however, an accumulation rate of 10 cm per year occurs near the mouth (at Sturgeon Bank) of the main channel (Barrie and Currie, 2000).
Sedimentary processes of the semi-enclosed areas
Sediments are supplied to a semi-enclosed basin either by advection and deposition from a slackening water current or as a dense near bottom mud flow (NBMF) set in motion by bottom stress. Sediment supply through advection and deposition to a semi-enclosed basin has been described by Fettweis (1995). There are two conditions: (1) a salinity difference between the basin and the main channel that is large enough to generate a density current; (2) a high turbidity in the main channel. These conditions lead to the formation of fluid mud deposits, and the grain-size spectrum shows an uninterrupted continuous profile. Sediments, however, may also settle from an episodic dense NBMF as described by Lin and Mehta (1989). Sediment layers resulting from this process show interruption and discontinuity with the underlying layer as a result of the erosional effect of the NBMF. In the case of the study sites, NBMF between the channel and semi-enclosed areas have not been reported so far. However, bottom stress sufficiently strong to generate a NBMF can also result from pressure waves generated by ships (Schoellhamer, 1996). Furthermore it was shown by Van Rijn (1993) that wave induced pressure variation is able to fluidise the top layer of a consolidated mud bed and to generate a thin NBMF with sediment concentrations exceeding 100 g l−1. This NBMF consists of dispersed fluid mud moving in between an overlying suspension and an underlying settled bed layer. With a decreasing shear (Lin and Mehta, 1989), the moving mud flow will settle rapidly adding new material on top of the previously settled layer.
The formation of mud deposits is therefore a combined result of layers formed by settling from a suspension (vertical movement) and layers formed from near bottom sediment transport (horizontal movement). When the vertical movement is important, the deposition shows either a uniform or an upward continuous fining of the grain size. However, when the horizontal movement becomes prevalent, the deposition exhibits either an upward coarsening or fining of the grain size depending on the bottom sediment sources. As a consequence, even during a short period of time, as for instance within the same half tide, several layers with different sediment textures can be formed (e.g. Reineck and Wunderlich, 1969). Sediments that are resuspended and newly deposited but not yet compacted are more susceptible to resuspension than undisturbed bottom sediments (Schoellhamer, 1996). As a result, the occurrence and the thickness of re-worked sediment layers differ from one event to the other.
Materials and methods
Undisturbed sediment samples were collected using a lightweight gravity UWITEC-corer with a tube diameter of 10 cm. This corer encloses a 20 to 50 cm sediment column overlain by at least 30 cm of bottom water. A special piston device was used to accurately recover samples from the water-sediment interface and from different sediment depth intervals. There were 2 to 3 duplicate cores taken at each sampling site for different physical-and chemical analyses. The cores taken between high and low runoff conditions were within a sampling area of 2 m2. Samples for chemical analysis were extruded under an oxygen free environment, and under ambient conditions for the other analyses. Sediment was extruded on board, at 1 cm intervals for the topmost 6 cm and at 1 or 2 cm intervals for the lower 6–10 cm layer, and immediately transferred to acid washed polyethylene bottles. Water samples were taken at 2 m below the water surface, at 30 to 100 cm above the bottom, and at the water-sediment interface at each station. All the samples were immediately deep frozen using either liquid nitrogen or dry ice on board and stored at −20°C until analysis.
Radiography and bulk density
An undisturbed sediment sample for radiograph and bulk density measurements was taken at slack tide using a large cylindrical box corer (60 cm diameter, 60 cm high). A radiograph of an undisturbed vertical slice (20 × 20 × 2 cm) sub-sample, taken from the large cylindrical box corer, was acquired using an X-ray radiation of 80 keV, 5 mA on Agfa Structurix film. The bulk density of a small cylindrical sediment core (8 cm in diameter), sampled near the vertical slice from the large cylindrical box corer, was measured by comparing the absorption of gamma rays emitted by a 241Am source to the absorption of these gamma rays in reference samples with known density.
The grain-size analyses were performed using a method described in Wartel et al. (1995). Approximately 10 g lyophilised and homogenised sediment sample was prepared by removing salts, organic matter and carbonates using hydrogen peroxide and hydrochloric acid respectively. A stable suspension was obtained after rinsing and adding 5 ml of a peptising agent. The coarse fraction (75 to 2000 μ m) was separated by wet sieving on a 75 μm sieve, then dried at 105°C, and finally dry sieved. The grain-size distribution of the fine fraction (2–75 μ m) was obtained using the Sedigraph 5100 coupled to a Mastertech 51. The precision for 10 consecutive measurements on aliquots of the same sample was around 1% for every grain-size fraction.
All samples were processed under N2 gas environment and thawed at 0 to 1°C. Sediment samples were submitted to KCl extraction. The optimal wet sediment weight to 1 M KCl volume ratio for making a sediment-slurry was experimentally determined to be 1:10. Slurries of 1 g wet sediment in 10 ml 1 M KCl solution were mixed for 30 min on a reciprocal-shaker, and centrifuged at 3,000 × g for 10 min at 0°C. The supernatant was then filtered through a 0.45 μm filter and collected into acid-washed polyethylene bottles. DIN concentrations of the sediment samples were determined immediately on the filtered supernatant solution (Armstrong et al., 1967; Bremner and Mulvaney, 1982). After thawing, water samples were filtered through 0.45 μm filters, and DIN were then measured using standard colorimetric methods (Armstrong et al., 1967; Wood et al., 1967; Koroleff, 1969; Strickland and Parsons, 1972; Bremner and Mulvaney, 1982). The detection limit was 0.05 μM for DIN. The precision for 6 consecutive measurements on aliquots of the same sample was about 1% for ammonium and was less than 2% for nitrate and nitrite.
Total organic matter content was determined by loss on ignition at 430°C for 24 hrs (Gale and Hoare, 1991). Sediments were dried to a constant weight at 105°C prior to combustion. The combustion of the recent and clay-rich sediments at 430°C for 24 hrs can minimise destruction and weight loss of carbonates and dehydration of clays (Cattol, 1962), and the longer time span is more effective against stronger organic compounds such as cellulose. The precision for 10 consecutive measurements on aliquots of the same sample was about 2% for OM.
Description of the sediments
The grain-size vertical profiles showed that the bottom sediments of the study areas were mainly stratified (Figure 2). Bottom sediments in both estuaries showed cohesiveness with a clay fraction (less than 4 micrometers) of more than 20% of the total grain-size distribution. Bottom sediments in the Scheldt were much finer (mean grain-size of 0.8 ± 0.4 μm with 63 ± 5% of clay) than those in the Fraser (mean grain-size of 7 ± 4 μm with 27 ± 6% of clay). The fine fraction (silt plus clay) exhibited a minor variation (around 5%) with depth in the Scheldt, while in the Fraser, the variation was larger, from 20% at the downstream station to 40% at the upstream station (Figure 2). The difference of fine fraction between low and high runoff was less than 5% in the Scheldt and around 10% in the Fraser (except for 50% difference at the Fraser downstream). It is noticed that grain-size of the bottom sediments in both estuaries showed close values to their respective suspended sediments in the water column. The mean grain-size of the Scheldt suspended sediments (1.7 ± 2.1 μm) was even slightly higher than the mean value of the Scheldt bottom sediments (0.8 ± 0.4 μm). A mean grain-size for the Fraser suspended sediments between Deas Slough (the midstream station) and Sturgeon Bank (the downstream station) of 11 ± 3μm was reported by Stecko and Bendell-Young (2000). This value was close to the mean grain-size of the Fraser bottom sediments (7± 4μm, not including the sandy-mud sediments of the downstream station at high runoff, otherwise it would be 11 ± 4 μm). Hence, the sediment properties suggested that the sedimentary deposits at the study areas were derived almost exclusively from suspensions with a very narrow grain-size range.
Organic matter in bottom sediments
OM content in the sediments was substantially higher in the Scheldt (8% to 14%) than in the Fraser (1% to 5%; Figure 3). The highest OM content in the Scheldt was found at the upstream station during low runoff, and moved to the midstream station during high runoff. The highest OM value in the Fraser was always located at the midstream station, and OM content was lower during high runoff, although the difference with low runoff was not as important as it was in the Scheldt.
Dissolved inorganic nitrogen
DIN distributions in the overlying water and in the sediments varied substantially between and within both estuaries (Tables 2 and 3). DIN concentrations in the water column were up to 740 μM in the Scheldt while in the Fraser never exceeded 12 μM (Table 2). DIN concentrations in the sediments were coherently much higher in the Scheldt (up to 5,100 μ M) compared with those in the Fraser (below 712 μM; Table 3). Ammonium was the dominant form of nitrogen in the sediments of both estuaries. Nitrate predominated in the Scheldt waters, while in the Fraser waters both ammonium and nitrate were the important forms of nitrogen.
Ammonium concentrations in the Scheldt water varied from 20 to 300 μM (Table 2), and were higher at high runoff than at low runoff, except for the downstream station where the values were of the same order of magnitude. In the Fraser, by contrast, ammonium concentrations never exceeded 8 μ M, and were lower during high runoff, except for the downstream station where the value was twice as high. Ammonium concentrations in the bottom sediments ranged from 150 μM to 10,000 μM in the Scheldt, and from 100 μM to 1,300 μM in the Fraser (Figure 4). Since these values were much higher than those in the overlying water, a net upward supply of ammonium can be expected. The gradient of ammonium concentration in the sediments of the Scheldt was weak during high runoff and exhibited relatively small vertical variations at the uppermost 10 cm, whereas, an obvious ammonium increment with depth was observed during low runoff. The ammonium concentration in the sediments of the Fraser increased steeply with depth, and the gradient was relatively comparable regardless high and low runoff periods. The ammonium vertical profiles in the sediments of both estuaries displayed a few noticeable truncations, showing a discontinuity in concentration-evolution with depth.
Nitrate concentrations of the Scheldt and the Fraser waters differed by two orders of magnitude (Table 2). In the sediments, Scheldt nitrate concentration ranged from 25 μM to 460 μM, but were all below 12 μM in the Fraser (Figure 5). In both estuaries, nitrate concentrations were high in the upper 2–5 cm and then decreased pronouncedly below this depth. It is noticed that in the Scheldt the decreased values remained more or less constant (50–85 μM) with depth at all stations. The nitrate concentrations in the sediment were higher at high runoff in both estuaries. It is observed that most of the nitrate vertical profiles in the sediments of both estuaries exhibited fluctuations with more than one peak value, while usually only one high value just below the water-sediment interface at the top of sediment layer is expected (e.g. Revsbech and Jorgensen, 1986).
Nitrite concentrations in water columns were 0.4–31 μM in the Scheldt, whereas in the Fraser they were near or below the detection limit (Table 2). In the sediments, nitrite concentrations had a range of 5–90 μM in the Scheldt, but never exceeded 4 μ M in the Fraser (Figure 6). The nitrite vertical profiles in the sediments of both estuaries exhibited some “anomalies” with a few peaks down to 6–8 cm below the water-sediment interface, while normally one peak below the water-sediment interface at the top of sediment layer is expected (e.g. Revsbech and Jorgensen, 1986).
Stratification of the vertical profiles
Every grain-size vertical profile in this study shows that the sediments are characterised as cohesive, though may be considered macroscopically as unvarying, the existence of stratification appears from the results (Figure 2). Layering is observed as abrupt changes in physical-and chemical properties (Figures 2, 3, 4, 5, 6).
In the case where a sediment layer formed from continuous sedimentation, OM attached to the suspended particles settles to the bottom, partly mineralises and partly contributes to the mean OM content in the bottom sediments. Due to mineralization, the OM content of bottom sediments is expected to decrease with time, and ammonium concentrations to increase steadily with depth in the sediments, which act as a source of ammonia to the overlying water column. Nitrate concentrations usually are high just below the water-sediment interface (top millimetre in the sediments) and decrease very rapidly with depth as a result of oxygen depletion. The sediments are a sink of nitrate for the water column. Usually the surface oxic layer is less than 1mm deep in organically rich estuarine muds, to possibly a few millimetres in sands (e.g. Revsbech et al., 1980; Revsbech and Jorgensen, 1986). These OM, nitrate and ammonium vertical conventional profiles have been very well documented and are widely known (e.g., data and figures in: Blackburn and Henriksen, 1983). However, the results of this study reveal the deviations from the conceptual OM and DIN vertical profiles in the sediments showing up as more than one peak value as well as truncations. There are studies (e.g., Jenkins and Kemp, 1984; Nedwell et al., 1999) reporting some intermediate high concentrations of nitrate varying between 2–6 cm in the estuarine sediments. Yet these intermediate high values have not been well explained or have been tentatively postulated as indications of three possibilities: either nitrate diffusion from overlying water, or infaunal irrigation, or nitrate production in the sediments.
In the case of the fluid mud of the study areas, the exchange of the nutrients between the sediments and the water column due to diffusion is probably of minor importance since the diffusing between layers is strongly inhibited by biological processes and by flocculation (Paterson, 1989; Paterson et al., 1990; Bennett et al., 1999). In this study no infaunal were found after examination of some selected samples from the collected cores. Given the data of a few peaks of nitrate concentrations down to several centimetres in the very anoxic sediments (measured redox potential ranging from −100 to −250 mV), in situ nitrate production is not very likely.
The conceptual hypotheses can not explain the observed stratification of the vertical DIN profiles in this study. The deviations from the conceptual OM and DIN vertical profiles in the sediments point to processes other than biogeochemical variations. An alternative hypothesis that could explain the observations is that a change in hydrodynamic process or in other words a predominant sedimentary process possibly together with biogeochemical ones may affect the vertical distribution of OM and DIN in sediments.
The well-documented “conventional” or “regular” OM and DIN vertical profiles in the sediments are the effect of a sediment layer formed from a continuous sedimentation. Nevertheless, in case a sediment layer forms from a discontinuous sedimentation, which is common in the estuarine system, these profiles can then be different. Although nutrients dynamics in sediments have been studied extensively (e.g. Blackburn and Henriksen, 1983; Jenkins and Kemp, 1984; Henriksen and Kemp, 1988; Boderie et al., 1993; Rysgaard et al., 1999), recent sedimentary history and short-term processes in estuarine systems caused by resuspension-sedimentation cycles have not yet been considered.
In estuaries a bottom sediment layer is always in an ephemeral stage and may either be resuspended into the water column or be consolidated onto the bottom bed depending on the hydrodynamic conditions. It is proposed here that these hydrodynamic conditions prevailing during deposition are reflected in the truncations of the grain-size spectrum. Inversely, the change in grain-size spectrum along a sediment core is an indicator for the changing hydrodynamic conditions during deposition. The data of this study suggest that the formation of the sediment layers resulting from short-term sedimentation events can affect the vertical DIN distributions in the sediments and the exchange of DIN with the overlying water. Three scenarios are discussed in detail hereafter, case 1 demonstrating typical stratification of sediment layers; case 2 and case 3 explaining the integration of the sedimentary process with the DIN distributions.
Case 1: Stratified sediment layers
Sediments in the study areas are cohesive and at the uppermost 10 cm are mostly stratified. Layering can be observed from the grain-size distributions (Figure 2) and the bulk density radiograph of sediment core (Figure 7). In the case of fluid mud, the diffusional exchange of nutrients between the layers, including exchange with the water column, is rare (Paterson, 1989; Paterson et al., 1990; Bennett et al., 1999). The poor exchange between the top sediment layer and the overlying water is illustrated by the radiograph of a bottom sediment core taken at the Scheldt midstream station (Figure 7). The radiograph of the uppermost 10 cm of the fluid mud clearly reveals layering with respect to the distribution and the size of the gas voids. No trace of infaunal irrigation can be observed. The bulk density profile corresponds to the stratification of the gas voids. Discontinuities in the void size can be recognised at 3 cm and at 6 cm below the sediment surface. More in particular, it can be seen that near the water-sediment interface a few millimetres thick mud layer evidently prevents the gas in the underlying sediments from escaping into the water column. This radiograph indicates not only the existence of stratification in the fluid mud but also the fact that there is little gas migration between the successive layers. If the observed layering is a result of resuspension followed by rapid deposition of the fluid mud as explained above, then deviations from the “regular”, which means without resuspension, OM and DIN profiles may occur. Removing the top layer and redepositing a new layer combined with poor exchange between layers may result in a break or truncation in the OM and DIN profiles. Also the time span of re-worked layers can be deduced from the stratification of OM and DIN profiles if the denitrification rate is accurately known.
Case 2: Fast sedimentation and DIN
The formation of a sediment layer, as introduced earlier, is a combined result of alternating settling from sediment transport vertically as well as horizontally. Even during a short period of time, as for instance within the same half tide, several layers with different sediment compositions can be formed. Such a fast deposition can happen due to the tidal activity and regardless of the river runoff. Figure 8 exhibits the vertical profiles of the grain-size, OM and DIN in the sediments at the Scheldt upstream station at low runoff. The grain-size profile showed one obvious discontinuity: δ (at 6 cm), separating sediment layers A (0–6 cm) and B(6–10 cm). The sedimentation pattern of layer A was obviously different from layer B. Layer A consisted of two sub-layers, an upper part (0–2 cm) had an upward coarsening grain size and decreasing OM content, while a lower part (2–6 cm) had an upward fining grain size and increasing OM content. Layer B showed a uniform grain size and OM content decreased with depth. DIN vertical distribution conformed to these sedimentary records. Although ammonium concentrations increased with depth, the gradient was not smooth and a truncation point could be observed at interface δ. The nitrate profile showed a high value in the upper part of layer A and a small peak near the top of layer B right below the interface δ. The nitrite profile exhibited peaks below each nitrate high value indicating that the nitrite was generated by denitrification just below the nitrate maximum. It is worth noting that the vertical profiles of ammonium and nitrate for every layer separately did show a “conventional” shape as described earlier. Sedimentary records of grain size, OM and DIN illustrate that the uppermost 10 cm of the bottom sediments were deposited in a very short time, at least short enough to “preserve” some nitrate (60 μM) at 6–7 cm below the actual water-sediment interface in spite of the prevailing strong anoxic conditions in the Scheldt. There are no data on denitrification rate available for the Scheldt estuarine sediments. S. Van Damme (Department of Biology, University of Antwerp, Belgium, pers. comm.) found that denitrification rate increased with decreasing salinity, and that a denitrification rate of 120 μmol m−2 h−1 (high tide) occurred on tidal flats near the Durme mouth (20 km seaward of the upstream station, Figure 1A). Besides, Jenkins and Kemp (1984) and Rysgaard et al. (1999) reported denitrification rates over 200 μmol m−2 h−1 in sediments of the Patuxent Estuary (USA) and of the Randers Fjord (Denmark) respectively. Also, Oremland et al. (1984) reported denitrification rates for shaken estuarine sediment slurries ranging from 17 to 280 μmol m−2 h−1 (in the absence of added nitrate and for salinities of 8–25). Considering the Scheldt upstream freshwater station at low runoff and under strong anoxic conditions, muddy sediments with denitrification rates of 130-280 μmol m−2 h−1 based on the literature data (cf. above), and applying measured porosity of 0.8, it can be calculated that it will take 7 to 15 hours to reduce a nitrate concentration of 300 μM at the water-sediment interface to a concentration of 60 μM at 6–7 cm in the sediments. The discontinuity between sediment layers A and B suggests that the sedimentation was interrupted, and that a former water-sediment interface occurred at δ in a time span possibly from half to one tidal cycle (a time span of 7 to 15 hours), or in other words, that the actual sedimentation at this sampling site was of the order of 6 to 7 cm per tidal cycle.
Case 3: Coupling of resuspension-sedimentation and DIN
Analyses of the sedimentary record at the Scheldt midstream station at high runoff are shown in Figure 9. Although the grain size and OM content were uniform, DIN profiles revealed two discontinuities: δ1 (at 2 cm) and δ2 (at 4 cm), separating sediment layers: A (0–2 cm), B(2–4 cm) and C (4–10 cm). Nitrate concentrations were highest in layer B. The high nitrate concentration found at 4 cm below the water-sediment interface can be explained by the sedimentation history. The constant grain-size profile indicates that these three layers were deposited rapidly from a uniform suspension. The nitrate concentration of layer B, almost 400 μ M, was only 10% less than the nitrate concentration of 440 μM occurring at the water-sediment interface. Shortly after deposition, and before the nitrate in layer B could be reduced to nitrite, layer A settled on top of layer B. The low nitrate in layer A (around 230 μM, or about half the value at the water-sediment interface) suggests its deposition from a near bottom mud flow (NBMF) derived from the mixing of locally resuspended settled mud low in nitrate and overlying water high in nitrate. The fast succession of two sedimentary events, sedimentation from a uniform suspension followed shortly by deposition of locally resuspended sediments, also explains the unvaried OM profile and the uniform ammonium concentrations in layers A and B, because not enough time was available for the decomposition of OM. In layer C, a “conventional” DIN profile shaped out: a slightly high nitrite concentration (about 70 μM) right below the interface δ2 (at 6 cm) and ammonium concentrations increasing with depth. As explained in case 2, this “conventional” DIN profile suggests that δ2 was a former water-sediment interface. Furthermore, the relatively low ammonium concentration near interface δ2 and the higher values of nitrate and nitrite in layer B indicate that there was only a short time interval between the formation of layers B and C. At least its duration was too short to mineralise a measurable amount of OM to ammonium. The steady ammonium depth profile suggests a sedimentation rate which exceeds the mineralisation of OM. The discontinuities in DIN profiles indicate that sedimentation alternated with the truncation of freshly deposited layers due to local resuspension. Besides, the weak gradients for the ammonium profiles at all three Scheldt stations at high runoff (Figure 4) indicate that resuspension was more significant during that period.
The occurrence of short-term episodic sedimentary events like resuspension-redeposition cycles was also observed in the Fraser Estuary. Vertical stratification of sediment layer, OM and DIN were shown in all Fraser profiles. At low runoff, discontinuities occurred in the sedimentary record of all three stations. The differences in grain-size among the layers were more distinguished than in the Scheldt. At high runoff, the stratification in the sedimentary records is even better recognised (Figure 2). For instance, at all three stations, peaks of nitrate concentrations appeared at several centimetres below the water-sediment interface (Figure 5), and next to each horizon with a nitrate peak nitrite high values followed (Figure 6) indicating the ongoing denitrification processes at each individual layer. The DIN stratification in the Fraser at high runoff reflects very dynamic and fast sedimentation processes as for instance an example shown in the Figure 10.
Episodic short-term erosion-resuspension-deposition cycles are among the most important estuarine processes, especially where navigation and engineering activities are intensive. A uniform sediment deposit does not always resonate a steady sedimentation under calm conditions. Neither does a fast resuspension-redeposition cycle necessarily show up as an alternation of layers resulting from variations in grain-size or OM content. The findings of the present study suggest a possible important relationship between these short-term sedimentary processes on the one hand and the vertical OM and DIN patterns in the sediments on the other. Sediment resuspension followed by fast sedimentation leads to a redistribution of DIN in the sediments, so that episodic resuspension-sedimentation events may contribute in shaping the vertical DIN-profile in the sediments, and if so, the DIN profile can inform on the short-term sedimentary events. Furthermore, though it can be argued, next to biogeochemical processes, the OM and DIN vertical distribution patterns are also governed by the sedimentary dynamics in the estuary. The same sedimentary processes even in different estuaries affect OM and DIN distributions in an equivalent way. The results indicate that physical factors, influencing the time scale of sediment recycling, may predominate over benthic biogeochemical factors in determining DIN behaviour. Depending on its strength, the available short-term physical forces can, at times, possibly “over-ride” or dominate biological-chemical processes. The potential implications of the findings of this study are relevant to other estuarine research, for instance, to the study of ecological communities especially those benthic habitats which abundance and species composition are susceptible to stresses that are induced by physical processes and nutrient dynamics across the water-sediment interface. The hypothesis developed in this paper corresponds closely to the important features of most estuaries. Variations from this hypothesis can be expected because of local factors, yet may not differ in nature, but differ in degree.
We thank the captains and crews of the Belgian research vessels “VEREMANS” and “SCALDIS” for their help during both cruises on the Scheldt Estuary. Sincere thanks also go to the captains and crews of Canadian Coast Guard Hovercrafts “SRN6 045” and “SIYAY” for their friendly and cooperative help during two sampling cruises on the Fraser Estuary. We gratefully thank T. Pedersen, B. Elner, V. Barrie, P.J. Harrison, B. Mueller, M. Church, H. Schreier and K. Hall for their support and providing a pleasant and stimulating scientific climate during our work in Canada. We thank C. A. Dyck, M. Hilmer and K. Fergesen who assisted in the lab work at the University of British Columbia, Canada.