Discussion

The data presented here represent only a preliminary look at the data set generated during our study of the dynamics of the Chesapeake Bay ETM, but they suggest a scenario in which dynamically variable flocculation and disaggregation are integral aspects of particle dynamics, particle trapping, and ecosystem function in this region. Based on this data and on additional analysis of the larger data set, we have formulated a conceptual model of ETM particle dynamics, presented in Figure 10.12. This model serves both to organize the data presented here and as a hypothesis to be tested through future data collection and analysis.

According to this model, an assemblage of inorganic and organic particles enters the estuary as smaller flocs and primary particles with the inflowing freshwater, approaches the ETM, and is swept up and over the leading edge of the salt front. Salinity increases due to mixing, enhancing particle stickiness and aggregation due to electrochemical flocculation, phytoplankton cell lysis, sloppy feeding by zooplankton, and bacterial production. In surface waters, floc size is limited by sinking losses and turbulent disaggregation. Large flocs that sink into the low energy pycnocline

FIGURE 10.12 Conceptual model of particle dynamics in the Chesapeake Bay ETM. See text for discussion. Wide arrows indicate residual current velocities, averaged over tidal cycles. Narrow arrows indicate direction of particle transport. Key: relatively unflocculated inorganic (p) and organic particles ( ETM (gray dome-shaped area); salt front (black contour lines); particle stickiness (shading); flocculated particles robust resuspended aggregates

FIGURE 10.12 Conceptual model of particle dynamics in the Chesapeake Bay ETM. See text for discussion. Wide arrows indicate residual current velocities, averaged over tidal cycles. Narrow arrows indicate direction of particle transport. Key: relatively unflocculated inorganic (p) and organic particles ( ETM (gray dome-shaped area); salt front (black contour lines); particle stickiness (shading); flocculated particles robust resuspended aggregates

are preserved by the lack of turbulent shear and accumulate due to reduced settling speeds as they encounter denser water; they may also continue to grow through floc-culation due to differential settling. Flocs that sink below the pycnocline are partially disaggregated and transported toward the ETM convergence by the estuarine circulation. Within the ETM convergence, more robust resuspended flocs/aggregates interact with fresher particles, carrying some of them to the bottom. Zooplankton also feed on organic material and bacteria in the flocs and produce relatively robust fecal pellets with high settling speeds. An unknown number of relatively fine particles is recycled from the ETM zone into surface or pycnocline waters. Although some finer particles wash through the ETM under high flow or unfavorable floccu-lation conditions, most flocculated particles are trapped. In the Chesapeake ETM, these trapped particles are eventually buried due to high sedimentation rates, though they may be removed later by dredging. The schematic in Figure 10.12 is presented in the context of an ETM at the limit of salt intrusion (Figure 10.5), but it is also applicable with small modification to an ETM at a local axial convergence (Figure 10.6). The modification is primarily that changes in particle stickiness might not play as strong a role, but changes in turbulence, particle size, and particle concentration associated with the frontal zone should be just as important as at the limit of salt.

There are several differences between this model and the traditional idea that flocculation occurs once through an increase in electrochemical attraction when freshwater particles first encounter salt. One-time flocculation cannot explain the existence of the observed large flocs in the pycnocline, with finer particles both above and below, nor can it explain the concentration of both large and small particles near bottom in the center of the ETM. These phenomena imply active exchange between large and small floc populations, at least at certain times and places. The precise mechanisms that control flocculation and disaggregation in the Chesapeake Bay ETM cannot be distinguished at this time, but it is quite likely that they vary in time and space. It is also quite likely that an equilibrium flocculation state is seldom reached in this dynamic environment, especially near the bottom within the ETM. This is because the time scales of flocculation and disaggregation are most likely equivalent to or longer than the time scales of tidal variability in near-bottom turbulence, particle concentration, and large aggregate resuspension/deposition.30,61 Finally, traditional thinking does not adequately allow for the influences of biogenic stickiness and fecal pellet production, which are likely to be significant factors in ETM particle dynamics. For example, Sanford et al.6 reported a seasonal variability in settling velocity that was repeated, albeit without exactly the same timing, in the settling tube data collected in 2001 and 2002. This is in spite of the lack of seasonal variability in primary particle size (Table 10.1).

Fractal scaling of the relationship between floc size and settling velocity (Equation 10.1) has been applied to particles from many different aquatic environments in recent years,26,34,54,63,64 but this was the first attempt that the authors have made to apply it to Chesapeake Bay particles. The novelty of the analysis for us was because of technology limitation; development of the VISTA eliminated that barrier, and we hope to pursue improvements to the video technique and accompanying analyses in the future. While the fractal representations presented here only accounted for approximately half of the observed variability, fractal scaling nevertheless presents itself as a powerful tool for summarizing the characteristics of estuarine particles in two parameters, facilitating intercomparison and modeling. For example, the agreement between the Ems fractal relationship quoted by Winterwerp61 and the fractal relationship of our surface water samples shown in Figure 10.9 suggests a similarity that is currently leading us to apply Winterwerp's flocculation model and parameteriz-ations to the Chesapeake Bay ETM. Another candidate model is that of Jackson,27,65 who also bases his formulation explicitly on fractal scaling. Thus, it behooves estuar-ine particle researchers to identify the fractal scales that describe their systems, along with sources of variability in those scales. Dyer and Manning,34 for example, have characterized the influences of different mechanisms of floc formation and destruction on fractal properties of the resulting particle populations. This approach may help to diagnose sources and mechanisms of flocculation in other environments through comparison of fractal scales.

In this chapter, we have shown that both the fractal scales and the dynamic variability of near bottom, resuspended particles in the Chesapeake ETM separate them from particles further up in the water column. In particular, it appears that direct resuspension of relatively robust large flocs is the rule rather than the exception, but that a combination of simple settling of these same flocs along with interactions with other particles controls deposition. This same phenomena of direct resuspension of large flocs has also been observed on the continental margin by Thomsen and Gust66 and in lower Chesapeake Bay by Fugate and Friedrichs,36 both observations from low to moderate energy environments with active benthos. It is in direct opposition to the expectation that high levels of shear stress should erode primary particles only, and/or that high turbulent shears in the bottom boundary layer should immediately destroy large flocs. In fact, Fugate and Friedrichs found that resuspended floc sizes in the bottom boundary layer of a more energetic and turbid, but less biologically active estuary were governed by the expected inverse relationship between local turbulent shear and floc size. Thus, the presence of directly resuspended large flocs/aggregates may indicate the dominance of biogenic particle packaging. Alternately, it may represent a threshold of turbulent energy below which resuspended flocs survive and above which disaggregation occurs30; this threshold is likely specific to the strength of the flocs in question. The distinction may not be important, since biological domination of the sediment-water interface may be inversely correlated with the level of physical energy and sediment mixing.67

Ultimately, of course, fine sediment transport is much more influenced by settling velocity than it is by particle size, such that knowing particle size alone is of limited utility. New techniques such as the LISST and in situ macro-photography have made it much easier to determine distributions of particle size, but measurement of settling velocity and its variability is still difficult. Readily available techniques such as the Valeport settling tube have been legitimately questioned, while specialized, more accurate video settling techniques are limited in availability and require painstaking analysis. That is why defining relationships between settling speed and particle size and understanding how these relationships respond to environmental variability are so important. It is also why intercomparison and standardization of particle size and settling speed measurement techniques32,33 are critical.

We have presented and compared two methods for estimating particle size, and three methods for estimating settling speed, with generally encouraging results. The VISTA video technique is as close to a standard as we could achieve, though it needs refinement. In particular, we need to increase resolution and further reduce the potential for floc breakup during sampling and the presence of background motion inside the closed tube. The reasonable agreement between the Valeport settling tube and VISTA settling speed estimates is quite encouraging, given the history of problems with other bottom withdrawal settling tube measurements noted above. This may be due to several factors specific to our implementation of the technique, including almost always sampling the pool of tidally resuspended flocs (i.e., break-up resistant) very near the bottom, sampling low enough TSS concentrations that interactions between particles were negligible, and carefully insulating the tube from external temperature fluctuations. This echos the findings of Dyer et al.,32 who concluded that adopting well-controlled protocols for settling tubes improved the consistency of results. An attractive feature of bottom-withdrawal settling tubes in general is that they also allow investigation of the biogeochemical characteristics of particles segregated by settling speed. Finally, we have explored a simple technique using calibrated LISST transmissometer and volume concentration data to estimate the variability of settling speed. The results are encouraging, though slightly slower than other measurements. They should be regarded as tentative but deserving of further study.

0 0

Post a comment