Table of Contents
- Section 1
- Section 2
- Section 3
- Section 4
- Section 5
- Section 6
- Section 7
- List of Figures
- List of Tables
The Effect of Zebra Mussels on Cycling and Potential
Bioavailability of PCBs: Case Study of Saginaw Bay
2.1 Model Development
Conceptual Approach for Multi-Stressor Aquatic Ecosystem Model
The following section presents the development of a model to simulate the cycling and bioaccumulation of PCBs by zebra mussels. As shown in Figure 2.1, the overall approach included coupling of the Saginaw Bay Phytoplankton-Zebra Mussel Model (SAGZM) developed by Limno-Tech, Inc. (LTI 1995, 1997) with a sorbent-hydrophobic organic chemical model similar to the one used by DePinto et al. (1997) and a bioaccumulation model (Thomann and Connolly 1984; Thomann 1989; Thomann et al. 1992; Gobas 1993) to produce a single Nutrient-Phytoplankton-Zebra mussel-PCB Model (SAGZM/PCB). The resulting coupled framework was termed as a Multi-Stressor Aquatic Ecosystem Model. The Multi-class Phytoplankton model, which was used in SAGZM, is a modified version of original models developed by Bierman and Dolan (1981, 1986a, 1986b). The following sections discuss the major building blocks of the overall model formulation (Figure 2.1).
2.1.2 Saginaw Bay Multi-class Phytoplankton Model
Saginaw Bay Multi-class Phytoplankton Model (Saginaw Bay eutrophication model) forms the basis of biotic particle production for the overall model and the link to the nutrient stressor. The model included state variables for phytoplankton biomass in five different functional groups (Figure 2.2): diatoms, greens, non-N2-fixing blue-greens, N2-fixing blue-greens and “others”.
The “others”consists primarily of dinoflaggellates and cryptomonds. There are important physiological differences among these five groups of phytoplankton and a typical successional pattern of their appearance is marked. The growing season begins in spring with diatoms, progresses to greens and “others”, and finally leads to the development of blue-greens in late summer and fall. The nutrients included in the model were phosphorus, nitrogen, and silicon (Figure 2.2). Internal nutrient pool kinetics was used to describe the processes of phytoplankton nutrient uptake and growth. The model also included two different functional types of zooplankton (herbivorous and carnivorous) and implicitly included higher order predation. The details of the model are described by Bierman and Dolan (1981, 1986a, 1986b).
2.1.3 Zebra Mussel Bioenergetics Model
The zebra mussel bioenergetics model, which has been used in SAGZM, synthesizes the best current information in the scientific literature to simulate the following processes: the effect of zebra mussel on phytoplankton, solids and nutrient concentrations; and the effect of ambient phytoplankton, abiotic solids, and herbivorous zooplankton and ambient temperature on zebra mussel filtration and growth. In this model, mussels were classified into three categories according to their age: <1, 1-2, and >2 year old. State variables for this model are the average weight of individuals in each of the cohort class. This zebra mussel bioenergetics model was adapted from the work of Schneider (1992) and much of the parameterization was taken from the study of Walz (1978). The model simulates the growth (in size), and water quality impacts of a pre-specified density of zebra mussel cohort classes over a one year period.
Coupled Phytoplankton Zebra Mussel Model (SAGZM)
The coupled phytoplankton zebra mussel model (SAGZM) was developed by Limno-Tech., Inc. (1995, 1997) to address the water quality issues related to zebra mussel invasion. In this model the zebra mussel bioenergetics model was integrated with the Saginaw Bay eutrophication model by coupling to its phytoplankton (all five groups of biotic solids), abiotic solids (added as a new state variable to the eutrophication model), herbivorous zooplankton and nutrient state variables (Figure 2.2). In addition to their normal transport and transformation processes, it was assumed that all five phytoplankton groups, herbivorous zooplankton, detritus and abiotic solids may be filtered from the water column by zebra mussels. Along with these particulate forms, the zebra mussels also take up the unavailable nutrients associated with them. Luxury nutrients of available nutrients stored in the phytoplankton (an accounting capability unique to a variable stoichiometry model such as this one) are returned immediately to the water column available pool. Previously, this model has been used to estimate the impacts of zebra mussel dynamics on phytoplankton and nutrients in the Bay.
2.1.5 Coupled Phytoplankton Zebra Mussel PCB Mass Balance Model (Multi-Stressor Aquatic Ecosystem Model) - SAGZM/PCB
SAGZM/PCB was build upon SAGZM. New state variables for particle dynamics were added to SAGZM to represent various types of solids that are important for PCB dynamics. SAGZM considers two types of solids in water column; biotic solids (algae in five functional groups and zooplankton) and abiotic solids, whereas only abiotic solids in sediments.
A new state variable, water column detritus solids was added to represent variable organic carbon content in solids. The sources for those solids in the water column were excretion of zooplankton, decay of algae, and resuspension of sediment detritus solids. All biotic, abiotic, and detritus solids settle to the sediments, albeit at different rates. The distinction between different types of solids has important ramifications for toxic contaminant partitioning and fate.
Another state variable, sediment detritus solids was added to represent whole spectrum of solids in sediments. The sources of sediment detritus included feces and pseudofeces of zebra mussels that account for rejection and excreted particulates and settling of water column detritus solids. Figure 2.2 shows the modified-coupled phytoplankton zebra mussel mass balance model with three types of solids in the water column (biotic, abiotic, and detritus) and two types of solids (abiotic and detritus) in the sediments.
The coupling of all the above models is very realistic, as phytoplankton and PCBs are related through organic content, so their behavior is expected to be coupled. Organic carbon, the principal sorbent for HOCs, is cycled largely by biotic processes such as production, grazing, respiration, and decay. This coupling would not only improve model realism and accuracy, but it would also allow the model to simulate the scenario to understand how nutrient control may impact toxic chemicals (such as PCBs) in the ecosystem. In order to simulate these multiple impacts which occur in a real complex aquatic environment, the model integrated the impacts of the following:
- nutrient loading,
- PCB loading,
- zebra mussel activity and physical forcing functions to simulate the response of phytoplankton (production, biomass, and functional group succession),
- zebra mussel (growth, particle processing, and PCB processing and bioaccumulation),
- PCB levels (concentration, phase distribution in water column and sediment).
Simultaneously, the model accounts for hydrology and hydraulic transport.
In model formulation, the traditional modeling approach for HOCs has been adopted i.e. first conduct a mass balance on suspended solids and then to assign a constant organic carbon fraction to these solids in order to compute HOC fractionation through equilibrium partitioning (Thomann and DiToro 1983). An important assumption in this approach has been that the solids are conserved and are subject to settling, resuspension, and deep burial. However, in this modeling context, it was assumed that the sediment detritus solids can undergo kinetic transformation and the repartitioning occurs.
The development of PCB dynamic model consisted of a sorbent dynamic model that explicitly represented seven sorbent state variables. Five of those sorbent state variables are biotic representing live phytoplankton and the remaining two represent the particulate detritus and abiotic solids (Figure 2.3). The division of these solids was on the basis of their fraction of organic carbon, which bind different contaminants to them. All the five groups of phytoplankton are rich in organic carbon, typically containing approximately 40 percent organic carbon by dry weight (Bierman and Dolan 1981). The organic carbon content is the highest in phytoplankton, followed by detritus and abiotic solids.
The basic assumption employed in the model was that the mass of contaminant does not change in the system and that no transformation of contaminant takes place. The contaminant was assumed to partition among water, phytoplankton groups, and solids in the water column. Assuming that the linear partitioning theory is applicable, the PCB fractions associated with each type of solids can be calculated. The total chemical concentration (CT) can be calculated as:
Where CT is the total PCB concentration in the water column, Cd is the apparent dissolved PCB concentration, CL is the PCB associated with phytoplankton each group (L=1 to 5), and Czoopl. is the PCB associated with each group of zooplankton (zoopl.=1 to 2). CAbiotic and CDetritus represent the PCBs sorbed to abiotic and detritus solids, respectively. The fraction of dissolved chemical, fd, can be calculated as:
Where KL is the partition coefficient for PCBs for all the five phytoplankton groups. KAbiotic and KDetritus are the partition coefficients for PCBs sorbed to the abiotic and detritus solids. [L], [Abiotic], and [Detritus] are the concentrations of biotic solids (five phytoplankton functional groups), abiotic and detritus solids, respectively.
The dissolved organic carbon (DOC) was not modeled explicitly. Even though DOC constitute a major fraction of total organic carbon, but studies conducted by Eadie et al. (1990) and Bierman et al. (1992) have shown that its ability to bind HOCs is substantially lower than particulate forms of organic carbon.
the above solids dynamics and PCB partitioning to all types of solids
based on a chemical-specific organic carbon partition coefficient (Koc)
and a solid-specific fraction of organic carbon (foc), the PCB
transport and fate follows the conceptual diagram shown in Figure
The bioaccumulation model predicts average ∑PCB concentration in biota for given ∑PCB concentration in water. It considered two trophic levels: phytoplankton and zooplankton. The model simulates ∑PCB concentration per gm wet weight of individuals, which includes 5 groups of phytoplankton, 2 groups of zooplankton, and 3 cohort classes of zebra mussels.
The bioaccumulation model was
based on a chemical mass balance for each organism in a defined food web. The
general form of the equation that simulates the contaminant body burden of an
organism i of the food web consuming organisms j (j=1 through n), was:
where subscript i refers to a predator, and j refers to its prey
ni = concentration of PCB in organism i [Mchem/Mwet]
kui = uptake rate of PCB from water [L3/ Mwet-t]
CT = concentration of total PCB in the water [Mchem/L3]
fd = dissolved (bioavailable) fraction of chemical
pij = feeding preference factor of organism i for organism j
(Normalized to 1 for all n prey)
nj = concentration of PCB in organism j [Mchem/Mwet]
aij = chemical assimilation efficiency across gut [Mchem absorbed/Mchem ingested]
Cij = food Consumption rate [Mprey /Mpredator-t]
Ki = apparent depuration rate [t-1]
Gi = organism growth rate [t-1]
Equation (3) relates the rate of change of chemical concentration within organism i to the uptake directly from water, the uptake through the food chain transfer, the elimination via respiration and excretion, and dilution due to growth of the organism.
The details of mass balance equations for the detritus solids both in sediments and water column and PCBs in water column, sediment, biota, and zebra mussel are given in Appendix. The mass balance equations on nutrients and biomass of biotic solids can be found in a paper by Bierman and Dolan (1981). Zebra mussel bioenergetics model is described in Limno-Tech., Inc. (1995).