Assessing a Ten-Fold Increase in the Chesapeake Bay
Native Oyster Population
A Report to the EPA Chesapeake Bay Program
July 2005
Carl F. Cerco and Mark R. Noel
US Army Engineer Research and Development Center, Vicksburg MS
-------
Abstract
The Chesapeake Bay Environmental Model Package (CBEMP) was used to assess the
environmental benefits of a ten-fold increase in native oysters in Chesapeake Bay. The CBEMP
consists of a coupled system of models including a three-dimensional hydrodynamic model, a
three-dimensional eutrophication model, and a sediment diagenesis model. The existing CBEMP
benthos submodel was modified to specifically represent the Virginia oyster, Crassostrea
virginica. The ten-fold oyster restoration is computed to increase summer-average, bottom,
dissolved oxygen in the deep waters of the bay (depth > 12.9 m) by 0.25 g m3. Summer-average
system-wide surface chlorophyll declines by 1 mg m3. Filtration of phytoplankton from the
water column produces net removal of 30,000 kg d"1 nitrogen through sediment denitrification
and sediment retention. A significant benefit of oyster restoration is enhancement of submerged
aquatic vegetation. Calculated summer-average biomass improves by 25% for a ten-fold increase
in oyster biomass. Oyster restoration is most beneficial in shallow regions with limited exchange
rather than in regions of great depth, large volume and spatial extent.
Point of Contact
Carl F. Cerco, PhD, PE
Research Hydrologist
Mail Stop EP-W
US Army ERDC
3909 Halls Ferry Road
Vicksburg MS 39180 USA
601-634-4207 (voice)
601-634-3129 (fax)
cercoc@wes.army.mil
-------
1 Introduction
More than twenty years ago, grazing by benthos was implicated as a
controlling process on phytoplankton concentration in tidal waters (Cloern 1982,
Cohen et al. 1984). Officer et al. (1982) identified criteria for regimes in which
benthic control is possible. They were:
1. Shallow water depths in the range of 2 to 10 m;
2. A large and widespread benthic filter feeding population;
3. Partially-enclosed regions of substantial size with poor
hydrodynamic exchange;
4. Adequate nutrient supplies; and
5. Regions that show relatively constant and low phytoplankton
levels.
A link between decimation of the oyster population and deteriorating
water quality in Chesapeake Bay was proposed by Newell (1988). Newell
calculated the 19th century oyster population could filter the entire volume of the
bay in less than a week and suggested an increase in the oyster population could
significantly improve water quality by removing large quantities of particulate
carbon. Gerritsen et al. (1994) largely countered Newell's suggestion. They
noted that benthic filter feeders can be dominant consumers in shallow portions
of the bay but are suppressed in deeper portions. Processes leading to
suppression include hydrodynamic limits and hypoxia. Gerritsen et al. concluded
that use of filter-feeding bivalves to improve water quality in large estuaries is
limited by the depth and width of the estuary.
Recent research on the role of oysters in Chesapeake Bay has focused on
processes by which oysters influence their immediate environment rather than on
system-wide effects. Newell et al. (2002) provided experimental evidence that
denitrification of nitrogen in oyster feces may enhance nitrogen removal in
estuaries. They examined the effect of light on algal biomass and nutrient fluxes
at the sediment-water interface and suggested that clarification of the water
column by filter feeders may provoke a shift to an ecosystem dominated by
benthic primary production. Porter et al. (2004) placed oysters in experimental
mesocosms. Their work largely supported the suggestions by Newell et al.
(2002). The found that oysters shifted processes to the sediment by decreasing
phytoplankton biomass and increasing light penetration to the bottom. Increased
Chapter 1 Introduction
1
-------
light penetration stimulated mic rophytobenthos, which diminished nutrient
regeneration from the sediments. They found, however, that high bottom shear
stress eroded the microphytobenthos and cautioned that, under high bottom shear
conditions, nutrient regeneration from the sediments may increase. Most
recently, Newell and Koch (2004) employed a model to examine the interactions
between oysters, turbidity, and seagrass density. They predicted that restoration
of oysters has the potential to reduce turbidity in shallow estuaries and facilitate
efforts to restore seagrasses.
Our own interest in oysters stems from the "Chesapeake 2000"
agreement. The agreement, signed by the executives of the Commonwealth of
Pennsylvania, the State of Maryland, the Commonwealth of Virginia, the District
of Columbia, the US Environmental Protection Agency, and the Chesapeake Bay
Commission, rededicates the individuals and entities to the "restoration and
protection of the ecological integrity, productivity, and beneficial uses of the
Chesapeake Bay system." The agreement sets specific goals including:
Restore, enhance and protect the finfish, shellfish and other
living resources, their habitats and ecological relationships to
sustain all fisheries and provide for a balanced ecosystem.
The agreement lists methods to achieve this goal including:
By 2010, achieve, at a minimum, a tenfold increase in native
oysters in the Chesapeake Bay, based on a 1994 baseline.
and
By 2004, assess the effects of different population levels of filter
feeders such as menhaden, oysters and clams on Bay water
quality and habitat.
The environmental effects of a ten-fold increase in population of native
oysters were assessed by incorporating oysters into the Chesapeake Bay
Environmental Model Package (CBEMP), a comprehensive mathematical model
of physical and eutrophication processes in the bay and its tidal tributaries. This
report is the primary documentation for the assessment.
The Chesapeake Bay Environmental Model Package
Three models are at the heart of the CBEMP. Distributed flows and
loads from the watershed are computed with a highly-modified version of the
HSPF model (Bicknell et al. 1996). These flows are input to the CH3D-WES
hydrodynamic model (Johnson et al. 1993) that computes three-dimensional
intra-tidal transport. Computed loads and transport are input to the CE-QUAL-
ICM eutrophication model (Cerco and Cole 1993) which computes algal
biomass, nutrient cycling, and dissolved oxygen, as well as numerous additional
constituents and processes. The eutrophication model incorporates a predictive
sediment diagenesis component (DiToro and Fitzpatrick 1993).
Chapter 1 Introduction
2
-------
The first coupling of these models simulated the period 1984-1986.
Emphasis in the model application was on examination of bottom-water anoxia.
Circa 1992, management emphasis shifted from dissolved oxygen, a living-
resource indicator, to living resources themselves. In response, the
computational grid was refined to emphasize resource-rich areas (Wang and
Johnson 2000) and living resources including benthos (Meyers et al. 2000),
zooplankton (Cerco and Meyers 2000), and submerged aquatic vegetation (Cerco
and Moore 2001) were added to the model. The simulation period was extended
from 1985 to 1994.
Model improvements to address the issues raised by the Chesapeake
2000 Agreement started soon after the agreement was signed. The computational
grid was further refined and plans were made to incorporate new living resources
into the model. At the same time, regulatory forces were shaping the direction of
management efforts. Regulatory agencies in Maryland listed the state's portion
of Chesapeake Bay as "impaired." The US Environmental Protection Agency
added bay waters within Virginia to the impaired list. Impairments in the bay
were defined as low dissolved oxygen, excessive chlorophyll concentration, and
diminished water clarity. Management emphasis shifted from living resources
back to living-resource indicators: dissolved oxygen, chlorophyll, and clarity. A
model recalibration was undertaken, with emphasis on improved accuracy in the
computation of the three key indicators.
A revision of the CBEMP was delivered in 2002 (Cerco and Noel 2004)
and used in development of the most recent nutrient and solids load allocations in
the bay. This version of the model is used to examine the impact of the tenfold
increase in native oysters. The 2002 CBEMP employs nutrient and solids loads
from Phase 4.3 of the watershed model (Linker et al. 2000). (Documentation
may be found on the Chesapeake Bay Program web site
http://www.chesapeakebay.net/modsc.htm.) Nutrient and solids loads are
computed on a daily basis for 94 sub-watersheds of the 166,000 km2 Chesapeake
Bay watershed and are routed to individual model cells based on local watershed
characteristics and on drainage area contributing to the cell. The hydrodynamic
and eutrophication models operate on a grid of 13,000 cells. The grid contains
2,900 surface cells (~4 km2) and employs non-orthogonal curvilinear coordinates
in the horizontal plane. Z coordinates are used in the vertical direction, which is
up to 19 layers deep. Depth of the surface cells is 2.1 m at mean tide and varies
as a function of tide, wind, and other forcing functions. Depth of sub-surface
cells is fixed at 1.5 m. A band of littoral cells, 2.1 m deep at mean tide, adjoins
the shoreline throughout most of the system. Ten years, 1985-1994, are
simulated continuously using time steps of ~5 minutes (hydrodynamic model)
and -15 minutes (eutrophication model).
References
Bicknell, B., Imhoff, J., Kittle, J., Donigian, A., Johanson, R., and Barnwell, T.
(1996). "Hydrologic simulation program - FORTRAN user's manual for
release 11," United States Environmental Protection Agency
Environmental Research Laboratory, Athens GA.
Chapter 1 Introduction
3
-------
Cerco, C., and Cole, T. (1993). "Three-dimensional eutrophication model of
Chesapeake Bay." Journal of Environmental Engineering, 119(6), 1006-
10025.
Cerco, C., and Meyers, M. (2000). "Tributary refinements to the Chesapeake Bay
Model," Journal of Environmental Engineering, 126(2), 164-174.
Cerco, C., and Moore, K. (2001). "System-wide submerged aquatic vegetation
model for Chesapeake Bay," Estuaries, 24(4), 522-534.
Cerco, C., and Noel, M. (2004). "The 2002 Chesapeake Bay eutrophication
model," EPA 903-R-04-004, Chesapeake Bay Program Office, US
Environmental Protection Agency, Annapolis, MD.
Cloern, J. (1982). "Does the benthos control phytoplankton biomass in south San
Francisco Bay 1," Marine Ecology Progress Series, 9, 191-202.
Cohen, R., Dresler, P., Phillips, E., and Cory, R. (1984). "The effect of the
Asiatic clam, Corbicula fluminea, on phytoplankton of the Potomac
River, Maryland," Limnology and Oceanography, 29(1), 170-180.
DiToro, D., and Fitzpatrick, J. (1993). "Chesapeake Bay sediment flux model,"
Contract Report EL-93-2, US Army Engineer Waterways Experiment
Station, Vicksburg, MS.
Gerritsen, J., Holland, A., and Irvine, D. (1994). "Suspension-feeding bivalves
and the fate of primary production: An estuarine model applied to
Chesapeake Bay," Estuaries, 17(2), 403-416.
Johnson, B., Kim, K., Heath, R, Hsieh, B., and Butler, L. (1993). "Validation of
a three-dimensional hydrodynamic model of Chesapeake Bay," Journal
of Hydraulic Engineering, 199(1), 2-20.
Linker, L., Shenk, G., Dennis, R, and Sweeney, J. (2000). "Cross-media models
of the Chesapeake Bay watershed and airshed," Water Quality and
Ecosystem Modeling, 1(1-4), 91-122.
Meyers, M., DiToro, D., and Lowe, S. (2000). "Coupling suspension feeders to
the Chesapeake Bay eutrophication model," Water Quality and
Ecosystem Modeling, 1(1-4), 123-140.
Newell, R. (1988). "Ecological changes in Chesapeake Bay: Are they the result
of overharvesting the American oyster (Crassostrea virginica
Understanding the estuary - Advances in Chesapeake Bay Research.
Publication 129, Chesapeake Research Consortium, Baltimore, 536-546.
Newell, R., Cornwell, J., and Owens, M. (2002). "Influence of simulated bivalve
biodeposition and microphytobenthos on sediment nitrogen dynamics,"
Limnology and Oceanography, 47(5), 1367-1379.
Chapter 1 Introduction
4
-------
Newell, R., and Koch, E. (2004). "Modeling seagrass density and distribution in
response to changes in turbidity stemming from bivalve filtration and
seagrass sediment stabilization," Estuaries, 27(5), 793-806.
Officer, C., Smayda, T., and Mann, R. (1982). "Benthic filter feeding: A natural
eutrophication control." Marine Ecology Progress Series, 9, 203-210.
Porter, E., Cornwell, J., and Sanford, L. (2004). "Effect of oysters Crassostrea
virginica and bottom shear velocity on benthic pelagic coupling and
estuarine water quality." Marine Ecology Progress Series, 271, 61-75.
Wang, H., and Johnson, B. (2000). "Validation and application of the second-
generation three-dimensional hydrodynamic model of Chesapeake Bay,"
Water Quality and Ecosystem Modeling, 1(1-4), 51-90.
Chapter 1 Introduction
5
-------
2 The Oyster Model
Introduction
The ultimate aim of eutrophication modeling is to preserve precious
living resources. Usually, the modeling process involves the simulation of
living-resource indicators such as dissolved oxygen. For the "Virginia Tributary
Refinements" phase of the Chesapeake Bay modeling (Cerco et al. 2002), a
decision was made to initiate direct interactive simulation of three living resource
groups: zooplankton, benthos, and SAV.
Benthos were included in the model because they are an important food
source for crabs, finfish, and other economically and ecologically significant
biota. In addition, benthos can exert a substantial influence on water quality
through their filtering of overlying water. Benthos within the model were
divided into two groups: deposit feeders and filter feeders (Figure 1). The
deposit-feeding group represents benthos that live within bottom sediments and
feed on deposited material. The filter-feeding group represents benthos that live
at the sediment surface and feed by filtering overlying water. The primary
reference for the benthos model (HydroQual, 2000) is available on-line at
http://www.chesapeakebay.net/modsc.htm. Less comprehensive descriptions
may be found in Cerco and Meyers (2000) and in Meyers at al. (2000).
The benthos model incorporates three filter-feeding groups: 1) Rangea
cunecita, which inhabit oligohaline and lower mesohaline portions of the system;
2) Macoma baltica, which inhabit mesohaline portions of the system; and 3)
Corbiculciflumineci, which are found in the tidal fresh portion of the Potomac.
These organisms were selected based on their dominance of total filter-feeding
biomass and on their widespread distribution. The distributions of the organisms
within the model grid were assigned based on observations from the Chesapeake
Bay benthic monitoring program
(http://www.chesapeakebay.net/data/index.htm). Oysters were neglected in the
initial application of the benthos model. The primary reasoning was that oyster
biomass was considered negligible relative to the most abundant organisms.
Oysters
The oyster model builds on the concepts established in the benthos
model. The existing benthos model was left untouched. The code was
duplicated and one portion was modified for specific application to native
Chapter 2 The Oyster Model
1
-------
oysters, Crassostrea virginica. The original model assigned one of the three
species exclusively to a model cell. In the revised model, oysters may coexist
and compete with the other filter feeders. The fundamental state variable is
oyster carbon, quantified as mass per unit area. The minimum area represented is
the quadrilateral model cell, which is typically 1 to 2 km on a side. Oyster
biomass and processes are averaged over the cell area. Oysters filter particulate
matter, including carbon, nitrogen, phosphorus, silica, and inorganic solids from
the water column. Particulate matter is deposited in the sediments as feces and
pseudofeces. Respiration removes dissolved oxygen from the water column
while excretion returns dissolved nitrogen and phosphorus.
Particulate carbon is removed from the water column by the filtration
process. Filtration rate is affected by temperature, salinity, suspended solids
concentration, and dissolved oxygen. The amount of carbon filtered may exceed
the oyster's ingestion capacity. In that case, the excess of filtration over
ingestion is deposited in the sediments as pseudofeces (Figure 2). A portion of
the carbon ingested is refractory or otherwise unavailable for nutrition. The
unassimilated fraction is deposited in the sediments as feces. Biomass
accumulation (or diminishment) is determined by the difference between carbon
assimilated and lost through respiration and mortality. Respiration losses remove
dissolved oxygen from the water column. Mortality losses are deposited to the
sediments as particulate carbon.
The nutrients nitrogen and phosphorus constitute a constant fraction of
oyster biomass. Particulate nitrogen and phosphorus, filtered from the water
column, are subject to ingestion and assimilation. Assimilated nutrients that are
not accumulated in biomass or lost to the sediments through mortality are
excreted to the water column in dissolved inorganic form. All filtered particulate
silica is deposited to the sediments or excreted to the water column. A fraction
(~ 10%) of filtered inorganic solids is deposited to the sediments. The fraction is
determined by the net settling velocity specified in the suspended solids
algorithms. The remainder is considered to be resuspended.
The mass-balance equation for oyster biomass is:
— = a-Fr-P0C-IF{\-RF)-0-BM-0-p-0 (1)
dt
in which:
O = oyster biomass (g C m2)
a = assimilation efficiency (0 < a < 1)
Fr = filtration rate (m3 g"1 oyster carbon d"1)
POC = particulate organic carbon in overlying water (g m3)
IF = fraction ingested (0 < IF < 1)
RF = respiratory fraction (0 < RF < 1)
BM = basal metabolic rate (d1)
B = specific mortality rate (d1)
t = time (d)
Chapter 2 The Oyster Model
2
-------
The assimilation efficiency is specified individually for each form of particulate
organic matter in the water column. The respiratory fraction represents active
respiratory losses associated with feeding activity. Basal metabolism represents
passive respiratory losses.
Filtration
Filtration rate is represented in the model as a maximum or optimal rate
that is modified by ambient temperature, suspended solids, salinity, and dissolved
oxygen:
in which:
f(T) = effect of temperature on filtration rate (0 < f(T) < 1)
f(TSS) = effect of suspended solids on filtration rate (0 < f(TSS) < 1)
f(S) = effect of salinity on filtration rate (0 < f(S) < 1)
f(DO) = effect of dissolved oxygen on filtration rate (0 < f(DO) < 1)
Frmax = maximum filtration rate (m3 g"1 oyster carbon d"1)
Bivalve filtration rate, quantified as water volume cleared of particles per
unit biomass per unit time (Winter 1978), is typically derived from observed
rates of particle removal from water overlying a known bivalve biomass
(Doering et al. 1986, Doering and Oviatt 1986, Riisgard 1988, Newell and Koch
2004). Since particle retention depends on particle size and composition
(Riisgard 1988, Langdon and Newell 1990), correct quantification of filtration
requires a particle distribution that represents the natural distribution in the study
system (Doering and Oviatt 1986). Filtration rate for our model was based
primarily on measures (Jordan 1987) conducted in a laboratory flume maintained
at ambient conditions in the adjacent Choptank River, a mesohaline Chesapeake
Bay tributary that supports a population of native oysters. These were
supplemented with laboratory measures conducted on oysters removed from the
same system (Newell and Koch 2004). Jordan reported weight-specific
biodeposition rate as a function of temperature, suspended solids concentration
and salinity. The biodeposition rate represents a minimum value for filtration
since all deposited material is first filtered. Filtration rate was derived:
in which:
WBR = weight-specific biodeposition rate (mg g1 dry oyster weight hr"1)
TSS = total suspended solids concentration (mg L"1)
Filtration rate was converted from L g1 DW h"1 to model units based on a
carbon-to-dry-weight ratio of 0.5.
The observed rates indicate a strong dependence of filtration on
temperature (Figure 3) although the range of filtration rates observed at any
Fr = f(T) •/(TSS) ¦ f (S) ¦ f(DO) ¦ Fr max
(2)
(3)
Chapter 2 The Oyster Model
3
-------
temperature indicate the influence of other factors as well. The maximum
filtration rate and the temperature dependence for use in the model are indicated
by a curve drawn across the highest filtration rates at any temperature:
Fr = Fr max • e~Kts' ^Topt^ (4)
in which:
Frmax = maximum filtration rate (0.55 m3 g"1 oyster carbon d"1)
Ktg = effect of temperature on filtration (0.015 °C~2)
T = temperature for optimal filtration (27 °C)
Suspended Solids Effects. The deleterious effect of high suspended solids
concentrations on oyster filtration rate has been long recognized although the
solids concentrations induced in classic experiments, 102 to 103 g m3 (Loosanoff
and Tommers 1948), are extreme relative to concentrations commonly observed
in Chesapeake Bay. We formed our solids function by recasting Jordan's data to
show filtration rate as a function of suspended solids concentration (Figure 4).
The experiments indicate three regions. Filtration rate was depressed when
solids were below ~ 5 gm m"3 and above ~ 25 gm m3, relative to filtration rate
when solids were between these two levels. The observations suggest oysters
reduce their filtration rate when food is unavailable or when filtration at the
maximum rate removes vastly more particles than the oysters can ingest. We
visually fit a piecewise function to Jordan's data (Figure 4) supplemented with an
approximation of Loosanoff and Tommers' results:
f('rSS) = 0.1 when TSS < 5 g m3
f(TSS) =1.0 when 5 g m3 < TSS < 25 g m3
ffrSSj = 0.2 when 25 g m3 < TSS < 100 g m3
jri'SSj = 0.0 when TSS > 100 g m3
Salinity Effects. Oysters reduce their filtration rate when ambient salinity falls
below ~20% of the oceanic value (Loosanoff 1953) and cease filtering when
salinity falls below ~10% of the oceanic value. The form and parameterization
of a relationship to describe these experiments is arbitrary. We selected a
functional form (Figure 5) used extensively elsewhere in the CBEMP:
f(S) = 0.5 • (l + tanh (S - KHsoy)) (5)
in which:
S = salinity (ppt)
KHsoy = salinity at which filtration rate is halved (7.5 ppt)
Dissolved Oxygen. Hypoxic conditions (dissolved oxygen < 2 g m3) have a
profound effect on the macrobenthic community of Chesapeake Bay. Effects
range from alteration in predation pressure (Nestlerode and Diaz 1998) to species
shifts (Dauer et al. 1992) to near total faunal depletion (Holland et al. 1977). In
the context of the benthos model, effects of hypoxia are expressed through a
Chapter 2 The Oyster Model
4
-------
reduction in filtration rate and increased mortality. The general function from the
benthos model (Figure 6), based on effects from marine species, was adapted
unchanged for the oyster model:
f(DO)
1
f
1 + exp 1.1
DO^-DO^
DOhx-DOqxj
(6)
V
in which:
DO = dissolved oxygen in overlying water (g m3)
DOhx = dissolved oxygen concentration at which value of function is one-half
(1.0 gm3)
DOqx = dissolved oxygen concentration at which value of function is one-fourth
(0.7 gm3)
This logistic function has the same shape as the tanh function used to quantify
salinity effects (Figure 5). The use of two parameters, DOhx and DOqx, allows
more freedom in specifying the shape of the function than the tanh function,
based on the single parameter KHsoy, allows.
Ingestion
Oyster ingestion capacity must be derived indirectly from sparse
observations and reports. In the report on his experiments, Jordan (1987) states
"at moderate and high temperatures and low seston concentration (< 4 mg/L)
nearly all biodeposits were feces" (page 54). This statement indicates no
pseudofeces was produced; all organic matter filtered was ingested. Elsewhere in
Jordan (1987) we find that ~ 75% of seston is organic matter and the filtration
rate at 4 g seston m3 is ~ 0.1 m"3 g"1 oyster C d"1 (Figure 4). The ingestion rate
must be at least the amount of organic matter filtered. Conversion to model units
indicates an ingestion rate of:
4 g seston 0.75 organic gC 0.1 nr _ 0.12 g C ingested
m total 2.5 g seston g C d g oyster C d
Tenore and Dunstan (1973) present a figure showing feeding rate and
biodeposition. The difference between feeding and deposition must be ingestion.
The largest observed difference is 19 mg C g1 DW d"1 or 0.038 g C ingested g1
oyster C d"1 (utilizing a carbon-to-dry-weight ratio of 0.5). No pseudofeces was
produced during their experiments so the derived ingestion rate is not necessarily
a maximum value.
In reporting on the removal of algae from suspension, Epifanio and
Ewart (1977) noted that large amounts of pseudofeces were produced when algal
suspensions exceeded 12 |_ig mL1. These results indicate the amount removed
from the water column when algal suspensions were less than 12 |_ig mL"1, ~ 4 to
17 mg algal DW g1 oyster total weight d"1, was ingested. The 15 g total weight
Chapter 2 The Oyster Model
5
-------
oysters in Epifanio and Ewart's experiments has a dry weight of 0.27 g (Dame
1972). The minimum ingestion rate is then:
4mgalgalDW 15 g I'W g oyster DW g algal C 0.18 g C ingested
g oyster TW 021 g DW 0.5 g oyster C 2500 mg DW g oyster C d
Analogous unit conversions yield 0.76 g C ingested g1 oyster C d"1 for a removal
rate of 17 mg algal DW g1 oyster total weight d1.
Summary of these analyses indicates the order of magnitude for ingestion
rate is 0.1 g C ingested g1 oyster C d1. The value 0.12 g C ingested g1 oyster C
d"1 was employed in the model based on our evaluation of Jordan's experiments.
Assimilation
The fraction of ingested carbon assimilated by oysters depends on the
carbon source. The assimilation of macrophyte detritus can be as low as 3%
(Langdon and Newell 1990) while the assimilation of viable microphytobenthos
is 70% to 90% (Cognie et al.). Tenore and Dunstan (1973) observed that oysters
assimilated 77% to 88% of a mixed algal culture. Specification of assimilation
for the oyster model is shaped by the nature of the eutrophication model. The
eutrophication model considers three forms of particulate organic carbon:
phytoplankton, labile particulate organic carbon, and refractory particulate
organic carbon. Assimilation of phytoplankton is specified as 75%, based on
citations above. The labile and refractory particulate organic carbon are detrital
components. These are mapped to three G classes of organic matter (Westrich
and Berner 1984) employed in the sediment diagenesis model (DiToro 2001).
The Gl, labile, class has half-life of 20 days. The G2, refractory, class has a
half-life of one year. The G3 class is inert within time scales considered by the
model. Model labile particulate organic carbon maps to the Gl class and is
assigned an assimilation efficiency of 75%, corresponding to phytoplankton.
Model refractory particulate organic carbon combines the G2 and G3 classes and
is assigned an assimilation efficiency of zero.
Respiration
Two forms of respiration are considered: active respiration, associated
with acquiring and assimilating food, and passive respiration (or basal
metabolism). This division of respiration is consistent with models of predators
ranging from zooplankton (Steele and Mullin 1977) to fish (Hewett and Johnson
1987). Active respiration is considered to be a constant fraction of assimilated
food. Basal metabolism is represented as a constant fraction of biomass,
modified by ambient temperature:
BM =BMr ¦eKTbmr-{T-Tr) (7)
in which:
BM = basal metabolism (d1)
BMr = basal metabolism at reference temperature (d1)
Chapter 2 The Oyster Model
6
-------
T = temperature (°C)
Tr = reference temperature (°C)
KTbmr = constant that relates metabolism to temperature (°C1)
The rate of basal metabolism depends on organism biomass (Winter
1978, Shumway and Koehn 1982). The average oyster in Jordan's (1987)
experiments, upon which our filtration rates are based, is 2.1 g DW. Allometric
relationships (Shumway and Koehn 1982) indicate basal metabolism for a 2.1 g
DW oyster at 20 °C is 0.002 to 0.005 d"1, depending on salinity. A graphical
summary presented by Winter (1978) indicates metabolic rate for a 2 g DW
oyster at 20 °C is 0.009 d"1. Winter noted a 1 g DW mussel requires 1.5% of its
dry tissue weight daily as a maintenance ration. Based on these reports, the value
0.008 d"1 was employed for basal metabolism at a reference temperature of 20 °C.
Parameter KTbmr was assigned the value 0.069 "C1. equivalent to a Q10 of 2,
typical of measured rates in oysters (Shumway and Koehn 1982).
The respiratory fraction was assigned through comparison of computed
oxygen consumption with metabolism in active oyster reefs (Boucher and
Boucher-Rodoni 1988, Dame et al. 1992). The value RF = 0.1 was determined.
A comparable value of 0.172 (specific dynamic activity coefficient) was assigned
to herbivorous fish in Chesapeake Bay (Luo et al. 2001).
Mortality
The model considers two forms of mortality. These are mortality due to
hypoxia and a term that considers all other sources of mortality including disease
and harvest. Although bivalves incorporate physiological responses that render
them tolerant to hypoxia, extended periods of anoxia result in near-extinction
(Holland et al. 1977, Josefson and Widbom 1988). Casting the results of
experiments and observations into a relationship that quantitatively relates
mortality to dissolved oxygen concentration incorporates a good deal of
uncertainty in functional form and parameterization. The effect of hypoxia on
oyster mortality, adopted from the benthos model, employs two concepts. The
first is the time to death under complete anoxia. This time to death is converted
to a first-order mortality rate via the relationship:
ln( 1/100)
hmr = — - (8)
ttd
in which:
hmr = mortality due to hypoxia (d1)
ttd = time to death for 99% of the population (14 d)
The mitigating effect on mortality of dissolved oxygen concentration
greater than zero is quantified through multiplication by (7 -f(DO)) in which
f(DO) is the logistic function that expresses the effects of hypoxia on filtration
rate (Equation 6). This functionality increases mortality as dissolved oxygen
concentrations become low enough to affect filtration rate (Figure 6). When
dissolved oxygen is depleted, filtration rate approaches zero and mortality is at its
Chapter 2 The Oyster Model
7
-------
maximum. As parameterized in the model, effects on filtration and mortality are
negligible until dissolved oxygen falls below ~ 2 g m"3 (Figure 6). The time to
death for 99% of the population exceeds 90 days when dissolved oxygen exceeds
1.4 g m3 (Figure 7). Under this scheme, some fraction of the oyster population
can survive an entire summer of hypoxia provided dissolved oxygen exceeds 1.4
g m3. No significant portion of the oyster population will survive summer
hypoxia for dissolved oxygen concentrations below 1.4 g m3.
Mortality from all other sources, primarily disease and harvest, is
represented by a spatially uniform and temporally constant first-order term.
Magnitude of the term is specified to produce various system-wide population
levels with the model. The order of magnitude can be derived from Jordan et al.
(2002) who reported the 1990 total mortality of "market stock" oysters in
northern Chesapeake Bay was 0.94 yr-1 (or 0.0026 d1). Of this total, 0.22 yr-1 (or
0.0006 d1) was natural mortality. The balance was fishing mortality.
Nutrients
Model oysters are composed of carbon, nitrogen, and phosphorus in
constant ratios. In the original benthos model (HydroQual 2000), the carbon-to-
nitrogen mass ratio of bivalves was set at 5.67:1; the phosphorus-to-carbon mass
ratio was 45:1. Composition data for bivalves is not abundant. Calculations by
Jordan (1987), based on earlier work by Kuenzler (1961) and Newell (1982),
yield a carbon-to-nitrogen mass ratio between 4.8:1 and 6.9:1 and a phosphorus-
to-carbon mass ratio of 66:1. The nitrogen composition values encompass the
value used in the model. The phosphorus composition value differs from the
model but no context exists to judge if the difference is significant.
The oyster model differs substantially from the original benthos model in
the way nutrients are assimilated and processed. In the original model, nutrients
are assimilated and excreted in constant ratios equivalent to the oyster
composition. If assimilated carbon is in excess relative to assimilated nitrogen or
phosphorus, the excess carbon is converted to feces and the bivalves are
effectively nutrient limited. Computed bivalve growth is:
G = min [Cassim, Nassim ¦ SFCN, Passim ¦ SFCP] (9)
in which:
G = bivalve biomass accumulation (g C m2 d"1)
Cassim = carbon assimilation rate (g C m2 d"1)
Nassim = nitrogen assimilation rate (g N m2 d"1)
SFCN = bivalve carbon-to-nitrogen ratio (g C g1 N)
Passim = phosphorus assimilation rate (g P m2 d"1)
SFCP = bivalve carbon-to-nitrogen ratio (g P g1 N)
If the carbon-to-nitrogen ratio in assimilated food, Cassim/Nassim, exceeds the
ratio in bivalve composition, SFCN, then biomass accumulation is proportional
to the rate of nitrogen assimilation. Similarly, when the ratio Cassim/Passim >
SFCP, biomass accumulation is proportional to phosphorus assimilation. The
Chapter 2 The Oyster Model
8
-------
algal phosphorus-to-carbon ratio in the eutrophication model (Cerco and Noel
2004) is 57:1 for spring diatoms and 80:1 for other algae. Since these ratios
exceed SFCP, growth of bivalves feeding on algae will be limited by the
phosphorus content of the algae rather than the amount of carbon assimilated.
Algal composition does not provide a complete picture of the tendency
for nutrient limitation of bivalve growth since modeled bivalves utilize detritus
well as algae. Initial applications of the oyster model indicated, however, that
phosphorus limitation of oyster growth did occur. Nutrient limitation was
eliminated through two methods. First, oyster phosphorus composition was
thinned out; carbon-to-phosphorus ratio was increased to 90:1. More
significantly, a mass balance approach to nutrient utilization and excretion was
adopted. Biomass accumulation was modeled as carbon assimilation less
respiration loss while nutrient excretion was calculated as the amount of
assimilated nutrients not required for biomass accumulation.
Model Parameters
Parameter values for the oyster model are summarized in Table 1.
Chapter 2 The Oyster Model
-------
Table 1
Parameters for Oyster Model
Parameter
Definition
Value
Units
Frmax
maximum filtration rate
0.55
m3 g"1 oyster carbon d"1
Topt
optimum temperature for filtration
27
°C
Ktg
constant that controls temperature
dependence of filtration
0.015
°C"2
KHsoy
salinity at which filtration rate is halved
7.5
ppt
BMR
base metabolism rate at 20 °C
0.008
d"1
KTbmr
constant that controls temperature
dependence of metabolism
0.069
°C"1
Tr
reference temperature for specification
of metabolism
20
°c
RF
respiratory fraction
0.1
0 < RF < 1
DOhx
dissolved oxygen concentration at
which value of logistic function is one-
half
1.0
gm"3
DOqx
dissolved oxygen concentration at
which value of logistic function is one-
quarter
0.7
gm"3
ttd
time to death for 99% of the population
14
d
3alg
assimilation efficiency for phytoplankton
0.75
0 < a < 1
Slab
assimilation efficiency for labile organic
matter
0.75
0 < a < 1
3ref
assimilation efficiency for refractory
organic matter
0.0
0 < a < 1
Imax
maximum ingestion rate
0.12
g prey C g1 C d"1
SFCN
carbon-to-nitrogen ratio
6
g C g"1 N
SFCP
carbon-to-phosphorus ratio
90
g C g"1 P
Chapter 2 The Oyster Model
10
-------
References
Boucher, G., and Boucher-Rodoni, R. (1988). "In situ measurement of
respiratory metabolism and nitrogen fluxes at the interface of oyster
beds." Marine Ecology Progress Series, 44, 229-238.
Cerco, C., and Meyers, M. (2000). "Tributary refinements to the Chesapeake Bay
Model," Journal of Environmental Engineering, 126(2), 164-174.
Cerco, C., Johnson, B., and Wang, H. (2002). "Tributary refinements to the
Chesapeake Bay model, ERDC TR-02-4, US Army Engineer Research
and Development Center, Vicksburg, MS.
Cerco, C., and Noel, M. (2004). "The 2002 Chesapeake Bay eutrophication
model," EPA 903-R-04-004, Chesapeake Bay Program Olfice, US
Environmental Protection Agency, Annapolis, MD.
Cognie, B., Barille, L., and Rince, Y. (2001). "Selective feeding of the oyster
Crassostrea gigas fed on a natural microphytobenthos assemblage,"
Estuaries, 24(1), 126-131.
Dame, R., (1972). "Comparison of various allometric relationships in intertidal
and subtidal American oysters," Fishery Bulletin, 70(4), 1121-1126.
Dame, R., Spurrier, J., and Zingmark, R. (1992). "In situ metabolism of an oyster
reefJournal of Experimental Marine Biology and Ecology, 164, 147-
159.
Dauer, D., Rodi, A., and Ranasinghe, J. (1992). "Effects of low dissolved oxygen
events on the macrobenthos of the lower Chesapeake Bay," Estuaries,
15(3), 384-391.
DiToro, D. (2001). Sediment Flux Modeling, John Wiley and Sons, New York.
Doering, P., and Oviatt, C. (1986). "Application of filtration rate models to field
populations of bivalves: an assessment using experimental mesocosms,"
Marine Ecology Progress Series, 31, 265-275.
Doering, P., Oviatt, C., and Kelly, J. (1986). "The effects of the filter-feeding
clam Mercenaria mercenaria on carbon cycling in experimental
ecosystems," Journal of Marine Research, 44, 839-861.
Epifanio, C., and Ewart, J. (1977). "Maximum ration of four algal diets for the
oyster Crassostrea virginica Gmelin." Aquacullure. 11, 13-29.
Hewett, S., and Johnson, B. (1997). "A generalized bioenergetics model of fish
growth for microcomputers," WIS-SG-87-245, Wisconsin Sea Grant
College Program, University of Wisconsin, Madison.
Chapter 2 The Oyster Model
11
-------
Holland, A., Mountford, N., and Mihursky, J. (1977). "Temporal variation in
upper bay mesohaline benthic communities: I. The 9-m mud habitat,"
Chesapeake Science, 18(4), 370-378.
HydroQual. (2000). "Development of a suspension feeding and deposit feeding
benthos model for Chesapeake Bay," produced by HydroQual Inc. under
contract to the U.S. Army Engineer Research and Development Center,
Vicksburg MS.
Jordan, S. (1987). "Sedimentation and remineralization associated with
biodeposition by the American oyster Crassostrea virginica (Gmelin),"
Ph.D. diss., University of Maryland, College Park.
Jordan, S., Greenhawk, K., McCollough, C., Vanisko, J., and Homer, M. (2002).
"Oyster biomass, abundance, and harvest in northern Chesapeake Bay:
Trends and forecasts," Journal of Shellfish Research, 21(2), 733-741.
Josefson, A., and Widbom, B. (1988). "Differential response of benthic
macrofauna and meiofauna to hypoxia in the Gullmar Fjord Basin,"
Marine Biology, 100, 31-40.
Kuenzler, E. (1961). "Phosphorus budget of a mussel population," Limnology
and Oceanography, 6, 400-415.
Langdon, C., and Newell, R. (1990). "Utilization of detritus and bacteria as food
sources by two bivalve suspension-feeders, the oyster Crassostrea
virginica and the mussel Geukensia demissa,"Marine Ecology Progress
Series, 58, 299-310.
Loosanoff, V., and Tommers, F. (1948). "Effect of suspended silt and other
substances on rate of feeding of oysters," Science, 107, 69-70.
Loosanoff, V. (1953). "Behavior of oysters in water of low salinities,"
Proceedings of the National Shellfish Association, 43:135-151.
Luo, J., Hartman, K., Brandt, S., Cerco, C., and Rippetoe, T. (2001). "A
spatially-explicit approach for estimating carrying capacity: An
application for the Atlantic menhaden (Brevoortia tyrannus) in
Chesapeake Bay," Estuaries, 24(4), 545-556.
Meyers, M., DiToro, D., and Lowe, S. (2000). "Coupling suspension feeders to
the Chesapeake Bay eutrophication model," Water Quality and
Ecosystem Modeling, 1, 123-140.
Nestlerode, J., and Diaz, R. (1998). "Effects of periodic environmental hypoxia
on predation of a tethered ploychaete, Glycera Americana: implications
for trophic dynamicsMarine Ecology Progress Series, 172, 185-195.
Newell, R. (1982). "An evaluation of the wet oxidation technique for use in
determining the energy content of seston samples," Canadian Journal of
Fisheries and Aquatic Science, 39, 1383-1388.
Chapter 2 The Oyster Model
12
-------
Newell, R., and Koch, E. (2004). "Modeling seagrass density and distribution in
response to changes in turbidity stemming from bivalve filtration and
seagrass sediment stabilization," Estuaries, 27(5), 793-806.
Riisgard, H. (1988). "Eficiency of particle retention and filtration rate in 6
species of Northeast American bivalves." Marine Ecology Progress
Series, 45, 217-223.
Shumway, S., and Koehn, R. (1982). "Oxygen consumption in the American
oyster Crassostrea virginica," Marine Ecology Progress Series, 9, 59-68.
Steele, J., and Mullin, M. (1977). "Zooplankton dynamics." The sea. E.
Goldberg, I. McCave, J. O'Brien, J. Steele eds., Volume 6, Wiley-
Interscience, New York, 857-890.
Tenore, K., and Dunstan, W. (1973). "Comparison of feeding and biodeposition
of three bivalves at different food levels." Marine Biology, 21, 190-195.
Westrich, J., and Berner, R. (1984). "The role of sedimentary organic matter in
bacterial sulfate reduction: The G model testedLimnology and
Oceanography, 29, 236-249.
Winter, J. (1978). "A review of the knowledge of suspension-feeding in
lamellibranchiate bivalves, with special reference to artificial aquaculture
systemsAquaculture, 13, 1-33.
Chapter 2 The Oyster Model
13
-------
Figure 1. Benthos model schematic.
Reapuatioii.
excretion
FJtiiiliuii
Ingestion Assimilation
P^udofecet
Feces
Biomass
Accumulation
f
Moitalitv
Figure 2. Processes affecting filtered material.
Chapter 2 The Oyster Model
14
-------
0.7 -
Jordan
0.6 -
Newell&Koch
Model
0.5 -
o
..
U)
0.4 -
"HS,
TS
co
0.3 -
* *
E
0.2 -
0 I*
*•
0.1 -
" ^ \ ~
1 * :
n _
8 * * *
U t
I T 1 1 T
s
0
5 10 15 20 25
30
Degrees C
Figure 3. Effect of Temperature on filtration rate.
0.6
0.5
TS
O
O)
CO
0.4
0)
•4—1
CO
cc
0,3
¦2 0.2
ffl
u_
0.1
»
• \
-------
0.8
0.6
o
s
±±
il
c
o
° 0 4
-------
DO (g rrf3)
Figure 7. Effect of dissolved oxygen on time to death for 99% of population.
Chapter 2 The Oyster Model
17
-------
3 Biomass Estimates
Introduction
The Chesapeake 2000 Agreement calls for a tenfold increase in native
oysters in the Chesapeake Bay, based on a 1994 baseline. At the commencement
of this study, no estimate of the baseline oyster population existed. Evaluation of
the existing population and its distribution was required before the effects of
proposed increases could be examined. Since our model is based on mass
balance, population estimates took the form of mass rather than number of
individuals. We use the terms "biomass" to indicate total weight of oysters e.g.
kg C and "density" to indicate weight per unit area e.g. g C m2.
Existing Biomass
Virginia
Density estimates for Virginia were provided by Dr. Roger Mann, of
Virginia Institute of Marine Science (VIMS), in October 2003. Estimates were
based on patent tong surveys. The EPA Chesapeake Bay Program Office
(CBPO) provided VIMS with model grid coordinates. Patent tong samples were
averaged for each model cell and results were provided as g DW/m2. Number of
samples per cell varied from 4 to more than 50. Estimates were provided for one
to five individual years in the interval 1998-2002. The coefficient of variation
(CV, defined as standard deviation/mean) for inter-annual density estimates in
individual cells (one or two km on a side) ranged from 0.11 to 1.67 with a
median value of 0.69. The CV of the inter-annual total biomass was 0.088. The
area of cells containing oysters was 377 km2.
Maryland
Biomass and spatial distribution for Maryland were based on the
recommendation of Dr. Roger Newell of the University of Maryland Center for
Environmental Science. Dr. Newell recommended recent biomass estimates
(Jordan et al. 2002) should be uniformly distributed across the historical oyster
habitat denoted in the "Yates" surveys (Yates 1911). The areas and locations of
named oyster bars were obtained by the CBPO and bar areas were assigned to
model cells. Total area of named oyster bars was 1330 km2. Mean biomass for
the period 1991-2000, 5.7 x 10s g DW, was obtained from Jordan et al. (2002).
A mean density of 0.43 g DW m2 (total biomass / total area) was assigned to the
Chapter 3 Biomass Estimates
1
-------
bar area in each model cell. Since the bar area was usually less than the cell area,
cell density was adjusted so that biomass per cell matched biomass of bars within
the cell. The area of cells containing oysters was 3696 km2.
Other Filter Feeders
Examination of the effects of oyster restoration requires consideration of
existing filter feeders. Observations from the bay-wide benthic database
(http://www.chesapeakebay.net/data/index.htm) were analyzed by HydroQual
(2000) as part of the initial benthos modeling effort. The analysis indicates
suspension feeding bivalves are distributed primarily in the upper bay and
tributaries (Figure 1). Average bivalve densities in the upper bay are commonly
an order of magnitude or more greater than the present density of oysters. The
arithmetic densities computed by HydroQual are perhaps influenced by a few
large density values; median densities might present a more realistic picture.
Still, the data support the decision to neglect Maryland oysters in the original
benthos model. In the lower bay, the existing oyster density is substantial
relative to other bivalves in the lower Rappahannock River and in a limited
portion of the James River. The decision to neglect existing oysters in these
rivers should be revisited. Recent research (Thompson and Schaffner 2001)
indicates polycheate filter feeders, with reported densities ~ 6 g C m"2, may also
play a substantial role in the lower bay.
Summary
The oyster density and distribution are distinctly different in the
Maryland and Virginia portions of the bay (Figure 2). In the northern, Maryland,
portion, lower densities are distributed over a wide area. In the southern,
Virginia, portion, high densities are concentrated in limited areas, primarily in the
lower James and Rappahannock Rivers. Oyster biomass in Virginia is five times
the biomass in Maryland (Table 1) but distributed across an order of magnitude
less area. We were puzzled by the limited distribution in Virginia, especially
since maps and other information we obtained indicated a wider distribution of
lease holdings and restoration areas. We were assured by Dr. Roger Mann that
much of the leased area is unproductive and that biomass outside the areas
reported to us is negligible. Our estimate of Maryland biomass is roughly half
the biomass from two other independent estimates (Table 1). Our estimate of
Virginia biomass is three times the biomass from an alternate independent
estimate (Table 1).
Modeled Biomass
Model oyster density is dynamically computed based on environmental
conditions including temperature, dissolved oxygen, salinity, and food supply.
The densities are not specified as model inputs. Rather, they must be calculated
as a function of model parameters and computed conditions. The calculation,
rather than specification, of density ensures that oysters are not placed where
conditions do not support their specified density. We initially attempted to
calculate target oyster densities through dynamic variation of the mortality
function. Mortality in each model cell was adjusted upwards or downwards as
Chapter 3 Biomass Estimates
2
-------
Table 1
Oyster Biomass Estimates
Source
Maryland, kg C
Virginia, kg C
Comments
This study
287,000
1,170,000
Maryland from Jordan
et al (2002). Virginia
from Roger Mann
(personal
communication).
Newell (1988)
550,000
400,000
Uphoff (2002)
570,000
Year 2000 exploitable
biomass based on
skipjack catch per
effort
calculated density exceeded or fell below specified levels. This process
successfully capped density at target levels but many cells would not support
existing density or a tenfold increase. The problem originated with the attempt to
calculate target densities within individual cells. The calculated conditions in
many cells would not support the target densities. Consequently, we switched to
a strategy in which a bay-wide target biomass was specified. A uniform bay-
wide mortality rate was prescribed that produced the target biomass. The
mortality rate was obtained through a trial-and-error process in which various
rates were prescribed and the calculated biomass was examined.
The spatial distributions of biomass and density are conveniently
examined through aggregation of individual model cells into Chesapeake Bay
Program Segments (CBPS). Program segments are subdivisions of the bay
determined by mean salinity, natural boundaries, and other features. Our analysis
is based on the original (circa 1993) segmentation (Table 2, Figure 3) in which
the bay is divided in 35 segments with a median area of 150 km2.
Computed density and biomass vary on intra-annual and inter-annual
bases (Figure 4). Variations within the annual cycle are largely driven by
temperature. Highest densities are computed in late summer and in fall, after a
season of filtering at peak rates (Figure 5). Variations from year to year (Figure
6) are largely driven by runoff. Variations in runoff may enhance or diminish
computed biomass, depending on local factors. Years with high runoff coincide
with large nutrient loads that result in high phytoplankton abundance. The
advantages produced by abundant food may be offset, however, by increased
anoxia and by sub-optimal salinity.
Baseline Estimates
First-order estimates of the density and biomass of existing bivalve filter
feeders can be obtained from the latest application of the CBEMP (Cerco and
Noel 2004). This benthos component of this model was originally calibrated to
match the observed density in the bay-wide benthic database (HydroQual 2000).
Subsequent review (Schaffner et al. 2002) indicated the model tends to over-
predict suspension-feeding density in the lower to mid-bay (where density is low)
and under-predicts or approximates suspension-feeding density in the upper bay
Chapter 3 Biomass Estimates
3
-------
and tributaries (where density is high). Still, the model biomass is a useful
baseline, especially in the absence of alternate bay-wide abundance estimates.
Table 2
Chesapeake Bay Program Segmenl
ts that Support Oysters
CBPS
Designation
State
CB2
Upper Chesapeake Bay
Maryland
CB3
Upper Central Chesapeake Bay
Maryland
CB4
Upper Middle Chesapeake Bay
Maryland
CB5
Lower Chesapeake Bay
Maryland - Virginia
CB6
Western Lower Chesapeake Bay
Virginia
CB7
Eastern Lower Chesapeake Bay
Virginia
EE1
Eastern Bay
Maryland
EE2
Lower Choptank River
Maryland
EE3
Tangier Sound
Maryland - Virginia
ET4
Chester River
Maryland
ET5
Choptank River
Maryland
ET6
Nanticoke River
Maryland
ET7
Wicomico River
Maryland
ET8
Manokin River
Maryland
ET9
Big Annemessex River
Maryland
LE1
Lower Patuxent River
Maryland
LE2
Lower Potomac River
Maryland
LE3
Lower Rappahannock River
Virginia
LE5
Lower James River
Virginia
RET1
Middle Patuxent River
Maryland
RET2
Middle Potomac River
Maryland
WE4
Mobjack Bay
Virginia
WT6
Magothy River
Maryland
WT7
Severn River
Maryland
WT8
South River
Maryland
Autumn is the season when individual oysters attain maximum biomass
and when most population surveys, on which our estimates are based, are
conducted. For comparison with estimates of existing oysters, we averaged the
calculated autumn (September - November) bivalve density and biomass from
ten years (1985 - 1994). The density comparisons are averaged across total
bottom area in each CBPS. The resulting densities are less than individual
observations or averages across oyster bars since area not suited for bivalves is
included in the average. In most portions of the bay, the calculated density of
existing bivalve filter feeders vastly exceeds the estimated density of oysters
(Figure 7). Notable exceptions are in the Rappahannock (LE3) and James (LE5)
where existing oysters exceed other bivalve filter feeders. Oysters also
predominate in two Eastern Shore tributaries (ET8, ET9) and in the lower
Chapter 3 Biomass Estimates
4
-------
western shore of the mainstem (CB6). These segments are characterized by the
virtual absence of other bivalves rather than by abundant oysters, however.
Biomass comparisons (Figure 8) reflect the density comparisons. Oyster biomass
exceeds other bivalve biomass in the lower Rappahannock and James Rivers.
Oysters are virtually the only bivalves in the two noted Eastern shore tributaries
(ET8, ET9) and in the lower western shore of the bay (CB6).
These comparisons have implications for the overall modeling effort and
for the present work. As noted previously, the decision to ignore oysters in the
model, until now, was a valid one, with the exception of the lower Rappahannock
and James Rivers. For the present study, the model runs with no oysters provide
an acceptable baseline for comparison with tenfold population increase since the
oysters comprise only a small fraction of filter-feeding biomass throughout most
of the bay.
Tenfold Increase
The model run for examination of the tenfold population increase, called
for in the Chesapeake 2000 Agreement, was determined through a recursive
process in which mortality rate was varied until the desired biomass was
obtained. Intra- and inter-annual variations in computed biomass made an exact
multiplier of existing oyster biomass impossible to obtain. We settled on
comparison of computed autumn (September - November) biomass with
population estimates since most surveys are conducted in the fall. We compared
the mean of ten computed years, 1985-1994, with the estimates of existing
population. We settled on a first-order mortality rate of 0.015 d"1, which
produced a mean biomass 13-times the estimated existing biomass (Table 3).
Biomass in individual years varied by roughly 50% above and below the mean.
We refer to this run as the "tenfold increase" although the magnitude and spatial
distribution of the increase varies. The southern, Virginia, portion of the bay
receives only a fourfold biomass increase while the northern, Maryland, portion
increases nearly 50-times. The disparity in multipliers reflects the disparity in
initial biomass distribution. An implication of this model run is that, under
existing conditions, the northern portion of the bay suffers higher mortality from
harvest and disease than the southern portion since imposition of a uniform
mortality rate results in greater biomass in the north than in the south. Estimates
of the present population indicate the opposite trend. With the tenfold increase,
oysters become the dominant filter feeders in the system (Figures 9, 10) although
other bivalves predominate in a few segments that provide marginal oyster
habitat. Also worth noting is a decline in bivalve biomass, as much as 50%,
throughout much of the bay (Figure 11).
Historical Biomass
As one part of sensitivity analyses, we computed the biomass of oysters
with no mortality from harvest or predation. Limitations to biomass in this run
were food availability, respiration, and mortality from hypoxia. The computed
biomass (Table 3) that resulted approached the pre-1870 biomass estimated by
Newell (1988). This run is documented as an example of improvements that
could result from full restoration of historic oyster biomass.
Chapter 3 Biomass Estimates
5
-------
Table 3
Estimated and Modeled Oyster Biomass, kg C
Maryland
Virginia
Total
Existing, estimated
287,005
1,099,339
1,386,344
Historic (Newell 1988)
94,000,000
Tenfold, model
14,107,500
4,374,953
18,482,453
Historic, model
69,749,506
17,165,230
86,914,736
Equivalent Settling and Removal Rates
The influence of oysters on the environment is a function of their
density, filtration rate, and local geometry. The product of density and filtration
rate has units of length/time (velocity) and is denoted here as "Equivalent
Settling Rate":
Woys = --\0-FrdA (1)
A J
in which:
Woys = equivalent settling rate (m d"1)
A = area over which rate is computed (m2)
O = oyster density (g C m2)
Fr = filtration rate (m3 g"1 oyster carbon d"1)
The equivalent settling rate can be viewed as the velocity at which particles are
transferred from the water column into the oyster bed. Higher velocities indicate
more rapid removal. However, the distance to be covered (depth) affects
removal as well as velocity. Geometry is brought into the characterization
through calculation of "Equivalent Removal Rate":
Roys =--\—-0-FrdA (2)
A J D
in which:
Roys = equivalent removal rate (d1)
D = local depth (m)
The equivalent removal rate can be viewed as a decay rate of material in the
water column. High removal rates indicate the bivalves clear the water column
rapidly. The inverse of the equivalent removal rate is an "Equivalent Residence
Time": the time required for the bivalves to filter the water column once.
Under existing conditions, highest settling rates are in smaller tributaries;
lower settling rates prevail in the mainstem bay and in the portions of major
western tributaries that adjoin the bay (Figure 12). The tenfold biomass increase
Chapter 3 Biomass Estimates
6
-------
(Figure 13) and the historic biomass (Figure 14) shift the highest settling rates to
the lower portions of the western tributaries and to the upper mainstem of the
bay. Median settling velocity increases by an order of magnitude from present
modeled conditions to historical conditions (Table 4).
Under existing conditions, the ranking of residence times corresponds to
the ranking of settling rates (Figure 15). Shortest residence times (highest
turnover rates) are in tributaries. More lengthy residence times prevail in the
lower portions of western tributaries and in the mainstem bay. The effects of
geometry influence the rankings under conditions of oyster restoration (Figures
16, 17). Several of the large-volume segments which rank high in terms of
settling rate rank lower when their depth is incorporated into the index of
potential bivalve influence. Overall, the median residence time of individual
CBPS's diminishes from 18 days under computed existing conditions to less than
three days under historic oyster densities (Table 4).
Table 4
Median Settling Rates, Removal Rates, and Residence Times
Settling, m d"1
Removal, d"1
Residence, d
Existing Conditions
0.15
0.04
18.3
Tenfold Oyster Increase
0.62
0.19
5.3
Historic Conditions
1.44
0.38
2.6
References
Cerco, C., and Noel, M. (2004). "The 2002 Chesapeake Bay eutrophication
model," EPA 903-R-04-004, Chesapeake Bay Program Office, US
Environmental Protection Agency, Annapolis, MD.
HydroQual. (2000). "Development of a suspension feeding and deposit feeding
benthos model for Chesapeake Bay," produced by HydroQual Inc. under
contract to the U.S. Army Engineer Research and Development Center,
Vicksburg MS.
Jordan, S., Greenhawk, K., McCollough, C., Vanisko, J., and Homer, M. (2002).
"Oyster biomass, abundance, and harvest in northern Chesapeake Bay:
Trends and forecasts," Journal of Shellfish Research, 21(2), 733-741.
Newell, R. (1988). "Ecological changes in Chesapeake Bay: Are they the result
of overharvesting the American oyster (Crassostrea virginica
Understanding the estuary - Advances in Chesapeake Bay Research.
Publication 129, Chesapeake Research Consortium, Baltimore, 536-546.
Schaffner, L., Friedrichs, C., and Dauer, D. (2002). "Review of the benthic
processes model with recommendations for future modeling efforts," A
Report from the Benthic Process Model Review Team, EPA Chesapeake
Bay Program, Annapolis MD. (Available at
http://www.chesapeakebay.net/modsc.htm)
Chapter 3 Biomass Estimates
7
-------
Thompson, M., and Schaffner, L. (2001). "Population biology and secondary
production of the suspension feeding polychaete Chaetopterus cf.
variopedatus: Implications for benthic-pelagic coupling in lower
Chesapeake Bay," Limnology and Oceanography, 46(8), 1899-1907.
Uphoff, J. (2002). "Biomass dynamic modeling of oysters in Maryland,"
Maryland Department of Natural Resources, Annapolis. (Unpublished
manuscript provided by the author).
Yates, C. (1911). "Survey of the oyster bars by county of the State of Maryland,"
Department of Commerce and Labor, Coast and Geodetic Survey,
Washington DC.
Chapter 3 Biomass Estimates
8
-------
J9 75
39.50
39. ,>5 "
39 00
38 75 "
38 50
38.25
3801
37 75 *
37 50
37 go
36 3
Z
p-SaTM
Virginia Heach
77.50
-77 00
-76 50
-76.00
-75.50
-75 00
Figure 1. Density of existing bivalve filter feeders (from HydroQual 2000)
Chapter 3 Biomass Estimates
9
-------
Figure 2. Present oyster density in Chesapeake Bay
Chapter 3 Biomass Estimates
-------
Figure 3. Chesapeake Bay Program Segments
Chapter 3 Biomass Estimates
-------
3500
OY20L
SPEED EE2
i
I
1
f\ 1
3000
-
1 h
2500
"e
q2000
O
E
1500
- j
: /
A II 4
A h
M
1000
7
i i i
500
y i j \ i u v
, .y ,y ,y .
1 \ i \ j \ j
y \j V/1 v
jJ
1 2 3 4 5
Years
6 7 8 9
10
Figure 4. Calculated oyster density in the lower Choptank River, 1985-1994
2500
2000
1500
£
O
cn
1000
500
December- March - May June-August September -
February November
Figure 5. Seasonal-average calculated oyster density in the lower Choptank River
Chapter 3 Biomass Estimates
12
-------
3500
3000
2500
"E 2000
a
D)
E 1500
1000
500
I I I I
I I
1985 1986 1987 1988 1989 1990 1991 1992 1993 1994
Figure 6. Calculated autumn oyster density in lower Choptank River
3.5
2,5
E
o
o>
1.5
0.5
¦I Oysters
¦ Bivalve Suspension
Feeders
; . I iLtl
# ,<$• ,<&¦ ,<& A A A A A 4y 4? <# /> ,# a®3
Figure 7. Estimated density of existing oysters and bivalve filter feeders
Chapter 3 Biomass Estimates
13
-------
1,200,000
1,000,000
: Oysters
: Bivalve Suspension
Feeders
800,000 -
o
oj 600,000 -
¦J*
400,000 -
200,000
#gif f i $$$$ & £ 4? £ £ e £ 4? £ «///#
Figure 8. Estimated biomass of existing oysters and bivalve filter feeders
7 -
r Oysters
® lJ Bivalve Suspension
Feeders
5 -
r, 4 -
'E
o
Figure 9. Calculated density of oysters and bivalve filter feeders under the nominal
tenfold increase in oyster biomass
Chapter 3 Biomass Estimates
14
-------
4000000
3500000
I i Oysters
Bivalve Suspension Feders
3000000 -
2500000 -
ra 2000000 -
1500000 -
1000000
500000 -
, if !,!¦¦[
_
COCOCQCQCQtQLULUUJ
O O O O U U LU LU LLi
'q- ur>
I— I— I— I— I— b~
LU LU LU LU LU LU
Figure 10. Calculated biomass of oysters and bivalve filter feeders under the
nominal tenfold increase in oyster biomass
1200000
1000000
" Existing
with tenfold oyster increase
800000 -
o
o) 600000 -
J*
400000 -
200000 -
JL J'
J1
~1
OOOOOOIlllUlUllllllllllJllJIU
Figure 11. Effect of tenfold increase in oyster biomass on biomass of other bivalve
filter feeders
Chapter 3 Biomass Estimates
15
-------
0.4 -
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0 -T-
X.*3 S*b AV >>N /V sJ\ xib y/b X*V ~fb v>Jb /A x /y A**5 A® sfr -b a6> aS
-------
_ „
sfo 'h Ah x^ <5-^ x.^ >A xfb y/s ~ V # *
«
Q
90 -
80 -
70 -
60 -
50 -
40 -
30 -
20 -
10 -
^ <4? £ 4? 4? <& S
-------
30 -
27 -
24 -
21 -
18 -
15 -
12 -
9 -
6 -
3 -
Figure 16. Time for oysters and bivalves to filter the water column under the
tenfold increase in oyster biomass
30
27 -
24 -
21 -
18 -
15 -
12 -
9 -
6 -
, n, i, n, i,; , n, , J h t 11 IJI p
^ 4s $ <£•* & & 4* $ «# S
Figure 17. Time for oysters and bivalves to filter the water column under historic
conditions
Chapter 3 Biomass Estimates
18
-------
4 Oyster Effects on Water
Quality
Introduction
Oysters affect the environment on a variety of spatial scales ranging from
their immediate surroundings outwards to the entire water body. The effects are
considered here on three scales. The first is the smallest that can be resolved in
the model, the model cell. Cell areas are ~ 106 m2, an order of magnitude larger
than typical Maryland oyster bars. Since modeled oysters are uniformly
distributed within cells, however, the processes in cells occupied by oysters are
comparable to processes in bars containing similar densities of oysters. The
second spatial scale is the regional scale represented by Chesapeake Bay
Program Segments (CBPS). Program segments (Figure 1) are subdivisions of the
bay determined by mean salinity, natural boundaries, and other features. Median
area is ~ 1.5 x 10s m2, of which only a fraction is occupied by oyster bottom.
The third scale is system-wide, an area of 1 x 1010 m2, as represented by the
model grid.
We selected three of the 35 CBPS for detailed examination of oyster
effects on the regional scale. The selected segments (Figure 1) provide a range
of geometry (Table 1) and environmental conditions. CB4 is a mainstem bay
segment with the greatest volume, surface area, and depth of the selected
segments. Due to the depth, only 70% of the area is suitable for oyster habitat, as
determined by the historic Yates surveys. Perhaps the most significant
characteristic of the segment is the regular occurrence of summer bottom-water
anoxia. EE2 is an eastern embayment that encompasses the mouth of the
Choptank River. Volume is an order of magnitude less and depth is half of the
selected mainstem segment. Virtually all of EE2 is suitable oyster habitat.
Minimum dissolved oxygen concentration in bottom water occasionally falls
below 3 g m3 but persistent anoxia does not occur. Segment ET9 is the Big
Annemessex River, located on the Maryland eastern shore. Despite the name,
the Big Annemessex is the smallest of the three selected segments, separated by
an order of magnitude in volume and area from EE2. Average depth is roughly
half the depth in the lower Choptank River. Virtually all the segment provides
suitable oyster habitat and minimum dissolved oxygen concentration exceeds 6 g
m3.
Chapter 4 Oyster Effects on Water Quality
1
-------
Table 1
Regional Characteristics
Region
Volume, 109m3
Area, km2
Mean Depth, m
Fraction oyster
bottom
CB4
10.8
966
11.2
0.71
EE2
1.8
334
5.3
1.00
ET9
0.1
33
2.8
0.8
Local Effects
Biomass-Specific Effects
Effects on the local scale can be normalized by oyster biomass or by
surface area. Biomass-specific results allow comparisons to published rates in
Chesapeake Bay and elsewhere. For examination of biomass-specific effects,
we selected a cell at a depth of 6.7 m within the lower Choptank River, CBPS
EE2 (Figure 1). This region supports a viable oyster population and represents
the environment from which oysters were drawn for the experiments of Jordan
(1987) and Newell and Koch (2004).
Biomass-specific filtration rates, computed within the model based on
the simulated environment, agree closely with the experiments on which the rates
were based as well as with other independent measures and calculations (Table
1). Order-of-magnitude similarity prevails between modeled and measured
respiration and ammonium excretion (Table 1). An interesting contrast occurs
with carbon deposition (Table 1). The model agrees well with Jordan's measures
but departs from other reports. The modeled and measured filtration and
respiration measures are comparable across systems because these are primarily
functions of oyster physiology. Carbon deposition is influenced by local organic
carbon concentration as well as by physiological processes and, consequently,
can only be compared when local organic carbon concentrations are similar.
Areal-Based Effects
The regional and system-wide effects of oyster restoration are best
understood by first isolating the local impacts of oysters. This is accomplished
by examining sediment diagenetic processes and fluxes between the bottom
sediments, oysters, and water column for a range of oyster densities. The basis
for comparison is the 2002 version of the model (Cerco and Noel 2004), which
included no oysters. This is compared to multiple model runs with oysters,
conducted at various mortality rates, that produced a range of oyster densities.
Three cells are considered, one each from CB4, EE2, and ET9. All values are
annual averages across the ten simulated years.
Benthic Algae. Benthic algae (Figure 2) are non-existent in the CB4 (3.7 m
depth) and EE2 (6.7 m depth) cells in the absence of oysters. The shallow ET9
cell (2.1 m depth) supports viable benthic algae at zero oyster density. Density of
benthic algae increases in all cells concurrent with oyster density as oysters clear
the water column of suspended solids. The enhancement of benthic algae is
consistent with experimental results (Newell et al. 2002, Porter et al. 2004)
Chapter 4 Oyster Effects on Water Quality
2
-------
although only the ET9 cell sustains algal density we calculate is sufficient to
influence nutrient exchange at the sediment-water interface (Cerco and Noel
2004). The model state variable is algal carbon. Most observations are of
Table 1
Modeled and Observed Biomass-Specific Oyster Effects
Property
Rate
Source
Comments
Filtration rate,
m3 g"1 oyster
Crf1
0.24
Model
Summer average
0.22
Jordan
(1987)
Mean value, T > 20 °C
0.26
Newell and
Koch (2004)
Average of measures at 20 and 25 oC
0.027 to 0.33
Epifanio and
Ewart (1977)
For algal suspensions > 1 g C m"3
0.27
Riisgard
(1988)
Calculated for a 2.1 g DW oyster at 27 to 29
°C
Respiration
rate, g DO g"1
oyster C d
0.04
Model
Summer average
0.03 to 0.06
Boucher and
Boucher-
Rodini (1988)
Spring and summer rates
0.017
Dame et al.
(1992)
Annual average
0.02
Dame (1972)
1 g DW oyster at 20 to 30 °C
Ammonium
excretion, mg
N g"1 oyster C
d"1
1.43
Model
Summer average
<0.1
Hammen et
al. (1966)
Ammonium plus urea
2.8 to 3.88
Boucher and
Boucher-
Rodini (1988)
Spring and summer rates, includes urea
0.8
Srna and
Baggaley
(1976)
1 g DW oyster at 20 °C
4.8 to 7.9
Magni et al
(2000)
Ruditapes and musculista
Carbon
deposition, g
C g"1 oyster C
d"1
0.088
Model
Summer average
0.099
Jordan
(1987)
Mean value, T > 20 °C
0.03
Haven and
Morales-
Alamo (1966)
0.002 to
0.012
Tenore and
Dunstan
(1973)
Depends on C concentration, range is 0.1 to
0.7 g C m"3
Chapter 4 Oyster Effects on Water Quality
3
-------
chlorophyll. Using a carbon-to-chlorophyll ratio of 50 (Gould and Gallagher
1990) indicates annual-average computed benthic algal chlorophyll is 30 to 40
mg m2 in the ET9 cell.
Carbon and Oxygen Fluxes. The introduction of oysters results in
biodeposition of carbon to the sediments (Figure 3). Carbon deposition due to
gravitational settling (Figure 4) is simultaneously diminished as particulate
carbon that previously settled is instead filtered. Total carbon deposition (Figure
5) is diminished by the introduction of oysters indicating that the minimum
computed density is sufficient to reduce net production of particulate carbon in
the water column. The amount of carbon removed by filtering (Figure 6) levels
off as oyster densities increase beyond the initial value. Several cells indicate
diminished filtration at the highest oyster densities. We attribute the level
filtration to an equilibrium between carbon supplied, through transport and
production, and carbon removed. As oyster density increases, biodeposition
decreases. At higher densities, larger fractions of the carbon filtered are lost
through respiration or retained as biomass. Total carbon deposition, through
settling and biodeposition, decreases continually in response to increased oyster
density.
Increasing oyster densities are accompanied by continual increases in
respiration (Figure 7) and decreases in diagenetic sediment oxygen consumption
(Figure 8). As noted in the biomass-specific results, respiration is largely a
function of oyster density, independent of location. The increased respiration is
more than offset by decreased sediment oxygen consumption so that total oxygen
consumption decreases as oyster density increases (Figure 9).
Nitrogen. Fluxes of particulate nitrogen reproduce the pattern
established for carbon. The introduction of oysters produces biodeposits to the
sediments. As oyster density increases, both biodeposition and settling decrease.
Biodeposition decreases because a greater fraction of nitrogen filtered is lost
through respiration or retained as biomass. Settling decreases because formation
of particulate nitrogen in the water column, through algal activity, is diminished
by oyster predation.
The introduction of oysters diminishes the release of diagenetically-
produced sediment ammonium (Figure 10). Diminished ammonium release is
partially offset by excretion from oysters but the net impact of oysters is reduced
net release to the water column, especially at highest densities (Figure 11). Two
processes contribute to the reduction in diagenetic ammonium release. The role
of reduced nitrogen deposition is obvious. Enhanced sediment nitrification to
nitrate is also apparent, as evidenced by enhanced sediment denitrifiction of
nitrate to nitrogen gas (Figure 12). Denitrification is also enhanced by the flux of
nitrate from the water column into the sediments; nitrate no longer used in algal
production diffuses into the sediments instead. The net effect of oysters on total
nitrogen is removal from the water column via enhanced denitrification and
retention in the sediments (Figure 13).
Phosphorus. Oyster effects on particulate phosphorus follow the pattern
established for carbon and nitrogen. Introduction of oysters results in
biodeposition, which is partially offset by diminished gravitational settling. As
Chapter 4 Oyster Effects on Water Quality
4
-------
oyster density increases, both biodeposition and settling decrease. Biodeposition
decreases because a greater fraction of phosphorus filtered is lost through
respiration or retained as biomass. Settling decreases because formation of
particulate phosphorus in the water column, through algal activity, is diminished
by oyster predation.
The net effect of oysters on dissolved phosphorus contrasts with nitrogen
and is site-specific. At two sites, release of diagenetically-produced phoshorus
diminishes as oyster density increases while at the third site release of diagenetic
phosphorus is largely independent of oyster density (Figure 14). The two sites at
which release diminishes support the largest densities of benthic algae so
interception of diagenetic phosphorus release is suggested. At the site with least
benthic algae, EE2, oyster phosphorus excretion adds to the constant diagenetic
flux so that net release of dissolved phosphorus to the water column increases
(Figure 15) and net retention in the sediments decreases (Figure 16). At the other
two sites, excretion offsets algal uptake so the net flux is nearly constant and
retention in the sediments increases as a non-linear function of oyster density.
Regional Effects
Three model runs are considered: 1) no oyster restoration, derived from
the 2002 version of the model; 2) a tenfold increase in oyster biomass; and 3)
historic oyster density. Quantities selected for analysis include:
• Summer-average bottom dissolved oxygen,
• Summer-average surface chlorophyll,
• Summer-average light attenuation,
• Summer-average SAV biomass,
• Annual-average surface algal carbon,
• Annual-average net primary production,
• Annual-average particulate carbon deposition,
• Annual-average sediment oxygen demand,
• Annual-average surface total nitrogen,
• Annual-average particulate nitrogen deposition,
• Annual-average sediment diagenetic ammonium flux,
• Annual-average net nitrogen removal (denitrification plus burial),
• Annual-average surface total phosphorus,
• Annual-average particulate phosphorus deposition,
• Annual-average sediment diagenetic phosphorus release, and
• Annual-average net phosphorus removal
Our convention for surface concentration is the average over the upper 6.7 m of
the water column, roughly the depth of the surface mixed layer in the mid-bay.
Bottom dissolved oxygen is represented by all waters below 12.9 m in CB4 and
below 6.7 m in EE2. Due to shallow depth, the surface mixed layer coincides
with the bottom in ET9. Results are averaged across the entire regional area and
across all model years.
Chapter 4 Oyster Effects on Water Quality
5
-------
CB4
Water quality standards in Chesapeake Bay are based on dissolved
oxygen, chlorophyll, and water clarity. The ten-fold oyster increase improves
summer-average, bottom, dissolved oxygen in this mainstem segment by less
than 0.5 g iri3 (Figure 17). Simulation of historic oyster densities improves
dissolved oxygen by roughly 1 g m3. Computed surface chlorophyll is reduced
by 30% for a ten-fold increase in oyster density and is halved when oysters are
restored to historic densities (Figure 18). Light attenuation is reduced by roughly
20% for a ten-fold increase in oyster densities and by roughly 40% when oysters
are restored to historic densities (Figure 19).
The improvements in dissolved oxygen and chlorophyll are effected by
reductions in net primary production (Figure 20). A 20% reduction in production
accompanies the ten-fold increase in oyster density. A reduction of nearly 40%
results from restoration of historic densities. The water clarity improvements,
effected by removal of phytoplankton and other solids from the water column,
produce increases in computed SAV biomass of 33% to more than 100% (Figure
21).
Restoration of oysters increases net nitrogen removal (Figure 22),
through denitrification and sediment retention, by 20% to 50% although the
reduction in surface total nitrogen concentration is only 10% to 15% (Figure 23).
When averaged over the region, the effect of oyster restoration is increased
phosphorus retention in the sediments (Figure 24). Net removal increases by a
third for a ten-fold increase in oyster density and doubles when oysters are
restored to historic densities. Phosphorus concentration in the water column
corresponds with net removal rates more closely than nitrogen (Figure 25).
Surface total phosphorus concentration is reduced by 20% to 40%.
EE2
Improvements in summer-average, bottom, dissolved oxygen at the
mouth of the Choptank are consistent with the mainstem segment: less than 0.5 g
m"3 for a ten-fold increase in oyster density and roughly 1 g m3 for restoration to
historic densities (Figure 26). Percentage reductions in surface chlorophyll
(Figure 27) and light attenuation (Figure 28) also correspond closely with the
adjacent mainstem segment as do the reductions in net primary production
(Figure 29) and improvements in SAV (Figure 30).
ET9
Computed dissolved oxygen concentration in the eastern shore
embayment declines by 0.5 g iri3 as a consequence of oyster restoration (Figure
31). The decline in dissolved oxygen reflects diminished dissolved oxygen
production associated with the 40% to 60% reduction in net primary production
(Figure 32). Reductions in summer surface chlorophyll exceed the reductions in
annual net production (Figure 33). The ten-fold increase in oyster density
induces a 60% decrease in summer surface chlorophyll while restoration to
historic densities induces a greater then 70% decrease. Light attenuation in this
region decreases by a third to nearly a half (Figure 34). Corresponding increases
Chapter 4 Oyster Effects on Water Quality
6
-------
in SAV greatly exceed the responses in other segments (Figure 35). SAV
biomass nearly triples for a ten-fold increase in oyster density and increases by
greater than a factor of four for restoration to historic oyster densities.
Regional Budgets
Nutrient budgets were constructed for each of the regions for the three
subject model runs. Results are annual averages across all model years. Terms
in the budgets are:
• Point Source - Direct inputs from municipal and industrial facilities
• Distributed - Loads to the region from the adjacent watershed
• Atmospheric - Loads to the water surface
• Transport - Net loads from the upstream region. For CB4, this is
adjacent mainstem region CB2. For EE2, this is the Choptank River
segment ET5. No upstream segment exists for ET9.
• Net Removal - Accumulation in the bottom sediments plus
denitrifi cation
• Incremental - Increase in net removal due to oysters
Nitrogen transport down the mainstem of the bay dwarfs all other
sources and sinks in CB4 (Figure 36). In view of the enormity of nitrogen
transported in relative to the amount removed by oysters, the ability of oyster
restoration to impact this segment at all is remarkable. This budget suggests the
impact of oysters on phytoplankton is through direct grazing rather than through
nutrient removal that results in limits to phytoplankton growth. Although
nutrient removal can be viewed as an ecosystem service, direct grazing should be
regarded as the primary service. More phosphorus is removed in CB4 than flows
in from upstream and local sources (Figure 37). The deficit is made up by net
phosphorus transport from downstream, as indicated by our earliest model (Cerco
and Cole 1994) and by bay nutrient budgets (Boynton et al. 1995). As with
nitrogen, the incremental nutrient removal by oysters is small relative to the net
transport along the bay axis.
Incremental nutrient removal by oysters in EE2 is significant relative to
other regional sources and sinks. Under the restoration scenarios, net nitrogen
(Figure 38) and phosphorus (Figure 39) removal exceed the local sources
indicating nutrient import from the adjacent mainstem segment.
Nitrogen loading and net removal in segment ET9 are closely balanced
under existing conditions (Figure 40). As with EE2, enhanced removal via oyster
restoration results in nitrogen import from the adjacent Tangier Sound. This
segment imports phosphorus under existing conditions (Figure 41). Net import is
enhanced under conditions of oyster restoration.
System-Wide Effects
The methods, properties examined, and budgeting from the regional
analyses are extensible to the entire system. We consider the system to extend
from the fall lines of major tributaries to the mouth of the bay. We were
Chapter 4 Oyster Effects on Water Quality
7
-------
requested to make two supplementary model runs for the sponsor. These
combined the 2002 model (Cerco and Noel 2004) with the nutrient and solids
loads from the recent allocation. One run was completed without oysters. The
second run incorporated the ten-fold oyster restoration. Since the results of those
runs have not been documented, we summarize them here.
Summer-average dissolved oxygen concentration is considered for all
portions of the bay greater than 12.9 m depth. Dissolved oxygen increases by
0.25 g m3 for the ten-fold oyster restoration and by 0.8 g m3 for restoration to
historic levels (Figure 42). By way of comparison, the dissolved oxygen
improvement attained by the allocation loads exceeds the improvement attained
by oyster restoration to historic levels. Allocation loads combined with ten-fold
oyster restoration provide the greatest level of improvement, more than 1 g m3
over current levels. System-wide, summer, surface chlorophyll concentration
declines by more than 1 mg m3 for a ten-fold increase in oyster biomass and by
2.5 mg m3 for restoration to historic levels (Figure 43). As with dissolved
oxygen, the allocation loads provide greater benefit than oyster restoration with
improved benefits from both load reductions and oyster restoration.
The improvements in dissolved oxygen and chlorophyll are effected by
reductions in net primary production (Figure 44). A 14% reduction in system-
wide production accompanies the ten-fold increase in oyster density. A reduction
of 25% results from restoration of historic densities. The allocation loads
provide greater reductions in algal production than any level of oyster restoration
and greatest reductions accompany load reductions and oyster restoration.
The water clarity improvements that accompany oyster restoration
(Figure 45) produce increases in computed system-wide SAV biomass of 25% to
more than 60% (Figure 46). The historic levels of oysters result in the greatest
improvements in SAV, suggesting local solids removal can be more effective
than indirect controls on organic solids effected through nutrient controls. Still,
the allocation loads produce larger improvements than the proposed ten-fold
increase in oyster biomass.
Load reductions produce greater reductions in total nutrients than oyster
restoration. The allocation loads diminish system-wide surface total nitrogen by
0.27 g m3 (Figure 47) and total phosphorus by 0.011 g m3 (Figure 48) with
marginal additional reductions accomplished by load reductions combined with
oyster restoration. The maximum nutrient reductions accomplished by oyster
restoration are 0.11 g m3 total nitrogen and 0.009 g m"3 total phosphorus. These
results contrast the different strategies for phytoplankton control. The allocation
loads reduce phytoplankton through nutrient reductions. Oyster restoration
controls phytoplankton by direct grazing; nutrient reductions are a by-product of
algal removal.
System-wide nutrient budgets can be constructed that parallel the
regional budgets. In this case, transport is the net flux at the mouth of the bay.
Negative transport indicates nutrient loss to the ocean; positive transport
indicates nutrient import from the ocean Ten-fold oyster restoration removes
30,000 kg d"1 total nitrogen from the system (Figure 49). Oysters at historic
levels remove 54,000 kg d1. Ten-fold oyster restoration removes 4,000 kg d1
Chapter 4 Oyster Effects on Water Quality
8
-------
total phosphorus from the system (Figure 50). Oysters at historic levels remove
5,000 kg d"1. By way of comparison, the ten-fold restoration removes loading
roughly equivalent to direct atmospheric deposition. These are 25,000 kg d"1
total nitrogen and 1,900 kg d1 total phosphorus.
References
Boucher, G., and Boucher-Rodoni, R. (1988). "In situ measurement of
respiratory metabolism and nitrogen fluxes at the interface of oyster
beds." Marine Ecology Progress Series, 44, 229-238.
Boynton, W., Garber, J., Summers, R, and Kemp, W. (1995). "Inputs,
transformations, and transport of nitrogen and phosphorus in Chesapeake
Bay and selected tributaries," Estuaries, 18(1B), 285-314.
Cerco, C., and Cole, T. (1994). "Three-dimensional eutrophication model of
Chesapeake Bay," Technical Report EL-94-4, US Army Engineer
Waterways Experiment Station, Vicksburg, MS.
Cerco, C., and Noel, M. (2004). "The 2002 Chesapeake Bay eutrophication
model," EPA 903-R-04-004, Chesapeake Bay Program Office, US
Environmental Protection Agency, Annapolis, MD.
Dame, R., (1972). "Comparison of various allometric relationships in intertidal
and subtidal American oysters," Fishery Bulletin, 70(4), 1121-1126.
Dame, R., Spurrier, J., and Zingmark, R. (1992). "In situ metabolism of an oyster
reefJournal of Experimental Marine Biology and Ecology, 164, 147-
159.
Epifanio, C., and Ewart, J. (1977). "Maximum ration of four algal diets for the
oyster Crassostrea virginica Gmelin." Aquacullure. 11, 13-29.
Gould, D., and Gallagher, E. (1990). "Field measurements of specific growth
rate, biomass, and primary production of benthic diatoms of Savin Hill
Cove, BostonLimnology and Oceanography, 35, 1757-1770.
Hammen, C., Miller, H., and Geer, W. (1966). "Nitrogen excretion of
Crassostrea virginica," Comparative Biochemistry and Physiology, 17,
1199-2000.
Haven, D., and Morales-Alamo, R. (1966). "Aspects of biodeposition by oysters
and other invertebrate filter feeders," Limnology and Oceanography, 11,
487-498.
Jordan, S. (1987). "Sedimentation and remineralization associated with
biodeposition by the American oyster Crassostrea virginica (Gmelin),"
Ph.D. diss., University of Maryland, College Park.
Chapter 4 Oyster Effects on Water Quality
9
-------
Magni, P., Montani, S., Takada, C., and Tsutsumi, H. (2000). "Temporal scaling
and relevance of bivalve nutrient excretion on a tidal flat of the Seto
Inland Sea, Japan," Marine Ecology Progress Series, 198, 139-155.
Newell, R., and Koch, E. (2004). "Modeling seagrass density and distribution in
response to changes in turbidity stemming from bivalve filtration and
seagrass sediment stabilization," Estuaries, 27(5), 793-806.
Riisgard, H. (1988). "Eficiency of particle retention and filtration rate in 6
species of Northeast American bivalves." Marine Ecology Progress
Series, 45, 217-223.
Srna, R., and Baggaley, A. (1976). "Rate of excretion of ammonia by the hard
clam Mercenaria mercenaria and the American oyster Crassostrea
virginica,"Marine Biology, 36, 251-258.
Tenore, K., and Dunstan, W. (1973). "Comparison of feeding and biodeposition
of three bivalves at different food levels." Marine Biology, 21, 190-195.
Chapter 4 Oyster Effects on Water Quality
10
-------
Chapter 4 Oyster Effects on Water Quality
11
-------
2500
2000
p'
1500 - /
r
4
CT
if)
o
CD
If 1000
co
500
Berithic Algae
¦ ET9
CB4
EE2
10
g Oyster C / sq m
15
20
Figure 2. Effect of oysters on benthic algae.
Particulate Carbon Biodeposits
0.18 -
0.16 -
0.14 -
¦o 0.12 -
£ 0.10 -
^ 0.08 -
O
CT0.06 -
0.04 -
0.02 -
0.00
• ET9
CB4
EE2
10
g oyster C / sq m
15
20
Figure 3. Effect of oysters on particulate carbon biodeposition.
Chapter 4 Oyster Effects on Water Quality
12
-------
Particulate Carbon Settling
0.9
0.2 -
0.1 -
0.0 -
0 5 10 15 20
g Oyster C / sq m
Figure 4. Effect of oysters on gravitational settling of particulate carbon.
Total Carbon Deposition
0.8
0.7
0.6
0.5
g"0.4
E
cr
to
O
or
0.3
0.2
0.1
0.0
10
g oyster C / sq m
¦ ET9
CB4
EE2
15
20
Figure 5. Effect of oysters on total carbon deposition.
Chapter 4 Oyster Effects on Water Quality
-------
Particulate Carbon Filtration
0.25
0.20
0.15
cr
o o.io
OJ
0.05
f-
¦ ET9
CB4
EE2
0.00
0 5 10
g Oyster C / sq m
Figure 6. Effect of oysters on particulate carbon filtration.
15
20
Oyster Respiration
X
¦ ET9
CB4
Choptank Deep
10
g oyster C / sq m
15
20
Figure 7. Effect of oysters on areal respiration.
Chapter 4 Oyster Effects on Water Quality
14
-------
Sediment Oxygen Demand
g Oyster C / sq m
Figure 8. Effect of oysters on sediment oxygen demand.
Total Oxygen Consumption
2.0 n
ra0.6 -
0.4 -
0.2 -
0.0 n 1 1 1
0 5 10 15 20
g oyster C / sq m
Figure 9. Effect of oysters on total benthic oxygen consumption.
Chapter 4 Oyster Effects on Water Quality
15
-------
Sediment-Water Ammonium Flux
"o 40 v
£
« 30 -
¦ ET9
CB4
EE2
E 20
10
10
g oyster C / sq m
15
Figure 10. Effect of oysters on sediment-water ammonium flux.
w 30
E 20
10
Net Dissolved Nitrogen Flux
-ET9
i- CB4
EE2
40 - \
20
10
g oyster C / sq m
15
Figure 11. Effect of oysters on net benthic dissolved nitrogen flux.
20
Chapter 4 Oyster Effects on Water Quality
-------
20
Denitrificatiori
tr..
w 15
E 10
¦ ET9
CB4
EE2
0 5 10
g oyster C / sq m
Figure 12. Effect of oysters on sediment denitrification.
15
20
Net N Removal
T3
35
30
25
E 20
cr
CO
z 15
o>
E
10
i
10
g oyster C / sq m
¦ ET9
CB4
EE2
15
20
Figure 13. Effect of oysters on net sediment nitrogen removal.
Chapter 4 Oyster Effects on Water Quality
-------
Sediment-Water P Flux
3,0 -
-1.0 J
g oyster C / sq m
Figure 14. Effect of oysters on sediment-water dissolved phosphorus flux. Positive
flux is release to the water column.
Net Dissolved P Flux
3.0
¦ ET9
CB4
EE2
10
g oyster C / sq m
15
20
Figure 15. Effect of oysters on net benthic dissolved phosphorus flux.
Chapter 4 Oyster Effects on Water Quality -| g
-------
Net P Removal
0.0 -I 1 1 1
0 5 10 15 20
g oyster C / sq m
Figure 16. Effect of oysters on net sediment phosphorus removal.
Dissolved Oxygen (>12.8m)
2.5 -I
2.0
1.5
d
0.5
o.o -I— £; i
No Oysters Ten_fold Historic
Figure 17. Effect of oysters on summer-average, bottom, dissolved oxygen in CB4
Chapter 4 Oyster Effects on Water Quality
-------
d
OS
3
10
9
8
7
6
5
4
3
2
1
0
Surface Chlorophyll
Wo Oysters
Ten fold
Historic
Figure 18. Effect of oysters on summer-average, surface, chlorophyll in CB4.
1.0
0.9
0.8
0.7
0.6
| 0.5
0.4
0.3
0.2
0.1
0.0
Light Attenuation
No Oysters
Ten fold
Historic
Figure 19. Effect of oysters on summer-average light attenuation in CB4.
Chapter 4 Oyster Effects on Water Quality
-------
Net Primary Production
1 .0 n
0.8
1
s05
(J
W
0.3
Figure 20. Effect of oysters on annual-average net phytoplankton primary
production in CB4.
SAV Biomass
200
180
160
140
120
100
80
60
40
20
0
No Oysters
Ten fold
Historic
Figure 21. Effect of oysters on summer-average SAV biomass in CB4.
Chapter 4 Oyster Effects on Water Quality
-------
Figure 22. Effect of oysters on net benthic nitrogen removal in CB4.
Figure 23. Effect of oysters on annual-average, surface, total nitrogen in CB4.
Chapter 4 Oyster Effects on Water Quality
22
-------
Figure 24. Effect of oysters on net benthic phosphorus removal in CB4.
Figure 25. Effect of oysters on annual-average, surface, total phosphorus in CB4.
Chapter 4 Oyster Effects on Water Quality
23
-------
Dissolved Oxygen 06.7m)
Nd Oysters
Ten fold
Historic
Figure 26. Effect of oysters on summer-average, bottom, dissolved oxygen in EE2.
9
8
7
6
¦d 5
OS
= 4
3
2
1
0
Surface Chlorophyll
No Oysters
Ten fold
Historic
Figure 27. Effect of oysters on summer-average, surface, chlorophyll in EE2.
Chapter 4 Oyster Effects on Water Quality
24
-------
Light Attenuation
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
No Oysters
Ten fold
Historic
Figure 28. Effect of oysters on summer-average light attenuation in EE2.
Wet Primary Production
1.0
0.8
¦o
E
£
o
a
0.5
0.3
0.0
No Oysters
Ten fold
Historic
Figure 29. Effect of oysters on annual-average net phytoplankton primary
production in EE2.
Chapter 4 Oyster Effects on Water Quality
-------
SAV Biomass
700
600
_ 500
o
| 400
| 300
C
° 200
100
0 -I ^ i ^ r
No Oysters Ten_fold Historic
Figure 30. Effect of oysters on summer-average SAV biomass in EE2.
Dissolved Oxygen (<6.7m)
7 -I
6
5
d 4
2
1
0 -I 1 T
No Oysters Ten_t'old Historic
Figure 31. Effect of oysters on summer-average dissolved oxygen in ET9.
Chapter 4 Oyster Effects on Water Quality
-------
Net Primary Production
0.3
0.6 j=j
1
£ 0 4
u
a
0.2 —
0.0 -I , ,
Wo Oysters Tenjold Historic
Figure 32. Effect of oysters on annual-average net phytoplankton primary
production in ET9.
12
10
8
ra 6
3
4
2
0
Figure 33. Effect of oysters on summer-average chlorophyll in ET9.
Surface Chlorophyll
¦
Wo Oysters Ten fold Historic
Chapter 4 Oyster Effects on Water Quality
-------
Light Attenuation
1.2 -I 1
1.0
0.8
¦§ 0.6
0.4
0.2
0.0 -I 1 1 1
No Oysters Ten_foid Historic
Figure 34. Effect of oysters on summer-average light attenuation in ET9.
SAV Biomass
120 -i
100
80
60
40
20
0 1 1 1
No Oysters Ten_fold Historic
Figure 35. Effect of oysters on summer-average SAV biomass in ET9.
Chapter 4 Oyster Effects on Water Quality
c
o
¦e
O
8
c
c
o
-------
CS
30000
25000
20000
15000
10000
5000
0
142,978
Nitrogen in CB4
139.082
136.903
. . Point Source
liiiiiiiT
:::::::::::
J Distributed
::::::::::
::::::::::::
III
] Atmospheric
::::::::::
III
III
] Transport
::::::::::
III
III
a Net Removal
::::::::::
111
11
_ Incremental
11
111
11
11
111
11
11
Existing Ten-Times Historic
Figure 36. Effect of oysters on nitrogen budget in CB4.
Phosphorus in CB4
! Point Source
! Distributed
! Atmospheric
i Transport
= Net Removal
! Incremental
Existing Ten-Times
Figure 37. Effect of oysters on phosphorus budget in CB4.
Historic
Chapter 4 Oyster Effects on Water Quality
29
-------
Nitrogen in EE2
_J Point Source
U Distributed
-Atmospheric
Transport
H Net Removal
. : Incremental
Existing
Ten-Times
Historic
Figure 38. Effect of oysters on nitrogen budget in EE2.
Phosphorus in EE2
! . Point Source
~ Distributed
: Atmospheric
Transport
3 Net Removal
II. Incremental
Existing
Ten-Times
Historic
Figure 39. Effect of oysters on phosphorus budget in EE2.
Chapter 4 Oyster Effects on Water Quality
-------
Nitrogen in ET9
600
~ Point Source
~ Distributed
II Atmospheric
Transport
g Net Removal
: : Incremental
- 300 II 1
200 M ¦ I ¦ ¦
100 4 11 |
0 i ' i i ' -
Existing Ten-Times Historic
Figure 40. Effect of oysters on nitrogen budget in ET9.
Phosphorus in ET9
I Point Source
J Distributed
. i Atmospheric
i Transport
=3 Net Removal
J Incremental
0 -
Existing Ten-Times Historic
Figure 41. Effect of oysters on phosphorus budget in ET9.
Chapter 4 Oyster Effects on Water Quality
31
-------
Dissolved Oxygen 012.8m)
4.0
3.0 —
d
? 2.0
1.0 —
0.0
No Oysters Ten_fold Historic Allocation + Allocation
Oysters
Figure 42. Effect of oysters on system-wide summer-average, bottom, dissolved
oxygen.
Surface Chlorophyll
8 -|
7
6
5 —
3
2 _
1 -
o -I ! ! "
No Oysters Ten_fold Historic Allocation + Allocation
Oysters
Figure 43. Effect of oysters on system-wide, summer-average, surface chlorophyll.
Chapter 4 Oyster Effects on Water Quality
32
-------
Net Primary Production
0.8 1
0.6
1 1 [
g o.4 —
o
0.2 —
0.0
No Oysters Ten_fold Historic Allocation + Allocation
Oysters
Figure 44. Effect of oysters on system-wide, annual-average, net phytoplankton
primary production.
Liqht Attenuation
1 4 1
1.2 ——
1.0 ==-
*" 0.6 —
0.4 —
0.2 —
0 0 -| , 1 ¦ , 1 I —
No Oysters Ten_fold Historic Allocation + Allocation
Oysters
Figure 45. Effect of oysters on system-wide, summer-average, light attenuation.
Chapter 4 Oyster Effects on Water Quality
-------
SAV Biomass
20000 -1
18000 r-
16000
c 14000 ——
12000
10000
1 8000
J 6000
4000
2000
0 -I— ,— t—
No Ten_fold Historic
Oysters
Allocation Allocation
+ Oysters
Figure 46. Effect of oysters on system-wide, summer-average, SAV biomass.
Total Nitrogen
0.3
0.S
!oj
0.2
0.0 -1=^ T
No Oysters Ten_fold Historic
T 1
Allocation + Allocation
Oysters
Figure 47. Effect of oysters on system-wide, annual-average, surface, total nitrogen.
Chapter 4 Oyster Effects on Water Quality
34
-------
Total Phosphorus
0.04
0.03
| 0.02 -
0.01
0.00
No Oysters Ten_fold Historic
T 1 1
Allocation + Allocation
Oysters
Figure 48. Effect of oysters on system-wide, annual-average, surface, total
phosphorus.
400000
300000
200000
100000
-100000
-200000
-300000
-400000
Figure 49. Effect of oysters on system-wide nitrogen budget.
System-Wide Nitrogen
~ Point Source
~ Distributed
~ Atmospheric
~ Transport
¦ Net Removal
~ Incremental
n
Exis
tin;
1
Ten-"
"im
^ ' Hisl
ori
J
Chapter 4 Oyster Effects on Water Quality
35
-------
System-Wide Phosphorus
O)
30000
20000
10000
0
-10000
-20000
-30000
-40000
-50000
I—r
::::::::::::
::::::::::::
Existing!!
Ten-Tim
Hi
Histori
-------
5 Discussion and Conclusions
Analysis of the oyster modeling is like peeling the proverbial onion.
There's always another layer to be examined. Every insight produces two more
questions. Sufficient model runs have been conducted to resolve the oyster issue
raised by the Chesapeake Bay 2000 Agreement:
By 2004, assess the effects of different population levels of filter
feeders such as menhaden, oysters and clams on Bay water
quality and habitat.
Additional examination of the runs can be conducted and fruitful insights remain
to be obtained. The production of this report is motivated by the need to produce
tangible, citable, documentation of the work completed to date.
Oyster restoration will, no doubt, benefit the bay environment. Our
analyses indicate the chief benefit will be restoration of SAV, brought about by
filtration of solids from the water column. The most significant conclusion from
our work, however, is that oyster restoration is no panacea for the host of
environmental problems that plague the bay. Oyster restoration should be
viewed as one of many contributions to remediation of the bay's problems.
Our work did not target specific regions of the bay with specific levels of
restoration. Rather, target levels for system-wide biomass were attained and the
spatial distribution of oysters was calculated dynamically based on computed
environmental factors including salinity, suspended solids, and available food.
Potential spatial distribution was limited to historic oyster beds. As a result of
our approach, the modeled ten-fold increase in oyster biomass multiplied oysters
in the Maryland portion of the bay by 50 times while the Virginia portion of the
bay received only a four-fold increase, primarily in the lower James and
Rappahannock Rivers. Consequently, our ten-fold increase probably exaggerates
the benefits to be obtained by ten-fold increases in local oyster densities in the
northern bay.
Our work indicates a ten-fold oyster increase will improve summer-
average, bottom, dissolved oxygen by ~ 0.3 g m"3 in the portion of the mainstem
plagued by the worst anoxia. Oyster restoration alone is not likely to bring the
deep channel of the mainstem into compliance with dissolved oxygen standards.
A dissolved oxygen increase of 0.3 g m3 has economic value when traded off
against the costs of nutrient controls. Some portions of the bay that marginally
violate dissolved oxygen standards will marginally meet the standards when
improved by 0.3 g iri3. System-wide, the combination of oyster restoration
Chapter 5 Discussion and Conclusions
1
-------
and the recent nutrient allocations are calculated to increase summer-average,
bottom, dissolved oxygen by ~ 1.1 g m3.
Multiple reasons can be offered for the absence of more significant
dissolved oxygen response to oyster restoration. The obvious explanation is that
oysters are found in the shoals rather than over the deep trench. Phytoplankton
production over the trench remains free to settle to bottom waters and contribute
to anoxia. A more subtle explanation lies in the origins of mainstem anoxia.
Oxygen depletion in the upper bay does not originate solely with excess
production in the overlying waters. Rather, oxygen depletion is accumulated as
net circulation moves bottom water up the channel from the mouth of the bay.
This mechanism was originally proposed by Kuo et al. (1991) for the
Rappahannock River and has been shown to apply to the mainstem bay as well
(Cerco 1995). Improvement in upper bay dissolved oxygen requires reduction in
lower bay oxygen demand. The oyster restoration strategy does nothing to
diminish oxygen demand in the lower bay and, consequently, has limited impact
on the upper bay.
Our work indicates oyster restoration removes both nitrogen and
phosphorus from the bay water column. Nitrogen removal is more significant
than phosphorus removal since nitrogen is the nutrient that contributes to excess
algal production in the portions of the bay occupied by oysters (Fisher et al.
1992, Mai one et al. 1996). We calculate the ten-fold increase in oyster biomass
removes 30,000 kg d1 total nitrogen from the system via enhanced denitrification
and retention in the sediments. This removal can be put into perspective by
noting the Susquehanna River provides ~ 150,000 kg d"1 total nitrogen to the
mainstem while point sources in the Baltimore vicinity provide ~ 15,000 kg d"1
(Cerco and Noel 2004). Oyster restoration may substitute for a major upgrade in
point-source controls but does not offset the larger distributed loading from the
watershed.
The comparison above does not address timing. Loads from the
watershed arrive largely during spring runoff and occasionally as autumn tropical
storms. Removal via oysters occurs during the warm months concurrent with
peak algal production. This issue introduces the question of primary "services"
provided by oysters. We suggest the primary service is direct grazing on algae.
Rather than quantifying the amount of nitrogen removed by oysters, we should
ask what load reductions produce reductions in algal biomass equivalent to the
reductions from grazing. Nutrient removal is a byproduct of grazing. In order
for nutrient removal to have value, it must be shown that the removal enhances
limits to algal production. The model can provide insights in this regard and
additional examination is warranted.
Our model provides unique capability to address oyster restoration in the
bay. We believe ours is the first approach to combine detailed representation of
the bay geometry with mechanistic representations of three-dimensional
transport, water-column eutrophication processes, sediment diagenetic processes,
and dynamic computation of oyster biomass. Due to the large number of
computed interactions, exact quantification of benefits such as S AV biomass
improvement involves uncertainty. We believe, however, our basic findings
regarding the nature and magnitude of restoration benefits are valid. Our results
Chapter 5 Discussion and Conclusions
2
-------
are consistent with the earlier findings of Officer et al (1992) and Gerritson et al.
(1994) and with the recent findings of Newell and Koch (2004). Benthic controls
of algal production are most effective in shallow, spatially-limited regions. In
these shallow regions, oyster removal of solids from the water column enhances
adjacent SAV beds. The ability to influence deep regions of large spatial extent
is limited by the location of oysters in the shoals and by exchange processes
between the shoals and deeper regions.
The potential improvements obtained by oyster restoration are also
limited by factors not considered in the model. Disease is an obvious limitation.
Habitat destruction has also been suggested as an impediment (Rothschild et al.
1994). We recommend that oyster restoration be targeted to specific areas with
suitable environments and that resulting environmental improvements be viewed
on similar, local scales.
References
Cerco, C. (1995). "Response of Chesapeake Bay to nutrient load reductions,"
Journal of Environmental Engineering, 121(8), 549-557.
Cerco, C., and Noel, M. (2004). "The 2002 Chesapeake Bay eutrophication
model," EPA 903-R-04-004, Chesapeake Bay Program Office, US
Environmental Protection Agency, Annapolis, MD.
Fisher T, Peele E, Ammerman J, Harding L (1992). Nutrient limitation of
phytoplankton in Chesapeake Bay. Mar Ecol Prog Ser 82:51-63
Gerritsen, J., Holland, A., and Irvine, D. (1994). "Suspension-feeding bivalves
and the fate of primary production: An estuarine model applied to
Chesapeake Bay," Estuaries, 17(2), 403-416.
Kuo, A., Park, K., and Moustafa, Z. (1991). "Spatial and temporal variabilities of
hypoxia in the Rappahannock River, Virginia," Estuaries, 14(2), 113-
121.
Mai one T, Conley D, Fisher T, Glibert P, Harding, Sellner K (1996). Scales of
nutrient-limited phytoplankton productivity in Chesapeake Bay.
Estuaries 19:371-385
Newell, R, and Koch, E. (2004). "Modeling seagrass density and distribution in
response to changes in turbidity stemming from bivalve filtration and
seagrass sediment stabilization," Estuaries, 27(5), 793-806.
Officer, C., Smayda, T., and Mann, R. (1982). "Benthic filter feeding: A natural
eutrophication control." Marine Ecology Progress Series, 9, 203-210.
Rothschild, B., Ault, J., Goulletquer, P., and Heral, M. (1994). "Decline of the
Chesapeake Bay oyster population: a century of habitat destruction and
overfishing," Marine Ecology Progress Series, 111, 29-39.
Chapter 5 Discussion and Conclusions
3
-------
Chapter 5 Discussion and Conclusions
------- |