&EPA
United States Office of Research and Development (ORD) EPA/600/B-18/241
Environmental Protection August 2018
Agency
AQUATOX (RELEASE 3.2)
MODELING ENVIRONMENTAL FATE
AND ECOLOGICAL EFFECTS IN
AQUATIC ECOSYSTEMS
VOLUME 2: TECHNICAL DOCUMENTATION
-------
{This Page Left Blank, Back of Cover}
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
AQUATOX (RELEASE 3.2)
MODELING ENVIRONMENTAL FATE
AND ECOLOGICAL EFFECTS IN
AQUATIC ECOSYSTEMS
VOLUME 2: TECHNICAL DOCUMENTATION
Richard A. Park1
and
Jonathan S. Clough2
AUGUST 2018
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT (ORD)
OFFICE OF SCIENCE AND TECHNOLOGY
WASHINGTON DC 20460
^co Modeling (Retired), Diamondhead MS 39525
2Warren Pinnacle Consulting, Inc., Warren VT 05674
l
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
DISCLAIMER
This document describes the scientific and technical background of the aquatic ecosystem model
AQUATOX, Release 3.2. Anticipated users of this document include persons who are interested
in using the model, including but not limited to researchers and regulators. The model described
in this document is not required, and the document does not change any legal requirements or
impose legally binding requirements on EPA, states, tribes or the regulated community. This
document has been approved for publication by the Office of Science and Technology, Office of
Water, U.S. Environmental Protection Agency. Mention of trade names, commercial products or
organizations does not imply endorsement or recommendation for use.
ACKNOWLEDGMENTS
This model has been developed and documented by Richard A. Park of Eco Modeling and by
Jonathan S. Clough of Warren Pinnacle Consulting, Inc.. The work was funded with Federal
funds from the U.S. Environmental Protection Agency, Office of Science and Technology under
contract numbers 68-C-01-0037 to AQUA TERRA Consultants, Anthony Donigian, Work
Assignment Leader; and EP-C-12-006 to Horsley Witten Group, Inc., Nigel Pickering, Work
Assignment Leader. Integration of Interspecies Correlation Estimation (Web-ICE) was made
possible due to the work of US. EPA Office of Research and Development Gulf Breeze, the
University of Missouri-Columbia, and the US Geological Survey.
Release 3.2 was developed under contract HHSN316201200013W, Task Order EP-G16H-01256
"Scientific Models, Applications, Visualizations, Computational Science, and Statistical Support
(SMAVCS3)," under contract to with CSRA LLC, with Henry Helgen as TDD Lead.
The assistance, advice, and comments of the EPA work assignment manager, Maijorie Coombs
Wellman of the Standards and Health Protection Division, Office of Science and Technology
have been of great value in developing this model and preparing this report. Further technical
and financial support from David A. Mauriello, Rufus Morison, and Donald Rodier of the Office
of Pollution Prevention and Toxics is gratefully acknowledged. Marietta Echeverria, Office of
Pesticide Program, contributed to the integrity of the model through her careful analysis and
comparison with EXAMS. Amy Polaczyk and Marco Propato, Warren Pinnacle Consulting,
made valuable contributions to the model formulation and testing. Release 3.2 was developed
with the thoughtful assistance of Rajbir Parmar of the National Exposure Research Laboratory
(ORD), and Brenda Rashleigh of the USEPA Environmental Effects Research Laboratory
(ORD).
Release 2 of this model underwent independent peer review by Donald DeAngelis, Robert
Pastorok, and Frieda Taub; and Release 3 underwent peer review by Marty Matlock, Damian
Preziosi, and Frieda Taub. Their diligence is greatly appreciated.
11
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
The addition of nearshore marine environments within the AQUATOX model was developed
and documented by Eldon Blancher of Sustainable Ecosystem Restoration, LLC, and by Richard
A. Park of Eco Modeling and Jonathan S. Clough of Warren Pinnacle Consulting, Inc. under
subcontract to University of Southern Mississippi; and by Scott Milroy of the University of
Southern Mississippi. Special thanks to the members of our technical-review team for their
suggestions and oversight of our progress:
• Joshua Bergeron, University of Southern Mississippi
• Justin Blancher, Sustainable Ecosystem Restoration, LLC
• Cori Gavin, Sustainable Ecosystem Restoration, LLC
• Meg Goecker, Sustainable Ecosystem Restoration, LLC
• Monty Graham, University of Southern Mississippi
• Read Hendon, University of Southern Mississippi
• Robert Leaf, University of Southern Mississippi
• Stephan O'Brien, University of Southern Mississippi
• Stephen Parker, Adaptive Management Services, LLC
• Chet Rakocinski, University of Southern Mississippi
• Kelly Robinson, Oregon State University
• Tom Strange, Sustainable Ecosystem Restoration, LLC
• Jerry Wiggert, University of Southern Mississippi
• Katherine Woodard, University of Southern Mississippi
111
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
TABLE OF CONTENTS
DISCLAIMER ii
TABLE OF CONTENTS iv
PREFACE ix
1. INTRODUCTION 1
1.1 Overview 1
1.2 Background 4
1.3 The Multi-Segment Version 6
1.4 The Estuary Model 6
1.5 The PFA Model 7
1.6 AQUATOX Release 3.2 Overview 7
Release 3.1 plus Update 9
Release 3.2 Update 10
1.7 Comparison with Other Models 11
1.8 Intended Application of AQUATOX 12
2. SIMULATION MODELING 14
2.1 Temporal and Spatial Resolution and Numerical Stability 14
2.2 Results Reporting 17
2.3 Input Data 18
2.4 Sensitivity Analysis 19
2.5 Uncertainty Analysis 21
2.6 Calibration and Validation 26
3. PHYSICAL CHARACTERISTICS 43
3.1 Morphometry 43
Volume 43
Bathymetric Approximations 46
Dynamic Mean Depth 49
Habitat Disaggregation 49
3.2 Velocity 50
3.3 Washout 51
3.4 Stratification and Mixing 52
Modeling Reservoirs and Stratification Options 56
3.5 Temperature 57
3.6 Light 58
Hourly Light 60
3.7 Wind 61
3.8 Multi-Segment Model 62
Stratification and the Multi-Segment Model 64
State Variable Movement in the Multi-Segment Model 64
3.9 "Marine" Site Type 65
iv
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
4. BIOTA 67
Guild Organization 68
Anti-Extinction Code 68
4.1 Algae 69
Light Limitation 73
Adaptive Light 78
Nutrient Limitation 79
Internal Nutrients Model 81
Current Limitation 84
Adjustment for Suboptimal Temperature 85
Algal Respiration 87
Photorespiration 88
Algal Mortality 89
Sinking 90
Washout and Sloughing 92
Detrital Accumulation in Periphyton 96
Chlorophyll a 96
Phytoplankton and Zooplankton Residence Time 97
Periphyton-Phytoplankton Link 98
4.2 Macrophytes 99
4.3 Animals 103
Consumption, Defecation, Predation, and Fishing 106
Refuge from Predation 107
Adaptive Food Preferences 112
Respiration 114
Excretion 117
Nonpredatory Mortality 118
Stocking and Harvesting of Animals 119
Suspended Sediment Effects 119
Gamete Loss and Recruitment 135
Washout and Drift 137
Vertical Migration 138
Migration Across Segments 139
Anadromous Migration Model 140
Promotion and Emergence 141
4.4 Oysters 142
4.5 Aquatic Dependent Vertebrates 144
4.6 Steinhaus Similarity Index 145
4.7 Biological Metrics 145
Trophic Level 149
Invertebrate Biotic Indices 150
5. RE MINERALIZATION 152
5.1 Detritus 152
Detrital Formation 156
Colonization 157
Decomposition 159
v
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
Sedimentation 162
5.2 Nitrogen 166
Assimilation 168
Nitrification and Denitrification 169
Ionization of Ammonia 171
Ammonia Toxicity 173
5.3 Phosphorus 174
5.4 Nutrient Mass Balance 176
Variable Stoichiometry 176
Nutrient Loading Variables 177
Nutrient Output Variables 177
Mass Balance of Nutrients 178
5.5 Dissolved Oxygen 185
Diel Oxygen 189
Lethal Effects due to Low Oxygen 190
Non-Lethal Effects due to Low Oxygen 196
5.6 Inorganic Carbon 197
5.7 Modeling Dynamic pH 200
5.8 Modeling Calcium Carbonate Precipitation and Effects 203
6. INORGANIC SEDIMENTS 205
6.1 Sand Silt Clay Model 205
Deposition and Scour of Silt and Clay 207
Scour, Deposition and Transport of Sand 210
Suspended Inorganic Sediments in Standing Water 212
6.2 Multi-Layer Sediment Model 213
Suspended Inorganic Sediments 215
Inorganics in the Sediment Bed 215
Detritus in the Sediment Bed 217
Pore Waters in the Sediment Bed 217
Dissolved Organic Matter within Pore Waters 218
Diffusion within Pore Waters 219
Sediment Interactions 220
7. SEDIMENT DIAGENESIS 223
7.1 Sediment Fluxes 225
7.2 POC 228
7.3 PON 230
7.4 POP 230
7.5 Ammonia 230
7.6 Nitrate 232
7.7 Orthophosphate 233
7.8 Methane 234
7.9 Sulfide 236
7.10 Biogenic Silica 237
7.11 Dissolved Silica 238
vi
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
8. TOXIC ORGANIC CHEMICALS 240
8.1 Ionization 247
8.2 Hydrolysis 248
8.3 Photolysis 250
8.4 Microbial Degradation 252
8.5 Volatilization 253
8.6 Partition Coefficients 256
Detritus 256
Algae 259
Macrophytes 260
Invertebrates 261
Fish 261
8.7 Nonequilibrium Kinetics 262
Sorption and Desorption to Detritus 263
Bioconcentration in Macrophytes and Algae 264
Macrophytes 264
Algae 265
Bioaccumulation in Animals 268
Gill Sorption 268
Dietary Uptake 271
Elimination 272
Bioaccumulation Factor 276
Linkages to Detrital Compartments 277
8.8 Alternative Uptake Model: Entering BCFs, Kl, and K2 277
8.9 Half-Life Calculation, DT50 and DT95 278
8.10 Chemical Sorption to Sediments 279
8.11 Chemicals in Pore Waters 281
8.12 Mass Balance Capabilities and Testing 283
8.13 Perfluoroalkylated Surfactants Submodel 285
Sorption 285
Biotransformation and Other Fate Processes 285
Bioaccumulation 285
Gill Uptake 286
Dietary Assimilation 287
Depuration 288
Bioconcentration Factors 289
8.14 Aggregation of Organic Chemicals 290
9. ECOTOXICOLOGY 292
9.1 Lethal Toxicity of Compounds 292
Interspecies Correlation Estimates (ICE) 292
Internal Calculations 294
9.2 Sublethal Toxicity 297
9.3 External Toxicity 300
10. ESTUARINE SUBMODEL 303
10.1 Estuarine Stratification 303
vii
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
10.2 Tidal Amplitude 304
10.3 Water Balance 305
10.4 Estuarine Exchange 306
10.5 Salinity Effects 307
Mortality and Gamete Loss 307
Other Biotic Processes 307
Sinking 308
Sorption 310
Volatilization 310
Reaeration 310
Migration 312
10.6 Nutrient Inputs to Lower Layer 312
11. REFERENCES 313
APPENDIX A. GLOSSARY OF TERMS 334
APPENDIX B. USER SUPPLIED PARAMETERS AND DATA 337
viii
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
PREFACE
The Clean Water Act- formally the Federal Water Pollution Control Act Amendments of 1972
(Public Law 92-50), and subsequent amendments in 1977, 1979, 1980, 1981, 1983, and 1987-
calls for the identification, control, and prevention of pollution of the nation's waters. Data
submitted by the States to the U.S. Environmental Protection Agency's WATERS (Watershed
Assessment, Tracking & Environmental ResultS) database (http://www.epa. gov/waters/)
indicate that a very high percentage of the Nations waters continue to be impaired. As of early
2009, of the waters that have been assessed, 44% of rivers and streams, 59% of lakes, reservoirs
and ponds, and 35% of estuaries were impaired for one or more of their designated uses. The
five most commonly reported causes of impairment in rivers and streams were: pathogens,
sediment, nutrients, habitat alteration and organic enrichment/dissolved oxygen depletion. In
lakes and reservoirs the five most common causes were mercury, nutrients, organic
enrichment/dissolved oxygen depletion, metals, and turbidity. In estuaries the five most
common causes were pathogens, mercury, organic enrichment/oxygen depletion, pesticides and
toxic organics. Many waters are impaired for multiple uses, by multiple causes, from multiple
sources.
New approaches and tools, including appropriate technical guidance documents, are needed to
facilitate ecosystem analyses of watersheds as required by the Clean Water Act. In particular,
there is a pressing need for refinement and release of an ecological risk methodology that
addresses the direct, indirect, and synergistic effects of nutrients, metals, toxic organic
chemicals, and non-chemical stressors on aquatic ecosystems, including streams, rivers, lakes,
and estuaries.
The ecosystem model AQUATOX is one of the few general ecological risk models that
represents the combined environmental fate and effects of toxic chemicals. The model also
represents conventional pollutants, such as nutrients and sediments, and considers several trophic
levels, including attached and planktonic algae, submerged aquatic vegetation, several types of
invertebrates, and several types of fish. It has been implemented for experimental tanks, ponds
and pond enclosures, streams, small rivers, linked river segments, lakes, reservoirs, linked
reservoir segments, and estuaries.
IX
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 1
1. INTRODUCTION
1.1 Overview
The AQUATOX model is an open-source1 general ecological risk assessment model that
represents the combined environmental fate and effects of conventional pollutants, such as
nutrients and sediments, and toxic chemicals in aquatic ecosystems. It considers several trophic
levels, including attached and planktonic algae and submerged aquatic vegetation, invertebrates,
and forage, bottom-feeding, and game fish; it also represents associated organic toxicants
(Figure 1). It can be implemented as a simple model (indeed, it has been used to simulate an
abiotic flask) or as a truly complex food-web model. Often it is desirable to model a food web
rather than a food chain, for example to examine the possibility of less tolerant organisms being
replaced by more tolerant organisms as environmental perturbations occur. "Food web models
provide a means for validation because they mechanistically describe the bioaccumulation
process and ascribe causality to observed relationships between biota and sediment or water"
(Connolly and Glaser 1998). The best way to accurately assess bioaccumulation is to use more
complex models, but only if the data needs of the models can be met and there is sufficient time
(Pelka 1998).
It has been implemented for experimental tanks, ponds and pond enclosures, streams, small
rivers, linked river segments, lakes, reservoirs, linked reservoir segments, and estuaries. It is
intended to be used to evaluate the likelihood of past, present, and future adverse effects from
various stressors including potentially toxic organic chemicals, nutrients, organic wastes,
sediments, and temperature. The stressors may be considered individually or together.
The fate portion of the model, which is applicable especially to organic toxicants, includes:
partitioning among organisms, suspended and sedimented detritus, suspended and sedimented
inorganic sediments, and water; volatilization; hydrolysis; photolysis; ionization; and microbial
degradation. The effects portion of the model includes: sublethal and lethal toxicity to the
various organisms modeled; and indirect effects such as release of grazing and predation
pressure, increase in detritus and recycling of nutrients from killed organisms, dissolved oxygen
sag due to increased decomposition, and loss of food base for animals.
AQUATOX represents the aquatic ecosystem by simulating the changing concentrations (in
mg/L or g/m3) of organisms, nutrients, chemicals, and sediments in a unit volume of water
(Figure 1). As such, it differs from population models, which represent the changes in numbers
of individuals. As O'Neill et al. (1986) stated, ecosystem models and population models are
complementary; one cannot take the place of the other. Population models excel at modeling
individual species at risk and modeling fishing pressure and other age/size-specific aspects; but
recycling of nutrients, the combined fate and effects of toxic chemicals, and other
interdependencies in the aquatic ecosystem are important aspects that AQUATOX represents and
that cannot be addressed by a population model.
1 To download the AQUATOX Source code go to the "Help" menu and select "About" and click on the "Source
Code" button.
1
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 1
Figure 1. Conceptual model of ecosystem represented by AQUATOX
Nutrients
Organic
toxicant
Partitioning
Oxygen
Death, Ingestion
Organic matter
Suspended and
bedded sedim ents
Settling, scour
Plants
Phytoplankton
Attached algae
M acrophytes
Ingestion
Animals
Invertebrates
Fish
o
*=
D
o.
Any ecosystem model consists of multiple components requiring input data. These are the
abiotic and biotic state variables or compartments being simulated (Figure 2). In AQUATOX
the biotic state variables may represent trophic levels, guilds, and/or species. The model can
represent a food web with both detrital- and algal-based trophic linkages. Closely related are
driving variables, such as temperature, light, and nutrient loadings, which force the system to
behave in certain ways. In AQUATOX state variables and driving variables are treated similarly
in the code. This provides flexibility because external loadings of state variables, such as
phytoplankton carried into a reach from upstream, may function as driving variables; and driving
variables, such as temperature, could be treated as dynamic state variables in a future
implementation. Constant, dynamic, and multiplicative loadings can be specified for
atmospheric, point- and nonpoint sources. Loadings of pollutants can be turned off at the click
of a button to obtain a control simulation for comparison with the perturbed simulation.
2
-------
AQUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 1
Figure 2. State variables in AQUATOX as implemented for Cahaba River, Alabama.
Forage Fish
Bottom Fish
stoneroller
Piscivore
bass
shiner,
bluegill
Zoobenthos
grazers
mayfly,
riffle beetle
Zoobenthos
susp. feeders
caddisfly
Predatory
Invertebrate
crayfish
stonefly
Zoobenthos
molluscs
snail, mussel,
Corbicula
Periphyton diatom,
Phytoplankton
Macrophyte
Periphyton
blue-green
diatom,
green
moss
green
Phosphate
Ammonia
Nitrate & Nitrite
Carbon Dioxide
Oxygen
Labile
Diss. Detritus
Disso ved
Refractory
Susp. Detritus
Labi e
Refractory
Diss. Detritus
Org. Toxicants
Susp. Detritus
atrazine
Labile
Sed. Detritus
Buried Refrac.
Sed. Detritus
Refractory
Sed. Detritus
Sediments
sand, silt, clay
(or TSS)
The model is written in object-oriented Pascal using the Delphi programming system for
Windows. An object is a unit of computer code that can be duplicated; its characteristics and
methods also can be inherited by higher-level objects. For example, the organism object,
including variables such as the LC50 (lethal concentration of a toxicant) and process functions
such as respiration, is inherited by the plant object; that is enhanced by plant-specific variables
and functions and is duplicated for four kinds of algae; and the plant object is inherited and
modified slightly for macrophytes and moss. This modularity forms the basis for the remarkable
flexibility of the model, including the ability to add and delete given state variables interactively.
AQUATOX utilizes differential equations to represent changing values of state variables,
normally with a reporting time step of one day. These equations require starting values or initial
conditions for the beginning of the simulation. If the first day of a simulation is changed, then
the initial conditions may need to be changed. A simulation can begin with any date and may be
for any length of time from a few days, corresponding to a microcosm experiment, to decades,
corresponding to an extreme event followed by long-term recovery.
The process equations contain another class of input variables: the parameters or coefficients
that allow the user to specify key process characteristics. For example, the maximum
consumption rate is a critical parameter characterizing various consumers. AQUATOX is a
3
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 1
mechanistic model with many parameters; however, default values are available so that the
analyst only has to be concerned with those parameters necessary for a specific risk analysis,
such as characterization of a new chemical. In the pages that follow, differential equations for
the state variables will be followed by process equations and parameter definitions.
Finally, the system being modeled is characterized by site constants, such as mean and
maximum depths. At present one can model lakes, reservoirs, streams, small rivers, estuaries,
and ponds- and even enclosures and tanks. The "Generalized Parameter" screen is used for all
these site types, although some, such as the hypolimnion and estuary entries, obviously are not
applicable to all. The temperature and light constants are used for simple forcing functions,
blurring the distinctions between site constants and driving variables.
Table 1. Model Overview Summary (also see Section 2.1)
Category:
Summary:
Notes:
Reporting Time Step
Daily or Hourly
time-step over which equations are solved
Differentiation
Variable time-step Runge Kutta
(with fixed time-step option)
smaller step sizes than the reporting time-
step may be utilized to reduce relative error
Output Averaging
Variable
editable by user
Conceptual Approach
Kinetic; biomass model
no longer a fugacity option for chemicals;
individual organisms are not modeled
Horizontal Spatial
Resolution
Point model, or ID and 2D with
linked segments
modeled units can be a lake, river, reservoir,
stream segment, estuary, or enclosure
Vertical Spatial
Resolution
Vertically stratified water
column when relevant
user-specified or model calculated dates of
stratification
Sediment Bed
Multiple sediment bed options
active layer only, multi-layer sediments,
sediment diagenesis submodels
Boundary Conditions
Inflows and outflows of all state
variables (dissolved oxygen,
nutrients, biota, detritus, and
toxic organics)
water inflow, point sources, nonpoint
sources, direct precipitation, separate
tributary inputs
Ecological Complexity
Variable—user can model
representative groups or
individual species
can model abiotic conditions or single
macrophyte species in a water tank up to
dozens of plant and animal species in a
complex river or reservoir system
Chemical Complexity
Zero to 20 organic chemicals
biotransformation to daughter products
may be modeled
Mass Balance Tracking
For nutrients and chemicals
1.2 Background
AQUATOX 3.2 is an update to AQUATOX Release 3 (see section 1.6 below for a list of updates
in Release 3.2). AQUATOX Release 3 was the result of an effort to combine all of the various
versions of AQUATOX into a single consolidated version. Models that were combined to
produce Release 3 included:
• AQUATOX Multi-Segment version
AQUATOX Estuarine Version
4
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 1
AQUATOX PFA Model (Perfloroalkylated Surfactants)
Each of these versions is discussed in a separate section below.
AQUATOX is the latest in a long series of models, starting with the aquatic ecosystem model
CLEAN (Park et al., 1974) and subsequently improved in consultation with numerous
researchers at various European hydrobiological laboratories, resulting in the CLEANER series
(Park et al., 1975, 1979, 1980; Park, 1978; Scavia and Park, 1976) and LAKETRACE (Collins
and Park, 1989). The MACROPHYTE model, developed for the U.S. Army Corps of Engineers
(Collins et al., 1985), provided additional capability for representing submersed aquatic
vegetation. Another series started with the toxic fate model PEST, developed to complement
CLEANER (Park et al., 1980, 1982), and continued with the TOXTRACE model (Park, 1984)
and the spreadsheet equilibrium fugacity PART model. AQUATOX combined algorithms from
these models with ecotoxicological constructs; and additional code was written as required for a
truly integrative fate and effects model (Park, 1990, 1993). The model was then restructured and
linked to Microsoft Windows interfaces to provide greater flexibility, capacity for additional
compartments, and user friendliness (Park et al., 1995). The current version has been improved
with the addition of constructs for sublethal effects and uncertainty analysis, making it a
powerful tool for probabilistic risk assessment.
This technical documentation is intended to provide verification of individual constructs or
mathematical and programming formulations used within AQUATOX. The scientific basis of
the constructs reflects empirical and theoretical support; precedence in the open literature and in
widely used models is noted. Units are given to confirm the dimensional analysis. The
mathematical formulations have been programmed and graphed in spreadsheets and the results
have been evaluated in terms of behavior consistent with our understanding of ecosystem
response; many of those graphs are given in the following documentation. The variable names in
the documentation correspond to those used in the program so that the mathematical
formulations and code can be compared, and the computer code has been checked for
consistency with those formulations. Much of this has been done as part of the continuing
process of internal review. Releases 2 and 3 of the AQUATOX model and documentation have
undergone successful peer reviews by external panels convened by the U.S. Environmental
Protection Agency. Release 3 has also been described in the peer-reviewed literature (Park et al.
2008).
Release 3 has significant additional capabilities compared to Release 2.2:
The full source code is now available. To download the open-source codebase use the
"Source Code" button in the model's "About" window.
Link to WEB-ICE (Interspecies Correlation Estimates) database and graphics
Sediment diagenesis based on the Di Toro model
Optional hourly time step with diel oxygen, light, and photosynthesis;
Low oxygen effects
Toxicity due to ammonia
Suspended and bedded sediment effects on organisms; % embeddedness
5
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 1
Calcium carbonate precipitation and removal of phosphorus
Adaptive light limitation for plants
Linked periphyton and phytoplankton compartments
Conversions for many units in input screens
User-specified seasonally varying thermocline depth
User-specified reaeration constant in addition to alternative estimation procedures
Improved CBOD to organic matter estimation
Estuarine reaeration incorporating salinity
Sensitivity analysis with tornado diagrams
Correlation of variables in uncertainty analysis
Sediment oxygen demand (SOD) output
Enhanced graphics including log plot, duration and exceedance graphs, and threshold
analysis
Option to export all graphs to Microsoft Word
Output of statistics for all graphed model results
Output of trophic state indices and ecosystem bioenergetics such as gross primary
productivity and community respiration
Integrated user's manual and context-sensitive help files
1.3 The Multi-Segment Version
The AQUATOX Multi-Segment version was developed and applied for the EPA Office of Water
in support of the Modeling Study of PCB Contamination in the Housatonic River. Capabilities
introduced with this version include the linkage of individual AQUATOX segments into a single
simulation. Segments can be linked together in a manner that allows feedback into the upstream
segment or a one-way "cascade" linkage can be created. More information about the physical
characteristics of linked segments may be found in Section 3.8 of this document.
Additionally, a sediment submodel was added to the AQUATOX model to enable tracing the
passage of toxicants within a multi-layered sediment bed. Specifications for this multi-layer
sediment model may be found in section 6.2 of this document.
1.4 The Estuarine Submodel
The Risk Assessment Division (RAD), EPA Office of Pollution Prevention and Toxics, is
responsible for assessing the human health and ecological risks of new and existing chemicals
that are regulated under the Toxic Substances Control Act (TSCA). RAD has partially funded
AQUATOX from its initial conceptualization. Many of the industrial chemicals regulated under
TSCA are discharged into estuarine environments.
6
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 1
Therefore, AQUATOX's capabilities were enhanced by adding salinity and other components
(including shore birds) that would be needed to simulate an estuarine environment. The estuarine
version of AQUATOX is intended to be an exploratory model for evaluating the possible fate
and effects of toxic chemicals and other pollutants in estuarine ecosystems. The model is not
intended to represent detailed, spatially varying site-specific conditions, but rather to be used in
representing the potential behavior of chemicals under average conditions. Therefore, it is best
used as a screening-level model applicable to data-poor evaluations in estuarine ecosystems.
Complete documentation for the AQUATOX estuarine submodel may be found in Chapter 10 of
this document.
1.5 The PFA Submodel
The bioaccumulation and effects of a group of chemicals known as perfluorinated surfactants has
been of recent interest. There are two major types of perfluorinated surfactants: perfluoro-
alkanesulfonates and perfluorocarboxylates. The perfluorinated compounds of interest as
bioaccumulators are the perfluorinated acids (PFAs). Perfluoroctane sulfonate (PFOS) belongs
to the sulfonate group and perfluorooctanoic acid (PFOA) belongs to the carboxylate group.
These persistent chemicals have been found in humans, fish, birds, and marine and terrestrial
mammals throughout the world. PFOS has an especially high bioconcentration factor in fish. The
principal focus was on PFOS because of its prevalence and the availability of data. Because both
chemical classes contain high- and low-chain homologs, AQUATOX will be useful in estimating
the fate and effects of a wide range of molecular weight components where actual data are not
available for every homolog.
Complete documentation for the AQUATOX Perfloroalkylated Surfactants model may be found
in Section 8.13 of this document.
1.6 AQUATOX Release 3.2 Overview
Additional capabilities are available in Release 3.2 as compared to Release 3. Some highlights
follow:
Addition of sediment-diagenesis "steady-state" mode to significantly increase model
speed;
Modification of denitrification code in order to simplify calibration and to achieve
alignment with other models;
Enabled importation of equilibrium CO2 concentrations to enable linkage to C02SYS
and similar models;
• New CBOD to organic matter conversion relying on percent-refractory detritus input;
Input and output BOD is clarified to be "carbonaceous" BOD.
Floating plants refinements
Added floating option for plants other than cyanobacteria (formerly known as
"blue-green algae)
7
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 1
Converted the averaging depth for floating plants to the top three meters to more
closely correspond to monitoring data
• Floating plants now explicitly move from the hypolimnion to the epilimnion when
a system is stratified.
Modifications to PFA (perfluoroalkylated surfactants) model to increase flexibility:
Uptake rates (Kls) and elimination rates (K2s) are visible and editable for animals
and plants
• New interface to estimate animal Kls and K2s as a function of chain length
Improved gill-uptake equation for invertebrates.
Bioaccumulation and toxicity modeling improvements:
Optional alternative elimination-rate estimation for animals based on Barber
(2003);
Updated ICE (toxicity regressions), based on new EPA models released in
February 2010 and improved AQUATOX ICE interface;
Addition of output of Kl, K2, and BCF estimates.
Improved sensitivity and uncertainty analyses
"Output to CSV" option for uncertainty runs so that complete results for every
iteration may be examined;
Option for non-random sampling for "statistical sensitivity analyses";
A "reverse tornado" diagram (effects diagram) that shows the effects of each
parameter change on the overall simulation;
• Nominal range sensitivity analysis has been added for linked segment
applications.
Database Improvements
AQUATOX database search functions dramatically improved.
"Scientific Name" field added to Animal and Plant databases.
Interface and Data Input Improvements
Software and software installer is 64-bit OS compatible;
Added an option in the "Setup" screen to trigger nitrogen fixation based on the N
to P ratio.
Addition of output variables to clarify whether photosynthesis is sub-optimal due
to high-light or low-light conditions.
Time-varying evaporation option in the "Site" screen with linkage from the
"Water Volume" screen
Grid mode within a study so that all animal, plant, and chemical parameters in a
study can be examined, edited, and then exported to Excel
Added capability to input time-series loads of organisms based on fish stocking
Updated HSPF WDM file linkage to be more generally applicable (does not
require use of WinHSPF).
Enabled hourly loadings for the following variables: all nutrients, CO2, Oxygen,
Inorganic suspended sediments (sand/silt/clay), TSS, Light, Organic Matter
8
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 1
"Graph Setup" window now enabled for linked-mode graphics.
• Other minor interface improvements.
Documentation for each of these enhancements may be found in this technical documentation
volume or in the User's Manual.
Release 3.1 plus Update
In 2014, EPA Release 3.1 plus was released with several additions to the model. Most
importantly, the option to model nutrient limitation in plants based on internal rather than
external nutrients was added. The internal concentration of P and N in each plant is tracked with
a separate state variable (See Internal Nutrients in section 4.1 of this document). This allows for
luxury uptake of nutrients during high-nutrient periods and expenditure of nutrient stores during
lower-nutrient periods. Concentrations of internal nutrients and derivative rates may be output
from these new state variables as well as the nutrient-to-organism ratio for each plant. Internal-
nutrient simulations maintain nutrient mass balance throughout.
Other enhancements are listed below:
• The capability to load and save observed data sets to a file, to move these data from one
study to another, was added to the interface.
• Improving the use of cached loading and saving of data has significantly optimized
loading and saving of large "aps" or "als" files.
• In non-estuary segments, the sinking of plants and suspended detritus is now affected by
salinity and the density of water when relevant.
• New outputs are produced for net-primary productivity, pelagic-invertebrate biomass,
benthic- invertebrate biomass, and fish biomass.
• The default line-thickness on graphs is now 2 pixels to improve visibility.
• The maximum respiration rate and maximum consumption rate fields are now auto-
calculated when allometric consumption or allometric respiration models are utilized.
This helps to ensure parameters are producing a reasonable allometric model.
• Low-light limitation and high-light limitation (part of plant "rates" outputs) now show
1.0 if they are not the limiting factor, which is more intuitive than 0.0.
• "Gameteloss" is limited so that it is not allowed to exceed the total percent gametes of the
organism.
• Moving waters (rivers and streams) are not predicted to ice over until their average water
temperature drops below 0 degrees centigrade.
• A macrophyte mortality calculation bug was fixed; the model was not adding the
"mortality coefficient" effect properly
• A few other minor interface glitches have been fixed to improve model usability.
9
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 1
Release 3.2 Update
In 2017, EPA Release 3.2 was produced. This version replaces the obsolete paradox database
management system (used in previous versions) with SQLite databases. In addition, the model
now can write all model inputs and outputs in ASCII format by saving files with a "*.txt"
extension. This allows users to view and change model inputs without using the graphical user
interface. Another model upgrade is a command-line version, which allows users to execute and
manipulate the model using a DOS command prompt. For information about how to use each of
these updates, please see the Release 3.2 User'sManual.
In addition to the interface changes discussed above, changes were made to the model to
represent the "nearshore marine environment" as discussed here.
Nearshore Marine Environment
AQUATOX Release 3.2 was designed to extend the existing AQUATOX estuarine version to
include improved capabilities for situations encountered in the nearshore marine environment.
Several changes were required to model food webs in the marine environment. The most notable
updates include:
• Additional equations to model the physical complexity of oyster reefs and the marsh-edge
environment;
• The capability to model size-classes of oysters and crabs within the model
• New invertebrate-modeling capabilities including allometric bioenergetics equations and
burrowing refuge from predation; and,
• As discussed in Chapter 4, to better represent marine-biology conventions, the guilds
used by AQUATOX to characterize these state variables were reorganized.
Differences from Release 3.1 plus
Most model simulations created in Release 3.1 and Release 3.1 plus produce identical or nearly-
identical results in Release 3.2. A few differences are visible in some studies, however. Nutrient
quantities may be slightly different because of a change in the animal-respiration equation (100).
This can have some ripple effects, especially in systems with long retention times. Other
changes to the process code that could affect model results follow:
• The BCF equation (1) has been changed to take into account metabolism of organic
chemicals.
• A bug was fixed when the user has selected the "Calculate BCF" option Alternative
Uptake Mode. The wrong equation was being used to calculate BCF in this seldom-used
model option. See section 8.8 for more information.
• Carrying capacity was not utilized as a parameter for benthic invertebrates previously,
but now is considered a hard cap based on habitat limitations. See equation (109b).
10
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 1
1.7 Comparison with Other Models
The following comparison is taken from Park et al. (2008):
The model is perhaps the most comprehensive aquatic ecosystem simulation model available, as
can be seen by comparison with other representative dynamic models being used for risk
assessment (Table 2). All the models, with the exception of QSim and CASM, are public
domain. The closest to AQUATOX in terms of scope is the family of CATS models developed
by Traas and others (Traas et al., 1996; Traas et al., 1998; Traas et al., 2001); these
ecotoxicology models have simple representations of growth and are not as suitable as
AQUATOX for detailed analyses of eutrophication effects. CASM (DeAngelis et al., 1989;
Bartell et al., 1999) is similar to CATS, with simplified growth terms, but it lacks a toxicant fate
component. QUAL2K (Chapra et al., 2007) and WASP (Di Toro et al., 1983; Wool et al., 2004)
are water quality models that share many functions with AQUATOX, including benthic algae
(Martin et al., 2006); WASP also models fate of toxicants. The hydraulic and water quality
models EFDC (Tetra Tech Inc., 2002) and HEM3D (Park et al., 1995a) are often combined;
EFDC has also been used to provide the flow field for linked segments in AQUATOX, resulting
in a similar representation. AQUATOX, QUAL2K, WASP, and EFDC include the sediment
diagenesis model for remineralization (Di Toro, 2001). WASP and the bioaccumulation model
QEAFdChn (Quantitative Environmental Analysis, 2001) have been combined in the Green Bay
Mass Balance (GBMB) study (U.S. Environmental Protection Agency, 1989), which Koelmans
et al. (2001) considered to be more accurate for portraying bioaccumulation than AQUATOX.
However, GBMB does not include an ecotoxicology component. BASS (Barber, 2001) is a very
detailed bioaccumulation and ecotoxicology model; it provides better resolution than
AQUATOX in modeling single species, but so far it has only been applied to fish and does not
include ecosystem dynamics. The German model QSim (Schol et al., 1999; Schol et al., 2002;
see also Rode et al., 2007) has detailed ecosystem functions and has been applied in studying
impacts of both eutrophication and hydraulics on river ecosystems. Similar to AQUATOX, it
has been used to analyze relationships between plankton and mussels and impacts of oxygen
depletion. Further comparison of models can be found in a book by Pastorok et al. (2002).
11
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER I
Table 2. Comparison of AQUATOX with other representative dynamic models used for risk
assessment (Park et al. 2008).
State variables and
processes
AQUATOX
CATS
CASM
Qual2K
WASP7
EFDC-HEM3D
QEAFdChn
BASS
QSim
Nutrients
X
X
X
X
X
X
X
Sediment diagenesis
X
X
X
X
Detritus
X
X
X
X
X
X
X
Oinohred oxygen
X
X
X
X
X
X
DO effects on biota
X
X
pH
X
X
X
NH4 toxicity
X
sand/silt/clay
X
X
X
sediment effects
X
Hydraulics
X
X
Heat budget
X
X
X
X
Salinity
X
X
X
Pbytoplankton
X
X
X
X
X
X
X
Periphyton
X
X
X
X
X
X
Macropbytes
X
X
X
X
Zooplanlcton
X
X
X
X
Zoobenthos
X
X
X
X
Fish
X
X
X
X
X
Bacteria
X
X
Pathogens
X
X
Organic toxicant fate
X
X
X
X
Organic toxicants in
Sediments
X
X
X
X
Stratified sediments
X
X
X
Phytoplankton
X
X
Periphyton
X
X
Macrophytes
X
X
Zooplankton
X
X
X
Zoobenthos
X
X
X
Pish
X
X
X
X
Birds or other animals
X
X
Ecotoxicity
X
X
X
X
Linked segments
X
X
X
X
X
X
1.8 Intended Application of AQUATOX
AQUATOX is intended to be used at any one of several levels of application. Like any model, it
is best used as one of several tools in a weight-of-evidence approach. The level of required
precision, rigor, data requirements and user effort depend upon the goals of the modeling
exercise and the potential consequences of the model results.
Perhaps its most widespread use is as a screening-level model requiring few changes to default
studies and parameters. In fact, it was originally developed as an evaluative model to assess the
fate and effects of pesticides and industrial organic chemicals in representative or "canonical"
environments; these include ponds and pond enclosures, experimental streams, and a
representative estuary. It is especially useful in taking the place of expensive, labor-intensive
mesocosm tests. It has been calibrated and validated with data from pond enclosures,
experimental streams, and a polluted harbor. In one early application, AQUATOX was driven
with predicted pesticide runoff into a farm pond adjacent to a corn field using the field model
PRZM. Also, with little effort the model can provide insights into the potential impacts of
invasive species and the possible effects of control measures, such as pesticide application, on
the aquatic ecosystem.
In recent years AQUATOX has been applied as part of the process of developing water quality
12
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 1
targets for nutrients, and comparing model-derived values with regional criteria developed
empirically. This application has involved setting up the model and calibrating with available
data for rivers and reservoirs receiving nutrients from wastewater treatment plants, agricultural
runoff, and background "natural" loadings. It has been our experience that this entails a
substantial level of effort, especially if the system is spatially heterogeneous, which then requires
application of linked segments. A certain amount of site-specific biotic, water quality and flow
data is required, as well as pollutant loading data, for calibration. However, once the model is
set up and calibrated for a site, it is relatively easy to represent a series of loading scenarios and
determine threshold nutrient levels for deleterious impacts such as nuisance algal blooms and
anoxia. This process is facilitated by the fact that the model has been calibrated across nutrient,
turbidity, and discharge gradients, resulting in robust parameter sets that span these conditions.
This is important because the intent of setting water quality targets is to model ecological
communities under changing conditions as a result of environmental management decisions; this
would give better assurance that the sometimes costly nutrient reduction actions would render
the desired environmental result.
The most intensive, time-consuming application of AQUATOX is in environmental remediation
projects, such as SUPERFUND. Because of the likely litigation and the potential for costly
remediation, this level of application requires site-specific calibration and validation using
quality-assured data collected specifically for the model. In dynamic systems, linkage to an
equally well calibrated and validated hydrodynamic model is essential to represent, for example,
burial and exhumation of contaminated sediments. Several of the more powerful features of the
model, such as the linked segments and IPX layered-sediment submodel, were developed for this
type of application. Unfortunately, the one remediation application performed by the model
developers cannot be published because of continuing litigation.
13
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
2. SIMULATION MODELING
2.1 Temporal and Spatial Resolution and Numerical Stability
AQUATOX Release 3 is designed to be a general, realistic
model of the fate and effects of pollutants in aquatic
ecosystems. In order to be fast, easy to use, and verifiable, it
was originally designed with the simplest spatial and
temporal resolutions consistent with this objective. Release 3
may still be run as a non-dimensional point model. However,
unlike previous versions of AQUATOX, in Release 3 spatial
segments may be linked together to form a two- or three-
dimensional model if a more complicated spatial resolution is
desired.
The model generally represents average daily conditions for a well-mixed aquatic system. Each
segment in a multi-dimensional run is also assumed to be well-mixed in each time-step.
AQUATOX also represents one-dimensional vertical epilimnetic and hypolimnetic conditions
for those systems that exhibit stratification on a seasonal basis. Multi-segment systems also can
be set up with vertical stratification. Furthermore, the effects of run, riffle, and pool
environments can be represented for streams. Results may be plotted in the AQUATOX output
screen with the capability to import observed data to examine against model predictions.
While the model is generally run with a daily maximum time-step, the temporal resolution of the
model can also be reduced to an hourly maximum time-step. This capability was added so that
AQUATOX can represent diel oxygen. See sections 3.6 and 5.5 for more information on how
this choice of hourly time-step affects AQUATOX equations. The reporting step can be as long
as several years or as short as one hour; results are integrated to obtain the desired reporting time
period.
According to Ford and Thornton (1979), a one-dimensional model is appropriate for reservoirs
that are between 0.5 and 10 km in length; if larger, then a two-dimensional model disaggregated
along the long axis is indicated. The one-dimensional assumption is also appropriate for many
lakes (Stefan and Fang, 1994). Similarly, one can consider a single reach or stretch of river at a
time.
Usually the reporting time step is one day, but numerical instability is avoided by allowing the
step size of the integration to vary to achieve a predetermined accuracy in the solution. (This is a
numerical approach, and the step size is not directly related to the temporal scale of the
ecosystem simulation.) AQUATOX uses a very efficient fourth- and fifth-order Runge-Kutta
integration routine with adaptive step size to solve the differential equations (Press et al., 1986,
1992). The routine uses the fifth-order solution to determine the error associated with the fourth-
order solution; it decreases the step size (often to 15 minutes or less) when rapid changes occur
and increases the step size when there are slow changes, such as in winter. However, the step
size is constrained to a maximum of one day (or one hour in hourly simulations) so that short-
term pollutant loadings are always detected. The reporting step, on the other hand, can be as
Simulation Modeling: Simplifying
Assumptions
• Each modeled segment is well-
mixed
• Model is rim with a daily or hourly
maximum time-step.
• Results are trapezoidally integrated
14
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
long as several years or as short as one hour; results are integrated to obtain the desired reporting
time period.
As an alternative, the user may specify an exact step size that is used thoughout the simulation.
This is similar to the way that many models solve the differential equations. The disadvantage is
that the accuracy of the solution may not be maintained. However, it is useful under some
circumstances and is discussed more fully later in this section.
The temporal and spatial resolution is in keeping with the generality and realism of the model
(see Park and Collins, 1982). Careful consideration has been given to the hierarchical nature of
the system. Hierarchy theory tells us that models should have resolutions appropriate to the
objectives; phenomena with temporal and spatial scales that are significantly longer than those of
interest should be treated as constants, and phenomena with much smaller temporal and spatial
scales should be treated as steady-state properties or parameters (Figure 3; also see O'Neill et al.,
1986). AQUATOX uses a longer time step than dynamic hydrologic models that are concerned
with representing short-term phenomena such as storm hydrographs, and it uses a shorter time
step than fate models that may be concerned only with long-term patterns such as
bioaccumulation in large fish.
Figure 3. Position of ecosystem models such as AQUATOX in the spatial-temporal hierarchy of models.
Rule-based habitat models
succession, urbanization, sea-level rise
(
Ecosystem models
Population models
High-resolution
process models
flood hydrograph
V
diurnal pH
Changing the permissible relative error (the difference between the fourth- and fifth-order
solutions) of the simulation can affect the results. The model allows the user to set the relative
error, usually between 0.005 and 0.01. Comparison of output shows that up to a point a smaller
error can yield a marked improvement in the simulation, although execution time is longer. For
example, simulations of two pulsed doses of chlorpyrifos in a pond exhibit a spread in the first
pulse of about 0.6 [j,g/L dissolved toxicant between the simulation with 0.001 relative error and
the simulation with 0.05 relative error (Figure 4); this is probably due in part to differences in the
timing of the reporting step. However, if we examine the dissolved oxygen levels, which
combine the effects of photosynthesis, decomposition, and reaeration, we find that there are
pronounced differences over the entire simulation period. The simulations with 0.001 and 0.01
relative error give almost exactly the same results, suggesting that the more efficient 0.01 relative
error should be used; the simulation with 0.05 relative error exhibits instability in the oxygen
15
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
simulation; and the simulation with 0.1 error gives quite different values for dissolved oxygen
(Figure 5). The observed mean daily maximum dissolved oxygen for that period was 9.2 mg/L
(U.S. Environmental Protection Agency, 1988), which corresponds most closely with the results
of simulation with 0.001 and 0.01 relative error.
Figure 4. Pond with chlorpyrifos in dissolved
phase.
06/19/88 06/30/88 07/12/88
06/24/88 07/06/88 07/18/88
0.001 0.01
0.05
0.1
Figure 5. Same as Figure 4 with Dissolved
Oxygen.
12
11
-10
o
i
1
ft
f V
\
I1 I
. . n A. A /
^ v y v v v v
06/19/88 06/30/88 07/12/88
06/24/88 07/06/88 07/18/88
— 0.001 — 0.01 • 0.05 - 0.1
A common use of AQUATOX is to determine the impact of a perturbation in a perturbed
simulation when compared with a control simulation. For example, the model is often run with
and without a potentially toxic organic chemical, and a percent difference graph is plotted
showing how the two simulations differ. Of particular interest is whether there are likely to be
significant differences in state variables or other environmental indices at very low
concentrations of the chemical. Because a simulation with the toxicant may require a decrease in
step size to capture the dynamics of the fate of the toxicant as opposed to a simulation without
the toxicant, there may be a mismatch in the step sizes of the two simulations, and the
simulations may differ solely on the basis of the difference in numerical resolution. Although
decreasing the relative error may decrease the mismatch, there may still be a difference that
prevents determination of the "no effects" level of the chemical.
In the example that follows, toxicity of PFOS has been turned off by setting all LC50 and EC50
parameter values to 0. In Figure 6 the default variable step size option has been used with a very
small relative error. In Figure 7 the simulation is the same except the constant step size option
has been used, and it is readily apparent that there is no difference between the perturbed and
control runs.
16
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
Figure 6. Percent differences in fish biomass between perturbed simulation with PFOS (toxicity turned
off) and control simulation without PFOS, using variable step size (relative error = 0.0001). The
differences are purely artifacts of the numerical method.
Conasauga River, GA (Difference)
50.0
40 o
30.0
20.0
LU
-30.0
-40.0
-50.0
3/3/2008 6/1/2008 8/30/2008 11/28/2008 2/26/2009
Figure 7. There is no difference in fish biomass between perturbed simulation with PFOS (toxicity turned
off) and control simulation without PFOS using a constant step size (0.1 day).
Conasauga River, GA (Difference)
5.00E+01 j
4.00E+011
3.00E+011
2.00E+011
lu I
z 1.00E+01
LU I
CL I
\H. 0.00E+00
LL I
° -1.00E+01
o"- I
-2.00E+01
-3.00E+01
-4.00E+01
-5.00E+01'
3/3/2008 6/1/2008 8/30/2008 11/28/2008 2/26/2009
2.2 Results Reporting
The AQUATOX results reporting time step may be set to any desired frequency, from a fraction
of an hour to multiple years. The Runge-Kutta differential equations solver produces a series of
results of variable frequency; this frequency may be either greater than or less than the reporting
time-step. To standardize AQUATOX output, the user has two options, the trapezoidal
Stoneroller
Spotted Bass
Catfish
Shiner
Bluegill
Stoneroller
Spotted Bass
Catfish
Shiner
^—Bluegill
17
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
integration of results (default) or the output of "instantaneous" concentrations. Using either of
these options, AQUATOX will produce output with time-stamps that match the reporting time-
step precisely.
When instantaneous concentrations are requested (in the model's setup screen) AQUATOX
returns output precisely at the requested reporting time-step through linear interpolation of the
nearest Runge-Kutta results that occur before and after the relevant reporting time-step.
When results are trapezoidally integrated, AQUATOX calculates results by summing all of the
trapezoids that can be produced by linear interpolation between Runge-Kutta results and dividing
by the results-reporting step-size to get an average result over the reporting step. In Figure 8, for
example, the areas of the four shaded trapezoids are summed together and this sum is divided by
the results reporting step to achieve an average result over that reporting step. When trapezoidal
integration is selected, AQUATOX output is time-stamped at the end of the interval over which
the integration is taking place. For example, if a user selects a 366.25 day time-step, the results
at the end of the first year will be reflective of all time-steps calculated within that year.
figure 8. An example of trapezoidal integration.
Results-
Reporting Step
Runge Kutta
Step Size
(Variable)
Results may be plotted in the AQUATOX output screen including the capability to import
observed data to examine against model predictions.
2.3 Input Data
AQUATOX accepts several forms of input data, a partial list of which follows:
• Point-estimate parameters describing animals, plants, chemicals, sites, and
remineralization. Default values for these parameters are generally available from
included databases (called "libraries"). The full list of these parameters, their units, and
18
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
their manner of reference in the interface, this document, and the source code may be
found in Appendix B of this document.
• Time series (or constant values) for nutrient-inflow, organic matter-inflow, and gas-
inflow loadings.
• Time series for inorganic sediments in water, water volume variables, and the pH, light,
and temperature climates.
• Time series of chemical inflow loadings and initial conditions.
• A feeding preference matrix must be specified to describe the food web in the simulation.
• Additional parameters may be required depending on which submodels are included (e.g.
additional sediment diagenesis parameters.)
• Nearly all point-estimate parameters may be represented by distributions when the model
is run in uncertainty mode (see section 2.5).
For more discussion of AQUATOX data requirements please see the "Data Requirements"
section in the AQUATOX Users Manual (or in the context sensitive help files included with the
model software). Furthermore, a Technical Note on Data Requirements is available.
For time-series loadings, when a value is input for every day of a simulation, AQUATOX will
read the relevant value on each day. If missing values are encountered by the model, a linear
interpolation will be performed between the surrounding dates. If the AQUATOX simulation
time includes dates before or after the input time-series the model assumes an annual cycle and
tries to calculate the appropriate input value accordingly. Please see the "Important Note about
Dynamic Loadings" in the AQUATOX Users Manual (integrated help-file) for a complete
description of this process.
2.4 Sensitivity Analysis
"Sensitivity" refers to the variation in output of a
mathematical model with respect to changes in the values of
the model inputs (Saltelli 2001). It provides a ranking of the
model input assumptions with respect to their relative
contribution to model output variability or uncertainty (U.S.
Environmental Protection Agency 1997).
AQUATOX includes a built-in nominal range sensitivity
analysis (Frey and Patil 2001), which may be used to examine
the sensitivity of multiple model outputs to multiple model
parameters. The user first selects which model parameters to
vary and which output variables to track. The model iteratively steps through each of the
parameters and varies them by a given percent in the positive and negative direction and saves
model results in an Excel file.
A sensitivity statistic may then be calculated such that when a 10% change in the parameter
results in a 10% change in the model result, the sensitivity is calculated as 100%.
Simplifying Assumptions:
• Parameters are treated as
independent
• Feeding preference matrices are not
included
• Sensitivity is compared for the last
step of the simulation
Caution
• 10% change is appropriate, a large
change can exceed reasonable
values and give misleading results
19
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
\ResultPos-ResultBaseKn\ + \Result -ResultBa A iqq
Sensitivity = i :
2 • | Result Baseline | PctChanged
where:
Sensitivity = normalized sensitivity statistic (%);
Result scenario = averaged AQUATOX result for a given endpoint given a positive
change in the input parameter, a negative change in the input
parameter or no change in the input parameter (baseline)
PctChanged = percent that the input parameter is modified in the positive and
negative directions.
Sensitivity is computed for the last time step of the simulation, so one usually sets the reporting
time step to encompass a year or the entire period of the simulation. For each output variable
tracked, model parameters may be sorted on the average sensitivity (for the positive and negative
tests) and plotted on a bar chart. The end result is referred to as a "Tornado Diagram." Tornado
diagrams may automatically be produced within the AQUATOX output window (Figure 9).
When interpreting a tornado diagram, the vertical line at the middle of the diagram represents the
deterministic model result. Red lines represent model results when the given parameter is
reduced by the user-input percentage while blue lines represent a positive change in the
parameter. An "effects diagram" that illustrates the effects of a single parameter change on all
tracked outputs can also be created. See the User's Manual (or context-sensitive help) for more
information on how to create and interpret these types of output.
When sensitivity analysis is run on a multi-segment model, the user must choose either
parameters that are relevant to all segments (e.g. animal or plant parameters) or individual
segments (e.g. state-variable initial conditions, or the segment's physical characteristics). The
segment for which the parameter is relevant may be selected in the sensitivity analysis setup
window (see the User's Manual for more information) Any number of global or segment-
specific parameters may be selected for a single sensitivity-analysis; output files will be written
for each segment in the simulation.
20
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
Figure 9. An example tornado diagram showing calculated sensitivity statistic.
Sensitivity of Chironomid (g/m2 dry) to a 20% change in the 15 most sensitive (tested) parameters
220% - Site: Ave. Epilimnetic Temperature (deg C)
163% - Chironomid: Maximum Temperature (deg. C)
142% - Site: Epi Temp. Range (deg C)
137% - Peri, Green: Max Photosynthetic Rate (1/d)
115% - Temp: Multiply Loading by
102% - Sphaerid: Maximum Temperature (deg. C)
87.2% - Peri, Green: Optimal Temperature (deg. C)
74.8% - Chironomid: Respiration Rate: (1 Id)
68.7% - Mayfly (Baetis: Respiration Rate: (1 / d)
66.7% - Water Vol: Initial Condition (cu.m)
54.2% - Shiner: Optimal Temperature (deg. C)
42.1% - Phyt, Blue-Gre: Max Photosynthetic Rate (1/d)
33.8% - Mayfly (Baetis: Max Consumption (g / g day)
27.9% - Largemouth Ba2: Maximum Temperature (deg. C)
27.2% - Peri High-Nut: Maximum Temperature (deg. C)
=1
1
1.5 2 2.5 3 3.5 4 4.5 5
Chironomid (g/m2 dry)
2.5 Uncertainty Analysis
There are numerous sources of uncertainty and variation in
natural systems. These include: site characteristics such as
water depth, which may vary seasonally and from site to site;
environmental loadings such as water flow, temperature, and
light, which may have a stochastic component; and critical
biotic parameters such as maximum photosynthetic and
consumption rates, which vary among experiments and
representative organisms.
In addition, there are sources of uncertainty and variation with
regard to pollutants, including: pollutant loadings from
runoff, point sources, and atmospheric deposition, which may
vary stochastically from day to day and year to year; physico-chemical characteristics such as
octanol-water partition coefficients and Henry Law constants that cannot be measured easily;
chemodynamic parameters such as microbial degradation, photolysis, and hydrolysis rates,
which may be subject to both measurement errors and indeterminate environmental controls.
Increasingly, environmental analysts and decision makers are requiring probabilistic modeling
approaches so that they can consider the implications of uncertainty in the analyses. AQUATOX
21
Uncertainty Analysis: Strengths
• Use of Latin hypercube sampling
is more efficient than brute-force
Monte Carlo analysis
• Nearly all variables and
parameters may be represented as
distributions
• Variables can be correlated
Simplifying Assumptions:
• Feeding preference matrices are
not included
• Modeled correlations cannot be
perfect (e.g. 1.0) due to limitations
of the Iman & Conover method
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
provides this capability by allowing the user to specify the types of distributions and key
statistics for almost all input variables. Depending on the specific variable and the amount of
available information, any one of several distributions may be most appropriate. A lognormal
distribution is the default for environmental and pollutant loadings. In the uncertainty analysis,
the distributions for constant loadings are sampled daily, providing day-to-day variation within
the limits of the distribution, reflecting the stochastic nature of such loadings. A useful tool in
testing scenarios is the multiplicative loading factor, which can be applied to all loads.
Distributions for dynamic loadings may employ multiplicative factors that are sampled once each
iteration (Figure 10). Normally the multiplicative factor for a loading is set to 1, but, as seen in
the example, under extreme conditions the loading may be ten times as great. In this way the
user could represent unexpected conditions such as pesticides being applied inadvertently just
before each large storm of the season. Loadings usually exhibit a lognormal distribution, and
that is suggested in these applications, unless there is information to the contrary. Figure 11
exhibits the result of such a loading distribution.
Figure 10. Distribution screen for point-source loading of toxicant in water.
Distribution Information
T H20: Mult. Point Source Load by
Distribution Type:
C Triangular
C Uniform
C Normal
<•" Lognormal
Distribution Parameters:
Mean
(• Probability C Cumulative Distribution
Std. Deviation 0.6
In an Uncertainty Run:
(• Use Above Distribution
C Use Point Estimate
X Cancel
Choice of distribution: A sequence of increasingly informative distributions should be
considered for most parameters. If only two values are known and nothing more can be
assumed, the two values may be used as minimum and maximum values for a uniform
distribution (Figure 12); this is often used for parameters where only two values are known. If
minimal information is available but there is reason to accept a particular value as most likely,
perhaps based on calibration, then a triangular distribution may be most suitable (Figure 13).
Note that the minimum and maximum values for the distribution are constraints that have zero
probability of occurrence. If additional data are available indicating both a central tendency and
spread of response, such as parameters for well-studied processes, then a normal distribution
22
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
may be most appropriate (Figure 14). The result of applying such a distribution in a simulation of
Onondaga Lake, New York, is shown in Figure 15, where simulated benthic feeding affects
decomposition and subsequently the predicted hypolimnetic anoxia. Most distributions are
truncated at zero because negative values would have no meaning (Log Kow is one exception).
Figure 11. Sensitivity of bass (g/m2) to variations in loadings of dieldrin in Coralville Lake, Iowa.
Largemouth Ba2 (g/m2
4/17/2009 2:16:20 PM
0.9
0.7
10/24/1969
12/23/1969
2/21/1970
4/22/1970
6/21/1970
8/20/1970
Mean
Minimum
Maximum
Mean - StDev
Mean + StDev
Deterministic
Figure 12. Uniform distribution for Henry's Law Figure 13. Triangular distribution for maximum
constant for esfenvalerate. consumption rate for bass.
23
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
Figure 14. Normal distribution for maximum consumption rate for
the detritivorous invertebrate Tubifex.
Distribution Information
D invert: Max Consumption: (g/gd)
Distribution Type:
( Triangular
r Uniform
(• Normal
P Lognormal
0.04 -
0.01 -
0.00
0.0173
U 4a:i
Distribution Parameters
Mean
Std. Deviation
(* Probability r Cumulative Distribution
In an Uncertainty Run:
c* Use Above Distribution
r Use Point Estimate
X Cancel
Figure 15. Sensitivity of hvpolimnetic oxygen in Lake Onondaga to
variations in maximum consumption rates of detritivores.
14
^12
en 10
£ 8
Z
o 6
>< 4
2
0
01/01/89 09/24/89 06/17/90
05/14/89 02/04/90 10/28/90
— Minimum Mean
— Maximum Deterministic
Efficient sampling from the distributions is obtained with the Latin hypercube method (McKay
et al., 1979; Palisade Corporation, 1991). Depending on how many iterations are chosen for the
analysis, each cumulative distribution is subdivided into that many equal segments. Then a
24
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
uniform random value is chosen within each segment and used in one of the subsequent
simulation runs. For example, the distribution shown in Figure 14 can be sampled as shown in
Figure 16. This method is particularly advantageous because all regions of the distribution,
including the tails, are sampled. A non-random seed can be used for the random number
generator, causing the same sequence of numbers to be picked in successive applications; this is
useful if you want to be able to duplicate the results exactly. The default is twenty iterations,
meaning that twenty simulations will be performed with sampled input values; this should be
considered the minimum number to provide any reliability. The optimal number can be
determined experimentally by noting the number required to obtain convergence of mean
response values for key state variables; in other words, at what point do additional iterations not
result in significant changes in the results? As many variables may be represented by
distributions as desired. Correlations may be imposed using the method of Iman and Conover
(1982). By varying one parameter at a time the sensitivity of the model to individual parameters
can be determined in a more rigorous way than nominal range sensitivity offers. This is done for
key parameters in the following documentation.
Figure 16. Latin hypercube sampling of a cumulative
distribution with a mean of 25 and standard deviation of 8
divided into 5 intervals.
1
0.8
0.6
0.4
0.2
0
0
3.17
1.58
An alternate way of presenting uncertainty is by means of a biomass risk graph, which plots the
probability that biomass will be reduced by a given percentage by the end of the simulation
(Mauriello and Park 2002). In practice, AQUATOX compares the end value with the initial
condition for each state variable, expressing the result as a percent decline:
where:
Declined ,-^Ll ,00
I StartVal J
(2)
Decline
EndVal
percent decline in biomass for a given state variable
value at the end of the simulation for a given state variable (units
depend on state variable);
25
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
StartVal = initial condition for given state variable.
The results from each iteration are sorted and plotted in a cumulative distribution so that the
probability that a particular percent decline will be exceeded can be evaluated (Figure 17). Note
that there are ten points in this example, one for each iteration as the consecutive segments of the
distribution are sampled.
Figure 17. Risk to bass from dieldrin in Coralville Reservoir, Iowa.
—Largemouth Ba2
A
TO
SI
O
60.0
10.0
-50 -40 -30 -20 -10 0 10
Percent Decline at Simulation End
Biomass Risk Graph
4/17/2009 2:17:42 PM
100.0
Uncertainty analysis can also be used to perform statistical sensitivity analysis, which is much
more powerful than the screening-level nominal range sensitivity analysis. Parameters are tested
one at a time using the most appropriate distribution of observed parameter values. The time-
varying and mean coefficient of variation can be calculated in an exported Excel file using the
mean and standard deviation results for a particular endpoint. Examples will be published in a
separate report.
2.6 Calibration and Validation
Rykiel (1996) defines calibration as "the estimation and adjustment of model parameters and
constants to improve the agreement between model output and a data set" while "validation is a
demonstration that a model within its domain of applicability possesses a satisfactory range of
accuracy consistent with the intended application of the model." A related process is
verification, which is "a subjective assessment of the behavior of the model" (Jorgensen 1986).
The terms are used in those ways in our applications of AQUATOX.
Endpoints for comparison of model results and data should utilize available data for various
ecosystem components, preferably covering nutrients, dissolved oxygen, and different trophic
26
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
levels, and toxic organics if they are being modeled. Although AQUATOX models a complete
food web, often the only biotic data available are chlorophyll a values. The model converts
biomass predictions to chlorophyll a values to facilitate comparison. Likewise, Secchi depth is
computed from the overall extinction coefficient for comparison with observed data.
Verification should consider process rates to confirm that the results were obtained for the
correct reasons (Wlosinski and Collins 1985). Rate information that can be assessed for
reasonableness and compared with observations includes sediment oxygen demand (SOD), the
fluxes of phosphorus, nitrogen, and dissolved oxygen, and all biotic process rates. These can be
presented in tabular and graphical form in AQUATOX.
There are several measures of model performance that can be used for both calibrations and
validations (Bartell et al. 1992, Schnoor 1996). The primary difficulty is in comparing general
model behavior over long periods to observed data from a few points in time with poorly defined
sample variability. Recognizing that evaluation is limited by the quantity and quality of data,
stringent measures of goodness of fit are often inappropriate; therefore, we follow a weight-of-
evidence approach with a sequence of increasingly rigorous tests to evaluate performance and
build confidence in the model results:
Reasonable behavior as demonstrated by time plots of key variables—is the model
behavior reasonable based on general experience? Are the end conditions similar to the
initial conditions? This is highly subjective, but when observed data are lacking or are
sparse and restricted to short time periods it provides a limited reality check (Figure 18,
Figure 19).
Visual inspections of data points compared to model plots—do the observations and
predictions exhibit a reasonable concordance of values (Figure 20, Figure 21)? Visual
inspection can also take into consideration if there is concordance given a slight shift in
time.
Do model curves fall within the error bands of observed data (Figure 22)? Alternatively,
if there are limited replicates, how do the model curves compare with the spread of
observed data?
Do point observations fall within predicted model bounds obtained through uncertainty
analysis? This has the limitation of being dependent on the precision of the model; the
greater the model uncertainty, the greater the possibility of the data being encompassed
by the error bounds (Figure 23).
Regression of paired data and model results—does the model produce results that are free
of systematic bias? What is the correlation (R )? See Figure 24, which corresponds to
the results shown in Figure 20.
Overlap between data and model distributions based on relative bias (rB) in combination
with the ratio of variances (F)—how much overlap is there (Figure 25)? Relative bias is
a robust measure of how well central tendencies of predicted and observed results
correspond; a value of 0 indicates that the means are the same (Bartell et al. 1992). The F
test is the ratio of the variance of the model and the variance of the data. A value of 1
indicates that the variances are the same.
27
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
Do the observed and predicted values differ significantly based on their cumulative
distributions (Figure 26)? The Kolmogorov-Smirnov statistic, a non-parametric test, can
be used; however, the two datasets should represent the same time periods (for example,
one should not compare predicted values over a year with observed values taken only
during spring and summer).
Figure 18. Predicted biomass patterns for animals in a hypothetical farm pond in Missouri.
Daphnia (mg/L dry)
Mayfly (Baetis (g/m2dry)
Gastropod (g/m2 dry)
Shiner (g/m2 dry)
Largemouth Bas (g/m2 dry)
Largemouth Ba2 (g/m2 dry)
5/16/1994 8/14/1994 11/12/1994 2/10/1995
FARM POND, ESFENVAL (CONTROL)
Run on 01-25-09 10:48 AM
Figure 19. Sediment oxygen demand predicted for Lake Onondaga, using Di Toro sediment diagenesis
option; this is an example of using rates for a reality check.
1/12/1989 5/12/1989 9/9/1989 1/7/1990 5/7/1990 9/4/1990
| — SOD(gQ2/m2d)|
1/2/1991
ONONDAGA LAKE, NY (CONTROL) Run on 06-20-08 2:47 PM
(Hypolimnion Segment)
28
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
Figure 20. Comparison of predicted and observed (Oliver and Niemi 1988) PCB congener
bioaccumulation factors in Lake Ontario lake trout.
Lake Trout
11
10
9
< 8
CO
9 7
6
5
4
Pred/Obs = 0.97 +/-1.03
Observed
Predicted
5 6 7 8
Log KOW
9
Figure 21. Predicted biomass and observed numbers of chironomid larvae in a
Duluth, Minnesota, pond dosed with 6 |xg/L chlorpyrifos.
0.40
0.36
0.32
0.28
^0.24
"O
cm 0.20
CD
0.16
0.12
0.08
0.04
0.00
CHLORPYRIFOS 6 ug/L (PERTURBED)
Run on 01-25-09 3:09 PM
1000
-900
-800
700
-600
500
-400
-300
-200
-100
¦ Chironomid (g/m2dry)
Obs. Chironomids (no./sample)
6/27/1986
7/27/1986
8/26/1986
29
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
Figure 22. Predicted and observed benthic chlorophyll a in Cahaba River, Alabama;
bars indicate one standard deviation in observed data.
T= 100
B
1/1/01
7/20/01
2/5/02
8/24/02
Figure 23. Visual comparison of the envelope of model uncertainty, using two standard deviations for
each of the nutrient loading distributions, with the observed data for chlorophyll a in Lake Onondaga,
NY.
120
-Min Chloroph (ug/L)
Mean Chloroph (ug/L)
-Max Chloroph (ug/L)
Det Chloroph (ug/L)
-Obs Chi
30
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
Figure 24. Regression shows that the correlation between predicted and observed (Oliver and Niemi
1988) PCB congener bioaccumulation factors in Lake Ontario trout may be very good, but the slope
indicates that there is systematic bias in the relationship. See Figure 20 for another presentation of these
same results.
LAKE ONTARIO TROUT
10
9
8
7
6
5
5
6
7
8
9
10
Obs Log BAF
31
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
Figure 25. Relative bias and F test to compare means and variances of observed data and predicted
results with AQUATOX. The isopleths correspond to the probability that the distributions of predicted
and observed, as defined by the combination of the rB and F statistics, are similar. The isopleths assume
normal distributions.
Statistical Comparison of Means and Variances
0.5
10.05-
(Pred - Ob s
obs
rB = 0.242, F = 0.400
predicted and observed
distributionsare similar
06
a«
p S pred
S2„bs
0.2
-6
-2
¦4
2
4
Figure 26. Comparison of predicted and observed chlorophyll a in Lake Onondaga, New
York (U.S. Environmental Protection Agency 2000). The Kolmogorov-Smirnov p statistic =
0.319, indicates that the distributions are not significantly different.
100 :
, '
90 ¦;
2? 80 J
a> 70 i
/ /
60 :
>/
^5 50 J
jj
| 40 ¦:
I
^ 30
° 20 ¦:
10 ^
t/ 1
o i
| I I I I I I I I I | I I I I I I 1 1 1 | 1 1 1 I 1 I I I I | I I I I I I I I I | 1 I 1 1 1 1 1 1 1 | 1 1 1 1 1 1 1 1 I |
0 20 40 60 80 100 120
Chlorophyll a (ug/L)
Observed Predicted
32
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
Data are often too sparse for adequate calibration at a given site. However, AQUATOX can be
calibrated simultaneously across sites using an expanded state variable list representative of a
range of conditions and using the same parameter set. In this way the observed biotic data can be
pooled and the resulting state variable and parameter sets, being applicable to diverse sites, are
assured to be robust. This is an approach that we have used on the Cahaba River, Alabama (Park
et al. 2002); on three dissimilar rivers in Minnesota (Park et al. 2005); and on 13 diverse reaches
on the Lower Boise River Idaho (CH2M HILL et al. 2008). The Minnesota rivers application is
discussed below.
Time series of driving variables for the Minnesota rivers were obtained from several sources
with varying degrees of resolution and reliability. Results of watershed simulations with HSPF
(Hydrologic Simulation Program- Fortran, a watershed loading model) were linked to
AQUATOX, providing boundary conditions (site constants and drivers) for the Blue Earth and
Crow Wing Rivers (Donigian et al. 2005). HSPF was not run for the Rum River; however, a
U.S. Geological Survey (USGS) gage is located at the sample site and both daily discharge and
sporadic water quality data were available from the USGS Web pages (search on "National
Water Information System"). AQUATOX interpolates between points, and this feature was used
to compute daily time series of nutrient concentrations from USGS National Water Information
System (NWIS) observed data. Total suspended solids (TSS) are critical because the daily light
climate for algae is affected. Therefore, we derived a significant relationship by regressing TSS
against ln-scaled discharge and used that to generate a daily time series for the Rum River
(Figure 27).
33
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
Figure 27. TSS at Rum River: a) linear regression against daily flow at gage; b) resulting simulated daily
time series (line), and observed values (symbols).
400000 800000 1200000 1600000 2000000
Daily MeanFlow (m3/d)
140
120 -
100 -
05/99
08/99
08/00
01/99
12/99
04/00
12/00
After calibration we evaluated the efficacy of generating daily time series for TN using a
regression of TN on discharge. The relationship is statistically significant and yielded a more
realistic time series than the interpolation with sparse data that we had used (Figure 28).
34
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
However, calculation of the different limitations on photosynthesis indicates that N is not
limiting in the Rum River (Figure 29), so we kept the simpler approach and did not repeat the
calibration (see section 4.1 for an explanation of the reduction factor as an expression of nutrient
limitation) . TP did not exhibit a statistically significant trend with discharge (R2 = 0.124) so the
simple interpolation was also kept.
Figure 28. TN at Rum River site: a) lndinear regression against daily flow at gage; b) interpolated TN
observations (red) and time series (black) estimated from discharge regression.
a)
1.8
1.6
1.4
-J 1.2
I 0.S
0.6
0.4
0.2
0
R2 = 0.6861
0 500000 1000000 1500000
Discharge
2000000
b)
Est Til
ObsTN
35
-------
AQUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
Figure 29. Predicted nutrient limitations for the dominant algal group in the Rum River.
Note that N is not limiting.
0.9
0.8
0.7
0.6
0.5
! 0.5
0.4
0.3
0.2
0.1
Rum R. 18 MN (CONTROL)
Run on 01-27-09 7:35 AM
3/21/1999 7/19/1999 11/16/1999 3/15/2000 7/13/2000 11/10/2000
Peri High-Nut N_LIM (frac)
¦ Peri High-Nut P04_LIM (frac)
¦ Peri High-Nut CQ2_LIM (frac)
In almost all cases parameter values were chosen from ranges reported in the literature (for
example, Le Cren and Lowe-McConnell 1980, Collins and Wlosinski 1983, Home and Goldman
1994, Jorgensen et al. 2000, Wetzel 2001). However, because these often are broad ranges and
the model is very sensitive to some parameters, iterative calibration was necessary for a subset of
parameters in AQUATOX. Conversely, some parameters have well established values and
default values were used with confidence. A few parameters such as extinction coefficients and
critical force for sloughing of periphyton are poorly defined or are unique to the AQUATOX
formulations and were treated as "free" parameters subject to broad calibration. For example,
some periphyton species are able to migrate vertically through the periphyton mat, and others
have open growth forms; therefore, they could be assigned extinction coefficient values without
regard to the physics of light transmission through biomass fixed in space. As noted earlier,
sensitivity analysis can help determine how much attention needs to be paid to individual
parameters. Sensitivity analysis of five diverse studies has shown that the model is sensitive to
optimal temperature (TOpt) for algae and fish, maximum photosynthesis (PMax) for algae, %
lost in periphytic sloughing, and log octanol-water partition coefficient (KOW). It is advisable to
perform sensitivity analysis when the initial calibration is complete in order to identify
parameters and driving variables requiring additional attention. Although not used in this
application, if modeling a toxic chemical, there are several published sources (for example,
Lyman et al. 1982, Verscheuren 1983, Schwarzenbach et al. 1993), and there are a couple
excellent online references, including the US EPA ECOTOX site and the USDA ARS Pesticide
Properties Database, which can be found with an Internet search engine.
Calibration of AQUATOX for the Minnesota rivers used observed chlorophyll a as the primary
target for obtaining best fits. Because there were only five to eight sestonic chlorophyll a
observations in each of the two target years and only one benthic chlorophyll a observation at
36
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
each location, calibration adequacy was evaluated subjectively, based on generally expected
behavior (e.g. blooms occurring during summer) and approximate concordance with observed
values (in terms of both magnitude and timing), as determined through graphical comparisons of
model output and data (Figure 30).
The central tendencies are similar for predicted and observed distributions for all three sites, as
shown by the relative bias (Figure 31). Despite the fluctuations in predicted chlorophyll a, the
predicted and observed variances are similar for the Crow Wing River and Rum River
simulations. Predicted periphyton sloughing events played a major role in determining the
timing of chlorophyll a peaks in both simulations. The variance in predicted values is too high in
the Blue Earth River simulation, where summer peak concentrations in 1999 appear to be
overestimated by a factor of about two. The reason for this is not known, but may be related to
inherent uncertainties in the simulated flow and TSS values, the sparseness of water chemistry
sampling data, and/or limitations of model algorithms. Given the wide range in degree of
enrichment among these three rivers, and the fact that the model was calibrated against all three
data sets using a single set of parameters, a two-fold error during one period of the Blue Earth
River simulation seems to be acceptable. The combined probability that the Blue Earth River
predictions and observations have the same distribution, based on both central tendency and
dispersion, is greater than 0.8. For the purpose of this analysis, we judged the calibration to be
adequate for the three rivers.
37
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
Figure 30. Observed (symbols) and calibrated AQUATOX simulations (lines) of chlorophyll a in three
Minnesota rivers: a) Blue Earth at mile 54, b) Rum at mile 18, c) Crow Wing at mile 72. Note the order-
of-magnitude range in scale among the figures.
a)
400
350 -
300
250 -
2?00 -
§f50 -
<^00 H
^50
b)
O)
~30 -\
CO
—'20 -\
10
0
01/99 05/99 08/99 12/99 04/00 08/00 12/00
70
60 -
50 -
40 -
r ip
11 II
ill
r ,u
ll 1 I
fllM
Mm
1—Jl
111'
I
01/99 05/99 08/99 12/99 04/00 08/00 12/00
30
25 -
20 -
§>15 -
-£Z
o
.110 -
5 -
0
i
1,1
y
/} ¦
01/99 05/99 08/99 12/99 04/00 08/00 12/00
38
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
Figure 31. Overlap between model and data distributions based on relative bias and ratio of variances, F;
1 = Blue Earth River, 2 = Crow Wing River, 3 = Rum River. Isopleths indicate the probability that the
predicted and observed distributions are the same, assuming normality.
<0
0.5
i
6
0.20
0.80
4
2
1
0.8
0.6
OA
0.2
0J
—6
-4
-2
0
2
4
6
Relative Bias
The calibrated algal model was also applied to three dissimilar sites on the Lower Boise River,
Idaho, without modification from the Minnesota calibration. This provided additional
verification of the generality of the parameter set. The three sites cover a broad range of nutrient
and turbidity conditions over 90 km. Eckert is a low-nutrient, clear-water site upstream of Boise;
Middleton receives wastewater treatment effluent and is a nutrient-enriched, clear-water site; and
Parma is a nutrient-enriched, turbid site impacted by irrigation return flow from agricultural
areas. Although the model overestimated periphyton at the Eckert site, the fit of the initial
application (Figure 32) provided an excellent basis for further river-specific calibration.
39
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
Figure 32. Predicted (line) and observed (symbols) benthic chlorophyll a (a) at Eckert Road, (b) near
Middleton, (c) near Parma, Lower Boise River, Idaho, using Minnesota parameter set.
a)
200
o, 150
E
O)
£ 100
50
35796
36296
36796
b)
200
^ 150
0)
E ioo
50
1/1/9S
*
5/16/99
9/27/00
250
200
^ 150
O)
£ 100
50
ll
.
fii
.
j i
L
f /V J/\
IvJ
/i
\J ArsJ
Wliv/l
1
1/1/98
5/16/99
9/27/00
40
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
As a limited validation, the calibrated model was applied to a site on the Cahaba River south of
Birmingham, Alabama, with modifications to only two parameters, critical force for periphyton
scouring and optimal temperature for algae. The Crow Wing and Rum Rivers have cobbles and
boulders and are more sensitive to higher current velocities than the bedrock outcrops in the
Cahaba River. Not only is the bedrock stable, it also provides abundant crevices and lee sides
that are protected refuges for periphyton. For these reasons greater water velocity is expected to
be required to initiate periphyton scour in the Cahaba River than in the Crow Wing and Rum
Rivers, thus the critical force (Fcrit) for scour of periphyton was more than doubled in the
Cahaba River simulation. Also, between Minnesota and Alabama one would expect different
local ecotypes in resident algal species, with differing adaptations to temperature. Based on
professional judgment, the optimum temperature values (Topt) for green algae and cyanobacteria
were therefore increased by 5°C to 31°C and 32°C respectively. The resulting fit to observed
data (Figure 22) was good. Furthermore, the fish and zoobenthos fits were acceptable (Figure 33,
see also Figure 70). Note that the bluegill are predicted to exhibit ammonia toxicity in 2001, an
observation made possible by viewing biotic process rates. (Within rates graphs, animal
mortality rates may be broken down into their various constituents, see (112)).
Figure 33. Predicted and observed fish in Cahaba River, Alabama; predicted shiner mean biomass = 0.6
g/m2 compared to observed 0.5 g/m2.
1.4
1.3
1.2
1.1
1.0
0.9
>*
-a 0.8
CM
£0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
Cahaba River AL (CONTROL)
Run on 01-26-09 12:07 PM
\ \
/ \ 1 \
/ \ / \
3 / \ / \
-—A. /
L A
i ^
L— ¦r'V^ _
Shiner (g/m2 dry)
Bluegill (g/m2 dry)
- Stone roller (g/m2 dry)
Smallmouth Bas (g/m2 dry)
Smallmouth Ba2 (g/m2 dry)
Obs stone rollers (g/m 2 dry)
Obs shiners (g/m2 dry)
Obs bluegill (g/m2 dry)
Obs bass (g/m2 dry)
8/26/2000 2/24/2001 8/25/2001 2/23/2002 8/24/2002
In another validation, published PCB data from New Bedford Harbor, Massachusetts, were used
to verify the generality of the estuarine ecosystem bioaccumulation model. The observed
concentrations of total PCBs in the water and bottom sediments in the Massachusetts site were
set as constant values in a simulation of Galveston Bay, Texas. The predicted PCB
concentrations in the various biotic compartments at the end of the simulation were then
compared to the observed means and standard deviations in New Bedford Harbor (Figure 34).
Considering that the sites and some of the species were different, the concordance in values
provides a validation of the model for assessing bioaccumulation of chemicals in a "canonical"
or representative estuarine environment.
41
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 2
Figure 34. Predicted and observed concentrations of PCBs in selected animals based on ecosystem
calibration for Galveston Bay, Texas and exposure data (Connolly 1991) for New Bedford Harbor,
Massachusetts.
10
*0?
OA
C
o
2 0.1
+•>
c
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
3. PHYSICAL CHARACTERISTICS
3.1 Morphometry
Volume
Volume is a state variable and can be computed in several
ways depending on availability of data and the site
dynamics. It is important for computing the dilution or
concentration of pollutants, nutrients, and organisms; it
may be constant, but usually it is time varying. In the
model, ponds, lakes, and reservoirs are treated differently
than streams, especially with respect to computing
volumes. The change in volume of ponds, lakes, and
reservoirs is computed as:
Morphometry: Simplifying
Assumptions
• Base flow equation assumes a
rectangular channel
• Site shapes are represented by
idealized geometrical
approximations
• Mean Depth may be held constant
or user varying depth may be
imported
dVolume
dt
Inflow - Discharge - Evap
(3)
where:
dVolume dt
Inflow
Discharge
Evap
derivative for volume of water (m3/d),
inflow of water into waterbody (m3/d),
discharge of water from waterbody (m3/d), and
evaporation (m3/d), see (3).
AQUATOX cannot successfully run if the volume of water in a site falls to zero. To avoid this
condition, if the site's water volume falls below a minimum value (which is defined as a fraction
of the initial condition using the parameter "Minimum Volume Frac." from the "Site" screen), all
differentiation of state variables is suspended (except for the water volume derivative) until the
water volume again moves above the minimum value. Differentiation of all state variables then
resumes.
A time series of evaporation may be entered in the "Site" screen in units of cubic meters per day.
Otherwise, evaporation is converted from an annual value for the site to a daily value using the
simple relationship:
Evap
MeanEvap
365
¦ 0.0254 ¦ Area
(4)
where:
Evap
MeanEvap
365
0.0254
Area
mean daily evaporation (m /d)
mean annual evaporation (in/yr),
days per year (d/yr),
conversion from inches to meters (m/in), and
area of the waterbody (m2).
43
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
The user is given several options for computing volume including keeping the volume constant;
making the volume a dynamic function of inflow, discharge, and evaporation; using a time series
of known values; and, for flowing waters, computing volume as a function of the Manning's
equation. Depending on the method, inflow and discharge are varied, as indicated in Table 3.
As shown in equation (2), an evaporation term is present in each of these volume calculation
options. In order to keep the volume constant, given a known inflow loading, evaporation must
be subtracted from discharge. This will reduce the quantity of state variables that wash out of the
system. In the dynamic formulation, evaporation is part of the differential equation, but neither
inflow nor discharge is a function of evaporation as they are both entered by the user. When
setting the volume of a water body to a known value, evaporation must again be subtracted from
discharge for the volume solution to be correct. Finally, when using the Manning's volume
equation, given a known discharge loading, the effects of evaporation must be added to the
inflow loading so that the proper Manning's volume is achieved. (This could increase the
amount of inflow loadings of toxicants and sediments to the system, although not significantly.)
Table 3. Computation of Volume, Inflow, and Discharge
Method
Inflow
Discharge
Constant
InflowLoad
InflowLoad - Evap
Dynamic
InflowLoad
DischargeLoad
Known values
InflowLoad
InflowLoad - Evap + (State - KnownVals)/dt
Manning
ManningVol - State/dt + Discharge + Evap
DischargeLoad
The variables are defined as:
InflowLoad
DischargeLoad
State
KnownVals
dt
ManningVol
user-supplied inflow loading (m /d);
user-supplied discharge loading (m3/d);
"3
computed state variable value for volume (m );
time series of known values of volume (m3);
incremental time in simulation (d); and
volume of stream reach (m3), see (4).
Figure 35 illustrates time-varying volumes and inflow loadings specified by the user and
discharge computed by the model for a run-of-the-river reservoir. Note that significant drops in
volume occur with operational releases, usually in the spring, for flood control purposes.
44
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
Figure 35. Volume, inflow, and discharge for a 4-year period
in Coralville Reservoir, Iowa.
03
"O
6.0E+07
5.0E+07
4.0E+07
3.0E+07
2.0E+07
1.0E+07
0.0E+00
I
y
V. .
2.5E+08
2.0E+08
1.5E+08
1 0E+08
5.0E+07
=3
O
Oct-74 Oct-75 Nov-76 Dec-77
Apr-75 May-76 Jun-77 Jul-78
0.0E+00
— Inflow
Discharge — Volume
The time-varying volume of water in a stream channel is computed as:
McmningVol = Y ¦ CLength ¦ Width
where:
Y
CLength
Width
dynamic mean depth (m), see (5);
length of reach (m); and
width of channel (m).
(5)
In streams the depth of water and flow rate are key variables in computing the transport, scour,
and deposition of sediments. Time-varying water depth is a function of the flow rate, channel
roughness, slope, and channel width using Manning's equation (Gregory, 1973), which is
rearranged to yield:
Y
f V/J
O ¦ Manning
t]Slope ¦ Width
(6)
where:
0
Manning
Slope
Width
flow rate (m /s);
1/3
Manning's roughness coefficient (s/m );
slope of channel (m/m); and
channel width (m).
The Manning's roughness coefficient is an important parameter representing frictional loss, but it
is not subject to direct measurement. The user can choose among the following stream types:
45
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
concrete channel (with a default Manning's coefficient of 0.020);
dredged channel, such as ditches and channelized streams (default coefficient of 0.030);
and
natural channel (default coefficient of 0.040).
These generalities are based on Chow's (1959) tabulated values as given by Hoggan (1989). The
user may also enter a value for the coefficient.
In the absence of inflow data, the flow rate is computed from the initial mean water depth,
assuming a rectangular channel and using a rearrangement of Manning's equation:
QBaxjDepoTjsh^wM (?)
Manning
"3
QBase = base flow (m /s); and
where:
Idepth = mean depth as given in site record (m).
3 3
The dynamic flow rate is calculated from the inflow loading by converting from m /d to m /s:
where:
0 = M^ (8)
86400
"3
Q = flow rate (m /s); and
Inflow = water discharged into channel from upstream (m3/d).
Bathymetric Approximations
The depth distribution of a water body is important because it determines the areas and volumes
subject to mixing and light penetration. The shapes of ponds, lakes, reservoirs, and streams are
represented in the model by idealized geometrical approximations, following the topological
treatment of Junge (1966; see also Straskraba and Gnauck, 1985). The shape parameter P
(Junge, 1966) characterizes the site, with a shape that is indicated by the ratio of mean to
maximum depth.:
p = 6.0.^!L-3.0 (9)
ZMax
Where:
ZMean = mean depth (m);
ZMax = maximum depth (m); and
P = characterizing parameter for shape (unitless); P is constrained
between -1.0 and 1.0
Shallow constructed ponds and ditches may be approximated by an ellipsoid where Z/ZMax =
0.6 and P = 0.6. Reservoirs and rivers generally are extreme elliptic sinusoids with values of P
46
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
constrained to -1.0. Lakes may be either elliptic sinusoids, with P between 0.0 and -1.0, or
elliptic hyperboloids with P between 0.0 and 1.0. Not all water bodies fit the elliptic shapes, but
the model generally is not sensitive to the deviations.
Based on these relationships, fractions of volumes and areas can be determined for any given
depth (Junge, 1966). The AreaFrac function returns the fraction of surface area that is at depth Z
given Zmax and P, which defines the morphometry of the water body. For example, if the water
body were an inverted cone, when horizontal slices were made through the cone looking down
from the top one could see both the surface area and the water/sediment boundary where the slice
was made. This would look like a circle within a circle, or a donut (Figure 36). AreaFrac
calculates the fraction that is the donut (not the donut hole). To get the donut hole, 1 - AreaFrac
is used.
AreaFrac = (1-P)- —— + P ¦ f (10)
ZMax ZMax
6.0¦ — 3.0 ¦ (1.0-P)- - f-2.0-P- - f
VolFrac = ZMax ZMax ZMax q
3.0 + P
where:
AreaFrac = fraction of area of site above given depth (unitless);
VolFrac = fraction of volume of site above given depth (unitless); and
Z = depth of interest (m).
For example, the fraction of the volume that is epilimnion can be computed by setting depth Z to
the mixing depth. Furthermore, by setting Z to the depth of the euphotic zone, where primary
production exceeds respiration, the fraction of the area available for colonization by macrophytes
and periphyton can be computed:
FracLit = (I-PJ-ZE»P><°ttc + p (ZE„pho
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
If the site is an artificial enclosure then the available area is increased accordingly:
„ r. , r. Area + EnclWallArea
rracLittoral = rracLit
Area
otherwise
(13)
where:
FracLittoral
ZEuphotic
Area
EnclWallArea
FracLittoral = FracLit
= fraction of site area that is within the euphotic zone (unitless);
= depth of the euphotic zone, is assumed to be 1% of surface light
and calculated as 4.605/Extinct (m) see (40);
= site area (m ); and
area of experimental enclosure's walls (m ).
Figure 37. Area as a function of depth Figure 38. Volume as a function of depth
RESERVOIR (P = -0.6)
1 3 5 7 9 11 13 15 17 19 21 23 25
2 4 6 8 10 12 14 16 18 20 22 24
DEPTH (m)
RESERVOIR (P = -0.6)
1 3 5 7 9 11 13 15 17 19 21 23 25
2 4 6 8 10 12 14 16 18 20 22 24
DEPTH(m)
If a user wishes to model a simpler system, the bathymetric approximations may be bypassed in
favor of a more rudimentary set of assumptions via an option in the "site data" screen.
When the user chooses not to "use bathymetry"
• the system is assumed to have vertical walls;
• the system is assumed to have a constant area as a function of depth;
• the system's depth may be calculated at any time as water volume divided by surface
area.
This option may be useful when linking data from other models to AQUATOX as the horizontal
spatial domain of AQUATOX remains unchanged over time. However, a system will not
undergo dynamic stratification based on water temperature unless the more complex bathymetric
approximations are utilized ((8) to (11)).
48
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
Dynamic Mean Depth
AQUATOX normally uses an assumption of unchanging mean depth (i.e., mean over the site
area). However, under some circumstances, and especially in the case of streams and reservoirs,
the depth of the system can change considerably over time, which could result in a significantly
different light climate for algae. For this reason, an option to import mean depth in meters has
been added. A daily time-series of mean depth values may be imported into the software (using
an interface found within the "site" screen by pressing the "Show Mean Depth Panel" button.) A
time-series of mean depth values can be estimated given known water volumes or can be
imported from a linked water hydrology model.
The user-input dynamic mean depth affects the following portions of AQUATOX:
Light climate, see (43);
Calculation of biotic volumes for sloughing calculations, see (74);
Calculation of vertical dispersion for stratification calculations, Thick in equation (18);
Calculation of sedimentation for plants & detritus, Thick in (165);
Oxygen reaeration, see (190);
Toxicant photolysis and volatilization, Thick in (320) and (331).
Habitat Disaggregation
Riverine environments are seldom homogeneous. Organisms often exhibit definite preferences
for habitats. Therefore, when modeling streams or rivers, animal and plant habitats are broken
down into three categories: "riffle," "run," and "pool." The combination of these three habitat
categories make up 100% of the available habitat within a riverine simulation. The preferred
percentage of each organism that resides within these three habitat types can be set within the
animal or plant data. Within the site data, the percentage of the river that is composed of each of
these three habitat categories also can be set. It should be noted that the habitat percentages are
considered constant over time, and thus would not capture significant changes in channel
morphology and habitat distribution due to major flooding events.
These habitats affect the simulations in two ways: as limitations on photosynthesis and
consumption and as weighting factors for water velocity (see 3.2 Velocity). Each animal and
plant is exposed to a weighted average water velocity depending on its location within the three
habitats. This weighted velocity affects all velocity-mediated processes including entrainment of
invertebrates and fish, breakage of macrophytes and scour of periphyton. The reaeration of the
system also is affected by the habitat-weighted velocities.
Limitations on photosynthesis and consumption are calculated depending on a species'
preferences for habitats and the available habitats within the water body. If the species
preference for a particular habitat is equal to zero then the portion of the water body that contains
that particular habitat limits the amount of consumption or photosynthesis accordingly.
49
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
HabUatUmit = j (14)
where:
HabitatLimitspecies = fraction of site available to organism (unitless), used to limit
ingestion, see (91), and photosynthesis, see (35), (85);
Preference habitat = preference of animal or plant for the habitat in question
(percentage); and
Per cent habitat = percentage of site composed of the habitat in question
(percentage).
2
It is important to note that the initial condition for an animal that is entered in g/m is an
indication of the total mass of the animal over the total surface area of the river. Because of this,
density data for various benthic organisms, which is generally collected in a specific habitat type,
cannot be used as input to AQUATOX until these values have been converted to represent the
entire surface area. This is especially true in modeling habitats; for example, an animal could
have a high density within riffles, but riffles might only constitute a small portion of the entire
system.
3.2 Velocity
If the user has site-specific velocity data, this may be entered on the "site data" screen in units of
cm/s. Otherwise, velocity is calculated as a simple function of flow and cross-sectional area:
Velocity -
AvgFlow
XSecArea 86400
¦100
(15)
where
Velocity
AvgFlow
XSecArea
86400
100
velocity (cm/s),
average flow over the reach (m3/d),
cross sectional area (m ),
s/d, and
cm/m.
where:
Inflow
Discharge
AvgFlow
Inflow + Discharge
2
(16)
flow into the reach (m /d);
flow out of the reach (m3/d).
It is assumed that this is the velocity for the run of the stream (user entered velocities are also
assumed to pertain to the run of the screen). No distinction is made in terms of vertical
differences in velocity in the stream. Following the approach and values used in the DSAMMt
model (Caupp et al. 1995), the riffle velocity is obtained by using a conversion factor that is
50
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
dependent on the discharge. Unlike the DSAMMt model, pools also are modeled, so a
conversion factor is used to obtain the pool velocity as well (Table 4).
Flows (Q = discharge)
Run
Velocity
Riffle
Velocity
Pool Velocity
Q < 2.59e5 m3/d
1.0
1.6
0.36
2.59e5 m3/d 7.77e5 m3/d
1.0
1.0
0.66
Figure 39. Predicted velocities in an Ohio stream according to habitat.
4on
350 -
3on -
V)
F
?5n -
u
¦>,
?no -
4—1
u
o
150 -
at
>
ion -
50 -
CN CO
LO CD
00 O
~ Run
- Riffle
Pool
CDCDCD0)0)0)CD0)0)0)G)0)
(D(7)(7)(D(7)(T>(J)CD(DCD(J)(D
3.3 Washout
Transport out of the system, or washout, is an important loss term for nutrients, floating
organisms, and dissolved toxicants in reservoirs and streams. Although it is considered
separately for several state variables, the process is a general function of discharge:
where:
Washout
State
Washout
Discharge
Volume
¦ State
(17)
loss due to being carried downstream (g/m -d), and
3
concentration of dissolved or floating state variable (g/m ).
51
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
3.4 Stratification and Mixing
Thermal stratification is handled in the simplest form
consistent with the goals of forecasting the effects of
nutrients and toxicants. Lakes and reservoirs are
considered in the model to have two vertical zones:
epilimnion and hypolimnion (Figure 40); the metalimnion
zone that separates these is ignored. Instead, the
thermocline, or plane of maximum temperature change, is
taken as the separator; this is also known as the mixing
depth (Hanna, 1990). Dividing the lake into two vertical
zones follows the treatment of Imboden (1973), Park et al.
(1974), and Straskraba and Gnauck (1983). The onset of
stratification is considered to occur when the mean water
temperature exceeds 4 deg. and the difference in
temperature between the epilimnion and hypolimnion
exceeds 3 deg.. Overturn occurs when the temperature of the epilimnion is less than 3 deg.,
usually in the fall. Winter stratification is not modeled, unless manually input. For simplicity,
the thermocline is generally assumed to occur at a constant depth. Alternatively, a user-specified
time-varying thermocline depth may be specified, see the section on modeling reservoirs below.
Figure 40. Thermal stratification in a lake; terms defined in text
Epilimnion
Thermocline
Thick
VertDispersion
Hypolimnion
There are numerous empirical models relating thermocline depth to lake characteristics.
AQUATOX uses an equation by Hanna (1990), based on the maximum effective length (or
fetch). The dataset includes 167 mostly temperate lakes with maximum effective lengths of 172
to 108,000 m and ranging in altitude from 10 to 1897 m. The equation has a coefficient of
determination r = 0.850, meaning that 85 percent of the sum of squares is explained by the
regression. Its curvilinear nature is shown in Figure 41, and it is computed as (Hanna, 1990):
log(MaxZMix) = 0.336 ¦ \o%(Length) - 0.245 (18)
where:
Stratification: Simplifying
Assumptions
• Two vertical zones modeled;
metalimnion is ignored
• Flowing waters are assumed not to
stratify
• Stratification occurs when vertical
temperature difference exceeds
three degrees
• Winter stratification is not modeled
• Thermocline occurs at constant
depth except when user enters time
series
• Wind action is implicit in vertical
dispersion calculations
52
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
MaxZMix = maximum mixing depth under stratified conditions (thermocline
depth) for lake (m); and
Length = maximum effective length for wave setup (m, converted from user-
supplied km).
Figure 41. Mixing depth as a function of fetch
MAXIMUM MIXING DEPTH
20-
£15 —
100
11500
22900
34300
5800 17200 28600 40000
LENGTH(m)
Wind action is implicit in this formulation. Wind has been modeled explicitly by Baca and
Arnett (1976, quoted by Bowie et al., 1985), but their approach requires calibration to individual
sites, and it is not used here.
Vertical dispersion for bulk mixing is modeled as a function of the time-varying hypolimnetic
and epilimnetic temperatures, following the treatment of Thomann and Mueller (1987, p. 203;
see also Chapra and Reckhow, 1983, p. 152; Figure 42):
VertDispersion = Thick ¦
HypVohime
rrit-1 1
' hypo ~ -a hypo
ThermoclArea ¦ Deltat T'epi - T\m
(19)
'po J
where:
VertDispersion =
Thick =
vertical dispersion coefficient (m /d);
distance between the centroid of the epilimnion and the centroid of
the hypolimnion, effectively the mean depth (m);
53
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
HypVolume
ThermoclArea =
Delicti =
rji t-1 rji t+1 —
hypo > hypo
rp t rp t
1 epi > * hypo ~
volume of the hypolimnion (m );
area of the thermocline (m );
time step (d);
temperature of hypolimnion one time step before and one time step
after present time (deg. C); and
temperature of epilimnion and hypolimnion at present time
(deg.C).
Stratification can break down temporarily as a result of high through flow. This is represented in
the model by making the vertical dispersion coefficient between the layers a function of
discharge for sites with retention times of less than or equal to 180 days (Figure 43), rather than
temperature differences as in equation 11, based on observations by Straskraba (1973) for a
Czech reservoir:
VertDispersion
1.37 ¦ 104 ¦ Retention'2 269
(20)
and:
Retention
Volume
TotDischarge
(21)
where:
Retention =
Volume =
TotDischarge =
retention time (d);
volume of site (m'); and
a
total discharge (m /d).
Figure 42. Vertical dispersion as a function of temperature differences
25
20
O
ui
£ 15
a 10
\. vert, c
ispersion
/
/
\
N
onset of
stratificati
/
Dn /
100
10
¦o
E
1
I LL.
0.1
0.01
UJ
o
o
z
o
(/)
3
0.001
12/30 02/28 04/29 06/28 08/27 10/26 12/25
DAY
— Epilimnion Temp. Hypolimnion Temp.
Vert. Dispersion (sq m/d) — 4 degrees
54
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
Figure 43. Vertical dispersion as a function of retention time
VERTICAL DISPERSION
100
180 162 144 126 108 90 72 54 36
171 153 135 117 99 81 63 45 27
RETENTION TIME (d)
The bulk vertical mixing coefficient is computed using site characteristics and the time-varying
vertical dispersion (Thomann and Mueller, 1987):
BulkMixCoejf
VertDispersion ¦ ThermoclArea
Thick
(22)
where:
BulkMixCoejf = bulk vertical mixing coefficient (m /d),
ThermoclArea = area of thermocline (m ).
Turbulent diffusion of biota and other material between epilimnion and hypolimnion is computed
separately for each segment for each time step while there is stratification:
_ BulkMixCoejf , ,
1 U1 OLJljj ^ ^ ( CoflC compartment, hypo ~ COTIC compartment, epi)
Volume
epi
T -hD'ff - BulkMixCoejf
/ III ODljf jlvpn ^ ( C OTIC compartment, epi ~ C OTIC compartment, hypo)
I olumeinpo
(23)
(24)
where:
TurbDijj
Volume
Cone
turbulent diffusion for a given zone (g/m -d);
volume of given segment (m ); and
concentration of given compartment in given zone (g/m3).
55
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
The effects of stratification, mixing due to high throughflow, and overturn are well illustrated by
the pattern of dissolved oxygen levels in the hypolimnion of Lake Nockamixon, a eutrophic
reservoir in Pennsylvania (Figure 44).
Figure 44. Stratification and mixing in Lake Nockamixon,
Pennsylvania as shown by hypolimnetic dissolved oxygen
onset of
stratification
overturn
cn
E 10
CD
high
throughflow
01 /01 /82 03/07/82 05/11 /82 07/15/82 09/18/82 11122182
Modeling Reservoirs and Stratification Options
Stratification assumptions and equations based on lake characteristics may not be appropriate for
modeling reservoirs. Moreover, a lake may have a unique morphometry or chemical
composition that renders inappropriate the equations presented above. For this reason, a
"stratification options" screen is available (through the "site" screen or "water-volume" screen)
that allows a user to specify the following characteristics of a stratified system:
• a constant or time-varying thermocline depth;
• options as to how to route inflow and outflow water; and
• the timing of stratification.
Water volumes for each segment are calculated as a function of the overall system volume and
the thermocline depth (see (10)). Because of this, if a time-varying thermocline depth is
specified, water from one segment must usually be transferred into the other segment, along with
the state variables within that water. In this manner, specifying a time-varying thermocline depth
has the potential to promote mixing between layers. Alternatively, using the linked-mode model,
two stratified segments may be specified with water volumes that are calculated independently
from the thermocline depth; see section 3.8 for more details about stratification in linked-mode.
By default, AQUATOX routes inflow and outflow to and from both segments as weighted by
volume. For example, if the hypolimnion has twice as much volume as the epilimnion, twice as
56
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
much inflow water will be routed to the hypolimnion as to the epilimnion (and twice as much
outflow water will be routed from the hypolimnion). The user has the option to route all inflow
and outflow waters to and from either segment. In this case, all of the nutrients, chemicals, and
other loadings within the inflow water will be routed directly to the specified segment and will
not be transferred to the other segment except through turbulent diffusion or overturn.
Atmospheric and point-source loadings are assumed to be routed to the epilimnion in all cases
(unless a linked-mode model is used in which case more flexibility is present).
Additionally, if a user has information about the timing of stratification, this may be specified on
the "stratification-options" entry screen. This can be used to specify winter stratification, for
example, or precise periods of stratification for each year modeled. If only one year of
stratification dates are entered and multiple years are modeled, all years are assumed to stratify
and overturn on the dates specified in the user input (regardless of the year specified).
3.5 Temperature
Temperature is an important controlling factor in the model. Virtually all processes are
temperature-dependent. They include stratification; biotic processes such as decomposition,
photosynthesis, consumption, respiration, reproduction, and mortality; and chemical fate
processes such as microbial degradation, volatilization, hydrolysis, and bioaccumulation. On the
other hand, temperature rarely fluctuates rapidly in aquatic systems. Default water temperature
loadings for the epilimnion and hypolimnion are represented through a simple sine
approximation for seasonal variations (Ward, 1963) based on user-supplied observed means and
ranges (Figure 45):
,,, . , „ TempRange
lemperatur e = lempMean + (-1.0
• (sinff. 0174533 -(0.987 ¦ (Day + PhaseShift) - 30))))]
where:
Temperature = average daily water temperature (deg. C);
TempMean = mean annual temperature (deg. C);
TempRange = annual temperature range (deg. C),
Day = day of year (d); and
PhaseShift = time lag in heating (= 90 d).
Observed temperature loadings should be entered if responses to short-term variations are of
interest. This is especially important if the timing of the onset of stratification is critical, because
stratification is a function of the difference in hypolimnetic and epilimnetic temperatures (see
Figure 42). It also is important in streams subject to releases from reservoirs and other point-
source temperature impacts.
57
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
3.6 Light
Light is important as the controlling factor for
photosynthesis and photolysis. The default incident light
function formulated for AQUATOX is a variation on the
temperature equation, but without the lag term:
Light: Simplifying Assumptions
• Ice cover is assumed when the
average water temperature drops
below 3 degrees centigrade.
• Photoperiod is approximated by
Julian date (day of year)
• Average daily light is the program
default, although hourly light may
be simulated
Solar = LightMean + LightRcmge ^ ^ 74533 ¦ Day -1.76)- FracUght
(26)
Fracngh, =1 0 - 0-98 {Canopy)
where:
Solar =
LightMean =
LightRange =
Day =
Frac Light
Canopy =
average daily incident light intensity (ly/d);
mean annual light intensity (ly/d);
annual range in light intensity (ly/d);
day of year (d, adjusted for hemisphere);
fraction of site that is shaded; and
user input fraction of site that is tree shaded.
The derived values are given as average light intensity in Langleys per day (Ly/d = 10
kcal/m -d). An observed time-series of light also can be supplied by the user; this is especially
important if the effects of daily weather conditions are of interest. For standing water, if the
average water temperature drops below 3 deg.C, the model assumes the presence of ice cover
and decreases transmitted light to 15% of incident radiation. (This has changed from 33% in
Release 2.2.) This reduction, due to the reflectivity and transmissivity of ice and snow, is an
average of widely varying values summarized by Wetzel (2001). New to Release 3.2, for
moving water (streams and rivers), the average water temperature must drop below 0 deg. C
before ice cover is assumed. For estuaries, average water temperature must fall below -1.8 deg.C
before the model assumes ice cover due to the influence of salinity.
Shade can be an important limitation to light, especially in riparian systems. A user input
"fraction of site subject to shade from a canopy" parameter can be entered either as a constant or
as a time-series within the "Site" input screen. This parameter can be left as zero for no shading
effects on light. Transmission of light through a riparian (stream-side) canopy is a combination
of diffuse and direct transmission (Canham et al. 1990). The average of four forest types from
closed hemlock to open spruce (and cypress) forests is 2% of incident radiation (Canham et al.
1990). Detailed studies in a Midwestern mixed deciduous forest confirm this value for the
summer months, although transmission increased to 40% in winter (Oliphant et al. 2006). A
58
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
value of 2% transmission for a closed canopy is used in AQUATOX. If the density of canopy
varies during the year, then a time-series should be provided, keeping in mind that the 2%
transmission will still apply to the fraction of canopy that is indicated.
Photoperiod is an integral part of the photosynthesis formulation. It is approximated using the
Julian date following the approach of Stewart (1975) (Figure 46):
where:
Photoperiod
Photoperiod =
A
Day =
12 +A- cos (380 ¦ —^ + 248)
365
24
(27)
fraction of the day with daylight (unitless); converted from hours
by dividing by 24;
hours of daylight minus 12 (d); and
day of year (d, converted to radians).
A is the difference between the number of hours of daylight at the summer solstice at a given
latitude and the vernal equinox, and is given by a linear regression developed by Groden (1977):
where:
Latitude
Sign
A = 0.1414 ¦ Latitude - Sign ¦2.413
latitude (deg., decimal), negative in southern hemisphere; and
1.0 in northern hemisphere, -1.0 in southern hemisphere.
(28)
Figure 45. Annual Temperature
Figure 46. Photoperiod as a Function of Date
TEMPERATURE IN A MIDWESTERN POND
35
w 25
77 153 229 305
39 115 191 267 343
JULIAN DAY
: 0.65
£ 0 6
Q
£ 0.55
Q
o 0.45
c
o
=5 04
ra
L"- 0.35
1 53 105 157 209 261 313 365
27 79 131 183 235 287 339
Julian Date
— Latitude 40 N Latitude 40 S
59
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
Hourly Light
When the model is run with an hourly time-step, solar radiation is calculated as variable during
the course of each day. The following equation is used to distribute the average daily incident
light intensity over the portion of the day with daylight hours.
Solar,
K
Solar
daily
hourly
2 Photoperiod
sin
FracDayPassed -
TC •
1-Photoperiod
2
Photoperiod
(29)
• Frac
Light
where:
Soldi hourly
Soldi daily
Photoperiod
FrdcDdyPdssed
Frdc Ligilt
solar radiation at the given time-step (ly/d);
average daily incident light intensity (ly/d), see (25);
fraction of the day with daylight (unitless); see (26);
fraction of the day that has passed (unitless)
fraction of site that is un-shaded, (frac., 1.0-user input shade);
A user may enter a constant or time-series shade variable in the site window ("Fraction of Site
that is Shaded"). When this input is utilized then the FrdcLight variable is calculated.
Figure 47: Average light per day is distributed during daylight hours
in a semi-sinusoidal pattern based on photoperiod.
1200
1000
800
>-
_l
600
400
200
0
10
20
30
40
50
Hours
60
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
3.7 Wind
Wind: Simplifying Assumptions
• If site data are not available a
Fourier series is used to represent
wind loadings
Wind is an important driving variable because it determines
the stability of blue-green algal blooms, affects reaeration
or oxygen exchange, and controls volatilization of some
organic chemicals. Wind also can affect the depth of
stratification for estuaries. Wind is usually measured at
meteorological stations at a height of 10 m and is expressed
as m/s. If site data are not available, default variable wind speeds are represented through a
Fourier series of sine and cosine terms; the mean and twelve additional harmonics seem to
effectively capture the variation (Figure 48):
Wind CosCoeff 0 Z CosCoeff n- Cos
Freqn ¦ 2n ¦ Day
365
+ SinCoeffo'Sin
Freq ¦ 2n ¦ Day"]
365
(30)
where:
Wind
CosCoeffo
CosCoeffn
Day
SmCoeffn
Freq„
wind speed; amplitude of the Fourier series (m/s);
cosine coefficient for the 0-order harmonic, which is the mean wind
speed (default = 3 m/s);
cosine coefficient for the nth-order harmonic;
day of year (d);
sine coefficient for the nth-order harmonic;
selected frequency for the nth- order harmonic.
This default loading is based on an annual cycle of data taken from the Buffalo, NY airport.
Therefore, it has a 365-day repeat, representative of seasonal variations in wind. Frequencies
were selected to ensure that the standard deviation of the Fourier series and the data were closely
matched. The frequency of wind-speeds of less than three meters per second were also precisely
matched to observed data as well as the periodicity of wind-events. The Fourier approach is
quite useful because the mean can be specified by the user and the variability will be imposed by
the function.
If ice cover is predicted, wind is set to 0. A user also may input a site-specific time series, which
may be important where the timing of a cyanobacteria bloom or reaeration is of interest.
61
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION CHAPTER 3
Figure 48. Default wind loadings for Onondaga Lake with mean = 4.17 m/s.
1/2/1991
— Wind (m/s) |
6.0
5.4
4.8
4.2
3.6
3.0
2.4
1.8
1.2
.6
.0
1/12/1989 5/12/1989 9/9/1989 1/7/1990 5/7/1990 9/4/1990
ONONDAGA LAKE, NY (CONTROL) Run on 04-24-08 9:07 PM
(Epilimnion Segment)
3.8 Multi-Segment Model
AQUATOX Release 3 includes the capability to link AQUATOX segments together, tracking
the flow of water and the passage of state variables from segment to segment. Some general
guidelines for using this model follow:
All linked segments must have an identical set of
state variables. (State variables that do not occur in
one segment may be set to zero there.)
Parameters pertaining to animal, plant, and
chemical state variables (i.e. "underlying data") are
considered global to the entire linked system. If
the user changes one of these parameters in one
segment, this parameter changes within all
segments.
On the other hand, "site" parameters, initial
conditions, and boundary conditions are unique to
each segment.
State variables can pass from segment to segment through active upstream and
downstream migration, passive drift, diffusion, and bedload.
Mass balance of all state variables is maintained throughout a multi-segment simulation.
There are two types of linkages that may be specified between individual segments, "cascade
links" and "feedback links." A cascade link is unidirectional; there is no potential for water or
Multi-Segment Model: Simplifying
Assumptions
• All linked segments have an
identical set of state variables
• Each segment is well mixed
• Linkages between segments may be
unidirectional or bidirectional
• Dynamic stratification does not
apply; stratified pairs of segments
must be specified by the user
62
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
state variable flow back upstream. Segments that are linked together by cascade linkages are
solved separately from one another moving from upstream to downstream. This is particularly
useful when modeling faster flowing rivers and streams.
A feedback link allows for water or state variables to flow in both directions. For bookkeeping
purposes, water flows are required to be unidirectional (i.e. entered water flows over a feedback
link must not be negative). However, two feedback links may be specified simultaneously (in
opposite directions) to allow for bidirectional water flows. Feedback links may also be subject to
diffusion; a diffusion coefficient, characteristic length, and cross section must be entered for
diffusion to be calculated, see (32). Segments that are linked together by feedback links are
solved simultaneously. There may only be one contiguous set of segments linked together by
feedback linkages within a simulation (i.e. the model will not solve a "feedback" set of segments
followed by downstream cascade segments followed by more feedback segments below that.)
Figure 49 gives an example of a simulation in which cascade segments and feedback segments
are both included. In this case, AQUATOX solves the simulation from the top down, solving
each segment 1-4, 6, and 6b individually before moving on to solve the feedback segments
simultaneously. Finally, segments 11-14 are be solved individually using the results from the
simultaneous segment run.
Figure 49: An example of feedback and cascade segments linked together.
©
Feedback Seg.
©
Cascade Seg.
\
Feedback Link
Cascade Link
63
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
Stratification and the Multi-Segment Model
Dynamic stratification as described in section 3.4 does not apply to the multi-segment model.
Instead, a user may specify two linked segments as a stratified pair. In this case, the segments
must be linked together with a feedback linkage. A "stratification" screen within each segment's
main interface allows a user to specify whether a segment is part of a stratified pair and, if so,
whether it is the epilimnion or the hypolimnion segment.
When two segments are set up as stratified together, the thermocline area is defined by the user-
entered cross section between. Annual cycles of stratification and overturn may be specified
using the time varying water flows and dispersion coefficients. As was the case in the dynamic
stratification model, fish automatically migrate to the epilimnion in the case of hypoxia in the
lower segment. Sinking phytoplankton and suspended detritus in the epilimnion segment fall
into the designated hypolimnion segment. The light climate of the bottom segment is limited to
that light which penetrates the segment defined as the epilimnion.
When the linked system has enough specified throughflow between the epilimnion and
hypolimnion segment, it is considered to be "well mixed." This is defined as when the average
daily water flow between segments is greater than 30% of the total water volume in both
segments. In this case, fish are assumed to have an equal preference to both segments and they
migrate to equality in a biomass basis. (This allows fish to return to the hypolimnion if it had
earlier been vacated due to anoxia.) Another implication of a well-mixed stratified system (in
linked mode) is that a weighted average of light climate is used when calculating plant
productivity. The calculation of LightLimit for plants (38) is based on a thickness-weighted
average of algal biomass and sediment throughout the entire thickness of the system. This
prevents unreasonable model results due to the light climate in a very thin epilimnion, for
example. Because the system is well-mixed, suspended algae should instead be subject to the
light climate throughout the water column.
State Variable Movement in the Multi-Segment Model
To maintain mass balance, all state variables that are subject to washout or passive drift are also
added to any downstream linked segments. The calculation for this process is as follows:
Washin = Y Washoutup"ream ' VolumeuPstream ' FracWash^^
^ TT 1 ^ '
upstream links UnitDowstream Segment
In the case of toxicants that are absorbed to or contained within a drifting state variable, the
following equation is used:
„r V- WaShOUtCarner-PPBcarner -Volume ^-FracWash,^
WashinToxCarrier = ^ 7Ti (32)
V OniTYlP
upstream links ' ^Dowstream Segment
64
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
where:
Washin =
Washout UpStream
VoltlfflC Segment
FracWashmsLink =
Washin ToxCamer
W^ashoUt Carrier
PPB Carrier
le-6
inflow load from upstream segment (unit/Ldownstream'd);
washout from upstream segment (unit/Lupsfream'd), see (16);
volume of given segment (m3);
fraction of upstream segment's outflow that goes to this
particular downstream segment (unitless);
inflow load of toxicant sorbed to a carrier from an upstream
Segment (Hg/Ldownstream'd);
washout of toxicant carrier from upstream (mg/Lupsiream'd);
concentration of toxicant in carrier upstream (|j,g/kg), see (310);
units conversion (kg/mg)
This Washin term is added to all derivatives for state variables that are suspended in the water
column and subject to drift or "washout."
Dissolved state variables are subject to diffusion across feedback links.
Diffusion
ThisSeg
DiffCoeff ¦Area
CharLength
(c.
°nCOtherSeg
ConcThjsSeg ^
(33)
where:
DiffusionmsSeg = gain of state variable due to diffusive transport over the feedback link
between two segments, (unit/d);
DiffCoeff = dispersion coefficient of feedback link, (m2 /d);
Area = surface area of the feedback link (m2);
= characteristic mixing length of the feedback link, (m);
CharLength
ConC Segment
= concentration of state variable in the relevant segment, (unit/m );
3.9 "Marine" Site Type
A "marine" site type has been added to the AQUATOX list of site types, bringing the full list to
"pond," "lake," "stream," "reservoir," "enclosure," "estuary," and "marine." The marine site
type was required because the "estuary" site type includes assumptions of salt-wedge
stratification that are not always appropriate for the nearshore marine environment (for more
information about the estuarine submodel, please see chapter 10.
If a "marine" site type has been selected the following characteristics of a simulation apply:
• Salinity must be included as a state variable;
• Nitrification and denitrification are not assumed limited by dissolved oxygen (i.e. DO
Correction is set to 1.0 in Equation (174) is and to 0.0 in equation (175))
• Reaeration is calculated using the AQUATOX estuarine reaeration code that takes into
account salinity, wind effects, and water velocity (see equation (445))
65
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 3
Stratified segments are labeled "upper" and "lower" rather than "epilimnion" and
"hypolimnion."
The temperature at which water is assumed to freeze is calculated using Equation (1)
below from UNESCO (1983).
Tf = -0.0575-5 + 1.710523-KT3 ¦ S^2 -2.154996• 1(T4-S2 (23b)
where:
Tf = freezing temperature of saline water (deg C); and
S = salinity (ppt).
66
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
4. BIOTA
The biota consists of two main groups, plants and animals;
each is represented by a set of process-level equations. In
turn, plants are differentiated into algae and macrophytes,
represented by slight variations in the differential
equations. Algae may be either phytoplankton or
periphyton. Phytoplankton are subject to sinking and
washout, while periphyton are subject to substrate limitation and scour by currents. Bryophytes
and freely-floating macrophytes are modeled as special classes of macrophytes, limited by
nutrients in the water column. These differences are treated at the process level in the equations
(Table 5). All are subject to habitat availability, but to differing degrees.
Table 5. Significant Differentiating Processes for Plants
Plant Type
Nutrient
Lim.
Current
Lim.
Light
Lim.
Sinking
Washout
Sloughing
Breakage
Habitat
Phytoplankton
~
~
~
~
~
Periphyton
~
~
~
~
~
Benthic
Macrophytes
~
~
~
Rooted-
Floating
Macrophytes
~
~
Free-Floating
Macrophytes
~
~
~
~
Bryophytes
~
~
~
~
Animals are subdivided into invertebrates and fish; the invertebrates may be "plankton
invertebrates," "nekton invertebrates," "benthic insects," or other "benthic invertebrates." These
groups are represented by different parameter values and by variations in the equations. Insects
are subject to emergence and therefore are lost from the system, but benthic invertebrates are not.
Fish may be represented by both juveniles and adults, which are connected by promotion. One
fish species can be designated as multi-year with up to 15 age classes connected by promotion.
Differences are shown in Table 6.
In addition to the directly-modeled animal categories above, a bioaccumulative endpoint such as
bald eagle, dolphin, or mink that feeds on aquatic compartments can be simulated. This
compartment is defined by feeding preferences, biomagnification factor, and clearance rate (see
section 4.4).
Biota: Simplifying Assumptions
• Biomass is simulated but not
numbers of individual organisms
• Responses are simulated as
averages for the entire group
67
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Table 6. Significant Differentiating Processes for Animals
Animal Type
Washout
Drift
Scour
Entrainment
Emergence
Promotion/
Recruitment
Multi-year
Plankton Invert.
~
Nekton Invert.
~
~
*
Benthic Invert.
~
~
*
Benthic Insect
~
~
~
Fish
~
~
~
* Oysters or size-class predatory invertebrates.
Guild Organization
To better generalize the aquatic ecosystem and to represent marine-biology conventions, the
guilds used by AQUATOX to characterize animals have been reorganized in this version.
Specifically, "shredders" have been renamed to "deposit feeders" and the "grazer" compartment
has been split into "suspended feeders" and "deposit feeders." The animal compartments
available for food web modeling are as follows:
Guild Name
Number of
Compartments
Suspension Feeders
6
Deposit Feeders
3
Veliger
2
Spat
2
Clams/Adult Oysters
4
Snails
2
Small Predatory Invert.
2
Predatory Invertebrate
4
Small Forage Fish
2
Large Forage Fish
2
Small Bottom Fish
2
Large Bottom Fish
2
Small Game Fish
4
Large Game Fish
4
Age-Class Fish
15
Anti-Extinction Code
Plants or animals with non-zero initial conditions are assumed to be "seeded" in the case that
their biomass drops to zero. This allows for species recovery in the aftermath of a physical or
68
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
chemical shock to a system. Each time a plant or animal has a biomass that falls to below 1x10'
10 • • 7
mg/L that organism is re-seeded with a loading of 1x10" mg/L.
4.1 Algae
Plants: Simplifying Assumptions
• Photosynthesis is modeled as a maximum observed rate multiplied by reduction factors. Hie reduction factors are
assumed to be independent of one another.
• There are two options for modeling nutrient effects on plants. Intracellular storage of nutrients may be modeled as a new
option to Release 3.2; otherwise constant stoichiometry within species is assumed and nutrient limitation is calculated as
a function of nutrients in the water column.
• For each individual nutrient, saturation kinetics is assumed
• Algae exhibit a nonlinear, adaptive response to temperature changes
• Low temperatures are assumed not to affect algal mortality
• The ratio between biovolume and biomass is assumed to be constant for a given growth form
• Constant chlorophyll a to biomass ratios are assumed within algae groups
Phytoplankton-specific
• Phytoplankton other than cyanobacteria are assumed to be mixed throughout the well-mixed layer unless specified as
"surface floating."
• In the event of ice cover, all phytoplankton will occur in the top 2 m
• Sinking of phytoplankton is modeled as a function of physiological state
• Phytoplankton are subject to downstream drift as a simple function of discharge
• To model phytoplankton (and zooplankton) residence time, an implicit assumption may be made that upstream reaches
included in the "Total River Length " have identical environmental conditions as the reach being modeled
Cyanobacteria-specific
• By default cyanobacteria are specified as "surface floating" in which case they are assumed to be located in the top 0.1 m
unless limited by lack of nutrients or sufficient wind occurs in which case they are located within the top 3 m. This
default assumption (that cyanobacteria float) can be changed by the user.
• The averaging depth for "surface floating" plants is three meters to more closely correspond to monitoring data.
• Cyanobacteria are not severely limited by nitrogen due to facultative nitrogen fixation (if N less than Vi KN)
Periphyton-specific
• Periphyton are limited by slow currents that do not replenish nutrients and carry away senescent biomass
• Periphyton are assumed to adapt to the ambient conditions of a particular channel
• Periphyton are defined as including associated detritus; non-living biomass is modeled implicitly
Macrophyte-specific
• Macrophytes occupy the littoral zone
• Rooted macrophytes and benthic macrophytes are not limited by nutrients but are assumed to take up necessary nutrients
from bottom sediments (located outside the AQUATOX domain)
• Rooted floating macrophytes are differentiated from benthic macrophytes in that rooted-floating macrophytes are
assumed to occur near the surface and are not limited by low light
• Non-rooted, floating macrophytes are limited by nutrients but not by low light. These macrophytes can wash out of a
system.
• Bryophytes are limited by nutrients, can tolerate low light, and contain a high percentage of refractory material
69
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION CHAPTER 4
3 2
The change in algal biomass—expressed as g/m for phytoplankton, but as g/m for periphyton—
is a function of the loading (especially phytoplankton from upstream), photosynthesis,
respiration, excretion or photorespiration, nonpredatory mortality, grazing or predatory mortality,
sloughing, and washout. As noted above, phytoplankton also are subject to sinking. If the
system is stratified, turbulent diffusion also affects the biomass of phytoplankton.
dBiomassphto
= Loading + Photosynthesis - Respiration - Excretion
dt
- Mortality - Predation ± Sinking ± Floating (34)
-Washout + Washin±TurbDiff + Diffusion Seg + ^()UK^
dBiomass
dt
Pen = Loading + Photosynthesis - Respiration - Excretion
-Mortality - Predation + SedPeri - Slough
(35)
where:
dBiomass/dt
Loading
Photosynthesis
Respiration
Excretion
Mortality
Predation
Washout
Washin
Sinking
Floating
TurbDiff
Diffusionseg
Slough
Sedperi
change in biomass of phytoplankton and periphyton with respect to
3 2
time (g/m -d and g/m -d);
3 2
boundary-condition loading of algal group (g/m -d and g/m -d);
3 2
rate of photosynthesis (g/m -d and g/m -d), see (35);
3 2
respiratory loss (g/m -d and g/m -d), see (63);
3 2
excretion or photorespiration (g/m -d and g/m -d), see (64);
3 2
nonpredatory mortality (g/m -d and g/m -d), see (66);
3 2
herbivory (g/m -d and g/m -d), see (99);
"3
loss due to being carried downstream (g/m -d), see (129);
loadings from upstream segments (linked segment version only,
g/m3 d), see (30);
loss or gain due to sinking between layers and sedimentation to
bottom (g/m3-d), see (69);
loss from the hypolimnion or gain to the epilimnion due to the
floatation of "surface-floating" phytoplankton. 100% of "surface-
floating" phytoplankton that arrive in the hypolimnion through
loadings or water flows are set to immediately float.
turbulent diffusion (g/m3-d), see (22) and (23);
gain or loss due to diffusive transport over the feedback link
"3
between two segments, (g/m -d), see (32);
Scour loss of Periphyton or addition to linked Phytoplankton, see
(75); and
Sedimentation of Phytoplankton to Periphyton, see (83).
70
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 50 and Figure 51 are examples of the predicted changes in biomass and the processes that
contribute to these changes in a eutrophic lake. Note that photosynthesis and predation dominate
the diatom rates, with respiration much less important during the growing season.
Figure 50. Predicted algal biomass in Lake Onondaga, New York
ONONDAGA LAKE, NY (PERTURBED) Run on 04-23-08 2:59 PM
(Epilimnion Segment)
¦ Cyclotella nan (mg/L dry)
¦ Greens (mg/L dry)
¦ Phyt, Blue-Gre (mg/L dry)
¦ Cryptomonad (mg/L dry)
1/12/1989 5/12/1989 9/9/1989
1/7/1990 5/7/1990 9/4/1990
1/2/1991
Figure 51. Predicted process rates for diatoms in Lake Onondaga, New York
ONONDAGA LAKE, NY (PERTURBED) Run on 04-23-08 2:59 PM
(Epilimnion Segment)
110
99
88
77
66
Ij 55
u
L
44
33
22
11
a
*)
A
ll
L
1
\
¦ Cyclotella nan Photosyn (Percent)
¦ Cyclotella nan Respir (Percent)
- Cyclotella nan Excret (Percent)
¦ Cyclotella nan Other Mort (Percent)
¦ Cyclotella nan Predation (Percent)
Cyclotella nan Washout (Percent)
¦ Cyclotella nan Sediment (Percent)
Cyclotella nan TurbDiff (Percent)
Cyclotella nan SinkToHypo (Percent)
3/11/1989
9/9/1989
3/10/1990
9/8/1990
71
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Photosynthesis is modeled as a maximum observed rate multiplied by reduction factors for the
effects of toxicants, habitat, and suboptimal light, temperature, current, and nutrients:
Photosynthesis = PMax ¦ PProdLimit ¦ Biomass ¦ HabitatLimit ¦ SaltEffect (36)
The limitation of primary production in phytoplankton is:
PProdLimit = LtLimit ¦ NutrLimit ¦ TCorr ¦ FracPhoto (37)
Periphyton have an additional limitation based on available substrate, which includes the littoral
bottom and the available surfaces of macrophytes. The macrophyte surface area conversion is
2 2
based on the observation of 24 m periphyton/m bottom (Wetzel, 1996) and assumes that the
observation was made with 200 g/m macrophytes.
PProdLimit = LtLimit ¦ NutrLimit ¦ VLimit ¦ TCorr ¦ FracPhoto
¦ ( FracLittoral + SurfAreaConv ¦ BiomassMacroPhyteS)
(38)
where:
Pmax
LtLimit
NutrLimit
Vlimit
TCorr
HabitatLimit
SaltEffect
FracPhoto
FracLittoral
SurfAreaConv
BiomaSSMacro
Biomass peri
= maximum photosynthetic rate (1/d);
= light limitation (unitless), see (38);
= nutrient limitation (unitless), see (55) and (55b) ;
= current limitation for periphyton (unitless), see (56);
= limitation due to suboptimal temperature (unitless), see (59);
= in streams, habitat limitation based on plant habitat preferences
(unitless), see (13).
= effect of salinity on photosynthesis (unitless);
= reduction factor for effect of toxicant on photosynthesis (unitless),
see (421);
= fraction of area that is within euphotic zone (unitless) see (11);
= surface area conversion for periphyton growing on macrophytes (0.12
m2/g);
= total biomass of macrophytes in system (g/m ); and
= biomass of periphytic algae (g/m ).
Under optimal conditions, a reduction factor has a value of 1; otherwise, it has a fractional value.
Use of a multiplicative construct implies that the factors are independent. Several authors (for
example, Collins, 1980; Straskraba and Gnauck, 1983) have shown that there are interactions
among the factors. However, we feel the data are insufficient to generalize to all algae;
therefore, the simpler multiplicative construct is used, as in many other models (Chen and Orlob,
1975; Lehman et al., 1975; J0rgensen, 1976; Di Toro et al., 1977; Kremer and Nixon, 1978; Park
et al., 1985; Ambrose et al., 1991). Default parameter values for the various processes are taken
primarily from compilations (for example, J0rgensen, 1979; Collins and Wlosinski, 1983; Bowie
et al., 1985); they may be modified as needed.
72
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Light Limitation
Because it is required for photosynthesis, light is a very important limiting variable. It is
especially important in controlling competition among plants with differing light requirements.
Similar to many other models (for example, Di Toro et al., 1971; Park et al., 1974, 1975, 1979,
1980; Lehman et al., 1975; Canale et al., 1975, 1976; Thomann et al., 1975, 1979; Scavia et al.,
1976; Bierman et al., 1980; O'Connor et al., 1981), AQUATOX uses the Steele (1962)
formulation for light limitation. Light is specified as average daily radiation. The average
radiation is multiplied by the photoperiod, or the fraction of the day with sunlight, based on a
simplification of Steele's (1962) equation proposed by Di Toro et al. (1971). The equation is
slightly different when the model is run with a daily versus an hourly time-step:
LtLimitnailv = 0.85 : ^^ (39)
Extinct ¦ (DepthBottom - DepthTop )
e ¦ Photoperiod ¦ (LtAtDepthDaily - LtAtTopDaily ) ¦ PeriphytExt
'thBonom-DepthTop,
ULimit = 6 • (LtAtDePthHourly ~ LtAtT°PHourly ) ' ^ !phytic
"Vl HOUrly Extinct ¦ (DepthBottom - DepthTop )
where:
LtLimitrimeStep = light limitation (unitless);
e = the base of natural logarithms (2.71828, unitless);
Photoperiod = fraction of day with daylight (unitless), see (26);
Extinct = total light extinction (1/m), see (40), (41);
Depth Bottom = maximum depth or depth of bottom of layer if stratified (m); if
periphyton or macrophyte then limited to euphotic depth;
Depth Top = depth of top of layer (m);
LtAtTop = limitation of algal growth due to light, (unitless) see (44), (45);
LtAtDepth = limitation due to insufficient light, (unitless), see (43);
PeriphytExt = extinction due to periphyton; only affects periphyton and
macrophytes (unitless), see (42).
Because the equation overestimates by 15 percent the cumulative effect of light limitation over a
24-hour day, a correction factor of 0.85 is applied to the daily formulation (Kremer and Nixon,
1978). When AQUATOX is run with an hourly time-step, the correction factor of 0.85 is not
relevant, nor the inclusion of photoperiod.
Light limitation does not apply to free-floating macrophytes as these are assumed to be located at
the surface of the water.
Even when the model is run with an hourly time-step, two algal equations utilize the daily light
limit equation (38) as most appropriate. First, when calculating algal mortality, the stress factor
for suboptimal light and nutrients (68) is expecting the input of daily light limitation (i.e. the
plants do not all die each night). Secondly, when calculating the sloughing of benthic algae (75)
73
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION CHAPTER 4
the calculation of suboptimal light is calibrated to daily light limitation, not the instantaneous
absence or presence of light (i.e. sloughing is not more likely to occur when it is dark).
Extinction of light is based on several additive terms: the baseline extinction coefficient for water
(which may include suspended sediment if it is not modeled explicitly), the so-called "self-
shading" of plants, attenuation due to suspended particulate organic matter (POM) and inorganic
sediment, and attenuation due to dissolved organic matter (DOM):
(41)
Extinct = Water Extinction + PhytoExtinction + ECoeffDOM ¦ DOM
+ ECoeffPOM ¦ ILPartDetr + ECoeffSed ¦ InorgSed
where:
WaterExtinction = user-supplied extinction due to water (1/m);
PhytoExtinction = user-supplied extinction due to phytoplankton and macrophytes
(1/m), see (41), (42);
"3
ECoeffDOM = attenuation coefficient for dissolved detritus l/(m g/m );
DOM = concentration of dissolved organic matter (g/m3), see (143) and
(144);
ECoeffPOM = attenuation coefficient for particulate detritus l/(mg/m );
"3
PartDetr = concentration of particulate detritus (g/m ), see (141) and
(142);
ECoeffSed = attenuation coefficient for suspended inorganic sediment
l/(mg/m3); and
InorgSed = concentration of total suspended inorganic sediment (g/m ), see
(244)
For computational reasons, the value of Extinct is constrained between 5"19 and 25. Self-shading
by phytoplankton, periphyton, and macrophytes is a function of the biomass and attenuation
coefficient for each group. Extinction by periphyton is computed differently because it is not
depth-dependent but rather pertains to the growing surface:
and
PhytoExtinction = (ECoeffPhytoalga ¦ Biomassaiga) (42)
PeriPhytExt = eLpen ECoeffPhytoper, ¦ Biomasspen) (43)
where:
EcoeffPhytoaiga = attenuation coefficient for given phytoplankton or macrophyte
(1/m-g/m3),
EcoeffPhytoperi = attenuation coefficient for given periphyton (1/m-g/m ),
Biomass = concentration of given plant (g/m3 or g/m2), and
The light limitation at depth is computed by:
74
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
LtAtDepthTimeStep ~ e
f Tjpht-r, o -ExtinctEpi-DepthTop
TimeStep Extinct VSeg ' DepthBottom
LightSat-LightCorr
(44)
Light limitation at the surface of the water body is computed by:
Light TimeStep
LtAtTopTimeStep £ LightSat-LightCorr
(45)
and light limitation at the top of the hypolimnion is computed by:
TimeStep - Extinct£pj ¦DepthTop
IA A17 op j-ime^lep ~ Q LightSat-LightCorr
(46)
where:
LtAtTop
LtAtDepth
Extinct
Light TimeStep
LightCorr
LightSat
limitation of algal growth due to light, (unitless multiplier, 0 being
no limitation, 1 being 100% limitation)
limitation due to insufficient light, (unitless, see LtAtTop)
overall extinction of light in relevant vertical segment (1/m), (40)
photosynthetically active radiation (ly/d), (46);
Correction factor, 1.0 for a daily time-step, 1.25 for an hourly
time-step. LightSat is increased by 25% to account for
instantaneous solar radiation as opposed to daily averages;
light saturation level for photosynthesis (ly/d).
Phytoplankton not specified as "surface floating" are assumed to be mixed throughout the well
mixed layer, although subject to sinking. However, healthy cyanobacteria (and some other algal
species) tend to float. Therefore, if the phytoplankton is specified as "surface floating" and the
nutrient limitation is greater than 0.25 (Equation (55)) and the wind is less than 3 m/s then
DepthBottom for surface floating algae is set to 0.1 m to account for buoyancy. Otherwise it is
set to 3 m to represent downward transport by Langmuir circulation. When calculating self-
shading for surface-floating algae the model accounts for more intense self shading in the upper
layer of the water column due to the floating concentration of algae there. The Extinct term in
equation (43) is multiplied by the segment thickness and divided by the thickness over which the
floating algae occur so that the more intense self-shading effects of these algae concentrated at
the top of the system are properly accounted for. Rather than average the biomass of "surface
floating" plants over the entire water column, the biomass is normalized to the top 3 m to more
closely correspond with monitoring data.
Under the ice, all phytoplankton are represented as occurring in the top 2 m (cf. LeCren and
Lowe-McConnell, 1980). As discussed in Section 3.6, light is decreased to 15% of incident
radiation if ice cover is predicted. Approximately half the incident solar radiation is
photosynthetically active (Edmondson, 1956):
LightTjmestep = SolarTimeStep ¦ o. 5 (47)
75
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
where:
SolarnmeStep = daily light intensity on a daily (25) or hourly (28) basis (ly/d).
The light-limitation function represents both limitation for suboptimal light intensity and
photoinhibition at high light intensities (Figure 52). When considered over the course of the
year, photoinhibition can occur in very clear, shallow systems during summer mid-day hours
(Figure 54), but it often is not a factor when considered over 24 hours (Figure 55).
To help understand the occurrence of photoinhibition as opposed to insufficient light, two new
output "photosynthetic limitation variables" are available—"LowLt LIM" and "HighLt LIM."
These output variables are same as the overall light limitation factor (LtLIM) but are modified
to indicate photoinhibition as opposed to insufficient light. When low-light limitation causes
light conditions to be sub-optimal then the "high-light limitation" is set to zero. When
photoinhibition is occurring then the "low-light limitation" is set to zero. To determine this
difference AQUATOX differentiates the equations used to produce the curves in Figures 50 and
51 (see (38) and (39)) and and determines whether the current light is greater than or less than
the maximum value.
It is also worth noting that in simulations with a one-day time step, the light limitation factor
(Lt LIM) represents a daily light limitation and is therefore subject to the photoperiod. In other
words, if the sun is shining only 50% of the day, the maximum the LtLimit can be is 0.5. This is
because Lt LIM is a limitation on the maximum daily photosynthesis rate for a plant which
would be based on 24-hours of light exposure.
The extinction coefficient for pure water varies considerably in the photosynthetically-active
400-700 nm range (Wetzel, 1975, p. 55); a value of 0.016 (1/m) correspond to the extinction of
green light. In many models dissolved organic matter and suspended sediment are not
considered separately, so a much larger extinction coefficient is used for "water" than in
AQUATOX. The attenuation coefficients have units of l/m-(g/m3) because they represent the
amount of extinction caused by a given concentration (Table 7).
Table 7. Light Extinction and Attenuation Coefficients
WaterExtinction
0.02 1/m
Wetzel, 1975
E CoeffPhy to diatom
0.14 l/m-(g/m3)
calibrated
E CoeffPhytO blue-green
0.099 l/m-(g/m3)
Megard et al., 1979 (calc.)
ECoeffDOM
0.03 l/m-(g/m3)
Effler et al., 1985 (calc.)
ECoeffPOM
0.12 l/m-(g/m3)
Verduin, 1982
ECoeffSed
0.17 l/m-(g/m3)
Straskraba and Gnauck,
1985
All coefficients may be user-supplied in the plant or site underlying data.
76
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 52. Instantaneous Light Response Figure 53. Daily Light Response Function
Function
Figure 54. Mid-day Light Limitation Figure 55. Daily Light Limitation
Diatoms in 0.5-m Deep Pond
V, 0.98
E 0.96
^ 0.94 -
'§ 0.92
¦i 0.9
0.88 -
200 250 300 350 400 450 500
Light (ly/d)
Diatoms in 0.5-m Deep Pond
"O
o>
500
450
:400
350
aj
J 300
w 250
200
3 53 103 153 203 253 303 353
Julian Date
0.98
I 0.45
t 0.4 '
o
ro 0.35
J 0.3
0.25 -
200 250 300 350 400 450
Average Light (ly/d)
Diatoms in 0.5-m Deep Pond
500
450
'400
~ 350
<1)
| 300
w 250
200
3 53 103 153 203 253 303 353
Julian Date
0.55
0.5
0.45
0.4
0.35
0.3
0.25
Light — Limitation
The Secchi depth, the depth at which a Secchi disk disappears from view, is a commonly used
indication of turbidity. It is computed as (Straskraba and Gnauck, 1985):
Secchi = — (48)
Extinction
where:
Secchi = Secchi depth (m).
This relationship also could be used to back-calculate an overall Extinction coefficient if only the
Secchi depth is known for a site.
It should be noted that although Secchi depth can be computed for the hypolimnion segment,
based on the suspended material, it is a relatively meaningless value for the hypolimnion and
generally should be ignored. Light extinction in the hypolimnion is calculated based on the light
that has first filtered through the epilimnion as shown in equation (45).
77
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
As a verification of the extinction computations, the calculated and observed Secchi depths were
compared for Lake George, New York. The Secchi depth is estimated to be 8.3 m in Lake
George, based on site data for the various components (Figure 56). This compares favorably
with observed values of 7.5 to 11 (Clifford, 1982).
Adaptive Light
Saturating light can be specified as a constant for each plant taxonomic group (classic
AQUATOX approach) or it can be adaptive based on Kremer and Nixon (1978) and similar to
the approach used in EFDC. The adaptive light saturation is the weighted average of
photosynthetically active solar radiation (PAR) at the optimal depth for growth of a given plant
group, using an approximation based on the user-specified light saturation and site solar radiation
and turbidity at the beginning of the simulation:
Figure 56. Contributions to light extinction in Lake George NY.
Sediment (0.00%)
rWater (6.97%)
§$i§Ss>^~Phytoplankton (1.59%)
POM (26.13%)
DOM (65.32%)
LightSatCalc = ().l{LighlHisl]) + ().2{LighlHisl)+ 0. \(fjgh/His/,)
(49)
LightHistn = PAR ¦ e{-Extmct 'ZOpt'
(50)
where:
LightSatCalc
LightHistn
adaptive light saturation (Ly/d)
photosynthetically active radiation at optimum depth for plant
growth n days prior to simulation date (Ly/d)
photosynthetically active radiation, Solar * 0.5 (Ly/d)
incident solar radiation (Ly/d)
total light extinction computed dynamically (40).
PAR
Solar
Extinct
78
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
If the LightSatCalc is greater or less than the user-entered maximum and minimum light
saturation coefficients ("Plant underlying data" screen) then the LightSatCalc is set to the user-
entered maximum or minimum. This LightSatCalc variable is then used in the LtAtDepth and
LtAtTop calculations (43)-(45).
ZOpt
\n(LightSat / MaxDailyLight)
Plant
- Extinct
(51)
Init.Cond
where:
ZOpt
Plant
LightSat =
MaxDailyLight =
ExtinctinuCond
optimum depth for a given plant (a constant approximated at the
beginning of the simulation in meters);
user entered light saturation coefficient (Ly/d);
maximum daily-averaged incident solar radiation for one
calendar year forward from the start date (Ly/d);
initial condition total light extinction (unitless);
Nutrient Limitation
There are several ways that nutrient limitation has been represented in models. Algae are
capable of taking up and storing sufficient nutrients to carry them through several generations,
and models have been developed to represent this. However, if the timing of algal blooms is not
critical, intracellular storage of nutrients can be ignored, constant stoichiometry can be assumed,
and the model is much simpler. Therefore, based on the efficacy of this simplifying assumption,
nutrient limitation by external nutrient concentrations has traditionally been used in AQUATOX,
as in many other models (for example, Chen, 1970; Parker, 1972; Lassen and Nielsen, 1972;
Larsen et al., 1974; Park et al., 1974; Chen and Orlob, 1975; Patten et al., 1975; Environmental
Laboratory, 1982; Ambrose et al., 1991). New to Release 3.1 and beyond, internal nutrient
concentrations may be modeled in AQUATOX; see the section on internal nutrients below.
When modeling nutrient limitations with external nutrients, for an individual nutrient, saturation
kinetics is assumed, using the Michaelis-Menten or Monod equation (Figure 57); this approach is
founded on numerous studies (cf. Hutchinson, 1967):
PUn.it = PhOSph°nS (52)
Phosphorus + KP
Nitrogen
NLimit
CLimit
Nitrogen + KN
Carbon
(53)
Carbon + KC02 (54)
where:
PLimit = limitation due to phosphorus (unitless);
79
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Phosphorus = available soluble phosphorus (gP/m );
KP = half-saturation constant for phosphorus (gP/m3);
NLimit = limitation due to nitrogen (unitless);
Nitrogen = available soluble nitrogen (gN/m3);
KN = half-saturation constant for nitrogen (gN/m );
CLimit = limitation due to inorganic carbon (unitless);
Carbon = available dissolved inorganic carbon (gC/m ); and
KC02 = half-saturation constant for carbon (gC/m3).
Figure 57. Nutrient limitation
MICHAELIS-MENTEN RELATIONSHIP
DIATOMS
0.8
b 0.6
o- 0.4
0.2
half-saturation
0.00 0.01 0.03 0.04 0.05 0.07 0.08 0.09
PHOSPHATE (mg/L)
Nitrogen fixation in cyanobacteria is handled by setting NLimit to 1.0 if Nitrogen is less than half
the KN value. Otherwise, it is assumed that nitrogen fixation is not operable, and NLimit is
computed as for the other algae. AQUATOX also provides an option to trigger nitrogen fixation
as a function of an input parameter, the ratio of inorganic N to inorganic P, which may be
selected and specified in the "Study Setup" screen. When the ratio falls below the threshold,
nitrogen fixation is assumed to occur; the default threshold N:P is 7. When internal nutrients are
modeled, uptake of nitrogen is set to its maximum rate due to nitrogen fixation when the internal
nutrient concentration falls below half of its internal half saturation coefficient. See (55b) and
(55f) below.
Concentrations must be expressed in terms of the chemical element. Because carbon dioxide is
computed internally, the concentration of carbon is corrected for the molar weight of the
element:
Carbon = C2C02 ¦ C02 (55)
where:
C2C02 = ratio of carbon to carbon dioxide (0.27); and
CO2 = inorganic carbon (g/m3).
When modeling with internal or external nutrients, AQUATOX uses the minimum limiting
80
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
nutrient, whereby the Michaelis-Menten equation is evaluated for each nutrient, and the factor
for the nutrient that is most limiting at a particular time is used. This is the approach used in
many similar models (for example, Larsen et al., 1973; Baca and Arnett, 1976; Scavia et al.,
1976; Smith, 1978; Bierman et al., 1980; Park et al., 1980; Johanson et al., 1980; Grenney and
Kraszewski, 1981; Ambrose et al., 1991). The overall nutrient limitation is calculated as
follows:
Alternative formulations used in other models include multiplicative and harmonic-mean
constructs, but the minimum limiting nutrient construct is well-founded in laboratory studies
with individual species.
Internal Nutrients Model
It is well known that many algae are able to take up nutrients even when not required for
growth—so-called "luxury uptake." MS.CLEANER, a precursor to AQUATOX, used internal
nutrients (Collins 1980), but this approach was not used in the original AQUATOX because of
memory limitations at the time. The present version of AQUATOX has the option of modeling
internal nutrients based on the approach of QUAL2K (Chapra et al. 2007) and WASP7
(Ambrose et al. 2006, Martin et al. 2006). When internal nutrients are specified, NLimit and
PLimit are calculated as a function of the internal nutrient concentration in plants, with nitrogen
fixation by cyanobacteria being a special case:
NutrLimit = mm(PLimit, NLimit, CLimit)
(56)
where:
NutrLimit
reduction due to limiting nutrient (unitless).
NLimit = 1
Min N Ratio
N Ratio
(55b)
PLimit = 1
Min P Ratio
P Ratio
If the plant is cyanobacteria and
N Ratio < (0.5 • NHalfSatinternai) then NLimit =1.0.
where:
N Ratio or P Ratio
NHalfSati„ternal
Min N/P Ratio
internal nutrient concentration over biomass, (g/g AFDW);
half-saturation constant for intracellular N (mg/mg AFDW);
N Ratio or P Ratio at which growth ceases, a user-input ratio,
(g/g AFDW);
81
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION CHAPTER 4
Internal nutrients are calculated with independent derivatives for each relevant plant as follows
—-^u^entphytoPlankton = loading + Uptake - Mortality - Respiration - Excretion
dt
-ZpredPredationpredAiga ± Sink ± Floating-Washout (55c)
+ Washin± TurbDiff + Diffusion Seg + Slough 13
—^utr^entperiPhyton = Loaciing + Uptake - Mortality - Respiration - Excretion
dt (55d)
" Zpred Predationpred, Aiga + SedimentationPhytoplaMon - Slough
where:
NutrientAiga
N20
Loading
Uptake
Mortality
Predation
Sinking Loss
Sinking Gain
Floating Loss
Floating Gain
Washout
Washin
TurbDiff Loss
TurbDiff Gain
Diffusion
Slough
Sedimentation
Respiration
Excretion
concentration of nutrient within plant compartment, ((Jg/L);
nutrient to organism ratio, ((_ig nutrient/mg organism);
external loadings ¦ N20; assumes external loadings have same
stoichiometry as current biomass, ([Jg/L d);
uptake of nutrients from the water column, see (55e) and (55g),
(M-g/L d);
mortality of algal biomass ¦ N20 ([Jg/L d);
predation of algal biomass ¦ N20 ([Jg/L d);
sinking loss of algal biomass ¦ N20 ((J-g/L d);
sinking gain of algal biomass • N20other Segment (|ag/L d);
floating loss of algal biomass ¦ N20 ((J-g/L d);
floating gain of algal biomass ¦ N20other Segment ((Jg/L d);
washout of algal biomass ¦ N20 ((Jg/L d);
gain from upstream segment of algal biomass,
Washout Other Segment ' N20 Other Segment ((Jg/L d),
turbulent diffusion loss of algal biomass ¦ N20 ((Jg/L d);
turbulent diffusion gain of biomass ¦ N20other Segment ((Jg/L d);
diffusion Linked Segment ' N20 LinkedSegment ((Jg/L d),
sloughing loss of periphyton biomass ¦ N20periphyton ((Jg/L d);
sedimentation of phytoplankton biomass ¦ N20 ((Jg/L d);
dark respiration of algal biomass ¦ N20 ((Jg/L d);
photo respiration of algal biomass ¦ N20 ((Jg/L d);
AQUATOX displays internal nutrients in plants as a concentration associated with overlying
water ((Jg/L), and as nutrient-to-organism ratios of grams of nutrient per gram of AFDW organic
matter.
Uptake of nutrients is modeled as follows:
82
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
PhytoUpN = MaxNUptake ¦ biomass • le3 •
( ammonia + nitrate V
NHalfSat
Internal
NHalfSat + ammonia + nitrate I NHalfSatlnternal + (NRatio -MinNRatio)
(55e)
If the plant is cyanobacteria and is fixing nitrogen then uptake is assumed to occur at the
maximum rate.
PhytoUpN = MaxNUptake ¦ biomass • le3 (55f)
Uptake of phosphorus is modeled with a similar formulation used for the uptake of nitrogen:
PhytoUpP = MaxPUptake ¦ biomass • le3 •
f TSP V
PHalfSat
Internal
PHalfSat + TSP 1PHalfSatInternal + {PhosRatio - MinPRatio)
(55g)
where:
Phyto UpNutrient
MaxNutrientUptake
NutrientHalfSat
NutrientHalfSatMemai
biomass
le3
uptake of internal nutrients ((_ig of nutrient/L d);
the maximum uptake rate for the nutrient (mg/mg AFDW-d);
half-saturation constant for external nutrient (|_ig nutrient/L);
half-saturation constant for intracellular nutrient (mg/mg AFDW);
algal biomass (mg/L); and
units conversion ([j,g/mg).
Some additional observations about the internal nutrients option follow:
• While the internal-nutrient model allows stoichiometry of plants to vary over time,
animal and suspended-organic-matter stoichiometry remain constant in the model at this
time.
• The internal-nutrient model is not utilized for benthic or rooted macrophytes, which are
assumed to get nutrients from sediments and are not assumed to have nutrient limitation
for that reason.
• Boundary-condition loadings of plants are assumed to have the same nutrient
characteristics as plants currently in the water body.
83
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Current Limitation
Because they are fixed in space, periphyton also are limited by slow currents that do not
replenish nutrients and carry away senescent biomass. Based on the work of Mclntire (1973)
and Colby and Mclntire (1978), a factor relating photosynthesis to current velocity is used for
periphyton:
where:
VLimit = min(7, RedStillWater
VelCoeff ¦ Velocity
-)
VLimit =
RedStillWater =
VelCoeff =
Velocity =
1 + VelCoeff ¦ Velocity
(57)
limitation or enhancement due to current velocity (unitless);
user-entered reduction in photosynthesis in absence of current
(unitless);
empirical proportionality coefficient for velocity (0.057, unitless);
and
flow rate (converted to m/s), see (14).
VLimit has a minimum value for photosynthesis in the absence of currents and increases
asymptotically to a maximum value for optimal current velocity (Figure 58). In high currents
scour can limit periphyton; see (75). The value of RedStillWater depends on the circumstances
under which the maximum photosynthesis rate was measured; if PMax was measured in still
water then RedStillWater = 1, otherwise a value of 0.2 is appropriate (Colby and Mclntire, 1978).
Figure 58. Effect of current velocity on periphyton photosynthesis,
ce
1
LL
m 0.8
0.6
z 0.4
lu 0.2
0
40
100
120
VELOCITY (cm/s)
84
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION CHAPTER 4
Adjustment for Suboptimal Temperature
AQUATOX uses a general but complex formulation to represent the effects of temperature. All
organisms exhibit a nonlinear, adaptive response to temperature changes (the so-called
Stroganov function). Process rates other than algal respiration increase as the ambient
temperature increases until the optimal temperature for the organism is reached; beyond that
optimum, process rates decrease until the lethal temperature is reached. This effect is
represented by a complex algorithm developed by O'Neill et al. (1972) and modified slightly for
application to aquatic systems (Park et al., 1974). An intermediate variable VTis computed first;
it is the ratio of the difference between the maximum temperature at which a process will occur
and the ambient temperature over the difference between the maximum temperature and the
optimal temperature for the process:
(TMax + Acclimation) - Temperature
V1 = (5")
(TMax + Acclimation) - (TOpt + Acclimation)
where:
Temperature = ambient water temperature (deg. C);
TMax = maximum temperature at which process will occur (deg. C);
TOpt = optimal temperature for process to occur (deg. C); and
Acclimation = temperature acclimation (deg. C), as described below.
Acclimation to both increasing and decreasing temperature is accounted for with a modification
developed by Kitchell et al. (1972):
Acclimation = XM ¦ [1 - e(~KT'ABS(T^a'^-™4))] (59)
where:
XM = maximum acclimation allowed (2.0 deg. C);
KT = coefficient for decreasing acclimation as temperature approaches Tref
(value is 0.5 and unitless);
ABS = function to obtain absolute value; and
TRef = "adaptation" temperature below which there is no acclimation (deg. C).
The mathematical sign of the variable Acclimation is negative if the ambient temperature is
below the temperature at which there is no acclimation; otherwise, it is positive.
If the variable VT is less than zero, in other words, if the ambient temperature exceeds (TMax +
Acclimation), then the suboptimal factor for temperature is set equal to zero and the process
stops. Otherwise, the suboptimal factor for temperature is calculated as (Park et al., 1974):
TCorr = VTX1 ¦ e(XT' "4T" (60)
where:
85
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
xr_WT2-(l + Jl + 40/YT f
400
where:
and,
where:
WT = \n(Q10) -((TMax + Acclimation) - (TOpt + Acclimation)) (62)
YT = \n(Q10) ¦ ((TMax + Acclimation) - (TOpt + Acclimation) + 2) (63)
Q10 = slope or rate of change per 10°C temperature change (unitless).
This well-founded, robust algorithm for TCorr is used in AQUATOX to obtain reduction factors
for suboptimal temperatures for all biologic processes in animals and plants, with the exception
of decomposition and plant respiration. By varying the parameters, organisms with both narrow
and broad temperature tolerances can be represented (Figure 59, Figure 60).
86
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION CHAPTER 4
Figure 59. Temperature response of cyanobacteria Figure 60. Temperature response of diatoms
STROGANOV FUNCTION
BLUE-GREENS
TOpt
1
Q10
0.4
"TRef
0.2
TMax
o
o
TEMPERATURE (C)
STROGANOV FUNCTION
DIATOMS
Top'
1
TRef
0.4
0.2
J" Max
o
0
TEMPERATURE (C)
Algal Respiration
Endogenous or dark respiration is the metabolic process whereby oxygen is taken up by plants
for the production of energy for maintenance and carbon dioxide is released (Collins and
Wlosinski, 1983). Although it is normally a small loss rate for the organisms, it has been shown
to be exponential with temperature (Aruga, 1965). Riley (1963, see also Groden, 1977) derived
an equation representing this relationship. Based on data presented by Collins (1980), maximum
respiration is constrained to 60% of photosynthesis. Laboratory experiments in support of the
CLEANER model confirmed the empirical relationship and provided additional evidence of the
correct parameter values (Collins, 1980), as demonstrated by Figure 61:
Respiration = ResplO ¦ 1.045
(T emperature - 20)
• Biomass
(64)
where:
Respiration =
Resp20 =
1.045
Temperature =
Biomass =
dark respiration (g/m -d);
user input respiration rate at 20°C (g/g-d);
exponential temperature coefficient (/°C);
ambient water temperature (°C); and
plant biomass (g/m3).
This construct also applies to macrophytes.
87
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 61. Respiration (Data From Collins, 1980)
DARK RESPIRATION
LU
< 0.6
on
= 0.4
20
TEMPERATURE(C)
Photorespiration
Algal excretion, also referred to as photorespiration, is the release of photosynthate (dissolved
organic material) that occurs in the presence of light. Environmental conditions that inhibit cell
division but still allow photoassimilation result in release of organic compounds. This is
especially true for both low and high levels of light (Fogg et al., 1965; Watt, 1966; Nalewajko,
1966; Collins, 1980). AQUATOX uses an equation modified from one by Desormeau (1978)
that is the inverse of the light limitation:
Excretion = KResp ¦ LightStress ¦ Photosynthesis
where:
Excretion
KResp
Photosynthesis =
(65)
release of photosynthate (g/m3-d);
coefficient of proportionality between excretion and
photosynthesis at optimal light levels (unitless); and
photosynthesis (g/m3-d), see (35),
and where:
where:
LtLimit
LightStress = 1 - LtLimit
light limitation for a given plant (unitless), see (38).
(66)
Excretion is a continuous function (Figure 62) and has a tendency to overestimate excretion
slightly at light levels close to light saturation where experimental evidence suggests a constant
relationship (Collins, 1980). The construct for photorespiration also applies to macrophytes.
88
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 62. Excretion as a fraction of photosynthesis
EFFECT OF LIGHT ON PHOTORESPIRATION
to
<2 0.09
X
^ 0.08
O 0.07
I—
o
X 0.06
CL
O 0.05
O
<
OH
0.04
DIATOMS IN POND
200 250
300 350 400
LIGHT (ly/d)
450 500
Algal Mortality
Nonpredatory algal mortality can occur as a response to toxic chemicals (discussed in Chapter
8) and as a response to unfavorable environmental conditions. Phytoplankton under stress may
suffer greatly increased mortality due to autolysis and parasitism (Harris, 1986). Therefore, most
phytoplankton decay occurs in the water column rather than in the sediments (DePinto, 1979).
The rapid remineralization of nutrients in the water column may result in a succession of blooms
(Harris, 1986). Sudden changes in the abiotic environment may cause the algal population to
crash; stressful changes include nutrient depletion, unfavorable temperature, and damage by light
(LeCren and Lowe-McConnell, 1980). These are represented by a mortality term in AQUATOX
that includes toxicity, high temperature (Scavia and Park, 1976), and combined nutrient and
light limitation (Collins and Park, 1989):
Mortality = (KMort + ExcessT + Stress) ¦ Biomass + Poisoned
(67)
where:
Mortality = nonpredatory mortality (g/m -d);
Poisoned = mortality rate due to toxicant (g/m -d), see (417);
KMort = intrinsic mortality rate (g/g-d); and
Biomass = plant biomass (g/m ),
and where:
and:
ExcessT
(Temperature - TMax)
2
Stress — 1 - e
-EMort ¦ (1 - (NutrLimit ¦ LtLimit))
(68)
(69)
89
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
where:
ExcessT = factor for high temperatures (g/g-d);
TMax = maximum temperature tolerated (° C);
Stress = factor for suboptimal light and nutrients (g/g-d),
Emort = approximate maximum fraction killed per day with total limitation
(g/g-d);
NutrLimit = reduction due to limiting nutrient (unitless), see (55)
LtLimit = light limitation (unitless), see (38).
Exponential functions are used so that increasing stress leads to rapid increases in mortality,
especially with high temperature where mortality is 50% per day at the TMax (Figure 61), and, to
a much lesser degree, with suboptimal nutrients and light (Figure 64). This simulated process is
responsible in part for maintaining realistically high levels of detritus in the simulated water
body. Low temperatures are assumed not to affect algal mortality.
Figure 63. Mortality due to high temperatures
Figure 64. Mortality due to light limitation
iRO.6
I- 0.4
X 0.2
TMax
24 26 28 30 32 34 36 38 40
Temperature
ALGAL MORTALITY
DIATOMS
0.03
0.028
L 0.024
E 0.022
0.02
0.018
200 250 300 350 400 450
LIGHT (ly/d)
Sinking
Sinking of phytoplankton, either between layers or to the bottom sediments, is modeled as a
function of physiological state, similar to mortality. Phytoplankton that are not stressed are
considered to sink at given rates, which are based on field observations and implicitly account
for the effects of averaged water movements (cf. Scavia, 1980). Sinking also is represented as
being impeded by turbulence associated with higher discharge (but only when discharge exceeds
mean discharge):
, KSed MeanDischarge 0 ^ , ,, ,,,
Sink = SeaAccel ¦ Densityractor ¦ Biomass (70)
Depth Discharge
where:
o
Sink = phytoplankton loss due to settling (g/m -d);
90
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
KSed
Depth
MeanDischarge
Discharge
DensityFactor
Biomass
= intrinsic settling rate (m/d);
= depth of water or, if stratified, thickness of layer (m);
= mean annual discharge (m /d);
= daily discharge (m3/d), see Table 3;
= if salinity is modeled, correction factor for water densities based on
salinity and temperature, see (442); and
= phytoplankton biomass (g/m ).
The model is able to mimic high sedimentation loss associated with the crashes of phytoplankton
blooms, as discussed by Harris (1986). As the phytoplankton are stressed by toxicants and
suboptimal light, nutrients, and temperature, the model computes an exponential increase in
sinking (Figure 65), as observed by Smayda (1974), and formulated by Collins and Park (1989):
where:
SedAccel ^ESed • (1 - LtLimit • NutrLimit • TCorr • FracPhoto)
(71)
SedAccel
ESed
LtLimit
NutrLimit
FracPhoto
TCorr
increase in sinking due to physiological stress (unitless);
exponential settling coefficient (unitless);
light limitation (unitless), see (38);
nutrient limitation (unitless), see (55); and
reduction factor for effect of toxicant on photosynthesis (unitless),
see (421);
temperature limitation (unitless), see (59).
Figure 65. Sinking as a function of nutrient stress
SINKING IN POND
DIATOMS, DEPTH = 3 m, BIOMASS = 1
o: 0.14
v 0.12
0.00 0.02 0.03 0.05 0.06 0.08 0.10
PHOSPHATE (g/cu m)
91
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION CHAPTER 4
Washout and Sloughing
Phytoplankton are subject to downstream drift. In streams and in lakes and reservoirs with low
retention times this may be a significant factor in reducing or even precluding phytoplankton
populations (LeCren and Lowe-McConnell, 1980). The process is modeled as a simple function
of discharge:
Disc hciv£€
Wash0Ut phytoplankton = ' Bi°maSS (72)
where:
Washout phytoplankton = loss due to downstream drift (g/m -d),
Discharge = daily discharge (m3/d);
-3
Volume = volume of site (m ); and
Biomass = biomass of phytoplankton (g/m3).
Periphyton often exhibit a pattern of buildup and then a sharp decline in biomass due to
sloughing. Based on extensive experimental data from Walker Branch, Tennessee (Rosemond,
1993), a complex sloughing formulation, extending the approach of Asaeda and Son (2000), was
implemented. This function was able to represent a wide range of conditions better (Figure 66
and Figure 67).
Washout periphyton = Slough + DislodgePeri JOX (73)
where:
Washoutperiphyton = loss due to sloughing (g/m -d);
Slough = loss due to natural causes (g/m -d), see (75); and
Dislodgeperi, Tax = loss due to toxicant-induced sloughing (g/m -d), see (427).
92
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 66. Comparison of predicted biomass of periphyton, constituent algae, and observed biomass of
periphyton (Rosemond, 1993) in Walker Branch, Tennessee, with addition of both N and P and removal
of grazers in Spring, 1989.
—Diatoms(g/m2) —Oth alg{g/m2) ^—Periphyton X Observed
93
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 67. Predicted rates for diatoms in Walker Branch, Tennessee, with addition of both N and P and
removal of grazers in Spring, 1989. Note the importance of periodic sloughing. Rates expressed as g/nr d.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
~ Photosythesis ~ Respiration ~ Excretion ~ Mortality ¦ Predation ¦ Sloughing
Natural sloughing is a function of senescence due to suboptimal conditions and the drag force of
currents acting on exposed biomass. Drag increases as both biomass and velocity increase:
DragForce = Rho ¦ DragCoeff ¦ Vel~ ¦ (BioVol ¦ Unit Area f/s ¦ IE -6
(74)
where:
DragForce
Rho
DragCoeff
Vel
BioVol
UnitArea
1E-6
drag force (kg m/s );
density of water (kg/m3);
drag coefficient (2.53E-4, unitless);
velocity (converted to m/s) see (14);
3 2
biovolume of algae (mm /mm );
unit area (mi#);
2 2
conversion factor (m /mm ).
Biovolume is not modeled directly by AQUATOX, so a simplifying assumption is that the
empirical relationship between biomass and biovolume is constant for a given growth form,
based on observed data from Rosemond (1993):
94
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION CHAPTER 4
Biomass ,
Biovoloia - . /lt, r , ' ZMecm
2.08E-9 (75)
Biomass .
Biovoha ———— • ZMean
8.57E-9
where:
BiovolDia
Biovolfu
Biomass
ZMean
3 2
biovolume of non-filamentous algae (mm /mm );
biovolume of filamentous algae (mmW);
biomass of given algal group (g/m );
mean depth (m).
Suboptimal light, nutrients, and temperature cause senescence of cells that bind the periphyton
and keep them attached to the substrate. This effect is represented by a factor, Suboptimal,
which is computed in modeling the effects of environmental conditions on photosynthesis.
Suboptimal decreases the critical force necessary to cause sloughing. If the drag force exceeds
the critical force for a given algal group modified by the Suboptimal factor and an adaptation
factor, then sloughing occurs:
If DragForce > Suboptimal0rg ¦ FCrit0rg ¦ Adaptation
then Slough = Biomass ¦ FracSloughed
else Slough = 0
(76)
where:
Suboptimal org
FCritorg
Adaptation
Slough
FracSloughed
factor for suboptimal nutrient, light, and temperature effect on
senescence of given periphyton group (unitless);
critical force necessary to dislodge given periphyton group (kg
m/s2);
factor to adjust for mean discharge of site compared to reference
site (unitless);
biomass lost by sloughing (g/m3);
fraction of biomass lost at one time, editable.
where:
Suboptimal0rg = NutrLimit0rg ¦ LtLimit0rg ¦ TCorr0rg ¦ 20
If SuboptimalQrg > 1 then Suboptimal0rg = 1
(77)
NutrLimit
LtLimitorg
TCorr
20
nutrient limitation for given algal group (unitless) computed by
AQUATOX; see (55);
light limitation for given algal group (unitless) computed by
AQUATOX; see (38); and
temperature limitation for a given algal group (unitless) computed
by AQUATOX; see (59).
factor to desensitize construct.
95
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
The sloughing construct was tested and calibrated (U.S. E.P.A., 2001) with data from
experiments with artificial and woodland streams in Tennessee (Rosemond, 1993, Figure 66).
However, in modeling periphyton at several sites, it was observed that sloughing appears to be
triggered at greatly differing mean velocities. The working hypothesis is that periphyton adapt to
the ambient conditions of a particular channel. Therefore, a factor is included to adjust for the
mean discharge of a given site compared to the reference site in Tennessee. It is still necessary
to calibrate FCrit for each site to account for intangible differences in channel and flow
conditions, analogous to the calibration of shear stress by sediment modelers, but the range of
calibration needed is reduced by the Adaptation factor:
Vel2
Adaptation =
0.006634 (78)
where:
Vel = velocity for given site (m/s), see (14);
0.006634 = mean velocity2 for reference experimental stream (m/s).
Detrital Accumulation in Periphyton
In phytoplankton, mortality results in immediate production of detritus, and that transfer is
modeled. However, for purposes of modeling, periphyton are defined as including associated
detritus. The accumulation of non-living biomass is modeled implicitly by not simulating
mortality due to suboptimal conditions. Rather, in the simulation biomass builds up, causing
increased self-shading, which in turn makes the periphyton more vulnerable to sudden loss due
to sloughing. The fact that part of the biomass is non-living is ignored as a simplification of the
model.
Chlorophyll a
Chlorophyll a is not simulated directly. However, because chlorophyll a is commonly measured
in aquatic systems and because water quality managers are accustomed to thinking of it as an
index of water quality, the model converts phytoplankton biomass estimates into approximate
values for chlorophyll a. The ratio of carbon to chlorophyll a exhibits a wide range of values
depending on the nutrient status of the algae (Harris, 1986); cyanobacteria often have higher
values (cf. Megard et al., 1979). Conversion factors between phytoplankton and chlorophyll a
are now editable on a species by species basis within each plants "underlying data." In the
absence of species-specific data, AQUATOX uses default values of 45 jagC/jag chlorophyll a for
cyanobacteria and a value of 28 for other phytoplankton as reported in the documentation for
WASP (Ambrose et al., 1991). The values are more representative for blooms than for static
conditions, but managers are usually most interested in the maxima. Results are presented as
total chlorophyll a in (J,g/L; therefore, the computation is:
96
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
ChlA = YJ
BiOfTlClSSphytoplankton CToOfg
where:
ChlA
Biomass
CToOrg
CToChla
1000
CToChla
Phytoplankton
1000
(79)
estimated biomass as chlorophyll a ((J,g/L);
biomass of given alga (mg/L);
ratio of carbon to biomass (0.526, unitless);
ratio of carbon to chlorophyll a (g carbon/g chl a); and
conversion factor for mg to (_ig (unitless).
Periphytic chlorophyll a is computed as a conversion from the ash-free dry weight (AFDW) of
periphyton; because periphyton can collect inorganic sediments, it is important to measure and
model it as AFDW. The conversion factor is based on the observed average ratio of chlorophyll
a to AFDW for the Cahaba River near Birmingham, Alabama (unpub. data) and also based on
data published in Biggs (1996)and Rosemond (1993).
Perichlor = AFDW -5.0
(80)
where:
PeriChlor
AFDW
periphytic chlorophyll a (mg/m );
ash free dry weight (g/m ).
Moss chlorophyll a is output for all plants designated with the plant type "Bryophytes." In this
case, ash free dry weight is multiplied by 8.91 to get the estimate of chlorophyll a in mg/m2
(Stream Bryophyte Group, 1999, p. 160). Total benthic chlorophyll a is also output in units of
mg/m2 (the sum of periphyton and moss chlorophyll a).
Phytoplankton and Zooplankton Residence Time
Phytoplankton and zooplankton can quickly wash out of a short reach, but they may be able to
grow over an extensive reach of a river, including its tributaries. Somehow the volume of water
occupied by the phytoplankton needs to be taken into consideration. To solve this problem,
AQUATOX takes into account the "Total Length" of the river being simulated, as opposed to the
length of the river reach, or "SiteLength" so that phytoplankton and zooplankton production
upstream can be estimated. This parameter can be directly entered on the "Site Data" screen or
estimated from the watershed area based on Leopold et al. (1964).
TotLength = 1.609 • 1.4 • (Watershed ¦ 0.386)06 (81)
97
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
where:
TotLength =
Watershed =
1.609
0.386
total river length (km);
land surface area contributing to flow out of the reach (square km);
km per mile;
square miles per square km.
If Enhanced Phytoplankton Retention is not chosen (or the total length or watershed area is
entered as zero,) the phytoplankton and zooplankton residence time equations are not used and
Equations (71) and (129) are used to calculate washout. In this case, the phytoplankton
residence time is equal to the retention time of the system.
Otherwise, to simulate the inflow of plankton from upstream reaches plankton upstream loadings
are estimated as follows:
f
Loading upstream = Washout
biota
Washout
\
biota
TotLength / Site Length
(82)
where:
Loading upstream
Washoutblota
TotLength
SiteLength
= loading of plankton due to upstream production (mg/L);
= washout of plankton from the current reach (mg/L);
= total river length (km);
= length of the modeled reach (km).
An integral assumption in this approach is that upstream reaches included in the total river length
have identical environmental conditions as the reach being modeled and that plankton production
in each mile up-stream will be identical to plankton production in the given reach. Residence
time for plankton within the total river length is estimated as follows:
residence
Volume ( TotLength
Discharge \ SiteLength
(83)
where:
tresidence
Volume
Discharge
TotLength
SiteLength
residence time for floating biota within the total river length (d);
volume of modeled segment reach (m3); see (2);
discharge of water from modeled reach (m /d); see Table 3;
total river length (km);
length of the modeled reach (km).
Periphyton-Phytoplankton Link
Periphyton may slough or be physically scoured, contributing to the suspended (sestonic) algae;
this may be reflected in the chlorophyll a observed in the water column. Periphyton may be
linked to a phytoplankton compartment so that sestonic chlorophyll a reflects the results of
periphyton sloughing. One-third of periphyton is assumed to become phytoplankton and two
98
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
thirds is assumed to become suspended detritus in a sloughing event. The default is linkage to
detritus with a warning.
Additionally, when phytoplankton undergoes sedimentation it will now be incorporated into the
linked periphyton layer if such a linkage exists. If multiple periphyton species are linked to a
single phytoplankton species, biomass is distributed to periphyton weighted by the mass of each
periphyton compartment. (A single periphyton compartment cannot be linked to multiple
phytoplankton compartments.)
SedPenplvtonA = SinkPhyto ^aSSpe"P"yto"A (84)
Mass A a LinkedPeri
where:
SedpenphytonA = sedimentation that goes to periphyton compartment A:
Sinkphyto = total sedimentation of linked phytoplankton compartment, see (69);
MassperiPhyton a = mass of periphyton compartment A;
MassAiiLinkedPeri = mass of all periphyton compartments linked to the
relevant phytoplankton compartment.
If no linkage is present, settling phytoplankton are assumed to contribute to sedimented detritus.
4.2 Macrophytes
Submersed aquatic vegetation or macrophytes can be an important component of shallow aquatic
ecosystems. It is not unusual for the majority of the biomass in an ecosystem to be in the form of
macrophytes during the growing season. Seasonal macrophyte growth, death, and
decomposition can affect nutrient cycling, and detritus and oxygen concentrations. By forming
dense cover, they can modify habitat and provide protection from predation for invertebrates and
smaller fish (Howick et al., 1993); this function is represented in AQUATOX (see Figure 73).
Macrophytes also provide direct and indirect food sources for many species of waterfowl,
including swans, ducks, and coots (Jupp and Spence, 1977b).
AQUATOX represents rooted macrophytes as occupying the littoral zone, that area of the bottom
surface that occurs within the euphotic zone (see (11) for computation). Similar to periphyton,
the macrophyte compartment has units of g/m . In nature, macrophytes can be greatly reduced if
phytoplankton blooms or higher levels of detritus increase the turbidity of the water (cf. Jupp and
Spence, 1977a). Because the depth of the euphotic zone is computed as a function of the
extinction coefficient {ZEuphotic = 4.605/Extinct), the area predicted to be occupied by
macrophytes can increase or decrease depending on the clarity of the water.
The macrophyte equations are based on submodels developed for the International Biological
Program (Titus et al., 1972; Park et al., 1974) and CLEANER models (Park et al., 1980) and for
the Corps of Engineers' CE-QUAL-R1 model (Collins et al., 1985):
99
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
and:
dBiomass
dt
Loading + Photosynthesis - Respiration - Excretion
-Mortality - Predation - Breakage
+ WashoutFreeFloat - WashinFreeFloat
(85)
Photosynthesis = PMax ¦ LtLimit ¦ TCorr ¦ Biomass ¦ FracLittoral
¦ NutrLimit ¦ FracPhoto ¦ HabitatLimit
(86)
where:
dBiomass/dt
Loading
Photosynthesis
Respiration
Excretion
Mortality
Predation
Breakage
PMax
LtLimit
TCorr
HabitatLimit
FracLittoral
NutrLimit
FracPhoto
Washout
FreeFloat
Washin
FreeFloat
= change in biomass with respect to time (g/m -d);
= loading of macrophyte, usually used as a "seed" (g/m -d);
2
= rate of photosynthesis (g/m -d);
= respiratory loss (g/m -d), see (63);
= excretion or photorespiration(g/m -d), see (64);
= nonpredatory mortality (g/m -d), see (87);
2
= herbivory (g/m -d), see (99);
= loss due to breakage (g/m -d), see (88);
= maximum photosynthetic rate (1/d);
= light limitation (unitless), see (38);
= correction for suboptimal temperature (unitless), see (59);
= in streams, habitat limitation based on plant habitat preferences
(unitless), see (13);
= fraction of bottom that is in the euphotic zone (unitless) see (11);
= nutrient limitation for bryophytes or freely-floating macrophytes
(unitless), see (55);
= reduction factor for effect of toxicant on photosynthesis (unitless),
see (421);
= washout of freely floating macrophytes, see (86); and
-3
= loadings from linked upstream segments (g/m d), see (30);
They share many of the constructs with the algal submodel described above. Temperature
limitation is modeled similarly, but with different parameter values. Light limitation also is
handled similarly, using the Steele (1962) formulation; the application of this equation has been
verified with laboratory data (Collins et al., 1985). Periphyton are epiphytic in the presence of
macrophytes; by growing on the leaves they contribute to the light extinction for the
macrophytes (Sand-Jensen, 1977). Extinction due to periphyton biomass is computed in
AQUATOX, by inclusion in LtLimit. For rooted macrophytes, nutrient limitation is not modeled
at this time because macrophytes can obtain most of their nutrients from bottom sediments
(Bristow and Whitcombe, 1971; Nichols and Keeney, 1976; Barko and Smart, 1980). Bryophytes
and freely floating macrophytes assimilate nutrients from water and are subject to nutrient
limitation.
100
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Release 3 includes free-floating macrophytes. These macrophytes are assumed to be floating at
the upper layer of the water column and therefore are not subject to light limitation.
Furthermore, free-floating macrophytes are not subject to the FracLittoral limitation to
macrophyte photosynthesis (85). On the other hand the washing of macrophytes out of the
system is affected by the carrying capacity for the species:
WashoUt freefloat
where:
( If! /hi j '/ A A/ i i i /, < / /1
1-
KCap / ZMean - State ) Discharge
\
KCap / ZMean
Volume
¦ ¦ State (87)
Washout freefloat = loss due to being carried downstream (g/m3 -d),
State = concentration of dissolved or floating state variable (g/m ),
KCap = carrying capacity (g/m2);
ZMean = mean depth from site underyling data (m);
Discharge = discharge (m3/d), see Table 3; and
"3
Volume = volume of site (m ), see (2);
Simulation of macrophyte respiration and excretion utilize the same equations as algae; excretion
in rooted macrophytes results in "nutrient pumping" because the nutrients are assumed to come
from the sediments but are excreted to the water column . Non-predatory mortality is modeled
similarly to algae as a function of suboptimal temperature (but not light). However, mortality is
a function of low as well as high temperatures, and winter die-back is represented as a result of
this control; the response is the inverse of the temperature limitation (Figure 68):
Mortality = [KMort + Poisoned + (1- e~EMort'(1 ~TCorr) ^)j. Biomass (88)
where:
KMort = intrinsic mortality rate (g/g-d);
Poisoned = mortality rate due to toxicant (g/g-d) (417), and
EMort = maximum mortality due to suboptimal temperature (g/g-d).
Sloughing of dead leaves can be a significant loss (LeCren and Lowe-McConnell, 1980); it is
simulated as an implicit result of mortality (Figure 68).
2 Because nutrients are not usually explicitly modeled in bottom sediments, macrophyte root uptake can result in
loss of mass balance, particularly in shallow ponds. The optional sediment diagenesis model does include nutrients
but linkage to macrophytes through root uptake has not yet been specified and implemented. However, the total
mass of nutrients taken into the water column through macrophyte uptake can be tracked as a model output (N and P
"Root Uptake" in kg).
101
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 68. Mortality as a function of temperature
MACROPHYTE MORTALITY
> 0.06
< 0.04
01
O 0.02
10 20 30 40
TEMPERATURE (C)
Macrophytes are subject to breakage due to higher water velocities; this breakage of live material
is different from the sloughing of dead leaves. Although breakage is a function of shoot length
and growth form as well as currents (Bartell et al., 2000; Hudon et al., 2000), a simpler construct
was developed for AQUATOX (Figure 69):
Velocity - VelMax
where:
Breakage
Velocity
VelMax
Gradual
UnitTime
Biomass
Breakage
Gradual ¦ UnitTime
- ¦ Biomass
(89)
macrophyte breakage (g/m -d);
current velocity (cm/s) see (14);
velocity at which total breakage occurs (cm/s);
velocity scaling factor (20 cm/s);
unit time for simulation (1 d);
macrophyte biomass (g/m2).
Figure 69. Breakage of macrophytes as a function of current
velocity; VelMax set to 300 cm/s.
N n£"
Velocity (cm/s)
©
The Breakage formulation also applies to freely floating macrophytes and may be considered
entrainment in periods of high flow. As such, VelMax should be set to a relatively high value for
these organisms.
102
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Bryophytes (mosses and liverworts) are a special class of macrophytes that attach to hard
substrates, are stimulated by and take up nutrients directly from the water, are resistant to
breakage, and decompose very slowly (Stream Bryophyte Group, 1999). Nutrient limitation is
enabled when the "Bryophytes" plant type is selected, just as it is for algae. The model assumes
that when a bryophyte breaks or dies the result is 75% particulate and 25% dissolved refractory
detritus; in contrast, other macrophytes are assumed to yield 62% labile detritus. All other
differences between bryophytes and other macrophytes in AQUATOX are based on differences
in parameter values. These include low saturating light levels, low optimum temperature, very
low mortality rates, moderate resistance to breakage, and resistance to herbivory (Arscott et al.,
1998; Stream Bryophyte Group, 1999). Because in the field it is difficult to separate bryophyte
chlorophyll from that of periphyton, it is computed so that the two can be combined and related
to field values:
MossChlor = ^(BryoCotiv ¦ Biomciss(90)
where:
MossChlor = bryophytic chlorophyll a (mg/m );
2 2
BryoConv = conversion from bryophyte AFDW to chlorophyll a (8.9 mg/m : g/m );
Biomasssryo = biomass of given bryophyte (AFDW in g/m2).
Currents and wave agitation can both stimulate and retard macrophyte growth. These effects
will be modeled in a future version. Similar to the effect on periphyton, water movement can
stimulate photosynthesis in macrophytes (Westlake, 1967); the same function could be used for
macrophytes as for periphyton, although with different parameter values. Jupp and Spence
(1977b) have shown that wave agitation can severely limit macrophytes; time-varying breakage
eventually will be modeled when wave action is simulated.
4.3 Animals
Animals: Simplifying Assumptions
• Ingestion is represented by a maximum consumption rate adjusted for conditions of food, temperature, sublethal toxicant
effects, and habitat preferences
• Reproduction is implicit in the increase in biomass
• Macrophytes can provide refuge from predation
• AQUATOX is a food-web model including prey switching based on prey availability
• Specific dynamic action (the metabolic "cost" of digesting and assimilating prey) is represented as proportional to food
assimilated
• Unless spawning dates are entered by the user, spawning occurs as a function of water temperature
• Zooplankton and fish will migrate vertically from an anoxic hypolimnion to the epilimnion
• Promotion from one size class of fish to the next is estimated as a fraction of total biomass growth
Zooplankton, benthic invertebrates, benthic insects, and fish are modeled, with only slight
differences in formulations, with a generalized animal submodel that is parameterized to
represent different groups:
103
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
dBiomass = j ()Cl(j + Consumption - Defecation - Respiration - Fishing
dt
- Excretion -Mortality - Predation - GameteLoss ± DiffusionSeg
- Washout + Washin ± Migration - Promotion + Recruit - Entrainment
(91)
GrowthRate = Consumption - Defecation - Respiration - Excretion
where:
dBiomass/dt =
Load =
Consumption =
Defecation =
Respiration =
Fishing =
Excretion =
Mortality =
Predation =
GameteLoss =
Washout =
Washin =
Diffusionseg =
Migration =
Promotion =
Recruit =
Entrainment =
GrowthRate =
change in biomass of animal with respect to time (g/m -d);
biomass loading, usually from upstream, or calculated from user-
"3
supplied fish stocking data (g/m -d);
"3
consumption of food (g/m -d), see (98);
"3
defecation of unassimilated food (g/m -d), see (97);
"3
respiration (g/m -d), see (100);
"3
loss of organism due to fishing pressure (g/m -d), user input fraction
fished multiplied by the biomass.
"3
excretion (g/m -d), see (111);
nonpredatory mortality (g/m -d), see (112);
"3
mortality from being preyed upon (g/m -d), see (99);
"3
loss of gametes during spawning (g/m -d), see (126);
-3
loss due to being carried downstream by washout and drift (g/m -d),
see (129) and (130);
loadings from linked upstream segments (g/m3 d), see (30);
gain or loss due to diffusive transport over the feedback link between
"3
two segments, pelagic inverts, only (g/m -d), see (32);
"3
loss (or gain) due to vertical migration (g/m -d), see (133);
"3
promotion to next size class or emergence (g/m -d), see (136);
"3
recruitment from previous size class (g/m -d), see (128);
"3
entrainment and downstream transport by floodwaters (g/m -d)
(132).
estimated growth rate as a function of derivative terms, output in
units of percentage per day when animal's "rates output" is turned
on.
The change in biomass (Figure 70) is a function of a number of processes (Figure 71) that are
subject to environmental factors, including biotic interactions. Similar to the way algae are
treated, parameters for different species of invertebrates and fish are loaded and available for
editing by means of the entry screens. Biomass of zoobenthos and fish is expressed as g/m2
"3
instead of g/m .
Growth rates have been part of AQUATOX output since Release 3.0. However, this output has
always been reported in units of "percent (of current biomass) per day." To better understand
total predicted secondary and tertiary productivity, and to effectively compare model results
against literature estimates, as of Release 3.2 growth rates are also output in "g AFDW/ m2-day."
104
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 70. Predicted changes in biomass in a stream
Cahaba River AL (CONTROL)
Run on 03-8-08 10:13 AM
2 26 2000 8/26/2000 2/24/2001 8/25/2001 2/23/2002 8/24/2002
Obs Corbicula (g/m2dry)
¦ Caddisfly,Tric (g/m2 dry)
- Corbicula (g/m2 dry)
Mussel (g/m2 dry)
Obs Snails (g/m2 dry)
Mayfly (Baetis (g/m2 dry)
Gastropod (g/m2 dry)
Obs Mayfly (g/m2 dry)
Obs Caddis fly (g/m2 dry)
Figure 71. Predicted Process Rates for the Invasive Clam Corbicula, Expressed as Percent of Biomass;
Yellow Spikes are Entrainment During Storm Events; Consumption Depends on Sloughing Periphyton..
Corbicula Load (Percent)
Corbicula Consumption (Percent)
Corbicula Defecation (Percent)
Corbicula Respiration (Percent)
Corbicula Excretion (Percent)
Corbicula Scour_Entrain (Percent)
Corbicula Predation (Percent)
Corbicula Mortality (Percent)
Corbicula GameteLoss (Percent)
18.0
4->
c
(LI
u 15.0
(LI
Q.
12.0
Cahaba River AL (CONTROL)
Run on 03-8-08 10:13 AM
12/5/2000
12/5/2001
12/5/2002
105
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Consumption, Defecation, Predation, and Fishing
Several formulations have been used in various models to represent consumption of prey,
reflecting the fact that there are different modes of feeding and that experimental evidence can be
fit by any one of several equations (Mullin et al., 1975; Scavia, 1979; Straskraba and Gnauck,
1985).
Ingestion is represented in AQUATOX by a maximum consumption rate, adjusted for ambient
food, temperature, oxygen, sediment, and salinity conditions, and reduced for sublethal toxicant
effects and limitations due to habitat preferences of a given predator:
Ingestion preypred = CMaxpred • SatFeeding ¦ TCorrpred • FoodDilution
(92)
• HabitatLimit ¦ ToxReduction ¦ HarmSS ¦ SaltEffect ¦ OlEffectFrac ¦ Biomasspred
where:
"3
IngestionPrey, pred = ingestion of given prey by given predator (g/m -d);
"3
Biomass = concentration of organism (g/m -d);
CMax = maximum feeding rate for predator (g/g-d);
SatFeeding = saturation-feeding kinetic factor, see (93);
TCorr = reduction factor for suboptimal temperature (unitless), see Figure 59;
FoodDilution = factor to account for dilution of available food by suspended
sediment (unitless), see (120);
ToxReduction = reduction due to effects of toxicant (see (424), unitless); and
HarmSS = reduction due to suspended sediment effects (see (116), unitless);
SaltEffect = effect of salinity on ingestion rate (unitless), see (440);
02EffectFrac = effect of reduced oxygen on ingestion (unitless), see (205); and
HabitatLimit = in streams, habitat limitation based on predator habitat preferences
(unitless), see (13).
The maximum consumption rate is sensitive to body size, so an alternative to specifying CMax
for fish is to compute it using an allometric equation and parameters from the Wisconsin
Bioenergetics Model (Hewett and Johnson, 1992; Hanson et al., 1997):
CMax = CA ¦ Mean WeightCB (93)
where:
CA = maximum consumption for a 1-g fish at optimal temperature (g/g-d);
MeanWeight = mean weight for a given fish species (g);
CB = slope of the allometric function for a given fish species.
Many animals adjust their search or filtration in accordance with the concentration of prey;
therefore, a saturation-kinetic term is used (Park et al., 1974, 1980; Scavia and Park, 1976):
Preference , ¦ Food
SatFeedinz = - prey, pred (94\
Lpreyf Preference prey pred • Food) + FHaljSatpred
106
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
where:
Preference
Food
FHalfSat
preference of predator for prey (unitless);
available biomass of given prey (g/m3);
= half-saturation constant for feeding by a predator (g/m ).
The food actually available to a predator may be reduced in two ways:
Food = (Biomassprey - BMinpred) • RefugeMacro ¦ RefugeQyster ¦ RefugeMarsh ¦ Refitge
Burrow
(95)
where:
BMin =
Refl,ge Oysler ~
RefugeMarsh
Refuges urrow
minimum prey biomass needed to begin feeding (g/m ); and
reduction factor for prey hiding in macrophytes or seagrass (unitless),
see (95);
reduction factor for prey hiding among oysters (unitless), see (95);
reduction factor for prey hiding in marsh edge; may also enhance
predation if marsh is disintegrating (unitless), see (95b);
reduction factor for prey's capability to burrow into sediment,
characterized by "BurrowIndex" parameter (unitless), see (95c).
Search or filtration may virtually cease below a minimum prey biomass {BMin) to conserve
energy (Figure 72), so that a minimum food level is incorporated (Parsons et al., 1969; Steele,
1974; Park et al., 1974; Scavia and Park, 1976; Scavia et al., 1976; Steele and Mullin, 1977).
However, some filter feeders such as cladocerans (for example, Daphnia) must constantly filter
because the filtratory appendages also serve for respiration; therefore, in these animals there is
no minimum feeding level and BMin is set to 0.
Figure 72. Saturation-kinetic consumption
BASS CONSUMPTION
BASS BIOMASS = 1 g/cu m
0.03 i—CMax-
0.025
0.015
U 0.005
2.65 5.3 7.95 10.6 13.25 15.9 18.55
PREY BIOMASS (g/cu m)
Refuge from Predation
Although AQUATOX is an ecosystem biomass model, it has some capability to represent habitat
characteristics. In particular, the model can account for the function of macrophytes or seagrass
107
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
as a habitat structure that provides refuge from predation. This refuge is represented by a factor
related to the macrophyte biomass that is original with AQUATOX (Figure 73):
Refuge,
where:
HalfSat =
BiomassMacro =
BiomassoySter =
Macro/ Oyster
1--
Biomass
Macro/ Oyster
Biomass
1Macro/ Oyster
HalfScit
half-saturation constant (20 g/m ),
biomass of macrophyte (g/m ), or
biomass of oyster (g/m ).
(96)
Figure 73. Refuge from predation
100 200 300
MACROPHYTE BIOMASS
400
Similarly, oyster beds provide structural refuge whereby smaller animals may hide in crevices
and dead shells. Using the same equation (95), but based on biomass of oysters with a nominal
HalfSat value of 80 g/m3, we get a slightly different form of reduction in the predator-prey
relationship (Figure 74). The parameter may be fine-tuned based on published studies (Lenihan
1999, Grabowski and Powers 2004).
A boolean parameter "Can Seek Refuge" in the animal underlying data specifies whether an
animal can seek refuge within macrophytes or oyster beds. Many fish and invertebrates can, but
not infauna. Similarly, a boolean parameter "/.v a Visual Feeder" in the animal underlying data
specifies whether an animal's prey can effectively seek refuge.
108
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION CHAPTER 4
Figure 74. Refuge from predation
1
0.9
0.8
*: 0.7
0.6
0.5
£ 0.4
ai
§> 0.3
2 0.2
0.1
0
0
100
200
300
400
500
Biomass
Refuge(Macrophyte) Refuge(Oysters)
Fractal Dimension of Marsh Edge
Another type of refuge is provided by tidal creeks and the diffuse interface between marsh and
water. The irregularity of the saltmarsh-water interface as a refuge capacity can be represented as
a fractal dimension. This construct was originally developed for the Sea Level Affecting Marshes
Model (SLAMM) (Park et al. 1989) and was approved by a peer review panel in 1990
(http://www2.epa.gov/exposure-assessment-models/peer-review-aquatox). Parameter values for
brown and white shrimp were based on catch statistics provided by the National Marine Fisheries
Service Galveston Lab (see also Zimmerman et al. 1991) and results were reported in
Congressional testimony by Park (1991).
If the shoreline-marsh interface were a straight line it would have a fractal dimension of 1.0,
corresponding to the Euclidean integer dimension; this is characteristic of coastlines subjected to
erosional retreat. Healthy marshes have a fractal dimension in excess of 1.0 (intermediate
between the Euclidean dimensions of a line and a surface). Examples of marsh areas with
calculated fractal dimensions are taken from Grand Bay on the border between Mississippi and
Alabama (Figure 75). With inundation due to a relative rise in sea level, as the interface
becomes more irregular due to disintegration of the marsh, the fractal dimension becomes larger.
Eventually, the marsh may break up into scattered remnants with a fractal dimension <1.0
(intermediate between the Euclidean dimensions for a line and a point). These relationships,
with the concomitant increase in refuge and even the loss of refuge, can be represented by a
slightly different equation.
Refuge
Marsh
1 + Coeff
FractalD Ma,Sh + Coeff
(95b)
109
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
where:
FractctlDMarsh = fractal dimension of marsh-water interface;
Coeff = fractal dimension Refuge coefficient (-0.5 to 100 with the lowest values
providing the strongest Refuge effect).
Refuge(Marsh)
Coefficient of -0.15
2
1.8
1.6
1.4
1.2
1
0.8
0.6
0.5
0.75
Fractal Dimension of Marsh Edge
1.25
1.5
Figure 1. Refuge Factor as a Function of Marsh-Edge Fractal Dimension
The fractal dimension is computed based on the co-occurrence of marsh and water in the same
GIS cells. Figure 75 shows three examples of computed fractal dimensions.
110
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 75. Examples of marsh areas with calculated fractal dimensions from Grand Bay, MS
M ' 5 V • V \
MHRBf • 1^3# ^ *
t
FD=1.52
FD=1.34
A \2> FD=1.33
wm
7 \
* i 'tR oiij^ Rigoiets island
:h>ar FD=1.47
- ppf
j m *
¦ -H\
v «
y
'i E
FD=1.01
} South Rigolels
Burrowing Refuge for Invertebrates
Many marine benthic invertebrates have the capability to elude predators due to burrowing. To
ensure that organisms with deep burrows are not subject to excess predation a burrowing refuge
construct has been added to AQUATOX.
Refuge
£
= 1-
Burrowlndex
Barrowlndex + HaljSatSmmw j
(95c)
111
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
where:
Re fuge Burrow
Burrowlndex =
Ha lfSat Burrow
unitless multiplier to represent burrowing refuge, see (95);
animal-specific parameter (unitless) with 0 representing no burrowing refuge;
half saturation coefficient for burrowing (unitless), set to 3.2.
The Burrowlndex parameter may be calibrated on an animal-specific basis to best represent the
extent of burrowing that reduces predation pressure on an organism (Figure 2). Some example
starting points for calibration are listed below:
• Blue Crab (Callinectes) is a limited burrower so it would have a Burrowlndex of 0.5;
• Sand Crab (Emeritd) creates shallow burrows so it would have a Burrowlndex of 1.0;
• Donax may be assumed to have more burrowing refuge than Emerita so it would have a
Burrowlndex of 2.0;
• Ghost Shrimp (Ccillichirus) are deep burrowers so they would have the maximum
Burrowlndex of 5.0.
Refuge(Burrow)
§
0.8
o
k_
3
0.6
-Q
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
The preference factors are normalized so that if a potential food source is not modeled or is
below the BMin value, the other preference factors are modified accordingly, representing
adaptive preferences:
Pref ,
J prey, pred
irvjvrvr^pnd- —
where:
"isSf (97)
PreferencepreyiPred = normalized preference of given predator for given prey
(unitless);
PrefPrey, pred = initial preference value from the animal parameter screen
(unitless); and
SumPref = sum of preference values for all food sources that are present
above the minimum biomass level for feeding during a
particular time step (unitless).
Similarly, different prey types have different potentials for assimilation by different predators.
The fraction of ingested prey that is egested as feces or discarded (and which is treated as a
source of detritus by the model, see (153) and (154)), is indicated by a matrix of egestion
coefficients with the same structure as the preference matrix, so that defecation is computed as
(Park et al., 1974):
Defecationpred = Zprey ( EgestCoeff prey pred ¦ Ingestion prey pred + IncrEgest ¦ IngestNoTox ) (98)
where:
-3
Defecation pred = total defecation for given predator (g/m -d);
"3
Ingestion prey, pred = ingestion of given prey by given predator (g/m d) (91);
EgestCoeJfprsy; pred = fraction of ingested prey that is egested (unitless); and
IncrEgest = increased egestion due to toxicant (see Eq. (425), unitless);
Ingesl voiox = ingestion excluding toxic effects, calculated as Ingestion
"3
divided by ToxReduction (see Eq. (424), g/m -d).
Consumption of prey for a predator is also considered predation or grazing for the prey.
Therefore, AQUATOX represents consumption as a source term for the predator and as a loss
term for the prey:
Consumption pred = X prey (Ingestion prey pred) (99)
Predation prey = Z pred
(Ingestion preypred) (100)
where
"3
Consumptionpred = total consumption rate by predator (g/m -d); and
Predationprsy = total predation on given prey (g/m -d).
Fishing pressure is represented simply as a fraction of biomass removed each day. A potential
future model enhancement could allow for temporally variable fishing pressures to better reflect
harvesting seasons. Fish may also be stocked within a modeled system by entering time series in
grams per day or grams per meter squared per day.
113
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Respiration
Respiration can be considered as having three components (Cui and Xie, 2000), subject to the
effects of salinity:
Respirationpred = [StdResp pred + ActiveResp pred + SpecDynActionpreJ ¦ SaltEffect (101)
where:
"3
Respiratioripred = respiratory loss of given predator (g/m -d);
"3
StdResppred = basal respiratory loss modified by temperature (g/m -d); see
(101);
respira
metabc
SaltEffect = effect of salinity on respiration (unitless), see (440).
"3
ActiveRespPred = respiratory loss associated with swimming (g/m -d), see (104);
-3
SpecDynActionpred = metabolic cost of processing food (g/m -d), see (110); and
Standard respiration is a rate at resting in which the organism is expending energy without
consumption. Active respiration is modeled only in fish and only when allometric (weight-
dependent) equations are used, so standard respiration can be considered as a composite
"routine" respiration for invertebrates and in the simpler implementation for fish. The so-called
specific dynamic action is the metabolic cost of digesting and assimilating prey. AQUATOX
simulates standard respiration as a basal rate modified by a temperature dependence and, in fish,
a density dependence (see Kitchell et al., 1974):
StdResppred = BasalResppred ¦ TCorr pred ¦ Biomasspred ¦ DensityDep (102)
where:
BasalRespprs(i = basal respiration rate at optimal temperature for given predator
(g/g-d); parameter input by user as "Respiration Rate" or computed
as a function of the weight of the animal (see below);
TCorr pred = Stroganov temperature function (unitless), see Figure 59;
Biomasspred = concentration of predator (g/m3); and
DensityDep = density-dependent respiration factor used in computing standard
respiration, applicable only to fish (unitless). See (109)
As an alternative formulation, respiration in fish or invertebrates can be modeled as a function of
the weight of the fish using an allometric equation (Hewett and Johnson, 1992; Hanson et al.,
1997):
StdResppred = BasalResppred ¦ MeanWeight predB""d ¦ TFnpred ¦ Biomasspred ¦ DensityDep (103)
where:
MeanWeight pred = mean weight for a given fish (g);
RBpred = slope of the allometric function for a given fish;
114
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
//'//prod = temperature function (unitless).
The allometric functions are based on the well known Wisconsin Bioenergetics Model and, for
convenience, use the published parameter values for that model (Hewett and Johnson, 1992;
Hanson et al., 1997). Weight-based bioenergetic functions have been extended to invertebrates
by several authors (for example, Brylawski and Miller 2003, Adamack et al. 2012) and that
capability is available as an enhancement in the present version of AQUATOX.
The basal respiration rate in that model is expressed as g of oxygen per g organic matter of fish
per day, and this has to be converted to organic matter respired:
BasalResp pred = RApred-1-5 (104)
where:
RAPred = basal respiration rate, characterized as the intercept of the allometric
mass function in the Wisconsin Bioenergetics Model documentation
(g 02/g organic matter -d);
1.5 = conversion factor (g organic matter/g O2).
Swimming activity may be large and variable (Hanson et al., 1997) and is subject to calibration
for a particular site, considering currents and other factors:
ActiveResppred = Activitypmd ¦ Biomasspred (105)
where:
ActivityPred = activity factor (g/g-d).
Activity can be a complex function of temperature. The Wisconsin Bioenergetics Model
(Hewett and Johnson, 1992; Hanson et al., 1997) provides two alternatives. Equation Set 1 uses
an exponential temperature function:
TFn = e(RQ'Temp) (106)
where:
RQ = the Q10 or rate of change per lOdeg. C for respiration (1/deg. C);
Temp = ambient temperature (deg. C).
This is coupled with a complex function for swimming speed as an allometric function of
temperature (Hewett and Johnson, 1992; Hanson et al., 1997):
Activity pred = e(RT0,Vel)
If Temp > RTL Then Vel = RKl ¦ MeanWeightRK4 (107)
Else Vel = ACT ¦ MeanWeightRK4 ¦ e(BACT' Temp)
where:
RTO = coefficient for swimming speed dependence on metabolism (s/cm);
115
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
RTL = temperature below which swimming activity is an exponential function of
temperature (deg. C);
Vel = swimming velocity (cm/s);
RK1 = intercept for swimming speed above the threshold temperature (cm/s);
RK4 = weight-dependent coefficient for swimming speed;
ACT = intercept for swimming speed for a 1 g fish at deg. C (cm/s); and
BACT = coefficient for swimming at low temperatures (1/deg. C),
Equation Set 2 uses the Stroganov function used elsewhere in AQUATOX:
TFn = TCorr (108)
and activity is a constant:
Activity = ACT (109)
where:
TCorr = reduction factor for suboptimal temperature (unitless), see (59);
ACT = activity factor, which is not the same as ACT'm Equation Set 1 (g/g-d).
Respiration in fish increases with crowding due to competition for spawning sites, interference in
feeding, and other factors. This adverse intraspecific interaction helps to constrain the
population to the carrying capacity; as the biomass approaches the carrying capacity for a given
species the respiration is increased proportionately (Kitchell et al., 1974):
, IncrResp ¦ Biomass
DensityDep = 1 + - (110)
KCap / ZMean
where:
IncrResp = increase in respiration at carrying capacity (0.5);
KCap = carrying capacity (g/m2);
ZMean = mean depth from site underyling data (m).
With the IncrResp value of 0.5, respiration is increased by 50% at carrying capacity (Kitchell et
al., 1974), as shown in Figure 76. This density-dependence is used only for fish, and not for
invertebrates.
Prior to AQUATOX Release 3.2, the benthic invertebrate "carrying capacity" parameter has had
little impact on simulations. With some marine-benthic invertebrate species, (oysters, for
example) available substrate becomes a limiting factor. For this reason, the model now enforces
a hard-cap at the benthic invertebrate carrying capacity by increasing mortality if the biomass
exceeds that level:
M°rtKCap = m[n(Bi0mClSSBenthos ~ KCaPBenthos , 0) (109b)
where:
MortKCap = mortality of zoobenthos due to exceeding available substrate (g/m2);
116
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
BiomctssBenthos = biomass of zoobenthos (g/m2); and
KCapBenthos = user input carrying capacity for this site and zoobenthos species (g/m2).
Figure 76. Density-dependent factor for increase in respiration as fish
biomass approaches the carrying capacity (10.0 in this example).
Biomass (g/m3)
As a simplification, specific dynamic action is represented as proportional to food assimilated
(Hewett and Johnson, 1992; see also Kitchell et al., 1974; Park et al., 1974):
SpecDynAction , = KResp , • (Consumption , - Defecation ,)
r s pred ± vred ^ ± vred J vred/
pred
pred
pred'
(111)
where:
KRe Sppred
proportion of assimilated energy lost to specific dynamic action
(unitless); parameter input by user as "Specific Dynamic Action;"
Consumptionpred = ingestion (g/m -d) see (98); and
3
Defecationpred = egestion of unassimilated food (g/m -d), see (97).
Excretion
As respiration occurs, biomass is lost and nitrogen and phosphorus are excreted directly to the
water (Home and Goldman 1994); see (169) and (183). Ganf and Blazka (1974) have reported
that this process is important to the dynamics of the Lake George, Uganda, ecosystem. Their
data were converted by Scavia and Park (1976) to obtain a proportionality constant relating
excretion to respiration:
where:
Excretion prec/
KExCVpred
Respirationp,ed
Excretionp,-ed ~ KExcr pred' Respiration
pred
(112)
excretion rate (g/m -d);
proportionality constant for excretion:respiration (unitless);
respiration rate (g/m -d), see (100).
117
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Excretion is approximately 17 percent of respiration, which is not an important biomass loss
term for animals, but it is important in nutrient recycling. All biomass lost due to animal
excretion is assumed to convert to dissolved labile detritus, see (151).
Nonpredatory Mortality
Nonpredatory mortality is a result of both environmental conditions and the toxicity of
pollutants:
Mortality pred = Dpred ¦ Biomasspred + Poisonedpmd+MortAmm0nia
+ MortLo,02 + M°rtSedEffectS + M°rtSatinrty + M°rtKCap ^ ^
where:
Mortality pred = nonpredatory mortality (g/m -d);
Dpred = environmental mortality rate; the maximum value of (113) and
(114), is used (1/d);
Biomasspred = biomass of given animal (g/m3);
"3
Poisoned = mortality due to toxic effects (g/m -d), see (417);
"3
MortAmmonia = ammonia mortality, (g/m -d), see (179);
-3
MortLowo2 = low oxygen mortality, (g/m -d), see (203);
MortsedEffects = mortality from suspended sediments, (g/m -d), see (115)
Mortsalinity = mortality from salinity, (g/m -d), see (112); and
MortKCap = mortality from benthic invertebrate exceeding available substrate,
(g/m3-d), see (109b).
Under normal conditions a baseline mortality rate is used:
D pred = KMort pred (114)
where:
KMort pred = normal nonpredatory mortality rate (1/d).
An exponential function is used for temperatures above the maximum (Figure 77):
!Temperature - TMaxpred
Dpred KMort pred + ^
where:
Temperature = ambient water temperature (°C); and
TMaxpred = maximum temperature tolerated (°C).
118
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 77. Mortality as a function of temperature
MORTALITY OF BASS
V
1 T
<
a
tr
O
co
mi
Q.
D
III
0.6
I
_l
04
O
1—
O
0.2
<
Cf
LL
0 ^
TMax
3 6 9 12 15 18 21 24 27 30 33 36 39 42
TEMPERATURE
The lower lethal temperature is often 0°C (Leidy and Jenkins, 1976), so it is ignored at this time.
Stocking and Harvesting of Animals
Given the importance of anthropogenic impacts to both fish and invertebrate populations within
the nearshore marine environment, an interface has been added to allow time-series stocking or
... 9
removal for all animals. These data may be input in units of g/(m day) (stocking or removal) or
percent/day (removal).
The result of fish stocking is an increase in the "load" portion of equation (90). The result of fish
removal is a decrease in the "load" rate that can become a negative number. The input of
"stocked" organisms is assumed to have no organic-chemical burden. Removal of toxicants or
nutrients from the system due to fishing or invertebrate removal is tracked with the "fishing loss"
mass-balance tracking variables. This procedure ensures the accounting for all mass balances of
nutrients and chemicals.
Suspended Sediment Effects
The approach used to quantify lethal and
sublethal effects of suspended sediments is
based on logarithmic models described by
Newcombe (for example, Newcombe 2003).
Summary of Sediment Effects:
• Mortality
• Reduction in feeding
• Dilution of food by sediment particles
• Stimulation of invertebrate drift
• Loss of spawning and protective habitat
interstices
They take the form of:
LethalSS = Slope SS ¦ In (XV) + InterceptSS + SIopeTime ¦ \ri(TExp)
(116)
119
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
where:
LethalSS
SlopeSS
ss
InterceptSS
SlopeTime
TExp
cumulative fraction killed by given exposure to a given suspended
sediment concentration (fraction/d)
slope for sediment response (unitless)
suspended inorganic sediment concentration (mg/L)
intercept for suspended sediment response (unitless)
slope for duration of exposure (unitless)
duration of exposure (d)
Unfortunately, there is a dearth of quantitative data on response to sediments. Therefore, the
responses are grouped according to sensitivity, and parameters for surrogate species are used.
The user can specify different parameter values; the values given below are provided as defaults.
For sublethal effects, avoidance behavior is noted at SS of about 100 mg/L (Doisy and Rabeni
2004); however, this could only be used as a cue for migration in the model and has been ignored
at this time.
Reduction in feeding occurs in game fish due to visual impairment (Crowe and Hay 2004). SS
of 25 mg/L seems to be threshold for response (Rowe et al. 2003). The general equation (115) is
used to represent a decrease in food due to turbidity, but without the exposure factor because the
response is instantaneous:
HarmSS = SlopeSS ¦ In (XV) + InterceptSS
(117)
where:
HarmSS
SlopeSS
SS
InterceptSS =
reduction factor for impairment of visual predation (unitless)
slope for suspended sediment response (-0.36, unitless)
suspended inorganic sediment concentration (mg/L). If TSS is
modeled see (244) otherwise, the sum of inorganic sediments in
the water column (e.g. Sand Silt Clay)\
intercept for suspended sediment reponse (2.11, unitless)
The equation is parameterized using data for coho salmon with 1-hr exposure (Berry et al. 2003).
It was verified with numerous other qualitative observations for salmon, Arctic grayling, and
trout (Berry et al. 2003). This equation is used for all visual-feeding fish, especially game fish.
The user has the option of turning on this factor
Figure 78. Reduction in feeding by coho salmon (Oncorhynchus kisutch)
due to suspended sediments. Data from (Berry et al. 2003).
120
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Reduced Feeding in Salmon
y = -0.3617Ln(x) + 2.11
¦2 0.8
R = 0.97
o
3
"§ 0.6
co _ .
5000 mg/L SS. Generally, these
are benthic species exposed to the flocculent zone and bottom sediments. The general equation
(115) is parameterized to accommodate the 24-hr lethality observations and is extended to other
times of exposure by fitting to observed 48-hr lethal responses:
LethalSS = 1.62 • In (SS) - 14.2 + 3.5 • In (TExp) (118)
where:
LethalSS = cumulative fraction killed by given exposure to a given suspended
sediment concentration (fraction/d)
TExp = time of exposure to given level of suspended sediment (d)
SS = minimum suspended inorganic sediment concentration over
exposure time (mg/L). If TSS is modeled see (244) otherwise, the
sum of inorganic sediments in the water column (e.g.
Sand+Silt+Clay).
The parameters are based on the benthic estuarine fish spot (Leiostomus xantfmrus), using data
compiled in Berry et al. (2003).
121
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Due to lack of data beyond 48 hours, this equation is applied using one- and two-day exposure
times only. The maximum effect is chosen from these two equation results.
Figure 79. Lethality of suspended sediments to spot (Leiostomus xanthurus),
a tolerant species, based on data compilation of Berry et al. (2003).
Spot Mortality
1.00
0.80
* 0.60
"" 0.20
0.00
0
2000
4000
6000
8000 10000 12000
SS (mg/L)
~ SS 24-hr ¦ SS 48-hr ¦ SS 24-hr Est
~ SS 48-hr Est Log. (SS 24-hr) Log. (SS 48-hr)
Sensitive
This category represents those species having 250 mg/L < 24-hr LCio <5000 mg/L SS. Small
estuarine species seem to be highly sensitive to suspended sediment (Figure 84). The general
parameters are based on a composite fit to data for bay anchovy, menhaden, and Atlantic
silversides taken from a compilation by Berry et al. (2003). The equation is:
LethalSS = 0.34-ln(XS) - 1.85 + O.l-ln(TExp) (119)
This equation is applied using one- and two-day exposure times along with effects from one,
two, and three weeks exposure. The maximum effect is chosen from these multiple calculations.
Figure 80 illustrates the response curve for white perch. The equation exhibits good extension to
juvenile rainbow trout with a 28-d exposure to SS (Figure 81) and Chinook salmon with a 1.5-d
exposure (Figure 82). In both cases the equation is slightly over-protective, but that is
considered appropriate.
122
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 80. Lethality of suspended sediments to white perch (Morone americana),
a sensitive species, based on data compilation of (Berry et al. 2003).
¦o
«
¦2.
c
o
+3
O
cv
White Perch Mortality, 24-hr
v = 0.341 Into -1.851
0 1000
~ Observed
2000
SS (mg/L)
3000
4000
Figure 81. Lethality of suspended sediments to juvenile rainbow trout (Oncorhynchns
mykiss) using parameters for sensitive species. Data from Berry et al. (2003).
Juvenile Rainbow Trout Mortality, 19-d,
1.00
0.80
1 0.60
£
O
o
0.40
0.20
0.00
-0.20
200
400
600
800
1000
SS (mg/L)
~ Obs
Est
¦Log. (Obs)
123
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 82. Lethality of suspended sediments to Chinook salmon (Oncorhynchns
tshawytscha) using parameters for sensitive species. Data from Berry et al. (2003).
Chinook Salmon Mortality, 36-hr
1.10
0.90
I 0.70
£
c 0.50
o
S 0.30
(0
0.10
"°-10 0 10000 20000 30000 40000 50000
SS (mg/L)
~ Obs 1.5 d ¦ Est 1.5 d Trend
Highly Sensitive
This category represents those species having a 24-hr LCio < 250 mg/L SS. Small estuarine
species seem to be highly sensitive to suspended sediment (Figure 84). The general parameters
are based on a composite fit to data for bay anchovy, menhaden, and Atlantic silversides taken
from a compilation by (Berry et al. 2003). The equation is:
LethalSS = 0.328-ln(^) - 1.375 + O.l-ln(TExp) (120)
This equation is applied using one and two day exposure times along with effects from one two
and three weeks exposure. The maximum effect is chosen from these multiple calculations.
124
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 83. Three-dimensional plot of equation for highly sensitive fish
Although not verified with observed data from longer exposure periods, the equation appears to
be robust; it yields reasonable predictions of mortality for a range of SS concentrations and
exposure periods.
125
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 84. Response of bay anchovy to SS. Data from (Berry et al. 2003)
Composite of Highly Sensitive Spp., 24-hr
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
y = 0.3278Ln(x)- 1.3751
~—~
R = 0.7125
0 200 400 600 800 1000 1200
TSS
Sediment Effects on Filter Feeders
Sediments can clog filter-feeding apparatuses in invertebrates and some fish. A 25% reduction
in feeding in Daphnia occurs with SS of 6 NTU (-22 mg/L) (Henley, 2000); rotifers are not
affected (Rowe et al. 2003). Equation (116) can be parameterized to reflect the Daphnia
response (SlopeSS = -0.46 and InterceptSS = 2.2, Figure 85).
Figure 85. Reduction in feeding by Daphnia due to suspended sediments.
Points represent LC75 and supposed LC50, and LC5 values.
Reduced Feeding in Daphnia
1
0.9
c
0.8
o
"5
0.7
3
¦o
0.6
U
OH
0.5
w
w
0.4
F
0.3
«5
X
0.2
0.1
0
20 40 60 80
SS (mg/L)
100
120
126
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Increased turbidity can inhibit feeding by mussels; 600 to 750 mg SS/L reduced clearance rates
in several mussel species (Henley et al. 2000). This can be used to parameterize Equation (116)
(,SlopeSS= -0.47 andInterceptSS = 3.1, Figure 86).
Figure 86. Reduction in clearance of sediment by freshwater mussels due to
suspended sediments. Points represent supposed LC95, LC50, and LC10 values.
o
o
s
¦o
£
(O
(O
ra
X
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Reduced Clearance in Mussels
200
400
SS (mg/L)
600
800
Reduced pumping was observed at SS > 1000 mg/L in the Eastern oyster (Berry et al. 2003).
This too can be used to parameterize Equation (116). (SlopeSS = -0.61 and InterceptSS = 4.72,
Figure 87).
127
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 87. Reduced pumping in Eastern oysters (Crassostrea virginicci).
Points represent supposed LC90, LC50, and LC5 values.
1
0.9
£
0.8
o
o
0.7
3
¦o
0.6
u
on
0.5
(O
(O
0 4
F
0.3
«5
X
0.2
0.1
0
Reduced Pumping in Oysters
500 1000 1500
SS (mg/L)
2000
2500
A related factor, which is treated separately in the model, is the degree to which there is dilution
of food by inorganic particles, offset by selective sorting of particles and feeding (Henley, 2000).
Mytilus edulis, the blue mussel, and Crassostrea virginica, the Eastern oyster, actively sort
particles; their food intake should not be affected by SS until very high levels that clog the filter
feeding mechanism are reached. In contrast, there is limited selective feeding among many
clear-water clams, including the surf clam Spisula solidissima, the Iceland scallop Chlamys
istandica, and probably many of the endangered freshwater mussels (Henley, 2000). The
dilution of available food for both filter feeders and grazers decreases as a proportionate function
of sediment corrected for the degree to which there is selective feeding (Figure 88):
FoodDilution =
Food
Food + Sed ¦ Proportion ¦ (1 - Sorting)
(121)
where:
FoodDilution =
Food =
Sed
Sorting =
Proportion =
factor to account for dilution of available food by suspended
sediment (unitless)
preferred food for filter feeders (mg/L) and for grazers (g/m2) (see
(94))
suspended sediment for filter feeders (mg/L) and deposited
sediment for benthic grazers (g/m )
degree to which there is selective feeding (unitless)
proportionality constant, set to 0.01 for snails and grazers and set
to 1.0 for all other organisms, (unitless)
128
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
To account for the fact that snails and grazers feed on periphyton above the depositional surface,
a proportionality constant is utilized for those organisms.
The intermediate variable Sed depends on the computation of suspended sediment for filter
feeders and the computation of deposited sediment for benthic grazers. If the optional sediment
transport submodel (Section 6.1) is used then:
Sed = (Conc{Silt) + Conc(Clay)) or
Sed = {Deposit {Silt) + Deposit {Clay)) ¦ Vol / Surf Area ¦ 1000 -1.0
(122)
where:
Sed
Concsed
1000
Depositsed
Volume
SurfArea
1.0
suspended sediment for filter feeders (taxa = 'Susp. Feeder' or
'Clam' in units of mg/L or g/m3) and deposited sediment for
benthic grazers (taxa = 'Sed Feeder' or 'Snail' or 'Grazer' in units
of g/m2);
concentration of suspended silt or clay (mg/L) (224);
conversion factor for kg to g;
amount of sediment deposited (kg/m day) (230);
water volume, (m3);
surface area, (m ); and
days' accumulation of sediment (day)
If the sediment transport submodel is not used and TSS is used as a driving variable then
suspended sediment is computed for filter feeders. Additionally, when TSS is used as a driving
variable, deposited sediment {Sed) is calculated using the relationship shown in Figure 91.
Sed
Suspended
= InorgSed
(123)
Sed'Deposited = 0270WnorgSed60dav)- 0.072
60 day >
where:
Sed
InorgSed =
InorgSed60day =
food dilution equation input (120), (mg/L or g/m );
suspended inorganic sediment computed from TSS (mg/L) (see
(244));
60 day average of suspended inorganic sediment computed from
TSS (mg/L) (see (244))
129
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 88. The FoodDilntion factor as a function of TSS with Food kept constant at 10
mg/L and with Sorting set to 0 and 0.5.
Sediment Dilution Factor
1.00 -
— °-90"
| 0.80-
^ 0.70 -
3 0.60 -
| 0.50 -
I 0 40
§ 0.30 -
8 0.20 -
LL.
0.10 -
0.00 -
0 10 20 30 40 50 60 70
TSS
Sorting = 0 Sorting = 0.5
Continued high levels of SS can cause mortality in oysters as shown in Figure 89. However, this
can be interpreted as the natural consequence of reduced filtration as predicted by
parameterization of (115). Therefore, oyster mortality due to SS is not simulated separately.
Figure 89. Response of oysters to SS. Data from (Berry et al. 2003).
Eastern Oyster Mortality, 12-d Exposure
¦a
0)
c
o
o
<3
1.00
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
1000
2000
3000
4000
SS (mg/L)
130
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Sediment Effects on Grazers
Sediment reduces preference of New Zealand mud snails and mayflies for periphyton (Suren
2005), which is ignored by the model. More important, the food quality of periphyton declines
linearly with increasing fine sediment content (Broekhuizen et al. 2001). This is represented as
food dilution by (120).
Riffle areas are degraded or lost by deposition of fine sediment, including sand (Crowe and Hay
2004). A 12-17% increase in fines in riffles areas resulted in 27-55% decrease in mayfly
abundance; this did not affect chironomids and simulids, and riffle beetles actually increased
(Crowe and Hay 2004). Drift rates doubled from 2.3%/d to 5.2%/d with a 16% increase in fine
interstitial sediments; chironomids and caddisflies were affected (Suren and Jowett 2001). This is
represented by a function in which the deposition rate is compared to a trigger value beyond
which there is accelerated drift:
Drift = Dislodge ¦ Biomass
(124)
where:
Drift
Dislodge
loss of zoobenthos due to downstream drift (g/m -d); and
fraction of biomass subject to drift per day (unitless).
Nocturnal drift is a natural phenomenon:
Dislodge = AvgDrift ¦ AccelDrift
(125)
where:
AvgDrift
fraction of biomass subject to normal drift per day (unitless).
AccelDrift = e
,(Deposit -Trigger)
(126)
where:
AccelDrift
factor for increasing invertebrate drift due to sediment deposition
(unitless);
total rate of inorganic sediment deposition (kg/m day), (125b);
deposition rate at which drift is accelerated (kg/m2 day).
Deposit
Trigger
131
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 90. AccelDrift as a function of depth-corrected
sediment deposition with Trigger = 0.2.
Drift as a Function of Sedimentation
2.500
$ 2.000
«
+¦»
5 1.500
5 1.000
o
| 0.500
0.000
0.2 0.4 0.6 0.8
Deposit (kg/m2 d)
1.2
The model computes daily sediment deposition rate based on suspended sediment using the
following relationship:
Deposit = 2.70 • l n (SS )
(125b)
where:
Deposit
SS
total rate of inorganic sediment deposition (kg/m day), (125b);
suspended inorganic sediment concentration (mg/L). If TSS is
modeled see (244) otherwise, the sum of inorganic sediments in
the water column (i.e. Scmd+Silt+CIay)',
132
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 91. Relationship of one-day sedimentation
to average TSS; data from Larkin and Slaney (1996)..
1.40
1.20
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 92. Relationship of 60-day sedimentation to average TSS;
data from (Larkin and Slaney 1996).
500.00
400.00
T3
O
ID
m
E
3?
~ 300.00
c
a>
E
Tg 200.00
TS
a>
« 100.00
o
a.
ai
Q
0.00
y = 108.lln(x) - 28.898
R2 = 0.9712
~71.86
~ 366.68
23.07
224.04
'221.13
1 10 100
Mean Suspended Sediment (mg/L)
A similar measure of fines is embeddedness, which is the extent to which sand, silt, and clay fill
the interstitial spaces among gravel and cobbles (Osmond et al. 1995). Good spawning substrate
is characterized as less than 25% embedded (Flosi et al. 1998). The data that allow us to predict
percent fines also yield an estimate of percent embeddedness (Figure 93), and that relationship is
used in the model. Although the training data only go to 34% embeddedness, the log
relationship using averaged data allows the regression to extend to any reasonable level of
suspended sediment. The user can enter an observed "baseline embeddedness" in the site record,
and that can be used as an initial condition. A corresponding embeddedness threshold value can
be entered in the animal record. If that value is exceeded then exclusion can be assumed
(mortality = 100%). Although this functionality is intended for salmonids, it can also apply to
other fish such as sculpins and to invertebrates that hide in the interstices. In practice, the
maximum 60-day moving average of suspended sediments is used to compute the percent
embeddedness; if the initial percent embeddedness is exceeded then the new simulated percent
embeddedness is used. The possibility of scour from a high-discharge event resetting the percent
embeddedness is ignored.
134
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 93. Relationship of 60-day percent embeddedness to average TSS;
data from (Larkin and Slaney 1996).
40%
35%
* 30%
l/>
a>
-§ 25%
0>
? 20%
JU
E 15%
* 10%
~ 26%
y=0.0777ln(x)-0.0208
5%
2%
0%
0
50
100
150
200
Mean Suspended Sediment (mg/L)
Gamete Loss and Recruitment
Eggs and sperm can be a significant fraction of adult biomass; in bluegills these can be 13
percent and 5 percent, respectively (Toetz, 1967), giving an average of 9 percent if the
proportion of sexes is equal. Because only a small fraction of these gametes results in viable
young when shed at the time of spawning, the remaining fraction is lost to detritus in the model.
There are two options for determining the date or dates on which spawning will take place. A
user can specify up to three dates on which spawning will take place. Alternatively, one may use
a construct that was modified from a formulation by Kitchell et al. (1974). As a simplification,
rather than requiring species-specific spawning temperatures, it assumes that spawning occurs
when the temperature first enters the range from six tenths of the optimum temperature to 1° less
than the optimal temperature. This is based on a comparison of the optimal temperatures with the
species-specific spawning temperatures reported by Kitchell et al. (1974). Depending on the
range of temperatures, this simplifying assumption usually will result in one or two spawnings
per year in a temperate ecosystem when a simple sinusoidal temperature function is used.
However, the user also can specify a maximum number of spawnings.
The loss rate for gametes is estimated for both fish and invertebrates as a function of user-
specified intrinsic gamete mortality and increased mortality due to effects of organic toxicants,
low oxygen, and salinity on adults (FracAdults being a function of carrying capacity).
135
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION CHAPTER 4
If (0.6 • TOpt) < Temperature < (TOpt -1.0) then
GameteLoss = (GMort + IncrMort + OlEffectFrac) ¦ FracAdults ¦ PctGamete (\21)
¦ SaltMort ¦ Biomass
else GameteLoss = 0
where:
Temperature
ambient water temperature (°C);
TOpt
optimum temperature (°C);
GameteLoss
loss rate for gametes (g/m3-d);
GMort
gamete mortality (1/d);
IncrMort
increased gamete and embryo mortality due to toxicant (see (426),
1/d);
02EffectFrac
calculated fraction of gametes lost at a given oxygen concentration
and exposure time (1/d), see (205);
PctGamete
fraction of adult predator biomass that is in gametes (unitless); and
FracAdults
fraction of biomass that is adult (unitless);
SaltMort
effect of salinity on gamete loss rate (unitless), see (440); and
Biomass
biomass of predator (g/m3).
As the biomass of a population reaches its carrying capacity, reproduction is usually reduced due
to stress; this results in a population that is primarily adults. Therefore, the proportion of adults
and the fraction of biomass in gametes are assumed to be at a maximum when the biomass is at
the carrying capacity (Figure 94):
Figure 94. Correction for population-age structure
BASS
PctGamete = 0.09, GMort = 0.1
^ 0.6
0.5
S0.4
w 0.3
lu 0.2
y 0.1
KCap
BIOMASS
136
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
FracAdults = 1.0-
^ Capacity ^
v KCap / ZMean j
(128)
if Biomass > KCap / ZMean then Capacity = 0 else Capacity = KCap / ZMean - Biomass
where:
KCap = carrying capacity, the maximum sustainable biomass (g/m );
ZMean = mean depth from site underyling data (m).
Spawning in large fish results in an increase in the biomass of small fish if both small and large
size classes are of the same species. Gametes are lost from the large fish, and the small fish gain
the viable gametes through recruitment:
Recruit (I -(GMort + IncrMort)) ¦ FracAdults ¦ PctGamete ¦ Biomass (129)
where:
-3
Recruit = biomass gained from successful spawning (g/m -d).
Washout and Drift
Downstream transport is an important loss term for invertebrates. Zooplankton are subject to
transport downstream similar to phytoplankton:
Washout = DlscharSe . Biomass (130)
Volume
where:
-3
Washout = loss of zooplankton due to downstream transport (g/m -d);
"3
Discharge = discharge (m /d), see Table 3;
Volume = volume of site (m3), see (2); and
Biomass = biomass of invertebrate (g/m ).
Likewise, zoobenthos exhibit drift, which is detachment followed by washout, and it is
represented by a construct that is original with AQUATOX:
WashoutZoobenthos = Drift = Dislodge ¦ Biomass (131)
where:
"3
Drift = loss of zoobenthos due to downstream drift (g/m -d); and
Dislodge = fraction of biomass subject to drift per day (unitless), see (131) and
(132)
Nocturnal drift is a natural phenomenon:
Dislodge = AvgDrift (132)
137
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
where:
AvgDrift
fraction of biomass subject to normal drift per day (unitless).
Animals also are subject to entrainment and downstream transport in flood waters. In fact,
annual variations in fish populations in streams are due largely to variations in flow, with almost
100% loss during large floods in Shenandoah National Park (NPS, 1997). A simple exponential
loss function was developed for AQUATOX:
where:
Entrainment =
Biomass =
MaxRate =
Vel
VelMax =
Gradual =
Vel-VelMax
Entrainment = Biomass ¦ MaxRate ¦ e Gradua'
entrainment and downstream transport (g/m -d);
biomass of given animal (g/m );
maximum loss per day (1/d);
velocity of water (cm/s), (14);
velocity at which there is total loss of biomass (cm/s); and
slope of exponential, set to 25 (cm/s).
(133)
Figure 95. Entrainment of animals as a function of stream velocity
with VelMax of 400 cm/s
T3
CD
C
"as
£ 0.6
c
o
0.4
o
n
i_
LL.
0.2
0
100
200
300
400
500
Velocity (cm/s)
Entrainment is not applied to pelagic invertebrates as these organisms already passively wash out
of a system during a flood event (129).
Vertical Migration
When presented with unfavorable conditions, most animals will attempt to migrate to an adjacent
area with more favorable conditions. The current version of AQUATOX, following the example
of CLEANER (Park et al., 1980), assumes that zooplankton and fish will exhibit avoidance
behavior by migrating vertically from an anoxic hypolimnion to the epilimnion. AQUATOX
assumes that EC50gromh is the best indicator of when the species has become so intolerant of the
oxygen climate that it is going to migrate. This also allows more tolerant species to spend more
time in the hypolimnion and less tolerant species to migrate earlier. The assumption is that
anoxic conditions will persist until overturn.
138
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
The construct calculates the absolute mass of the given group of organisms in the hypolimnion,
then divides by the volume of the epilimnion to obtain the biomass being added to the
epilimnion:
where:
If VSeg = Hypo and Anoxic
(134)
Migration
HypVolume ¦ Biomass pred,
pred, hypo
VSeg
Hypo
Anoxic
Migration
HypVolume
EpiVolume
BiomaSSprcol, hypo
EpiVolume
= vertical segment;
= hypolimnion;
= boolean variable for anoxic conditions when O2 < EC50grOwth;
"3
= rate of migration (g/m -d);
= volume of hypolimnion (m3), see Figure 36;
= volume of epilimnion (m3), see Figure 36; and
= biomass of given predator in hypolimnion (g/m3).
In the estuarine model, fish will also migrate vertically based on salinity cues (see Section 10.5).
In the multi-segment version of AQUATOX, fish will vertically migrate to achieve equality on a
biomass basis if the system becomes well mixed (see Section 3.).
Migration Across Segments
To simulate seasonal migration patterns animals may be set up to move from one segment to
another during a multi-segment model run. Animals may migrate to or from a segment on any
date of the year to represent an appropriate seasonal pattern; however, reaches must be linked
together with "feedback links" for migration to be enabled. The user must specify the date on
which migration occurs, the fraction of the state variable's concentration expected to migrate,
and the segment(s) involved. The calculation of state variable movement to and from each
segment must be normalized to the volume of water in the destination segment:
MigrationFromSeg = ConeSourceSeg ¦FracMoving
(135)
Migration
ConcSourceSeg-VolumeSourceSeg ¦FracMoving
ToSeg
Volume
(136)
'Destination
where:
Migration FromSeg
Migration ToSeg
ConCsegment
Volume Segment
FracMoving
loss of state variable in source segment (mg/LsourceSeg'd);
gain of state variable in destination segment (mg/LoestinationSeg d);
concentration of state variable in given segment (mg/L);
volume of given segment (m );
user input fraction of animals migrating on given date (unitless);
139
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Anadromous Migration Model
A new option in AQUATOX is to model the migration of fish into and out of the main study area
in order to approximate anadromous migration behavior. Anadromous fish live most of their
adult life in saltwater, but they return to freshwater to spawn, and juveniles grow for a few
months to a few years before going to saltwater; during their time in freshwater they may be
exposed to and bioaccumulate organic toxicants. Chinook salmon and Pacific lamprey are two
species in the animal database that can be used with this model. The anadromous migration
component is a fairly simple model that holds off-site fish in what is assumed to be a clean
"holding tank." No additional exposure of the fish to the toxicant is predicted to occur while off-
site, but growth dilution and depuration of toxicant is assumed to occur.
To get to this model, a size-class fish must be modeled and then an "Anadromous" button will
appear in the fish loading options screen. Inputs for this model are shown below
• Day-of-year of migration (integer)
• Fraction of biomass migrating (fraction)
• Day-of-year of adult return (integer)
• Years spent off site (integer)
• Mortality fraction (fraction)
Based on these parameters and the weight of the juvenile and adult organisms, the biomass
returning to the freshwater study area may be calculated as follows.
Biomass u.
Adult
Biomass D i
Juvenile
{\-Mor
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Promotion and Emergence
Although AQUATOX is an ecosystem model, promotion to the next size class is important in
representing the emergence of aquatic insects, and therefore loss of biomass from the system,
and in predicting bioaccumulation of hydrophobic organic compounds in larger fish. The model
assumes that promotion is determined by the rate of growth. Growth is considered to be the sum
of consumption and the loss terms other than mortality and migration; a fraction of the growth
goes into promotion to the next size class (cf. Park et al., 1980):
Promotion = KPropred' (Consumption - Defecation - Respiration - Excretion - GameteLoss) (137)
where:
Promotion = rate of promotion to the next size class or insect emergence (137)
(g/m3-d);
KPro = fraction of growth that goes to promotion or emergence (0.5,
unitless);
"3
Consumption = rate of consumption (g/m -d), see (98);
"3
Defecation = rate of defecation (g/m -d), see (97);
"3
Respiration = rate of respiration (g/m -d), see (100);
"3
Excretion = rate of excretion (g/m -d), see (111); and
"3
GameteLoss = loss rate for gametes (g/m -d), see (126).
This is a simplification of a complex response that depends on the mean weight of the
individuals. However, simulation of mean weight would require modeling both biomass and
numbers of individuals (Park et al., 1979, 1980), and that is beyond the scope of this model at
present. Promotion of multi-age fish is straightforward; each age class is promoted to the next
age class on the first spawning date each year. The oldest age class merely increments biomass
from the previous age class to any remaining biomass in the class. Of course, any associated
toxicant is transferred to the next class as well. Recruitment to the youngest age class is the
fraction of gametes that are not subject to mortality at spawning. Note that the user specifies the
age at which spawning begins on the "multi-age fish" screen.
Insect emergence can be an important factor in the dynamics of an aquatic ecosystem. Often
there is synchrony in the emergence; in AQUATOX this is assumed to be cued to temperature
with additional forcing as twice the promotion that would ordinarily be computed, and is
represented by:
If Temperature > (0.8 • TOpt) and Temperature < {TOpt -1.0) then
(138)
Emergelnsect = 2 • Promotion
where:
Emergelnsect = insect emergence (mg/L-d);
Temperature = ambient water temperature (°C); and
TOpt = optimum temperature (°C);
141
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Because emergence is a function of the organism's growth rate, if the temperature passes through
the optimal temperature interval while the growth rate of the organism is zero or below zero,
emergence of insects does not occur.
Size Classes for Crabs
The capability to model fish with size classes (juvenile and adult) has been part of AQUATOX
nearly since the model's creation. However, this capability had not been extended to
invertebrates. Size classes have now been added to the AQUATOX model to better reflect the
differences in life history for some predatory invertebrates (crabs in particular). This change
allows the model to represent differential bioenergetics within life stages, and the different
vulnerability of these organisms to predation pressures and organic toxicants.
No modifications were made to the fish size-class model described here. However, the model's
interface now includes two small predatory invertebrate compartments that may be linked to the
larger predatory invertebrate categories.
4.4 Oysters
New to AQUATOX Release 3.2 is a size-class model for oysters, containing four specified life
stages for oysters (Figure 96).
Figure 96. Schematic of AQUATOX Oyster Size-Class Model
Oyster Model
Recruitment
(Spawning) Promotion (settling)
Promotion Promotion
Toxicant Burden passed with size-class promotion and recruitment
Oysters are modeled using the same differential equations that govern all animals within
AQUATOX; clams and other bivalves have always been included in the model. Oysters are
unique in AQUATOX, however, due to their life-history classifications and the promotion and
recruitment of biomass from one biotic compartment to the next.
142
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Oyster veligers are assumed to be suspended in the water column and are modeled with mg/L
units. Oyster veligers can be modeled alone in non-oyster-reef habitats. If seed oysters or sack
oysters are included then it is assumed that an oyster reef is specifically being modeled. Rules
for promotion and recruitment between size classes follow.
• Veliger promotion to Spat
o If salinity is sufficient (between 5 and 30 ppt) then it is assumed that veliger
successfully settles over its 3-week life span. Cake (1983) cites multiple sources
(Carriker 1951, Davis 1958, Calabrese & Davis 1970) who indicate veligers will
settle at salinities 5-30 ppt.)
o 5% of biomass is assumed to settle each day and is assumed "competent" spat.
This rate is estimated using a 20-day settlement period consistent with the veliger
life span cited in Galtsoff (1964) cited in Cake (1983) and Bahr and Lanier (1981)
cited in VanderKooy (2012).
• Spat promotion to Seed
o Spat are generally parameterized with high mortality rates, including predation.
VanderKooy (2012) cites several papers which collectively estimate spat to seed
mortality to range from 15-100%.
o 50% of biomass growth is assigned to the seed category. This is the same
promotion algorithm that has historically been used for fish size-class promotion
in AQUATOX (136).
o This promotion assumption is a simplification of a complex response that depends
on the mean weight of the individuals. However, simulation of mean weight
would require modeling both biomass and numbers of individuals (Park et al.,
1979, 1980) which is presently beyond the scope of this model.
• Seed promotion to Sack oyster
o Like spat promotion, 50% of biomass growth is assigned to the sack category.
• Sack spawning to Veliger (recruitment)
o As a simplifying assumption, Sack are the exclusive contributors to overall
spawned biomass. Menzel (1951) is cited in VanderKooy (2012), indicating that
oysters in the Gulf can become sexually mature within 4 weeks of settlement (i.e.
Seed can also contribute to recruitment). However, VanderKooy (2012) states
that "the number of gametes released during each spawn is directly correlated
with oyster size and gonadal development (Davis and Chanley 1955, Galtsoff
1964, Thompson et al. 1996) For this reason, it is assumed that Sack are the
dominant contributors.
o Spawning is triggered when salinity is greater than 10 ppt and temperature is
greater than 20 degrees C. VanderKooy (2012) claim that "salinity fluctuations do
not appear to play a significant role in controlling spawning in oysters, however,
salinities below 5-6 ppt can inhibit gametogenesis (Butler 1949, Loosanoff
143
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
1953)." However, there is support for the 10 ppt threshold in Cake (1983), who
claims "optimal spawning salinity ranges from 10 to 30 parts per thousand (ppt)."
o With regards to temperature, Cake (1983) cites multiple sources (Ingle 1951,
Menzel 1955, Hayes and Menzel 1981) who indicate mass spawnings typically
occur when temperatures reach and/or exceed 25 degrees C. However, there is
ample support for the 20 degree C threshold, as VanderKooy (2012) cites multiple
authors (Butler 1949, Loosanoff 1953, Schlesselman 1955, Hofstetter 1977, 1983)
who indicate the threshold is indeed 20 degrees C.
o The "percent gamete" parameter determines the extent of biomass expelled. This
is spread over 275 days of spawning, based on Gulf-of-Mexico data. Cake (1983)
cites (Butler 1954), who claims that spawning generally occurs year-round in the
Gulf of Mexico, but does not occur during the three-month span from December
to February. A site-specific number of spawning days may be specified for other
locations.
o The Gamete loss equation (126) splits spawning between viable veliger biomass
and non-viable organic matter (detritus).
4.5 Aquatic-Dependent Vertebrates
Herring gulls and other shorebirds were added to AQUATOX Release 3 as a bioaccumulative
endpoint—not as a dynamic variable but as a post-processed variable reflecting dietary exposure
to a contaminant. In fact, the endpoint can be used to simulate bioaccumulation for any aquatic
feeding organism, such as bald eagles, mink, and dolphins, provided that the organism feeds
exclusively on biotic compartments modeled within AQUATOX. The user can specify a
biomagnification factor (BMF) and the preferences for various food sources so that alternate
exposures can be computed. Dietary preferences are input as fraction of total food consumed by
the modeled species and are normalized to 100% when the model is run.
The concentration of each chemical is based on the chemical concentration in prey at a given
time-step.
PPBBird,^ = Z{Prefr,v ¦ BMFrm ¦ PPBPrvIJ (139)
2 = 1
where:
PPBBirdfoxicant = estimated concentration of this toxicant in bird or other
(i^g/kg);
BMFtox = biomagnification factor for this chemical in bird or other
(unitless);
PPBpreyjox = concentration of this chemical in prey ([j,g/kg), see (310).
Uptake of toxicant is assumed to be instantaneous, but depuration of the chemical is governed by
the user-input clearance rate. If the concentration of chemical is declining in shorebirds (due to
organism
organism
144
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
the concentrations of the chemical declining in prey), the lowest the chemical concentration in
birds can fall to at any time is calculated as follows:
PPBBirdLowestTox = PPBBirdToxt_x (1 - C/earTox)A7' (140)
where:
PPBBirdLoWest Tox = lowest conc. of this toxicant in gulls or other organism at this time-
step ((J-g/kg);
PPBBirdrox,t-i = concentration of this toxicant in in the previous time-step ([j,g/kg);
Clear fox = clearance rate for the given toxicant, (1/day)
4.6 Steinhaus Similarity Index
Within the differences graph portion of the output interface, a user may select to write a set of
Steinhaus similarity indices in Microsoft Excel format. The Steinhaus index (Legendre and
Legendre 1998) measures the concordance in values (usually numbers of individuals, but
biomass in this application) between two samples for each species. Typically it is computed from
monitoring data from perturbed and unperturbed, or reference, sites. When calculated by
AQUATOX it is a measure of the difference between the control and perturbed simulations. A
Steinhaus index of 1.0 indicates that all species have identical biomass in both simulations (i.e.,
the perturbed and control simulations); an index of 0.0 indicates a complete dissimilarity
between the two simulations.
The equation for the Steinhaus index is as follows:
2 • £ mm(Biomass t control , Biomassperturbed )
£ = —T
YJ(V/omassi conlrol + Biomass , perturbed)
=i (141)
where:
S = Steinhaus similarity index at time t;
Biomass i_COntroi = biomass of species i, control scenario at time t;
Biomass i_perturbed = biomass of species i, perturbed scenario at time t.
A time-series of indices is written for each day of the simulation representing the similarity on
that date. Separate indices are written out for plants, all animals, invertebrates only, and fish
only.
4.7 Biological Metrics
Ecological indicators are defined as primarily biological and are measurable characteristics of the
structure, composition, and function of ecological systems (Niemi and McDonald 2004). The
term "indicator" as used by Niemi and McDonald is a rather broad one, and includes two terms
145
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
often used within the biocriteria program, "metric" and "index". A biological metric is a
numerical value that represents a quantitative community parameter, such as species diversity,
or percent EPT (see below). A multimetric index is a number that integrates several metrics to
express a site's condition or health, such as an IBI (Index of Biological Integrity). AQUATOX
has the ability to calculate numerous metrics, some of which can be compared to similar metrics
derived from monitoring data. However, there are limitations in the application of many such
metrics that reflect the differing capabilities of simulation models as opposed to field studies.
Models can predict continuing complex responses to changing conditions, while field
measurements usually represent snapshots of existing conditions with limited empirical
predictive power. Aquatic models have limited taxonomic resolution and usually represent
biomass; most metrics and indices applied in the field are based on detailed taxonomic
identifications and involve counting the numbers of individual organisms per sample. Therefore,
only a subset of possible indicators can be implemented with AQUATOX; however, given the
biologic realism of the model, the list is much more extensive than for other models.
Biotic metrics and indices have been widely used for several decades, stimulated in part by
inclusion in rapid bioassessment protocols (RBP) by the US EPA (Plafkin et al. 1989). Most are
applicable to streams and wadeable rivers (Barbour et al. 1999), though there is a suite of indices
(the trophic state indices) that were developed as a measure of eutrophication in lakes. Metrics
can be calculated for algae, which indicate short-term impacts; macroinvertebrates, which
integrate short-term impacts on localized areas; and fish, which are indicators of long-term
impacts over broad reaches (Barbour et al. 1999).
Ecological indicator measures fall into several well defined categories. Those metrics that are
presently calculated in AQUATOX are shown below in boldface; the others enumerated here can
be calculated offline using exported Excel output files:
• Composition—many metrics related to community composition are suitable for
simulation with AQUATOX by selecting the appropriate "Benthic metric designation"
category on the underlying data screen; they include:
o % EPT (the following three combined) (Barbour et al. 1999)
¦ % Ephemeroptera (mayfly larvae) (Maloney and Feminella 2006)
¦ % Plecoptera (stonefly larvae) (Barbour et al. 1999)
¦ %Trichoptera (caddisfly larvae) (Barbour et al. 1999)
o % chironomids (midge larvae) (Barbour et al. 1999)
o % oligochaetes (aquatic worms) (Barbour et al. 1999)
o % Corbicula (invasive Asian clam) (Barbour et al. 1999)
o % Eunotia (interstitial diatom characteristic of low-nutrient conditions) (Lowe et
al. 2006)
o % cyanobacteria (cyanobacteria characteristic of high-nutrient, turbid
conditions) (Trimbee and Prepas 1987).
• Trophic—these include metrics that can be calculated from AQUATOX output:
o Trophic Level (Odum, 1971)
o Periphytic chlorophyll a (Barbour et al. 1999)
o Sestonic chlorophyll a (Barbour et al. 1999)
146
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
o % predators (can apply to both macroinvertebrates and fish) (Barbour et al. 1999)
o % omnivores (best applied to fish in AQUATOX) (Barbour et al. 1999)
o % forage or insectivorous fish (Barbour et al. 1999)
o Pelagic Invt. Biomass (mg/L), Benthic Invt. Biomass (g/m2), Fish Biomass
(g/m ) sum of biomass within these three animal categories.
• Trophic state—surrogates for lake and reservoir algal biomass adjusted to a common
scale (Gibson et al. 2000):
o TSI(TN) (total nitrogen)
o TSI(SD) (Secchi depth)
o TSI(CHL) (chlorophyll a)
o TSI(TP) (total phosphorus)
• Ecosystem bioenergetic—whole ecosystem metrics:
o Gross primary productivity, GPP (g 02/m2 d) (Odum 1971), more meaningful
if expressed as an annual measure (g 02/m2 yr) (Wetzel 2001)
o Net primary productivity, NPP (g 02/m d) GPP minus dark respiration
o Community respiration, R (g 02/m2 d) (Odum 1971), more meaningful if
expressed as an annual measure (g 02/m yr) (Wetzel 2001)
o P/R (ratio of GPP to community respiration) (Odum 1971)
o Turnover time (P/B, ratio of GPP to biomass in days) (Odum 1971)
In addition to those listed above, there are several ecological indicators that are not suitable for
simulation modeling in general or for AQUATOX in particular:
• Richness—these are based on numbers of observed taxonomic groups and are not
suitable for simulation modeling;
• Tolerance/intolerance—based on number of tolerant or intolerant species and therefore
unsuitable for modeling;
• Life cycle—percent of organisms with short or long life cycles, not easily modeled with
AQUATOX.
The trophic state indices are applicable to lakes and reservoirs. They are lognormal-transformed
values that attempt to convert environmental variables to a common value representing algal
biomass (Gibson et al. 2000):
Secchi Depth (m): TSI (SD) = 60 - 14.41 ln(SD)
Chlorophyll a (|ig/L): TSI (CHL) = 9.81 ln(CHL) + 30.6
Total Phosphorus (mg/L): TSI (TP) = 14.42 ln(TP) + 4.15
Total Nitrogen (mg/L): TSI(TN) = 54.45 + 14.43 ln(TN)
The user can specify over what time period the indices are averaged. This enables better
comparison with field-derived TSIs, which are generally calculated from samples taken during
the growing season. Obviously, chlorophyll a is the best representation of algal biomass, and that
metric should generally be used in determining the trophic state of a lake or reservoir (Table 8).
However, comparing the TSIs is also informative (Table 9).
147
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
The bioenergetic metrics are widely used by ecologists and have practical value as indicators of
accumulating organic matter (Odum 1971) and response to watershed disturbance (Dale and
Maloney 2004).
Table 8. Changes in Temperate Lake Attributes According to Trophic State
(Gibson et al. 2000, adapted from Carlson and Simpson 1996).
TSI
Value
11
TP
nw
Attribute*
Water Supply
RacreaHon
Fisheries
<30
>8
<6
digotrophy: Clear
water, oxygen
throughout fw year in
the hypoBmriiort
Sainwiid
fisheries
dominate
30-40
8-4
6-12
Hypolimnia of shallower
lakes may become
anoxic
Saimonitl
fisheries in
deep lakes
40-50
4-2
12-24
Mesolropftf: Water
moderately clear but
increasing probability of
hypoRmnetic anoxia
doling summer
Iron and manganese
evident during the
summer. THM
precursors exceed
0.1 mgrt. and turbkfity
>1 NTU
Hypolimnetic
anoxia
results in
loss of
salmonjds.
Walleye my
pratominate
50-80
2-1
24-48
Eubophy: Anoxic
hypoRmnia, macrophyte
Iron, rnar^anese,
taste, and odor
Warm-water
fisheries
only, Bass
may be
dominant
00-7:
0.5-1
48-96
Blue-green alp»
dominate, algal scums
and maciophyte
pfobtems
Weeds, alia!
scums, and
tow
and boJna
70-80
0.25-
0.5
§§-102
Hypereubophy (light
Imited). Dense algae
and maciopSiftas
>80
<0.25
1tt2~
384
Algal scums, few
macrophytes
Rough fish
dominate,
summer fish
Mis possible
148
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Table 9. Conditions Associated with Various Trophic State Index Variable Relationships
(Gibson et al. 2000).
Relationship Between TS1 Variables
Conditions
TSI «'CKi =TS«CK.) = TSKSDj
Algae dominate light attenuation
TSIiCHL: > ~SliSD)
Large particulates, such as Aphanizomenon Hakes,
dominate
TSIiTP} = TSI{SD) > TSI(CHL)
Nonalgal parttculates or color dominate light attenuation
TSIiSDj = TSI(CHL) >TSI|TP)
Phosphorus limits, algal biomass (TWTP ratio greater than
33:1)
TSI (TP) > "SKCHL) = TSI(SD)
Zooplamfcton grazing, nitrogen, or some factor other than
phosphorus limits algal biomass
Trophic Level
Trophic Level output has been added for each animal to clarify how its feeding preferences have
been translated into feeding practices within the dynamic simulation (Odum, 1971). The
unitless trophic level is calculated within AQUATOX as follows. First, algae and organic matter
are assigned trophic levels of 1.0. For animals, the idealized trophic level is assigned based on
the position of their prey in the food chain; organisms that are exclusively herbivores or
detritivores are assigned to a trophic level of 2.0; organisms that are exclusively predators will be
assigned to a trophic level of 3.0 or higher. The trophic level can then be calculated for each
time step as a function of the trophic level of its prey being consumed at that time step (based on
actual prey availability):
TLpredator
1](tlp,v +
1.0 ^Frac
Prey
(140b)
where:
TLPredator = calculated trophic level of the predator (unitless);
TLprey = calculated trophic level of one prey item (unitless);
Fracprey = fraction of the prey item consumed compared to all food being consumed in
that time step by the predator (unitless).
Fractional trophic-level outputs are likely based on the complexity of the foodweb. If a
heterotroph is not feeding in a given time step it is assigned to a trophic level of 2.0.
It can be useful to calculate an average trophic level over a year, weighting by the organism's
growth rate at each time step. This method provides the best estimate of the trophic level that
represents the organism's biomass; it ensures that dormant periods of non-feeding or low-feeding
behavior do not bias the trophic level derived.
149
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Invertebrate Biotic Indices
As noted above, some invertebrate biotic indices can be readily computed by AQUATOX with
one caveat: they are based on relative biomass rather than numerical density (number of
individuals representing a taxonomic group). The simplifying assumption is that weights of
individuals are roughly comparable. Of course, this is not actually the case, but individual
weights vary greatly depending on the growth stage, so use of biomass has less error.
Computation of benthic invertebrate indices by AQUATOX requires that the taxonomic
affiliations be designated as:
• oligochaete (worm)
• chironomid (midge) and other fly larvae
• mayfly
• stonefly
• caddisfly
• beetle
• mussel
• other bivalve
• amphipod
• gastropod
• other.
With these designations AQUATOX can compute % EPT (Ephemeroptera or mayflies,
Plecoptera or stoneflies, and Trichoptera or caddisflies) as a percentage of the total biomass of
benthic invertebrates. Mussels are excluded from the computation in AQUATOX because the
potential biomass of a single individual may exceed that of all other invertebrates. The EPT are
usually the most sensitive aquatic insect orders, so this index is often useful. A detailed study of
Fort Benning, Georgia streams showed that %Ephemeroptera, %Plecoptera, and %Trichoptera
were significantly inversely correlated with the degree of disturbance in the watershed (Maloney
and Feminella 2006, Mulholland et al. 2007). The index has been used in evaluating remediation
(Purcell et al. 2002). The user is cautioned to ensure that the benthic metric chosen is
appropriate for the region; e.g. one wouldn't expect Plecoptera to be prevalent in Florida, due to
their temperature preferences.
Chironomids (midge larvae) are generally tolerant (Maloney and Feminella 2006), so the %
Chironomids index is useful as an indicator of disturbance (Figure 97). Another index that might
indicate disturbance is the computed value of % Oligochaetes; however, that metric has exhibited
mixed results in Georgia (Maloney and Feminella 2006).
150
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 4
Figure 97. Example of % Chironomid index computed for Upatoi Creek, Fort Benning, Georgia;
observed values are courtesy of George Williams
Upatoi Creek Ft Benning GA(Control)
• Obs %Chiro (percent)
100
100
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
12/6/1999
12/5/2001
12/5/2003
151
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
5. REMINERALIZATION
5.1 Detritus
For the purposes of AQUATOX, the term "detritus" is used to
include all non-living organic material and associated
decomposers (bacteria and fungi). As such, it includes both
particulate and dissolved material in the sense of Wetzel (1975),
but it also includes the microflora and is analogous to
"biodetritus" of Odum and de la Cruz (1963) . Detritus is
modeled as eight compartments: refractory (resistant) dissolved,
suspended, sedimented, and buried detritus; and labile (readily
decomposed) dissolved, suspended, sedimented, and buried
detritus (Figure 98). This degree of disaggregation is considered
necessary to provide more realistic simulations of the detrital
food web; the bioavailability of toxicants, with orders-of-
magnitude differences in partitioning; and biochemical oxygen demand, which depends largely
on the decomposition rates. Buried detritus is considered to be taken out of active participation
in the functioning of the ecosystem. In general, dissolved organic material is about ten times that
of suspended particulate matter in lakes and streams (Saunders, 1980), and refractory compounds
usually predominate; however, the proportions are modeled dynamically.
Detritus: Simplifying
Assumptions
• Refractory detritus does not
decompose directly but is
converted to labile detritus
through colonization
• Detrital sedimentation is
modeled with simplifying
assumptions (unless the
sediment submodel for streams
is included)
• Biomass of bacteria is not
explicitly modeled
Figure 98. Detritus compartments in AQUATOX
detr.
fm.
detr.
fm.
-o
Refractory
Dissolved
Labile
^fm.
Dissolved
decomp.
-£>
Refractory
Suspended
colonization
colonization
ingestion
detr. _
Refractory
Sediments
sedimentation
colonization
burial
exposure
ingestion
scour
Labile
Suspended
_ detr.
^^NrrT
ingestion
decomp
-r>*
sedimentation
V
o+
Labile
Sediments
n
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
The concentrations of detritus in these eight compartments are the result of several competing
processes:
CiStl ST)J^G u Da tv
—— = Loading + DetrFm - Colonization - Washout + Washin
dt (142)
- Sedimentation - Ingestion + Scour ± Sinking ± TurbDiff ± DiffusionSeg
dSuspLabDetr , _ 7 .
= Loading + Detrrm + Colonization - Decomposition
dt
- Washout + Washin- Sedimentation - Ingestion + Scour (143)
± Sinking ± TurbDiff + DiffusionSeg
dDissRe frDc tv
= Loading + DetrFm - Colonization - Washout + Washin
dt (144)
± TurbDiff ± DiffusionSeg
dDissLabDetr , . . ,jr , ,jr , ,
= Loading + Detrrm - Decomposition - Washout + Washin
dt (145)
± TurbDiff ± DiffusionSeg
dSedRefrDetr = /ot/6///7p- + DetrFm + Sedimentation + Exposure
dt 8 F (146)
- Colonization - Ingestion - Scour - Burial
dSedLabileDetr , _ , .
= Loading + Detrrm + Sedimentation + Colonization
dt 5 (147)
- Ingestion - Decomposition - Scour + Exposure - Burial
dBuriedRefrDetr_ = Sedimentation + Burial - Scour - Exposure (148)
dt
dBuriedLabileDetr
dt
Sedimentation + Burial - Scour - Exposure
where:
(149)
dSuspRefrDetr/dt = change in concentration of suspended refractory detritus
"3
with respect to time (g/m -d);
dSuspLabileDetr/dt = change in concentration of suspended labile detritus with
respect to time (g/m -d);
dDissRefrDetr/dt = change in concentration of dissolved refractory detritus
with respect to time (g/m -d);
153
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
dDissLabDetr/dt
dSedRefrDetr/dt
dSedLabileDetr/dt
dBuriedRefrDetr/dt
dBuriedLabileDetr/dt
Loading
DetrFm
Colonization
Decomposition
Sedimentation
Scour
Exposure
Burial
Washout
Washin
Diffusionseg
Ingestion
Sinking
TurbDiff
change in concentration of dissolved labile detritus with
"3
respect to time (g/m -d);
change in concentration of sedimented refractory detritus
with respect to time (g/m -d);
change in concentration of sedimented labile detritus with
"3
respect to time (g/m -d);
change in concentration of buried refractory detritus with
"3
respect to time (g/m -d);
change in concentration of buried labile detritus with
"3
respect to time (g/m -d);
loading of given detritus from nonpoint and point sources,
"3
or from upstream (g/m -d);
-3
detrital formation (g/m -d);
-3
colonization of refractory detritus by decomposers (g/m -d),
see (155);
-3
loss due to microbial decomposition (g/m -d), see (159);
transfer from suspended detritus to sedimented detritus by
"3
sinking (g/m -d); in streams with the inorganic sediment
model attached see (235), for all other systems see (165);
-3
resuspension from sedimented detritus (g/m -d); in streams
with the inorganic sediment model attached see (233), for
all other systems see (165) (resuspension);
transfer from buried to sedimented by scour of overlying
"3
sediments (g/m -d);
transfer from sedimented to deeply buried (g/m -d), see
(167b);
"3
loss due to being carried downstream (g/m -d), see (16);
loadings from upstream segments (g/m d), see (30);
gain or loss due to diffusive transport over the feedback
-3
link between two segments, (g/m -d), see (32);
loss due to ingestion by detritivores and filter feeders
(g/m3-d), see (91);
detrital sinking from epilimnion and to hypolimnion under
stratified conditions, see (165); and
transfer between epilimnion and hypolimnion due to
turbulent diffusion (g/m3-d), see (22) and (23).
As a simplification, refractory detritus is considered not to decompose directly, but rather to be
converted to labile detritus through microbial colonization. Labile detritus is then available for
both decomposition and ingestion by detritivores (organisms that feed on detritus). Because
detritivores digest microbes and defecate the remaining organic material, detritus has to be
conditioned through microbial colonization before it is suitable food. Therefore, the assimilation
efficiency of detritivores for refractory material is usually set to 0.0, and the assimilation
efficiency for labile material is increased accordingly.
154
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Sedimentation and scour (resuspension) are opposite processes. In shallow systems there may be
no long-term sedimentation (Wetzel et al., 1972), while in deep systems there may be little
resuspension. In the classic AQUATOX model, sedimentation is a function of flow, ice cover
and, in very shallow water, wind based on simplifying assumptions. Scour and exposure of
organic matter are applicable only in streams where they are keyed to the behavior of clay and
silt. Scour as an explicit function of wave and current action is not implemented, however, the
capability to link to hydrodynamic models is provided. See chapter 6 for a discussion of the
various inorganic sediment models and their implications to organic sediments.
Within AQUATOX, the user must specify the percentage particulate and percentage refractory
for each source of organic matter. Table 10 presents some guidance on populating these variables
based on Allan (1995), Hessen and Tranvik (1998), and Wetzel (2001). These percentages can
be specified as constant variables or by using a time-series.
Table 10. Suggested detrital boundary conditions based on literature and in the absence of data
Particulate
Refractory
OM conc.
Ecosystem
%
%
(mg/L)
Oligotrophic lakes
10%
90%
4
Eutrophic lakes
15%
86%
24
Forested streams
20%
60%
5
Rivers
30%
60%
14
Blackwater stream
5%
95%
26
AQUATOX simulates detritus as organic matter (dry weight); however, the user can input data
as organic carbon or carbonaceous biochemical oxygen demand (CBOD) and the model will
make the necessary conversions. Organic matter is assumed to be 1.90 • organic carbon as
derived from stoichiometry (Winberg 1971). The conversion from BOD includes the
simplifying assumption that any BOD data input into the model are primarily based on
carbonaceous oxygen demand:
OM = CBOD
f CBODSCBOl),
02Biomass
where:
(148b)
CBOD5 CBODt,
1
100% - PercenlRefh
(148c)
Time J
and:
OM
CBOD
organic matter input as required by AQUATOX (g OM/m );
carbonaceous biochemical demand 5-day from user input (g O2
/m3);
155
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
CBOD5 CBODu = CBOD5 to ultimate carbonaceous BOD conversion factor, also
defined as CBODu:CBOD5 ratio;
PercentRefr Time = user-defined percent refractory matter for given source of organic
matter, may be a time series; and
02Biomass = ratio O2 to organic matter (OM). (remineralization parameter, the
default is 0.575 based on Winberg (1971));
AQUATOX has always assumed that user-input BOD5 loadings are primarily composed of
carbonaceous oxygen demand but this assumption has been made more explicit in Release 3.1.
The equations above are used by AQUATOX when converting initial conditions and loadings in
CBOD5 to organic matter, when estimating CBOD5 from organic matter for simulation output,
and when linking HSPF BOD data. Equations (148b) and (148c) are new to AQUATOX
Release 3.1 and beyond; a warning message is displayed if an older study that utilizes BOD
loadings is imported into the current version.
Detrital Formation
Detritus is formed in several ways: through mortality, gamete loss, sinking of phytoplankton,
excretion and defecation:
DetrFfnsuspRefrDetr = SHota (Mort 2detr, Hota • Mortality Uota) (150)
DetrFmDlsSRefrDetr = Zb,ota(Mort 2detrMota ¦ Mortalitybiota) + ^blota(Excr 2detrMota ¦ Excretion) (151)
DetrFmDissLabUeDetr = I^brota (Mort 2detrMota • Mortality bwta) + Zblota (Excr 2detrMota ¦ Excretion) (152)
DetrFmSuspLdbOeDetr = Ebiota (M°n 2detr, biota ' M°rtalUy blota) + Mammals GameteLoSS (153)
F)etrFlTlsedLabileDetr X pred 2detr, pred ' ^ ^efeCLillO}} ) + ^compartment(Sinkingcgm^artment) (154)
DetrFmSedRefrDetr = £pred(Def 2detr, pred ' Defecationpmd)
\ /
Xcompartment (Sedimentation compartment ¦ PlantSinkToDetr)
where:
DetrFm
Mort2fetr, biota
Excr2 (fetr, biota
Mortality hlota
Excretion
GameteLoss
formation of detritus (g/m -d);
fraction of given dead organism that goes to given detritus
(unitless);
fraction of excretion that goes to given detritus (unitless), see
Table 11;
death rate for organism (g/m -d), see (66), (87) and (112);
excretion rate for organism (g/m -d), see (64) and (111) for plants
and animals, respectively;
loss rate for gametes (g/m -d), see (126);
156
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Def2 detr, biota
Defecationpred =
Sedimentation =
PlantSinkToDetr =
fraction of defecation that goes to given detritus (unitless);
"3
defecation rate for organism (g/m -d), see (97);
"3
loss of phytoplankton to bottom sediments (g/m -d), see (69); and
labile and refractory portions of phytoplankton (unitless, 0.92 and
0.08 respectively).
A fraction of mortality, including sloughing of leaves from macrophytes, is assumed to go to
refractory detritus; a much larger fraction goes to labile detritus. Excreted material goes to both
refractory and labile detritus, while gametes are considered to be labile. Half the defecated
material is assumed to be labile because of the conditioning due to ingestion and subsequent
inoculation with bacteria in the gut (LeCren and Lowe-McConnell, 1980); fecal pellets sink
rapidly (Smayda, 1971), so defecation is treated as if it were directly to sediments.
Phytoplankton that sink to the bottom (that are not linked to periphyton compartments) are
considered to become detritus; most are consumed quickly by zoobenthos (LeCren and Lowe-
McConnell, 1980) and are not available to be resuspended.
Table 11. Mortality and Excretion to Detritus
Algal
Mortality
Macrophyte
Mortality
Bryophyte
Mortality
Animal
Mortality
Dissolved Labile Detritus
0.27
0.24
0.00
0.27
Dissolved Refractory Detritus
0.03
0.01
0.25
0.03
Suspended Labile Detritus
0.65
0.38
0.00
0.56
Suspended Refractory Detritus
0.05
0.37
0.75
0.14
Algal
Excretion
Macrophyte
Excretion
Bryophyte
Excretion
Animal
Excretion
Dissolved Labile Detritus
0.9
0.8
0.8
1.0
Dissolved Refractory Detritus
0.1
0.2
0.2
0.0
Colonization
Refractory detritus is converted to labile detritus through microbial colonization. When bacteria
and fungi colonize dissolved refractory organic matter, they are in effect turning it into
particulate matter. Detritus is usually refractory because it has a deficiency of nitrogen
compared to microbial biomass. In order for microbes to colonize refractory detritus, they have
to take up additional nitrogen from the water (Saunders et al., 1980). Thus, colonization is
nitrogen-limited, as well as being limited by suboptimal temperature, pH, and dissolved oxygen:
where:
Colonization = ColonizeMax ¦ DecTCorr ¦ NLimit ¦ pHCorr
¦ DOCorrection ¦ RefrDetr
"3
Colonization = rate of conversion of refractory to labile detritus (g/m -d);
(156)
157
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
ColonizeMctx
Nlimit
DecTCorr
pHCorr
DOCorrection
RefrDetr
maximum colonization rate under ideal conditions (g/g-d);
limitation due to suboptimal nitrogen levels (unitless), see (157);
the effect of temperature (unitless), see (156);
limitation due to suboptimal pH level (unitless), see (162);
limitation due to suboptimal oxygen level (unitless), see (160);
and
concentration of refractory detritus in suspension, sedimented, or
dissolved (g/m ).
Because microbial colonization and decomposition involves microflora with a wide range of
temperature tolerances, the effect of temperature is modeled in the traditional way (Thomann and
Mueller, 1987), taking the rate at an observed temperature and correcting it for the ambient
temperature up to a user-defined, high maximum temperature, at which point it drops to 0:
DecTCorr = Thetaemp TObs where
Theta = 1.047 if Temp >19° else
Theta =1.185- 0.00729 ¦ Temp
(157)
If Temp > TMax Then DecTCorr = 0
The resulting curve has a shoulder similar to the Stroganov curve, but the effect increases up to
the maximum rate (Figure 99).
Figure 99. Colonization and decomposition as an effect of temperature
TEMPERATURE (C)
158
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
The nitrogen limitation construct, which is original with AQUATOX, is parameterized using an
analysis of data presented by Egglishaw (1972) for Scottish streams. It is computed by:
NLimit
N-MinN
N -MinN + HalfSatN
(158)
where:
N
MinN
HalfSatN
Ammonia
Nitrate
N = Ammonia + Nitrate
total available nitrogen (g/m );
(159)
minimum level of nitrogen for colonization (= 0.1 g/m );
half-saturation constant for nitrogen stimulation (=0.15 g/m );
concentration of ammonia (g/m3); and
-3
concentration of nitrite and nitrate (g/m ).
Although it can be changed by the user, a default maximum colonization rate of 0.007 (g/g-d) is
provided, based on Mclntire and Colby (1978, after Sedell et al., 1975). The rates of
decomposition (or colonization) of refractory dissolved organic matter are comparable to those
for particulate matter. Saunders (1980) reported values of 0.007 (g/g-d) for a eutrophic lake and
0.008 (g/g-d) for a tundra pond. Anaerobic rates were reported by Gunnison et al. (1985).
Decomposition
Decomposition is the process by which detritus is broken down by bacteria and fungi, yielding
constituent nutrients, including nitrogen, phosphorus, and inorganic carbon. Therefore, it is a
critical process in modeling nutrient recycling. In AQUATOX, following a concept first
advanced by Park et al. (1974), the process is modeled as a first-order equation with
multiplicative limitations for suboptimal environmental conditions (see section 4.1 for a
discussion of similar construct for photosynthesis):
Decomposition = DecayMax ¦ DOCorrection ¦ DecTCorr ¦ pHCorr ¦ Detritus (160)
where:
Decomposition
DecayMax
DOCorrection
DecTCorr
pHCorr
Detritus
loss due to microbial decomposition (g/m -d);
maximum decomposition rate under aerobic conditions (g/g-d);
correction for anaerobic conditions (unitless), see (160);
the effect of temperature (unitless), see (156);
correction for suboptimal pH (unitless), see (162); and
concentration of detritus, including dissolved but not buried (g/m3).
Note that biomass of bacteria is not explicitly modeled in AQUATOX. In some models (for
example, EXAMS, Burns et al., 1982) decomposition is represented by a second-order equation
using an empirical estimate of bacteria biomass. However, using bacterial biomass as a site
159
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
constant would constrain the model, potentially forcing the rate. Decomposers were modeled
explicitly as a part of the CLEAN model (Clesceri et al., 1977). However, if conditions are
favorable, decomposers can double in 20 minutes; this can result in stiff equations, adding
significantly to the computational time. Ordinarily, decomposers will grow rapidly as long as
conditions are favorable. The only time the biomass of decomposers might need to be
considered explicitly is when a new organic chemical is introduced and the microbial assemblage
requires time to become adapted to using it as a substrate.
The effect of temperature on biodegradation is represented by Equation (156), which also is used
for colonization. The function for dissolved oxygen, formulated for AQUATOX, is:
^ ^ KAnaerobic
DOLorrection = r actor + (1 - r actor) (161)
DecayMax
where the predicted DO concentrations are entered into a Michaelis-Menten formulation to
determine the extent to which degradation rates are affected by ambient DO concentrations
(Clesceri, 1980; Park et al., 1982):
Faclor = Oxygen (|62)
HalfSatO + Oxygen
and:
Factor = Michaelis-Menten factor (unitless);
3 3
KAnaerobic = decomposition rate at 0 g/m oxygen (g/m -d or (j,g/L-d); Set to
"3
0.3 g/m -d for microbial degradation of sediments. For chemicals,
(160) is also used and the "rate of anaerobic microbial degr." from
the chemical underlying data is used (KMDegrAnaerobic).
Oxygen = dissolved oxygen concentration (g/m ); and
HalfSatO = half-saturation constant for oxygen (g/m3) (0.5 g/m3 in the water
"3
column or 8.0 g/m for sedimented detritus).
DOCorrection accounts for both decreased and increased (Figure 100) degradation rates under
anaerobic conditions, with KAnaerobic/DecayMax having values less than one and greater than
one, respectively. Detritus will always decompose more slowly under anaerobic conditions; but
some organic chemicals, such as some halogenated compounds (Hill and McCarty, 1967), will
.3
degrade more rapidly. Half-saturation constants of 0.1 to 1.4 g/m have been reported (Bowie et
al., 1985); a value of 0.5 g/m3 is used in the water column and a calibrated value of 8.0 g/m3 is
used for the sediments to force anoxic conditions.
160
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Figure 100. Correction for dissolved oxygen
V)
V)
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Figure 101. Limitation due to pH
EFFECT OF pH
0
3.0 4.1 5.2 6.3 7.4 8.5 9.6 10.7
2
Sediment oxygen demand (SOD in g 02/m d) is also calculated by taking the sum of detrital
decomposition and then multiplying by 02Biomass (the ratio of oxygen to organic matter). This
can be compared with SOD values derived from the optional sediment diagenesis model
(Chapter 7).
Sedimentation
Depending upon which options the user chooses, sedimentation (i.e., the sinking of suspended
particles to the sediment bed) is calculated differently (Table 12). When the inorganic-sediment
model (sand-silt-clay) is included, the sedimentation and deposition of detritus is assumed to
mimic the sedimentation and resuspension of silt (see (235) and (233)). If the multi-layer
sediment model is included (using user-input erosion and deposition time-series) the
sedimentation of detritus is calculated using the deposition velocity for cohesives (assumed to be
a surrogate for organic matter) as follows:
When the inorganic-sediment model or the multiple-layer sediment model are not included in a
simulation (i.e. "classic" AQUATOX formulations are used), the sedimentation of suspended
particulate detritus to bottom sediments can be modeled using simplifying assumptions (165).
The constructs are intended to provide general responses to environmental factors, but they could
be considerably improved upon by linkage to a hydrodynamic model (currently only available
with the multi-layer sediment model).
Sedimentation
DepVel
— State
(165)
Thick
Sedimentation
¦ Decel ¦ State ¦ DensityFactor
(166)
Thick
where:
3
Sedimentation = transfer from suspended to sedimented by sinking (g/m -d), if
negative is effectively Resuspension (see below);
162
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
KSed = sedimentation rate (m/d);
DepVel = user input time-series of deposition velocities for cohesives (multi-
layer model only; m/d);
Thick = depth of water or thickness of layer if stratified (m);
Decel = deceleration factor (unitless), see (166);
State = concentration of particulate detrital compartment (g/m3); and
DensityFactor = if salinity is modeled, correction factor for water densities based on
salinity and temperature, see (442).
Table 12: Summary of Detrital Deposition and Resuspension in AQUATOX
Deposition of Suspended Detritus & Phytoplankton
Assumption
Equation
"Classic" AQUATOX model
Sedimentation is a function of Mean Discharge
(165)
Sand-Silt-Clay submodel
Follows "silt" as calculated by the Sand-Silt-Clay
submodel
(235)
Multi-layer Sediment Model
Follows "cohesives" class, (which may be user input
or calculated using the sand-silt-clay model)
(164);
(235)
Sediment Diagenesis
Choice of "Classic" AQUATOX or Sand-Silt-Clay
assumptions
Resuspention of Sedimented Detritus
Assumption
Equation
"Classic" AQUATOX model
Resuspension is a function of Mean Discharge
(165)
Sand-Silt-Clay submodel
Follows "silt" in inorganic sediments model
(233)
Multi-layer Sediment Model
Follows "cohesives" class, (which may be user input
or calculated using the sand-silt-clay model)
(167);
(233)
Sediment Diagenesis
Resuspension is not enabled.
If the discharge exceeds the mean discharge then sedimentation is slowed proportionately (
163
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Figure 102):
If TotDischarge > MeanDischarge then
MeanDischarge
where:
TotDischarge
MeanDischarge
Decel
TotDischarge
else Decel = 1.0
(167)
total epilimnetic and hypolimnetic discharge (m /d); and
mean discharge, recalculated on an annual basis at the beginning of
-3
each year of the simulation (m /d).
164
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Figure 102. Relationship of decel to discharge with a mean discharge of 5 nrVs.
1.2
Mean
discharge
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
InitialCondition = initial condition of detritus (g/m )
5.2 Nitrogen
In the water column, two nitrogen compartments, ammonia
and nitrate, are modeled. Nitrite occurs in very low
concentrations and is rapidly transformed through nitrification
and denitrification (Wetzel, 1975); therefore, it is modeled
with nitrate. Un-ionized ammonia (NH3) is not modeled as a
separate state variable but is estimated as a fraction of
ammonia (177). In the sediment bed, if the optional sediment
diagenesis model is included (see chapter 7), nitrogen is
explicitly modeled; otherwise inorganic nitrogen in the
sediment bed is ignored, but organic nitrogen is implicitly
modeled as a component of sedimented detritus.
Nitrogen: Simplifying Assumptions
• Nitrite is not explicitly modeled
• Both nitrogen fixation and
denitrification are subject to
environmental controls; therefore,
the nitrogen cycle is represented
with considerable uncertainty.
• Lethal effects from un-ionized and
ionized ammonia are assumed
additive.
• Ammonia makes up stoichiometric
imbalances between trophic levels.
In the water column, ammonia is assimilated by algae and macrophytes and is converted to
nitrate as a result of nitrification:
dAmmonia
dt
where:
dAmmonia/dt
Loading
Remineralization =
Nitrify
Assimilation
Washout
Washin
Diffusionseg
TurbDiff
Loading + Remineralization - Nitrify - AssimilationAmmoma
- Washout + Washin ± TurbDiff ± DiffusionSeg + FluxDiagemsis
(169)
Flux
Diagenesis
= change in concentration of ammonia with time (g/m -d);
"3
= loading of nutrient from inflow (g/m -d);
-3
= ammonia derived from detritus and biota (g/m -d), see (169);
= nitrification (g/m -d), see (174);
"3
= assimilation of nutrient by plants (g/m -d), see (171);
"3
= loss of nutrient due to being carried downstream (g/m -d), see (16)
= loadings from linked upstream segments (g/m3d), see (30);
= gain or loss due to diffusive transport over the feedback link between
-3
two segments, (g/m -d), see (32);
= depth-averaged turbulent diffusion between epilimnion and
"3
hypolimnion if stratified (g/m -d), see (22) and (23);
-3
= potential flux from the sediment diagenesis model, (g/m -d), see (273)
Remineralization includes all processes by which ammonia is produced from animal, plants, and
detritus, including decomposition and excretion required to maintain variable stoichiometry (see
Table 14):
166
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Remineralization = PhotoResp + DarkResp + AnimalResp + AnimalExcr
+ DetritalDecomp + AnimalPredation + NutrRelDefecation
+ NutrRelPlantSink + NutrRelMortality + NutrRelGameteLoss
+ NutrRelColonization + NutrRelPeriScour
(170)
where:
PhotoResp
DarkResp
AnimalResp
AnimalExcr
DetritalDecomp
AnimalPredation
NutrRelDefecation
NutrRelPlantSink
NutrRelMortality
NutrRelGameteLoss
NutrRelColonization
NutrRelPeriScour
algal excretion of ammonia due to photo respiration (g/m -d);
"3
algal excretion of ammonia due to dark respiration (g/m -d);
"3
excretion of ammonia due to animal respiration (g/m -d);
animal excretion of excess nutrients to ammonia to maintain
"3
constant org. to N ratio as required (g/m -d);
nitrogen release due to detrital decomposition (g/m -d);
change in nitrogen content necessitated when an animal consumes
"3
prey with a different nutrient content (g/m -d), see discussion in
"Mass Balance of Nutrients" in Section 5.4;
ammonia released from animal defecation (g/m -d);
ammonia balance from sinking of plants and conversion to detritus
(g/m3-d);
ammonia balance from biota mortality and conversion to detritus
(g/m3-d);
ammonia balance from gamete loss and conversion to detritus
(g/m3-d);
ammonia balance from colonization of refractory detritus into labile
"3
detritus (g/m -d);
ammonia balance when periphyton is scoured and converted to
"3
phytoplankton and suspended detritus, (g/m -d);
Nitrate is assimilated by plants and is converted to free nitrogen (and lost) through
denitrification:
dNitrate
dt
where:
Loading + Nitrify - Denitrify - AssimNitrate ~ Washout + Washin
± TurbDiff ± DiffusionSeg +Flux
dNitrate I dt
Washin
Diffusionseg
Loading
Denitrify
FluXDiagenesis
Diagenesis
(171)
change in concentration of nitrate with time (g/m -d);
loadings from linked upstream segments (g/m3 d), see (30);
gain or loss due to diffusive transport over the feedback link between
two segments, (g/m -d), see (32);
user entered loading of nitrate, including atmospheric deposition;
"3
denitrification (g/m -d), see (175);
-3
potential flux from the sediment diagenesis model, (g/m -d), see
(273)
167
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Free nitrogen can be fixed by cyanobacteria. Both nitrogen fixation and denitrification are
subject to environmental controls and are difficult to model with any accuracy; therefore, the
nitrogen cycle is represented with considerable uncertainty.
Figure 103. Components of nitrogen remineralization
Free N (not in
model domain)
Decomposition
Denitri
ication
Excretion
Nitrification
Assimilation
Assimilation
Fixation
Detritus
Plants
Animals
Plants
Ammonia
Nitrate
& Nitrite
AQUATOX will estimate and output total nitrogen (TN) in the water column. Total nitrogen is
the sum of ammonia and nitrate in the water column as well as nitrogen associated with
dissolved and suspended particulate organic matter and phytoplankton (see section 5.4 for further
details).
Assimilation
Nitrogen compounds are assimilated by plants as a function of photosynthesis in the respective
groups (Ambrose et al., 1991):
AssimilationAmmoma = Implant (Photosynthesispkmt ¦ N20rgpkmt ¦ NH4PreJ) (172)
Assimilationmtrate = Ep;®,, (Photosynthesispkmt ¦ N20rgpkmt -(1- NH4Pref» (173)
When internal nutrients are modeled, the equations are slightly different
Assimilation Ammonia = Orient (Phyto UpN ¦ It'3 • NH4Pref) (171b)
168
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Assimilation Ammonia = Y.piant (PhytoUpN ¦ le3 ¦ (1 - NH4Pref))
(172b)
where:
Assimilation
Photosynthesis
N20rg Plant
PhytoUpN
NH4Pref
assimilation rate for given nutrient (g/m -d);
"3
rate of photosynthesis (g/m -d), see (35);
fraction of photosynthate that is nitrogen (unitless, user input as
part of plant underlying data);
"3
uptake of internal nutrients (mg/m -d), see (55e);
ammonia preference factor (unitless).
Only 23 percent of nitrate is nitrogen, but 78 percent of ammonia is nitrogen. This results in an
apparent preference for ammonia. The preference factor is calculated with an equation
developed by Thomann and Fitzpatrick (1982) and cited and used in WASP (Ambrose et al.,
1991):
NH4Pref
where:
N2NH4 ¦ Ammonia ¦ N2N03 ¦ Nitrate
(KN + N2NH4 ¦ Ammonia) ¦ (KN + N2N03 ¦ Nitrate)
(174)
N2NH4 ¦ Ammonia ¦ KN
(N2NH4 ¦ Ammonia + N2N03 ¦ Nitrate) ¦ (KN + N2N03 ¦ Nitrate)
N2NH4 = ratio of nitrogen to ammonia (0.78);
N2N03 = ratio of nitrogen to nitrate (0.23);
KN = half-saturation constant for nitrogen uptake (g N/m3);
-3
Ammonia = concentration of ammonia (g/m ); and
Nitrate = concentration of nitrate (g/m3).
For algae other than cyanobacteria, Uptake is the Redfield (1958) ratio; although other ratios (cf.
Harris, 1986) may be used by editing the parameter screen. At this time nitrogen-fixation by
cyanobacteria is represented by using a smaller uptake ratio, thus "creating" nitrogen. Nitrogen
fixation is not tracked explicitly as a separate rate in the plant's derivative.
Nitrification and Denitrification
Nitrification is the conversion of ammonia to nitrite and then to nitrate by nitrifying bacteria; it
occurs at the sediment-water interface (Effler et al., 1996) and in the water column (Schnoor
1996). The maximum rate of nitrification is reduced by limitation factors for suboptimal
dissolved oxygen and pH, similar to the way that decomposition is modeled, but using the more
restrictive correction for suboptimal temperature used for plants and animals:
Nitrify = KNitri ¦ DOCorrection ¦ TCorr ¦ pHCorr ¦ Ammonia (175)
where:
169
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Nitrify =
KNitri =
DOCorrection =
TCorr =
pHCorr =
Ammonia =
nitrification rate (g/m -d);
maximum rate of nitrification (m/d);
correction for anaerobic conditions (unitless) see (160);
correction for suboptimal temperature (unitless); see (59);
correction for suboptimal pH (unitless), see (162); and
concentration of ammonia (g/m ).
If the Sediment Diagenesis model is used, the KNitri value may need to be decreased to account
for sediment nitrification being represented separately. The nitrifying bacteria have narrow
environmental optima; according to Bowie et al. (1985) they require aerobic conditions with a
pH between 7 and 9.8, an optimal temperature of 3Cf, and minimum and maximum temperatures
of 10"and 60"respectively (Figure 101, Figure 102).
Figure 104. Response to pH, nitrification
Figure 105. Response to temperature, nitrification
EFFECT OF pH
O 0.4
STROGANOV FUNCTION
NITRIFICATION
TMax
10 20 30 40 50
TEMPERATURE (C)
Denitrification is the conversion of nitrate and nitrite to free nitrogen and occurs as an anaerobic
process. However, only a small part of the denitrification occurs at the sediment-water interface
and it can also occur in the water column due to "anoxic microsites" such as the interior of
detrital particles (Di Toro 2001). Therefore, AQUATOX follows the convention of other models
in representing denitrification as a bulk process (by combining sediment and water-column
denitrification). This approach is a change from earlier model versions, including AQUATOX
Release 3.0, where denitrification processes at the sediment-water interface and in the water
column were considered separately. Low oxygen levels enhance the denitfification process
(Ambrose et al., 1991):
Denitrify = KDenitri ¦(!- DOCorrection) • TCorr • pHCorr • Nitrate
where:
(176)
Denitrify
KDenitri
TCorr
pHCorr
denitrification rate (g/m -d);
user-input maximum rate of denitrification (1/d);
effect of suboptimal temperature (unitless), see (59);
effect of suboptimal pH (unitless), see (162); and
170
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Nitrate
concentration of nitrate (g/m ).
KDenitri might need to be reduced when the sediment diagenesis model is included, because
denitrification in the sediment bed is explicitly tracked within that model (see (278))
Furthermore, denitrification is accomplished by a large number of reducing bacteria under
anaerobic conditions and with broad environmental tolerances (Bowie et al., 1985; Figure 103,
Figure 104).
Figure 106. Response to pH, denitrification
EFFECT OF pH
O 0.4
Figure 107. Response to temperature, denitrif.
STROGANOV FUNCTION
DECOMPOSITION
Tlvlax
20 30 40 50
TEMPERATURE (C)
Ionization of Ammonia
The un-ionized form of ammonia, NH3, is toxic to invertebrates and fish. Therefore, it is often
singled out as a water quality criterion. Un-ionized ammonia is in equilibrium with the
ammonium ion, NH4+, and the proportion is determined by pH and temperature. It is useful to
report NH3 as well as total ammonia (NH3 + NH4+).
The computation of the fraction of total ammonia that is un-ionized is relatively straightforward
(Bowie et al. 1985):
1
FracNH 3 = — ;(
\ _|_ IQPkh-pH
NH3 = FracNH3 • Ammonia
2729.92
pkh = 0.09018 +
'/Kelvin
(177)
(178)
(179)
where:
FracNH3
pkh
NH3
Ammonia
TKelvin
fraction of un-ionized ammonia (unitless);
hydrolysis constant;
un-ionized ammonia (mg/L);
total ammonia (mg/L) see (168);
temperature (°K).
171
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
The relative contributions of temperature and pH can be seen by graphing the fraction of un-
ionized ammonia against each of those variables in simulations of Lake Onondaga (Figure 108
and Figure 109). As inspection of the construct would suggest, un-ionized ammonia has a linear
relationship to temperature and a logarithmic relationship to pH, which causes it to be sensitive
to extremes in pH.
Figure 108. Fraction of un-ionized ammonia roughly following temperature.
Fraction NH3
0.07
0.06
25
0.05
20
0.04
Frac NH3
Temp
0.03
0.02
0.01
Sep-88 Apr-89 Oct-89 May-90 Nov-90 Jun-91
Figure 109. Fraction of un-ionized ammonia affected by extreme values of pH.
Fraction NH3
0.07
8.3
8.2
0.06
0.05
0.04
Frac NH3
7.9
0.03
7.8
7.7
0.02
0.01
7.6
7.5
Sep-88 Apr-89 Oct-89 May-90 Nov-90 Jun-91
The construct was verified with the same set of data from Lake Onondaga as was used for the pH
verification (Effler et al. 1996), see section 5.7. It fits the observed data well (Figure 110).
172
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Figure 110. Comparison of predicted and observed fraction of NH3 for Lake Onondaga, NY.
Data from (Effler et al. 1996).
Fraction NH3
¦Frac NH3
-Obs frac NH3
¦Poly. (Obs frac NH3)
Feb- Apr- May- Jul-89 Aug- Oct- Dec-
89 89 89 89 89 89
Ammonia Toxicity
Lethal effects of ammonia on animals have been implemented in AQUATOX based on Update
of Ambient Water Quality Criteria for Ammonia (U.S. Environmental Protection Agency, 1999).
Based on this document, it is preferable to base toxicity on total ammonia, taking into account
the contributions from the un-ionized and ionized ammonia (LC50u and LC50i):
LC 50,
LC 50
t, 8
R
+ -
1
1 + 10^"8 1 + 108"^
R
1 + 10
pH f—pH
(180)
LC 50,
LC 50
t, 8
R
+ -
1
1 + 10^"8 1 + 108"^
1
1 + 10
pH—pHT
(181)
where:
LC50„
LC50,
LC50total ammonia
LC50 for the unionized concentrations of ammonia
LC50 for the ionized concentrations of ammonia.
LC50U + LC501
173
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION CHAPTER 5
pHj = transition pH at which LC50 is the average of the high- and low-
pH intercepts (7.204);
R = shape parameter defined as the ratio of the high- and low-pH
intercepts (0.00704), along with pHT, defines the shape of the
curve;
LC50t § = user-input LC50totalammonia at 20 degrees centigrade and pH of 8.
LC50 parameters derived with the equations above are then applied to the external toxicity
formulation (see section 9.3, equations (429)-(431)). The slope of the Weibull curve is a
constant 0.7 for both forms of ammonia. This value produces the best general match of data
from Appendix 6 from the Ammonia Criteria update (U.S. Environmental Protection Agency,
1999). Lethal effects from un-ionized and ionized ammonia are assumed to be additive.
5.3 Phosphorus
The phosphorus cycle is simpler than the nitrogen cycle.
Decomposition, excretion, and assimilation are important
processes that are similar to those described above. As was
the case with ammonia and nitrate, if the optional sediment
diagenesis model is included (see Chapter 7), flux of
phosphate from the sediment bed may be added to the water
column, especially under anoxic conditions. Additionally,
sorption to calcite may have a significant effect on phosphate predictions in high pH systems due
to precipitation of calcium carbonate. This optional formulation is important to adequately
simulate marl lakes.
Phosphorus: Simplifying
Assumption
• Total bioavailable soluble
phosphorus is modeled
• A constant sorption rate for calcite
is used
• Soluble phosphorus makes up
stoichiometric imbalances between
trophic levels.
dPhosphate , ,,r ,
= Loadmg+Reminerahzation - AssimilationPhosphate-Washout
ut (182)
+ Washin ± TurbDiff ± DiffusionSeg - SorptionP + Flux Dmgenesis
Assimilation = ZPkmt (Photosynthesispkmt ¦ UptakePhosphorus) (183)
where:
3
dPhosphate dt = change in concentration of phosphate with time (g/m -d);
Loading = loading of nutrient from inflow and atmospheric deposition
(g/m3-d);
Remineralization = phosphate derived from detritus and biota (g/m -d), see (183);
Assimilation = assimilation by plants (g/m -d);
TurbDiff = depth-averaged turbulent diffusion between epilimnion and
hypolimnion if stratified (g/m -d), see (22) and (23);
174
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Washout
Washin
Diffusionseg
SorptionP
FIUX£)iagenesis
Photosynthesis
Uptake
loss of nutrient due to being carried downstream (g/m -d), see (16)
loadings from linked upstream segments (g/m3d), see (30);
gain or loss due to diffusive transport over the feedback link
-3
between two segments, (g/m -d), see (32);
rate of sorption of phosphorus to calcite (mgP/L-d), see (218);
"3
potential flux from the sediment diagenesis model, (g/m -d), see
(273)
rate of photosynthesis (g/m -d), see (35), and
fraction of photosynthate that is phosphate (unitless, 0.018).
As was the case with ammonia, Remineralization includes all processes by which phosphate is
produced from animal, plants, and detritus, including decomposition, excretion, and other
processes required to maintain mass balance given variable stoichiometry (see Table 15):
Remineralization = PhotoResp + DarkResp + AnimalResp + AnimalExcr
+ DetritalDecomp + AnimalPredation + NutrRelDefecation
+ NutrRelPlantSink + NutrRelMortality + NutrRelGameteLoss
+ NutrRelColonization + NutrRelPeriScour
(184)
where:
PhotoResp
DarkResp
AnimalResp
AnimalExcr
DetritalDecomp
AnimalPredation
NutrRelDefecation
NutrRelPlantSink
NutrRelMortality
NutrRelGameteLoss
NutrRelColonization
NutrRelPeriScour
algal excretion of phosphate due to photo-respiration (g/m -d);
"3
algal excretion of phosphate due to dark respiration (g/m -d);
"3
excretion of phosphate due to animal respiration (g/m -d);
animal excretion of excess nutrients to phosphate to maintain
constant org. to P ratio as required (g/m -d);
phosphate release due to detrital decomposition (g/m -d);
change in phosphate content necessitated when an animal consumes
prey with a different nutrient content (g/m -d), see discussion in
"Mass Balance of Nutrients" below;
-3
phosphate released from animal defecation (g/m -d);
phosphate balance from sinking of plants and conversion to detritus
(g/m3-d);
phosphate balance from biota mortality and conversion to detritus
(g/m3-d);
phosphate balance from gamete loss and conversion to detritus
(g/m3-d);
phosphate balance from colonization of refractory detritus into
labile detritus (g/m -d);
phosphate balance when periphyton is scoured and converted to
-3
phytoplankton and suspended detritus, (g/m -d);
At this time AQUATOX models only phosphate available for plants; a correction factor in the
loading screen allows the user to scale total phosphate loadings to available phosphate. A default
value is provided for average atmospheric deposition, but this should be adjusted for site
175
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
conditions. In particular, entrainment of dust from tilled fields and new highway construction
can cause significant increases in phosphate loadings. As with nitrogen, the default uptake
parameter is the Redfield (1958) ratio; it may be edited if a different ratio is desired (cf. Harris,
1986).
AQUATOX estimates and outputs total phosphate (TP) in the water column. TP is the sum of
dissolved phosphate in the water column as well as phosphate associated with dissolved and
suspended particulate organic matter and phytoplankton(see section 5.4 for further details).
5.4 Nutrient Mass Balance
Variable Stoichiometry
The ratios of elements in organic matter are allowed to
vary among but not within compartments. This is
accomplished by providing editable fields for N:organic
matter and P:organic matter for each compartment.
Furthermore, the wet to dry ratio is editable for all
compartments; it has a default value of 5.
In order to maintain the specified ratios for each
compartment, the model explicitly accounts for processes
that balance the ratios during transfers, such as excretion
coupled with consumption and nutrient uptake coupled
with detrital colonization. Nutritional value is not
automatically related to stoichiometry in the model, but it
is implicit in default egestion values provided with various food sources. Table 13 shows the
default stoichiometric values suggested for the model, although these can be edited.
Table 13: Default stochiometric values in AQUATOX
Compartment
Frac. N
(dry)
Frac. P
(dry)
Reference
Refrac. detritus
0.002
0.0002
Sterner & Elser 2002
Labile detritus
0.079
0.018
Redfield (1958) ratios
Phytoplankton
0.059
0.007
Sterner & Elser 2002
Cyanobacteria
0.059
0.007
same as phytoplankton for now
Periphyton
0.04
0.0044
Sterner & Elser 2002
Macrophytes
0.018
0.002
Sterner & Elser 2002
Cladocerans
0.09
0.014
Sterner & Elser 2002
Copepods
0.09
0.006
Sterner & Elser 2002
Zoobenthos
0.09
0.014
same as cladocerans for now
Minnows
0.097
0.0149
Sterner & George 2000
Shiner
0.1
0.025
Sterner & George 2000
Perch
0.1
0.031
Sterner & George 2000
Smelt
0.1
0.016
Sterner & George 2000
Bluegill
0.1
0.031
same as perch for now
Trout
0.1
0.031
same as perch for now
Bass
0.1
0.031
same as perch for now
176
Nutrient Mass Balance: Simplifying
Assumptions
• Stoichiometry within each model
compartment is constant over time
• Free nitrogen is not tracked within
AQUATOX
• Nutrients taken up by macrophyte
roots come from sources that are
outside the modeled system
• Mass balance may fail if total
nutrients in the water column drop
to zero (due to inter-organism
interactions)
• Ammonia loadings are assumed to
be 12 to 15% when total nitrate
loadings are input by the user.
• Dissolved nutrients make up
stoichiometric imbalances between
trophic levels.
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Nutrient Loading Variables
Often water quality data are given as total nitrogen and phosphorus. In order to improve
agreement with monitoring data, AQUATOX can accept both loadings and initial conditions as
"Total N" and "Total P." This approach is made possible by accounting for the nitrogen and
phosphorus contributed by suspended and dissolved detritus and phytoplankton and back-
calculating the amount that must be available as freely dissolved nutrients. The precision of this
conversion is aided by the model's variable stoichiometry. For nitrogen:
N =N -N -N (185)
Dissolved Total SuspendedDetritus SuspendedPlants V /
where:
-3
NDissolved = bioavailable dissolved nitrogen (g/m d); see (170);
"3
NTotal = loadings of total nitrogen as input by the user (g/m d);
NSuspendedDetritus = nitrogen in suspended detritus loadings (g/m3 d);
NSuspendedPlants = nitrogen in suspended plant loadings (g/m3 d).
When Total N inputs are used, ammonia is assumed to be a fixed percentage of bioavailable
dissolved nitrogen, based on the type of input:
• Inflow waters: Ammonia content of dissolved inorganic nitrogen = 12%
• Point sources: Ammonia content of dissolved inorganic nitrogen = 15%
• Non-point sources: Ammonia content of dissolved inorganic nitrogen = 12%
These percentages are based on professional judgement, they are averages from several large
data sets. However, if the user wishes to use a different percentage, separate ammonia and
nitrate data sets can be derived from the Total N time-series and input individually.
In acknowledgment of the way it is used in the model, the phosphorus state variable is
designated "Total Soluble P." Phosphorus that is not bioavailable (i.e. immobilized phosphorus
and acid-soluble phosphorus) may be specified using the FracAvail parameter as shown here:
TSP = FracAvail(PTotal ) — PSuspendedDetritus ~ ^SuspendedPlants (1^6)
where:
"3
TSP = bioavailable phosphorus (g/m d); see (181);
FracAvail = user-input bioavailable fraction of phosphorus;
PTotal = loadings of total phosphorus (g/m3 d);
PSuspendedDetritus = phosphorus in suspended detritus loadings (g/m3 d);
pSuspendedPlants = phosphorus in suspended plant loadings (g/m3 d).
Nutrient Output Variables
In order to compare model results with monitoring data, total phosphorus, and total nitrogen are
calculated as output variables. This approach is accomplished by the reverse of the calculations
177
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
for the loadings: the contributions of the nutrient in the freely dissolved state and tied up in
phytoplankton and dissolved and particulate organic matter are calculated and summed.
Carbonaceous biochemical oxygen demand (CBOD5) is estimated considering the sum of
detrital decomposition. The contributions from phytoplankton and labile dissolved and
particulate organic matter are included using an oxygen to biomass conversion factor entered in
the remineralization record.
Mass Balance of Nutrients
Variables for tracking mass balance and nutrient fate are included in the output as detailed
below. Phosphorus and Nitrogen balance mass to machine accuracy. To maintain mass balance,
nutrients are tracked through many interactions.
The mass balance and nutrient fate tracking variables are:
Nutrient Tot. Mass: Total mass of nutrient in the system in kg
Nutrient Tot. Loss: Total loss of nutrient from system since simulation start, kg
Nutrient Tot. Washout: Total washout since simulation start, kg
Nutrient Wash, Dissolved: Washout in dissolved form since simulation start, kg
Nutrient Wash, Animals: Washout in animals since start, kg
Nutrient Wash, Detritus: Washout in detritus since start, kg
Nutrient Wash, Plants: Washout in plants since start, kg
Nutrient Loss Emergel: Loss of nutrients in emerging insects since start, kg
Nutrient Loss Denitrif.: Denitrification since start, kg
Nutrient Burial: Burial of nutrients since start, kg
Nutrient Tot. Load: Total nutrient load since start, kg
Nutrient Load, Dissolved: Dissolved nutrient load since start, kg
Nutrient Load as Detritus: Nutrient load in detritus since start, kg
Nutrient Load as Biota: Nutrient load in biota since start, kg
Nutrient Root Uptake: Load of nutrients into sytem via macrophyte roots since start. (Macrophyte root
uptake is currently assumed to occur from below the modeled sediment layer), kg
Nitrogen Fixation: Load of nitrate into system since start via nitrogen fixation, kg
Nutrient MB Test: Mass balance test, total Mass + Loss - Load: Should stay constant
Nutrient Exposure: Exposure of buried nutrients
Nutrient Net Layer Sink: For stratified systems, sinking since start, kg
Nutrient Net TurbDiff: For stratified systems, Turbdiff since start, kg
Nutrient Net Layer Migr.: For stratified systems, migration since start, kg
Nutrient Total Net Layer: Net nutrient movement to or from paired vertical layer, kg (This value is the
sum of sinking, turbulent diffusion and migration. This quantity also accounts for nutrient
transport caused by water movement when the thermocline depth changes.)
Nutrient Mass Dissolved: Total mass of dissolved nutrient in system, kg
Nutrient Mass Detritus: Total mass of nutrient in detritus in system, kg
Nutrient Mass Animals: Total mass of nutrient in animals in system, kg
Nutrient Mass Plants: Total mass of nutrient in plants in system, kg
It is important to make careful note of the units presented in the list above. Load and loss terms
are calculated in terms of "kg since the start of the simulation," total mass units are "kg at the
current moment."
Simplified diagrams of the nitrogen and phosphorus cycles can be found in Figure 108 and
178
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Figure 109. A full accounting of the 18 nutrient linkages and all external loads and losses for
nitrogen and phosphorus is also provided in Table 14 and Table 15.
There are instances in which nutrients can be moved to and from compartments that are not in
the model domain. For example, when NO3 undergoes denitrification and becomes free nitrogen
the free nitrogen is no longer tracked within AQUATOX. An example of nutrients entering the
model domain comes with the growth of macrophytes. Rooted macrophytes are not limited by a
lack of nutrients in the water column as nutrients are derived from the sediment. Therefore,
when photosynthesis of macrophytes produces growth, the nutrient content within the leaves of
the macrophytes is assumed to originate from the pore waters of the sediments. However, this
implicit "nutrient pumping" is tracked in the mass balance output. Nitrogen fixation is another
addition of nutrients from outside of the model domain that is tracked with the mass balance
output varaible called "N fixation."
Additionally, some simplifications are required as a result of dietary imbalances. For example,
herbivores generally have higher nutrient concentrations than the plants that they are consuming.
When biomass is converted from a plant into an animal through consumption the imbalance has
to be satisfied to maintain mass balance. Sterner and Elser (2002) state: "There is no single way
that consumers maintain their stoichiometry in the face of imbalanced resources." As a
simplification, AQUATOX takes nutrients from the dissolved water-column compartments to
make up this difference (see AnimalPredation in (169)). However, these same herbivores ingest
plants with higher nutrient concentrations than the fecal matter that they defecate. When
biomass is converted from plants to detrital matter through defecation the model simulates a
release of nutrients into the water column (see NutrRelDefecation in (169)). These two
simplifying algorithms, therefore, balance each other for the most part, and such interactions will
have only a minor effect on predicted water-column nutrient concentrations. Likewise, nutrient-
poor refractory detritus is converted to labile detritus through microbial colonization and growth;
this is stimulated by uptake of nutrients from the water column (Sterner and Elser 2002) and is
represented in the model.
179
-------
Figure 111
Nitrogen Cycle in AQUATOX
00
o
animals
mortality, defecation, gameteloss
ingestion
ingestion
mortality
detritus
plants
excretion,
respiration
assimilation
excretion, respiration
decomp.
macrophyte
root uptake
Washout
Loadings
dissolved in water
nitrification
NH
NO
free nitrogen
(not in model domain)
N in pore waters
(outside model domain)
denitrification
-------
Table 14
Nitrogen Mass Balance: Accounting
N03
link
NH4
link
Detritus, Sed.
Refractory
link
Detritus, Sed.
Labile
link
Detritus,
Dissolved
link
Load
external load
Load
external load
Load
external load
Load
external load
Load
external load
Nitrif
from NH4
a
Nitrif
to N03
a
Defecation
from animal
e
Defecation
from animal
e
Decomp (labile)
to NH4
d
DeNitrif
external loss
Assimil
to plant
b
Plant Sedmtn
from plant
f
Plant Sedmtn
from plant
f
Mortality
from anim/plt
k
N03Assim
to plant
b
Excretion
from anim/plt
c,o
Colonz
to SedLabDetr
g
Colonz
from SedRefrDetr
g
Colonz
DissRefr->PartLab
g
Washout
external loss
Respiration
from anim/plt
m,n
Predation
to Animal
h
Predation
to Animal
h
Excretion
from anim/plt
I
TurbDiff
layer accountg
DetritalDecomp
from LabileDetr
d
Sedimentation
from PartRefrDetr
i
Decomp
to NH4
d
Washout
external loss
Washout
external loss
Scour
to PartRefrDetr
j
Sedimentation
from PartLabDetr
i
TurbDiff
layer accountg
TurbDiff
layer accountg
Burial
external loss
Scour
to PartLabDetr
j
Exposure
external load
Burial
external loss
Exposure
external load
Detritus,
Particulate Refr.
link
Detritus, Particulate
Labile
link
Algae
link
Macrophytes
link
Animals
link
Load
external load
Load
external load
Load
external load
Load
external load
Load
external load
mortality
from anim/plt
k
Decomp
to NH4
d
Photosyn
from N03, NH4
b
Photosyn
root uptake, external
Consumption
from anim/plt
h
Colonz
to PartLabDetr
g
mortality
from anim/plt
k
Respiration
to NH4
m
Respiration
to NH4
m
Defecation
to sed detr
e
Washout
external loss
GamLoss
from Animal
q
Photo Resp
to diss detr, NH4
l,c
Photo Resp
to diss detr, NH4
l,c
Respiration
to NH4 if req.
n
Predation
to Animal
h
Colonz
from Diss.PartRefr
g
Mortality
to Diss / Part Detr
k
Mortality
to Part Detr
k
Excretion
to NH4 if req.
0
Sedimentation
to SedRefrDetr
i
Washout
external loss
Predation
to Animal
h
Predation
to animal
h
TurbDiff
layer accountg
Scour
from SedRefrDetr
j
Predation
to Animal
h
Washout
external loss
Breakage
to detr., as mort
k
Predation
to animal
h
SinkToHyp
layer accountg
Sedimentation
to SedLabDetr
i
Sedimntn (Sink)
to Sed Detr
f
Mortality
to Part Detr
k
SinkFromEpi
layer accountg
Scour
from SedLabDetr
j
TurbDiff
layer accountg
GameteLoss
to PartLabDetr
q
TurbDiff
layer accountg
SinkToHypo
layer accountg
SinkToHypo
layer accountg
Drift
external loss
SinkFromEpi
layer accountg
SinkFromEpi
layer accountg
Entrain
external loss
TurbDiff
layer accountg
Sloughing
to detr., phytoplk
r
Promotion
to animal
P
ToxDislodge
to detr., as mort
k
Recruit
from animal
P
Emergel
external loss
Migration
layer accountg
Linkage Notes
a Denitrification from NH4 to N03.
b An appropriate quantity of N03 and NH4 are taken into a plant as part of photosynthesis so that mass balance is maintained,
c When excretion & respiration takes place in plants and animals, all nitrogen lost goes directly to dissolved NH4.
d Labile detritus breaks down and the nutrient content is released as NH4.
e Defecation is split into sedimented-labile and sed-refr detritus 50-50. Excess nitrogen is released as NH4.
Plants sink and are split into sedimented-labile and sed-refr detritus (92-08). Excess nitrogen is released as NH4.
Refractory detritus converts into labile detritus. Any nitrogen imbalance is balanced using NH4 in water.
Animals eat plants and detritus. Animal homeostasis (const, org to N ratio) is managed through Respiration & Excretion.
Suspended sediment sinks and joins bottom sediment. Any change in N between phases is made up using dissolved NH4.
Bottom sediment is scoured up and joins suspended sediment. Any change in N between phases is made up using dissolved NH4.
Animals and plants die and are divided up among suspended and dissolved detritus. Excess nitrogen is released as NH4.
Plants excrete organic matter to dissolved detritus. Excess Nitrogen is released as NH4.
Plant respiration, nutrients are released to NH4
Animal respiration, nutrients are relased to NH4 to maintain animal constant org. to N ratio as required.
Animal excretion of excess nutrients to NH4 to maintain constant org. to N ratio as required.
If young and old age-classes have different ratios, a warning is raised. Prom/Recr takes place outside derivatives so ratios must match.
Through gameteloss, biomass is converted to Part Lab Detr. Excess Nitrogen is released as NH4.
1/3 of periphyton sloughing goes to phytoplankton, 2/3 to detritus as mortality. Nutrients are balanced between compartments.
-------
Figure 112
Phosphorus Cycle in AQUATOX
00
Loadings
mortality, defecation, gameteloss
ingestion
mortality
detritus
assimilation
decomp.
plants
excretion, respiration
phosphate
dissolved in water
animals
ingestion
excretion,
respiration
macrophyte
root uptake
Washout
P in pore waters
(outside model domain)
-------
Table 15
Phosphorus Mass Balance: Accounting
Total Soluble P
link
Detritus, Sed.
Refractory
link
Detritus, Sed.
Labile
link
Detritus,
Dissolved
link
Load
external load
Load
external load
Load
external load
Load
external load
Assimilation
to plant
b
Defecation
from animal
e
Defecation
from animal
e
Decomp (labile) to tsp d
Excretion
from anim/plt
c,o
Plant Sedmtn
from plant
f
Plant Sedmtn
from plant
f
Mortality
from anim/plt k
Respiration
from anim/plt
m,n
Colonz
to Sed Lab Detr
g
Colonz
from SedRefrDetr
g
Colonz
DissRefr->PartLab g
DetritalDecomp
from LabileDetr
d
Predation
to Animal
h
Predation
to Animal
h
Excretion
from anim/plt I
Washout
external loss
Sedimentation
from PartRefrDetr
i
Decomp
to TSP
d
Washout
external loss
TurbDiff
layer accountg
Scour
to PartRefrDetr
j
Sedimentation
from PartLabDetr
i
TurbDiff
layer accountg
Burial
external loss
Scour
to PartLabDetr
j
Exposure
external load
Burial
external loss
Exposure
external load
Detritus,
Particulate Refr.
link
Detritus,
Particulate
link
Algae
link
Macrophytes
link
Animals
link
Load
external load
Load
external load
Load
external load
Load
external load
Load
external load
mortality
from anim/plt
k
Decomp
to TSP
d
Photosyn
from TSP
b
Photosyn
root uptake, external
Consumption
from anim/plt
h
Colonz
to PartLabDetr
g
mortality
from anim/plt
k
Respiration
to TSP
b
Respiration
to TSP
m
Defecation
to sed detr
e
Washout
external loss
GamLoss
from Animal
q
Photo Resp
to diss detr, TSP
l.c
Photo Resp
to diss detr, TSP
l.c
Respiration
to TSP if req.
n
Predation
to Animal
h
Colonz
from Diss,PartRefr
g
Mortality
to Diss / Part Detr
k
Mortality
to Part Detr
k
Excretion
to TSP if req.
l,0
Sedimentation
to SedRefrDetr
i
Washout
external loss
Predation
to Animal
h
Predation
to animal
h
TurbDiff
layer accountg
Scour
from SedRefrDetr
j
Predation
to Animal
h
Washout
external loss
Breakage
to detr., as mort
k
Predation
to animal
h
SinkToHyp
layer accountg
Sedimentation
to SedLabDetr
i
Sedimntn (Sink)to Sed Detr
f
Mortality
to Part Detr
k
SinkFromEpi
layer accountg
Scour
from SedLabDetr
j
TurbDiff
layer accountg
Gamete Loss
to PartLabDetr
q
TurbDiff
layer accountg
SinkToHypo
layer accountg
SinkToHypo
layer accountg
Drift
external loss
SinkFromEpi
layer accountg
SinkFromEpi
layer accountg
Entrain
external loss
TurbDiff
layer accountg
Sloughing
to detr., phytoplk
r
Promotion
to animal
P
ToxDislodge
to detr., as mort
k
Recruit
from animal
P
Emergel
external loss
Migration
layer accountg
oo
Linkage Notes
b An appropriate quantity of phosphorus is taken into a plant as part of photosynthesis so that mass balance is maintained,
c When excretion & respiration takes place in plants and animals (organic matter becomes DOM) additional P lost goes directly to dissolved P.
d Labile detritus breaks down and the nutrient content is released as dissolved P.
e Defecation is split into sedimented-labile and sed-refr detritus 50-50. Excess phosphorus is released as dissolved P.
f Plants sink and are split into sedimented-labile and sed-refr detritus (92-08). Excess phosphorus is released as dissolved P.
g Refractory detritus breaks down into labile detritus. Any P imbalance is balanced using dissolved P in water,
h Animals eat plants and detritus. Animal homeostasis (const, org to P ratio) is managed through Respiration & Excretion,
i Suspended sediment sinks and joins bottom sediment. Any change in P between phases is made up using dissolved P.
j Bottom sediment is scoured up and joins suspended sediment. Any change in P between phases is made up using dissolved P.
k Animals and plants die and are divided up among suspended and dissolved detritus. Excess phosphorus is released as dissolved P.
I Plants and animals excrete organic matter to dissolved detritus. Excess phosphorus is released as dissolved P.
m Plant respiration, nutrients are released to dissolved phosphorus.
n Animal respiration, nutrients are relased to dissolved P to maintain animal constant org. to P ratio as required,
o Animal excretion of excess nutrients to P to maintain constant org. to P ratio as required.
p If young and old age-classes have different ratios, a warning is raised. Prom/Recr takes place outside derivatives so ratios must match,
q Through gameteloss, biomass is converted to Part Lab Detr. Excess phosphorus is released as dissolved P.
r 1/3 of periphyton sloughing goes to phytoplankton, 2/3 to detritus as mortality. Nutrients are balanced between compartments.
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
In some cases, when concentrations of nutrients in the water column drop to zero, perfect mass
balance of nutrients will not be maintained. Nutrient to organic matter ratios within organisms
do not vary over time, therefore transformation of organic matter (e.g. consumption, mortality,
sloughing, and sedimentation) occasionally requires that a nutrient difference be made up from
the water column. If there are no available nutrients in the water column, a slight loss of mass
balance is possible.
The mass associated with each component can be plotted, as in Figure 113.
Figure 113 Distribution of predicted mass of nitrogen in Lake Onondaga NY.
ONONDAGA LAKE, NY (PERTURBED) Run on 03-23-09 3:29 PM
(Epilimnion Segment)
N Mass Dissolved (kg)
N Mass Susp. Detritus (kg)
N Mass Animals (kg)
N Mass Plants (kg)
N Mass Bottom Sed. (kg)
1.0E+6
1.0E+3
1/12/1989 5/12/1989 9/9/1989 1/7/1990 5/7/1990 9/4/1990 1/2/1991
184
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
5.5 Dissolved Oxygen
Oxygen is an important regulatory endpoint; very low levels
can result in mass mortality for fish and other organisms,
mobilization of nutrients and metals, and decreased
degradation of toxic organic materials. Dissolved oxygen is
usually simulated as a daily average and does not account for
diurnal fluctuations (however, see Diel Oxygen below). It is
a function of reaeration, photosynthesis, respiration,
decomposition, and nitrification:
Oxygen: Simplifying Assumptions
• Reaeration is set to zero if ice cover
is predicted
• Cyanobacteria blooms limit the
depth of oxygen reaeration
dOxygen
dt
Loading + Reaeration + Photosynthesized - BOD Respiration
- NitroDemand - Washout + Washin ± TurbDiff ± Diffusion
Seg
(187)
Photosynthesized = 02Photo ¦ IiPlant( Photosynthesis plant)
BOD = 02Biomass ¦ i^Detntus(DecompositionDetntus) )
(188)
(189)
NitroDemand = 02N ¦ Nitrify
(190)
where:
dOxygen/dt
Loading
Reaeration
Photosynthesized
02Photo
BOD
NitroDemand
Washout
Washin
Diffusionseg
02Biomass
Photosynthesis
Decomposition
£ Respiration
02N
Nitrify
change in concentration of dissolved oxygen (g/m -d);
"3
loading from inflow (g/m -d);
"3
atmospheric exchange of oxygen (g/m -d), see (190);
-3
oxygen produced by photosynthesis (g/m -d);
ratio of oxygen to photosynthesis (1.6, unitless);
"3
instantaneous biochemical oxygen demand (g/m -d);
"3
oxygen taken up by nitrification (g/m -d);
"3
loss due to being carried downstream (g/m -d), see (16);
loadings from linked upstream segments (g/m3d), see (30);
gain or loss due to diffusive transport over the feedback link
"3
between two segments, (g/m -d), see (32);
ratio of oxygen to organic matter (unitless);
-3
rate of photosynthesis (g/m -d), see (35), (85);
rate of decomposition (g/m -d), see (159);
"3
sum of respiration for all organisms (g/m -d), (63) and (100);
ratio of oxygen to nitrogen (unitless); and
"3
rate of nitrification (g N/m -d) see (174).
Reaeration is a function of the depth-averaged mass transfer coefficient KReaer, corrected for
ambient temperature, multiplied by the difference between the dissolved oxygen level and the
saturation level (cf. Bowie et al., 1985):
185
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Reaeration = KReaer ¦ (02Sat - Oxygen)
(191)
where:
Reaeration
KReaer
02Sat
Oxygen
-3
mass transfer of oxygen (g/m -d);
depth-averaged reaeration coefficient (1/d);
saturation concentration of oxygen (g/m3), see (198); and
"3
concentration of oxygen (g/m ).
For reaeration in estuaries, see Chapter 10 and equation (445).
In conditions where ice cover is assumed, as well as in the hypolimnetic segment of a stratified
simulation, Reaeration is generally set to zero. However, to prevent excessive oxygen buildup
under these conditions, oxygen is not allowed to exceed two times saturation (02Sat). Any
oxygen buildup beyond two times saturation is added to Reaeration as a loss term.
KReaer may be entered as a constant value within the site's "underlying data." Alternatively,
AQUATOX will calculate KReaer based on the site-type and other characteristics. In standing
water KReaer is computed as a minimum transfer velocity plus the effect of wind on the transfer
velocity (Schwarzenbach et al., 1993) divided by the thickness of the mixed layer to obtain a
depth-averaged coefficient (Figure 111):
Algal blooms can generate dissolved oxygen levels that are as much as 400% of saturation
(Wetzel, 2001). However, near-surface cyanobacteria blooms, which are modeled as being in
the top 0.1 m, produce high levels of oxygen that do not extend significantly into deeper water.
An adjustment is made in the code so that if the cyanobacteria biomass exceeds 1 mg/L and is
greater than other phytoplankton biomass, the thickness subject to oxygen reaeration is set to 0.1
m. This does not affect the KReaer that is used in computing volatilization (see section 8.5).
In streams, reaeration is a function of current velocity and water depth (Figure 112) following the
approach of Covar (1978, see Bowie et al., 1985) and used in WASP (Ambrose et al., 1991).
The decision rules for which equation to use are taken from the WASP5 code (Ambrose et al.,
1991).
If Vel < 0.518 m/sec:
KReaer
(4E-4 + 4E-5• Wind2) • 864
Thick
(192)
where:
Wind
864
Thick
wind velocity 10 m above the water (m/sec);
conversion factor (cm/sec to m/d); and
thickness of mixed layer (m).
TransitionDepth = 0
(193)
else:
186
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
TransitionDepth = 4.411- Vel2'9135 (194)
where:
Vel = velocity of stream (converted to m/sec) see (14); and
TransitionDepth = intermediate variable (m).
If Depth < 0.61 m (but > 0.06), the equation of Owens et al. (1964, cited in Ambrose et al., 1991)
is used:
KReaer = 5.349 ¦ Vet67 • Depth'185 (195)
where:
Depth = mean depth of stream (m).
Otherwise, if Depth is > TransitionDepth, the equation of O'Connor and Dobbins (1958, cited in
Ambrose et al., 1991) is used:
KReaer = 3.93- Vet50 ¦ Depth150
Else, if Depth < TransitionDepth but not <0.60 m, the equation of Churchill et al. (1962, cited in
Ambrose et al., 1991) is used:
KReaer = 5.049 ¦ Vel097 • Depth167 (196)
In extremely shallow streams, especially experimental streams where depth is < 0.06 m, an
equation developed by Krenkel and Orlob (1962, cited in Bowie et al. 1985) from flume data is
used:
234 -(U- Slope)0408
KReaer = 5 (197)
j_j-0.66 V '
where:
U = velocity (converted to fps);
Slope = longitudinal channel slope (m/m); and
H = water depth (converted to ft).
If reaeration due to wind exceeds that due to current velocity, the equation for standing water is
used. Reaeration is set to 0 if ice cover is expected (i.e., when the depth-averaged temperature <
3deg. C).
187
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Figure 114. Reaeration as a Function of Wind Figure 115. Reaeration in Streams
EFFECT OF WIND
OXYGEN, DEPTH = 1 m
10
8
6
4
2
0
12
16
0
4
8
2 6 10 14
WIND (m/s)
Reaeration is assumed to be representative of 20 deg. C, so it is adjusted for ambient water
temperature using (Thomann and Mueller 1987):
where:
KReaer / =
Kreaei'20 =
Theta =
Temperature =
Tf T~i rrii , (Temperature - 20)
TvivCClCrt KJ\6C16T20 ' 1 rlCtQ
Reaeration coefficient at ambient temperature (1/d);
Reaeration coefficient for 20deg. C (1/d);
temperature coefficient (1.024); and
ambient water temperature (deg. C).
(198)
In Release 3, oxygen saturation is calculated using the formulation of Thomann and Mueller
(1987, p 277), see also APHA et al (1995). Oxygen saturation is calculated as a function of
temperature (Figure 113), salinity (Figure 114), and altitude (Figure 118):
02Sat = AltEffect ¦ exp
where
-139.3441
1.57570E + 5
TKelvin
8.62195E + 11
6.6423 IE+ 7 1.2438E + 10
- + -
TKelvirr
TKelvirf
TKelvin4
o 10-754 2140.7
-S\ 0.017674 + -
TKelvin TKelvin2
AltEffect =
100 - (0.0035-3.28083 • Altitude)
100
(199)
and where:
AltEffect
TKelvin
S
Altitude
Fractional reduction in oxygen saturation due to the effects of altitude
(Thomann and Mueller 1987, from Zison et al. 1978);
Kelvin temperature;
salinity driving variable, set to zero if not included in model (ppt); and
site specific altitude (m).
188
DEPTH
(m)
VELOCITY (m/sec)
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Figure 116. Saturation as a Function of Temp.
Oxygen Saturation
Salinity = 0 ppt, Altitude = 0 m
14.0
12 n
E.
i10
« a
Figure 117. Saturation as a Function of Salinity
4.0
10 20 30
Temperature (C)
40
10.0
Oxygen Saturation
Temperature = 20 C, Altitude = 0 m
J
9.5 1
9.0 -j
E
i
8.5 -j
I
!
00
b
7.5 -|
7.0 4
10 20
Salinity (ppt)
30
40
Figure 118. Saturation as a function of altitude
"SS
£_
o
10 -
9
8
7
6
5
Oxygen Saturation
Temperature = 20 C, Salinity = 0 ppt
500 1000 1500
Altitude (m)
2000
Diel Oxygen
Significant fluctuations in oxygen are possible over the course of each day, particularly under
eutrophic conditions. This type of fluctuation may now be captured within AQUATOX when
the model is run with an hourly time-step. If the model is run with a larger reporting time step
(but an hourly integration time-step) the minimum and maximum oxygen concentrations will be
output on the basis of the hourly results.
The instantaneous light climate (28) affects the photosynthesis within the system and this, in
turn, affects the amount of oxygen released into the water column (187). To assist in this
simulation, hourly oxygen loadings may be input into AQUATOX if such data are available.
Alternatively, the effects of oxygen loadings and washout may be turned off, assuming that
upstream processes governing oxygen are producing water concentrations identical to the current
stream segment being modeled; in this way, in-stream processes can be analyzed without being
dominated by upstream loadings.
189
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
AQUATOX can also output minimum and maximum predicted oxygen levels if the "data storage
stepsize" is greater than the "simulation step size." In other words if a simulation is set up with
an hourly step size but results are integrated daily, AQUATOX will plot the mininum and
maximum hourly prediction during that day.
Lethal Effects due to Low Oxygen
AQUATOX represents both lethal and non-lethal effects from low concentrations of dissolved
oxygen. The US EPA saltwater criteria document suggests the following general model for
estimating time to mortality based on data from two species of saltwater juvenile fish, one
species of juvenile freshwater fish, and three species of saltwater larval crustaceans (U.S.
Environmental Protection Agency, 2000, Equation 9):
LCTtme = sl°Pe conc. ' HLC24hours) + Intercept mnc: (200)
exptime exptime
where:
LCnme = Lethal Concentration for a given percentage of a population
over the given duration (mg/L);
Slope co„, = 0.19b LC24hours +0.064 (201)
exptime
and
Intercept conc = 0.392 • LC24hours + 0.204 (202)
exptime
To produce a general model of low oxygen effects, concentrations at which different percentages
are killed (holding duration constant) also need to be related to one another. That is to say, a
model that relates LC5 to LC50 to LC95 must be produced. Examining available data (Figure
119 to Figure 121), a linear model seems appropriate
LCFracdumtjon = Slope conc ¦ LCKnownduratlon + Intercept conc (203)
pctkilled pctkilled
where:
LCFracduration = concentration at which given percentage of organisms are killed
estimated from a known lethal concentration (holding duration
190
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
constant).
LCKnownduration = known lethal concentration for a given percentage of organisms at
the given duration.
Further examination of available data indicates different slopes for different species (Figure
122). Most important, however, is that for all species, the range of slopes is quite narrow,
ranging from -0.001 to -0.01. This indicates that for all species and all durations, the range at
which mortality occurs due to insufficient oxygen is quite narrow. For this reason, the
intermediate value of -0.007 was chosen as it is likely to reproduce available data reasonably
well. This is preferable to having a user input this slope as these data are unlikely to be available
to most users. Given a known lethal concentration at a known duration and using this slope, the
Intercept can be calculated see (204).
Figure 119. Menhaden percent killed vs. 02 exposure concentration
Menhaden Pet Killed vs. Cones
atvarious exposure times (hours)
~ 2
1.8
8
A 16
y=-0.0098x+ 1.5156
1.6
Linear(8)
Linear(72)
~
0.4
0.2
0
0
20
40
60
80
100
Percent Killed
191
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Figure 120. Blue Crab percent killed vs. 02 exposure concentration
BlueCrabPct Killed vs. Cones, at various
exposure times (hours)
2
1
y=-0.007x +1.047
0.8
Linear(6-12)
Linear(14-24)
0.4
y =-0.005x + 0.766
0.2
0
0
50
100
150
Percent Killed
Figure 121. Spot percent killed vs. 02 exposure concentration
Spot Pet. Killed vs. Cones at various
exposure times (hours)
0.9
0.8
y =-0.0019x +0.7678
0.7
0.6
0.5
y = -0.0016x + 0.5778
0.4
0 20 40 60 80 100
Percent Killed
192
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Figure 122. Slope vs. species type
Slope Conc./PctKilled vs. Species Type
0
-0.002
-a
aj
5 -0.004
•*-»
O
Q.
£ -0.006
o
o
g. -0.008
o
OT
-0.01
-0.012
¦
~ Menhaden, Blue Crab
~
~
¦ Spot
~~
~
~ ~
20 40 60 80
duration
100 120
Combining equations (199) to (202), given a user input 24-hour lethal concentration (in the
"Animal underlying data" screen), the model can calculate the fraction killed at a given duration
and at a given concentration.
PctKilled = ¦
^02Conc - 0.204 + 0.064 • \n(/ixpTime)
0.191 • In (ExpTime) + 0.392
- Intercept
pctkilled
-0.007
(204)
where:
Intercept CO)IC = f/'KnownJmMon + 0.007 • PctKilledKnown
conc.
pctkilled
(205)
and:
PctKilled
O2C0TIC
ExpTime
LCKnOWn duration
PctKilled Known
estimated percent killed at a given oxygen concentration and
exposure time;
concentration of oxygen (mg/L);
exposure time (hours);
user input lethal concentration (24-hour) (mg/L);
user input percentage for lethal concentration (percentage);
The model presented in equation (203) requires a user to input 24-hour lethal concentration as
193
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
this is the basis for the general model presented in the saltwater criteria document. If a user has a
lethal concentration at a different duration, the user must estimate the 24-hour lethal
concentration, bearing in mind that the relationship between exposure time and lethal
concentrations is usually logarithmic in nature (Figure 123). There are insufficient data to
develop a general model that will estimate 24-hour lethal concentrations given different user
input durations.
AQUATOX tracks oxygen concentrations over the previous 96 hours from the current time-step.
The oxygen effects model is then applied with the durations shown below:
• 1 hour, 4 hours, 12 hours (when model is run with hourly time-step only)
• 1 day, 2 days, 4 days (relevant to both hourly and daily time-steps)
AQUATOX finds the minimum oxygen concentration over each of these time-periods and
applies it to equation (203). The maximum percent killed over all of the durations tested is then
applied to the animal biomass by increasing mortality (equations (90) and (112)).
Figure 124 shows an example of a three-dimensional response surface produced by this model.
This is a model of low oxygen lethality for Atlantic menhaden produced by entering a 24-hour
LC95 of 0.61 mg/L. Figure 125 shows model predictions using a 24 hour LC50 of 3 mg/L
overlaid on a figure from the U.S. Environmental Protection Agency's 1986 Quality Criteria for
Water. This plot shows that the default value of 3 mg/L works well for many species, but for
white bass, for example, the LC50 should be set to a lower concentration.
Figure 123. LC50 to exposure time based on data from U.S. EPA 2000
LC50to Exposure Time Relationships;
Menhaden and Spot
y = 0.0762ln(x) +0.6332
y = 0.0526ln(x) +0.4773
0.4
~ Spot 88 mm
0.3
¦ Atlantic menhaden 132 mm
0.2
Log. (Spot88mm)
0.1
Log. (Atlantic menhaden 132mm)
0
0
20
40
Exposure Time, Hours
60
80
194
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
(Exposure
Time in
Hours)
Figure 124. Example of low 02 lethality model- menhaden response surface
195
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Figure 125. 96-hour model predictions (in red) compared against continuous exposure data
(Data from U.S. EPA 1986, model set up using a 24-hour LC50 of 3.0mg/L)
o
>
Z3
GO
~o
O
o
CO
I 20
IOO
go
SO
AO
20
1 I 6 [ I I [
zP
~
IF?
I
0 Largemouth Boss
.~Black Crappie
White Sucker
^7 White Bass
• Northern Pike
¦ Channel Catfish
a Walleye
~ Smallmouth Bass
1 till
6 7 8 9 to
Dissolved Oxygen (mg/L)
Non-Lethal Effects due to Low Oxygen
The same three dimensional model used for lethal effects is utilized to calculate non-lethal low
oxygen effects (functions of exposure level and time.) In this case, EC50 reproduction affects
the fraction of gametes that are lost and EC50 growth affects consumption rates.
f
OlEffectFrac =
0-,Conc - 0.204 + 0.064 • In(ExpTime)
0.191- In {ExpTime) + 0.392
-0.007
A
- Intercept
conc.
pctkilled
(206)
and:
Intercept conc = EC50„„ + 0.007 • 50 (207)
pctkilled
where:
02EffectFrac = calculated fraction of gametes lost or reduction in growth rate at a
given oxygen concentration and exposure time;
02Conc = concentration of oxygen (mg/L);
196
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
ExpTime = exposure time (hours);
EC50duration = user input 50% effect concentration (24-hour) (mg/L);
02EffectFrac is then applied to ingestion (91) and gamete loss (126).
5.6 Inorganic Carbon
Many models ignore carbon dioxide as an ecosystem
component (Bowie et al., 1985). However, it can be an
important limiting nutrient. Similar to other nutrients, it is
produced by decomposition and is assimilated by plants; it
also is respired by organisms:
Carbon Dioxide: Simplifying
Assumptions
• Atmospheric exchange is treated
similar to that for oxygen.
• For saltwater systems, an alternative
option is to import a time-series of
equilibrium C02 levels.
dC02
dt
Loading + Respired +Decompose - Assimilation - Washout
+ Washin± C02AtmosExch ± TurbDiff + DiffusionSeg
(208)
where:
Respired = C02Biomass • Z0rgm (RespirationQrgamsm)
Assimilation = IiPla„t( PhotosynthesisPIant ¦ UptakeC02)
(209)
(210)
Decompose = C02Biomass ¦ Y.DetnU DecompDetntm)
(211)
and where:
dC02 dt
Loading
Respired
Decompose
Assimilation
Washout
Washin
Diffusionse
C 02A tmosExch
C02Biomass
Respiration
Decomposition
Photosynthesis
UptakeC02
change in concentration of carbon dioxide (g/m -d);
loading of carbon dioxide from inflow (g/m -d);
carbon dioxide produced by respiration (g/m -d);
carbon dioxide derived from decomposition (g/m -d);
assimilation of carbon dioxide by plants (g/m -d);
loss due to being carried downstream (g/m -d), see (16);
loadings from linked upstream segments (g/m3d), see (30);
gain or loss due to diffusive transport over the feedback
link between two segments, (g/m -d), see (32);
interchange of carbon dioxide with atmosphere (g/m3-d);
ratio of carbon dioxide to organic matter (unitless);
rate of respiration (g/m -d), see (63) and (100);
rate of decomposition (g/m -d), see (159);
rate of photosynthesis (g/m -d), see (35); and
ratio of carbon dioxide to photosynthate (= 0.53).
197
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Carbon dioxide also is exchanged with the atmosphere; this process is important, but is not
instantaneous: significant undersaturation and over saturation are possible (Stumm and Morgan,
1996). The treatment of atmospheric exchange is similar to that for oxygen:
C02AtmosExch = KLiqC02 ¦ (C02Sat - C02)
(212)
In fact, the mass transfer coefficient is based on the well-established reaeration coefficient for
oxygen, corrected for the difference in diffusivity of carbon dioxide as recommended by
Schwarzenbach et al. (1993):
where:
C 02A tmosExch
KLiqC02
C02
C02Equil
KReaer
MolWt02
MolWtC02
KLiqC02 = KReaer
/ \ 0.25
f MolWtO, '
KMolWtC02 j
(213)
interchange of carbon dioxide with atmosphere (g/m -d);
depth-averaged liquid-phase mass transfer coefficient (1/d);
concentration of carbon dioxide (g/m );
equilibrium concentration of carbon dioxide (g/m3), see (213);
depth-averaged reaeration coefficient for oxygen (1/d), see (191)-
(195);
molecular weight of oxygen (=32); and
molecular weight of carbon dioxide (= 44).
Keying the mass-transfer coefficient for carbon dioxide to the reaeration coefficient for oxygen
is very powerful in that the effects of wind (Figure 123) and the velocity and depth of streams
can be represented, using the oxygen equations (Equations (191)-(195)).
Figure 126. Carbon dioxide mass transfer
EFFECT OF WIND
CARBON DIOXIDE, DEPTH = 1 m
10
8
6
4
2
0
0 2 4 6 8 10 12 14 16
WIND (m/s)
Based on this approach, the predicted mass transfer under still conditions is 0.92, compared to
the observed value of 0.89 ± 0.03 (Lyman et al., 1982). This same approach is used, with minor
modifications, to predict the volatilization of other chemicals (see Section 8.5). Computation of
equilibrium of carbon dioxide is based on the method in Bowie et al. (1985; see also Chapra and
198
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Reckhow, 1983) using Henry's law constant, with its temperature dependency (Figure 124), and
the partial pressure of carbon dioxide:
C02Equil = C02Henry ¦ pC02
where:
2385.73
C02Henry = MC02 ¦ 10 TKelvin
-14.0184 + 0.0152642 • TKelvin
(214)
(215)
TKelvin = 2 73.15 + Temperature
(216)
and where:
C02Equil
C02Henry
pC02
MC02
TKelvin
Temperature
equilibrium concentration of carbon dioxide (g/m );
"3
Henry's law constant for carbon dioxide (g/m -atm):
atmospheric partial pressure of carbon dioxide (= 0.00035);
mg carbon dioxide per mole (= 44000);
temperature in deg.K, and
ambient water temperature (deg. C).
Figure 127. Saturation of carbon dioxide
CARBON DIOXIDE SATURATION
12 16.5 21 25.5 30
TEMPERATURE (C)
The equilibrium CO2 equations described above cannot be applied to a seawater system as the
chemistry in seawater is significantly different from freshwater. Over the years, several models
and constants used to describe the dissociation of carbon dioxide in seawater have been proposed
by investigators.
For saline conditions, the equilibrium parameters of the CO2 system can be derived by using
C02SYS (Yuan, 2006) or C02calc (USGS, 2010) and the results used as inputs for C02Equil in
the AQUATOX simulation. Within these models, the user needs to provide two of the five
measurable CO2 system parameters: Total alkalinity (TA), Total carbon dioxide (TC02), pH
and Partial pressure of carbon dioxide (pC02) or fugacity of carbon dioxide (fC02); along with
temperature (T), pressure (P) and salinity (S). The user can then select appropriate constants
199
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
from proposed literature values and the program will calculate the remaining carbonate system
parameters including a time-series of CO2 concentrations in water.
For maximum flexibility, AQUATOX has an interface that will accept these time-series of
C02Equil. In this manner, the user can select the most appropriate model for their site and
import these values into the AQUATOX interface. A time-series of pH can also be estimated by
these ocean-water chemistry models.
5.7 Modeling Dynamic pH
Dynamic pH is important in simulations for several reasons:
o pH affects the ionization of ammonia and potential
resulting toxicity;
o pH affects the hydrolysis and ionization of organic
chemicals which potentially has effects on chemical
fate and the degree of toxicity;
o pH also affects the decay of organic matter and denitrification of nitrate which could
eventually feed back to the animals;
o if pH exceeds 7.5, calcite precipitation can take place which has a significant effect
on the food-web.
A user-input time-series of pH levels may be used to drive the model or AQUATOX can
calculate pH levels.
Many models follow the example of Stumm and Morgan (1996) and solve simultaneous
equations for pH, alkalinity, and the complete carbonate-bicarbonate equilibrium system.
However, this approach requires more data than are often available, and the iterative solution of
the equations entails an additional computational burden—all for a precision that is unnecessary
for ecosystem models. The alternative is to restrict the range of simulated pH to that of normal
aquatic systems and to make simplifying assumptions that allow a semi-empirical computation of
pH (Marmorek et al. 1996, Small and Sutton 1986). That is the approach taken for AQUATOX.
The computation is good for the pH range of 3.75 to 8.25, where the carbonate ion is negligible
and can thus be ignored. (Any predictions above 8.25 are truncated to 8.25 and any predictions
below 3.75 are set to 3.75.) The derivation is given by Small and Sutton (1986), with a
correction for dissolved organic carbon (Marmorek et al. 1996). It incorporates a quadratic
function of carbon dioxide; and it is a nonlinear function of mean alkalinity and the
concentration of refractory dissolved organic carbon (humic and fulvic acids), by means of an
inverse hyperbolic sine function:
( Alkalinity - 5.1 • DOC\
pHCalc = A + B ¦ ArcSmH
v C J (217)
Dynamic pH: Simplifying
Assumptions
• Simple semi-empirical formulation
• Computation is good for the pH
range of 3.75 to 8.25
200
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION CHAPTER 5
where:
pHCalc
ArcSinH
Alkalinity
DOC
5.1
A - - Log^j Alpha
B = 1 / ln(l 0)
C = 2- yj Alpha
Alpha = H2C03 * • CC02 + pkw
H2C03 * ~~ lo-(e,,-o.on,.T + o,oon.T.Tyo,2
where:
H2C03 * =
first acidity constant;
CC02
CO2 expressed as [j,eq/L; see (207) multiplied by conversion factor
of 22.73 (ueq/mg);
pkw
ionization constant for water (le-14);
T
temperature (°C); see (24);
0.92
correction factor for dissolved CO2.
Calibration and verification of the construct used data from nine lakes and ponds in the National
Eutrophication Survey (U.S. Environmental Protection Agency, 1977), two observations on Lake
Onondaga, NY, from before and after closure of a chlor-alkali plant (Effler et al., 1996), and one
observation in a river (Figure 128). The correction factor for CO2 was obtained by fitting the
data to the unity line, but ignoring the two highest points because the construct does not predict
pH above 8.25.
pH;
inverse hyperbolic sine function;
mean Gran alkalinity ([j,eq CaCOs/L);
refractory dissolved organic carbon (mg/L); sum of (143), (144);
average [j,eq of organic ions per mg of DOC;
201
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION CHAPTER 5
Figure 128. Comparison of predicted and observed pHs from selected lakes.
Observed vs. Predicted pH
x
0¦ 8.5
6.0 7.0 8.0
Predicted pH
9.0
10.0
The construct also was verified using time-series data from Lake Onondaga, NY (Figure 129).
The observed data were interpolated from the 2-m depth pH isopleths on a graph (Effler et al.
1996), introducing some uncertainty into the comparison.
Figure 129. Comparison of predicted and observed pH values for Lake Onondaga, NY.
Data from (Effler et al. 1996).
Predicted pH, Lake Onondaga NY
-AQUATOX
—Observed
Poly. (Observed)
202
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
5.8 Modeling Calcium Carbonate Precipitation and Effects
Precipitation of calcium carbonate (mostly calcite in
freshwater), with the potential for sorption and removal of
phosphorus, is modeled as an extension of the pH
approach. The prediction of pH in AQUATOX does not
extend past 8.25 because the carbonate-bicarbonate system
becomes dominant. We use a predicted pH of 7.5 as a
threshold for precipitation of calcium carbonate in
freshwater ecosystems. Almost all calcite is formed
biogenically, primarily by plants using bicarbonate as a source of carbon (McConnaughey et al.
1994). Even "whitings" (sudden precipitation of fine-grained calcite) have been shown to be a
consequence of cyanobacteria photosynthesis (Thompson et al. 1997). Calcareous plants are
characterized by pH polarization with acidic and alkaline poles; calcification occurs at the
alkaline pole (McConnaughey et al. 1994). Proton generation leads to formation of twice as
much CO2 than is used in the process, providing CO2 that is immediately taken up for
photosynthesis. As a result, calcification and photosynthesis use equivalent moles of C, as
shown by both theory and experiments (McConnaughey et al. 1994). Three chemical reactions
represent this process:
Calcite Precipitation: Simplifying
Assumptions
• Biogenic origin
• pH of 7.5 is considered as a
threshold for precipitation
• Dissolved phosphate sorbs to
calcium carbonate but desorption is
not modeled
Ca2+ + CO2 + H20 CaC03 + 2H+
2H+ + 2HC03" 2C02 + 2H20
Ca2+ + 2HC03" -> CaC03 + C02 + H20
Not all plants can use bicarbonate. However, it is difficult to generalize; mosses do not and
many chrysophytes (golden algae) do not. Evidence suggests that other groups, including
greens, cyanobacteria, diatoms, and macrophytes, have species that do use bicarbonate and that
these will dominate in alkaline systems.
The algorithm simulates precipitation of calcite as being the molar equivalent to photosynthesis
of most plants and as occurring when the threshold pH of 7.5 is reached:
If PH >— 7.5 then CalcitePcpt - CICalcite
Photosynthesis
PlantSubset
C20M
(218)
where:
pH =
CalcitePcpt =
C2Calcite =
Photosynthesis =
PlantSubset =
C20M
pH calculated by Eq. 204 or observed time series;
calcite precipitated (mg calcite/L • d);
stoichiometric constant for C and calcite (8.33, g calcite /g C);
rate of photosynthesis for a subset of plants (g/m • d);
all plants except Bryophytes and Other Algae;
stoichiometric constant for C and organic matter (1.9, g C/g OM).
Precipitated calcite is protected, in part, by sorbed organic material. Therefore, it is assumed to
be insoluble—an assumption also made in the sediment diagenesis model (Di Toro 2001).
Because the settling rate is fast, it is also assumed that the calcite goes directly to the sediment.
203
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 5
Phosphorus is adsorbed to the surface and coprecipitates with calcium carbonate (Wetzel 2001).
The rate of coprecipitation seems to be dependent on the rate of calcite precipitation (Otsuki and
Wetzel 1972). However, the sorption is weak and can be reversed easily (Murphy et al. 1983).
Therefore, the default partition coefficient (300 L/kg) is based on equilibration experiments with
sediments from a marl lake (Van Rees et al. 1991).
SorptionP = KDPCalcite ¦ Phosphate ¦ CalcitePcpt ¦ \e - 6
(219)
where:
SorptionP
KDPCalcite
Phosphate
1 e-6
rate of sorption of phosphorus to calcite (mgP/L • d);
partition coefficient for phosphorus to calcite (L/kg);
concentration of phosphorus in water (mg P/L) (see (181));
conversion factor (kg/mg).
Ironically, precipitation is impeded by phosphorus levels that are too high. The threshold for
inhibition is about 30 mg-P/L (Neal 2001). Furthermore, dissolved organic matter also can
inhibit precipitation, with 120 mg C/L being the threshold (Neal 2001). However, these
concentrations are so high that they are ignored in the model.
204
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 6
6. INORGANIC SEDIMENTS
Inorganic sediments can have significant effects on light climate and inorganic sediment effects
on biota can also be explicitly modeled (see the section on suspended sediment effects starting
on page 119). Release 3 of AQUATOX contains four levels of inorganic sediment submodels:
• a very simple model based on a regression relationship between sediment deposition and
total suspended sediments, see (122). This approach should be used when the only
inorganic sediment data available are TSS. Add the "TSS" state variable to use this
option.
• a simple inorganic sediments submodel described in Section 6.1. This model can be used
to estimate the scour and deposition of inorganic sediments at a site as a function of water
flows; therefore it is only applicable to streams and rivers. This model requires additional
data about the types of inorganic sediments (i.e., sand, silt, or clay) and their average rate
of scour and deposition under different water-flow regimes. This model may be selected
under the sediment menu by choosing "Add Sand Silt Clay Model."
• a complex multiple-layer sediment submodel described in Section 6.2. This model can
be used to estimate the sequestration of organic toxicants within the deeper layers of the
sediments and the potential for scour of such toxicants from these deep layers. This
submodel should be linked to a hydrodynamic model to calculate the scour and
deposition of sediments in the modeled segment. This model may be selected under the
sediment menu by choosing "Add Multi-Layer Sediment." Additional layers may also be
added or removed using the options listed under the sediment menu.
• a sediment diagenesis model described in Section 7. This model provides a more
sophisticated accounting of the decay of organic matter and remineralization in an
anaerobic sediment bed and the effects on sediment oxygen demand. The diagenesis
model assumes a depositional environment; scour of sediments is not incorporated. This
model may be selected under the sediment menu by choosing "Add Sediment
Diagenesis."
Within an AQUATOX simulation it is also possible to ignore the effects of inorganic sediments
on ecosystem characteristics altogether, by including none of the models listed above. However,
the model will always track the remineralization of organic material within the sediment bed and
the water column.
6.1 Sand Silt Clay Model
The original version was contributed by Rodolfo Camacho of
Abt Associates Inc. AQUATOX simulates scour, deposition
and transport of sediments and calculates the concentration of
sediments in the water column and sediment bed within a
river reach. For running waters, the sediment is divided into
three categories according to the particle size:
• sand, with particle sizes between 0.062 to 2.0
millimeters (mm),
Sand, Silt, Clay: Simplifying
Assumptions
• River reach is short and well-mixed
• Channel is rectangular
• Daily average flow regime
determines scour, and deposition
• Model for streams / rivers only
205
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 6
• silt (0.004 to 0.062 mm), and
• clay (0.00024 to 0.004 mm).
Wash load (primarily clay and silt) is deposited or eroded within the channel reach depending on
the daily flow regime. Sand transport is also computed within the channel reach. The river
reach is assumed to be short and well mixed so that concentration does not vary longitudinally.
Flow routing is not performed within the river reach. The daily average flow regime determines
the amount of scour, deposition and transport of sediment. Scour, deposition and transport
quantities are also limited by the amount of solids available in the bed sediments and the water
column.
Within the bed, the mass of sediment in each of the three sediment size classes is a function of
the mass in the previous time step, and the mass of sediment in the overlying water column lost
through deposition, and gained through scour:
MassBedsed = MassBedsed t=-i + (DepositSed - ScourSed)' VolumeWater' TimeStep (220)
mass of sediment in channel bed (kg);
mass of sediment in channel bed on previous day (kg);
amount of suspended sediment deposited (kg/m d); see (230);
"3
amount of silt or clay resuspended (kg/m d); see (227);
volume of stream reach (m ); see (2); and
derivative time-step (d).
sediment size classes are calculated as:
where:
MassBedsed =
MassBedsed, t = -i
Depositsed
ScOtir Sed
Volume Water
TimeStep =
The volumes of the respective
Tr , _ MassBedSed
VOltllflC Sed
Rhosed
(221)
where:
Volumesed = volume of given sediment size class (m );
MassBedsed = mass of the given sediment size class (kg);
Rhosed = density of given sediment size class (kg/m3);
Rhosand = 2600 (kg/m3); and
Rhosiit, Clay = 2400 (kg/m3).
The porosity of the bed is calculated as the volume weighted average of the porosity of its
components:
BedPorosity = TFracsed ¦ P°rositySed (222)
where:
BedPorosity = porosity of the bed (fraction);
Fracsed = fraction of the bed that is composed of given sediment class; and
Porositysed = porosity of given sediment class (fraction).
206
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION CHAPTER 6
The total volume of the bed is calculated as:
BedVolume = ^umesand + Volume sat + Volumeciay (223)
1 - BedPorosity
where:
-3
BedVolume = Volume of the bed (m ).
The depth of the bed is calculated as
,, , BedVolume
BedDepth = (224)
ChannelLength ¦ ChannelWidth
where:
BedDepth = depth of the sediment bed (m);
ChannelLength = length of the channel (m); and
ChannelWidth = width of the channel (m).
The concentrations of silt and clay suspended in the water column are computed similarly to the
mass of those sediments in the bed, with the addition of loadings from upstream and losses
downstream:
KgLoad„,
Concsed = + ConcSed,t=-i + ScourSed - DepositSed - Washsed (225)
{J - o64UU
where:
"3
Concsed = concentration of silt or clay in water column (kg/m );
Concsed, t = -i = concentration of silt or clay on previous day (kg/m );
KgLoadsed = loading of clay or silt (kg/d);
Q = flow rate (m3/s);
86400 = conversion from m3/s to m3/d;
"3
Scour sed = amount of silt or clay resuspended (kg/m ); see (227);
Depositsed = amount of suspended sediment deposited (kg/m3); see (230); and
Washsed = amount of sediment lost through downstream transport (kg/m );
see (231).
The concentration of sand is computed using a totally different approach, which is described in
the section on Sand below.
Deposition and Scour of Silt and Clay
Relationships for scour and deposition of cohesive sediments (silts and clays) used in
AQUATOX are the same as the ones used by the Hydrologic Simulation Program in Fortran
(HSPF, U.S. Environmental Protection Agency, 1991). Deposition and scour of silts and clay
are modeled using the relationships for deposition (Krone, 1962) and scour (Partheniades, 1965)
as summarized by Partheniades (1971).
207
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 6
Shear stress is computed as (Bicknell et al., 1992):
Tau = H20Density ¦ Slope ¦ HRadius (226)
where:
Tau = shear stress (kg/m2);
H20Density = density of water (1000 kg/m3);
Slope = slope of channel (m/m);
and hydraulic radius (HRadius) is (Colby and Mclntire, 1978):
HRadius V' Wld'h (227)
2-Y + Width
where:
HRadius = hydraulic radius (m);
Y = average depth over reach (m); and
Width = channel width (m).
Resuspension or scour of bed sediments is predicted to occur when the computed shear stress is
greater than the critical shear stress for scour:
if Tau > TauScoursed then
0 Erodibility„ , f Tau ^ (228)
Scour^ 1
Sed ~y
TauScoursed
J
where:
"3
Scoursed = resuspension of silt or clay (kg/m d);
Erodibilitysed = erodibility coefficient (0.244 kg/m2 d); and
TauScoursed = critical shear stress for scour of silt or clay (kg/m2).
The amount of sediment that is resuspended is constrained by the mass of sediments stored in the
bed. An intermediate variable representing the maximum potential mass that can be scoured is
calculated; if the mass available is less than the potential, then scour is set to the lower amount:
Check sed ~ Scour sed' Volumewater (229)
if Mass sed - Check sed then
0 _ Masssed (230)
Scour Sed
VoluWieWater
where:
Checksed = maximum potential mass (kg); and
208
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 6
Masssed = mass of silt or clay in bed (kg).
Deposition occurs when the computed shear stress is less than the critical depositional shear
stress:
if Tau < Taiil)ep,. ,il then
f [7'w ¦SecPerDay
DepositSed = Cone
' Sed
l-e
Tau
TauDePsed
(231)
where:
Deposit sed
TauDepsed
Concsed =
VTsed
SecPerDay =
amount of sediment deposited (kg/m day);
critical depositional shear stress (kg/m2);
"3
concentration of suspended silt or clay (kg/m );
terminal fall velocity of given sediment type (m/s); and
86400 (seconds / day).
The terminal fall velocity is specified in the site's underlying data.
Downstream transport is an important mechanism for loss of suspended sediment from a given
stream reach:
where:
Washsed
day);
Disch
CoTIC Sed
SegVolume
Washsed -
Disch ¦ Concsed
SegVolume
(232)
amount of given sediment lost to downstream transport (kg/m
discharge of water from the segment (m /day);
concentration of suspended sediment (kg/m );
volume of segment (m ).
When the inorganic sediment model is included in an AQUATOX stream simulation, the
deposition and erosion of detritus mimics the deposition and erosion of silt. The fraction of
detritus that is being scoured or deposited is assumed to equal the fraction of silt that is being
scoured or deposited. The following equations are used to calculate the scour and deposition of
detritus:
Frac Scour Detritus -Frac Scour sm ~ Scour sm' ^umesilt (233)
Masssm
where:
FracScour
Scour sat
Volumesiit
Scour Detritus - Frac Scour
Detritus
Cone
AllSedDetritus
1000
fraction of scour per day (fraction/day);
amount of silt scoured (kg/m3 day) see (227);
volume of silt initially in the bed (m );
(234)
209
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 6
Masssnt = mass of silt initially in the bed (kg);
ConcAiisedDetritus = all sedimented detritus (labile and refractory) in the stream bed
(kg/m3);
ScourDetritus = amount of detritus scoured (g/m3 day); and
1000 = conversion of kg to g.
The equations for deposition of detritus are similar:
Frac DepositionL
Frac Deposition
Depositionm ¦ 1000
Silt
Cone silt
DepositionDetntus = Frac DepositionDetntus ¦ Cone SuSPDetntuS
where:
(235)
(236)
Depositionsiit
amount
Cone sut
amount
FracDeposition
fraction
ConC SuspDetritus
amount
and
DepositionDetritus
amount
amount of detritus deposited (g/m day).
Scour, Deposition and Transport of Sand
Scour, deposition and transport of sand are simulated using the Engelund and Hansen (1967)
sediment transport relationships as presented by Brownlie (1981). This relationship was selected
because of its simplicity and accuracy. Brownlie (1981) shows that this relationship gives good
results when compared to 13 others using a field and laboratory data set of about 7,000 records.
PotConc
Sand
0.05-
Rho
Velocity ¦ Slope
where:
PotConCSand
Rho
RhOSand
Velocity
Slope
1 ) Sand
TauStar
RhOSand-Kh° \ RhOsand-Rho „ , nnn
ll ' 5 ' Vsand/1UUU
V Rho
¦ -J TauStar (237)
potential concentration of suspended sand (kg/m );
density of water (1000 kg/m3)
"3
= density of sand (2650 kg/m );
flow velocity (converted to m/s);
slope of stream (m/m);
mean diameter of sand particle (0.30 mm converted to m); and
dimensionless shear stress.
The dimensionless shear stress is calculated by:
Rho
TauStar
Rhosand ~ Rho
¦ HRadius ¦ ¦
Slope
DSand/1000
(238)
210
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 6
where:
HRadius = hydraulic radius (m).
Once the potential concentration has been determined for the given flow rate and channel
characteristics, it is compared with the present concentration. If the potential concentration is
greater, the difference is considered to be made available through scour, up to the limit of the
bed. If the potential concentration is less than what is in suspension, the difference is considered
to be deposited:
Check Sand PotCoflCsand VolUTYlSwater
(239)
MassSusp Sand = Cone sand • VolumeWater (240)
TotalMasssand = MmnSuap Sand + MassBedSand
(241)
if Checksand ^ MassSuspSand then
DepositSand = MassSuspSand - Checksand (242)
ConCSand PotConCSand
if Checksand ^ TotalMasssand then
MassBedsand= 0
(243)
„ _ TotalMasssand
COnCsand
VolumeWater
if Checksand > MassSuspSand and < TotalMasssand then
Scoursand= Checksand - MassSuspSand (244)
_ MassSuspSand + Scoursand
COnCSand
VolumeWater
211
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 6
Suspended Inorganic Sediments in Standing Water
At present, AQUATOX does not compute settling of inorganic sediments in standing water or
scour as a function of wave action. However, suspended sediments are important in creating
turbidity and limiting light, especially in reservoirs and shallow lakes. Therefore, the user can
provide loadings of total suspended solids (TSS), and the model will back-calculate suspended
inorganic sediment concentrations by subtracting the simulated phytoplankton and suspended
detritus concentrations:
where:
InorgSed = TSS - Z Phyto - X PartDetr (245)
"3
InorgSed = concentration of suspended inorganic sediments (g/m );
TSS = observed concentration of total suspended solids (g/m );
Phyto = predicted phytoplankton concentrations (g/m3); and
PartDetr = predicted suspended detritus concentrations (g/m ).
A radio button on the TSS loadings screen is used to specify whether user-input TSS loadings are
"total suspended (inorganic) sediments" or "total suspended solids." If "inorganic sediments"
are specified then equation (244) is not required as the TSS loading is not assumed to include
phytoplankton or organic matter.
212
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 6
6.2 Multi-Layer Sediment Model
As an alternative to the simple sand-silt-clay model
described above (section 6.1), AQUATOX also includes a
complex multiple layer sediment transport model. This
model can simulate up to ten bottom layers of sediment.
Within each sediment layer, the state variables consist of
inorganic solids, pore waters, labile and refractory
dissolved organic matter in pore waters, and sedimented
detritus. Nutrient concentrations are not modeled in the
pore waters of the sediment layers, although dissolved
organic matter is. Each of these state variables can also have up to twenty organic toxicant
concentrations associated with it. The AQUATOX sediment transport component is summarized
in Figure 130.
Multi-Layer Sediment Model:
Simplifying Assumptions
• Top layer is "active layer" that
interacts with the water column
• Individual sediment layers are well-
mixed
• Density of each sediment layer
remains constant
• Hardpan barrier assumed at the
bottom of the svstem
Figure 130: Components of the AQUATOX sediment transport model and units.
Susp. Inorg.
Solids mg/Lwc
£ _
^ O
r~ x
H H
U U
Inorganic
Solids g/m2
S H
3tI *
n n
u u
Inorganic
Solids g/m2
S H
3tI *
Kfl H
II II
Inorganic
Solids g/m2
S H
31 S
Water Col.
£ —|
m3
O
% x
Pore Water
(5 H
m3/m2
i— o
-a X
Pore Water
m3/m2
<5 H
j=8
Pore Water
m3/m2
<5 H
j=8
s
DOM in Water
Col. mg/Lwc
Tox
M0/I-WC
Susp. Detr mg/L
r~ x
DOM in Pore
Water mg/Lpw
<5 H
,-8
Sed Detr mg/L
S H
^ O
f *
DOM in Pore
Water mg/Lpw
<5 H
,-8
Buried Detr g/m2
<5 H
3^ °
DOM in Pore
Water mg/Lpw
S H
j=8
Buried Detr g/m2
(5 H
3^ °
The AQUATOX sediment submodel was designed to be nearly identical in concept to IPX (In-
Place Pollutant eXport) version 2.7.4 (Velleux et. al 2000). Erosion and deposition cause
changes in the mass of sediments in the top or "active" layer. When the active layer becomes too
large or too small, a conveyor-belt action takes place moving all of the layers up or down intact
("pez dispenser" action). Because all layers are assumed well-mixed, moving partial layers up
and down and then recalculating concentrations within sediment layers would result in too much
mixing throughout the sediment layers (and advection of pollutants from the bottom layer to the
top). During development, the AQUATOX sediment submodel was closely tested against the
IPX model and precisely reproduced results from that model.
213
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 6
Within AQUATOX, inorganic sediments in layered sediments are represented as three distinct
state variables: cohesives (clay), non-cohesives (silt), and non-cohesives2 (sand). These
correspond to the variables described in Section 6.1.
For each inorganic compartment, the sediment transport model accepts daily input parameters for
interactions between the top sediment layer and the water column. These interactions are input
as daily scour and daily deposition for each inorganic sediment type in units of grams per day.
The model also requires deposition and erosion velocities for cohesive inorganic sediments.
These inputs are then used to calculate the deposition and erosion of organic matter within the
system.
AQUATOX assumes that the density of each sediment layer will remain constant throughout a
simulation. Because of this, the volume and thickness of the top bed layer will vary in response
to deposition and erosion. Additionally, the surface area of the multi-layer sediment bed is set to
remain constant. Even if the sediment surface at a site grows or shrinks due to water volume
changes, this model tracks sediments under the initial-condition surface area.
When the top layer has reached a maximum thickness, it is broken into two layers. Other layers
in the system are moved down one layer without disturbing their concentrations or thicknesses.
This allows the model to maintain a toxicant concentration gradient within the sediment layers
during depositional regimes. Similarly, when the top layer has eroded to a minimum size, the
layer beneath it is joined with the active layer to form a new top layer. In this case, lower layers
are moved up one level, without changing their concentrations, densities, or thicknesses. More
details about these processes can be found in section on sediment layer interactions below.
At the bottom of the system, a hardpan barrier is assumed. The model, therefore, has no
interaction beneath its lowest layer. If enough erosion takes place so that this hardpan barrier is
exposed, no further erosion will be possible. Deposition can, however, rebuild the sediment
layer system. This hardpan bottom prevents the artificial inclusion of "clean" sediment and
organic matter into the model's simulation during erosional events. Because it is a barrier and
not a boundary, it prevents loss of toxicant to the system under depositional regimes.
AQUATOX writes output data for a fixed number of sediment layers. When, due to deposition,
a layer is buried below the fixed number of sediment layers, AQUATOX keeps track of that
layer, but does not write daily output. That deep layer is stored in memory and state variables in
that layer have the potential to move back into the system later due to erosion. When, due to
erosion, there are fewer than the fixed number of sediment layers, AQUATOX writes zeros for
all layers below the hardpan barrier.
Pore water moves up and down through the sediment system when layers move upward and
downward in the system. Substances dissolved in pore water also move through the system as a
result of diffusion.
214
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 6
Suspended Inorganic Sediments
As mentioned above, inorganic sediments are broken into three sets of state variables based on
particle size. Each of these three inorganic sediment types are found in the water column as well
as in each modeled sediment layer.
For inorganic sediments suspended in the water column, the derivative looks as follows:
dSuspSediment
dt
= Loading + Scour - Deposition - Washout + Washin
(246)
where:
dSuspSediment/dt
Loading
Scour
Deposition
Washout
Washin
change in concentration of suspended sediment (g/m d);
inflow loadings (excluding upstream segments) (g/m3d);
-3
scour from the active sediment layer (g/m d);
deposition to the active sediment layer (g/m3d);
"3
loss due to being carried downstream (g/m d), see (16);
loadings from upstream segments (g/m d), see (30);
There are two options for specifying deposition to and scour from the active layer when using the
multi-layer sediment option. Deposition and scour can be simulated by a hydrodynamic model
and imported into AQUATOX. In this case, for each of the three categories of suspended
sediment, deposition to and scour from the active layer are input to AQUATOX as a daily time
series in units of g/d. These inputs are converted into units of g/m3 d by dividing by the volume
of the segment.
Alternatively, based on user specification, the model can calculate deposition and scour using the
sand-silt-clay model specifications, see (230), (227). In the "Edit Sediment Layer Data" dialog,
where cohesives or non-cohesives are being input there is a checkbox that states "use sand-silt-
clay model" to toggle between these two options.
Unlike the simple sediment model, suspended sediments can sorb organic toxicants when the
multi-layer sediment model is run. More specifications about sorption of organic chemicals to
inorganic sediments can be found in Section 8.10 of this document.
Inorganics in the Sediment Bed
Inorganic sediments are found in each sediment layer that is modeled. The derivative, however
is relevant only for the active (top) layer.
dBottomSediment ^ . . 0 n n
= Deposition - Scour + Bedload - Bedloss (247)
215
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 6
where:
dBottomSediment/
Scour
Deposition
Bedload
Bedloss
change in concentration of sediment in this bed layer (g/m d);
movement to the water column (g/m d);
deposition from the water column (g/m2d);
bedload from all upstream segments (g/m d). Only relevant for
the active layer of sediment, see (247);
loss due to bedload to all downstream segments (g/m d). Only
relevant for the active layer of sediment, see (248).
Deposition and scour are input into the model in units of g/d. These inputs are divided by the
area of the system to get units of g/m2d.
Bed load is input as a loading in g/d for each link between two segments, if multiple segments
are being modeled. This process is only relevant for the top layer of sediment modeled. The
total bed load for a particular segment can be calculated by summing the loadings over all
incoming links.
BedLoad = ^
BedLoad
Upstreamlink
AvgArea
(248)
where:
BedLoad
BedLoadijpStreamlink
AvgArea
total bedload from all upstream segments (g/m d);
bedload over one of the upstream links (g/d);
average area of the segment (m );
Similarly, total bed loss is the sum of the loadings over all outgoing links:
BedLoSS \ ^^^^^^^Upstreamlink
AvgArea
(249)
BedLoss =
BedLoSS Downstfeamlink
AvgArea =
total bedloss to all downstream segments (g/m d);
bedload over one of the downstream links (g/d);
average area of the segment (m2);
As mentioned above, the derivative presented is relevant only for the active layer. Inorganic
sediments below the active layer do move up and down through the system as a result of
exposure or deposition. However, these sediments move as a part of their entire intact layer
when the active layer has reached its maximum or minimum level.
When the top layer reaches a minimum thickness, the layer below the active layer is added to the
active layer to form one new layer. The inorganic sediments within these two layers do undergo
mixing during this process.
216
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 6
Detritus in the Sediment Bed
State variables tracking sedimented labile and refractory detritus are also included in each layer
of sediment that is simulated. The equations for sedimented detritus in the active layer are the
same as those for "classic" AQUATOX.
Like inorganic sediments, buried detritus below the active layer only moves up and down in the
system when its layer moves up and down intact. Therefore, detritus found below the active
layer has a very simple derivative:
dBuriedDetritus
= -Decomp (250)
dt
where:
dBuriedDetritus/dt = change in concentration of sediment on bottom (g/m d);
Decomp = microbial decomposition in (g/m d) see (159).
Pore Waters in the Sediment Bed
Pore water quantities are also tracked in the sediment bed. The derivative for pore waters is
quite straightforward:
where:
dPoreWater „ .
= ( rain - LossTln
dt Up Up
(251)
dPoreWater/dt = change in volume of pore water in the sediment bed normalized
GainUp
Loss Up
3 2
per unit area (m /m d);
3 2
gain of pore water from the water column above (m /m d);
loss of pore water to the water column above (m3/m2 d);
In the active layer, pore waters are assumed to move into the water column when scour occurs.
To keep the bed density constant, the loss of pore waters can be solved as follows:
Loss
Up
X
•Sedim ents
(ErodeSedDensitySed) - (ErodeSed / BedDensity)
(1 / BedDensity) — 1 e- 6
(252)
where:
Loss Up
Erode sed
Density sed
BedDensity
loss of pore water to the water column above (cm / d);
scour of this sediment to the water column above, (g/d);
-3
density of this sediment (g/m );
density of the active layer (g/m3);
217
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 6
le-6
one over the density of water (m /g);
Pore waters are taken from the water column when deposition occurs. Keeping the density
constant, the gain of pore waters can be solved as follows:
Gain,
Up
I
Sediments
where:
Gain up
Deposit sed
Density sed
BedDensity
le-6
{DepositSedDensitySed) - {DepositSed / BedDensity)
(1 /BedDensity) - le-6
(253)
gain of pore water from the water column above (cm /d);
deposit of this sediment from the water column, (g/d);
-3
density of this sediment (g/m );
density of the active layer (g/m3);
one over the density of water (m /g);
When the active layer becomes too large it becomes split into two layers. During this split, the
new second layer is assumed compressed to the density of the old second layer. This
compression results in squeezing of pore water out into the water column. Details of this process
can be found in the section on sediment layer interactions, below.
Dissolved Organic Matter within Pore Waters
Another state variable tracked within the sediment bed is dissolved organic matter within pore
waters. Dissolved labile and refractory detritus within pore waters are tracked as separate state
variables. Like other dissolved detritus, these variables use units of mg/L. However, it is
important to note that these are liters of pore water and not liters in the water column.
dDOMv
dt
= GainDOMUp -LossDOMUp ±DiffDown ±DiffUp - Decomp
¦Up
'Up
(254)
where:
dI)()Mi-orciyaler dt -
GainDOMjjp
LossDOMup
Diff jP, DiffDown =
Decomp
change in concentration of DOM in pore water in the sediment bed
normalized per unit area (mg/L^ d);
active layer only: gain of DOM due to pore water gain from the water
column (ing/L7,M d);
active layer only: loss of DOM due to pore water loss to the water
column (mg/L/7Md);
diffusion over upper or lower boundary (mg/Lpwd), see (256);
microbial decomposition in (mg/Lpwd), see (159).
The increase of DOM due to pore water gain from the water column is simply the volume of
218
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 6
water that is moving from the water column above multiplied by the DOM concentration in the
above sediment layer. However, the concentration then needs to be normalized for the volume
of pore water in the current segment:
GainDOMj = ConcDOM n , • GainPWL
AvgArea- le3 N
Up I Pore Water Vol,
(255)
where:
GainDOMjjp
(mg/LpWd);
ConcDOM n_i
GainPW,
Up
AvgArea
1 e3
PoreWaterVol
= gain of DOM due to pore water gain from the layer above
= concentration of DOM in above layer (mg/LUnper water);
= gain of pore water from above (rn:\ippCr water/m d);
= average area of the segment (m );
"3
= units conversion (L/m );
= pore water volume (L);
The loss of DOM in pore water to the water column is a simpler equation due to the fact that
there are no units conversions necessary:
LossDOM J, = ConcDOM
Up f
LossPWUp '
v Pore WaterConc j
(256)
where:
LossDOMup
ConcDOM n
LossPW,
Up
= loss of DOM in pore water to the layer above (mg/Lpw • d);
PoreWaterConc =
concentration of DOM in this layer (mg/LpW);
loss of pore water to above layer (m3pw/m d);
pore water concentration (m3pw/m2);
Because diffusion and decomposition of DOM in pore water occur throughout the system, not
just the active layer, the above derivative is relevant for the whole system. DOM in pore water
also moves up and down through a system when its layer moves intact due to erosion or
deposition.
Diffusion within Pore Waters
AQUATOX calculates the diffusion of dissolved organic matter within pore waters in the
sediment layers. This calculation requires that porosity be included in the diffusion equation:
219
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 6
Diffusionv =
DiffCoeff ¦ Area ¦ AvgPor
CharLength ¦ AvgPor
Cone,,
Cone
down
Porosity uv Porosity downJ
(257)
where:
Diffusion up =
DiffCoeff =
Area =
AvgPor =
CharLength =
ConC Layer
PorOSity Layer =
gain of DOM due to diffusive transport over the upper boundary of the
sediment layer, (g/d);
dispersion coefficient, (m2 /d);
interfacial area of the upper boundary of the sediment layer (m );
average porosity of the two layers. If the boundary is a sediment/water
boundary, AvgPor is the porosity of the sediment, (fraction);
characteristic mixing length, see text below, (m);
concentration of the relevant segment, (g/m );
porosity of the relevant layer (fraction).
For the characteristic mixing length, AQUATOX uses the distance between two benthic segment
midpoints. For pore water exchange with a surface water segment, the characteristic mixing
length is taken to be the depth of the surficial benthic segment
Equation (256) is also used to calculate the diffusion of toxicants within pore waters. In this
case, the units of DiffusionUp are mg/d rather than g/d and the concentrations of toxicants
within the layers are in units of [j,g/L rather than mg/L.
Sediment Interactions
The mass of the top sediment layer increases and decreases as a result of deposition and scour.
Because the density of this layer remains constant, the volume and thickness of the top sediment
layer also increases and decreases. When the thickness of the top sediment layer reaches its
maximum, as defined by the user, the upper bed is split horizontally into two layers. The top of
these two layers maintains the same density it had before the layer was split up. It is assigned the
initial condition depth of the active layer.
The lower level is assumed to be compressed to the same density as the level below it. This
compression results in pore water being squeezed into the water column. The volume that is lost
as a result of this compression can be solved as follows:
VolumeLost =
BedMassVx eCompress - (DensityLower ¦ BedVolPieCompress
where:
VolumeLost
BedKLaSS PreCompress
BedVolpreCompress
Density iower
le6
)
le6 - Density
(258)
Lower
volume of active layer lost due to compaction (m );
mass of the new second layer before compression (g);
volume of the new second layer before compression (m );
density of the layer below the active layer (g/m3);
"3
density of water (g/m )
220
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 6
The above equation also provides the quantity of pore water squeezed into the water column
because the compression of the active layer is entirely the result of pore water being squeezed
out. Toxicants, dissolved organic matter, and toxicants associated with dissolved organic matter
in the pore water also move into the water column as a result of this compression. If there is
only one layer in the system when the splitting of the active layer takes place, Densitylower is
assumed to be the initial condition density of the second layer in the system.
The volume of a sediment layer is defined as follows:
V SedMass
BedVoln = —
BedDensity
(259)
where:
BedVoln
SedMass
BedDensity
volume of bed at layer n (m );
mass of sediment type (g);
density of bed (g/m );
The porosity of a sediment layer is defined as:
FracWater„ = 1 - "V
n SedTypes
f ConcSed
DensitySed y
(260)
where:
Frac Water „
ConCsed
Sedtypes
Density sed
= porosity of the sediment layer (fraction);
concentration of the sediment (g/m3);
all organic and inorganic sediments
density of the sediment (g/m3);
When the thickness of the top sediment layer reaches a minimum, as defined by the user, the two
top layers combine into one new active layer. The density of this new active layer is the
weighted average of the densities of the combined layers.
NewBedDensity =
VolumeLayer2-DensityLayer2 + VolumeLqyerl •DensityLqyerl
Volume Layer2 + Volume Layerl
(261)
where:
NewBedDensity
VolumeiayerN
DensityLayerN
density of new joined bed (g/m );
"3
volume of layer that was initially layer 1 or 2 (m )
density of layer that was initially layer 1 or 2 (g/m3);
The height of the new layer is the sum of the heights of the two layers being joined.
221
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 6
The bottom of the system is composed of a hardpan barrier. When this bottom is exposed, no
further erosion can take place. When deposition occurs on this hardpan bottom, it is rebuilt with
the density of the layer that existed previously. If enough deposition occurs so that two layers
are created, the new second layer is compressed to the density of the original second layer.
If a system starts with exposed hardpan as an initial condition, the user must still specify the
density of the top layer so that AQUATOX knows what density to create the top layer with. If
the user specifies a density for the second layer, this will be used when enough deposition occurs
so that two layers are created.
222
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 7
7. SEDIMENT DIAGENESIS
AQUATOX has been modified to include a representation of
the sediment bed as presented in Di Toro's Sediment Flux
Modeling (2001). This optional sediment submodel tracks the
effects of organic matter decomposition on pore-water
nutrients, and predicts the flux of nutrients from the pore
waters to the overlying water column based on this
decomposition. It is a more realistic representation of nutrient
fluxes than the "classic" AQUATOX model. It includes silica,
which will be modeled as a nutrient for diatoms in a later
version.
The model assumes a small aerobic layer (LI) above a larger
anaerobic layer (L2). For this reason, it is best to apply this
optional submodel in eutrophic sites where anaerobic
sediments are prevalent.
Because AQUATOX simulates organic matter with
stoichiometric ratios for nutrients and Di Toro's model
simulates separate organic nutrients, the organic-nutrient
relationships are redefined for the sediments. The additional 21 state variables added when the
sediment diagenesis model is enabled (and one driving variable) are as follows:
• POC (Particulate Organic Carbon) in sediment: three state variables to represent three
reactivity classes (see below). A component of the particulate organic matter (POM) that
settles from the water column into the anaerobic layer (Layer 2) and decomposes.
• PON (Particulate Organic Nitrate) in sediment: as with POC, three state variables to
represent three reactivity classes in the anaerobic layer. Another component of POM.
• POP (Particulate Organic Phosphate) in sediment: as with POC, three state variables to
represent three reactivity classes in the anaerobic layer. The third modeled component of
POM.
• Ammonia: two state variables to represent two layers. Formed by the decomposition of
PON, this process is also called the diagenesis flux. Ammonia in sediment undergoes
nitrification and flux to or from the water column.
• Nitrate: two state variables (in Layers 1 and 2). Formed by nitrification of ammonia in
the sediment bed. Undergoes denitrification and flux to or from the water column.
• Orthophosphate: two state variables (in Layers 1 and 2). Formed by the decomposition
of POP in sediment (diagenesis flux). Flux to or from the water column is predicted but
may be limited by strong P sorption to oxidated ferrous iron in the aerobic layer.
• Methane: (Layer 2) Methane is formed due to the decomposition of POC in the sediment
bed under low-salinity conditions. Methane undergoes oxidation resulting in increased
sediment oxygen demand.
• Sulfide: two state variables (in Layers 1 and 2). Hydrogen sulfide (H2S) is formed,
rather than methane under saline conditions. Sulfide in sediment may undergo burial,
flux to the water column, or oxidation (increasing SOD).
Sediment Diagenesis Model:
Simplifying Assumptions
• Model assumes a depositional
environment (no scour is modeled).
• Two layers of sediment are
modeled.
• Aerobic (top) layer is quite thin
• Model is best suited to represent
predominantly anaerobic
sediments.
• Deposition of particulate organic
matter moves directly into Layer 2.
Particulate organic matter in Layer
1 assumed to be negligible and is
not modeled
• Hie fraction of POP and PON
within defecated or sedimented
matter is assumed equal to the ratio
of phosphate or nitrate to organic
matter for given species.
• All methane is oxidized or lost.
223
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 7
Biogenic Silica: Silica in sediment is modeled using three state variables. Silica
deposited from the water column is bioavailable or "biogenic silica" and is modeled in
Layer 2. Biogenic silica can then either undergo deep burial or dissolution to dissolved
silica.
Dissolved Silica: two state variables (in Layers 1 and 2). Produced when biogenic silica
breaks down due to dissolution. Available Silica in Layer 2 and Silica in Layers 1 & 2.
Dissolved silica may undergo burial or flux to the water column.
COD: Driving variable for chemical oxygen demand in the water column that affects the
flux of sulfide to the water column.
Figure 131: Simplified schematic of the AQUATOX sediment diagenesis model
(Diagram does not include Silica, Sulfide or COD)
Water Column
Organic Matter
Aerobic
Anaerobic
Flux to
Water
fn(Oxygen)
Flux to
Water
Flux to
Water
Phosphate
Ammonia
Nitrification
G1..G3
POC
SOD
Nitrate
Mineralization
CH4
Oxidation
~SOD
PON
Mineralization
Mineralization
Ammonia
POP
Deeply Buried
Phosphate
i
Nitrate
Denitri-
fi cation
Particulate organic matter in the sediment bed (POC, PON, and POP) is divided into three
reactivity classes as follows:
• Gi - reactivity class 1, equivalent to labile organic matter
• G2 - reactivity class 2, equivalent to refractory organic matter
• G3 - reactivity class 3, nonreactive
Within the system of equations governing these state variables, sediment oxygen demand (SOD)
is a function of specific chemical reactions following the decomposition of organic matter.
224
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 7
Specifically the oxidation of methane or sulfide and the nitrification of ammonia increases the
predicted SOD . This in turn has effects on the amount of oxygen present in the water column.
The amount of oxygen in the water column, however significantly affects the nitrification of
ammonia (275).
To optimize the solution of this feedback loop, an iterative solution is utilized to calculate SOD
in each time-step, (see Eq 263) An initial value of SOD (SODj„mai) is estimated. (In the first
time-step, SODInitiai is calculated by the model based on sediment initial conditions, in later time-
steps the SODjnitiai is assumed to equal the SOD in the previous time-step.) Based on SODj„mai,
the concentrations of ammonia, nitrate, and sulfide or methane can be calculated by the model
Then, using those nutrient concentrations, a new estimate of SOD may be obtained. This
becomes the new "initial" estimate of SOD until the initial estimate and "new" estimate of SOD
converge (to within the relative error set in the AQUATOX setup screen).
This iterative solution is likely not mandatory within AQUATOX as the water column model is
not decoupled from the sediment diagenesis model (all differential equations are solved
simultaneously.) However, by including this iterative solution, the solution for SOD is not a
limiting factor when setting the variable differentiation time-step.
Most implementations of Di Toro's model solve state variables in the thin aerobic upper layer
(Layer 1) using an assumption of steady-state. This option was added to AQUATOX Release
3.1. A checkbox at the top of the diagenesis initial conditions screen can be selected for running
the model in this manner. Initial tests of the steady-state model produce results nearly identical
to non-steady-state model results and the model runs up to ten times faster.
However, precise balancing of the mass of nutrients is not generally possible when the steady-
state model is incorporated. If there are two interacting state variables and one is solved with a
steady-state solution and the other is solved using differential equations, the conservation of
mass is not possible. (For example, when solved under steady state, the nutrient mass in Layer 1
will change based on the conditions prior to the time-step but that nutrient mass is not explicitly
added to or subtracted from another state variable.)
It would be advisable to simulate a site with steady state turned on during model calibration and
off for production runs if balancing the mass of nutrients is important. When the steady-state
model is not utilized, the state variables in sediment Layer 1 are solved using differential
equations. The thickness of Layer 1 (a user input variable) might therefore have a significant
effect on model run time, with larger layer thicknesses resulting in shorter run-times.
7.1 Sediment Fluxes
State variables in the two model layers are subject to a number of fluxes to and from other
modeled and unmodeled compartments. Fluxes in the model include:
• Diffusion of the dissolved component of state variables to and from the water column;
• Diffusion of the dissolved component of the state variables between layers;
• Burial of the state variables below the lower layer and out of the modeled system; and
225
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION CHAPTER 7
• Particulate mixing of the two layers and resultant exchange of state variable.
To calculate these fluxes, the diffusion velocity between layers must be solved as well as a
particle mixing velocity between the two layers and a surface mass transfer coefficient.
Diffusion Velocity Between Layers
Diffusion between layers is specified by a diffusion coefficient, provided by the user and
adjusted for the water temperature in the system. Enhanced diffusive mixing due to bioturbation
is not currently included in the AQUATOX implementation, though direct mixing by
bioturbation is.
KL='d m
I \ /-) Temp -20
^ ' " Dd
h2 (262)
KL = diffusion velocity between layers (m/d);
Dd = diffusion coefficient for pore water (m2/d);
6>od = constant for temperature adjustment for Dd (unitless);
Temp = temperature of water (deg. C); and
H2 = depth of sediment layer 2 (m).
Particle Mixing Between Layers (Bioturbation)
In a departure from Di Toro's model, particle mixing between layers is a direct function of the
modeled benthic biomass in the system. Di Toro's formulation uses the assumption that benthic
biomass is proportional to the labile carbon in the sediment. As AQUATOX calculates benthic
biomass explicitly, this simplifying assumption is not required and a direct empirical relationship
based on benthic biomass is utilized.
j Q(Log(Benthi cJBiomass) - 2.778151
0)
le-4
where:
1,2
(263)
CO 12 ~
Benthic Biomass =
H2
le-4
particle mixing velocity between layers (m/d)
sum of benthic invertebrate biomass (g/m dry);
depth of sediment layer 2 (m); and
2 2
pore water concentration (m /cm );
226
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 7
Figure 132: Relationship derived from Di Toro, 2001, Figure 13.1A
"Diffusion coefficient for particle mixing versus benthic biomass"
Biomass (g dry/m2)
Additionally, the calculation of benthic biomass by AQUATOX includes benthic invertebrate
mortality due to low oxygen conditions and recovery when oxygen concentrations rise. Because
of this, Di Toro's benthic stress model incorporating accumulated stress and dissipation of stress
is not required nor included within AQUATOX.
Surface Mass Transfer Coefficient
Di Toro has advanced the idea that the diffusive surface mass transfer coefficient can be
successfully related to the sediment oxygen demand (Di Toro et al. 1990). The resulting
equation is as follows.
SOD
s =
Oxygen Water
(264)
SOD = CSOD +NSOD
where:
s
SOD
CSOD
NSOD
( )X\ fflater
= surface diffusive transfer (m/d)
= sediment oxygen demand (g O2 / m d);
= carbon based sediment oxygen demand (g O2 / m2 d) see (287) or
(291);
= sediment oxygen demand due to nitrification (g O2 / m d) see (275),
converted into oxygen equivalent units (1.714 g02/gN);
= overlying water oxygen conc. (g O2 / m3) (186).
227
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 7
As shown above, SOD is the sum of the carbon based sediment oxygen demand and sediment
oxygen demand due to nitrification..
7.2 POC
Particulate Organic Carbon in sediment is assumed to be located exclusively in the second layer
of sediment. Three state variables are utilized to represent three reactivity classes (Gi through
G3). POC is a component of the particulate organic matter that settles from the water column
into the anaerobic layer and decomposes; it is also subject to consumption by detritivores. In this
case, the POC uptake from that predation must be calculated separately from the POP and PON.
dPOC
Sediment
dt
where:
Deposition
Mineralization
Burial
Predation
Detr20C
= Deposition - Mineralization - Burial - (Predation / DetrlOC) (265)
-3
= deposition from water column (g C/ m d) see (266);
-3
= decomposition (g C/ m d) see (267) ;
= deep burial below modeled layer (g C/ m3 d) see (265);
"3
= predation by detritivores (g C/ m d) see (99); and
= detrital organic matter is assumed to be 1.90 • organic carbon as
derived from stoichiometry (Winberg 1971).
For all state variables burial is solved as a function of the user input burial rate w2:
w.
where:
Burial
POM
w2
H„
Burial = POM ¦
burial below modeled layer (g C/ m d); and
POP, POC, or PON (g C/ m3);
user input burial rate (m/d); and
depth of sediment layer n (m).
(266)
Burial from the top layer is added to the second layer, whereas burial from the second layer is
considered deep burial out of the modeled system.
Deposition is solved as
Deposition
POM Gi
^ Def ¦ Def2P0MOi + ^Sed- SedlPOM,
Algae&Detritus
Gi
Animals
vol.
vol.
(267)
sediment
where:
Deposition pom_a= deposition of G, reactivity class of POP, POC, or PON from water
column (g OM/ m3 d);
Def = defecation of animals, see (97) (g OM/m3water d);
Def2POMoi = fraction of POP, POC, or PON reaction class G, in defecated matter;
228
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 7
Sed
Sed2POMGl
VolwaJer
Vol sediment
sedimentation of plants or detritus, see (165), (g OM/m water d);
fraction of POP, POC, or PON reaction class G, in sedimented algae
or detritus (unitless);
water volume (m3); and
-3
sediment volume (m );
Assigning fractions of defecation to the relevant POM class (i.e., determining Def2POMoi) is a
two-part process. First, the fraction of POM, POC, or PON in the defecated material must be
determined. Second, each fraction must be again multiplied by a fraction to assign it to the three
reactivity classes (Gi to G3). In this manner, particulate organic matter is separated into nine
different state variables in the sediment.
The fractions of POP and PON within defecated matter are assumed to equal the ratios of
phosphate or nitrate to organic matter for sedimented labile detritus; these are editable
parameters ("remineralization" screen). The fraction of POC within defecated matter is set to
52.6% (Winberg 1971). Defecated matter is split evenly between reactivity classes Gi and G2.
with no defecation assigned to the non-reactive G3 class (Def2SedLabile=0.5).
Similarly, assigning fractions of sedimentation to reactivity classes is a two-part process. As
before, the fraction of POP and PON within sedimented matter is assumed equal to the ratio of
phosphate or nitrate to organic matter for the given species or detritus (editable parameters). The
fraction of POC within sedimented matter is again set to 52.6% (Winberg 1971). The amount of
refractory detritus that is converted to reactivity class G3 is a user entered parameter. The rest of
the refractory detritus is assigned to G2 and labile detritus becomes Gi. 92% of sinking plants
are assumed to be labile (Gi) with no sinking algae being converted to the non-reactive
compartment (G3).
The decomposition of organic matter is calculated as a first-order reaction with an exponential
temperature sensitivity built in:
MineralizationPOM Gi = P0MOi ¦ KPOM Gi ¦ 0POM GiTemp~20 (268)
where:
Mineralization pom Gi =
POMa
KpOM_Gi =
OpOMjGi =
Temp =
decomposition of G; reactivity class of POP, POC, or PON in
the sediment bed (g/m3 d);
-3
concentration of POM in reactivity class G, (g/m );
decay rate of POM class (1/d);
exponential temperature adjustment for decomposition of POM
class G; (unitless); and
temperature (deg.C).
Feeding on Gi is calculated based on preferences for labile detritus and feeding on G2 is based
on preferences for refractory detritus; these are set in the animal data screens. As a simplifying
assumption, there is no feeding on nonreactive G3.
229
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 7
7.3 PON
Particulate Organic Nitrogen in sediment is also assumed to be in the second layer of sediment.
Three state variables are utilized to represent three reaction classes (Gi through G3).
dPON
Sediment
dt
where:
= Deposition - Mineralization - Burial - Predation ¦ N20rg
Deposition =
Mineralization =
Burial =
Predation =
N20rg =
(269)
deposition from water column (g N/ m d) see (266);
decomposition to ammonia (g N/ m3 d) see (267) ;
"3
deep burial below modeled layer (g N/ m d) see (265);
predation by detritivores (g N/ m3 d) see (99); and
user input conversion factor between N and refractory or labile
detritus (g N / g OC).
7.4 POP
Particulate Organic Phosphate in sediment is solved in a very similar manner to POC and PON.
Mineralization rates may be different, however.
dP()I\
Sediment
dt
where:
Deposition
Mineralization
Burial
Predation
P20rg
= Deposition - Mineralization - Burial - Predation ¦ P20rg
(270)
deposition from water column (g P/ m d) see (266);
decomposition to orthophosphate (g P/ m3 d) see (267) ;
-3
deep burial below modeled layer (g P/ m d) see (265);
predation by detritivores (g P/ m3 d) see (99); and
user input conversion factor between P and refractory or labile
detritus (g P / g OC).
7.5 Ammonia
Ammonia in the sediment is solved using two state variables to represent the two layers.
Ammonia is formed by the decomposition of PON. Ammonia in sediment undergoes
nitrification, burial, and flux to or from the water column. The ammonia in each state variable is
the sum of dissolved and particulate ammonia. The fraction that is dissolved is solved below in
equation (274). The ammonia differential equations are as follows:
dAmmonia
dt
L2 Sed
= Diag _ Flux - Burial + Flux2 Anaerobic
(271)
230
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 7
dAmmonia
XI Sed
dt
= -Nitrification - Burial - FluxlWater - Flux2Anaerobic
(272)
where:
DiagFlux
Burial
Flux2Anaerobic
Flux2Water
Nitrification
= decomposition of PON, see (267) ;
"3
= burial below relevant layer (g N/ m d) see (265);
= flux to layer 2 from layer 1 (g N/ m3 d, may be negative) see (272) ;
"3
= flux to water from layer 1 (gN/m d, may be negative); see (273);
= conversion to nitrate (g N/ m3 d) see (275);
Flux2 Anaerobic = -
%2 (
fp2Conc2 ~fp\Conc
l)+KL(fc
"d2Conc2 -fd\Conc\)]/HLayer
l)]A
(273)
where:
CO 1,2
KL
fp,layer
fd,layer
ConC layer
-^layer
particle mixing velocity between layers (m/d), see (262);
diffusion velocity between layers (m/d), see (261);
particulate fraction in layer 1 or 2 (unitless); see (274)
dissolved fraction in layer 1 or 2 (unitless); see (274)
"3
total concentration of state variable in layer (g/m ); and
depth of layer being evaluated (m);
FluxlWater = s{fd
(274)
where:
5 = surface diffusive transfer (m/d); (263)
fdi = dissolved fraction in layer 1;
"3
Cone layer = total concentration of state variable in layer (g/m ); and
Hi = depth of layer 1 (m);
The fraction of ammonia that is dissolved in each layer is calculated as follows:
/;
d ammonia, layer
1 + m layer ' NH 4
A
p ammonia, layer
= 1-/,
d ammonia, layer
(275)
where:
fd ammonia,layer
W layer
KdNH4
fp ammonia,layer
dissolved fraction in layer;
user-input solids concentration in layer (kg/L);
editable partition coefficient for ammonium (L/kg); and
particulate fraction in layer.
231
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 7
Ammonia in the top layer is converted to nitrate in the presence of oxygen, resulting in sediment
oxygen demand. Since the nitrification reaction requires oxygen, no nitrification is assumed to
occur in the lower anaerobic layer. Nitrification in the aerobic layer is calculated as follows:
Nitrification =
DOu
V
\2-KM02 +DOwc j
KM,
^KMNH4+NH 4!
^.2 QTemp-20
( NH4j ^
(276)
where:
Nitrification
DOwc.
KMNH4
KM02
s
NH4i
Hi
e
Temp
conversion of ammonia to nitrate (g N/m d);
dissolved oxygen in the water column (g/m );
user-input nitrification half-saturation coefficient for ammonium
(g N/m3);
user-input nitrification half-saturation coefficient for oxygen
(g 02/m3);
reaction velocity for nitrification (m/d); (user-input, differentiating
between fresh and salt water)
surface diffusive transfer (m/d); (263)
concentration of ammonia in layer 1 (g/m3); (168)
user-input depth of layer 1 (m);
user-input exponential temperature adjustment for nitrification
(unitless); and
temperature (deg.C).
7.6 Nitrate
Nitrate is formed by the nitrification of ammonia in the top layer of the sediment bed. Nitrate in
sediment undergoes denitrification, burial and flux to or from the water column.
dNitrate L2Sed
dt
= -Burial - Denitr + Flux2Anaerobic (277)
dNitrate LlSed
— = Nitrification - Denitr - Burial - FluxlWater - Flux! Anaerobic (278)
"3
Burial = burial to layer below modeled layer or out of the system(g N/ m d)
where:
Burial =
see (265);
232
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 7
-3
Flux2Anaerobic = flux to layer 2 from layer 1 (g N/ m d, may be negative) see (272) ;
Flux2Water = flux to water from layer 1 (g N/ m3 d, may be negative); see (273);
"3
Nitrification = conversion of ammonia to nitrate (g N/ m d), see (275);
Denitr = denitrification of nitrate to free nitrogen (g N/ m3 d), see (278);
Nitrate is assumed to be dissolved in the sediment bed so fd = 1.0 and fp = 0.0.
Denitrification is solved as follows
„:-emr'-"(NO 3 ^
DCflitV layer, NO3 NO3
layer
TT
^ layer J
(279)
where:
k layer, No3 = user-input reaction velocity for denitrification given the layer and
salinity regime (m/d);
6 = user-input exponential temperature adjustment for denitrification
(unitless); and
5 = surface diffusive transfer (m/d); (263)
H layer = depth of layer (m);
NO3iayer = concentration of nitrate in layer (g/m3); and
Temp = temperature (deg.C).
7.7 Orthophosphate
Phosphate in the sediment is solved using two state variables to represent the two layers. Like
ammonia, the phosphate in each state variable represents the sum of dissolved and particulate
phosphate.
dP04
L2Sed
dt
= Diag _ Flux - Burial + Flux2 Anaerobic
(280)
dP04
LISed
dt
= -Burial - FluxlWater - Flux2Anaerobic
(281)
where:
Diag Flux
Burial
decomposition of POP, see (267) ;
burial to layer below modeled layer or out of the system(g P/ m d)
see (265);
Flux2Anaerobic = flux to layer 2 from layer 1 (g P/ m d, may be negative) see (272) ;
Flux2Water
= flux to water from layer 1 (g P/ m d, may be negative); see (273);
When oxygen is present in the water column, the diffusion of phosphorus from sediment pore
233
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 7
waters is limited. This is due to strong P sorption to oxidated ferrous iron in the aerobic layer
(iron oxyhydroxide precipitate). Under conditions of anoxia, phosphorus flux from sediments
increases significantly.
Di Toro incorporates the effect of oxygen on phosphate flux into his model by making the
dissolved fraction of phosphate a function of oxygen in the water column. When the oxygen in
water decreases below a critical threshold the partition coefficient for phosphate is increased by a
user-entered factor. As the oxygen goes to zero, the partition coefficient is smoothly reduced to
the anaerobic coefficient using an exponential function:
if DOwc>DOCnt
,P04
then KdPQ4, = KdPQ4 2AKdPQ4,
(282)
DOw
else KdPQ4l = Kdp042AKdp04lDo
Crit,P04
Partitioning of phosphate between the dissolved and particulate forms will affect on the flux of
phosphate to the water column (273).
1
' d phosphate, layer
1 + miayer ' Kdpo 4 layer
(283)
where:
fd phosphate,layer
W layer
Kdpo4,2
kKdpo4,i
DOwc
DO Crit,P04.
dissolved fraction in layer (unitless);
user-input solids concentration in layer (kg/L); and
partition coefficient for phosphate in layer 2 (L/kg);
fresh or saltwater factor to increase the aerobic (Li) partition
coefficient of PO4 relative to the anaerobic (L2) coeff (unitless);
"3
dissolved oxygen in the water column (g/m ), see (186); and
critical oxygen concentration for adjustment of partition coefficient
for inorganic P (g/m );
7.8 Methane
Methane is formed due to the decomposition of POC in the sediment bed under low-salinity
conditions. Methane undergoes oxidation resulting in increased sediment oxygen demand.
dMethane L2Sd
— = Diag _ FluxMethane - Flux2WaterMethane - OxidationMethane (284)
dt
where:
Methane used = methane in the anaerobic layer expressed in oxygen equivalence
units (g 02equiv / m3)
Diag Flux = decomposition of POC in freshwater, adjusted for the organic carbon
234
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 7
"3
lost due to denitrification (g 02equiv / m d) see (284);
Flux2Water = methane flux to water (g 02equiv / m3 d), see (288); and
Oxidation = oxidation of methane (CSOD) (g 02equiv / m3 d) see (287);
In the manner of Di Toro, methane and sulfide are tracked in units of oxygen equivalents (g
02equiv / m ) to easily balance the model's computations.
In fresh water conditions, decomposing POC is converted to methane which is tracked in oxygen
equivalents. In salt water, decomposing POC becomes sulfide. However, some POC is lost due
to denitrification and does not decompose:
Diag_FluxMethane Sulfide = Mineralizationpocj - 2.86• Denitrification (285)
where:
DiagCFlllXMethane, Sulfide
= decomposition of POC in water, adjusted for the organic carbon lost
"3
due to denitrification (g 02equiv / m d);
Mineralizationpoc = decomposition of POC in freshwater, (g POC / m3 d) see (267) ;
"3
Denitrification = denitrification of nitrate, (g N/ m d) see (278);
32/12 = conversion between POC and oxygen equivalents; and
2.86 = conversion between Nitrate and oxygen equivalents;
Oxidation of methane is solved as a function of the saturation concentration of methane in pore
water.
CH 4 sat = 100^1 + ^^j 1.02420^ (286)
CSODMax = mm (^2KL ¦ CH 4 „, • Diag _ FluxMethane, Diag _ FluxMethane) (287)
CSODMax
^ f r\ Temp-20
1-sech ^*CH4 ' CH4
OxidationMethane = (288)
2
where:
CH4Sat
Zmean
Temp
CSOD Max
KL
Diag_FIUXMethane ~
= saturation concentration of methane in pore water (g 02equiv / m );
= mean depth of water column above the sediment bed (m);
= temperature (deg.C);
= maximum oxidation flux (g 02equiv / m d);
= diffusion velocity between layers (m/d); (261)
diagenesis flux of methane to water column, adjusted to be in units
235
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 7
OxidatioriMethane
sech
s
K-CH4
6ch4
h2
of (§ 02equiv / m d),
oxidation of methane (g 02equiv / m3 d);
hyperbolic secant function
surface diffusive transfer (m/d); (263)
reaction velocity for methane oxidation(m/d);
exp. temperature adjustment for methane oxidation (unitless); and
depth of layer 2 (m); (methane mass arbitrarily tracked on the second layer)
All methane is assumed to be oxidized or to escape from the sediment to water. Thus the
derivative for methane will remain at zero and the solution for the flux to water can be solved as
follows:
Flux2WaterMetham = Diag_FluxMethane - OxidationMethane
(289)
where:
Diag Flux
Oxidation
= decomposition of POC in freshwater, adjusted for the organic carbon
"3
lost due to denitrification (g 02equiv / m d), see (284);
= oxidation of methane (g 02equiv / m3 d), see (287);
7.9 Sulfide
Sulfide is formed, rather than methane, under saline conditions. Sulfide in sediment may
undergo burial, flux to the water column, or oxidation, which increases SOD.
dSulfide
L2Sed
dt
= Diag _ FluxSulflde - Burial + Flux2 Anaerobic
(290)
dSulfide
L\ Sed
dt
= -Oxidation - Burial - FluxlWater - Flux2Anaerobic
(291)
where:
Sulfide $ec{
DiagFlux suifick
Burial
Flux2Anaerobic
Flux2Water
Oxidation
= sulfide concentration in layer n of sediment, (g 02equiv / m );
= decomposition of POC in salt water, adjusted for the organic carbon
lost due to denitrification (g 02equiv / m d), see (284);
= burial to layer below modeled layer or out of the system (g 02equiv /
m3 d); see (265);
= flux to layer 2 from layer 1 (g 02equiv / m3 d, may be neg.) see (272) ;
= flux to water from Li (g 02equiv / m3 d, may be neg.) (Note the
driving var. "COD" represents the water col. conc. of sulfide.) see
(273);
= oxidation of sulfide in the active layer;
236
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 7
a:
H2S,d
fdl+K.
2
H2S,p
•/>
Temp-20
H2S
OxidationSulfide = ConcH2SLl ¦
DOu
\2KMH2SDO j
s-Hi
(292)
where:
Oxidation sulfide
Concms.u
KH2S,d
Kms.p
DOwc.
KMh2S,DP
0H2S
s
Hi
oxidation of sulfide (g 02equiv / m d);
concentration of sulfide in layer 1 (g 02equiv / m3);
reaction velocity for dissolved sulfide oxidation (m/d);
reaction velocity for particulate sulfide oxidation (m/d);
dissolved oxygen in the water column (g/m3);
sulfide oxidation normalization constant for oxygen (g 02/m );
exp. temperature adjustment for sulfide oxidation (unitless);
surface diffusive transfer (m/d); and
depth of layer 1 (m);
The fraction of sulfide that is dissolved in each layer is calculated as follows:
1
fc
d sulfide, layer
1 + miayer ' ^^H2S, Layer
(293)
where:
fd sulfide,layer
Wl layer
KdNH4
= dissolved fraction in layer;
= solids concentration in layer (kg/L); and
= partition coefficient for sulfide for layer (L/kg);
The particulate fraction of sulfide in each layer is calculated as one minus the dissolved fraction.
7.10 Biogenic Silica
Silica in sediment is modeled using three state variables. Silica associated with diatoms and
deposited from the water column is biogenic silica and is modeled in Layer 2. Biogenic silica
can then either undergo deep burial or dissolution to dissolved silica.
dBiogenic Silica, ,
= Deposition - Dissolution - Burial (294)
dt
where:
¦j
Deposition = deposition from water column (g Si/ m d) see (294);
"3
Dissolution = dissolution of biogenic silica (g Si/m d)
Burial = deep burial below modeled layer (g Si/ m3 d) see (265); and
237
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 7
Deposition of silica is a function of the sinking of diatoms:
DepositionSi = f ^ Sed ¦ FracSilica vo^water
\Diatoms J V®^sediment
(295)
where:
Depositionsi
FracSilica
Sed
Volwafer
Vol sediment
= deposition of silica from water column (g Si/ m d);
= user-input fraction of silica in diatoms, (unitless);
"3
= sedimentation of diatoms, see (165), (g OM/m water d);
= water volume (m3); and
-3
= sediment volume (m );
Biogenic silica can undergo dissolution to dissolved silica. This reaction can also operate in
reverse:
/
Dissolution = KSi0Si
Temp-20
Cone
\
Avail Si
yConcAvailSl +KM
PSi j
(^5'at fd,silica, L2 ¦Concsmca,L2) (296)
where:
Dissolution
KSi
Osi
ConC yaj\layer
KMPSl
Si Sat
fd silica,layer
dissolution of biogenic silica (g Si/ m d);
user-input reaction velocity for dissolved silica dissolution (1/d);
user-input exponential temperature adjustment for silica dissolution
(unitless);
"3
concentration of available silica or silica in layer 2 (g Si/ m );
user input silica dissolution half-saturation constant for biogenic
"3
silica (g Si/m );
user-input saturation concentration of silica in pore water (g Si/m3);
dissolved fraction of silica in layer.
7.11 Dissolved Silica
Dissolved silica is produced when biogenic silica breaks down due to dissolution, and could
potentially be modeled as a limiting nutrient for diatoms in a later version of AQUATOX.
Dissolved silica (referred to hereafter as "silica") is modeled in two layers:
dSilica L2Sed
= Dissolution - Burial + Flux! Anaerobic (297)
dt
dSilica n,
= -Burial - FluxlWater - FluxlAnaerobic (298)
dt
where:
238
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 7
"3
Dissolution = dissolution of biogenic silica (g Si / m d), see (295);
Burial = burial to layer below modeled layer or out of the system (g Si / m3
d); see (265);
Flux2Anaerobic = flux to layer 2 from layer 1 (g Si/ m3 d, may be negative) see (272) ;
Flux2Water = flux to water from layer 1 (g Si/ m d, may be negative); see (273);
Similar to inorganic phosphate, dissolved oxygen causes a barrier to silica flux to the water
column. This is modeled by increasing the partition coefficient by a factor when the dissolved
oxygen decreases below a critical threshold.
if DOwc > DOCrit Si then KdSi l KdSi 2AKdSi j
else KdSi x = KdSi 2AKdSi 1 DoCrit Si
(299)
J"d Si .layer , , - , (300)
l + miayer-KdSUayer
where:
fdsilica,.layer = dissolved fraction in layer (unitless);
niiayer = solids concentration in layer (kg/L); and
Kdsi,2 = partition coefficient for silica in layer 2 (L/kg);
AKdsij = fresh or saltwater factor to increase the aerobic (Li) partition
coefficient of silica relative to the anaerobic (L2) coeff (unitless);
DO wc ~ dissolved oxygen in the water column (g/m3); and
DOCrit,si. = critical oxygen concentration for adjustment of partition coefficient
for silica (g/m3);
239
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION CHAPTER 8
8. TOXIC ORGANIC CHEMICALS
The chemical fate module of AQUATOX predicts the
partitioning of a compound between water, sediment, and
biota (Figure 133), and estimates the rate of degradation of the
compound (Figure 134). Microbial degradation,
biotransformation, photolysis, hydrolysis, and volatilization
are modeled in AQUATOX. Each of these processes is
described generally, and again in more detail below.
Nonequilibrium concentrations, as represented by kinetic
equations, depend on sorption, desorption, and elimination as
functions of the chemical, and exposure through water and
food as a function of bioenergetics of the organism.
Equilibrium partitioning is no longer represented in
AQUATOX except as a constraint on sorption to detritus and
plants and as a basis for computing internal toxicity.
Partitioning to inorganic sediments is not modeled unless the
multi-layer sediment model is included.
Microbial degradation is modeled by entering a maximum biodegradation rate for a particular
organic toxicant, which is subsequently reduced to account for suboptimal temperature, pH, and
dissolved oxygen. Biotransformation is represented by user-supplied first-order rate constants
with the option of also modeling multiple daughter products. Photolysis is modeled by using a
light screening factor (Schwarzenbach et al., 1993) and the near-surface, direct photolysis first-
order rate constant for each pollutant. The light screening factor is a function of both the diffuse
attenuation coefficient near the surface and the average diffuse attenuation coefficient for the
whole water column. For those organic chemicals that undergo hydrolysis, neutral, acid-, and
base-catalyzed reaction rates are entered into AQUATOX as applicable. Volatilization is
modeled using a stagnant two-film model, with the air and water transfer velocities approximated
by empirical equations based on reaeration of oxygen (Schwarzenbach et al., 1993).
Toxic Organic Chemicals:
Simplifying Assumptions
• Kinetic model of toxicant fate
• Photolysis in sediments is not
included
• A generalized equation is used to
calculate partitioning of polar
compounds
• Direct sorption onto the body of an
animal is ignored
• The exchange of toxicant through
the gill membrane is assumed to be
facilitated by the same mechanism
as the uptake of oxygen
• Estimation of the elimination rate
constant k2 may be made based on
logKow with two alternative
formulations available
• Biotransformation occurs at a
constant rate throughout a
simulation
240
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER:
Figure 133. In-situ uptake and release of chlorpyrifos in a pond, dominated by plants
CHLORPYRIFOS 6 ug/L (PERTURBED)
Run on 05-2-08 4:33 PM
!-> 215
6/27/1986 7/12/1986 7/27/1986 8/11/1986 8/26/1986 9/10/1986
»T1 H20 GillSorption (Percent)
»T1 H20 Depuration (Percent)
¦T1 H20 DetrSorpt (Percent)
»T1 H20 Decomp (Percent)
T1 H20 DetrDesorpt (Percent)
T1 H2Q PlantSorp (Percent)
Figure 134. In-situ degradation rates for chlorpyrifos in pond
-0.5
4->
c
(LI
y -1.0
-1.5
-2.0
-2.5
-3.0
CHLORPYRIFOS 6 ug/L (PERTURBED)
Run on 05-2-08 4:33 PM
f
—
—
—
—
—
T1 H20 Hydrolysis (Percent)
T1 H20 Photolysis (Percent)
T1 H20 MicroMet (Percent)
T1 H2Q Volatil (Percent)
6/27/1986 7/12/1986 7/27/1986 8/11/1986 8/26/1986 9/10/1986
241
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
The mass balance equations follow. The change in mass of toxicant in the water includes
explicit representations of mobilization of the toxicant from sediment to water as a result of
decomposition of the labile sediment detritus compartment, sorption to and desorption from the
detrital sediment compartments, uptake by algae and macrophytes, uptake across the gills of
animals, depuration by organisms, and turbulent diffusion between epilimnion and hypolimnion:
d Toxicantwater = Loading + ^LMeOetr (DeCOmpOSitWTl LMeDetr • PPBLabileDetr ' le-6)
at
+ £ De sorption DetrTox + £ Depuration0rg- £ SorptionSedTox
- Y^GillUptake -MacroUptake- £ AlgalUptakeAlga (301)
- Hydrolysis - Photolysis - MicrobialDegrdn + Volatilization
-Discharge + BiotransformMicmbIn ± TurbDiff ±DiffusionSeg
± PorewaterAdvection ± DiffusionSedimenf Washout + Washin
The equations for the toxicant associated with the two sediment detritus compartments are rather
involved, involving direct processes such as sorption and indirect conversions such as defecation.
However, photolysis is not included based on the assumption that it is not a significant process
for detrital sediments:
d ToxicantSedLabileDetr = ^()rpjj()n _ DeSOipHol) + (Colonization ¦ PPBSedRefrDetr ' 1 e - 6)
+ £/¦«,/ £/¦«„ (l)ef2SedI,abile ¦ DefecationToxPred Pre)
- (Resuspension + Scour + Decomposition) ¦ PPBsediabUeDetr ' 1 e - 6
^Pred^nS^S^°npred SedLabileDetr ' PPB SedLabileDetr ' lC"6 (302)
+ Sedimentation ¦ PPBsuspLab ileDetr 1 C ~ 6
+ T(Sed2I)etr ¦ SinkPhyto• PPBPhyto• le-6)
- Hydrolysis - MicrobialDegrdn - Burial + Expose
± BiotransformMicrobial
d ToXiCClTitSedRefrDetr ry ,• ,•
— = Sorption - Desorption
+ Y.Predl.Preyk1 -Def2SedLabile) ¦ DefecationToxPredPrJ
- {Resuspension + Scour + Colonization) ¦ PPB SedRefrDetr ' le-6
Hpred^nSeStiOnpred SedRefrDetr ' PPB SedRefrDetr ' lC"6 (303)
+ {Sedimentation + Scour) ¦ PPBsuspRefrDetr • le-6
+ Y.{Sed2I)etr ¦ SinkPhyto • PPBPhyto • le-6)
- Hydrolysis - MicrobialDegrdn - Burial + Expose
± BiotransformMicmbial
242
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
Similarly for the toxicant associated with suspended and dissolved detritus, the equations are:
d ToXi CClTlt SuspLabileDetr j / ¦ o - • I \ . • if' / ¦
= Loading + Sorption - Desorption + WashinToxCarrier
dt
+ Xorg (fMort2Detr ¦ MortalityQrg + GameteLoss0rg ) ¦ PPB0rg ¦ 1 e-6)
- (Sedimentation + Deposition + Washout + Decomposition
"^liPred
Ingestion Pre(i
SuspLabileDetr
) ' PPBSuspLabileDetr ' 1 e -6 (304)
+ Colonization ¦ PPBsusVRefrDetr ¦ 1 e-6 ± BiotransformMjcrobjal
+ (Resuspension + Scour) ¦ PPBsedLMieDetr • 1 e-6 ± SedToHyp
- Hydrolysis - Photolysis - MicrobialDegrdn ± TurbDiff ± DiffusionSeg
^ Toxicant SuspRefrDetr / I ¦ ry ¦ > \ ..
——— = Loading + Sorption - Desorption
+ iLtj JMorQRef ¦ MortalityQrg ¦ PPB0rg • 1 e-6)
- (Sedimentation + Deposition + Washout + Colonization
± BiotransformMicmbial + £Pred Ingestion SuspRefrDetr) • PPBSuspRefrDetr • 1 e -6 (305)
+ (Resuspension + Scour) ¦ PPBsedRefrDetr ¦ 1 e-6
± SedToHyp - Hydrolysis - Photolysis - MicrobialDegrdn
± TurbDiff ± DiffusionSeg + WashinToxCarrier
—^ox^can^DlssLaMeDetr = Loading + Sorption - Desorption + SumExcrToxToDissorg
C11
+ 'Lorg(Mort2Detr ¦ MortalityQrg ¦ PPB0rg ¦ 1 e-6)
- (Washout + Decomposition) ¦ PPBDissLatueDetr ¦ 1 e-6 (306)
± BiotransformMjcmbjal - Hydrolysis - Photolysis
- MicrobialDegrdn ± TurbDiff ± DiffusionSeg + WashinToxCarrier
±PorewaterAdvection ± Diffusion
Sediment
dToxicantDissRefrDetr = Loadjng + Sorption - Desorption + SumExcToxToDiss0rg
dt
+ lLor/Moi'l2Ref' MortalityQrg ¦ PPB0rg ¦ 1 e-6)
- (Washout + Colonization) ¦ PPB DissRefrDetr ¦ 1 e -6 (307)
± BiotransformMjcmbjal - Hydrolysis - Photolysis
- MicrobialDegrdn ± TurbDiff ± DiffusionSeg + WashinToxCarrier
±PorewaterAdvection± Diffusion
Sediment
243
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
When the simple sediment model is run, there are no equations for buried detritus, as they are
considered to be sequestered and outside of the influence of any processes which would change
the concentrations of their associated toxicants. When the multi-layer sediment model is
included, equations for toxicants in pore waters and toxicants in buried sediments may be found
in sections 8.10 and 8.11.
Toxicants associated with algae are represented as:
—^()X'canl u"'a = Loacnng + AlgalUptake - Depuration ± TurbDiff ± Diffusion s
dt
+ WashinToxCamer - (Excretion + Washout + TpredPredationPred,Aiga+ Mortality (308)
+ Sink ± SinkToHypo ± Floating) ¦ PPBAiga • 1 e -6 ± BiotransformAlga
Macrophytes are represented similarly, but reflecting the fact that they are stationary unless
specified as free-floating:
^ ToXICantMacrophyte i /• i / j - i j > . /1 ¦ .
— = Loading + Macro Uptake - Depuration - (Excretion
+ Tpred Predationpred, Macro + Mortality + WashoutFreeFloating + Breakage) (309)
• PPBMacro • 1 e -6 ± BiotransformMacrophyte + WashinToxCamerFreeFloat
The toxicant associated with animals is represented by an involved kinetic equation because of
the various routes of exposure and transfer:
d Toxicant. Loading + (jmuptake + Y.rr,yD„,Urul, + TurbDiff
- (Depuration + TpredPredationPred,Ammai + Mortality + Spawn (310)
± Promotion + Drift + Migration + Emergelnsect) ¦ PPBAmmai • 1 e -6
± BiotransformAmmal + WashinToxCarmr
where:
Toxi can t Water
Toxicant SedDetr
Toxi Cant SuspDetr
ToxicantDissDetr
ToxicantAiga
Toxicant Macrophyte
Toxicant Animal
PPPsedDetr
244
toxicant in dissolved phase in unit volume of water ((J,g/L);
mass of toxicant associated with each of the two sediment
detritus compartments in unit volume of water ((J,g/L);
mass of toxicant associated with each of the two suspended
detritus compartments in unit volume of water ((J,g/L);
mass of toxicant associated with each of the two dissolved
organic compartments in unit volume of water ((J,g/L);
mass of toxicant associated with given alga in unit volume
of water ((J,g/L);
mass of toxicant associated with macrophyte in unit
volume of water([j,g/L);
mass of toxicant associated with given animal in unit
volume of water ((J,g/L);
concentration of toxicant in sediment detritus ([j,g/kg), see
(310);
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION CHAPTER 8
PPB SuspDetr
PPB' DissDetr
PPBAlga
PPBMacrophyte
PPBAnimal
1 e-6
Loading
TurbDiff
Washin
Wash in ToxCarrier
Diffusionseg
DiffilSionSediment
P or ew aterMiction
Hydrolysis
Biotransform Microbial
Biotransform org
Photolysis
MicrobialDegrdn
Volatilization
Discharge
Burial
Expose
Decomposition
Depuration
Sorption
Desorption
Colonization
concentration of toxicant in suspended detritus ([j,g/kg);
concentration of toxicant in dissolved organics ([j,g/kg);
concentration of toxicant in given alga ([j,g/kg);
concentration of toxicant in macrophyte ([j,g/kg);
concentration of toxicant in given animal ([j,g/kg);
units conversion (kg/mg);
loading of toxicant from external sources ([j,g/L-d);
depth-averaged turbulent diffusion between epilimnion and
hypolimnion (^g/L-d), see (22) and (23).
"3
loadings from linked upstream segments (g/m d), see (30);
inflow load of toxicant sorbed to a carrier from an upstream
segment (ng/L d), see (31);
gain or loss due to diffusive transport over the feedback
link between two segments, (ng/L-d), see (32);
gain or loss due to diffusive transport to porewaters in the
sediment (ng/L-d), see (256);
gain or loss of toxicant to porewater due to scour or
deposition of sediment (|j,g/LpW-d), see (394), (395);
rate of loss due to hydrolysis ([j,g/L-d), see (313);
biotransformation to or from given organic chemical in
given detrital compartment due to microbial
decomposition ([j.g/L-d), see (375);
biotransformation to or from given organic chemical within
the given organism ([j,g/L-d); (375)
rate of loss due to direct photolysis ([j.g/L-d), see (320);
assumed not to be significant for bottom sediments;
rate of loss due to microbial degradation ([j,g/L-d), see
(326);
rate of loss due to volatilization ([j,g/L-d), see (331);
rate of loss of toxicant due to discharge downstream
([j,g/L-d), see Table 3;
rate of loss due to deposition and resultant deep burial
([j,g/L-d) see (167b);
rate of exposure due to resuspension of overlying sediments
(ng/L-d), see (227);
rate of decomposition of given detritus (mg/L-d), see (159);
elimination rate for toxicant due to clearance ([j,g/L-d), see
(362), (363), and (372);
rate of sorption to given organic or inorganic compartment
(ng/L-d), see (350);
rate of desorption from given organic or inorganic
compartment ([j,g/L-d), see (351);
rate of conversion of refractory to labile detritus (g/m -d),
see (155);
245
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
DefecationToxpred, prez
Def2SedLabile
Resuspension
Scour
Sedimentation
Deposition
Sed2Detr
Sink
Breakage
Mortalityorg
Mort2Detr
GameteLoss
Mort2Ref
Washout or Drift
SedToHyp
Ingestionpred, Prey
Predationpred, Prey
ExcToxToDissorg
Excretion
SinkToHypo
AlgalUptake
MacroUptake
GillUptake
rate of transfer of toxicant due to defecation of given prey
by given predator ([j,g/L-d), see (379);
fraction of defecation that goes to sediment labile detritus,
= 0.5;
rate of resuspension of given sediment detritus (mg/L-d)
without the inorganic sediment model attached, see (165);
rate of resuspension of given sediment detritus (mg/L-d); in
streams with the inorganic sediment model attached, see
(233);
rate of sedimentation of given suspended detritus (mg/L-d);
without the inorganic sediment model attached, see (165);
rate of sedimentation of given suspended detritus (mg/L-d)
in streams with the inorganic sediment model attached, see
(235);
fraction of sinking phytoplankton that goes to given detrital
compartment;
loss rate of phytoplankton to bottom sediments (mg/L-d),
see (69);
loss of macrophytes due to breakage (g/m -d), see (88);
nonpredatory mortality of given organism (mg/L-d), see
(66), (87), and (112);
fraction of dead organism that is labile (unitless);
"3
loss rate for gametes (g/m -d), see (126);
fraction of dead organism that is refractory (unitless);
rate of loss of given toxicant, suspended detritus or
organism due to being carried downstream (mg/L-d), see
(16), (71), (72), (130), and (131);
rate of settling loss to hypolimnion from epilimnion
(mg/L-d). May be positive or negative depending on
segment being simulated, see (69);
rate of ingestion of given food or prey by given predator
(mg/L-d), see (91);
predatory mortality by given predator on given prey
(mg/L-d), see (99);
toxicant excretion from plants to dissolved organics
(mg/L-d);
"3
excretion rate for given organism (g/m -d), see (64), (111);
rate of transfer of phytoplankton to hypolimnion (mg/L-d).
May be positive or negative depending on segment being
modeled, see (69);
rate of sorption by algae ((J,g/L - d), see (360);
rate of sorption by macrophytes ((J,g/L - d), see (356);
rate of absorption of toxicant by the gills ((J,g/L - d), see
(365);
246
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
DietUptakeprey = rate of dietary absorption of toxicant associated with given
prey ([j,g/L-d), see (369);
Recruit = biomass gained from successful spawning (g/m -d), see
(128);
Promotion = promotion from one age class to the next (mg/L-d), see
(136);
Migration = rate of migration (g/m -d), see (133); and
Emergelnsect = insect emergence (mg/L-d), see (137).
The concentration in each carrier is given by:
PPB, = ToxState'—7e6 (311)
Carrier State i
where:
PPBi = concentration of chemical in carrier i ([j,g/kg);
ToxStatei = mass of chemical in carrier i ((J,g/L);
CarrierState = biomass of carrier (mg/L); and
le6 = conversion factor (mg/kg).
8.1 Ionization
Dissociation of an organic acid or base in water can have a significant effect on its environmental
properties. In particular, solubility, volatilization, photolysis, sorption, and bioconcentration of
an ionized compound can be affected. Rather than modeling ionization products, the approach
taken in AQUATOX is to represent the modifications to the fate and transport of the neutral
species, based on the fraction that is not dissociated. The acid dissociation constant is a measure
of the strength of the acid or base, and is expressed as the negative log, pKa, and the fraction that
is not ionized is:
Nondissoc = (312)
-^ + 70
where:
Nondissoc = nondissociated fraction (unitless).
If the compound is a base then the fraction not ionized is:
Nondissoc = (313)
1 + 10
Note: If pKa is set to zero then ionization is ignored (i.e. NonDissoc is set to 1.0).
When pKa = pH half the compound is ionized and half is not (Figure 132). At ambient
environmental pH values, compounds with a pKa in the range of 4 to 9 will exhibit significant
dissociation (Figure 133).
247
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION CHAPTER :
Figure 135. Dissociation of pentachlorophenol Figure 136. Dissociation as a function of pKa at
(pKa = 4.75) at higher ph values an ambient pH of 7
1
0.8
0.6
0.4
0.2
0
2
4
6
8
10
/
/
/
/
/
/
/
y
0 2 4 6 8 10 12 14
pka
8.2 Hydrolysis
Hydrolysis is the degradation of a compound through reaction with water. During hydrolysis,
both a pollutant molecule and a water molecule are split, and the two water molecule fragments
(H+ and OH") join to the two pollutant fragments to form new chemicals. Neutral and acid- and
base-catalyzed hydrolysis are modeled using the approach of Mabey and Mill (1978) in which an
overall pseudo-first-order rate constant is computed for a given pH, adjusted for the ambient
temperature of the water:
Hydrolysis = KHyd ¦ Toxicant Phase (314)
where:
and where:
KHyd
KAcidExp
KBaseExp
KUncat
Arrhen
KHyd = (KAcidExp + KBaseExp + KUncat) ¦ Arrhen (315)
= overall pseudo-first-order rate constant for a given pH and
temperature (1/d);
= pseudo-first-order acid-catalyzed rate constant for a given pH
(1/d);
= pseudo-first-order base-catalyzed rate constant for a given pH
(1/d);
= the measured first-order reaction rate at pH 7 (1/d); and
= temperature adjustment (unitless), see (319).
In neutral hydrolysis reactions, the pollutant reacts with a water molecule (H20) and the
concentration of water is usually included in KUncat. In acid-catalyzed hydrolysis, the hydrogen
ion reacts with the pollutant, and a first-order decay rate for a given pH can be estimated as
follows:
248
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
KAcidExp = KAcid ¦ Hlon
where:
and where:
KAcid
Hlon
pH
rPH
HIon = 10
acid-catalyzed rate constant (L/mol-d);
concentration of hydrogen ions (mol/L); and
pH of water column.
(316)
(317)
Likewise for base-catalyzed hydrolysis, the first-order rate constant for a reaction between the
hydroxide ion and the pollutant at a given pH (Figure 137) can be described as:
KBaseExp = KBase ¦ OHIon
where:
and where:
KBase
OHIon
pH -14
OHIon = 10
= base-catalyzed rate constant (L/mol • d); and
= concentration of hydroxide ions (mol/L).
Figure 137. Base-catalyzed hydrolysis of pentachlorophenol
(318)
(319)
c ncn rva
6.00E-03
?
O 5.95E-03
£ 5.90E-03
<
cc
5.85E-03
5.80E-03
2
0
A
e
p
8
H
1
Hydrolysis reaction rates are adjusted for the temperature of the waterbody being modeled by
using the Arrhenius rate law (Hemond and Fechner 1994). An activation energy value of 18,000
cal/mol (a mid-range value for organic chemicals) is used as a default:
Arrhen
En
En
Q \ R • KelvinT R • TObs
(320)
where:
En
R
Arrhenius activation energy (cal/mol);
universal gas constant (cal/mol • Kelvin);
249
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
KelvinT
TObs
temperature for which rate constant is to be predicted (Kelvin); and
temperature at which known rate constant was measured (Kelvin).
8.3 Photolysis
Direct photolysis is the process by which a compound absorbs light and undergoes
transformation:
For consistency, photolysis is computed for both the epilimnion and hypolimnion in stratified
systems. However, photolysis is not a significant factor at hypolimnetic depths and is also
ignored in sediments.
Ionization may result in a significant shift in the absorption of light (Lyman et al., 1982;
Schwarzenbach et al., 1993). However, there is a general absence of information on the effects
of light on ionized species. The user provides an observed half-life for photolysis, and this is
usually determined either with distilled water or with water from a representative site, so that
ionization may be included in the calculated lumped parameter KPhot.
Based on the approach of Thomann and Mueller (1987; see also Schwarzenbach et al. 1993), the
observed first-order rate constant for the compound is modified by a light attenuation factor for
ultraviolet light so that the process as represented is depth-sensitive (Figure 138); it also is
adjusted by a factor for time-varying light:
Photolysis = KPhot ¦ ToxicantPhase
(321)
where:
Photolysis
KPhot
rate of loss due to photodegradation ([j,g/L-d); and
direct photolysis first-order rate constant (1/day).
KPhot = PhotRate ¦ ScreeningFactor ¦ LightFactor
(322)
where:
PhotRate =
ScreeningFactor
LightFactor =
direct, observed photolysis first-order rate constant (1/day);
a light screening factor (unitless), see (322); and
a time-varying light factor (unitless), see (323).
250
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
Figure 138. Photolysis of pentachlorophenol as a function of
light intensity and depth of water
O 0.4
X 0.2
100 200 300 400 500 600 700
LIGHT INTENSITY (ly/d)
-DEPTH (m)
-1.5 — 2 --2.5
— 0.5
A light screening factor adjusts the observed laboratory photolytic transformation rate of a given
pollutant for field conditions with variable light attenuation and depth (Thomann and Mueller,
1987):
ScreeningFactor
RadDistr 1 - exp
(- Extinct • Thick)
RcidDistrO Extinct ¦ Thick
(323)
where:
RadDistr
RadDistrO
Extinct
Thick
radiance distribution function, which is the ratio of the average
pathlength to the depth (see Schwarzenbach et al., 1993) (taken to
be 1.6, unitless);
radiance distribution function for the top of the segment (taken to
be 1.2 for the top of the epilimnion and 1.6 for the top of the
hypolimnion, unitless);
light extinction coefficient (1/m) not including periphyton, see
(40);
thickness of the water body segment if stratified or maximum
depth if unstratified (m).
The equation presented above implicitly makes the following assumptions:
quantum yield is independent of wavelength; and,
the value used for PhotRate is a representative near-surface, first-order rate constant for
direct photolysis.
251
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
The rate is modified further to represent seasonally varying light conditions and the effect of ice
cover:
T ^ SolarO
LightFactor = (324)
Ave Solar
where:
SolarO = time-varying average light intensity at the top of the segment (ly/day); and
AveSolar = average light intensity for late spring or early summer, corresponding to
time when photolytic half-life is often measured (default = 500 Ly/day).
If the system is unstratified or if the epilimnion is being modeled, the light intensity is the light
loading:
SolarO = Solar (325)
otherwise we are interested in the intensity at the top of the hypolimnion and the attenuation of
light is given as a logarithmic decrease over the thickness of the epilimnion:
SolarO = Solar • exp^(326)
where:
Solar = incident solar radiation loading (ly/d), see (25); and
MaxZMix = depth of the mixing zone (m), see (17).
Because the ultraviolet light intensity exhibits greater seasonal variation than the visible
spectrum (Lyman et al., 1982), decreasing markedly when the angle of the sun is low, this
construct could predict higher rates of photolysis in the winter than might actually occur.
However, the model also accounts for significant attenuation of light due to ice cover (see
section 3.6) so that photolysis, as modeled, is not an important process in northern waters in the
winter.
8.4 Microbial Degradation
Not only can microorganisms decompose the detrital organic material in ecosystems, they also
can degrade xenobiotic organic compounds such as fuels, solvents, and pesticides to obtain
energy. In AQUATOX this process of biodegradation of pollutants, whether they are dissolved
in the water column or adsorbed to organic detritus in the water column or sediments, is modeled
using the same equations as for decomposition of detritus, substituting the pollutant and its
degradation parameters for detritus in Equation (159) and supporting equations:
MicrobialDegrdn = KMDegrdnphase ¦ DOCorrection ¦ TCorr ¦ pHCorr
¦ Toxicant ph^e
where:
MicrobialDegrdn =
KMDegrdn =
(327)
loss due to microbial degradation (g/m -d);
maximum aerobic microbial degradation rate, either in
water column or sediments (1/d), in sediments this is
assumed to be four times the user-entered value for water:
252
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
DOCorrection
TCorr
effect of anaerobic conditions (unitless), see (160);
effect of suboptimal temperature (unitless), see (59);
effect of suboptimal pH (unitless), see (162); and
concentration of organic toxicant (g/m3).
pHCorr
Toxicant
Microbial degradation of toxicants proceeds more quickly if the material is associated with
surficial or particulate sediments rather than dissolved in the water column (Godshalk and Barko,
1985); thus, in calculating the loss due to microbial degradation in the sorbed phase, the
maximum degradation rate is converted by the model to four times the user entered maximum
chemical degradation rate in the water (Max. Rate of Aerobic Microbial Degradation). The
model assumes that reported maximum microbial degradation rates are for the dissolved phase; if
the reported degradation value is from a study with additional organic matter, such as suspended
slurry or wet soil samples, then the parameter value that is entered should be one-fourth that
reported.
8.5 Volatilization
Volatilization is modeled using the "stagnant boundary theory", or two-film model, in which a
pollutant molecule must diffuse across both a stagnant water layer and a stagnant air layer to
volatilize out of a waterbody (Whitman, 1923; Liss and Slater, 1974). Diffusion rates of
pollutants in these stagnant boundary layers can be related to the known diffusion rates of
chemicals such as oxygen and water vapor. The thickness of the stagnant boundary layers must
also be taken into account to estimate the volatile flux of a chemical out of (or into) the
waterbody.
The time required for a pollutant to diffuse through the stagnant water layer in a waterbody is
based on the well-established equations for the reaeration of oxygen, corrected for the difference
in diffusivity as indicated by the respective molecular weights (Thomann and Mueller, 1987, p.
533). The diffusivity through the water film is greatly enhanced by the degree of ionization
(Schwarzenbach et al., 1993, p. 243), and the depth-averaged reaeration coefficient is multiplied
by the thickness of the well-mixed zone:
KLiq = KReaer ¦ Thick ¦ MolWtO
MolWt
Nondissoc
1
(328)
where:
MolWt02
MolWt
Nondissoc
KLiq
KReaer
Thick
water-side transfer velocity (m/d);
depth-averaged reaeration coefficient for oxygen (1/d), see (191)-
(195);
thickness of the water body segment if stratified or maximum
depth if unstratified (m);
molecular weight of oxygen (g/mol, =32);
molecular weight of pollutant (g/mol); and
nondissociated fraction (unitless), see (311).
253
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
Likewise, the thickness of the air-side stagnant boundary layer is also affected by wind. Wind
usually is measured at 10 m, and laboratory experiments are based on wind measured at 10 cm,
so a conversion is necessary (Banks, 1975). To estimate the air-side transfer velocity of a
pollutant, we used the following empirical equation based on the evaporation of water, corrected
for the difference in diffusivity of water vapor compared to the toxicant (Thomann and Mueller,
1987, p. 534):
where:
KGas
Wind
0.5
Moimrno
KGas = 168- MolWtH2-
O
v 0.25
MolWt
¦ Wind ¦ 0.5
(329)
air-side transfer velocity (m/d);
wind speed ten meters above the water surface (m/s);
conversion factor (wind at 10 cm/wind at 10 m); and
molecular weight of water (g/mol, =18).
The total resistance to the mass transfer of the pollutant through both the stagnant boundary
layers can be expressed as the sum of the resistances- the reciprocals of the air- and water-phase
mass transfer coefficients (Schwarzenbach et al., 1993), modified for the effects of ionization:
KOVol KLiq KGas ¦ Henry Law ¦ Nondissoc
(330)
where:
(m/d);
KOVol = total mass transfer coefficient through both stagnant boundary layers
and where:
HenryLaw =
Henry =
HLCSaltFactor=
R
TKelvin =
HenryLaw
Henry ¦ HLCSaltFactor
R ¦ TKelvin
(331)
Henry's law constant (unitlessj;
3 1
Henry's law constant (atm m mol" );
Correction factor for effect of salinity (unitless), see (444).
3 1
gas constant (=8.206E-5 atm m (mol K)~ ); and
temperature in °K.
The Henry's law constant is applicable only to the fraction that is nondissociated because the
ionized species will not be present in the gas phase (Schwarzenbach et al., 1993, p. 179).
The atmospheric exchange of the pollutant can be expressed as the depth-averaged total mass
transfer coefficient times the difference between the concentration of the chemical and the
saturation concentration:
Volatilization = - • (ToxSat - Toxicant^) (332)
lhick
254
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
where:
Volatilization = interchange with atmosphere ([j,g/L-d);
Thick = depth of water or thickness of surface layer (m);
ToxSat = saturation concentration of pollutant in equilibrium with the gas
phase ((J-g/L), see (332); and
Toxicantwater = concentration of pollutant in water ((J,g/L).
The saturation concentration depends on the concentration of the pollutant in the air, ignoring
temperature effects (Thomann and Mueller, 1987, p. 532; see also Schnoor, 1996), but adjusting
for ionization and units:
ToxSat = Toxicant air 10Q0 p33^
HenryLaw ¦ Nondissoc
where:
Nondissoc = nondissociated fraction (unitless).
"3
Toxicantair = gas-phase concentration of the pollutant (g/m ); and
Theoretically, toxicants can be transferred in either direction across the water-air interface.
Often the pollutant can be assumed to have a negligible concentration in the air and ToxSat is
zero. However, this general construct can represent the transferral of volatile pollutants into
water bodies. Volatilization might become negative if toxicant concentrations are high in the air,
and concentrations in the water column may increase as a result of this interchange. Because
ionized species do not volatilize, the saturation level increases if ionization is occurring.
The nondimensional Henry's law constant, which relates the concentration of a compound in the
air phase to its concentration in the water phase, strongly affects the air-phase resistance.
Depending on the value of the Henry's law constant, the water phase, the air phase or both may
control volatilization. For example, with a depth of 1 m and a wind of 1 m/s, the gas phase is
100,000 times as important as the water phase for atrazine (Henry's law constant = 3.0E-9), but
the water phase is 50 times as important as the air phase for benzene (Henry's law constant =
5.5E-3). Volatilization of atrazine exhibits a linear relationship with wind (Figure 136) in
contrast to the exponential relationship exhibited by benzene (Figure 137).
255
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
Figure 139.
Wind
Atrazine KOVol as a function of
VOLATILIZATION OF ATRAZINE
4E-05
1 2.5E-05
5 2E-05
O 1.5E-05
* 1E-05
10.5 14 17.5 21 24.5 28
WIND (m/s)
Figure 140.
Wind
Benzene KOVol as a function of
VOLATILIZATION OF BENZENE
12 15 18 21
WIND (m/s)
24 27
-AQUATOX
Schwarzenbach etal., 1993
8.6 Partition Coefficients
Although AQUATOX is a kinetic model, steady-state partition coefficients for organic pollutants
are computed in order to place constraints on competitive uptake and loss processes in detritus
and plants, speeding up computations. Bioconcentration factors also are used in computing
internal toxicity in plants and animals. They are estimated from empirical regression equations
and the pollutant's octanol-water partition coefficient.
Detritus
Natural organic matter is the primary sorbent for neutral organic pollutants. Hydrophobic
chemicals partition primarily in nonpolar organic matter (Abbott et al. 1995). Refractory detritus
is relatively nonpolar; its partition coefficient (in the non-dissolved phase) is a function of the
octanol-water partition coefficient (N = 34, r = 0.93; Schwarzenbach et al. 1993):
where:
KOMf>efrDetr
KOW
KOMRefrDetr = 1-38- ROW
detritus-water partition coefficient (L/kg); and
octanol-water partition coefficient (unitless).
(334)
Detritus in sediments is simulated separately from inorganic sediments, rather than as a fraction
of the sediments as in other models. When the multi-layer sediment model is not included,
refractory detritus is used as a surrogate for sediments in general; and the sediment partition
coefficient KPSed, which can be entered manually by the user, is the same as KOMRefrDetr.
Equation (334) and the equations that follow are extended to polar compounds, following the
approach of Smejtek and Wang (1993):
256
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
KOM RefrDetr = 1-38- KOW082 ¦ Nondissoc
+ (1- Nondissoc) ¦ lonCorr ¦1.38 ¦ KOW0'82
where:
Nondissoc = un-ionized fraction (unitless); and
lonCorr = correction factor for decreased sorption, 0.01 for chemicals that are
bases and 0.1 for acids, (unitless).
Using pentachorophenol as a test compound, and comparing it to octanol, the influence of pH-
mediated dissociation is seen in Figure 141. This relationship is verified by comparison with the
results of Smejtek and Wang (1993) using egg membrane. However, in the general model Eq.
(334) is used for refractory detrital sediments as well.
Figure 141. Refractory detritus-water and octanol-water partition coefficients for pentachlorophenol as a
function of pH
1E6
1E5
1E4
1E3
1E2
1E1
3
4
5
6
7
8
9
PH
PCP KOM for refractory detritus
—- Un-ionized PCP - octanol/water
There appears to be a dichotomy in partitioning; data in the literature suggest that labile detritus
does not take up hydrophobic compounds as rapidly as refractory detritus. Algal cell membranes
contain polar lipids, and it is likely that this polarity is retained in the early stages of
decomposition. KOC does not remain the same upon aging, death, and decomposition, probably
because of polarity changes. In an experiment using fresh and aged algal detritus, there was a
100% increase in KOC with aging (Koelmans et al., 1995). KOC increased as the C/N ratio
increased, indicating that the material was becoming more refractory. In another study, KOC
doubled between day 2 and day 34, probably due to deeper penetration into the organic matrix
and lower polarity (Cornelissen et al., 1997).
Polar substrates increase the pKa of the compound (Smejtek and Wang, 1993). This is
represented in the model by lowering the pH of polar particulate material by one pH unit, which
changes the dissociation accordingly.
The partition equation for labile detritus (non-dissolved) is based on a study by Koelmans et al.
(1995) using fresh algal detritus (N = 3, r2 = 1.0):
257
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
KOC
LabPart
23.44-KOW
0.61
(336)
In the model, the equation is generalized to polar compounds and transformed to an organic
matter partition coefficient:
KOMLabDetr = (23.44 ¦ KOW06' ¦ Nondissoc
(1 - Nondissoc) ¦ IonCorr ¦ 23.44 ¦ KOW0'61) ¦ 0.526
(337)
where:
KOC LabPart
KOMLab£)efr
IonCorr
0.526
partition coefficient for labile particulate organic carbon (L/kg);
partition coefficient for labile detritus (L/kg);
correction factor for decreased sorption, 0.01 for chemicals that are
bases and 0.1 for acids, (unitless); and
conversion from KOC to KOM (g OC/g OM).
O'Connor and Connolly (1980; see also Ambrose et al., 1991) found that the sediment partition
coefficient is the inverse of the mass of suspended sediment, and Di Toro (1985) developed a
construct to represent the relationship. However, AQUATOX models partitioning directly to
organic detritus and ignores inorganic sediments, which are seldom involved directly in sorption
of neutral organic pollutants. Therefore, the partition coefficient is not corrected for mass of
sediment.
Association of hydrophobic compounds with colloidal and dissolved organic matter (DOM)
reduces bioavailability; such contaminants are unavailable for uptake by organisms (Stange and
Swackhamer 1994, Gilek et al. 1996). Therefore, it is imperative that complexation of organic
chemicals with DOM be modeled correctly. In particular, contradictory research results can be
reconciled by considering that DOM is not homogeneous. For instance, refractory humic acids,
derived from decomposition of terrestrial and wetland organic material, are quite different from
labile exudates from algae and other indigenous organisms.
Humic acids exhibit high polarity and do not readily complex neutral compounds. Natural
humic acids from a Finnish lake with extensive marshes were spiked with a PCB, but a PCB-
humic acid complex could not be demonstrated (Maaret et al. 1992). In another study, Freidig et
al. (1998) used artificially prepared Aldrich humic acid to determine a humic acid-DOC partition
coefficient (n = 5, r2, = 0.80), although they cautioned about extrapolation to the field. Landrum
et al. (1984) found that KOC values for natural dissolved organic matter were approximately one
order of magnitude less than for Aldrich humic acids (Gobas and Zhang 1994); incorporating
that factor into the equation of Freidig et al. (1998) yields:
KOCRefrDOM = 2.88-KOW067 (338)
where:
KOCRefrDOM = refractory dissolved organic carbon partition coefficient (L/kg).
Until a better relationship is found, we are using a generalization of this equation to include polar
compounds, transformed from organic carbon to organic matter, in AQUATOX:
258
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
KOMnefrDOM = (2-88 ¦ KOW067 • Nondissoc
(I - Nondissoc) ¦ IonCorr ¦ 2.88 ¦ KOW0'67) • 0.526
where:
KOMRefrDoM = refractory dissolved organic matter partition coefficient (L/kg).
Algae
Nonpolar lipids in algae occur in the cell contents, and it is likely that they constitute part of the
labile dissolved exudate, which may be both excreted and lysed material. Therefore, the stronger
relationship reported by Koelmans and Heugens (1998) for partitioning to algal exudate (n = 6, r2
= 0.926) is:
KOCladoc = 0.88 ¦ KOW (340)
which we also generalized for polar compounds and transformed:
KOM LabooM = (0.88-KOW- Nondissoc
(341)
+ (1 - Nondissoc) ¦ IonCorr ¦ 0.88 ¦ KOW) ¦ 0.526
where:
KOCLabDOC = partition coefficient for labile dissolved organic carbon (L/kg); and
KOMLabDOM = partition coefficient for labile dissolved organic matter (L/kg).
Unfortunately, older data and modeling efforts failed to distinguish between hydrophobic
compounds that were truly dissolved and those that were complexed with DOM. For example,
the PCB water concentrations for Lake Ontario, reported by Oliver and Niimi (1988) and used by
many subsequent researchers, included both dissolved and DOC-complexed PCBs (a fact which
they recognized). In their steady-state model of PCBs in the Great Lakes, Thomann and Mueller
(1983) defined "dissolved" as that which is not particulate (passing a 0.45 micron filter). In their
Hudson River PCB model, Thomann et al. (1991) again used an operational definition of
dissolved PCBs. AQUATOX distinguishes between truly dissolved and complexed compounds;
therefore, the partition coefficients calculated by AQUATOX may be larger than those used in
older studies.
Bioaccumulation of PCBs in algae depends on solubility, hydrophobicity and molecular
configuration of the compound, and growth rate, surface area and type, and content and type of
lipid in the alga (Stange and Swackhamer 1994). Phytoplankton may double or triple in one day
and periphyton turnover may be so rapid that some PCBs will not reach equilibrium (cf. Hill and
Napolitano 1997).
Hydrophobic compounds partition to lipids in algae, but the relationship is not a simple one.
Phytoplankton lipids can range from 3 to 30% by weight (Swackhamer and Skoglund 1991), and
not all lipids are the same. Polar phospholipids occur on the surface. Hydrophobic compounds
preferentially partition to internal neutral lipids, but those are usually a minor fraction of the total
lipids, and they vary depending on growth conditions and species (Stange and Swackhamer
259
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
1994). Algal lipids have a much stronger affinity for hydrophobic compounds than does octanol,
so that the algal BCFiiPid > K0w (Stange and Swackhamer 1994, Koelmans et al. 1995, Sijm et
al. 1998).
For algae, the approximation to estimate the dry-weight bioaccumulation factor (r = 0.87),
computed from Swackhamer and Skoglund's (1993) study of numerous PCB congeners, is:
log (BCF AigJ = 0.41 + 0.91 ¦ LogKOW (342)
where:
BCFAiga = partition coefficient between algae and water (L/kg).
Rearranging and extending to hydrophilic and ionized compounds:
BCFAiga = 2.57 ¦ KOW093 • Nondissoc
+ (1 - Nondissoc) ¦ IonCorr -0.257 ¦ KOW0'93
Comparing the results of using these coefficients, we see that they are consistent with the relative
importance of the various substrates in binding organic chemicals (Figure 140). Binding
capacity of detritus is greater than dissolved organic matter in Great Lakes waters (Stange and
Swackhamer 1994, Gilek et al. 1996). In a study using Baltic Sea water, less than 7% PCBs
were associated with dissolved organic matter and most were associated with algae (Bjork and
Gilek 1999). In contrast, in a study using algal exudate and a PCB, 98% of the dissolved
concentration was as a dissolved organic matter complex and only 2% was bioavailable
(Koelmans andHeugens 1998).
The influence of substrate polarity is evident in Figure 139, which shows the effect of ionization
on binding of pentachlorophenol to various types of organic matter. The polar substrates, such
as algal detritus, have an inflection point which is one pH unit higher than that of nonpolar
substrates, such as refractory detritus. The relative importance of the substrates for binding is
also demonstrated quite clearly.
Macrophytes
For macrophytes, an empirical relationship reported by Gobas et al. (1991) for 9 chemicals with
LogKOW,s of 4 to 8.3 (r2 = 0.97) is used:
log (BCF Macro) = 0.98 • LogKOW - 2.24 (344)
Again, rearranging and extending to hydrophilic and ionized compounds:
BCF Macro = 0. 00575 ¦ KOW098 • (Nondissoc + 0.2) (345)
260
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
Invertebrates
For the invertebrate bioconcentration factor, the following empirical equation is used for
nondetritivores, based on 7 chemicals with LogKOWs ranging from 3.3 to 6.2 and
bioconcentration factors for Daphniapulex (r = 0.85; Southworth et al., 1978; see also Lyman et
al., 1982), converted to dry weight:
log (BCF
Invertebrate
) (0.7520 ¦ LogKOW - 0.4362) ¦ WetToDry
(346)
where:
BCFInvertebrate
WetToDry
partition coefficient between invertebrates and water (L/kg); and
wet to dry conversion factor (unitless, default = 5).
Extending and generalizing to ionized compounds:
BCF Vertebrate = 0.3663 ¦ KOWa752° ¦ (Nondissoc + 0. 01)
(347)
For invertebrates that are detritivores the following equation is used, based on Gobas 1993:
FracLipid
D/^77 _ r
Invertebrate
1
where:
FracOC
¦ ¦ KOMnefrDetr ' (NotldiSSOC + 0. 01)
(348)
Detritus
BCFinvertebrate = partition coefficient between invertebrates and water (L/kg);
FracLipid = fraction of lipid within the organism;
FracOC Detritus = fraction of organic carbon in detritus (= 0.526);
KOMRefrDetr = partition coefficient for refractory sediment detritus (L/kg), see (334).
Figure 142. Partitioning to Various Types of
Organic Matter as Function of Kow
o
1E10
1E9
1E8
1E7
1E6
1E5
1E4
1E3
1E2
6 7
Log KOW
humic acids algae exudate
algal detritus » refr. detritus - sediments
Figure 143. Partitioning to Various Types of
Organic Matter as a Function of pH
o
1E6
1E5
1E4
1E3
humic acids
exudate
6
PH
algae octanol/water
algal detritus • refr. detritus
Fish
Fish take longer to reach equilibrium with the surrounding water; therefore, a nonequilibrium
bioconcentration factor is used. For each pollutant, a whole-fish bioconcentration factor is based
261
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
on the lipid content of the fish extended to hydrophilic chemicals (McCarty et al., 1992), with
provision for ionization:
KBmSh= Lipid ¦ WetToDry ¦ KOW ¦ (Nondissoc + 0.01)
(349)
where:
KB Fish
Lipid
WetToDry
partition coefficient between whole fish and water (L/kg);
fraction of fish that is lipid (g lipid/g fish); and
wet to dry conversion factor (unitless, default = 5).
The bioconcentration factor is adjusted for the time to reach equilibrium as a function of the
clearance or elimination rate and the time of exposure (Hawker and Connell, 1985; Connell and
Hawker, 1988; Figure 144):
BCF
Fish
KB
Fish
(>¦
(-Depuration • TElapsed)
)
(350)
where:
BCFFish -
TElapsed =
Depuration =
quasi-equilibrium bioconcentration factor for fish (L/kg);
time elapsed since fish was first exposed (d); and
clearance, which may include biotransformation, see (372) (1/d).
Figure 144. Bioconcentration factor for fish as a function
of time and log KOW
<
or
LU
O
z
O
O
o
CO
1E7
I I | | | I I I I | |
tog KO
A/ = 8
log KOW = 3
0 200 400 600 800 1000 1200
DAY
8.7 Nonequilibrium Kinetics
Often there is an absence of equilibrium due to growth or insufficient exposure time, metabolic
biotransformation, dietary exposure, and nonlinear relationships for very large and/or
superhydrophobic compounds (Bertelsen et al. 1998). Although it is important to have a
knowledge of equilibrium partitioning because it is an indication of the condition toward which
systems tend (Bertelsen et al. 1998), it is often impossible to determine steady-state potential due
262
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
to changes in bioavailability and physiology (Landrum 1998). For example, PCBs may not be at
steady state even in large systems such as Lake Ontario that have been polluted over a long
period of time. In fact, PCBs in Lake Ontario exhibit a 25-fold disequilibrium (Cook and
Burkhard 1998). The challenge is to obtain sufficient data for a kinetic model (Gobas et al.
1995).
Sorption and Desorption to Detritus
Partitioning to detritus appears to involve rapid sorption to particle surfaces, followed by slow
movement into, and out of, organic matter and porous aggregates (Karickhoff and Morris,
1985). Therefore attainment of equilibrium may be slow. Because of the need to represent
sorption and desorption separately in detritus, kinetic formulations are used (Thomann and
Mueller, 1987), with provision for ionization:
Sorption = k lDetr ¦ Toxiccmtwater ¦ (Nondissoc + 0.01)
v /
• Org2C ¦ Detr ¦ UptakeLimit ¦ 7e - 6
Desorption = k 2Detr • ToxicantL
(352)
where:
Sorption
kl Detr
Nondissoc
Toxicant Water
Org2C
Detr
le-6
Desorption
k— Detr
UptakeLimit
Toxicant Detr
rate of sorption to given detritus compartment ([j,g/L-d);
sorption rate constant (user-editable, default value of 1.39 L/kg-d),
see (355);
fraction not ionized (unitless), see (311);
concentration of toxicant in water ((J,g/L);
conversion factor for organic matter to carbon (= 0.526 g C/g
organic matter);
mass of each of the detritus compartments per unit volume (mg/L);
units conversion (kg/mg);
rate of desorption from given sediment detritus compartment
(ng/L-d);
desorption rate constant (1/d), see (354);
factor to limit uptake as equilibrium is reached (unitless) see (352);
and
mass of toxicant in each of the detritus compartments ((J-g/L).
In order to limit sorption to detritus and algae as equilibrium is reached, UptakeLimit is
computed as:
UptakeLimitCarner = Toxicantw^' kPcamer ~ PPBcamer (353)
Toxicant Water ' kpCarrier
where:
UptakeLimit carrier = factor to limit uptake as equilibrium is reached (unitless);
263
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
kpCarrier = partition coefficient (KOM) or bioconcentration factor (BCF) for
each carrier (L/kg), see (333) to (342);
PPBCarrier = concentration of toxicant in each carrier ([j,g/kg), see (310).
Desorption of the detrital compartments is the reciprocal of the reaction time, which Karickhoff
and Morris (1985) found to be a linear function of the partition coefficient over three orders of
magnitude (r2 = 0.87):
— ~ 0.03 -24- KOM (354)
k2
So k2 is taken to be:
1 39
k2 = — (355)
KOM
where:
KOM = detritus-water partition coefficient (L/kg OM, see section 8.6); and
24 = conversion from hours to days.
Because the kinetic definition of the detrital partition coefficient KOM is:
kl
KOM =— (356)
kl
the sorption rate constant kl is set by the user (Kl Detritus). The default value is 1.39 L/kg-d.
Bioconcentration in Macrophytes and Algae
Macrophytes: As Gobas et al. (1991) have shown, submerged aquatic macrophytes take up and
release organic chemicals over a measurable period of time at rates related to the octanol-water
partition coefficient. Uptake and elimination are modeled assuming that the chemical is
transported through both aqueous and lipid phases in the plant, with rate constants using
empirical equations fit to observed data (Gobas et al., 1991), modified to account for ionization
effects (Figure 145, Figure 146):
Macro Uptake = kl ¦ Toxicant water' StVar Piant -le-6 (357)
Depurationplant = k2 ¦ Toxicant Piant (358)
Wo <359)
0.0020 +
KOW ¦ Nondissoc
If the user selects to estimate the elimination rate constant based on KOW (see section 8.8), the
following equation is used:
264
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
k2
1.58 + 0.000015 ¦ KOW ¦ Nondissoc
(360)
where:
MacroUptake
Depuration piant
StVar piant
1 e-6
Toxicant piant
kl
k2
KOW
Nondissoc
uptake of toxicant by plant (pg/L-d);
clearance of toxicant from plant (pg/L-d);
biomass of given plant (mg/L);
units conversion (kg/mg);
mass of toxicant in plant (pg/L);
sorption rate constant (L/kg-d);
elimination rate constant (1/d).
octanol-water partition coefficient (unitless); and
fraction of un-ionized toxicant (unitless).
Figure 145. Uptake rate constant for macrophytes
(after Gobas et al., 1991)
500
400
300
200
100
0
0
Log KOW
— Predicted ¦ Observed
Figure 146. Elimination rate constant for macrophytes
(after Gobas et al., 1991)
D
0.6
\-
\
\
CM U A
"*"0 3
V
\
A
\
0
2
— F
4
3redi
Log
cted
6
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
content (0.2% in Myriophyllum spicatum as observed by Gobas et al. (1991), which affect both
uptake and elimination of toxicants. However, the approach used by Gobas et al. (1991) in
modeling bioaccumulation in macrophytes provides a useful guide to modeling kinetic uptake in
algae.
There is probably a two-step algal bioaccumulation mechanism for hydrophobic compounds,
with rapid surface sorption of 40-90% within 24 hours and then a small, steady increase with
transfer to interior lipids for the duration of the exposure (Swackhamer and Skoglund 1991).
Uptake increases with increase in the surface area of algae (Wang et al. 1997). Therefore, the
smaller the organism the larger the uptake rate constant (Sijm et al. 1998). However, in small
phytoplankton, such as the nannoplankton that dominate the Great lakes, a high surface to
volume ratio can increase sorption, but high growth rates can limit internal contaminant
concentrations (Swackhamer and Skoglund 1991). The combination of lipid content, surface
area, and growth rate results in species differences in bioaccumulation factors among algae
(Wood et al. 1997). Uptake of toxicants is a function of the uptake rate constant and the
concentration of toxicant truly dissolved in the water, and is constrained by competitive uptake
by other compartments; also, because it is fast, it is limited as it approaches equilibrium, similar
to sorption to detritus :
AlgalUptake = kl ¦ UptakeLimitAlga ¦ ToxState ¦ Carrier ¦ 1 e - 6
(361)
where:
AlgalUptake =
kl =
UptakeLimitAiga =
ToxState =
Carrier =
le-6
rate of sorption by algae ([j,g/L-d);
uptake rate constant (L/kg-d), see (361);
factor to limit uptake as equilibrium is reached (unitless), see
(352);
concentration of dissolved toxicant ((J-g/L);
biomass of algal compartment (mg/L); and
conversion factor (kg/mg).
The kinetics of partitioning of toxicants to algae is based on studies on PCB congeners in The
Netherlands by Koelmans, Sijm, and colleagues and at the University of Minnesota by Skoglund
and Swackhamer. Both groups found uptake to be very rapid. Sijm et al. (1998) presented data
on several congeners that were used in this study to develop the following relationship for
phytoplankton (Figure 147):
kl = (362)
1.8 E- 6 + l/(KOW • (Nondissoc + 0.01))
Because size-dependent passive transport is indicated (Sijm et al., 1998), uptake by periphyton is
set arbitrarily at ten percent of that for phytoplankton.
Depuration is modeled as a linear function; it does not include loss due to excretion of
photosynthate with associated toxicant, which is modeled separately:
Depuration = k2 ¦ State (363)
where:
266
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
Depuration =
State =
k2 =
elimination of toxicant (pg/L-d);
concentration of toxicant associated with alga (pg/L); and
elimination rate constant (1/d).
As a simplifying assumption, the depuration rate for periphyton is assumed to be two orders of
magnitude less:
Depuration = k2 ¦ State ¦ 0.01
(364)
The elimination rate in plants may be input in the toxicity record by the user or it may be
estimated using the following equation based in part on Skoglund et al. (1996). Unlike
Skoglund, this equation ignores surface sorption and recognizes that growth dilution is explicit in
AQUATOX (see Figure 148):
where:
k-. \ ac
LFrac
WetToDry
k2
2.4E + 5
Algae
(KOW ¦ LFrac ¦ WetToDry)
(365)
desorption rate constant (1/d);
fraction lipid (wet weight), entered in the "chemical toxicity''
screen; and
translation from wet to dry weight (user input).
Figure 147. Algal sorption rate constant as a function
of octanol-water partition coefficient
FIT TO DATA OF SUM ET AL. 1998
600000
500000
=? 400000
~ 300000
100000
4 6
LOG KOW
Obs K1
10
Pred K1
267
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
Figure 148. Rate of elimination by algae as a function of
octanol-water partition coefficient
r
V
\
0 2 4 6 8 10
Log KOW
Bioaccumulation in Animals
Animals can absorb toxic organic chemicals directly from the water through their gills and from
contaminated food through their guts. Direct sorption onto the body is ignored as a simplifying
assumption in this version of the model. Reduction of body burdens of organic chemicals is
accomplished through excretion and biotransformation, which are often considered together as
empirically determined elimination rates. "Growth dilution" occurs when growth of the
organism is faster than accumulation of the toxicant. Gobas (1993) includes fecal egestion, but
in AQUATOX egestion is merely the amount ingested but not assimilated; it is accounted for
indirectly in DietUptake. However, fecal loss is important as an input to the detrital toxicant
pool, and it is considered later in that context. Inclusion of mortality and promotion terms is
necessary for mass balance, but emphasizes the fact that average concentrations are being
modeled for any particular compartment.
Gill Sorption: An important route of exposure is by active transport through the gills (Macek et
al., 1977). This is the route that has been measured so often in bioconcentration experiments
with fish. As the organism respires, water is passed over the outer surface of the gill and blood is
moved past the inner surface. The exchange of toxicant through the gill membrane is assumed to
be facilitated by the same mechanism as the uptake of oxygen, following the approach of
Fagerstrom and Asell (1973, 1975), Weininger (1978), and Thomann and Mueller (1987; see
also Thomann, 1989). Therefore, the uptake rate for each animal can be calculated as a function
of respiration (Leung, 1978; Park et al., 1980):
GillUptake = KUptake ¦ Toxicant Wate,¦ • FracWaterCohwm (366)
WEfjTox ¦ Respiration ¦ OlBiomass
KUptake (367)
Oxygen ¦ WEff02
where:
GillUptake = uptake of toxicant by gills ((J,g/L - d);
268
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
KUptake = uptake rate (1/d);
Toxicantwaer = concentration of toxicant in water ((J,g/L);
FracwaterCoiumn = fraction of organism in water column (unitless), differentiates from
pore-water uptake if the multi-layer sediment model is included;
WEjflox = withdrawal efficiency for toxicant by gills (unitless), see (367);
Respiration = respiration rate (mg biomass/L-d), see (100);
02Biomass = ratio of oxygen to organic matter (mg oxygen/mg biomass; 0.575);
Oxygen = concentration of dissolved oxygen (mg oxygen/L), see (186); and
WEff02 = withdrawal efficiency for oxygen (unitless, generally 0.62);
The oxygen uptake efficiency WEff02 is assigned a constant value of 0.62 based on observations
of McKim et al. (1985). The toxicant uptake efficiency, WEjfTox, can be expected to have a
sigmoidal relationship to the log octanol-water partition coefficient based on aqueous and lipid
transport (Spacie and Hamelink, 1982). This is represented by an inelegant but reasonable,
piece-wise fit (Figure 149) to the data of McKim et al. (1985) using 750-g fish, corrected for
ionization:
If LogKOW <1.5 then
WEffTox = 0.1
If 1.5 < LogKOW >3.0 then
WEjfTox = 0.1 + Nondissoc • (0.3 • LogKOW - 0.45)
If 3.0 < LogKOW <6.0 then
6 (368)
WEjfTox = 0.1 + Nondissoc • 0.45
If 6.0 < LogKOW < 8.0 then
WEjfTox = 0.1 + Nondissoc • (0.45 - 0.23 • {LogKOW - 6.0))
If LogKOW >8.0 then
WEjfTox = 0.1
where:
LogKOW = log octanol-water partition coefficient (unitless); and
Nondissoc = fraction of toxicant that is un-ionized (unitless), see (311).
269
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
Figure 149. Piece-wise fit to observed toxicant uptake data;
Modified from McKim et al., 1985
80
60
z
n 40
LOG KOW
Ionization decreases the uptake efficiency (Figure 147). This same algorithm is used for
invertebrates. Thomann (1989) has proposed a similar construct for these same data and a
slightly different construct for small organisms, but the scatter in the data does not seem to
justify using two different constructs.
Figure 150. The Effect of Differing Fractions of Un-
ionized Chemical on Uptake Efficiency
0
0 2 4 6 8 10
Log KOW
The user input FracwaerCoinmn parameter is only relevant if the multi-layer sediment model is
included. If so, this parameter determines how much gill uptake comes from the water column
and how much from the pore waters of the active layer. Gill uptake from pore waters is
calculated as follows and added to gill uptake from the water column:
GiUUptakePoreWater = KUptake ¦ Toxicant PoreWater
• (l FraCWaterCohmni)' tt j (^69)
VolumeWaterCol
where:
270
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
GillUptake =
ToxicantporeWater
Volume PoreWater
Volume WaterCol
uptake of toxicant by gills ((J,g/LwaterCoi - d);
concentration of toxicant in pore waters (^g/LporeWater);
volume of pore water (Lporewater); and
volume of water column (LWaterCoi)-
Dietary Uptake: Hydrophobic chemicals usually bioaccumulate primarily through absorption
from contaminated food. Persistent, highly hydrophobic chemicals demonstrate
biomagnification or increasing concentrations as they are passed up the food chain from one
trophic level to another; therefore, dietary exposure can be quite important (Gobas et al., 1993).
Uptake from contaminated prey can be computed as (Thomann and Mueller, 1987; Gobas,
1993):
where:
DietUptake p = KDprey ¦ PPBprey • - 6
(370)
KDprey = GutEffTox ¦ GutEffRed ¦ Ingestion
Prey
(371)
and:
DietUptake prey
KDprey
PPBprey
1 e-6
GutEffTox
GutEffRed
Ingestion prey
uptake of toxicant from given prey ((_ig toxicant/L-d);
dietary uptake rate for given prey (mg prey/L-d);
conc. of toxicant in given prey ([j,g toxicant/kg prey), see (310);
units conversion (kg/mg);
efficiency of sorption of toxicant from gut (unitless);
reduction in GutEffTox due to non-lethal effects, see (371); and
ingestion of given prey (mg prey/L-d), see (91).
Gobas (1993) presents an empirical equation for estimating GutEffTox as a function of the
octanol-water partition coefficient. However, data published by Gobas et al. (1993) suggest that
there is no trend in efficiency between LogKOW 4.5 and 7.5 (Figure 151); this is to be expected
because the digestive system has evolved to assimilate a wide variety of organic molecules.
Therefore, the mean value of 0.62 is used in AQUATOX as a constant for small fish. Nichols et
al. (1998) demonstrated that uptake is more efficient in larger fish; therefore, a value of 0.92 is
used for large game fish because of their size. Invertebrates generally exhibit lower efficiencies;
Landrum and Robbins (1990) showed that values ranged from 0.42 to 0.24 for chemicals with
log KOWs from 4.4 to 6.7; the mean value of 0.35 is used for invertebrates in AQUATOX.
These values cannot be edited at this time. (Note, the PFA model uses a relationship to chain
length, see (403) and (404).)
271
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
Figure 151. GutEffTox constant based on mean value for data
from Gobas et al., 1993
1
>.
O
c
0
1 0-75
C
o
o. 0.5
o
00
_Q
<0.25
c
ro
0)
4.5
5.5 6 6.5
Log KOW
7.5
Guppies
Goldfish — Mean = 0.63
One potential non-lethal effect of toxicant exposure is an increase in the rate of egestion, see
(425). If GutEffTox is kept constant at the same time that the egestion rate is increased, toxicant
concentrations will increase too much within organisms (biomass falls but toxicant uptake
remains constant). To avoid this problem, and to reflect that the rate of toxicant uptake is more
a function of assimilated rather than total ingested food, the GutEffTox must be reduced by the
same quantity that assimilated food is decreased.
GutEffRed = 1 - RedGrow
(372)
where:
GutEffRed = reduction in GutEffTox due to toxicant induced increased egestion
(unitless);
RedGrow = factor for reduced assimilation of food in animals (unitless); see
(422).
Despite this adjustment, if overall species growth rates become negative due to the reduced
assimilation of food in animals, toxicant concentrations in animals will still increase (a process
that is best conceived as the opposite of growth dilution.)
Elimination: Elimination or clearance includes both excretion (depuration) and
biotransformation of a toxicant by organisms. Biotransformation may cause underestimation of
elimination (McCarty et al., 1992). An overall elimination rate constant is estimated and
reported in the toxicity record. The user may then modify the value based on observed data; that
value is used in subsequent simulations. If, known, biotransformation also can be explicitly
modeled.
For any given time the clearance rate is:
Depuration Anmal = k2 ¦ Toxicant Amma • TCorr
272
(373)
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
where:
DepurationAnimai
k2
Toxicant Animal
TCorr
clearance rate ([j,g/L-d);
elimination rate constant (1/d);
mass of toxicant in given animal ((J-g/L); and
correction for suboptimal temperature (unitless), see (59).
If the multi-layer sediment model is included, the amount of depuration that goes to the water
column vs. the active layer of pore waters is determined by the user input "Frac. in Water
Column" parameter.
Estimation of the elimination rate constant k2 is based on a slope related to log Kow and an
intercept that is a direct function of respiration, assuming an allometric relationship between
respiration and the weight of the animal (Thomann, 1989), and an inverse function of the lipid
content in a construct unique to AQUATOX:
If WetWt < 5 g then
Log k2 = -0.536 ¦ Log Kow ¦ Log NonDissoc + 0.065 ¦
WetWtRB
LipidFrac
(374)
else
Log k2 = -0.536 ¦ Log Kow ¦ Log NonDissoc + 0.116 ¦
WetWtRB
LipidFrac
(375)
where
WetWt
RB
Kow
NonDissoc
LipidFrac
octanol-water partition coefficient (unitless);
fraction of toxicant that is un-ionized (unitless), see (311);
fraction of lipid in organism (g lipid/g organism wet);
mean wet weight of organism (g);
allometric exponent for respiration (unitless).
273
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
Figure 152. Depuration rate constants for invertebrates and fish
based on AQUATOX "classic" formulation (equations 373 and 374)
CM
U)
o
4
3
2
1
0
-1
-2
-3
-4
K2 for Various Animals
4 5 6
Log KOW
¦ Daphnia
Diporeia
~ 10-g fish
A Eel obs
- Eel
Linear (10-g fish pred)
Linear (Daphnia pred)
Linear (Diporeia pred)
In AQUATOX Release 3.1 and after, an alternative k2 estimation procedure is available based on
Barber (2003):
K_ C- WeM*"" (374b)
LipidFrac ¦ Kow
where
C
WetWt
LipidFrac
Kow
constant of 445 for fish and 890 for invertebrates;
mean wet weight of organism (g);
fraction of lipid in organism (g lipid/g organism wet);
octanol-water partition coefficient (unitless);
Barber's (2003) formulation is based on uptake rates divided by LipidFracxKow fas a surrogate
for BCF). The uptake rate equation utilized is based on an allometric analysis of 517 data points,
though there is a high degree of uncertainty in this relationship. Figure 153 shows that the
AQUATOX and Barber formulations have different relationships between predicted elimination
rates and Kow. Our testing suggests that some studies benefit from one uptake formulation and
some benefit from the other; however, at this point there is no general guidance as to which
formulation to use in a given application.
274
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
Figure 153. k2 predictions by Log Kqw for a lOOg fish with 5% lipid
Barber K2
AQUATOX K2
Biotransformation: Biotransformation can cause the conversion of a toxicant to another
toxicant or to a harmless daughter product through a variety of pathways. Internal
biotransformation to given daughter products by plants and animals is modeled by means of
empirical rate constants provided by the user in the "Chemical Biotransformation" screen:
Biotransformation = Toxicantorgamsm • BioRateConstorgamsm, tax (376)
where
Biotransformation = rate of conversion of chemical by given organism ((J,g/L d),
BioRateConst = biotransformation rate constant to a given toxicant,
provided by user (1/day)
with the model keeping track of both the loss and the gains to various daughter compartments.
A simplifying assumption of the model is that biotransformation occurs at a constant rate
throughout a simulation.
Biotransformation also can take place as a consequence of microbial decomposition. The
percentage of microbial biotransformation from and into each of the organic chemicals in a
simulation can be specified, with different values for aerobic and anaerobic decomposition. The
amount of biotransformation into a given chemical can then be calculated as follows for aerobic
conditions:
BiotransformMicrobln = TorgTOX MicrobialDegradn0rgTox ¦ FracAerobic ¦ Frac0rgTOX (377)
275
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
and for anaerobic conditions:
Biotransform
Microb In
^OrgTox
Microbial DegradnQrgTox ¦ (1 - FracAerobic) ¦ Frac
OrgTox
(378)
where:
Biotransform Microb in
MicrobialDegradn
Frac Aerobic
Frac orgTox
Biotransformation to a given organic chemical in a given
detrital compartment due to microbial decomposition ((J,g/L d);
total microbial degradation of a different toxicant in this detrital
compartment ((J,g/L d) see (326);
fraction of the microbial degradation that is aerobic (unitless),
see (378); and
user input fraction of the organic toxicant that is transformed to
the current organic toxicant (inputs can differ depending on
whether the degradation is aerobic or anaerobic).
To calculate the fraction of microbial decomposition that is aerobic, the following equation is
used:
FracAerobic
Factor
DOCorrection
(379)
where:
Factor =
DOCorrection =
Michaelis-Menten factor (unitless) see (161);
effect of oxygen on microbial decomposition (unitless) see (160).
Bioaccumulation Factor: Customarily, bioaccumulation is expressed as a bioaccumulation
factor (BAF), which is the ratio of the concentration in the organism to that in the water. The
BAF can be expressed as a wet-weight, dry-weight, or lipid-normalized basis (Gobas and
Morrison 2000). In AQUATOX, the BAFs are output as both wet-weight and wet-weight lipid-
normalized values. The concentration in an organism is wet-weight, and the lipid fraction is
input by the user as a wet-weight value:
BAFup* = Log 10
( PPBOr„ ' ToXjCCOtt Water ^
FracLipid
BAFWet = Log 10 (PPB
Organism
/ Toxicant Water)
(378b)
where:
PPB Organism
FracLipid
Toxicant Water
concentration of toxicant in given animal ((J-g/kg wet);
fraction of organism that is lipid (g lipid/g organism wet); and
concentration of toxicant in water ((J,g/L);
276
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
Linkages to Detrital Compartments
Toxicants are transferred from organismal to detrital compartments through defecation and
mortality. The amount transferred due to defecation is the unassimilated portion of the toxicant
that is ingested:
DefecationTox = 2/KEgestPredPrey ¦ PPBprey • 1 e - 6) (380)
KEgestPre* Prey (i~ GutEffTox • GutEffRed) ¦ Ingestion PmdPmy (381)
where::
DefecationTox = rate of transfer of toxicant due to defecation (p,g/L-d);
KEgestpreci, prey = fecal egestion rate for given prey by given predator (mg
prey/L-d);
PPBprey = concentration of toxicant in given prey (p,g/kg), see (310);
1 e-6 = units conversion (kg/mg);
GutEffTox = efficiency of sorption of toxicant from gut (unitless); and
GutEffRed = reduction in GutEffTox due to non-lethal effects, see (371) ;
IngestionPred, Prey = rate of ingestion of given prey by given predator (mg/L-d), see
(91)
The amount of toxicant transferred due to mortality may be large; it is a function of the
concentrations of toxicant in the dying organisms and the mortality rates:
MortTox = 2/MortalityQrg ¦ PPB0rg • le6) (382)
where:
MortTox = rate of transfer of toxicant due to mortality (^g/L-d);
Mortality org = rate of mortality of given organism (mg/L-d), see (66), (87) and
(H2);
PPBorg = concentration of toxicant in given organism ([j,g/kg), see (310); and
1 e-6 = units conversion (kg/mg).
8.8 Alternative Uptake Model: Entering BCFs, Kl, and K2
When performing bioaccumulation calculations, the default behavior of the AQUATOX model is
to allow the user to enter elimination rate constants (K2) for all plants and animals for a
particular organic chemical. K2 values may also be estimated based on the Log Kow of the
chemical. Uptake in plants is a function of Log K0w while gill uptake in animals is a function of
respiration and chemical uptake efficiency. The AQUATOX default model works well for a
wide variety of bioaccumulative organic chemicals, but some chemicals that are subject to very
rapid uptake and depuration are not efficiently modeled using these relationships; the rapid rates
create stiff equations that require shorter time-steps for solution. In addition, because of the
rapid rates, the chemical does approach equilibrium quickly.
277
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
For this reason, an alternative uptake model is provided to the user. In the chemical toxicity
record, the user may enter two of the three factors defining uptake (BCF, Kl, K2+Km) and the
third factor is calculated using the below relationship (Gobas and Morrison 2000, p204) (note, if
the option to estimate the K2 depuration rate based on BCF and Kl is selected, the elimination
rate is estimated as the K2 parameter and the metabolism rate is considered to be zero.) :
BCF = — (383)
(K2 + KJ
where: BCF = bioconcentration factor (L/kg dry);
Kl = uptake rate constant (L/kg dry day);
K2 = elimination rate constant (1/d);
Km = metabolism or biotransformation, see (375).
Given these parameters, AQUATOX calculates uptake and depuration in plants and animals as
kinetic processes.
Uptake = Kl ¦ ToxState ¦ Biomass • 1 e - 6
Depuration = K2- ToxState
where: Uptake = uptake rate within organism (|_ig/L day);
Kl = uptake rate constant (L/kg dry day);
ToxState = concentration of toxicant in organism in water (ng/L)
Biomass = concentration organism in water (mg/L)
le-6 = (kg/mg)
Depuration = loss rate within organism (|_ig/L day);
K2 = elimination rate constant (1/d).
(384)
(385)
Dietary uptake of chemicals by animals is not affected by this alternative parameterization.
8.9 Half-Life Calculation, DT50 and DT95
AQUATOX estimates time to 50% (half-lives, DT50s) and time to 95% chemical loss (DT95s)
independently in bottom sediment and in the water column. Estimates are produced at each
output time-step depending on the average loss rate during that time-step in that medium.
HydrolysisWater + Photolysis + MicrobialWater + Washout + Volatilization. + Sorption
L°SSWater =
ATaSS water
278
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
L0SSSed = '
MicrobialSed + Hydrolysis Sed + Desorption + Scour\
'Sed
Mass
Sed
(387)
where:
LoSSMedia
Hydrolysis Media
Photolysis
Microbial Media
Washout
Volatilization
Sorption
-MaSSMedia
Desorption
Scour
loss rate within media (1/d);
hydrolysis rate in given media (|J,g/L d), see (313);
photolysis rate in the water column (|_ig/L d), see (320);
rate of microbial metabolism in given media (|~ig/L d), see (326);
rate of toxicant washout from the water column (|_ig/L d); see (16);
rate of chemical volatilization in the water column (|_ig/L d), see (331);
sorption of toxicant to detritus, plants, and animals (|~ig/L d), see (350);
mass of chemical in the media (ng/L);
desorption of toxicant from bottom sediment, (ng/L d) see (351);
resuspension of toxicants in bottom sediments, (|~ig/L d) see (233).
Loss rates are converted into time to 50% and 95% loss using the following formulae for first-
order reactions:
DT50Media = 0.6931 LossMedia
DT95Media = 2.996! LossMedia
(388)
(389)
where: DT50Medm
D T9 5 Media
LoSSMedia
time in which 50% of chemical will be lost at current loss rate (d);
time in which 95% of chemical will be lost at current loss rate (d);
loss rate within media (1/d);
8.10 Chemical Sorption to Sediments
When the complex multi-layer sediment model is included, chemicals can sorb to and desorb
from suspended inorganic sediments based on user input rates that are applied to the model's
equations for sorption (249), and desorption (250). To activate this model, required rates are:
K1 uptake rate constant L/kg dry day
K2 depuration rate constant 1/day
Kp partition coefficient L/kg dry
The derivative for toxicants sorbed to inorganic sediments is similar to that for suspended
organics:
279
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
dToxicant
SuspSed
dt
= Load - Microbial + Sorption -Desorption
SuspSed
- (.Deposition + Washout) ¦ PPB
+ (Washin ¦ PPB suSpsedupstream• le-6)
+ (Scour • ppBBottomSed ¦ le-6)
le-6
(390)
where:
Toxicant suspsed
Load
Microbial
Sorption
Desorption
Deposition
Washout
Washin
Scour
toxicant in relevant suspended sediment size-class ((J-g/L);
loading of toxicant from external sources ([j,g/L-d);
rate of loss due to microbial degradation ([j,g/L-d), see (326);
rate of sorption to given compartment ([j,g/L-d), see (350);
rate of desorption from given compartment ([j,g/L-d), see (351);
rate of sedimentation of given suspended detritus (mg/L-d) in
streams with the inorganic sediment model attached, see (230);
rate of loss of from sediment being carried downstream (mg/L-d),
see (16)
rate of gain from sediment carried in from any upstream linked
segments (mg/L-d), see (30);
rate of resuspension of given sediment (mg/L-d), see (227);
Chemicals also are tracked within inorganic sediments in the multi-layer sediment bed:
dTnxirrmf
BottomSed = Sorption -Desorption - Microbial
dt
+ (Deposition ¦ PPB suspsed ' 1 e - 6) + BedLoadJox
-(Scour • PPB BottomSed ¦ le-6)- Redl,osslox (391)
where:
Toxicant BottomSed
Microbial
Sorption
Desorption
Deposition
Scour
BedLoadfox
= toxicant in bottom sediment (relevant sediment size-class
l^g/m2);
rate of loss due to microbial degradation ((J,g/m -d), see (326);
rate of sorption to given compartment ((J,g/m -d after units
conversion), see (350);
rate of desorption from given compartment ((J,g/m -d after units
conversion), see (351);
rate of sedimentation of given suspended detritus ((J,g/m -d after
units conversion) in streams with the inorganic sediment model
attached, see (230);
rate of resuspension of given sediment ((J,g/m -d after units
conversion), see (227);
rate of bed load of given toxicant (|ig/m -d), see (391);
280
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
BedLoss
Tox
rate of bed loss of given toxicant ((J,g/m -d), see (392).
In several cases above, units need to be converted from (j,g/L-d to (J,g/m -d when moving from
sediment suspended in the water column to bed sediment. This is done by multiplying by water
volume and then dividing by the sediment bed surface area. Toxicant mass balance has been
verified to be conservative through this process.
Toxicant movement due to bedload and bedloss are straightforward calculations:
BedLoadTox = ^
Bed Load,
Upstreamlink
AvgArea
PPB • 1e - 3
UpstreamBed
(392)
where:
BedLoadfox
B edLoad IJpstreamlink
AvgArea
PPB UpstreamBed
le-3
toxicant bedload from all upstream segments ((J,g/m -d);
bedload over one of the upstream links (g/d);
average area of the segment (m );
toxicant concentration in the relevant upstream link ([j,g/kg)
units conversion (kg/g)
Similarly, total bed loss is the sum of the loadings over all outgoing links:
BedLossTox = £
BedLoss
\
Downstreamlink t PPQ • 16 3
AvgArea Bed
J
(393)
BedPossfox
Bedljjss iIOKnsireaini,nijc
AvgArea
PPB Bed
le-3
toxicant bedloss from current segment ((J,g/m -d);
bedloss over one of the downstream links (g/d);
average area of the segment (m );
toxicant concentration in the current segment ([j,g/kg)
units conversion (kg/g)
8.11 Chemicals in Pore Waters
When the complex multi-layer sediment model is included, pore waters may contain toxic
organic chemicals. Chemicals in pore waters are separated into those that are freely dissolved
and those that are complexed to dissolved organic carbon within the pore waters.
dPo^CiCafTtpreelyDissolvedP.W. s ' -1 ¦ T 1 ¦ . r \ -rr i i \ -rr i \
= GainPoxUp - LossloxUp ± DiffDown ± DiffUp + Decomp
- GillUptake - Microbial - Sorption + Desorption + Depuration
281
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
where:
Toxicant FreefyDissalvedP. W.
GainToxjjp
LossToxjjp
Diffup, DiffDown
Decomp
GillUptake
Depuration
Microbial
Sorption, Desorption
change in concentration of pore water in the sediment bed
normalized per unit area (ng/L^ d);
active layer only: gain of toxicant due to pore water gain from
the water column (jig/L^'d), see (394);
active layer only: loss of toxicant due to pore water loss to the
water column (jig/L^ • d), see (395);
diffusion over upper or lower boundary (|j,g/Lpwd), see (256);
freely dissolved toxicant gain due to microbial decomposition
of organic matter (|j,g/Lpwd), see (159);
active layer only: uptake of toxicant into organisms that
reside at least partially in the sediment (|j,g/Lpwd) (365);
active layer only: excretion of toxicant by organisms that
reside at least partially in the sediment (|j,g/Lpw d), (362);
loss of toxicant in pore waters due to microbial degradation
(Hg/LPw d) see (326);
sorption to and desorption from organic matter and inorganic
matter in the current layer (|j,g/Lpw d). (350), (351)
GainToxUp =
GainUp • AreaSedLayer • ConcToxWateiCol • le3
Volume
(395)
PoreWater
LossToxUp =
LossUp ¦ AreaSedLayer • ConcToxPoreWater • le3
Volume
(396)
PoreWater
where:
GainToxjjp
LossToxup
Gaini p Loss
Up
A rea^edLayer
(lone 7 'ox \ jeLiia
VGlume poreWater
le3
gain of toxicant in pore water from the water col. (|j,g/Lpw d);
loss of toxicant in pore water to the water column above
(Hg/Lpwd);
gain or loss of pore water from the water column above (m3/m2 d);
see (252), (251);
sediment layer area (m2);
concentration of toxicant in relevant media (|_ig/L);
pore water volume (Lpw);
units conversion (L/m ).
Chemicals also sorb to dissolved organic matter within pore waters:
282
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
rlTnxirrmi
= GainDOMToxUp - LossDOMToxUp ± DiffDown ± DiffUp
- (Decomp ¦ PPB • le - 6) - Microbial - Sorption + Desorption
where:
GainDOMTox uP
LossDOMTox i;r
Dijfup, DiffDown
Decomp
Microbial
Sorption
Desorption
gain of toxicant sorbed to DOM from the water column
(Hg/Lpw d) see (394);
loss of toxicant sorbed to DOM in pore water to the water column
above (|j,g/LpW d) see (395);
diffusion over upper or lower boundary (|j,g/LpWd), see (256);
Decomposition of DOM (|j,g/LpWd), see (159);
loss of toxicant sorbed to DOM due to microbial degradation
(Hg/Lpw d) see (326);
sorption to DOM (|j,g/LpW d). (350)
desorption from DOM (|j,g/LpW d). (351)
8.12 Mass Balance Capabilities and Testing
A chemical mass balance testing capability was added to the code during the development of the
estuarine version of AQUATOX. This capability ensured that all linkages between stratified
layers were properly developed with no loss of mass balance. New PFA (perfluorinated acid)
formulations were also tested for mass balance with this capability. Current testing indicates that
AQUATOX balances chemical mass to machine accuracy.
The chemical mass balance testing comprehensively tracks the mass of all chemical loadings and
losses to the system. Chemical mass balance is explicitly tested with this capability; mass
balance of state variables containing chemicals is implicitly tested. The Chemical MBTest output
variable keeps track of all chemical by the following equation:
MBTest = Chemical Mass + Chemical Loss - Chemical Load - Net Layer Exchange (398)
In this manner, the MBTest will stay constant (within machine accuracy) throughout a simulation
if mass balance is being maintained. However, the chemical mass balance function does not
work if the "Keep Freely Dissolved Contaminant Constant" option is selected within the setup
screen.
The chemical mass balance capability also provides a chemical tracking capability that allows
the user to see exactly what is happening to the chemical within the system. Chemical fate may
be tracked using the following output categories (all units are in kilograms):
Chem. MBTest: Mass balance test as described above, see (397).
Chem. Mass: Total chemical mass in the system including chemicals within
283
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
biota.
Chem. Loss + Mass: Chemical loss plus chemical mass in the system.
Chem. Tot Wash:
Chem. WashH20:
Chem. WashAnim:
Chem. WashDetr:
Chem. WashPlnt:
Chem. Tot Loss:
Chem. Hydrol:
Chem. Photol:
Chem. Volatil:
Chem. MicrobMet:
Chem. BioTrans:
Chem. Emergel:
Chem. Fishing Loss:
Washout of chemical from the system since the simulation start.
The sum of the below four categories:
Washout of chemical dissolved in water
Washout of chemical in drifting animals.
Washout of chemical in suspended & dissolved detritus.
Washout of chemical in plants
Total loss of chemical from the system since the simulation start.
The sum of the following eight categories plus washout:
Chemical loss due to hydrolysis.
Chemical loss due to photolysis.
Chemical loss due to volatilization.
Chemical loss due to microbial metabolism.
Chemical loss due to biotransformation.
Chemical loss due to the emergence of insects.
Chemical loss due to fishing.
Chem. Tot Load:
Chem. H20 Load:
Chem. Detr Load:
Chem. Biota Load:
Total loading of chemical into the system since the simulation
start. The sum of the following three categories:
Load of chemical directly into water.
Load of chemical within detritus loadings.
Load of chemical within plant and animal loadings.
Net LayerExch:
Chem. Net Sink:
Chem. Net Entrain:
Chem. Net TurbDiff:
Chem. Net Migrate:
Chem. Delta Thick:
Net of layer exchange between the other layer in the system. The
sum of the below five categories:
Net sinking from upper to lower layer.
Net entrainment of chemical.
Net turbulent diffusion of chemical.
Net migration of chemical in animals.
Chemical movement due to changes in the thickness of the two
layers.
284
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
8.13 Perfluoroalkylated Surfactants Submodel
As mentioned in the introduction (section 1.5), the perfluorinated compounds of interest as
bioaccumulators are the perfluorinated acids (PFAs). Perfluoroctane sulfonate (PFOS) belongs
to the sulfonate group and perfluorooctanoic acid (PFOA) belongs to the carboxylate group. Due
to their use in industrial manufacturing, these persistent chemicals are found in humans, fish,
birds, and marine and terrestrial mammals throughout the world. PFOS has an especially high
bioconcentration factor in fish.
Sorption
Perfluorinated surfactants are quite different from hydrocarbon surfactants. The nonpolar
perfluorocarbon tail repels both water and oil, and the perfluorinated surfactants are much more
active than their hydrocarbon counterparts (Moody and Field 2000). A field is provided for the
user to input a value for the organic matter partition coefficient ("Kom for Sediments"); this
empirical approach was taken in lieu of sufficient theory to support a mechanistic formulation.
Sorption to algae and macrophytes are also modeled empirically ("BCF for Algae" and "BCF for
Macrophytes" parameters).
Biotransformation and Other Fate Processes
PFOS and other related chemicals are anionic surfactants and, as such, they are not subject to
volatilization. However, the worldwide detection of PFOS suggests that there are one or more
precursors that are volatile. Therefore, a fate model for these compounds would not be complete
if it were not able to represent the movement and transformation of significant precursors to
PFOS and other bioaccumulative fluorinated organics. In particular, some fluorinated
compounds are subject to biodegradation of the nonfluorinated portion (Key et al. 1998, Moody
and Field 2000, Giesy and Kannan 2001); these can yield both volatile and nonvolatile
biotransformation products (Key et al. 1998). For example, A'-EtFOSE alcohol is subject to
microbial degradation, yielding 92% PFOS and 8% PFOA (Lange 2000 cited in Cahill et al.
2003). AQUATOX has the capability of representing biotransformation from one congener or
homolog to one or more others when there are sufficient data to parameterize that part of the
model.
Bioaccumulation
PFOS and PFOA and similar compounds bioaccumulate differently than PCBs and chlorinated
pesticides (Kannan et al. 2001). The perfluorinated compounds of interest as bioaccumulators
are the acids. At least for PFOS the salts dissociate instantaneously at neutral pH (OECD 2002).
Perfluorinated acids (PFAs) are oil repelling and are taken up by protein rather than lipids
(Kannan quoted in Scientific American, March, 2001). Therefore, their kinetics cannot be
modeled as functions of the octanol-water partition coefficient. Instead, relationships based on
perfluoroalkyl chain length (Martin et al. 2003a) are used.
285
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
Gill Uptake
Data on PFAs were insufficient at the time this submodel was first developed (2005) to
determine withdrawal efficiencies and explicitly include respiration such as is done for other
organic compounds simulated by AQUATOX. Based on the data of (Martin et al. 2003a), the
uptake rate for all but the longest chain-length carboxylates can be represented as:
kl = SizeCorr-10-512U+01164-chamLength (399)
where
kl = uptake transfer rate (L/kg d);
ChainLength = length of perfluoroalkyl chain (integer).
If chain length exceeds 11, the value for 11 is used. These data were based on 5-g trout, and
uptake is implicitly a function of respiration, which is sensitive to size. A size correction is
based on a standard allometric relationship and the reciprocal of that value for a 5-g fish:
SizeCorr = MeanWeight"" (400)
Sizeref
where
SizeCorr = allometric correction for size (unitless);
MeanWeight = mean wet weight of organism (g);
RB = allometric exponent for respiration (unitless);
SizeRef = reference value (0.7248).
The respiration rate decreases with larger sizes. The allometric exponent RB is assigned values
based on the Wisconsin Bioenergetics Model (Hewett and Johnson 1992). If RB = -0.2 then the
correction for a 10-g fish is 0.63, that is, uptake is 63% that of the fish for which the kl values
were determined; the correction for a 100-g fish is 55% of the reference; and the correction for a
1000-g fish is 35%. For invertebrates RB is assigned a value of -0.25 (Moloney and Field 1989).
Although there are only two data points for sulfonates (Martin et al. 2003a), the trend defined by
those points provides an approximation:
kl = SizeCorr ¦lO-6 00+0-96&cha"lLength (401)
However, the kl values were determined from the first few observed uptake values and not the
observations just before the depuration phase of the experiment. Adjusting the intercept actually
provides a better fit to the overall experiment:
kl = SizeCorr ¦ 10 "
%5+0.966-ChainLength (402)
286
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
Uptake rates (kl values) must be estimated and entered in the "chemical toxicity" screen and
estimates can be modified with user-supplied values if appropriate.
With the greater intercept, the sulfonates are taken up more rapidly than the carboxylates, as
shown in Figure 154. Gill uptake is calculated as:
GillUptake = ToxicantWater ¦ kl ¦ StVarAnjmal ¦ \e - 6
(403)
where:
GillUptake
WetToDry
SizeCorr
Toxicant Water
kl
Sft a\ Animal
1 e-6
uptake of toxicant by gills (ng/L d);
conversion factor for wet to dry weights (5);
allometric correction for size (unitless), see (400)(400);
concentration of toxicant in water ((J,g/L);
uptake transfer rate (L/kg d);
biomass of given animal (mg/L);
units conversion (kg/mg).
Figure 154: Predicted and observed uptake transfer rates for carboxylates and sulfonates.
8 10 12
Perfluoroalkyl Chain Length
Pred carboxylate
¦ Obs carboxylate
Pred sulfonate
x Obs sulfonate
Dietary Assimilation
Martin et al. (2003b) found that assimilation of PFAs was quite efficient, exceeding that for the
normal hydrophobic chemicals. However, many of the calculated values reported (Martin et al.
2003b) exceeded 1.0, so the observed assimilation efficiencies were normalized to a maximum
of 1.0, and equations were derived for uptake from the gut (GutEffTox). If a carboxylate:
log GutEff = - 0.91 + 0.085 • ChainLength r1 = 0.897 (404)
If a sulfonate:
287
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
GutEff = - 0.68 + 0.21 • ChainLength r2 =1.0 (2 points) (405)
In the absence of information on other organisms, these equations are used for all animals.
Figure 155: Gut assimilation efficiency as a function of chain length.
1.20
>1
1.00
o
c
a>
o
F
0.80
m
c
o
0.60
4--
I
0.40
v>
v>
<
0.20
0.00
¦Pred carboxylate
¦Pred/Obs sulfonate
Normalized
carboxylate
7 9 11
Perfluoroalkyl Chain Length
13
Depuration
Based on regression of published data from experiments with juvenile trout (Martin et al. 2003a,
Martin et al. 2003b), carboxylate depuration can be estimated as:
k2 = SizeCorr ¦ 1 cr0 0873 " 0 1207'chainLensth r2 = 0.98 (406)
where:
k2 = depuration rate (1/d).
SizeCorr = allometric correction for size (unitless), see (400);
Only four data points are available for two sulfonate compounds (Martin et al. 2003a, Martin et
al. 2003b); but they indicate that depuration is much slower than for carboxylates. The model
extrapolates from those two pairs of points, but this estimation procedure should be used with
caution (Figure 156):
k2 = ^Co/r-lO-0 733 - 007 ™6"^ r2 = 0.84 (407)
where:
k2 = depuration rate (1/d).
SizeCorr = allometric correction for size (unitless), see (400);
Because uptake is so efficient in the gut, depuration may be largely across the gills. If this is true
288
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
then depuration rate can be related to respiration rate, providing a correction for size.
In the absence of any data, this approach to modeling depuration is extended to invertebrates.
When data become available on depuration of PFAs in invertebrates, this series of constructs
may be modified. Depuration rates (k2s) must be estimated and entered in the "chemical
toxicity" screen and estimates can be modified with user-supplied values if appropriate.
Figure 156. Depuration rate as a function of perfluoroalkyl chain length.
¦ Obs Caboxylate
— Pred Carboxylate
Obs Sulfonate
Pi ed Sulfonate
5 7 9 11 13
Perfluoroalkyl Chain Length
0.12
1
0.08
0.06
0.04
0.02
Available data indicate that concentrations of PFOS in wildlife are less than those known to
cause toxic effects in laboratory animals (Giesy and Kannan 2001). AQUATOX provides a
means of factoring in toxicity data as they become available for aquatic species.
Bioconcentration Factors
The steady-state bioconcentration factor (BCF) for carboxylates, used to compute time-
dependent toxicity, can be estimated by (Martin et al. 2003a):
log BCF = - 5.724 + 0.9146 • ChciinLength r2 = 0.995 (408)
where
BCF = bioconcentration factor (L/kg).
Similar to uptake, the slope for the BCFs of sulfonates closely parallels that of carboxylates but
with a different intercept (Figure 157):
log BCFsulfonate = - 5,195 + 1,03 • ChainLength r2 = 1.0 (2 points) (409)
289
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
For compounds with perfluoroalkyl chain lengths in excess of 11, it is assumed that the BCF is
the same as that for chain length 11, as suggested by the outlier (Figure 157).
Figure 157. Bioconcentration factors as functions of perfluoroalkyl chain length.
Obs Carboxylate
a Obs Sulfonate
Perfluoroalkyl Chain Length
8.14 Aggregation of Organic Chemicals
When modeling entire classes of organic chemicals (e.g. Total PCBs), it is often advantageous to
break up the classes into individual compounds or bins (binning individual analytes by
octonol/water partition coefficient or Kow for example). In this manner, the bioaccumulation
and effects of each portion of the chemical class can be governed by its unique chemical
properties. To enable this process, the modeling of up to 20 individual compounds (or Kow bins)
has always been a capability since AQUATOX 3.0.
However, this type of modeling can create a mismatch when comparing model results to data. If
data are collected by chemical class (e.g. TPCB), individual bins must be summed together
before performing comparisons. This was always possible to do within AQUATOX but it
required a time-consuming export of data into Excel to perform these calculations and
comparisons.
Additionally, chemical toxicity data can sometimes be expressed on a single aggregative class
basis rather than an individual analyte basis (e.g. a site-specific LC50 for TPCB). Unless the
modeled bins are aggregated within the model, these types of toxicity data cannot be used.
To better support the modeling of complex groupings of organic chemicals, AQUATOX Release
3.2 has the capability to model one chemical compartment as an aggregated combination of the
other compartments. To trigger this capacity, an organic toxicant must be added to the "Tl"
compartment and the checkbox in the Setup window under Toxicant Modeling Options that
reads "Tl is an aggregate of all other toxicants in study" must be checked.
290
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 8
When this occurs, the following equations become relevant for the toxicant in water and biota
20
T\water
Water
(408b)
i=2
20
TXorganism ^ ^Organism
1=2
(408c)
where:
Tir
concentration of chemical i in carrier, ug/L or PPB;
Derivatives for the chemical in T1 become irrelevant as it is set as a function of the derivatives of
all of its individual bins (T2 to T20).
Chemical toxicity data may be entered for this aggregated chemical compartment. When this
occurs, though, chemical toxicity parameters must be left as blank for the individual analytes or
double counting of toxic effects will occur. The choice of whether to use aggregated chemical
toxicity data or analyte-specific toxicity data may be made on an organism-by-organism basis.
291
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 9
9. ECOTOXICOLOGY
Unlike most ecological models, AQUATOX contains an
ecotoxicology submodel that computes both lethal and sublethal
acute toxic effects from the concentration of a toxicant in a
given organism. Furthermore, because AQUATOX is an
ecosystem model, it can simulate indirect effects such as loss of
forage base, reduction in predation, and anoxia due to
decomposition following a fish kill.
User-supplied values for LC50, the concentration of a toxicant
in water that causes 50% mortality, form the basis for a
sequence of computations that lead to estimates of the biomass
of a given organism lost through lethal toxicity each day. The
sequence, which is documented in this chapter, is to compute:
• the internal concentration causing 50% mortality for a
given period of exposure;
• the internal concentration causing 50% mortality after an
infinite period of time based on an asymptotic concentration-response relationship;
• the time-varying lethal internal concentration of a chemical;
• the cumulative mortality for a given internal concentration;
• the biomass lost per day as an increment to the cumulative mortality.
The user-supplied EC50s, the concentrations in water eliciting sublethal toxicity responses in
50% of the population, are used to obtain factors relating the sublethal toxicities to the lethal
toxicity. Because AQUATOX can simulate as many as twenty toxic organic chemicals
simultaneously, the simplifying assumption is made that the toxic effects are additive.
9.1 Lethal Toxicity of Compounds
Interspecies Correlation Estimates (ICE)
Often LC50 data will only be available for one or two of the many species that a user wishes to
include in a simulation. To alleviate this problem, a substantial database of regressions
(Interspecies Correlation Estimation, ICE) is available as developed by the US. EPA Office of
Research and Development, the University of Missouri-Columbia, and the US Geological
Survey (Asfaw and Mayer, 2003). At this time the Web-ICE database has over 2000 regressions
with over 100 aquatic species as "surrogates" (Raimondo et al. 2007). Regressions may be made
on the basis of species, families, or genera. The database also includes goodness of fit
information for regressions so their suitability for a given application may be ascertained. Only
statistically significant regressions are included in the database.
Using the ICE database and the following regression equation, the model can be parameterized
to represent a complete food web.
Ecotoxicology: Simplifying
Assumptions
• Toxic effects of multiple chemicals
are additive
• Sublethal effects levels of
chemicals may be estimated as a
fraction of lethal effects levels
• Regressions from one species to
another are available regardless of
the mode of action
• Hie external toxicity model
assumes immediate toxic effect to a
level of external exposure
• Cumulative toxicity considers
differing tolerances in a population,
but ignores inherited tolerance
• Resistance to lower doses is
conferred for the lifetime of an
animal and for one year for a plant.
292
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 9
Log LC50
Estimated
Intercept + Slope ¦ Log LC50(
Observed
(410)
where:
LC50 Estimated
estimated LC50 ((J,g/L);
intercept for regression ((J,g/L);
slope of the regression equation;
observed LC50 ((J,g/L).
Intercept
Slope
L C5 0 Observed
The ICE database is integrated into the AQUATOX user interface. A link is provided to the
Web-based (Web-ICE) site so that the user can alternatively use the web tool. The steps that a
user can take to use ICE within AQUATOX to estimate unavailable LC50 data are as follows:
• Invoke the ICE interface from the AQUATOX "Chemical Toxicity Parameter" screen;
• Choose from the six available ICE databases (species, genus, and family by either
scientific names or common names);
• Either choose a "surrogate species" that matches a species for which there is observed
LC50 data, or start with a "predicted species" that matches a species that you wish to
model;
• The list box that you did not select from in the previous step will narrow to reflect the
available surrogate or predicted species that match with your selection. Select a choice
from this list box as well. If you wish to start over again, you may select the "show all"
button next to this list box.
• Examine the goodness of fit for your model and evaluate whether it is appropriate for
your purposes. Where there are multiple surrogates for the desired predicted species,
compare the statistics and choose best surrogate/predicted pair;
• Apply the model by assigning the surrogate and predicted species to species within the
chemical's toxicity record.
Experimentally derived toxicity data for individual species should be used when available.
However, ICE may then be used to estimate toxicity for species that have not yet been studied
given a particular chemical. There are uncertainties in this estimation procedure, but the model
helps to track these uncertainties. When the ICE model is invoked, data about the goodness of fit
and confidence interval are copied back into the "LC50 comment" field. Overall model
uncertainty resulting from this estimation can then be numerically quantified—these goodness of
fit data can be utilized within an iterative AQUATOX uncertainty analysis (see section 2.5).
293
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 9
Internal Calculations
Toxicity is based on the internal concentration of the toxicant in the specified organism. Many
compounds, especially those with higher octanol-water partition coefficients, take appreciable
time to accumulate in the tissue. Therefore, length of exposure is critical in determining toxicity.
The same principles apply to organic toxicants and to both plants and animals.
The internal lethal concentration for a given period of exposure can be computed from reported
lethal toxicity data based on the simple relationship suggested by an algorithm in the FGETS
model (Suarez and Barber, 1992):
InternalLCSO = BCF ¦ LC50 (411)
where:
InternalLC50 = internal concentration that causes 50% mortality;
BCF = bioconcentration factor (L/kg), see (342) to (349); and
LC50 = concentration of toxicant in water that causes 50% mortality ((J-g/L).
For compounds with a LogKOWin excess of 5 the usual 96-hr toxicity exposure does not reach
steady state, so a time-dependent BCF is used to account for the actual internal concentration at
the end of the toxicity determination. This is applicable no matter what the length of exposure
(Figure 158, based on Figure 144).
Figure 158: Bioconcentration factor as a function of time and KOW
The internal concentration causing 50% mortality after an infinite period of exposure, LCInfinite,
can be computed by:
LCInfinite = InternalLC50 • (/ - (412)
where:
k2 = elimination rate constant (1/d); and
294
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 9
ObsTElapsed = exposure time in toxicity determination (h).
Essentially this equation determines the asymptotic toxicity relationship and provides the model
with a constant toxicity parameter for a given compound.
The model estimates k2, see (364) and (354), assuming that this k2 is the same as that measured
in bioconcentration tests; good agreement has been reported between the two (Mackay et al.,
1992). The user may then override that estimate by entering an observed value. The k2 can be
calculated off-line based on the observed half-life:
k2 = 01692_ (413)
11/2
where:
t'/2 = observed half-life.
Based on the Mancini (1983) model, the lethal internal concentration of a toxicant for a given
exposure period can be expressed as (Crommentuijn et al. (1994):
LethalConc = LCIf"Ue (414)
j k2 • TElapsed
where:
LethalConc = tissue-based concentration of toxicant that causes 50% mortality
(ppb or (J-g/kg);
LCInfinite = ultimate internal lethal toxicant concentration after an infinitely
long exposure time (ppb);
TElapsed = period of exposure (d).
The longer the exposure the lower the internal concentration required for lethality.
Exposure is limited to the lifetime of the organism:
if TElapsed > LifeSpan then TElapsed = LifeSpan (415)
where:
LifeSpan = user-defined mean lifetime for given organism (d).
Based on an estimate of time to reach equilibrium (Connell and Hawker, 1988),
if TElapsed > then
F k2 (416)
LethalConc = LCInfinite
The fraction killed by a given internal concentration of toxicant is best estimated using the time-
dependent LethalConc in the cumulative form of the Weibull distribution (Mackay et al., 1992;
see also Christensen and Nyholm, 1984):
295
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 9
( PPB Shape
CumFracKilled = 1 - gi LethalConc J (417)
where:
CumFracKilled = fraction of organisms killed per day (g/g d),
PPB = internal concentration of toxicant (pg/kg). see (310); and
Shape = parameter expressing variability in toxic response (unitless).
As a practical matter, if CumFracKilled exceeds 95%, then it is set to 100% to avoid complex
computations with small numbers. By setting organismal loadings to very small numbers, seed
values can be maintained in the simulation.
This formulation is preferable to the empirical probit and logit equations because it is simple and
yet based on mechanistic relationships. The Shape parameter is important because it controls the
spread of mortality. The larger the value, the greater the distribution of mortality over toxicant
concentrations and time. Mackay et al. (1992) found that a value of 0.33 gave the best fit to data
on toxicity of 21 narcotic chemicals to fathead minnows. This value is used as a default in
AQUATOX, but it can be changed by the user. Although mercury is not currently modeled, data
on MeHg toxicity shows that the Shape parameter may take a value less than 0.1 (Figure 154).
Hgppb : LetfiaConcHg for Day 637
Figure 159. The effect of Shape in fitting the observed (McKim et al., 1976)
cumulative fraction killed following continued exposure to MeHg
1000 1200 1400 1600
q0.3
uj
_i
£ 0.6
z
O
530.4
<
a:
^ 0.2
0
0 200 400 600 800
DAYS
The biomass killed per day is computed by disaggregating the cumulative mortality. Think of
the biomass at any given time as consisting of two types: biomass that has already been exposed
to the toxicant previously, which is called Resistant because it represents the fraction that was
not killed; and new biomass that has formed through growth, reproduction, and migration and
has not been exposed to a given level of toxicant and therefore is referred to as N01/resistant.
Then think of the cumulative distribution as being the total CumFracKilled, which includes the
FracKiiled that is in excess of the cumulative amount on the previous day if the internal
296
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 9
concentration of toxicant increases. A conservative estimate of the biomass killed at a given
timeis computed as:
Poisoned = Resistant ¦ Biomass ¦ FracKilled + Nonresistant ¦ CumFracKill (418)
where:
Poisoned
Resistant
FracKilled
Nonresistant
biomass of given organisms killed by exposure to toxicant at given
time (g/m3 d);
fraction of biomass not killed by previous exposure (frac);
fraction killed per day in excess of the previous fraction (g/g d);
biomass not previously exposed; the biomass in excess of the
resistant biomass (g/m3) = (1-Resistant) Biomass.
New biomass is considered vulnerable, ignoring the possibility of inherited tolerance. It is
assumed for purposes of risk analysis that resistance is not conferred for an indefinite period. In
animals elapsed exposure time is capped at the average life span, which is a parameter in the
animal record. However, it is assumed that resistance persists in the population until the end of
the growing season. Macrophytes can live for an entire growing season, and algae usually
reproduce asexually as long as conditions are favorable. However, winter die-back does occur in
most macrophytes, and many algae will switch to sexual reproduction under unfavorable
conditions, especially triggered by light and temperature. As a simplifying assumption for both
animals and plants, in the northern hemisphere January 1 is taken as being the date at which
exposure and resistance are reset; in the southern hemisphere (denoted by negative latitude in the
site record) July 1 is the reset date. On this date, the variables Resistant, FracKilledPrevious, and
TElapsed are all set to zero.
9.2 Sublethal Toxicity
Organisms usually have adverse reactions to toxicants at levels significantly below those that
cause death. In fact, the lethal to sublethal ratio is commonly used to quantify this relationship.
The user supplies observed EC50 values, which can then be used to compute AFs (application
factors). For example:
where:
EC50Growth
AFGrowth
LC50
AFGrowth
EC50Growth
LC50
(419)
external concentration of toxicant at which there is a 50%
reduction in growth ((J,g/L);
sublethal to lethal ratio for growth (unitless); and
external concentration of toxicant at which 50% of population is
killed ((J-g/L).
If the user enters an observed EC50 value, the model provides the option of applying the
resulting AF to estimate EC50s for other organisms. The computations for AFPhoto and
AFRepro are similar:
297
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 9
AFPhoto
EC50Photo
LC50
(420)
AFRepro =
EC50Repro
LC50
(421)
where:
EC50Photo
AFPhoto
EC50Repro
AFRepro
external concentration of toxicant at which there is a 50%
reduction in photosynthesis ((J-g/L);
sublethal to lethal ratio for photosynthesis (unitless);
external concentration of toxicant at which there is a 50%
reduction in reproduction ((J,g/L); and
sublethal to lethal ratio for reproduction (unitless).
Because of the nature of these application factors, sublethal effects cannot be calculated (using
internal calculations) unless LC50 parameters are included in the model.
Similar to computation of lethal toxicity in the model, sublethal toxicity is based on internal
concentrations of a toxicant. Often sublethal effects form a continuum with lethal effects and the
difference is merely one of degree (Mackay et al., 1992). Regardless of whether or not the mode
of action is the same, the computed factors relate the observed effect to the lethal effect and
permit efficient computation of sublethal effects factors in conjunction with computation of
lethal effects. Because AQUATOX simulates biomass, no distinction is made between reduction
in a process in an individual and the fraction of the population exhibiting that response. The
commonly measured reduction in photosynthesis is a good example: the data only indicate that a
given reduction takes place at a given concentration, not whether all individuals are affected.
The factor enters into the Weibull equation to estimate reduction factors for photosynthesis,
growth, and reproduction:
PPB
FracPhotO — e\ LethalConc ¦ AFPhoto
1/Shape
(422)
RedGrOWth 1 - p I LethalConc ¦ AFGrowth
PPB
1/Shape
(423)
where:
RedRepro — 1 - e
PPB
\ 1/Shape
LethalConc ¦ AFRepro
(424)
FracPhoto = reduction factor for effect of toxicant on photosynthesis (unitless);
RedGrowth = factor for reduced growth in animals (unitless);
RedRepro = factor for reduced reproduction in animals (unitless);
PPB = internal concentration of toxicant ([j,g/kg), see (310);
LethalConc = tissue-based conc. of toxicant that causes mortality ([j,g/kg), see (413);
298
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 9
= sublethal to lethal ratio for photosynthesis (unitless, default of 0.10);
= sublethal to lethal ratio for growth in animals (unitless, default of 0.10);
= sublethal to lethal ratio for reproduction in animals (unitless, default of
0.05);
= parameter expressing variability in toxic response (unitless, default of
0.33).
The reduction factor for photosynthesis, FracPhoto, enters into the photosynthesis equation (Eq.
(35)) and it also appears in the equation for the acceleration of sinking of phytoplankton due to
stress (Eq. (69)).
The variable for reduced growth, RedGrowth, is arbitrarily split between two processes,
ingestion (Eq. (91)), where it reduces consumption by 20%:
ToxReduction = 1-(0.2 ¦ RedGrowth) (425)
and defecation (Eq. (97)), where it increases the amount of food that is not assimilated by 80%:
IncrEgest = (1 - EgestCoeff prify pred) ¦0.8¦ RedGrow th (426)
These have indirect effects on the rest of the ecosystem through reduced predation and increased
production of detritus in the form of feces.
Embryos are often more sensitive to toxicants, although reproductive failure may occur for
various reasons. As a simplification, the factor for reduced reproduction, RedRepro, is used
only to increase gamete mortality (Eq. (126)) beyond what would occur otherwise:
IncrMort = (1 - GMort) ¦ RedRepro (427)
By modeling sublethal and lethal effects, AQUATOX makes the link between chemical fate and
the functioning of the aquatic ecosystem- a pioneering approach that has been refined over the
past twenty years, following the first publications (Park et al., 1988; Park, 1990).
Sloughing of periphyton and drift of invertebrates also can be elicited by toxicants. For example,
sloughing can be caused by a surfactant that disrupts the adhesion of the periphyton, or an
invertebrate may release its hold on the substrate when irritated by a toxicant. Often the response
is immediate so that these responses can be modeled as dependent on dissolved concentrations of
toxicants with an available sublethal toxicity parameter, as in the equation for periphyton
sloughing:
Toxicant
DislodgePen Tox = MaxToxSlough ¦ — ^r— BiomassPen (428)
ToxicantWater + EC 50 Dlslodge
where:
-3
Dislodgeperi, Tox = periphyton sloughing due to given toxicant (g/m d);
AFPhoto
AFGrowth
AFRepro
Shape
299
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 9
MaxToxSlough =
Toxicant Water =
EC50 Dislodge
Biomass
Peri
maximum fraction of periphyton biomass lost by sloughing due to
given toxicant (fraction/d, 0.1);
concentration of toxicant dissolved in water ((J,g/L); see (300);
external concentration of toxicant at which there is 50% sloughing
(Hg/L); and
biomass of given periphyton (g/m3); see (33).
Likewise, drift is greatly increased when zoobenthos are subjected to stress by sublethal doses of
toxic chemicals (Muirhead-Thomson, 1987), and that is represented by a saturation-kinetic
formulation that utilizes an analogous sublethal toxicity parameter :
Dislodge Tox =
ToxicantWater - DriftThreshold
ToxicantWater - DriftThreshold + EC50
(429)
Growth
where:
Toxi can t Water ~
DriftThreshold =
EC50 Growth ~
concentration of toxicant in water ((J,g/L);
the concentration of toxicant that initiates drift ((J-g/L); and
concentration at which half the population is affected ((J-g/L).
These terms are incorporated in the respective periphyton washout (72) and zoobenthos drift
(130) equations.
9.3 External Toxicity
Chemicals that are taken up very rapidly and those that have an external mode of toxicity, such
as affecting the gills directly, are best simulated with an external toxicity construct. AQUATOX
has an alternative computation for CumFracKilled, when calculating toxic effects based on
external concentrations, using the two-parameter Weibull distribution as in Christiensen and
Nyholm (1984):
CumFracKilled = 1 - exp(-kzEta)
(430)
where:
z = external concentration of toxicant (|_ig/L);
CumFracKilled = cumulative fraction of organisms killed for a given period of exposure
(fraction/d), applied to equation (417);
k and Eta = fitted parameters describing the dose response curve.
Rather than require the user to fit toxicological bioassay data to determine the parameters for k
and Eta, these parameters are derived to fit the LC50 and the slope of the cumulative mortality
curve at the LC50 (in the manner of the RAMAS Ecotoxicology model, Spencer and Ferson,
1997):
300
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 9
k =
-ln(0.5)
LC50ma
(431)
Eta =
- 2 • LC50 ¦ slope
ln(0.5)
(432)
where: slope = slope of the cumulative mortality curve at LC50 (unitless).
LC50 = concentration where half of individuals are affected (|_ig/L),
AQUATOX can assume that each chemical's dose response curve has a distinct shape, relevant
to all organisms modeled. In this manner, a single "slope factor" parameter describing the shape
of the Weibull curve can be entered in the chemical record rather than requiring the user to
derive slope parameters for each organism modeled. (Note, this is different than the shape
parameter used for internal toxicity.) However, animal and plant-specific slope factors may also
be entered in the animal and plant chemical-toxicity databases. If these values are left blank or
zero, the value from the chemical record is used. Otherwise the organism-specific factor is used.
The units for this factor are the same as those for the chemical underlying data (the slope at
EC50 multiplied by the EC50 in ug/L).
As shown below, the slope of the curve at the LC50 is both a function of the shape of the
Weibull distribution and also the magnitude of the LC50 in question. Figure 160 shows two
Weibull distributions with identical shapes, but with slopes that are significantly different due to
the scales of the x axes.
Figure 160. Weibull distributions with identical shapes, but different slopes.
Weibull Distribution, LC50=1, Slope=1
100%
3
E
3
o
50%
0.5 1 1.5
Concentration
Weibull Distribution, LC50=100, Slope=0.01
100%
o
at
0)
>
3
E
3
o
50%
¦Weibull
-Slope
50 100 150
Concentration
200
301
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 9
For this reason, rather than have a user enter "the slope at LC50" into the chemical record,
AQUATOX asks the user to enter a "slope factor" defined as "the slope at LC50 multiplied by
LC50." In the above example, the user would enter a slope factor of 1.0 and then, given an
LC50 of 1 or an LC50 of 100, the above two curves would be generated.
When modeling toxicity based on external concentrations, organisms are assumed to come to
equilibrium with external concentrations (or the toxicity is assumed to be based on external
effects to the organism).
Unlike the internal model, application factors are not used to estimate sublethal effects when
calculating external toxicity. Therefore, EC50 parameters do not need to be paired with LC50
values to calculate sublethal effects with this model.
302
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 10
10. ESTUARINE SUBMODEL
The estuarine version of AQUATOX is intended to be an
exploratory model for evaluating the possible fate and effects
of toxic chemicals and other pollutants in estuarine ecosystems.
The model is not intended to represent detailed, spatially
varying site-specific conditions, but rather to be used in
representing the potential behavior of chemicals under average
conditions. Therefore, it is best used as a screening-level
model applicable to data-poor evaluations in estuarine
ecosystems. However, it can be calibrated for different
estuaries.
Hourly tidal fluctuations are not included in the model; the
native AQUATOX time-step is one day. Because of this, the
overall water volume of the estuary may be assumed to remain
constant over the entire simulation. The simplifying
assumption is that the water volume of the estuary is not
sensitive to the freshwater inflow. The volumes and depths of
the fresh layer and the salt wedge do vary as a function of the
daily average tidal range and freshwater flows.
If simulation of spatially-explicit, site-specific estuarine
conditions is desired, then a multi-segment model can be
implemented with flow among segments provided by an external hydrodynamic model. The
numerous effects of salinity described below in the context of the estuarine submodel are also
applicable to the classic and multi-segment versions of AQUATOX.
AQUATOX Estuarine Submodel:
Simplifying Assumptions
• Estuary is a single segment that
always has two well-mixed layers
• The estuary has freshwater inflow
from upstream and saltwater inflow
from the seaward end (salt-wedge)
• Water flows at the seaward end are
estimated using the salt-balance
approach
• Effects of salinity on sorption are
minor and are not modeled
• Hourly tidal fluxes are not modeled
• Daily average volume of the
estuary is assumed to remain
constant over time
• The surface area of the lower layer
is the same as the upper layer
• Nutrient concentrations in
inflowing seawater are assumed to
be constant
• Possible salinity effects on
microbial degradation, hydrolysis,
and photolysis are ignored.
10.1 Estuarine Stratification
As a general case, the estuarine system is assumed to always have two layers, although at times
the layers may be essentially identical because of respective thicknesses and turbulent diffusion.
The two layers are assumed to be a function of and to vary with freshwater loadings and daily
tidal ranges. The fraction of depth in the upper layer is adjusted to account for changing volumes
due to entrainment (flow of water from lower to upper layer; see section 10.4), with a value of
1.5 based on inspection of published observations. If ResidFlow > 0 then:
„ , ,, ResidFlow
rreshwaterHead =
Area
(433)
FreshwaterHead
FracUpper = 1.5- —
TidalAmplitude + FreshwaterHead
where:
303
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 10
FreshwaterHead = height of freshwater (m/d);
ResidFlow
Area
FracUpper
Tidal Amplitude
inflow residual flow of fresh water minus daily evaporation,
"3
(m /d) user inputs;
area of the estuary taken at mean tide (m2).
fraction of mean depth that is upper layer (unitless).
tidal amplitude (m), see (434);
If ResidFlow < = 0 then FracUpper is taken as having a nominal value of 0.05.
The thicknesses of the two layers, and therefore the volumes of the two layers, may be calculated
as a function of FracUpper.
where:
ThickUpper
FracLower
ThickLower
MeanDepth
VolumeUpper
VolumeLower
Area
ThickUpper = FracUpper ¦ MeanDepth
ThickLower = MeanDepth - ThickUpper
VolumeUpper = FracUpper ¦ Area
VolumeLower = FracLower ¦ Area
= thickness of the upper layer (m);
= 1 - FracUpper; see (432);
= thickness of the lower layer (m);
= mean depth of the estuary (m);
= volume of the upper layer (m3);
-3
= volume of the lower layer (m );
= area of the estuary taken at mean tide (m2).
(434)
As shown in the formulations above, layer thicknesses are a function of the daily predicted tidal
range. Given that the estuary's average daily volume is assumed to remain constant, to maintain
mass-balance of water AQUATOX moves water from one layer to the next when thicknesses
change. (This same movement of water occurs when the user specifies a variable thermocline
depth in a stratified lake or reservoir, see section 3.4 on "Modeling Reservoirs and Stratification
Options") In order to maintain biomass, nutrient, and toxicant mass-balance AQUATOX also
transfers state variables located in the moving water from one layer to the next. This transfer can
cause minor fluctuations that are visible in some estuarine-version results (e.g. wave-like patterns
in fish biomass predictions.) Such fluctuation is predominantly an artifact of the simple manner
in which AQUATOX models estuarine water volume.
10.2 Tidal Amplitude
Tidal amplitude is calculated using the general equation found in the Manual of Harmonic
Analysis and Prediction of Tides (U.S. Department of Commerce 1994):
Tidal Amplitude = "V
r /—tCon.
r AmpCon ¦ Nodefactorc
Con.,Year
cos((SpeedCon ¦ Hours) + EquilConJear - EpochCon ) I (435)
304
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 10
where:
Tidal Amplitude
Con.
Ampcon.
Nodefactor
Speed
Hours
Equil
Epoch
one-half the range of a constituent tide (m);
eight constituents of tidal range listed below;
user-input amplitude for each constituent (m);
node factor for each constituent for each year, hard-wired into
AQUATOX for 1970-2037 (deg.);
speeds of each constituent in (deg./hour), hard-wired into
AQUATOX for each relevant constituent;
time since the start of the year (hours);
equilibrium argument for each constituent for each year in degrees
for the meridian of Greenwich, hard-wired into AQUATOX for
1970-2037 (deg.);
user input phase lag for each constituent (deg.).
AQUATOX requires Amplitudes and Epochs for the following eight constituents of tidal range
for the modeled esturary, generally available for download from NOAA databases. These
"primary" constituents were found to have the largest effect on tidal range and will predict tidal
range to the precision as required by the estuarine submodel:
M2 - Principal lunar semidiurnal constituent
S2 - Principal solar semidiurnal constituent
N2 - Larger lunar elliptic semidiurnal constituent
K1 - Lunar diurnal constituent
01 - Lunar diurnal constituent
SSA - Solar semiannual constituent
SA - Solar annual constituent
PI - Solar diurnal constituent
10.3 Water Balance
Water balance is computed using the salt balance approach (Ibanez et al. 1999):
Saltwater Inflow -
ResidFlow
SalinityLower/SalinityUppper -1
(436)
Outflow =
ResidFlow
where:
Saltwater Inflow =
Outflow
ResidFlow
1 - SalinityUpper / SalinityLower
(437)
water entering estuary from mouth of estuary, usually into lower
level but may be into upper level if evaporation exceeds freshwater
inflow (m3/d);
water leaving estuary at mouth (m3/d);
residual flow of fresh water; may be negative if evaporation
305
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 10
exceeds freshwater inflow (m /d);
SalinityLower = salinity of lower layer at mouth of estuary (psu or %o);
SalinityUpper = salinity of upper layer at mouth of estuary (psu or %o);
Programmatically, the system is modeled as a single constant-volume segment with two layers
and with freshwater inflow from upstream and saltwater inflow from the seaward end. Ice cover
is not assumed on top of estuaries unless the average water temperature falls below -1.8 deg.C.
10.4 Estuarine Exchange
Saltwater inflow occurs to replace water that is admixed (entrained) from one layer (usually the
lower) to the other layer, producing the observed salinities of the two layers at the mouth of the
estuary. (Note that this use of the term "entrainment" differs from the downstream entrainment
of organisms, e.g. (132).) This circulation is much greater than any longitudinal mixing (see
Thomann and Mueller 1987). Therefore, effectively, entrainment is the equivalent of
Saltwater Inflow^ but its derivation is informative:
,, j Saltwater Inflow
EntrainVel =
Area
VertAdvection = EntrainVel ¦ Thick (438)
,, . VertAdvection ¦ Area
Entrainment =
Thick
where:
EntrainVel = entrainment velocity of lower layer into upper layer (m/d);
VertAdvectiveDisp = vertical advective dispersion (m /d);
Entrainment = vertical flow as derived above (m3/d).
Transport of suspended and dissolved substances from the lower layer to the upper layer can then
be computed. In a truly stratified estuary turbulent diffusion will be minimal, so we will set the
bulk mixing coefficient (BulkMixCoeff) to 0.1 m /d following the example of Koseff et al.
(1993). However, when wind exceeds 3 m/s Langmuir circulation sets up with downwelling and
upwelling extending to about 3 m. Therefore, if the thickness of the upper layer is less than 3 m
and the wind speed is greater than 3 m/s, then bulk mixing is increased by a factor of 5.
Turbulent diffusion can then be computed for each dissolved and suspended compartment:
306
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 10
BulkMixCoeff . . (
TurbDiff = —— Langmuir ¦ [Cone
Volume
— OTIC
compartment, lower compartment .upper )
TurbDifflower = BulkMlxCoeff . Langmuir ¦ {Cone
(439)
Volume,
— C1 one
compartment, upper compartment, lower )
lower
If ThickUpper < 3 and Wind > 3 then Langmuir = 5 else Langmuir = 1
where:
TurbDiff
BulkMixCoeff
Langmuir
Volume upper
Volume lower
Cone
= turbulent diffusion (g/m -d);
= bulk mixing coefficient (0.1 m /d);
= factor for greater mixing when wind equals or exceeds 3 m/s
(unitless);
= volume of the upper layer (m3);
"3
= volume of the lower layer (m );
= concentration of given compartment in a given layer (g/m3).
10.5 Salinity Effects
Mortality and Gamete Loss
Salinity that is less than or greater than threshold values increases mortality and gamete loss:
where:
(440)
if SalMin < Salinity < SalMax then SaltMort = 0
if Salinity < SalMintYim SaltMort = SalCoeffX ¦ QSam"-SaU"'*
if Salinity > SalMax then SaltMort = SalCoeffl ¦ eSalM»-s'UMax
SalMin = minimum salinity below which effect is manifested (%o);
Salinity = ambient salinity (%o);
SalMax = maximum salinity above which effect is manifested (%o);
SaltMort = mortality due to salinity (1/d);
SalCoeffl = coefficient for effect of low salinity (unitless);
SalCoeff2 = coefficient for effect of high salinity (unitless);
e = the base of natural logarithms (2.71828, unitless).
SaltMort is then applied to mortality (112) and gamete loss (126). The model assumes
reproduction is affected because eggs and sperm are not viable in abnormal salinities.
Other Biotic Processes
Salinity beyond the range of tolerance for a particular process, including photosynthesis,
ingestion, and respiration, will reduce the process:
307
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 10
where:
if SalMin < Salinity < SalMax then SalEjfect = 1
if Salinity < SalMin then SalEffect = SalCoeff 1 •
if Salinity > SalMax then SalEffect = SalCoeff 2 ¦ eSaMa^SaIini^
SaltEffect = effect of salinity on given process (unitless).
(441)
In general, the ranges of tolerance of abnormal salinities in animals, going from least tolerant to
most tolerant, affects reproduction, ingestion, respiration, and mortality in that order (Figure
161). Respiration decreases because gill ventilation is depressed. SaltEffect is applied to
ingestion (91), respiration (100), and photosynthesis (35) as appropriate.
Figure 161. Effects of salinity on various animal processes.
c
o
o
3
"D
0)
ft
0)
1.2
1
0.8
0.6
Si 0.4
ra
ft
0.2
Effects of Salinity
--
--
--
1 1 1 1 1 1
V
--
10 20
30 40 50
Salinity (ppt)
60 70
5
4.5
4 _
3.5 5
3 ~
>
2.5 ^
2
1.5
1
0.5
0
80
ra
t:
o
¦Ingestion GameteLoss Respiration
¦Mortality
Sinking
Sinking of phytoplankton and suspended detritus also is affected by salinity, more so than by
temperature (Figure 159, Figure 160). However, because ambient salinity and temperature affect
sinking by controlling density, we will compute a density factor based on the effects of both
compared to the salinity and temperature of the observed sinking rate (Thomann and Mueller
1987):
308
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 10
WaterDensity = 1 + ¦< 10
(28.14 - 0.0735 • Temperature - 0.00469 • Temperature2)
+ (0.802 - 0.002 • Temperature)¦ (Salinity - 35)
(442)
WaterDensity
reference
DensityFactor =
WaterDensity ambmit
If salinity is not included in AQUATOX as a state variable, DensityFactor is set to 1.0.
(443)
where:
WaterDensity reference
WaterDensity ambient
Temperature
DensityFactor
Sink
KSed
Thick
Sink = . DensityFactor
Thick
(444)
density of water at temperature and salinity of observed sinking
rate (kg/L);
density of water at ambient temperature and salinity (kg/L);
temperature of water (°C);
correction factor for water densities other than those at which
sinking rates were observed (unitless);
sinking rate of given suspended compartment (g/m -d);
intrinsic settling rate (m/d);
thickness of water layer (m).
Figure 162. Correction factor for sinking as
a function of temperature.
Effect of Temperature on Sinking
1.01
0.995
0.99
_ 1.005
0 5 10 15 20 25 30 35 40 45
Temperature
Figure 163. Correction factor for sinking as a
function of salinity.
Effect of Salinity on Sinking
1.01
1
i-
0.99
0
0
TO
0.98
Ll
>
0 9/
¦*->
£
0.96
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 10
Sorption
The influence of seawater or "salting out" does not cause major changes in sorption of organic
compounds (Schwarzenbach et al. 1993). It varies with the compound, with greater effect on
polar compounds, but is seldom measured. Therefore, it will be ignored at the present time.
Volatilization
Volatilization is affected by salinity, and can be represented by a linear increase in the Henry's
Law constant (Eqn. 330). At 35%o salinity the average increase in the constant across tested
organic compounds is 1.4 compared to that of distilled water (Schwarzenbach et al. 1993).
Applying this relationship:
HLCSaltFactor = 1 + 0.01143 - Salinity (445)
Estuarine Reaeration
Reaeration is affected by salinity, especially through calculation of the saturation level (02Sat).
Salinity is included in the present formulation for 02Sat. Computation of the depth-averaged
reaeration coefficient (KReaer) requires determination of the effects of both tidal velocity and
wind velocity. Thomann and Fitzpatrick (1982, see also (Chapra 1997)) combine the two in one
equation:
„ ^ JVelocity 0.728 • V Wind -0.317-Wind + 0.0372 -Wind2 ,AA„
KReaer = 3.93 - —- H (446)
Thick3/2 Thick
The daily average tidal velocity can be computed by a variation of a formulation presented by
(Thomann and Mueller, 1987), substituting the spring tide harmonic for the diurnal harmonic:
Velocity =
f
ResidFlowVel + TidalVel ¦
l + O.S-sinf^^
I 12
86400
(447)
ResidFlowVel =
OutFlow
XSecArea
XSecArea = Depth ¦ Width
(448)
(449)
310
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 10
TidalVel =
TidalPrism
XSecArea
(450)
TidalPrism = 2.0- Amplitude ¦ Area
(451)
where:
Velocity =
Wind =
ResidFlowVel =
Outflow =
TidalVel =
Day =
XSecArea =
Depth =
Width =
TidalPrism =
Amplitude =
Area =
water velocity (m/s);
wind velocity (m/s), see (29);
residual flow velocity of fresh water (m/d);
water leaving estuary at mouth (m3/d), see (436);
mean tidal velocity (m/d);
day of year (d);
cross-sectional area of estuary (m );
mean water depth (m);
width of estuary (m);
the difference in water volume between low and high tides (m3);
tidal amplitude (m), see (434);
area of site (m2).
Figure 164. Daily average water velocities based on freshwater flow and tidal flow.
Freshwater and Tidal Velocities
0.25
0.20
15
10
0.05
0.00
0
5
10
15
20
25
30
Days
Salinity can significantly affect atmospheric exchange of carbon dioxide. For saline systems, the
equilibrium parameters of the CO2 system should be obtained using C02SYS (Yuan, 2006) or
C02calc (USGS, 2010) and the results used as inputs for COlEquil in the estuarine AQUATOX
311
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 10
simulation. See section 5.6 for more information about this implementation.
Migration
Fish and pelagic invertebrates will also migrate vertically when the salinity level is not favorable.
Favorable salinity is defined as the range of salinity in which no ingestion effects occur for the
animal (from the minimum to the maximum salinity tolerances for ingestion). If the salinity of
the current segment is outside that range, and the salinity of the other segment is within the range
of favorable salinity, the animal is predicted to migrate vertically to the other segment.
Entrainment for pelagic invertebrates (movement due to water movement from the lower layer to
the upper layer as predicted by the salt balance model, see (437)) will also be set to zero if the
salinity in the upper layer is outside of the favorable range. This can have significant effects on
shrimp populations, for example.
10.6 Nutrient Inputs to Lower Layer
Nutrient concentrations in ocean water flowing into the lower layer are set to temporally constant
levels, the assumption being that the chemical composition of seawater remains relatively
uniform. Nutrients and gasses in seawater may be edited using a button available in the initial
conditions and loadings screen for each relevant variable. The default nutrient and gas
composition of seawater are set as follows:
•
Ammonia:
0.02
mg/L
(Data from Galveston Bay, TX)
•
Nitrate:
0.05
mg/L
(Data from Galveston Bay, TX)
•
Phosphate:
0.03
mg/L
(Data from Galveston Bay, TX)
•
Oxygen:
7.0
mg/L
(Default oxygen inflow to lower segment)
•
C02 :
90.0
mg/L
(Anthoni, 2006)
312
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
11. REFERENCES
Adamack, A. T., C. A. Stow, D. M. Mason, L. P. Rozas, and T. J. Minello. 2012. Predicting the
effects of freshwater diversions on juvenile brown shrimp growth and production: a
Bayesian-based approach. MarEcol Prog Ser Vol. 444:155-173.
Allan, J.D., 1995, Stream Ecology: Structure and Function of Running Waters: London,
Chapman & Hall, 388 p.
Ambrose, R. B., Jr., J. L. Martin, and T. A. Wool. 2006. WASP7 Benthic Algae - Model Theory
and User's Guide, Supplement to Water Quality Analysis Simulation Program (WASP)
User Documentation Page 32. U.S. Environmental Protection Agency, Washington, DC
Ambrose, R. B., Jr., T. A. Wool, J. L. Martin, J. P. Connolly, and R. W. Schanz. 1991.
WASP5.X, A Hydrodynamic and Water Quality Model—Model Theory, User's Manual,
and Programmer's Guide., Environmental Protection Agency, Environmental Research
Laboratory, Athens, Georgia.
Anthoni, Dr. J. Floor, 2006, The Chemical Composition of Seawater, www.seafriends.org.nz/
oceano/seawater.htm
APHA. 1995. Standard methods. 19th Edition. American Public Health Association,
Washington, DC.
Bahr, L.M. and W.P. Lanier. 1981. The ecology of intertidal oyster reefs on the South Atlantic
coast: A community profile. U.S. Fish and Wildlife Service, Biology Service Program
81(15): 1-105
Barber, M. C. 2001. Bioaccumulation and Aquatic System Simulator (BASS) User's Manual,
Test Version 2.1 beta. Pages 76. U.S. Environmental Protection Agency, Athens, GA.
Barber, M. C. 2003. A Review and Comparison of Models for Predicting Dynamic Chemical
Bioconcentration in Fish. Environmental Toxicology and Chemistry, 22(9): 1963-1992.
Barbour, M. T., J. Gerritsen, B. D. Snyder, and J. B. Stribling. 1999. Rapid Bioassessment
Protocols for Use in Streams and Wadeable Rivers: Periphyton, Benthic
Macroinvertebrates and Fish, Second Edition. EPA 841-B-99-002, U.S. Environmental
Protection Agency; Office of Water, Washington, D.C.
Bartell, S. M., G. Lefebvre, G. Kaminski, M. Carreau, and K. R. Campbell. 1999. An Ecological
Model for Assessing Ecological Risks in Quebec Rivers, Lakes, and Reservoirs.
Ecological Modelling 124: 43-67.
Bartell, S. M., R. H. Gardner, and R. V. O'Neill. 1992. Ecological Risk Estimation. Lewis
Publishers, Boca Raton, Florida.
Berry, W., N. Rubinstein, B. Melzian, and B. Hill. 2003. The Biological Effects of Suspended
and Bedded Sediments (SABS) in Aquatic Systems: A Review. Pages 58. U.S.
Environmental Protection Agency, Narragansett, RI.
313
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
Biggs, B. J. F. 1996. Patterns in Benthic Algae of Streams. Pages 31-56 in R. J. Stevenson, M. L.
Bothwell, and R. L. Lowe, eds. Algal Ecology: Freshwater Benthic Ecosystems.
Academic Press, San Diego.
Bowie, G. L., W. B. Mills, D. P. Porcella, C. L. Campbell, J. R. Pagenkopf, G. L. Rupp, K. M.
Johnson, P. W. H. Chan, and S. A. Gherini. 1985. Rates, Constants, and Kinetics
Formulations in Surface Water Quality Modeling. U.S. Environmental Protection
Agency, Athens GA.
Broekhuizen, N., S. Parkyn, and D. Miller. 2001. Fine Sediment Effects on Feeding and Growth
in the Invertebrate Grazers Potamopyrgus antipodarum (Gastropoda, Hydrobiidae) and
Deleatidium sp. (Ephemeroptera, Leptophlebiidae). Hydrobiologia 457: 125-132.
Brylawski, B. J., and T. J. Miller. 2003. Bioenergetic modeling of the blue crab (Callinectes
sapidus) using the Fish Bioenergetics (3.0) computer program. Bulletin Marine Science
72:491-504.
Butler, P.A. 1949. Gametogenesis in the oyster under conditions of depressed salinity. Biol.
Bull. 96(3): 263-269.
Butler, P. A. 1954. Summary of our knowledge of the oyster in the Gulf of Mexico. U.S. Fish
Wildlife Service Bull. 89: 479-489.
Cahill, T. M., I. Cousins, and D. Mackay. 2003. General Fugacity-Based Model to Predict The
Environmental Fate of Multiple Chemical Species. Environ. Toxicology Chemistry 22:
483-493.
Cake, E.W., Jr. 1983. Habitat suitability index models: Gulf of Mexico, American Oyster.
U.S. Department of the Interior, Fish & Wildlife Service, Report FWS/OBS-82/10.57
Calabrese, A. and H.S. Davis. 1970. Tolerances and requirements of bivalve molluscs. Helgol.
Wiss. Meeresunters. 20: 553-564.
Canham, C. D., J. S. Denslow, W. J. Piatt, J. R. Runkle, T. A. Spies, and P. S. White. 1990. Light
regimes beneath closed canopies and tree-fall gaps in temperate and tropical forests.
Canadian Jour. Forest Research 20:620-631.
Carriker, M.R. 1951. Ecological observations on the distribution of oyster larvae in New Jersey
estuaries. Ecol. Monogr. 21(1): 19-38.
CH2M HILL, Eco Modeling, Warren Pinnacle Consulting, and Boise City Public Works. 2008.
Numeric Nutrient Criteria Development for the Lower Boise River Using the
AQUATOX Model. U.S. Environmental Protection Agency, Office of Water,
Washington, DC.
Chapra, S. C. 1997. Surface Water Quality Modeling. McGraw-Hill, New York NY.
Chapra, S. C., G. J. Pelletier, and H. Tao. 2007. QUAL2K: A Modeling Framework for
Simulating River and Stream Water Quality, Version 2.07: Documentation and Users
Manual. Civil and Environmental Engineering Dept., Tufts University, Medford, MA.
Clarke, D. G., and D. H. Wilber. 2000. Assessment of Potential Impacts of Dredging Operations
Due to Sediment Resuspension. Pages 14. U.S. Army Engineer Research and
Development Center, Vicksburg, MS.
314
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
Collins, C. D. 1980. Formulation and Validation of a Mathematical Model of Phytoplankton
Growth. Ecology 6:639-649.
Collins, C. D., and J. H. Wlosinski. 1983. Coefficients for Use in the U.S. Army Corps of
Engineers Reservoir Model, CE-QUAL-R1. Tech. Rept. E-83-, Environmental
Laboratory, U.S. Army Engineer Waterways Experiment Station, Vicksburg, Miss.
Connolly, J. P. 1991. Application of a food chain model to polychlorinated biphenyl
contamination of the lobster and winter flounder food chains in New Bedford Harbor.
Environ. Sci. Technol. 25:760-770.
Crowe, A., and J. Hay. 2004. Effects of Fine Sediment on River Biota. Pages 35. Cawthron
Institute, prepared for Motueka Integrated Catchment Management Program, Nelson,
New Zealand.
Davis, H.C. 1958. Survival and growth of clam and oyster larvae at different salinities. Biol.
Bull. 114(3): 296-307.
Davis, H.C., and P.E. Chanley. 1955. Spawning and egg production of oysters and clams. Biol.
Bull. 110: 117-128.
DeAngelis, D. L., S. M. Bartell, and A. L. Brenkert. 1989. Effects of nutrient cycling and food-
chain length on resilience. American Naturalist 134: 778-805.
Di Toro, D. M. 2001. Sediment Flux Modeling. Wiley-Interscience, New York.
Di Toro, D. M., I. I. Fitzpatrick, and R. V. Thomann. 1983. Water Quality Analysis Simulation
Program (WASP) and Model Verification Program (MVP) - Documentation.
Hydroscience, Inc. for U.S. EPA, Duluth, MN.
Di Toro, D. M., P. Paquin, K. Subburamu, and D. A. Gruber. 1990. Sediment oxygen demoand
model: methane and ammonia oxidation. Journal Environmental Engineering ASCE 116:
945-986.
Doisy, K. E., and C. F. Rabeni. 2004. Effects of Suspended Sediment on Native Missouri Fishes:
A Literature Review and Synthesis. Department of Fisheries and Wildlife, University of
Missouri, Columbia MO.
Donigian, A. S., I. T. Love, I. S. Clough, R. A. Park, I. N. Carleton, P. A. Cocca, and I. C.
Imhoff 2005. Nutrient Criteria Development with a Linked Modeling System: Watershed
and Ecological Model Application and Linkage, in 2005 TMDL Conference. Water
Environment Federation, Philadelphia PA.
Effler, S. W., C. T. Driscoll, S. M. Doerr, C. M. Brooks, M. T. Auer, B. A. Wagner, I. Address,
W. Wang, D. L. lohnson, I. liao, and S. G. Dos Santos. 1996. 5. Chemistry. Pages 263-
283 in S. W. Effler, ed. Limnological and Engineering Analysis of a Polluted Urban
Lake. Springer, New York.
Egglishaw, H.I. 1972. An Experimental Study of the Breakdown of Cellulose in Fast-Flowing
Streams. In Melchiorri-Santolini, U., and I.W. Hopton, Eds., Detritus and Its Role in
Aquatic Ecosystems, Proceedings of an IBP-UNESCO Symposium, Mem. 1st. Ital.
Idrobiol. 29 Suppl.:405-428.
315
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
Elser, J. J., R. W. Sterner, A. E. Galford, T. H. Chrzanowski, M. P. Stainton, and D. W.
Schindler. 2000. Pelagic C:N:P Stoichiometry in a Eutrophied Lake: Responses to a
Whole-Lake Food-Web Manipulation. Ecosystems 3: 293-307.
Engelund, F. and E. Hansen. 1967. A Monograph of Sediment Transport in Alluvial Streams.
Teknisk Vorlag, Copenhagen, Denmark.
Environmental Laboratory. 1982. CE-QUAL-R1: A Numerical One-Dimensional Model of
Reservoir Water Quality; A User's Manual. Instruction Report E-82-1, U.S. Army
Engineers Waterways Experiment Station, Vicksburg, Miss.
Fagerstrom, T., and B. Asell. 1973. Methyl Mercury Accumulation in an Aquatic Food Chain,
A Model and Some Implications for Research Planning. Ambio, 2(5): 164-171.
Fagerstrom, T., R. Kurten ,and B. Asell. 1975. Statistical Parameters as Criteria in Model
Evaluation: Kinetics of Mercury Accumulation in Pike J is ox lucius. Oikos 26:109-116.
Flosi, G., S. Downie, I. Hopelain, M. Bird, R. Coey, and B. Collins. 1998. California Salmonid
Stream Habitat Restoration Manual, Third Edition. Pages 495. CA Department of Fish
and Game, Inland Fisheries Division, Sacramento, CA.
Fogg, G.E., C. Nalewajko, and W.D. Watt. 1965. Extracellular Products of Phytoplankton
Photosynthesis. Proc. RoyalSoc. Biol., 162:517-534.
Ford, D.E., and K.W. Thornton. 1979. Time and Length Scales for the One-Dimensional
Assumption and Its Relation to Ecological Models. Water Resources Research
15(1): 113-120.
Freidig, A.P., E.A. Garicano, and F.I.M. Busser. 1998. Estimating Impact of Humic Acid on
Bioavailability and Bioaccumulation of Hydrophobic Chemicals in Guppies Using
Kinetic Solid-Phase Extraction. Environmental Toxicology and Chemistry, 17(6):998-
1004.
Frey, H.C. , and S. R. Patil. 2001. Identification and Review of Sensitivity Analysis Methods.
Paper read at Sensitivity Analysis Methods, lune 11-12, 2001, at North Carolina State
University, Raleigh NC.
Galtsoff, P.S. 1964. The American Oyster Crassostrea virginica Gmelin. U.S. Department of
the Interior, Fish & Wildlife Service, Fishery Bulletin 64.
Ganf, G.G, and P. Blazka. 1974. Oxygen Uptake, Ammonia and Phosphate Excretion by
Zooplankton in a Shallow Equatorial Lake (Lake Goerge, Uganda). Limnol. Oceanog.
19(2):313-325.
Giesy, J. P., and K. Kannan. 2001. Global Distribution of Perfluorooctane Sulfonate in Wildlife.
Environ. Sci. Technol. 35: 1339-1342.
Gilek, M., M. Bjork, D. Broman, N. Kautsky, and C. Naf. 1996. Enhanced Accumulation of PCB
Congeners by Baltic Sea Blue Mussels, Mytilus edulis, with Increased Algae Enrichment.
Environmental Toxicology and Chemistry, 15(9): 1597-1605.
Gobas, F.A.P.C. 1993. A Model for Predicting the Bioaccumulation Hydrophobic Organic
Chemicals in Aquatic Food-webs: Application to Lake Ontario. Ecological Modelling,
69:1-17.
316
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
Gobas, F.A.P.C., E.J. McNeil, L. Lovett-Doust, and G.D. Haffner. 1991. Bioconcentration of
Chlorinated Aromatic Hydrocarbons in Aquatic Macrophytes (Myriophyllum spicatum).
Environmental Science & Technology, 25:924-929.
Gobas, F. A. P. C., and H. A. Morrison. 2000. Bioconcentration and Biomagnification in the
Aquatic Environment. Pages 189-231 in R. S. Boethling and D. Mackay, eds. Handbook
of Property Estimation Methods for Chemicals. Lewis, Boca Raton, Florida
Gobas, F.A.P.C., M.N. Z-Graggen, X. Zhang. 1995. Time response of the Lake Ontario
Ecosystem to Virtual Elimination of PCBs. Environmental Science & Technology,
29(8):2038-2046.
Gobas, F.A.P.C., Xin Zhang, and Ralph Wells. 1993. Gastrointestinal Magnification: The
Mechanism of Biomagnification and Food Chain Accumulation of Organic Chemicals.
Environmental Science & Technology, 27:2855-2863.
Gobas, F. A. P. C., and X. Zhang. 1994. Interactions of Organic Chemicals with Particulate and
Dissolved Organic Matter in the Aquatic Environment. Pages 83-91 in J. L. Hamelink, P.
F. Landrum, H. L. Bergman, and W. H. Benson, editors. Bioavailability:
Physical,
Chemical, and Biological Interactions. Lewis Publishers, Boca Raton FL.
Godshalk, G.L., and J.W. Barko. 1985. Chapter 4, Vegetative Succession and Decomposition in
Reservoirs. In D. Gunnison (ed.), Microbial Processes in Reservoirs, Dordrecht: Dr. W.
Junk Publishers, pp. 59-77.
Grabowski, J. H., and S. P. Powers. 2004. Habitat complexity mitigates trophic transfer on oyster
reefs. Marine Ecology Progress Series 277:291-295.
Groden, W.T. 1977. Modeling Temperature and Light Adaptationm of Algae. Report 2, Cenyer
for Ecological Modeling, Rensselaer Polytechnic Institute, Troy, New York, 17 pp.
Gunnison, D., J.M. Brannon, and R.L. Chen. 1985. Chapter 9, Modeling Geomicrobial
Processes in Reservoirs. In D. Gunnison (ed.), Microbial Processes in Reservoirs.
Dordrecht: Dr. W. Junk Publishers, pp. 155-167.
Hanna, M. 1990. Evaluation of Models Predicting Mixing Depth. Can. J. Fish. Aquat. Sci.,
47:940-947.
Harris, G.P. 1986. Phytoplankton Ecology: Structure, Function and Fluctuation. Chapman and
Hall, London, 384 pp.
Hawker, D.W. and D.W. Connell. 1985. Prediction of Bioconcentration Factors Under
Non-Equilibrium Conditions. Chemosphere 14(11/12): 1835-1843.
Hayes, P.F. and R.W. Menzel. 1981. The reproductive cycle of early setting Crassostrea
virginica Gmelin in the northern Gulf of Mexico, and its implications for population
recruitment. Biol. Bull. 160: 80-88.
Hemond, H. F., and E. J. Fechner. 1994. Chemical Fate and Transport in the Environment.
Academic Press, New York.
317
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
Henley, W. E., M. A. Patterson, R. J. Neves, and A. D. Lemly. 2000. Effects of Sedimentation
and Turbidity on Lotic Food Webs: A Concise Review for Natural Resource Managers.
Reviews in Fisheries Science 8: 125-139.
Hessen, D. O. and L. J. Tranvik, editors. 1998. Aquatic Humic Substances: Ecology and
Biochemistry. Springer-Verlag, Berlin.
Hewett, S.W., and B.L. Johnson. 1992. Fish Bioenergetics 2 Model. Madison, Wisconsin:
University of Wisconsin Sea Grant Institute, 79 pp.
Hill, I.R., and P.L. McCarty. 1967. Anaerobic Degradation of Selected Chlorinated Pesticides.
Jour. Water Poll. Control Fed. 39:1259.
Hill, W.R., and Napolitano, G.E. 1997. PCB Congener Accumulation by Periphyton, Herbivores,
and Omnivores. Archives Environmental Contamination Toxicology, 32:449-455.
Hofstetter, R.P. 1977. Trends in population levels of the American oyster, Crassostrea virginica
Gmelin on public reefs in Galveston Bay, Texas. Technical Series Number 10. 90 pp.
Texas Parks and Wildlife Department, Coastal Fisheries Branch, Austin, Texas.
Hofstetter, R.P. 1983. Oyster population trends in Galveston Bay 1973-1978. Management Data
SeriesNumber 51. Texas Parks and Wildlife Department, Coastal Fisheries Branch,
Austin, Texas.
Hoggan, D.H. 1989. Computer-Assisted Floodplain Hydrology and Hydraulics, McGraw-Hill
New York, 518 pp.
Home, A.J., and C.R. Goldman. 1994. Limnology - 2nd edition. McGraw-Hill, New York, 576
pp.
Howick, G.L., F. deNoyelles, S.L. Dewey, L. Mason, and D. Baker. 1993. The Feasibility of
Stocking Largemouth Bass in 0.04-ha Mesocosms Used for Pesticide Research.
Environmental Toxicology and Chemistry, 12:1883-1893.
Hrbacek, J. 1966. A Morphometrical Study of Some Backwaters and Fish Ponds in Relation to
the Representative Plankton Samples. In Hydrobiological Studies 7, J. Hrbacek, Ed.,
Czechoslovak Academy of Sciences, Prague, p. 221-257.
Hudon, C., S. Lalonde, and P. Gagnon. 2000. Ranking the Effects of Site Exposure, Plant
Growth Form, Water Depth, and Transparency on Aquatic Plant Biomass. Can. J. Fish.
Aquat. Sci. 57(Suppl. 1):31-42.
Hutchinson, G.E. 1957. A Treatise on Limnology, Volume I, Geography, Physics, and
Chemistry. John Wiley & Sons, New York, 1015 pp.
Hutchinson, G.E. 1967. A Treatise on Limnology, Volume II, Introduction to Lake Biology and
the Limnoplankton. Wiley & Sons, New York, 1115 pp.
Ibanez, C., J. Saldana, and N. Prat. 1999. A Model to Determine the Advective Circulation in a
Three Layer, Salt Wedge Estuary: Application to the Ebre River Estuary. Estuarine,
Coastal and Shelf Science 48: 271-279.
Iman, R. I., and W. J. Conover. 1982. A distribution-free approach to inducing rank correlation
among input variables. Communications in Statistics B11: 311-334.
318
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
Imboden, D.M. 1973. Limnologische Transport- und Nahrstoffmodelle. Schweiz. Z. Hydrol.
35:29-68.
Ingle, R.M. 1951. Spawning and setting of oysters in relation to seasonal environmental changes.
Bull. Mar. Sci. Gulf Caribb. 1: 111-135.
lohanson, R.C., J.C. Imhoff, and H.H. Davis, Ir. 1980. Users Manual for Hydrological
Simulatiuon Program Fortran (HSPF). U.S. Environmental Protection Agency, Athens
Environmental Research Laboratory, EPA-600/9-80-015, 678 pp.
lorgensen, L. A., S. E. lorgensen, and S. N. Nielsen. 2000. ECOTOX: Ecological Modelling and
Ecotoxicology. in. Elsevier Science.
l0rgensen, S.E. 1976. A Eutrophication Model for a Lake. Ecol. Modelling, 2:147-165.
l0rgensen, S.E., H.F. Mejer, M. Friis, L.A. l0rgensen, and J. Hendriksen (Eds.). 1979.
Handbook of Environmental Data and Ecological Parameters. Copenhagen:
International Society of Ecological Modelling.
I0rgensen, S. E. 1986. Fundamentals of Ecological Modelling. Elsevier, Amsterdam.
lunge, C.O. 1966. Depth distributions for quadratic surfaces and other configurations. In:
Hrbacek, I. (Ed.): Hydrobiological Studies. Vol. 7, Academia, Prague, pp. 257-265.
lupp, B.P., and D.H.N. Spence. 1977a. Limitations on Macrophytes in a Eutrophic Lake, Loch
Leven I. Effects of Phytoplankton. Journal Ecology, 65:175-186.
lupp, B.P., and D.H.N. Spence. 1977b. Limitations on Macrophytes in a Eutrophic Lake, Loch
Leven II. Wave Action, Sediments, and Waterfowl Grazing. Journal Ecology, 65:431-
446.
Kaller, M. D., and K. J. Hartman. 2004. Evidence of a Threshold Level of Fine Sediment
Accumulation for Altering Benthic Macroinvertebrate Communities. Hydrobiologia 518:
95-104.
Kannan, K., J. Koistinen, K. Beckman, T. Evans, J. F. Gorzelany, K. J. Hansen, P. D. lones, E.
Helle, M. Nyman, and J. P. Giesy. 2001. Accumulation of Perfluorooctane Sulfonate in
Marine Mammals. Environ. Sci. Technol. 35: 1593-1598.
Karickhoff, S.W., and K.R. Morris. 1985. Sorption Dynamics of Hydrophobic Pollutants in
Sediment Suspensions. Environmental Toxicology and Chemistry, 4:469-479.
Key, B. D., R. D. Howell, and C. S. Criddle. 1998. Defluorination of Organofluorine Sulfur
Compounds by Pseudomonas sp. Strain D2. Environ. Sci. Technol. 32: 2283-2287.
Kitchell, J.F., J.F. Koonce, R.V. O'Neill, H.H. Shugart, Ir., I.I Magnuson, and R.S. Booth. 1972.
Implementation of a Predator-Prey Biomass Model for Fishes. Eastern Deciduous Forest
Biome , International Biological Program, Report 72-118. 57 pp.
Kitchell, I.F., I.F. Koonce, R.V. O'Neill, H.H. Shugart, Ir., I.I Magnuson, and R.S. Booth. 1974.
Model of fish biomass dynamics. Trans. Am. Fish. Soc. 103:786-798.
Koelmans, A.A., S.F.M. Anzion, and L. Lijklema. 1995. Dynamics of Organic Micropollutant
Biosorption to Cyanobacteria and Detritus. Environmental Science & Technology,
29(4):933-940.
319
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
Koelmans, A.A., and E.H.W. Heugens. 1998. Binding Constants of Chlorobenzenes and
Polychlorobiphenyls for Algal Exudates. Water Science Technology, 37(3):67-73.
Koelmans, A. A., A. Van der Heidje, L. M. Knijff, and R. H. Aalderink. 2001. Integrated
Modelling of Eutrophication and Organic Contaminant Fate & Effects in Aquatic
Ecosystems. AReview. Water Research 35: 3517-3536.
Koseff, J. R., J. K. Holen, S. G. Monismith, and J. E. Cloern. 1993. Coupled effects of vertical
mixing and benthic grazing on phytoplankton populations in shallow, turbid estuaries.
Journal of Marine Research 51: 843-868.
Kremer, J.N., and S.W. Nixon. 1978. A Coastal Marine Ecosystem. Springer-Verlag, New
York, N.Y., 217 pp.
Krenkel, P.A., and G.T. Orlob. 1962. Turbulent Diffusion and the Reaeration Coefficient.
Proc. ASCE, Jour. San. Eng. Div., 88 (SA 2):53-83.
Krone, R. B. 1962. Flume Studies of The Transport of Sediment in Estuarial Shoaling
Nrocesses: Final Report, Hydraulic Engr. and San. Engr., Research Lab., University of
California at Berkeley.
Lam, R.K., and B.W. Frost. 1976. Model of Copepod Filtering Responses to Changes in Size
and Concentration of Food. Limnol. Oceanogr. 21:490-500.
Landrum, P. F., S. R. Nihart, B. J. Eadie, and W. S. Gardner. 1984. Reverse-phase Separation
Method for Determining Pollutant Binding to Aldrich Humic Acid and Dissolved
Organic Carbon of Natural Waters. Environ. Sci. Technol. 18:187-192.
Lange, C. C. 2000. The Aerobic Biodegradation of N-EtFOSE Alcohol by the Microbial Activity
Present in Municpal Wastewater Treatment Sludge. 3M Environmental Laboratory, St.
Paul, MN.
Larkin, G. A., and P. A. Slaney. 1996. Calibration of a Habitat Sedimentation Indicator for Use
in Measuring the Effectiveness of Watershed Restoration Treatments. Pages 14. Province
of British Columbia, Ministry of Environment, Lands and Parks, and Ministry of Forests,
Vancouver, B.C., Canada.
Larsen, D.P., H.T. Mercier, and K.W. Malueg. 1973. Modeling Algal Growth Dynamics in
Shagawa Lake, Minnesota, with Comments Concerning Nrojected Restoration of the
Lake. In E.J. Middlebrooks, D.H. Falkenborg, and T.E. Maloney (Eds.). Modeling the
Eutrophication Process. Logan, Utah: Utah State University, pp. 15-32.
Le Cren, E.P., and R.H. Lowe-McConnell (Eds.). 1980. The Functioning of Freshwater
Ecosystems. Cambridge: Cambridge University Press, 588 pp.
Legendre, P., and L. Legendre. 1998. Numerical Ecology. Elsevier Science BV, Amsterdam.
Lehman, J.T., D.B. Botkin, and G.E. Likens. 1975. The Assumptions and Rationales of a
Computer Model of Phytoplankton Population Dynamics. Limnol. and Oceanogr.
20(3):343-364.
Leidy, G.R., and R.M. lenkins. 1977. The Development of Fishery Compartments and
Population Rate Coefficients for Use in Reservoir Ecosystem Modeling. Contract Rept.
320
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
CR-Y-77-1, U.S. Army Engineer Waterways Experiment Station, Vicksburg Mississippi,
134 pp.
Lenihan, H. S. 1999. Physical-biological coupling on oyster reefs: how habitat structure
influences individual performance. Ecological Monographs 69:251-275.
Leung, D.K. 1978. Modeling the Bioaccumulation of Pesticides in Fish. Report N. 5, Center for
Ecological Modeling, Resselaer Polytechnic Institute, Troy, N. Y.
Liss, P.S., and P.G. Slater. 1974. Flux of Gases Across the Air-Sea Interface. Nature, 247:181-
184.
Loosanoff, V.L. 1953. Behavior of oysters in water of low salinity. Proceedings of the National
Shellfisheries Association (1952): 135-151.
Lyman, W.I., W.F. Reehl, and D.H. Rosenblatt. 1982. Handbook of Chemical Property
Estimation Methods. McGraw-Hill, New York.
Maaret, K., K, Leif, and H. Bjarne. 1992. Studies on the Partition Behavior of Three Organic
Hydrophobic Pollutants in Natural Humic Water. Chemosphere, 24(7):919-925.
Mabey, W., and T. Mill. 1978. Critical Review of Hydrolysis of Organic Compounds in Water
Under Environmental Conditions. J. Phys. Chem. Ref Data, 7:383-415.
Macek, K.I., M.E. Barrows, R.F. Frasny, and B.H. Sleight III. 1977. Bioconcentration of 14C-
Pesticides by Bluegill Sunfish During Continuous Aqueous Exposure. In Structure-
Activity Correlations in Studies of Toxicity and Bioconcentration with Aquatic
Organisms, G.D. Veith and D. Konasewick, eds.
Mackay, D., H. Puig, and L.S. McCarty. 1992. An Equation Describing the Time Course and
Variability in Uptake and Toxicity of Narcotic Chemicals to Fish. Environmental
Toxicology and Chemistry, 11:941 -951.
Maloney, K. O., and I. W. Feminella. 2006. Evaluation of single- and multi-metric benthic
macroinvertebrate indicators of catchment disturbance over time at the Fort Benning
Military Installation, Georgia, USA. Ecological Indicators 6:469-484.
Mancini, I.L. 1983. A Method for Calculating Effects on Aquatic Organisms of Time Varying
Concentrations. Water Res. 10:1355-1362.
Marmorek, D. R., R. M. MacQueen, C. H. R. Wedeles, I. Korman, P. I. Blancher, and D. K.
McNicol. 1996. Improving pH and Alkalinity Estimates for Regional-scale Acidification
Models: Incorporation of Dissolved Organic Carbon. Can. J. Fish. Aquat. Sci 53: 1602-
1608.
Martin, J. L., R. A. Ambrose, and T. A. Wool. 2006. WASP7 Benthic Algae - Model Theory and
User's Guide. U.S. Environmental Protection Agency, Athens, Georgia.
Martin, J. W., S. A. Mabury, K. R. Solomon, and D. C. G. Muir. 2003a. Bioconcentration and
Tissue Distribution of Perfluorinated Acids in Rainbow Trout (Oncorhyncus mykiss).
Environmental Toxicology and Chemistry 22: 196-204.
Martin, J. W., S. A. Mabury, K. R. Solomon, and D. C. G. Muir. 2003b. Dietary Accumulation
of Perfluorinated Acids in Juvenile Rainbow Trout (Oncorhynchus mykiss).
321
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
Environmental Toxicology and Chemistry 22: 189-195.
Mauriello, D. A., and R. A. Park. 2002. An adaptive framework for ecological assessment and
management. Pages 509-514 in A. E. Rizzoli and A. J. Jakeman, eds. Integrated
Assessment and Decision Support. International Environmental Modeling and Software
Society, Manno Switzerland.
Mayer, F. L., Ir., and M. R. Ellersieck. 1986. Manual of Acute Toxicity: Interpretation and
DataBase for 410 Chemicals and 66 Species of Freshwater Animals: U.S. Department of
Interior Fish and Wildlife Service, Resource Publication 160; Wasjington, D.C.
Mayio, A.E., and G.H. Grubbs. 1993. Nationwide Water-Quality Reporting to the Congress as
Required Under Section 305(b) of the Clean Water Act. In National Water Summary
1990-97, Water Supply Paper 2400; Washington, D.C.: U.S. Geological Survey, pp. 141-
146.
McCarty, L.S., G.W. Ozburn, A.D. Smith, and D.G. Dixon. 1992. Toxicokinetic Modeling of
Mixtures of Organic Chemicals. Environmental Toxicology and Chemistry, 11:1037-
1047.
McConnaughey, T. A., I. W. LaBaugh, D. O. Rosenberry, R. G. Striegl, M. M. Reddy, P. F.
Schuster, and V. Carter. 1994. Carbon Budget for a Groundwater-fed Lake: Calcification
Supports Summer Photosynthesis. Limnol. Oceanog. 39: 1319-1332.
Mclntire, C.D. 1968. Structural Characteristics of Benthic Algal Communities in Laboratory
Streams. Ecology 49(3):520-537.
Mclntire, C.D. 1973. Periphyton Dynamics in Laboratory Streams: a Simulation Model and Its
Implications. Ecological Monographs 43(3):399-419.
Mclntire, C.D., and I.A. Colby. 1978. A Hierarchical Model of Lotic Ecosystems. Ecological
Monographs 48:167-190.
McKay, M.D., W.I. Conover, and R.I. Beckman. 1979. A Comparison of Three Methods for
Selecting Values of Input Variables in the Analysis of Output from a Computer Code.
Technometrics 211:239-245.
McKim, I.M., G.F. Olson, G.W. Holcombe, and E.P. Hunt. 1976. Long-Term Effects of
Methylmercuric Chloride on Three Generations of Brook Trout (Salvelinus fontinalis):
Toxicity, Accumulation, Distribution, and Elimination. Journal Fisheries Research
Board Canada, 33(12):27226-2739.
McKim, I.M., P. Schmeider, and G. Veith. 1985. Absorption Dynamics of Organic Chemical
Transport Across Trout Gills as Related to Octanol-Water Partition Coefficient.
Toxicology and Applied Pharmacology, 77:1 -10.
Megard, R.O., W.S. Comles, P.D. Smith, and A.S. Knoll. 1979. Attenuation of Light and Daily
Integral Rates of Photosynthesis Attained by Planktonic Algae. Limnol. Oceanogr.,
24:1038-1050.
Menzel, R. W. 1955. Some phases of the biology of Ostrea eguestris Say and a comparison with
Crassostrea virginica Gmelin. Publ. Inst. Mar. Sci. Univ. Tex. 4(1): 69-153.
Menzel, R.W. 1951. Early sexual development and growth of the American oyster in Louisiana
322
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
waters. Science 113: 719-721.
Moloney, C. L. and J. G. Field. 1989. General allometric equations for rates of nutrient uptake,
ingestion, and respiration in plankton organisms. Limnol. Oceanogr. 35:1290-1299.
Moody, C. A., and J. A. Field. 2000. Perfluorinated Surfactants and the Environmental
Implications of their Use in Fire-Fighting Foams. Environ. Sci. Technol. 34: 3864-3870.
Muirhead-Thomson, R.C. 1987. Pesticide Impact on Stream Fauna with Special Reference to
Macroinvertebrates. Cambridge: Cambridge University Press, 275 pp.
Mulholland, P. J., J. W. Feminella, B. G. Lockaby, and G. L. Hollon. 2007. Riparian Ecosystem
Management at Military Installations: Determination of Impacts and Evaluation of
Restoration and Enhancement Strategies. Oak Ridge National Laboratory, Oak Ridge
TN.
Mullin, M.M. 1963. Some Factors Affecting the Feeding of Marine Copepods of the Genus
Calanus. Limnol. Oceanogr. 8:239-250.
Mullin, M.M., E.F. Stewart, and FJ. Foglister. 1975. Ingestion by Planktonic Grazers as a
Function of Concentration of Food. Limnol. Oceanog. 20:259-262.
Murphy, T. P., K. I. Hall, and I. Yesaki. 1983. Coprecipitation of Phosphate with Calcite in a
Naturally Eutrophic Lake. Limnol. Oceanog. 28: 58-69.
Nalewajko, C. 1966. Photosynthesis and Excretion in Various Planktonic Algae. Limnol.
Oceanogr., 11:1-10.
Neal, C. 2001. The potential for phosphorus pollution remediation by calcite precipitation in UK
freshwaters. Hydrology and Earth System Sciences, 5: 119-131.
Newcombe, C. P. 2003. Impact Assessment Model for Clear Water Fishes Exposed to
Excessively Cloudy Water. Journal of the American Water Resources Association
(JAWRA) 39: 529-544.
Nichols, D.S., and D.R. Keeney. 1976. Nitrogen Nutrition of Myriphyllum spicatum: Uptake
and Translocation of 15N by Shoots and Roots. Freshwater Biology 6:145-154.
O'Connor, D.J., and J.P. Connolly. 1980. The Effect of Concentration of Adsorbing Solids on
the Partition Coefficient. Water Research, 14:1517-1523.
O'Connor, D.J., and W.E. Dobbins. 1958. Mechanism of Reaeration in Natural Streams. ASCE
Transactions, pp. 641-684, Paper No. 2934.
O'Connor, D.J., J.L. Mancini, and J.R. Guerriero. 1981. Evaluation of Factors Influencing the
Temporal Variation of Dissolved Oxygen in the New York Bight, Phase II. Manhattan
College, Bronx, New York
O'Neill, R.V. 1969. Indirect Estimation of Energy Fluxes in Animal Food Webs. Jour.
Theoret. Biol., 22:284-290.
Odum, E. P. 1971. Fundamentals of Ecology. Third edition. W.B. Saunders Co., Philadelphia.
Odum, E.P., and A. A. de la Cruz. 1963. Detritus as a Major Component of Ecosystems. Amer.
Inst. Biol. Sci. Bull., 13:39-40.
OECD. 2002. Co-operation On Existing Chemicals: Hazard Assessment of
323
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
Perfluorooctaneulfonate (PFOS) and Its Salts. Organization for Economic Co-operation
and Development.
Oliphant, A. J., J. W. Rose III, C. S. B. Grimmond, and H.-P. Schmid. 2006. Observations of
Canopy Light Penetration and Net Ecosystem Exchange of CO2 under different sky
conditions in a Mid-Western Mixed Deciduous Forest, in 27th Conference on
Agricultural and Forest Meteorology. American Meteorological Society, San Diego.
Oliver, B. G., and A. I. Niemi. 1988. Trophodynamic Analysis of Polychlorinated Biphenyl
Congeners and Other Chlorinated Hydrocarbons in the Lake Ontario Ecosystem. Environ.
Sci. Technol. 22:388-397.
O'Neill, R.V., D.L. DeAngelis, I.B. Waide, and T.F.H. Allen. 1986. A Hierarchical Concept of
the Ecosystem. Princeton University Press, Princeton, N J.
O'Neill, R.V., R.A. Goldstein, H.H. Shugart, and I.B. Mankin. 1972. Terrestrial Ecosystem
Energy Model. Eastern Deciduous Forest Biome, International Biological Program
Report 72-19.
Osmond, D. L., D.E. Line, I.A. Gale, R.W. Gannon, C.B. Knott, K.A. Bartenhagen, M.H.
Turner, S.W. Coffey, I. Spooner, I. Wells, I.C. Walker, L.L. Hargrove, M.A. Foster, P.D.
Robillard, and D.W. Lehning. 1995. WATERSHEDSS: Water, Soil and Hydro-
Environmental Decision Support System.
Otsuki, A., and R. G. Wetzel. 1972. Coprecipitation of Phosphate with Carbonates in a Marl
Lake. Limnology & Oceanography 17: 763-767.
Owens, M., R.W. Edwards, and I.W. Gibbs. 1964. Some Reaeration Studies ion Streams.
Internat. Jour. Air Water Poll. 8:469-486.
Palisade Corporation. 1991. Risk Analysis and Simulation Add-In for Lotus 1-2-3. Newfield New
York, 342 pp.
Park, K., A. Y. Kuo, I. Shen, and I. M. Hamrick. 1995. A Three-Dimensional Hydrodynamic-
Eutrophication Model (HEM-3D): Description of Water Quality and Sediment Process
Submodels. Special Report in Applied Marine Science and Ocean Engineering No. 327.
Park, R.A. 1978. A Model for Simulating Lake Ecosystems. Center for Ecological Modeling
Report No. 3, Rensselaer Polytechnic Institute, Troy, New York, 19 pp.
Park, R. A., and C. D. Collins. 1982. Realism in Ecosystem Models. Perspectives in Computing
2: 18-27.
Park, R.A. 1984. TOXTRACE: A Model to Simulate the Fate and Transport of Toxic
Chemicals in Terrestrial and Aquatic Environments. Acqua e Aria, No. 6, p. 599-607 (in
Italian).
Park, R.A. 1990. AQUATOX, a Modular Toxic Effects Model for Aquatic Ecosystems. Final
Report, EPA-026-87; U.S. Environmental Protection Agency, Corvallis, Oregon.
Park, R. A. 1991. Global Climate Change and Greenhouse Emissions. Pages 171-182
Subcommittee on Health and Environment, U.S. House of Representatives, Washington
DC.
324
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
Park, R.A. 1999. Evaluation of AQUATOXfor Predicting Bioaccumulation ofPCBs in the Lake
Ontario Food Web. In: AQUATOX for Windows: A Modular Fate and Effects Model for
Aquatic Ecosystems-Volume 3: Model Validation Reports. U.S. Environmental
Protection Agency 2000. EPA-823-R-00-008
Park, R.A., B.B. MacLeod, C.D. Collins, J.R. Albanese, and D. Merchant. 1985. Documentation
of the Aquatic Ecosystem MINI. Cleaner, A Final Report for Grant No. R806299020.
U.S. Environmental Protection Agency, Environmental Research Laboratory, Athens,
Georgia. 85 pp.
Park, R.A., B.H. Indyke, and G.W. Heitzman. 1981. Predicting the Fate of Coal-Derived
Pollutants in Aquatic Environments. Paper presented at Energy and Ecological
Modelling symposium, Louisville, Kentucky, April 2023, 1981. Developments in
Environmental Modeling 1. 7 pp.
Park, R.A., C.D. Collins, C.I. Connolly, J.R. Albanese, and B.B. MacLeod. 1980.
Documentation of the Aquatic Ecosystem Model MS.CLEANER, A Final Report for
Grant No. R80504701, U.S. Environmental Protection Agency, Environmental Research
Laboratory, Athens, Georgia. 112 pp.
Park, R.A., C.D. Collins, D.K. Leung, C.W. Boylen, J.R. Albanese, P. deCaprariis, and H.
Forstner. 1979. The Aquatic Ecosystem Model MS.CLEANER. In State-of- the-Art in
Ecological Modeling, edited by S.E. Jorgensen, 579-602. International Society for
Ecological Modelling, Denmark.
Park, R.A., C.I. Connolly, J.R. Albanese, L.S. Clesceri, G.W. Heitzman, H.H. Herbrandson, B.H.
Indyke, J.R. Loehe, S. Ross, D.D. Sharma, and W.W. Shuster. 1980. Modeling
Transport and Behavior of Pesticides and Other Toxic Organic Materials in Aquatic
Environments. Center for Ecological Modeling Report No. 7. Rensselaer Polytechnic
Institute, Troy, New York. 163 pp.
Park, R.A., C.I. Connolly, J.R. Albanese, L.S. Clesceri, G.W. Heitzman, H.H. Herbrandson, B.H.
Indyke, J.R. Loehe, S. Ross, D.D. Sharma, and W.W. Shuster. 1982. Modeling the Fate
of Toxic Organic Materials in Aquatic Environments. U.S. Environmental Protection
Agency Rept. EPA-6OO/S3-82-028, Athens, Georgia.
Park, R.A., D. Scavia, and N.L. Clesceri. 1975. CLEANER, The Lake George Model. In
Ecological Modeling in a Management Context. Resources for the Future, Inc.,
Washington, D.C.
Park, R.A., J.J. Anderson, G.L. Swartzman, R. Morison, and J.M. Emlen. 1988. Assessment of
Risks of Toxic Pollutants to Aquatic Organisms and Ecosystems Using a Sequential
Modeling Approach. In Fate and Effects of Pollutants on Aquatic Organisms and
Ecosystems, 153-165. EPA/600/9-88/001. Athens, Ga.: U.S. Environmental Protection
Agency
Park, R.A., R.V. O'Neill, J.A. Bloomfield, H.H. Shugart, Jr., R.S. Booth, J.F. Koonce, M.S.
Adams, L.S. Clesceri, E.M. Colon, E.H. Dettman, R.A. Goldstein, J.A. Hoopes, D.D.
Huff, S. Katz, J.F. Kitchell, R.C. Kohberger, E.J. LaRow, D.C, McNaught, J.L. Peterson,
D. Scavia, J.E. Titus, P.R. Weiler, J.W. Wilkinson, and C.S. Zahorcak. 1974. A
Generalized Model for Simulating Lake Ecosystems. Simulation, 23(2):30-50.
Reprinted in Benchmark Papers in Ecology.
325
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
Park, R.A., T.W. Groden, and C.J. Desormeau. 1979. Modifications to the Model CLEANER
Requiring Further Research. In Perspectives on Lake Ecosystem Modeling, edited by D.
Scavia and A. Robertson. Ann Arbor Science Publishers, Inc., 22 pp.
Park, R. A., M. S. Trehan, P. W. Mausel, and R. C. Howe. 1989. The Effects of Sea Level Rise
on U.S. Coastal Wetlands. Pages 1-1 to 1-55. in J. B. Smith and D. A. Tirpak, editors.
The Potential Effects of Global Climate Change on the United States, Appendix B - Sea
Level Rise. U.S. Environmental Protection Agency, Washington, D.C.
Park, R. A., E. C. Blancher, S. A. Sklenar, and J. L. Wood. 2002. Modeling the Effects of
Multiple Stressors on a Use-Impaired River, in Society of Environmental Toxicology and
Chemistry, Salt Lake City.
Park, R. A., and J. S. Clough. 2005. Validation of AQUATOX with Nonylphenol Field Data
(Unpublished Report). U.S. Environmental Protection Agency, Washington, DC.
Park, R. A., J. S. Clough, M. C. Wellman, and A. S. Donigian. 2005. Nutrient Criteria
Development with a Linked Modeling System: Calibration of AQUATOX Across a
Nutrient Gradient. Pages 885-902 in TMDL 2005. Water Environment Federation,
Philadelphia, Penn.
Park, R. A., J. S. Clough, and M. C. Wellman. 2008. AQUATOX: Modeling Environmental Fate
and Ecological Effects in Aquatic Ecosystems. Ecological Modelling 213: 1-15.
Parker, R.A. 1972. Estimation of Aquatic Ecosystem Parameters. Verh. Internal Verein.
Limnol. 18:257-263.
Parsons, T.R., R.J. LeBresseur, J.D. Fulton, and O.D. Kennedy. 1969. Production Studies in the
Strait of Georgia II. Secondary Production Under the Fraser River Plume, February to
May, 1967. Jour. Exp. Mar. Biol. Ecol. 3:39-50.
Partheniades, E. 1965. Erosion and Deposition of Cohesive Soils. ASCEJour. Hydrol. Div. pp.
105-138.
Partheniades, E. 1971. "Erosion and Deposition of Cohesive Materials". In River Mechanics, H.
W. Shen Ed. Chapter 25. Water Resources Publications, Littleton, Colorado.
Pastorok, R. A., S. M. Bartell, S. Ferson, and L. R. Ginzburg, editors. 2002. Ecological Modeling
in Risk Assessment. Lewis, Boca Raton, Florida.
Patten, B.C., D.A. Egloff, and T.H. Richardson. 1975. Total Ecosystem Model for a Cove in
Lake Texoma. In B.C. Patten (Ed.) Systems Analysis and Simulation in Ecology. Vol. III.
New York: Academic Press, pp. 205-241.
Plafkin, J. L., M. T. Barbour, K. D. Porter, S. K. Gross, and R. M. Hughes. 1989. Rapid
bioassessment protocols for use in streams and rivers: Benthic macroinvertebrates and
fish EPA 440-4-89-001, U.S. Environmental Protection Agency, Office of Water
Regulations and Standards, Washington, D.C.
Press, W.H., B.P. Flannery, S.A. Teukolsky, and W.T. Vetterling. 1986. Numerical Recipes:
The Art of Scientific Computing. Cambridge University Press, Cambridge, U.K. 818 pp.
Purcell, A. H., C. Friedrich, and V. H. Resh. 2002. An Assessment of a Small Urban Stream
Restoration Project in Northern California. Restoration Ecology 10:685-694.
326
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
Quantitative Environmental Analysis. 2001. Documentation: Bioaccumulation Model
QEAFDCHN v. 1.0. Pages 21. QEA, LLC, Montvale, NJ.
Raimondo, S., Vivian, D.N., Barron, M.G., 2007. Web-based Interspecies Correlation
Estimation (Web-ICE) for Acute Toxicity: User Manual. Version 1.1. EPA/600/R-07/071,
U.S. Environmental Protection Agency, Gulf Breeze, FL.
Redfield, A.C. 1958. The Biological Control of Chemical Factors in the Environment.
American Scientist 46:205-222.
Riley, G.A. 1963. Theory of Food-Chain Relations in the Ocean. The Sea, 2.
Rode, M., U. Suhr, and G. Wriedt. 2007. Multi-objective calibration of a river water quality
model—Information content of calibration data. Ecological Modelling 204: 129-142.
Rosemond, A.D. 1993. Seasonality and Control of Stream Periphyton: Effects of Nutrients,
Light, and Herbivores. Dissertation, Vanderbilt University, Nashville, Tenn., 185 pp.
Rowe, M., D. Essig, and B. lessup. 2003. Guide to Selection of Sediment Targets for Use in
Idaho TMDLs. Pages 46. Idaho Department of Environmental Quality, Boise, Idaho.
Rykiel, E. I., Ir. 1996. Testing ecological models: the meaning of validation. Ecological
modelling 90:229-244.
Saltelli, A. 2001. Sensitivity Analysis for Importance Assessment. Paper read at Sensitivity
Analysis Methods, lune 11-12, 2001, at North Carolina State University, Raleigh NC.
Sand-Iensen, K. 1977. Effects of Epiphytes on Eelgrass (Zostera marina L.) in Danish Coastal
Waters. Marine Technology Society Journal 17:15-21.
Sathyendranath S, Stuart V, Nair A, Oka K and others (2009) Carbon-to-chlorophyll ratio and
growth rate of phytoplankton in the sea. Mar Ecol Prog Ser 383:73-84
Saunders, G.W. 1980. 7. Organic Matter and Decomposers. In E.P. Le Cren and R.H. Lowe-
McConnell (Eds.), The Functioning of Freshwater Ecosystems. Cambridge: Cambridge
University Press, pp. 341-392.
Scavia, D. 1979. Chapter 6 The Use of Ecological Models of Lakes in Synthesizing Available
Information and Identifying Research Needs. In D. Scavia and A. Robertson (Eds.)
Perspectives on Lake Ecosystem Modeling. Ann Arbor, Michigan: Ann Arbor Science,
pp. 109-168.
Scavia, D. 1980. An Ecological Model of Lake Ontario. Ecological Modelling 8:49-78.
Scavia, D., and R.A. Park. 1976. Documentation of Selected Constructs and Parameter Values
in the Aquatic Model CLEANER. Ecological Modelling 2(l):33-58.
Scavia, D., B.J. Eadie, and A. Robertson. 1976. An Ecological Model for Lake Ontario-Model
Formulation, Calibration, and Preliminary Evaluation. Tech. Report ERL 371-GLERL
12, National Oceanic and Atmospheric Administration, Boulder, Colorado.
Schlesselman, G.W. 1955. The gulf coast oyster industry of the United States. Geographic
Review 45(4): 531-541.
Schnoor, J. E. 1996. Environmental Modeling: Fate and Transport of Pollutants in Water, Air,
327
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
and Soil. John Wiley & Sons, Inc., New York.
Schol, A., V. Kirchesch, T. Bergfeld, and D. Miiller. 1999. Model-based analysis of oxygen
budget and biological processes in the regulated rivers Moselle and Saar: modelling the
influence of benthic filter feeders on phytoplankton. Hydrobiologia 410: 167-176.
Schol, A., V. Kirchesch, T. Bergfeld, F. Scholl, J. Borcherding, and D. Miiller. 2002. Modelling
the chlorophyll content of the River Rhine - interaction between riverine algal production
and population biomass of grazers, rotifers and zebra mussel, Dreissena polymorpha.
Intyernational Review of Hydrobiology 87: 295-317.
Schwarzenbach, R., P. M. Gschwend, and D. M. Imboden. 1993. Environmental Organic
Chemistry. John Wiley & Sons, New York.
Sedell, J.R., F.J. Triska, and N.S. Triska. 1975. The Processing of Conifer and Hardwood
Leaves in Two Coniferous Forest Streams: I. Weight Loss and Associated Invertebrates.
Herh. Internal Verein. Limnol., 19:1617-1627.
Sijm, D.T.H.M., K.W. Broersen, D.F de Roode, and P. Mayer. 1998. Bioconcentration Kinetics
of Hydrophobic Chemicals in Different Densities of Chlorella Opyrenoidosa.
Environmental Toxicology and Chemistry 17:9:1695-1704.
Skoglund, R.S., K. Stange, and D.L. Swackhamer. 1996. A Kinetics Model for Predicting the
Accumulation of PCBs in Phytoplankton. Environmental Science and Technology
30:7:2113-2120.
Small, M. J., and M. C. Sutton. 1986. A Regional pH-Alkalinity Relationship. Water Research
20: 335-343.
Smayda, T.J. 1971. Some Measurements of the Sinking Rate of Fecal Pellets. Limnology and
Oceanography 14:621-625.
Smayda, T.J. 1974. Some Experiments on the Sinking Characteristics of Two Freshwater
Diatoms. Limnology and Oceanography 19:628-635.
Smejtek, P., and S. Wang. 1993. Distribution of Hydrophobic Ionizable Xenobiotics Between
Water and Lipid Membranes: Pentachlorophenol and Pentachlorophenate. A Comparison
with Octanol-Water Partition. Archives of Environmental Contamination and Toxicology,
25(3):394.
Smith, D.J. 1978. WQRRS, Generalized Computer Program for River-Reservoir Systems. U.S.
Army Corps of Engineers, Hydrologic Engineering Center (HEC), Davis, California
Users Manual 401-100, 100A, 210 pp.
Southworth, G.R., J.J. Beauchamp, and P.K. Schmieder. 1978. Bioaccumulation Potential of
Polycyclic Aromatic Hydrocarbons in Daphniapulex. Water Res., 12:973-977.
Spacie, A., and J.L. Hamelink. 1982. Alternative Models for Describing the Bioconcentration of
Organics in Fish. Environmental Toxicology and Chemistry, 1:309-320.
Stange, K., and D.L. Swackhamer. 1994. Factors Affecting Phytoplankton Species-Specific
Differences in Accumulation of 40 Polychlorinated Biphenyls (PCBs). Environmental
Toxicology and Chemistry, 13(11): 1849-1860.
328
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
Steele, J.H. 1962. Environmental Control of Photosynthesis in the Sea. Limnol. Oceanogr.,
7:137-150.
Steele, J.H. 1974. The Structure of Marine Ecosystems. Harvard University Press, Cambridge,
Massachusetts, 128 pp.
Steele, J.H., and M.M. Mullin. 1977. Zooplankton Dynamics. In E.D. Goldberg, I.N. McCave,
J.J. 0=Brien, and J.H. Steele (Eds.), The Sea Vol. 6: Marine Modeling, New York:
Wiley-Interscience, p. 857.
Stefan, H.G., and X. Fang. 1994. Dissolved Oxygen Model for Regional Lake Analysis.
Ecological Modelling 71:37-68.
Sterner, R. W., and J. J. Elser. 2002. Ecological Stoichiometry: The Biology of Elements from
Molecules to the Biosphere. Princeton University Press, Princeton NJ.
Sterner, R.W., and N. B. George. 2000. Carbon, Nitrogen, and Phosphorus Stoichiometry of
Cyprinid Fishes. Ecology 81: 127-140.
Stewart, D.C. 1975. Mathematical Modelling of the Ecosystem of Lough Neagh. Ph.D.
Dissertation, Queen's University, Belfast, Northern Ireland.
Straskraba, M. 1973. Limnological Basis for Modeling Reservoir Ecosystems. In Ackermann,
W.C., G.F. White, and E.B. Worhtington (eds.) Man-Made Lakes: Their Problems and
Environmental Effects. Geophys. Momogr. Series Vol. 17, London, pp. 517-538.
Straskraba, M. and A.H. Gnauck. 1985. Freshwater Ecosystems: Modelling and Simulation.
Developments in Environmental Modelling, 8. Elsevier Science Publishers, Amsterdam,
The Netherlands. 309 pp.
Stream Bryophyte Group. 1999. Role of bryophytes in stream ecosystems. Journal of the North
American Benthological Society, 18, 151-184.
Stumm, W., and J. J. Morgan. 1996. Aquatic Chemistry: Chemical Equilibria and Rates in
Natural Waters. John Wiley & Sons, New York.
Suarez, L.A., and M.C. Barber. 1992. PIRANHA Version 2.0, FGETS Version 3.0-11 User's
Manual, In PIRANHA Pesticide and Industrial Chemical Risk Analysis and Hazard
Assessment. Athens, Georgia: U.S. Environmental Protection Agency.
Suren, A., M., and J. G. Jowett. 2001. Effects of Deposited Sediment on Invertebrate Drift: an
Experimental Study. New Zealand Journal of Marine and Freshwater Research 35: 725-
737.Suren, A., M. 2005. Effects of Deposited Sediment on Patch Selection by Two
Grazing Stream Invertebrates. Hydrobiologia 549: 205-218.
Suter, G.W., II, A.E. Rosen, and E. Linder. 1986. 4. Analysis of Extrapolation Error. User's
Manual for Ecological Risk Assessment. Oak Ridge National Laboratory, ORNL-6251,
pp. 49-81.
Swackhamer, D.L., and R.S. Skoglund. 1991. The Role of Phytoplankton in the Partitioning of
Hydrophobic Organic Contaminants in Water. In Baker, R.A., ed., Organic Substances
and Sediments in Water Vol. 2 C Processes and Analytical, Lewis: Chelsea MI, pp. 91-
105.
329
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
Swackhamer, D.L., and R.S. Skoglund. 1993. Bioaccumulation of PCBs by Algae: Kinetics
versus Equilibrium. Environmental Toxicology & Chemistry, 12:831-838.
Tetra Tech Inc. 2002. Draft User's Manual for Environmental Fluid Dynamics Code Hydro
Version (EFDC-Hydro). U.S. Environmental Protection Agency, Atlanta, GA.
Thomann, R. V., and J. Mueller. 1987. Principles of Surface Water Quality Modeling and
Control. HarperCollins, New York, NY.
Thomann, R.V. 1989. Bioaccumulation Model of Organic Chemical Distribution in Aquatic
Food Chains. Environmental Science & Technology, 23:699-707.
Thomann, R.V., and J.J. Fitzpatrick. 1982. Calibration and Verification of a Mathematical
Model of the Eutrophication of the Potomac Estuary. Prepared for Department of
Environmental Services, Government of the District of Columbia, Washington, D.C.
Thomann, R.V., D.M. Di Toro, R.P. Winfield, and D.J. O'Connor. 1975. Mathematical
Modeling of Phytoplankton in Lake Ontario, Part 1. Model Development and
Verification. Manhattan College, Bronx, New York, for U.S. Environmental Protection
Agency EPA-600/3-75-005.
Thomann, R.V., J. Segna, and R. Winfield. 1979. Verification Analysis of Lake Ontario and
Rochester Embayment Three-Dimensional Eutrophication Models. Manhattan College,
Bronx, New York, for U.S. Environmental Protection Agency.
Thomann, R.V., J.A. Mueller, R.P. Winfield, and C.-R. Huang. 1991. Model of Fate and
Accumulation of PCB Homologues in Hudson Estuary. Jour. Environ. Engineering,
117(2): 161-178.
Thompson, J. B., S. Schultze-Lam, T. J. Beveridge, and D. J. Des Marais. 1997. Whiting Events:
Biogenic Origin Due to the Photosynthetic Activity of Cyanobacterial Picoplankton.
Limnology and Oceanography 42: 133-141.
Thompson, R.J., R.I.E. Newell, V.S. Kennedy, and R. Mann. 1996. Reproductive processes and
early development. Pages 335-370 In: V.S. Kennedy, R.I.E. Newell, and A.F. Eble (eds.)
The Eastern Oyster Crassostrea virginica. Maryland Sea Grant College, University of
Maryland, College Park, Maryland.
Titus, J.E., M.S. Adams, P.R. Weiler, R.V. O'Neill, H.H. Shugart, Jr., and J.B. Mankin. 1972.
Production Model for Myriophyllum spicatum L. Memo Rept. 72-19, U.S. International
Biological Program Eastern Deciduous Forest Biome, University of Wisconsin,
Madison, 17 pp.
Toetz, D.W. 1967. The Importance of Gamete Losses in Measurements of Freshwater Fish
Production. Ecology. 48:1017-1020.
Traas, T. P., J. A. Stab, P. R. G. Kramer, W. P. Cofino, and T. Aldenberg. 1996. Modeling and
Risk Assessment of Tributyltin Accumulation in the Food Web of a Shallow Freshwater
Lake Environ. Sci. Technol. 30: 1227-1237.
Traas, T. P., J. H. Janse, T. Aldenberg, and J. T. Brock. 1998. A Food Web Model for Fate and
Direct and Indirect Effects of Dursban 4E (Active Ingredient Chlorpyrifos) in Freshwater
Microcosms. Aquatic Ecology 32: 179-190.
330
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
Traas, T. P., J. H. Janse, P. J. Van den Brink, and T. Aldenberg. 2001. A Food Web Model for
Fate and Effects of Toxicants and Nutrients in Aquatic Mesocosms. Model Description.
RIVM, Bilthoven, The Netherlands.
UNESCO 1983 Algorithms for computation of fundamental properties of seawater, Unesco
technical papers in marine science 44.
U.S. Department of Commerce, 1994, Manual of Harmonic Analysis and Prediction of Tides.
Special Publication No. 98, Revised (1940) Edition (reprinted 1958 with corrections;
reprinted again 1994). United States Government Printing Office, 1994.
U.S. Environmental Protection Agency, 1977. Various reports on Lake Chemung and Lake
Allegan, MI; White Bear Lake, MN; Saratoga Lake, NY; Sebasticook Lake, ME; and
Bantam Lake, Aspinook Pond, and Hanover Pond, CT. National Eutrophication Survey
Working Napers. U.S. Environmental Protection Agency, Washington, D.C.
U.S. Environmental Protection Agency, 1986. Quality Criteria for Water, 1986. Environmental
Protection Agency, Washington, D.C., Office of Water Regulations and Standards,
EPA/440/5-86/001, 398 pp.
U.S. Environmental Protection Agency, 1988. The Effects of Chloropyrifos on a Natural
Aquatic System: A Research Design for Littoral Enclosure Studies and Final Research
Report. U.S. Environmental Protection Agency, Environmental Research Laboratory,
Duluth, Minnessota, 194 pp.
U.S. Environmental Protection Agency. 1989. Green Bay/Fox River Mass Balance Study U.S.
Environmental Protection Agency Great Lakes National Program Office, Chicago, IL.
U.S. Environmental Protection Agency, 1991. Hydrological Simulation Program - FORTRAN -
User's Manual for Release 10 (Pre-release Draft Version). U.S. EPA Technology
Development and Applications Branch in cooperation with USGS Water Resources
Division, Office of Surface Water. By Bicknell, B.R., I.C. Imhoff, I.L. Kittle, A.S.
Donigian, and -R.C. lohanson.
U.S. Environmental Protection Agency, 1995. Great Lakes Water Quality Initiative Technical
Support Document for the Procedure to Determine Bioaccumulation Factors. EPA-820-
B-95-005, U.S. Environmental Protection Agency, Washington, D.C.
U.S. Environmental Protection Agency. 1997. Guiding Principles for Monte Carlo Analysis.
Risk Assessment Forum. Washington, DC: U.S. Environmental Protection Agency.
U.S. Environmental Protection Agency, 1999, Update of Ambient Water Quality Criteria for
Ammonia, September 1999, U.S. EPA Office of Water, U.S. EPA Office of Science and
Technology Washington, D.C.
U.S. Environmental Protection Agency, 2000a. Ambient Aquatic Life Water Quality Criteria for
Dissolved Oxygen (Saltwater)
U.S. Environmental Protection Agency, 2000b. Progress in Water Quality: An Evaluation of the
National Investment in Municipal Wastewater Treatment: Appendix B. EPA-832-R-00-
008, http://www.epa.gov/owmitnet/wqualitv/app-b .pdf.
U.S. Environmental Protection Agency, 2002. National Water Quality Inventory: 2000 Report.
EP A-841 -R-02-001.
331
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
U.S. Environmental Protection Agency, 2005. AQUATOXFor Windows: A Modular Fate and
Effects Model For Aquatic Ecosystems Release 2.1 Addendum To Release 2 Technical
Documentation.
USGS, 2010. C02calc: A User-Friendly Seawater Carbon Calculator for Windows, Mac OS X,
and iOS (iPhone), by L.L. Robbins, M.E. Hansen, J. A. Kleypas, and S.C. Meylan. Open-
File Report 2010-1280.
VanderKooy, S. (editor). 2012. The oyster fishery of the Gulf of Mexico, United States: A
regional management plan (2012 revision). Publication No. 202, Gulf States Marine
Fisheries Commission, Ocean Springs, Mississippi.
Van Rees, K. C. J., K. R. Reddy, and P. A. Moore, Jr. 1991. Lake Okeechobee Phosphorus
Dynamics Study: Biogeochemical Processes in the Sediments, Chapter 7: Phosphorus
Exchange Between Sediment and Overlying Water. Pages 7-1 to 7-26. University
Florida, Soil Science Department, Gainesville, Fl.
Velleux, M., S. Westenbroek, J. Ruppel, M. Settles, and D. Endicott. 2000. A User's Guide to
IPX, the In-Place Pollutant Export Water Quality Modeling Framework, Version 2.7.4.
U.S. Environmental Protection Agency, Office of Research and Development, National
Health and Environmental Effects Research Laboratory, Mid-Continental Ecology
Division-Duluth, Large Lakes Research Station, Grosse lie, Michigan. 179 pp.
Verduin, 1982. Components Contributing to Light Extinction in Natural Waters: Method of
Isolation. Arch. Hydrobiol., 93(3):303-312.
Verscheuren, K. 1983. Handbook of Environmental Data on Organic Chemicals, Second edition.
Van Nostrand Reinhold, New York.
Ward 1963, ASCE 1989, 6:1-16
Watt, W.D. 1966. Release of Dissolved Organic Material From the Cells of Phytoplankton
Species in Natural and Mixed Populations. Proceedings of the Royal Society, London, B
164:521-525.
Weininger, D. 1978. Accumulation of PCBs by Lake Trout in Lake Michigan. Ph.D.
Dissertation, University of Wisconsin, Madison, 232 pp.
Westlake, D.F. 1967. Some Effects of Low Velocity Currents on the Metabolism of Aquatic
Macrophytes. Journal Experimental Botany 18:187-205.
Wetzel, R.G., P.H. Rich, M.C. Miller, and H.L. Allen. 1972. Metabolism of Dissolved and
Particulate Detrital Carbon in a Temperate Hard-water Lake, in U. Melchiorri-Santolinii
and J.W. Hopton (eds.) Detritus and Its Role in Aquatic Ecosystems, Mem. 1st. Ital.
Idobiol., 29(Suppl): 185-243.
Wetzel, R.G. 1975. Limnology, W.B. Saunders, Philadelphia, 743 pp.
Wetzel, R.G. 2001. Limnology: Lake and River Ecosystems. San Diego: Academic Press, 1006
pp.
Whitman, W.G. 1923. The two-film theory of gas absorption. Chem. Metal. Eng. 29:146-148.
Winberg, G. G. 1971. Symbols, Units and Conversion Factors in Studies of Freshwater
Productivity. Pages 23. International Biological Programme Central Office, London.
332
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
CHAPTER 11
Wlosinski, J. H., and C. D. Collins. 1985. Confirmation of the Water Quality Model CE-QUAL-
R1 Using Data from Eau Galle Reservoir, Wisconsin. Army Engineers Waterways
Experiment Station, Vicksburg, Mississippi.
Wood, L.W., P. O.Keefe, and B. Bush. 1997. Similarity Analysis of PAH and PCB
Bioaccumulation Patterns in Sediment-Exposed Chironomus tentans Larvae.
Environmental Toxicology and Chemistry, 16(2):283-292.
Wool, T. A., R. B. Ambrose, J. L. Martin, and E. A. Comer. 2004. Water Quality Analysis
Simulation Program (WASP) Version 6.0 DRAFT: User's Manual. US Environmental
Protection Agency - Region 4, Atlanta GA.
Yuan, 2006. The Development of the Web Based C02SYS Program. Masters Thesis. University
of Montana, Autumn 2006.
Zimmerman, R. J., T. J. Minello, E. F. Klima, and J. M. Nance. 1991. Effects of accelerated sea-
level rise on coastal secondary production. Pages 110-124 in Coastal wetlands. ASCE.
333
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
APPENDIX A. GLOSSARY OF TERMS
Taken in large part from: The Institute of Ecology. 1974. An Ecological Glossary for Engineers
and Resource Managers. TIE Publication #3, 50 pp.
Abiotic nonliving, pertaining to physico-chemical factors only
Adsorption the adherence of substances to the surfaces of bodies with which they
are in contact
Aerobic living, acting, or occurring in the presence of oxygen
Algae any of a group of chlorophyll-bearing aquatic plants with no true leaves,
stems, or roots
Allochthonous material derived from outside a habitat or environment under
consideration
Algal bloom rapid and flourishing growth of algae
Alluvial of alluvium
Alluvium sediments deposited by running water
Ambient surrounding on all sides
Anaerobic capable of living or acting in the absence of oxygen
Anoxic pertaining to conditions of oxygen deficiency
Aphotic below the level of light penetration in water
Assimilation transformation of absorbed nutrients into living matter
Autochthonous material derived from within a habitat, such as through plant growth
Benthic pertaining to the bottom of a water body; pertaining to organisms that
live on the bottom
Benthos those organisms that live on the bottom of a body of water
Biodegradable can be broken down into simple inorganic substances by the action of
decomposers (bacteria and fungi)
Biochemical oxygen
demand (BOD) the amount of oxygen required to decompose a given amount of organic
matter
Biomagnification the step by step concentration of chemicals in successive levels of a food
chain or food web
Biomass the total weight of matter incorporated into (living and/or dead)
organisms
Biota the fauna and flora of a habitat or region
Chlorophyll the green, photosynthetic pigments of plants
Colloid a dispersion of particles larger than small molecules and that do not
settle out of suspension
Consumer an organism that consumes another
Copepods a large subclass of usually minute, mostly free-swimming aquatic
crustaceans
Crustacean a large class of arthropods that bear a horny shell
Decomposers bacteria and fungi that break down organic detritus
Detritus dead organic matter
Diatom any of class of minute algae with cases of silica
Diurnal pertaining to daily occurrence
334
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
Dynamic equilibrium
Ecology
Ecosystem
Emergent
Environment
Epilimnion
Epiphytes
Equilibrium
Euphotic
Eutrophic
Fauna
Flood plain
Flora
Fluvial
Food chain
Food web
Forage fish
Habitat
Humic
Hydrodynamics
Hypolimnion
Influent
Inorganic
Invertebrate
Limiting factor
Limnetic zone
Limnology
Littoral zone
Macrofauna
Macrophytes
Nutrients
Omnivorous
Organic chemical
Overturn
Oxygen depletion
Parameter
a state of relative balance between processes having opposite effects
the study of the interrelationships of organisms with and within their
environment
a biotic community and its (living and nonliving) environment
considered together
aquatic plants, usually rooted, which have portions above water for part
of their life cycle
the sum total of all the external conditions that act on an organism
the well mixed surficial layer of a lake; above the hypolimnion
plants that grow on other plants, but are not parasitic
a steady state in a dynamic system, with outflow balancing inflow
pertaining to the upper layers of water in which sufficient light
penetrates to permit growth of plants
aquatic systems with high nutrient input and high plant growth
the animals of a habitat or region
that part of a river valley that is covered in periods of high (flood) water
plants of a habitat or region
pertaining to a stream
animals linked by linear predator-prey relationships with plants or
detritus at the base
similar to food chain, but implies cross connections
fish eaten by other fish
the environment in which a population of plants or animals occurs
pertaining to the partial decomposition of leaves and other plant material
the study of the movement of water
the lower layer of a stratified water body, below the well mixed zone
anything flowing into a water body
pertaining to matter that is neither living nor immediately derived from
living matter
animals lacking a backbone
an environmental factor that limits the growth of an organism; the factor
that is closest to the physiological limits of tolerance of that organism
the open water zone of a lake or pond from the surface to the depth of
effective light penetration
the study of inland waters
the shoreward zone of a water body in which the light penetrates to the
bottom, thus usually supporting rooted aquatic plants
animals visible to the naked eye
large (non-microscopic), usually rooted, aquatic plants
chemical elements essential to life
feeding on a variety of organisms and organic detritus
compounds containing carbon;
the complete circulation or mixing of the upper and lower waters of a
lake when temperatures (and densities) are similar
exhaustion of oxygen by chemical or biological use
a measurable, variable quantity as distinct from a statistic
335
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION APPENDICES
Pelagic zone
Periphyton
Oxidation
Photic zone
Phytoplankton
Plankton
Pond
Population
Predator
Prey
Producer
Production
Productivity
Productivity, primary
Productivity, secondary
Reservoir
Riverine
Rough fish
Sediment
Siltation
Stratification
Substrate
Succession
Tolerance
Trophic level
Turbidity
Volatilization
Wastewater
Wetlands
Zooplankton
open water with no association with the bottom
community of algae and associated organisms, usually small but densely
set, closely attached to surfaces on or projecting above the bottom
a reaction between molecules, ordinarily involves gain of oxygen
the region of aquatic environments in which the intensity of light is
sufficient for photosynthesis
small, mostly microscopic algae floating in the water column
small organisms floating in the water
a small, shallow lake
a group of organisms of the same species
an organism, usually an animal, that kills and consumes other organisms
an organism killed and at least partially consumed by a predator
an organism that can synthesize organic matter using inorganic materials
and an external energy source (light or chemical)
the amount of organic material produced by biological activity
the rate of production of organic matter
the rate of production by plants
the rate of production by consumers
an artificially impounded body of water
pertaining to rivers
a non-sport fish, usually omnivorous in food habits
any mineral and/or organic matter deposited by water or air
the deposition of silt-sized and clay-sized (smaller than sand-sized)
particles
division of a water body into two or more depth zones due to
temperature or density
the layer on which organisms grow; the organic substance attacked by
decomposers
the replacement of one plant assemblage with another through time
an organism's capacity to endure or adapt to unfavorable conditions
all organisms that secure their food at a common step in the food chain
condition of water resulting from suspended matter, including inorganic
and organic material and plankton
the act of passing into a gaseous state at ordinary temperatures and
pressures
water derived from a municipal or industrial waste treatment plant
land saturated or nearly saturated with water for most of the year;
usually vegetated
small aquatic animals, floating, usually with limited swimming
capability
336
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
APPENDIX B. USER-SUPPLIED PARAMETERS AND DATA
The model has many parameters and internal variables. Most of these are linked to data structures such as ChemicalRecord, SiteRecord, and
ReminRecord, which in turn may be linked to input forms that the user accesses through the Windows environment. Although consistency has
been a goal, some names may differ between the code, the user interface, and the technical documentation
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
ChemicalRecord
Chemical
Underlying Data
For each chemical simulated, the following
Parameters are required
Chemical
ChemName
N/A
Chemical's Name. Used for Reference only.
N/A
CAS Registry No.
CASRegNo
N/A
CAS Registry Number. Used for Reference only.
N/A
Molecular Weight
MolWt
MolWt
Molecular weight of pollutant
g/mol
Dissociation Constant
pka
pKa
Acid dissociation constant
negative log
Solubility
Solubility
N/A
Not utilized as a parameter by the code.
ppm
Henry's Law Constant
Henry
Henry
Henry's law constant
atm m3 mol-1
Vapor Pressure
VPress
N/A
Not utilized as a parameter by the code.
mm Hg
Octanol-water partition
coefficient
LogKow
LogKow
Log octanol-water partition coefficient
unitless
KPSED
KPSed
KPSed
Detritus-water partition coefficient
L/kgOC
KOMRe&DOM
KOMRefrDOM
KOMRe&DOM
Reftractory DOM to Water Partition Coefficient
L/kgOM
Uptake Rate (K1) Detritus
K1 Detritus
K1 Detr
Uptake rate constant for organic matter, default of 1.39
L/kg dry day
Cohesives K1
CohesivesKl
K1
Uptake rate constant for cohesives
L/kg dry day
Cohesives K2
CohesivesK2
K2
Depuration rate constant for cohesives
day1
Cohesives Kp
CohesivesKp
Kp
Partition coefficient for cohesives
L/kg dry
Non-Cohesives K1
NonCohKl
K1
Uptake rate constant for non-cohesives class 1
L/kg dry day
Non-Cohesives K2
NonCohK2
K2
Depuration rate constant for non-cohesives class 1
day1
Non-Cohesives Kp
NonCohKp
Kp
Partition coefficient for non-cohesives class 1
L/kg dry
337
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
Non-Cohesives2 K1
NonCoh2Kl
K1
Uptake rate constant for non-cohesives class 2
L/kg dry day
Non-Cohesives2 K2
NonCoh2K2
K2
Depuration rate constant for non-cohesives class 2
day1
Non-Cohesives2 Kp
NonCoh2Kp
Kp
Partition coefficient for non-cohesives class 2
L/kg dry
Activation Energy for
Temperature
En
En
Arrhenius activation energy
cal/mol
Rate of Anaerobic Microbial
Degradation
KMDegrAnaerobic
KAnaerobic
Decomposition rate at 0 g/m3 oxygen
1/d
Max. Rate of Aerobic
Microbial Degradation
KMDegrdn
KMDegrdn
Maximum (microbial) degradation rate
1/d
Uncatalyzed hydrolysis
constant
KUnCat
KUncat
The measured first-order reaction rate at ph 7
1/d
Acid catalyzed hydrolysis
constant
KAcid
KAcid
Pseudo-first-order acid-catalyzed rate constant for a given ph
L/mol • d
Base catalyzed hydrolysis
constant
KBase
KBase
Pseudo-first-order rate constant for a given ph
L/mol • d
Photolysis Rate
PhotolysisRate
KPhot
Direct photolysis first-order rate constant
1/d
Oxidation Rate Constant
OxRateConst
N/A
Not utilized as a parameter by the code.
U mol d
Weibull Shape Parameter
Weibull_Shape
Shape (Internal Model)
Parameter expressing variability in toxic response; default is 0.33
unitless
Weibull Slope Factor
WeibullSlopeF actor
Slope Factor (External
Model)
Slope at EC50 multiplied by EC50
slope • ug/L
Chemical is a Base
ChemlsBase
Compound is a base
True if the compound is a base
True/F alse
This Chemical is a PFA
IsPFA
Compound is a PFA
True if the compound is a perfluorinated surfactant
True/F alse
Type of PFA
PFAType
carboxylate / sulfonate
Sulfonate group and carboxylate group
carboxylate /
sulfonate
Perfluoralkyl Chain Length
PFAChainLength
ChainLength
Length of perfluoroalkyl chain
Integer
Kom for Sediments (PFA)
PFASedKom
Kom for Sediments
Organic matter partition coefficient for the PFA
L/kg
338
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
USER INTERFACE
INTERNAL
TECH DOC
Description
UNITS
BCF for Algae (PFA)
PFAAlgBCF
BCF for Algae
Bioconcentration Factor for the PFA to algae
L/kg
BCF for Macrophytes (PFA)
PFAMacroBCF
BCF for Macrophytes
Bioconcentration Factor for the PFA to macrophytes
L/kg
Use BCF to Estimate Uptake
BCFUptake
SiteRecord
Site Underlying
Data
For each water body simulated, the following
Parameters are required
Site Name
SiteName
N/A
Site's Name. Used for Reference only.
N/A
Max Length (or reach)
SiteLength
Length
Maximum effective length for wave setup
km
Vol.
Volume
Volume
Initial volume of site (must be copied into state var.)
m3
Surface Area
Area
Area
Site area
m2
Estuary Site Width
SiteWidth
Width
Width of estuary
m
Mean Depth
ZMean
ZMean
Mean depth, (initial condition if dynamic mean depth is selected)
M
Maximum Depth
ZMax
ZMax
Maximum depth
M
Ave. Temp, (epilimnetic or
hypolimnetic)
TempMean
TempMean
Mean annual temperature of epilimnion (or hypolimnion)
°C
Epilimnetic Temp. Range (or
hypolimnetic)
TempRange
TempRange
Annual temperature range of epilimnion (or hypolimnion)
°C
Latitude
Latitude
Latitude
Latitude
Deg, decimal
Altitude (affects oxygen sat.)
Altitude
Altitude
Site specific altitude
m
Average Light
LightMean
LightMean
Mean annual light intensity
Langleys/day
Annual Light Range
LightRange
LightRange
Annual range in light intensity
Langleys/day
Total Alkalinity
AlkCaC03
N/A
Not utilized as a parameter by the code.
mg/L
Hardness as CaC03
HardCaC03
N/A
Not utilized as a parameter by the code.
mg CaC03 / L
Sulfate Ion Cone
S04Conc
N/A
Not utilized as a parameter by the code.
mg/L
339
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
Total Dissolved Solids
TotalDissSolids
N/A
Not utilized as a parameter by the code.
mg/L
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
Enclosure Wall Area
EnclWallArea
EnclWallArea
Area of experimental enclosures walls; only relevant to enclosure
m2
Mean Evaporation
MeanEvap
MeanEvap
Mean annual evaporation
inches / year
Extinct. CoeffWater
ECoeffWater
ExtinctH20
Light extinction ofwavelength 312.5 nm in pure water
1/m
Extinct. Coeff Sediment
ECoeffSed
ECoeffSed
Light extinction due to inorganic sediment in water
l/(mg/m3)
Extinct. Coeff DOM
ECoeffDOM
ECoeffDOM
Light extinction due to dissolved organic matter in water
l/(mg/m3)
Extinct. Coeff POM
ECoeffPOM
ECoeffPOM
Light extinction due to particulate organic matter in water
l/(mg/m3)
Baseline Percent
Embeddedness
BasePercentEmbed
baseline embeddedness
Observed embeddedness that is used as an initial condition
percent (0-100)
Minimum Volume Frac.
Min_Vol_Frac
Minimum Volume Frac.
Fraction of initial condition that is the minimum volume of a site
frac. of Initial
Condition
Auto Select Eqn. for
reaeration
UseCovar
Covar
Boolean to determine whether user is entering reaeration
coefficient
boolean
Enter KReaer
KReaer
KReaer
Depth-averaged reaeration coefficient
1/d
Total Length
TotalLength
TotLength
Total river length for calculating Nhytoplankton retention
km
Watershed Area
WaterShedArea
Watershed
Watershed area for estimating total river length (above)
km2
Fractal Dimension
FractalD
FractalDMarsh
Fractal dimension of marsh-water interface for the site.
unitless
Fractal D. Refuge
Coefficient
FD_Refuge_Coeff
Coeff
Fractal dimension Refuge coefficient (-0.5 to 100 with the lowest
values providing the strongest Refuge effect).
unitless
Elalf Sat Oyster Refuge
HalfSatOysterRefuge
HalfSat (eqn. 95)
Half-saturation constant for oysters in terms of providing refuge
from feeding
g/m2
M2, Amplitude & Epoch
amplitude 1, kl
M2
Estuary Only - principal lunar semidiurnal constituent
m, deg. Local
Siderial Time
(LST)
S2, Amplitude & Epoch
amplitude2, k2
S2
Estuary Only - principal solar semidiurnal constituent
m, deg. LST
N2, Amplitude & Epoch
amplitude3, k3
N2
Estuary Only - larger lunar elliptic semidiurnal constituent
m, deg. LST
K1, Amplitude & Epoch
amplitude4, k4
Kl
Estuary Only - lunar diurnal constituent
m, deg. LST
340
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
Ol, Amplitude & Epoch
amplitude5, k5
Ol
Estuary Only - lunar diurnal constituent
m, deg. LST
SSA, Amplitude & Epoch
amplitude6, k6
SSA
Estuary Only - solar semiannual constituent
m, deg. LST
SA, Amplitude & Epoch
amplitude7, k7
SA
Estuary Only - solar annual constituent
m, deg. LST
PI, Amplitude & Epoch
amplitude8, k8
PI
Estuary Only - solar diurnal constituent
m, deg. LST
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
SiteRecord (Stream-
Specific)
Site Underlying
Data
For each stream simulated, the following
Parameters are required
Channel Slope
Channel_Slope
Slope
Slope of channel
m/m
Maximum Channel Depth
Before Flooding
Max_Chan_Depth
Max_Chan_Depth
Depth at which flooding occurs
m
Sediment Depth
SedDepth
SedDepth
Maximum sediment depth
m
Stream Type
StreamType
Stream Type
Concrete channel, dredged channel, natural channel
Choice from List
use the below value
U seEnteredManning
Do not determine Manning coefficient from streamtype
true/false
Mannings Coefficient
EnteredManning
Manning
Manually entered Manning coefficient.
s / m1/3
Percent Riffle
PctRiffle
Riffle
Percent riffle in stream reach
%
Percent Pool
PctPool
Pool
Percent pool in stream reach
%
SiteRecord (Sand-Silt-
Clay Specific)
Site Underlying
Data
For each stream with the inorganic sediments
model included, the following Parameters are
required
Silt: Critical Shear Stress for
Scour
tssilt
TauScourSed
Critical shear stress for scour of silt
kg/m2
341
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
Silt: Critical Shear Stress for
Deposition
tdep_silt
TauDepSed
Critical shear stress for deposition of silt
kg/m2
Silt: Fall Velocity
FallVel_silt
VTSed
Terminal fall velocity of silt
m/s
Clay: Critical Shear Stress
for Scour
ts_clay
TauScourSed
Critical shear stress for scour of clay
kg/m2
Clay: Critical Shear Stress
for Deposition
tdep_clay
TauDepSed
Critical shear stress for deposition of clay
kg/m2
Clay: Fall Velocity
FallVel_clay
VTSed
Terminal fall velocity of clay
m/s
ReminRecord
Remineralization
Data
For each simulation, the following Parameters are
required (pertaining to organic matter)
Max. Degrdn Rate, labile
DecayMax_Lab
DecayMax
Maximum decomposition rate
g/g'd
Max Degrdn Rate, Refrac
DecayMax_Refr
ColonizeMax
Maximum colonization rate under ideal conditions
g/g'd
Temp. Response Slope
Qio
Q10
Not utilized as a parameter by the code.
Optimum Temperature
TOpt
TOpt
Optimum temperature for degredation to occur
°C
Maximum Temperature
TMax
TMax
Maximum temperature at which degradation will occur
°C
Min. Adaptation Temp
TRef
TRef
Not utilized as a parameter by the code.
°C
Min pH for Degradation
pHMin
pHMin
Minimum ph below which limitation on biodegradation rate
occurs.
PH
Max pH for Degradation
pHMax
pHMax
Maximum ph above which limitation on biodegradation occurs.
PH
KNitri, Max Rate of Nitrif.
KNitri
KNitri
Maximum rate of nitrification
1/day
KDenitri Bottom (max.)
KDenitri_Bot
KDenitriBottom
Maximum rate of denitrification at the sed/water interface
1/day
KDenitri Water (max.)
KDenitri_Wat
KDenitriWater
Maximum rate of denitrification in the water column
1/day
P to Organics, Labile
P20rgLab
P20rgLab
Ratio of phosphate to labile organic matter
fraction dry weight
N to Organics, Labile
N20rgLab
N20rgLab
Ratio of nitrate to labile organic matter
fraction dry weight
342
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
P to Organic s, Refractory
P20rgRefr
P20rgRefr
Ratio of phosphate to refractory organic matter
fraction dry weight
N to Organic s, Refractory
N20rgRefr
N20rgRefr
Ratio of nitrate to refractory organic matter
fraction dry weight
P to Organics, Diss. Labile
P20rgDissLab
P20rgDissLab
Ratio of phosphate to dissolved labile organic matter
fraction dry weight
N to Organics, Diss. Labile
N20rgDissLab
N20rgDissLab
Ratio of nitrate to dissolved labile organic matter
fraction dry weight
P to Organics, Diss. Refr.
P20rgDissRefr
P20rgDissRefr
Ratio of phosphate to dissolved refractory organic matter
fraction dry weight
N to Organics, Diss. Refr.
N20rgDissRefr
N20rgDissRefr
Ratio of nitrate to dissolved refractory organic matter
fraction dry weight
02 : Biomass, Respiration
02Biomass
02Biomass
Ratio of oxygen to organic matter
unitless ratio
CBODu to BOD5
conversion factor
BOD5_CBODu
N/A
Not utilized as a parameter by the code.
unitless ratio
02: N, Nitrification
02N
02N
Ratio of oxygen to nitrogen
unitless ratio
Detrital Sed Rate (KSed)
KSed
KSed
Intrinsic sedimentation rate
m/d
Temperature of Obs. KSed
KSedTemp
TemperatureReference
Reference temperature of water for calculating detrital sinking rate
deg. c
Salinity of Obs. KSed
KSedSalinity
Salinity Reference
Reference salinity of water for calculating detrital sinking rate
%o
P04, Anaerobic Sed.
PSedRelease
N/A
Not utilized as a parameter by the code.
g/m2-d
NH4, Aerobic Sed.
NSedRelease
N/A
Not utilized as a parameter by the code.
g/m2-d
Wet to Dry Susp. Labile
Wet2DrySLab
Wet2DrySLab
Wet weight to dry weight ratio for suspended labile detritus
ratio
Wet to Dry Susp. Refr
Wet2DrySRefr
Wet2DrySRefr
Wet weight to dry weight ratio for suspended refractory detritus
ratio
Wet to Dry Sed. Labile
Wet2DryPLab
Wet2DryPLab
Wet weight to dry weight ratio for particulate labile detritus
ratio
Wet to Dry Sed. Refr.
Wet2DryPRefr
Wet2DryPRefr
Wet weight to dry weight ratio for particulate refractory detritus
ratio
KD, P to CaC03
KDPCalcite
KD_P_Calcite
Partition coefficient for phosphorus to calcite
L/kg
ZooRecord
Animal Underlying
Data
For each animal in the simulation, the following
Parameters are required
Animal
AnimalName
N/A
Animal's Name. Used for Reference only.
N/A
Animal Type
Animal_Type
Animal Type
Animal type (fish, pelagic invert, benthic invert, benthic insect)
Choice from List
Taxonomic Type or Guild
Guild_Taxa
Taxonomic type or guild
Taxonomic type or trophic guild
Choice from List
343
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
Toxicity Record
ToxicityRecord
N/A
Associates animal with appropriate toxicity data
Choice from List
Half Saturation Feeding
FHalfSat
FHalfSat
Half-saturation constant for feeding by a predator
g/m3
Maximum Consumption
CMax
CMax
Maximum feeding rate for predator
g/g'd
Min Prey for Feeding
BMin
BMin
Minimum prey biomass needed to begin feeding
g/m3 or g/m2
Sorting: selective feeding
Sorting
Sorting
Fractional degree to which there is selective feeding
Unitless
Burrowlndex
Burrow_Index
Burrowlndex
animal-specific parameter with 0 representing no burrowing
refuge;
Unitless
CanSeekRefuge
CanSeekRefuge
Can Seek Refuge
can this animal, as prey, seek refuge in macrophytes, seagrass, or
oyster bed?
Boolean
Is a Visual Feeder
Visual_Feeder
Is a Visual Feeder
Does this animal feed based on vision, thereby being impeded by
animals seeking refuge in macrophytes, etc.?
Boolean
Susp. Sed. Affect Feeding
SuspSedFeeding
Option to use eqn.
Does suspended sediment affect feeding
Boolean
Slope for Sed. Response
SlopeSSFeed
SlopeSS
Slope for sediment response
Unitless
Intercept for Sed. Resp.
InterceptSSFeed
Intercepts S
Intercept for sediment response
Unitless
Temp Response Slope
Q10
Q10
Slope or rate of change in process per 10°C temperature change
Unitless
Optimum Temperature
TOpt
TOpt
Optimum temperature for given process
°C
Maximum Temperature
TMax
TMax
Maximum temperature tolerated
°C
Min Adaptation Temp
TRef
TRef
Adaptation temperature below which there is no acclimation
°C
Endogenous Respiration
EndogResp
EndogResp
Basal respiration rate at 0° C for given predator
day1
Specific Dynamic Action
KResp
KResp
Proportion assimilated energy lost to specific dynamic action
Unitless
Excretion:Respiration
KExcr
KExcr
Proportionality constant for excretion:respiration
Unitless
N to Organic s
N20rglnit
N20rg
Fixed ratio of nitrate to organic matter for given species
fraction dry weight
P to Organic s
P20rglnit
P20rg
Fixed ratio of phosphate to organic matter for given species
fraction dry weight
Wet to Dry
Wet2Dry
Wet2Dry
Ratio of wet weight to dry weight for given species
Ratio
Gamete : Biomass
PctGamete
PctGamete
Fraction of adult predator biomass that is in gametes
Unitless
Gamete Mortality
GMort
GMort
Gamete mortality
1/d
344
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
Mortality Coefficient
KMort
KMort
Intrinsic mortality rate
1/d
Sensitivity to Sediment
SensToSediment
Sensitivity Categories
Which equation to use for mortality due to sediment
"Zero," "Tolerant,"
"Sensitive,"
"Highly Sensitive"
Ortanism is Sensitive to
Percent Embeddedness
SenstoPctEmbed
N/A
If this checkbox is checked then the organism will be sensitive to
the sites calculated embeddedness as a function of TSS
Boolean
Percent Embeddedness
Threshold
PctEmbedThreshold
embeddedness threshold
value
If the site's calculated embeddedness exceeds this value, mortality
for the organism is set to 100%
percent (0-100)
Carrying Capacity
KCap
KCap
Carrying capacity
g/m2
Average Drift
AveDrift
Dislodge
Fraction of biomass subject to drift per day
fraction / day
Trigger: Deposition Rate
Trigger
Trigger
deposition rate at which drift is accelerated
kg/m2 day
Frac. in Water Column
FracInWaterCol
FraC Water Column
Fraction of organism in water column, differentiates from pore-
water uptake if the multi-layer sediment model is included
Fraction
VelMax
VelMax
VelMax
Maximum water velocity tolerated
cm/s
Removal due to Fishing
Fishing_Frac
fraction fished
Daily loss of organism due to fishing Pressure
Fraction
Mean lifespan
LifeSpan
LifeSpan
Mean lifespan in days
Days
Fraction that is lipid
FishFracLipid
LipidFrac
Fraction of lipid in organism
g lipid/g org. Wet
Mean Wet Weight
MeanWeight
WetWt
Mean wet weight of organism
g wet
Low 02: Lethal Cone
02_LethalConc
LCKnownduratlon
Concentration where there is a known mortality over 24 hours
mg/L (24 hour)
Low02: Pet. Killed
02_LethalPct
PctKilledKj,,,™
The percentage of the organisms killed at the lcknown level above.
Percentage
Low02: EC50 Growth
O2_EC50growth
EC50 duration
Concentration where there is 50% reduction in growth over 24
hours
mg/L (24 hour)
Low02: EC50 Reproduction
O2_EC50repro
EC50 duration
Concentration where there is 50% reduction in reproduction over
24 hours
mg/L (24 hour)
Ammonia Toxicity: LC50,
Total Ammonia (pH=8)
Ammonia_LC50
LC50t8
LC50totalammonia at 20 degrees centigrade and ph of 8
mg/L (ph=8)
Salinity Ingestion Effects
Salmin_Ing, SalMax_Ing,
Salcoeffl_Ing, Salcoeff2_Ing
SalMin, SalMax,
SalCoeffl, SalCoeff2
Parameters used to calculate the effects of the current level of
salinity on ingestion for the given animal
%o, %o, unitless,
unitless
345
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
Salinity Gamete Loss Effects
Salmin Gam, SalMax Gam,
Salcoeffl Gam, Salcoeff2 Gam
SalMin, SalMax,
SalCoeffI, SalCoeff2
Parameters used to calculate the effects of the current level of
salinity on gamete loss for the given animal
%o, %o, unitless
Salinity Respiration Effects
Salmin_Rsp, SalMax_Rsp,
Salcoeffl _Rsp, Salcoeff2_Rsp
SalMin, SalMax,
SalCoeffI, SalCoeff2
Parameters used to calculate the effects of the current level of
salinity on respiration for the given animal
%o, %o, unitless,
unitless
Salinity Mortality Effects
Salmin_Mort, SalMax_Mort,
Salcoeffl _Mort,
Salcoeff2_Mort
SalMin, SalMax,
SalCoeffI, SalCoeff2
Parameters used to calculate the effects of the current level of
salinity on mortality of the given animal
%o, %o, unitless,
unitless
Percent in Riffle
PrefRiffle
PreferenceElabitat
Percentage of biomass of animal that is in riffle, as opposed to run
or pool
%
Percent in Pool
PrefPool
PreferenceElabitat
Percentage of biomass of animal that is in pool, as opposed to run
or riffle
%
Fish spawn automatically,
based on temperature range
Auto Spawn
Does AQUATOX calculate Spawn Dates
true/false
Fish spawn of the following
dates each year
SpawnDatel..3
User entered spawn dates
Date
Fish can spawn an unlimited
number of times...
UnlimitedSpawning
Allow fish to spawn unlimited times each year
true/false
Use Allometric Equation to
Calculate Maximum
Consumption
UseAllom_C
Use allometric consumption equation
true/false
Intercept for weight
dependence
CA
Allometric consumption parameter
real number
Slope for weight dependence
CB
Allometric consumption parameter
real number
Use Allometric Equation to
Calculate Respiration
UseAllom_R
Use allometric consumption respiration
true/false
RA
RA
Intercept for species specific metabolism
real number
RB
RB
Weight dependence coefficient
real number
346
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
Use "Set 1" of Respiration
Equations
UseSetl
Use "Set 1" of Allometric Respiration Parameters
true/false
RQ
RQ
RQ
Allometric respiration parameter
real number
RTL
RTL
RTL
Temperature below which swimming activity is an exponential
function of temperature
°C
ACT
ACT
ACT
Intercept for swimming speed for a lg fish
cm/s
RTO
RTO
RTO
Coefficient for swimming speed dependence on metabolism
s/cm
RK1
RK1
RK1
Intercept for swimming speed above the threshold temperature
cm/s
BACT
BACT
BACT
Coefficient for swimming at low temperatures
1/°C
RTM
RTM
Not currently used as a parameter by the code
RK4
RK4
RK4
Weight-dependent coefficient for swimming speed
real number
ACT
ACT
Intercept of swimming speed vs. Temperature and weight
real number
Preference (ratio)
TropMnt.Pref[ ]
Prefprey,pred
Initial preference value from the animal parameter screen
Unitless
Egestion (frac.)
TropMnt.Egest[ ]
EgestCoeffprey,pred
Fraction of ingested prey that is egested
Unitless
PlantRecord
Plant Underlying
Data
For each Plant in the Simulation, the following
Parameters are required
Plant
PlantName
Plant's name. Used for reference only.
N/A
Plant Type
PlantType
Plant Type
Plant type: (Phytoplankton, Periphyton, Macrophytes, Bryophytes)
Choice from List
Plant is Surface Floating
SurfaceFloating
SurfaceFloating
Is this plant surface floating and therefore subject to a shallowlight
climate as well as excluded from the hypolimnion.
Boolean
Macrophyte Type
Macrophyte_T ype
Macrophyte Type
Benthic, rooted floating, free-floating
Choice from List
Taxonomic Group
T axonomic_T ype
Taxonomic Group
Taxonomic group
Choice from List
Toxicity Record
ToxicityRecord
N/A
Associates plant with appropriate toxicity data
Choice from List
347
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
Saturating Light
LightSat
LightSat
Light saturation level for photosynthesis
ly/d
Use Adaptive Light
UseAdaptiveLight
Adaptive Light
Choice whether to use adaptive light construct
Boolean
Max. Saturating Light
MaxLightSat
user-entered maximum
Maximum light saturation allowed from adaptive light equation
ly/d
Min. Saturating Light
MinLightSat
user-entered minimum
Minimum light saturation allowed from adaptive light equation
ly/d
P Half-saturation
KP04
KP
Half-saturation constant for phosphorus
gP/m3
N Half-saturation
KN
KN
Half-saturation constant for nitrogen
gN/m3
Inorg C Half-saturation
KCarbon
KC02
Half-saturation constant for carbon
gC/m3
Temp Response Slope
Q10
Q10
Slope or rate of change per 10°C temperature change
Unitless
Optimum Temperature
TOpt
TOpt
Optimum temperature
°C
Maximum Temperature
TMax
TMax
Maximum temperature tolerated
°C
Min. Adaptation Temp
TRef
TRef
Adaptation temperature below which there is no acclimation
°C
Max. Photosynthesis Rate
PMax
PMax
Maximum photo synthetic rate
1/d
Photorespiration Coefficient
KResp
KResp
Coefficient of proportionality between. Excretion and
photosynthesis at optimal light levels
Unitless
Resp Rate at 20 deg. C
Resp20
Resp20
Respiration rate at 20°C
g/g'd
Mortality Coefficient
KMort
KMort
Intrinsic mortality rate
g/g'd
Exponential Mort Coeff
EMort
EMort
Exponential factor for suboptimal conditions
g/g'd
P to Photosynthate
P20rg
P20rg
Initial ratio of phosphate to organic matter for given species
fraction dry weight
N to Photosynthate
N20rg
N20rg
Initial ratio of nitrate to organic matter for given species
fraction dry weight
Light Extinction
ECoeffPhyto
EcoeffPhyto
Attenuation coefficient for given alga
l/m-g/m3w
Wet to Dry
Wet2Dry
Wet2Dry
Ratio of wet weight to dry weight for given species
Ratio
Fraction that is lipid
PlantFracLipid
LipidFrac
Fraction of lipid in organism
g lipid/g org. Wet
N Half-saturation Internal
NHalfSatlnternal
NHalfSatinternal
half-saturation constant for intracellular nitrogen
gN/gAFDW
P Half-saturation Internal
PHalfSatlnternal
PHalfSatlnternal
half-saturation constant for intracellular phosphorus
gP / gAFDW
348
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
N Max Uptake Rate
MaxNUptake
MaxNUptake
the maximum uptake rate for nitrogen
gN/gAFDW-d
P Max Uptake Rate
MaxPUptake
MaxPUptake
the maximum uptake rate for phosphorus
gP / gAFDW-d
Min N Ratio
Min_N_Ratio
MinNRatio
the ratio of intracellular nitrogen at which growth ceases
gN/gAFDW
Min P Ratio
Min_P_Ratio
MinPRatio
the ratio of intracellular phosphorus at which growth ceases
gP / gAFDW
Phytoplankton:
C: Chlorophyll a
Plant_to_Chla
CToChla
ratio of carbon to chlorophyll a
g carbon/g chl. a
Phytoplankton:
Sedimentation Rate (KSed)
KSed
KSed
Intrinsic settling rate
m/d
Phytoplankton: Temperature
of Obs. KSed
KSedTemp
Temperature Reference
Reference temperature of water for calculating Nhytoplankton
sinking rate
deg. C
Phytoplankton: Salinity of
Obs. KSed
KSedSalinity
Salinity Reference
Reference salinity of water for calculating Nhytoplankton sinking
rate
%o
Phytoplankton: Exp.
Sedimentation Coeff
ESed
ESed
Exponential settling coefficient
Unitless
Macrophytes: Carrying
Capacity
Carry _Capac
KCap
Macrophyte carrying capacity, converted to g/m3 and used to
calculate washout of free-floating macrophytes
g/m2
Macrophytes: VelMax
Macro_VelMax
VelMax
Velocity at which total breakage occurs
cm/s
Periphyton: Reduction in
Still Water
Red_Still_Water
RedStillWater
Reduction in photosynthesis in absence of current
Unitless
Periphyton: Critical Force
(FCrit)
FCrit
FCrit
Critical force necessary to dislodge given periphyton group
newtons (kg m/s2)
Percent Lost in Slough Event
PctSloughed
FracSloughed
Fraction of biomass lost at one time
%
Percent in Riffle
PrefRiffle
PrefRiffle
Percentage of biomass of plant that is in riffle, as opposed to run or
pool
%
Percent in Pool
PrefPool
PrefPool
Percentage of biomass of plant that is in pool, as opposed to run or
riffle
%
Salinity Photosyn. Effects
Sahnin_Phot, SalMax_Phot,
Salcoeffl_Phot,
Salcoeff2_Phot
SalMin, SalMax,
SalCoeffl, SalCoeff2
Parameters used to calculate the effects of the current level of
salinity on photosynthesis for the given plant
%o, %o, unitless,
unitless
Salinity Mortality Effects
Salrnin_Mort, SalMax_Mort,
Salcoeffl _Mort,
SalMin, SalMax,
SalCoeffl, SalCoeff2
Parameters used to calculate the effects of the current level of
salinity on mortality for the given plant
%o, %o, unitless,
349
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
Salcoeff2_Mort
unitless
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
AnimalToxRecord
Animal Toxicity
Parameters
For each Chemical Simulated, the following
Parameters are required for each animal simulated
LC50
LC50
LC50
Concentration of toxicant in water that causes 50% mortality
|xg/L
LC50 exp time (h)
LC50_exp_time
ObsTElapsed
Exposure time in toxicity determination
H
K2 Elim rate const
K2
K2
Elimination rate constant
1/d
K1 Uptake const
K1
K1
Uptake rate constant, only used if "Enter Kl" option is selected
L / kg dry day
BCF
BCF
BCF
Bioconcentration factor, only used if "Enter BCF" option is
selected
L / kg dry
Biotrnsfm rate
BioRateConst
BioRateConst
Percentage of chemical that is biotransformed to
Specific daughter products
1/d
EC50 growth
EC50_growth
EC50Growth
External concentration of toxicant at which there is a 50%
reduction in growth
Hg/L
Growth exp (h)
Growth_exp_time
ObsTElapsed
Exposure time in toxicity determination
H
EC 50 repro
EC50_repro
EC50Repro
External concentration of toxicant at which there is a 50%
reduction in reprod
|xg/L
Repro exp time (h)
Repro_exp_time
ObsTElapsed
Exposure time in toxicity determination
H
Mean wet weight (g)
Mean_wet_wt
WetWt
Mean wet weight of organism
G
Lipid Frac
Lipid_frac
LipidFrac
Fraction of lipid in organism
g lipid/g wet wt.
Drift Threshold (|xg/L)
Drift_Thresh
Drift Threshold
Concentration at which drift is initiated
Hg/L
LC50 Slope
LC50_Slope
Slope Factor (External
Model)
Animal-specific Slope at LC50 multiplied by LC50. If left blank
or zero, the value from the chemical record is used.
slope • ug/L
TPlantT oxRecor d
Plant Toxicity
Parameter
For each Chemical Simulated, the following
Parameters are required for each plant simulated
EC 50 photo
EC50_photo
EC50Photo
External concentration of toxicant at which there is 50% reduction
in photosynthesis
Hg/L
350
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
EC50 exp time (h)
EC50_exp_time
ObsTElapsed
Exposure time in toxicity determination
H
EC50 dislodge
EC50_dislodge
EC50Dislodge
For periphyton only: external concentration of toxicant at which
there is 50% dislodge of periphyton
Hg/L
K2 Elim rate const
K2
K2
Elimination rate constant
1/d
K1 Uptake const
K1
K1
Uptake rate constant, only used if "Enter Kl" option is selected
L / kg dry day
BCF
BCF
BCF
Bioconcentration factor, only used if "Enter BCF" option is
selected
L / kg dry
Biotrnsfm rate
BioRateConst
BioRateConst
Percentage of chemical that is biotransformed to
Specific daughter products
1/d
LC50
LC50
LC50
Concentration of toxicant in water that causes 50% mortality
Hg/L
LC50 exp.time (h)
LC50_exp_time
ObsTElapsed
Exposure time in toxicity determination
H
Lipid Frac
Lipid_frac
LipidFrac
Fraction of lipid in organism
g lipid/g org. Wet
LC50 Slope
LC50_Slope
Slope Factor (External
Model)
Plant-specific Slope at LC50 multiplied by LC50. If left blank or
zero, the value from the chemical record is used.
slope • ug/L
TChemical
Chemical
Parameters
For each Chemical to be simulated, the following
Parameters are required
Initial Condition
InitialCond
Initial Condition
Initial Condition of the state variable
Hg/L
Gas-phase conc.
Tox_Air
Toxicantair
Gas-phase concentration of the pollutant
g/m3
Loadings from Inflow
Loadings
Inflow Loadings
Daily loading as a result of the inflow of water
Hg/L
Loadings from Point Sources
Alt_Loadings[Pointsource]
Point Source Loadings
Daily loading from point sources
g/d
Loadings from Direct
Precipitation
Alt_Loadings[Direct Precip]
Direct Precipitation Load
Daily loading from direct precipitation
g/m2 -d
Nonpoint-source Loadings
Alt_Loadings[NonPointsource]
Non-Point Source Loading
Daily loading from non-point sources
g/dTox_AirGas-
phase
concentrationg/m3
Biotransformation
BioTrans[ ]
Biotransform
Percentage of chemical that is biotransformed to specific daughter
products
%
351
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
TRemineralize
Nutrient Parameters
For each Nutrient to be simulated, 02 and CO2,
the following Parameters are required
Initial Condition
InitialCond
Initial Condition
Initial Condition of the state variable (TotP or TotN optional)
mg/L
Loadings
Inflow Loadings
Daily loading as a result of the inflow of water (TotP or TotN
optional)
mg/L
Loadings from Point Sources
Alt_Loadings[Pointsource]
Point Source Loadings
Daily loading from point sources
g/d
Loadings from Direct
Precipitation
Alt_Loadings[Direct Precip]
Direct Precipitation Loa
Daily loading from direct precipitation
g/m2 -d
Non-point source loadings
Alt_Loadings[NonPointsource]
Non-Point Source Loading
Daily loading from non-point sources
g/d
Fraction of Phosphate
Available
FracAvail
Fraction of phosphate loadings that is available versus that which
is tied up in minerals
Unitless
TSedDetr
Sed. Detritus
Parameters
For the Labile and Refractory Sedimented
Detritus compartments, the following Parameters
are required
Initial Condition
InitialCond
Initial Condition
Initial Condition of the labile or refractory sedimented detritus
g/m2
Initial Condition
TT oxicant.InitialCond
Toxicant Exposure
Initial Toxicant Exposure of the state variable, for each chemical
lig/kg
Loadings from Inflow
Loadings
Inflow Loadings
Daily loading of the sedimented detritus as a result of the inflow of
water
mg/L
(Toxicant) Loadings
TToxicant.Loads
Tox Exposure of Inflow L
Daily parameter; Toxicant Exposure of each type of inflowing
detritus, for each chemical
lig/kg
352
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
TDetritus
Susp & Dissolved
Detritus
For the Suspended and Dissolved Detritus
compartments, the following Parameters are
required
Initial Condition
InitialCond
Initial Condition
Initial Condition of suspended & dissolved detritus, as organic
matter, organic carbon, or biochemical oxygen demand
mg/L
Initial Condition: %
Particulate
Percent_Part_IC
Percent of Initial Condition that is particulate as opposed to
dissolved detritus
Percentage
Initial Condition: %
Refractory
Percent_Refr_IC
Percent of Initial Condition that is refractory as opposed to labile
detritus
Percentage
Inflow Loadings
Loadings
Inflow Loadings
Daily loading as a result of the inflow of water
mg/L
Dissolved / Particulate
Breakdown
Percent_Part
Percent Particulate Inflow,
Point Source, Non-Point
Source
Three constant or time-series parameters; % of each type of
loading that is particulate as opposed to dissolved detritus
Percentage
Labile / Refractory
Breakdown
Percent_Refr
Percent Refractory Inflow,
Point Source, Non-Point
Source
Three constant or time-series parameters; % of each type of
loading that is refractory as opposed to labile detritus
Percentage
Loadings from Point Sources
Alt_Loadings[Pointsource]
Point Source Loadings
Daily loading from point sources
§ organic matter^
Nonpoint-source Loadings
(Associated with Organic
Matter)
Alt_Loadings
Non-Point Source Loading
Daily loading from non-point sources
§ organic matter^
(Toxicant) Initial Condition
TT oxicant.InitialCond
Toxicant Exposure
Initial Toxicant Exposure of the suspended and dissolved detritus
lig/kg
(Toxicant) Loadings
(associated with Organic
Matter)
TToxicant.Loads
Tox Exposure of Inflow
Loading
Daily parameter; Toxicant Exposure of each type of inflowing
detritus, for each chemical
lig/kg
353
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
TBuried Detritus
Buried Detritus
For Each Type of Buried Detritus, the following
Parameters are required
Initial Condition
InitialCond
Initial Condition
Initial Condition of the labile and refractory buried detruitus
Kg/cu. m
(Toxicant) Initial Condition
TT oxicant.InitialCond
Toxicant Exposure
Initial Toxicant Exposure of the labile and refractory buried
detritus , for each chemical simulated
Kg/cu. m
TPlant
Plant Parameters
For each plant type simulated, the following
Parameters are required
Initial Condition
InitialCond
Initial Condition
Initial Condition of the plant
mg/L or g/m2 dry
Loadings from Inflow
Loadings
Inflow Loadings
Daily loading as a result of the inflow of water
mg/L or g/m2 dry
(Toxicant) Initial Condition
TT oxicant.InitialCond
Toxicant Exposure
Initial Toxicant Exposure of the plant
lig/kg
(Toxicant) Loadings
TToxicant.Loads
Tox Exposure of Inflow L
Daily parameter; Toxicant exposure of the Inflow Loadings, for
each chemical simulated
lig/kg
TAnimal
Animal Parameters
For each animal type simulated, the following
Parameters are required
Initial Condition
InitialCond
Initial Condition
Initial Condition of the animal
mg/L or g/sq.m
also expressed as
g/m2
Loadings from Inflow
Loadings
Inflow Loadings
Daily loading as a result of the inflow of water
mg/L or g/sq. m
(Toxicant) Initial Condition
Ttoxicant.InitialCond
Toxicant Exposure
Initial Toxicant Exposure of the animal
lig/kg
(Toxicant) Loadings
TToxicant.Loads
Tox Exposure of Inflow L
Daily parameter; toxic exposure of the Inflow Loadings, for each
chemical simulated
lig/kg
Preference (ratio)
TrophlntArray.Pref
Prefprey, pred
For each prey-type ingested, a preference value within the matrix
of preferences
Unitless
Egestion (frac.)
TrophlntArray.ECoeff
EgestCoeff
For each prey-type ingested, the fraction of ingested prey that is
egested
Unitless
354
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
TVolume
Volume Parameters
For each segment simulated, the following water
flow parameters are required
Initial Condition
InitialCond
Initial Condition
Initial Condition of the water volume .
m3
Water volume
Volume
Volume
Choose method of calculating volume; choose between Manning's
equation, constant volume, variable depending upon inflow and
discharge, or use known values
cu. M
Inflow of Water
InflowLoad
Inflow of Water
Inflow of water; daily parameter, can choose between constant and
dynamic loadings
m3/d cu m/d
Discharge of Water
DischargeLoad
Discharge of Water
Discharge of water; daily parameter, can choose between constant
and dynamic loadings
m3/d
Site Characteristics
Site Characteristics
The following Parameters are required
Site Type
SiteType
Site Type
Site type affects many portions of the model.
Pond, Lake,
Stream, Reservoir,
Enclosure, Estuary
Frac. of Site that is Shaded
Shade
user input shade
Fraction of site that is shaded, time-series
Fraction
Water Velocity
DynVelocity
user entered velocity
Optional, time series of run velocities
cm/s
Site Mean Depth
DynZMean
user entered mean depth
Optional, time series of mean depth for site
M
Temperature
Temperature
Temperature Parameters Required
Initial Condition
InitialCond
Initial condition
Initial temperature of the segment or layer (if vertically stratified
°C
Could this system stratify
could system vertically stratify
true/false
Valuation or loading
Temperature of the segment. Can use annual means for each
stratum and constant or dynamic values
°C
355
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
Wind
Wind
Wind parameters required
Initial Condition
InitalCond
Initial wind velocity 10 m above the water
m/s
Mean Value
Mean Value
Mean wind velocity
m/s
Wind Loading
Wind
Wind
Daily parameter; wind velocity 10 m above the water; 1, can
choose default time series, constant or dynamic loadings
m/s
Light
Light
Light Parameters Required
Initial Condition
Light
Light
ly/d
Loading
Loadsrec
Daily parameter; avg. light intensity at segment top; can choose
annual mean, constant loading or dynamic loadings
Photoperiod
Photoperiod
Photoperiod
Fraction of day with daylight; optional, can be calculated from
latitude
hr/d
pH
pH
pH Parameters Required
Initial Condition
InitialCond
Initial pH value
PH
State Variable Valuation
PH
PH
pH of the segment; can choose constant or daily value.
PH
Mean alkalinity
alkalinity
alkalinity
mean Gran alkalinity (if dynamic pH option selected)
|xeq CaC03/L
Sand / Silt / Clay
TSediment
Inorganic Sediment
Parameters
If the inorganic sediments model is included in
AQUATOX, the following Parameters are required
for sand, silt, and clay
Initial Susp. Sed.
InitialCond
Initial Condition
Initial Condition of the sand, silt, or clay
mg/L
Frac in Bed Seds
FracInBed
FracSed
Fraction of the bed that is composed of this inorganic sediment.
Fractions of sand, silt, and clay must add to 1.0
Fraction
Loadings from Inflow
Loadings
Inflow Loadings
Daily sediment loading as a result of the inflow of water
mg/L
356
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
Loadings from Point Sources
Alt_Loadings[Pointsource]
Point Source Loadings
Daily loading from point sources
g/d
Loadings from Direct
Precipitation
Alt_Loadings[Direct Precip]
Direct Precipitation Loa
Daily loading from direct precipitation
Kg-d
Non-point source loadings
Alt_Loadings[NonPointsource]
Non-Point Source Loading
Daily loading from non-point sources
g/d
Multi-Layer Sediment
Model
Global SedData
Multi-layer
Sediment
Parameters
If the multi-layer sediment model is included in
AQUATOX, the following general parameters are
required
Densities [Organic and
Inorganic Components]
Densities
Density Sed
Density of each organic and inorganic component of the sediment
bed.
g/cm3
Multi-Layer Sediment
Model
Active Layer SedData
Multi-layer
Parameters
If the multi-layer sediment model is included these
parameters are required for the active layer only
Max Thickness of Active
Layer
MaxUpperThick
user defined maximum
thickness
Maximum thickness of the active layer before it becomes split into
multiple layers
M
Min Thickness of Active
Layer
BioTurbThick
user defined minimum
thickness
Minimum thickness of active layer before it is added to the layer
below it
M
Cohesives, NonCohesives,
Daily Scour
LScour
Erode Sed
Scour of this sediment to the water column above
g/d
Cohesives, NonCohesives,
Daily Deposition
LDeposition
DepositSed
Deposit of this sediment from the water column
g/d
Cohesives only, Erosion
Velocity
LErodVel
ErodeVel
User input time-series of cohesives erosion velocities, used to
calculate scour of organics
m/d
Cohesives only, Deposition
Velocity
LDepVel
DepVel
User input time-series of cohesives deposition velocities, used to
calculate deposition of organics
m/d
Multi-Layer Sediment
Model
Each Layer SedData
Multi-layer
Parameters
If the multi-layer sediment model is included these
parameters are required for each layer modeled
Thickness
BedDepthIC
thickness
Initial thickness of each modeled layer
M
Diffusion Coefficient for top
of sediment layer
UpperDispCoeff
DiffCoeff
Dispersion coefficient
m2/d
357
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
Pore Water Init. Cond.
TPoreWater.InitialCond
Concsed Initial Cond.
Concentration of pore water initial condition
m3 water / m2
RDOM, LDOM PoreW,
Initial Cond
TDOMPorewater.InitialCond
Concsed Initial Cond.
Concentration of refractory or labile DOM in pore water, initial
condition
g/m3
Cohesives, NonCohesives,
Initial Cond
TBottomSediment.InitialCond
Concsed Initial Cond.
Concentration of inorganic sediments in the layer, initial condition
gm2
R Detr Sed, L Detr Sed,
Initial Cond
TBuriedSed.InitialCond
Concsed Initial Cond.
Concentration of refractory and labile organic sediments in the
layer, initial condition
g/m2 dry
Chemical Exposures
[Component] T ox.InitialCond
Toxicant BottomSed Initial
Cond.
Concentration of relevant toxicant in element of sediment layer
jxg/L pore water,
ug/kg solids
Trophic Interactions,
BCFs for Shorebirds
Gull Parameters
Shorebirds
If the shorebird model is included in a simulation,
the following Parameters are required
Preference (ratio)
GullPref
prPf
1 IClpj-gy^ pred
For each prey-type ingested, a preference value within the matrix
of preferences
Unitless
Biomagnification Factor
GullBMF
bmfTox
Biomagnification factor for this chemical in gull
Unitless
Clearance Rate
GullClear
ClearTox
Clearance rate for the given toxicant in gulls
day1
Link Between Two
Segments
TSegmentLink
Multi-Segment
Model
If the multi-segment model is used for a simulation,
the following Parameters are required for each link
between segments
Type of Link
LinkType
two types of linkages
Indicates whether linkage is unidirectional or bidirectional
"cascade" or
"feedback"
Link Name
Name
Used for the user to keep track of linkages
String
FromSeg, ToSeg
FromID, ToID
Used for the model to keep track of linkages
existing segments
Characteristic Length
CharLength
CharLength
Characteristic mixing length, feedback links only
M
Water Flow Data
WaterFlowData
Discharge
Time-series of water flow from one segment to the next
m3/d
Dispersion Coeff
DiffusionData
DiffusionThlsSeg
Time-series of dispersion coefficients between two segments,
feedback links only
m2/d
358
-------
AOUATOX (RELEASE 3.2) TECHNICAL DOCUMENTATION
APPENDICES
USER INTERFACE
INTERNAL
TECH DOC
DESCRIPTION
UNITS
XSection of Boundary
XSectionData
Area
Time-series of cross sectional areas between two segments,
feedback links only
m2
BedLoadInl0gsmics
BedLoad
Bedloadupstreamlink
Time-series of bedload from the upstream segment to the
downstream segment
g/d
359
------- |