|
PRODUCTIVITY MEASUREMENTS AND SIMULATION MODELS
OF A SHALLOW ESTUARINE ECOSYSTEM
RECEIVING A THERMAL PLUME
AT CRYSTAL RIVER, FLORIDA
By
WADE HAMPTON BARNES SMITH
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA IN PARTIAL
FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1976
ACKNOWLEDGEMENTS
Many of the ideas presented here were shaped in
discussions among the entire systems ecology group of
the Department of Environmental Engineering Sciences.
Special acknowledgements go to major professor H. T.
Odum for his stimulation, guidance, and the distinct
privilege of participating in this program. My super-
visory committee included P. L. Brezonik, T. Bullock,
J. Ewel, and S. C. Snedaker.
This work was supported, by contract No. GEC 159,
918-200-188.19 (Models and Measurements for Determining
the Role of the Power Plants and Cooling Alternatives
at Crystal River, Florida) between the Florida Power
Corporation and the University of Florida Systems Ecology
Program, Department of Environmental Engineering Sciences,
H. T. Odum, principal investigator.
Many people helped in the field and with data
workup: J. Bevis, N. Black, W. Boynton, C. High, D.
Hinck, M. Homer, M. Kemp, M. Lehman, H. McKellar, A.
Merriam, F. Ramsey, and D. Young. Analog computers used
in this study were maintained by A. Copsey, and J. Murphey,
who also provided programming assistance. Progress in
this study was much facilitated by K. Garrison, J. Johnson,
D. McMullin, and W. Trowell of the Florida Power Corpora-
tion.
Use of the R. V. Susio was provided by the State
University System Institute of Oceanography.
iii
TABLE'OF CONTENTS
Page
ACKNOWLEDGEMENTS . . . . ii
LIST OF TABLES . . . ... vii
LIST OF FIGURES . . . . x
ABSTRACT . . . . . xvii
INTRODUCTION . . . . 1
System Adaptation, Environmental Impact,
and Thermal Loading of the Estuary at
Crystal River, Florida . . 5
Models for Gaining an Overview of the
Estuary and Power Plant at Crystal
River . . 9
Previous Studies of Thermally Affected
Aquatic Ecosystems . .. .... 16
Description of Study Area at Crystal
River . . . . 21
Other Studies of the Crystal River
Region . . . . 31
Previous Simulation Models of Marine Eco-
systems, Diurnal Oxygen Dynamics,
Temperature, and the Effects of Power
Plants on Ecosystems . . 33
Plan of Study . . 39
METHODS . . . . .
Metabolic Measurements . . .
Other Field Measurements . .
DATA ASSEMBLED FROM OTHER PHASES OF THE CRYSTAL
RIVER PROJECT AND ELSEWHERE . .
Energy Sources and Inflows Affecting the
Inner Bay . . . .
Stocks of the Inner Bay . .
RESULTS
DISCUSS
Metabolism Measurements . . .
Model Diagrams for Comparing Ecosystems
Affected and Unaffected by the
Discharge Plume . . .
Simulation Model of Diurnal Properties
of the Inner Bay Ecosystem . .
Simulation Model of Seasonal Properties
of the Inner Bay Ecosystem . .
ON . . . . .
Seasonal Patterns of the Ecosystems at
Crystal River . . .
Comparisons of the Ecosystems at Crystal
River and Adaptation to the Thermal
Discharge . . . .
Predictions of the Effect of the Operation
of Unit Three at Crystal River ..
Energy Costs of Alternatives to Estuarine
Cooling of the Thermal Discharge at
Crystal River . . .
Page
41
41
62
69
69
78
116
116
149
154
201
223
223
239
252
253
Page
APPENDICES . . . . . 256
A EXPLANATION OF THE ENERGY SYMBOLS
USED IN THIS STUDY . ... .257
B GRAPHICAL ANALYSES OF DIURNAL STUDIES
OF COMMUNITY METABOLISM IN THE INNER
BAY AFFECTED BY THE THERMAL DISCHARGE
PLUME AND IN THE FORT ISLAND AND
HODGES ISLAND AREAS AWAY FROM THE
INFLUENCE OF THE THERMAL DISCHARGE 262
C INITIAL AND MAXIMUM VALUES OF STOCKS AND
FLOWS, HEAT BUDGET CALCULATIONS, CAL-
CULATION OF TRANSFER COEFFICIENTS,
SCALED EQUATIONS, POTENTIOMETER SET-
TINGS, FUNCTION GENERATOR SET-UP,
AND ANALOG COMPUTER PATCHING DIAGRAM
FOR DIURNAL SIMULATION MODEL OF
INNER BAY . . ... 311
D INITIAL AND MAXIMUM VALUES OF STOCKS
AND FLOWS, CALCULATION OF TRANSFER
COEFFICIENTS, SCALED EQUATIONS,
POTENTIOMETRIC SETTINGS, FUNCTION
GENERATOR SET-UP, AND ANALOG COM-
PUTER PATCHING DIAGRAM FOR SEASONAL
SIMULATION MODEL OF THE INNER BAY 356
E DOCUMENTATION OF DATA USED IN SUMMARY
DIAGRAMS OF SUMMER STOCKS AND FLOWS
FOR THE INNER DISCHARGE BAY AND
SOUTH INTAKE AREA . . .. 407
LITERATURE CITED . . . . 414
BIOGRAPHICAL SKETCH . . . ... 426
LIST OF TABLES
Table Page
1 Results of a technique test of the
Winkler method to determine the effect
of the presence or absence of acid in
fixed bottles which have been stored
for eight hours before titration. 44
2 Seasonal comparison of average wind speed
at Crystal River site. 75
3 Record of metabolism for the inner dis-
charge bay as measured by diurnal free
water oxygen changes and light and dark
bottles. 117
4 Record of metabolism for the Fort Island
and Hodges Island areas away from the
influence of the power plant discharge as
measured by diurnal free water oxygen
changes and light and dark bottles. 121
5 Diffusion rates measured in the power
plant discharge and Fort Island study
areas. 146
6 Average extinction coefficients for light
penetration of water on the inner dis-
charge bay affected by the power plant
discharge plume and unaffected areas to
the north and south. 148
7 Differential equations for diurnal model
of inner bay given in Figure 40. 157
vii
Table Page
8 Differential equations for seasonal
model of inner bay system given in
Figure 55. 202
9 Comparison of gross primary produc-
tion and total respiration measured
at Crystal River with some values from
other areas in Florida and similar
systems elsewhere. 225
C-1 Documentation of values used for forcing
functions, standing stocks, and exchange
rates in the diurnal simulation model of
the inner bay. 312
C-2 Initial and maximum values of storage
for diurnal simulation model of inner
bay. 329
C-3 Initial and maximum values of forcing
functions for simulation model of inner
bay. 330
C-4 Calculation of radioactive, evaporative,
and convective heat losses for use in
diurnal simulation model of inner bay. 331
C-5 Calculation of transfer coefficients for
diurnal simulation model of inner bay. 333
C-6 Equations of Table 7 scaled for simulation
of diurnal model of the inner bay given
in Figure 40. 339
C-7 Scaling of terms associated with photo-
synthesis in equations in Table 10 for
diurnal simulation model of the inner bay. 347
C-8 Potentiometer settings for initial run
of diurnal simulation model of inner bay. 349
C-9 Potentiometer settings for the EAI 580
variable diode function generator used to
produce the tidal volume exchange function
given in Figure 42 for the diurnal model
of the inner bay. 352
viii
Table Page
D-1 Documentation of values used for stand-
ing stocks and exchange rates in the
seasonal model of the inner bay. 356
D-2 Initial and maximum values of forcing
functions and storage for seasonal
simulation model of inner bay. 376
D-3 Calculation of transfer coefficients for
seasonal simulation model of inner bay. 377
D-4 Equations of Table 11 scaled for simula-
tion of seasonal model of the inner bay
given in Figure 54. 387
D-5 Scaling of terms associated with photo-
synthesis in equations in Table 8 for
seasonal simulation model of the inner
bay. 398
D-6 Potentiometer settings for initial run
of simulation of the seasonal model of
the inner bay. 402
D-7 Potentiometer settings for EAI 580 vari-
able diode function generator used to
produce the seasonal cycle of sunlight
given in Figure 55 for the simulation
of the seasonal model of the inner bay. 404
E-1 Documentation of numbers appearing on
Figure 38 of the inner discharge bay
ecosystem affected by the thermal dis-
charge of the power plant. 408
E-2 Documentation of numbers appearing on
Figure 38 of the south intake area eco-
system unaffected by the thermal plume of
the power plant. 411
LIST OF FIGURES
Figure Pag
1 Location of Florida Power Corporation's
power plants near Crystal River, Florida,
in relation to the major features of the
regional coastline. 3
2 Energy diagrams of producer and consumer
modules indicating the push-pull effects
of temperature on internal processes. 8
3 Aggregated energy diagram of the main
features believed important in the eco-
system of the inner discharge bay at
Crystal River. 12
4 Energy diagram of the ecosystem of the
inner discharge bay, which includes much
of the complexity omitted from Figure 3. 15
5 Bathymetry of power plant discharge area
at Crystal River. 23
6 Thermally affected area showing location
of the shallow inner bay system dominated
by the seagrass, Halodule wrightii, and
the deeper outer bay system. 27
7 Typical daily tidal cycle at Crystal River
site indicating unequal high and low tides. 29
8 Model of factors affecting oxygen dynamics
in water. 46
9 Example of graphical format for calculation
of community metabolism at Fort Island, 24-25
August, 1973, using full diurnal curve of
oxygen. 50
Figure Page
10 Graphical format for calculation of
community metabolism using dawn-dusk-
dawn data. 57
11 Comparison of community metabolism
estimates obtained from complete
diurnal measurements of oxygen versus
estimates obtained from dawn-dusk-
dawn calculations made using the same
data. 60
12 Example of two experiments to determine
oxygen diffusion coefficients by measur-
ing the rate of return of oxygen into a
nitrogen-filled dome floating on the
water's surface. 65
13 Examples of submarine photometer measure-
ments of light penetration through the
water column taken at Fort Island away
from the influence of the power plant
discharge plume and in the inner bay
influenced by the plume. 68
14 Average daily insolation by month at
Tampa, Florida. 71
15 Wind direction by season at Crystal River
site. 74
16 Monthly mean air temperature at Tampa,
Florida. 77
17 Monthly mean precipitation at Tampa,
Florida. 80
18 Weekly averages of surface water temper-
atures for the plume-affected inner dis-
charge bay and ambient water of the south
intake area. 83
19 Weekly average of electricity generated by
power units at Crystal River, and weekly
average intake and discharge water temper-
ature for unit 1. 86
Figure Page
20 Average diel water temperatures meas-
ured during community metabolism studies
of the inner discharge bay and the Fort
Island and Hodges Island control areas. 88
21 Diurnal patterns of electricity generated,
water temperatures at three locations,
and tidal stage in the discharge area of
May 24-27, 1974. 91
22 Average salinities measured on the inner
discharge bay and Fort Island and Hodges
Island study areas during the community
metabolism studies. 94
23 Seasonal patterns of benthic macrophytes
in the thermally affected inner bay and
inshore portion of the south intake area. 97
24 Map of summer standing crop of attached
macrophytic plants in the region near the
Crystal River power plants. 100
25 Seasonal diversity of benthic macrophytes
in the inner discharge bay and the south
intake area. 102
26 Seasonal record of biomass of benthic
macroinvertebrates in the inner discharge
bay and south intake areas. 105
27 Seasonal record of biomass of fish caught
with drop nets in the inner discharge bays
and south intake areas. 107
28 Carbon, nitrogen, and phosphorus measure-
ments at the mouth of the discharge canal
and a station in the south intake area. 110
29 Measurements of live chlorophyll-a and
phytoplankton biomass at a station in the
south intake area and at the mouth of the
discharge canal. 114
xii
Figure Page
30 Daytime net photosynthesis and night
respiration in the inner discharge bay
affected by the thermal plume and the
Fort Island and Hodges Island area
away from the influence of the power
plant. 125
31 Daytime net photosynthesis plus night
respiration as a measure of gross
primary production in the inner dis-
charge bay affected by the thermal plume
and the Fort Island and Hodges Island
areas away from the influence of the
thermal plume. 127
32 All daytime net photosynthesis and night
respiration values from Tables 6 and 7
and Figure 30 plotted on 12-month graph. 129
33 All daytime net photosynthesis plus night
respiration values from Tables 6 and 7
and Figure 31 plotted on 12-month graph. 131
34 Average oxygen values from all summertime
diurnal measurements taken in the inner
discharge bay and Fort Island control bay. 135
35 Seasonal averages of daytime net photo-
synthesis and night respiration in the
inner discharge bay and control areas. 138
36 Seasonal averages of daytime net photo-
synthesis plus night respiration as a
measure of gross primary production for
plume-affected inner bay discharge area
and unaffected control areas. 141
37 Seasonal trends of the ratio of daytime
net photosynthesis divided by night
respiration for plume-affected inner bay
area and unaffected Fort Island and Hodges
Island areas. 144
38 Summary energy diagram of summer stocks and
flows for the inner discharge bay. 151
xiii
Figure Page
39 Summary energy diagram of summer stocks of
biomass or material and flows of energy
and organic matter for the south intake
area away from the influence of the power
plant discharge. 153
40 Energy diagram for simulation model of
inner discharge bay emphasizing the
diurnal properties of the system. 156
41 Computer plots of forcing functions of
tidal volume exchange, depth, offshore
oxygen, and offshore water temperature
used in the diurnal simulation model. 163
42 Simulation results of diurnal model of
inner bay with coefficients set as
originally scaled. 166
43 Data gathered from the inner bay during
the community metabolism study of June
21-22, 1973, against which the simulation
of the model of Figure 20 was compared. 170
44 Solar insolation for June 21, 1973, as
recorded by a pyranometer located at the
Crystal River power plant site. Total
radiation received is indicated. 172
45 Simulation results of diurnal model of
inner bay with original scaling, but
sunlight reduced to a daily total
similar to June 21-22, 1973. 174
46 Simulation results of diurnal model of
inner bay with equal amounts of canal
and offshore water contributed to the
inner bay on a rising tide. 178
47 Simulation results of diurnal model of
the inner bay with two parts canal water
to one part offshore water contributed to
the inner bay on a rising tide.
48 Simulation results of diurnal model of
inner bay with canal water alone being
contributed to the inner bay on a rising
tide. 182
xiv
Figure Page
49 Simulation results of diurnal model
of the inner bay with a 7C dif-
ferential of discharge canal water
over ambient water and a mixing ratio
on a rising tide of one part canal
water to one part offshore water. 185
50 Simulation results of diurnal model of
the inner bay with a 7C differential
of discharge canal water over ambient
water and a mixing ratio of 2 parts
canal water to 1 part offshore water
on a rising tide. 187
51 Simulation results of diurnal model of
the inner bay with a 70C differential
of discharge canal water over ambient
water and with canal water alone flow-
ing onto the inner bay on a rising tide. 192
52 Simulation results of diurnal model of
the inner bay with no discharge of
cooling water from the power plant dis-
charge canal and original scaling of
insolation. 194
53 Simulation results of diurnal model of the
inner bay with no discharge from the
power plant discharge canal and insolation
reduced to one-half original scaling. 196
54 Simulation results of diurnal model with
timing of occurrence of high and low tide
reversed from previous runs. 198
55 Energy diagram of simulated model of
inner discharge bay emphasizing seasonal
properties of the ecosystem. 200
56 Seasonal patterns of insolation and tempera-
ture used as forcing functions in the
seasonal model of the inner bay ecosystem. 200
57 Simulation results with initial scaling
of seasonal model of the inner bay. 210
Figure Page
58 Simulation results of seasonal model
of the inner bay with seasonal pattern
of temperature increased 3C. 214
59 Response of seasonal simulation model
of the inner bay to increased tempera-
ture and turbidity. 217
60 Response of seasonal simulation model
of the inner bay to decreased turbidity
and a seasonal temperature range as
originally scaled. 220
61 Response of seasonal simulation model
of inner bay to decreased turbidity
and a seasonal temperature range of
180C 36C. 222
62 Energy diagram and analog computer patch-
ing diagram of simulation model of
producer module with temperature affec-
ting both photosynthetic and respiratory
pathways. 231
63 Simulation results of model of producer
module in Figure 62 with seasonally vary-
ing light and temperature. 233
64 Simulation response of seasonal model of
the inner discharge bay to the addition
of pathways of exchange of fish and fish
larvae with offshore waters. 237
65 Simulation results of seasonal model of
inner bay as modified in Figure 64 with
larger photosynthetic coefficient added
in winter. 241
C-1 Analog computer patching diagram of
scaled equations given in Tables C-6 and
C-7 for the diurnal simulation model of
the inner bay. 354
D-1 Analog computer diagram of scaled equa-
tions given in Tables D-4 and D-5 for the
seasonal simulation model of the inner
bay. 406
xvi
Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
PRODUCTIVITY MEASUREMENTS AND SIMULATION MODELS
OF A SHALLOW ESTUARINE ECOSYSTEM
RECEIVING A THERMAL PLUME
AT CRYSTAL RIVER, FLORIDA
By
Wade Hampton Barnes Smith
August, 1976
Chairman: Howard T. Odum
Major Department: Environmental Engineering Sciences
The effects of the heated discharge of two power
plants on the receiving estuarine ecosystem near Crystal
River on the west coast of Florida were investigated with
measurements and simulation models to help understand
relationships and predict the consequence of a third
power plant under construction. Energy circuit models
emphasizing diurnal and seasonal aspects of ecosystem
responses were used to assess the effect of power plant
operation on estuarine ecosystems. Field measurements
were taken in the discharge-affected and -unaffected
areas nearby.
Benthic populations dominated total metabolism
in both systems. Community gross primary production
varied seasonally in.the unaffected areas from a winter
xvi
low of 3.3 g 02/m2.day to a spring high of 8.8 g 02/m2-day.
In the discharge area it was relatively constant, remain-
2
ing about 4 g 02/m -day in all seasons. Phytoplankton
production normally was about 5 percent of total produc-
tion in the unaffected areas and about 23 percent in the
discharge area. In the spring its contribution increased
greatly to 25 percent in the unaffected area and 70 per-
cent in the discharge area.
Total biomass was less in the discharge than
in unaffected areas. Lower standing stock of primary
producers and benthic invertebrates in the discharge
area accounted for almost all the difference.
Diversity was lower in the discharge bay than in
the unaffected area. Mixed macroalgae and seagrasses
were the dominant benthic producers in the unaffected
areas, while the seagrass Halodule wrightii was virtually
the only species in the discharge bay. Species diversity
was lower for oyster reef organisms, and fewer species
of fish were caught in drop nets in the discharge bay
than in the unaffected bay.
A shift toward more cycling of material and energy
through the phytoplankton and filter feeders and away
from the benthic components of the system may have
occurred in the discharge area as an adaptation to the
thermal plume.
xviii
Simulation of the model of diurnal system proper-
ties with coefficients representing those for discharge
conditions gave patterns similar to those measured in the
discharge bay. The model was relatively insensitive to
adjustments in water temperature within the range expected
in the future at Crystal River. A change in the quantity
of daily insolation produced a larger change in model
response.
The simulation model of seasonal system proper-
ties was also more sensitive to light than to water
temperature. Increasing temperature alone increased
primary production and total respiration somewhat, espe-
cially in the spring. Fish and invertebrate biomass
remained the same, while detrital storage declined,
perhaps indicating their importance as an energy source
for offsetting increased respiratory drains on consumers
because of increased temperature. Increases in light
alone greatly increased system storage and flows, sug-
gesting the importance of turbidity in controlling
metabolism in the discharge area. Increasing temperature
but decreasing light lowered metabolism.
Adjustments to the seasonal model tested the
theory that systems with prominent seasonal pulses may
be exploited by populations that move in during the period
of plenty, experience rapid exponential growth, and then
xix
move away. With some migration the fish stock could main-
tain itself in a stable oscillating yearly pattern.
Results of other adjustments to the seasonal model
suggested that seasonal substitution of species of primary
producers may be the most effective way to make maximum
use of available energies at all times of the year.
INTRODUCTION
This is a study of shallow, benthic-dominated
estuarine ecosystems on the Florida west coast, one of
which received a thermal discharge from two electric
generating stations of the Florida Power Corporation.
Water was drawn from the deeper offshore Gulf of Mexico,
passed through the power plant condenser system, where
its temperature was increased about 5C, and discharged
onto the shallow inshore coastal area (Figure 1). This
study was made to increase understanding of the structure
and function of estuarine ecosystems, the relationship of
individual parts to the functioning of the total pattern
as an integrated unit, and the effects of temperature
change. Specifically investigated was the nature of an
estuarine system which had been receiving thermal effluent
for six years. How does an entire ecosystem adjust and
adapt to these new energy conditions imposed on it? How
does the new system serve as an interface between the
economy of man and that of nature?
Energy diagrams were drawn to organize, summarize
and synthesize data in models, and as a conceptual tool
Figure 1. Location of Florida Power Corporation's power plants near Crystal
River, Florida, in relation to the major features of the regional
coastline. Oyster bars are indicated by dotted outlines.
for illustrating ideas about the ecosystem at Crystal River
and the role of thermal loading in shaping its pattern.
Total metabolism, including photosynthetic production and
total respiration, was measured as the primary indicator
of the main system functions. Using simpler models, com-
puter simulations of seasonal and diurnal trends were run
and compared with measured data. This study was part of
a much larger project funded by the Florida Power Corpora-
tion evaluating questions related to the impact of its
power plants at Crystal River on the adjacent estuarine
ecosystems. As required by our contract, efforts were made
to summarize data from other studies in developing an over-
view of the ecosystem.
As fossil fuels for powering man's economy become
scarcer and more expensive, the need increases for recognizing,
utilizing, and protecting the important work contributions
of nature in support of man's economy, and establishing
effective feedback pathways from man to protect his life
support system. A regional system of man and nature which
allows its natural components to contribute work services
in support of the overall pattern may avoid unnecessary
technological constructions and be most successful in
utilizing all available energies when use of the environ-
ment constitutes more useful work than is lost by environ-
mental impact. For example, should coastal and estuarine
waters be used for cooling of the thermal effluent from
electric generating stations, or is it necessary to build
technological alternatives such as cooling towers for this
purpose?
System Adaptation, Environmental Impact,
and Thermal Loading of the Estuary at
Crystal River, Florida
At Crystal River, and wherever thermal effluents
flow into an ecosystem, potential energy is carried with
it. This energy, like all other energy sources impinging
there, is available for doing work in the environment,
although its exact way of doing work may not be known
(Odum, 1974b).
Other changes besides thermal loading are caused
by power plant installations. At Crystal River the con-
struction of dikes and the pumping of cooling water through
the canal system (see Figure 1) may modify the current
and flushing characteristics of the surrounding waters.
Turbidity in the discharge area may have increased because
of scouring of the discharge canal and from sediments
carried through the power plant from the intake canal
where they had been stirred up by barge traffic. Canal
construction caused changes in drainage patterns and mor-
phology of adjacent salt marshes.
The most important impact, however, may be the
effect of higher temperature on biological processes.
Adaptation and acclimation of metabolism of individual
organisms to offset temperature changes is well estab-
lished (Bullock, 1955). Much less is known about the re-
sponse of whole ecosystems to changed thermal regimes.
How does system structure adapt so that the new pattern
that emerges is best coupled to the changed thermal
regime? What is the nature of this new system linked to
man's technology? These questions may begin to be answered
by observing such system properties as total community
metabolism, species diversity, and seasonal patterns.
4The effect of thermal loading on biological pro-
cesses may be important at both the level of individual
organisms and the ecosystem (Kelley, 1971; Odum, 1974b;
McKellar, 1975). Since all processes are stimulated equally
below the threshold of rapid thermal enzyme destruction,
temperature acts to increase processes building structure
as well as those degrading it (Figure 2). "For a plant
(Figure 2a), the dark reactions of photosynthesis may be
stimulated as much as respiratory pathways, so that the
overall effect on biomass may be neutral. However, if
photosynthesis is limited by energy and material shortages
so that respiratory losses are not offset, biomass may
decline. The same holds true for a consumer (Figure 2b).
If the metabolic pathways of digestion and rebuilding of
structural animal biomass are affected at the same rate
as those degrading this biomass, metabolism increases but
Figure 2. Energy diagrams of producer and consumer
modules indicating the push-pull effects of
temperature on internal processes. See
Appendix A for meaning of symbols.
(a) Producer module with temperature acting
on both photosynthetic primary produc-
tion and respiration processes.
(b) Consumer module with temperature acting
on processes of biomass formation through
food gathering, digestion, and assimila-
tion as well as the respiratory degrada-
tion of biomass.
(b)
FOOD
SOURCE
I
~__ _I_
_-
the amount of biomass is unaffected. However, if food is
limited the population loses mass because it cannot compen-
sate for respiratory losses.
For the ecosystem, if it is to compete at the new
steady state, respiration degrading structure induced by
higher temperatures must be compensated for by the larger
push of increased rates of production of system structure.
-Can accelerated cycling of nutrients from increased res-
piration offset nutrient limitations to primary production?
Is this increased production enough to supply energy demands
of larger consumers?
Models for Gaining an Overview of the
Estuary and Power Plant at Crystal River
Proposed in Figures 3 and 4 are the energy circuit
models at different levels of complexity of the system of
estuary and power plants at Crystal River. Their purpose
is to organize in overview concepts of system structure,
processes, pathways, interactions, and relationships.
Inherent in the diagrams are patterns important on both
daily and seasonal time scales. An explanation of symbols
used in this dissertation is given in Appendix A. More
complete discussions and additional symbols are given by
Odum (1971, 1972, 1974a, 1975).
Simplified Model of the Inner Bay
Given in Figure 3 is a model diagram of the inner
bay ecosystem from which details have been eliminated
leaving only the basic system structure of water storage,
benthic macrophytes, consumer populations and tidal ex-
changes with the saltmarsh, offshore, and canal ecosystems.
On a rising tide surface water from the power plant dis-
charge canal is forced onto the inner bay by the damming
effect of water flowing on from offshore and the increasing
height of head of the approaching wave of the tide. On
a falling tide canal water flows directly down the channel
beside the inner bay, where it receives water flowing off
the bay. Diffusive oxygen exchange with the atmosphere
occurs driven largely by turbulence induced by tidal
exchange. Gains of heat result from solar insolation
and atmospheric longwave radiation. Heat losses occur
from conduction, back radiation, and evaporation.
Primary production occurs in the phytoplankton
and the benthic macrophytes, which take up nutrients and
oxygen from the water column while returning oxygen and
organic matter to it. Storages of organic matter are in
the water column and sediments, which are consumed by
populations of microbes. Two classes of consumers are
shown. In the water column are free living animals feed-
ing on detritus, phytoplankton, benthic invertebrates, and
Figure 3. Aggregated energy diagram of the main fea-
tures believed important in the ecosystem
of the inner discharge bay at Crystal
River. Details within the compartments
have been omitted to emphasize basic sys-
tem structure and function in overview.
Symbol shown as T indicates a connection
from heat sinks (- ). See Appendix A for
definition of symbols.
S SALT MARSH
S SUBSYSTEM I
EVAPORATION
EXCHANGE WITH
CANAL ONLY ON
RISING TIDE
INNER BAY I
WATER COLUMN
CONTAINING HEAT,
NUTRIENTS, TURBIDITY,
ORGANIC MATTER,
OXYGEN, AND
PHYTOPLANKTON
BENTHIC
MACROPHYTES
AND EPIPHYTES
each other. Many of the larger members of this compart-
ment migrate seasonally to and from the offshore regions.
Benthic invertebrates and oysters feed largely on detritus
and phytoplankton. Nutrients are regenerated into the
water column from all respiratory pathways.
Detailed Model of the Inner Bay
In Figure 4, more of the complexity of detail
within the compartments has been added to the model of
Figure 3. Sunlight penetrating the water column of the
inner bay is attenuated by turbidity, shading of phyto-
plankton biomass, and the natural extinction properties
of water. Primary production utilizes the remaining
light, and is concentrated in the benthic macrophytes
and their associated epiphytes with a smaller contribution
from the phytoplankton. This production moves to the higher
trophic levels primarily through a storage of detritus and
its associated microbes in the sediments. A much smaller
amount is stored in the water column. A small amount is
transferred by direct grazing of epiphytes. Larger con-
sumer populations are represented by benthic inverte-
brates and oysters, zooplankton and larval forms, shrimps
and crabs, and resident and migratory fish. Seasonal
migratory movements of shrimps, crabs, and migratory fish
stocks are indicated. All respiratory pathways are shown
returning nutrients into the water column storage. Various
Figure 4. Energy diagram of the ecosystem of the inner
discharge bay, which includes much of the
complexity omitted from Figure 3. Pathways
of oxygen uptake and temperature effects
have been abbreviated for clarity. Pathways
from storage of heat and oxygen labeled T
and 0, respectively, are assumed to be con-
nected with similarly labeled pathways on
work gates and consumer modules. Pathway
marked as T indicates a connection from
heat sink symbols ( ).
exchanges with the adjacent saltmarsh, power plant canal,
and immediate offshore ecosystems occur with the rise and
fall of the tide.
Heat in the water is lost and gained through physi-
cal processes as well as advective exchanges. Gains occur
from solar insolation, atmospheric longwave radiation, and
heat generated from all biological processes. Losses re-
sult from back radiation and evaporation. Conduction is
a gain or loss depending on the direction of the gradient
between air and water. Oxygen has a diffusive exchange
with the air driven largely by water turbulence.
In summary, this detailed model serves to emphasize
issues related to the interactions of power plants with
estuaries, helps the reader visualize the system studied
at Crystal River, summarizes initial understanding of its
characteristics, helps to plan the research program, and
provides a basis for simpler models for simulation.
Previous Studies of Thermally Affected
Aquatic Ecosystems
Most work on the effect of temperature on life
processes has been at the level of the whole organism
or at smaller (e. g. subcellular) levels; less work has
been concerned with its effect on whole ecosystems.
Perhaps because of their simplicity the most thoroughly
studied ecosystems to date have been thermal spring
ecosystems. System structure has been discussed by most
authors (Brock, 1967a, 1967b, 1969; Brock and Brock, 1969;
Kullberg, 1966; Stockner, 1967, 1968; Wiegert and Fraleigh,
1972). Zonation of algal or bacterial mat communities
associated with temperature gradients, both down and across
spring runs, with a vertical zonation of structure at any
given point were the main characteristics of these eco-
systems. Filamentous bacteria were dominant in the hotest
portion of the stream, being replaced by blue-green algae
as the water cooled. Green algae, in turn, replaced the
blue-green at still lower temperature cooling. Species
diversity was very low overall, tending to increase down
the temperature gradient.
Community metabolism in thermal springs has also
been measured (Brock, 1967b; Duke, 1967; Phinney and
Mclntire, 1965; Stockner, 1968; Wiegert and Fraleigh, 1972).
Values measured generally fell within ranges reported
for many other types of aquatic ecosystems. Brock (1970)
reviewed work on high temperature systems.
The work reported above is mostly on springs with
temperatures in excess of 45C, which is generally above.
the thermal limits for enzyme destruction of most organisms.
Available work on thermally affected systems within temper-
ature ranges more normally encountered in nature has mostly
involved microcosm studies. Allen and Brock (1968) re-
ported that microcosms held at a range of temperatures
from 20C to 75C and all seeded alike from a wide variety
of sources; each developed its own characteristic combina-
tion of species. Beyers (1962) found only small responses
in community metabolism to 36-hour increases in temperature.
Davis (1971), studying experimental estuarine ecosystems
contained in large plastic swimming pools, found increased
gross community primary production and respiration during
spring, summer, and fall in those heated 4-60C above con-
trols. Kelley (1971) studied high-nutrient freshwater micro-
cosms subject to constant low, constant high, and fluctuat-
ing temperature regimes. Mean values of net production
and night respiration over the study period were higher in
those microcosms which had higher mean temperatures. Vari-
ous aspects of the biology of Par Pond, a freshwater reactor-
cooling pond at the Savannah River Plant, South Carolina,
have been studied for a number of years by investigators
at the Savannah River Ecology Laboratory of the University
of Georgia (Gibbons and Sharitz, 1974b).
A general assessment of research related to the
environmental effects of the operation of power plants
is difficult because much of it is contained in reports
to Federal agencies concerned with licensing, and is gen-
erally unavailable for review. Zieman (1970) has reported
on the early effects of the operation of power plants at
Turkey Point on Biscayne Bay, near Miami, Florida. Condi-
tions of flow rate and temperature rise of the cooling water
were very similar to those at Crystal River. The receiving
ecosystem was dominated by a mixture of macroalgae and sea-
grasses (mostly Thallassia testudinum). By the end of the
second summer of operation 50-60 acres of bay bottom adja-
cent to the mouth of the discharge canal had been denuded
of this community and replaced by a blue-green algal mat
community. An additional 70-75 acres had some Thallassia,
but were still devoid of macrophytic algae, while 160-170
more acres exhibited some stress to the existing macro-
algae populations.
Other available power plant data have dealt with
more northern situations involving phytoplankton.-dominated
ecosystems. The effects of increased temperature on pri-
mary production were usually measured by the uptake of
carbon-14, often in bottles held in illuminated light
boxes. Results have been mixed. Several studies in-
volving both estuarine and freshwater cooling systems
have found stimulation of photosynthesis in the cooler
months and a depressing effect in the warmest months
(Morgan and Stross, 1969; Smith et al., 1974). Tilly
(1974), using carbon-14 measurements incubated in situ
in Par Pond, South Carolina, found primary production to be
somewhat greater in the surface water at the warmer station.
This tendency was more pronounced during the warm months of
the year. Gurtz and Weiss (1974), also using carbon-14
methods, found inhibition of photosynthesis at all times
of the year. A trend toward greater inhibition at higher
ambient water temperatures was suggested by the data.
Only several reports appear to be available on
aspects of ecosystems which have been adapting to power
plant discharges for a number of years. North (1968)
studied the discharge area affected since 1957 at Morro
Bay, California. He found abundance and diversity of
plants and animals to be reduced in a transitional
region over a distance of approximately 200 m from the
end of the discharge canal. Recovery to conditions
typical of the area occurred in a relatively short hori-
zontal distance of 10 m at the end of the transitional
region. J. R. Adams et al. (1974) could find no differ-
ence in intertidal sandy beach populations located near
the discharge versus ones further away.
Few power plant studies appear to have synthesized
the diverse data into an overview of the ecosystem respond-
ing as an integrated unit to the new set of environmental
forcing functions. Emphasis has generally been placed
on individual aspects of power plant operation, such as
entrainment through the condenser cooling system and
entrapment on the screens protecting the cooling water
intake pumps (Jensen, 1974c), or on individual species
or components of the ecosystem. Typical studies might
examine mortality of phytoplankton from passage through
the condenser system, diversity and biomass of benthic
organisms and fish in the discharge area, or primary
production of the phytoplankton component of the ecosystem.
Often these studies have been done in the laboratory.
Chesapeake Science, volume 10 (1969), and proceedings of
symposia edited by Gibbons and Sharitz (1974a) and Esch
(in press) contain many papers of this type.
Several studies have been published which contain
most research results for a particular power plant in one
volume (Jensen, 1974a, 1974b; Central Electricity Generat-
ing Board). Discussion of results, however, is by sub-
systems with little attempt to synthesize the findings
with text, diagrams, or simulation models into a picture
of the functioning of the whole ecosystem.
Description of Study Area at Crystal River
The power plant site in Citrus County (Figure 1)
is on the low wave energy portion of the Florida west
coast as defined by Tanner (1960). The shallow sloping
bottom (46 km to the 5 fathom contour) is part of the
drowned karst topography of this portion of west central
Florida. The topography of the immediate offshore region
is a series of shallow basins separated by oyster reefs
(Figure 5). Freshwater sources influencing the area are
the Crystal River 4.8 km to the south (mean flow 1500
m /min; 400,000 gpm), and the Withlacoochee River and
Cross Florida Barge Canal 6.4 and 5.8 km to the north,
respectively, with a combined flow of 2150 m /min
(570,000 gpm).
Figure 5. Bathymetry of power plant discharge area at Crystal River. Location
of inner bay has been circled. Contour interval is 1.0 feet. Datum
based on mean sea level. (Adapted from Rodgers et al., 1974)
5 4-.45 INNER
-' ,_ H
~~ DISCHARGEE
5 5
49
N I4 M7 A
NAUTICAL MILE RODGERS, eF C/,i 1974)
The power plants are on the landward edge of a
tidal saltmarsh dominated by Juncus roemerianas bordered
on the seaward edge by a narrow fringe of Spartina
alterniflora. Two units were in operation during this
study--unit 1 since July, 1966, and unit 2 since November,
1969--giving a combined total output of 897 megawatts
electrical (MWe). A nuclear powered unit of 885 MWe
output was under construction. The two operating units
cycle water for once-through cooling at a combined flow
of 2430 m3/min (640,000 gpm) through canals dredged
across the shallow offshore region and saltmarsh.
Maximum condenser temperature rise is 6.10C.
The power plant intake canal extends approxi-
mately 4.8 km into the Gulf with an average depth of
6-7 m and a width of about 75 m, serving also as the
passageway for delivery of fuel oil in barges by large
ocean-going tow boats. Cooling water passes down the
canal at about 8 cm/sec before being pumped through
the power plant condensers, where its temperature rises
5-60C. The discharge canal is about 1.6 km long with
an average depth and width of about 4.5 m and 50 m,
respectively. The smaller cross-sectional area causes
the stream velocity to be about twice that in the intake
canal. The residence time of water masses in the canal
system is about 20 hours for the intake canal and about
3.5 hours for the discharge canal.
Two types of bay systems are affected by the
thermal plume (Figure 6). Immediately adjacent to the
saltmarsh is the shallow bay of this study averaging
about 1 m in depth, composed of a mixture of bottom
covered with seagrass, some oyster reef associations,
and areas of sand and mud. Seaward of a row of oyster
bars is a deeper outer basin of about 2 m average depth
in which the plankton and reef ecosystems become impor-
tant. The "bays" referred to here are actually the
immediate landward edge of the Gulf of Mexico.
The plume-affected inner bay of this study is
a shallow benthic seagrass-dominated system composed
almost exclusively of Halodule (Diplanthera) wrightii
during the warm months, while in the winter of 1972-73
mixed Ectocarpaceae proliferated and covered much of the
bottom area. It did not return during the milder winter
of 1973-74. The unequal semi-diurnal tide (see Figure
7) has an average tidal amplitude of about one meter
exposing much of the bay bottom on the lowest of the
two daily low tides, and draining the entire bay on
the lowest spring tides. In addition, strong northerly
winds associated with passages of cold fronts in winter
occasionally push the regional water mass offshore and
drain the bay and the nearby coastal area for several
days. With normal weather and tides, the heated plume
moves back and forth across portions of the bay in
Figure 6.
Thermally affected area showing location of
the shallow inner bay system dominated by the
seagrass, Halodule wrightii, and the deeper
outer bay system. Lettered dots indicate
inner bay locations of remote telemetry
buoys maintained by Florida Power Corporation
for recording water temperatures. Location
lettering is as designated by the Florida
Power Corporation.
0 0.5
KILOMETERS
a
o
o ra,,' '
OUTER
B BAY
\Le :/
O
O
Figure 7. Typical daily tidal cycle at Crystal River site
indicating unequal high and low tides. Ampli-
tude changes were taken from tide tables (U. S.
Department of Commerce, 1972) for June 12,
1973.
2.0
E
F1.0
a
O
0
TIME OF DAY
0600 1200 1800
response to the tidal cycle. The shallow areas near
the power plant which were unaffected by the thermal
plume were dominated by a diverse mixture of benthic
macroalgae and seagrasses.
Areas away from the influence of the power plant
discharge at Fort Island, Hodges Island, and in the south
intake area (Figure 1) were used as comparison areas.
The south intake area was located immediately south
of the southern intake canal dike. Measurements taken
there by others included stocks of fish, benthic inverte-
brates, benthic macrophytes, zooplankton, sediment organic
content, and nutrient concentrations. The benthic macro-
phytic producers were a diverse mixture of macroalgae
and seagrasses.
Total metabolism measurements were made at Fort
Island and Hodges Island. Most measurements were made
in a funnel-shaped bay south of Fort Island. This area,
which was somewhat deeper than the inner discharge bay
area, was characterized by a benthic flora similar to
the south intake area. The extreme clarity of the water
influenced by the nearby Crystal River allowed much
greater light penetration to the bottom as measured with
a submarine photometer than in the power plant discharge
area. Hodges Island to the north of the Withlacoochee
River (Figure 1) was away from freshwater influences.
This bay had more turbid water with little growth of
benthic macrophytic plants.
Other Studies of the Crystal River Region
Little work is available on the Crystal River
region prior to power plant construction. Dawson (1955)
provided data on oyster populations and hydrography,
including measurements at stations now well within the
influence of the power plant.
After construction and operation of the plants
were initiated, many studies were made as part of the
larger research program undertaken by Florida Power
Corporation. Benthic seagrasses and algae were inven-
toried by Steidinger and Van Breedveld (1971), while
quantitative measurements of biomass were made by Van
Tyne (1974). Benthic invertebrates were inventoried
by Lyons et al. (1971) and measured quantitatively by
Evink and Green (1974).
Trawl samples of fish were reported by Grimes
(1971), Grimes and Mountain (1971), and Mountain (1972).
C. A. Adams (1974) analyzed data on fish caught in
dropnets from the shallow inshore areas, while Carr and
Adams (1973) discussed the food habits of juvenile fish
in the beds of benthic seagrasses and macrophytes.
Homer (1975) studied seasonal patterns of tidal creek
fishes.
Trace metal content of oysters from the intake
and discharge canals was reported by Grimes (1971) and
Mountain (1972). Biomass, diversity, and metabolism of
oyster bars were measured by Lehman (1974a, 1974b).
Blue crab movements in the intake canal area
were monitored by Adams, Oesterling, and Snedaker (1974).
Nutrients, chlorophyll, and phytoplankton numbers and
diversity were measured by Gibson (1975). Zooplankton
biomass and diversity were studied by Maturo (1974).
Fish and other organisms trapped on the screens
protecting the condenser water intake pumps were monitored
by Adams, Bilgere, and Snedaker (1974). Entrainment of
larval fish and zooplankton through the condenser system
was measured by Maturo (1974) and Snedaker and Johnson
(1975).
Total community metabolism was measured and
studied with simulation models for the saltmarsh by
Young (1974), for the oyster reefs by Lehman (1974a,
1974b), for the deeper outer bays by McKellar (1974, 1975),
and for the power plant canals by Kemp (1974). A larger
scale analysis of the energetic costs associated with
estuarine cooling compared to technological alternatives
was done by Odum (1974b), Odum et al. (1974), and Kemp
et al. (1975).
Physical measurements of the hydrography of the
area were reviewed by Carder (1975). These were used
by Klausewitz (1973) for verification of a computer
simulation model of the behavior of the thermal discharge
plume. Bedient (1972) simulated the flushing of water
from the discharge canal as it related to dispersion of
radioactive wastes in the discharge water. Swindler
(1973) examined the sedimentology of the region between
the Crystal River and Withlacoochee River. Cottrell
(1974) studied sediment composition and sedimentation
rates in the more immediate plant area.
Previous Simulation Models of Marine Ecosystems,
Diurnal Oxygen Dynamics, Temperature, and the
Effects of Power Plants on Ecosystems
Several previous attempts at modeling marine
ecosystems have appeared in the literature. Chen and
Orlob (1972) developed an extensive simulation of the
San Francisco Bay and Delta region incorporating spacial
as well as temporal elements. The geographical region
was divided into a network of nodes and connecting path-
ways. Mass balance equations were used to transfer
materials between nodes with tidal dynamics as the
forcing function. Up to 22 parameters could be con-
sidered: dissolved oxygen, biochemical oxygen demand,
alkalinity, pH, temperature, nitrogen (three forms),
phosphorus, suspended sediment, three types of algae,
zooplankton, three types of fish, and benthic animals.
For conservative elements, only terms for diffusion,
advection, input, and output were included in the mass
balance. For biological elements, appropriate terms
for rates of growth, respiration, mortality, and chemical
transformations were added. Temperature linearly affected
respiratory pathways of fish and zooplankton, and affected
both photosynthesis and respiration of algae. Growth rate
coefficients were based on Michaelis-Mention kinetics.
Model calibration to real data was presented for only
several parameters with fit being quite good. Subsequent
runs evaluated the effect on the bay of proposed regional
sewage treatment and water diversion alternatives.
Steele (1974) simulated a simple model of the
North Sea using storage of nutrients, phytoplankton,
zooplankton weight, and zooplankton numbers. Sunlight
was considered nonlimiting and was omitted as a forcing
function, so that changes in phytoplankton biomass were
a function only of nutrients, mixing below the thermo-
cline, and zooplankton grazing. Nutrient cycling was
included as excretion by zooplankton respiration.
Equation terms for nutrient uptake and zooplankton
grazing were derived from observed experimental data
and were given the form of Michaelis-Menton kinetics.
Brylinsky (1972) performed a sensitivity analysis
on a model of the English channel, which included storage
of phytoplankton, zooplankton, benthic fauna, pelagic
fish, demersal fish, and bacteria. Photosynthesis was
considered a constant external input. Pathways of
I
exchange between compartments were linear and controlled
solely by the donor compartment. Since nutrients were
not included as a variable, cycling was not a model fea-
ture. It was stated that the model was not intended to
be realistic, but, instead, to illustrate the applica-
tion and usefulness of the tool of sensitivity analysis.
An early attempt to simulate diurnal oxygen
dynamics of an ecosystem was made by Odum, Beyers, and
Armstrong (1963) using a passive analog circuit. Results
supported the theoretical discussion of the effect of
a small organic storage capacity in the nannoplankton
on the measurement of primary production in tropical
seas.
Several authors have obtained very good fit
for data from microcosms to relatively simple models
of their diurnal properties. Sollins (1970), studying
a blue-green algal mat, followed oxygen through compart-
ments of producers, consumers, detritus, dissolved oxy-
gen, CO2 (total in solution), atmospheric oxygen, and
water. All flows between compartments were controlled
by the upstream compartment only (donor control). Using
a square-wave regime of light input, the model produced
simulated curves of oxygen very similar to measured
curves and their rates of change.
Kelley (1971) included only storage of carbon
dioxide and labile and structural organic matter in his
simulation of a nutrient-rich freshwater microcosm of
mixed plankton. Since his study was partially concerned
with the effects of temperature, it was included in a
push-pull fashion as an action on every pathway. Rates
of flow between compartments were otherwise controlled
only by the donor compartments, as in the model by
Sollins. Excellent fit was obtained to the measured
oxygen data.
Nixon and Odum (1970) considered only storage
of organic material and nutrients in a model of hyper-
saline algal mat community. Transient responses of this
very simple model were compatible with those observed
in the microcosm.
Simulations based on the more variable data
gathered from open ecosystems in nature have been
carried out. A model of Bissel Cove, Rhode Island
(Nixon and Oviatt, 1973) was basically an oxygen balance
consisting of a single storage of oxygen with inputs
from primary production of plankton, macroalgae, and
benthic microflora. Respiratory oxygen losses occurred
to producers, sediments, detritus, shrimp, and fish.
Diffusion exchange with the atmosphere and tidal exchange
with a constant oxygen source were losses or gains
depending on the saturation level of the water and the
stage of the tide. Rates of oxygen losses or gains for
each pathway were empirically derived from regression
equations calculated from observed data. No feedback
or cycling pathways were included. Model response fit
reasonably well to observed diurnal curves of oxygen.
Boynton (1975) simulated a river-dominated
estuary to examine issues of river discharge schedules
and potential effects of human development on nearby
lands and its relation to an oyster fishery existing
in the bay. Using a simplified energy symbol model,
diurnal curves of oxygen very similar to data measured
in the area were obtained.
Several simulations have included temperature
actions. An early one emphasizing the effect of temper-
ature as an exponential function on zooplankton popula-
tions of the North Sea was done by Riley (1946, 1947).
Odum (1975) translated these equations into models
using the energy circuit language.
Hall (1974) briefly reported on a simulation
model of the effect of power plants on the striped
bass fishery of the Hudson River. Details of the model
were not given.
Odum (1974b) discussed some general principles
regarding temperature and system responses, including
the push-pull effect on both ordering and disordering
processes. Examples were given of simulations of equa-
tions proposed by Eyring and Eyring (1963) and Morowitz
(1968) which incorporated the push-pull feature of
temperature action.
38
Nixon and Oviatt (1973) included temperature
actions only on respiratory pathways in their simula-
tion model of Bissel Cove. As a result, a decline in
oxygen was predicted as the effect of the action of a
hypothetical power plant on the cove.
Miller (1974) simulated the effect of maintain-
ing mangrove vegetation in power plant canals to aid in
cooling the water before recirculation through the power
plant. Increased, but not severely detrimental, water
stress was predicted for the trees.
Several simulation models of other ecosystems
at Crystal River have been run. Young (1974) observed
increased photosynthesis, respiration, and live and dead
standing crop in simulations of the effect of elevated
water temperatures on the fringing Spartina saltmarshes.
Lehman (1974b) simulated the intertidal oyster reefs.
i Model responses included faster turnover rates for
plume-affected conditions. Simulations of effects of
adding thermal waters of another power plant suggested
reduced seasonal variation of reef standing stocks.
Kemp (1974), in a preliminary simulation of the com-
munity of fish, plankton, and benthos of the power plant
intake canal, found fish stocks to be most sensitive to
water flow rates and immigration. Plankton was rela-
tively insensitive to most parameters, being controlled
principally by concentrations carried in from outside
the canal. McKellar (1975) simulated the outer bay of
the discharge area (see Figure 6). Raising the water
temperature to that measured in the discharge area
produced only small increases in total metabolism and
some component storage. Water exchanges were shown to
be a stabilizing influence by dampening large fluctua-
tions in zooplankton, phosphorus, and detritus. Simula-
tion of the conditions expected with future power plants
produced no large changes in total community metabolism.
Plan of Study
The structure and function of the thermally
affected inner bay ecosystem at Crystal River and
unaffected areas to the north and south were determined
from field measurements of biomass of organisms and
system metabolism, and from the behavior of ecosystem
simulation models evaluated with these and other data.
The conceptual model shown in Figure 4 was developed as
an overview to show the relation of the main energy
exchanges with the outside, and of the main storage
of the inner bay ecosystem among themselves. Simpler
models which aggregated the main stocks and processes
were simulated on an analog computer.
Total community metabolism was determined from
diurnal changes in free-water oxygen concentrations
and was used as an indication of the ability of the
ecosystem to process the energies available to it.
Comparison of metabolism of the thermally affected
area with areas away from the influence of the power
plants indicated the degree to which these processing
abilities had been altered. Measurements were taken
from June, 1972 through May, 1974 representing all
seasons and establishing general seasonal trends of
metabolism.
Efforts using bottle experiments were made to
partition total metabolism between its planktonic and
benthic components. Measurements were made of penetra-
tion of light through the water column.
Models were evaluated with data obtained in this
study and gathered concurrently by other researchers,
with other supporting measurements, information from
the literature, and some necessary calculations and
assumptions. These models were translated directly into
a set of differential equations, which were programmed
for analog computer simulation. Simulation runs were
made with coefficients set for conditions with and without
the influence of the power plant. Results were compared
to the observed data. Sensitivity of the models was
examined with respect to changes in water temperature
and ratios of discharge canal water to offshore water
mixing on the inner bay. Finally, simulations were run
with conditions expected when the new power plant begins
operation.
METHODS
Metabolic Measurements
Community metabolism was measured with diurnal
sampling of oxygen following Odum and Hoskins (1958),
Odum and Wilson (1962), and Odum (1967), and an abbrevi-
ated method using dawn-dusk-dawn oxygen samples (McConnell,
1962). Oxygen was measured by the azide modification of
the Winkler technique (Amer. Publ. Health Assoc., 1971),
but adapted for use with smaller sample collection
bottles.
Mini-Winkler Field Kit and Winkler Method
Modification
Because of the large number of samples to be
processed and the need for compactness, a mini-Winkler
field kit developed at the University of Texas Institute
of Marine Sciences was used in this study. Standard
flat-topped 125-ml reagent bottles were used for sample
collection in place of 300-ml BOD bottles. Samples were
fixed with 0.5 ml of manganous sulfate and azide reagent
carried in dropping bottles in the field kit. After
acidification with 0.5 ml concentrated sulfuric acid,
100-ml subsamples were titrated with 0.012 N sodium
thiosulfate. This normality allowed direct reading of
milliliters of titrant as mg/l of oxygen.
Variability between replicate pairs of oxygen
samples could have arisen from many sources. Since the
small reagent bottles used were inexpensive, variation
in their individual volumes was expected. A test of a
54-bottle subsample of those in use gave an average
volume of 122.8 ml with a standard error of 0.22.
Because each bottle was filled from a separate sample
of bay water taken 30 seconds to one minute apart, vari-
ations due to water mass differences could also have
occurred. Other sources of variation could have included
differences in reagent volumes added and differences in
sample volumes titrated.
Actual differences in titrant volume encountered
between replicate pairs of samples were small, however.
Based on a subsample of 486 replicate pairs, 72.6 percent
differed by 2 drops (0.1 ml) or less. Since titrant
volume was generally in the range of 4-8 ml, this gave
an average error of 1.3-2.5 percent. Loss of accuracy
due to increased sources of variability was, therefore,
considered minimal, and was far outweighed by convenience
in handling in the field. More samples could be processed,
permitting better estimates for the whole bay.
Significance of Delay in Fixing Winkler Bottles
with Acid
A test was made of the effect of an eight-hour
delay in adding acid to the sample bottles in the Winkler
analysis of oxygen. Thirty bottles were filled with
thoroughly mixed salt water from a bucket, and immediately
fixed with the manganous sulfate and azide reagents. Ten
bottles were picked at random, acidified, and titrated
within 30 minutes. The remaining bottles were split
into two groups, one group of 10 bottles receiving acid,
while the other did not. Both groups were stored in the
dark for eight hours. At the end of that time, acid was
added to the bottles which had not received it earlier,
and both groups were titrated. Table 1 gives the results
of the three treatments. Differences between treatments
were significant (95% level), but were considered too
small to have any significant effect on the measurements.
Complete Diurnal Sampling of Oxygen
The calculation of total community primary pro-
duction and respiration from free-water measurements
of oxygen is based on the model given in Figure 8.
As indicated, the oxygen concentration in the water
column at any moment and changes in concentration with
time are a function primarily of the production of oxygen
during photosynthesis, its consumption in respiratory
Table 1.
Results of a technique test of the Winkler
method to determine the effect of the
presence or absence of acid in fixed bottles
which have been stored for eight hours before
titration. Each treatment contained 10
bottles. Results are given in milliliters
of titrant.
Bottles fixed, Bottles fixed and Bottles fixed
acidified, and acidified immedi- immediately;
titrated ately; titrated acidified and
immediately 8 hours later titrated 8 hours
later
Average 5.45 5.43 5.48
Std. Error 0.02 0.01 0.01
Figure 8. Model of factors affecting oxygen dynamics
in water.
SATURATION
--- ADVECTIVE
---- EXCHANGE
PHOTOSYNTHESIS
RESPIRATION
I ________
processes, gains or losses because of advective exchange
with adjacent water masses, and diffusive exchanges with
the atmosphere. The contribution to oxygen dynamics of
the nonbiological processes of advection and diffusion
may be corrected for if their magnitudes are known or can
be estimated. Subtracting their effect allows a calcu-
lation of changes resulting only from the biological
processes of photosynthesis and respiration, and, thus,
a calculation of production and respiration.
After correcting for diffusion and advection,
any gain in oxygen concentration during daylight hours
would be a consequence of the greater production of oxygen
in photosynthesis than its concurrent use in respiration,
thereby providing a measure of net primary production.
At night, when there would be no production of oxygen
by photosynthesis, the rate of oxygen decline would be
an estimate of community respiration. By assuming a
similar respiration rate for daylight hours (which would
be a conservative assumption), an estimate of the rate
of gross primary production may be obtained by adding
daytime photosynthesis and night respiration.
Stations were sampled approximately every three
hours over a 24-hour period. Two buckets of surface
water were collected 30 seconds to one minute apart at
each station, and sample bottles were filled from the
bottom by siphoning through rubber tubing. Late night
samples were sometimes stored without acidification for
titration the following morning (see above for effect
on Winkler analysis). Time, temperature, salinity, and
depth were noted at each station.
Because of the large tidal flushing, advection
of water masses from outside areas was at first thought
to be important. In order to assess this effect on the
diurnal oxygen curve in the study areas, four or five
stations were sampled in the early part of the project.
Analysis indicated a general similarity in the daily
increase and decrease of oxygen at all stations, sug-
gesting that advection was from areas of similar metabo-
lism. Thus, errors introduced by advection were thought
to be small, and the number of stations was usually re-
duced to two or three to meet field schedules.
Diurnal metabolism graphs were constructed using
a standard format (Figure 9) to allow easy visual com-
parison among all diurnal samples taken at Crystal River
as well as with others in the literature (Odum and
Hoskins, 1958). The data were analyzed several different
ways as the study progressed. At first, a graph for each
station was plotted and analyzed separately. Later, all
points from separate stations were plotted on one graph,
but only the mean curve was analyzed (Figure 9). Each
oxygen point was the average of duplicate Winkler
analyses. Oxygen per square meter (Figure 9c) was
Figure 9. Example of graphical format for calculation of
community metabolism at Fort Island, 24-25
August, 1973, using full diurnal curve of oxy-
gen. Open circles represent average of measure-
ments at four stations, each of which are shown
as solid points. (See text for detailed dis-
cussion of [g] and [h].)
(a) Oxygen concentration.
(b) Depth.
(c) Areal oxygen obtained by multiplying (a)
and (b).
(d) Temperature.
(e) Salinity.
(f) Percent saturation of oxygen calculated
using oxygen values in (a).
(g) Rate-of-change of oxygen. Dotted line is
rate-of-change of (c). Solid line with
solid dots ( -- ) is rate-of-change cor-
rected for depth changes. Solid line with
open circles (o--o) is rate-of-change
curve corrected for diffusion using coef-
ficient values given across the top of the
diagram. Units of diffusion coefficients
are g02/m2/hr./l00% deficit.
(h) Rate-of-change of oxygen. Solid line with
solid dots ( -- ) is rate-of-change of
(a) multiplied by average depth at each
hour. Solid line with open circles
(o---o) is curve corrected for diffu-
sion using same coefficients as in (g).
z
I-
0
o
u
Go
0 o
000 1200 1800
TIME OF DAY
-( C)
(d)
Il "- L % I I IIVIL
o (g) HOTOSYT
o I
+0.5
n- o
0 E 77-
z -0.5 -_
wI
x I
o -1.0 V
I NIGHT RESPIRATIC
1.5
+1.0 I I
(h)
w
0 DAYTIME NET
z PHOTOSYNTHESIS
< + 0.5-
o E
I77
I 0.5-
x
o NIGHT RESPIRATION-
1.0 I I I
0600 1200 1800
TIME OF DAY
Figure 9 continued
obtained by multiplying oxygen concentration (Figure 9a)
by depth at that time. Percent saturation (Figure 9f)
was calculated for the temperature and salinity at each
time using the formula of Truesdale et al. (1955). The
divergence of Truesdale's saturation values from those
presented in Standard Methods (Amer. Publ. Health Assoc.,
1955) was reviewed by Churchill et al. (1962), who showed
deviations at temperatures less than 250C. Maximum
deviations, however, were less than 5% of the values
from Standard Methods, so the errors incurred in this
study by using Truesdale's values were considered small.
An oxygen rate-of-change curve (Figure 9g) was
constructed from the graph of average oxygen per square
meter. The amount of change of oxygen during each hour
was measured and plotted on the half hour. This raw
curve reflected changes in oxygen concentration under
one square meter due to changing depth from tidal exchange
and diffusive exchange with the atmosphere, as well as
photosynthesis and respiration. The effect of changing
depth was eliminated by multiplying the incremental depth
change for each hour by the average oxygen concentration
during that hour. This value was added to the rate
curve if the tide was falling or subtracted if the tide
was rising.
The final adjustment to the rate-of-change curve
was for oxygen lost or gained by diffusion between the
water and atmosphere (see more complete discussion on
page 62). In general, in the discharge bay only a falling
tide from a high high to a low low stage had a sufficient
current producing a diffusion rate large enough to make
an appreciable correction in the metabolism calculation.
Both rising and falling tidal current velocities were
greater in the control areas making diffusion corrections
more important at all tidal stages. For daytime net
photosynthesis the average difference between the area
under the curves adjusted and unadjusted for diffusion
in the inner bay was 8 percent. At the Fort Island
control area it was 24 percent, while at Hodges Island
(only two measurements) it was 2 percent. Any diffusion
estimate that was incorrect for the discharge bay would
have a relatively small effect on the metabolism calcu-
lation. At Fort Island the effect would be only somewhat
larger.
This laborious method was later modified; average
oxygen concentration, temperature, depth, salinity, and
percent saturation were plotted as before, but the
area-based oxygen curve was not calculated. The rate-
of-change curve (Figure 9h) was obtained by multiplying
the hourly rate-of-change of oxygen concentration by the
average depth at that hour giving the rate-of-change
on an areal basis. The adjustment for diffusion was
made as before.
In all methods the final rate-of-change graph
showed the rise of oxygen resulting from net photosyn-
thesis during the day, and decrease because of respira-
tion at night. Net daytime photosynthesis was taken as
the area under the rate-of-change curve above the zero
rate-of-change line. Nighttime respiration was taken
as the area under the rate-of-change curve below the
zero rate-of-change curve (Figure 9g and 9h).
Dawn-Dusk-Dawn Measurements
In order to gain more data as a check on day-
to-day variability of total metabolism and to reduce
the amount of field labor involved, the dawn-dusk-dawn
method (McConnell, 1962) was used after the first year.
The low point of oxygen at dawn, the high point at dusk,
and the low point the following dawn were measured as
a short-cut method of approximating the true diurnal
curve. Experience in the field showed that the time
of the minimum and maximum was not always at dawn or
dusk. Clouds in the east at sunrise tended to delay
the onset of rising oxygen by an hour or more. Simi-
larly, afternoon thunderstorms often caused the downturn
of oxygen well before dusk. Even on clear days full
diurnal curves showed that oxygen concentration often
would not increase any more in the last two hours before
sunset. The times of dawn and dusk sampling, then, were
often adjusted to the prevailing conditions. Dawn
samples were delayed if the morning was cloudy in the
east. Dusk samples were generally taken about 1-1/2
hours before dusk.
Water samples were drawn, fixed, and titrated
as described before. Diurnal graphs of averaged data
were drawn in the same way as for full diurnals (Figure
10) but, of course, used only three points. Straight
lines were used to connect points for oxygen, tempera-
ture, and percent saturation. Because depth was impor-
tant to the metabolism calculation, the actual daily
pattern was estimated from the observed measurements
and the expected tidal amplitudes for the Crystal River
area published in the U. S. Department of Commerce
tide tables. Because the daily pattern of salinity
change was complex, no attempt was made to extrapolate
between the measured values.
With the three-point dawn-dusk-dawn method,
net production and/or night respiration would be under-
estimated if the minimum and maximum points of oxygen
were not sampled when they occurred. The method also
used fewer replications so that any one unusual measure-
ment would have a greater effect on the calculation of
metabolism. McKellar (1975) gives a more complete
discussion of errors associated with the method.
Figure 10. Graphical format for calculation of community
metabolism using dawn-dusk-dawn data. Open
circles are the average of measurements at
individual stations indicated by solid dots.
Numbers across top of the rate-of-change
graph are diffusion coefficients.
Z E
(\
>- cJ
XO
0
I
E 2.0
0 1.0
W
ar
FU
0E
W
I-
I--
z
I-
Lir
2
I -
cr0
x
0
36 -
34-
+0.5
0.0
-0.5
1200 1800
I
I I I
0
0 -
0600
An analysis of the difference in metabolism
estimates calculated by the dawn-dusk-dawn and full
diurnal curve methods is given in Figure 11. Data points
were read from a subsample of the graphs of full diurnal
curves of oxygen as if that day had been sampled by the
dawn-dusk-dawn method,and daytime net photosynthesis
and night respiration were calculated. Daytime net
photosynthesis would have been underestimated by the
dawn-dusk-dawn method three times in the inner bay by
an average of 33 percent and overestimated twice by a
small amount. Agreement was better at Fort Island and
Hodges Island but would have been over- or underestimated
by up to 25 percent.
Night respiration by the dawn-dusk-dawn method
was only an average of 58 percent of that calculated
by the full diurnal curve method in the inner discharge
bay. At Fort Island the three-point method was only
75 percent of the full curve method on three occasions,
while the full curve value was only 88 percent of the
three-point value two times.
McKellar (1975) for the outer discharge and
control bays at Crystal River found the dawn-dusk-dawn
method to underestimate gross production values (daytime
net production plus night respiration) usually by less
than 10 percent. The average difference between the
two methods was not significant at the 0.05 level.
Figure 11. Comparison of community metabolism estimates
obtained from complete diurnal measurements
of oxygen versus estimates obtained from
dawn-dusk-dawn calculations made using the
same data.
(a) Daytime net photosynthesis.
(b) Night respiration.
-o
LI'
N
01
0
-j
z
UJ
>
Q:
'3
0
-J
0
2
*0
N
,I
Q
-J
>
0
DAWN-DUSK-DAWN,
6 8
g 02/m2day
/N
CHARGE BAY
-AND .
ISLAND
'
6 8
g 02/m -day
DAWN- DUSK-DAWN,
Eley (1970) found that dawn-dusk-dawn estimates averaged
91 percent of gross production and 87 percent of total
respiration in eight laboratory microcosms and 71 percent
of gross production and 52 percent of total respiration
in Keystone Reservoir, Oklahoma when compared to the full
diurnal curve analysis. In this study 61 percent of the
metabolism measurements from the inner bay and 68 percent
from the outer bay were made with the dawn-dusk-dawn
method. Since the apparent underestimation was largest
in the inner bay, these values may be conservative esti-
mates.
Effects of Advection on Calculation of Metabolism
If an increase in oxygen occurred at night because
of advection, an artifact in the rate-of-change curve
was produced which made it appear as if photosynthesis
was occurring. Net production would be overestimated
because the nighttime gain in oxygen would be added to
the actual net production occurring during daylight hours.
Night respiration would be underestimated because the
area of positive oxygen gain would not be counted in the
calculation of respiration. By measuring this omitted
area, night respiration was found to be underestimated
by an average of 1.5 g/m2.day on the full diurnal curves
from the inner bay.
Light and Dark Bottle Measurements
Light and dark bottle studies were made in the
later stages of the project to estimate metabolic com-
ponents of the water column as apart from the metabolism
of the sediments and larger consumer organisms. Bottles
(300 ml, BOD) were suspended at about 0.5 m depth by
small chains secured to a four-foot length of 3/4-inch
PVC pipe floated at each end by a plastic milk carton.
Generally, five replicates each of both light and dark
bottles were put out as soon as the dawn diurnal run
was completed, and picked up at the same time the follow-
ing day. Fixation and titration were as in American Public
Health Association (1971), except that only a 100-ml sub-
sample was titrated because of the 0.0125 N thiosulfate
used. The increase in the light bottle was taken as 24-
hour net production, the decrease in the dark bottle was
taken as 24-hour respiration, and the sum of the oxygen
gained plus that used up was taken as gross photosynthesis.
Other Field Measurements
Diffusion Measurements
At Crystal River the rate of diffusion of oxygen
into and out of the water column tended to be largely
a function of tidal current velocity. Diffusion was
measured at various stages of the tidal cycle using a
small nitrogen-filled plastic dome, which floated on the
water surface (Hall, 1970, based on original work of
Copeland and Duffer, 1964). An oxygen probe measured
the return of oxygen into the dome from the water under
the normal conditions of underwater circulation. A
linear regression was calculated from the raw data.
Although the increase in oxygen in the dome is not
linear, the early response approximates a straight line.
The diffusion rate as g/m2/hr/100 percent deficit was
calculated from the linear regression, area of water
surface covered, volume of the dome, and the observed
saturation value of dissolved oxygen in the water. This
was the maximum rate of diffusion into oxygen-free water
or out of water 200 percent saturated with oxygen.
Figure 12 shows a typical diffusion measurement.
Because of the small number of measurements
taken, assigning diffusion rates to time periods on
the graph was a combination of actual measured values
and estimates based on field experience with the general
magnitudes of tidal currents at different stages of the
tidal cycle in the study areas. The actual diffusion
correction for each hour was calculated by multiplying
the maximum rate selected for that hour by the actual
saturation deficit during that hour.
Figure 12.
Example of two experiments to determine oxygen
diffusion coefficients by measuring the rate
of return of oxygen into a nitrogen-filled
dome floating on the water's surface. Line
through points was obtained by calculating
a linear regression. Meter was calibrated
to give a reading of 10 in air. Data obtained
at Fort Island study area.
30
LU
-j
<0
c) JUNE 25, 1973
.J FALLING TIDE (HIGH HIGH
I TO LOW LOW )
: y = 0.175 + 0.021 x
U- 2 r2= 0.998
LL 20
--
0
z
Z 10
LU
CL
Sy=0.139+ 0.0 85 x
2= 0.988
30 60 90 120
TIME FROM START, MINUTES
U- '0 60 90
LTIM FRO STARITIN MINUTESWLO
Light Penetration of the Water Column
Light penetration through the water column was
measured with a submarine photometer (Tsurumi Precision
Instrument Co., S/N 88130). Light intensity was mea-
sured at 0.1-meter depth intervals from the surface
to the bottom and compared to a deck cell reading insola-
tion incident to the water surface. Results were graphed
on semi-log paper (Figure 13). The extinction coefficient
was calculated was
ln (11/12)
K = I
Z2 Z1
where I1 was light intensity at the shallower depth
(Z1) and 12 was light intensity at the deeper depth
(Z2). K was in units of meter- .
Figure 13.
Examples of submarine photometer measurements
of light penetration through the water column
taken at Fort Island away from the influence
of the power plant discharge plume and in
the inner bay influenced by the plume. Lines
through points were fitted by eye. k, extinc-
tion coefficient.
PERCENT
10
O -
OF SURFACE LIGHT INTENSITY
50 100
DATA ASSEMBLED FROM OTHER PHASES OF THE
CRYSTAL RIVER PROJECT AND ELSEWHERE
One of the major objectives of the overall
research program at Crystal River was to synthesize
the knowledge of the forcing functions outside of the
system and the storage and process operating within
the system. To this end, records of many of these
variables from other phases of the project and elsewhere
are included here to provide a total view of the estuarine
ecosystem. These data are used for obtaining values for
the model simulations and in determining if the simula-
tion results are reasonable.
Energy Sources and Inflows
Affecting the Inner Bay
Seasonal and diurnal patterns of some of the
external factors shown in Figure 4 are given below.
Sunlight
In Figure 14 is the average daily insolation by
month measured at Tampa, Florida, 97 km to the south of
Crystal River (Water Information Center, Inc., 1974).
Figure 14. Average daily insolation by month at Tampa,
Florida (Water Information Center, Inc.,
1974).
8000
0
cu
^ 6000
0 4000
.J
0
0
C( 0
I. --
I I I I I I I i I
I I I I I I
J F M AM J JY A S O N D
I
I
Peak insolation months (about 6000 Kcal/m2-day) were
April and May at the very end of the winter-spring dry
season. Daily summer values were lower due to frequent
cloudiness from convective storms.
Wind Direction and Speed
Wind rose diagrams by season are given in Figure
15 (Fla. Power Corp., 1972). Summer winds were pre-
dominantly westerly and easterly as influenced by the
large-scale circulation about the shifting position of
the subtropical high-pressure system and by the more
local regional land-sea breeze system. With the change
in the fall and winter to weather patterns dominated by
frontal systems, the predominant wind direction shifted
to northerly directions. Average wind speed as given
in Table 2 (Fla. Power Corp., 1972) was lowest in the
summer and highest in fall and winter due to the strong
winds associated with frontal passages.
Ambient Air Temperature
In Figure 16 are monthly mean, mean maximum,
and mean minimum daily temperatures at Tampa, Florida
(Fla. Power Corp., 1972). Diurnal variation was smallest
during the summer months when the climate was primarily
under the influence of the subtropical high pressure
system, and frontal systems usually remained well north
Figure 15. Wind direction by season at Crystal River site.
Bars are percent of readings occurring from
each compass bearing (Florida Power Corpora-
tion, 1972).
DEC., JAN., FEB. MAR. ,APR., MAY
JUN.,JUL., AUG.
SEP!,OCT., NOV.
Table 2. Seasonal comparison of average wind speed at
Crystal River site (Fla. Power Corp., 1972)
Average wind speed, mph
11.1
9.5
12.0
12.0
Annual average
Season
Spring
Summer
Autumn
Winter
11 .4
Figure 16. Monthly mean air temperature at Tampa,
Florida (Water Information Center, Inc.,
1974).
77
100
100. i i -- i -- -j,...l...L ,j -- 1 1
EXTREME HIGH,.-'' AVERAGE DAILY MAXIMUM --.-
MONTHLY MEAN
80
SL- AVERAGE DAILY MINIMUM
o .. --- -.,
cr 60
.. / .__
J 40-
CL
/ 1
20 EXTREME LOW
O JFM I i J I J I
J F M A M J JYA S 0 N D
of the area. Minimum temperatures dropped sharply in
October as cold fronts began penetrating into Florida,
and remained low through the winter when the climate was
characterized by cold air advection following frequent
frontal passages.
Precipitation
Monthly mean precipitation at Tampa is presented
in Figure 17 (Fla. Power Corp., 1972). About 60 percent
of the yearly rainfall occurred from June through Sep-
tember and was associated with showers and thunderstorms
in tropical air masses. During the extensive eight-month
dry period extending through May, precipitation was mainly
associated with frontal systems.
Stocks of the Inner Bay
Assembled below are data on stocks of organisms
and other quantities important within the inner bay
system.
Water Temperatures
Weekly average water temperatures at various
locations in the discharge canal, discharge study area,
and intake area during the course of this study are given
in Figure 18. Buoy locations are given in Figure 6.
Weekly average electricity generated by units 1 and 2
Figure 17. Monthly mean precipitation at Tampa,
Florida (Water Information Center, Inc.,
1974).
15 --- --
.)
<:
C
o
". 10 -
O
t-
0. L L J J L. I J._
J F M A M J JYA S O N D
|
RSS
HIDE
