Title: Effects of elevated turbidity and nutrients on the net production of a tropical seagrass community /
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00097405/00001
 Material Information
Title: Effects of elevated turbidity and nutrients on the net production of a tropical seagrass community /
Physical Description: vii, 135 leaves : ill., map ; 28 cm.
Language: English
Creator: Caldwell, John William
Publication Date: 1985
Copyright Date: 1985
Subject: Seagrasses   ( lcsh )
Environmental Engineering Sciences thesis Ph. D
Dissertations, Academic -- Environmental Engineering Sciences -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Thesis: Thesis (Ph. D.)--University of Florida, 1985.
Bibliography: Bibliography: leaves 129-134.
Additional Physical Form: Also available on World Wide Web
General Note: Typescript.
General Note: Vita.
Statement of Responsibility: by John William Caldwell.
 Record Information
Bibliographic ID: UF00097405
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000867588
notis - AEG4447
oclc - 014364425


This item has the following downloads:

PDF ( 4 MBs ) ( PDF )

Full Text








This research was funded by Contract No. 99200-7213

from the Florida Department of Transportation to

F.J. Mature, Jr. and J.W. Caldwell. I wish to thank

Mr. Gary Evink and Mr. Bill Jordan of FDOT for their as-


I thank my chairman, Dr. Thomas L. Crisman, and

co-chairman, Dr. Frank J. Maturo, for their helpful discus-

sions, suggestions and support during this work. I thank

Dr. Clay Montague for his insight and input into the modeling

effort, Dr. Frank Nordlie for his discussions of productivity

questions and use of his laboratory equipment, the late

Dr. John McCreary for his instruction in particle settling

rates, and Dr. Gabriel Bitton for his agreeing to serve on

my committee at the eleventh hour.

Numerous former students have provided major help in

both the field and laboratory work. They include Bob and

Gail Knight, Kate Benkert, Art Watson, the late Bill Coggins,

Michelle Hosack, Karen Ogren, Jim Hill and Jose Gallo. I

especially thank John Higman and Jerri Ann Blair for their

unselfish help and support during all phases of this endeavor.

I wish to thank CH2M HILL for making its resources

available during the final phases of this dissertation. I

thank Tere Myers for her typing of this manuscript and

Larry Foley for drafting the figures.

My family deserves major credit for their continued

support and reinforcement of my educational desires over the

years. I sincerely thank my wife, Jo, for her encouragement

and support during both my graduate degrees.

Finally, I thank my son, John Edward, who provided me

with the final push to finish this work.



ACKNOWLEDGEMENTS. ................................. ii

ABSTRACT. ......................................... v


I PROBLEM STATEMENT........................ 1


Introduction................................ 5
Materials and Methods .................... 8
Results .................................. 21
Discussion....... ......................... 49
Summary ................................. 55


Introduction ........................... 58
Materials and Methods .................... 59
Results .................................. 65
Discussion................................. 71
Summary ....................... ........... 75


Introduction ............................ 77
Materials and Methods .................... 79
Results ...................................... 106
Discussion............................... 116
Summary ..................... ............. 123

V SUMMARY OF FINDINGS ...................... 125

REFERENCES......................................... 129

BIOGRAPHICAL SKETCH .............................. .. 135


.dat OIL

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy




December 1985

Chairman: Thomas L. Crisman
Co-chairman: Frank J. Maturo, Jr.
Major Department: Environmental Engineering Sciences

Dredging effects on seagrass communities in the Florida

Keys were examined by 1) comparing impacts on net production

resulting from dredging and natural weather events, 2) de-

termining changes in community photosynthetic efficiency,

3) evaluating shading and nutrient effects on net produc-

tion, and 4) developing a systems dynamics model.

Net community production was estimated during numerous

meteorological and dredging events using the Odum-Hoskins

oxygen technique in flow-through field microcosms. In other

experiments, shading and nutrients (phosphorus, nitrate, and

ammonia) were manipulated to simulate dredge plume conditions.

The model examined the relationships between seagrass biomass,

water column and sediment nutrients, detritus, and consumers.

The greatest depression in net community production re-

sulted from severe thunderstorms and dredging events,

respectively. Net community production measured two years

after dredging showed an approximate four-fold decrease.

In field microcosm experiments, significant interaction

occurred between shading and nutrient concentration.

Significant metabolic reduction occurred due to shading,

even with higher nutrient conditions. All but the lowest

concentration resulted in significantly increased production

in the light. Qualitative comparison with a control showed

an enhancement of production only in the light.

The model of seagrass production was most sensitive to

changes in nutrient-seagrass relationships, seagrass pro-

duction estimates, and seagrass-light interactions. Recovery

of seagrass biomass following numerous dredging events

(3.5 years) was longer than that from the estimated total

annual thunderstorms encountered (1 year) but shorter than

recovery from hurricane events (4.1 years).

The effects of short-term dredging on net community

production as shown by field experimentation and the model

were less severe than some weather events because of dif-

ferences in duration and intensity. Lowered post-dredging

photosynthetic efficiency, as defined by unit biomass pro-

duction per unit light, was a result of degraded system

function possibly from sediment scour from the dredge cut.

The major deleterious effect of dredging resulted from

shading, although some production enhancement could occur

from nutrient release. The model suggested that dredging

effects were prolonged because of current scour from the

dredge cut, that recovery time was comparable to hurricane

events, and that more investigation is needed in sea-

grass-nutrient and light-production relationships.


Replacement of the bridges in the Florida Keys, some of

which have been in place since the early 1900's, is a major

road construction project being carried out by the Florida

Department of Transportation (FDOT). For the most part,

construction of the new bridges has been adjacent to the

older structures. As a part of the construction activities,

the need arose to barge bridge span sections to the site for

inclusion in the structure. This activity required the

dredging of a channel through shallow marine grass flats to

allow the span barge access.

Considerable interest was expressed by the Florida

Department of Environmental Regulation (FDER) concerning the

effects of elevated turbidity from these construction

activities on the nearby biological communities. The highly

flocculent nature of the sediments, the strong tidal currents

known in the area and the close proximity of potentially

sensitive marine ecosystems all combined to create concerns

regarding the environmental effects of the dredging

activities. Of special interest were the extensive seagrass

beds (primarily, Thalassia testudinum), the dominant community

closest to the planned dredging activities.

Early observations of high turbidity caused by weather

events such as cold front passage and violent tropical thunder-

storms indicated these ecosystems are subjected normally to

periodic high-intensity turbidity pulses. The extensive

areas of seagrasses present in the area suggested this com-

munity, in particular, might be adapted to these pulses and

therefore would not be impacted severely by similarly inter-

mittent dredging activities in the vicinity. Obviously,

outright removal of the seagrass by the dredge would be the

most severe impact. This was not considered due to the

small amount removed compared to the total acreage present.

Only those impacts relating to effects from the dredge plume

were considered.

Therefore, dredge plume effects should be 1) comparable

in intensity to those encountered during natural weather

events, 2) short term, depending on the longevity of the

dredge plume, 3) deleterious to net community production as

a consequence of shading, 4) beneficial to net production

arising from nutrient release from the sediments, and

5) dissipated quickly and show no measureable effect on net

community production on a long-term basis.

The objective of this study was to test these hypotheses

by 1) a comparison of the response of net community production

(=metabolism) to a dredge plume and natural weather events,

2) analysis of net community metabolism in seagrass commu-

nities prior to and two years after being subjected to a

dredge plume, 3) field experimentation with two major impacts

believed to be associated with the dredging event, namely

shading and nutrient addition, and 4) modeling of the sea-

grass ecosystem to identify factors that are most important

in controlling change in seagrass production, to compare

long-term response and recovery of the seagrass model fol-

lowing various perturbations, and to point out areas of

research requiring further investigation.

Field measurements of community metabolism utilized the

upstream/downstream oxygen technique of Odum and Hoskins

(1958). This method, which is used throughout the field

experiments in the determination of net community production,

is explained in Chapter II. This technique has received

considerable criticism (Bittaker and Iverson 1976; Phillips

and McRoy 1980) on the basis that diffusion is largely

uncontrolled and only roughly estimated and, because of

large internal lacunae, macrophytes tend to store oxygen in-

ternally which is not released, therefore providing an under-

estimate of production. In this study, the issue of diffusion

was eliminated through the use of covered microcosms which

entrained the water and prevented gaseous exchange with the

atmosphere. Underestimates of seagrass production were not

an issue here since it was the response of the total community

that was of interest. Although internal recycling of oxygen

more than likely did occur in the seagrass blades, it is

believed this was held to a minimum as a result of the use

of flow-through microcosms which maintained water movement

through the plants (Westlake 1967; Fonseca et al. 1982).

A final caveat regarding the model is included here.

A misconception of model output exists when it is believed

that output should reproduce nature exactly and that the

model can always be used to prescribe policy. The major

reasons why most models cannot be used for these purposes

involves the lack of information required to fully document

even the simplest of models. This necessitates the use of

estimates for internal model constants that could lead to

misleading model output. Further, differences in inter-

pretation of how the "real world" operates in its myriad

relationships can also provide a source of error.

Perhaps the greatest benefit derived from the model is

the learning experience which evolves during the model

development process. From the very beginning of model

construction, the modeler is forced to think about how

components operate internally and connect to other

components. This evaluation process, which then has to be

translated into precisely defined mathematical relationships,

is the real power of modeling.

The following sections are divided into four chapters.

Chapter II includes an analysis of short- and long-term dred-

ging effects on seagrass community production. Chapter III

contains field investigation of two of the believed major

impacts relating to a dredge plume, and Chapter IV contains

results and analysis of the simulation model. A summary

chapter (Chapter V) is provided at the end which contains an

overview of the major findings of this work.



The effects of elevated turbidity resulting from

resuspension of sediments during dredging operations are

major concerns in shallow coastal marine environments. Many

of these effects, which can be detrimental or beneficial,

have been discussed in reviews by Stern and Stickle (1978)

and Johnston (1981). Specific investigations of dredging

events relevant to this study have addressed the areas of

light attenuation effects on primary production in the water

column (Stross and Stottlemeyer 1965; Biggs 1968), light

limitations of benthic primary production (Taylor and Solomon

1968; Cambridge and McComb 1984), reduction in community

metabolism as a result of high turbidity (Odum and Wilson

1962; Odum 1963), resuspension of nutrients from a dredge

plume (Odum 1963; Stross and Stottlemeyer 1965; Flemer et

al. 1968; Flemer 1970; Sherk 1971), and the recovery of a

benthic ecosystem from a dredging event (Conner and Simon


During replacement of the bridges in the Florida Keys

dredging of shallow marine areas was needed to facilitate

building of the new structures. Because of concerns for

turbidity-related impacts on nearby seagrass communities,

documentation and examination of impacts resulting from the

proposed bridge construction-related dredging activities

were required.

Early field observations of the dredge in operation

indicated that the dredge plume was of fairly short temporal

duration following dredge stoppage and that little or no burial

of the surrounding community by dredged sediments occurred.

Furthermore, observations of areas already dredged indicated

that strong tidal currents present in the area would also

contribute to turbidity levels.

It appeared that the resultant turbidity activities

would impact the seagrass ecosystems in two stages: an

initial, intensive pulse during the dredging event and a

longer, lower level, gradually decreasing turbidity plume

resulting from current scour of sediment from the dredge

cut. It was expected that the magnitude and duration of the

initial turbidity pulse would be largely a function of the

level of dredging activity, localized current regimes and

the sediment grain size. The larger the volume of dredging

activity, the more sediment that would be released into the

water column; the higher the water currents, the greater the

potential dispersion of the dredge plume; and the smaller

the grain size the greater the distance of dispersion by

these water currents. During dredging, currents would tend

to disperse the plume, thereby lessening its effect on the

immediate vicinity through dilution, but also causing a

larger area to be impacted.

The rationale for this study included the testing of

the hypothesis that, on a short-term basis, the effects on

seagrass net community metabolism were comparable to those

caused by weather events normally encountered by the

communities. It was hypothesized that over the long-term

period no appreciable degradation in net community metabolism

would occur, although an initial decline might be shown due

to current scour from the dredge cut. The objectives of

this study, then, consisted of a comparison of the reduction

in net community metabolism caused by dredging and naturally

occurring weather events and a comparison of pre- and post-

dredging photosynthetic efficiencies to evaluate any residual

effects due to dredging operations.

Because of the observed erratic nature of a dredge plume

coupled with quickly changing weather, a measurement technique

was required which could monitor response of the seagrass

community to frequently changing light. Measurement of com-

munity metabolism using a flow-through microcosm was selected

because it was believed that this technique would provide a

quantifiable index of the whole system response to the per-

turbations of interest (Walsh et al. 1982). Also, preliminary

measurements indicated this technique would provide 1) a

rapid response to short-term changes in the community's light

environment, and 2) a movable apparatus enabling examination

of several different areas within the community as the dredge

proceeded along its course. As an aid in interpretation of

any effects measured due to dredging, natural

meteorological events were also monitored in an identical

manner for comparison. As many different dredging and

weather events as possible were monitored, compared, and

ranked based on their level of impact on net community pro-


Materials and Methods

The study was conducted at the western end of 7-Mile

Bridge in the middle Florida Keys (Figure 1). Construction

of the new structure paralleled the old bridge to the south,

necessitating the dredging of a channel through the shallow

areas to a depth of approximately 1.5 m at mean low water

(trLW) and a width of 46 m to facilitate the movement of the

span and piling barges. A clamshell dredge (crane with a

bucket) excavated the sediment which was then barged to

upland disposal sites. Dredging was intermittent throughout

its period of operation because mud barges were usually

exchanged only during high tide and the clamshell dredge

required frequent repair.

The benthic biotic community in these areas is character-

ized by extensive beds of turtle grass (Thalassia testudinum)

along with less dense patches of manatee grass (Halodule

wrightii) and shoal grass (Syringodium filiforme). Inter-

spersed between these grass areas are sea rod (Plexaura

homomalla), sea fan (Gorgonia ventalina), and sponge com-

munities in the deeper swash channels. Other nearby areas

that are extremely shallow and often exposed at low tide are

dominated by dense areas of the finger coral, Porites sp.,

*-V, I


N I( -

..r ,.. -.


Figure 1. Map showing sampling location in the Florida Keys.




clumps of coralline algae, and patches of Thalassia


Strong tidal currents are a dominant physical feature

of this area. These currents likely serve as major exchange

mechanisms between adjoining communities by importing and

exporting materials. In addition to materials exchange,

these currents can also function as a rapid metabolic waste

removal mechanism.

Estimates of metabolism of the seagrass-dominated com-

munities were made using the upstream/downstream oxygen

method of Odum and Hoskins (1958). Flow-through microcosms

(=tunnels) which were placed over the seagrass community

were constructed of a polyvinyl chloride (PVC) framework

covered with 6 mil clear polyethylene (Figure 2). The tunnel

length was approximately 4 m with an average height of

0.35 m. Residence time of water passing through the chamber

was regulated with baffle plates applied to each tunnel end.

Each plate contained numerous holes which were plugged with

corks until a desired current speed was attained. Current

regulation through the system was important because residence

times which were too short resulted in oxygen differences

below the detection limit of the oxygen monitoring equipment.

Residence times approaching 20 min were found to provide

adequate upstream/downstream oxygen differences under most

situations while still allowing flushing of the system.

Microcosms were sampled by pumping water from either

the upstream or downstream end to the sensing sonde of a

To Sensing Probe;
Dissolved Oxygen
(Upstream) ,


To Sensing Probe:
Dissolved Oxygen-

Surface, .

Baffle Plate (Typ. Opp. End)

Figure 2. Diagram showing field set-up of flow-through


Hydrolab, Inc., Model 6000 Water Quality Surveyor (Figure 2).

Triplicate dissolved oxygen measurements were collected by

alternating these with single temperature and conductivity

recordings collected during each sampling of a tunnel end.

Other routine measurements noted included hourly readings of

solar input (Lamda Instruments, Li-Cor Model 175), and percent

transmission of light through water (Hydroproducts, Inc.,

Model 612S Transmissometer).

The average residence time of water in the tunnel was

determined at each station from several timed dye flow studies.

It was assumed that the residence time of water in the tunnel

represented an average flow time over the entire tidal cycle.

Tunnels were sampled on a frequency of five minutes. In

other words, given a 20 min residence time, the upstream end

would be sampled every 5 min for fifteen minutes. The 5 min

sampling frequency at the lower end of the tunnel then began

at the 20 min interval (from the start of the sampling cycle)

and was repeated every 5 minutes. In this manner, a plug of

water was followed as it entered the tunnel and then was

sampled again at the exiting end in order to determine the

change in oxygen over the time interval. This procedure was

repeated throughout several consecutive diurnal periods and

varying environmental conditions.

The microcosm was set up parallel to the north-south

tidal currents, approximately 50 m south of the bridge and

slightly ahead of the dredge's path which paralleled the

bridge as it proceeded to the west. This enabled the

collection of background data to ensure proper working of

the microcosm prior to sampling during the dredging event.

As the dredge moved to a position directly north of the

tunnel, the system was monitored to record the metabolic

response of the community to the presence of the dredge plume

in the water column. Response of the seagrass community to

weather disturbances was monitored in an identical manner.

All dissolved oxygen levels were corrected for tempera-

ture and conductivity using the following relationship:

CDO=1-[(3.439+0.0361/(T+22.12)) X (C/1000)]

where CDO=corrected dissolved oxygen

T=temperature in C

C=conductivity in mhos/cm at 250 C

constants in the equation are specific for the

monitoring instrument and were supplied by the

manufacturer (Hydrolab, Inc., Austin, Texas).

Diffusion of oxygen through the polyethylene plastic

covering the tunnels was evaluated according to the


Q = D x S ((PI-P2)/T)

where Q=oxygen diffusion through polyethylene (ccO2/cm2/sec)

D=diffusion constant for polyethylene

S=solubility constant for polyethylene

P1,P2=partial pressure of oxygen inside and outside of

tunnel, respectively

T=thickness of plastic

(From Crank and Park 1968)

Calculations indicated less than 1% of the ambient oxy-

gen diffused through the plastic during an average 20 min

residence time, so this diffusion type was ignored.

The oxygen production values were calculated according

to the following equation:


where ROC=rate of oxygen change (gODm-2hr- )

DOD=dissolved oxygen at the downstream tunnel end

DOU=dissolved oxygen at the upstream tunnel end

D=mean tunnel depth (0.35m)

T=residence time of water in the tunnel in factors of

an hour

A stylized diurnal rate of change curve is shown in

Figure 3. A rapid increase is shown in the morning hours

with a peak occurring shortly after mid-day. Net production

was usually confined to the time period from approximately

2 hours after sunrise to 2 hours before sunset. Integration

of the area under the curve yields the average community

metabolism for any desired time period. Net community

metabolism was calculated as all areas under the curve with

metabolism greater than 0 g 2m-2hr-1 (the portion of the

curve above the 0.0 rate line). Negative oxygen production

values after sunrise and before sunset were subtracted from

the total net production value.

Dredging and weather effects on net community metabolism

were compared on the basis of the percent reduction of net

community metabolism by each type of occurrence. This

U V)


LU Net
o Community
< / Production

-u - .- -.. . .- - - -. . 0 .0



Figure 3. Ideal production curve showing area of net
community production.

comparison was evaluated over two time intervals. First, an

assessment was made of the effect at the time of the event

based on a comparison of a projection of the net community

metabolism level had the event not occurred and the actual

measured values during the perturbation. The projection was

made by connecting the data points at the beginning and end

of the perturbation time period (Figure 4). The integrated

solar values at hourly intervals are plotted across the top

of each productivity curve for comparison. The reduction in

net metabolism was calculated by the following:


where RED=reduction in net community metabolism

resulting from the event

P=projected metabolism had the event not occurred

for the discrete event time interval (based on

straight-line connection of points between the

last measured value before the perturbation and

the first measurement after the perturbation).

A=actual measured net metabolism during the event

time interval

For comparison between events, a percent reduction for

each event was determined by

%RED=(P/(P + A) x 100

where %RED=per cent reduction in net community

metabolism based on the timing of the event

P and A =from the previous equation

/ RED = P/P +A
0 % RED= (P/P+A) x 100
%DTR (P/DT) x 100


Figure 4. Ideal production curve showing the method
used to calculate the reduction in net
production due to perturbations,

P=projected metabolism calculated from the

previous equation

The second comparison involved calculation of the

event's effect in relation to the daily net community

metabolism. In this case, the percent reduction in daily

net metabolism was calculated by:

%DTR=(P/DT) x (100)

where %DTR=per cent reduction in daily net community


R=reduction in net metabolism, calculated as shown


DT=daily net community metabolism

Values for all events measured were compared and ranked

in descending order according to the severity of reduction

in net community metabolism.

Natural and man-made turbidity events were categorized

and compared as shown in Table 1. As many of these events

as possible were monitored to compare their effects on net

community production.

To determine possible long-term effects on seagrass net

community metabolism due to dredging, photosynthetic

efficiencies were calculated from data gathered during the

summer of the main dredging operation and compared with

similar data collected two years later. Average net

production values from each time interval were normalized

for solar radiation, water depth and clarity, and biomass

standing stocks.


Table 1. Categories of natural and man-made turbidity events


Severe thunderstorms
early forenoon

Severe thunderstorms
late afternoon



dredging (with silt cur-

dredging (with no silt

tug plume (caused by pro-
Peller scour from barge
tugs in shallow seagrass

Normal thunderstorms
local rain, light blockage

Distant thunderstorms
no local rain, sun blocked
by distant storm

Overcast, rain

Scattered clouds

Solar radiation reaching the community was calculated

from light transmission data measured in the field and by

the following equation:
LD=LS x e x D

where LD=Light at the desired water depth (watts m-2
LS=Light at the surface (watts m-2)

D=Depth of the water column in meters

L= -In T


where T =fraction of percent transmission of light

PL-optical pathlength of the transmissometer

Standing stock biomass was determined from 4-inch

diameter PVC cores collected in the field from each tunnel

to a depth of approximately 45 cm. Four cores containing

above- and below-ground biomass from each microcosm were

field-sieved through a 500 micron mesh screen, preserved

with 5% formalin, stained with rose bengal, and transported

to Gainesville for sorting and weighing. All samples were

sorted in the laboratory and all recognizable above- and

below-ground plant material removed. The plant material was

dried at 1051C for approximately one week to determine dry

weight. The sample was then heated to 550'C for 1 hour to

determine loss on ignition (ash-free weight).

Productivity data were normalized by dividing the

average net production by the calculated solar radiation

reaching the community and by the dry weight and ash-free

biomass estimates. This yielded the net photosynthetic

efficiency (g02g- dry and ash-free biomass watt-1) of the

seagrass community prior to and two years after the dredging



Effects of weather and dredging on net community pro-

duction are shown in the figures described below. Hourly

integrated solar radiation values are plotted in the upper

part of each figure. The duration of each event as well as

the amount of reduction of metabolism is indicated on each

productivity curve. The productivity curves often show highly

variable swings in oxygen values which are not reflected in

the solar curves. Integration of the solar curves in the

field on an hourly basis, however, allowed direct comparison

of solar changes with changes in the production graphs in

all instances of major perturbation. The duration of

low-level perturbations were determined from field logs.

During a dredging perturbation, the effects of background

"noise" due to variable solar condition have not been

removed. The response of the community during this time

interval was attributed solely to the disturbance being


The effect of a severe thunderstorm (white squall) oc-

curring early in the afternoon is indicated in Figure 5 as

the solar radiation peaked at approximately 1300 hours and

then dropped drastically. Most of the afternoon net produc-

tion was eliminated by this event, including the midday peak.

This resulted in a 68.5% reduction in the net metabolism

! 500

6 8

14 16

20 22 24


7 % RED= 68.5
- % DTR = 39.9
S 1.0


-1.5 TSE Severe Thunderstorm, Early



6 8 10 12 14 16 18 20 22 24
Time of Day
Figure 5. Reduction in net community production as a
result of a severe thunderstorm occurring in
the early afternoon. Solar radiation is shown
at the top of the figure. Abbreviations used are
explained in the text.

during the event and 39.9% across the entire day.

The effects of hurricane-associated thunderstorms on

net community production are shown in Figures 6 and 7. As

indicated in the figures, intense weather disturbances such

as these with highly variable solar input can cause extreme

fluctuations in production and can drive the oxygen rate

curve negative. In Figure 6, depression of net metabolism

during the time the storms occurred ranged from 58.6 to

62.9%, while the effect on the daily value went from a low

of 12.0% to a high of 46.1%. Solar values peaked at noon,

remained level until 1700 and showed a rapid decline

thereafter. An occurrence of net production is shown at

night (after 2200 hrs). This will be discussed in the

following section. Generally low and variable solar input

is shown in Figure 7 which is reflected in large swings in

the production curve. Intense weather systems such as these

cause a loss of approximately 50% of the daily net production.

The effect of a severe thunderstorm (white squall) oc-

curring late in the afternoon is illustrated in Figure 8.

Approximately 57% of the net production and 22.5% of the

daily net production were lost during the thunderstorm.

Solar insolation showed a sharp decline from the start of

the thunderstorm (1600 hours) until its end at 2000 hours.

The effect of a distant thunderstorm which is defined

as a storm which occurs in the distance usually in late

afternoon, blocking the sun but causing no local rainfall,

is shown in Figure 9 and 10. The reduction in metabolism

6 8 10 12 14 16 18 20 22 24


() (2) (I) % RED = 62.9
.5 0% DTR = 46.1

(2) % RED = 58.6
S1.0 % DTR = 12.0




S -1.0

-1.5 H Hurricane Related Thunderstorms



6 8 10 12 14 16 18 20 22 24
Time of Day

Figure 6. Reduction in net community production as a
result of hurricane-related thunderstorms.
Solar radiation is shown at the top of the figure.
Abbreviations used are explained in the text.


1000 Solar

6 8 10 12 14 16 18 20

0.50- H (1) % RED = 64.4
()I % DTR = 31.3
(2) % RED =76.9
% DTR 18.3

7 0.30




U 0.10


IH Hurricane Related Thunderstorms

6 8 10 12 14 16 18 20

Time of Day
Figure 7. Reduction in net community production as a
result of hurricane-related thunderstorms.
Solar radiation is shown at the top of the
figure. Abbreviations used are explained in
the text.


1000- Solar
0 Radiation
: 500.

6 8 10 12 14 16 18 20 22 24


2.0- (1) %RED 24.1 DC TSL
%DTR 16.9 ) (2)
(2) % RED= 5Z 9
.5 % DTR = 22.5





-1.5 DC- Dredge Plume with Silt Curtain
TSL Severe Thunderstorm, Late



6 8 10 12 14 16 18 20 22 24
Time of Day
Figure 8. Reduction in net community production as a result
of dredging with a silt curtain and a severe
thunderstorm occurring late in the afternoon.
Solar radiation is shown at the top of the figure.
Abbreviations used are explained in the text.

6 8 10 12 14 16 18 20 22 24

%RED= 9.6
%DTR 2.7


TD Thunderstorm, Distant

6 8 10 12 14 16 18

Time of Day

20 22 24

Figure 9. Reduction in net community production as a
result of a distant thunderstorm. Solar radiation
is shown at the top of the figure. Abbreviations
used are explained in the text.



2 500 -




-C -


I -



6 8 10 12 14 16 18

% RED = 58.8


TD Thunderstorm, Distant

6 8 10 12 14 16
Time of Doy

18 20 22 24

Figure 10. Reduction in net community production as a
result of a distant thunderstorm. Solar radiation
is shown at the top of the figure. Abbreviations
are explained in the text.

<, 1.0


0 0






20 22 24

in Figure 9 is 9.6% during the event and 2.7% for the day.

The response of the production curve during this time period

is reflected in the solar curve which shows a slightly

steeper decline between 1500 and 1600 hours as compared to

values shown for other time periods. Between 1600 hours and

1730 hours the decline is not as great. Reasons for

fluctuation in the curve between 1100 and 1500 hours were

not noted in the field log and are not included in the

analysis. A more severe reduction (58.8%) resulting from a

similar thunderstorm is shown in the curve in Figure 10.

Figures 11 and 12 demonstrate the effect of normal

(short duration with local rainfall) thunderstorms on

metabolic activity. Effects of these events at the time the

storm occurred were very similar on both days (approximately

21%). Overall effects on the day's production were also

very similar, ranging from 6.2-6.8%. Unfortunately, solar

records were lost for both days.

Although it is a rare weather event in the Florida Keys

in the summertime, an overcast, rainy day produces the re-

sponse shown in Figure 13. This event caused a reduction

in net community metabolism of 32.1% during the time it oc-

curred and an overall reduction in the daily net production

of 6.9%. Other variation shown in the production curve is a

result of gradual cloud build-up but is not included in the

analysis as the exact timing of the events as they occurred

were not recorded.

Scattered clouds (Figures 14-16) were the final weather

(I) % RED= 48.0
% DTR 9.7
(2) % RED = 44.1
% DTR= 17.2
(3) % RED =21.7
% DTR = 6.2



S 1.5

E 1.0




5 o'




6 8 10 12 14 16 18
Time of Doy

20 22 24

Figure 11. Reduction in net community production as a
result of a tug plume, dredging with no silt
curtain, and a normal thunderstorm. Abbreviations
used are explained in the text.


TP Tug Plume
DNC Dredge Plume with no Silt Curtain
TN Thunderstorm, Normal


"~/ Y




1.5- TN
S% RED 21.8
N %DTR 6.8
E 1.0-



~ -LO

-1.5- LEGEND

TN Thunderstorm, Normal


6 8 10 12 14 16 18 20 22 24

Time of Day
Figure 12. Reduction in net community production as a
result of a normal thunderstorm. Abbreviations
used are explained in the text.

1000 Solar

5006 8

6 8 10 12 14 16 18 20 22 2,

% RED = 32.1
/ DTR= 6.9





0 0.5

- 0.5



6 8 10 12 14 16
Time of Day

18 20 22 24

Figure 13. Reduction in net community production as a
result of overcast, rainy conditions. Solar
radiation is shown at the top of the figure.
Abbreviations used are explained in the text.

OR Overcast, Rainy


1000- Solar
S? Radiation


6 8 10 12 14 16 18 20 22 24



SC % RED 15.9
1.5 % DTR 5.8

1.0 -



c -IC

-.5 SC Scaered Clouds



6 8 10 12 14 16 18 20 22 24
Time of Day
Figure 14. Reduction in net community production as a
result of scattered clouds. Solar radiation
is shown at the top of the figure. Abbreviations
used are explained in the text.


1000' Solar

6 8 10 12 14 16 18 20 22 24



1.5 DC SC
(1) (2) (1) % RED= 50.1
L % DTR = 23.8

S 1.0- (2) % RED = 22.3
E % OTR = 10.3



-0 .5

( -1.0-

-1.5 DC Dredge Plume with Silt Curtain
SC- Scattered Clouds



6 8 10 12 14 16 18 20 22 24
Time of Day

Figure 15. Reduction in net community production as a
result of dredging with a silt curtain and
scattered clouds. Solar radiation is shown
at the top of the figure. Abbreviations used
are explained in the text.


1000 Solar

6 8 10 12 14 16 18 20 22 24

2.5 -
(1) (2) (1) % RED 16.0
2.0 (2) % RED = 10.1


3 0.5



-1.5 DC Dredge Plume with Silt Curtain
SC Scattered Clouds



6 8 10 12 14 16 18 20 22 24
Time of Day
Figure 16. Reduction in net community production as a
result of dredging with a silt curtain and
scattered clouds. Solar radiation is shown
at the top of the figure. Abbreviations used
are explained in the text.

category observed and resulted in a reduction of 10.1-22.3%

during the period in which this event was recorded and a

5.8-10.3% alteration of the daily amount. Effects of this

weather type are not well documented in the solar record

since the clouds pass over at a frequency greater than the

hourly intervals over which sunlight was integrated. Again,

field logs were used to document the occurrence of this weather


Dredging effects on net community metabolism are highly

variable as shown in Figures 8, 11, and 15-23. Reduction in

net production during dredging with a silt curtain

(Figures 15-21) ranged from a low of 10.0% to a high of 74.9%.

Reduction in daily net metabolism showed a low value of 1.3%

and a high of 23.8%. Solar radiation showed no major fluc-

tuation during these events. Of this group, Figure 20 showed

evidence of a small amount of net production occurring in

the dark. This will be discussed in the following section.

The effect of dredging with no silt curtain is shown in

Figures 11, 22 and 23. This event is also extremely variable

in its effects, the reduction in net community metabolism

ranging from a low of 44.1% (Figure 11) to a high of 98.0%

(Figure 22). A decline in solar radiation is indicated in

Figure 22 during this dredging event which reinforces the

effect of dredging. The effect on daily net production was

less drastic, showing a low of 17.2% and a high value of 43.1%.

The effects of another dredging-related event, shown in

Figures 11, 18, and 24, is the turbidity plume resulting

1000 Solh
I, Rac
f 500


18 20 22 24


% RED= 23.5
%DTR = 21.6




1.0 -


" O




6 8 10 12 14 16 18 20

Time of Day

22 24

Figure 17. Reduction in net community production as a
result of dredging with a silt curtain. Solar
radiation is shown at the top of the figure.
Abbreviations used are explained in the text.


10 12 14 16



DC Dredge Plume with Silt Curtain



S Radiation

S 500

6 8 10 12 14 16 18 20 22 24


2.0 DOC- --TP
(1) (2)

1.5 (1) %RED 74.9
/% DTR = 6.8
(2) %RED= 17.1
S 1.0 % DTR = 1.9

S 0.5-


-1.5 DC Dredge Plume with Silt Curtain
TP- Tug Plume



6 8 10 12 14 16 18 20 22 24
Time of Day
Figure 18. Reduction in net community production as a
result of dredging with a silt curtain and a
tug plume. Solar radiation is shown at the
top of the figure. Abbreviations used are
explained in the text.

1000 Solar



6 8 10 12

14 16 18

20 22 24


% RED = 35.2


DC Dredge Plume with Silt Curtain

8 10 12 14 16
Time of Day

18 20 22 24

Figure 19. Reduction in net community production as a
result of dredging with a silt curtain. Solar
radiation is shown at the top of the figure.
Abbreviations used are explained in the text.


3 0.5





1000 Solaor


6 8 10 12 14 16 18 20 22 24

% RED= 10.0
% DTR = 1.3






S o-


s -,,0-




6 8 10 12 14 16 18 20 22 24
Time of Day
Figure 20. Reduction in net community production as a
result of dredging with a silt curtain. Solar
radiation is shown at the top of the figure.
Abbreviations used are explained in the text.


DC Dredge Plume with Silt Curtain

18 20 22 24


DC % RED = 28.3



r o.



-1.5 DC Dredge Plume with Silt Curtain



6 8 10 12 14 16 18 20 22 24
Time of Day
Figure 21. Reduction in net community production as a
result of dredging with a silt curtain. Solar
radiation is shown at the top of the figure.
Abbreviations used are explained in the text.

6 8 10 12

o Solar

6 8 10 12 14 16 18 20 22 2'


%RED= 98.0
% DTR 43.1




r 1.0


- 0.5




8 10 12 14 16 18 20
Time of Day

22 24

Figure 22. Reduction in net community production as a
result of dredging without a silt curtain.
Solar radiation is shown at the top of the
figure. Abbreviations used are explained in
the text.


DNC Dredge Plume with no Silt Curtain



6 8 b 12 14 16

Time of Day

18 20 22 24

Figure 23. Reduction in net community production as a
result of dredging without a silt curtain. Solar
radiation is shown at the top of the figure.
Abbreviations used are explained in the text.




S 1.0,







6 8 10 12 14 16 18 20 22 24

% RED = 42.6
% DTR = 5.8


DNC- Dredge Plume with no Silt Curtain

1000- Solar
N Radiation

0 -


~ 1.0






10 12 14

16 18 20 22 24

% RED = 23.6


TP- Tug Plume

6 8 10 12 14

16 18 20 22 24

Time of Day
Figure 24. Reduction in net community production as a
result of a tug plume. Solar radiation is shown
at the top of the figure. Abbreviations used
are explained in the text.

from the passage of tug boats moving work barges through the

shallow areas. This activity resulted in a range of reduction

9.7-48.0% during the time interval in which they occurred

and an overall daily reduction of 1.9-17.2%. A drop in sun-

light also occurred during this time interval.

A summary ranking all turbidity events monitored is

presented in Table 2. The three events in both categories

(time interval and daily metabolism reduction) resulting in

the greatest reduction in net community metabolism are

naturally occurring weather events: a severe pre-noon

thunderstorm, hurricane-related thunderstorms, and a severe

late thunderstorm, respectively. Although some events

changed in order between the two time comparisons (time of

occurrence and reduction in the total metabolism), dredging

events generally reduce metabolism less than thunderstorms.

Less intense weather events and dredging events (tug plumes)

rank lowest of the categories examined.

In the comparison involving the time period in which

the event occurred, the high variability of the measurements

associated with both dredging types (with and without a

curtain) indicates the erratic nature of this activity.

Distant thunderstorms also show high variability in their

reduction of metabolism. Reduction of daily net production

remained highly variable for only one dredging event (no

curtain) but variability of the metabolism reduction was

decreased substantially when a silt curtain was in place.

I .-I
^ 3 !

0 '0 00 a,

0 n>, -a I,-
e c se

00 a, 0. -I
989 e
1%s >

I a, _a C ,


a, a, a, a, a, a, a,

,1 -

I a, r

0 >>

0;-u> 0E* 0 a,

N a,- S a,>.-S
V : > f
K~ u1 lO3 r


Sa, a, a, a, a

I a, a, an a


aS o

Hurricane-related thunderstorm effects were also highly

variable throughout the day as were tug plume-associated


A comparison of net photosynthetic efficiencies prior

to the dredging events and two years later is shown in

Table 3.

The average depths of all the stations are similar,

ranging from 0.9 to 1.4 m. Mean percent light transmission

of the water column was approximately 40% higher in the post-

dredging period, 50.6 as compared to 36.0 during the summer

of dredging. Higher variability of the post-dredging water

clarity was due to the presence of strong sustained winds

during the early stages of the sampling effort. Surface

solar insolation also averaged slightly higher in the post-

dredging sampling period and was also less variable. The

amount of solar insolation reaching the plants was also ap-

proximately 2 times higher in the post-dredging period, due

mainly to the increased water clarity shown during this

time. Mean dry weight and ash-free biomass estimates for

the two periods were remarkably similar although some

variability was noted from two separate samplings within the

pre-dredging interval. Net community production estimates

were approximately 2 times lower in the post-dredging


Comparison of net production rates per gram of both dry

weight and ash-free weight of standing stock were both

* o0I


F 0 0 00 0 0

000 00 00~a 000000000 0
00 o.O -00 0 0 0;0 00020


i" ':"



" 8

approximately 2 times higher during the sampling period un-

affected by dredging. Final efficiency values (solar inso-

lation included) were approximately 4 times greater in the

communities unaffected by turbidity due to dredging.


The rapid response of autotrophic communities in clear-

water environments to changes in sunlight has been shown in

previous work (Knight, 1983). The similar response shown by

the marine seagrass communities in this study indicates that

they are also very highly responsive to changes in their

environment. Permanent changes in water quality (including

light) would be expected to result in alteration of the

primary producer communities. Turbidity impacts associated

with human activities, such as dredging, would be expected

to be most severe on those primary producers less tolerant

of lower light conditions or higher siltation rates.

Several authors have shown loss of seagrass habitat and

degradation of benthic communities resulting from increased

turbidity, nutrients, toxic pollutants or storms (Grady

1981; Birch and Birch 1984; Bulthius et al. 1984).

A commonly used approach in environmental studies is to

monitor changes in a parameter of interest over time as a

result of some perturbation. Seldom are attempts made to

place observed changes in perspective by comparing these

with naturally occurring events. However, when making com-

parisons of this type, caution must be exercised to ensure

that the natural and man-made perturbations contain similar

components. In the present study, the weather events used

for comparison are believed to be similar to dredging events

in that light blockage occurs in each case.

In the present study of short-term effects, both the

event intensity and timing appear to be the major factors

influencing the ranking of all the perturbation events. The

highest ranking activities, all weather events, achieved

this level because they had a high intensity level of light

blockage and were of sufficient duration to eliminate most

of the net primary productivity for that day. Occurrence of

the severe thunderstorm early in the daylight period caused

elimination of much of the net photosynthesis during this

time of rapid photosynthetic increase.

The dredging activities, both with and without silt

curtains, were intermediate in effects compared to natural

events. Reduction of net productivity resulting from these

two dredging events was approximately similar during the

time in which they occurred and in their effect on the daily

production level. However, the variability of the data re-

sulting from dredging with no curtain was much higher compared

to the value determined during dredging with silt curtains

in place. This could indicate that in this high current

area the presence of a silt curtain does cause a more even

dispersion of the dredge plume. The plume emanating from

the dredge within the curtain may tend to lose some of the

larger particles and at the same time be more evenly dis-

persed in the water column from turbulent mixing as it flows

under the curtain.

In general, the major short-term effects of dredging

(both with and without a silt curtain) on seagrass net

community production fall somewhere in the middle of the

spectrum of meteorological events that the system encounters

normally. This agrees with the findings of Tramontano and

Bohlen (1984) in which effects of a dredging operation in an

estuary were short-lived and less severe than those due to

storms. One of the reasons the dredging events in this

study achieved their ranking is that in practice the

dredging operations were intermittent because of tidally

dictated work schedules, equipment malfunctions and

transportation of sediment to shore.

Recommendations for future dredging in this area should

consider maintenance of this intermittent nature of the

operation. Development of a dredging scheme which would

provide continuous dredging would very likely elevate

dredging activities to the top of the present list because

continuous dredging throughout all periods of the day would

increase the intensity of dredging operations and eliminate

most, if not all, of the net primary production.

Net productivity measured during nighttime hours is a

phenomenon not uncommon in oxygen-based productivity

calculations (Caldwell and Odum 1980). The net production

of oxygen measured at this time can be explained by several

sources of error inherent in a microcosm system of this

nature. Nighttime net production may be a result of release

of oxygen from the sediments as a result of translocation of

oxygen from the leaves to the plant rhizomes during peak

production periods of the day. Zieman (1975) has shown that

this can happen on a somewhat unpredictable basis and is not

well correlated with any particular peak production level.

Another explanation could be that either highly oxy-

genated patches of water or water of oxygen content similar

to that which entered the upstream end may be entering the

downstream end of the tunnel. This could occur from

turbulent flow as a result of the overlying water column

passing along the top and sides of the tunnel and then

swirling in the tunnel. Although the sampling hose was

assumed to be inserted into the end of the tunnel far enough

to avoid this problem, it is possible a small amount of this

water from outside the tunnel could enter and be picked up

by the downstream hose. If this happened, the water from

the downstream could be higher in oxygen content than the

upstream end, therefore falsely indicating photosynthetic

activity. The magnitude of such an error is small if one

assumes it is similar to the magnitude of the 2,200 hours

occurrence in Figure 6.

Seagrasses such as Thalassia testudinum have lacunae in

the leaves where gases are stored during peak photosynthesis

periods (Phillips and McRoy 1980). It is possible that stored

oxygen is released in a burst from the leaves later in the

day, elevating the oxygen readings to give the appearance

that photosynthesis had taken place.

Finally, the residence time of water in the tunnel was

determined crudely with dye and represents an average flow

time during a tidal cycle. Since tidal flow is a dynamic

process which would continuously alter the residence time of

water in the tunnel, sampling of the upstream and downstream

portions of each slug of water as it passes through is more

than likely not completely synchronized. This probably is

not critical when large differences occur during the day,

but may become a factor when respiration is fairly balanced

during the nighttime hours.

Thus, it appears the microcosm method of upstream/down-

stream oxygen measurement utilized for calculation of net

community metabolism is not designed for detection of minute

changes in oxygen production or consumption. The system is

designed to be a simple field monitoring device able to detect

major changes in seagrass community function.

Criticisms of the oxygen method of seagrass community

metabolism measurement (Bittaker and Iverson 1976; Phillips

and McRoy 1980) regarding oxygen storage within the plant

itself and problems with advective diffusion were not con-

sidered appropriate for the present study. The absolute

production of the seagrass itself was not critical; the

parameter of interest was the response of the entire com-

munity to natural and man-made perturbations, regardless of

potential internal gas storage or recycling. Advective

diffusion was eliminated through the use of covered

microcosms. The measurement of oxygen flux provided

a rapid response to fluctuations in the light regime and

served as an index of ecosystem function. Kemp and Boynton

(1980) have shown that community metabolism measurements

utilizing oxygen techniques where horizontal oxygen

diffusion is controlled are a useful index of ecosystem

function and therefore can be used to compare natural and

perturbed conditions. Additionally, all of the methods of

currently used seagrass productivity measurement suffer from

some disadvantages peculiar to the method which prevented

their use in this study. Standing stock measurements, leaf

area index, and leaf growth measurement all require long

periods of time for obtaining results and are not readily

adaptable to the measurement of short-term phenomena.

Tagged carbon methods suffer from the same criticism as

oxygen methods in that it is likely that recycling of carbon

products does occur (Bittaker and Iverson 1976).

The pre- and post-dredging measurements of photosynthetic

efficiency provide a different perspective on possible effects

relating to dredging in the area. Net community production,

when adjusted for pre- and post-dredging biomass, water

clarity and solar differences, has been substantially reduced.

Sampling during the post-dredging interval was accomplished

during a period of higher solar radiation (low number of

storms). Further, water clarity was also higher in the post-

dredging period. These two parameters, when coupled together,

show that more light was reaching the plants in the period

of measurement following dredging operations. Both dry weight

and ash free biomass estimates also are very similar, so

decreased biomass can not account for lowered net productivity.

There are several plausible explanations for lowered

post-dredging photosynthetic efficiency. Epiphytic organisms

are perhaps more susceptible to dredging effects such as

elevated turbidity or high siltation which may result in

excessive abrasion, clogging or covering. Since this

component can contribute a substantial amount to the net

community metabolism while being a small component of the

standing stock biomass (Pomeroy 1974), this portion of the

community could have been depressed severely and not be

indicated by a drastic change in biomass estimates. This

segment of the community was not separately measured, however.

On the other hand, the impairment of the photosynthetic

ability of the community may have occurred uniformly within

each community component but at a lower level and not to the

extent of reducing standing stocks. In either case, the net

community production is lower. Since no measurements were

collected in either the interim period or over a longer

term, it is difficult to determine if the community is

gradually rebounding from an impacted state or whether it is

on a gradual decline. It is apparent, however, that water

clarity has returned to pre-dredging levels. Only

additional measurements over a longer time period will

document the long-term adjustment of the seagrass community.


1. Reduction in net community production due to elevated

turbidity from natural and anthropogenic events was

compared by estimating the reduction during the time

period in which the event occurred and also by comparing

this reduction to the daily production. The greatest

reduction was shown by severe summer and hurricane-

related thunderstorms. Dredging activities in which no

silt curtain was used were next in the ranking followed

by distant thunderstorms and dredging with a silt

curtain, respectively. Overcast conditions, tug

plumes, normal thunderstorms and scattered clouds,

respectively, completed the ranking. Severe and

hurricane-related thunderstorms also ranked highest in

the percentage reduction of the daily production,

followed in order by dredge plumes with a curtain,

overcast conditions, dredging with no curtain,

scattered clouds, normal thunderstorms, tug plumes, and

distant thunderstorms, respectively.

2. Comparison of pre- and post-dredging net community

production normalized for solar radiation and standing

stock biomass showed an approximate 4-fold decrease two

years after dredging. Severe impairment of photosynthesis

of the epiphytic community or low-level impairment of

all community components were offered as explanations

for this finding.

3. Occasional measurement of nighttime production of

oxygen in the microcosm was attributed to inaccuracies

inherent in the microcosm technique. It was suggested


that the method may be most useful for detection of

gross changes in highly productive systems rather than

small differences when production is minimal.



Two of the major components of a dredge plume, shading

and elevated nutrients from resuspended sediments, have been

the subject of numerous investigations which have dealt with

these phenomena on an individual basis and not as interacting

entities. A review of the pertinent literature pertaining

to effects of dredging-associated turbidity and shading on

marine primary production is contained in Chapter II. Relevant

previous research includes investigation in four areas: ni-

trogen sources, requirements and limitations for macrophytes

(Patriquin 1972; McRoy et al. 1973; Capone and Taylor 1977;

Stirling and Wormald 1977; Khalid et al. 1978; Capone et

al. 1979; Barko and Smart 1981; Capone 1982; Strom and Biggs

1982; Short 1983), forms of nitrogen preferred by seagrasses

(Fisher et al. 1982; lizumi et al. 1982; Dortch and Conway

1984), phosphorus requirements of macrophytes (McRoy and

Barsdate 1970; Carnigan and Kalf 1982), and enrichment studies

using various forms of nitrogen and phosphorus (Orth 1977;

Parker 1982; Thursby 1984; Dawes et al. 1984; Roberts et al.

1984; Ulrich and Barton 1985).

The present study investigates the short-term effects

on net seagrass community production resulting from two facets

of a dredging event: 1) shading by the dredge plume

and 2) addition of nutrients into the water column as a result

of nutrient release from resuspended sediments. Early field

observations of a very rapid community response to shading

suggested the hypothesis that these communities could be

responsive to other changes such as addition of nutrients

and also might be able to absorb and quickly utilize these

nutrients from the water column when they are available.

Hence, higher nutrient levels might cause an immediate

increase in community metabolism. Further, this metabolic

enhancement could perhaps offset reductions in metabolism

due to shading.

The objectives of this study were 1) to establish the

effects of shading and nutrient addition on net seagrass

community metabolism separately and in combination, 2) to

determine the relative importance of each factor by measure-

ment of the levels of response of net community metabolism,

3) to help determine whether the two factors cause an overall

reduction or increase in net community production, and 4) if

both increases and decreases do occur, determine the over-all

direction of influence of a dredge plume. The objectives

were investigated in the field using flow-through microcosms

to determine net community metabolism from upstream/downstream

oxygen measurements.

Materials and Methods

Seagrass community metabolism was determined using the

upstream/downstream oxygen method of Odum and Hoskins (1958).

General methods, including equipment, oxygen data corrections,

and calculation were identical to those described in the

previous chapter. Variations of the general scheme will be

included as appropriate.

Field experimentation involved the application of two

treatments, namely shading and nutrients, to the microcosm

at various levels and in as many different combinations as

possible over several diurnal periods.

Four microcosms, positioned in parallel, each approxi-

mately 1-meter wide and 4-meters long, were used for the

experimental field apparatus. Three of these tunnels were

used exclusively for nutrient addition (ammonia, nitrate,

and phosphorus) and one tunnel remained as a measure of base-

line conditions. The same tunnel was used for each nutrient

throughout the experiments. The microcosms were covered and

anchored as described in Chapter II.

Baffle plates at each tunnel end were used to control

water movement through the tunnel so that sufficient resi-

dence time existed to measure oxygen change within the

detection limits of the oxygen sensor. Rubber hoses were

inserted into each end of each tunnel and connected to a

valve system on the boat. By opening and closing the

valves, water was selectively pumped from the desired tunnel

end to a Hydrolab Model 8000 (Hydrolab Corporation, Austin,

Texas) for determination of dissolved oxygen, temperature,

and conductivity in the same manner as described in

Chapter II.

The shading portion of the experiment was carried out

by shading the tunnels with nursery shade cloth having a

mesh size approximating 60% sun blockage (40% light passage).

Three different nutrients were added: nitrate, ammonia, and

phosphorus. The levels of nutrients added consisted of four

concentrations designed to bracket the concentration of nu-

trients actually measured in a dredge plume in the field

(Table 4). Ammonia and phosphorus were determined

colorimetrically using the procedures described in Stickland

and Parsons (1972). The four concentrations consisted of

1 x 10-5 mol 1-1, 1 x 10-7 vmol 1-1, 1 x 109 umol -1, and

1 x 101 umol 1 A concentrated solution was added to

the tunnel to achieve the final concentration based on the

assumptions of a water residence time of 20 minutes and a

uniform mixing of water throughout the tunnel.

Shading of each microcosm was accomplished by attaching

the nursery shade cloth to one end of the tunnel and

unrolling it to the other end and tying down the sides.

When not in use, the shade cloth was rolled up and secured

to one end of the microcosm.

Nutrients were added with a gravity feed system that

consisted of 5-gallon carboys lashed to the top of a truck

innertube that floated at the surface. The jugs were

connected to the tunnels with tygon tubing that was attached

to the upstream end of the tunnel through a horizontal

diffuser in the baffle plate. This diffuser, which consisted

of four polyvinylchloride (PVC) nipples, allowed dispersal

o 0 0

CQ r-l


o o a

H i~- l
to r

to t
I I o

C 0 to1r-





3- 0
.g <3 3 23

C) oF $4
M 03 r-i
C! o
t D=

of the nutrients in the tunnel as evenly as possible. Nu-

trient flow from the jugs was controlled by in-line PVC ball

valves. The flow required to release the contents of the

jug into the tunnel to achieve the desired final concentration

was determined through field experiments with dye release.

Residence time of the water in each microcosm was

determined through several timed dye studies to be an average

of approximately 20 minutes. This residence time was used

during the entire tidal cycle. Initially, the water in each

tunnel was sampled for oxygen, conductivity, and temperature

at 5-minute intervals beginning with the upstream end. At

the end of 20 minutes, the downstream end was sampled which

allowed sampling of the plug after the 20-minute residence

time. With the number of tunnels being nearly simultaneously

monitored, this allowed only one measurement per tunnel per

hour. Sampling effort was therefore increased to a frequency

of 1 per minute. Each separate upstream end was sampled at

1-minute intervals for 4 minutes. Sampling was suspended

for 1 minute out of every 5 to allow readjustment of

sampling so they occurred at the same minute of every hour.

At the beginning of minute 20, the valves were switched so

that the downstream ends of all four tunnels were sampled in

a similar sequence beginning with Tunnel 1. This enabled

the downstream sampling to be timed to coincide with the

20-minute residence time of water. This 1-minute interval

sampling continued during the entire experiment over an

approximate 14- to 18-hour day.

The experimental treatments were applied over 4-hour

time blocks throughout the day. These time blocks covered

the time period from 7 to 10 A.M., 10 A.M. to 1 P.M., 1 to

3 P.M., and 3 to 7 P.M. The combined treatment of nutrients

and shade were applied to the experimental chambers in as

many combinations as possible over each 4-hour test period

throughout the experiment. At the end of 2 weeks, the tunnels

were moved and the experiment was replicated.

Data were analyzed using the general linear model (GLM)

procedure of the Statistical Analysis System (SAS) at the

Northeast Regional Data Center (NERDC) at the University of

Florida. This variation of the ANOVA procedure was utilized

because of the unbalanced design of the experiment. The

dependent variable calculated in this set of experiments was

the rate of change of oxygen (community metabolism)

calculated from the difference equation (see Chapter II)

utilizing the upstream and downstream oxygen values from the

microcosms. The independent variables, or treatments were

1) shading, with two levels of shading on and shading off;

2) nutrient addition with three types of nutrients corres-

ponding to nitrate, ammonia, and phosphorus, each with

four levels of concentration; and 3) station differences

with two levels corresponding to each of the replicate

sampling episodes.

The procedure for analysis of the ANOVA results

consists of a step-wise examination of significance levels

(a=.05) of F values resulting from the least square

determination of the general linear models procedure. The

first step involves determination of the significance level

(a=.05) of the interaction term which contains all of the

main effects, in this case shading, the type of nutrient

added, and the concentration of the nutrient. If the

interaction term is not significant, then all two-way

interactions of the main effects are evaluated. If none of

these interactions are significant, then each main effect is

inspected separately. In this manner, significance in the

model is attributed to the main effects acting either singly

or in combination.

The ANOVA compared only the tunnels receiving nutrients

and shading. This allowed determination of differences

between nutrient type, concentration, and shading effects.

A comparison was also made between these microcosms and a

control. This allowed a qualitative assessment of nutrient

and shading effects compared to background.


Comparison of treatment effects between the experimental

tunnels showed that the 3-way interaction term was not signi-

ficant (Table 5). Examination of all the 2-way interaction

terms showed significant differences (p=.003) only in the

shade-nutrient concentration term (Table 5) indicating a

significant interaction between shading and the concentration

of the nutrients added. Evaluation of the nutrient main

4I b

S0 0 0 0 Q)

So co o o o
o o o o o

(U -i D to n ito
-' .- .N o o

0 o o o o ^

4 n m m00
4- (N 00 0 > '


0 (0 o -4 l
1 0 ) 0 0 0
4 4- 4-I 4+ 0 14

00U 0 0U-O c 0
-H H 4 N 4 N -H 4)
4 r r o

44 *d I' H 0 -) 0
40 4 4 -' 40 -H 40
04 4 0 04- 0d 0 14 0 > 4o
1 O I0 a) a1 0 a

00 4 H14 H 14J U)
00 4J 4 41 4-1 04 0 4 0
4(0 0 0 0 0 0H 0o

Fa u Zu 4 (- 0a Q 1
ri c u 4O r 41 (a :$ a) 11
0 4 3 () 0 =! cI C a a
"E C w Z U Z m H M0 4 U

effect term showed no significance, indicating there was no

significant difference in the response variable due to the

type of nutrient being added so this term was dropped from

further analysis.

The results of combining all nutrient concentrations

and comparing the net production with shade on and off are

shown in Table 6. In the unshaded condition, the effects

due to the addition of different concentrations of nutrients

were highly significant (p=.0001), i.e., different concentra-

tions of nutrients caused a different response. No sig-

nificance was noted with the shaded tunnels.

The effects of each nutrient concentration under shaded

and unshaded conditions are shown in Table 7. The nutrient

concentration in the unshaded microcosms was significant at

all levels except the lowest concentration

(1 x 10-0 mol 1 ). The nutrient concentration of

1 x 10 umol 1- showed the largest difference (0.416)

between shading treatments. This was followed in order by

each of the increasing concentrations.

A qualitative comparison of mean net production between

an unshaded control and communities exposed to the three

nutrient types under shaded and unshaded conditions is

contained in Table 8. Net production of the unshaded

seagrass was higher than the control for all three nutrient

types. In the shaded case, the experimental microcosms were

lower than the control although the seagrass subjected to

nitrate addition showed considerably higher production than


Table 6. ANOVA results of combined nutrient concentrations
under shaded and unshaded conditions.

Treatment Na DF F Value

Unshaded 144 3 12.84

Shaded 119 3 0.54

Probability Significance

0.0001 Yes

0.6608 No

number of observations

degrees of freedom

ca= 0.05




^ 0 i-l10 0


ui o o o o

5 "^

E Q'

i r13

% U
~I ~I


C4 i

I .1



C '
0 4-
d C

1 c 1





YI ~i -m


the other two shaded microcosms.

To test for differences between plots, the microcosms

were moved and the experiment replicated. Table 9 shows no

significant differences existed between stations.


The metabolic response of these clear-water seagrass

communities to changes in their physical environment is very

rapid, sometimes on the order of minutes as determined from

field observations described in Chapter II. Because of this

rapid response time, the present experiments were designed

to examine the response of the community to two components

(shading and nutrient addition) of a dredge plume.

The response of the community to shading was not un-

expected as effects of light blockage on net community

production has been demonstrated earlier (see Chapter II).

Clearly, of the effects investigated, shading causes the

most drastic alteration in net community production.

The response of seagrass communities to nutrient inputs

is much less clear. Seagrass nutrient dynamics are poorly

understood and the results are often contradictory. In the

analysis of the effects of nutrient addition in the present

investigation, similar responses of the community to all

three nutrient types are not unique. Harlin and Thorne-Miller

(1981) showed that both nitrogen and phosphorus stimulated

growth of Zostera marina in the field, suggesting an almost

simultaneous deficiency in both nutrients. Thursby (1984)

showed that Ruppia in the summer contained low tissue


Table 9. ANOVA results of comparison between stations.

Treatment Na DF F Value Probabilityc Significance


263 1 3.38


number of observations

degrees of freedom

Probability a= 0.05

concentrations of both nitrogen and phosphorus, indicating

this species was deficient in both these nutrients. Ulrich

and Barton (1985) found that Phraqmites australis shoot growth

was influenced by both nitrate and phosphorus application.

Roberts et al. (1984) showed that, under conditions of nu-

trient enrichment by nitrogen, additional phosphorus may be

needed for maximum growth. These highly productive tropical

systems may have evolved very efficient relationships with

nitrogen fixing heterotrophs (Patriquin and Knowles 1972)

such that nitrogen is generally in excess. Similar nitrogen

abundance could explain the response of this system to added


McRoy and Barsdate (1970) showed that phosphorus uptake

is a function of the metabolic activity of the plant. Periods

of high productivity such as occur in seagrass communities

may momentarily deplete phosphorus stocks. Direct application

of phosphorus to the community could push productivity to

higher levels.

Interaction effects similar to those found in this

study between shading and nutrient concentrations have been

documented by Dawes et al. (1984) who showed that for an

alga in nutrient-rich water, photosynthetic rates were

significantly different when it was exposed to light at two

different intensities. McRoy and Barsdate (1970) showed

that light was required for uptake of phosphorus, implying

that phosphorus is utilized only under adequate light


The highest response in net community production was

found in the unshaded microcosm at concentration 1 x 10-9

Pmol 1-, indicating this may be an optimum nutrient level

for seagrass production. The negative production values

recorded for the highest nutrient concentration with shading

may indicate that photosynthetic inhibition is occurring at

the higher concentration, although this concentration was

much lower than those used by Harlin and Thorne-Miller

(1981) which did not produce inhibition. The absence of a

significant difference at the lowest concentration indicates

that in the sunlight the concentration is too low to enhance


Comparison of the experimental tunnels with the control

shows some production enhancement occurs in the sunlight

when nutrients are added as compared to a baseline. This is

consistent with work by others (Harlin and Thorne-Miller

1981; Thursby 1984; Ulrich and Barton 1985; and Roberts et

al. 1984) in that several nutrients, for various reasons,

can be nearly simultaneously limiting. It is also possible,

that with nutrient additions higher than those experienced

during a dredging event, productivity may be increased such

that a significant difference could occur. Effects of high

nutrient loadings are fairly well documented for algae (Tewari

1972; Waite and Mitchell 1972; Guist and Humm 1976; and Ho

1979) showing that excessive growth occurs under these con-

ditions. However, in the lower level nutrient additions

related to dredging events, the trend is toward enhancement

of production only in the sunlight. Under low light con-

ditions, the addition of nitrate caused the highest

community metabolism compared to other nutrients. This is

consistent with the findings of lizumi et al. (1982) who

showed that although ammonia is the most easily utilized

form of nitrogen because of ammonia toxicity, nitrate is

taken up more readily.

The most significant effect caused by a dredge plume is

the reduction in net production due to shading. Although

some enhancement is indicated resulting from increased

nutrient availability from resuspended sediments, this is

overshadowed by shading effects. If a mechanism existed

which enabled the plume to dissipate quickly and allow

nutrients to remain available, the present study indicates

production would be increased. This is highly unlikely,

however, in the study area since rapid plume dispersal

mechanisms, such as high currents, would also quickly

dissipate any elevated nutrients.


1. Shading effects were the major impact of a dredge

plume, causing a reduction in seagrass net community


2. Response of the community to shading and addition of

nutrients by a dredge plume was a result of interaction

between shading and the concentration of nutrients

added. Net production was higher in the sunlight in

the presence of nutrients with an optimum production at


-9 -1
a concentration of 1 x 10 pmol 1-. There was no

significant difference in net production between the

three types of nutrients (nitrate, ammonia, and

phosphorus) utilized.

3. Comparison of experimental and control communities'

response to nutrient addition and shading showed

enhancement of production only in unshaded microcosms

and reduction in net production due to shading.



Although seagrass ecosystems are among the most

productive in the world (Phillips and McRoy 1980), few

published accounts exist which examine with a simulation

model their primary and secondary production, nutrient

dynamics, decomposition, and response to outside influences.

Thayer et al. (1975) constructed an energy flow diagram of a

Zostera marina community and showed that organic matter was

exported although they neglected primary production of the

epiphyte and benthic diatom communities. Verhagen and

Nienhuis (1983) modeled the production, seasonal biomass

changes, and distribution of the eelgrass Zostera marina.

Seasonal oscillations observed in eelgrass production were

believed to be a function of light, water temperature,

currents, and the age of eelgrass blades. According to

their model, the height of the eelgrass shoots could be

explained by space limitation, lack of below ground biomass,

and insufficient light. They indicated that the combination

of vertical, horizontal, and year-to-year variation of the

community could be best explained by a parameter they had

not used in earlier versions of the model: production of

seeds and growth of the eelgrass shoots from seeds as

opposed to the more common sprouting from rhizomes. Short


(1980) developed a simulation model of a seagrass production

system and examined the effects of temperature, light

limitation, tidal current interactions, respiration, and

mechanical damage on seagrass production. In general, he

found good correlation between the simulation and empirical

data with regard to standing crop, blade length, rhizome

length, and production.

In the present study, a deterministic model was used

which showed explicit feedback loops coupling the various

system components. This type model was believed adequate

to examine cause and effect relationships in a natural

seagrass ecosystem and to compare their response to natural

and man-made perturbations. This systems dynamics model

(Forrester 1968) showed feedbacks explicitly, portrayed

information as well as material flows, and included linear

and nonlinear relationships. The microcomputer program

utilized to run the model was Simu-Dyn, a DYNAMO simulation

program developed by Dr. Clay L. Montague.

The objective of this model was to examine five seagrass

community components: standing stock biomass of the seagrass

community (including epiphytes), detritus, nutrients, (water

column and sediment), and consumers. The response of these

components to outside influence was evaluated according to

the following: 1) system response under a standard model run

environment, which is a run with all constants set to a

pre-determined, documented level, 2) effects of systematic

alteration of model constants, 3) response to natural

perturbations such as thunderstorms and a hurricane and

4) changes in output from dredging events.

Model output from the standard run provides a baseline

for comparison with each run of the sensitivity analysis.

The sensitivity analysis (systematic alteration of constants)

indicates those model constants responsible for the greatest

change in the model output, i.e., those constants to which

the model is most sensitive to change. This analysis can

also point out specific model relationships requiring further

work (Overton 1977). Model response to perturbations

provides insite into the effects of elevated turbidity due

to thunderstorms, hurricanes, and dredging perturbations.

Finally, model output can be used to estimate recovery time

of parameters of interest following perturbations.

Materials and Methods

The modeling procedure can be subdivided into eight

component processes. These are: 1) development of an

influence diagram, 2) construction of a model diagram with

all appropriate equations, 3) production of a standard model

run, 4) testing of model parameters through sensitivity

analyses, 5) observation and interpretation of model

response to perturbations, 6) field validation, 7) implemen-

tation in the real world, and 8) evaluation of the

implementation while continuing to revise the model.

Information contained in the present model analysis includes

those processes through Step 5.

Influence Diagram

Closed cycles of influence between model components are

shown in Figure 25. Arrows indicate cause and effect (i.e.,

influence): the influencing parameter is indicated at the

tail of the arrow while the parameter that is influenced is

shown at the arrow head. Positive and negative signs at

each arrow head indicate the direction of influence, i.e., a

positive sign indicates that as the influencing parameter

changes, the affected parameter also changes in the same di-

rection. Conversely, a negative sign indicates an influence

in the opposite direction, for example, a decrease in one

parameter will cause an increase in another. Positive and

negative signs indicated within the small dashed arrow

inside a loop show the overall sign of the particular loop

of interest. Positive loops reinforce changes, negative

loops oppose change and yield stability. Parameters shown

outside the closed cycle of influence are those which affect

loop components but are not a part of the closed loop

system. These parameters are added as they contain

perturbations and energy sources of interest as well as

influencing parameters within the closed cycle.

Five loops are shown in the model closed cycle of

influence. These loops vary in size and consist of influences

between the following components. The loops are numbered to

correspond with Figure 25: 1) water column nutrients-seagrass,

2) sediment nutrients-seagrass, 3) consumers-detritus,

4) sediment nutrients-consumers-detritus-seagrass, and






r o


S/ u
-, m


0 1
u (D

Li l

O ^''

/ *r4

5) seagrass-detritus-consumers-water column nutrients. Loop 1

shows a negative influence between seagrass and water column

nutrients. As a seagrass community grows, the available

water column nutrients decrease through uptake by the plants

during photosynthesis (Patriquin 1972). In addition, the

model also contains an equation structure that allows some

leakage of nutrients back into the water column (McRoy and

Barsdate 1970) either through passive plant release or active

pumping. The same arrangement of nutrient uptake and

release is also suggested for Loop 2. Loop 3 is a negative

influence as the availability of detritus and consumers

tends to keep the other in check. As the food source

(detritus) increases, the consumer population also begins to

grow due to the higher food availability. However, with

growth of the consumer populations, demand for food

increases and the food stocks are consumed. As food stocks

dwindle, the growth rate of the consumers slows down. This

relationship, as opposed to Loops 1 and 2, can cause

oscillations in both populations which resemble the classical

predator-prey oscillations shown by Lotka and Volterra (from

Pianka 1978). Because this loop is short (two accumulations

only) information delays between these components are minimal

and response time is short.

Delays in Loop 4 are longer. Loop 4 is also positive,

meaning the changes within the loop components tends to

reinforce the original loop direction. Increases in

sediment nutrients can increase seagrass biomass (Patriquin

1972; Penhale and Smith 1977) which causes more detritus to

be formed due to the death of more seagrass. More food

availability stimulates growth of the consumer population

which causes more nutrients to be released into the sediments

(Mann 1982).

Loop 5 is also a positive loop. This loop structure is

very similar to Loop 4 with the substitution of water column

nutrients for the sediment nutrient accumulation. Both

Loops 4 and 5, because they contain more components, could

show a longer response time to changes within the loop,

especially if one of the components were limited as to its

magnitude of response during a given cycle. External

influences which are shown outside the closed cycles of

influence consist of perturbations which affect nutrients

and water clarity. Dredging influences water clarity by

increasing turbidity as dredging intensity increases

(McCarthy et al. 1974). With increasing turbidity, available

light decreases which slows the growth of the seagrass com-

munity. Dredging also increases the water column nutrients

due to resuspension of sediment nutrients from the dredge

plume (Tramontano and Bohlen 1984). No increase in sediment

nutrients from the dredging process was included because it

was assumed nutrients released to the water column were re-

moved rapidly, leaving only negligible amounts to settle

out. Loss from the sediment nutrient pool during dredging

was also not considered because only communities in close

proximity to the dredge were evaluated. Weather

disturbances such as thunderstorms affected a larger area

and impacted both water column and sediment nutrients. During

storm events, sediment nutrients were stirred up from the

bottom into the water column, increasing the nutrient levels.

Turbidity increases due to storm activities caused a reduction

of available light.

Model Flow Diagram

The flow diagram for the model is presented in Figure 26.

Material flows are represented by solid lines and information

flows by dotted lines. Constants are indicated by the

symbol 1, numbered, described, and referenced in Table 10.

Auxiliaries, which represent convenient additional calculation

steps, are shown as larger open circles. The rectangular

boxes indicate levels or accumulations of material. Arrows

going into and out of each box from left to right represent

the inflow and outflow to the accumulation, respectively.

"Valves" shown on each arrow represent controllers of rates

of inflow and outflow; parameters which affect an accumulation

impinge on a rate of inflow or outflow and are shown with

dashed lines drawn to the rate controllers. In this manner,

all feedback loops and influences, constants, auxiliaries,

material flows, information flows, and interactions are in-


Model Rationale

The rationale for the model is explained on the basis

of the relationship of each accumulation to its component

parts. The units of all constants are shown in Table 10.



- -



S Om

a ,



S I a ^z
E/ / as m
Sz / /


-5- _7 0 1

22I ^ w

% a e
~~L~ E B

"f "
uU ~~~

o I
0 0

1 1 um
I I .
uO s o *

i- l E- 'O B 3






= i:



c N 3 ^ 13. c
4 0 i 0 01 0 o l
H 0 .4 9 0 H 0 OH>
3 H 0. 4 0 H 0 H
H 0 H 000 000 +0+ - 00
H 00 H'a 4 HH a 0.0
0 O 4 0 0 4.' 4 4 0 4 '0

i-lr r -1
I. 0 ..l 000. >H. 0. 0

4. 0. 00 0 .0 00... C .

i I S^ I 1 8l 1. i1i
a s, s "'s 3~ 8 s
a ' E ?^
0 0 0 0 4 P i e
s ~ ~ a i J ^- .-
0< 00.0 0 0 0^ 00" 0~r 0h 0'
0 ^ 0- .- 0 0 0 0 0
0 0$ 000 .. 0 0.0 0

0 I 0 0 0 0 0 0 0

i I
* ' s
s 0 V' 0 00 '0 0 0.0 0 0 1
~ 5 & > .i0 00 10 I 0 '0 s '0^
0 0000 00. 00 00 d 00 0 0.. 0

8 0 '0
I 0 0 |

0| .0 0 0 0 0 0 "
0 0 1 0 '0

~ g S| | E
8 I
19 o

0 ? 0

00. 0 0 0 00~

0 6 0 n

I g. g

0n 0 -i 0ui
f in 3o in*o 4

T? 3 3 'a

3 Bo
F : C

U 7 Q ^
*" >^

Ri a


21 I

4J a
G e

5 s
8" "
C: 1


0. 3

*Ig-a S
Is ssg


o r .

01 ^

W >,

i |

E o' 'a

""-3 S^

0- N -

d., "
d I
$ ~. "
d~iil i
a41~ B

am e
4 4R 44 4'

2 i i 5 .
it .g s k 8
4' 4 4 a 4. 4
4' 4.44 4. 44 4.44 4

l *- Q c '4 4 n i


o 4


U 3

o .


o 0
L8 I


" o
'. "

~ B
5 9
3 ~

`" "


N =fi

Accumulation units are included in the text description.

Seagrass Community Biomass Compartment

Accumulation of seagrass community biomass (g m-2 day-1

in the modeled ecosystem was considered to be a function of

several major inflows and outflows. Inflow includes the

photosynthetic production of organic matter through the use

of light and uptake of nutrients available from water column

and sediments. These parameters were coupled by calculating

an arithmetic mean of the contribution of the available

nutrients to seagrass production (pmol 1-1 day-1) and then

calculating the geometric mean of this value and the

contribution to community production by light. In this

fashion, if one factor was severely limiting, production

became very low. Use of an arithmetic mean by itself would

allow some production to occur even if one of the parameters

was absent entirely, eliminating the concept of limiting

nutrients. This combined estimate was then multiplied by a

theoretical productivity maximum. This inflow represented an

estimate of the maximum possible production given the

existing level of available light and nutrients. The

expression was then converted to standing stock biomass

(g m-2) by dividing by maximum standing stock (g m-2), a

photosynthetic quotient of oxygen: carbon (g02 gC- ), and a

carbon to organic matter conversion gC g- biomass). This

resulted in a final value for the specific seagrass rate of
biomass production (growth) in units of g biomass m g

standing stock-.

Outflow from this compartment was attributed to three

components: loss of seagrass due to breakage by wind and

waves during storms (g m-2 day- ), routine die-off

(g m-2 day- ) and respiratory losses (g02 m-2 day- ). Blade

loss during storms was estimated to be 1% per event. Routine

die-off (death rate) was equivalent to complete annual turn-

over. Maintenance loss was estimated from field measurements

of oxygen change (gO, m-2 day-1) (as described in Chapter II)

and converted to standing stock biomass in the same manner

as the inflow measurements.

The detritus component (g m-2) in the model represented

the net accumulation resulting from three inflows and two

outflows. The three input factors included imported

detritus from nearby areas due to tidal currents (Ogden

1980), dead seagrass from routine die-off in the seagrass

compartment, and detritus gained from stirring of the

adjacent areas by thunderstorms with subsequent import by

currents. This latter amount was estimated from field

observations that the gain of detritus during a thunderstorm

from an adjacent square meter of seagrass appeared to equal

approximately 50 percent of the total in the adjacent area

less 10% which settled out quickly. This thunderstorm

mechanism also stirred up an equivalent amount of detritus

from the modeled square meter. This resulted in

simultaneous detritus import and export.

The major outflow from the detritus compartment was the

amount of detritus consumed or decomposed by consumers (Klug

1980). This outflow was calculated as a linear function of

the number of consumers present and their detritus consumption
rate (g consumer ) as shown in Figure 27a. The slope of

this curve (al) is not a constant; it is a variable which is

a rectangular hyperbolic function of the amount of detritus

present (Figure 27b). As the amount of detritus increases,

consumption also increases but approaches a point of satura-

tion where further addition of detritus causes little increase

in consumption.

This association of a linear curve with a variable slope

described by a hyperbolic function is a recurring relationship

throughout this model. It will be redescribed in each

pertinent section to delineate the specific relationships

included in the model, but the rationale is the same. Use

of this convention allows the limitation of the linear

relationship by the saturation curve. In this manner, rates

are not allowed to increase without limit. As the rate

begins to increase, saturation is approached and the slope

of the linear curve begins to get smaller, thus flattening

the curve.


The inflow and outflow to this compartment were

characterized using one relationship for each case. Inflow,

or growth of the population, was described as the growth of

new consumers formed as a linear function of the number of

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs