ABSTRACT
Rates of oxygen depletion per biomass were determined
separately for four groups of organisms: fish, invertebrates,
healthy algae, and necrotic algae. Rates of oxygen addition per
biomass were determined for healthy algae at varying degrees of
solar irradiance. These rates were incorporated into a computer
program and used to predict oxygen levels in tidepools with known
amounts of biomass of the above groups of organisms. The model
was calibrated using input values from one pool, and tested using
values from two other pools. The model predicted oxygen levels
reasonably close to actual field values for night depletion and
morning addition, but predicted a continuous rise in tidepool
oxygen during the afternoon, when actual tidepool oxygen
concentrations tend to level off. When manipulated over a
midnight to noon time course, the model highlighted the
importance of algae to both addition and depletion of oxygen in
tidepools. Other factors that showed significant effects on
ambient oxygen levels included invertebrate population,
daylength, and local topography.
INTRODUCTION
The tidepool is an interesting ecosystem to model because
physico-chemical parameters such as dissolved oxygen or
temperature display much greater diurnal fluctuation there
compared to variation seen in the ocean (See Figs. 1 & 2).
Temperatures in these pools at low tide during the day reach
levels never experienced in the ocean the whole year round, and
this limits the number of species that may exist there. For
instance, two species of tidepool cottids, Clinocottus globiceps
and Oligocottus maculosis, have upper temperature limits of 26°c,
and so they are only found in tidepools on the Pacific coast from
San Francisco northward, as temperatures in Monterey Bay
tidepools exceed this threshold (See Fig. 2). The cottid
Oligocottus snyderi resides in pools at Hopkins Marine Station,
and has an upper temperature limit of 30°C (Morris, 1960), well
above the maximum level recorded during a sunny day in June (Fig,
2).
While temperature can limit the number of species in
tidepools during the day, dissolved oxygen levels may limit the
number of species at night, when algae becomes a consumer
of oxygen instead of a producer. As Fig. 1 shows, oxygen levels
may dip to near zero in a tidepool by dawn, and fish cannot
survive at zero oxygen for long. Fig. 1 also demonstrates that
oxygen levels vary significantly from pool to pool, and that they
may decrease even during the day. It appears from this natural
variation that a number of factors determine ambient oxygen
factors comprises the focus of this investigation.
Here a computer simulation model is constructed to elucidate
factors controlling oxygen levels and the relative significances
of their control. Models are observer-defined abstractions of the
real world, and though they may only approximate what is
happening in an ecosystem, they can be valuable research tools
for a number of reasons (Wetzel and Wiegert, 1983). First of all,
they provide an easy way to "manipulate" the environment without
causing damage, and may be used to forecast potentially harmful
impacts of a questionable environmental alteration. Additionally,
a model can condense years into minutes, further assisting in the
long-term prediction of impacts. Models also determine sensitive
or controlling parameters and may be used to identify areas where
information is lacking or inadequate. These characteristics point
to computer modeling as a sound research vehicle for this
investigation into what factors, biological or physical,
determine ambient oxygen levels in tidepools.
MATERIALS AND METHODS
We identified sixteen tidepools on the coastline of Hopkins
Marine Station, located on the southern end of Monterey Bay, as
feasible potential study sites. Factors which went into
determining feasibility included size (less than one square meter
surface area), elevation (less than two meters above MLLW), and
separation from any interconnected pool networks. Each pool was
numbered and marked with surveyor's tape glued to the rock.
Tidepool surface areas were determined by approximating the
surface as a combination of triangles and rectangles. Volumes
were estimated by removal and measurement of the tidepool water.
We determined each tidepool's elevation above MLLW through the
use of a surveyor's transit.
Extensive field measurements were made for six randomly
chosen pools. I measured dissolved oxygen and temperature on site
with a Yellow Springs Instruments Model 54 oxygen meter, and
samples were taken and removed to the lab for determination of pH
and salinity. The readings spanned three days and nights, and
were obtained at high and low tides and at midpoints between
tides. These measurements were designed to identify any
correlations between tidepool physico-chemical parameters. I
calculated linear regressions of oxygen against corresponding
temperature readings (n=69) and pH readings (n=58) to note
whether such correlations existed.
An experiment was conducted to determine the degree of
mixing that takes place in a tidepool. The experiment was carried
out on a single oval-shaped tidepool on a calm day (wind speed =
0-5 mph) and on a windy day (w.s. - 15-20 mph) to note whether
air turbulence affects mixing. I performed the experiment on a
horseshoe-shaped pool as well to determine the effect of pool
shape on mixing. A wire grid marked off in inches was placed over
the tidepools and anchored against the rocks. I gently pipeted
1/2 ml of a 1 g/1 solution of fluorescein dye into the tidepool
at the center of the grid, starting the stopwatch simultaneously,
I recorded the time elapsed when dye crossed the marked lines on
the grid in the north, south, east, and west directions. Average
tidepool horizontal mixing velocity (THMV) was then determined by
dividing distance by the time recorded in all directions. These
values were added and then divided by the number of readings to
arrive at an estimate of average THMV.
To establish oxygen depletion rates for the simulation
model, rates of oxygen depletion by the different groups of
organisms were determined in the lab. About 100g of fish and
600g of invertebrates (e.g. crabs, snails, anenomes, urchins)
were captured and brought to the lab. I bubbled carbon dioxide
into an unmarked tidepool to facilitate fish capture. The
invertebrate species were sampled in roughly the same ratio as
seen in the tidepools. About 500g of healthy algae was collected
from the intertidal zone at low tide, mostly composed of species
found in the sample tidepools, and it was divided into two
samples. About 300g of upcast necrotic (rotting) algae was
collected from the West Beach of Hopkins Marine Station. Total
biomass of the fish, invertebrates, necrotic algae, and the two
samples of healthy algae were determined on a Mettler 1 kg
balance. The number of each kind of fish and invertebrate species
was recorded (Note: the calcium carbonate or chitin exoskeletons
of the invertebrates were included in the measurement of
biomass).
Seven aquaria were prepared for the experiment by filling
each with 20 liters of sea water and nothing else. They were
situated away from direct sunlight, and the algae tanks were
covered with black plastic to prevent photosynthesis from
occurring. I took oxygen and temperature readings as well as
samples for pH and salinity determination before loading any of
the groups of organisms into their respective aquaria. Using an
Aquarium Systems aquarium heater, I heated one tank to 25°C to
measure temperature effects on the oxygen consumption rate of
healthy algae. One aquarium was left empty as a control for
microbial influence upon oxygen levels, pH, and salinity. The
invertebrate tank was covered with a screen mesh to prevent the
animals from escaping. After the different samples were placed in
their aquaria, readings and samples were taken every three hours
for twelve hours (the algae tanks were sampled every four hours
for twelve hours). From these data I determined instantaneous
oxygen depletion rate per ambient oxygen level for fish,
invertebrates, algae at 25°C, algae at 15°C, necrotic algae, and
microorganisms [in umole 0 / (g biomass X hr. X ppm ambient o,))
(See Appendix 1).
Experiments were also carried out to determine rates of
oxygen addition by algae. These experiments were begun at 12
noon, so that oxygen addition rates obtained represented those at
maximum solar irradiance, which occurs at about 1 pm. About 1 kg
of healthy algae was collected in the manner described above.
Four aquaria were then prepared: one for direct sunlight exposure
and three for gradations of diffuse sunlight exposure. The
interiors of the aquaria were lined with black plastic to mimic
tidepool conditions. Without taking this precaution the aquaria
would receive radiant input from the sides as well as the top, a
situation that does not occur in tidepools. Two of the aquaria
were lined with the plastic to the 20 liter waterline. One of
these was placed in direct sunlight and one in a fully shaded
area to mimic the diffuse sunlight a trench tidepool would
receive. A third aquarium was lined with plastic to the level
where the algae resting on the bottom barely showed above the
plastic line, thus imitating diffuse sunlight exposure in an open
pool. A fourth aquarium was lined at a level in between the two
prior plastic levels, approximating the amount of diffuse
sunlight a semi-trench pool would receive. The latter two aquaria
were also placed in a fully shaded area.
All four aquaria were filled with 20 liters of filtered sea
water and initial oxygen levels were recorded. The collected
algae were divided into four samples, each of which was weighed
and loaded into one of the four prepared aquaria. Oxygen
readings were made at 15, 30, 45, 60, 90, 120, and 150 minutes.
Oxygen addition rates were calculated in umole O/(g biomassX
hour) for each of the four conditions represented by the
experimental aquaria (See Appendix 2).
To provide field data to test the oxygen model against, and
to demonstrate the stresses that the tidepool ecosystem must
endure, I took hourly oxygen and temperature readings from three
sample tidepools and from the ocean off Agassiz Beach, Hopkins
Marine Station. One of these pools, Tidepool 3, represented an
open pool, while #4 and #7 represented a trench pool and a semi¬
trench pool, respectively (See Fig. 3). These hourly readings,
taken on three separate days, combined to form one diurnal cycle.
The weather was virtually identical on the three days - sunny and
clear - and the readings fit together smoothly (See Figs. 1 and
2).
Before modeling oxygen flux, oxygen sources, sinks, and
pathways had to be clearly defined. A tidepool oxygen model
schematic was constructed for this purpose (See Fig. 4). Data
from the oxygen depletion and addition experiments provided
values to be assigned to the arrows in the diagram. The
dependency of various fluxes on the ambient oxygen level itself
implied by the data - is depicted in the schematic in the form of
dashed-line informational controls, and was entered into the
computer program within the flux equations (See Appendices 1 and
4). Oxygen addition rates were assumed to be dependent upon solar
irradiance, which can be modeled as:
L = Lmax X SIN(/daylength X time of day)
Since photosynthetic rates do not increase linearly with
irradiance, but rather in a fashion described in Fig. 5, oxygen
addition was assumed to ascend during the morning (and descend
during the evening) proportionally to the square root of the
sine, which seems to accurately model the increase in oxygen levels
observed during early morning in directly sunlit pools (See Fig.
10).
I introduced an ad hoc equation for diffusion of oxygen to
the atmosphere at levels above supersaturation (28 ppm):
Rate = (K) (Surface Area) ([0) - 7.966)
((o) in uM/1, S.A. in sq. m)
Such a development in the model can necessitate additional model
inputs - in this example, tidepool surface area. A full listing
of the oxygen model inputs can be found in Table 1. Note that the
times of sunrise and sunset are necessary inputs to define the
daylength, and subsequently influence oxygen addition in the
model.
To test the model, characteristics of real tidepools had to
be determined, for example the biomass of different groups of
organisms. Each of the sixteen sample tidepools were censused for
fish biomass.
To estimate invertebrate biomass I collected average-sized
individuals of different species found in the pools (e.g.,
Tegula, Anthopleura, etc.). The weights of these individuals were
recorded and used to convert the number of individuals to
biomass. I counted individuals of the different species in each
test pool.
Two types of healthy algal biomass had to be estimated in
the test pools: foliose and crustose. To estimate foliose
biomass, I collected fronds of algal species found in the test
pools, weighed them for reference, and counted the number of
similar-sized fronds found in the pools. I repeated this
procedure for the determination of necrotic algal biomass. I
collected samples of crustose algae (e.g., corallines, Hilden¬
brandia spp.) by hammer and chisel and measured the area of the
larger pieces. The weights of these pieces were determined, and
an area:biomass conversion factor was developed. Exposed area of
crustose algae was determined in each test pool, and estimates of
biomass were derived from these values.
Phytoplankton content of each of the sixteen sample
tidepools was determined by spectrophotometric analysis for
chlorophyll in the sample water. This survey was performed to
determine whether enough phytoplankton biomass was present to
affect net oxygen addition in the pools.
The model parameters were fine-tuned by using Tidepool 3
10
model input values. The diffuse sunlight coefficient of 0.178
umole 0/(g X hr.) was adjusted to 0.070 to provide for a tighter
fit to the field data. This adjustment indicates that
experimental conditions for the determination of algal oxygen
addition did not accurately mimic actual diffuse solar irradiance
occurring in the tidepools. The diffuse sunlight coefficient for
a semi-trench pool determined in the algal oxygen addition
experiment was approximately half that of the open pool [0.083
umole 0/ (g X hr.)], and this value was adjusted to 0.037,
using the same scale above. Once the model predicted oxygen
levels close to actual field levels along the time course of a
low tide in Tidepool 3 it was tested with the specified values
for Tidepools 4 and 7 for validation.
As soon as the model had achieved some degree of accuracy
with regard to all three test pools, model inputs were
manipulated to simulate changes in the tidepool ecosystem, and
predicted impacts on the oxygen levels were noted. Examples of
such manipulations include the removal or addition of fish,
removal of half the invertebrates or algae, and the addition of
drifting necrotic algae.
RESULTS
Linear regression analysis of 58 simultaneous oxygen and pH
readings yielded a product-moment correlation coefficient of
r=.932, which at 56 degrees of freedom is significant to the
Pe.O01 level. This highly significant correlation allowed for the
formulation of a linear equation predicting pH levels at any
11
given oxygen concentration (in ppm):
pH = 7.34434 + 0.093187 X [01
In the computer program, predicted oxygen levels are converted
into an estimation of pH using the above equation along with 953
confidence limits. The derivation of these limits comprises
Appendix 3. See also Fig. 6.
Oxygen and temperature were also found to be significantly
correlated, but not as directly (See Fig. 7). High oxygen
concentrations correspond to high temperature, however, which
goes against the physical relationship between the two
parameters, whereby oxygen is less soluble in water at higher
temperatures. Salinity was found to be uniform throughout the
tidepools.
The tidepool mixing experiment yielded higher average
horizontal mixing velocities during windy days compared to calm
days: 0.0556 inches/second compared to 0.0308 ips. During the
windy day test, the dye was untrackably dilute after about four
minutes, whereas the dye on the calm day could be tracked for 6-8
minutes. Additionally, on the windy day, fluorescein was
distributed evenly over the oval-shaped pool (t7) while in a
horseshoe-shaped pool (+3) the dye tended to drift toward one
end.
Oxygen depletion rates by fish, invertebrates, and algae are
depicted in Fig. 8, as are the weights of each group observed.
The fish consumed about as much oxygen per time as the other
groups, even though much less biomass was present. The control
tank showed no net depletion, so microorganisms were assumed to
have negligible effects upon oxygen levels compared to the other
12
groups of organisms. Algae at 25°C showed a greater depletion rate
than algae at 15°C. The initial steep slopes of the lines in Fig.
8 demonstrate what appears to be a dependency of depletion rate
on ambient oxygen levels. A sample calculation of this dependency
is given in Appendix 1. Only necrotic algae continued to deplete
oxygen linearly with time at oxygen levels less than 3 ppm.
Oxygen depletion constants for each group of organisms are listed
in Table 4 on a per-mass basis.
The two-hour algal oxygen addition experiments yielded
linear addition rates at all levels of sunlight exposure (see
Fig. 9). The calculation of addition coefficients (in umole o,/Ig X
hrj) with respect to exposure type was straightforward and is
presented in Appendix 2.
Field values of fish biomass and estimates of invertebrate
and algal biomass in the three model test pools (3,4,7) are
listed in Table 2. Other model inputs are also listed therein.
When entered into the model, these values lead to predictions
that follow actual field readings quite closely during the night
and morning (Figs. 10, 11, 12). However, during the afternoon,
the model continues to rapidly add oxygen due to photosynthesis,
while field values for tidepool oxygen show a tendency to level
off (See Fig. 1 to note this leveling effect). Increasing the
mass-flux coefficient, K, helps to dampen the model's erroneous
afternoon oxygen climb, but it also dampens the morning oxygen
rise which otherwise fits rather closely to the field readings.
Hence the model in its present form seems to predict night to
morning levels with some accuracy, but loses its usefulness from
13
about solar noon on.
The phytoplankton biomasses determined for the sample
tidepools were very low compared to their algal biomasses. The
average value found was 0.227 mg chlorophyll per pool - hardly
significant compared to the amount of macrophytic chlorophyll in
any given pool (over 500 g).
Manipulations of the model led to some interesting results
(see Table 3). Neither removal of fish nor addition (up
to the highest biomass found in the 16 sample tidepools) resulted
in any significant changes in oxygen levels. The most significant
results occurred when half the algal biomass was removed,
indicating the importance of algae in terms of affecting both
oxygen addition and depletion. Daylength was also found to be
important in determining oxygen concentrations. All other factors
held constant, the same pool during the second week of December
had oxygen levels decline further than in June because of the
longer night.
A tendency for Tegula to migrate above the waterline of
pools at night was noted during diurnal surveys. Pachygrapsus
crabs also enter and exit the pools often. These phenomena were
observed in Tidepool 4, which contained 145g of snails and crabs.
Removing these animals from the pool would hence mimic their
migratory behavior. However, when this was done in a model
simulation, the oxygen still fell to about the same level as when
the snails and crabs were present.
An upcast heap of Macrocystis came to rest in the trench of
Tidepool 4 during mid-May for about a week, until another high
tide carried it away again. Toward the end of its stay it was
14
emitting a strong odor, and no fish were noticeable in the pool.
Addition of 1000g necrotic algae (probably a gross underestimate
of actual levels) into the modeled pool resulted in a decline to
zero ppm oxygen 1 1/2 hours after separation from the ocean, not
gaining any oxygen until the next high tide. This simulation
points to drifting necrotic algae as a potential factor
controlling tidepool oxygen concentrations.
Tidepool 7 is largely encrusted by coralline algae
(estimated 1357.6g of 2041.8g total algae present there).
Simulation of a specific parasite which kills off all the
coralline algae, assuming no necrotic oxygen utilization, results
in oxygen levels significantly higher at dawn and significantly
lower at the next high tide (see Table 3). This simulation once
more demonstrates how important algae are to tidepool oxygen
depletion as well as addition.
Simulation of the removal of a large rock wall from the east
side of Tidepool 7 yielded predictions of oxygen levels which
differed significantly from levels with an intact wall present.
This alteration transforms the semi-trench pool into an open
pool, and allows direct sunlight to reach the pool about an hour
earlier. The higher oxygen levels experienced in the morning
without the rock wall demonstrate the importance of local
topography in determining oxygen concentrations during the day.
DISCUSSION
The relationship of oxygen levels to pH and temperature in
the field tidepool readings was striking. However, it would be a
mistake to conclude that a causal relationship exists, for
instance, between pH and oxygen levels in the tidepools simply
because they correlate significantly. One hypothesis regarding
this correlation is that a third input is controlling both
parameters separately. According to this hypothesis, the
photosynthesis-respiration exchange of CO, and O, gases is
responsible:
« PHOTOSYNT. (DAY)
(NIGHT) RESPIR.
0 + C(H,O) «= H,O + CO «=» HCO3-+ H+
In this way, when oxygen levels are high due to photosynthetic
activity, carbon is being fixed in the algae, depleting ambient
carbon dioxide stores, pulling overall equilibrium to the left of
the equation, lowering H" concentrations and making the solution
less acidic - the trend seen in the field and in the laboratory
(See Fig. 6).
Similarly, the correlation between oxygen and temperature
depicted in Fig. 7 can be viewed as an influence of one outside
factor upon two independent physico-chemical parameters. Solar
irradiation causes water temperature to rise, and it also causes
dissolved oxygen to rise due to its photosynthetic effect.
The discrepancies between predicted and actual values of
oxygen depicted in Figs. 10-12 are minor, but significant enough
to merit further development of the model. More importantly, the
model yet fails to predict the afternoon leveling effect
aforementioned. In this way, the model helps to point out what
16
information is lacking or inadequate, as some mechanism after
solar noon is either enhancing oxygen depletion or stemming algal
oxygen production. To fine tune the model further, one needs to
consider the fundamental assumptions of the model in its current
form. Some assumptions are listed in Table 5. The assumption that
temperature remains constant at lab conditions of 15-17°c is
invalid for tidepools, as demonstrated by the field data. Further
investigations into temperature effects on oxygen depletion rates
of the different groups of organisms could perhaps yield insight
into the afternoon leveling effect. For instance, Table 4 reveals
a clear positive correlation of algal oxygen consumption with
temperature.
Additionally, oxygen depletion was found to be dependent on
ambient oxygen levels, but the experiment that provided this
information was conducted at oxygen levels of 0-10 ppm with
readings taken every 3 hours, and therefore did not provide a
complete picture of ambient oxygen effects on depletion rates in
the tidepools (no readings above 10 ppm, regularly exceeded in my
tidepools). To obtain a more accurate picture of these effects,
an experiment could be conducted that began at levels of 25-30
ppm, with readings taken every 1/2 hour, which would offer better
estimates of instantaneous oxygen depletion rate per ambient
oxygen concentration.
The model also assumes uniform mixing in tidepools, which
may or may not be the case in different situations. The tidepool
mixing experiment showed that pool shape and air turbulence can
affect overall mixing rates. Field readings performed early in
17
the investigation showed that oxygen levels could vary
substantially in one pool - the extreme example being Tidepool 1,
a horseshoe-shaped pool which showed a gradient from 5.3 ppm at
one end to 7.2 ppm at the other at 3:00 pm on a sunny day.
Gradients present in pools on a much smaller scale, say along a
transect beginning at an algal frond and extending
perpendicularly outward, could not be detected with the Model 54
oxygen meter, as the probe had to be continuously moved to
complete a measurement. A study on microzonation of oxygen levels
within tidepools using a more precise technique of measurement
could lead to the determination of contours of oxygen
concentrations in tidepools. The Model 54 meter was also
inadequate for this investigation because the maximum oxygen
level it could measure was 20 ppm, a value regularly exceeded in
my tidepools.
Other assumptions of the model which may affect its accuracy
include the assumptions that oxygen depletion and addition rates
are similar among the different groups of organisms analyzed. The
model could benefit from a species by species analysis of oxygen
depletion and addition as well as a more precise field method of
biomass estimation.
The model in its present form does not account for seasonal
or geographic variations in such parameters as solar irradiance
and temperature, and this renders it very specific to the
Monterey Bay area. Field investigations conducted at various
locations during various seasons could provide a more accurate
view of tidepool oxygen dynamics. Additionally, much more testing
is necessary before this model can be deemed accurate. More
18
diurnal field readings and biomass data for more tidepools would
be necessary to adequately test and tune the model. One can be
confident in a model's predictive power only when the model is
tested in such a rigorous manner.
In spite of the many shortcomings of the model, some
conclusions could be drawn. The simulations pointed to algae as
the most important controlling factor in terms of both oxygen
addition and depletion. Additionally, the model suggests that
fish play an insignificant role in affecting ambient oxygen
levels in tidepools off of Hopkins Marine Station. These two
findings appear to be fundamentally derived from the biomass of
the groups of organisms present in tidepools as opposed to their
depletion rate per biomass, as fish certainly consume more oxygen
per time per biomass than algae, but the typical tidepool
contains so much more algal biomass, as shown in Table 2, that
the net depletion due to fish is insignificant.
Manipulating the model also helped to show the importance of
other factors in determining tidepool oxygen levels, such as
daylength, local topography, and random ecological events such
as upcast drifting necrotic algae. The next logical
step would be to remove half of the algae in one of the test
pools to see if oxygen levels are really affected in the way the
model predicts.
LITERATURE CITED
Morris, Robert W., "Temperature, Salinity, and Southern Limits of
Three Species of Pacific Cottid Fishes," Limnology and
Oceanography, 5:175-9, 1960.
Sokal, Robert B., and F. James Rohlf, Biometry, W.H. Freeman &
Co., New York, 1981.
Wetzel, Richard L., and Richard G. Wiegert, "Ecosystem Simulation
Models," from Carpenter, E. J., and D. Capone, eds.
Nitrogen in the Marine Environment, Academic Press, New
York, 1983.
ACKNOWLEDGEMENTS
I would like to thank Mark W. Denny for his invaluable
assistance and advice regarding this project, as well as for his
steadfast encouragement and patience. I also would like to thank
Ladd Johnson and Dick Zimmerman for their willingness to take
time out of their busy schedules to assist me in my research.
Lastly, but certainly not least, I would like to extend my
deepest thanks to Kelly Jensen, my field partner, for her time,
efforts, patience, intellectual stimulation, and good sense of
humor.
20
FIGURE LEGENDS
Figure 1: Oxygen field readings conducted over a diurnal cycle on
5/31/89 and 6/4/89. This graph demonstrates the fluctuations of
oxygen found in tidepools relative to the ocean, variations among
tidepools in oxygen concentration, and the afternoon leveling
effect unaccounted for by the simulation model.
Figure 2: Temperature field readings conducted over a diurnal
cycle on 5/31/89 and 6/4/89. This graph shows temperature
fluctuations found in tidepools during the day which exceed any
temperature extremes found in the ocean.
Figure 3: Local topography of the three model test pools. All
three sketches are views from the south-south west. These
sketches demonstrate how topography can affect the amount of
diffuse sunlight that may irradiate a tidepool.
Figure 4: Schematic of the computer simulation model. Dashed
boxes indicate factors not accounted for in the model. Large
rectangles indicate biological oxygen sources and sinks.
Triangles depict ambient oxygen concentration.
Figure 5: Rough mathematical relationship of photosynthetic
output to solar irradiance (Denny, 1989).
Figure 6: Scatter plot and linear regression of simultaneous
21
oxygen and pH readings (n=58) from the field and the lab showing
the strong correlation used to derive the pH predictive capacity
of the computer program.
Figure 7: Scatter plot and linear regression of simultaneous
oxygen and temperature field readings (n=69), indicating a strong
correlation.
Figure 8: Oxygen depletion by different groups of organisms.
Rates of depletion are listed in Table 4 on a per-mass basis.
Figure 9: Oxygen addition by algae at varying degrees of solar
irradiance.
Figure 10: Model test run on the calibration pool, Tidepool 3 (an
open pool), conducted over a low tide beginning in the middle of
the night and extending through to late morning.
Figure 11: Model test run on a validation pool, Tidepool 4 (a
trench pool), conducted over the same low tide as Tidepool 3.
Figure 12: Model test run on a validation pool, Tidepool 7 (a
semi-trench pool), conducted over the same low tide as the others.
22
APPENDIX 1

SONPLE, CALCULATION OF OXYGEN DEFLETICN
RATE PER AMBIENT CXYGEN CONCENTRATICN



s)
1
instantaessat u -Oky/(lX.
- 6.X (20 1 in aguarium
- 18 umoles y  (110.2 ima  hr.
= O.16333 umoles /g  hr.)
At Oxygen Level of: (3.2 - 5.5) / 2 - 5.85 ppm
4t X's reprasent midpoints between data points to which the calculated
instantaneous rates correspond.
** Dots ara the experimental data points.
APPENDIX2
SAMFLE CALCULATICN CF ALGAL OXYGEN ADDITICH
IN UMOLES OXYGEN PER  BICMASS FER HCUR
UNDER UARIOUS LEVELS OF SOLAR IRRADIANCE
Linear regreesions were calculated for each af the
data setsas they appearedoeaddingge

in each experimental aguaritm.
Direct Sunlight:
Eitpe af regreesion -.Se amolehr.)
Rateaf Additin ope
-m
Diffuse Sunlight (Trench Feol):
Slope af regreesion - G.164705 umle OE /(1X hr.)
Rate of Addition = Slope X (20 1 / 245.0 g blomass)
- 0.01 umola hr.)
Diffuse Sunlight (Open Pool):
Slope of regression = 2.05029 umale 02 / (1 X hr.)
Rate of Addition = Slope X (20 1 / 230.0 g biomass)
- 0.178 umle  h.)
Diffuse Sunlight (Semi-Trench Fool)
Slope of regression = 1.001754 umole C2 / (1 X hr.)
Rate of Addition = Slope X (20 1 / 235.g biomass)
= 0.083 umole 02 /(qXhr.)
APPENDIX 3
Linear Regressien Eqvation:
- 0.093187 o t 7 37139
XIndepender
varalle (Lo.1)
AFR
dependat
[=.732
Varfable (pH)
Jp. - S6 n-58
ttest for significance ef
(o.073187-0
t- Sope- -
= 77.23
0.004 876
Sslope
t 3.20
Nate: sope ande
ott ho The lineat
P.00
rgsin pte
781 Canfdence limik for pl Espnales
T latus S.tware,
2X- 449.4
DY-467.85
X= 7.748
Vr.74
x4404.04
2
2Xy - 3110.746
(X)  921.905
X
(9 - 30.8844
Zy - 2)
(EX)(29) - 85915
Lry - ZXy-
Explahed Variaten!

(85.915)
29
-8006
921.765
2x
Onexplawed Varate!
= 2229 = 30.8841-8.000 - 22.88
Xx
w SZx - 22.82 - 0.4085
V.)
5
h-2
Sadard tr.f Pestate fon
4 (K.-3)
59 =
464-7748)
s n
5.4085
2x
721.965
Cerfidaee lait
Ly - pt  (to)sy)
la - ph- t)(s)
11
APPENDIX4
JEM MODEL
O
++ TIDEFOCL OX
SY STEUEN MOCRE
JUME 7. 1780
S:PRINT"TIDEFCOL OXVG
N MCLEL"
10 CL
PFINT
— — —
SOLUED GXYGEN LEUELS TN
TILEF
FOCLS FECM THE
15 ERINT"
'THIS MCDEL WILL PREDICT DIS!
ION.
AM IO THE TIME OF SE-INUNDA
SEFARATION FEOM THE OC
IME OF
-I
E LIMITE TAN
CORFD
CONETFUCT 7
15 PRINT:FRINT"THE MODEL WILL ALSC!
MOXYGEN AND H LEVELE. ANFL
ICN EE
ERVED FIELD COFREL
OF H BASED ON OE
- = 28."
T"ELEASE IHEUT THE FELLCHTME ALUET
NNDHEL
EINTERI
IMEE
a-


Tidenel Volume

MEL
aaa-
ttaaat

NEE
O

F
a
MEUI"Time of Re-inundation by Oeean heuryninute H."m,e
E

I"Iime af Sanrise hoarminute'

IMEUI"Iime of Sunset (heuraminute!':s.
47 INEUT"EStimate Leyel of Cloudiness (-Cles -npletaly Overcast;
SIF W35 THEM 57
70 PRIMT"Now, Enter the range during the day in which the Tidepool is entirely
illuminated by Direct Sunlight."
75 INFUT"Baginning Time (hour,minute, r F)":EH,EM,ABT
IMEUT"Ending Time (hour.minute.aor
EE
IMEMI"Local Topography Factor (0-Open, 1-Semi-Tranch, E-Trench)";G
2IF TG2 THEN
TG=O THEN TF=.07
74 IF7
THEM TF=.67
95 IF TG=1
96 IF TG=2 THEN TF=.022
100 ERINT"Now, enter the time increment you'd wish to have owygen readings pred:
(in minutes).
cted at
110 INPUT
IHEIMUATICN
115 INEUT"WQULD YOU LIKE A FRINTOUT MADE
117 IF P="Y" THEN PT=1
140 HI=H:MI=M:AIt=AS:GOSUB 5000:T-71
000:TM=T
145 HI=HH:MI=MM:A1S=AAS:GOSUB
5000:TB=T1
SO HI=BH:MI=BM:AIS=ABS:GOSUB
5000:TE=I1
155 HI=EH:MI=EM:A1S=ACS:GOSUB
160 HI=RH:MI=RM:A1S="A":GOSUB
SO00:TR=T1
165 HI=SH:MI=SM:A1S="P":GOSUB
5000:TS=T1
US="*4.*4“
N
TI=1
CLS
00
205 IF TDTN THEN T6=1
OH 952 CONFIDENCE LIMIT!
210 PRINT,,"
225 PRINT"TINE", "OXYGEN (ppm)", "pH", "LOUER", "UPPER
O M2=T-(60(INT(T/80)))
H2=INT(T/0)
FM=O
IF H2I2 THEN HA-H2-12:FM=1
37 IF H2-O THEN H2=12
239 IF H2=12 THEN PM=1
245 PRINT HE;:PRINT USING":44";M2;
247 IF PM=O THEN FRINT"am", ELSE FRINT"pm",
248 IF OHO THEN O=O
249 GO
U 00
2EO FRINT O/V,:FRINT USING US;FH;:FRINT TAB(4S);:FRINT USING US:LI::FRINT TAEC
);:FRINT USING USL
260 X1=0:X2-0:XS=O:X4=0:XS-O:X-0
270 IF PT=O THEM 27E
PRINT HE;:LFRINT USING":44":M2
272
27
IF FM=O THEN LFRIMT"am", ELSE LFRIN
"m"
274
PRINT O/V.:LFRINT USING"44.44"; (7.34434+(.673187(0/0))
275
IE TETH THEM 999
I=II
280
00
T=T+1
IFT14 THEM
TH=1 THEN 310
304
F TTH THEN TETM
3
I1=1/60
310
—.
REM FISH DEFLEION
D1=.007014+(.580404(0/))
1=DIAFEII
2
230 O=O-X1
HEM INVEETE
-RIEERATE DEPLETION
IF D/VSS THEN D2=.008(079)
TO 345
FO/OS4.S THEN D2=.00435(/):C
34
344
2=.00(0/V)
X2=D2IEI1
345
35
C=O-X2
EALTHY ALGAE
360
IF TTR AND TKTS THEN 380
365
REM DEFLETICN IN DAEE
IF O/V39.S THEN DS=.00727(0/0):GOTO 372
369
371 DS=.00S(O/)
(3=D3I1kAB
1=0-X3
375
37
TATR THEN 450
IF TSTB AND TKTE THEN 400
380
ssa
IN DIFFUSE SUNLIGHT
FEM ADDITION
S4=SIN((3.14157274/(TS-TR))(T-TR))
383
38
Q=0+X4
388
X4=(TF(S40.7))I1AB
389
Q=0+X4
GOTO 450
390
399
MM ADDITION IN DIRECT SUNLIGHT
SS=SIN((3.14159274/(TS-TR))(T-TR)
400
27-(W((.3
7-TF)/5)))(SS.7))I14B
5=.
405
410 C=OX5
450
NECROTIC ALGAE DEPLETION
REM
455
X6=.OS3IINB
O=C-X6
OXYGEN DIFFUSION TO ATMOSFHERE

DX=.35SA((0/V)-7.766)
466
IF DXO THEN 475
468
470
C=O-DX
475
IF TETN THEN 230
IF TTI THEM 300
480

GOTO
797
END
FH=(7.34434+(.093187(0/9)))
000
SY=(.408((1/8)(((0/)-7.748)2)/21.7
400
Li-FH-(2.0042)
4010
2-FH+(2.0042)
4011
RET
402
1=0:IF AIS-"A" THEMO
o
IF HIE12 THEM TI(HIO)-MI ELSE TI((1S-HI0+MI
SOOt
TO EOBE
01 IF HI=12 THEN TIEMI ELSE TI(HIOHI
SOS RETURN
MODEL INPUTS
TIDEPOOL OXYGEN
FISH BIOMASS
INVERTEBRATE BIOMASS
HEALTHY ALGAE BIOMASS
NECROTIC ALGAE BIOMASS
TIDEPOOL VOLUME
TIDEPOOL SURFACE AREA
INITIAL HIGH TIDE OXYGEN LEVEL
TIME OF SEPARATION FROM OCEAN
TIME OF RE-INUNDATION BY OCEAN
TIME OF SUNRISE
TIME OF SUNSET
ESTIMOTE OF CLOUDINESS LEVEL (O=CLEAR, 5-COMPLETELY OVERCAST)
IIHE RONGE IN WHICH TIDEPOOL IS IRRADIATED BY DIRECT SUNLIGHT
LOCOL TOPOGRAPHY FACTOR (O=OPEN, 1=SEMI-TRENCH, 2-TRENCH)
I
02
O m
7
12
19


7
a-
—
Table 3: Some Manipulations of the Model
Maximum OI
Minimum LOe
Manipulation
What It
Represents
at 5:45 am
at 10:45 am
TIDEPOOL 3
—
20.4
3.0
None
——
3.0
20.4
Removal of Fish
20.0
2.9
ii of
Highest Fish Mass
Found in 15 Sample
2.g Fis
Tidepools.
12.7
——
Remaval af 1/2
4.1
Algal Biomass
———
22.3
3.5
Remal of 1/
Inyert. Biomass
2.2 (7:4am)
Canrise at 7:0am Same Tidepool.
17.1
in December
TIDEFCOL
—---——
S.2
.3
Nae
0.9
8.2
Migration of
Removal of 1459
Tegula and
Invertebrates
Pachygrapsus
at Night
Upcast Heap
O.O
.O
itn o
f Drifting
Necrtic Algae
Macrocystis
TIDEFOOL
—--—
2.7
10.-
None
S.3
Parasite ills Off
Remoal of
4.2
Al1 Crustose
1357.6g Algae
Coralline Algae
In Tidepool
17.3
Removal of Large
2.7
Cpen Fool
emi
Rock Wall From
Insteadof
at ie f ol
Trench Pool:
with irect
Sunlight
Irradiating
at 7:Soam
Table 4: Per-Nass Oxygen Depletion Constants
Derived From Experimental Data
Depletion Rate (umoles Oxygen 7 fg X hr. X LGxygenl in ppal)
Fish
O.020404
9.006497
Invertebrates
Algae at 15
O.00S754
lgae at
O.011128
Necrati lgae
.O5 umel ehr.
(independent of Cxygan
SOME ASSUMPTIONS OF THE MODEL
STEADY STATE RATES OF OXYGEN ADDITION AND DEFLETION
UNIFOEM MIXING OF TIDEFOOL
UNIFOEM OXYGEN DEFLETION FATES AMONG ALL INVERTEEFATE SFECIES
UNIFOEM OXYGEN DEFLETION RATES AMONG ALL FISH EFECIES
UNIFOEM OXYGEN DEFLETION FATES AMONG ALL SFECIES OF ALGAE
UNIFOEM OXYGEN ADDITION RATES AMOMG ALL SFECIES OF ALGAE
OXYGEN CONSUMFTION RATES INCREASE WITH OXYGEN CONCENTRATION
ALL INVERTEBRATES STAY IN FOOL FOR DURATION OF LOW TIDE
COMSTANT TEMPERATURE OF 15-17 DEGREES C
MICROBIAL ACTIVITY HAS NEGLIGIBLE EFFECT
OXYGEN DIFFUSION TO ATMOSPHERE DUE TO SUFEF-SATUFATION
IS CONTRIBUTING TO DEPLETION AT HIGHER OXYGEN LEVELS
V
DISSOLVED OXYGEN (ppm)
S





D
Z

G)
Z
Z
0
8+
H
x0

E
TEMPERATURE (C)
S +
5
O




+
2 8

Ae
O

7
2
I
Z
G)
0
3
C
+
S
8
—
7

3
SOLAR RRADIANCE
pH
aatataavaaaaaaaa-


o-
04
1
D

o
D

( -


1
D
p
)

+

-

9
51 -

-

TEIAPERATURE (degrees C)
tatakakataaaaaavaa-


&
5

6

1Z

9
o+

O¬
o.
G.
I
Z
2

O
O 0

DISSOLVED OXYGEN (ppm)
o+
DISSOLVED OXYGEN (ppm)
N

6
5

—

G)
2
9
X
30

+
N +

4
6

Z
—
T


0
DISSOLVED OXGEN (pm

L
25
9
-

30
UISSOLVEU CXYGEH (pp
n


T

9
E


-
H

Z
—
—


—
+
N-
-
n
DISSOLVED ONGEN (pom)

-

P


a

L 4

Z

C

—