EPA R3 73 018
Ecological Research Series
APRIL 1973
An in Situ Evaluation of
Nutrient Effects in Lakes
Office of Research and Monitoring
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and
Monitoring, Environmental Protection Agency, have
been grouped into five series. These five broad
categories were established to facilitate further
development and application of environmental
technology. Elimination of traditional grouping
was consciously planned to foster technology
transfer and a maximum interface in related
fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
-------
EPA-R3-73-018
April 1973
AN IN SITU EVALUATION OF NUTRIENT
EFFECTS IN LAKES
By
Robert A. Jordan
Michael E. Bender
Project 16010 HIU
Project Officer
Charles F. Powers
National Eutrophication Research Program
Environmental Protection Agency
National Environmental Research Center
Corvallis, Oregon
Prepared for
OFFICE OF RESEARCH AND MONITORING
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402
Price $2,00 domestic postpaid or $2.25 GPO Bookstore
-------
EPA Review Notice
This report has been reviewed by the
Environmental Protection Agency and
approved for publication. Approval
does not signify that the contents
necessarily reflect the views and
policies of the Environmental
Protection Agency, nor does mention
of trade names or commercial products
constitute endorsement or recommendation
for use.
-------
ABSTRACT
A method for performing in situ nutrient enrichment
experiments on natural lalce~phytoplankton communities
was developed and evaluated. One set of experiments
in which it was employed was designed to detect
limiting nutrients and to provide a basis for pre-
dicting future experiment results. Productivity
increased in response to all three of the treatment
variables used, N, P, and EDTA, but response patterns
varied from experiment to experiment. Individual
species responded differently to different treatments,
and interactions among the treatment variables were
important in shaping the community responses to
mixtures of two or three variables. The most consis-
tent features of the productivity results were incor-
porated into a "most probable response pattern," which
was partially validated by a second series of experi-
ments .
The second experiment series was also used to test the
ability of NTA to stimulate phytoplankton productivity.
Stimulation was continually obtained.
In a third series of experiments sewage effluents were
tested in parallel with N and P. Varying degrees of
overlap between the species complexes responding to the
sewage and to the N and P treatments were found.
Recommendations for the use of in situ enrichment ex-
periments in eutrophication studTes are presented.
This report was submitted in fulfillment of Project
Number 16010 HIU under the partial sponsorship of the
Water Quality Office, Environmental Protection Agency.
111
-------
CONTENTS
Section Page
I Conclusions and Recommendations 1
II Introduction 9
III Methods 21
IV Experiments to Identify Limiting
Nutrients and Evaluate Predictive
Potential 59
V Experiments to Test a Potential
Environmental Contaminant 105
VI Experiments to Evaluate Stimulation
by Sewage Effluents 113
VII Containment Effects 119
VIII Acknowledgements 139
IX References 141
X Publications 153
XI Appendices 155
v
-------
FIGURES
No. Page
1 Representative Summer Photosynthesis and
Light Intensity Profiles in Third Sister
Lake, 1969 25
2 Productivity Profiles, Third Sister Lake,
1970 26
3 Light and Temperature Profiles, Third
Sister Lake, 5-6-71 27
4 Productivity Profiles, Third Sister Lake,
5-6-71 28
5 Productivity and Incident Light Levels,
Third Sister Lake, 5-6-71 29
6 Comparison of Means and 95% Confidence
Intervals for Batches of 1 ml Ampoules
and Series of 1 ml Shots from an Automatic
Syringe 37
7 Activity of ^C Stock Solutions as Moni-
tored by Scintillation Counts on Days
of use 39
8 Productivity Series, Third Sister Lake,
8-15-69 43
9 Productivity Series, Third Sister Lake,
6-3-71 45
10 Sampling Methods Employed in Experiments
4A and 4B 47
11 Comparisons of Means and 951 Confidence
Intervals for Groups of Lake Water Samples
Taken from a Jug by Four Different Sampling
Methods 49
12 Effect of Nonuniform Jug Filling in
Experiment 5 50
13 Removal of Jug Filling Effect 52
14 Subsequent Experiments: Results of Improved
Jug Filling Technique 53
VI
-------
No.
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Experiment 8: Treatment Effects
Experiment 17: Treatment Effects
Experiment 12, Productivity Response
Patterns
Experiment 12, Growth Response Patterns I
Experiment 12, Growth Response Patterns II
Examples of NP Interactions
Comparison of Productivity Response
Patterns, Experiments 1-20
Spring Thermal Conditions in Third Sister
Lake - 1969, 1970
Fall Thermal Conditions in Third Sister
Lake - 1968, 1969
Numbers of Species in Common in Each Pair
of Experiments - 1968, 1969
Sorensen's Index Values for Total Species
Complexes - 1968, 1969
Numbers of Responding Species in Common
in Each Pair of Experiments - 1968, 1969
Sorensen's Index Values for Responding
Species - 1968, 1969
Anabaena wisconsinense Biomass Estimates -
1968, 1969
Ankistrodesmus falcatus Biomass Estimates -
1968, 1969
Cryptomonas ovata Biomass Estimates -
1968, 1969
Chroomonas acuta Biomass Estimates -
1968, 1969
Rhabdoderma sigmoidea Biomass Estimates -
Page
62
63
67
68
71
74
76
79
80
84
85
88
89
91
92
93
94
1968, 1969 96
VII
-------
No.
33
34
35
36
37
38
39
40
41
42
43
44
45
Synedra rumpens Biomass Estimates -
1968, 1969
Lyngbya limnetica Biomass Estimates -
1968, 1969
Chrysidalis sp. Biomass Estimates -
1968, 1969
Gomphosphaeria lacustris Biomass Esti-
mates - 1968, 1969
Aphanocapsa elachista Biomass Estimates -
1968, 1969
Productivity Response Patterns vs
Responding Species - 1969
Interaction Between NTA and Zn, Experi-
ment 25
Sewage Experiments: Productivity
Response Patterns
Productivity Trends in Control Jugs -
1968, 1969
Biomass Estimates for Neutral Species vs
Complete Communities
Comparison of Productivity Profiles:
Two Columns vs Two Lake Stations
Productivity Trends in Control Tubes
Productivity Trends: Control Jars vs
Page
97
98
99
100
101
103
109
115
121
126
127
129
Control Jugs 131
Vlll
-------
TABLES
No. Page
1 2x2x2 Replicated Factorial Design 21
2 Design of Experiment 1 31
3 Experiment 1 (June 1968) : Raw Data 33
4 Standardization of 14C Ampoules by
Scintillation Counting 36
5 Comparison of Replicability of Ampoules
and Syringe Shots 35
6 Conversion of CPM to Carbon Fixed 41
7 Experiment 3 (8-7/8-10-68): Raw Data -
Effect of Exposure Rack on 14C Uptake 48
8 Evaluation of Sensitivity of Method 54
9 Summary of Experimental Designs,
Experiments 1-20 60
10 Design of Experiment 12 66
11 Summary of Productivity Response
Patterns, Experiments 1-20 72
12 Frequency of Responses, Experiments 1-20 72
13 Tests for Similarities among "Near"
versus "Distant" Experiments 77
14 Ambient Lake Conditions: Experiments 1-20 81
15 Occurrence of Phytoplankton Species in
Initial Communities - 1968, 1969 83
16 Sums of Sorensen's Index Values -
1968, 1969 86
17 Responding Species - 1968, 1969 87
18 Summary of Productivity Response
Patterns, Experiments 22-34 106
19 Frequency of Responses, Experiments 22-34 106
IX
-------
N£. Page
20 Comparison of NTA and EDTA Effects,
Experiments 22-34 108
21 Experiment 22 Species Response Patterns,
Treatment Means (ug/1) 111
22 Experiment 29 Species Response Patterns,
Treatment Means (yg/1) 111
23 Sewage Experiments: Background Data 113
24 Sewage Experiments: Summary of Treatment
Effects 116
25 Comparison of Lake and Control Productiv-
ity Measurements, 1970 and 1971
Experiments 122
26 Sewage Experiments: Summary of Contain-
ment Effects 124
27 Sewage Experiments: Summary of Treat-
ment and Containment Effects 124
28 Experiment 18, 10-13-69 - 10-18-69 132
29 Experiment 18: Biomass Estimates (yg/1) 133
30 Physico-chemical Conditions in Control
and High Treatment Jugs 135
-------
SECTION I
CONCLUSIONS AND RECOMMENDATIONS
PROCEDURAL ASPECTS OF _IN SITU NUTRIENT ENRICHMENT
EXPERIMENTS
Statisti(:al D
Growth of phytoplankton in samples of natural waters
was frequently stimulated by more than one nutrient
at a given time, and interactions among different
nutrient treatments frequently appeared. Therefore
it was concluded that in experiments intended to de-
termine what may be limiting phytoplankton growth in
a natural system, treatments with several potentially
important nutrients should be employed. The factorial
statistical design is particularly useful in multivar-
iate experiments of this type because it permits iden-
tification of interactions as well as independent
effects .
Response Measurement
The ultimate response of interest in an enrichment
experiment is algal growth. Since growth is difficult
to measure directly in natural communities productivity
is used as a rapidly obtainable index of growth. In
this study the validity of productivity as a growth
index was verified in numerous comparisons between pro-
ductivity and cell count data. However, because of
published reports of inconsistencies between productiv-
ity and cell count results the relationship should be
tested routinely in any study employing in. situ nutrient
enrichment experiments. Experiments shouTd be conducted
for several days in order to allow detectable growth
responses to occur and in order for changes in produc-
tivity response patterns to be followed.
Sjms itivity
The sensitivity of an in situ enrichment experiment
in detecting treatment effects depends strongly \ipon
the degree of replicability attained between experimen-
tal units receiving identical treatments. In this
study, within- treatment replicability was continually
improved so that in some of the later experiments pro-
ductivity elevations of as little as 10% over the con-
trol levels represented statistically significant treat-
ment effects .
-------
INTERPRETATION OF RESULTS
Variability in Response Patterns within
Experiments
In this study, day to day changes in productivity re-
sponse patterns within an experiment were found to
reflect changes in the relative contributions of in-
dividual species to community productivity. Therefore,
all productivity responses detected in an experiment,
whether or not they persisted throughout the experiment,
were assumed to contribute to the accumulation of
algal biomass and were regarded as legitimate treat-
ment effects.
Variability in Response Patterns between
Experiments
Within series of similar experiments conducted in the
ice free seasons of 1968, 1969, and 1970 certain fea-
tures of the response patterns were relatively consis-
tent, while other features varied from experiment to
experiment. The consistent features appeared to re-
late to physio-chemical environmental conditions, while
the variations could be attributed to changes in the
species composition of the lake phytoplankton. The
composition of the total community shifted gradually
from experiment to experiment, while the composition of
the fraction of the community that responded to the
nutrient treatments shifted abruptly. Species that re-
sponded to treatments in one experiment did not neces-
sarily respond when they were present in others. It
was concluded that the best way to overcome the vari-
ability among experiments was to conduct enough experi-
ments within any temporal series to determine tne
consistent features, and use these features to charac-
terize the average response.
TREATMENT EFFECTS
Selectivity of_ Treatment Effects
When a treatment effect is detected by productivity
measurements it does not necessarily mean that the
entire algal community responded to the treatment.
Actually, nutrient treatments in in. situ enrichment
experiments are highly selective, and in many experi-
ments in this study only one species was found that
could account for the productivity responses. Produc-
tivity measurements detect the quantitative component
-------
of community response; cell counts are necessary to
determine the qualitative component.
Individual species did not necessarily respond to
treatments in all experiments in which they were present.
There was some suggestion, however, that a given species
was more likely to respond during a period of ascendancy
in the natural community than during periods of rela-
tively stable population size. Thus it did not often
happen that the species shifts in the experiment contain-
ers resulted in dominance by a normally rare form, a
result reported by Thomas (1964) .
The fact that treatment effects were highly selective
implies that it cannot be concluded that a nutrient
that stimulates productivity in an in situ enrichment
experiment is limiting community productivity. It may
be limiting the growth of the species that respond to
it. This point is particularly evident when it is con-
sidered that species that did not respond to the nutri-
ent treatments often were able to continue growing in
the control communities, in the absence of enrichment.
In Third Sister Lake species of bluegreen algae respon-
ded more frequently than species of other groups, im-
plying that enrichment of the lake could favor dominance
by bluegreens.
Importance ojf Interactions
Interactions among the nutrient treatments frequently
appeared in the productivity data. The most consistently
important interaction was synergism between N and P.
Species counts verified that the interactions occurred
on the species level as well as on the community level.
Interactions of the types observed undoubtedly contri-
bute to the overall effects of nutrient mixtures enter-
ing natural waters.
Stimulation by_ Chelators
Both EDTA and NTA stimulated productivity in most of
the experiments in which they were employed. Their
effects were generally independent of the effects of
accompanying N and P treatments. When both chelators
were tested at equal molarities in the same experiment
their effects usually differed in intensity. EDTA,
whose complexes with metals have higher stability con-
stants than those of NTA, usually stimulated more strong-
ly. Occasionally one compound stimulated and the other
did not, implying that their effects may have differed
-------
in mechanism as well as in intensity.
APPLICATIONS OF DJ SITU ENRICHMENT EXPERIMENTS
Identification o_f Limiting Nutrients and
Prediction of~FuTure Responses
One application of in s itu enrichment experiments is
to identify growth-limiting nutrients and to develop
background data on the responses of the phytoplankton
community to additions of those nutrients, that can be
used to predict future responses. In this study the
results of two years of experiments were used to de-
velop a "most probable response pattern," and this
pattern was compared to the results of a third year of
experiments.
The experiments of the first two years indicated that
all three of the nutrient variables employed could
stimulate algal growth, and that their order of impor-
tance as growth stimulators was P>EDTA>N. The predic-
tion that this order would be maintained in the follow-
ing experiment series was successful, but more specific
predictions regarding seasonal variations of the re-
sponse patterns were not. Thus it was concluded that
in situ enrichment experiments could be used to identify
Important controlling nutrients in natural systems, but
their results could be used to make only very general
predictions concerning when or how additions of these
nutrients are likely to stimulate algal growth.
From analyses of the behavior of individual species in
the experiments it was concluded that treatment effects
on the species level could not be predicted as readily
as treatment effects on the community level. One reason
is that similar productivity response patterns were
obtained in different experiments in which different
species groups responded to the treatments. A second
reason is that individual species did not necessarily
respond similarly in different experiments in which they
were present.
Testing p_f a. Potential Environmental
Contaminant
This application was attempted for NTA, and it was found
that this substance could stimulate algal productivity
fairly consistently at a treatment level of .252 mg/1.
Side experiments indicated that NTA stimulation was due
more probably to a chelation mechanism than to its
-------
utilization as a nitrogen source.
The effects of the NTA treatments varied from experi-
ment to experiment. Therefore in this as well as in
the first application of in situ enrichment experi-
ments, it is important thaT a series of experiments
be performed if the "average" effect is to be deter-
mined. Several experiments are necessary, also, to
avoid the possibility of concluding from one or two
experiments with negative results that the treatment
is biologically inert, when under other conditions or
for other species it may be active.
Interpreting Stimulation by a_ Natural
Nutrient Mixture
A third application of in. situ enrichment experiments
is to interpret the stimulatory effect of a natural
nutrient mixture in terms of its components. This
was attempted for samples of sewage effluents in
several experiments. Effects of the sewage treatments
tended to exceed the effects of parallel treatments
with N and P. It was found that because of the high
selectivity of treatment effects the extent to which
N and P contributed to the sewage effects could not be
reliably determined by comparing only productivity
responses to the treatments. In one experiment respon-
ses to treatments with sewage and with a known nutrient
mixture were similar In terms of productivity, but
when individual species growth responses were deter-
mined it was found that entirely different groups of
species responded to the two treatments. Two other ex-
periments revealed different degrees of overlap between
the species complexes responding to the sewage and to
the nutrient mixtures. Thus when the effects of treat-
ments are compared in this way, species counts must be
done to avoid drawing erroneous conclusions from the
productivity results.
Although the high selectivity of treatment effects
complicates the comparison of productivity responses to
two treatments, it is the main reason why in situ bio-
assays can be used very effectively in interpreting
stimulatory effects of complex nutrient mixtures. It
is conceivable that if enough of the components of the
mixture \\rere tested, separately and in combinations,
most of the stimulatory effects of the mixture could be
explained.
Examination of the species data for one of the experi-
ments suggested that interactions between nitrogen and
-------
phosphorus, on the species level, contributed to the
overall sewage effect.
CONTAINMENT EFFECTS
Community Level versus Species^ Level
Containment effects appeared on the community level as
declines in productivity with time in the experimental
vessels or as changes in productivity in the vessels
relative to productivity in the lake. On the species
level the containment effects were highly selective,
with different species exhibiting many different patterns
of response. Responses often changed from experiment
to experiment for the same species. The resultant of
the individual species responses was a shift in the
species composition of the experimental communities.
A species that did not decline but remained stable in a
control community was not necessarily neutral to con-
tainment, since it may have increased in the lake. Thus
the true measure of a containment effect is the differ-
ence in the fate of the species in the jug from its
fate in the lake.
Responses t£ Treatments and Containment
Species that responded to the nutrient treatments were
usually neutral to containment or responded positively
to it, but occasional positive treatment effects in
conjunction with negative containment effects appeared.
Consistent Sensitivity tp_ Containment
Of the algae that exhibited negative containment effects,
the taxonomic group that was the most sensitive was the
Cryptophyta. One species, Chroomonas acuta, was present
in the initial community of almost every experiment,
but almost invariably disappeared from the experimental
vessels by the end. Since this species was often very
abundant in the initial communities it can be concluded
that the jug environment, at least for Third Sister Lake
phytoplankton, tended to make abundant species rare more
often than it made rare species abundant.
Dependence on Type of Container
Containment effects depend to some degree on the type
container employed. In this study species responded
in different ways to a closed 19 liter jug suspended in
-------
the lake and to an open 4 liter jar incubated in a
water bath beside the lake.
Rate p_£ Development
Containment effects develop rapidly, often within a few
hours of the start of incubation.
Effects on Application p_f Results
Superficial analysis of results, ignoring containment
effects, is sufficient for approaching general questions
such as whether a treatment nutrient is limiting some
member of the community or could stimulate productivity
in the natural system. Because of the uncertainties
involved, however, in. si-tu enrichment experiments em-
ployed in eutrophication studies should be used in
conjunction with other evaluation procedures, such as
studies of nutrient dynamics in the natural system or
physiological assays for nutrient limitation, and
should not be used as the sole basis for conclusions.
-------
SECTION II
INTRODUCTION
THE ENRICHMENT PROBLEM
The biological productivity of a body of water is
influenced by many factors (Rawson 1939). Among these
are its morphometry, the climate in which it is found,
and its content of dissolved and suspended matter,
loosely referred to as its "water quality," Morph-
ometry and climate tend to change relatively slowly,
but water quality can change rapidly in response to
changes in the surrounding drainage basin. Man is an
increasingly widespread and potent effector of changes
in drainage basins, but he has failed to control his
activities sufficiently to avoid unplanned consequences
in neighboring waters. As a result numerous pollution
problems have developed that reduce the suitability
of natural waters for indigenous organisms and for use
by man.
One consequence of uncontrolled human activities can
be the enrichment of a water body with plant nutrients,
a process referred to as "eutrophication" (Stewart
and Rohlich 1967). The relationship between human ac-
tivities and the release of nutrients into natural
waters is most easily established where sewage effluents
enter lakes and streams (Hasler 1947) , less readily
confirmed where fertilizers applied to agricultural
soils are washed in (Stanford e_t a 1. 1970) , and even
more obscure where natural terFestTial nutrient cycles
are destroyed, permitting soil nutrients to escape into
drainage waters (Bormann and Likens 1970).
Eutrophication is not inherently bad, and in fact it
has been done deliberately in many instances in attempts
to increase fish production. However, there have been
enough cases of inadvertent fertilization resulting in
undesirable changes in natural waters that nutrient en-
richment is now recognized as one of our most serious
water quality problems.
Most of the unfavorable consequences of eutrophication
are not direct effects of the nutrients themselves,
but side effects of resultant excessive growths of
aquatic plants (Lee 1970). Excessive algae in drinking
water supplies can increase the clogging rate of sand
filters, increase the chlorine demand of the water, and
contribute color, tastes, and odors to the finished
water. In recreational lakes algal accumulations can
-------
impair aesthetic qualities as well as cause deterior-
ation of fisheries. In the long run eutrophication
accelerates the rate at which lakes fill up with sedi-
ment, thereby hastening their extinction.
NUTRIENT MANAGEMENT
The pressure for corrective action has become sufficient-
ly intense for government agencies to become involved
(e.g. U.S. Congress 1970, Bueltman e£ aa. 1969), and
it has been encouraged by recent reports of the rapid
improvement of Lake Washington following elimination of
the inflow of sewage treatment plant effluents (e.g.
Edmondson 1970). In this case the complete effluents
were diverted to a different drainage basin, solving the
problem by exporting it rather than by instituting im-
proved methods of nutrient management.
In most cases effluents cannot be exported but must be
managed more effectively within a drainage basin. The
ultimate approach will most probably have to be re-
cycling of waste water into drinking water, without re-
leasing wastes into natural waters, but this is in the
future. The nutrient control method that is the most
compatible with contemporary wastewater management
practice, the foundation of which is the production and
discharge of an effluent, is removal of nutrients prior
to effluent release. This can be done by reducing the
levels of nutrients entering wastewaters, the approach
that is the basis of the current pressure for removal
of phosphates from detergents, and by improving waste-
water treatment to remove the nutrients that cannot be
initially excluded.
It is generally believed that not all of the nutrients
that enter natural waters as a result of human activities
are responsible for the resultant problems. This is be-
cause not all nutrients are present in natural waters in
the proportions required by plants. Natural supplies of
some nutrients are in great excess relative to supplies
of other nutrients, and their abundance is therefore un-
likely to be limiting to plant growth. Wastewater
inflows are more likely to enhance aquatic plant growth
by supplying nutrients that are naturally scarce than
nutrients that are naturally abundant, and it is logical
to conclude that removal of the naturally scarce nutrients
from wastewater should improve conditions in receiving
waters.
A further point that must be considered, however, is
10
-------
that there are many sources of nutrient enrichment
other than waste-waters, and some of them, such as
rainfall, are impossible to control. Consequently it
is not sufficient to identify the scarcest nutrient in
a given water body and remove it from wastewater in-
flows, but it is also necessary to compare the extent
to which the total inputs of different nutrients can
be reduced by removing them from wastewater.
An example of the integration of these two aspects of
enrichment, nutrient limitation and nutrient controlla-
bility, is provided by a study of eutrophication in the
Potomac estuary (Jaworski et_ al. 1971). Nitrogen was
found to be more influentiaT tTTan phosphorus in con-
trolling phytoplankton growth, but phosphorus input
could be reduced more effectively than nitrogen input
by improved waste treatment. Therefore in the proposal
for future nutrient management phosphorus removal from
effluents was emphasized more strongly than nitrogen
removal.
As long as the approach to nutrient management is to
be selective, emphasizing control of only the key sub-
stances, there is a need for methods that can be used
to evaluate each body of water for which nutrient con-
trol is desired. In order to alleviate existing prob-
lems techniques are needed for (1) identifying limiting
nutrients, (2) identifying nutrient sources, (3) iden-
tifying the substances in nutrient sources that are
stimulating growth, and (4) evaluating the effective-
ness of nutrient management practices. In order to pro-
tect aquatic communities from future problems methods
are needed for (1) evaluating the sensitivity of an
aquatic system to additional enrichment and (2) screen-
ing potential future environmental contaminants for
possible effects.
RESEARCH METHODS
Limiting nutrients can be identified with some degree of
confidence by investigating nutrient changes in the sys-
tem of interest. For example phosphorus, rather than
nitrogen, was identified as the key nutrient in the
eutrophication of Lake Washington (Edmondson 1970) be-
cause its decline subsequent to nutrient diversion was
more closely correlated with the decline in phytoplank-
ton biomass, whereas nitrogen was found to be more in-
fluential than phosphorus in the coastal waters off New
York City, where nitrogen levels decline more rapidly
than phosphorus levels with distance from the harbor
(Ryther and Dunstan 1971). In the former case, however,
11
-------
the limiting nutrient was identified after its input
had been curtailed.
Several techniques applicable to eutrophication studies
employ algae as analytical tools. One group of these
can be classed as biochemical assays, because they ex-
ploit various biochemical traits of algae to assess
their nutritional state. For example limitation of
growth by phosphate can be inferred from (1) high ac-
tivity of alkaline phosphatase enzymes that are pro-
duced by phosphate deficient cells and function by re-
leasing inorganic phosphate from organic compounds or
(2) essential absence of the intracellular stores of
surplus inorganic phosphates that are characteristic
of cells growing in an environment rich in phosphate
(Fitzgerald and Nelson 1966). In a similar way nitro-
gen limitation can be inferred from (1) high rates of
ammonia utilization in the dark (Fitzgerald 1968) or
(2) low activity of the nitrate reductase enzyme
(Eppley e_t_ al. 1969). However, these techniques are
difficultf~tcT~apply to natural diverse algal communities
because of the physiological variations that occur
among species. Thus a "high" enzyme level for one
species may be "low" for another species. Also, dif-
ferent species within a community may be limited by
different nutrients at the same time (Fitzgerald 1969),
precluding any conclusion about what is limiting the
community. Nonetheless they have been used successfully
to complement other types of measurements, such as in
the Potomac study (Jaworski et^ al. 1971) where high
rates of ammonia utilization inThe dark in conjunction
with low alkaline phosphatase activity and the presence
of stored phosphate in algal samples supported the
conclusion that nitrogen was limiting in the reaches
where ambient nitrogen levels were low.
A second group of approaches employing algae, which can
be applied to most of the questions that arise in eu-
trophication studies, consists of the productivity
bioassays, or enrichment experiments. In these techni-
ques algal growth in response to a treatment is measured,
and the methods vary according to the algae used (single
species cultures or samples of natural communities) and
the growth conditions (batch or continuous culture,
laboratory or in situ incubation). The Joint Industry-
Government Task~~Force on Eutrophication (Bueltman et^ al.
1969) has been developing a series of standardized "algal
assay" methods, concentrating so far on a bottle test
employing separate cultures of four standard species
grown under laboratory conditions (Environmental Pro-
tection Agency 1971). The procedural and statistical
12
-------
aspects of the method have been intensively studied,
but the problems of using the results to evaluate or
predict events in natural waters have not yet been
resolved.
The second method that is under study is similar to the
first, only it employs continuous rather than batch
culture conditions. The third method, however, differs
fundamentally from the other two in employing samples
of natural algal communities as the test organisms.
In enrichment experiments employing single species cul-
tures the results are easy to interpret because the only
difference between a treatment a.rd a control culture is
the nutrient treatment itself. However the results are
difficult to apply to natural waters where environmen-
tal conditions are different, cell densities are gen-
erally lower, and many different species are present
(Lund and Tailing 1957). This last difference is par-
ticularly important in interpreting negative results
of single species bioassays, and its importance relates
to the fact that nutrient requirements vary greatly
among different algal species (Chu 1942, 1943; Rodhe
1948). If a sample of lake water is filtered to remove
the natural algae, inoculated with a test species, and
spiked with a treatment substance, stimulation of
growth can be safely interpreted as an indication that
the test substance is biologically active and that it
could stimulate algal growth in the natural system. If
growth in the culture is not stimulated, can it be con-
cluded that growth would not be stimulated in the
natural system? Not necessarily, for some other species
present in the natural community could respond.
When samples of the natural community are enclosed in
containers, treated with nutrients, and incubated in. situ;
natural cell densities are maintained, numerous species
are present, and environmental conditions such as light
and temperature are fairly natural. Thus it seems that
results of this type of experiment would be much easier
to apply to the natural system than are the results of
single species laboratory tests. It is true that the
difference between a treatment and a control experimen-
tal unit in an i_n situ enrichment experiment is still
just the nutrient" treatment. However, there are dif-
ferences between the environment in the control unit
and the environment in the natural system that act as
"treatments," and that are not evaluated by the normal
statistical procedure of comparing control units with
treatment units. An in. situ enrichment experiment is a
nested procedure in which the true control is the
13
-------
natural water body, the first set of treatments, which
is applied similarly to all experimental units, is a
consequence of containment, and the second set of
treatments consists of the nutrients added according
to the experimental design. The magnitudes of the
effects of containment influence the validity of using
the responses to the nutrient treatments in evaluating
or predicting events in the natural system.
Although the in situ technique is not as "direct" an
approach to studying algal nutrition in natural systems
as it seems on the surface, it still has appeal mainly
because it employs the natural species assemblage in
the system being studied. Thus, for reasons discussed
above, it should be more sensitive in detecting treat-
ment effects than the single species laboratory method,
and more conservative in protecting natural waters.
Enrichment experiments employing samples of natural
phytoplankton communities have been used by many in-
vestigators to study nutrient limitation in natural
systems (e.g. Goldman 1960a, Biesinger 1967, Thomas
1964, Wetzel 1966, McLaren 1969, Kemmerer 1968,
Hutchinson 1941, Gachter 1968, Schelske and. Stoermer
1971, Powers e_t al_. 1972, Moss 1969, Hamilton 1969,
Tranter and Newell 1963, Menzel and Ryther 1961), but
few thorough evaluations of the methods have been per-
formed, and few guidelines for their application are
available. If an in_ situ technique is to be incor-
porated into the standard Algal Assay Procedure, sta-
tistical properties and containment effects need par-
ticular attention so that "the results can be applied
with judgment to field conditions" (Bueltman et al.
1969).
OBJECTIVES
The initial objective of this study was to develop a
method for performing in. situ enrichment experiments
with samples of natural lake phytoplankton. Statisti-
cal properties were to be evaluated, and effects of
containment were to be investigated.
The resulting method was to be used in a lake for three
interrelated studies. In the first of these a series
of similar experiments was to be performed and the re-
sults analyzed in an effort to evaluate (1) the ability
of the method to detect limiting nutrients and (2) its
predictive potential. In the second a potential en-
vironmental contaminant was to be evaluated for its
14
-------
ability to stimulate algal growth. In the third the
method was to be employed in an effort to interpret the
stimulatory effects of some natural nutrient mixtures
in terms of the effects of identifiable components.
It was hoped that from the results of these studies
guidelines for the utilization of in situ enrichment
experiments in the field of eutropKTcation could be
developed.
CHOICE OF VARIABLES
Identifying Limiting Nutrients and Evaluating Predic-
tive Potential
In most of the enrichment experiments performed by pre-
vious investigators nitrogen or phosphorus or both
have had stimulatory effects (e.g. Goldman and Carter
1965, Edmondson and Edmondson 1946-7, Menzel and Ryther
1961, McLaren 1969, Kalff 1971, Kemmerer 1968, Wetzel
1966, Hutchinson 1941, Gachter 1968, Lange 1971,
Thomas 1964, Powers e_t al. 1972). As there is little
doubt that these two elements have major roles in de-
termining the production, periodicity, and species
composition of primary producer populations (Lund 1965) ,
they were natural choices for use as variables in the
present study. Treatment additions were chosen to
cause ecologically reasonable increases in N and P
over the ambient lake levels.
In a number of enrichment experiments in which nitrogen
and phosphorus were tested independently and in mix-
tures, their effects in the mixtures were found to be
interdependent. For example in an experiment by
Hutchinson and Riley (Hutchinson 1941) phosphorus or
nitrogen treatments added to separate jugs of Linsley
Pond water stimulated phytoplankton growth, but stim-
ulation by a mixture of the two nutrients greatly ex-
ceeded the sum of the effects of the separate additions.
This was one of the first published examples of a
synergistic interaction between nitrogen and phosphorus,
exhibited on the community level. In a more recent
study by Goldman and Armstrong (1969) Lake Tahoe phy-
toplankton samples were treated with several combina-
tions of nitrate and phosphate levels. Maximum growth
of the dominant species, Fragilaria crotonensis,
occurred in samples receiving 20 yg/1 of nitrate-N and
2 yg/1 of phosphate-P. The same phosphorus treatment
added along with 10 yg/1 of nitrogen produced little or
no response. In this case the nitrogen-phosphorus
interaction was detected on the species level.
While these and other studies demonstrated that nitrogen
15
-------
and phosphorus treatments could interact synergistically
in stimulating algal growth they did not reveal the
mechanism of the interaction. Evidence of a direct
physiological link had appeared in some work by Ketchum
(1939) who found that the rate of phosphate absorption
by cultures of Nitzschia closterium was dependent upon
the concentration of nitratein the medium. A reverse
influence was suggested later by the discovery by Eppley
et al. (1969) that the activity of nitrate reductase
extracted from Ditylum brightwellii was dependent upon
the concentration of phosphate present.
From these two lines of evidence, enrichment experi-
ments and physiological studies, it was concluded that
interactions between nitrogen and phosphorus may occur
whenever nutrient mixtures containing available forms
of both of these elements enter natural waters. There-
fore it was decided to employ mixtures as well as in-
dependent additions of nitrogen and phosphorus as treat-
ments in the present study, so that interactions could
be detected.
Although nitrogen and phosphorus are undoubtedly impor-
tant contributors to the effects of natural nutrient
mixtures on algal growth, efforts to attribute the
stimulatory abilities of such mixtures entirely to
their nitrogen and phosphorus contents have generally
been unsuccessful. Rodhe (1958) reported that in exper-
iments involving additions of small amounts of phos-
phate, nitrate, and hypolimnetic water to samples of
natural phytoplankton communities lt+C uptake after 24
hours was increased by as much as 80-0 by the hypo-
limnetic water while N and P never caused increases
greater than 301. Goldman and Armstrong (1969) re-
ported similar results in experiments employing treat-
ments with N, P, and water from enriched streams enter-
ing Lake Tahoe: the stream samples produced more
stimulation than their N and P contents could explain.
In both of these studies it was hypothesized that un-
known organic compounds present in the nutrient mix-
tures could have been responsible for the growth stim-
ulation above that produced by N and P.
One mechanism by which certain organic compounds can
influence algal nutrition is chelation of micronutrients
(Saunders 1957), and it has been demonstrated that
chelating agents added with or without micronutrients
to samples of natural waters can stimulate algal growth
(Schelske 1960, Johnston 1964). In a study reported
by Wetzel (1966) additions of EDTA (ethylene diamine
tetraacetic acid) to water samples from marl lakes in
16
-------
Indiana interacted indirectly with added phosphate,
resulting in stimulation of algal productivity. The
EDTA presumably chelated excess cations that other-
wise would have rendered the phosphate unavailable by
precipitation. Because of the possible role of such
chelators in natural nutrient sources, and because of
the demonstrated effects of EDTA in enrichment experi-
ments it was decided to include EDTA as a third varia-
ble, in addition to N and P, in this phase of the
present study.
Testing a Potential Environmental Contaminant
In recognition of the necessity for the complete or
partial removal of phosphates from their products, the
detergent industries have been evaluating potential al-
ternative builders for a number of years (U.S. Congress
1969). One compound, NTA (nitrilotriacetic acid),
has been under serious study since the early 1960's
(Duthie 1972), and consequently has been under scrutiny
by government agencies in the U.S. and Canada as a
potential environmental contaminant.
Because of the potential importance of this compound,and
in consideration of the results of the experiments with
the related compound EDTA, it was decided in the spring
of 1970 to initiate experimentation with NTA. A sample
of trisodium nitrilotriacetate was obtained for this
purpose from R.N. Sturm of the Proctor and Gamble
Company, Research and Development Department, on May
4, 1970.
It has been estimated that under full utilization by
the detergent industry NTA would be produced at the
rate of over one billion pounds per year, and could en-
ter sewage treatment plants at levels of 8 mg/1
(Shumate e_t al_. 1970) to 20 mg/1 (Hamilton 1972).
Average levels in surface waters have been anticipated
to approximate .05 mg/1 (Sturm and Payne 1971), but
actual levels, of course, would vary widely depending
on proximity to sources.
The potential for NTA to exert an environmental impact
would depend on its biodegradability in sewage treat-
ment plants and in natural waters, and on the nature
of its degradation products. The compound was found by
Thompson and Duthie (1968) to be readily broken down
in sewage treatment plants, where it is claimed to de-
grade to C02, H20, and inorganic nitrogen compounds.
Bunch and Ettinger (1967) suggested that the rapid
17
-------
breakdown of NTA by sewage organisms was related to the
presence of only one nitrogen atom in the molecule,
since several other chelators, including EDTA, contain-
ing two or more nitrogens persisted much longer under
their test conditions. Biodegradation of NTA in Ohio
River water was studied by Thompson and Duthie(1968),
who found that after acclimatization periods of 8-12
days organisms in river water samples could remove
doses of NTA in 2-6 days. Similar results were obtained
by Warren and Malec (1972) for Detroit and Meramec
River waters.
Tests of the effects of NTA on algal growth by several
investigators have yielded a variety of results. Studies
employing single species cultures (Christie 1970,
Sturm and Payne 1971) indicated little or no effect of
NTA treatments. Studies employing natural species
assemblages (Mitchell 1971, Sakamoto 1971, Goldman 1972),
however, frequently detected stimulatory effects, dem-
onstrating the greater sensitivity of multispecies
communities used in testing potential pollutants. The
effects in these studies were attributed to enhancement
of the availability of iron chelated by the NTA. This
hypothesis was supported in the study by Sakamoto
(1971) , in which NTA added to lake water increased the
amount of iron passing through a filter.
Moreover, interest in studying the ability of NTA to
stimulate algal growth was by no means unique to this
investigation. Objective evaluation of its potential
environmental significance should be based on a large
amount of data, covering many different systems in many
different geographical regions. In this sense there is
no duplication of effort.
Interpreting Stimulation by a Natural Nutrient
Mixture
It is well established that sewage treatment plant
effluents characteristically contain biologically avail-
able forms of nitrogen and phosphorus (Mackenthun et al.
1964), organic compounds that have chelating proper^
ties (Vallentyne 1957) , and numerous other biologically
active substances. Therefore it was decided to conduct
a series of side experiments in which effects of known
nutrient mixtures and effects of sewage effluents could
be compared, permitting assessment of the contributions
of the components of the known mixtures to the stimula-
tory effects of the effluent samples.
Effluent samples for this purpose were obtained from
18
-------
the Ann Arbor, Michigan sewage treatment plant, a
secondary treatment plant employing the activated
sludge process. Samples were taken from a point
just preceding the chlorination unit.
STUDY SITES
Third Sister Lake
All but one of the enrichment experiments were per-
formed in Third Sister Lake, Washtenaw County,
Michigan. This lake has a surface area of 3850 sq
m (Eggleton 1931) and a maximum depth of 16.5 m,
and its drainage area is partly forested and partly
agricultural land. It stratifies sharply in the
summer, and the hypolimnetic oxygen becomes severely
depleted during this period. The phytoplankton
community is continually dominated by cryptophyte
flagellates and a mixture of species of green and
bluegreen algae. Further background data will appear
in the discussion of results.
Crystal Lake
One experiment was performed in Crystal Lake, Benzie
County, Michigan. With a surface area of almost
4 x 10* sq m and a maximum depth of almost SO m
(Michigan Conservation Department 1940), this lake
provided a very different research environment from
that of Third Sister Lake. This lake is regarded as
a valuable recreational asset to the State of Michigan,
based largely on the high clarity of its water result-
ing from low phytoplankton productivity. The dominant
species at the time of the experiment were diatoms.
19
-------
SECTION III
METHODS
REQUIREMENTS OF THE FIELD METHOD
Statistical
Design
A review of the pertinent literature indicated that
interactions among the three variables N, P, and EDTA
could occur when they were applied in mixtures to ex-
perimental phytoplankton communities. Therefore a
statistical design capable of detecting interactions
as well as independent treatment effects was needed.
The complete factorial design, illustrated in Table 1
satisfied this requirement.
TABLE 1
2x2x2 REPLICATED FACTORIAL DESIGN
Variable
Nitrate-N
Phosphate-P
EDTA
No. of Replicates
Dose (M-g/l)
C
0
0
/~l
CL
500
r\
C^
)
5
0
2
500
2
£-
C
0
/"I,
)
500
2
^
^
5
0
2
500
2
The design presented in the table consists of three
variables, each at two treatment levels. Eight com-
binations of the different treatments are possible, and
all of these were included in the design. Each combin-
ation was applied to at least two experimental units,
which for our purposes were identical lake water samples
After responses of the experimental communities to the
treatment combinations were measured, the results were
subjected to analysis of variance procedures which
separated all possible individual or interactive effects
and indicated their statistical significance.
In order for analysis of variance techniques to be
applicable the data must satisfy three conditions: (1)
They must be normally distributed. (2) The variance
due to experimental error must be separable from that
due to treatment effects. (3) The error variances of
-------
the different treatment communities must be homogeneous.
Raw data which do not possess these properties can be
changed to other forms such as logarithms or square
roots to make them suitable for analysis (Barnes 1952).
In the present study, the separability of error and
treatment variances was assured by providing duplicate
experimental units for each treatment combination
(Table 1). The raw data from each experiment were tested
for normality and variance homogeneity by plotting the
standard deviations within the pairs of treatment
replicates versus the treatment means. The corner test
(Walker and Lev 1953) was performed on each plot to
check for dependence of the standard deviation on the
mean. Normally distributed data exhibit no relation-
ship between the standard deviation and the mean, so
when the plot showed this lack of relationship transfor-
mation was regarded as unnecessary. When a relation-
ship was observed,log transformation was performed to
eliminate it.
Computer facilities at the University of Michigan and
at the Virginia Institute of Marine Science were utilized
in the analysis of the data.
Incubation of Experimental Communities
Container. Once the statistical design had been selec-
ted the physical problems of applying it to natural
phytoplankton communities were confronted. I_n situ en-
richment experiments have been performed employing
vessels ranging in size from 500 ml flasks (Goldman
1960 a) to 4000 liter plastic bags (Schelske and
Stoermer 1971). Small containers are objectionable be-
cause they limit the size and number of subsamples that
can be taken and because of their high surface area to
volume ratios, which can exaggerate the effects of
containment. Extremely large containers, on the other
hand, are difficult to manipulate and, more important,
are difficult to mix. Thorough mixing of the contents
of a container is necessary if subsamples are to be
representative.
A further point that was considered in selecting a con-
tainer was the problem of distributing the experimental
community uniformly among the treatment vessels at the
beginning of an experiment. Uniform distribution was
necessary in order to assure that treatment effects
would not be obscured by initial discrepancies. It was
concluded from a study by Jackson and Bender (1964)
that this could be accomplished best by initially
22
-------
isolating, in a single tank, all of the lake water to
be used. The contents of the tank could then be mixed
continuously while the treatment vessels were filled
from it.
The experimental vessels selected were 19-liter (5 gal-
lon) Pyrex(K) glass jugs. They were large enough to
permit repeated subsampling and to, hopefully, avoid or
retard the development of at least some containment
effects. They \^ere small enough to be conveniently
handled and to permit a large number of them to be
filled from a reasonable sized tank. Choice of the jugs
was strengthened by a report by Abbott (1966) who con-
cluded that his set of 18 jugs exhibited sufficient sta-
tistical replicability for use in aquatic research.
Depth. In previous studies experimental vessels have
been incubated at the lake surface (Goldman 1960 b,
Biesinger 1967) as well as at various depths in the
water column. For the present study the lake surface
was rejected for incubation because of reports in the
literature (e.g. Edmondson 1956, Wetzel 1966) of de-
pression of photosynthesis by surface light intensities.
This phenomenon was subsequently demonstrated for Third
Sister Lake phytoplankton on numerous occasions.
The alternative was subsurface incubation at some depth
in the epilimnion. It was desirable to operate in the
epilimnion because it is in this layer that most of the
photosynthetic activity occurs in a natural lake. Fur-
thermore there is a depth interval in this layer within
which the rate of photosynthesis is maximal, or light
saturated, during most of the day, and it is there that
nutrient deficiencies are most likely to develop.
Assuming that nutrient deficient algae would be the most
responsive to experimental nutrient treatments, it was
decided that the zone of light saturation would be op-
timal for incubation.
For the initial experiments the depth of the shoulders
of the jugs was set at 1 m below the lake surface. This
choice was based on a statement by Edmondson (1956) that
the maximum rate of photosynthesis will lie between 0,3
and 1.7 m in a "least clear" lake and between 2.0 and
6.4 m in a "most clear" lake. It was estimated that
Third Sister Lake was closer to "least" than to "most"
clear.
Subsequent 14C productivity profiles confirmed that 1 m
was \vithin the zone of light saturation in Third Sister
Lake on most clear days. In 1969 profiles were
23
-------
determined from measurements made at 1 m intervals,
and the maxima usually appeared near 1 m depth, with
depression of uptake at the surface (Figure 1). In 1970
four near-surface profiles were measured in order to
more clearly define the regions of light inhibition and
saturation (Figure 2), Samples were taken from jugs
suspended at 1 m depth and were incubated at 20 cm in-
tervals from the surface to 2 m, for 4 hours bracketing
midday. Except for one day that was extremely overcast
the depth interval occupied by the jugs was within the
zone of light saturation. Surface depression occurred
on all days.
In 1971 a series of 5 productivity profiles was measured
on one day to delineate the depth zones of light inhibi-
tion, saturation, and limitation, and to indicate the
variations of these phenomena with time. Figure 3 pre-
sents the temperature and percent light transmittance
profiles for that day. Light penetration was measured
with a submarine photometer (G.M. Manufacturing and In-
strument Company, model 268 WA 310). Figure 4 shows the
productivity profiles plotted separately for each time
interval of incubation. In Figure 5 productivity at
each depth is plotted as a function of time and, con-
currently, of incident surface light intensity. Light
intensity was measured with a Solar Radiation Recorder
(Weather Measure Corporation, model R401), and the value
plotted in each time interval is the maximum recorded
during that interval. Light inhibition is evident at
the surface, where the shape of the productivity distri-
bution in Figure 5 approximates the mirror image of the
incident light curve. Light saturation during the cen-
tral 3 time intervals is evident at 1 m and 2 m, while
light limitation, indicated by productivity distributions
that closely follow the light intensity curve, shows up
at 4 m and below.
Response Measurements
Method. The final major question to be resolved concerned
the types and frequencies of response measurements to be
made. The actual response of interest was algal growth,
but this is difficult to measure directly at cell den-
sities characteristic of natural waters (Lund and Tailing
1957). Consequently, numerous indirect methods have
been employed, including chlorophyll changes (Hutchinson
1941) , oxygen production (Edmondson § Edmondson 1946) ,
ll|C uptake rates (Goldman 1960a, 1960b, 1964; Goldman
and Wetzel 1963; Goldman and Carter 1965; Schelske 1960;
Biesinger 1967; Wetzel 1966), and combinations of these
(McLaren 1969: chlorophyll and 14C uptake, Kemmerer
24
-------
PERCENT OF MAXIMUM
20 40 60 80
100
o.
bJ
O
13
15
CI4 UPTAKE, 6-21-69,% OF MAXIMUM
--LIGHT INTENSITY, 7-3-69,% OF
SURFACE INTENSITY
Figure 1. Representative Summer Photosynthesis and Light
Intensity Profiles in Third Sister Lake, 1969
25
-------
0
20-
40-
6O-
80-
o
UJ
0 140-
160-
180-^
200-
mg C/(m3 x 4hr.J
10 20 3O 40 50
5-12-70
overcast
10 20 30 40
50
\J
20-
40-
60-
~ 80-
E
-100-
X
£ 120-
UJ
°I40-
»60-
I8O-
200-
"^
\
5-27-70
partly cloudy
/
> -p
j
I8
mg C/(m3x 4hr.J
0
20-
40 -I
60-
-80-
u
-100-
t
UJ
°I40-
160-
180-
20O-
mg C/( msx 4 hr }
JO 20 30 4O 50
5-21-70
clear
j
U
6
s
30 4O 50
6-19-70
partly cloudy
tng C/ (nr»3x 4hr.)
Figure 2. Productivity Profiles, Third Sister Lake, 1970
26
-------
LIGHT (Percent Transmittance)
10 20 30 40 50 60 70 80 90 100
i i i i i i i i i
II -
12-
13-
14-
15-
16'
o
Light
O Temperature
4 5 6 7 8 9 10 II 12 13 14 15
TEMPERATURE (°C)
Figure 3. Light and Temperature Profiles, Third Sister
Lake, 5-6-71
27
-------
C14Productivity (mg C/ (mjx2.5hr.))
10
K)
00
2030
T
| 0600-0830
1600-1830
TIME INTERVAL
Figure 4. Productivity Profiles, Third Sister Lake, 5-6-71
-------
El.2
Z o .8-1
X S .4H
o i
15-
5-
25-
7 15-
ir>
oj 5
X
J l5
» 5H
E
15
5-
o
3
-o
O
25-
15-
5-
5-
TIME
0600-0830 10830-1100
INTERVAL
[ 1100- 1330 | 1330-1600
| 1600-1630 |
Om
I
m
2m
I
i
S i! ^_
I
i
4m
I.
I
6m
I
1
I
I
1
., fi
8m
Figure 5. Productivity and Incident Light Levels,
Third Sister Lake, 5-6-71
29
-------
1968: chlorophyll and oxygen production, Gachter 1968
and Schelske and Storermer 1971: lkC uptake and
species changes). The obvious preference for the lt+C
method is attributable to its rapidity and sensitivity
to small changes in population size or photosynthetic
activity. In some instances responses measured by
14C uptake differences have been detected within min-
utes or hours after nutrient addition--much more rap-
idly than they could appear as significant increases in
cell numbers (Goldman 1960b, Wetzel 1966). However,
extension of these immediate responses to prediction
of actual population increase is risky since the sub-
stances added may give only transitory stimulation to
the photosynthetic process (Dugdale 1967). Also,
measurement of only the immediate responses in terms
of 1LlC uptake precludes detection of responses that
might appear after a lag period, such as those observed
by Menzel et al. (1963).
Schedule. Because of the large designs to be used in
thisstudy and the desire to detect small responses to
low level nutrient additions the 14C method of response
measurement was the most attractive. Considering that
transitory effects or lag periods might arise in the
responses it was decided to conduct each experiment for
several days so that a series of l^C measurements could
be performed. To make certain that differences in
these measurements actually meant differences in pro-
duction initial and final plankton counts for repre-
sentative experimental units were to be performed.
In skeleton form the three components of the method at
this point can be summarized as: (1) a factorial
statistical design, (2) iji situ incubation of experi-
mental communities in 19-liter jugs at 1 m depth and
(3) measurement of responses by ^C uptake and plankton
examination techniques. The next section will examine
in detail the development of specific procedures. This
development process was inseparable from the evaluation
of experimental results, and data will be drawn freely
from specific experiments when needed.
DEVELOPMENT OF FIELD PROCEDURES
Exp e jr imejn t ^1: June 1968
Objectives. From the many literature accounts of en-
richment experiments the only generalization that
could be drawn was that an added nutrient was likely
to cause a detectable response. This response would
occur if some segment of the phytoplankton community
30
-------
TABLE 2
DESIGN OF EXPERIMENT 1
Variable
Nitrate-N
Phosphate-P
EDTA
No. of Replicates
Dose (M-g/l)
0
0
0
3
100
20
500
1
2500
1
50
500
1
2500
1
250
20
500
1
2500
T_
^
>
500
1
0
2500
1
found the enriched environment more favorable than the
control environment for growth, and would be detected
if the measurements were made at the appropriate time.
Thus the first experiment was approached with the ex-
pectation of answering the following basic questions:
(1) Was the proposed method capable of detecting re-
sponses of Third Sister Lake phytoplankton communities
to low level additions of nitrate, phosphate, and EDTA?
(2) Did the response pattern vary from day to day? (3)
Did the nutrients elicit responses independently or in-
teractively? (4) Did replicate experimental units be-
have identically?
Design. Since the first question was the most critical
It was decided that in this initial trial it was more
important to cover a range of doses, extending upward
to higher concentrations than were to be employed rou-
tinely, than to satisfy the statistical requirement of
full replication. Thus if the high level treatments
caused responses and the low level treatments did not,
it would be concluded that the proposed low level doses
were insufficient rather than that the nutrient variables
themselves were insignificant in Third Sister Lake.
Accordingly, the design presented in Table 2 was chosen
for experiment 1.
Eight jugs were to receive different nutrient combina-
tions, while three control jugs were to be included to
provide base line measures of productivity and of rep-
licability among jugs receiving identical treatments.
Low and high level doses of nitrate and phosphate were
intended to be approximately two and five times the am-
bient lake levels. The EDTA levels were intended to
fall within the range of dissolved organic matter in
lakes quoted by Hutchinson (1957, p. 883).
31
-------
Procedures. The experiment was begun on June 12, 1968.
Lake water was pumped from 1 m depth into a rectangular
75 gallon (284 liter) translucent polyethylene tank un-
til about 250 liters were accumulated. The contents of
the tank were then mixed continuously with a wooden
paddle while the water was dispensed into the eleven
jugs. The nutrient solutions were added in small volumes
by pipetting into each jug before it was completely full,
so that the turbulence accompanying the final lake xvater
addition would provide thorough mixing. A float con-
sisting of a 4 foot length of 2 x 2 with a one gallon
plastic jug tied to each end was then attached to each
jug to support it so that its shoulders were 1 m below
the lake surface.
Carbon-14 uptake measurements were made on June 12, 13,
15, and 19. Sampling was accomplished by pulling each
jug from the water, shaking it, and pouring about 1.2
liters into a dispensing funnel. While the contents
of the funnel were stirred, four 300 ml BOD bottles (con-
sisting of two light bottles, one dark bottle, and one
chemical sample bottle) were filled from it. Each lkC
incubation bottle received the tracer in an injection
of 2 ml of solution, prepared with an activity of approx-
imately 1 microcurie per ml and stored in a sealed
ampoule. After injection the bottles were placed on
their sides in two racks, each having a capacity of 18
bottles, and the racks were suspended at 1 m depth for
4 hours, from 1000 to 1400. While the productivity
samples were incubating the chemical samples ivere ana-
lyzed for pH and total alkalinity, so that 14C counts
could be converted to carbon fixed at a later date. At
the end of incubation the productivity bottles were
placed in the dark and duplicate 100 ml volumes from
each bottle were filtered under suction through separate
47 mm, ,45y pore size membrane filters (Millipore Corp.).
The filters were rinsed with small volumes of distilled
water, dried, glued to planchets, and counted for dup-
licate 2 minute periods in a proportional, gas flow,
ultrathin window counter (Beckman-Sharp Laboratories
Low Beta II). After the last sampling run the jugs were
emptied and acid cleaned to remove all organic matter
and adsorbed nutrients so that no carryover would affect
the next experiment.
To prepare the results for analysis the raw lkC counts
for each bottle were averaged, and the dark bottle
average subtracted from each of the two corresponding
light bottle averages. This yielded two independent ob-
servations (bottle means) for each jug for each sampling
day. The means (jug means) and standard deviations of
-------
TABLE 3
EXPERIMENT 1 (June 1968): RAW DATA
(JUG MEANS AM) STANDARD DEVIATIONS
IN COUNTS PER MINUTE)
Treatment Jugs
Date
6-12
6-13
6-15
6-19
Date
6-12
6-13
6-15
6-19
NOn -N
lOO^g/1 *
s
250ug/l z
s
100ng/l *
, T
250ng/l
s
100ng/l T
s
250UR/1 *
s
100|ig/l *
/ *"
S
Control A
x- 516.3
s 4-525
* 417-3
s 25.03
x 303.1
s 39.60
T 167.2
s 17.96
POI.-P:
EDTA:
0.5mg/l
552.3
88.10
415.2
59-^7
637.1
71.70
667.0
7.283
1120.
129.0
1127.
29.91
837-6
78.49
1885.
156.2
Control
507.7
31.11
204.8
107.1
4l4.0
120.9
140.4
170.9
20 i-ig/1
EDTA:
2.5mg/l
620.0
21.85
559-6
66.33
548.8
562.0
145.0
1276.
22.13
1240.
216.1
1223
55-86
l48o.
138.6
POU-P:
EDTA:
0.5mg/l
555-2
59.54
517.3
26.94
654.8
95-81
754.2
4o.i6
851.2
157-4
1665.
119.4
948.8
132.8
1716.
52.33
50 M-g/1
EDTA:
2.5rr£/l
591-2
57.06
370.0
39-24
625.3
16.55
693.8
11.67
1013.
60.88
2166.
3-o4l
1031.
45.25
1858.
269.5
B Control C Grand Control
425.4
78.70
385.0
88.18
637.2
30.41
240.2
53-25
Mean
483.2
50.15
335-7
114.5
451-4
170.2
182.6
51.66
33
-------
these data appear in Table 3. The two bottle means
were independent observations from a jug, but they
were not independent observations from a treatment
since they both came from a single treatment vessel.
Fully independent observations, suitable for variance
analysis in which all treatment effects could be re-
solved, would have to come from two different jugs
receiving the same treatment. Nonetheless, for the
purposes of this first trial it was decided that bottle
means would be subjected to analysis of variance so
that some separation of treatment effects could be
achieved. A corner test of the jug means and standard
deviations indicated that no transformation was neces-
sary.
Results. The results of the data analysis will be in-
cluded in a later section, and only the general features
of the response pattern will be described here. It is
obvious from Table 3 that the phytoplankton communities
responded strongly to the nutrients added, even at the
lower levels. The response was not immediate, however,
but built gradually to a maximum 3 to 7 days from the
start of incubation. In general replicate samples from
the same jug exhibited similar uptake rates although
some pairs differed from one another by several hundred
counts per minute. The control means tended to differ
widely. On the first three sampling days the standard
deviation associated with the grand mean of the three
control jugs exceeded most of the standard deviations
between the bottle means, implying poor between-jug
replicability relative to between-sample replicability.
The results were encouraging and enlightening. The
response pattern showed that the nutrients employed were
highly stimulatory, at the lower as well as the higher
treatment levels. The lt+C method of response measure-
ment was sufficiently sensitive to detect this stimula-
tion. The pattern changed with time and confirmed the
need for several lt*C runs during each experiment. Fi-
nally, the observations on replicability between sam-
ples and between jugs called for complete replication of
future experiments, with at least two jugs for each
treatment, and secondarily for refinement of sampling
methods. The latter process extended into a continuing
effort to identify and remove sources of extraneous
variance from all parts of the method.
The importance of extraneous variance can be appreciated
by briefly considering the significance test employed
in analysis of variance. A series of calculations re-
sults in a set of mean squares, one for each treatment
34
-------
and one representing the component of the total
variance contributed by experimental error. Ratios
are then calculated, each with a treatment mean square
as the numerator and the error mean square as the de-
nominator, and the significance of a treatment effect
depends upon the size of its ratio. The ability of the
method to detect subtle treatment effects, then, de-
pends just as much upon the size of the denominator as
it does upon the size of the numerator, so that any pro-
cedural refinements that can reduce the experimental
error will enhance the sensitivity of the method. The
many manipulations involved in carrying out an enrich-
ment experiment provide many opportunities for the in-
troduction of error. The following discussion will
treat each source of error that was explored and will
conclude with a description of the final form of the
method employed in most of the experiments performed
after 1968.
Sources p_f Statistical Error
1UC Technique. The basic response measurement employed
waslkC uptake. Before other sources of error could
be identified a reliable method of preparing, storing,
calibrating, and introducing the tracer had to be de-
veloped.
The first procedure attempted involved the preparation
of ampoules containing 1 or 2 ml of a stock ll+C solution.
The stock solution was prepared by diluting with dis-
tilled water a commercial preparation obtained from
Nuclear-Chicago, so that the theoretical final activity
of the solution was 1 microcurie per ml. Nonlabelled
carrier was added (.35 g NaHCOs/1) to reduce the pro-
bability of escape of 1^C02 molecules, and the pH was
adjusted to 9 so that virtually no free C02 would be
present. Ampoules were filled by pipetting from the
final solution, and after flame sealing were autoclaved
to prevent bacterial growth (Vollenweider 1969, p. 54).
Submersion of the ampoules in a solution of II methylene
blue during autoclaving permitted detection of leaks
(Steemann-Nielsen 1958). These ampoules were used in
three experiments before a sample of them was tested
for uniformity, by scintillation counting.
The testing procedure involved diluting the contents of
an ampoule to 100 ml with pH 9 buffer, adding .5 ml of
the diluted solution to a scintillation medium consist-
ing of 7 ml of scint toluene and 2.5 ml of NCS, Qy and
counting for several 2 minute periods in a scintillation
counter (Packard Tri-Carb Liquid Scintillation Spec-
trometer, Model 3375). Table 4 shows the results for
35
-------
TABLE 4
STANDARDIZATION OF llfC AMPOULES
BY SCINTILLATION COUNTING
Batch
1
2
3
4
Ampoule Batch Standard 95% Confidence
mean (cpm) mean (cpm) Deviation Interval
(423.7)
14382.0 13760.6 878.7 5866.4-21654.8
13139-3
10899.2 9978. if 1361.1+ 7812.4-12144.4
11020.0
807^.7
9919.8
3359-2 3370.1 271.6 2695-4-4044.8
3104.1
3647.0
2168.8 2325.9 302.1 1575.5-3076.3
2674.2
2134.6
cpm = counts per minute
TABLE 5
COMPARISON OF REPLICABILITY OF AMPOULES AM) SYRIKGE SHOTS
Ampoules
Syringe
Shots
Ampoule
Batch Mean (cpm)
4 2976.2
2191.6
3790.4
1393-8
Group Shot
Mean (cpm)
1 1850.0
1880.0
1833-2
1962.2
2 1935-5
1847.2
1969.8
1945.0
Batch
Mean (x)
2588.0
Group
Mean (x)
1881.4
1924.4
Std. Dev.
(s)
1029.5
Std. Dev.
(s)
57-3
53.4
95"£ Confidence
Interval
949.9-4226.!
95^ Confidence
Interval
1790.4-1971.4
1839.^-2009.4
36
-------
Q
O
X
tu A
l-
o
I>
O
O c
o: S
Figure 6.
A*l ML AMPOULE
S = SYRINGE
\-
-X-
-X-
900 1300 1700 2100 2500 2900 3300 37OO 4100
SCINTILLATION COUNTS (CPM)
Comparison of Means and 95% Confidence Intervals
for Batches of 1 ml Ampoules and Series of 1 ml
Shots from an Automatic Syringe
37
-------
the four ampoule batches prepared. The great variability
within batches, summarized by the 951 confidence inter-
vals, indicated that 11+C addition error was probably
a major source of between-sample variance. Thus it was
decided to test an alternative procedure: injecting
into the sample bottles directly from the stock solu-
tion. Table 5 contrasts scintillation counts for a batch
of 4 ampoules with counts for 2 series of 4-1 ml shots
from an automatic syringe, diluted and counted in the
same way. The second group of shots was taken four days
after the first group, from the same stock, and the re-
sults indicated that stability of a bulk stock, at least
for a short period, was excellent. Figure 6 compares
the confidence intervals for the syringe shots with
those for two batches of ampoules.
It was concluded that direct injection into samples from
a stock solution involved less chance of introducing
experimental error than did the use of ampoules. Other
investigators report precisely the opposite finding
(Doty and Oguri 1958). Fogg (1958), however, reported
that he used a stock solution and maintained it by pass-
ing it through a membrane filter and determining its
activity before each use. Goldman (1960b) reported
using an automatic syringe.
The method that was chosen for routine operation in-
volved injection of 1 ml volumes of tracer from a stock
solution into the sample bottles, using an automatic
syringe fitted with a Teflon(R) needle. Prior to injec-
tion a 1 ml volume of lake water was displaced from
each sample bottle so that insertion of the stopper after
injection would not splash out any 14C. After the last
sample was injected a final shot was added to a volu-
metric flask containing 99 ml of distilled water. Three
3 ml aliquots of the diluted tracer were then pipetted
into scintillation vials containing 20 ml of a medium
prepared by dissolving 100 g of napthalene and 6 g of
PPO (2,5-diphenyloxazole) in 1 liter of dioxane. These
samples were counted for duplicate 5 minute periods in
the scintillation counter. It was determined from
counting samples of this type prepared from a solution
of known activity that 40290 cpm corresponded to 1 mi-
crocurie/ml of stock solution. Thus by scintillation
monitoring of the tracer used in each experiment it was
possible to determine its activity during each sampling
run. Figure 7 shows that activity varied only slightly
from run to run within a llfC batch.
Two other modifications were made in the lkC method to
reduce error. First, 125 ml Pyrex (^)sample bottles were
substituted for the 300 ml BOD bottles so that the
38
-------
Q.
O
(O
O
O
35000
34000
33000
32000
31000
30000
c.
O
§ 29000
_l
P 28000
z
g 27000
26000
25000
EXPT. 19
EXPT. 9
1
4567
TIME (DAYS)
8
10 II
Figure 7. Activity of * "* Stock Solutions as Monitored
by Scintillation Counts on Days of Use
39
-------
entire samples could be filtered. This was done to
reduce filtration error as well as the time required
for filtration. It should be noted that several
attempts were made to detect effects of the length of
time samples were stored before filtration (up to 3
hours between the first and the last sample in some
cases), but no effects were ever found. The samples
were held in closed boxes, without the addition of
preservative.
The second modification involved extending the counting
period to 10 minutes to improve counting accuracy. For
example, in the case of a sample counting at 1000 cpm,
the 0.95 error for a 10 minute count is 18 cpm compared
to 45 cpm for a 2 minute count (U.S. Dept. of HEW
1960, p. 131).
Rinsing of the filters with dilute HC1 to remove inor-
ganic ^^C absorbed in the filter and adsorbed to
detritus was considered, but was rejected on the basis
of a report by McAllister (1961) that the decontamin-
ation process was more likely to increase the error of
the measurement than to decrease it. Fogg (1958) had
preceded this finding by expressing his opinion that
rinsing filters with distilled water rather than IIC1
was "satisfactory in removing inorganic li4C, from
freshwater phytoplankton at least."
Loss of 14C from phytoplankton stored dry on filters
has been reported by Wallen and Geen (1968) . Since
most of the loss that they detected occurred within the
first 24 hours of storage, and most of our samples were
stored at least that long before counting, this source
of error was not assessed.
Carbon-14 results were converted to carbon fixed by the
equation of Saunders et al. (1962): P = I. x C x f, in
-Q
which P is photosynthesis in mg C per cubic meter, r is
uptake of radioactive carbon in counts per minute, R is
the total available radioactive carbon in counts per
minute, C is the total available inorganic carbon in
mg/m3, and f is the isotope correction factor. R is
further defined as microcuries of radioactivity used
x efficiency of counter x correction for Millipore ab-
sorption effect x disintegrations per minute per micro-
curie. Table 6 contains the values for these factors
and the final equation that was used in these studies.
During lkC runs the samples were usually incubated in
the lake for a period of 4 hours. Longer incubation
40
-------
TABLE o
CONVERSION OF CPM TO CARBOH FIXED
Conversion Equation
f
R
Micro-curies used
Counter efficiency
Millipore absorption factor
dpm/mic rocur ie
Final equation for Third
Sister Lake
p = - x C x f (Saunders et al.
R 1962)
Q
Photosynthesis in mg C/nr
cpm counted, if entire sample is
filtered
20.h x 103 mg C/m3 (Third Sister
Lake)
1.06
Total available radioactive carbon,
in cpm
Scintillation cpm
40290 cpm/microcurie
0.25
0.838
2.22 x 10
,6
P = r x
1873
Scint. cpm
periods were rejected on the basis of reports that 14C
uptake rates tend to decline due to bottle effects in
samples incubated for more than a few hours (Barnett
and Hirota 1967, Vollenweider and Nauwerck 1961). While
bottle effects were undesirable, it was felt that the
incubation period should be significantly longer than
the period required for sampling the jugs and adding the
llfC. Thus differences among sample productivities due
to differences in sampling time or injection time
would, hopefully, be overcome.
In order to test the influence of incubation time on 11+C
uptake rate a special experiment, experiment IS, was
performed in August 1969. The experiment was designed
to compare four exposure durations: 2.5, 5, 15, and 24
hours. This was accomplished by exposing four stimul-
taneous series of samples taken from the same plankton
41
-------
community periodically over the course of the day.
In addition to samples taken directly from the lake,
samples were also taken from jugs filled the previous
day and incubated in the lake. The first batch of
samples began incubation at 0600 on August 15, 1969.
This batch consisted of eight light bottles and four
dark bottles from each of two jugs and two lake stations,
48 bottles in all. All samples were taken from 1 m
depth and were incubated at that depth. At the end of
2.5 hours one set of bottles from each of the four
sampling stations was replaced by fresh samples, and
immediately filtered. The fresh samples were taken from
the same two lake stations, but from two previously
unsampled jugs. After 2.5 hours more the second batch
of 2.5 hour samples and the first batch of 5 hour samples
were similarly replaced. This process continued until
2100 so that, altogether, six 2.5 hour sets, three 5
hour sets, and one 15 hour set of samples were processed.
At 0600 the next day the 24 hour set was filtered. The
results are plotted in Figure 8.
Each bar is a summation of means, with its total length
representing the total amount of carbon fixed by a given
series of samples over the course of the day. There
were differences in uptake rates between series exposed
for different lengths of time, but there were also
large differences between samples taken directly from
the lake and samples taken from jugs. Significant
differences between jug and lake samples within expo-
sure periods are indicated.
Comparison of the two 2.5 hour series shows that fresh
lake samples were more productive than were jug samples
for 2 of the time intervals. Comparison of the two 5
hour series shows greater productivity by the jug sam-
ples for one time interval, and in the 15 and 24 hour
series lake and jug samples diverged even more widely.
Comparison of the 2.5, 5, and 15 hour lake series shows
a very great reduction in uptake rate with exposure
time, and all differences among these three series are
significant (.05 level, Tukey's Test). A similar com-
parison of the jug series shows a mild depression, and
none of the differences are significant.
The reduced sensitivity to bottle effects of the jug
samples relative to the open lake samples implies that
the phytoplankton communities that had been imprisoned
in the jugs since the previous day had become adapted
to the enclosed environment. They had already responded
to bottle effects, and had somewhat compensated for them.
Therefore samples from the jugs could tolerate longer
42
-------
6CH
50-
40-
10
6
\
O 30-
o>
E
§ 20-|
o
O
ol lo-
o
ot
JC
o
Q) (A
JC O>
O 3
M
O>
2.5 hr. 5 hr. 15 hr. 24 hr.
exposures exposures exposures exposures
U5 hr. total) (15 hr. total)
* Adjacent bars or segments of bars significantly
different at .05 level.
Figure 8. Productivity Series, Third Sister Lake, 8-15-69
43
-------
exposure periods in productivity bottles before ex-
periencing further bottle effects, than could fresh
lake samples. A more accurate interpretation of the
apparent adaptation to the jug environment will be
discussed in a later section.
In a similar experiment conducted in 1971 (Figure 9)
incubation periods of 1, 2, 3, 4?|5 and b hours were
compared. Little difference in ii+C uptake rates
snowed up among the different periods, except that
the rates for samples incubated only 1 hour were sig-
nificantly lower than the maximum [4 hour) rates (.05
level, Tukey's Test). Tiiis difference implies that
there was a lag period between the start of incubation
and the onset of lkC uptake. The only significant
difference between jug and lake samples was for the
1 hour time interval (.05 level), implying that the^
lag period may have been related to adaptation to the
bottle environment. As in experiment 15, jug samples
were preadapted to containment.
In processing the productivity data all counts were con-
verted to mg carbon fixed per m3 per 4 hour incubation
period, rather than per day, since as Vollenweider and
;\auwerck (1961) pointed out the distribution of photo-
synthesis in relation to time during a day is assymetric,
with a maximum in late morning. Thus multiplying the
result of a 4 hour incubation by an arbitrary factor
and calling the product a daily rate would merely intro-
duce error.
With the use of a reliable lkC technique other parts of
the overall method could be tested and improved.
Sampling Technique. The best method for subsampling
from the jugs into the 14C incubation bottles would be
the method which provided the best replicability be-
tween samples. The initial method of shaking and pour-
ing was undesirable because of the danger of accident
as much as because of its inherently sloppy nature. An
improved method was developed which was based upon
agitation of the jug contents with a plunger and re-
moval of samples by suction. The major objection to
this method was the insertion of foreign materials into
the jugs since Doty and Oguri (1958) reported that a
variety of materials including plywood, tygon, neoprene,
plastic hose material, and metal pumps depressed photo-
synthesis in samples which contacted them, while bottles,
buckets, and funnels made of plastic had no effect.
Thus it was attempted to assure the inertness of the
materials to be inserted, by constructing the plunger
44
-------
12-
0-
X
to
e
o
E
'>
o
o
w
Q.
8-
6-
4-
2-
SAMPLES FROM JUGS
SAMPLES FROM LAKE
i
4
2 3
INCUBATION PERIOD (hr.)
Figure 9. Productivity Series, Third Sister Lake, 6-3-71
45
-------
from the top of a plastic bottle and a broom handle
coated with epoxy paint, and by using a short length
of rubber tubing for sample removal.
Several versions of the suction method were compared
to determine which one provided the best replicability
(see Figure 10). In the first trial two methods were
compared. In the first method four bottles were filled
simultaneously from a suction apparatus consisting of
a suction pump and 4 liter trap on one end, the tube
and plunger combination on the other end, and a mani-
fold of tubes and stoppers in the middle. Water flowed
from the jug through the tube and into the bottles, with
overflov; accumulating in the trap. Sufficient water
was collected in the trap to fill a second series of
four bottles. These bottles were filled one by one by
dispensing from the trap in a manner similar to dis-
pensing from the filling funnel employed in the shaking
and pouring method. Two series of four bottles were
filled by each method, and 1UC uptake was measured by
the improved technique. Means and 95"o confidence limits
were calculated for each set of 4, and these are pre-
sented in the top graph of Figure 11, labelled Experi-
ment 4A. Simultaneous refers to the first method, and
indirect sequential refers to dispensing from the trap.
Neither method performed very well. The lack of over-
lap of the confidence intervals of the two methods was
interpreted to mean that overflowing of the simultaneous
bottles to fill the trap caused concentration of
plankton in those bottles, and consequent reduction in
the population available for filling the indirect sequen
tial bottles.
In the second trial (Experiment 4B) the simultaneous
method was replaced by the direct sequential method, in
which water flowed through the tube and into a single
bottle which was linked in turn by a second tube to
the trap. This time overflow was prevented. Two sets
of four bottles were filled sequentially in this way.
The indirect sequential method was again employed, but
with water flowing directly to the trap, and a third
method, the batch sequential method, was added. This
method was similar to the indirect sequential method
except that the lkC was added to the contents of the
trap before they were distributed among the bottles. As
Figure 11 shows this method worked poorly compared to
the indirect sequential method, which in turn was far
excelled by the direct sequential method. Consequently
in all subsequent experiments the direct sequential
sampling method was employed, with agitation of the con-
tents of the jug by plunging and prevention of overflow
46
-------
V
1"
f
?/
L
XP
i (pUMP|
SIMULTANEOUS
|
J
J
S
uc
1
1
\f
1 L
TRAP.
1 *
B Q Q Q
'UMP
INDIRECT SEQUENTIAL
I
^
1 (
o
JUG
I1
,
L
L rlL, ,11, rJU I
1 (1 II 11 ITRAP
t
PUMP
DIRECT SEQUENTIAL
Figure Id Sampling Methods Employed in Experiments 4A and 4B
47
-------
TABLE 7
EXPERIMENT 3 (8-7/8-10-68): RAW DATA
EFFECT OF EXPOSURE RACK OK ^C UPTAKE
(COUNTS PER MINUTE, JUG MEANS)
Mutrient
Date Treatment Rack 1 Rack 2 Rack 3
8-8
8-10
1
2
3
1
2
3
337-3
459-6
713.2
x 503-4
132.4
488.8
188.2
x 269.8
707.3
817.1
837.2
787.2
324.4
365.8
567.5
419.2
723.4
716.1
746.0
728.5
366.7
498.1
541.0
468.6
from the sample bottles.
Samp1e Incubation Technique. In the first three experi-
ments two or three separate racks, each with a capacity
of 18 BOD bottles, were employed for incubating the
ltfC samples at 1 m depth. In experiment 3 the distri-
bution of the bottles among the three racks was con-
trolled. It was found that consistent differences
developed between samples on different racks (Table 7) ,
presumably because their exposure periods and orienta-
tions toward the sun were slightly different. It was
concluded that one large rack capable of holding all
the sample bottles would remove a significant source
of extraneous variance. This rack was constructed and
employed in all subsequent experiments.
Jug Filling Technique. The 75 gallon (284 liter) rec-
tangular tank used in the jug filling process in the
first small experiments limited the size of the design
that could be employed to a 2 x 5 replicated factorial.
However, it was desired to perform larger experiments,
based on 2x2x2 or 2x3x2 replicated designs, so
this tank was replaced by a 150 gallon (568 liter)
cylindrical polyethylene tank. The larger tank was
first used in experiment 5, a 2 x 3 x 2 experiment, per-
formed in October and November of 1968. For mixing
4S
-------
H
5SIM.
IC
1.5.
CO
SI M. = SIMULTANEOUS
I.S. * INDIRECT SEQUENTIAL
-X-
I X 1
f . ~J( __ j
I--X-H
90 100 110 120 130 140 150 160 170 180 190
EXPT. 4A
z
3 1*8.
0.
s
< B.S.
BO
.0.
D.S.«DIRECT SEQUENTIAL
I.S.« INDIRECT SEQUENTIAL
B.S.* BATCH SEQUENTIAL
I \j>..._ i
1 /"V 1
1 X~H
|^ _ _ V/ ___ 1
I "7T 1
t* ^(- i
i y,_ i
140 ISO 160 170 180 190 200 210 220 230 240
EXPT. 4B' CI4 UPTAKE (CPM)
Figure 11. Comparison of Means and 95% Confidence Intervals
for Groups of Lake Water Samples Taken from a
Jug by Four Different Sampling Methods
49
-------
700,
6OO
500
400
I
o 300
200
100
EXPT. 5=10-28
Y« 644.6-9.946 X
r « -.800 »«
CORNER TEST«-23w
II 13 IS 17 19 21 23 25
JUG FILLING ORDER
900,
800
700
6OO
^500
Q.
°40Oi
3OO
20O
100
EXPT.5-IO-3O
Y« 768.4-IO.84 X
r*-.778 ««
CORNER TEST33«»
3 5 7 9 II 13 15 17 19 21 23 25
JUG FILLING ORDER
Figure 12. Effect of Nonuniform Jug Filling
in Experiment 5
50
-------
the tank during filling of the jugs a larger plunger
was constructed from a broom handle and several large
plastic bottles. It was thought that in a cylindri-
cal tank vertical currents had to be induced to keep
the phytoplankton in suspension. Unfortunately, the
plunger broke and was replaced by the wooden paddle
early in the filling process. The result of the con-
sequent inadequate mixing is presented in Figure 12,
in which 14C uptake during the first two sampling
runs is plotted versus jug filling order. The re-
gression slopes and corner tests are highly significant
indicating that the jugs filled early received more
plankton than did the jugs filled later. Since the
water was dispensed into the jugs via a tube passing
through the wall of the tank just above the bottom,
this nonuniform distribution of plankton was clearly
due to settling of organisms in spite of the mixing
effort.
Because of the closeness of the slopes of the two re-
gression lines subtraction of the first day uptake
values from the second day values removed the slope
and resulted in data that could be analyzed for treat-
ment effects (Figure 13). The second graph in this
figure shows that after several more days of incubation
the treatment effects had overcome the initial discrep-
ancies, the slope had disappeared, and these final data
could also be analyzed normally. Before the next ex-
periment a more sturdy plunger was constructed, and as
the plots in Figure 14 show jug filling problems were
no longer encountered.
Summary ojf Statistical Performance.
Most of the major sources of statistical error discussed
in the preceding sections were identified and reduced
during the first 5 experiments, performed in 1968. In
Table 8 the statistical performance of the method is
summarized for each experiment in terms of the 5% least
significant difference (Steel and Torrie 1960, p. 106)
divided by the control mean. When more than one pro-
ductivity run was performed in an experiment the small-
est ratio obtained is tabulated.
The ratio for experiment 1, .238, is smaller than for
the other 1968 experiments. This is because the experi-
ment 1 least significant difference was computed from
a data analysis performed on replicate sample values
from single jugs, rather than on replicate jug means.
Experiment 2 was the first experiment employing 2 sep-
arate jugs for each treatment combination, and its Isd
51
-------
300
,200
a
o 100
EXPT. 5'(IO-30MlO-28)
Y-123.8-.8947 X
r»-.!089
CORNER TEST--2
3 5 7 9 II 13 15 17 19 21 23 25
JUG FILLING ORDER
EXPT. 5'II-4
I2OO
HOG
1000
900
800
700
a 600
o
500
400
300
200
IOO
0
r -.0109
CORNER TEST + 1
. . *
, 35 7 9 II 13 15 17 19 21 23 28
JUG FILLING ORDER
Figure 13. Removal of Jug Filling Effect
52
-------
1200!
1100
I000|
900
800
700
a 600
o
500
400
300
EOO
IOO
0
EXPT. 6=4-25
Y« 519.7 + 6.816 X
r = .1721
CORNER TEST= + 6
9 II 13 15 17 19 21 23 25
JUG FILLING ORDER
EXPT.7'5-27
400
300
JL 200
o
100
O
Y« 184.7 + 3.531
r «.3775
CORNER TEST*
X
+ 3
I 35 7 9 II 13 15 17 <9
JUG FILLING ORDER
Figure 14. Subsequent Experiments: Results of
Improved Jug Filling Technique
53
-------
TABLE 8
EVALUATION OF SENSITIVITY OF METHOD
Experiment
Number
1
2
3
k
5
6
7
8
9
16
17
19
20
22
23
2k
25
29
30
31
32
Error df
8
8
9
6
12
8
8
8
8
8
8
8
h
8
10
8
12
8
8
9
k
ylo isd
x Control
.238
2.20
1.01
.358
-535
.298
.193
.151
.280
-239
.1^9
.091
.106
.130
.085
.135
.151
.Ik8
.110
.077
.126
.176
is consequently the highest obtained. The reduction in
the ratio from experiment 2 to experiment 3 is attribu-
table to the inclusion of lkC sample exposure rack as
a variable in the analysis, and consequent reduction in
error variance.
The greatest proportional reduction in statistical error
occurred between experiments 3 and 4. This was due to
several modifications in the method, including the sub-
stitution of a single 1<+C incubation rack for the three
racks employed previously, the use of a common stock
solution of J1*C rather than the ampoules, and use of
the batch sequential sampling method. The direct se-
quential sampling method was used for the first time
in experiment 5, but the jug filling problem more than
negated the benefit of this improvement.
The improved jug filling method was introduced in exper-
iment 6, and from that point on the statistical quality
of the method remained essentially constant. The slight
54
-------
improvement between experiments 6 and 7 was due partly
to substitution of 125 ml reagent bottles for the 500
ml BOD bottles used for the productivity samples, and
partly to the use of two aliquots of water to fill each
jug. In experiment 6 the jugs were completely filled
in succession, while in experiment 7 they were all filled
to half capacity, and then to full capacity. Beginning
with experiment 17 each jug was filled in thirds.
As a result of the various improvements introduced into
the method, after the initial developmental phase pro-
ductivity means exceeding the control means by about
15% routinely represented treatment effects significant
at the .05 level. In some experiments the sensitivity
limit was below 10% of the control productivity levels.
Further improvement in sensitivity could probably be
achieved by using more than 2 experimental units per
treatment combination, but the degree of improvement
possible would not justify the expense and effort in-
volved.
SUMMARY OF_ METHOD
The final form of the method, incorporating all improve-
ments, can be summarized as follows:
Setting up an Enrichment Experiment
The total volume of water to fill all the jugs required
was pumped from 1 m depth into a 150 gallon cylindrical
polyethylene tank. The contents of the tank were mixed
continuously with a Plexiglass (&) plunger while the jugs
were filled by gravity flow. Jugs were filled with
three volumes of water, each volume approximating one
third of the capacity of a jug. During the jug filling
process four samples were taken, two of which were
membrane filtered and frozen for later chemical analysis,
and two of which were preserved with Lugol's solution
(Saraceni and Ruggiu 1969, p. 7) for later plankton
counting. After the jugs had been filled, the randomly
assigned nutrient mixtures were added by pipetting, while
the jug contents were continuously mixed with a plastic
plunger. From this point until the jugs were suspended
at the incubation depth, they were shielded with black
polyethylene bags to prevent light shock.
ltfC Uptake Measurement
On at least two separate days during the experiment the
jugs were raised one by one and samples were removed by
55
-------
the direct sequential sampling method into one dark
and tvv'o light 125 ml glass stoppered bottles. Carbon-
14 solution was added by injecting a 1 ml aliquot via
an automatic syringe into each bottle, following re-
moval of 1 ml of lake water to prevent splashout with
reinsertion of the stopper. The bottles were incubated
on their sides on a plywood rack suspended in the lake
at 1 m depth, where they remained for 4 hours. After
incubation the entire samples were membrane filtered;
and the filters were rinsed with distilled water, dried,
glued to planchets, and counted for 10 minute periods
in a beta counter. The raw counts were converted to
mg C fixed per m3 per 4 hours, using a formula pre-
sented earlier and measurements of the activity of the
tracer obtained at time of use.
Background Environmental Measurements
During the course of an experiment measurements of
water temperature, using a thermistor unit, and Secchi
disk transparency were made, to supplement the 14C
data. Chemical analyses performed on the stored samples
included total dissolved phosphate, using persulfate
oxidation (U.S. Dept. Int. 1969) to liberate bound
phosphate and employing ascorbic acid as the reducing
agent (Murphy and Riley 1962), and dissolved nitrate,
employing the salicylate method (Schering 1931).
Periodic analyses of fresh samples including ortho as
well as total phosphate measurements, were performed
to verify the results for the stored samples. The in-
strument used for reading the per cent transmittance of
the" developed samples was a Beckman DK-2 spectrophoto-
meter, with a 10 cm path length. The limits of the
sensitivity of this instrument were 1 yg PCK-P/l and
10 vg N03-N/1.
Plankton Counting
The plankton samples were concentrated by settling and
were observed with an inverted microscope (Unitron,
model BN-13). Counts of the important phytoplankton
species (i.e., nannoplankters occurring in at least half
of the 400 power fields observed and larger organisms
that could be effectively enumerated in 200 or 100
power transects or in 100 power full scans) were made
for selected samples. Individuals were measured, and
cell numbers were converted to biomass based on calcu-
lated cell volumes and assumption of a density of 1 g/cc
for all species except diatoms, which were assigned
1.1 g/cc based on data presented by Hutchinson (1967,
p. 248). References used in species identification were
56
-------
Desikachary (1959), Huber-Pestalozzi (1938, 1941, 1942,
1961, 1968), Patrick and Reiner (1966), Prescott (1962),
and West and West (1908).
Statistical Analysis
Carbon-14 uptake data were tested for normality and
transformed if necessary to base 10 logarithms before
being subjected to analysis of variance. Correlation
analyses were performed to detect relationships between
differences in plankton counts and differences in ]lfC
uptake measurements obtained for control and high treat-
ment populations. This data analysis was intended to
detect treatment effects and to identify the algal
species that had experienced these effects.
57
-------
SECTION IV
EXPERIMENTS TO IDENTIFY LIMITING NUTRIENTS
AND EVALUATE PREDICTIVE POTENTIAL
DESIGNS
This series of 13 experiments was conducted in the
ice free seasons of 1968 and 1969, and ended with an
experiment in the spring of 1970. The experimental
designs are summarized in Table 9. In experiments
1-5 several designs were tested while the statistical
properties of the method were being explored. Experi
ments 6-19 were all of similar designs, based on a
2x2x2 factorial setup with 2 replicate experimen-
tal units per treatment combination. Experiment 20
combined a 2 x 2 factorial with a one way design.
Addition levels were varied somewhat from experiment
to experiment.
PRODUCTIVITY RESULTS: APPENDIX A
The tables in Appendix A summarize the treatments, pro-
ductivity results, and variance analyses for the 13
experiments. In experiments 1-5 extraneous sources of
variance tended to obscure the treatment effects.
Nonetheless, the major features of the response patterns
are discernible. Experiments 6-20 employed the im-
proved method and were more sensitive in detecting
treatment effects. Experiments 10-15 and 18 were in-
tended to investigate various features of the method,
and their results are discussed elsewhere.
INTERPRETATION OF_ RESPONSE PATTERNS
Stable versus Variable Patterns
Examination of the variance analyses (Appendix A) re-
veals that in some of the experiments the response
patterns developed immediately and remained relatively
unchanged in all of the 11+C runs conducted, while in
other experiments the response pattern changed signif-
icantly from the first to the last 14C run. Experi-
ment 8 provides an example of the first, "stable" type
of response pattern, and its productivity results are
plotted in Figure 15. The pattern consisted of a pri-
mary stimulatory effect of P that was enhanced by N,
and it appeared in all 3 sets of productivity
-------
TABLE 9
SUMMARY OF EXPERIMENTAL DESIGNS
EXPERIMENTS 1-20
Experiment Starting
Date
Variables
Treatment Addition
Design
1
2
3
U
5
6
6-12-68 N03-N
POij-P
EDTA
7-18-68 MO^-H
POij-P
EDTA
8-7-68 EDTA
9-19-68 WOk-N
Poly-P
10-27-68 NOo-I\T
PO,-P
*T
EDTA
U-23-69 NOo-N
POl^-P
EDTA
.100 mg/1
.250 mg/1
.020 mg/1
.050 mg/1
.500 mg/1
2.50 mg/1
0 mg/1
.100 mg/1
0 mg/1
.020 mg/1
0 mg/1
.500 mg/1
0 rag/1
.500 mg/1
2.5 mg/1
0 mg/1
.100 mg/1
0 mg/1
.050 mg/1
.100 mg/1
0 mg/1
.100 mg/1
0 Kg/1
.010 mg/1
.050 mg/1
0 rag/1
.500 mg/1
0 mg/1
.025 mg/1
0 mg/1
.005 mg/1
0 mg/1
.500 mg/1
2x2x2 Factorial,
unreplieated,
+ 3 control jugs
2x2x2 Factorial,
2 replicates
per cell
1 way design,
3 replicates
per cell
2x3 Factorial,
2 replicates
per cell
2x3x2 Factorial,
2 replicates
per cell
2x2x2 Factorial,
2 replicates
per cell
60
-------
TABLE 9 (Continued)
Experiment Starting Variables
Date Treatment Addition
7 5-26-69 KO--S
-
PO^-P
EDTA
8 6-^-69 I3H3-N
FOj^-P
EDTA
9 7-1-69 NOo-N
PO^-P
EDTA
16 8-21-69 NO-^-K
"
PO^-P
EDTA
17 9-13-6"9 NOo-N
j
PO,,-P
-t
EDTA
19 10-25-69 NOo-N
j
POh-P
EBTA
OQ ]J_OO_TQ >;Oi-K
j
POk-P
EDTA
0 mg/1
.025 mg/1
0 mg/1
.005 mg/1
0 mg/1
.500 mg/1
0 mg/1
.025 mg/1
0 mg/1
.005 mg/1
0 mg/1
. 500 mg/1
0 mg/1
.025 mg/1
0 mg/1
.005 .mg/1
0 mg/1
.050 mg/1
0 mg/1
.100 mg/1
0 mg/1
.003 mg/1
0 mg/1
.500 mg/1
0 mg/1
.025 mg/1
0 mg/1
.005 mg/1
0 mg/1
500 mg/1
0 mg/1
.025 mg/1
0 mg/1
.005 mg/1
0 mg/1
.500 mg/1
0 mg/1
.025 mg/1
0 mg/1
. 005 mg/1
0 mg/1
.382 mg/1
Design
2x2x2 Factorial,
2 replicates
per cell
2x2x2 Factorial,
2 replicates
per cell
2x2x2 Factorial,
2 replicates
per cell
2x2x2 Factorial,
2 replicates
per cell
2x2x2 Factorial,
2 replicates
per cell
2x2x2 Factorial,
2 replicates
per cell
2x2 Factorial,
2 replicates pel-
cell "for N and ?j
Independent
treatment of EDTA
61
-------
FIGURE 15
EXPERIMENT 8: TREATMENT EFFECTS
M6 C/(M3X4HR)
Source of
c
ftt
to
t
> 10 20 30 40 50 60
i i i i i i
NH3+P+C |
NH3*P |
NH3+C |
NH3 |
P4-C |
P |
C J
Control 1
vuriuiion naiio
NH3-N 22.80**
P 705.30**
C .003
NP 21.53**
NC .02
PC 1.16
NPC .03
NH3=25jug NH3-N/I
P = 5jugP04-P/l
C = .SmgEDTA/l
(O
o
i i i i i i
NHB*P*C ( NH3-N
NHS+P | £
NH3+C
NH3
NP
tl/*
NC
PC
P*C | NPC
p 1
C
Control
H.82*K
1 27.06* »
.08
10.18*
1.17
.43
1.50
UJ
u
CL
ffi
8 !
NH3+P+C 1
NH3+P |
NH3^C |
NH3 |
P+C |
p 1
c 1
Control |
NH3-N
P
C
NP
NC
PC
NPC
7.20 X
29.94**
.04
3.22
1.03
.86
1.70
(0
I
I
a>
**-Significant at the .01 level.
X -Significant ot the .05 level.
62
-------
FIGURE 16
EXPERIMENT 17: TREATMENT EFFECTS
MG C/(M3X4HR)
Source of
Vor lotion
c
) 5 10 15 20 25 30 3!
i i i i i I
N+P+C 1
N+P I
N+C |
N 1
P+C |
P 1
c I
Control |
N 67. 58* ail
PA9 a^»&
tc.w t/fltm
C 76.23**
NP 29.61 **
NC -4O
PC M8
NPC .37
N= 25jugN08-N/l
P s 5jug P04 -P/t
C= .5mgEDTA/l
i i i i i i i
N+P+C |
N+P I
N+C
N 1
P+C |
P 1
c 1
Control 1
N
P
C
NP
NC
PC
NPC
53.00 **
.004
14.13**
.001
3.43
1.74
1.27
z
o
o
u
UJ
(t
O)
<0
I
<£
i
0)
0>
-------
measurements. Experiment 17, in contrast, exemplifies
the type that varied with time (Figure 16). On the
first day of measurement a main effect of EDTA was
clearly evident, as was a synergistic interaction
between N and P. On the second day the EDTA effect
still appeared, but an independent N effect had replaced
the NP interaction. Finally, by the third day the EDTA
effect had disappeared leaving independent but weakly
expressed N and P effects.
In the case of a stable response pattern, interpre-
tation of the results is clear--the one set of effects
observed constitutes the response of the experimental
community to the treatments. When the pattern varies
with time, however, the nature of the true response is
uncertain and interpretation of the results is more
complicated. Brief discussions of two studies reported
in the literature will illustrate this point.
In the first study Biesinger (1967) applied a series
of micromrtrient treatments to samples of Alaskan lake
phytoplankton. Carbon-14 productivity measurements
made after 12 hours of incubation detected stimulation
by Li, Co, V, B, and Mn. After 24 hours of incubation
only Co, V, and Fe effects were still present, while
after 48 hours only the samples treated with Fe showed
an effect. The author concluded that many substances
added to'natural waters can increase 14C uptake in
short term experiments, and that longer term experiments
are necessary to reach adequate conclusions about
limiting factors.
In a study of nutrient limitation of Sargasso Sea phyto-
plankton, Menzel and Ryther (1961) detected stimulation
by Fe treatments, but the Fe effects lasted only 24
hours in the absence of added N and P. Still it was
concluded that Fe was the limiting nutrient. In further
experiments which lasted up to 9 days (Menzel e_t al.
1963) it was found that a mixed treatment of N^P, and
Fe produced a rapid response, but that after a lag period
a comparable response to N and P, added without Fe,
developed. Addition of Al to the N+P treatment also
accelerated the response, so it was concluded that the
metals somehow exerted catalytic effects on the samples
treated with N and P. In this study, as well as in the
work by Biesinger, the interpretation of results was
strongly influenced by the durations of the experiments.
1_'*C Uptake versus Growth
One other basic factor that contributes to the
64
-------
uncertainty in interpreting response patterns that are
expressed exclusively in terms of llfC uptake is that
ltfC uptake is not always directly related to growth.
This applies to long term as well as to transitory re-
sponses. In a study reported by Goldman and Armstrong
(1969), for instance, nitrate and phosphate treatments
were applied to samples of pelagic Lake Tahoe phyto-
plankton. In one experiment cell counts increased in
response to treatments with P alone, but 11+C uptake
declined below control levels. In another experiment
the reverse occurred: 14C uptake was stimulated to
levels 50% above controls, but cell counts did not ex-
ceed the control counts. If it is assumed that ll*C
uptake is used as a response variable in enrichment
experiments because it is a rapid and convenient way to
measure population growth, then the relationship be-
tween the two processes should be confirmed frequently.
Experimental Evaj^uat^on: Experiment 12
Questions Asked. In the present study two important
questions were posed that required answers before con-
clusions could be derived from the results. (1) Should
the early effects that subsequently disappeared be re-
garded as transitory stimulations of 14C uptake similar
to those encountered by Biesinger (1967), and therefore
be discounted in favor of the final patterns that de-
veloped? (2) Could productivity responses in general
be interpreted to signify growth responses? Some in-
sights into these questions have been provided by a
detailed analysis of the species growth responses that
occurred in a special experiment, experiment 12, con-
ducted in 1969.
Design and Procedures. This experiment was conducted
in CrystaT Lake, Benzie County, Michigan beginning July
22, 1969. The design is shown in Table 10 which in-
dicates the treatment additions, the number of replicate
jugs per cell, and the ambient nutrient levels.
The field procedures were the same as those employed in
the Third Sister Lake experiments. Responses were de-
tected on three separate days following the start of
the experiment by measuring carbon-14 productivity of
subsamples from the jugs by the standard technique.
Counts were made for the 15 most important species, and
the numerical results were converted to biomass esti-
mates by multiplying by volume and density estimates.
65
-------
TABLE 10
DESIGM OF EXPERIMENT 12
Variable
N03 - N
PO^ - P
EDTA
Number of
Replicates
Dose (jJ.g/1)
0
0
0
2
500
2
5
0
2
500
2
50
0
0
d.
500
2
5
0
2
500
^
Ambient concentrations NCn - M : 30 p.g/1
Total dissolved P : 6.5 M.g/1
Productivity Results versus Growth Responses. The re-
measurements appear in Figure
the clarity of this and of
suits of the productivity
17. In order to maximize
the other figures in this
significantly different (
the averages were plotted
they applied. On July 24
section, values that were in-
05 level) were averaged, and
for all treatments to which
two days after the start of
the experiment, the response pattern indicated stimu-
lation by EDTA alone but not by either N or P alone.
However, the mixture of N and P did cause stimulation,
and the statistical analysis revealed a significant
NP interaction. Three days later, on July 27, an in-
dependent P effect appeared, as did the NP
but the EDTA effect did not. Instead, two
interactions involving EDTA appeared. One
apparent blockage of the P effect, and the
enhancement of the effect of the NP mixture
interaction,
types of
of these was
other was
. This
last effect showed up more strongly on July 30, while
both of the two way interactions had disappeared leaving
the independent P effect. Thus this experiment belonged
to the second type discussed above, with three differ-
ent sets of productivity measurements yielding three
different response patterns.
Various elements of the productivity response patterns
can be elucidated by considering growth responses of
individual species within the experimental communities,
Figure 18. The response pattern of Synedra nana, one
of the two dominant species, embodies most of the fea-
tures shown in the productivity patterns: an indepen-
dent P effect, blockage of this effect by EDTA. and
enhancement of the effect by N + EDTA. The coefficient
of correlation between the biomass estimates for this
species and the productivity results on July 30 is .956.
66
-------
201
15-
10
5
0
X
Kl
E
2 On
I 5-
O
o. I OH
E
>-
h-
> 0
H
O
O
Q.
0-
0
JULY 24, 1969
i
i
JULY 27, 1969
Ill
I
JULY 30, 1969
i
I
I
N
NP PE NPE NE
Figure 17, Experiment 12, Productivity Response
Patterns
67
-------
lOOO-i
750-
500-
250-
o»
UJ
t-
CA
UJ
CO
<
o
QQ
50-
Synedro nono
Frggiloria
crotonensis
IOH
o
ton
o
IOH
0
2H
0
H
Synedro radians
Nitzschio $p.
HVVl
1
1
1
i
1
I
Achnonthas sp.
Synechocystig
oqualilus
Rhodomonas minuta
0-1 ESO CSSTC
I C N E P NP PE NPE NE
Figure I&. Experiment 12, Growth Response
Patterns I
68
-------
Fragilaria crotonensis, the second dominant species,
shows stimulation by P alone, enhancement of the P
effect by N, and blockage of the P effect by EDTA.
These are three of the features of the productivity
response pattern of July 27, and the correlation be-
tween the Fragilaria biomass estimates and the July
27 productivity results is .977.
If the growth patterns of Synedra nana and Fragilaria
crotonensis are pooled the correlation betwe~en bfom"as"s
and productivity on July 30 is increased to .96], and
all but one of the features of the productivity pattern
are accounted for. What remains is the response to
the PE mixture, that appeared on July 24 as well as
July 30. This can be covered by including the next
three species, Synedra radians, Nitzschia sp., and
Achnanthes sp., since for them the effeTcY of P was not
blocked by EDTA. Synedra radians and Nitzschia sp.
responded to P alone, with" no mollification oT~the effect
by either N or EDTA, while Achnanthes sp. responded
more strongly to the NP mixture.Addition of the bio-
mass estimates for these three species to the sums of
the estimates for Synedra nana and Fragilaria
crotonensis raises thecorrelation between biomass and
productivity on July 30 to .973.
These correlation analyses indicate that the relative
contributions of individual species to community pro-
ductivity changed with time. Thus on July 27 most of
the productivity responses could be attributed to
Fragilaria crotonensis, although the influence of
Synedra nana had begun to appear, at least in the re-
sponse" to the NPE mixture. By July 30 dominance of
the productivity pattern had shifted to Synedra nana,
but five species in all were required to account for
all of the important productivity results. Thus each
set of productivity measurements provided instantaneous
community response estimates integrated over the active
species, while the final biomass determinations provided
individual species response estimates integrated over
time. Since the day to day changes in the productivity
response patterns reflected genuine changes in the
growth activities of individual species, all of the pro-
ductivity patterns observed were legitimate components
of the community response to the nutrient treatments.
Thus in this experiment and in each Third Sister Lake
experiment, all of the productivity responses obtained,
regardless of their persistence, were included in
summarizing the results.
In experiment 12 there were two additional species,
Synechocystis aquatilus and Rhodomonas minuta (Figure 18) ,
69
-------
that responded positively to nutrient treatments.
Both of these responded only to the NP mixture. Eight
other species, shown in Figure 19, either responded
negatively to the jug environment, or negatively to
some of the nutrient treatments, or not at all. Most
of these were minor species, but three of them
Cyclotella ocellata, Cryptomonas ovata, and Cyclotella
stelligera, accounted for significant fractions of
community biomass.
The variety of response patterns observed on the species
level in this experiment exemplifies the ability of
nutrient enrichments to alter the species composition
of phytoplankton communities. Species shifts in response
to nutrient treatments have been observed in numerous
other enrichment studies (Menzel et al. 1963, Thomas
1964, Barlow et al. 1971, Gachter~T9FS", Schelske and
Stoermer 1972J7 and were encountered in the Third Sister
Lake series of experiments, to be discussed.
The complex response patterns exhibited by some of the
species in experiment 12, e.g. Synedra nana and
Fragilaria crotonensis for which positive and negative
interactions among treatments appeared, indicate that
more than one nutrient treatment can influence a given
species at a given time in a natural community. The
growth responses of these two species, as well as all of
the others, were totally dependent upon the treatment
additions of P, while the other two treatments acted
only to modify the P effects. In this sense P was the/
nutrient limiting species growth, and consequently
community productivity, in this experiment.
PRODUCTIVITY RESPONSE PATTERNS
Summary o_f Treatment Effects
Table 11 is an attempt to summarize the productivity
response patterns obtained in experiments 1-20. The
treatments listed under independent effects caused sig-
nificant stimulation of productivity in the presence or
absence of the other treatment substances. Independent
effect in this sense does not necessarily mean that
the treatments did not interact with other substances
in the lake water, since for EDTA chelation of ambient
trace metals was probably the basis for its effect.
When two or more substances had independent effects,
these effects were essentially additive in mixtures of
the substances. Interactions occurred when a treatment
substance that did not have an independent effect
70
-------
50-
25-
0
25-
0-
2-
0-
Cyclotella ocellata
UJ
0)
UJ
en
CO
O
m
0-
25
0-
10-
0
I-
0
5
0
I
1
1
II
I
1
s
Cryptomonos ovata
1
Microcvstis
oeruginoso
1
I
1
I
Anabaena spiroides
I
Cyclotello
stelligero
Glenodinium Borqei
Cryptomonos eroso
1
Pediostfum Boryonum
1
I 1
N
P NP PE NPE NE
Figure 19. Experiment 12, Growth Response
Patterns II
71
-------
TABLE 11
SUMMARY OF PRODUCTIVITY
RESPONSE PATTERNS,
EXPERIMENTS 1-20
Month: Apr May
1968 Expt:
Independent
effects
Interactions
Jun
1
N
EDTA
PN
Jill Aug Sep
23 k
P
EDTA
NP
Oct
5
NP
Notes
N, P EDTA
Not Not
Tested Tested
1969 Expt:
EDTA
Independent
effects
Interactions
Notes
1970 Expt: 20
Independent P
effects EDTA
7
P
HP
N EDTA
8
P
NP
9
N
P
EDTA
16
P
EDTA
17
P
EDTA
NP
19
P
NP EDTA
NH3-N
employed
TABLE 12
FREQUENCY OF RESPONSES,
EXPERIMENTS 1-20
Treatment No. of Expts. No. of Expts. No. of Expts. No. of Expts.
Tested with with with
Independent Interactions No Effect
Effects Only
N
P
EDTA
12
12
12
2
8
7
6
3
2
h
1
3
72
-------
enhanced the otherwise independent effect of a second
treatment (e.g. N in experiment 8, Figures 15 and 20),
or when two treatments that had no effect when applied
separately had an effect when applied in mixtures
(e.g. N and P in experiment 17, Figures 16 and 20).
Table 12 summarizes Table 11 indicating the overall im-
portance of each of the treatments. Of the three treat-
ment substances P stimulated productivity in the largest
number of experiments. In 8 experiments it acted in-
dependently, and in 3 its action depended upon the pres-
ence of added N. EDTA produced independent effects in
7 experiments, while N stimulated productivity only
twice in the absence of added P. Thus if the three
treatments are ranked according to their abilities to
stimulate primary productivity in the experimental sys-
tem the order is P, EDTA, and N.
The results of these 13 experiments could be pooled
without further analysis in order to predict productiv-
ity responses in future experiments. The prediction,
which might be called the "most probable response
pattern," would include an independent effect of EDTA
and an independent effect of P that would be enhanced
in the presence of added N. Examination of Table 11,
however, indicates that this theoretical pattern ac-
tually appeared in only one experiment, number 17, and
that substantial variation in response patterns occurred
from experiment to experiment.
Variations Among Experiments
Literature Examples. Variability seems to be a common
feature in the results of studies in which series of
experiments have been conducted on a single natural
system. In the most general sense Rodhe (1958) attri-
buted season to season changes in enrichment effects
to changes in environmental conditions. He observed
that nutrient treatments were most stimulatory during
the period of thermal stratification, when nutrient
availability to the photic zone was reduced, and least
stimulatory during the fall circulation. However, ex-
periment to experiment changes in enrichment effects
during a season have been more difficult to explain in
terms of environmental changes.
In one study Lange (1971) employed a procedure similar
to the PAAP technique (Bueltman e_t al_. 1969) to test
for limiting nutrients in water samples from Lake Erie.
Fifteen experiments were performed, most of them in a
biweekly series from April-October 1969. In each
73
-------
EXPERIMENT 8
17
25-
24-
~ 23-
x
H>
22-
21-
0 20-
o»
£ 19-
18-
17
EXPERIMENT 17
X
r-
Po
Figure 20. Examples of NP Interactions
74
-------
experiment four cultured algal species were grown
separately in filtered lakewater samples to which
numerous pure and mixed nutrient treatments were
added. Responses were determined by cell counts.
There were response variations among experiments with-
in species and among species within experiments, and
no obvious correlations between the responses and am-
bient nutrient levels emerged. Conclusions about the
relative roles of the different nutrients as limiting
factors in the la"ke were based upon the number of
experiments in which each one stimulated. The nutrient
that stimulated the most frequently was N (39/60 cases),
and it was concluded therefore that N was in adequate
supply in the smallest number of samples tested. Based
on these results the "most probable response" to a
single nutrient in future experiments would be to N,
but the probability would be only .65. One advantage
claimed for the PAAP bottle test over in situ tests is
reproducibility of results, since species composition
can be eliminated as a variable by employing the same
species in the same stage of growth in every experiment.
If the results of Lange (1971) are representative,
however, reproducibility seems to be an elusive quality
when using the culture technique as well as when using
the in situ technique.
Biological conditions have been invoked in some
studies, in attempting to account for varying response
patterns. Goldman and Wetzel (1963) reported that bio-
assay responses in Clear Lake, California related to
the overall activity levels of the phytoplankton pres-
ent. During periods of high primary productivity
treatments were stimulatory, and when productivity was
low no treatment effects occurred. In studies reported
by Goldman (1960b) and Powers e_t al. (1972) , direct
examination of the phytoplankton communities present
in different experiments revealed that the species com-
position changed from experiment to experiment. Changes
in the experimental response patterns were thought to
have related to the species shifts.
In attempting to interpret the results of the Third
Sister Lake series of enrichment experiments, an effort
will first be made to determine if the variability among
experiments was random or if gradual changes occurred.
If the changes were gradual, it seems more logical to
expect to relate them to changes in environmental condi-
tions or in the species present, than if the changes
were random.
75
-------
Experiment Number
19 -
1 7
M
1 £
±0 -
-
D
o
7
1
f.
c;
9
3 -
2_
1 -
7q
75 ,^>
,pq Tqn -7^
1 0 °5 50 ^7^
_?q .9^ .sn .7^ -50
50 75 ^75 .nn '?q -qn
1 .n .qn .75 .75 .00 .25 .50
.50 .50 .50 .25 .25 -50 .25 -50
,.50
50
.25
75
75
.25
.50
.50
.00
1
123^567
8 9 16 17 19
Experiment Number
Figure 21. Comparison of Productivity
Response Patterns, Experiments 1-20
76
-------
TABIIE 13
TESTS FOR SIMILARITIES AMONG "HEAR"
VERSUS "DISTANT" EXPERIMENTS
Experiment Sums of Comparison. Values
Near Distant
2
5
6
7
8
9
16
IT
19
Overall Sum:
3.00
3.00
1.75
2.00
2.00
1.75
2.50
2.50
2.00
20.50
1.25
1-25
1.75
3-00
3-00
75
1.75
2.25
2.00
17.00
ercomparison p_f_ Experiments . The comparisons appear
TrTTTgure21.FoT the purpose of this figure each pro-
ductivity response pattern was summarized in terms of
four segments: independent responses to N, to P, and
to EDTA, and an interaction. Experiments 3, 4, and 20
were omitted because they did not include all treatment
combinations. Pairs of experiments in which all 4 seg-
ments were similar, such as experiments 2 and 5 (See
Table 11), in which an interaction between N and P showed
up but in which no independent effects occurred, were
awarded a 1. Pairs with dissimilar segments were award-
ed fractions from 0 to .75 based on the number of similar
segments divided by 4. In experiments 8 and 16, for ex-
ample, the responses to N and P were similar while the
responses to EDTA and the interaction differed. The
comparison value is .5.
These comparison values can be used to test for a ten-
dency for experiments conducted close together in time to
have more similar results than experiments that were more
widely spaced. For this test two sums were obtained
for each experiment (Table 13). The first is the sum
of the values obtained by comparing the experiment in
question with the four other experiments that were closest
to it in time. The second is the sum of the values for
the four experiments farthest from this experiment in
time. For the purpose of this test experiment 1 was
omitted so that each experiment could be compared to an
even number of others. The sums of these suras are 20.5
77
-------
for the "near" experiments versus 17.0 for the
"distant" experiments, indicating that "near" experi-
ments were more similar than "distant" experiments.
This suggests that changes in response patterns from
experiment to experiment were tied to gradual changes
occurring in the lake system, rather than being totally
random variations.
Environmental Conditions. Table 14 summarizes the am-
bientnTnvironmental conditions associated with the
starting dates of experiments 1-20. The nutrient data
indicate that nitrate was available at the working depth
of 1 m for the entire summer of 1969, while orthophos-
phate declined gradually from a spring peak and reached
the detection limit in late fall. These data seem to
be consistent with the relative roles of N and P as
treatments in the enrichment experiments (Table 11).
Phosphorus additions consistently stimulated productivity
in experiments 7 through 19, during the period of de-
clining ambient P, while N additions were of secondary
importance and usually stimulated only in conjunction
with P.
Stimulation by N was entirely absent from the two spring
experiments, numbers 6 and 20. This seems to relate
to the ambient nitrate levels, which were at seasonal
maxima in both of these experiments. In experiment 20
the ambient nitrate level was much higher than in ex-
periment 6, and this may relate to the slower warming
rate of the surface waters in 1970 compared to 1969.
As shown in Figure 22 a more stable thermal structure
had developed prior to experiment 6 than prior to ex-
periment 20, with consequent reduction in the nutrient
pool of the photic zone and reduction in nitrate levels.
This line of reasoning, however, does not explain why
the ambient orthophosphate level for experiment 20 was
so much lower than for experiment 6. This difference,
however, is consistent with the occurrence of stimula-
tion by phosphate additions in experiment 20 and the
absence of a P effect in experiment 6.
Reference to Table 11 indicates that stimulation by EDTA
was essentially absent in the two fall experiments, 5
and 19. Figure 23 shows that prior to both of these
experiments there had been periods of steadily declining
surface water temperatures. The consequent erosion of
the thermocline permitted reintroduction into the epi-
limnion of quantities of suspended and dissolved material
that had been accumulating in the thermocline during
the summer. This material may be assumed to have in-
cluded trace metals arid dissolved organic compounds with
78
-------
16-
14-
12-
10-
,° 8H
£ 6J
Q_
UJ
Q 4-
£ 2
START OF EXPERIMENT 6
Id
| .8-,
o: 16-
Ld
a.
5 I4H
UJ
»-
12-
10-
8-
6-
4-
TEMPERATURE
p o-
b o--
1 I
ill
10 15 20
APRIL- MAY, 1969
25
I T
30
START OF EXPERIMENT 20
TEMPERATURE
-0
-2
UJ
2
IJ
O
O
s
cr
UJ
i
h-
a.
O
I
h-
Q.
UJ
a
DEPTH
10 15 20
APRIL-MAY, 1970
i
25
I '
30
-0
-2
Figure 22. Spring Thermal Conditions in Third Sister
Lake - 1969, 1970
79
-------
START OF EXPERIMENT 5
10
15 20 25 30
OCT. -NOV., 1968
UJ
oc
I- 18-1
cc
Ul
a.
2
UJ
t-
16-
14-
12-
10-
8-
6-
4-
START OF EXPERIMENT 19
TEMPERATURE
Q_
O
X
h-
Q.
UJ
10
Figure 23.
I
15
{ i i i i i i r
20 25
T»
30
|T
4
-5
-7
8
OCT. -NOV., 1969
Fall Thermal Conditions in Third Sister
Lake - 1968, 1969
80
-------
TABLE 14
AMBIENT LAKE CONDITIONS: EXPERIMENTS 1-20
Expt.
No.
1
2
3
k
5
6
7
8
9
16
17
19
20
Starting
Date
6-12-68
7-18-68
8-07-68
9-19-68
10-27-68
U-23-69
5-26-69
6-0*1-69
7-01-69
8-21-69
9-13-69
10-25-69
1+-23-70
Temp . at
1 m Depth
(°c)
26
21
10
15
18
19
25
25
21
10
10
Light
Extinction
Coefficient
.90
1.20
.98
-69
.U6
78
78
1.17
Nte-N
(W5/D
56
50
to
ii-i
1*2.5
to
38
62
230
POl^-P,
Ortho
10
10
10
6
7-5
5
k
<1
<1
5
(W5/D
Total
10
7
7-5
5
8
6.5
2
chelating abilities, thus reducing the sensitivity of
the system to treatment with additional chelators.
In two other experiments, 2 and 8, EDTA effects were
absent with no apparent relation to normal seasonal
changes within the lake. Prior to experiment 2, how-
ever, on June 25, 1968 a massive inflow of surface run-
off water occurred as a consequence of an unusually
severe series of rainstorms. The runoff carried
sufficient suspended particles to render the entire
lake brown in color and to reduce light penetration to
essentially nil. The particles gradually settled out,
and the lake was visually normal by the start of experi-
ment 2, on July 18. Nonetheless it is possible that
allochthonous organic compounds carried in with the
runoff had remained in sufficient concentrations to make
the treatment additions of HDTA superfluous. It is
also possible, of course, that stimulation by EDTA did
occur but \vas obscured by the high error variance that
showed up in experiment 2. This experiment, as mentioned
in the methods section, was the least sensitive in de-
tecting treatment effects of all the experiments per-
formed (Table 8).
A further consistency that can be observed between ex-
periments performed at similar times in two different
81
-------
years is the absence of a nitrate effect, either inde-
pendent or interactive, in experiments 4 and 16. Nothing
in the ambient lake data suggests an explanation for
this pattern.
Examination of ambient physico-chemical conditions has
offered explanations for several features of the response
patterns: (1) the greater stimulatory ability of phos-
phate treatments relative to nitrate treatments in the
summer, (2) the absence of nitrate effects in the spring
experiments, (3) the occurrence of a phosphate effect
in experiment 20, but not in experiment 6, and (4) the
essential absence of EDTA effects in the fall experiments,
5 and 19. The physical and chemical data provide little
help in interpreting the variations in response patterns
among the summer experiments. Changes in the species
composition of the phytoplankton community may be par-
tially responsible for the variations, and will be
examined next.
Phytoplankton--Communities. In experiment 12 it was
found theft most of the teatures of the productivity re-
sponse patterns, including the changes in these patterns
that occurred from day to day, could be explained in
terms of the growth response patterns of individual
phytoplankton species. The productivity response pat-
terns for experiments 1-19 can be assumed to be, likewise,
integrals of the various growth responses of the species
present. Shifts in the species composition of the phy-
toplankton community, as well as changes in the physio-
logical condition of constant members of the community,
can be expected to have caused changes in the productiv-
ity response patterns from experiment to experiment.
Table 15 summarizes the species composition of the
phytoplankton communities present at the beginning of
each of the 1968 and 1969 experiments. It will be noted
that several species were present in nearly all of the
experiments while the majority were present only occasion-
ally, sometimes at characteristic times of the year.
Many rare species were undoubtedly missed, so this table
should be regarded as a compilation of the occurrence
data for the major species.
In Figure 24 the experiments are intercompared on the
basis of the number of major species that overlapped in
each possible pair of experiments. Figure 25 presents
these comparisons in terms of Sorensen's Index (Sorensen
1948) values, in which the number of overlapping species
is doubled and divided by the total number of species
present in both experiments. The range of possible
82
-------
TABLE 15
OCCURRENCE OF PHYTOPLANKTOW SPECIES IN
INITIAL COMMUNITIES - 1968, 1969
Species
123
Experiment Number
5 6 7 8 9 16 17
19
Chroomonas acuta x xxxxxxx
Cryptomonas ovata x x x x x x
Cryptomonas erosa x x x x x
Ankistrodesmus falcatus xxxxxxxxx
Crucigenia tetrapedia x x x x x x
Tetraedron minimum x x x x x x x
Tetraedron caudatum x x x
Oocystis parva x x x x
Pediastrum tetras x x x
Rhabdoderma sigmoidea x xxxxxxx
Lyngbya limnetica x x x x x x x
Chroococcus dispersus xxxxxxxxx
Gomphosphaeria lacustris x x x
Synedra rumpens x x x x x x
Fragilaria crotonensis x x x x
Asterionella formosa x x x
Gonium pectorale x x x x
Cosmarium truncate Hum x x
Scenedesmus bijuga x
Synedra radians x x
Snn.ed.ra. acus x x x x
Anabaena wisconsinense x x x
Cylindrospermum stagnale x x
Spirulina major x x
Aulosira sp. x x
Microcystis incerta x x x x
Elaktothrix gelatinosa x x x
Chrysidalis sp. x x
Ochrcmonas sp. x
Microcystis aeruginosa x
Sphaerocystis Schroeteri x x x
Oscillatoria rubescens x
Chroococcus minutus x
Fragilaria capucjna x
Pediastrum Boryanum x
Aphanothece nidulans x x
Aphanocapsa elachista
Aphanizomenon flos -aquae
Chlamydomonas pseudopertyi
Glenodiniura pulviscvilus
Coelastrum microporum
Oscillatoria tenuis
Asterococcus limneticus
Anabaena Scheremetievi
x
x
x
x
x
x
x
x
x
x
x
x
x
x
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
83
-------
Experiment Number
1Q '
-ir
-LO-
Q
-
7
f.
O -
c; _
J
Q _
2_
1 -
1 f, f.
1C 11 £
10 11 8 7
T T 7 ft Q f,
J-J- / o y D
TQ Q R Q11 7
_____ TPn"? n ft n~in Y
6c; Q o )i )i !i o
20 1± 12 13 9 11 11 11 8
(Iff 1 1 1 1 1
123^56789
7
9
5
8
8
11
11
1
11
i
16
16
7
6
5
10
7
2
6
t
17
17
13
5
7
7
7
7
6
3
8
l
19
Experiment Number
Figure 2k. Numbers of Species in
Common in Each Pair of Experiments
1968, 1969
84
-------
!H
Q)
£1
%
0)
.^
h
-------
TABLE 16
SUMS OF SORENSEN'S I3DEX VALUES - 19^8, 1969
Year
1968
1969
Expt.
1
2
3
4
5
6
7
8
9
16
17
19
Overall Sum:
1969
6
7
8
9
16
17
19
Overall Sum:
Hear
2.J4
1.52
2.82
2.62
2.50
2.40
2.30
2.46
2.04
2.46
2.36
2.38
28.60
1.58
1.84
1.64
1.34
1.78
1.76
1.72
11.66
Distant
2.56
1.20
2.52
2.64
2.62
2.04
2.68
2.86
2.28
1.88
1.66
1.86
26.80
l.o4
70
1.24
1.18
1.04
.90
l.o4
7.14
values is therefore from 0 to 1. The maximum value
that shows up is .72, while the minimum is .06.
In Table 16 the "near" and "distant" values of Soren-
sen's Index are summed for each experiment in a way
similar to the treatment of the productivity compari-
sons in Table 13. When all experiments are included,
the grand "near" sum is only slightly higher than the
grand "distant" sum, while when only the 1969 experi-
ments are used the near sum is substantially greater.
This discrepancy implies that continual shifts in species
composition occurred within a year, but species complexes
recurred in the separate years.
Not all of the species present in a given experiment re-
sponded to the treatments in the experiment, and often
the number of responding species was a small minority.
In Table 17 the responding species are listed, and the
experiments in which they responded to nutrient
86
-------
TABLE 17
EESPOIvDIIJG SPECIES - 1968, 1969
Experiment Number Code Letter
1 2 3 li 5 6 7 8 9 16 IT 19 for Figui-e 38
Species
Ankistrodesmus falcatus x x x x
Rhabdoderma sigmoidea x x
Chroococcus dispersus x
Synedra rumpens x x x
10 (j. Sphere x
Synedra radians x x
Anabaena visconsinense x x
Elaktothrix gelatinosa x
Lyngbya limiietica x x
Crypt omonas ovata x x
Chrysidalis sp. x
Ochromonas sp. x
Chroomonas acuta x x
Synedra acus x
Microcystis incerta x
Aphanothece nidulans x
Gomphosphaeria lacustris x x
Aphanocapsa elachista x x
Aphanizomenon flos -aquae x
Chlajnydomonas pseud ope rtyi x
Microcystis aeruginosa x
Oscillatoria tenuis x
E
C
G
P
N
D
Q
A
B
F
H
I
J
K
0
L
M
R
S
treatments are indicated. The largest number of re-
sponding species was 6, in experiment 19, while in
several experiments only one species responded. Compari
son of this table with Table 15 indicates that changes
in responding species were much more abrupt than were
changes in the total species complex. This observation
exemplifies how subtle changes in environmental condi-
tions can induce shifts in the species composition of
natural phytoplankton communities.
Figure 26 intercompares the experiments on the basis of
the number of overlapping responding species, while
Figure 27 expresses the comparisons in terms of Soren-
sen's Index. Most of the index values are zero and are
omitted. This general lack of overlap of responding
species supports the hypothesized role of changes in
species occurrence or physiological state in causing
changes in response patterns among experiments.
87
-------
Experiment Number
J-?-
If -
9_
-
5 -
q _
J
? _
1 -
5 1
1 1
k °
^
1
..,.,.! 1
o 1 1
51 121 l
1 2 3 U 5 6 7 8 9 16 17 19
Experiment Number
Figure 26. Numbers of Kesponding
Species in Common in Each Pair of
Experiments - 1968, 1969
88
-------
m
0)
0)
JH
0)
19-
17 -
16
9 -
8 -
7 -
6 -
.33
33
.40
.4U
3
2
1
.29
-22
33
.U4 .22
.18
1
1
^ 5 6 7 8 9
Experiment Mumber
16 17 19
Figxire 27. Sorensen's Index Values
for Responding Species - 19^8^
89
-------
In Appendix B quantitative data for the most important
species in experiments 1-19 are tabulated. In most of
the experiments samples from initial, final control,
and final high treatment communities were counted, and
the counts were converted to biomass estimates. Correla-
tion analyses between the biomass estimates and the
final productivity values were performed when possible,
to indicate those species whose growth responses were
most influential in shaping the productivity response
patterns. Certain of these data are useful in further
attempts to interpret consistencies and shifts in the
productivity patterns among experiments.
Phytoplankton-- Individual Species. It was noted that
in experiments 4 and 16 no nitrate effects appeared, and
that no environmental change that could explain this
was apparent. It was determined from examining the
species present that in both of these experiments
heterocyst-forming bluegreen algae were important mem-
bers of the phytoplankton community. Figure 28 shows
that one of these organisms, Anabaena wisconsi ne nse,
had its peak in abundance in Augustand September of
both 1968 and 1969, and that it responded to treatments
in experiments 3 and 16. Examination of Table 15 re-
veals that other lieterocyst-formers, Cylindrospermum
stagnale, Aulosira sp., and Aphanizomenon flos-aquae
were present in one or both of these experiments, and
absent at other times of year. Thus it is reasonable
to hypothesize that these organisms, heterocyst-formers
and therefore potential nitrogen fixers, could have*
contributed fixed nitrogen to the lake system in the
late summer of both 1968 and 1969, thereby eliminating
the impact of the nitrate added as treatments at these
times.
The biomass estimates for a number of other species
have been plotted in Figures 29-57, and a variety of
temporal distribution patterns appear. The first three
species, Ankistrodesmus falcatus, Cryptomonas oyata,
and Chroomonas acuta were present in all ornearly all
of experiments 1-19. Ankistrodesmus maintained con-
stant low population levelsin the lake except for the
spring and early summer of 1969 when a population peak
occurred. Responses to treatments were confined to the
early summer and late fall of both years. Cryptomonas
population levels were low except in October of both
years., when it peaked and responded to treatments.
Chroomonas differed from the first two species in that
its lake populations were generally high, at least for
most of 1969, while its final population levels in the
experimental jugs were low, if not nonexistent.
90
-------
CD
u-
8-
6-
4-
1 = Initial
C = Control
H = High Treatment
* = Response to Treatment
°~* 'Exp. 1 ' !Exp. 21
Jun. Jul.
4-
2-
n.
p
^
*
\
y
^
r/
i
\
1
^ 1 C H
Exp. 3 Exp.41 ^Exp. 51
Aug. Sep. Oct.
1968
1
C H 1 C H
F70 ?
I
Exp. 6 Exp.7
Apr. May
Exp. 8
Jun Jul.
1969
Aug. Sep. Ocl
Figure 28. Anabaena wisconsjnense Biomass Estimates-
1968, 1969
91
-------
^
o>
°
ICH ICH ICH JCH ICH
« V
^ r$\
rixpTl1 'Exp 2* ^ Exp ^ 'EXP41 Exp 51
Jun. Jul. Aug. Sep Oct
1968
160-
140-
120-
~
3
^ 80~
5
0 60-
CD
4O-
20-
'/
S
f
s
s
s
r
y
f
s
r
/
/
/
/
/
/
^
s
/
X
y
y
y
y
y
y
y
*
/
/
y
/
/
/
/
/
y
/
/
/
/
/
/
/
/
/
/:
/
$
f^
/
1 = Initial
C = Control
71
y
y
y
y
y
y
y
y
y
y
y
y
y
/,
f
/j
/
/,
y
X,
f.
/*
fy
s
/
^
f
y
y
Ex p. 6
H = High Treatment
» = Response to Treatment
171
y
y
y
/
y
V
^
Iffad ICH ICH ICH ICH ICH
/*y / M-*
Exp.7 'Exp-81 'Exp. 9 ' 'Exp.161 'Exp I71 'Exp.191
APR MAY JUN. JUL. AUG. SEP. OCT.
1969
igure 29. Ankistrodesmus falcatus Biomass Estimates
1968, 1969
92
-------
400-
32O-
"o>
"-'24O-
cn
en
^ 160-
0
GQ
80-
0-
*
1 = Initial
C = Control
H = High Treatment
* = Response to Treatment
ICH ICH
^Exp. 1 *Exp. 21 'Exp. 3* 'Exp. 4
^
//
\
Y<
f /'
7,
fa
'f.
//
Y/
S J\
\
^
'/),
f i1
^
^
'/,
\
Exp. 5
Jun. Jul. Aug. Sep. Ocf
1968
240-
o>
3 !60~
en
I 8°-
CD
O-
ICH ICH ICH ICH
^
%
V,
y.
^
y
*
**)
^
y,
^
i
, i , . , i i r^^l i r fc*^
Exp. 6 'Exp. 7' Exp. 8 Exp. 9
Apr. MOV J"n- Jui. Aug. Sep. Ocl
1969
Figure 30. Lryptomonas ovata Biomass Esti^^tes-1968, 1959
93
-------
5-
v
o»
^ 0-
40-i
35-
_ 30-
V.
o»
^ 25-
-------
Treatment effects that appeared for this species in
experiments? and 19 did not represent actual growth of
the populations in response to the nutrient treatments.
Instead, they showed up because treatment populations
declined less than the"control populations. This or-
ganism was evidently sensitive to the jug environment,
and will be discussed further in the section dealing
with containment effects.
The next three species, Rhabdoderma sigmoidea, Synedra
rumpens, and Lyngbya limnetica were significant respond-
ing species in both 1968 and 1969, but were not present
at all times during both years. Rhabdoderma and
Synedra responded in the early summer of both years,
while Lyngbya responded in the fall of 1968 and early
summer of 1969.
The last three figures present the biomass data for
three species that were essentially limited in their
occurrence in the experimental communities to i-hort time
periods in 1969. Chrysidalis sp. showed up in experi-
ment 6 and was one of the two species that responded
to the standard treatment mixture in that experiment.
Its population in the lake had declined drastically by
the start of experiment 7, and was not seen after that.
Gomphosphaeria lacustris was seen in two experiments
in 1968, but did not appear in countable quantities un-
til experiment 9. It responded in that experiment and
in experiment 16, then disappeared. Aphanocapsa
elachista showed up in experiment 16, reached high
levels in experiment 17, then declined, after respond-
ing in both of these experiments.
These examples illustrate the fact that the successional
changes of responding species were almost as sudden as
was implied in Table 17. Species were generally pres-
ent in more experiments than those in which they respon-
ded to treatments, but usually preceded or followed
their responses to treatments with bursts of growth in
the lake. Figure 38 serves to compare the gradually
changing productivity response pattern in 1969 with the
abruptly changing complex of responding species. It
seems surprising that the changes in response patterns
were as gradual as they were in the face of the general
lack of overlap of responding species.
It is implied in the preceding two discussions that
changes in the physico-chemical environment and changes
in the phytoplankton species composition both contribu-
ted to the variations that appeared in productivity re-
sponse patterns. It seems reasonable to propose that
the environmental conditions served to shape the major
-------
en
en
!*-
O
CD
-
I C
CTV
Y/Y/
*
\
y
/
Exp. I
I C H
F/P/I
p/vvi/O
'Exp. 2' 'Exp. 3' 'Exp. 4' Exp. 5
Jun Jul. Aug. Sep. Od
I968
lO-i
8-
^*
\
?6-
BIOMASS C
ro **
i
0
\
A
77
/i
'/
V
y,
Exp.
i
6
I
7
7
y
A
y.
y.
yt
y
y.
y
/
^
y
y
y
y
/
y
y
y
y
y
y
y.
y
/
\
/.
Exp.7
1 '- Initial
C = Control
H = High Treatment
* = Response to
Treatment
1 d& 1 C H
17^4
YV '/ //
V'/'/' . ...n -, 1
'Exp.8 'Exp. 9' 'Exp.161 'Exp.lT 'Ex p. 19
Apr. May Jun. Jul. Aug. Sep. Oct.
1969
Figure 32. Rhabdoderma sigmoidea Biomass Estimates -
1968, 1969
96
-------
80-
-«
"^
§* 60-
-*
en
< 40-
O
OQ
20-
0-
i C
f'/L
*
y
y
y
y
y
v
y
y,
y
/,
A
y
/
y
y,
Exp. 1
*
y
1 f* \Jt 1 (^ W
i \j y\ ion
'/// J7-I
'Ex p. 2 'Exp. 3^ 'Exp. 41 'Exp. 5'
Jun. Jut. Aug. Sep. Oct.
1968
80-
C^ 60-^
CO 40-
o
- 20-
CD
0-
\
T-//
//,
//,
/ /,
//
?X
^
k\\\\\\\V
^
/
y
/
/
/
/
f .
s
k
f
/.
/t
£xp.6
1 = Initial
C = Control
H - High Treatment
* - Response to Treatment
1 CH 1 CH
*
'Exp.T1 lExp. 8 'Exp. 91 'Exp.161 lExp. 1?' 'Exp. »9l
Apr. May Jun. Jul. Aug. Sep. Oct.
1969
Figure 33. Synedra rumpens Biomass Estimates - 1968
1969
97
-------
120-
100-
^ 80-
^
>
~
i i i
o o o
10
^
320-
yy
CO
1 o-
Y
VJs *
^S / S
)^y H* 1 C o I C H
/S/V< *rf&Ss
7y
*
I
*
'/,
^
/
/
/
X\\\\X\X\\\\\\\X3
^
Ex p. 5
Oct .
5 Exp.6 'Exp. 7 'Exp. 8' 'Exp. 91 'Exp. I61 'Exp. I71 'Exp. I91
Apr. May Jun Jul. Aug. Sep
(969
Oct
Figure 34. Lyngbya limnetica Biomass Estimates -
1968, 1969
98
-------
- lOOi
GO
3
CO co-
CO ou
0
m o-
200-
"^
°* 150-
^
0 0
O ID
SSVWOI8
\
y
^/
1 'Exp.l ' 'Exp.21 'Exp. 31 'Exp.41 'Exp.51
Jun. Jul. Aug. Sep. Oct.
1968
*
r?i
C
/ /
^
4,
y
//
'/
S j
/>
^
y
V
1 = Initial
C = Control
H = High Treatment
* = Response to Treatment
1 C H
°Exp.6 Exp.7' 'Exp.8l 'Exp.9' 'Exp.16' 'Exp.17' 'Exp. 19
A
Figure
pr. May Jun. Jul. Aug. Sep. Oct.
1969
35. Chrysidalis sp. Biomass Estimates-1968 , 1969
99
-------
^ I.O-i
3s
-------
2>
o o-1
1 Exp. 1 ' ' Exp. 2' ' Exp. 3' Exp. 41 'Exp. 5'
00 Jun. Jul. Aug. Sep. Oct.
1968
70-|
60-
50-
1 "
o»40_
>
CO
CO
< 30-
O
CD
20-
10-
-
1 = Initial
C = Control
H = High Treatment
*= Response to Treatment
1 C H
*
Exp. 6' 'Exp.71 'Exp.81 'Exp. 91 Exp, 16
Apr. May Jun. Jul. Aug.
^
V;
y
/;
/,
/,
/;
//
\
\
\
I
%
y,
/;
Y,
/S
%
y.
i
\
\
*
y
y
//
\
y.
i
y.
y.
y.
y.
y
\
\
^
^,
7xp.l7
Sep.
1 C H
r ' ' ' ' i
Exp. 19
Oct.
1969
Figure 37. Aphanocapsa elachista Biomass Estimates-1968 . 1969
101
-------
features of the response patterns, such as the relative
importance of P and N in the present series and certain
seasonal aspects of the treatment effects, while species
changes were reflected in the short term changes between
successive experiments.
Use in. Predicting Future Responses
After examining the factors that influenced the varia-
tions in the productivity response patterns, it can be
concluded that for In situ enrichment bioassays to be
used as predictive fcTols series of experiments in each
season in the water body of interest should be conducted.
One, or perhaps a seasonal series of "most probable re-
sponse patterns" could then be proposed if clear-cut
relationships appeared between ambient conditions and
the experimental results.
From the set of experiments performed in Third Sister
Lake in 1968 and 1969, "most probable response patterns"
for the spring, summer, and fall can be proposed. In
the spring,treatments with inorganic P or N are less
likely to stimulate primary productivity than are organic
chelators. In the summer, as inorganic P declines in
the epilimnion, phosphate treatments become consistently
stimulatory. Stimulation by inorganic nitrogen treat-
ments generally is dependent upon the presence of added
P, except in late summer when heterocysted bluegreens
become active, and N treatments are no longer effective.
Stimulation by chelators is fairly consistent through-
out the summer, but is reduced in the fall as the
thermocline is pushed downward, and decomposition pro-
ducts are released into the epilimnion.
Data concerning the species involved in enrichment ex-
periments can considerably amplify predictions based
strictly on productivity responses. This is because the
composition of the phytoplankton as well as its quantity
influences whether or not water quality problems de-
velop from increased productivity. Of the species of
algae that responded to treatments in the experiments
discussed, 11 were bluegreens, 4 were greens, 3 were
diatoms, 2 were cryptophytes, and 2 were chrysomonads.
Of the 44 major species that were identified in initial
phytoplankton communities, 17 were bluegreens, 16 were
greens, 6 were diatoms, 3 were cryptophytes, 2 were
chrysomonads, and 1 was a dinoflagellate. Thus the
nutrient treatments tended to shift the community struc-
ture from approximate equality between green and blue-
green species to a dominance of bluegreen species.
Moreover responses by bluegreens occurred at most times
102
-------
EXPERIMENT NUMBER
6 78 9 16 17 19
t-P
zo
UJUJ
UJ
or
N
r
E
NP
j N E
PE
NPE
tn
UJ
o
UJ
Q_
CO
O
-------
of the year, while responses by the other groups
tended to be limited to the spring, early summer, or
fall. In the late summer of 1969, July, August, and
September, the responding species were almost exclu-
sively bluegreens, one of which, Aph an i. zome no n f los -
aquae, is a well known nuisance b1o om f orme r. Blooms
of planktonic bluegreen algae have not been a problem
in Third Sister Lake, but these results imply that in-
creases in the nutrient input to the photic zone in
the summer could favor bloom formers and lead to
nuisance conditions. In the next section, subsequent
experiments designed to test the stimulatory ability
of NTA will be discussed. Nitrate and phosphate treat
ments were also employed in some of these experiments,
permitting their comparison with the predictions
derived from the initial series.
104
-------
SECTION V
EXPERIMENTS TO TEST A POTENTIAL
ENVIRONMENTAL CONTAMINANT
DESIGNS
In 1970 and 1971 nine :Ln. situ enrichment experiments
were performed with NTA~~as a treatment. The designs,
productivity results, and data analyses are tabulated
in Appendix C. The NTA treatment level employed in
most of the experiments was .252 mg/1 (1.3 x 10"6M).
In addition to NTA, treatments with N, P, and EDTA were
frequently included while treatments with glycine,
vitamin B-12, Zn, and Mo were included in one experi-
ment. Productivity in the open lake was monitored dur-
ing most of the experiments, but discussion of these
results will be deferred until the section on contain-
ment effects.
PRODUCTIVITY RESULTS
Comparison with 1968, 1969 Results
Table 18 summarizes the productivity response patterns,
while Table 19 indicates the frequency of occurrence
of each possible independent effect. Considering the
complete response patterns, it will be noted that the
three treatments N, P, and NTA each achieved indepen-
dent stimulation of productivity in the majority of the
experiments in which they were included. Responses to
EDTA occurred in only half of its experiments. The con-
sistent effects of P and chelator treatments in this
series, therefore, repeated the main results of the
1968-69 series. The major change in the 1970-71 exper-
iments was the greater significance of the nitrogen
treatments, which achieved independent effects rather
than the secondary effects in conjunction with phos-
phorus that occurred previously.
In both 1968 and 1969 EDTA failed to stimulate produc-
tivity in the fall experiments, and this was inter-
preted to relate to the onset of overturn. In 1970
thermal conditions at the time of experiment 30 were
similar to conditions in the previous two falls, yet
both NTA and EDTA effects appeared, casting doubt on
the hypothesis that thermocline erosion reduced the sen-
sitivity of the system to added chelators.
105
-------
TABLE 18
SUMMARY OF PRODUCTIVITY
RESPONSE PATTERNS,
EXPERIMENTS 22-3^
1970
Month :
Expt:
Independent
effects
May
22
NTA
EDTA
May
23
N
May
2k
P
NTA
Jun
25
N+P
NTA
EDTA
Zn(-)
Aug
29
N
NTA
EDTA
Oct
30
N
P
NTA
EDTA
Interactions
Notes
NTA Zn
1971 Expt:
P not P, EDTA
Tested not
Tested
31
Independent
effects
Interactions
Notes
N
P
NTA
N, P N, P
not not
Tested Tested
TABLE 19
f
FREQUENCY OF RESPONSES
EXPERIMENTS 22-3^
Treatment No. of Expts. No. of Expts. No. of Expts. No. of Expts
Tested with with with
Independent Interactions No Effect
Effects Only
N
P
NTA
EDTA
Glycine
B 12
Zn
Ko
7
5
9
8
1
1
1 .
1
5
H
6
k
0
0
1
0
0
0
0
0
0
0
0
0
2
1
q
k
1
1
0
1
106
-------
The differences between the results of the 1970-71
experiment series and the previous series indicate
that year to year variations in response patterns could
be greater than anticipated on the basis of the 1968-
69 data. Nonetheless the major elements of the "most
probable response pattern" formulated from the experi-
ments of the first two years would be altered little
by incorporation of the later results. They would
simply be broadened to include a greater likelihood of
independent stimulation by nitrogen and the probability
of stimulation by chelators throughout the ice free
season.
Combining all of the experiment results and ranking the
treatments on the basis of frequency of effects, the
order of importance in stimulating productivity in Third
Sister Lake is still (1) P, (2) chelators, and (3) N.
Thus a recommendation for nutrient management in the
Third Sister Lake watershed would stress control of P
inputs, but would advise that all inputs be prevented
if possible since the phytoplankton of the lake is
evidently highly sensitive to stimulation by alloch-
thonous substances of many types.
With regard to formulating predictions on the basis of
in situ enrichment experiments the results indicate that
only the most general predictions can be made with a
high degree of confidence. Predictions of the influen-
ces of particular environmental conditions or species
assemblages on response patterns can be attempted, but
can be expected to fail more often than predictions of
average community responses over entire seasons. Thus
nutrient management policies that relax restrictions
on inputs during periods of supposedly reduced sensi-
tivity to enrichment are less desirable than policies
that assume that the most probable response pertains at
all times.
NTA Effects
Turning to the effects of NTA, it is apparent from
Tables 18 and 19 that this compound could stimulate pro-
ductivity as consistently as had EDTA throughout the
study. Table 20 summarizes the relative effects of these
two compounds in the experiments in which both were em-
ployed. In most of the experiments stimulation by EDTA
treatments was either similar to or stronger than that
achieved by NTA treatments of equal molarity, while in
one experiment NTA stimulated but EDTA did not. The
occurrence of differences in the effects of the two sub-
stances is not surprising, since the stability constants
107
-------
TABLE 20
COMPARISON OF UTA. AND EDTA
EFFECTS, EXPERIMENTS 22-3^
Experiment
22
2U
25
29
30
31
32
Treatments
Stimulating
Productivity
EDTA,
EDTA,
EDTA,
EDTA,
EDTA,
MTA
HTA
NTA
ETA
KTA
ETA
Selative
Effects
EDTA
EDTA
EDTA
EDTA
EDTA
> NTA
> NTA
> NTA
= NTA
= NTA
Neither
for LPTA-netal complexes
complexes for all metals
are stronger than
(Pollard 1966) .
for NTA-Metal
It seemed most probable that NTA treatments were stim-
ulating productivity by means of a chelation mechanism,
since the experiments were terminated before biode-
gradation was likely to occur. Also, results of studies
with cultures of estuarine phytoplankton (Erickson
et al. 1970) indicated lack of utilization of NTA as a
nitrogen source by algae. Nonetheless the possibility
of its utilization as a nitrogen source existed, and
this was tested in experiments 22, 23 and 24. In these
experiments side treatments with inorganic nitrogen,
half as ammonia and half as nitrate, were included at
levels comparable to the nitrogen levels added as NTA
(Appendix C, Tables 1, 2, and 3). In no case did the
have effects comparable to the NTA
In experiments 22 and 24 NTA stim-
and nitrogen did not, and in experi-
occurred.
nitrogen treatments
effects (Table 18).
ulated productivity
ment 23 the reverse
While these results confirmed that the NTA was not uti-
lized as a nitrogen source, they did not prove that
chelation was the mechanism actually involved. A more
direct test was employed in experiment 25, in which
treatments with the metals Zn and Mo
conjunction with the NTA treatments.
occurred, either when added alone or
NTA, but NTA stimulated productivity
The Zn treatments, however, strongly
tivity, and this inhibition was partially canceled by
the addition of NTA (Figure 39). A similar interaction
were included in
No effect of Mo
in combination with
when added alone.
inhibited produc-
10-8
-------
70-
60-
50-
O
CJ>
e
>.
>
3
O
w
0_
40-
30-
20-
10-
NTA Present
NTA Absent
o-H-
Zn,
Zn,
Figure 39. Interaction between NTA and Zn, Experiment 25
109
-------
between Zn and EDTA was found by Glooschenko and Moore
(1971) and was postulated by them as the mechanism by
which EDTA and citrate stimulated productivity of
phytoplankton in water samples from polluted Hamilton
Harbor. Since there are no known inputs of toxic
metals to Third Sister Lake, it is more logical to
assume that the chelate effects observed in the present
experiments were due to enhanced availability of trace
metal nutrients, rather than sequestration of toxicants
In additional parts of experiment 25 treatments with
vitamin B-12 and with glycine, a possible breakdown
product of NTA (Thompson and Duthie 1968) , were tested
for independent effects, but none were found (Appendix
C, Table 4).
SPECIES COUNTS
NTA Effects
Limited plankton counts were performed for two of the
NTA experiments in an attempt to confirm that produc-
tivity stimulation signified actual growth stimulation.
In experiment 22 three major species were counted
(Table 21). Both Chroomonas acuta and Cryptomonas
ovata declined in the control jugs, but declined less
in tne jugs treated with NTA and EDTA. Thus the treat-
ment effects for these species consisted of differen-
tial survivals rather than differential growth rates.
Ankistrodesmus falcatus, however, did exhibit an actual
growth response, but only to EDTA. Thus in this exper-
iment it could be claimed that the NTA effect was an
artifact of the method since it appeared for species
whose major responses were to the jug environment.
In experiment 29 (Table 22) a similar problem of inter-
pretation arises, although the response pattern is
more complicated. Anabaena wisconsinense declined in
the control jugs, less in the jugs treated with N, and
not at all in those treated with P. In response to NTA,
however, actual growth above the initial population
level occurred. Thus NTA caused a growth response, but
since the species involved was evidently strongly af-
fected by the jug environment it could be argued that,
once again, the NTA effect was an artifact. In the
same experiment Ankistrodesmus falcatus grew to similar
population levels in all experimental units.
110
-------
TABLE 21
EXPERIMENT 22
SPECIES RESPONSE PATTERNS
TREATMENT MEANS (
Species
Chroomonas
acuta
Crypt omonas
ovata
Ankistrodesmus
falcatus
Significant
Treatment
Initial Control NTA^
167.0
22-7
31.2
Simple Effects
Chroomonas
acuta
33-7 53-6
6.71 7-85
28.6 29.3
(Tukey's Test)
Species
Crypt omonas
ovata
EDTA^
72.6
10.02
41.8
Ankistrodesmus
falcatus
- Control
** *
EDTAo -
o - o
o - Control
TABLE 22
EXPERIMENT 29
SPECIES RESPONSE PATTERNS
TREATIvIENT MEANS (M
Species
Treatment
Initial Control N P HTA
2.06 .246 1.46 1-90 3.62
Anabaena
wisconsinense
Ankistrodesmus -322 .628 .730
falcatus
Significant Simple Effects (Tukey's Test)
Species
Anabaena Ankistrodesmus
wisconsinense falcatus
Comparison
KTA - Control **
ETA - M *"*
WTA - P *
N - Control *
M - p
P - Control *
111
-------
Comparison with 1968, 1969 Results
The species counted in these two experiments had been
significant members of the responding species complexes
in experiments 1-19. Ankistrodesmus falcatus, it will
be recalled, was presentin allof the1968 and 1969
experiments, peaked in the spring, and responded in the
early summer and fall. The 1970 data are consistent
with this pattern, with the occurrence of high popula-
tion levels and of a response in a May experiment versus
low population levels and lack of response in an August
experiment. Chroomonas acuta and Cryptomonas ovata
both declined in the jugs as had been the case, at
least for Chroomonas acuta, in all of the 1968 and 1969
experiments"! Finally, Anabaena wisconsinense, which
had appeared in the late summer in both 1968 and 1969,
appeared again in August 1970.
In this year, in contrast to the preceding two, nitrogen
treatments stimulated productivity in the presence of
this species. Thus one other specific prediction that
nitrogen treatments would not stimulate in the presence
of heterocyst forming bluegreen algae? that was included
in tne "most probable response pattern" must be elimina-
ted or at least reduced from a probability to a possi-
bility.
112
-------
SECTION VI
EXPERIMENTS TO EVALUATE STIMULATION BY SEWAGE EFFLUENTS
DESIGNS
During the course of the complete study five side
experiments were performed in which volumes of mem-
brane filtered (.45v pore size) secondary sewage
treatment plant effluents were employed as treatments.
The results of the effluent treatments were then com-
pared to the results of treatments with mixtures of
N and P, which in two cases were comparable to the
levels added in the effluents. The productivity re-
sults and data analyses are tabulated in Appendix D.
The results of three of these experiments, 6-S, 25-S,
and 34-S, have been analyzed in particular detail, in-
cluding extensive species counts, and this discussion
will be restricted to them. The other two experiments,
24-S and 29-S, evaluated only in terms of productiv-
ity, were similar in that the responses to the sewage
additions exceeded the responses to the standard
nutrient mixtures .
TABLE 23
SEWAGE EXPERIMENTS: BACKGROUND DATA
Experiment
6-S
25-S
3^-S
April
June
June
'69
70
'71
Lake Levels
K03 -N POk-P
50 10
40 6
~o 7
Treatments
Mixture
H 03-K POk
25
"s O
' '
210 36
-P
5
-x
c
EDTA
Hg/1
500
0
0
Sewage
NO^-H POk-P
M-g/1 M-g/1
(325 ml-»-19 if
"(200 ml-*19 1)
.3 23
(200 ml-KL9 1)
210 368
Table
me n t s
23
to
presents
be discussed
background
data for the three
In experiment 6-S the
P04 levels in the known nutrient mixture were
approximate
experi-
N03 and
chosen to
half the
ambient
of N03 and P04 added in the
In experiments 25-S and 34-S
mixtures were similar to the
lake levels. The levels
ewage were not determined.
the levels added in the NP
levels added in the sewage
..itrogen forms other than nitrate in the sewage were
rot considered, and therefore the inorganic nitrogen
;ovals in the NP mixtures were not strictly comparable
113
-------
to the levels in the sewage. Thus in the comparisons
to follow, conclusions about the relative contributions
of inorganic nitrogen to the sewage effects will not
be attempted.
PRODUCTIVITY RESULTS
Figure 40 shows the results of the productivity measure-
ments, expressed in mg C/(m3 x 4 hr) , and plotted ver-
sus time since the start of each experiment. All
differences that are shown on a given day are signi-
ficant at the .05 level, and most at the .01 level
according to Tukey's test (Steel and Torrie 1960).
Values that were not significantly different have been
averaged and appear as one point. In experiments 25-S
and 34-S productivity in the open lake was monitored
during the experiments, and these results appear as
the dotted lines.
In experiments 6-S and 34-S the productivity levels
of the experimental units treated with sewage and with
the nutrient mixture followed similar trends with time,
and ultimately converged. In experiment 25-S the re-
sponse to the mixture declined rapidly, while the re-
sponse to sewage intensified before it declined.
The only certain conclusion that can be drawn from the
productivity data is that sewage was more stimulatory
to community productivity than was the nutrient mix-
ture in all three experiments. Since effects on
community structure are not revealed it cannot be con-
cluded whether the sewage and mixture treatments affect-
ed the same species to different degrees, or completely
different species, or some of the same species and
some different species. Such qualitative species in-
formation would be required if the bioassay were intend-
ed to determine the extent to which nitrate and phos-
phate in the sewage contributed to its overall effect.
The divergence of the trends of control productivity
and lake productivity in experiments 25-S and 34-S im-
plies that the algae in the jug communities responded
to containment as well as to the nutrient treatments.
This aspect will be discussed in the section dealing
with containment effects.
114
-------
200-
150-
100-
50-
eo
o
EXPERIMENT 6-S
N + P + EDTA
i I I I T I I I ' 1 I
O
0>
E
(S)
O
z
UJ
\-
>
o
O
O
tr
Q_
150-
100-
50-
0
EXPERIMENT 25-S
300-
250-
200-
I 50-
100-
50-
/ \
/ N
EXPERIMENT 34-S
A s ^N + P
LAKE
CONTROL
0-H r
1 1 1 1 1 1
I 2 3 4 5 6 7 8 9 10 II 12
DAYS AFTER START OF EXPERIMENT
Figure 40. Sewage Experiments; Productivity Response
Patterns
115
-------
TABLE 2k
SEWAGE EXPERIMENTS:
SUMMARY OF TREATMENT EFFECTS
(llumber of Species Showing Each Effect)
Response Pattern Experiment Number
JJo Treatment Effect
ydxture* Effect Only
Sewage Effect Only
Both Effects
Same Magnitude
Sewage > Mixture
6-S
5
2
6
0
0
25-S
9
0
7
h
2
3^-S
2
0
2
5
2
*Mixture = N + P + EDTA in Experiment 6-S
K + P in Experiments 25-S and
SPECIES COUNTS
To determine whether the treatment effects were general
or selective, biomass data for individual species were
analyzed. Mean biomass estimates and the results of
variance analyses of the individual estimates are pre-
sented in Appendix E.
Table 24 summarizes the results for the mixture and
sewage treatments in terms of the number of species
that exhibited each possible treatment effect. In ex-
periment 34-S 2 species responded to neither treatment,
2 others responded to sewage alone, and 7 others re-
sponded to both treatments, 5 of them to the same de-
gree. Thus there was a strong qualitative overlap
between the two treatment effects.
In experiment 2S-S 9 species did not respond, 7 re-
sponded only to sewage, while only 6 responded to both.
The overlap, therefore, was not as strong as in experi-
ment 54-S. Reference to Figure 40 will reveal that the
different degrees of species overlap relate closely to
the different degrees of similarity between the sewage
and mixture effects observed in terms o£ productivity
in these two experiments. Nine species responded
116
-------
differently in experiment 25-S while only 4 responded
differently in experiment 34-S.
Returning to Table 24, we find that in experiment 6~S
there was no species overlap at all--the species either
did not respond , responded only to the mixture or only
to the sewage. The productivity data in Figure 40, how-
ever, contain no evidence of this qualitative discrep-
ancy.
In addition to data for the treatment units receiving
N + P and sewage, the experiment 34-S table (Appendix
E, Table 5 ) includes species data for units receiving
P alone and N alone. Factorial analysis of variance
techniques were employed to test for interactions be-
tween N and P on the species level and detected positive
interactions for Chroomonas acuta, Elaktothrix gelatinosa,
Sphaerocystis schroeteri, Synedra sp., Cryptomonas ovata,
and, to some extent, Crucigenia rjscj^angula.ris. (IT the
other species, Aphanothece nidulans responded only to P,
while Ankistrodesmus falcatus responded only to N, but
not in the presence of P.TTTese data further exemplify
the variety of patterns that different species may exhi-
bit in response to components of nutrient mixtures, that
was shown for experiment 12 in section IV.
SEWAGE VERSUS MIXTURE EFFECTS
The comparisons of the productivity and individual species
responses to the nutrient mixture with the responses to
the sewage treatments in these three experiments indicate
that in all cases factors in addition to those included
in the known mixture contributed to the effect of the
sewage. Presumably additional substances suspected as
potential contributors to the sewage effect could be added
to the mixture until eventually most or all of the stimu-
lation by the sewage could be accounted for. As indicated
by the additional species data for experiment 34-S inter-
actions at the species level are likely to be important
in shaping the overall community response to any nutrient
mixture.
The varying degrees of overlap between the species com-
plexes responding to the sewage and those responding to
the nutrient mixtures emphasize the necessity for basing
comparisons of treatment effects on species data as well
as productivity patterns, whenever it is to be concluded
whether or to what degree the components of one treatment
contributed to the effects of another treatment. This is
due to the high selectivity of treatment effects in enrich
ment experiments employing mixed communities of algae.
117
-------
SECTION VII
CONTAINMENT EFFECTS
BACKGROUND
Thus far the results of the i_n situ enrichment experi-
ments conducted in this study~ha've been discussed
chiefly with regard to their internal consistencies.
The results of individual experiments have been compared,
as have the major response patterns that have shown up
in series of experiments. The question that has been
asked is: To what extent can the results of future ex-
periments be predicted on the basis of the results of
past experiments? This is one step removed from the ul-
timate question that must be considered in evaluating
the utility of the method: To what extent can the re-
sponses of the community in the natural system to future
enrichment be predicted on the basis of past experiments?
To deal with this question data on the response of the
natural plankton community to the experimental environ-
ment are required.
A plankton community enclosed in an :Ln situ incubation
vessel is subjected to several important environmental
alterations. First and probably most important is a re-
duction in turbulence from that normally experienced in
the open water. Turbulence affects the organisms by
producing relative motion of the lake water past them,
which in turn enhances nutrient availability by main-
taining a steep concentration gradient between the en-
vironment and the cell (Hutchinson 1967, p. 293). The
importance of turbulence to algal growth was demonstrated
by Fogg and Than-Tun (1960) , who found that growth of
Anabaena cylindrica in culture was doubled when the
flasks were shaken at 90 oscillations per minute instead
of 65, and was prevented at 140 oscillations per minute.
A second set of environmental changes can result from
the presence of the surface of the vessel (Lund and
Tailing 1957). Bacterial growth on and solute exchange
with the walls of the vessel can alter the nutrient en-
vironment to an extent that should be related to the
material composing the vessel and to the ratio of the
surface area of the vessel to the volume it contains.
Also the vessel surface presents a barrier to the ex-
change of solutes with the outside environment, thus
permitting further modifications of the interior environ-
ment as nutrients are utilized and metabolic wastes ac-
cumulate .
119
-------
Thirdly, the quality and quantity of light reaching the
enclosed algae will be modified by passing through the
walls of the container. Glass absorbs more light than
does water, especially in the shorter wavelengths. This
absorption can enhance productivity by protecting the
plankton from excessive light intensity on bright days,
or it can depress productivity on cloudy days. Plastic
and quartz containers modify light conditions less than
do glass containers (Soeder and Tailing 1969).
OBSERVED EFFECTS
Productivity Declines
Some indications of the influence of containment on the
behavior of the communities employed in the 1968 and
1969 series of enrichment experiments can be obtained
by examining the productivity results for control comm-
unities in individual experiments. In Figure 41 the
mean productivity of the control jugs involved in each
experiment is plotted versus time since the beginning
of the experiment. The apparent trend observed in 10
of the 12 experiments is decline of productivity \vith
time. The two exceptions were experiments performed in
the spring and late fall of 1969, while those that
showed decline were performed in the summer and early
fall, implying a relationship to season.
While the productivity trends for the control jugs in
these experiments suggest that the jug environment de-
pressed community productivity, this could not be con-
cluded for certain because productivity in the open lake
was not concurrently monitored. Beginning in the spring
of 1970 comparisons between jug and lake productivity
were made, and the results appear in Table 25.
Reference to this table reveals that in four experiments
(22, 23, 31, and 32) productivity in the control jugs
maintained levels that were comparable to the levels
achieved by the lake samples throughout the experiments.
In experiment 24 control productivity exceeded lake pro-
ductivity by almost 10 mg C/(m3 x 4 hr) on both days of
measurement. This was probably due to a decline of
the lake community subsequent to the start of the experi
ment. The experiment was set up in the morning of May
25, and by noon all of the jugs had been filled and sus-
pended at the incubation depth. In the afternoon a
severe rainstorm washed in large amounts of silt that
raised the turbidity of the lake and apparently carried
the phytoplankton downward as it settled.
120
-------
EXPT.6
I
23456789 10 II
DAYS SINCE START OF EXPERIMENT
Figure 41. Productivity Trends in Control Jugs- 1968. 1969
121
-------
TABLE 25
COMPARISON OF LAKE AND CONTROL PRODUCTIVITY MEASUREMENTS,
1970 AND 1971 EXPERIMENTS
Experiment Date
20
22
23
24
25
29
30
31
32
34
4-24-70
4-25-70
4-26-70
4-27-70
4-28-70
5-3-70
5-13-70
5-15-70
5-20-70
5-21-70
5-26-70
5-27-70
6-18-70
6-19-70
6-22-70
8-22-70
8-23-70
8-24-70
8-25-70
8-26-70
10-25-70
10-27-70
10-29-70
5-19-71
5-20-7X
5-21-71
5-26-70
5-27-70
5-28-70
6-10-71
6-11-71
6-12-71
6-14-71
Control
mgC/(m3
110.8
124.4
122.4
118.6
120.4
55.8
41.9
52.4
56.2
42.0
42.2
33-2
61.9
35.4
17-2
32.5
56.2
46.o
4o.8
61.6
69.9
51.4
18.1
22.7
26.6
29.4
39-4
37-5
36.3
69.8
31-9
30.0
25.0
Lake '.
x4hr) (
114.6
130.4
139. ^
154.5
118.9
77.2
49.2
57-2
57-5
-
33-3
24.8
52.3
52.9
50.8
_
100.0
87.5
71.2
69.3
85.3
88.2
36.1
21.3
29.0
28.1
36.2
33-3
29-9
50-3
37-2
53-7
61.5
Lake Lakt.
Control Higii Treatment
1.03
1.05
1.13
1.30
.99
1-38
1.17
1.09
1.02
-
79
i ^
.84
1.49
2.95
-
1.78
1.91
1.74
1.12
1.22
1.72
1.99
-9^
1.09
.96
-92
.88
.82
.72
1.17
1.79
2.46
993
.936
1.02
1.15
.800
1.10
.924
i.o4
.908
.643
.544
-5>2
1.21
1.39
l.ll
1.08
.984
.748
.960
.938
1.21
.873
.976
.900
.881
.826
.860
.455
.246
.233
.314
1 ?
_L t-t t
-------
In the remaining experiments (20, 2b, 29, 30, and 54)
control productivity and lake productivity diverged. In
three of these experiments (25, 30, and 34) the diver-
gence increased with time as the control productivity
declined and the lake productivity remained relatively
stable.
Species Shifts
The species data for experiments 25-S and 34-S were
discussed in the preceding section with regard to respon-
ses to the nutrient treatments. Responses to contain-
ment were also evident for a number of species (Appendix
b, Tables 5 and 5) and these can be examined in an
attempt to interpret the productivity changes that oc-
curred in the 2 experiments. The numbers of species ex-
hibiting each possible containment effect are summarized
in Table 26. In both experiments there were species
that responded positively as well as negatively to con-
tainment, yet the resultant effect at the community
level was a decline in productivity in both cases.
Two of the species that exhibited negative containment
effects in experiment 25-S were Chroomonas acuta and
Cryptomonas oyata_, species that had been found in pre-
vious experiments to be intolerant of the jug environ-
ment. The third species, Cryptomonas erosa, is closely
related to the first two. All tHTree of these species
declined somewhat in the open lake, but essentially died
out in all of the jugs.
The fourth species that responded negatively to contain-
ment was a species of Synedra (sp. 1) that increased in
all jugs, but did not increase in the control jugs as
much as it did in the open lake. Aphanothece nidulans
in experiment 34-S exhibited this same type of pattern,
which was also interpreted as a negat.:/~e containment
effect. Implicit in these latter evaluations is the
assumption that the final lake populations were sampled
from the same water mass as were the initial populations
--an assumption that is open to criticism. However, con-
sidering the small size of Third Sister Lake, the pro-
tection from wind afforded by the surrounding forest,
and the fact that many species did not change in abundance
between the initial and final lake samples, such an
assumption is not highly unreasonable in this case. In
larger bodies of water containment effects would probably
have to be evaluated, exclusively by comparing initial
lake and final control samples which would, however, con-
fuse containment effects with increases or declines in
populations that would have occurred naturally.
123
-------
TABLE 2c
SEWAGE EXPERIMENTS:
SUMMARY OF CONTAINMENT EFFECTS
(Number of Species Shewing Each Effect)
Response Pattern
Experiment Number
No Containment Effect
Positive Containment Effect
Negative Containment Effect
25-S
13
5
7
3
1
TABLE 2f
SEWAGE EXPERIMENTS: SUMMARY OF
TREATMENT AND CONTAINMENT EFFECTS
Experiment Number:
Containment Effect:
Response Pattern
N+P Effect Only
Sewage Effect Only
Both Effects
Same Magnitude
Sewage > Mixture
No Treatment Effect
0
5
3
E 8
5
25-S
+ -
2
1
1 1
k 1
1 3
z
0
7
h
2
13
9
0
i
3
1
5
2
3^-S
+ - z
0
1 2
115
1 2
3 1 9
2
In experiment 34-S Chrooraonas acuta maintained control
populations little changed From the initial lake popu-
lations, indicating that even the historically most
sensitive species did not always follow its normal de-
cline.
Table 27 summarizes the complete species data for ex-
periments 25-S and 34-S, showing for each species the
combination of treatment and containment effects that
it exhibited. Most of the species that responded to
the nutrient treatments were either neutral to contain-
ment or were favored by it. However, in each experiment
there was one species that responded to a treatment
but responded negatively to containment.
The patterns presented in Table 27 exemplify the true
124
-------
response of a natural phytoplankton community in an
ill si^u enrichment experiment. The responses measured
on tlTecommunity level in terms of productivity or of
any other gross measurement represent the summation of
a set of individual species responses to containment
and to the nutrient treatments. The declines in con-
trol productivity detected in experiments 1-19 were
thus, for the most part, secondary effects of species
shifts in response to the jug environment.
The results of experiment 15 (See Section III) indicated
that the enclosed plankton communities adapted rapidly
to the jug environment. It is now apparent that the
"adaptation" did not involve some physiological adjust
ment shared by all species, but rather resulted from
selective removal of the species most sensitive to con-
tainment. Since experiment 15 was performed at a time
when Chropmonas acuta comprised about 50% of the phyto-
pi anktfoirTTTomass, it is logical to hypothesize that the
adaptation observed was due to rapid selective removal
of this species from the jug communities.
Since the containment effects in experiments 25-S and
34-S were highly species specific, it is possible to
subtract out the sensitive species and determine if the
interpretations of the experiment results are affected.
This is done in figure 42, in which biomass estimates
for the neutral species were summed and plotted on the
left, and sums of all the species biomass estimates were
plotted on the right. The superficial conclusion, that
sewage was more stimulatory than the nutrient mixtures,
is not changed in either experiment. Thus if the ques-
tions asked of an enrichment experiment are simply
whether a given treatment might affect phytoplankton
growth, or which of two treatments might be more potent,
the answers appear to be similar whether or not contain-
ment effects are evaluated. Response measurement in
terms of productivity is adequate as long as it is con-
firmed that productivity signifies growth. More detailed
questions, such as whether two treatments have similar
effects, require more detailed analyses of individual
species responses.
Experiments with Alternative Containers
Up to this point, containment effects have been discussed
as if they were one set of effects common to all con-
tainers. Evidence has accumulated in this study as well
as in others to indicate that containment effects depend
to some extent on the type of container employed.
125
-------
o>
CO
CO
o
CD
300n
250-
200-
150-
100
50
0
1750
1500
1250
1000
750-
500-
250-
0
EXPERIMENT 25-S
EXPERIMENT 34
S
CON MIX SEW
NEUTRAL TO
CONTAINMENT
CON MIX SEW
TOTAL
Figure 42. Biomass Estimates for Neutral Species vs
Complete Communities
126
-------
500
400
1
x
M
S
x,
o
0
300
200
100
SAMPLES FROM COLUMNS
SAMPLES FROM LAKE STATIONS
cot. i
COLUMNS
INSTALLED
6-19 6-21 6-24
DATE (1969)
6-27
Figure 43. Comparison of Productivity Profiles:
Two Columns vs Two Lake Stations
127
-------
Polyethylene Columns. The long term trend of produc-
tivity declinein the jugs was noticed long before
plankton counts were made that related this trend to
selective elimination of particular species. It was
felt that if surface effects were responsible for this
property, perhaps larger containers with smaller sur-
face area to volume ratios would perform better. Ac-
cordingly some experiments were performed with large
polyethylene columns (Goldman 1962). Substitution"of
these vessels for the jugs would also have added the
dimension of depth to the experimental design which
with the jugs was restricted to a single depth.
Data on the long term productivity of populations en-
closed in large bags are rare in the literature. The
.'nost complete record appears in Gachter's paper (1968),
in which lie reported a gradual increase followed by a
gradual decline in productivity in a control vessel
over an 18 day period. The productivity never declined
below the initial level. McLaren (1969) found that pro-
ductivity in a fertilized column greatly exceeded that
in the open lake over a three week period. However, no
data could be found directly comparing productivity in
a control column with that in open water.
The first column experiment in the present study was
performed early in June, 1969. Carbon-14 productivity
profiles were measured using samples taken from within
a single column and from a single lake station. The
column was constructed from 6 mil polyethylene film, 0,5
m in diameter by 10 m long, open at both ends, and was
suspended in Third Sister Lake with its upper end about
10 cm above the water surface. Profiles were measured
three times during a 12 day period, and showed a gradual
decline in productivity in the column relative to the
lake. A second experiment was performed employing dup-
licate columns and duplicate lake stations. The results,
which appear in Figure 43, shotted not only that produc-
tivity declined in the columns relative to the lake, but
also that the columns diverged from one another. Thus
the columns proved to be no better than the jugs at main-
taining the photosynthetic activity of phytoplankton
communities, and they showed that they would perform
much worse as experimental units in a statistical design.
A more detailed discussion of these results is found in
Bender and Jordan (1D70) .
Glass Tubes and Jars. Information about the behavior of
two other types of containers emerged indirectly from
efforts to develop a screening method for testing a
large variety of nutrients for potential use in jug experi
ments. One container was a glass tube sealed to a glass
128
-------
28
24
J 20
2»
-------
plate at one end and open at the other, with dimensions
10 cm diameter by 40 cm high. The experimental lakewater
volume added to each tube was two liters. The second
container was a glass jar with dimensions 16 cm diameter
by 24 cm high, and with a mouth of 8 cm diameter. Two
or three liters of lake water were added. During an
experiment these vessels were incubated alongside the
lake in a water bath through which lake water was pumped
for cooling. The mouths of the containers were left
open.
Figure 44 shows productivity trends in control tubes for
the three experiments in which they were employed. Pro-
ductivity declined with time, and replicability between
duplicate tubes receiving identical treatments was poor.
The jars were then tested in hopes that they \vould at
least provide better replicability. These vessels were
employed in two experiments, 16 and 18.
Direct comparison of the behavior of control jars with
that of control jugs was possible in experiment 16 since
both sets of vessels were filled Ciojn the same tank of
lake water, and 14C measurements were run simultaneously.
The surprising results appear in Figure 45. Productiv-
ity of samples from the control jars more than doubled
between the first and second measurements, while pro-
ductivity in jug samples dropped almost 50%.
In experiment IS, a 2 x 3 design employing I\ and P
(Table 28), jars alone were employed, and productivity
again increased with time. Examination of plankton
samples revealed that Chroomonas acutj., which had never
increased in numbers in a control~Tug~b u t almost always
declined or disappeared, had actually increased in the
control jars (Table 29). The productivity results of
experiment 18 differed from the results of the jug ex-
periments of the same series in that the former indica-
ted no stimulation by the P treatments but stimulation
only by added N, in contrast to the P effects consis-
tently obtained in the jug experiments. This could be
due in part to the use of lower treatment levels of P
and higher levels of N than were normally used in the
jug experiments.
The difference in the fates of Chroomonas acuta in the
jar environment and in the jug environment indicates
that this species was not intolerant of containers in
general, but rather of the jug container in particular.
Differential responses of phytoplankton species to dif-
ferent types of incubation containers were reported by
Thomas (1961), who traced species changes in flasks and
in 7 m plexiglass columns, all suspended in a lake.
130
-------
52
48
44
40
36
QC 19
I 3Z
x 28
M
£24
o
o 20
16
12
8
-TREATMENT VESSELS'JUGS
-TREATMENT VESSELS /^EXPT 16
= JARS /
EXPT.I8
~~ "~~ *
EXPT. 17
12345
DAYS SINCE START OF EXPERIMENT
Figure 45. Productivity Trends: Control Jars vs
Control Jugs
131
-------
Treatment
N
Control
Source
K
P
NP
Comparison
N - Control
-P2
Control
- N
Control
- N
P- -
P -
TABLE 28
EXPERIMENT 18
10-13-69 - 10-18-69
Design
Enrichment
K03 - N
- K
- P
Concentration
333
+
150 ng/1
1-0 Hg/1
1.7 ng/1
Productivity- Results
10-114--69
10-16-69
X
19-5
18.6
16.8
15-4
15-6
13-5
X/Control
1.44
1.38
1.24
l.l4
1.16
1.00
X
4l.4
31-5
30.3
20.2
19-9
20. 4
X/Control
2.03
1.54
1.49
.99
.98
1.00
ANOVA
10-1^-69
F
17-79**
3.09
10-16-69
F
2.91
2.88
Significant Simple Effects
10-14-69 10-16-69
F F
7.62*
10.6*
8.2* 35-4**
132
-------
CABLE 29
EXPERIMENT 18: BIOMASS ESTIMATES
Species Treatments
Initial Control N+P
15.7IM 38X
6.08 24.3 22.1
36.4 29.1 49-1
16.7 18.3
.154 .194
2-70 3.90
54.3 67-5
Cryptomonas ovata 50.8 75«2
Cryptcmonas erosa 48.8 37.^
Total 213.5 276.2 400.8
MG C/(irPx4hr) 23-8 35-1
Oscillatoria rubescens disappeared from the flasks but
notfrom the columns, while Fragilaria crotonensis and
Mougeotia sp. grew better in tlTe flasks than in the
columns.
The jar environment differed from the jug environment
in four major ways: (1) greater surface area to volume
ratio, (2) access to sunlight unaltered by passage
through water or glass, (3) opportunity for gas exchange
with the atmosphere, and (4) opportunity for wind-
induced turbulence. Which one or which combination of
these factors favored Chroomonas is uncertain, espec-
ially since an abundant population of this organism
(10.9 yg/1) was found under 30 cm of snow-covered ice
on February 3, 1970. From this observation it appears
that Chroomonas acuta does not require high light in-
tensities, gas exchange, or high turbulence, and its
abundant occurrence in the open lake water in the summer
implies no need for close association with a surface.
If pure speculation is employed briefly, interspecific
interactions such as predation or competition can be
proposed to explain the different responses of this
species to the two environments. Both of these hypo-
thetical mechanisms depend upon the reduced turbulence
in the jug environment, which could allow planktonic
133
-------
organisms to sink toward the bottom. Flagellates such
as Chroomonas acuta could be expected to settle less
rapidly than nonmotile species, and could, after a
sufficient period of time, be the major remaining com-
ponents of the phytoplankton in the open water in the
jugs. Swimming grazers would then feed selectively
upon these flagellates because they were the most
available prey.
If the Chroomonas cells settled with the other species,
competition for nutrients could occur in the concen-
trated mass of organisms accumulated at the bottom of
the jug. If Chroomonas were a poor competitor, it could
decline as a result.Calculations by Hurlburt (1970)
indicated that competition for nutrients in plankton
communities could not occur unless cell densities ex-
ceeded 3 x 108 per liter, at which point nutrient de-
pleted zones surrounding cells could overlap. In Third
Sister Lake cell densities rarely exceeded 1 x 106 per
liter. Therefore to achieve densities conducive to
competition all of the organisms present in a 19 liter
jug at a density of 1 x 10s per liter would have to
settle into a bottom layer of approximately 60 ml
volume. This seems unlikely, particularly since the
jug contents were thoroughly mixed during sampling for
each lkC run.
Studies of the Jug Environment
During the course of the study various attempts were
made to identify factors that contributed to the con-
tainment effects exhibited by the jugs. While none of
these efforts produced an explanation for the response
of Chroomonas acuta to the jug environment, they will
be discussed to indicate the approaches that were tried.
It was mentioned at the beginning of this section that
the chemical environment within a jug could change with
time, due to the lack of interchange of water with the
outside. In experiments 1-6 alkalinity and pH were
frequently measured in the jugs (Table 30), Neither
parameter appeared to change with time within a jug or
to differ significantly between treatment and control
jugs.
It was mentioned that radiation passing through the jug
walls could be modified and consequently modify the jug
environment. One result could have been a greenhouse
effect, but when temperatures inside the jugs were com-
pared to the ambient lake temperatures no differences
were found (Table 30).
134
-------
TABLE 30
PHYSICO-CHEMICAL CONDITIONS IN
CONTROL AND HIGH TREATMENT JUGS
Parameter
ALK*
PH
A_,K
ALK
pH
Temp.
ALK
pH
Temp.
ALK
pH
ALK
pH
c
Jug
Control
HHH
Control
HHH
Control
ffiiil
Control
II
Control
H
Control
H
Contiul
HH
Control
HH
Control
HH
Control
HHH
Control
HHH
Control
HHH
Control
HHH
Expt.
1
1
1
1
2
2
3
3
3
3
3
3
k
k
k
k
k
k
5
5
5
5
6
6
6
6
Productivity Run
1
8?. h
86.0
81.0
8l.O
78.0
78.5
8.16
8.13
26.1
26.1
78.0
76.0
8.20
8.13
20.5
20.5
82.0
81.0
86.0
86.0
2
87.7
Qh.Q
79-0
80.0
77-0
78.0
8.26
a. zk
25.6
25.5
78.0
76.0
8.10
8.10
21.6
21.3
83.2
82.0
7.81
7-85
89.0
83.0
8.6
3 U
86.0 86.0
86.0 87.0
8.2
8.2
80.0
80.0
(identical to
(identical to
86.0 81.0
81*. 0 82.0
a.k 8.65
8.7 8.7
5
lake)
lake)
84.0
86.0
8.26
8.50
* mg CaCO^/l
135
-------
There were variations in wall thickness among the jugs,
which ranged in weight from 10 Ib to 16 Ib 12 oz. It
was at one time suspected that light could have been
modified more by passing through the thicker walls
than through the thin walls, and that this could have
contributed to variance among the jugs. In experiment
4 wall thickness (inferred from jug weight) was in-
cluded as a testable variable, and no effects on the
productivity results were found.
Since reduction in turbulence was suspected as a factor
contributing to the containment effect, various mixing
regimes were tested to see if they altered productivity
results. In experiment 30 an extra pair of control jugs
was included that was not sampled until the day of the
last ll*C run. These jugs remained undisturbed in the
lake for 5 days while the other jugs in the experiment
were sampled on days 2, 3, and 5. The productivity re-
sults for the unmixed controls were insignificantly
different from the results for the normal controls on
day 5. In experiment 32 half of the jugs were mixed
three times per day while the other half were mixed once
a day. Again no productivity differences appeared.
The possibility that the sulfuric acid-chromate solution
used to clean the jugs was leaving toxic residues was
tested in experiments 31 and 32, in which jugs cleaned
with this solution were compared with jugs cleaned with
HC1 and with water only. No productivity differences
appeared.
Finally, in experiments 30 and 32 attempts were made to
remove the zooplankton from certain jugs so that preda-
tion effects could be studied. Results were erratic,
indicating that the technique used for zooplankton re-
moval (straining the lake water through a # 20 net) was
disrupting the phytoplankton community as well.
SIGNIFICANCE OF CONTAINMENT EFFECTS
Although the series of enrichment experiments conducted
in this study produced many examples of containment
effects and demonstrated that they can be highly and
consistently selective for particular species, notably
cryptomonads, the mechanisms of the effects were not
explained. Attempts to evaluate the general significance
of containment effects were more successful, as shown
in the analysis of the data for experiments 25-S and
34-S. These results indicated that containment effects
do not totally negate all value claimed for in situ
136
-------
enrichment experiments in interpreting or predicting
responses to enrichment in the natural system, but
rather they limit the precision of the predictions that
can be made. Substances may be tested to determine if
they might affect algal growth in a given system, but
conclusions about how much growth could result from a
given amount introduced into the natural system cannot
be drawn.
137
-------
SECTION VIII
ACKNOWLEDGEMENTS
The field portions of this study were conducted at the
University of Michigan. Numerous faculty members and
students in The Department of Environmental Health,
School of Public Health assisted with the project, and
are acknowledged with sincere thanks.
The cooperation of the Department of Forestry, School
of Natural Resources, who permitted the use of Third
Sister Lake was greatly appreciated. Deserving of par-
ticular gratitude is Professor Frank F. Hooper, Depart-
ment of Wildlife and Fisheries, who initially suggested
the use of this ideal research site.
Data analysis facilities made available at the Virginia
Institute of Marine Science were invaluable in the final
stages of the project.
The support of the Water Quality Office, Environmental
Protection Agency, and the help provided by Dr. Kenneth
W. Malueg and Dr. Charles F. Powers, the Grant Project
Officers, is acknowledged with sincere thanks.
139
-------
SECTION IX
REFERENCES
1. Abbott, W. 1966. Microcosm studies on estuarine
waters I. The replicability of microcosms.
J. Wat. Pollut. Control Fed. 38: 258-270.
2. Barlow, J.P., W.R. Schaffner, and F. de Noyelles,
Jr. 1971. Experimental Phytoplankton Success-
ion. Presented at the 162nd annual meeting of
the American Chemical Society, Washington, B.C.,
September 12-17, 1971. (abstract WATR-58).
3. Barnes, K. 1952. The use of transformations in
marine biological statistics. J. Cons. perm.
int. Explor. Mer 18: 61-71.
4. Barnett, A.M. and J. Hi rota. 1967. Changes in
the apparent rate of 1J+C uptake with length of
incubation period in natural phytoplankton
populations. Limnol. Oceanogr. 12: 549-353.
5. Bender, M.E. and R.A. Jordan. 1970. Plastic en-
closure versus open lake productivity measure-
ments. Trans. Am. Fish. Soc. 99: 607-610.
6. Biesinger, K.E. 1967. Micronutrients as possible
factors limiting primary productivity in cer-
tain Alaskan lakes. Ph.D. Thesis, Univ. of
Mich.
7. Bormann, F.H. and G.E. Likens. 1970. The nutri-
ent cycles of an ecosystem. Sci. Amer. 223:
92-101.
8. Bueltman, C.G., et al. 1969. Provisional algal
assay procedure. New York: Joint Industry/
Government Task Force on Eutrophication: 62 pp
9. Bunch, R.L. and M.B. Ettinger. 1967. Biode-
gradability of potential organic substitutes
for phosphates. Proc. 22nd industrial waste
conf., Purdue Univ.: 393-396.
10. Christie, A.E. 1970. Trisodium nitrilotriacetate
and algae. Water and Sewage Works 117: 58-59.
141
-------
11. Chu, S.P. 1942. The influence of the mineral
composition of the medium on the growth of
planktonic algae. Part I. Methods and cul-
ture media. J. Ecol. 30: 284-325.
12. Chu, S.P. 1943. The influence of the mineral
composition of the medium on the growth of
planktonic algae. Part II. The influence of
the concentration of inorganic nitrogen and
phosphate phosphorus. J. Ecol. 31: 109-148.
13. Desikachary, T.V. 1959. Cyanophyta. New
Delhi, Indian Council of Agricultural
Research: 686 pp.
14. Doty, M.S. and M. Oguri. 1958. Selected fea-
tures of the isotopic carbon primary produc-
tivity technique. In Measurements of primary
production in the sea, Rapp. P.-v Reun. Cons.
perm. int. Explor. Mer 144: 47-55.
15. Dugdale, R.C. 1967. Nutrient limitation in
the sea: Dynamics, identification and
significance. Limnol. Oceanogr. 12: 685-695.
16. Duthie, J.R. 1972. Detergents: nutrient con-
siderations and total assessment. In Likens,
G.E. (ed.). Nutrients and eutrophication:
the limiting-nutrient controversy. Amer.
Soc. of Limnol. and Oceanogr. Special Symposia
Vol. I: 205-216.
17. Edmondson, W.T. 1956. The relation of photo-
synthesis by phyuoplankton to light in lakes.
Ecology 37:
18. Edmondson, W.T. 1970. Phosphorus, nitrogen, and
algae in Lake Washington after diversion
of sewage. Science 169: 690-691.
19. Edmondson, W.T. and Y.K. Edmondson. 1946-7.
Measurement of production in fertilized salt-
water. J. Mar. Res. 6: 228-245.
20. Eggleton, F.E. 1931. A limnological study of
the profundal bottom fauna of certain fresh-
water lakes. Ecol. Monogr. 3: 231-331.
142
-------
21. Environmental Protection Agency. National
Eutrophication Research Program. 1971.
Algal assay procedure bottle test. Washing-
ton, B.C.: U.S. Govt. Printing Office, 82pp.
22. Eppley, R.W., J.L. Coatsworth and L. Solorzano.
1969. Studies of nitrate reductase in marine
phytoplankton. Limnol. and Oceanogr. 14:
194-205.
23. Erickson, S.J., I.E. Maloney, and J.H. Gentile.
1970. Effect of nitrilotriacetic acid on
the growth and metabolism of estuarine phyto-
plankton. J. Wat. Pollut. Control Fed. 42:
R329-R33S.
24. Fitzgerald, G.P. 1968. Detection of limiting
or surplus nitrogen in algae and aquatic weeds.
J. Phycol. 4: 121-126.
25. Fitzgerald, G.P. 1969. Field and laboratory
evaluations of bioassays for nitrogen and
phosphorus with algae and aquatic weeds.
Limnol. and Oceanogr. 14: 206-212.
26. Fitzgerald, G.P. and T.C. Nelson. 1966. Extrac-
tive and enzymatic analysis for limiting or
surplus phosphorus in algae. J. Phycol. 2:
32-37.
27 Fogg, G.E. 1958. Extracellular products of phy-
toplankton and estimation of primary production.
In Measurements of primary production in the
sea, Rapp. P.-v. Reun. Cons. perm. int. Explor.
Mer 144: 56-60.
28. Fogg, G.E. and Than-Tun. 1960. Interrelations
of photosynthesis and assimilation of elemen-
tary nitrogen in a blue-green alga. Proc.
Roy. Soc. B, 153: 111-127.
29. Gachter, R. 1968. Phosphorhaushalt and(|plank-
tische Primarproduktion im Vierwaldstattersee
(Horwer Bucht). Schweiz. Z. Hydrol. 30: 1-66.
30. Glooschenko, W.A. and J.E. Moore. 1971. The
effect of citrate upon phytoplankton in Lake
Ontario. Presented at the 162nd annual meeting
of the American Chemical Society, Washington
D C., September 12-17, 1971 (abstract WATR-72).
143
-------
31. Goldman, C.R. 1960a. Molybdenum as a factor
limiting primary productivity in Castle Lake,
California. Science 132: 1016-1017.
32. Goldman, C.R. 1960b. Primary productivity and
limiting factors in three lakes of the Alaska
Peninsula. Ecol. Monogr. 30: 207-230.
33. Goldman, C.R. 1962. A method of studying nu-
trient limiting factors in situ in water
columns isolated by polyetfhylene film. Limnol.
Oceanogr. 7: 99-101.
34. Goldman, C.R. 1964. Primary productivity and
micronutrient limiting factors in some North
American and New Zealand lakes. Verh. int.
Verein. theor. angew. Limnol. 15: 365-374.
35. Goldman, C.R. 1972. The role of minor nutrients
in limiting the productivity of aquatic
ecosystems. In Likens, G.W. (ed.). Nutrients
and eutrophication: the limiting-nutrient con-
troversy. Amer. Soc. of Limnol. and Oceanogr.
Special Symposia Vol. I: 21-33.
36. Goldman, C.R. and R. Armstrong. 1969. Primary
productivity studies in Lake Tahoe, Cali-
fornia. Verh. int. Verein. theor. angew.
Limnol. 17: 49-71.
37. Goldman, C.R. and R.C. Carter. 1965. An investi-
gation by rapid carbon-14 bioassay of factors
affecting the cultural eutrophication of Lake
Tahoe, California-Nevada. J. Wat. Pollut.
Control Fed. 37: 1044-1059.
38. Goldman, C.R. and R.G. Wetzel. 1963. A study of
the primary productivity of Clear Lake, Lake
County, California. Ecology 44: 283-294.
39. Hamilton, D.H. 1969. Nutrient limitation of
summer phytoplankton growth in Cayuga Lake.
Limnol. and Oceanogr. 14: 579-590.
40. Hamilton, R.D. 1972. The environmental accept-
ability of NTA: current research and areas
of concern. In Likens, G.E. (ed.). Nutrients
and eutrophication: the limiting-nutrient
controversy. Amer. Soc. of Limnol. and
Oceanogr. Special Symposia Vol. I: 217-221.
144
-------
41. Hasler, A.D. 1947. Eutrophication of lakes by
domestic drainage. Ecology 28: 383-395.
42. Huber-Pestalozzi, G. 1938. Das Phytoplankton
des Susswassers, 1. Teil. Allgemeiner Teil,
Blaualgen, Bakterien, Pilze. Stuttgart, E.
Schweizerbart'sche Verlagsbuchhandlung: 342 pp
43. Huber-Pestalozzi, G. 1941. Das Phytoplankton
des Susswassers, 2. Teil, 1. Halfte.
Chrysophyceen, Farblose Flagellaten, Heter-
okonten. Stuttgart, E. Schweizerbart'sche
Verlagsbuchhandlung: 365 pp.
44. Huber-Pestalozzi, G. 1942. Das Phytoplankton
des Susswassers, 2. Teil, 2. Halfte. Diatomen.
Stuttgart, E. Schweizerbart'sche
Verlagsbuchhandlung: pp. 366-549.
45. Huber-Pestalozzi, G. 1961. Das Phytoplankton
desf)Susswassers, 5. Teil. Chlorophyceae
(Grunalgen) Ordnung: Volvocales. Stuttgart,
E. Schweizerbart'sche Verlagsbuchhandlung:
744 pp. + 158 Tafeln.
46. Huber-Pestalozzi, G. 1968. Das Phytoplankton
des Susswassers, 3. Teil. Cryptophyceae,
Chloromonadophyceae, Dinophyceae. Stuttgart,
E. Schweizerbart'sche Verlagsbuchhandlung:
322 pp.
47. Hulburt, E.M. 1970. Competition for nutrients
by marine phytoplankton in oceanic, coastal,
and estuarine regions. Ecology 51: 475-484.
48. Hutchinson, G.E. 1941. Limnological studies in
Connecticut IV. The mechanisms of inter-
mediary metabolism in stratified lakes. Ecol.
Monogr. 11: 22-60.
49. Hutchinson, G.E. 1957. A Treatise on Limnology
Vol. I. Geography, Physics, Chemistry. New
York, Wiley: 1015 pp.
50. Hutchinson, G.E. 1967. A Treatise on Limnology
Vol. II. Introduction to Lake Biology and
the Limnoplankton. New York, Wiley: 1115 pp,
145
-------
51. Jackson, H.W. and M.E. Bender. 1964. Freshwater
Plankton No. 1 - Report of a study conducted
by the Analytical Reference Service Training
Program. U.S. Dept. H.E.W., Publ. Health Serv.,
Bureau of State Services, Division of Water
Supply and Pollution Control, R.A. Taft Sanit.
Eng. Center, Cincinnati, Ohio.
52
Jaworski, N.A., D.W. Lear, Jr., and 0. Villa, Jr.
1971. Nutrient management in the Potomac
estuary. In Likens, G.E. (ed.). Nutrients and
eutrophication; the limiting-nutrient contro-
versy. Amer. Soc. of Limnol. and Oceanogr.
Special Symposia Vol. I: 246-269.
53. Johnston, R. 1964. Sea water, the natural
medium of phytoplankton II. Trace metals
and chelation, and general discussion. J. mar.
biol. Ass. U.K. 44: 87-109.
54. Kalff, J. 1971. Nutrient limiting factors in
an Arctic tundra pond. Ecology 52: 655-659.
55. Kemmerer, A.J. 1968. A method to determine
fertilization requirements of a small sport
fishing lake. Trans. Am. Fish. Soc. 97:
425-428.
56. Ketchum, B.H. 1939. The absorption of phosphate
and nitrate by illuminated cultures of
Nitzschia closterium. Am. J. Bot. 26: 399-407.
57. Lange, W. 1971. Limiting nutrient elements in
filtered Lake Erie water. Wat. Res. 5: 1031-
1048.
58. Lee, G.F. 1970. Eutrophication. U. of Wis. Water
Resources Center, Eutrophication Information
Program, Occasional Paper No. 2: 39 pp.
59. Lund, J.W.G. 1965. The ecology of the freshwater
phytoplankton. Biol. Rev. 40: 231-293.
60. Lund J.W.G. and J.F. Tailing. 1957. Botanical
limnological methods with special reference
to the algae. Bot. Rev. 23: 489-583.
61. Mackenthun, K.M., W.M. Ingram, and R. Porges.
1964. Limnological aspects of recreational
lakes. U.S. Public Health Service Publication
No. 1167. Washington, D.C.: U.S. Govt.
Printing Office,176 pp.
146
-------
62. McAllister, C.D. 1961. Decontamination of
filters in the ltfC method of measuring marine
photosynthesis. Limnol. Oceanogr. 6: '447-450.
63. McLaren, I.A. 1969. Primary production and
nutrients in Ogac Lake, a landlocked fjord on
Baffin Island. J, Fish. Res. Bd. Can. 26:
1561-1576.
64. Menzel, D.W., E.M. Hulburt, and J.H. Ryther.
1963. The effects of enriching Sargasso Sea
water on the production and species composition
of the phytoplankton. Deep-Sea Res. 10:
209-219.
65. Menzel, D.W. and J.H. Ryther. 1961. Nutrients
limiting the production of phytoplankton in
the Sargasso Sea, with special reference to
iron. Deep-Sea Res. 7: 276-281.
66. Michigan Conservation Department. Division of
Fisheries. Institute for Fisheries Research.
1940. Lake inventory map - Crystal Lake.
67. Mitchell, D. 1971. Algal bioassays for estimating
the effect of various chemical compounds upon
the planktonic algal population. Presented
at the 162nd annual meeting of the American
Chemical Society, Washington, D.C., September
12-17, 1971 (abstract WATR-67).
68. Moss, B. 1969, Limitation of algal growth in
some central African waters. Limnol. and
Oceanogr. 14: 591-601.
69. Murphy, J. and J.P. Riley. 1962. A modified
single solution method for the determination
of phosphate in natural waters. Anal. Chim.
Acta. 27: 31-36.
70. Patrick, R. and C.W. Reimer. 1966. The Diatoms
of the United States Exclusive of Alaska and
Hawaii, Vol. 1. Philadelphia, Acad. Nat. Sci.:
688 pp.
71. Pollard, R.R. 1966. Amino acid chelating agents
in detergent applications. Soap and Chemical
Specialties 42: 58-62.
147
-------
72. Powers, C.F., D.W. Schults, K.W. Malueg, R.M.
Brice, and M.D. Schuldt. 1972. Algal respon-
ses to nutrient additions in natural waters.
II. Field experiments. In Likens, G.E. (ed.).
Nutrients and eutrophication: the limiting-
nutrient controversy. Amer. Soc. of Limnol.
and Oceanogr. Special Symposia Vol. I: 141-
154.
73. Prescott, G.W. 1962. Algae of the Western Great
Lakes Area. Dubuque, Wm. C. Brown Co.: 977 pp.
74. Rawson, D.S. 1939. Some physical and chemical
factors in the metabolism of lakes. A.A.A.S.
Bull. 10: 9-26.
75. Rodhe, W. 1948. Environmental requirements of
fresh-water plankton algae. Experimental
studies in the ecology of phytoplankton. Symb.
bot. upsal. 10: 1-149.
76. Rodhe, K. 1958. The primary production in lakes:
Some results and restrictions of the 14C
Method. In Measurements of primary produc-
tion in the sea, Rapp. P.-v Reun. Cons. perm.
int. Explor. Mer 144: 122-128.
77. Ryther, J.H. and W.M. Dunstan. 1971. Nitrogen,
phosphorus, and eutrophication in the coastal
marine environment. Science 171: 1008-1013.
78. Sakamoto, M. 1971. Chemical factors involved in
the control of phytoplankton production in
the experimental lakes area, northwestern
Ontario. J. Fish. Res. Bd. Can. 28: 203-213.
79. Saraceni, C. and D. Ruggiu. 1969. Techniques
for sampling water and phytoplankton. In
Vollenweider, R.A. (ed.). A Manual on Methods
for Measuring Primary Production in Aquatic
Environments, I.B.P. Handbook No. 12.
Philadelphia, F.A. Davis: p. 7.
80. Saunders, G.W. 1957. Interrelations of dissolved
organic matter and phytoplankton. Bot. Rev.
23: 389-409.
81. Saunders, G.W., F.B. Trama, and R.W. Bachmann.
1962. Evaluation of a modified Clk technique
for shipboard estimation of photosynthesis in
large lakes. Publs. gt. Lakes Res. Div. 8: 61 pp
148
-------
82. Schelske, C.L. 1960. Iron,
organic
and
matter ,
other factors limiting primary productivity
in a marl lake. Science 136: 45-46.
83. Schelske, C.L. and E.F. Stoermer. 1971. Eutro
phication, silica depletion, and predicted
changes in algal quality in Lake Michigan.
Science 173: 423-424.
84
85
86
Schelske, C.L. and E.F. Stoermer. 1972. Phos-
phorus, silica, and eutrophication of Lake
Michigan. In Likens, G.E. (ed.). Nutrients
and eutrophication: the limiting-nutrient
controversy. Amer. Soc. of Limnol. and
Oceanogr. Special Symposia Vol. I: 157-170.
Schering, K. 1931. Dosage des Nitrates dans
1'eau potable. Annales de Chimie Analytique
et de Chimie Appliquee 36: 188.
et de
Shumate, K.S., J.E. Thompson,
and C.L. Dean. 1970. NTA
ted sludge - field study.
Control Fed. 42: 631-640.
J.D. Brookhard,
removal by activa-
J. Wat. Pollut.
87
88
89
Soeder, C.J. and J.F. Tailing. 1969. The enclo-
sure of phytoplankton populations. In
Vollenweider, R.A. (ed.). A Manual on Methods
for Measuring Primary Production in Aquatic
Environments, I.E.P. Handbook No. 12. Phila-
delphia, F.A. Davis: p. 65.
Sorensen, T. 1948. A method of establishing
groups of equal amplitude in plant sociology
based on similarity of species content and
its application to analyses of the vegetation
of Danish Commons. Biol. 'Skr. 5: 34 pp.
Stanford, G., C.B. England, andA.W. Taylor.
1970. Fertilizer use and water quality. U.S.
Dept. of Agriculture, Agricultural Research
Service Pub. ARS 41-168: 19 pp.
90
Steel, R.G.D. and J.H. Torrie. 1960
and Procedures of Statistics with
erence to the Biological Sciences
McGraw-Hill: 481 pp.
Principles
Special Ref-
New York,
149
-------
91. Steemann-Nielsen, E. 1958. Experimental methods
for measuring organic production in the sea.
In Measurements of primary production in the
sea, Rapp. P.-v Reun. Cons. perm. int. Explor.
Her 144: 38-46.
92. Stewart, K.M. and G.A. Rohlich. 1967. Eutro-
phication - a review. California State Water
Quality Control Board, Pub. 34: 188 pp.
93. Sturm, R.N. and A.G. Payne. 1971. Environmental
testing of trisodium nitrilotriacetate:
bioassays for aquatic safety and algal stimu-
lation. Presented at the 162nd annual meeting
of the American Chemical Society, Washington,
B.C., September 12-17, 1971 (abstiact WATR-69),
M.
94. Thomas, E.A. 1961. Vergleiche uber die Plankton-
produktion in Flaschen und im Plankton-Test-Lot
nach Thomas. Verh. int. Verein. theor. angew.
Limnol. 14: 140-146.
95. Thomas, E.A. 1964. Nahrstoffexperimente in
Plankton-Test-Loten (1958). Verh. int.
Verein. theor. angew. Limnol. 15: 342-351.
96. Thompson, J.E. and J,R. Duthie. 1968. The bio-
degradability and treatability of NTA. J.
Wat. Pollut. Control Fed. 40: 306-319.
97. Tranter, D.J. and B.S. Newell. 1963. Enrichment
experiments in the Indian Ocean. Deep-Sea
Res. 10: 1-10.
98, U.S. Congress. House. Committee of Government
Operations. Conservation and Natural Resources
Subcommittee. 1970. Phosphates in detergents
and the eutrophication of America's waters.
Washington, D.C.: U.S. Govt. Printing Office,
318 pp.
99. U.S. Dept. H.E.W. Publ. Health Serv. Bureau of
State Services. Division of Radiological
Health. 1960. Radiological Health Handbook.
Washington, D.C.: 468 pp.
100. U.S. Dept. Int. F.W.P.C.A. Div. of Water Quality
Res. Analytical Quality Control Lab. 1969.
FWPCA Methods for Chemical Analysis of Water
and Wastes. Cincinnati: 280 pp.
150
-------
101.
102.
103.
104.
105.
Vallentyne, J.R. 1957. The molecular nature of
organic matter in lakes and oceans, with
lesser reference to sewage and terrestrial
soils. J. Fish. Res. Bd. Can. 14: 33-82.
Vollenweider, R.A. 1969. Preparation of working
solutions. In Vollenweider, R.A. (ed.). A
Manual on Methods for Measuring Primary Pro-
duction in Aquatic Environments, I.B.P. Hand-
book No. 12. Philadelphia, F.A. Davis: p. 54.
Vollenweider, R.A. and A. Nauwerck. 1961. Some
observations on the Clk method for measuring
primary production. Verh. int. Verein. Theor.
angew. Limnol. 14: 134-139.
H. and J. Lev. 1953. Statistical Infer-
Walker,
ence. New
Winston.
York: Holt, Rinehart, and
106.
Walleri, D.G. and G.H. Geen. 1968. Loss of radio-
activity during storage of C1^-labelled phyto-
plankton on membrane filters. J. Fish. Res.
Bd. Can. 25: 2219-2224.
Warren, C.B. and E.J. Malec. 1972. Biodegrada-
tion of nitrilotriacetic acid and related
imino and amino acids in river water. Science
176: 277-279.
107.
108.
West, F.L.S. and G.S. West. 1908.
of the British Desmidiaceae, Vol
the Ray Society: 274 pp.
A Monograph
, III. London,
Wetzel, R.G. 1966. Productivity and nutrient
relationships in marl lakes of northern
Indiana. Verh. int. Verein. theor. angew.
Limnol. 16: 321-332.
151
-------
SECTION X
PUBLICATIONS
1. Bender, M.E. and R.A. Jordan. 1970. Plastic
enclosure versus open lake productivity
measurements. Trans. Am. Fish. Soc. 99: 607-
610.
2. Jordan, R.A. 1970. Factorial enrichment experi-
ments in a southeastern Michigan lake. Ph. D.
thesis, Univ. of Mich.
5. Jordan, R.A. and M.E. Bender. 1972. Stimulation
of phytoplankton growth by mixtures of phosphate
nitrate, and organic chelators. Water Res.
6: (in press).
153
-------
SECTION XI
APPENDICES
Appendix
A Productivity Results,
B
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
1
7
4~l
3
4
5
6
7
8
9
10
11
12
13
Experiment
Experiment
Experiment
Experiment
Experiment
Experiment
Experiment
Experiment
Experiment
Experiment
Experiment
Experiment
Experiment
Experiments ]
1,
2,
3,
4,
5,
6,
7,
8,
9,
16,
17,
19,
20,
6-12-68 -
7-18-68 -
8-7-68 -
9-19-68 -
10-27-68-
4-23-69 -
5-26-69 -
6-4-69 -
7-1-69 -
8-21-69 -
9-13-69 -
10-25-69-
4-23-70 -
.-20
6-19-68
7-22-68
8-10-68
9-23-68
11-4-68
5-5-69
6-2-69
6-11-69
7-7-69
8-25-69
9-18-69
11-4-69
5-3-70
Page
157
158
160
162
164
165
167
169
171
173
175
177
179
181
Phytoplankton Species Biomass Estimates,
Experiments 1-19 183
Table 1: Experiment 1: Biomass Estimates
(yg/D 184
Table 2: Experiment 2: Biomass Estimates
(yg/D 185
Table 3: Experiment 3: Biomass Estimates
(yg/D 186
Table 4: Experiment 4: Biomass Estimates
(yg/1) 187
Table 5: Experiment 5: Biomass Estimates
(pg/1) 188
Table 6: Experiment 6: Biomass Estimates
(yg/D 189
Table 7: Experiment 7: Biomass Estimates
(yg/D 190
Table 8: Experiment 8: Biomass Estimates
(yg/1) 191
Table 9: Experiment 9: Biomass Estimates
(yg/1) 192
Table 10: Experiment 16: Biomass Estimates
(yg/D 193
Table 11: Experiment 17: Biomass Estimates
(yg/D 194
Table 12: Experiment 19: Biomass Estimates
(yg/D 195
155
-------
Appendix Pa§.e.
C
D
E
Productivity Results, Experiments 22-34
Table 1: Experiment 22, 5-12-70 - 5-15-70
Table 2: Experiment 23, 5-19-70 - 5-21-70
Table 3: Experiment 24, 5-25-70 - 5-29-70
Table 4: Experiment 25, 6-17-70 - 6-22-70
Table 5: Experiment 29, 8-21-70 - 8-26-70
Table 6: Experiment 30, 10-24-70- 10-29-70
Table 7: Experiment 31, 5-17-71 - 5-21-71
Table 8: Experiment 32, 5-24-71 - 5-28-71
Table 9: Experiment 34, 6-9-71 - 6-14-71
Productivity Results, Sewage Experiments
Table 1: Experiment 6-S,
4-23-69 - 5-5-69
Table 2: Experiment 24-S,
5-25-70 - 5-29-70
Table 3: Experiment 25-S,
6-17-70 - 6-22-70
Table 4: Experiment 29-S,
8-21-70 - 8-26-70
Table 5: Experiment 34-S,
6-9-71 - 6-14-71
Phytoplankton Species Biomass Estimates,
Sewage Experiments
Table 1: Experiment 6-S: Species Biomass
Estimates, Treatment Means-yg/1
Table 2: Experiment 6-S: Species (Cont.),
Treatment Effects (Tukey's Test)
Table 3: Experiment 25-S: Species Biomass
Estimates, Treatment Means-yg/1
Table 4: Experiment 25-S: Species (Cont.),
Significant Treatment Effects
(Tukey's Test)
Table 5: Experiment 34-S: Species Biomass
Estimates, Treatment Means-yg/1
Table 6: Experiment 34-S: Species ANOVA
197
198
199
200
202
204
206
208
209
210
213
214
215
216
217
218
219
220
221
222
223
224
225
156
-------
APPENDIX A
Productivity Results
Experiments 1-20
157
-------
TABLE 1
EXPERIMENT 1
6-12-68 - 6-19-68
Enrichment
p0i| -
- P
EDTA
Design
Concentration
100 p.g/1
250 iig/1
20 ng/1
50 jig/1
5 mg/1
2.5 mg/1
reatment
K2+P-|_+EDTA2
Control
Productivity Results - mgC/(ia3x4hr)
6-13-68
6-15-68
6-19-68
X
65.6
71.4
53-2
63.1
59-2
63.2
52.0
6o.3
31.8
X/Control
2.06
2.24
1.67
1.98
1.86
1.99
1.63
1.90
1.00
X
204.9
157.5
117.2
106.6
95.8
80.5
120.8
106.0
42.7
X/Control
4.80
3.69
2.74
2.50
2.24
1.88
2.83
2.48
1.00
X
175-8
162.3
11*0.0
178.2
97.6
93.2
115.8
79.2-
17-3
X/Control
10.16
9.38
8.09
10.30
5.64
5-39
6.69
4.58
1.00
ANOVA
Source
N
P
EDTA
KP
K EDTA
P EDTA
NP EDTA
6-13-68
F
2.37
5-19
4.26
.84
.12
.57
.01
6-15-68
F
69.86**
16.19**
74.28**
1.62
2.92
2.73
6-19-68
F
110.96**
-36
.40
.88
6.52
58
10.63*
158
-------
EXPERIMENT 1 (Continued)
Significant Simple Effects
6-13-68 6-15-68 6-19-68
Comparison F F F
_ _ 59- ^
- N1+P2+EDTA1 1^9.1** 29.0**
98.7** 37-1**
- N1+P1+EDTA]_
- N2-)-P-L+EDTA1 21.'
Nl+P2't":Er)TA2 ~ " ~
N1+P1+EDTA2 -
N-L+Pg+EDTAg -
K2+P1+EDTA2 - N2H
._ _ K2+p2+EI)TAi
* .05 level
** .01 level
159
-------
TABLE 2
EXPERIMENT 2
7-18-68 - 7-22-68
Design
Enrichment Concentration
1-103 ~ 1: 10° ^g/1
- P 20 i-ig/i
. 5 ng/1
EDTA
Productivity Results - ingC/(ni3x4hr)
Treatment
1HP+EDTA
K+P
i;+EDTA
T"
P+E3TA
p
EDTA
Control
7-20-68
X X/Control
290.2 6.5U
95-4 2.15
94.4 2.13
59-0 1-33
14. 3 -32
49.4 i.n
76.2 1.72
44.4 i.oo
7-21-68
X X/Control
13^-6
73-5
174.0
65.0
8.0
61.8
64.6
X
131
132
l48
56
82
70
86
8
7-2P-66
X/Control
.3
.4
.0
. 6
(-"
£.
.8
.0
9
AKOVA
Source
T,"
i\
P
EDTA
UP
II EDTA
P EDTA
KP EDTA
7-20-68
F
7.42*
.28
1.26
8.71*
4.88
1.78
1.97
7-22-6C
17
7.02*
3-57
6.50*
1.45
2.51
5.88*
.70
160
-------
EXPERIMENT 2 (Continued)
Significant Simple Effects
Comparison
K - Control
N+P - P
N+EDTA - EDTA
N+P+EDTA - P+EDTA
P - Control
P+N - W
P+EDTA - EDTA
P+N+EDTA ~ K+EDTA
EDTA - Control
EDTA+P - P
EDTA+E - K
EDTA+K+P - W+P
7-20-C
F
7-22-68
F
9.85*
10.C*
13. 6**
* .05 level
** .01 level
161
-------
TABLE 3
EXPERIMENT 3
8-7-68 - 8-lo-<
Design
Enrichment
EDTA
]_
Concentration
.5 mg/1
2.5 mg/1
Productivity Results -
8-8-68
Treatment
EDTA2
Control
X
72.4
' 62.8
55.7
X/Cont.
1.30
1.13
8-10-68
X X/Control
42.6
26.0
1-57
Source
Treatments
ANOVA
8-8-68
F
.820
8-10-68
F
1.28
Productivity Results -
8-8-68
Treatment
Rack 1
Rack 1
Rack 1
Rack 2
Rack 2
Rack 2
Rack 3
Rack 3
Rack 3
EDTAo
EDTA^
Control
EDTAg
EDTA-^
Control
EDTA2
EDTA-i
Control
X
67-5
43.5
31-9
79-2
77-3
66.9
70.6
67.7
68.4
X/Control
2.12
1.36
1.18
1.16
1.03
99
X
17.8
46.2
12.5
53.7
34.6
30.7
51.2
47.1
34.7
8-10-68
X/Control
.42
70
1.55
1.13
1.48
1.36
162
-------
EXPERIMENT 3 (Continued)
MOVA
8-8-68 8-10-68
Treatment F F
Rack 5.98- 14.94*
EDTA 2.08 4.33*
Significant Simple Effects
8-8-68 8-10-68
Comparison F F
R0 Control - R-^ Control 6.08*
R^ EDTA, - R-, EDTAn 5-66^
R" £DTA0 - R-, EPTA^ 10-9**
<- d -i- ^
O Control - R-L Control
^ EDTA-, - R] EDTA-j^
O EDTAo - RT EDTA0 9-6*
3
-------
TABLE k-
EXPERIMENT h
9-19-68 - 9-23-^
Enrichment
N03 - N
Poly - P-L
Poly - P2
Design
Concentration
100 ^ig/1
50 M-g/1
100 jig/1
Productivity Results - mgC/(nHxUhr)
Treatment
N+P2
N+P
Control
9-21-68
X
93-8
81.1+
50.0
86.1+
8l.it
X/Control
1.82
1.58
97
1.67
1.58
51.6
X
86.8
73.8
29.2
72.6
77. k
1.00
9-23-68
X/Control
2.140
2.0k
.81
2.01
2.14
1.00
ANOVA
Source
N
P
HP
9-21-68
F
.09
35.96**
.03
9-23-68
F
.29
128.143**
14.96
Comparison
N - Control
P-j^ - Control
P-j^+N - N
P2 - Control
P+N - N
Significant Simple Effects
9-21-68 9-23-68
19.5**
22.25**
25.5**
37-5**
72.5**
108.2**
6l.l4**
150.0**
* .05 level
** .01 level
164
-------
TABLE 5
EXPERIMENT 5
10-27-68 - 11-4-68
Enrichment
WO^ - N
POi, - P2
- P
EDTA
Design
Concentration
100 ng/1
50 ug/1
10 ng/1
5 mg/1
Productivity Results - mgC/(m3x4hr)
(!0-30-68)-(lO-28-68) 11-M
Treatment
N+P2+EDTA
N+EDTA
K
P2+EDTA
EDTA
Control
Cation
Control
X
6.5
8.0
3-1
2.6
6
7
8
6.2
2.7
X/Control
09
09
78
.69
.58
.80
.60
.84
1-38
91
i.oo
.60
X
48.9
47.0
47.0
28.9
15-7
12.8
12.6
12.7
15.2
13-9
13-5
14.3
X/Cont:
3.42
3.29
3.29
2.02
1.10
90
.88
.89
1.06
97
94
1.00
13.7
.96
Source
N
P
EDTA
NP
N EDTA
P EDTA
NP EDTA
ANOVA
(lO-30-68)-(lo-28-68)
F
13.5&**
9.05**
54
9.05**
.08
.23
3.06
11-4-68
188.79**
46.33**
7.25*
51.29**
6.90*
4.08
2.68
165
-------
Significant Simple Effects
(10-30-68 )-(lo-28-68) n-U-68
Comparison F F
W - Control
N+P! - P., 18.25**
N+EDTA -jSDIA
- PESETA 10.9* 82.3**
- P2 95.5**
N+P2+EDIA - P2+EDTA 18.2** 107. 0**
P-j_ - Control
- N 12.1**
- N+EDTA 16.5** 79.7**
* .05 level
** .01 level
P2 - Control
P2+N - N 19.0** 9^.7**
P-«I+EDTA - N+EDTA 4.65* 90.0**
- N+P1 26.5**
166
-------
TABLE 6
EXPERIMEKT 6
11-23-69 - 5-5-6Q
Design
Treatment
E+P+EDTA
K+P
N+EDTA
N
P+EDTA
P
EDTA
Control
Cation
Control
Treatment
N+P+EDTA
N+P
N+EDTA
N
P+EDTA
P
EDTA
Control
Cation
Control
Pi
X
120.0
87.8
96.8
54-7
91.8
70.5
93.5
71.6
75.2
X
81.6
70.1
69.0
44.8
6i.4
63.6
68.0
51.1
42.0
Enrichment Concentration
NO- - N 25 y.g/1
PO^ - P 5 Hg/1
EDTA .5 mg/1
oductivity Results - mgC/(m:)x4hr^
4-25-69
X/Control
1.68
1.23
1.35
.76
1.28
.98
1.31
1.00
1.05
5-2-69
X/Control
1.60
1-37
1.35
.86
1.20
1.24
1.33
1.00
.82
4-28-69
X
73-6
57.2
61.1
42.2
58.8
43.0
53.5
44.6
50.0
X/Control
1.65
1.28
1.37
95
1.32
.96
1.20
1.00
1.12
)
4-30-69
X
88.2
76.4
84.2
57-7
8o.4
70.2
74.8
59-7
53-3
X/Control
1.48
1.28
l.4l
97
1.35
1.18
1.25
1.00
.89
5-5-69
X
75.2
82.0
65.8
53-2
62.6
64.3
66.6
6o.2
45.7
X/Control
1.25
1.36
1.09
.88
l.o4
1.07
l.ll
1.00
76
167
-------
EXPERIMENT 6 (Continued)
Source
H
P
EDTA
NP
W EDTA
P EDTA
NP EDTA
ANOVA
11-25-69 4-28-69 4-30-69 5-2-69 5-5-69
F
.99
2.78
13.39**
3-38
94
.11
.09
F
1.55
1.31
4.82
75
.15
.25
.12
F
1.60
5.18
13.92**
1.15
.59
1.31
34
F
2.72
11.08*
14.71**
5.89*
2.59
5-78*
.24
F
1.46
4.27
-32
4.21
.002
2.20
38
Comparison
N - Control
N+P - P
N+EDTA - EDTA
M+P+EDIA - P-HEDTA
P - Control
P+N - N
P+EDTA - EDTA
P+N+EDTA - N+EDTA
EDTA - Control
EDTA+P - P
EDTA+H - N
EDTA4W+P - N+P
Significant Simple Effects
4-25-69 4-28-69 4-30-69 5-2-69 5-5-1
14.6**
6.94-*
6.45*
9.65* 13-5**
* .05 level
** .01 level
168
-------
TABLE 7
EXPERIMENT 7
5-26-69 - 6-2-69
Enrichment
P01+ - P
EDTA
Design
Concentration
25 Hg/1
5 Hg/1
.5 mg/1
Productivity Results - mgC/(urxHhr)
Treatment
N+P+EDTA
N+P
N+EDTA
N
P+EDTA
P
EDTA
Control
Cation
Control
X
6l.9
55-5
57-9
U5-7
55-8
52.1
Li8.4
50.5
U5.0
5-27-69
X/Control
1.23
1.10
1.15
.90
1.10
1.03
.96
1.00
_ 5-29-69
X X/Control
2.13
1.68
1.7^
1.03
1.21
63-8
50.2
52.0
30.9
36.3
29-3
30.8
29.9
26.4
1.03
1.00
.88
6-2-(
X
2U.1
22.6
23-9
16.3
2^.0
22.8
17.9
X/Cont:
1.68
1.58
1.67
1.1U
1.68
1.59
1.25
18.8
1.00
1.31
AN OVA
Source
N
P
EDTA
NP
N EDTA
P EDTA
WP EDTA
5-27-69
F
2.85
7*^5*
5.78*
.32
k.12
.00
1.89
5-29-69
F
1+6.92**
12.13**
17.19**
6.50*
6.72*
.02
1.75
6-2-69
F
5-02
37.26**
16.23**
5.57*
1.59
6.26*
1.20
169
-------
EXPERIMENT 7 (Continued)
Significant Simple Effects
5-27-69 5~29-69 6-2-69
Comparison F F F
N - Control
N+P - P 16.6**
N+EDTA - EDTA 16.7** 17.8**
N+P+EDTA - P+EDIA 28.4**
P - Control 2*1.5**
P-fJJ - N 5.5* 1*1.1** 13.3**
P+EDTA - EDTA 12.2**
P+N+EDTA - N+EDTA
EDTA - Control
EDTA+P - P
EDIA+W - N 8.6* 16.7** 19-7**
EDTA+N+P - E+P 6.9*
* .05 level
** .01 level
170
-------
TABLE 8
EXPERIMENT 8
6-1+-69 - 6-11-69
Enrichment
Mb - K
EDTA
Design
Concentration
25 Hg/1
.5 mg/1
Productivity Results -
Treatment
K+P+EDTA
N+P
N+EDTA
N
P+EDTA
P
EDTA
Control
Cation
Control
Source
K
P
EDTA
UP
K EDTA
P EDTA
HP EDTA
X
54.8
55-8
28.2
27.0
46.5
147.5
27.8
27.2
25.0
6-6-69
X/Control
2.01
2.05
i.o4
99
1.71
1-75
1.02
1.00
.92
6-6-69
F
22.80**
705.30**
.003
21.53**
.02
1.16
.03
6-9-69
X
42.8
148.9
19-2
18.2
3^.7
32.1
18.5
18.0
17-
X/Control
2.38
2.72
1.07
1.01
1
1
93
78
1.03
i.oo
99
AWOVA
6-9-69
F
11.82**
127.06**
.08
10.18*
1.17
1.50
6-11-69
X
26.0
30.7
19.0
17.6
23.0
21.3
17-4
16.8
X/Control
1.55
1.83
1.13
1.05
1.37
1.27
i.oU
1.00
16.9
l.oi
6-11-69
F
7.20*
29.94**
.04
3.22
1.03
.86
1.70
171
-------
EXPERIMENT 8 (Continued)
Significant SIxiplc Effect,';
Comparison
K - Control
N+P - P
K+EDTA - EDTA
N+P+EDTA - P+EDTA
P - Control
P+E - K
P+EDTA - EDTA
P+N+EDTA - N+EDTA
EDTA - Control
EDTA+P - P
EDTA+N - N
EDTA+N+P - N+P
6-6-69
F
21.8*'*
21.8**
129**
21*6**
6-9-69
F
20.2**
lit. 2**
67.!**
18.9*-*
39.^'-*
6-11-69
F
11. ₯*
2.14**
* .05 level
** .01 level
172
-------
TABLE 9
EXPERIMENT 9
7-1-69 - 7-7-69
Design
Treatment
N+P+EDTA
N+P
M+EDTA
.U
P+EDTA
P
EDTA
Control
Cation
Control
Source
K
P
EDTA
IIP
U EDTA
P EDTA
NP EDTA
Enrichment
K03 - E
POj^ - P
EDTA
Productivity
7-2-69
X X/Control
70.7 1.69
70.3 1.68
56.3 1.34
45.1 1.08
65.0 1.55
hQ.6 1.16
49.3 1.18
41.9 l.oo
36.9 .88
7-2-69
F
194.87**
503.60**
175.18s*
48.20**
11.08*
.12
35.11-
Concentration
25 Hg/1
5 t-Lg/l
50 ng/1
Eesults - mgC/(m x4hr)
y-4-69
X X/Control X
26.1 1.80 12.2
26.5 1.83 11.5
24.6 1.70 11.5
20.6 1.42 8.93
17-3 1.19 13-2
14.7 1.01 10.2
14.1 .97 9.10
14.5 1.00 5.84
12.5 .86 5.17
AMOVA
7-4-69
F
173.02**
14.43**
4.43
1.98
-23
.20
6.77*
7-7-69
X/Control
2.09
1-97
1.97
1.53
2.26
1-75
1.56
1.00
.88
7-7-69
F
9.67*
39-42**
26.07**
7-51*
2.59
1.22
.80
173
-------
EXPERIMENT 9 (Continued)
Significant Simple Effects
7-2-69 7-^-69 7-7-69
Comparison F F F
N - Control 18.76** 10.78**
N+P - P 227.35** 69.55**
N+EDTA - EDTA 23.10** 5*1.96** 6.60*
N+P+EDTA - P+EDTA 33.92** 38.69**
P - Control 21.31** 21.31**
P+N - N 305.67** 17.05** 7.66*
P+EDTA - EDTA 118.'"
P+N+EDTA - N+EDTA
EDTA - Control 26.51** 11-99**
EDTA+P - P 130.33** 10.57*
EDTA+N - N 59.98** 7.91!* 7.55*
EDTA+N+P - N+P
* .05 level
** .01 level
174
-------
TABLE 10
EXPERIMEM1 l6
8-21-69 - 8-25-69
Enrichment
KOo - N
- P
EDTA
Design
Concentration
100 (J.g/1
3 US/1
.5 mg/1
Productivity Results -
Treatment
N+P+EDTA
N+P
N+EDTA
N
P+EDTA
P
EDTA
Control
Cation
Control
X
1+2.2
31-7
25-7
17-7
39-5
28.3
27.1
20.3
20*3
X/Cont:
2.08
1.56
1.27
.87
1.95
1.39
1.33
1.00
1.00
8-25-69
X
23-7
15-8
15-0
11.2
26.7
18. 5
15-1
12.3
10.1
X/Cont rol
1.93
1.28
1.22
.91
2.17
1.50
1.23
1.00
.82
AWOVA
Source
N
P
EDTA
NP
R EDTA
P EDTA
HP EDTA
8-22-69
F
25
71^.95**
5.76*
.02
2.66
.21
8-25-69
F
U.71
99-39**
52.97**
2.11
.04
9-38*
.19
175
-------
EXPERIMENT 16 (Continued)
Significant Simple Effects
8-22-69 8-25-69
Comparison F F
N - Control
N+P - P
N+E3TA - EDTA
N+P+EDTA - P+EDTA
P - Control ll*.6** 15-9**
P-fH - N k$.0** 8.93*-
P+EDTA - EDTA 3^-7** 55-8**
P+K+EDTA - H+EDTA 6l.O** 30.8**
EDTA - Control 10-5**
EDTA+P - P 28.k** 27-6**
EDTA+B - N li+.6** 5-87*
EDTA+N+P - K+P 2
* .05 level
** .01 level
176
-------
TABLE 11
EXPERIMENT 1?
9-13-69 - 9-18-69
Design
Enrichment
- P
EDTA
Concentration
25 Hg/1
5 m?/l
5 rag/1
Productivity Results - mgC/(m:)x4hr)
9-14-69
Treatment
K+P+EDTA
N+P
N+EDTA
W
P+EDTA
P
EDTA
Control
Cation
Control
X
30.
24.
23-
18.
22.
17-
21.
17-
13.
X/Control
3
8
3
3
6
2
1
5
1+
1
1
1
1
1
1
1
73
.42
33
.05
.29
.98
.21
.00
77
X
27
23
30
20
17
15
17
14
13
9-16-69
X/Control
7
.6
.8
.4
.3
.1
.5
.8
9
1.
1.
2.
1.
1.
1.
1.
1.
87
59
08
38
17
02
18
00
94
X
12
14
10
9
10
u
9
8
8
9-18-69
9
.6
3
.82
74
.5
.28
.96
.11
X/Control
1 ^-T-\-
1.63
1.15
1.10
1.20
1.28
l.o4
1.00
.91
AWOVA
9-14-69
Source
K
P
EDTA
TIP
K EDTA
P EDTA
KP EDTA
F
6? . 58**
42 . 93**
76,25**
29.61-*
.41
1.18
.37
9-16-69
F
53-00**
.001
1.27
9-18-69
F
13.48**
33-99**
73
3-15
.17
2.86
.36
177
-------
EXPERIMENT 17 ( Continued)
Significant Simple Effects
9-1U-69 9-16-69 9-18-69
Comparison F F F
N - Control
N+P - P 1+5.5** 10.5* 10.6*
H+EDTA - EDTA 25.2**
H+P+EDTA - P+EDTA 1*5-5** 15-7**
P - Control 7.15*
P+N - N 32.If** 2U.5**
P+EDTA - EDTA
P+N+EDTA - N+EDTA 38.h** 7-15*
EDTA - Control 10.2*
EDTA+P - P 23.M-*
EDTA+N - K 19.4** 16.0**
EDTA+N+P -
* .05 level
** .01 level
178
-------
TABLE 12
EXPERIMENT 19
10-25-69 - 11-4-69
Des ign
Enrichment Concentration
- II 25 M-g/1
- P 5 iig/1
5 mg/1
EDTA
Productivity Results - mgC/(in3x4hr)
10-26-69
Treatment
K+P+EDTA
K+r
K+EDTA
N
P+EDTA
P
EJ)TA
Control
Cation
Control
X X/Control
13-7
14.2
12.1+
12. 4
14.4
14. 7
13-2
12.9
12.9
1.06
1.10
.96
.96
1.12
l.lU
1.02
1.00
1.00
10-28-69
X X/Control
5-22
5.U6
4.21
3.94
5.55
5.30
4.21
4.00
3.51
1.31
1.37
1.05
99
1.39
1.32
1.05
1.00
.88
10-30-69
X
6.51
6.48
4.82
4.46
6.58
6.54
4.6i
4.34
4.o6
X/Control
1.50
1.49 .
1.11
1.03
1.52
1-51
1.06
1.00
.94
11-2-69
Treatment
I\i+P+EDTA
E+P
K+EDTA
E
P+EDTA
P
EDTA
Control
Cation
Control
X
4.58
4.38
2.71
3-01
3-99
5.04
2.94
3-37
2.70
X/Control
1.36
1.30
.80
.89
1.18
1.50
.87
1.00
.80
11-4-69
X
30.4
23-5
18.0
17.1
22.5
23.1
18.4
17.9
15.4
X/Control
1.70
1.31
1.01
.96
1.26
1.29
1.03
1.00
.86
179
-------
EXPERIMENT 19 (Continued)
AM OVA
Source
N
P
EDTA
KP
M EDTA
P EDTA
I:P EDTA
10-26-26
F
5-91*
37. 57**
.28
.02
.26
1.37
.00
10-28-69
jj
.05
23-77**
.20
.01
17
.19
.27
Significant Simple
Comparison
N - Control
N+P - P
N+EDTA - EDTA
N+P+EDTA - P+EDTA
P-Control
P+K - H
P+EDTA - EDTA
P+M+EDTA - 1M+EDTA
10-26-69
F
12.7-**
5-71*
13-5**
6.5*
10-28-69
F
6,02*
8.36*
6.4o*
10-30-69
F
.03
51.3****
.41
.18
.01
.26
.01
Effects
10-30-69
F
16.0**
13 . 4**
13.1**
9.44*
11-2-69
F
.64
51.26**
3-58
.40
2.72
.02
1.85
11-2-69
F
16.1**
10.8*
5-9*
20.1**
11-4-69
F
6.55*
99-33**
7.42*
11.19**
7-57*
3.o4
6.21*
11-4-69
F
30.9**
13 . 5**
23.0**
8.64*
75.6**
EDTA - Control
EDTA+P - P
EDTA-H-J - N
EDTA-+-K+P - N+P
24.1**
* .05 level
** .01 level
180
-------
TABLE 13
EXPERIMENT 20
4-23-70 - 5-3-70
Enrichment
N03 - N
PO^ - P
EDTA
Design
Concentration
25 tig/1
5 Hg/1
.382 mg/1
Productivity Results - mgC/(m3x4hr)
4-24-70
Treatment
EDTA
IHP
K
P
Control
Lake
x :
115-1
113-2
96.9
ni.4
110.8
114.6
X/Cont:
1.04
1.02
.87
1.01
1.00
1.03
X
4-25-70
X/Control
139-2
133.6
119.8
134.1
124.4
1.12
1.07
96
1.08
1.00
130.4 1.05
4-27-70
Treatment
EDTA
ii+P
n
p
Control
Lake
X
134.2
131.4
114.7
135.0
118.6
154.5
X/Control
1.13
l.ll
97
1.14
1.00
1.30
4-28-70
X
148.9
136.2
113-4
121.8
120.4
118.9
X/Control
1.24
1.13
.94
1.01
1.00
99
4-26-70
X
139-0
135-0
115-4
l4o.3
122.4
139-4
X
70.4
69.0
56.2
65.4
55.8
77-2
X/Control
1.13
1.09
.93
1.14
1.00
1.13
5-3-70
X/Control
1.26
1.24
1.01
1.17
1.00
1.38
Source
N
P
HP
4-24-70
F
3.55
6.93
6.06
4-25-70
F
.387
7.90*
.240
ANOVA
4-26-70
F
3-41
32.18**
.066
4-27-70
F
.342
6.78
.ooo4
4-28-70
F
.722
7.72*
5.98
5-3-70
F
.784
24.81**
.502
181
-------
Comparison
H - Control
E+P - P
P - Control
P+M - N
EXPERIMENT 20 (Continued)
Significant Simple Effects
70 U-25-70 U-26-70 U-27-70 14-28-70 5-3-70
F F F F F
18.25*
16.6*
EDTA
F
2.16
ANOVA
ij-25-70 U-26-70 ii-27-70 1^-28-70 5-3-70
r F F I' F
8.22 9-05 ^.814 51-97* 1^-51
Lake: Significantly higher than control on 5-3-70 (Tukey's Test)
* .05 level
** .01 level
182
-------
APPENDIX B
Phytoplankton Species Biomass Estimates
Experiments 1-19
183
-------
TABLE 1
EXPERIMENT 1: BIOMASS ESTIMATES (ug/1)
Species
Chroococcus dispersus
Ankistrodesmus falcatus
Rhabdoderma sigmoidea
Sy_nedra rump ens
Scene desmus biiuga
Chroomonas acuta .
10y sphere
Total
MG C/(m3 x 4hr)
Treatments
Initial
10.4
1.06
.944
7.06
T
T
2.01
21.5
Control
3.12
1.48
1.15
1.41
T
4.01
11.2
15.8
n+p+c
7.14
7.91
2.63
85.4
.578
T
74.2
177.8
79.2
N+p+c
6.10
9.89
2.16
68.5
1.60
145.
233.2
178.2
n+P+c
2.97
9.32
1.28
43.8
.578
71.2
129.1
93.2
n+p+C
6.05
11.7
1.68
29.7
1.60
60.2
110.9
115.8
N+P+C
11.8
18.6
23.2
17.7
8.25
105.3
184.8
175.8
Corr
Coef 1
.677
.842*
.548
.265
.925**
CO
1 _
** _
* _
T -
C,c -
Correlation coefficient computed with ll*C uptake on last day of experiment
Significant at the .01 level
Significant at the .05 level
Not numerous enough for reliable estimate
Chelator (EDTA); high and low levels, respectively
-------
TABLE 2
EXPERIMENT 2: BIOMASS ESTIMATES (yg/1)
oo
Species
Chroococcus dispersus
Anlclstrodesmus falcatus
Synedra rumpens
Synedra radians
Synedra acus
Total
MG C/(m3 x 4hr)
Control
x
7 07
7.32
3.26
4 54
17.2+
44.4
Treatir
N+C
15.2
3. 35
11.0
4.35
2.45
36.4
94.4
tents
N+P
16.9
2.23
23.6
6.74
4.95
54.4
95.4
N+P+C
13.3
2.19
22.3
11.3
4.21
53.3
290.2
Correlation
Coefficients
-.279
.629
.963*
.126
x - Obscured by high detritus content
-------
TABLE 3
EXPERIMENT 3: BIOMASS ESTIMATES (yg/1)
Species
^ Chroococcus dispersus
oo Ankistrodesmus falcatus
CT Anabaena wisconsinense
(Jylindrospermum siagnaie
Spirulina major
Aulosira sp.
Synedra rumpens
Synedra radians
Total
MG C/(m3 x 4hr)
Treatments
Initial
(a)1
6.07
.211
1.13
.130
.031
.001
.372
T
7.91
(b)
10.9
.633
2.79
.318
.005
.014
.124
T
14.8
Control
(a)
7.00
1.64
3.25
.155
.045
1.49
T
13.6
68.4
(b)
7.06
1.08
3.34
.214
.048
.010
.248
T
12.0
66.9
c
(a)
7.00
1.42
4.51
.160
.087
2.11
T
15.3
77.3
(b)
7.70
.994
7.18
.263
.092
.010
1.62
T
17.9
67.7
C
(a)
9.66
1.44
6.33
.174
.049
2.61
T
20.3
67.5
(b)
8.09
.832
4.93
.141
.038
1.99
T
16.0
79.2
Corr.
Coef .
.195
-.486
.509
-.201
.287
.610
(a) and (b) are labels for two separate plankton samples
-------
TABLE 4
EXPERIMENT 4: BIOMASS ESTIMATES (yg/1)
Species
Chroococcus dispersus
Elaktothrix gelatinosa
Microcystis incerta
Anabaena wisconsinense
Cylindrospermum stagnale
Lyngbya limnetica
Aulosira sp.
Spirulina major
Oocystis parva
Cosmarium truncatellum
Chroomonas acuta
Total
MG C /(m3x 4hr)
Treatments
Initial
(a)
6.46
1.15
2.59
10.4
.182
54.5
8.90
.125
5.99
12.7
1.47
104.5
(b)
8.02
.650
3.37
7.01
.409
37.2
11.6
.282
8.27
14.0
1.20
92.0
Control
(a)
2.98
.146
3.26
5.40
.162
111.8
6.14
.075
3.92
54.7
188.6
33.7
(b)
7.54
.190
4.63
6.69
.554
133.9
12.5
.209
6.28
58.0
230.5
38.7
N+P
(a)
3.64
.312
3.86
2.40
.361
126.7
4.35
.315
4.64
29.8
176.4
82.8
(b)
3.61
.281
3.95
4.07
.442
138.0
5.59
.151
5.92
50.4
212.4
90.8
Correlation
Coefficients
-.388
.949*
.049
-.815
.239
.570
-.615
.479
.189
CC.-3
. D J -J
-------
TABLE 5
EXPERIMENT 5: BIOMASS ESTIMATES (pg/1)
Species
Chroococcus dispersus
Aiikistrodesmus fa leaf us
Rhabdoderma sigmoidea
Lyngbya limnetica
Crucigenia tetrapedia
Chroomonas acuta
Cryptomonas bvata
Total
MG C/(m3 x 4hr)
Initial
16.5
1.30
.337
17.5
1.54
3.34
187.2
227.7
Treatments
Control
(a)
20.0
2.71
.899
27.8
2.89
230.0
284.3
17.1
(b)
14.6
4.59
.337
29.4
1.73
181.9
232.6
11.6
N+P-t-C
(a)
21.1
8.24
.337
82.8
2.50
519.0
634.0
47.7
(b)
86.1
7.42
1.01
120.0
5.78
315.6
535.9
50.4
Correlation
Coefficients
.680
.894
.217
.956*
.679
.784
oo
-------
TABLE 6
EXPERIMENT 6: BIOJ'ASS ESTIMATES (yg/1)
Species
Chroococcus miriutus
Chroococcus dispersus
Ankistrodesmus falcatus
Rhabdoderma sigmoidea
Lyngbya limnetica
Oscillatoria runescens
Synedra rurapens
Fragilaria crotonensis
Fragilaria capucina var. mes.
Fragilaria capucina (35 y)
Fragilaria capucina (70 y)
Tetraedron minimum
Chrysidalis ,sp.
Ochromonas sp .
Chroomonas acuta
Crypt omon as e T o s a
Total
MG C/(m3 x 4hr)
Treatments
Initial
(a)
57.4
22.8
77.0
2.10
15.0
.290
34.2
.097
.085
.135
.023
5.74
74.2
8.8
8.50
10.3
309.0
(b)
42.0
18.7
79.2
2.10
20.4
.278
43.5
.105
.338
.593
.093
8.05
126.0
17.6
10.3
19.6
388.9
Control
(a)
112.0
30.1
143.0
10.5
19.8
.054
95.2
.692
1.20
.131
12.4
7.00
26.4
458.5
62 .0
(b)
96.6
13.9
169.4
4.20
24.0
.090
43.5
.321
1.91
.523
11. 8
12.6
44.0
422.8
58.5
N+P+C
(a)
92.4
17.0
151.8
4.20
30.6
.391
73.0
.553
1.41
.150
10.4
254.8
308.0
944 .7
72.5
(b)
81.2
20.1
129.8
6.30
26.4
.202
73.5
.637
1.06
.126
13.4
212.8
325.6
891.1
7 7 . 8
Corr ,
Coeff .
-.766
.080
-.748
-.181
.654
.678
.269
.501
-.699
.176
.913*
.958*
-------
TABLE 7
EXPERIMENT 7: BIOMASS ESTIMATES (yg/1)
Species
Chroococcus dispersus
£ Ankistrodesmus falcatus
o Rhabdoderma sigmoidea
Lyngbya iimnetaca
Synedra rumpens
Synedra acus
Chrysidalis sp .
Chroomonas acuta
Cryptomonas ovata
Cryptomonas erosa
Total 9
MG C/(m"x 4hr)
Treatments
Initial
(a)
.753
16.8
2.70
4.27
.063
.313
15.9
37.2
T
T
78.0
(b)
.621
30.6
3.48
4.65
.063
.313
20.1
40.1
T
T
100.0
Control
(a)
.770
45.8
8.26
7.27
.069
1.57
.467
2.14
66.3
15.1
(b)
.687
25.0
4.21
6.52
.056
1.88
.467
4.41
43.2
13.7
N+P+C
.642
55.9
9.38
19.9
.075
.940
.467
6.69
94.0
23.8
N+P
.764
65.0
9.89
18.5
.213
2.51
1.63
6.55
105.1
22.6
Correlation
Coefficients
-.254
.866
.818
.998**
.564
-.119
.490
.845
-------
TABLE 8
EXPERIMENT 8: BIOMASS ESTIMATES (yg/1)
Species
Chroococcus dispersus
Ankistrodesmus falcatus
Rhab do derma sigmoidea
Lyngbya limnetica
Microcystis incorta
Aphariothece nidfulans
Synedra. rumpens
Synedra ac.u.s.
Cruciqenia tetrapedia
Sphaerocvstis Schroeteri
Chroomonas acuta
Total
MG C/(m3 x 4hr)
Treatments
Initial
9.02
.120
1.01
.257
.009
.089
.048
.198
.915
T
35.7
47.4
Control
(a)
1.52
2.60
.144
T
7.32
.051
.257
1.46
2.76
16.1
18.0
(b)
2.12
3.27
.174
T
3.24
.110
.257
1.69
1.85
12.7
15.6
N+P+C
(a)
.636
2.73
.038
.193
20.2
.286
1.21
1.32
1.38
28.0
24.9
(b)
.742
1.05
T
.916
22.2
.176
1.14
1.56
2.52
30.3
27.1
Correlation
Coefficients
-.959
-.784
-.999
.997**
.719
.958**
-.488
n on
. u ou
-------
TABLE 9
EXPERIMENT 9: BIOMASS ESTIMATES (yg/1)
Species
Chroococcus dispersus
Ankistrodesmus falcatus
Rhabdoderma sigmoidea
Microcystis incerta
Gomphosphaeria lacustris
Qocystis parva
Chro.omonas a cut a
Cryptomonas ovata
Total
MG C/(m3 x 4hr)
Treatments
Initial
(a)
5.49
T
T
.740
.168
1.60
37.2
T
45.2
(b)
5.15
T
T
.304
.151
2.10
30.1
T
37.8
Control
(a)
4.32
T
T
.203
1.55
2.58
T
T
8.65
6.52
(b)
3.67
T
T
.249
1.03
1.71
T
T
6.66
5.17
N+P+C
4.83
T
T
.539
1.93
2.73
T
T
10.0
12.6
N+P
4.39
T
T
.399
2.81
2.18
T
T
9.78
11.9
Correlation
Coefficient
.840
.917*
.826
.533
o
r-o
-------
TABLE 10
EXPERIMENT 16: BIOMASS ESTIMATES (yg/1)
Species
Microcystis incerta
ID Aphanocapsa elachista
"" Gomphosphaeria lacustris
Anabaena wisconsinense
Aphanizomenon flos-aquae
Cylindrospermum stagnale
Chi amy domon as pseudopertyi
Tetraedron cauaatum
Oocystis p_arva
Sph aero cyst is Schroeteri
Chroomonas acuta
Cryptomonas ovata
Total
MG C/(m3 x 4hr)
Treatments
Initial
.017
.380
T
.039
.131
.273
.120
2.29
14.2
10.2
27.1
.603
55.4
Control
(a)
.192
2.77
.032
.408
1.77
.304
6.06
9.49
8.54
10.0
.160
.720
40.4
12.7
(b)
.055
1.98
.008
.328
2.19
.144
6.58
10.4
12.5
11.2
T
1.08
46.5
11.9
N+P+C
(a)
.336
4.82
.256
.712
3.44
.312
20.6
5.90
4.05
12.0
T
1.08
53.5
22.6
(b)
.184
4.20
.120
.896
3.88
12.5
32.9
14.2
9.31
10.5
.968
3.62
93.3
25.0
Correlation
Coefficients
.613
.922*
.772
.988**
.974**
.695
.964**
.160
-.564
.323
-------
TABLE 11
EXPERIMENT 17: BIOMASS ESTIMATES (ug/D
Species
Aphanocapsa elachista
Anabaena wisconsinense
Ankistrodesmus Falcatus
Microcystis incerta
Tetraedron caudatum
Oocystis garya
Sphaerocystis Schroeteri
Asterococcus limneticus
Glenodinium pulvisculus
Chroomonas acuta
Total
MG C/(m3 x 4hr)
Treatments
Initial
(a)
44.3
x
1.41
T
.748
T
4.00
6.72
T
4.32
61.5
(b)
49.2
x
1.14
T
.416
T
5.44
4.48
T
6.72
67.4
Control
(a)
62.7
1.79
.377
4.77
.276
5.26
7.80
161.3
6.22
T
250.5
8.54
(b)
52.3
1.68
.474
4.82
1.41
6.29
7.40
156.8
7.44
T
238.6
9.39
N+P+C
(a)
62.1
1.53
.406
7.11
.744
4.11
3.28
67.2
3.25
T
149.7
12.2
(b)
80.7
1.56
.293
4.64
1.33
6.29
6.80
156.8
4.11
T
262.5
13.6
Correlation
Coefficients
.752
.432
-.052
-.539
-.394
-.832
x = not counted
-------
TABLE 32
EXPERIMENT 19: BIOMASS ESTIMATES (yg/1)
Species
Microcystis aeruginosa
£ Microcystis incerta
V! Oscillatoria tenuis
Oscillatoria rubescens
Anabaena Scheremetievi
Ankistrodesmus falcatus
Synedra radians
Chroomonas acuta
Cryptomonas ovata
Total
MG C/(m3 x 4hr)
Treatments
Initial
(a)
32.3
7.80
.080
.440
.680
.190
4.32
31.0
110.6
187.4
(b)
32.8
6.96
.280
.024
.680
.220
4.88
31.0
146.0
222.8
Control
(a)
30.6
10.2
.600
.420
.808
3.52
8.64
.600
172.6
227.3
18.1
(b)
29.0
7.68
.400
.120
.200
2.78
9.60
.800
192.2
242.8
17.7
N+P+C
(a)
45.5
11.3
1.04
.760
.320
5.80
22.9
1.40
220.8
309.8
29.8
(b)
49.8
8.64
1.28
.080
.280
5.91
21.4
3.60
235.8
326.8
31.0
Corr .
Coeff .
.994**
.332
.971**
.140
.845
.984**
.986**
.941*
-------
APPENDIX C
Productivity Results
Experiments 22-34
197
-------
TABLE 1
EXPERIMENT 22
5-12-70 - 5-15-70
Productivity Results - mgC/(m?x1ihr)
Treatment
N: 92 tig/1 N03-N+
92 ng/1 NH3-N
NTA3: 2.52 mg/1
NTA2; 252 ng/1
NTAj: 25.2 jig/1
EDTA3: 3.82 mg/1
EDTA2: 382 ng/1
EDTAx: 38.2
Control
Lake
_ 5-13-JO
X X/Control
45.8
53-3
1»6.6
1^.0
59.2
53.7
U5.6
41.9
49.2
1.09
1.27
1.11
.89
1.41
1.28
1.09
1.00
1.17
_ 5-15-70
X X/Control
56.2
54.8
55.4
56.0
56.6
52.4
55.4
52.4
57.2
1.07
1.05
1.06
1.07
1.08
1,00
1.06
1.00
1.09
Source
Treatments
ANOVA
5-13-70
F
12.69**
5-15-70
F
52
Significant Simple Effects (Tukey's Test)
5-13-70 5-15-70
Comparison
NTA3 -> Control
EDTA^ - Control
EDTA2 - Control
*
**
* .05 level
** .01 level
198
-------
TABLE
EXPERIMENT 23
5-19-70 - 5-21-70
Productivity Results - mgC/(nPxVnr)
5-20-70
5-2L-70
Treatment
N3 : 92 M-g/1 K 0-j -K+
92 M-g/1 NH3-K
N2, 18.5 Mg/i K03-m-
18.5 Mg/1 NH3-II
l^: U-5 M-g/1 ft 0-^-13+
4.5 M-g/1 MH3-N
NTAg: 2.5 mg/1
NTA5: 1.25 mg/1
NIA^: 500 ng/1
NTA3 : 250 M-g/1
NTA2: 125 M-g/1
NTA1: 25 ktg/1
Control
Lake
X X/Control
63. if
61.2
65.0
58.7
53.2
6l.O
57.0
6o.O
57.3
56.2
57-5
1.13
1.09
l.Ol*
I. Ok
1.09
1.01
1.07
1.02
1.00
1.02
x :
60.6
53. ^
19:°6
ho'.6
ho. -2
^5.3
h.i . 8
39-0
Jr2.0
X/Cont:
1.W+
1.27
1.17
99
.97
.96
1.01
1.00
93
1.00
ANOVA
Source
Treatments
5-20-70
F
3-57*
5-21-70
F
15 . 11**
Significant Simple Effects (Tukey's 'V
5-20-70 5-21-70
Comparison
K3 - Control
1^ - Control
-X-*
* .05 level
** .01 level
199
-------
TABLE 3
EXPERIMENT 2k
5-25-70 - 5-29-70
Treatment
N+P+NTA
N+P
N+NTA
N
P+NTA
P
NTA
Control
EDTA
Lake
Design
Enrichment
N03-N
PO^-P
NTA
N03-N+NH3-N
EDTA
C on c ent rat i on
25 Mg/1
5 Mg/1
.252 mg/1
18 Mfi/1
.382 mg/1
Productivity Results - mgC/(nPx4hr )
5-26-70
X X/Control
51.8 1.23
51.5 1.22
44.8 1.06
37.4 .89
54.0 1.28
51,2 1.21
45.2 1.07
42.2 1.00
42.3 l.oo
41.3 .98
33.3 .79
5-27-70
X X/Control
45-6 1.37
41.8 1.26
39.8 1.20
31.6 .95
42.7 1-29
41.2 1.24
38.2 1.15
33.2 1.00
35.6 1.07
32.2 .97
24.8 .75
ANOVA
Source
N
P
NTA
NP
N NTA
P NTA
NP NTA
5-26-70
F
2.09
62.94**
7.50*
.432
.155
2.15
1.92
5-27-70
23
388
96**
10.35*
-326
.872
1.87
.024
200
-------
EXPERIMENT 24 (Continued)
Significant Simple Effects
5-26-TO 5-27-70
Comparison F F
N-Control
N+P - P
N+NTA - NTA
N+P+NTA - P+NTA
P - Control 13.6** 7.65*
P+N - N 31-9** 12.2**
P+NTA - NTA 12.9**
P-ttJ+MTA - N+WTA 8.0*
NTA - Control
NTA+P - P
NTA+N - II 9-19*
NTA+N+P - N+P
* .05 level
** .01 level
201
-------
TABLE 1*
EXPERIMENT 25
6-17-70 - 6-22-70
Enrichment
PCty-P
NTA
EDTA
Glycine
Zn
Mo
B 12
Design
Concentration
2.57 Hg/1
23 Hg/1
.252 mg/1
.382 mg/1
.098 mg/1
1*2.8 ng/1
9-6 [ig/1
1.0 mg/1
Productivity Results - mgC/(m3xUhr)
Treatment
B 12
N-HP
Glycine
NTA + EDTA
EDTA
NTA
Zn
NTA + Zn
Mo
NTA + Mo
Control
Lake
6-
X
70.6
9*** 7
6U.O
72.7
78.2
7!*. 8
16.6
1*6.1*
59.6
72.6
61.9
52.3
18-70
X/Control
l.ll*
1.53
1.03
1.17
1.26
1.21
.27
.75
.96
1.17
1.00
.81*
6-19-70 f
X
1*0.6
1*3.5
1*1.2
1*7-0
1*6.8
1*0.1*
11.3
28.0
36.3
1*3.2
35- >*
52.9
X/Control
1.15
1.23
1.16
1.33
1.32
l.ll*
.32
.79
1.03
1.22
1.00
1.1*9
TT
36.1*
22.0
17-9
17.2
50.8
6-22-70
X/Control
2.12
1.28
l.Ol*
1.00
2.95
202
-------
EXPERIMENT 25 (Continued)
6-18-70
ANOVA
6-19-70
Source
NTA
Zn
NTA Zn
Source
NTA
MO
NTA Mo
Source
NTA
EDTA
NTA EDTA
Source
Treatments
F
87.86**
262.97**
13.83**
F
18.23**
.53
.0003
F
.71*1*
2.71+
^.55
F
39.69**
F
378.22**
1061*.22**
111.81***
F
15.69**
1.5l*
.1*07
F
.925
11.52*
.819
F
239-37**
Significant Simple Effects (Tukey's Test)
6-18-70 6-19-70
Comparison
N+P - Control
* .05 level
** .01 level
203
-------
TABLE 5
EXPERIMENT 29
8-21-70 - 8-26-70
Design
Enrichment
- N
- P
NTA
EDTA
Concentration
25
5
.252 mg/1
.382 mg/1
Productivity Results - mgC/(m3x14hr)
8-22-70
Treatment
EDTA
N+P+NTA
N+P
N+NTA
N
P+NTA
P
NTA
Control
Lake
X
51-3
48.2
35-3
42.7
37.0
43.4
34.6
39.7
32.5
X/Control
1.58
1.49
1.09
1.31
1.14
1.33
1.06
1.22
1.00
8-23-70
X
99-1
90.4
65.8
85.1
62.8
81.4
61.8
74.4
56.2
100-0
X/Control
1.17
1.6l
1.17
1.51
1.12
1.45
1.10
1.32
1.00
1.78
8-24-70
X
80.0
81.8
58.0
77-0
56.2
69.8
49.1
65.4
46.0
87.8
X/Control
1.74
1.78
1.26
1.67
1.22
1.52
1.07
1.42
1.00
1.91
8-2570
Treatment
EDTA
N+P+NTA
N+P
N+NTA
N
P+NTA
P
NTA
Control
Lake
X
67.2
72.4
57.0
69.0
58.2
61.2
48.9
59-2
1*0.8
71.2
X/Cont.
1.65
1.77
1.1*0
1.69
1.43
1.50
1.20
1.45
1.00
1.74
_ 8-26-70
X X/Control
92.8
61.6
69.3
1.51
1.00
1.12
204
-------
EXPERIMENT 29 (Continued)
AN OVA
Source
ft
P
ITTA
KP
K NTA
p KTA
KP NT A
8-22-70
F
5-13
2.86
36.1*9**
.095
.209
2.42
,940
8-23-70
F
17-52*-*
8.31*
138.19**
.376
1.63
.282
.008
8-24-70
₯
50 . 4l**
5.1*5*
199.06**
.025
.567
.496
.092
8-25-70
F
61. 4i**
4-39
92.67**
1.69
.580
.061
3-20
Significant Simple Effects
Comparison
N - Control
N+P - P
N+NTA - NT A
N+P+NTA - P+NTA
P - Control
P+N - H
P+KTA - NTA
P+N+NTA - N+NTA
NTA - Control
NTA+P - P
HTA+N - N
NTA+N+P - N+P
Tukey's Test
EDTA - Control
Lake - Control
-22-70
F
8-23-70
F
8-95*
6.22*
8-24-70
F
11.5**
8 8*
W \J
"1 ll . OX"^~
17-3**
8-25-70
F
34.4**
7.44*
10.9*
14.2**
6.32*
9.47*
**
25.4**
29 . 4**
38.4**
**
**
4l . 7**
ij.y 4 c,**
51.7**
62.8**
**
**
8.0*
17-2**
13.2**
O1"? piV V
d.( .O**
-X-
**
* .05 level
** .01 level
205
-------
TABLE 6
EXPERIMENT 30
10-24-70 - 10-29-70
Enrichment
N03 - N
POj^ - P
KTA
EDTA
Design
Concentration
25 WE/1
5 HB/1
.252 mg/1
.382 mg/1
Productivity Results - mgC/(m3x4hr)
Treatment
EDTA
N+P+WTA
N+P
P + NTA
p
NTA
Control
Lake
10-25-70
X
75-3
88.8
81.1
75.2
75.1
85.8
85.3
76.1
69.9
85.3
X/Control
1.08
1.27
1.16
1.08
1.07
1.23
1.22
1.09
1.00
1.22
10-27-70
X
63.3
94.1
7^-7
72.1
64.0
62.6
57.7
57-1
51.4
88.2
X/Control
1.23
1.83
1.45
1.4o
1.24
1.22
1.12
1.11
1,00
1.72
10-29-70
X
21.1
29.8
24.4
28.6
22.8
22.2
18.4
21.8
18.1
36.1
X/Control
1.16
1.65
1.35
1.58
1.26
1.23
1.02
1.20
1.00
1.99
ANOVA
Source
N
P
NTA
Iff
N NTA
P NTA
NP NTA
10-25-70
F
.236
45.66**
4.85
.672
.034
.085
4.05
10-27-70
F
79.42**
27,12**
19.87**
5-98*
3.96
1.55
2.01
10-29-70
F
31.24**
.618
17.30**
.206
.624
.005
.012
206
-------
EXPERIMENT 30 (Continued)
Significant Simple Effects
Comparison
N - Control
N+P - P
N+NTA - NTA
N+P+NTA - P+NTA
P - Control
P+K - K
P+NTA - NTA
P+N+NTA - N+NTA
NTA - Control
NTA+P - P
NTA+N - N
NTA+N+P - N+P
Lake - Control
Control - C
10-25i70
F
10-27-70
F
8.72*
21.6**
8.54*
6.28*
26.6**
20.6**
-*-*
-**-
**
10-29-70
F
7-07*
9.05*
11.4**
6.6*
5.74*
**
.05 level
.01 level
207
-------
TABLE 7
EXPERIMENT 31
5-17-71 - 5-21-71
Design
Enrichment
NTA
EDTA
Concentration
.252 mg/1
.382 mg/1
Treatment
NTA
EDTA
Control
Lake
Productivity Results - mgC/(m3x^hr)
5-19-71
X X/ Control
2h.h 1.07
25.5 1-12
22.7 1-00
21.3 -9^
5-20-71
X
29.7
29.1
26.6
29.0
X/Control
1.12
1.09
1.00
1.09
5-21-71
X
31.2
29.5
28.1
X/Control
1.06
1.00
1.00
.96
Significant Simple Effects (Tukey's Test)
5-19-71 5-20-71 5-21-71
Comparison
NTA - Control
EDTA - Control
Control - Lake
.05 level
208
-------
TABLE 8
EXPERIMENT 32
5-214-71 - 5-28-71
Des ign
Enrichment
NTA
EDTA
Concentration
.252 mg/1
.382 mg/1
Productivity Results - mgC/(m3x4hr)
5-26-70 5-27-70
5-28-70
Treatment
NTA
EDTA
Control
Lake
X
41,1
41. if
39-4
36.2
X/Control
1.04
1.05
1.00
.92
X
40.3
4l.2
37.8
33-3
X/Control
1.07
1.09
1.00
.88
X
34.8
34.0
36.3
29.9
X/Control
96
.94
1.00
.82
Significant Simple Effects (Tukey's Test)
5-26-70 5-27-70 5-28-70
Comparison
NTA - Control
EDTA - Control
Control - Lake
* .05 level
209
-------
TABLE 9
EXPERIMENT 34
6-9-71 - 6-14-71
Design
Enrichment
NO? - N
P0j| - P
NTA
EDTA
Concentration
.198 mg/1
.350 mg/1
.252 mg/1
.382 mg/1
Productivity Results - mgC/(m3x4hr)
6-10-71
Treatment
N+P
N
P
NTA
EDTA
Control
Lake
X
110
69
79
65
65
69
50
X/Control
.5
.2
.3
.8
.0
.8
.3
1.
*
1.
.
.
1.
.
58
99
14
94
93
00
72
X
151.
57-
36.
30.
28.
31.
37.
6-11-71
6-12-71
X/Control
2 4.
6 1.
6 1.
0
6
9 1.
2 1.
74
8l
15
94
90
00
17
X
230.
47-
4o.
26.
26.
30.
53.
6
0
8
3
6
0
7
6-14-71
Treatment
N+P
N
P
NTA
EDTA
Control
Lake
X
195
28
38
24
22
25
6l
X/Control
7.69
1-57
1.36
.88
.89
1.00
1.79
X/Control
.8
.0
.2
.5
9
.0
.5
7-
1.
1.
.
.
1.
2.
83
12
53
98
92
00
46
ANOVA
6-10-71
Source
N
P
NP
F
23
66
26
-97**
.15**
.06**
6-11-71
F
224.
110.
90.
6-12-71 6-14-71
F
44** 1799.02**
37**
07**
1591.23**
1258.65**
F
64.84**
82.26**
60.08**
210
-------
EXPERIMENT 31* (Continued)
Significant Simple Effects
6-10-71 6-11-71 6-12-71 6-ll|-71
Comparison F F F F
N - Control 15.0* 2^.0**
N+P - P 50.1*-* 299.0** 3030.0** 12*1.7**
P - Control 9.68**
P+N - N 87.7** 199-0** 2822.0** lUl.if**
Tukey's Test
Lake - Control * * **
* .05 level
** .01 level
211
-------
APPENDIX D
Productivity Results
Sewage Experiments
213
-------
TABLE 1
EXPERIMENT 6-S
V23-69 - 5-5-69
Sewage Addition 325 ml to 19L
Productivity Results - mgC/(m3x4hr)
Treatment
Sewage
N+P+EDTA
Control
Treatment
Sewage
N+P+EDTA
Control
4-25-69
X X/Control
196.8 2.75
120.0 1.68
71.6 l.GC
5-2-69
X X/C on ;;rol
81.6 i.6c
81.6 l.6o
51.1 1.00
4-28-69
X X/Control
L14.4 2.56
73-6 1.65
44.6 i.oo
5-5-69
X X/Control
48.9 .81
-5-2 1.25
60.2 i.oo
4-50-69
X X/Control
97.0 1.62
88.2 1.48
59-7 1-00
AKOVA
Significant Simple Effects (Tukey's Test)
4-25-69 4-28-69 4-30-69 5-2-69 5-5-69
Comparison
Sewage - Control
N+P+EDTA - Control
Sewage - N+P+EDTA
*
*
* Significant at .05 level.
214
-------
TABLE 2
EXPERIMENT 2^-S
5-25-70 - 5-29-70
Sewage Addition 200 ml to 19L
Productivity Results - mgC/(m3x^hr)
5-26-70 5-27-70
Treatment X X/Control X X/Control
Sewage
N+P
Control
ANOVA
Significant Simple Effects (Tukey's Test)
5-26-70 5-27-70
Comparison
Sewage - Control ** **
N+P - Control **
Sewage - K+P ** **
** Significant at .01 level.
77.6
51.5
1+2.2
±.8k
1.22
1.00
105 . 6
ill. 8
33.2
3*18
1.26
l.oo
215
-------
TABLE 3
EXPERIMENT 25-S
6-17-70 - 6-22-70
Sewage Addition 200 ml to 19L
Treatment
Sexrage
N+P
Control
Productivity Results - mgC/(nPx^hr)
6-18-70
X X/Control
109.2
9^.7
61.9
1.78
1-53
LOO
6-19-70
X X/Control
169.2 ^.78
*8-5 1-23
35-^ LOO
6-22-70
X X/Control
98.2 5-71
36. h 2.12
17.
1.00
ANOVA
Significant Simple Effects (Tukey's Test)
6-18-70 6-19-70 6-22-70
Comparison
Sewage - Control
N+P - Control
Sewage - N+P
**
**
**
**
**
*
**
* Significant at .05 level.
** Significant at .01 level.
216
-------
TABLE k
EXPERIMENT 29-S
8-21-70 - 8-26-70
Sewage Addition ijQO ml to 19L
Productivity Results -
8-22-70
8-23-70
Treatment
Sewage
N+P+NTA
Control
X
38.2
J+8.2
32.5
X/Control
1.18
1.^9
1.00
X
75-3
90. l+
56.2
X/Control
1.3*+
1.6l
1.00
X
88.1
81.8
1+6.0
8-25-70
8-26-70
Treatment
Sewage
N+P+NTA
Control
X
102. h
72.1+
1+0.8
X/Control
2.51
1.77
1.00
X
168.9
92.8
61.6
X/Cont:
2.7*+
1-51
1.00
8-2H-70
X/Control
1-92
1.78
1.00
Comparison
Sewage - Control
N+P+NTA - Control
Sewage - N+P+NTA
ANOVA
Significant Simple Effects (Tukey's Test)
8-22-70 8-23-70 8-2U-70 8-25-70 8-26-70
*
*
* Significant at .05 level.
** Significant at .01 level.
217
-------
TABLE 5
EXPERIMENT 3^-S
6-9-71 - 6-14-71
Sewage Addition 200 ml to 19L
Productivity Results - mgC/(m3xUhr)
_ 6-10-71 _6-U-Jl _6-12^71
Treatment X X/Control X X/Control X X/Control
Sewage 139-2 1.99 2^5-0 7-68 323.8 10-79
N+P 110.5 1.58 151.2 h.7^ 230.6 7.69
Control 69.8 1.00 31.9 1.00 30.0 1.00
_ 6-llt-71
Treatment X X/Control
Sewage Ilk.k 6.98
N+P 195.8 7-83
Control 25.0 1.00
AN OVA
Significant Simple Effects (Tukey's Test)
6-10-71 6-11-71 6-12-71 6-1*1-71
Comparison
Sewage - Control ** ** ** **
N+P - Control ** ** ** -x-x-
Sewage - N+P * ** **
* Significant at .05 level.
** Significant at .01 level.
218
-------
APPENDIX E
Phytoplankton Species Biomass Estimates
Sewage Experiments
219
-------
TABLE 1
EXPERIMENT 6-S
Species Biomass Estimates
Treatment Means - ug/1
Species
Ochr omonas
sp.
Chrysidalis
sp.
Synedra
rumpens
Fragilaria
capucina
Fragilaria
crotonensis
Ankistrodesmus
f ale at us
Tetraedron
minimum
Chroococcus
djspersus
Chroococcus
minutus
Lyngbya
limnetica
Ehabdoderma
sigmoidea
Crypt omonas
ercsa
Chroomonas
Initial Lake
13-2
100.1
38.8
.211
.101
78.1
6.90
20-3
17.7
2.10
15-0
9-9
Control
35.2
9.80
69. 4
1.56
.506
156.2
12.1
22.0
21.9
7.35
0
0
H+P+EDTA
316.8
233-8
73.2
1.24
.595
i4o.8
11.9
18.6
86.8
28.5
5-25
0
0
Sewage
32.4
10.8
200.3
11.55
8.23
314.0
33.8
246.4
123.2
27.7
2.10
0
0
acuta
220
-------
Specie;
TABLE 2
6-S
Species (Continued)
Treatment Effects (Tukey's Test)
Comparison
Sewage - Control Sewage - 1+P+EDTA K-s-P -Co
Synedra
runrpens
Fragilaria
capucina
Fragilaria
erotoneiisijs
Ankis trodesmus
falcatus
Tetraedron
Chroococcus
dispersus
Chroococcus
minutus_
Lyngbya
iicmetica
Rliabdodenaa
signioidea
Crypt OEionas
erosa
Chroomonas
acuta
* Significant at .05 level
** Significant at .01 level
221
-------
TABLE 3
EXPERIMENT 25-S
Species Bicmass Estimates
Treatment Means - y.g/1
Species
Coelosphaer ium
Kuet z ing ianum
Oscillatoria
sp.
Mallomonas
sp.
Crucigenia
tetrapedia
Synedra
sp. 1
Synedra
sp. 2
Crucigenia
rectangular is
Sphae r ocy st is
Schroeteri
Microcystis
incerta
Scenedesmus
"bijuga
Ankistrodesmus
falcatus
Coelastrum
micrqporum
K ephr ocyx ium
sp.
Chrooaonas
acuta
Cryptcmonas
oyata
Crypt cmonas
erosa
Oocystis
submarina
Ankistrodesmus
falcatxis var.
Ankis trode smus
falcatus var.
Syneehocystis
aquatllus
Aphanocapsa
elachista
Tetraedron
Initial Lake
20.70
5-32
.01
.100
.001
.108
7-50
5.64
.280
.2^5
.01
.072
.211
^3-75
58.85
23-50
56.35
.017
acicularis
.1^2
2
1.42
7.67
.025
Final Lake
36.49
3-90
.01
.081
.153
.108
5.44
4.44
.112
.228
.01
.145
.036
22.87
27.66
6.58
37-17
.024
.104
2.53
15-^5
.036
Control
30.90
8.56
1.15
.108
.034
.237
5.44
2.69
.168
.216
.008
.362
.355
.228
.456
.518
31.^5
.018
.068
2.76
7-33
.058
i\T+p
59-20
27.35
5.85
.174
1.06
1.77
8.94
8.36
.352
.542
.021
.434
.598
.019
.342
.032
27.00
.035
.090
2.74
19-05
.114
Sewage
73. Op
25-50
7.10
.218
4.43
3.88
57.0
23.3
7.26
3.01
.081
3.18
1.109
0
.233
.027
26.85
.047
.085
4.78
13-50
.147
caudatian
222
-------
Species
TABLE k
EXPERIMENT 25-S
Species (Continued)
Significant Treatment Effects (Tukey's Test)
Comparison
Sewage - Control Sewage - K+P W+P - Control
C oe1osphae r ium
Kuetzingianum
Oscillatoria
sp.
Mallomonas
sp.
Crucigenia
tetrapedia
L'ynedra
sp. 1
Synedra
sp. 2
Crucigenia
rectangularis
Sphaerocystis
Schroeteri
Microcystis
incerta
Scenedesmus
bijuga
Arikistrodesmus
falcatus
Coelastrum
microporurn
Kephrocytium
sp.
* .05 level
** .01 level
#*
**
223
-------
Species
Chroqmonas
acuta
Elaktothrix
gelatinosa
Sphaerocystis
Schroeteri
Aphanothece
nidulans
Synedra
sp.
Cryptomonas
ovata
Crucigenia
rectangularis
Oocystis
sp.
Chlamydomonas
sp.
Ankistrodesmus
falcatus
Aphani zomenon
flos-aquae
TABLE 5
EXPERIMENT 3^-S
Species Biomass Estimates
Treatment Means - pg/1
Initial Final Control P
Lake Lake
3M.5 98.6 136.2
130.0
.381
2-75
.218
43-3
136.8
5-29
3.91
.580
.551
2.1K)
148.7
372
4.29 2.78
24.9 65.9
0 .436 .709
24.7 24.2 31-5
1.90 106.2 157.4
144.8 320.8 319.6
3.58 2.88
5-36 3.96
1.01 .706
3.78
3-99
.832
N
185.6
.382
3-78
27.4
50.2
130.8
290.6
4.06
7-85
1.26
K+p
636.2
937
8.15
53-0
2.18
119.4
220.4
288.5
5.23
.895
Sewage
676.0
1.16
9.02
57-0
181.9
304.8
73.3
6.11
.738
224
-------
TABLE 6
EXPERIMENT 34-S
Species ANOVA
Species
Chroomonas
acuta
ANOVA
Source F
p
HP
655.7 **
453.6 **
324.9 **
Elaktothrix
gelat inosa
Source F
N 11-27 *
NP 8.52 *
Spnaerocystis
Schroeteri
Source
E
P
HP
F
6.87
2.39
10.06
Aphanothece
nidulans
Source F
N -67
P 27-4 **
NP 1-49
Simple Effects
Comparison F
K - Control 28.6 **
N+P - P 951 **
P - Control
P+-H - H 771 **
Sewage - Control **
Sewage - K+P
E+P - Control **
Comparison F
N - Control
H+P - P 18.9 *
P - Control
P-HS - N 19-1 *
Sewage - Control *
Sewage - N+P
N+P - Control *
Comparison F
N - Control
N+p _ p 16.8 *
p - Control
p+W - N 11.1 *
Sewage - Control *
Sewage - N+P
N+P - Control *
Comparison F
N - Control
N+P - P
P - Control 20.8 *
P+N - N 8.05 *
Sewage - Control *
Sewage - N+P
N+P - Control *
225
-------
EXPERIMENT
Species ANOVA (Continued)
Species
Synedra
sp.
ANOVA
Source F
N
P
NP
Cryptomonas
ovata
Chlamydomonas
sp.
Ankistrodesmus
falcatus
22.3^ **
39.1*14. **
20.79 *
Source F
N 3^.58 **
P 15.58 *
NP 10.17 *
Source F
N 2.25
P .075
NP .612
Source F
N 21.17 **
P 24.20 **(-)
HP 2.23
Simple Effects
Comparison F
N - Control
N+P - P 43.0 **
P - Control
P+N - N 58.8 **
Sewage - Control **
Sewage - N+P
N+P - Control *
Comparison F
N - Control
N+P - P 1*0.8 **
P - Control
P+N - ft 25.4 **
Sewage - Control **
Sewage - N+P *
N+P - Control **
Comparison F
N - Control
N+P - P
P - Control
P+N - N
Sewage - Control **
Sewage - N+P **
N+P - Control
Comparison F
N - Control 18.5 *
N+P - P
P - Control
P+N - N 20.5 *
Sewage - Control
Sewage - N+P
N+P - Control
226
-------
EXPERIMENT
Species ANOVA (Continued)
Species
Aphanizomenon
flos-aquae
Crucigenia
rectangularis
Oocystis
sp.
ANOVA
Source
N
P
NP
Source
N
P
NP
F
1.26
3.01
.021
F
18.0? *
U6.68 **
3-51
Simple Effects
Comparison F
N - Control
N+P - P
P - Control
P+N - N
Sewage - Control
Sewage - N+P
K+P - Control
Comparison F
N - Control
H+P - P 18.6 *
P - Control 12.2 *
P+N - N 37.7 **
Sewage - Control **
Sewage - N+P
N+P - Control
-*
*
Source F
N 2.3^
P .007
WP .001
Comparison F
N - Control
N+P - P
P - Control
P+N - N
Sewage - Control *
Sewage - N+P *
N+P - Control
* Significant at .05 level
** Significant at .01 level
227
ftl).S. GOVERNMENT PRINTING OFFICE:1973 514-155/303 1-3
-------
SELECTED WATER
RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
! /. Report ffo.
Z,
W
-f. Thle
An In Situ Evaluation of Nutrient Effects in Lakes
5. heport Is Organ- itioa EPA
75. Supplementary Nates
Type Kept, and
Period Covered
Final, 1968-1972
Environmental Protection Agency report
number, EPA-R3-73-018, April 1973.
16. Abstract
A method for performing in situ nutrient enrichment experiments on natural lake
phytoplankton communities was developed and evaluated. One set of experiments in which
it was employed was designed, to detect limiting nutrients and to provide a basis for
predicting future.experiment results. Productivity increased in response to all three
of the treatment variables used, W, P, and EDTA, but response patterns varied frcan
experiment to experiment. Individual species responded differently to different
treatments, and interactions among the treatment variables were important in shaping
the community responses to mixtures of two or three variables. The most consistent
features of the productivity results were incorporated into a "most probable response
pattern," which was partially validated by a second series of experiments.
The second experiment series was also used to test the ability of HTA to stimulate
phytoplankton productivity. Stimulation was continually obtained.
In a third series of experiments sewage effluents were tested in parallel with H and P.
Varying degrees of overlap between the species complexes responding to the sewage and
to the N and P treatments were found.
Recommendations for the use of ija situ enrichment experiments in eutrophication studies
are presented.
'7s, Descriptors
*Eutrophication, *Nutrients, *Phytoplankton, Primary Productivity, Nitrogen,
Phosphorus, Chelation, Lakes
l"b. Identifiers
*Enrichment Experiments, *Factorial Designs, Individual Species Responses
77c. COWRR Field & Group Q5C
IS. Availability
29. Security Ciass.
~0.' Set LT/fy C' >-s.
21.
No. of
Pages
Send To:
WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON. D C. 2O24O
I institution Virginia Institute of Marine Science
Robert A» Jordan
------- |