WATER POLLUTION CONTROL RESEARCH SERIES • 16010 DON 02/72
    Eutrophication Factors
in North Central Florida  Lakes
   U.S. ENVIRONMENTAL PROTECTION AGENCY

-------
          WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Series describes the
results and progress in the control and abatement of pollution
in our Nation's vaters.  They provide a central source of
information on the research, development and demonstration
activities in the Environmental Protection Agency, through
inhouse research and grants and contracts with Federal,  State,
and local agencies, research institutions, and industrial
organizations.

Inquiries pertaining to Water Pollution Control Research
Reports should be directed to the Chief,  Publications Branch
(Water), Research Information Division, R&M,  Environmental
Protection Agency, Washington, B.C. 20460.

-------
            EUTROPHICATION FACTORS
        IN NORTH CENTRAL FLORIDA LAKES
                      By


H. D. Putnam, P. L. Brezonik, & E. E. Shannon
    Environmental Engineering Department
            University of Florida
            Gainesville, Florida
                    for  the

      Office of Research and Monitoring
        Environmental  Protection Agency
             Project # 16010 DON
               February 1972

-------
                          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.
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20102 - Price $1 25
                              ii

-------
                                ABSTRACT

       A small Florida lake has been receiving a regimen of nutrient
addition equivalent to 500 mg/m3-yr N and 43 tag/m3-yr P since 1967.
Data has been accumulated through 1969.  The effect on the lacustrine
ecosystem of various biogenes includes production by primary producers,
species diversity of plankton and certain production estimates at the
secondary trophic level using natural populations of planktivorous fish.
Plankton production using isotopic carbon is ca. 58 grms/m2-yr;  Species
diversity is slowly changing to a mixed chlorophycean and yellow-green.
Biomass of benthic green filamentous types has increased slightly.  Nu-
trient addition has had little influence on zooplankton production.

       Related studies on 53 other regional lakes have been done using
a multi-dimensional hybrid concept as defined by several trophic state
indicators.  This trophic state index has provided a means for ranking
the lakes on an arbitrary scale.  Cluster analysis utilizing pertinent
characteristics resulted in classification of other lakes.

       Land use patterns and population characteristics were determined
photographically and N and P budgets estimated.  Using multiple regres-
sion and canonical analysis, several significant relationships were
found between lake trophic state, lake basin, land use, and population
characteristics.  In general, trophic state of lakes can be expressed
as a simple relationship incorporating K and P influx rates.

       This report was submitted in fulfillment of Grant #16010DON
under the  (partial) sponsorship of the Water Quality Office, Environmental
Protection Agency.
                                   iii

-------
                                 CONTENTS

Section

I            Conclusions and Recommendations

II           Introduction

III          Rate of Nutrient Addition to Anderson-Cue Lake

IV           Physical Characteristics of the Research Lakes
                and Drainage Basins

V            Chemical Studies

VI           Biology

VII          Trophic State Studies of Lakes in North Central
                Florida

VIII         Acknowledgements

IX           References

X            Publications
  1

  3

  5


 13

 23

 45


 91

129

133

141
                                     v

-------
                                  FIGURES

                                                                  Page

IV-1       LOCATION MAP OF ANDERSON-CUE AND McCLOUD LAKES          14

IV-2       TOPOGRAPHY OF ANDERSON-CUE LAKE                         15

IV-3       VIEW OF ANDERSON-CUE LAKE LOOKING NORTHWEST
           CHECKING RAIN GAGE AT LAKE SITE
           UNLOADING SEWAGE EFFLUENT INTO STORAGE TANK
           CHECKING HYGROTHERMOGRAPH AT LAKE SITE                  17

IV-4       WIND ROSE, ANDERSON-CUE LAKE, OCTOBER, 1967-
           SEPTEMBER, 1968                                         18

IV-5       GROUND AND SURFACE WATER LEVELS — ANDERSON-CUE
           LAKE                                                    20

V-l        DISSOLVED OXYGEN AND WATER TEMPERATURE IN ANDERSON-
           CUE LAKE                                                28

V-2        DISSOLVED OXYGEN AND WATER TEMPERATURE IN McCLOUD
           LAKE                                                    28

V-3        TOTAL ORGANIC NITROGEN IN ANDERSON-CUE AND McCLOUD
           LAKES                                                   29

V-4        AMMONIA IN ANDERSON-CUE AND McCLOUD LAKES               29

V-5        NITRATE IN ANDERSON-CUE AND McCLOUD LAKES               31

V-6        ORTHO-PHOSPHATE IN ANDERSON-CUE AND McCLOUD LAKES       31

V-7        TOTAL PHOSPHATE IN ANDERSON-CUE AND McCLOUD LAKES       32

V-8        LATERAL VARIATIONS OF AMMONIA IN ANDERSON-CUE LAKE
           JANUARY, 1968                                           36

V-9        LATERAL VARIATIONS OF ORTHO-PHOSPHATE IN ANDERSON-
           CUE LAKE, JANUARY, 1968                                 36

V-10       AMMONIA AND PHOSPHORUS CONCENTRATIONS AROUND NUTRIENT
           OUTFALL -- JANUARY 1969                                 37

V-ll       DIURNAL VARIATIONS IN ANDERSON-CUE LAKE                 39


                                    vi

-------
                              FIGURES (cont)

                                                                   Page

V-12       NITROGEN IN ANDERSON-CUE SEDIMENTS                       40

V-13       PHOSPHATE IN ANDERSON-CUE SEDIMENTS                      40

V-14       PERCENT VOLATILE SOLIDS AND TOTAL IRON IN ANDERSON-
           CUE SEDIMENTS                                            41

V-15       TEMPORAL CHANGES OF AQUEOUS AMMONIA IN LAKE WATER        43

V-16       TEMPORAL CHANGES OF AQUEOUS ORTHO-PHOSPHATE IN
           LAKE WATER                                               43

VI-1       ANNUAL VARIATION OF PRIMARY PRODUCTIVITY IN ANDERSON-
           CUE AND MeCLOUD LAKES                                    54

VI-2       DIURNAL VARIATIONS IN PRIMARY PRODUCTION DURING 1968     58

VI-3       ANNUAL VARIATION OF CHLOROPHYLL a IN ANDERSON-CUE
           AND McCLOUD LAKES                                        62

VI-4       HORIZONTAL DISTRIBUTION OF CHLOROPHYLL a IN ANDERSON-
           CUE LAKE                                                 64

VI-5       CARBON-14 FIXATION BY PHYTOPLANKTON STIMULATED WITH
           100 ygm PHOSPHORUS                                       67

VII-1      AIACHUA COUNTY LAKES                                     94

VII-2      DRAINAGE BASINS AND GEOLOGIC MAP OF AIACHUA COUNTY       96

VII-3      CORRELATION BETWEEN SODIUM AND CHLORIDE FOR 33 LAKES    108

VII-4      CORRELATION BETWEEN CALCIUM AND BICARBONATE ALKALINITY
           FOR 33 LAKES                                            108

VI1-5      CORRELATION BETWEEN TOTAL NITROGEN AND TOTAL PHOS-
           PHATE                                                   110

VII-6      CORRELATION BETWEEN COD AND COLOR                       110

VII-7      TURBIDITY VERSUS THE INVERSE OF SECCHI DISC             111

VII-8      PRIMARY PRODUCTION VERSUS TOTAL ORGANIC NITROGEN        111

VII-9      RELATIONSHIP BETWEEN PRIMARY PRODUCTION AND DEPTH       113

VII-10     RELATIONSHIP BETWEEN CHLOROPHYLL a AND PRIMARY
           PRODUCTION                                              113


                                   vii

-------
                              FIGURES (cont)

                                                                   Page

VII-11     LOCATION OF LAKES IN LOWER OKIAWAHA RIVER BASIN         115

VII-12     LOCATION MAP SHOWING MAJOR LAKE STUDY AREAS             121

VII-13     DENDOGRAM OF COLORED AND CLEAR LAKES CLUSTERED WITH
           RESPECT TO SEVEN TROPHIC INDICATORS                     125

VII-14     ANNUAL NITROGEN AND PHOSPHORUS LOADING RATES VERSUS
           MEAN DEPTH FOR THE FIFTY-FIVE LAKES                     126
                                   viii

-------
                                  TABLES
III-l      ESTIMATED NUTRIENT BUDGET FOR LAKE MENDOTA,
           WISCONSIN                                                 8

III-2      SOURCES AND SINKS FOR THE NUTRIENT BUDGET OF A LAKE       9

III-3      AMOUNT AND NUTRIENT CONTENT OF RAINFALL AT ANDERSON-
           CUE LAKE, 1968                                           10

III-4      PARTIAL NUTRIENT BUDGET FOR ANDERSON-CUE IAKE, 1968      11

IV-1       ANDERSON-CUE LAKE -- HYDROLOGICAL DATA                   21

V-l        CHEMICAL COMPOSITION OF ANDERSON-CUE AND MeCLOUD
           LAKES AND RAINFALL                                       25

V-2        TEMPERATURE AND DISSOLVED OXYGEN:   ANNUAL AVERAGES
           AT THREE DEPTHS IN STATIONS 4, 7, 11                     26

V-3        COMPARISON OF AVERAGE CONCENTRATIONS OF SOME BIO-
           GENIC PARAMETERS AT THREE DEPTHS IN STATIONS 7 AND 11    35

V-4        COMPARISON OF AVERAGE CONCENTRATIONS OF SOME BIO-
           GENIC PARAMETERS AT THE ROUTINE SAMPLING STATIONS        36

VI-1       GROUPS OF MICROSCOPIC ALGAE AND PROTOZOA IN DE-
           TAILED ANALYSES OF FIVE DATES IN 1967-68 IN McCLOUD
           AND ANDERSON-CUE LAKES, AND THE NUMBER OF OCCURRENCES    51

VI-2       COMPARISON OF AVERAGE CONCENTRATIONS OF SOME BIO-
           GENIC PARAMETERS IN EXPERIMENTAL LAKES 1968-70           53

VI-3 A     PRIMARY PRODUCTIVITY VALUES FOR ANDERSON-CUE IAKE        55

VI-3 B     PRIMARY PRODUCTIVITY VALUES FOR McCLOUD LAKE             56

VI-4       CHLOROPHYLL a LEVELS IN ANDERSON-CUE AND McCLOUD
           LAKES       ""                                            61

VI-5       ALGAL GROWTH RESPONSE TO NUTRIENTS                       65

VI-6       AMMONIA UPTAKE BY ALGAE IN ANDERSON-CUE WATER            66
                                    ix

-------
                               TABLES (cont)
VI-7       STANDING CROP ESTIMATES OF ANDERSON-CUE AND
           McCLOUD LAKES                                            69

VI-8       EXTRACTIVE PHOSPHATE AND ALKALINE PHOSPHATASE            73

VI-9       FOOD OF LABIDESTHES SICCULUS                             74

VI-10      DETERMINATION OF ZOOPLANKTON WEIGHTS IN ANDERSON-
           CUE AND McCLOUD LAKES                                    75

VI-11      ZOOPLANKTON DENSITIES AND CALCULATED BIOMASS FOR
           AUGUST 1967                                              76
VI-12
VI-13
VI-14
VI-15
VII-1
VII-2
VII-3
VII-4
VII-5
VII-6
VII-7
VII-8
ZOOPLANKTON DENSITIES AND CALCULATED BIOMASS FOR
AUGUST 1968
A PARTIAL FAUNAL LIST FROM LAKES ANDERSON-CUE
AND McCLOUD
SOME REPRESENTATIVE RATES OF SECONDARY PRODUCTION
IN AQUATIC ECOSYSTEMS
MONTHLY PREDATION PRESSURE BY E. FUSIFORME AND H.
FORMOSA IN LAKES ANDERSON-CUE AND McCLOUD 1969-70
LAKES IN AIACHUA COUNTY
PARAMETERS MEASURED IN LAKE STUDY
TROPHIC CHARACTERISTICS OF LARGE LAKES IN AIACHUA
COUNTY
TROPHIC CHARACTERISTICS OF MEDIUM-SIZE AIACHUA COUNTY
LAKES
TROPHIC CHARACTERISTICS OF SMALL LAKES IN AIACHUA
COUNTY
TROPHIC CHARACTERISTICS OF PONDS AND LAKELETS IN
AIACHUA COUNTY
PHYSICAL CHARACTERISTICS OF LAKES IN OKIAWAHA BASIN
CHEMICAL AND BIOLOGICAL CONDITIONS IN OKIAWAHA LAKES
77
79
84
87
93
98
99
101
103
105
114

           OCTOBER, 1968                                           117

-------
                               TABLES (cont)

                                                                   Page
VII-9      COMPARISON OF MAJOR IONS IN LAKE APOPKA IN 1924
           AND 1968                                                118

VII-10     MEANS, STANDARD DEVIATIONS AND CORRELATION MATRIX
           FOR SEVEN TROPHIC STATE INDICATORS                      122

VII-11     FLORIDA LAKES RANKED WITH RESPECT TO THE TROPHIC
           STATE INDEX (TSI)                                       123
                                      xi

-------
                                SECTION I

                     CONCLUSIONS AND RECOMMENDATIONS

Conclusions

1.   Anderson-Cue Lake has been enriched with nutrients and sewage
     since March, 1967.  Lake response during this period has been
     negligible for most of the routine chemical and biological para-
     meters.  As the data show, using the mean for all years, a sig-
     nificant difference does occur for chlorophyll a, but among no
     other parameters.  We are beginning to visually observe large
     increases in the periphyton especially attached filamentous
     chlorophyceae.  These are species of Mougeotia, Mougeotopsis
     and Spirogyra predominantly.  Extensive submersed growths of
     these organisms have been blanketing the littoral zone for the
     past 18 months with a periodicity from about April through early
     July.  The crop declines in late summer.  Although as yet these
     are qualitative observations we are fairly confident they re-
     present a significant algal response to enrichment.  It appears
     that at the present level of eutrophy in Anderson-Cue Lake pro-
     duction estimates of only phytoplankton may be misleading and
     that more reliability regarding the effects of lake enrichment
     may be gained by examining the productivity of other segments
     of the plant community as well.  This means that Florida lakes
     of this type undergo a subtle alteration in the early stages
     of eutrophication which do not respond to the usual methods of
     detection.  Examinations of other primary producer compartments
     would be fruitful in this respect.

2.   Trophic state indices can be established in a quantitative manner
     for lakes in this geographic area.  The value of being able to
     determine a lake's trophic state is far reaching in establishing
     future quality standards for surface waters.  Whether or not the
     indices are completely applicable to lakes in other geographic
     areas is not known at the present time.  However, preliminary
     trials indicate good correlation.

3.   Results from studies of the second trophic level have shown the
     mean value  to be 50 k cal m"2 year"^-.  This amount is within
     the expected range for lakes exhibiting oligotrophic character-
     istics.  Thus far effects of enrichment on the zooplankton have
     been minimal.
                                 -1-

-------
     Attempts  to use  littoral  zooplankton as  indicators of trophic
     state must be carefully carried out to reduce masking effects
     produced  by other variables.  In this study fluctuations in lake
     level obscured changes in the species composition and diversity
     of  zooplankton induced by eutrophication.  Primary production
     and  temperature  are among the most important environmental fac-
     tors related to  reproduction of littoral zooplankton.  Correla-
     tions between population  density and primary production most
     likely could be  improved  if the organic matter contribution of
     periphyton was considered.
Recommendations

       This three year study was composed of background data and
nutrient enrichment information.  In particular, the TSI (Trophic
State Index) has provided a mathematical formulation of lake eutro-
phication.  The experimental lake has shown practically no change
during the nutrient enrichment program.  Basically, a review of the
three year data produced one question, "What ecological compartment
has been significantly affected by the N, F and sewage additions?"
Only two alternatives appear possible:  (1) the additives were readily
absorbed by the sediment and/or (2) the additives were utilized by
the plants acting as a biological pump to remove the nutrients from
the sediment.  Neither of these possibilities has been investigated.

       Generally, primary production rates are based exclusively on
the phytoplankton.  In determining trophic states such a procedure
appears to be misleading since there are other photosynthetic com-
partments which need to be considered.  For example, periphyton and
macrophytes may proliferate in the littoral zone when the influx of
nutrients increases.  Increased growths of these plants have been
observed in Anderson-Cue Lake and likely are a result of the enrich-
ment regimen.

       Sediment analyses are not normally a parameter of water quality
surveys.  However, the history of lakes have been revealed by core
sampling as an entity.  While the past conditions of Anderson-Cue are
not of particular interest at this time, the role of nutrient exchange
and benthic algal growth in the sediment-water interface has emerged
as a major point of concern in attempting to forecast early warning
signals of lake deterioration.

       Thus, the above discussion brings forth the recommendation of
continued study of these lakes by looking at new areas where the first
warning signs of degradation may be observed.

       Specifically, this study showed Anderson-Cue to be phosphorous
limited.  Recent investigations at the EPA  facility in Athens, Georgia,
indicate carbon as another enrichment parameter.  Carbon content anal-
ysis in the Anderson-Cue community would provide much needed informa-
tion.  Additionally, the natural nutrient input to these lakes must be
defined.  So far only rain water has been examined.


                                 -2-

-------
                               SECTION II

                              INTRODUCTION

       Florida has a vast and valuable resource of fresh water considering
the springs and nearly 8,000 (greater than one acre) lakes found with-
in the state.  Practically all of these surface waters are useful in a
recreational sense and for this reason Florida appeals greatly to
tourists everywhere within this country and Canada.  Fishing, boating
and various contact water sports are enjoyed by both residents and out-
of-state visitors throughout the year.  Therefore, the conservation
of this fresh-water resource is most important to the state's economy.

       However, since water is so intimately involved in the total well
being of the environment, impairment of aquatic systems in varying de-
grees will affect all the biota including man within a particular eco-
system.  Essentially, the water quality of lakes and other fresh-water
resources mirrors the status of the total environment.

       Over the years, Florida lakes have been enriched gradually with
nutrient salts from the land.  Encroaching urbanization and intensive
agricultural practices have, however, increased nutrient addition to
lakes on an unprecedented scale in recent years.  This enrichment has
accelerated the eutrophication of surface water thereby shortening the
lives of lakes and generally impairing the quality of the water.

       This problem, which can be reflected nationally, is acute in
Florida.  The shallow lake basins, long hours of sunlight and mild
winter temperatures are some of the factors which make surface water
particularly susceptible to the effects of enrichment and lead to sus-
tained algal blooms throughout the year.  The most classic example in
Florida is Lake Apopka near Orlando.  This is a 30,000-acre lake
(12,145 hectares) which has been extensively enriched by fertilizers
from bordering citrus and winter vegetable farms, municipalities, and
citrus processing plants.  A hyacinth eradication program employing
herbicides over the last 20 years has left a flocculant bottom layer
of undecomposed plant residues.  These unconsolidated sediments accord-
ing to a recent Federal Water Quality Administration report (1968)
cover 90 percent of the lake bottom.  Fish reproduction in Lake Apopka
is prevented by the lack of suitable areas for spawning and by a per-
sistent anaerobic environment.

       A similar process is occurring in many other lakes within the
state.  Although the visible effects of eutrophication are well docu-
mented, very little real knowledge exists regarding the interplay of
environmental parameters during lake enrichment.  Ultimately management
systems for whole drainage basins must be devised if the eutrophication
                                -3-

-------
problem  is  to be dealt with effectively.  First, however, it  is neces-
sary  to  understand eutrophication  in quantitative terms and in part to
find  the most effective combinations of enriching substances  and de-
termine  how these relate for example,  to the physical environment of
lake morphology, climate and various edaphic factors.  The present level
of knowledge relative to cycling of biogenic materials is primitive and
more experimental work is badly needed in this area.  Recent  work by
Kerr e_t al  (1970) may shed some insight on this problem.  Ultimately to
offer the maximum use of a lake to those living within the basin we
must know what enrichment stress can be placed on surface water without
measurably  impairing its quality.

       This can be brought about only by long-term research.  Pnojects
such as described herein, using whole  lakes as experimental units, are
few in this country.  More are needed especially in varying geographic
locations if we are to understand completely the eutrophication process.

       The  site for most of this study was in sandy, scrub-oak terrain
near Melrose, Florida, about 25 miles east of Gainesville.  There are
numerous lakes in this area.  Two of these (Anderson-Cue Lake and
McCloud Lake) located on private property were selected through the
cooperation of the owners.  The isolated location of these lakes assured
freedom from outside interference and urban or agricultural influence.
Considerable effort was exerted in 1966 to establish a field  station at
the lake site and to install appropriate instrumentation.  Background
data on the chemistry and biology of the lakes were obtained  in order
to be certain of their similarity and trophic state.  Nutrient enrich-
ment of Anderson-Cue Lake has been continuous since March, 1967, and
during this time both routine monitoring and special studies have been
carried out in the two lakes comprising the experimental system.

       In 1968 eutrophication research was extended to other  lakes of
various types and exhibiting varying trophic stages in the north central
Florida area.  For comparative purposes eutrophic lakes in the Oklawaha
chain were also included.

       Useful information is gained from programs of this kind since
it has helped point the way to the development of a realistic index of
trophic state for Florida lakes.  In addition valuable base line infor-
mation is obtained concerning water quality in north central Florida.
From studies of this kind will emerge patterns for lake management.

       Since lakes are slow to change, a considerable lag may ensue
before lakes respond to restorative measures.  Therefore, it is essential
to regulate the pollution stress on lakes so as to maintain a desirable
level of water quality at a stage where preventive measures will suffice.
Hopefully ecosystem management models of whole lake basins will provide
the means to accomplish these objectives.

-------
                               SECTION III

             RATE OF NUTRIENT ADDITION TO ANDERSON-CUE LAKE

Previous Fertilization Studies

       One approach to studying eutrophication is to artificially
enrich (eutrophy) a lake at a controlled rate and measure all the para-
meters which define trophic state.  The problem then becomes a matter
of relating the response of a lake  (in terms of trophic structure) to
the degree and rate of nutrient enrichment.  The lake thus serves as a
general model for the process; this approach has been used in the pre-
sent study.  Intentional fertilization of lakes for scientific purposes
is not a new concept.  Stewart and Rohlich (1967) recently reviewed
previous experiments on lake fertilization.  Einsele (1941) reported
one of the first experiments on lake fertilization.  He applied slug
doses of superphosphate to a small German lake in 1937 and 1938.  Tem-
porary increases in the phytoplankton of the lake were found but the
lake soon returned to normal.  A number of investigators (e.g. Ball,
1948a, b; Langford, 1950; Nelson and Edmondson, 1955; Hooper and Ball,
1964) have attempted to increase the productivity of fish ponds by
adding fertilizer.  These attempts have had only moderate success.  In
most fertilization experiments, nutrients have been added in high slug
doses rather than continuously.  While temporary effects have been
noted, the ponds or lakes usually returned to their original conditions
in short periods of time.

       Induced circulation by aeration or actual pumping of hypolimnion
water into the epilimnion has been proposed by Easier (1957) and
Hooper e_t al (1952) to render nutrients within a lake more available
to phytoplankton.  However, unless the sediments were actually dis-
turbed by aeration, the long-term effect of this process would seem
to be the reverse of that intended.  Aeration would tend to keep the
hypolimnion oxygenated and maintain an oxidized microzone at the sedi-
ment water interface which decreases the release of nutrients from the
sediments.
Nutrient Loading Rates Into the Experimental Lake

       Because of the nature and purpose of previous fertilization
efforts, results from these studies are not directly applicable to
the problem of cultural eutrophication.  Man-induced eutrophication
is characterized by a more or less continuous addition of nutrients
to a lake from such sources as sewage effluent and urban and agri-
cultural runoff, while most previous fertilization attempts have used
sporadic or one-time applications of fertilizer.  Generally, sufficient
background data were obtained to describe the trophic state of a lake
                                 -5-

-------
and  its natural temporal variations.  The study reported here is viewed
as a  long-term effort  to follow the effects of a controlled nutrient
input on a lake's trophic state.  Two small lakes are involved; one lake
is serving as a control and the other lake is being artificially en-
riched by continuous and controlled addition of nutrients.  A variety of
routine chemical and biological data on the lakes is being collected
along with routine physical and climatic data in order to identify fac-
tors affecting the rate and severity of eutrophication.

       Chemical and biological measurements during 1966 and early 1967
established the oligotrophic nature of Anderson-Cue Lake.  In March,
1967, nutrient additions to the lake were begun.  It was decided to add
sufficient nitrogen to raise the total N content of the water 0.50 mg N/l
over a period of one year (assuming all nitrogen would stay in solution).
This is equivalent to 500 mg/m3-yr or about 10 mg/m -week.  The volume
of Anderson-Cue Lake was estimated to be 248,000 m^.  Thus a weekly load-
ing of 2.48 kg N was desired.  This was achieved by adding 21.2 Ibs of
ammonium chloride to 300 gallons of sewage effluent, which was trucked
out to the lake site and fed from a holding tank into the lake with a
chemical feed pump at a rate of 1.8 gal/hr.  The nutrient outfall is
located 2 ft below the surface and 200 ft off the scuth shore of the
lake in about 10 ft of water.

       It was decided to increase the total phosphorus content of the
lake by 0.0427 mg P/l in one year.  This is equivalent to 42.7 mg/m^-yr
or 0.854 mg/nr-week.  For the lake a loading rate of 0.212 kg P/week is
indicated.  This was achieved by adding 2.47 Ib of NasPO^ to the sewage
effluent each week.

       Originally it was planned to add only sewage effluent to the
experimental lake.  This would have been feasible if Berry Pond
(one acre surface, maximum depth, 13 ft) could have been used as ori-
ginally planned, but its trophic characteristics obviated this plan.
Nearly a million gallons of sewage effluent would have to be trans-
ported to Anderson-Cue Lake annually for the desired nutrient loading
rate.  Thus logistics precluded the use of sewage effluent alone, and
it became necessary to enrich the effluent with nitrogen and phosphorous
compounds.

       After two years, the method of nutrient addition was changed to
direct application of the chemicals to the lake by towing a burlap bag
behind a motor boat.  The towing pattern was simply to make a large
circle in the middle of the lake.  This method was considered an im-
provement since a residue was often found in the holding tank.
Nutrient Budgets for the Experimental and Other Lakes

       The above nutrient addition rates compare closely with those
estimated for the nutrient budget of Lake Mendota, Wisconsin, by Lee
e£ a_l (1966) .  This eutrophic lake receives a heavy influx of nutrients
from agricultural drainage, but ground water and atmospheric precipi-
tation also make important contributions.  The nitrogen loading of Lake
                                 -6-

-------
                                                            3
Mendota was estimated to be 556,000 Ib per year, or 534 mg/m -yr.  Of
this quantity, Brezonik and Lee (1968) have estimated that two-thirds
or 360,000 Ib remain in the lake and is deposited in the sediments,
while the remainder is lost through the outlet and by denitrification.
The phosphorus budget for Lake Mendota was estimated to be 44,900 Ib
per year or 42.7 mg P/m-^-yr.  Table III-l summarizes the computed nu-
trient budget for Lake Mendota.

       Relatively few other lake nutrient budgets have been established.
Mortimer (1939) constructed a nitrogen balance for Lake Winder-mere
(England).  He found a close balance between input and output -- 326
and 318 metric tons, respectively.  Hutchinson (1957) felt that nu-
trient inflow and outflow normally would balance closely in oligo-
trophic lakes, but not in eutrophic lakes.  Rohlich and Lea (1949)
reported an extensive nutrient balance on Lake Mendota, Wisconsin.
Of the estimated 156 metric tons of nitrogen entering the lake annually,
only 41 tons left through the surface outlet.  Corresponding values for
phosphate were phosphate were 16.4 and 11.6 metric tons.  Partial nu-
trient budgets were determined for the lower Madison lakes by Sawyer
e_t a_l (1945).  However, only soluble phosphorus and inorganic nitrogen
inputs were measured rather than total values, and the usefulness of
the results is thus impaired.  Aside from nutrient balances on Lake
Tahoe (McGauhey e£ al, 1963) and Lake Washington (Edmondson, 1968) no
other definitive nutrient budgets are known.  Less detailed budgets
have been drawn for a few other lakes--for example, Lake Pure in Denmark
(Berg, 1958), Castle Lake in California (Goldman, 1961) and western
Lake Erie (Curl, 1959).  Various aspects of nitrogen and phosphorus
budgets and sources have been treated by numerous workers.  Brezonik
(1968), Feth (1966), and Fruh (1967) have reviewed these studies in
considerable detail.

       Table III-2 lists the most common nutrient sources and sinks for
lakes.  Only some of these are applicable to the study lakes.  Artifi-
cial enrichment represents the most significant nutrient source for
Anderson-Cue Lake.  The possible natural sources of nitrogen are bio-
logical fixation, atmospheric precipitation, air-borne particulates and
surface and subsurface runoff.  Preliminary results indicate the rain-
fall directly on the lake surface is the most important natural source.
Nitrogen fixation has not yet been measured in the lake, but the near
absence of blue-green algae in the biota of the lake implies that it
does not occur.  While bacterial fixation is possible, available carbon
substrates are low and indicate the source is probably negligible.
Contributions from runoff appear to be small.  The amount of runoff
draining into the lake is apparently low, and the soil is so nutrient
depleted that rainfall runoff would pick up little or no additional
nutrients in passing through and over the soil.

       Measurements of the nutrient content of the rainfall were made
periodically in 1967 and 1968.  A summary of the results are shown in
Table III-3.  The nitrogen content of rainfall appears to be quite
variable.  However, these results can be combined with the rainfall
amounts  (see Section IV, Table IV-1) to yield an estimate of the total
nitrogen contribution of rainfall to Anderson-Cue Lake.  For 1968, 44 kg
                                 -7-

-------
                                                             Table Ill-l

                                  ESTIMATED NUTRIENT BUDGET FOR LAKE MENDOTA, WISCONSIN'- *
Source
Municipal and
industrial waste-water

Urban runoff

Rural runoff

Precipitation on
lake surface
Ground water

Nitrogen fixation
Marsh drainage
Appro*. Total
Nitrogen
Contribution
kg-

21,200
(total)
13,700
(soluble)
23,500
(soluble)

43,900
113,000

36,100

252,000
Contribution
%

8

5

9


17
45

14
—
100
Phosphorus
Contribution
kg-

7,750
(total)
3,680
(soluble)
9,100
(soluble)
64 to
3,460
274




21,240"
Contribution
%

36

17

42


2
2



—
100
Sink*
Outlet loss
Denitrification

Fish catch

Weed removal

Ground water
recharge
Sediments and
other losses5



Nitrogen
Lost
kg.
41,300
28,000

11,300

3,250




168,000


252,000
Lost
%
16.4
11.1

4.5

1.3


—

66.7


100
1 After Lee et al (1966) and Brezonik and Lee (1988)
* The values presented are only rough approximations
* This total based on 455 kg. per year ot phosphorus in precipitation on the lake surface
* Phosphorus sinks have not been determined but outlet loss and sediment deposition probably account for most of the phosphorus
' By difference between total sinks (assumed to equal total sources) and sum of aU other calculated sinks.  Other sinks are probably small, and
 sediment deposition accounts for most of the nitrogen in this category.

-------
                               TABLE III-2

           SOURCES AND SINKS FOR THE NUTRIENT BUDGET OF A LAKE
SOURCES
1.   Air-borne
        Rain water
        Aerosols and dust
        Leaves and miscellaneous debris

2.   Surface
        Agricultural runoff and drainage
        Urban storm water runoff
        Marsh drainage
        Runoff and drainage from uncultivated land
        Domestic waste effluents
        Industrial waste effluents
        Wastes from boating activities

3.   Underground
        Natural groundwater
        Subsurface agricultural and urban drainage
        Subsurface drainage from septic tanks near lake shore

4.   In situ
        Nitrogen fixation
        Sediment leaching
SINKS

Effluent loss
Groundwater recharge
Fish caught or removed
Weed harvesting
Insect emergence
Evaporation (aerosol formation from surface foam)
Denitrification
Sediment deposition of detrital particles
Inorganic precipitation (for calcium phosphate, and some trace metals)
     and deposition into sediments
                                 -9-

-------
                               Table 111-3

            AMOUNT AND NUTRIENT CONTENT OF RAINFALL AT
                       ANDERSON-CUE LAKE, 1968
Date
2-19
2-26
3-4
3-11
4-15
5-6
5-13
5-27
6-24
7-5
7-25
R-O
(J j£l
8-19
9-3
9-16
10-11
10-19
11-11
19.9
Amount1
Inches
1.00
0.25
0.35
1.20
0.65
0.45
0.75
3.35
6.65
4.00
4.65
0 c=;
V/»Od
2.10
10.85
4.30
2.85
3.70
3.60
2.9.5
TON
mg/1
__
—
—
0.57
0.67
0.64
0.33
0.11
0.24
0.39
0.01
n ^0
U.QU
0.34
0.07
0.07
0.12
0.11
0.23
0.54
NH:,-N
mg/1
0,46
0.23
0.80
0.86
0.33
0.10
0.0
0.02
0.02
0.05
0.01

0.14
—
0.11
0.05
0.21
0.07
0.18
NO3-N
mg/1
0.40
0.94
0.26
0.24
0.27
0.30
0.15
0.05
0.14
0.09
0.08
ft OQ
\j.£\j
0.22
0.09
0.16
0.11
0.04
0.05
0.11
o-PO4
/*g/l
230
20
25
18
2.2
28
20
. .
9
10
8

33
__ _
5
12
4
25
18
t-PO<
Mg/1
__
—
—
—
—
—
—
__
30
30
—
9(1
£t\J
70
20
—
—
_
30
—
1 Total amount of rain in period from date on previous line to the date listed on
 the line.  First line indicates rainfall from Jan. 1 to Feb. 19; last line indicates
 amount of rain from Nov. 12 to the end of the year.
                              -10-

-------
                               TABLE III-4

           PARTIAL NUTRIENT BUDGET FOR ANDERSON-CUE IAKE,  1968
Sewage and nutrient mixture
Rainfall on lake surface

                 Total
NITROGEN

      kg
      124
       44

      168
mg/1 of lake water
      0.67
      0.24

      (T91
       Other possible nitrogen sources not completely evaluated are
nitrogen fixation, groundwater seepage, subsurface runoff and air-
borne particulates (leaves, etc.)-  Present information indicates all
these except perhaps the last were insignificant in 1968.
Sewage and nutrient mixture
Rainfall on lake surface

                 Total
                               PHOSPHORUS
     10.60
      2.67

     13.27
mg/1 of lake water
       0.057
       0.014

       0.071
       Other possible phosphorus sources are groundwater, subsurface
runoff and air-borne particulates.  Present information indicates the
first two were insignificant, but there is insufficient information to
evaluate the last source.


 This is the concentration which would result if the amount of nutrient
 in column 1 were diluted to the volume of the lake (approximately 150
 acre-feet or 185,000m3).
nitrogen was added to the lake by rainfall directly on the lake surface.
(This value corresponds with 49.4 kg nitrogen added to the lake by
rainfall during 1967, as previously computed by Brezonik and Putnam, 1968).
By comparison, about 124 kg N was added by the nutrient mixture.  Ac-
tually a greater disparity between these two sources exists than is
indicated by the magnitude of the two numbers.  The rainfall contribu-
tion is diluted in a large volume of water, whereas the nutrient mix-
ture is highly concentrated and contributes an insignificant amount of
water to the lake.  Phosphate analyses from 1968 indicate a wide range
of phosphate in rain water.  The data indicate that about 2.7 kg P was
contributed by rainfall in 1968.  This compares with 10.6 kg added to the
lake in the nutrient mixture.  These results are summarized as a nu-
trient budget for Anderson-Cue lake in Table III-4.
                                -11-

-------
Sampling Program

       The duration and frequency of the sampling program is detailed
in Section V (Chemical Studies) and Section VI  (Biology).  The tests
conducted and the methods utilized are displayed in Appendix A and B.

       Generally, background data collection began in January, 1966 on
a bi-monthly basis.  Nutrient addition was initiated in March, 1967.
Bi-monthly sampling was conducted through January, 1970.

       Several special- investigations were conducted in conjunction
with these regular investigations.  The special studies included the
weekly collection of samples for zooplankton analyses from January, 1967
until December, 1968.  This study correlated population dynamics with
productivity.  A subsequent bi-weekly research program dealt with the
interactions of zooplankton and their fish predators.  This research
continued from June, 1968 through April, 1970.

       An intensive program was undertaken to develop a quantitative
trophic state index using multivariate statistical techniques.  The
sampling schedule used in this study was designed to provide informa-
tion on the average chemical, biological and physical characteristics
of fifty-five central Florida lakes over a one year period.

       Systematic sampling of the fifty-five lakes commenced in June,
1969.  It was decided to sample all fifty-five lakes at four-month
intervals for a period of one year.  Hence, the subsequent fifty-five
lake sampling periods were in October, 1969, February, 1970 and June,
1970.  From the fifty-five lake experimental group a sub-group of nine-
teen lakes was selected.  These lakes were sampled at two-month inter-
vals so that in addition to being sampled with the main group they were
sampled in August, 1969, December, 1969 and April, 1970.  The nineteen
lakes were selected on the basis of being representative of several
trophic types of lakes and/or being of special interest.  It was felt
that this sub-group of lakes could be used to reflect seasonal trends
in lake characteristics without sampling the entire fifty-five lakes on
a closer time interval.
                                -12-

-------
                               SECTION IV

                        PHYSICAL CHARACTERISTICS
                OF THE RESEARCH LAKES AND DRAINAGE BASINS
General
       Anderson-Cue and McCloud Lakes (see Figure IV-1) are located in
a region of high sand hills with many circular to elliptical basins
which have resulted from solution of the underlying limestone.  Both
lakes have small drainage basins with no influent or effluent streams.

       The tops of many of the surrounding hills reach elevations of 190
to 220 ft, MSL.  Westward at Melrose the terrain changes from sand hills
to the Okefeenokee Terrace, a poorly drained terrace 140 to 160 ft, MSL.
Eastward, beyond Baywood, the sand hills are bound by lower marine ter-
races.  The immediate area of the research lakes is the Trail Ridge
portion of the Central Highlands.

       Three hydrographic surveys of Anderson-Cue Lake have been made
since the fall of 1965.  Echo soundings were made in November, 1965;
a stadia survey (including topography of the basin) was made in July-
August, 1967; and echo soundings were again made in March, 1968.  Using
these data a topographic map of the lake and surrounding area was pre-
pared and is shown in Figure IV-2.  The highest level shown on the map
was the shoreline which stood at 125.78 ft, MSL, in March, 1966, at
which time the lake had a surface area of 19.3 acres and a volume of
approximately 201 acre-feet.  When it was decided to use McCloud Lake
instead of Berry Pond as the control body of water in late 1966 the
volume of water in McCloud Lake exceeded that in Anderson-Cue Lake by
approximately 15 percent.  This information was obtained by stadia
survey.

       Yearly excess precipitation over evaporation (30 year record)
is 12 to 18 in. in the xeric hills surrounding Anderson-Cue lake.
(These data are not applicable to the research lake itself.)  The ex-
cess precipitation and runoff percolate downward through breaks in the
sands and clays of an aquifuge that overlies the Floridan aquifer.  The
influent drainage has resulted in a subsidence karst landscape which
forms a principal recharge area in North Florida for the Floridan
artesian system.  Analyses of the water level data for Anderson-Cue
Lake, the surrounding water table data, and the rainfall and evaporation
records indicate that there is very little contribution from surface
and subsurface runoff to the lake.  Sands covering the basin are porous.
Downward seepage rates are high and surface runoff is exceedingly low.
                                 -13-

-------
•»w
                                                                       ?<&
                Figure IV-1 . Lex:otion Map of And«f»on-Cue and McCloud Lake*.

-------
Figure IV-2. Topography of Anderson-Cue Lake.
                      -15-

-------
Geology

       Materials exposed in the sand hills are largely of two types:
very fine surface sands and the underlying kaolinic gravels, sands and
sandy clays.  These sediments are known as the Citronelle Formation.
Well borings show an aquifuge of from 80 to 100 ft of phosphatic sands,
sandy clays and clays lying below the surface.  These materials are
known as the Hawthorne Formation of Lower and Middle Miocene Age.  Under-
lying the Hawthorne Formation is the Floridan aquifer, the upper portion
of which is the Ocala Limestone of Eocene Age.

       The piezometric surface of the water in the Floridan aquifer is
approximately 90 ft above MSL in the vicinity of Anderson-Cue Lake.
The porous sand and gravel of the Citronelle Formation contain a perched
water table above the aquifuge—the Hawthorne Formation.  Anderson-Cue
Lake is itself a perched lake.  The lake level is the result of a bal-
ance between precipitation, evaporation and outflow into the water
table aquifer and Floridan aquifer.

       The vegetation in both lake basins is sparse and primarily scrub
oak, indicative of poor nutrient conditions.  There is no human habi-
tation in either basin.  The major source of nutrients for the lakes in
their natural states appears to be from the atmosphere via precipitation
and air-borne particulates.
Instrumentation

       A Gurley water level recorder with staff gage and a recording
rain gage were installed at Anderson-Cue Lake in February, 1966.

       In September, 1967, an Aerovane wind recorder and Foxboro
hygrothermograph were installed.  The transmitter for the wind recorder
was mounted on a pole approximately 150 ft from the south shore of the
lake and three feet above the water surface.  Examples of some of the
instrumentation are shown in Figure IV-3.
Meteorological and Hydrological phenomena

       Anderson-Cue Lake lies in a shallow valley oriented in a NNE to
SSW direction and is surrounded by scrub oak and pine trees.  These
characteristics have a marked effect on the air-flow over the water
surface.  The air speed in general is calm to light (0-7 mph).  The
prevailing winds are from 30 to 60 degrees (NNE to NE) and from 210
to 240 degrees (SSW to SW).  When a tropical storm or frontal system
passes over or close to NE Florida the wind direction is influenced by
such phenomena and higher wind speeds are recorded.  A wind rose for the
period October, 1967 to September, 1968, is shown in Figure IV-4.
                                 -16-

-------
View of Anderson-Cue Lake
   Looking Northwest
Checking Rain Gage at Lake Site
 Unloading Sewage Effluent
    into Storage Tank
   Checking Hygrothermograph
         at Lake Site
                               Figure !V-3.
                                 -17-

-------


                     SCALE: 1mm # 3 HOURS








       1-3MPH



Figure \V-4. Wind Rose, Anderson-Cue Lake, October 1967 - September 1968.
                          -18-

-------
       The only significant currents in shallow Anderson-Cue Lake are
wind currents.  In bodies of water larger than Anderson-Cue Lake such
surface water currents flow to the right of the wind and set up a clock-
wise circulation.  In Anderson-Cue Lake, however, these currents cause
a pileup of water on the leeward shore which is returned by fanouts in
both clockwise and counterclockwise directions.

       Many factors must be considered in attempting to explain the
fluctuations of the lake level and the water table in the research
area.  These include evaporation, precipitation, and flow to the water
table aquifer and Floridan aquifer.  The most complex of these factors
is evaporation.  The question arises as to what percent of time in a
certain period was the variation of dewpoint temperature with height
such as to lead to condensation on or evaporation from the lake sur-
face.  An inversion of the dewpoint will develop if the surface acts as
a heat sink to remove water vapor.  This condition can be expected during
clear nights when there is strong radiation from the ground; during
times of high relative humidity; and during times of build-up of sur-
face inversions which occur frequently in the Anderson-Cue Lake area.
In fact, about 50 percent of the time the water vapor flux is directed
downward.  This reversal of evaporation is evident during non-daylight
hours when winds are persistently less than 7 mph and relative humidity
greater than 90 percent.

       It is interesting to note from the rain gage records of the past
year that during the periods when the vapor flux is directed upward
(generally from 0800 hr to 1800 hr) approximately 0.12 to 0.15 of an
inch of water is evaporated daily.  This indicates the large amount of
evaporation from the lake surface that can be expected unless the amount
of precipitation plus condensation received when the lake acts as a
heat sink can overcome the evaporation losses and losses to the water-
table aquifer and Floridan aquifer.

       Water in the water-table aquifer is unconfined so that its surface
is free to rise and fall with the variance in rainfall.  Rainfall on
the Anderson-Cue Lake basin for the period March 28, 1966, through
June 30, 1968, was deficient by 13.85 in. (the closest "departure from
normal" data are accumulated at Gainesville, approximately 20 miles to
the west).  For the period November 2, 1967 through June 30, 1968 the
deficiency was 8.56 in.  This is shown in Figure IV-5 (refer to Figure
IV-2 for location of test wells).  Because the piezometric surface of
the Floridan aquifer is below the level of the lake, water cannot move
from the Floridan aquifer to the lake.  The net groundwater flow during
the period of study has been composed only of outflow to the water-
table aquifer and to the Floridan aquifer, the greater flow being to
the water-table aquifer east of the lake.  Note the level of Well No. 4
in Figure IV-5.

       As shown in Table IV-1, for the period March 28, 1966 through
November 30, 1968, evaporation losses exceeded rainfall by 8.35 in.
and approximately 65 acre feet of lake water was lost to the aquifers.
                                -19-

-------
125
123 -
122
121
120
119
    **§
    =4 U-
                         _c
                         30
                         •d
     ' <
     1 Oi
         \
         (N
.E

                                                    §
                CO


      \
              x%
                             X  V
                                     \\
            Well No.
            1
            2
            3
            i,
—-—-—.-   Lake Level
                                           \ ^
                                             \   '

                        See Fig. IV-2 for well locations.
                                                          f
                                                                           ^•••••*
                        Figure IV-5. Ground and Surface Water Levels —Anderson-Cue Lake.

-------
                                TABLE IV-1
                    ANDERSON-CUE LAKE HVTDROLOGICAL DATA
 Dates

 3/28/66
 4/26/66
.5/25/66
 6/22/66
 7/21/66
 8/18/66
 9/13/66
 10/11/66
 11/08/66
 12/06/66
 1/03/67
 3/28/67
 5/02/67
 5/31/67
 6/27/67
 3/01/67
 9/06/67
 10/03/67
 11/02/67
 12/03/67
 1/04/68
 1/31/68
 2/29/68
 3/31/68
 4/30/68
 5/31/68
 6/30/68
 7/30/68
 8/31/68
 9/30/68
 10/31/68
 11/30/68
 12/30/68
 1/31/69
 2/28/69
 3/31/69
 4/30/69
 5/31/69
 6/30/69
 7/31/69
          (1)
Lake Evaporation (in.)
                                            (2)
                                      Rainfall  (in.)
              (est)
            Summary:   186.73 (1)
                     -171.69 (2)
                       15.04 in.
        (3)
Lake Level (ft-MSL^

       125.78
       125.33
       125.04
       124.69
       124.25
       124.43
       124.41
       124.51
       124.21
       123.75
       123.61
       123.61
       122.99
       122.83
       123.07
       123.33
       123.81
       123.66
       123.46
       122.99
       123.07
       122.63
       122.17
       121,59
       120.83
       120.49
       120.02
       119.54
       120.45
       120.75
       121.09
       121.19
       121.09
       120.87
       120.89
       120.67 (eat)
       120.45
       120.49
       120.01
       120.87
      -(55.40 in.)
                                    (3)
                                          55.40
                                         -15.04
                                          40.36 in. or approximately
                                                56 acre/ft.
NOTE:   Lake evaporation is computed from data collected at the U.S.
Weather Bureau evaporation station at Gainesville, Florida.  Pan coeffi-
cients are those used for" Lake Okeechobee, Florida: Kohler, M.A., 1954,
Lake and Pan Evaporation in Water Loss Investigations-Lake Hefner Studies,
Technical Report: U.S. Geological Survey Prof. Paper 269, p. 128.
                                    -21-

-------
For this period, the residence time of water in Anderson-Cue Lake has
been calculated to be 5.43 years.  The same calculation holds true for
McCloud Lake (control) which rises and falls at the same time and in
the same proportion as Anderson-Cue Lake.
                                -23-

-------
                                SECTION V

                            CHEMICAL STUDIES

Introduction
       Trophic state is manifested by a variety of chemical and
biological parameters.  This section will summarize the routine chemi-
cal data obtained on the two study lakes; biological results will be
presented in the following section.  Results for Anderson-Cue Lake ex-
tend for a period of four years-from 1966 to the present.  McCloud
Lake has been sampled routinely since the beginning of 1967.  During
1966 and 1967 sampling was approximately bi-weekly, especially for the
important nutrient parameters.  Sampling has been on a monthly basis
since January, 1968, because short term variations have been found to
be rather small.  Monthly sampling has also allowed more time for
other special studies.  The objective of routine sampling is to define
changes in the chemical and biological composition of the lake as it
undergoes controlled eutrophication.  To accomplish this requires a
representative sampling program with careful attention to possible
temporal, lateral and depth variations in composition.

       Three permanent sampling stations were located in Anderson-Cue
Lake.  Stations 4 and 7 are in the centers of the lake's two basins,
and Station 8 is on the south shore in about 3 feet of water.  The lo-
cation of these stations is shown in Figure IV-1.  Two permanent sta-
tions are located in McCloud Lake, Station 11 near the center of the
lake, and Station 12 near the north shore in 3 feet of water.  Samples
were taken at three depths  (top, middle and bottom) at Stations 4, 7
and 11, and at mid-depth at the shore stations (8 and 12).  For more
detailed determinations of  lateral variability, a sampling grid of
about 50 stations was established on Anderson-Cue Lake; these are in-
dicated in the isopleth maps shown in later discussions of these studies.
Parameters measured routinely  (bi-weekly or monthly) include dissolved
oxygen, pH, conductivity, acidity, dissolved and suspended solids,
ortho and total phosphate,  total and particulate organic nitrogen,
ammonia, nitrite, and nitrate.  In addition, data have been routinely
collected on physical conditions such as water temperature and Secchi
disc transparency.  Other major and minor chemical constituents have
been determined less  frequently.  These  include chloride, sulfate,
calcium, magnesium, sodium, potassium, silica, iron, manganese, chemical
oxygen demand and biochemical  oxygen demand.  Chemical characterization
of  lake sediments has included determination of percent volatile solids,
total organic nitrogen, ammonia, total phosphate, iron, and manganese
(see Appendix A for methods).
                                -23-

-------
 Chemical  Characteristics

       The  two  lakes are  typical of  the small  lakes in the Trail Ridge
 portion of  the  Central Highlands of  Florida.   Table V-l summarizes the
 chemical  characteristics  of  the lakes.  Few significant changes in
 gross chemical  composition have been noted during the period of record;
 hence the values  in Table V-l are mean values  for each lake during the
 period of record.  Both lakes are colorless, low in dissolved solids
 and extremely soft.  The waters are  acidic, with typical pH values
 ranging between 4.6 and 5.5.  The waters have  little buffer capacity and
 essentially no  alkalinity.   Consequently, acidity titrations have been
 used to estimate  total C02.  Specific conductance has increased in
 Anderson-Cue Lake from about 25 ymho cm~l to about 40 ymho cm~l over the
 last two  years.  Corresponding increases in McCloud Lake have been less-
 from 30 to  35 ymho cm~*.  Some of the increase would seem to be the re-
 sult of excess  evaporation over precipitation during the period; nutrient
 additions were  probably responsible  in part for the increase in the ex-
 perimental  lake.

       The  low  dissolved solids and  ionic content of the lakes are
 indicative  of the waters' origin, i.e. atmospheric precipitation.
 Table V-l lists some comparative values for the chemical composition of
 rain water at the lake site.  Concentrations of major ions compare
 reasonably close for the lakes and rain water.  The ionic content of
 rain water varies considerably, but  the values in Table V-l represent
approximate ranges for the various ions.  The data are too sparse for
 reliable estimates of mean rainfall  composition, which would be useful
 in deriving a chemical model for the lake waters from rainfall compo-
 sition and possible chemical interactions between rain water runoff and
 soil constituents.  The primary reason for studying the composition of
 rain is to determine its significance as a nutrient source.  The rain-
 fall values for nitrogen and phosphorus species in Table V-l are mean
values revealing the importance of rain as a nutrient source, especially
 for lakes unaffected by cultural sources.

       While concentrations of major ions are not likely to limit primary
production in either lake, the paucity of several is likely to select
against certain types of organisms.   Low silica probably is a contribu-
 tor to the small diatom populations; low calcium and magnesium indicate
 the waters are unsuitable for macrophytes like Chara and some algae
which prefer hard water.  The low pH of these lakes is undoubtedly a
contributing factor for the  low populations of blue-green algae, but
encourages the maintenance of a desmid population.   The relatively low
nutrient  levels favor organisms like Dinobryon and Synura,  which pre-
 fer such environments (Hutchinson,  1967 ).

-------
                               TABLE V-l

                  CHEMICAL  COMPOSITION OF
                   AND McCLOUD LAKES AND
ANDERSON-CUE
RAIN WATER
Constituent Anderson-Cue Lake
Specific conduct.
PH
Acidity as CaCO»
Cl"
S04=
Na
K+
Ca+2
Mg+2
Silica
Total org. N
Particulate org. N
NH3-N
N03--N
Ortho phosphate
Total phosphate
COD
BOD
Sus. solids
Turbidity
31.
4.
3.
6.
5.
2.
0.
0.
0.
0.
0.
0.
0.
0.
0.
65
93
20
25
4
49
51
74
58
14
47
26
234
0014
067
0.0084
0.017
10.7
1.02
5.9
11.1
McCloud Lake
32.29
4.85
3
5
5
2
0
0
0
0
0
0
0
0
0
0
0
10
0
5
8
.50
.93
.0
.81
.25
.61
.57
.10
.42
.21
.105
.0012
.041
.006
.012
.7
.86
.2
.8
Rain Water
10-30
5.3-6.8
1.
0.
0.
0.
1.
0.

0.

0.
0.
0.
0.
0.




74
8
29-1.
13-0.
01-2.
06-0.

32

208
005
209
027
033





85
21
06
35











 All values in mg/1 except specific conductance (ytnho con  ) and pH.
 Nitrogen species are in mg N/l and phosphate in rag P/l.

 Mean values for all samples from the mid-lake stations during 1967
 and 1968.

3
 Nutrient (and Chloride) concentrations are mean values for all deter-
 minations in 1968; other values represent range encountered in one to
 four determinations during 1967-68.
                                  -25-

-------
                                TABLE V-2

                    TEMPERATURE AND DISSOLVED OXYGEN:
          ANNUAL AVERAGES AT THREE DEPTHS IN STATIONS, 4, 7, 11
                                1                                 2
Station &            Temperature                  Dissolved Oxygen
Depth          1967      1968      1969       1967      1968      1969

4   Top        23.61     22.11     23.05      7.39      7.94      9.08
    Mid        23.06     21.74     22.55      7.39      7.96      8.05
    Bottom     22.91     21.57     22.66      7.27      7.48      8.20

7   Top        23.42     22.13     23.09      7.66      8.04      8.48
    Mid        23.13     21.83     22.71      7.56      8.03      8.34
    Bottom     22.90     21.08     22.43      7.37      7.30      8.18

11  Top        23.49     22.65     23.28      7.58      8.07      8.35
    Mid        22.78     21.98     22.49      7.50      8.02      8.25
    Bottom     21.93     21.73     22.04      6.26      6.90      7.28

 Temperature in C
2
 Dissolved Oxygen in mg/1
Temperature and Dissolved Oxygen

       Neither lake shows much evidence for stable thermal stratification
at any time of the year.  Table V-2 summarizes average temperature and
dissolved oxygen at the three depths sampled for Stations 4, 7, and 11.
The maximum difference in average annual temperature from top to bottom
was about 1.5° C for Station 11 in 1967; differences at the other sta-
tions have been 1°C or less.  Somewhat larger differences at Station 11
have been found from top to bottom on particular days.  During the period
of high water in summer of 1967, 3°C differentials were sometimes found,
but the changes occurred in the bottom few feet, and most of the lake
was freely circulating.  Temperature profiles in Anderson-Cue Lake are
normally within one degree Celsius from top to bottom.  During periods
of intense warming and calm weather, temporary stratification could
occur in either lake, but we have not yet found such conditions in over
three years of sampling.  Water temperatures range from about 12°C in
winter to about 32°C in mid-summer.

       Dissolved oxygen profiles also show little change with depth.
Average differences at Stations 4 and 7 in 1967 were only 0.12 and
0.29 mg/1,  respectively.  Slightly greater differences occurred in
1969; 0.75 and 0.95 mg/1 at Stations 4 and 7, respectively; these may
reflect the somewhat greater production and standing crop in Anderson-
Cue in 1969.  Vertical differences at Station 11 were greater (1.2 -
1.3, mg/1) than the Anderson-Cue results for the three years, corrobor-
ating the greater vertical stability of this lake.  There is no evidence
                                -26-

-------
of oxygen depletion in the bottom water of either lake at any time,  but
considering the lack of thermal stratification and oligotrophic condi-
tions, this is not surprising.  Seasonal variations in dissolved oxygen
largely reflect changes in solubility with temperature.  Figures V-l
and V-2 show the average temperature and dissolved oxygen values for
1967, 1968, and 1969 at Stations 7 and 11, respectively.  Oxygen values
generally were near saturation, but a tendency toward slight under-
saturation is noted.  Rates of photosynthesis and respiration in either
lake are too slow to markedly influence dissolved oxygen, but this should
change in Anderson-Cue Lake as nutrient additions are continued.
Variations in Biogenic Compounds

       In routine chemical analyses, greatest attention has been
centered on nitrogen and phosphorus compounds, which presumably are
most critical for primary production, and on other substances whose
concentrations are affected by the activity of organisms.  Both lakes
were extremely poor in nutrients before enrichment began.  Ammonia
ranged between 0.02 and 0.06 mg N/l; nitrate was less than 0.04 mg N/l,
and total organic nitrogen averaged about 0.3 mg N/l.  Ortho-phosphate
was often undetectable and averaged less than 5 yg P/l.  Total phosphate
exhibited similarly low concentrations.  The above concentration ranges
are for 1966 and early 1967, before nutrient enrichment of Anderson-Cue
Lake.  Enrichment began in March of 1967 and is still continuing.  Fig-
ures V-3 to V-7 show the seasonal variations in nitrogen and phosphorus
forms in the two lakes from January, 1967 to December, 1969.  The points
on each plot represent mean values for the mid-lake stations in each
lake.  Differences between the two lakes may have been somewhat greater
than the plots indicate.  Occasional high nutrient concentrations were
encountered in the bottom sample of McCloud Lake; these may have re-
sulted from stirring the sediment during sample collection.  These are
included in the averages, but are probably not representative of the
lake as a whole.

       Seasonal patterns in both lakes are rather similar.  With the
exception of total organic nitrogen  (Figure V-3), nutrient concentra-
tions are'consistently higher in Anderson-Cue compared to McCloud Lake.
Total organic nitrogen does not appear to exhibit a marked seasonality
in either lake.  The' high and erratic nature of  the data in mid-1967 is
partially the result of analytical difficulties subsequently corrected.
Concentrations were usually slightly higher in the experimental lake
during 1968 and 1969.

       The effect of enrichment on ammonia concentrations is clearly
illustrated in Figure V-4.  Ammonia has been consistently higher in
the experimental lake throughout 1967, 1968, and 1969, but differences
became much more pronounced during 1968.  There does not seem to be a
major seasonal influence on ammonia; rather concentrations fluctuate
considerably from month to month.  It  is interesting to note that the
                                  -27-

-------
ro
oo
         I -
                            0,0,
                       1	1	1	1	1	I	1	1	1	1	1	1	1	1	i ._  i  i   i   i   ii
             J  F  M  A   *   I  JAS   ONDJF   HAMJJASONOJFMAUJ  JAS   OND
                            IW7                                IMS                               list

                        Figure V-l.  Dissolved Oxygen and Water Temperature  in Anderson-Cue Lake
                                          Me Cloud U*.
          E "
          I »
           I
           <5
           1
                                             1 - 1 - 1 - 1 - 1 - 1
                                                                                       1 - 1 - 1 - 1
               JfllAMJJASOHDJFtlAIIJJASONO.JFMAHJJASONO
                              1967                                 l«6                                tlfj


                            Figure V-2.  Dissolved  Oxygen and Water Temperature  in McCloud Lake

-------
N5
VO
I
                 FMAUJJASONOJFyA
                          Figure V-3.  Total Organic Nitrogen in Anderson-Cue and McCloud Lakes
                            J  JASONDJ   F  M  A  M  J   J
                                   Figure V-4. Ammonia  in Anderson-Cue  and McCloud Lakes

-------
biological forces within the lake can exert an over-riding influence on
the general trend toward increased concentrations resulting from nutrient
input.  A decrease from about 0.34 mg NHo-N/1 to 0.06 mg NH^-N/l occur-
red in Anderson-Cue Lake during February, 1968.  This corresponded to a
winter bloom of Dinobryon and Synura.  After the bloom, ammonia increased
to 0.40 mg N/l within the next month.  A similar but smaller decrease
and subsequent increase occurred contemporaneously in McCloud Lake.
Ammonia was being continuously fed into the experimental lake at a rate
equivalent to a 0.1 mg N/l average increase in the water for the two
month period (February and March).  With some fluctuations ammonia con-
tinued to increase in Anderson-Cue Lake during 1968 while values in
McCloud Lake showed a much smaller trend.  Ammonia in Anderson-Cue Lake
has now increased to levels commonly considered indicative of eutrophy.

       Nitrate seasonal patterns (Figure V-5) are nearly identical for
the two lakes, but the experimental lake has consistently higher values
(by about 0.01 to 0.09 mg NCL-N/l).  The peak concentrations for both
lakes in 1968 occurred in early February; minimum values were found in
late May.  The seasonal pattern probably can be explained by uptake of
nitrate during the late winter bloom and inhibition of nitrification at
warm summer temperatures.  Nitrate concentrations have been below
0.10 mg N/l except during the winter of 1968.  No winter maximum was
found in 1967.  In 1969, peak concentrations occurred in early March
and minimum values occurred in June and July for both lakes.

       Ortho-phosphate concentrations in both lakes (Figures V-6) are
normally quite low and follow a similar seasonal pattern.  Peak con-
centrations occur from late spring to mid-summer.  Values were below
10 yg P/l in both lakes during winter and early spring and again in
fall of all three years.  The low phosphate values (except during sum-
mer) indicate phosphorus is probably the limiting eutrophying factor in
Anderson-Cue Lake.  The seasonal pattern of ortho-phosphate is rather
the opposite of that found in north temperate lakes and is somewhat un-
expected.  This pattern (summer maximum) has been found for ammonia in
some Polish lakes (Karcher, 1939).  Data on seasonal variations of phos-
phate in other unstratified lakes are rather sketchy.  There seems to
be a consistent trend in the total phosphate of both lakes to higher
values in summer and minimum values in early winter.  With few exceptions
concentrations were higher in the experimental lake than in the control,
but the differences were usually not striking.  Concentrations increased
greatly in both lakes in early fall of 1968 and reached maximum values
of 93 and 84 yg P/l (average concentrations) in Anderson-Cue and McCloud
respectively, in mid-October.  Data for this period in 1967 are not
available for comparison.  In August, 1969, total phosphate concentra-
tions reached maximum values of 31 and 23 yg P/l (average concentrations)
for Anderson-Cue and McCloud respectively.  Such a rapid and large in-
crease in total phosphate and subsequent rapid decline would seem to
imply an important role for the sediments as a phosphorus source and
sink since they alone would seem capable of providing such amounts to
the water.
                                -30-

-------
                                     J  J   ASONDJFUAHJ  J   »   S   0  N  B
          Figure  V-5.  Nitrate in Anderson-Cue  and McCloud Lakes
35

30

JS

X

15

10
                                             	McOwd
 \
             A
A  •  J  J   A  •  0  H  B   JfK.H       *
      IM7                               1961
                                                         J  F  U  *
                                                                      J  J   *
                                                                                 0  N  0
       Figure V-6. Ortho-Phosphate in Anderson-Cue and  McCloud Lakes

-------
I
Ui
10
                              Figure V-7.  Total Phosphate in Anderson-Cue and McCloud Lakes

-------
       During the first 21 months of nutrient enrichment (through
December, 1968), approximately 217 kg nitrogen and 18.5 kg phosphorus,
mostly as ammonia and ortho-phosphates, were added to Anderson-Cue Lake
through the nucrient outfall.  This was sufficient to increase the N
and P levels in the lake by 0.87 and 0.082 mg/1, respectively, at the
lake's volume in 1967, if all the nutrient material remained in the
lake.  The actual nutrient increase should have been larger because of
the decrease in lake volume in 1968.  Inspection of Figures V-3 to V-7
shows this theoretical situation clearly does not apply.  Increases in
total N and P concentrations do not approach these levels and much of
the added nutrient evidently was deposited in the sediments or was lost
through ground water seepage.  This has been found to be the case in
other lakes where nutrient budgets have been constricted.  This would
seem to imply an important role for sediment regeneration of nutrients
in the eutrophication process.  Possibly the onset of deleterious con-
ditions in the eutrophication process is contingent upon exhaustion of
the sediment's capacity to retain nutrients.

       Considerably greater detail is known about the variations of
chemical species in the lakes than was presented above.  The data in
Figures V-l to V-7 represent mean values for each lake on a particular
date.  Data were collected from the three depths at each mid-lake sta-
tion on each date.  Table V-3 summarizes results for some biogenic
elements at the three depths; the data indicate the lakes are well mixed
vertically (as implied by the temperature and dissolved oxygen data in
an earlier section), and the vertical differences in chemical species
were usually very small.  Differences between the stations in each lake
are also small for most chemical species.  Occasionally, parameters
such as total organic nitrogen have exhibited significantly higher values
at the shore stations, apparently because of slough-off from littoral
vegetation.  Table V-4 is a summary of some data from the three stations
in Anderson-Cue Lake and the  two stations in Me Cloud Lake.  Because of
the large number of samples even the small differences shown for some
parameters are statistically  significant, but it seems unlikely that the
differences would be ecologically significant or that they would change
one's opinion about the lakes1 homogeneity.

       A detailed study of the lateral variations in ammonia and ortho-
phosphate was conducted in January, 1968, and gives further evidence of
the experimental lakes' comparative homogeneity.  This is not to say
that there are no differences at all.  Figure V-8 shows a slight trend
for higher ammonia near the southern shore.  In general, ortho-phosphate
was higher in shore areas than in the  lake center  (Figure V-9), but
values for the southern shore were  the lowest in the lake.  The high
ortho-phosphate values in the northwest portion of the lake probably
represent a minor source of pollution  from cattle grazing in this area
during this period.  The results indicate that the routine stations are
representative of the conditions throughout the lakes within the  limits
of accuracy desired for this  project.  A detailed survey of the area
surrounding the nutrient outfall was conducted in January, 1969.  The
results shown in Figure V-10  imply  rapid mixing in the lake since no
concentration gradients resulting from nutrient additions were  found.
                                 -33-

-------
                                   TABLE V-3

                 COMPARISON OF AVERAGE CONCENTRATIONS OF SOME
                     BIOGENIC PARAMETERS AT THREE DEPTHS
                            IN STATIONS 7 AND 11.
Parameter* Station N*
pH



TOH



PON



HH3




O-P04


t-P04



7

11

7

11

7

11

7

11


7

11
7

11

19
10
19
10
18
10
18
10
11
9
11
9
20
10
20
10
21
10
21
10
18
9
18
9
Year
1967
1968
1967
1968
1967
1968
1967
1968
1967
1968
1967
1968
1967
1968
1967
1968
1967
1968
1967
1968
1967
1968
1967
1968
Top
Mean S.D.
4.67
5.20
4.90
4.97
0.41
0.33
0.40
0.30
0.15
0.17
0.14
0,17
0.104
0.363
0.052
0.136
4.0
11.5
3.9
7.5
9.3
24.6
7.7
10.6
0.50
0,49
1.12
0.31
0.15
0.06
0.19
0.06
0.08
0.05
0.23
0.06
0.057
0.199
0.033
0.083
4.6
8.8
5.8
5.4
9.9
29.0
8.8
6.1
Mid
Mean S.D.3
4
5
4
4
0
0
0
0
0
0
0
0
0
0
0
0
5
12
3
8
12
20
9
12
.75
.08
.75
.93
.49
.37
.43
.37
.23
.22
.19
.19
.126
.368
.060
.133
.5
.0
.3
.4
.8
.1
.7
.4
0.64
0.24
1.29
0.34
0,20
0.09
0.16
0.07
0.20
0.07
0.16
0.07
0.112
0.179
0.036
0.068
7.0
8.9
4.9
5.9
7.4
9.3
8.2
5.4
Bottom
Mean S.D. _
4.
5.
4.
4.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
6.
13.
3.
11.
11.
24.
13.
18.
74
27
70
92
65
38
50
38
35
21
26
21
102
393
093
179
1
4
1
9
1
0
4
1
0.72
0.35
1.25
0.48
0.48
0.09
0.22
0.11
0.45
0.08
0.27
0.10
0.069
0.191
0.077
0.097
8.2
11.3
5.0
12.7
8,2
7.4
9.5
15.7
  Values for total organic nitrogen (TON), particulate organic  nitrogen  (PON)  and
  ammonia in mg N/l; ortho and total phosphate in yg' P/l.

^ "H * number of determinations.  Data from 1968 are for January to October, while
  data from 1967 are for entire, year.

  S.D. • standard deviation • U (x-x)^.   These are standard deviations of the results
                             I  M—l
  from the particular station and depth for the given year and  reflect the annual
  variability rather than the' analytical precision of the  test.  They are presented
  primarily to indicate the former rather than for further statistical testing,
                                     -34-

-------
                                   TABLE V-4

                 COMPARISON OF AVERAGE  CONCENTRATIONS OF  SOME
                          BIOGENIC PARAMETERS  AT  THE
                           ROUTINE SAMPLING STATIONS
                            Anderson-Cue Lake          McCloud Lake
                       Station   Station   Station   Station  Station
Parameter
PH
Acidity
TON
?N
NH3N
N03N
0-P04
t-P04
COD
BOD
Year
1967
1968
1969
1967
1968
1969
1967
1968
1969
1967
1968
1969
1967
1968
1969
1967
1968
1969
1967
1968
1969
1967
1968
1969
1968
1969
1967
1968
1969
4
4.70
5.10
4,80
3.5
2.8
3.5
0.54
0.39
0.35
0.38
0.21
0.22
0.101
0.356
0.190
0.043
0.097
0.075
3.9
12.1
4.3
11.7
24.2
19.4
8.4
5.64
0.86
1.14
0.74
7
4.72
5.18
4.83
3.6
3.0
3.3
0.52
0.36
0.29
0.24
0.20
0.18
0.111
0.374
0.215
0.040
0.090
0.078
5.2
12.3
2.6
11.1
22.8
16.4
9.7
5.64
0.89
0.80
0.38
8
4.76
5.15
4.77
3.3
2.9
2.9
0.62
0.59
0.27
0.37
0.24
0.16
0.112
0.337
0.198
0.041
0.088
0.079
4.2
13.3
2.9
11.6
22.8
19.3
20.7
3.76
1.83
1.33
0.19
11
4.78
4.94
4.98
4.1
3.3
2.9
0.44
0.35
0.34
0.20
0.19
0.13
0.068
0.149
0.070
0.024
0.058
0.022
3.7
9.2
3.7
10.3
13.7
15.0
11.4
5.64
0.96
0.58
0.48
12
4.64
5.03
4.98
3.1
2.6
2.4
0.48
0.49
0.39
0.26
0.29
0.22
0.064
0.121
0.074
0.019
0.064
0.025
2.9
7.6
3.1
9.2
15.3
20.0
8.5
3.76
0.73
1.54
0.16
^•Acidity in mg/1 as CaCOs;  ortho  and total phosphate  in  yg P/l; all other values
   in mg/1.  PN stands for  particulate  organic  nitrogen.
                                        -35-

-------
figure V-B.  Loftroi Verionons ot Amir.onio in Andeisoo-Cuo
            Lcke, Jc/iucry, J9cS, Concentrations in rr-3 N/l.
            Lccctions &f routine wmplinj stations shewn In
            b!oc!cs.
                                                  Fl8«r» V-». LoUral Vcrictiow of Ort!-o-Fh«ishc!o In
                                                             Andcisc.n-Cuc Loli«, January, I*o3.
                                                             Concanffotiew In c« P/l, Lccotlon of
                                                             I«U»IM sampling itoliora shown In blocU
                                           -36-

-------
u>
                                                                                           o   '•*-
                                            Figure V-10. Ammonia and Phosphorus Concentrations
                                                        Around Nutrient Outfall  — January 1969.

-------
       Several diurnal studies in which samples were taken every hour
or every two hours over a 24 hour period have also been made, but in most
cases the small variations appear random rather than cyclic.  A few par-
ameters do show a measurable diurnal variation; Figure V-ll illustrates
this behavior for several biogenic species during a diurnal study from
January 31 to February 1, 1968.  As Anderson-Cue Lake becomes further
enriched and more productive, biogenic species will undoubtedly show
greater diurnal periodicity, and studies of the lake's diurnal varia-
tions may be a useful indicator of the lake's advancing eutrophy.
Sediment Studies

       Sediments may exert considerable influence on the eutrophication
process, both as nutrient sources and as nutrient sinks.  Furthermore,
they can exert large oxygen demands on the overlying water and provide
a home and substrate for organisms which spend all or part of their life
cycles as benthos.  A variety of studies has been undertaken to provide
a better understanding of the role of lake sediments in eutrophication.

       As a first step, chemical characteristics and variations in
sediment types have been determined for the study lakes.  Representative
results for Anderson-Cue Lake are shown in Figures V-12 through V-14.
Several sediment types occur in the lake:   near shore, the bottom is
sand covered with a thin layer of loose detritus and periphyton.  In
parts of the deep regions, peat-like sediments are evident with much
fibrous and undecomposed plant material.  In other areas the sediments
are darker and finer grained, more like the ooze or sapropel of alka-
line lakes.  The sediments of McCloud Lake have not been as well charac-
terized, but peat-like sediments are less in evidence there.  The
results indicate that sediments from these lakes are actually higher in
nitrogen and phosphorus than sediments from some eutrophic lakes.  For
example, sediments from Lake Mendota, Wisconsin, have from 200 to 1200
phosphorus and 2000 to 14,000 ppm total organic nitrogen (Hasler, 1963).
The sediments in this alkaline lake are over 30 percent precipitated
calcium carbonate, whereas those in the lakes of this study are composed
largely of organic matter.  The absence of carbonate deposits in the
sediments of the study lakes permits volatile solids determinations to
approximate the organic content of these sediments.  The high values
in Figure V-14 indicate the largely organic nature of these sediments.

       The sediments in the study lakes are obviously enriched with
nitrogen and phosphorus compared to the overlying water and thus re-
present potential nutrient sources.  However, most of the nitrogen is
present in the organic form rather than as free ammonia (Figure V-12),
and most of the phosphorus is also bound (presumably organic) rather
than free ortho-phosphate (Figure V-13).  Leaching and incubation
studies undertaken to reveal the importance of sediments in nutrient
storage and release must consider the wide variations in available nu-
trient content of the various sediments in the lake as well as the
                                 -38-

-------
w
vo
i
                                16.0 r
                             u
                             o
                             a.

                             v
                                15.0
                                14.0
  4)


 Vt ^"*


5(5
                                10.0
                          c— co
                             O
                                 2.0
                                  1.0
                           "-  U
                          o<
                                                                                             Temp.
                                                                                                                                   0.5
                                                                                                       - 0.25
                                                                                                                                       o


                                                                                                                                       a
                                                                                                                                       'c
J	L
                                                       _L
J	L
J	L
                                                                            J_
                                                                                                                    J	L
                                          12N          4PM

                                               Jan. 31
                                          8PM
                                        12M




                                       Time
                                 4AM           SAM

                                      Feb.  1
                                        12N
                                                       Figure V-ll.  Diurnal Variations in Anderson-Cue Lake.

-------
                      N
                                                                  N
Figure V-12, Nitrorjcn in Anderson-Cue Scsli-Tcnts. Top
            nuTtber is QT.frc.nta; l:o.Vm number is total
            or.y!n!c nitro^n. Values in mg N/9 dry \vt
                                                   Figuro V-13. PliospMare in Anttcnon-Cjc Sadimcnn. Top
                                                                nuin'^or is oriho phosphoU, bollofl ntir.Uf it
                                                                total phcsphotc value! in /ifl P/g dry wt. of
                                                                sediments.
                                          -40-

-------
                    N
Figure V-14. Percent Volatile Solids (Top Number) and
              Tofal Iron (Bottom Number) In Anderson-Cue
              Sediments.  Iron values in mg Fe/a dry wt. of
              sediment.
             -41-

-------
varying opportunities for transport  from sediment  to water afforded at
different  locations  in  the  lake.  For example, bottom currents are com-
paratively slow in the  center or deep portions of  the lakes and nutrient
exchange may be a diffusion-controlled phenomenon.  On the other hand,
the thin sediment layer in  the sandy littoral areas is frequently and
easily mixed with the overlying water by wind-generated currents and
waves and by movement of fish.  Nutrient release from these sediments is
probably controlled  by metabolic rates rather than by physical factors.
The overall question of nutrient exchange between  sediments and water
(in either direction) is extremely complex and a completely satisfac-
tory answer is perhaps beyond the "present state of the art".

       A variety of  sediment exchange experiments, including laboratory
and in situ studies, have been undertaken or are presently underway or
planned.  Some initial results of a  laboratory incubation study are
shown in Figures V-15 and V-16.  Sediment from the middle of Anderson-
Cue Lake was incubated  in 10 liter bottles under varying conditions, and
ammonia and ortho-phosphate in the over-lying water was followed over a
period of 20 days.  About 0.5 liters of sediment was placed in each
bottle.  Incubation was conducted in the laboratory at 22°C under condi-
tions of (artificial) laboratory lighting.  One bottle (A) was incubated
under the above conditions; a second bottle was treated similarly ex-
cept that light was  excluded (C), a  third bottle (B) was continuously
purged with nitrogen to maintain anoxic conditions and a fourth bottle
(D) was stirred to keep the sediment mixed with the water.  Changes in
the ammonia content  of the overlying water are shown in Figure V-15.
Anoxic conditions allowed more ammonia leaching in bottle (B) than in
the oxygenated control  (A).  However, stirring was more effective and
maximum ammonia was  released immediately.  A similar situation occurred
with ortho-phosphate (Figure V-16).  Anoxia induced the release of
considerably more phosphate than oxygenated conditions in the control.
The dark bottle showed consistently  lower ortho-phosphate than in the
control, but the trend was similar.  Again mixing was most effective
in liberating phosphate.  The maximum phosphate was released almost
immediately and the concentration declined about 50 percent over the 20
day incubation.  It  is obviously premature to extrapolate these results
to the in situ role  of sediments in nutrient recycling.  But it would
seem that mixing is  the most effective mechanism to release sediment
nutrients into the water if that were desired.   Alternately it is appar-
ent that the single most effective means of limiting release is preven-
tion of mixing.  In deep lakes relatively little mixing occurs.  However*
during periods of high winds, sufficient currents may be generated in
shallow lakes and littoral zones of deeper lakes to stir sediments with
the water and release considerable amounts of nutrients.
                               -42-

-------
.6
.5
.4
.3
                                                    D (xO.5)
                           10
15
20
                             Days
   Figure V-15. Temporal Changes of Aqueous Ammonia In Lake Water
                Incubated With Sediment Under Varying Conditions:
                A, control;  B, anoxic; C, dark;  D, stirred.
                                                    20
                             Days
   Figure V-16. Temporal Changes of Aqueous Ortho-Phosphate in
               Lake Water Incubated With Sediment Under Varying
               Conditions (see Figure V-15 for key).
                       -43-

-------
                               SECTION VI

                                 BIOLOGY

Introduction

       Biological studies of Anderson-Cue and McCloud Lakes have
included:   (1) the species, succession and productivity of algal forms,
(2) the standing stock of phytoplankton, (3) limiting nutrients for
phytoplankton growth, (4) standing crop estimates of littoral vegetation,
and (5) production estimates of the zooplanktivorous fish Labidesthes
sicculus.  In addition monthly diurnal variations in primary productivity
have been noted along with plant pigment levels as a measure of algal
biomass.  Two special 10 day studies dealt with an analysis of environ-
mental factors affecting primary production (Brezonik e_t ad, 1969).  The
methods for these procedures are outlined in Appendix B.
Methods Developed

       Except for the special techniques described below, all biological
observations in this study were carried out using standardized tech-
niques such as outlined in Standard Methods for the Examination of Water
and Wastewater (1965) and FWQA Analytical Techniques for the National
Eutrophication Research Program (June, 1969).  Special techniques to
determine population dynamics and productivity of zooplankton were de-
veloped and are described below.

       Zooplankton samples were taken from January, 1967 through
December, 1968, in Lake Anderson-Cue and from May, 1967, through
December, 1968, in Lake McCloud.  From January to August, 1967, samples
were taken bi-weekly with a Wisconsin closing plankton net (125 meshes/
inch) towed horizontally for a known distance at three depths corres-
ponding to just below the surface, mid-depth, and just above the bottom.
Aliquots from these samples were counted; counts were adjusted to
No./m3 and integrated over depth to obtain No./mz.  After August, 1967,
weekly samples were taken with a vertical-tow plankton net (125 meshes/
inch) pulled from bottom to surface.  When using the vertical-tow net,
samples were taken from three stations in the lake and each sample con-
centrated to 35 ml.  The three samples were combined, mixed thoroughly,
and a sub-sample of 35 ml taken from the mixture for preservation with
formalin and for counting.  This procedure provided a physical means
of averaging samples from three areas of the lake while counting only
one sample, thus providing a more representative sample from the lake.

-------
       Zooplankters were counted using a compound microscope at a
magnification of 21 diameters.  Each sample was shaken thoroughly; a 1 ml
aliquot was then taken with a graduated syringe and placed in a Sedge-
wick-Rafter counting chamber.  Three such aliquots were counted for each
sample and all zooplankters in the chambers were identified and counted.
Occasionally, when phytoplankton was especially numerous, it was neces-
sary to dilute a sample before counting.  Counts from samples taken with
a vertical-tow net were converted directly to No./m2 based on the area of
the net mouth; e.g., a net with a 0.2m2 mouth area towed from bottom to
surface would collect the plankton under 0.2m2 of lake surface.
       Biomass determination.  To determine the mean individual biomass
of a zooplankton species, individuals were sorted from a sample under a
dissecting microscope, blotted, dried under vacuum desiccation, and
weighed on a calibrated quartz helix.  From 5 to 200 dry individuals
were weighed at a time to get an adequate deflection of the helix.  To
avoid bias in unconsciously selecting only the larger individuals,
several drops of plankton sample were placed on a watch glass and all
individuals of a species were removed for drying.  The total indivi-
duals weighed of any species were taken from several samples in case a
species might average larger in some samples than in others.  Due to the
small size and relative scarcity of planktonic rotifers in these lakes,
only the biomass of the most abundant species, Keratella americana,
could be determined.  For conversion of population estimates to biomass,
all other rotifers were considered to be the same size as K. americana.
The obvious error involved is quite small in terms of total zooplankton
biomass since K. americana, when abundant, comprised only ca. 7 percent
of the total biomass.

       After the mean dry weight/individual of a species had been
determined, the species biomass for any sample date was calculated by
multiplying individuals/m2 by rag/individual.  The species biomasses
were summed to obtain total zooplankton biomass.
       Secondary production.  An estimate of yearly production was made
for each zooplankton species except that rotifers other than K. americana.
were lumped.  In order to put limits on secondary production, three
calculations were made:   a minimum estimate, a maximum estimate, and a
"best estimate".  The minimum estimate for a species was obtained by
summing the net positive change in population size over a year, then
multiplying by the average mass per individual of the species.  The
other two estimates were unrelated to the minimum estimate, except that
all were based on the same population data and biomass data.
       The classical sigmoid growth curve for a population is described
                    dN _ rN(K-E
                    dt       K
by the equation     
-------
       N  =  No.  of  individuals

       t  =  time

       r  =  instantaneous rate  of  increase

       K  =  carrying capacity of the environment.

 If N  is very small relative to K  the expression K-N simplifies  to 1.0
                                                 K
 and the resulting equation, dN =  rN describes the logarithmic phase of
                            dt
 the growth  curve.  In this study  N was considered to be very much smaller
 than  K for  several reasons :

       1.   The relative density  of zooplankton in oligotrophic  lakes
            such as those studied is very low.  For example, maximum
            cladoceran densities  in Anderson-Cue and McCloud are in the
            order of 50-100/liter.  Ward (1940) reported up to  2,000
            cladocera/1 in a small pond and Borecky (1956) reported
            3,500 cladocera/1  in  Pymatuning Reservoir.

       2.   Density dependent  effects on laboratory populations of
            Daphnia were not seen by Frank, e_t a_l (1957) until densi-
            ties of 1,000-2,000/1 were reached.

       3.   A plot of r vs. N  for data from Anderson-Cue shows no ten-
            dency for r to decline as N gets larger.  If N were approach-
            ing K, r should be approaching zero.

       The  equation dN = rN can be written in its integrated form:
                    dt
        T" t"
Nj- * Noe  , and taking natural logarithms:   In Nt = rt + In No,
which when  In Nt is plotted against t, gives a straight line with r as
 the s lope .

       In this study logarithms of population were plotted against time
and each slope was considered  to  be an estimate of r over that time
period.  For any species the maximum positive slope observed was con-
sidered to approximate rm, the intrinsic rate of increase for that
species.   To determine maximum production,  each species was considered
to be reproducing at rm at all times during the year so that observed
differences between r and rm were considered to be due entirely to a
variable death rate.   The maximum productivity at any time was calculated
from the equation:
       Pm=Nt rm B'
       n  = maximum productivity
       Km
       N  = population size at time t

-------
       r  = maximum observed instantaneous rate of increase
        m

       B  = average biomass per individual .

Values of Pm were integrated over a year's time to get maximum yearly
production.

       The "best estimate" of production was determined similarly to
maximum production except that all positive r values were used to de-
termine productivity:   p = N  r  B, where:

       N  and B are the same as above

       r  = the observed rate of increase at time t

       p  = productivity

Productivity was integrated over time to get a "best estimate" for
yearly production.  Finally production values for all species were
summed to get total zooplankton production.
       Fredaticm.  Labidesthes sicculus was considered to be the chief
zooplankton predator in these lakes.  The population size of Ij. sicculus
was estimated using the Peterson mark-recapture method as described by
Ricker (1958).  Fish were captured individually at night with dipnets
and each fish was immediately marked by clipping a pectoral fin and
released.  Since L. sicculus will lie at the surface in shallow water
on a dark night, the marking procedure was fairly simple.  The brief
period of handling ensured minimum damage to the fish.  Recaptures were
made after one week.

       At numerous times during the year, samples of _L. sicculus were
collected with a dipnet or seine and preserved in 10 percent formalin
for later analysis.  To determine food habits, each fish was measured
and its gut contents analyzed.  The entire gut was removed, carefully
pulled apart, and washed with a few drops of water in a Sedgewick-Rafter
counting chamber.  All recognizable organisms in the gut were counted
under 21X magnification with frequent use of higher magnification to
check identification.  The counts of zooplankton in the gut were con-
verted to total mass of zooplankton eaten by multiplying the number of
each species by the mean individual biomass for the species.
       Statistics.  Statistical tests were used to evaluate apparent
trends in the results.  Since sampling frequency was arbitrarily chosen,
samples were considered to be random with respect to population sizes
or biomass.  Total biomass values were assumed to have an approximately
normal distribution.  In determining correlation coefficients between
species, only samples in which both species occurred were used, as in-
                                 -48-

-------
elusion of zero values would constitute a significant departure from a
normal distribution and thus invalidate the test statistic "r".  All
statistical tests used are described by Mendenhall (1967).  A signifi-
cance level of 0.05 was used throughout.

       In a related, but independent study, the interactions of littoral
zooplankton and their fish predators were studied.  The analysis scheme
used in this study is described below.

       Zooplankton collections were made in each lake every two weeks.
Samples were taken in the inner zone (water depth = 0.5-1.0m) with a
Van Dorn water sampler (volume = 1.93 1) after vigorous stirring to
dislodge organisms from the vegetation and insure random dispersal in
the water column.  These were concentrated to 32 ml with a plankton net
and preserved in one percent formalin.  Three one ml aliquots were
counted.  All cladocera were identified to species.  Rotifers were iden-
tified to genera.  Copepoda and their nauplii were counted, but not
identified.  Counts were converted to organisms per m2 by multiplying
the organisms per m  by depth (m) at the sampling site.

       Etheostoma fus 1forme and Heterandria formosa were collected every
two weeks with a dipnet in the same area as the plankton sample.  Five
fish of each species were taken from each lake and preserved in 10 per-
cent formalin.  Due to the small size of the fishes and the lack of a
definitive stomach in H. formosa, the entire gut contents were removed
and counted.  Although total length of each fish examined was recorded,
all sizes of both species of fishes fed on the same organisms.

       The contribution of each species or group of zooplankters to the
diet was expressed as its percentage of the total organisms eaten.  Due
to the large number of species and the difficulty in measuring the bio-
mass of individual zooplankters, the contribution of species or groups
in terms of proportion of total biomass was not determined.  Bi-weekly
fish gut analyses were pooled and reported as monthly averages.

       During the second year of the study, fish densities were estimated
by pulling a wire mesh dredge through specified areas in the littoral
at various times during the year.
       Gut clearance rate.  Laboratory experiments were conducted to
determine gut clearance time at 15°C, 20°C, and 30°C for both fishes.
Following two weeks of acclimation fish were placed in a feeding cham-
ber containing zooplankton.  After 30 minutes they were returned to
their original aquaria.  At intervals of 1 or 1/2 hour, five fish were
removed and their gut contents were examined, until the microorganisms
were in the lower portion of the gut and were no longer identifiable.

-------
       Algal succession.  The variety and succession of algal forms have
been followed both along the marginal shallow bottom and among the lit-
toral vegetation of both lakes.  Comparisons have been made with the kinds
and numbers in the open surface water and with those close to the water-
sediment interface at maximum depth.  This early work which began in 1965
was reported previously (Lackey and Lackey, 1967) and serves as a base-
line for future changes in the microbiota as nutrient enrichment con-
tinues.

       Both lakes are similar in many respects.  For example Synura
uvella occurs in both lakes and typically is more abundant in the deep
open water areas.  Many species of colorless Euglenophyceae are found
in the lakes and each body of water supports a bloom of Dinobryon at
least once during the late winter.  Both support a varied dinoflagellate
flora.  Generally the low population of photosynthetic forms indicates
limited algal growth substances and the low numbers of zooflagellata and
ciliates as well as the paucity of open water euglenids indicates the
low organic content of the lake waters.

       As the experimental lake continues to eutrophy the greatest
fluctuation in species will be those which occur in very small numbers.
A question to be answered is whether the crop of Dinobryon, Synura,
Peridinium umbonatum and Stentor amethystinus will increase considerably
as enrichment proceeds or whether species now encountered infrequently
will increase and supplant the present common organisms.

       Rhode as discussed in Hutchinson (1967 ) has reported the
disappearance of Dinobryon with increasing phosphorus levels.  It will
be interesting to note the effect of rising phosphorus levels in
Anderson-Cue Lake on the indigenous Dj.nobryon. population.  Blooms of
these chrysophytes occur annually between December and February.  Data
for plankton analysis are presented in the Appendix.  A summary of this
phase of the research showing the number of species and frequency of
occurrence is outlined in Table VI-1.  Generally the total species of
each lake is similar.  This is so because nutrients have been added so
that biotic changes can occur gradually and it is still early to note
marked differences.  Some species variation obviously does exist be-
tween the two lakes, but presently we can not determine whether this is
a reflection of nutrient enrichment.  The plankton data as presented
should be considered as representative for the soft, acid, sand bottom
lakes found in north central Florida.

       Appendix C reflects the percent occurrence of species during
1967 and 1968.  The data are qualitative since population and biomass
estimates were infrequently carried out.  The extensive list of species
includes littoral as well as open water forms and therefore forms not
usually associated with the open water plankton are included.  A great
many inshore forms are part of the periphyton.  This community although
an important part of the lake biota, often is not considered in limno-
logical investigations.  However, the periphyton can be the most pro-
ductive plant community in a lake system as was observed by Wetzel
(1962) in the study of Borax lake in California.
                                -50-

-------
                       TABLE VI-1

GROUPS OF MICROSCOPIC ALGAE AND PROTOZOA IN DETAILED ANALYSES
 ON FIVE DATES IN 1967-68 IN McCLOUD AND ANDERSON-CUE  LAKES,
               AND THH NUMBER OF OCCURRENCES.
Organism Group
Sulfur Bacteria
Blue Green Algae
Green Algae
Volvocales
Euglenophyceae
Cryptophyceae
Chrysophyceae
Chloromonadida
Dinoflagellata
Bacillarieae (Diatoms)
Rhizopoda
Zooflagellata
Ciliata
Torals
Total
Soecies
2
26
58
7
39
5
23
4
17
7
31
22
59
300
McCloud Anderson-Cue
No.
Species
1
24
48
7
23
5
23
2
17
5
21
18
42
236
No.
Occur. 	
3
23
117
13
40
18
33
6
36
8
40
23
82
472
No.
Species 	
2
21
39
2
29
5
18
4
14
4
24
15
41
218
No.
Occur.
4
33
101
5
50
14
24
8
31
6
33
19
82
410

-------
Effect of Nutrient Enrichment

       Anderson-Cue has been enriched with nutrients since March, 1967.
The 34 month regimen included loading with 300 gallons of secondary sew-
age effluent with supplemental nitrogen and phosphorus to achieve 2.48 kg
N and .212 kg P weekly.  As of August, 1970, approximately 387 kg N and
33.00 kg P have been added.  Lake response during this time has been
negligible for most of the routine chemical and biological parameters.
Table VI-2 shows yearly averages for some biogenic parameters in both
lakes for the period 1967-1969.  A mean for February and April 1970, is
also included.  As the data show, using the mean for all years, a sig-
nificant difference does occur for chlorophyll a_, but among no other
parameters.  However, averages for individual years relate increased
ammonia and primary production as well as plant pigments as response
parameters.

       Figure VI-1 and Table VI-3 A,B summarize phytoplankton production
levels for a three year period beginning in 1967.  The curves for the
two lakes with regard to their productivity—approximate one another
rather closely showing a similar summer—early fall maximum.  Annual
production estimates in Anderson-Cue Lake show essentially no differences
with Lake McCloud.  The level of fixed carbon at the end of 1969 was
58 grams/m  for Anderson-Cue compared to 60 grams for the control lake.
Bioassays using C1^ techniques as well as Provisional Algal Assay Pro-
cedures (PAAPl show phosphorus to be a principle limiting nutrient in
both lakes.

       Certain comparisons can be made with other nearby lakes in
Alachua County which demonstrate eutrophic conditions.  Productivity
determinations were made on these lakes in November, 1968, using an en-
closed box with a constant light source at an ambient temperature of 20°C.

       Hawthorne                    55.45*
       Newnan's                     53.55
       Bivens Arm                   77.54
       Orange                       43.01
       Wauberg                     124.30
       McCloud                      21.83
       Anderson-Cue                 11.52
                    3
       *mg C  fixed/m  -hour

       These  data  indicate that  both  lakes  in the experimental system
are  less  productive and  that Anderson-Cue will require much more nu-
trient enrichment  to  reach a  level comparable to other eutrophic lakes
in this area.   Interestingly  enough comparative productivity  of  Lake
Apopka and Lake Dora  in  Orange and Lake Counties which are recognized
hypereutrophic  lakes  in  central  Florida is  400 mg/c3-hr and  1000 mg C/nr*-
hr respectively.   Both of  these  lakes support continuous algal blooms
throughout  the  year.   In addition the lakes are virtually useless for
recreational  purposes.
                                  -52-

-------
                              TABLE VI-2
            COMPARISON OF AVERAGE CONCENTRATIONS OF SOME
         BIOGENIC PARAMETERS IN  EXPERIMENTAL LAKES 1968-1970
Anderson-Cue McCloud


Parameter Year

PH


Conda


TON


NH3-N


N03-N


PO^-P


Total P


PPb


Chi ac

1967
1968
1969*
1970
1967
1968
1969
1970*
1967
1968
1969
1970*
1967
1968
1969
1970*
1967
1968
1969
1970*
1967
1968
1969
1970*
1967
1968
1969*
1970
1967
1968
1969
1970*
1967
1968
1969*
1970
Annual <,
.30 '36
.23
.11
•39 1R
.20 l18
.02
.04
.08 0?
.10 '°7
.07
.005
:S5
.002
.011
.026 Q1,
.017 'QU
.020
92.43
157'44 134 83
154.61 134'83
—
2.04
?:S
2.15


4.75
5.03
4.89
5.42
_.
38.15
35.00
34.50
.45
.34
.32
.26
.07
.11
.06
.26
.02
.04
.02
.02
.004
.011
.003
.001
.010
.014
.017
.015
107.18
140.08
152.79
—
1.63
1.30
2.43
2.41



5.02


35.88


.34


.12


.02


.005


.014


133.35


1.94

 average of February and April
apnho cm'
b                        2
 primary production mgc/m -day
Cchlorophyll a mg/m
                                -53-

-------
                             Figure VI-1
E
i
>>
o
Q
O
9
£
o
O
O
o>
E
o
Q
K
iZ
o
9
500

400

300

200

100

 0

   f

700

600

500

4OO

300

200

100

 0
           1968
                                          Anderson-Cue
                                          McCloud
           1969
                                                Anderson-Cu«
                                         ,'  ,	McCloud
               A
           I     2    34    56    7   8    9   10   II   12
                                Month

         ANNUAL VARIATION  OF PRIMARY  PRODUCTIVITY  IN

                ANDERSON-CUE  AND WcCLOUD LAKES

-------
                          TABLE  VI-3 A

        PRIMA1W PRODUCTIVITY VALUES  FOR ANDKRSON-CUE LAKE
 Dale
    nig C fixed/day
Surface     5'
                                     10'
                                                mg C  fixed/day
1/24/67
1/31/67
2/28/67
3/14/67
3/28/67
4/11/67
4/25/67
5/9/67
5/23/67
6/6/67
6/20/67
7/7/67
7/19/67
8/1/67
9/12/67
9/26/67
10/11/67
11/28/67
1/9/67
2/21/68
3/19/68
4/29/68
5/16/68
6/18/68
7/23/68
8/27/68
9/16/68
10/14/68
11/11/68
12/11/68
1/6/69
2/3/69
3/14/69
3/31/69*
4/28/69
5/26/69
6/23/69''=
7/23/69
8/18/69
9/15/69
10/21/69
11/18/69*
12/15/69
1/14/70
1967 Avt!ra£O
1968 Average
1969 Average.
7.18
2.30
19.70
20.80
7.47
34.80
47.30
9.53
28,90
31.60
68.40
96.00
127.50
33.80
40.10
31.90
22.70
17.00
17.40
33.10
26.70*'
35.91
105.63
153.34
113.16
128.52
160,90
151.08
16.29
15.64
40.40
108.65
20.39
28.20
32.19
64.40
274.31
157.32
161.73
64.10
12.17
8.43
1.12
14.74



9.21
9.39
9.23
6.24
14.60
25.60
32.40
14.00
27.50
27.20
47.70
57.50
98.10
66.70
14.40
15.04
51.30
21.70
18.40
31 . 80
15 . 52*
34.45'
59.60
72.01
107.36
41.76
140.90
83.88
13.61
18.19
73.30
113.51
22 . 32
34.43
14.76
74.95
226.97
43.06
27.00
40.38
7.21
11.91
1.22
14.95



12.30
9.70
6.91
13.24
5.90
21.80
22.00
7.33
8.62
23.70
10.30
42.80
70.30
33.50
6.80
35.60
43.50
15.80
6.93
48.50
4.47*
1.40
9 . 32
6. /b
5.66
9.70
8.62
11 ! 76 •
4.18
16.25
17.06
37.68
9.06
9.64
1.33
1.39
41.16
20.98
8.90
7.04
2.93
11.91
0.92
5.50



	 ^ f^y-* *• *"
29 . 84
26.44
32.56
27.28
38.40
81.40
100.80
37.60
75.60
84 . 00
138.80
185.20
303.20
183.60
53.60
62.60
141.60
61.20
50.20
101.00
48.20
90.00
181.60
2e0.80
294.00
152.00
397.00
252 .50
39.00
53.00
181.40
308.40
60.30
93.10
46.20
191.20
635.00
182.50
137.50
116.00
21.68
33.00
3.45
37.32
92.43
157.44
154.61
jj ciii j^x \-y p / t.
402
382
•J\Jt*
375
203
ft\J J
540
652
703
400
461
483
456
607
504
225
560
573
•f* / *J
461
294
42'»
t,'J'i
K "> /
*-*• : "•! *
6f:0
56t
560
580
380
476
110
315
300
215
392
490
573
358
556
522
606
266
221
340
361
212



*Valucs were obtained from diurnal data,  calculated by
 incubation periods at each depth and multiplying by hours of
 daylight.                                 '
                             -55-

-------
                         TABLK VI-3 B



        PMIJAUY rr.OnUCl'IVlTY VALUES 1-OK McCLOUD  LAKE
                 mg C fixed/day—ra^              mg C fixed/day
Pate	Surface;	V	10_V	Rci?L
2/7/67
2/3.4/67
2/20/67
3/14/67
3/28/67
4/13/67
4/25/67
5/9/67
5/23/67
6/6/67
6/20/67
7/7/67
8/1/67
8/29/67
9/12/67
9/26/67
1/9/68
2/22/68
3/19/68
4/29/68
6/18/68
7/?Vf.8
8/27/68
9/16/68
10/14/68
11/11/68
12/11/68
1/6/69
2/3/69
3/14/69
3/ 31/69 *
4/28/69
5/7.6/69
6/23/69
7/23/69
8/18/69
9/15/6?
10/21/69
11/18/69*
12/15/69
1/14/70
1967 Average
1968 Average
1961) Average
8.28
3.27
13.80
5.21
24.92
24.90
2.00
10.50
30.86
59.50
59.56
32.69
68.03
6.49
53.78
109 . 59
7.70
2.10
3.14'=
23.14
99.12
76,73
282.26
142.10
57.48
23.48
10.99
5.63
15.66
5.87
4.17
26.47
47.47
139.26
231.29
198.21
9.76
20.06
6.53
2.35
6.44



15.50
3.53
13.90
16.34
13.64
30.16
13.26
17.34
31.33
43.72
96.23
49.90
45.48
14.43
29.33
150.57
18.44
5.60
3.99*
22.21
61.56
54.92
138.48
113.12
47,16
24.98
11.04
6.58
12.20
4.98
5.28
38.70
43.58
117.69
251.85
144.61
9.14
17.58
3.37
1.43
7.58



7.14
3.77
7.07
13,17
36.62
23.33
6.37
23.58
22.44
67.20
169.06
37.00
19.07
7.55
13.21
37.41
15.10
4.81
3.35*
15.30
28.38
32.98
64.78
69.35
30.00
12.01
8.18
5.55
8.53
5.44
4.82
29.92
46.50
75.35
277.10
86.53
4.20
9.35
3.37
3.47
7.16



41.10
10.72
39.36
39.20
55.00
87.30
33.28
52.80
92.00
151.80
312.00
139.00
137.00
37.64
94.60
392.00
50.40
15.64
11.54
64.20
192.00
J68.80
460.00
337.00
139.00
70.30
32.00
18.60
37.50
15.40
15.38
98.00
136.40
353.00
778.00
440.50
25.76
48.44
12.56
6.70
21.64
107.3.8
140.08
152.79
                                                                    278
                                                                    452

                                                                    375
                                                                    203
                                                                    540
                                                                    652
                                                                    703
                                                                    400
                                                                    461
                                                                    483
                                                                    456
                                                                    504
                                                                    575
                                                                    225
                                                                    560
                                                                    294
                                                                    320
                                                                    499
                                                                    574
                                                                    560
                                                                    560
                                                                    580
                                                                    380
                                                                    476
                                                                    110
                                                                    335
                                                                    300
                                                                    215
                                                                    392
                                                                    490
                                                                    573
                                                                    358
                                                                    556
                                                                    522
                                                                    606
                                                                    266
                                                                    221
                                                                    340
                                                                    361
                                                                    212
*Valnes were  obtained  from  diurnal  data,  calculated by averaging
 incubation periods at. each depth and multiplying by hours of day-
                              -56-

-------
       The comparative production estimates of lakes provide information
regarding the trophic status which in turn can be useful in establishing
recreational and use potential of surface water.  This technique is es-
pecially worthwhile in multi-lake studies where surface waters are being
examined on a regional basis.  In this kind of investigation comparative
productivity estimates can quickly determine lakes where algal growth is
high and where ecosystem management procedures could be used effectively.

       Diurnal variations in integral photosynthesis were recorded
monthly during 1968.  These data are presented in Figure VI-2.  (Hori-
zontal lines indicate rain.)  Rather obviously marked variations in
integral photosynthesis occurred during 1968 in both lakes.  Maximum
productivity in Anderson-Cue Lake was noted in September when fixation
of carbon amounted to 65 mg/m^-hr.  Highest photosynthesis between the
two lakes was observed in McCloud.  Fixation rates here reached 80 mg
C/m^-hr.at the end of August.

       Variations in the vertical distribution of chlorophyll a_ between
McCloud and Anderson-Cue Lakes are presented in Table VI-4 and Figure
VI-3.  The data are inclusive beginning in January of 1967.  Generally
the mean value for chlorophyll a at the surface, 5 and 10 foot depths in
Anderson-Cue Lake (1.99, 2.00, 2.12 mg/nr) were comparable to McCloud
Lake (1.53, 1.89, 1.46 mg/m3) during the first year.   The standing crop
of phytoplankton increased significantly during 1968 in Anderson-Cue
water, but remained virtually the same as before in McCloud.  Average
values of 4.12, 4.12, and 4.41 mg chl a per m3 were noted in Anderson-
Cue Lake at the three depths while McCloud Lake supported an average
of 2.14, 2.02, and 2.24 mg chl a per m3.

       Considering seasonal effects and based on integral values
chlorophyll a during 1967 gradually increased from a range of 1-5 mg/m
during the first four months to levels greater than 15 mg/m^ in Anderson-
Cue during mid-summer and fall, (Figure VI-3).  Both lakes had approxi-
mately the same phytoplankton biomass based on pigment analyses except
for peaks during July and October in Anderson-Cue Lake.  At these times
chlorophyll in the experimental lake reached a maximum in excess of 15
mg chl a_ per m^.

       The 1968 data show the divergent characteristics of the two
lakes with regard to pigment levels.  Chlorophyll a_ was 5 mg/m^ or
higher at all times during the year.  Three peaks were evident.  The
highest (>25 mg/m^) occurred in February during a winter flowering of
Dinobryon.  The others as indicated in Figure VI-3 were observed in May
and late July.  The spring bloom coincided with growths of Synura.

       It is worthwhile noting the significant algal growth response to
added nutrients as evidenced in the elevated chlorophyll values.  Most
frequently chlorophyll a was greater than 10 mg/m^.  These data support
productivity data in that these two parameters show the response to nu-
trient influx far more rapidly than changes in species diversity.
                                 -57-

-------
 I  15


 .*  10
 u
 o>
               1/23
            244 Lang leys
                       _L
_L
          0800  1000  1200  1400   1600
                        2/27
                 15


                 10


                  5


                  0
                   449 Lang leys
                       0900   1100   1300  1500  1700
CM
 e   4
 U

     ,&


     0
               3/19
                                              15 -
             499 Langleys
                 10

                  5

             •^   o
            _L
 _L
_L
           0900 1100   1300   1500   1700  1900

                        Time EST
                        4/29
                        1000  1200  1400  1600  1800
                                  Time EST
      EST   Eastern Standard Time
      EOT   Eastern Daylight Time
           ———  Anderson-Cue

           	McCloud
                                       - Clear

                                       - Overcast
                Figure VI-2.  Diurnal Variations in Primary Production During 1968.
                                              -58.

-------
f
 L 30
TJ
 V
£ 20
u
 O>
 E 10
            5/16
    0 -    600 Lang leys
           I	I
                              I	I
         0900  1100  1300  1500  1700
                                                     6/18
                                                    1000  1200  1400  1600  1800
    40
's
 I  30
£  20
u
            7/23
           560 Lang leys
                                            80


                                            60


                                            40


                                            20
         0900  1100  1300   1500   1700   1900

                      Time EOT
                                                     8/27
                                                   0900  1100  1300 1500  1700  1900

                                                               Time EOT
                                    Figure VI-2. (Continued)
                                     -59-

-------
   60
   50
U
 en
   30
   20
   10
 9/16
                                   40


                                   30



                                   20


                                   10



                                    0
         0900  1100  1300  1500  1700  1900

                   TlmeEDT
10/14
                                                     476 Langleys
                                         0900  1100  1300   1500   1700  190


                                                     Time EOT
 u
 o>
 11/11
              110 Langleys
                                   I
0800  1000  1200  1400   1600

           Time EST
                                                    12/11
                                            315 Langleys
                                                  0900  1100  1300  1500  1700

                                                             Time EST
                                  Figure VI-2. (Continued)
                                   .60.

-------
Date
                          TABLE VI-4

                 CHLOROPHYLL  a LEVELS
              IN ANDERSON-CUE AND McCLOUD LAKES
    Anderson-Cue
Surf     5'
10'
      McCloud
Surf     5'       10'
1/17/67
1/24/67
1/31/67
2/7/67
2/14/67
2/28/67
3/14/67
3/28/67
4/11/67
ft/25/67
5/9/67
5/23/67
6/6/67
6/20/67
7/7/67
7/19/67
8/1/67
8/29/67
9/12/67
9/26/67
10/11/67
1/9/68
2/21/68
3/19/68*
4/29/68*.
5/16/68*
6/18/68*
7/23/68*
8/27/68*
9/16/68*
10/14/68*
11/11/68*
12/11/68*
1/6/69
2/3/69
3/14/69
3/31/69*
4/28/69
5/26/69
6/23/69*
7/23/69
8/19/69
9/15/69
10/21/69
11/18/69*
12/15/69
1967 Average
1968 Average
1969 Average
1.09
1.08
1.09
0.54

0.83
0.25
1.38
2.44
1.77
2.91
1.92
1.76
2.03
2.86
4.20
1.70
3.15
3.02
4.22
1.64
2.57
7.74
3.00
4.62
5.07
3.57
5.50
4.60
4.18
4.56
2.07
1.92
6.74
3.82
1.80
1.10
4.44
4.36
4.45
4.66
5.17
3.75
1.84
1.88
1.88



0.54
0.25
1.38
1.12
	
1.12
1.38
1.38
1.10
1.67
2.57
1.83
1.87
1.45
1.02
5.73
1.45
4.22
2.80
4.45
2.77
0.77
8.77
3.75
3.27
5.06
3.24
7.13
4.66
4.05
4.95
1.90
1.92
7.74
3.82
	
1.40
3.71
4.39
5.18
4.24
5.56
3.75
1.65
2.02
2.25
2.04
5.17
7.35
0.83
0.25
1.12
1.12
_—
1.08
1.60
1.37
1.34
1.57
2.55
1.67
1.97
1.89
2.03
5.64
1.36
2.57
2.80
7.14
2.55
4.47
8.41
5.42
2.56
4.06
3.70
6.97
3.73
4.39
5.50
1,87
1.85
6.08
4.02
	
1.29
3.35
5.45
5.41
9.19
6.67
4.70
1.65
2.15
3.28


2
- rn
	
0.54
0.54
0.67
0.80
1.60
	
0.66
1.55
2.43
2.48
1.97
1.78
1.47
2.38
0.33
1.48
3.15
2.20
	
1.23
0.77
0.68
0.91
__—
1.77
2.18
5.66
2.62
2.47
3.30
1.98
0.72
1.18
0.60
0.51
2.10
1.86
4.37
6.24
4.71
1.88
2.11
1.54
0.96



	

0.25
0.54
0.83
1.84
1.34
	
0.99
0.99
1.90'
3.18
1.96
1.98
2.03
2.70
2.57
3.39
3.41
2.20
_—
0.10
1.12
0.73
0.84
____.
1.68
2.17
5.30
2.33
2.82
3.02
2.07
0.74
0.94
	 	
0.48
2.02
1.73
4.01
6.24
4.32
1.44
2.25
1.33
1.03
1.63
1.30
2.43

_— .
0.25
0.83
0.80
1.08
0.54
	 .
0.57
1112
1.90
2.56
1.86
2.32
0.68
1.49
1.59
2.48
2.81
1.99
	 _
__
1.12
0.86
1.01
_— ...
1.93
2.34
4.81
2.60
2.69
2.85
2.14
0.97
0.75
	 _
0.40
2.03
2.19
4.73
6.56
5.02
1.77
2.10
1.22
0.90



4.VW? .'VVtSJLelJJ«          : r " -"*                             — " "-*
*Diurnals - Chlorophyll a values are an average of  approxiaately 4  samples
 per day
                               -61.

-------
                             Figure  VI-3
N_
25


20


15


10


 5


 0
          1967
                                               Anderson-Cue
                                          	McCloud
          1968
    25


    20


    15


    10


     5


     0
                                           Anderson-Cue

                                           McCloud
 «x
 9
20


15


10
          1969
Anderson-Cue

McCloud
                                 J	L
     "     I    2    3   4    5    6    7    8    9    10   II    12

                                Month

  ANNUAL VARIATION OF  CHLOROPHYLL  a.  IN  ANDERSON-CUE

                     AND McCLOUD LAKES
                              -6?..

-------
       McCloud  Lake during  the  last half of  1968 had an  increased
 phytoplankton biomass reaching  a peak  in late August.  During the last
 two  years  the control lake  always supported  a maximum population of
 phytoplankton during August and September.

       Horizontal chlorophyll a distribution in Anderson-Cue Lake was
 determined  four times in  1968 from February  to the end of May.  These
 measurements were made only in  surface water.  No correlative production
 estimates were  completed.   As indicated in Figure VI-4 phytoplankton
 was  unevenly distributed  in surface water on all sampling periods.
 Chlorophyll a levels were always highest in  the near-shore environment
 ranging from 6  to greater than  15 mg/m .  On three occasions these
 occurred in the same extreme southern part of the lake.  Causes for the
 plankton patchiness are likely  a combination of wind effects and local
 areas of enrichment especially  in the shallow water where animals visit
 for  drinking.   Two experiments  for coliform  distribution confirmed
 animal movement in the shallow  water along the lake margin.

       Limiting nutrients for phytoplankton  growth were determined
 essentially using Goldman's technique  (1964,  1965) for bioassay.  Many
 results using this method are varied and interpretation is difficult
 to make.  Generally three qualitative judgements can be made.  A parti-
 cular substance may be limiting to plant growth, result in no change
 or elicit an inhibitory response.  The results of these bioassays are
 shown in Table  VI-5.  Considering Anderson-Cue Lake, phosphorus limited
 primary production 92 percent of the time during 1967 trials.  The
 phytoplankton did not respond to nitrogen additions indicating that this
 substance was not a limiting growth factor,

       Verification of the  nitrogen experiments has been carried out
 using a laboratory technique proposed by Fitzgerald (1968).  The pro-
 cedure essentially employs an exposure of ammonia nitrogen to a popula-
 tion of organisms and ammonia uptake is followed over a set time period.
 The results from this procedure (Table VI-6) showed that the natural
 plankton from Anderson-Cue lake were not nitrogen deficient.  These
 data reflect the rising ammonia levels in the experimental lake as nu-
 trient enrichment progresses (see Section V).

       The  reasons why phosphorus concentrations in the water do not
 reflect those added to the lake are obscure.  Certainly the supply is
 insufficient for the optimum growth of much of the phytoplankton.
 Phosphorus  loss may be to the sediments although we have been unable to
 determine this as yet.

       Substances other than N and P appear to be limiting during
various times of the year.  However, the  difficulties encountered with
 this particular method make refined judgements questionable.  A reliable
 standardized algal growth potential procedure is badly needed at this
 time for all those doing eutrophication research.
                               -63-

-------
2/15/68
HH! >lOmg/M3
|    1 5-10ma/M3
3/1V68
     >10.6mg/M3
     9.5-10.5mg/M3
|   |  8.1-9.
     7.0-8.
 4/9/68
      >l5mg/M3
      10.1-15mg/M3
 [   I  4.1-10mg/M3
      0-4ma/M3
      4.5-6.2mg/M3
      3-4.5mg/M3
 Figure VI-4.  Horizontal  Distribution of  Chlorophyll a
                Ande r s on-Cue Lake.

-------
            TABLE VI-5
ALGAL GROWTH RESPONSE TO NUTRIENTS
      Anderson-Cue Lake 1967
1/4 1/24 3/14 3/28 4/11 5/23 6/6
N---XX--
P + + + - + + +
N,P + +
Fe X + X X X
Si X + X
S - X
Vitamins - - -
Trace
Metals - XX X
EDTA X - X
McCloud Lake 1967
2/14 3/14 3/28 4/11 6/6 6/20 8/1
N -X 	
P +XXX-- +
N,P + X +
Fe . + _ + ___
Si + + - + - X
S - + + + --X
Vitamins - X
Trace
Metals + + - - + - X
EDTA + - + + X X
+ = Limiting X • Inhibitory
6/20 8/1 8/29 9/12
X X
XXX
X
8/29 9/12
+
X X
+
+ +
+ +
+ +
+
- «= No Change
             -65-

-------
                               TABLE VI-6

              AMMONIA UPTAKE BY ALGAE IN ANDERSON-CUE WATER

                                   ,               Filamentous  _
Time  (min.)               Chlorella                Chlorophyceae
                                 *                          *
      0                      1.24                       0.83

    +15                      1.30                       0.81

    +30                      1.35                       0.84

    +45                      1.36                       0.92

    +60                      1.38                       0.89

    +90                      	                       0.97


 laboratory stock culture
2
 filamentous green algae
 probably Mougeotia collected in Anderson-Cue Lake.

 NH-N mg/1
       Figure VI-5 shows the results of phosphorus addition to natural
phytoplankton in Anderson-Cue water over a period of 120 hours.  The
experiment was carried out to determine lag time in phosphorus uptake
by phytoplankton.  It may be seen that maximum stimulation of algae by
phosphorus as detected by labeled C-14 carbonate fixation occurs in 72
hours.

       This particular time dependent bioassay method is very useful as
a tool which can be used to follow the enrichment of single lakes by
observing the change in slope of the response line at successive time
periods.  In addition comparative relationships among lakes within a
region can be made.  As can be seen from Figure VI-5 the greatest de-
mand for phosphorus occurs during the summer in Anderson-Cue Lake and
decreases considerably during the fall when biological activity is
lower.  Uptake of P in McCloud Lake is lower than Anderson-Cue and proba-
bly is due to a lower biomass of plankton organisms.  Maximum uptake by
organisms in the control lake occurs in 48 hours.

       Goldman (1965) has noted photosynthetic stimulation from the
addition of various compounds and elements in trace quantities.  Lakes
showing micronutrient deficiencies frequently can be found and such
substances as vitamins and various cations doubtless play an important
role in algal growth cycles.  Sources of these materials are provided
to lakes by (1) tributary streams, (2) runoff from land, or (3) inter-
change from sediments.  Lakes near urban centers could very well re-
ceive some micronutrients from rain falling through polluted air.
                                -66-

-------
        50
	 Anderson-Cue
. __ McCloud
 x
I
        40
        30
        20
        10
                                                         Oct. '68
                                                                         Oct.  '68
                      24
                48
  72
Hogrs
96
120
              Figure VI-5 Carbon - 14 Fixation by Phytoplankton Stimulated with
                          lOOjugm Phosphorus.
                                      -67-

-------
       Bioassay data on the two lakes show that cations limited algal
photosynthesis in trials carried out on McCloud Lake.  No similar pattern
was observed in Anderson-Cue Lake during the same period.  Two times
during the study vitamin additions stimulated algal growth in McCloud
Lake with no correlative response in Anderson-Cue water.  It seems un-
likely that differences between the two lakes could be attributable to
variation in nutrient loading since enriched sewage effluent was not
added to Anderson-Cue Lake until March, 1967.  Therefore, at the present
time differences in algal growth response to micronutrients appear to be
inherent between the two lakes.
Characteristics of Littoral Zone

       Littoral plant growth as expressed in dry weight has remained
constant in both lakes.  Since this area experienced an extended drought
during the winter and early spring months of the year 1967-1968, both
lakes dropped considerably, exposing former submerged stations.  Those
stations out of water were not cut in December.  A second problem en-
countered was that of migrant cows feeding in these areas.  Adequate
fencing now prevents this from occurring.  Table VI-7 contains the re-
sults of this part of the project.  As might be expected plant growth
was greatest during the spring and summer months and least in the fall
and winter.  Average dry weight per square meter of littoral growth in
Anderson-Cue Lake amounted to 50 grams which represented an increase
from 1967 of 20 grams/m^.  Nutrients added to the lake expectedly would
enhance the growth of marginal vegetation, but it is too soon to say
that the 1968 data reflect the influence of induced eutrophication.

       The yegetation of the two lakes was studied to determine the kinds
of plants, and the density and growth features of these plants in the
littoral zones.  McCloud Lake was studied to find contrasts, if any, be-
tween it, without pollution, and Anderson-Cue Lake with well recorded
pollution.

       As mentioned earlier, both of the lakes decreased in depth and
size in 1967-1968, and the receding waters left a wide zone where formerly
water flooded the shore.  In both cases, the littoral was generally di-
vided into three zones as far as the kinds of plants and vegetation were
concerned.  These were:   (1) The Upper Zone with some shrubs present
or evidence that they were formerly present; (2) The Middle Zone where
grasses, sedges, and other herbs are abundant, but very few of the true
aquatic plants occur; and (3) The Lower Zone where distinctly hydrophy-
tic herbs and some aquatic plants occur.  This Lower Zone has, due to
the receding water, a number of the plants usually common in open water,
such as the water lilies and the maidencane grass.

       These three zones around McCloud Lake were more distinct and had
a denser population of plants than around Anderson-Cue Lake.  Except for
the Lower Zone near the water the same species of abundant plants were
                                -68-

-------
            Table VI-7




STANDING CROP ESTIMATES OF ANDERSON-CUE AND McClOUD LAKES
Plant
Station Dry \Vt. (g)
March • May, J%7
Amlenon-Cuc
1 17.5749
3 12.2894
S 18.9235
6 44.36.53
7 5.2773
McCIoud
1 H.
-------
present in both littoral belts.  The notable difference was an obvious
decrease in density and the number of kinds of plants present in the
Lower Zone along the waters edge of Anderson-Cue Lake, as compared to
McCloud Lake and to the density of plants in the Middle Zone of Anderson-
Cue Lake.  This part of the Lower Zone that shows injury and death of
some of the Lower Zone plants is especially notable because some of the
plant species were not injured by what is presumed to be the toxic ef-
fects of the pollution and there are a few, such as Ludwegia alternifolj£'
which were only abundant in this "polluted zone".

       The following is a brief and general list of the plants in each
of these three littoral zones, and also a list to show the plants of the
Lower Zone of Anderson-Cue Lake which died or became less abundant.
               GENERAL PLANT COMPOSITION OF LITTORAL ZONES
                     McCloud and Anderson-Cue Lakes
Upper Zone
       Hypericum fascilculatum
       CephaIanthus occidentalis
       Spartina bakerii
       Andropogon spp.
       Pluchea sp.
       Eupatorium capillifolium
       Juncus effusus
       and other upland herbs
Middle Zone
       Sabatia campanulata
       Lachnocaulon minus
       Pluchea foetida
       Fuirena breviseta
       Juncus sp.
       Manisuris tuberculosa
       Cyperus spp.
       Xyris pallescens
       Lachnocaulon sp.
       Panicum hemitomon, and many other grasses and herbs
       Most of these do not stand flooding
Lower Zone
       Webateria (proliferating sedge, very small)
       Sagittaria graminea
       Fuirena spp.
       Erioaulon compressum
       Utricularia subulata
       Utricularia cornuta
       Cyperus spp.
                                -70-

-------
 Lower Zone  Cont.

        Lachnocaulon minus
        Sabatia campanulata
        Lycopodium spp.
        Xyris  pallescens
        Panicum hemitomon
        Leersia sp.
        Ludwegia alternifolia
        Cephalanthus occidentalis  (small)

                CHANGES IN  "POLLUTED  PART" OF LOWER ZONE

 Plants  that seem  definitely killed  or reduced in size and number

 Sagittaria graminea                     Lycopodium spp.
 Lachnocaulon  minus                       Fuirena spp.
 Xyris pallescens                         and some of the Utricularia sp.
 Sabatia campanulata

 Plants  that survived, and in some areas seem to grow better in the
 "polluted part" of  this Lower Zone.

Websteria sp.                            Panicum hemitomon
Leersia sp.                              Ludwegia alternifolia
Utricularia sp.

Of these, the  Ludwegia seems to grow  taller and denser.

        This "polluted" part of the  Lower Zone is widest, about 5-7 feet
along the southwest  shore of Anderson-Cue Lake.  It is narrower along
 the northeast  part  of this  lake, being about 2-4 feet wide.  The dif-
 ferences in width of the probably polluted zone may be due to more
prevailing winds  that make higher water in the Southwest shore area.
This  shore vegetation needs to be studied more intensely to find causes
and extent of  this death of plants.   Similar surveys of the kinds of
plants  and vegetation zones should  be made 3-4 times a year.
.Provisional Algal Assay Procedure (PAAP)

       In the laboratory, preparation was made to adapt the provisional
algal assay procedure (PAAP), as proposed by the FWPCA's Joint Industry/
Government Task Force on Eutrophication.  All three of the recommended
assay organisms are growing on a lighted shaker table at 24°C.  Up to
this time, Selenastrum capricornutum has been used as the experimental
organism.  The other two organisms, Microcystis aeruginosa and Anabaena
.flos-aquae. will be similarly tested.
                                 -71-

-------
       Five methods of assaying algal growth are being evaluated.  These
include cell count, radiocarbon uptake, dry weight, light absorbance,
and adenosine triphosphate (ATP) level.  Of thase methods, absorbance
and radiocarbon uptake are best suited for quantifying Selenastrum
growth.  Dry weight and ATP determinations are not sensitive enough using
the inoculum size presecribed.  Water samples from an unproductive and
a productive lake have been subjected to the PAAP.  The results obtained
were as expected i.e., the test organism grew faster and reached higher
numbers in the productive lake water than in the unproductive lake water.
Further tests of this nature are now being conducted, using several dif-
ferent lakes as well as sewage treatment plant effluent.

       Laboratory work on the PAAP study is being done routinely.  To
date, all four of the methods of growth analysis outlined in the pro-
cedure have been tested with all three organisms.  We have found only
two of these methods, cell counting and absorbance measurements, to be
reliable as tools for assaying growth because this study involves work-
ing with cultures of low cell number.  However, another method, chlorophyll
a (Creitz and Richards, 1955), has been tried and found to be sensitive
enough to estimate growth at these low cell numbers.  The possible use
of the Coulter Counter instead of the S-R cell to increase the precision
of cell counting is planned.  Having established the assay methods, ex-
perimental work with lake water and lake sediment will be continued and
enhanced.

       Preliminary tests using sediment collected from Anderson-Cue Lake
and inoculated with the organism Selenastrum demonstrated a significant
difference in the growth rates of the organisms in flasks containing
PAAP media plus sediment leached water versus PAAP media only.  Increas-
ing the volume of sediment leached water added to each culture flask did
not increase the growth rate.
       Extractive phosphate and alkaline phosphatase.  Both parameters
were measured in the apical portions of the littoral vegetation as well
as attached filamentous chlorophyceae.  Table VI-8 presents some pre-
liminary data.  The data obtained thus far would indicate the growth to
be limited by the lack of available phosphorus.
Secondary and Tertiary Trophic Levels

       The research bearing on the determination of secondary and
tertiary trophic levels in both lakes has involved work with zooplankton
and the planktivorous brook silverside (Labidesthes sicculus).  This
fish is a significant element in the ecosystem of both Anderson-Cue and
McCloud Lakes feeding chiefly in pelagic areas on zooplankton and small
insects.  It is very selective in what it eats, the size of the prey
being definitely correlated with fish size.  Juvenile Labidesthes feed
                                -72-

-------
                                TABLE VI-8

               EXTRACTIVE PHOSPHATE AND ALKALINE PHOSPHATASE

                         Extractive Phosphate     Alkaline Phosphatase
 Plant                   nig P/100 mg dry wt.      Units/mg/hr	

 Mougeotia (outfall)             0.17                     300.

 Mougeotia (N.E. end)            0.01                      60.

 Ludwegia                        0.10                      16.

 Websteria                       0.03                      16.

 St.  Johns Wort                  0.04                       3.

 Utricularia                      0.02                       9.

 Mayaca                           0.01                       7.
 chiefly on rotifers  and  copepod nauplii while larger individuals  eat
 cladocera, adult copepods and occasional small insects.   (Table VI-9
 represents a partial spectrum of zooplankton ingested by Labidesthes
 based  on examination of  gut  contents).   Labidesthes  reproduce  in  spring
 and  early summer and few individuals  live longer than a  year,  so  the
 food requirements of the population will change  from summer  to winter as
 the  fish grow.   Thus the food availability  is related not so much to the
 total  biomass of zooplankton as to the  biomass of acceptable food items
 during a given  season.   The  delicate  balance involved will be expected
 to change as  the lake becomes more eutrophic.

        Zooplankton samples have been  collected bi-weekly with a tow net
 and  microscopic  counts have  been made according  to Standard Methods for
 _the  Examination  of Water and Wastewater (1965).  Individuals of each
 important species were picked out under a dissecting microscope, dried
 in a vacuum desiccator and weighed on a calibrated quartz helix (Table
 VI-10).   In order to get an  adequate  deflection  of the helix, groups
 of 5-200 individuals are weighed at a time and an average mass calculated,
 Average mass per individual  can be multiplied  by number per square meter
 of lake surface  to obtain the biomass of the  organism  in gm/m2.  Caloric
 Values  for species of zooplankters are  determined using a microbomb
 calorimeter.  Thus several years of bi-weekly  data on  zooplankton popu-
 lations  can be readily converted into biomass  or energy content.

        Data for  conversion of  zooplankton numbers to biomass can be
 found  in Appendix F.  Not all data are as yet available for  conversion
 of zooplankton numbers to biomass.  The accompanying Tables  VI-11 and
VI-12,  showing the results from August  1967 and 1968, are examples of
 the  type  of information which are available for each month of the  inves-
 tigation.  Additional population densities (No./m2)  for 1967  and 1968
are  presented in Vol. II, Table 2-9,  Appendix F.
                                -73-

-------
                               TABLE VI-9

                      FOOD OF IABIDESTHES SICCULUS

Labidesthes   Standard length 19 mm   Approximate dry weight 10 mg

       Gut Contents              Number     Weight
Bosmina
Tropocyclops
Larval Cope pods
Chydorus
Labidesthes Standard
Gut Contents
Bosmina
Small Cyclops
Larval Cope pods
Chydorus
Diaptomus
Diptera
19
7
11
4
length 29 mm
22.5
3.5
1.6
4.8
Total weight
(dry) of food
32.4 yg

Approximate dry weight 23 mg
Number Weight (ug)
30
113
27
2
1
1
35.5
56.8
3.9
2.4
3.6
45.0
Total weight
(dry) of food
147 U8



Labidesthes   Standard length 37 mm   Approximate dry weight 49 mg

       Gut Contents	Number	Weight (ug)
Bosmina
Small Cyclops
Larval Cope pods
Diaptomus
Daphnia
Diaphanosoma
14
30
3
14
8
1
16.5
15.1
0.4
50.4
31.4
1.4
                                                          Total weight
                                                          (dry) of food.
                                                          115 yg

-------
                         TABLE VI-10

           DETERMINATION OF ZOOPLANKTON WEIGHTS IN
                ANDERSON-CUE AND McCLOUD LAKES
                                                 Range          Avg.  Dry
                  No. of           Total         Dry Weight     Weight
Organism De
Keratella
Larval Copepods
Tropocyclops
Bosmina
Diaphanosoma
Diaptomus
Daphnia
Holopediura
Mesocyclops
Chaoborus
'terminations
4
2
5
5
6
10
7
2
10
5
Individuals
320
200
125
97
110
100
70
18
95
35
(micrograms)
0.036-0,053
0.145-0.148
0.371-0.608
0.73-1.38
0.872-1.82
3.11-4.53
1.49-5.77
4.5-6.6
4.81-10.8
14.7-84.6
(micrograms;
0.0435
0.146
0.503
1.182
1.423
3.603
3.922
5.549
8.47
45.6
Weights in this table were obtained by weighing groups of from 5-200
individuals of each species on a calibrated quartz helix.
                               -75-

-------
                         TABLE VI-11

        ZOOPLANKTON DENSITIES AND CALCULATED BIOMASS
                       FOR AUGUST 1967
Anderson-Cue Lake
                      3 Aug.
                      11 Aug.
             26 Aug.

Tropocyclops
Mesocyclops
Diaptomus
Larval Copepods
Bosmlna
Daphnia
Diaphanosoma
Holopedium
Keratella
No/m2
4,715
3,169
11,038
18,987
1,227
3,156
879
0
9,922
mg/m2
2.372
26.841
39.770
2.772
1.450
12.378
1.251
0.0
0.432
No/m2
2,160
1,452
5,179
27,836
5,706
3,876
0
0
9,065
mfc/m2
1.086
12.298
18.660
4.064
6.744
15.202
0.0
0.0
0.394
No/m2
19,742
13,271
22,484
74,147
18,050
761
0
0
26,699
mR/m2 __
9.930
112.405
81.010
10.825
21.335
2.984
0.0
0.0
1.161
Total
         87.266            58.448
Avg. Blomass (mg/m2) 128.45
                  239.650
McCloud Lake

Cyclops (sp. 1)
Cyclops (sp. 2)
Diaptomus
Larval Copepods
Bosmina
Daphnia
Diaph'anosoma
Holopedium
Keratella

Total
5,625
3,781
9,628
36,030
6,503
105
367
0
41,370
2.829
32.025
34.690
5.260
7.688
0.412
0.522
0.0
1.800
5,371
3,611
8,713
47,800
10,540
79
26
0
66,190
2.702
30.585
31.393
6.979
12.458
0.310
0.037
0.0
2.879
         85.226
                  Avg. Biomass (mg/m2) 86.28
87.343
                               -76-

-------
                         TABLE VI-12

         ZOOPLANKTON DENSITIES AND CALCULATED BIOMASS
                       FOR AUGUST 1968
Anderson-Cue Lake
                      12 Aug
                        19 Aug.
              26 Aug.

Cyclops (sp. 1)
Cyclops (sp. 2)
Diaptomus
Larval Copepods
Bosmina
Daphnia
Diaphanosoma
Holopedium
Keratella
Total
No/m2
24,200
7,991
144,514
192,000
1,142
228
0
0
58,901

mg/m2
12.173
67.684
520.684
28.032
M.350
0.894
0.0
0.0
2.562
''633.379
No/m2
34,496
16,666
92,690
90,635
685
228
228
0
42,464

mg/m2
17.351
141,161
333.962
13.233
0.810
0.894
0.324
0.0
1.847
509.583
No/m2
43,847
44,385
291,058
194,756
7,801
807
4,035
0
538

mg/m2
22.055
375.941
1048.682
28.434
9.221
3.165
5.742
0.0
0.023
1493.263
                  Avg. Biomass 878.74 mg/m2
McCloud Lake

Cyclops (sp. 1)
Cyclops (sp. 2)
Diaptomus
Larval Copepods
Bosmina
Daphnia
Diaphanosoma
Holopedium
Keratella

Total
36,528
11,187
25,570
86,982
27,168
1,142
12,785
228
2,511
18.374
94.754
92.129
12.699
32.113
4.479
18.193
1.265
0.109
23,743
19,405
34,702
101,137
24,656
0
20,775
0
1,826
11.943
164.360
125.031
14.766
29.143
0.0
29.563
0.0
0.079
27,976
28,514
91,998
80,162
112,980
0
35,625
0
0
14.072
241.514
331.469
11.704
135.542
0.0
50.694
0.0
0.0
          274.135
Avg. Biomass 478.00 mg/m2
374.885
784.995
                               -77-

-------
       Populations of many zooplankton species have shown tremendous
fluctuations over the time period of the study.  Some of these fluctua-
tions are seasonal in nature or weather-induced but others may be
correlated with changing physio-chemical parameters of the lake.  De-
finite trends caused by increased enrichment have not yet been established.

       The change in L. sicculus population in Anderson-Cue Lake has been
very spectacular.  The population has dropped from an estimated 130,000
in November, 1967, to virtually zero at present.  The McCloud population
has shown no corresponding decline.

       Ij. sicculus has previously been shown to be very susceptible to
high turbidities often accompanying pollution, but Anderson-Cue turbi-
dities have not risen nearly as high as those in some local lakes where
L;. sicculus is abundant.  Possibly the decline was caused by the ex-
tremely low water level and concomitant lack of littoral vegetation
suitable for spawning.

       Table VI-13 contains a partial faunal list from the study lakes.
This information was provided by Paul and Karolyn Maslin whose research
dealing with secondary and tertiary production levels formed the basis
for two dissertations reproduced in Vol. II as Appendix A and B.  In-
vestigations of this kind are very important as virtually no information
is available for the southeast U.S. regarding zooplankton species di-
versity, production rates and changes in species composition as influ-
enced by the eutrophication process.  The most significant conclusions
drawn from these works are described below.
       Equilibrium between zooplankton and environment.  Slobodkin (1954)
demonstrated that a fairly long time period, on the order of 40 days, is
required for a single cladoceran species, Daphnia obtusa, to reach an
equilibrium population is a constant laboratory environment and has sug-
gested that natural populations may never reach equilibrium before
further environmental changes occur.  In the light of the present study
several additions may be made to Slobodkin's hypothesis.  First, in a
mixed culture or natural community the environment of any one species
includes all of the other species as well as the physical parameters.
If some environmental change occurs, all populations will tend to ad-
just to the new conditions.  The adjustment is not instantaneous but
requires a time period of at least one and probably several generations.
As any one species population changes in adjustment to the environment
it will automatically alter the environments of all other species.  Thus
the time required for a community to equilibrate is greatly extended.
If other changes should occur before equilibrium is reached, the commun-
ity will be in perpetual nonequilibrium.  Forces will always be present
pushing the populations toward equilibrium, but equilibrium can never
be attained, since the equilibrial size of any population will be con-
stantly changing.
                                 -78,

-------
                           TABLE VI-13

       A PARTIAL FAUNAL LIST FROM LAKES ANDERSON-CUE AND McCLOUD
FISHES:
            Chaenobryttus gulosus
            Etheostoma fusiforme
            Fundulus chrysotus*
            Fundulus lineolatus*
            Gambusia affinis
LIMNETIC CLADOCTERA:
            Bosmina coregoni
            Daphnia arabigua
LITTORAL CLADOCERA:
                                               Heterandria formosa
                                               Labidesthes sicculus
                                               Lepomls macrochirus
                                               Micropterus salmoldes
                                               Diaphanosoma brachyurum
                                               Holopedium amazonicum
                                               Eurycercus lamellatus*
                                               Graptoleberis testudinata
                                               Ilyocryptus splnifer
                                               Macrothrix rosea
                                               Monospilus dispar
                                               Pleuroxus sp. (Hastatus?)
                                               Simocephalus expinosus*
            Acroperus harpae
            Acantholeberis curvirostris
            Alona affinis
            A.  costata
            A.  guttata
            A.  quadrangularis
            Alonalla globulosa
            Anchistropus minor*
            Camptocercus rectirostris
            Ceriodaphnia pulchella
            Chydorus biocornutus*
            C.  piger
            C.  sphaericus

*indicates species found only in Lake McCloud

LIMNETIC COPEPODS:                         LITTORAL COPEPODS:
            Diaptomus floridanus
            Cyclops bicuspidatus
            Tropolyclops prasinus
LIMNETIC ROTIFERS:
            Conochiloides sp.
            Conochilus sp.
            Keratella americana
            K.  taurocephala
            Pedalia sp.
            Polyarthra sp.
                                               Cyclops exilis
                                               Eucyclops speratus


                                           LITTORAL ROTIFERS:

                                               Brachionus sp.
                                               Keratella serrulate
                                               Lecane spp.
                                               Macrochaetus sp.
                                               Monommata sp.
                                               Monostyla sp.
                                               Trichocerca spp.
                                   -79-

-------
       Demonstration of cause-effect relationships between environmental
variables and population size or group biomass is difficult because of
the nonequilibrial nature of the relationship, as well as the lag re-
sponse described by Slobodkin (1954) and Edmondson (1965).  The lag
period necessary for any species population to adjust to an environmental
change is dependent upon temperature and food (Ingle, e_t al, 1937;
Hazelwood and Parker, 1961; Elbourne, 1966).  Accordingly, no two species
can be expected to have the same lag time, and lag time for any one
species will not be constant.  Even if variations in a single environmen-
tal factor are responsible for changes in zooplankton biomass, correla-
tion of biomass with the particular factor may be poor.  Comparison of
taxonomic groups, or even separate species, with environmental factors
should give truer correlations but even these will be blurred by vari-
ations in time lag with environmental changes.
       Competition patterns of zooplankton species.  The correlation
coefficients between populations of the different cladocera showed no
tendency for any species to decline while any other species was increas-
ing.  On the contrary, some showed positive correlations (D. ambigua to
H. amazonicum; JB. coregoni to D. brachyurum).  The pattern of dominance
in cladocera, however, showed that a given set of environmental condi-
tions might be distinctly more favorable to one species than to another.
Even though two species were positively correlated, dominance of one over
the other at a given time was almost certain.  In the summer of 1968 in
Anderson-Cue, all cladocera disappeared from the lake for a short time.
As they began to come back, three species were seen.  J3. cor eg on ia and
£. ambigua appeared together, with both populations increasing rapidly
but Bosmina. increased more rapidly than Daphnia, exceeding it in numbers
by an order of magnitude as both populations began to level off.  D.
brachyurum appeared about two weeks later than the other two but multi-
plied at a rate nearly equivalent to Bosmina's initial rate and soon
achieved a population roughly equal to Bosmina.  The Bosmina population
then decreased by about an order of magnitude while Diaphanosoma de-
creased only slightly, thus acquiring dominance.

       The above example shows two things.  First, competition probably
exists between these species.  Second, the effects of competition can
be obscured by changes in the environment.  When cladocera started to
come back into the system all three were favored, just as populations
of two or more species inoculated into a new medium would increase ini-
tially before competition became important.  After the populations
reached a "normal" level a distinct tendency for one to dominate was
displayed.
                                -80.

-------
         Food-predator-climate relationships.   The  presence  of competition
  but not competitive exclusion is  explained by the non-equilibrium nature
  of the  ecosystem.   If an equilibrium is  reached,  the  outcompeted  species
  will be eliminated, but  environmental changes effectively  give  the  losing
  species a new start.   If the environment is  sufficiently transitory
  species which would compete  strongly in  a laboratory  culture may  show a
  distinctly positive correlation.   In the event that biological  equili-
  brium should  be  reached,  causing  exclusion of some species   nearly  all
  fresh-water plankters  have some sort of  resting stage which  allows  them
  to develop a  new population  at a  later time  (Hutchinson, 1967).

         The smaller  coefficients of variation  for  groups as compared to
  species and the  better interlake correlation  of total and group biomass
  as  compared to species biomass indicate  that  factors acting  to control
  the  populations are acting on the whole  trophic level rather than on
  individual  species.  The good correlation (r  = 0.686) between zooplankton
 biomass in  the two  lakes indicates that  climatic  factors (i.e., factors
 which would affect both lakes) play a large part  in regulating'zooplank-
  ton populations, but the mechanism for action of climate on zooplankton
  is not intuitively obvious.

        While the summer decline was statistically significant in both
 lakes, the reason it occurred is less easily demonstrated.   When the
 summer data were ignored the  zooplankton biomass followed  the tempera-
 ture curve. A possible explanation for the  summer drop would be that
 temperature had passed the optimum for species present.  However,  zoo-
 plankton biomass returned to  normal and,  indeed,  to its highest value
 while temperature was still above  the level  of the initial  drop.  Also
 the recovery  from  the summer low  began while temperature was at maxi-
 mum.  The summer low could not be  due to  lack of food  because both
 primary  productivity and  chlorophyll concentration of  phytoplankton
 were higher during  the summer.

       A study of the  reactions  of individual species  to the  summer
 discontinuity  provides  an insight  into its cause.   Only the  larger
 zooplankters showed  a  significant decline during the summer.   Inter-
 mediate  species  showed  essentially no response, while very  small species
 such as  the rotifers increased significantly.  Also, rarer  species of
 rotifers were  much more likely to be  seen during the summer biomass  low.
 A further  clue  is provided by the  fact that larval copepods did  not
 decline;  they  even increased  slightly.  Evidently  conditions were  favor-
 able  for copepod  reproduction since fewer adults were able  to produce
 as many or more young.  Accordingly the decline in adults must have been
 due  to an  increased  death rate.  This  in  turn  points to an increased
 predation rate in summer with apparent selection for the larger forms
 producing results similar to  those described by Brooks and Dodson  (1965).

       An explanation for increased predation during summer is readily
available:   Labidesthes sicculus and Lepomis macrochirus are spawning
and the young of both are much more dependent upon zooplankton than are
 the adults.  Hubbs (1921) and  Werner  (1969) have discussed  the migration
                                -81.

-------
of fry of these species into the limnetic zone where they feed upon
zooplankton, particularly microcrustacea.  McLane (1955) reported that
in Florida Ij. sicculus spawns throughout the summer while _L. macrochirus
spawns from May to October with most intensive spawning in June.  Since
the bluegill fry spend ca. 1.5 months in the limnetic zone (Werner, 1969),
their period of heaviest predation falls exactly during the summer zoo-
plankton low.  The briefer summer decline in Anderson-Cue, 1968, as
opposed to 1967 or McCloud, 1968, also supports the predation mechanism;
the JL. sicculus population in Anderson-Cue, 1968, for some unknown reason,
was very much reduced by compari    with the 1967 population or the
McCloud population.

       When the period of summer decline is ignored, total zooplankton
biomass is closely correlated with both temperature and primary produc-
tivity.  Since temperature and primary productivity are strongly corre-
lated throughout the year, it seems reasonable that the influence of
climate on zooplankton biomass, as evidenced by the close correlation
between lakes, is due to two principal factors, neither of which acts
directly on the zooplankton.  (1) Temperature limits primary production
(given a reasonably constant nutrient supply), thus limiting the food
available to zooplankton and, accordingly, the zooplankton during most
of the year.  (2) The onset of fish reproduction, with its concomitant
predation surge, is determined by climatic factors, such as temperature
and photoperiod, and thus is synchronized in both lakes.

       In these lakes zooplankton as a group is food-limited during
most of the year.  During that period competition for food should be
important.  In the summer, however, when zooplankton biomass is at its
lowest and primary productivity is at its highest, no competition for
food should exist.  Thus the summer is the period when the rotifers,
relatively predation immune due to their small size, develop their
largest populations.  Also several rarer species of rotifers are much
more common during the summer.  Their apparent relationship to tempera-
ture merely reflects their reaction to the decreased competition during
the summer.  The same general explanation pertains to the change in
frequency of Daphnia and Bosmina dominance with warm temperatures.  Tem-
perature is not directly involved; Daphnia, being larger, receives
greater predation pressure than Bosmina.
       Secondary production estimates.  The estimations of secondary
production and efficiency are, admittedly, very rough.  The range from
minimum to maximum estimate covers about an order of magnitude.  How-
ever, the ranges for the 1968 values are reduced from that for 1967,
probably reflecting the greater accuracy obtained with more frequent
sampling.  The range of efficiencies determined includes from a rea-
sonably low efficiency to one impossibly high.  The efficiencies asso-
ciated with the "best estimate" are in line with the general ecology
efficiency of 8-12 percent (Slobodkin, 1968).
                                -82-

-------
       An advantage  of  this method  of  estimating  zooplankton  production
 is  that  contributions made by each  species  or higher  taxonomic group can
 be  assessed.   Thus  in both Anderson-Cue and McCloud,  1968, copepods
 accounted for  ca. 75 percent of  the  production, cladocera contributed
 ca.  24 percent, and  rotifer species  present coupled with their sporadic
 occurrence accounted for  their low  contribution to production.

       There have been  a  few previous  determinations  of aquatic secondary
 production.  As pointed out in the  introduction, most of these were only
 rough estimates.  The most important studies are summarized in Table VI-14.
 For  comparative purposes  all values are listed as originally  reported
 and  then as converted to  kcal m~2 year-1.   Of the previous workers only
 McAllister reported  any boundaries  to  production.  The range  of values
 from the present study  is quite  similar to  that reported by McAllister
 and  includes all of  the other values for standing fresh waters.  The
 "best estimate" of mean annual zooplankton  production in Lakes Anderson-
 Cue  and  McCloud is ca.  50 kcal m-2.  As one would expect for  oligotrophic
 lakes, this value lies  toward the low  side  of the range for standing
 waters.
       Zooplankton abundance factors.  Of the environmental parameters
examined, fluctuations in primary productivity and temperature had the
greatest effect on littoral zooplankton abundance.  This pattern was
especially seen in cladocera and rotifera where primary productivity
and temperature would be correlated with the rate of reproduction, and
thus, changes in population densities of both groups.

       In the multiple regression analyses copepoda were not highly
correlated with the three environmental parameters.  According to
Hutchinson (1967), variations in copepod egg number are less clearly
associated with environmental conditions than in cladocera.  Also, the
immature stages of the life cycle presumably give copepods a greater
range of filtrable food, thus making them more adjustable to varying
conditions than cladocera.

       Both Anderson-Cue and McCloud had their highest density of
cladocera in the summer, 1968, corresponding to high primary produc-
tivity.  While both showed a decreased density in the fall, Anderson-
Cue had a high density of cladocera from December through February and
McCloud maintained low populations.  This accompanied a high primary
productivity.

       Primary productivity in the two lakes was significantly correlated
(r - 0.511, p - 0.01) in 1968-1969.  However, in 1969-1970, this corre-
lation was higher (r» 0.907, p = 0.001), indicating a greater similarity
between the lakes.
                               .83-

-------
I
00
•P"
I
                                                    TABLE VI-14

                      SOME REPRESENTATIVE RATES OF SECONDARY PRODUCTION IN AQUATIC ECOSYSTEMS
Investigator

Odum and Yount
   (1954)

Stress, et al
   (1961)

Straskraba
   (1963)

Ilkowska, et al
   (1966)

McAllister
   (1969)

Present study
                                Organisms
crustacean
zooplankton

littoral
zooplankton

limnetic
zooplankton

marine
zooplankton

limnetic
zooplankton
                Ecosystem
Reported Rate
                                                                         —9    —1
                                most            Silver Springs,   240 g m zyear
                                herbivores      Florida
                                                bog lake


                                                fishpond
                                                fertile area
                                                Pacific

                                                oligotrophic
                                                lakes
8 g m 2year~


1.7 g N m~2year""1
1.5-23.2 g C m"2year-l


4-52 g m~2year~1
Converted to
kcal m"
                                                             1200
40
53
                                                two eutrophic     27 and 40 g nT^ear""1      135 and 200
                                                lakes
15-232


20-260

-------
        In  1969-1970 both  lakes again had high  simmer cladoceran  populations
and decreasing  fall populations.  Differences  between  cladoceran densi-
ties were  probably due  to an  interaction between environmental parameters.
Decreasing primary productivity and temperature and habitat changes due
to increasing lake level  resulted in a decrease in population density  in
both lakes in the fall.   However, the greater  abundance of C. sphaericus
in McCloud than in Anderson-Cue from January through March,~~1970, sug-
gests  that the  better developed littoral zone  of McCloud was not as
strongly affected by lake level fluctuations.  Benthic and vegetative
inhabiting species had  decreased population densities  in both lakes.
       Lake  level.  Fluctuations  in  lake level and the resultant effects
on vegetation  in  the  littoral  zone had the greatest  influence on species
diversity.   The general physical  condition of sand lakes produces drastic
changes  in lake levels.  During this study the above characteristic was
observed.  It may be  expected  that the variations found in the littoral
zone, as affected by  drought, would  influence the species diversity.
Since sampling depth  was kept constant throughout the study, vegetation
in the sampling area  was reduced  in  the fall due to  the rapid increase
in lake  level, slow macrophyte growth and delayed aufwuchs colonization.
The continued  increase in lake level might also have been responsible for
the lower values  of primary productivity in both lakes in the spring
of 1970 compared  to the spring of 1969.  The pattern of species diversity
was best related  to the changes in lake level, since diversity was low
as lake  level  increased and was high when lake level stabilized or was
dropping.  This would affect most chydoridae species since they are
normal inhibitors of  littoral vegetation.  Exceptions may be found in
those species either  inhabiting the  sediment water interface or the
limnetic zone  (A. quadrangularis and C. sphaericus).
       Prey selection.  Littoral cladocera were the most important food
items for Etheostoma fus iforme and Heterandria formosa.  When littoral
cladocera were scarce E. fusiforme fed mainly on copepoda and nonplank-
ton, while H. formoaa chiefly ate rotifers.  Food selection was deter-
mined by size and accessibility of prey.  Cladocera were the most important
items in the diet of both fishes.  Energetically, this seems reasonable,
since littoral cladocera are poor swimmers compared to copepods and
would be easier to capture.  Because of their small size, rotifers would
be energetically important food sources only when numberous.

       When cladocera were scarce, E_. f us if orme consumed copepods and
nonplankton, while H. formosa ate mainly rotifers and some nonplankton.
Rotifers were a large component of the diet even when they were not
abundant.
                                 -85-

-------
        In Anderson-Cue E. fuslforme had a mean electivity  index of  -0.43
 for copepods, while H. formosa had a value of -0.49.   In McCloud the
 mean electivity  index of 12. fusiforme was +0.12 and that for H. formosa
 was -0.28.  This  implies that E. fusiforme was somewhat better at ob-
 taining copepods  than H. frarmpsa, but they were not highly selected by
 either  fish.  Also IS. fusiforme  in McCloud was dependent on copepods
 for a longer time than in Anderson-Cue, due to the lower cladoceran
 populations.

       As evident from the low levels of predation pressure (Table VI-15))
 the two fishes had an abundant food supply when cladocera were numerous.
 Because of partitioning of resources when cladocera were scarce, compe-
 tition did not occur.  The period when their feeding habits were most
 similar (when their niches overlapped) was when food supply was not
 limiting.

        Prey selection appeared to be determined by a combination of
 factors.  Both fishes had high mean electivity indices for the same
 species of cladocera.  These were generally large and/or benthic species.
£• fusiforme always had higher indices for the smaller vegetative in-
 habiting species, but these were usually negative.  Mean indices for £.
 sphaericus were similar, indicating an equal availability.

        If the average percent of the diet made up by each cladoceran
 species is examined, the results are similar.  The large benthic species
averaged 5.0 percent of the diet of E. fusiforme and 2.4 percent of the
diet of H. formosa.  Small species which inhabit the vegetation consti-
 tuted 0.9 percent of the diet of IS. fusiforme and 1.6 percent of the diet
 of H. formosa.  Thus, the large benthic species were more important to
both fishes than the small vegetative inhabiting species.

       Size and accessibility seemed to be equally important factors in
 food selection.  Accessibility refers to the behavior of the prey, such
as concealment or ease of escape.  It also refers to characteristics
 of the predator which equip it for catching certain prey.  This is simi-
 lar to the definition of Ivlev (1961), although he considers accessibil^^
 to be more a property of the prey.  To a slight extent, size is a com-
 ponent of accessibility, since larger plankters are usually better
 swimmers.  Also, an organism may be too large or too small to be eaten.
 In this case all organisms studied were within a suitable size range
 for both fishes.

       Large benthic cladocera were selected by both fishes over small
 species.  However, since IS. fusiforme rests on the bottom these cladocef*
may be more accessible, which would explain higher electivity indices.
Among the rotifera and copepoda, accessibility seemed to be more impor-
 tant than size.  H. formosa selected rotifera over copepoda, perhaps be"
cause it was not as efficient as IS. fusiforme at catching copepods.
                                -86.

-------
                             TABLE VI-15

Monthly Predation Pressure by E_. fusiforme and H. fonnosa in Lakes
Anderson-Cue (A-Q) and McCloud (McC) , 1969-1970.
Organism Jul Aug Sep Oct Nov
A. curvirostris
A. harpae
A. qcadrangularis
C. piger
C. sphaericus
G. testudinaria
I. apinifer
H. rosea
P. striatus
S. exspinosus
Lecane spp.
Monostyla sp.
Copepoda
A-Q
McC
A-Q
McC
A-Q
McC
A-Q
McC
A-Q
McC
A-Q
McC
A-Q
McC
A-Q
McC
A-Q
McC
A-Q
McC
A-Q
McC
A-Q
McC
A-Q
McC

0.05
0.008
0.20
0.08
0.08
+
0.09
0.12
0.006
4-
0.34
0.03
0.021
0.52
0.08
0.63
0.04
0.05

0.05
0.03
0.35
0.04
0.02
+
4-
0.89
0.05
4-
0.64
++
0.02
0.03
0.93
0.17
0.31
0.21
0.054
0.14
0.05
0.16

0.10
0.08
0.038
0.01
0.24
0.09
•H-
0.10
•H-

0.22
4-
0.07
0.10
0.08
0.09
0.04
0.003
4-
4-
0.058
0.12
0.002
0.19
0.10
0.008
0.25
4-
0.008
0.28
+
0.17
0.17
0.41
0.23
0.03
0.024

+
0.01
0.01
4-
0.49
0.43
0.012
4-
4+
0.027
4-
4-
0.01

0.033
0.016
Dec
4-
0.11
4-
0.04
0.05
+
4-
0.36
0.37
0.072
0.01

4-
0.24
4-
4-'
0.105
0.12
0.007
0.16
0.006
0.03
Jan
0.24
0.68
0.59
4-
0.52
0.09
0.10
0.82
0.20
0.03
4-
0.044
4-
4-
4-
0.02
0.24
0.05
0.06
0.03
0.04
Feb
4-
0.
0.
0.
0.
0.
3.
0.
4-
0.


0.
0.
4-
0.
0.
0.
0.
025
62
49
087
045
56
05
024


02
014

014
006
01
018
Mar
4-
0.17
4-
0.107
4-
0.25
0.13
0.36
4-
+


+
4-
0.016
0.02
0.007
0.032
4-  - eaten, but not present in plankton sample.
4-f - large quantity eaten, but not present in plankton sample.
                               .87-

-------
       Abundance, of course, plays some role.  If an organism is so
rare that it is seldom encountered, it would not seem to be energetically
important.  However, small cladocera constituted sizeable portions of the
diet when scarce.  Thus, abundance was not a significant factor in prey
selection.
       Predation pressure.  The feeding of jS. fusiforme and H. formosa
on the littoral zooplankton is only a small part of the total predation.
All of the littoral fishes feed to some extent on them.  Also, species
of game fish and pelagic fishes spend a portion of their lives in the
littoral.  Bluegill fry, in particular, feed heavily on littoral cladocera
in these lakes (personal observations).

       The values of predation pressure were usually low.  The highest
value was on £. sphaericus, 3.56 percent, in February, 1970.  Dodson
(1970) measured removal rates of Daphnia due to salamander and midge
larvae predation in a similar fashion.  When his values are expressed
as percent removal per day, they are of the same order of magnitude as
the values for £. sphaericus in this study.  Hall (1964) estimated total
predation pressure on Daphnia in the summer to be 21.9 percent per day.
He concluded that predation was the most important factor controlling
population size in the summer.

       The fact that gut clearance rate increases with temperature and
that fish densities are greatest in the summer indicates a greater pre-
dation in the summer, which corresponds to the time of highest number
of species and abundance of littoral zooplankton.  This strengthens the
observation that the littoral zooplankton are not controlled by El.
fusiforme and H. formosa.   However, bluegill fry enter the littoral in
late August from the limnetic zone.  Since this is a time of decreasing
zooplankton populations, they may exert a much stronger predation
pressure.

       The effect of predation on prey populations has been examined.
Slobodkin (1961) speculated that where 'two species are limited by pre-
dation, they can coexist if they differ from each other ecologically to
some significant degree.  The stability of a system is increased by
ecological diversification of the species.  Connell (1961) found that
predation tended to decrease competition between two species of barna-
cles.  Paine (1966) suggested that animal species diversity is related
to the number of predators in a system and their efficiency in pre-
venting single species from monopolizing some limiting resource.

       Since many species of littoral zooplankton are approximately
the same size and utilize the same food supply, they would compete when
these resources were limiting.  By keeping prey species populations be-
low the level of resource limitation, predation could play an important
role in maintaining a high species diversity.
                                .88.

-------
       Eutrophication effects^  It was difficult to demonstrate a cause-
effect relationship between changes in littoral zooplankton and enrichment
of Lake Anderson-Cue.  Brooks (1969) mentioned that an increase in stand-
ing crop of herbivores was an expected consequence of enrichment.  The
total zooplankton was generally greater in Anderson-Cue than in McCloud.
However, a decrease in abundance was seen in both lakes the second sampl-
ing year.  Changes in abundance, species composition and diversity were
more likely correlated with changes in the littoral zone as a result of
lake level fluctuations than with enrichment.

       Whiteside and Harmsworth (1967) found a high correlation between
transparency and species diversity in Danish and Indiana lakes.  Since
chydorids are typically inhabitants of the macrophytic vegetation, they
suggest that this habitat extends to greater depths in clearer lakes.
Lesser penetration of light in more enriched lakes due to phytoplankton
blooms and dissolved organic matter would decrease habitat diversity and
thus, species diversity.  Whiteside (1970) divided the chydorid associa-
tions into three areas:   the littoral vegetational, the littoral benthonic
and the limnetic.  He found that in lakes of extreme eutrophy, limnetic
species (C_. sphaericus, A. rectangula) comprised most of the chydorids.
The benthic species were not as much affected by eutrophication as those
which inhabited the littoral vegetation.

       Changes in habitat diversity can occur as a result of long-term
phenomena, such as eutrophication, or short-term phenomena, such as lake
level fluctuations.  Both would have similar effects on diversity and
density of littoral zooplankton.  Also, changes in the composition of
zooplankton may result from selective predation.  In order to use lit-
toral zooplankters as indicators of eutrophication, the effects of
changes in environmental parameters, as well as selective predation, on
the population dynamics of littoral zooplankters must be understood.

       The effects of environmental variables on littoral zooplankton
populations are complex.  Because of their effect on the rate of repro-
duction, primary productivity and temperature are probably the most
important controlling factors.  However, the correlations between popu-
lation and density and primary productivity could be improved if the
contribution of periphyton was considered.  According to Fryer (1968),
periphyton is an important food source for several species of Chydoridae.

       This and earlier studies have shown that changes in habitat
diversity affect population density, as well as diversity of littoral
zooplankton.  Although lake level fluctuations were monitored, this did
not give an adequate reflection of changes in the littoral zone.  To
determine the effects of eutrophication on littoral zooplankton, quanti-
tative measurements of changes in macrophytes and aufwuchs are needed,
since these would be more indicative of changes in habitat diversity.

-------
       Predation by _E. fus 1 forme and H. formosa does not appear to control
population densities of littoral zooplankters in these lakes.  However,
a measure of total predation might show this to be an important factor,
especially when game fish fry are present in the littoral zone.
                                -90-

-------
                                SECTION VII

          TROPHIC  STATE  STUDIES  OF  lAKES  IN NORTH  CENTRAL  FLORIDA

       A  survey of  the  physical, chemical and  biological  features  of
 lakes  in  north central  Florida  was  initiated in 1968.  This was under-
 taken  for several specific  objectives:    (1) to assess the present
 trophic quality of  lakes  in this region;  (2) to provide baseline data for
 future studies on rates of  changes  in the quality of  these lakes;  (3) to
 gather sufficient information to evaluate the  appropriateness  of present
 trophic state criteria  in sub-tropical lakes;  and (4) to  provide neces-
 sary data to construct  an equation  or index  (or indices)  for trophic
 states in sub-tropical  lakes.

       Sampling programs have been  initiated on lakes in  three regions
 of north  central  Florida:    (1) Alachua  County (the location of Gaines-
 ville); (2) the Central Highlands  in Putnam and Clay  Counties; and (3) the
 lower Oklawaha River basin  in Lake, Seminole,  and Marion  Counties.  The
 most extensive sampling program is  being conducted on the lakes in Alachua
 County (for reasons of  convenience  and also because of the wide range of
 lake types available within the county).  Because of  the  larger number of
 lakes in  the Central Highlands  region, only selected  lakes are included
 in this survey.   The two model  lakes described in earlier sections of
 the report are located within this  region.  The lakes in  this area are
 somewhat  different  from and  generally higher in quality than lakes in
Alachua County.   These  lakes are becoming increasingly important recre-
 ationally as locations of homes for vacationers and retirees.  Conse-
 quently it was considered important to study the  trophic nature of these
 lakes.  The lower Oklawaha  River basin contains some of the largest and
most enriched lakes in Florida.  These lakes are  important for recrea-
 tion, and  in the  past were  considered among the best bass fishing  lakes
 in the state.  Considerable  evidence of rapid eutrophication has been
 found in  some of  the lakes within recent years, and agricultural runoff
has been  implicated as a major contributor of nutrients.   In view  of
 these facts it was  felt that any study of eutrophication would be  in-
complete without  some information on these lakes.   Numerous studies have
already been made on the Oklawaha lakes, and studies by groups such as
 the Florida Game and Fresh Water Fish Commission are continuing.    It is
not intended that the present study duplicate these efforts but comple-
ment them.  Because biological  information needed in trophic state indices
 is not available  for the lakes, limited sampling  on six of the largest
 lakes in  this region has been undertaken.

       Both qualitative and quantitative studies of the trophic state of
 lakes in north central Florida were performed.   The  results of the quali-
 tative study, using present  trophic state criteria,  illustrated the need
                                 -91-

-------
for a new, more quantitative approach  to  trophic  state classifications.
In this section both the qualitative and  the quantitative  trophic  state
studies are discussed.  The quantitative  study resulted  in the  formula-
tion of the Trophic State  Index, a novel  classification  criteria utili-
zing multivariate techniques.
Qualitative Trophic State Study

       This section describes  initial attempts to classify the  lakes
using present trophic state criteria.  It  is divided into two parts.
Part one describes the trophic state of the Alachua County lakes while
the second describes the trophic state of  the lakes in the Oklawaha
River Basin.
       Alachua County lakes.  All the significant  lakes in the county
are included in the survey; criteria for significance  include size  (all
lakes larger than about 3 hectares were sampled) and economic or recre-
ational value (any lake with a public or private access road or a per-
manent dwelling on its shore was considered significant in this regard).
Thirty-three lakes in or partly in Alachua County were found to meet
these criteria.  The locations were determined from U.S.G.S. topographic
and county road maps.  The  lakes and their locations are listed in
Table VII-1 and shown in Figure VII-1.  Not all the lakes are named,
and some of the names are known only to local residents.  In some cases
conflicting names appear on different maps; we have assumed the most
recent map to be correct in these instances.  Several  other lakes appar-
ently meeting the criteria are indicated on topographic maps but were
either swamps or completely dried land at the time of  sampling.

       Previous limnological efforts in Alachua County have been rather
sparse.  Because of its unusual characteristics, Lake Mize has been the
subject of several studies, including Harkness and Pierce (1940) and
Nordlie (1967).  Odum (1953) reported some values for dissolved phos-
phorus in Lake Mize and several other lakes in the county.  Nordlie
(1967) studied primary production and its limiting factors in Bivens
Arm, Lake Mize and Newnan's Lake.  The nitrogen cycle, nitrogen fixation
and organic color in Lake Mize have been studied by Brezonik (1969).
Miscellaneous data on several of the larger lakes can be found in Florid*
Geological Survey publications (e.g. Clark e_t a_l, 1962, 1964).

       It is apparent from Figure VII-1 that the lakes are not uniformly
distributed throughout the county.  Most of the lakes are located in the
eastern half of the county, and the four major lakes are in the eastern
third.  The terrain surrounding the lakes is remarkably varied, consider
ing the small geographical area.  Most of the lakes in the eastern part
of the county have heavily  forested shorelines, but the type of forest
varies for different lakes.  The large lakes have an outlet and one or
more inlet streams, but many of the small lakes have neither.
                                -92-

-------
                                       Table YTI-I
                               LAKES IN ALACHUA COUNTY
 Number    Name
Location
Depth1
1 Depth in meters
"Area in hectares
1 Symbols are as follows:
       O —  oligotrophic
       M •= mesotrophic
       E •= eutrophic
       HE «- hypereutrophic
       D •* dystrophic
       S «= senescent
       C •>> high organic color in lake water
Area*
Type'
Santa Fc Basin
1
2
3
4
5
Orange
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Santa Fe
Little Santa Fe
Hickory Pond
Altho
Cooter Pond
Creek Basin
Elizabeth
Clearwatcr
Hawthorn
Little Orange
Unnamed
Moss Lee
Jeggord
Still Pond
Lochloosa
Orange
Palatka
Newnan's
Mize
trout
North of Melrose
North of lake (1)
West of lake (2)
East of Waldo
Norlh central part of county

Southwest of Melrose
Southeast of Melrose
Hawthorn
Southeast of Hawthorn
South of lake (9)
Southeast of lake (10)
South of Hawthorn
East of Lochloosa
Southeast part of county
Southeast part of county
South of lake (17)
East of Gainesville
Northeast of Gainesville
Southeast of Gainesville
8.5
6.5
4
4.5
3

2.5
3
3.5
3
4

4

3
3
1
2
22
1.5
1760
464
32
228
124

57

35
241
50
57
85

2484
3105
25
2562
0.9
15
OC
OC
OC
OC
M

MC
O
EC
MC
MC

D-MC

E
E
D-S
HEC
D-OC
D-EC
No Surface Drainage Region
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Meta
Unnamed
Bivens Arm
Alice
Clear
Unnamed
Unnamed
Unnamed
Kanapaha
Watermelon Pond
Long Pond
Burnt Pond
Wauberg
Tuscawilla
Northwest Gainesville
South Gainesville
South Gainesville
U. of F. campus
Southwest Gainesville
West of Gainesville
South of lake (27)
Northwest of Gainesville
Southwest of Gainesville
Southwest part of county
West of Micanopy
West of lake (32)
Northwest of Micanopy
South of Micanopy
1.5
3
2
2
2
2

5
1
2
1.5
2.5
5
2
3.9
1.8
70
87
4.3
7.4
2.2
5.4
84
632

23
103
64
M
E
HE
S
EC
M
D-MC
M
E-S
D-OC
D-OC
EC
E
MC
                                       -93-

-------
Figure VII-1. Alachua County Lakes
              -94-

-------
       On the basis of surface runoff, the county can be divided  into
 three  zones as shown  in Figure VII-2.  The southeastern third of  the
 county lies within the Orange Creek Basin; fourteen county lakes  are in
 the basin, but not all are connected by surface streams.  The Santa Fe
 River  basin in northern Alachua County has only five lakes but consider-
 able swampland.  The  southwestern part of the county, comprising  some 300
 square miles, is an area from which there is no surface outflow.  The few
 small  streams terminate in limestone sinkholes.  Fourteen lakes are lo-
 cated  in this region, but most of them are small ponds in the vicinity
 of Gainesville.

       Figure VII-2 also indicates surface distribution of geological
 formations in Alachua County.  A brief description of these surface
 formations (summarized from Clark ejt al, 1962) is presented below.  The
 Hawthorne Formation is a marine deposit of Miocene Age  consisting of
 thick and sandy clays interbedded with phosphatic limestone.  This forma-
 tion is exposed in central, northern and eastern Alachua County.  Higher
 terrace deposits outcrop in a 70 square mile area in northern Gainesville
 and north central Alachua County.  These deposits are of Pleistocene Age
 and consist of fine to medium sands, clayey sands, and multi-colored
 clays.  The Citronelle and Alachua Formations are of Pliocene Age.  A
 small area of Citronelle outcrops in eastern Alachua County.  This is a
 nonfossiliferous deltaic deposit of sand, gravel and clayey sand.  The
Alachua Formation is exposed to western Alachua County and consists of
 terrestrial deposits of white, gray and colored sands, clayey sands and
 some multi-colored clays.  Vertebrate fossils, limestone and phosphate
 pebbles and boulders are scattered throughout the deposit.  The Ocala
 Group of Eocene Age limestones is at the surface in western and southern
Alachua County.  In much of the area shown in Figure VII-2 as having the
 Ocala Group at the surface, a thin layer of residual sands and clays
 covers the limestone.  The Ocala Group consists of coquina, hard  and
 soft limestones, and  dolomitic limestones.  The Ocala Group underlies all
 of Alachua County and is the main part of the Floridan aquifer.   Because
 of the underlying Ocala limestones, Alachua County can be described as
 generally having a karst topography.  Filled and open sinks, sinkhole
 lakes, solution pipes and lakes and prairies are typical of such areas,
and the county has these features in abundance.  Further details about
 the drainage basins and general landforms in Alachua County can be found
 in Clark e_t al, (1964).

       Large areas of the county are covered with slash pine for  pulp
 purposes, and extensive swamplands exist.  Particularly in the northern
and eastern areas, the swamplands are forested with cypress and other
 trees.  Consequently much of the surface water in the county is highly
colored with organic exudates and plant degradation products.   While
agriculture in the broad sense occupies an important place in the economy
 of the county, relatively little land is devoted to annually harvested
crops.  Corn and watermelons are perhaps the largest crops and are grown
 in widely scattered areas.  Some tobacco is grown in the northwest part
 of the county, and a few orange groves are located in the south-central
                                -95-

-------
                                                                                    AlocKua
                                                                                   Formation
                                                                                     Ocala
                                                                                     Group



                                                                                    Hawthorn*
                                                                                    Formation
                                                                                   ij^cr ferrae*
                                                                                   deposits
                                                                                   CifroneJIe
                                                                                   format i on
Figure VII-2.   Drainage Bas.ins  and  Geologic  Map  of Alachua County
                  (After  Clark e_t  al.,  1962)
                                       -96.

-------
  part near Orange Lake.   Considerable land is devoted to cattle grazing
  In general,  agriculture  would seem to have little  effect on the quality
  of surface waters in the county.   The effects of geology and surface
  land use  on  the  water quality of  individual  Alachua  County  lakes will  be
  discussed in later paragraphs.  The  physical,  chemical,  and biological
  parameters measured in each  lake  are enumerated  in Table VII-2.

         The lakes covered in  this  study range  in  size  from 0.9  hectares
  (Lake Mize)  to 3100 hectares  (Orange Lake).   Only eleven of the  lakes
  are  larger than  100 hectares, and  only four  of these  are larger  than
  1,000 hectares.   Most of the  lakes are quite  shallow, Lake  Mize  being
  the  only  significant exception.  This  small  lake is a limestone  solution
  sink and  has a depth of  at least 25  meters.   Santa Fe Lake  with  a maxi-
 mum  depth around 8 meters is  the second deepest  lake  in  the county.  A
 majority  of  lakes have maximum depths  of 3-4 meters,  but about a third,
  including some of the larger  lakes have maximum depths of 2 meters or
  less.  Bathymetric maps  are available  for only three  (3) county  lakes:
 Newnan's, Mize,  and Orange.  The depths recorded in Table VII-1 are the
 maximum values found at  several widely scattered sampling stations on
 each lake.  More detailed soundings may locate deeper areas in some lakes,
 but we believe the values in Table VII-1 to be closely representative of
 the maximum depths.

        Because of their shallow depths, few of the  lakes in Alachua County
 show significant thermal stratification.   Lake Mize is the only known
 monomictic lake,  having a well developed  thermocline  from March to Novem-
 ber (Brezonik,  1969).  Santa  Fe Lake  may  stratify at  least for brief
 periods  in spring but data are not yet available  to prove this.  Even
 shallow  lakes will stratify briefly during calm periods  of intense
 warming.   Such  conditions occur most  commonly in  spring,  but can occur
 in fall  and winter also.   Evidence  for temporary  stratification has  been
 found in a number of lakes, but  temperatures  differ from top to bottom
 usually  by only a few degrees  Celsius,  which  is insufficient to prevent
 mixing by  normal  winds.

       Conditions in three of  the  four large  lakes  (greater  than 1,000
 hectares)  in  Alachua County leave much to  be  desired.  Relevant data
 concerning trophic conditions  in these  lakes  are  summarized  in  Table VII-3.
 Whether conditions have been affected by cultural influences is  not yet
 certain, but  natural factors,  especially geological and morphological
 seem  adequate to  explain  most  of the  conditions.  Santa Fe Lake  is the
 only  oligotrophic lake in this size category.   Its maximum depth  (8
 meters) is not particularly impressive, but it  is by far  the deepest of
 the larger lakes.  An important factor  relating to the lake's oligo-
 trophy is  its small  drainage basin, the majority of which is forest.  A
moderate number of cottages and homes dot the shoreline,  but no other
 urban or agricultural eutrophying influences are evident.  Santa  Fe Lake
 lies wholly within the Hawthorne Formation, which contains phosphatic
clay deposits, but this circumstance  is apparently mitigated by the small
drainage basin.   Phosphate and inorganic nitrogen levels in Santa Fe
Lake are relatively high compared to the critical  levels suggested to
                                -97-

-------
Lake depth
Lake area
Temperature
A.   Water

Acidity
Alkalinity
Calcium
C.O.D.
Chloride
Color
Conductance
Fluoride
Iron

B.   Sediments
                               TABLE VII-2
                    PARAMETERS MEASURED IN LAKE STUDY
                                Physical
                     Secchi disc transparency
                     Land use in lake basin
                     Shoreline development
                                Chemical
Magnesium
Manganese
Organic nitrogen
Ammonia
Nitrite
Nitrate
Oxygen, dissolved
PH
Orthophosphate
Total phosphate
Potassium
Silica
Sodium
Suspended solids
Total solids
Sulfate
Turbidity
Ammonia                                    Percent volatile solids
Organic nitrogen                           Iron
Total phosphate                            Manganese
Sediment type (visual classification as peat, muck, etc.)
                               Biological
Algal identification and counts
Chlorophylls a, b, and £
Total carotenoids
Primary production
                     Species diversities indices
                        of algae
                     Visual classification of
                        vegetation surrounding lake
                                -98-

-------
                         TABLE VII-3

TROPHIC CHARACTERISTICS OF LARGE LAKES IN ALACHUA COUNTY1'2


Parameter3	Santa Fe L.   Newnan's L.   Orange L.   Lochloosa L.
Secchi disc
Turbidity
Sp. conductance
Acidity
Alkalinity
pH
COD
Color
TON
Airanonia-N
Nitrate-N
Ortho P
Total P
Chloride
Sodium
Calcium
Iron
Manganese
Silica
Organism count
Chlorophyll a^
Primary production
2.4
5.4
48
0.6
2.6
6.5
16
130
0.44
0.22
0.0
0.001
0.11
10.4
7.5
1.3
0.02
0.008
0.11
28
5.56
13.5
0.6
5.8
60
0.9
9.0
7.6
79
490
1,13
0.02
0.0
0.006
0.05
10.8
9.8
4.0
0.06
0.003
0.68
2525
8.62
53.6
0.6
6.0
67
0.5
15
7.6
56
280
1.09
0.02
0.0
0.005
0.05
9.5
8.1
6.0
0.02
0.050
0.57
2234
9.68
43.0
0.8
7.9
90
0.5
24
8.3
55
200
1.17
0.01
0.0
0.004
0.05
10.6
9.6
10.0
0.02
0.002
0.16
5670
12.5
35.6
  Lakes  larger than 1,000  hectares.

  Data from December,  1968,  sampling;  data from June,  1968  show
  similar trends.   Organisms counts  are  from June  samples since counts  from
  December samples  are not yet  available.   They are presented  for
  comparative purposes among themselves  only and should not be com-
  pared  with the  other chemical and  biological  results (from December).
  Each number represents a single  determination on a composite sample
  taken  from three  stations  in  the lake.

  Units  for the parameters are  in  mg/1 except as follows:    Secchi
  disc transparency in meters;  specific  conductance in ymho cm~l;
  acidity and alkalinity in  mg/1 as  CaCQy,   pH  in  pH units;  color in
  mg/1 as P +;  organism counts  in  numbers/ml; chlorophyll a. in mg/m3;
  and primary production in  mg  C/m3~ hr  on composite  samples  run in
  a  laboratory incubator.
                              -99-

-------
stimulate blooms in north temperate lakes, but concentrations are lower
than in the other large lakes.  The biological data are all indicative of
oligotrophic conditions; color (which limits light transmissions) and
depth are probably contributing factors to this condition.

       Newnan's, Orange, and Lochloosa Lakes are in the Orange Creek
drainage basin, and all show considerable evidence of eutrophy.  New-
nan's Lake lies within the Hawthorne Formation outcrop, and its major
influent stream, Hatchet Creek, drains a large area of the formation
north of the lake.  Water flows from the southern end of Newnan's Lake
through a creek, to a man-made canal and a meandering river into Orange
Lake.  The northern half of Lake Lochloosa and most of its drainage
basin lie within the Hawthorne Formation.  Orange and Lochloosa Lakes are
connected by Cross Creek, but the major flow from both lakes is through
outlets to Orange Creek to the southeast.  The high trophic conditions
of these lakes are at least partially the result of edaphic considerations.
There are no major cultural sources of nutrients for these lakes although
fertilized pasture land is drained by Hatchet Creek and the towns of
Orange Lake and Mclntosh probably release some sewage into Orange Lake.

       Newnan's Lake is especially peccant; its extreme shallowness must
be a contributing factor.  The maximum depth shown on the bathymetric
map is 3.6 meters (12 feet), but the mean depth is under 2 meters.  With
the large surface area, winds must be quite effective in mixing sediments
with the overlying water.  Orange and Lochloosa Lakes are only slightly
deeper, and the geological and hydrological situation implies they are
following shortly behind Newnan's Lake on the same avenue of degradation
and extinction.  All three lakes had profuse algal blooms in June of 1968.
The November standing crops were much lower though still blooming accord-
ing to Lackey's (1945) definition (500 organisms/ml).  Essentially all
the criteria (see Table VII-3) reaffirm the advanced eutrophy of the
three lakes.  The large lakes in the Orange Creek basin produce an abun-
dance of fish and are popular with sport fishermen.  The advanced eutrophy
suggests game fishing may be in a somewhat precarious position, and take-
over by rough and trash fish could conceivably be imminent, particularly
in Newnan's Lake.  Any plans to increase the man-made sources of nutrients
to these lakes should be viewed askance.

       Results from the medium sized lakes are summarized in Table VII-4.
These lakes are widely scattered geographically and present an interesting
spectrum of trophic conditions.  With the exception of Lake Wauberg, the
lakes are in good condition.  Wauberg is in an advanced state of eutrophy
with frequent and obnoxious algal blooms.  The reasons for this condi-
tion are not entirely clear.  The lake is not excessively shallow
depth is at least 5 meters) and receives no urban runoff, sewage effluent
or large amount of agricultural runoff.  The drainage basin seems small
and is in the Ocala Limestone Formation.  In past years the lake was used
extensively for picnicking and swimming; a semi-private beach and res-
taurant were located on the lake.  But the potential nutrient additions
from these sources seem inadequate to explain the present conditions.
The University of Florida now owns the land around the lake and operates
it as a camp and recreational facility for the university community.
                               -100-

-------
                                           TABLE VII-4

                 TROPHIC  CHARACTERISTICS  OF MEDIUM-SIZE AIACHUA COUNTY LAKES '
Parameter

Seccht disc
Turbidity
Sp. conductance
Acidity
Alkalinity
pH
COD
Color
TON
Ammonia -N
Nitrate-N
Ortho-P
Total P
Chloride
Sodium
Calcium
Iron
Manganese
Silica
Organism count
Chlorophyll a_
Primary production
Altho

1.7
3.3
49
1.3
3.2
6.5
22
250
0.62
0.23
0.03
0.002
0.09
10.1
7.2
2.0
0.01
0.008
0.30
633
4.84
10.3
Cooter

1.2
4.5
63
1.4
4.8
6.9
87
250
1.12
0.45
0.01
0.003
0.16
7.6
5.7
2.2
0.01
0.009
0.81
431
21.9
87.0
Little
Santa Fe

1.3
6.3
50
3.2
1.6
5.7
31
385
0.57
0.26
0.02
0.000
0.10
9.7
7.0
1.3
0.03
0.014
0.31
150
3.28
6.6
Little
Orange

1.0
4.4
54
4.9
1.1
5.9
68
495
0.89
0.16
0.00
0.005
0.04
8.8
6.6
2.2
0.04
0.019
0.31
2352
4.11
12.7
Tuscawilla

1.2
3.3
55
6.0
12.8
6.6
52
625
0.85
0.03
0.03
0.024
0.28
7.7
6.0
4.0
0.03
0.003
2.26
341
5.76
12.2
Watermelon
3

4.0
32
2.76
0.0
4.8
23
285
0.86
0.20
0.03
0.001
0.05
6.2
4.5
0.5
0.01
0.042
0.08
125
6.44
5.33
Wauberg

0.9
10.1
59
0.0
15.7
8.3
25
125
1.71
0.09
0.00
0.005
0.15
8.7
9.4
6.0
0.00
0.004
0.17
9288
30.1
124.3
 Lakes between 100 and 1,000 hectares

2
 See Table VII-3 for sampling dates and units of parameters

3
 Disc visible to bottom (1.6m)

-------
       Lake Altho and Little Santa Fe Lake are classified as colored
oligotrophic.  The two lakes are connected by an improved canal, and to-
gether with Santa Fe Lake, they form the headwater of the Santa Fe River.
Santa Fe and Little Santa Fe Lakes in a sense are one lake with two basins
separated by peninsulas which constrict the water to a short channel
several hundred yards wide.  Both Altho and Little Santa Fe are moderately
deep (for this area) with maximum depths of 5 and 6.5 meters, respectively-
Both have relatively small drainage basins and no sources of cultural
enrichment other than a relatively small number of summer cottages and
homes.  Little Orange Lake, in the Orange Creek basin, is somewhat simi-
lar to the above two lakes, although it is more highly colored (about 500
ppm in November) and shallower (maximum depth 3.5 meters).  The lake is
surrounded by forest and is dotted with cottages, but there are apparently
no major cultural influences.  The lake is somewhat difficult to classify-
Its high color implies dystrophy, but the lake has moderate nutrient
levels and a varied population of diatoms, blue-green algae and green
algae.  The lake has been tentatively classified mesotrophic-colored.

       Watermelon Pond is an irregularly shaped lake in the southwest
area of the county.  The lake has a moderate color, and except for its
shallowness and low pH, the data conform to the usual criteria for oli-
gotrophy.  Cooter Pond is in the Santa Fe drainage basin but is not
connected to the river by a permanent surface outlet.  The lake is now
surrounded by pasture, but until recently groves of tung trees were cul-
tivated on the land.  The general conditions now suggest mesotrophic to
eutrophic conditions.  The shallowness (2.5 meters maximum depth) implies
a small capacity to assimilate more nutrients, and the lake appears to
be susceptible to further eutrophication by agricultural runoff and cat-
tle wastes.  Tuscawilla Lake occupies a gently sloping depression in a
rather flat region south of Micanopy.  The lake is shallow and in periods
of drought (such as spring of 1968) its area shrinks considerably.  The
lake is highly colored and has a low ionic content.  Nitrogen levels are
moderate but phosphorus is high, and an abundance of macrophytes is
found in the littoral zone.  The chemical conditions suggest mesotrophy,
but the shallowness and extensive macrophytes imply the lake may be
approaching senescence.

       Thirteen lakes scattered throughout Alachua County are in the size
range 10-100 hectares.  Included are lakes in all classifications; con-
ditions relevant to trophic state are summarized in Table VII-5.  Space
does not permit discussion of all these lakes, but it is pertinent to
note that three of the four Gainesville area lakes in this size class
show evidence of cultural effects.  Bivens Ann receives urban runoff
from a small stream in the southern part of Gainesville and has probably
been affected by University of Florida experimental cattle farms on its
northwest shore.  The lake has an interesting history; the U.S.G.S.
topographic map of Gainesville dated 1895 shows Bivens Arm connected
with Alachua Lake (Payne's Prairie), both of which were then dry.  The
latter never became a (water-filled) lake again, but Bivens Arm is now
a permanent lake separated from Payne's Prairie by built-up roads.
                               -102-

-------
                                           TABLE  VII-5
                   TROPHIC CHARACTERISTICS OF SMALL LAKES  IN ALACHUA COUNTY
1,2
                  Bivens         Eliza-  Haw-    Hick-  Jeg-    Kana-        Moss

Parameter  Alice  Ann     Burnt   beth    thorne  ory    gord    paha  Long  Lee   Palatka  Trout   #10











H
o
V*
1











Secchi
disc
Turbidity
Sp. cond.
Acidity
Alkalinity
PH
COD
Color
TON
Ammonia -N
Nitrate-N
Ortho-P
Total P
Chloride
Sodium
Calcium
Iron
Manganese
Silica
Organism
count
Chlor a
Prim.
prod.
3
^
5.3
533
13.4
178
7.4
12
130
0.33
0.01
0.00
0.070
0.59
16.7
16.0
71
0.01
0.003
12.7

16
1.60

0

0.9
10.5
263
0.0
139
8.5
65
75
1.19
0.01
0.00
0.020
0.32
15.3
15.0
50
0.01
0.003
1.42

2600
11.0

77.5

0.9
4.9
67
3.7
15
6.8
66
620
1.87
0.47
0.00
0.039
0.36
9.6
6.5
10.0
0.04
0.007
0.64

1372
21.4

54.4

0.9
3.9
38
2.1
2.0
5.9
48
155
0.72
0.08
0.07
0.005
0.11
7.5
5.4
1.9
0.5
0.019
0.68

625
2.98

0.58

1.2
3.9
185
1.1
86
8.0
37
95
1.22
0.05
0.01
0.003
0.08
11.2
11.0
33.0
0.00
0.003
0.27

26,209
19.4

55.5

2.1
2.8
39
0.9
1.8
6.6
21
160
0.65
0.13
0.00
0.017
0.04
10.0
6.0
0.9
0.01
0.025
0.30

119
6.36

7.52

1.2
7.5
53
2.7
0.6
5.2
26
345
0.41
0.11
0.01
0.005
0.12
8.6
7.7
0.5
0.03
0.027
1.56

3824
4.28

4.26

0.6
7.2
166
4.0
70.5
7.3
78
185
3.84
0.96
0.03
0.014
0.42
8.0
10.9
19.0
0.02
0.01
0.09

11,736
9.48

26.9
3
•**
3.3
16
6.4
0.0
5.1
38
240
0.89
0.05
0.02
0.001
0.04
4.4
2.5
0.6
0.01
0.033
0.18

—
2.56

1.42

2.0
5.4
43.0
2.8
1.2
5.9
46
278
.79
.04
.10
.010
.043
8.9
4.9
1.8
.00
.011
.017

1
3.14

12.9
3
•j
3.3
22
7.4
0.0
4.8
72
185
0.86
0.14
0.00
0.002
0.08
3.8
3.1
0.4
0.01
0.005
0.71


3.34

3.36

0.6
1.8
41
8.9
1.2
5.3
64
990
1.55
0.40
0.11
0.14
0.16
7.9
6.0
1.6
0.1
0.05
1.80


10.8

10.5

1.6
4.4
44
5.0
2.5
6.0
48
305
0.81
0.06
0,01
0.027
0.10
8.8
6.6
1.7
0.02
0.012
0.35


3.29

17.4
 Lakes between 10 and 100 hectares.

 ?
 "See Table VII-3 for details on sampling dates and units of parameters.


 Disc visible to bottom.

-------
Sediment cores would provide considerable information on the nature of
Bivens Arm when it was connected with Alachua Lake and the ontogeny of
the lake since it was separated and refilled.  The lake may have been
eutrophic all along; but it seems likely that the above nutrient sources
have intensified conditions in the recent past.

       Lake Alice has been classified senescent.  This lake was once
eutrophic, but in recent years nearly all of the lake's surface has been
covered with water hyacinths.  The lake is shallow, and decaying vege-
tation produces obnoxious odors.  Treated sewage from the University of
Florida waste treatment plant enters the east side of the lake, but nu-
trient concentrations in the water do not reflect this source of enrich-
ment.  The extensive hyacinth growths are evidently effective nutrient
removers.  The lake is almost devoid of phytoplankton as a result of the
light cover provided by the hyacinths, the relatively low nutrient con-
centrations and perhaps antibiotic effects of the macrophytes.  A large
volume of cooling water from the University steam plant enters Lake
Alice daily.  This undoubtedly has an important effect on the lake's
biota, and probably prevents planktonic populations by a flushing effect.
In 1970 a restoration program was carried out on Lake Alice which era-
dicated all of the hyacinths from the lake surface.  This program hope-
fully will improve the lake's quality of surface water.

       Lakes Kanapaha and Hawthorne are also culturally enriched.  The
former lake is connected with Kanapaha Sink, the terminus of Hogtown
Creek in Gainesville; the latter lake receives some sewage from the town
of Hawthorne.  Burnt Pond is apparently naturally eutrophic.  Its shallot"
ness indicates the lake may not be far from the senescent state.  Most
of the other lakes in this size category are in relatively good condition.
High organic color and relative isolation from urban and agricultural
development characterize these lakes.

       Nine of the lakes in this survey are smaller than 10 hectares.
There are also innumerable small ponds (one hectare or less) scattered
throughout the county (particularly around Gainesville).  These are
usually quite shallow and can hardly be considered as lakes in the usual
sense of the word.  The nine lakes chosen in this survey include all
the large lakes in this size category (i.e. lakes between 3 and 10 hec-
tares) and a few smaller lakes chosen because of their unusual charac-
teristics or potential significance in a broad recreational sense.  The
trophic conditions of these lakes are summarized in Table VII-6.  Lake
Mize is perhaps the most interesting of these small lakes, because of
its great depth (about 25 meters).  The lake exhibits the usual charac-
teristics of dystrophy--high color and acidity, low pH, few organisms,
etc.  The lake is remarkably similar morphologically and chemically to
Lake Mary, Wisconsin, except for the latter*s mesomixes.  Lake Mize is
located in the University of Florida Austin Gary Memorial Forest, which
one might presume would augur the preservation of this unusual lake.
Unfortunately an enclosure for maintaining and raising waterfowl was re-
cently allowed to be built on the shore, and now over fifty ducks
contribute untreated excrement directly to  the water.

-------
                                          TABLE VII-6
              TROPHIC CHARACTERISTICS OF PONDS AND LAKELETS IN ALACHUA COUNTY
1,2
Parameter
Secchi disc
Turbidity
Sp. cond.
Acidity
Alkalinity
PH
COD
Color
TON
Ammonia -N
§ Nitrate-N
g Ortho-P
yx Total P
Chloride
Sodium
Calcium
Iron
Manganese
Silica
Organism
Count
Chlor a
Prim. prod.
Clear
0.8
10.0
106
0.0
27.6
8.9
44
345
1.38
0.02
0.00
0.007
0.17
8.2
7.0
10.0
0.04
0.005
1.68

5,896
11.4
69.1
Clearwater
1.8
2.8
33
1.5
0.8
5.4
14
25
0.59
0.10
0.00
0.003
0.05
7.4
5.2
0.7
0.01
0.023
0.14

9
3.32
0.33
Meta
1.4
6.5
86
1.4
27.4
7.6
40
90
0.83
0.09
0.00
0.001
0.06
8.2
7.0
10.0
0.00
0.003
0.14

15,160
5.51
3.59
Mize
1.0
3.9
45
5.8
1.4
5.6
34
260
0.51
0.26
0.01
0.005
0.09
9.2
5.3
0.8
0.05
0.010
0.98

88
24.2
7.46
Still
#21
0.8
10.6
275
0.3
128
8.2
50
165
1.33
0.02
0.00
0.004
0.25
17.7
15.0
48
0.00
0.003
2.91

10,840
39.0
235
#25
3
0.9
106
0.0
50.4
8.3
17
75
0.73
0.07
0.03
0.001
0.06
3.4
3.1
20.0
0.00
0.016
0.57

163
0.18
6.06
#26

0.9
38
9.8
0.6
5.3
29
500
0.74
0.03
0.01
0.004
0.12
9.1
6.3
0.9
0.05
0.015
1.60

303
6.86
2.09
#27
1.8
3.9
43.0
9.0
5.7
6.0
28
220
0.56
0.04
0.00
0.012
0.20
7.7
5.6
2.1
0.02
0.014
0.66

18
12.7
41.0
Still
Pond
.7
3.8
40
1.3
.6
6.0
20
70
.58
.07
.02
.006
.016
7.8
5.5
1.0
.00
.004
.002

«-.
1.81
0.2
Lakes smaller than 10 hectares.

See Table VII-3 for details of sampling dates and units of parameters,

Disc visible to bottom.

-------
       Lake #27 (unnamed) is unusual in that it is completely covered
with duckweed (Lemna) normally indicative of high nutrient conditions.
The lake is isolated from urban and agricultural effects and otherwise
exhibits oligotrophic characteristics.  The lake is moderately deep
(for this area)--maximum depth is at least 5 meters.  There does not
seem to be a satisfactory explanation for the extensive duckweed growth
in this lake and its virtual absence in all the other lakes of this sur-
vey.  Clear lake was evidently named before the onset of a eutrophic con-
dition which contradicts its name.  The lake is surrounded by homes with
septic tanks and has become eutrophic only in recent years (Furman, per-
sonal communication).  An extremely dense Anabaena bloom was present in
the June, 1968, sample and the lake oscillates between relatively clear
and bloom conditions.  Lake Meta  (in northwest Gainesville) also shows
some signs of cultural enrichment from surrounding homes, but to a
lesser extent.

       Even cursory  inspection of the chemical and biological results
obtained reveals graphic disparities in trophic conditions among the
lakes.  In fact inspection of the lakes themselves illustrates this in
a dramatic albeit crude way.  An attempt to classify the lakes according
to the usual trophic types is included in Table VII-1.  Some geographi-
cal patterns in trophic type are  evident but not altogether conclusive.
Lakes in the Santa Fe River basin are oligotrophic and somewhat colored.
Many of the small lakes in the Orange Creek basin are also oligotrophic
and most are highly  colored, but  the large lakes are eutrophic.  Lakes
in  the southwestern  part of the county are quite varied with respect  to
trophic levels.  Some of the small  lakes in urban and suburban Gaines-
ville show evidence  of eutrophy;  cultural influences may have exerted
some stress on  these lakes.

       The trophic criteria do not  give clear  indications  of  trophic
state  in all cases.  Some  of the  smaller  lakes have both oligotrophic
and eutrophic  characteristics, and  many  lakes  have  dystrophic character-
istics along with oligotrophic or eutrophic conditions.  The  lakes with
oligotrophic and eutrophic  criteria are  classified  as mesotrophic; Lake
Elizabeth  and  Cooter Pond  are examples.  These lakes may be  in a  trans-
itional state  between  the  two  trophic  levels.   Alternatively, these may
be  examples  of the  inadequacy  of  trophic  criteria developed  for  temper-
ate lakes  when applied  to  subtropical lakes.   The inadequacy of  the
dystrophic classification  is amplified by the  results  from this  study.
Over  half  of  the  lakes  show some  signs of dystrophy,  i.e.  high  color,
 low pH,  low  ionic  solids.   However, high color is not  always correlated
with  other criteria  for  dystrophy.   Some colored lakes  have  neutral  or
alkaline  pH  (e.g.  Lake  Kanapaha)  or high calcium and alkalinity (Lake
#21).   The range  of nutrient concentrations  and plankton  in  these lakes
 is  from unproductive to hypereutrophic.   It  is obvious  that  one class
 (dystrophy)  is insufficient to describe  all these lakes,  even though
 each has  several  criteria  indicating dystrophy.  Hansen's  (1962)  dicho-
 tomic classifications  where lakes are classified as colored  or  clear an
 each of these has the full range  of trophic  states (oligotrophy to eu-
 trophy) are  an improvement over classical typology.  However, eutrophi-
                                .106.

-------
cation in colored and clear lakes may proceed along different courses,
and trophic states in the two classes may not be strictly parallel
(e.g. different biota and trophic structures may result from eutrophica-
tion in the two classes).  Lakes showing partial dystrophy in this study
have been classified colored-oligotrophic, colored mesotrophic, etc.,
as appropriate.  A few lakes with rather extreme dystrophic conditions
(e.g. Lakes Mize, Palatka, Jeggord, Long Pond) have been tentatively
classified as dystrophic along with an alternate classification such as
colored-oligotrophic.  These are subject to change as the classification
scheme becomes further developed and refined.

       The shallow lakes present especially difficult problems in
classification.  Lakes with depths of two meters or less are suscepti-
ble to take-over by macrophytes like hyacinths.  Thus they could quickly
become swamps or bogs and must be near the senescent stage.  The present
trophic state of these lakes range from dystrophy (Palatka), mesotrophy
(Tuscawilla), hypereutrophy (Bivens Arm) to senescence (Alice).  It is
not clear whether a lake like Tuscawilla has already been through a
eutrophic stage and is now in a decreased state of production heading
toward senescence or it never reached eutrophic conditions and will pass
into senescence without doing so.  If lake classifications are to imply
anything regarding a lake's ontogeny, and its probable future develop-
ment, it is important that such questions be resolved.  A corollary to
the problem of shallow lakes is the problem of lakes with extensive
shallow areas but some deeper holes.  For example, Newnan's Lake has
large areas with a depth of 1 meter or less but has several small regions
nearly 4 meters deep (mean depth is about 2 meters).  The shallow areas
could give rise to extensive growths of rooted or attached macrophytes
and develop into swamp or bog, which would indicate senescence, but the
deeper areas could remain lacustrine for much longer.  In fact, consider-
able difficulty has been encountered in controlling water hyacinths
(not an attached plant) in recent years, which implies the lake, or a
large portion of it could become senescent in the near future.

       The data in Table VII-3 to VII-6 show a number of interesting
correlations.  Figure VII-3 indicates a positive correlation between
sodium and chloride concentrations in the 33 lakes, implying a common
origin for the two ions.  However, the estimated line of best fit de-
viates significantly from the theoretical line for NaCl dissolution with
an excess of sodium over chloride.  Most of the chloride presumably
originates as sodium chloride, probably from atmospheric precipitation
(from marine aerosols) and cultural sources.  But other salts and
weathering of clays must also act as sources of sodium.  High sodium
and chloride concentrations correspond with lakes having known pollu-
tion sources (e.g. Alice, Bivens Arm, and Hawthorne).

       Calcium and bicarbonate alkalinity are highly correlated (Figure
VII-4).  The slope of the line of best fit approaches the theoretical
line for dissolution of calcium carbonate:

                     CaC03 + C02 + H20 > Ca+2 + HCO,
                               -107-

-------
     18


     16


     14


     12
E
&-   10
       Theoretical     /
        for NaCI     ,
       r  /
    175
    150
 |  10°


2
<   50

     25

      0
10
                               8     10     12
                                Sodium ^ppin;
                                Figure  VII-3
                               14
20    30
                               40     50     60
                              Calcium (ppm)
                              Figure  VII-4
                                           16    18
70    80
                        -108.

-------
 implying  this  to  be  the  source  of calcium and alkalinity  in  the  lakes.
 Most  of the  lakes have low concentrations of these  ions;  only nine  lakes
 have  calcium concentrations  of  10 ppm  (as Ca) or greater.  The major
 natural sources of water for the lakes  in the county are  atmospheric
 precipitation, and surface and  sub-surface runoff through perched water
 tables in the  sandy  soil; hence the  low calcium concentrations.  High
 calcium suggests  a lake  receives ground water from  the Floridan  aquifer
 (a  limestone stratum), which means the  lake is either spring-fed or re-
 ceives waste water from  supplies using  well water from the aquifer.
 Most  of the  lakes with high  calcium  have known sources of waste  water
 (e.g. AlicejBivens Arm,  Clear,  Kanapaha, and Hawthorne).  Since  high
 nutrient  concentrations  are  associated  with such waters,  these lakes
 are highly enriched.  Thus for  most  of  the lakes in Alachua  County,
 calcium ion appears  to be a  good indicator of cultural eutrophication.
 The relatively high  calcium  content  of  Lochloosa Lake probably results
 from  Magnesia Spring, which  drains into Lochloosa Creek about  5  miles
 north of  the creek's confluence with the lake.

       There was  little  apparent correlation between total nitrogen and
 total phosphate in the December, 1968,  sampling of the 33 lakes  as indi-
 cated by  the scatter diagram in Figure  VII-5.  However, lakes  with the
 most  extreme N/P  ratios  are  seriously polluted (e.g. Alice and Kanapaha)
 and there  are some indications  that  lakes in a particular region have
 similar N/P ratios.  For example, lakes in the Santa Fe River  headwaters
 (Altho, Santa Fe,  Little Santa  Fe) have similar ratios which are gener-
 ally  higher than  the larger  lakes in the Orange Creek Basin.   The dashed
 line  in Figure VII-5 indicates  the usually quoted optimum N/P  ratio
 of 15:1 by atoms  for algal growth.   Points below the line suggest an
 excess of  nitrogen relative  to phosphorus; most of the lakes belong in
 this category.

       Chemical oxygen demand (COD)  and color show no correlation for
 the 33 lakes (Figure VII-6).  Evidently there is a considerable  varia-
 bility in  non-colored organic matter in the waters.   The highest color/
 COD ratio was 17  (Lake #26).  Eight  of  the lakes have ratios of  10-13
 and the remainder  of the waters have lower ratios, suggesting  large
 amounts of non-colored organic matter.  There is some evidence that low
 color/COD  ratios  indicate organic pollution; for example, Bivens Arm
had the lowest color/COD ratio  (1.15), but other apparently unpolluted
 lakes (e.g. Palatka)  also had low ratios (2.6).

       A plot of  turbidity versus the inverse of Secchi disc reading
 (Figure VII-7) also shows considerable  scatter,  implying other factors
control Secchi disc visibility.   Inspection of the data suggest color
has an important effect;  Trout Lake,  with a color of 990 ppm, had the
highest ratio (0.93)  of  (Secchi disc)1/turbidity.  A multiple regression
of Secchi disc against turbidity and  color,  after removal of Secchi disc
 readings affected by shallow bottoms, would probably account for most
of the variance.
                                -109-

-------
  I 0.6
0  0.5
s.
o


O
   0.4
    o.:
    o.
          •

         Alice
                       *   X
                       X
                             X
                 X
           x   • •   •
              •  /•./••
                                                        X
                                                     X
                                                   «
                                           X
                                              X
                                      X
                                         X
            0.5    1.0    1.5     2.0    2.5    3.0    3.5    4.0

                          Total Nitrogen (ppm as N)

                          Figure VII-5
                                                               Kanapaha i
                                                             4.5    i.O




£
*rt
O
§.
£.
7
6







800

700
600
500


400

300


200

100

)
J990

• •
•••
_ • • •


"" •
• •
•
"" • •
• ' •
• •
•
• • •
.
*
• • i i tilt 	 1 	 1 	 1 —
° 10 20 30 40 50 60 70 80 90 100
                                       COD (ppm)
                                       Figure VII-6
                                 -no-

-------
«    2.0
     1.6
.o
'u>

>

8
u
o
V
     1.2
              Trout L.
                   .  . /
                   • v
                         6      8     10    12


                         Turbidity (ppm as SFCfc)

                         Figure VII-7
                                                 14

>H
-•1
X
|
iry Production
E
£




160
140
,»
100
80
60
40

20

0
-
(235)
.

"
-
(3.84)
•MM
*
* . . % * * *
1*1 L 4. i i i i ii i
0.5 1.0 1.5 2.0 2.5
                                        Ton (mg/l)

                                        Figure  VII-8
                             -111-

-------
       A wide range of primary production and chlorophyll a data are
encountered in the 33 lakes.  Rates of the former ranged from negligible
(Lake Alice) to 235 mg C/m^-hr (Lake #21).  In order to facilitate com-
parison of rates in the lakes, all incubations were carried out under
constant light and temperature conditions in a laboratory incubator.  In
general, highest rates of primary production are associated with lakes
showing eutrophic chemical characteristics.  However, there does not
seem to be any simple quantitative relationship between primary produc-
tion values and concentrations of nutrients.  For example, Figure VTI-8
shows the scatter which occurs when primary production is plotted vs
total organic nitrogen.  Plots of primary production vs inorganic nitro-
gen and phosphate show similar scatter.

       There is also no simple relationship between primary production
and depth (Figure VII-9), as Rawson (1955) found for the Great Lakes
and some Canadian lakes he studied.  There is some uncertainty concern-
ing the meaning of depths used to plot Figure VII-9 (see discussion
above).  Since bathytnetric maps are not available for these lakes, mean
depths are unknown; maximum depths for most lakes are probably somewhat
greater than the values used here.  Some of the variance may be removed
with accurate measures of mean depth, but primary production values
would still seem to be grouped into classes rather than along any linear
or curvilinear relationship.  The horizontal dashed line in Figure VII-9
represents an arbitrary division between high and low productivity lakes.
Points above this line correspond with lakes classed as mesotrophic or
eutrophic.  The dashed vertical line represents an arbitrary division
between deep and shallow lakes; most of the lakes would be classified
as shallow productive or shallow unproductive on this basis.

       Chlorophyll a_ concentrations also show a qualitative correlation
with other criteria for trophic state (i.e., high chlorophyll a_ con-
centrations are associated with lakes classified as eutrophic or meso-
trophic).  However, quantitative correlations with nutrient concentrations
are not evident.  The relationship between chlorophyll a_ and primary
production also shows considerable scatter (Figure VII-10), although
there would seem to be a significant correlation between the two para-
meters.
       Lakes in the Oklawaha River Basin.  The region northwest of
Orlando consists of rolling hills occupied by thousands of acres of
citrus trees.  The area has an abundance of lakes situated in its val-
leys, including a chain of large lakes in the southern part of the
Oklawaha River Basin.  These lakes have been popular for recreation and
in the past at least one (Lake Apopka) was famous nationally for bass
fishing.  The quality of several of the lakes has declined radically in
recent years and concern has been expressed regarding the cultural eu-
trophication of the entire chain.  Implicated as an important nutrient
                                -112-

-------
I4U
»
CO
5 100
f
^ 80
c
*r
o
•3 60
ct
1 4°
£
Oft
ZU





250'
1
^ 200
-f
CO
§ 1?5
f
1 15°
•g 125
<£
0s 100
et
75

50



25
0
C
(235)
t

-
•.
*

-

I
•
t_^w
•
1 . ^ ' 	 	 (22)
i / * i f i i i i i i
12 34567891-
Depfh (Meters)
Figure VII-9

-



x
s
/
•'
X
xx
x«x
• /
• X"
. x '• •
X
• X-
X .
X
• xx? *
.••%••! i 1 I ? | | | |
) 4 8 12 16 20 24 28 32 36 40
     Chlorophyll a (Mg/M3)
      Figure VII-10
-113-

-------
source is agricultural runoff-primarily from vegetable farms  in reclaimed
marsh and wetlands, but also from citrus groves; domestic waste effluent
from towns on the lakes' shorelines and waste from citrus concentrate
processing plants are also important nutrient sources.

       The location of the lakes included in this survey is shown  in
Figure VII-11 and physical data on the lakes is summarized in Table VII-7-
                               TABLE VII-7
           PHYSICAL CHARACTERISTICS OF LAKES IN OKIAWAHA BASIN
Lake
Apopka
Dora
Eustis
Griffin
Harris
Weir
Area
(ha)
12,200
2,060
2,740
3,455
7,090
2,420
Depth
(m)
2
3.5
5
3
5.5
7.5
Surface
Temp.
°C
19.8
16.7
24.9
19.8
22.9
25.7
Bottom
Temp.
°C
19.6
16.6
24.7
18.9
20.0
24.0
Lake2
Type
HE
HE
E
E
E
M-E
 Maximum depth found during sampling.  In the case of Lake Apopka,
 this value also represents the mean depth; maximum depth of this lake
 is 6 m in a small hole near the southern shore  (see Schneider and
 Little, 1968) for a bathymetric map).  A map of Lake Weir is also
 available (Kenner, 1964) and indicates a maximum depth of 10 m in a
 small hole near the southern shore.

2
 See Table VII-1 for list of abbreviations.
Lake Apopka is the largest of these lakes, and also the first lake  in
the Oklawaha chain.  This lake is probably the most famous of the chain*
but is now considered as the furtherest advanced in the eutrophication
process.  The lake has been the subject of numerous investigations and
reports (see Burgess, 1964; Huffstutler e_t al^ 1965; Schneider and
Little, 1968 for details on the recent history of the lakes).  Lake
Apopka drains into Lake Dora through the Beauclair Canal, and the
latter lake drains into Lake Eustis.  Lake Harris also drains into Lake
Eustis, which drains into the last important lake of the chain, Lake

-------
                                                101  234
                                                     Miles
                                     LAKE J0EUSTIS
                                    EUSTIS
                                              LAKE  A POPKA

                                                    34

                                               \
LEESBUR6
                                                   WINTER GARDEN
    Figure VII-11:     Location of Lakes in Lower Oklawaha River Basin.
                               -115-

-------
Griffin.  While not directly in the chain, Lake Weir, about 10 miles
northwest of Lake Griffin, is within the Oklawaha drainage basin.  None
of these lakes has received the attention accorded to Lake Apopka, but
some prior studies are available.  The Florida Game and Fresh Water Fish
Commission is performing a detailed study on the trophic states of
LakesApopkas Eustis, and Weir, and their data will be useful in develop-
ing and analyzing trophic criteria.

       The six lakes were sampled for the chemical and biological
parameters shown in Table VII-2.  The results are summarized in Table VII"
8.  The trophic characters of the lakes are revealed quite unambiguously
by these initial results.  All the lakes in the Oklawaha chain are eu-
trophic.  Lake Weir is in considerably better condition than the lakes
in the chain, but it is on the border between mesotrophy and eutrophy.
Lakes Apopka and Dora have the most advanced cases of eutrophy; all
vant chemical and biological data indicate a high degree of enrichment
and productivity.  The shallowness of Lake Apopka implies the lake could
pass from hypereutrophy into senescence if macrophytes (e.g. hyacinths)
are allowed to take over.  Lakes Eustis and Griffin present slightly
improved characteristics compared to the first two lakes.  The chemical
and biological conditions of Lake Harris are the best of the five lakes
in the chain.  The trophic criteria definitely indicate a eutrophic
state, but perhaps not the hypereutrophy of the other lakes.  Perhaps
some of the water quality improvements in Lakes Eustis and Griffin re-
sult from dilution of enriched Lake Dora effluent by relatively nutrient
poor Lake Harris water.

       The water chemistry of Lake Weir differs considerably from the
other lakes.  Alkalinity and hardness are low in Lake Weir and pH is
near neutrality.  Major cations and anions are moderate to low and a
specific conductance of 135 jjmho cm-1 reflects this fact.  Lakes in the
chain have high alkalinities and hardness and pH values near 9 or above-
Major ion concentrations are high and specific conductance ranges from
230-335 ymho cm"1.  The correlation between high trophic level and high
concentrations of major ions in these lakes and relatively low trophic
level and low ionic content in Lake Weir might imply that eutrophica-
tion involves an increase in dissolved solids and a change from soft to
hard water.  In fact dissolved solids has been considered a trophic
criterion and was used as such by Beeton (1965) and others.  A certain
degree of correlation would be expected between increasing nutrients
and increasing dissolved solids where enrichment results from domestic
waste effluent and the town's water supply is hard ground water.  This
is the case with Lake Apopka, where the city of Winter Garden obtains
its drinking water from wells.  The meager background chemical data on
the lake indicate some increases in dissolved solids and hardness in
the period of record, but the lake water was relatively hard and high
in dissolved solids over 40 years ago.  Black and Brown (1951) reported
a chemical analysis of Lake Apopka from August, 1924, made by the U. S.
Geological Survey.  Table VII-9 compares their results for major ions
with the results from this study.  While the lake was probably quite
productive in 1924, serious degradation in water quality has occurred
                              -116.

-------
                            TABLE VII-8


                 CHEMICAL AND BIOLOGICAL CONDITIONS
                  IN OICLAWAHA LAKES, OCTOBER,  1968'

Constituent2  Apopka     Dora	Harris    Eustis
Griffin     Weir
Diss. 0,
Cond,
PH
Alkalinity
COD
Color
Secchi Disc
Turbidity
Sus. solids
TON
NH,-N
Ortho PO4
N08-N
NOa-N
Total P
Si03
Fe (total
Mn (total)
Ca«
Mr2
Na*
K.*
ci-
SO4-
Chora
Prim. prod.
11.2
330
9.5
126
113
30
0.3
27
43
3.8
0.18
0.016
0.004
0.09
0.24
6.4
T
T
25.1
14.7
11.7
2.9
21
10.2
34.1
0.386
11.4
335
9.5
130
157
43
0.3
28
77
5.3
0.17
0.108
0.005
0.11
0.39
0.05
0.01
0.003
30.5
14.2
18.1
3.4
22.3
12.3
72.1
1.02
9.7
230
8.8
91
48
13
0.8
20
13
1.7
0.19
0.010
0.004
0.05
0.035
3.4
0.01
0.006
23.1
6.1
9.7
0.9
12.8
4.9
17.7
0.037
8.78
275
8.9
104
112
28
0.6
25
46
4.1
0.20
0.014
0.006
0.24
0.22
1.4
0.01
0.005
26.3
10.7
14.5
2.4
18.7
7.1
33.1
0.274
8.4
290
8.9
112
78
25
0.5
28
27
3.0
0.18
0.014
0.007
0.09
0.20
0.40
0.01
0.003
28.9
9.6
14.0
2.1
17.5
6.7
45.4
0.183
7.60
135
7.2
15.1
25.
5
1.2
3
22
0.97
0.25
0.006
0.004
0.05
0.021
0.34
T
0.005
1.4
3.1
15.5
1.0
26.8
4.8
8.4
0.011
1 Results are average values for composite samples from three stations in each
 lake.

* Chemical species in mg/1 except as follows: alkalinity in mg/1 as CaCO3, nitro-
 gen .species in mg N/l, phospnorus  species  in mg P/l,  color in mg/1 as Ft.
 specific conductance in /anno cm-i, Secchi 'disc visibility in meters, chlorophyll
 a in mg/ms.

•Primary production in mg C fixed/1-hr.  Composite lake-water samples were
 run in a laboratory incubator.
                              -117-

-------
                               TABLE VII-9
      COMPARISON OF MAJOR IONS IN LAKE APOPKA IN 1924 AND 1968
Constituent
Total dissolved solids
Calcium
Magnesium
Sodium and Potassium
Iron
Bicarbonate
Sulfate
Chloride
Silica
19242
129
18
8.3
11
0.17
89
4.5
17
6.7
19683
184
25.1
14.7
13.4
T «0.01)
152
10.2
21
6.4
 Results in mg/1; potassium expressed in terms of sodium (Na).

2
 From Black and Brown (1951); original data by the U.S. Geol. Surv.
 (Water Supply Paper 596-G), Washington, D.C. (Collins and Howard,
 1927).

3
 Data from this study.
only in the last 15-20 years.  Thus increasing hardness and dissolved
solids would not seem to be a necessary feature of cultural eutrophi-
cation.  Quite possibly eutrophication in Lake Weir may proceed along a
different course not involving increases in solids and hardness--as has
apparently been the case with lakes in the Orange Creek Basin (e.g.
Newnan's Lake).

       It is difficult to draw firm conclusions regarding trophic state
from one set of samples.  Various parameters may respond to seasonal
changes differently in each lake.  With this limitation in mind, we
would tentatively systematize the trophic conditions of these lakes in
the following sequence:
                                 -118-

-------
       Lake Apopka = hypereutrophic—near senescence

       Lake Dora = hypereutrophic

       Lakes Bustis and Griffin = eutrophic

       Lake Harris = moderately eutrophic

       Lake Weir = mesotrophic to moderately eutrophic

The above sequence is somewhat relative since it is difficult to compare
lakes  in different regions in a quantitative manner.  Eutrophy in Lake
Apopka is considered more advanced than in Lake Dora because of the ex-
treme  shallowness of the former lake.  The chemical and biological
results are similar for the two lakes; in fact Lake Dora had more ex-
treme  values for some trophic criteria.  But it is somewhat deeper and
because of this perhaps more amenable to restoration.
       Conclusions.  The results obtained in this phase of the
eutrophication study substantiate the complexities involved in defining
and quantifying lacustrine trophic states.  The chemical and biological
information obtained on each lake has enabled us to classify the lakes
qualitatively as oligotrophic or eutrophic, etc.  The correlations among
different trophic indicators are generally good for these lakes--on a
qualitative basis, but simple regressions of one criterion versus another
in all cases showed large amounts of scatter or unexplained variance.
Trophic state and the interrelations among trophic criteria are multi-
variate functions.  Quantification of these phenomena requires more
sophisticated mathematical analysis than those employed above.  Stochas-
tic modeling procedures seem to be particularly appropriate to the
problem of quantifying trophic indices with masses of data collected
from different lakes.  Multivariate techniques, such as canonical corre-
lation, multiple discriminant analysis and factor analysis (principal
component analysis), were used to synthesize quantitative measures of
trophic state for sub-tropical lakes.  This is discussed in the following
section.
Quantitative Trophic State Study

       In establishing the trophic state index the apparent response
of lakes to nutrient enrichment has been envisioned as a multi-dimen-
sional hybrid concept described by several physical, chemical and
biological trophic state indicators and appropriate multivariate sta-
tistical methods have been used to analyze the relationships between
nutrient enrichment and trophic state.  One aspect of these studies was
concerned with formulating a Trophic State Index (TSI) that would pro-
vide a means for quantifying trophic state on a numerical scale.  Such
                               -119-

-------
an  index  is desirable  for  the  following reasons:    (1) a  TSI would
facilitate identification  and  comparison  of  lakes,  (2) in the dynamic
process of trophic state change, a TSI provides a method  for deter-
mining the trophic state of a  lake at any particular  time and evaluation
of  the index over time should  provide information on  the  direction and
rate of lake succession, (3) a TSI should be used in  quantifying  the
response  of the  lake ecosystem and its environment  (watershed conditions
influencing nutrient enrichment).  Such relationships should in turn be
useful in lake-basin management models, where  it may  be of particular
interest  to know how much  nutrient enrichment will  cause  significant
change in lake water quality.

       In formulating  the  TSI  seven trophic  state indicators were
considered as describing the concept of trophic state; viz. primary
production (PP), chlorophyll a (CHA.), total  phosphorus (TP), total or-
ganic nitrogen (TON), conductivity (COND), inverse  Secchi disc transpar-
ency (1/SD), and the inverse of a cation  ration  fCa  + Mg.  (1/CR)
                                                 W  + K  '
(after Pearsall, 1922).  Inverses of the  last two parameters were used
so  that each indicator exhibited a positive  response  to eutrophication
(i.e. the indicator value  increases with  increasing eutrophy).  Data for
these indicators had been  obtained from an extensive  one  year study of
the chemical, biological,  physical, and morphometric  characteristics of
fifty-five Florida lakes contained within the areas indicated in Fig-
ure VII-12.  The means, standard deviations, and the  correlation matrix
between the seven indicators are summarized  in Table  VII-10.  A princi-
pal component analysis was performed on the correlation matrix and the
first principal component  extracted.  The first principal component was
modified  by adding a constant  to prevent negative values  and interpreted
as  the Trophic State Index.  The method of principal  component analysis
is described by Morrison (1967).  The formula for evaluating TSI is
given by:

       TSI - .919(1/SD) +  .800(COND) + .896(TON) +  .738(TP) +

             .942(PP) + .862(CHA) + .635(1/CR) + 5.190             (1)

where the terms in parentheses refer to standardized  indicator values
since the units of the indicators differ.  Standardization involves
the subtraction of the mean value of the  indicator  from the raw data
value and dividing by the  indicator standard deviation.   The appropriate
means and standard deviations are from Table VII-10.  Calculation of
the TSI then is a straight forward process of substituting the stan-
dardized  trophic state indicator values for a lake  into equation (1).
The TSI's were calculated  for  fifty-five Florida lakes and ranked in
descending order of magnitude  in Table VII-11.  In addition, the TSI's
of Lake Tahoe, Lake Erie, and  Lake Superior were calculated and in-
serted in Table VII-11.  Secchi disc values  influenced by organic color
were color corrected by a multiple regression procedure not described
here.
                               -120-

-------
                      30° -00'
                           STUDY AREAS
                                             .EGENO
                                              STUDY AREA
VICINITY  MAP  SHOWING  MAJOR
     LAKE  STUDY AREAS
                                                     BEACH
                                                        VILLC
                     FIGURE VII-12

   LOCATION  MAP SHOWING MAJOR  LAKE  STUDY AREAS
                         -121-

-------
                               TABLE VII-10

            MEANS, STANDARD DEVIATIONS, AND CORRELATION MATRIX
                    FOR SEVEN TROPHIC STATE INDICATORS

                                                      Standard
         o
Indicator                     Mean                    Deviation
  L/SD                          .84                      .77
  COND                        93.1                    101.3
  TON                          1.02                      .82
  TP4                           .125                     .177
  PP                          44.8                     82.3
  CHA                         16.9                     19.8
  1/CR                         1.47                     1.63
Correlation Matrix:

         1/SD      COND      TON       TP4       PP        CHA       1/CR

1/SD     1.000      .617      .880      .542      .927      .784      .502
COND               1.000      .582      .762      .654      .540      .560
TON                          1.000      .500      .890      .788      .474
TP4                                    1.000      .576      .553      .440
PP                                               1.000      .859      .478
CHA                                                        1.000      .402
1/CR                                                                 1.000
a
 Key to indicator abbreviations in text
                                   -122-

-------
                               TABLE VII-11

    FLORIDA LAKES RANKED WITH RESPECT TO THE TROPHIC STATE  INDEX (TSI)
             Lake
                           TSI
(a)   HYPEREUTROPHIC GROUP    (> 10.0)

     Lake Apopka"(34)a         22.1
     Lake Dora (35)            18.5
     Unnamed 20 (20)           18.5
     Biven's Arm Lake (23)     14.7
     Lake Griffin (38)         13.7
     Lake Kanapaha (28)        13.5
     Lake Alice (22)           10.7
     Lake Bustis (37)          10.5

(b)   EUTROPHIC GROUP         (7.0-9.9)

     Lake Erie*                 9.8
     Lake Hawthorne (8)         9.1
     Clear Lake (24)            8.8
     Burnt Pond (31)            8.3
     Lake Wauberg (32)          7.4
     Lake Newnans (17)          7.1

(c)   MESO-EUTROPHIC GROUP    (4.0-6.9)
     Unnamed 25 (25)
     Lake Harris (36)
     Unnamed 27 (27)
     Cooter Pond (5)
     Lake Lochloosa (14)
     Lake Tuscawilla (33)
     Calf Pond (19)
     Orange Lake (15)
     Lake Mize (18)
                           6.4
                           6.3
                           5.8
                           5.3
                           5.2
                           4.8
                           4.6
                           4.3
                           4.2
(d)  OLIGO-MESOTROPHIC GROUP (2.0-3.9)
Lake Superior*
Watermelon Pond  (29)
Lit. Orange Lake  (9)
Lake Weir (39)
Palatka Pond  (16)
Lake Elizabeth (6)
Unnamed 10 (10)

aLake number  in parentheses

 Included for comparison
                                3.7
                                3.6
                                3.4
                                3.3
                                3.2
                                3.2
                                3.2
              Lake            TSI

     Seville's Pond (26)      3.1
     Lake Meta (21)           3.1
     Lake Jeggord (12)        2.8
     Long Pond (30)           2.8
     Lake Moss Lee (11)       2.8
     Lake Clearwater (7)      2.6
     Lake Altho (4)           2.5
     Hickory Pond (3)         2.5
     Lake Santa Fe (1)        2.5
     Lake Suggs (51)          2.3
     Lake Lit. Santa Fe (2)   2.3
     Lake Adaho (48)          2.2
     Wall Lake (46)           2.1
     Lake Winnott (53)        2.0

(e)   OLIGOTROPHIC GROUPS    (< 2.0)

     Still Pond (13)          1.9
     Kingsley Lake (40)       1.9
     Lake Geneva (44)         1.8
     Lake Tahoe*              1.7
     Lake Gallilee (55)       1.6
     Lake Anderson-Cue (50)   1.5
     Swan Lake (45)           1.5
     Lake Brooklyn (43)       1.5
     Lake McCloud (49)        1.5
     Cowpen Lake (54)         1.5
     Long Lake (52)           1.3
     Sand Hill Lake (41)      1.3
     Magnolia Lake (42)       1.3
     Lake Santa Rosa (47)     1.3
                                 -123-

-------
       In order to determine logical groups of lakes whose members
possess similar trophic state characteristics the fifty-five lakes were
subjected to a cluster analysis considering the seven trophic indicators.
The principles and methods of cluster analysis are described by Sokal
and Sneath (1963).  The hierarchal clustering patterns (dendrograms) are
shown in Figure VII-13.  One dendrogram is for the colored lakes and the
other for the relatively clear lakes.  (It was found for Florida lakes
that fundamental distinctions could be made between lakes on the basis
of organic color as suggested by Hansen (1962)).  The value of the ob-
jective function on the abscissa of the dendrograms is used as a measure
of similarity.  In general, an increase in the value of the objective
function at which a function is made represents a decrease in similarity
among lakes of the newly formed group, thus lakes of greatest similarity
are joined first and so on.

       Several groups of lakes were interpreted in terms of classical
trophic-state nomenclature as:   Groups 1, 2, and 3 combined - oligo-
mesotrophic, Groups 4, 5, and 6 combined meso-eutrophic, Group A -
oligo-trophic, Group B - mesotrophic and Group C - eutrophic.  The ver-
tical dashed lines in Figure VII-14 were used for the discriminant
analysis phase of the studies.

       Using the groups of lakes delineated by the cluster analysis,
five tentative TSI ranges were separated and interpreted.  These ranges
and groups are indicated in Table VII-11.

       Some features of the TSI ranking in Table VII-11 deserve mention.
Lake Alice, a lake with profuse growth of water hyacinths, has been
ranked as hypereutrophic whereas on the basis of plankton productivity
alone, it may well be considered as being oligotrophic.  However, in
the TSI, the high nitrogen, phosphorus, and conductivity indicator values
were sufficient to counteract the low primary production and chlorophyll
a indicator values.  Anderson-Cue Lake (as previously discussed) has
been subjected to artificial nutrient enrichment for three years, but
has demonstrated little response with respect to the seven trophic state
indicators, with the exception of chlorophyll a.  As previously pointed
out the lake has responded to increased periphyton growths.  Because none
of the seven indicators are sensitive to this response, the lake still
registers a low TSI (1.5), which is identical to the TSI of McCloud
Lake (the control).

       The TSI formulated for Florida lakes appears to apply reasonably
well in ranking other lakes; for example,  Lake Tahoe, Lake Erie, and
Lake Superior have been positioned where one might expect them to fall*
The TSI, undoubtedly requires further testing and refining and it may
well be that some trophic state indicators could be added or deleted
from the index formula [Equation (1)] .  However, the multivariate
methods used for deriving the TSI and for grouping lakes seem most lo-
gical and hold considerable potential as tools for studying other aspects
of the eutrophication process.
                                -124-

-------
                                                      VALUE OF THE OtUECTIVE FUNCTION
                                                       -6OO   -500  -10O  -300 _ -SCO
(41) SANO HILL
(42) MAGNOLIA
(SlJGAtLICE
( ADAliO
150 suces
(9) LIT. ORANGE
(t2) JEGCORO
(31) BURNT
CS3) TUSCftWILLA
(5) COOTER
(19) CALF
(1£) PAUATKA
 t>0) LOWS POHO
 (zc) SEVILLE'S
(27J HO. £7
 (17) NEV/NANS
 08) HIZE
 (201 KANAPAMft
                                  FIGURE. VIJ-13

   DENDOGRM OF COLORED  (UPPKR)  AND CLEAR  (LOWER)  LAKES CLUSTERED
                WITH RESPECT TO  SEVKN  TROPHIC INDICATORS
                                     -125-

-------
  10.0
 Z
 O
 e

 o

 I
   1.0
 i>>
 o
 o
 t
                                    EUTROPHIC '
       (VOUEWYEIDE.R)
                      KE-KESO- EUTROPHIC
                      f EUTROPHIC
                      HE-HYPER  EUTROPHIC
             1.0
                VEAN  DEPTH (METERS)
                                        	|
                                         10.0
                                                   LO
                                                  O
                                                  o

                                                  s
                                                  g
PHOSPH
                                                   JO I
                                                             Zi>C)
                                                                               li(MQ
                                                                                   EUTROPHIC LAKES
                                                                                    (FLORIDA)
                                                                          0- OLICOTROPHIC
                                                                          OM-OLIGO • UESOT ROPH 1C
                                                                          ME-MESO-EUTROPHIC
                                                                          HE-HYPEft EUTROPHIC
                                                       -OLICOTROPHIC LAKES
                                                         (VOLLENWEIOER)
                                                              1.0
                                                                                          10.0
                                                                MEAN DEPTH (METERS)
FIGURE VII-14:
ANNUAL NITROGEN  (UPPER)  AND PHOSPHORUS  (LOWER)  LOADING
RATES  VERSUS  MEAN DEPTH FOR THE FIFTY-FIVE  LAKES
                                              -126-

-------
Trophic State and Nutrient Budget Relationships

       In another phase of the eutrophication studies, partial nitrogen
and phosphorus budgets were calculated for each of the fifty-five lakes.
Literature values for the relative contributions of nitrogen and phos-
phorus from the various sources were used in the budget calculation.
[For a detailed description of the calculations see Shannon (1970)]  .
As a whole hypereutrophic and eutrophic conditions were associated with
high nitrogen and phosphorus loadings.  The relationships between mean
depth, lake trophic state and nitrogen or phosphorus loadings are ex-
pressed graphically in Figure VII-14.  The approach is identical to that
taken by Vollenweider (1968) and his oligotrophic and eutrophic areas
are labeled and denoted by solid lines.  Lake coordinates with respect
to mean depth and nitrogen or phosphorus loading are denoted by the
lake number (Table VII-11) and each lake's trophic state (from TSI
ratings of Table VII-11) is given in parentheses.  It appeared that
Vollenweider's critical loading areas were too low and that Florida lakes
were able to assimilate more nitrogen and phosphorus than suggested by
Vollenweider before becoming mesotrophic or eutrophic.  Accordingly re-
gions more applicable to Florida lakes were delimited by dashed lines and
labeled.

       As methods of evaluating nutrient budgets and determining limiting
nutrients become more concise diagrams such as those in Figure VII-14
should offer an avenue for predicting the trophic state of a lake for a
given nutrient budget.
                              -127-

-------
                              SECTION VIII

                            ACKNOWLEDGEMENTS

       Special thanks go to Mr. Carl Swisher of Melrose on whose property
the research lakes are located.  Without the availability of these lakes,
the project field work would have been impossible.  The cooperation of
Mr. Swisher will, we feel, shorten the time for the accumulation of
knowledge adequate to cope with the serious problems of rehabilitating
many eutrophic lakes in Florida.  Special thanks also go to Colonel
Harold Ashley of Melrose for his untiring efforts on behalf of the pro-
ject.  Colonel Ashley, a former member of the State of Florida Game and
Fresh Water Fish Commission is a dedicated conservationist who leaves
no stones unturned in his efforts to see that Florida's natural resources
are not destroyed.

       During the progress of this study many agencies and individuals
rendered essential and valuable assistance and advice.  Many technical
reports, papers and communications were contributed which provided im-
portant background material for the staff.  The following list illus-
trates some of the sources from which cooperation was received.
                                -129-

-------
                            FEDERAL AGENCIES
Department of the Interior

   Environmental Protection Agency
   Office of Water Resources Research
   Geological Survey

                             STATE AGENCIES

Trustees of the Internal Improvement Fund
Game and Fresh Water Fish Commission
Dr. James B. Lackey
Adm. Anthony L. Danis
Dr. John H. Davis
Dr. H. D. Putnam


Dr. P. L. Brezonik


Dr. W. H. Morgan


Dr. E. E. Shannon


Dr. Jackson L. Fox


Mr. Roger Yorton


Mrs. Zena Hodor


Mr. Arley DuBose


Mr. Frank Browne


Mr. Glen Brasington
                               CONSULTANTS
Melrose, Florida
Melrose, Florida
Gainesville, Florida
                          PROJECT STAFF MEMBERS
Environmental Engineering
Department

Environmental Engineering
Department

Environmental Engineering
Department

Environmental Engineering
Department

Environmental Engineering
Department

Environmental Engineering
Department

Environmental Engineering
Department

Environmental Engineering
Department

Environmental Engineering
Department

Environmental Engineering
Department
                                -130.

-------
Mr. Thomas Salmon
Mr. Bill Van Veldhuisen
Mr. Samuel Richardson
Mr. Roger King
Mrs. Jeanne Dorsey
Mrs. Terrie Woodfin
Mrs. Effie Galbraith
Miss Shirley Jordan
Mr. Truman Perry

Dr. Roy Me Ca Id in


Prof. Thomas deS. Furman


Dr. H. K. Brooks

Dr. Paul Maslin


Dr. Karolyn Maslin


Dr. J. H. Davis
                                 OTHERS
Environmental
Department

Environmental
Department

Environmental
Department

Environmental
Department

Environmental
Department

Environmental
Department

Environmental
Department

Environmental
Department
Engineering


Engineering


Engineering


Engineering


Engineering


Engineering


Engineering


Engineering
Melrose, Florida

Environmental Engineering
Department

Environmental Engineering
Department

Geology Department

Chico State College
Chico, California

Chico State College
Chico, California

Botany Department
                                -131-

-------
                                SECTION IX

                               REFERENCES

American Public Health Association.  1965.  Standard methods for the
   examination of water and wastewater, 12th ed. New York.

Ball, R. C.  1948a. Fertilization of natural lakes in Michigan. Trans.
   Amer. Fish. Soc. 78, 145-55.

	.  1948b. Fertilization of lakes -- good or bad?  Mich. Conserv.
   17 (9), 7-14.

Beeton, A. M.  1965.  Eutrophication of the St. Lawrence Great Lakes.
   Limnol. Oceanogr. 10, 240-54.

Berg, K.  1958.  Investigations on Pure Lake 1950-1954.  Limnologlcal
   studies on cultural influences.  Folia Limnol. Scand. 10, 1-189.

Black, A. P. and Brown, E.  1951.  Chemical character of Florida's
   waters — 1951.  State Bd. Conservation, Div. Water Surv. and
   Research.  Paper No. 6, 119 pp.

Borecky, G.  W.  1956.  Population density of the limnetic cladocera of
   Pymatuning Reservoir.  Ecology 37: 719-27.

Brezonik, P. L.  1968.  The dynamics of the nitrogen cycle in natural
   waters.  Ph.D. thesis, University of Wisconsin, Madison.

                and Lee, G. F.  1968.  Denitrification as a nitrogen
   sink in Lake Mendota, Wis., Envir, Sci. Technol.  2, 120-25.

  	 and Putnam, H. D.  1968.   Eutrophication:   Small Florida
   lakes to study the process. Proc.  17th South.  Water Resources and
   Poll, Contr. Conf., Univ. of North Carolina, April, 1968.  pp 315-333.

                 1969.  Eutrophication:   The process and its  potential
   for modeling, Proc.  Eutrophication Workshop Southeast,  St.  Petersburg,
   Fla.,  Univ.  of Fla.  Water Resources Center, Publ.  pp 290.
              ., Morgan, W. H., Shannon, E. E., and Putnam, H. D.  1969.
Eutrophication factors in north central Florida lakes.  Water Resources
Research Center.  Publ. No. 5, Bull. Series No. 134, Fla. Eng. and Indus.
Exper. Sta.  pp 56-66.

Brooks, J. L. and Dodson, S. I. 1965.  Predation, body size and composition
of plankton.  Science 150: 28-36.


                                  -133-

-------
Brooks, J. L.  1969.  Eutrophication and changes in the composition of
   the zooplankton.  Symposium on eutrophication:  causes, consequences,
   correctives.  Nat. Acad. Sci., Washington, D. C.  pp 236-255.

Burgess, J. E.  1964.  Summary report of Lake Apopka.  Florida State Bd.
   of Health, 23 pp. (mimeo).

Clark, W. E., Musgrove, R. H., Menke, C. D. and Cagle, J. W., Jr.  1962.
   Interim report on the water resources of Alachua, Bradford, Clay and
   Union Counties, Florida.  Fla. Geol. Surv., Inf. Arc. No. 36. 92 pp.

            , Musgrove, R. H., Menke, C. D., and Cagle, J. W., Jr.
   1964.  Water resources of Alachua, Bradford, Clay, and Union Counties,
   Florida.  Fla. Geol. Surv. Rept. of Invest. No. 35, 170 pp.

Collins, D. W. and Howard, C. S.  1927.  Chemical character of Florida
   waters. U.S. Geol. Surv. Wat. Supp. Paper 596-G, Washington, B.C.

Connell, J. H.  1961.  The influence of interspecific competition and
   other factors on the distribution of the barnacle Chthamalus stellatus.
   Ecology 42: 710-723.

Creitz, G. I. and Richards, F. A.  1955.  The estimation and character-
   ization of plankton populations by pigment analysis.  III. A note on
   the use of "Millipore" membrane filters in the estimation of plankton
   pigments.  Jour. Marine Res. (Sears Foundation) 14, 211-216.

Curl, H.  1959.  The origin and distribution of phosphorus in western
   Lake Erie.  Limnol. Oceanogr. 4, 66-76.

Dodson, S. I.  1970.  Complementary feeding niches sustained by size-
   selective predation.  Limnol. Oceanog. 15: 131-137.

Edmondson, W. T.  1965.  Reproductive rate of planktonic rotifers as
   related to food and temperature in nature.  Ecol. Mang. 35: 61-111.

	.  1968.  Water quality management and lake eutrophica-
   tion:  The Lake Washington case.  In water resources and public
   policy.  T. H. Campbell and R. 0. Sylvester (eds). Univ. of Washington
   Press, Seattle, pp 139-178.

Einsele, W.  1941.  Die Umsetzung von zugefuhrtem, anorganischem Phosphat
   im eutrophen See und ihre Ruckwirkungen auf seinen Gesamthaushalt.
   Z. Fisch. 39, 407-488.

Elbourne, C.A.  1966.  The life cycle of Cyclops strenuus strenuus
   Fischer in a small pond.  J. Anim. Ecol. 35: 333-347.

Feth, J. H.  1966.  Nitrogen compounds in natural waters-a review.
   Water Resources Res. 2, 41-58.

Fitzgerald, G. P.  1968.  Personal communication.  Univ. of Wis.


                                  -134-

-------
 Frank,  P. W.,  Boll,  C.  G.,  and  Kelly,  R. W.   1957.   Vital  statistics  of
    laboratory  cultures  of Daphnia  pulex DeGeer as related  to  density.
    Physiol.  Zool.  30: 287-305.

 Fruh, E. G.  1967.   The overall picture of eutrophication.  Jour. Water
    Poll. Contr.  Fed. 39, 1449-1463.

 Fryer,  G.   1968.   Evolution and adaptive radiation  in  the  Chydoridae
    (Crustacea: Cladocera):  a  study in  comparative functional  morphology
    and  ecology.  Phil.  Iran.  Royal Soc., Series B.  254: 221-385.

 Goldman, C.  R.   1961.   The  contribution of alder trees  (Alnus tenuifolia)
    to the primary  productivity  of  Castle Lake, California.  Ecology
    42,  282-288.

 	.   1964.   Primary  productivity and micronutrient limiting
    factors  in some North American  and  New Zealand lakes.  Verh. Internat.
    Verein. Lunnol. 15,  352-359.

 	.   1965.   Micronutrient limiting factors and  their detection
    in natural phytoplankton populations, p. 121-135.   In C. R. Goldman
    (ed) Primary  Productivity  in Aquatic Environments.  Mem. 1st. Ital.
    Idrobiol., 18 Suppl., Univ.  of  Calif. Press, Berkeley.

 Hall, D. J.  1964.  An  experimental approach to the dynamics of a natural
    population of Daphnia galeata Mendptae.  Ecology 45: 94-112.

 Hansen, K.  1962.  The  dystrophic  lake type.   Hydrobiologia 19, 183-191.

 Harkness, W. J. K. and  Pierce,  E. L.   1940.  The limnology of Lake Mize,
    Florida.  Quart. J.  Florida Acad. Sci. 5,  96-116.

 Hasler, A. D.  1957.   Natural and artificially (air-ploughing) induced
   movement of radioactive  phosphorus  from the muds of lakes.   UNESCO
    Int. Conf. Radioisotopes 4,  658-675.

	.  1963.   Wisconsin-1940-1961.  In Limnology in North
   America, D. G. Frey (ed) Univ. Wisconsin Press, Madison, pp 55-94.

Hazelwood, D. H. and Parker, R. A.  1961.  Population dynamics of some
   freshwater zooplankton.  Ecology 42: 266-274.

Hooper, F. F. and Ball, R. C.  1964.  Responses of a marl lake to
   fertilization.  Trans. Amer. Fish Soc. 93,  164-173.

            , Ball, R.  C. and Tanner,  H.  A.  1952.  An experiment in the
   artificial  circulation of a small michigan lake.   Trans.  Amer.
   Fish Soc.  82, 222-241.
                                  -135-

-------
Hubbs, C. L.  1921.  An ecological study of the life history of the
   freshwater artherine fish Labidesthes sicculus.  Ecology 2: 262-276.

Huffstutler, K. K., Burgess, J. E. and Glenn, B. B.  1965.  Biological,
   physical, and chemical study of Lake Apopka, 1962-1964.  Florida
   State Board of Health, 78 pp (mimeo).

Hutchinson, G. E.  1957.  Treatise on limnology. Vol. I. Wiley, New York.
   pp 1015.

	.  1967.  Treatise on limnology, Vol II.  Wiley, New
   York, pp 1115.

Ilkowska, A. H., Gliwicz, Z., and Spodniewska, I.  1966.  Zooplankton
   production and some trophic dependences in the pelagic zone of two
   Massurian lakes.  Verh. int. Verein. Theor. angew. Limnol. 16: 432-440.

Ingle, L., Wood, T. R. and Banta, A. M.  1937.  A study of longevity,
   reproduction, and heart rate in Daphnia longispina as influenced by
   limitation in quantity of food.  J. Exper. Zool. 76: 324-352.

Ivlev, V. S.  1961.  Experimental ecology of the feeding of fishes.
   Yale Univ. Press, New Haven, Connecticut.  302 pp.

Karcher, F. H.  1939.  Untersuchungen uber den Stickstoffhaushalt in
   ostpreussischen Waldseen.  Arch. Hydrobiol.  35, 177-266.

Kenner, W. E.  1964.  Maps showing depths of selected lakes in Florida.
   Fla. Geol. Surv. Inf. Circ. 40, 82 pp.

Kerr, P. C., Paris, D. F. and Brockway, D. L.  1970.  The interrelation
   of carbon and phosphorus in regulating heterotrophic and autotrophic
   populations in an aquatic ecosystem.  Proc. 25th Industrial Waste
   Conf., Purdue Univ.   (in press).

Lackey, J. B.  1945.  Plankton productivity of certain southeastern
   Wisconsin lakes as related to fertilization.  II. Productivity,
   Sewage Wks. J. 17, 795-802.

Lackey, J. B. and Lackey, E. W.  1967." A partial checklist of Florida
   freshwater algae and protozoa with reference to McCloud and Cue Lakes.
   Water Resources Research Center Publ. No. 3, Bull. Series No. 131,
   Fla. Eng. and Indus. Exper. Sta.

Langford, R. R.  1950.  Fertilization of lakes in Algonquin Park,
   Ontario.  Trans. Amer. Fish Soc. 78, 133-144.

Lee, G. F. and Fruh, E. G. 1966.  The aging of lakes.  Indust. Water
   Engin. 3 (2), 26-30.
                                 -136-

-------
McAllister, C. D.  1969.  Aspects of estimating zooplankton production
   from phytoplankton production.  J. Fish Res. Bd. Can. 26: 199-220.

McGauhey, P. H., Eliassen, R., Rohlich, G. A., Ludwig, H. F., and
   Pearson, E. A. 1963.  Comprehensive study on the protection of water
   resources of Lake Tahoe basin through controlled waste disposal.
   Prepared for the Board of Directors, Lake Tahoe Area Council, Al
   Tahoe, Calif,  pp 157.

McLane, W. M.  1955.  The fishes of the St. Johns River System. Ph.D.
   dissertation. University of Fla., Gainesville, Fla.  pp 376.

Mendenhall, W.  1967.  Introduction to probability and statistics.
   2nd Ed. Wadsworth Publishing Co., Inc. Belmont, Calif. 393 pp.

Morrison, D. F.  1967.  Multivariate statistical methods.  McGraw Hill
   Book Co., New York, pp 338.

Mortimer, C. H.  1939.  The work of the Freshwater Biological Association
   of Great Britain in regard to water supplies.  The nitrogen balance of
   large bodies of water.  Offie. Circ. British Waterworks Assoc. 21,
   1-10.

Nelson, P. R. and Edmondson, W. T.  1955.  Limnological effects of fer-
   tilizing Bare Lake, Alaska.  Fishery Bull. Fish Wildl. Serv. U.S.
   102, 414-436.

Nordlie, F. G.  1967.  Chemical and biological dynamics in two solution
   lakes.  Final Rept. to Fed. Wat. Poll. Contr. Admin. Grant No.
   WP-00530 (Mimeo).

Odum, H. T.  1953.  Dissolved phosphorus in Florida waters, Fla. Geol.
   Surv. Rept. Invest. 9, 1-40.

Paine, R. T.  1966.  Food web complexity and species diversity.  Amer.
   Nat. 100: 65-75.

Pearsall, W. H.  1922.  A suggestion as to factors influencing the dis-
   tribution of free-floating vegetation, J. of Ecology 9:241.

Rawson, D. S.  1955.  Morphometry as a dominant factor in the produc-
   tivity of large lakes.  Verh. int. Verein. Limnol. 12, 164-175.

Ricker, W. E.  1958.  Handbook of computations for biological statistics
   of fish populations.  Roger Dunamel, F.R.S.C.  Ottawa, Canada.

Rohlich, G. A. and Lea, W. L.  1949.  The origin and quantities of plant
   nutrients in Lake Mendota.  Report Univ. Wisconsin Lake Investigation
   Committee, Madison, Wis.  (mimeo, 8 pp).
                                 -137-

-------
Sawyer, C. N., Lackey, J. B. and Lenz, A. T.  1945.  Investigation of
   the odor nuisances in the Madison Lakes, particularly Lakes Monona,
   Waubesa, and Kegonsa from July, 1943 to July, 1944.  Report to
   Governor's Committee, Madison, Wis.  92 pp.

Schneider, R. F. and Little, J. A,  1968.  Characterization of bottom
   sediments and selected nitrogen and phosphorus sources in Lake Apopka,
   Florida.  U.S. Dept. of Interior, Fed. Wat. Poll. Contr. Admin.,
   Southeast Water Lab. Tech. Rept. (mimeo).

Shannon, E. E.  1970.  Eutrophication-trophic state relationships in
   north and central Florida lakes.  Dissertation presented to the
   Graduate Council of the University of Florida.

Slobodkin, L. B.  1954.  Population dynamics in Daphnia obtusa Kurz.
   Ecol. Monogr. 24: 69-88.

                  1961.  Growth and regulation of animal populations.
   Holt, Rinehart and Winston, Inc., New York.

   	.  1968.  How to be a predator.  Am Zoologist 8: 43-51.
Sokal, R. R. and Sneath, P. H.  1963.  Principles of numerical taxonomy,
   W. H. Freeman, San Francisco and London.

Stewart, K. M. and Rohlich, G. A.  1967.  Eutrophication-a review.
   Calif. State Water Qual. Control Bd., Publ. No. 34, 188 pp.

Straskraba, M.  1963.  Share of the littoral region in the productivity
   of two fishponds in Southern Bohemia.  Rospravy CSAV, rada MPV,
   73, 13: 63 pp.

Stross, R. G., Neess, J. C. and Hasler, A. D.  1961.  Turnover time and
   production of planktonic Crustacea in limed and reference portion of
   a bog lake.  Ecology 42: 237-245.

United States Department of the Interior.  1969.  Federal Water Pollution
   Control Administration.  Analytical techniques for the national eu-
   trophication research program.  Pacific Northwest Laboratory Corvallis,
   Oregon.  June,  pp 141.

Vollenweider, R. A.  1968.  Scientific fundamentals of the eutrophication
   of lakes and flowing waters, with particular reference to nitrogen
   and phosphorus as factors in eutrophication, O.C.E.D. Report, DAS/CSI
   68.27.

Ward, E. B.  1940.  A seasonal population study of pond entomostraca in
   the Cincinnati region.  Amer. Nat. 23: 635.
                                 -138-

-------
Werner, R. G.  1969.  Ecology of limnetic bluegill (Lepomis macrochirus)
   fry in Crane Lake, Indiana.  Amer. Nat. 81: 164-181.

Wetzel, R. G.  1962.  A comparative study of the primary productivity of
   higher aquatic plants, periphyton, and phytoplankton in a saline lake.
   Ph.D. Dissertation, Univ. of Calif. Davis,  pp 156.

Whiteside, M. C.  1970.  Danish chydorid cladocera:  modern ecology and
   core studies.  Ecol. Monog.  40: 79-118.

                and Harmsworth, R. V.  1967.  Species diversity in
   chydorid (Cladocera) communities.  Ecology 48: 664-667.
                                 -139-

-------
                                SECTION X.

                               PUBLICATIONS

 Shannon,  E.  E.   1970.   Eutrophication-trophic state relationships in
    north  and central Florida lakes.   Dissertation presented to the
    Graduate  Council of the  University of Florida.

 Maslin, P. E.   1969.   Population dynamics and productivity of zooplankton
    in  two sandhills lakes.   A dissertation presented to the Graduate
    Council of the  University of Florida

 Maslin, K. R.   1970.   The interactions of littoral zooplankton and their
    fish predators.   A  dissertation presented  to the Graduate Council of
    the University  of Florida.

 Lackey, J. B. and  Lackey, E. W.   1967.   A partial checklist of Florida
    fresh-water  algae and protozoa with reference to McCloud and Cue
    Lakes.  Water Resources  Research  Center.   Publ.  No.  3,  Bull. Series
    No. 131,  Fla. Eng.  and Indus.  Exper.  Sta.

 Putnam, H. D.,  Shannon, E.  E.,  and Brezonik,  P.  L.   1970.   Lake eutro-
    phication:   a critical look at the  problem.   Presented  at the
    International Symposium  on  Hydrogeochemistry and Bio-chemistry.
    Tokyo, Japan.

 Brezonik, P. L., Morgan, W.  H.,  Shannon,  E. E.,  and Putnam,  H.  D.   1969.
    Eutrophication  factors in north central Florida  lakes.   Water Re-
    sources Research Center.  Publ. No.  5,  Bull.  Series  No.  134, Fla.  Eng.
   and Indus. Exper. Sta. pp 101.

 Brezonik, P. L.  1968.  Application  of mathematical models  to the
   eutrophication process.   Proc. llth  Conf.  Great  Lakes Reg.  16-30.
    Internat. Assoc.  Great  Lakes  Res.

 Brezonik, P. L. and Putnam, H. D.  1968.   Eutrophication:   small Florida
    lakes as models to  study the process.   Proc.  17th Southern Water  Res.
   and Poll. Control Conf.

Brezonik, P.  L.  1970.  Limnological features of  north  central  Florida
    lakes.   Presented at the 33rd Annual meeting of  the Amer.  Soc. of
   Limn,  and Ocean.  Kingston, R. I.

Brezonik,  P.  L. and Shannon, E. E.   1971.  Eutrophication:  cause-effect
   relationships in Florida lakes,  submitted to  the Jour, of Sanit.  Eng.
   Di.v.,  Amer. Soc. of Civil Eng.

    •US. GOVERNMENT PRINTING OFFICE:  H7J-4M-4B3/M

                                 -141-

-------
   SELECTED WATER
   RESOURCES ABSTRACTS
   INPUT TRANSACTION FORM
         I, Report No.
3. Accession No.
                             w
   4. Title
    Eutrophication Factors In North Central Florida Lakes
   7. Author(s) putnam> H. D., Morgan,  W.  H.,  Brezonik, P. L.,
             Shannon, E. E., Maslin, P.  E.
   9. Organization
               orIda University
             Environmental Engineering Department
             Gainesville, Florida
                             5. Report Date
                             6.   .  .
                            *$,' Performing Organization
                               Report No.        ,

                            W. Project No.
                                 16010 DON
  12. Sponsoring Organization

  IS. Supplementary Notes
                            11. Contract! Grant No.
                                                                   'W.  Type of Report and ,'
                                                                       Period Covered
  16. Abstract A. small Florida lake has  been receiving a regimen of nutrient  addition
  equivalent to 500 mg/m^-yr N and 43 mg/m3-yr P since 1967.  Data has  been accumulated
  through 1969.  The effect on the lacustrine ecosystem of various biogenes includes
  production by primary producers, species diversity of plankton and  certain production
  estimates at the secondary trophic  level using natural populations  of planktivorous
  fish.   Plankton production using isotopic carbon is ca. 58 grms/m^-yr.   Species diver-
  sity  is slowly changing to a mixed  chlorophycean and yellow-green.  Biomass of benthic
  green  filamentous types has increased  slightly.  Nutrient addition  has had little in-
  fluence on zooplankton production.

            Related studies on 53 other  regional lakes have been done using a multi-
  dimensional hybrid concept as defined  by several trophic state indicators.  This
  trophic state index has provided a  means for ranking the lakes on an  arbitrary scale.
  Cluster analysis utilizing pertinent characteristics resulted in classification of
  other  lakes.

            Land use patterns and population characteristics were determined photogra-
  phically and N and P budgets estimated.   Using multiple regression  and canonical
  analysis, several significant relationships were found between lake trophic state, lake
  basin,  land,use, and population characteristics,. In general, trophic, state of lakes can
 •Be expressed go a oimplc relationonip  Incorporating N ana P influn  rotoc.	
  17a. Descriptors
            *Eutrophication, *Limnology,  *Mathematical Model, *Essential Nutrients,
            *Primary Productivity, Water  Quality,  Trophic Level, Aquatic Algae,  Fish
            Populations


  17b. Identifiers
            Anderson-Cue Lake
            Melrose, Florida
  17c. CO WRR Field & Group  05B.05C
  18. Availability
  Abstractor  Z,  K.  Hodor
                Send To:
                                                       WATER RESOURCES SCIENTIFIC INFORMATION CENTER
                                                       U.S. DEPARTMENT OF THE INTERIOR
                                                       WASHINGTON. D. C. 20240
institution Univers ity of Florida
WRSIC 102 (REV. JUNE 1971)
                                          GPO 913.281

-------