EPA 660/3-73-002
 . .  iq,-                          Ecological  Research Series
NITROGEN  SOURCES  AND CYCLING

IN NATURAL WATERS
                                     Office of Research and Development

                                     U.S. Environmental Protection Agency

                                     Washington, D.C. 20460

-------
            RESEARCH REPORTING SERIES
Research reports of the  Office  of  Research  and
Monitoring,  Environmental Protection Agency, have
been grouped into five series.  These  five  bread
categories  were established to facilitate further
development  and  application   of   environmental
technology.   Elimination  of traditional grouping
was  consciously  planned  to  foster   technolcgy
transfer   and  a  maximum  interface  in  related
fields.  The five series are:

   1.  Environmental Health Effects  Research
   2.  Environmental Protection Technology
   3.  Ecological Research
   4.  Environmental Monitoring
   5.  Socioeconomic Environmental studies

This report has been assigned  to  the  ECOLOGICAL
RESEARCH  series.   This series describes research
on the effects of pollution on humans,  plant  and
animal   species,  and  materials.   Problems  are
assessed   for   their   long-   and   short-term
influences.    Investigations  include  formation,
transport, and pathway studies  to   determine  the
fate  of  pollutants and their effects.  This work
provides the technical basis for setting standards
to  minimize   undesirable   changes  in   living
organisms   in   the   aquatic,   terrestrial  and
atmospheric environments.

-------
                                         EPA 660/3-73-002
                                         July 1973
               NITROGEN SOURCES AND  CYCLING
                     IN NATURAL WATERS
                            By

                    Patrick L.  Brezonik
                   University of  Florida
    Department of Environmental  Engineering Sciences
                Gainesville, Florida  32601

               Research Grant No. 16010-DCK
                       Project Officer

                   Dr.  Charles F. Powers
        National Eutrophication Research Laboratory
    Pacific Northwest  Environmental Research  Laboratory
                 Corvallis,  Oregon  97330
                        Prepared for

             OFFICE OF RESEARCH AND MONITORING
           U.S.  ENVIRONMENTAL PROTECTION  AGENCY
                   WASHINGTON, D.C.  20460
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price $2.35

-------
            EPA Review Notice
This report has been reviewed by the Environ-
mental Protection Agency, and approved for
publication.  Approval does not signify the
views and policies of the Environmental Pro-
tection Agency, nor does mention of trade names
or commercial products constitute endorsement
or recommendation for use.
                      ii

-------
                             Abstract
Sources of nitrogen were reviewed to determine their significance
in lacustrine budgets.  Nutrients in rainfall were found significant
although their variability obviates precise conclusions.  Using litera-
ture values for nutrient export from various land uses, nutrient bud-
gets were calculated for 55 Florida lakes.  Critical N and P load-
ing rates (above which eutrophication is likely)  were estimated from
the calculated budgets and lake trophic conditions.

Algal fixation in two eutrophic Florida lakes was studied in detail;
the total annual N fixed and factors affecting the occurrence of
fixation were evaluated.  A survey of fixation in 55 Florida lakes
showed significant fixation only in eutrophic lakes.  Bacterial fixa-
tion in the anoxic hypolimnion of a small lake contributed substantial
nitrogen to the lake, and N fixing activity was found in both estuarine
and lacustrine sediments.  The acetylene reduction assay for N fixa-
tion was evaluated; short incubations were found essential.  Reduction
was light dependent and N2 acted as a competitive inhibitor.

A preliminary experiment suggested that lacustrine sediments act as
ammonia buffers; estuarine sediment sorbed ammonia strongly with little
tendency to release ammonia to the water.  Interferences from high
organic color were evaluated for automated inorganic N and P analyti-
cal methods.  Various amino acids were also shown to interfere with
the indophenol ammonia procedure.

This report was submitted in fulfillment of Project Number 16010 DCK,
under the sponsorship of the Office of Research and Monitoring,
Environmental Protection Agency.
                                ill

-------
                             CONTENTS


Section                                                        Page

  I    Conclusions                                               1

  II   Recommendations                                           3

  III  Introduction                                              5

  IV   Review of the Nitrogen Cycle in Natural Waters            7

  V    Transport of Nitrogen Into Lakes                         21

  VI   Nitrogen Fixation as an ln_ Situ Nitrogen Source
       for Natural Waters.   I. Algal Fixation in Lakes          39

  VII  Nitrogen Fixation as an I_n Situ Nitrogen Source
       for Natural Waters.   II. Bacterial Fixation in
       Lakes and Sediments                                       63

  VIII Other In Situ Nitrogen Sources and Sinks                 93

  IX   Analytical Investigations                               109

  X    Acknowledgements                                        139

  XI   References                                              141

  XII  Publications and Patents                                157

  XIII Appendix                                                159

       A.  Sampling and Analytical Methods
       B.  Routine Data Collected on Bivin's Arm and
    N      Newnan's Lake during Nitrogen Fixation Study
       C.  Enrichment and Isolation Procedures for Nitrogen
           Fixing Agents in Lake Water and Sediments

-------
                                FIGURES
                                                                   PAGE
NO.

 1     Simplified nitrogen cycle showing main molecular
       transformations                                               °

 2     Nitrogen cycle reactions in an idealized stratified
       lake                                                          9

 3     Bathymetric map of Newnan's Lake                             44

 4     Bathymetric map of Bivin's Arm                               45

 5     Nitrogen levels, nitrogen fixation rates and primary
       production in Bivin's Arm                                    47

 6     Nitrogen levels, nitrogen fixation rates and primary
       production in Newnan's Lake                                  49

 7     Temperature, primary production and nitrogen fixation
       during Aphanizomenon bloom in Newnan's Lake                  50

 8     Areal variations  in primary production and nitrogen
       fixation  in Newnan's Lake, 8 December 1969                   53

 9     Diel variations  in nitrogen fixation, primary production
       and light  intensity in Newnan's Lake, 16 December 1969       55

 10     Diel variations  in primary production and related para-
       meters  in Newnan's Lake, 21 April 1971                        56

 11     Bathymetric map  of Lake Mize, Florida                        64

 12     Liinnological  characteristics of Lake Mary, Wisconsin         65

 13     Limnological  characteristics of Lake Mize, Florida           65

 14     Temperature profiles in  Lake Mize during 1969 and 1970       69

 15     Dissolved  oxygen profiles  in Lake Mize during 1969  and  1970  70

 16     Depth profiles  of nitrogen  fixation  in Lake Mize during
       1969 and  1970                                                72

 17     Total hourly  fixation  in Lake Mize during 1969  and  1970     74

 18     Depth distribution of  acetylene  reduction rates  in
        Waccasassa  Estuary  sediments
                                                                    77
                                   vi

-------
19     Areal variations of acetylene reduction in Waccasassa
       Estuary sediments                                            78

20     Lineweaver-Burk plot showing effect of N  on acetylene
       reduction by estuary sediments                               81

21     Lineweaver-Burk plot showing effect of N~ on acetylene
       reduction by Lake Kanapaka sediments                         89

22     Sediment C/N ratio vs.  trophic state ranking of lakes        98

23     Total organic nitrogen vs. total phosphorus in Florida
       lake sediments                                               99

24     Decline in aqueous ammonia levels in aquaria containing
       undisturbed estuary sediments                               100

25     Ammonia uptake by fresh and sterilized estuary sediments    102

26     Ammonia uptake by Bivin's Arm sediments                     105

27     Ammonia leached from Bivin's  Arm sediments                 105

28     Typical chromatogram showing elution order of acetylene,
       ethylene and other gases                                    110

29     Effect of incubation time on acetylene reduction by
       estuary sediments                                           114

30     Effect of N2 and light on time course of acetylene re-
       duction by natural population of Anabaena                   115

31     Theoretical Lineweaver-Burk plot for competitive
       inhibition                                                  117

32     Competitive inhibition of acetylene reduction in Anabaena
       population by N~                                            117

33     Inhibition of acetylene reduction in Anabaena population
       by carbon monoxide                                          119
         S
34     Lineweaver-Burk plot of CO inhibition data from Figure 33   119

35     Ethylene production by natural Anabaena population vs.
       acetylene concentration at various levels of NoO addition   120

36     Immediate effect of ammonia on ethylene production by
       Anabaena                                                    123

37     Effect of ammonia on ethylene production by Anab aena
       24 hours after ammonia addition                             124
                                    VI1

-------
38     Effect of organic color on apparent nitrate concentration
       corrected by method of blank substraction                  128

39     Effect of organic color on apparent nitrite concentration
       corrected by method of blank subtraction                   129

40     Effect of organic color on apparent ammonia concentration
       cannot be corrected by method of subtraction               130

41     Effect of increasing organic color concentration on calibration
       curves for alkaline-phenol ammonia procedure               132

42     Absorption spectra for products of alkaline phenol-ammonia
       procedure at various levels of organic color               134
                                Vlll

-------
                                TABLES
No.                                                                Page

 1     Sources and sinks for the nitrogen budget of a lake          22

 2     Nitrogen and phosphorus contents of various nutrient         24
       sources

 3     Variations in nutrient content of rainfall in Gainesville
       area                                                         28

 4     Nutrient content of rainfall in. coastal and rural areas      30

 5     Temporal variations in nutrient contents of rainshowers      31

 6     Contribution of rainfall to nutrient budget of Anderson-
       Cue Lake                                                     33

 7     Regression analyses of trophic state index (TSI) vs.
       nitrogen and phosphorus loading rates                        35

 8     Critical loading rates for nitrogen and phosphorus           36

 9     Summary of the  occurrence of nitrogen  fixation in Florida
       lakes                                                        40

 10     Nitrogen fixation rates in selected Florida lakes            ^2

 11     Chemical characteristics of Newnan's Lake and Bivin's Arm    46

 12     Areal variations in physical and biogenic parameters in
       Newnan's Lake,  14 April, 1969                                51

 13     Summary of areal and vertical variations in nitrogen
       fixation and related parameters in Newnan's Lake,
       8 December, 1969                                             54

 14     Multiple regression analysis of primary production  and
       limiting factors in Newnan's Lake during Aphanizomenon
       bloom                                                        59

 15     Contributions of nitrogen fixation to  nitrogen budgets
        of Newnan's Lake  and  Bivin's Arm
                                                                     61
                                      ix

-------
16     Nitrogen fixation rates in Lakes Mary, Wisconsin,  and Mize,       67

       Florida, in summer, 1968
17
18
19
20
21
22
23
24
25
26
27
28
29
Chemical characteristics of Lake Mize
Changes in nutrient levels of Lake Mize from 1968 to 1970
Effect of added organic substrates on rates of acetylene
reduction by estuary sediments
Acetylene reduction in Bivin's Arm sediments, August, 1968
Lake surveyed for sediment nitrogen fixation
Acetylene reduction rates in Florida lake sediments
Stratification of acetylene reduction in Florida lake sediments
Effect of organic substrates on rates of acetylene reduction
by Lake Kanapaha sediments
Sediment characteristics of north central Florida lakes
Ammonia uptake by estuary sediments in short-term shaker-
flask experiment
Replicate data on algal acetylene reduction
Replicate data on acetylene reduction by estuary sediments
Ammonia effects on nitrogenase in natural population of
Anabaena
68
71
81
84
85
86
87
90
95
103
112
113
122
 30     Application of external compensation method  of  nitrite and

        nitrate samples containing varying color  concentrations



 31     Response of two automated ammonia methods to free  amino
          . j                                                            136
        acids

-------
                             SECTION I

                            CONCLUSIONS
1.  Rainfall contributes substantial amounts of nitrogen to lakes.  The
great variability in concentrations obviates simple assessment of the sig-
nificance of this source, and experimental evaluation for each  lake seems
necessary for accurate results.  Nutrient concentrations decline rapidly
with length of rainfall, indicating the role of precipitation as an atmos-
pheric cleansing agent.

2.  Using literature values for nutrient export from various land uses
and easily evaluated watershed characteristics, one can compute nutrient
budgets for lakes that, if not totally accurate,are nevertheless meaningful
and useful in assessing the eutrophication potential of a lake.  This
approach, with refined export values would seem especially useful for small
but still recreationally important lakes, where the time and expense of
experimental nutrient budget evaluation cannot be justified.

3.  Nitrogen fixation by blue green algae in eutrophic surface waters is
of doubtless ecological significance, but in terms of contributions to
the total nitrogen budgets of lakes fixation probably represents a  rela-
tively minor source.  Fixation is found only in moderately or highly
eutrophic waters where it likely exascerbates  the problem by supplying
more nitrogen during times of relative depletion.  However, if external
nutrient supplies are  decreased below critical (eutrophying) levels, fix-
ation itself is unlikely  to cause continued problems.

4.  Bacterial fixation  in anoxic lake waters is now an established  fact,
and in some cases it can supply significant amounts of nitrogen  to  a lake
system.  Fixation in sediments is also established, and heterotrophic
bacteria have been  identified  as the responsible agents in  the sediments
studied here.  The  ecological  significance  of  this phenomenon  remains
puzzling, but the occurrence of fixation  suggests  that nitrogen  is  rela-
tively unavailable  at  least in some sediments.

5.  Non-enzymatic chemical equilibria, in particular  sorption  reactions,
play an important,  perhaps dominant role  in controlling ammonia  inter-
change betVeen sediments and water.  Sorption  of ammonia onto  sediments
is, at least in some cases, rapid and largely  irreversible.

6.  The acetylene reduction method  is a useful, precise and simple  method
for assaying nitrogen  fixation in environmental samples.  Short  incuba-
tions  (approximately one hour) should be  used  for  accurate  rate  determi-
nations.  Acetylene reduction  by algae is strongly light dependent.
Molecular nitrogen  inhibits acetylene reduction, and  a pattern of compe-
titive inhibition is obtained  from  Lineweaver-Burk plots.   This  fact is
useful in establishing the enzymatic nature of acetylene reduction  in
environmental samples.   Atmospheric levels  of  N9 decrease acetylene
reduction rates bv  about 25-30 percent, and if only approximate  results

-------
are desired, removal of Nn may not be necessary.

7.  The interference of organic color in automated nitrite, nitrate and
ortho phosphate analyses can be simply compensated for by the method of
blank subtraction, but this cannot be done with the indophenol ammonia
method.  Organic color or some constituent associated with color (e.g.
iron) causes positive interferences with this method which cannot be corrected
by eliminating a key reagent and measuring the color "blank."  Further, a
variety of  amino acids also react as ammonia in the indophenol method,
in some cases giving greater responses than equivalent ammonia-N levels.
This method should  thus be used with caution, especially when free amino
acids are suspected in a water sample.  When high color is present, a
alternate ammonia procedure would be preferable.

-------
                            SECTION II

                          RECOMMENDATIONS
1.   Various diffuse sources of nitrogen, such as agricultural runoff,
sanitary landfill sites, and urban runoff, need much further study to
assess their quantitative significance in lake nutrient budgets.

2.   The extent of denitrification in wetlands and sediments needs to be
evaluated to properly assess the importance of this reaction as a nitro-
gen sink for lakes.

3.   The relationships between nitrogen-fixing blue-green algae and other
bloom forming algae need further investigation in terms of the transfer
of  nitrogen from fixers to non-fixers.

4.   The role and ecological significance of nitrogen fixing activities
in lacustrine and estuarine sediments need further exploration.

5.   The role of sediments as nitrogen reservoirs or sinks needs further
clarification.  Experiments involving laboratory model systems and
in situ methods should be undertaken.

-------
                            SECTION III

                           INTRODUCTION
Of the major nutrient cycles in natural waters, the nitrogen cycle is
perhaps the most interesting, most complex and least understood from a
quantitative point of view.  The geocycle of nitrogen is largely a biochemi-
cal phenomenon; in natural waters it is nearly wholly so.  As such, the
nitrogen cycle, like the carbon and phosphorus cycles, is inextricably
related to aquatic organic productivity.  While many elements and com-
pounds are required for biosynthesis, nitrogen and phosphorus have long
been considered the prime limiting nutrients for primary production;
recent evidence suggests carbon may also limit production in some situa-
tions.  The great recent concern over cultural eutrophication has stimu-
lated much new research into the chemistry and biochemistry of these
nutrients in aquatic systems, into quantifying the sources and sinks of
nutrients and  into the dynamics of nutrient uptake and release.  This
report discusses  these subjects with emphasis on the cycle of nitrogen in
natural waters.

The nitrogen cycle in natural waters has been studied over a long period
of time.  Qualitative aspects of the cycle are well known, but quantitative
studies are much  less advanced.  For example, in situ rates of nitrogen
cycle  reactions have only  recently been determined in a  few lakes using
1^N.   Rates of ammonia  release  from  lacustrine sediments are virtually
unknown,  and determination of  the nitrogen contributions from various
sources is still  in  a crude and elementary state.  The  significance  of
natural and uncontrollable cultural  sources  like urban  and agricultural
runoff must be placed in  proper perspective  with the  one nutrient source
most  often cited  in  the popular press,  i.e.  domestic  waste effluent.  If
uncontrollable sources  are alone sufficient  to effect  eutrophy  in a  lake,
then a priori decisions to limit waste effluent as a nutrient source may
be highly costly  to  the taxpaper without  the anticipated benefits in
improved  water quality.   On the other  hand,  it may be found  that reduction
of nutrieYits in controllable sources may  significantly  improve  or completely
restore water  quality.   In any event,  decisions  to  limit nutrient  flux by
constraining certain sources must arise from knowledge  of  the  entire nitro-
gen budget  rather than  from ignorance  or  speculation.   To  understand the
factors controlling  the concentrations and temporal  variations  of nitro-
gen  forms in natural waters requires two  types  cf  research  efforts.   The
first involves quantitative studies  of the sources  and  sinks  comprising
the  nitrogen budget  of  a  water body; the  second  involves determination
of  the internal dynamics  or in situ turnover rates  of the  different  nitro-
gen  forms in  the water  body.

The  overall  objective of  the project was  to study  the factors  controlling
 the  concentration and forms of nitrogen in natural waters.   To this  end

-------
the project consisted of two principal phases:

     1)  delineation of the relative importance of certain source and
sink components in nitrogen budgets and;

     2)  investigation  of  the  internal cycling of nitrogen compounds
within a body  of water.  The study was conducted on lakes of varying
trophic states (from oligotrophic to hypereutrophic) in the region around
Gainesville, Florida, and  thus, has special relevance to lakes in humid,
subtropical  regions like the southeastern United States where information
of  this type is especially sparse.  However, the results should also have
a general  significance  in  other geographical areas, for many aspects of
the nitrogen cycle  are  similar in widely varying climates.

-------
                             SECTION IV

            REVIEW OF THE NITROGEN CYCLE IN NATURAL WATERS
Nitrogen occurs in the biosphere in a variety of forms ranging in oxida-
tion state from +5 to -3.  Inorganic nitrogen is present primarily as
highly oxidized nitrite and nitrate, as reduced ammonia and as molecular
nitrogen.  A variety of intermediate gaseous oxides of nitrogen are im-
portant in atmospheric chemistry but not in natural waters.  Naturally
occurring organic nitrogen consists primarily of amino and amide (pro-
teinaceous) nitrogen along with some heterocyclic compounds     present
as cellular constituents, as non-living particulate matter, as soluble
organic compounds, and as inorganic ions in solution.  All these forms
are interrelated by a series of reactions known collectively as the "nitro-
gen cycle," which portrays the flow of nitrogen from inorganic forms in
soil, air and water into  living systems and then back again into inor-
ganic forms.

This cyclic phenomenon can be illustrated diagramatically  in various
ways; Figure  1 presents  a simplified reaction sequence of  the interconver-
sions between organic nitrogen and  the raain inorganic forms, and it
labels  the principal  reactions.  This figure indicates the central role
of ammonia in the  cycle  as the link between the organic and inorganic
phases.  The major reactions of the nitrogen cycle  are (1) ammonia and
 (2) nitrate assimilation,  (3) ammonification,  (4) nitrification,  (5)
denitrification,  and  (6)  nitrogen  fixation.  Reactions (1) and  (2)
represent  and major pathways for conversion of  inorganic nitrogen into
organic (cellular) forms.  Reaction  (3) represents a series of  complex
reactions  breaking down  organic forms  (proteins,  amino acids, nucleic
acids,  etc.)  into  ammonia.  Reaction  (4), nitrification,  is  the  aerobic
oxidation  of  ammonia  to  nitrite and nitrate.   Reactions  (1)  -  (4) can
be considered to  comprise a potentially  closed system of  reactions
 (the internal cycling of nitrogen  among  its various forms  within  a body
of water).  Reactions  (5)  and  (6)  are  respectively  in situ sinks  and
sources for nitrogen  within given  ecosystem.   Nitrogen fixation  (reac-
tion 6)  is\he reduction of molecular  (atmospheric) nitrogen to  ammonia
and then to cellular  (organic) nitrogen.  Molecular nitrogen is  not nor-
mally  considered  part of the nitrogen  reserve  in  an ecosystem since it
 (N9) is not utilizable by most  organisms;  thus fixation  is a source of
nitrogen to the  exosystem.  Denitrification  (reaction 5)  is  a nitrogen
sink since it  converts utilizable  nitrogen  in  the forms  of nitrate  and
nitrite to molecular  nitrogen  under anoxic  conditions.

Figure  2 illustrates  the nitrogen  cycle  as  it  may occur  in an  idealized
 stratified lake.   With  the exception  of  ammonia exchange  with  sediments
 it  is  apparent  that  all  reactions  are  biologically mediated.  By far  the

-------
oo
 COMPLEX
 HIGH WEIGtHV
 ORG/tf
2	xPURINES
	PYRIMIDINES
PEPTI
     AMINO ACIDS


       AMINES
                                                             UREA
                                                                                                   1NO.
                       FIGURE 1,  SIMPLIFIED  NITROGEN  CYCLE SHOVING MAIN MOLECULAR TRANSFORMATIONS:

                           1, NITRATE ASSIMILATION,  2,  AMMONIA ASSIMILATION, 3,  AMMONIFICATION/
                                 ^i, NITRIFICATION/ 5,  DENITRIFICATION, 6,  i\!ITROGEN FIXATIO?]

-------
                                                         EPILIMNION
                                                         INORGANIC N
                                                         ASSIMILATION
                                                         THERMXLINE
                                                        •lYPOLIfttlON
                                                        SEDIMENT
FI"URE 2,  NITROGEN  CYCLE REACTIONS IN AN IDEALIZED STRATIFIED LAKE,
      "OTE T-iAT 30TJ  AER03IC AND A^AEROT;iIC TRAiXS
         ARE WO'v':  I\| T:t L'V^OLI^;iOr;,  II! A REAL LAKE
         HEY ^OULD OF COURSE :JOT OCCUR SIMULTA'IEOUSLY,
          REDRAWN AND ADAPTED FRO" \UZ;!3T30V (1Q59) ,

-------
greatest influx of inorganic nitrogen into organisms results from ammonia
and nitrate assimilation.  These reactions predominate in surface waters
and are mediated primarily by phytoplankton and macrophytes.  While nitrate
tends to be the predominant inorganic nitrogen form in surface waters,
there is considerable evidence  that  ammonia is the preferred form for plank-
tonic assimilation since  it is  already at the reduction level of organic
nitrogen.  Nitrate can be used  by most plants (Strickland, 1965, and
others  cite some flagellates as exceptions and a few algae have been
reported to prefer nitrate over ammonia  (Proctor, 1957)). Organisms using
nitrate as their nitrogen source must first reduce it to the level of
ammonia before  incorporating it into organic forms, which process re-
quires  a reduction system including  the  enzyme nitrate reductase.  This
inducible enzyme is  present in  algal cells only when nitrate is being
used  as the nitrogen source (Eppley, Coatsworth et_ a.l_. , 1969), which
suggests a mechanism for  determining the form of nitrogen an algal pop-
ulation is using.

Reduction of  nitrate also requires the oxidation of organic matter to
C02>  and this oxidation-reduction sequence is apparently not linked to
the  respiratory chain and to ATP synthesis.  Thus utilization of nitrate
is in a sense energetically wasteful.

Ammonia is a  weak base:

                  NH,+    £  NH,,    +  H+ ,  pKa = 9.3.
                     4        J aq.
Thus  in poorly  buffered waters  where photosynthesis may raise the pH
as high as 10-11,  the predominant form is NH-j     while in well buffered
and  in  low productivity systems the  cationic form (NH, ) dominates.
The  fact that ammonia assimilation can continue at acidic pH values
(where  [NHo   ]  is negligible)  implies that the cation itself is trans-
ported  through  the cell membrane.  At high concentrations NHo is toxic
to organisms  and this fact may  explain the reports of earlier workers
that  ammonia  is  a less suitable source of nitrogen for algal cultures
than  is nitrate.  Growth  media  frequently contain 5 or more mg N/l (as
ammonia or nitrate) .  If  CC>2 uptake by photosynthesis is sufficient to
raise the pH  to  10 or so, then  the NH, could act as a toxicant.  On the
other hand assimilation of ammonia and its incorporation into organic
matter  tends  to  lower the pH   (Fogg, 1965) since NH  is a base and its
removal tends to increase acidity.  Whichever phenomenon is in fact res-
ponsible for  the inhibitory effects  noted with high ammonia concentra-
tions,  it is  clear that a well  buffered  system is essential to maintain
proper  growth conditions  in ammonia  containing media.

Rates of ammonia and  nitrate assimilation in aquatic environments have
recently been determined  by several  workers using 15N tracer techniques.
Such studies  are relevant to an increased understanding of the metabolism
of the  aquatic  community  since  the availability and utilization of
nitrogen profoundly  influences  and in some cases controls primary pro-
duction.  Of  interest in  this regard is  such information as the rates
of inorganic  nitrogen turnover, the  relative nutritional importance of
the various nitrogen  forms, the effect of concentration on rates of
utilization,  and the  relationship between nitrogen assimilation and
                                     10

-------
primary production.  Althougn the   N tracer approach to studying aquatic
nitrogen cycle dynamics is only about 10 years old, considerable gains
have already been made through its use.  The nitrogen cycle can no
longer be viewed as a slow, primarily seasonal cycle.  Turnover times
for inorganic nitrogen are on the order of days or even hours in many
waters, which  implies the essentiality of rapid organic nitrogen decom-
position or a constant influx of new nitrogen into the system.

Dugdale and Dugdale (1965) were the first to report In situ ammonia
and nitrate assimilation rates in surface waters.  Using ' bN techniques
they found two main pulses of assimilation in Sanctuary Lake, Pennsyl-
vania.  During the first pulse, in June, ammonia was assimilated most
rapidly with a maximum rate of about 110 yg N/l.-day.  The maximum
nitrate assimilation rate during the same period was about 40 yg N/l-day.
A second pulse of assimilation in early September utilized primarily N~.
Some of the high assimilation rates reported may be an artifact of
methodology.  For example, the nitrate concentration at the time of
the June assimilation maximum was only about 10 yg N/l.  The amounts of
  NO-j" and   NH~ added to samples were not specified, but a large amount
of tracer relative to the initial unlabeled nitrogen would stimulate
assimilation beyond the normal rate.

Rates of ammonia and nitrate assimilation in Lake Mendota, Wisconsin,
during 1966 were reported by Brezonik (1968, 1971a).   Ammonia assimi-
lation rates were greater than nitrate assimilation rates in all cases,
and highest assimilation rates in the surface water occurred in late
spring and late summer.  No correlation was found between ammonia and
nitrate concentrations and assimilation rates.  Depth profiles of
ammonia assimilation during  holomixis and early stratification were
fairly uniform and no trends were indicated.  However, by early June
a pronounced stratification of assimilation was found with high rates
in the  epilimnion and much lower rates in the hypolimnion.

In nearly every assimilation study published thus far ammonia was by far
more important than nitrate as a nitrogen source.  In view of the ease
with which ammonia can be assimilated this is not surprising, but prior
to these tracer studies most marine biologists considered nitrate to
be the only significant nitrogen source in the sea.  It now seems cer-
tain that both marine and fresh water algae derive most of their nitro-
gen from ammonia often in spite of higher nitrate concentrations.  There
may be a certain amount of "wheel-spinning" involved in ammonia assi-
milation; rather than representing new production, at least part of
the uptake may derive from the necessity to recapture nitrogen from
compounds which apparently continually leak through cell walls (Whitta-
ker and Feeney, 1971; Hellebust,  1965; Stewart, 1963).  Dugdale and
Goering (1967) derived a simple model for the nitrogen cycle in marine
surface waters in which they considered primary production associated
with ammonia assimilation to be "regenerated" production,and primary
production associated with nitrate assimilation was regarded as "new"
production.   Only the latter would be available for export to higher
trophic levels since nitrate input from the deep water is the principal
source of nitrogen to the surface.  Ammonia in sea water is the product
of short term regeneration, and primary production associated with its
                                   11

-------
assimilation is essentially that needed  to maintain the standing crop.
Extension of this approach to  the more complicated nitrogen cycles in
shallow lakes would be of perhaps questionable validity.

An exception to the general rule of more rapid ammonia than nitrate
assimilation was recently reported by Goering &t_ al. (1970) in the eastern
subtropical Pacific.  Phytoplankton in a nutrient-rich discontinuity
layer at the bottom of the euphotic zone were found to use nitrate pri-
marily while algae in the impoverished surface water used ammonia.

Dugdale  (1967) has developed  a model for nutrient limitation in the sea
in which assimilation of ammonia and nitrate were assumed to follow
Michaelis-Menten kinetics:
                                     VS
                               v =
                                   Kt+S    ,


 where  v is  the  rate  of  nutrient  assimilation; V a maximum rate, constant
 under  a given set  of conditions;  S is nutrient concentration; and Kt a
 transport or half-saturation constant with units of concentration.  The
 numerical value of Kfc is  equal to the nutrient concentration at which
 v is one-half the  maximum for the system  ( V/2).  Subsequent l5N measure-
 ments  on natural populations by  Maclsaac  and Dugdale (1969) and direct
 uptake measurements  on  algal cultures by  Eppley, Rogers _et^ a^Li (1969)
 have verified this model  and estimated Kt values for various marine
 surface waters.  Kt  values calculated by  the former authors ranged from
 less than 0.2 to 4.2 ug.-atom/1,  for nitrate and from 0.1 to 1.3 yg.-atom/1,
 for ammonia. Lower  transport constants correlated with oligotrophic
 (nutrient-depleted)  areas while  higher K  values were found in nutrient
 rich waters. It was thus suggested that  phytoplankton in oligotrophic
 waters are  adapted to low ambient concentrations and can assimilate
 nutrients more  rapidly  under these conditions than can phytoplankton
 from nutrient-enriched  regions.

 The reverse of  assimilation is ammonification, whereby organic nitrogen
 is returned to  the inorganic nitrogen pool as ammonia.  This is a compli-
 cated  process involving several  mechanisms.  Early workers considered
 only bacterial  decomposition of  soluble organic nitrogen and organic
 detritus  as important (vonBrand  et. al. ,  1937), but more recent studies
 have shown  significant  excretion of ammonia and amino acids by zooplank-
 ton feeding on  phytoplankton and detritus.  Johannes (1969) suggested this
 as the  dominant  mechanism of ammonification in surface waters  and
 reviewed  previous  work  indicating that net zooplankton release amounts
 of dissolved nitrogen and phosphorus equal to  their total body content
 of these  nutrients in 20  to 200  hours.  A third mechanism for ammoni-
 fication  is  direct autolysis after cell death, which may account for 30
 to 50 percent of the  nutrients released from plant and animal material
 (Johannes,  1969; Krause,  1964; Golterman, 1960).  The excretion of amino
 acids and other  small nitrogenous substances from photosynthesizing
 cells  (Hellebust,  1965) and their subsequent direct uptake by other algae
 or bacteria represents  a  short-circuiting complication of ammonification
 pathways, the significance of which is not yet understood.
                                     12

-------
Current information and theories suggest that bacterial decomposition
of organic matter accounts for a minor portion of nutrient regeneration
in marine surface waters and in shallow lakes (although this is probably
not not true in systems receiving large organic inputs from sewage out-
falls) .   In soils and bottom sediments and in anoxic hypolimnia of
lakes, nutrient regeneration by bacteria and fungi is the dominant pro-
cess.  Ammonification is obviously important in reviewing a limited
supply of nitrogen for assimilation and growth of primary producers.

In situ ammonification rates can be measured by an isotope dilution
technique (Zilversmit e_t_ _al. , 1943; Dugdale, 1965; Brezonik, 1968).
Assuming steady state conditions for assimilation and ammonification,
15N label added to the ammonia pool will gradually become diluted as
both labeled and unlabeled ammonia is assimilated but comparatively
unlabeled N is returned to the ammonia pool by ammonification of organic
N.  The decrease in l5N enrichment in the ammonia pool can be quantified
and related to the rate of ammonification.  Because of the difficulties
encountered in l5N measurements, only a few in situ ammonification
rates in lakes have been reported.  Alexander (1970) reported that
maximum ammonification rates in Smith Lake, Alaska, were associated
with a spring bloom of Anabaena and with the period of maximum ammonia
assimilation.  Relatively high rates were measured in Lake Mendota,
Wisconsin  (Brezonik, 1968, 1971a) with a tendency for higher rates in
the bottom waters.  From steady-state kinetics, ammonia turnover times
were calculated to range from 7 to 62 hours in Lake Mendota surface
waters during the summer.  Because steady state conditions do not
exactly apply and because of the errors inherent in the methodology,
these calculations are only approximate (Brezonik, 1971a); nonetheless
they suggest a much more rapid cycling of nitrogen than measurement of
concentration changes alone would suggest.

Ammonia is oxidized to nitrite and nitrate in the process of nitrifica-
tion by a select group of aerobic autotrophic bacteria which obtain
their energy by nitrogen oxidation and  their cellular carbon by reduc-
tion of C0?.  A variety of heterotrophic bacteria, actinornycetes and
fungi (Schmidt, 1954; Hatcher and Schmidt, 1971) have also been reported
capable of nitrification, generally  at  much  slower rates, but  the  signi-
ficance of these organisms  in  aquatic nitrification is not well known.
The biochemistry and metabolism  of nitrifying organisms has been reviewed
by Painter  (1970) particularly with  reference to  its  occurrence in sewage
treatment.  The significance of  nitrification in  the  nitrogen  cycle  lies
in the conversion of labile  ammonia  (which tends  to be lost from solu-
tion by sorption onto sediments  and  by  volatilization at  high  pH)  to  a
more stable form  (nitrate).  On  the  other hand, nitrate can be reduced
to molecular nitrogen by  the process of denitrification,  and thus  nitri-
fication has a  second and opposite role —  that of producing the reac-
tants for  this  nitrogen sink.  The oxidation of ammonia to nitrate re-
quires almost 4.5 mg 02 per mg  of  ammonia -  N oxidized.   Thus  nitrifi-
cation can  act  as an important  oxygen sink  in streams which receive high
ammonia concentrations  from unnitrified sewage effluents.  Attempts  to
quantify  and model these  effects have recently been  described  by a
number of workers  (Stratton and  llcCarty,  1967; Wezernak and Gannon,
 1969: Thomann et  al., 1971).
                                  13

-------
In  situ nitrification rates can be estimated by  adding  15NE^  to  a water
sample  and determining the amount of label in the  nitrate  and nitrite
fractions after incubation (Brezonik, 1968).   The  latter is accomplished
by  reducing the oxidized nitrogen to ammonia with  Devarda's alloy
 (Bremner and Keeney, 1966), and subsequent conversion of ammonia to N?
by  hypobromite oxidation (Rittenberg   «!t_ al., 1939; Neess et_ al_. , 1962)
 for isotope ratio analysis.  Dugdale and Goering (1967) were  unable to
 detect nitrification in the surface waters of the  Sargasso Sea using
 incubations as long as a week, and it appears that  regeneration  of
 nitrate in the sea occurs primarily below the euphotic  zone.   A  few
 nitrification rates were reported for Lake Mendota by Brezonik (1968,
 1971a).  During 1966 rates increased with depth  in spring but were very
 low in the surface water during summer.   The fact  that  nitrate levels
 invariably decline in lake surface waters during spring and summer
 suggests that nitrifying bacteria are unable to  compete with  algae for
 ammonia, and nitrification is thus felt  to be of  minor significance
 in surface waters during the growing season (Brezonik,  1968).  However,
 Brezonik and Lee (1968) also found that  substantial nitrification evi-
 dently occurred at mid-depths (7-17 m.)  in Lake  Mendota, Wisconsin,
 during late spring and early summer.   Largest  increases occurred in the
 12-15 meter zone where nitrate increased from less  than 0.2 mg.  N/l. in
 June to 0,5-0.7 mg. N/l. in late July.   Nitrification in the  hypolimnion
 is  especially significant since the water eventually becomes  anoxic
 during late summer, and denitrification  takes  place.  Thus nitrifica-
 tion increases the importance of denitrification as a nitrogen sink in
 Lake Mendota.  Vollenweider (1963)  has  also reported nitrification at
 mid-depths during summer stratification  in Lake  Orta, Italy.

 In  the process of denitrification nitrate is  used  as a  terminal  electron
 acceptor by facultative and anaerobic bacteria in  the absence of oxygen.
Nitrite is formed as the first intermediate in the  process, and  nitrous
oxide can sometimes be formed along with molecular nitrogen although it
is  not  an essential intermediate.  Since the principal  end product (N2)
is  a nitrogen form not utilizable by most organisms, this  reaction acts
as  a nitrogen sink and can assume  importance in the nitrogen balances
of  lakes and other aquatic systems subject to intermittent anoxia,
 (e.g. soils, Delwiche, 1965).  The biochemistry  of  denitrification has
been reviewed extensively (Delwiche, 1956, 1965; Kessler,  1965;  Painter,
1970),  who describe the long and considerable controversy  over the re-
quirement of completely anoxic conditions for this  reaction to occur.
The consensus now is that denitrification occurs only when oxygen is
absent  from the system or at least sufficiently  low enough to allow
anoxic  microzones to develop.

Denitrification may account in part for the difficulty  encountered in
obtaining nitrogen balances in waste treatment plants (Wuhrmann, 1954;
Symons  et al.,  1965) .   The reaction may occur in anoxic microzones of
activated sludge floe or in anoxic settling basins. Nitrification
followed by intentional denitrification has been proposed  as  a means
of  nitrogen removal from biological waste treatment plants (Johnson and
Schroepfer,  1964;  Wuhrmann, 1964) .  Denitrification has been  shown to
be  an important nitrogen sink in lakes (Brezonik and Lee,  1968;  Goering
and V.  Dugdale, 1966).   Its role in the nitrogen cycle  of  the oceans is
                                 14

-------
less defined but there is no doubt of its occurrence in some oceanic
environments (Goering and R. Dugdale, 1966; Goering, 1968).

Not all the nitrate that is reduced is lost to the system as N_.   A
significant but variable amount of assimilatory reduction to ammonia and
organic nitrogen occurs simultaneously (but different organisms may be
involved in the two processes).  This was first noted by this author
in denitrification experiments with sewage sludge digesters; only 50
percent of added nitrate was recovered as N2 (Brezonik and Lee, 1966).
An early study of denitrification in lakes (Goering and V. Dugdale,
1966) suggested that assimilatory nitrate reduction was unimportant
based on low measured rates of ammonia production from labeled nitrate into
organic nitrogen.  Brezonik and Lee (1968) found that reduction to ammonia
and organic nitrogen occurred at comparable but somewhat lower rates
than denitrification in Lake Mendota, Wisconsin.  Of the 4.43 x 107g.,
nitrate-N present in the lake hypolimnion below 14 m. in mid summer of
1966, an estimated 2.81 x 107g. were lost by denitrification while the
balance (1.62 x 107g.) was reduced to ammonia and organic nitrogen.
Goering (1968) also reported ratios of N2-N formed/N03~-N lost ranging
from 1/10 to 8/10 in the eastern Pacific, indicating that assimilatory
reduction is of great but variable importance.

Denitrification in lake sediments may be an important nitrate sink in
ground water seeping into lakes as Keeney et al. (1971) have recently
described for Lake Mendota, Wisconsin.  Finally, denitrification may be
be  the mechanism whereby  the high N/P ratios of fresh waters are trans-
formed to the lower ratios  encountered in estuarine waters  (Stumm, 1971).
No  In situ  experiments  of  denitrification have been conducted in  such
environments, but if density  stratification is sufficient to produce
anoxic or near anoxic  conditions  in  the bottom layers of an estuary,  this
reaction could occur.

Probably more attention  has been  devoted  to nitrogen fixation  than to  all
the other nitrogen cycle  reactions combined.  This  process  is  important
at  several  levels.  Geochemically nitrogen  fixation is  essential in main-
taining a nitrogen balance  in the biosphere, which  would  otherwise become
depleted as a consequence of  denitrification  over  a time  scale of
millions of years.  In agriculture the  reaction  is  important  in maintaining
or  increasing soil  fertility.   In natural waters nitrogen  fixation acts  as
a source of nitrogen  and permits  continued organic  production when the
supply of  fixed  nitrogen becomes  depleted.  A variety  of  organisms are
capable of  nitrogen  fixation  including  a number  of  blue-green  algae,
apparently  all  photosynthetic bacteria,  various  aerobic bacteria  (e.g.
Azotobacter), anaerobic  bacteria  (e.g.  Clostridium) , many  facultative
bacteria  (but only  under anoxic conditions (Wilson, 1969)),legume root
nodules and nonleguminous root nodulated plants  such as Podocarpus anr*
the alder  tree  (Alnus  sp.).   Most studies of  nitrogen  fixation in  natural
waters  to  date  have  emphasized the role of filamentous, heterocystous
blue-green  algae.   The occurrence of Azotobacter and Clostridium in  the
Black  Sea  (Pshenin,  1959, 1963)  and  in  other  aquatic habitats  has  been
established,  and photosynthetic bacteria have been found  in dense  popu-
lations  in a  variety  of  lakes (e.g.  Triiper and Genovese,  1968). However
the presence  of  bacteria capable  of  fixation  does  not  necessarily  imply
                                 15

-------
the actual occurrence of  fixation in any  given environment.  On the other
hand these organisms have  the potential  for  fixation presumably because
it is useful to them at  least in some habitats and  it would be surprising
if fixation bv these organisms were not  found in  the environment.  Most
biologists have de-emphasized the role of these organisms  (especially
the heterotrophic bacteria)  in nitrogen  fixation  because they have felt
that insufficient oxidizable carbohydrate (or other substrate) would
be available  in the environmental Stewart (1966,  1968,  1970) has reviewed
the  agents  and occurrence of nitrogen  fixation in the biosphere in greater
detail.

Nitrogen fixation  is  generally  considered to be an  adaptive process used
by organisms  only when the supply of  fixed nitrogen is  depleted.  Nitrogen
 fixing alyae  usually  bloom in lakes  only after nutrients have been de-
pleted by blooms  of  other algae  (i.e.  late summer in temperate lakes).
However, contrary  to  earlier opinions,  small to moderate concentrations
 cf ammonia do not  necessarily inhibit  fixation although synthesis of
 the  enzyme nitrogenase is repressed  at  high levels.  Stewart (1969)
 suggests that the  levels of combined nitrogen nitrogen  in most natural
 ecosystems are insufficient to  inhibit  fixation immediately or even to
 persist long enough  for existing nitrogenase to be  diluted out.  Low
 levels of combined nitrogen may  actually be advantageous to nitrogen-
 fixing plants by  enabling more  efficient and healthy growth than could
be  achieved on N2  alone.

 Reduction of N2  to the level of  ammonia requires  energy and a source of
 reduced hydrogen,  both of which  may  be  obtained either  from chotesynthetic
 production or from oxidation of  organic  carbon.   Nitrate reduction to
 ammonia similarly  requires energy obtainable by oxidizing organic sub-
 strates.  While  the  net reactions  in each case are  exergonic, they are
 nevertheless energy  sinks since  the  cells apparently have no means of
 trapping and storing the energy  (as  ATP)  given off by the  reactions.   To
 this extent use  of nitrate and of molecular nitrogen rather than ammonia
 for  assimilation  is  wasteful of  energy.

Molybdenum has been  shown to be  a specific requirement  for organisms
using  nitrate or  fixing nitrogen,  and  it apparently is  a constituent  of
the  enzymes directly involved in the processes (nitrate reductase and
nitrogenase,  respectively).  This  element has been  shown to be a limiting
nutrient in various  soils and at least  one lake has demonstrated a
deficiency.   Goldman  (1960) found a  planktonic growth response when
molybdenum was added  to Castle  Lake,  California;  this Mo limitation
was  presumably linked to nitrogen metabolism.  Nitrogen fixing organisms
also require  iron  in  greater amounts than is required by organisms
growing on  fixed nitrogen, and cobalt  (or vitamin B^)  has also been
implicated  as  an essential nutrient  for fixation.

The  biochemical pathway of nitrogen  fixation is poorly  understood.
A long history of  efforts to isolate intermediate reduction products is
recorded in the literature, but  thus far only  ammonia — the final in-
organic intermediate  before conversion to organic nitrogen compounds —
has  been definitely shown to be  a step in the process.  Because of this
distinct lack  of success in finding  intermediates,  most biochemists
                                  16

-------
currently favor a pathway in which the intermediates are tightly bound
to the enzyme active site.  Burris (1966, 1969) has reviewed the bio-
chemistry of fixation in several informative articles.

Suggestions that nitrogen fixation occurs in lakes date back at least
as far as far as Hutchinson (1941), who thought that Anabaena was res-
ponsible for an increase in fixed nitrogen in Linsley Pond (Connecticut)
during the summer.  The first in situ measurements of lacustrine fixa-
tion were reported by Dugdale et_ ad. (1959).  These workers developed the
15N technique (Neess _et_ al_.  1962) and demonstrated the occurrence of
low rates of fixation in several lakes.  More thorough studies were
reported by these workers for several Alaskan lakes (Dugdale and Dug-
dale, 1961) where only very low rates were detected; in Sanctuary Lake,
Pensylvania, (Dugdale and Dugdale, 1962)  where a maximum rate of 130 yg
N fixed per liter per day was associated with a late summer bloom of
Anabaena; in Lakes Mendota and Wingra, Wisconsin, (Goering and Neess,
1964); in Smith Lake, Alaska (Dugdale, 1965); and in the Sargasso Sea
(Dugdale e_t £l. , 1964).   High rates of fixation have been consistently
correlated with blooms of blue-green algae, primarily Anabaena,
Gloeotrichia, and Aphani z omen on in lakes, Trichodesnium in subtropical
marine waters and Calothrlx in shallow temperate marine environments
(Stewart, 1965).

The expense and difficulties associated with !5N tracer techniques pre-
cluded routine and detailed investigations of nitrogen fixation in
aquatic environments until recently when a simple, indirect method of
assessing nitrogen fixation rates was developed (Stewart _e_t al. , 1967) .
This method utilizes the fact that the nitrogen fixing enzyme complex
reduces acetylene to ethylene, the production of which can be sensitively
and easily determined by gas chromatography.  Acetylene (H-C=C-H) is
isoelectronic with molecular nitrogen (N=N), and the nitrogenase system
is incapable of distinguishing between the two molecules.  In fact a
number of other similarly shaped molecules such as cyanide (C~N)~  and
carbon monoxide (C=0) also react at or attach           to the nitro-
genase active site and thus act as enzymatic inhibitors of fixation
(Burris, 1969).  Based on the fact that nitrogenase reduces acetylene
to ethylene (H2C=CH2) , Dilworth (1966) proposed that di-imide (HN=NH)
is the first intermediate in the reduction of nitrogen.  Of the various
compounds that react at the nitrogenase active site, acetylene is best
suited for indirect assay of nitrogenase activity.  Neither acetylene
nor its redaction product is toxic to the organisms.  Furthermore
acetylene reduction yields a single, easily measured product; (ethylene
is not further reduced to ethane by the enzyme).  Finally the affinity
of nitrogenase for acetylene is large, in fact larger than for its
normal substrate.  Thus even in the presence of molecular nitrogen,
acetylene is preferentially reduced.  The advantages of the acetylene
reduction assay over 15N techniques are both a greatly simplified metho-
dology and a greater sensitivity.  These properties have facilitated
such fundamental studies on aquatic nitrogen fixation as the effects of
sunlight intensity, detailed depth profiles, and diel variations (Stewart
et_ jil. , 1967; Rusness and Burris, 1970).  In addition the fact that
acetylene reduction yields a gaseous product has enabled the study of
                                  17

-------
low fixation rates in lacustrine and estuarine sediments (Brooks et al.,
1971; Keirn and Brezonik,  1971); see Section VI.

In work supported by  this  project  Brezonik  and Harper (1969) reported
evidence for nitrogen fixation  in  anoxic  lake hypolimnia, and Stewart
(1969) has  similarly  reported  on  fixation  in anoxic waters of a Nor-
wegian fjord.  Brooks (1969), Brooks ejt _aJL. , (1971) and Keirn and Brezonik
(1971) in work described  later  in  this report have also found low but
measurable  rates  of nitrogen fixation in  estuarine and lacustrine sedi-
ments, and  Howard  et^  jil_.  (1970) have reported low fixation rates in Lake
Erie sediments.   Thus the known distribution of nitrogen fixation in
aquatic habitats  has  been extended considerably by recent work.

The  rate  at which  newly fixed nitrogen is transferred from agents of
fixation  to other aquatic organisms is of great interest.  Jones and
Stewart  (1969) have  reported that  extracellular nitrogen liberated by
the  marine  nitrogen  fixer, Calothrix scopuloruci, is assimilated by
marine  algae,  fungi,  and  bacteria  and could serve as the sole nitrogen
source  for  a  species  of Chlorella. Based on limiting nutrient bio-
assays,  Fitzgerald (1969) concluded that  transfer of nitrogen newly fixed
by Aphanizomenon  is  of  little significance  to colonies of Microcystis
grox-jing  in  the same water, but  this is perhaps an overextension of the
data.  It seems  clear from Fitzgerald's results that Aphanizomenon is
unable  to supply  Microcystis with  sufficient nitrogen to satisfy the
organisms,  but the actual magnitude of supply is still unclear.

With the  exception of sediment-water interactions (to be discussed in  a
later section),  the  major nitrogen cycle  reactions of natural waters
have been discussed  above.  As  presently  understood the nitrogen cycle
is almost exclusively a biological phenomenon.  However, the possible
importance  of chemical reactions  and of heretofore undetected  (but
potential)  biological transformations should not be overlooked.  A
variety  of  other  reactions are  thermodynamically feasible, and some of
these could have  a substantial  impact on  the overall cycle.  The mechanisms
of No formation  from fixed nitrogen have  been the subject of a long con-
troversy  and  many pathways have been proposed and rejected.  Formation
of N? directly from  ammonia has been hypothesized to occur in  sludge
digestion (Malina, 1961;  Crane, 1962) and anoxic lake waters  (Koyama,
1964) , but  others  have found no evidence  for such a reaction  (Brezonik
and  Lee,  1966; Wijler and Delwiche, 1954).  The latter authors used  l5N
methods  to  show  that  all N  in denitrification derives from oxidized
nitrogen  (NO^" ,NC>2~)  sources.  The reaction of nitrous acid with amino
groups under  acid conditions:

            RCHCOOH + HN02 	>  RCHOHCOOH + N? + H,,!)

(commonly known as the  Van Slyke  reaction)  has been proposed by some
workers  to  be  significant in acid lakes  (Hutchinson, 1957) and soils
(Reuss and  Smith,  1965) but has been rejected by others  (Bremner and
Nelson,  1968;  Nelson  and  Bremner, 1969).   However  the latter  authors  have
provided  evidence  for significant chemical  decomposition of nitrite  to
a  variety oC  gaseous  products in  soils.   Bremner and Nelson  (1969) demon-
strated  that  nitrite  decomposition in sterilized soils is due  to reaction

-------
of nitrite with soil organic matter and to self decomposition of nitrous
acid (2HNO- = NO +  NOo + H~0)  under acid conditions.  The nitrite-or-
ganic matter reaction was found to involve phenols and polyphenols such
as lignins and tannins.  The mechanism is thought to involve reaction
of nitrous acid under slightly acid conditions (pH 5) with phenolic
compounds to form nitrosophenols which tautomerize to quinone oxiir.es.
The latter are decomposed by nitrous acid to form ^ and ^0 (Austin,
1961).   This reaction could be of importance in the nitrogen cycle of
natural waters.  Colored lakes are high in polyphenolic substances
(tannins, lignins, "humic acid") and generally have acid pH values.
While nitrite levels in natural waters are normally low, such a reaction
could act to decompose nitrite as rapidly as it if formed.

While photochemical reactions are highly important in the atmospheric
chemistry of nitrogen, photochemistry is presently thought to play an
insignificant role in aquatic nitrogen transformations.  Since most photo-
chemical reactions are induced by UV light, such processes should be
limited to a narrow surface layer.  Photochemical nitrification in the
sea was proposed by various workers in the 1930's (ZoBell, 1935;
Rakestraw and Hollaender, 1936; Cooper, 1937) partly as a result of
difficulties in culturing marine nitrifiers (Waksman et al., 1933), but
Hamilton  (1964) concluded that such a reaction is of no significance
in the marine nitrogen cycle.  Hamilton found slight photoreduction of
nitrite but  concluded  that  this too was insignificant.

Several  thermodynamically possible nitrogen reactions have never been
shown  to  occur as biological  phenomenon.  For example, N? could be used
as a terminal electron acceptor in organic carbon oxidation with the
release  of sufficient  energy  for  organism growth:

             glucose +  4N2 + 8H+ + 6H20 = 6CO  + 8NH*  , AG° = -84.5 kcal/mole,

The  above  reaction of  course  occurs when heterotrophic organisms fix
nitrogen,  but  it  is not  thought  that  any of  this  energy is trapped and
retained by  nitrogen  fixing cells for  other  uses.  However,  it  is  con-
ceivable  that  this reaction could be  a source  of  energy for  some anaerobic
bacteria.  Oxidation  of  molecular nitrogen  to  nitrate by  molecular oxygen
is also  exergonic at  pH  7:

             N2 +  ~°2  + H2°  = 2N03~ +  2H+ (w)'  AGw =  ~15-2 kcal/mole.

With the  large reservoir of available reactants  for  this  process,  it  is
surprising that no organisms have evolved  to take advantage  of  the situa-
tion.  However, this  is  probably  fortunate  considering  the nature  of  the
product  (i.e.  nitric  acid).
                                   19

-------
                             SECTION V

                 TRANSPORT OF NITROGEN INTO LAKES
The concentration and forms of nitrogen in a lake at a given time are a
product of input rates, of the interconversion reactions occurring within
the lake and the rates of loss via outflow, denitrification and sediment
deposition.  Lakes act in general as nutrient traps, implying an accumu-
lation of nitrogen in bottom sediments as a lake gradually fills in.  Both
natural and cultural nitrogen sources can be significant in lacustrine
nitrogen budgets.  The natural nutrient input to a lake is a function of
its drainage basin geochemistry (i.e. the potential for nutrient leach-
ing from soils and substratum), drainage basin size, hydrology, precipi-
tation patterns, etc. (Brezonik et_ jil_. , 1969).  Superimposed on these
factors are a multitude of human or cultural factors, which can be ex-
pressed in terms of drainage basin land use patterns and population
characteristics.  Quantitative information on lacustrine nitrogen budgets
and the significance of individual sources is sparse, but concern over
cultural eutrophication has stimulated much needed measurements of the
nitrogen (and phosphorus) contributions from various sources along with
development of elementary nutrient models.

Table  1 lists the possible nitrogen sources and sinks which must be con-
sidered in calculating the nitrogen budget for a lake.  The classical
approach to evaluation of a lake's nutrient budget is actual measurement
of nutrient concentrations and flows  for each source over a reasonable
time span  (e.g. one year).  Obviously this requires a large time and man-
power  expenditure, especially  for large lakes, and some diffuse sources
(e.g.  ground water seepage) may not be amenable to direct or accurate
evaluation.  Because of these  difficulties it is probable that complete
measurements of all sources and sinks have not been accomplished for any
single lake although approximate nutrient budgets have been established
for a  few American and European lakes  (see Brezonik ejt^ jil_. , 1969, and
Vollenweider, 1968, for reviews).  An alternative to actual measurement
is development of a simple simulation model of nutrient transport based
on knowledge of the lake's drainage basin  size, population, hydrology and
land use patterns and on  literature  values for nutrient contributions
from the various sources.  This approach was  taken by Lee e_t_ _al_. (1966)
in their well-known budget for Lake Mendota,  Wisconsin.  Actual measure-
ments  of nutrient contributions are  undoubtedly more accurate  than  litera-
ture estimates, but as the literature on nutrient export rates from various
land uses  and other sources becomes  more complete, the accuracy of
literature-based nutrient budgets will increase.  The accuracy of such
nutrient budgets depends  not  only on  the accuracy of the original measure-
ments  but  also on their applicability to the  particular system under
consideration.  However considering  the expense and difficulties of direct
                                  21

-------
Table 1.  Sources and Sinks for the Nitrogen Budget of a Lake
                           Sources
 1.  Surface
     Agricultural  (cropland) runoff
       and drainage
     Aniraal waste  runoff
     Marsh drainage
     Runoff from uncultivated  and
       forest  land
     Urban storm water  runoff
     Domestic  waste  effluents
     Industrial  waste effluent
     Wastes from boating  activities

  2.  In.  situ

     Nitrogen  fixation
     Sediment  leaching
3.   Airborne

    Rainwater
    Aerosols and dust
    Leaves  and miscellaneous
     debris
4.  Underground

    Natural ground
    Subsurface agricultural and
      urban drainage
    Subsurface drainage from
      septic tanks near lake
      shore
                             Sinks
     Effluent  loss
     Groundwater  recharge
     Fish harvest
     Weed harvest
     Insect  emergence
     Volatilization  (of N
    Evaporation (aerosol) forma-
      tion from surface foam)
    Denitrification
    Sediment deposition of detritus
    Sorption of ammonia onto
      sediments
                               22

-------
measurement, literature-based models seem the only reasonable approach
for many lakes receiving cultural stress.  Thus refinement of nutrient
contribution values from the various types of sources is of great impor-
tance.

Clearly experimental evaluation of the quantitative importance of the
myriad sources of nitrogen was beyond the scope of the present project.
Rather it was concerned primarily with the importance of in situ sources
and sinks, especially nitrogen fixation and sediments (see Sections VI-
VIII).  However, in conjunction with a concurrent study of eutrophication
in Florida lakes a significant effort has been made to gather information
from the literature regarding the significance of natural and cultural
sources and to apply this information to the development of nitrogen
budgets and simple nutrient-eutrophication models for Florida lakes.  In
addition the significance of rainfall as a nutrient source received some
experimental attention during this project.

The significance of particular nitrogen  sources has  received  considerable
attention  in various recent reviews, and an  exhaustive  literature  review
is not  in  order herein.  For details the reader should  consult  the ori-
ginal literature or the  specific reviews cited below.   This section will
present a  summary  of the nitrogen  (and to a lesser extent  phosphorus)
contributions  from various  sources  and describe the  use of such data  in
developing nutrient budgets  and enrichment models.   Feth  (1966)  reviewed
the sources and  concentration  range of nitrogen in natural waters  empha-
sizing  natural sources  and  a geochemical viewpoint.  Agricultural  sources,
especially fertilizer and soil losses have been treated by Biggar  and
Corey (1969),  and  C. F.  Cooper (1969) has  reviewed nutrient export from
forest  land.   Weibel  (1969)  reviewed his  and other studies on urban nutrient
sources.   Schraufnagel j2t_ al.  (1967)  and Lee et_ jil_.  (1966) reviewed the
natural and cultural sources of  nitrogen especially  as  applied  to  lakes
and streams in Wisconsin, and Vollenweider (1968) has  provided  a broad
ranging review on  both  nitrogen and phosphorus sources  for  lakes.

Table 2 summarizes the  ranges of nitrogen and  phosphorus  contributions from
various natural  and  cultural sources and provides further references  which
have  evaluated or  reviewed  their significance.  Nutrient  contributions
from  undisturbed  forest land are generally considered to  be  small, but
fertilization  and  clear-cutting tend to increase  nutrient export (Cole
and Gessel, 1965;  Bormann e_t^ aJL  1968).   For example,  the former authors
found that^nitrogen  output  via percolotion water  from a Douglas fir
forest  increased  from 0.54  kjj./ha. to 0.69 and 1.04  kg./ha.  when plots
were  fertilized with  urea and ammonium sulfate,  respectively.  Clear-cutting
increased  nitrogen export to 0.96 kg,/ha., and this  practice also tends
to increase runoff and  sediment transport.  Bormann ^t a_l_. (1968) have
found large alterations in the nitrogen cycle of  a watershed after its
hardwood forest was  leveled and regrowth prevented by herbicide applications.
Greatly increased  nitrate levels were noted in stream water (greater than
lOmg. N/l.),  and nitrogen export rose from less  than 2 kg./ha.-yr. to more
than  6.0.   Thus properly managed and undistrubed forests retain their
accumulated nutrients  efficiently and not surprisingly should be of little
concern in eutrophication control.  However management and harvest prac-
tices can alter this  state of affairs and produce considerable nutrient export,
                                   23

-------
Source
Natural
              Table 2. Nitrogen and Phosphorus Contents
                     of Various Nutrient Sources 1
                     References
Forest runoff
Forest Percolation
water
Swamp and marsh
runoff
Meadow land and
runoff
Precipitation on
lake surface
Aquatic birds
Leaves and pollen
Cultural
Domestic sewage
Agricultural areas
Citrus
Pasture
Cropland
1.3-5.0
1.5-3.4
0.54
?
?
.58
.18-. 98
.48-. 95
?
39.4
22.4
8.5
1-5
5-120
.084-. 18
.83-. 86
.034
9
?
.044
.015-. 060
.09-. 18
?
0.80
0.18
0.18
0.15-0.75
0.22-1.0
Cooper, 1969;
Sylvester, 1961
Cole & Gessel, 1965
Vollenweider,1968
Lee et al. 1971

Vollenweider,1968
Brezonik e_t al . , 1969
Vollenweider, 1968
Hutchinson, 1957, Feth
1966; Veibel et al. ,1966b
Sanderson, 1954
Paloumpis and Starrett,
1960; Gates, 1963
Vollenweider, 1968
Goldman, 1961, Hynes
Kaushik, 1969, Richard-
son, et al. , 1970
Vollenweider, 1968 Saw-
yer, 1947; Mackenthun et al.
1964;Englebrecht&Morgan, 1959
Montelaro,1970
Miller, 1955
Vollenweider , 1968
Johnson ,et .al , 1965 Moe
1967; Biggar&Corey, 1969
Farm animals,
    feedlots
                     Vollenweider,  1968;  B. A.
                     Stewart,  1970;  Loehr,
                     1969; Miner et_ al  1966;
                     Hutchinson & Viets,  1969
Urban runoff
8.8
1.1
Weibel ej: al, 1964;1966ab
Weibel, 1969; Palmer,
1950; Sawyer, ^t al, 1945
Septic tanks
                      Polta,  1969;  Patterson
                      et  al,  1971
Marsh and landfill
    drainage	
                      Quasim,  1965
     units in kg/ha-yr. except rainfall  (g/m^-lake area-yr.), birds
(kg/duck-yr.), and domestic sewage  (kg/capita-yr.) .
                                  24

-------
Undisturbed meadowland is probably camparable to forestland in retention
of nutrients, but intensively used pastureland appears to be a somewhat
greater source.  In many areas pasture land is fertilized to increase the
yield of grass, and this fertilizer coupled with animal wastes undoubtedly
increases nutrient export rates.  Swamp and marshland have frequently been
cited as important both in trapping nutrients before they reach a lake
and in contributing to the nutrient budget of a lake (Benson, 1965; Lee
1970a), but quantitative information is sparse.  Recently Lee et. al. (1971)
described the annual nutrient balance in several Wisconsin marshes as
in approximate steady state with the nitrogen and phosphorus income equal
to the outflow.  During growing season in these temperate marshes nutrient
inflow is retained by the marsh but these nutrients appear to be exported
during a short period of high flow in spring.  On the other hand, draining
a marsh and converting it to dry land tends to release large amounts of
nitrogen and phosphorus as a result of soil oxidation and mineralization.
Pollen was recently shown to be an insignificant source of nitrogen for
Lake Tahoe (Richerson et. al., 1970), but it may be of greater importance
(a few percent or so) in small alpine or subalpine lakes with limited
nutrient inflow and greater shoreline to volume ratio.  Leaves are known
to exert a significant effect on water quality in streams (Slack and Felz,
1968), but quantitative  assessment of their role in lacustrine nutrient
budgets  is a difficult matter.

Nutrient contributions of water  fowl  are  not well known but may be  signi-
ficant in  some cases,  especially where large rookeries  and migratory bird
preserves  occur on  small water  bodies.  A community of  about 50 mallard
ducks maintained  for  research purposes on the  shore of  Lake Mize,  a  small,
deep lake  near Gainesville,  Florida,  was recently  estimated  to have  increased
the  nitrogen and  phosphorus  loading rates beyond those  suggested  as  cri-
tical for  eutrophication (Brezonik,  1971b).  Duck  contributions were cal-
culated  to represent  50  and  75  percent of  the  nitrogen  and phosphorus  loading
rates of 8.1 and  0.73 g./m?-yr., respectively.  Waste production  of  small
and  medium sized  birds (ducks,  chickens) have  been assessed in  relation
to pollution from commercial breeding, production,  and  processing (San-
derson,  1954;  Gates,  1963).   The role of wild  aquatic birds  in  the nutrient
regime of  a  water body may be more that  of  cycling agents  than  of direct
sources  (i.e.  much of their food may be  taken  from the  lake  itself - fish-
or from tt^e  immediate watershed).   However,  such  activities  can still stim-
ulate increased  production and  apparent  eutrophication  by  accelerating
nutrient cycling  and degrading a normally  more stable (organic)  nutrient
reservoir.

Agricultural sources of nitrogen are difficult to summarize because of
large local  variations in soil retention capacities,  in irrigation and.
fertilization  practices, and in differences associated with climate and
type of crops  grown.   However,  some data are available  to  indicate the sig-
nificance  of agricultural runoff as a nutrient source.   Johnston et. al.
 (1965) found large percentages of nitrogen applied to crop lands in the
San  Joaquin  Valley of California appeared in the  tile drainage effluents.
From 9 to  70 percent and phosphate concentrations ranged from 0.053 to
                                    25

-------
0.23 ppm., but even these  comparatively  low concentrations are sufficient
to stimulate algal growths.  Moe et^ aL_.  (1967)  found  considerably smaller
losses of nitrogen from soils  in southern Indiana.  A maximum of 15 per
cent of  applied  fertilizer (NH^NCs)  was  found  in surface runoff after
simulated heavy  rains.

Waste  output  for the  common farm animals (cattle, hogs, sheep, horses)
are  reasonably well known (Vollenweider, 1968;  B. A.  Stewart, 1970),
and  total  animal populations are readily available  in areal units at
least  as small as counties.  However,  the amount of animal waste which
enters and is transported by the source  waters  to a lake depends largely
on local circumstances, and few quantitative figures  are available.  Ammonia
volatilized from cattle feed lots has  been shown to contribute signifi-
cantly to  the nitrogen budgets of down wind lakes (Hutchinson and Viets,
1969).  For example,  this source was estimated  to add about 0.6 mg.
NH3~N/l.-yr.  to  Seely Lake, Colorado,  a  small  lake  2  km. down wind  from
a  90,000  unit cattle feed lot.  Ground  water  in areas of cattle feed lots
is frequently contaminated with high nitrate levels,  but present data are
insufficient to  allow prediction of nitrogen fluxes for individual  situa-
 tions .

Nitrogen contributions from sewered populations are well known on a per
capita basis, and the total nitrogen concentration  in raw and treated
domestic sexvage  is probably sufficiently constant to  permit evaluation
of this source for nutrient budget purposes from readily available  sewage
 flow and population records.  Non-sewered human waste contributions, as
 from septic tank drainage, are far more  difficult to  evaluate.  Septic
 tanks  are undoubtedly important factors  in the  cultural eutrophication of
many recreational lakes.  However there  are almost  no data available to
quantify this source  (see Patterson et al. 1971 for a detailed discussion).
Urban  runoff composition has been studied in detail in a few places
 (Weibel et_ al. ,  1964, 1966a; Palmer, 1950; Sawyer e_t  al. 1945), and storm
drainage flows are more widely known,  but the  general applicability of
existing information  outside the original study areas is unknown.

Rainfall has recently received attention as a  source  of nutrients,  especially
nitrogen.   Menzel and Spaeth (1962) correlated  the  ammonia content  of
Sargasso Sea surface  water with the rainfall of the previous five  days,
and  Gambell (1963)  found rainfall to be a major source of nitrate  and sul-
fate in certain  Virginia and North Carolina streams.  Chalupa  (1960) found
significant rainfall  contributions of phosphate to  a  Czech reservoir.
Parker (1968)  has found bloom stimulating concentrations of vitamin Bio
in rainfall of the St.  Louis area.  Weibel ej^ al.  (1964, 1966a,b)  reported
extensive  studies on  rainfall and urban and rural runoff as nutrient sources
in the Cincinnati,  Ohio, region. Cincinnati rainfall  averaged more  than
one  mg/1 in total nitrogen and 0.24 mgP/1 in hydrolyzable phosphate.
Reviews  on rainfall as a nutrient source have been  compiled by Feth (1966) ,
Hutchinson (1957)  and Vollenweider (1968), and if previous studies  have
confirmed  one point,  it is that nitrogen concentrations in rainfall are
(unfortunately)  highly variable.  Both nitrate and  ammonia occur in signi-
ficant amounts but  the latter is usually higher.  In  the tropics nitrate
is said  to be somewhat more important than in temperate rains but  even there
                                  26

-------
ammonia is usually higher (Hutchinson, 1957).  Total nitrogen concen-
trations frequently approach 1/mg. l./l., and both natural and cultural
sources are responsible.  The correlation of high rainfall ammonia levels
with alkaline soils and of low rainfall ammonia content in regions with
acid soils suggests that sorption of ammonia onto soil clays may be an
important factor in the hydrospheric nitrogen cycle (Feth, 1966).  Al-
though a number of cultural sources (ammoniated fertilizers, nitrogen
oxides in auto exhaust) would seem to make important contributions to
atmospheric fixed nitrogen, there is conflicting evidence regarding the
correlation of high rainfall nitrogen levels with areas of industrial
or cultural activity (see Vollenweider, 1968, and Feth, 1966, for details).
Careful analysis of historical records to discern possible long term
trends toward increased rainfall nitrogen resulting from intensified
inorganic fertilizer use, from increases in vehicular traffic and internal
combustion engines (sources of nitrogen oxides), or from other cultural
sources has not been reported.

While relatively large  amounts of data have been collected on nitrogen
in rainfall in several  areas of the country, little information is avail-
able for  the Florida peninsula.  Because of the variabilities in concen-
trations  and amounts, it is not yet possible to make reliable estimates
of rainfall contributions to nutrient budgets without a thorough field
study.  To further knowledge concerning these variabilities several
studies on nutrients in rain were conducted during the project.  The
great variations in amount  of rainfall within a small geographical area
are  indicated in Table  3.   r.ight  stations scattered 8 miles long were set
up in August, 1953, and 9 stations were similarly established during the
rainy season in August,  1%9.  Rainfall was collected for 7 days in 1968
and  during 3 days of intermittent showers in 1963, in pyrex bottles
through 4 inch dianeter glass 'funnels  (see Appendix A for analytical
details).  The results  from 1968  show an inverse correlation between rain-
fall amounts and nutrient  content,  in agreement with the role of rainfall
as a cleansing agent for the  atmosphere.  The marked variations  in rain-
fall amounts even in this  small  geographic araa were surprising.  Whether
the  sampling stations would have  such  large  differences  over longer periods
is not known, but presumably  the  differences would become small  over the
period of a year or so.  Ammonia  and  nitrate were much higher in the 1963
collections\than in I960, which may be  related  to differences in amounts
of rainfall during  the  days preceding sampling  in the  two years.  Since
rainfall  seems to  cleanse  the  atmosphere  a  large  amount  of  rain  prior
to a collection period  would  result in  lower concentrations  in  collected
samples.  The ratio of  nitrate to annaonia varied  markedly  among  samples  in
both years- this would  seem to  reflect  the  importance  of local  terrestrial
sources such as dust and airborne fertilizer since  the  larger scale  atmos-
pheric movements  and patterns  of  circulation would  tend  to  homogenize
rail!fall  composition over  as small  an area  as  sampled  here.   It is  interest--
ing  to note  that  orthophosphate  concentrations  in  the  samples were with
few  exceptions higher  than the  critical concentration  said  to stimulate
blooms  in lakes.   Thus, rainfall  might  stimulate blooms  rather  than  providing
a dilution effect;  this seems  especially  likely during summer when surface
waters  are nutrient depleted.

-------
          Table 3.  Variations in Nutrient Content  of
                      Rainfall in Gainesville Area
Sampling
 Station
Volume
Collected5
Ortho PO-P
        4
t-P04-P
NH -N
NO -N
1
2
3
4
5
700
370
150
120
115
 35
 27
                               August, 1968
0.03
0.23
0.40
0.13
0.07
0.06
0.29
          0.08
          0.79
          0.28
          0.81
          0.27
          0.79
          1.26
        0.004
        0.024
        0.006
        0.007
        0.007
        0.018
        0.008
                               August, 1969
0.25
0.41
0.40
0.52
0.52
0.94
1.11
A
B3
C
D
E
F
G
H3
I
250
135
290
20
130
180
130
290
170
0.006
0.043
0.009
0.018
0.011
0.018
0.004
0.021
0.002
.033
0.65
0.125
—
0.076
0.061
0.045
0.32
0.022
0.35
0.10
0.20
0.05
0.14
0.02
0.05
0.10
0.15
0.02
0.09
0.11
0.07
0.07
0.07
0.05
0.05
0.06
Concentrations in mg/1.

2Volume in ml collected in 1 liter pyrex bottle with 4 inch diameter glass
  funnel over 7 day period in 1968 and 3 day period in 1969.

3Sample contained large amounts of particulate matter evidently from
  dry fallout
                                   28

-------
While Gainesville is not heavily industrialized, the high nutrient levels
in Table 3 may be partly the result of man's activities.  To determine the
nutrient concentrations in rural rainfall, a series of sampling stations
were set up near Cedar Key, Florida, a small town on the Gulf of Mexico
about 65 miles from Gainesville.  Eight stations were set up within the
town and on a line from the Gulf up to 20 miles from the coart.  Results
for a rainfall in late August, 1968, are shown in Table 4.  Ammonia and
nitrate were considerably lower in these rain samples indicating possible
cultural effects in the Gainesville samples.  However, the Gainesville
rains were brief intense showers, while the Cedar Kay rain was steady fine
rain lasting for 24 hours.  Also the former samples were collected over
a week's time while the latter samples represent 24 hours of rain.  The
effect of saline Gulf water on the ionic content (especially Na ) of rain-
fall is clearly seen in the first two stations on Cedar Key.  These were
the only stations with detectable ammonia, but nitrate and phosphate show
no trends with distance from the coast.

Previous studies (Putnam and Olson, 1960; Weibel jBt_ _a_l. , 1964) have des-
cribed rainfall as a cleansing agent for the atmosphere.  Nutrient and
particulate concentrations in precipitation were generally found to decrease
during the course of an extended rainfall and to increase in rain with the
length of antecedent drought.  Several experiments were conducted in this
project to study the washing out of airborne nutrients over short periods
of time during rainfalls.  Rainwater was collected on the roof of the
Environmental Engineering Building at various time intervals (five minutes
to one hour) after the beginning of showers on several occasions during
August and September, 1969.  Samples were collected in pyrex bottles with
a large funnel and preserved with mercuric chloride for analysis of  nitro-
gen  and phosphorus forms.

Results from three rain showers  (Table 5) indicate a rapid cleansing of
nutrients from the atmosphere.  A  fourth rainfall  (Table  5) did not show
such marked decreases;  this was  a  light steady  rain rather than an intense
shower, and it was preceded by  a heavy rain on  the previous day.  Most  or
the  phosphorus in rain was present  as  ortho-phosphate.  Initial concentra-
tions were considerably higher  than the threshold  levels  for algal bloom
problems, but after  ten minutes  or  so, the  concentrations dropped to
0-lOyg P/l.  Organic nitrogen and  ammonia were  the predominant  forms of
nitrogen  in the rain, and  initial  levels were  comparable  to those in eu-
trophic waters.  Both forms  generally  decreased to O.lmg  N/l or  less
within an hour.  Nitrate  levels  were lower — tlo more than 0.08 ing N/l
initially -- and declined  to  trace  levels  rapidly.  A  trace of  nitrite
was  found on only one occasion.  The  rapid  decline of  nutrient  forms in
rain showers suggest  that  the nutrients may originate  as  low altitude
aerosols.  If this  is the  case,  the high  initial concentrations probably
reflect local,  cultural  activity (e.g. auto exhaust,  airborne  fertilizer
from plowed fields,  etc.), whereas  the low  concentrations afterward may
reflect natural or  general  levels  of  nutrients  in  the  atmosphere.

It should be apparent from the  above  results  that  estimation of  total
nutrient  contributions  to  a  lake via  rainfall  is not  so simple  a  matter
as one might have  first  thought.   In  particular, the  use  of literature
values for nitrogen  concentrations  is  likely  to produce inaccuracies
                                 29

-------
                                 Table 4.   Nutrient Content of Coastal and Rural Rainfall1
OJ
o
Location2
Cedar Key a
Cedar Key b
1
3
5
10
15
20
Amount 3
580
530
490
490
640
770
1070
710
KH3-N
0.08
0.03
0.00
0.00
0.00
0.00
0.00
0.00
NO-3-N
0.05
0.05
0.04
0.06
0.03
0.05
0.05
0.06
0-PO -P
4
0.012
0.036
0.010
0.007
0.027
0.028
0.024
0.010
Ca+2
1.34
0.61
—
—
0.30
0.31
0.22
0.31
W +2 »T +
Mg Na
3.05 11.0
0.24 2.0
—
—
0.05 0.26
0.04 0.38
0.05 0.45
0.03 0.31
K+
1.7
0.3
—
—
0.08
0.15
0.18
0.12
                     Concentrations in mg/1.

                     2Stations were located at the waters'  edge on Cedar Key (a), on the island interior (b),
                      and on a line approximately due east  from Cedar Key with the distances (in miles)  from
                      the coast as noted.

                     3Amount in ml collected in 1 liter pyrex bottles with 4" diameter glass funnels during
                      24 hour period.

-------
Table 5.   Temporal Variations of Nutrients in Rain Showers
           in Gainesville,  August and September, 1969:
Date
Time2
o-PO ,-P
4
t-PO,-P
4
TON
NH -N
NO^-N
4 Aug.  Type of rain:   thunder shower following weekend of dry weather
           0-5
           5-10
          10-20
          20-30
          30-60
0.056
0.032
0.010
0.005
0.005
0.058
0.032
0.015
0.005
0.005
1.85
1.00
0.83
0.10
0.92
0.84
0.09
0.09
0.08
0.08
0.00
0.00
0.15
0.10
0.00
10 Sept. Type of rain:  shower following day with trace of rain
           0-5
           5-10
          10-20
          20-30
          30-60
0.080
0.031
0.010
0.00
0.00
0.150
0.031
0.015
0.00
0.00
0.85
0.75
0.10
0.15
0.09
0.85
0.75
0.10
0.10
0.10
0.08
0.05
0.01
0.01
0.01
 16  Sept. Type of rain:   thunder shower following day with light shower
0-5
5-10
10-20
20-30
30-60
0.038
0.010
0.010
0.00
0.00
                               0.045
                               0.010
                               0.010
                               0.00
                               0.00
                       1.60
                       1.30
                       0.0
                       0.0
                       0.0
                    0.50
                    0.40
                    0.18
                    0.0
                    0.13
                 0.05
                 0.01
                 0.00
                 0.00
                 0.00
 3  Sept.  Type  of  rain:   light  and  steady  following  day  with  2  inches of rain
> 0-5
5-10
10-20
20-30
30-60
0.008
0.005
0.00
0.00
0.00
0.008
0.005
0.00
0.00
0.00
0.15
0.10
0.00
0.00
0.10
0.10
0.10
0.08
0.00
0.00
0.08
0.005
0.00
0.00
0.00
       concentrations in mg/1
  JTime in minutes  after start of rainfall
                                31

-------
because of  the large  local  and  temporal variations which  characterize
the nitrogen  content  of  rain.   For phosphorus  in rainfall there are almost
no data available  in  the literature,  so this  flux must be evaluated
experimentally.  In conjunction with  a concurrent project on eutrophica-
tion  factors  in  Florida  lakes,  the total nitrogen and phosphorus  contri-
butions  of  rainfall to the nutrient budget of  a  small lake were determined
in  1968.   A recording rain gauge was  established at  the lake site 30 miles
east  of  Gainesville,  and samples were collected  at weekly intervals for
ammonia,  nitrite,  nitrate and phosphate analyses. Rain was collected  in
a pyrex bottle with a 4-inch diameter glass funnel.  In order  to  prevent
contamination of collected rainfall by dust,  leaves, dead insects and
 other particulate matter a wad of glass wool was placed over the  funnel
 drain.  Results of this study are summarized  in  Table 6.   Because of the
 inverse correlation between nitrogen   content  and amount  of rainfall,
use of a simple mean value of total N for all  measurements times  the volume
 of rainfall on the lake surface would not be  very accurate.  Consequently
 nitrogen contributions were calculated for each  sampling  increment and
 these results then summed.  Rainfall  is a significant although not
 dominant nutrient supplier to Anderson-Cue Lake. If the  artificial nutrient
 input to this lake is not considered, rainfall then  is seen as an even more
 important  source, especially for oligotrophic  lakes  in nutrient impoverished
 watersheds (as is Anderson-Cue Lake).

 In summary, a large volume of raw data on nutrient sources is  available,
 but generalizations on  the importance of various sources  can in only  few
 cases be estimated.  A  few point sources (e.g. domestic sewage) are
 reasonably well characterized, and some diffuse  sources (e.g.  urban runoff)
 can probably be estimated with fair accuracy  using literature  values.
 However the quantitative aspects of nutrient  budgeting are still  relatively
 undeveloped.

 From a management point of view it is more important to determine the  contri-
 butions from man-made sources since they are  more readily controlled  than
 natural sources.  However the importance of natural  sources must  be known
 in order to determine whether control of man-made sources will result  in
 significant improvements in water quality.  The most significant  cultural
 sources of nitrogen are domestic sewage, agricultural  (cropland)  runoff,
 animal wastes from farming operations, and urban runoff.   Nitrogen fixation,
 rainfall,  ground water, and natural stream flow are  probsbly  the  most  sig-
 nificant natural sources.  Undoubtedly sediment deposition,  flow  through
 the outlet (in lakes) and denitrification are the largest sinks  for nitro-
 gen in natural waters.  Outlet losses would seem to  be  directly propor-
 tional to  the nitrogen  concentration within the body of water. Denitri-
 fication is limited in lakes to those that stratify  thermally  and  lose their
 oxygen in  the bottom water.

 The collected information on nutrient export  from various sources has  been
 used in conjunction with land use and population data  to  compute  nitrogen
 and phosphorus budgets  for 55 lakes  in north  and central  Florida  (Shannon,
 1970; Shannon and Brezonik, 1971a).  Because  this work was performed  pri-
 marily in  relation to a concurrent project (Brezonik,  1971c),  detailed
 results are not presented here.  However,  the findings are of  interest
 in relation to the preceding discussion  of nitrogen sources,  and  a summary
                                    32

-------
Table 6.  Contribution of Rainfall to Nutrient
               Budget of Anderson-Cue Lake
      Total rainfall in 1968:   136.7 cm (53.8 in)
      Lake area:
      Lake volume:
                  7.7 x 10* m2
                  1.85 x 105 m3
Nutrient
Rainfall nutrients in g/ra:

           Mean
Range
TON
NH3-N
NOo~'N
Ortho PO^-P
Total PO^-P
0.31
0.21
0.21
0.009
0.033
0.07-0.67
0.01-0.86
0.04-0.94
0.002-0.033
0.020-0.070
                         Budget
N
Source kg g/m3
Rainfall1 49.7 0.27
Artificial nutrient
mixture2 124 0.67
Other sources
(surface runoff)3 134 0.72
Total 308 1.66
Percent of total
contributed by rain 16
P
kg g/m3
3.4 0.019

10.6 0.057

4.4 0.024
18.4 0.100

18
 1Rainfall nitrogen levels were inversely proportional to amount
 of rain collected; budget value was calculated from total N
 measured at each collection times the amount of rain collected
 since last analysis and summing these amounts over the entire
 lakV surface.  Since relatively few total phosphate analyses
 were run and no inverse correlation with rainfall amount could
 be discerned, rainfall phosphorus in the budget was calculated
 from the mean for all samples (0.033 g/m3) times the total 1968
 rainfall volume over the lake surface.

 2Ammonium chloride and sodium phosphate dissolved in sewage as
 a controlled enrichment experiment (Brezonik et al., 1969).
 Calculated from watershed (land use) patterns from Shannon (1970)
 and Shannon and Brezonik (1971a).
                            33

-------
is presented below.  Watershed  land  use  patterns were  obtained  from recent
aerial photographs, lake  and watershed areas were measured from U.S.G.S.
topographic maps,  and population estimates  for each watershed were obtained
by counting the  dwellings visible on aerial photographs.  Nitrogen budgets
were  then  calculated  from appropriate export  rates  (i.e. kg N exported
per hectare of  a particular land use per year, kg N per  capita  per year etc.)
and the  measured watershed parameters.   Budgets were calculated per unit
area  of  lake  surface  and  also  per volume of lake water.  On an  areal basis
nitrogen loading rates  ranged  from 1.2 g./m.  -yr.,  for a small  oligotrophic
lake  (Swan Lake, Putnam County) to over  90  g./m.2-yr., for Lake Alice, a
shallow, hypereutrophic lake which receives 1.5-2.0 mgd  of treated sewage
from  the University of  Florida waste treatment plant.  Expressed on a
volumetric loading basis, nitrogen extremes were 0.18  to 106 g./m.J-yr.

Loading  rates generally correlated with  lake  trophic conditions.  In order
to study relationships  between lake conditions and watershed characteris-
tics  (e.g. nutrient loading rates),  a numerical index  (TSI) was developed
from  values  for 7 quantitative trophic indicators (e.g.  primary pro-
duction, total N and N  levels, Secchi disc  transparency).  Annual mean
values for the parameters from the 55 Florida  lakes and  the multi-variate
statistical  Technique of  principal component  analysis were used in this
derivation (Shannon and Brezonik, 1971b).   Increasing  values of the index
corresponded  to increasingly eutrophic conditions, and various  analyses
of the results suggest  the index does in fact  quantify the concept of
trophic  state.

Regression analyses of  the trophic state index (TSI) vs. N and  P
loading  rates (expressed  on both areal and  volumetric bases) were per-
formed using  an additive  model (TSI = f(N + P) and a multiplicative model
 (TSI  = f(N +  P) .  Some  representative results  are given  in Table 7.
In that  phosphorus was  incorporated into most  of the stepwlse regression
equations first and was hence  the most important variable in a  statisti-
cal sense, it might be  inferred that phosphorus rather than nitrogen is
the controlling input for eutrophication in Florida lakes.  However,
regression analyses are inherently eianirieal,  and such inferences must
be approached with ~" ''.-.:..

Of great interest in  control of cultural eutrophication  is the  develop-
ment  of  critical loading  rates for nitrogen and phosphorus, above which
one can  expect  eutrophic  conditions to ensue.   Sawyer  (1947) was the
first to propose quantitative  guidelines of this sort.   Based on data
from  Wisconsin  lakes  he suggested that 0.015  mg/1 of inorganic  phosphorus
and 0.3  mg/1  of  inorganic nitrogen at the spring maximum are critical
levels,  above which algal blooms can normally be expected.  In  the ab-
sence of any  other studies, these values have been widely quoted and
applied  to many  tvpes of  lakes in diverse geologic  and climatic situa-
tions.   Recently Vollenweider  (1968) analyzed the available data on
nutrient loading rates  and corresponding trophic conditions and proposed
permissible and  critical  loading rates  for nitrogen and  phosphorus as
a function lake  mean  depth. Using the  regression equations shown in
Table 7, similar critical loading rates  were  estimated for Florida
lakes (see Shannon and  Brezonik, 1971a,  or Brezonik,  1971c for  details
of this  derivation).  Results  are shown  in Table 8, and  VollenweiderTs
                                    34

-------
                                Table 7.  Regression Analyses of Trophic  State  Index  (TSI)
                    vs Nitrogen and Phosphorus Loading Rates for 55 Florida Lakes   (from Brezonik,  1971a)
                                                                                   Multiple    Percent Variance
                           Loading Rate                                        ,    Correlation  Explained by
           Model              Units                 Equation            F Ratio    Coefficient  Equation
           Additive
             (1)           per unit lake   TSI = 0.62(N  ) + 10.06(P  )   46.44        0.804        64.5
                              ,-                        o L»           o L»
                           surface area

             (2)           per unit lake   TSI = 26.1(P  ) + 0.90(N  )    43.20        0.793        62.9
                             1                         VL          V LJ
                           volume
Ln

           Multiplicative
                                                            / o      o r\
            (3)            Per unit lake   TSI = 0.84(P_T ) ' ' (NST ) *       14.08       0.600        36.0
                                                       SL      3L,
                           surface area

            (4)            per unit lake   TSI = 1.08(Ptrr ) '42(N  ) ^     15.64       0.620        38.5
                             ,                         VL      V L
                           vo lume
           'Abbreviations:  TSI = trophic state index (ditnensionless) ; Ng-^ and Pg-^ = nitrogen  and  phosphorus
                            surface loading rates in g/m2-yr; Ny^ and Py^ = nitrogen and phosphorus  volumetric
                            loading rates in g/m3-yr.

            All significant at the 99% confidence level.

-------
       Table 8.  Critical Loading Rates for Nitrogen and Phosphorus
                      Loading       Permissible Loading Dangerous Loading
  Reference          Rate Units           (up to)	(in excess of)
	N	P	N     -     P
Shannon and
Brezonik 1971a   Volumetric (g/m3-yr)   .86    .12      1.51       .22

Ibid.            Areal  (g/m2-yr)       2.0     .28      3.4        .49

Vollenweider
(1968)a          Areal  (g/m2-yr)       1.0     .07      2.0        .13
aFor lakes with mean  depths of 5 m or less.
                                    36

-------
(1968)  values for shallow lakes (z < 5m.)  are given for comparison.
While differences are evident between the two analyses, the estimates
are of  the same magnitude.  Florida lakes appear capable of assimilat-
ing somewhat greater quantities of nutrients before becoming mesotro-
phic or eutrophic than is indicated by Vollenweider's values.

The above values should not be regarded as final.  In a sense they are
quite unrefined because of uncertainties and even complete ignorance
regarding nutrient export rates from various types of terrain and land
uses.  However, as these values are experimentally determined the em-
pirical analysis described here should become more precise, and it
offers  the potential for quickly and economically predicting water
quality conditions from watershed characteristics.
                                    37

-------
                             SECTION VI

 NITROGEN FIXATION AS AN _IN SITU NITROGEN SOURCE FOR NATURAL WATERS

                     I.   ALGAL FIXATION IN LAKES
Aside from the external nitrogen sources discussed previously, several
mechanisms within a lake can effect an increase or decrease in the total
N content of the lake.   Nitrogen fixation is an in situ source, while
denitrification is a sink.   Sediments must act as nitrogen sinks over
the geological life span of a lake, but over shorter periods of time they
may act as sources to the overlying water.  The role of sediments as
a source or sink and of denitrification is discussed in Section VIII.

A wide variety of studies were undertaken to assess the significance
of nitrogen fixation as a nitrogen source in lacustrine systems.  For
the purpose of this report the studies can be grouped into four phases:
1.) an extensive survey of nitrogen fixation in the 55 Florida lakes
described in the previous section, 2.) an in depth study of fixation
in two eutrophic lakes, Bivin's Arm and Newnan's Lake, 3) investiga-
tions on bacterial fixation in lakes, and 4.) investigation of the nitro-
gen fixing properties of sediments.  The  first two phases are described
in this section, the latter two in Section VII.  Analytical procedures
used throughout these studies are described in Appendix A.

In order to estimate the overall significance of nitrogen fixation as
a nutrient source, fixation measurements  were included as part of a broad
ranging study of the limnology and trophic conditions in Florida lakes
(Brezonik, 1971c).  The 55 lakes included in  the study were of widely
varying trophic and chemical conditions  (Shannon and Brezonik, 1971c) ,
and most are located within a 40 mile  radius  of Gainesville.  All 55
lakes were sampled 4 times, and 19 of  them were sampled  7 times over
one year.  For ease in  sample handling,  all  incubations were  carried
out under ^Laboratory conditions (22°C  in the  light) in a shaker-bath;
samples were assayed on the same day  as  they  were  collected.  This pro-
cedure undoubtedly yielded results somewhat  different  from  in situ
data, but laboratory incubation permitted many more samples to be pro-
cessed.  For assay and  comparative purposes,  the method  is  well suited.

Assuming  the lakes represent a  cross-section of  Florida  lakes as  a whole,
it is pertinent  to consider  the frequency of  nitrogen  fixation  in them.
Out  of a  total  of  272  lake  samplings  (not including separate  samples
taken in  depth  profiles on  some of the lakes),  42  samples from  15
different  lakes  gave positive  fixation rates  using the acetylene  reduc-
tion method  (Table 9).  L?ke Mize,  in x^hich  fixation was  found  only  in
the  anoxic hypolimnetic waters is  not included  here, but the 7 samplings
                                     39

-------
      Table 9.    Summary of the Occurrence of
                Nitrogen Fixation in Florida Lakes
Number of lakes sampled:                                 55

Number of lakes in wbich fixation was detected:           ^5

Total lake samplings (not including separate
  depth profiles on some lakes) :                        272

Number of positive rates detected:                       42
Range of measured rates  (nM C2H^/l-hr) :                 1-262
                         (ng N/l-hr):                   93-2450

Frequency distribution of fixation:    //positive samples     #lakes
                                              1                5
                                              2                5
                                              3                0
                                              4                2
                                              5                0
                                              6                2
                                              7                1

Nitrogen fixing algae detected:  Anabaena sp. , Aphanizomenon
                                   40

-------
of Newnan's Lake and Bivin's Arm conducted as part of the 55 lake study
are included.   The infrequent sampling routine probably missed fixation
in 4 or 5 more lakes which seem likely environments for sporadic
nitrogen fixation (based on their trophic conditions and the nature of
their plankton).  Low (undetectable) fixation may be more widespread or
more frequent  in lakes giving occasional high rates, but no attempt was
made to concentrate the algae by filtration or centrifugation before
incubation.  It is doubtful whether such low rates (as requiring organism
concentration  for detection) would be ecologically significant.  Fixa-
tion was detected on only one occasion in 5 of the 15 lakes and in 5
other lakes it was found twice.  Overall fixation is thus not a typical
feature of the lakes.  However, in five lakes it apparently occurs
commonly; rates in these lakes are summarized in Table 10.  Rates ex-
pressed as equivalent N fixed in ng. NH-j-N/l.-hr. ranged from just
detectable (about 0.5) to a high of 175, but most rates were low (less
than 10).  However even rather low rates can contribute to a lake's
nitrogen budget if the rates continue for a long period.  For example,
fixation was found in Lake Dora on all 7 sampling dates ar.d in Lake
Hawthorne on 6 out of 7 sampling dates.  Mean fixation rates in the two
lakes are 116  and 127 np,. N/l.-hr.  respectively.  Assuming this mean
rate occurs over a year and that fixation occurs on the average for 8
hours a day, fixation would then contribute about 0.34 and 0.37 mg. N/l.
to Lakes Dora  and Hawthorne, respectively, on an annual basis.

Detectable fixation was found in the lake surface waters only in the
presence of blue-green algae, and the phenomenon seems limited to eutro-
phic or mesotrophic lakes.  None of the oligotrophic Trail Ridge lakes
gave detectable fixation, and species of algae capable of fixation were
not found in their plankton.  Anabaena and Aphanizomenon were the only
two nitrogen fixing algae aetected in the lakes, and the former was by
far the more common.

The seasonal pattern of nitrogen fixation in temperate lakes has been
well established  (Dugdale and Dugdale, 1962; Goering and Neess, 1964),
and fixation is primarily a late summer phenomenon.  In subarctic lakes
significant fixation is apparently  limited to parts of the short ice-
free season (Billaud, 1968).   In tropical and subtropical lakes seasonal
cycles are less pronounced  and  fixation is possible year-round, but
no data are available to substantiate  this point.  Consequently, a
detailed ye^r study  of nitrogen fixation was conducted on two highly
eutrophic  lakes near Gainesville in order to  (1)  determine seasonal
variations in nitrogen fixation,  (2) estimate the  total  contribution
of fixation to  the nitrogen budgets of  these  lakes,  (3)  determine the
environmental factors controlling  fixation,  and  (4)  relate fixation
activities to other  biological cycles  and phenomena  in  the lakes, viz.
nutrient cycles,  primary  production, algal blooms.

The two  lakes chosen for  the  detailed  study  are  Newnan's Lake  and Bivin's
Arm, both  of which  are  located near Gainesville,  Florida, and  are within
10 miles of each  other.   Both  lakes are highly  eutrophic, but  the chemical
characteristics and  causes  of  these conditions  in each  lake  are quite
different.  Newnan's Lake is  the  catchment  for  a large pine  forest and
                                41

-------
Table 10.  Nitrogen Fixation Rates in Selected Florida Lakes1
Date
1969
June
August
October
December
Bivin'
Arm

67.2
1020
0
0
s Lake
Dora

19.6
280
26.6
133
Lake Newnan '
Hawthorne Lake

134
392
9.8
49

0
0
19.6
384
' s Unnamed
#20

0
0
95
47



.2
.6
1970
February
April
June
165
16.8
440
57.4
88.2
210
151
506
0
133
30.8
0
0
39.2
26.6
ZA11 rates in ng N/l-hr, derived from nmoles ethylene produced/1-hr,
using theoretical factor of 1.5 moles ethylene produced per mole
ammonia fixed.
                           42

-------
and partially swampy area.   The drainage basin north of the lake is
rich in phosphatic minerals (Clark et al.,  1964),  and while some
cultural enrichment may enter the lake as runoff  from fertilized pas-
ture and leachate from City of Gainesville sanitary landfill operations
northeast of the lake, Newnan's Lake appears to be largely naturally
eutrophic.   The bathymetry of Newnan's Lake is shown in Figure 3; its
large surface area (2433 ha.) and shallow depth (z = 2m) coupled with
its loose,  unconsolidated organic sediments insure frequent wind
generated resuspension of the surficial sediments  and their recircula-
tion to the surface waters.  On the other hand, Bivin's Arm receives
much of its nutrient budget from cultural sources, including a stream
which drains a large urbanized area and into which poorly treated
sewage has  in the past been dumped, septic tank drainage, and cattle
wastes from University of Florida experimental farms on its western edge.
A large rookery on the northern shore of the lake may supply substantial
nutrients and effect a more rapid cycling of nutrients in the lake.
This lake is also shallow  (z = 1.5m; see Figure 4 for a bathymetric map),
and even in the absence of cultural inputs the lake would probably be
highly productive.  In their presence the lake can only be described
as grossly enriched.  Algal blooms are present continuously through
the year. Secchi disc transparencies greater than 0.6m  (2 ft.) are
uncommon, and the water frequently has a rather sickly green color.

Table 11 summarizes the chemical characteristics of these two lakes.
Newnan's Lake is highly colored from the swamp and pine forest drainage,
and its water is both low  in conductance and in hardness.  The pH of  the
lake is neutral to slightly  alkaline, but the  low alkalinity implies
a poor buffer capacity.  A rather high iron content  (average  .23 mg/1)
is associated with the high  color, but manganese is very  low as  are  other
heavy metals.  Total  phosphate in Newnan's Lake, while high and  re-
flecting eutrophy, is less than one-fourth  the level  in Bivin's  Arm.
Bivin's Arm has considerably more  dissolved solids  and  hardness  which
probably arise from cultural sources.  Its  pH  is usually  greater than
that of Newnan's Lake and  its  alkalinity is much greater.   Organic  color
is comparatively  low  and  that which  is present is  at  least  partially
autochthonous.

The biota  of  these  two  lakes  is  somewhat different  although blue-green
algae are  dominant  in both.  Aphanizomenon  is  the  predominant bloom former
during winter in  Newnan's  Lake whereas Anabaena nay bloom in  summer;
Microcystis  is  commonly  present  in abundance  in both lakes.   Anabaena
is the  common nitrogen  fixing  bloom former  in Bivin's Arm,  but  the  flora
of this  lake  are  much more diverse than  in  Newnan's Lake, and several pulses
of diatom  blooms  also occur  during the  year (Harper,  1971).   Water  hya-
cinths  (Eichornia crassipes)  have  been  a serious  problem in both lakes,
and  their  surfaces  have been  nearly covered  by  this species  within the
recent  past.   Chemical  spraying now keeps  the growth of hyacinths in
 check,  but the effect of the herbicides  employed  on other parts  of  the
ecosystems is a matter of question and  concern.

Figure  5 shows the concentrations  of major nitrogen forms,  rates of
nitrogen fixation and primary production in Bivin's Arm from May, 1969,

-------
1
N
           FIGURE 3,   SATiYfTETRIC MA" OF  :Tr'Av/c: LAKE,
         ALACHUA COUNTY,  FLORIDA,   DE^TH Ca^TOURS IK FEET,
                               44

-------
1
"
                                        FIGURE L\,   BATi-iYMETRIC MAP OF 3WIN'S ARM/


                                      ALACHUA COUNTY,  FLORIDA,  DEPTH CONTOURS IN FEET,

-------
Table 11.  Chemical Characteristics of Newnan's Lake and Bivin's Arm1
         Parameter2         Newnan's Lake        Bivin's Arm
Dissolved oxygen
PH
Alkalinity (as CaCO )
Acidity (as CaC03)
Conductivity (y mho cm"1)
Turbidity (JTU)
Color (as Pt)
Total organic N
Total P
Cl~
S04=
Na +
K+
Mg+2
Ca+2
Fe
Mn
COD
8.6
7.8
7.3
1.0
65.7
4.8
235
1.55
0.10
11
3
8.4
0.6
1.5
5.3
0.23
0.002
68
10.0
9.0
105
0
267
11
48
1.94
0.45
16
7
11.5
1.7
4.7
30.3
0.04
0.002
60
          values  for  six samplings  (every 2 months) over the period
     June, 1969 to April, 1970.

     Concentrations in mg/1 except  as noted.
                                 46

-------
1.0 _
                                                                                M
            FIGURE 5.   INORGANIC  NITROGEN  LEVELS, NITROGEN  FIXATION AND
      PRIWRY PRODUCTION  IN BIVIN's ARM/ FLORIDA, FROM MAY, 1969  TO MAY, 1970,

-------
tc May, 1970.  Routine  data  for  each  sampling data are  listed in
Appendix B.  Fixation occurred at  moderate  to high rates  throughout
the summer of  1969;  a maximum of 330  nmoles ethylene/l.-hr.  (equivalent
to about   3.1]jg. N/l.-hr.) was found  in  late August.  Lower  rates were
noted  during winter, and fixation  was detected on about 50 percent of
the sampling dates.  Fixation rates in this lake do not seem to be signi-
ficantly  correlated  with either  primary  production or nitrogen concen-
trations.  A rather  definite seasonal trend in primary production is
apparent,  with peak  rates in summer (June to October) and  lower rates
from  November  to May, but the pattern of algal activity is really much
more  complex,  being  characterized  by  a pronounced short-term variation
in both production and  nutrient  levels.

Seasonal patterns  of nitrogen forms,  fixation rates and primary pro-
duction in Newnan's  Lake are displayed in Figure 6, and routine data
 collected on each  sampling date  are listed  in Appendix B.  In spite of
 the geographical proximity and similar enriched conditions in the two
 lakes, their patterns  of fixation  are quite dissimilar.   Except for
one high rate  of nitrogen fixation in July, 1969, during  a bloom of
Anabaena and Microcystis and for very low rates on several occasions
 in spring and  fall,  fixation in  Newnan's Lake is associated  with a
 dense winter and early  spring bloom of Aphanizomenon. In winter of
 1969-1970 the  bloom occurred in  two pulses.  The first pulse began in
mid-December and died  out in early January, apparently because of ad-
verse weather  (an abrupt cold wave and an extended period  of heavy rain)
The population stayed  low in January  and gradually increased during
 February to  high levels through  most  of  March.  Maximum fixation rates
 during the Aphanizomenon bloom were about 40 nmoles ethylene/l.-hr,
 roughly equivalent to  0.37yg. N/l.-hr.  Fixation was highly  correlated
with primary production during  the winter bloom period  (Figure 7) in
Newnan's Lake, but no  obvious correlation exists between  fixation and
 concentrations of nitrogen forms.

 It should be noted that algal fixation occurred in both lakes in the
 presence of moderate to high (i.e. for lake waters) ammonia  levels.  In
 fact the maximum rate  measured  in Bivin's Arm occurred when  the ammonia
 concentration was  0.5  mgN/1. High rates in Newnan's Lake occurred at
 ammonia levels as  high as 0.6 mgN/1.   Thus  while fixation occurs fre-
 quently on nutrient depleted waters,  it  is  not limited  to them.  Pre-
vious studies  have demonstrated  simultaneous assimilation of ammonia,
 nitrate and  molecular nitrogen  (Dugdale  and Dugdale, 1965; Billaud,
 1968); no doubt this also occurs in the  lakes studied here.

Several measurements of areal variability in nitrogen fixation and
related parameters were made on Newnan's Lake during the  project in
order to evaluate the homogeneity (or lack  thereof)  of  biogenic para-
meters in the  lake and hence determine the  number  of sampling stations
required for adequate representation of  the lake.  Table  12  presents
 the results  of an areal study conducted  on  April  19, 1969.   Top, mid
and bottom samples were composited at 10 stations  scattered  around  the
entire lake.  All  samples were  collected within  one  hour  and brought  to
shore for processing.   Primary  production  and nitrogen  fixation samples
                                  48

-------
 MJJASONDJF
                                  MONTH
  FIGURE 6.  INORGANIC NITROGEN CONCENTRATIONS, NITROGEN FIXATION, AND
PRiriARY PRODUCTION IN NEWMAN'S LAKE/ FLORIDA FROM MAY, 1369 TO fW, 1970
                                  49
M

-------
Ul
o
              f
              c
              o
              >,
              i
                280
                240
                200
                160
                120
                80
                40
                 0
                               I
                           I
                                                                                           I
                                                   28
                                                   24
                                                   20
                                                   16
                                                   12
                         11      21

                             DEC,
31    10
     20
JAN,
9     19

 FE3,

  DATE
1      11

 MAR,
21    31

     APR,
                            FIGURE 7,  TEMPERATURE, PRIMARY PRODUCTION AND NITROGEN  FIXATION
                                IN NEMAN'S  LAKE DURING '-/INTER (1953-1970) APHANIZQMENON BLOOM,
                                                                                                       c
                                                                                                       o
                                                        PH

                                                        C
                                                        


-------
       Table 12.  Areal Variations in Physical and Biogenic Parameters in Newnan's Lake, 19 April, 1969
                                       Station
Mean and 95%
Parameter1
Sfc. Temp.
Secchi
Disc
Tot. depth
D. 0.
pH
Alk.
TON
NH3-N
NO^-N
o-PO^-P
t-PO^-P
Prim. Prod.
1
22.0

0.6
1.35
6.40
7.05
10.4
1.51
0.61
0.07
0.023
0.167
20.8
2
22.4

0.6
1.5
6.00
7.05
10.7
1.64
0.71
0.03
0.010
0.144
32.4
3
22.4

0.6
2.0
5.80
6.90
9.95
1.55
0.78
0.03
0.010
0.156
43.5
4
22.5

0.6
2.0
5.80
7.05
10.0
1.68
0.82
0.02
0.010
0.184
71.0
5
22.3

0.6
2.6
5.90
7.00
10.0
1.40
0.84
0.02
0.012
0.152
41.7
6
22.3

0.6
1.8
5.70
7.00
10.1
1.77
0.76
0.02
0.010
0.108
92.0
7
22.5

0.6
1.8
4.50
6.95
9.71
1.88
1.06
0.02
0.012
0.188
51.1
8
22.0

0.6
1.8
5.90
7.00
10.1
1.48
0.77
0.03
0.019
0.160
38.8
9
22.1

0.6
1.8
5.50
7.00
10.3
1.46
0.94
0.04
0.025
0.140
16.5
10
22.4

0.6
1.8
6.50
7.05
10.4
1.44
0.92
0.00
0.020
0.160
51.1
Confidence Coefficient o
Interval Variation (7<,
22.3

0.6
—
5.8 ± 1.0
7.00 (median)
10.2 ± 0.5
1.58 ± 0.29
0.82 ± 0.24
0.028 ± 0.003
0.015 ± 0.011
0.162 ± 0.029
46.0
—

—
—
28.3
—
8.3
30.4
47.6
17.9
118
28.7

Temp, in °C; Secchi disc visibility and total depth in meters;  pH =  -log  (H  );   alkalinity  in  mg/1  as  CaCO  ;
primary production in mg C/m3-hr;  all other values  in mg/1.

-------
shore for processing.  Primary production  and  nitrogen  fixation samples
were then returned  to  the lake for incubation  at  a near shore  station.
Although nitrogen fixation was not occurring in  the  lake  on  this day,
the data are  nevertheless useful in describing the variability in bio-
logical  conditions.  The variations in biogenic  parameters was much
larger  than in the  physical parameters and in  pH and alkalinity.  Pri-
mary  production had by far the greatest variability  - from  21  to 92 mg
C/m-hr.   Newnan's  Lake  obviously cannot be considered  homogeneous, and
any one sampling station may not be representative  of the entire lake
at a given time.  In order to obtain representative  data  and still keep
the sampling and analyses within reasonable limits,  it  was  decided to
 composite 6-8 stations scattered at random around the lake  into one
sample for the routine bi-weekly analyses.  Thus the data in Figure 6
 represent determination on a composite lake water sample  and are felt
 to reflect closely the average conditions in the lake at  a  given time.
Because Bivin's Arm is a much smaller and presumably better  mixed lake,
 it was not felt necessary to composite so many stations.  Consequently
 the data in Figure 5 represent composite samples fros three  stations
 and 3 depths.

 A second areal variation study was conducted on  Newnan's  Lake  December  8,
 1969, at the onset of the Aphanizomenon bloom. Again 10 stations were
 sampled and nitrogen fixation was  found at each  station.  Figure  8 shows
 the distribution of primary production and nitrogen  fixation rates at
 the 10 stations; the former varied in a much smaller range  on  this date
 than in the April  study.  Nitrogen fixation was  much more variable  than
 primary production and  the two rates do not seem correlated.  Table  13
 presents depth variations  found at two stations  and  summarizes the
 variations found in nitrogen  fixation, primary production and  chlorophyll.
 It is interesting  to note  that areal variations  with the  latter
 biomass measure  were    somewhat  smaller than those  of  primary produc-
 tion.  The large vertical  differences in these parameters in so  shallow
 a water column  are probably  the result of self-shading.  Note  that
 chlorophyll  decreased less than did the activity parameters -  primary
 production and  nitrogen fixation  which are light  dependent phenomena.

 Diel variations  in Newnan's Lake  were measured on thr?e occasions.
 •\pril 21,  196?, December 16,  1969 and April 20,  1971.  Nitrogen fixa-
 tion was found  in  the lake only in the December  study.   The April,  1971,
 study apparently just missed  the  annual Apnanizomenon bloom, which  was
 quite evident in the  lake  a  few days  earlier.   Figure 9 shows  the time
 course of primary  production,  nitrogen  fixation and light intensity
 during the December  16,  1969,  study.  Sampling began at 7:00 A.  M.  and
 continued at  about one  and a  half hour  intervals until dark.  Primary
 production reached a  maximum value at mid-morning and decreased markedly
 during the afternoon, at least  partly because an increased   cloud cover
 reduced light intensity.   Peak nitrogen fixation lagged behind photo-
 synthesis and occurred  in  early afternoon.  Although sampling was not
 continued after  dark  the data suggest  that nitrogen  fixation  is limited
 to daylight hours  or  at least occurs  at much  lower rates during the night,
 Figure 10 presents some results of a  more detailed diel study conducted
 in April,  1971.  Unfortunately the lake was not  fixing nitrogen at this
                                   52

-------
FIGURE 8,  AREAL VARIATIONS IN NB^IAN'S LAKE/ DECEfBER 8, 1969,
    STATION NUMBERS IN PARENTHESES; UPPER VALUE: PRIORY
      PRODUCTION IN m§ C/m3-hr; LOWER VALUE1. NITROGEN
                   FIXATION IN W N/m3-hr.
                              53

-------
               Table 13.  Vertical and Areal Variations in
Nitrogen Fixation and Related Parameters in Newnan's Lake, 8 December 1969
         Depth     Primary    Nitrogen
           m.    Production   Fixation    Chlor. a_.   Temp,
                 mg C/m3-hr   nM N/l-hr   mg/m3       °C

                                Station 4

         Sfc         38         20         70          13.8
         1           19         14         69          13.3
         2           20         10         64          13.3

                                Station 8

         Sfc         38         19         70          14.8
         2           19         18         71          13.6
         4           17          6         64          13.3
                       Summary of Areal Variations1

         Mean        36.3       18.1       67.4
         Range     31.0-44.1  9.8-33.6   60.7-75.1
         1Means of ten stations, eight of which were samples
          composited with depth and two of which (stations 4 and
          8 above) were samples taken at discrete intervals.
          Mean values over depth for these two stations were used
          for calculation of the areal means.
                                   54

-------
0.7
                             PRIMARY PRODUCTION
    0700
                                                                  1730
                FIGURE 9,  DIEL VARIATIONS IN NEWNAN'S LAKE/
                            DECE^ER 16, 1969,
                                   55

-------
  26
~  24
 G
 3
 «  22
 
-------
time but the data are interesting in illustrating the large changes
in biogenic parameters that can occur on a daily basis in eutrophic
lakes.  The lake was very calm during this study, and a slight,  tem-
porary temperature stratification occurred during the day.   Dissolved
oxygen values rose during daylight hours but fell much below saturation
after nightfall.  The small decline in oxygen after several hours of
sunlight seems to be a case of photorespiration, as recently described
by Odum (1971).  The total change in oxygen content during the day was
dramatic; considering the calm conditions most of the increase would
appear to arise from photosynthesis rather than gas transfer from the
atmosphere.  Nutrients declined somewhat irregularly during the day-
light hours, but the ratios of carbon, nitrogen and phosphorus net assi-
milation during the period of peak photosynthesis can be approximated
from the data in Figure 10.  During the period 9:30 A. M. to 4:15 P. M.
photosynthesis resulted in a net increase of 2.1 mgO~/l in the surface
waters.  Before and after this period photosynthesis was less than or
about equal to respiration as the dissolved oxygen values indicate.
During the period of maximum photo synthesis a net decrease of 0.16  mg
NHy-N/l and 0.022 mg ortho PO^-P/1 occurred in the surface water.  Con-
verting the oxygen increase to equivalent carbon fixed indicates a net
of about 0.75 mgC/1 assimilated.  On a molecular basis then these
values indicate net C:N:P  uptake ratios of 100:18:1*1 on a molar basis,
xtfhich ratios are quite similar to the average C:N:P composition ratios
reported for algal cells  (i.e. 106:16:1).

The rather large diel changes in biogenic parameters  reflect the dynamic
nature of nutrient cycling especially in eutrophic lakes.  These changes
also imply that routine sampling for seasonal or other long term studies
should be conducted at approximately the same time of day.  Otherwise
changes resulting from diel cycles may erroneously be thought of as
reflecting longer term variations.

The severity and duration of  the annual Aphanizomenon bloom in Newnan's
Lake is apparently quite  variable.   In winter 1968-1969  the bloom occurred
from late December to mid February,  and at  its height was more intense
than during the following year.  Two samples  taken near  shore on January 20,
1969, were thick with clumps  of Aphanizomenon,  and yielded  fixation  rates
of 70 and  163  nM C2H,/l.-hr., which  values  are  considerably in excess
of the maximum values for the winter 1969-1970 bloom.  On  the other  hand
massive concentrations of the algae  near  shore  frequently  result  from
sustained winds, so the above values are  not  necessarily representative
of fixation rates in  the  entire  lake during the  January, 1969, bloom.
The bloom  of Aphanizomenon  in 1971  appeared much  later  (late March)  and
was of much shorter duration  than  in the  previous  year.  These differences
are no doubt associated with  differences  in weather  conditions (rainfall,
temperature) from year to year.  However  this qualitatively simple
explanation would be  very difficult  to  verify in a quantitative manner,
and attempts to correlate the bloom occurrence  and duration with  climate
data  collected during this  study were not  fruitful.   Perhaps  a detailed
analysis of  past weather  records  and bloom conditions (if  they were
available)  could yield an empirical  explanation,  but  that  is beyond  the
scope of this  project.  While the  details  and absolute amounts of fixa-
tion  in  the  lakes may vary  yearly,  the  pattern  seems  qualitatively  similar
                                 57

-------
from year to year.  Thus  the  data presented here are regarded as typical,
and in this sense can be  utilized for more general conclusions on the
importance of fixation  in Florida lakes.

An obvious but unanswered question  about  the  patterns of  fixation in these
two lakes concerns  its  occurrence primarily as  a winter phenomenon
in one  lake but  as  a  summer phenomenon  in the other lake  in spite their
proximity and essentially identical climate.  The  large volume of data
(Appendix B) collected  during the Aphanizomenon bloom in  Newnan's Lake
during  winter,  1969-1970  has  been scrutinized and  subjected to statis-
tical analysis  in order to elicit trends  and  potential causal factors
for  the appearance  and  eventual demise  of the bloom.  However there does
not  appear  to be any  clear answer,  and  bloom  formation and disappearance
 (hence  also  the  occurrence of nitrogen  fixation) may be controlled by
a subtle interaction  of physical and chemical factors, or by biological
causes  that were not  measured.   Several possibilities are suggested:
initiation  of  the bloom by temperature  changes  (e.g. an increase in
 temperature  following a cold  period during which Microcystis and the
other algae  present in  abundance during the bulk of the year are elimi-
nated) ; changes  in light intensity  or in photoperiod, or  perhaps a com-
bination of  temperature and light conditions  are necessary.  There do
not  appear  to  be any  substantial chemical changes  in the  lake prior to
bloom onset, but the  possibility of a biologically induced change in an
   organic  growth factor or in a required trace  metal such as iron or moly-
bdenum  should  not be  dismissed.  Possible reasons  for bloom disappearance
 are  even more  numerous.  This strain of algae may have rather narrow
 temperature  requirements, and the pulse shown in Figure 7 can be ex-
 plained on  this  basis.   When  the lake temperature  declined because of
 a severe cold  spell in  early  January,  1970  (Figure 7), the bloom rapidly
died out.   Only  after the temperature rose above 14°C in  late February
were the algae  able to  reach  dense  bloom conditions again, but once the
 temperature  rose above  about  20° in late March  the algae  again died out
or alternatively were no longer able to compete successfully against
other forms.   In support of this hypothesis a multiple regression ana-
lysis (Table 14) of primary production  vs. temperature, light, ammonia
and  ortho phosphate in  the lake during  this bloom  found temperature
to be the most  important variable  (statistically).  In fact using a
multiplicative model  (logarithmic transformation)  temperature alone
explained nearly as much of the variance as die! all four  variables com-
bined.   However  temperature alone is probably too  simple  an explana-
tion  and there are  discrepancies in the argument.  There  is no informa-
tion  in the literature  suggesting that  Aphanizomenon has  such a narrow
temperature tolerance (Hammer,  1964).   In fact  blooms of  this algae in
temperate lakes  (e.g.  in Wisconsin) are common  in mid and late summer
when  water temperatures are between 20  and  25°C.   Further, the drop in
Newnan's Lake temperature in  January 1970 was also associated with large
quantities of rain  which  raised the lake level  as much as three feet.
Consequently the Aphanizomenon bloom was both diluted and transported
out of  the lake  in  the  unusually high out-flows generated by the rain.
Alternative explanations  for  the blooms demise  also exist, including
the possibility  of  nutrient exhaustion  (probably a trace  nutrient since
inorganic phosphate remained  low but not depleted  during  the entire bloom
                                  58

-------
             Table 14.  Limiting Factors for Primary
 Production in Newnan's Lake, Florida During Aphanizomenon Bloom
                  Multiple Regression Analysis
Dependent Variable:    Primary Production
Independent Variables: Light, Temperature, NH-j-N, Ortho-PO^-P


1.  Additive Model:

    Y = a1X1 + a2X2 + a^ + a^


    PP = 4.76(Temp)  + 4.32(o-P04> - 6.05(Light) - 69.4(NH3~N)

                Var.             „            Increase
               Entered	R	In R2
                 Temp.0.65790.6579
                 o-PO,         0.6812          0.0233
                 Light         0.6953          0.0141
                 NH?	   0.6995	0.0042

               *   9
                 Rz - measure of portion  of variance accounted
                 for by  regression
 2.  Multiplicative Model:

    Y = X blX b2X b3X b4
         J-   2.   j   4
     InY = b  InX  + b2lnX2 + b3lnX3 +
    PP =  1.65 (Temp)  -  0.26(Light)  -  0.14(o-P04> + 0.07(NH3-N)
Var.
Entered
Temp.
Light
o-P04
NH -N
R2
0.9416
0.9425
0.9432
0.9433
Increase
In R2
0.9416
0.0008
0.0007
0.0001
                               59

-------
and nitrogen would not  seem a limiting element  to  a nitrogen  fixing
bloom).  Viral attack is  currently  a popular explanation among aquatic
scientists  for otherwise  unexplainable bloom disappearances, as is
autointoxication  (i.e.  synthesis  of a substance by  the algae which
causes their  own  death),  but neither possibility has ever been proved
to  cause bloom death in the environment.   In brief, the complexities
of  algal bloom cycles are still beyond simple explanation and  compre-
hension.  While many causes can be  conjectured, unfortunately  they
remain only that.

Seasonal patterns of nitrogen fixation, nutrient concentrations, pri-
mary production and  other biogenic  parameters are poorly defined in
these Florida lakes  in  comparison with classic cycles reported for
temperate  lakes.   Rather, production, nutrients and algal populations
are characterized by a  bloom-crash  cycle  throughout the year.  Pro-
nounced differences  were  noted between sampling dates even when sampling
was done  as frequently  as twice a week.  The large measured changes in
chemical  and biological parameters  from one sampling to the next were
matched by  obvious visual changes in the  lakes, and many of the rapid
 fluctuations can  be  ascribed to rapidly changing vjeather conditions.

The diel  variations  coupled with  the large oscillations in fixation
 from one  sampling date  to the next  obviate accurate assessment of the
total quantities  of  nitrogen fixed  in the lakes during a year.  How-
ever, the  data in Figure  5 and &  should permit at least a rough estimate.
Assuming  nitrogen fixation is primarily a daylight phenomenon, and during
periods of  fixation  it  occurs in  the lakes for about 8 hours a day at
the rates  shown  in Figures 5 and  6, the nitrogen fixed in the  2 lakes
over the  year study  was calculated, and the results are summarized in
Table 15.   While  the total N fixed  is large in both lakes, fixation
represents  a relatively small contribution to the total nitrogen bud-
gets previously  calculated for these lakes by Shannon (1970) (see Sec-
tion V).   Based  on the  survey of  Florida  lakes described earlier, it is
unlikely  that nitrogen  fixation would play a substantially greater
role in the nutrient budgets of other Florida lakes.  Kence, nitrogen
fixation  in lakes seems to be of  greater  significance in an ecological
sense than  as a nutrient  source.
                                     60

-------
       Table  15.   Contributions  of  Fixation  to  the
      Nitrogen Budgets  of Newnan's  Lake  and  Bivin's Arm



Newnan's Lake
Bivin's Arm

N Fixed

0.15
0.52
External
N Supply
g/m3yr
2.60
6.86
Total Percent of
N Supply Total Supplied
by Fixation
2.75 5.5
7.38 7.0
'•Calculated from watershed characteristics (land use and population
 patterns)  by Shannon and Brezonik (1971a) - see Section V.
                               61

-------
                             SECTION VII

 NITROGEN FIXATION AS AN IN_ SITU NITROGEN SOURCE FOR NATURAL WATERS

          II.   BACTERIAL FIXATION IN LAKES AND SEDIMENTS
While numerous studies have identified and isolated nitrogen fixing
bacteria from lakes and marine waters (see Section IV), most previous
investigators have considered heterotrophic (bacterial) fixation to
be of little consequence.  A lack of oxidizable carbohydrate has usually
been cited as the limit-rrrg—factor for heterotrophic fixation in both
soils and natural waters ( Stewart  1969).  Various reports (see Stewart,
1966, for a review) indicate that heterotrophic fixation under labora-
tory (pure culture) conditions is highly inefficient in terms of nitro-
gen fixed per unit of carbon oxidized.  However it is always dangerous
to extrapolate highly artificial laboratory results to in sjLjtja con-
ditions.  Studies conducted as a part of this project indicate that
heterotrophic (bacterial) fixation can be important in certain anoxic
lacustrine environments as well as in sediments.  In addition the ability
of photosynthetic bacteria to fix nitrogen is well known (Stewart, 1966)
and while these organisms occupy a rather special niche (i.e. they
require anoxic conditions and low light intensity) and are not generally
abundant in the environment, they may fix substantial amounts of nitro-
gen in favorable habitats.  For example, Stewart  (1969) correlated
significant uptake of :5N2 at a depth of 7m. in  a Norwegian fjord with
the presence of the photosynthetic bacterium Pelodictyon.  Triiper and
Genovese (1968) have  reported high concentrations of photosynthetic
bacteria occurring at intermediate depths in certain  Italian lakes.

The first evidence for  bacterial nitrogen fixation in  lakes was re-
ported by Brezonik and  Harper (1969)  in work supported by this project.
They found fixation in  two dystrophic lakes chosen for their extensive
anoxic environments.  Lake Mary, Wisconsin, is  a  colored meromictic
lake, permanently  anoxic below 5 meters  (maximum  depth = 21 meters).
Lake Mize, Florida, is  a dystrophic,  highly colored  lake with a maxi-
mum depth around  25 meters and a surface area of  0.9 ha.  The morpho-
metry of Lake Mize (Figure 13) does not  promote good  vertical circula-
tion, but the lake is monomictic with a  circulation  period  from October
or November  to the end  of  February or early March depending on weather
conditions.  Oxygen is  lost  during the  long period of  stratification,
and the  lake Is anoxic  below about 3  to  5 meters  from June  to November.

Lake Mary was sampled June 28, 1968,  and Lake Mize on July  11, and
August  21,  1968.   Profiles of some pertinent lake conditions are  shown
in Figures  12 and 13  for Lakes Mary  and  Mize, respectively.  Rates
of acetylene  reduction  at  various depths  in the two  lakes appear  in
                                 63

-------
•
.
                                                      FIGURE 11


    BATHYMETRIC WP OF LAKE MIZE, ALACHUA COUNTY, FLORIDA. CONTOURS IN PETERS,  MAP COURTESY OF F, G, NORDLIE,

-------
                Temperature (°C) ; Dissolved  0~  (mg/1)
                 6           12           18            24
                              Dissolved Oo (mg/1)
                                    4        " 6
                                Temperature (°C)
                                   16         24
                                                                                                              32
  10
p.
cu
Q
  15  _
  20
                                                          20
                                                                                                              I
                 80
320
                     160          240
                  Color   (ppm  as  ft)
FIGURE 1?, DEPTH PROFILES FOR  TEMPERATURE (o),
    DISSOLVED OXYGEN (•) AMD COLOR  (A)  IN
                               28
60
            L^.!<-  MARY,  !'/ISCOMSr-L JUNE   /
       ARRO1-'  INDICATES DEPTH OF SECCHI DISC VISIBILITY,
                   120          180
             Color      (nr>^ as Pt)
FIGURE 15, DEPTH PROFILE?  FOR TEMPERATURE (o)/
    DISSOLVED OXYGEN  (•) AMD  COLOR  (A)  IN
    LAKE MIZE, AUGUST 1, 1%8,  ARRC"''  IMDJ-
    CATES DEPTH OF3ECCHI   DISC  VISIBILITY,
                                                                                                               40

-------
Table 16.  Rates measured  in  this  initial  study were  generally low
compared with  fixation  rates  found in  some eutrophic  surface waters
containing blue-green algae.   The  highest  rate in Lake Mary occurred
near  the bottom, but  in Lake  Mize  maximum  activity was noted at a
depth of  15  m.

A  detailed  study  of nitrogen  fixation  in Lake Mize was conducted
during  1969  and 1970  to determine  the  extent and significance of
bacterial fixation in the  lake (Keirn  and  Brezonik, 1971).  Table 17
presents  a summary of the  general  chemical characteristics of Lake
Mize.  The lake has a high but variable color, which  is evidently
 leached from pine needle litter in the drainage basin, and has typi-
 cal characteristics of  dystrophy,  i.e. low values of  pH, calcium,
 and primary productivity.   Figures 14  and  15 present  temperature and
 dissolved oxygen profiles  of  the lake  for  1969 and 1970.  In 1969 the
 lake was only slightly  thermally stratified in February but a marked
 depletion of dissolved  oxygen was  found in the hypolimnion.  By April
 anoxic  conditions had developed below  7 meters and by July no oxygen
 was present below 3 meters.  Stratification began somewhat later in
 1970, but dissolved oxygen was partially depleted in  the bottom waters
 during  February.   Anoxic conditions occurred below 5  meters by early
 April and below 3 meters by early  June; it is also interesting to note
 that dissolved oxvgen values  were  low  even in surface waters during
 the summer of 1970.

 Conditions in Lake Mize have  changed during the last  several years as
 a result of an increased nutrient  loading.  In fall of 1968 an enclosure
 housing about 50 ducks was placed  at the north shore  of the lake.  In
 response to enrichment, a lush growth  of emergent grass occurred
 along the previously nearly bare shoreline and the rate of primary
 production increased markedly.  Rates  ranging from 1.9 to 10.4 mg.
 C/m.3-hr. were found prior to 1970, but samples taken in April and June,
 1970, yielded rates of 42.3 and 36.2 mg. C/m.3-hr., respectively.  Con-
 centrations of nitrogen and phosphorus also increased substantially
 from 1968 to 1970 (Table 18), presumably reflecting duck enrichment.
 Low concentrations of ammonia occurred throughout the water column in
 spring  of 1969 and an inverse clinograde distribution occurred in summer.
By September 1969 epilimnetic ammonia  concentrations  surpassed those
 in the  bottom water.   Concentrations were higher throughout 1970 and
 generally showed more complex profiles.  Total organic nitrogen was
 rather  uniform throughout the water column during the two years, and
nitrate levels remained low  (typically < 0.05 mg. N/l.) but somewhat
variable during the study.

Nitrogen fixation as measured by the acetylene reduction method was
 found during summer stratification in 1969 and  1970.  Figure  16
illustrates  the relationship of fixation rate to depth during  the
periods of fixation;  it is noted that  fixation occurred only  in  the
hypolimnion during both years.  Rates  were highest during  the  summer
of 1969,  and the depth at which peak fixation occurred became  greater
 over the course of this summer.  This  tendency was not  found  during
 1970 when fixation started later than 1969 and  ended  sooner.   Why  a
                                 66

-------
                           Table 16.  Acetylene Reduction in Lakes Mary and Mize
Lake Mary- June 28
Depth l
meters
0 a
2 b
5 b
7 b
10 b
15 b
20 b
TKN2
mg/1
0.93
3.1
1.8
2.9
2.5
3.3
4.5
nM C2H4/
1-hr.
0.54
0.36
0.89
0.98
1.70
1.16
5.00
, 19£8
ng N/ 3
1-hr.
5.0
3.3
8.3
9.2
15.8
10.8
46.6
Lake Mize-July 11, 1968
Depth1
meters
0 a
1.5 a
3.0 a
6.1 a
6.1 a
9.1 b
18.3 b
TON2
mg/1
0.56
0.56
0.63
0.65
0.65
0.70
0.69
nM G2H4/
1-hr.
17.8
14.3
12.5
11.6
8.9
33.0
10.7
ng N/ 3
1-hr.
166
133
117
108
83
308
100
Lake Mize-A.ugust 21
Depth a
meters
0 a
2 a
5 a
1 b
10 b
15 b
20 b
TKN 2
mg/1
0.38
0.60
0.69
0.73
0.77
1.29
0.66
nM C2H4/
1-hr
0
0
0
2.98
3.28
8.93
2.68
, 1968
ng N/ 3
1-hr
0
0
0
27.8
30.6
83.3
25.0
1 Samples were purged with oxygen-argon mixture (a)  or helium (b)  before  incubation.


2TKN = total Kjeldahl nitrogen;  TON = total organic nitrogen,  both  in  mg N/l.

q
 Calculated from nanomoles of ethylene produced per liter-hour assuming  theoretical  ratio  of

 1.5 moles ethylene produced per mole of ammonia fixed.

-------
Table 17.  Chemical Characteristics of Lake Mize, Florida1
     pH                                            5.03
     Acidity (mg CaCO,/!)                         29.4
     Alkalinity (mg CaCOj/1)      _                1.7
     Specific conductance  (ymho cm  )             52.9
     Color  (mg Pt/1 at pH  8.3)                   434
     Turbidity (JTU)                               1.4
     Chloride (mg/1)                              10.3
     Sulfate (mg/1)                                5.0
     Na  (mg/1)                                     6.7
     K (mg/1)                                      0.32
     Mg  (mg/1)                                     0.94
     Ca  (mg/1)                                     3.4
     Fe  (mg/1)                                     2.4
     Mn  (yg/1)                                    19
     F (mg/1)                                      0.02
     SiO   (mg/1)                                   3.19
     COD  (mg/1)                                   69
     Total  solids  (mg/1)                          77
     Suspended solids  (mg/1)                       2
     1Average of Data  collected from June 1968 to June 1970
      (Shannon  1970).
                              68

-------
(X
Q
 3
 6

 9
L2
.15
18
                   FIGURE 14,  TEMPERATURE PROFILES  IN LAKE MIZE DURING  1969 AND 1970

-------
FIGURE 15. DISSOLVED OXYGEN PROFILES IN LAKE MIZE DURING 1369 AND 1970,

-------
Table 18.  Changes in Nutrient Levels in Lake Mize,  1968-19701
                    August, 1968   June,  1969    June,  1970
Parameter2
Total organic N
Ammonia-N
Ortho phosphate-P
Total phosphate-P
Top
0
0
0
0
.25
.20
.000
.005
Bottom
0
0
0
0
.63
.40
.063
.063
Top
0
0
0
0
.82
.03
.018
.087
Bottom
0
0
0
0
.66
.07
.030
.059
Top Bottoi
1
0
0
0
.35
.50
.058
.15
1.02
0.28
0.10
0.19
 Values represent means of top 3 meters and bottom 15 meters,
 or approximately the mean epilimnetic and hypolimnetic
 concentrations.

2Values in mg/1 as N or P.
                           71

-------
   0
   9   .
g-12
Q
  15
  18
                 1.0         2.0         3.0
              Nitrogen Fixation (yg N/l-hr)
6
>w'

4J
o.
o
12  .
15
                                                            18
                                                                                I
6-12-70

6-18-70 •
6-30-70 •

7-9-70  A
                  0.5             1.0
                 Nitrogen Fixation  (pg N/l-hr)
                1.5
                  FIGURE 16,  DEPTH PROFILES OF NITROGEN FIXATION  IN LAKE MIZE DURING  1969 AND 1979,

-------
a shorter  period of fixation occurred in 1970 is not completely clear;
however the  noted nutrient enrichment may have been a factor.  Maximum
nitrogen fixation rates ranged from 0.08 to 3.26 Ug. N/l.-hr.; the
upper value  is  within the range of rates reported for blue-green algae
in lakes (Dugdale and Dugdale, 1962; Goering and Neess,  1964;  Billaud,
1968) .

Nitrogen fixation in Lake Mize is apparently restricted  to mid-summer
stratification.  No fixation was detected in samples assayed in April,
September, October, and December 1969 as well as February, April and
May 1970.  The  annual cycle of nitrogen fixation for 1969 and  1970 is
illustrated  in  Figure 17, which shows total hourly nitrogen fixation
in the lake  over the year.  It is apparent that fixation was much
greater during  1969 both in duration and rate.  Values for total nitro-
gen fixation per hour were estimated by summing the product of the
volumetric fixation rate at each depth times the calculated lake volume
for each depth.  The annual input of nitrogen to Lake Mize by  fixation
can be extrapolated from the data in Figure 17.  Calculation of yearly
rates from a few one-hour incubations is undoubtedly risky, but even
an approximate  value would be instructive.  Assuming that during the
period of  fixation the reaction occurred 24 hours a day  at the rates
shown in Figure 16, the total nitrogen contribution in 1969 was 39.2 kg.
and in 1970  was 9.6 kg., or 1.14 and 0.28 g./m.3 of lake water per year,
for 1969 and 1970, respectively.  Shannon and Brezonik (197la) com-
puted a nitrogen budget for Lake Mize (excluding fixation) of  2.05
g./m.3-yr, (see Section V).  Thus fixation represents about 56 and 14
percent of the  total nitrogen income from other sources  in 1969 and
1970, respectively, and at least for this admittedly unusual lake, bac-
terial nitrogen fixation is a highly significant nutrient source.

Because of the  unusual nature of the above results an intensive effort
was undertaken  to determine the agents of fixation in Lake Mize.  Bac-
teriological samples were collected aseptically at the same depths
assayed for  nitrogen fixation on several occasions, and samples from
those depths where fixation occurred were subjected to enrichment culture
for groups of nitrogen-fixing organisms.  Three enrichment schemes
were utilized covering the spectrum of likely microbial agents:  1) an-
aerobic or facultative heterotrophic bacteria, 2) photosynthetic bacteria,
3) yeasts  and fungi.  Details of the enrichment and isolation procedures
are given in Aopendix C.

Enrichment techniques indicate that at least two nitrogen fixing groups
of bacteria, one heterotrophic, the other autotrophic, exist in the
depths of Lake  Mize.  Six isolates of gram positive spore forming rods
which grew anaerobically but not aerobically on nitrogen  free media
were isolated from water samples taken at 3, 5 and  7 meter depths on
June 18 and  at  5 and 7 meters on July 9.  A  total of seven transfers
of the anaerobic isolates were made, each time to nitrogen free media,
and the tests for acetylene reduction were positive at each step.  Cul-
ture of the  six isolates grown in liquid media for  24 hours gave acety-
lene reduction rates in the range 1.5 to 10 nmole C2H? reduced/mg
organism N-hr.
                                73

-------
a

00
•w

0)
4-1
tfl
o
X
•H
PM


0)
00
O
t-i
4-1
•H
SB

01

n)
cfl
4-J
O
      30
     20
     10
          Jan   Feb    Mar    Apr   May    June    July    Aug  Sept  Oct   Nov    Dec

                                              MONTH


                   FIGURE 17,  HOURLY  RATE OF NITROGEN FIXED IN ENTIRE VOLUME

                               OF LAKE MIZE DURING 1969 AND 1970,

-------
Colonies  of purple sulfur bacteria began to appear after 4 weeks
incubation in water from seven meters depth.  Growth began at the
glass cellulose interface, which hampered observation of morpholo-
gical types.   Transfers of pigmented colonies were made into liquid
media containing only Na7S and other inorganic salts at pH 8.5 and incu-
bated in  the light.  Phase contrast microscopy showed two morphologi-
cal forms, a long motile, spiral shaped organism which we identified
as a species of Thiospirilium and a small motile rod, apparently a
species of Chromatium.  The latter form predominated in mixed cultures
and grew  rapidly under the culture conditions.

No growth on nitrogen free media occurred within 7 weeks in any of
the 65 yeast and fungal isolates picked from the June 18 and July 9
samples.   However, growth was observed within 2 weeks for all isolates
incubated in the test medium plus nitrogen.

The agents responsible for fixation in Lake Mize are probably hetero-
trophic bacteria.  Fixation occurs only in a region of low or no light
and no dissolved oxygen.  The fact that fixation is maximum at inter-
mediate depths suggests photosynthetic bacteria may be the agents,
and they  are in fact present in the subsurface layers of this lake.
However,  the high rates of acetylene reduction found in Lake Mize sam-
ples incubated in the dark would tend to rule out photoautotrophic
forms as  dominant agents of fixation since photosynthesis apparently
provides  the source of energy for fixation by these forms.  Rhodospi-
rillum rubrum, for example, has been shown to fix nitrogen anaerobi-
cally in  the light, but intact cells of this bacterium cease fixing
immediately when placed in the dark (Pratt and Frenkel 1959).  Further,
photosynthetic bacteria were isolated from only one of the sampled
depths at which fixation occurred (7 meters).  The long lag before
significant growths of these organisms were noted in enrichment media
and the fact that microscopic examination of raw lake water and seston
retained  on Millipore filters failed to reveal the presence of photo-
synthetic bacteria imply very low populations in the lake.

On the other hand, heterotrophic growth on a nitrogen-free medium was
rapid for samples from the various depths at which fixation occurred.
The taxonomic characteristics of the isolates (i.e. gram-positive, spore-
forming,  obligate anaerobes, capable of utilizing N2 as their sole N
source are those of the genus Clostridium.  Whether this is the only
heterotrophic bacterium fixing nitrogen in Lake Mize cannot be answered
by the present study.  Other enrichment procedures may isolate differ-
ent forms; Arthrobacter has been suggested as another likely agent
(J. Sieburth, personal communication).  No aerobic or facultative bac-
teria and no fungi or yeasts were found which could grow on nitrogen
deficient media.  Cyanophyceans are not likely to be agents of fixa-
tion in Lake Mize since algal nitrogen fixation is related to photo-
synthesis and conditions at the depths of fixation indicate no oxygen
production.  No blue-green algae were found in samples examined micro-
scopically.  The observed stratification of fixation and its annual
periodicity may be influenced by such factors as a requirement for low
light and narrow tolerance to high or two sulfide concentrations  or
Eh values, which form gradients in the anoxic zone of this lake.
                                 75

-------
Sediments are  generally  considered to be relatively enriched in nutrients
and as such  are  unlikely environments for nitrogen fixation.  However,
studies supported by this project have detected fixation in both estua-
rine and  lacustrine sediments from subtropical and tropical environments.
In addition  the  report of Howard £t al. (1970) on nitrogen fixation in
Lake Erie sediments indicates this phenomenon is not unique to warm
climates.

Sediments were collected at various locations in the Waccasassa Estuary
and in various lakes with a Viomemade coring device.  Cores were returned
 to the laboratory and sectioned for profile analyses of nitrogen fixation.
 Since the sediments were anoxic, purging of N~ was accomplished with
helium (see Apendix A for procedured details).

 The optimum incubation  time for the estuarine sediments was determined  to
 be one hour.  The amount of ethylene production was linear for incubation
 up to that  time, but the rate decreased markedly after that (see Section IX)
 Rates of ethylene production in embayment sediments ranged from 0.00 to
 0.54 nM  C~Hx/g.  dry wt.-hr. for all the determinations (over 60) made
 in this  study.   A distinct layering was found within a typical sediment
 core (Figure  18).   In the flocculent,  unconsolidated 1 to 2 cm at the
 core surface, no Cft^ reduction was found at  any time.  Significant re-
 duction  rates were  consistently noted  in the  next  2 to 5 cm of consoli-
 dated, grey-black  ooze.  From 5 to  20  cm the  typical core consists of
 coarse organic  material, which overlies the  limerock substrate.  Acety-
 lene reduction  was  consistently found  in the  bottom zones of sediment
 cores, but  rates were low compared with those in the upper portions.

 An areal survey of  the  Waccasassa embayment  sediments showed acetylene
 reduction to  be a  consistent phenomenon in the  upper sediments.  Dupli-
 cate cores  were taken at eight stations in the  embayment, and the upper
 2 to 5 cm portion  of  each was blended  to obtain a  homogenous sample.
 Acetylene reduction was then run  on a  portion of the blended sediment.
 The distribution of ethylene production rates in the estuary sediments
 is shown in Figure  19.   A range  of  0.175 to  0.54 nM C^L/g- dry wt.-hr.
 was found,  but  most of  the values were near  the mean rate of 0.33
 nM C~H,/g.  dry  wt.-hr.   Expressed in  terms of the  equivalent amount of
 nitrogen that would have been  fixed (using the  theoretical  ratio of 1.5
 moles C2H,  produced per mole NH-  fixed)  the  results indicate a range of
 1.63 to  5.0 ng  N/g. dry wt.-hr.

 Several  experiments were performed  to  evaluate  the environmental condi-
 tions for fixation  in sediments  and to define the  agents of fixation.
 The effect  of a  nitrogen-free  oxygen atmosphere vs. an anoxic  (helium)
 environment was  determined in  one experiment on five sediment samples.
 In all but  one  case,  the rate  of  ethylene product  was slightly higher
 under a  helium  atmosphere, but the  differences  were not as  great as
 expected if strict  anaerobes were the  fixing agents.  Probably the or-
 ganic sediments  exerted a sufficient  oxygen  demand to maintain anoxic
 microzones  in the samples even in an  oxygen  atmosphere.  It was also
 found that  exposure of  incubating samples  to light decreased the acety-
 lene reduction  activity.  The mean rate  of ethylene production in five
 replicates  incubated in the  light was  only one  third of the mean value
                                    76

-------
     FLOC

                                          ETHYLENE PRODUCTION
                                          (ng C2H^/g dry wt.-hr)
                                                  1.62 - 11.4
COARSE
ORGAN ICS  5~20 cm
0.03 - 0.40
        FIGURE 13,  SECTION OF CORE  ILLUSTRATING
        VARIATION OF  ACETYLENE REDUCTION WITH
          DEPTH  IT! VACCASASSA ESTUARY  SEDIMENTS,
          RANGE  OF ETHYLENE PRODUCED REPRESENTS
        AT LEAST SIX  DETERMINATIONS  IN EACH LAYER,


-------
          FIGURE 19, DISTRIBUTION OF ETHYLENE PRODUCTION RATES
     WITHIN WACCASASSA ESTUARY SEDIMENTS (in nB G2H4/§ dr
DUPLICATE CORES AT EACH OF EIGHT STATIONS,   STATION 11 WAS IN MUD FLAT,
                                 78

-------
for replicates of the same sediment incubated in the dark.   The reasons
for this cannot be fully stated; however, it is well known that many
non-photosynthetic bacteria have pigments and that these organisms can
be inhibited by visible radiation.

Since the agents of ethylene production (hence nitrogen fixation)  in
the sediments are presumably anaerobic or facultative bacteria, addi-
tion of easily assimilable carbonaceous substrate to sediment samples
should enhance their activity.  Several experiments have given somewhat
conflicting results in this regard.  In an early experiment acetate
gave apparent stimulation of acetylene reduction at concentrations of
0.02 and 0.2 M but glucose gave no response.  However no acetate stimu-
lation was found in three later experiments in the range 10"1* to 1C"1 M,
and in fact 10"1 M acetate and butyrate actually inhibited fixation by
about 30 percent.  No evidence for glucose stimulation or inhibition
was found in any of the four enrichment experiments in the concentration
range lQ~k - 10~1 M, but sucrose gave a definite stimulation (50-100
percent increases over controls) in each of the three experiments  in
which it was added (Table 19) .  These results are similar to responses
of lacustrine sediments to organic additions (described later in this
section).

Several experiments were performed to evaluate the impact of inhibitors
on acetylene reduction.  In 30-40 ml. sediment slurries fixation was
inhibited by 1 ml. of fifty percent trichloroacetic acid or 3 ml.  of
saturated mercuric chloride solution.  If acetylene reduction is re-
lated to nitrogen fixing activity, it would be expected that molecular
nitrogen would inhibit the rate of acetylene reduction.  Since acetylene
is a competitive inhibitor of N~  (Schb'llhorn and Burris, 1967), the con-
verse should be true, i.e. added molecular nitrogen should reduct the
rate of acetylene reduction according to the competitive enzyme inhi-
bition pattern (see Cleland (1963) for a complete discussion of enzyme
kinetics).  Thus a reciprocal (Lineweaver-Burk) plot of reaction velo-
city (ethylene production rate) versus substrate  (acetylene) concentra-
tion should be linear, and added  inhibitor  (molecular nitrogen) should
give the classical competitive inhibition plot (see Section IX).  To
verify this for Waccasassa sediments, two sets of samples were set up;
one series was exposed to various concentrations of acetylene in a
helium (nitrogen-free atmosphere, and in another series using the same
sediment, samples were exposed to various acetylene concentrations in
a 70% N2 - 30% C02 atmosphere.  A Lineweaver-Burk (reciprocal) plot of
the ethlene production rate vs. acetylene concentration (Figure 20)
shows the classical competitive inhibition pattern.  The maximum velo-
city (V   ) as obtained from the y-intercept of the Lineweaver-Burk plot
is approximately the same for both curves implying that high concentra-
tions of acetylene negate the inhibitory effect of N2-  Competitive
inhibition patterns are given by  enzyme inhibitors that react at the
same active site on the enzyme as the substrate.

The organisms most likely to be responsible for the above phenomena
would seem to be .anaerobic bacteria such as Clostridium sp.  An
enrichment and isolation procedure was conducted (see Appendix C for
details on methodology) to determine whether such organisms could  be
                                  79

-------
    Table 19.   Effect of Organic Substrate Addition
     on Ethylene Production by Estuarine Sediment
                        Relative Response1
                   Concentration Added (Molar)
Substrate
Acetate
Butyrate
Glucose
Sucrose
Hf1
-
-
0
+
1CT2
0
0
0
+
1C' 3
0
0
0
0
i
-------
00
               0.12
               0.10
           M
           rC

           '
           •n
           oo
u
W)
c
               0.08
               0.06
            >  0.04
               0.02
                                                           I
                                      0.2
       0.4

I/Substrate (cc
                                                                                      atm    O
                                                                                   He atn
                                                                  0.6
0.8
                                                                        -1
                       FIGURE 20,  LINECAVER-BURK PLOT OF ETHYLENE PRODUCTION VS ACETYLENE CONCENTRATION
                            INDICATES COMPETITIVE INHIBITION BY N2 IN WACCASASSA ESTUARY SEDIMENT,

-------
detected in Waccasassa  sediments.   Three  isolates  were  obtained x^hich
could grow in  a  nitrogen-free medium and  reduce  acetylene  to  ethylene
(hence presumably  fix molecular nitrogen).   The  characteristics of  the
final isolates vere highly indicative of  a  Clostridium  (or clostridium-
like) culture; i.e. the isolates were gram-positive,  strict anaerobic,
large rods, which  ferment sugars (and fix molecular nitrogen).  Further
physiological  tests would of course be necessary to establish specific
taxonomy  of  the cultures, especially down to the species  level.   How-
ever the  taxonomy  of marine and estuarine anaerobes is  not highly
developed,  and to  place the cultures in a particular genus would  be
questionable  at this time.  Nonetheless,  this work has  demonstrated
that "clostridia-like"  nitrogen-fixing bacteria are present in Wacca-
sassa Estuary  sediments.  The results do  not exclude the  possibility
of other  nitrogen-fixing forms being present in the sediment  since  the
isolation method used was somewhat selective for clostridia-like bac-
teria.

Sufficient evidence has been obtained to  conclude x^ith  reasonable
assurance that the acetylene reduction activity is directly related
 to nitrogen fixing organisms in the sediments.  All lines  of  evidence
point to  ethylene production as a biological phenomenon; both trich-
 loioacetic acid and mercuric chloride completely inhibit  acetylene
 reduction.  Organic substrate (in some cases) affects the  rate of ethy-
 lene production.  The fact that N» acts as a competitive  inhibitor  of
 acetylene reduction strongly suggests that ethylene is  produced by
nitrogenase.   Nitrogen-free media produced growths of gram positive
 spore-forming rods from sediments under an atmosphere of  pure nitrogen,
 and a pure culture similar to Clostridium sp. was isolated from the
 sediment  and was shown capable of  nitrogen fixation by acetylene re-
 duction

 A rough estimate of the total amount of nitrogen fixed  in the 2-5 cm
 layer of sediments in the Waccasassa can be obtained from the data  pre-
 sented in Figure 19.  Assuming that the mean value of 3.1 ng  N fixed/g
 sediment-hr.   is a  reasonable estimate of the nitrogen fixation rate in
 this stratum  of sediment  throughout the estuary, the amount of nitrogen
 fixed on an annual basis  is 37 yg N cm2-yr. or for the  entire 7 km2
 estuary,  2.6  x 103 kg N/yr. These values are based on an average  sedi-
 ment dry weight of 0.455  g/cm3 in the 2 to 5 cm zone.  These  extrapo-
 lated values  are obviously rather crude,  but they serve to illustrate
 the point that the seemingly low rates found in the sediments in  fact
 represent significant amounts of nitrogen fixed on an annual  basis.

 Because sediments  are normally thought to be enriched in nutrients, we
were somewhat surprised at our initial results indicating nitrogen  fixa-
 tion in Waccasassa estuary sediments.  Free ammonia was determined  on
 sediments from 14 widely scattered stations in the estuary (Brooks,
 1969); a range of 0.01 to 0.37 and mean of 0.06 mg NH^-N/g sediment was
 found.   Not all this ammonia would necessarily be available to micro-
 organisms; much or perhaps all of it could be loosely sorbed  to solid
material  in the sediment.  Some evidence was found to support this  hypo-
 thesis (see Section VIII); Waccasassa Estuary sediment  was shown  to
 contain 3 to 5 percent clay mineral and to sorb ammonia rapidly from
solution.
                                  82

-------
The first indication that lacustrine sediments supports nitrogen fixing
activities was uncovered in  1968 when a short core from Bivin's Arm
yielded positive rates which decreased rapidly with depth  (Table 20).
Sediment cores were obtained from 16 lakes in north central Florida in
September, 1969, and from these and 9 additional lakes in June and July
of 1970 to determine whether lacustrine sediment fixation is a more
widespread phenomenon.  These lakes were all sampled periodically in
conjunction with other studies  (Shannon and Brezonik, 1971c; Brezonik,
1971c), and their water and  sediment characteristics are thus known.
A list of the 25 lakes and their trophic states is presented in Table  21
along with pertinent sediment characteristics.  Three general types
of sediment occur in the lakes:  A) brown flocculent or unconsolidated
material often with undecomposed plant remains, B) black, jelly-like
sediment, and C) sandy sediment.

Short cores (30-50 cm) from  the lakes either blended in their en-
tirety and an aliquot taken  for measurement of nitrogen fixation or
segmented into several zones which were then tested for fixation indi-
vidually.  Another aliquot of sediment was dried and weighed, and results
were reported as rate of fixation per gram dry weight of sediment.  Seven
of the twenty-five lake sediments showed apparent nitrogen fixation in
one or both surveys (Table 22) .  Four of the seven sediments were peat-
like (Type A), and except for Lake Apopka,  these also gave the highest
rates of ethylen production.  Peat sediments might be expected to have
a lower available nitrogen content than muck sediments.  Sediment from
lake #20 appeared tobbe a mixture of sand and sludge.  This shallow lake
(maximum depth is 3.5 m) receives some domestic waste effluent and is
at times anoxic below 1 meter.  Sediments from Lakes Apopka, Bivin's Arm,
Kanapaha and Orange emitted  a musty or moldy odor when blended, while
the other sediments were either odorless or gave off a hydrogen sulfide
odor.  Lake Alice was the only sediment with a pronounced H~S odor that
showed significant fixation.  This lake receives a large proportion of
its inflow as treated sewage effluent.  Five cores taken from this lake
in 1969 showed fixation, but no fixation was found in a core taken during
the 1970 survey.

In order to determine the layers of sediment most active in nitrogen fixa-
tion, profiles of acetylene  reduction rates were measured on sediment
cores from Orange Lake in the 1969 survey and on all lakes sampled in the
1970 survey.  The loose, unconsolidated nature of most of the sediments
exhibiting nitrogen fixation precluded detailed segmenting of the cores,
and only the core from unnamed lake 20 was easily divisible into segments.
Table 23 presents the 1969 profiles for Orange Lake and profiles from
the 4 lakes exhibiting fixation in the 1970 survey.  Highest rates were
found in the upper layers; this trend was most marked when the rates
are considered per gram dry  weight of sediment rather than per milligram
of sediment nitrogen.  Sediment from lake #20, while not as active as
the other three sediments when compared on a sediment dry weight basis,
showed comparable activity on per sediment nitrogen basis.

Sediment samples from lakes  in the Peten region of Guatemala surveyed
in conjunction with another  project during the summers of 1969 and 1970
also showed nitrogen fixation by the acetylene reduction method.  Loca-
tions and descriptions of these  lakes  are given by Fox ejt.  al.  (1970).
                                    83

-------
             Table 20.  Acetylene Reduction in
       Bivin's Arm Sediment Collected in August, 1968
                                                Ethylene Production
  Sample                     Treatment1        nM C^/g. dry wt-hr.

Ekman dredge -
 Surface to                Control (TCA) He2           0.043
 3" mixture                              He            2.09
                                      OojAr            0.54
                           500 mg NH3~N/1 added, He    0.27
Core
Surface
Surface
3" deep
5" deep
7" deep
9" deep

0 Ar
He
He
He
He
He

0.31
2.15
0.25
0.21
0.10
0.09
1Control treated with  1 ml 50%  trichloroacetic acid before adding
 acetylene.  Samples purged either with helium (He) or with gas
 mixture (20% 02, 0.03% CO-, balance Ar) to remove dissolved N .

2Equal to "blank" reading, i.e. background ethylene present in
 acetylene.

-------
     Table 21.  Lakes Surveyed for Sediment Nitrogen Fixation1
Hypereutrophic                 Eutrophic                 Mesotrophic

Apopka (B)                     Hawthorne  (A)             Cooter Pond (B)
Unnamed 20 (D)                 Clear  (B)                 Lochloosa (B)
Bivin's Arm (A)                Wauberg  (A)               Calf Pond (A)
Griffin (B)                    Newnan's  (A)              Orange (A)
Alice (A)
Eustis (B)
Kanapaha  (A)


Oligotrophic                   Ultraoligotrophic

Watermelon Pond  (B)            Anderson-Cue  (A)
Unnamed 10  (A)                 Gallilee  (B)
Jeggord (3)                    Cowpen  (B)
Moss Lee  (B)                   Sand  Hill (C)
Altho (B)                      Swan  (C)


Classification  into trophic class  after Shannon and  Brezonik (1971b).

2Sediment  types  found:

 A-Brown  flocculant  unconsolidated  material with or without
    undecomposed  plant remains.
 B-Black  Jelly like  sediment.
 C-Sandv  bottom.
 D-This sediment was a mixture of sand and brown, foul-smelling sludge.
                             85

-------
      Table 22.  Nitrogen Fixation Rates in Florida Lake Sediments
Lake
Sediment
Characteristics
     Nitrogen Fixation Rate1
  1969             19702
ng.N/g-hr. ng.N/g-hr.  Ug.N/g N-hr.
Bivin's Arm
Kanapaha
Orange
Moss Lee
Apopka3
Alice
Unnamed 20
A
A
A
B
B
A
D
28
36
9.5
1.8
17
1.2"
0
22
59
28
0
-
0
16
2.0
1.4
1.3
0
-
0
23
 Calculated  assumed  a C^H^/NH  molar production ratio of 3/2.

 2Maximum rate  found  in profile  (Table  23).

 30ne dredge  sample Nov.  3,  1969.

 "* Average of  5  cores.
                                86

-------
Table 23.  Stratification of Nitrogen
 Fixation in Florida Lake Sediments
Stratum
(cm)
Orange Lake •
1969
°~37 Shallow Water Core
37-50
0-36 Deep Water Core
36-50
1970
0-33
33-42
Lake Kan ap ah a
1970
0-5
5-15
15-25
25-38
Lake Bivin's Arm
1970
0-2
2-5
5-10
10-15
20-25
Lake Unnamed 20
1970
0-2
2-4
4-6
6-8
8-10
10-12
12-15

15-20
20-25
Nitrogen Fixation
ng N/g - hr. yg N/;


1.09
0.33
0.98
0.29

28
1.1


59
10
3
0


14
22
18
5
0


3
12
16
3
4
6
0


0
Rate
g N-hr


	
	
	
	

1.3
0.78


1.4
1.4
0.37
0


1.4
2.0
1.9
0.8
0


0.51
2.1
2.3
0.71
0.25
0.24
0
o

0
                 87

-------
In 1969, Laguna Petenxil  sediment  fixed  at  a  rate  of  3.1  yg. N/g.
sediment N-hr., while  green  flocculant surface sediment from Laguna
Echixil fixed at  a  rate of 5 yg. N/g. sediment-hr.  Sediments  from
two Peten lakes  (Eckixil  and Sal Peten)  were  sampled  in detailed pro-
files in 1970, and  high rates of fixation were found  in both lakes (up
to 2.4 yg. N/g.  sed. N-hr.).   Sal  Peten  sediment exhibited  three distinct
bands of fixation.  The near surface sediments of  these lakes  are un-
usual in being composed of flocculent, pigniented organic  particles.  The
green to pink color of the sediment apparently derives from undecomposed
algae and perhaps partially  from photosynthetic bacteria.

Because the  acetylene  reduction technique is  an indirect  assay for nitro-
gen fixation and because  fixation  had never before been measured in a
sedimentary  environment,  it  was felt essential to  demonstrate  that the
measured ethylene production indeed reflected nitrogen fixing  activities.
This was accomplished  by  a variety of experiments.  First,  biological
poisons were shown  to  inhibit acetylene  reduction.  Addition of 1 ml of
50% trichloroacetic acid  or  3 ml of saturated mercuric chloride to sedi-
ment samples (25 ml slurries)  immediately and completely  stopped ethy-
lene production.  Second, if acetylene reduction is related to nitrogen
fixation, addition  of  N2  should inhibit  the rate of acetylene  reduction
as was  shown for Waccasassa  Estuary sediments and  for nitrogen fixing
algae  (see Section  IX).   A similar experiment was  performed with sediment
from Lake Kanapaha, a  shallow, highly eutrophic lake  near Gainesville,
Florida.  Varying amounts of  acetylene were added  to  serum  bottles con-
taining 25 ml of sediment slurry which had  either  been purged  with an
02~Ar mixture to eliminate N   or left unpurged, and the samples were
incubated for 1.5 hours at 22°C in the dark.  A Lineweaver-Burk (recipro-
cal) plot of the resulting rates vs. substrate'(acetylene)  concentrations
(Figure 21)  fit  a pattern of  competitive inhibition of ethylene produc-
tion by N£.  This corroborates the experiment on Waccasassa Estuary sedi-
ment discussed previously and is in agreement with Schollhorn  and Burris'
(1967)  finding that acetylene is a competitive inhibitor  of nitrogen fix-
ation.

Heterotrophic nitrogen fixation is known to depend on the availability
of organic carbon.  Table 24  shows the results of  an  experiment designed
to show the  effect  of  various  organic substrates on acetylene  reduction
in lake sediment.   Twenty-five ml.  samples  of sediment slurry were purged
and injected with acetylene in the usual manner except that prior to
injection of acetylene sufficient  carbon source was added to give the
concentrations shown.  Sucrose definitely stimulated  acetylene reduction
while the 0.10 M  concentrations of sodium acetate  and butyrate caused
inhibition.  The  above experiments strongly  support a  biological media-
tion of ethylene  production from acetylene  which is directly related to
nitrogen fixing  activities.  Fixation in both estuarine and lacustrine
sediments is most likely mediated by heterotrophic bacteria, and sucrose
additions stimulated fixation in both sediments,    Photosynthetic bacteria
are obviously unlikely agents for  the observed activities since fixation
extends into fairly deep layers of sediment and blue-green  algae are
rare or absent in the  Waccasassa Estuary and  in some  of the lakes whose
sediments show fixation.  Blue-green algae  cannot be  completely ruled
out in all cases, for  Hoare at al.  (1971) have demonstrated that Nostoc
                                  88

-------
   5.0
J-i
,e
   4.0
£  3.0

G
o
o
3
TD
O
S  2.0
w
   1.0
                        O
                                                   0.68 atm N2   •


                                                   0.00 atm N0   O
                 0.2
0.4
0.6
0.8
                        1/S  =  I/(Acetylene Added)   (cc C2H2)
                                                            -1
i.o
        FIGURE 21,  COMPETITIVE  INHIBITION OF ETHYLEME PRODUCTION

         (BY NITROGENASF.) CAUSE BY !io  IN  U\KE KANAPAHA SEDIMENT,
                              89

-------
Table 24.   Effect of Organic Substrates on Rates  of
          Acetylene Reduction by Lake Kanapaha  Sediment
              Relative Rate  of Acetylene Reduction1
         Substrate
10
10
10 3M
                                                 10~"M
Glucose
Sucrose
Pyruvate
Acetate
Butyrate
1.2
1.9
1.1
0.8
0.8
1.1
1.7
1.0
1.1
1.0
1.1
1.2
1.0
1.2
1.1
1.0
1.0
1.0
1.0
1.1
         Relative to acetylene reduction rate of 0.50 nmoles
          C2H2 reduced/mg N-hr. (average of 4 controls) .  Each
          value is the mean of duplicate samples.
                             90

-------
and other species can metabolize heterotrophically even under anoxic
conditions.  However, bacteria are at least as likely an explanation
for fixation under these circumstances.

While nitrogen fixation does not occur universally in lake sediments,
its occurrence at least sometimes in 7 of the 25 Florida lake sediments
suggests that the phenomenon is more than just an environmental curiosity.
The occurrence of fixation in the three Guatemala sediments tested in
Lake Erie and Waccasassa Estuary sediments indicates a fairly widespread
distribution of low nitrogen fixation rates in the sediments underlying
natural waters.  The range of fixation rates measured in Florida sedi-
ments was 0.33 - 59 ng. N/g. sediment-hr. in surface layers and 0.02 -
1.1 ng. N/g. sediment-hr. in the bottom strata of 30-50 cm. cores.
Comparable rates were found in Guatemalan lake sediments and in sediments
from the Waccasassa Estuary.  The rates for Lake Erie sediments (Howard
et al., 1970) are also in the same range.

It would be of great interest to determine why fixation occurs in some
sediments but not in others.  Examination of sediment characteristics
of the Florida lakes shows no obvious correlation between the occurrence
of sediment fixation and sediment ammonia, total organic nitrogen, phos-
phorus, or percent volatile solids, carbon or nitrogen.  Fixation acti-
vities should be inversely related to ammonia concentration.  On the
other hand, even though sediments may be apparently  nitrogen-rich, much
of the ammonia may be sorbed onto clays and other particles and may
unavailable to microorganisms (see Section VIII).

Increased  activity in the upper layers of sediment probably reflects
higher concentrations of oxidizable substrate in these  layers.  Depth
profiles of  fixation were not nearly so narrow in the lake sediment cores
as those found in Waccasassa Estuary sediments, where fixation was con-
fined  largely  to the 2-5 cm. stratum.  The broader depth distribution is
undoubtedly  related  to wind-induced mixing of relatively unconsolidated
sediments  in the shallow lakes.  This hypothesis is  supported by  the
results for lake #20, which had a sharper stratification and which also
has a  fairly compact sediment.  Stratification of fixation is much more
pronounced when  the  rates are expressed per gram dry weight sediment  than
when expressed per unit  of  organic nitrogen in the sediment.

Nitrogen fixed in the sediments will be only  partially  released to the
overlying  water, depending  on the degree  of mixing effected by wind
action.  In  the  loose unconsolidated sediments such  as  found in Lakes
Kanapaha,  Orange and Bivin's Arm, this may be considerable;  in compacted
sediments  release to the overlying water  may  be  controlled by much slower
diffusion  processes.  Extrapolation  of the sediment  fixation rates in
Tables 22  and  23 to  annual  amounts indicates  that the  process may contri-
bute substantial quantities of nitrogen  to the  lake  basin  as a whole  and
may in this  sense be geochemically significant.
                                  91

-------
                            SECTION VIII

             OTHER IN SITU NITROGEN SOURCES AND SINKS
Although sediments are generally described as nutrient traps,  this  is
only true as a net reaction.  Transfer of nutrients between water and
sediment is at least partially reversible and to an extent cyclic rather
than one-way.  Over geological time sediments build-up acts as a net
sink, but on a seasonal or shorter time basis the possibility  of sedi-
ments acting as nutrient sources must also be considered.   Nitrogen is
lost to sediments by deposition of particulate matter (detritus, silt)
and by sorption of ammonia onto clays in or being deposited onto the
sediments.  Nitrogen can be released from sediments by activities of
burrowing animals, decomposition of organic nitrogen to ammonia and
its diffusion into the overlying water, and desorption of  ammonia from
clays and other sorbents in the sediment.

Nitrogen exchange at the sediment-water interface has received scant
attention compared to phosphate exchange.  Several classic studies
have discussed the role of oxygen and redox potential in controlling
diffusion of ammonia from sediments.  Mortimer (1941, 1942) reported
that little ammonia is released from sediments as long as  an oxidized
microzone exists but that large quantities of this (and other  nutrients)
are liberated when this surface layer is reduced.  He felt that this
general increase could be explained by sorption of ammonia onto a com-
plex organic floe containing iron (III).  The apparently low rate of
nitrogen release from oxidized sediments could also be partially ex-
plained by nitrification at the sediment surface causing nitrate rather
than ammonia release, and by the fact that ammonia is unlikely to
accumulate in oxygenated water as it will either be oxidized or assi-
milated.  The role of the oxidized microzone in influencing sediment-
water nutrient cycling would be a fruitful area for future research.

For purposes of nutrient exchange, lake sediments can be divided into
at least two main types based on the presence or absence of a defined
water-sediment interface.   In the former case nutrient exchange may
be limited by rates of diffusion, the presence of an  oxidized surface
layer, and the activities of burrowing animals  (Gahler, 1969).  Many
shallow, eutrophic Florida  lakes have no defined sediment-water inter-
face.  A flocculant suspension covers the bottom in these  lakes with
gradual compaction from thin "soup"  to consolidated sediment occurring
over a depth of perhaps several feet.  In this case considerable ex-
change may be effected by wind generated currents and turbulence,
causing the sediment suspension to mix with  overlying water.  Rates
of anaerobic decomposition  of organic sediment are probably a primary
factor limiting nutrient exchange in such sediments.
                                93

-------
Thus,sediment-water  nitrogen interactions  involve both biological and
chemical processes,  and rates of exchange  may be further controlled by
physical  (e.g.  diffusion)  phenomenon.   Qualitative aspects of biologi-
cal  decomposition in sediments are fairly  well understood (e.g. Foree
et_ al. ,  1970;  Lee, 1970b;  Rittenberg et aJL. ,  1955) and need no further
elaboration here.  The importance of burrowing animals (e.g. worms,
larvae,  crustaceans) in aerating sediments and in physically trans-
porting nutrients from the sediment and excreting them into the over-
lying water was suggested (Brooks, 1969) in work supported by this
project.  When surface sediment from the Waccasassa Estuary (Florida)
was  covered with estuary water in laboratory aquaria, a defined sedi-
 ment-water interface and oxidized surface layer quickly arose, and with-
 in several days extensive burrowing and tube building by amphipods and
 annelids were evident  through the aquarium glass wall.  However quan-
 titative data on regeneration rates induced by  this mechanism are
 lacking.

 Ammonia may be sorbed  onto  clays  and organic colloids in sediments, in
 which case  an equilibrium between aqueous ammonia and ammonia in sedi-
 ments would be set  up.  Principles  of  soil chemistry and cation adsorp-
 tion processes in soils should be applicable  to the study of this phe-
 nomenon in  sediments  (Toetz,  1970;  Toth and Ott,  1970).  Carritt and
 Goodgal  (1954) listed  clays,  gels of ferric hydroxide and silicic acid,
 humus colloids,  polymorphic inorganic  and organic complexes and sur-
 faces of  living  and dead  organisms  as  possible  adsorbents active in
 sediments.  If sorption plays an important role in ammonia equilibrium
 between sediment and water, it  should  be  possible to describe the
 equilibriun between ammonia adsorbed and  ammonia concentration in the
 interstitial  water  by  one of the well-known  sorption isotherms such as
 the Langmuir  or  Freundlich.  Since  other  cations compete with ammonia
 for sorbent binding sites,  the  extent  of  ammonia adsorption should be
 affected by changes in concentrations  of  other  ions in  the bulk solu-
 tion.   On the other hand  a considerable amount  of "trapping" may occur
 such  that the sorbed ions are not freely  in  contact with bulk or insti-
 tial  water but have migrated by intraparticle diffusion.  Such a situa-
 tion would render the sorbed ammonia relatively immune  to changes in
 composition of the  bulk solution, in effect  making aamonia  sorption
 irreversible.

 A problem in  defining the role  of sorption  in sediment-nutrient inter-
 actions is to separate the effects  of  biological activities.  Use of
 poisoned or irradiated sediments is possible,  but there is  always a
 danger in changing the chemical nature of sediment by  such  treatment.
 The importance of chemical processes can  be  inferred  if rapid uptake
 or  release occurs during  short  (several hour)  incubations since bio-
 logical decomposition and assimilation are  inherently slow  processes.
 A further problem also arises in relating the results  of  laboratory
 shaker-flask experiments  on sorption and  leaching to  in situ  conditions.

 The high concentrations of  ammonia which typically arise in anoxic lake
 hypolimnia during summer  stratification have long been known  (Domo-
 galla et^ cd. , 1926; Hutchinson, 1957). Release of  ammonia  from  sedi-
 ments has been a widely accepted mechanism for these  increases, but
 the decomposition of sinking organisms and  detritus  offers  an alternate
                                   94

-------
explanation for at least some of the increase.  Attempts to determine
the relative importance of these two mechanisms have not yet been
successful.  Rates of detrital deposition in sediment traps and changes
in the elemental composition of detritus as it sinks have been reported
(Kleerekoper, 1953), and sediment leaching experiments involving labo-
ratory incubations have been described by Mortimer (1941, 1942), Sawyer,
et al. (1945) and others.  But the quantitative significance of sediments
HLS 1m ammonia source remains elusive.  Recently Fitzgerald (1970) des-
cribed experiments in which algae grown in contact with sediment from
Lake Mendota, Wisconsin, maintained  a phosphorus deficiency in spite
of the high phosphate level of the sediments.  The results qualitatively
suggest that phosphate  incorporation into sediments is largely irre-
versible.  Whether this  also applies to nitrogen in sediments and if
so, under what conditions, remains unknown.

In summary  the processes of nitrogen cycling  and interchange in sedi-
ments are'complicated and  poorly understood.   The mechanisms whereby
nitrogen  is  exchanged between water  and sediments are probably known,
and in some  cases  qualitative  rankings  can be given  to  their importance.
However almost no  information  is available to establish  the in_ situ
rates and controlling  factors  for  these processes.

The chemistry  of  nitrogen  in  Florida sediments has been  studied  in two
phases.   The chemical  characteristics  of  sediments  from the 55  lakes
described in previous  sections have  been  evaluated  to provide  a  basis
for later studies on the dynamics  of nutrient exchange.   In conjunction
with  a  separately funded project  the sediments were  analyzed  for total
organic  nitrogen, free ammonia, ortho  and total phosphate,  iron, man-
ganese,  volatile solids, C,  H, and N contents by elemental analysis,
 chlorophyll and carotenoid degradation products, benthos, and gross
physical description.   Of primary concern here are  the  first  two analy-
 ses,  but the other parameters are important  in describing the sediments
 and'may later be useful in relating exchange rates  and  capacities to
 sediment type.

 Results for surface sediments collected by Ekman dredge from the lakes
 are presented in Table  25.  For comparative purposes the type of sedi-
 ment and organic content are listed along with nitrogen content.  Methods
 used in the analyses are described  in Appendix A.  Highest concentra-
 tions of sediment ammonia were generally associated with enriched or
 polluted lakes (e.g. Hawthorne, Dora, Griffin) or with undisturbed lakes
 having a high sediment  organic content.  Ammonia concentrations ranged
 from less than 10 ppm  to  over 1000  ppm on a  dry weight basis.  Total
 nitrogen values ranged  from 0 to 4.0 percent on a dry weight basis, but
 low levels  (less  than  one percent)  were  generally associated with low
 carbon and  low volatile solid  samples.   For  example, no  nitrogen was
 detected  in the  Swan Lake  (Putnam County) sample, but  this sediment
 was composed  largely of sand  and had only 1.2 percent  carbon and 2.3
 percent  volatile  solids.  This lake is also  very clear  and can  be con-
 sidered  ultra-oligotrophic.   Highest values  of total nitrogen were
 associated  with  high volatile  solids  and carbon and usually with eutro-
 phic or  polluted conditions.   Lake  Hawthorne had the highest  sediment
 nitrogen;  this  eutrophic  lake receives sewage effluent from  the town  of
                                  95

-------
Table 25.  Sediment Characteristics of North Central Florida Lakes
Lake1
Sante Fe
Little Santa Fe
Hickory Pond
Altho
Cooter
Elizabeth
Clearwater
Hawthorne
Little Orange
Unnamed (#10)
Moss Lee
Jeggord
Still Pond
Lochloosa
Orange
Palatka
Newnan's
Calf
Unnamed (#20)
Meta
Alice
Bivin's Arm
Clear
Unnamed (#25)
Beville's Pond
Unnamed (#27)
Kanapaha
Watermelon
Long Pond
Burnt
Wauburg
Tuscawilla
Apopka
Dora
Harris
Eustis
Griffin
Weir
Kingsley
Sumter Lovry
Magnolia
Brooklyn
Geneva
Sediment Volatile
Type2 Solids%
DBr,Bl-N-Si
DBr,Bl-N-Si
DBr-O-S, Si
DBr-0-S,Si
DBr-O-S, Si, Or
DBr-T-S,Or
Br-0-S,Or
LBr-N-S,Or
DBr-T-S,Or
LBr-N-S,Or
DBr-T-S
DBr-T-S
LBr-0-S,D
Br-O-S, Or
Br-0-S,Or,P
Br-N-D
Br-O-P
Br-N-P,D
G-O-S
G-0-S,D
Br-O-P, D
Br-O-P, D
Br-O-S
Br-O-S ,D
Br-O-P ,D
Br-O-P, D
Br-O-P, D
B1-N-M,S
B1-0-M,D
DBr-0-S,M,D
Bl-0-S,Or
G, Br-O-S, D
DBr-O-S ,M
DBr-O-S ,M
DBr-O-S ,M
DBr-O-S ,M
DBr-O-S ,M
G, Br-O-S, Si
Bl, Br-O-S, Si
Bl-0-S,Si,D
Bl-0-S,Si
Bl-0-S,Si
Bl-O-S
18.7
58.8
6.8
61.5
51.5
13.2
8.1
82.4
15.0
88.3
87.3
62.2
24.7
15.3
47.9
95.1
52.8
84.3
25.7
29.6
43.0
65.7
4.9
2.1
84.4
41.7
76.5
64.6
49.6
34.5
39.2
52.7
52.2
59.7
32.5
56.4
63.4
54.7
34.4
24.4
32.2
7.4
16.6
Ammonia-N
Pg/g
30
30
10
40
70
10
10
1010
80
180
180
20
140
10
50
520
320
10
310
30
40
570
10
10
270
50
240
60
60
60
460
80
250
480
250
150
820
150
30
10
30
10
< 10
Carbon3 Nitrogen3
% %
11.5
45.2
4.6
36.6
26.6
8.1
6.1
44.9
8.2
49.1
49.6
42.3
1.6
8.7
26.0
48.2
28.1
50.1
12.8
18.3
14.8
35.5
2.3
1.2
46.2
23.6
40.3
43.4
27.3
18.6
23.6
27.3
29.7
27.4
16.9
26.9
33.7
31.8
18.3
14.2
14.9
3.3
8.9
0.7
2.4
0.3
1.8
2.1
0.4
0.4
4.0
0.6
3.5
3.5
1.4
0.1
0.9
2.4
3.1
2.4
3.5
0.9
1.3
1.2
3.2
0.2
0.1
3.9
1.9
3.6
2.9
2.1
1.3
1.7
2.1
2.4
2.2
1.6
2.3
2.7
2.4
1.1
0.7
0.7
0.2
0.6
C/N
wt . /wt .
16.6
18.7
18.4
20.6
12.7
19.6
15.3
11.1
14.8
14.1
14.3
29.6
12.4
9.8
11.0
15.5
11.9
14.2
14.4
14.4
12.7
11.2
11.7
13.7
11.9
12.7
11.2
14.8
12.8
14.2
13.6
13.1
12.5
12.5
10.6
11.7
12.3
13.3
17.4
20.0
22.6
15.0
15.8
                                   96

-------
Lake
                          Table 25.  (Contd.)
Sediment   Volatile Ammonia-N3   Carbon3  Nitrogen3  C/N
 Type2      Solids%  yg/g          %         %      wt./wt,
Swan
Wall
Santa Rosa
Adaho
McCloud "
Anderson-Cue
Suggs
Long
Cowpen
Gallilee
Bl-O-S
Br-0-S,P
LBr-N-S
DBr-0-S,Si
DBr-0-S,Si
DBr-0-S,Si,D
DBr-0-S,Si
DBr-N-S
Bl-O-S, Si
Bl-O-S, Si
2.3
66.8
14.9
58.9
67.4
85.3
40.5
31.6
31.4
49.2
<10
30
410
30
510
80
50
150
10
20
1.2
41.5
0.4
35.9
38.4
50.2
6.4
0.7
18.9
29.4
0.0
2.6
0.02
1.8
2.4
2.7
0.4
0.03
1.2
1.8
__
16.3
21.0
19.5
16.5
18.5
17.3
24.3
16.0
16.5
'Lakes Santa Fe to Tuscawilla are in Alachua County;  Lakes Apopka to Weir
 are in the Oklawaha River Basin; remaining lakes  are in  the Trail Ridge
 (sandhill) region of the central highlands.  See  Brezonik et_  al. (1969)
 for lake locations.

2Sediments typed according to color, odor, and texture:   Br=brown, DB=dark
 brown, LBr=light brown, Bl=black, G=gray, 0=H S odor, N=No  odor, S=sand,
 Si=silt, Or=fine organic matter, P=peat, D=debris (roots,  leaves, etc.),
 M=muck.

3Expressed on a dry weight basis.
                                  97

-------
Hawthorne  (Alachua  County).

Only one lake had a C/N ratio less than ten (on a weight-weight basis) .
The total  range in C/N ratios was 9.8 to 29.6, but most of the ratios
were between 11 and 18.  Thus all the lake sediments are relatively
enriched  in carbon compared to nitrogen.  There is a slight trend toward
lower  C/N  ratios in the more eutrophic lakes, but the data are highly
scattered.  Figure 22 presents a plot of sediment C/N ratio vs. a rela-
 tive ranking of trophic state (Shannon and Brezonik, 1971b) for the 55
 lakes.  Similarly a plot of total organic nitrogen vs. total phosphorus
 in the lake sediments  (Figure 23) shows no trend toward a constant
 ratio; N/P ratios vary from 0.7 to over 40 in these sediments.  The
 "ideal" N/P ratio  (by weight) in organisms is about 7 (assuming the fre-
 quently cited elemental ratios C:N:P of 105:16:1 by atoms).  Most of
 the sediments have N/P ratios much larger  than  7, indicating relative
 nitrogen  enrichment.   Thus on this basis alone  one would conclude that
 phosphorus is the more likely limiting  nutrient in most Florida lakes.
 However too much emphasis should  not be placed  on these figures.  The
 ecological significance of nutrient  ratios in sediments remains obscure;
 certainly the activity or mobility of  the  species (i.e. their relative
 abilities for recycling) are  more important  in  determining what is limi-
 ting  than are  static  concentration ratios.

 The second phase of  sediment  studies  involved  evaluation of the rates
 of nitrogen  exchange  between sediments  and the  overlying water.  Most
 of the work  on this  aspect has  been  conducted  on estuarine sediments
  (from the Waccasassa  Estuary on the  Florida  Gulf Coast),   (Brooks,
  1969).  Results from a series of ammonia  exchange experiments demon-
  strate the significance  of  sediments as nutrient sinks  and indicate
  that  at  least  some sediments have a  greater  tendency  to remove nutrients
  from  solution  rather  than  to release them to the overlying water.

  An  initial experiment with  estuarine sediment incubated in aquaria
  indicated that ammonia levels tended to decline in  the  overlying water
  with  time whether the aquatic microorganisms were eliminated by various
  sterilization  techniques  or not.  These preliminary results suggested
  that  these sediments act  as a nutrient trap  probably by simple sorption
  onto  clays or  organic colloids and that from a net  viewpoint  release of
  ammonia  to the overlying water is unimportant.   Several further experi-
  ments were conducted to  verify this  supposition.

  Three sediment-water aquaria were set up  as  closed  systems in which
  any  free ammonia evolved  as  a gas could be trapped  and  measured.
  Ammonia  free air was  bubbled through the  water to keep  it  in  slow cir-
  culation but the sediments  were left undisturbed and developed a sharp
  sediment-water interface  and oxidized microzone in  a few days.  After
  the  aquaria were equilibrated ammonia was added to  two  aquaria to raise
  aqueous  concentrations by  about 0.1 and  0.5 mg N/l respectively, and
  concentrations in the water were monitored over a six day  period  (Fig-
  ure 24).   In the aquarium with no addition,  ammonia remained  relatively
  constant, but  in both aquaria with added  ammonia there  was a  rapid  loss
  from the water with final  values in  the vicinity of 0.10 mg N/l.  In
  no  case  was a  significant  amount of  ammonia  stripped from the water;
                                     98

-------
   25
                                                                     O
I-



fc
CO
ILJ
UJ
—
a
20
   15
   10
       o o
         O
                    o  o
                              o
                         00
                             o
       oo°o°
                  o
                                     o
                                  00
                                             o
                                              o
                                                  o
                                   o
                                       o o
                                    o
                           o
                               o
                                              o
                                                    o
                                                       o
                                                             o
                                                                     o
                                                                      o
                                                            o
                                                       o
                                                              o o
                                                               o
                 10
                           20
30
40
                                     Trophic State Rank
50
60
          FIGURE 22,  CARBON TO NITROGEN RATIO  IF; 55 LAKE SEDIMENTS AS A  FUNCTION  OF  A
             RELATIVE TROPHIC STATE RANKING (FROM BREZONIK l?71c),  MOST  EIFTROPHIC
              LAKE AT LOWEST END OF RANK; MOST OLIGOTROPHIC LAKE AT HIGMEST  E^SD,

-------
0.40
4-1
| °'32
(G
3
O
X
o, 024
en
o
iH
to
0
H
0.16

0.08
0.0
O '
/ 0
o/
o/
/
/
/ o
o f
' ° /
' 0 0 0
,00 0
/ 0 0 O
0 'Q£ 0 ° ° 0 g o $
° /* 0 °
               0.8         1.6          2.4         3.2

                       Total Nitrogen (% dry wt.)
4.0
FIGURE 23,  TOTAL PHOSPHORUS VS TOTAL NITROGEN IN 55 FLORIDA LAKE

  SEDIMENTS,  DASHED LINE REPRESENTS IDEAL N:P RATIO OF 7:1
   (VIT.AfT,) IN MICROORGANISMS, VALUES BELOW THE LINE INDICATE
     RELATIVE ENRICHMENT OF N IN SEDIMENT; SEDIMENTS OCCURRING
              ABOVE THE LINE ARE RELATIVELY ENRICHED IN P,
                              100

-------
o
             0.6
             0.5
              0.2
              0.1
                                                                      O  Control

                                                                      C  0.1 mg N/l added

                                                                      •  0.5 mg N/l added
                                                           Day

                           FIGURE 2L\,  UPTAKE OF AMMONIA BY F.STUARINE SEDIMENTS IN AQUARIA,
                           V,'AS  COMTINUALLY ilIXED BY BUBBLING AIR 3UT SEDIMENTS WERE UNDISTURBED,

-------
the maximum ammonia  caught  in the acid  trap was  5vig  as  N,  and  the  pH
levels in all  three  aquaria remained  in the range  8.0-8.3  through-
out the experiment.   Aside  from sediment uptake, the only  likely sink
would be assimilation by algae and bacteria.   The  water was filtered
before being added to the aquaria, so initial  organism  levels  were low,
and there was  no visual evidence for  growths either  in  the water or
on the aquaria walls.  While not perhaps unequivocal, the  results
strongly suggest th&t sediments from  the Waccasassa  Estuary act  to sorb
ammonia from solution.

One question raised by the  above experiments was whether ammonia uptake
by sediments was biologically mediated  or a strictly chemical  pheo-
memon.  A  sediment-water system which was sterilized by cobalt irra-
diation gave  similar rates  of ammonia uptake as  an unsterilized  system
did  (Figure 25).  The results fit neither a first  nor second order
rate  expression but could be approximated by two zero order expressions:
a rapid  initial uptake (k=1.3mg/1-day)  followed  by a slower rate  (k=0.03
mg/l-day)  occurring after several days.  This  result is reminiscent of the
kinetics  of phosphate sorption by estuarine sediments reported by  Carritt
and  Goodgal (1954), uptake  of pesticides by clays  (Haque et. al.,  1968),
and  sorption of alkyl benzene sulfonates by activated carbon (Weber and
Rumer, 1965).   In all these cases the rates can  be explained by a  mechanism
 involving rapid sorption onto the particle (sediment)  surface  followed by
 a slower  (rate limiting) intra particle diffusion  controlled process.

 Several short term ammonia  sorption-leaching  experiments also  indicated
 the non-biological nature of the phenomenon with rapid  uptake  of  ammonia
 from the aqueous phase to the sediments.  Table  26 summarizes  the  results
 from a three hour experiment in which 300 ml.  of estuary water was added
 to four flasks, and  22 g. sediment was  added  to  two  of  these.   An  identical
 amount of ammonium chloride solution  was added to one of the pure  water
 and one of the sediment water flasks, and samples  were immediately taken
 from all four flasks  for ammonia determinations.  The flasks were  kept
 in a constant temperature shaker-bath and replicate  analyses were  also
 made after one and three hours.  No change was noted in the water  only
 flasks and in the sediment-water flask  without added ammonia during the
 three hours, but a rapid uptake of ammonia occurred  in the sediment
 flask with added ammonia.  About ninety-four  percent was lost  from the
 water within  the three hour period and  over  sixty percent was  sorbed  onto
 the sediment  immediately as indicated by the  0 hour  concentration of  2.7mg
 N/l in this flask compared 6.8mg N/l  in the  flask with added ammonia  but
 without sediment.

 The above experiment was also repeated using estuary water from which the
 background ammonia was  removed by boiling and with sediment samples from
 various depths  in a  core (interface  (top cm),  middle (1-10 cm) and the
 bottom of the  core  (below  10  cm).  In  no instance was free ammonia detected
 in the water  phase  at the  termination  of the experiment.  Uptake of various
 ammonia concentrations  (1.5  to  4.5 mg  N/l) by 10 to 30 g. fresh sediment
 was also determined  using  shaken flasks  as above with a contact tine of 6 h
 hours.  Free  ammonia levels  in  the water after incubation ranged from O.-O
 to 1.8 mg N/l,  but  results did  not conform to the Freundlich isotherm.
 Finally, several  experiments  were conducted
                                   102

-------
                                           O  Fresh Sediment

                                              Sterilized Sediment
                              Time   (days)

FIGURE 25, DECRFASE IN A^FiOUS AMMOTJIA CONCENTRATION '/ITH  TF'E  IM
        AQUARIA V'lTH FRESH AMD STERILIZED ESTUARY  SEDU^NT

-------
                         Table 26.
     Uptake of Ammonia by Waccasassa Estuary Sediment
         in Short-Term Shaker-Flask Experiment1
Estuary
Water"2
Control (no sediment)
5 min
1 hr
3 hr
0.
0.
0.
0.
11
11
11
10
Estuary Wat
Added NHj
4
6.80
2.70
1.06
0.35
1From Brooks (1969).   Four flasks had  300 ml estuary water added.
 Ammonium chloride solution was  added  to  2 flasks in identical
 amounts and 22 g fresh sediment was added to one of these.  Of
 the 2 remaining flasks, one was a  control and the other had 22 g
 sediment added to it.

2Ammonia concentrations in mg N/l;  controls (no sediment) remained
 constant throughout  the experiment.
                             104

-------
with blended sediments.  In these cases as much as 0.34 mg NH^-N could
be obtained per gram (dry weight) of freshly homogenized sediment.

Ammonia within soil (and presumably) sediment exists both in a form
exchangeable by other  cations such as sodium, calcium or potassium and
in the fixed or non-exchangeable state (Bremner, 1965; Jackson, 1958).
In an estuarine environment with high concentrations of sodium, magne-
sium, potassium and calcium present, it might be predicted  that  exchange-
able ammonia on the sediments would be small, and the absorbed  ammonia
might be present in the non-exchangeable form.  This agrees with the
above results in which little tendency for ammonia leaching was ob-
served.  However it should be noted that the concentrations of sodium
and other cations in a marine or estuarine environment do not approach
the levels used in standard soil chemistry to displace exchangeable
ammonia  (Jackson, 1958).  Bremner  (1965) has indicated that grinding
of sediments would release a portion of the fixed ammonia.  This was
found to be the case with Vaccasassa Estuary sediments homogenized in
a Waring blender.  Probably some of this ammonia resulted from dis-
ruption of cellular material.

The above experiments  indicate  that ammonia uptake and release by
Waccasassa Estuary sediments is largely controlled by strictly chemical
(sorption) phenomena.  X-ray diffraction analysis of  the sediment showed
the presence of small  amounts of clays, primarily of  a kaolinitic nature,
but further experiments  are necessary  to describe this sorption mathe-
metically.

Rapid  losses of ammonia  from solution  have also been  found with lake
sediments in some preliminary experiments.  Figure  26 shows  the ammonia
remaining in solution  vs.  amount of  ammonia added to  sediment  (dry weight)
from Bivin's Arm, a highly  eutrophic  lake near Gainesville,  Florida.
In  this  experiment 10  ml of  fresh  sediment, equivalent  to  0.4  g. dry
weight, was suspended  in 100 ml demineralized water  in  125 ml  Erlenmeyer
flasks,  which were continuously shaken on a shaker  table  for 2  hours.
At  low  concentrations  of added  ammonia the  levels after  incubation were
higher  than the added  levels;  for  example at  0.1 mg  N added  0.45 mg N
 (both  per 100 ml  sample)  was measured  in  solution after  incubation.
But  at  higher levels  of  added  ammonia  the amounts left were  relatively
small  (e.g. with  5.0  mg  N added 0.92 mg N remained  in the  100  ml sample).
These  results can be  interpreted more  readily in light  of  Figure 27
which  plots  the  amount of ammonia  leached from sediment  vs.  the amount
of  sediment  added to  100 ml of  deionized water.   Thus,  with 0.4 g.  (dry
weight)  sediment  suspended in  100  ml of water,  0.37 mg annonia N is
 released, but the amount of leaching is highly dependent on the sediment
 /water ratio.   At 2.0 g. (dry weight)   per  100 ml of water (a 5-fold
 increase)  only  0.65  mg ammonia  N is released (a less than 2-fold in-
 crease).  These results  seem to suggest that  sediments  from Bivin's Arm
 act as a buffer to  maintain a  certain level of ammonia in their overlying
                                   105

-------
 9.0 .-
           10    20     30    40    50

           Ammonia Added (mg N/l)
                                         DO
                                         a
                                         -a
                                         0)
                                         .o
                                         ij
                                         o
                                         tfi
                                         -a
                                         c
                                         o
                                            6.0 n
                                            4.0
2.0
                                           -2.0
                                           -4.0
0
         1.0    2.0    3.0    4.0

        Ammonia Added (mg  N)
                                                                              5.0
  7.0
25

sc
         FIGURE 26, AMMONIA UPTAKE BY LAKE SEDIMENT:  (A)  AMMONIA CONCENTRATION
          REMAINING AFTER 2 HOURS INCUBATION AS  FUNCTION  OF INITIAL AQUEOUS
            CONCENTRATION; (B) NET SORPTION OF AMMONIA (AMDUNT ADDED MINUS
             AMOUNT REMAINING) vs INITIAL AMOUNT OF AMMONIA ADDED TO WATER
o
M
01
   5.0
                             O
o
E
   3.0
   1.0
                 0.4         0.8          1.2        1.6         2.0

                            Sediment, g. (dry weight)

                    FIGURE 27. AMMONIA LEACHED AS A FUNCTION OF AMOUNT

                    OF LAKE SEDIMENT ADDED TO 100 ML DEIONIZED WATER
                                       106

-------
(or interstitial) waters.  When water low or entirely depleted in ammonia
is placed in contact with these sediments, they tend to release ammonia
to the water, but when more concentrated ammonia solutions are contacted
with the sediments they sorb the ammonia decreasing solution concentra-
tion toward some equilibrium value.

Obviously much remains to be done  to characterize the equilibria and
kinetics of ammonia exchange between sediments and water.  However,  the
above experiments indicate that rapid and large exchanges apparently
mediated by simple sorption phenomena are possible for nitrogen as well
as for phosphorus.  While the nitrogen cycle within the water column
is nearly exclusively a biochemical phenomenon, non-enzymatic processes
may be more important within sediments and at their interface with water.

While sediments  can act as either  in situ sources or sinks, denitrifi-
cation acts solely as a sink.  As  described in Section IV, this reaction
occurs in water when oxygen is depleted and is mediated by facultative
bacteria who use nitrate as a terminal electron acceptor (in the place
of 02) with the  major reduction product being N2.  However, varying
amounts of ammonia and organic nitrogen are apparently always produced
in situ simultaneous with denitrif ication, as  has been shown by a num-
ber of investigations (see Section IV) including one supported by this
project (Tang. 1968).  In the latter study from 70 to 80 percent of
added nitrate nitrogen was recovered as N~ in  laboratory sludge digesters,
and the remainder was reduced  to  ammonia and  organic nitrogen.  This
phenomenon occurred even at high ammonia  levels where nitrate assimi-
lation would seen a superfluous nitrogen  source.  It is not possible to
ascertain from such mixed culture  experiments  as these and others
(B-ezonik and Lee, 1968; Goering,  1968; and Keeney £t al., 1971) whe-
ther  the biological agents of denitrification  are the same as those
reducing nitrate to the  level of ammonia, but  from an ecological view-
point this is perhaps unimportant. What  is important however is that
both  reactions occur whenever oxygen  is depleted in  a nitrate containing
environment.  Thus denitrification (i.e.  nitrogen lost to  the system)
is not equivalent to nitrate  lost  and  any attempts  to assess  the sig-
nificance of denitrification  as  a  nitrogen sink must take  this  into
account.

Denitrification  in Florida  lakes would seem  to be of minor importance
although there may be unique  situations where  this  statement  is not
valid.  In the study  of  55  lakes  in north central Florida (Brezonik,
197Ic) only  about 8-9 of  them appear  to  develop  stable summer strati-
fication, and of these  only  4  develop  anoxic  conditions  in a  signifi-
cant  volume  of  their hypolimnia.   In  addition  several other  lakes
develop  temporary stratification and  lose their  oxygen in the bottom
half  to one  meter,  probably  because of a  high  sediment oxygen demand.
Denitrification  would be possible  in  these  regions;  however from a
quantitative point  of view  they  represent only a  minor  fraction of  the
water volume in  the  lakes.   Further nitrate  concentrations in Florida
lakes are  typically  low.  Most  of  the waters  are  soft and acidic, and
nitrification  is not  favored by  such  environments.   Hence most of the
inorganic nitrogen  in Florida lake waters occurs  as ammonia.   In those
lakes which  stratify  nitrate levels greater  than 0.10 mg N/l  are rare.
                                107

-------
Thus, denitrification does not appear to be a significant nitrogen sink
in Florida lakes and it vas felt that a detailed evaluation of its
occurrence was not worth pursuing.   On the other hand, denitrification
may be important in areas of Florida where nitrate rich ground water
seeps through organic and anaerobic sediments or where surface drainage
enriched  in nitrate from agricultural fertilization passes through
anoxic soils or sediments.  However, this aspect was beyond the scope
of the present project.
                                   108

-------
                                SECTION IX

                        ANALYTICAL INVESTIGATIONS
While research on the analytical aspects of nitrogen in natural waters
was not a primary function of this project, some work was undertaken to
evaluate procedures used in the main phases of this research.  Because
the acetylene reduction method was used extensively in the nitrogen fix-
ation studies and since this is a relatively new method (first reported
by Stewart et_. at. , 1967), an evaluation of the procedure was considered
essential.  Also, a large number of inorganic nitrogen (ammonia and
nitrate) analyses of natural waters were performed during this project
and also in conjunction with other investigations on lake eutrophication
conducted in  this laboratory.  A number of difficulties were encountered
..-;ith the usual automated procedures for these constituents, especially
interferences arising from the high color content of many Florida lakes.
Consequently  a study of these procedures was undertaken to devise methods
of overcoming thesa interferences.  The acetylene reduction assay ralies
on the  fact that  acetylene acts as a  competitive inhibitor of nitrogen
with the enzyme nitrogenase, which reduces it to ethylene.  Production
of the  latter gas can be quantified by flame ionization gas chromato-
graphy. and the method, though an indirect assay, is rapid, highly
sensitive, and inexpensive.  Laboratory experiments were  conducted  to
determine optimum incubation conditions, to evaluate the  method as  an
indirect assay for nitrogen  fixation,  and to study environmental con-
ditions affecting nitrogen  fixation by blue-green algae.

The basic procedures used  for  acetylene reduction assays  of  sediments
and waters are described in  Appendix  A.  Figure  28 shows  a typical
chromatogram  obtained  from one cc. of gas phase  in an  incubated sample.
The column of Poropak R cleanly separated ethylene from  any  traces  of
ethane  (sometimes present  in the  acetylene)  and  from the  large peak
of acetylene.  h'arly in the  development of  this  technique for  assays
of lo*T  fixation rates  in sadiments  it was  recognized that controls  were
necessary with each set- of samples.   These  consisted of  samples carried
through the basic sequence  (Appendix  A)  except  that  1  nil  of  50 percent
trichloroacetic acid was added before inserting  the  acetylene.  Con-
trols were necessary because  rather wide variations  of background
ethylene occurred in the   purified"  acetylene  used.  Ethylene  levels
were especially high in  full tanks, but contaminant  levels declined
with use.  In order  to  avoid complications  arising  from  very high
backgrounds  it was  ^our.d expedient  to waste  about  the  first  third  of
a  t ank.

Sediments  for these  evaluation studies wsra  obtained from the  Wacea-
sassa Estuary (Gulf  Coast  of Florida) and  from Lake  Kanapaha,  near
Gainesville/Florida.   Water samples  for  this  phase  were obtained  from
                                    109

-------
           CMROMATOGRAPH
           RESPONSE  	 ^
APPROXIMATE TIME OF TRAVEL (MIM)
                          o

                          0.5
                                             n
                                            "*
                                            cr«
2.0
                                                2.5

4.0
INTEGRATOR
TRACING
         FIGURE 28,  TYPICAL CHTOMATOGRAM  ILLUSTRATING  IDEMTIFICATION
     OF PEAKS AND MEASUREMENT OF PEAK AREA FOR ACETYLENE REDUCTION ASSAY,

-------
a small pond on the University of Florida golf course, which conveniently
has a dense bloom of Anabaena (2 species) throughout the summer.   Table
27 presents replicate data for 3 pond samples incubated for one hour
and 8 samples incubated for 2 hours under laboratory conditions.     Mean
C2H^ production from the two hour incubations was 51.0 ± 1.9 nmoles
c2H4/m£ N-hr. at the 95% confidence level.  The mean one hour incuba-
tion rate was 84.2 nmoles/mg N-hr.  The replication is obviously within
acceptable limits for biological phenomena.  However, the rates decreased
rather significantly with longer incubation times, probably because of
bottle effects.  Possibly, removal of N2 from the organisms causes nitro-
gen starvation over this period of time.  Table 28 presents results of
a replication using Waccasassa Estuary sediments.  A mean of 11.5 ng,
ethylene per g. dry wt. sediment-hr. with a relative standard devia-
tion of 3.6 percent was found for six replicates, again indicating the
methodogy has adequate precision.  Figure 29 shows the effect of incu-
bation tine on the amount of ethylene produced by estuary sediments.
Up to about one hour the rate appears linear, but after that it drops
markedly.  Thus the efficacy of short incubations is apparent.

The effect of light and the presence of N2 on the time course of acety-
lene reduction was also studied on water from Golf Course Pond.  One
set of samples was purged with the 02-C02-Ar mixture to remove N2;
another set was left unpurged.  A set of purged and unpurged samples
were each incubated in the light and in the dark for times ranging from
0.5 to 3 hours (Figure 30).  Correlation of nitrogen fixation rates
with primary production is evident from the much higher rates of ethy-
lene production in the light than in the dark.  However, fixation is
evidently not completely dependent on photosynthesis to supply reduced
hydrogen (NADPH2) for fixation; otherwise fixation would have ceased
nearly immediately after light was removed.  Reduced hydrogen apparently
is also supplied by respiration (oxidation of organic compounds).  In
all cases, the rate of ethylene production decreased with longer incu-
bation times but the effect was most dramatic in the light incubated,
unpurged set.  Rates were linear for about an hour in the light and
for about the same time in the dark although variability in the dark
incubation data do not permit great accuracy in this conclusion.  The
decline in production with time could result from bottle effects, nitro-
gen starvation (in purged samples), organism shock from manipulation
or from incubation conditions (since temperature and light conditions
in the lab differed from the natural conditions).  Removal of N2 did
not significantly affect the rates measured in dark incubations, but
these rates were low and not as precise.  In the light, the presence
of N2 has a depressant effect, especially over longer incubation times.
However the magnitude of the depression is not too large to preclude
the direct addition of acetylene to unpurged samples if field conditions
prevent purging.  Based on Figure 30 results obtained in this manner
would be 25 to 30 percent too low (assuming incubations of 0.5 to 1
hour) but this error may be acceptable in surveys or in difficult field
circumstances where transport of purging tanks would be cumbersome or
otherwise undesirable.

The effect of N2 on acetylene reduction was studied in greater detail
in a further experiment on Golf Course Pond water.  Since C2H2 is
                                 111

-------
    Table 27.  Replicate Data on Nitrogen Fixation by
Natural Population of Anabaena sp.  from Golf Course Pond


Sample No.
1
2
3
4
5
6
7
8
9
10
11
Incubation
Time
(Hours)
1
I
I
2
2
2
2
2
2
2
2

C2H, Produced
Nanomoles
12.0
11.2
9.7
11.6
13.7
13.6
16.6
16.9
16.9
12.8
15.6

Fixation Rate
nrooles 02*1 //mg N-hr
85.7
86.2
80.8
48.3
48.9
52.3
55.3
52.8
52.8
53.3
52.0
 1See text for details of incubation and statistical analysis
 of results.
                           112

-------
Table 28.  Replication of Acetylene
   Reduction by Estuary Sediments
Sample
Dry Weight
g.
4.827
4.631
4.971
4.961
4.582
4.760



Ethylene
Nanomoles
2.00
1.83
2.16
2.08
1.83
1.91



Production
nM/g. dry wt
0.414
0.396
0.436
0.418
0.400
0.403
x = 0.410
s2 = 0.0062
                  113

-------
M
JZ


u  20
-o

00
 CM
u

00
c
o

^
o
u
PL,

0)

0>
W
   10
    0
                                                   t
                                                                                                O
                           i.o
                                                  2.0


                                         Incubation Time (hr)
                                                                         3.0
                                                                                               4.0
               FIGURE 29,  EWLENE  PRODUCTION VS, TIME FOR WACCASASSA ESTUARY SEDIMENTS

-------
 100
0)
4J
td
P
o
t-i
03
  80
oo
0^60
'O
Q)
0
3
T3
O
J-i
0)
c
4J
40
   20
                              1.0
                                                   2.0
                                           Tine (hr)
                                                                               3.0
       FIGURE 3H, TIME COURSE OF ACETYLENE REDUCTION  BY NATURAL ANA?AEf'A "OPULATIO'

            UNDER CONDITIONS OF LIGHT AND DAR!<  IM  T-E PRESENCE OF 1,78 ATM :^

                 (opE^! CIRCLES) AND  r^ T-F. A~SE;CE OF ',q (CLOSED CIRCLE),

-------
apparently competitive  inhibitor  of N^  the converse should be  true,
and a reciprocal  (Lineweaver-Burk) plot  of reaction velocity vs.  sub-
strate concentration should  give  a competitive  inhibition  pattern
(Figure  31) .   This  pattern was  found  for the natural Anabaena popu-
lation in  the pond  (Figure 32).   In this experiment varying amounts
(05   1.0, 3.0,  5.0 ml) of acetylene  (the substrate) were  added to
serum bottles with  25 ml of  pond  water both in  the presence of  inhi-
bitor N? (0.78 atm) and its  absence  (by  purging air  from the bottles
with  the 02-CO?-Ar  gas mixture).   Samples were  incubated for  one hour
 in the light at 22°C.  Competitive inhibition patterns (Cleland, 1963)
 are given by inhibitors that combine  with or  react at  the  same  active
 site on the enzyme as the normal  substrate.   The molecular similarity
 of N9 and C7H9 suggests they would react at  the same site on  nitroge-
 nase and  that each should thus competitively  inhibit reaction of the
 other.  Schollho'rn and Burris (1966,  1967)  found competitive  inhibi-
 tion of N2 reduction by acetylene, but  Burris (1969)  reported that
 Hwang and Burris (1968) did not.

 Differentiating between competitive  inhibition  and other patterns
 (e.g. non-competitive  inhibition) is  not always a simple or unequi-
 vocal matter.  The distinguishing feature of  the  former pattern is
 that  the  reciprocal  plots for different inhibitor levels intersect on
 the  ordinate; in the  latter pattern the plots intersect to the  left
 of the  ordinate.   The  physiological significance  of this is that in
 competitive  inhibition the maximum reaction velocity  (V_ ) is  un-
 affected  by  the  presence  of inhibitor.   Vmax is obtained from  the
 y-intercept  of the 1/v vs. 1/S curve on a Lineweaver-Burk plot;  at
 1/S  = 0,  S  = °°  or  in physical  terms S is at a saturating concen-
 tration'and  v is at  a maximum level  ( = V   ).  Since  the y-intercept
 is unaffected by inhibitor  in competitive inhibition  (see Figure 31),
 V    is  unaffected,  meaning  that very  high (saturating)  levels of
 substrate can completely  overcome inhibitor effects.   If  inhibitors
 change  the Lineweaver-Burk  plot  y-intercept values, Vmax  is changed,
 and  even  saturating levels  of substrate cannot completely overcome
 inhibitor effects.   Note  however, that  placement  of lines through  1/v
 vs.  1/S data is  subject to  a  certain amount of experimental error.
 Whether or  not curves for varying inhibitor levels actually intersect
  exactly on the y-intercept  is  always subject to some  uncertainty,
  the magnitude of which increases with scatter  of  the  data.

  What is perhaps  more significant about  Figure  32  than whether  the  data
  yield an exact pattern of competitive  inhibition  is that  acetylene
  reduction rates  can be described by  Michaelis-Menten  kinetics  since
  the reciprocal (Lineweaver-Burk) plots  are linear (within experimental
  error).  This fact supports the  enzymatic basis  for acetylene  reduction,
  and the  fact that N2 inhibits (in some  way)  the rate  of acetylene re-
  duction  is  further evidence that ethylene production  from acetylene
  is mediated by nitrogenase and is therefore  an indirect assay  of nitro-
  gen  fixation.  There is at present  no  evidence in the literature for
   the production of ethylene from acetylene by any non-nitrogen fixing
  organisms;  however  the possibility that such organisms might exist
  should be kept  in mind.  These  techniques  (i.e. determination of N2
   inhibition  and  of  fit  to enzyme (Lineweaver-Burk or Michaelis Menten)
                                  116

-------
                                                                        1 = 0
                                     I/(Substrate)


      FIGURE 31. THEORETICAL LINEWEAVER-BURK PLOT FOR COMPETITIVE INHIBITION

  OF ENZYME REACTION RATE, WHERE  h/  I2/  13 REPRESENT INCREASING INHIBITOR LEVELS,
     0.08 P-
 i
•z.
     0.06
sc
E
    0.04
    0.02
                         0.5            1-0

                            l/(Substrate)          (cc C2H2)"



              FIGURE 32,  COMPETITIVE INHIBITION OF ACETYLENE REDUCTION

                      IN NATURAL ANABAENA SP, POPULATION BY ^2'

                                        117

-------
kinetics) should be  useful  in  relating ethylene production  in  an
environmental sample to nitrogen  fixation, and they were  used  for
this purpose in  the  studies of fixation  in sediments  (Section  VII).

As a further evaluation of  the acetylene reduction method,  the effects
of two  other nitrogenase inhibitors on C2H, production were studied
with Golf Course Pond water containing Anabaena sp. Both  carbon mono-
xide and nitrous oxide are  reported to inhibit nitrogen  fixation.
Presumably,  they should also inhibit nitrogenase  reduction  of  C^^2
to C^i,- In one experiment varying amounts of 02^ and CO  were added
to pond samples  and  incubated  in  the light for one hour.  The  results
 (Figure 33)  indicate that CO is a very potent inhibitor of  acetylene
reduction,  and  the data seem to fit a Lineweaver-Burk plot  (Figure 34).
However, for some inexplicable reason the curves  for  various CO levels
intersect to the right of the  y-axis.  The author is  unaware of any
enzyme  kinetic  expressions  yielding this pattern, and this  aspect  of
the  result  would seem an artifact.  Schbllhorn and Burris (1966, 1967)
 found  competitive inhibition of C2H2 reduction by CO, but Hwang and
Burris  (1968)  did not.  Perhaps the intact organism systems are too
 complex to  consistently yield  the simple, pure enzyme inhibition
patterns.   The  inhibition of C2H2 reduction by CO is  good evidence for
 the  biochemical  nature of the  reduction, but the  data are inconclusive
 regarding the  nature of the enzyme inhibition pattern and hence the
mechanism of inhibition.

 In another  experiment, samples of Bivin's Arm lake water  were  purged
 to remove ^j  and varying amounts of C2H2 and N20 were added.   The re-
 sults  indicated a typical Michaelis-Menten response to added C2H2» but
 no significant  inhibition by N^O  was found in the range 0.04 to 0.20
 atm (Figure 35).  There is  no  information in the  literature on the
 range  and nature of  ^0 inhibition of nitrogenase, but Hardy et al.
 (1968)  reviewed other studies  showing that nitrogenase reduces N^O to
 N2-   Possibly  ^0 is much more loosely bound by nitrogenase than is
 C2H2,  so that  even at the lowest  C£H2 concentration and highest ^0
 concentration,  C2H2  may essentially saturate the  enzyme active site,
 at least in comparison to ^0.  The lack of inhibition by ^0  cannot
 be explained by  solubility  differences,  for N20 is somewhat more solu-
 ble_than C2H2,  respective Henry's Law constants at 20°C being  1.98 x
 10"* and 1.21  x 10~3.  Regardless of the explanation  for  these results,
 ^0  would not  appear useful in inhibition studies with nitrogenase.
 It is well known that nitrogenase  is an adaptive enzyme  and  that  cells
 fix nitrogen only when other  available sources  (^3, NOp  are  depleted.
 In this sense, ammonia can be considered to inhibit fixation.  The
 question arises whether ammonia  inhibits the activity of existing
 nitrogenase or represses synthesis of new enzyme.  Another pertinent
 question concerns the level of ammonia which produces enzyme inhibition
 or repression of enzyme synthesis.  Recent data by Hardy ej^  al.  (1968)
 indicates that ammonia does not  directly inhibit fixation but  merely
 represses synthesis of nitrogenase enzyme.  According to Stewart  (1968)
 levels of ammonia in the environment are unlikely to be  high enough
 to inhibit fixation immediately, but the quantitative aspects  of  this
 question are poorly understood.  The immediate  and longer term effect
                                 118

-------
    60
~  ^5
 p*
 i
    30
 c
 c
|   15
 cu
 c
 01
                                                      1 cc C?H9
                                                      3 cc
                                                      5 cc C H
                                                              2
     0
        0           1.0           2.0         3.0         4.0        5.0
                            Carbon Monoxide (cc. added)
   FIGURE 33,   INHIBITION OF ETHYLENE PRODUCTION IN NATURAL ANABAENA
     POPULATION BY  CARSON MONOXIDE AT SEVERAL LEVELS OF SUBSTRATE (ACETYLENE)
s:
c
   0.4
   0.3
P.  0.2
c-
                                               No  Co     0
                                               1 cc CO   9
                    0.2          0.4          0.6         0.8
                          1/S  =  I/(Acetvlene)   (cc C^)"1

            FIRIRE 34, REpLOT OF DATA  FROM  FIGURE 33 ACCORDING TO
                           Llf-JE'/EAVER-BURK  EXPRESS ION,
                              119
                                                                     1.0

-------
 100
4-J
C3
u
   80
CO
a.
€,60
o
c
o
 o
P.-;
 0)
IH
£  20
                 1.0
4.0
                     2.0         3.0
                      Acetylene (cc added)
FIGURE 35, ETHYLENE PRODUCTION RATES FOR NATURAL AWBAENA
5.0
       DOPULATION VS ACETi^E AT FOUR LEVELS OF NITROUS OXIDE
   SHO's' ^ SIGNIFICANT INHIBITION,   SYhBOLS:   CLOSED CIRCLES, 0 ATM
  OPFN CIRCLES/  0.04 ATM N;  DOTS,  0.12 ATM r; TRIANGLES, 0,20 A
                               120

-------
of ammonia on nitrogen fixation by a natural Anabaena population
(from Golf Course Pond) was studied in an attempt to clarify the
matter.  Two liter samples from the pond were inoculated with
0.10 and 0.50 ing NH-^-N/l and one control (no ammonia added) was also
set up.  Samples were  taken from each bottle for acetylene reduction
assay within 20 minutes of ammonia addition, and a second set of sam-
ples were assayed after the bottles had been incubated in situ for
24 hours.  The pond had an especially heavy bloom of Anabaena and
Microcystis species at this time.  Samples for ammonia analysis were
taken initially, immediately after ammonia addition and at the end of
24 hours.  Results from this experiment are presented in Table 29 and
Figures  36 and 37.  Ammonia (up  to 0.5 mg N/l) has no direct inhibitory
effect on C2H2 reduction;  i.e. the initial samples showed no decrease
in C2H^  formation with increasing ammonia concentration.  There was a
depression in C2H^ production  possibly related to ammonia concentration
after  24 hours incubation; however,  the control bottle also showed a
lower  rate after the  twenty-four hour incubation.  This was probably
caused by bottle "wall effects," e.g. death of some N fixing organisms.
Furthermore  the differences in acetylene  reduction between  the control
and  ammonia  enriched  samples  after  24 hours were much larger (on a rel-
tive basis)  for the  lowest acetylene concentrations  (Ice/bottle) than
for  the  highest acetylene  levels.   Experimental  errors tend to be  lower
at higher activities  and higher  substrate levels, so more weight should
be put on these values.  Lineweaver-Burk  plots tend  to emphasise the
differences  found  at  lowest  rates and substrate  levels, where  precision
is poorest,  and in this  sense such  plots  can  be  misleading.  Thus  the
effect of ammonia  on nitrogen fixation after  24  hours is  somewhat  am-
biguous.  A  slight depression seems evident,  even  considering  the  above
cautionary  comments,  but  the  difference in rates between  the  control  and
the  sample with  0.5 mg N/l added was only about  8  percent (for the 5cc
acetylene samples).   These results  therefore  suggest that ambient
ammonia  levels (which in lakes are  infrequently  higher than 0.5 mg
NHo-N/1) should not  directly  influence nitrogen  fixation  rates at  least
over short periods  of time.

The  second analytical study performed in  relation  to  this  project  con-
cerned the evaluation of automated  inorganic  nitrogen methodologies,
particularly in reference  to  interferences  from  organic  color  and
dissolved free amino  acids.   As  part of  our investigations  of  trophic
conditions in Florida lakes  (Brezonik,  1971c)  and  in relation  to the
studies  of nitrogen in these  lakes  conducted  as  part of  this project,
a large  number of  ammonia, nitrite  and nitrate analyses have been  per-
formed in this laboratory  over the  past  several  years.  Because  of the
volume of work involved,  analyses were performed by  automated  (Technicon
Auto Analyzer) techniques.  In the  nitrite and nitrate  procedures
chemistries  employed in  Standard Methods  (A.P.H.A.,  1965)(i.e.,  the
diazotization of  sulfanilic acid and a-naphthylamine by nitrous  acid
to form  a pink  dye and the well-known brucine method for  nitrate)  were
automated  directly.

The  Standard Methods procedures for ammonia are  either  inappropriate
for  the  waters being analyzed (viz.  direct Nesslerization)  or  are un-
suitable for automation  (viz.  distillation and Nesslerization  or titra-
                                   121

-------
         Table 29.   Ammonia Inhibition  of Nitrogenase in a
                        Natural Population of Anabaena sp.
                Initial Effect
                                                     Effect After 24 Hours
Ammonia
Added1
None
(Control)


0.1


0.5


Measured C2H2
Ammonia Added
0.00 0.5
1.0
3.0
5.0
0.09 1.0
3.0
5.0
0,48 1.0
3.0
5.0
Production
Rate of C2H^3
19.2
34.2
102
111
33.9
127
117
31.7
104
123
Measured ^2^?
Ammonia Added
0.03 1.0
3.0
3.0
5.0
0.06 1.0
3.0
5.0
0.09 1.0
3.0
5.0
Production
Rate of C2H^3
2.40
5.28
58.0
67.8
21.3
42.7
60.0
18.5
38.1
62.3
mg N/l

cc. acetylene to 70 ml serum bottle with  25 ml water sample.
3nM C-H, /1-hr . ;  also equal to  nM
 water had 1 mg  particulate N/l.
                                       particulate N-hr. since the
                                     122

-------
                                      O  Control

                                      •  0.1 mg N/l added

                                      A  0.5 mg N/l added
             0.5            1.0

                  I/(Substrate)
(cc
1.5

^-l
                        2.0
    FIGURE 35, ADDITION OF AfWNIA TO NATURAL AMA3AENA SP,
POPULATION HAS NO IMMEDIATE EFFECT ON RATE OF NITROGEN FIXATION,
                         123

-------
                  0.25
           0.50
l/(Substrate)
                                         (cc
 0.75


T N-l
1.00
  FIGURE 37, EFFECT OF AMMONIA ON NITROGEN FIXATION BY NATURAL ANABAENA
      SP, POPULATION AFTER INCUBATION WITH ADDED NH3 FOR 24 HOURS,
         OPEN CIRCLES/ NO AMMONIA ADDED; TRIANGLE/ 0,10 MG N/L
         ADDED; CHOSED CIRCLES/ 0,50 MG fVL ADDED,  DASHED LINE
REPRESENTS INITIAL LINEVJEAVER-BURK LINE FOR ALL SAMPLES (SEE FIGURE
                               124

-------
N ==0 + HC1  (fast)
tion) .   Rather the most common automated method for ammonia employs
the Berthelot or indophenol reaction in which ammonia, hypochlorite
and alkaline phenol react to form a product believed related to indo-
phenol.  The product is intensely blue with an absorption maximum
around 630 nm.  Methods employing this reaction are highly sensitive,
but because of its complex chemistry, irreproducible results often are
obtained by manual procedures.  Bolleter et_ jil_. (1961) have proposed
the following reaction sequence;

     1)  NH3 + HOC1 t NH2C1 + H20   (fast)

     2)  C1H2N ~\ ^   } QH + 2HOC1 -> Cl-N =(   VO + 2H20 + 2HC1  (slow)

     3)  HO-=+ Cl-N =

                                                      (fast)

                                      indophenol blue
But the overall reaction is  doubtless more complex than this.  Weich-
selbaum et_ al. (1969) noted  that  the reaction has a "very odd behavior"
with respect to temperature,  order  and timing of reagents additions,
reagent concentrations, and  so  forth.  They felt that many side and
competing reactions  (some irreversible) result in a situation where the
same equilibrium  condition is never reached twice.  Many of the pro-
blems associated with this method are overcome by use of the Auto Ana-
lyzer since all samples and  standards are treated identically with res-
pect to order and  timing of  reagent addition, temperature, reaction
time before measurement, etc.   With the Auto Analyzer this method
essentially becomes  a kinetic analysis, i.e. the concentration of
ammonia in a sample  is determined by the amount  reacted after a cer-
tain  time, not by the  amount of product  that would be  formed  at equili-
brium.  However,  because  of  this fact,  the  method  is particularly  sus-
ceptible  to  errors caused  by catalysts  or  inhibitors, which nay occur
in  samples but not in standards, causing variable acceleration or  dec-
celeration of  reaction rate.  Iron, chromium  and manganese  are known to
catalyze  the  reaction, while oxidants  such  as  persulfate  inhibit  color
development.   It  has also  been  reported (Russell,  1944)  that  certain
amino  acids  respond  similarly to ammonia in this procedure.

A less  common  automated  ammonia procedure  is  the Grasshof method.   This
sensitive method  was developed  for  sea water  analysis  (Grasshof,  1966)
and is  listed  as  Technicon Auto Analyzer "Industrial Method  42-69W." The
chemistry of  this procedure  is  based on the formation  of  a  bromamine
intermediate  from the  reaction  of ammonia  and  bromine.  The  intermediate
reacts  with  iodide to  form iodine,  which combines  with  starch to  form
the familiar  starch-iodine blue complex with  an absorption  maximum at
570 nm.   The  method  is  attractive in certain  respects,  but  it has  a
highly  critical  and  narrow pH range.   This  is  not  an overwhelming pro-
blem in seawater  analysis  since seawater is well buffered within  a
narrow  range,  but it is  a  problem in fresh  waters.  For example,  typical
                                    125

-------
pH values of the 55 Florida  lakes  ranged from about 4.5  to  9.5  or more
(Shannon and Brezonik,  1971c).  The Grasshof method was  tried  in this
laboratory for  some time,  but  satisfactory  results could not be  obtained
with Florida waters,  and the procedure was  abandoned.

Organic  color  is  a common but  highly  variable constituent  of Florida
lakes.   Many of the oligotrophic  lakes in the sand hills are essentially
colorless, but in lakes receiving swamp or  pine  forest  runoff  color
concentrations may reach several  hundred units  (on the  Pt.  scale).  In
fact  color levels up  to 700 units have been found in  Lake  Mize.   Since
the analytical procedures for  inorganic nitrogen analysis  all  involve
measurement  of color  produced by  a chemical reaction, it was  felt essen-
 tial  to evaluate the  influence of naturally occurring organic  color on
 these tests.   Empirical evidence  indicated  at an early  date that color
was affecting the results of inorganic nitrogen  analyses.   In  particular
 more problems were encountered with the  ammonia  analyses than  with nitrite
 and nitrate.   Assuming no interaction between organic color and the
 reagents in the various nitrogen analyses,  one would  expect the degree
 of interference from organic color to be  greatest for nitrate  and least
 for ammonia since organic color absorbs  most strong  at  the yellow end
 of the visible spectrum (around 400 nm) .  The absorption maxima for the
 nitrate (brucine), nitrite  (diazotization)  and   ammonia (indophenol)
 procedures are 410,520 and  630 nm., respectively.

 To apply compensation methods to these  analyses  requires careful consi-
 deration of the reactions involved.  In external compensation, whereby
 a "key  reagent" is removed  from the analytical  stream to provide a color
 blank,  the "key reagent" must fulfill several criteria.  Removing it
 from the system must not  affect the baseline,  and since organic color
 intensity is pH dependent,  reagent removal must not  alter the reaction
 mixture pH.  Finally,  the key reagent must not  react with organic color,
 and its removal obviously must completely eliminate color formation in
 the reactions  involving the analyte.  For the most part, selecting rea-
 gents which met these  criteria was accomplished by trial and error.
 It was  found that sodium hypochlorite was the most suitable reagent to
 remove  in the  indophenol  procedure for ammonia.  Removal of brucine
 from the mixed reagent of the nitrate procedure was  found satisfactory,
 and in  the case of nitrite  analysis, removal of  the N  (1-naphthyl) -
 ethylenediamine reagent met the  established criteria.   In order  to keep
  the  concentration of the  remaining reagents unaltered,  deionized water
 was  used  to replace  the key reagent when color blanks were run.

 To determine  whether the  selected key reagents  allowed  satisfactory
  correction of  analyses by external compensation  (i.e.  color blank  sub-
  traction),  a  series  of experiments were undertaken with the various
  analytical  procedures.   For each analytical procedure  a standard or
  calibration  curve was  prepared in the usual manner using  the  range of
  concentrations normally selected for the particular  constituent.   A
  second, similar curve  was prepared by adding an identical  amount of
  organic color concentrate to  each standard, thus obtaining a  series of
  artificial colored waters having various concentrations of one  of  the
  inorganic nitrogen constituents.  The color concentrate was obtained
  by vacuum evaporation at a temperature  of   about 40°C  of  colored lake
                                    126

-------
water from Lake Mize in a manner similar to that of Black and Christ-
man (1963).   The original sample was concentrated about 50-fold and
then filtered through a 0.45 ym Millipore filter and passed through
a mixed bed ion exchange resin to  remove concentrated inorganic ions.
The "purified11 concentrate was then stored at 4°C in the dark in filled
glass containers to minimize oxidation and photodecomposition.  The
inorganic nitrogen analysis was repeated on each of the color ' spiked'1
samples resulting higher absorbance values for each sample compared to
the absorbances of corresponding concentrations in the standard solu-
tions without color.  Kext, the key reagent was removed from the analy-
tical stream and color blanks were run on each sample.  The results for
each constituent were plotted on one graph and a curve representing
the difference between the  absorbance values of color "spiked
standards and their corresponding  color blanks was computed and also
drawn on each graph.

The results for nitrate, nitrite and ammonia are shown respectively
on Figures 38, 29, and 40.  In  the case of nitrite and nitrate it is
apparent that subtraction  of  the color blank values from the absor-
bances of the color "spiked"  samples results in curves which coincide
exactly with the calibration  curve using pure standards.  This is good
evidence that no interaction  takes place between the key reagents and
organic color, and the results  indicate that the external mathod of
compensating for effects of organic  color  can be successfully applied
to nitrate and nitrite analyses by these procedures.  The validity
of this correction technique  for nitrate and nitrite  assays was veri-
fied by attempting recovery of  a known  amount of standard  from solutions
of varying color.  Solutions  of  0.10 mg NOj" - N/l were spiked with
color concentrate  in  levels ranging from 17  to 405 units,  and nitrite
standards were spiked with color  to give final concentrations  ranging
from 12 to 287 units.  The analyses  and color blanks  were  determined,
and the latter values subtracted  from  the  former  gave  the  corrected
value.  Table  30 summarizes the results and  shows  that  application of
external  compensation yields  good  agreement  with known  concentrations.

Results from  the ammonia experiment (Figure  40) were  at  considerable
variance with  the  above  data.   Subtraction of  the  color blank curve
froa the  absorbance  curve  for the  spiked  standards  produced a curve
 (D) that was still higher than the calibration  curve.  Further the
slopes of the  color  "spiked"  curves were  greater  than that of the  cali-
bration curve, yielding  an increasing divergence  between them at  in-
creasing  ammonia  levels.   Thus, the differences  between the calibration
curve and the  curves  for the  colored samples cannot be explained
solely by assuming that  some  ammonia was  added  to the test solutions
via the color  concentrate.  Since all samples  received the same amount
of  color  concentrate,  the diverging nature of the curves remains  un-
explained.   Interaction of the organic color or of some constituent
associated with  organic  color with the reagents seemed likely.  This
interaction  is further  shown by the data shown in Figure 41.  These
results were  obrained by preparing six series of standards solutions.
Each  series was  given a different color,  ranging from 0 units for
series A  to  244  units for series  F; within a given series of ammonia
 standards (ranging from 0 to 1.0  mg N/l)  each solution had an equal amount
 of  color.   The increasing divergence of the curves as color increased
                                    127

-------
   0.28
   0.24
   0.20
O 0.16
o
C/3
  0.12
  0.08
  0.04 -
           0.0
                                       O "Spiked" Colored Water  (A)

                                       * "Color Blank"  (B)

                                       * Calibration Curve and (A)-(B)
0.2         0.4         0.6

       Nitrate (np N/l)
                                                             0.8
          FIGURE 38,  EFFECT OF ORGANIC COLOR ON APPARENT NITRATE IN
           BRUCINE TEST CAN BE ACCURATELY CORRECTED FOR BY .METHOD
             OF BLANK SUBTRACTION,  COLOR LEVEL IN TEST SOLUTIONS
                      WAS 63 UNITS  (PT SCALE) AT p,H 8,3,
                                 128

-------
0.16
0.14 ,
0.12
 0.10
 0.08
 0.06
 0.04
 0.02 _
  O  "Sulked" Colored Water (A)

  3  "Color Blank" (B)

      Calibration Curve and (A)-(B)
Nitrite
                                           6           8
                                              (yg N/l)
  FIGURC 7\  EFFECT OF  ORGANIC COLOR ON APPARENT NITRITE CONCENTRATION
     CAN BE ACCURATELY  CORRECTED FOR BY METHOD OF BLANK SUBTRACTION,
    ORGANIC COLOR IN TEST SOLUTIONS WAS W UNITS (FT SCALE)  AT PH t.*>,
                                  129

-------
  ,36
   ,32
   ,28
  .24
  .20
LU
CJ

   .16
   .12
   .08
   .04
                                                  A = Calibration Curve
                                                  B = "Spiked" Colored water
                                                  C = "Color Blank"
                                                  D = B-C
          0.0
0.2       0.4
     Ammonia
                                                               1.0
                                 0.6       0.8
                                    (mg N/l)

  P 40, FFRCT OF ORGANIC COLOR ON APPARENT AWONIA CONCENTRATION
CANNOT BE ELIMINATED BY BLANK SUBTRACTION,  ORGANIC COLOR LEVEL
    IN TEST SOLUTIONS WAS 1P5 UNITS (PT SCALE) AT pH 8,3,
                                    130

-------
Table 30.  Application of External Compensation
           Method to Nitrate and Nitrite Samples
         Containing Varying Color Concentrations
Synthetic
Water


A
B
C
D
E
F
G
H


A
B
C
D
E
F
G
Color

Units

0
17
25
50
84
119
208
405


0
12
21
50
130
173
287
Known

N03-N,
mg N/l
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
N02 -N ,
yg N/i
5.0
5.0
5.0
5.0
5.0
5.0
5.0
Apparent

N03-N,
mg N/l
0.10
0.10
0.12
0.15
0.16
0.18
0.25
0.38
N02 -N ,
yg N/I
5.0
5.8
6.0
6.9
9.7
11.2
17.2
Color
Blank
as N03-N,
mg N/l
0.00
0.02
0.03
0,05
0.07
0.10
0.17
0.30
as N02-N,
yg N/I
0.0
0.8
0.9
2.0
4.8
6.1
12.2
Corrected

N03-N,
mg N/l
0.10
0,08
0.09
0.10
0.09
0.08
0.08
0.08
N02-N
yg N/I
5.0
5.0
5.1
4.9
4.9
5.1
5.0
 All  analytical  values  are  the  average  of  three replicates
                           131

-------
   0.36
   0.32
   0.28
   0.24
   0.20
o
t/5
   0.16
   0.12
  0.08  -
   0.04  -
    0.0
                                                                     B


                                                                     A
                                                             Color Units
           0,0   0.1  0.2   0.3  0.4  0.5  0.6  0.7   0.8  9.9   1.0

                                  Ammonia  (mg N/l)


              FIGURE 41,  EFFECT OF  INCREASING COLOR CONCENTRATION ON

             CALIBRATION  CURVES FOR ALKALINE-PHENOL AMMONIA PROCEDURE
                                     132

-------
supports the suoposition that organic color (or some associated con-
stituent) interacts with the reagents in a manner which precludes
correction by external compensation.  In effect addition of organic
color concentrate caused the analytical reaction to proceed toward com-
pletion at a rate increasing in proportion to the amount ef color plus
the concentration of ammonia present.

A. possible explanation for these effects derives from consideration of
the nature of the analytical reaction  (see p. 1?5) for the indophenol
technique and the nature of organic color.  The slow or rate determin-
ing step of the indophenol reaction involves the reaction of monochlora-
nine with phenol  (step 2) .  Organic color is known to be aromatic in
character with many phenolic groups present (Christman and Ghasserai, 1966).
Thus, some of these phenolic groups could possibly react with
monochloramine under the reaction  conditions.  In effect, the presence
of organic color would increase the concentration of phenol, shift the
equilibrium of step ?. to the right, and thus, accelerate the rate deter-
mining step.  As mentioned earlier, the Auto Analyzer measures the
reaction products in this test before  the reaction reaches completion.
Thus, varying amounts of color would cause various responses to the same
ammonia concentration.

If organic color  reacted in the manner just described it might be
supposed that the formation of extraneous color products  (i.e. con-
taining substituted phenols from organic color) would cause a shift in
the absorption spectrum of the reaction products.  This possibility was
studied by measuring the  spectra for several reaction mixtures, each
containing 0.5 me y^-N/l but with color concentrations from 0 to  378
units  (?t. scale).  However,  the abosrption maximum  for all samples was
the same  (63<~t nm.), and the shapes of  the spectra were identical
(^igure 42).  Thus, the possibility of side reactions with phenolic
groups  from  organic color  is  not substantiated, but  it cannot be  ruled
out solely on this experiment.

An alternate, perhaps simpler  explanation  for  the results with ammonia
and organic  color is the  presence  of  the catalyst in the  color concentrate.
It is well known  that organic  color is frequently associated with high
iron levels, and  in  fact  hake  v±z& water has  a  mean  iron  concentration
of 2.4  me/1  (Table  17).   Since  this iron is associated with the  organic
color,  perhaps as P chalate,  or  even  as  an  integral  part  of the  color
molecule, nassine the concentrate  through  an  ion  exchange  resin would
not separate all  the iron  from the concentrate.   As  mentioned  previously,
the indophenol procedure  is  catalyzed  by iron,  and  in fact  some  refer-
ences  on  the indophenol procedure  call for  the  use  of sodium nitro-
prusside  (i.e. sodium pentacyanonitrosyloferrate)  to catalyze  the reaction,
 (e.g.  see Weichnelbaum  zt_.  .<•_.,  1%°) .   This  reagent was  not  generally
used in cur  analyses since  sensitivity ^s  sufficient without  it.  Fur-
ther,  solutions  of  this  species  are unstable  and  have a  relatively short
life tin-p,  and  the  appropriate concentration  for  proper  catalysis is
fairlv critical.  That  iron  associated with  organic color is  the cause
Of  t-n^ r>roblem with  the  indophenol procedure  was  indicated  by  some
recent experiments  in  our laboratory.   Addition of  the sodium  nitro-
prusside  catalvst seems  to eliminate  th=  interference due to  color
 (o^c/^nt for  ti'is  interference  due to absorption ^y color  itself,  which
                                    133

-------
                                             AMMONIA LEVEL:  0,5  mg N/l
  1.10
  1.00
  0.90
  0.80
  0.70
 ;0.60   -
c
C/5
   0.50 -
  0.40
  0.30  .
  0.20
  0.10  -
  0.00
                                                      83
                                                     249
                                                     378
                                                      0   Color Units
                                                            it     n
it    it

II    U
       480  510    540   570   600   630   660    690    720    750   780   810
                               Wavelength  (nm)

               FIGURE 42, ABSORPTION SPECTRA  FOR PRODUCTS OF ALKALINE
             PHENOL-AMMONIA PROCEDURE AT VARIOUS LEVELS OF  ORGANIC COLOR
                                      134

-------
can be corrected by the external compensation method).  However, further
work will be required to verify this and  to determine the optimum level
of catalyst to add.

Free amino acids have been known, to exist in lake water since the work of
Peterson et. al. (1925) on Lake Mendota, Wisconsin.  Since then low levels
have been reported in English  lakes (Fogg and Westlake, 1955) and in marine
surface waters  (e.g. Degens  et. al., 1964).  The excretion of free amino
acids by zooplankton is a well known phenomenon  (Webb and Johannes, 1967),
and blue-green algae have also been shown to excrete nitrogenous materials
including peptides, amides and amino acids especially in enriched lakes
should not be unexpected.  Limited analyses of amino acids in Florida lakes
(Yorton, 1971) indicated the presence  of  low levels  (1.5-12 yg/1) of total
amino acids, but no definitive work has been undertaken.  Because of the
chemically similar behavior  of amino acids to aamonia in various reactions,
a brief study was undertaken to determine whether any amino acids would
react in the ammonia procedures employed  in our  laboratory.

Solutions of 15 common  amino acids and urea were prepared so that equiva-
lent N concentrations were approximately  0,5 mg/1 in each.  Aliquots of
each solution were analyzed  for ammonia nitrogen by microdistillation
followed by Nesslerization,  and all were  found to be free of ammonia.
Aliquots were also digested  by snicro-Kjeldahl methods to determine the
total nitrogen content  of each.  Each  solution was then analyzed by both
the indophenol and the  Grasshof methods for ammonia nitrogen.  The results
shown in Table  31 indicate that both methods are susceptible to considerable
errors from various amino acids.  The  Grasshof procedure is much more prone
to error than the indophenol method, but  all the amino acids tested gave
at least some response  in both methods.   Interestingly urea was not found
to react in either method.

In the Grasshof method  leucine and methionine give approximately the
same response as ammonia.  Considering only amino  groups as reactive,
histidine and argine also fall in  this category  since only one-third
and one-half, respectively,  of their nitrogen is in  amino  groups.  The
other amino acids gave  responses ranging  from 23 to  290 percent of
equivalent ammonia nitrogen  in the  Grasshof method.   In  the  indophenol
method glycine  gave a response equivalent to that  of  asimonia,  and
threonine gave  a somewhat  greater  response;  the  other amino  acids  gave
responses ranging from  4  to  70 percent of the equivalent ammonia nitro-
gen.

The responses greater  than  100 percent can be explained by  the fact
that the procedures do  not measure  equilibrium  concentrations,  but only
the extent of reaction  after a certain time.  The  amino  acids  giving
responses greater  than  100 percent  evidently react more  rapidly than
ammonia; hence  relatively more product has  been  formed  at  the  time of
measurement than is  the case for  an equivalent  ammonia  solution.   The
extent of reaction  of non-predictable  from the  amino acid  structure
or isoelectric  point.   The  reasons  for the relative  responses  of  the
various  amino acids  in  these procedures would be an  interesting study,
but this was beyond  the scope  of  the  present project.
                                   135

-------
                          Table 3.1.  Response of Automated Ammonia
                             Procedures to Free Amino Acids
                                         Grasshoff Method
Alkaline-Phenol Method
Amino Acid
Alanine
Glycine
Arginine
Asparagine
Aspartic Acid
Cysteine
Histidine
Leucine
Lysine
Methionine
Phenylalanine
Proline
Serine
Threonine
Valine
Urea
Free NH
Present
<.001
<.001
<.001
<.001
<.001
<.001
<.001
<.001
<.001
<.001
<.001
<.001
<.001
<.001
<.001
<.001
Total N
Present
.495
.476
.405
.48
.48
.33
.35
.495
.374
.47
.462
.445
.478
.495
.47
.445
Apparent
NH -N
0.60
1.38
.18
.11
.28
.10
.11
.53
.74
.49
.52
1.03
.59
.25
.60
.08
% of Avail-
able N
121
290
44
23
58
31
31
107
198
104
113
231
122
51
128
18
Apparent
NH -N
.23
.47
.02
.02
.06
.05
.03
.07
.05
.31
.15
.17
.34
.60
.04
00
% of Avail-
able N
46
99
5
4
13
15
9
14
13
66
32
38
70
121
9
00
All values in mg N/liter.

-------
The need for better automated procedures for ammonia should be
apparent from the studies  reported above.  The interference resulting
from color or color correlated species is significant and not correct-
able by usual compensation methods.  The source of the interference is
still not known with  certainty, but  iron is the most attractive candi-
date.  If this is substantiated,  at  least this error can be eliminated
by swamping out the natural  iron  in  the sample with added catalyst
(sodium nitroprusside).  However,  the interferences resulting from
amino acids may be more  serious.   The normal levels of free amino
acids in natural waters , compared  to the levels of ammonia are not
well known, but it should  be noted that ammonia concentrations in the
ug N/l range are of great  interest in limiting nutrient studies.  Limi-
ted information suggests that  free amino acids occur in waters in these
concentrations, and the  introduction of serious errors in low level
ammonia analyses is thus a distinct  possibility.
                                      137

-------
                             SECTION X

                         ACKNOWLEDGEMENTS
The able assistance of Michael A. Keirn  and Roger A. Yorton in con-
ducting the research culminating in  this  report is gratefully acknow-
ledged.  Mr. Keirn conducted  or directed  most of the nitrogen fixation
studies and was responsible for the  bacterial isolations.  Mr. Yorton
performed or directed the various chemical analyses throughout the
project.

Carol C. Harper worked on initial phases  of the nitrogen fixation
studies, and Roger King  assisted in  various aspects of sampling and
analysis.  Glenn T. Brasington was most  cooperative in performing
sampling and other duties especially during diurnal and other extended
studies.  The assistance of W. II. Morgan  in fiscal matters greatly
simplified administration of  this grant.  The principal investigator
was Patrick L. Brezonik, who  directed the project and wrote this report,

The support of the project by the Office  of Research and Monitoring,
Environmental Protection Agency, and the  cooperation and assistance
of C. Powers, grant project officer,  are  acknowledged with sincere
thanks.
                                 139

-------
                             SECTION XI

                             REFERENCES
Alexander, V. 1970. Relationships between turnover rates in the bio-
       logical nitrogen cycle and algal productivity.  Proc.  25th
       Industrial Waste Conf., Purdue Univ., Eng. Exper. Sta.  Bull.
       (in press).

American Public Health Association.  1965.  Standard methods  for the
       examination of water and wastewater, 12th ed. New York.  769  p.

Austin, A. T.  1961.  Nitrosation in Organic Chemistry; Sci.  Prog.
       49_, 619.

Benson, D.  1965.  How much is a marsh worth?  New York State Con-
       servationist Bull. No. L-132, N. Y. State Conserv. Dept.

Biggar, J. W. and R. B. Corey.  1969.  Agricultural drainage  and
       eutrophication.  In Eutrophication:  Causes, Consequences,
       Correctives, Nat. Acad. Sci., Wash., D.  C., p. 404.

Billaud, V. A.  1968.  Nitrogen fixation and the utilization  of
       other inorganic nitrogen sources in a subarctic lake.   J. Fish
       Res. Bd. Canada 25, 2101-10.

Black, A. P. and R. F. Christiaan.  1963.  Chemical characteristics
       of fulvic acids.  J. Amer. Water Works Assoc. 55, 753-770.

Bolleter, W. T., C. J. Bushman, and P. N. Tidwell.  1961.  Spectro-
       phonetric determination of ammonia as indophenol.  Anal.
       Chem. 13, 592-594.

Bormann, F. H., G.  E. Likens, D, W. Fisher, and R. S. Pierce.  1968.
       Nutrient loss accelerated by clear-cutting of a forest eco-
       system.  Science 159, 882-884.

von Brand, T., N. W. Rakestraw, and C. E. Renn.  1937.  The experi-
       mental decomposition and regeneration of nitrogenous organic
       matter in sea water.  3iol. Bull., Woods Hole 72, 285.

Bremner, J. M.  1965.  Inorganic forms of nitrogen.  In Methods of
       soil analysis, part 2.  Chemical and microbiological properties,
       CT A. Black  (ed.).  Amer. Soc. Agron., Madison, Wis. p. 1179-1237,
                                141

-------
Brernner, J. "'.. and r?. ?.. Keeney.   1966.   ^termination and isotope-
       ratio analysis of different forms of nitrogen in soils:   4:
       Exchangeable aamoniun, nitrate and nitrite by direct dis-
       tillation methods.  Soil Sci.  Soc. Am.  Proc. 30, 5^3-587.

Bremner, J. M. and T>. W. Nelson.   1968.   Chemical decompostion of
       nitrate in soils.  Trans.  9th  Internat. Coner. Soil Science,
       Vol. II,
Brezonik, P. L.  1°6C*.  Dynamics of the nitrogen cycle in natural
       waters.  Ph.n. Thesis, University of Wisconsin, Madison.

Brezonik, ?. L.  1971a.   Nitrogen:  sources and transformations in
       natural waters.  Proc. Symp. Nutrients in natural waters,
       161st Hat. Amer.  Chem. Soc. x'eet., Los Angeles, April, 1971,
       J. Wiley, Inc. (in press).

Brezonik, P. L.  In71b.   Nutrient sources for Lake >-ize.  Rept. to
       Univ. Fla. School of Forestry, .Austin Carey Forest Committee,
       February, 1^71, Gainesville, Fla.  (mimeo) .

Brezonik, P. L.  1971c.   Trophic states of Florida lakes.  Proj .
       conipl. rept.  to Office of Water Pvesearch, University of Florida
       Water Resources Publ. ,#13.

Brezonik, P. L.  and  C.  L. Harper.  196Q .  Nitrogen fixation in some
       anoxic lacustrine environments.  Science 164 , 1277-79.

Brezonik, P. L.  and  G. F. Lee.   1966.  Sources of elemental nitrogen
       in fermentation gases.   Internat.  J. Air Water Poll. 10, 145-160.

Brezonik, P. L.  and  G. F. Lee.   1968.  ^nitrification  as a nitrogen
       sink in Lake Mendota, Wisconsin.   Fnvir. Sci. Technol. 2., 120-125.

Brezonik, P. L. , W. II. Morgan,  E. F. Shannon, and K. D. Putnam.  1969.
       Eutrophication factors in north central Florida  lakes.   Bull.
       Series 134, Eng.  Indus t.  Expt. Sta., University  of Florida,
       Gainesville, Florida, 101pp.

Brooks, P.. TI., Jr.   1969.  In situ nitrogen dynamics within an  estuarine
       embayment.  Ph.D. thesis. University of Florida, Gainesville, Fla.

Brooks, R., P. L. Brezonik, l\.  D. Putnam, and M. A. Keirn.  Nitrogen
       fixation  in an estuarine environment:  the Waccasassa on the
       Florida Gulf  Coast.  Limnol. Oceanogr.  16:701-710.

Burris, R.  H.   1966.  Nitrogen  fixation.  In Plant biochemistry,
       J. Bonner and J. Varner  (eds.), Academic Press,  New York,
       pp.  961-979.
                                    142

-------
Burris, R. H.  1969.  Biochemistry of nitrogen fixation.  Proc.  Roy.
       Soc. 3 172, 339-354.

Carritt, D. E. and S. Goodgal.  1954.  Sorption reactions and some
       ecological implications.  Deep-Sea Res. !_, 224-243.

Chalupa, J.  I960.  Eutrophication of reservoirs by atmospheric phos-
       phorus.  Sbornik Vysoke  Skoly Chen. Technol. 29, 406.

Christman, R. F. and M. Ghassemi.  1966.  Chemical nature of organic
       color in water.  J.  Amer. Water Works Assoc. 14, 921-927.

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

Cleland, W. W.   1963.  The kinetics  of enzyme-catalyzed reactions with
       two or more  substrates or products. II. Inhibition:  nomencla-
       ture  and  theory.   Biochim. Biophys. Acta  6]_, 173-187.

Cole,  D. W.  and  S.  P. Gessel.   1965.  Movement of  elements  through a
       forest soil  as influenced by  tree  removal  and nutrient  addi-
       tions.   In Forest-soil  relationships  in North America, C. T.
       Youngberg  (ed.),  Oregon  State Univ. Press,  Corvallis, p. 95.

Cooper,  C. F.   1969.  Nutrient  output from managed forests.  In
       Eutrophication:   causes, consequences,  correctives.   Nat. Acad.
        Sci.,  Wash., D.  C. , p.  446.

Cooper,  L. H. N.   1937.   The nitrogen cycle  of  the sea, J.  Mar. Biol.
       Assoc. U.  K. 22_,  183.

Crane, J.  S.   1962.  M.  S. thesis,  Univ. Wisconsin, Madison.

Creitz,  G. I.  and F.  A.  Richards.   1955.  The estimation and charac-
       'terization of plankton  populations by pigment analysis.  III.
        A note on the use of "Millipore" membrane filters in the
        estimation of plankton  pigments.   J.  Marines Res.  14, 211-216.

De.ge.ns   E. T.,  J.  H. Reuter, and K.  N.  F. Shaw.   1964.   Biochemical
       ' compounds in offshore California sediments and seawaters.
        Geochin. Cosnochin. Acta 28_, 45-65.

 Delwiche  C. C.  1956.   Denitrification.  In Inorganic nitrogen meta-
        bolism, J. McElroy  and  B. Glass  (eds.), Johns Hopkins Press,
        Baltimore, Md.

 Delwiche  C  C.  1965.  The cycling of  carbon and nitrogen in  the bio-
        sphere.  In Microbiology and soil fertility, C. M. Gilmour and
        0" N  Allen  (eds.),  Oregon State Univ. Press, Corvallis, p. 29.
                                  143

-------
Dilxrorth, M. J.  1966.  Acetylene reduction by N-fixing preparations
       from Clostridium pasteurianum.  Biochim. Biophys. Acta 127,
       285-294.

Domogalla, B. P., E. 3. Fred, and W. H. Peterson.  1926.  Seasonal
       variations in the ammonia and nitrate content of lake waters.
       J. Amer. Water Works Assoc. 15, 369-385.

Dugdale, R. C.  1967.  Nutrient limitation in the sea:  dynamics,
       identification, and significance.  Linmol. Oceanogr. 12, 685-95.

Dugdale, R. C. and V. A. Dugdale.  1961.  Sources of phosphorus and
       nitrogen on Afognak Island.  Limnol. Oceanogr., 6_, 13-23.

Dugdale, R. C., V. A. Dugdale, J. Neess, and J. J. Goering.  1959.
       Nitrogen fixation in lakes.  Science, 130, 859-860.

Dugdale, R. C. and J. J. Goering.  1967.  Uptake of new and regenerated
       forms of nitrogen in primary productivity.  Limnol. Oceanogr.
       _12, 196-206.

Dugdale, R. C. and J. J. Goering, and J. H. Pvyther.  1964.  High nitro-
       gen fixation rates in the Sargasso Sea and the Arabian Sea.
       Linmol. Oceanogr., 9^, 507-510.

Dugdale, V. A.  1965.  Inorganic nitrogen metabolism and phytoplankton
       primary productivity in a subarctic lake.  Ph.D. thesis,
       Univ. Alaska, College.

Dugdale, V. A. and R. C. Dugdale.   1962.  Nitrogen metabolism in lakes.
       II.  Role of nitrogen fixation in Sanctuary Lake, Pennsylvania.
       Linmol. Oceanogr. ]_•> 170-177.

Dugdale, V. A. and R. C. Dugdale.   1965.  Nitrogen metabolism in lakes.
       III.  Tracer studies of the  assimilation of inorganic nitrogen
       sources.  Limnol. Oceanogr.  10, 53-57.

Englebrecht, R. S. and J. J. Morgan.  1959.  Studies on the occurrence
       and degradation of condensed phosphates in surface waters.  Sew.
       Industr. Wastes 31, 458.

Eppley, R. W. , J.  L. Coatsworth, and L. Solorzano.  1969.  Studies of
       nitrate reciuctase in marine phytoplankton.  Limnol. Oceanogr.
       L4, 194-205.

Eppley, R. W., J.  X. Rogers, and J. J. McCarthy.  1969.  Half-satura-
       tion constants for uptake of nitrate and ammonium by marine
       phytoplankton.  Limnol. Oceanogr. 14, 912.

Feth, J. H.  1966.  Nitrogen compounds in natural water - a review.
       Water Resources Research, 2, 41-58.
                                 144

-------
Fitzgerald, G. P.   1969.   Some  factors  in  the  competition or anta-
       gonism among bacteria, algae,  and aquatic weeds.  J. Phycol.
       1, 351-359.

Fitzgerald, G. P.   1970.   Aerobic  lake  muds for the removal of phos-
       phorus from  lake waters.  Liranol. Oceanogr. 15, 550-555.

Fogg, G- E.  1965.  Algal  cultures  and  p_hy top lank ton ecology.  Univ.
       Wisconsin Press, Madison.

Fogg, G. E. and D.  F. West lake.  1955.  The importance of extracellular
       products of  algae in  fresh waters.  Verb. Internat. Verein.
       Limnol. XII, 219-231.

Foree, E. G. , W. J. Jewell,  and P.  L. McCarty. 1970.  Proc. 5th Internat.
       Water Poll.  Research  Conf.,  San  Francisco, Calif., July.
       Pergamon Press (in  press) .

Fox, J. L., P. L. Brezonik,  H. D. Putnam,  and  S. C. Snedaker.  1970.
       A study of the aquatic and terrestrial  ecology of selected
       karst lakes  and basins near  Flores, Guatemala.   Kept, to Univ.
       Florida.  Biomed. Sci. Support Grant Comm.  47 p. (mimeo).

Gahler, A. R.  1969.  Sediment water  interchange.   In Proc. Eutrophi-
       cation biostimulation assessment workshop,  Univ. Calif.,
       Berkeley, Sanitary  Enginering  Research Lab, and Federal Water
       Quality Lab, Corvallis, Ore.,  pp. 243-257.

Gambell, A. W. , Jr.  1963.   Sulfate and nitrate content of precipita-
       tion over parts of  North Carolina and Virginia.  Geol. Surv.
       Prof. Paper  475-C,  209-214.

Gates, G. A.  1963.  Treatment of Long  Island duck farm wastes.  J.
       Water Poll.  Contr.  Fed. 35,  1569.

Goering, J. J.  1968.  Denitrification  in the oxygen minimum layer
       of the Eastern tropical Pacific Ocean.   Deep-Sea Res. 15,  157-164.

Goering, J. J. and  R. C. Dugdale.   1966.  Denitrification rates in an
       island bay in the equatiorial Pacific Ocean.  Science _L5j4, 505-6.

Goering, J. J. and  V. A. Dugdale.  1966.  Estimation of rates of
       denitrification in  a  subarctic lake.  Limnol.  Oceanogr.  11, 113-17.

Goering, J. J. and  J. C. Neess.   1964.  Nitrogen fixation in two
       Wisconsin Lakes.   Limnol. Oceanogr., j^,  530-539.
                                 145

-------
Goering, J. J. , D. D. Wallen and R. M. Nauman.  1970.  Nitrogen uptake
       by phytoplankton in the discontinuity layer of the eastern sub-
       tropical Pacific Ocean.  Limnol. Oceanogr. 15, 789-796.

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

Goldman, C. R.  1967.  Integration of field and laboratory experiments
       in  primary productivity studies.  In Estuaries, G. Lauff (ed.),
       Amer. Assoc. Adv. Sci., Wash., D. C., pp. 346-352.

Golterman, H.  L.  1960.  Studies on the cycle of elements in  fresh
       waters.  Acta. Botan. Neerl. 9^:1-58.

Grasshof, K.   1966.  Results and possibilities of automated analysis  of
       nutrients in seawater.  In Automation in analytical chemistry,
       Proc. Technicon Symposium, Mediad, New York,  N. Y. pp. 573-579.

Grau, F. H. and P. W. Wilson.  1962.  Physiology of  nitrogen  fixation
       by Bacillus polymyxa. J. Bacteriol.  86, 490.

Hamilton, R. D.  1964.  Photochemical processes  in  the inorganic nitro-
       gen cycle of the sea.  Limnol. Oceanogr. 9_,  107-111.

Hammer, U. T.  1964.  The succession  of "bloom" species  of blue green
       algae and some causal factors.  Verb.  Internat. Verein. Limnol.
       JL5_, 829-836.

Haque, R. , Lindstrom, F. T., Freed, V. H.,  and R. Sexton.  1968.  Kinetic
       study of the sorption of 2,  4-D  on  some clays.  Environ. Sci.
       Technol. 2. 207-211.

Hardy, R. W. F., R. D. Holsten, E. K. Jackson, and  R. C. Burns.  1968.
       The acetylene-ethylene assay for N~   fixation:  laboratory and
       field evaluation.  Plant Physiol. 43_,  1185-1207.

Harper, C. L.  1971.  Ph.D.  thesis, University of Florida, Gainesville.

Hatcher, H. J. and E. L. Schmidt.   1971.  Nitrification  of aspartate
       by Aspergillus flavus.  Appl. Microbiol. 21,  181-186.

Hellebust, J.  A.  1965.  Excretion of some  organic  compounds  by marine
       phytoplankton.  Limnol. Oceanogr. 10,  192-206.

Hoare, D, S.,  L. 0. Ingram, E. L. Thurston,  and P.  Walkup.  1971.  Dark
       heterotrophic growth of an endophytic blue-green algal.  Arch.
       Mikrobiol. _78, 301-321.

Howard, D. L., J. I. Frea, R. M. Beeister,  and P. R.  Dugan.   1970.
       Biological nitrogen fixation in Lake  Erie.   Science 169, 61-62.
                                146

-------
Hutchinson,  G. E.   1941.   Limnological  studies in Connecticut.  IV.
       The mechanism of  intermediary metabolism in stratified lakes.
       Ecol. Monogr.  11,  21-60.

Hutchinson,  G. E.   1957.   Treatise  on limnology, vol. I. Wiley, New York.

Hutchinson,  G. L.  and F.  G.  Viets.  1969.  Nitrogen enrichment of
       surface water by  absorption  of ammonia volatilized from
       cattle feedlots,  Science  166, 514.

Hwang, J. C. and R.  H. Burris.   1968.   Binding  sites of nitrogenase.
       Federation  Proc.  24.  639.

Hynes, H. B. N,, and Kaushik, N. K.  1969.  The relationship between
       dissolved nutrient salts  and protein production in submerged
       autumnal leaves, Verh. Intern. Verein. Liianol. 17, 95.

Jackson, M. L.  1958.  Soil  chemical analysis.  Prentice Hall, Engle-
       wood Cliffs,  New Jersey.

Johannes, R. E.  1969.  Nutrient regeneration in lakes and oceans.
       In Advances  in microbiology  of the sea,, M. R. Droop and
       E. J. F. Wood (eds.).  Academic  Press, New York. Vol. 3^:203-212.

Johnson, W. J. and  G.  J.  Schroepfer.  1964.  Nitrogen removal by
       nitrification and  denitrification.  J. Water Poll. Contr. Fed.
       J36, 1015-1036.

Johnston, W. R. , F.  Ittihadieh, R. M. Daum, and A. F. Pillsbury.  1965.
       Nitrogen and phosphorus in tile  drainage effluent.  Soil Sci.
       Soc. Amer.  Proc. 29,  287-89.

Jones, K. and W. D.  P. Stewart.  1969.  Nitrogen turnover in marine
       and brackish  habitats. III.  Production of extracellular
       nitrogen by  Calothrix scopulorum, J. Mar. Biol. Assoc., U. K.
       .49, 701.

Kahn, L. and F. T.  Brezenski.  1967.  Determination of nitrate in
       estuarine waters.   Automatic determination using a brucine
       method.  Envir. Sci.  Technol., I: 492-494.

Keeney, D. R. , R. L.  Chen, and D. A. Graetz.  1971.  Importance
       of denitrification and nitrate reduction in sediments to
       the nitrogen  budget of lakes.  Nature 233:  66-67.

Keirn, M. A. and P.  L. Brezonik.  1971.  Nitrogen fixation by bacteria
       in Lake Mize,  Florida, and in some lacustrine sediments.
       Limnol. Oceanogr.  16^  (5), 720-731.
                                 147

-------
Kessler.  1964.  Nitrate assimilation by plants.  Ann. Rev. Plant
       Physiol. 15: 57-72.

Kleerekoper, H.  1953.  The mineralization of plankton.  J. Fish.
       Res. Bd. Canada 10: 284-291.

Koyama, T.  1964.  Discussion of Paper III-10.  Proc. 2nd Internal.
       Conf. Water Poll. Res., Tokyo. Pergamon Press, New York.

Krause, H. R.   1964.  Zur Chemie und Biochemie der Zersetzung von
       Siisswasserorganismen, unter besonderer Berucksichtigung des
       Abbaues  der organischen Phosphorkomponenten, Verh. Internat.
       Verein.  Limnol. 15, 549.

Kuznetsov, C.  I.   1959.  Die Rolle von Mikroorganisms im Stoffkreislauf
       der Seen.   Berlin Verlag.

Lackey, J. B.,  E.  W.  Lackey, and G. B. Morgan.   1965.  Taxonomy  and
       ecology  of  the sulfur bacteria.  Bull. Ser. 119 Engineering
       Industr. Exper. Sta., Univ. Florida,  Gainesville.

Lee,  G. F.   1970a.  Eutrophication.  Occasional  Paper No.  2, Water
       Resources Center, Univ. Wisconsin,  Madison.

Lee,  G. F. 1970b.  Factors  affecting the  transfer of materials between
       water  and sediments.  Literature Review  No. 1., Water Resources
       Center,  Univ.  Wisconsin, Madison.

Lee,  G. F. (Chairman), et.  al.   1966.  Report on the nutrient sources
       of Lake  Mendota.  Nutrient  sources  subcommittee of  the Lake
       Mendota  problems  committee, University of Wisconsin, Madison
       (iiimeo) .

Lee,  G. F., E.  Bentley,  and R. Amundson.   1971.   Effect of marshes
       on water quality.  Presented  at XVIII Congress  of Internat.
       Soc. Limnol.,  Leningrad, U. S. S.  R., August.   Verh. Internat.
       Verein.  Liranol. 18 (in press).

Loehr, R.  1969.  Animal wastes — a national problem.  J. Sanit.
       Eng. Div., Amer. Soc. Civil Eng. 95.  189-221.

Mac Isaac, J. J. and  R. C. Dugdale.  1969.   The  kinetics of nitrate
       and ammonia uptake by natural populations of marine phytoplank-
       ton.  Deep-Sea Res. 16, 45-57.

MacKenthun, K.  M., W. M. Ingram, and R. Porges,  Limnological aspects
       of Recreational lakes. Publ. Health Service Publ. 1167  (1964).

Malina, J.   1961.  Ph.D. thesis, Univ. Wisconsin, Madison.
                                 148

-------
Menzel, D. W. and J. P.  Spaeth.  1962.  Occurrence, of ammonia in
       Sargasso Sea Waters  and in rain water at Bermuda.  Limnol.
       Oceanogr. , ]_, 159-162.

Metcalfe, G. and M. Brown.   1957.  Nitrogen fixation by a new species
       of Nocardia.  Jour.  Gen. Microbiol. 17, 567-572.

Miller, C. E. 1955.  Soil fertility.  Wiley, New York.

Miner, J. R. , R. I. Lipper,  L. R. Fina, and J. W. Funk.  1966.
       Cattle feedlot  runoff — its nature and variation.  J. Water
       Poll. Contr. Fed. 38, 1582.

Moe, P. G. ,  J. V. Mannering, and C. B. Johnson.  1967.  Loss of
       fertilizer nitrogen  in surface runoff water.  Soil Sci. 104,
       389-94.

Montelaro, J.  1970.   Personal communication to E. E. Shannon.

Mortimer, C. H.  1941,  1942. The exchange of dissolved substances
       between mud  and water in lakes.  Ecology 29 , 280-329;
       J30, 147-201.

Murphy, J. and J. P. Riley.   1962.  A modified single solution
       method for the  determination of phosphate in natural waters.
       Analyt. Chim. Acta  2.7, 31-36.

Neess, J. C., R. C. Dugdale, V. A. Dugdale, and J. J. Goering.   1962.
       Nitrogen metabolism in lakes.  I.  Measurement of nitrogen
       fixation with  15N.   Limnol. Oceanogr.  7_, 163-169.

Nelson, D. W. and J. M.  Bremner.   1969.   Soil Biol. Biochem. JL,  229.

Odum, H.  T., Nixon, S.  W.  and DiSalvo, L. H.  1971.  Adaptations
       for photoregenerative cycling,  In The Structure and Function
       of Fresh Water  Microbial Communities.  John Cairns, Jr.  (ed) ,
       1-30.

Painter,  H.  L.  1970.   A review of  literature on  inorganic nitrogen
       metabolism in microorganisms.  Water Research  4^  393-450.

Palmer,  C. L.   1950.   The  pollutional effects of  storm-water  overflows
       from  combined  sewers.  Sewage  Industr. Wastes  22,  154.

Parker,  B. C.   1968.   Rain as a source  of vitamin  BI2-   Nature  219,
       617-18.

Parsons,  T.  and J.  D.  H. Strickland.   1963.   Discussion  of  spectro-
       photometric  determination  of marine-plant  pigments, with revised
       equations  for  ascertaining chlorophylls  and carotenoids.
       J. Marine  Res.  21,  155-162.
                                 149

-------
Patterson, W. , Minear, R. A.,  Nedved, T.  K.   1971.  Septic tanks
       and the environment.  State of Illinois Inst.  for Environ.
       Quality, Document No. 71-2.

Paloumpis, A. A., and Starret, W. C.  1960.   An ecological study of
       benthic organisms of three Illinois River flood plain lakes.
       Am. Midland Natur. 64, 406.

Perry, J. H. (ed.).  1950.  Chemical Engineering Handbook, 3rd Ed.
       McGraw-Hill, New York.

Peterson, W. H., E. B. Fred, and B. P. Doraogalla.  1925.  The
       occurrence of amino acids and other nitrogen compounds in
       lake water.  J. Biol. Chem. 63. 287-295.

Polta, R. C.   1969.  Septic tank effluents, in Water pollution by
       nutrients — sources, effects and control.  WRRC, Bull. No.  13,
       Univ. of Minn., Minneapolis, Minn.

Pratt, D. C. and A. W. Frenkel.  1959.  Studies on nitrogen fixation
       and photosynthesis  of Rhodospirillum rubrum.  Plant. Physiol.,
       ^4: 333-337.

Proctor,  V. W.  1957.  Preferential assimilation of nitrate ion by
       H aema tococcus pluvialis.  Amer. J. Bot. 44, 141-143.

Pshenin,  L. N.  1959.  Quantitative distribution of nitrogen-fixing
       bacteria and their  ecology  in  the region of the Zernov
       phyllophora field in the  Black Sea.  Microbiologiia 28, 927-932.

Pshenin,  L. N.  1963.  Distribution and ecology of Azotobacter in
       the Black Sea.  In  Symposium on marine microbiology, C. H.
       Oppenheimer (ed.).  Thomas  Springfield, 111.

Putnam, H. D. and T. A. Olson.   1960.  An investigation of nutrients
       in western Lake Superior.   School of Public Health, Univ. of
       Minnesota, Minneapolis, 24 pp. (mimeo).

Quasim, S. R.  1965.  Chemical characteristics of seepage water from
       simulated landfills.  Ph.D. Dissertation, Univ. of West Virginia.

Rakestraw, N. W. and A. Hollaender.  1936.  Photochemical oxidation
       of ammonia in sea water.  Science 84, 442-443.

Reuss, J. 0. and R.  L. Smith.  1965.  Chemical reactions of nitrites
       in arid soils.   Soil Sci. Soc. Amer. Proc. 29, 267-270.

Richerson, P. J., G. A. Moshiri, and G.  L. Godshalk.   1970.  Certain
       ecological aspects of pollen dispersion in Lake Tahoe (California-
       Nevada) .  Limnol. Oceanogr. 15, 149-153.
                                150

-------
Rittenberg, D., A. S. Keston, F. Roseburg, and R. Schoenheimer.   1939.
       Studies in protein metabolism.  II.  The determination of nitro-
       gen isotopes in organic  compounds.  J. Biol. Chem. 127, 291-299.

Rittenberg, S. C., K. 0. Emery, and W. L. Orr.  1955.  Regeneration
       of nutrients in sediments of marine basins.  Deep-Sea Res. J3,
       23-45.

Rusness , B. and R. H. Burris.   1970.  Acetylene reduction (nitrogen
       fixation) in Wisconsin lakes.  Limnol. Oceanogr. 15, 808.

Russell, J. A.  1944.  The  colorimetric estimation of small amounts of
       ammonia by the phenol-hypochlorite reaction.  J. Biol. Chem.
       156, 457-461.

Sanderson, W. W.  1954.  Studies of the character and treatment of
       wastes from duck  farms,  Proc.  8th Ind. Waste Conf., Purdue
       Univ. Ext. Series. 83, (3B), 170.

Sawyer, C. N.  1947.  Fertilization of lakes by agricultural and
       urban drainage.   J.  New  England Water Works Assoc. 61, 109.

Sawyer, C. N. , J. B. Lackey,  and A. T. Lenz.  1945.  Investigation
       of  the odor nuisance occurring in the Madison Lakes, particu-
       larly Lakes Monona,  Waubesa, and Kegonsa from July 1943 to
       July  1944.  Report to  the Governor's Comm., State  of Wisconsin,
       Madison.

Schb'llhorn, R. and R. H. Burris.   1966.  Study of intermediates  in
       nitrogen fixation.   Federation Proc.  24, 710.

Schb'llhorn, R. and R. H. Burris.   1967.  Acetylene as  a  competitive
       inhibitor  of N.  fixation.   Proc. Nat. Acad. Sci.  U. S. 58,
       213-216.

Schmidt, E. L.  1954.  Nitrate  formation by  a soil fungus.   Science 119,
       187-189.

Schraufnagel, F.  H.  (Chairman), R.  3. Corey,  A. D. Hasler, G. F.  Lee,
       and T. L.  Wirth.   1967.   Excessive water  fertilization.   Rept.
       to  Water Subcommittee of Natural  Resources  Committee  of  State
       Agencies,  Madison, Wis.  (miraeo).

Shannon, E.  E.   1970.   Eutrophication-trophic  state  relationships in
       north-central  Florida lakes.   Ph.D.  thesis, Univ.  Florida,
       Gainesville,  257p.

Shannon, E.  E.  and  P.  L. Brezonik.   1971a.   Relationships between lake
        trophic  state  and nitrogen  and phosphorus  loading rates.   Sub-
       mitted to  Environ.  Sci.  Technol.
                                    151

-------
Shannon, E. E, and P. L.  Brezonik.  1971b.   Eutrophication:   cause-effect
       relationships in Florida lakes.   J.  Sanit. Eng.  Div.,  Amer.  Soc.
        Civil Eng. 9J7 (in press).

Shannon, E. E. and P. L.  Brezonik.  1971c.   Limnological characteristics
       of north and central Florida lakes.   Limnol. Oceanogr. 16 (in press)

Slack, K. V. and H. R. Felz.  1968.  Tree leaf control on low flow
       water quality in a small Virginia stream.  Envir. Sci. Technol.
       ,2, 126-130.

Stewart, B. A.  1970.  Volatilization and nitrification of nitrogen
       from urine under simulated cattle feedlot conditions.   Envir.
       Sci. Technol. _4, 579.

Stewart, W. D. P.  1963.   Liberation of extracellular nitrogen by two
       nitrogen-fixing blue-green algae.  Nature ZOO, 1020-1021.

Stewart, W. D. P.  1965.  Nitrogen turnover in marine and brackish
        environments   I.  Nitrogen fixation Ann. Botany, London 22,
        229-239.

Stewart, W. D. P.   1966.  Nitrogen fixation in  plants.  University
        London Press,  London.

Stewart, W. D. P.   1968.  Nitrogen input into  aquatic ecosystems.
        In Algae,  man and  the environment, D. Jackson  (ed.).  Syracuse
        Univ.  Press,  Syracuse, New York.

Stewart, W. D. P.   1969.  Biological and ecological aspects  of nitro-
        gen  fixation  by free living microorganisms.  Proc. Royal Soc.
        3172,  367-388.

Stewart, W. D. P.   1970.  Algal fixation of atmospheric nitrogen.
        Plant  and  Soil 32, 555-588.

Stewart, W. D. P., G. P.  Fitzgerald, and R. H.  Burris.  1967.   In situ
        studies on nitrogen  fixation using the  acetylene reduction
        technique.  Proc.  Nat. Acad. Sci., U. S.  58, 2071-2078.

Stratton,  F.  E.  and  P. L. McCarty.  1967.  Prediction of  nitrifica-
        tion effects  on the  dissolved oxygen balance of  streams. Envir.
        Sci. Technol.  1: 405-410.

Strickland, J. D.  H.   1965.  Production of organic matter in the
        primary stages of  the marine food chain.   In Chemical oceanography
        J.  P.  Riley and D.  Skirrow (eds.).  Academic Press, New  York.
                                    152

-------
Stumm, W.  1971.  The acceleration of the hydrogeochemical cycling
       of phosphorus.  Syinp, on changing chemistry in the sea,
       Goteborg, Sweden, Nobel Symp. Ser., Wiley, New York (in  press).

Sylvester, R. 0.  1961.  Nutrient content of drainage water from
       forested, urban and agricultural areas.  In Algae and metropo-
       litan wastes, U. S. Publ. Health Ser. Rept. W 61-3.

Syiaons, J. M. , S. R. Weibel, and G. G. Robeck.  1965.  Impoundment
       influences on water quality.  J. Amer. Water Works Assoc.  57,
       51-74.

Tang, Tsye-Lang.  1968.  Methane formation and associated nitrogen cycle
       reactions in sediments and sludges.  M.S. thesis, University
       of Florida, Gainesville.

Technicon Corp.  1969.  Ammonia in water and waste water.  Technicon
       Industr. Meth. Bull. 19-69w, Tarrytown, New York.

Thomann, R. V., D. J. O'Connor, and D. M. DiToro.  1971.  The effect
       of nitrification on the dissolved oxygen of streams and
       estuaries.  Envir. Eng. Sci. Prog. Manhattan College, Bronx,
       N. Y. (mimeo).

Toetz, D. W.  1970.  Experiments on the adsorption of ammonium ions
       by clay particles in natural waters.  Water Resources Res. 6_,
       979-980.

Toth, S. and A. Ott.  1970.  Characterization of bottom sediments:
       cation exchange capacity and exchangeable cation status.
       Envir. Sci. Technol. 4_, 935-939.

Triiper, H. G. and S. Genovese.  1968.  Characterization of photo-
       synthetic sulfur bacteria causing red water in Lake Faro.
       Limnol. Oceanogr. 13, 225-232.

Vollenweider, R. A.  1963.  Studi sulla situazione attuale del
       regime chimico e biologico del Lago  di Orta.  Mem. 1st.
       Ital. Idrobiol. 16, 21-125.

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

Waksman, S. A., M. Hotchkiss,  and L. C. Carey.   1933.   Bacteria
       concerned in  the cycle  of nitrogen in  the  sea.   Biol. Bull.
       65,  137-167.
                                  153

-------
Webb, K. L. and R. E. Johannes.  1967.  Studies of the release of
       dissolved  free amino acids by marine zooplankton.  Limnol.
       Oceanogr. _12, 376-382.

Weber, W.  J.  and  R. R. Rumer , Jr.  1965.  Intraparticle transport
       of  sulfonated alkybenzenes in a porous solid:  diffusion with
       nonlinear  adsorption.  Water Resources Res. ^L, 361-373.

Weibel,  S. R.  1969.  Urban drainage as a factor in eutrophication.
       In  Eutrophication:   causes, consequences, correctives.  Nat.
       Acad.  Sci., Wash.,  D.  C.

Weibel,  S. R.,  R. B. Weidner, A. G. Christiansen,  and R. G.  Anderson.
        1964.  Urban  land runoff  as a  factor in  stream pollution.
        J.  Wat.  Poll. Contr.  Ted.  36_,  914-24.

Weibel,  S. R.,  R. B. Weidner, A.  G. Christiansen,  and R. G.  Anderson.
        1966a.  Characterization,  treatment,  and disposal of  urban
        storm water.   Proc. Third Internat.  Conf. Water  Poll. Res.,
        Munich,  Germany, Paper 1-15.

 Weibel,  S. R.,  R. B.  Weidner, J.  M.  Cohen,  and A.  G.  Christiansen.
        1966b.  Pesticides and other contaminants in rainfall and
        runoff.   J.  Araer. Water Works Assoc.  58, 1075-84.

 Weichselbaum, T.  E., J. C. Hagerty,  and H.  B. Mark, Jr.   1969.
        A  reaction rate method for ammonia and blood urea nitrogen
        utilizing a pentacyanonitrosyloferrate catalyzed Berthelot
        reaction.   Anal. Chen. 41. 848-850.

 Wezernak,  C.  T. and J.  J. Gannon.  1969.   A field study of inorganic
        nitrogen oxidation rates.  Proc. 24th Industr.  Waste Conf.
        Purdue Univ., Eng. Ext. Ser.  No. 135., Lafayette, Indiana.
        pp. 53-68.

 Whittaker, R. H.  and P. P. Feeny.   1971  Allelochemics:  chemical
        interactions  between species.   Science 171, 757-770.

Wijler,  j.  and C.  C.  Delwiche.  1954.   Investigations on the denitri-
        fying process in soil.  Plant and Soil j^, 155-169.

 Wilson,  P. W.  1969.   First steps in nitrogen fixation.  Proc.  Royal
        Soc. B172, 319-325.

 Wuhnnann,  K.   1954.   High rate activated sludge treatment and its
        relation to  stream sanitation.  I.   Pilot plant studies.
        Sewage Industr.  Wastes 26. 1-27.

 Wuhnnann,  K.   1964.   Nitrogen removal in sewage treatment processes.
        Verh.  Internat.   Verein.  Linnol. 15. 580-596.
                                 154

-------
Yorton, 1.  1971.  The effects of amino acids and organic color on
     automated nutrient analyses of natural waters.  M.S. thesis,
     University of Florida, Gainesville.

Zilvermit, D. 3., C. Entenman, and M. C. Fishier.  1943.  On the
     calculation of "turnover time" and ''turnover rate'1 from
     experiments.  J. Gen. Physiol. 26, 325.

ZoBell, C. E.  1935.  The  assimilation of ammonium nitrogen by marine
     phytoplankton.  Proc. Nat. Acad. Sci. U. S. 21, 517.
                                    155

-------
                                 SECTION XII

                          PUBLICATIONS  AND PATENTS
The following publications  have  resulted  from work supported wholly or
in part by this project:

Brezonik, P. L.   1971.  Nitrogen:   Sources  and  transformations in natural
waters.  Presented  at  164th National  American Chemical Society Meeting,
Symposium on Nutrients  in Natural  Waters, Los Angeles, Calif., April,
1971, proceedings to be published  by  J. Wiley,  Inc.

Brezonik, P. L. and C.  L. Harper.   1969.  Nitrogen fixation in some
anoxic lacustrine environments.  Science  164, 1277-1279.

Brooks, ?,. ".., P. L. Brezonik, H.  D.  Putnam, and M. A. Keirn.  1971.
Nitrogen fixation in an estuarine  environment:  The Uaccasassa on the
Florida Gulf coast.  Limnol.  Oceanogr. I6_   (5), 701-710.

Keirn, M. A. and P. L.  Brezonik.   1971.  Nitrogen fixation by bacteris
in Lake Mize, Florida,  and  in some lacustrine sediments.  Limnol.
Oceanography 1_6_   (5),  720--731.

Keirn, M. A. and P. L.  Brezonik.   Significance  of nitrogen fixation in
the surface waters of eutrophic  lakes.  In  preparation for submission
to Ecology.

Shannon, II. L. and P. L.  Brezonik.  1972.  Relationships  between lake trophic
state and nitrogen and  phosphorus  loading rates.  Environ. Sci. Technol.
6_, 719-725.

Tang, Tsye-Lang.  1968.   Methane formation  and  associated nitrogen cycle
reactions in sediments  and  sludges.   M.S. thesis, University of Florida,
Gainesville.

Yorton, R. A.  1°71. Effects of  color and araino acids on automated
nutrient analyses.  M. 3.  thesis, University of  Florida, Gainesville.

Yorton, R. A. and P. L. Brezonik.   Interferences in the analyses of
inorganic nitrogen and  phosphate in natural waters and procedures for
correction.  "'anuscript in  preparation for  submission to J. Am. Wat.
Works Assoc.
                                    157

-------
                             SECTION XIII

                              APPENDIX A

                   SAMPLING AND ANALYTICAL METHODS
1.  Sampling Procedures.

Water samples for routine chemistry were taken by Van Dorn sampler;
bacteriological samples from the depths were taken aseptically in BOD
bottles attached to a special apparatus which opens the bottle from
the surface at the pull of a string.  Sediment cores 30 to 50  cm.  in length
were collected in plexiglass tubes.  Surface sediments were collected
with an Ekman dredge.

2.  Routine Chemical Analyses.

Acidity, alkalinity, and pH were run potentiometrically in the labora-
taory within a few hours after collection of samples in BOD bottles
to avoid contact with air.  Dissolved oxygen was determined by the
Winkler-azide method (A.P.M.A. 1965).  Cations were measured by atomic
absorption spectrophotometry.  Other determinations were done according
to Standard Methods  (A.P.H.A. 1965).

3.  Nutrient Analyses.

Inorganic nitrogen and  phosphorus  forms were determined with a Tech-
nicon Auto Analyzer  on  samples preserved with one ml saturated mercuric
chloride per liter of sample.  Ammonia was determined using modifica-
tions (Yorton, 1971) of the  phenol-hypochlorite method  (Technicon Corp.,
1969).  Nitrate was  determined by  the automated brucine technique
(Kahn and Brezenski, 1967),  and nitrite by adaption of  the Standard
Methods diazotization procedure to the Auto Analyzer.   Ortho phosphate
was measured by the  Murphy and Riley  (1962) single reagent method. Flow
diagrams and details of the  automated version of these  methods are pre-
sented by Yorton  (1971).  Total organic nitrogen was determined by the
macro-Kjeldahl method according to Standard Methods; recovered ammonia
was measured by titration or by the Auto Analyzer.  Particulate organic
nitrogen was measured by micro or  macro-Kjeldahl on solids scraped from
an MgCO~ mat placed  on  a 0.45 ym.  Millipore filter.  Total phosphate
analyses were performed by persulfate-sulfuric acid autoclaving for one
hour at 15 psi followed by manual  determination of ortho phosphate
by the Murphy and Riley (1962) method.

4.  Biological Methods.

The acetylene reduction technique  (Stewart et al., 1967) was used to
measure rates of  nitrogen fixation. The technique used on water
                                  159

-------
samples was described by Brezonik and Harper (1969) and on sediments
by Brooks e_t_ _al. (1971).  Sediment and water samples (20 to 25 ml) were
placed in 70 nl serum bottles, capped and purged with helium or, in
the case of waters containing dissolved oxygen, with a gas mixture
(20% 0?, 0.03% C02, balance Ar) .  Following this 5 cc. of gas phase was
remove? with a gas tight syringe and replaced with 5 cc purified acety-
lene  (Matheson Co.)-  After incubation (usually for 1-2 hours) activity
was stopped by adding 1 cc. 50% trichloroacetic acid with a needle and
syringe.  Gas separation and ethylene analysis was accomplished with a
Varian-Aerograph 600 D gas chromatograph with a hydrogen flame ioniza-
tion  detector and a 1/8" x 6'  Poropak R column at room temperature.
Newnan's Lake and Bivin's Arm  samples were incubated in situ.  Other
lake  and sediment samples were brought to the laboratory for processing
and incubation  at 22°C in a water bath-shaker.  Sediments and anoxic
water samples were incubated in the dark; aerobic samples were incubated
under daylight-type fluorescent lighting.  Controls run with each set
of samples were  carried through the identical procedure except that
1 ml  50% TCA was added before  acetylene addition.

Primary  production was  assayed by  11+C  techniques with in situ incubation
for 1 to  3 hours and  Geiger counting of the  1!*C incorporated  into par-
ticulate matter (filtered onto 0.45 vim. Millipore  filters).  Chloro-
phyll a was  measured  by standard  acetone  extraction and spectrophoto-
metry TCreitz  and Richards,  1955)  using Parsons and Strickland's  (1963)
equations  to  calculate  chlorophyll  concentration.

5.  Sediment  Analyses.

Volatile solids were  determined by  Standard  Methods  (A.P.H,A.,  1965).
Total organic  nitrogen  and  free ammonia  determinations were made  on
fresh-sediments diluted with  demineralized water  into a  thin  slurry.
Ammonia was  distilled from  the slurry  buffered  at  pH  8.3 with sodium
bicarbonate  and the  recovered  ammonia  measured  by  titration.  Total
organic nitrogen was  determined similarly by micro-Kjeldahl digestion.
Results were  converted  to a dry weight basis by drying and  weighing a
known volume  of the sediment  slurry.   Elemental analyses were accom-
plished with  a  Perkin-Elmer C-H-N  analyzer in oven (60°C) dried  sedi-
ment  samples which were pulverized  to  a homogenous mixture  in a  mortar
and pestle.  Total phosphate was  run on  fresh sediment slurries  using
persulfate oxidation  and  the Murphy  and Riley (1962) manual method  for
ortho phosphate.

References cited in this  appendix  are  listed in Section X.
                                 160

-------
                                      Appendix B




Routine Data Collected on Newnan's Lake and Bivin's Arm During Nitrogen Fixation Study




                                      Bivin's Arm
Parameter
Air Temp.
Surf, water
temp.
Cloud cover
Wind
Secchi Disc.
Dis. 07
L
surface
bottom
PH
Alkalinity
as CaC03
Tot. org. N
NH3-N
NC>2-N
N03-N
Ortho P04-P
Total PO,-P
Prim prod.
Nofixed (a)
Z.
(b)
21
May
29.0
27.3

20
0-2
0.6

10.5
8.4
8.9
97

2.7
0.70
0.008
0.03
0.5
1.0
256
0
0
6
June
25.7
28.2

100
5
0.3

13.0
12.5
9.5
88

3.3
0.11
0.003
0.00
0.042
0.37
1070
7.2
0.067
25
June
32.0
31.1

0
0-2
0.45

12. 1
12.0
9.2
95

4.0
0.80
0.001
0.00
0.063
0.56
790
114
1.08
9
July
__
—

90
0
0.3

10.5
10.0
9.5
113

5.4
0.0
0.001
0.00
0.091
0.58
2760
25
0.24
29
July
36.0
31.5

80
5
0.3

12.3
10.6
9.9
115

2.9
—
—
—
—
—
1480
46
0.43
7
Aug
38.4
31.3

80
0
0.6

13.3
9.4
9.5
145

2.30
0.21
—
0.01
0.3
0.61
117
109
1.02
21
Aug
27.6
27.6

5
10
0.6

12.6
12.1
9.5
100

3.0
0.48
0.000
0.00
0.25
0.49
1960
330
3.06
3
Sept
28.6
27.5

90
5
0.75

3.8
3.7
8.7
100

1.7
0.76
0.003
0.04
0.27
0.49
100
46
0.43
18
Sept
33.9
27.2

30
5
0.45

9.9
9.8
9.8
125

3.25
0.62
—
0.03
0.23
0.72
1580
0
0
2
Oct
25.3
25.9

100
5
0.9

4.6
4.1
8.8
96

1.1
0.38
0.006
0.21
0.19
0.54
166
0
0
16
Oct
32.0
29.0

5
0
0.45

15.0
14.4
9.6
105

2.4
0.35
0.000
0,00
0.13
0.54
1220
42
0.39
30
Oct
20.5
20.5

100
10-15
0.45

10.0
9.4
9.2
155

2.6
0.73
0.004
0.06
0.30
0.58
420
0
0
13
Nov
16.4
17.8

5
5
0.45

10.8
11.0
8.9
160

2.4
0.42
0.006
0.00
0.02
0.51
81
0
0

-------
                                              Bivin's Arm
Parameter*
Air temp.
Surf, water
temp.
Cloud cover
Wind
Secchi disc.
Dis . O^ surf.
bottom
pH
Alkalinity
Tot. org. N
NH3-N
N02-N
N03-N
Ortho PO^-P
Total P04-P
Prim. prod.
N2 fixed (a)
(b)
24
Nov
23.3
15.8

10
0-3
0.45
11.4
11.4
8.8
—
1.8
0.00
0.000
0.00
0.15
0.41
47
0
0
11
Dec
16.3
15.3

15
5-10
0.3
8.8
8.6
8.4
110
1.8
0.07
0.001
0.00
0.16
0.51
324
12.9
.12
2
Jan
10.5
12.8

100
10
0.45
8.4
8.1
9.0
100
2.0
0.23
0.000
0.03
0.055
0.21
236
15.7
.15
12
Jan
10
9.3

100
5
0.85
10.2
10.8
8.0
110
1.0
0.16
0.036
0.14
0.30
0.54
84
0
0
21
Jan
8.5
12.9

0
5
0.75
9.5
9.4
7.9
120
1.8
1.0
0.045
0.67
0.33
0.50
650
13.4
.12
4
Feb
10
13.0

0
5
0.45
8.8
8.7
8.3
80
1.2
0.08
0.000
0.14
0.031
0.63
168
0
0
18
Feb
15.5
16.1

95
0-5
0.6
11.8
10.9
8.9
100
1.7
0.00
0.005
0.03
0.43
0.62
25
17.7
.17
9
Mar
21
17.8

15
5
0.45
6.9
6.7
8.4
115
1.8
0.20
0.006
0.02
0.48
0.70
68
0
0
31
Mar
24.5
24.5

30
10
0.75
8.5
6.5
8.4
98
1.0
0.22
0.002
0.03
0.40
0.62
263
0
0
7
Apr
22.7
21.9

5
5
0.75
7.9
8.6
8.0
105
1.1
0.34
0.006
0.03
0.23
0.73
81
1.8
.016
28
Apr
27.5
26.8

75
0
0.55
12.2
8.7
8.9
120
1.7
0.13
0.000
0.02
0.43
0.59
345
3.6
.034
12
May
26.8
25.3

50
0-5
0.45
9.9
8.2
9.0
185
2.1
0.55
0.000
0.01
0.21
0.65
400
0
0
26
May
28.8
24.8

10
5
0.55
7.5
7.5
8.8
110
1.6
0.56
0.000
0.00
0.13
0.52
585
0
0
iTemperature °C; cloud cover in percent; wind in mph; Secchi disc in m.;  all concentrations in mg/1 of
species noted; primary production in mg C/m -hr. ; nitrogen fixation (a)  in nM ethylene/m^-hr;  and nitrogen
fixation (b) in ng N/l-hr.  assuming a theoreticl molar ratio of 1.5 moles ethylene produced per mole of
ammonia.

-------
                                                    Newnan's Lake
                                                         1969
01
OJ
Parameter
Air temp .
Water temp.
surf.
bottom
Cloud cover
Wind
Secchi Disc.
Dis Oo surf.
bottom
PH
Alkalinity
as CaC03
Acidity
as CaC03
Tot. org. N
NH3-N
NO^-N
NO^-N
Ortho PO,-P
Total P04-P
Prim. prod.
N2 fixed (a)
(b)
20
May
26.6

28.0
25.1
,
—
0.6
8.6
8.1
8.8
8.4

0

1.9
0.69
.006
0.16
0.012
0. 16
223
0
0
3
June
26.0

27.0
27.0
0
—
0.4
8.3
8.0
7.1
2.9

1.4

1.8
0.80
.006
0.22
0.028
0.15
15
0
0
24
June
32.7

31.7
28.9
40
0
0.45
8.7
8.0
8.6
6.8

0

1.5
0.07
.001
0.00
0.015
0.15
169
0
0
8
July
31.2

31.5
31.0
20
5-10
0.3
7.7
7.6
7.7
10

0

1.7
0.07
.002
0.00
0.09
0.25
229
0
0
24
July
32.5

30.0
29.7
80
10
0.45
8.0
8.3
8.0
12.5

0

2.1
—
—
—
—
—
331
76.5
0.71
6
Aug
36.4

29.1
27.9
20
5
0.6
9.1
6.5
9.1
8.5

0

1.2
0.03
—
0.02
0.009
0.10
247
0
0
22
Aug
31.1

30.8
30.5
30
—
0.45
7.9
7.9
7.5
7.5

—

1.8
0.18
.000
0.03
0.010
0.14
67
0
0
9
Sept
24.1

28.1
28.0
90-100
5-10
0.45
8.0
7.9
7.4
7.2

—

1.5
0.04
.000
0.11
0.004
0.09
23
0
0
17
Sept
37.1

32.3
26.5
50
0
1.05
7.8
7.2
7.1
7.2

1.9

1.3
0.02
.000
0.00
0.000
0.15
54
0
0
25
Sept
26.8

27.5
27.0
20
—
0.6
6.2
6.2
6.6
7.2

3.0

1.3
0.10
.000
0.01
0.000
0.14
86
0
0
14
Oct
32.0

28.5
26.0
40
—
0.6
8.0
6.3
7.2
7.7

2.1

1.3
0.29
.001
0.02
0.003
0.08
63
21
0.02
23
Oct
23.1

24.6
23.8
45
10
0.6
8.0
7.9
7.1
8.6

1.7

2.3
0.32
.003
0.06
0.010
0. 12
48
0
0
7
Nov
20.9

16.9
16.4
60
5
0.45
9.4
8.6
6.9
7.2

2.5

1.4
0.48
.003
0.16
0.012
0.10
39
0
0
         See footnote at end of table for parameter units.

-------
                                          Newnan's Lake
1969
Parameter
Air temp.
Water temp. surf.
" " bottom
Cloud cover
Light intensity
Wind
Secchi disc.
Dis. Oa surf.
" " bottom
pH
Alkalinity as CaCOs
Acidity as CaCOa
Tot. org. N
NH3-N
NOj-N
NO-3-N
Ortho P04-P
Total PO^-P
Chlorophyll a
Prim, production
N fixation (a)
(b)
20 4
Nov Dec
26.3 15.7
14.3 13.4
14.0 13.1
100 30
—
10 10
0.55 0.45
9.9 9.3
10.0 10.5
7.1 7.8
8.2 8.7
1.6 0.6
1.7 2.4
0.33 0.60
0.002 —
0.05 0.09
0.007 0.010
0.15 0.12
—
36 70
0 41
0 0.38
8
Dec
19.3
14.5
13.4
5
—
0-5
0.4
11.1
—
8.4
8.9
0
2.0
0.31
—
0.02
0.004
0.17
66.6
34
24.5
0.23
11
Dec
14.4
15.0
15.0
25
6080
5
0.25
9.7
—
7.0
8.7
2.3
2.1
0.06
0.000
0.06
0.007
0.14
67.9
119
20.3
0.19
16
Dec
15.7
13.7
13.5
40
5000
5
0.45
11.9
—
8.5
8.6
0
3.9
0.23
0.000
0.02
0.009
0.14
78
41
28.6
0.27
18
Dec
18.5
13.9
12.5
20
6400
0-5
0.5
—
— —
—
—
—
1.6
0.09
0.001
0.05
0.007
0.05
10.4
160
31.2
0.29
23
Dec
13.2
13.0
12.6
10
6400
0
0.45
11.4
—
7.8
—
—
2.0
0.19
—
0.14
0.013
0.15
—
77
11.4
0.11
26
Dec
13.0
13.8
13.8
0
7040
15
0.4
10.4
—
7.0
12
2
2.1
0.19
—
0.14
0.005
0.19
10.4
11.2
15.2
0.14
30
Dec
22.0
14.6
14.4
60
4160
10
0.25
—
—
—
—
2.2
2.2
0.18
—
0.05
0.012
0.19
10.9
35
22.1
0.21
1970
2
Jan
10.6
13.6
13.6
100
510
10
0.25
10.1
—
6.8
1.2
2.8
0.7
0.20
—
0.07
0.005
0.15
10.0
4.8
5.7
0.053
7
Jan
9.0
12.5
12.5
75
2720
20
0.3
10.1
_i_
6.7
1.5
4.6
0.7
0.02
0.000
0.05
0.005
0.21
16.5
15.5
6.5
0.06
13
Jan
7.0
8.2
8.1
100
640
10
0.4
11.4
—
6.6
6.1
3.2
1.8
0.10
0.003
0.05
0.007
0.14
73
43
2.7
0.025
20
Jan
11.9
12.8
12.5
75
720
0
0.4
12.4
—
6.9
0
2.2
2.2
—
0.000
0.04
0.013
0.050
60
32
0.5
0.004
footnote at end of table for parameter units.
                                                                                     Cont'd.

-------
                                                     Newnan's Lake

                                                          1970
Ui
Parameter
Air Temp.
Water temp. surf.
bottom
Cloud cover
Light intensity
Wind
Secchi Disc.
Ms Oy surf.
hot turn
PH
AlkaliniLy
Acidity as CaGO.,
Tot. org. N
NH3-N
NC>2~N
NO--N
Orf-ho P04-P
Total PO^-P
Chlorophyll a
Prim, production
N2 fixation (a)
(b)
30
Jan
11.0
13.7
13.4
70
2880
15
0.45
12.6
—
6.8
0
2.7
1.7
0.23
. 000
0.07
0.007
0.060
57
—
2.8
0.025
3
Feb
12.5
14.2
14.2
100
1030
15-20
0.45
8.5
—
7.0
1.6
2. 1
1.1
0.10
.000
0.05
0.090
0.11
67
29
—
""
16
Feb
18.9
13.9
12.1
100
1020
5
0.45
10.5
—
7.1
5.8
1.2
0.7
0.06
.000
0.02
0.008
0.09
49
—
14.3
0.13
27
Feb
16.4
16.4
13.0
0
3860
0
0.7
8.9
—
6.7
8.0
2.9
1.0
0.21
.001
0.05
0.011
0.09
78
85
17.9
0.17
4
Mar
19.2
17.5
16.3
60
4970
5
0.75
9.9
—
7.3
12.7
3.2
1.4
0.15
.000
0.05
0.006
—
58
117
23
0.21
11
Mar
22.4
19.0
18.5
40
8000
10
0.8
8.0
—
7.4
16.0
4.0
1.1
0.33
.006
0.02
0.012
0.08
62
281
39
0.36
20
Mar
28.5
18.3
17.2
50
800
10
0.8
10.2
—
6.7
0.4
2.9
1.2
0.06
.002
0.11
0.007
0.12
51
281
27.6
0.26
26
Mar
20.1
18.5
18.2
100
1280
0
0.75
10.2
—
7.2
6.3
0.9
1.5
0.08
.000
0.05
0.001
0.12
68
270
11.8
0.11
2
Apr
23.0
22.5
22.5
100
—
25
.35
7.5
7.5
6.9
4.2
1.8
1.3
0.28
.003
0.03
0.010
0.15
55
22
0
0
7
\pr
15.2
19.1
19.0
0
—
5
0.6
7.5
6.7
6.8
6.5
2.2
1.2
0.40
.003
0.04
0.017
0.13
87
197
3.2
0.03
24
Apr
30.4
27.0
25.4
95
6400
—
0.5
7.0
6.8
6.8
3.5
1.9
1.2
0.79
.008
0.03
0.042
0.064
68
29
1.5
0.014
6
May
27.1
26.5
25.5
50
7000
5
1.0
8.6
8.0
7.2
10
2.5
1.1
0.0
.000
0.03
0.008
0.12
43
129
1.6
0.015
22
May
31.8
27.0
26.0
40
6400
15
0.5
8.8
8.4
7.5
6.4
0.6
1.9
0.25
.001
0.08
0.011
0.056
80
185
0
0
2
June
52.1
28.2
28.0
100
5000
5
0.8
8.2
8.0
7.0
6.3
2.1
2.0
0.36
.003
0.02
0.011
0.063
101
65
0
0
         ^Temperature in °C; cloud cover in percent; wind in mph: Secchi disc in meters; chemical concentrations
         in mg/1 of species noted; chlorophyll a in mg/m  primary production in mg C/m -hr.; nitrogen fixation
         (a) in nM &2 H^/m -hr.; and nitrogen fixation  (b) in ng N/l-hr. assuming a theoretical producation
         ratio of 1.5 moles ethylene produced per mole  of ammonia fixed.

-------
                              APPENDIX C

       ENRICHMENT AND ISOLATION PROCEDURES FOR NITROGEN FIXING
                 AGENTS IN LAKE WATER AND SEDIMENTS
Water samples were collected from various depths in Lake Mize asepti-
cally by lowering a closed, autoclaved BOD bottle to the appropriate
depth and opening the stopper with a string.   Lake water samples  were
subjected to enrichment schemes for heterotrophic anaerobic,  faculta-
tive, or aerobic bacteria, photosynthetic bacteria, and yeasts and
fungi.  In addition, samples were examined microscopically for blue-
green algae and photosynthetic bacteria but visual examination of all
samples yielded negative results.

Heterotrophic nitrogen fixing bacteria, both aerobic and anaerobic,
were enriched intially using a nitrogen free, modified Winogradsky's
medium  (Grau and Wilson, 1962) with sucrose as the carbon source.
Aliquots from each depth were mixed with equal amounts of double strength
medium  and incubated at 20°C both aercbically and anaerobically under
one atm No.  Control aliquots were carried through the same procedure
with the exception that 1.0 g/1 NH.C1 was present in this medium.  After
growth  was noted by increased turbidity, transfers were made to fresh
nitrogen free media.  After three transfers streak plates were made on
to the  same medium plus 1.5% agar.

Enrichment for photosynthetic bacteria was done with three media.
Athiorhodaceae (non-sulfur purple bacteria) were enriched by placing
100 ml  of lake water from each depth into A BOD bottle and filling with
a medium consisting of sodium acetate (2 g/1), NH,CL (1 g/1), l^HPO^
(0.5 g/1), MgCl2 (0.1 g/1), yeast extract (0.05 g/1), and adjusted to
pH 7.0.  The bottles were stoppered and incubated at 25°C under 250
foot candles of continuous fluorescent light.  Enrichment for Thiorho-
daceae  (green and purple sulfur bacteria) was accomplished by inocu-
lating  200 ml of lake water into each of two BOD bottles containing
100 ml  of a sterile slurry of cellulose  (3 g.), CaSO^ (3 g.)  , NH^,C1
(0.139  g.), KH2P04 (0.33 g.), and Ha2S•7H20 (0.07 g.), one adjusted to
pH 7.3  (for green sulfur bacteria) and the other adjusted to pH 8.5
(for purple sulfur bacteria).  The bottles were capped and incubated as
described above for the Athiorhodaceae.  Controls, consisting of sedi-
ment samples from a lake known to contain these groups (Lake Alice,
Gainesville, Florida (Lackey et_ al. , 1964), were carried through the
procedure to assure that the scheme was valid.

Enrichment for nitrogen fixing yeasts of fungi was made by filtering
aliquots of water through 0.45 urn. membrane filters which were incu-
bated at 25°C and 35°C on Sabouraud dextrose  agar.  All morphological
types  (about 30% of the total colonies or 65  isolates) were  picked
                                   166

-------
after 4 days growth,  gram stained  and  inoculated  into  a nitrogen de-
ficient medium modified  from Metcalfe  and Brown  (1957), containing
0.3% sodium benzoate;  glucose,  sucrose,  and  mannitol glucose as car-
bon sources at 5.0  g/1;  phosphate,  and trace metals and buffered to
pH 7.2.  The isolates  were also inoculated into  a control medium (i.e.
the above medium plus  nitrogen  as NH/C1).  All cultures were incubated
in the dark at 25°C.

Anaerobic bacteria  such  as occur in the  genus _Clp_st_ri_dium were consi-
dered the most likely  organisms responsible  for  fixation in the anoxic
estuarine sediments,  and the following enrichment and  isolation scheme
for clostridia-like nitrogen-fixing organisms was used to determine
whether such organisms occur in Waccasassa Estuary sediments.  A sedi-
ment sample from the  2-5 cm.  depth  was heated at  80°C  for 10 minutes
to destroy vegetative  forms and then inoculated  into a nitrogen-free
salt medium containing 4 g/1 of one of the following carbon sources:
acetate, maltose, mannitol, sorbitol or  sucrose.   All  of these except
acetate have been reported as fermentable carbon  sources for Clostridium
pasteurianum.  After  incubation at  room  temperature (22-25°C) under
a pure nitrogen atmosphere for  three days, growth was  apparent (by in-
creased turbidity and  gas production)  in all samples except the medium
containing acetate  as  carbon source.   Most rapid  growth occurred ini-
tially with mannitol;  however in subsequent  transfers  sorbitol gave
quickest response.  A  test for  acetylene reduction activity in the
mannitol culture gave  positive  results.   Gram studies  showed an abun-
dance of gram-positive rods accompanied  by a few  gram-variable coccoid
forms.  Many of the rods showed sub-terminal swellings characteristic
of developing spores.

The enrichment cultures  were incubated five  more  days  to promote
sporulation, then heated to 80°C for 10  minutes  and transferred to
basal salts media containing sorbitol.  Transfers were incubated under
an Np atmosphere at room temperature for 48  hours, and another transfer
was made to similar media.  Following  incubation  for 5 days, a third
transfer was made.  Micro-Kjeldahl  analysis  of each batch of fresh
media showed no detectable ammonia  or  organic M.   A test for acetylene
reduction on the 8  bottles (duplicates of the 4  original carbon sources
showing growth) from  the third  transfer  gave positive  results for 3
bottles.

Streak plates were  made  onto the nitrogen-free basal salts medium plus
sorbitol in 2 percent  agar from the incubated material of the third
transfer, and the plates were incubated  for  72 hours in an anerobic
jar (N^ atmosphere).   Colonies  on  the  plates were all  similar in
appearance and did  not display  chromogenesis.  Ten colonies which were
gram-positive rods  were  picked  from plates streaked with media from
bottles which exhibited  acetylene  reduction  and were transferred to
basal salts plus sorbitol.  Each isolate was grown under both aerobic
and anaerobic (^ atmosphere) conditions.  Four  colonies grew only
anaerobically; the  remainder grew  under  both conditions.  The latter
were tested for acetylene reduction and  gave negative  results.  The
four obligate anaerobic  isolates were  also tested for  acetylene reduc-
tion, and three of  the isolates showed positive  rates.
fiU.S. GOVERNMENT PRINTING OFFICE: 1973 546-303/35 1-3

-------
 SELECTED WATER
 RESOURCES ABSTRACTS
 INPUT TRANSACTION FORM
                     1. Report No.
                                2.
                                                                   3. Accession No.
                                         w
 4.  Title
       NITROGEN SOURCES AND  CYCLING IN NATURAL WATERS
  7. AuthoT(s)
       Patrick L.  Brezonik
  9. Organization
       University of Florida
       Department of Environmental Engineering
       Gainesville, Florida

 12. Sponsoring Organization

 35. Supplementary Notes

      Environmental  Protection Agency  report number,
      EPA-660/3-73-002, July 1973.
                                         5. Repoit Date

                                         6.
                                         8, Performing Organization
                                           Report No.

                                         JO. Project No.
                                            16010 DCK
                                         11.  Contract/Grant No.
                                         13.  Type of Report and
                                            Period Covered
 16. Abstract   Sources of nitrogen were reviewed  to  determine their significance in lake
 nitrogen budgets.  Nutrients in rainfall were evaluated and found to be significant.
 Nitrogen and phosphorus budgets were calculated  for  55  Florida lakes and critical load-
 ing  rates established by  comparing calculated budgets with data on trophic state.
             Nitrogen  fixation by Cyanophyceae  was studied in detail in two eutrophic
 Florida lakes for one year.   Also a survey  of fixation  in Florida lakes was conducted
 and  fixation found only  in eutrophic lakes.   Bacterial  fixation was found to contribute
 significant nitrogen to  the anoxic hypolimnon of a small stratified lake.  Nitrogen
 fixation was found in both lacustrine and estuarine  sediments.
             Sediments of  55 lakes were characterized  chemically and results suggest that
 such sediment may act as  an ammonia buffer, sorbing  ammonia at high concentrations and
 releasing it to ammonia  depleted water.  Estuarine sediment sorbed ammonia strongly
 but  failed to release it  to overlying water.
             The acetylene  reduction assay for  nitrogen  fixation was evaluated.  Inter-
 ferences in automated nutrient determinations due to organic color were studied and
 simple color correction  found for nitrite,  nitrate and  orthophosphate but not  ammonia
 as  determined by the indophenol method.  Amino acids also interfered with the  ammonia
 analysis. (Keirn-Florida)
  17a. Descriptors
  17b. Identifiers
              *Eutrophication, *Limnology,  *Nutrients, *Nitrogen Fixation
               Water  Quality, Trophic Level,  Blue-Green Algae, Bacteria
               Florida Lakes, Waccassa Estuary
  17c. COWRR Field & Croup
  18.  Availability
19. Security Class.
   (Report)
                          20. Security Class.
                             (Page)
21. No. of
   Pages

22. Price
                                                        Send To:
                             WATER RESOURCES SCIENTIFIC INFORMATION CFNTFR
                             US DEPARTMENT OF THE INTERIOR
                             WASHINGTON. D. C 20210
  Abstractor  Michael Keirn
                                       \Institution   University of Florida
WRSIC 102 (REV. JUNE 1971)

-------