EPA-600/3-76-094a
October 1976
Ecological Research Series
ALGAL NUTRIENT AVAILABILITY AND
LIMITATION IN LAKE ONTARIO DURING IFYGL
Parti
Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Duluth, Minnesota 55804
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
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.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/3-76-094a
October 1976
ALGAL NUTRIENT AVAILABILITY AND LIMITATION
IN LAKE ONTARIO DURING IFYGL
Part I. Available Phosphorus in Urban Runoff
and Lake Ontario Tributary Waters
by
William F. Cowen
and
G. Fred Lee
University of Texas at Dallas
Richardson, Texas 75080
Contract No. R-800537-02
Project Officer
Nelson Thomas
Large Lakes Research Station
Environmental Research Laboratory-Duluth
Grosse He, MI 48138
ENVIRONMENTAL RESEARCH LABORATORY - DULUTH
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
DULUTH, MINNESOTA 55804
-------
DISCLAIMER
This report has been reviewed by the Environmental Research Laboratory,
Duluth, Minnesota, U.S. Environmental Protection Agency, and approved for
publication. Approval does not signify that the contents necessarily
reflect the views and policies of the U.S. Environmental Protection Agency,
nor does mention or trade names or commercial products constitute endorse-
ment or recommendation for use.
11
-------
FOREWORD
Our nation's freshwaters are vital for all animals and plants, yet our
diverse uses of water for recreation, food, energy, transportation, and
industry physically and chemically alter lakes, rivers, and streams. Such
alterations threaten terrestrial organisms, as well as those living in water.
The Environmental Research Laboratory in Duluth, Minnesota develops methods,
conducts laboratory and field studies, and extrapolates research findings.
to determine how physical and chemical pollution affects aquatic
life
to assess the effects of ecosystems on pollutants
to predict effects of pollutants on large lakes through use of
models
to measure bioaccumulation of pollutants in aquatic organisms that
are consumed by other animals, including man
This project was conducted as part of the International Field Year for
Great Lakes Research and consisted of three separate parts, all directed
toward providing information needed to assess the factors limiting algal
growth in Lake Ontario and the amounts of nitrogen and phosphorus in tributary
drainage which would likely become available in the lake. Part I is concerned
with a comprehensive study of the amounts of phosphorus entering Lake Ontario
from U.S. tributaries which will likely become available in the lake. Parti-
cular attention is given to the particulate and organic forms of phosphorus
in the major U.S. tributaries to the lake. Part II is concerned with a study
of the amounts of available nitrogen entering Lake Ontario from the U.S.
tributaries. Part III is concerned with the factors limiting algal growth
in Lake Ontario and in the major U.S. tributaries. This report presents
Part I of this study. Parts II and III are published as separate reports by
the Environmental Protection Agency under the following titles:
Part II: Nitrogen Available in Lake Ontario Tributary Water
Samples and Urban Runoff from Madison, Wisconsin
Part III: Algal Nutrient Limitation in Lake Ontario during IFYGL
Donald I. Mount, Ph.D.
Director
Environmental Research Laboratory
Duluth, Minnesota
iii
-------
ABSTRACT
Urban runoff, precipitation, and river samples from Madison, Wisconsin
and New York State were analyzed for various phosphorus forms and the por-
tion of each phosphorus form available for algal use. Total phosphorus,
soluble phosphorus, particulate phosphorus, and soluble reactive forms were
measured. In addition, acid extractable, base extractable, and anion ex-
change resin extractable inorganic phosphorus was determined on the part-
iculate fractions.
Algal assay procedures were used to assess portions of the various
phosphorus fractions available for Selenastrum capricornutum growth. Avail-
ability of particulate phosphorus in urban runoff from Madison, Wisconsin
was highly variable ranging from 8 to 55 percent. Genessee River basin
urban runoff had from less than 1 to 24 percent of its particulate phosp-
horus available. Particulate phosphorus from the Niagra, Genesee, Oswego,
and Black Rivers showed only 6 percent or less available to this alga.
Autoclaving the samples increased the amount of particulate phosphorus
available. Precipitation samples usually showed less than 9 percent of the
total phosphorus available to Selenastrum capricornutum. Total phosphorus
available for algal growth from New York tributaries was highly variable.
About 39, 24, and 15 percent of particulate phosphorus in urban run-
off from Madison could be extracted by acid, base, and anion exchange. Re-
sults from urban areas in the Genesee River basin in New York were similar.
Resin extractions in long-term aerobic dark incubations produced results
similar to short-term tests, indicating that physical and chemical rather
than microbial mineralization processes were probably the ket factors regul-
ating the release of inorganic P from the runoff particules to the solution
phase.
This report was submitted in fulfillment of Contract No. R-800537-02
under the sponsorship of the Environmental Protection Agency. Work was
completed as of June, 1975.
IV
-------
CONTENTS
PART I
Sections
I
II
III
IV
V
VI
VII
VIII
IX
X
Introduction
Conclusions
Recommendat ions
Literature Review
Sampling of Urban Runoff and Lake Ontario
Tributaries
Analytical Methods
Results
Discussion
References
Appendices
Page
1
3
8
10
47
60
117
165
208
217
v
-------
PART I
FIGURES
No. Page
1 Lake Ontario and Major Tributaries 34
2 Urban Runoff Sampling Stations in Madison, 50
Wisconsin
3 Column for Solution of Phosphate from Dowex 78
1-X8 Resin
4 Recovery of Added Phosphate from Dowex 1-X8 79
Resin by IN Na2SO. Leaching
5 Absorbance of Selenastrum Cultures in AAP 88
Medium vs. Time
6 Standard Curve for Cell Counts of S. capri- 89
cornutum after 18 Days vs. Orthophosphate
Concentration (0-600 ygP/1)
7 Absorbance vs. Dry Weight of S. capricor- 94
nutum
8 Standard Curve for Absorbance of S. capri- 95
cornutum after 18 Days vs. Orthophosphate
Concentration (0-600 ygP/1)
9 Hypothetical Case for Underestimation of 9g
Available P in a Filtered Sample Bioassay
10 Standard Curves for Absorbance of S.capri- 99
cornutum vs. Orthophosphate Concentration
in AAP Medium or in Spikes Added to Sample
No. 52 (Oswego R.) Water (Autoclaved, Fil-
tered)
.11 Correlation between Absorbance and Cell 101
Counts for S. capricornutum in AAP Medium
12 Absorbance vs. Cell Counts for S. capri- 102
cornutum Grown in Autoclaved, Filtered River
Waters Supplemented with IX AAP(-P) Medium
vi
-------
FIGURES (continued)
No.
13 Standard Curves for S. capricornutum Cultured 108
in 50 ml and 125 ml flasks
14 Procedures in Handling and Analysis 116
of Samples
15 Percent of PP Extracted from Madison 119
Urban Runoff Particles by Chemical Methods
16 Bioassay of Soluble P in Madison Urban 134
Runoff Sample B-9
17 Percent of PP Extracted from Genesee R. 141
Basin Particles by Chemical Methods
18 Comparison of Chemical and Biological 173
Extraction of PP (Madison Urban Runoff
Samples)
19 Comparison of Chemical and Biological 185
Extraction of PP (Genesee R. Basin
Samples)
20-23 Comparison of Chemical and Biological 201
Tests for Determining the Availability
of TP in New York River Waters
20 Niagara River 202
21 Genesee River 203
22 Oswego River 204
23 Black River 205
VII
-------
PART I
TABLES
No. Page
1 Composition of Chicago Street Litter 16
2 Dust and Dirt Loads in Chicago 16
3 Estimated Percent of Total P as "Soluble 31
Available P" in Urban Runoff
4 Phosphorus Yields in Urban Stormwater 31
Runoff
5 Major Sources of Phosphorus to Lake 33
Ontario (1966-1967)
6 Sources of Phosphorus to the Niagara River 35
7 Madison Stormwater Runoff Record 48
8 Madison Urban Runoff Sampling Stations , 49
9 Description of the Urban Areas Sampled for 51
Runoff Phosphorus
10 Location of the EPA New York Rain Gages 55
11 Location of Runoff Sampling Stations in the 56
Genesee River Basin, N.Y.
12 Land Use Distribution in Sub-Basins of the * 57
Genesee River Basin, N.Y.
13 Effect of Filter Type on the DRP Analysis 64
of New York River Waters
14 Precision of DRP Analyses Using Marked Test 66
Tubes
15 Efficiency of Scraping Particulate Matter 70
from Membrane Filters
Vlll
-------
TABLES (continued)
No. Page
16 Efficiency of the Persulfate Digestion Method 72
as Compared to a Perchloric Acid Method
17 Estimated Precision of Chemical Analyses for 74
Soluble and Particulate Phosphorus Forms
18 Recovery of Added Orthophosphate Spikes in 74
DRP and TSP Analyses
19 Recovery of Added Orthophosphate from Dowex QQ
1-X8 Anion-Exchange Resin by IN Na^SO^
Leaching
20 Recovery of Orthophosphate Spikes Added to 82
Acid, Base, and Resin Extraction Solutions
before Equilibration with Sample Particles
21 Composition of Standard AAP Algal Medium 85
22 Recovery of Orthophosphate Spikes, Added to 92
AAP(-P) Medium Containing Particulate Matter,
in Selenastrum Growth Bioassays
23 Comparison of TSP in Culture Flasks and Bio- 104
assay Results, as Computed from Direct Cell
Counts and from Uncorrected A7I-n Data
24 Recovery of Orthophosphate Spikes Added to 105
Autoclaved, Filtered Oswego R. Sample No. 52,
as Calculated Using Corrected Net A^rr, Bio-
TN J_ / 0 U
assay Data
25 Recovery of Orthophosphate Spikes Added to 106
Autoclaved, Filtered River Waters, as Cal-
culated Using Corrected Net A?5Q Bioassay
Data
26 Hydrolysis of Sodium Tripoly Phosphate (TPP) 112
by Murphy-Riley Color Reagent at 27°C
27 Hydrolysis of 200 ygP/1 TPP in Chloroformed 112
Distilled Water at 20°C
28 Dark Incubation of Standard 200 ygP/1 Ortho- 114
phosphate Solutions with One Gram of Dowex
1-X8 Resin at 20°C
29 Dark Incubation of Chloroformed Standard 114
100 ygP/1 Orthophosphate Solutions at 20°C
IX
-------
TABLES (continued)
No. Page
30 Effect of Storage Time in One-Gallon Cubi-
tainers on the DRP Concentrations of Dis-
tilled Water Solutions
31 Phosphorus Forms in Madison Urban Runoff 12°
Samples
32 Extraction of Madison Urban Runoff Particles j_2i
with Acid, Base, and Anion-Exchange Resin
33 Extraction of Madison Urban Runoff Particles 124
by Selenastrum in Algal Bioassays
34 DRP Changes During Dark Incubation of Runoff 126
Particles in Lake Water
35 Summary of Net Mean DRP Released from Runoff 127
Particles to Lake Water
36 Recovery of Inorganic P from Spiked Urban I29
Runoff Sample D-ll by Anion-Exchange Resin
Extraction
37 Maximum DRP Values in Test Flasks with Anion- 13°
Exchange Resin During Dark Incubations of
Madison Urban Runoff
38 Bioassay of Filtered Urban Runoff Samples 132
with Selenastrum
39 Phosphorus Forms and Algal-Available PP in 135
Madison Snow Samples Collected April 10, 1973
40 Phosphorus Forms in New York Rain Gage Samples I37
41 Bioassay of Unfiltered New York Rain Gage j_3g
Samples with Selenastrum
42 Phosphorus Forms in Genesee River Basin ^-3^
Samples
43 Extraction of Genesee R. Basin Sample Par- 142
tides with Acid, Base, and Anion-Exchange
Resin
x
-------
TABLES (continued)
No. Page
44 Extraction of Genesee R. Basin Sample Par- I1*3
tides by Selenastrum in Algal Bioassays
45 Maximum DRP Values in Test Flasks with
Anion-Exchange Resin During Dark Incubations
of Genesee R. Basin Samples
4-6 Phosphorus Forms in New York River Water
Samples and Lake Ontario River Mouth Samples
47 Extraction of Genesee River Particles with
Acid, Base, and Anion-Exchange Resin
4-8 DRP Changes During Dark Incubation of Genesee
R. Sample No. 42 Particles in Lake Ontario
Water
49 Extraction of New York River Water Particles
by Selenastrum in Algal Bioassays
50 DRP Changes in a Chloroformed Oswego R. Sample
(No. 55) ,with and without added Sodium Tripoly
Phosphate (TPP)
51 DRP Changes in Chloroformed River Waters, with
and without Added Sodium Tripoly Phosphate
(TPP)
52 Maximum DRP Values in Chloroformed New York 155
River Waters after 1 to 16 Days of Dark Incu-
bation
53 Changes in the DRP Fraction of TP as a Result I57
of Cold Storage Followed by Dark Incubation
with Chloroform
54 DRP Changes During Dark Incubation of Equal 158
Volumes of Genesee R. (No. 16) and Genesee R.
Mouth (No. 17) Water Samples
55 Maximum Percent of TP Observed as DRP During 159
Dark Incubations of New York River Water
Samples
56 Effect of Autoclave Treatment on the Soluble
P Forms in River Waters
XI
-------
TABLES (continued)
No. Page
57 Comparison of Soluble P and Algal-Available 162
P in Autoclaved, Filtered River Waters
58 Comparison of TSP and Algal-Available P, 164
Calculated Using Corrected A75Q Data in
Autoclaved, Filtered River Waters
59 Comparison of Resin Extractable PP^ for 170
Direct Short-Term Extraction of PP and for
Long-Term Dark Incubations of Unfiltered
Runoff
60 Comparison of Chemically and Biologically 182
Determined Phosphorus Forms in New York
Rain Gage Samples
61 Estimated Percent of PP Available, As Cal- 190
culated from Treatments of Unfiltered River
Water Samples
62 Comparison between DRP, TSP, and Algal- 195
Available P in Autoclaved, Filtered New York
River Waters
63 Percent of River Water PP Available to Sele- 196
nastrum in Direct Bioassays of Autoclaved PP,
As Compared to Calculated PP Availability in
Bioassays of Autoclaved, Filtered River Water
64 Comparison of Chemical and Biological Tests 199
for Determining the Availability of TP in
New York River Waters
xii
-------
ACKNOWLEDGMENTS
This investigation was conducted primarily at the
University of Wisconsin, Madison. In addition to support
given for this investigation by the U.S. Environmental
Protection Agency, support was also given by the Depart-
ment of Civil and Environmental Engineering at the Univer-
sity of Wisconsin and the Institute for Environmental
Sciences at the University of Texas at Dallas.
Collection of samples from the Oswego River was
performed under the direction of Richard B. Moore, of the
Lake Ontario Environmental Laboratory, State University
College, Oswego, N.Y. Collection of rain gage water was
performed by the staff of the U.S. EPA laboratory,
Rochester, N.Y., under the direction of Donald Casey -
Genesee River basin samples were sent to Madison by the
personnel of the New York State Department of Environmental
Conservation, Albany, N.Y., under the supervision of
Patricia Boulton. Thanks are also due to Nelson Thomas,
project officer for the U.S. EPA, for his assistance in
sample procurement.
The aid of Ann Matlack, Scott Ramos, Mary Jane, Obert,
and Eunice Adamany in the laboratory phase of this research
is greatly appreciated. Ms. Adamany also assisted in the
compilation of data and typing of many bioassay data
tables.
This report is essentially the same as the Ph.D,
thesis of William F. Cowen for the University of Wisconsin.
Xlll
-------
SECTION I
INTRODUCTION
The continuing trend toward urbanization in the United
States will pose at least a dual threat to the quality of
surface water resources with respect to nutrient inputs.
The first threat will come from point sources of plant
nutrients, such as sewage treatment plant effluents or indus-
trial wastes. The second threat will come in the form of
diffuse sources of nutrients, such as storm water runoff.
Both threats will contribute to the fertilization, or
eutrophication, of surface waters.
A logical approach to the control of eutrophication
will require that sources of nutrients be evaluated in
terms of quantity and quality. Since phosphorus has been
implicated as a probable limiting nutrient in many waters,
the loads of total phosphorus discharged by urban runoff
have been measured by several workers. These studies have
shown that large amounts of phosphorus are present in urban
storm water drainage. A few of these studies have included
measurements of "dissolved reactive" or "soluble" phosphorus
in an attempt to define the fraction of total phosphorus
which might be available for algal growth in the receiving
water. However, there is still little information on the
availability of phosphorus in the insoluble forms found in
urban runoff. The purpose of the studies reported here was
to investigate the availability of phosphorus forms in urban
runoff and to apply the methods developed for urban systems
-------
to the study of phosphorus availability in tributary
waters to Lake Ontario as part of the International Field
Year for the Great Lakes.
-------
SECTION II
CONCLUSIONS
1. Particulate phosphorus (PP) of 0.45 micron size or
larger, isolated from Madison urban runoff, showed group
mean values of acid extractable inorganic P from PP
(extractable PP.) which ranged from 33 to 46 percent of
PP. The corresponding range for sodium hydroxide extract-
able PP.^ was 22 to 27 percent of PP, and for anion-
exchange resin extractable PP. the range was 13 to 17 per-
cent of PP.
2. Since the group mean values of the chemical extrac-
tions represented the chemical nature of PP from the
various land uses sampled for urban runoff, the relatively
narrow ranges of group mean values for a given type of
chemical extraction indicated that the PP forms trans-
ported by surface runoff from different land uses in
Madison were similar. Possibly, the predominant type of
PP in the urban runoff samples was derived from a common
source, such as dustfall or eroded soil.
3. Dark incubations of runoff PP in lake waters or dark
incubations of unfiltered runoff itself with added anion-
exchange resin indicated that physical-chemical processes
such as desorption or dissolution were more important
factors in the release of inorganic P to solution than was
microbial mineralization of PP. This conclusion was the
-------
result of the fairly close agreement between resin
extractable PP. in relatively long-term as compared to
short-term aqueous incubation systems containing PP and
resin.
4. Bioassays of PP from Madison runoff generally showed
availability values which were intermediate between the
acid and resin extractable PP. values for a given sample.
An overall bioassay range of 8 to 55 percent of PP was
found in 13 samples, 10 of which were derived from runoff
draining residentially zoned areas of Madison. The over-
all average for all samples tested was 30 percent of PP
available to Selenastrum capricornutum in 18 days.
5. Because the resin and bioassay tests probably repre-
sented the closest approximations to the true availability
of PP in the receiving water, their mean values of 15 and
30 percent, respectively, represent a reasonable estimate
of the availability of PP in the Madison samples. This
range corresponds also to the range of group means
reported from base extractions, 22 to 27 percent of PP.
Because of particle settling and possibly poor mixing of
the runoff particles in the receiving water, these values
should be regarded as upper bounds for the availability
of PP forms in the receiving water.
6. Soluble P forms in urban runoff, as defined by 0.45
micron filtration, appeared to be subject to overestima-
tion of their algal-available fraction of total soluble
phosphorus (TSP) in the Selenastrum bioassays, possibly
due to the inherent problems related to the construction
of a valid standard curve. However, some of the samples
showed less algal-available P than even the chemically
measured dissolved reactive phosphorus (DRP) in the runoff
filtrates. In these samples, the presence of colloidal,
acid-soluble P forms could have inflated the chemical
-------
results from DRP analysis. Care should be exercised in
the interpretation of chemical analyses of apparently
soluble P forms in runoff, unless the results have been
checked with bioassay data.
7. Bioassay of three samples of PP from Madison snow
showed PP availability in the range of <2 to 23 percent
of PP. The lowest availability value was found in the PP
from snow collected in the commercial district of central
Madison, where the automobile traffic was very heavy.
Consequently, the presence of heavy metal (i.e. lead)
toxicity in the algal assay was considered a possible
explanation for the lack of growth in this sample.
8. Some of the rain gage samples from the state of
New York contained large quantities of phosphorus forms
in the <0.45 micrometer size fraction which did not react
with the color reagent for DRP nor were used by Selenastrum
for growth. In terms of total P availability in the rain
waters, only 3 of the 13 samples bioassayed showed 10 per-
cent or more of the TP to be available to Selenastrum.
Accurate bioassays of TP were not possible for five of the
rain samples because of their low TP levels.
9. Urban runoff samples from two stations in the Genesee
River basin in New York showed group mean values of acid,
base, and resin extractable PP. as follows: acid, 30 to
48 percent of PP; base, 18 to 30 percent, and resin, 11
to 25 percent. These values were close to the ranges
reported for Madison urban runoff PP samples.
10. The bioassay of PP forms from the Genesee R. basin
samples generally showed values which were less than or
equal to the resin extraction values, unless the particles
were autoclaved before bioassay. Based on resin extraction
and bioassay data averages from all samples, probably 16
-------
percent of the PP or less would be expected to become
available in the receiving water. The base extraction
data predicted a value of about 22 percent of PP-
11. Samples of runoff particles from cropland, pasture,
and brushland did not appear to differ significantly from
runoff particles from one of the urban stations (Dansville,
No. 7), in terms of the PP. extracted by acid, base, or
resin. Particles from the other urban station (Rochester
East, No. 2) appeared to show higher percentages of chem-
ically extracted PP. than did the particles from all other
stations.
12. Chemical extractions of Genesee R. particles showed
variable fractions of extractable PP.. However, algal
bioassays of these particles indicated that less than 6
percent of the PP was available to Selenastrum in 18 days.
Autoclaving of the particles resulted in significant
increases in the fraction of PP available to Selenastrum.
13. None of the PP samples from New York tributaries to
Lake Ontario tested by algal growth assays showed more than
6 percent of PP to be available for algal growth on a
short-term basis (assay of natural particles). On a long-
term basis (assay of autoclaved particles), the assays
showed that perhaps 26 to 57 percent of PP may become
available in Lake Ontario after death and autolysis of the
native organisms associated with the PP in the rivers.
14. Chloroform treatment of unfiltered river waters demon-
strated the release of DRP from PP forms, in most of the
samples tested, indicating the presence of algae and/or
zooplankton in the sample PP. The maximum DRP contribution
from PP as a result of chloroforming was estimated to repre-
sent from 0 to 86 percent of PP in the Niagara R. samples,
3 to 18 percent of PP in the Genesee R. samples, 0 to 60
percent of PP in the Oswego R. samples, and 10 to 20
percent of PP in the Black R. samples. Dark incubations
-------
of unfiltered river water generally showed lower values of
DRP contributed by PP than were seen in the chloroform
tests, although there were some exceptions to this rule.
15. Because of the possible resorption of DRP to particu-
late matter after autoclaving, the direct algal assay of
autoclaved PP forms in the Algal Assay Procedure - P
medium appeared to give a more realistic estimate of the
expected PP availability than did the autoclave-filtration
bioassay on whole water samples, where the contribution
of PP was calculated from the bioassay of the autoclaved
sample filtrate and the initial TSP value of the sample.
16. Total P availability showed a wide range of values in
the Niagara and Genesee R. samples collected during the
study because of changing proportions of PP and TSP in the
samples. In the case of the Oswego and Black River samples,
however, an estimated TP availability of about 50 to 80
percent (Oswego R.), and 20 to 50 percent (Black R.) was
found.
-------
SECTION III
RECOMMENDATIONS
This study has shown that for the Lake Ontario drain-
age basin and to some extent the urban drainage entering
Lake Mendota in Wisconsin a substantial part of the total
phosphorus present in urban and rural drainage is not
likely to become available for algal growth. It appears
that based on this study approximately 20 percent on the
average of the non-soluble orthophosphate present in
surface water drainage is available for algal growth.
These results raise serious questions about the validity of
the commonly used approach that the total phosphorus enter-
ing a lake or stream from diffuse sources will become
available for aquatic plant growth. As a result of these
studies, the following are recommended:
1. The best available estimate of aquatic plant available
phosphorus entering a given water course be computed by
determining the soluble orthophosphate plus 20 percent
of the phosphorus that is measured as the difference
between soluble orthophosphate and total phosphorus.
2. Because of the impetus currently being given for a
national phosphate control program, studies of the
type conducted in the Lake Ontario drainage basin and
in Madison, Wisconsin whould be expanded to include
samples from all of the major rivers in the U.S. before
any kind of national phosphate control program is
initiated. There is little point in spending large
-------
amounts of money to control unavailable phosphorus.
The ion exchange incubation technique and algal assay
procedure used in this investigation should be used to
evaluate potentially available phosphorus. Other
chemical methods such as mild acid or basic extraction
procedures overestimate the amount of available phos-
phorus and therefore should not be used.
-------
SECTION IV
LITERATURE REVIEW
INTRODUCTION TO THE CHEMISTRY OF PHOSPHORUS IN NATURAL
WATERS
Phosphorus (P) is present in a multitude of forms in
natural waters. However, the majority of these forms con-
tain the phosphorus atom in combination with oxygen as the
_3
orthophosphate (POU) group. This group is also called
inorganic P or soluble orthophosphate, and it exists
-1 -2
primarily as H?POU or HPOU at the pH values commonly
encountered in natural waters. Soluble orthophosphate is
readily removed from solution by algal cells and can be
directly utilized for cell growth. Hence, soluble ortho-
phosphate is "readily available" to algae. All other forms
of phosphorus are relatively less available to algae than
is soluble orthophosphate, unless the forms can be converted
to soluble orthophosphate at a sufficiently rapid rate such
that the rate of algal growth is not limited by the rate of
conversion. These other relatively "less available" forms
include particulate inorganic, condensed, and organic P;
and soluble condensed and organic P.
Particulate inorganic P includes the inorganic P found
in discrete mineral compounds such as apatite (Ca (OH)?
(P01|)6), wavellite (Alg (OH) g (PO^ ) 2 ) , and strengite
(FePO^. 2H2) ) . The group also includes inorganic P associ-
ated with amorphous , poorly soluble iron or aluminum oxides
10
-------
and hydroxides which may be in combination with suspended
soil particles in the water. The availability of parti-
culate inorganic P forms to algae depends upon sorption-
desorption and solution-precipitation processes, in which
inorganic P is transferred between the solid and solution
phases.
Particulate organic P forms of biological interest
generally have inorganic P in combination with carbon as
t
a phosphate ester: -C-OPO . The cells of living and dead
organisms, as well as organic P esters adsorbed or precipi-
tated on solid particles are included in the particulate
organic P class. In order to become available for algal
growth, particulate organic P compounds must be hydrolyzed
i
to form organic -COH compounds plus soluble inorganic P.
Two mechanisms are possible. The organic P compound may be
released to solution by desorption, cell lysis, or dis-
solution reactions, with subsequent hydrolysis of the
organic P in the solution phase, or the particulate organic
P compound may be hydrolyzed and the inorganic P product
released to solution.
It is clear that soluble organic P compounds form a
pool of potentially available P for any algal species with
the enzymes (phosphatases) necessary for the hydrolysis of
the compounds. Soluble condensed phosphates would form a
similar pool of potentially available P- Condensed phos-
phates are composed of inorganic P groups linked by P-O-P
bonds, and they can be hydrolyzed by chemical and enzymatic
means, producing inorganic P in solution. Condensed phos-
phates associated with particulate matter would become
available for algal growth only after reactions similar to
those described above for particulate organic P forms.
Because the chemical characterization of all the P
compounds in natural waters is extremely difficult, the P
11
-------
forms have instead been classified in terms of routine
laboratory operations. Soluble forms are defined by fil-
tration through a small pore-size filter, generally O.M-5
microns in pore size. The soluble forms are then classi-
fied according to chemical reactivity to a colorimetric
test with acid molybdate. Compounds which react positively
are called dissolved reactive P (DRP). This class includes
soluble inorganic P or orthophosphate, and possibly
colloidal, acid-soluble particulate inorganic P forms or
soluble organic P forms which are hydrolyzed in the acidic
colorimetric test (Chamberlain and Shapiro, 1973).
Generally, however, DRP is felt to be a good estimate of
"readily available" phosphorus in a water sample.
Estimation of soluble orthophosphate plus soluble con-
densed phosphates is generally provided by treatment of a
filtered sample with hot acid before colorimetric analysis
for soluble inorganic P. The result of such a test is often
called soluble hydrolyzable P. Digestion of a filtered
sample with perchloric or persulfuric acids releases in-
organic P from organic P compounds besides hydrolyzing con-
densed phosphates. Thus, total soluble P (TSP) is measured
with such digestion methods.
Hot acid treatment of an unfiltered sample results in
an estimation of total hydrolyzable P, which includes
soluble hydrolyzable P forms plus insoluble condensed phos-
phates and particulate inorganic P dissolved by the acid
treatment. The particulate organic P forms can also be
measured if the unfiltered sample is treated by a perch-
loric or persulfuric acid digestion, resulting in an
estimation of total P (TP). Subtraction of TSP from TP
gives an estimate of the particulate (organic and inorganic)
P (PP) in the sample.
12
-------
SOURCES OF PHOSPHORUS TO URBAN RUNOFF
Precipitation, dustfall, vegetation, street litter, and
eroded soil probably account for most of the phosphorus
carried from urban areas to lakes and streams by stormwater
runoff. Each of these sources will be discussed below, in
terms of their importance and of the relative availability
of the phosphorus forms which they might contribute.
Precipitation
Very little data has been collected on the concentra-
tion of phosphorus in urban precipitation. Weibel et al.
(1966) measured an average concentration of 80 ugP/1 as
total hydrolyzable P in precipitation collected in
Cincinnati. Kluesener (1971) reported an average of 32
ygP/1, with a range of 8 to 90 ugP/1 as total P for rain-
fall near Lake Wingra in Madison, Wisconsin. In the same
study, dissolved reactive P was measured at about 25 ugP/1.
Since dissolved reactive P would be expected to be "readily
available" for algal growth, about 78 percent of the total
P measured in rainfall by Kluesener was apparently "readily
available." Tamm (1951) measured a total P level of about
30 ygP/1 in rainfall collected just outside of Stockholm,
Sweden.
In terms of phosphorus yields, urban rainfall was esti-
mated by Sonzogni and Lee (1972) to contribute 0.20 Ibs
P/acre/year of total P and 0.16 Ibs P/acre/year of dissolved
reactive P. Chalupa (1960) noted a much lower yield for a
reservoir in Czechoslavakia. He computed that 0.008 Ibs
P/acre were brought in by rainfall over a seven month period,
Allen et_ al. (1968) reported a range of 0.18 to 0.88 Ibs
P/acre/year for rainfall at stations 10 to 29 km from large
towns and industries in Great Britain.
13
-------
Dustfall
Much of the phosphorus in urban precipitation may be
derived from the inclusion of atmospheric dust particles.
During dry periods these particles settle out on urban
areas, where they await transport by runoff events. A
recent study by the American Public Works Association
(APWA, 1969) in Chicago recorded an average of 36.9 tons
of dustfall/mi2/month. The dust was 20 to 40 microns in
diameter and varied in yield throughout the year; August
had the lowest and March the highest yields. Sartor and
Boyd (1972) collected composite samples of dry materials
from the streets of Milwaukee, Bucyrus (Ohio), and
Baltimore. Approximately 56 percent of the total P in the
composites was associated with particles less than 43
microns in diameter. They assumed that particles of this
size range could have been produced by industrial stacks
and vents' or raised as dust on construction sites and dis-
persed by air currents. In Seattle, Johnson et al. (1965)
placed jars of water on roofs throughout the city to trap
dustfall. The water-soluble P leached from the dustfall
ranged from 0 to 0.45 Ibs P/acre/year. Kluesener (1971)
trapped dustfall to Lake Wingra and calculated yields of
about 0.7 Ibs P/acre/year for total P and 0.1 Ibs P/acre/
year for dissolved reactive P. Thus, only about 14 percent
of the total P in the dustfall was probably "readily
available" phosphorus.
A possible source of dustfall phosphorus in urban
atmospheres may be lead halophosphate, which has been
detected in automobile exhaust (Anonymous, 1971; reviewed
by Kluesener, 1971). However, Bryan (1970) was unable to
show a positive correlation between lead and phosphorus in
urban runoff from Durham, North Carolina. Thus, the signi-
ficance of this source of phosphorus is still in question.
14
-------
Street Litter
After settling out of the atmosphere, dustfall becomes
just one of the components of urban street litter. The
American Public Works Association (APWA, 1969) research
group classified street litter into five groups, namely:
rags, paper, dust and dirt, vegetation, and inorganic
matter. From June 19 to August 29, the Chicago street
litter had the composition shown in Table 1 (a). The dust
and dirt was the predominant type of litter. During this
period there were only small differences in the total load
of litter between high and low population density areas
and the city-wide average. In October, however, the litter
composition changed considerably, as shown in Table 1 (b).
The great increase in the relative amount of vegetation
apparently resulted from fallen leaves. During October,
the high population density wards showed significantly less
litter than the low density wards, as expected from the dif-
ferences in the amount of forestation between the two types
of urban areas.
The seasonal pattern of street litter loading in the
Chicago studies showed a minimum base loading in the summer
(June - September). Winter residues added extra loading in
spring (March - June), and leaves added to the base load in
October - November. During most of the year, October -
November excepted, the dust and dirt fraction of street
litter was the most significant component, ranging from 45
to 83 percent of the total litter by weight. In terms of
absolute loads, total litter varied from 0.5 to 8.0 Ibs/day/
100 ft. of curb. The loads of dust and dirt varied with the
type of urban land use, as shown in Table 2. The resi-
dential areas showed increasing amounts of dust and dirt as
the population density increased, and industrial areas showed
the highest loads. Sartor and Boyd (1972) found similar
15
-------
characteristics, for total solids loading as a function of
land use, except that commercial areas showed lighter loads
than residential areas.
Table 1. COMPOSITION OF CHICAGO STREET LITTER
Litter in Various Materials
Rags Paper Dust and Dirt Vegetation Inorganic Matter
(% by weight)
(a) June 19 - August 29, 1967
0.2 4.7 72.0 11.1 12.0
(b) October, 1967
0.1 2.4- 36.5 55.0 6.0
Source: APWA, 1969
Physically, dust and dirt was operationally defined in
the Chicago study as that material passing a 1/8 inch mesh
cloth. Laboratory leaching tests in a mixing apparatus
demonstrated that approximately three percent by weight was
soluble in water. The phosphorus content of the leachate
averaged 17 yg P/g of solids leached; however, the phos-
phorus content of the particles was not measured and thus
no estimates of the availability of particulate P could be
made.
Vegetation
Table 1 (b) demonstrates the importance of vegetation
to street litter loads during the fall of the year.
Table 2. DUST AND DIRT LOADS IN CHICAGO
Land Use Dust and Dirt Loads
~(lbs/day/100 ft. of Curb)
Single Family Residential 0.7
Multiple Family Residential 2.3
Commercial 3 . 3
Industrial 7.8
Source: APWA, 1969
16
-------
The contribution of phosphorus from detached vegetation
(litter) will be discussed below. It is known, however,
that living vegetation also has the capacity to contribute
phosphorus to runoff, through contact with precipitation.
Tamm (1951) noted an increase in the phosphorus concentra-
tion of rainfall as a result of passage through a canopy
of pine needles or birch leaves. The rain which passed
through the canopy, called throughfall, was enriched in
phosphorus by three to ten times over the level of phos-
phorus in the incident rainfall. These tests were made in
an area just outside Stockholm, Sweden. In tests conducted
in an experimental forest, the enrichment of phosphorus in
throughfall was only about two times the control level, for
pine and spruce canopies. Tamm concluded that the higher
values near Stockholm may have been due to urban dustfall
accumulations on the leaves of the trees. Carlisle et al.
(1966b) analyzed rain and throughfall for an oak canopy.
They reported a phosphorus yield for throughfall of 1.1 Ibs
P/acre/year compared to 0.38 Ibs P/acre/year for rainfall
controls. Throughfall yields were higher in the period of
vegetative growth than in the leafless period, and in May
they detected very high yields of phosphorus in throughfall,
apparently due to pollen and new leaves. The new leaves
were shown to have a higher phosphorus content (0.7 percent)
than leaves from the rest of the year (0.2 percent). Will
(1959) studied conifer forests in New Zealand and computed
throughfall yields of O.U2 to 3.9 Ibs P/acre/year compared
to rainfall yields of 0.21 to 0.54 Ibs P/acre/year.
Upon detachment of leaves or seeds from trees, the
phosphorus contained in the vegetative litter becomes a
potential source of phosphorus to urban runoff. Inorganic
phosphorus occurs in plants primarily in the vacuolar sap
-2 1
as soluble orthophosphate (HPO^ and ^PO^ ), and concen-
trations in the sap may be hundreds of times greater than
17
-------
the concentration that exists in the soil solution (Lawton,
1961). Organic phosphorus occurs as nucleotides, phos-
phate-sugar esters, phospholipids, nucleoproteins, and
nucleic acids. Seeds are known to contain large amounts
of organic phosphorus, particularly inositol hexaphosphate,
or phytic acid. In vegetative tissues, however, inorganic
phosphorus is usually predominant.
Indirect evidence for the phosphorus contribution of
vegetative litter to urban runoff has been supplied by
Kluesener (1971), who noted high concentrations of phos-
phorus in runoff samples collected in May and November.
The high spring concentrations were attributed to the
leaching of tree seeds by runoff. Although lawn and garden
fertilizers were being applied at this time of the year,
their contribution was discounted by Kluesener because they
were being applied to pervious surfaces where the yield of
runoff was insignificant. Carlisle et al. (1966a) pointed
out that although small, non-leafy materials such as insect
frass, bird scales, and male flowers that fall in the spring
accounted for only 14.7 percent of the dry weight of the
total annual litter-fall in an oak stand, these materials
contributed 40.2 percent of the phosphorus in all types of
vegetative litter.
The November peak in runoff phosphorus seen by Kluesener
(1971) was attributed to the rainfall or runoff leaching of
piles of leaves in the street gutters. Cowen and Lee (1973)
leached oak and poplar leaves with distilled water in lab-
oratory columns. Dissolved reactive phosphorus was readily
leached from the leaves; when oak leaves were cut to expose
veins to the water, the yield of dissolved reactive phos-
phorus was increased from about 240 yg P/g for whole leaves
to 650 yg P/g for the cut leaves (oven-dry weights).
Increases in the time of soaking prior to the column leaching
18
-------
were also shown to increase phosphorus yields. Surface
films of moisture on leaves collected after a rainstorm
were found to contain dissolved reactive phosphorus, which
could be carried off the leaves in subsequent runoff
events.
Other plant leaching studies by Timmons e_t al. (1970)
have demonstrated that the leaching of soluble phosphorus
from plants is greatly enhanced by drying or freezing treat-
ments. Since leaf litter is often subjected to freezing
over the winter months in many urban areas, these results
indicate that high concentrations of dissolved phosphorus
should be found in spring melt waters. The leaching
studies indicate that careless disposal of grass clippings
or leaf litter by urban residents can greatly enhance the
yield of available phosphorus from these materials. Street
gutters or storm sewer catchment basins are clearly the
worst possible disposal sites, as they enhance the leaching
of phosphorus into runoff. Sartor and Boyd (1972) reported
concentrations of 1,100 to 2,200 yg P/l (total P) in super-
natant water from Milwaukee and Baltimore catch basins.
Eroded Soil
Exposed land in the vicinity of impervious surfaces
such as streets or sidewalks has been shown to yield large
amounts of sediment to runoff. "Urbanization" by definition
means the construction of new homes, roads, and other
structures on previously unused or rural lands. Such
activities expose soil to the erosive forces of wind and
rainfall. Wolman and Schick (1967) compared the sediment
loads in streams draining natural or farm lands with loads
in streams draining urban areas near Washington, D.C. and
Baltimore. The sediment concentrations in rural streams
were 2,000 mg/1 or less, while urban streams from conr-
struction areas varied from 3,000 to over 150,000 mg/1.
19
-------
Thompson (1970) evaluated the rates of erosion for parts
of Detroit and found a rate of 69 tons/acre/year for areas
under development, compared with an average rate for the
city of only about 3 tons/acre/year.
Storm sewer loads of sediment have been investigated
by Benzie and Courchaine (1966), in a comparison of separate
and combined sewer system discharges in Michigan. The con-
centrations of suspended solids in the separate system were
10 times those in the combined system, and the solids were
primarily clay resulting from construction and development
activities in the drainage basin. Evans et al. (1968)
tested the change in the total-P of runoff samples (23 to
2700 yg P/l) during various periods of settling in the
laboratory and found that total P was reduced by 30 percent
in five hours' time, while suspended solids were reduced by
70 percent during the same time period. This agrees with
the findings of Sartor and Boyd (1972), that about 44 per-
cent of the total-P by weight in dry street contaminants
was associated with particles over 43 microns in size
(referred to as "sand" on the basis of size), and these
particles made up 74 to 92 percent of the dry matter by
weight in the composite samples from Milwaukee, Bucyrus,
and Baltimore. The other 56 percent of the total-P was
associated with lighter (less than 43 micron) particles,
which have been discussed above in terms of dustfall, and
which would not settle as readily as "sand" size material.
The availability of phosphorus associated with eroded
soils is a complicated problem, involving both (a) physical-
chemical and (b) biological reactions.
Physical-chemical reactions of soils--
Physical-chemical reactions of importance in soils are
adsorption-desorption and precipitation-dissolution. Adsorp-
tion and precipitation reactions between soil constituents
20
-------
and soluble inorganic or organic phosphates result in phos-
phate fixation by the soil and a corresponding decrease in
the concentration of soluble P forms in the water.
The form of phosphorus in the soil will depend upon the
pH, microbial activity, and oxidation-reduction potential
of the soil. In acid soil, inorganic P is usually combined
with iron and aluminum; Williams et al. (1971a) concluded
that, in the non-calcareous lake sediments which they
studied, inorganic P was probably held by a short range
order iron complex. Similarly, Williams et_ al. (1958)
indicated the importance of amorphous iron and aluminum in P
fixation by soils. In neutral or alkaline soils, inorganic
P may be held by fertilizer-soil reaction products or as an
apatite mineral. Sorption or precipitation of inorganic P
by calcium carbonate has been suggested as another mech-
anism (Williams et al., 1971b).
Because clays may contain up to 50 percent of the
total-P in some soils, and because clays are readily eroded
(Scarseth and Chandler, 1938) fixation of phosphorus by
this soil fraction should be quite important in urban
oxide and hydrous oxide coatings on clay mineral surfaces.
Consequently, the mechanisms of inorganic P fixation on
clays may be similar to those in acid soils. The combina-
tion of high P content and susceptibility to erosion results
in a high concentration of phosphorus in eroded soil rela-
tive to the undisturbed soil. Massey and Jackson (1952)
reported concentrations of ph 3-extractable inorganic P in
eroded soils which were 3.4 times the concentration of pH
3-extractable inorganic P in the undisturbed soils, in
samples from four locations in Wisconsin.
A similar enrichment would be expected for the organic
P forms in eroded soils, as up to 85 percent of the total-P
21
-------
in soils may be in organic forms (Alexander, 1961), and
concentrations of organic matter in eroded soils have been
reported to be 2,1 times the concentrations found in undis-
turbed soils (Massey and Jackson, 1952).
The principal organic phosphorus compounds in eroded
soils are those derived from the tissues of plants and
micro-organisms; nucleic acids, phospholipids, and inositol
phosphates. The fixation of nucleic acids by clay minerals
has been studied by Goring and Bartholomew (1952), who con-
cluded that these compounds were adsorbed by a reversible,
cation dependent reaction. The nucleic acids have been
estimated to be present in soils at levels of about five to
ten percent of the total organic phosphorus (Anderson, 1967)
Phospholipids are also a minor fraction of the organic
phosphorus, probably about 0.5 to 2.5 percent of the total
(Alexander, 1961).
The major identified organic phosphorus forms in soils
are the inositol phosphates, Inositol is a six-carbon
cyclitol with several isomers, all of which can form phos-
phate esters containing one to six phosphate groups. The
hexaphosphates form iron and aluminum salts which are
insoluble in acid media; consequently, the fixation of
these compounds is probably analogous to that of inorganic
phosphate. Jackman and Black (1951a) found that in the
presence of an excess of the respective cations, iron
phytate (myoinositol hexaphosphate) was insoluble from pH
2.5 to 8.0 and aluminum phytate was insoluble from pH 3.0
to 9.0. Calcium phytate, in contrast, was insoluble above
pH 6.0 and magnesium phytate was insoluble above pH 9.7.
The salts of the lower inositol phosphate esters were more
soluble than those of the phytates, but still less soluble
than salts of orthophosphate. Anderson and Aldridge (1962)
showed strong sorption of phytic acid on boehmite, soil
22
-------
clay, and montmorillonite at pH 3.0 to H.O. Similarly,
Goring and Bartholomew (1950) noted that the maximum
sorption of inositol hexaphosphate on bentonite, kaolinite,
and illite occurred at pH 3.5 to 4.5, which agreed with the
minimum of the solubility versus pH curves for the iron and
aluminum salts of the inositol hexaphosphate. The same
behavior was noted with orthophosphate.
The reverse of P-fixation reactions in soils are
desorption and dissolution, which result in an increase of
soluble P for possible algal uptake. The balance between
P-fixation and P release reactions is partly related to the
soluble P concentration existing in the vicinity of the
soil particle. Eroded soils are expected to act as phos-
phate "buffers" by removing or adding P to solution in
response to changes in P concentration. Such behavior has
been studied by Taylor and Kunishi (1971), Kunishi et al.
(1972), and Ryden et al. (1972) for eroded agricultural and
urban soils. These authors equilibrated soils in laboratory
systems containing various initial concentrations of soluble
inorganic P. They then calculated the amount of P (in ppm
of dry soil) adsorbed or desorbed by the soil and plotted
these values against the final concentrations of soluble
inorganic P in solution, after equilibration. The inter-
section of the curve with the concentration axis gave the
equilibrium phosphate concentration, which was the concen-
tration in the water at which no net P uptake or release
would take place.
Even more important than the equilibrium phosphate
potential, however, was the slope of the curve. The slope
was an index of the capacity of the soil to buffer the
phosphorus concentration in a solution. Taylor and Kunishi
(1971) defined the biologically available P in the soil as
the amount of P which must be removed from the soil in
23
-------
bringing the soil into equilibrium with an arbitrarily
chosen concentration of P in the water phase. If the
chosen concentration of soluble inorganic P were very low,
as in the epilimnion of a productive lake in late summer,
the biologically available P measured in the test-might well
be close to the entire amount of potentially available P in
the soil.
The kinetics of such buffering action have been inves-
tigated by Li e_t al_. (1972), using lake sediments equili-
3 2
brated with soluble inorganic P-' They reported that a
major portion (45 to 87 percent) of the exchangeable native
sediment P participated in a rapid exchange reaction char-
acterized by a first-order (with respect to P) rate constant
ranging from 7.4 to 46 hours ~ . The exchange process was
graphically resolved into three first-order reactions of
differing rates. Similarly, Amer et al. (1955) have used
P
32 and anion-exchange resin equilibrations of soils to
demonstrate the existence of at least three reactions occur-
ring simultaneously but with different rates of P exchange.
The fastest reaction was completed in about 15 minutes, the
intermediate reaction in two to three hours, and the slowest
reaction was still continuing at the end of 72 hours. Their
resin-adsorption data for four soils showed that between 51
and 75 percent of the P eventually adsorbed by the anion-
exchange resin from the soils in 72 hours was picked up by
the resin in the first two hours of the tests. Such data
demonstrates the time dependence of soil availability term-
inology such as "readily available" and "potentially avail-
able."
Biological reactions of soils
The studies of phosphate exchange cited above did not
distinguish the role of microorganisms in the reactions
from the physical-chemical factors, although the latter
24
-------
were clearly considered to be the dominant factors influen-
cing the results of P exchange experiments. However, micro-
organisms have been shown to remove inorganic P from solution
in competition with lake muds (Phillips, 1964), and the avail-
ability of both inorganic and organic P forms in soils may be
determined, on a long-term basis, by microbial solubilization
reactions.
For example, many types of microorganisms in soils are
capable of solubilizing inorganic phosphorus compounds which
are insoluble under normal soil conditions. Alexander (1961)
estimated that 10 to 50 percent of the bacterial isolates
which he tested could solubilize calcium phosphates. The
bacteria apparently produced organic acids, or in the case
of chemoautotrophs, nitric or sulfuric acids which dissolved
the calcium phosphates. Rodina (1963) stated that bacteria
capable of liberating phosphorus from organic compounds and
those able to transform insoluble compounds into soluble ones
were always present in the detritus of the lakes tested.
Recent work by Harrison et a_L. (1972) on Klamath Lake sediments
showed similar results. The order of microbial solubilization
was CaHP04> Ca3(P04)2> FeP04> Mg(P04)2>Al PO^. A chelation
mechanism was proposed, in which organic ac^ds produced by the
bacteria complexed the cations to release the phosphate to the
water -
Microbial reactions of importance to the availability of
organic P compounds are the hydrolysis reactions catalyzed
by phosphatase enzymes. These reactions split off inorganic
P from its esters. If the ester was insoluble, the inorganic
P may not be released to solution but instead may be bound by
the same mechanisms which held the ester. In this case the
availability of the organic P compound would show no change.
In the case of dissolved esterified P, however, the action of
a phosphatase enzyme would produce soluble inorganic P for
immediate uptake by an algal cell.
25
-------
The investigation of soil-bound organic P hydrolysis
has demonstrated the resistance of these forms of enzymatic
attack. This is especially true for the phytins, which are
hydrolyzed more slowly than phospholipids and nucleic acids
(Alexander, 1961). Jackman (1949) felt that it was not the
level of phytase enzyme activity, but rather the low solu-
bility of phytin that limited its decomposition in his
tests. Likewise, Greaves et al. (1963) estimated that 50
percent of the microorganisms present in soil and on plant
roots possess an active phytase enzyme. Bower (1949) dis-
covered a decrease in enzyme activity when kaolin or bent-
onite was added to a mixture of nuclease enzyme and various
nucleic acid derivatives. The extent of decreased activity
could not be explained solely by the adsorption of the
enzyme by the clays, so that nucleic acid derivatives must
also have been sorbed, accounting for the rest of the
inactivation.
Two other factors besides solubility which can affect
the enzymatic hydrolysis of organic phosphorus are temper-
ature and pH. Bower (1949) found higher rates of hydro-
lysis at 35 C than at 25 C in soils. According to Thompson
and Black (1949), the release of inorganic P responded
slightly to increases in temperature below 30°C, while
above this temperature the rates were markedly influenced
by rising temperature.
The effect of pH on the hydrolysis or organic P in soils
is complicated by its effects on enzyme kinetics and on
organic P solubility. Jackman and Black (1951b) investi-
gated these effects with inositol phosphates and phytase
enzymes at different pH values in soils. Iron and aluminum
salts showed very little hydrolysis in the pH range of 3.0
to 7.0 where these salts are slightly soluble. In the case
of calcium and magnesium salts, the enzyme activity and
26
-------
hydrolysis rates both showed a maximum at pH 5.0 to 6.0,
decreasing rapidly on either side of this pH range. The
solubility of the calcium and magnesium salts was low above
this pH range, and the salts were completely soluble below
this pH range. The hydrolysis rate was apparently maximum
at the pH values where both solubility and enzyme activity
could be achieved. Halstead (1964) used disodium phenyl-
phosphate as a substrate to assay the phosphatase activity
of soil suspensions in buffered solutions. An acid mineral
soil showed activity which gradually increased from pH 2.0
to 7.0. An organic soil gave peaks of activity between
pH 5.0 and 9.5, indicating the presence of both acid and
alkaline phosphatases.
The eventual fate of hydrolyzed inorganic P released
from organic P forms in runoff will depend upon the compo-
sition of the organic residues undergoing microbial
decomposition. Should the concentration of phosphorus in
these residues exceed the relative concentration required
for heterotrophic bacterial nutrition, the excess phosphate
will appear as inorganic P in solution (net mineralization)
This phosphorus would be "readily available" to algae in
the water receiving the runoff. Alexander (1961) calcu-
lated that at least 0.2 percent of the dry weight of
decomposing carbonaceous matter should be phosphorus, if
net mineralization is to occur. At lower concentrations,
net immobilization will occur, as the phosphorus becomes
incorporated into microbial tissue. Some of this phos-
phorus will be released later as soluble inorganic P after
death and lysis of the microorganisms.
27
-------
PHOSPHORUS CONCENTRATIONS IN URBAN RUNOFF
In early studies on urban runoff, the characteristic
high levels of suspended solids, biochemical oxygen demand,
and coliform counts were sufficient evidence to alert water
pollution control workers to the problems of surface runoff,
Later workers began to include nutrient analyses in their
research, as the concern for eutrophication became stronger.
Sylvester and Anderson (1964) summarized their data
from street gutter samples as follows: soluble P (soluble
reactive P) had a median value of 22 ygP/1, and total P had
a median value of 155 ygP/1. Since "soluble reactive"
probably measures readily available phosphorus, about 14-
percent of the total P in their samples was in a readily
available form.
Weibel et al. (1966) reported the range of P concentra-
tions in their Cincinnati study to be 7 to 2,430 ygP/1 as
total hydrolyzable phosphorus, with a mean value of 370
ygP/1. About 62 percent of the total hydrolyzable P was
soluble hydrolyzable P.
Bryan (1970) selected a 1.67 square mile section of
Durham, North Carolina, for sampling urban runoff. Phos-
phorus concentrations were 27 to 157 ygP/1 soluble P (mean
50 ygP/1) and 177 to 806 ygP/1 total P (mean 400). The
ratio of soluble P to total P mean values was 0.13 or 13
percent. Although Bryan did not clearly indicate whether
soluble P was measured as reactive P or hydrolyzable P, the
computed ratio does give an indication of the available
proportion of total P.
Measurements made in Tulsa, Oklahoma (Avco Economic
Systems Corporation 1970) indicated a range of 180 to 1160
ygP/1 for dissolved reactive P, with a mean of 380 ygP/1.
Total P was not measured. Burm et_ al. (1968) sampled storm
sewers in Ann Arbor, Michigan, and reported mean annual
28
-------
values of 270 ygP/1 for soluble P and 1670 ygP/1 for total
P. They did not describe the procedure used to measure
soluble P.
Kluesener (1971) investigated urban runoff P concentra-
tions in Madison, Wisconsin, and obtained an annual average
of 570 ygP/1 as dissolved reactive P and 980 ygP/1 as total
P; indicating that about 58 percent of the total P in his
runoff samples was apparently in readily available forms.
If all the values reported in runoff studies as dis-
solved reactive, soluble, and soluble hydrolyzable are
considered to be estimates of P forms which are eventually
convertible to soluble orthophosphate in a receiving water,
then these values are estimates of "soluble available P."
Table 3 compares the runoff studies on the basis of the
percentage of total P reported as "soluble available P."
About 33 percent was obtained as an overall average from
all the studies listed, indicating that perhaps 67 percent
of total P is in particulate forms of comparatively unknown
algal availability. The concentrations of soluble P forms
listed in the table indicate that, without even considering
particulate P forms, urban runoff probably contains avail-
able P at levels far above the concentration of 10 ygP/1
in lake waters cited by Sawyer (1947) as being an upper
limit before nuisance growths of algae will occur. Hence,
urban runoff would be exerting a phosphorus fertilizing
effect on most lake waters, although the magnitude of the
effect would depend upon the relative volumes of the lake
and of the runoff.
29
-------
EFFECT OF URBAN LAND USE ON RUNOFF PHOSPHORUS
Based on the small number of areas investigated, phos-
phorus yields for urban runoff appear to range from 0.21
to 1.5 Ibs P/acre/year for total P and 0.4 to 2.7 Ibs
P/acre/year for soluble P, as listed in Table 4 . The
differences among the areas investigated cannot be ,
explained completely based on the area type. Bryan (1970)
was not able to relate land used to any of the various
pollution parameters in his studies. Moreover, the studies
of Kluesener (1971) and Weibel et_ al. (1964, 1966) on
residential and residential-light commercial areas,
respectively, showed P yields which were in close agreement
with the yield from an area containing 20 percent commercial
and industrial land uses besides residential, public, insti-
tutional, and unused lands (Bryan, 1970). Hence, the
relationship between phosphorus yield and urban land use is
still not clear.
The distribution of phosphorus sources, in contrast,
does seem to be related to land use. The Chicago study by
the American Public Works Association (1969) ranked the
various urban land used in terms of their dust and dirt
accumulations in the following order: industrial >
commercial > multiple family residential > single family
residential. Sartor and Boyd (1972) found the following
loading intensities for total P in the form of dry street
surface contaminants (Ibs/curb mile): industrial, 3.43;
residential, 1.07; and commercial, 0.29. These values
reflected the variations in total solids loading intensities,
rather than the phosphorus content of the solids, which did
not seem to be significantly different in the various land
uses.
30
-------
Table 3. ESTIMATED PERCENT OF TOTAL P AS "SOLUBLE
AVAILABLE P" IN URBAN RUNOFF
Burm et al .
(1968)
Kluesener
(1971)
Sylvester &
Anderson (1964)
Bryan (1970)
Weibel et al.
(19661
(A)
"Soluble Avail.
Pa" (ygP/1)
270
570
22
50
230
(B) a
Total P
(ygP/1)
1,670
980
155
400
370
Average
(A)/(B)
(percent )
16
58
14
13
6_2_
33
Mean values except for Sylvester £ Anderson (1964),
where median values were reported.
Table 4. PHOSPHORUS YIELDS IN URBAN
STORMWATER RUNOFF
Author
P Yield as:
Soluble P Total P
(Ibs. P/ac*re/yr)
Weibel (1969)
Kluesener (1971)
Jaworski & Hetling (1970)
Bryan (1970)
Avco (1970)
Owen £ Johnson (1966)
0.6
0.4 to 2.7
0.94
1.0
0.21
1.1
1.5
31
-------
PHOSPHORUS IN LAKE ONTARIO TRIBUTARIES
Niagara River
Table 5 shows the five largest waste sources to Lake
Ontario, as compiled in a 1969 report to the International
Joint Commission (IJC) investigating the lower Great Lakes.
The most significant source of phosphorus to Lake Ontario
is the Niagara River, with an estimated total P input of
7,700 to 8,000 tons per year, primarily from Lake Erie and
from municipal wastes entering the river from Canada and
the U.S. (Table 6 )
The mean flow of the Niagara River is 195,000 cfs (cubic
feet per second), or 85 percent of the flow to the lake
from all tributaries. Figure 1 shows the locations of
the major urban areas on the river, which are responsible
for increasing the concentration of phosphorus in the
river by. 71 percent between Lake Erie and Lake Ontario
(IJC, 1969).
Genesee River
The Genesee River originates in the hills of northern
Pennsylvania and flows north across New York to its mouth
on Lake Ontario at Rochester, New York. About three-
quarters of the 2,446 square mile drainage basin is used
for agriculture. The combination of phosphorus from land
drainage in the basin and industrial and municipal wastes
from the Rochester metropolitan area (into the river or
directly into the lake) results in a phosphorus contri-
bution to the lake of about nine percent of the total P
input (Table 5 ).
Oswego River
The Oswego River drains about 5,000 square miles of
north central New York, including most of the Finger Lakes
area. The land uses in the basin are similar to those of
the Genesee basin: cropland with pasture, woodland, and
32
-------
Table 5. MAJOR SOURCES OF PHOSPHORUS TO LAKE ONTARIO
(1966-1967)
% of Total P
Source from all sources
Niagara River, including municipal 56
and industrial sources from the
Buffalo-Niagara Falls area.
Metro Toronto region, including 13
all local municipal, industrial, and
tributary discharges to Lake Ontario.
Metropolitan Rochester area, including 9
all municipal and industrial waste
sources and the Genesee River.
St. Catherine area, including municipal 5.2
and industrial waste sources and Twelve
Mile Creek.
Hamilton area, including municipal, 2.3
industrial, and tributary discharges
to Hamilton Harbour.
(Source: IJC, 1969)
T *7
33
-------
Ol
Ji.
SOURCE: NOAA( 1971)
0 10 20 30 40 50
STATUTE MILES
Figure I. Lake Ontario and major tributaries
-------
Table 6. SOURCES OF PHOSPHORUS TO THE NIAGARA RIVER
Source
(short tons/year)
Lake Erie
New York
State
Municipal Wastes
Industrial Wastes
Land Drainage
2,000
150
50
Province
of
Ontario
Municipal Wastes
Industrial Wastes
Land Drainage
330
80
50
Unaccountable
Total
540
7,700e
(Source: IJC, 1969)
aA value of M-,800 short tons per year has been estimated
for Lake Erie input in 1972 (Great Lakes Water Quality,
1972); this new value would change the total in the table
to 8,000 tons per year-
35
-------
forest. The 1969 IJC report estimated a stream loading
from municipal treatment plants and untreated sources in
the basin equivalent to a population of 475,000 persons.
Black River
The Black River drainage basin covers an area of about
1,930 square miles on the western slope of the Adirondack
Plateau. The upper reaches of the basin are mountainous
and wooded, while the region near Lake Ontario is similar
to the agricultural lands of the Genesee or Oswego basins.
In terms of phosphorus loads to Lake Ontario, the 1969
IJC report listed the New York rivers in the following
order of increasing loads: Black R., 181 tons P/yr.;
Genesee R., 314 tons P/yr.; Oswego R., 619 tons P/yr.; and
Niagara R., 7,700 tons P/yr. The annual total P input from
all sources (tributaries plus direct discharges to the lake)
with planned P reduction programs, has been estimated at
17,800 tons in 1972 (Great Lakes Water Quality, 1972), com-
pared to an estimated 13,700 tons in 1969 (IJC, 1969).
Since the lake is presently in a stage of nutrient content
intermediate between oligotrophy (nutrient-poor) and meso-
trophy (intermediate nutrient enrichment), future increases
in the phosphorus and nitrogen loadings from tributaries
and direct waste discharges could cause major changes in
algal production in the lake. Already, the inshore waters
near Toronto and the mouths of the Niagara, Genesee, and
Oswego Rivers show serious growths of the green alga
Cladophora and the presence of benthic (bottom-dwelling)
organisms which are normally characteristic of eutrophic
(nutrient-rich) waters (IJC, 1969).
36
-------
Phosphorus Loads and Stream Discharge
The rate of both sediment and phosphorus loading from
a river basin is not constant throughout the year, but
instead correlates with the stream hydrograph. During
periods of higher than normal flow, the base flow of the
stream is supplemented with surface runoff waters, which
generally carry much higher concentrations of suspended
sediment and total P than does the base flow. Since both
the flow and the average P concentration are increased,
the yield per unit time (flow times concentration) must
also increase. Owen and Johnson (1966) found highly signi-
ficant correlations between daily yield of total P and
streamflow from six watersheds (urban and agricultural) in
the metropolitan Toronto region of the Lake Ontario drain-
age basin. The association between high total P concentra-
tions and stream discharge from an agricultural watershed
was noted by Shannon and Lee (1966) in their studies on
Black Earth Creek, Wisconsin. They reported peak concen-
trations of about 900 ugP/1 of total P and 500 ugP/1 of
soluble orthophosphate plus soluble condensed phosphates,
indicating that perhaps 55 percent of the total P was in
particulate forms during the period of snowraelt in the
watershed.
In addition to inputs of phosphorus from runoff during
high flow periods, stream loads are also increased by high
rates of streambank erosion and resuspension of streambed
sediments up into the wash load carried by the water above
the bottom of the stream. Since phosphorus is associated
with the streambank and wash load sediments, the increased
sediment loads during high flow periods should result in
increased particulate phosphorus loads during these periods
In terms of percent of total annual loading of particulate
phosphorus to Lake Ontario, the particulate matter carried
37
-------
into the lake during the high spring flow period should be
the most important input, and availability studies on this
material would probably be more significant than studies
on material which enters the lake in smaller quantity during
the rest of the year-
ANALYSES OF AVAILABLE PHOSPHORUS
Chemical Methods
Since soluble orthophosphate, or soluble inorganic P,
is readily available for algal growth in natural waters,
the question of availability as applied to a given phos-
phorus form can be reduced to the question of how rapidly
and to what extent that phosphorus form can supply soluble
orthophosphate to an algal cell.
Consequently, attempts have been made to accurately
measure soluble orthophosphate, as the best means of
estimating the true value of "readily available P" in a
water sample. One of the most widely used procedures is
the molybdenum blue procedure with an antimony catalyst and
ascorbic acid as a reducing agent (Murphy and Riley, 1962).
This procedure is subject to arsenate interferences
(Chamberlain and Shapiro, 1969), and the acidic conditions
of the test may hydrolyze any very labile organic phosphate
esters in the sample (Rigler, 1966, 1968). Chamberlain and
Shapiro (1973) have concluded, however, that hydrolysis of
organic phosphate esters in 0.45 micron pore-size filtered
samples would probably not cause significant errors in the
assay of soluble orthophosphate in natural surface lake
waters.
Another possibility of error would be the presence of
acid-labile inorganic phosphate of particulate nature but
small enough to pass the 0.45 micron pore-size membrane
filters generally employed for soluble orthophosphate
analysis. Chamberlain (1968) reported that about 90 percent
38
-------
3 2
of the P in 0.45 micron pore-size membrane-filtered lake
water (prepared from lake water to which soluble inorganic
32
P had been added) could be removed by refiltering the
water through a 0.01 micron pore-size membrane filter, and
3 2
about 80 percent of the P could be removed by ultra-
centrifugation.
Some investigators have attempted to evaluate the
molybdenum blue procedures by comparison with algal bio-
assays. Chamberlain and Shapiro (1969) assayed soluble
orthophosphate with a 30-second extraction procedure and
with a one-hour uptake test with Microcystis aeruginosa,
and found no serious discrepancies between the results in
arsenate-free Minnesota lakes. Kuenzler and Ketchum (1962)
and Rigler (1966), however, found discrepancies between the
32
results obtained using isotope partitioning of PCL
(between algal cells and the growth medium) and the results
obtained from chemical analyses of the inorganic P in
solution. In Rigler's (1966) case, the discrepancies were
10 to 100 fold, probably as a result of the very low con-
centrations of soluble inorganic P involved. In contrast,
Walton and Lee (1972) found essentially no difference
between chemically measured and algal-available ortho-
phosphate in Lake Mendota water, sediment extracts, and
algal extracts.
The estimation of available phosphate in runoff soil
suspensions is complicated by the presence of significant
amounts of inorganic P sorbed to the soil particles. Direct
analysis of the runoff by colorimetric procedures for
soluble orthophosphate would require a preliminary filtra-
tion step, which would prevent the estimation of potentially
available inorganic P on the particles filtered off. Since
some of this particulate inorganic P may become available
to algae through dissolution or desorption reactions in the
39
-------
receiving water, the errors in neglecting soil-bound
inorganic P would be quite significant. Agronomists have
devised several procedures for estimating the content of
plant-available phosphate in soils; these procedures pro-
vide data which can be correlated with greenhouse or field
tests of plant growth, in order to assess the phosphate
status of the soil in terms of agricultural productivity.
Various chemical extraction procedures have been found to
correlate well with plant growth on various types of soils.
Some of the extractants used are water, carbon dioxide
saturated water, acids, bases, salts, and various buffered
solutions.
Another method of estimating the pool of potentially
available inorganic P in soils is radioisotopic exchange
3 2
with inorganic P. Li et al. (1972) found that the amount
32
of inorganic P which would exchange with inorganic P
comprised from 19 to 43 percent of the total native inor-
ganic P in the sediments. Amer et al. (1955) compared
3 2
P methods to methods where the soil P was adsorbed by
anion-exchange resins. Quantities of P adsorbed from soils
by resin were less than those quantities which equilibrated
3 2
with P during the same time intervals. However, the
correlation coefficient for P adsorbed by the resin in two
hours and P-availability to plants in the greenhouse was
0.95.
The ion-exchange resin method has been developed into a
practical laboratory procedure for testing soils for
available P, through the efforts of Cooke and Hislop (1963)
and Hislop and Cooke (1968). They reported that the degree
of exchange was temperature dependent, with a two-fold
increase in P availability between 10 and 30°C. Time was
not a critical factor after 12 hours of equilibration of
soil and resin in aqueous suspension (Cooke and Hislop,
40
-------
1963). Lathwell el: al. (1958) compared resin methods with
chemical extractive methods. The resin methods gave the
highest degree of correlation with crop response data,
especially with respect to phosphorus uptake by plants.
Kunishi et_ al. (1972) used the method of Hislop and Cooke
(1968) to estimate the available P sorbed by eroded soils
from an agricultural drainage basin, during actual runoff
events. About 5 to 20 percent of the amount adsorbed
remained available to the resin.
In contrast to the soil P availability tests which are
based on correlations with plant growth or phosphorus
uptake, other tests have been developed for the purpose of
quantitating the chemical forms of soil P. These tests
contribute to an understanding of the processes by which
phosphorus in soils is fixed or released. One of the
earliest attempts at a scheme of soil phosphorus fraction-
ation was that of Chang and Jackson (1957), who classified
phosphates in soils into four main groups: calcium phos-
phate, aluminum phosphate, iron phosphate, and reductant-
soluble phosphate extractable after removal of the first
three forms. Their extraction scheme employed successive
extractions of the soil with ammonium fluoride (0.5N),
sodium hydroxide (0.1N), sulfuric acid (0.5N), and sodium
citrate-sodium dithionite solution. Each extractant was
intended to measure one of the four main groups of phosphate
when the extractions were carried out in the proper sequence
However, their characterization of iron, aluminum, and
calcium phosphates as discrete chemical forms probably is
not entirely accurate because of the presence of amorphous,
poorly soluble phosphates resulting from sorption or preci-
pitation reactions on soil surfaces (Stumm and Morgan, 1970)
Several modifications of the Chang and Jackson (1957)
methods have been proposed in recent years. Williams et al.
41
-------
(1971b) found that with calcareous lake sediments the Chang
and Jackson procedure underestimated the total amount of
inorganic P solubilized during the ammonium fluoride and
sodium hydroxide extractions, because of resorption of the
inorganic P by calcium carbonate or calcium fluoride. They
proposed a fractionation scheme for calcareous soils based
on successive single extractions with sodium hydroxide,
citrate-dithionite-bicarbonate, and hydrochloric acid.
Shukla ejt al. (1971) used both acid ammonium oxalate and
citrate-dithionite-bicarbonate to show that sorption of
inorganic P was apparently controlled by amorphous iron
oxides in both calcareous and non-calcareous lake sedi-
ments .
In summary, the soil P availability tests developed
for routine use are operationally defined, although they
may be correlated with soil P availability measured
biologically. The procedures for measuring the various
forms of phosphorus in soils may also be related to bio-
logical availability, if the P form being measured is the
same form responsible for phosphorus release under the
proper conditions. It is clear that in both types of
tests, some correlative relationship to plant (or algal, in
the case of lake sediments) bioassays must exist if the
tests are to have practical uses.
Biological Methods
Generally, the chemical methods of measuring available
phosphorus measure dissolved inorganic P plus inorganic P
extracted from soils or other insoluble particles. The
contribution from organic P forms in solution or in the
soils are neglected as insignificant compared to inorganic
P. This is probably a valid assumption in the case of
runoff or river waters carrying large suspended soil loads.
In cases where much of the phosphorus may also exist as
42
-------
vegetal matter or as microbial cells (bacteria, algae,
and zooplankton), biological techniques are needed. Three
basic techniques have been used: batch mineralization
systems, cell-free systems, and algal bioassays.
Waksman et al. (1937) used batch mineralization systems
to estimate the extent of nutrient regeneration in sea
water. Samples of water were stored in darkness to allow
bacterial growth and subsequent die-off. During incuba-
tion, inorganic P in solution was measured in order to
follow the release of available P with time. Renn (1937)
conducted similar experiments and demonstrated that P
assimilation occurred during the first few days of the
3
test, when the bacterial numbers were increasing to 10
to 10 times their normal population levels in natural
sea water. P regeneration followed rapidly after die-off
of the bacteria. When a carbon source such as glucose was
added, the time required for P regeneration was much
longer than in normal sea water. Consequently, the
efficiency of bacteria as agents of phosphorus regeneration
is related to the supply of other nutrients besides
phosphorus in the water. If these nutrients are in short
supply relative to phosphorus, P regeneration will be
rapid, as the life span of the bacteria will be short due
to the nutrient limitations.
If chloroform or toluene is added to soils or natural
waters, the microbial cells will lyse and release their
cellular phosphates and phosphatase enzymes into solution.
This cell-free method results in a much more rapid rate of
inorganic P production than in batch mineralization
systems, since microbial P assimilation reactions are
prevented by the sterilizing chemicals. Thompson and Black
(1949) showed a three-fold increase in the apparent hydro-
lysis of soil organic P in toluene-treated samples compared
43
-------
to untreated controls. Berman (1970) incubated Lake
Kinneret water with chloroform to show the release of
soluble inorganic P from the dissolved organic or parti-
culate P in the lake. About 53 percent of the total P in
the samples was converted to dissolved reactive P in three
to five days. However, his samples probably contained
mainly algal-P and zooplankton-P, hence his results would
not be expected to be characteristic of an urban runoff
system with large suspended soil loads. A similar study
by Golterman (1960) showed that almost 65 percent of the
total cell P was released as dissolved reactive P in a
28-day incubation of chloroform-treated Scenedesmus cells.
In the presence of soils, any inorganic P released
via cell lysis or phosphatase enzyme action may be sorbed
by the soils and never measured in solution. Acid extrac-
tions have been used by several workers to remove the
sorbed inorganic P. Rogers (1942) used 10 percent hydro-
chloric acid, and Thompson and Black (1949) and Bower
(1949) used IN sulfuric acid in tests with toluene.
Direct algal bioassay of available P in soils has been
performed by Fitzgerald (1970a), who incubated P-starved
Cladophora in direct contact with aerobic lake muds. The
(filamentous) alga was then separated from the mud for
cellular P analysis. Boiling water extractions of the
cells showed that, at most, only one percent of the total
sediment P was available. Golterman et_ al. (1969), in
contrast, demonstrated that the phosphate taken up from
their lake muds by Scenedesmus was about 5 to 30 percent
of the total P in the muds. Similarly, Spear (1970) con-
ducted long-term dark aerobic incubations of muds from the
same lake (Mendota) sampled by Fitzgerald, and found
phosphorus release. Tests of phosphorus precipitated by
iron or calcium in systems where unicellular Chlorella was
44
-------
separated from the insoluble P forms by dialysis tubing or
membranes showed that such P sources were available for
algal growth (Fitzgerald, 1970a,b). This was also demon-
strated by Golterman ejt al. (1969), with Scenedesmus grown
in direct contact with FePO^ (added in dry form) or hydro-
xyapatite. Apparently the mechanisms governing P avail-
ability in lake muds are not explained by simple
insolubility properties alone.
The direct algal bioassay of filtered waters has been
used by Skulberg (1964) and Baalsrud (1967) in Europe.
Their work eventually led to the Algal Assay Procedure (AAP)
Bottle Test (Environmental Protection Agency, 1971). This
procedure called for filtration of the water sample before
inoculation with the test alga, although the sample could
be autoclaved prior to the filtration if an estimate of
the available P in bacterial, algal, and zooplankton cells
were desired. The filtration step was required in order
to eliminate competition between the natural organisms
and the inoculum (test) alga and to eliminate interferences
in the measurement of algal growth by light absorbance or
fluorescence procedures. Such interferences would be
caused by detritus or soil particles in the natural water
samples. Golterman et_ al. (1969) were able to quantitate
Scenedesmus obliquus cells in the presence of lake muds by
making microscopic cell counts, thus eliminating the need
to alter the natural system by filtering or autoclaving.
Another way to avoid such alterations has been demonstrated
by Goldman et_ al. (1969), who measured the rate of carbon-
14 dioxide uptake by natural plankton populations in
unfiltered samples. They reported this method to be more
sensitive than cell counts for the quantitation of algal
population changes. The carbon-14 procedure was also used
by Plumb (1973) in measurements of the response of very
45
-------
dilute populations of Lake Superior phytoplankton to
taconite tailings.
SUMMARY
A review of the literature has indicated that urban
runoff carries a high concentration of phosphorus, much of
which is in the particulate phase. The fraction of the
particulate P which might be available for algal growth
has not been thoroughly investigated, however, either in
the runoff or in river waters containing a combination of
runoff and wastewater inputs.
46
-------
SECTION V
SAMPLING OF URBAN RUNOFF AND LAKE ONTARIO
TRIBUTARIES
MADISON URBAN RUNOFF
Samples of urban runoff were collected from several
sites in the city of Madison, Wisconsin (population 171,769
in 1970), from August, 1972 to March, 1973 (Table 7.)
Table 8 lists the sampling stations and their locations.
The type of urban land use in the vicinity of each station
is given in Table 9 . Three major types of land use were
selected: residential, commercial, and urban construction.
Runoff from industrial land uses was not sampled, because
such areas constituted a small fraction of the total
Madison urban area. Although construction sites also
account for a small portion of the urban area, the high
yields of sediment from urban construction activities
(Wolman and Schick, 1967; Thompson, 1970) likely contri-
bute to appreciable phosphorus transport.
Residential population densities in Madison generally
decrease along a line from the state capitol square (see
Figure 2 ; on the isthmus between Lakes Mendota and
Monona) to the western boundary of the city. Thus, the
locations of sampling stations H, D, B, and A shown in
Figure 2 were in areas zoned R6, R5, R2, and Rl, areas
of decreasing population density, respectively (see Table
9)^ In terms of forestry, the area around station B on
47
-------
Table 7. MADISON STORMWATER RUNOFF RECORD
Runoff Event
No. Date
1 August 11, 1972
2 August 19, 1972
3 August 23, 1972
4 September 19, 1972
5 September 20, 1972
6 October 20, 1972
7 October 22, 1972
8 December 30, 1972
9* January 17, 1973
10 January 18, 1973
11 February 1, 1973
12 March 5, 1973
*Runoff was from snowmelt alone, with no precipitation.
48
-------
Table 8.
MADISON URBAN RUNOFF SAMPLING STATIONS
Station
Location
A
B
D
H
Inlet grate of the open storm sewer in
the median strip of Whitney Way, near
the Montauk Place Intersection
Inlet grate of the open storm sewer in
the median strip of Manitou Way, near
the Tumalo Trail intersection
Outlet of the storm drain pipe under the
Water Chemistry Laboratory driveway, on
the U. of Wis. campus
Street gutter inlet grate at the inter-
section of King, Butler, and Wilson
Streets; on Wilson Street
Street gutter near the corner of Island
Drive and Masthead Streets; on Island
Drive
Outlet of the large storm sewer pipe in
Law Park which drains the capitol square
area and is located on line with
Pinckney Street extended
Street gutter inlet grate at Xrhe end of
Broom Street near the intersection of
Broom Street and John Nolan Drive
Street gutter on Whitney Way, near the
Sheboygan Street Intersection
49
-------
rj
Figure 2. Urban runoff sampling stations in Madison,Wisconsin
-------
Table 9. DESCRIPTION OF THE URBAN AREAS SAMPLED
FOR RUNOFF PHOSPHORUS
Sampling Madison Zoning Code Description of Land Uses
Station for Drainage Area in the Drainage Area
A Rl A single family residential
area, with some low density
multiple family dwellings
B R2 An area much like Rl, except
that less useable open
space per dwelling is
allowed
D R5 The University of Wisconsin
campus area around Bascom
Hill
H R6 A high density residential
area located near the center
of the city, with low use-
able open space allowed per
dwelling unit
E and G C4- The central commercial
district of Madison, around
the capitol square
F Rl, R5, Cl and C2 A residential-light com-
mercial area; construction
activity with soil erosion
until sod was put down
I R2 A residential area; con-
struction activity with
soil erosion
51
-------
Manitou Way appeared to have the highest density of trees
of all the residential stations. The lowest density of
trees was in the area near Station H (Broom Street).
The commercial area selected comprised the central
business district of Madison, located around the capitol
square. This section of the city is characterized by
heavy automobile and pedestrian traffic, and most of the
land area is covered by impervious (street, sidewalk, and
roof) surfaces. Two stations were selected for collection
of runoff from the commercial area (Table 9). Station G
was sampled one time, on September 20, 1972; all other
samples were collected from Station E which could be
sampled more conveniently than Station G.
Madison urban construction sites were chosen for
sampling on the basis of observed soil erosion during
runoff events in the summer of 1972. Station F was sampled
from August, 1972 to January, 1973. The last sample taken,
January 17, was taken after the site had been sodded to
prevent further erosion. Hence, this sample (No. F-9)
probably represented the runoff from a residential-light
commercial area more closely than runoff from a construc-
tion site. Station I was selected at a construction site
about one mile from Station F, as an alternate site for
samples of construction site runoff collected after
January 17, 1973.
All runoff samples were collected as grab samples in
one-gallon acid-washed polyethylene cubitainers. Shallow
streams of runoff in street gutters were sampled with a
partially collapsed cubitainer, so that the mouth of the
container could be placed as close to the gutter surface
as possible. Repeated grab samples collected in this
manner were combined in a second cubitainer until a gallon
of sample water was collected. Most samples were collected
52
-------
at inlet grates, where the runoff was falling over a ledge
and could be readily caught in a cubitainer. Because the
depths of the runoff streams were usually quite shallow
and the flows fairly turbulent, the particulate matter in
the samples was assumed to be representative of the
particle size distribution being transported in the runoff.
Observation of samples allowed to stand a short time after
collection often showed the presence of relatively heavy
sediment, indicating that the sampling procedures were
probably not introducing a serious bias by collecting only
light particulate matter.
Previous urban runoff studies (Kluesener, 1971; Weibel
et al., 1964) have shown that grab samples taken at differ-
ent times during the runoff event may vary both in the
phosphorus concentration and in the ratio of soluble to
particulate P forms. The size distribution of the parti-
culate matter transported might also be expected to vary,
and this could cause a qualitative change in the nature of
the particulate P forms in the samples, hence also affect
the measured phosphorus availability of particulate P
forms. It was hoped that by sampling a given site more
than once over a relatively long period of time, some of
these differences in particle size distribution might be
taken into account in an overall average availability value
for that site.
PRECIPITATION SAMPLES
In an effort to assess the algal growth potential of
dustfall and precipitation, snow samples were collected in
Madison and rain gage samples were received from IFYGL
personnel in the state of New York.
Madison snow samples were collected from the following
three sites on April 10, 1973: 1) the roof of the City-
County Building near the state capitol, 2) a lawn across
53
-------
Commercial Avenue from the Oscar Mayer meat packing plant,
and 3) a hill just inside the entrance to the Picnic Point
Park on the University of Wisconsin campus. Surface
layers of snow 0 to 3 cm in depth were scraped into plastic
bottles which had been cut to form scoops. The scoop-
bottles were covered with aluminum foil and stored at 4 C
until analysis.
New York rain gages were established by the Environ-
mental Protection Agency (EPA) Laboratory, Rochester, N.Y.,
at the locations shown in Table 10. The notations "open"
or "closed" in the table indicate the collection of bulk
precipitation (dustfall plus rainfall) or of rainfall
alone, respectively. Samples were collected during May and
June, 1973, and were shipped to Madison in plastic bottles
without addition of preservatives.
GENESEE RIVER BASIN SAMPLES
In cooperation with the New York State Department of
Environmental Conservation, water samples from streams in
the Genesee River Basin were sent to Madison for analyses
of phosphorus availability. Table 11 lists the locations
of the sampling stations along with the predominant land
use in the sub-basin drained by each stream. A detailed
description of land uses within the sub-basins, expressed
in terms of acreage and percent of total area, is given in
Table 12. The sub-basins are listed according to pre-
dominant land use. Two of the stations were located in
urban areas, while the others were located in agricultural,
forested, or brushland areas. All samples were collected
and sent to Madison in one-gallon cubitainers enclosed in
an insulated box with two to three small bottles of frozen
water for refrigeration. The samples were generally
collected on a biweekly schedule and were received in
Madison 1 to 9 days after collection.
54
-------
Table 10. LOCATION OF THE EPA NEW YORK RAIN GAGES
Sample
No.*
601-0
602-C
603-
604-
605-
C
C
C
Precipitation Collection Rain Gage
Period Location
May
May
May
May
May
1
1
1
1
1
- May
- May
- May
- May
- May
3
3
3
3
3
o,
o,
o,
o,
o,
197
197
197
197
197
3
3
3
3
3
Macedon,
N.
Skeneateles
Oswego, N
Brockport
.Y
?
University
Y.
, N.Y
N.Y.
of
606-C
May 1 - May 30, 1973
Rochester, N.Y.
Cape Vincent,N.Y,
601-C
602-C
603-0
604-C
605-C
606-C
608-0
June
June
June
June
June
June
June
1
1
1
1
1
1
1
- June
- June
- June
- June
- June
- June
- June
30,
30,
30,
30,
30,
30,
30,
1973
1973
1973
1973
1973
1973
1973
Macedon, N.Y.
Skeneateles, N.Y.
Oswego, N.Y.
Brockport, N.Y.
University of
Rochester, N.Y.
Cape Vincent, N.Y.
Clarence, N.Y.
0 = Open at all times
C = Closed except during precipitation events
(Source: Casey, 1973)
55
-------
Table 11. LOCATION OF RUNOFF SAMPLING STATIONS
IN THE GENESEE RIVER BASIN, N. Y.
Station
No.
1
2
3
H
5
6
7
Station^Location ,
U.S.G.S. Map Name
Bryon
Rochester East
Geneseo
Geneseo
Geneseo
Springwater
Dansville
Predominant
Land Use
Cropland
Urban
Beginning of
pasture
Pasture
Beginning of
pasture
Forest
High density
Stream
Sampled
Spring Creek
Allen Creek
--
Jaycox Creek
Briggs Gully
__
Dansville
Andover
residential
Beginning of
high density
residential
Brushland
East Valley
Creek
*United States Geological Survey
(Source: Boulton, 1972)
56
-------
Table 12.
LAND USE DISTRIBUTION IN SUB-BASINS
OF THE GENESEE RIVER BASIN, N. Y.
Land Use
No. Acres
%of
Total Area
Cropland - Spring Creek
(Batavia N. and Bryon)
Cropland
Brush
Forest
Bogs and Wooded
High intensity
agriculture
Pasture
Misc.
TOTAL
Wetlands
7063
1046
720
761
2047
739
157
12533
57
8
6
6
16
6
1
100
Pasture
Forest
Brush
Agriculture
TOTAL
Pasture - Jaycox
(Geneseo)
1168
Creek
- inactive
31
16
1223
Brush
Forest
Agriculture
Misc.
TOTAL
Brushland - East Valley Creek
(Andover and Greenwood)
2571
788
1409
36
4804
Forest - Briggs Gully
(Springwater and Bristol Springs)
Forest 2016
Brush 622
Cropland 20
Bogs and Wooded Wetland 179
Misc. 2_
TOTAL 2839
95
1
3
1
100
53
16
30
1
100
71
22
1
6
0
100
57
-------
Table 12. (continued)
LAND USE DISTRIBUTION IN SUB-BASINS
OF THE GENESEE RIVER BASIN, N.Y.
T ~TT^ M A % of Total
Land Use No. Acres Area
Urban - Allen Creek
(Rochester E. and Mendon Ponds)
Residential, Public 8366 52
Outdoor .recreation
Commercial
Agriculture 3771 23
Forest and Brush 3194 19
Industrial 288 2
Misc. 646 4
TOTAL 16265 100
Residential (High Density)
(Dansville)
Residential 159 91
(high density)
Urban (downtown) 15
TOTAL 174
(Source: Boulton, 1972)
LAKE ONTARIO TRIBUTARY SAMPLES
Four New York tributaries to Lake Ontario were
selected for sampling (Figure 1 arrows). Grab samples
were collected periodically to dtudy the forms of phos-
phorus carried in suspension by the rivers. No attempt
was made to sample the bed load (sediment transported along
the stream bottom). Several samples of river water were
collected by personnel of the State Universities of New
York at Buffalo and Oswego. Other samples were collected
by University of Wisconsin students and staff during the
1973 period of high spring flow at the following locations:
58
-------
The Niagara River was sampled at two locations, one
above Niagara Falls at Beaver Island State Park (Grand
Island), and one below the falls at the mouth of the river,
near the Fort Niagara Coast Guard Station on Lake Ontario.
The Beaver Island Park samples were taken from the west
branch of the river, not far from the southern tip of
Grand Island. The Fort Niagara samples were taken from the
New York bank of the river, about 50 yards from the mouth.
The Genesee River was sampled from the Route 104 bridge,
or from the navigation locks near the bridge, in the city
of Oswego. The Route 104 bridge is located less than a
mile from the river mouth.
The Black River was sampled from the south bank near
the Route 180 bridge at Dexter, New York, which is located
about one mile from Black River Bay on Lake Ontario.
The spring flow samples were collected from the sur-
face of the rivers with a plastic bucket, which was either
lowered from a bridge or thrown out from shore. All
samples were shipped to Madison on commercial airlines;
samples sent by workers in New York were receoved 4 to 9
days after collection, while samples collected by
University of Wisconsin students and staff were received
within two days of collection. All samples were stored
at 4 C until analyzed; generally, such storage lasted
2 to 7 days.
59
-------
SECTION VI
ANALYTICAL METHODS
All water samples in this investigation were analyzed
by chemical and biological methods designed to estimate
the algal-available fraction of the particulate P, soluble
P, or total P in the samples. This chapter presents the
methods used to quantitate the phosphorus forms and to
estimate their availability. The rationale for the
methods and an evaluation of them is also presented.
CHEMICAL ANALYSES OF PHOSPHORUS FORMS
Useful estimates of phosphorus availability in natural
waters may be obtained from chemical analyses of phos-
phorus forms where such analyses have been correlated with
the results of algal growth bioassays, as in the studies of
Walton and Lee (1972), and Chamberlain and Shapiro (1969).
Chemical methods serve to qualify total phosphorus meas-
urements by providing estimates of the fraction of total P
available to algae.
In this investigation, the following seven operation-
ally defined phosphorus forms or fractions were selected
for analysis:
1) Dissolved reactive phosphorus (DRP)
2) Total soluble phosphorus (TSP)
3) Particulate phosphorus (PP)
4) Total phosphorus (TP)
60
-------
5) Particulate inorganic phosphorus (PP.) extracted
by:
a) Acid
b) Base
c) Anion-exchange resin
The first four forms in the above list were determined to
quantitate the phosphorus forms in the samples. TP gives
the upper limit of available P. DRP includes mainly dis-
solved orthophosphate and hence gives an estimate of the
readily available P. TSP includes DRP plus other soluble,
less readily available forms such as condensed phosphates
or organic phosphate esters. PP consists of the total
inorganic plus organic phosphate associated with parti-
culate matter; hence PP is less readily available than DRP.
The PP extraction procedures were selected from the
literature (Wentz, 1967; Williams et_ a_l. , 1967; and Hislop
and Cooke, 1968) to approximate the fraction of PP which
might be available to algae in a body of water- Although
these procedures may extract both organic and inorganic P
forms from the PP, the analytical methods used to deter-
mine the quantity of P in the extract measures mainly
inorganic P. Consequently, the chemically measured P
extracted from PP is termed the "PP. extracted" by acid,
base, or resin.
PP. extracted by acid includes calcium phosphates and
much of the iron- and aluminum-bound phosphates. PP.
extracted by base includes phosphates bound to iron or
aluminum but not those associated with calcium. Neither
method measures phosphates occluded in oxides of iron
(reductant-soluble P). Consequently, the phosphates
measured are probably those precipitated or adsorbed on
the surfaces of soil particles or bound in discrete mineral
phases. If surface-bound phosphate is the major fraction
61
-------
extracted, then the extractable PPi should be correlated
with available P. Surface-bound phosphates should be
more readily available than occluded or slowly-soluble
discrete phase phosphates (Chang and Jackson, 1957).
Since neither the acid nor the base extractions should
cause appreciable hydrolysis of organic phosphates, the
PP. in the extracts should be primarily derived from
inorganic P forms associated with particulate matter.
The PP. extracted by resin is comprised of that
inorganic P on the particles which can equilibrate fairly
rapidly with the inorganic P in the solution phase around
the particles. By decreasing the concentration of inor-
ganic P in the solution phase to low levels, the anion-
exchange resin forces a reversal of phosphate sorption and
precipitation reactions. The PP. transported to the resin
"sink" is expected to approximate the inorganic phosphate
which would be released from the particles upon dilution
in a phosphorus-deficient receiving water.
Dissolved Reactive Phosphorus
DRP was determined by the colorimetric method of
Murphy and Riley (1962) after filtration of the samples
to remove particulate matter. The color reagent was added
to the filtrates in the ratio of 3 ml of reagent to M-0 ml
of filtrate, and the mixture was diluted to 50 ml, as
specified by Murphy and Riley (1962). Alternatively, 4 ml
of color reagent was added to 20 ml of sample without
dilution. In either case, standards, blanks, and ^samples
were treated alike to compensate for the different volume
ratios used.
Filtration for the initial DRP values of all samples
was performed with O.U5 micron pore-size Millipore
membrane filters, which had been presoaked in dilute HC1
(1 to 2 ml concentrated HC1 per 500 ml of water) for at
Millipore Corp., Bedford, Mass.
62
-------
least one day before use. The filters were rinsed and
stored in distilled water until used. Chemical extracts
of PP or dark-incubated water samples were filtered
through either membrane or No. 934AH Reeve-Angel glass
fiber filters , of undefined pore size. The objective
of filtration was to obtain a clarified sample for the
colorimetric analysis of soluble inorganic P. In most
cases, glass fiber filters were used for river waters
and membrane filters for urban runoff. After December,
1972, however, all filtrations were performed with
membrane filters. Table 13 shows the effect of filter
type on the colorimetric analysis of New York river
waters. Significant differences between the results
with glass fiber and membrane filters at the 95 percent
confidence level were found for two of the samples
tested, while only one sample (No. 42) showed a signifi-
cant difference at the 99 percent confidence level. In
terms of the fraction of total P represented by the two
mean values for sample No. 42, however, the difference
was only 2 percent (23 and 21 percent of total P for
glass fiber and membrane filter methods, respectively).
Because of the wide range of phosphorus concentra-
tions encountered in these investigations, contamination
of volumetric glassware, especially pipettes, was con-
sidered a potential problem. To minimize this problem,
the use of pipettes was restricted by using calibrated
test tubes for the measurement of filtrate volumes and
as vessels for performing the DRP color reaction on the
filtrates. Since the color reagent effectively
"cleaned" the test tubes by a complexation reaction
a'Reeve Angel Co., Clifton, N.J.
63
-------
Table 13, EFFECT OF FILTER TYPE ON THE DRP ANALYSIS
OF NEW YORK RIVER WATERS
DRP Concentration in Filtrate
No. 934AH Reeve Angel 0.4-5 micron pore
glass fiber filters size Millipore
Sample
No. 40
Niagara R.
mean value
std. deviation
3
3
5
5
3
4a
0.7
(ugP/1) filters
3
3
3
3
3
3a
0
No. 42 Genesee R. 33 31
35 31
35 31
35 32
35 31
mean value 35a'b 31a'b
std. deviation 0.9 0.4
No. 44 Black R.
mean value
std. deviation
No. 47 Oswego R.
mean value
std. deviation
6
6
6
6
6
6
0
57
55
57
61
--
58
6.5
6
6
6
6
6
6
0
55
57
57
57
57
57
0.9
a
The mean yalues for glass fiber and Millipore filters
were significantly different at the 95% confidence
level.
b
level.
The mean yalues for glass fiber and Millipore filters
were significantly different at the 99% confidence
64
-------
with the phosphate, only a few rinses with distilled
water were required to wash the tubes for the next set
of samples. A set of 25 X 200 mm test tubes was cali-
brated at 20, 40, and 50 ml volumes by pipetting water
into the tubes and marking the meniscus with a black
line on the outside of the tubes. Table 14 presents
the results of replicate DRP analyses made by filling
ten randomly selected tubes to the 20 ml marks with a
river water filtrate. Four ml of Murphy and Riley
(1962) reagent were added to each tube with a repro-
ducible-volume Repipeta so that the results of the
analyses would reflect the variability of the sample
measurement step. The coefficient of variation was
only 1 percent in this test. Volumetric measurement of
40 ml with the calibrated test tubes would have even
less variability because the relative volume error
would be halved in comparison to measurement of 20 ml
volumes.
The absorbance of the color resulting from addition
of the Murphy and Riley reagent to 20 or 40 ml of sample
was measured after about 30 minutes on a Beckman Model
DU spectrophotometer with either a 5 or 10 cm absorp-
<2
tion cell, or on a Spectronic 20 spectrophotometer
with either a ^-inch or 1-inch absorption cell. Each
set of samples was accompanied by a series of standards
and a reagent blank, plus a membrane filter bank as
necessary. All blanks, samples, and standards were
measured against distilled water at 882 nm wavelength.
The standard solutions were prepared with volumetric
glassware.
,Labindustries Co., Berkeley, Cal.
cBeckman Instruments, Inc., Fullerton, Cal.
Bausch and Lomb Co., Rochester, N.Y.
65
-------
Table 14. PRECISION OF DRP ANALYSES
USING MARKED TEST TUBES
Sample Replication No.
1
2
3
4
5
6
7
8
9
10
DRP
(ueP/1)
60
62
62
62
62
62
62
62
62
62
mean value 62
standard deviation 0 . 6
coefficient of
variation 1%
66
-------
Total Soluble Phosphorus
TSP was determined in the same filtrates used for
DRP analysis. A 20 ml sample was transferred into a
calibrated test tube or pipetted into a 50 ml Erlenmeyer
flask, and 0.2 ml of a strong sulfuric acid solution
(30 ml of concentrated H^SO^ diluted to 100 ml with
water) added. A freshly prepared solution of potassium
persulfate was made by adding 5 grams of the salt to
100 ml of water and warming on a hot plate to just dis-
solve the salt. A 3 ml aliquot of this solution was
added to each acidified sample. The flasks or digestion
tubes were capped with aluminum foil and autoclaved for
about one hour at 15 pounds per square inch (psi)
pressure. The samples were then allowed to cool, and a
predetermined volume of 3N NaOH was added to neutralize
the digestion acid. DRP was determined after dilution
of the mixture to 50 ml with water and 8 ml of color
reagent.
Total Phosphorus
TP was determined in unfiltered samples using a
digestion procedure similar to that us,ed for TSP,
except that a filtration step was sometimes required
after the digestion. Residual particulate matter was
filtered out with glass fiber or membrane filters before
neutralization of the samples with 3N NaOH. If neutral-
ization preceded filtration, much of the phosphorus was
lost in the filtration step. Apparently iron hydroxides
were formed at the neutral pH and were responsible for
sorption of phosphate during the subsequent filtration.
Standards and blanks prepared for the TSP analysis and
carried through the persulfate digestion procedure
described above were also used to quantitate TP. All
67
-------
volumetric measurements of samples to be analyzed for
TP were made with graduated cylinders in an effort to
obtain representative suspensions of particulate matter.
Particulate Phosphorus
PP, the total insoluble inorganic and organic
phosphorus in a sample, was determined either by calcu-
lation (TP-TSP) or by direct persulfate digestion of
particles separated from the bulk solution by membrane
filtration.
Analysis by calculation was performed by subtract-
ing the mean value of three replicate TSP determinations
from the mean value of three replicate TP determinations
made on the same sample. The results obtained in this
manner were generally used to quantitate the amount of
particulate phosphorus taken for chemical extraction.
In these extractions, the sample particulate matter was
deposited on a membrane filter by passing the sample
through the filter. The filter plus particles were
then placed into extraction solution and scraped clean
with a metal spatula. Since the filters were soaked
for at least 5 minutes, even fine material retained in
the pores of the filter may have been extracted to some
extent. As an approximation, it was assumed that all
of the PP deposited on the filter ((TP-TSP) X volume
filtered) from the sample was subjected to extraction,
so the PP. found in the extract could be expressed as a
percent of the calculated PP on the filter.
Particulate matter for algal bioassays was isolated
from the bulk solution in the same manner as described
above. The particles deposited on the filters (non-
filtrable particles) were scraped into P-free algal
medium, and the filters were removed from the medium
68
-------
before inoculation with algae. Thus, some of the sample
particles may have been trapped in the filters and hence
not subjected to the bioassay. The true concentration
of PP in the algal medium would then differ from the con-
centration computed by difference (TP-TSP). The true
concentration was best estimated in this case by perform-
ing a TP (persulfate digestion) analysis on the suspension
of particles in the medium or on a similarly prepared
suspension in distilled water.
Table 15 presents a comparison of the two proce-
dures for estimating the total PP in a sample. A
"scraping efficiency" ratio was computed as the ratio of
PP found by digestion of scraped particles divided by PP
determined from calculation (TP-TSP). An overall
average ratio of 0.95 was found from the 19 samples
tested. Low values (<0.80) were seen in three samples.
Two of the samples, No. 50 and No. 56 had very low PP
values, and PP determined by calculation may have been
subject to a relatively large error (Appendix A). The
low value for the other sample (No. 59) may have been due
to an experimental error in the recorded volume of sample
filtered to obtain the particles for direct digestion, as
this volume was required along with the analytical con-
centration of PP found by direct digestion, to compute
the value shown in column two of Table 15 (59 ygP/1).
In general, the results shown in this table indicate that
most of the particulate phosphorus in the bulk sample was
transferred to the bioassay medium or into the chemical
extraction solutions.
The terms TP, TSP, and (total) PP, which have been
used in connection with persulfate digestions, were
operationally defined by that digestion procedure. The
69
-------
Table 15. EFFICIENCY OF SCRAPING PARTICULATE MATTER
FROM MEMBRANE FILTERS
Sample PP (ygP/l)a
CD
Determined from ,
Sample scraped particles
(2)
Determined by
calculation
(TP-TSP)
Ratio ,
E-ll
D-ll
H-ll
A-12
B-12
D-12
1-12
Genesee R,
No. 34
No. 42
No. 51
No. 58
Oswego R.
No. 43
No. 52
No. 59
Black R.
No. 44
No. 53
No. 60
Niagara R.
No. 50
No. 56
Madison Runoff
138
453
253
194
408
524
1540
145 0.95
441 1.03
262 0.96
214 0.91
421 0.97
561 0.93
1419 1.09
New York Rivers
387
101
62
128
56
51
59
24
24
68
12
13
360 1.08
105 0.96
62 1.00
150 0.86
50 1.12
48 1.06
88 0.67
19 1.26
25 0.96
75 0.91
19 0.63
26 0.50
Average 0.95
Mean values of triplicate determinations of PP, or of
TP and TSP used in the calculation (TP-TSP)
Values obtained from direct persulfate digestion of
particles in distilled water or (-P) algal medium of
volume v, where the particles were derived from a vol-
ume V of original sample. Thus:
PP(in table)" PP(from digestion)X (v/V)
70
-------
persulfate digestion procedure used routinely in the
studies reported here was essentially that given in
Standard Methods (1971). Although it was realized that
this procedure may not have measured all of the phos-
phorus in some types of samples, the method was used
because of its wide usage in other studies of urban
runoff and nutrient budgets for lakes, as well as in
related International Field Year for the Great Lakes
(IFYGL) research. The calculated values of percent
availability may be applicable to other workers' TP or
PP data, if all the data are based on a common digestion
procedure.
Table 16 shows that the persulfate method was not
significantly less effective than a method using a
stronger oxidizing agent (perchloric acid in a ternary
mixture with nitric and sulfuric acids( for one of the
samples tested (urban runoff No. D-12). For the other
sample tested (No. 1-12, collected near an urban constr-
uction site), the perchloric acid method gave a TP value
about 8 percent higher than the TP value from the per-
sulfate procedure. With a solution of inositol phos-
phates supplied by Dr. W. Weimer , both methods appeared
to be equally effective. In contrast to the oxidizing
methods, a simple hot (one hour at 100°C) acid hydrolysis
with the sulfuric acid used in the persulfate method was
very ineffective in a TSP analysis of the inositol phos-
phates. Based on the tests reported in Table 16 , the
aDr. Walter C. Weimer was a graduate student at the
University of Wisconsin Water Chemistry Laboratory at
the time of these investigations. His present address
is Pacific Northwest Laboratories, Battelle Boulevard,
Richland, Wash. 99342.
71
-------
persulfate method probably underestimates TP, TSP, or
PP by 8 percent or less relative to the perchloric acid
method.
Table 16. EFFICIENCY OF THE PERSULFATE DIGESTION METHOD
AS COMPARED TO A PERCHLORIC ACID METHOD
1.) Particles Filtered from Madison Runoff
PP (ygP/1)
. Perchloric ,
Sample No. Persulfate digestion acid digestion
D-12
mean values x.
std. deviation
(x9 - x, ) not
1-12
mean values x.
std. deviation
145
152
150
145
152
= 149 x2 =
s = 4 s =
significant at the 95%
410
405
400
395
400
= 402 x2 =
sn = 6 So =
175
130
145
177
140
153
21
confidence level
407
455
455
407
460
437
27
(x2 - x-j^) significant at the 95% confidence level
(x2 - x,) = 8 percent of x~
2.) Inositol Phosphate Solution
A. Persulfate Digestion61
Replicate No.
1
2
3
4
range
TSP
(ygP/1)
398
398
400
398
398-400
72
-------
Table 16 (cont'd). EFFICIENCY OF THE PERSULFATE DIGESTION
METHOD AS COMPARED TO A PERCHLORIC ACID
METHOD
B. Perchloric Acid Digestion*3
Replicate No. TSP
(ygP/1)
1 385
2 402
C. Sulfuric Acidc
range 385-402
TSP - 18 ygP/1
Method given in text.
b20 ml sample + 1 ml cone. HNOg/HCLO^(60%)/cone. H^O^
predigestion of 20 ml sample W/l ml cone. HNO~.
c d
20 ml sample + 0.2 ml of H9SOusoln. given in text;
heated 1 hr. at 100°C. A *
Table 17 summarizes the estimated random errors of
each of the analytical methods used to characterize the
phosphorus forms in runoff and stream samples. The
statistical methods used in these estimates are given in
Appendix A, Coefficients of variation (cv) of 5.5 per-
cent or less were found for the methods of direct
analysis. PP determined by calculation as TP-TSP would
have a cv which would be partially dependent upon the
value of PP. If this value were about the same as
(>_ 90 percent of) TP, the cv for PP would be about 2 per-
cent, as calculated using the cv values for TP and TSP
in Table 17. The cv for PP would be increased to 4 and
9 percent of PP if PP were decreased to one-half and one-
fourth of TP, respectively. The approximate accuracy of
the analyses was tested by adding orthophosphate spikes
to samples used for DRP or TP analyses (Table 18), The
average recovery of a 25 ygP/1 spike in the DRP analyses
was 23 ygP.l, or 92 percent. The same percent spike
recovery was found for a 500 ygP/1 spike in a TP analysis
of urban runoff Sample B-9.
73
-------
Table 17. ESTIMATED PRECISION OF CHEMICAL ANALYSES FOR
SOLUBLE AND PARTICULATE PHOSPHORUS FORMS
Procedure
DRP
TSP
TP
PP
Number
of Mean
Replicates Value
10
10
10
6a
62
63
2140
129
Coefficient
Range Standard of
Deviation Variation
. (iio-P '1 "1
*> Hgr / X )
60-62
61-65
2000-2220
117-136
0.6
1.6
76
7.1
1.
2.
3.
5.
0%
5%
6%
5%
Replicates from a suspension of particles in AAP (-P)
algal medium
Table 18. RECOVERY OF ADDED ORTHOPHOSPHATE SPIKES
IN DRP AND TSP ANALYSES
A. DRP: Samples
filtered before
DRPa
In In
Sample No. sample 25
56-Niagara R.
57-Niagara R.
58-Genesee R.
59-Oswego R.
60-Black R.
501-13-Genesee
R. Basin
(ygP/1)
37
3
46
57
11
adding
spike
sample +
yg/1 spike Spike
57
29
68
77
35
0 25
(ygP/1)
20
26
22
20
24
25
mean values 23
Recovery
(% of
added P)
80
104
88
80
96
100
92
Mean of three replicates
74
-------
Table 18 (cont'd.). RECOVERY OF ADDED ORTHOPHOSPHATE SPIKES
IN DRP AND TSP ANALYSES
B. Total P: Sample B-9 (Unfiltered Madison Urban Runoff)
TP Concentration
In
sample
In sample +
500 ygP/1 spike
(UgP/D
mean values: x,
difference:
(x2 -
1076
1080
1028
1080
1028
1058
+ 462 ygP/1
1544
1494
1540
1510
1512
1520
% of added spike recovered: 462/500 x 100 = 92%
Particulate Inorganic P (PP.)Extracted by Chemicals
PP.^ Extracted by Acid
Particles were deposited on a membrane filter by passing
a known volume of sample through the filter. The particles
and the filter were placed into a 50 ml Erlenmeyer flask with
20 ml of acid extraction solution, the filter was scraped with
a metal spatula, and the flask was then swirled by hand for 5
to 15 minutes at room temperature. The filter was left in the
flask during extraction. The acid extraction solution
contained 8.1 ml of concentrated HC1 and 1.3 ml of concentra-
ted H2S04, diluted to 2 liters with distilled water (Wentz,
1967). The calculated acid concentration was 0.083N.
75
-------
The acid extracts were analyzed for DRP after filtra-
tion through glass fiber or membrane filters, directly
into 25 X 200 mm calibrated test tubes. Prior to 1973, the
entire contents of the extraction flask were poured into
the filter apparatus, followed by a few small volumes of
water used to rinse the flask. This procedure was simpli-
fied in later tests, by simply taking a 15 ml aliquot from
the extract for filtration. With either procedure, the
filtrate was diluted to 50 ml with water and eight ml of
color reagent for the analysis of extracted PPj_- Standards
and blanks were made with 15 (or 20) ml of extraction
solution, to compensate for any pH effects on color
development caused by the extraction acids. An average
filter blank of 3 ygP/1 was calculated from extractions of
20 washed membrane filters, run periodically during 1972
and 1973. This average blank was subtracted from the
acid extraction PP. values.
PP. Extracted by Anion-Exchange Resin
"^""1
Particles isolated from a water sample onto a membrane
filter were scraped into 100 ml of distilled water in a 125
ml Erlenmeyer flask or standard BOD (Biochemical Oxygen
Demand) bottle; the membrane filter was left in the test
flask. One gram of the chloride form was added to the flask,
which was then sealed with Parafilm and shaken mechanically
for 24 hours at 15 to 22°C on a wrist-action shaker. After
the equilibration period, the resin beads were allowed to
settle and a 20 ml portion was decanted from the supernatant
liquid for filtration and DRP analysis.
American Can Co., Neenah, Wise.
76
-------
The suspension of particulate matter and resin remain-
ing in the test flask was poured through a CRC Micro Sieve
(U.S. Series No. 60: 250 microns). Since the resin had been
dry-sieved before use (only retained beads were saved for
use), and because of resin swelling, almost all of the resin
was retained on the wire sieve during the wet-sieving opera-
tion, while the particulate matter from the sample was
assumed to pass through the sieve quantitatively, leaving
only the resin beads on the wire mesh. A few small rinses
with water were used to rinse any residual resin out of the
test flask onto the sieve.
The slurry of wet resin on the sieve was scraped into
a long-stem funnel, which was fitted with a pinch clamp and
filled with IN Na_SCL, as shown in Figure 3. The resin
was allowed to soak in the funnel for at least 2 hours before
the pinch clamp was opened to begin the elution, with fresh
IN NaSCL added to the reservoir as needed. A 100 ml volume
of effluent was collected in a volumetric flask, at an
elution rate of approximately 3 ml/minute. The contents of
the volumetric flask were mixed well, and a 40 ml aliquot
was taken for analysis of resin-bound inorganic P. The
aliquot was diluted to 50 ml with water and 8 ml of color
reagent for this analysis.
Table 19 summarizes the percent recovery of phosphate
spikes added to 1 gram of resin. The average percent
recoveries are plotted in Figure 4. This graph was used
to find the correction factor necessary for the calculation
of resin-bound inorganic P (Appendix A) . The inorganic P
bound to the resin and the supernatant DRP of the test
flask were added to get the apparent "available" P in the
flask.
Chemical Rubber Co., Cleveland, Ohio
77
-------
Figure 3. Column for solution of phosphate from Dowex I
X8 resin
IN Na2S04
SOLUTION-
DOWEX RESIN-
GLASS WOOL PLUG-
FLEXIBLE TUBING
WITH PINCH CLAMP»
78
-------
Figure 4. Recovery of added phosphate from Dowex
X8 resin by I N Na2S04 leaching
100
90
80
or
UJ 70
0 60
UJ
* 50
i
r
S 40
O
Q;
UJ 30
Q.
20
10
S J" o <**
M 9 ° 1
$-
-
-
O - RANGE
-
i i i i i i I I 1 i
10 20 30 40 50 60
P ADDED TO RESIN,
70 80 90 100 '
79
-------
Table 19 . RECOVERY OF ADDED ORTHOPHOSPHATE FROM
DOWEX 1-X8 ANION-EXCHANGE RESIN BY
IN Na2S04 LEACHING
Ortho P
Added to Number of Average % Recovery % Recovery
Resin Determinations of Added P Range
(ygP/g resin)
2 3 83 80-84
5 4 87 81-90
10 3 95 93 - 96
21 11 92 89 - 94
22 3 91 87 - 93
25 1 92 92
35 3 100 98 - 102
50 5 98 95 - 102
75 3 99 98 - 99
100 5 101 96 - 103
80
-------
All standards and reagent blanks were prepared with
the same IN Na^SO^ solution used to elute the resin. A
resin blank was run with 1 gram of resin in 100 ml of dis-
tilled water and was leached along with the samples. The
inorganic P in the resin blank eluant was subtracted from
that in the sample eluants. The reagent blank was used to
correct the standards.
Table 20 summarizes a series of tests which were run
to estimate the efficiency of acid, base, and resin extrac-
tions. Orthophosphate spikes were added to the extraction
solutions prior to introduction of the particulate matter,
except in the case of the resin extraction, where the spike
was added to a suspension of the particles in water a few
minutes before adding the resin. The mean values showed
recoveries of 77 to 100 percent for these methods. The low
value for the base extraction of Sample No. 34 may have been
due to resorption of phosphate in the extraction solution
by calcium carbonate (Williams et al., 1971b). No such
explanation was available for the apparent low recovery
from the acid extraction of Sample No. 502-14. The ranges
of experimental results with both spiked and unspiked
Sample No. 502-14 were unfortunately large relative to the
size of the added spike, so that the apparent recovery was
not measured precisely. As a conservative estimate of the
recovery of acid- or base-extractable P with the methods
employed here, about 75 percent was recovered. Generally,
all analyses of extractable PP^ and the other P forms dis-
cussed above were performed in triplicate.
ALGAL ASSAYS OF PARTICULATE PHOSPHORUS
Direct estimation of the available phosphorus in
natural particulate matter was made by growing Selenastrum
capricornutum in phosphorus-free medium containing the
particulate matter. The procedure consisted of four steps:
81
-------
Table 20. RECOVERY OF ORTHOPHOSPHATE SPIKES ADDED TO
ACID, BASE, AND RESIN EXTRACTION SOLUTIONS
BEFORE EQUILIBRATION WITH SAMPLE PARTICLES
A. 0.083N Acid Extractions
1.) Sample No. 42 (Genesee R.) Particles
DRP Concentrations in Extracts
after Equilibration with Particles
Unspiked extraction Spiked extraction
solution solution ( + 200 ygP/1)
279
292
308
x-j^ = 293
(ygP/1)
480
504
496
x2 = 493
mean values
difference: (x2 - x ) = 200 ygP/1
% of added spike recovered: 200/200 x 100 = 100%
2.) Sample No. 502-14 (Genesee R. Basin Sample)
Particles
DRP Concentrations in Extracts
after Equilibration with Particles
Unspiked extraction Spiked extraction
solution solution (+100 ygP/1)
311
332
337
(ygP/1)
381
409
421
mean values x, = 327 y. = 404
difference: (x. - x ) = 77 ygP/1
% of added spike recovered: 77/100 x 100 = 77f
-------
Table 20 (cont'd). RECOVERY OF ORTHOPHOSPHATE SPIKES
ADDED TO ACID, BASE, AND RESIN EX-
TRACTION SOLUTIONS BEFORE EQUILI-
BRATION WITH SAMPLE PARTICLES
B. 0.1N NaOH Extraction of Sample No. 34 (Genesee R.)
Particles
DRP Concentration in Extracts
after Equilibration with Particles
Unspiked extraction Spiked extraction
solution solution ( + 238 ygP/1)
46 224
58 248
50 237
mean values x, = 51 x? = 236
difference: (JL - iL) = 185 ygP/1
% of added spike recovered: 185/238 x 100 = 78%
C. Anion-Exchange Resin Extraction of Sample No. 507-14
(Genesee R. Basin Sample) Particles
Total DRP Concentration in Supernatant and
in Na2SO^ Extract of Resin after Equili-
bration of Resin and Particles
Unspiked extractionSpiked extraction
solution solution
(H20 + resin) (H20+100 ygP/1 = resin)
Xn =
27
30
28
28
(ygP/1)
126
128
127
x0 = 127
mean values
difference:
% of added spike recovered: 99/100 x 100 = 99%
difference: (x - x ) = 99 ygP/1
O 7.
O J
-------
1.) preparation of the inoculum, 2.) preparation of the test
flasks, 3.) growth of the algae, and 4.) quantitation of the
algal growth. Steps No. 1 and No. 3 were general procedures
which were also used for soluble P bioassays (discussed
below). Steps No. 2 and No. 4 were performed differently in
assays of particulate P than in assays of soluble P.
Preparation of the Algal Inoculum
A unialgal culture of Selenastrum capricornutum,
initially obtained from Dr. G. P. Fitzgerald , was sub-
cultured at approximately biweekly intervals by transfer-
ring a small volume of mature cell culture to a liter of
freshly prepared, sterile growth medium. The medium used
for the stock culture transfers was an enriched Algal Assay
Procedure (AAP; EPA, 1971) medium containing three times
the nitrate and phosphate levels of AAP medium.
The standard AAP medium composition is given in
Table 21. Six stock solutions of the individual macro-
nutrient salts were made with reagent grade chemicals dis-
solved in glass distilled water at 1000 times the final
concentration in AAP medium. A single stock micronutrient
solution was made by combining all of the micronutrient
salts in glass distilled water at 1000 times their final
concentrations in the AAP medium. The complete medium was
made by combining 1 ml volumes of all stock solutions and
diluting to 1 liter with glass distilled water -
The algal inoculum for growth assays was washed from
its stock culture medium by an alternating sequence of
three centrifugations (1500 rpm for 20 min.) and three re-
suspensions of the pelletized cells in about 10 ml of fresh
AAP medium, minus phosphate (AAP(-P)). Following the third
Water Chemistry Laboratory, University of Wisconsin,
Madison
84
-------
Table 21. COMPOSITION OF STANDARD AAP ALGAL MEDIUM
A. Macronutrients
Compound
NaNOg
K2HPCV
MgCl2
MgSO^ . 7
CaCl2 . 2
NaHC03
Concentration
(mg/1)
25.500
1.044
5.700
H20 14.700
H20 4.410
15.000
Element
N
P
Mg
S
C
Ca
Na
K
Concentration
(mg/1)
4.200
0.186
2.904
1.911
2.143
1.202
11.001
0.469
B. Micronutrients
Compound
H3B°3
MnCl2
ZnCl2
CoCl2
CuCl2
Na0MoO., .
2 4
FeCl3
Na2EDTA .
Concentration
(yg/1)
185.520
264.264
32.709
0.780
0.009
2H00 7.260
2
96.000
2H20 300.000
Element
B
Mn
Zn
Co
Cu
Mo
Fe
Concentration
(yg/D
32.460
115.374
15.691
0.354
0.004
2.878
33.051
(Source: EPA, 1971)
85
-------
resuspension, the cell suspension was counted on a hemo-
cytometer under a microscope, to determine the dilution
needed to achieve the proper volume of suspension and cell
4.
concentration in the suspension. One ml of a 27 X 10
cells/ml inoculum cell suspension was required for each
assay test flask. Except in the transfer of stock cultures,
sterile techniques were not used in the handling of the
algal cells or the preparation of algal assay media.
Preparation of the Test Flasks
All assay cultures were grown in 50 ml Erlenmeyer
flasks cleaned with hot, concentrated nitric acid and
rinsed with regular distilled water (six times), followed
by glass distilled water (four times).
Sample particles isolated on membrane filters, as
described above, were scraped into AAP(-P) medium, and the
suspension was shaken to distribute the particles uniformly
in the medium. Five 25 ml aliquots were taken with a 25 ml
graduated cylinder and poured into the culture flasks. The
remainder of the suspension was saved for PP analysis. One
ml of glass distilled water and 1 ml of inoculum cell sus-
pension were added to each flask to give a final culture
volume of 27 ml, with an initial algal cell count of
4
1 X 10 cells/ml. Comparison tests were also run with
autoclaved particulate matter in AAP(-P) medium. The sus-
pensions were autoclaved for 15 minutes at 15 psi, cooled,
and bubbled with carbon dioxide, if necessary, to adjust
the pH to 7-8 before use in a bioassay.
Standard culture flasks contained 25 ml of AAP(-P)
medium plus 1 ml of inoculum cell suspension and 1 ml of
standard KH PO^ solution (made with glass distilled water),
which was prepared from the same primary standard solution
used for chemical analyses. Since the volume of the
standard cultures, like the sample cultures, was 27 ml, the
86
-------
standard Kf^PO^ solutions contained 27 times their final
concentrations in the assay flasks. Blanks were prepared
in the same manner as the standards, except that 1 ml of
water was substituted for 1 ml of standard P solution.
Generally, five replicate flasks were run for blanks,
standards and samples. All flasks were stoppered with
cotton plugs.
Growth of the Algae
The culture flasks were incubated under approximately
400 foot-candles of continuous light provided by cool white
fluorescent lamps in a constant temperature room set at
20°C. Under the lights, the temperature was 22+2°C. The
flasks were swirled daily by hand during their incubation.
Tests were run to determine the proper incubation
period for the bioassays with Selenastrum. Figure 5 shows
the results of one of the tests, where the average light
absorbance (750 nm) of five standard flasks at each phos-
phorus level was followed versus time of incubation. The
average absorbance values at all levels of P appeared to
have leveled off between 15 and 18 days. After 18 days,
some of the cultures showed gradual increases in absorbance.
Based on such data, an 18 day incubation period was chosen
for the growth assays. Available phosphorus from particu-
late matter was thus operationally defined by the population
of algae which the particulate phosphorus could produce in
a phosphorus-free medium in 18 days, compared to the growth
from orthophosphate in the same medium. Figure 6 shows
the relationship between algal growth after 18 days and the
initial concentration of orthophosphate in the cultures. Up
to 200 ygP/1, phosphorus appeared to be clearly limiting the
algal growth reached in 18 days. Above 200 ygP/1, however,
the slope of the curve changed but did not appear to reach
the well-defined plateau (slope^O) expected if phosphorus
87
-------
Figure 5. Absorbonce of Selenastrum cultures in AAP medium vs. time
CO
CO
ORTHOPHOSPHATE
CONCENTRATIONS
0.5
0.4
E
c
O
UJ
o
O Blank
A 10 /ig P/l
O 25 ^g P/l
D 50/ig P/l
h» 100 ^g P/l
A200/tg P/l
300 /ig P/l
0.3
K 0.2
O
tn
CD
O.I
0
12 16
TIME ( days)
-------
Figure 6. Standard curve for cell counts of S. copicornutum after
18 days vs. orthophosphate concentration (0-600/igP/l)
1200
CO
ID
E
x.
CO
UJ
o
800
400
100 200 300 400
ORTHOPHOSPHATE./igP/l
500
600
-------
were no longer limiting the growth. A possible explanation
may be that phosphorus was still nutritionally limiting,
but self-shading at high culture densities restricted the
light intensity for part of the culture in each flask,
causing a partial growth limitation and the rounding off of
the curve in Figure 6 . However, to avoid such effects,
attempts were made to work in the range of 0 to 200 ygP/1.
Quantitation of Algal Growth
Following the incubation period, the culture flasks
were shaken and, if necessary, scraped with a spatula to
loosen cells from the flask walls. A small volume was
withdrawn and placed into a hemocytometer counting chamber
under a microscope. The Selenastrum cells in the chamber
were quantitated by counting 10 fields of at least 50 cells
per field. An exception to this procedure was made for the
-4-
blanks or 10 ygP/1 standards, where 10 fields of 10 ml
volume were counted. A second exception was made for bio-
assays of unfiltered New York precipitation samples in which
very little algal growth was apparent. These assay
cultures were compared to standards with 10 ygP/1 by meas-
urement of light absorbance at 750 nm wavelength (A75Q).
Since the samples contained particulate matter which added
to the absorbance from algae, the apparent growth in the
cultures was biased positively. If such positively biased
A750 values were found to be less than the A75Q values for
the 10 ygP/1 standards, then the actual growth in the
samples must have been less than the growth in the 10 ygP/1
cultures, and the available P concentration in the samples
was reported as less than 10 ygP/1.
The 10 hemocytometer field counts from each sample
flask were averaged, and the mean field count was compared
to a standard curve like that in Figure 6 to find the
apparent available P in the flask. Standard curves were
90
-------
drawn through points which represented the average cell
count from five replicate flasks, where each flask cell
count was the average of ten hemocytometer field counts.
Thus, each point was derived from 50 field counts from
each phosphorus level. As shown in Appendix A , a
"smallest detectable" phosphorus concentration was calcu-
lated from the mean and standard deviation of the blank
flasks. Sample flasks with cell counts lower than a
specified cutoff value ("C" in Appendix A ) were assigned
an apparent P value of less than the smallest detectable P
concentration, although the lack of growth may have been
due to an inhibition unrelated to the presence or absence
of phosphorus.
Table 22 shows the recovery of orthophosphate spikes
(50 ygP/1) added to two suspensions of river water particles
which had been autoclaved before assay in AAP(-P) medium.
Recoveries of only 76 to 86 percent were found for the two
samples, which had PP concentrations of 79 to 115 ygP/1,
respectively, before autoclaving. The autoclaving procedure
may have solubilized some of this particulate phosphorus,
but evidently the PP remaining sorbed some of the added
orthophosphate spike, and rendered it unavailable to
Selenastrum. The different recoveries for the two samples
may have been related to the different origins of the par-
ticulate matter, namely the Niagara and Genesee Rivers. As
a comparison test, 25 ygP/1 spikes were added to an auto-
claved, filtered Niagara River sample which had been
supplemented with AAP (-P) medium, and whose growth was
quantitated by cell counts, as in the particulate P bio-
assays. The spikes were quantitatively recovered in this
test, as there was no particulate matter in the test flasks.
91
-------
Table 22.
RECOVERY OF ORTHOPHOSPHATE SPIKES, ADDED TO
AAP(-P) MEDIUM CONTAINING PARTICULATE MATTER,
IN SELENASTRUM GROWTH BIOASSAYS
PP cone.
Sample No. in cultures In
(ygP/1)
No. 50- 79
autoclaved
Niagara R.
particles
mean values x-, =
difference: (x~ -
Apparent Available Pa
sample In sample
(ygP/1)
46
43
46
51
41
45 x2 =
x ) = 38 ygP/1
% of added P recovered: 38/50 x 100
No. 51- 115
autoclaved
Genesee R.
particles
mean values x, =~
52
51
45
43
44
~4~T~ x =
' * " o
+50 ygP/1 spike
88
78
82
85
83
= 76%
__
91
93
85
90
f-1
(x2
of added P recovered:
difference:
= ^3 ygP/l
43/50 x 100 = 86%
Comparison Test: Sample No. 56 (Niagara R.), Autoclaved,
Filtered; and Supplemented with AAP(-P)
Nutrients before Assay
Apparent Available
In Sample
7
7
8
8
5
mean values x, = 7
difference: (x? - xn ) =
% of added P recovered:
In sample +
(ygP/1)
X2
25 ygP/1
Pa
25 ygP/1 spike
31
34
30
34
32
32
25/25 x 100 - 100%
All test cultures were quantitated by comparing cell counts
of sample cultures to standard curves of cell count vs. P
concentration in AAP standard cultures.
92
-------
ALGAL ASSAYS OF SOLUBLE PHOSPHORUS
For estimation of the potentially available fraction
of TP in river waters, unfiltered samples were first auto-
claved for 15 minutes at 15 psi, then were cooled and
filtered through 0.4-5 micron pore size membrane filters.
If necessary, carbon dioxide was bubbled through the
filtrates to bring their pH to 7-8.
Twenty ml aliquots of filtrate from autoclaved river
water, containing less than 200 ygP/1 of TSP, were pipetted
into five replicate 50 ml culture flasks. Five ml of
freshly prepared 5X AAP(-P) medium was added to each flask
to give 25 ml of solution with IX AAP(-P) plus the nutrients
present in the sample water. Phosphorus was shown to be
limiting to the growth of Selenastrum in AAP medium in the
range of 0 to 200 ygP/1 (Figure 6). Therefore, phosphorus
had to be limiting in this range in a medium containing
AAP(-P) nutrients plus the nutrients in the sample water.
One ml of glass distilled water and 1 ml of inoculum cell
suspension was added to each flask to give a final volume
of 27 ml. Standard and blank flasks for soluble P assays
were identical to those for particulate P assays, and the
same culture conditions were used for both procedures.
For the assay of soluble P in urban runoff samples,
two dilutions were used. Twenty ml of filtered runoff was
enriched with 5 ml of 5X AAP(-P) medium, or 5 ml of filtered
runoff was enriched with 20 ml of 5/4X AAP(-P) medium. In
either case, 1 ml of glass distilled water and 1 ml of
inoculum cell suspension was added to give a final volume
of 27 ml.
Following the 18-day growth period, the cultures were
shaken and their absorbance (A), at 750 nm wavelength was
measured in a 1-inch Spectronic 20 cell. Figure 7
demonstrates the optical properties of this quantitation
93
-------
Figure 7. Absorbance vs. dry weight of S. caphcornutum
50 100 150 200
S.CAPRICORNUTUM (mg/i)
250
94
-------
method. Two different stock Selenastrum cultures were
quantitated by dry weight measurements, and various dilu-
tions and concentrations of the cultures were quantitated
by A75Q measurement. The A75Q values were plotted versus
dry weight values computed from the dry weight concentra-
tions of the two stock cultures. There was a linear
relationship (dashed lines in Fig. 7) between dry weight
and absorbance in the two tests shown, up to about A = 0.350
At higher absorbance values, there was a negative deviation
from linearity, as the absorbance per unit of algae
decreased.
As demonstrated by the standard curve of A7D.Q (18-day
old cultures) versus phosphorus concentration in Figure
8, the apparent final growth levels for high values of P
did not follow the same linear relationship as did the
growth at lower levels of P. The observed negative devia-
tion from the initial linear function may have been
partially due to the optics of culture measurement in the
Spectronic 20 (see Figure 7) , where dense cultures show
less absorbance per unit weight of algae than do dilute
cultures. In addition, some physical growth factor such
as light may have limited the growth at high population
densities, where self-shading of cells becomes important.
The absence of a well-defined plateau above 200 ygP/1
indicates that P may still have been a limiting nutrient
above that P level.
Each standard curve was constructed with average Ajcn
values from five replicate AAP standard cultures at each P
level. Minimum detectable values of available P were
calculated as shown in Appendix A.
In the bioassays of particulate matter in AAP(-P)
medium, an assumption was made that the particles acted
only as sources or sinks of available P and did not
95
-------
Figure 8. Standard curve for absorbance of S.capricornutum after ISdays
vs. ort ho phosphate concentration (0-600/zgP/l)
0.4
e
c
O
to
f-
UJ
O
CO
a:
o
CO
m
0.3
0.2
O.I
100 200 300 400
ORTHOPHOSPHATE./igP/l
500
600
-------
significantly alter the composition of the algal growth
medium, so that comparisons between particulate suspension in
AAP C-P) medium and standard AAP medium cultures containing
orthophosphate could be made. In the case of the bioassay
of filtered natural water samples, however, supplementation
of the filtrates with AAP (-P) nutrients created a new
medium, containing both AAP (-P) nutrients and the nutrients
originally present in the filtrate. Although both the
supplemented sample medium and the AAP standard medium may
have been limited by phosphorus, their response curves
(growth as indicated by A7t-n vs P) may not have been the
same shape, as illustrated in Figure 9 . In the hypothe-
tical case shown in the figure, comparison of the sample
culture growth to the AAP standard curve (dotted lines)
would result in an underestimation of the available P in
the sample.
An actual partial sample response curve was drawn by
adding orthophosphate spikes to Sample No. 52 (autoclaved,
filtered, and supplemented with AAP (-P)), as shown in
Figure 10 Since the phosphorus level in the sample was
fairly high, no information about the lower portion of the
response curve was given by this experiment. 'However, the
upper side of the curve was different from the upper portion
of the AAP standard curve, so the curves were probably
not alike at lower values of P. The unspiked sample
showed an apparent available P value of 76 ugP/1 when its
average A7,-n value was compared to the AAP standard curve
(dotted lines). This value was an overestimate, since the
TSP in the culture flasks was measured as 54- ygP/1 before
the bioassay. Likewise, the sample with a 25 ugP/1 spike
showed a P value of 51 ygP/1 above the unspiked sample,
an overestimate of the spike by 26 ygP/1.
97
-------
Figure 9. Hypothetical case for underestimation of available P in a filtered
sample bioassay.
0.5
0.4
6
c
O
10
UJ
O
z
<
QQ
o:
o
en
00
0.3
0.2
O.I
OBSERVED GROWTH
0
AAP STANDARD CURVE
SAMPLE +AAP(-P)CURVE
(HYPOTHETICAL)
-AVAILABLE P ESTIMATED FROM AAP CURVE
ACTUAL AVAILABLE P
i i i i
100 200 300 400
ORTHOPHOSPHATE,/igP/l
500
600
-------
ID
IO
Figure 10. Standard curves for absorbance of S. Capricornutum vs.
orthophosphate concentration in AAP medium or in spikes
added to sample #52 (Oswego River)water(Autoclaved,Filtered)
0.4 -
OO-O AAP STANDARDS
STANDARD SPIKES ADDED TO SAMPLE
#52 ( 20/27 DILUTION FOR ALL)
APPARENT AVAILABLE P
I
0 100 200 300
ORTHOPHOSPHATE.^gP/l CONCENTRATION IN AAP MEDIUM OR
ADDED TO #52 SAMPLE
-------
In order to determine whether the optical properties
of 18-day old Selenastrum cultures grown in sample water
supplemented with AAP(-P) and in AAP medium were different,
correlation curves relating A7l-n and cell counts were
drawn. Figure 11 shows the correlation found for AAP
o
standard cultures. A least-squares slope of 5.6 X 10
absorbance units/cell count unit was obtained, with a
correlation coefficient of r=0.92. In contrast, a series
of Niagara, Genesee, and Oswego River samples showed a
8
slope of 8.0 X 10 absorbance units per cell count unit,
with a correlation coefficient of r=0.98 (Figure 12).
The ratio of the sample to standard slopes (8.0/5.6=
1.43) indicated that for a given cell count, the net A?gQ
for a sample culture would be about 1.43 times the net
A7rn from an AAP standard culture with the same final cell
count. When the net mean A7C-n value for Sample No. 52
(Figure 6.8) was divided by the slope ratio of 1.43, a
corrected net mean An en value was obtained.
Mean observed A75Q, sample No. 52 = 0.151
Mean blank culture A 5Q = 0.010
Net mean A75Q, sample No. 52 = 0.151-0.010 = 0.141
Corrected net mean A75Q, sample No. 52 = 0.141/1.43
= 0.099
Addition of the blank A?5 value to the corrected sample
A'? 50 value gave an absorbance value which was then com-
pared to an AAP standard curve of A7t.n versus P
/ 0 U
concentration:
Corrected net mean Ay + blank A = 0.099 + 0.010
= 0.109 (Standard curve P value for 0.109 was 53 ygP/1)
The resulting P value was in agreement with the TSP level
of 54 ygP/1 in the culture flask prior to assay. Unfor-
tunately, the DRP level in the culture flasks was not
measured.
100
-------
Figure II. Correlation between absorbance and cell counts for
S. capricornutum in AAP medium
E
c.
O
in
LJ
o
0.3
0.2
CO
a:
o
go.i
<£
SLOPE' b = 5.6 X IO~8A750/CELLS/ml
INTERCEPT 0 = 0.009 A750
100 200 300
CELLS/ml (x I0~4 )
400
500
-------
o
t-o
0.4
E 0.3
c
O
m
o 0.2
m
oc
o
CO
m
O.I
Figure 12. Absorbance vs. cell counts for S. capricornutum grown in autoclaved,
filtered river waters supplemented with Ix AAP-P medium
SAMPLES
#51
O #47
A #52 (ASSAYED MAY 31)
#52 (ASSAYED JUNES)
D #56
+ #56 + 25^gP/l SPIKE
SLOPED = 8.0X |0"8A750/CELLS/ml
INTERCEPT: a = 0.015 A750
_j_
i
100
CELLS/ml ( X 10
200
-4 \
300
-------
The preceding method of calculation essentially trans-
formed the growth response of sample cultures to a cell
count basis of quantitation, with the assumption that a
unit of P in a sample culture produced the same concentra-
tion of cells as a unit of P in a standard AAP culture.
This assumption was tested by counting assay Samples No. 56
and No. 56 + 25 ygP/1 spike with a hemocytometer (see
Table 22); complete recovery of the spike was found in
this test. Table 23 shows a comparison between TSP and
bioassay results as measured by uncorrected A7[-n data or
by hemocytometer cell counts. Bioassay results above TSP
were considered overestimates of P availability, since TSP
represented the theoretical maximum value of available P.
In all the tests shown in Table 23 except the very
dilute Niagara River Sample No. 56, the uncorrected A,,,-.-.
data overestimated available P. Direct cell counting, in
contrast, overestimated available P in only one of the
samples (No. 51). Standard AAP calibration curves of cell
count versus P concentration were used to quantitate those
(autoclaved, filtered) samples which were counted on a
hemocytometer.
As a further test of the cell count basis 'for bioassay
calculations, the A7t-n data from spikes added to Sample
No. 52 (Figure 10 ) were corrected as described above. The
results are presented in Table 24. Recoveries of 116 and
92 percent were found for 25 and 50 ugP/1 spikes, respect-
ively, while higher spike levels were recovered less
completely. The low recoveries at high spike levels was
probably due to growth limitation by other factors. For
other samples, the use of corrected absorbance values
yielded P recoveries about 12 percent or less from the
expected recovery (Table 25).
103
-------
Table 23.
COMPARISON OF TSP IN CULTURE FLASKS AND
BIOASSAY RESULTS, AS COMPUTED FROM DIRECT
CELL COUNTS AND FROM UNCORRECTED A?50 DATA
TSP in Mean Apparent Available P from:
H Culture A750a
Sample Flask data
No. 56
(Niagara R.
No. 51
(Genesee R.
No. 47
(Oswego R. )
No. 52
(Oswego R. ,
No. 52
(Oswego R. ;
( iicrP/T )
13 7
83 139
53 61
54 69
Assayed May 31)
54 76
Assayed June 5)
Cell b
counts
7
92
41
50
50
Uncorrected A~5- data was compared to AAP standard curves
of A^gg vs. P concentration to find apparent available P.
Sample cell counts were compared to AAP standard curves
of cell count vs. P concentration to find apparent avail-
able P.
All samples were autoclaved, filtered, and supplemented
with AAP(-P) nutrients before assay.
104
-------
Table 24.
o
On
RECOVERY OF ORTHOPHOSPHATE SPIKES ADDED TO AUTOCLAVED, FILTERED
OSWEGO R. SAMPLE NO. 52, AS CALCULATED USING CORRECTED NET Ay50
BIOASSAY DATA
Sample
No.
No.
No.
No.
No.
52
52 -
52 -
52 -
52 -
I- 25
I- 50
l- 100
^ 200
Ay
ygP/i
ygP/i
ygP/i
ygP/i
Net
50
0.
0.
0.
0.
0.
mean a Corrected net mean Apparent
measured A,,,.., + blank An .. b Available
/ o U / b U
141
217
259
309
331
0
0
0
0
0
.109
.162
.191
.226
.241
(ygP/1)
53
82
99
122
134
P Spike Recovery
(ygP/l)(% of added P)
-
2
4
6
8
-
9
6
9
1
--
116
92
69
40
Mean A,-,,-,-, measured minus the A7t-n average of the blank cultures
Net mean A7J-n divided by 1.43
-------
Table 25.
RECOVERY OF ORTHOPHOSPHATE SPIKES ADDED TO
AUTOCLAVED, FILTERED RIVER WATERS, AS
CALCULATED USING CORRECTED NET A75Q
BIO AS SAY DATA
Net mean
A750 a
Sample measured
No. 56
(Niagara R. )
No. 56 +
25 ygP/1
No. 57
(Niagara R. )
No. 57 +
25 ygP/1
No. 58
(Genesee R. )
No. 58 +
25 ygP/1
No. 59
(Oswego R. )
No. 59 +
25 ygP/1
No. 59 (JgX)C
No. 59 (igX)C +
0.015
0.103
0.007
0.098
0.136
0.222
0.154
0.238
0.075
0.181
Corrected, net Apparent
mean A7,-n + Available
blank Ay50 P
0.014
0.076
0.009
0.073
0.100
0.160
0.122
0.172
0.057
0.131
( i i i-r
5
32
3
30
42
70
52
76
24
57
Spike
recovery
P/~l ")
27
_
27
mm mm
28
__
24
33
25 ygP/1
Average P recovery 28 ygP/1
P added 25 ygP/1
Average % of added P recovered 112%
Mean A75Q measured minus the A?[-n average of the blank
cultures
Net mean A? divided by 1.43
Autoclaved, filtered sample No. 59 was diluted 1+1 with
HO before the bioassay
106
-------
In summary, the results of bioassays with autoclaved
river waters were quantitated in two ways: 1.) with uncor-
rected A~5Q data, and 2.) with absorbance data corrected
to a cell count basis. The correlation curve in Figure 12
was made with only a few samples from the Genesee (No. 51),
Niagara (No. 56 and No. 56 + 25 ygP/1 spike), and the
Oswego (No. 47 and No. 52) Rivers. Consequently, there is
not extensive data to show that the ratio of 1.43, between
sample and standard culture net mean A-rQ values holds for
all of the samples from these sources which were bioassayed.
Nor is there any evidence that the autoclaved Black River
or filtered Madison urban runoff samples would show the
same ratio, since no samples from these sources were in-
cluded in the regression analyses outlined above. Conse-
quently, none of the Black River or filtered Madison runoff
A7rn bioassay data were corrected with the ratio method
before comparison to standard AAP curves of Anrn vs. P
concentration. In all bioassays, the criteria given in
Appendix A were used to reject outliers among the replicate
culture flasks.
As a check to insure that the culture volume and flask
size were not restrictive to algal growth due to a low
surface to volume ratio, a comparison growth test was made
with a surface volume ratio suggested by the EPA (1971).
The 27 ml culture volume/50 ml flask ratio used in the
studies reported here was compared to the 42 ml culture
volume/125 ml flask ratio suggested for standard AAP tests
(EPA, 1971). Figure 13 shows that the results of the two
procedures were similar. As noted above, the flasks were
swirled daily to provide exchange of gases between the
culture and the atmosphere.
107
-------
Figure 13. Standard curves for S. copricornutum cultured in 50ml and
125 ml flasks
0.41-
o
oo
O-O-O 27 ml CULTURE/50 ml ERLENMEYER FLASK
A-A-A 42 ml CULTURE/125 ml ERLENMEYER FLASK
100 200 300
©RTHOPHOSPHATE CONCENTRATION, fj.qP/\ IN AAP MEDIUM
-------
ALGAL ASSAYS OF UNFILTERED RAIN GAGE SAMPLES
The combined available phosphorus from soluble and
particulate forms in a rain sample was estimated by growing
Selenastrum in an unfiltered sample. Twenty ml of sample
or an aliquot of sample diluted to 20 ml was enriched with
5 ml of 5X AAP(-P) medium, and 1 ml of water plus 1 ml of
inoculum cell suspension was added to give a final volume
of 27 ml. After the 18 day growth period, the Selenastrum
was counted on a hemocytometer, and quantitated as in the
assays of particulate phosphorus.
DARK INCUBATIONS OF UNFILTERED SAMPLES
Samples of unfiltered water were stored at 15 to 22°C
in darkness for up to 50 days, in order to follow the
changes in DRP as an indicator of net inorganic phosphorus
release or immobilization in the samples. Three types of
dark incubation were compared: 1.) sample alone, 2.) sample
plus anion-exchange resin, and 3.) sample plus chloroform.
The general procedures used for these tests are described
in this section.
Samples Incubated Alone
Three acid-washed (1:1 cone. HC1:water by volume)
125 ml Erlenmeyer flasks or standard BOD bottles were
rinsed with sample and allowed to drain. One hundred ml
of well-mixed sample was then measured with a graduated
cylinder and poured into each flask. The flasks were
plugged with cotton and covered with a black plastic
shroud for storage in the algal culture room on an unlighted
shelf. Each test flask was swirled daily by hand to mix
the samples and provide adequate aeration. After approxi-
mately 13, 25, and 50 days of storage, the flasks were
sampled by filtering a well-mixed portion through a glass
fiber or membrane filter. The filtrate (20 ml) was col-
lected in a calibrated 25 X 200 mm test tube, for subsequent
DRP analysis. The remaining sample in the test flask was
109
-------
saved for further incubation if necessary -
Samples Incubated with Resin
Six flasks containing 100 ml of sample were given 1
gram of Dowex 1X8 resin per flask. Three flasks with 1
gram of resin in 100 ml of water were set up as resin
blanks. The resin was air-dried and dry-sieved as described
above for resin extraction of PP, before use in the incuba-
tions. On each scheduled sampling date, two of the flasks
with sample plus resin and one of the resin blank flasks
were analyzed by the same procedure used in the extraction
of PP by resin.
Two variations of the general sample + resin test were
attempted. In the first variation, river water (300 ml)
was mixed with 300 ml of Lake Ontario water collected near
the mouth of the river. Bottles with this mixture were
incubated for 100 days, with and without resin. Bottles
with lake water alone and with resin were carried along as
controls, so that the available P from the river water
could be calculated. On the sampling dates, the bottles
without resin were sampled as described above for "sample
alone." The contents of the bottles with resin were poured
through a No. 60 sieve and placed in a large funnel to catch
the liquid which passed through the sieve for further incu-
bation. The resin on the sieve was extracted with IN
Na^SO as outlined above. After taking an aliquot from
the liquid which had passed through the sieve (for super-
natant DRP analysis), 1 gram of fresh resin was placed in
the remaining liquid for the next period of incubation. A
volumetric correction factor was applied to each resin-
bound DRP value to account for the volume of sample which
was extracted by the resin. The cumulative resin-bound
DRP at each point during the incubation was added to the
prevailing supernatant DRP level at that point in time, to
find the cumulative "available" P as a function of time.
110
-------
In the second variation of the general procedure,
particles isolated on membrane filters were scraped into
100 ml of water taken from the lake which was expected to
receive the runoff or river water carrying the particles.
Three flasks without resin and six flasks with resin were
incubated, for the suspensions of particles and for the
lake water controls. In all other respects, the procedure
for this test was the same as that used for unfiltered
samples + resin.
Samples Incubated with Chloroform
One ml of reagent grade chloroform (CHC1~) was added
to 100 ml of unfiltered sample of river water in an acid-
washed test flask. The concentration of chloroform was in
excess of the saturation concentration, as evidenced by a
bubble of chloroform in the test flasks. Three replicate
flasks were run for each sample, under conditions of
temperature and darkness already specified for samples
without chloroform. The bottles were capped with aluminum
foil and were sampled for DRP after approximately 1, 7, and
14 days of storage, by removal of a well-mixed 20 ml aliquot
for filtration. Additional one-mi portions of chloroform
were added as necessary to maintain saturation during the
dark incubations.
In some of the incubations, the presence of phosphatase
activity in the samples was determined by addition of 100 or
200 UgP/1 as sodium tripoly phosphate (TPP) to a test flask
containing 100 ml of chloroformed sample. The volume of
the TPP spike was usually only 1 ml, so no correction was
made for the small change in volume caused by the TPP
spike. Table 26 shows the expected hydrolysis of TPP
by the Murphy-Riley color reagent in the DRP analysis at
ambient room temperature (about 27 C).
Ill
-------
Table 26. HYDROLYSIS OF SODIUM TRIPOLY PHOSPHATE (TPP)
BY MURPHY-RILEY COLOR REAGENT AT 27°C
DRP in Percent of P
10,000 ygP/1 in TPP released
Time TPP solution by hydrolysis
15
30
45
19
min.
min.
min.
hr.
(ygP/1)
139
174
228
3850
1.39
1.74
2.28
38.5
In the 30-minute time period normally required for
complete color development, only 1.7 percent of the TPP
was hydrolyzed in the acidic color reagent (4 ml) plus
sample (20 ml) mixture. The hydrolysis of TPP in chloro-
formed distilled (not autoclaved) water over 14 days was
found to average 10 percent of the TPP present (Table 27)
Table 2.7. HYDROLYSIS OF 200 ygP/1 TPP IN CHLOROFORMED
DISTILLED WATER AT 20°C
Bottle No.
1
2
3
Mean values
DRP (ygP/1)
1 day
2
2
2
2
after:
7 days
12
14
11
12
14 days
32
15
13
20
No interferences from the chloroform were noted in the
analysis of DRP, since in all cases the filtrates were
undersaturated with respect to chloroform due to the
vacuum filtration. Also, standards run in the same manner
as the samples did not appear to differ from the standards
run without chloroform.
112
-------
In all dark incubations run before April 1, 1973, the
test flasks were acid washed (1:1 HC1:water) before adding
the sample, with no pretreatment of the flasks. Later
incubations were run after the glassware had been pre-
rinsed with sample water. Table 28 shows the recovery
of orthophosphate in distilled water solutions incubated
with anion-exchange resin. Three bottles with resin were
incubated, and one bottle was analyzed at each time indi-
cated. Recoveries were in the range of 92 to 98 percent,
although a recovery of 92 percent at 200 ygP/1 represented
a loss of 16 ygP/1, probably sorbed on the glass walls of
the test bottle. In a similar test (Table 29), 100 ugP/1
orthophosphate solutions in chloroformed water were
incubated for 14 days. Recoveries of 99 to 101 percent
were found in this test. In both of these tests, the glass^
ware was pre-rinsed with the test solution before incuba-
tion. Although losses from distilled water systems may
not be the same as losses from samples, these results
indicate that available P measured in long-term tests may
be underestimates of the P actually released.
The possible sorption of orthophosphate by the plastic
cubitainers used for sample transport and storage was also
tested, as shown in Table 30. At the 50 ygP/1 level,
the sorption of orthophosphate from distilled water
solution appeared to be less than 6 percent after 50 days
in a cubitainer washed either with HC1 and water, or with
water alone. Since the HCl-water method was used to clean
all cubitainers before use on sampling trips, Table 30
indicates that losses of DRP to the walls of the container
should not be significant.
113
-------
Table 28.
DARK INCUBATION OF STANDARD 200 ugP/1 ORTHO-
PHOSPHATE SOLUTIONS WITH ONE GRAM OF DOWEX
1-X8 RESIN AT 20°C
DRP on Resin (ygP/1)'
Bottle No. 1 Bottle No. 2 Bottle No. 3
(stored 13 days) (stored 27 days) (stored 50 days)
mean values
% of added P
recovered
190
190
190
95
195
190
201
195
98
185
182
184
92
Values given are replicate analyses of the Na~SO^ resin
leachate from each bottle, corrected for P recovery in
the leaching procedure (see Figure 4). Supernatant DRP
in the test bottles was undetectable.
Table 29. DARK INCUBATION OF CHLOROFORMED STANDARD
100 ygP/1 ORTHOPHOSPHATE SOLUTIONS AT 20°C
Bottle No.
1
2
3
4
Initial
100
100
100
100
DRP Concentration
(ygP/1)
+CHC1 7 days +CHC1_ 14
o o
100
97
99
99
101
101
101
99
days
114
-------
Table 30. EFFECT OF STORAGE TIME IN ONE-GALLON
CUBITAINERS ON THE DRP CONCENTRATIONS
OF DISTILLED WATER SOLUTIONS
Test A: Cubitainer rinsed with 6N HC1, followed by six
rinses with distilled water before storage
Test B: Cubitainer rinsed six times with distilled water
before storage
Both tests: Volume of solution = 2.5 1, stored at 4 in
darkness
Time of
Storage
(Days)
0
1
9
13
42
DRP Concentration
Test A
51
51
51
50
50
Test B
52
52
53
50
49
aMean of duplicate determinations
A general summary of the procedures used in the hand-
ling and analysis of samples is given in Figure 14. Not
all of the unit processes shown were attempted for each
sample, because of difficulties in scheduling tests on
samples arriving at random times in the laboratory. How-
ever, an attempt was made to use at least two different
procedures for estimates of available P in particulate
matter or in unfiltered river water samples.
115
-------
Figure 14. Procedures in handling and analysis of samples
PARTICULATE
MATTER
STORAGE AT 4°C IN:
FIELD
STORAGE
AT
4°C
MILLIPORE //~2-
FILTRATION/ /^y
|
-------
SECTION VII
RESULTS
In this chapter the results of chemical and biological
tests of phosphorus availability are summarized. The data
have been divided into four major sections, based on the
origin of the water samples:
Madison Urban Runoff
Madison and New York Precipitation
Genesee River Basin Samples
New York River Samples
The Madison urban runoff samples were studied for the pur-
pose of developing the techniques necessary for analyzing
samples collected from New York rivers and runoff in con-
nection with the International Field Year on the Great
Lakes (IFYGL). In addition, the results of the Madison
runoff studies were intended for comparison with the
results from the urban runoff samples collected in the
Genesee River basin.
Within each major section, the phosphorus forms in the
samples are reported first, followed by the results of PP,
TSP, and/or TP availability tests. Since the primary
objective of these tests was to determine the fraction of
these forms which might be available to algae, the results
of all the tests are reported in relative terms, as per-
centages of the PP, TSP, and/or TP in the samples.
117
-------
MADISON URBAN RUNOFF
Phosphorus Forms
The concentrations of analytically defined phosphorus
forms in Madison urban runoff samples are given in Table .31.
Total phosphorus (TP) levels of 59 to 2930 ygP/1 were found
in the samples, and particulate phosphorus (PP) forms repre-
sented 13 to 97 percent of sample TP. Since these samples
were taken as grab samples during runoff events, the con-
centrations in Table 31 should not be taken as representa-
tive of average concentrations in runoff from a particular
land usage or time of year. Rather, the data were obtained
for investigation of the availability of soluble or
particulate P in representative samples.
Chemical Extractions of PP
The concentrations of inorganic P extracted from parti-
culate forms (extractable PP.) by acid, base, or anion-
exchange resin are presented in Table 32. By comparison
with the total PP concentration in each sample, the PP.
extracted by each reagent was expressed as a percent of
sample PP. All samples from a given land use class (see
Table 9) were grouped together in Table 32, and a mean
value was computed for each group and extraction method.
These group mean values are displayed graphically in
Figure 15. The ranges shown for each group and extraction
method were taken from the data in columns six through eight
of Table 32.
An important feature of Figure 15 is the relative
similarity seen between the various land uses, 'for any
given chemical extraction method. For example, the group
mean values for acid extractable PP. all fall within a range
of 33 to i|6 percent of PP. The corresponding range for base
extractions is 22 to 27 percent, and for resin extractions
13 to 17 percent. The relative order of extraction yield
118
-------
Figure 15. Percent of PP extracted from Madison urban runoff particles by
chemical methods
100 r
90 -
MEAN AND RANGE, % OF PP EXTRACTED
a 80
UJ
i
rr
i
X 60
UJ
2: so
^
0 40
^^
y 30
cc
LU
Q- 20
10
o
-
L
_
\
. <
-
A -ACID J
T EXTRACTION 1
_L J
j
1
i
§ T
_
L
i
.- BASE JL- RESIN
EXTRACTION V EXTRACTION
-L
^
i
i
I
Y
1 T
9
1
_^
/
k
<
^
^
\
(
.
)
\
i
Rl R2 R5 R6 C4 CONSTRUCTION
(UW CAMPUS) SITES
URBAN LAND USE CLASSES
-------
Table 31- PHOSPHORUS FORMS IN MADISON URBAN
RUNOFF SAMPLES
Phosphorus (uRP/l)a .
Sample
A-l
B-l
D-l
A- 2
B-2
D-2
E-2
A- 3
B-3
D-3
E-3
A- 4
B-M
F-4
D-5
6-5
A- 6
B-6
D-6
E-6
F-6
H-6
A- 7
B-7
D-7
E-7
F-7
H-7
A- 8
B-8
D-8
A- 9
B-9
F-9
D-10
E-10
H-10
D-ll
E-ll
H-ll
A-12
B-12
D-12
1-12
Date Collected
Aug. 11,
It
It
Aug. 19,
M
tl
II
Aug. 23,
11
II
II
Sept. 19
ti
IT
Sept. 20
M
Oct. 20,
M
11
ii
M
n
Oct. 22,
It
II
It
II
tl
Dec. 30 ?
it
IT
Jan. 17,
M
It
Jan. 18,
ti
it
Feb. 1,
M
n
Mar. 5,
tl
TT
11
1972
1972
1972
, 1972
, 1972
1972
1972
1972
1973
1973
1973
1973
DRP
80
144
94
134
162
205
37
80
112
168
41
186
240
77
219
33
748
1430
213
195
161
415
119
576
151
45
300C
74
354
410
328
868
812
486
244
33
110
252
30
110
264
347
247
30
TSP
87
163
105
143
183
230
45
85
119
180
45
195
260
78
251
41
899
1640
418
202
202
451
120
580
167
46
280
77
366
435
352
979
900
552
272
41
133
257
34
115
275
353
257
46
TP
201
290
291
203
282
353
59
118
233
679
74
283
438
2930
433
131
1030
2060
711
249
1090
539
172
735
252
73
645
146
589
699
696 ,
1180°
1110°
755H
740^
136^
299d
698
179
377
489
774
818
1460
ppD
114
127
186
60
99
123
14
33
114
499
29
88
178
2850
182
90
130
420
293
47
890
88
52
155
85
27
365
69
223
264
344
203e
210
203e
468
95e
166e
441
145
262
214
421
561
1410
bMean values from duplicate or triplicate analyses
Unless otherwise noted, PP was determined by calculation
^ J..JT ~ J. o L '/
cThis value of DRP was higher than the TSP, hence probably
,in error; the TSP value given was used to compute PP
Determined by calculation from direct PP analysis (PP + TSP)
Q
Determined directly on (membrane) nonfiltrable particles
120
-------
Table 32.
EXTRACTION OF MADISON URBAN RUNOFF PARTICLES
WITH ACID, BASE, AND ANION-EXCHANGE RESIN
Sample
A-l
A- 2
A- 3
A- 4
A- 6
A- 7
A- 8
A- 9
A-12
B-l
B-2
B-3
B-4
B-6
B-7
B-8
B-9
B-12
D-l
D-2
D-3
D-5
D-6
D-7
D-8
D-10
D-ll
D-12
H-6
H-7
H-10
H-ll
PP
(ugP/i
114
60
33
88
130
52
223
203
214
127
99
114
178
420
155
264
210
421
186
123
499
182
293
85
344
468
441
561
88
69
166
262
PP.
) Aci
31.
62
16
121
75.
55.
59.
72
41
104
98.
201
U.
__
39
105
11
115
186
314
27
15
__
150
CugP/l)Extracteda PP.(%
d Base Resin Acid
Residential
31
20
5.8
8 19.8
38
11
80
1 56
1
Group Mean
Residential
34
26
27
5 47.1
113
39
60
6 48
--
Group Mean
(Rl)
16
1
6
8.8
21
2
49
29
39
Values
(R2)
35
11
23
34
54
19
51
32
72
Values
__
__
36
48
30
54
37
li
38
34
17
26
40
47
48
35
of PP)Extracted
Base
27
33
17
22
30
21
36
28
--
27
27
26
24
26
27
25
23
23
25
Resin
14
2
17
10
16
4
22
15
18_
13
28
11
20
19
13
12
19
15
17_
17
of Wis . Campus (R5)
40
26
129
38
70
8.0
98
105
--
Group Mean
Residential
22
13
78.9
Group Mean
16
27
79
22
33
6
52
54.1
89
109
Values
(R6)
16
3
26.7
53
Values
21
35
13
33
--
42
56
33
26
21
il
35
22
21
26
21
24
9
28
24
22
26
18
__
19.
25
9
22
16
12
12
7
15
12
20
2_0
16
18
5
16
19.
15
121
-------
Table 32. EXTRACTION OF MADISON URBAN RUNOFF PARTICLES
WITH ACID, BASE, AND ANION-EXCHANGE RESIN
Sample
E-2
E-3
G-5
E-6
E-7
E-10
E-ll
F-4
F-6
F-7
1-12
F-9b
PP PP. (ygP/l)Extracteda
(ygP/1) Acid Base Resin
14
29
90
47
27
95
145
2850
890
365
1410
203
37
13
5
87
1580
357
124
805
94
Commercial
5.1
14
.9 16
9.7
.5 4.3
.4 35
Group Mean
Construction
155
262
137
--
Group Mean
.8 67
(C4)
3
6.0
12
4.5
3
14
23
Values
Sites
301
148
76
188
Values
31
PP. (%
Acid
42
28
21
12
38
55
40
34
_57_
46
47
of PP)Extracted
Base
37
49
18
20
16
24
27
6
30
38
25
33
Resin
21
21
13
10
10
15
16_
15
11
17
21
13
16
16
of triplicate determinations; the A-6 base extract value
is the mean value of duplicate determinations.
This sample was collected from station F after the con-
struction site had been sodded to prevent erosion.
122
-------
for the three methods was acid > base > resin, as shown by
the group mean values in Figure 15,
The data from Sample F-9 were not included in the
previously described averaging processes, as this sample
was not readily classified into one of the land use classes
shown. Sample F-9 was collected from a construction site
which had been covered with sod before sample collection.
In comparison, samples F-U, F-6, and F-7 were collected from
the same site while it was "open" and seriously eroding.
The changed character of the site was indicated by the
relatively low concentration of PP in Sample F-9 (Table 31) .
However, the proportions of extractable PP. were not signi-
ficantly different between Sample F-9 and Samples F-4, F-6,
and F-7.
Extraction of PP by Selenastrum
As a check on the chemical extraction results, several
bioassays with Selenastrum were run on Madison urban runoff
particles in AAP (-P) medium. The results of these assays
are given in Table 33. Taken as a general group, the resi-
dential samples (R1-R6 zoning classes) exhibited PP avail-
ability of 23 to H5 percent. This is a range of sample
means, from the bioassay data compiled in Tables B.I to B.5,
Appendix B. Unfortunately, not enough bioassay data were
collected from the other land use classes for computation
of their group availability ranges. In the commercial
sample (E-ll) and open construction site sample (1-12), the
proportion of PP extracted by Selenastrum appeared to be
lower than in the residential Samples. In contrast, sample
F-9 from the "covered" construction site exhibited the
highest proportion of PP available (55 percent) of all the
samples tested.
123
-------
Table 33, EXTRACTION OF MADISON URBAN RUNOFF PARTICLES
BY SELENASTRUM IN ALGAL BIOASSAYS
Sample
A- 8
A- 9
A-12
B-8
B-9
B-12
PP
Assayed
(ygP/1)
201b
185
179
130b
97
189
Apparent Available
Mean Valuec
(ygP/1)
Residential
60
65
80
Residential
37
39
59
P
Std. deviation Mean Value
(ygP/1) (
(Rl)
3
8
11
(R2)
5
4
6
% of PP)
30
35
45
28
40
31
U. of Wis. Campus (R5)
D-8
D-ll
D-12
H-ll
E-ll
1-12
F-9d
157b
209
121
469
254
356
94
43
49
31
Residential
142
Commercial
19
Construction
58
52
3
8
2
(R6)
6
(C4)
6
Sites
5
5
27
23
26
30
8
16
55
Unless otherwise noted, all PP values were determined
directly on aliquots of the assayed suspensions.
These values were calculated by difference (TP-TSP) as
shown in Table B.2 of Appendix B,
(2
Mean values of four to six replicate culture flasks.
This sample was collected after the construction site had
been sodded to prevent erosion.
124
-------
Incubation of Runoff Particles in Receiving Waters
In an effort to describe the effect of runoff particles
on the phosphorus chemistry of their receiving waters, a
short-term dark incubation of (membrane) nonfiltrable parti-
cles was made in either Lake Mendota or Monona water. Anion-
exchange resin was added to some of the test flasks in order
to study the behavior of the particles in lake waters with
low DRP concentrations.
Table 34- shows the DRP changes seen during the dark
incubations. In flasks with resin, the DRP value reported
in the table is the "resin-extractable" DRP (R-DRP), equal
to the sum of the DRP bound to the resin and the DRP in the
supernatant solution above the resin and natural particles.
The net mean DRP contributed by the particles was obtained
by subtraction of the appropriate mean lake water control
DRP value from the mean DRP results in flasks with parti-
cles plus lake water. These net values are given in
Table 35. In flasks without resin, the net DRP contribu-
tions were positive for sample D-ll particles but negative
for particles from samples E-ll or H-ll. Comparison tests
with anion-exchange resin showed positive net DRP contribu-
tions from the particles to the resin. In the latter tests,
the resin lowered the concentration of DRP in the receiving
waters from 66 and 80 ugP/1 in Lake Mendota and Monona
waters, respectively, down to about 2 ygP/1. Comparison of
the maximum net mean R-DRP values seen in these tests with
the PP concentrations in the flasks (Table 35) gave the
following percents of PP contributed by the particles:
Maximum percent of PP
Sample particles extractable by resin
D-ll 21
E-ll 8
H-ll 10
125
-------
Table 34, DRP CHANGES DURING DARK INCUBATION OF RUNOFF
PARTICLES IN LAKE WATER
Test sample
L. Mendota water
100 ml L. Mendota
+ particles from
D-ll
L. Monona water
Initial
DRPa
(ugP/D
66
mean values
water 66
25 ml
mean values
80
mean values
100 ml L. Monona water 80
+ particles from
E-ll
100 ml
mean values
100 ml L. Monona water
+ particles from
H-ll
100 ml
mean values
DRP
Sample
13 days
82
87
il
84
88
90
ii
90
96
103
100
100
69
70
Zi
70
63
67
65
65
(ugP/D after
only
26 days
88
88
ii
88
94
98
99^
97
96
104
104
101
79
82
81
81
86
87
89
87
incubation*3
Sample +
13 days
74
75
Zi
74
98
102
96
99
97
94
94_
97
105
103
105
104
122
115
115
117
resin
26 days
77
78
80.
78
101
98
97_
99
94
95
ii
96
108
105
112
108
126
138
126
130
aDark Incubation was begun February 7, 1973.
DRP in flasks with resin includes DRP in solution plus DRP on the resin.
-------
Table 35. SUMMARY OF NET MEAN DRP RELEASED FROM RUNOFF
PARTICLES TO LAKE WATER (See Table 34)
Runoff sample
particles
in
in
in
D-ll
L. Mendota
water
E-ll
L. Monona
water
H-ll
L. Monona
water
PP
from runoff
(ygP/1)
116
118
119
Mean 118
162
159
153
Mean 158
335
322
330
Mean 329
Net Mean DRP (ygP/1) from Particles
Sample only Sample + resin
13 days 26 days 13 days 26 days
6 9 25 21
-30 -20 7 12
-35 -14 20 34
Calculated for a given incubation time and treatment + resin by subtracting mean
DRP values for lake water samples from mean DRP values from samples with lake
water + particles.
-------
The relationship between orthophosphate, resin, and
runoff was also investigated by spiking sample D-ll with
orthophosphate, as shown in Table 36. The samples were
shaken for 24 hours after spiking, then were analyzed for
DRP. The samples, including the unspiked control, demon-
strated phosphorus uptake (Table 36-a ) of 16 to 41 ygP/1.
To see if the phosphorus lost to the particles could be
recovered with anion-exchange resin, one gram of resin was
added to each flask, which was then shaken continuously for
another 24 hours. The total DRP recovered by the resin is
given in part (b) of Table 36, Comparison of these values
with the expected DRP concentrations in the flasks after
spiking (Table 36 -c) showed recoveries of 100 to 111
percent.
Dark Incubations of Unfiltered Runoff
Samples of unfiltered runoff were stored in darkness
for up to 50 days, with anion-exchange resin added to some
of the samples as a sink for any phosphorus released from
the runoff particles. The data from these tests are com-
piled in Tables C.I to C.6 of Appendix C. Comparison of
the resin-extractable DRP (R-DRP) values to the DRP values
in flasks without resin generally showed the R-DRP levels to
be approximately equal to or greater than the DRP levels in
flasks without resin. In the resin flasks, the R-DRP values
at 50 days were generally close to the 25-day incubation
values in the early tests. Consequently, later tests were
carried out for only about 26 days of incubation.
Table 37 lists the maximum observed R-DRP values for
each sample incubated. As an estimate of the contribution
of inorganic P to the resin from the particulate matter in
the samples, the following equation (1) was employed:
128
-------
Table 36. RECOVERY OF INORGANIC P FROM SPIKED URBAN
RUNOFF SAMPLE D-ll BY ANION-EXCHANGE RESIN
EXTRACTION
A. DRP 24 hours after spiking 100 nl of sample D-ll
(Initial sample DRP =212 ygP/1)
IDRP (ugP/1)
Flask code Added in spike Expected cone. Observed cone.
A 0 212 191
B 50 262 246
C 98 310 281
D 192 404 363
B. DRP after equilibration of 80 ml of spiked sample with
anion-exchange resin for 24 hours
DRP (ygP/1)
Flask code
A
B
C
D
In solution
8
11
12
14
On resin
205
269
332
414
Total recovered
213
280
344
428
C. Recovery of expected DRP by resin equilibration
Flask code
A
B
C
D
DRP (ygP/
Expected
212
262
310
404
1)
Recovered
213
280
344
428
%
Recovered
100
107
111
106
aEach code letter represents four replicate test flasks:
all DRP values are mean values of the four flasks.
129
-------
Table 37.
MAXIMUM DRP VALUES IN TEST FLASKS WITH ANION-
EXCHANGE RESIN DURING DARK INCUBATIONS OF
MADISON URBAN RUNOFF
Initial
concen-
tration
Sample
A- 8
A- 9
A-12
B-4
B-7
B-8
B-9
B-12
DRP
TSP
(ugP/1)
354 366
868 979
264 275
240 260
576 580
410 435
812 900
347 353
Incuba-
tion
period
Max.obs.
R-DRPa in
incubation
(days) (ygP/1)
Residential (Rl
50 410
50 970
21 311
Residential (R2
50
50
25
25
21
318
648
467
883
392
Inorganic P released
from PP
(Max. observed values)
(ygP/1)
)
44
-9
36
)
58
68
32
-17
39
(% of PP)e
20
-4
17
33
44
12
-8
9
U. of Wis. Campus (R5)
D-8
D-10
D-ll
D-12
H-ll
E-ll
F-7
F-9d
1-12
aR-DRP =
sample
328
244
252
247
110
30
300
486
30
DRP
352
272
257
257
115
34
280
552
46
in
50 422
50 342
26 316
21 336
Residential (R6)
26 105
Commercial (C4)
13 36
Construction Sites
C 50 395
50 556
21 142
solution plus DRP bound
70
70
59
79
-10
2
115
4
96
to resin
20
15
13
14
-4
1
31
2
7
ygP/1 of
^Calculated by: R-DRP - Initial sample TSP
Used in calculation of DRP from PP, although it is a lower
dvalue than DRP &
This sample was.collected after the construction was sodded
to prevent erosion
PP values used to calculate these data are given in
Table 31.
130
-------
Estimated inorganic _ Maximum observed Initial Sample
P released from PP R-DRP TSP
Equation (1) is only an estimate because the difference
between the initial sample DRP and TSP concentrations (dis-
solved unreactive-P) cannot be assumed to be converted
completely to DRP by hydrolysis during the incubation, nor
can the particles be assumed to contribute DRP equal to the
entire difference between the maximum R-DRP observed and the
initial sample DRP. However, it can be stated that at_ least
the difference between the maximum observed R-DRP and the
sample TSP had to come from the runoff particles. The last
column in Table 37 expresses the estimated contributions
from the particles in terms of percent of PP. The data show
that in some samples (A-9, B-9, F-9, H-ll, and E-ll) the
maximum R-DRP levels were similar to the initial TSP levels
in the samples (values from Table 31 ). For these samples,
the necessity of invoking a possible contribution of
inorganic P from PP was unnecessary. In other samples,
however, R-DRP exceeded TSP, and calculated values of 12 to
44 percent of PP were found for the estimated contribution
of inorganic P from runoff particles.
Bioassay of Soluble P in Runoff
The results of algal bioassays of filtered runoff
samples are given in Table 38. These data are compiled
from Tables B.6 to B.10 of Appendix B and are based on
uncorrected A7[-n data. Table 38 shows that the urban
runoff filtrates generally showed high relative TSP
availability. The results were compared to sample TSP,
since TSP represented the maximum possible available P
level. In several residential samples, the apparent per-
cent of TSP available to Selenastrum exceeded 100 percent.
One such sample, B-9 showed 206 percent of TSP available.
Spikes of 50 and 100 ygP/1 were added to this sample,
131
-------
Table 38. BIOASSAY OF FILTERED URBAN RUNOFF SAMPLES
WITH SELENASTRUM
P in assay Apparent algal-available
flasks _ in assay flasks
Sample
A- 8
A- 9
A-12
B-9a
B-9b
B-12
D-8
D-10
D-ll
D-12
H-10
H-ll
E-10
E-ll
DRP
(UgP/
66
161
49
150
64
61
45
47
46
81
20
24
22
TSP
1)
68
181
51
167
167
65
U.
65
50
48
48
98
21
30
25
Mean value
(ygP/1)
Residential (Rl)
73
>200
84
Residential (R2)
>200
343
116
Std.dev.
(ygP/1)
7
--
10
10
2
Mean Value
(% of TSP
107
>110
165
>120
206
179
of Wis. Campus (R5)
74
57
47
71
Residential (R6)
46
12
Commercial (C4)
5
7
6
1
4
7
4
3
1
3
114
114
98
148
47
57
17
28
Construction Sites
1-12
F-9C
22
90
34
102
12
139
1
6
35
136
aAssayed Jan. 19, 1973
Assayed Mar. 6, 1963
c
This sample was collected after the construction site had
been sodded to prevent erosion.
Mean values based on uncorrected A^r.-. data
132
-------
producing the response curve shown in Figure 16. The
growth response with the 50 ygP/1 spike indicated that
phosphorus was still limiting growth, although the next
50 ygP/1 increment clearly showed a smaller growth response
per unit of added P (see 100 ygP/1 spike data point). If
the response curve for the spiked sample is transposed to
the chemically determined value of TSP (167 ygP/1) and com-
pared to the standard curve drawn from AAP standard cultures,
the reason for the overestimation of available P in the
sample becomes readily apparent. The absorbances of the
samples (which were spiked with AAP(-P) nutrients) were
higher than the AAP standard curve shown in the figure.
Thus, the apparent available P (arrows) was much higher
than even the theoretical upper limit, TSP.
Of the 15 different samples tested, 10 showed available
P of at least 98 percent of TSP. The five remaining samples
had available P concentrations of less than 60 percent of
TSP; their concentrations of available P were even less
than the chemically measured DRP levels. Four of the
samples with low soluble P availability came from commercial
or high-density residential areas and one was from an open
construction site.
PRECIPITATION SAMPLES
Madison Snow
Table 39 (a) shows the concentrations of P forms found
in three samples of Madison snow collected April 10, 1973.
DRP and TSP were essentially identical in these samples, and
PP forms were clearly dominant. By filtering large volumes
of melted snow, PP concentrations of 82 to 131 ygP/1 were
obtained for Selenastrum growth bioassays (Table B.ll,
Appendix B. As shown in Table 39 (b), less than 25 percent
of the PP in the snow samples was available to Selenastrum
in 18 days. The sample collected from the roof of the
133
-------
Figure 16. Bioossay of soluble P in Madison urban runoff sample #B-9
E
c
O
UJ
O
z
-*
ABSORB/
0.6
0.5
i
0.4
0.3
0.2
O.I
CURVE TRANSPOSED
TO TSP- 167/igP/l
O-O-O AAP STANDARD CURVE
A-A-A SAMPLE B-9 -l-AAP(-P)
NUTRIENTS + ORTHO-P
SPIKES
APPARENT AVAILABLE P
CONCENTRATION
0 100 200 300
ORTHOPHOSPHATE,
400
-------
Table 39. PHOSPHORUS FORMS AND ALGAL-AVAILABLE PP IN
MADISON SNOW SAMPLES COLLECTED APRIL 10, 1973
A. Phosphorus Forms
Sampling Phosphorus
Sample site DRP
1. Picnic Point Park 3
2. City-County Bldg. 3
TSP
2
4
(ygP/1)
TP
32
37
PP
30
33
Roof
3. Near Meat Packing 3 0 44
Plant
B. Bioassay of PP Forms with Selenastrum
PP Apparent Available P
Sample Assayed Mean value Std.deviation Mean value
(ygP/1)(ygP/1)(ygP/1)(% of PP)
1. 82 14 4 17
2. 118 <2 - <2
3. 131 30 3 23
135
-------
City-County building in the commercial district of Madison
did not show detectable growth in any of the culture flasks.
New York Rain Gage Samples
A summary of the P forms in samples from open and
covered rain gages in the State of New York is given in
Table 40. Unlike urban runoff these precipitation samples
had very high TSP concentrations along with low DRP levels.
(In most urban runoff samples, TSP and DRP were similar).
Because of the small volumes of rain water and their low
concentrations of PP, unfiltered samples were bioassayed to
estimate the available fraction of TP rather than of PP.
Bioassays of the unfiltered rain water (Table B.12, Appendix
B) showed that in those samples with "abnormally" high TSP
concentrations, available P was, like DRP, only a small
fraction of TSP (Tables 4-1 and 40). In contrast, those
samples with DRP close to TSP (June 601 and 604) had avail-
able P levels only 21 ugP/1 or less lower than TSP.
On a TP basis, the data in Table 41 show that only
three samples had available P concentrations which were
10 percent or more of TP (May 601, June 601, and June 604).
Except for the June 604 sample, the available P in these
three samples was lower than even the sample DRP values.
GENESEE R. BASIN SAMPLES
Phosphorus Forms
Although many samples were received from the Genesee
R. basin in the State of New York, only those listed in
Table 42 were selected for PP availability studies. The
other samples were too dilute for such analyses. As shown
in Table 42, the concentrations of PP in the samples
studied varied from 17 to 2110 ygP/1 and from 35 to 99
percent of TP.
136
-------
Table 40. PHOSPHORUS FORMS IN NEW YORK RAIN GAGE SAMPLES
Sample location Month Phosphorus (ygP/l)a
Code No.D
601-0
602-C
603-C
604-C
605-C
606-C
601-C
602-C
603-C
604-C
605-C
606-C
608-0
Collected
May
11
it
I!
11
11
June
I!
IT
11
11
11
11
DRP
60
8
< 1
< 1
< 1
1
52
4
5
48
< 1
1
1
TSP
245
62
2
2
304
401
63
6
82
58
6
6
318
TP
420
86
4
6
350
439
72
10
106
64
9
12
346
PP
175
24
2
4
46
38
9
4
24
6
3
6
28
All values are mean values of two or three replicate
analyses except May 601, 602, 603, and 604 TSP values,
where single analyses are reported.
The sample location code number refers to the EPA sampling
site, as given in Table 10. The letters "C" and "0" refer
to the condition of the rain gages during dry periods
(Closed or Open).
137
-------
Table 41. BIOASSAY OF UNFILTERED NEW YORK RAIN GAGE
SAMPLES WITH SELENASTRUM
Sample
Sample TP
(
May-601-0
May-602-C
May-603-C
May-604-C
May-605-C
May-606-C
June-601-C
June-602-C
June-603-0
June-604-C
June-605-C
June-606-C
June-608-0
:ygP/i)
420
86
4
6
350
439
72
10
106
64
9
12
346
Apparent Available
Mean value Std. deviation
(UgP/D* (ugP/1)
42 4
4 0.5
<14
<14
<14
< 3
42 4
<14
4 1
58 11
<14
<14
<14
P
Mean value
(% of PP)
10
5
-
-
<4
<:L
58
-
4
90
-
-
<4
These values were calculated from Table B.12 of Appendix B
by multiplying the bioassay results in Table B.12 by the
sample dilution correcti.on factor of 27/20.
138
-------
Table 42
PHOSPHORUS FORMS IN GENESEE RIVER BASIN SAMPLES
Sample
402-6
402-8
404-8
407-8
409-8
402-9
409-9
502-1
502-7
507-7
502-8
504-8
507-8
507-9
502-10
502-11
507-11
501-12
502-12
507-12
501-13
507-13
501-14
502-14
507-14
Date Collected
Oct. 6, 1972
Nov. 3, 1972
Nov. 2, 1972
tt
ti
Nov. 15, 1972
Nov. 14, 1972
Dec. 15, 1972
Mar. 22, 1973
it
April 4, 1973
it
April 3, 1973
April 17, 1973
May 1, 1973
May 16, 1973
May 15, 1973
May 30, 1973
tt
tt
June 12-13, 1973
tt
June 26, 1973
it
June 25, 1973
Phosphorus (ugP/l)a
DRP
72
70
182
27
1
55
14
27
24
19
54
6
2
15
37
2
3
43
2
1
5
6
59
9
TSP
77
78
193
29
8
66
26
33
26
4
27
60
4
2
22
46
5
12
55
4
4
9
7
66
10
TP
118
188
350
361
131
112
2140
60
69
39
59
452
29
27
77
81
22
60
165
32
31
239
38
129
284
PP
41
110
157
332
123
46
2110
27
43
35
32
392
25
25
55
35
17
48
110
28
27
230
31
63
274
values of duplicate or triplicate determinations
139
-------
Chemical Extractions of PP
The concentrations of inorganic P extracted from PP
forms (extractable PP.) by acid, base, and resin are given
in Table 43. These concentrations were compared to sample
PP concentrations and expressed as a percent of PP, then
averaged as a group on the basis of sampling station.
Stations No. 1, 4, 7, and 9 showed group mean values for
acid extractable PP. in the range of 22 to 30 percent, while
the corresponding value for Station No. 2 was 48 percent.
Figure 17 shows that in general the group mean values for
the base and resin extractions of all stations other than
No. 2 were also lower than the corresponding values for
Station No. 2. For any given station, the relative yields
from the chemical extraction methods were in the order:
acid > base > resin. The ranges shown for each group and
extraction method were taken from the data in columns six
through eight of Table 43.
Extraction of PP by Selenastrum
Generally, less than 25 percent of the PP incubated
in AAP(-P) medium was available to Selenastrum in 18 days,
even if the particulate suspensions were autoclaved before
the bioassays (Table 44). Only one value over 25 percent
was recorded (34 percent), in a bioassay run after auto-
claving the particles from Sample No. 502-14. These bioassay
data are compiled from Tables B.13 to B.28, Appendix B .
Dark Incubations of Unfiltered Samples
Table 45 shows the data collected from very limited
dark incubations of Genesee R. basin samples, as summarized
from Tables C.2 to C.12, Appendix G. The R-DRP values in
all three of the samples tested were greater than the sample
TSP values. Calculated estimates of the inorganic P con-
tribution from PP ranged from 7 to 28 percent of PP.
140
-------
Figure 17. Percent of PP extracted from Genesee R. basin particles by chemical methods
100 p
90
080
UJ
0 70
a:
x 60
UJ
Q_
0_ 50
u.
o
40
1-
§30
(£.
UJ
0- 20
10
MEAN AND RANGE, % OF PP EXTRACTED
i ACID I BASE 5 RESIN
fC EXTRACTION 1 " EXTRACTION I" EXTRACTION
-
I
T
"
i
NO. 1
-j-
1 I
1 1
1 i
1 1
| |
1 1
I 1
i 1
i i
1 1
- !* T 11
~*\ |T! 5
I x Ji Q
i i i
NO 2 NO. 4 NO. 7 NO. 9
(BYRON) (ROCHESTER EAST) (GENESEO) (DANSVILLE) (ANDOVER)
SAMPLING STATIONS
-------
Table 43.
EXTRACTION OF GENESEE R. BASIN SAMPLE PARTICLES
WITH ACID, BASE, AND ANION-EXCHANGE RESIN
Sample
501-12
501-13
501-14
402-6
402-8
402-9
502-7
502-8
502-11
502-12
502-14
404-8
504-8
407-8
507-7
507-8
507-9
507-11
507-12
507-13
507-14
409-8
409-9
PP
(ugp/D
48
27
31
41
110
46
43
32
35
110
63
157
392
332
35
25
25
17
21
230
274
123
2110
PP. (ygP/l)Extracteda PP.(% of PP) Extracted
Acid Base Resin Acid
Station No. 1 (Byron)
11.9 5.1 0.5 25
4.8 5.0 2 18
6.9 3.0 3 22_
Group Mean Values 22
Station No. 2 (Rochester East)
17.3 11 11 42
43.8 32.0 27 40
27.7 17 10 60
15.9 12
8.0 5.6
16.1 10 10 46
53.5 29.6 30 49
32.7 17 16 52
Group Mean Values 48
Station No. 4 (Genesseo)
46 29 27 30
112 75 69 29
Group Mean Values 30
Station No. 7 (Dansville)
116 60 35 35
11.8 2 34
8.0 1.9
7.3 2.6 3 29
4.8 2.7 2.8 28
6.2 5 3 29
64.5 34 31 28
82.8 28.7 15 30
Group Mean Values 30
Station No. 9 (Andover)
28.6 20 9 24
742 445 228 35
Group Mean Values 30
Base
11
18
i£
13
28
29
36
37
25
30
27
27_
30
18
19
18
18
__
32
10
16
22
15
11
18
16
21
18
Resin
1
7
11.
6
26
25
23
29
18
30
27
2_5
25
17
17_
17
11
5
8
13
16
16
13
_5_
11
7
11
9
Mean values of duplicate or triplicate determinations
142
-------
Table 44.
EXTRACTION OF GENESEE R. BASIN SAMPLE PARTICLES
BY SELENASTRUM IN ALGAL BIOASSAYS
Sample
501-12
501-13CA)
501-14
501-14(A)
502-1d
502-1C
502-7
502-8
502-10
502-11
502-12
502-14
502-14(A)
504-8
PP a
Apparent Available P
Assayed Mean value Std . deviation
(UgP/1) (
Station
146
103
52
54
Station No.
43
46
137
100
328
164
81
127
128
Station
57
ygP
No
< 4
17
< 5
11
2
9
10
30
21
79
< 2
< 6
< 2
44
No.
4
Station No.
507-7
507-8
507-11
507-12
507-13(A)
507-14
507-14(A)
61
64
68
210
224
444
481
0
2
< 2
< 3
17
12
50
/!)£ (ygP/1)
. 1 ( Byron )
« K
0.4
__
0.8
(Rochester East)
5
5
5
5
31
2
4 (Genesseo)
0.7
7 (Dansville)
_ _
1
2
2
2
Mean value
a of PP)
< 3
16
<10
20
21
22
22
21
24
< 1
< 7
< 2
34
7
< 36
3
< 3
< 1
8
3
10
dUnless otherwise noted, these values were determined
directly on aliquots of the assayed suspensions.
Mean values of four to six replicate culture flasks
CAssayed on Dec. 21, 1972
dAssayed on Jan. 9, 1973; PP was calculated by difference
(TP-TSP), as shown in Table B.2 of Appendix B.
Calculated assuming a minimum detectable P concentration
of 2 ygP/1
(A) = Autoclaved suspensions of particles
143
-------
Table 45. MAXIMUM DRP VALUES IN TEST FLASKS WITH ANION-
EXCHANGE RESIN DURING DARK INCUBATIONS OF
GENESEE R. BASIN SAMPLES
Sample
402-8
407-8
507-13
Initial
Concen-
tration
DRP TSP
(ygP/1)
70 78
27 29
5 9
Incuba-
tion
Period
(days)
50
50
50
Max.obs.
R-DRPa in
incubation
(ygP/1)
109
104
26
Inorganic P released
from PPb
(max. observed
(ygP/1) (%
31
75
17
values)
of PP)C
28
22
7
R-DRP = DRP in solution plus DRP bound to resin (ygP/1 of
sample)
^Calculated by: R-DRP - Initial sample TSP
i
'PP values were taken from Table 42 for this computation.
144
-------
NEW YORK RIVER SAMPLES
Phosphorus Forms
The phosphorus forms measured in samples of New York
tributaries to Lake Ontario and in samples of lake water
collected near the mouths of the Genesee and Oswego Rivers
are given in Table 46. PP in the samples ranged from
5 to 360 ugP/1 and from 23 to 94 percent of TP. Emphasis
was placed on the estimation of the algal-available fraction
of both PP and TP in all studies with river water samples.
Chemical Extractions of Genesee R. PP
The results of chemical extractions of Genesee R.
particles are reported in Table 47. The PP. extracted by
acid was in all cases more than the PP. extracted by base
or resin. However, the range of the percent of PP extracted
by acid was quite broad, from 21 to 79 percent. The combined
range for base and resin extractions was less broad, from
6 to 31 percent, with close agreement between the results
of the two methods. Table 48 shows the data collected
from a dark incubation of Sample No. 42 (Genesee R.) parti-
cles in Lake Ontario water of initial DRP equal to 19 ygP/1.
In test flasks with resin, a maximum net value of 18 percent
of PP was extracted by the resin after 24 days of dark
incubation. The corresponding maximum value for flasks
without resin was 14 percent of PP. Both values are
similar to the values noted for base and resin extractable
PP. in Sample No. 42 (Table 47).
Extraction of PP by Selenastrum
Membrane-nonfiltrable particles from New York rivers
were bioassayed in AAP(-P) medium both in their natural
forms or after autoclaving. Table 49 shows that natural
particles did not yield more than 6 percent of their phos-
phorus to Selenastrum cells for growth. The presence of
several species of natural algae was noted in many of the
samples during the hemocytometer cell counting of the
145
-------
Table 46.
PHOSPHORUS FORMS IN NEW YORK RIVER WATER
SAMPLES AND LAKE ONTARIO RIVER MOUTH SAMPLES
Sample
No. River
16
17
22
23
24
25
26
27
28
29
31
32
33
34
35
36
40
41
Genesee
Genesee R.
Mouth
Oswego R.
Mouth
Oswego
Oswego R.
Mouth
Black
Oswego
Niagara
(Ft. Niagara)
Oswego
Oswego
Oswego
Phosphorus (ygP/1^
Date Collected
Aug. 2, 1972
ii
Aug. 7, 1972
ii
Sept. 1, 1972
Aug. 28, 1972
I!
Feb. 26, 1973
March 2, 1973
March 12, 1973
March 28, 1973
Niagara April 6, 1973
(Beaver I. Park)
Niagara
(Ft. Niagara)
Genesee
Oswego
Black
11
April 7, 1973
it
11
Niagara April 30, 1973
(Beaver I. Park)
Niagara
(Ft. Niagara)
11
DRP
43
12
58
49
41
14
79
4
68
78
43
2
5
26
47
7
6
4
TSP
82
20
62
58
50
19
87
8
71
82
49
5
10
26
52
12
10
8
TP PP
167
43
93
96
88
53
154
18
93
106
95
30
34
386
105
34
15
22
85
23
31
38
38
34
67
10
22
24
46
25
24
360
53
22
5
14
(Continued)
146
-------
Table 46. PHOSPHORUS FORMS IN NEW YORK RIVER WATER
SAMPLES AND LAKE ONTARIO RIVER MOUTH SAMPLES
Sample
No.
42
43
44
47
49
50
51
52
53
54
55
56
57
58
59
60
River
Genesee
Oswego
Black
Oswego
Date Collected
May 1, 1973
ii
ti
May 14, 1973
Niagara May 27, 1973
(Beaver I. Park)
Niagara
(Ft. Niagara)
Genesee
Oswego
Black
Oswego
Oswego
Niagara
(Ft. Niagara)
ii
May 28, 1973
ii
ii
May 31, 1973
June 4, 1973
June 16, 1973
Niagara "
(Beaver I. Park)
Genesee
Oswego
Black
June 17, 1973
ii
ii
Phosphorus
DRP
40
38
9
41
2
1
104
50
5
40
35
26
3
49
46
13
TSP
45
46
15
50
6
7
111
56
16
51
45
33
7
58
59
24
(ygp/i)a.
TP
150
96
34
98
51
26
173
104
41
87
96
59
86
204
147
99
ppu
105
50
19
48
45
19
62
48
25
36
51
26
79
146
88
75
aAll values are mean values of triplicate determinations
except Sample No. 17 and No. 18 DRP values, which are mean
values of duplicate determinations
bDetermined by difference (TP-TSP)
147
-------
Table 47.
EXTRACTION OF GENESEE RIVER PARTICLES WITH
ACID, BASE, AND ANION-EXCHANGE RESIN
Sample PPa
No. (ygP/1)
34
42
51
58
360
108b
62
146
PPi (ygP/1 Extracted
Acid
284
59
27
30
Base
41
19
17
17
Resin
32
25
19
8
PP.(% of PP) Extracted
i
Acid
79
55
44
21
Base
11
18
28
12
Resin
9
23
31
6
Unless otherwise noted, these values were determined by
calculation (TP-TSP)
Determined directly on a suspension of particles in
distilled water
148
-------
Table 48.
DRP CHANGES DURING DARK INCUBATION OF GENESEE RIVER SAMPLE
NO. 42 PARTICLES IN LAKE ONTARIO WATER
Test sample
L. Ontario
water
100 ml L.
Ontario water
+ particles
from 100 ml
Sample No. 4 2
Initial
DRP
(ygP/l)a
19
Mean values
19
Mean values
DRP(ugP/l)
Sample Only
13 days
2
3
7
4
17
19
11
18
24 days
2
2
1
2
11
13
15
13
50 days
1
1
1
1
10
1
!_
4
after incubation
Sample + resin
13 days
5
5_
5
22
22
22
24 days
4
4
4
24
20
22
50 days
1
4
2
10
16
13
Net Mean values
(DRP from No.42 particles)
14
11
17
18
11
aThe dark incubation was begun May 7, 1973
The PP contributed to the test flasks by Sample No. 42 particles was 101 ygP/1;
therefore, the Max. % of Sample No. 42 PP found as DRP = 18/101 X 100 = 18% in
flasks with resin, and 14/101 X 100 = 14% in flasks without resin
-------
Table 49.
EXTRACTION OF NEW YORK RIVER WATER PARTICLES
BY SELENASTRUM IN ALGAL BIOASSAYS
Sample No.
50
50(A)
56(A)
34
42
51
SKA)
58(A)
43
47
52
52(A)
59(A)
44
53
53(A)
60(A)
PP a
Apparent Available P
Assayed Mean value
(ygP/1) (
Niagara
79
79
87
119
187
115
115
237
309
236
141
141
163
133
66
66
189
UgP/1)
R. at Ft.
<4
45
29
Genesee R.
3
<12
<3
47
85
Oswego R.
<2
<5
<3
62
52
Black R.
4
<3
30
49
Std. deviation
(ygP/1)
Niagara
4
2
1
--
4
18
--
6
5
1
--
7
5
Mean value
(% of PP)
<5
57
33
2
<6
<3
41
36
<1
<2
<2
44
32
3
<5
45
26
Determined directly on aliquots of suspensions of particles
in AAP(-P) medium
(A) = Autoclaved suspensions of particles
150
-------
cultures. In one case (Sample No. 58), the Selenastrum
could not be counted because the cells formed clumps with
the natural algae.
In contrast to the relatively low availability of PP
in natural particles, the autoclaved particles yielded
from 26 to 57 percent of their PP to Selenastrum. These
data were taken from Tables B.13 to B.28, Appendix B.
Total P Availability
Unfiltered river water samples were treated by three
different procedures, in an effort to estimate the percent
of TP which might become available in Lake Ontario.
Chloroform addition--
Table 50 shows the changes in DRP which occurred in
Sample No. 55 (Oswego R.) as a result of chloroform addition
and subsequent dark incubation. After only one day, the
DRP concentration was increased from 35 ygP/1 to 51 ygP/1.
Since 51 ygP/1 was greater than the initial TSP value of
45 ygP/1, some of the DRP increase must have come from the
PP in the sample. Taking the highest mean value observed
(59 ygP/1) as an estimate of the potential available P in the
sample, the estimated DRP contribution from the* 51 ygP/1 of
PP in the sample was calculated as 59-45=14 ygP/1, or 14/51
X 100 percent of PP (27 percent). The test flasks with 100
ygP/1 of a condensed phosphate spike, in the form of sodium
tripoly phosphate (TPP), showed complete hydrolysis of TPP to
DRP in seven days (160 ygP/1 DRP, compared to 59 ygP/1 in the
controls). That this hydrolysis was probably due to natural
enzymes in the river water was shown by the relatively slow
hydrolysis rate seen in the "chemical hydrolysis" control
flasks containing 100 ygP/1 TPP in chloroformed, distilled
(not autoclaved) water; only 12 percent of the TPP was
hydrolyzed after seven days.
151
-------
Table 50.
DRP CHANGES IN A CHLOROFORMED OSWEGO R. SAMPLE
(NO. 55), WITH AND WITHOUT ADDED SODIUM TRIPOLY
PHOSPHATE (TPP)
A. Initial sample No. 55 data: DRP = 35 ygP/1
TSP = 45 ygP/1
TP = 96 ygP/1
B. Dark incubation with chloroform
Test sample
No. 55 water
No. 55 water
+ TPP
Dist. water
+ TPP
Flask
1
2
3
1
2
3
1
2
3
DRP (ygP/1)
No. 1 day
49
52
11
Mean values 51
75
76
79
Mean values 77
2
2
_2
Mean values 2
after incubation
7 days
59
57
62
59
146
169
166
160
12
14
11
12
Test flasks were prepared as follows:
3 flasks with 200 ml of No. 55 + 2 ml CHC1,
2 ml CHClg + 2 ml of
3 flasks with 200 ml of No. 55 +
10 mgP/1 sodium tripoly-P (TPP)
3 flasks with 200 ml of dist. water + 2 ml CHC1
of 10 mgP/1 sodium tripoly-P (TPP)
+ 2 ml
152
-------
Table 51 summarizes the results of tests with river
and river mouth samples which had been stored at 4 C for
16 to 30 days prior to the chloroform addition. Since the
objective of these tests was to estimate the long-term
potential available P in the river waters, any biological
changes occurring naturally in the samples before chloro-
form addition were not considered harmful to the test
objectives. In this test, the DRP levels in some flasks
decreased during incubation with chloroform. Consequently,
average DRP values from three replicate flasks at each
incubation time were compared, and the maximum average DRP
level in the sample was selected as the best estimate of
the potential available DRP level in the sample. The DRP
values in Table 51 indicated the presence of phosphatases,
since in most cases the DRP level in the (single) TPP-spiked
flask was higher than the mean of the three unspiked flasks.
Because of the losses in DRP, especially after about 25 days
of incubation, later tests were run for only 7-16 days.
Table 52 lists the maximum DRP values seen in chloro-
form tests with all river samples, as compiled from Appendix
D,. * The following mean values and ranges were computed from
the data in Table 52, expressed as a percent of TP:
River Mean Range
Niagara R. -Ft. Niagara 63 41-91
Niagara R. - Beaver I. Pk. 51 26-80
Genesee R. 38 9-71
Oswego R. 68 52-81
Black R. 45 36-51
It is readily apparent from these numbers that no general
trend was seen for TP availability in the Niagara or Genesee
River samples, while the Oswego and Black R. samples showed
somewhat more consistent behavior in the tests.
153
-------
Table 51. DRP CHANGES IN CHLOROFORMED RIVER WATERS, WITH
AND WITHOUT ADDED SODIUM TRIPOLY PHOSPHATE (TPP)
A. Initial period of storage at 4 C in
Sample No.
22-Oswego R. Mouth
23-Oswego R.
24-Oswego R. Mouth
25-Black R.
26-Oswego R.
Determined prior to
B. Dark incubation with
Initial DRPa
(ygP/1)
58
49
41
14
79
storage period
chloroform at
darkness
Storage Period
(days )
30
30
16
16
16
17-23°C
DRPa after incubation
(mean values
Sample No. 1 day 8 days
No. 22 64
65
64
No. 22 + TPPa 77
No. 23 64
69
65
No. 23 + TPPa 86
No. 24 59
59
59
No. 24 + TPPa 69
No. 25 20
19
19
No. 25 + TPPa 33
No. 26 95
97
95
No. 26 + TPPa 112
69 ygP
(64) 70 (70)
73
146
72
(66) 52 (63)
65
79
44
(59) 52 (46)
41
90
15
(20) 16 (16)
18
68
46
(96) 96 (75)
84
144
in jparentheses )
25 days
/:L) 75
74 (73)
69
104
76
51 (63)
63
87
3
40 (22)
24
137
16
4 (15)
24
117
42
96 (72)
79
152
50 days
72
70 (66)
55
101
66
14 (39)
36
72
12
58 (36)
37
127
24
12 (24)
37
100
49
100 (66)
49
56
One ml of 10 mgP/1 sodium tripoly phosphate + 100 ml sample
water
Samples were filtered through glass fiber filters for DRP
analysis
154
-------
Table 52. MAXIMUM DRP VALUES IN CHLOROFORMED
NEW YORK RIVER WATERS AFTER 1 TO 16 DAYS
OF DARK INCUBATION
Sample
27
33d
41Q
50
56
rt
40a
49
57
34 ,
42d
51
58
s-\
22
(r.
23^
24b
(r.
26L
28
29
31
35,
43d
52
54
55
59
V
25b
36 ,
44d
53
60
No. TP
CygP/
18
34
22
26
59
15
51
86
386
150
173
204
93
mouth )
96
88
mouth)
154
93
106
95
105
96
104
87
96
147
53
34
34
41
99
Maximum DRP observed during
1) (ugP/1) (%
Niagara R. at Ft. Niagara
10
14
20
19
31
Niagara R. at Beaver I. Park
12
24
22
Genesee R.
36
51
122
77
Oswego R.
73
66
59
96
69
86
49
67
76
72
58
59
91
Black R.
24
15
17
21
36
incubationa
of TP)
56
41
91
73
52
80
47
26
9
34
71
38
79
69
67
62
74
81
52
64
79
69
67
62
62
45
44
50
51
36
, Mean values from triplicate flasks
Stored 16 days at 4°C before test
^Stored 30 days at 4 C before test
Stored 96 days at 4 C before test
155
-------
The effect of preliminary storage at 4 C before chloro-
form addition was studied with one set of samples as shown
in Table 53. The Niagara R. sample (No. 40) from the
Beaver I. station, along with the Genesee and Black R.
samples, showed little or no change in their DRP concentra-
tions during a 96-day cold storage at 4 C. In contrast,
the other Niagara R. (No. 41) sample showed an increase of
DRP from 18 to 73 percent of TP. Also, the Oswego sample
showed an increase of DRP from 40 to 72 percent of TP
during the cold storage period. All samples showed in-
creases in DRP as a result of chloroform treatment, but the
increases were slight for the Genesee and Oswego R. samples.
Substantial percentage changes were seen in the DRP levels
of the Niagara and Black R. samples after addition of
chloroform and subsequent incubation for only seven days.
After seven days the available P was essentially constant
or decreased relative to the value at seven days.
Dark Incubation with Resin--
Table 54 shows the results of a dark incubation of a
1:1 mixture of Genesee R. and river mouth water incubated
for 100 days, with and without anion-exchange resin. Using
the values obtained at 100 days, the calculated DRP from
the river water in the two bottles with resin was:
76 ugP/1 X 2 - 35 yg P/l = 111 yg P/l, or
70 percent of TP
The calculated value for the river water in the flask with-
out resin was 116 ygP/1, or 69 percent of TP, which was not
considered significantly different from the value with the
resin.
The results of other incubations, in which the river
water was not diluted by lake water, are given in Table 55
compiled from Tables C.9 to C.13 of Appendix C. The follow-
ing ranges and mean values were found for the maximum ob-
served mean DRP values in the incubations: (percent of TP)
156
-------
Table 53. CHANGES IN THE DRP FRACTION OF TP AS A
RESULT OF COLD STORAGE FOLLOWED BY
DARK INCUBATION WITH CHLOROFORM
DRP (% of sample TP)
Stored 4°C, then
Initial After 96 days chloroformed:
Sample No. Value at 4 C in dark 1 day 7 days 14 days
40 40 47 67 80 67
(Niagara R. at
Beaver I. Pk.)
41 18 73 82 91 86
(Niagara R. at
Ft. Niagara)
42
(Genesee R. )
43
(Oswego R. )
44
(Black R. )
27
40
26
27
72
26
29
76
32
33
77
47
34
79
50
aMean values of triplicate test flasks
157
-------
Table 54. DRP CHANGES DURING DARK INCUBATION OF EQUAL
VOLUMES OF GENESEE R. (NO. 16) AND
GENESEE R. MOUTH (NO, 17) WATER SAMPLES
A.
B.
Initial Sample Data
Sample
16
17
Phosphorus (ygP/1)
DRP TSP TP
43 82 167
12 20 43
Dark Incubation (Each value represents one test bo
DRP(ygP/l after incubation
Test Sample 10
300 ml No. 17 water
300 ml No. 17 water
days 25 days 50 days 100
£tle>
days
32 31 30 34
70 73 75 75
+ 300 ml No.16 water
300 ml No.17 water 33 33 39 35
+ anion-exch. resin
300 ml No.17 water 73 77 77 76
+ 300 ml No.16 water 69 72 75 75
+ anion-exch. resin
DRP in bottles with resin (Ig) includes DRP in solution
plus DRP on the resin.
Calculations:
1) Bottles with resin
No. 17 DRP at 100 days =35 ugP/1
No. 17/No. 16 mixture DRP at 100 days = 76 ygP/1
DRP from No. 16=(2X76)-35= 117 ygP/1
2) Bottles without resin
No. 17 DRP at 100 days = 34 ygP/1
No. 17/No. 16 mixture DRP at 100 days = 75 ygP/1
DRP from No. 16 = (2 X 75) - 34 = 116 ugP/1
158
-------
Table 55. MAXIMUM PER CENT OF TP OBSERVED AS DRP
DURING DARK INCUBATIONS OF NEW YORK RIVER WATER SAMPLES
Sample Maximum Observed DRP in Incubation a
No. TP Sample Sample + Resin Sample Sample + Resin
ygP/1 ygP/1 f6- ~* ^s
(% of TP)
Niagara R. at Ft. Niagara
33
41
50
56
32
49
57
16
34
42
51
58
T~,
23b
24°
(r.
26
31
35
43
52
54
59
C*i
25°
36
44
53
60
34
22
26
59
30
51
86
167
386
150
173
204
96
88
mouth)
154
95
105
96
104
87
147
53
34
34
41
99
12
5
8
30
Niagara
6
14
10
116
31
42
109
71
72
61
114
47
64
66
63
49
67
29
10
6
7
24
8
4
5
26
R. at Beaver I.
4
7
9
Genesee R.
117
63
39
110
77
Oswego R.
76
56
108
64
73
37
55
51
55
Black R.
26
3
7
8
22
35
23
31
51
Park
20
28
12
69
8
28
63
34
75
69
74
50
61
69
61
56
46
55
29
18
17
24
24
18
19
44
13
14
10
70
16
26
64
37
79
64
70
67
70
39
53
59
37
49
9
21
20
22
'DRP in flasks with sample + resin includes DRP in solu-
tion plus DRP on the resin. All values are mean values
from duplicate or triplicate test flasks, except sample
no. 16 (see Table 54).
^Stored 35 days at 4°C before test.
GStored 21 days at 4 C before test.
159
-------
Niagara R. -
Beaver I.
Genesee R.
Oswego R.
Black R.
20
40
62
29
12-28
8-69
46-75
17-55
River Mean Range (no resin) Mean Range (resin)
Niagara R. - 35 23-51 2.6 18-44
Ft. Niagara
12 10-14
43 16-70
60 37-79
24 9-49
In some cases, the flasks with resin showed lower values
of DRP than did the flasks without resin, so both values
were reported in the summary given above. Except in the
case of the Niagara River (Beaver I.) samples, the ranges
of TP availability were too broad to allow any generali-
zations to be made.
Autoclave treatment and bioassay--
Table 56 shows the forms of phosphorus found in auto-
claved river water samples in comparison to the forms
present in the original samples. One significant discrep-
ancy was found between the TSP after autoclaving and the
TSP before autoclaving. Sample No. 56 had 33 ygP/1 TSP
initially and only 17 ygP/1 after autoclaving. The DRP
dropped from 26 to 6 ygP/1. Smaller TSP losses were seen
in Samples No. 27, No. 40, No. 41, No. 28, No. 29, and No.
31. All of these samples were from the Niagara R. or
Oswego R. Since the purpose of autoclaving the samples was
to estimate the maximum potentially available P in the
water, the bioassay values for these samples are probably
under estimates of the maximum potentially available P
because of possible phosphorus fixation reactions which
occurred before the samples were filtered for bioassay.
Table 57 presents the bioassay data which were
collected for the purpose of determining whether the TSP
measured chemically was actually available for algal
growth. The data in this table, based on uncorrected
160
-------
Table 56. EFFECT OF AUTOCLAVE TREATMENT
ON THE SOLUBLE P FORMS IN RIVER WATERS
Initial Soluble
Sample
27
33
41
50
56
32
40
49
57
34
42
51
58
28
29
31
35
43
47
52
59
36
53
60
No . DRP
Niagara
4
5
4
1
26
Niagara
2
6
2
3
26
40
104
49
68
78
43
47
38
41
50
46
7
5
13
TSP
(UgP/1)
R. at Ft.
8
10
8
7
33
P Soluble P after Autoclaving
DRP
(ygP/1)
Niagara
__
2
6
TSP
6
13
6
16
17
R. at Beaver I. Park
5
10
6
7
Genesee R,
26
45
111
58
Oswego R,
71
82
49
52
46
50
56
59
Black R.
12
16
24
__
1
5
»
__
103
52
__
52
48
18
30
8
6
16
15
40
44
113
68
64
76
40
72
53
72
73
76
20
37
51
161
-------
Table 57. COMPARISON OF SOLUBLE P AND
ALGAL-AVAILABLE P IN AUTOCLAVED, FILTERED RIVER WATERS
P after Autoclaving, Filtering (ygP/1)
Sample No. DRP TSP Algal-Available Pa Bioassay Std. Dev
Niagara
27
33
41
50
56
2
6
6
13
6
16
17
Niagara
32
40
49
57
34
42
51
58
28
29
31
35
43
47^
52b
52C
59
36
53
60
1
5
103
52
52
48
18
30
8
6
16
15
40
44
113
68
64
76
40
72
53
72
73
73
76
20
37
51
R. at Ft. Niagara
<2
7
<2
<4
9
R. at Beaver I. Park
<2
<2
<4
4
Genesee R.
42
40
187
81
Oswego R.
86
124
54
124
54
82
93
103
94
Black R.
12
23
38
--
1
0
1
1
3
23
0.4
5
9
3
8
8
4
2
4
3
4
4
3
Mean values of Selanastrum bioassays, using uncorrected
data for quantitation of apparent algal-available P. The
bioassay data in this table were derived from the data in
Appendix B by multiplying the data in the appendix by
27/20 to account for the sample dilution in the bioassay
flasks.
^Assayed May 31, 1973.
'Assayed June 5, 1973.
162
-------
^750 <^a~':a indicate that in the Niagara R. and Black R.
samples, less P was available to Selenastrum than pre-
dicted by the TSP value measured after autoclaving. In
the Niagara and Black R. samples whose DRP data was
measured after autoclaving, the bioassay values fell
between the DRP and TSP concentrations, as expected if
there were no gross over or underestimations of the
available P. However, apparent overestimations were
found in several Oswego and Genesee R. samples where
apparent available P was greater than the TSP measured
after autoclaving.
As noted in Section VI, the correction of A75(, data
to a cell count basis should account for the optical
differences between cells grown in sample water supple-
mented with AAP(-P) and in AAP(-P) itself. These cor-
rections were made in Table B.29 of Appendix B to produce
the data given in Table 58. The following mean values
and ranges were calculated from the data:
Per Cent of TP
River Mean Range
Niagara-Ft. Niagara <12 9-15
Niagara-Beaver I. Pk. < 8 5-<13
t
Genesee R. 32 8-72
Oswego R. 60 38-84
Black R. 43 35-56
The values from the Black R. samples were not corrected
values, since none of the Black R. samples had been used
to construct the correlation curve shown in Figure 12.
Only the Niagara and Black R. samples showed relatively
consistent bioassay results, in terms of the per cent of
TP available after autoclaving and filtering. The ranges
for the other rivers were very broad.
163
-------
Table 58, COMPARISON OF TSP AND ALGAL-AVAILABLE P,
CALCULATED USING CORRECTED A7rn DATA
IN AUTOCLAVED, FILTERED RIVER WATERS
Sample
No.
27
33
41
50
56
32
40
49
57
34
42
51
58
28
29
31
35
43
47K
52o
52C
59
36
53
60
P
TSP
6
13
6
16
17
8
6
16
15
40
44
113
68
64
76
40
72
53
72
73
73
76
20
37
51
(ygP/D _
Algal-Available P
Niagara R. at
<2
5
<2
<4
7
Niagara R. at
<2
<2
<4
4
Genesee
30
31
124
57
Oswego
61
88
36
88
39
57
63
72
65
Black R
12
23
38
TSP
Ft.
P (% of TP)
Algal-Available P
Niagara
33 <11
38 15
27 < 9
62 <15
29 12
Beaver I . Park
27
40
31
17
R.
10
29
65
33
R.
69
72
42
68
55
74
70
70
d42
59
90
52
< 7
<13
< 8
5
8
21
72
28
66
83
38
84
41
58
60
69
44
35
56
38
Mean values of Selenastrum bioassays, using A7,-n
corrected as explained in the text. The correct*
_ _ data
explained in the text. The corrected
A^rn data are given in Table B.29 of Appendix B , along
with the apparent available P values in the assay
flasks. The assay flasks values were multiplied by
27/20 (to correct for sample dilution in the bioassay
,flask) to obtain the data given in this table.
^Assayed May 31, 1973
Assayed June 5, 1973
Black R. samples were not quantitated with corrected data;
the values given in this table are taken from Table 57
data.
164
-------
SECTION VIII
DISCUSSION
MADISON URBAN RUNOFF
Stormwater samples collected in Madison, Wisconsin,
from August 1972 to March 1973 showed a wide range of
total phosphorus concentrations, from 59 ugP/1 (see Table
31 ). These concentrations were probably the result of
the following factors: (1) the different street solids
loading intensities expected for different urban land
usages (APWA, 1969; Sartor and Boyd, 1972), (2) the
changes in TP concentration with time during the runoff
event (Kluesener, 1971; Weibel et al., 1964), and (3) the
additional phosphorus inputs from autumn leaf-fall (Cowen
and Lee, 1973) or from construction activities (Ryden et
al., 1972), which modify the normal solids loading inten-
sity for a given urban land use. Regardless of the abso-
lute values of TP in the samples, however, the soluble
and particulate P forms in the samples were expected to
qualitatively represent the nature of the materials trans-
ported to the Madison lakes in urban runoff.
Particulate P Availability
Mean values of acid and base extractable inorganic P
(PP.) from PP contained in urban runoff ranged from 6 to
60 per cent of PP, with considerable overlap of the values
for the two extraction methods (see Figure 15.). The
165
-------
overlap was expected, since the acid extraction should
dissolve calcium phosphate completely and aluminum and
iron bound phosphate considerably, while the base ex-
traction should dissolve aluminum and iron bound phos-
phate, but not calcium phosphate (Chang and Jackson,
1957). Thus, especially if the content of calcium phos-
phates in the particles was low, the two extraction pro-
cedures could yield similar results. The degree of
hydrolysis of organic phosphates with these procedures
was not evaluated but could include both hydrolysis
during the extraction and in the acidic color reagent
used in the Murphy and Riley (1962) colorimetric test for
DRP. The latter step was considered by Chamberlain and
Shapiro (1973) to cause litte hydrolysis of glucose-1-
phosphate, which is reported to be a very labile ester
(Weil-Malherbe and Green, 1951). Consequently, hydrolysis
was likely minimal in the DRP analysis procedure. How-
ever, since the filtered extracts were often stored (4OC)
overnight before the analysis for inorganic P, and be-
cause of the long duration of the base extraction (over-
night), some hydrolysis of organic esters may have
occurred in the extraction medium.
The results shown in Table 32 and Figure 15 indi-
cated that the range of mean values for acid extractable
PP. in residential runoff samples was greater than the
range of mean values for base extractable PP. in the same
samples. The greater variability between samples of the
acid extractable fraction may have been due to differences
in the relative amounts of calcium-bound P in the partic-
ulate matter, since calcium-bound P would be extracted by
the acidic treatment but not by extraction with base.
Further, the acid extractable PP. likely approaches the
total inorganic P content, as is the case for many lake
sediments (Williams et al., 1971b).
166
-------
When the group mean values from the various urban
sampling sites were compared for a given extraction
method, there appeared to be very little difference be-
tween the various land uses sampled. As shown in Figure
15, all group mean values for acid extractable PP. fell
within the range of 33 to 46 per cent of PP, while all
group mean values for base extractable PP. fell into the
range of 22 to 27 per cent of PP. Possibly, the domi-
nant form of PP in all samples was derived from a common
source input, such as dustfall or eroded soil. Sartor
and Boyd (1972), in a study which included industrial
land uses, noted that while the various types of urban
areas exhibited different loading intensities of total P
in dry street surface contaminants (industrial > residen-
tial > commercial), the P(\ content of the solids did not
appear to differ appreciably among the various land use
types (0.103 to 0.14-2 per cent by weight).
In contrast to the treatment of runoff PP with acid
and base, which dissolved certain phosphatic components
of the PP, the resin extraction was a relatively mild
treatment, designed to measure only the inorganic P
which was involved in exchange reactions between the solid
and solution phases. Thus, the resin method was expected
to produce lower values of extractable PP. than the acid
or base methods. This was actually the case, as demon-
strated by the group mean values shown in Figure 15.
The range of mean values for sample replicates was 2 to
28 per cent of PP (Table 32 ), but the group mean values
for the various land uses were in the range of 13 to 17
per cent of PP.
Long-term (26 days) dark incubations of particles
from runoff samples D-ll, E-ll, and H-ll in unfiltered
lake waters containing anion-exchange resin showed resin
167
-------
extractable PP-^ values of 21, 8, and 10 percent of PP,
respectively (Tables 34 and 35 ). In comparison, the
short-term (24 hour) resin extraction of these particles
in distilled water gave extractable PPj_ values of 20, 16,
and 20 percent of PP, respectively (Table 32 ). Since
the DRP levels in both systems were maintained low by the
resin, the major differences between the systems were the
presence of lake water organisms and chemical constit-
uents, and the time allowed for equilibration. Consider-
ing the errors inherent in the lake water experiment due
to subtraction of control values from test values, the
two procedures yielded similar results. This would indi-
cate that physical-chemical sorption-desorption and precip-
itation-dissolution reactions were more important in
controlling the release of inorganic P than were biological
processes, such as DRP release from autolysis of microbial
cells in the PP. The increases in DRP observed in lake
water control flasks (no runoff particles added) between
1 and 13 days demonstrated mineralization of detrital
phosphorus to DRP, as reported in several investigations
involving dark incubations of sea water (Waksman et al.,
1937; Renn, 1937).
Long-term dark incubations of runoff particles in
lake waters without anion-exchange resin demonstrated
DRP removal from the lake water by sample E-ll and H-ll
particles, while the particles from sample D-ll showed
essentially no net contribution or removal of DRP. As
discussed by Taylor and Kunishi (1971) and Ryden
-------
DRP level in the receiving water, any results obtained
from tests with resin would be applicable only to receiving
waters with low DRP levels, such as the epilimnion waters
of the Madison lakes in the late summer. Also, since the
resin uptake might simulate DRP uptake by algae, the resin
tests might be representative of low DRP waters which
exhibit a high algal demand for P.
The addition of orthophosphate spikes to unfiltered
runoff sample D-ll demonstrated that the particles in the
runoff were capable of rapid fixation of inorganic P
(Table 36). Resin extraction of the spiked runoff sample
replicates showed that all of the added spike was recovered
from the solution phase and the solid phase of the runoff.
Since these tests were performed in a total elapsed time
of just 48 hours, biological release of DRP from autolysis
of organisms was considered unimportant, although uptake
of the orthophosphate spikes by microorganisms could have
occurred (Phillips, 1964). However, the recovery of
phosphate which had been fixed by the particles indicated
that sorption-desorption processes were probably more
important than biological reactions. The' slight excess
(6 to 11 percent) recovered P over 100 percent from the
resin extraction was likely derived from exchangeable P
originally associated with the PP in the runoff.
In contrast to the short-term (24 hour) direct
extractions of PP with resin (Table 32 ), long-term (up
to 50 days) dark incubation of unfiltered runoff (Table 37)
was designed to allow biological reactions to occur in the
runoff particles. As indicated in Table 59, the estimated
inorganic P released from PP (Maximum R - DRP - Initial
sample TSP) in the long-term incubations were significantly
(10 or more percent) larger than the short-term direct
169
-------
Table 59. COMPARISON OF RESIN EXTRACTABLE PP^
FOR DIRECT SHORT-TERM EXTRACTION OF PP AND
FOR LONG-TERM DARK INCUBATIONS OF UNFILTERED RUNOFF
Resin Extractable PP^
(% of PP)
Calculated-^ from long-
Direct short-term term dark incubations
Sample No . Extraction of PPa of unfiltered runoff
A- 8
A- 9
A-12
B-4
B-7
B-8
B-9
B-12
D-8
D-10
D-ll
D-12
E-ll
H-ll
F-7
F-9
1-12
22
15
18
19
12
19
15
17
15
12
20
20
16
20
21
16
13
20
-4
17
32
44
12
-8
9
20
15
13
14
1
-4
31
2
7
aMean values, from Table 32
^Calculated using equation (1) in text and data from Table 37
extraction values in only 3 of the 17 samples tested
(samples B-4, B-7, and F-7). In five samples, the results
of the long-term dark incubations were significantly lower
than the results of the short-term PP extractions. Both
test systems contained anion-exchange resin, to promote
desorption of inorganic P from the PP forms. However, in
the long-term dark incubations of unfiltered runoff the
soluble chemical constituents of the runoff were present
along with the sample PP and resin. In the direct resin
170
-------
extraction of PP over 24 hours, only the resin and sample
PP were present in the distilled water medium. Possibly
the soluble chemical constituents of the runoff may have
competed for the available sorption sites on the resin,
thus reducing the capacity of the resin for inorganic P
adsorption. Such competition was absent in the distilled
water-resin extraction system.
Although the unfiltered runoff samples were incubated
for long periods, no significant release of DRP was detected
in most samples, either from desorption or mineralization
reactions, such as hydrolysis of organic P or bacterial
release of inorganic P as a result of death and cell lysis.
Nine of the samples showed values which were in agreement
(to within 10 percent of PP) for the long-term and short-
term test systems. Since the short-term system evaluated
only the physical-chemical reactions of the PP, while the
long-term system evaluated both physical-chemical and bio-
logical reactions of the PP, agreement of the two test
systems indicates that physical-chemical processes are the
key factors controlling the release of inorganic P to
solution from particulate P forms. Only in the three
samples (B-U, B-7, and F-7) where the results of long-
term incubations significantly exceeded the results of
the short-term tests was there evidence for significant
inorganic P release from biological reactions in the run-
off particles.
Direct biological assessment of the algal-available
fraction of PP was sought by growing Selenastrum in a
P-free algal medium with the runoff particles as the sole
source of P to the algae (Table 33). An overall range
of 8 to 55 percent of PP was found for the 13 samples
tested, mostly from residential sites. The mean value for
171
-------
all samples was 30 percent, a value which was intermediate
between the ranges of group mean values from the acid and
base extractions of PP, 33 to 46 and 22 to 27 percent of
PP, respectively. However, such comparisons are not strictly
valid, since some of the samples were not analyzed by all
three methods. Figure 18 compares the results of the
various PP extraction methods and the algal bioassay on a
sample-by-sample basis. The ranges shown represent the
range of the individual replicate determinations from all
methods on a given sample.
With three exceptions, the bioassay results fell
within a range established by the mean values of the acid
and resin extractions. In six of the 10 samples tested by
bioassay and base extractions, the results agreed to within
10 percent (of PP) or less. In contrast, the bioassay and
acid extraction results were within 10 percent in only 4 of
13 samples. The studies of Golterman et al. (1969) and
Fitzgerald (197 Ob) have indicated that some relatively
insoluble iron and calcium phosphates could be used for
growth by Selenastrum, Chlorella, and Scenedesmus cells.
Thus, some agreement between the results of bioassays of
runoff PP and extractions of PP which dissolve iron and
calcium phosphates might have been expected from the reports
in the literature. The resin extraction data was in
essential agreement with bioassay results (10 percent or
less) for 7 of the 13 samples tested, so apparently the
resin and base extraction methods were measuring "available"
PP more often than was the acid extraction method.
The test most likely to predict the true availability
of PP forms in the receiving water cannot be definitely
determined from the data reported here. The factors
which will affect the eventual contribution of inorganic
172
-------
100
90
0 80
UJ
I-
O 70
o:
X 60
UJ
Si 50
u.
O
40
g 30
or
UJ
0- 20
10
Figure 18. Comparison of chemical and biological extraction of PP
( Madison urban runoff samples)
MEAN VALUES OF: A = ACID EXTRACTION
B = BASE EXTRACTION
R^RESIN EXTRACTION
S = SELENASTRUM GROWTH
ASSAY
I
RANGE OF
= DATA FROM
ALL METHODS
A-8 A-9 A-12 B-8 B-9 B-12 D-8 D-ll D-12 H-ll E-l
SAMPLE NO.
1-12 F-9
-------
P from PP will include, besides the chemical forms of PP
in the runoff, the temperature and DRP concentration of
the receiving water, the steepness of the lake basin at
the point of discharge, the rate of sediment deposition
and mixing in the area of deposition, and finally, the
density and diameter of the particles themselves.
Large, high density particles, such as those observed
in samples of runoff from both residential and urban con-
struction areas, would rapidly settle out of the water
column in the absence of turbulence at the point of entry
into the lake. If the littoral zone of the lake is not
steep at this point, the particles will remain in the
photic zone of the lake, where they could release phosphorus
to the water for phytoplankton growth, or to rooted
aquatic plants. Martin et al. (1969) have shown that lake
muds are the principal source of nutrients for the (rooted)
aquatic weed, Najas. The temperature and DRP level of the
lake water, as well as the degree of sediment-water mixing
would also affect the fraction of PP contributed to the
lake water as available P. Cooke and Hislop (1963) reported
that the degree of exchange of phosphorus from coils to
anion-exchange resin was temperature dependent, with a
two-fold increase in phosphorus extractability between
10 and 30°C. A temperature of 15-22°C was used in the
studies with Madison runoff and Lake Ontario tributaries,
to simulate natural water temperatures.. The DRP level of
the receiving water would affect the sorption or desorption
of inorganic P from soil particles, as predicted from the
phosphate sorption curves of Ryden ejt al. (1972), which
related the concentration of inorganic P in the aqueous
phase to the change in concentration of inorganic P on
the particles in equilibrium with the aqueous phase. Low
174
-------
DRP levels in the receiving water would increase the
potential desorption of inorganic P. However, the reverse
(sorption) at high DRP levels in the water may not
necessarily hold for some eroded soils. Ryden ejt al. (1972)
found that soil from an Al horizon was capable of releasing
an appreciable amount of inorganic P to solution, even in
the presence of relatively high levels of added inorganic
P. A dissolution reaction of a slightly soluble soil-ferti-
lizer P reaction product in this horizon was suggested as
the reason for the P release characteristics of the Al
horizon. If temperature and DRP concentration are suitable
for P release, the actual release may still be limited by
the degree of mixing of the particles with the water column.
Also, further sedimentation would hinder release by covering
previously deposited material, as in the foreset beds of
a delta.
Small, low density particles would remain in suspension
in the photic zone for relatively long periods of time
compared to large, high density particles. Therefore, the
small particles should be more important to the chemical
and biological availability predicted in the tests discussed
above. Particles remaining in suspension would be in close
contact with planktonic algae, as in the algal bioassay,
and if the soluble inorganic P level of the receiving
water were very low, algal uptake might approximate the
phosphorus availability predicted by the resin extractions
of PP forms in distilled water- Evidence from other
studies indicates that much of the PP in runoff is
associated with fine particles. Sartor and Boyd (1972)
reported that up to 55 percent of the P by weight in dry
street contaminants was associated with particles less
175
-------
than "43 microns in diameter. The loadings of phosphorus
from dustfall reported by Kluesener (1971) and Johnson et al.
(1965) of 0.7 Ibs. total P/acre/year and 0 to 0.45 Ibs.
water-soluble P/acre/year, respectively, are indications of
the potential contribution of small particles to total
runoff phosphorus. Much of the fine particulate matter,
falling on impervious surfaces, is readily transported by
surface runoff, which has been estimated to carry an
annual load of 1 Ib. total P/acre from all source inputs
of phosphorus in the urban environment (Kluesener, 1971).
Thus, the urban loading of dustfall P and yield of runoff
total P are seen to be different by perhaps no more than
a factor of two or less (0.45 to 0.7 compared to 1.0).
Evans et_ al. (1968) were only able to reduce the TP content
of runoff samples by 30 percent in 5 hours of settling
time in the laboratory, although the suspended solids
concentration was reduced by 70 percent in the tests.
Because of all the considerations discussed above,
the results of the chemical and biological tests of PP
availability are probably overestimates of the true PP
availability in the receiving water. However, the bio-
assay and short-term direct resin extractions of PP are
probably the best estimators of the true behavior of the
particles, based on theoretical considerations. The mean
values of PP availability determined by these two tests
have been summarized for all samples tested, as follows:
Mean Percent of PP "Available"
Test Range Ave. of Mean Values
Resin Extraction 2-28 15
Selenastrum Bioassay 8-55 30
176
-------
Soluble P Availability
The bioassays of membrane-filtered runoff showed
three classes of response (Table 38): (1) available P
greater than 114 percent of TSP, (2) available P between
98 and 114 percent of TSP, and (3) available P less than
60 percent of TSP. In membrane-filtered samples, the TSP
value represented the maximum possible concentration of
available P, since it was considered improbable that any
phosphorus available for algal growth would not be measured
by the persulfate digestion method for TSP. Samples
whose available P levels fell within 98 to 114 percent of
TSP (A-8, D-8, D-10, and D-ll) apparently contained
soluble P compounds whose biological availability roughly
matched their liability to persulfate digestion.
Samples whose available P was less than TSP (H-10,
H-ll, E-10, E-ll, and 1-12) apparently contained dissolved
P forms which were measured by the chemical digestion pro-
cedure, but were not available for algal growth. Since
the available P levels in all these samples were less
than even the DRP levels, the chemical method for DRP may
have overestimated the available P in these filtrates.
Alternatively, the bioassay may have been in error, due
to the difficulty in establishing a standard curve for the
bioassay of soluble P as discussed below. The DRP in
the filtrates may have included some colloidal inorganic
P, which could pass the filter and be dissolved by the
acid conditions of the DRP analysis and the very acidic
conditions of the persulfate digestion for TSP. The bio-
logical availability of such forms would likely be less
than that of dissolved orthophosphate-phosphorus. Chamber-
lain (1968) found that approximately 90 percent of the
32P in membrane-filtered (0.45 micron pore size) lake
177
-------
3 2
waters, prepared with P before filtration, could be
removed by refiltering the lake water through a 0.01
micron pore size membrane filter- In a parallel test,
3 2
80 percent of the P could be sedimented from the membrane
filtered lake water by ultracentrifugation. Other possible
sources of nonavailable "soluble" P include condensed or
organic P compounds which are hydrolyzed in the chemical
analyses but are not available to the test algae in the
bioassay. Chamberlain and Shapiro (1969) have pointed
out that arsenate may also appear as "available P"
(phosphate) in the chemical test for DRP. None of these
sources of error can be ruled out until more data is
collected on the soluble P forms in urban runoff.
Six of the runoff samples tested showed algal-avail-
able P higher than 114 percent of TSP (Sample B-9 is
included in this group, with available P reported as
greater than 110 percent of TSP). Since this was theoret-
ically impossible, an error in the bioassay procedure
was indicated. In all the bioassays of runoff soluble P,
the filtrates were supplemented with concentrated AAP (-P)
medium such that in the assay flasks there would be at
least the same concentration of AAP (-P) nutrients as
were present in the AAP standard flasks. However, unlike
the standard AAP cultures, the sample cultures also con-
tained the nutrient salts from the original runoff,
diluted somewhat by the added AAP (-P) spike and algal
inoculum. Thus, even though phosphate limitation existed
in both the sample and standard cultures, the nutrient
balance differed and produced differences in the optical
properties of the cells. In Figure 16, the curve
of A~5Q versus added orthophosphate spikes for sample
B-9 was transposed to begin at 167 ygP/1 for the unspiked
sample, since 167 ugP/1 was the TSP concentration in this
178
-------
sample culture. It could have been placed at any point
less than the TSP value; the same conclusion would have
been reached, namely that the true response curve repre-
sented by the extrapolation of the spiked sample response
curve back to 0.0 ygP/1 was to the left of the standard
response curve. Thus, at any given A750 value for a
sample culture, the standard curve would overestimate
the available P concentration in the sample culture flask.
Because of the problems involved with high values of
apparently available TSP, definite conclusions about the
relative availability of the TSP forms in Madison urban
runoff could not be made. However, the low results from
some of the samples do indicate that the chemical analyses
of soluble P forms should not be accepted as accurate
measurements of available P unless they have been correlated
with biological tests.
PRECIPITATION SAMPLES
Madison Snow Samples; PP Availability
Chemical analyses of Madison snow samples for phos-
phorus forms showed that most of the phosphorus was in a
particulate form, probably from dustfall on the snow or
crystallization of the snow around atmospheric dust par-
ticles. The biological availability of the snow PP
ranged from less than 2 to 23 percent of PP (Table 39).
The dustfall samples collected by Kluesener (1971) showed
about 14 percent of the total P in the dustfall to be
leachable as DRP. Since by definition "dry fallout" or
dustfall is entirely particulate in nature, Kluesener
was actually measuring the percent of dustfall PP which
could be leached as DRP. Consequently, his data are com-
parable to the bioassay results reported here.
179
-------
The snow sample which showed less than 2 percent of
PP to be algal-available was collected in a commercial
section of Madison (the roof of the City-County building),
where the automobile traffic is very heavy- Sartor and
Boyd (1972) reported that the concentrations of zinc,
copper, and lead were highest in street surface contami-
nants from commercial areas, as compared to materials
from industrial or residential areas. Thus, the^possibility
of toxicity in the bioassay of the commercial area PP must
be considered as an explanation of the apparent low P
availability in this sample. These results are only
intended as preliminary estimates of the PP availability
of dustfall. Unfortunately, more samples of this type
were not collected during the winter of 1972-1973.
New York Rain Gage Samples; TP Availability
These samples displayed a unique mixture of phosphorus
forms, as demonstrated by the data in Table 4-0. Several
samples showed very high TSP levels along with relatively
low DRP concentrations. Such data indicated that the
filtration procedure may have included phosphorus forms
which were dissolved but not reactive, or which were not
truly dissolved species in the filtrates analyzed for DRP
and TSP. Generally, in runoff waters, the concentrations
of DRP and TSP were of comparable magnitude (see Table 31).
The presence of colloidal, acid soluble phosphorus forms
would not in itself explain the differences seen, unless
the colloidal forms required oxidation for release of
their phosphate to solution as inorganic P.
The bioassay of unfiltered rain gage waters with
"abnormal" TSP concentrations showed that the algal-
available P levels were comparable to the DRP concentra-
tions, and hence much lower than the TSP (or TP) levels.
180
-------
Table 60 compares the chemical and biological analyses
of the rain gage samples. If the samples with extremely
low TP values are disregarded, only two samples, June 601
and June 604, contained algal-available P concentrations
of the same order of magnitude of DRP and TSP. Since
only two sets of rain gage samples were received from
New York state, any conclusions relating TP availability
(range: <1 to 90 percent of TP) to the location or con-
dition of the gage (open or closed) would be extremely
difficult. The results of this short study serve to
point out that care must be taken in the interpretation
of chemical analyses of the phosphorus forms, especially
TSP, in such samples.
GENESEE RIVER BASIN SAMPLES; PP AVAILABILITY
Land use in the Genesee R. basin study varied from
pasture land to high density residential area. The latter
was of interest in comparison to the Madison urban runoff
study. Samples from the non-urban land use areas in the
Genesee R. basin were frequently so dilute that direct
extractions of the PP in these samples were difficult.
Consequently, most of the estimates of PP availability
were made on samples from the residential areas, Stations
No. 2 (Rochester East), and No. 7 (Dansville).
The group mean values of acid extractable PPj_ from
Stations No. 2 and No. 7 were 48 and 30 percent of PP
respectively. These values were just outside the range
of the group mean values from the Madison urban runoff
PP, 33 to 46 percent, although the individual sample
values showed ranges which overlapped (Figures 15 and 17).
181
-------
Table 60. COMPARISON OF CHEMICALLY AND BIOLOGICALLY
DETERMINED PHOSPHORUS FORMS IN NEW YORK RAIN GAGE SAMPLES
Sample No.
May 601-0
May 602-C
May 603-C
May 604-C
May 605-C
May 6Q6-C
June 601-C
June 602-C
June 603-0
June 604-C
June 605-C
June 606-C
June 608-0
DRP
60
8
0
0
0
1
52
4
5
48
0
1
1
TSP
245
62
2
2
304
401
63
6
82
58
6
6
318
TP Algal-Available Pa
p/i
420
86
4
6
350
439
72
10
106
64
9
12
346
42
4
^14
< 14
<14
<3
42
<14
4
58
< T Lf.
< 1_U
^14
aMean values, from Table 41
The group mean values of the base extractable PP. from
Stations No. 2 and No. 7 were 30 and 18 percent of PP,
respectively. These values were lower than the acid ex-
traction values for the same samples and just outside the
range of the group mean values from base extractable PP^
in the Madison samples, 22 to 27 percent. Comparison of
Figures 15 and 17 shows that there was an overlap in
the ranges of the base extraction data from the Madison
and Genesee basin samples.
The resin extractable PP^_ data from the Genesee "R.
basin Stations No. 2 and No. 7 showed group mean values
of 25 and 11 percent of PP, respectively. In comparison,
the group mean values from the Madison samples fell into
182
-------
the range of 13 to 17 percent. As with acid and base
extraction data, however, there was considerable overlap
between the ranges of the resin data shown in Figures 15
and 17.
Like the PP in Madison samples, the PP from Stations
No. 2 and No. 7 of the Genesee R. basin had group mean
values of extractable PP.j_ in the order of acid > base >
resin. However, unlike the Madison PP samples, whose
group mean values for a given extraction method all agreed
to within 13 percent (of PP) or less, the acid extraction
group means for Stations No. 2 and No. 7 differed by 18
percent. The differences between the group mean values
of the base and resin extractions of PP from Stations
No. 2 and No. 7 were 12 and 7 percent, respectively. In
all methods of PP extraction, the Station No. 2 materials
showed the highest values of extractable PP-j_.
The results of the extractions of PP forms contained
in runoff from Stations No. 1, No. 4, and No. 9 were close
to the results of extractions of PP from Station No. 7
samples, even though the runoff from the former stations
was derived from cropland (No. 1), pasture (No. 4), and
brushland (No. 9).
Long-term dark incubations of unfiltered runoff
Samples No. 402-8, No. 407-8, and No. 507-13 with resin
yielded the following estimated inorganic P released
from PP (R-DRP-Initial sample TSP), as compared to the
results of short-term (24 hr.) direct extractions of PP
with resin (see Table 43):
183
-------
Percent of PP Extracted by Resin
Long-term Dark Short-term Direct
Sample No. Incubation of Runoff PP Extraction
402-8 (Rochester East) 28 25
407-8 (Dansville) 22 11
507-13 (Dansville) 7 13
The long-term incubation (50 days) values were close
to the values from the short-term resin extractions of PP
in distilled water, except in the case of Sample No. 407-8.
The larger values from the long-term incubation of this
sample may have been the result of mineralization pro-
cesses occurring in the particles during the dark incubation,
The direct bioassay of PP by Selenastrum is compared
to the other extractions of PP in Figure 19- In all
cases where both a resin extraction (short-term, in
distilled water) and a Selenastrum bioassay were run on
the PP, the bioassay value (S) was essentially equal to
or less than the resin extraction value. Since in most
cases, the base extraction values were close to the resin
extraction values, the bioassay values agreed nearly as
well with the base extraction as with the resin extraction
values. For simplification, bioassay values listed as
"less than" a given cutoff P concentration (Table 44)
were plotted at the cutoff value in Figure 19 . Acid
extraction values were generally significantly higher
than the bioassay values, which ranged from <1 to 24
percent of PP. The acid extraction range was, in contrast,
18 to 60 percent of PP.
Bioassays on particles which had been autoclaved in
their AAP (-P) medium suspensions showed a slightly
184
-------
CO
C_n
100 r
90
80
Q,
0_
UJ
o
IT
UJ
Q.
70 -
o
UJ
h-
o
IT
X 60
UJ
50
40
30
20
Figure 19. Comparison of chemical and biological extraction of PP
( Genesee R. basin samples)
MEAN VALUES OF: A = ACID EXTRACTION
B = BASE EXTRACTION
R= RESIN EXTRACTION
S =SELENASTRUM GROWTH
ASSAY (* INDICATES AUTOCLAVED
PARTICLES)
I
RANGE OF
= DATA FROM
ALL METHODS
501- 501- 501- 502- 502- 502- 502- 502- 504- 507- 507- 507- 507- 507- 507-
12 13 14 7 8 II 12 14 8 7 8 II 12 13 14
SAMPLE NO.
-------
higher availability range (8 to 34 percent of PP) than
did "natural" particles. In one of the samples tested
by both bioassay procedures, (No. 507-14), the increase
in available PP due to autoclaving was only 7 percent of
PP. Figure 19 shows that even after autoclaving, the
new bioassay value of 10 percent(S*) was not much dif-
ferent from the resin extraction, base extraction, or
original bioassay value. It appears that biological
phosphate was not released in large quantity as a result
of the autoclaving, or the particles resorbed most of the
inorganic P which should have been released from lysed
microbial cells in the PP. Physical-chemical phosphorus
fixation mechanisms would seem to be dominant in such
samples as those from Station No. 7 (Dansville- high
density residential). Sample No. 507-13 also showed a
bioassay value after autoclaving which was close to the
resin and base extraction values. Samples of PP from
other stations (No. 501-13, No. 501-14, and No. 502-14),
in contrast, tended to show bioassay values after auto-
claving which were higher than the resin extraction
values or the bioassay values for "natural" particles.
In summary, the expected percentage of available
PP in the Genesee R. basin samples (as estimated by
the short-term direct resin extraction, base extraction,
and bioassay (unauto.claved) mean values from all samples
tested) would be:
Mean Percent of PP "Available"
Test Range Average of mean values
Resin Extraction 1-30 16
Base Extraction 10-37 22
Selenastrum Bioassay
-------
NEW YORK RIVER SAMPLES
Particulate P Availability
Direct chemical extraction of PP forms isolated
from river water samples was attempted only for Genesee
R. samples. Table 47 demonstrated that the chemical
nature of the PP forms in this river was quite variable
in the different samples obtained during the spring of
1973. The short-term direct resin extraction data
for Genesee R. Sample No. 42 showed 23 percent of PP
extracted, which compared fairly closely to the results
of the long-term dark incubation of Sample No. 42 par-
ticles in Lake Ontario water containing anion-exchange
resin (18 percent of PP; Table 48).
Selenastrum growth assays of the PP from Genesee R.
samples showed that less than 6 percent of the PP was
available to Selenastrum. The apparent contradiction
between these data and the short-term resin extraction
data, which showed 6 to 31 percent of PP resin extrac-
table, may be explained by available P uptake by com-
peting organisms native to the water samples, in the
algal growth assays. Several species of diatoms were
/
observed during the microscopic counting of Selenastrum
cells in the 18-day growth assays. Although the short-
term resin extractions were also performed in a lighted
room, the 24 hr. resin equilibration period was likely
too short for significant growth of natural algae,
compared to the 18-day algal assays, so that very little
competition for dissolved inorganic P was expected from
natural algae. Also, the resin extractions were per-
formed in a distilled water medium, which was not as
conducive to algal growth as was the AAP (-P) algal
assay medium.
The competition for DRP was eliminated in the algal
187
-------
assays by autoclaving the suspensions of PP in AAP (-P)
medium, before inoculating with the test organism. Auto-
claved particles from two Genesee R. samples, No.58 and
No.51, showed 36 and 41 percent of PP available, re-
spectively, compared to 6 and 31 percent for short-term,
resin extractable PP.. The higher values for autoclaved
PP bioassays compared to short-term resin extractions
may be partially due to release of cellular phosphates
from native microorganisms lysed during the autoclaving
procedure. These phosphates would not be extractable
by the anion-exchange resin. Alternatively, some of the
P forms in the autoclaved suspensions may have been
available to algae via phosphatase enzyme hydrolysis,
yet not measured as DRP in the resin extraction procedure.
Samples of PP from the Niagara, Oswego, and Black
Rivers also showed less than 6 percent of the PP to be
available to Selenastrum, unless the PP suspensions were
autoclaved prior to bioassay. Autoclaved suspensions of
PP from these rivers showed 26 to 57 percent of PP
available to Selenastrum. Because of the lysis of cellular
biomass in the autoclaving procedure, the results of
such assays should be regarded as indications of the long-
term availability of the PP forms in Lake Ontario.
The bioassays of "natural" particles demonstrated
that the particles themselves were not inert, but were
capable of holding available P during incubations under
light. Much phosphorus may have been "biologically
available" from the PP during these assays, but it could
not be measured by looking only at the growth of the
inoculated test organism. The short-term resin extrac-
tion procedure attempted to measure only the organic
P bound to the PP by physical-chemical forces, hence
that P which would be released during periods of low DRP
levels in Lake Ontario. By not allowing the released P
-------
to be taken up by a growing population of native algae,
as in the bioassay of "natural" particles, the resin
procedure probably gave a better estimate of the readily
available P than did the bioassay of "natural" PP.
Several treatments of unfiltered river water were
performed in an effort to gain some insight into the
long-term availability of the total phosphorus in the
samples. By using equation (1), below, the contribution
of DRP from PP in the samples was estimated:
Maximum observed Initial value Estimated con-
(1) Algal-available P - of sample TSP = tribution of DRP
or DRP from sample PP
It should be noted that this equation would underesti-
mate the contribution of algal-available P or DRP
from PP in those samples where the initial dissolved
unreactive P concentration (TSP - DRP) was high and was
not used for algal growth or converted to DRP during the
test. In those cases, DRP from the particles equal to
any unavailable dissolved unreactive P would not be
counted by equation (1) in the estimation of available
P from PP. However, at least the P computed from (1)
had to come from PP. The contributions calculated by
equation (1) were expressed as a percentage of sample PP,
in order to compare the calculated values with the results
of direct bioassays of PP.
Table 61 lists the calculated PP availability
figures, from (a) chloroform treatment of river waters,
(b) long-term dark incubation of river waters with and
without anion-exchange resin, where the values reported
in Table 8.3 was computed using either the value with
or without resin in equation (1), whichever value was
larger, and (c) AAP bioassays of autoclaved, filtered
river waters, using corrected Arn data from Table 58.
189
-------
Table 61. ESTIMATED PERCENT OF PP AVAILABLE, AS CALCULATEDa
FROM TREATMENTS OF UNFILTERED RIVER WATER SAMPLES
Sample No.
27
33
41
50
56
32
40
49
57
16
34
42
51
58
23
26
28
29
Percent of PP Available as :
DRP after fc DRP or R-DRP c
chloroforming after dark incubation
20
17
86
63
0
__
40
40
19
__
3
6
18
13
21
13
0
17
Niagara R. at Ft. Niagara
__
8
0
5
0
Niagara R. at Beaver I. Park
4
18
38
Genesee R.
41
10
0
0
13
Oswego R.
47
40
Algal-available P
in AAP bioassayd
0
0
0
0
0
0
0
0
0
_
1
0
21
0
__
0
25
(continued)
-------
Table 61. ESTIMATED PERCENT OF PP AVAILABLE, AS CALCULATEDa
FROM TREATMENTS OF UNFILTERED RIVER WATER SAMPLES
Percent of PP Available as :
DRP after , DRP or R-DRP Algal-available P
Sample No. chloroforming after dark incubation0 in AAP bioassay"-
31
35
43
47
52
54
55
59
0
28
60
33
20
28
36
28
40
40
15
0
9
0
68
0
15
15-33
7
Black R.
25
36
44
53
60
15
14
10
20
16
29
0
0
0
0
__ _
0
28
19
Calculated from the maximum mean observed values from the various
treatments, using equation (1) of the text.
Calculated from data in Table 52.
/»
Value given is the larger of the two values found with and without anion
exchange resin added to the samples; calculated w/Table 55 data
Calculated with data from Table 58.
-------
The value of 60 percent of PP available was net or ex-
ceeded only in four samples. However, it was difficult
to draw any conclusions about these results because
of the variability between samples from a given river.
Only the chloroform-treated Genesee R. and Black R.
samples appeared to show consistent results.
The PP availability calculated from the chloroform
treatment of Genesee R. Samples No.51 and No.58 was
18 and 13 percent, respectively, compared to values of
41 and 36 percent, respectively, for direct bioassays
of autoclaved PP (Table 49) . Sorption of some of
the phosphorus released by chloroform-induced lysis may
have been responsible for the low values from chloroform
treatment of these samples. The dark incubations of
unfiltered Genesee R. samples also showed very low
values of calculated DRP from PP, even though most of
the values in Table 61 were from test flasks, containing
anion-exchange resin (except Sample No.42, where the
incubation without resin produced slightly higher results),
The resin apparently had a negligible effect on the
long-term indications, however. Sorption by the suspended
soils may have been effective in competing with the
resin for phosphate. In the single test of Sample No.16
Genesee River water (300 ml) mixed with Lake Ontario
water (300 ml), relatively high proportion (40 percent)
of the PP was found to be converted to DRP. This sample
was collected in late summer of 1972, while all the other
samples from the Genesee R. were collected in spring, so
that possibly the nature of Sample No.16 was completely
different from that of the spring samples.
The Black R. PP forms appeared to be quite consis-
tent in terms of the percent of PP released as DRP after
chloroform treatment. However, the values from the
dark incubation systems (with and without resins) were
192
-------
low, except in the case of Sample No.25. This sample
was collected in late summer of 1972. Again the diff-
erence could possibly be related to the time of coll-
ection, although the values from chloroform treatment
were similar for all samples from the Black River.
Both the chloroform treatment and long-term dark
incubations without resin were designed to measure the
potentially available pool of phosphorus associated
with microbial cells. The long-term dark incubations
with resin were intended to measure, in addition, the
inorganic P extractable from suspended soil particles.
In nine of the 25 samples tested by dark incubation
(with and without resin) , the calculated DRP contri-
bution from PP was zero, and the presence or absence of
resin did not significantly affect the results for most
of the 25 samples (Table 55). Only in seven of the 25
samples did the dark incubation values exceed the chlor-
oform treatment values.
The algal-available P values measured in auto-
claved, filtered samples (AAP test, EPA, 1971) appeared
to be overestimates of the true available P, for
reasons given in Section VI. After correction of the
bioassay A7t-n values to a cell count basis to obtain
/ o u
corrected bioassay data, the corrected results (Table
58) were used in equation (1) to obtain the data
shown in Table 61). In many samples, only slight
changes in the TSP levels were noted as a result of
autoclaving (Table 56), although such treatment should
have produced roughly as much soluble P as the chloroform
treatment because both methods resulted in microorganism
cell lysis. In one sample, Niagara R. No.56, the TSP
showed a marked decrease as a result of the autoclave
treatment. Thus, the possibility exists that resorption
of some of the P released by autoclaving of particulate
193
-------
P forms in the sample could have caused some of the
low values shown in Table 61.
Another possible reason for the low values from
bioassays is the availibility of the TSP forms released
during autoclaving. Table 62 shows the chemically
measured DRP and TSP values compared to the bioassay
values of some autoclaved, filtered river waters. In
the case of the Niagara and Black R. samples, and Genesee
Sample No.58, it appears that the bioassay values were
closer to the DRP values than to the TSP values, in-
dicating that perhaps some of the "soluble" P released
from autoclaving was biologically not available to
Selenastrum. Since these forms could only be measured
after persulfate digestion, they may have been colloidal
or soluble organic P esters, perhaps refractory cellular
debris.
The net result of these factors was a relatively
low calculated PP availability (equation 1) from bio-
assays of autoclaved, filtered river waters (AAP test)
in comparison to direct bioassays of autoclaved suspensions
of PP forms in AAP (-P) medium. This comparison is
shown in Table 63. The combination of (a) resorption
of soluble P released during autoclaving (and thus its
loss via the filtration prior to bioassay) and (b) the
apparently low availability of the soluble, unreactive
P forms remaining after autoclaving was likely respon-
sible for some of the low results found with the stan-
dard AAP test and equation (1). The direct algal bio-
assay of autoclaved PP forms did not involve a filtra-
tion prior to bioassay, so that any resorbed P was still
potentially available to the test algae, as was the in-
organic P originally associated with the PP forms.
However, in both assay procedures, the considerations
expressed in part (b), above, would apply, since in both
194
-------
Table 62. COMPARISON BETWEEN DRP, TSP, AND
ALGAL-AVAILABLE P IN AUTOCLAVED, FILTERED
NEW YORK RIVER WATERS
Sample No.
DRP TSP Algal
fi, ,,r> /T \
-Available Pa
Niagara R. at Ft. Niagara
50
56
49
57
51
58
52
59
53
60
2 16
6 17
Niagara R. at Beaver
1 16
5 15
Genesee R.
103 113
52 68
Oswego R.
52 73
48 76
Black R.
18 37
30 51
<4
7
I. Park
<4
4
124
57
63-67
65
23
38
aCorrected values for all samples except those of the
Black R.; values from Table 58.
195
-------
Table 63. PERCENT OF RIVER WATER PP AVAILABLE TO
SELENASTRUM IN DIRECT BIOASSAYS OF AUTOCLAVED PP,
AS COMPARED TO CALCULATED PP AVAILABILITY IN
BIOASSAYS OF AUTOCLAVED, FILTERED RIVER WATER
Percent of PP Available to Selenastrum
Direct bioassay Calculated3 from bioassay of
of autoclaved PP autoclaved, filtered sample
Sample No. from river water of river water
50
56
51
58
52
59
53
60
Calculated
the text.
Niagara R. at Ft. Niagara
57 0
63 0
Genesee R.
41 21
36 0
Oswego R.
44 15-33
32 7
Black R.
45 28
26 19
from data in Table 58 and equation (1) of
n _ i_ n _ 1 1 r\
196
-------
methods the soluble P forms remaining after autoclaving
are the same. Thus, the differences seen in Table 63
between the two methods must have been the result of the
filtration step used in the standard AAP test.
The foregoing comparison indicates that the standard
AAP test (EPA, 1971) for estimating the potential avail-
able P from both soluble and particulate P forms in
natural waters may underestimate the contribution of
available P from the insoluble forms. The direct assay
of particulate forms isolated from the parent sample
and autoclaved in AAP(-P) medium probably is a better
procedure for determining the potential available P
from PP, especially since both the sample and standard
cultures contain essentially the same growth medium,
AAP(-P), whereas in the standard AAP test, the samples
contain AAP(-P) medium plus the nutrient salts from
the river water.
Total P Availability
The maximum observed values of DRP or algal-
available P from the various treatments of river waters
were also expressed as a percentage of the TP in the
samples. The direct bioassays of autoclaved PP were
fitted into this scheme by calculating the "A" values,
as shown in equation (2):
A = estimated available P in sample T TP X 100%
where:
est. avail. P = sample DRP + %PP avail.
to Selenastrum X ,
Sample PP after
autoclaving
This equation assumes that the dissolved unreactive P
(TSP - DRP) in the samples will not contribute signifi-
cantly to the available fraction of TP. Since, in
197
-------
general, TSP - DRP was a relatively small part of the
sample TP, the results calculated from equation (20 would
not be greatly affected if this assumption were false. It
can be stated that at_ least the P computed with equation
(2) should be available.
Table 64 compares the availability computed from all
the treatments of river waters, as well as the calculated
"A" values from direct algal assays of autoclaved PP- Al-
though the "S" value of TP availability from bioassay of
autoclaved, filtered water had been shown (Table 65 ) to
underestimate the PP availability of Samples Nos. 51, 52,
and 59, the TP availability measured by the "S" and "A"
values were in fairly close agreement for these samples.
Since the PP in these samples was only a fraction of TP,
differences in the measured PP availability would cause
smaller relative differences in TP availability. In the
Niagara R. samples, great differences in the "S" and "A"
values were seen, because of the factors discussed above
for the autoclaving-filtering ("S") procedure.
Figures 20 "to 2 3 present the data from Table 64 in
a graphic form. "S" values listed in Table 64 as "less
than X" were plotted in the figures as a horizontal line
at the value X, with an arrow labeled with an "S" pointing
toward zero.
The Niagara R. samples (Figure 20) showed more varia-
bility in the maximum observed TP availability than did the
other rivers. Chloroform treatment produced values of 26
to 91 percent of TP, while the dark incubations showed 12
to 51 percent of TP as DRP, and the AAP bioassays showed
very low values of 15 percent or less. The "A" value for
Sample No. 50 fell within the interval between the chloro-
form value and the dark incubation value, while the "A"
value for Sample No. 56 was slightly higher than the
198
-------
Table 64. COMPARISON OF CHEMICAL AND BIOLOGICAL TESTS FOR DETERMINING
THE AVAILABILITY OF TP IN NEW YORK RIVER WATERS
Maximum Mean DRP (% of TP)
Initial
Sample No. DRP
27
33
41
50
56
32
40
49
57
16
34
42
51
58
22 (r. mouth)
23
24 (r. mouth)
26
22
15
18
4
44
7
40
4
4
26
7
27
60
24
62
51
47
51
After chloro- After dark. After
form added incubation autoclaving
56
41
91
73
52
80
47
26
9
34
71
38
79
69
67
62
Niagara
35
23
31
51
Niagara
20
28
12
70
16
28
64
37
79
69
74
R. at Ft. Niagara
--
8
10
R. at Beaver I. Park
2
6
Genesee R.
60
26
Oswego R.
Algal-Available P (% of TP)
S
Value0
< 11
15
< 9
< 15
12
< 7
< 13
< 8
5
8
21
72
28
A
Value
46
59
75
50
-------
o
o
Table 64. COMPARISON OF CHEMICAL AND BIOLOGICAL TESTS FOR
DETERMINING THE AVAILABILITY OF TP IN NEW YORK
RIVER WATERS
Sample No.
28
29
31
35
43
47
52
54
55
59
Initial
DRP
73
74
45
45
40
42
48
46
36
31
Maximum Mean
After chloro-
form added
74
81
52
64
79
69
67
62
62
DRP (% of TP)
After dark,
. b
incubation
___
67
70
69
61
59
46
Algal- Available P (% of TP)
After
autoclavinq
__
50
33
S
ValueC
66
83
38
84
41
58
60-69
44
A
Value
68
50
Black R.
25
36
44
53
60
26
21
26
12
13
45
44
50
51
36
55
29
21
20
24
44
30
35
56
38
39
33
Data from Table 52.
"Value given is the larger of the two values found with and without anion exchange resin added
^to the samples; data from Table 55.
"S values are the results of Selenastrum bioassays of autoclaved, filtered river water; data
from Table 58.
A values are derived from equation (2) in the text, using PP bioassay (autoclaved PP) data
from Table 49.
-------
Figures 20 to 23. Comparison of chemical and biological tests for
determining the availability of TP in New York
river waters
KEY:
(D INITIAL DRP IN THE SAMPLE ( Mean Value)
A DRP AFTER CHLOROFORM ADDITION! Maximum Observed Mean Value)
D DRP AFTER DARK INCUBATION ( Maximum Observed Mean Value )
DRP AFTER AUTOCLAVING ( Mean Value)
(?) S VALUE; MEAN VALUE OF SELENASTRUM BIOASSAY OF AUTOCLAVED,
FILTERED RIVER SAMPLE (see Table 64 )a
@ A VALUE ; COMPUTED FROM BIOASSAY OF PP FORMS IN AAP ( -P )
SUSPENSIONS ( see Table 64 )
QValues of S in Table 64 which are listed as
-------
Figure 20.
a.
l-
u_
o
LJ
O
cr
LJ
Q-
100
90
80
70
60
50
40
30
20
10 -
32 40 49
BEAVER I. STATE PARK
57
27
33 41
FORT NIAGARA
50
56
NIAGARA RIVER SAMPLE NO.
-------
Figure 21.
a.
i
1
ii
O
1-
LU
O
a:
bJ
0.
100
90
80
70
60
50
40
30
20
10
T i
©
($
j
j
A
O T-© $
i
©J-©
i i i i i
16 34 42 51
GENESEE RIVER SAMPLE NO.
58
-------
Figure 22.
100
90
80
70
0,
U. 60
O
H 50
LJ
DC 40 -
LJ
a.
30 -
20 -
10 -
© Q)
22 24 23 26 28 29 31 35 43 47 52 54 55 59
(R. MOUTH)
OSWEGO RIVER SAMPLE NO.
-------
to
o
a.
\-
I00r
90
80
70
60
Z 50
LU
O
CE 40
LU
CL
30
20
10
JL
Figure 23.
_L
_L
J_
25 36 44 53
BLACK RIVER SAMPLE NO.
60
-------
chloroform value. The initial DRP values in these samples
were less than 50 percent of TP; in all samples tested by
chloroform or dark incubation treatments the initial DRP
was exceeded by the DRP after treatment.
Like the Niagara R. samples, the Genesee R. samples
(Figure 21 ) showed a wide range of observed TP availability
because of the changing nature of the PP in the samples and
the changing concentrations of soluble P forms in the sam-
ples. Chloroform treatment showed a range of 9 to 71 per-
cent of TP as DRP, similar to the bioassay "S" value range
of 8 to 72 percent of TP. The calculated "A" values were
within 13 percent of the chloroform or dark incubation va-
lues, for Samples No. 51 and No. 58.
The Oswego R. samples (Figure 22 ) generally showed
fairly consistent TP availability, with chloroform values
between 52 and 81 percent of TP and dark incubation values
between 46 and 79 percent. Except in the case of Samples
No. 31 and No. 43, the "S" values generally were in the
same range as the chloroform or dark incubation values.
The range of "S" values for all samples except No. 31 and
No. 43 was 58 to 83 percent. The "A" values fell into the
interval between the chloroform values and the dark incu-
bation values, for Samples No. 52 and No. 59. The two
river mouth samples did not appear to differ significantly
from the river samples.
Black R. samples (Figure 23) showed relatively con-
sistent TP availability, with chloroform values between
36 and 51 percent of TP, and dark incubation values between
20 and 55 percent of TP. The "S" values agreed closely
with the chloroform treatment values, but were higher than
the dark incubation values. The "A" values fell between
the chloroform and dark incubation values for Samples No. 53
and No. 60.
206
-------
In summary, the Oswego and Black River samples tested
for TP availability showed relatively consistent values be-
tween samples collected at various times during these stu-
dies. The Niagara and Genesee samples showed greater varia-
bility, such that an intensive sampling schedule would have
to be established to adequately characterize the availabili-
ty of P forms in these rivers throughout even the spring
flow period. These studies indicate that generalizations
about either PP or TP availability in the New York rivers
would have to be weighed to account for the TP or PP load-
ing to Lake Ontario for the time interval during which the
measurements of availability were made. Only in this way
could an accurate nutrient budget, based on [(concentra-
tion) X (% availability) X (flowV] data be constructed.
207
-------
SECTION IX
REFERENCES
Alexander, M. 1961. Introduction to Soil Microbiology.
New York, John Wiley and Sons, Inc. 472 p.
Allen, S.E., A. Carlisle, E.J. White, and C.C. Evans.
1968. The Plant Nutrient Content of Rainwater. J.
Ecology. 5_6_: 497-504.
Amer, F., D.R. Bouldin, C.A. Black, and F.R. Duke. 1955.
Characterization of Soil Phosphorus by Anion Exchange
Resin Adsorption and P32 Equilibration. Plant and
Soil. 6_: 391-408.
American Public Health Association, American Water Works
Association, Water Pollution Control Federation.
1971. Standard Methods for the Examination of Water
and Wastewater. 13th ed. New York, American Public
Health Association. 874 p.
American Public Works Association-Research Foundation.
1969. Water Pollution Aspects of Urban Runoff. U.S.
Dept. of Interior, FWPCA Publication No. WP-20-15.
272 p.
Anderson, G. 1967. Nucleic Acids, Derivations, and
Organic Phosphates. In: Soil Biochemistry, A.D.
McLaren and G.H. Peterson (eds).New York, Marcel
Dekker, Inc. p. 67-90.
Anderson, G. and E.Z. Arlidge. 1962. The Adsorption of
Inositol Phosphates and Glycerophosphates by Soil
Clays, Clay Minerals and Hydratecl Sesquioxides in
Acid Media. J.Soil Sci. 13: 216-224.
Anonymous (No Title). 1971. Environment.: 1J3: 28-29.
AVCO Economic Systems Corp. 1970. Storm Water Pollution
from Urban Land Activity. U.S. Dept. of Interior,
FWQA Publication No. 11034FKL 07/70. 325 p.
208
-------
Baalsrud, K. 1967. Influence of Nutrient Concentrations
on Primary Production. In: Pollution and Marine
Ecology, T.A. Olson and F.J. Burgess (eds.).New York,
Interscience, Inc. p. 159-169.
Benzie, W.J. and R.J. Courchaine. 1966. A Study of
the Discharges from Separate Storm Sewers and Com-
bined Sewers. J. WPCF. 38: 419-421.
Berman, T. 1970. Alkaline Phosphatases and Phosphorus
Availability in Lake Kinneret. Limnol. Oceanogr.
15: 663-674.
Boulton, P. Personal Communication (New York State Dept.
of Environmental Conservation, Albany, N.Y.),
July 31, 1962.
Bower, C.A. 1949. Studies on the Forms and Availability
of Soil Organic Phosphorus. Iowa Agr- Exp. Sta.
Res. Bull. 362: 963-996.
Burm, R.J., D.R. Krawczyk, and G.L. Harlow. 1968.
Chemical and Physical Comparison of Combined and
Separate Sewer Discharges. J. WPCF. 40: 112-126.
Bryan, E.H. 1970. Quality of Storm Water Drainage from
Urban Land Areas in North Carolina. Univ. of North
Carolina Water Resources Research Inst., Publication
No. 37. Chapel Hill. 75 p.
Carlisle, A., A.H.F. Brown, and E.J. White. 1966a.
Litter Fall, Leaf Production and the Effects of
Defoliation by Tortix Viridana in a Sessile Oak
(Quercus Petrea) Woodland. J. Ecol. 54: 65-75.
Carlisle, A., A.H.F. Brown, and E.J. White. 1966b. The
Organic Matter and Nutrient Elements in the Precip-
itation Beneath a Sessile Oak (Quercus Petrea)
Canopy. J. Ecol. _5_4: 87-98.
Casey, D.J. Personal Communication (Environmental Protec-
tion Agency, Rochester, N.Y.) July 11, 1973.
Chalupa, J. 1960. Eutrophication of Reservoirs by
Atmospheric Phosphorus. Sci. Pap. Inst., Chem.
Technol. Fuel and Water (Czech.). 4_: 295-308,
Water Poll. Abstracts. 35: 660.
Chamberlain, W.M. 1968. A Preliminary Study of the
Nature and Importance of Soluble Organic Phosphorus
in the Phosphorus Cycle of Lakes. Ph.D. Thesis.
Univ- of Toronto, Canada.
209
-------
Chamberlain, W. and J. Shapiro. 1969. On the Biological
Significance of Phosphate Analysis: Comparison
of Standard and New Methods with a Bioassay. Limnol.
Oceanogr. IMj 921-927.
Chamberlain, W. and J. Shapiro. 1973. Phosphate
Measurements in Natural WatersA Critique. In:
Environmental Phosphorus Handbook, E.J. Griffith,
A. Breton, J.M. Spencer, and D.T. Mitchell (eds.).
New York, John Wiley and Sons. p. 355-566.
Chang, S.C. and M.L. Jackson. 1957. Fractionation of
Soil Phosphorus. Soil Sci. 84: 133-144.
Cooke, I.J. and J. Hislop. Io63. Use of Anion Exchange
Resin for the Assessment of Available Soil Phosphate.
Soil Sci. 3b_: 308-312.
Cowen, W.F. and G.F. Lee. 1973. Leaves As a Source of
Phosphorus. Environ. Sci. and Technol. 7_: 853-854.
Evans, F.L., E.E. Geldreich, S.R. Weibel, and G.G.
Robeck. 1970. Treatment of Urban Stormwater Runoff.
J. WPCF. 40: R162-R170.
Fitzgerald, G.P. 1970a. Aerobic Lake Muds for the Removal
of Phosphorus from Lake Waters. Limnol. Oceanogr.
15_: 550-555.
Fitzgerald, G.P. 1970b. Evaluation of the Availability of
Sources of Nitrogen and Phosphorus for Algae. J.
Phycol. 6_: 239-247.
Goldman, C.R., M.G. Tunzi, and R. Armstrong. 1969. C
Uptake As a Sensitive Measure of the Growth of Algal
Cultures. In: Proceedings of the Eutrophication-
Biostimulation Assessment Workshop, June 19-21, 1969,
E.J. Middlebrooks, T.E. Maloney, C.F. Powers, and
L.M. Kaack (eds.). U.S. Federal Pollution Control
Administration, Corvallis, Oregon. p. 158-170.
Golterman, H.L. 1960. Studies on the Cycle of Elements
in Fresh Water. Acta Botanica Neerlandica. 9: 1-58
Golterman, H.L., C.C. Bakels, and J. Jakobs-Mogelin. 1969.
Availability of Mud Phosphates for the Growth of
Algae. Verh. Internat. Verin. Limnol. 17: 467-479.
Goring, C.A.I., and W.V. Bartholomew. 1950. Microbial
Products and Soil Organic Matter, III. Adsorption of
Carbohydrate Phosphates by Clays. Soil Sci. Soc. Am.
Proc. 15: 189-194.
210
-------
Goring, C.A.I., and W.V. Bartholomew. 1952. Adsorption of
Monomucleotides, Nucleic Acids, and Nucleoproteins
by Clays. Soil Sci. 74: 149-164.
Greaves, M.P., G. Anderson, and D.M. Webly. 1963. A Rapid
Method for Determining Phytase Activity of Soil
Microorganisms. Nature. 200: 1231-1232.
Great Lakes Water Quality. 1972. Agreement between U.S.
and Canada, signed at Ottawa Apr. 15, 1972. Treaties
and other International Acts Series No. 7132.
Halstead, R.L. 1964. Phosphatase Activity of Soils as
Influenced by Lime and Other Treatments. Can. J. Soil
Sci. 44: 137-143.
Harrison, M.J., R.E. Pacha, and R.Y. Morita. 1972. Solu-
bilization of Inorganic Phosphates by Bacteria Isolated
from Upper Klamath Lake Sediment. Limnol. Oceanogr.
17: 50-57.
Hislop, J. and I.J. Cooke. 1968. Anion Exchange Resin as a
Means of Assessing Soil Phosphate Status: A Laboratory
Technique. Soil Sci. 105: 8-11.
International Joint Commission of the Pollution of Lake Erie,
Lake Ontario, and the International Section of the St.
Lawrence River. 1969. Vol. 3. Lake Ontario and the
International Section of the St. Lawrence River.
(Report to the IJC by the Internat. L. Erie Water Pol-
lution Board and the Internat. L. Ontario-St. Lawrence R.
Water Pollution Board). 329 p.
Jackman, R.H. 1949. Mineralization of Inositol-bound Phosphorus
in Soil. Ph.D. Thesis, Iowa State Univ., Ames'.
Jackman, R.H. and C.A. Black. 1951a. Solubility of Iron,
Aluminum, Calcium and Magnesium Inositol Phosphates at
Different pH Values. Soil Sci. 72_: 179-186.
Jackman, R.H. and C.A. Black. 1951b. Hydrolysis of Iron,
Aluminum, Calcium, and Magnesium Inositol Phosphates by
Phytase at Different pH Values. Soil Sci. 7_2_: 261-266.
Jaworski, N.A. and L.J. Hetling. 1970. Relative Contributions
of Nutrients to the Potomac River Basin from Various
Sources. U.S. Dept. of Interior, FWPCA, Middle Atlantic
Region. Tech. Rep. No. 31. 36 p.
Johnson, R.E., A.T. Rossano, and R.O. Sylvester. 1965. Dust-
fall as a Source of Water Quality Impairment. Nat. Symp.
on Sanitary Engineering Research, Development and Design.
Univ. Park, Pa.
211
-------
Kluesener, J.W. 1971. Nutrient Transport and Transformations
in Lake Wingra, Wisconsin. Ph.D. Thesis, Univ. of Wis. ,
Madison.
Kuenzler, E.J. and B.H. Ketchum. 1962. Rate of Phosphorus
Uptake by Phaeodactylum tricornutum. Biol. Bull. 12 3 :
134-145.
Kunishi, H.M., A.W. Taylor, W.R, Heald, W.J. Gburek, and
R.N. Weaver. 1972. Phosphate Movement from an Agri-
cultural Watershed During Two Rainfall Periods. J. Agr.
Food Chem. 2Q_: 900-901.
Lathwell, D.J., N. Sanchez, D.J. Lisk, and M. Peech. 1958.
Availability of Soil Phosphorus as Determined by Several
Chemical Methods. Agron. 50: 366-369.
Lawton, K. 1961. Plant Nutrition and the Utilization of
Fertilizer Phosphorus. In: Phosphorus and Its Com-
pounds, Vol. II. Technology, Biological Functions and
Applications, J.R. Van Wozer (ed.). New York, Inter-
science Publishers, Inc. p. 1461-1545.
Li, W.C., D.E. Armstrong, J.D.H. Williams, R.F. Harris, and
J.K. Syers. 1972. Rate and Extent of Inorganic Phos-
phate Exchange in Lake Sediments. Soil Sci. Soc. Am.
Proc. ^6_: 279-285.
Martin, J.B., Jr., B.N. Bradford, and H.G. Kennedy. 1969.
Factors Affecting the Growth of Najas in Pickwick
Reservoir- Nat. Fertilizer Develop. Center, TVA, Muscle
Shoals, Ala. 47 p.
Massey, H.F. and M.L. Jackson. 1952. Selective Erosion of
Soil Fertility Constituents. Soil Sci. Soc. Am. Proc.
I_6_: 353-356.
Murphy, J. and J.P. Riley. 1962. A Modified Single Solution
Method for the Determination of Phosphate in Natural
Waters. Anal. Chim. Acta. 2T_: 31-36.
National Oceanographic and Atmospheric Administration. 1971.
National Ocean Survey, Lake Survey Center: Lake Ontario
Chart No. 2, Ann Arbor, Michigan.
Owen, G.E. and M.G. Johnson, 1966. Significance of Some Fac-
tors Affecting Yields of Phosphorus from Several Lake
Ontario Watersheds. Great Lakes Research Division, Univ.
of Mich., Ann Arbor. Publ. No. 15. p. 400-410.
212
-------
Phillips, J.E. 1964. The Ecological Role of Phosphorus in
Waters with Special Reference to Microorganisms. In:
Principles and Applications in Aquatic Microbiology,
H. Heukelekian, and N.C. Dondero (eds.).New York,
John Wiley and Sons. p. 61-81.
Plumb, R.H., Jr. 1973. Effects of Taconite Tailings on
Water Quality. Ph.D. Thesis, Univ. of Wis., Madison.
Renn, C.E. 1937. Bacteria and the Phosphorus Cycle in the
Sea. Biol. Bull. 72: 190-195.
Rigler, F.H. 1966. Radiobiological Analysis of Inorganic
Phosphorus in Lake Water. Verh. Internal Verein.
Limnol. 16: 465-470.
Rigler, F.H. 1968. Further Observations Inconsistent with
the Hypothesis that the Molybdenum Blue Method Measures
Orthophosphate in Lake Water. Limnol. Oceanogr. 13;
7-13.
Rodina, A.G. 1963. Microbiology of Detritus of Lakes.
Limnol, Oceanogr. 8_: 388-393.
Rogers, H.T. 1942. Dephosphorylation of Organic Phosphorus
Compounds by Soil Catalysts. Soil Sci. 54: 439-446.
Roth, C.B., M.L. Jackson, and J.K. Syers. 1969. Deferration
Effect on Structural Ferrous-Ferric Iron Ratio and CEC
of Vermiculites and Soils. Clays, Clay Min. 17: 253-
264.
Ryden, J.C., J.K. Syers, and R.F. Harris. 1972. Nutrient
Enrichment of Runoff Waters by Soils. Phase I: Phos-
phorus Enrichment Potential of Urban Soils in the City
of Madison. U. of Wis. Water Resources Center, Madison,
Wis. Tech. Rep. No. OWRR A-038-Wis. 70 p.
Sartor, J.D. and G.B. Boyd. 1972. Water Pollution Aspects of
Street Surface Contaminants. U.S. Environmental Pro-
tection Agency, Washington D.C. Rep. No. EPA-R-72-081.
326 p.
Sawyer, C.N. 1947. Fertilization of Lakes by Agricultural
and Urban Drainage. New Eng. Water Works Assn. 61:
109-127.
213
-------
Scarseth, G.D. and W.V. Chandler., 1938. Loss of Phosphate
from a Light-Textured Soil in Alabama and Its Relation
to Some Aspects of Soil Conservation. Agron. J. 30:
361-371}.
Shannon, J.E. and G.F. Lee. 1966. Hydrolysis of Condensed
Phosphates in Natural Waters. Air Water Poll. 10: 735-
756.
Shukla, S.S., J.K. Syers, J.D.H. Williams, D.E. Armstrong,
and R.F. Harris. 1971. Sorption of Inorganic Phosphate
by Lake Sediments. Soil Sci. Soc. Am. Proc. 35; 244-
249.
Skulberg, O.M. 1964. Algal Problems Related to the Eutro-
phication of European Water Supplies and a Bioassay
Method to Assess Fertilizing Influences of Pollution on
Inland Waters. In: Algae and Man, D.F. Jackson, (ed.).
New York, Plenum Press, p. 262-299.
Sonzogni, W.C. and G.F. Lee. 1972. Nutrient Sources for Lake
Mendota-1972. Univ. of Wis., Madison. Water Chemistry
Program Report. 49 p.
Spear, R.D. 1970. The Release of Phosphorus from Lake Sedi-
ments. Ph.D. Thesis. Univ. of Wis., Madison.
Stumm, W. and J.J. Morgan. 1970. Aquatic Chemistry. New
York, Wiley-Interscience. 583 p.
Sylvester, R.O. and G.C. Anderson. 1964. A Lake's Response
to Its Environment. J. San. Eng. Div. Proc. ASCE. 90:
SAI, 1-22.
Tamm, C.O. 1951. Removal of Plant Nutrients from Tree Crowns
by Rain. Physiol. Plantarum. 4_: 184-188.
Taylor, A.W. and H.M. Kunishi. 1971. Phosphate Equilibria
on Stream Sediment and Soil in a Watershed Draining
and Agricultural Region. J.Agr. Food Chem. 19: 827-
831.
Thompson, J.R. 1970. Soil Erosion in the Detroit Metropolitan
Area. J. Soil Water Conservation. 25: 8-10.
Thompson, L.M. and C.A. Black. 1949. The Mineralization of
Organic Phosphorus, Nitrogen and Carbon in Clarion and
Webster Soils. Soil Sci. Soc. Am. Proc. 14: 147-151.
214
-------
Timmons, D.R., R,F. Holt and J,J. Latterell. 1970. Leach-
ing of Crop Residues as a Source of Nutrients in
Surface Runoff Water. Water Resources Research, 6:
1367-1375. ~
U.S. Environmental Protection Agency. 1971. Algal Assay
Procedure Bottle Test. National Eutrophication Res.
Program, Corvallis, Oregon. 82 p.
Waksman, S.A., J.L.Stokes, and M.R. Butler. 1937. Relation
of Bacteria to Diatoms in Sea Water. J. Marine Biol.
Assoc. 22: 359-373.
Walton, C.P. and G.F. Lee. 1972. A Biological Evaluation of
the Molybdenum Blue Method for Orthophosphate Analysis.
Verh. Internat. Verein. Limnol. 18: 676-684.
Weibel, S.R. 1969. Urban Drainage as a Factor in Eutro-
phication. In: Eutrophication: Causes, Consequences,
Correctives. Nat. Acad. of Science, Washington, D.C.
p. 383-403.
Weibel, S.R., R.J. Anderson, and R.L. Woodward. 1964. Urban
Land Runoff as a Factor in Stream Pollution. J. WPCF.
36: 914-924.
Weibel, S.R., R.B. Weidner, J.M. Cohen, and A.G. Christiansen.
1966. Pesticides and Other Contaminants in Rainfall
and Runoff. J. AWWA. 5_8_: 1075-1084.
Weil-Malherbe, H. and R.H. Green. 1951. The Catalytic
Effect of Molybdate on the Hydrolysis of Organic Phos-
phate Bonds. Biochem. J. 49: 286-292.
Wentz, D.A. 1967. Available Phosphorus in Lake Sediments.
M.S. Thesis. Univ. of Wis., Madison. 106 p.
Will, G.M. 1959. Nutrient Return in Litter and Rainfall
Under Some Exotic Confier Stands in New Zealand. N.Z.
J. Agric. Research. _2_: 719-734.
Williams, E.G., N.M. Scott, and M.J. McDonald. 1958. Soil
Properties and Phosphate Sorption. J.Sci. Food Agr.
_9_: 551-559.
Williams, J.D.H., J.K. Syers, D.E. Armstrong, and R.F.
Harris. 1971a. Characterization of Inorganic Phos-
phate in Noncalcareous Lake Sediments. Soil Sci. S.oc.
Am. Proc. 35: 556-561.
215
-------
Williams, J.D.H., J.K. Syers, R.F. Harris, and D.E.
Armstrong. 1971b. Fractionation of Inorganic Phos-
phate in Calcareous Lake Sediments. Soil Sci. Soc.
Am. Proc. 35: 250-255.
Williams, J.D.H., J.K. Syers, and T.W. Walker. 1967.
Fractionation of Soil Inorganic Phosphate by a Modi-
fication of Chang and Jackson's Procedure. Soil Sci.
Am. Proc. _3]L: 736-739.
Wolman, M.G. and A.P. Schick. 1967. Effects of Construc-
tion on Fluvial Sediment, Urban and Suburban Areas of
Maryland. Water Resources Res. 3: 451-4-6H.
216
-------
SECTION X
APPENDICES*
Page
A Statistical Formulae and Other Special 218
Calculations
B Bioassays of Soluble and Particulate 221
Phosphorus with S_. Capri cor nut urn
C Dark Incubations of Unfiltered Runoff 267
and River Water Samples
D Chloroform Treatment of New York River 283
Water Samples
':In order to reduce printing costs, appendices have not
been included in this report. They may be obtained from
Project Officer, Nelson A. Thomas, Large Lakes Research
Station, 9311 Groh Road, Grosse lie, Michigan 48138.
217
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/3-76-094a
2.
3. RECIPIENT'S ACCESSIOr+NO.
4. TITLE AND SUBTITLE
ALGAL NUTRIENT AVAILABILITY AND
LIMITATION IN LAKE ONTARIO DURING IFYGL
Part I. Available Phosphorus in Urban Runoff
5. REPORT DATE
October* 1 Q7fi
6. PERFORMING ORGANIZATION CODE
d Lak
T-rih
7. AUTHOR(S)
William F. Cowen*and G. Fred Lee
VU.S. Army Medical Bioengineering Research and
^ T -1- ^j- De-|-rirk . Frederi nk Md .
8. PERFORMING ORGANIZATION
D
_L
_E
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Center for Environmental Studies
University of Texas at Dallas
Richardson, Texas 75080
10. PROGRAM ELEMENT NO.
1BA608
11. CONTRACT/GRANT NO.
Contract
R-800537-02
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
Environmental Research Laboratory-Duluth
Office of Research and Development
U.S. Environmental Protection Agency
Duluth, Minnesota 558Q4
Final Report
14. SPONSORING AGENCY CODE
EPA-ORD
15. SUPPLEMENTARY NOTES
is. ABSTRACT Samples of Madison urban runoff, precipitation from Madison and
New York State were analyzed for various forms of phosphorus to esti-
mate the algal-available fraction of each of these P forms. Urban run-
off particulate P forms from Madison runoff showed acid extractable in-
organic P in the range of 33 to 46% of the particulate P. Ranges for the
OH'and for exchange resin extractable inorganic P were 22 to 27 and 13
to 17 % of particulate P, respectively. Runoff from urban areas in the
Genesee R. basin (N.Y.) showed acid, base, and resin extractable inor-
ganic P in the ranges of 30 to 48, 18 to 30 and 11 to 25% of particu-
late P, respectively, in general agreement with the Madison samples.
Inorganic P extracted from particulate P by resin in long-term aerobic
dark incubations was similar to that extracted by the resin in short-
term tests, indicating that physical and chemical rather than microbial
mineralization processes were probably the key factors regulating the
release of inorganic P from the runoff particles to the solution phase.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COS AT I Field/Group
Precipitation (meteorology), Runoff,
Phosphorus.
Madison Wise, New York
State, Lake Ontario
06F
07B
13. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. OF PAGES
232
20. SECURITY CLASS (This page)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
218
U.S. GOVERNMENT PRINTING OFFICE: 1976-757-056/5ltl)lt Region No. 5-I
------- |