905R87100
00522
                      VARIATION AMONG AND WITHIN PROCEDURES FOR
                      BIOAVAILABLE PHOSPHORUS

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
VARIATION AMONG AND WITHIN PROCEDURES FOR ESTIMATION

             OF BIOAVAILABLE PHOSPHORUS
                         by
                   Thomas C. Young
                 Joseph V. De Pinto
                   Bryan J. Hughes
  Department of Civil and Environmental Engineering
                 Clarkson University
                 Potsdam, NY  13676
             EPA Grant Number R005761-01
                   Project Officer


                  Marcella Gewirth
         Great Lakes National Program Office
                      Region V
              Chicago, Illinois  60605
                     March, 1987
       U.S. Environmental Protection Agency
        Great  Lakes  National Program Office
                  GLNPO  Library

-------
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
I
I
I
I
I
                                  DISCLAIMER


     This report has been reviewed by the Great Lakes National Program Office,
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 of trade names
or commercial products constitute endorsement or recommendation for use.

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
                                    FORWARD


     The Environmental Protection Agency was established to coordinate
administration of the major Federal programs designed to protect  the quality
of our environment.


     An important part of the Agency's effort involves the search for
information about environmental problems, management techniques,  and new
technologies through which optimum use of the nation's land and water
resources can be assured and the threat pollution poses to the welfare of  the
American people can be minimized.


     The Great Lakes National Program Office (GLNPO) of the United States
     Environmental Protection Agency was established in Region V,  Chicago,  to
     focus attention on the significant and complex natural resource
     represented by the Great Lakes.


     GLNPO implements a multi-media environmental management program drawing
     on a wide range of expertise represented by universities, private firms,
     State, Federal, and Canadian governmental agencies, and the  International
     Joint Commission.  The goal of the GLNPO program is to develop programs,
     practices and technology necessary for a better understanding of the
     Great Lakes Basin ecosystem and to eliminate or reduce to the maximum
     extent practicable the discharge of pollutants into the Great Lakes
     system.  GLNPO also coordinates U.S. actions in fulfillment  of the Great
     Lakes Water Quality Agreement of 1978 between Canada and the United
     States of America.


     We hope that the information and data contained herein will  help planners
and managers of pollution control agencies make better decisions  for carrying
forward their pollution control responsibilities
                                   Peter L.  Wise
                                   Director
                                   Great Lakes National Program Office
                                               11

-------
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
I
I
I
                                   AB STRACT

     One bioassay and five chemical extraction procedures for estimation of
biologically available particulate phosphorus (BAPP) were applied to 12
samples of aquatic particulate matter from the lower Great Lakes region.  The
determinations were made to provide a basis for comparing or converting
estimates among the procedures.  Although the procedures extracted widely
differing amounts of phosphorus (P), the results indicate that accurate
comparisons or regression conversions may be made among procedures for most
samples, but not for all samples.   In fresh samples amounts of P extracted by
the procedure of De Pinto were consistently closest in magnitude to the
amounts taken up by algae during the bioassays.   On both fresh and archived
samples the procedure of Armstrong gave results that overestimated but
correlated most closely with the bioassay results.

     The effects of storage time (0 to 9 days) and temperature (4, 22, and
45 C) on concentrations of soluble reactive P (SRP) and BAPP were examined
using unfiltered water samples from two rivers (Maumee and Huron) in the
western basin of Lake Erie.  The results showed significant changes in
concentrations of the two forms of P for most combinations of time and
temperature of holding, and the major changes occurred at 45 C for samples
from both rivers.  It was evident from the results that any period of sample
storage could affect the reliability of SRP estimates in river samples held at
45 C.  The results support the procedures recommended by USEPA and Standard
Methods for the handling of water samples collected for P analysis.  Observed
changes in SRP during storage, however, were offset partially by inverse
changes in BAPP, indicating that for storage times of no more than 9 days, the
total bioavailable P (BAP) of water samples (SRP+BAPP) may be partially
conserved.
                                     111

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
                                   CONTENTS

Foreword                                                                    ii
Abstract                                                                   iii
List of Figures                                                              v
List of Tables                                                              vi
Acknowledgements                                                           vii
    1.  Introduction                                                         1
    2.  Conclusions                                                          6
    3.  Recommendations                                                      9
    4.  Methods and Materials                                               11
    5.  Results and Discussion                                              18
References                                                                  39
Appendices                                                                  41
    A.  Description of Chemical Extraction Procedures                      A.I
    B.  Concentrations of SRP during Storage Effects Study                 B.I
    C.  Concentrations of BAPP during Storage Effects Study                C.I
    D.  Concentrations of Extractable and Bioassay Determined BAPP         D.I
    E.  Statistical Analyses:  Analysis of Variance Tables                 E.O
        E.I Double Split-Plot ANOVA of Storage Effects on SRP            E.I.I
        E.2 Double Split-Plot ANOVA of Storage Effects on BAPP           E.2.1
        E.3 Double Split-Plot ANOVA of Storage Effects on
            BAPP/TSS                                                     E.3.1
        E.4 ANCOVA of Source Variables and SRP During Storage            E.4.1
        E.5 Factorial ANOVA of Chemical Extraction Data                  E.5.1
        E.6 Correlations Among Chemical Extraction and Bioassay
            Results                                                      E.6.1
    F.  Comparison of Field and Laboratory Filtration Units for
        SRP Analysis                                                       F.I
                                               IV

-------
                                    FIGURES



Number                                                                    Page
I


I


I


I


•          1   Location  of sampled  systems  in  the lower Great Lakes region.           13


            12   Concentration of  SRP in Huron River samples as a function of
                temperature and  holding time during storage effects study.            20


•          3   Concentration of  SRP in Mautnee  River samples as a function of
                temperature and  holding time during storage effects study.            21


•          4   Regression of incremental changes in BAPP on SRP during
                storage  effects  study.                                                24

I
            5   Amounts of P measured by each procedure, averaged over all
mm              sediment samples.                                                     27


            6   Amounts of P measured on each sample, averaged over all
•              analytical procedures.                                                28
 7   Comparison of extractable P among all sediment samples.                29





 8   Illustration of bioassay results for determination of

      algal-available P.                                                     32
I



I


           19   Comparison of chemical and biological measurements of
                BAPP for all samples.                                                 35



|        10   Comparison of chemical and biological measurements of BAPP.            38



I



I



I



I

-------
 I
 I
 I
                                              TABLES
 •         Number                                                                    Page
 •          1    Methods  for  determination  of BAPP  in the Great Lakes  region.           12
            12    Chemical characteristics of water  samples collected for  storage
                effects study.                                                        19
 •          3    Summary  of BAPP  analyses,  ug P/g.                                      25
 •          4    Algal  uptake of  sediment P during  bioassay experiments,  ug P/g.        31
            15    Average  algal-available P  released by sediments during bioassay
                experiments, ug P/g  (from Martin  et_ al. 1983).                        33
 I          6    All pair-wise comparison of extraction procedures.                     37
 I
 I
 I
 I
 I
 I
I
I
I

-------

                               ACKNOWLEDGEMENTS

     The cooperation and assistance of many persons were essential in
completing this project.  Special thanks are due to Dr.  Bill Richardson,  Dr.
Mike Mullin, and other U.S.  EPA personnel at the ERL-Duluth Large Lakes
Research Station (Grosse lie, MI) for the use of their facilities during the
storage effects investigation; to Dr. Dave Baker for his willingness in
supplying the manpower necessary to collect and send samples from the Ohio
Rivers for use during the study of extraction methods; to Mr.  Dave Payne of
U.S. EPA Region V Quality Assurance Office for assistance in designing the
analytical program; and to Ms. Marcella Gewirth, our Project Officer, from the
Great Lakes National Program Office, for her interest in the project and
guidance during its implementation.
                                     Vll

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
                                   SECTION 1

                                 INTRODUCTION
BACKGROUND

     For at least two decades, eutrophication of freshwater lakes and rivers
has been an issue of national concern.  A result of nutrient enrichment of
water, eutrophication causes excessive production of algae and aquatic plants,
depletion of dissolved oxygen, deterioration of fishery quality, and
impairments to other beneficial uses of water resources.  High nutrient inputs
generally may be traced to intensified agriculture, urbanization, and direct
discharges of a variety of wastes, including industrial wastewater, raw
sewage, and municipal wastewater treatment plant effluents.  Of the various
nutrients that these sources provide to receiving lakes and rivers, phosphorus
(P) generally has been determined as the key element giving rise to eutrophic
conditions.

     In response to the growing problem of eutrophication in the lower Great
Lakes  (Lakes Erie and Ontario), P discharges from point sources, such as
wastewater treatment plants, have been reduced.  Specifically, the
International Joint Commission has recommended the reduction of total P
concentrations in municipal wastewater effluents to 1.0 mg P/L and is
considering further reductions (LJC 1970, 1972, 1978).  Additionally, on the
recommendation of the IJC  (1978), detergent P levels in the lower Great Lakes
basin have been reduced to help reduced P loads to the lakes.  This has been
accomplished by legislative actions of most, though not all, states and
provinces in the basins of the lakes.

     It is important to realize that the total, analytically-defineable
quantity of P in a sample of natural water includes a variety of P forms, and
these are not equally effective in stimulating growth of algae and other
aquatic organisms.  Orthophosphate, or soluble, monomeric inorganic P
(H P0-x  ), is the principal form of P that is transported across
cell membranes and, as such, is immediately available for direct uptake by
algae and other primary producers (Wetzel 1983).  As a consequence, elevated
levels or availability of orthophosphate may be cited as the proximate cause
of eutrophic conditions in most cases.

     Much of the P contained in municipal treatment plant effluents, often
over 70 percent, occurs as orthophosphate (analytically approximated as the
soluble, molybdate-reactive fraction of total P, ie. SRP), and these sources
have been shown (Young and De Pinto 1982) to contain relatively high levels of
biologically available P (BAP).  In contrast, however, only a small fraction
of the P in unpolluted fresh waters, usually less than 10 percent, is present
as orthophosphate (Wetzel 1983).   Nonetheless, as shown by De Pinto et
al. (1981) among other investigators, substantially more P than the
orthophosphate fraction in natural waters is biologically available.
Presumably, this can occur only through conversion of various soluble and
insoluble particulate-bound forms to orthophosphate prior to biological
uptake.

-------
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
I
     Recent studies have indicated that much of the P load entering the lower
Great Lakes is from diffuse or non-point sources.  In fact, it has been
estimated that 50 percent and 40 percent of the total P loads to Lakes Erie
and Ontario, respectively, are from diffuse sources (Chapra and Sonzogni
1979).  Most of this P is contained in suspended particulate matter and has an
unknown availablility. Yet, in any body of water, P availability is a function
of several independent but partially intercorrelated variables which include
current trophic status; abundance and diversity of biota; climate; thermal
regime; morphometry; hydrology; mixing dynamics; orthophosphate concentration;
concentration, size, shape, density, and distribution of forms of P associated
with suspended solids  (Logan et al. 1979).

     In the face of such complexity, effective management of eutrophication
through BAP load control requires some simplification of the problem.  One
approach to simplification has been to treat BAP as a state variable for
mathematical simulation of water quality.  This approach considers BAP as a
quantity that may be measured and modeled much as any other state variable,
such as dissolved oxygen or total P, and has given generally superior
predictions of water quality for Lake Erie compared to a model that did not
distinguish among forms of BAP (De Pinto e_t al. 1984).  To calibrate and
verify any such model, however, requires methods for analysis of BAP.  Both
steps would be necessary before a model involving BAP could be exploited for
its potential management benefits.
STATEMENT OF THE PROBLEM

     Making a conceptual distinction between total P and BAP is much easier
than making an analytical one.  This happens because measurement of truly
bioavailable P is tied to quantification of a biological response.  Thus,
measurement of BAP requires the use of bioassay methods rather than chemical
analysis.  Unfortunately, the bioassays required for measurement of BAP
generally require more time to perform and are less precise than the routine
analytical procedures required for measurement of the constituents commonly
used to characterize water quality.  Simply stated, present knowledge
concerning the forms and mobility of P in aquatic systems does not permit
direct measurement of specific, discrete forms of P in water or sediments that
can be equated with BAP.

     For example, it is well accepted that soluble BAP depends principally on
the orthophosphate fraction, which usually is analytically approximated by
measurement of soluble (molybdate-)reactive phosphorus (SRP).  However, data
published by De Pinto ert al. (1980) and Young e_t al. (1982) concerning
bioavailability of wastewater P demonstrate that soluble BAP substantially
exceeds that which may be measured as SRP.  Moreover, the problem is
compounded in the case of particulate BAP (BAPP) owing to the variety of
physical and chemical phases that may be present.  Awareness of the need for
and the difficulties associated with bioassays, particularly with respect to
estimation of BAPP,  has led to a growing body of research directed toward
development of empirical methods for quantification.  For BAPP measurement,
several investigators recently have used chemical extraction techniques.

     In general BAPP extraction methods have been adapted from procedures
developed by agronomists for assessment of soil fertility as related to crop
production (Chang and Jackson 1957).  Methods for BAPP analysis generally
derive from the work of Williams and co-workers (Williams et al.  1971), who
modified soil procedures for application to aquatic sediments.  A common
feature of most sediment BAPP measurement schemes is an extraction with strong

-------
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
I
I
I
I
base, NaOH.  Unfortunately, the specific procedural modifications that have
been used in connection with aquatic sediments have tended to be unique to
each investigation.  This has resulted in the development of a variety of P
extraction methods and, consequently, varying estimates of BAPP.  Furthermore,
comparisons between the various procedures has received little attention
(Williams e_t al. 1980).  In addition, few investigations have included
bioassay determinations as a reference for comparisons between extraction
procedures.  When bioassays have been performed,  other aspects of the methods
have varied as much as the chemical extraction methods.  This methodological
variability has prevented any rational synthesis of the results of various
investigations on a specific system.  Thus, a need exists to examine the
comparability of the various methods currently used to quantify BAPP.

     Once methodological variances among procedures for estimation of BAPP are
resolved and a procedure is selected, it becomes possible to implement
management decisions that focus on controlling BAP.  With respect to
implementation, however, a matter of significant practical consequence
concerns specification of appropriate procedures for handling samples in the
quantities required for monitoring available P levels in the lower Great
Lakes.  Given the large number of samples that would be required, varying
periods of time may elapse between sample collection and analysis.  Such
holding periods, and the conditions of holding, may affect the concentrations
of all forms of P in the water, including available forms, through various
chemical and microbial transformations.  Examples of these transformations
include microbial mineralization of organic P-containing compounds,
immobilization of phosphate by microbial uptake,  solubilization of inorganic
phosphoric acid salts, formation of insoluble precipitates, and solid-solution
partitioning phenomena involving the interfaces between sample solution and
suspended solids or the container.

     Despite an awareness of the problem, relatively little is known about the
effects of storage on the BAP content of water samples.  Examination of the
effects of storage on orthophosphate (SRP) and BAPP can provide guidance on
sample handling requirements for collection of accurate data in a field
program that involves estimation of BAP forms on a large scale.
OBJECTIVES AND SCOPE

     Stated in general terms, the objectives of this investigation focused on
comparing a variety of procedures for estimation of BAPP in the lower Great
Lakes region and on determining the rate and extent to which storage affects
orthophosphate (SRP) and BAPP concentrations in natural water samples.
Specifically, the objectives included the following elements:

    1.  Compare five chemical extraction procedures for estimation of
        BAPP.

    2.  Compare chemically defined estimates of BAPP to those obtained
        by bioassay measurements.

    3.  Determine the rate and extent to which concentrations of SRP are
        affected by time of storage,  temperature during storage, sample
        source,  and source related variables (initial pH,  suspended solids,
        total P,  conductivity).

     For the first objective, a series of 12 sediment samples,  collected from
diverse aquatic  systems around the lower Great Lakes, were analyzed using five

-------
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
I
I
I
I
chemical extraction procedures for estimation of BAPP.  The procedures that
were compared are used currently by individuals engaged in research in the
Great Lakes region.  The chemical procedures for estimation of BAPP included
those used by Dr. D.E. Armstrong et_ al. (1979) at the University of Wisconsin;
Dr. D.B. Baker at the Water Quality Laboratory, Heidelberg College; Canada
Center for Inland Waters (Mayer and Williams 1980); and two variants of the
procedure of De Pinto et_ al.  (1981) in use at Clarkson University.

     Bioavailability bioassays using algae and sediments as the sole P supply
for growth were performed on  the same sediment samples as those analyzed to
meets the requirements of the first objective.  This was done to provide
reference estimates of BAPP levels for inclusion in the analysis of
comparability among the extraction procedures.  The bioassays were performed
using techniques that were developed and tested during previous investigations
(De Pinto et. al. 1981, De Pinto 1982, Young et. al. 1982, Young and De Pinto
1982).

     Meeting the third objective required a combined field and laboratory
examination of the rate and extent to which storage causes changes in the
concentrations of BAPP and immediately-available soluble P (SRP) in water
samples that contain particulate matter.  Experimental variables that were
controlled included the sample source and time and temperature of holding;
variables related to sample source were not controlled but were treated as
covariates in the analysis of the fixed effects.  Two different rivers were
selected to provide a broad range of concentrations of P and suspended solids
for incubation at three temperatures: 4, 22, and 45 C; and holding times up to
nine days.  The conditions at the upper limits (nine days, 45 C) were selected
to simulate sample holding under extreme summer conditions as might be
encountered during part of an extensive monitoring effort.

     By evaluation of the relationship between several common methods for
chemical measurement of BAPP and how they compare to bioassay measurements,
the results of this investigation provide a basis for understanding the limit
on accuracy that applies to efforts to integrate past and future research
concerning determination of BAPP.  Further, the results that pertain to the
effects of storage will yield needed information on sample handling
requirements for collection of accurate data on BAP.

     Analysis of the results from the proposed study have not been directed
toward support of a "best" procedure for chemically measuring BAPP.  The term
"best" when applied to BAPP methods can assume several connotations:  accurate
or unbiased,  most consistent or precise, or, perhaps,  easiest to implement,
logistically.   At best, "best" for a monitoring system may not be the "best"
for surveillance, or the best for field- or laboratory-based research.
Rather,  the intent of the research has been to provide a systematic
examination of the similarities and differences that exist between the results
of the various procedures,  including bioassay methods, for determination of
BAPP.   The results, therefore, constitute a guide for procedure selection that
permits an informed choice insofar as that may be done by considering the
potential analytical difficulties associated with each method, the degree of
correlation to be expected between the selected procedure and other methods,
and correlation of each method with truly available P as measured by
bioassay.

     This investigation has focused on characterization of BAPP in samples
from fluvial systems in the lower Great Lakes region and the results so
obtained are intended primarily to be used in that area.  Nonetheless, the
chemical extraction methods proposed for investigation are based on

-------

fundamental aspects of the chemistry of P in soils and sediments.
Furthermore, the variety of sources of particulate matter tested has provided
generality for the conclusions of this study.  Consequently, application of
the results should not be overly constrained by unique geographical or
geological features of the lower Great Lakes region.

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
                                   SECTION 2

                                  CONCLUSIONS
COMPARISONS AMONG EXTRACTION PROCEDURES

     The major hypothesis involving the chemical extraction procedures for
estimation of BAPP stated that the procedures did not differ with respect to
amounts of P each would extract from samples of sediment.  The results of this
investigation, tested by analysis of variance, demonstrated significant and
generally consistent differences in amounts of P extracted from 12 sediment
samples by four of five procedures for estimation of BAPP in aquatic samples.
Absolute differences in P extracted among the procedures were strongly
dependent on individual sample characteristics.  The wide variation in total P
levels among the sediments was the main determinant of sample influence.

     Extraction results for the two variants of the De Pinto procedure
differed by an average of less than 7 ug P/g for the 12 sediments, the total P
of which averaged 1277 ug P/g.  This difference was not significant (p>0.05).
Thus, use of filtration rather than high-speed centrifugation for solid-liquid
separation prior to color development did not appear affect the results of the
analysis to an appreciable extent.

     Ranked according to the average fraction of total sediment P extracted by
each, the procedures and their approximate proportions, in parentheses, would
be ordered:

   De Pinto/Filtr. ~ De Pinto/Centrif.  < Baker < Armstrong < CCIW < Total P
         (1.0)             (1.0)         (2.0)     (2.8)     (3.6)   (6.4)
     Analysis of the extraction data from all samples indicated that the the
amount of sediment P extracted by any given procedure was not, in all cases, a
simple proportion of that extracted by any of the other procedures.  Rather,
the results obtained from a given procedure depended, to some degree, on
factors specific to individual samples.   In the most severe case, results that
were reproducible but quite anomalous were obtained when the De Pinto-based
procedures were applied to sediments that had been held in storage for several
years.  The other extraction procedures, however, gave characteristic
results.   In general, therefore,  it may  be concluded that a simple regression
equation may serve to convert estimates  of BAPP by one procedure into
equivalent estimates by another procedure as long as the original estimates
are for freshly collected samples.

     A high degree of intercorrelation existed among the extraction results
obtained by the methods of Baker,  Armstrong, and CCIW for all the samples in
this study.  The results obtained by these procedures also correlated
significantly with the total P levels of the sediment samples.  This suggests
that the three procedures may extract P  from approximately the same
physicochemically bound fraction or fractions of total P, though with
different efficiencies.  Since these procedures were strongly intercorrelated

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
for all samples, regression equations could be used to convert BAPP estimates
between each of the three procedures with acceptable accuracy for both fresh
and stored samples.

     None of the procedures investigated during this project was obviously
superior to the others as an estimator of algal bioassay-determined BAPP.
Amounts of P extracted from freshly collected sediments by the De Pinto
procedures were consistently closest in magnitude to the amounts taken up by
algae during the bioassays and could, therefore, be considered the most
accurate predictors of BAPP among the methods tested.  On the other hand,
while the Baker, Armstrong, and CCIW procedures gave extractable P levels that
generally overestimated the bioassay results, they correlated most closely
with the bioassay results for both fresh and stored samples.  As a
consequence, these three procedures were the most precise for prediction of
BAPP using regression equations.  Among the various procedures,  that of Baker
was the simplest to perform and easily could be adapted for use in most
moderately-equipped analytical laboratories; however, the extraction methods
of De Pinto were the most reproducible.
THE EFFECTS OF SAMPLE STORAGE ON AVAILABLE PHOSPHORUS

     Water samples from two rivers held under controlled conditions of
temperature for varying periods showed significant changes in sample
concentrations of SRP, BAPP by the method of Baker (1983), and BAPP per unit
of suspended solids.  The changes depended simultaneously on all the
experimental factors: sample source, temperature of holding, and time of
holding.  This means that SRP concentrations in samples from the Huron River
changed over time and as a function of temperature in a manner that did not
parallel that observed in samples from the Maumee River.  Changes in SRP
concentrations for temperatures of 4 and 22 C were small for samples from both
rivers for all holding periods, and, with the exception of the 9 day holding
period at 4 C for the Huron River samples, none of the differences were
significant for samples held at these temperatures.  Samples held at 45 C,
however, demonstrated significant changes within the first 0.5 day period of
incubation in samples from both rivers.

     The direction of the change in SRP concentrations was similar for samples
from both rivers when incubated at 45 C; in both cases the concentration of
SRP decreased to very low levels.  At 4 and 22 C,  however, the direction of
change in SRP was dissimilar between the rivers; in the Huron River samples
SRP increased,  while in the Raisin River samples,  SRP did not change enough to
establish a reliable trend.

     The data on effects of storage on BAPP and BAPP per unit weight of
suspended solids were analyzed to test hypotheses similar to those for SRP.
The results indicated significant changes in BAPP concentrations occurred
during .storage and that the changes were dependent simultaneously on all of
the main effect variables: sample source, temperature of holding, and time of
holding.  Normalization of the BAPP for suspended solids concentration did not
alter these conclusions.

     Incremental changes in concentrations of SRP and BAPP during sample
storage were tested to determine whether changes in one "available" fraction
would be predictable from changes in another.  The results showed a
significant negative correlation between the two quantities, which indicates
that losses of SRP during storage were balanced to an extent by increases in
BAPP.  Based on a regression analysis of the incremental changes,

-------

approximately 90 percent of SRP concentration losses and gains were reflected
in reciprocal gains and losses of BAPP in the water samples.  Since, as noted
above, different procedures for extraction of BAPP from sediments gave
different results,  it is possible that a value other than 90 percent would
have resulted had these samples been analyzed by one of the other procedures.
Nonetheless, it appears that changes in SRP during storage may be accompanied
by simultaneous and opposite changes in BAPP.

     Changes in SRP during storage were evaluated by an analysis of covariance
to determine whether initial pH, conductivity, and concentrations of total
suspended solids, total P, and SRP in field filtered subsamples were related
to changes in SRP during each storage period.  Among these variables, only the
concentration of SRP in samples immediately after collection was a consistent
factor relating to changes in SRP during storage.  Samples with high initial
concentrations of SRP tended to be those that lost more SRP during storage;
this was characteristic of the samples from the Maumee River, on the average.
Effects attributable to initial SRP, however, were not significant for more
than two days after storage was begun.

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
                                   SECTION 3

                                RECOMMENDATIONS
COMPARISONS AMONG EXTRACTION PROCEDURES

     This investigation has shown that it is feasible to use regression
relationships to transform BAPP estimates among several methods when freshly
collected sediments have been analyzed.  This practice, however, is not
recommended for sediments that have been held in storage for long periods.

     The anomalous results observed for stored samples underscore a need for
additional research to improve knowledge of factors that give rise to varying
estimates of BAPP by different chemical and biological analytical methods.   It
is recommended that additional research be undertaken that focuses on
fundamental properties of BAPP and development of appropriate methods for its
quantitation.

     Inasmuch as the concept of BAPP has assumed a significant role in
development of Great Lakes water quality management strategies, it is
recommended that a single procedure for BAPP measurement be adopted for common
use within the Great Lakes community.   To do otherwise would confine BAPP to
research functions only and would preclude its application and acceptance for
monitoring and surveillance activities, even though it has been largely out of
concern for the latter that the concept has developed.  Since estimates of
BAPP depend on the procedures selected, any successful program that contains
surveillance or monitoring activities must employ a single, standardized
method to allow the data to have application beyond the components of the
specific investigation.

     Since the concept of fractioning total P into available and unavailable
portions has been demonstrated to be realistic in the laboratory, the next
step should be a field scale test of the concept.  It is recommended that a
monitoring and research program be undertaken to demonstrate the feasibility,
potential value, and logistical problems of using BAP as a management tool  for
improvement of water quality.  The program should focus on one of the Great
Lakes,  or a significant segment thereof, and include collection of data on  BAP
inputs, outputs, and changes in storage within the system.  It is emphasized
that a field-oriented program like this would mandate use of a single
procedure for analysis of BAP to insure transfer of compatible data for mutual
use among program participants.

THE EFFECTS OF SAMPLE STORAGE ON AVAILABLE PHOSPHORUS

     The data presented on the effect of sample storage on concentrations of
BAP suggest that for temperatures in the range of 4 to 45 C, no holding time
is truly satisfactory for all samples if they are to be analyzed for SRP and
BAPP.  Thus, it is recommended that water samples collected for analysis of

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
BAP forms be refrigerated and the analysis be performed as soon as possible
(24-48 hours).  These recommended storage conditions do not differ
significantly from those given in Standard Methods (APHA, 1981) and by the
USEPA (1976).


     If samples for BAP determination cannot be analyzed immediately or
refrigerated, some method of preservation should be selected.  Otherwise,
concentrations of the quantities of interest are subject to redistribution
between solution and solid phases in the samples.  It is interesting to note,
however, that the redistribution of fractions appears to be somewhat
conservative on the average, since loss from the SRP fraction may be picked up
partially during analysis as a gain in the BAPP fraction.  Thus, an
approximation to the total quantity of BAP (particulate and soluble) in a
water sample may be estimable from a stored sample if both SRP and BAPP are
determined and summed, and if the storage period is not longer than a few
days.  Additional research on the nature and extent of BAP redistribution
during sample storage could be helpful in establishing sound alternatives for
preservation or immediate analysis.
                                     10

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
                                   SECTION 4

                             METHODS AND MATERIALS
COMPARISONS AMONG EXTRACTION PROCEDURES

Study Design

     Replicate analyses on a series of sediment samples with a range of P
biological availability provided the basis for comparisons among BAPP
extraction procedures.  The procedures compared during this investigation are
described in a detailed, stepwise manner in Appendix A and include those of
Armstrong et al. (1979), Baker (1983), Canada Center for Inland Waters - CCIW
(Mayer and Williams 1982), and De Pinto e_t al. (1981); the procedure of De
Pinto et al. (1981) was studied as two procedures differing only by use of
centrifugation or filtration of the extractant solution from the sediments
immediately prior to neutralization for color development.  The procedures all
differ in some respects but they have in common a period of contact between
sediments and NaOH.  Table 1 is a summary of the procedures and unique points
about each.

     In addition to these procedures, algal bioassays were performed to
provide estimates of BAPP as a control against which each of the chemical
procedures could be compared and for use as a covariate in comparing the
procedures among themselves.  The bioassay procedure involved use of the DCDA
(Dual Culture Diffusion Apparatus) of De Pinto (1982), and a detailed
description of the DCDA bioassay procedure is given in Appendix A.

Sampling

     For BAPP comparisons, river water and bottom sediment samples were
collected from seven major tributaries to the Lower Great Lakes.  These
included the Cuyahoga, Maumee, and Sandusky Rivers of Ohio; Cattaraugus Creek,
and the Genesee and Oswego Rivers of New York; and the Raisin River in
southeastern lower Michigan.  The location of the sampling sites is shown in
Figure 1.

     Sampling of the Ohio and Michigan rivers began mid-March 1984 and
continued through mid-April of that year.  Samples were collected by personnel
from the Water Quality Laboratory at Heidelberg College, (Tiffin, Ohio).
Collections were made as surface grab samples near the center of the main
channel during major storm runoff events.  Bottom sediment samples were taken
during early May 1984 from each of the three New York rivers.  Also included
in the sample set were two archived bottom sediment samples collected during
the fall 1984 from the southwestern shore of Lake Erie, near Monroe,
Michigan.  To complete the set of 12 samples, a standard river sediment sample
from the National Bureau of Standards (SRM 1645 River Sediment) was included.
                                     11

-------

TABLE 1.  METHODS FOR DETERMINATION OF BIOAVAILABLE PARTICIPATE
          PHOSPHORUS IN USE IN THE GREAT LAKES REGION.
Procedure   Extractant
               Extractant:Solid
                    Ratio
               Comment**
Armstrong
Baker
CCIW
0.1 N NaOH+
1.0 N NaCl

0.1 N NaOH+
1.0 N NaCl

CDB(reduc-
tant) then
0.1 N NaOH
De Pinto*   0.1 N NaOH
2000:1
1250:1
 500:1
                     600:1
Approximates NAIF


Solids on filters
during extraction

Sequential;
Sum = NAIP
               Correlations with
               BAPP as determined
               by algal bioassay
*   The method of De Pinto includes both a filtration and a
    centrifugation separation of solids and extractant.

**  NAIP = Non-apatite Inorganic Phosphorus
    BAP = Bioavailable Phosphorus
                                  12

-------

                                                                     d
                                                                     o
                                                                    •H
                                                                     60
                                                                     0)
                                                                     CO
                                                                     
-------
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
I
I
Sample Handling and Preparation

     For each of the Ohio and Michigan rivers, approximately 10 liters of
river water were collected and placed in acid-cleaned polyethylene
cubitainers.  Upon return to the Water Quality Laboratory at Heidelberg
College, the cubitainers were stored at 4 C.  Within a few days of collection,
the samples were placed in insulated shipping containers (Trans-Temp) with
frozen ice-packs and shipped to Clarkson University by Greyhound Bus.

     Upon receipt at Clarkson, the cubitainers were mixed thoroughly and the
contents filtered under positive pressure through 0.4 micron pore-diameter
polycarbonate membranes (Nuclepore).  The non-filterable residue retained on
the filter was removed using a flexible spatula and resuspended in sufficient
P-free synthetic medium (Martin 1983) to yield a solids concentration of
approximately 10 mg/ml.  The resuspended "concentrates" of non-filterable
residue were refrigerated at 4 C until analysis.   Martin (1983) observed that
such residue concentrates stored at 4 C lost an average of less than two
percent of total sediment P to the resuspension medium over a period of
several months.

     Sediments from the Genesee and Oswego Rivers, Lake Erie, and Cattaraugus
Creek were collected as bottom grab samples.  The entire contents of the
sediment grabs were mixed to homogeneity and an aliquot of each was
transferred into an acid-cleaned polyethylene jar.  The jars then were placed
on ice in a cooler and transported by automobile to Clarkson University.  Upon
arrival at the Clarkson laboratory, the bottom sediment samples were
resuspended in the same P-free medium as the non-filterable residue samples
described previously to solids concentrations of approximately 10 mg/ml.  To
improve analytical precision in a few of the more heterogeneous samples
(Cattaraugus and Genesee), pebbles, sticks,  coarse and fine sand fractions
were removed from the sediment concentrates by sedimentation prior to
placement into storage at 4 C.

Analytical Program

     Prior to performing the BAPP extractions it was necessary to determine
the suspended solids concentration of each of the sediment concentrates.  This
was done using standard methods (APHA 1981), and involved vacuum-assisted
filtration of a known volume of sample through a dried and tared,  0.45 micron
pore-diameter nitrocellulose membrane for collection of the non-filterable
residue.  After drying to constant weight at 103  C,  the residue was determined
by weight change of the filter plus residue on a Mettler Model A30 analytical
balance.

     Measurement of P in water was performed colorimetrically,  using the
method of Murphy and Riley (1962)  as described in Standard Methods (APHA
1981).  Sample pretreatments included filtration through 0.45 micron
pore-diameter nitrocellulose membranes to separate soluble fractions from the
total sample for determinations of SRP and sulfuric acid-peroxydisulfate
digestions for conversion of combined P to orthophosphate for determinations
of total P.   Both of these steps were performed,  when required,  prior to color
development.
                                     14

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Data Analysis

     The data on BAPP were analyzed by factorial analyses of variance and
covariance with the extraction procedures and the samples as main effect
variables in the analysis.  The factorial design was selected rather than
blocking on individual samples since it was felt that the procedures might not
yield proportionately similar amounts of extractable P among samples but that
sample-procedure interaction could be significant and should be tested.   Total
P and algal bioassay determined BAPP initially were used as covariates,  but
they were found to be redundant with each other and could not improve the
efficiency of tests for differences among procedures and subsequently were
dropped from the analysis.  Additionally, all-pairwise correlation analyses
and regression analyses were performed to clarify the relationships between
the procedures.
EFFECTS OF SAMPLE STORAGE ON AVAILABLE PHOSPHORUS

Study Design

     Investigation of the effects of short term storage on the concentration
of orthophosphate and BAPP in river water was conducted over a period of three
weeks beginning mid-March 1984.  The work was performed at a temporary
laboratory site established at the U.S. Environmental Protection Agency Large
Lakes Research Station at Grosse lie, Michigan (ERL-Duluth).

     Over a period of nine days, water samples were collected daily from the
study rivers and transported to the temporary laboratory.   Subsamples were
filtered immediately upon collection in the field for determination of initial
SRP concentrations.   Upon arrival at the laboratory, each sample was split
into 15 subsamples for incubation at temperatures of 4, 22, and 45 C for
periods of 0.5, 1.0, 2.0, 4.0, and 9.0 days.  At initiation and at termination
of each incubation period, subsamples were sacrificed for determination of SRP
and BAPP by the method of Baker (1983).

River Selection

     The rivers selected for the storage effects study were the Huron and
Maumee Rivers.  They were chosen after review of data from the Corps of
Engineers on suspended sediment and orthophosphate concentrations in
tributaries to the lower Great Lakes.  The review was undertaken with the
intent of selecting two rivers with widely differing concentrations of both
materials to maximize the possibility that the investigation would include
samples of both low (<0.05 mg P/L) and high (>0.10 mg P/L) concentrations of P
and suspended solids.   Other considerations for selection were: accessibility
from either Grosse lie,  MI or Tiffin, OH (either was a potential site for
location of a temporary laboratory) and contribution of significant flow to
the lake.

     Because of its very large flow and high concentrations of SRP and
suspended solids,  the Maumee River was an obvious choice.   The other stream,
therefore, had to be one that generally contained low levels of these
materials. The Huron River (MI) and Honey Creek (OH) appeared to be the best
alternatives,  based on the above criteria.   Due to lower levels of soluble
reactive P in the Huron River, it ultimately was selected as the second
                                     15

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
river.  The data that led to this selection is summarized in the following:

     For the period 3 March 1977 through 27 June 1977, the mean daily maximum
concentration of SRP and suspended solids  (standard deviations in parentheses)
for the three rivers named above were:

Mauniee River:  0.112 mg P/L (0.029 mg P/L) and 213.4 mg SS/L (316.7 mg SS/L)
Honey Creek:   0.063 mg P/L (0.029 mg P/L) and  30.8 mg SS/L ( 76.3 mg SS/L)
Huron River:   0.025 mg P/L (0.019 mg P/L) and  24.7 mg SS/L ( 10.7 mg SS/L)

Sampling

     On three days during the period 21-29 March 1984, ten samples of
approximately ten liters each were collected from both the Maumee and Huron
Rivers.  Upon collection the water was placed in acid-cleaned polyethylene
cubitainers.  Each cubitainer was immediately subsampled and the aliquot was
filtered through a 0.4 micron pore-diameter polycarbonate membrane filter.  A
hand pump was used to provide vacuum during filtration.  The filtrate was
transferred to a clean 125 ml polyethylene bottle for later analysis of SRP,
and the bottle was placed on ice in a cooler.  The remainder of the sample
also was placed in the dark and cooled for transport to the temporary
laboratory.

     Upon arrival at the laboratory, the contents of each cubitainer were
mixed thoroughly and 100 ml aliquot subsamples were transferred to a series of
15 new, acid-cleaned 125 ml polyethylene bottles for incubation at
temperatures of 4, 22, and 45 C and a second series of four similar bottles
for determinations of initial concentrations of total and soluble reactive P
and suspended solids.  Incubation of the subsamples was done in darkened,
thermostatted enclosures and temperature was monitored by frequent
observations with a mercury thermometer.

     Determinations also were made of initial pH and conductivity.  The pH was
determined using an Orion Model 501 ion analyzer with a glass-AgCl combination
pH electrode.  Conductivity was measured using a YSI Model 31 conductivity
bridge.

     During the storage effects study, all orthophosphate concentrations were
determined colorimetrically using a Baush and Lomb Spectronic 710
spectrophotometer.  Color development was performed using the method of Murphy
and Riley (1962) as described in Standard Methods (APHA 1981) and was done
after filtration of the samples through 0.4 micron pore-diameter polycarbonate
membranes (Nuclepore).  Preliminary tests on standard orthophosphate solutions
and river water showed the laboratory filtration apparatus and the field
filtration units gave nearly identical results (Appendix F).  Initially a set
of standard solutions ranging from 0 to 200 ug P/L were analyzed with each
group of samples to provide a standard curve against which to read sample
concentrations.

     All orthophosphate concentrations throughout the study were calculated
from this initial standard curve.  Subsequent analyses were performed using
the Youden "A-B" technique for calibration control.   Included for
implementation of the technique were two deionized water blanks,  a 100 ug P/L
orthophosphate standard,  an "A" solution with an orthophosphate concentration
of approximately 20 ug P/L,  and a "B" solution with an orthophosphate
                                     16

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
concentration of approximately 160 ug P/L.


     At elapsed time intervals of 0.5, 1.0, 2.0, 4.0, and 9.0 days from the
beginning of incubation, one subsample (in 125 mL bottle) from each was
removed from each incubator and filtered under vacuum through a 0.4 micron
pore-diameter polycarbonate filter.  The filtrate was analyzed immediately for
SRP.  The non-filterable residue was retained on its filter and placed in a
clean 125 ml polyethylene bottle for storage by refrigeration until later
chemical extraction of BAPP.


     The BAPP extractions and analyses of total P were performed upon return
to Clarkson University from the temporary laboratory at Grosse lie, Michigan.
The extraction method used was similar to the NaOH/NaCl procedure of Baker
(Table 1, p.15; and Appendix A) modified by substituting a 0.4 micron
pore-diameter polycarbonate membrane for the 0.6 micron pore-diameter
polyvinyl chloride membrane recommended by Baker (1983).  The latter no longer
are manufactured and preliminary testing showed the polycarbonate membranes
would withstand extraction by 0.1 N NaOH for the required period of shaking.


Data Analysis


     Effects of storage conditions on concentrations of SRP and BAPP were
tested by analyses of variance and covariance.  The experimental arrangement
permitted analysis as a double split-plot design since the samples were split
with respect to treatments  (temperature of incubation) and split within
treatments for repeat measurement (time of incubation).  Other designs, such
as latin squares or randomized complete blocks also could have been used
effectively for analysis of the experiment; however, the double split-plot
model required fewer assumptions about the nature of the processes affecting
the samples and was conservative with respect to the partitioning of
variance.  Covariates used to improve the efficiency of the analysis included
total P, total suspended solids, pH, conductivity,  and field filtered SRP.
                                     17

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
                                   SECTION 4

                            RESULTS AND DISCUSSION


EFFECTS OF SAMPLE STORAGE ON AVAILABLE PHOSPHORUS

     The results of total and SRP analyses, as well as suspended solids, pH,
and conductivity measurements from both Huron and Maumee river samples are
presented in Table 2.  These data indicate distinct differences between
samples from the two rivers.  For example, the 10 Huron River samples were
characterized by relatively low levels of total P (67.6-82.7 ug P/L), SRP
(A.2-9.5 ug P/L), and suspended solids (10-26 mg/L).  In contrast,  the levels
of these parameters were considerably higher for samples from the Maumee
River.  With respect to pH and conductivity, however, the Maumee values were
lower than those from the Huron River.

     The response of both Huron and Maumee river samples during storage for as
long as nine days at three different temperatures (4, 22, and 45 C) is shown
graphically in Figures 2 and 3.  Each data point represented on these figures
is the average response of 10 samples collected during the study.  A listing
of SRP measurements for all treatment combinations is given in Appendix B.

     Examination of the Maumee River plot shows essentially no change in SRP
concentration for samples stored at 4 and 22 C.  In contrast, Huron River
samples stored at these same temperatures decreased in concentration for the
first two days of storage, then increased somewhat throughout the remainder of
the study.  Johnson e_t al. (1975) found SRP levels increased rapidly but
subsequently decreased in stream samples stored at 5 C for 30 days.  Others
have also observed increases in the soluble inorganic P content of river water
stored at 4 and 23 C for 12 weeks (Klingaman and Nelson 1976).  Heron (1962)
found that lake water samples stored at room temperature exhibited an increase
in P concentration during the first 10 hours of storage.  This increase was
followed by decreasing concentrations for the remainder of the study, a
sequence of concentration changes similar to those observed during the present
study on samples from the Huron River.  Additionally, Murphy and Riley (1956)
reported that the SRP of two seawater samples stored for one month at 20 C had
increased from 13.5 to 15.7 and 25.7 to 27.2 ug P/L, respectively.

     After a substantial initial increase in SRP concentration, within 12
hours of storage at 45 C, samples from the Huron River rapidly decreased to
extremely low levels.  Similarly, the Maumee River samples lost a significant
amount of SRP during the first 4 days of storage, but unlike the Huron samples
increased slightly between 4 and 9 days.   It is apparent from these data that
the kinetics of SRP transformations did not follow simple rate laws over the
incubation periods.  The treatment combination giving rise to the most
consistent response was that for the Maumee samples incubated at 45 C.  For
this group of samples the half-life of SRP was approximately 1 d over the
first 4 days of incubation.  This is equivalent to a first order loss rate of
                                      18

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
TABLE 2.  CHEMICAL CHARACTERISTICS OF WATER SAMPLES COLLECTED FOR STORAGE
         EFFECTS STUDY.
Source and
Sample Date of
Number Collection
Huron #1
Huron #2
Huron #3
Huron #4
Huron #5
Huron #6
Huron #7
Huron #8
Huron #9
Huron #10
Maumee # 1
Maumee #2
Maumee #3
Maumee #4
Maumee #5
Maumee #6
Maumee #7
Maumee #8
Maumee #9
Maumee #10
21Mar84
21Mar84
21Mar84
26Mar84
26Mar84
26Mar84
26Mar84
29Mar84
29Mar84
29Mar84
21Mar84
21Mar84
21Mar84
26Mar84
26Mar84
26Mar84
26Mar84
29Mar84
29Mar84
29Mar84
Total
Phosphorus
(ug P/L)
81.1
79.5
67.6
77.2
79.6
70.8
80.4
69.2
78.8
82.7
467.7
392.7
409.2
308.6
319.2
317.6
331.2
412.2
425.7
425.7
Soluble
Reactive
Phosphorus
(ug P/L)
4.9
4.2
4.6
9.5
9.4
9.4
8.7
9.5
8.8
8.9
92.2
95.5
93.5
94.4
73.3
73.1
73.5
71.6
70.5
76.6
Suspended
Solids pH
(mg/L)
24
21
26
11
17
18
16
10
11
13
189
199
170
127
133
130
139
198
204
215
8.15
8.17
8.16
8.09
8.09
8.11
8.10
8.07
8.07
8.07
7.77
7.77
7.77
7.81
7.81
7.82
7.81
7.86
7.86
7.86
Conductivity
(umho/cm)
540
500
520
610
580
600
600
570
560
560
330
320
330
355
340
345
350
345
350
340
                                     19

-------

                                               *- o
1/d 6ri 'uojiejjueouoo
                                                                    C
                                                                    03
                                                                    0)
                                                                    S-i
                                                                    OJ
                                                                    o
                                                                   •rl
                                                                    4-1
                                                                    O
                                                                    C

                                                                   4-1
                                                                    03
                                                                    cd
                                                                    en
                                                                    o>
                                                                    6  3
                                                                    C3  4-1
                                                                    en  en

                                                                    S-i  en
                                                                    01  4-1
                                                                    >  o
                                                                   •H  a)
                                                                   Pi  "4-1
                                                                      u-i
                                                                    C  OJ
  01
  oo
  n)
                                                                    3
                                                                   P3
•H 4-1
CO
                                                                      o
                                                                      4-1
                                                                      cn

                                                                      oc
                                                                      C
                                                                      H
                                                                   4-i  M
                                                                   O  3
                                                                    O  01
                                                                    •H  6
                                                                    4-1 -H
                                                                   4-1  00
                                                                   s  c
                                                                   01 -H
                                                                   CJ 13
                                                                   C rH
                                                                   O  O
                                                                   O ^
                                                                   CN


                                                                    OJ
                                                                    }-<


                                                                    bO
                                                                   •H
                           20

-------
  I
  I
  I
  I
  I
  I
  I
  I
  I
  I
  I
 I
 I
 I
 I
 I
I
I
I
6r1
                                                                                                           "d
                                                                                                           C
                                                                                                           nj
                                                                                                           S-i
                                                                                                           3
                                                                                                           -,
                                                                                                           6 TJ
                                                                                                           cd  3
                                                                                                           en  4-1
                                                                                                               en

                                                                                                           d)  en

                                                                                                          •H  a
                                                                                                          Pi  a)
                                                                                                              4-1
                                                                                                           0) <4-l
                                                                                                           a)  a)

                                                                                                           3  0)
                                                                                                           cd  oo
                                                                                                          S  cfl
                                                                                                          Pi   00
                                                                                                          C/}   C
                                                                                                              •H
                                                                                                          M-l   S-
                                                                                                           O   3
                                                                                                              T3
                                                                                                           C
                                                                                                           O   0)
                                                                                                          •H   6
                                                                                                          4-1  -H
                                                                                                          4J  00
                                                                                                          C  C
                                                                                                          a)  -H
                                                                                                          a  T3
                                                                                                          c  i-t
                                                                                                          o  o
                                                                                                          CU
                                                                                                          S-i

                                                                                                          oo
                                                                                                          •H
                                     21

-------

approximately 0.7/d.  The variability of the transformation, however,
precluded a systematic analysis of reaction rate or order for the stored
samples.

     Although Figures 2 and 3 illustrate the general trends of SRP
concentration during storage, statistical analysis of the data provides a more
objective interpretation of storage effects.  Analysis of variance (Appendix
E.I) indicated that significant changes in SRP concentration occurred during
storage, and that the changes depended simultaneously on three variables:
sample source (river), temperature during storage, and time of storage.  This
means that changes in SRP concentrations in Huron River samples, caused by
varying time and temperature of storage, did not parallel the changes observed
in samples from the Maumee River under similar conditions of holding.  For
instance, changes in SRP concentration for temperatures of 4 and 22 C were
small for samples from both rivers for all storage periods, and with the
exception of the 9 day storage period at 4 C for the Huron River samples, none
of the differences were significant (p>0.05) for samples held at these
temperatures.  However, samples from both rivers stored at 45 C demonstrated
significant changes within the first 12 hours of incubation.

     The direction of net change in SRP concentration was similar for samples
from both rivers when incubated at 45 C for periods longer than 12 hours.  In
both cases the concentration of SRP decreased to very low levels.  At 4 and 22
C, however, the direction of change was dissimilar between the rivers.  In the
Huron River samples, SRP decreased initially then increased, and the amount of
change overall was small though significant (p<0.05).  In the Maumee River
samples, on the other hand, SRP did not change enough to establish a reliable
trend at these temperatures.

     It has been widely accepted that increases in SRP concentration during
storage are likely due to bacterial or enzymatic decomposition of organic P
compounds (Murphy and Riley 1956; Gilmartin 1967; Thayer 1970; Johnson et
al. 1975; Klingaman and Nelson 1976).   In fact, Heron (1962) demonstrated that
a viable bacterial population is essential to bring about changes in the P
concentration of lake water samples low in orthophosphate (2.6 ug P/L).
Decreases in the levels of orthophosphate,  on the other hand,  are thought to
be the result of either utilization of P by a developing bacterial population
(Gilmartin 1967),  or sorption reactions (Johnson et al.  1975).  No data were
collected on microbial populations during the storage effects study reported
here, but no unusual features would be expected for the study systems.

     Concentrations of BAPP in the stored water samples are given in Appendix
C.  These data were analyzed statistically to test the effects of storage on
BAPP (Appendix E.2) and BAPP per unit weight of suspended solids (Appendix
E.3).  The results of the analysis of variance were similar to those derived
from the SRP data and indicate that significant changes in BAPP concentration
took place during storage.  As was true for changes in SRP, the changes in
BAPP depended simultaneously on three variables: sample source (river),
temperature during storage, and time of storage.  Normalization of the BAPP
data for suspended solids concentration did not alter these conclusions.

     Incremental changes in concentrations of SRP and BAPP during sample
storage were tested to determine whether changes in one bioavailable fraction,
SRP,  would be predictable from changes in another, BAPP.   The results showed a
                                     22

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
significant negative correlation between the two quantities  (r=-0.562 for
N=238), which indicates that a loss of SRP during storage was balanced to an
extent by an increase in BAPP (Figure 4).  Based on a regression analysis of
the incremental changes, approximately 90 percent of SRP concentration losses
and gains were reflected by reciprocal gains and losses in BAPP.  It is
possible, however, that a value other than 90 percent would have resulted if
the samples had been analyzed by a procedure other than Baker's  (1983).
Nevertheless, it appears that changes in SRP during storage may be accompanied
by concomitant and opposite changes in BAPP.  To an extent, therefore, the sum
of the concentrations of SRP and BAPP may be conservative in water samples
that are stored for brief periods, possibly up to nine days.  Thus, it appears
that an approximation of the total quantity of BAP (particulate and soluble)
in a water sample may be estimated from a stored sample if both SRP and BAPP
are determined and summed.

     Additional statistical analyses were performed to determine the extent
that initial pH, conductivity, and concentrations of total suspended solids,
total P, and SRP in field filtered samples were related to changes in SRP
during storage.  An analysis of covariance (Appendix E.4) indicated that the
only factor tested that related significantly (p<0.05) to changes in SRP
during storage was the concentration of SRP in field filtered samples.  As
might be expected intuitively, samples that lost more SRP during storage
tended to be those with high initial concentrations.   This was particularly
characteristic of samples from the Maumee River stored at 45 C.  However, the
analysis demonstrated that effects attributable to intitial SRP were not
significant for more than two days after storage was begun.

     From this investigation it can be concluded that for temperatures in the
range of 4-45 C, no storage time was wholly satisfactory for Huron River
samples analyzed for SRP and BAPP.  Similarly, storage of Maumee River samples
at 45 C permitted significant changes to occur in SRP and BAPP levels.
However, storage at 4 and 22 C appeared effective for inhibition of these
changes in the Maumee samples.  Nonetheless, in the absence of direct evidence
that samples from a given system do not require special storage precautions,
it is recommended that water samples collected for analysis of BAP be stored
under refrigeration and that the analysis be performed as soon as possible.
This procedure is equivalent to that recommended by USEPA (1976) and Standard
Methods (APHA 1981).  If the samples cannot be analyzed immediately or placed
in cold storage, then effective methods of preservation should be sought and
implemented.


COMPARISON OF BIOAVAILABLE PARTICULATE PHOSPHORUS PROCEDURES

Chemical Extraction Methods

     The results of the BAPP extractions on 12 sediment samples for all
procedures are summarized in Table 3.   The data presented are mean values
expressed as ug of P extracted per gram of sediment.   Results of the total P
analyses are also included in the table.  A complete listing of replicate
measurements for all samples is given in Appendix D.

     The results of this investigation,  tested by analysis of variance
(Appendix E.5), demonstrated significant, consistent differences in amounts of
P extracted from various sediment samples by four of the five chemical


                                     23

-------
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
I
I
I
                                   3
                                   i-J
                                   0)

                                   CO
                                   4J
                                   o
                                   0)

                                   M-l
                                   CD

                                   OJ
                                   bO

                                   S-i
                                   o
                                   60
                                   C
                                  S-i
                                  PM
                                  Pi
                                  PQ
                                  U)
                                  OJ
                                  oc
1/d 6ri
    HO^NH-V
                                  a)

                                  a)
                                  u
                                  o
                                  fl
                                  •H
                                  O
                                  •rl
                                  CO
                                  CO
                                  00
                                  a)
                                  cu
                                  S-i
                                  00
                                  •H
24

-------
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
I
TABLE 3.  SUMMARY OF BIOAVAILABLE PHOSPHORUS ANALYSES, ug P/g.
SOURCE
MAUMEE R.
CUYAHOGA R.
SANDUSKY R.
RAISIN R.
MAUMEE R.
SANDUSKY R.
L. ERIE
L. ERIE
OSWEGO R.
GENE SEE R.
CATTARAUGUS
CREEK
NBS STANDARD
SEDIMENT
SAMPLE
1
2
3
4
5
6
7
8
9
10
11
12
TOTAL
112.
1124.
987.
1261.
836.
900.
2209.
2472.
1007.
607.
423.
985.
P
3
3
9
5
8
0
9
9
2
5
9
6
CCTCENT
228.7
397.4
296.0
137.8
189.1
263.6
415.1
264.8
223.9
91.3
30.2
38.9
CCTFILT
225.
385.
289.
134.
183.
255.
396.
250.
219.
85.
29.
43.
0
1
2
9
1
9
5
1
2
6
7
2
BAKER
424.2
492.3
323.3
429.2
221.4
260.9
1197.0
1295.8
226.1
124.8
33.7
108.9
ARMSTRONG
368.
509.
344.
454.
222.
276.
1559.
1790.
231.
128.
33.
115.
7
2
3
6
8
3
4
2
8
9
8
9
CCIW
580.6
691.1
592.7
635.8
411.4
525.2
1929.7
2230.4
334.3
223.9
79.9
291.4
BIOASSAY
309.0
399.5
227.7
	
233.0
183.5
1526.7
2040.3
106.0
56.7
2.7
	
                                       25

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
extraction procedures.  Extraction results for the two variants of the
De Pinto procedure (CCTCENT, CCTFILT on Figure 5) differed by an average of
less than 7 ug P/g for the 12 sediments, which was not significant (p>0.05).
Thus, the use of filtration rather than high-speed centrifugation for
solid-liquid separation prior to color development gave no appreciable bias to
the results of the analysis.

     Among the samples, total P ranged from 418 to 2482 ug P/g, with the
highest values measured in the samples from Lake Erie and the lowest measured
in samples from Cattaraugus Creek.  Ranked according to the average fraction
of total sediment P extracted by each, the procedures would be ordered:

   De Pinto/Filtr. ~ De Pinto/Centrif. < Baker < Armstrong < CCIW < Total P
         (1.0)             (1.0)         (2.0)     (2.8)     (3.6)   (6.4)

This relationship is shown graphically in Figure 5.

     The amount of P extracted by the various procedures depended
interactively on sample-specific factors.  This means the P extracted from
sediments by one or more of the procedures was not a simple, consistent
proportion of that extracted by each of the other procedures.  From this
finding the conservative conclusion may be drawn that it is not always
possible to apply simple factors such as regression coefficients to convert
estimates of BAPP by one procedure into estimates by another procedure.  Doing
so could result in some biased BAPP estimates.  This conclusion is too
simplistic, however,  as the following more detailed consideration of the
results will show.

     This procedure-sample dependency may be illustrated by examining the
extent of correlation between the BAPP procedures.  A complete table of
correlation coefficients for all possible procedure pairings is presented in
Appendix E.6.

     The results obtained by the methods of Baker, Armstrong, and CCIW showed
a high degree of intercorrelation (r>0.992 for N=12).  Further, the procedures
correlated strongly (r>0.975 for N=12) with the total P levels of the sediment
samples (Figure 6).  This suggests the procedures of Baker, Armstrong,  and
CCIW may extract P from the same physicochemically bound fraction or fractions
of total P, but do so with varying efficiencies.

     The extraction results obtained by the two modifications of the De Pinto
method also showed significant positive correlation (p_<0.05) with each of the
other extraction procedures and total P, but the magnitude of the correlation
was less (0.566
-------An error occurred while trying to OCR this image.

-------
1
1
1

1


I OT
• w
^^^
r-M
1.
g :
i s!
O x
IPS •
P-4 "^
1 ft*
O |
1 —
m
^JI^J ^F*
o g
|{/2 —
HH g
| «J x * B-
V '- \ \

N. •
» * \
"« E \ v
— ' *° \ \
CO \ \ \
\ \
'•-. \ \
\ \
"•••-. x^\
'. ^N ^
\ \
i i x'r>
1 i x -. •*>
II X- MB
1 o
Z v '
i_ o >: x
• ^«^ ^^
o o «$ o: o 2
o o m •< o m
D + ^ < X >

i i i i i i i i i i i
• * *••* •»»*
IN r- i- *- w- OOOO
3/Huj 'HclVfl



28
















•
rc

*^^
^^^^
•;
1


C










en
QJ
0 3
. -d
C>4 i)
CJ
o
ex
i— i
. ^
- . 'T-t
(^ ^
03
n
M
i— i
- N *

\ I
GO -a
• •• -,
Ecu
00
r • 2
r- QJ
. CL?
0)
I— 1
, — — ex
_ ^ <13 1
T— ^— ' ^
o ^
I— 03
1 cu
C
_ CD °
d "S
3
M

0)
^t" CX4
•• *
o ^
o
w
— jj
C
3
O
O<^
<<
»
^O

0)
SJ
3
60
•H




-------


                                CL
                                E
                       oo
                       c
                       o
                 "«   ^  H
                 CQ   <  O

                 D   +   •»
             i
            (N
                  i
                 (N
                        i
                        c
1     I     i     I
1      1
                                                                            ro
                                                                            6
                                                                                   DO
                                                                                  \
                                                                                  QL.
                                                                                 UJ
                                                                                 O
                                                                                 o
                                                                                 o
                                                                                 Q-

                                                                                 _o;
                                                                                 JD
                                                                                  03
                                                                                           en
                                                                                           a)
                                                                                           rH
                                                                                           a.

                                                                                           a]
                                                                                           en
                                                                                           C
                                                                                           a)

                                                                                           •rl
                                                                                           T3
                                                                                           0)
                                                                                           en
                                                                                            Ml
                                                                                            C
                                                                                            o

                                                                                            CO
                                                                                           O
                                                                                           cfl
                                                                                           i-i
                                                                                           C
                                                                                           o
                                                                                           en
                                                                                           •H
                                                                                           j-i
                                                                                           cfl

                                                                                           i-
                                                                                           O
                                                                                           U
                                                                                           r^.

                                                                                           OJ
                                                                                           Jj
                                                                                           3
                                                                                           60
                                                                                           •H
                                             29

-------
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
I
I
I
procedure-sample interaction noted in the analysis of variance.  Still,
deletion of the Lake Erie samples from the data set reduced the magnitude but
did not eliminate the statistical significance of the interaction between
samples and procedures (p<0.0001).  This means conversion of BAPP estimates by
a linear regression procedure would incur statistically significant bias,
though possibly inconsequential for management purposes, for some
sample-procedure combinations, even if the Lake Erie samples were unique and
could be disregarded as outliers.

     Concern over the irregular extraction results obtained for the Lake Erie
sediments by the De Pinto procedures led to futher tests of the sediments
using the procedures.  These tests revealed that the Lake Erie samples were
extremely sensitive to solution:sediment ratio used for the extraction.  For
example, when Lake Erie #2 (Sample #8) was re-extracted using the reagents of
the De Pinto method, but employing a solution to sediment ratio comparable to
the Baker and Armstrong procedures (approximately 2000:1),  1300 ug P/g was
extracted compared to 250 ug P/g originally extracted with the lower ratio
(500:1).  The higher amount,  1300 ug P/g, is virtually identical to the BAPP
estimate obtained by the Baker procedure for that sample (Table 3 and Appendix
D).  Clearly the estimate of BAPP from the stored Lake Erie sediments was
sensitive to the solid:solution ratio used during the extraction.

Comparison With Bioassays

     The results of 14 day algal bioassays for 10 sediment samples are
summarized in Table 4.  The results shown are mean values of triplicate
bioassay measurements performed on each sample.  The total P concentrations of
the sediments placed in the DCDA's are also presented.  Bioassays were not
performed, however, using either the Raisin River or the National Bureau of
Standards samples.   It was believed that the high levels of heavy metals
(Cr=29,600 ug/g, Zn=1720 ug/g, Pb=714 ug/g) in the NBS sample would adversely
affect algal growth and bias  the results; lack of sufficient sample volume
precluded analysis of the Raisin River sediment sample.

     All of the samples,  except those from Lake Erie, released essentially all
BAPP (algal-available P)  within 5.0 days.  Release of P from the Cattaraugus
Creek and Genesee River sediments was extremely rapid and was virtually
complete within 2.0 days.  The Lake Erie samples,  on the other hand, continued
to release considerable amounts of P throughout the two week bioassay
analysis.  The amount of P determined by bioassay to be bioavailable ranged
from 2.6 to 2100 ug P/g or from 0.6 to 84.9 percent of total sediment P.  As
an illustration of the nature of the BAPP bioassays,  cumulative P release is
plotted in Figure 8 as a function of time for bioassays performed on Samples
#1 and 7 (Cuyahoga River and  Lake Erie #1).

     A comparison of bioassay results with those obtained by Martin et
al. (1983),  who used the same bioassay procedure as reported here on some of
the same rivers, is presented in Table 5.  The results shown are mean values
of triplicate bioassay measurements performed on several samples from each
river.   Martin et^ al. (1983)  calculated ultimate BAPP values from data on
cumulative P uptake by algae  over time.  The calculation assumed the rate of
uptake depended only on the rate of release of BAPP from the sediment
particles and the amount  of BAPP initially present on the particles at the
start of the bioassay.  In this investigation, calculation of ultimate BAPP
                                     30

-------

TABLE 4.  AVAILABLE PHOSPHORUS RELEASED BY SEDIMENTS DURING BIOASSAY
          EXPERIMENTS, ug P/g.
SAMPLE
TOTAL P
    DAY OF ALGAE HARVEST

2.0     5.0     9.0    14.0
CUYAHOGA

CATTARAUGUS

L. ERIE #1

L. ERIE #2

GENESEE

MAUMEE #1

MAUMEE #2

OSWEGO

SANDUSKY #1

SANDUSKY #2
  1124.3   335.4    45.7

   423.9     2.6

  2209.9   696.3   467.2

  2472.9  1054.9   601.9

   607.5    55.8     0.6

  1123.3   255.8    53.0

   836.8   207.6    25.7

  1007.2    96.5     8.2

   987.9   195.4    32.4

   900.0   125.3     8.2
              245.5

              281.0
  7.4



117.6

153.6
                        1.2
            CUMULATIVE P UPTAKE

            ULTIMATE   % TOTAL P
 388.5

   2.6

1550.0

2100.0

  56.4

 308.8

 233.4

 105.9

 227.8

 183.5
34.5

 0.6

70.1

84.9

 9.3

27.5

27.9

10.5

23.2

20.4
                                       31

-------

                 ,-»-»-    ooooooooo
   I  I  I  I   I
CM O> 80 h*  
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
 TABLE 5.  AVERAGE AVAILABLE PHOSPHORUS RELEASED BY
           SEDIMENTS DURING BIOASSAY EXPERIMENTS,
           ug P/g  (FROM MARTIN ET AL., 1983).
SOURCE
 NUMBER

   OF

SAMPLES
TOTAL P
CUMULATIVE P UPTAKE


ULTIMATE   % TOTAL P
CUYAHOGA          4    1314.0


CATTARAUGUS       5     559.0


L. ERIE #1        1    2656.0


L. ERIE #2        1    3044.0


GENESEE           1     900.0


MAUMEE           11    1308.0


SANDUSKY         17    1145.0
                         449.2


                          38.8


                        1435.0


                        1482.4


                         173.8


                         337.3


                         247.1
                           33.9


                            7.7


                           54.0


                           48.7


                           19.3


                           25.0


                           21.4
                                  33

-------

release was done only for the samples assayed from Lake Erie, since they
continued to release P throughout the 14 day procedure.  Calculation of
ultimate BAPP for the other samples was not necessary since they had released
essentially all BAPP by the end of 5.0 days.  For these samples, the amounts
of P reported as ultimately available in Table 4 correspond to the total
measured uptake by the algae during the bioassays.

     Comparison of the bioassay results from this investigation with those
obtained by Martin e_t al. (1983) revealed highly similar results for the three
Ohio rivers (Cuyahoga, Maumee, Sandusky).  However, the Cattaraugus and
Genesee river sediments released somewhat less P than that reported by Martin
et al.  (1983).  This discrepancy may be due to the fact that Martin e_t al.
analyzed water column suspended sediments, while in this study measurements
were performed on bottom sediments collected from the two New York rivers.

     Comparison of the two samples from Lake Erie revealed a pronounced
dissimilarity in the findings from each investigation.  Martin et al. (1983)
found that 48.7 and 54.0 percent of the total sediment P was released from
these same samples.  On the other hand, results from the present investigation
showed that the samples contained amounts of ultimate BAPP corresponding to
70.1 and 84.9 percent of total P, respectively (actual measured levels were
69.1 and 84.6 percent, respectively).  The reasons for the wide differences in
results between the two investigations are unknown; however, the duration of
storage between the analyses of Martin et al. (1983) and those done for this
investigation (approximately 4 years) may have permitted changes in the
P-retaining properties and BAPP levels of the sediments.  This will be
discussed subsequently.

     In order to assess how well the chemical extraction procedures compared
with BAPP determined by the algal bioassays, correlation coefficients were
calculated between the bioassay measured BAPP and that estimated from chemical
measurements (Appendix E.6).  The results of the comparisons are shown
graphically in Figure 9.  The methods of Baker,  Armstrong,  and CCIW all showed
a high degree of correlation (r>0.981 for N=10)  with the bioassay data,
implying that these procedures could be used for accurate prediction of BAPP.
Comparisons involving algal-determined BAPP and the two variations of the
De Pinto method were strongly biased by the results for the Lake Erie samples
and showed poor correlations as a consequence.

     As noted in comparing the chemical procedures for BAPP, the Lake Erie
samples gave atypical results, at least by comparison with the results for the
other samples as analyzed by the various procedures.  This anomaly may have
been a consequence of the age of the Lake Erie sediment samples, which had
been held approximately four years in refrigeration storage.  Although the
ageing mechanisms were not investigated, two processes may have been
involved.  The first would include microbial oxidation of relatively
unavailable organic P to relatively bioavailable,  inorganic P during storage.
Accompanying this process would be a second one: slow oxidation and higher
ordering of Fe oxides which would increase levels of occluded non-apatite
inorganic P.  This latter fraction is not generally considered to be
extractable by the procedure used by De Pinto,  but is extractable by the
procedure of CCIW, and apparently is by that of Armstrong (Armstrong et
al. 1979) and possibly Baker (1983).
                                     34

-------
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
I





UjL
W
tf
5s
° E
0 1
O X
PS •

"o
p^ *™
O o
JO
^r
o g
OT o
*^ o
- . X \<
Q. \ \ *
E \ N \
i \ '• \ \
1 \ '••- \ \ I
^ \D > '. X \ <3 ^ •»
'\ \ \
\ \
\ \ \
\ \
\ \
\ ^
\
\ \ \
\
i » \ \
Z \

__| LU _J Q; (_ O >• ^'V
 '• V
n-i-i-Sss a tv^
ooo •• »
a >•.
a + o^xt* n *
a
i i i i i i i i i i i i n
| 0Q (O ^^ f^ ^^ QQ JQ ^^ ^^ ^J JQ Ifl ^^
ICMCMCMCMCM*— T-»— »— r-OOO























il
* 1
^t
^M
"3L
'• ^Vii
i ^
CM C
6 c



^ft*
" CM

-

q
" CM
-
ID



. °1



GO
" CD




" 6



o
o
i
i
3/gw
                                      w
                                      CD
                                      CO
                                      CO

                                      O
                                     ho
CL
Q.
                                     CD
                                            ex
                                            6
                                            cfl
                                            CO
                                            Pu

                                            
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Interconversion of BAPP Estimates

     In view of the solution:sediment sensitivity of the stored Lake Erie
samples and the possibly related anomalous extraction results, the analysis of
correlation between the extraction procedures was recast with a focus on
freshly collected samples.  This analysis is summarized in Table 6. The
tabulated data show strong intercorrelation (r^0.85, p^O.Ol) among all the
BAPP procedures, including chemical and biological estimates, plus total P.
As stated earlier, this suggests that these BAPP estimation methods quantitate
similar physicochemically-bound fractions of total P, but do so with different
efficiencies.  The strength of the intercorrelations indicates that
interconversion of BAPP estimates is possible with good accuracy and precision
as long as fresh sediment samples are used for initial BAPP estimates.  Also,
as shown in Figure 10, all the procedures correlate well with the algal
bioassay data, which means transformation of a chemical BAPP estimate to a
biological one is feasible for all methods, again with the caveat of using
fresh sediments for analysis.

     Regression equation slopes and intercepts, developed for interconversion
of BAPP estimates using the data acquired and procedures tested in this study,
also are given in Table 6. With respect to prediction of algal-available P,
the procedures of De Pinto gave results that were closest in magnitude to the
values determined by bioassay.  This is indicated in Table 6 by the values of
the slopes relating bioassay estimates as a dependent variable (BIOASSY as a
Row Header) to either of the De Pinto procedures (CCTCENT or CCTFILT as a
Column Header); each slope is essentially 1.0.  Thus, these procedures could
be considered to be the most accurate predictors of algal BAPP among those
tested.  For these same data, however, the procedures of Baker, Armstrong, and
CCIW were the most precise predictors of algal BAPP, since they correlated
most highly with the bioassay results.

     Using a similar all-pairwise approach but considering both fresh samples
and the stored ones from Lake Erie (Appendix E.6),  the procedure of Armstrong
is shown by its excellent correlation (r=0.994, N=10) to be the most precise
predictor of the bioassay results.   However,  the strong intercorrelation
between the Armstrong, Baker, and CCIW (and algal BAPP) estimates means any of
these chemical methods for BAPP estimation would be precise for both fresh and
stored samples.  Thus, it is recommended that either the Baker, Armstrong, or
CCIW procedures be selected when samples that have been stored must be
chemically analyzed for BAPP.  When freshly collected bottom or suspended
sediments are to be analyzed, however, it appears that any of the five
extraction procedures may be used,  and the results interconverted between each
procedure with sufficient accuracy to suit a wide range of purposes.
                                     36

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
TABLE 6. ALL PAIRWISE COMPARISON OF EXTRACTION PROCEDURES.*
                              EXTRACTION PROCEDURE**
EXTRACTION
PROCEDURE
     TP   CCTCENT  CCTFILT  BAKER   ARMSTRNG   CCIW   BIOASSY
TP
          0.8902
          0.0030
          1.9210
          463.3
          0.8947
          0.0027
          1.9799
          462.4
          0.9266
          0.0009
          1.5357
          471.7
          0.9143
          0.0015
          1.5416
          468.7
CCTCENT
CCTFILT
 0.8902
 0.0030
 0.4125
-146.5

 0.8947
 0.0027
 0.4043
-145.3
0.9999
0.0001
1.0253
0.6375
0.9044
0.0020
0.6946
32.11
0.9561
0.0002
0.7470
17.45
 0.9999
 0.0001
 0.9751
-0.5670
          0.9064
          0.0019
          0.6789
          30.33
          0.9567
          0.0002
          0.7290
          16.31
          0.9005
          0.0023
          1.0778
          413.0

          0.9425
          0.0005
          0.5228
         -9.699

          0.9437
          0.0004
          0.5104
         -10.34
BAKER
0.9266
0.0009
0.5591
-226.6
0.9044
0.0020
1.1777
10.11
0.9064
0.0019
1.2103
10.263
ARMSTRNG
CCIW
BIOASSY
 0.9143
 0.0015
 0.5423
-210.7

 0.9005
 0.0023
 0.7523
-229.4

 0.8452
 0.0082
 0.4483
-203.1
 0.9561
 0.0002
 1.2237
 1.348

 0.9425
 0.0005
 1.6992
 64.52

 0.8666
 0.0054
 0.9918
-23.49
 0.9567
 0.0002
 1.2556
 1.926

 0.9437
 0.0004
 1.7446
 65.09

 0.8669
 0.0053
 1.0174
-22.97
 0.9869
 0.0001
 0.9701
 9.020

 0.9516
 0.0003
 1.3176
 82.92

 0.9581
 0.0002
 0.8421
-32.00
 0.98691   0.9516
 0.0001   0.0003
 1.0040   0.6873
-2.208   -32.11

          0.9657
          0.0001
          0.6856
         -30.24
       \
 0.9657
 0.0001
 1.3603
 70.11
          0.8452
          0.0082
          1.5938
          573.9

          0.8666
          0.0054
          0.7572
          71.34

          0.8669
          0.0053
          0.7387
          68.93

          0.9581
          0.0002
          1.0900
          56.49

          0.9500
          0.0003
          1.0624
          62.87

          0.9340
          0.0007
          1.4713
          150.7
 0.9500
 0.0003
 0.8495
-34.91
 0.9340
 0.0007
 0.5929
-65.11
*  Results averaged within samples by procedure prior to analysis; data
   not included from Lake Erie bottom sediment, Raisin River, and NBS
   SRM.  For all statistical significance tests, N = 8 and  r|  > 0.707
   for ,p < 0.05.  All correlations significant at p < 0.01.

** Format of Tabulated Data:
        Coefficient of Correlation (r)
        Probability of Zero Correlation (Ho: r=0)
        Slope of Bivariate Regression Lines (bl),  [(ug/g)/(ug/g)]:
             Column headers are independent variables
             Row headers are dependent variables
        Intercept of Bivariate Regression Line  (bo),  [ug/g]
                                         37

-------
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
I
CO

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
                                  REFERENCES


APHA 1981. Standard methods. 15th ed. American Public Health Assoc.
  Washington, DC.


Armstrong, D.E., Perry, J.R., and Flatness, D. 1979. Availability of
  pollutants associated with suspended or settled river sediments which gain
  access to the Great Lakes.  Final Report, Vol. 11, Menomenee River Pilot
  Watershed Study.  Water Resources Center, University of Wisconsin-Madison.
  96 pp.


Baker, D.B. 1983. Tributary loading of bio available P into Lakes Erie and
  Ontario.  Final Report, Grant No. R005708-01; U.S.EPA GLNPO-Chicago, IL.
  44-45.


Chang, S.C. and Jackson, M.L. 1957.  Fractionation of soil P.  Soil Sci. 84:
  133-144.


Chapra, S.C., and Sonzogni, W.C. 1979. Great Lakes total P budget for the
  mid-1970's.  J. Water Pollut. Control Fed. 51: 2524-2533.


De Pinto, J.V. 1982.  An experimental apparatus for evaluating kinetics of
  available P release from aquatic particulates.  Water Res. 16: 1065-1070.


De Pinto, J.V., Young, T.C. et al. 1980.  Phosphorus removal in lower Great
  Lakes municipal treatment plants.  EPA-600/2-80-117, MERL, Cincinnati, OH.
  147 pp.


De Pinto, J.V., Young, T.C., and Martin, S.C. 1981.  Algal-available P in
  suspended sediments of lower Great Lakes tributaries.  J. Great Lakes Res.
  7: 311-325.


Gilmartin, M. 1967. Changes in inorganic phosphate concentration occurring
  during seawater sample storage.  Limnol. Oceanogr. 12: 325-328.


Heron, J. 1962. Determination of phosphate in water after storage in
  polyethylene.  Limnol. Oceanogr. 7: 316-321.


IJC. 1970. Pollution of Lake Erie, Lake Ontario, and the international section
  of the St.  Lawrence River. International Joint Commission, Windsor, ONT.


	. 1972.  Great Lakes water quality agreement with annexes and texts and
  terms of reference, between the United States and Canada, signed at Ottawa,
  April 15, 1972. International Joint Commission, Windsor, ONT.


	. 1978.  Great Lakes water quality agreement of 1978, with annexes and
  terms of reference, between the United States and Canada, signed at Ottawa,
  November 22, 1978. International Joint Commission, Windsor, ONT.
                                     39

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Johnson, A.H., Bouldin, D.R., and Hergert, G.W.  1975.  Some observations
  concerning preparation and storage of stream samples  for dissolved  inorganic
  phosphate analysis.  Water Resour. Res.  11: 559-562.


Klingaman, E.D., and Nelson, D.W. 1976. Evaluation of  methods  for preserving
  the levels of soluble inorganic P and nitrogen in unfiltered water  samples.
  J. Eviron. Qual. 5: 42-46.


Logan, T.J., Verhoff, F.H., and De Pinto,  J.V. 1979. Biological availability
  of total P.  LEWMS Report, U.S. Army Corps of Engineers-Buffalo, NY.


Martin, S.C., De Pinto, J.V., and Young, T.C. 1983. Biological availability of
  sediment P inputs to the lower Great Lakes.  Final Report, Grant No.
  CR807155; U. S.EPA/ERL-Duluth, MN  (LLRS-Grosse He, MI); 170 pp.


Murphy, J., and Riley, J.P. 1956. The storage of seawater samples for the
  determination of dissolved inorganic phosphate.  Anal. Chim. Acta.  14:
  318-319.


Murphy, J., and Riley, J.P. 1962. A modified single solution method for the
  determination of phosphate in natural waters.  Anal.  Chim. Acta. 27: 31-36.


Salisbury, D.W., De Pinto, J.V., and Young T.C., 1983.  Impact of
  algal-available P on Lake Erie water quality: mathematical modeling. Final
  Report, Grant No. CR807155-03; U.S.EPA/ERL-Duluth, MN (LLRS-Grosse  He, MI);
  72 pp.


Thayer, G.W. 1970. Comparison of two storage methods for the analysis  of
  nitrogen and P fractions in estuarine waters.  Chesapeake Sci. 11:  155-158.


USEPA. 1976. Methods for chemical analysis of water and wastes.
  EPA-625/6-74-003a, Environmental Monitoring and Support Laboratory, US
  Environmental Protection Agency, Cincinnati, OH.


Wetzel, R.G. 1983. Limnology. Saunders. New York. 766 pp.


Mayer, T. and Williams, J.D.H. 1981.  Modified procedure for determining the
  forms of P in freshwater sediments.  Technical bulletin No.  119.   NWRI,
  Canada Centre for Inland Waters.  Burlington, ONT.


Williams, J.D.H. e_t al_. 1971.  Fractionation of inorganic phosphate in
  calcareous lake sediments.  Soil Sci. Soc. Amer. Proc. 35: 250-255.


Williams, J.D.H. e_t al. 1980.  Availability to Scenedesmus quadricauda of
  different forms of P in sedimentary materials in the  Great Lakes.
  Limnol. Oceanog. 25: 1-11.


Yaksich, S.M. et al. 1982.  Lake Erie nutrient load, 1970-1980.  LEWMS,
  U.S.Army Corps of Engineers, Buffalo District, NY.


Young, T.C., De Pinto, J.V. et ad. 1982.  Algal availability of P in municipal
  wastewater.  J. Water Pollut. Control Fed. 54: 1505-1516.
                                     40

-------
APPENDICES
  41

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
                     APPENDIX A


DETAILED DESCRIPTIONS OF SEDIMENT ANALYTICAL METHODS
                        A.I

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
APPENDIX A.  Detailed Descriptions of Sediment Analytical Methods


THE NAOH EXTRACTION METHOD OF DE PINTO (1982)

Introduction

     Given in the following is a specific description of a procedure for the
measurement of a fraction of the particulate phosphorus in a water sample that
shows close correlation with the quantity of the particulate phosphorus which
is biologically available.  In general,  the procedure requires collection of a
relatively large volume of raw water which contains particulate matter.  The
particles are then collected by filtration onto a membrane filter and
resuspended in a phosphorus-free medium to form a sediment concentrate.
Aliquots of the sediment concentrate are put in contact with approximately 0.1
N NaOH for a specified period.  Then, the particles are separated from
solution by filtration and the filtrate is analyzed for orthophosphate by
standard colorimetric methods.  The concentration of orthophosphate in the
NaOH extract, divided by the concentration of solids during the extraction,
yields the estimated quantity of available particulate phosphorus per unit dry
weight of solids.  Multiplication of this value by the suspended solids
concentration of the original water sample equals the concentration of
available particulate phosphorus in the original sample.  The analysis should
be performed in duplicate.

Sample Collection

     The volume of sample that should be collected is dependent on the
suspended solids concentration of the water being sampled.  In all cases
collect a volume of sample that will yield a minimum of 0.50 ug of sediment.
With careful handling of the sample, 0.50 ug of sediment will be adequate to
perform suspended solids measurements on both the raw sample and the sediment
concentrate that will be prepared, and to perform duplicate NaOH extractions
on the sediment.  Low solids concentrations, for example, 20 to 100 mg/1, will
naturally require a larger volume of sample than will a solids concentration
of 200 to 1,000 mg/1.  Regardless of the solids concentration it is wise to
collect more sample than is needed. Samples should be stored in polyethylene
bottles or carboys, transported in cooled containers, and stored at 4 C.
Samples should be held for processing no longer than 48 hrs.

Cleansing of Glassware and Apparatus

   ,  Wash all glassware, storage containers, and other apparatus that will
come in contact with the sediments and solutions with 1+1 HC1 and rinse
thoroughly with deionized water.  Avoid the use of detergents containing
phosphate.  Wash the membrane filters to be used to separate the NaOH extract
from the sediments by either soaking the membranes in deionized water for 24
hrs  (50 membranes/2 liters ), or soaking the membranes in deionized water (50
membranes/2 liters ) for 3-1 hour periods, changing the deionized water after
each period (Standard Methods, 15th Edition, Section 424A, p. 412).
                                     A.2

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Apparatus


a.  Van Dorn sampler or equivalent


b.  Polyethylene bottles or carboys.


c.  Vacuum filtration unit and vacuum source; side-arm
    flasks, 125 ml and 1,000 ml or larger.


d.  0.4 micron pore diameter polycarbonate membranes.


e.  Sediment Pipets:  Prepare these by enlarging the tip openings of
    graduated Mohr pipets or volumetric transfer pipets depending on the
    type needed.  Generally, a 5 to 10 ml graduated Mohr type will be
    needed.  Use a sediment pipet whenever pipeting a suspended solids
    sample.


f.  Centrifuge tubes with screw-cap tops calibrated to 50.0 ml.
    Alternatively, any plastic or glass container with a water-tight cap
    may be used, provided that it has been calibrated to contain 50.0 ml
    and can accommodate vigorous shaking without leaking for 17 hours.


g.  Drying oven with thermostatic control capable of maintaining 103 C.


h.  Desiccator


i.  Analytical Balance, capable of weighing to 0.1 mg.


j.  Rotary shaker table or equivalent, with adjustable speed control.


k.  Spectrophotometer with infrared phototube for use at 880 nm.


1.  Magnetic stir table and stir bars.


Reagents


a.  1+1 HC1 acid wash solution.


b.  0.125 N NaOH extractant solution:  Dissolves 5.0 g of NaOHin 1,000 ml
    of deionized water.


c.  Stock phosphate solution 1.00 ml  50.0 ug P0,_p.  Dissolve 219.5 mg
    of anhydrous potassium dihydrogen phosphate CKH pQ j £n ^ QQQ m-j_
    of deionized water.


d.  Standard phosphate solution, 1.00 ml  1.00 ug P0,_p.

   . Dilute 10.00 ml of the 1.00 ml  50.0 ug P04_p stock'phosphate
    solution to 500 ml.


e.  Resuspension Medium:  0.1  N NaCl: Dissolve 5.8 NaCl in 1,000 ml of
    deionized water.


f.  Phenolphthalein indicator  solution, aqueous.


B'  H2S04 Neutralization acid,  5 N: Use the 5 N H2S04 solution



                                     A.3

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
    prepared for the combined reagent.

h.  Solutions for combined reagent:

       Sulfuric Acid, 5 N
       Potassium antimonyl tartrate
       Ammonium molybdate
       Ascorbic acid, 0.1 M

    Prepare these solutions as outlined in Standard Methods, 15th Edition,
    1980, Section 424F., p. 420: Ascorbic Acid Method.

i.  Combined reagent:  Combine the reagents above as outlined in Standard
    Methods, ibid.

Preparation of Sediment Concentrate

     Initially, it is necessary to determine the suspended solids
concentration of the raw sample using the method outlined in Standard Methods
(15th Edition, 1980, Section 209G, p. 97).  An exception to the Standard
Methods procedure is that 0.4 micron pore diameter polycarbonate membranes
should be substituted for glass fiber filters.  Always ensure that the sample
is well mixed before removing aliquots for solids measurement,  and take for
analysis a volume of sample that will yield 25 to 50 mg of dry nonfiltrable
residue on the membrane.

     From the solids measurement, calculate the volume of raw sample
equivalent to at least 0.40 g of dry sediment and filter this volume of raw
sample through a 0.4 micron pore diameter polycarbonate membrane.  More than 1
membrane may be required due to pore blockage as the sediment accumulates on
the membrane.  The filtration time and the number of membranes necessary will
be a function of the suspended solids concentration of the raw sample and the
grain size distribution of the sediment particles.

     After filtering the required volume of raw sample, transfer the sediment
from the membranes to a 125 ml Erlenmeyer flask by scraping the bulk of the
sediment from the membrane with a spatula. Rinse any sediment sticking to the
spatula and remaining on the membranes into the concentrate flask with the
medium for resuspension, 0.1 N NaCl dispensed through a polyethylene squirt
bottle.  Also, recover any sediment that sticks to the filtration apparatus on
the edge that comes in contact with the membrane.  When rinsing the spatula
and membranes take care not to exceed the volume of suspension medium needed
to approximate a 10 tag/ml suspended solids concentration.  After the collected
solids have been transferred to the erlenmeyer flask, add additional 0.1 N
NaCl to yield a final suspended solids concentration of approx. 10 mg/ml.  For
example, if enough raw sample is filtered to yield 0.40 g of sediment, the
final volume of suspension media should be 40 ml.  Likewise, if enough raw
sample is filtered to yield 0.50 g of sediment, the final volume of suspension
media would be 50 ml.

     When the concentrate is prepared, add a magnetic stir bar, stopper or
seal the flask with parafilm, and mix the concentrate to homogeneity on a
magnetic stir table.  The sediment concentrate then may be further processed
or stored at 4 C up to 60 days.
                                     A.4

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Suspended Solids Determination on Sediment Concentrate

     Determine the solids concentration of the concentrate by the method used
for the raw sample.  For an estimated solids concentration of 10 rag/ml, filter
3 to 4 ml of the concentrate. Remove aliquots by sediment pipet from the
concentrate while it is actively being mixed on a magnetic stir table.

NaOH Extraction

     From the suspended solids value of the concentrate, calculate the volume
of concentrate equivalent to 0.1 g of dry sediment.  This volume should not be
less than 6.0 ml or exceed 14.0 ml.

     With a sediment pipet transfer the volume of concentrate equivalent to
0.1 g dry solids to a 50 ml centrifuge tube.  Fill the tube to the 50 ml mark
with 0.125 N NaOH.  Cap the tube and mix its contents continuously for 17 hrs
at room temperature with a mechanical shaking device such as a rotary shaker
table. Adjust the rate of mixing so that it is just fast enough to keep the
sediments well mixed within the tube.  The use of 0.1 g of sediment in a 50 ml
centrifuge tube is not mandatory.  However, the solutiontsolid ratio during
the extraction should be kept within the range of 500 to 600:1.  Other
sediment aliquots and extraction volumes may be employed at the analyst's
discretion if the solution:solid ratio is maintained within this range.

     Following the 17 hrs of mixing, divide the NaOH extract into
approximately equal portions and separate the NaOH extract from the sediments
of one portion using vacuum filtration through a pre-washed 0.4 micron pore
diameter polycarbonate membrane and into a dry 125 ml side-arm flask.  On the
other portion, separate the sediments and extract by high-speed centrifugation
(~32,000xg) for 30 minutes.

Measurement of Reactive NaOH-Extractable Phosporus

     The orthophosphate concentration of the NaOH extract generally falls in
the range: 200-800 ug P/L.  Make dilutions of the extract as determined by its
expected phosphorus concentration and the light path length used with the
spectrophotometr.   A table containing light path lengths and the range of
phosphorus concentrations that can be measured at each light path length is
presented in Standard Methods, 15th Edition, 1980, Section 424 F,  p. 420.
Pipet the appropriate volume of filtered extract into a 50 ml volumetric
flask.

     To the volumetric flask containing the required aliquot of extract add 1
drop of phenolphthalein indicator.   The extract should turn red.  Add 5 N
 zSO^ dropwise (swirling the flask between drops) until the red color
is discharged.  Fill the volumetric to the 50 ml mark with deionized water and
transfer the contents of the volumetric to a dry 125 ml Erlenmeyer flask.  Add
8.0 ml of the combined reagent to the Erlenmeyer and mix thoroughly by
swirling the flask.  Allow at least 10 minutes for color development and
measure the absorbance of the sample at 880 nm within 30 minutes of the
combined reagent addition (Standard Methods, 15th Edition,  1980, Section 424
F, p.  420).  Use a deionized water-combined reagent blank as the reference
solution.
                                     A.5

-------
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
I
Preparation of Standard Curve


     Prepare a series of 5 phosphorus standards that will bracket the expected
phosphorus concentration in the diluted extract.  Prepare these standards by
making dilutions of the 1 ug P/ml standard phosphorus solution in 50 ml
volumetric flasks.  Also prepare 2-50 ml deionized water blanks.   Once the
dilutions have been made transfer the contents of the volumetrics to dry 125
ml Erlentneyer flasks, add 8.0 ml of the combined reagent to each standard and
blank, and mix thoroughly by swirling.  Allow 10 minutes for color development
and measure the absorbance of the samples against the reagent blanks at 880 nm
within 30 minutes of the mixed reagent addition.
                                     A.6

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
THE NAOH EXTRACTION METHOD OF BAKER (1983)


a.  Filter 50 ml of a well-mixed sample through a Millipore BD (polyvinyl
    chloride) 0.6 micron pore-diameter membrane.  (Millipore BDWP 04700).
    Keep filtrate and filter.


    For this step, the volume filtered should be adjusted to permit
    collection of between 2 and 20 mg of sediment on the filter.
    For this reason,  suspended sediment analysis should preceed setting up
    this test.


b.  Place the filter in a French Square bottle.  A French Square may be
    replaced with any glass screw cap bottle which will hold 25 ml, have
    enough room for agitation, and fit on a shaker.


c.  Add 25 ml of 0.1 N NaOH in 1.0 N NaCl to the French Square.


d.  Cap the French Square and place it on a shaker bath for 18 hours.


e.  At the end of the 18 hour period,  filter the extract through a
    Millipore AP prefilter (Millipore AP25 04700) and into a 125  ml
    Erlenmeyer flask.


f.  Add 25 ml of 0.1 N sulfuric acid to neutralize the extract, and mix
    well.


g.  Determine the soluble reactive phosphorus content of the neutralized
    extract by the modified single-reagent method (APHA, 1980).
                                     A.7

-------
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
I
 I
THE NAOH EXTRACTION METHOD OF ARMSTRONG ET AL. (1979)


a.  By high-speed centrifugation (13,000 rpm, fixed-angle rotor with 50 ml
    bottles), of a suitable volume of water sample, or previously
    concentrated sample (see Method A, above), collect approximately 15 mg
    suspended sediment (dry weight) into a 50 ml (nominal volume)
    polypropylene centrifuge bottle.


    To determine the exact weight of suspended solids collected in the
    tube, analysis of suspended solids on the sample prior to
    centrifugation, would be required to calculate the liquid volume needed
    for centrifugation.


b.  Add 30 ml of 0.1 N NaOH in 1.0 N NaCl to the sediments in the
    centrifuge bottle.  Cap the bottle and place on a shaker table for 18
    hours.


c.  At the end of the 18 hour period of extraction, separate the sediments
    from the extract, first by high-speed centrifugation (as above), then
    filter the extract through a 0.4 micron pore-diameter membrane
    (polycarbonate).


d.  Neutralize the filtered extract with 2 N sulfuric acid,  using
    phenolphthalein as an indicator.


e.  Determine the soluble reactive phosphorus content of the neutralized
    extract by the modified single-reagent method (APHA, 1980).
                                     A.8

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
THE NON-APATITE EXTRACTION METHOD OF WILLIAMS AMD MAYER (1980)


CDS Extractable Inorganic Phosphorus


    a.  Weigh 0.1 g of freeze-dried particulate matter into a 100 ml
        centrifuge bottle.


    b.  Add 50 ml of 0.22 M Na-citrate/0.11 M Na-bicarbonate reagent.


    c.  Immerse tubes in 85 C water bath.  After 15 minutes add 1.0 g
        Na-dithionite.  Maintain samples in water bath at 85 C for a
        further 15 minutes stirring frequently with a glass rod.


    d.  Centrifuge extract for 15 minutes at 2000 rpm.


    e.  Transfer supernatant quantitatively into 100 mL volumetric flasks,
        leaving residue undisturbed in the tubes.


    f.  Add 25 mL of 1.0 M NaCl solution to each tube.  Wash residue well
        by vortex mixing.


    g.  Centrifuge the extract at 2000 rpm for 10 minutes.


    h.  Transfer extract to volumetric flask as in (e).  Save residue for
        NaOH extraction.


    i.  Add 1 mL FeCl3 to each flask<


    j.  Allow flasks to stand exposed to atmosphere for 2-3 days (covered
        with filter paper or paper towels) until development of
        yellowish-brown color indicates that the oxidation of the dithionite
        is complete.


    k.  Filter sample, adjust pH to neutral. Make up to 100 mL and analyze
        for ortho-P.


    1.  Standard of 500 ug of P and distilled water blank also carried
        through the procedure.
NaOH Extractable Inorganic Phosphorus


    a.   Add 50 ml of 1 N NaOH to the residue from the procedure for
        extraction of CDB-P (D.I. above).


    b.   Stopper, shake well and place tubes on moon walker apparatus over
        night.


    c.   Centrifuge for 15 minutes at 2000 rpm.


    d.   Transfer 20 mL aliquot of the supernatant solution to another
        centrifuge tube.


    e.   Add 10 mL of 3 N HC1 to centrifuge tube containing 20 mL aliquot.
                                     A.9

-------
I
•
I
I
I
               f.  Mix with vortex mixer and centrifuge for  15 minutes at 2000 rpm.
              g.  Transfer a  10 mL aliquot of  the clarified supernatant to a 50 mL
•                volumetric  flask.
              h.  Adjust pH to neutral with 0.1 N HC1 and make up to 50 mL.
|            i.  Analyze for ortho-P.
•            j.  Discard remainder of NaOH extract.
              k.  Standard of 50 ug of P and distilled water blank also carried
                  through the procedure.

                  NOTE:  Non-Apatite P   NaOH-Extractable P + CDS Extractable P
I

I

I

I

I

I

I

I

I

I
                                              A. 10

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
THE BIOASSAY METHOD OF DE PINTO ET AL.U981) AND DE PINTO (1982)
    Using the two chambered, continuously-stirred, Dual Culture Diffusion
Apparatus (DCDA) of De Pinto (1982), sediments for bioassay analysis of
available phosphorus are introduced into a darkened, "phosphorus-release"
vessel and phosphorus- starved algae are placed in a transparent "algal-assay"
vessel. The two vessels are clamped together such that the
vessel contents are separated only by a 0.4 micron pore-diameter membrane
filter.  The membrane allows diffusion of soluble substances,  such as
phosphorus,  from one vessel to the other, but it keeps the algae and
sediments in their respective vessels.  The rate of diffusion across the
membrane does not limit the rate of movement of phosphorus from the
sediments to the algae, since it has been found to be large relative to
typical release rates for natural sediments.

    The algae to be used for the bioassays will be harvested from
phosphorus-starved, seven day cultures of Selenastrum c ap r i c o rnut urn.
The algal medium to be used for all bioassay work will be one modified from
that of Guilliard and Lorenzen (1972), with the only modification being that
involving phosphorus additions.  For growing phosphorus-starved cultures,
the usual total phosphorus content of algal growth media is approximately 30
ug P/L.

    No attempt will be made to prevent growth of the natural bacterial
flora that accompanies the sediments upon innoculation into the release
vessel.  Bioassay incubations will be done at ambient temperature in an air
conditioned room (23 C +/- 10 ) with lighting supplied by approximately 4000
lux of cool-white fluorescent lamps.

    After an initial three-day incubation period,  which is lengthened to
seven and ten days later in the test,  the contents of the assay vessel are
harvested and immedieatly replaced with another seven-day, phosphorus-
starved culture of bioassay algae.  The cycle of harvent and replacement of
assay vessel contents is continued for a period of 21 to 30 days,  depending
on the extent and rate of available phosphorus release from the particulate
matter.

    The total phosphorus content of the initial and subsequently harvested
or seeded algal cultures is known from measurements made directly on the
cultures of interest.  By assuming algal accumulation of phosphorus in the
assay vessel is the result of available phosphorus release from the
particulate matter and subsequent transport of that phosphorus through the
membrane partition to the algae,  followed by algal uptake, and by invoking
the principle of mass conservation, the rate and extent of available
phosphorus release may be calculated over each sampling interval.   Past
studies (De Pinto et al., 1981) have shown the release of sediment-bound
phosphorus to be approximately first-order, dependent on the amount of
phosphorus that may ultimately be released for algal uptake.
                                    A.11

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
                   APPENDIX B
CONCENTRATION OF SRP DURING STORAGE EFFECTS STUDY
                      B.I

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
APPENDIX B


CONCENTRATION OF SOLUBLE REACTIVE PHOSPHORUS (ug P/L) IN WATER SAMPLES

COLLECTED FROM THE HURON RIVER AS A FUNCTION OF TIME AND TEMPERATURE

OF INCUBATION.


Field Zero
Incubation
11.
9.
10.
10.
17.
10.
10.
11.
11.
11.
8
9
2
8
5
8
5
1
5
1
Incubation
11.
9.
10.
10.
17.
10.
10.
11.
11.
11.
8
9
2
8
5
8
5
1
5
1
Incubation
11.
9.
10.
10.
17.
10.
10.
11.
11.
11.
8
9
2
8
5
8
5
1
5
1
Elapsed
Lab Zero 0.
Temperature =
4.9 2.
4.2 2.
4.5 3.
9.5 7.
9.4 8.
9.5 8.
8.7 8.
9.5 7.
8.8 8.
8.9 7.
Temperature =
4.9 3.
4.2 2.
4.5 2.
9.5 9.
9.4 8.
9.5 7.
8.7 7.
9.5 10.
8.8 10.
8.9 9.
Temperature =
4.9 14.
4.2 18.
4.5 15.
9.5 4.
9.4 4.
9.5 13.
8.7 6.
9.5 24.
8.8 25.
8.9 24.
Time
5
4C
7
6
2
2
7
3
7
8
0
8
22C
0
6
6
5
4
8
3
0
3
9
45C
8
9
9
0
3
6
7
3
0
3
of Incubation (days)
1.

2.
2.
2.
6.
7.
7.
7.
8.
8.
8.

3.
3.
2.
7.
6.
10.
5.
11.
10.
12.

3.
2.
2.
2.
1.
1.
2.
2.
4.
2.
0

6
6
7
8
0
3
0
4
9
4

2
0
7
5
7
8
4
6
3
4

5
7
1
9
3
8
7
7
8
4
2.

2.
2.
2.
6.
6.
6.
6.
8.
8.
7.

3.
4.
3.
8.
10.
4.
11.
6.
13.
13.

2.
2.
2.
0.
2.
1.
2.
1.
0.
0.
0

6
1
4
7
2
2
4
4
1
0

8
3
5
7
0
3
1
2
8
5

9
1
4
8
1
3
3
0
8
8
4.

3.
3.
4.
7.
8.
7.
7.
10.
9.
9.

8.
8.
8.
16.
14.
8.
17.
8.
9.
9.

3.
3.
2.
2.
2.
3.
2.
0.
0.
0.
0

5
5
3
3
0
6
8
2
7
7

6
7
7
5
3
6
6
0
4
9

8
2
6
6
9
2
1
7
8
7
9.

8.
9.
8.
14.
17.
16.
16.
17.
15.
15.

7.
7.
8.
16.
15.
10.
11.
9.

21.

2.
1.
1.
1.
1.
0.
1.
0.
0.
0.
0

7
5
7
6
6
8
2
1
1
9

8
2
9
2
2
8
9
1

4

9
5
1
1
1
8
0
4
7
5
                                     B.2

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
APPENDIX B


CONCENTRATION OF SOLUBLE REACTIVE PHOSPHORUS (ug P/L) IN WATER SAMPLES
COLLECTED FROM THE MAUMEE RIVER AS A FUNCTION OF TIME AND TEMPERATURE
OF INCUBATION.

Field

Zero
Incubation
94.
96.
96.
75.
75.
77.
79.
73.
75.
72.
7
9
3
7
1
0
2
8
1
2
Incubation
94.
96.
96.
75.
75.
77.
79.
73.
75.
72.
7
9
3
7
1
0
2
8
1
2
Incubation
94.
96.
96.
75.
75.
77.
79.
7.3.
75.
72.
7
9
3
7
1
0
2
8
1
2
Elapsed
Lab Zero 0.
Temperature =
92.2 92.
95.5 91.
93.5 91.
74.5 73.
73.3 75.
73.1 74.
73.5 74.
71.6 67.
70.5 70.
76.6 71.
Temperature =
92.2 89.
95.5 90.
93.5 90.
74.5 74.
73.3 76.
73.1 74.
73.5 74.
71.6 73.
70.5 72.
76.6 74.
Temperature =
92.2 77.
95.5 77.
93.5 70.
74.5 42.
73.3 45.
73.1 46.
73.5 44.
71.6 81.
70.5 84.
76.6 83.
Time
5
4C
1
2
4
0
4
4
6
6
9
7
22C
8
6
6
7
0
3
4
0
5
4
45C
6
1
9
3
9
9
0
1
2
9
of Incubation (days)
1.

90.
92.
91.
75.
75.
76.
76.
70.
71.
70.

90.
91.
91.
74.
75.
74.
73.
70.
73.
73.

36.
41.
41.
32.
23.
22.
21.
39.
47.
48.
0

4
5
8
7
8
3
5
6
7
0

9
5
4
7
5
4
0
5
8
5

4
0
5
3
6
8
6
0
8
3
2.

92.
91.
90.
73.
78.
74.
74.
74.
73.
84.

87.
86.
86.
71.
72.
71.
71.
66.
70.
72.

21.
21.
22.
14.
13.
10.
8.
1.
31.
28.
0

0
4
7
3
2
9
3
6
3
9

6
1
6
7
0
6
2
2
8
2

2
4
2
9
6
0
6
8
4
7
4.

87.
89.
91.
74.
73.
74.
74.
73.
74.
71.

89.
88.
90.
71.
70.
72.
70.
72.
73.
72.

22.
13.
21.
4.
3.
2.
2.
1.
16.
25.
0

2
6
7
4
3
3
3
2
3
7

5
8
9
3
8
2
1
8
2
5

0
9
1
9
8
3
3
8
3
5
9.

93.
92.
91.
75.
81.
76.
75.
73.
72.
71.

91.
89.
91.
74.
70.
69.
72.
65.
74.
73.

29.
33.
31.
1.
1.
1.
2.
24.
23.
24.
0

1
5
8
8
5
0
7
8
5
6

2
8
4
9
6
7
0
7
4
2

5
1
8
8
9
6
3
9
6
6
                                     B.3

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
                    APPENDIX C
CONCENTRATION OF BAPP DURING STORAGE EFFECTS STUDY
                       C.I

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
APPENDIX C


CONCENTRATION OF BAPP  (ug P/L) AND  SUSPENDED  SOLIDS  (mg/L)  IN
WATER SAMPLES COLLECTED FROM THE HURON RIVER  AS A FUNCTION  OF
TIME AND TEMPERATURE OF INCUBATION.
Initial
Suspended
Solids
Incubation
24
21
26
11
17
18
16
10
11
13
Incubation
24
21
26
11
17
18
16
10
11
13
Incubation
24
21
26
11
17
18
16
10
11
13
Elapsed Time
0.50
Temperature
36.62
42.90
36.62
50.37
51.30
44.00
43.53
39.94
48.37
43.12
Temperature
41.76
39.76
43.76
46.33
49.75
35.15
44.78
42.80
43.92
42.49
Temperature
25.63
20.49
28.20
56.74
44.78
33.90
35.46
18.14
20.53
21.01
1.00
= 4C
44.96
43.64
38.09
44.76
39.95
44.76
48.28
40.90
38.91
41.97
= 22C
39.55
43.93
41.30
33.22
32.74
44.28
41.87
27.75
33.71
38.61
= 45C
41.45
34.59
40.28
36.70
36.59
55.66
50.21
38.30
37.84
42.13
of Incubation (days)
2.00

43.74
45.14
35.80
43.66
43.20
38.44
52.27
38.23
37.92
32.93

34.26
35.24
33.70
35.06
37.83
26.92
32.76
25.46
30.75
24.05

38.86
36.63
35.24
34.29
38.44
43.20
49.65
32.15
39.79
30.44
4.00

42.21
46.67
34.26
33.06
44.07
41.13
49.04
36.67
35.89
39.03

25.20
22.41
27.57
22.83
28.26
19.57
37.87
17.19
13.73
24.73

31.19
26.73
30.91
28.57
37.09
33.99
40.04
24.89
40.44
38.24
9.00

21.62
35.04
33.40
33.13
30.17
26.60
38.31
25.58
30.31
30.15

32,50
26.69
30.12
17.89
19.48
24.42
21.78
24.80

16.45

55.76
43.83
42.64
37.65
31.12
44.18
53.20
39.83
24.40
49.90
                                     C.2

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
APPENDIX C


CONCENTRATION OF BAPP  (ug P/L) AND  SUSPENDED  SOLIDS  (mg/D  IN

WATER SAMPLES COLLECTED FROM THE MAUMEE RIVER AS A FUNCTION OF
TIME AND TEMPERATURE OF INCUBATION.
Initial
Suspended
Solids
Incubation
189
199
170
127
133
130
139
198
204
215
Incubation
189
199
170
127
133
130
139
198
204
215
Incubation
189
199
170
127
133
130
139
.198
204
215
Elapsed Time
0.50
Temperature
125.54
130.39
128.96
95.88
98.36
96.03
97.90
136.67
139.69
142.39
Temperature
137.38
134.39
128.96
89.04
85.16
91.06
97.58
141.12
138.42
128.71
Temperature
129.82
135.24
142.81
107.68
106.90
100.69
117.78
130.78
121.08
126.96
1.00
= 4C
122.78
127.31
113.73
100.52
97.64
89.31
102.77
121.35
129.46
130.23
= 22C
105.55
117.96
118.11
90.91
98.12
81.45
98.12
120.44
124.57
115.85
= 45C
164.39
180.89
130.96
125.68
152.12
140.58
134.49
137.11
137.11
136.96
of Incubation (days)
2.00

120.59
116.68
118.36
92.51
88.36
85.60
89.84
136.21
133.57
135.28

109.15
127.42
119.47
92.36
94.81
87.29
83.60
124.69
130.92
122.66

160.06
167.17
134.67
130.60
147.96
155.33
148.73
204.60
166.59
186.06
4.00

119.05
118.56
121.84
90.28
95.87
96.80
94.94
125.31
133.01
117.29

108.32
114.68
118.71
84.08
84.86
79.43
83.31
116.35
120.28
122.95

153.08
185.23
149.43
151.85
155.57
150.14
157.43
194.61
169.15
147.31
9.00

120.35
129.75
119.16
95.31
98.89
92.67
101.37
137.05
129.33
145.86

108.87
110.36
119.61
86.47
91.27
87.07
80.39
123.82
123.03
117.52

147.64
137.20

134.66
161.71
140.57
159.53
138.39
191.60
139.02
                                     C.3

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
                        APPENDIX D
CONCENTRATIONS OF EXTRACTABLE AND BIOASSAY DETERMINED BAPP
                              D.I

-------
1
1



1




1

1


1

1





1
1




1
1
1
1
1
1

APPENDIX D
RESULTS OF BAPP ANALYSES, ug P/g

SOURCE SAMPLE TOTAL P CCTCENT CCTFILT
MAUMEE R. 1 1128.6 225.9 222.1
1118.0 231.5 227.8
CUYAHOGA R. 2 1135.7 399.3 385.1
1131.2 397.4 385.1
1134.3
1107.2
SANDUSKY R. 3 977.6 294.5 288.7
998.2 297.5 289.7
RAISIN R. 4 1271.1 137.3 133.5
1271.1 138.2 136.3
1247.8
1255.9
MAUMEE R. 5 825.5 189.1 183.1
848.1 189.1 183.1
SANDUSKY R. 6 888.8 263.6 256.9
911.1 263.6 254.9

L. ERIE 7 2169.0 409.4 388.3
2279.6 420.8 404.6
2195.6
2195.6
L. ERIE 8 2470.4 271.7 257.0
2468.4 258.0 243.3
2470.2 213.7
2482.7 234.8







D.2




BAKER ARMSTRONG CCIW BIOASSAY
422.6 365.1 563.8 310.0
425.7 372.4 597.4 299.0
489.6 500.2 670.4 367.0
492.3 509.2 691.1 350.0
449.0

324.3 341.9 581.2 236.0
322.3 346.8 604.3 219.0
428.6 459.1 628.5
429.7 450.1 643.1


219.6 221.5 405.9 231.0
223.2 224.2 416.9 233.0
235.0
265.6 271.8 520.0 164.0
256.2 280.9 530.4 203.0

1201.5 1618.9 1928.3 1462.0
1192.5 1527.7 1931.1 1665.0
1592.2 1453.0
1498.9
1300.3 1794.6 2245.0 1955.0
1291.3 1785.9 2215.8 2140.0
2026.0










-------
1
1
1

1
•
1
1




1
1

1

1


1
1
1
1
1
1
1

APPENDIX D




RESULTS OF BAPP ANALYSES, ug P/g

SOURCE SAMPLE
OSWEGO R. 9

GENESEE R. 10
CATTARAUGUS 11
CREEK
MBS STANDARD 12
SEDIMENT



NUMBER OF
DETERMINATIONS
AVERAGE
STANDARD
DEVIATION
MINIMUM
MAXIMUM










TOTAL P CCTCENT
994.5 223.4
1019.9 224.3
605.7 92.2
609.3 90.3
429.3 29.7
418.5 30.7
930.0 37.0
922.7 40.8
1001.1
966.5
1071.5
1022.1
36 24
1277.0 214.8
605.8 119.9

418.5 29.7
2482.7 420.8










CCTFILT BAKER ARMSTRONG CCIW BIOASSAY
225.3 227.3 231.5 322.9 106.0
213.1 224.9 232.1 345.7 112.0
85.6 124.2 130.9 218.2 52.0
85.6 125.3 126.8 229.6 61.0
29.7 33.7 32.9 79.7 5.0
29.7 33.7 34.7 80.0 1.0
43.7 111.8 118.6 289.7
42.7 106.3 114.4 293.1
109.2 115.6
108.2 115.0

26 26 28 24 29
209.4 403.5 550.5 709.7 518.6
110.4 383.7 583.8 639.7 667.0

29.7 33.7 32.9 79.7 1.0
404.6 1300.3 1794.6 2245.0 2140.0








D.3

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
 I
                   APPENDIX E.O
STATISTICAL ANALYSES:  ANALYSIS OF VARIANCE TABLES
                       E.O

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
 I
                  APPENDIX E.I
DOUBLE SPLIT-PLOT ANOVA OF STORAGE EFFECTS ON SRP
                                              E.I.I

-------
1
1
1

1

1
•
1

1

1

1





1




1
•
1



1
•

1
APPENDIX E.I
DOUBLE SPLIT-PLOT ANOVA OF STORAGE EFFECTS ON SRP

DEPENDENT VARIABLE:
SOURCE
MODEL

ERROR
CORRECTED TOTAL
MODEL F =
R- SQUARE
0.991530


SOURCE

RIVER
SAMPLE (RIVER)
TEMP
TEMP*RIVER
SAMPLE*TEMP (RIVER)
TIME
RIVER*TIME
TEMP*TIME
TEMP*RIVER*TIME

TESTS OF HYPOTHESES
AS AN ERROR TERM:

SOURCE
RIVER

TESTS OF HYPOTHESES
AS AN ERROR TERM:
SOURCE


SAMPLE (RIVER)
TEMP
TEMP*RIVER

SRP
DF
83

215
298
303.24
C.V.
10.5964


DF

1
18
2
2
36
4
4
8
8

USING THE TYPE


DF
1

USING THE TYPE

DF


18
2
2


SUM OF SQUARES
340981.47384995

2912.77553400
343894.24938395

ROOT MSB
3.68073270


TYPE III SS

225232.51210329
8876.76468743
46754.02848090
33983.69568510
3629.14174980
4510.69087345
2758.68393520
11217.64289398
3172.55765340





MEAN SQUARE
4108.

13.

PR > F

34.


F VALUE

16625.03
36.40
1725.52
1254.22
7.44
83.24
50.91
103.50
29.27

21052831

54779318

= 0.0001
SRP MEAN
73555184


PR > F

0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001

III MS FOR SAMPLE(RIVER)


TYPE III SS
225232.51210329



F VALUE
456.72



PR > F
0.0001

III MS FOR SAMPLE*TEMP (RIVER)

TYPE III SS


8876.76468743
46754.02848090
33983.69568510

F VALUE


4.89
231.89
168.55

PR > F


0.0001
0.0001
0.0001
E.I.2

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
 I
                   APPENDIX E.2
DOUBLE SPLIT-PLOT ANOVA OF STORAGE EFFECTS ON BAPP
                                              E.2.1

-------
1
1
1

1



1

1

1

1





1
•

1




1
1

1

1
1
APPENDIX E.2
DOUBLE SPLIT-PLOT ANOVA OF STORAGE EFFECTS ON BAPP

DEPENDENT VARIABLE:
SOURCE
MODEL

ERROR
CORRECTED TOTAL
MODEL F =
R- SQUARE
0.977633


SOURCE

RIVER
SAMPLE (RIVER)
TEMP
TEMP*RIVER
SAMPLE*TEMP (RIVER)
TIME
RIVER*TIME
TEMP*TIME
TEMP*RIVER*TIME


TESTS OF HYPOTHESES
AS AN ERROR TERM

SOURCE
RIVER

TESTS OF HYPOTHESES
AS AN ERROR TERM
SOURCE
SAMPLE (RIVER)
TEMP
TEMP*RIVER

RNAOHP
DF
83

214
297
112.69
C.V.
10.5249


DF

1
18
2
2
36
4
4
8
8


USING THE


DF
1

USING THE

DF
18
2
2


SUM OF SQUARES
659716.79449932

15093.66124195
674810.45574128

ROOT MSE
8.39828121


TYPE III SS

565189.37586957
29045.08712331
25113.66489230
19824.03399125
5052.63030753
658.26984750
2177.31410798
10415.28314372
3599.76112477




MEAN SQUARE
7948.39511445

70.53112730

PR > F = 0.0001
RNAOHP MEAN
79.79439597


F VALUE PR > F

8013.33 0.0001
22.88 0.0001
178.03 0.0001
140.53 0.0001
1.99 0.0015
2.33 0.0568
7.72 0.0001
18.46 0.0001
6.38 0.0001


TYPE III MS FOR SAMPLE (RIVER)


TYPE III SS
565189.37586957



F VALUE PR > F
350.26 0.0001

TYPE III MS FOR SAMPLE*TEMP (RIVER)

TYPE III SS
29045.08712331
25113.66489230
19824.03399125

F VALUE PR > F
11.50 0.0001
89.47 0.0001
70.62 0.0001
E.2.2

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
 I
                     APPENDIX E.3
DOUBLE SPLIT-PLOT ANOVA OF STORAGE EFFECTS ON BAPP/TSS
                                              E.3.1

-------
1
1
1

1

1
•
1

1

1

1





1
•






1


APPENDIX E.3
DOUBLE SPLIT-PLOT ANOVA OF STORAGE EFFECTS ON BAPP/TSS

DEPENDENT VARIABLE:
SOURCE
MODEL

ERROR
CORRECTED TOTAL
MODEL F =
R- SQUARE
0.928683


SOURCE

RIVER
SAMPLE (RIVER)
TEMP
TEMP*RIVER
SAMPLE*TEMP (RIVER)
TIME
RIVER*TIME
TEMP*TIME
TEMP*RIVER*TIME

TESTS OF HYPOTHESES
AS AN ERROR TERM

SOURCE
RIVER

TESTS OF HYPOTHESES
AS AN ERROR TERM

BAPP/TSS
DF
83

214
297
33.57
C.V.
21.2050


DF

1
18
2
2
36
4
4
8
8

USING THE TYPE


DF
1

USING THE TYPE



SUM OF SQUARES
303.65548909

23.31876430
326.97425339

ROOT MSE
0.33010027


TYPE III SS

199.25856678
66.30799813
5.77614755
3.57659199
3.84485061
3.94091205
4.72058190
8.54298049
5.57195851

III MS FOR SAMPLE


TYPE III SS
199.25856678








PR >




F VALUE

1828.63
33.81
26.50
16.41
0.98
9.04
10.83
9.80
6.39

(RIVER)


F VALUE
54.09



MEAN SQUARE
3.65849987

0.10896619

F = 0.0001
NORM MEAN
1.55671292


PR > F

0.0001
0.0001
0.0001
0.0001
0.5071
0.0001
0.0001
0.0001
0.0001




PR > F
0.0001

III MS FOR SAMPLE*TEMP (RIVER)



SOURCE
DF
TYPE III SS
F VALUE
PR
SAMPLE (RIVER)
TEMP
TEMP*RIVER
18
2
2
66.30799813
5.77614755
3.57659199
34.49
27.04
16.74
0.0001
0.0001
0.0001
                                   E.3.2

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
 I
                  APPENDIX E.4
ANCOVA OF SOURCE VARIABLES AND SRP DURING STORAGE
                                              E.4.1

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
 I
APPENDIX E.4
ANCOVA OF SOURCE VARIABLES AND SRP DURING STORAGE
STORAGE PERIOD:
SOURCE
RIVER
TEMP
TEMP*RIVER
TP
TSS
PH
COND
FIELD
STORAGE PERIOD:
SOURCE
RIVER
TEMP
TEMP*RIVER
TP
TSS
PH
COND
FIELD
STORAGE PERIOD:
SOURCE
RIVER
TEMP
TEMP*RIVER
TP
TSS
PH
COND
FIELD
TIME = 0 TO TIME
DF
1
2
2
1
1
1
1
1
TIME = 0 TO TIME
DF
1
2
2
1
1
1
1
1
TIME = 0 TO TIME
DF
1
2
2
1
1
1
1
1
=0.5 DAYS
TYPE III SS
90.54002949
84.52749000
1551.86614333
61.30429782
4.79320779
43.79826328
112.32486374
343.16361263
= 1.0 DAYS
TYPE III SS
4.40797891
7574.06262333
5208.31260333
12.05587451
0.60905374
0.56131775
1.95032945
120.93768348
=2.0 DAYS
TYPE III SS
0.81960001
14501.62623000
10469.99858333
19.86049355
4.22025522
0.91340833
6.82838717
167.72751549

F VALUE
1.66
0.78
14.23
1.12
0.09
0.80
2.06
6.30

F VALUE
0.19
162.37
111.66
0.52
0.03
0.02
0.08
5.19

F VALUE
0.04
367.39
265.25
1.01
0.21
0.05
0.35
8.50

PR > F
0.2035
0.4661
0.0001
0.2941
0.7681
0.3744
0.1575
0.0155

PR > F
0.6657
0.0001
0.0001
0.4756
0.8723
0.8774
0.7737
0.0272

PR > F
0.8394
0.0001
0.0001
0.3207
0.6458
0.8306
0.5591
0.0053
                                              E.4.2

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
APPENDIX E.4


ANCOVA OF SOURCE VARIABLES AND SRP DURING STORAGE
STORAGE PERIOD:
SOURCE
RIVER
TEMP
TEMP*RIVER
TP
TSS
PH
COND
FIELD
STORAGE PERIOD:
SOURCE
RIVER
TEMP
TEMP*RIVER
TP
TSS
PH
COND
FIELD
TIME = 0 TO TIME =
DF
1
2
2
1
1
1
1
1
TIME = 0 TO TIME =
DF
1
2
2
1
1
1
I
1
4.0 DAYS
TYPE III SS
0.46361422
17898.25454333
11890.68314333
64.90952628
41.65454603
84.46284448
9.25050180
13.34825249
9.0 DAYS
TYPE III SS
82.59397964
17832.48047204
8065.69424048
49.65860240
6.87744634
9.77979789
0.58474204
11.21721784

F VALUE
0.03
666.13
442.54
4.83
3.10
6.29
0.69
0.99

F VALUE
3.59
388.06
175.52
2.16
0.30
0.43
0.03
0.49

PR > F
0.8534
0.0001
0.0001
0.0327
0.0845
0.0155
0.4107
0.3238

PR > F
0.0640
0.0001
0.0001
0.1481
0.5868
0.5172
0.8739
0.4881
                                              E.4.3

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
 I
 I
               APPENDIX E.5
FACTORIAL ANOVA OF CHEMICAL EXTRACTION DATA
                                               E.5.1

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
 I
APPENDIX E.5


FACTORIAL ANOVA OF CHEMICAL EXTRACTION DATA
DEPENDENT VARIABLE
SOURCE
MODEL
ERROR
CORRECTED TOTAL
MODEL F =
R- SQUARE
0.998431
SOURCE
PROC
SAMPLE
PROC* SAMPLE
: BAP
DF
69
85
154
784.15
C.V.
6.3011
DF
5
11
53

SUM OF SQUARES
41498802.79288702
65193.38577504
41563996.17866206

ROOT MSE
27.69442199
TYPE III SS
4355958.29279255
23266129.09088487
9385673.82885074

MEAN SQUARE
601431.92453459
766.98100912

PR > F = 0.0001
BAP MEAN
439.51529032
F VALUE PR > F
1135.87 0.0001
2757.70 0.0001
230.89 0.0001
                                               E.5.2

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
 I
 I
                       APPENDIX E.6



CORRELATIONS AMONG CHEMICAL EXTRACTION AND BIOASSAY RESULTS
                                               E.6.1

-------
1
1







1
•



1

1
•







1
•



1
•

1

1




APPENDIX E.6
CORRELATIONS



AMONG



CHEMICAL













EXTRACTION AND B IDAS SAY RESULTS


All Pairwise Comparisons, all
VARIABLE


TP
CCTCENT
CCTFILT
BAKER
ARMSTRNG
CCIW
BIOASSY
Correlation



TP 1.
0


CCTCENT 0 .
0

CCTFILT 0.
0

BAKER 0.
0

ARMSTRNG 0.
0

CCIW 0.
0

BIOASSY 0.
0

N


12
12
12
12
12
12
10
MEAN


1161.7333
214.7333
208.1250
428.1333
502.9917
710.5333
508.5100
Coefficients /
TP


00000
.0000
12

57817
.0489
12
56583
.0552
12
97713
.0001
12
97472
.0001
12
97808
.0001
12
97402
.0001
10
CCTCENT


0.57817
0.0489
12

1,00000
0.0000
12
0.99964
0.0001
12
0.64719
0.0229
12
0.59088
0.0431
12
0.60550
0.0369
12
0.49848
0.1425
10
STD DEV


598.59102
125.07213
119.79774
406.77749
566.63827
668.01307
692.59261

data.


SUM MINIMUM


13940.
2576.
2497.
5137.
6035.
8526.
5085.
Type I Error Rate
CCTFILT


0.56583 0.
0.0552 0
12

0.99964 0.
0.0001 0
12
1.00000 0.
0.0000 0
12
0.63261 1.
0.0273 0
12
0.57468 0.
0.0506 0
12
0.59012 0.
0.0434 0
12
0.47754 0.
0.1628 0
10


800 423
800 30
500 29
600 33
900 33
400 79
100 2
/ Number
BAKER ARMSTRNG


97713
.0001
12

64719
.0229
12
63261
.0273
12
00000
.0000
12
99365
.0001
12
99250
.0001
12
98115
.0001
10


0.97472
0.0001
12

0.59088
0.0431
12
0.57468
0.0506
12
0.99365
0.0001
12
1.00000
0.0000
12
0.99718
0.0001
12
0.99353
0.0001
10


.9000
.2000
.7000
.7000
.8000
.9000
.7000




MAXIMUM


2472
415
396
1295
1790
2230
2040


.9000
.1000
.5000
.8000
.2000
.4000
.3000
of Observations
CCIW


0.97808
0.0001
12

0.60550
0.0369
12
0.59012
0.0434
12
0.99250
0.0001
12
0.99718
0.0001
12
1.00000
0.0000
12
0.99058
0.0001
10
BIOASSY


0.
0


0.
0

0.
0

0.
0

0.
0

0.
0



97402
.0001
10

49848
.1425
10
47754
.1628
10
98115
.0001
10
99353
.0001
10
99058
.0001
10
1.00000
0

.0000
10
E.6.2

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
 I
                             APPENDIX F
COMPARISON OF FIELD AND LABORATORY FILTRATION UNITS FOR SRP ANALYSIS
                                               F.I

-------
I
I
APPENDIX F
COMPARISON OF FIELD AND LABORATORY FILTRATION UNITS FOR SRP ANALYSIS
I
I
I
I
I
I
I
I
I
I
I
 I
     Sample     Filtration Unit   Sample Absorbance   Concentration (ug  P/L)

I   Maumee #9    Field                0.471                74.2
     Maumee #9    Laboratory           0.471                74.2
                    Difference =  0.000                 0.0
Huron #9     Field                0.128                19.0
Huron #9     Laboratory           0.122                20.0
                    Difference =  0.006                  1.0
                                               F.2

-------
TECHNICAL REPORT DATA
'Pkssr rrcJ Jitiirin tii":< tj> thcrc-ir*,, ,'u /, >rt ro"I/~ Yr;'ij"
i pEcoRT NC
2
4 7.T..E A\2 S. S- 7 _E
Variation Among and Within Procedures for Estimation
of Bioavailable Phosphorus
7 AUTl-iORiS)
Thomas C. Young, Joseph V.
DePinto, Bryan J. Hughes
9 PERFORMING ORGANIZATION NAME AMD ADDRESS
Department of Civil and Environmental Engineering
Clarkson University
Potsdam, NY 13676
12. SPONSORING AGENCY NAME AND ADC
Great Lakes National Progi
^.United States Environments
230 South Dearborn Street
Chicago, IL 60604
RESS
am Office/Region 5
il Protection Agency
3 RECIPIENT'S ACCESSiON-NO
b REPORT DATE
March, 1987
6 PERFORMING ORGANIZATION CODE
8 PERFORMING ORGANIZATION REPORT N
10 PROGRAM ELEMENT NO.
11 CONTRACT/GRANT NO.
R005761-01
13. TYPE OF REPORT AND PERIOD COVEREC
Final Report 10/83-12/84
14 SPONSORING AGENCY CODE
 15. SUPPLEMENTARY NOTES Presented at 49th Ann.  Conf.  Am.  Soc.  Limnol.  Oceanog.; at Great
 Lakes '86, Ann. meeting of Int. Assoc. Great  Lakes Res.;  and EPA Symposium on
 Characterization of Sludges, Sediments...  (May  1986;  Cincinnati,  OH) and to be
 16. ABSTRACT puDiisned -in symposium proceedings  oy  Aoin.
      Sediments were  assayed for biologically available particulate phosphorus  (BAPP)
 by bioassay  and  chemical  extractions to permit comparisons among the procedures.
 Although  the procedures extracted  widely differing amounts of phosphorus  (P),  accurate
 estimate  conversions,  using regression relationships,  could be made among procedures
 for freshly  collected  samples.   Samples stored for several years gave anomalous
 extractable  P results  by  some procedures.   In fresh samples amounts of P extracted by
 the De Pinto procedure were consistently closest in magnitude to the amounts taken up
 by algae  during  bioassays.   In both fresh and stored samples the Armstrong procedure
 gave results that  correlated most  closely with bioassays.
      Effects of  storage time (0 to 9 days) and temperature (4, 22, and 45 C) on
 soluble reactive P (SRP)  and BAPP  were examined using unfiltered samples from  the
 Maumee and Huron Rivers.   Significant changes occurred for most combinations of time
 and temperature.   Major changes occurred at 45 C; any period of sample storage could
 affect the reliability of SRP estimates in samples held at 45 C.  The results  support
 standard  procedures  for handling  of water samples collected for P analysis.  Observed
 changes in SRP during  storage,  however, were offset partially by inverse changes in
 BAPP, indicating that  for storage  times of no more than 9 days, the total bioavailable
 P  (BAP) of water samples  (SRP+BAPP) may be partially conserved.
17
                               KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                             b IDENTIFIERS'OPEN ENDED TERMS
  COS AT i Field'Croup
 Phosphorus, sediments, river,  nutrients,
 fluvial, suspended solids,  bioavailability,
 soluble, particulate
IB DISTRIBUTION STATEMENT
                                              19 SECURITY CLASS /This Report,
21 NO OF PAGES
  84 pages
                                             20 SECURITY CLASS iThispag-.-,
                                                                        22 PRICE
EPA Form 2220-1 (9-73)

-------
                                                         INSTRUCTIONS

    1    REPORT NUMBER
        ln^cT' thi I f'A rtp^r' r»"rr>c: d^ it appear^ or, the cc^cr of [V. pi.Mi.^ti'T

    2    LEAVE BLANK

    3    RECIPIENTS ACCESSION NUMBER
        Reserved fo;  usi b>  ea^ report recipient

    4    TITLE AND  SUBTITLE
        Titlt should ndi-a'i. Je.ir;\ snd brirfh the subjert coverage of the report and be di<.pL> c; p:o~ir
        Up-: ot otherwise ^'  ear  Indicate the bjs,, or, which n wa; selected 'i g  dcit o> iss^t Juu o"
        approve',  dale ofpriparctior;. etc /

    <&.   PERFORMING ORGANIZATION CODE
        Leave blank

    7.   AUTHORIS)
        Give name(s) in conventional order /John K. Doe, J Robert Doc. etc I  List author's affiliation if it differ1, from the performing organi-
        zation.

   V.~ PERFORMING ORGANIZATION REPORT NUMBER
        Insert if performing organization wishes to assign trm number

    9.   PERFORMING ORGANIZATION NAME AND ADDRESS
        Give name, street, cit>. state, and ZIP code. List no more than t*o lesels of an organizational hirearchy.

    10.  PROGRAM ELEMENT NUMBER
        Use the program element number under which the report was prepared  Subordinate number; mav be included in parenthe>es

    11.  CONTRACT/GRANT NUMBER
        Insert contract or grant number under which report was  prepared.

    12.  SPONSORING AGENCY NAME AND ADDRESS
        Include ZIP code

    13.  TYPE OF REPORT AND PERIOD COVERED
        Indicate interim final, etc., and if applicable, dates covered.

    14.  SPONSORING AGENCY CODE
        Leave blank.

    15.  SUPPLEMENTARY NOTES
        Enter information not included elsewhere but useful, such as  Prepared in cooperation with. Translation of Presented at conference of.
        To be pubbshed in. Supersedes. Supplements, etc.                                                    ~

    16.  ABSTRACT
        Include a  brief '200 words Of less> factual summary of the most significant information contained in the report-  If the report contain* a
        significant bibliography or literature  survey. mention it here.

    17.  KEY WORDS AND DOCUMENT ANALYSIS
        (ai DESCRIPTORS - Select from the Thesaurus of Engineering and Scientific Terms the proper authorized term* that identify  the major
        concept of the research and are sufficient!) specific and precise to be used as  index entries for cataloging

        (b> IDLNTIFIERS AND OPEN-ENDED TERMS - L'se identifiers for project name*, code names equipment designators etc.  Use open-
        ended terms written in descriptor form for those subjects for which no descriptor exists

        (c) COS^iTI FIELD GROUP - Field and group assignments are to be taken from the 1965 COSATI Sub.iect Category List  Since  the rra-
        icnty of documents arc mvi'irfKrtrMir.vy in r.iture the Primary F.eld Group aiMg.-i ic,u'>r vuli be specific discipline, area of human
        erJea'. :r  e;i>p; of pr>sicai object   The applications)  will be cross-referenced  with secondary Fieid Group assignments that vul! ]olio\\
        the  primary postmg(s)

    18.  DISTRIBUTION STATEMENT
        Denote releasabilju to the  public or limitation for reasons other than security for example  "Release Unlimited." Cite an\ a\ailobiliti to
        the  public, with address  and price  ,

    19. & 20. SECURITY CLASSIFICATION
        DO NOT  submit classified reports tc the National Technical Information service.

    21.  NUMBER OF PAGES
        Insert the total number of pages, including this one and unnumbered pages, but exclude distribution bst. if an>.

    22.  PRICE
        Insert the price set b> the National Technical Information Service or tht Government Printing Office, if known
EPA Form 2220-1 (9-73) (Reverie)

-------