J

magi) WIN WHWHHTIW

N
312910087 9140

 

 

 

 

 

 

 

1 " W
manner
Michigan 3mm
1 University
\ f
This is to certify that the
dissertation entitled

PHOSPHATE EXCHANGE BETWEEN
LITTORAL SEDIMENT AND LAKE WATER

presented by

Robert Paul Glandon

has been accepted towards fulfillment
of the requirements for

_Bh_._D_.__degree in Eisherieis. Wildlife

0%722/170010

*ajor pk ‘ofessor

Date October 20, 1982

 

MS U is an Affirmative Action/Equal Opportunity Institution 042771

 

MSU

 

 

 

RETURNING MATERIALS:
Place in book drop to

 

 

ALW~

 

 

LJBRARJES remove this checkout from
“ your record. FINES Win
be charged if book is
returned after the date
stamped below.
HS 2 6 199?

 

 

PHOSPHATE EXCHANGE BETWEEN
LITTORAL SEDIMENT AND LAKE WATER

BY

ROBERT PAUL GLANDON

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Fisheries and Wildlife

1982

ABSTRACT

PHOSPHATE EXCHANGE BETWEEN
LITTORAL SEDIMENT AND LAKE WATER

BY
Robert Paul Glandon

Soluble phosphate is an important form of phosphorus in
the environment. Current models, which consider a largely
hydrostatic sediment-water interface, hold that aerobic sedi-
ments act as a barrier to sediment phosphate flux to over-
lying water. However, differences between littoral sediment
interstitial and lake water phosphate levels indicate poten-
tial for diffusive flux. Further, water movement that re-
sults in perfusion of epilimnetic water or resuspension of
littoral sediment particulates, can modify the particle-
solution phosphate system. Information concerning mechanisms
of particle response to changes in ambient phosphate or vol-
ume of ambient water can enhance models of phosphate exchange
under hydrodynamic conditions.

Interstitial phosphate of sediments underlying 0.75-
1.00 m water in a shallow lake was variable in space. Lev-
els were generally greater than those of overlying water

and gradients could be used to estimate static-water phos-

phate eff:
solution c
water, mes
proached I
Particulai
rected by
of associz

Sedir
sampling 5
of particx
ponse of 1
levels imI
phosphate
ambient p1
phate sor}
condition;
described
sorption 1
concepts «
ulate res,
depends 0

water, an

ting from

Robert Paul Glandon
1x through surficial sediment layers. Assuming
ilcium activity and pH similar to that of overlying
1 soluble phosphate of surficial sediments ap-
cedicted levels in equilibrium with hydroxyapatite.
a response to ambient perturbation could be di-
phosphate mineral equilibria and limited by kinetics
ted phosphate exchange.

ent material was taken from the site of interstitial
ad a portion was treated to modify characteristics
late surfaces. Experimental observations of res-
articulates to alterations in lake water phosphate
licated controlling mechanisms other than calcium
solubility. Langmuir-type analysis suggested that
osphate level is governed by the reservoir of phos-
ed on sediment particles. Relative to experimental
, particle-solution phosphate relationships were
using quantified parameters related to phosphate
aximum and binding energy. A model, incorporating
rawn from experimental data, indicates that partic-
onse to perturbation of a sediment-phosphate system
phosphate activity of interstitial and infusing

particle weight to solution volume ratio resul-

iisturbance.

DEDICATION

To my son, Kevin

(5
T3 I

ACKNOWLEDGEMENTS

I would like to express my sincere appreciation to Dr.
Clarence D. McNabb whose open-minded approach has been inspiring.

Thanks to Drs. Boyd G. Ellis and Bernard D. Knezek for
their direct and indirect contributions to this effort.

Many of my fellow students deserve thanks, particularly
Ted R. Batterson, Frederick C. Payne, and Mehdi Siami. Each has
contributed to my graduate experience.

My wife, Nancy, deserves thanks for everything.

My parents deserve special recognition for their con-
tinous encouragement.

This study was supported in part by funds from the U.S.
Environmental Protection Agency, and the Michigan Agricultural

Experimentation Station at Michigan State University.

(-3
V5.

«4.

TABLE OF CONTENTS

Page
LIST or TABLES ... ....... . ..... ........................ v
LIST OF FIGURES ....................................... vii
INTRODUCTION .......................................... 1
MATERIALS AND METHODS ................................. 7
RESULTS ............................................... 16'
DISCUSSION ............ ..... ........................... 43
LITERATURE CITED ...................................... 69
APPENDIX 1 ............................................ 74
APPENDIX 2 ............................................ 79
APPENDIX 3 ............................ ..... ........... 81

APPENDIX 4 . ..... ............... ....................... 84

iv

Table

1.

LIST OF TABLES

Page

Calculated and measured net phosphate exchange

(uM P04 g'l) between whole sediments and in-

fusing lake water of different initial phosphate
activity (uM P04 1'1). Values assume equilibrium
conditions with ionic strength 3 x 10’ M and

pH 7.60 . Negative Sign indicates net sediment
sorption of phosphate ........................... 63

Calculated and measured net phosphate exchange

(UM P04 1’1) between fine sediments and infusing
lake water of different initial phosphate act-
ivity (uM P04 171). Calculated exchange is

based on experimentally estimated ambient cross-
over activity of 18.63 UM P04 1‘1 (A), and on a
hypothetical crossover activity of 22 EM P04

1' (B). Values assume equilibrium conditions

with ionic strength 3 x 10'3 M and pH 7.75 ...... 64

Effect of measured additions of arsenate (added
as Na2HASO4-7H20) on determinations of phosphate
in standard (KH2P04) preparations ............... 75

Effects of addition of arsenate reductant on
determinations of phosphate in distilled water
solutions containing measured additions of phos-
phate and arsenate .............................. 78

Concentration of soluble reactive and total phos-
phate, iron, and calcium in shallow sediment
collections and lake water used in processing

fine sediment. Corresponding values represent
estimates of the amount of each component de-
posited during drying (at 800C) of 96 g of sedi-
ment material ................................... 80

Filterable iron (uM l—l) in lake water of differ-
ent initial phosphate levels following 0.5, 2.0,
and 48 hours exposure to whole sediments. Lake
water, prior to sediment addition, contained

0 734 uM Fe 1'1 ............................ . ..... 82

Filterable iron (UM l-l) in lake water of differ—
ent initial phosphate levels following 0.5, 2.0,

Table

Page

and 48 hours exposure to fine sediments. Lake
water prior to sediment addition contained
0.84211MFe1-1 O.......OOOOOOOOOOOOOOO0.0.000... 83

Solubility products of calcium phosphate com-
pounds and dissociation constants of phosphoric
acid species and water at 25°C used in the
development of the solubility diagram of Figure

10 ......OOOOOOOOOIOOOO......OOOOOOOOOOOOOO....O. 85

vi

Figure

1.

LIST OF FIGURES

Page

Content of total interstitial phosphorus (mean
i 1 standard error) with depth in shallow sedi-
ments of the south basin of Lake Lansing in the
interval from July - October, 1978 ................ 18

Net reactive phosphate exchange between sedi-

ment and lake water following 2 hours (A-A) and

48 hours (x-X) exposure to different initial

reactive phosphate levels. Initial ambient

levels (uM P0 1‘1) are indicated in parentheses.
Horizontal aXIs shows final ambient reactive
phosphate level. Panel A results from.whole
sediments. Panel B results from fine sediments .. 22

Net Reactive phosphate sorbed by whole sediments
(solid curve) and fine sediments (dashed curve)

from lake water to different initial reactive
phosphate levels, following 2 hours (A-A) and 48
hours (X-X) exposure. Horizontal axis indicates
resultant ambient reactive phosphate level.
Associated initial levels of ambient reactive
phosphate (uM P04 1’1) are indicated in paren-

theses ........................................... 27

Relationship between release of soluble nonre-
active phosphate and release of reactive phosphate

by sediments following 0.5 hours (0-0), 2 hours

(A-A), and 48 hours (x-x) exposure to lake water

with low initial reactive phosphate levels. Solid
line, results from whole sediments; dashed line

from fine sediments .............................. 30

Relationship between release of soluble nonreact-

ive phosphate and sorption of reactive phosphate

by sediments following 0.5 hours (0-0), 2 hours

(A-A), 48 hours (X-X) exposure to lake water

with high initial reactive phosphate levels.

Solid line, result from whole sediments; dash line
from fine sediments ........... ..... .............. 32

Ambient calcium (mean and 1 standard deviation)
and pH (mean and range) with duration of whole
(A-A) and fine (0-0) sediment exposure to lake
water phosphate solution. The dashed lines

vii

Figure Page

10.

11.

indicate hypothesized course of change in ambient
calcium within initial 30 minutes sediment-lake

water exposure. Time 0 indicates point of sediment
introduction to lake water......... ...... ........... 35

Net reactive phosphate released from fine sedi-
ments under conditions of ambient phosphate
accumulation.(A—A), and cumulative phosphate
release from fine sediments to ambient lake
water replaced at 15 minute intervals (0-0)._l
Lake water initially contained 0.10 uM P04 1 .
Values are expressed as a mean of six replicates

t 1 standard error ................................ 40

Relation between pH and relative proportions of
soluble reactive phosphate species. Calcula-

tions based on dissociation constants of phos-

phoric acid, mono-, and dihydrogen phosphate

ions in dilute solution ........................... 46

Influence of ionic strength (u) on activity
coefficient (7) of mono— and dihydrogen phos-

phate for solutions of ionic strength less

than 0.01 molar. V was calculated from: log y

= -0.5085 22 /fi_ ................................. 49

Solubility diagram for certain calcium phosphate
compounds in sediments at 25°C. Levels of solu-
ble reactive phosphate were calculated by consid-
ering appropriate solubility equilibria and _3
assuming calcium levels maintained at l x 10

molar (solid lines) and 3 x 10‘3 molar (dashed
lines. (-) refer to ambient reactive phosphate
levels following 48 hours exposure of whole
sediments to high initial levels of ambient
phosphate. (X) same, but for fine sediment ex-
posure. Phosphate activities were converted to
concentrations for comparison with observed levels
of total interstitial phosphate of south basin
sediments ..... ......... ........................... 52

Calculated relationship between phosphate sorbed

by whole and fine sediments and ambient phos-

phate activity. Each isotherm was developed

from estimations of Langmuir constants B and K.

Ionic strength 3 x 10‘3 M and equilibrium with

respect to particulate-solution phosphate ex-

change were assumed. The pH was taken as 7.60

and 7.75 for whole and fine sediments, respec-

tively .................................. . ......... 60

Figure

12.

Net exchange of particle associated phosphate

as a function of phosphate activity of infusing
medium and of interstitial water in an ideal-
ized sorption-desorption exchange system.
Letters indicate phosphate activity level
(A<B<C) and superscripts indicate infusing
media. The dashed curve indicates sediment
response to infusing phosphate activity level
(3’) if sediment : solution is reduced as result
of perturbation ................................

Percent over-estimation of phosphate concen-

tration as a function of the ratio of phosphate
to arsenate in solution ........................

ix

Page

67

77

INTRODUCTION

The general relationship, developed from a large number
of lakes, between annual loading of total phosphorus from
external sources and mean annual chlorophyll-a content of the
open water, highlights the singular importance of phosphorus
to primary production in lacustrine systems (Vollenweider,
1980). However, the empirical nature of the relationship
limits identification of mechanisms underlying variability
between lakes with similar hydrologic characteristics. In
some of these, fluctuations in autotrophic production are
observed in the absence of temporally corresponding changes
in allochthonous phosphorus input. Detailed lake studies,
based on comparisons of total phosphorus import and export
over time, generally indicate that exchange between sediments
and overlying water is an important component of phosphorus
dynamics. While phosphorus budget analyses suggest that an-
nual import and export differences result from net sedimenta-
tion, examination of water column phosphorus over time
frequently shows periods of net sediment phosphorus release
(Hutchinson, 1957). The extent and ecological impact of phos—
phorus regeneration from sediments is highly influenced by
morphometric and associated hydrodynamic features of the lake.

The influential early work of Mortimer (1941, 1942),
subsequently verified by many investigators, signaled the

l

importance of oxidation state of surficial sediments in in-
fluencing the direction of net flux of sediment phosphorus.
Chemical reduction of inorganic phosphate binding metal oxides
and hydroxides, particularly those of iron, leading to liber-
ation of soluble phosphate to overlying anoxic water represents
a potentially significant mechanism of phosphate release,
particularly from deep-water sediments. However, any realized
impact of that on biological species dependent on water column
phosphate levels is contingent upon the relative volume of
anoxic water, area of underlying sediment surface, and fre-
quency and extent of circulation to zones of maximum phosphate
utilization. By contrast, a majority of lakes in the north
temperate region are Shallow, and characterized by a relatively
large volume of oxygenated water overlying extensive littoral
sediments (Wetzel, 1975). In these systems, the significance
of sediment phosphate release depends to a larger degree on
extent of diffusion-driven transport across an aerobic sedi-
ment-water interface, and on direct phosphate release from

the surface of particles brought into the water column by
turbulent mixing.

The potential of sediment particulates to influence
soluble phosphate following transport to media of low ambient
levels has been evidenced in phosphorus uptake by phytoplankton
cultured axenically in media initially containing only sedi-
ment phosphate (Golterman, et al. 1969; Williams, et a1. 1980).
Further, sediments introduced to water containing relatively

high levels of soluble phosphate may cause ambient phosphate

reduction (Carritt and Goodgal, 1954; Taylor and Kunishi,
1970). Particle-solution phosphate exchange in sediments is
largely controlled by chemical equilibria (Hayes, 1954:
Pomeroy, 1965; Li, et a1. 1972).

Distribution of soluble phosphate in sediment-solution
systems has been considered with respect to solubility equi-
libria of certain aluminum, iron, and calcium phosphate
minerals (Stumm and Leckie, 1971; Stumm and Morgan, 1981).
Solubility product calculations suggest that excess accumula-
tion or depletion of soluble phosphate is buffered by pre-
cipitation or dissolution of phosphate minerals, principally
variscite (AlPO4), strengite (FeP04), dicalcium phosphate
(CaHP04-2H20), octacalcium phosphate (CaH(P04)3), or
hydroxyapatite (Calo(P04)6(OH)2). However, solubility-pH
relationships suggest that variscite and strengite are unstable
at pH levels greater than 6, and Bache (1963) suggests that
these minerals are unlikely to persist in solid form at pH
levels greater than 3. Syers, et al. (1973) conclude that
while it is highly unlikely that discrete phase iron and
aluminum phosphates exist in sediments, hydroxyapatite is
commonly found. In lakes that experience decalcification of
the water column the calcium phosphate system is believed to
be one of the principle control mechanisms for the distribution
of phosphate to overlying water (Stumm and Leckie, 1971).

Williams, et a1. (1970); Ryden, et a1. (1973) and others
suggest that the balance between soluble and particulate

phosphate in sediments is predominantly influenced by the

4

extent of association of the phosphate anion with preexisting
solids, rather than through formation and dissolution of in-
soluble precipitates. Support for the dominance of sorp-
tion/desorption reaction with sediment solids stems in part
from differences between observed rate and extent of phosphate
exchange (Hayes, et a1. 1952; Li, et al. 1972) and that ex-
pected from relevent precipitation reactions (Lindsay and
Moreno, 1960; Stumm and Leckie, 1971).

Proposed mechanisms of phosphate sorption involve the
nature of the solid surface. Chen, et a1. (1973), working
with well characterized clay minerals, proposed that a pH
dependent electrostatic attraction between surface aluminate
groups and solution phosphate anions could account for their
sorptive ability. Edzwald (1977) presents evidence that metals
associated with clays, probably as hydrous metal oxide coat-
ings, have an overriding influence on phosphate sorption.

This is supported by observed correlations between content of
amorphous iron, aluminum, and silica compounds and phosphate
sorption capacity of bottom sediments of the Maumee River
(McCallister and Logan, 1978). Strong correlations between
surficial sediment phosphorus levels and extractable amorphous
iron oxides indicate that iron plays a particularly important
role in this relationship (Shukla, et a1. 1971; Williams, et
a1. 1971).

Soluble phosphorus of interstitial water is the sediment
phosphorus fraction with the highest chemical mobility (Syers,

et al. 1973). In shallow, surficial sediments, it appears

5

that phosphate precipitation reactions with calcium or sorp-
tion reactions, principally involving amorphous metal oxides,
can have a significant influence on interstitial phosphorus
levels. These chemical reactions can also determine partic-
ulate response to changes in those levels. These considera-
tions may be contrasted with those that characterize aerobic
sediments as a barrier preventing phosphorus release to over-
lying water (Ruttner, 1963; Wetzel, 1975). The latter imply
uni-directional association of phosphate with sediment partic-
ulates and ignore the operational significance of the reciprocal
nature of the particle-solution phosphate relationship.
Phosphate exchange reactions may result in interstitial phos-
phorus levels that are greater than those of overlying water:
indicating gradients along which diffusive flux may occur.
Further, resuspension of littoral sediment particulates by
wind-induced currents has been commonly observed (Davis and
Brubaker, 1973). This has the potential to alter the particle-
solution relationship and result in particle-induced changes

in water column phosphate levels.

The aim of this work is to evaluate the influence of
Shallow sediment particulates on soluble phosphorus levels.
More specifically, particle influenced levels of soluble phos-
phorus were examined through measurements of interstitial
phosphorus in a hydrodynamically active littoral zone. Phos-
phate exchange characteristics and capacity of particulates
to modify changes in ambient phosphate levels were examined

in experimental sediment particle-lake water systems. Results

6
were evaluted with respect to the role of precipitation
reactions, through considerations of calcium phosphate equi-

libria; and sorption reactions, through Langmuir analysis.

MATERIALS AND METHODS

Lake Lansing, the site of this study, exemplifies a
system in which the functional role of shallow sediments in
phosphorus dynamics can be of major significance. Approxi-
mately 80% of the lake volume is contained above the three
meter contour and is well mixed by wind-driven circulation
throughout the summer months. The oxygen content of the
circulating stratum remains near equilibrium saturation with
the atmosphere. During the mid to late summer period, littoral
and epilimnetic water attains saturation with respect to cal-
cium carbonate. Seasonal dynamics of this compound in Lake
Lansing are marked by evidence of a mid-summer decline in
soluble calcium levels.

Interstitial phosphorus profiles of shallow sediments
of the south basin of Lake Lansing were measured in_§itu
periodically during the summer and early fall of 1978. The
sampling device consisted of a compartmented dialysis sampler
fitted with a biologically inert polycarbonate membrane (Nu-
cleopore, 0.2 um pore size). Each sampler was machined from
plexiglass into a tapered block encasing twenty rows of ports,
with four ports per row. Each port was 2.5 cm in diameter
and 13 ml in volume. To prepare samplers for placement in
the lake, acid washed ports were filled with ion-free water

and covered with a sheet of membrane that was held in place

7

8

by a plexiglass cover fastened to the sampler with nylon
screws. Holes in the cover matched open surfaces of the ports.
Time required for equilibration of soluble phosphorus
across the membrane was determined in the laboratory by plac-
ing a prepared sampler in a bath of phosphate solution con-

4 1'1. Daily measurements of port-water

taining 9.7 uM P0
phosphate were compared to those of bath water. After nine
days, port phosphate levels equaled those of the bath and that
period was taken as the minimum duration of sampler exposure
in the lake.

Samplers were placed in the aqueous sediments of the
east and west littoral zones of the south basin at a water
depth of 0.75 - 1.00 m. Sampler position relative to the
sediment-water interface was maintained by securing samplers
to a cross bar connected to aluminum conduit that was driven
deep into the sediment. Clamps connecting the cross bar to
the conduit were tightened at the desired sampler position.
Samplers were left in place for ten to fourteen days.

At the end of the period, samplers were returned to the
laboratory. Particulate material clinging to or near the mem-
brane surface was removed to prevent sample contamination.
Measured volumes of port water were pipetted from sampler
ports and colorimetrically analyzed for total phosphorus fol-
lowing persulfate digestion (A.P.H.A., 1976).

Complemetary information concerning the response of
Sediment particulates to changes in ambient phosphate was

gathered through laboratory experiments. Interstitial

phosphorus levels may be viewed as an approximate inte-
gration of particulate response to physico-chemical changes
experienced throughout the period of measurement. These
provided a means of estimating diffusive phosphorus flux
along interstitial gradients. Estimation of sediment re-
sponse to a given alteration in ambient phosphate level
required direct examination.

Sediment material used in the laboratory was extracted
from south basin with an Eckman dredge at a water depth of
0.5 - 0.75 m. Sediments were taken from the west littoral
zone near the point of interstitial water sampling. Small
amounts of litter of terrestrial origin, and fragments of
viable aquatic macrophytes in samples were discarded. Excess
water was removed from representative subsamples of material
by concentrating sediment particulates on 0.5 um glass-fiber
filters with the assistance of a mechanically generated vac-
uum of 12 psi. Filter-retained sediment, designated as whole
sediment in this study, was refrigerated in a wet state at
4°C. Experiments were conducted within thirty-six hours of
collection.

Laboratory work consisted of following particle induced
changes in total and reactive filterable phosphate in lake
water as a function of initial level of dissolved reactive
phosphorus in that water. Five grams of wet sediment material
was placed into each of eighteen, 300 m1 Erlenmeyer flasks
containing 100 m1 of filtered (0.5 um) lake water. Dry weight

of added sediment was 15.2% of total sediment weight. Reaction

10

flasks differed with respect to initial levels of soluble
reactive phosphate. Particulate response to six initial
phosphate levels, ranging from 0.1 to 303 uM P04 1-1, was
examined. Each reactive phosphate level above that of lake
water was obtained by adding the appropriate quantity of
phosphate salt (KH2P04) to a measured volume of lake water.
Aliquots were withdrawn from each batch solution for direct
analysis of total dissolved and reactive phosphorus prior to
sediment addition. Each sediment/solution phosphorus combi-
nation was prepared in triplicate so that the extent of
solution phosphorus change could be measured following each
of three pre-selected exposure intervals. The intention was
to identify the steady state or "quasi-equilibrium" condition
through observed decrease in the rate of change in experimental
systems.

Initital pH of each sediment/solution suspension was
recorded immediately following sediment introduction. A
calibrated Corning Model 6 pH meter and Orion Model 91-25
epoxy combination probe was used for these measurements.
Flasks were then covered with porous foam plugs to retard
evaporation and placed on an oscillating shaker table to maxi-
mize sediment-solution contact. Reaction vessels, represent-
ing each initial solution phosphate level, were removed from
the table after 0.5, 2.0, and 48 hour exposures. Final pH
was recorded and sediment particulates removed by filtration.
The 0.5 um filter (glass-fiber) was assumed to have separated

soluble from insoluble particulate components of sediment-

11

solution systems. Filtrate was analyzed for content of total
and reactive phosphate.

Dissolved reactive phosphate was determined colorimetri-
cally using the single reagent ascorbic acid method without
prior persulfate digestion of filtrate (A.P.H.A., 1976). Re-
sults of these measurements have been taken to indicate levels
of ionic phosphate in solution. The difference between total
phosphorus content of filtrate, determined following persul—
fate digestion, and reactive phosphorus content was taken as
a measure of the soluble nonreactive fraction. Soluble non-
reactive phosphorus may include dissolved condensed and organic
phosphates (A.P.H.A., 1976); in natural systems not heavily
influenced by wastewater, the soluble nonreactive fraction
is typically dominated by a dissolved organic phosphate com-
ponent (Clesceri and Lee, 1965; Sonzoni, et a1. 1981).

Lake Lansing sediments were previously determined to
contain high levels of total arsenic (Batterson, 1980). Since
any solubilized arsenate represents a potential positive in-
terference in phosphate measurements (Shapiro, 1973), all
samples and standards were subjected to an arsenate reductant
(Johnson, 1971). The reductant was shown to be effective in
analytical results presented here (Appendix 1).

Experimental conditions were intended to reflect part-
icle-solution phosphorus exchange in an aerobic environment.
It was assumed that chemical reduction of iron could influence
sediment response and that a functional change in redox po-

tential would be indicated by an increase in soluble iron

12

content. In order to monitor transformation of ferric to
soluble ferrous form, solution iron content was measured prior
to and following sediment exposure. In an effort to distin-
guish relative importance of pertinent precipitation and
sorption mechanisms, solution calcium levels were also mea-
sureda The intention was to incorporate calcium phosphate sol-
ubility expressions in the analysis of observed response of
experimental systems. Aliquots of filtrate from each sedi-
ment-solution suspension were acidified (4ml double distilled
nitric acid per liter; pH = 2) and stored for later analysis
of iron and calcium content and comparisons with lake water
content prior to sediment addition. Soluble calcium and iron
was determined by atomic abSorption spectrOphotemtry. Lan-
thanum chloride was added to calcium samples and standards to
eliminate phosphate interference. (U.S.E.P.A., 1979).

To estimate soluble phosphorus, iron, and calcium con-
tributions of the aqueous component of wet sediment, weighed
subsamples were dried in a forced-air oven at 100°C to constant
weight. Weight loss upon drying was coupled to measured le-
vels of total and reactive phosphorus. iron, and calcium in
the filtrate obtained during sediment preparation to estimate
mass of these materials introduced with the aqueous sediment
component. Water volume of whole sediments was considered in
determining total aqueous volume in experiments.

Reactions between particulate and solution components
are influenced by the chemical nature of the particulate sur-

face (Stumm and Morgan, 1981). For example, carbonates may

13

be a much less energetic phosphate retention agent than hy-
drated metal oxides (Shukla, et a1. 1971; McCallister and
Logan, 1978). Sediments used in the second phase of the
laboratory work here were processed to alter particle surface
characteristics. Specifically, processing was intended to
enhance the relative influence of carbonate precipitates on
particle-induced changes in solution phosphate. Observations
of particle-solution phosphate exchange were contrasted with
those of whole sediments.

For this work, small sized particulate fractions were
obtained by passing a total of five liters of sediment through
a series of three standard U.S.G.S. sieves. Sieves were
stacked in order of decreasing pore size; the smallest (227
um) was placed over an enamel collecting tray. A total of
fifteen liters of filtered lake water, of known phosphorus,
calcium, and iron content was introduced with sediment to
the top-most sieve. Added lake water sorted and carried fine
particles through the series. Water in the enamel collecting
tray was evaporated in a forced-air oven at 80°C. Remaining
particulates were removed from the tray and stored until use.
Estimated quantities of formerly soluble phosphorus, calcium,
and iron deposited during evaporation are presented here
(Appendix 2).

Certain effects of fine sediment processing were ex-
pected from considering potential changes in aqueous chemical
composition during evaporative concentration. The reaction

path of natural waters during isothermal evaporation depends

14

on the Ca2+ : HC03-, which is reflected in the alkalinity and

calcium content of water prior to evaporation (Stumm and Morgan,
1981). The alkalinity of lake water used in processing fine

3

sediment was 2.4 x 10- eq 1-1. Calcium content was 7.98 x

-4 1

10 moles l- . Predicted calcium carbonate precipitation

and increase in solution pH during evaporation of water, char-

2+ 1-1) < alkalinity (in equivalents

acterized by 2 (moles Ca
1-1), were expected to have important effects on sorptive
characteristics of exposed particles. The effect of increasing
pH is, in many cases, to decrease phosphate sorption mediated
by metal oxides and hydroxides associated with surfaces of
particulate materials (Edzwald, 1977). Further, dessication
may have a denaturing effect on metastable metal oxides, such
as highly hydrated ferric hydroxide (Ponnamperuma, 1967).

This may underly reduced phosphate sorption capability, ob-
served elsewhere, of sediment particulates following drying
(cf. Harter, 1968; also Fitzgerald, 1970). The treatment like-
ly changed particle surface characteristics through alterna-
tion of particle-associated ferric hydroxide and through de-
position of carbonate precipitates.

Exchange of filterable total and reactive phosphate be-
tween fine sediments and lake water was examined in a manner
Similar to that employed with whole sediment. One gram dry
material was added to 100 ml lake water phosphate solutions.

In a separate experiment, extent of net phosphate re-
lease from fine sediment particulates following timed exposure

to low initial levels of lake water phosphate was examined.

15

Here, 0.5 9 fine sediment was added to each flask containing
100 ml filtered lake water. The sediment-water suspension
was filtered after fifteen minutes exposure. Filter-retained
sediment was reintroduced to a second 100 ml aliquot of fil-
tered water by rinsing the filter in that volume. In total,
each initial 0.5 g of material was transferred six times.
Soluble reactive phosphate was measured following each expo-
sure. For comparison, 0.5 g sediment was added to each of
several flasks containing 100 ml lake water. A sample was
filtered following 15, 30, 45, 90, and 120 minute exposures
in order to follow the course and extent of phosphate re-
lease in the absence of medium renewal, ie. under conditions
of ambient phosphate accumulation. Each experiment was

replicated six times.

RESULTS

Levels of total soluble phosphorus in water equili-
brated with interstitial water in Lake Lansing sediments can
be characterized as highly variable (Figure 1). Variability
in the data prevented discriminating between profiles of the
east and west sampling locations in the south basin. Simi-
larly, temporal differences between phosphorus profiles at a
given location could not be discerned. As a result, data
were combined with respect to depth to obtain mean intersti-
tial phosphorus characterizing south basin sediments over the
midsummer to early fall period.01== 8 to 10 per depth).

Combined soluble reactive plus nonreactive phosphorus

1

ranged from 1.9 uM P04 1- near the sediment-water interface

to 4.5 uM P0 1"1 at a depth of 58 cm below the sediment sur-

4
face.' These data contrast with those reported elsewhere
(Holdren, et a1. 1977; Carignan, et al. 1981) in that phos—
phorus levels here were relatively low and profiles lacked

an apparent steep interstitial phosphorus gradient. Rapidly
increasing levels of soluble phosphorus with depth in the
sediment have been attributed to the presence of a redox bound-
ary, below which phosphate-binding iron compounds are reduced
(Mortimer, 1971). However, literature observations reflected

conditions of experimental sediment cores or sediment underly-

ing comparatively quiescent deeper water. Relatively high

16

17

Figure 1. Content of total interstitial phosphorus (mean

i 1 standard error) with depth in shallow sediments

of the south basin of Lake Lansing in the interval
from July - October, 1978.

Depth In Ssdtmsntdcm)

HOSSSHBKB

8

 

18

Surface

 

 

 

 

 

' ‘vfi
L X
7

 

0.00 0.40 080 I20 |.60 2.00 2.40 2.80 3.20 3.60 4.00 4.40 4.” 5.20 560 600 6.20

Total Interstitial Phosphorus ( uM I'-‘O4 I")

19

variability in interstitial phosphorus examined here may re-
flect effects of particulate resuspension or ambient medium
displacement by seepage or turbulent movement of overlying
water. This implies that interstitial phosphate gradients
and associated diffusive flux were variable within the sedi-
ment.

The expected flux of phosphorus from sediment to over-

lying water due to molecular diffusion can be calculated from:

--92
F — ¢( Ddz
where D is the phosphate diffusion coefficient, 10-6 cm2

sec.1 (Stumm and Leckie, 1971) and dC is the ambient reactive

phosphate concentration difference across a distance, dz. The
value of 4 for south basin sediments was near 0.96 (cf. Siami,
1981), and reflects the volume fraction of sediment occupied
by water (Stumm and Morgan, 1981).

Driving force of molecular diffusion is the concentration
gradient of soluble reactive phosphate within the region of
interest. Since data obtained were the sum of soluble react-
ive and organic phosphorus, the reactive fraction could not be
separately determined. If it is assumed that the level of sol-
uble organic phosphorus was relatively constant along the dif-
fusion path (cf. Holdren, et a1. 1977), the data may be used
to approximate the soluble reactive phosphorus gradient. Total
interstitial phosphorus at 2 cm ranged from 0.581 to 3.871 uM
PO4 1-1. For the region extending from the interface to a depth

of 2 cm, the diffusion path, dz, was 1.0 cm; measured to

the center of the 0-2 cm layer. Using constants given above,

20

the rate of phosphate diffusion from this depth to overlying

water (mean 0.10 uM PO 1-1) could have ranged from 0.40 to 3.13

uM PO4 m"2 day-1.

4

Estimated maximum phosphate flux of 3.13 uM PO4 m-2 day-1

could increase phosphate content of l m3 lake water by 0.003

1 day-l, or approximately 3% of mean epilimnetic

uM P04 1'
levels. By comparison, if turbulent movement of overlying wa-
ter resulted in displacement of interstitial solution (at 3.9
uM PO4 1-1) to a depth of one centimeter, epilimnetic phosphate
could be increased by 30% (when 0 = 0.96). Further, phosphate
release from surfaces of particulates swept from the sediment
surface can contribute to phosphate levels of overlying water.
Capacity of sediment solids to buffer interstitial phosphate

or respond to changes in ambient levels is an important char—
acteristic of particle-solution phosphate relationships that
can give perspective to diffusive flux considerations present-
ed here.

Results of experiments indicated a dynamic relationship
between particulate and ambient reactive phosphate (Figure 2).
The direction and extent of net movement of reactive phosphate
between sediments and ambient lake water was depedent upon
sediment type. On a dry weight basis, whole sediment particu-

1

lates released 0.3 uM P04 9. during 48 hour exposure to initial

ambient phosphate of 0.13 uM PO 1-1. The result was an ambient

l

4

level of 2.4 uM P04 1. Extent of net release was restricted

at higher initial levels of ambient phosphate, and direction

of particulate-solution phosphate exchange reversed following

Figure 2.

21

Net reactive phosphate exchange between sediment

and lake water following 2 hours (A- A) and 48 hours
(x-x) exposure to different initial reactive phosphate
levels. Initial ambient levels (uM P04 1’1 ) are
indicated in parentheses. Horizontal axis shows

final ambient reactive phosphate level. Panel A
results from whole sediments. Panel B results from
fine sediments.

Net Reactive Phosphate Exchanged (11M F’O4 q-I dry weight)

0.40
0.3 0
0.20
0. I 0
0.00
0. I0

0.20

0.30
0.40

0.60
0.40
0.20
0.00
0.2 0
0.40
0.60
0.8 O
I .00
I .2 0

22

 

 

 

 

 

 

 

' A
. 4 x (6.23)
I-
_ 4 (152)
2.00 3.00 4.00

. B (32.26).(
F
- x‘(0.0'n

u (2 (3 :4 I5 (5 (7 (e (9 20 2: 22 23 24 25 26 27 23

Final Ambient Reactive Phosphate (1.1M F’O4 2Q")

23

sediment exposure to 6.23 uM PO4 1-1. Differences between 2

and 48 hour results have been taken to reflect experimental
error imposed by uncontrolled variables.

In contrast to whole sediments, introduction of fine part-
iculate material to lake water resulted in extensive net phos-

1 was liberated to phos-

phate release. Nearly 1.2 uM P04 9-
phate-poor lake water, roughly four fold that of whole sediments.
By comparison, this tendency was only moderately restricted
at higher initial ambient phosphate levels. Net release was
reduced by 9% in response to a ten fold increase in initial

1

lake water phosphate to 1.58 uM P0 1- , one fourth the inhi-

4
bition of whole sediment response to a similar ambient increase.
This contrast between sediment types was reflected in the rate
of development of the relationships of Figure 2. Under lake
water conditions promoting sediment phosphate release, over

80% of the 48 hour fine sediment response occurred within 30
minutes; almost twice the short term progress of whole sedi-
ments. 0n the other hand, over 90% of the final whole sediment
phosphate uptake from lake water occurred within 30 minutes,
compared to 65% for fine sediments. ‘

The whole sediment-lake water phosphate relationship im-
plies that no net exchange would result from sediment exposure
to lake water bearing 3.2 uM P04 1.1. Limited by the experi-
mental conditions employed, this "crossover" concentration can
be taken as an estimate of the ambient reactive phosphate level

in equilibrium with whole sediment particulates prior to use

for experimentation (McCallister and Logan, 1978; Oloya and

24

Logan, 1980). The experimentally determined crossover phos-
phate level for whole sediments was in the range of Summer-fall
interstitial phosphorus measured in the south basin surficial
sediments.

Crossover concentration can be considered a reference
point in the phosphate exchange relationship permitting an es-
timate of the direction of net phosphate exchange, if the ame
bient level is altered. The relationship between the points
can be used to estimate extent of exchange and expected new
crossover value (Taylor and Kunishi, 1971). A greater slope
in the relationship identifying a particular equilibrium cross-

over value would indicate a greater sediment-mediated resistance

to change from that ambient level (ie. buffer capacity of the

sediment-lake water system). Capacity of sediment particulates
to buffer ambient phosphate would depend on extent of sediment
surface exposure relative to ambient water volume. For a given
sediment in suspension, this can be approximated by the sediment
weight to solution volume ratio (sediment : solution). Under
whole sediment : solution used (approximately 8 g dry weight

per liter; considered similar to sediment density near the in-
terface of the lake) buffering capacity was calculated from:

uM P04 9"1 exchanged x g sediment 1'1

 

crossover phosphate _ phosphate concentration
concentration of new or infusing medium

(modified from Mayer and Gloss, 1980). The equation yields the
ratio of sediment phosphate exchanged, expressed as uM PO4 1-1,

to the perturbation from crossover phosphate concentration,

25

also expressed as uM PO4 1-1. Data of Figure 2 indicate that

under experimental conditions, the capacity of whole sediments
to buffer ambient phosphate at 3.2 uM P04 1"1 was near 0.79.

In contrast to whole sediments, the strong tendency of
fine sediments to release phosphate was reflected in the com-
paratively high crossover value of 23 EM P04 1-1, roughly seven
fold that of whole sediments. The calculated capacity of fine
sediments to buffer or resist changes in ambient phosphate
near this relatively high level was 0.5.

Sediment response to relatively high initial levels of
ambient phosphate can yield information on capacity of sedi-
ments to retain phosphate (Williams, et a1. 1970) and potential
of sediments to supply phosphate to overlying water (Woodruff
and Kamprath, 1965). The amount of phosphate sorbed from re-
latively high initial lake water levels was dependent upon
sediment type (Figure 3). Relatively high phosphate release
from fine sediments (Figure 2) suggests that a portion of the
difference between the curves of Figure 3 may be due to dif— .‘
ferences in initial phosphate content of sediment particulates.
For either sediment, the changing Slope of the relationship
between net phosphate sorbed and corresponding ambient phos-
phate suggests that the degree of particulate response to am-
bient perturbation depends on the pre-existing equilibrium
phosphate level. The relationships indicate reduced capacity
of particulates to buffer high ambient levels, and suggest that
the rate of change of diffusive flux of interstitial phosphate

is a function of ambient phosphate level.

Figure 3,

26

Net reactive phosphate sorbed by whole sediments
(solid curve) and fine sediments (dashed curve)

from lake water of different initial reactive
phosphate levels, following 2 hours (A-A) and 48
hours (X-X) exposure. Horizontal axis indicates
resultant ambient reactive phosphate level.
Associated initial levels of ambient reactive
phosphate (uM P04 1'1) are indicated in parentheses.

27

:0. .8 :3 2292: 3.88m .835 SE
cemenNONN QNSNOQ 00. oh 00. on. 028. ON. 0: 8. cm on Oh Om on O? 0» ON 0.

..c~.wn.\ (a gnaw.

adofl

 

“OhthnN-

2

N.
n.
v.

 

0.

(MM up 0-6 '04 w") pquos «nucleons Mucous

28

The near linear relationship describing fine sediment
response suggests a higher capacity of particulates to retain
phosphate. This could reflect effects of greater weight-nor-
malized surface area (Olsen and Watanabe, 1957). However,
particle associated phosphate was considerably lower at ob-
served ambient levels compared to whole sediments, possibly
indicating a less energetic particle-phosphate bond. By com-
parison, only 50% of 48 hour fine sediment sorption occurred
within the first 30 minutes of exposure, while over 90% of
total phosphate sorbed by whole sediments occurred in that
interval.

Net exchange of reactive phosphate dominated the movement
of total soluble phosphorus between particulates and solution
for both sediment types examined (Figures 4 and 5). The
amount of net reactive phosphate release to lake water had no
apparent effect on the extent of net soluble nonreactive phos-
phate release from sediment(Figure 4). While fine sediments

1 .
' near seven

released approximately 0.35 uM nonreactive P04 9-
fold that of whole sediments, the reactive : nonreactive phos-
phate release ratio ranged from 2 : l to 6 : 1 for both sedi-
ments. The variation in the release ratio was principally due
to variation in the reactive phosphate fraction.

Under conditions of net reactive phosphate uptake from
solution, nonreactive phosphate release apparently increased
for both sediment types (Figure 5). Although the relationship

between reactive phosphate uptake and nonreactive phosphate

release was fairly insensitive for both sediments, it suggests

29

Figure 4. Relationship between release of soluble nonreactive
phosphate and release of reactive phosphate by
sediments following 0.5 hours (O-O), 2 hours (A-A),
and 48 hours (X-X) exposure to lake water with low
initial reactive phosphate levels. Solid line,

results from whole sediments; dashed line from fine
sediments.

30

l2

 

 

 

l 0
<
< 'X _
I x .93;
I. .9
3
Q '00 s05 >s
,. e
I -00 0»
l° 0"
0..
«IN:
3
103
a)
E
«'0 SE.
.2
D
is;
L. .... g
.3
o
O
J ng
0X
0 ‘-
l_ l l 1 L
'0. ‘2 '°. N -:
O O O O O

(mbgem Mp |.bl'od w ‘0‘) pesoeiea etoquoqd Panacea-mu

29

Figure 4. Relationship between release of soluble nonreactive
phosphate and release of reactive phosphate by
sediments following 0.5 hours (O-O), 2 hours (A-A),
and 48 hours (X-X) exposure to lake water with low
initial reactive phosphate levels. Solid line,
results from whole sediments; dashed line from fine
sediments.

30

 

 

 

4
q 'x ..
Ix 022
I. .9
3
4 lo0 '0’ ,i
r e
I ~00 E»
|° °"
0.
.1»:
3.
.49“:
U)
.3
.... 0'2.
.2
0
“0‘5.
8
a
at” g
a:
o
8
#610:
odx
0 ‘-
L l J l 1
If! ‘1’ '0. 0! -.
o o’ o o o

(NOIOM Mp .-b’od w or) posoelea eioquoqd mucosa-non

31

Figure 5. Relationship between release of soluble nonreactive
phosphate and sorption of reactive phosphate by
sediments following 0.5 hours (O-O), 2 hours (A-A),
48 hours (X-X) exposure to lake water with high
initial reactive phosphate levels. Solid line,
result from whole sediments; dash line from fine
sediments.

32

.222. to .-o 40.. 23 933 2289... 388m

 

E m. N. : O. o m b m n v m N .
u u n J d '1 4 d u d 10 u d e. d N
x
.a m.
o a
a
D
mw
x LNG M
o d
m
d
.3 W
x .0...
ill 0 a
41x1 4 o \«A m.
\ .8 w
0 \\ m.
\ .4...
\ .3 w
\ ..d
\ . m
x x \ 5....
llll'llll'ik 10.0 p
..A.
q 0
m
.3 M

 

 

33

that mechanisms of sediment uptake of reactive phosphate may
involve displacement of phosphate-containing organic molecules.
The hypothesis that reactive phosphate added to soil systems
displaced sorbed organic phosphate was not supported by the
data of Wier and Black (1968). According to Latteral, et a1.
(1971), however, reactive phosphate added to and sorbed by
sediments resulted in a substantial increase in the proportion
of organic to reactive phosphate in equilibrium solution. This
was interpreted as being caused by displacement of organic
phosphorus from sediment sorption sites by the more strongly
sorbed reactive phosphate.

Interpretation of changes in solution phosphate levels
following sediment introduction partially rests on observations
of other system variables that may influence results. Assess-
ing potential impact of pH-dependent dissolution/precipitation
reactions involving calcium phosphate compounds, or changes in
oxidation-reduction potential leading to solubilization of
phosphate binding sediment compounds is critical to a discussion
of mechanisms of sediment-solution phosphate exchange. Toward
this end, solution calcium, pH, and iron were monitored.

The general directions of change in ambient pH and calcium
were similar in both whole and fine sediment-lake water systems
(Figure (H. Further, for each sediment type, trends in ambient
calcium and pH were similar at all phosphate levels examined,
ie. there were no correlations between the extent of ambient
calcium increase or pH decrease and initial ambient phosphate,

or extent of inorganic phosphate exchange. Measurements of

Figure 6.

34

Ambient calcium (mean and 1 standard deviation) and
pH (mean and range) with duration of whole (A-A)

and fine (o-o) sediment exposure to lake water
phosphate solutions. The dashed lines indicate hy-
pothesized course of change in ambient calcium within
initial 30 minutes sediment-lake water exposure. Time

0 indicates point of sediment introduction to lake
water.

35

(pm 1‘ ..3’ “low ) “WINDS W‘NWV
o o o o
9' 0' as as

r I I j

 

 

,r——--1‘r

1"""’—'

onion
5555

L

Hd "1°!qu

.‘q
p.

 

¢

 

Duration 01' Sediment-Lake Water Exposure (Hours)

36

soluble calcium and ambient pH were grouped to obtain a mean
for each exposure interval.

Lake water pH increased from 7.4 to 7.95 and 8.10 upon
introduction of whole and fine sediment (time 0). Solution
calcium increased over lake water levels by about 20 uM 1-1
within 30 minutes sediment-lake water exposure. The dashed
lines indicate hypothesized course of solution calcium change;
suggesting that increases in soluble calcium corresponded to
observed pH increases (of. Turner and Clark, 1956). These
trends may reflect response of calcium carbonate equilibria
and as such were largely independent of observed phosphate
reactions.

Calcium carbonate solubility is a function of pH and
carbon dioxide content of water, the relationship can be de-
scribed by:

8 pCa = pH + a log PCO2 - 4.93

where pCa is the negative log of the molar concentration of

calcium, and FCC is the partial pressure of carbon dioxide.

2

At equilibrium, partial pressure of aqueous CO2 is equal to

the partial pressure of atmospheric CO2 (Turner and Clark, 1956).
Assuming lake water used for experimentation was in equi-

librium with atmospheric CO2 prior to sediment addition, PCO2

= 0.0003 atm, and 5 log PCO = —l.761. At pH 7.4, predicted

2
aqueous calcium level would be 3.8 x 10"2 molar. Measured

levels were near 1 x 10-3 M. This indicates a tendency for
calcium carbonate, in contact with lake water, to dissolve.

Dissolution of CaCO3, and subsequent increase in CO32-, could

37

incur a pH increase. A resultant disequilibrium between aqueous
and atmospheric CO2 would promote reinvasion of CO2 from the
atmosphere causing a gradual pH decrease in solution. The lab—
oratory results suggest that the rate of potential calcium car—
bonate dissolution early in the exposure period was greater

than the rate of invasion of atmospheric CO The observation

2.
that pH-Ca equilibrium was not attained in the duration of the
experiments (ie. final calcium levels were lower than predict-
ed) may reflect insufficient calcium reserves or reduction in
the rate of dissolution of CaCO3 coated with adsorbed organic
compounds (Wetzel, 1971, 1972). These considerations suggest
that observed calcium increases may not involve dissolution of
phosphate containing minerals.

Filterable iron content did not change measureably in the
sediment-lake water system (Appendix 3), suggesting that redox
potentials did not fall below a critial level for chemical re-
duction of phosphate-binding iron oxides. While observed levels
of soluble iron were low relative to analytical precision, they
were high relative to predicted concentrations based on solu-
bilities of iron oxides and hydroxides. Assuming particulate
iron passing the 0.5 um filter was minimal, the observed iron
levels may reflect presence of naturally occurring metal che-
lating compounds in solution.

The reciprocal nature of phosphate exchange implies, for
example, that the extent of particulate release of phosphate,
in response to a reduction in ambient levels, is modified by

the correspounding increase in solution phosphate. This suggests

38

the importance of water movement characteristics on the extent
of phosphate release from sediment particulates. Turbulent
displacement of sediment particles to the relatively phosphorus-
poor water column could maximize phosphate release from particle
surfaces.v

Figure 7 shows that phosphate release from fine sediments
was greater when ambient phosphate was maintained at relative;
ly low levels by lake water replacement. The dashed line of
the figure indicates that following 2 hours exposure to 100 ml
lake water, 0.5 g sediment released approximately 1.28 uM
PO4 g-l; corresponding to an ambient phosphate level of 6.5
uM PO4 1-1. While sediment impact on ambient phosphate level
was reduced from that observed with l 9 material per 100 ml
(Figure 2) weight-normalized phosphate release was higher.

This could indicate inhibiting effects of increased ambient
phosphate on rate of net particulate release. Fine sediments
released a total of approximately 1.57 uM P04 9.1 in response;
to ambient medium renewal.

Contrast in sediment response shown in Figure 7 was dom-
inated by differences in phosphate release during the 15 - 30
minute exposure interval. Within the initial 15 minute exposure
interval, sediments released 1.0 uM P04 9.1 resulting in ambi-

1

ent level near 5.1 uM P04 1. . Materials exposed to renewed

medium, containing 0.1 uM P0 1.1, released an additional 0.38

4
0M PO4 g-l, while continued exposure to 5.1 uM P04 1"1 restrict—
ed release to 0.13 MM PO4 g_l. The ratio of differences in

phosphate released to differences in ambient phosphate within

Figure 7.

39

Net reactive phosphate released from fine sediments
under conditions of ambient phosphate accumulation
(A-A), and cumulative phosphate released from fine
sediments to ambient lake water replaced at 15
minute intervals (O-O). Lake water initially
contained 0.10 uM P04 1‘1. Values are expressed

as a mean of six replicates :1 standard error.

ON.

0:

8.

305558386 .883 8.0.. (ESE—com to 522.5

cm on Oh cm on O? on

d d I d u I

 

l

L

 

U

08: m
0

u.

A

0
3... W
U.

D

N
08.. m
m

s

O

0.

con. I
w

w
32%
A

m

. K
o8. m
9

m.

N

41

the 15 - 30 minute exposure period was:

0.38 - 0.13

5.10 - 0.10 "" 0°05 '

 

This ratio declined during subsequent exposure intervals, in—
dicating effects of diminishing quantities of desorbable phos-
phate associated with particles exposed to renewed medium.

The cumulative total of phosphate released to renewed
medium, 1.57 uM P04 9.1, was considered an estimate of de-
sorbable particulate phosphate at time 0; and measured release
during exposure intervals as a reduction in that amount. The
data conformed to first order kinetics.

The general form of phosphate exchange can be expressed
as:

particulate - PO solution - P04

4

Wtivfi‘

2

where kl and k are exchange rate constants. The rate of part-

2
iculate -lake water phosphate exchange can be expressed by:

dP _ _
jg — klP kZS

where P is particulate P04 and S is solution P04. If solution
phosphate is reduced to levels below those supporting steady
state and maintained low through ambient replacement, then the
rate of change of P in approaching equilibrium with phosphate-

poor lake water may be approximated from:

dP _
-a—t- — klP

Applied here, the equation indicates that the rate of change of
particulate desorbable phosphate, g; , is equal to a constant

times the amount on the particle, and

42

—5 = -kldt .
Integrating both sides the equation becomes a linear ex-
pression,
1n P = -klt + P .
The intercept P is the estimated desorbable particulate phos-
phate at t = 0. Specifically, for the data presented here,

1) = -o.059 + 0.324 .

ln(uM P04 9'
The linearity of the data obtained (r2 = 0.995) implies that
observed release rates were a function of particle content.
The difference between the calculated intercept of 1.383 uM

P04 g-l .324

(e = 1.383) and estimated 1.57 uM may reflect error
involved in the implicit assumption that phosphate release
within the initial 15 minutes interval was negligibly affected
by ambient phosphate levels. The slope, k1 was an estimate of
the rate of change of particulate phosphate and calculated as
0.059 min-1. While the release rate was specific to particle
and environmental characteristics, the suggestion that it was
a function of the amount of desorbable phosphate associated
with particles exposed to a phosphate-poor environment may be

k
generally applicable when Ea is small.
1

DISCUSSION

In contrast to overlying water, inorganic phosphorus
frequently constitutes a major portion of total phosphorus in
sediments (Sommers, et a1. 1970). Phosphate originating from
mineralization of sedimented organic matter or flux from deep—
er anaerobic sediment layers can result in interstitial phos-
phate levels greater than those of overlying water. Relatively
high interstitial phosphate can result in gradients, such as
the one measured here. Because phosphate gradients represent
potential for release of sediment phosphate to overlying water,
the nature of mechanisms involved in developing and maintain-
ing gradients deserves attention.

Estimating the potential of phosphorus liberation from
littoral sediments to overlying water can be approached by
examining the effects of certain phosphate reactions between
solid and solution sediment components. In addition to the
impact of phosphate sorption/desorption reactions, equilibria
characterized by solubility of solid Ca-phosphate precipitates
can influence levels of interstitial phosphate, particularly
in sediments of lakes that experience epilimnetic calcium pre-
cipitation (Stumm and Leckie, 1971). Evidence of a midsummer
decline in Lake Lansing littoral and epilimnetic calcium, sug—
gests presence of calcium in Shallow sediments and warrants

examination of the potential impact of Ca-phosphate solubility

43

44

equilibria on observed phosphorus levels in sediment solution.

Since calcium phosphate equilibria expressions involve
interactions between activities of specific ionic phosphate
species and other soluble components of the aqueous medium,
it is useful to consider factors influencing these variables
with respect to the data presented. Soluble reactive phos-
phorus is considered to exist as one or more of the forms of
phosphoric acid, ie. H3PO4, H2P04—, HPO42-, and PO43-. The
forms dominating soluble phosphate levels depends on pH. Under
pH conditions of the work discussed here, concern is with H2PO4-
and HPO42- (Figure 8). If soluble reactive phosphorus measure-
ments used in this study are taken as estimates of ionic phos-
phate levels, the amount of each form present can be determined
from measurements of ambient pH. An additional level of con-
sideration is drawn from observations of physical chemistry that
the reactivity of a given ion may not be entirely determined
by its molar concentration. The concept is contained in the
Law of Mass Action which states that 'at a given temperature,
the rate of a chemical reaction is proportional to the active
masses of the reactants', indicating that activity of soluble
phosphate species is a relevant parameter when considering
chemical reactions.

Concept of activity accounts for differences between
observed and ideal behavior of solutes. When greater than

10’4

molar, the ionic strength of a given solution has been
empirically determined to be an important factor modifying the

effects of the concentration of the ion of interest.

45

Figure 8. Relation between pH and relative prOportions of
soluble reactive phosphate species. Calculations
based on dissociation constants of phosphoric acid,
mono-, and dihydrogen phosphate ions in dilute

solution.

Ill '

P0

Hpo:

H 2 Po;-

HsPO4

46

 

 

 

I.0

0.9 t
0.7 '

I l l
w. «o. a
O O O

ewquoud emooeu

0.3 '

‘t
o
e

.3 L
N ".
O O

0.
0

names IO uouoou wow

I4

I3

l2

l0

pH

47

Quantification of this modifying factor, expressed in the activity

coefficient, takes into account the concentration and valence of
all ions in solution. For solutions of ionic strength less
than 0.01 molar, the activity coefficient, y, can be calculated
from:

log Y = -A 22/5—
where A is the temperature-dependent Debye-Huckel constant,
equal to 0.5085 at 25°C, 2 is the valence of the ion of inter-
est, and u is the ionic strength of the solution, u== %(.§
molar concentration of ion 1 times the valence of ion i l-l
squared). Solute activity is equal to the concentration times
the activity coefficient (Tinoco, et a1. 1978).

Activity of soluble reactive phosphate is a function of
pH dependent ionic distribution (Figure 8) and solution ionic
strength (Figure 9).

Determination of ionic strength is most accurately made
with knowledge of all ions of solution. However, it can be
approximated with measurements of dominant forms (Stumm and
Morgan, 1981). Ionic strength was approximated here as 3 x molar
calcium levels (after Lindsay and Moreno, 1960). Based on mea-
surements of soluble calcium of Lake Lansing epilimnion, ionic
strength was taken as 3 x 10-3molar. Griffin and Jurinak (1973),
working with 124 river water samples and 27 soil extracts, found
an excellent correlation (r2 = 0.992) between ionic strength
and electrical conductance, u = 0.013 EC, where EC is expressed

as millimhos cm-l. Electrical conductance of Lake Lansing epi-

. . . . . -l . . .
limnion varied around a mean of 0.26 mllllthS cm , Indicating

48

Figure 9. Influence of ionic strength (u) on activity
coefficient (7) of mono- and dihydrogen phosphate
for solutions of ionic strength less than 0.01
molar. V was calculated from: log y = -0.5085
22 /E—

 

.49

I.000 r _

3300'

.700 "

 

 

.600

l l

500 4.00 3.00 2.00

Ionic Strength 0f Solution (- I00 males .9." i

50

an ionic strength estimate near that made with calcium.

Incorporating information presented above and appro-
priate solubility expressions, the expected levels of total
soluble reactive phosphorus in equilibrium with dicalcium phos-
phate, octacalcium phosphate, and hydroxyapatite can be cal-
culated (Figure 10; also of. Appendix 4). Calculations yield
activities of soluble phosphate which have been converted to 4
concentrations (using assumptions presented above) for compari-
son with data collected in this study. Although calcium phos-
phates other than those presented in Figure 10 may form in
sediments, lack of information concerning their solubilities
precludes inclusion in the Figure. Fluoroapatite which gener-
ally has a lower solubility than hydroxyapatite in the pH range
of interest, has been excluded for lack of information concern-
ing solution levels of fluoride ion.

Significance of the solubility diagram of Figure 10 stems
from the potential of sediment interstitial water phosphorus
levels to approach equilibrium with the stable calcium phos-
phate minerals considered. Measured soluble phosphate values
falling above a given line indicate supersaturation with respect
to that compound and suggest that precipitation is possible.
Soluble phosphate levels below the line indicate undersaturation
and that the corresponding compound would dissolve. Figure 10
suggests that dicalcium phosphate and octacalcium phosphate
form or remain stable only at high levels of soluble phosphate.

If soluble phosphate levels decreased, as a result of diffusion,

biogenic uptake, or association with hydroxyapatite, these

Figure 10.

51

Solubility diagram for certain calcium phosphate
compounds in sediments at 250C. Levels of soluble
reactive phosphate were calculated by considering
appropriate solubility equilibria an assuming
calcium levels maintained at l x 10‘ molar (solid
lines) and 3 x 10'3 molar (dashed line). (°)

refer to ambient reactive phosphate levels following
48 hours exposure to whole sediments to high initial
levels of ambient phosphate. (X) same, but for

fine sediment exposure. Phosphate activities were
converted to concentrations for comparison with
observed levels of total interstitial phosphate of
south basin sediments.

Interstitial Reactive Phosphate Concentration(-Ioo moles P041")

 

5 6 7 8

52

   
      
 

dicalcium phosphate dihydrate
00 H P04- 2 H20

 

octacalcium phosphate
Ca4HI (20413-3 H20

hydroxyapatite
CaDIPO‘Ié 0H I2

P
b

l

b

‘01.

Ambient pH

53

minerals would tend to dissolve. Using measured levels of to-
tal interstitial phosphorus of sediments as an estimate of
soluble reactive phosphate, surficial sediment solution levels
ranged from about 0.581 to 3.871 uM 1-1, corresponding to 6.24
and 5.41, respectively, on the vertical scale. Based on the
assumptions used to calculate the isotherms, sediments would ex-
perience dissolution of any dicalcium or octacalcium phosphate
present and precipitation of hydroxyapatite at pH levels above
7. Calcium increases above 1 x 10.3 molar would have the effect
of shifting the lines parallel (to the ones indicated) and
downward (cf. predicted hydroxyapatite isotherm under conditions
of 3 x 10"3 molar calcium, dashed line of Figure 10). In
general, solution pH, calcium, and ionic strength can greatly
influence equilibrium phosphate levels. Observed variability

in interstitial phosphorus could result from differences in the
magnitude of these parameters over short distances within the
sediment.

Considerations of interstitial water buffering of solu-
ble phosphate imposed by calcium phosphate equilibria suggest
that equilibrium levels can be highly influenced by physico-
chemical parameters. In the dynamic environment of Shallow
water sediments, an important variable modifying sediment im-
pact on soluble phosphate is the rate of response to ambient
changes. In general, only the most soluble phosphate compounds
react fast enough to determine solution phosphate activity un-
der environmental conditions that may change within short time

intervals. Kinetic information concerning the approach of

 

 

54

ambient phosphate to equilibria with calcium phosphate minerals in

sediments is, at best, qualitative. Dicalcium phosphate is con—

sidered the most reactive of the compounds discussed here,
with reaction kinetics suggesting that equilibrium could be
obtained within hours of ambient phosphate perturbation
(Moreno, et a1. 1960). Hydroxyapatite is the most insoluble
of the forms presented. The pH and temperature dependent rate
of formation involves slow transformation of amorphous calcium
phosphate into crystalline form. Reported rates of apatite
formation vary from two weeks (Stumm and Leckie, 1971) to over
three months (Moreno, et a1. 1960). It appears that at pH
levels greater than 7, apatite phosphate may influence the
direction and endpoint of ambient phosphate changes. However,
attainment of equilibrium levels will likely be restricted to
largely undisturbed systems where slow dissolution and precip-
itation reactions have adequate time to dominate ambient phos-
phate levels.

Laboratory work reported here shows that under aerobic
conditions a major portion of sediment-induced change from
initial phosphate levels occurred within 30 minutes of expo-
sure. The rate of change following 2 hours exposure to all
initial phosphate levels examined was greatly reduced. Quasi-
equilibrium phosphate levels observed in the laboratory follow-
ing 2 and 48 hours exposure to high initial levels of ambient
phosphate are plotted at the corresponding mean final pH
(Figure 10). Solution ionic strength, calcium, and temperature

in experiments were similar to those assumed when calculating

55

the isotherms. While a few points fall near predicted phos-
phate levels in equilibrium with octacalcium phosphate, viewed
as a set, the data appear to be independent of the isotherm.
Moreno, et a1. (1960) observed that approximately one month
was required for formation of the mineral under a similar pH
and temperature regime. In the laboratory work presented here,
calcium and phosphate activities did not exceed the solubility
product of the more responsive dicalcium phosphate mineral.
These observations indicate that 2 and 48 hour ambient phosphate
concentrations stabilized at levels other than those predicted
by solubility equilibria of calcium phosphate minerals consid-
ered and implicate other controlling mechanisms.

Sediment induced changes in ambient reactive phosphate
have been attributed to phosphate association with pre-existing
solids rather than through formation of insoluble precipitates.
The Langmuir adsorption isotherm has been commonly used to
quantitatively describe particulate induced changes in ambient
reactive phosphate for soil (cf. Ellis and Knezek, 1972) and
sediment (cf. Syers, et a1 1973) systems. In contrast to sol-
ubility equilibria, Langmuir analysis describes the net effect
of participating particulate-solution phosphate associations.

The Langmuir equation was originally based on the ki-
netic theory of gases. The same equation has been applied to
adsorption of ions from solution by solids (Olsen and Watanabe,
1957). The analysis stems from the general consideration that
observed solution levels result from the net effect of desorp-

tion and sorption reactions acting Simultaneously (Shapiro and

56

Fried, 1959). Individually, the rate of release of phosphate

from sediments is a function of sorbed particulate phosphate,

X

m
’1)

rate of release = kl(m

where k1 is a constant. Rate of phosphate sorption on parti-
culate surfaces is a function of solution phosphate activity,
(P), and the difference between phosphate sorbed and a sediment-
specific sorption maximum, B:
rate of sorption = k2(P)(B-%) .
At equilibrium, the rate of release equals the rate of sorption:
kl(%) = 1020:) (Ia-g) .

A linear transformation of the equilibrium expressions yields:

(P) = 1 + (P)
x KB B
m k2
where K is a sediment-specific constant, equal to E— , and is

1

related to the binding of particulate surfaces for sorbed phos-
phate.

The Langmuir constants, B and K, can be used to quanti-
tatively describe the relationship between sorbed and equilib-
rium ambient phosphate for a given sediment-solution system.
Values for the constants(;?n be obtained from experimental ob-

servations if a plot of x versus (P) is linear. The Slope
m

1

and intercept are equal to l and respectively.

i.
B KB '
Since E-of the Langmuir relationship refers to amount
of phosphate sorbed, measurements of solution phosphate losses
Should be adjusted for the amount sorbed prior to sediment-solu-

tion exposure. In laboratory work with whole sediments, this

57

adjustment was taken as the amount released following long-term

exposure to phosphorus-poor lake water (0.305 uM P04 9.1). This
was added to measured solution phosphate losses to estimate %-.

Corresponding ambient phosphate activity was estimated by as-

suming ionic strength to be 3 x 10-3

molar (ie. 3 x observed
soluble calcium) and pH of 7.6 . Under these conditions, mea-
sured concentrations of soluble reactive phosphate can be con-
verted to activities by multiplying by 0.82 . A plot of values
obtained following long-term whole sediment exposure to the
three highest levels of initial ambient phosphate resulted in
a straight line with a slope of 0.058, B = 17.179 uM P04 9.1;
and an intercept of 2.125, K = 0.027 liters per uM P04.

Estimation of the constants K and B for fine sediment
material was made by adding 1.38 uM P04 9.1, the quanitity of
desorbable phosphate estimated from analysis of P04 released
to renewed ambient medium, to measured solution phosphate losses.
At an ambient pH of 7.75 and ionic strength of 3 x 10"3 molar,
the conversion factor used to obtain ambient phosphate activity
was 0.81 . The constant B, indicating sorption maximum, was
estimated as 47.091 uM PO4 g-1. The constant K, related to
phosphate binding energy, was 0.002 liters per uM P04.

The sorption maximum of fine material was over twice that
calculated for whole sediments, and may reflect the greater
proportion of small particles ( < 227 pm in diameter). The
larger surface area per unit weight of material implies a greater

number of sites for phosphate sorption. An interesting feature

of the analysis is the apparent ten fold reduction in the value

58

of the constant related to the binding energy of processed fine
sediments. This may indicate that expected carbonate precipi-
tation on particulate surfaces during preparation of fine sedi-
ments altered phosphate sorption characteristics. Green, et a1.
(1978) found a negative correlation between value of the sorption
energy parameter and calcite content of Maumee River sediments.

Figure 11 illustrates the relationship between sorbed
phosphate and ambient activity, based on experimental estimates
of Langmuir constants B and K. The sediment-specific relation-
ships suggest that particulate-solution phosphate exchange would
result from perturbation of equilibrium ambient phosphate act-
ivity (resulting from infusion of epilimnetic water through
surficial sediments, for example). The direction and extent
of particulate response would depend on phosphate activity of
infusing medium, and sediment weight/solution volume ratio.

An equation that quantitatively describes sediment re-
sponse to perturbation from equilibrium ambient phosphate can

be obtained from the Langmuir expression:

flitL-I-fl
x_ KB B
m
_1 1
- B ( R + P)
1
— 1 l
E‘s—p‘rfl’)
m
x = B(P)
m I
E + (P) (Equation 1).

If y is taken as the net quantity of phosphate exchanged

with sediment, then g-- y is the level of sorbed sediment

Figure 11.

59

Calculated relationship between phosphate sorbed
by whole and fine sediments and ambient phosphate
activity. Each isotherm was developed from esti-
mations of Langmuir constants B and K. Ionic
strength 3 x 10'3 M and equilibrium with respect
to particulate—solution phosphate exchange were
assumed. The pH was taken as 7.60 and 7.75 for
whole and fine sediments, respectively.

60

ON

...... 40.. .23 3.264 22.32... 2.88.... .835
8.8. 8.2. 8. o... 9.8. 8. o: 8. 8 8 2 8 8 0.. on 8 o.

q q 1 1 q u u q 1 d d 1

222.68 2....

382.com 22.3

 

 

    
 

O) O h 10 D V '0 N
(I-D’Od WV) (39qu aloudsoud mucosa

enmzo

 

..“2

61

phosphate following exchange. Ambient phosphate in equilibrium
with g'- y will depend on impact of phosphate mass exchanged

on solution activity:
B(P1 + AP)

 

E. =
m Y 1
... (P1 + AP)

K (Equation 2)

where P1 is the phosphate activity of infusing medium, and AP
is a function of the solid : solution following perturbation
and a conversion factor relating phosphate exchanged to result-
ant ambient phosphate activity.

Laboratory results obtained with whole sediments can be

used to illustrate an application of these considerations.

Estimated phosphate activity at crossover, 2.624 uM P04 1-1

(ie. 3.20 x 0.82 = 2.624) reflected an estimated sorbed phos-
phate level, %" of 1.137 uM P04 9-1 (from Equation 1). Per-
turbation of the sediment system by introduction to phosphate-

poor lake water (P1 = 0.13 x 0.82 = 0.107 uM PO 1-1) resulted

4
in net phosphate release from sediments. At the solid : solu-

tion of 0.8 g dry weight per 100 m1 lake water, each uM P04

released per gram would increase ambient phosphate activity by

6.56 uM PO 1'1 (ie. 8 um PO 1‘1 x 0.82 = 6.56). For this ex-

4 4
ample, AP of Equation 2 is equal to 6.56 g 2-1 times y; where
y is expressed in uM P04 9.1. Expected weight-normalized phos-

phate release, y, can be calculated from:

= 17.179 (0.107 + 6.56y)
37.037 + (0.107 + 6.56y) '

1

 

1.137 -

Expected phosphate release is 0.28 uM P04 9-

In general, calculated net phosphate exchange resulting

62

from whole sediment exposure to initial phosphate activities
examined agree reasonably well with observations of net exchange
(Table 1).

Employing analogous considerations to results obtained
with fine sediment, calculated net release following exposure
to initial phosphate activities used is somewhat less than that
observed (Table 2). This may reflect error in the estimated
crossover value since points used for its estimation were widely
separated (cf. Figure 2).

Influence of solid : solution on net phosphate exchange
is indicated from measurements of phosphate desorption from
0.5 9 fine sediment under conditions of ambient medium renewal.
In this experiment, 100 m1 lake water was renewed six times and

considered similar to a solid : solution of 0.83 g 1-1. Assum-

ing pre-disturbed ambient activity of 18.63 uM PO4 1-1, the
calculated net release of 1.58 uM P04 9.1 is comparable to
measured cumulative release of 1.57 uM P04 9.1. Similarly,
calculated desorption from 0.5 g to 100 m1 lake water is 1.22
pM P04 9.1 and comparable to observed release of 1.26 uM.

A few limitations of the preceded analysis should be
noted. Care should be exercised when applying constants. Any
experimental error experienced when obtaining data could lead
to serious deviations between calculated and operational values
of B and K. Further, an implicit assumption of Equation 2 is
that phosphate sorbed is entirely releaseable. This may not

be true. If sorbed phosphate forms metastable associations with

the particulate sorbent, hydrated ferric oxides for example,

63

Table 1. Calculated and measured net phosphate exchange
(uM P0 9.1) between whole sediments and infusing

lake water of different initial phosphate activity
(uM P04 1'1). Values assume equilibrium conditions

with ionic strength 3 x 10'3M and pH 7.60 .
Negative sign indicates net sediment sorption of

 

 

phosphate.
Initial Phosphate Calculated Measured
Activity Exchange Exchange
0.107 0.28 0.30
1.246 0.15 0.16
-0.27 -0.31

5.108

 

64

Table 2. Calculated and measured net phosphate exchange
(uM P04 g'l) between fine sediments and infusing
lake water of different initial phosphate activity
(uM P04 1'1). Calculated exchange is based on
experimentally estimated ambient crossover activity
of 18.63 uM P04 1'1 (A), and on a hypothetical

crossover activity of 22 uM P04 1' (B). Values
uilibrium conditions with ionic strength

 

 

assume gq
3 x 10' M and pH 7.75.
Initial Calculated Exchange -
Phosphate Measured
Activity A B Exchange
0.057 0.96 1.14 1.18
1.284 0.90 1.07 1.08
4.965 0.71 0.87 0.82

 

65

changes in the sorbent over time may affect its ability to re-
lease phosphate following perturbation. While these consider-
ations suggest that modifications of Equation 2 may be required
for quantitative predictions, the expression may be used to
qualitatively envision response of a sediment-solution system
to ambient perturbation.

A generalized representation of sorption/desorption con-
cepts discussed is presented in Figure 12. Based on observa-
tions of this study, the figure illustrates hypothetical interr
actions between sediment particulates, characterized by phosphate
binding energy (K) and sorption capacity (B), and infusing
media. The direction and extent of phosphate exchange depends
on equilibrium interstitial phosphate activity (A<B<C), phosphate
activity of infusing medium (A’<B’<C’), and solid : solution
of the disturbed system. For example, under conditions of
constant solid : solution, particulate release of phosphate
would be expected following infusion of medium of activity A’
into sediments previously resting under an interstitial equil-
ibrium phosphate level of B. A new interstitial equilibrium
could result and would be expected to fall somewhere between A
and B. On the other hand, if medium C’ disturbed sediments
resting at B, phosphate sorption would be predicted, and a new
interstitial equilibrium expected between B and C. No net ex-
change is expected if interstitial and infusing medium phosphate
activities are equal. The Figure also suggests that the extent
of particle-solution phosphate exchange is expected to increase

with increasing difference between infusing medium and

Figure 12.

66

Net exchange of particle associated phosphate as
a function of phosphate activity of infusing
medium and of interstitial water in an idealized
sorption-desorption exchange system. Letters
indicate phosphate activity level (A<B<C) and
superscripts indicate infusing media. The dashed
curve indicates sediment response to infusing
phosphate activity level (8’) if sediment :
solution is reduced as result of perturbation.

 

 

 

67

 

 

Equilibrium. Interstitial Phosphate Activity.

 

 

(__

uoudios o uoudioseo

owquoqd peiogoossv - opumd

68

interstitial phosphate activity levels.

The dashed line of Figure 12 indicates expected particulate
response if infusion of medium B’ resulted in a sediment : solu-
tion lower than that of the sediment environment. For example,
if the movement of infusing medium (with activity B’) were
great enough to lift surficial sediment particulates (previously
resting at C) from the interface, phosphate release would be
greater than that expected if particulates were not displaced.

Literature estimates of sediment contribution of phosphate
to overlying water are largely limited to estimates of diffusive
flux (Burns and Ross, 1972; Holdren, et a1. 1977). They repre-
sent a largely static relationship between lake and sediment
compartments in that the potential impacts of water movement
are not considered. Limnological thought is pervaded by the
idea that summer density stratification results in stagnation
of large lake-strata. There is little basis for this view.

Even deep water sediments are disturbed from several millimeters
to several centimeters (Davis, 1968, 1973; Wetzel, et al. 1972).
Considerations drawn from the data of this study suggest that
particle-solution chemical equilibria can interact with hydro-
mechanics at the shallow sediment-water interface to influence
rate and extent of sediment phosphate flux. Shallow sediments
must be considered a particle-solution phosphate exchange system
that functionally interacts with water movements to influence

phosphorus dynamics.

LITERATURE CITED

American Public Health Association. 1976. Standard Methods
for the Examination of Water and Wastewater 14th ed.
Washington, D. C., American Public Health Association,
1193 pp.

Bache, B.W. 1963. Aluminum and iron phosphate studies relat-
ing to soil. I. Solution and hydrolysis of variscite
and strengite. J. Soil Sci., 14: 113-123.

Batterson, T.R. 1980. Arsenic in Lake Lansing, Michigan.
Ph.D. Diss., Michigan State University, East Lansing,
Michigan.

Burns, N.M., and C.W. Ross. 1972. Oxygen-nutrient relation-
ships within the central basin of Lake Erie. In H.E.
Allen and J.P. Kramer, eds. Nutrients in Natural Waters.
New York, Wiley-Interscience. 457 pp.

Carignan, R., and R.J. Flett. 1981. Postdepositional mobility
of phosphorus in lake sediment. Limnol. Oceanogr.,
26: 361-366.

Carritt, D.E., and S. Goodgal. 1954. Sorption reactions and
some ecological implications. Deep-Sea Res., 1: 224-243.

Chen, Y.S., J.N. Bulter, and W. Stumm. 1973. Phosphate adsorp-
tion reactions with clay minerals. J. Colloid Interface
Science, 43: 421-436.

Clesceri, N.L. and G.F. Lee. 1965. Hydrolysis of condensed
phosphate -I: Non-sterile environment. Int. J. Air Wat.
Poll., 9: 723-742.

Davis, M.B. 1968. Pollen grains in lake sediments. Redeposi-
tion caused by seasonal water circulation. Science, 162:
796-799.

Davis, M.B. 1973. Redeposition of pollen grains in lake sedi-
ment. Limnol. Oceanog., 18: 44-52.

Edzwald, J.K. 1977. Phosphorus in aquatic systems: the role
of the sediments. In I.H. Suffet, ed. Fate of Pollutants

in the Air and WateF—Environments, Vol. 8. New York,
Wiley-Interscience. 214pp.

Ellis, B.G., and B.D. Knezek. 1972. Adsorption reactions of

69

7O

micronutrients in soils. In J.J. Mortvedt, P.M.
Giordano, and W.L. Lindsay, eds. Micronutrients in
Agriculture. Madison, Wisconsin, Soil Sci. Soc.
Amer., Inc., Chapter 4.

Fitzgerald, G.P. 1970. Aerobic lake muds for the removal of
phosphorus from lake waters. Limnol. Oceanogr., 15:
550-555.

Golterman, H.L., C.C. Bakels, and J. Jakobs-Mogelin. 1969.
Availability of mud phosphates for growth of algae.
Verh. Internat. Verein. Limnol., 17: 467-479.

Green, D.E., T.J. Logan, and N.E. Smeck. 1978. Phosphate
adsorption-desorption characteristics of suspended
sediments in the Maumee River Basin of Ohio. J. Environ.
Qual., 7: 208-212.

Griffin, R.A., and J.J. Jurinak. 1973. Estimation of activity
coefficients from the electrical conductivity of natural
aquatic systems and soil extracts. Soil Sci., 116:
26-30. -

Harter, R.D. 1968. Adsorption of phosphorus by lake sediment.
Soil Sci Soc. Amer. Proc., 32: 514-518.

Hayes, F.R. 1964. The mud-water interface. Oceanogr. Mar.

Hayes, F.R., J.A. McCarter, M.L. Cameron, and D.A. Livingstone.
1952. On the kinetics of phosphorus exchange in lakes.

Holdren, G.C., D.E. Armstrong, and R.E. Harris. 1977. Inter-
stitial inorganic phosphorus concentrations in Lakes
Mendota and Wingra. Water Res., 11: 1041-1047.

Hutchinson, G.E. 1957. A Treatise on Limnology. I. Geography,
Physics, and Chemistry. New York, John Wiley and Sons,
Inc., 1015 pp.

Johnson, D.L. 1971. Simultaneous determinations of arsenate
and phosphate in natural waters. Environ. 501-:
Technol., 5: 411-414.

Latterell, J.J., R.F. Holt, and D.R. Timmons. 1971. Phosphorus
availability in lake sediments. J. Soil and Water
Cons., 26: 21-24.

Li, W.C., D.E. Armstrong, J.D.H. Williams, R.F. Harris, and
J.K. Syers. 1972. Rate and extent of inorganic phosphate
exchange in lake sediments. Soil Sci. Amer. Proc.,

36: 279-285.

71

Lindsay, W.L., and E.C. Moreno. 1960. Phosphate phase equili-
bria in soils. Soil Sc1.Soc. Amer. Proc., 4:

177-182.

Mayer, L.M., and S.P. Gloss. 1980. Buffering of silica and
phosphate in a turbid river. Limnol. Oceanogr., 25:
12-22 0

McCallister, D.L., and T.J. Logan. 1978. Phosphate adsorption-
desorption characteristics of soils and bottom sediments
in the Maumee River Basin of Ohio. J. Environ. Qual.,

7: 87-92.

Moreno, E.C., W.E. Brown, G. Osbern. 1960. Stability of di-
calcium phosphate dihydrate in aqueous solutions and
solubility of octacalcium phosphate. Soil SCi- SOC-
Amer. Proc., 24: 94-98.

Mortimer, C.H. 1941. The exchange of dissolved substances
between mud and water in lakes (Parts I and II). J.
Ecol., 29: 280-329.

Mortimer, C.H. 1942. The exchange of dissolved substances
between mud and water in lakes (Parts III, IV, summary,
and references). J. Ecol., 30: 147-201.

Mortimer, C.H. 1971. Chemical exchanges between sediments
and water in the Great Lakes - speculation on probable
regulatory mechanisms. Limnol. Oceanogr., 16: 387-404.

Oloya, T.O., and T.J. Logan. 1980. Phosphate desorption from
soils and sediments with varying levels of extractable
phosphate. J. Environ. Qual., 9: 526-531.

Olsen, S.R., and F.S. Wateranabe. 1957. A method to determine
phosphorus adsorption maximum of soils as measured by
the Langmuir isotherm. Soil Sci. Soc. Amer. Proc.,

21: 144-149.

Pomeroy, L.R., E.E. Smith, and C.M. Grant. 1965. The exchange
of phosphate between estuarine water and sediments.
Limnol. Oceanogr., 10: 167-172.

Ponnamperuma, F.N. E.M. Tianco, and T. Loy. 1967. Redox
equilibria in flooded soils : I. The iron hydroxide
system. Soil Sci., 103: 374-382.

Ruttner, F. 1963. Fundamentals of Limnology. (Translat.
D.G. Frey and F.E.J. Fry.) Toronto, Univeristy of
Toronto Press, 295 pp.

Ryden, J.C., J.K. Syers, R.F. Harris. 1973. Phosphorus in
runoff and streams. Adv. in Agron., 25: 1-45.

72
Shapiro, J. 1973.

Determination of dissolved reactive phosphate
in natural waters.

In E.J. Griffith, A. Beeton, J.M.
Spencer, D.T. Mitchell, eds. Environmental Phosphorus
Handbook.

New York, Wiley-Interscience.
Shapiro, R.E. and M. Fried. 1959.

Relative release and re-
tentiveness of soil phosphates. Soil Soc. Amer.
23: 195-198.

357 pp.

PIOC. ,

1971. Sorption of inorganic phosphorus
by lake sediments. Soil Sci. Soc. Amer. Proc.,
35: 244-249.

Skukla, S.S., J.K. Syers, J.D.H. Williams, D.E. Armstrong,
and R.F. Harris.

Sommers, L.E., R.F. Harris, J.D.H. Williams, D.E. Armstrong,
and J.K. Syers. 1970. Determination of total inorganic
phosphorus in lake sediments. Limnol. Oceanogr., 15:
301-304.

Sonzoni, W.C., S.C. Chapra, D.E. Armstrong, and T.J. Logan.
1981.

Bioavailability of phosphorus imputs to lakes:
Significance to management Great Lakes Environmental
Planning Study, Contribution No. 40., 30 pp.

Stumm, W., and J.O. Leckie.

1971. Phosphate exchange with
sediments; Its role in the productivity of surface waters.
London, Proc.

5th Int. Wat. Pollut. Res. Conf. Pergamon
Press. III, 26: 1-16.

Stumm, W., and J.J. Morgan.

1981. Aquatic Chemistry. An
Introduction Emphasing Chemical Equilibria in Natural
Waters.

Second Edition. New York, Wiley-Interscience,
780 pp.

Syers, J.K., R.F. Harris, and D.E. Armstrong. 1973.

Phosphate
Chemistry in lake sediments. J. Environ. Qual., 2:
1-14 0

Taylor, A.W., and H.M. Konishi. 1971. Phosphate equilibria on
stream sediment and soil in a watershed draining an
agricultural region. J. Agr. Food Chem., 19:
Taylor, A.W., and H.M. Kunishi.

827-831.
1974.
phosphates from wastewater.

Soil adsorption of

In factors Involved in Land
Application of Agricultural and Municipal Wastes.
Beltsville, Maryland, U.S. Department of Agric.,

195 pp.
Tinoco, I., K. Saucer, and J.C. Wang. 1978.

Physical Chemistry.

Principles and Applications in Biological Sciences.

Englewood Cliffs, N.J., Prentice-Hall,
Turner, R.C. and J.S. Clark. 1956.

Inc., 624 pp.
The pH of calcareous soils.
Soil Sci., 82: 337-341.

73

United States Environmental Protection Agency. 1979. Methods
for Chemical Analysis of Water and Wastes. Environ.
Monitoring and Support Lab. (O.R.D.), Cincinnati, Ohio,
EPA-600 4-79-020.

Vollenweider, R.A. and J.J. Kerekes. 1980. Background and
summary results of OECD cooperative program on eutro-
phication. pp. 25-36. Proc. International symposium
on inland waters and lake restoration. Washington, D.C.,
U.S. EPA 440/5-81.

Wetzel, R.G. 1970. Recent and postglacial production rates of
a marl lake. Limnol. Oceanogr., 15: 491-503.

Wetzel, R.G. 1971. The role of carbon in hard-water marl lakes.
In G.E. Likens, ed. Nutrients and Eutrophication:
The Limiting-Nutrient Controversy. Special Symposium,
Amer. Soc. Limnol. Oceanogr., 1: 84-91.

Wetzel, R.G. 1975. Limnology. Philadelphia, W.B. Saunders
C0,, 743 pp.

Wetzel, R.G., P.H. Rich, M.C. Miller, and H.L. Allen. 1972.
Metabolism of dissolved and particulate detrital carbon
in a temperate hard-water lake. Mem. Ist. Ital.
Idrobiol., 29: 185-243.

Wier, D.R., and C.A. Black. 1968. Mineralization of organic
phosphorus in soils as affected by addition of inorganic
phosphorus. Soil Sci. Soc. Amer. Proc., 32: 51-55.

Williams, J.D., J.K. Syers, R.F. Harris, and D.E. Armstrong.
1970. Adsorption and desorption of inorganic phosphate
by lake sediments. Soil Sci. Soc. Amer. Proc., 35:
250-255.

Williams, J.D. J.K. Syers, S.S. Shukla, R.F. Harris, and D.E.
Armstrong. 1971. Levels of inorganic and total phosphorus
in lake sediments as related to other sediment parameters.
Environ. Sci. Technol., 5: 1113-1120.

Williams, J.D., H. Shear, and R.L. Thomas. 1980. Availability
to Scenedesmus quadricauda of different forms of phos-
phorus in sedimentary materials from the Great Lakes.
Limnol. Oceanogr., 25: 1-11.

 

Winter, T.C. 1978. Ground-water component of lake water and
nutrient budgets. Verh. Int. Ver. Limnol., 20: 438-444.

Woodruff, J.R. and E.J. Kamprath. 1965. Phosphorus adsorption
maximum as measured by the Langmuir isotherm and its
relationship to phosphorus availability. Soil Sci.

Soc. Amer. Proc., 29: 148-150.

74

Appendix 1. Effects of solution arsenate on phosphate
. determinations and effectiveness of arsenate
reductant in eliminating interference.

75

.H on Hmnwm UwEDwmm

om 0cm e0m< How mucmfiofimmmoo xufl>fluo¢ .H

 

 

 

om.vH mH.H mo.m no.m
oo.~m oo.H H~.H Ho.a
oa.~m v>.o om.o Hm.o
I wH.o I oo.o mm.a
oo.v mm.o mm.v~ ho.m
oo.¢a mN.o vm.v Hm.H
oH.v~ mH.o m¢.N Hm.o
t no.0 i oo.o mm.o
«om say a 403 2: son 22v 402 21c
mmmouocH m pmuommoz ta vom 23v coflumnucmocoo :oflumuucwocou
w CH wmmmHOCH mfium m m<
H.mc0wumummmum Avommmxv pumocmum ca mumnmmozm mo mcoHumcfiE
unmump so Ao~m5.v0m¢mmmz mm cowboy mumcmmum mo mcoflufioom pmuomwme mo Doommm .Hi< magma

76

Figure A-l. Percent over-estimation of phosphate concentration
as a function of the ratio of phosphate to
arsenate in solution.

 

 

77

¢N mm mm .N ON m.

h.

w.

.....02 237.16.. 5:.

n. V. n.

N.

0.

mo»

#IO

‘1'

no

 

d

J

[ml

1 a u

0.
ON
on
o¢
on
ow
ON
00
om

 

uo-ginnueauog owquoqd JO uogwwusadano %

78

 

 

Table A-2. Effects of addition of arsenate reductant on
determinations of phosphate in distilled water
solutions containing measured additions of
phosphate and arsenate.

Measured P04
Concentration Concentration
(M 1-1) (M 1'1)
w/o with 1
P04 AsO4 Reductant Reductant
0.00 0.00 0.00 0.00
0.33 0.03 0.00
1.33 0.16 0.00
0.81 0.00 0.84 0.74
0.33 1.00 0.77
1.33 1.55 0.77
1.61 0.00 1.61 1.58
0.33 1.84 1.58
1.33 2.68 1.71
8.07 0.00 8.07 8.07
0.33 8.39 8.07
1.33 9.29 8.03

 

l. 5 ml of arsenate reductant containing Na28205 and
Na252O3-5H20 were added to samples and standards
(cf. Johnson, 1971).

79

Appendix 2. Estimates of phosphorus, iron, and calcium added
to fine sediment particulates during the evaporation
and drying treatment.

80

.Hzon H How Do omm um cowumoneoo mcfl30a
IHom umoH mm3 unmfim3 who unmfiflpmm moan mo wmm mamumE«xoumm£ .usmflm3 wup unmafipmw
maons mo wow mmz AsmumEmao a“ E: bmmvv cofluomum unmaflomm mcflm mo unmflmz who .H

 

 

mmomfl made some whose was ssfloamo Hmuoe
nm.m~ «m.m mo.m mw.mH am.o cosH Hmuoe
vm.aa mm.oa ¢H.m m~.> me.o macromona sauce
oa.m mm.m am.a ~¢.~ GH.o mumnmmosm o>fluommm
121V 12:3 1 na 21c Axuc land 21V ucmnoosoo mansflom
Hmuoa ucmEfipmm no omHHou Hmumz mxmq

 

.HHMHHODME newswpmm mo 6 mm mo AUoom umv mcflhuo mCMMSU wmuflmommw ucmcomeoo

comm mo unsoEm may mo mmumeflumm uoommummu mosam> mcflccommwunoo .ucmefloom

mcflm mcfimmmooum :a com: Hmum3 mme can mcofluomaaoo ucmfifipmm 3oHHmnm cfl
Edfloamo com .couw .mumnmmonm Hmuou pom m>wuommn OHQSHOm mo cowumnuomocou .m|¢ canoe

81

Appendix 3. Total iron content of filtrate following exposure
of sediment particulates to different initial
levels of ambient phosphate.

82

Table A-4. Filterable iron (uM 1'1) in lake water of
different initial phosphate levels following
0.5, 2.0, and 48 hours exposure to whole sedi-
mentsl. Lake water, prin to sediment addition,
contained 0.734 uM Fe 1' .

 

Initial Lake Water Duration of Exposure (hours)
)

 

Phosphate (UM 1’ 0.5 2.0 48
0.13 0.895 0.985 0.985
1.52 0.698 0.985 0.842
6.23 0.734 0.842 0.842
31.29 0.734 0.842 0.806
122.30 0.734 0.842 0.734
302.90 0.859 0.895 0.698

 

1. Estimated contribution of filterable iron from aqueous
of whole sediment is 0.1 uM Fe 1'1.

83

Table A-5. Filterable iron (uM 1'1) in lake water of
different initial phosphate levels following
0.5, 2.0, and 48 hours exposure to fine sedi-
ments. Lake water prior to sediment addition
contained 0.342 uM Fe 1‘1.

 

Initial Lake Water Duration of Exposure (hours)
)

 

Phosphate (UM 1' 0.5 2.0 48
0.07 0.985 0.734 0.842
1.58 0.913 0.985 0.734
6.13 0.842 0.842 0.985
32.26 0.842 0.734 0.842
122.30 0.985 0.842 0.734
302.60 0.967 0.842 0.842

 

84

Appendix 4. Solubility and dissociation expressions, and
calculations used in the development of the
calcium phosphate solubility diagram of Figure 10.

85

 

 

.omma .ocmuoz pom >mm©cflq .mo .H
oo.¢H mom + mm mm Omm momma
mm.NH vommm I ¢Omm + an mmm Imvomm mvmcmmocm :mmoup>£ocoz
om.h wommmm I wanna + mm mmm Iwommm mumsmmonm :omoupxcflo
~H.~ eommmm I wommmm + no Hem eommm neon oflnonmnoee
oe.nHH zoom + «anew + nosed mm mimoceieoneoano whennensxonesm
Hm.m¢ wommm + mm + come Mm Ommm.mavomvmvmo momnmmozm Esfioamomuoo
om.e vomen + mom mm o~m~.¢ommno whenesefle

mumnmmonm EDHOHMOHQ
axm :oflmmmsmxm auflawnoaom manfiuom HMOflEmsu pcsomEou

 

.Hoa wusmfim mo Emummflo muHHHQSHOm
on» no unmEQon>mp may :a com: 00mm um nouns cam mmwoomm owom canonmmosm mo
mucmumcoo coaumfi00mmwp cam mpcoomaoo mumnmmonm Esfloamo mo muosoonm wuflaanoaom .mIm manna

86

CALCIUM PHOSPHATE SOLUBILITY CALCULATIONS

The following calculations of soluble reactive phosphate
concentration in equilibrium with calcium phosphate compounds
at pH 7 and 25°C illustrate the development of Figure 10.
Here, calcium concentration was taken as 1 x 10"3 molar, and

ionic strength as 3 x 10"3

M. Activity coefficients, y , were
calculated from:
log y = -A zz/u

where A is a Debye-Huckel constant equal to 0.5085 at 250C,

and z is the charge of the ion of interest. Accordingly, yCa2+
= 0.774, ynzpo4’ = 0.938, yHPo42' = 0.774, and ypo43‘ = 0.561.
'p' indicates negative log of the activity. For example, if
Ca2+ concentration = l x 10.3 M, Ca2+ activity = 1 x 10"3 (0.774)
= 7.74 x 10 '4; and pCa = 3.11.
A. Dicalcium phosphate dihydrate - CaHPO4-2HZO
6.56 = pCa + pHPO4
6.56 = 3.11 + pHPO4
3.45 = pHPO4 . (of. Table A-6)
Equilibrium levels of H2PO4- in solution at pH 7 can be obtained
from:
7.20 = pH + pHPO4 - pH2PO4
7.20 = 7.00 + 3.45 - pH2P04
3.25 = pH2PO4
2- _ _ -4 2- -l - _
pHPO - 3.45 - 3.55 x 10 M HPO 1 , and pH PO -
4 4 2 4
3.25 = 5.62 x 10"4 M H p0 ‘ 1'1.

2 4

87

Molar concentration x y = activity; activity/y = molar concen-

 

 

tration.

inn activity (M 1-1) .1. concentration (M 1-1)
HPO42’ 3.55 x 10-4 0.774 4.59 x 10-4
H2P04- 5.62 x 10‘4 0.938 5.99 x 10"4
p(reactive phosphate concentration) = 3(3é99 x 10.4 + 4.59 x 10-4)

B. Octacalcium phosphate - Ca4H(PO4)3-3H20

46.91 + 4pCa + pH + 3pPO4
46.91 = 4(3.11) + 7.00 + 3ppo4
9.16 =

pPO4 .

Equlibrium levels of HPO42- can be obtained from:

12.32 = pH + pPO4 - pHPO4
12.32 = 7.00 + 9.16 - pHPO4
3.84 = pHPO4 .

(HZPO4-) can be obtained from:

 

 

 

 

7.20 = 7.00 + 3.84 — szPO4
3.64 = szPO4 .
ion activity (M 1-1) Y concentration (M 1-1)
PO43‘ 6.92 x 10-10 0.561 1.23 x 10'9
HPO42' 1.45 x 10-4 0.774 1.87 x 10-4
H2P04- 2.29 x 10-4 0.938. 2.44 x 10"4
p(reactive phosphate concentration) = 3.26 .
C. Hydroxyapatite - CalO(PO4)6(OH)2
ion activity (M 1-1) Y concentration (M 1-1)
HPO42' 3.55 x 10"4 0.774 4.59 x 10'4
H2P04- 5.62 x 10‘4 0.938 5.99 x 10‘4
p( reactive phosphate concentration) p(5.99 x 10"4 + 4.59 x 10-4)

2.98

"1111111111"1111111111“