EPA-600/3-81-001
January 1981
PROJECT SUMMARY
MODIFICATIONS OF MODELS PREDICTING TROPHIC STATE OF LAKES
Adjustment of Models to Account for the
Biological Manifestations of Nutrients
by
S. C. Hern, V. W. Lambou, L. R. Williams
and W. D. Taylor
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
LAS VEGAS, NEVADA 89114
-------
ABSTRACT
The strong relationship between total phosphorus and phytoplankton bio-
mass in lakes has been clearly confirmed by researchers. What is now needed
to predict algal biomass for making better management decisions for individual
lakes is a quantitative understanding of the range in biomass (as measured by
chlorophyll a) per unit of phosphorus. This range extends over several orders
of magnitude.
To determine the environmental factors affecting the response of phyto-
plankton chlorophyll a. to total phosphorus concentration, we analyzed data
collected from 757 U.S. lakes. We found that light attenuation from inter-
ferences not related to chlorophyll .§_ can dramatically affect the quantity of
phytoplankton biomass in many U.S. lakes. The ratio of biologically avail-
able phosphorus to nitrogen is, 1n some cases, an important factor in
determining the amount of chlorophyll a_ produced per unit of phosphorus
present.
This report presents methods to modify nutrient ambient- and loading-
models that predict the trophic state of lakes to:
1. change the trophic classification based on an ambient total
phosphorus level to one based on the biological manifest-
ation of nutrients as measured by chlorophyll a^
2. allow determination of the critical levels of phosphorus
that will result in unacceptable levels of chlorophyll a_, and
3. account for the unique characteristics of a lake that
affect the amount of chlorophyll .a produced per unit of
phosphorus.
We found that, if chlorophyll a. were used as the trophic classification
criterion rather than total phosphorus, many U.S. lakes would be classified
lower, i.e., less eutrophic.
-------
INTRODUCTION
Phosphorus supply is considered to be the primary determinant of algal
community biomass and production in most temperate zone lakes. Researchers
clearly confirm the strong correlation between total phosphorus and phyto-
plankton biomass (as measured by chlorophyll a). What is presently needed to
predict algae biomass for management decisions on individual lakes is a
quantitative understanding of the range in biomass (chlorophyll .a) per unit
of phosphorus; this range extends over orders of magnitude.
Existing models predict ambient total phosphorus lake concentration from
tributary phosphorus loading data. These and similar models are widely used
to indicate the degree of eutrophy of lakes and evaluate the environmental
effects of ambient phosphorus levels in lakes and, accordingly, to make
decisions about lake management, e.g., the manipulation of ambient total
phosphorus concentration to produce a desired environmental effect. All these
models use levels of 10 and 20 micrograms per liter of ambient phosphorus con-
centrations to divide lakes into three standard trophic classifications
oligotrophic, mesotrophic, and eutrophicon the assumption that the relation-
ship of phytoplankton biomass to phosphorus is the same for all lakes.
However, the use and incorporation of phosphorus into phytoplankton bio-
mass varies significantly from lake to lake. The efficiency of use of phos-
phorus is largely dependent on the availability of light, sufficient supply
of other nutrients, and biological availability of the various phosphorus
species.
The use of these models to predict phytoplankton biomass from actual or
potential phosphorus concentrations in individual bodies of water could lead
to faulty management decisions if the factors affecting the use of phosphorus
are not taken into consideration. This study evaluates the factors affecting
the relationship of phytoplankton biomass to phosphorus levels and shows how
to modify models predicting the trophic state of lakes to take these factors
into account. The data base used to evaluate the factors was derived from
data collected in the National Eutrophication Survey during the spring, sum-
mer and fall of 1972 through 1975, and involved 757 of the lakes surveyed
throughout the 48 conterminous States.
-------
CONCLUSIONS
Previous workers have established a strong relationship between CHLA
(a measure of phytoplankton biomass) and total phosphorus in lakes. They
report an extremely high log-log product moment correlation coefficients
ranging from 0.85 to 0.98 between chlorophyll a. and total phosphorus in lakes.
The implication of these findings is that phosphorus is the element that
controls algal biomass. However, v/e believe that the relationship between
chlorophyll a_ and total phosphorus described by these workers represents the
situation under nearly ideal conditions, i.e., without major interferences.
In our study of 757 U.S. lakes, a log-log product moment correlation
coefficient of only 0.60 was found between chlorophyll a. and total phosphorus,
and the response ratio (i.e., the amount of chlorophyll a_ produced per unit
of total phosphorus) was found to vary greatly. We concluded that many U.S.
lakes do not reach maximum production of chlorophyll a^ because of interference
factors.
Interference factors that may prevent phyotplankton chlorophyll a. from
achieving maximum theoretical concentrations based upon ambient total phos-
phorus levels in a lake include availability of light, limitation of growth
factors other than total phosphorus components, domination of the aquatic
flora by vascular plants rather than phytoplankton, short hydraulic retention
time, and the presence of toxic substances. We found that light attenuation
from non-chlorophyll ji. related interferences can dramatically affect the
quantity of phytoplankton biomass present in lakes. Also, in some cases, the
ratio of biologically available phosphorus to nitrogen is an important factor
in determining the amount of chlorophyll a^ produced per unit of total phos-
phorus, while temperature is not an important factor.
te
Most of the trophic classification schemes for lakes use nutrient levels
rather than the biological manifestation of nutrients as measured by chloro-
phyll £ as the basis of classification. We found that, when classified on
the basis of chlorophyll a. rather than total phosphorus, 25 percent of the
757 lakes used in the study were classified lower, i.e., less eutrophic. If,
in fact, the large population of lakes used in this study is representative
of conditions throughout the U.S., and if the manifestations of nutrients
-------
rather than their absolute concentrations are the primary criteria for
beneficial water use, many communities could be spared the burden of costly
nutrient-removal programs suggested by phosphorus-based trophic classifi-
cations.
We developed methods to modify loading and ambient models that predict
the trophic state of lakes to (1) change the trophic classification based on
an ambient total phosphorus level to one based on the biological manifestation
of nutrients as measured by chlorophyll a^ (2) determine the critical levels
of total phosphorus which will result in an unacceptable level of chlorophyll
a^ so that the level of total phosphorus can be manipulated to achieve the
desired use of a given waterbody; and (3) account for the unique character-
istics of a lake that affect the amount of chlorophyll a. produced per unit of
total phosphorus.
RECOMMENDATIONS
The commonly used ambient and loading models predict the trophic state
of a lake from total phosphorus levels and assume that all lakes will respond
in the same manner to a given ambient total phosphorus concentration. Because
non-chlorophyll a. light interferences and other interferences in many U.S.
lakes significantly decrease the amount of chlorophyll a_ produced per unit of
total phosphorus, and since excessive algal growth or other manifestations of
nutrients are more important from a water quality standpoint than a trophic
classifiaction based on an arbitrary ambient total phosphorus level, we
recommend that ambient and loading models predicting trophic state be adjusted
to account for the amount of chlorophyll a^ produced per unit of total phos-
phorus in a lake.
-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring and Support
Laboratory-Las Vegas, U.S. Environmental Protection Agency, and approved for
publication. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
11
-------
CONTENTS
Page
Figures iv
Tables v
Abbreviations vi
Acknowledgment ix
Introduction 1
Conclusions 2
Recommendation 3
Materials and Methods 4
Lake selection 4
Field sampling methods 4
Nutrient analyses 7
Data management 7
Relationship of Phytoplankton Biomass to Phosphorus 11
Factors Affecting Response Ratios 16
Light 16
Nitrogen to phosphorus ratios 21
Temperature 24
Modification of Models to Predict Trophic State 25
Loading models 25
Ambient models 31
Results from using modified models 32
Literature Cited 36
iii
-------
FIGURES
Number Page
1. Number and location of lakes sampled during the years
indicated 5
2. The relationship between RA and TP in lakes based on Jones
and Bachmann's (1976) regression equation 13
3. Summer RA for 757 U.S. lakes 14
4. Plot of summer log TP and log CHLA for 757 U.S. lakes 15
5. Relationship of summer CHLA to summer SO for U.S. lakes 17
6. The relationship of summer RS to RA 19
7. The relationship between summer log RA and log TP based on
Jones and Bachmann's (1976) regression equation 27
8. Trophic classifications based on summer TP and CHLA
criteria 34
iv
-------
TABLES
Number Page
1. Seasonal Sampling Periods for 1972 through 1975 6
2. Lake Water Sample Analyses Summary 7
3. Analytical Methods and Precision of Laboratory Analyses 8
4. Reported Relationships Between CHLA and TP in Lakes 12
5. RA by Season 12
6. Mean Summer Parameter Values by Non-CHLA Light
Interference and RA Groups 20
7. Number and Percent of Lakes Classified According to
Summer N/P Ratios as Nitrogen-Limited, Transitional,
and Phosphorous-Limited by Non-CHLA Light Interference
and RA Group 23
8. The Relationship between Log T and Log RA by Season
in U.S. Lakes . . 24
9. Number of U.S. Lakes Sampled by NES Classified as
Oligotrophic, Mesotrophic, and Eutrophic by Summer
CHLA and TP Criteria 33
-------
LIST OF ABBREVIATIONS
PARAMETER ABBREVIATIONS
CHLA Chlorophyll a.
DP Dissolved Phosphorus
IN Ammonia-nitrite-nitrate-N
KN Total Kjeldahl-N
N/P Ratio of IN to TP
NO ~ Nitrite-nitrate-N
NH Ammonia-N
OP Dissolved orthophosphorus
RA Response ratio
SO Secchi disk
T Temperature
TP Total phosphorus
MODEL ABBREVIATIONS
A --- 4.77
AERA the mean summer RA for the lake corrected to what it Would
be at the 20 wg/1 level of TP, i.e., the ambient
eutrophic level
the mean summer RA for the lake corrected to what it would
be at the 10 yg/1 level of TP, i.e., the ambient
mesotrophic level
minimum mean ambient summer lake TP concentration
which will cause the lake to become eutrophic corrected to
account for the lake's RA
-------
-- the minimum mean ambient summer lake TP concentration
which will cause the lake to become mesotrophic corrected
to account for the lake's RA
ATPg the minimum mean ambient summer lake TP concentration
which will cause the lake to become eutrophic
the minimum mean ambient summer lake TP concentration
which will cause the lake to become mesotrophic
B -- -8
ERA a constant equal to 0.32 which is the RA predicted from 20
ug/1 of ambient TP
ETP a constant equal to 20, which is the theoretical minimum
ambient concentration of TP in ug/1 which will result in
eutrophic conditions in a lake and is the level which if not
equaled will result in meso- or oligotrophic conditions
HT hydraulic residence time
L phosphorus areal loading
MRA a constant equal to 0.23 which is the RA predicted from 10
yg/1 of ambient TP
MTP a constant equal to 10, which is the theoretical minimum
ambient concentration of TP in ug/1 which will result in
mesotrophic conditions in a lake and is the level which if
not equaled will result in oligotrophic conditions
OCHLA observed mean summer CHLA in a lake
ORA observed summer ambient RA in the lake
OTP -- observed summer ambient TP in the lake
OTTP observed mean tributary TP
OS observed SD value for a given lake
P hydraulic flushing rate
PS predicted SD
the minimum mean tributary TP concentrations in ug/1
which will cause a lake to be eutrophic at equilibrium
corrected to account for the lake's RA
TP£ the minimum mean tributary TP concentration in ug/1
which will cause a lake to be eutrophic
vii
-------
-- the minimum mean tributary TP concentration in ug/1
which will cause a lake to be mesotrophic at equilibrium
corrected to account for the lake's RA
the minimum mean tributary TP concentration in yg/1
which will cause a lake to be mesotrophic
R retention coefficient for phosphorus
RS residual SD, an index of non-CHLA-related light attenuation
Z -- mean depth
vili
-------
ACKNOWLEDGMENT
This work was partially supported by the Environmental and Water Quality
Operational Studies Program, Environmental Laboratory, U.S. Army Engineers
Waterways Experiment Station through Interagency Agreement No.
EPA-IAG-78-X0393 (Waterways Experiment Station Agreement No. WES-78-12).
1x
-------
INTRODUCTION
Phosphorus supply is considered to be the primary determinant of algal
community biomass and production in most temperate zone lakes (Schindler 1974,
Williams et al. 1977, Lambou et al. 1976, Dillon and Rigler 1974, Bachmann and
Jones 1974, Edmondson 1970). Kalff and Knoechel (1978) state that the strong
relationship between total phosphorus and phytoplankton biomass (chlorophyll
a) has been clearly confirmed and what is presently needed in order to predict
algal biomass for management decisions on individual lakes is a quantitative
understanding of the range in biomass (chlorophyll a) per unit of phosphorus
which extends over orders of magnitude.
Vollenweider (1968), Dillon (1975), and Larsen and Mercier (1976) have
developed models to predict ambient total phosphorus lake concentration from
tributary phosphorus loading data. These and other similar models are widely
accepted and used to indicate the degree of eutrophy of lakes and evaluate the
environmental effects of ambient phosphorus levels in lakes. These models are
currently being used to make decisions about lake management, e.g., the
manipulation of ambient total phosphorus concentration to produce a desired
environmental effect. All of the commonly used models utilize levels of 10
ug/1 of ambient phosphorus concentrations to divide lakes into three standard
trophic classificationsoligotrophic, mesotrophic, and eutrophic. The
assumption is made in the use of these somewhat arbitrary levels of 10 and 20
ug/1 of ambient lake phosphorus concentrations that the relationship of
phytoplankton biomass to phosphorus is the same for all lakes. However, the
utilization and incorporation of phosphorus into phytoplankton biomass varies
significantly from lake to lake. The efficiency of utilization of phosphorus
is largely dependent upon the availability of light, sufficient supply of
other macronutrients (e.g., nitrogen) and micronutrients, and biological
availability of the various phosphorus species.
The use of models to predict phytoplankton biomass from actual or
potential phosphorus concentrations in individual bodies of water could lead
to faulty management decisions if the factors affecting the utilization of
phosphorus are not taken into consideration. The purpose of this paper is to
evaluate the factors affecting the relationship of phytoplankton biomass to
phosphorus levels and to show how to modify models predicting the trophic
state of lakes to take these factors into account.
-------
CONCLUSIONS
1) The response ratio, i.e., the amount of chlorophyll ji produced per unit of
total phosphorus, varies greatly in United States lakes, even though there
is a strong correlation between total phosphorus and phytoplankton biomass
(as measured by chlorophyll aj in lakes.
2) Many United States lakes do not reach maximum production of chlorophyll ^
because of interference factors.
3) Light attenuation from'non-chlorophyll j^ related interferences can
dramatically affect the quantity of phytopla'nkton biomass present in
United States lakes.
4) The ratio of biologically available phosphorus to nitrogen is in some
cases an important factor in determining the response ratio in United
States lakes.
5) Temperature does not seem to be an important factor influencing the
response ratio in United States lakes.
6) Nutrient loading and ambient models which predict the trophic state of
lakes can be modified to: a) change the trophic classification based on
an ambient total phosphorus level to one based on the biological
manifestation of nutrients as measured by chlorophyll a^, b) determine the
"critical levels of total phosphorus which will result in unacceptable
levels of chlorophyll ji so that the level of total phosphorus can be
manipulated to achieve the desired use of a given water body and
c) account for the unique characteristics of a lake which affects the
amount of chlorophyll a^ which will be produced per unit of total
phosphorus.
7} Many United States lakes would be classified lower (i.e., less eutrophic)
if the biological manifestation of nutrients as measured by chlorphyll a^
is used as the basis of classification rather than total phosphorus.
8} If the manifestations of nutrients, rather than their absolute levels, are
considered the primary criteria for beneficial water use, many communities
could be spared the unnecessary burden of costly nutrient-removal programs
suggested by phosphorus-based trophic classifications.
-------
RECOMMENDATION
It is recommended that loading and ambient models predicting trophic state
be adjusted to account for the amount of chlorophyll a produced per unit of
total phosphorus in a lake because (1) non-chlorophylT £ light interferences
and other interferences in many United States lakes significantly decrease the
amount of chlorophyll a_ produced per unit of total phosphorus and (2)
excessive algal growth or other manifestations of nutrient enrichment are more
important from a water quality standpoint than a trophic classification based
on an arbitrary ambient total phosphorus level.
-------
MATERIALS AND METHODS
The data base used to evaluate the factors affecting the relationship of
phytoplankton biomass to phosphorus levels came from the 815 lakes (Figure 1)
sampled throughout the 48 contiguous United States during the spring, summer,
and fall of 1972 through 1975 by the National Eutrophication Survey (NES).
LAKE SELECTION
Lakes included in the NES were selected through discussions with State
water pollution agency personnel and U.S. Environmental Protection Agency
Regional Offices (U.S. EPA 1975). Screening and selection of lakes sampled in
1972 and 1973 strongly emphasized lakes with actual or potential accelerated
eutrophication problems. As a result, the selection was generally limited to
1akes:
(1) impacted by one or more municipal sewage treatment plant outfalls
either directly into the lake or by discharge to an inlet tributary
within approximately 40 kilometers of the lake;
(2) 40 hectares or larger in size, and
(3) with a mean hydraulic retention time of at least 30 days.
Specific selection criteria were waived for some lakes of particular interest
to the individual states. During 1974 and 1975, lakes were selected to
broaden the spectrum of water quality types represented (U.S. EPA 1975).
FIELD SAMPLING METHODS
Lake sampling was accomplished by two sampling teams, each consisting of a
limnologist, pilot, and sampling technician, operating from pontoon-equipped
Bell UH-1H helicopters. A mobile field laboratory provided analytical
support. With-few exceptions, each lake was sampled under spring, summer, and
fall conditions. The sampling periods for each season are given by sampling
year in Table 1.
Sampling site locations were chosen to define the character of the lake as
a whole and to investigate visible or known problem areas, e.g., algal blooms,
and sediment or effluent plumes (U.S. EPA 1975). The number of sites was
limited by the survey nature of the program and varied in accordance with lake
-------
0)
>
3
CO
c
§o
**3
(0
a
2
LJLJ
"3
c
03
o
!-
o
1.
(Q
O)
>>
(V
4->
0>
i"-
L.
3
o
o
o
01
0)
c
o
-------
TABLE 1. SEASONAL SAMPLING PERIODS FOR 1972 THROUGH 1975
Sampling
Year
1972
1973
1974
1975
Number of
Lakes
231
250
180
154
Spring
5/20-7/19
3/7-7/11
3/4-5/12
2/21-6/10
Season
Summer
7/20-9/24
7/5-9/18
5/13-9/5
6/11-9/3
Fall
9/15-11/16
9/19-11/4
9/11-11/14
9/4-12/11
size, morphological and hydrological complexity and practical considerations
of time, flight range, and weather.
The Survey helicopters were equipped with electric winches and
approximately 60 meters (m) of hollow-core, multi-conductor cable attached to
a submersible pump and an InterOcean Systems sensor package capable of making
in situ measurements of conductivity, temperature, optical transmissivity and
depth. Rack-mounted equipment located inside the helicopter provided analog
recording capabilities and a digital display of the sensor values. An echo
sounder, Secchi disk, sample containers and related equipment items necessary
for water sampling were also carried in each helicopter (U.S. EPA 1975).
After landing at the station the sensor-pump package was used to collect
data and water samples at various depths. Sampling depths were selected to
characterize the water column after inspection of analog strip-chart records
generated during the initial pass of the sensor package through the water
column.
At each sampling depth, water samples were collected for nutrient,
alkalinity, pH and dissolved oxygen determinations. For nutrient analyses,
two 130 milliliter (ml) polyethylene sample bottles were filled and sent to
the Environmental Monitoring and Systems Laboratory-Las Vegas (EMSL-LV) for
analysis.
At each sampling station, a chlorophyll a_ and a phytoplankton sample were
taken. These depth-integrated samples were uniform mixtures of water from the
surface to a depth of 4.6 m (15 feet) or from the surface to the lower limit
of the photic zone, representing 1 percent of surface incident light,
whichever was greater. If the depth at the sampling station was less than 4.6
m, the sample was taken from just off the bottom to the surface. Table 2
summarizes sample collection and handling procedures.
-------
TABLE 2. LAKE WATER SAMPLE ANALYSES SUMMARY*
Parameters
Temperature
Chlorophyll a_
Total phosphorus
Total Kjeldahl-N
Nitrite-Nitrate-N,
Amonia-N
Sample
Volume
4-ounce
(130 ml)'
4-ounce
(130 ml)
4-ounce
(130 ml)
Field
Treatment
Refrigerated,
dark
Unfiltered,
HgCl2
Filtered,
HgCl2
Where
Performed
Ijn situ
Field lab
EMSL-LV
EMSL-LV
Depth
Continuous
Photic zone
integration
Select levels
Select levels
*Modified from U.S. EPA (1975)
A Beckman Electromate portable pH meter and combination electrode were
used to determine pH as soon as possible following sample delivery to the
field laboratory at the end of the day's sampling. Dissolved oxygen samples
were titrated with phenylarsine oxide in the mobile field laboratory within 16
hours after collection. Chlorophyll a^ analyses were performed at the end of
each day in the trailer laboratory according to the fluorometric procedure
described by Yentsch and Menzel (1963). One of each pair of nutrient samples
was filtered through a type HA 0.45-micrometer Millipore membrane filter into
a clean, unused polyethylene bottle, recapped, and, along with its unfiltered
counterpart, forwarded to EMSL-LV for analysis.
NUTRIENT ANALYSES
The analytical methods utilized to process the samples at EMSL-LV are
outlined in Table 3. All of the analyses were performed utilizing adaptations
of automated procedures described in Mullins, et al. (1975).
DATA MANAGEMENT
The physical and chemical data were input to STORET (STOrage and
RETrieval), the U.S. EPA computer system which processes and maintains water
quality data. Data manipulation was as prescribed by Bliss, et al. (1975).
Basic calculations for parameters measured in sampled lakes were performed in
such a way as to give equal weight to: each depth sampled at a station; each
sampling station sampled on an individual lake during a sampling round; and
each sampling round on an individual lake during a sampling year. Mean
-------
TABLE 3. ANALYTICAL METHODS AND PRECISION OF LABORATORY ANALYSES
Parameter
Method
Precision
Total
Phosphorus
Nitrite-Nitrate-N
Ammonia-N
Kjeldahl-N
Persulfate oxidation followed by
colorimetric determination of
antimony-phosphomolybdate complex
5.0 yg/1 P or ± 5%
Cadmium reduction followed by the 10.0 ug/1 N or ± 5%
diazotization of sulfanilamide by
nitrite coupled with N-(l-naphthyl)-
ethylene diamine.
Alkaline phenol hypochlorite
reaction producing indcphenol
blue.
Acid digestion followed by the
above procedure for ammonia
nitrogen.
5.0 ug/1 N or ±
100.0 ug/1 N or ± 5%
parameter values for each sampling station were calculated as follows (Lambou,
et al. 1976):
Pari/D
(1)
where: Par- = mean value for a parameter at the j sampling station during
J a sampling round (lake-date).
Par.j = value for the i depth, and
D * the number of depths for which a parameter was measured at
the jtn sampling station during a sampling round.
Mean parameter values for each sampling round were calculated as follows:
Par
Paiys
(2)
1-1
8
-------
where: Par, = mean value for the k sampling round on a given lake, and
S = number of sampling sites.
Mean lake parameter values for a given sampling year were calculated as
fol1ows:
3
faT = J] faTk/3 (3)
k-1
where: Par = mean parameter value for a given sampling year.
Mean parameter values for a given sampling year were calculated only when
values were available for the first, second, and third sampling rounds.
Equations 1 and 2 were used to determine mean parameter values for total
phosphorus (TP), ammonia-N (NH), nitrite-nitrate-N (NO), ammonia-nitrite-
nitrate-N (IN), and total Kjeldahl-N (KN), all expressed in micrograms per
liter (yg/1); temperature (T) in degrees Celsius (°C), and Secchi disk (SD)
in m.
The ratio of IN/TP (N/P) for each sampling station was calculated as
fol 1 ows:
W - TN/TP (4)
However, at the Equation 1 level a total phosphorus value was deleted if
either nitrogen complement was missing. Round and yearly values were
calculated using Equations 2 and 3.
Unlike the above parameters measured at various depths, only one
chlorophyll ^ (CHLA) measurement was made at any individual sampling station
during a sampling round. Therefore,
CHLA. - CHLA (5)
J
where: CHLA. = the mean CHLA concentration in micrograms per liter (ug/1)
J for the jth sampling station during a sampling round, and
CHLA = the CHLA concentration in pg/1 for a depth-integrated water
sample*
-------
A response ratio (RA) i.e., the amount of CHLA per unit of TP was
calculated for each sampling round as follows:
(6)
_^^_i i. L
where: RA. = the response ratio for the k sampling round on a
given lake, while
CHLA. and TP. = the mean CHLA and TP for the kth sampling round
on a given lake.
10
-------
RELATIONSHIP OF PHYTOPLANKTON BIOMASS TO PHOSPHORUS
A strong relationship between CHLA (a measure of phytoplankton biomass)
and TP in lakes has been established by Jones and Bachmann (1976), Dillon and
Rigler (1974), Sakamoto (1966), and Carlson (1977). The log-log product
moment correlation coefficients reported ranged from 0.85 to 0.98 (Table 4).
The inplication of these findings is that phosphorus is the element that
controls algal biomass, however, the lakes utilized by the above authors in
deriving their relationships are largely phosphorus-limited. No appreciable
differences can be detected between the slope of regression equations (summer
CHLA regressed on summer .or spring TP) from lakes studied in Japan (Sakamoto
1966), Canada (Dillon and Rigler 1974) and North America (Jones and Bachmann
1976 and Carlson 1977). Because of the similarities between the various
regression equations derived from lakes studied throughout the world, we feel
that they accurately describe the relationship between CHLA and TP under
nearly ideal conditions, i.e., without major interferences.
The slopes of the regression lines from the four sources cited above are
all greater than 1.4 (Table 4) which indicates that proportional increases in
CHLA concentrations accrue at a faster rate than increases in TP concentra-
tions, i.e., the response ratio of CHLA to TP is greater at high concentra-
tions of TP than at low concentrations. This is also shown in Figure 2, where
the RA based on Jones and Bachmann's (1976) regression equation is presented.
However, the log-log regression equations derived from the lakes we sampled
suggest that RA decreases as total phosphorus increases (Table 5). Since
lake management decisions to control or manipulate phytoplankton biomass are
usually made on summer data and this was the period when we observed the
greatest potential response of the phytoplankton community to available
nutrients, only summer data are used throughout the remainder of this report.
A histogram of summer RA for 757 lakes and reservoirs located in the
conterminous United States is presented in Figure 3. Summer RA varied widely
among the various lakes ranging from 0.002 to 2.26 with a mean of 0.29
(Table 5).
Mean summer CHLA is plotted against mean summer TP for the 757 lakes we
studied (Figure 4). Superimposed on Figure 4 is the regression equation of
Jones and Bachmann (1976). The plots of the vast majority (80 percent) of
the 757 lakes shown in Figure 4 lie below the Jones and Bachmann regression
line. Of these, most plots occur at TP levels in excess of 20 micrograms per
liter. We believe that the reason most lakes do not reach maximum production
of CHLA, as described by the regression equation, is because of interference
factors.
11
-------
TABLE 4. REPORTED RELATIONSHIPS BETWEEN CHLA AND TP IN LAKES
Number of
Source Regression Equation* Lakes
Jones and
Bachmann (1976)t Log CHLA = -1
Dillon and
Rigler (1974)* Log CHLA = -1
Sakamoto (1966)*
(as calculated by
Dillon and Rigler) Log CHLA » -1
Carlson (1977)t Log CHLA - -1
This study t Log £HlA~ = -0
*CHLA and TP in ug/1
tSummer CHLA and summer TP
^Summer CHLA and spring TP
TABLE 5
RA
Mean
Standard deviation
Minimum value
Maximum value
Number of lakes
.09 + 1.46 Log TP 143
.136 + 1.45 Log TP 46
.13 + 1.58 Log TP 28
.06 + 1.45 Log TP 43
.11 + 0.64 Log ?F 757
. RA BY SEASON
Season*
Spring Summer
0.24 0.29
0.27 0.29
0.001 0.002
2.81 2.26
785 757
Log-Log
Product Moment
Correlation
Coefficient
r = 0.95
r = 0.95
r = 0.98
r = 0.85
r = 0.60
Fall
0.24
0.21
0.001
1.60
111
*A one way ANOVA and Student Newman Keul ^multiple range test indicated a
significant difference between summer values and fall and spring values at
the 0.05 level.
12
-------
c
(0
J=
O
(O
CO
0)
c
o
o
O)
-------
J"
1
J"
1
03
r-.
s-
o
I
o>
S § SSS 2 S 3 S 3
*n cw
r i ( r
14
-------
CJ
O1
o
o
<+-
<
CD
o
IS
to
C7>
O
O
CL.
*3-
O>
=3
V1K3
15
-------
FACTORS AFFECTING RESPONSE RATIOS
Factors which may prevent phytoplankton CHLA from achieving maximum
theoretical concentrations based upon ambient TP levels in a lake include:
1. availability of light;
2. limitation of growth by nutrients other than TP, e.g., nitrogen,
carbon, silica, etc.;
3. biological availability of the TP components;
4. domination of the aquatic flora by vascular plants rather than
phytoplankton;
5. temperature;
6. short hydraulic retention time; and
7. presence of toxic substances.
LIGHT
The relationship between SD and CHLA concentrations in lakes has been
described by a number of investigators (Edmondson 1970, Bachmann and Jones
1974, and Carlson 1977). After analyzing data from 147 lakes, Carlson (1977)
described a strong relationship (r = 0.93) between CHLA (yg/1) and SD (m),
i.e.:
In SD = 2.04 - (0.68 In CHLA) (7)
The lakes Carlson utilized were relatively free of non-chlorophyll-related
particles. The strength of the relationship Carlson described indicates that
light attenuation from any source other than CHLA-related particles was
homogenous in the population of lakes represented in his study.
A natural log transformation of our summer CHLA and SD resulted in a
correlation coefficient of only 0.56 (Figure 5). This indicates that light
attenuation due to non-CHLA-related suspensoids or dissolved color must be an
important factor in a number of lakes sampled in our study.
16
-------
s
O)
(13
«v
*b
1
CO
i.
O
C- Q.
O S
i- O
o u
o o
*"~
Q} C
c gj
5
I/) 0)
c en
on
CL c
T- O
c: s_
(U
C£
(U
-. * -
C3
to
cj
CM
I
(t/6rt) V1H3 "I
-------
If Carlson's regression equation accurately describes the relationship
between CHLA and SO in water bodies relatively free of non-CHLA-related
particles or dissolved color interferences, it can be utilized in conjunction
with observed CHLA concentrations and SD measurements to estimate light
attenuation due to non-CHLA-related particles or dissolved color. We did this
in the following manner:
RS = PS - OS (8)
where: RS = residual SD, an index of non-CHLA-related light attenuation;
PS = predicted SD, determined by substituting our mean ambient
summer CHLA value for a given lake into Equation 7; and
OS = the observed SD value for a given lake.
If the RS value is positive, non-CHLA-related suspensoids or color
interferences are expected. As light attenuation due to the non-CHLA-related
factors increases, RS values increase. For a RS value equal to 0, the light
attenuation is assumed to be mainly due to CHLA-related particles. We
calculated a mean summer RS value for each of our lakes by the method
presented above and plotted the values against the RA to examine the effects
of light interference upon the relationship of CHLA to TP (Figure 6).
Theoretically, if all of the light attenuation was due to CHLA-related
particles, then, according to Carlson's equation (Equation 7), no lake would
have a RS value below zero. Obviously, all lakes, including those
investigated by Carlson, have some light attenuation due to non-chlorophyll-
related particles and color interferences. Therefore, the true zero line
(where virtually all light interferences are due solely to CHLA-related
particles) is less than a RS value of zero. The true zero line should contain
lakes that extend over the full range of RA values, since many lakes are
relatively free of non-CHLA-related particles. An examination of Figure 6 and
a consideration of experimental error led us to the conclusion that the
population of lakes which extends over the full range of RA and lies between
RS values of 0.46 to -2.00 encompasses the true zero line. Since SD values
less than true zero are theoretically impossible, we believe the RS values
less than -2.00 in Figure 6 are due to experimental error, and are not used in
our analysis of the factors affecting RA's. With the above criteria, the
lakes can be divided into two groups by their RS values (Figure 6). One group
has RS values of >0.46 (high non-CHLA light interference} while the other
group has RS values between 0.46 and -2.00 (low non-CHLA light interference).
We divided the lakes into 3 RA groups: "very low" (0.00 to <0.50), "low"
(0.50 to <1.0), and "high" (H.O) (Figure 6). Our rationale for this
classification is that with a response of approximately 1.0 or above, all the
TP in the system is "theoretically" associated with cellular CHLA (Strickland
1960). Therefore, at a RA of 1.0 or greater, the algae are utilizing all
available phosphorus and incorporating it into cellular material. At a RA
18
-------
aouaj3j.jaq.UL
VIHO-iiou M6LH
*g s
g aouajaj.jaq.ui. q.u.5i
.0 VlHO-uou
~ to
c c
o
i. >
CU k»
c .e
o
CVJ
II u-
e= c
« L.
C V
O ^
rs C
a o
.C O
O 4-)
~4 C
II ^
ee
at
o
o
*
a a o o
u x
« « a a ~
! O« fc. fc.
I < WX MO
[U
u
fc. U I
hi «
U
w
o
< u
in
CSJ
Of
a>
a:
s_
LT)
O
(U
ai
-------
of 0.5, half of the TP is theoretically tied up in phytoplankton cells, while
the other half of the TP pool is extracellular. These divisions are somewhat
arbitrary, as the relationship between cellular phosphorus and CHLA can
dramatically change, depending upon environmental conditions and phytoplankton
community composition, e.g., luxury uptake by selected phytoplankton forms and
increased CHLA production per unit of biomass under low-light conditions. Even
though the divisions of the response ratio are somewhat arbitrary, they
provide a useful mechanism to compare major response groups.
Only 1 of 104 water bodies was found in the low or high response groups
which had a RS >0.46 m while 287 of 632 (45%) of the water bodies with a
very-low response had a RS >0.46 m. The low non-CHLA light interference,
very-low RA group had approximately double the mean TP concentration but
seven times the CHLA concentration as compared to the high non-CHLA light
interference group, very-low RA group (Table 6). This suggests that non-CHLA
light interference is an important factor in controlling the RA. The low
non-CHLA light interference very-low RA group had the highest mean TP
concentration (208 ug/1). The RS in the low non-CHLA light interference group
decreases from -0.19 for the very-low RA group to -0.35 and -0.64 for the low
and high RA groups, respectively (Table 6).
TABLE 6. MEAN SUMMER PARAMETER VALUES BY NON-CHLA LIGHT INTERFERENCE
AND RA GROUPS
Non-CHLA*
Light
Interference
LOW
RAt
Very-Low
Low
High
Nt
345
88
15
TP
(ug/D
208
75
161
KN
(ug/D
1234
1286
2003
CHLA
(ug/D
35.8
50.4
216.4
T
°C
22.0
23.2
24.0
RA
0.24
0.69
1.59
RS
(m)
-0.19
-0.35
-0.64
HIGH Very-Low 287 96 637 5.6 21.1 0.12 2.00
*Low non-CHLA light interference « -2.00 m RS _<0.46 m, while high non-CHLA
light interference » >0.46 m RS.
tVery-low response = <0.5 RA, low response = 0.5 >^ RA jQ.O, and high
response = >1.0 RA.
*N - the number of lakes per response and light interference groups.
However, the number of observations may be slightly less for a given
parameter due to missing values.
20
-------
Mean CHLA concentrations in the various non-CHLA light interference groups
seem to increase with increasing mean KN levels (Table 6). One exception to
this occurred in the high non-CHLA light interference, very-low RA group where
the mean CHLA level was only 5.6 u9/l but the corresponding KN concentration
was relatively high (637 u9/l)» This indicates that a large percentage of the
nitrogen present in this group is not associated with phytoplankton. This
same group had the highest amount of non-CHLA-related light interference as is
evident from its high RS (2.00). The nitrogen not associated with
phytoplankton is most likely caused by high levels of allochthonous organic
materials or resident bacterial or zooplankton populations in these lakes.
It is concluded that in many United States lakes, light attenuation from
other than CHLA-related interferences can dramatically affect the quantity of
phytoplankton biomass present.
NITROGEN TO PHOSPHORUS RATIOS
If a nutrient besides phosphorus is limiting phytoplankton growth, a low
RA should result. Nitrogen and phosphorus are frequently mentioned as the
nutrients most likely to limit growth of plants. The concept of a limiting
nutrient, as related to Liebig's Law of the Minimum, is that some nutrient,
least available relative to the growth requirements of a given organism,
imposes primary limitation on the growth of that organism.
Various estimates have been proposed as to what constitutes that ratio of
nitrogen to phosphorus at which the addition of either results in the
limitation by the other. Such estimates have been reported as low as 5/1 and
as high as 30/1 or more, by weight, generally centering about 12/1 to 14/1.
These are in reasonable agreement with theoretical needs based upon
stoichiometric equations of algal constituents (Vollenweider 1968).
While the limiting nutrient concept has some utility in allowing
prediction of the potential growth limits of laboratory monocultures, its
extrapolation into natural system studies should be approached warily. A
number of enrichment studies have found the effects of phosphorus and nitrogen
to be interdependent (Hutchinson 1941; Goldman and Armstrong 1969; Ketchum
1939; Powers et.al. 1972), influenced by the presence of trace organic
materials in the waters (Rodhe 1958), and dependent upon previous algal
culture exposure, i.e., prior luxury uptake of nitrogen or phosphorus.
It is not unlikely that a mixed natural phytoplankton population would
contain taxa with a range of optimal growth requirements and predisposing
nutritional status. Addition of either nitrogen or phosphorus could
potentially evoke a net increase in phytoplankton growth, especially in those
cases in which the ambient nitrogen/phosphorus ratio is intermediate between
the optimal growth requirements of the various phytoplankton elements.
Utilization of nitrogen/phosphorus ratios does not consider the possibilities
of other limitations, e.g., carbon, trace elements (iron, manganese, copper,
zinc, molybdenum, chlorine, cobalt, boron, silicon, sodium, vanadium and
iodine), co-factors, and vitamins. The presence of organic materials may
increase nutrient assimilation in some members of the community (Goldman
21
-------
and Armstrong 1969; Rodhe 1958), or inhibit it in others (Williams 1975).
Conditions may exist, over a range of nitrogen/phosphorus values, favoring
response to the addition of either phosphorus or nitrogen and representing
essentially "net co-limitation." Different species within a community may be
limited by different nutrients simultaneously (Fitzgerald 1964). A
theoretical basis for simultaneous co-regulation of the specific growth rate
of a single population by multiple nutrient is presented by Sykes (1974).
Lambou et al. (1976) divided lakes into phosphorus-limited [IN/dissolved
phosphorus (DP) >14], transition (1CKIN/DPXL4), and nitrogen-limited
(IN/DP<10) groups based upon the yearly mean IN/DP observed in the lake. They
state that, while admittedly arbitrary, the suggested divisions represent a
convenient means of comparing groups of lakes presumably representing "largely
nitrogen-limited" and largely "phosphorus-limited" populations, and a third
group representing a buffer between the first two groups. Their groupings
were determined by comparing the ratios of IN to dissolved orthophosphorus
(OP) in lake water upon which the limiting nutrient was determined by the
Algal Assay Procedure Bottle Test. This test (EPA 1971) utilizes the response
of the green alga Selenastrum capricornutum to nutrient spikes, usually
nitrogen and phosphorus, alone and in concert, to determine the growth-
limiting nutrient.
In our analysis of the lakes sampled by NES, because OP or DP was not
available for all of the lakes sampled, we used TP instead of DP in
calculating N/P ratios for dividing the lakes into phosphorus-limited
(IN/TP>14), transitional (10
-------
on
oo
o
N*^
L^
<:
CSL
a.
^^
z >-
CQ
Q£B
LU O
2£ LU
5J2
KH
gT«S
00 O
CD 3 o:
z as o
gf £
O 00
0 O O
o z z
< o. «c
a a LU
LJ_ LU
0") _J LU
oo
in
»*»«»*
S S? ro
r^ r>> in
en in co
I-v VO
CM en in
n co 1-1
to
«n to
in in
8
>x 1 1
1 1
in
JJ
<-< i i
~ i i
CM
CO
^^
O O
1^ v^ 1
* «"^
C3
CM
i ~ !
en vo
«n «T CD
"^
S S S
*H CM CM
"- *" *5
MO
IO f*^ CM
"Z* to" co
CM
^^ f^ OO
CD r» in
eo* 5r" eo*
f^ VO
CM
in co in
0- CO
j
i
>i f
BJ o -2"
* H» ^
VO
*
R
*^
CM
CM
1
CM
to
in
H^
in
in
VO
r*.
»
in
8
O
in
*»
j5
CM
S
^^
CM
O
CM
eo
IM
r*.
^f
S
to
^J
1**.
C9
eo
o"
in
00
^r
_,
d
o
H
t
yf
0)
i/i
^
S
J^
£1
_
e
ai
^
o
S1
*J
u
tfl
09 .C
Jt U
S
^» c
e "^
3 *
** *t
<0
Wl f
e
01 "0 Ol
u. .c
u *» at
« 01 C *»
o 4 at «
-els"
Q 8 * a. ji
ee in at
o e^ c
S £* -
US « |- .. M
tjl U v« 0)
"** * ** ""
M t r- TJ JS
v g -a *j
«c 5 jj ^ a
U ^ VI 7
LO , S
J= C L. J.
O Z O) 0 O
o>.e H-
» t. o a.
o L. vi a.
§<* 4J O "S.
.C Z
vi e a.
4) C
in ro o o
es "~ u u 1
v t*_ 01 ai E
_ 04^
«£ S s 01
a: i- 3 3 jc
at c e **
N ^
_ B I «
J 3
O C ** 4J CL
X " "E"! z
1? E ZZ -o
tti 3 it aj
> = zo. E
1 U H W
23
-------
then to 10.1 in the high RA group (Table 7). The ratio of biologically
available phosphorus to nitrogen forms, as theoretically represented by N/P
ratios, is in some cases an important factor in determining the RA ratios in
1akes.
TEMPERATURE
Solar radiation is required for photosynthesis and heat maintenance in the
environment. In general, approximately 50 percent of the total incident light
is transformed into heat and is extinguished in the first meter of water (Reid
1961). Van't Hoff's principle states that the rate at which biological
processes proceed is increased nearly two-fold with each 10°C rise in
temperature. The principle operates only within the range of tolerance of a
given species, and is further restricted by an optimal point within the
over-all range. Temperature is important to freshwater algae in determining
general geographical distribution and control rates of anabolic and catabolic
fluxes, as well as rates of the nutrient reflux, and it may affect
productivity by accelerating or retarding physiological processes (Smith 1950,
Mortimer 1969). However in our analysis of the data collected from the lakes
sampled by NES, we were not able to demonstrate any significant seasonal
relationships between T and RA (Table 8). The log-log product moment
correlation coefficients ranged from only 0.08 to 0.18. The implication of
these findings is that T is not an important factor influencing RA over the
range of temperature values (2.6 to 32.2°C) observed in the lakes.
TABLE 8. THE RELATIONSHIP BETWEEN LOG
IN U.S. LAKES
T AND LOG RA BY SEASON
Season
Spring
Summer
Fall
Product moment correlation
coefficients
Number of observations
0.11
769
0.18
754
0.08
747
24
-------
MODIFICATION OF MODELS TO PREDICT TROPHIC STATE
LOADING MODELS
The phosphorus loading mean depth relationship formulated by
Vollenweider (1968) has been widely accepted and used to indicate the degree
of eutrophy of lakes and evaluate the level of phosphorus loading to lakes.
Dillon (1975) pointed out that too little thought and criticism was given to
the limitations of the model by people using it. Subsequently, modifications
of the basic mass balance equation have been derived to predict mean ambient
lake phosphorus concentrations at equilibrium. Dillon (1975) utilizes
phosphorus areal loading (L), the retention coefficient for phosphorus (R),
the hydraulic flushing rate (P), and mean depth (Z) in a plot of the form
1 (1 - R) Vers1s z (9)
to estimate trophic state. Vollenweider (1975) revised his original formula
to include HT, hydraulic residence time, so that areal phosphorus loading (L)
is plotted against mean depth (Z) divided by HT. Larsen and Mercier (1976)
provide an alternative (to prior loading concepts) which avoids the criticism
of Edmondson (1970) that the effect of an increasing phosphorus load upon a
lake depends, in part, upon whether that increase results from increases in
influent flows, concentrations, or both. The Larsen/Mercier formula plots
mean tributary TP concentration against the phosphorus retention coefficient,
which they called R experimental, computed in the same way as Dillon's R,
i.e.,
R , i TP leaving system (10)
TP entering system
It is interesting to note that in applying the above-mentioned formulas,
each of the authors selected to use levels of 10 and 20 yg/1 of ambient lake
TP to divide lakes into the three standard trophic classifications
oligotrophic, mesotrophic, and eutrophic.
The most important public concern involves the manifestation of nutrients,
such as excessive algal growth, rather than phosphorus levels. However,
"critical" levels of total phosphorus need to be determined, so that
meaningful phosphorus control procedures can be accomplished on a given lake.
25
-------
The theoretical relationship between the summer RA and TP based upon Jones
and Bachmann's (1976) regression equation has already been presented in
Figure 2. A log-log transformation of the RA and TP concentration results in
a straight line function (Figure 7). This straight line function can be used
to provide a basis of comparison between the theoretical RA and the actual RA
for a given lake at a given TP level so that loading models can be modified
to:
1. change the trophic classification based on an ambient TP level to one
based on the biological manifestation of nutrients as measured by
CHLA,
2. determine the "critical" levels of TP which will result in an
unacceptable level of CHLA concentration so that the level of TP can
be manipulated to achieve the desired use of a given water body, and
3. account for the unique characteristics of a lake or reservoir which
effect the RA.
The Larsen and Mercier (1976) model predicts the mean tributary TP
concentration which would cause eutrophic or mesotrophic conditions as
follows:
fjf or
i-K,
(12)
where: TPp. a the minimum mean tributary TP concentration in ug/1 which will
cause a lake to be eutrophic at equilibrium,
TP"M = the minimum mean tributary TP concentrations in ug/1 which
will cause a lake to be mesotrophic at equilibrium,
ETP » a constant equal to 20, which is the theoretical minimum
ambient yg/1 of TP in a lake resulting in eutrophic conditions
and is the level which if not equaled or exceeded will result
in meso- or oligotrophic conditions, and
MTP = a constant equal to 10, which is the theoretical minimum
ambient yg/1 of TP in a lake resulting in mesotrophic
conditions and is the level which if not equaled or exceeded
will result in oligotrophic conditions.
26
-------
3.
C
CM
I
vo
I
c
fl3
ca
O)
c
o
c
o
01
(T3
.a
a.
i c
o
O 4J
ZJ
T3 CT
C Ol
< o
CD (/I
O
o
i.
CO
I
en
601
27
-------
The Larsen and Mercier equations (i.e., Equations 11 and 12) can be corrected
to account for the RA of a specific lake as follows:
ETP(ERA/AERA) (13)
TP.M MTP(MRA/AMRA) (14)
> ~
where: "TP.C = the minimum mean tributary TP concentrations in yg/1 which
will cause a lake to be eutrophic at equilibrium
corrected to account for the lake's RA,
TP~AM = t'ie mln''niuni mean tributary TP concentrations in yg/1 which
will cause a lake to be mesotrophic at equilibrium
corrected to account for the lake's RA,
ERA = a constant equal to 0.32 which is the RA predicted from
20 yg/1 of ambient TP utilizing Jones and Bachmann's (1976)
regression equation,
MRA = a constant equal to 0.23 which is the RA predicted from
10 yg/1 of ambient TP utilizing the Jones and Bachmann's
(1976) regression equation,
AERA = the mean summer RA for the lake corrected to what it would be
at the 20 yg/1 level of TP i.e., the ambient eutrophic level,
and
AMRA = the mean summer RA for the lake corrected to what it would be
at the 10 yg/1 level of TP, i.e., the ambient mesotrophic
level.
The ERA constant of 0.32 was determined from utilizing the ETP constant of
20 yg/1 of ambient TP in the Jones and Bachmann (1976) regression equation
given in Table 4 and which is as follows:
log yg/1 CHLA = -1.09 + 1.46 log yg/1 TP (15)
Substituting in 20 yg/1 for TP, log CHLA is equal to 0.81 and CHLA is equal to
6.4. Therefore the ERA is equal to 6.4/20 or 0.32. Similarly, the MRA
constant of 0.23 was determined utilizing the MTP constant of 10 yg/1 of
ambient TP.
28
-------
The AERA is determined from the following equation:
'«» «RA - []g Off I ft] ["» ETP - B] + A
where: ORA = the observed summer ambient RA in the lake,
DTP = the observed summer ambient TP in the lake,
A = -4.77 which is the log of the RA determined from Equation 15
utilizing a TP concentration at approximately 0 (since log 0
is undefined, an extremely low TP concentration, i.e.,
0.00000001 yg/1, was used to approximate 0 on the log scale),
and
B = -8 which is the log of the TP, i.e., 0.00000001 ug/1, which is
used to approximate 0 in Equation 15.
Substituting into Equation 16,
IOQ AERA = F1oq OR* * 4'771 9 30 - 4 77
109 AtKA ^ 1 Og OTP + 8 J '
The AMRA is determined from the following equation:
- B]
- '« «"" - " * A
Substituting into Equation 18,
IOQ AMRA = Hog ORA * 4.771 g _ , ?? (19)
log MMKA j^ log OTP + 8 J y 4>//
The constants used in Equations 16 and 18|are used to establish the slope of a
line (Figure 7) which begins at -4.77 (log RA) and -8 (log TP). Using the ORA
and the OTP the RA is adjusted using the relationship shown in Figure 7, which
was determined from the Jones and Bachmann (1976) regression equation
(Equation 15) to one which would cause eutrophic (AERA) or mesotrophic
conditions in the lake (AMRA).
A comparison of using the Larsen and Mercier equations (Equations 11 and
12) to predict trophic state with the modified equations to account for a
29
-------
lake's RA (Equations 13 and 14) can be made using the data we collected from a
lake during 1973. Our data showed that the lake had a:
OTP = 36.3 yg/1 ,
observed mean summer CHLA (OCHLA) = 6.3 yg/T ,
1-R = 0.71,
ORA' = 0.17, and
observed mean tributary TP (OTTP) = 57.3 yg/1.
Substituting into Equation 11 we find:
TPF 20 ~n , ,, (20)
E = oTTY = 28.2 yg/1
Since 28.2 yg/1 of TP represent the theoretical mean tributary concentration
which will cause the lake to be eutrophic under steady state conditions and
the OTTP is 57.3 yg/1 , the use of Equation 11 would classify the lake as
eutrophic. Substituting into Equation 13:
TPAE . 20(0.32^0.13) . 69-3 pg/1 (21)
Since 69.3 yg/1 is greater than 57.3 yg/1, we find if we use the modified
equation which accounts for the lake's RA, the lake would be classified as
eutrophic and could possibly be either mesotrophic or oligotrophic. To
determine whether it is mesotrophic or oligotrophic, we substitute into
Equation 14 as follows:
10(0-23/0.10) . 32. (22)
Since 32.4 yg/1 is less than 57.3 yg/1 we would classify the lake as
mesotrophic.
30
-------
AMBIENT MODELS
As with loading models, ambient models can be adjusted to account for the
observed RA in a lake. Using the commonly accepted convention of 10 and
20 ug/1 of ambient lake TP to divide lakes into three standard classifications
oligotrophic, mesotrophic, and eutrophic, we find that:
ATPE = 20 ug/1, and (23)
ATPM = 10 ug/1 (24)
where: ATPE = the minimum mean ambient summer lake TP concentration which
will cause the lake to become eutrophic and
ATP.. » the minimum mean ambient summer lake TP concentration which
will cause the lake to become mesotrophic.
Equations 23 and 24 can be corrected to account for the RA of a specific lake
as follows:
ATPAE = ETP(ERA/AERA), (25)
ATPAM = MTP(MRA/AMRA), (26)
where: ATP.r = the minimum mean ambient summer lake TP concentration which
will cause the lake to become eutrophic corrected to
account for the lake's RA and
ATP.JYJ = the minimum mean ambient summer lake TP concentration which
will cause the lake to become mesotrophic corrected to
account for the lake's RA.
Using the same data collected from a lake during 1973 which we previously used
as an example of adjusting loading models, we find that:
ATP... = 20(0.32/0.13) = 49.2ug/l,and (27)
AE
31
-------
ATPAM = 10(0.23/0.10) = 23.0 yg/1. (28)
Since the OTP was 36.3 yg/1, which is between the ATP/^ and ATP^n we would
classify the lake as mesotrophic using the correction for its RA; however,
using the normal convention of 20 yg/1 of ambient lake TP to determine the
eutrophic dividing line, we would have classified the lake as eutrophic.
RESULTS FROM USING MODIFIED MODELS
Equations 13, 14, 27, and 28 define trophic levels by the biological
manifestation of nutrients as measured by CHLA levels, rather than by TP
levels. Substituting 10 and 20 ug/1 of TP (the normally used dividing values
between oligotrophic, mesotrophic, and eutrophic) into Equation 15 [the Jones
and Bachmann (1976) regression equation to predict CHLA from TP], we arrive at
levels of 2.3 and 6.4 ug/1 of CHLA as the dividing values between
oligotrophic, mesotrophic, and eutrophic.
Trophic classifications based on mean ambient summer TP and CHLA levels of
757 lakes for which we had data are presented in Table 9 and Figure 8. The TP
criteria subdivided the lakes into the following trophic categories: 59
oligotrophic, 97 mesotrophic, and 621 eutrophic, while the CHLA criteria
subdivided the lakes into the following trophic levels: 78 oligotrophic, 197
mesotrophic, and 482 eutrophic (Table 9). Both trophic classification schemes
classified 19 water bodies as oligotrophic, 45 as mesotrophic, and 447 as
eutrophic, for a total agreement of 511 or 68 percent. Of the 246 water
bodies which were classified differently by the two schemes, 193 (25%) are
classified lower (less eutrophic), while only 53 (7%) are classified as higher
(more eutrophic) by the CHLA criteria than by the TP criteria (Table 9). If
the large population of lakes used for this comparison is, in fact,
representative of conditions throughout the United States, it can be expected
that some 25X of the nation's lakes would be classifed lower (less eutrophic)
using CHLA levels as a trophic classification criteria. And, if the
manifestations of nutrients rather than their absolute levels are considered
the primary criteria for beneficial water use, many communities may be spared
the unnecessary burden of costly nutrient-removal programs suggested by
phosphorus-based trophic classifications. Also, it should be emphasized that
eutrophication does not necessarily result in the degradation of water
quality. Whether or not the process adversely impacts water quality depends
on the beneficial use for which a lake is being managed and the degree of
nutrient enrichment.
The Vollenweider, Dillon, and Larsen/Mercier loading models for predicting.
ambient lake phosphorus concentrations and classifying lakes by trophic state
have been compared by Hern et al. (1978). The Larsen/Mercier and Dillon
models gave comparable results in ranking 39 lakes relative to known ambient
phosphorus concentrations. The Vollenweider model, which does not have a
phosphorus retention time capacity component, was unable to achieve the high
rank correlations found with the other models. The Larsen/Mercier model was
superior to the Dillon model as it required less information and produced the
32
-------
TABLE 9. NUMBER OF U.S. LAKES SAMPLED BY NES CLASSIFIED AS OLIGOTROPHIC,
MESOTROPHIC, AND EUTROPHIC BY SUMMER CHLA AND TP CRITERIA
Classified Classified by CHLA Criteriat
by TP
Criteria* Oligotrophic Mesotrophic Eutrophic Total
Oligotrophic 19* 18t 2* 59
Mesotrophic 19* 45* 33+ 97
Eutrophic 40* 134* 447* 621
Total 78 197 482 757
*TP criteria are <10 ug/1 TP » oligotrophic, 10 > yg/1 TP <20 = mesotrophic,
and >20 ug/1 TP » eutrophic.
tCHLA criteria as <2.3 ug/1 CHLA » oligotrophic, 2.3 >_ yg/1 CHLA <6.4 =
mesotrophic, and X>-4 ug/1 CHLA = eutrophic.
*Lakes which are classified trophically the same by both the CHLA and
TP criteria.
tLakes which are classified trophically higher (more eutrophic) by the CHLA
criteria than by the TP criteria.
*Lakes which are classified trophically lower (less eutrophic) by the CHLA
criteria than by the TP criteria.
highest Spearman rank correlation coefficient in the comparison study.
Therefore, only the application of the RA to the Larsen/Mercier model is
presented in this report. However, the same concept can be easily used to
modify any model which predicts ambient TP concentration by modifying the TP
levels used to divide lakes into trophic categories according to the lake's
RA. Likewise, if the commonly accepted convention of using TP criteria of 10
and 20 ug/1 is not suitable for a worker's purpose, the same concept can be
used to modify models based on other TP and CHLA criteria for trophic
classification.
The concept of modifying loading and ambient models to predict trophic
state was developed from a data base which contains only one summer sampling
round per lake. Because of the large population of lakes represented in our
data base we believe that the apparent trends in our data represent the actual
situations even though data for an individual lake may not be precise. We
suggest that water sampling be conducted intensively enough to accurately
estimate ambient summer CHLA and TP concentrations before a lake's observed RA
is used to modify loading or ambient models which predict trophic state.
33
-------
11 II 1^ » |- v g vww*Y*-*"*J*b***«^l^
0
Dj.qdoaa.n3
MM
^^
Cft<
L M«J
2:
< ~
*
«« *
,«
^
_ *
CL * * **
" *" " sT
c < «
O 4 < < «
^ ^
o ,. »<*«
g; U «c
<" S "T* 3*"»«"3
a S- "" - *****
P « »«
"» £S -
§ = -.- - s-" i
+* A * « ^ * 2 ^ I* ^
IQ « « B «
o " * 5 5*
C 0 S **J
f <* > « « U «
«" «^«««ucc
(o "5 <«£eoi«
^^ V S * O ^ to
O ee « ii t> «
o ao*
^_ . ^ < u a
,» <«.«»<
^^ tf ^ ^* a «
O »" < «c « u a «
J^ ««. W U W 1
» 2o. ***" 522
0 1 * <*5* rJ"
eg «L * 3t
^^ « «B 0 -
1 J= "
o a. -t:
«/> o < «
s *^ 25
1 'S t
a. a a.*/
o ^~
u ° -«
o>
5
0 0 O 0
. . « o
tO «3- f»J CJ
«*H
" hM"
-OSS^
*
«
^
^
^
«c
"* I
< M
^
u
C
* U
^ ^ C «
u w
< o
a « a
*
O-B
u
^ M *
U ««
a a
B < «
« ^ «* *
a wt.
«C « «C Cj
K U 1
« n
31:
f-H
00
o
J-J,
OLL)doa^.o6 L [0
^^
^^
cn 5-
<0 3 4J
xjk ^^ ^3
r- «/»
Q. t II
"! *"" *
(C 3. V
H O
1 «* 3.
p^ vxJ "" *^fc "
ZOO
o "S.co ca.
c o o
(0 5- CM i.
a. o »3
U, i Z
a. N/
*
(L/6rt) V1HD 601
34
-------
In summary, it is important to adjust for a lake's RA, when using loading
and ambient models which predict trophic state because:
1. non-CHLA light interferences and other interferences significantly
decrease most lakes' RA and
2. excessive algal growth or other manifestations of nutrient enrichment
are more important from a v/ater quality standpoint than a trophic
classification based on an arbitrary ambient TP level.
35
-------
LITERATURE CITED
Bachmann, R. W. and J. R. Jones. 1974. Phosphorus inputs and algal blooms in
'lakes. Iowa State J. Res. 49(2) :155-160.
Bliss, J. D., M. J. Friedland, and J. Hodson. 1975. Statistical Manipulation
of National Eutrophication Survey Data in STORET. National Eutrophication
Survey Working Paper No. 472. National Environmental Research Center, Las
Vegas, Nevada. 15 pp.
Carlson, R. E. 1977. A trophic state index for lakes. Llmnol . Oceanogr.
22:361-369.
Chiaudani, G. and M. Vighi. 1974. The N:P ratio and test with Selenastrum to
predict eutrophication in lakes. Water Res. 8:1063-1069.
Dillon, P. J. 1975. The phosphorus budget of Cameron Lake, Ontario: the
importance of flushing rate to the degree of eutrophy of lakes. Limnol .
Oceanogr. 20:28-39.
Dillon, P. J. and F. H. Rigler. 1974. The phosphorus-chlorophyll
relationship in lakes. Limnol . Oceanogr. 19:767-773.
Edmondson, W. T. 1970. Phosphorus, nitrogen and algae in Lake Washington
after diversion of sewage. Science. 169:690-691.
Fitzgerald, B. P. 1964. Detection of limiting or surplus nutrients in algae.
Progress report to NIH, 1961-1964 project. Working Paper No. 297. 48 pp.
Goldman, C. R. and R. Armstrong. 1969. Primary productivity studies in Lake
Tahoe, California. Verh. Int. Verein. Theor. Angew. Limnol. 17:49-71.
Hern, S. C. , V. W. Lambou, and L. R. Williams. 1978. Comparisons of models
predicting ambient lake phosphorus concentrations. Working Paper No. 704.
Environmental Monitoring and Support Laboratory, Las Vegas, Nevada. 9 pp.
Hutchinson, G. E. 1941. Limnological studies in Connecticut, IV. The
mechanisms of intermediary metabolism in stratified lakes. Ecol . Monoqr.
11:21-60.
Jones, J. R. and R. W. Bachmann. 1976. Prediction of phosphorus and
chlorophyll levels in lakes. J_. Water Pollut. Control Fed.
48(9}:2176-2182.
36
-------
Kalff, J. and R. Knoechel. 1978. Phytoplankton and their dynamics in
oligotrophic and eutropnic lakes. Ann. Rev. Ecol. Syst. 9:475-495.
Ketchum, B. H. 1939. The absorption of phosphate and nitrate by illuminated
cultures of Nitzschia closterium. Amer. jJ. Bot. 26:399-407.
Lambou, V. W., L. R. Williams, S. C. Hern, R. W. Thomas, and J. D. Bliss.
1976. Prediction of phytoplankton productivity in lakes. Iri:
Environmental Monitoring and Simulation, W. R. Ott (ed.).
EPA-600/9-76-016. U.S. Environmental Protection Agency, Washington, D. C.
pp. 696-700.
Larsen, D. P. and H. T. Mercier. 1976. Phosphorus retention capacity of
lakes. J. Fish. Res. Board Can. 33:1742-1750.
Mortimer, C. H. 1969. Physical factors with bearing on eutrophication in
lakes in general and in large lakes in particular. In: Eutrophication:
Causes, Consequences, Correctives. Nat. Acad. Sci., Washington, pp.
340-368.
Mullins, J. W., R. N. Snelling, D. D. Moden, and R. G. Seals. 1975. National
Eutrophication Survey: Data acquisition and laboratory analysis system
for lake samples. EPA-600/4-75-015. U.S. Environmental Protection
Agency, Environmental Monitoring and Support Laboratory, Las Vegas,
Nevada. 21 pp.
Powers, C. F., D. W. Schults, K. W. Malueg, R. M. Brice, and M. D. Schuldt.
1972. Algal responses to nutrient additions in natural waters, II. Field
experiments. In: G. E. Likens (ed.). Nutrients and eutrophication: the
limiting nutrient controversy. Amer. Soc. Limnol. Oceanogr., Publ.
1:141-154.
Reid, G. K. 1961. Ecology of Inland Waters and Estuaries. Reinhold
Publishing Corporation, New York. 3~75 pp.
Rodhe, W. 1958. The primary production in lakes: some methods and
restrictions of the l^C method. In: Measurements of primary production
in the sea, rapp. D-V Reun. Cons. Int. Perm. Explor. Mer. 144:122-128.
Sakamoto, M. 1966. Primary production by phytoplankton community in some
Japanese lakes and its dependence on lake depth. Arch. Hydrobiol.
62:1-28.
Schindler, D. W. 1974. Eutrophication and recovery of experimental lakes:
implications for lake management. Science. 84:897-899.
Smith, G. M. 1950. The Fresh-Water Alage of the United States. McGraw-Hill
Book Company, New York. 719 pp.
Strickland, J. D. H. 1960. Measuring the production of marine phytoplankton.
Fish. Res. Bd. £an. Bulletin No. 122. Ottawa. 172 pp.
37
-------
Sykes, R. M. 1974. Theory of multiple limiting nutrients. J.- Water Pollut.
Control Fed. 46(10):2387-2392.
U.S. Environmental Protection Agency. 1971. Algal assay procedure: bottle
test. U.S. Govt. Printing Off.: 1972-795-146/1. Region X. 82 pp.
U.S. Environmental Protection Agency. 1974. National Eutrophication Survey
methods for lakes sampled in 1972. National Eutrophication Survey Working
Paper No. 1. National Environmental Research Centers, Las Vegas, Nevada
and Corvallis, Oregon. 40 pp.
U.S. Environmental Protection Agency. 1975. National Eutrophication Survey
Methods, 1973-1976. Working Paper No. 175. Environmental Monitoring and
Support Laboratory, Las Vegas, Nevada, and Corvallis Environmental
Research Laboratory, Con/all is, Oregon. 91 pp.
Vollenweider, R. A. 1968. Scientific fundamentals of the eutrophication of
lakes and flowing waters, with particular reference to nitrogen and
phosphorus as factors in eutrophication. Organ. Econ. Coop. Dev., Paris
Tech. Rep. DAS/CSI/68.27:1-159.
Vollenweider, R. A. 1975. Input-output models with special reference to
phosphorus loading concept in limnology. Schweiz. I. Hydro!. 37:53-83.
Williams, L. R. 1975. The role of heteroinhibition in the development of
Anabaena flos-aquae water blooms. In: Proc. of the Biostimulation
Nutrient Assessment Workshop. 1973. Corvallis, Oregon.
Williams, L. R., V. W. Lambou, S. C. Hern, and R. W. Thomas. 1977.
Relationships of productivity and problem conditions to ambient nutrients:
National Eutrophication Survey findings for 418 eastern lakes. Working
Paper No. 725. Environmental Monitoring and Support Laboratory, Las
Vegas, Nevada. 20 pp.
Yentsch, C. S. and D. W. Menzel. 1963. A method for determination of
phytoplankton (chlorophyll and phaeophytin) by fluorescence. Deep Sea
Research. (10):221-231.
38
------- |