RESPONSE OF EUTROPHIC SHAGAWA LAKE, MINNESOTA,
USA, TO POINT-SOURCE, PHOSPHORUS REDUCTION
By
D. P. LARSEN, K. W. MALUEG, D. W. SCHULTS, AND R. M. BRICE
EUTROPHICATION AND LAKE RESTORATION BRANCH
U.S. ENVIRONMENTAL PROTECTION AGENCY
Introduction
Evidence has been compiled over the past several years indicating
the critical role the supply of phosphorus plays in controlling the
productivity of aquatic systems. Vollenweider (1973) has shown that
the trophic state of a great variety of lakes can be correlated with
areal total phosphorus loading, mean depth, and hydraulic retention
time. He has also shown that annual primary production of the
Laurentian Great Lakes can be related to their phosphorus loading
(Vollenweider, et al_., 1974). Vollenweider (1969, 1973) and Dillon
(1974) have derived mass balance equations to relate total phosphorus
concentration to total phosphorus loading, mean depth, hydraulic
flushing coefficient, and phosphorus deposition coefficient.
Sakamoto (1966) and Dillon (1974) have demonstrated a good correlation
between summer chlorophyll a_ concentrations and total phosphorus
concentration at spring overturn. Further, Jones has obtained a good
relationship between potential phosphorus concentration and average
summer chlorophyll a^ concentration for many lakes in Iowa as well as
for a number of other lakes.
Personal communication from Mr. John R. Jones, Iowa State University.
-------
Since the supply of phosphorus, to a large extent, controls the
productivity of aquatic systems, its supply can often be curtailed
to reduce the trophic state of adversely affected water bodies. A
municipal tertiary wastewater treatment plant, which has been constructed
at Ely, Minnesota, USA, reduced wastewater phosphorus concentration to
less than 50 yg/1. The result has been a 70% reduction in the loading
of phosphorus to nearby eutrophic Shagawa Lake during 1973-1974. This
paper documents the phosphorus loading reduction and describes changes
in the lake subsequent to that reduction.
Shagawa Lake
The Shagawa Lake basin was formed during the retreat of the Wisconsin
glacier some 10,000 years ago. It is one of numerous lakes located
in a sparsely populated region of northeastern Minnesota where iron
mining and lumbering first attracted settlers. Now the area primarily
attracts outdoor enthusiasts. The City of Ely began developing along
the southern shore of Shagawa Lake around the turn of the century,
attained a maximum population of about 6000 in the 1930's, and has
since declined to a relatively stable population of about 5000
year-round residents. Wastewater from Ely initially flowed into the
lake untreated, began receiving primary treatment in 1911, secondary
treatment in 1952, and tertiary treatment to remove phosphorus in 1973.
Shagawa Lake is characterized by three basins of about 13 m maximum
depth, a mean depth of 5.6 m, a mean volume of about 5.3 x 10 m ,
and a surface area of 925 ha (Figure 1). One major tributary,
Burntside River, and several minor tributaries enter the lake; there
is one outlet, Shagawa River. The hydraulic residence time is 8-9
months and the phosphorus residence time has been less than 6 months.
Surface water temperatures generally reach 22-24°C during summer months,
while bottom temperatures sometimes reach nearly 20°C. Stratification
-------
exists below 5-6 m during summer months but is subject to break-down
by passage of cold fronts and strong winds. The lake is ice-covered
approximately five months a year. Specific conductance is about 60
ymho/cm, and alkalinity is 22 mg/1 as CaCO-. Anaerobic conditions
have existed in the deep holes during summer stratification and during
late winter under ice-cover.
Large blooms of diatoms and green algae develop during spring, and
blue-green algal blooms dominate during the late summer. These blooms
are in marked contrast to those of the generally oligotrophic, nearby
lakes. Core analyses (Bradbury and Megard, 1972; Bradbury and
Waddington, 1973) suggest that a marked increase in productivity
occurred corresponding with the development of mining and lumbering
in the area. The eutrophic state of Shagawa Lake was attributed to
the supply of nutrients, particularly phosphorus, from the wastewater
of Ely (Powers, et al_., 1972; Smith, 1973; Malueg, ejt al_., in press).
Materials and Methods
Tributary hydraulic flow has been monitored daily since 1966, utilizing
gages installed and calibrated by the U.S. Geological Survey; wastewater
flow has been monitored hourly. Details of tributary and wastewater
flow, rainfall, evaporation, and ground-water are presented elsewhere
(Malueg, ejt ail_., in press).
Weekly point samples were obtained from the tributaries for analysis
of total phosphorus and orthophosphate phosphorus. Daily composite
samples were obtained from wastewater during the interval 1972 to 1974.
Prior to that time, grab samples were obtained at less frequent intervals.
Phosphorus loadings were determined as described in Malueg, e_t a\_. (in
press).
-------
Total phosphorus (TP) and orthophosphate phosphorus (OP) were measured
as described in Standard Methods (1968), and chlorophyll _§_ (uncorrected
for phaeo-pigments) was measured according to UNESCO (1966) methods.
Weekly samples for analysis of TP, OP, and chlorophyll j[ were obtained
at the surface and at 1.5 m depth intervals to the bottom in each of
the three basins.
Results
Loadi ng reducti on
The average annual load of wastewater and "natural" phosphorus for
1967-mid-1974 is summarized in Table 1. "Natural" phosphorus in the
present context encompasses all phosphorus not measured as wastewater
phosphorus and includes that contributed by tributaries, direct runoff
(estimated), and rainfall. The wastewater supply accounted for about
81% of the total phosphorus load from 1967 to 1972 and was reduced to
about 25% during 1973. The total phosphorus load to the lake was
reduced to about 30% of its previous level. During 1973 and 1974, the
lake had entered a phase of phosphorus washout, approximately 50% more
phosphorus leaving the lake than entering. The supply of total
phosphorus from all sources is summarized in Figure 2, delineating
the difference between wastewater and "natural" loads. There was a
significant decrease in the wastewater supply during the period
January-March, 1973, when the treatment plant underwent testing prior
to full-scale operation. A further decrease occurred in April, 1973,
when the plant commenced full-scale operation. Occasional spikes from
that time on indicate plant bypass due to temporary shutdown or
excessively high flow.
-------
Lake Changes
t
Total phosphorus (TP) concentration
The changes in concentration of TP in the upper 5.25 m (approximating
the average mixed depth during stratification and representing about
80% of the lake volume) are summarized in Figure 3. Prior to ice-out
(usually in late April), the mean concentration of TP increased during
the interval from 1971 to 1973. When the lake was ice-free (May -
early November), the differences were indistinguishable for 1971 and
1972, but a lower concentration occurred during the fall of 1973.
During the ice-free season, the usual pattern was a decline in TP
concentration to low values of 25-30 yg/1 in June followed by a sharp
increase in late summer to values near 90 yg/1. This increase was
primarily a result of release of phosphorus by the sediments. During
the ice-covered interval in 1974, the TP concentration was nearly
constant at about 20 yg/1, less than half that of 1973 and slightly
lower than that in 1971. An increase in concentration occurred during
ice breakup, as had been observed in previous years. Subsequent to
ice breakup, the concentrations were similar to those of previous years
but slightly lower, particularly in July when the concentrations were
nearly half those of the previous three years.
Displayed for comparison is the average concentration of TP in the
upper 5.25 m of Burntside Lake during 1972. Burntside Lake is a
lower mesotrophic to oligotrophic lake located upstream of Shagawa
Lake and provides about 70% of the flow to Shagawa Lake. Burntside
Lake serves as a useful control.
Orthophosphate phosphorus (OP) concentration
The average concentrations of OP in the mixed zone are summarized in
Figure 4. High concentrations, ranging from 10-35 yg OP/1, existed
-------
during ice-cover each year. The trend during this interval was
similar to that observed for TP, although during short intervals, the
concentrations observed in 1971 and 1972 were nearly the same as those
observed in 1973. The pattern observed each year after ice-out has
been a decline of OP during the spring algal bloom to levels less than
2 yg/1 for most of June. In early July, there has been an increase in
OP originating from the sediments. The increase was particularly
relevant in 1973 when the wastewater supply had been reduced, and the
lake had entered a period of phosphorus washout. The concentrations of
OP attained during July and August are not considered limiting to
algal growth.
In early 1974, one year after initiation of treatment, OP
concentrations were considerably lower than those observed in
previous years during the winter months. However, the nearly constant
concentration of 11 yg/1 was sufficient to support a large, spring
algal bloom. In contrast, OP concentrations in Burntside Lake were
consistently low, generally less than 2 yg/1 throughout the year.
Chlorophyll a concentration
The concentration of chlorophyll a^ has been used as an indicator of
algal biomass throughout this study. A spring bloom at ice-out has
occurred each year except 1973. In 1973, the ice-cover was more
transparent than in other years and the maximum of the early algal
bloom was reached before ice breakup. Fluctuations between 10 and
30 yg/1 existed throughout most of June and July followed by a
dramatic increase in early August. In 1972, this increase had occurred
about one month earlier. These summer blooms are a response to the
increased supply of nutrients from the sediments as indicated by the
changes in TP and OP during this interval. This was especially
evident in August, 1973, when essentially no phosphorus was supplied
from wastewater. The blooms generally began to decline in September,
responding to lower temperatures and decreased light levels.
-------
In 1974, the spring bloom was of a magnitude similar to that in 1971
and 1972, but the June and Ouly concentrations of chlorophyll a^ were
lower than had been observed in previous years. The low constant
pattern of chlorophyll a. in Burntside Lake stands in marked contrast
to that of productive Shagawa Lake.
Discussion
Sonzogni and Lee (1973), and Dillon (1974) have reviewed the response
of lakes to which the supply of nutrients has been curtailed, and, in
general, the response has been that which might be predicted by simple
hydraulic or phosphorus washout models. Examples include Lake
Washington, USA; Zellersee, Austria; and Little Otter Lake, Canada. A
notable exception has been Lake Samrnamish, USA, to which about 40% of
the phosphorus supply had been curtailed. Lake Trummen, Sweden, also
failed to respond significantly to diversion of large amounts of
nutrients (Bjork, 1972). The failure of predicted changes was
attributed to high internal phosphorus loads in both lakes.
Shagawa Lake appears to have reached a state in which internal loading,
especially during July and August, contributes a sufficient amount of
phosphorus to produce large crops of algae. Mass balance estimates
demonstrated that the internal load of phosphorus during July-August,
1971-1973, was sufficient to increase the average concentration in the
lake 1-2 yg/l/day (Larsen, in prep.). Figures 3 and 4 display this
increase, particularly in 1973, when during this interval, phosphorus
leaving the lake exceeded that entering from external sources. Most
of this internal load redeposited during fall circulation, and thus it
had little influence on annual calculations of the phosphorus budget.
-------
Although any attempt to predict the recovery of the lake will depend
upon the ability to predict the course of internal loading, it is
useful to compare the phosphorus mass in the lake with that predicted
by a mass balance model of the type described by Vollenweider (1973)
(Figure 6). This model assumes a well-mixed lake, no internal sources
of loading, and a constant phosphorus deposition coefficient.
For Shagawa Lake, a value for the phosphorus deposition coefficient was
calculated as the difference between the phosphorus loss coefficient
and the hydraulic loss coefficient. The phosphorus loss coefficient
was calculated as the quotient of the average annual loading of total
phosphorus to the lake and the average total phosphorus mass in the
lake for the two years 1971 and 1972. To project TP concentrations
in the lake, an average hydraulic and "natural" phosphorus loading
year was calculated using weekly values from 1972 and 1973; residual
wastewater phosphorus loadings were estimated from those obtained
after April, 1973. Weekly model calculations were made to compare
phosphorus and hydraulic washout models with observations in the lake.
In the absence of internal loading, both models suggest a rapid
recovery reaching a steady state within two years. The hydraulic
washout model better estimates the TP concentration during the
1973-1974 ice-covered interval. The ice-covered interval perhaps
best represents average conditions in the lake in the absence of large
influences of internal loading. Marked deviations from the models
emphasize the importance of internal loading particularly during the
summer period. During this period in 1973, the average concentration
of TP in the lake increased from 30 yg/1 to more than 80 yg/1,
as a result of internal loading. A refined prediction of
the recovery of the lake should include an estimate of the internal
loading of phosphorus, a proposition of considerable difficulty.
-------
Although not as rapid as could be expected in the absence of internal
loading, there is certainly evidence suggesting a favorable response
to reduced phosphorus loading to Shagawa Lake. Winter-time
concentrations of TP and OP were lower in 1974 than they have been
in previous years. The lake is in a phase of phosphorus washout,
having lost about twice as much phosphorus as gained in 1973 and
early 1974. The June-July concentrations of chlorophyll a_ in 1974
are at their lowest as compared with those of the previous three years.
Summary
The eutrophic state of Shagawa Lake has been attributed primarily to
the supply of phosphorus from the wastewater of nearby Ely, Minnesota,
USA. A recently constructed tertiary wastewater treatment plant has
reduced the phosphorus load to the lake by about 70%. As a result, the
lake has entered a phase of phosphorus washout. Average total and
orthophosphate phosphorus concentrations in the lake are lower than in
the past; chlorophyll a^ concentrations are only slightly lower. The
significant amount of internal phosphorus loading during summer months
suggests that the predicted achievement of a new equilibrium in about
two years, based on a phosphorus washout model, will not be attained.
Acknowledgments: The sampling program was carried out by Mark Schuldt,
Bob Randall, Paul Smith and their staffs. Howard Mercier assisted with
the data reduction. Marv All urn and Spencer Peterson provided useful
comments concerning manuscript revisions.
-------
References
American Public Health Association. 1965. Standard Methods for the
Examination of Water and Wastewater. 12th ed. APHA, New York.
769 p.
Bradbury, J. P. and R. 0. Megard. 1972. The stratigraphic record
of pollution in Shagawa Lake, northeastern Minnesota. Geol.
Soc. Am. Bull. 83:2639-2648.
Bradbury, J. P. and J. C. B. Waddington. 1973. The impact of
European settlement on Shagawa Lake, northeastern Minnesota,
USA. Univ. of Minnesota, Limnol. Res. Ctr. Contribution
No. 112. 31 p.
Bjork, S. 1972. Swedish lake restoration program gets results.
Ambio 1:153-165.
Dillon, P. J. 1974. The prediction of phosphorus and chlorophyll
concentrations in lakes. Ph. D. Thesis. University of Toronto.
330 p.
Malueg, K. W., D. P. Larsen, D. W. Schults and H. T. Mercier. 1974.
A six-year water, phosphorus, and nitrogen budget for a eutrophic
Minnesota lake prior to point source phosphorus removal. In
Press.
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, p. 141-154. IJK
G. E. Likens (ed.), Nutrients and eutrophication. Am. Soc.
Limnol. Oceanogr. Spec. Symp. 1.
Sakamoto, M. 1966. Primary production by phytoplankton community
in some Japanese lakes and its dependence on lake depth. Arch.
Hydrobiol. 62:1-28.
Smith, P. D. 1973. Studies on the effects of sewage effluents on
phytoplankton productivity in experimental ponds. M.S. Thesis.
University of Minnesota, Minneapolis, MN.
Sonzogni, W. C. and G. F. Lee. 1974. Diversion of wastewaters from
Madison Lakes. Journal of the Environmental Engineering Division
ASCE. 100:153-170.
-------
UNESCO. 1966. Monographs on Oceanographic Methodology. 1.
Determination of photosynthetic pigments in seawater.
United Nations Educational Scientific and Cultural
Organization. Paris. 69 p.
Vollenweider, R. A. 1969. Moglichkeiten und Grenzen elementarer
Modelle der Stoffbilanz von Seen. Arch. Hydrobiol. 66:1-36.
Vollenweider, R. A. 1973. Input-output Models. Schweiz. Z. Hydrol,
In Press.
Vollenweider, R. A., M. Munawar and P. Stadelmann. 1974. A
comparative review of phytoplankton and primary production in
the Laurentian Great Lakes. J. Fish Res. Bd. Can. 31:739-762.
-------
Ltst of Tables
Table 1. Annual total phosphorus supply to, and loss from,
Shagawa Lake, 1967-1973.
-------
Table 1. Annual total phosphorus supply to,
and loss from, Shagawa Lake, 1967-1974
Supply (kg)
Year
1967
1968
1969
1970
1971
1972
Average
1973
1974
(Jan. -May)
Wastewater
5245
5349
5449
5606
5460
5176
5380
543
70
"Natural "
797
1530
1328
1766
1379
1064
1310
1596
628
Wastewater
% of Total
87
78
80
76
80
83
81
25
10
Loss (kg)
Shagawa
Ri ver
2928
6203
5488
6145
4675
3144
4674
4308
1065
-------
List of Figures
Figure 1. Bathymetric map of Shagawa Lake.
Figure 2. Wastewater and "natural" total phosphorus supply
to Shagawa Lake, 1971-1974.
Figure 3. Average total phosphorus concentrations in upper
5.25 m of Shagawa (1971-1974) and Burntside (1972)
Lakes.
Figure 4. Average orthophosphate phosphorus concentrations in
upper 5.25 m of Shagawa (1971-1974) and Burntside
(1972) Lakes.
Figure 5. Average chlorophyll a^ concentrations in upper 5.25 m
of Shagawa (1971-1974) and Burntside (1972) Lakes.
Figure 6. Comparison of hydraulic and phosphorus washout models
with observations in Shagawa Lake, 1973-1974. Total
phosphorus concentrations are the average for the
entire lake.
-------
Ul
>
UJ
> 3 W
(E < (E
UJ . UJ
I- H
Z >- UJ
- ^ s
nr >
5 o; co
-------
OJ
m^
in
00
7--> <
7<
•s
L?
ro
o
10
O
O
O
lO
O O
in
o
o
o
in
graven snaoHdsoHd
-------
— CM ro ri- cvi
N-
(7)
O
O>
O
00
O
10
O
ro
O
C\J
sndOHdsoHd nvioi
-------
(l/6r1) sndOHdSOHd 31VHdSOHdOHldO
-------
(I/6H)
-------
in
ro
N-
0)
o
0>
o
00
o
(D
O
in
o
ro
O
OJ
sndOHdsoHd nvioi
------- |