s

363.7394
H2ebnd

EVALUATION OF THE BENEFITS OF NUTRIENT REDUCTIONS
ON ALGAL LEVELS IN THE CLARK FORK RIVER
FINAL REPORT

SUBMITTED
JUNE 30, 1991
BY

STATE DOCUMENTS COLLECTIOl.

JAn 2 8 2002

MONTANA STATE LIB;,, . v

1515 E eth AVr
HELEN.' ^-^ '^""' ■"

DR. VICKI WATSON, ASSOCIATE PROFESSOR
UNIVERSITY OF MONTANA

TO

MONTANA DEPT. HEALTH AND ENVIRONMENTAL SCIENCES
MR. GARY INGMAN, PROJECT EVALUATOR

MONTANA STATE LIBRARY

S 363 7394 H2ebnd 1 99 1 c.2 Watson
Evaluation ol benelits of nutrient reduc

0864 00077848

TABLE OF CONTENTS
BENEFITS OF THE MISSOULA PHOSPHATE BAN — EXECUTIVE SUMMARY i
NUTRIENT LIMITATION IN THE UPPER CLARK FORK — EXEC SUMMARY iii

BENEFITS OF THE MISSOULA PHOSPHATE BAN 1

TABLES 7

FIGURES 8

APPENDIX N — NUTRIENT SAMPLING ON THE CLARK FORK RIVER,
SUMMER 1990

APPENDIX T — UPPER RIVER RECIPROCAL TRANSPLANT STUDY

APPENDIX M — DESCRIPTION OF THE RIVER ALGAE ACCUMULATION MODEL
PERISIM

Digitized by the Internet Archive

in 2011 with funding from

IVIontana State Library

http://www.archive.org/details/evaluationofbene1991wats

BENEFITS OF THE MISSOULA P BAN

EXECUTIVE SUMMARY

Attached algae in the Clark Fork River exceed aesthetic nuisance
levels set by British Columbia and interfere with recreation.
Algal respiratory demands contribute to night time violations of
Montana water quality standards for dissolved oxygen. In an
effort to control algal levels by controlling nutrient loading,
the city of Missoula banned phosphate-based laundry detergents,
reducing P loading from this major point source almost 50%.

To evaluate the benefits of this ban, a study of river algae
accumulation was conducted. It was not possible to base this
evaluation on a comparison of preban and postban algal levels
because preban data were from a low flow year while post ban data
were from an average flow year. In addition preban data included
only one site above the point source (insufficient site
replication) , and the three down stream sites had nutrient levels
that approached or exceeded the point that saturates algal
standing crop. Little improvement was expected at these sites;
most improvement was expected further downstream where reduction
in lower nutrient levels should result in greater reductions in
algae. No preban data were available for these sites.

Hence the evaluation was accomplished by using a computer model
that simulates the accumulation of attached algae over the
summer. The model accurately predicted the accumulation of algae
measured on artificial substrates at four sites before the ban
and 7 sites after the ban. Then this validated model was used to
predict the accumulation of algae at several sites downstream of
the point source under preban and postban loading for both
average flow and 10 year low flow years. Reductions in algae
levels were expressed as the percent reduction in the peak
biomass and the cumulative biomass simulated over the summer.

Under average flow conditions, the model predicted that the P ban
would effect a 6 to 38 % reduction in peak and mean algal
biomass, depending on the site. Under low flow conditions, the
predicted reductions were slightly greater — 9 to 48 %. As
expected, greater reductions occurred further downstream.
Assuming the study sites are representative of the reaches they
bracket, the P ban produced at least a 25% reduction in algal
levels over 110 river miles. During low flow years, the ban was
predicted to effect at least a 40% reduction in 100 river miles.

British Columbia has set 100 mg/m2 of chlorophyll a as the level
of attached algae that represents an aesthetic nuisance. Assuming
for the moment that this level represents a criterion for algae,
peak summer levels in the Clark Fork exceed this criterion from
the headwaters down to the confluence with the Flathead. Peak
levels predicted by the model exceed 100 mg/m2 both before and
after the ban. However, mean summer levels from Harper's Bridge
to St. Regis exceed 100 mg/m2 before the ban and drop below this
level after the ban in average years. In low flow years only

Harper's Bridge and Huson move from above the criterion to below
it. Under all scenarios, the site just below the sewage plant
exceeds the criterion, and the Plains site is below the
criterion. If the Harper's Bridge, Huson and St. Regis sites are
representative of the reaches they bracket and the reach down to
the Flathead, the P ban reduced the mean summer chlorophyll level
below this nuisance criterion for 100 miles of river.

Given the role of attached algae in dissolved oxygen violations
and its possible role in foaming, such reductions would certainly
be ecologically significant. Would such reductions produce a
change in algal levels or foaming that casual observers could
perceive? This is unlikely. Algal levels are tremendously
variable from rock to rock within a site. So differences between
sites and years can only be perceived with very costly labor
intensive sampling programs. And the differences between sites
and years could be due to factors other than nutrients. Paying
such a cost to prove the benefits of the P ban in each site and
year is not justified.

Like rivers, people show a lot of variability. We all know people
who died of lung disease who did not smoke, and smokers who did
not die of lung disease. However, by compiling health and
mortality data from thousands of people, it was possible to
determine that there is a significant relationship between
smoking and lung disease. This relationship is used to predict
the increased risk of lung disease from various levels of
exposure to smoking. Based on this relationship, warnings are
issued against smoking, and smoking is banned in areas where it
would affect other people. We do not require proof that every
person who smokes will get lung disease before protecting people
from the risk associated with smoking.

Artificial stream studies have quantified the relationship
between nutrient and attached algae levels. This relationship was
incorporated into a realistic, well validated model that was used
to predict how river algal levels will likely respond to various
exposures to nutrients. Based on these predictions, the Missoula
P ban has greatly reduced the river's risk of producing nuisance
algae levels and the attendant water quality problems. The
benefit to cost ratio of this managment action is quite high.

The only management action that could rival the P ban in its
potential effect would be much more costly — land treatment of the
Missoula sewage in the summer time. If all these nutrients were
retained on the land, the resulting reduction in P and N loads to
the river would be enormous. The sites downstream of the sewage
plant might be expected to approach the site above the sewage
plant in nutrient and algal levels. Nutrients brought in by the
Bitterroot River and the pulp mill would result in some increase
in algae, but summer algal levels would be drastically reduced.
These reductions would further benefit chemical water quality
(and could be predicted using the model used in this study) ;
however, the potential impacts on the fishery of this reduction
in food base should be addressed also.

n

NUTRIENT LIMITATION IN THE UPPER CLARK FORK

EXECUTIVE SUMMARY

The upper Clark Fork exhibits massive growths of the filamentous
green alga Cladophora that interferes with fishing and
contributes to violations of the state DO standard. To evaluate
what nutrient reductions would be required to reduce Cladophora
levels significantly, Cladophora growth potential was assessed at
several sites on the river that exhibit different nutrient levels.

Cladophora requires more than a year to develop massive growths
on new substrates. Germinating zoospores must first establish a
holdfast and then grow upright filaments. The upright filaments
often break off, particularly over winter, and new filaments grow
from the holdfast in subsequent years. With each year the growths
appear more massive. To study the limitation of massive growths
in one growing season, rocks already well colonized with
Cladophora and just beginning the new year's growth of filaments
were collected from two sites. These rocks were then transplanted
to four other sites that differed in their levels of soluble
inorganic P and N. The four sites and their nutrient levels were:

Median (and range) of

SRP (ppb)

Nitrate (ppb)

N:P

Warm

Springs:

40 (20-90)

15 (<10-40)

0.375

Deer

Lodge :

5 (3-22)

80 (4-150)

16

Gold

Creek:

27 (11-37)

<10 (<10-30)

<0.37

Bear

Creek:

37 (16-46)

<10 (<10)

<0.27

In mid July, rocks were placed at similar depths and water
velocities and checked weekly when water samples were collected.
Half the rocks were harvested after one month and half after a
second month. Algal levels were expressed as chlorophyll a and
ash free dry weight (AFDW) per unit area.

In August, chlorophyll levels were slightly correlated with
median SRP and nitrate (slightly better with nitrate) . In
September, neither nutrient showed any correlation with
chlorophyll levels. In August, AFDW showed no significant
differences between sites, and in September, AFDW differences
were not correlated with median nutrient levels.

The ratio of N:P at these sites suggests that N is more likely to
limit algal growth than P at three sites (Warm Springs, Gold
Creek and Bear Creek) while P should be more limiting at Deer
Lodge. However, massive growths accumulated at Deer Lodge, Gold
Creek and Bear Creek but not at Warm Springs. The lack of massive
accumulations at Warm Springs may have been caused by heavier
grazing or by limitation by toxic metals or a shortage of some
trace element.

m

Hence, nutrient levels in the upper river appear to be
sufficiently high long enough that other factors account for much
of the variation in algal levels.

These results suggest that it may be necessary to reduce the
median nutrient concentrations below the 5ppb SRP observed at
Deer Lodge and below the nitrate detection limit (10 ppb)
observed at Gold Creek and Bear Creek before really significant
reductions in Cladophora levels are observed. A careful
assessment of point source contributions is necessary to
determine whether this would be possible by controlling point
sources alone.

It should be noted that Cladophora can store excess nutrients
during short periods of high nutrient levels to last through
fairly long periods of low nutrient levels. Hence if all the
sites had a few short periods of high nutrient levels, the
difference in the typical levels at these sites may not be that
relevant. Such short periods of high nutrient levels might not be
picked up even in weekly sampling. In addition, nutrient levels
in June might determine algal levels for the rest of the summer.
It may be that controlling these short periods of high nutrient
levels would result in reductions in algal levels at those sites
that typically have fairly low nutrient levels. Again, a careful
assessment of point source contributions may reveal whether this
is possible by controlling point sources only.

It may be that the most important factors controlling Cladophora
accumulations in some Montana rivers is the frequency and
duration of spring scouring flows and the hardness of the water.
The middle Clark Fork has nutrient levels as high or higher than
the upper river (although the N:P ratios are not so low) but is
not characterized generally by massive Cladophora growths (though
the alga is present) . The Middle Clark Fork water is softer and
has higher flows that last until later in the spring/ summer .
Increased reductions in flows associated with withdrawal of water
from the river will increase hardness and decrease scouring flows
(particularly toward the end of the high flow season) . This may
allow the Cladophora problem to move downstream. In recent years
I have observed that certain growths of Cladophora in the middle
river have become more massive, but this may be natural
variation.

IV

INTRODUCTION

Attached algae communities in the Clark Fork River sometimes
reach levels that interfere with beneficial uses of the river.
Nutrient addition experiments carried out in streamside
artificial streams have identified the levels of nitrogen and
phosphorus that saturate attached algae standing crop in these
waters. Most reaches of the Clark Fork most of the time are below
the nutrient levels that saturate standing crop, hence reductions
in nutrient levels may reduce attached algae levels in the river
(Watson et al. in press) .

Recently, the City of Missoula enacted a ban on the sale of
laundry detergents containing a significant amount of phosphate.
Since that time the soluble reactive phosphorus (SRP) load from
the Missoula sewage plant has been reduced almost 50%. In
recent years the Frenchtown pulp mill dramatically reduced P
loading (and N loading to a lesser extent) . This reduction, along
with the greater dilution by higher river flows, likely resulted
in lower instream SRP levels in 199 0 when compared to the drought
years that preceded the ban. The effect of such a change in
loading on river biota is often evaluated by comparing biotic
response above and below the point of loading both before and
after the change takes place. This is the classic impact study
design. Simple upstream-downstream comparisons are confounded by
the fact that the sites may differ for many factors besides the
one under study. Simple before-after comparisons are confounded
by year-to-year differences in many uncontrolled factors. Even
with the combined upstream-downstream before-after design, it is
difficult to evaluate the benefits of a single action when
numerous other factors change over time and space as well. In
addition, benefits may not occur immediately downstream or
immediately in time.

In the case of the effect of the Missoula phosphate ban, some
limited 'before the ban' algal accumulation data exists. Algal
accumulation on artificial substrates was measured above and
below the sewage plant and the pulp mill in 1987 and above and
below the sewage plant in 1988. However, the three sites below
the sewage plant have such high nutrient levels that nutrients
may still be near levels that saturate standing crop even after
the ban. Changes in nutrient levels near the point of saturation
have a small effect on algal growth rates which is detectable in
a controlled laboratory situation but are unlikely to be
detectable when measured under field conditions with their
greater sources of variation. The greatest benefits of the ban
are likely to accrue further down river where nutrient levels are
lower. Here changes in algae levels may be great enough to
measure; however, there is no preban data for these sites.

Another approach provides a better means of quantifying the
effects of the P ban on algal levels in the Clark Fork. A model
of attached algae accumulation developed by Watson has been
validated for the middle Clark Fork River using instream nutrient

and algae levels measured in 1988 and 1990. The model was then
used to predict accumulation of algal standing crop in different
parts of the river under both average flow and low flow
conditions given the SRP loads produced by the Missoula Sewage
Plant before and after the P ban.

ALGAL ACCUMULATION IN THE MIDDLE & LOWER CLARK FORK, 1987 & 1990

The primary purpose of monitoring algal accumulation on
artificial substrates was to obtain data that could be used to
validate the algal accumulation model. Some limited evaluations
of the artificial substrate data are made below.

Methods — Artificial substrates identical to those used in the
summer of 1987 (unglazed ceramic tiles) and similar to those used
in 1988 (styrofoam beadboard) were placed in the Clark Fork at
the same sites monitored in 1987: above and below the Missoula
waste water treatment plant WWTP (Russel Street and Schmidt site)
and above and below the mill (Bioassay shack and Ken Cyr ' s land
at Huson) . In addition substrates were placed at several sites
farther downriver (Alberton, Superior, St. Regis, and Plains —
the Superior substrates were vandalized soon after placement) .
Substrates were kept between 20-30 cm deep and in water flowing
0.3 m/s +/- O.lm/s. Middle river substrates were sampled weekly
and lower river substrates monthly from early July until the end
of September. Five to 10 replicate samples were collected each
time. Water samples were collected at the same time and analyzed
for soluble reactive P and for soluble inorganic nitrogen (unless
the WQB was sampling at that time) . Algal material on the
substrates was analyzed for chlorophyll a and for ash free dry
weight according to Standard Methods (1985) .

Results — Substrates placed upstream of the Missoula WWTP showed
fairly similar behavior in all three years (Figures, 1, 2, 3 —
chlorophll levels were somewhat higher in 1988, a severe drought
year) . The site below the Missoula WWTP reached similar peak
levels in all three years but certainly grew more slowly at the
beginning of the summer in 1990. The SRP and nitrate levels were
lower earlier in the summer and rose in late August and early
September, probably contributing to the rapid increase in
accumulation during that period. The stations above and below the
mill showed a rapid increase in biomass early in the summer (and
actually outstripped the below Missoula site early on) but levels
dropped off later in the summer. In early summer, these sites had
higher nitrate levels than the below Missoula site, but later
these nitrate levels dropped below those at the below Missoula
site (see nutrient data in Appendix N) . Low N levels and the drop
in algal standing crop observed in early summer at the below
Missoula site may have been caused by some high flows in upper
river tributaries which were somewhat damped out by groundwater
inflows and Bitterroot River inputs at the sites bracketing the
mill. Generally, algal standing crops at the sites bracketing the
mill were similar in 1990 and 1987 with the exception of the last
sampling below the mill which was much lower in 1990 than in
1987.

There is no preban data for the sites from Alberton down, but
after the ban the Alberton algal levels were similar to the below
mill levels (just a bit lower) , and the St. Regis levels were
similar to the above Missoula levels. The algal accumulation
levels at Plains (not shown in these figures; see next section)
were lower still. Hence these sites demonstrate that lower
accumulation levels are associated with the lower nutrient levels
that characterize these sites.

Conclusions — To facilitate comparing the maximum algal
accumulation levels attained in 1987, 1988 and 1990, the last
three to four samplings of the summer were averaged (Figure 4) .
Sites are referred to by number on graphs: 1,2 — above and below
the Missoula STP; 3,4 — above and below the mill; 5 — Alberton; 6 —
St. Regis; 7 — Plains. Again the site above the STP had similar
biomass in all years with 1988 a little higher. The site below
the mill accumulated significantly less chlorophyll in 1990,
while the site above the mill accumulated significantly less AFDW.

EVALUATION OF BENEFITS OF THE P BAN THROUGH USE OF AN ALGAE MODEL

Nature of the Algae Accumulation Model

The model used in this study is described in detail in Appendix
M. The following is a short conceptual description.

The algal accumulation model was developed to simulate the
accumulation of attached algae on the bed of a river. The model
allows algal biomass to 'grow' as a function of temperature,
light and nutrient level. Biomass is lost to respiration and
sloughing. The algal growth and respiration functions are fairly
typical formulations used in the literature. A maximum growth
rate (taken from the literature) measured under optimal
conditions of light, temperature and nutrient levels is reduced
in proportion to the difference between optimal conditions and
the actual ambient conditions. Respiration is handled in a
similar fashion but is a function of temperature only.

The above portions of the model could be used to simulate algal
growth in a lake as well as a river. In order to consider the
accumulation of attached algae in a river, the model must also
consider sloughing (the detachment and washing away of the
algae) . Sloughing has been modeled in a number of ways, from
having a fixed sloughing rate to having sloughing increase as
algal biomass approaches a level considered to be the maximum
biomass that can be supported by the conditions at the site under
study. Generally, this maximum biomass is set at the maximum
value ever observed in the system under study. In the model used
here, an improvement on the latter approach was used.

Studies of attached algae in artificial streams show that when
attached algae is allowed to colonize and grow on a bare surface,
it accumulates rapidly at first, then the rate of accumulation

slows and approaches zero. The processes of loss come to balance
the processes that add to the biomass and an equilibrium biomass
is reached. This equilibrium biomass is a function of nutrient
concentration. That is, the higher the nutrient concentration the
longer the algae will increase in biomass before accumulation
levels off. Hence the maximum biomass that can be sustained is a
function of nutrient level. The artificial stream studies of
Bothwell and Watson were used to develop a function that predicts
maximum biomass from soluble reactive phosphorus (SRP) levels.
Hence the maximum biomass that a site can support is changing
over time as a function of ambient SRP concentration. And again
as the algal biomass levels estimated by the model approach the
maximum level that can be supported, sloughing increases.

So in a nutshell, the model is given as input an initial low
level of biomass, and then is given on an hourly basis the light,
temperature and nutrient conditions at a particular site on the
river. Every hour, the model estimates the growth, respiration
and sloughing rates as described above and adds or subtracts an
appropriate amount of biomass from its previous estimate of the
biomass. Thus algal biomass gradually increases or decreases
depending on whether current conditions cause gains to exceed
losses or vice versa.

Validation of the Model

To validate a simulation model one must have several independent
sets of validation data. Validation data is a combination of both
the input data the model requires (in this case, light,
temperature and nutrient data) and the ouput data the model
produces (algal biomass levels over the summer) . In 1988 and 1990
both SRP and algal accumulation rates were measured at several
sites on the Clark Fork. (In 1987 algal accumulation rates were
measured but SRP was not, and in 1989 SRP was measured but algal
accumulation rates were not) .

The model was used to simulate algal accumulation at two sites in
1988 (above and below the Missoula sewage treatment plant or STP)
and at 7 sites in 1990 (above and below the STP, above and below
the pulp mill and at Alberton, below St. Regis and at Plains) .
The results of these validation runs appear in Figures 5 to 12 .
In all of these figures the model simulation results appear as a
line. The horizontal markers represent the 95% confidence
interval of the measured algal levels on artificial substrates in
the river. The line produced by the model does not pass through
every 95% confidence interval, but taken all together, the model
does a good job of predicting the differences in algal levels
between these sites and the general pattern of algal accumulation
over the growing season.

Estimating Nutrient Levels Under Different Loading & Flow Scenarios

To evaluate the benefits of the phosphate ban, it seemed most
reasonable to use the algae model to predict algal accumulation
rates under the nutrient conditions one would expect under the

following 4 scenarios:

Average flow conditions with P loadings before and after ban
Low flow conditions with P loadings before and after ban

In order to estimate the SRP levels that would be expected under
these conditions, it was necessary to develop a simple nutrient
model for the river. This model estimates the nutrient
concentration at each station by dividing the total loads added
to the upstream reach by the flows at the station in question. In
other words:

SRP @ DOWN = (SRP*Flow £ UP ± SRP*Flow of Trib or Eff )

(combined Flow of UP and Trib or Eff)

Where SRP is the concentration of SRP at downstream (DOWN) or
upstream stations (UP) or in tributaries (Trib) or effluents
(Eff) ; Flow is the volume of water passing a point per unit time.

Flow data were provided by the USGS office in Helena and included
the long term average flows for these sites and the 10th
percentile low flows (only 10% of flows are lower than these) .
Sewage plant loading is based on data available from the plant
(data from preban years were averaged as were post ban data) .
Inriver SRP concentrations measured at the above Missoula site
and in the Bitterroot and Flathead rivers were also averaged for
preban and post ban years.

This approach has a number of sources of error. It assumes we
have a good estimate of the loads and the flows of each reach. It
also assumes that phosphorus is conservative which it is not.
Phosphorus may be taken up or released by the algae in each reach
at any given time. Hence the downstream concentrations may be
greater or lesser than would be predicted by this strictly
conservative approach.

During the summer, nutrient levels along the river are generally
lower than predicted by this conservative model. This is to be
expected since algae are taking up nutrients from the water. The
percent of the incoming nutrient load retained in each reach over
several summers of available data was determined, and this
retained load was subtracted before estimating the concentration
below the reach. The above approach produced the nutrient levels
summarized in Figures 13 to 16.

Algal Accumulation Under Four Nutrient Scenarios

The above nutrient levels were used to produce four experimental
simulations for each of the following sites: above and below the
WWTP, above and below the mill, at St. Regis and Plains. Results
of these simulations appear in figures 17 to 22. Each figure
shows the results of two simulations — before and after the P ban.
Each site and each flow regime is depicted on a different figure.

Conclusions

To permit quick comparisons between pre and post ban simulations,
differences between these simulations were summarized in a single
number. First, the peak algal biomass of the summer was noted for
each pair of simulations, and the percent reduction between these
peaks was noted. In addition, the mean summer algal biomass was
determined, and the percent reduction between pre and post ban
means was calculated (Table 1) .

Generally, greater % differences are exhibited farther down the
river where nutrient levels are lower. The farther the nutrient
level is below saturation, the greater is the effect of a
nutrient loading reduction. In addition, algal differences are
generally greater in lower flow years than in average flow years
because the instream nutrient concentration change from pre to
post ban loading is greater in low flow years. One exception is
the site just below the Missoula WWTP. In low flow years the
nutrient concentrations both before and after the ban are so near
saturation that the reduction has less of an effect.

British Columbia has set 100 mg/m2 of chlorophyll a as the level
of attached algae that represents an aesthetic nuisance. Assuming
for the moment that this level represents a criterion for algae,
observed summer levels in the Clark Fork exceeded this criterion
from the headwaters down to the confluence with the Flathead
before the ban (Water Quality Bureau data) . Peak levels predicted
by the model exceed 100 mg/m2 before and after the ban. However,
predicted mean summer levels from Harper's Bridge to St. Regis
exceed this level before the ban and drop below it after the ban
in average flow years. In low flow years, only the mean biomass
at Harper's & Huson move from above the criterion to below it.
Under all scenarios, the site below the WWTP exceeds the
criterion, and the Plains site is below it. If the Harper's to
St. Regis sites are representative of the reaches they bracket
and the reach down to the Flathead, the P ban reduced mean summer
chlorophyll below this criterion for 100 miles of river (Table 2) .

ACKNOWLEDG EMENT S

This work was supported by the Clark Fork Monitoring Project with
funds from the Clean Water Act Section 525 Study Project. Access
to the river was kindly provided by Ken Cyr, Mr. and Mrs. Robert
Hammer, Mr. 'Schmidty' Schmidt and Stone Container Corporation.
The Missoula WWTP loading data were provided by Tim Hunter and
Joe Aldegarie of the City of Missoula. Clark Fork River flows
were provided by Mel White of the Helena Office of the USGS.

REFERENCES

APHA. 1985. Standard Methods for the examination of water and
waste water. 16 th edition.

Watson, V. J., P. Berlind, L. Bahls. in press. Control of algal
standing crop by P and N in the Clark Fork River. Clark Fork
River Symposium, Montana Academy of Science, April, 1990.

TABLE 1. PREDICTED REDUCTION IN ALGAL ACCUMULATION IN RESPONSE TO
THE MISSOULA PHOSPHATE DETERGENT BAN

FLOW
AVERAGE

SITE

ABV MSL
BEL MSL
ABV MILL
BEL MILL
ST. REGIS
PLAINS

RIVER PEAK SUMMER BIOMASS

MILES (mg CHL a/in2)

FROM BEFORE AFTER %

ABV MSL P BAN P BAN REDUCED

0 50 50 0

2 220 190 14

13 160 150 6

25 190 160 16

89 140 130 7

122 80 50 38

MEAN SUMMER BIOMASS

(mg CHL a/in2)

BEFORE AFTER %

P BAN P BAN REDUCED

40 40 0

152 139 9

117 88 25

123 92 25

112 80 29

66 41 38

LOW

ABV MSL
BEL MSL
ABV MILL
BEL MILL
ST. REGIS
PLAINS

0

2

13

25

89
122

50
230
180
200
170

70

50
210
150
170
130

40

0
9
17
15
24
43

34

154

106

103

85

36

34
138
69
60
44
20

0
10
35
42
48
44

TABLE 2. PREDICTED REDUCTION IN FREQUENCY OF OCCURRENCE OF NUISANCE ALGAL
CONDITIONS IN RESPONSE TO THE MISSOULA PHOSPHATE DETERGENT BAN

RIVER

FLOW

SITE

MILES
BELOW
ABV MSL

AVERAGE

ABV MSL

0

BEL MSL

2

ABV MILL

13

BEL MILL

25

ST. REGIS

89

PLAINS

122

LOW

ABV MSL

0

BEL MSL

2

ABV MILL

13

BEL MILL

25

ST. REGIS

89

PLAINS

122

% OF DAYS ALGAL LEVELS % REDUCTION

> 100 mg/in2 CHL IN FREQUENCY OF
BEFORE AFTER EXCEEDANCES OF

CHL CRITERIA

0

0

65

75

100

0

0

2

29

50

200

0

P BAN

PBAN

0

0

79

79

65

23

63

16

43

0

0

0

0

0

83

81

78

58

78

47

65

0

0

0

LIST OF FIGURES

OBSERVED ATTACHED ALGAL BIOMASS ACCUMULATION, MIDDLE CLARK FORK:

FIG 1. ash free dry weight accumulation (AFDW) , 1987 & 1990

FIG 2. chlorophyll a accumulation, 1987 & 1990

FIG 3. chlorophyll a accumulation, 1988

FIG 4. late summer averages, chlorophyll and AFDW, 1987-1990

VALIDATION RUNS OF THE ALGAL ACCUMULATION MODEL

(COMPARISON OF SIMULATED AND OBSERVED ALGAL LEVELS)

FIG 5. chlorophyll a, above & below Missoula WWTP, 1988

FIG 6. chlorophyll a, above Se below Missoula WWTP, 1990

FIG 7. ash free dry weight, above & below Missoula WWTP, 1990

FIG 8. chlorophyll, above & below mill (Harper Br. & Huson) , 1990

FIG 9. ash free dry weight, above & below mill, 1990

FIG 10. chlorophyll & ash free dry weight, Alberton, 1990

FIG 11. chlorophyll, St. Regis & Plains, 1990

FIG 12. ash free dry weight, St. Regis & Plains, 1990

MODEL SIMULATIONS OF SOLUBLE REACTIVE PHOSPHORUS LEVELS AT SIX
SITES ON THE RIVER FOR 2 FLOW SCENARIOS AND 2 LOADING SCENARIOS

SITES ARE: AM & BM = ABOVE AND BELOW MISSOULA WWTP;

HB & HU = ABOVE AND BELOW PULP MILL (HARPER'S BRIDGE & HUSON);
SR = ST. REGIS; PL = PLAINS.

FIG 13. AVERAGE FLOW, BEFORE BAN

FIG 14. AVERAGE FLOW, AFTER BAN

FIG 15. LOW FLOW, BEFORE BAN

FIG 16. LOW FLOW, AFTER BAN

ALGAL MODEL SIMULATIONS OF ALGAL BIOMASS BEFORE AND AFTER P BAN
FOR AVERAGE AND LOW FLOW CONDITIONS:

FIG 17. ABOVE MISSOULA

FIG 18. BELOW MISSOULA

FIG 19. ABOVE MILL (HARPER'S BRIDGE)

FIG 20. BELOW MILL (HUSON)

FIG 21. ST. REGIS

FIG 22. PLAINS

ALGAL ACCUMULATION, SUMMER 1987

MIDDLE CU\RK FORK RIVER

10

20 30 40 50 60

DAYS SINCE STARTUP (870715)

ALGAL ACCUMULATION, SUMMER 1990

I
CD

>
CE
Q
UJ
LU

tz

LL

I
en
<

MIDDLE CLARK FORK RIVER

20 30 40 50 60

DAYS SINCE STARTUP (900713)

FIGURE ]

ALGAL ACCUMULATION, SUMMER 1990

MIDDLE CLARK FORK RIVER

250

J

ABVMSL

-■-

BELMSL

ABVMILL

—I —
BELMILL

▲

ALB

STREG

10 20 30 40 50 60 70 80

DAYS SINCE STARTUP (900713)

ALGAL ACCUMULATION, SUMMER 1987

MIDDLE CU\RK FORK RIVER

250

10

20 30 40 50 60

DAYS SINCE STARTUP (870715)

FIGURE 2

ABVMSL
BELMSL
ABVMILL
BELMILL

70 80

ALGAL ACCUMULATION, SUMMER 1988

MIDDLE CUKRK FORK RIVER

250

20 30 40 50 60

DAYS SINCE STARTUP (880708)

ABVMSL

BELMSL

80

FIGURE 3

FIGURE 4

LATE SUMMER ALGAL BIOMASS, 1987-90

MIDDLE CLARK FORK RIVER

250

c7 200
E

E

ra 150

>-

i 100

o

_l

o 50

ABV MSL

BELMSL

—

+

BEL MILL

■4- ~

ABV MILL

—

± +

-t-

+

ALB

SR

PL

87 88 90 87 88 90 87 90 87 90 90 90 90

YEAR

35
30
25
20

15H
10

X

< 5

£

I

o

LU

cr

Q
LU

BELMSL

ABV MILL

BEL MILL

ABV MSL

ALB

+

SR

PL

"sT 90 87 90 87 90 87 go"

YEAR

"go 90 go"

MEAN

- 95% CON. LIM. -

SIMULATION OF ALGAL STANDING CROP
ABOVE MSL STP. SUMMER 1988

300
250
200
150
100
50h

MODEL

OBS 95%

CONFIM

§00 210 220 230 240 250 260 270
DAY OF YEAR

SIMULATION OF ALGAL STANDING CROP
BELOW MSL STP, SUMMER 1988

MODEL

OBS 95%

CONFIM

200 210 220 230 240 250 260 270
DAY OF YEAR

Figure 5

SIMULATION OF ALGAL STANDING CROP
ABOVE MSL STP, SUMMER 1990

300
250
200
150
100
50h

MOOEL

OBS. 95%

CX>FIM

S20 230 240 250 260
DAY OF YEAR

270 280

SIMULATION OF ALGAL STANDING CROP
BELOW MSL STP, SUMMER 1990

300
250
200
150
100
50-

220 230 240 250 260
DAY OF YEAR

MODEL

OBS. 95%

CONFiW

270 280

Figure 6

SIMULATION OF ALGAL STANDING CROP
ABOVE MSL STP, SUMMER 1990

I

50
40
30
20
10

MODEL
oes 95%
CONFUwl

§20 230 240 250 260 270 280
DAY OF YEAR

SIMULATION OF ALGAL STANDING CROP
BELOW MSL STP, SUMMER 1990

I

50
40
30
20
10

MODEL
OBS 95%

§20 230 240 250 260 270 280
DAY OF YEAR

Figure 7

SIMULATION OF ALGAL STANDING CROP
AT HARPER BRIDGE, SUMMER 1990

300
250
200
150
100-
50-

MODEL

OBS 95%

CXDNFIM

§10 220 230 240 250 260 270 280
DAY OF YEAR

SIMULATION OF ALGAL STANDING CROP
AT HUSON, SUMMER 1990

SlO 220 230 240 250 260 270 280
DAY OF YEAR

MODEL

OBS 95%

COM^iM

(

Figure 8

SIMULATION OF ALGAL STANDING CROP
AT HARPER BRIDGE, SUMMER 1990

50
40
30
20
10-

MCOEL
CBS 95%

§10 220 230 240 250 260 270 280
DAY OF YEAR

SIMULATION OF ALGAL STANDING CROP
AT HUSON, SUMMER 1990

50
40
30
20
lOh -

MODEL

OBS 95%

COt¥JJA

210 220 230 240 250 260 270 280
DAY OF YEAR

Figure 9

SIMULATION OF ALGAL STANDING CROP
AT ALBERTON, SUMMER 1990

300
250
200
150
100
50

MODEL

OBS 95%

OONFUA

§10 220 230 240 250 260 270 280
DAYOF YD«

SIMULATION OF ALGAL STANDING CROP
AT ALBERTON. SUMMER 1990 -

50
40
30
20
10

MODEL

OBS. 95%

CCfFlM

Si 0 220 230 240 250 260 270 280
DAY CF\EAR

Figure 10

SIMULATION OF ALGAL STANDING CROP
BELOW ST. REGIS, SUMMER 1990

I

.5X)

—

?fn

Mum

2U0

-

(TR 95%

150

^

CONFiM

1CX)

50

-

^— -— — — ^=—

s,

.

-^

_

0

220

230 240 250 260 270

280

DAY OF YEN?

SIMULATION OF ALGAL STANDING CROP

PLAINS, SUMMER 1990

300

250
200
150
100
50

MOOEL

OBSl 95%

CONFiM

220 230 240 250 280 270 280
DAYOTYEJiR

Figure 11

SIMULATION OF ALGAL STANDING CROP
BELOW ST. REGIS, SUMMER 1990

g

50
40
30
20

10

MODEL

CBS. 95%

CCNFiiA

SlO 220 230 240 250 260 270 280
DAYOFVEAR

SIMULATION OF ALGAL STANDING CROP

PLAINS, SUMMER 1990

50

!

40

i

30

g

20

10-

MODEL

OBS. 95%

CONFiM

§10 220 230 240 250 260 270 280
DAYOF YE^

Figure 12

FIGURE 13

SIMULATED SRP LEVELS, CLARK FORK RIVER
AVERAGE FLOW, BEFORE P BAN

15-Jun

1 4-Aug 29-Aug 1 3-Sep

AM

— I—
BM

— ^I^r-

HB

— B-

HU
SR
PL

DATE

FIGURE 14

SIMULATED SRP LEVELS, CLARK FORK RIVER
AVERAGE FLOW, AFTER P BAN

60

^ 50

CO
Q_
D- 40

O

Z

o
o

D-
DC

30

0
15-Jun

30-Jun 15-Jul

30-Jul
DATE

1 — I — I — r

14-Aug 29-Aug 13-Sep

AM

BM

HB

HU
SR

PL

FIGURE 15

SIMULATED SRP LEVELS, CLARK FORK RIVER
LOW FLOW, BEFORE P BAN

0 I I — I — \ — \ — i ' i — I — I — I — I — I —

15-Jun 30-Jun 15-Jul

T r

^^— I — I 1 — I — I — I — I 1 1 i I III "T^-^

30-Jul 14-Aug 29-Aug 13-Sep
DATE

AM

BM

HB

HU
— >e-
SR

PL

FIGURE 16

SIMULATED SRP LEVELS, CLARK FORK RIVER
LOW FLOW, AFTER P BAN

15-Jun 30-Jun 15-Jul

30-Jul 14-Aug 29-Aug 13-Sep
DATE

250

<N 200
^ 150

o

100
50

FIGURE 17

SIMULATED ALGAL ACCUMULATION
ABOVE MISSOULA, AVG FLOW SCENARIO

BEFORE AND AFTER P BAN

i

I I I I I
15-Jun 30-Jun

-1 — I I I — I — I — I I I —
15-Jul 30-Jul

DATE

14-Aug 29-Aug 13-Sep

250

<N 200
I' 150

I
O

100

50
0

SIMULATED ALGAL ACCUMULATION
ABOVE MISSOULA, LOW FLOW SCENARIO

BEFORE AND AFTER P BAN

-1 — I — T— 1 — H

-I— I — i— r

-1 — I — I — r-

-I— I — I — r

I I I I I I

15-Jun 30-Jun 15-Jul 30-Jul 14-Aug 29-Aug 13-Sep

DATE

250

FIGURE 18

SIMULATED ALGAL ACCUMUU\TION
BELOW MISSOULA, AVG FLOW SCENARIO

~i — I — I — I — I — I — I — I — I — I — I — I — I — r

15-Jun 30-Jun 15-Jul 30-Jul

DATE

14-Aug 29-Aug 13-Sep

SIMULATED ALGAL ACCUMULATION
BELOW MISSOULj^, LOW FLOW SCENARIO

250

BEFORE P BAN

AFTER P BAN

0 I I I I

-1 — I — I — r—\ — r— 1 — I — I — I — I — I — I — I — I — I — I — I — I — I — I — 1 — I — I — h

15-Jun 30-Jun 15-Jul 30-Jul 14-Aug 29-Aug 13-Sep

DATE

FIGURE 19

SIMULATED ALGAL ACCUMULATION
HARPER'S BRIDGE, LOW FLOW SCENARIO

250

15-Jun 30-Jun 15-Jul 30-Jul 14-Aug 29-Aug 13-Sep

DATE

250

E

200

150

<

100

_j
X

o

50

SIMULATED ALGAL ACCUMULATION
HARPER'S BRIDGE, AVG FLOW SCENARIO

BEFORE P BAN

AFTER P BAN

I I I I I I I

-1 — I — I — I — I — I — i-H — I — I — r-

1—1 — I — I — I — I — t— 1 — I — I — r-

15-Jun 30-Jun 15-Jul 30-Jul 14-Aug 29-Aug 13-Sep

DATE

FIGURE 20

SIMUUMED ALGAL ACCUMUU\TION
HUSON, AVG FLOW SCENARIO

250

E

200

CD

E

150

<

100

_j

X

50

Ot — I — I — I — I — I — I — I — I — I — I — I — I — I — I — I — I — I — I — I — I — I — I — 1 — I — I — I — I — I — I I

15-Jun 30-Jun 15-Jul 30-Jul 14-Aug 29-Aug 13-Sep

DATE

250

"^ 200

SIMULATED ALGAL ACCUMULATION
HUSON, LOW FLOW SCENARIO

15-Jun 30-Jun 15-Jul 30-Jul 14-Aug 29-Aug 13-Sep

DATE

250

FIGURE 21

SIMULATED ALGAL ACCUMULATION
ST. REGIS, AVG FLOW SCENARIO

14-Aug 29-Aug 13-Sep

DATE

250

SIMULATED ALGAL ACCUMULATION
ST. REGIS, LOW FLOW SCENARIO

"T — I — I — r

BEFORE P BAN

AFTER P BAN

15-Jun 30-Jun 15-Jul 30-Jul 14-Aug 29-Aug 13-Sep

DATE

FIGURE 22

250

SIMULATED ALGAL ACCUMUUMION
PUMNS, AVG FLOW SCENARIO

15-Jun 30-Jun 15-Jul

30-Jul 14-Aug 29-Aug 13-Sep
DATE

250

<^ 200
^ 150

< 100

50-

I
O

SIMUU\TED ALGAL ACCUMUUVTION
PLAINS, LOW FLOW SCENARIO

BEFORE P BAN

AFTER P BAN

i

/

U I I I I I I I I I I I I
15-Jun 30-Jun 15-Jul

-1 — I — I — I — I — 1 — I — I — r

30-Jul 14-Aug 29-Aug 13-Sep
DATE

APPENDIX N — NUTRIENT SAMPLING ON THE CLARK FORK RIVER, SUMMER 1990

METHODOLOGY

Between July 11 and October 2, 1990, water was sampled weekly at 12
sites on the upper and middle Clark Fork and bimonthly from 5 sites
on the lower river (site locations in Table 1) . Samples were not
collected during the weeks when the Montana Water Quality Bureau was
conducting its monthly sampling.

Sampling procedures followed EPA guidelines, including acid washing of
all eguipment that came in contact with samples. Immediately after
collection, samples were filtered in the field through 0.45 um
membrane filters. Samples were transported on ice to the lab (within 8
hours of collection) and then frozen until thawed by microwave just
before analysis. Filtered samples were analyzed in the Pi's lab at the
University of Montana for soluble reactive phosphorus or SRP (EPA
365.2 — colorimetric ascorbic acid) and for nitrate (EPA 353.3 —
color imetric, manual cadmium reduction) . Samples were analyzed for
ammonia at the Freshwater Laboratory of the UM Yellow Bay Station.

The limit of quantification (or LOQ) of SRP is 2 . 5 ug/1 +/- 10% or 10
ug/1 +/- 5%. The LOQ of nitrate is 10 ug/1 +/- 20% or 30 ug/1 +/-5%.
EPA QA samples were run with each batch of analyses and were always
within 95% of their accepted values. Like the river samples, these EPA
QA samples were frozen and thawed before analysis, indicating that
this method of handling did not substantially alter the nutrient
levels in the samples. Field blanks showed no measureable
contamination .

Table 2 contains the summer nutrient sampling data, including that
collected by WQB.

RESULTS

SRP — Figure IN shows the maximum, minimum and median SRP values from
July to October, 1990, at each site. The mainstem shows increases in
SRP downstream of the Deer Lodge sewage lagoons, Gold Creek, Flint
Creek and the Missoula sewage plant. Values at any one site vary 2 to
5 fold over time.

Figure 2N shows weekly SRP values at Warm Springs (WS) and Deer Lodge
(DL) . The wide ranging values at WS may have resulted from the
reclamation work being done on the ponds and creek channel. The rise
in SRP at DL on 8/21 and 8/28 occurred during high summer flows
following heavy rains. On 8/21 the river at DL was extremely muddy
(channel bottom could not be seen in 6 inces of water) . Deer Lodge and
the Bear Creek site (below Flint Creek and Drummond) were noticeably
more muddy than other sites on 8/21. The river was somewhat clearer on
8/28, but the flow continued to be high.

Fig. 3N shows weekly SRP values at 5 upper river sites that were
sampled by this project as well as by WQB. DL, Beavertail (BVT) , and
Turah (TU) all increased suddenly on 8/21, but sites above the Little
Blackfoot (ALBF) and below Gold Creek (GC) did not. ALBF and GC

decreased over time from the end of July to the end of September. This
trend may be the result of some combination of algal depletion,
changes in flow or in the Deer Lodge lagoon discharge.

Figure 4N provides a more detailed look at the stretch of river
between the confluence with the Little Blackfoot and Rock Creek. This
project sampled two additional sites not regularly sampled by WQB:
between the Little Blackfoot and Gold Creek (BLBF) and near Bear Creek
(BC) which is below Flint Creek. BLBF closely follows the trend at GC
mentioned earlier. The slight increase in SRP from BLBF to GC shows
the influence of Gold Creek when the creek's concentration exceeded
100 ug/1 SRP and its flow was greater than 15 cfs. After 8/21 the
Clark Fork's flow increases about 20% (USGS data) and Gold Creek flow
decreases by 50% (J. Carey, unpubl. data) . The SRP values at the BC
and BVT sites were well above upstream sites after mid August. USGS
data indicate that Flint Creek was high from late July to late August
and may help explain these high SRP values.

Figure 5N shows flow and SRP over time at GC. Flows alone cannot
explain the decrease in SRP levels over the summer at this site.

Figure 6N shows weekly SRP values at 5 middle river sites (ABVSTP,
BELSTP, ABVMILL, BELMILL, and ALB) . The first three sites downstream
of the Missoula STP showed increases in SRP in mid August. The
extremely high value (70 ppb) for the above the mill site seems
unlikely, but an increase at these sites at that time was corroborated
by the periphyton data. Before mid August algae were accumulating very
slowly at all three sites. Then in mid August all three sites showed a
dramatic increase in accumulation rates, followed by a decrease in
algal standing crops when nutrient levels dropped the following week.
Then the site below the STP exhibited rapid accumulation again as its
nutrient levels remained high while the sites above and below the mill
leveled off at more modest algae levels in keeping with their modest
nutrient levels.

NITRATE — Figure 7N shows maximum, minimum and median values of nitrate
from Warm Springs to Plains. Nitrate was generally at or below
detection except at Deer Lodge, below Missoula STP, below the
Bitterroot (ABVMILL) and below the mill (BELMILL) . Figure 8N shows
weekly nitrate at WS, DL, and ALBF. At Deer Lodge, nitrate levels are
inversely related to flows (Figure 9N) , supporting the hypothesis that
most of this nitrate originates in groundwater. The two outliers are
8/21 and 8/28, the days of high, turbid flow.

Figure ION shows nitrate levels in the middle river over the summer.
Nitrate above Missoula's STP were at or below detection. Nitrate
levels were much higher below the STP and were sometimes higher still
at below the Bitterroot (ABVMILL) . However, when ammonia values were
added to nitrates, the site below the STP always had the highest
soluble inorganic N levels (Figure UN) . The site below the mill
showed considerable variation which was not detected by the WQB
monthly sampling. The WQB sampling site below the mill is a bit
farther downstream than the site used by this study.

TABLE 1,

IDENTIFICATION OF SAMPLING SITES

SYMBOL NAME

WS Warm Springs

DL Deer Lodge

ALBF Above Little Blackfoot

BLBF Below Little Blackfoot

GC Gold Creek

DR Drummond

BC Bear Creek

BVT Beavertail

TU Turah

BF Blackfoot River

ABVSTP Above Missoula STP
or AM

BELSTP Below Missoula STP
or BM

ABVMILL Above pulp mill
or HB

BELMILL Below pulp mill
or HU

ALB Alberton

SU Superior

AFH Above Flathead River

FH

Flathead River

BFH Below Flathead River
or PL

LOCATION
** 0.5 miles below Warm Springs Bridge
** Business 190 bridge, south of town
** Three miles east of Garrison

One mile west of Pat's Place
** USGS station below Gold Creek

Campground below railroad bridge

BLM access 3 miles upstream

of bridge at Bear Creek
Beavertail campground

(near WQB's Bonita site)
Turah fishing access (WQB samples

at bridge just downstream)
0.5 miles above mouth (WQB site is

USGS station 5.6 mi NE of Bonner)

** below Russell Street bridge
** Shuffield site

* At mill's bioassay shack

(WQB site at Harper Bridge — HB)

* Ken Cyr's land at Huson (HU)

(WQB site downstream of RR bridge)

* 1 mile upstream of Petty Creek

(WQB site just upstream of creek)
** Bridge in downtown Superior

* 3 miles downstream of St. Regis

(WQB site at Hwy 200 bridge
just above confluence)

* Flathead River near Hwy 200 mile 88

(WQB site near reservation
boundary)

* Across river from Plains fairgrounds

(WQB site above Thompson Falls
Reservoir)

** Sample site same as WQB site

* Sample site near or in same river reach as WQB site,
so expected to have similar water quality

TABLE 2. CLARK FORK RIVER NUTRIENT DATA, SUMMER 1990

DATE

Site

Air T

H20 T

PH

SRP

N03

07/11/90

ws

35

24

8.4

0.045

0.04

07/17/90

ws

WQB

0.053

0.03

07/24/90

ws

21.0

19.0

8.1

0.058

0.02

08/01/90

ws

20.0

18.0

7.6

0.040

0.01

08/06/90

ws

WQB

0.030

0.01

08/14/90

ws

30.0

22.5

7.8

0.092

0.02

08/21/90

ws

18.0

16.5

8.5

0.067

0.01

08/28/90

ws

19.0

15.0

9.0

0.019

<

0.01

09/06/90

ws

25.0

18.0

9.0

0.044

<

0.01

09/12/90

ws

WQB

0.040

0.01

09/17/90

ws

16.5

15.0

8.6

0.043

<

0.01

10/02/90

ws

9.5

11.0

8.45

0.044

0.02

07/11/90

DL

31

22

7.8

0.015

0.03

07/18/90

DL

WQB

0.005

0.10

07/24/90

DL

21.5

17.5

7.9

0.013

0.11

08/01/90

DL

28.0

17.0

6.3

0.008

0.13

08/07/90

DL

WQB

0.006

0. 12

08/14/90

DL

30.0

20.0

7.6

0.006

0.08

08/21/90

DL

20.0

16.0

8.2

0.019

0.09

08/28/90

DL

23.0

14.0

8.2

0.022

0.06

09/06/90

DL

26.0

16.5

8.4

0.008

0.04

09/11/90

DL

WQB

0.003

0.04

09/17/90

DL

21.0

14.0

8.4

0.003

0.05

10/02/90

DL

9.5

10.0

8.2

0.004

0.15

07/11/90

ALBF

33

23

7.7

0.050

<

0.01

07/18/90

ALBF

WQB

0.068

0.03

07/24/90

ALBF

20.0

15.0

8.3

0.060

<

0.01

08/01/90

ALBF

18.0

15.0

7.2

0.052

<

0.01

08/07/90

ALBF

WQB

0.047

<

0.005

08/14/90

ALBF

27.0

19.0

7.8

0.045

<

0.01

08/21/90

ALBF

17.0

14.0

8.5

0.034

0.03

08/28/90

ALBF

15.0

13.0

8.2

0.024

<

0.01

09/06/90

ALBF

23.0

15.0

8.2

0.019

<

0.01

09/11/90

ALBF

WQB

0.012

<

0.005

09/17/90

ALBF

15.0

12.0

8.0

0.013

<

0.01

10/02/90

ALBF

7.0

9.5

8.5

0.014

<

0.01

TABLE 2. CLARK FORK RIVER NUTRIENT DATA, SUMMER 1990 (cont)

DATE

Site

Air T H20 T

PH

SRP

NO 3

07/11/90

BLBF

33

23

7.75

0.033

<

0.01

07/24/90

BLBF

22.0

18.0

8.1

0.034

<

0.01

08/01/90

BLBF

27.0

18.0

8.1

0.030

<

0.01

08/14/90

BLBF

32.0

21.0

7.7

0.024

<

0.01

08/21/90

BLBF

20.0

16.5

8.5

0.024

0.01

08/28/90

BLBF

23.0

16.5

8.5

0.020

<

0.01

09/06/90

BLBF

28.0

18.0

8.50

0.017

<

0.01

09/17/90

BLBF

19.0

15.0

8.6

0.014

<

0.01

10/02/90

BLBF

7.0

9.5

8.5

0.014

<

0.01

07/11/90

GC

35

23

7.65

0.031

<

0.01

07/18/90

GC

WQB

0.034

<

0.01

07/24/90

GC

22.5

18.5

8.0

0.037

<

0.01

08/01/90

GC

27.0

19.0

7.7

0.032

<

0.01

08/07/90

GC

WQB

0.029

<

0.01

08/14/90

GC

33.0

21.0

8.1

0.026

<

0.01

08/21/90

GC

21.5

17.0

8.6

0.026

<

0.01

08/28/90

GC

23.0

17.5

8.7

0.020

<

0.01

09/06/90

GC

27.0

19.0

8.70

0.018

<

0.01

09/12/90

GC

WQB

0.011

0.03

09/17/90

GC

19.5

16.0

8.8

0.013

<

0.01

10/02/90

GC

9.0

9.5

8.6

0.014

<

0.01

07/11/90

DR

34

23

7.5

0.038

<

0.01

10/02/90

DR

7.5

9.5

8.5

0.021

<

0.01

07/24/90

BC

23.0

19.0

8.1

0.037

<

0.01

08/01/90

BC

31.0

21.0

7.6

0.037

<

0.01

08/14/90

BC

37.0

24.0

8.0

0.043

<

0.01

08/21/90

BC

26.0

17.5

8.5

0.046

<

0.01

08/28/90

BC

28.0

18.5

8.6

0.043

<

0.01

09/06/90

BC

29.0

19.5

8.60

0.030

<

0.01

09/17/90

BC

23.0

17.0

8.8

0.021

<

0.01

10/02/90

BC

10.5

9.8

8.55

0.016

<

0.01

TABLE 2. CLARK FORK RIVER NUTRIENT DATA, SUMMER 1990 (cont)

DATE

Site

Air T

H20 T

PH

SRP

NOB

07/11/90

BVT

34

24

7.6

0.024

<

0.01

07/17/90

BONITA

WQB

0.016

<

0.01

07/24/90

BVT

22.0

19.0

8.3

0.020

<

0.01

08/01/90

BVT

29.0

21.5

7.9

0.017

<

0.01

08/05/90

BONITA

WQB

0.009

<

0.01

08/14/90

BVT

28.0

18.0

7.4

0.007

<

0.01

08/21/90

BVT

23.0

18.0

8.5

0.039

0.02

08/28/90

BVT

25.0

19.0

8.5

0.032

<

0.01

09/06/90

BVT

29.0

20.0

8.70

0.022

<

0.01

09/10/90

BONITA

WQB

0.016

<

0.01

09/17/90

BVT

21.0

17.0

8.8

0,009

<

0.01

10/02/90

BVT

10.5

11.5

8.55

0.007

<

0.01

07/11/90

TU

34

22.5

7.6

0.011

<

0.01

07/17/90

TU

WQB

0.006

<

0.01

07/24/90

TU

18.0

17.0

8.1

0.005

<

0.01

08/01/90

TU

31.0

20.0

7.7

0.008

<

0.01

08/14/90

TU

38.0

24.0

8.3

0.005

<

0.01

08/05/90

TU

WQB

0.006

0.03

08/21/90

TU

21.0

18.0

8.5

0.009

0.02

08/28/90

TU

28.0

18.0

8.6

0.016

<

0.01

09/06/90

TU

31.0

19.5

8.70

0.010

<

0.01

09/10/90

TU

WQB

0.007

<

0.01

09/17/90

TU

20.0

16.0

8.8

0.005

0.02

10/02/90

TU

10.0

11.0

8.7

0.004

<

0.01

07/12/90

BE

34

21

7.9

0.002

<

0.01

07/17/90

BE

WQB

0.002

<

0.01

07/24/90

BE

18.0

17.0

8.1

0.010

<

0.01

08/01/90

BE

31.0

19.0

7.6

0.002

<

0.01

08/05/90

BE

WQB

0.002

0.01

08/14/90

BE

37.0

22.0

7.8

0.003

0.01

08/21/90

BE

23.0

17.0

8.6

0.036

0.02

08/28/90

BE

29.0

17.5

8.6

0.002

<

0.01

09/06/90

BE

32.5

19.5

8.60

0.005

<

0.01

09/10/90

BE

WQB

0.001

<

0.01

09/17/90

BE

19.5

15.0

8.7

0.002

<

0.01

10/02/90

BE

10.0

9.0

8.6

0.001

<

0.01

TABLE 2. CLARK FORK RIVER NUTRIENT DATA, SUMMER 1990 (cont)

DATE

Site

Air T H20 T

PH

SRP

N03

07/12/90

ABVSTP

20

7.8

0.009

<

0.01

07/17/90

ABVSTP

WQB

0.006

<

0.01

07/26/90

ABVSTP

19.0

17.0

7.4

0.004

<

0.01

07/31/90

ABVSTP

21.0

20.0

7.7

0.006

<

0.01

08/05/90

ABVSTP

WQB

0.006

<

0.01

08/16/90

ABVSTP

33.0

23.0

7.7

0.008

0.01

08/23/90

ABVSTP

21.0

17.0

8.5

0.005

0.01

08/30/90

ABVSTP

23.0

17.0

8.5

0.006

<

0.01

09/06/90

ABVSTP

20.0

8.50

0.010

<

0.01

09/10/90

ABVSTP

WQB

0.004

<

0.01

09/21/90

ABVSTP

10.5

13.5

8.7

0.003

<

0.01

09/28/90

ABVSTP

15.0

0.003

<

0.01

07/12/90

BELSTP

20.5

7.7

0.015

0.05

07/17/90

BELSTP

WQB

0.015

0.05

07/26/90

BELSTP

15.0

16.0

7.6

0.023

0.04

07/31/90

BELSTP

23.0

21.0

7.7

0.024

0.04

08/05/90

BELSTP

WQB

0.012

0.04

08/16/90

BELSTP

35.0

22.0

7.8

0.031

0.05

08/23/90

BELSTP

21.0

18.0

8.4

0.026

0.03

08/30/90

BELSTP

18.0

17.0

8.4

0.019

0.02

09/06/90

BELSTP

19.0

8.65

0.030

0.03

09/10/90

BELSTP

WQB

0.025

0.09

09/21/90

BELSTP

15.0

14.0

8.7

0.028

0.07

09/28/90

BELSTP

16.0

8.2

0.033

0.05

07/12/90

ABVMILL

31

8

0.008

0.02

07/17/90

ABVMILL

WQB

0.006

0.02

07/26/90

ABVMILL

17.0

17.0

7.3

0.010

0.03

07/31/90

ABVMILL

23.0

21.0

7.6

0.008

0.05

08/05/90

ABVMILL

WQB

0.006

0.05

08/16/90

ABVMILL

31.0

22.0

7.6

0.073

0.05

08/23/90

ABVMILL

26.0

17.0

8.5

0.011

0.05

08/30/90

ABVMILL

24.0

16.0

7.6

0.005

0.04

09/06/90

ABVMILL

18.5

8.40

0.006

0.03

09/10/90

ABVMILL

WQB

0.005

0.02

09/21/90

ABVMILL

19.0

14.0

8.2

0.006

0.04

09/28/90

ABVMILL

13.5

7.9

0.010

0.03

07/13/90

BELMILL

22

7.9

0.005

<

0.01

07/17/90

BELMILL

WQB

0.006

<

0.005

07/26/90

BELMILL

16.0

16.0

6.4

0.010

0.07

07/31/90

BELMILL

25.0

20.5

7.6

0.015

0.04

08/06/90

BELMILL

WQB

0.005

0.01

08/16/90

BELMILL

35.0

22.0

7.6

0.028

0.03

08/23/90

BELMILL

25.0

17.0

8.5

0.014

0.05

08/30/90

BELMILL

18.0

12.0

12.0

0.010

0.03

09/06/90

BELMILL

19.5

8.60

0.007

<

0.01

09/10/90

BELMILL

WQB

0.007

<

0.005

09/21/90

BELMILL

22.0

15.5

8.7

0.007

0.02

09/28/90

BELMILL

15.0

■?

0.008

0.02

TABLE 2. CLARK FORK RIVER NUTRIENT DATA, SUMMER 1990 (cont)

DATE Site Air T H20 T pH SRP N03

37.5
29.0
25.0

26.5

35
19.5
26.0
26.0

35
22.5
22.0
26.0

35
20.0
22.0
26.0

32
19.5
21.5

26

SITES IDENTIFIED IN TABLE 1

AIR T = AIR TEMPERATURE, H20 T = WATER TEMPERATURE, BOTH IN DE
SRP = SOLUBLE REACTIVE PHOSPHORUS IN ppb
N03 = NITRATES AND NITRITES IN ppb

WQB = MONTANA WATER QUALITY BUREAU

07/13/90

ALB

07/17/90

ALB

WQB

07/31/90

ALB

08/06/90

ALB

WQB

08/23/90

ALB

09/06/90

ALB

09/10/90

ALB

WQB

09/21/90

ALB

09/28/90

ALB

07/13/90

SU

07/18/90

SU

WQB

07/31/90

SU

08/06/90

SU

WQB

08/23/90

SU

09/10/90

SU

WQB

09/21/90

SU

07/13/90

AFH

07/18/90

AFH

WQB

07/31/90

AFH

08/07/90

AFH

WQB

08/23/90

AFH

09/11/90

AFH

WQB

09/21/90

AFH

07/13/90

FH

07/18/90

FH

WQB

07/31/90

FH

08/08/90

FH

WQB

08/23/90

FH

09/11/90

FH

WQB

09/21/90

FH

07/13/90

BFH

07/18/90

BFH

WQB

07/31/90

BFH

08/08/90

BFH

WQB

08/23/90

BFH

09/11/90

BFH

WQB

09/21/90

BFH

21

7.9

0.008

<

0.01

0.005

<

0.01

21.5

7.4

0.012
0.006

0.03
0.01

18.0

7.5

0.011

0.05

20.5

8.40

0.007

<

0.01

0.007

<

0.01

16.0

8.7

0.006

<

0.01

15.0

8

0.008

<

0.01

23

7.9

0.006

<

0.01

0.003

<

0.01

20.5

7.7

0.014

<

0.01

0.005

<

0.01

17.0

7.5

0.024

0.05

0.005

<

0.01

16.5

8.8

0.006

<

0.01

22

7.6

0.005

<

0.01

0.003

<

0.01

20.5

7.4

0.005

0.01

0.006

<

0.01

20.0

7.1

0.005

<

0.01

0.003

<

0.01

15.5

8.7

0.004

0.02

25

7.8

0.005

<

0.01

<

0.001

<

0.01

21.0

7.4

0.032

<

0.01

0.002

<

0.01

19.5

7.5

0.004

<

0.01

0.001

<

0.01

19.0

8.9

0.002

0.02

25

7.6

0.002

0.01

0.007

<

0.01

21.5

7.3

0.002

<

0.01

0.007

<

0.01

20.0

7.2

0.006

0.01

0.009

<

0.01

18

8.7

0.002

0.02

CLARK FORK RIVER NUTRIENT DATA, SUMMER 1990

LIST OF FIGURES IN APPENDIX N

NUTRIENT LEVELS IN THE CLARK FORK RIVER, JULY-OCTOBER, 1990
SOLUBLE REACTIVE PHOSPHORUS (SRP)
NITRATES & NITRITES (N03)

FIG IN. MAXIMUM, MINIMUM, MEDIAN SRP, ALL SITES

FIG 2N. WEEKLY SRP VALUES AT WARM SPRINGS AND DEER LODGE

FIG 3N. WEEKLY SRP VALUES AT 5 UPPER RIVER SITES

FIG 4N. WEEKLY SRP VALUES AT 4 UPPER RIVER SITES,

INCLUDING 2 NOT SAMPLED BY THE WATER QUALITY BUREAU

FIG 5N. WEEKLY FLOW AND SRP AT GOLD CREEK

FIG 6N. WEEKLY SRP IN MIDDLE RIVER

FIG 7N. MAXIMUM, MINIMUM, MEDIAN N03 ALL SITES

FIG 8N. WEEKLY N03 LEVELS, WARM SPRINGS TO ABOVE LITTLE BLACKFOOT

FIG 9N. WEEKLY FLOWS AND N03 AT DEER LODGE

FIG ION. WEEKLY N03 IN MIDDLE RIVER

FIG UN. WEEKLY SOLUBLE INORGANIC NITROGEN (SIN), MIDDLE RIVER
(SIN = NITRATES, NITRITES AND AMMONIA)

SITES IDENTIFIED ON TABLE 1.

MEDIAN SRP VALUES VS RIVER MILE

WARM SPRINGS TO TURAH

20 40 60 80 100 120

CLARK FORK RIVER LOCATION IN MILES

140

MEDIAN SRP VALUES VS RIVER MILE

ABVSTP TO BELOW FLATHEAD

140 160 180 200 220 240 260

CLARK FORK RIVER LOCATION IN MILES

280

- MAXIMUM - MINIMUM

MEDIAN

Figure IN

WEEKLY SRP AT WS AND DL

7/11 7/17' 7/24 8/1 8/6' 8/14 8/21 8/28 9/6 9/11' 9/17 10/2
WEEKLY MONITORING (WQD = )

DL

-^4^

WS

Figure 2N

WEEKLY SRP IN UPPER CLARK FORK

0.07

7/11 7/17' 7/24 8/1 8/6" 8/14 8/21 8/28 9/6 9/11' 9/17 10/2
WEEKLY MONITORING (WQB = ' )

-»- DL
-X- TU

ALBF
BVT

^ GC

Figure 3N

WEEKLY SRP IN UPPER CLARK FORK

0.07

7/11 7/17' 7/24 8/1 8/6' 8/14 8/21 8/28 9/e 9/11' 9/17 10/2
WEEKLY MONITORING (WQB = ' )

BLBF

-fcj-

GC

-H- BC

BVT

Figure 4N

CLARK FORK AT GOLD CREEK

WEEKLY FLOW AND SRP

0.04

320

0.01

7/11 7/17' 7/24 8/1 8/6' 8/14 8/21 8/28 9/6 9/11 9/17
WEEKLY MONITORING

160

SRP

FLOW

Figure 5N

WEEKLY SRP IN MIDDLE RIVER

0.04

ABVMILL = .073 I

(D

CO

0.03

0.02

0.01-

7/11 7/17' 7/24 8/1 8/6" 8/14 8/21 8/28 9/6 9/11" 9/17 10/2
WEEKLY MONITORING (WQB = ' )

■*- ABVSTP — *— BELSTP ^^ ABVMILL
-s- BELMILL ^<- ALB

Figure 6N

MEDIAN NITRATE VS RIVER MILE

WARM SPRINGS TO TURAH

20 40 60 80 100 120

CLARK FORK RIVER LOCATION IN MILES

140

ABVSTP TO BELOW FLATHEAD

0.16

140 160 180 200 220 240 260

CLARK FORK RIVER LOCATION IN MILES

280

- MAXIMUM - MINIMUM

MEDIAN

Figure 7N

WEEKLY NITRATE AT WS, DL, AND ALBF

0.16

7/11 7/17 7/24 8/1 8/6' 8/14 8/21 8/28 9/6 d/11'9/17 10/2
WEEKLY MONITORING (WQB = ' )

WS

DL

-*^ ALBF Figure 8N

CLARK FORK AT DEER LODGE

FLOW VS NITFRATE

0.14

8/1

0.12

-J

O 0.1

z

uj 0.08

t 0.06

0.04

8^"
7/24
7/17'

8/14

8/21

MUDDY

B2f

9 17

0.02

9 1

^

7/11

20 40 60 80 100 120 140 160

FLOW IN CFS

Figure 9N

WEEKLY NITRATE IN MIDDLE CLARK FORK

0.09

7/11 7/17' 7/24 8/1 8/6' 8/14 8/21 8/28 9/6 9/11' 9/17 10/2
WEEKLY MONITORING (WQB = ' )

ABVSTP

BELSTP -^^ ABVMILL ^ BELMILL

Figure ION

140

SOLUBLE INORGANIC NITROGEN
SUMMER, 1990

Figure UN

190 200 210 220 230 240 250 260 270 280

DAY OF YEAR

APPENDIX T — UPPER RIVER RECIPROCAL TRANSPLANT STUDY

In summer, the upper Clark Fork is generally much higher in
phosphorus and lower in nitrogen than is the middle river
(exceptions include the high P levels below the Missoula WWTP and
the high N levels in Deer Lodge) . In addition, the algal
community is dominated by Cladophora. a filamentous green alga,
while the middle river is dominated by a mixed community of
diatoms. It has proven difficult to study Cladophora in
artificial streams and determine what nutrient levels limit
standing crop of this alga. Instream studies on artificial
substrates are hampered by the long time required for this alga
to become established on substrates. Hence we attempted to study
the response of this alga to different nutrient levels by
transplanting rocks already colonized with Cladophora to
different parts of the river. Sites were chosen that differed in
nitrogen and phosphorus levels.

In summer the Deer Lodge (DL) site is relatively high in nitrate
(40-120ppb) and low in phosphorus (5-20ppb) , while the Warm
Springs (WS) site is high in P (3 0-90ppb) and lower in nitrate
(10-30ppb) . Nitrate is usually below detection (<10ppb) at the
sites below Gold Creek (GC) and Bear Creek (BC) . Both these sites
have moderately high P levels (10-35 ppb) , but in 1990 GC had the
higher P levels early in the summer while the BC area had higher
P levels later in summer. If nutrients are the most important
factors limiting accumulation and if P is more limiting than N,
the Deer Lodge site should show the least accumulation, followed
by GC and BC with WS showing the most accumulation. If N is more
limiting at these levels, DL should show the most accumulation,
followed by WS, then GC and BC.

METHODS — In mid July, rocks colonized with Cladophora (and other
periphyton) were collected from the river bed just below the
confluence with Warm Springs creek (WS) and at Beavertail Hill
campground (BVT) . Rocks from both these source sites were
transported in river water in coolers to the four test sites
(WS,DL,GC,BC) . Here rocks were placed in shallow aluminum pans
which were dug into the substrate so the rocks were almost flush
with the surrounding substrate. A ring of vaseline was applied
to the lip of the pans to reduce movement of insects into the
pans. Pans were checked weekly and kept at 2 0-3 0cm deep in
rapidly flowing water (0.3m/s +/- O.lm/s).

Known areas of mixed attached algae community were sampled from
ten rocks from each source site at the beginning of the study (7-
17-90) . Ten rocks from each source site were collected from the
transplant pans at each test site after one and two months (mid
August and mid September) . In some cases only 5 rocks were
available because of loss of pans. The algal material was
analyzed for chlorophyll a and for ash free dry weight by
Standard Methods (1985) .

RESULTS — Initial algal biomass levels (on 7/17) and those present
after 1-2 months at test sites appear in figures Tl and T2 .

Initially, the Cladophora from WS was 3-6 inches long and bright
green. The Cladophora from BVT was only one to two inches long
and was covered with a thick mass of brown diatoms. Within a
month, the algae on the rocks from both sites took on the
appearance of the Cladophora at the test site (similar color and
growth form) . Over the first month, the WS rocks incubated at WS
increased in both chl a and AFDW (filaments had increased to 6 to
8 inches in length) ; however, biomass dropped dramatically in the
second month. By mid September, the rocks were stripped bare and
were covered with caddisflies and other aguatic insects. The BVT
rocks decreased in biomass at the WS site in both months. The
first month the algae simply appeared to be less healthy, the
second month these rocks also appeared to be stripped bare.

At the Deer Lodge site, the algae on the rocks from both WS and
BVT took on the rich dark green color and the very long growth
form (filaments up to 2 feet long) that characterized the rocks
at the Deer Lodge site. By mid August, the biomass on the WS
rocks increased at this site but the biomass on the BVT rocks did
not increase significantly. The Deer Lodge site was vandalized
before the September sampling.

At the Gold Creek site, the transplanted algae assumed the bright
green color and long filaments that characterized this site.
Initially, chlorophyll levels actually decreased at this site,
while AFDW increased on the WS rocks but not on the BVT rocks.
But by mid September, the WS rocks showed a significant increase
in biomass, while biomass levels on the BVT rocks were too
variable to make valid comparisons.

At the Bear Creek site, the transplanted algae took on the pale
yellowish green color and long filaments that characterized this
site. With the exception of an initial increase in AFDW on the WS
rocks, there was little change in biomass on the rocks incubated
at this site.

How could the filaments increase in length without increasing
significantly in biomass? The data presented here is based on
samples collected from 10 sq. cm. on the top center surface of
each rock. The remaining algal material on the rock was also
collected and the rock's upper surface area was estimated. This
latter approach yielded higher average values; however, these
measurements were much more variable and so generally did not
show significant increases either. More replicates would no doubt
have narrowed the confidence interval and shown more significant
differences. Forty rocks were placed at each of the 4 test sites
to allow two samplings of 10 from each source site. However,
losses of some pans caused only 5 replicates to be available for
later sampling at some sites. Finding and transporting more than
160 rocks with similar amounts of Cladophora would be difficult.

What can be made of all this? If one looks at AFDW, there was
little change in the BVT algae at all the sites (except for the
decrease at WS which may have been due to metals toxicity or
heavy macroinvertebrate grazing) . The AFDW of the WS algae showed

similar increases at all the sites in August, but in September
lost AFDW at WS & BC while gaining at GC. If one looks at
chlorophyll and the first month of incubation, higher nitrogen
sites (DL & WS) exhibited a significant increase in chlorophyll
on the WS rocks while the lower nutrient sites farther downstream
did not. However, the September sampling suggests that factors
other than nutrients were more important in determining
chlorophyll levels at this time. As for the BVT rocks, the very
great variability in two of the samplings make it difficult to
discern a pattern in these data. BVT data will be more closely
inspected for outliers to determine if the confidence intervals
can be narrowed.

It should be noted that with the exception of the final sampling
at WS, all the rocks visually appeared to have much more algal
biomass after incubation — Cladophora filaments were much longer
and more noticeable. This increase is observed in the WS data.
However, the heavy layer of diatoms on the BVT Cladophora seemed
to be lost, and the increase in Cladophora biomass did not appear
to offset the loss of the diatoms.

Because the loss of the diatoms obscures the increase in
Cladophora on the BVT rocks, I will concentrate on the WS rocks.
Perhaps the most interesting result is the similarity of the
response at WS and DL despite the very different N:P ratios of
these sites. The P levels at these sites are both high enough
that little difference in P limitation would be expected but N
levels are low enough at WS that N limitation of standing crop
might be expected. N levels did drop over the summer and perhaps
this partly explains the low September algal levels at WS.
Aquatic insects were much more noticeable at WS in September and
differential grazing may explain the different algae levels;
however, in late summer, 1989, there were more aquatic insects at
DL than at WS (Rokosch and Watson, unpublished data) .

The two downstream sites {GC & BC) had lower nitrate levels than
WS or DL but had P levels intermediate between these sites. Their
P levels were high enough to expect P saturation, however, so
their main response should be to the lower N levels. These sites
initially produced less chlorophyll than the two upper river
sites but not less AFDW. The loss of the DL site for the second
sampling is a problem — if this site had shown the highest biomass
levels at the second sampling, algal levels would have been
approximately proportional to nitrate levels. Certainly the
surrounding rocks at the DL site seemed to have the darkest green
and most massive Cladophora growths at this time, however, visual
evaluations are not always in agreement with quantitative sampling.

As for the effects of differential grazing at the sites — in late
summer 1989, the GC and BC sites had similar aquatic insect
abundances and biomass to the WS sites but less than at DL
(Rokosch and Watson, unpub. data). Perhaps the WQB's 1990 insect
data would shed light on whether differential grazing may help
explain these patterns.

So perhaps the clearest conclusion from all this is that by mid
August the WS rocks had accumulated more chlorophyll at the high
nitrate sites than at the low nitrate sites. P seemed to be
sufficient at all sites since the low P DL site accumulated more
than did higher P sites. It would seem necessary to reduce P
levels below the 10 ppb levels observed at DL to achieve much
reduction in algae levels through P limitation. A careful
evaluation of point sources is necessary to determine if this is
possible by point source control alone. As for limitiation by
nitrates, the relationship between algal levels and nitrate
levels was not that strong suggesting that other factors are
important in regulating algal accumulation. Reduction of nitrate
below detection levels might result in some reduction in
chlorophyll levels but did not appear to reduce AFDW levels.

The difficulty of drawing conclusions from instream studies given
the many confounding variables is obvious. But it is clear that
nutrient differences between these sites were not an overriding
factor in determining algal levels as might be expected where
nutrient levels are fairly high.

Acknowledgements

This work was supported by the Clark Fork monitoring project with
funds from the Clean Water Act Section 525 study. The Grant Kohrs
Historic Ranch site provided access to the river at Deer Lodge.

ALGAL BIOMASS ON WARM SPRINGS ROCKS

TRANSPLANTED TO SITES IN CLARK FORK

ax)

600

400

200

7-17 8-1"9-17 8-11 8-11 9-17 8-11 &-17
@WAR^ SP. @OEER LOGOLD CR QOEAR CR.

STES AND DATES BOMASS mPLYZED

ALGAL BIOMASS ON BEAVERTAIL ROCKS

TRANSPLANTED TO SITES IN CLARK PORK

E.

800

600

400

200

0

7-17 8-1-8-17 8-11 8-11 9-17 8-11 9-17
eWAfW SP. ©DEER LOGOLD CR @BE>R CR

SITES mo DATES aOMASS mfiLYZH)

Figure 11

g

ALGAL BIOMASS ON WARM SPMNGS ROCKS
TRANSPLANTED TO SITES IN CLARK PORK

3CX)

250
2CX)
150
100
50

7-17 8-1-9-17 8-11 8-11 9-17 8-11 9-17
@WAf*4 SP. (80EE? LOGOLD CR QBEAR CR

STTES NMD DATES BCMASS ANALYZED

ALGAL BIOMASS ON BEAVERTAIL ROCKS

TRANSPLANTED TO SITES IN CLARK PORK

g

300

250
200
150
100--
50-
0

-50

7-17 8-1-B-17 8-11 8-11 9-17 8-11 9-17
©WARM SP. ©DEER LOGOD CR @8EAR CR

SntS AMD DATES BCMASS AMALYZED

Figure 12

APPENDIX M — DESCRIPTION OF RIVER ALGAE ACCUMULATION MODEL

(PERISIM)

PERISIM is a computer program written in IBM BASICA which
simulates the accumulation of attached algal biomass on river
rocks over time. In its current form the model has been shown to
simulate with reasonable accuracy the accumulation of biomass of
the mixed diatom community that characterizes the middle and
lower Clark Fork River during the summer growing season.

The model simulates attached algal biomass accumulation by
estimating the production of new biomass (via photosynthesis) and
the loss of biomass (to respiration and sloughing) every hour.
Gains in biomass are added to the previous estimate of biomass
and losses are subtracted. Thus the mass of algal material
gradually increases or decreases depending on whether gains
exceed losses or vice versa. That is, the change in biomass over
time is simulated by the equation:

Bt = Bt-1 + Bt-1 * (rates of growth -respiration -sloughing)

where Bt = biomass at time t, Bt-1 = biomass at previous point in
time. The time step used in this model is one hour.

The ecological processes simulated are algal growth (via
photosynthesis) , respiration and sloughing (or loss of biomass
due to detaching of algae from the substratum and washing away) .
Each of these will be discussed separately.

Algal growth — Algal growth has been modeled with varying levels
of complexity. While some of the more complex methods are
considered to depict nutrient uptake and growth more
realistically, some of the simpler methods often produce
estimates that are as accurate (or more accurate) . In PERISIM,
algal growth is a function of available light, nutrients and
temperature. Under optimum conditions of light, nutrients and
temperature, algal biomass increases at a maximum exponential
rate (that is, some fraction of the existing biomass is added
each day) . When any of these factors is less than optimum, the
rate of increase is reduced by a specific formula. That is, the
growth rate is estimated by:

u = Umax * LD * ND * TD

where u is the rate of increase, umax is the maximum rate of
increase of which that particular community is capable, and LD,
ND and TD are the light, nutrient and temperature dependent
functions that reduce the maximum rate of increase to that
expected under these suboptimum conditions (Lehman et al. 1975) .

The light dependence formula is that used by Steele (1964, 1965)
which recognizes that photosynthesis increases with light up to a
point, then becomes saturated and finally inhibited at higher
light levels. This relationship is depicted by the formula:

LD or light dependence of growth = (IZ/IOPT) *exp{ 1-IZ/IOPT)

where IZ is the light at depth z in calories/cm2/day , lOPT is the
optimum light level. PERISIM estimates the surface light level
for that time of year and time of day at the latitude being
simulated. The light is then attenuated for the depth being
simulated, using the formula:

IZ = lo * exp(-nz)

where lo is the surface light and n is the extinction coefficient.

Obviously, growth exhibits an increasing and decreasing response
to temperature over a wide range of temperatures. But
temperatures in the Clark Fork rarely exceed the optimum
(around 2 5 C for many species) , hence growth may be represented
as a simple increasing function of temperature by the formula:

TD or temperature dependence of growth = 0.04*(T)

where T is the ambient water temperature in centigrade.

The effect of nutrients on algal growth has been modeled in
numerous ways. The formula used here is the simple Monod or
Michaelis Menten formulation which assumes that algal growth
rates can be estimated from ambient available nutrient levels.
When estimating accumulation rates of a mixed community this
method does as well or better than more complex methods (DiToro
1981) . Hence nutrient dependence is calculated as:

ND or nutrient dependence of growth = P/ (P+Kp)

where P is the concentration of soluble reactive P (SRP) in ppb
in the ambient water and Kp is the half saturation constant or
the concentration of P which produces half of the maximum growth
rate. Note that when P = Kp then ND is 0.5 and the growth rate is
1/2 of the maximum rate. When P is much > Kp, this term
approaches 1 and the growth rate approaches the maximum growth
rate. When P is much < Kp, this term approaches zero. However, Kp
is very low, 2ppb or less according to most research. Hence, when
ambient SRP is low relative to Kp, it is below detection.

Respiration is modeled as a simple function of temperature
similar to the temperature dependence of growth:

Respiration rate R = 0.04 * T * Kr

Where Kr is maximum daily respiration rate (0.1/day) and T is as
before. This equation produces respiration rates similar to those
produced by the equation developed by Graham et al. (1982) in
which R = 0.151*(0.025T + 0.1).

Sloughing is a function of water velocity and turbulence and the
vigor of the algal community. Artificial stream studies show
that, following colonization, algal biomass increases at a rapid
exponential rate than levels off as losses in biomass come to
balance gains. As long as environmental conditions do not change
greatly, a dynamic eguilibrium biomass is established that seems
to be a function of ambient nutrient level. That is, under higher
nutrient levels, algal biomass accumulates to a higher level
before leveling off than it does under lower nutrient levels.
Based on the work of Bothwell (1989) and Watson (1990) , a formula
was developed that described this relationship between ambient
nutrient level and the maximum biomass sustained at this dynamic
equilibrium:

Bmax = 10 + 60 * P/ (P+Kb)

where 10 g/m2 is the biomass sustained at P levels below
detection level, 60 is the maximum biomass sustained when P
saturates standing crop and Kb is the SRP level that produces a
standing crop that is about half of the maximum level.

As the biomass at a site approaches the maximum biomass that can
be sustained given the nutrient levels there, sloughing
increases. This is accomplished by the formula:

Sloughing rate = SLMAX * B/Bmax

where SLMAX is the maximum sloughing rate (set at 1/2 the
standing crop per day) . This approach is similar to that used by
Auer and Canale (1980, 1982) except that their Bmax is fixed
rather than a function of ambient nutrient levels. Actually, Auer
and Canale made two uses of Bmax to limit the standing crop of
Cladophora. As B approached Bmax, the growth rate slowed due to
shading, nutrient limitation and waste buildup and the sloughing
rate increased. The formulations used were:

growth dependance = 1- (B/Bmax)

sloughing rate = max rate (B/Bmax) for a given wind speed

Model Inputs

The model estimates the light levels for each day and hour. The
environmental data that the model requires is water quality data,
specifically, water temperature (in degrees centigrade) and
nutrient levels (ppb of SRP) . These parameters were measured
weekly during the summers of 1988 and 1990 on the Clark Fork. The
model requires daily values and another simple program (PERIFILE)
was developed to produce daily water quality input data files
from the weekly measurements. An initial biomass value must be
specified, typically between 1 and 3 g/m2 is used for
simulations. For validation runs, the amount of biomass observed
on artificial substrates after one week of colonization was used
as the initial biomass and the simulation was started on that
date. The model also requires the values of the rate constants
and other constants in the simulation equations. These are
summarized in table 1.

Model Outputs

PERISIM estimates ash free dry weight of attached algae per sq.
meter of river bed for each day of the simulation. Ash free dry
weight is then converted to chlorophyll a by a conversion factor
(CCF) . This factor is 150 for high nutrient sites (like Below
Missoula STP and Harper Bridge — which tend to be richer in
chlorophyll) , 300 for low nutrient sites (above Missoula, Plains)
and 200 for moderate nutrient sites (all others) .

Limitations of the model

The intent of this modeling exercise was to find the simplest
model that would estimate the seasonal changes in algal biomass
levels in the middle and lower Clark Fork with acceptable
accuracy. The model was made more complex only as seemed
necessary to obtain accurate reproductions of the river's
observed algal levels. As was stated, the current version of the
model is capable of reasonably accurate reproductions of the
observed algal levels of 1988 and 1990 (as seen in figures 5
through 12 of the main body of this final report) .

Many refinements could be added to make the model more realistic,
and these refinements may or may not increase the accuracy of the
model's predictions. These include growth limitation by multiple
nutrients (particularly adding limitation by nitrogen) and the
effects of grazing, scouring flows, turbidity, and cloudy
weather. Adding these and other refinements to the model does not
greatly complicate the model, but it does greatly complicate the
problem of obtaining input data for a given year and a given
site. The less data hungry the model, the more useful it is, as
long as it provides reasonably accurate predictions.

The refinement that would most improve the model's usefulness in
predicting algal accumulation in the Clark Fork would be adding
nitrogen limitation to the model, and this is planned for the
future. However, there is much less information available on
growth parameters for N than for P.

In addition, it would be useful to modify the model so that it
could simulate the accumulation of the perennial alga Cladophora.
This is a more complex task since one river rock may be starting
the growing season with no Cladophora while another may be
starting with a growth that is several years old. Validating a
Cladophora model is also more difficult because it is harder to
obtain validation data. One must measure accumulation on natural
substrates or on artificial substrates that have been in the
river for a year or more. These efforts are more labor intensive
and the data is much more variable than is true of measurements
of diatom community accumulation. Obviously, this is a long term
project.

References

Auer, M. T. and R. P. Canale. 1980. Phosphorus uptake dynamics as
related to mathematical modeling of Cladophora at a site on
Lake Huron. J. Great Lakes Res. 6(l):l-7.

Bothwell, M. L. 1989. Phosphorus limited growth dynamics of lotic
periphytic diatom communities: areal biomass & cellular growth
rate responses. Can. J. Fish. Aquat. Sci. 46 (6) : 1293-1301.

Canale, R. P. and M. T. Auer. 1982. Ecological studies and
mathematical odeling of Cladophora in Lake Huron: 5. Model
development & calibration. J. Great Lakes Res. 8 (1) : 112-125.

DiToro, D. M. 1980. Applicability of cellular equilibrium and Monod
theory to phytoplankton growth kinetics. Ecol. Modelling
8:201-18.

Graham, J. M. , Auer, M. T. , Canale, R. P. Hoffman, J. P. 1982.
Ecological studies and mathematical modeling of Cladophora
in Lake Huron: 4. Photosynthesis & respiration as functions
of light and temperature. J. Great Lakes Res. 8(1): 100-11.

Hutchinson, G. E. 1975. A treatise on limnology. John Wiley &
Sons.

Lehman, J. T. , D. B. Botkin, and G. E. Likens. 1975. The
assumptions and rationales of a computer model of
phytoplankton population dynamics. Limnol. Oceanogr. 20:343-64.

Spain, J. D. 1982. BASIC microcomputer models in biology.
Addison-Wesley. Publ. Co. London.

Watson, V. J. 1981. Seasonal phosphorus dynamics in a stratified
lake ecosystem: an evaluation of external and internal loading
and control. PhD dissertation. Botany, Univ. Wisconsin-Madison.

1983. Application of a seasonal phosphorus dynamics

model to a lake ecosystem: assessing lake loading tolerance,
pp. 575-592. In Lauenroth, W. K. , G. V. Skogerboe and M.
Flug(eds.) Analysis of ecological systems: state of the art of
ecological modeling. Elsevier Scientific Publ. Amsterdam.

1990. Improving the biotic resources of the upper Clark

Fork River. Streamside artificial stream studies. Report to
Montana Dept. Natural Resources and Conservation.

Whitton, B. A. 1967. Studies on growth of riverain Cladophora in
culture. Archiv. fur Mikrobiologie. 58:21-29.

RATE CONSTANTS AND OTHER PARAMETERS USED IN PERISIM

SYMBOL
M

VALUE USED
378 cal/cm2/day

VAR

249 cal/cm2/day

VARDL

4hrs

N

0.5

MUMAX

1 per day

(ie, doubles daily)

KI

10 cal/cm2/day

lOPT

15 cal/cm2/day

KPMU

2ppb

KPB

5ppb

SLMAX

0.5/day

KR

0.1/day

DEFINITION

mean annual dally light intensity

at 45 N latitude

seasonal variation of light intensity
either side of the mean

seasonal variation of daylength
either side of the mean

extinction coefficient of water

maximum growth rate of algae

SOURCE
Hutchinson 1975

half saturation constant for light
for photosynthesis

optimum light level for photosynthe

half saturation constant of phospho
for algal growth

half saturation constant of phospho
for algal standing crop

maximum daily sloughing rate
similar to 0.3/day used in

respiration rate coefficient
used in R = 0.04 • KR • T
produces values similar to
R = .151'(.025T + .l)usedby

Watson 1981, 1983
Whitton 1967

Watson 1981, 1983

explained
in text

Auer&Canalel980
Auer & Canale 1988

Watson 1981, 1983

Graham et al. 1982