 

 

«was

1002 LIBRARY
Michigan State
University

 

 

This is to certify that the
thesis entitled

THE EFFECTS OF LAMPRICIDE TREATMENTS ON
STREAM pH

presented by

REBECCA N. GANNON

has been accepted towards fulﬁllment
of the requirements for the

MS. /)degree In Fisheries and Wildlife

 

 

”I WJW/

 

Major Professor' 5 Signature
2r; Mam I” / 2657

Date

 

MSU is an affirmative-action, equal-opportunity employer

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

 

 

DATE DUE DATE DUE DATE DUE
nor 1 1 2009
113l09

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:/ClRC/DateDue.indd-p.1

THE EFFECTS OF LAMPRICIDE TREATMENTS ON STREAM pH
By

Rebecca N. Gannon

A THESIS
Subnnuedto
Michigan State University
in partial fulﬁllment of the requirements
for the degree of
MASTER OF SCIENCE
Department of Fisheries and Wildlife

2007

ABSTRACT
THE EFFECTS OF LAMPRICIDE TREATMENTS ON STREAM pH
By

Rebecca Nancy Gannon

Addition of lampricide 3-triﬂuoromethyl-4-nitrophenol (TF M) during sea
lamprey control treatments has been hypothesized to cause Signiﬁcant suppression of pH
levels in certain river systems negatively impacting ﬁsh populations. The goal of this
research was to evaluate possible mechanisms of pH suppression in order to adjust
chemical control treatment processes to allow for maximum killing effect on sea lamprey
(Petromyzon marinus) while protecting non-target resident ﬁsh species. The cause of the
noted pH suppression has been hypothesized to be related to the magnitude of plant
production within streams treated and the timing of chemical treatments. In order to
assess changes in plant production levels during a treatment, I measured diel variation in
dissolved oxygen, pH, temperature, and ﬂow rates before, during and after TFM
treatments. TF M additions in streams evaluated in 2005 and 2006 did not impact river
pH levels. This ﬁnding suggests that TFM applications in rivers with high plant biomass
do not or rarely Show the pH suppression earlier noted. Suppression appears to be the

result of Spatial and temporal pH variation among treated rivers.

ACKNOWLEDGMENTS

I would like to thank everyone who made this project possible. My advisor and
graduate committee chairman, Dr. Daniel Hayes, who helped me get through this time in
my life and supported me through the entire project. Dr. William Taylor, who brought
me in on this project and made it possible for me to get this degree. Dr. Merritt Turetsky,
for your guidance on everything plants. The staff of the US. Fish and Wildlife Service,
Ludington and Marquette Biological Stations and also the Department of Fisheries and
Oceans, Canada. In particular, Denny Lavis, Dorrance Brege, Brian Stephens, Alex
Gonzalez, and Jeff Sartor. Thank you also to Jason Krebill and Lois Misher who helped
me in the collection of data during my two ﬁeld seasons. I would not have been able to
collect my data or ﬁnish my project without all of their help, support and encouragement.
Michigan State University and the US. Fish and Wildlife Service provided the funding
for this project.

I would also like to thank my family and friends for their emotional and spiritual
support over the last three years. I would especially like to thank my sister Heather. She
has supported me my entire life and is always there when I need help, advice or a
shoulder to cry on. Thanks for being the best Sister a girl could ever ask for! My parents,
Dennis Gannon and Karen Tryc, my Sister Chelsea Childs, and my grandmother
Catherine Tryc for always pushing me towards my goals and supporting me throughout
my life. All of the wonderful friends that I met while at Michigan State. Thanks for all
of your encouragement and for keeping me grounded. Last, but certainly not least, I
thank God for all the blessings in my life and for helping me get through my time here. It

is through Him that all things are possible.

iii

TABLE OF CONTENTS

LIST OF TABLES .................................................................................... v
LIST OF FIGURES ................................................................................. vi
INTRODUCTION .................................................................................... 1
METHODS ............................................................................................ 7
RESULTS ............................................................................................ 16
DISCUSSION ....................................................................................... 39
REFERENCES ...................................................................................... 45

iv

LIST OF TABLES

Table l. The approximate number of TFM treatments (1980-2006) and ﬁsh kills that
have occurred in the United States (1994-2006) and in Canada (1991-2006). A
reportable kill is any treatment where more than 50 of any species have been
killed by TF M application. A sensitive Species is a species that is known to be
sensitive to TF M applications (e.g. other native lamprey, log perch, and mud
puppies). A non-sensitive species is a species that is not affected by normal TF M
levels (e.g., Chinook salmon and walleye) ............................................... 4

Table 2. Summary of study river attributes. Mean pH, dissolved oxygen (DO, mg/l),
temperature (C°), alkalinity (total alkalinity), discharge (cfs), total dissolved solids
(TDS, g/l) and miles of river treated with TF M. Rivers are arranged by ascending
pH levels ...................................................................................... 9

Table 3. Mean and die] amplitude of pH, dissolved oxygen, and temperature at each site
sampled. Averages are computed over entire sample period for each river. Diel
amplitude is computed as the mean difference in the daily minimum and
maximum .................................................................................... 18

Table 4. Summary of results obtained from General Linear Model (GLM) analysis of
differences in temperature, pH, and dissolved oxygen, comparing least square
means before, during, and after TFM exposure at each site .......................... 19

Table 5. Average river miles treated with TFM within the last 25 years. The
sample size is the number of water chemistry samples taken by the data loggers
deployed on each river ..................................................................... 20

Table 6. Average width, average depth, mean dry plant weight, dominate substrate, and
number of plots surveyed during plant biomass collection ........................... 36

Table 7. Laboratory plant experiment results. Mean and standard deviation for three
replications of each experiment (20% and 80% plant abundance) as well as the
control tank values are represented. Measurements were taken before, during and
after TFM was added to each experimental tank ....................................... 38

Table 8. Average light attenuation measurements taken on Bear Creek .................... 39

LIST OF FIGURES

Figure 1. Realtionship of pH and TF M (USFWS SLC Operating Manual). Sea lamprey
concentration is the minimum lethal concentration at which lamprey will die.
Brown trout concentration is a concentration at which TFM becomes toxic to
sensitive species such as brown trout. Treatment concentration is typically run at
75-85 % of the “brown trout” concentration .............................................. 3

Figure 2. Map of Michigan showing location of study rivers. Longitude and latitude
coordinates can be found in Appendix A ................................................. 8

Figure 3. Betsie River graphs of water pH, dissolved oxygen (mg/l) and temperature (C).
Bars indicate when TFM chemical block passed through the treatment logger
locations ..................................................................................... 22

Figure 4. Sturgeon River graphs of water pH, dissolved oxygen (mg/l) and temperature
(C). Bars indicate when TF M chemical block passed through the treatment logger
locations ..................................................................................... 24

Figure 5. Pere Marquette River graphs of water pH, dissolved oxygen (mg/l) and
temperature (C). Bars indicate when TFM chemical block passed through the
treatment logger locations ................................................................. 25

Figure 6. Muskegon River graphs of water pH, dissolved oxygen (mg/l) and temperature
(C). Bars indicate when TFM chemical block passed through the treatment logger

locations ..................................................................................... 27

Figure 7. Riﬂe River graphs of water pH, dissolved oxygen (mg/l) and temperature (C).
Bars indicate when TFM chemical block passed through the treatment logger
locations ..................................................................................... 29

Figure 8. White River graphs of water pH, dissolved oxygen (mg/l) and temperature (C).
Bars indicate when TFM chemical block passed through the treatment logger
locations ..................................................................................... 31

Figure 9. Bear Creek graphs of water pH, dissolved oxygen (mg/l) and temperature (C).
Bars indicate when TF M chemical block passed through the treatment logger
locations ..................................................................................... 32

Figure 10. Trail Creek graphs of water pH, dissolved oxygen (mg/l) and temperature (C).

Bars indicate when TFM chemical block passed through the treatment logger
locations ..................................................................................... 34

vi

Figure 11. Chippewa River graphs of water pH, dissolved oxygen (mg/l) and
temperature (C). Bars indicate when TFM chemical block passed through the
treatment logger locations ................................................................. 35

Figure 12. Relationship of mean dry plant density (g/mz) and mean diel pH amplitude
................................................................................................ 40

Figure 13. Relationship of mean dry plant density (g/mz) and mean diel DO amplitude
................................................................................................ 42

Figure 14. Relationship between mean diel dissolved oxygen amplitude and mean diel
pH amplitude ................................................................................ 43

vii

Sea lamprey (Petromyzon marinus) are parasitic predators that entered the Great
Lakes through the New York Finger Lakes (Crowe 1978). After the Welland Canal was
completed in 1829, sea lamprey were able to bypass Niagara Falls which had been a
natural barrier (Crowe 1978). By the 1940’s sea lamprey were present in all ﬁve Great
Lakes. Sea lamprey have wreaked havoc on the Great Lakes ﬁsheries Since the 1930’s,
contributing to reductions in the number of lake trout to the brink of extinction (Crowe
1978). In the late 1940’s, efforts by the American and Canadian governments were
initiated to control sea lamprey populations. The ﬁrst step was to understand this species
life cycle, and then to ﬁnd a method that would kill the lamprey with minimal harm to
other aquatic life. It was determined that the best control method was to attack the
lamprey when they are most concentrated in stream tributaries as spawning adults or as
juveniles (Crowe 1978). Chemical control was researched for several years with over
6000 chemicals tested. Between 1957 and 1958, several promising chemicals were tested
in the ﬁeld, two of which proved to be very effective, 3-triﬂuoromethyl-4-nitrophenol
(TF M) and 5-dichloro-4-nitrosalicylanilide (B73) (Crowe 1978).

Since 195 8, TFM has been used to control the populations of sea lamprey in
selected tributary rivers of the Great Lakes (Crowe 1978). The liquid chemical is
pumped directly into a stream at concentrations pre-determined to be lethal to sea
lamprey. The minimum lethal concentration (MLC) is determined by the pH and
alkalinity of the river, and is maintained for 9-12 hours continuously at each application
Site to ensure an effective treatment. Each block of chemical moves down the river based
on the velocity of the river. In some cases, it can take 3 to 4 days to treat a river and all

of its tributaries. AS the block of chemical moves downstream, the concentration of the

TFM decreases because of the additional ground and surface water inputs from
tributaries joining the main river. To maintain a lethal concentration, additional chemical
is added at “boost” Sites, in a leap frog manner along the river. Prior to treating a river
for sea lamprey, the water chemistry of the river is determined by measuring several
variables including; pH, dissolved oxygen, alkalinity, and temperature. These variables
are sampled at numerous Sites approximately every four hours for four to ﬁve days prior
to treatment. Of all the variables measured, pH by far has the most signiﬁcant impact on
the toxicity of TF M (Dawson 1975).

In the early stages of sea lamprey control, a treatment was not considered
successful unless teleost ﬁsh and sea lamprey were killed. However, for the past 25
years, non-target kills have become a cause for greater concern. For Example, ﬁsh kills
that occur while working in Michigan must be reported to the Michigan Department of
Environment Quality and the Michigan Department of Natural Resources. Under certain
conditions, ﬁsh kills can occur, even when the levels applied are thought to be within
operating range. The suspected cause of unexpected toxicity is thought to be due to pH
suppression due to the addition of TFM (Boogaard et al. 2005). Past ﬁsh kills have been
most prevalent in productive rivers, particularly rivers with high abundance of aquatic
macrophytes. It is believed that the addition of lampricide into a river system affects the
natural processes of photosynthesis (Dawson 1975) and respiration of macrophytes (Maki
1975), thereby altering pH.

In 1975, Dawson found that the toxicity of TFM in a river system is inversely
related with the pH of a river system (Figure 1). As pH decreases, the toxicity of TFM

increases (Dawson 1975). The toxicity of TFM is 5 times greater at a pH of 7.00 than at

18 ' Relationship of pH and TFM Toxicity

16'

—Sea Lamprey

d
‘
n

- Brown Trout

.I
N
A

T Treatment Concentration j ,

TFM (ppm)
3
\

 

 

 

o T T j T t V

7 7.25 7.5 7.75 8 8.25 8.5
pH Units

Figure l. Realtionship of pH and TF M (USFWS SLC Operating Manual). Sea lamprey concentration is
the minimum lethal concentration at which lamprey will die. Brown trout concentration is a concentration
at which TF M becomes toxic to sensitive species such as brown trout. Treatment concentration is typically
run at 75-85 % of the “brown trout” concentration.

a pH of 8.00 (McDonald 2006). In 1993, a prediction chart was created by the United
States Fish and Wildlife Service (USFWS) to determine the MLC for a treatment based
on the pH and alkalinity of a river (Scholeﬁeld et al. 1999). In highly productive rivers,
USFWS employees that TF M leads to an apparent suppression of the pH. Bill et. a1.
(1988) found that a reduction in pH by as little as 0.5 could cause TF M to become toxic
to ﬁsh during a lampricide treatment when this level would normally not produce non-
target mortality. Prior instances of suspected pH suppression have generally occurred in
limited stream reaches (i.e., less than ﬁve miles) and suppression occurs rapidly (i.e.,
within hours). As the pH of the river drops, a reduction in the amount of TF M being

applied is necessary. If the TFM concentration is not reduced, the level of chemical will

exceed the maximum lethal concentrations for that particular river. If TF M
concentration remains above the maximum level for more than a few hours, a ﬁsh kill
may result.

From 1980-2006, Canada and the United States conducted approximately 1502
TF M treatments during this time period (Table 1). Of those treatments, there were a total
of 88 reportable ﬁsh kills, representing only 6% of the treatments. A reportable ﬁsh kill
is any treatment in which 50 individual ﬁsh of any one species are killed. Each ﬁsh kill
is classiﬁed to be a kill of sensitive species or non-sensitive species and the cause of the
kill is also recorded (e.g., pH suppression or spawning). A sensitive species is a species
that is known to be affected by TF M applications. Fish kills of sensitive Species
accounted for only 5 %. A non-sensitive species is one that does not generally Show any
adverse affects from TF M addition. Fish kills of non-sensitive Species were only 1% of
the reportable ﬁsh kills. In many cases, sensitive species are native lamprey that are
unavoidably killed as a result of the TFM application. This is a very low percentage
considering the number of treatments conducted. Although it is a low percentage any

ﬁsh kill is considered problematic and Should be avoided at all cost.

 

 

USA Canada
Number of treatments 972 530
Number of reportable kills 28 60
Numberwrth a kill of 23 52
sensrtlve spec1es
Number with a kill of non- 5 8

sensitive species

 

Table 1. The approximate number of TF M treatments (1980-2006) and ﬁsh kills that have occurred in the
United States (1994-2006) and in Canada (1991-2006). A reportable kill is any treatment where more than
50 of any species have been killed by TF M application. A sensitive species is a species that is known to be
sensitive to TFM applications (e.g. other native lamprey, log perch, and mud puppies). A non-sensitive
species is a species that is not affected by normal TFM levels (e.g., Chinook salmon and walleye).

An additional concern is that TF M treatments that occur on rivers with a large
diurnal shift in pH and dissolved oxygen may result in ﬁsh kills. Thus one possible
mechanism for pH suppression is if TF M inhibits the production of oxygen or decreases
the level of pH or dissolved oxygen below their normal levels (Scholeﬁeld et a1. 1999).

Photosynthesis is the process by which plants take in CO; and water, and use
sunlight to produce 02 and glucose. When organisms respire Oz and glucose are used,
and C02, water, and energy are released. Photosynthesis and respiration rates affect the
concentration of CO; which then affects the pH balance. Photosynthesis uses dissolved
carbon dioxide, temporarily reducing the concentration of carbonic acid (H2CO3), leading
to higher pH. As C02 is released and dissolved into the river during respiration, the pH is
lowered. Thus, if TF M directly or indirectly inﬂuences the balance between respiration
and photosynthesis, it has the potential to shift river pH. One more complexity is that the
rate at which photosynthesis is the greatest is in the middle of the day, before noon, and
decreases after noon (Meyer 1939). In 1970, Hynes found that as macrophytes
photosynthesize during the day there is an increase in river pH. In the evening, when
respiration predominates, the pH of a river decreases. This change in macrophyte
processes causes the pH of a river to oscillate throughout a 24 hour period. Maki (1975)
and Maki and Johnson, (1976) found that TFM not only affects 02 production but also
plant growth and metabolism which may lower stream pH and dissolved oxygen.

Since lampricide treatments occur during the day and night, light availability may
also play a role in suppressing pH levels. It is unknown whether TF M affects
macrophyte physiology, or if the color of TFM (yellowish when mixed with water)

attenuates the wavelengths of sunlight needed to perform photosynthesis.

Photosynthetic activity varies continuously in river systems due to the natural variation
in light intensity (Meyer 1939). Light attenuation is also inﬂuenced by the amount of
dissolved organic matter and inorganic matter naturally in the river (Krause-Jensen
1998). This background attenuation will inﬂuence the fraction of light that is available to
the plants and in some cases can act as an inhibitor (Krause-Jensen 1998). Macrophytes
tend to photosynthesis best at wavelengths of solar radiation between 400 — 700 nm
(Hauer 1996). If TFM is inhibiting photosynthesis by blocking these wavelengths, then
the rate of photosynthesis may be reduced.

Being able to predict the cause(s) of pH suppression will help reduce the threat of
ﬁsh kills. The particular mechanism responsible (e. g., inhibition of photosynthesis or an
increase in respiration rates) is not of primary concern, but being able to determine the
conditions where pH suppression is likely is the ﬁrst goal of this research. If we can
understand the situations where pH suppression is likely, we may then be able to create a
prediction table or some other tool that will allow managers to adjust treatment levels to
avoid major ﬁsh kills. The goal of this project is to better understand the causes of pH
suppression in order to predict its occurrence and minimize its effects on non-target
species.

There are three main objectives in this study. The ﬁrst objective is to determine
the degree of pH suppression during applications of TFM. This will be done by
collecting water chemistry data and looking at the statistics before, during, and aﬁer TF M
application. The second objective is to determine if macrophyte abundance is related to
pH suppression during lampricide treatments. Plant samples will be collected in order to

estimate macrophyte abundance within each river. Not only will plant samples be

collected, but a lab experiment will also be conducted. This will allow me to determine
in, a controlled environment, if the addition of TFM affects normal plant functions. The
ﬁnal objective is to document the occurrence of any ﬁsh kills that occurred on the study

streams .

METHODS
River Selection

Rivers are typically treated for sea lamprey every three to ﬁve years, depending
on stream productivity. For this study, I selected all three rivers that had a history of
suspected pH suppression that were treated in Michigan and northern Indiana during
2005 and 2006. In addition, I sampled Six rivers that did not have a prior history of pH
suppression. Study rivers were chosen based on two criteria: the treatment schedule and
past history of pH suppression. Treatment schedules are released in February. Once the
list was released, I coordinated with treatment supervisors to determine which rivers ﬁt
my criteria of having a prior suspected pH suppression event and being treated during the
ﬁeld season. A total of nine rivers were sampled for this study (Figure 2). Not all of the
rivers sampled had a history of pH suppression. The Chippewa River, Bear Creek, and
the Muskegon Rivers have all had potential pH suppression events. The remainders of
the rivers chosen were considered control rivers. I felt that the natural river approach
would allow for a better view into what is actually occurring during a suppression event.
A lab experiment was also conducted but will not be focused on. V

Sampling occurred over two ﬁeld seasons (2005-2006). All rivers sampled were
in Michigan except for Trail Creek which is in northeast Indiana near the Michigan

border (Figure 2). The length of treated stretches of river ranged from 11.5 miles to 165

 

. A

 

f / Ii
2’/
I ,/ ‘

[fete Marquette R l . _

 

W
t‘ l
. f- 1‘ Chippewa R
/ .
Muskegon R

.'/'

I ‘ l

 

Figure 2. Map of Michigan showing location of study rivers.

miles. Average pH levels ranged from 7.46-8.27 and dissolved oxygen from 5.92-9.83

(Table 2). The alkalinity was relatively high throughout all rivers sampled with an

 

 

average of 148.

River Date Sampled pH DO Temp Alk CF S TDS Miles

Trail Creek 7/26/06-8/03/06 7.46 5.92 21.6 196 24.3 0.3 16.5
Sturgeon River 7/12/06-7/20/06 7.76 7.80 22.7 103 42.1 0.1 68.0

Bear Creek 8/22/06-8/30/06 8.05 8.18 15.7 139 80.0 0.2 47.0

Pm xgzjuette 8/08/06-8/ 17/06 8.19 i 8.36 18.9 136 320.7 0.2 165.0
Betsie River 9/04/06-9/11/06 8.22 9.28 15.6 171 136.6 0.2 1 1.5
Muskegon River 8/25/05-8/30/05 8.24 8.47 21.6 148 1233.8 0.2 88.0
Riﬂe River 9/19/06-9/28/06 8.24 9.83 13.21 191 123.7 0.3 110.0

White River 8/09/05-8/17/05 8.24 8.41 20.1 162 154.5 0.3 70.0

ChippewaRiver 7/13/05-7/21/05 8.27 7.95 25.7 185 132.2 0.3 73.0

 

Table 2. Summary of study river attributes. Mean pH, dissolved oxygen (DO, mg/l),
temperature (C°), alkalinity (total alkalinity), discharge (cfs), total dissolved solids (TDS,
g/l) and miles of river treated with TFM. Rivers are arranged by ascending pH levels.

The alkalinity ranged from 103 to 196 and ﬂuctuated little on a daily basis within
any river. The total dissolved solids average was 0.23 g/l and did not vary substantially
among rivers. The general size of each river varied widely with discharges ranging from
24 to 1234 CF S with an average of 485 CFS. Land use varied by location consisting of

rural, urban, industrial, and agricultural areas.

Data Logger Information
In most cases, four data loggers were placed throughout the river to be treated.
One logger, was placed above the TFM application point. This logger never came into

contact with TFM during the treatment, and was used as the control logger on each river.

The second logger, treatment 1, was placed downstream of the control. The remaining
loggers, treatment 2 and 3, were placed sequentially down the remainder of the river.
Locations for logger placement were determined based on the advice of chemical

control supervisors and access. Many of the rivers studied had little to no access other
than road crossings. Of the nine rivers sampled, four rivers (Muskegon River, Riﬂe
River, Betsie River, and the White River) had dams of various sizes. Treatment
application points were located above the dam on the Muskegon, Riﬂe, and the Betsie
River. To maintain a true control logger at dam locations, the logger was placed above
the application point which was typically in the pond area formed by the dam.

Hydrolab Datasonde data loggers were used to collect continuous water chemistry
data. Each data logger measured pH, dissolved oxygen, temperature, total dissolved
solids, and conductivity every ﬁfteen minutes. The data loggers were calibrated before
each use and calibration was checked every two days while loggers were deployed. If the
logger was not reading accurately, the unit was recalibrated. I found that it was necessary
to clean the loggers every two days with a tooth brush to remove buildup on the sensors
to maintain accuracy. Once the loggers were retrieved, each logger was cleaned with a
mild detergent. Data were collected prior to each treatment, during the treatment, and
post treatment. This allowed me to have reference data at every site as well as a logger
that was placed above the treatment.

Each data logger was calibrated at the beginning of each sampling period. Each
meter took approximately 45 minutes to calibrate. If recalibration was necessary, I found
it was generally only the pH probe that required recalibrating. The ﬁrst calibration was

of the pH probe. Using pH 7 and 10 standards, each storage container (a plastic screw

10

cap cup which holds storage solutions to protect the probes) was ﬁlled with a standard
and allowed to Sit for at least 10 minutes. After 10 minutes if the meter read within an
acceptable range (depended on the temperature of the solution and the correction factor
listed by the manufacturer) the probe was calibrated. Between each solution, the probe
and storage cup were rinsed with de-ionized water. The next step was to calibrate the
conductivity meter. The storage cup was ﬁlled with conductivity standard 1432, and
allowed to Sit for 10 minutes before the probe was calibrated. There is no calibration for
the total dissolved solids probe.

The ﬁnal step was to calibrate the luminescent dissolved oxygen (LDO) meter.
This was the most difﬁcult meter to calibrate. To calibrate the probe it was necessary to
completely dry the LDO probe and the remaining probes. It was then necessary to
stabilize the meter so that the probes were facing down. I used a titration stand that was
modiﬁed to hold the probe. Once the meter was stabilized, the storage cup was ﬁlled
with tap water to just below the probe, but not touching the probe. The probe was
allowed to sit until the meter had reached as close to 100% saturation as possible. This
was then checked against the Hydras3LT program provided with the data logger. Once
the LDO meter was standardized, the last step was to program the meter for data
collection.

Several problems occurred with instrumentation while in the ﬁeld. On the
Chippewa River, the control and the treatment logger #3 both failed. Even though the
control logger failed, there were data collected by each treatment logger prior to the TFM
treatment thus providing control data for evaluating the effects of TFM. Due to major

ﬂooding on Trail Creek, the control and treatment #3 loggers failed to collect data. Both

11

loggers were submerged in sediment for four days prior to being found and retrieved.
The Riﬂe River also had one logger malfunction, treatment logger #3. Bear Creek also
ﬂooded, and the control and treatment #2 loggers were both buried in sediment. These
loggers were not retrievable for four days. Due to the ﬂooding, the treatment of Bear
Creek with TFM was cancelled. Data were collected, but were not used to determine the

impact of TFM on stream water chemistry.

Hand Sampled Water Chemistries

Although the Hydrolab Dataloggers measured pH, dissolved oxygen, and
temperature, I coordinated with the chemical control crew from the USFWS Ludington
Biological Station to collect hand sampled data. These data were used to check the
accuracy of the data loggers. River pH, dissolved oxygen, temperature, and alkalinity
were measured every four hours on average, and hourly during treatments.

All hand samples were measured using USFWS Instrument Operating Procedures
(IOP). Temperature and pH readings were taken using a handheld Beckman pHI 240
meter (IOP: 007.3A & IOP: 007.1B). Dissolved oxygen was measured in mg/L using an
YSI model 55 dissolved oxygen meter (IOP: 008.1). Alkalinity was determined using the
sulfuric acid titration method as described in the USFWS Technical Operating Procedure

(TOP) (TOP: 005.1).

Non-Target Surveys
A non-target Species is any organism that is not a sea lamprey as they are the

target of TFM. Aﬁer TFM treatments occur, staff from the US. Fish and Wildlife

12

Service search along the river to collect, identify, and determine the cause of death of
any organism found dead along the river. This is typically done at road crossings because
of access limitations on some rivers. Two crew members are sent to each site, one
searching upstream 100 yards and the other downstream for 100 yards. In the case of

larger rivers, canoes are used.

Plant Biomass

Plant density and abundance were estimated using a quantitative sampling
approach. Prior to each collection, ten sites were randomly selected from sites that were
easily accessible or had road crossings. Once the ten sites were selected, each site was
then brought into ArcGIS and a random transect was generated. Each transect was
within 100 yards upstream or downstream of the access point. Transect locations were
uploaded into a GPS unit for accurate identiﬁcation. At each transect, the river width
was measured, and the total width was divided by tento provide sample locations for
each of the 10 plots along the transect. Each plot was surveyed using a 1m2 aluminum
square. The depth of each plot was measured (in meters) and dominant substrate was
visually determined. If the depth was over 0.5 meters, plant material was not collected,
but the species and density in percent were visually estimated. All plant material was
collected within each plot that was under 0.5 m in depth. The plant material was placed
' in a plastic Ziploc bag, and labeled as to the location within each transect. The plant
material was identiﬁed, or if I could not identify the plant, pictures and voucher samples
were taken. Once identiﬁed, the plant material was patted dry and a wet weight was

measured in grams. All plant material was then dried for 6 hours in a food dehydrator

13

until completely dry. Once the plant material was dry, a dry weight was measured, and

the sample was discarded.

Laboratory Plant Experiment

A laboratory experiment was set up to determine the effects of lampricide on
plant photosynthesis in a controlled environment. The test was run in a bioassay trailer
owned by the USFWS. The trailer is set up to run serial dilution toxicity tests, but was
modiﬁed for our purposes. The bioassay trailer is set up as a ﬂow through system and is
set up stream Side. I chose to use the Pere Marquette River, Mason County, MI, as the
site of my experiment. This Site was close to the USFWS station and was the easiest to
access. A total of 4 tanks were used: 3 treatment tanks and 1 control.

Plants were collected from the Lincoln River, Mason County, MI, because of the
low abundance of plant material in the Pere Marquette River. Plant specimens were
collected using a shovel. A 6 inch x 6 inch square, 1 inch deep section was harvested and
placed in a cooler with water for transport. Plants were then placed in each tank with a
cover of approximately 80%. A grid was drawn at the bottom of each tank and 80 % of
the squares were covered with plant material. The plants were left in the tanks for 12
hours in order to allow them to acclimate to the river water. After the 12 hour period,
each tank was equipped with a data logger.

All four data loggers were calibrated and programmed to take pH, dissolved
oxygen, temperature, conductivity, and total dissolved solids readings every 15 minutes.
Plant growth lights were used to simulate natural light. Using electric timers, the lights

were turned on at 0630 and shut off at 1930.

14

The day prior to TF M application, I collected stream chemistry data from 0900

to 1630 and was able to determine a natural shift in pH. The pH ranged from 8.10 to 8.22
with an alkalinity of 150. Based on this information, the calculated minimum lethal
concentration (MLC) of TFM was 3.5 ppm. I used a rate of 4.0 ppm which would be
typical during a treatment as we usually treat above the MLC. Each logger was set to run
12 hours prior to the addition of TFM. TF M application began at 0900 and was shut off
at 2100 that night. Each treatment tank measured at or near 4.0 ppm the entire test. Once
the chemical was stopped, the concentration of TFM in the treatment tanks began to
decline immediately and was gone within an hour.

The loggers continued to sample for 12 hours after the ﬁrst study was conducted.
After the 12 hours, 60% of the original 80% of plant material was removed leaving
approximately 20%. This allowed me to look at the effects of high and low densities.
The tanks were again allowed to acclimate for 12 hours. The second application of TFM

was executed identically to the ﬁrst.

Light Attenuation

Light measurement readings were taken using a Li-Cor 189 quantum radiometer
photometer with a Li-Cor Ll-l92SA underwater quantum sensor at a set of selected Sites
on tributaries to Bear Creek. The sensor was set to measure light intensity (umol/mz/sec)
of photosynthetically active radiation, between the wavelengths of 400-700nm. Five
control transects and ﬁve treatment transects were measured. The control transects were .
clear water 3 meters upstream of each treatment site application point. Each treatment

transect was 5 meters downstream of the TFM application point. This allowed for the

15

TF M to be completely mixed within the water column. Within each transect, ten plots
were established using the same method as was used for the plant biomass plots. Each
light reading was taken 0.5 meters from the water surface. The average depth of each

plot was 0.58m.

Analysis

Using SAS, logger information was sorted and summarized. A general linear
model (GLM) analysis was used to evaluate differences between sites and die] variation
within sites. I also looked at the averages of the three main variables (pH, dissolved
oxygen, and temperature) before, during, and after a chemical treatment. The mean diel
amplitude of pH, dissolved oxygen, and temperature were also calculated at each logger

Site. Averages were calculated for the entire sampling period by river and Site.

RESULTS
General Trends

Several water chemistry parameters (alkalinity, total dissolved solids (TDS), and
conductivity) varied little within each river, but pH, dissolved oxygen (DO), and
temperature varied substantially among Sites within individual rivers, and on a diel basis.
Because alkalinity, TDS, and conductivity varied little within a river, these parameters
are Simply reported as a mean across all sampling (Table 2). Although pH, DO and '
temperature varied among sites, most of the rivers studied Showed parallel behavior
among sites with strong diel patterns evident in these parameters. In most rivers, there
was substantial site-to-Site variation in pH, DO, and temperature. Site-to-site variation in

pH was particularly evident in the White River, Sturgeon River, and Pere Marquette

l6

River (Table 3). Site-to-site variation in dissolved oxygen was highest in the Chippewa
River, and was also very high in the Pere Marquette River. Site-to-site variation in
temperature was generally less than pH or DO, but was highest in the Sturgeon River.

Within day variation in pH, DO, and temperature was substantial on several rivers
(Table 3). The average diel amplitude was 0.35 pH units, 2.48 mg/l DO, and 319° C.
The Chippewa River showed the largest diel ﬂuctuations in both pH and DO, with pH
varying up to 0.76 units per day, and DO varying up to 10.56 mg/l per day. The
Muskegon River also showed higher than average diel ﬂuctuations in pH and DO, with
mean amplitudes ranging up to 0.55 pH units, and 4.21 mg/l per day.

Graphs of pH and DO over the entire study period for each river, provided in
individual river summaries, indicate that the presence of TF M did not result in noticeable
reductions in pH or DO. Table 4 shows the summary of results from the GLM analysis
on the differences between pre-, during, and after TFM data collected. The averages over
all river differences varied very little during the study periods. Temperature had the
biggest difference from a TF M treatment with an average difference of 0.48 degree
difference after TF M treatment occurred. As was stated before, temperature has no direct
effect on the toxicity of TFM but does affect the ability of water to hold oxygen. The
GLM analysis showed statistically signiﬁcant differences in pH before, during, and after
treatment, but the magnitude of these differences were not of biological signiﬁcance.
Relative departure of pH from the grand mean was -0.02 units before treatment, during
was 0.04 units and after having a average difference of -0.05 units. Dissolved oxygen

was consistent during the TFM treatment.

17

 

pH DO Temp

 

 

River Logger pH Amp DO Amp Temp Amp
Betsie River Control 8.22 0.43 9.36 2.97 15.25 2.91
Treatment 1 8.18 0.37 9.48 1.31 15.48 2.98
Treatment 2 8.23 0.30 9.46 1.42 15.64 2.90
Treatment 3 8.25 0.24 8.82 2.27 16.18 2.64
Chippewa River Control Failed
Treatment 1 8.17 0.25 7.14 2.94 25.21 2.90
Treatment 2 8.31 0.56 7.43 4.38 25.71 3.46
Treatment 3 8.33 0.76 9.27 10.56 26.30 5.07
Muskegon River Control 8.39 0.48 8.91 3.46 21.86 2.1 1
Treatment 1 8.10 0.55 8.29 4.21 21.26 3.94
Treatment 2 8.17 0.47 8.23 2.64 21.35 2.93
Treatment 3 8.30 0.17 8.44 1.32 21.82 2.54
White River Control 8.41 0.39 8.69 1.56 20.54 3.50
Treatment 1 8.26 0.39 8.48 2.32 20.85 2.73
Treatment 2 8.12 0.18 8.17 1.10 19.57 2.29
Treatment 3 8.19 0.22 8.34 1.53 19.59 2.46
Sturgeon River Control 7.53 0.34 7.43 2.51 20.75 7.44
Treatment 1 7.75 0.40 7.73 1.48 23.58 6.62
Treatment 2 7.83 0.39 8.24 1.64 23.14 3.32
Treatment 3 7.92 0.33 7.81 2.64 23.17 3.51
Pere Marquette River Control 7.93 0.20 7.87 1.47 19.29 2.51
Treatment 1 8.27 0.22 8.16 1.69 18.55 2.27
Treatment 2 8.21 0.46 9.03 2.89 18.78 3.02
Treatment 3 8.37 0.19 8.47 1.39 19.17 2.36
Riﬂe River Control 8.31 0.20 10.03 0.97 14.50 1.88
Treatment 1 8.10 0.29 9.91 1.63 12.46 1.94
Treatment 2 8.35 0.33 9.56 2.11 12.73 2.62
Treatment 3 Failed
Overall Average 8.16 0.35 8.57 2.48 19.72 3.19

 

Table 3. Mean and die] amplitude of pH, dissolved oxygen, and temperature at each site
sampled. Averages are computed over entire sample period for each river. Diel
amplitude is computed as the mean difference in the daily minimum and maximum.

18

.05 Some a 8:898 2E. $0 80m 28 A9 @030 Am: 80000 338 98:3 032 308088 .50sz 02,330
was .30 .BBSoQEB 5 80:80.06 0o max—ace @500 3002 30:5 0:250 :80 358% 3:52 00 $8830 .v «Baa.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

30.0 000.0 00.0 000 N00 5000 30.0 00.0.. V00 N00- 0500 0N0.0 0T0 N0- E0- mow80><
35.50 oz «BREED

028m
Howwoq mm 20E
00.0 000.0v _N.0 N_.0 50.0 0000 ~000v 2.0- 000- 00.0- 000.0 ~000v 0v; NA 50; Nm 20E
000.0v 00.0 0Nm0 ~000V 000 N40 30 000.0 _00.0v 00.0 N0.N 00.N 5 05M
30.0 000.0 00.0- 000- N0.0- 500.0 ~000v v0.0- 0V0- 500- 00.0 ~000v 00.0 $0 000 mm .2 .m
N000 ~000v 0N4- 050- 50.7 «:0 _00.0V 30- N0. 0N0- 0N00 _00.0v 00.0 00.0 00 Nm .5 .0
000.0 _00.0v 000- M20- mN0- 0N00 _00.0v 00.0- 0N0- 0N0- V000 V000 50 00.0 N50 10 .20
0_ 0.0 ~000v 00.0- 00.0 0V0- N000 _00.0v 00.0- 0N0- 0 m0- 500 _00.0v woN 00.0- me- mm .waBm
00N.0 000.0v 0:- 000 $0- 0500 _00.0v 0 m0- :.0- 0N0- va0 000.0v 000 N50.- SN- Nm .wuam
NN00 ~000V N00- 000- 00.0- 000.0 ~000v 90- 000- NN.0- 000.0 _00.0v 0v.N- V5.0- 00.N- a .330
20.0 ~000v 50.0- 000 00.0 050.0 ~000v 50.0- 00.0 0N0 0000 000.0V 00.N cm; 00.0 00 BE?
w000 ~000v 0V0 05.0 _0.0 500.0 _00.0v 00.0 $0 0N0 000.0 _00.0v 00; N0; :0 N0 BE?
0000 ~000v 00.0 000 00 .0 0000 _00.0v 5N0 NNO 50.0 500.0 ~000v 0N0- 0V0 @0- 5 8E?
0000 ~000V 00.0 00.0- 00.0 00 _ .0 500v N0 000 50.0 00 0 .0 _00.0v N00 000- 00.0.. mm 132
00.0 _00.0v 00.0 0V0 00.0 00 H0 000.0v Nv0 0N0 w _ .0 000.0 5wN0 v0.0 20 00 .0 N0 .0752
000.0 50.0 NA 5N0 00.0 N500 _00.0v Nv0 N00 0N0 000.0 _00.0v 00.0 00.0 00.0 a .0332
0VNO _00.0v 00.0 0: 0.0 500.0 _00.0v 00.0- :0 000- 02.0 000.0v 05.0- 000- 2.7 mm 330m
N000 _00.0v 2.0- NNO 50.0- 900.0 _00.0v 00.0 00.0 00.0- 05.0 000.0 3.0- NN.0- 0V0- Nm 330m
050 N000 M20- 500 000- 0000 _00.0v _0.0 :0 50.0 5N00 000.0v NNO- 000 N0- :0 230m
hm m < D m mm m < D m Am A < O m 85 52m

cowzxo 002085 $0 83800th

 

 

 

 

19

Before the treatment the average difference was 0.02 mg/l, during was 0.03 mg/l and
after the TFM block an average of 0.05 mg/l difference. Statistical comparison of the
difference between pre, during, and post TFM time periods suggest the reduction in pH
was no greater than 0.10 units, and the reduction in DO was no greater than 0.02 mg/l.
Given the magnitude of die] ﬂuctuations and site-to-site variation, 1 view this as a minor
impact on pH and DO.

Among the rivers studied, a ﬁsh kill was observed only on the Betsie River. As is
evident if Table 4, the pattern in pH and DO within the Betsie River do not appear to
have been inﬂuenced by the addition of TFM. More details concerning the ﬁsh kill and

water chemistry on the Betsie River are provided in the river summary.

Analysis

The differences between the pre-, during, and post-TF M treatment readings
Showed that there were signiﬁcant changes in pH, dissolved oxygen and temperature
readings. Although the results were signiﬁcant, the high number of samples that were
analyzed could account for the detection of minor differences. Table 5 shows the number
of miles of river treated with T FM and the number of samples collected with the data

loggers.

20

 

River Miles of River Treated Sample Size

 

Betsie 1 1.5 684
Sturgeon 68 737
White 70 742
Chippewa 73 750
Muskegon 88 476
Riﬂe 1 10 8 l 3

Pere Marquette 165 821

 

Table 5. Average river miles treated with TFM within the last 25 years. The
sample size is the number of water chemistry samples taken by the data loggers
deployed on each river.

Individual River Summaries
Betsie River

The Betsie River was the only river studied in which a ﬁsh kill occurred during
TF M treatment. The TFM treatment on the Betsie River was set up at Homestead Dam.
The ﬁsh kill occurred below Homestead Dam and continued to Grace Road which is
approximately 2 miles downstream. There were 438 salmon were collected from the
river and taken to the local trash dump. The treatment had begun at midnight and the ﬁsh
kill was not noted until the daylight hours. Once the ﬁsh kill was determined, the
remainder of the treatments were shutdown and the treatment was stopped.

Strong daily shifts in pH were seen during the sampling period (Figure 3). Trends
in pH were parallel among all Sites and within day, except treatment logger 3 had a
unique double daily peak in pH. Logger 3 was located approximately 12 river miles from
the control logger at Homestead Dam. Logger 3 had lower daily amplitude shifts in pH
(0.24 units), than the control logger, 0.43 pH units, but did not go higher than the control
logger pH. A natural drop in pH was detected on day ﬁve, lasting two days.

Dissolved oxygen was also nearly parallel, especially between treatment loggers 1

and 2. Dissolved oxygen was higher than average at the dam and lower than average at

21

8.6 a

8.5 -'

8.4 .-

8.3 '1

pH

8.2 '-

8.1 -'

    

 

 

7.9 U V "I' V U U V

12-

11.5-1

11‘

10.5 -

10-

9.5 ‘

 

 

9d

8.5 .

 

7.5 -' ._.._.._

 

 

201

— Control

— Treatment 1
— Treatment 2
_... Treatment 3

19-
18'1

17.-

\ ’1
16- l

. l 1 '3 ‘1 j ‘l’
15- 3, -' I ' \
‘-l \ ' \\
14 - -' \
'l
.\ /'
I \

8 7 8 9

   

 

Temp

   

13 -I ”2:...

 

 

12 v r
1 2 3 4

V V

5
Day of Study

Figure 3. Betsie River graphs of water pH, dissolved oxygen (mg/1) and temperature (C). Bars indicate
when TFM chemical block passed through the treatment logger locations.

22

I

treatment logger 3. A drop in dissolved oxygen levels was also noted on the ﬁfth and
sixth days coinciding with the drop in pH. The average oxygen concentration ranged
from 8.82 to 9.48 mg/l across the four sites sampled. Temperature trends were nearly
parallel across all sites. A large drop in temperature occurred on day six and continued to
decline for the remainder of the sampling period.
Although a ﬁsh kill occurred, the addition of TF M did not cause a noticeable
suppression of pH (Figure 3). The pH levels on the fourth day had reached a maximum

pH of 8.54. After the natural decline on the ﬁfth day, pH never went above 8.44 units.

.; A- -_. “.52.”

Sturgeon River L .
The Sturgeon River showed diel patterns representative of that observed in other

rivers (Figure 4). Die] patterns in pH were largely parallel, but there was a high degree of

variation among sites in the daily amplitudes of pH. The control data logger had a lower

overall pH during the sampling period, averaging 7.53 units. Trends in dissolved oxygen

were also nearly parallel among sites. The daily amplitude of DO variation was small

when compared to other rivers sampled, with an average amplitude shift of 2.07 mg/l.

Temperature was also parallel among the three treatment data loggers. The control

logger had a lower temperature overall, but a larger daily amplitude shift than the other

 

three loggers. The treatment loggers had a very large diel ﬂuctuation in temperature,

averaging 3.32 to 6.62° C per day.

Pere Marquette River
Water pH differed substantially between sites, with averages ranging between

7.93 and 8.37 units (Figure 5). The pH at the control site was lower than the remaining

23

8.2 - . ,.
8.1 - [A

8 u

7.9 -
7.8 -

I 7.7 -
7.6 -
7.5 ~
7.4 -

 

7.3 - —

 

 

 

7.2 i I t I I V

10-

 

 

 

34 u
33 ‘ —Control

32 : —Treatment 1
31 — Treatment 2
30 - “I“ Treatment 3

29-
28‘
27-
26- K
25:

24- K /\

23-l

I f . - \
22 d “‘5‘ \ l‘ a.
21 - K, V
20 - '

19 '
18 '
17 ‘
16 '
15 '
1‘ v I I I r v v 1

 

Temp

 

 

 

Day of Study

Figure 4. Sturgeon River graphs of water pH, dissolved oxygen (mg/1) and temperature (C). Bars indicate
when TF M chemical block passed through the treatment logger locations.

24

 

 

 

 

 

 

 

— Control

— Treatment 1
— Treatment 2
“- Treatment 3

 

 

Day of Study

Figure 5. Pete Marquette River graphs of water pH, dissolved oxygen (mg/l) and temperature (C). Bars

indicate when TFM chemical block passed through the treatment logger locations

25

loggers during the entire sampling period, but showed similar diel variation to the other
sites. The daily amplitude in D0 was similar among sites, only varying by 1.16 mg/l.
Treatment logger 2 had the highest within site variation ranging from 7.99-8.62 pH units.
Dissolved oxygen levels were comparable across sites. Treatment logger 2 had the
highest variation, with a daily amplitude shift averaging 2.89 mg/l. The remaining three
loggers had very little variation and were parallel. Temperature was parallel between

sites, and a drop in temperature was noted on days 5 and 6 across all sites.

Muskegon River

Water chemistry patterns in the Muskegon River were strongly affected by Croton
Dam, located at the upstream boundary of the TFM treatment (Figure 6). The control
logger was placed above Croton Dam as the treatment occurred at the dam. The water
that was being sampled by the logger was lake water from Croton Pond.

Treatment logger 1 and 2 both had relatively large amplitude diel shifts in pH of
0.55 and 0.47 pH units. Trends over time were similar between these two loggers, but
were less parallel than in the other rivers discussed so far. Logger 3 had very little
change in pH levels throughout the day and night with an average diel shift of only 0.17
pH units. Treatment logger 3 was in an area that was on average deeper than the other
sites and highly turbid.

Loggers l and 2 also showed parallel dissolved oxygen trends. Once again, low
amplitude diel shifts were recorded at treatment logger 3 with an average shiﬁ of 1.32
mg/l while logger 1 was 4.21 mg/l. Temperature readings also showed strong diel

patterns among all sites. The three treatment loggers were relatively parallel although

26

8.7
8.6
8.5
8.4

8.2 d

pH

8.1 ~
8 d
7.9 -

7.8 -‘

 

 

7.7

12

11

10

DO

 

 

 

25 - —00ntrol

— Treatment 1
— Treatment 2
24 . “-7 Treatment 3

\ /r~'\

18 1 . v u
3 4 5 6 7 8

Day of Study

Figure 6. Muskegon River graphs of water pH, dissolved oxygen (mg/l) and temperature (C). Bars
indicate when TF M chemical block passed through the treatment logger locations.

  

Temp

   

 

 

27

logger 3 had a lower amplitude shift of 2.54 degrees than the other two loggers at 3.94

and 2.93 degrees.

Rifle River

Treatment logger 3 on the Riﬂe River malfunctioned during sampling, and based
on trends over time, I suspect treatment logger 2 may have malfunctioned for part of the
time. (Figure 7) The control logger was above the dam on the West Branch of the Riﬂe.
There is a moderately large pond above the dam that the logger was sampling from as can
been seen in Figure 7. Water pH levels for logger l and 2 showed a diel pattern at ﬁrst
and after day six a pattern returned at logger 2 but was not as strong as before. The
average amplitude shifts for logger 1 and 2 are 0.29 and 0.33 pH units.

Dissolved oxygen levels were nearly parallel, although the control logger showed
the effects of the dam on DO levels. The diel amplitude swing in D0 was relatively
small, only varying by 1.63 mg/l at logger 1 and 2.11 mg/l at logger 2. Temperature was
moderately parallel among the two treatment sites. There was a large increase in

temperature on day ﬁve and then the temperature returns to original levels after two days.

White River

The diel pattern of pH levels in the White River was very pronounced, as is
visible in ﬁgure 8. The diel pattern is strong but there is a very low amplitude shift
ranging on average from 0.00 units to 0.06 units. As was seen on the Betsie River,
double diel pH peaks occurred at the control logger, which was located approximately

100 yards below the Hesperia Dam on the White River. Above the cement darn there is a

28

8.8 -
8.6 - ‘ . l t '.

8.4 a

   

7.8 -'

 

 

7.6 I I I f t V

12-

11-

 

 

 

 

 

17-

-— Control
1 6 - — Treatment 2

'-—- Treatment 3

r' ' 4 .1
1 3 . l . 2‘ L ’r\; i

12-- ‘. ‘ /\ r A
ix! .
V

Temp
d
a

1 1 d V x... r " ‘\

 

 

 

1o ‘7 V T
2 3 4 5 6 7 8 9 10 11

Day of Study

U V I I U '

Figure 7. Riﬂe River graphs of water pH, dissolved oxygen (mg/l) and temperature (C). Bars indicate
when TFM chemical block passed through the treatment logger locations.

29

16.25 hectare pond. Overall, pH was comparatively parallel among the treatment sites
with very little site variation with the average variation in diel amplitude being only 0.30
pH units. Oxygen levels were similar among all sites, ranging from 8.17 to 8.69 mg/l.
Temperature readings were a little more sporadic. Although there was still a noticeable
diel pattern, the parallel nature seen in the levels of pH and DO were not as visible.
Temperature at treatment logger 3 ﬂuctuated substantially ranging on average of 2.02

degrees, without an apparent pattern.

Bear Creek

Bear Creek endured heavy ﬂooding during the study period, with the main
ﬂooding event occurring on the fourth day of sampling. The control and treatment one
loggers were buried in sediment for several days before they could be retrieved. Due to
the heavy ﬂooding, this river was not treated for sea lamprey. Thus, data collected were
not statistically analyzed. Prior to ﬂooding, pH showed a diel pattern at ﬁrst but then
ﬂood events caused the pH to become erratic and there were no real patterns
reestablished until the eigth day of the study (Figure 9).

Dissolved oxygen diel patterns began the same as pH levels, but quickly were
affected by the rain. Only the treatment 4 logger was able to capture a reestablished
oxygen trend. Temperature was not affected to the extent that pH and dissolved oxygen
were interrupted. There was a noticeable diel cycle throughout the sampling period;
however, the rain affected the temperature during the ﬁfth day of the study. The control

logger temperature dropped dramatically from an average of 14.25 degrees to an

30

8.8 -
8.7 -
8.8 a
8.5 -
8.4 '

z 8.3 d
8.2 ..
8.1 -

3 .1

7.9 ‘

 

 

 

7.8

11-

10'

DO
:8

 

q
q
d
e

 

 

 

24!

23-

22‘

21-

20'

Temp

19-

18-l

17-

 

I

10

0|
‘11
Q
Q:

— Control

— Treatment 1
— Treatment 2
* Treatment 3

 

 

18

1

Day of Study

Figure 8. White River graphs of water pH, dissolved oxygen (mg/1) and temperature (C). Bars indicate
when TFM chemical block passed through the treatment logger locations.

31

 

 

 

11-

10- 9’

LC? .8 Mix}
0 7" , i

51
4c-

31
21
1
5

20-

 

 

 

 

— Control

— Treatment 1
— Treatment 2
' Treatment 3

19'

   
   
    

18‘

 

 

12 V I V I U

4 5 6 7 8 9
Day of Study

Figure 9. Bear River graphs of water pH, dissolved oxygen (mg/l) and temperature (C). Bars indicate
when TFM chemical block passed through the treatment logger locations.

32

average low of 12.69 degrees but began to return to normal temperature levels towards

the end of the sample period.

Trail Creek

Although data were collected, ﬂooding events during the sampling period caused
the loggers to malfunction and the data were not representative of typical water chemistry
levels. Treatment logger 3 and 4 were both buried in sediment for several days. Figure
10 shows the data that was collected by the loggers but this data was not analyzed
statistically. Measurement of pH and DO were very sporadic and no real pattern could be
determined. Temperature readings, however, show a strong diel pattern. Between the
control logger and the treatment 1 logger, a largely consistent pattern can be seen,
although the amplitude was rather minimal. The decrease in temperature between day 5

and 6 can be explained by the heavy rain event that occurred.

Chippewa River

The Chippewa River also had logger malfunctions (Figure 11). The control and
treatment 3 loggers malfunctioned. Logger 3 collected a couple days worth of data
before the malfunction occurred. During the ﬁrst three days, logger 3 showed large shifts
in daily pH levels with an average amplitude shift of 0.76 pH units. Treatment 1 logger
did not have a pattern that was distinguishable throughout the entire sampling period.
The average amplitude shift was 0.56 pH units (Table 3).

It was unfortunate that logger 3 failed because it showed dissolved oxygen diel
shifts over the ﬁrst two days that averaged 10.56 mg/l. Smaller amplitude shifts were

seen at treatment loggers 1 and 2 but they were still nearly parallel. The average shift in

33

7.8 '-

08-...0..- QC...

 

 

 

 

 

 

 

 

 

      

 

 

6 7 8 9 10
' .....
9 . r
i
8 q i
g
7 q “I.“ , J " l' 23 a
6 't ~ . . i I
5.. t .1 3 f5; I
a 4 . i‘ i ii i
‘1 a-
3 . it .‘ s1 3:
:1 ll ;
2 ' 1 I '._r t
:i‘ I
1 1 . . . -‘£ :
_..M..,.,. 1‘: i
o I I I I I I ll I
3 4 5 6 7 8 9 1O
25 -
—-Treatment2
"Treatment 3
24 '
a 23 ' ’2 ﬂ .« J\ . i h
E ' ‘ '/J‘\ ' l -'. i
O I
1-
22 'I
f .
I" ' /\«
21 J ’ \/
20 I I I I I II I
3 4 5 6 7 8 9 10

Day of Study

Figure 10. Trail Creek graphs of water pH, dissolved oxygen (mg/1) and temperature (C). Bars indicate
when TFM chemical block passed through the treatment logger locations.

34

7.8 -I

 

 

7.6 I I I I I I I I I

16-

14-

12-

DO

10-

 

 

 

10

30-

—- Treatment 1
29 ‘ -— Treatment 2
. --— Treatment 3

:2: .   /\t \\ ,
25 . \ / \K i . ‘

Temp

1
u

244

    

23-l

v.~o“"o>o~e \

 

 

 

22 I I I I I
0 2 4 6 8 1 0
Day of Study

Figure 11. Chippewa River graphs of water pH, dissolved oxygen (mg/l) and temperature (C). Bars
indicate when TF M chemical block passed through the treatment logger locations.

35

 

 

dissolved oxygen at logger 1 was 2.94 mg/l while logger 2 was 4.38 mg/l amplitude shift.
Temperature, however, was very parallel at loggers l and 2. The average temperature
shift at logger 1 was 2.90 and 3.46 degrees at logger 2. Logger 3 again had high diel shift

amplitude with an average of 5.07 degrees.

Plant Collections

I sampled plants on all nine rivers sampled for water chemistry (Table 6). A total
of 680 plots were surveyed. I could not sample all transects within each river due to
accessibility and/or depth constraints. The mean depth was 0.51m and ranged from 0.42
to 0.71m while the mean width was 22.6m ranging from 11.7-57.9m (Table 6). No plants
were collected on the Pere Marquette River or the Riﬂe River within the random
transects. Although no plants were within the random transects, there were plants within

the river systems. The mean plant biomass varied substantially among rivers. Bear

 

Creek had the highest mean dry weight with 2.56g/ m2 while the Pere Marquette and
Riﬂe Rivers contained no plants within transects. The dominant substrate was sand with

8 rivers being sandy and 1 river being primarily gravel (Muskegon River).

 

Mean Plant Dominant

 

River Width (n3 Depth (m) Weight (g/mz) Substrate N

Riﬂe 16.8 0.44 0.00 Sand 70
Betsie 17.2 0.46 0.55 Sand 100

Sturgeon 14.1 0.43 0.25 Sand 60

Chippewa 25.9 0.52 1.41 Sand 80

White 21.7 0.50 0.04 Sand 90

Muskegon 57.9 0.71 0.38 Gravel 80
Pere Marquette 15.8 0.59 0.00 Sand 100
Bear 1 1.7 0.42 2.56 Sand 100

 

Table 6. Average width, average depth, mean dry plant weight, dominate substrate, and
number of plots surveyed during plant biomass collection.

36

 

Laboratory Experiment

The controlled plant experiment yielded no significant evidence of pH
suppression. After running both the 80% and the 20% plant biomass tanks, no visible
sign of pH suppression was found. Throughout the experiment all tanks remained within
0.2 ppm of TF M for the entire 12 hour run. Table 7 shows the results of both
experiments. Mean water pH in the two treatment and control tank remained within 0.06 i
units throughout the entire experiment. Before and during the treatment period, pH’s
remained virtually the same, all being within 0.03 units. After the TFM application was E
shut off, the pH of the 80% plant trial was lower than the other two treatments by 0.06
units; however this was not a biologically signiﬁcant amount.

Dissolved oxygen varied throughout all experiments and in the control tank.
Average oxygen levels prior to TFM addition (8.57 mg/l in 20%) were lower than during
the treatment (8.68 mg/l in 20%). This could be accounted for because the treatment
period ran during the day when photosynthesis would be highest. After TF M was
stopped, both the 20% and the control loggers showed decreases in average dissolved
oxygen levels. The 80% treatment tank had an increase of 0. 12 mg/l in average dissolved

oxygen. Temperature readings did not seem to be affected by the addition of TFM. The

 

only noticeable difference was within the 80% logger. The average temperature dropped
from 19.12 to 17.97 during TFM application and then increased to 18.42 after TFM was
shut off.

Looking at the data it does not appear that TFM had a signiﬁcant affect on pH,

dissolved oxygen or temperature. Most of the variation can be explained by the fact that

37

photosynthesis is occurring during the TF M addition. This test was also run in a ﬂow
through system which takes water from the adjacent river, could explain the subtle

differences seen during this experiment.

 

pH
Before During After
Treatment Mean Std Dev Mean Std Dev Mean Std Dev
20% 7.97 0.06 7.95 0.07 8.04 0.06
80% 7.96 0.12 7.92 0.05 7.98 0.05
Control 7.98 0.06 7.95 0.04 8.03 0.04

 

Dissolved Oxygen
Before During After
Mean Std Dev Mean Std Dev Mean Std Dev
20% 8.57 0.43 8.68 0.20 8.50 0.21
80% 8.34 0.27 8.46 0.14 8.58 0.12
Control 8.27 0.15 8.45 0.17 8.25 1.00

 

Temperature
Before During Aﬁer
Mean Std Dev Mean Std Dev Mean Std Dev
20% 17.46 0.41 18.41 1.03 19.31 0.65
80% 19.12 0.43 17.97 0.26 18.42 0.18
Control 18.62 0.91 18.24 0.80 19.00 0.69

Table 7. Laboratory plant experiment results. Mean and standard deviation for three
replications of each experiment (20% and 80% plant abundance) as well as the control
tank values are represented. Measurements were taken before, during and after TFM was
added to each experimental tank.

Light Attenuation
Light measurements were collected on treated tributaries of Bear Creek (Table 8).
Although Bear Creek itself was not treated, due to ﬂooding, the tributaries were treated.

A total of ten sites were sampled, ﬁve control stretches and ﬁve treated stretches. The

38

 

results show that TF M does have an affect on the intensity of light that reaches 0.5m

below the waters surface.

Light intensity in the control stretches ranged from 542 to 678 umols/mZ/sec,
averaging 609.8 umols/mz/sec. Light intensity in the TFM treated stretches ranged from
123 to 189 umols/mz/sec, with a mean of 152.8 umols/mz/sec, resulting in a difference of
457 umols/mz/sec between the control stretch and the TF M treated reach. This is

evidence that the addition of TF M does have an effect on the intensity of light available

for plants to complete photosynthesis.

 

 

Control TFM Treated
Location Average Average
Big Beaver Creek 542 123
Lemon Creek 678 189
Little Bear Creek 652 176
Little Cedar Cr. Up 581 134
Little Cedar Cr. Down 596 142

 

Table 8. Average light attenuation measurements taken on Bear Creek.

DISCUSSION

Overall, the data loggers worked well collecting the data needed for this study.
The only thing that would have helped would be to have an anti-fouling device attached

to the logger. This would reduce the amount of maintenance that was necessary to

maintain the accuracy of the loggers.

As was seen in the multiple river graphs, diel patterns resulting from
photosynthesis and respiration were visible on all rivers. Within each river, there were
large diel amplitude shiﬁs, but there was also a lot of variation between sites along the

rivers. Many of the rivers sampled were rather large, the Muskegon with 88 miles treated

39

 

 

 

and the Chippewa with 73 miles and the Pere Marquette River with 165 miles treated.
The lower reaches in these rivers had turbid water and was typically warmer. This may
explain some of the site to site variation along a river.

None of the seven rivers that had a TFM treatment and that had

functioning loggers experienced pH suppression coinciding with TF M passage.
Although pH, dissolved oxygen, and temperature were variable, the effects of TFM were
not signiﬁcant. Although a ﬁsh kill occurred during my sampling period, pH suppression
does not appear to be the cause. Prior to treatment there appears to have been a natural
drop in pH levels. The treatment began around 12 midnight, started while the pH cycle
was at its lowest point. Unfortunately, it was not realized that a kill occurred until day

break when the carcasses were visible.

Mean Plant Density and Mean Diel pH Amplitude
0.55 -

0.5 -

P

a

on
I

P
1‘
1

0.35 -

Mean pH Amplitude

c
u
0

0.25 1

 

 

082 I I I I I I I I
0 0.2 0.4 0.6 0.8 1 1 .2 1 .4 1.6
Mean Dry Plant Density (glmz)

Figure 12. Relationship of mean dry plant density (g/mz) and mean diel pH amplitude.

40

 

LL].

Once this was discovered the treatment was cut short and the remainder of the
river was not treated at lethal levels of TF M. The Chinook salmon that did perish were in
spawning or post-spawning condition which is a stressful period in the ﬁsh’s life cycle.

The abundance of aquatic macrophytes collected was relatively low in all rivers
sampled. No plants were collected on the Riﬂe or Pere Marquette River, although there
were plants within the river, just not in the random transects. Figure 12 shows the
relationship between mean pH amplitude and mean dry plant density. There is a clear
relationship between plant abundance and pH levels, with mean pH amplitude increasing
with density of plants. As would be expected, there is also a relationship between mean
dissolved oxygen amplitude and mean plant density (Figure 13). The rise in mean plant
density has a positive effect on the mean DO amplitude as this is a byproduct of
photosynthesis. This was most evident in the Chippewa River where production and
plant biomass were very high compared to the other rivers sampled.

Across all rivers, I observed a positive relationship between mean dissolved
oxygen diel amplitude and mean pH diel amplitude (Figure 14). The relationship appears
to be non-linear, but the number of data points for high amplitude pH and DO variation is
limited and the apparent nonlinearity was driven by the Chippewa River.

Spatial and temporal variation in pH can be caused by many factors such as
springs, run-off, plant production, river substrate, industrial efﬂuent, and geology to name
a few (Kalff 2002). Because sea lamprey treatments occur in a wide variety of settings
and areas, this may occasionally create unique localized conditions. Michigan, Ohio,

Pennsylvania, New York, Wisconsin, and Canada all have very different geological make

up.

41

 

Mean Plant Density and Mean Diel Dissolved Oxygen

 

 

Amplitude
6 O
o 5 '
u
3
=‘e‘
< ‘ ‘
O
D
5
s 3 o
2 9 O
0 9
1}
1 I I I I I I I
0 0.2 0.4 0.6 0.8 1 1.2 1.4
Mean Dry Plant Density (glmz)

Figure 13. Relationship of mean dry plant density (g/mz) and mean diel DO amplitude.

This may account for the spatial differences in some areas based on the alkalinity
of the water. Where the geology is more calcareous, the alkalinity will tend to be higher
(Kalff 2002) which will tend to moderate natural ﬂuctuations in pH. This can also be the
case for ground water additions as they pass through the surrounding bedrock and
substrate.

The effects of impoundments were also visible from data on rivers with dams.
The Muskegon and Riﬂe Rivers most clearly showed this disturbance. Control loggers
were placed in the ponds above both of these dams because TF M treatments occurred at
the dam. The White and Betsie Rivers also had dams but did not show the same effect.
The White River dam in Hesperia is a smaller dam than that of the Muskegon or Riﬂe

Rivers, and treatment of the White River began below the dam.

42

 

 

Relationship of mean dissolved oxygen amplitude to mean pH

 

 

amplitude
0.55 ~
0 e
u 0.5 '
3
r:
75- o.45 -
E
‘3 O
:1: .
a. 0.4
c o
3 0.35 -
E ’ .
0.3 - e i 9‘7
e e
0.25 - -
I
g
3.
002 I I I I j I
1 2 3 4 5 6 7

Mean dissolved oxygen amplitude

Figure 14. Relationship between mean diel dissolved oxygen amplitude and

mean diel pH amplitude.

The control logger was placed below the dam which comes off a pond. The
control logger on the Betsie River was placed above a low-head barrier dam, and this
river reach ﬂows more like a river as the pond that feeds that area is farther upstream. On
both the White and Betsie Rivers a dual pH peak was seen on a daily basis. Both rivers at

one point come off of a pond, so there may be a cycle occurring that has yet to be

 

. explained.

Fish kills have historically occurred in short random stretches of river that are not
easily predicted. This more than likely has occurred due to natural spatial and temporal
pH ﬂuctuations. It is possible that localized pockets of pH suppression due to TFM
occur, but most likely TFM additions are exacerbating pre-existing variability in pH. In

all previous cases, pH suppression occurred in less than 5 miles of river. This may not be"

43

as detrimental in a river like the Pere Marquette where 165 miles of river were treated.
But in a river like the Betsie in which only 11.5 miles were treated, 5 miles of
suppression can be devastating. Of the 1,502 treatments that occurred over the last 15
years only 6% have had a reportable ﬁsh kill. 1 In the grand scheme of things this is a
relatively rare occurrence, and should be considered a great success compared to early
TFM treatments, which were not considered effective unless teleost ﬁsh were killed.
Although this is a very small percentage it is still desirable to minimize major ﬁsh kills.
Although the primary cause of pH suppression was not due to plant abundance, it maybe
useful to explore alternative causes of ﬁsh kills to avoid future problems. There are
several rivers (Cedar River, Upper Peninsula, MI and the Little Salmon River, New
York) which have been reported to experience pH suppression in nearly the same place
every time the river is treated for sea lamprey. It would be valuable to perform an
analysis of the affected area during the next treatment of these rivers, as the primary
ﬁndings of this study suggest that natural spatial and temporal pH variation may be the

primary cause of ﬁsh kills.

44

 

REFERENCES

Bills, T. D., and D. A. Johnson. 1992. Effect of pH on the toxicity of TFM to sea
lamprey larvae and nontarget species during a stream treatment. Great lakes
Fishery Commission, Technical Report 57, Ann Arbor.

Bills, T. D., L. L. Marking, G. E. Howe, and J. J. Rach. 1988. Relation of pH to toxicity
of lampricide TFM in the laboratory. Great Lakes Fishery Commission,
Technical Report 53, Ann Arbor.

Boogaard, M. A., C. S. Kolar, J. E. Rivera, R. J. Scholeﬁeld, and D. A. Johnson. 2005.
Great Lakes Fishery Commission Progress Report: Study of issues related to
stream pH and lampricide treatments. Great Lakes Fishery Commission, Progress
Report, Ann Arbor.

Crowe, W. R. 1978. Great Lakes Fishery Commission History, Program, and Progress.
Pages 10-11. Great Lakes Fishery Commission Bulletin, Ann Arbor.

Dawson, V. K., K. B. Cumming, and P. A. Gilderhus. 1975. Laboratory efﬁcacy of 3-
Triﬂuoromethyl-4-nitrophenol (TFM) as a lampricide. United States Fish and
Wildlife Service, Investigations in Fish Control 63: 1-13.

Hauer, R. F., and G. A. Lamberti. 1996. Methods in Stream Ecology. Academic Press,
California.

Hynes, H. B. N. 1970. The ecology of running waters. Univ. Toronto Press. The
Chaucer Press Ltd. Bungay, Suffolk, Great Britain.

Kalff, J. 2002. Limnology. Prentice Hall, Upper Saddle River, New Jersey.

Krause-Jensen, D., and K. Sand-Jensen. 1998. Light attenuation and photosynthesis of
aquatic plant communities. Limnology and Oceanography 43 (3): 396-407.

Maki, A. W., L. D. Geissel, and H. E. Johnson. 1975. Toxicity of the lampricide 3-
Triﬂuoromethyl-4-nitrophenol (TFM) to 10 species of algae. United States Fish
and Wildlife Service, Investigations in Fish Control 56: 1-17.

Maki, A. W., and H. E. Johnson. 1976. Evaluation of a toxicant on the metabolism of
model stream communities. J. Fish. Res. Board Can. 33: 2740-2746.

McDonald, G. and C. Kolar. 2006. Research to guide the use of lampricides for
controlling sea lamprey. Great Lakes Fishery Commission, Research Theme

Paper, Ann Arbor.

Meyer, B. S. 1939. The daily cycle of apparent photosynthesis in a submerged aquatic.
American Journal of Botany 29 (9): 755-760.

45

Scholeﬁeld, R. J ., K. T. F redricks, K. S. Slaght, and J. G. Seelye. 1999. Effects of the
lampricide 3-Triﬂuoromethyl-4-nitrophenol (TF M) on pH, net oxygen
production, and respiration by algae. Great Lakes Fishery Commission,
Technical Report 63, Ann Arbor.

46

 

   
 
 

   

Ill“ 1 ll 1 H H ll

23 02956 4220