" 'RRARY
_....‘.gan State
University

 

 

 

This is to certify that the

thesis entitled

Early Life History of Sea Lamprey Larvae:
Emergence, Dispersal, and Effects of Density
on Movements

presented by
Amy Lynne Derosier

has been accepted towards fulfillment
of the requirements for

M.S. degreeinFiSho & Wildlo

[W [rfi

Major profeisér

 

 

 

Date /<;2///2//[)/

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

EARLY LIFE HISTORY OF SEA LAMPREY LARVAE:
EMERGENCE, DISPERSAL, AND EFFECTS OF DENSITY ON MOVEMENTS.

By

Amy Lynne Derosier

A THESIS

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

MASTER OF SCIENCE

Department of Fisheries and Wildlife

2001

ABSTRACT

EARLY LIFE HISTORY OF SEA LAMPREY LARVAE:
EMERGENCE, DISPERSAL, AND EFFECTS OF DENSITY ON MOVEMENTS.

By

Amy Lynne Derosier

Since the invasion of sea lamprey into the Great Lakes, there has been a major effort to
control their populations. Managers want to decrease their reliance on chemical control
methods and increase their reliance on alternative control methods, such as: increased
trapping, continued use of barriers, and the sterile male release program. To determine
the potential effectiveness of these alternative control methods more information is
needed on the life cycle of sea lamprey, especially during the first year of life. I
investigated emergence of sea lamprey from nests, their dispersal away from nests during
their first growing season using field and microsatellite methods, and the relation
between density and movements after settlement in age-0 sea lamprey. Emergence
occurred over a short period of time, between 8 and 14 days and during the darkest hours
of the night, between 1200 and 0300. The numbers of prolarvae produced from nests was
quite variable and ranged from 87 to 20,713. Despite the potential for large numbers of
prolarvae emerging over a short period of time, densities of age-0 sea lamprey were low,
on average between 4 and 10 per m2. The low densities of age-0 sea lamprey seen in the
field may be in part due to their ability to disperse widely, at least 874 m. Density
independent dispersal or movements were seen during warmer temperatures, possibly
allowing individuals to establish low densities such that density-dependent dispersal seen

during cooler temperatures is not very prevalent.

ACKNOWLEDGMENTS

I would like to thank my advisor, Dr. Michael L. Jones, for being the mentor I needed
and for being a friend. I would also like to thank my committee, Drs. Dan Hayes and
Rich Men'itt, for providing equipment or lab space and for always making time when I
was looking for advice. I would like to thank the Great Lakes Fishery Commission for
providing the funding for this project and also the sea lamprey community for providing
such a supportive and stimulating environment. Thank you to the US Geological
Survey’s Hammond Bay Biological Station (HBBS) and the US Fish and Wildlife
Service’s Lundington and Marquette stations for providing equipment and fish. Special
thanks to Roger Bergstedt and Bill Swink for advice on experimental design and writing
and to the staff at HBBS for their misc. help and making my summers entertaining.
Thanks to Greg Wright and CORA for allowing me to work on Nunns Creek. I would
like to thank Dr. Scribner and the folks in his lab (especially Jennifer Warillow) for
helping me get over my fear of genetics. I had some great people work with me on my
project, this work could not have been done with out them: Adam Segerlind, Jesse
Gabbard, Steve Nimcheski, Brandon Schroder, Amy Morgan, John Netto, Maggie Kurtz,
Ben Drake, and Nate Nye. My lab mates: Steve Haeseker, Mike Steeves, John Netto,
Katherine Smith, and Amber Peters, always provided a great working atmosphere but
always knew when to lighten things up — Thanks! Thanks go to my family and friends,
who have always supported me — I rely on their love! And last but not least, thank you

Kevin for always knowing how to make me laugh and supporting me when I needed it.

iii

TABLE OF CONTENTS

LIST OF TABLES: ................................................................................. vi
LIST OF FIGURES: .............................................................................. vii
GENERAL INTRODUCTION:................................... .............................. 1
CHAPTER 1: SEA LAMPREY EMERGENCE FROM NESTS ........................ 5
INTRODUCTION ................................................................................................................ 5
METHODS ........................................................................................................................ 7
Study Streams ............................................................................................................... 7
Field sampling .............................................................................................................. 8
Laboratory experiments ............................................................................................. I 0
Secondary analysis ..................................................................................................... 1 1
Data analysis .............................................................................................................. I 1
RESULTS ..................................................................................................................... 12
Field sampling ............................................................................................................ 12
Laboratory experiment ............................................................................................... I 5
Secondary analysis ..................................................................................................... I 5
DISCUSSION ................................................................................................................... 16

CHAPTER 2: DISPERSAL OF AGE-0 SEA LAMPREY DURING THE FIRST

GROWING SEASON ............................................................................. 34
INTRODUCTION .............................................................................................................. 34
METHODS ...................................................................................................................... 36

Study design ................................................................................................................ 36
Study streams .............................................................................................................. 37
Field methods ............................................................................................................. 37
Microsatellite methods ............................................................................................... 39
Data analysis .............................................................................................................. 40
RESULTS ........................................................................................................................ 43
DISCUSSION ................. 46

CHAPTER 3: THE EFFECTS OF DENSITY ON THE MOVEMENTS OF SEA

LAMPREY LARVAE IN THE LABORATORY ........................................... 61
INTRODUCTION .............................................................................................................. 61
EXPERIMENTAL DESIGN: .............................................................................................. 62
EXPERIMENTAL METHODS: .......................................................................................... 63

Age-0 ammocoete trials: ............................................................................................ 63
Age 1 and older ammocoete trials: ............................................................................ 65 .

iv

Data analysis: ............................................................................................................ 65

RESULTS ........................................................................................................................ 66
Age-0: ......................................................................................................................... 66
Age-1 +.° ...................................................................................................................... 6 7

DISCUSSION ................................................................................................................... 67

REFERENCES: .................................................................................... 75

LIST OF TABLES

Chapter 1:

Table 1: Dates that nests were first located and the dates and numbers of male and female
spawners seen on nests in the Trout River. (* denotes that nest was re-worked but no
spawners were seen, - denotes no spawners were seen) ............................................ 22

Table 2. Total number of prolarvae collected in each nest for each stream, the total
number of prolarvae collected by stream, the mean number of prolarvae collected per

nest, and the standard deviation of prolarvae collected per nest. ............................... 22
Table 3: Mean (and standard deviation) for the beginning, peak, and end of prolarval

emergence (or burrowing) for all study components. ................................................ 23
Chapter 2:

Table 1: Mean densities of age-0 ammocoetes (per m2) within zones by nest and sampling
event. .......................................................................................................................... 51

Table 2: Summary of microsatellite loci used in analysis: H(O) is the observed
heterozygosity, H(E) is the expected heterozygosity, Exc1(l) is exclusionary power
without any known parent, Excl(2) is the same as Excl(1) but when one parent is
known. ........................................................................................................................ 51

Table 3: Efficiency values for Ogemaw Creek and the Carp River by zone and sampling
event. .......................................................................................................................... 52

Table 4: Estimated abundance’s of age-0 ammocoetes in sampling reaches below nests.52

Chapter 3:

Table 1: Number of ammocoetes that moved out of source trays and where they settled
for each density and month. ....................................................................................... 72

vi

LIST OF FIGURES

Chapter 1:

Figure 1: Schematic diagram of nest locations in the Trout River, nest numbers are in
bold, and inset showing river location (not to scale) .................................................. 24

Figure 2: Schematic diagram of nest locations in the Black Mallard River, nest numbers
are in bold, and inset showing river location (not to scale) ........................................ 25

Figure 3: Average daily stream temperatures in the Black Mallard River (A) and the
Trout River (B) during the spawning and emergence period and relative water level
during the emergence period. (*denotes late spawnings) .......................................... 26

Figure 4: Box plot of the percent of emergent prolarvae collected during set 1 (2100-2400
h), set 2 (2400-0300 h), and set 3 (after 0300 h). The whiskers Show the minimum
and maximum values, the box depicts the 25 and 75 % quartiles, and the dash the
median. ....................................................................................................................... 27

Figure 5: The number of prolarvae collected each day by nest for the Black Mallard River
and the average daily stream temperatures and relative water levels during the
emergence period. ...................................................................................................... 28

Figure 6: The number of prolarvae collected each day by nest for the Trout River and the
average daily stream temperatures and relative water levels during the emergence
period .......................................................................................................................... 29

Figure 7: Scatterplot of the number of prolarvae produced vs. the number of spawners
seen on nests in the Trout River. ................................................................................ 30

Figure 8: Distribution of lengths for emerging prolarvae for each stream, the Black
Mallard River (A) and the Trout River (B). ............................................................... 31

Figure 9: Scatterplot of lengths for emerging prolarvae over time in the Black Mallard
River (A) and in the Trout River (B) .......................................................................... 32

Figure 10: Boxplot of cumulative degree days for the peak of prolarval emergence for
each study component, the line is the mean, the box represents the standard deviation,

and the whiskers are the minimum and maximum values. ........................................ 33
Chapter 2:
Figure 1: Locations of study streams. .............................................................................. 53

vii

Figure 2: Schematic showing sampling zones. The near zone extends from the nest to 50
m downstream, the middle zone extends from 55 to 155 m downstream from the nest,
and far zone extends from 160 to 310 In downstream from the nest. There is a 5 m
buffer between each zone. .......................................................................................... 54

Figure 3: Schematic of adaptive sampling design. Figure A shows the Six initial random
plots sampled, boxes in gray indicates that at least one age-0 ammocoete was
collected, i.e. the condition was met. Figure B shows those same six random plots
with the additional sampled neighborhood plots in black. ......................................... 55

Figure 4: Histograms of ammocoetes collected in Ogemaw Creek (A) and the Carp River
(B) ............................................................................................................................... 56

Figure 5: Box plots of the mean densities of age-0 ammocoetes per m2 by zone for
Ogemaw Creek (white boxes) and the Carp River (shaded boxes). The box
represents the 25th and 75th quartile, the line the mediam and the whiskers the
minmum and maximum. ............................................................................................ 57

Figure 6: Box plots of the mean densities of age-0 ammocoetes per m2 by sample event
for Ogemaw Creek (white boxes) and the Carp River (shaded boxes). The box
represents the 25th and 75th quartile, the line the mediam and the whiskers the
minmum and maximum. ............................................................................................ 57

Figure 7: Histogram of the number of ammocoetes associated with each male parent for
upstream most nest (A) and downstream most nest (B) ............................................. 58

Figure 8: Histogram of values for randomly distributed nest-mate pairs, 1000
simulations. Asterisk denotes the number of nest-mate pairs in observed data. ....... 59

Figure 9: Scatterplot of average ammocoete lengths in mm in relation to the density of
the habitat patch in Ogemaw Creek (A) and the Carp River (B). The circles represent
ammocoetes during the first sampling event, the squares the second sampling event,

and the diamonds the third sampling event. ............................................................... 60
Chapter 3:
Figure 1: Diagram of raceway tanks and layout of the source and sink trays. ................ 73

Figure 2: Average percent of ammocoetes that moved out of the source tray in August
(n=11) and September (n=19) for each density. Error bars represent :l: 1 standard
error. ........................................................................................................................... 73

Figure 3: Percent of ammocoetes that did not move out of source tray by average daily
water temperature for all densities and ages. ............................................................. 74

viii

General Introduction:

Since the invasion of sea lamprey into the Great Lakes, there has been a major effort to
control their populations. The main control method used is a chemical called 3-
trifluoromethyl—4-nitrophenol (TFM), which is used to treat streams to kill larvae (Smith,
1971, Meyer and Schnick, 1983). This method is very effective and has been used since
the early 1960’s (Torblaa and Westman, 1980). However due to increasing costs of the
chemical and public concerns about using chemicals in streams, the Great Lakes Fishery
Commission wants to reduce the reliance on chemical controls (Strategic Vision of the
Great Lakes Fishery Commission, 1992 and 2001). To do this, managers will need to
rely more heavily on alternative control methods such as: increased trapping, continued

use of barriers, and the sterile male release program.

These alternative control methods may not be as effective as chemical control however.
Chemical controls target the larval stage of sea lamprey, whereas the alternative control
methods target reproduction. This difference in targeted life stages complicates effective
control. Chemical controls reduce the population just before the parasitic life stage,
while with using alternative control methods there are many life stages between when the

population is reduced and when they become parasites, allowing time for compensatory

mechanisms to play out. Compensatory mechanisms are density-dependent demographic
responses that effectively increase population growth at low densities and decrease
population growth at high densities. There is evidence to suggest that compensatory
mechanisms exist in sea lamprey populations (Smith, 1971; Purvis, 1979; Heinrich et al.,
1980; Morman, 1987; Murdoch etal., 1992), yet the evidence is far from conclusive
(Jones et al., in review). To effectively control sea lamprey populations using alternative
control methods, managers need to determine at what life stages compensatory
mechanisms exist and their magnitude. If sea lamprey populations are not able or only
slightly able to compensate, the effectiveness of the alternative control methods may not
be compromised. But if compensatory mechanisms are strong, control efforts using
alternative methods may not decrease the population sufficiently to justify the decision.
To determine the potential effectiveness of these alternative control methods more
information is needed on the demographics of sea lamprey particularly during

reproduction and larval life stages.

The life cycle of the sea lamprey is anadromous. Adult sea lamprey migrate into streams
in April and May and begin spawning when mean water temperatures reach around 11 °C
(Applegate, 1950). Sea lamprey spawn in cobble and gravel areas, generally riffles,
where water velocities are between 0.5 and 1.5 m/s (ManiOn and Hanson, 1980). Males
start the construction of nests, which consist of a depression and a perpendicular ridge of
cobble downstream; nest construction can take 1 to 3 days (Manion and Hanson, 1980).
Once a female joins a male, spawning commences. Eggs are released in the depression

and carried by the water current into the interstices of the cobble ridge (Applegate, 1950).

The adults kick up sand with their tails to help cement eggs into the ridge. Eggs hatch
after approximately 10 to 13 days when held at a constant water temperature of 18 °C
(Piavis, 1961); once hatched they are considered prolarvae. Prolarvae are able to burrow
17 to 33 days after fertilization. The burrowing stage is termed developmental stage 17
(Piavis, 1961), and is the point at which they are assumed to emerge from nests.
Prolarvae become larvae after burrowing and when the yolk is passed through the gut and
the digestive system is fimctional; this occurs 33 to 40 days after fertilization (Piavis,

1961).

The larval stage of the sea lamprey can last from 3 to 12 years depending on the
productivity of the stream. Larvae live in depositional areas of streams and filter-feed on
detritus (Sutton and Bowen, 1994). When larvae reach a length of around 120 mm they
begin to metamorphose into parasites. Following metamorphosis, they migrate
downstream into the lakes or ocean, between September and May, and begin feeding on
fish, such as sahnon and trout. They grow quickly and spend only about 12 to 20 months
feeding as parasites, after which time they become adults and migrate back into streams

to complete the life cycle.

The early life stages of fish are generally thought to heavily influence recruitment
(Cushing, 1996; Wootton, 1990). Hjort (1914) proposed that the greatest mortality
occurs during the early life of fish and coined this the critical period. Recent authors
suggest that this critical period occurs during transitions (Benoit and Pepin, 1999), such
as hatching, emergence, and the shift from endogenous (yolk) to exogenous (active)

feeding (Lasker, 1981; Diana, 1995). Sea lamprey fecundity is estimated to be high,

producing between 55,000 and 69,000 eggs per female (Manion and Hanson, 1980).
However, typical age-l and older larval densities are reported to be around 2 per m2

(Jones et al., in review), suggesting that mortality is high during the first year of life.

Little is known of the first year of life of the sea lamprey. Sea lamprey at this stage are
small (4-20 mm in length, 1-2 mm in width) and hence difficult to sample. The.
embryological development has been fiIlly described, as have some aspects of the
emergence stage. Because of the lack of information on age-0 sea lamprey and the
apparent importance of the first year of life in determining recruitment in fish, I chose to
explore three aspects of age-0 sea lamprey ecology. I chose to expand on the description
of emergence that has been reported by Applegate (1950) and Manion and McLain
(1971), by following emergence fiom 10 nests in each of two streams. I also investigated
age-0 sea lamprey dispersal or distributions during the first growing season, which has
not been done in the past. I also examined movements of larvae after settlement in

relation to density and temperature / season.

Chapter 1:
Sea lamprey emergence from nests.

Introduction

Recent authors suggested that the greatest mortality in fishes occurs during transitions
(Benoit and Pepin, 1999), such as hatching, emergence, and the shift from endogenous
(yolk) to exogenous (active) feeding (Lasker, 1981; Diana, 1995). Many freshwater fish
hatch within nests but do not emerge for some time after, often just before the switch
from endogenous to exogenous feeding. The timing and pattern of emergence will
therefore determine the conditions that larvae experience before and during their first
feeding. The timing of emergence also determines the length Of the first growing season
(Elliott and Hurley, 1998) and thus the size that age-0 fish reach before entering their first
winter. Hence, emergence can be a key stage in a fish species’ life history and

knowledge of this stage may provide insight into processes affecting early larval survival.

Although sea lamprey emergence from nests has been previously investigated, a full
description of this stage is lacking. Applegate (1950) followed only 3 nests from
fertilization to emergence, and reported the number of larvae emerging each day.
Emergence occurred 19 to 20 days after fertilization. Manion and McLain (1971)
monitored 48.nests but only reported the average (22 days) and the maximum (34 days)

number of days it took for prolarvae to emerge. These authors did not describe the

duration (Elliott, 1984; Snucins et al., 1992) or the diel pattern (F ield-Dodgson, 1988;
Kempinger, 1988) of sea lamprey emergence. If emergence is spread over a relatively
long period of time, then that species may be bet hedging (Hopper, 1999); the risk of all
larvae emerging during a period when environmental conditions are poor is lessened. On
the other hand, if emergence happens over a relatively short period of time, the emerging
larvae may experience reduced mortality due to swamping of predators. Because
lamprey larvae are more or less passively transported to burrowing habitats (Applegate,
1950), this latter strategy may also create localized high densities, resulting in intense

intra-specific competition, dependent upon the numbers emerging.

One common method used for reporting the duration of developmental stages is degree
days (Cassehnan, 1995; Karnler, 1992; Elliott, 1984; Kempinger, 1988; Ross and Merritt,
1978), but this has yet to be used for sea lamprey. This index combines both the numbers
of days development takes to reach different stages and the fluctuating temperatures to
which eggs and prolarvae are exposed during this period. This index can be easily
compared among locations and among species. In addition, if emergence can be
quantified using degree days, managers could use this information to more effectively

treat streams because eggs are not vulnerable to TFM (cite).

Applegate (1950) is the only author that reports the lengths of sea lamprey emerging from
nests. He provided a histogram of lengths for one nest (222 individuals, average length =
8.54 mm). Lengths of larvae at emergence are commonly reported (Field-Dodgson,

1988; Randall, 1982), and needs to be further explored in sea lamprey. Looking across

 

nests and streams may provide insight into the ecology of larvae at this life stage. Larger
larvae may enjoy a competitive advantage over smaller larvae. This advantage may be

caused by large eggs via larger females (Kamler, 1992).

The objective of this component of my study was to expand on the previous descriptions
of sea lamprey emergence. Specifically I set out to: (1) describe emergence relative to
spawning activity; (2) examine the diel pattern of emergence; (3) compare the numbers
and size (lengths) of prolarvae produced in individual nests and between streams; (4)
quantify the duration of the emergence period; and (5) determine if emergence timing can

be predicted using degree days.

To address these objectives, I conducted a field study, a laboratory experiment, and a
secondary analysis of published data. Ten nests, in each of two streams, were monitored
fiom nest construction to prolarval emergence. Eggs were raised in the laboratory until
prolarvae burrowed into a sand substrate (similar to Piavis, 1960), to test whether
prolarvae burrow at the same time as prolarvae emerge from nests. To supplement these
two components, data from Applegate (1950) and Piavis (1961) was reanalyzed to

calculate degree days.

Methods

Study Streams
A tributary to Lake Huron, the Trout River is located just north of Rogers City in Presque
Isle County, Michigan. It has a catchment size of 36.8 miz. There is a sea lamprey

barrier on the river; my study nests were located below the barrier within a 160 m reach

(Figure l). The instream habitat is suitable for both sea lamprey spawning (cobble,
gravel) and for larval rearing (depositional, fine sediments). This stream is regularly
treated with TFM, and was treated while eggs were incubating in nests but before

emergence began.

The Black Mallard River is also a tributary to Lake Huron and is about 20 mi north of
Rogers City. The catchment size is 27.4 rniz. The Black Mallard River is also a known
sea lamprey producer and is regularly treated with TFM. There is less spawning and
larval habitat in the Black Mallard River than in the Trout River and water levels are
often quite low by mid-surmner. Applegate (1950) also used this river as a study stream,
at that time it was named Carp Creek. Study nests were within a 175 In section of the

river (Figure 2).

Field sampling

A continuous temperature monitor (HOBO, Onset Computer Corp., Pocasset, MA) was
installed in each stream to record daily water temperatures (every 12 min) throughout the
Spawning and emergence period. The daily water levels during the emergence period

were recorded using a temporary staff gauge in each stream.

Study streams were walked daily, starting on May 20, 2000 to locate sea lamprey nests
and monitor spawning activity. Monitoring began before any nesting activity had taken
place in the study reach. When a nest was found it was marked with a flag and the

following data were recorded: 1) the date it was first located, 2) dates spawners were

present, 3) the sex of adults was recorded, and 4) it was noted if the nest looked like it
had been re-worked. When a nest is first built, it is easy to locate because the rocks
creating the nest have been turned over and are clean (no periphyton) and hence brighter
than the surrounding rocks. After a couple of days, the rocks in the nest become covered
in periphyton again. Therefore, if a nest is re-worked a few days after first being built, by
either the same adult or a different adult, it is noticeable. A sample of eggs was collected
a few days after a nest was found to confirm successfiil deposition and the stage of
embryological development (Piavis, 1961). Eggs were staged following Piavis’ (1961)
description of sea lamprey development and an expected hatch and emergence date was
calculated. Drift sampling began immediately after the expected hatch date but well

before the expected emergence date.

Prolarvae were collected below nests using a rectangular drift net, 45.5 x 15 cm, with a
mesh size of 350 pm. The drift nets were set less than 2 m downstream from 10 nests in
each stream, approximately every other night from June 8 to July 13, 2000 in the Black
Mallard River, and June 19 to July 14, 2000 in the Trout River. Young-of-the-year sea
lamprey ammocoetes drift mainly at night (Johnston, 1997; Bennett and Ross, 1995). To
confirm emergence followed this diel pattern, I conducted a preliminary study. Drift nets
were set below two nests in Weldon Creek, Mason County, Michigan, from July 5 to July
24, 1998. Of the 105 emergent prolarvae collected, only one was collected during the
daylight hours (0900 to 1600 h). Therefore, in both study streams, nets were set for three

3 h periods (9 h total) each night, beginning between 20:30-21 :30 (before dusk) and

ending between 05:30-07:30 hrs (after dawn). Prolarvae caught in each net were either

counted on site or preserved in 70% ethanol for later enumeration.

Prolarval lengths were measured using a dissecting scope attached to a monitor to digitize
prolarvae. The program OPTIMAS version 4.10 (BioScan, Inc., 1987-1993) was used to
calculate lengths. Approximately 1000 prolarvae were measured from each stream,

stratified by nest.

Laboratory experiments

Prolarvae were raised in the laboratory from sea lamprey spawned June 10 — 13 and June
16-17, 2000. Eggs and milt were removed fi'om ripe sea lamprey and spawned into a
beaker with approximately 1 liter of filtered Lake Huron water. After 30 minutes, beaker
water was changed several times and eggs were introduced into a 10 liter glass battery
jar; each spawning was placed in a different battery jar. Eggs were arranged in a single
layer. Battery jars were placed in an insulated water bath, and held at a constant
temperature of 18 °C (optimal temperature for development according to Piavis, 1961).

F our days after fertilization, dead eggs were removed. Water was changed with filtered

water every other day.

After eggs hatched, prolarvae were transferred to 100 ml beakers in the water bath. Each
beaker was filled with approximately 20 ml of sand and 80 ml of water. For each
spawning date, 10 prolarvae were introduced into each of three replicate beakers. Each
day the water was changed and the number of prolarvae remaining on the surface of the

sand was counted; those missing were assumed to have burrowed.

10

Secondary analysis

I reanalyzed data from Applegate (1950) and Piavis (1961) by converting daily water
temperatures into cumulative degree days to emergence (Applegate) or to stage 17
(Piavis). From Applegate’s study I used the data reported for three nests from the
Ocqueoc River, in Presque Isle County, Michigan: the date spawning was completed, the
mean date prolarvae were captured in plankton nets, and the number of prolarvae
collected. The appendix to his work listed mean daily stream temperatures. For the

Piavis publication, I used trials with rearing temperatures of 15.5, 18, and 21 °C. He

f

recorded the time to reach each developmental stage; I used stage 17 as a surrogate for

emergence.

Data analysis

I calculated cumulative degree days (CDD) using the following equation: DD=Z (t — 7),
where t is the average daily stream temperature (°C), and the constant, 7, is the
temperature below which no development occurs (Piavis, 1961). The estimated
fertilization date was used as the starting point for the calculation of crunulative degree

days.

All data were analyzed using parametric statistical methods in Statistica (Statsoft, Inc.
1998). I calculated the cumulative percent of prolarvae that emerged each day; those
dates and CDD that corresponded most closely to the 10th, 50‘“, and 90th cumulative
percentiles were defined as the start, peak, and end dates and CDD. All differences were

considered significant if the p-value was less than 0.05.

11

I tested the null hypotheses that: 1) prolarval lengths were the same for all nests within a
stream and between streams; 2) the beginning, peak, and end CDD for the emergence
period were the same for both streams; 3) the peak of emergence (CDD) was the same for
my field data and Applegate’s (1950) data; 4) the peak of burrowing (CDD) was the same
for my laboratory study and Piavis’s (1961) study; and 5) there was no difference
between the peak of emergence in the field studies and burrowmg in the laboratory
studies. Many of these hypotheses provide insight into whether CDD can be used to

predict prolarval emergence.

Results

Field sampling

Water temperatures were similar between the two streams during Spawning; average
daily temperatures ranged from 15-18 °C in the Black Mallard River (Figure 3A) and 13-
19 °C in the Trout River (Figure 3B). The average water temperature during emergence
was 20.6 °C (range 15.5-23.9 °C) in the Black Mallard River (Figure 3A) and 21.7 °C
(range 17.6-24.3 °C) in the Trout River (Figure 3B). The relative water stage increased in
the Black Mallard River just after drift sampling began due to heavy rains, then gradually
decreased over the sampling period (Figure 3A). The Trout River’s stage remained
relatively constant, with a slight decline over time (Figure 3B). The heavy rains,
corresponding rapid drop in water temperature, and increased flow in the Black Mallard

River, provided a contrast to the more stable Trout River.

12

 

Nests were found in the Black Mallard River from May 22 to May 30, but no spawners
were seen on nests. However, adults were seen late in the season. Upstream of nest 10 a
new nest was built by a male and two females on June 23 (mean water temperature = 20.3
°C), nest 9 was reworked on June 23, and nest 6 was reworked on July 1 (mean water
temperature = 21.3 0C). Spawning occurred in the Trout River study reach from May 23

to June 12; spawners were seen on most nests (Table 1).

Emergence tended to be concentrated during the middle (darkest hours) of the night.
Approximately 24 — 30 % of prolarvae emerged between 2100 h and 2400 h, 66 - 68 %

emerged between 2400 h and 0300 h, and 3 — 7 % of prolarvae emerged after 0300 h

(Figure 4).

The number of prolarvae produced from nests was quite variable and generally showed a
unimodal distribution (Figure 5 and 6). In the Black Mallard and Trout Rivers, 6,037
prolarvae (range 87 — 1,900 per nest) and 83,541 prolarvae (range 346 — 20,713 per nest)
were collected, respectively (Table 2). The maximum number of prolarvae collected in
one net over a 3 h period was 736 in the Black Mallard River and 8,348 in the Trout
River. For the Trout River, where adults were observed on nests, there is no relationship

between the number of spawners seen on a nest and the number of prolarvae produced

(Figure 7).

Prolarvae emerging from nests were on average smaller in the Black Mallard River than

those in the Trout River (p<0.05, t=17.77, df=2390). The average length of emerging

13

TI

prolarvae in the Black Mallard River was 7.96 mm (range 4.86 — 12.45 mm; Figure 8A)
and in the Trout River, 8.65 mm (range 4.69 — 12.80 mm; Figure 8B). Within streams
there were significant differences in lengths between nests (Black Mallard River:

p < 0.05, F9,313=12.28; Trout River: p < 0.05, F4,1333=188.11). No trend in prolarval
lengths was seen over time (Figure 9), although there appears to be more scatter or
variability in lengths as the emergence period progressed. Larger individuals were
collected later in the season but some of these animals looked as though they had been
feeding and hence were not emerging prolarvae but rather ammocoetes that were re-

distributing.

Emergence began in the Black Mallard River on average 23 days (range 18-26 days) after
the nest was constructed and at a mean CDD (i standard deviation) of 254 i 24 (Table
3). Cumulative degree days in the Black Mallard River were calculated using catches
before July 1, because prolarvae collected after were assumed to be from later spawnings
(June 23 and July 1) than the ones I was following (May 22 to May 30). The Trout River
began emerging 22 days (range 18-25 days) after fertilization and at an average CDD of
318 i 37, which was significantly greater than the Black Mallard River (p < 0.05, t: 4.58,

df=18).

The Trout River also took longer to reach the peak of emergence than the Black Mallard
River (p<0.05, t=5.09, df=18). The Black Mallard River peaked at an average of 26 days
(range 23-28 days) or 290 i 17 CDD (Table 3). The Trout River emerged 31 days (range

26-39 days) after fertilization or 370 i 47 CDD.

14

Emergence ended in the Black Mallard River at a mean of 29 days (range 23-35 days) or
338 i 28 CDD after fertilization. Emergence ended in the Trout River after 36 days (28-
41 days) and at an average of 432 i 53 CDD. Again, emergence in the Black Mallard

Creek ended earlier than the Trout River (p<0.05, t=4.9, df=l8) (Table 3).

On average, prolarvae in the Black Mallard River emerged over a 7 day period (range 3-
12 days) and a mean of 83 i 38 CDD. The Trout River prolarvae emerged over 8 days
on average (range 4-14 days) or 113 i 41 CDD. The emergence period was not different

between the two streams (days: p=0.28, t=1.12, df=18; CDD: p=0.11, t=1.67, df=18).

Laboratory experiment

Prolarvae began to burrow in the laboratory experiments at an average CDD of 261
(range 242 — 286), Similar to the field results, and finished burrowing at an average of
294 CDD (range 253 — 396). The peak of burrowing occurred at an average of 24 days
(range 22-28 days) after fertilization or 267 i 12 CDD. The CDD to peak emergence in
the Black Mallard and Trout River’s were greater than the peak of burrowing in the
laboratory experiments (Black Mallard: p<0.05, t=3.00, df=15; Trout River: p<0.05,

t=5.68, df=15), i.e. larvae in the study streams took more heat to emerge (Figure 10).

Secondary analysis

The peak of emergence in Applegate’s (1950) study nests occurred 19 to 20 days after

fertilization or 303 i 3 CDD. Prolarvae began emerging around 274 CDD and finished

15

emerging around 319 CDD. There were no significant differences between when 50% of
prolarvae emerged in Applegate’s nests and in the Black Mallard River nests (p < 0.25,
t=t.21, df=11). However in the Trout River, CDD’S to peak emergence was greater than
in Applegate’s data (p<0.05, t=2.43, df=11). Prolarvae in my laboratory experiments
burrowed earlier than prolarvae emerged in Applegate’s nests (p<0.05, t=4.86, dfi8)

(Figure 10).

In Piavis’s experiments, prolarvae reared in 15, 18, and 21 °C water temperatures began
burrowing at 214, 193, and 212 CDD, respectively. Using the middle day that stage 17
occurred as 50% burrowed, I calculated a mean CDD of 248 i 41. There were no
differences between the CDD to 50% burrowed in my laboratory experiments and
Piavis’s experiments (p = 0.2769, t=1.17, df=8). There was also no difference between
when prolarvae emerged in Applegate’s nests and when prolarvae burrowed in Piavis’s
experiments (p=0.08, t=2.29, df=4). Both of my study streams had greater CDD at peak
emergence than did Piavis’s experiments (Black Mallard: p<0.05, t=2.68, df=11; Trout

River: p<0.05, t=4.06, df=1 1).

Discussion

Water temperatures needed for sea lamprey spawning and emergence are quite broad.
The temperatures seen in my study streams during both spawning (13-19 0C) and
emergence (15-24 °C) are similar to those reported by other authors (1 1-24 °C:
Applegate, 1950; 10-18.5 °C: Manion and McLain, 1971). The late spawnings in the
Black Mallard River (20-21 °C) occurred with in the range of reported spawning

temperatures. Applegate (1950) also witnessed spawnings late into the season.

16

The distributions of catch over time are generally unimodal, but some nests showed
additional peaks suggesting late spawnings. As an example, nest 6 was reworked on July
1 and showed an additional peak in the emergence distribution (< 10 individuals) around
July 12. This date corresponds to 176 CDD and is likely the start of the emergence
period for the prolarvae spawned on July 1. Hence, the later peaks in the emergence

distributions were removed when analyzing CDD.

The patterns of emergence for sea lamprey are similar to those of other species.
Emergence occurs predominately during the darkest hours of the night (1200 — 0300 h)
and declines sharply after 0300 h. Studies sampling drift or movements of stream young-
of—the-year fishes have also reported increased drift around midnight and a decline
towards dawn. (Brown and Armstrong, 1985; Bennett and Ross, 1995; Johnston, 1997).
The overall duration of the emergence period is short, on average it took 8 to 14 days for
80% ofprolarvae to emerge; Applegate (1950) reported duration’s of 3 to 4 days. This
difference is likely due to water temperature variations; even slight changes in
temperatures over time can drastically change the heat accumulated. Chinook salmon
also exhibit a short emergence duration, where the majority of fry emerge, on average, in

11 days (F ield-Dodgson, 1988), despite a much longer incubation period.

The numbers of emergent prolarvae varied greatly from nest to nest and stream to stream.

The Trout River produced 5,569 prolarvae per female seen, on average and the Black

Mallard Creek produced 604 prolarvae per nest, on average. No correlation was evident

l7

between the numbers of females present and the numbers of prolarvae produced.
Applegate (1950) collected only 222 — 622 prolarvae from individual nests. Manion
(1968) dismantled l9 nests before emergence and reported 1,763 to 10,545 prolarvae per
nest. He also noted that on average, a single nest (one pair of spawners) produced 3,240
prolarvae and a double nest (two pairs of spawners) produced 7,531 prolarvae. It is
possible that the numbers of prolarvae collected in my study are under-estimates because
prolarvae may have drifted around the mouth of the drift nets. Although Applegate
(1950) constructed enclosed raceways around nests to collect all emerging prolarvae, his
numbers may also be under-estimates. The raceways may have changed the water flow
to the nest and perhaps increased nest mortality. Regardless, the numbers of prolarvae
produced in nests and streams varied greatly, but can be large. Variations in nesting
success may in part explain the large amount of density independent variation in

recruitment success observed by Jones et al. (in review).

With such recruitment variation of emerging prolarvae, it is possible that high densities
could occur after settlement. Due to their small size, prolarvae are not good swimmers
and rely on water currents to move them to burrowing habitats (Applegate, 1950). I
collected over 4,000 prolarvae in a single net in a 3 hr time period (maximum of 8,000 / 3
h) on multiple occasions. Again, it is possible that these are under-estimates of the true
numbers of prolarvae emerging. These large emergence events and the concentration of
emergence in a short period of the night suggest that large numbers of prolarvae could be

deposited in the same habitat patch. This overcrowding could potentially affect their

18

feeding rates (Yap and Bowen, in review), growth rates (Morman, 1987; Murdoch et a1,

1992), movements (Chapter 3) and hence survival.

The Trout River produced larger numbers of prolarvae per nest than the Black Mallard
River. This difference could be due to the heavy rains and subsequent high water levels
in the Black Mallard River that possibly decreased the efficiency of the drift nets.
However, the plotted data (Figure 5B-K) for each nest in the Black Mallard River,
suggests that the peak catches did not occur until well after the high waters started to
subside, indicating that this difference is not likely due to inefficient sampling. Therefore
the difference in production between my two study streams is more likely due to other

factors, such as egg or female size or water-quality.

In addition to producing more prolarvae, the Trout River prolarvae are on average larger
(8.65 mm, range 4.69 — 12.80 mm) than those in the Black Mallard River (7.96 mm,
range 4.86 — 12.45 mm) and they emerged later (Trout River: 370 CDD; Black Mallard:
290 CDD). Water temperatures between the two streams are relatively similar and hence
this seems unlikely to be the reason for these differences. One possibility is that larger
females spawned in the Trout River compared to the Black Mallard River. The Trout
River had more adults in the system possibly creating higher competition for mates,
allowing only the larger females to Spawn. Or possibly, the Trout River attracted larger
and better-conditioned females due to water-quality. The size of a female will influence
the size of the offspring produced (Benoit and Pepin, 1999; Kamler, 1992; Elliott and

Hurley, 1998) and possibly their survival (Diana, 1995). Larger eggs tend to have longer

19

developmental periods in fishes (Ware, 1975; Economou, 1991). Furthermore, the rate of
yolk absorption tends to be high in small eggs and lower in larger eggs (Kamler, 1992).
Therefore, prolarvae in the Black Mallard River possibly absorbed their yolk and had to
emerge earlier to begin exogenous feeding. In contrast, the Trout River prolarvae
possibly had more yolk reserves due to larger eggs and hence higher survival. These
results suggest that knowledge about the linkage between condition of parents (Oconnor,
in review), egg size, and timing of emergence may provide more insight into larval

recruitment.

With the additiOn of the secondary analysis, my evidence suggests that cmnulative degree
days to emergence are different in the field than in the laboratory. The Black Mallard
and Trout River’s cumulative degree days to emergence were higher than in my
laboratory study and Piavis’ study (1961 ). It took more degree days to reach peak
emergence in the field, when temperature fluctuated, than in the laboratory, when it was
held constant. Traditionally, emergence is estimated to occur about 22 days after
fertilization (Piavis, 1961). However, my results show that laboratory results are

generally under-estimates of what is happening in the field.

The main control strategy in sea lamprey management relies on the chemical treatment of
streams with TFM. While in the nests, eggs and possibly prolarvae are less vulnerable to
TFM. Due to time constraints of the control agents, streams are now treated without
regard to when the current year-class in a stream will emerge. For some costly-to-treat

streams, it would be useful and more cost effective to know when the majority of

20

prolarvae have emerged so that an extra year-class can be targeted, thereby increasing the
time interval between costly treatments. Other stream sections require frequent
treatments because they are close to lakes and hence there is concern that lentic
populations could accumulate. These lentic areas can not be treated effectively. My
results from Chapter 2 suggest that larvae disperse widely and soon after emergence. If
managers want to effectively treat these short stream sections they need to treat soon after
emergence when age-0 ammocoetes are vulnerable, but not before emergence is
complete. Knowledge of the timing of emergence is therefore important. The results
presented herein could be used to estimate the optimal timing for treatments. If managers
want to target an additional year-class it may be wise to wait to treat until 36 days or at
least 432 CDD alter the spawning period. Unfortunately, to apply these results
knowledge of the timing of spawning is needed; spawning surveys are uncommon on
lamprey producing streams. On the other hand, many rivers have adult traps on them,
and if the relationship between the numbers of adults collected in traps and the timing of
spawning were available, this emergence work could be used more broadly to better

target optimal treatment times.

21

 

Table 1: Dates that nests were first located and the dates and numbers of male and female
spawners seen on nests in the Trout River. (* denotes that nest was re-worked but no
spawners were seen, - denotes no spawners were seen)

 

Nest Located Active Males Females

 

 

 

 

 

 

 

 

 

 

1 6/01 6/01 1 2
6/04 1 1

6/07 *
2 6/01 6/01 1 3
3 5/25 5/25 1 1
5/31 1 l
4 5/25 5/25 1 1
5/31 1 1
5 5/30 - - -
6 5/25 5/25 1 1
6/05 1 1
7 5/30 - - -
8 5/23 5/30 1 2
6/05 1 1
9 5/23 5/31 1 1
10 5/25 6/07 1 l

 

Table 2. Total number of prolarvae collected in each nest for each stream, the total
number of prolarvae collected by stream, the mean number of prolarvae collected per
nest, and the standard deviation of prolarvae collected per nest.

 

 

 

Nest Black Trout
Mallard River
River

1 87 346

2 267 14,880

3 201 1,316

4 535 12,015

5 1,900 1,044

6 164 14,632

7 109 3,986

8 768 1 1,093

9 666 3,516
10 1,340 20,713
Total 6,037 83,541
Mean 604 8,354
Std Dev 600 7,190

 

22

Table 3: Mean (and standard deviation) for the beginning, peak, and end of prolarval
emergence (or burrowing) for all study components.

 

 

 

Field Studies Laboratory Studies
Black Trout River Applegate’s Laboratory Piavis’s Lab
Mallard Data Exp. Study
River
Beginning 254 i 24 318 i 37 -- - -
Peak 290 i 17 370 i 47 303 i 3 267 i 12 248 i 41
End 338 i 28 432 i 53 -- -- ~-

 

23

Figure 1: Schematic diagram of nest locations in the Trout River, nest numbers are in
bold, and inset showing river location (not to scale).

 

   

 

 

 

 

 

l
8) 2
5.2m ® Q
I v
Trout River
26 m
3 4
0 ® ® 3
V V
20 m ®5
V Flow —_
O.

 

 

 

7
30 m 3 ® Nest
®8 V Drift Net

 

 

 

 

 

10

 

 

24

Figure 2: Schematic diagram of nest locations in the Black Mallard River, nest numbers

are in bold, and inset showing river location (not to scale).

7m

«n1

10m

10

V, (8
V7

<®<®.
<69,

~200 m

 

25m

 

 

 

 

F low

25

 

 

 

‘V’ Black Mallard

 

 

® Nest
V Drift Net

 

 

 

 

Figure 3: Average daily stream temperatures in the Black Mallard River (A) and the
Trout River (B) during the spawning and emergence period and relative water level
during the emergence period. (*denotes late spawnings)

 

 

 

 

 

 

 

 

 

 

 

A.

a 26 - e . - - f ‘40
o .
V ‘35
a)
5 24’ I
g l A ‘30 E
8. 22» FEmergence M 1% 325 3':
i9 \r’v \l 3 .9
r. 20» . :20 cu
.9: * 3
m i t a 1
; Spawnin :15 a):

18» , .g
2‘ a l co
'5 . ' , 10 T)
D 16- l l i m
. l 1 .5
C) l t
9 5 l 1 1
a; 14 - - - ‘ o
4 5/14/00 6/3/00 6/23/00 7/13/00

Date
B.
5 26 4 ~ - - - - - fi 20
o ' 1
9 24' I
3 ‘16
m —
a 22» s
s l 3
o 20 ‘12 '03
l: Emergence . *5
§ 18- ' ‘ 38 i
3 .2
i; 16' . l l . I?
s .- 3...... . I l i , .. a:
a) 14' ' " ‘ . , ‘
r J ‘ I l !
g 12 . - - l l ' o
< 5/14/00 6/3/00 6/23/00 7/13/00
Date

26

Figure 4: Box plot of the percent of emergent prolarvae collected during set 1 (2100-2400
h), set 2 (2400-0300 h), and set 3 (after 0300 h). The whiskers Show the minimum and
maximum values, the box depicts the 25 and 75 % quartiles, and the dash the median.

80. . T
T .

 

 

 

 

 

 

 

 

 

 

 

20»_1_ i
if

0 . - , If!

4A

Set 1 Set 2 Set 3 Set 1 Set 2 Set 3
Black Mallard River Trout River

 

 

 

Percent emerged
A
9
—+
F—

 

 

 

 

 

 

 

 

27

Figure 5: The number of prolarvae collected each day by nest for the Black Mallard River
and the average daily stream temperatures and relative water levels during the emergence
period.

 

 

 

2 24 32
a. 28
“ 20 24
§ 20%
O 16 u.
l" 8
E 16 12 g
3 Rain 8
(3,3, 12‘= 4

l .0

 

 

Nests
n=164

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Neat 7
n = 109
O .
g
3 5° Nest 3 350 N0“ 8
H n 8 768
o. 40 n = 201 300
“5 250
o 30 20°
53 10 50
g of o
it: 180 200 .
:3: "m4 160 Neat 9
n a 535 n = 666
120 120
100
so so
so
40 40
20
0 o.

 

 

 

 

 

 

 

 

 

 

33:35::ggaa 325,333.23: ,§;§S, 3335333333:
EEEZEESEZEigézfied’éE ZZEZSZSEXEEQ‘zé'ngz-‘i
PPPPFNNNN FPFPPNNNN

Abundance of Prolarvae

Figure 6: The number of prolarvae collected each day by nest for the Trout River and the
average daily stream temperatures and relative water levels during the emergence period.

5326 20
l 16
123
a i
4 3
0

 

 

Stream Temperature (
I»: a: a; 8 E '33

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
   

 

 

 

 

 

 

 

 

 

 

12
160 6000
160 5000 Name
140 .
120 n M346 4000 n814,632
100 3000
80
60 2000
‘° 1000
20
0 0‘
12000 2000
10000 m2 1600 No.12”
n=14,880 "=3.
300° 1200
6000
600
4000
2000 . ‘°°
o 0
6000 1600 N.“
4000 No.14 n= 3616
n=12.015 12°°
3000
600
2000
1000 400
0‘ o _
240 10000
200 NW5 8000 Nestto
"“0“ n= 20.713
160
6000
120
80 4000
40 2000
°§§§§§§ g‘é i”: 3%}: °§§s§§§53= 5' =53
3 7 .. .1, ' .. 7 -: 7 '1 7 7
663363 ~ -- aaggggvafi ~ 3:3
Date

Figure 7: Scatterplot of the number of prolarvae produced vs. the number of spawners
seen on nests in the Trout River.

 

 

 

 

24000l
200001 . 3
w 3
$16000. 1
E O .
912000’ O
E . o .
o I
s 8000: a
E :
2 .
4°00 0
0 . O .
0 1 2 3 4 6

Number of Spawners

30

 

7.96 mn

Average

 

 

 

601

00
4|

130 - ..
120'
110'
1001

5:039“.

Figure 8: Distribution of lengths for emerging prolarvae for each stream, the Black

Mallard River (A) and the Trout River (B).

N.0_.
0.N_.
¢.N_.
0N?
0. Z.
N. E.
0.0?
#0..
0.0..
0.0
N0
0.0
v.0
0.0
0.\.
. NS
0.0

.,v.0

0.0
0.0
N.0
0.».
vé
, 0.?

Length in mm

 

8.65 rm 1

Average

 

7" V V T —v v v

 

 

5:039”.

N.0_,
0.N_.
v.N_.
0.N_.
0. S.
N. E.
0.0..

. v.09

0.0V
0.0
Na
0.0
V0
0.0
0.\.
NS

- 0.0

«.0
0.0
0.0
N.0
0.v
«xv
0.?

Length in mm

31

 

 

 

7/3/00 7/8/00 7/1 3l00
Date

6/28l00

6/23/00

 

 

 

 

 

Figure 9: Scatterplot of lengths for emerging prolarvae over time in the Black Mallard

River (A) and in the Trout River (B).

w
.W o 090 800 o
E 1
l
7
oo 093500 000 0 000 o
.o O OE . . o oo 8
0 00000 Eooo
0 000858880
o 0085998 0
o o 8000005588 0 0 E
0 880089000 3 go
o BEBBO I
80 o 7 oo
gee
o 90889
O no 009000000
8 ooBO
EEO
egg
0 o 0 .3 90000000000
m m GOES
0 3900956 o .m o99
ES 6 0 GE
0 o 9
o E EDS
E
00089809 0 9o
oooooo
OOBEBO 0
003550
88890 M
.3 oEHBBBO
1
W o 86o0
0000 o o
3080 8090 o
ooBOOUOo o
oeo
O
-r - r- - . rl-ll m - - - - - - 1-1
2 1 O 9 8 7 6 5 4, 4 3 2 1 O 9 8 7 6 5
1 1 1 6 1 4| 4| 1 1
EE 5 5000.. EE 5 £000..

32

6/1 8/00

Figure 10: Boxplot of cumulative degree days for the peak of prolarval emergence for
each study component, the line is the mean, the box represents the standard deviation,
and the whiskers are the minimum and maximum values.

 

500 a +
460 - T
420 1

380 *

 

 

340’

 

 

 

 

COD
1

300 l E __
zeo _ + E

220 -

 

 

 

 

 

 

 

 

 

 

 

 

180 A - -
Trout River Applegate's Data Piavis's Lab Study

Black Mallard River My Lab Study
Study Component

33

Chapter 2:
Dispersal 0f age-0 sea lamprey during the first growing season.

Introduction

Dispersal of larval fish away from spawning habitat seems likely to be a critical process
in determining the habitat conditions and level of intra-specific competition (Howard,
1960) experienced by young-of-the-year (age-0) fish and thus their survival rates. If
dispersal is passive, age-0 fish may be found in sub-optimal habitats and may clump
together (Robinson et al., 1998). If dispersal is active, individuals may become more
evenly spaced (Robinson et al., 1998) or may continue dispersing to find optimal habitats.
Competition may be a driving factor in determining recruitment if dispersal is limited.
Yet, physical transport processes may be more important if dispersal is wide or unlimited
(Economou, 1991). Of course a combination of these may be more likely. However, as
dispersal distances increase so does the risk of not finding suitable habitat (Economou,
1991), becoming prey, and depleting energy reserves (Kamler, 1992). Understanding the

extent to which fish disperse can provide valuable insight into recruitment processes.
Hardisty (1961) and Hansen and Hayne (1962) state that sea lamprey dispersal from nests

during their first growing season is a very vulnerable time, but that little is known about

this life stage. Densities are generally at a maximum at this stage and fish are at their

34

smallest and most vulnerable size. Manion and McLain (1971) report that ammocoetes
remain close to spawning areas and then gradually scatter. Beamish and Lowartz (1996)
mention that age-0 ammocoetes move to nursery habitats after emerging from nests and
then within a few months ‘join the remainder of the population’. It is suggested that
dispersal is largely dependent on heavy rainfalls (Hardisty, 1961), yet no study has
looked Specifically at this process in age-0 sea lamprey and no quantitative estimates for

distances moved have been reported.

Densities of age-0 ammocoetes have also not been documented, despite routine stream
assessments by the Sea Lamprey Control Program. Because of their small size, age-0
ammocoetes are not collected or reported. High densities decrease growth rates for age-1
and older ammocoetes in the laboratory (Morman, 1987; Murdoch et al., 1992) and in
the field (Weise and Pajos, 1998), but the effect of density on growth rates of age-0

ammocoetes is unknown.

Because age-0 sea larnpreys are not often sampled, methods and sampling designs are not
documented. I expected age-0 ammocoetes to be difficult to find and clumped when
found. Adaptive sampling is an irmovative approach to compensate for the potentially
inefficient sampling of rare and clustered populations (Thompson, 1992; Thompson and
Seber, 1996). Adaptive sampling allows for the flexibility to increase sampling intensity

in areas that have the object of interest, in this case age-0 ammocoetes.

35

The objectives for this study component were to: (1) document densities of age-0
ammocoetes; (2) determine if age-0 ammocoetes cluster close to nests of origin or if they
disperse widely; (3) look for evidence in the field of density affects on growth of age-0
ammocoetes; and (4) determine the efficiency of adaptive sampling for estimating

densities of age-0 ammocoetes.

Methods

Study design

To address these objectives, I needed to be able to associate ammocoetes collected in the
stream with a specific source nest. This could be accomplished by restricting nests to a
few widely separated locations or by marking prolarvae to uniquely identify their nest of
origin. I tried to restrict spawning areas by installing cages in streams and introducing
adults into them. However, this was unsuccessful, adults escaped and no nesting
occurred within these cages. Tagging prolarvae is not practical because of their size (4.5
- 12 mm). Therefore, I chose to introduce small numbers of adults above barriers in two
streams which otherwise were not accessible to sea lamprey. I then identified nests that
were isolated and conducted surveys to estimate age-0 densities at a range of distances
downstream of nests. To finther enhance my ability to associate ammocoetes with source
nests I also collected DNA samples from adults and larvae and used microsatellite

methods to assign larvae to parents and determine siblings.

36

Study streams

Ogemaw Creek is a tributary to the Rifle River in Ogemaw County, Michigan (Figure 1).
The stream has a barrier at the mouth that sea larnpreys cannot traverse. The stream is
approximately 6 m wide and 3 km long. The Carp River is a tributary to Lake Superior
and is located near Pancake Bay in Ontario, Canada (Figure l). The Carp River has a
mean width of approximately 10 m and a length of 8 km. It too has a barrier to sea

larnpreys. Twelve pairs of adults were stocked into each stream above the barrier.

Field methods

No previous studies have investigated age-0 ammocoete movements; therefore I
conducted a preliminary study to explore how far downstream of nests ammocoetes
might disperse. In Ogemaw Creek in 1998, I located nests and electrofished 32 randomly
selected transects. These data suggested that age-0 ammocoetes seemed to be within 200
m of nesting areas. Using these preliminary data, I created regions or zones below nests
to determine if ammocoetes clump near nests or if they disperse widely. 1 established
three sampling zones (Figure 2) downstream of two nests in each stream. The first
sampling zone was considered near the nest and started at the nest and extended
downstream 50 m. The next zone extended from 55 to 155 m downstream of a nest and
was called the middle zone. The third zone included habitats between 160 m and 310 m
downstream from the nest, and was called the far zone. A 5 m buffer between each zone
was not sampled. Catches of ammocoetes were compared between zones to determine

dispersal. If ammocoetes were principally collected in the near zone then they were

37

considered to be limited dispersers. If ammocoetes were commonly found in the far

zone, they were considered widely dispersed.

To determine successful deposition and embryological development stage (Piavis, 1961),
a sample of eggs was collected fiom each nest and preserved in 70% ethanol. The
developmental stage was used to determine the timing of the first sampling event, to be

sure that ammocoetes had left nests.

A small backpack gold-mining dredge (Keene Equipment, CO) with a 4 cm diameter
suction hose Was used to collect age-0 ammocoetes. Sample plots were 25 x 25 cm2 and
were delineated using a plexi-glass frame with solid sides and an open top and bottom, to
prevent escape of ammocoetes during dredging. Bottom sediments were excavated
using the dredge to a depth of approximately 6 cm and collected in a bucket with a 590
um screen bottom. Material that did not pass through the screen was hand-picked in the
field to collect ammocoetes. All age-0 ammocoetes were preserved in 70% ethanol, later
identified to species, and measured. Sampling occurred in Ogemaw Creek on July 20 —
22 (event 1), August 12 — 13 (event 2), and September 5 — 6 (event 3). Sampling
occurred in the Carp River on August 2 — 5 (event 1), August 23 — 24 (event 2), and

September 24 — 25 (event 3).

Initially, six random plots were sampled within each zone below nests. I conducted a
power analysis with data collected using the same dredge technique (L. O'Connor and J.
Kelso, Department of Fisheries and Oceans, Ontario, Canada, unpublished data). I

calculated that six samples would provide sufficient power (01=0.05 and B=0.3; Merritt

38

 

and Cummins, 1996) to detect a difference of twelve ammocoetes per In2 between zones.
Random sample plots were located by generating random transects within each zone and
then sampling the closest preferred larval habitat or type I habitat, which consists of a

mixture of sand, silt, and detritus, to the transect.

Following the adaptive sampling approach, if an age-O ammocoetes was collected in a
random plot, additional neighborhood plots were sampled. The neighborhood samples in
this study consisted of one 25 x 25 cm2 plot upstream and downstream of the random plot
(Figure 3). During the first sampling event, neighborhood plots were sampled until no
age-0 ammocoetes were collected; hence the number of neighborhood plots could consist
of more than two additional plots. This was changed to only one neighborhood plot
upstream and downstream during the second and third sampling event because of time

constraints.

Microsatellite methods

Microsatellite loci are ‘nuclear DNA with short repeated core sequences scattered
throughout the nuclear genome’ (Wirgin and Waldman, 1994). These sequences are
relatively prevalent in the genome and can be inherited, but do not seem to be used in
gene expression, that is they do not code for specific functions or protein products
(Scribner and Pearces, 2000). Because these sequences are not used in gene expression,
they do not appear to be under selective pressures and so generally have more variation
than gene expression sequences (Wirgin and Waldman, 1994). The number of different

alleles that is seen at a locus across individuals characterizes this variation. By using

39

multiple loci, relatedness can be assessed using likelihood estimates of parent and

offspring genotypes.

Fin clips were taken from all adults stocked into Ogemaw Creek and preserved in tissue
buffer. DNA was extracted from tissue samples of fin clips and age-0 ammocoetes using
Puregene® DNA extraction kits (Gentra Systems, Inc.). Additional ammocoetes were
collected for this analysis. Once extracted, sample concentrations were obtained using
flourometry and then diluted to 20 ng/ml. Polymerase chain reactions (PCR) were used
to amplify targeted microsatellite loci within the DNA. Primers used in the PCR include:
SLGA210F / Kim210R, F GT3, Sp1120, SLGA38F / 3SLGA38R, GISES, and GISB15.
Primers are the different reagent mixtures to cut and amplify. specific loci. These
flourescently labeled PCR products were run onto 6% denaturing polyacrylamide gels
and scanned using a FM B10 11 (Hitachi, Inc.) scanner to produce a picture of the alleles
associated with each loci. Each gel was run with not only my samples but also a ladder
marking sizes down the gel and individuals of known genotypes. From the gel pictures,

alleles were scored, thereby genotyping each individual.

Data analysis
Field data were analyzed using parametric statistical methods in Statistica (Statsoft, Inc.
1998). I explored the following questions: 1) are densities of age-0 ammocoetes different

between zones and 2) does density affect age-0 ammocoete length.

40

In adaptive sampling the mean of a zone (13) is calculated from (Thompson and Seber,

page 98, 1996):

where A,- denotes the network which includes random plot i,

m is the number of plots within network A.-, not including edge plots,

Yj is the number of ammocoetes in sample j, and

n is the number of initial random sample plots.
A network is a cluster of plots that includes the random and neighborhood sample plots.
An edge plot is a neighborhood sample plot where the condition was not met, i.e. no age-
0 ammocoetes were collected. If the random sample plot contains no ammocoetes then
m is equal to 1, m is also equal to 1 if the neighborhood sample plots do not contain any

age-0 ammocoetes.

The variance for a zone is calculated from (Thompson, p. 271, 1992):
.. (N— n)
varuz)=——— 1)le-

where N is the total number of possible sample plots within a zone. N was calculated by

multiplying the area of a zone by an estimate of the percent of type I habitat in the

41

stream, which is estimated from habitat transect data collected during larval assessments,

and then dividing by 0.0625 m2, which is the size of a single sample plot.

The microsatellite data were analyzed using CERVUS 2.0 (Marshall et al., University of
Edinburgh, 1998-2001), a likelihood-based parentage analysis program, to associate
ammocoetes with parents thereby determining siblings. In addition to determining the
maximum distance separating siblings, I used the data to address two other questions.
First, I wanted to detemrine if different families occur below different nests. 1f
ammocoetes do not disperse widely one would expect to find ammocoetes of different
mothers or fathers below different nests. To explore this question I compared the
distributions of siblings found below the two nests. Second, I wanted to determine if
siblings clump in habitat patches (i.e. are siblings found more often in the same habitat
patch than would be expected by chance?) To test this, I used a randomization test
(Sokal and Rohlf, 1995) to determine if the total number of siblings observed 00-
occurring in habitat patches was significantly greater than what could occur randomly.
The observed number of sibling pairs in each habitat patch was counted and summed
over all habitat patches. Then all the siblings from all patches were randomly re-assigned
to habitat patches and the total number of sibling pairs was counted; this simulation was
repeated 1000 times. I compared the observed number of sibling pairs in my data to the
simulated random distribution to determine the probability of obtaining the observed

number of sibling pairs by chance.

42

To determine whether the adaptive cluster sampling (acs) is more efficient than a Simple
random sampling (srs) design for age-0 ammocoete populations, the ratio of the acs
variance and the srs variance was calculated. If the ratio is less than 1 then the acs was
more efficient, whereas if the ratio is greater than 1 the simple random sampling design
with sample size n* was more efficient. The variance of a simple random sample is

calculated as (Thompson, page 275, 1992):

f N \
;(yi“ra)2 (N-n")
var(srs) = n * -1 (Nn*)

 

 

 

where n“ is the number of plots sampled, in this case the number of random plots

sampled.

Results

In Ogemaw Creek, a total of 141 plots were sampled and 40 age-0 ammocoetes (range
per plot: 0 — 5) were collected. Ten age-0 ammocoetes were collected in Ogemaw Creek
downstream of nest 1, and 30 age-0 ammocoetes downstream of nest 2. In the Carp
River, 172 plots were sampled and 251 age-0 ammocoetes (range per plot: 0 — 23) were
collected. In the Carp River, 160 and 91 age-0 ammocoetes were collected downstream
of nest 1 and nest 2, respectively. Out of the 313 plots sampled in both streams, age-O
ammocoetes were found in only 74 (Figure 4). These data do not follow a Poisson
distribution (x2 = 181.86, df=2, p<0.05), suggesting that age-0 ammocoetes are not

distributed randomly but are aggregated.

43

Mean densities of age-0 ammocoetes varied between 0 and 44 per m2 (Table 1). There
were no significant differences in age-0 ammocoete densities between zones or between
events (Two-way ANOVA: Zones — F237=0.47, p=0.63; Event — F227=1.66, p=0.21;
interaction — F437=0.45, p=0.77) (Figure 5). Yet the far zone in the Carp River does on
average have higher densities than the other two zones and there appears to be a

downward trend by the last sample event (Figure 6).

Six loci were used to determine parentage in Ogemaw Creek, however the number of
alleles within each locus was small, ranging from 2 to 8 (Table 2). Expected
heterozygosity values ranged between 0.32 and 0.726, values less than 0.5 are generally
not useful for large-scale parentage analysis (Cervus 1998-2001). Locus SLGA38 had a
relatively high null allele fi'equency (>005). A null allele is an allele that can not always
be detected with the primer being used because of mutations in the binding sites such that
the allele is not amplified sufficiently (Cervus, 1998-2001). Despite the high null allele
frequency estimate, SLGA38 did not cause major problems with mismatches in

genotyping between adults and ammocoetes and hence was kept in the analysis.

Ammocoetes from the second and third sampling events were genotyped. The low allelic
diversity made it difficult to identify siblings with much confidence. Eight females
produced the 42 ammocoetes collected. Cervus (1998-2001) analyzes parents separately.
When males were used as the unknown parent, nine ammocoetes were attributed to a
father with greater than 60% confidence, while only one of those was attributed to a

father at a confidence of 80% or greater. When females were used as the unknown

parent, twelve ammocoetes were attributed to a mother with greater than 60% confidence,
and three of those were attributed to a mother at a confidence of 80% or greater. Because
ammocoetes were attributed to mothers with greater confidences than the fathers, mothers
were used in the analysis. In those cases where confidences were below 60%, the most-
likely mother was used. Notwithstanding the low confidence in sibling identification, in

the analysis presented below I assume that parental assignments are correct.

Siblings were found in the near zone of the upstream nest and the far zone of the
downstream nest, by the second sampling event siblings were 820 m apart and by the
third sampling event 874 m apart. There is no evidence that distinct groups of siblings

are associated with the two different nests, their distributions are similar (Figure 7).

The randomization test suggests that pairs of siblings are more likely to be found in the
same habitat patch than would be expected by chance. The randomization’s produced a
mean number of sibling pairs around 9 (Figure 8), while my data indicated 20 sibling
pairs. Only four out of 1000 simulations produced 20 or greater sibling pairs, suggesting
that ammocoetes appear to aggregate with siblings. However, it must be emphasized

again that these results are conditional on a low confidence of assignment of parentage.

Ogemaw Creek ammocoetes were on average larger than those in the Carp River. In
Ogemaw Creek the average length of ammocoetes during the second sampling event was
12.69 mm and 17.21 mm in the third sampling event (sample event 1 was not recorded).

In the Carp River, ammocoetes averaged 11.40, 13.54, and 15.59 mm in length during the

45

first, second, and third sampling event, respectively. There were significant length
differences between streams during the second sampling event (F 1.96=7.04, p<0.05) but
not during the last sampling event (F 1 ,21=3.32, p=0.08); however only 5 ammocoetes
were collected in the Carp River at this time. There were no length differences between
zones in either of the streams except during the second sampling in the Carp River, the
average length of ammocoetes in the near zone was 14.03 mm, in the middle zone was
13.15 mm, and in the far zone was 12.89 mm (F 2,6g=3.66, p<0.05). There also appears to

be no strong relationship between density and length within streams (Figure 9).

Although the data suggest that age-0 ammocoetes aggregate, the adaptive-cluster
sampling (acs) was generally not more efficient than a simple random sampling (srs)
design. Efficiency values ranged from 0.29 to 10.03 (Table 3); those values < 1 indicate
that the adaptive sampling design was more efficient. On seven occasions the acs did
better than the srs, however on 17 occasions they were the same. Essentially, a simple
random sampling design with 6 plots estimated the population density just as accurately

as an adaptive cluster sampling design.

Discussion

Fine-scale densities of age-0 ammocoetes can be very high (384 / m2), although average
densities were low, 4 and 10 ammocoetes per m2 in Ogemaw Creek and the Carp River,
respectively. Densities reported here are not much higher than those seen for older
ammocoetes. Other authors report densities of age-1 and older ammocoetes in the field

between 0 and 10 per m2 (Manion and McLain, 1971, Jones et al., in review), most

46

yearling densities range between 0 and 2 per m2 (Jones et al., in review). Given the large
numbers of ammocoetes expected to emerge fiom nests (range 87 4— 20,713 prolarvae per
nest, as reported in Chapter 1) and the relatively small amount of habitat, I expected high
densities to be prevalent. Yet these results suggest that age-0 ammocoetes are generally
found in low densities. I estimated the overall abundance of age-0 ammocoetes by
multiplying the average density below nests by the amount of larval habitat (type 1 =
fine, depositional sediments). In Ogemaw Creek, estimated abundance’s are relatively
low (Table 4). This could be due to low nesting success as seen in the Black Mallard
River in Chapter 1, or it could be due to early mortality of recently emerged larvae, wide
dispersal, gear efficiency, or a combination of all four. Abundance estimates in the Carp
River (Table 4) are relatively high during the first two sampling events and suggest high
nesting success as was seen in the Trout River in Chapter 1. However, by the last
sampling event, the abundance estimate drops sharply suggesting either mortality later in
the season or continuing dispersal. Due to the study design used, i.e. introducing low
numbers of spawning adults, the densities reported here may be lower than in naturally

occurring populations.

Both the field and microsatellite data suggest that age-0 ammocoetes do not clmnp near
nests but disperse widely. These results suggest that siblings can disperse at least 874 m
apart in the first growing season; this is a minimum estimate of dispersal distance because
my study area only extended 974 m downstream of the upstream nest. Manion and
McLain (1971) suggested that age-0 ammocoetes initially remained close to spawning

areas and then generally scattered. In their study, adults were introduced into stream

47

reaches that were separated by either falls or installed dams. The most downstream reach
did not receive any adults. Ammocoetes were not found in the most downstream reach
until the next spring. The reach above the most downstream reach was 1.3 km long,
suggesting that ammocoetes dispersed less than that during their first growing season.
Hardisty (1961) suggested that age-0 ammocoetes are dependent on heavy rains to help
them disperse; however I do not believe this to be the case and present evidence in

Chapter 3 that ammocoetes disperse readily early in the first growing season.

The microsatellite work supported the field data, however, most of the ammocoetes were
assigned to parents with confidences below 60% and it seems likely that fewer females
spawned. Only three nests were located in the study area and upstream of the study area
in Ogemaw Creek. One pair of adults was seen spawning on a nest and all nests were
small. Generally when there are multiple adults on a nest the downstream end of the nest
become less horse-shoe shaped and more ridge shaped, this was not the case in Ogemaw
Creek. Furthermore, O’Connor et al. (in review) suggest that it is likely that less than

50% of adults spawn.

The low polymorphism of the loci used in my analysis severely limited the use of these
data. Other studies successfully assigning parents to offspring use loci that range in the
number of alleles fiom 2 to 30 (Norris et al., 2000), more commonly 20-30 alleles
(Letcher and King, 1999). The more variable the loci, generally the easier it is to assign
parentage, especially in a small population such as the ones created in this study. Norris

et a1. (2000) assigned parentage using only 4 loci and correctly identified parents 94.3%

48

of the time. However, sea lamprey show similar variability to the endangered wood
stork, which have on average 2 alleles at a loci (Van Den Bussche et al., 1999). It is
possible that this low variability of alleles is due to the founder effect, where relatively
few adults created the population. Research is continuing to identify other sea lamprey
loci that exhibit greater polymorphism. If such loci can be found, this method would be
invaluable by allowing researchers to explore in more detail the population dynamics of

sea lamprey.

I found no evidence for the suggestion that young ammocoetes segregate from the rest of
the population (Bearnish and Lowartz, 1996). In the plots fiom which I collected age-0

ammocoetes, about half (31 of 74) also had older ammocoetes in them.

Ogemaw Creek ammocoetes were larger than those in the Carp River and also grew more
between sampling events. Age-O ammocoetes grew about 2 mm between sampling
events in the Carp River, whereas they grew about 4.5 mm in Ogemaw Creek. The Carp
River had much higher densities than were found in Ogemaw Creek. The scatter plot of
length by density (Figure 7) does show a slight downward trend as densities increase in
the Carp River. Increasing density affects growth rates of older ammocoetes (Murdoch et
al., 1992; Mallatt, 1983; Morman, 1987) over longer periods of time. However, these
differences are just as likely due to cooler water temperatures in the Carp River and

possibly a shorter growing season because of its more northerly location.

49

Ammocoete distributions are aggregated, yet the adaptive cluster sampling design did not
decrease the variance of the density estimates more so than a Simple random design. This
inefficiency of the adaptive design is likely due to age-0 ammocoetes not aggregating to
the extent expected. They were not found in high concentrations; in Ogemaw Creek the
maximum number collected in a plot was five and in the Carp River, a maximum of 23
were collected in a plot. Therefore, a simple random sampling design is sufficient for
estimating densities of age-0 ammocoetes. However, one advantage to the adaptive
cluster design is the increased number of animals collected (Thompson, 1992). The
neighborhood samples in the Carp River had a higher catch per unit effort (0 — 3.12) than,
did the random. sample plots (0.12 — 0.95); in Ogemaw Creek the catch per unit efforts
between the neighborhood and random plots were essentially the same. Hence, if

individuals are needed for other parameters of interest, this design can be very effective.

The dredge was an effective device to sample the early age-0 ammocoetes and generally
did not kill them. However, processing the dredge material was time consuming and the
technique is fairly destructive to the habitat. For these reasons, I do not suggest using this
method as a sole means for sampling age-0 ammocoetes. Age-0 ammocoetes are
susceptible to the traditional AbP-2 electrofishing method by late summer (mid to late
August). With fine mesh paddles, electrofishing is a relatively effective way of sampling
these small fish. Hence, I suggest using the dredge method to capture the early age-0
ammocoetes for intense research activities and the AbP-2 electrofisher for routine

assessments.

50

 

 

 

20.0 600000 0000000 00300 0000000288 0800. 00.0 “0000—
0000 ”300.80 0005 0300 03:06:30 1800. .000 02020 00 00080: 0002
0000.0- 3N0 S 0.0 000.0 N00 0 0m000
0000.0- 000.0 000.0 0N0.0 000.0 0 Emma
0000.0+ 000.0 000.0 0N0.0 N00 0 00.0qu
030.0. 00 0 .0 0N0 .0 000 .0 000.0 N 0N 0 Em
0N00.0- <0 0 .0 000.0 000 .0 000.0 0 00.00
00 3.0+ 000.0 «N0 .0 N000 000.0 N 002.000
00000000 020:0 00
26:0 :62 @000 €30 000 SE 60662 063

 

.0305— 00 000.80 000 0003 0:0 Czexm 00 0800 0.0 mm aNzoxm 0:030 8505— 000 000505 0300 0000202008 00 Cxoxm
50000008000: 00000008 05 fl Amvm 56003820: 0020000 05 mm ADE ”£03000 5 00m: 002 05000008008 00 088800 ”N 0300.

 

 

 

 

00.0 3.0 SN 0 0 .00 NNN— 00.0 0N0 _ 0N0 00.N “00003.0

0»: 0 0 00.0N 00.00 00.00 00.0 0 00.0 N

00.0 0 00.N >0.N 00.00 00.00 5.34 00.00 00.N 0 0.80

00.0 00.0 00.0 00.00 0 00.0 00.N 0Y0 00.N N

0 00.0. 00.N 00.N 0 00.0 00.N 0 0 0 308000

0 m < O m < O m < 802 500:0
0 50>m N 0:020 _ ~00>m

 

0:05 0:20:30 000 000: 00 00:00 £505 58 0000 0000000880 0000 00 0006000 502 ”0 0300.

51

 

 

we :3 $0 0
NR 00% 005. _ .030 000
20 $0 08 N
80 w _ m 02 _ 0020 308000
0002
0 0:030 N 0:0>m 0 0:030 800:0

 

.800: 3200 000000: 0:20:00 5 0000000880 0-000 00 0.0800550 0000:50m ”v 0300.

 

 

 

2: -- -- 8.0 80 8.0 8.0 -- 8.: N

80 -- 8.— 2: 8.0 8.2 03 m3 2: _ 000

8: 8.0 80 2: -- 8.0 of 8.: of N

-- one 8.: 2: -- 2: 8.: -- -- _ 308000

0 m < 0 m < 0 m < 002 6800
m u=o>m N ~=o>m — «GP/m

 

.0:0>0 0::0800 0:0 0:00 00 0030 0000 05 0:0 000000 306000 000 m020> 00:20am ”0 0300.

2
5

Figure 1: Locations of study streams.

 

53

Figure 2: Schematic showing sampling zones. The near zone extends from the nest to 50
m downstream, the middle zone extends fiom 55 to 155 m downstream fi‘om the nest, and
far zone extends from 160 to 310 m downstream from the nest. There is a 5 m buffer
between each zone.

 

 

 

 

 

2
2
Near
50 m
V
Middle
155 m 4— 5 m Buffer zone
Far
310 m

 

 

 

 

54

Figure 3: Schematic of adaptive sampling design. Figure A shows the six initial random
plots sampled, boxes in gray indicates that at least one age-O ammocoete was collected,
i.e. the condition was met. Figure B shows those same six random plots with the
additional sampled neighborhood plots in black.

A' Random sample B'
plot — no
animals

collected

 

[:1

Neighborhood
sample plo ~

[I D E] I

[j
/'

Random sample
plot - animals
collected

 

 

 

 

55

Figure 4: Histograms of ammocoetes collected in Ogemaw Creek (A) and the Carp River

(B)-

No of obs

No of obs

 

112’

888?

0081313838193

 

 

 

 

 

o 1 2 3 A 5 6
Number of age-O ammocoetes collected per plot

 

135
126
117

 

 

 

.012 3 4 5 6 7 8 910111213141516171819202122232425

Number of age-0 ammocoetes collected per plot

56

Figure 5: Box plots of the mean densities of age—0 ammocoetes per m2 by zone for
Ogemaw Creek (white boxes) and the Carp River (shaded boxes). The box represents the
25th and 75‘h quartile, the line the mediam and the whiskers the minmum and maximum.

 

 

 

 

50

N

E

‘0 4o .
D.

m

2

8 30 -
O

O

E

m 20

C.’

8,

C

m

0

2 0 %T - .

A B C
Zone

Figure 6: Box plots of the mean densities of age-0 ammocoetes per m2 by sample event
for Ogemaw Creek (white boxes) and the Carp River (shaded boxes). The box represents
the 25th and 75‘h quartile, the line the mediam and the whiskers the minmum and
maximum.

 

50

‘° iéT %;

0 &

1 2 3
Sample event

2

Mean age-O ammocoetes per m

 

 

 

57

Figure 7: Histogram of the number of ammocoetes associated with each male parent for
upstream most nest (A) and downstream most nest (B).

 

No of obs

 

 

 

4 5 6 7
Parent (Mom)

 

No of obs

 

 

 

 

5 6 r 12
Parent (Mom)

 

58

Figure 8: Histogram of values for randomly distributed nest-mate pairs, 1000
simulations. Asterisk denotes the number of nest-mate pairs in observed data.

 

168 »
156 »
1441
132 .
120 -

No of obs
2

 

12’ l I t 1

 

 

0‘
2 3 4 5 6 7 8 9101112131415161718192021222324
Number of sibling pairs

59

Figure 9: Scatterplot of average ammocoete lengths in mm in relation to the density of
the habitat patch in Ogemaw Creek (A) and the Carp River (B). The circles represent
ammocoetes during the first sampling event, the squares the second sampling event, and
the diamonds the third sampling event.

21 - - - - - - . -

 

19'
17)
15»

13'

Average Length in mm

11» D

 

 

 

10 20 30 40 50 60 70 80 90

18 v v 1" ' vyv 1 vvvvv '1 'v *

 

171
16i
15»
14»
13in DU
12}

Average Length in mm
0 O
O

11»o
0 O

 

 

 

0 50 100 150 200 250 300 350 400 450

Density of age-0 ammocoetes per m

60

Chapter 3:
The effects of density on the movements of sea lamprey larvae
in the laboratory.

Introduction

The relation between growth and density of sea lamprey ammocoetes is density
dependent in aquaria and cages (Mallatt, 1983; Morman, 1987; Murdoch et al., 1992).
However, field observations are inconclusive as to the presence or degree of density
effects on growth (Jones et al., 2001). These contrary conclusions may be due to the
limitations of the enclosed experiments as they do not allow for dispersal. Dispersal may
be an important mechanism in lessening the demographic effects of density. To date, no

studies have investigated ammocoete dispersal and its relationship to density.

Density is an important mechanism determining dispersal in both fish and aquatic insects
(Hume and Parkinson, 1987; Fonseca and Hart, 1996; Kerans et a1, 2000). Ammocoetes
are benthic suspension feeders and burrow into soft sediments, a behavior similar to
suspension feeding stream insects. Stocked steelhead fi'y dispersal was shown to
generally increase with high densities (Hume and Parkinson, 1987). Several insect taxa
show density-dependent drifi (stoneflies: Walton et al., 1977; black flies: Fonseca and
Hart, 1996; Gersabeck and Merritt, 1979; caddisfly: Kerans etal., 2000). Kerans et a1.

(2000) investigated how density, velocity, and substrate affected dispersal in the

61

caddisfly Hydropsyche slossonae and found it to be conditional on season. Dispersal was
density dependent in the spring, however in the fall dispersal increased but was density
independent. Understanding the degree to which ammocoetes disperse and the factors
affecting dispersal may link laboratory and field studies on growth, and provide managers

with a better understanding of sea lamprey population dynamics and behavior.

I examined the following questions: 1) do ammocoetes move / disperse at any density; 2)
does density of ammocoetes affect their likelihood of moving; and 3) are there seasonal
differences in movement. I conducted trials using age-l and older ammocoetes (age 1+)
but concentrated on age-O ammocoetes. Densities are highest during the first year of life
and hence may be subjected to greater density effects than later stages. In addition, this

early life stage often determines population size in fishes (Trippel and Chambers, 1997).

Experimental Design:

Dispersal may be hard to discern from mortality (Le Cren, 1973). Studies in natural
systems often can only report the difference in numbers of animals from one time period
to the next and generally label this as mortality. Tagging methods can be used to monitor
dispersal in the field, however age-O ammocoetes are too small to use these techniques.
In laboratory aquarium experiments, only mortality can be assessed due to the

constrained environment.

Therefore, an experimental design was required to monitor movements of individual

animals or small groups of animals and to be able to manipulate densities. Because of the

62

difficulties of conducting this in the field, I designed a laboratory study (adapted from
F onseca and Hart, 1996) that:

I. allowed for movement of ammocoetes,

2. simulated naturalystream conditions,

3. and allowed easy monitoring of movement by creating source areas with

ammocoetes and sink areas without.

Ammocoetes were placed in replicate artificial streams at three randomly assigned
densities: 5, 15 , and 30 per experimental unit, equivalent to 80, 240, and 480 ammocoetes
per m2 respectively. The two lower densities are similar to those used in previously
published growth and density studies (Mallatt, 1983; Morman, 1987; Murdoch et al.,
1992). I conducted thirty replicates for each density of age-O ammocoetes between
August and September 2000. These two time periods were chosen to examine summer
and fall seasonal effects. Fifteen replicates were completed for each density of age-1+
ammocoetes fiom July to September. Fewer trials using age-1+ ammocoetes were
conducted because of the emphasis on age-O ammocoetes, and hence seasonal effects

were not tested.

Experimental methods:

Age-0 ammocoete trials:
Age-0 ammocoetes (8 — 26 mm in length) were collected using drift nets (350 um mesh,
for 3 h sets) and an electrofisher (AbP2-backpack electrofisher, University of Wisconsin)

in the Trout River, Presque Isle County, a tributary to Lake Huron, in northern Michigan.

63

Approximately 75 age-O ammocoetes (1 1—26 mm in length) were held in each of two 38
L (IO-gal) aquaria or holding tanks, filled with sand to a depth of 6 cm. These holding
tanks were kept in a water bath at a constant temperature of 18°C. The aerated water in
the holding tanks was replaced every other day with filtered Lake Huron water. Every
other day, ammocoetes were fed suspended sediments collected fiom depositional stream

habitats that were sieved through a 125 um mesh screen.

Three oval raceway tanks (F rigid Units, Inc., Toledo, Ohio) were used to create stream-
like conditions. 'The raceways were 2.9 m in length, 40 cm wide, and the overall unit
width was 1.6 m (Figure 1). A near-shore pump circulated unfiltered Lake Huron water
to produce low flow conditions (~15 cm/sec) in the raceways similar to field conditions.
The water was 17 cm deep and water temperature was recorded daily. Natural lighting
was not controlled, but was provided from windows around tanks. Four plastic trays
(25.5 x 28.5 x 9 cm) filled with 2 cm of beach sand were placed in each tank to provide
habitat for ammocoetes. One tray served as a source tray and the other three trays served
as sinks to collect dispersing ammocoetes. The sand was replaced with dry sand before
each trial. The trays were positioned 50-60 cm apart, with the first (source) tray 1 m

away from the water input (Figure l).

Ammocoetes were introduced into the source tray using an open top and bottom plexis-
glass introduction box (20 cm wide x 20 cm long x 22 cm tall), to ensure that all
ammocoetes burrowed into the sand before each trial started. The introduction box was

sunk into the sand and was only removed afier all ammocoetes had burrowed. Trials

1

began in late afternoon ( 1400 — 1700 h) and ended the next morning (0800 — 1200 h).
Preliminary observations indicated that ammocoetes moved only at night and that
movement tended to occur after a single night. Contents of trays were hand-sieved after
each trial to find ammocoetes. The number of ammocoetes found in each tray was
recorded. Ammocoetes were then returned to the holding aquarium and not reused until

2 days later to reduce stress on the animals.

Age 1 and older ammocoete trials:

Age-1+ ammocoetes (range 33 — 116 mm) were obtained from the US Geological
Survey’s Hammond Bay Biological Station. These ammocoetes were fed a yeast mixture
every other day, which is standard practice for raising age-1+ ammocoetes. Age-0
ammocoetes were not feed yeast because previous studies had shown that they did not
feed well on the yeast (W. Swink, USGS, Hammond Bay Biological Station, personal
communication). Methods for trials were the same as that described for age-0

ammocoetes (see above).

Data analysis:

The fraction of the initial population that were in each tray was computed. Attempts to
normalize these data using an arcsine square root transformation (N eter et al., 1996) were
unsuccessful. Hence, I analyzed the data by comparing the fi'action in the source and
sink trays among densities and season, using a Kruskal-Wallis non-parametric test
(Gibbons, 1993; Statistica, StatSofi, Inc., 1998), unless otherwise noted. I tested the null

hypotheses that: (1) the fraction of ammocoetes that moved out of the source tray was the

65

same for each density and season and (2) the fraction of ammocoetes that burrowed into
the three sink trays were the same for each density and season. P values less than <0.05

were considered significant.

Results

Age-0:

In August, age-0 ammocoete movement out of the source tray was not different between
the density trials (p>O.8, n=33) (Figure 2). The average percent of ammocoetes that
moved out of the source tray (:SE) was 27.3 i 7.8 % in the low-density trials, 20.6 :1; 4.8
% in the medium-density trials, and 26.4 i 5.4 % in the high-density trials. Of the
ammocoetes that moved, 83.1% were not found in any sink tray and assumed to have
been lost down the drain. In the low density trials, ammocoetes were found only in the
first sink tray. In the medium and high density trials, ammocoetes were found in all sink
trays but in decreasing numbers with increasing distance away from the source tray. The

numbers of ammocoetes found in sink trays were not significantly different from each

other within each density treatment (p>0.3, n=99) (Table 1).

In September, age-O ammocoete movement among the density trials (Figure 2) was
significantly different (p=0.03, n=57). The average percent of ammocoetes that moved
out of the source tray (:SE) was 4.2 i 1.9 % in low-density trials, 6.3 i 2.5 % in

medium-density trials, and 10.7 i 2.6 % in high-density trials. Of the ammocoetes that

moved in September, 30.1% were missing and assumed to have been lost down the drain.

The distribution of ammocoetes in the sink trays was similar to that seen during the

66

ll .

August trials. There were no significant differences (p>0.3, n=17l) in the number of

ammocoetes found in sink trays in the different density trials (Table 1).

Movement was generally less in September than in August (Figure 2), moreover, there
was a general decline in movement in all densities as temperatures decreased (Figure 3).
Of the ammocoetes that moved from the source tray, a significantly larger proportion
were found in sink trays in September than in August (Mann-Whitney u test, p < 0.0001,
September n=5 7, August n=33). Ammocoetes were rarely found swimming fieely in the

raceways at the end of a trial.

Age-1 +:

On average, 20 % of age-1+ ammocoetes moved regardless of density (p>0.9, n=45).
Age-O movements in August were not significantly different than the age-1+ movements
(Mann-Whitney U test, p=0.24), however age-O movements in September were
significantly different (Mann-Whitney U test, p<0.05). Again, there was a general
decline in movement in all densities as temperatures decreased (Figure 3). Of the age-1+
ammocoetes that moved, 6.8 % were missing, this was significantly less than in the age-O
August trials (Mann-Whitney U test, p<0.01), however it was not different than the

number missing in the age-O September trials (Mann-Whitney U test, p=0.10).

Discussion

Approximately a quarter (20 — 25 %) of ammocoetes moved out of the source tray in July

and August, regardless of density. Ammocoete movements in all densities were greater

67

when temperatures were warmer, suggesting that density-independent dispersal
predominates, regardless of age. Other fishes also have high drift or movement rates
soon after emergence (Atlantic sahnon, white sucker, and sea lamprey: Johnston, 1997;
multiple species: Copp and Cellot, 1988; Brown and Armstrong, 1985). These results
suggest a possible mechanism for the wide dispersal observed in chapter 2. The dispersal

process may assist in moving ammocoetes away from nests to discourage clumping.

My data suggest that dispersal differs with water temperatures and possibly season.
Because temperature and lighting were not controlled in the experiments, temperature is
presumed to be an indicator for seasonal changes, although other factors, such as day
length and barometric pressure, may confound these results. Kerans et al. (2000) found
that the caddisfly larvae Hydropsyche slossonae (fifth instar) exhibited density-
independent dispersal in the fall, and density-dependent dispersal in the spring. This
caddisfly pupates in late spring and being heaviest at this time do not disperse as
frequently. Age-0 ammocoetes showed the opposite pattern in that, dispersal seemed to
be controlled by density-independent processes in warm temperatures (e. g. summer) and
density-dependent processes in cooler temperatures (e. g. fall). Observations of age-1+
ammocoetes in warm temperatures were similar. Similar to H. slossonae, ammocoete
movements were greater when density-independent dispersal predominated. Life cycle
characteristics may also explain ammocoetes tendency for density-dependent dispersal in
cooler temperatures (e.g. September). Early winter is suggested to be a critical time for
fish as they seek to build up energy reserves to assist in overwinter survival at a time

when food supply is limited. Yap and Bowen (2001, in review) reported that the northern

68

brook lampreys available food and assimilation efficiency declined from May through
August, but increased briefly in September. Cunjak (1988) examined brown and brook
trout energy stores over winter and found that in early winter (November — December),
both lipid and serum protein levels decreased rapidly, even though food supply was
sufficient. Gardiner and Geddes (1980) reported similar results in Atlantic salmon fi'om
October to December. The amount of fat or energy stored in tissues before the early
winter period may determine the overwinter survival of fish. These observations suggest

that when food is more limiting, ammocoete dispersal may be more sensitive to density.

Body size is a major factor in determining overwinter survival, especially for young—of-
the-year fish because they are at their smallest size (Shuter and Post, 1990). Because sea
lampreys hatch in mid- to late-summer and grow relatively slowly, the fall may be a
crucial time for ammocoetes to feed. As previously mentioned, growth is affected by
density when ammocoetes (age-1+) are confined to tanks or cages (Mallatt, 1983;
Morman, 1987; Murdoch et al., 1992). Mallett (1983) found that even at high food
concentrations, increasing density inhibited ammocoete growth, and suggested that
ammocoetes might release “some growth-inhibiting compound into the surrounding
sediments”. Bowen (personal communication) proposes that ammocoetes’ assimilation
efficiency decreases with increasing density, causing a decline in growth rates. The
density-related-dispersal behavior shown in my experiments may increase or enhance
survival by allowing ammocoetes to more efficiently feed before winter by finding lower
densities of ammocoetes. Such dispersal may also provide decreased interactions

between ammocoetes during winter; these interactions could increase activity levels and

69

consequently deplete energy reserves faster, as shown in pumpkinseeds (Bernard and

Fox, 1997).

The average densities of age-0 ammocoetes reported in Chapter 2 (5 to 10 per m2) are
low compared to the low density used in these experiments (80 per m2). Densities were
reported as high as 384 per m2 (Chapter 2), but infrequently. It is possible that the
process of density-independent dispersal early in the season and ammocoetes ability to
disperse widely, allow individuals to establish low densities such that density-dependent
dispersal is not very prevalent. The densities reported in Chapter 2 from streams where
the spawning population size was kept very low may be much lower than is seen in

streams open to adult migrations, however.

A limitation of my experiments is the use of sub-optimal habitat (sand) in the source and
sink trays. It is possible that larvae may be able to withstand higher densities in more
optimal habitats (type 1 = fine, depositional sediments). However, because of the short
time period (1 night) over which I saw ammocoete movement and Mallett’s (1983)
observations of decreased growth even with high food concentrations, it seems unlikely
that more optimal habitat would have had any major effect to change these results. In
addition, the results of this study suggest that the ability to gain body mass, and possibly
the accumulation of lipids, as water temperatures decline may be an important factor in
overwinter survival. I proposed that larvae move more at high densities because they do
not grow freely, and furthermore suggest that this growth and hence accumulation of fats,

i.e. lipids, determines survival. If proven, percent lipids in age-0 ammocoetes in the fall

70

could provide a key variable in determining recruitment for that year-class. Further work

is needed to test this hypothesis.

71

Table 1: Number of ammocoetes that moved out of source trays and where they settled
for each density and month.

 

 

 

 

 

Destination
Month Density Number Total Number Sink Sink Sink Missing

of trials individuals moved tray 1 tray 2 tray 3

used

Low 11 55 15 l 0 0 14
August Medium 1 l 165 34 3 0 2 29
Hi 11 330 87 7 5 5 70
Low 19 95 4 l 0 0 3
September Medium 19 285 l 8 5 3 3 7
High 19 570 61 18 16 12 15
Summary Low 30 150 19 2 0 0 17
(August + Medium 30 450 52 8 3 5 36
September) lgh 30 900 148 25 21 17 85

 

72

Figure 1: Diagram of raceway tanks and layout of the source and sink trays.

 

 

2.9 m
<¥ 5
Standpipe
(drain)

 

 

D 1.6m 40cm

 

 

 

Wat input

Source tray El Sink trays (1-3)

 

 

 

Figure 2: Average percent of ammocoetes that moved out of the source tray in August
(n=1 1) and September (n=19) for each density. Error bars represent :1: 1 standard error.

 

40 - lAugust ISeptember
36 -
32 -
28 -
24 -
20 -
16 -
12 -

8 _

 

 

 

 

low medium high
Density J

 

73

Figure 3: Percent of ammocoetes that did not move out of source tray by average daily
water temperature for all densities and ages.

Percent of ammocoetes
that stayed in source plot

 

 

 

 

100’ I I~O e. IIoo I0 I .
II I I .
o oo o o «no coo I
90’ - I I «
09 IE] I i
I n
so. .. oo ‘
II n j
I o o ,
70» El I .
o I .
I: I .
60’ I ' OI co .
I i
50’ I
40~ .
30 * * ‘ 4 ‘
13 15 17 19 21 23

Average daily water temperature (C °)

74

25

0.06

DENSITY: 5
AGE: 0

DENSITY: 5
AGE: 1+
DENSITY: 15
AGE: 0
DENSITY: 15
AGE: 1+
DENSITY: 30
AGE: 0

DENSITY: 30
AGE: 1+

References:

Applegate, V.C. 1950. Natural history of the sea lamprey (Petromyzon marinas) in
Michigan. US. Fish and Wildlife Service Special Scientific Report Fisheries No.
55: 237 p.

Beamish, F.W.H. and Lowartz, S. 1996. Larval habitat of American brook lamprey.
Canadian Journal of Fisheries and Aquatic Sciences, 53: 693-700.

Bennett, R.M. and Ross, R.M. 1995. Stream drift of newly hatched sea lampreys in the
upper Delaware River. Journal of the Pennsylvania Academy of Science, 69(1):
7-9.

Benoit, H.P. and Pepin, P. 1999. Interaction of rearing temperature and maternal
influence on egg development rates and larval size at hatch in yellowtail flounder
(Pleuronectesfenugineus). Canadian Journal of Fisheries and Aquatic Sciences,
56: 785-794.

Bernard, G. and Fox, MG. 1997. Effects of body size and population density on
overwinter survival of age-0 pumpkinseeds. North American Journal of Fisheries
Management, 17: 581-590.

Brown, A.V. and Armstrong, ML. 1985. Propensity to drift downstream among various
species of fish. Journal of Freshwater Ecology, 3(1): 3-17.

Casselrnan, J .M. 1995. Survival and development of lake trout eggs and fly in eastern
Lake Ontario -- in situ incubation, Yorkshire Bar, 1989-1993. Journal of Great
Lakes Research, 21 (Supplement 1): 384-399.

Copp, G.H. and Cellot, B. 1988. Drift of embryonic and larval fishes, especially
Lepomis gibbosus (L.), in the Upper Rhone River. Journal of Freshwater
Ecology, 4(4): 419-424.

Cunjak, RA. 1988. Physiological consequences of overwintering in streams: the cost of
acclimatization? Canadian Journal of Fisheries and Aquatic Sciences, 45: 443-
452.

Cushing, DH. 1996. Towards a science of recruitment in fish populations. Ecology
Institute, Germany.

Diana, J .S. 1995. Biology and ecology of fishes. Cooper Publishing Group LLC,
Carmel, IN.

Economou, A.N. 1991. Is dispersal of fish eggs, embryos and larvae an insurance
against density dependence? Environmental Biology of Fishes, 31: 313-321.

75

Elliott, J .M. 1984. Numerical changes and population regulation in young migratory
trout Salmo trutta in a Lake District stream, 1966-83. Journal of Animal Ecology,
53: 327-350.

Elliott, J.M. and Hurley, M.A. 1998. Predicting fluctuations in the size of newly
emerged sea-trout fry in a Lake District stream. Journal of Fish Biology, 53:
1 120-1 133.

Field-Dodgson, MS. 1988. Size characteristics and diet of emergent chinook salmon in
a small, stable, New Zealand stream. Journal of Fish Biology, 32: 27-40.

Fonseca, D.M. and Hart, DD. 1996. Density-dependent dispersal of black fly neonates is
mediated by flow. Oikos, 75: 49-58.

Gardiner, W.R. and Geddes, P. 1980. The influence of body composition on the survival
of juvenile salmon. Hydrobiologia, 69 (1-2): 67-72.

Gersabeck, ER and Merritt, RW. 1979. Effect of physical factors on the colonization
and relocation behavior of immature black flies (Diptera, Simuliidae).
Environmental Entomology, 8(1): 34-39.

Gibbons, JD. 1993. Nonparametric statistics: an introduction. Sage Publications, Inc.,
Newbury Park, CA. p. 83.

Great Lakes Fishery Commission. 1992. Strategic vision of the Great Lakes Fishery
Commission for the decade of the 1990's. Ann Arbor, MI.

Great Lakes Fishery Commission. 2001. Strategic vision of the Great Lakes Fishery
Comrrrission for the first decade of the new rrrillennium. Ann Arbor, MI.

Hansen, M.J. and Hayne, D.W. 1962. Sea lamprey larvae in Ogontz Bay and Ogontz
River, Michigan. The Journal of Wildlife Management, 26(3): 237-247.

Hardisty, M.W. 1961. The growth of larval lampreys. Journal of Animal Ecology, 30:
357-371.

Heinrich, J .W., Weise, J .G., and Smith, BR. 1980. Changes in biological characteristics
of the sea lamprey (Petromyzon marinas) as related to lamprey abundance, prey
abundance, and sea lamprey control. Canadian Journal of Fisheries and Aquatic
Sciences, 37: 1861-1871.

Hj ort, J. 1914. Fluctuations in the great fisheries of Northern Europe viewed in the light
of biological research. Rapports et Procés-Verbaux des Rénunions, Conseil
International pour L'exploration de la Mer., 20: 1-20.

Hopper, KR. 1999. Risk-spreading and bet-hedging in insect population biology.
Annual Review of Entomology, 44: 535-560.

76

Howard, W.E. 1960. Innate and environmental dispersal of individual vertebrates.
American Midland Naturalist, 63: 152-161.

Hume, J .M.B. and Parkinson, EA. 1987. Effect of stocking density on the survival,
growth, and dispersal of steelhead trout fi'y (Salmo gairdneri). Canadian J oumal
of Fisheries and Aquatic Sciences, 44: 271-281.

Johnston, T.A. 1997. Downstream movements of young-of-the-year fishes in Catamaran
Brook and the Little Southwest Miramichi River, New Brunswick. Journal of
Fish Biology, 51: 1047-1062.

Jones, M.L., Bergstedt, RA., T wohey, M.B., Fodale, M.F., Cuddy, D.W., and Slade,
J .W. In review. Compensatory mechanisms in Great Lakes sea lamprey
populations. ,

Kamler, E. 1992. Early life history of fish: an energetics approach. Chapman & Hall,
London.

Kempinger, J -.J . 1988. Spawning and early life history of lake sturgeon in the Lake
Winnebago system, Wisconsin. American Fisheries Society Symposium, 5: 110-
122.

Kerans, B.L., Chesson, P.L., and Stein, RS. 2000. Assessing density-dependent
establishment and dispersal: an example using caddisfly larvae. Canadian Journal
of Fisheries and Aquatic Sciences, 57:1190-1199.

Lasker, R. 1981. Marine fish larvae. Washington Sea Grant Program, University of
Washington, Seattle, WA.

Le Cren, ED. 1973. The population dynamics of young trout (Salmo trutta) in relation to
density and territorial behaviour. Rapports et Procés-Verbaux des Rénunions,
Conseil International pour L'exploration de la Men, 164: 241-246.

Letcher, B.H. and King, TL. 1999. Targeted stock identification using multilocus
genotype 'familyprinting'. Fisheries Research, 43: 99-111.

Mallatt, J. 1983. Laboratory growth of larval lampreys (Lampetra (Entosphenus)
tridentata Richardson) at different food concentrations and animal densities.
Journal of Fish Biology, 22: 293-301.

Manion, P.J. 1968. Production of sea lamprey larvae from nests in two Lake Superior
streams. Transactions of the American Fisheries Society, 97: 484-485.

Manion, P.J. and Hanson, L.H. 1980. Spawning behavior and fecundity of lampreys

fi'om the upper three Great Lakes. Canadian Journal of Fisheries and Aquatic
Sciences, 37: 1635-1640.

77

Manion, P.J. and Mclain, AL. 1971. Biology of larval sea lampreys (Petromyzon
marinas) of the 1960 year class, isolated in the Big Garlic River, Michigan, 1960-
65. Great Lakes Fishery Commission Technical Report, October, p. 35.

Merritt, R.W. and Cummins, K.W. 1996. An introduction to the aquatic insects of North
America. Kendall / Hunt Publishing Company, Iowa.

Meyer, RP. and Schnick, RA. 1983. Sea lamprey control techniques: past, present, and
future. Journal of Great Lakes Research, 9(3): 354-8.

Morman, RH. 1987. Relationships of density to growth and metamorphosis of caged
larval sea lampreys, Petromyzon marinas Linnaeus, in Michigan streams. Journal
of Fish Biology, 30:173-81.

Murdoch, S.P., Docker, M.F., and Beamish, F.W.H. 1992. Effect of density and
individual variation on growth of sea lamprey (Petromyzon marinas) larvae in the
laboratory. Canadian Journal of Zoology, 70:184-188.

Neter, J ., Kutner, M.H., Nachtsheirn, C.J., and Wasserrnarr, W. 1996. Applied Linear
Statistical Models. Fourth edition, Times Mirror Higher Education Group, Inc.,
USA.

Norris, A.T., Bradley, D.G., Cunningham, ER 2000. Parentage and relatedness
determination in farmed Atlantic salmon (Salmo salar) using microsatellite
markers. Aquaculture, 182: 73-83.

Oconnor, L. and Kelso, J .R.M. In review. Upstream migration rates, emigration and
likelihood of reproduction by sea lamprey (Petromyzon marinas) populations in
Great Lakes streams.

Piavis, G.W. 1961. Embryological stages in the sea lamprey and effects of temperature
on development. US. Fish and Wildlife Service Fishery Bulletin, 61(182).

Purvis, H.A. 1979. Variations in growth, age at transformation, and sex ratio of sea
lampreys reestablished in chemically treated tributaries of the upper Great Lakes.
Great Lakes Fishery Commission Technical Reports, 35: 36 p.

Randall, R.G. 1982. Emergence, population densities, and growth of salmon and trout
fly in two New Brunswick streams. Canadian Journal of Zoology, 60: 2239-
2244.

Robinson, A.T., Clarkson, R.W., and Forrest, RE. 1998. Dispersal of larval fishes in a
regulated river tributary. Transactions of the American Fisheries Society, 127:
772-786.

Ross, DH. and Merritt, R.W. 197 8. The larval instars and population dynamics of five

species of black flies (Diptera: Sirnuliidae) and their responses to selected
environmental factors. Canadian Journal of Zoology, 56(8): 1633-1642.

78

 

Scribner, K and Pearces, J .M. 2000. Microsatellites: evolutionary and methodological
background and empirical applications at individual, population, and phylogenetic
levels. In Molecular Methods in Ecology, edited by A.J. Baker, Blackwell
Science Ltd., USA, p. 235-273.

Shuter, BJ. and Post, J .R. 1990. Climate, population viability, and the zoogeography of
temperate fishes. Transactions of the American Fisheries Society, 119: 314-336.

Smith, BR. 1971. Sea lampreys in the Great Lakes of North America. In The biology of
lampreys, volume 1, edited by M.W. Hardisty and LC. Potter, Acedemic Press,
NY, p 207-47.

Snucins, E.J., Curry, R.A., and Gunn, J .M. 1992. Brook trout (Salvelinusfontinalis)
embryo habitat and timing of alevin emergence in a lake and a stream. Canadian
Journal of Zoology, 70: 423-427.

Sokal, RR. and Rohlf, P.J. 1995. Biometry: the principles and practice of statistics in
biological research. W.H. Freeman and company, New York.

Sutton, T.M. and Bowen, SH. 1994. Significance of organic detritus in the diet of larval
lampreys in the Great Lakes basin. Canadian Journal of Fisheries and Aquatic
Sciences, 51: 2380-2387.

Thompson, SK. 1992. Sampling. John Wiley & Sons, Inc., New York.

Thompson, SK. and Seber, GAR 1996. Adaptive sampling. John Wiley & Sons, Inc.,
New York.

Torblaa, R.L. and Westman, R.W. 1980. Ecological impacts of larnpricide treatments on
sea lamprey (Petromyzon marinas) ammocoetes and metamorphosed individuals.
Canadian Journal of Fisheries and Aquatic Sciences, 37: 1835-1850.

Trippel, EA. and Chambers, RC. 1997. The early life history of fishes and its role in
recruitment processes. Editors: R.C. Chambers and EA. Trippel. Early life history
and recruitment in fish populations. Chapman and Hall, London.

Van Den Bussche, R.A., Harmon, S.A., Baker, R.J., Bryan Jr., A.L., Rodgers Jr., J .A.,
Harris, M.J., and Lehr Brisbin Jr., I. 1999. Low levels of genetic variability in
North American populations of the wood stork (Mycteria americana). The Auk,
116(4): 1083-1092.

Walton, Jr., 0. E., Reice, SR, and Andrews, R.W. 1977. The effects of density,
sediment particle size and velocity on drift of Acroneuria abnormis (Plecoptera).
Oikos, 28: 291-298.

Ware, D.M. 1975. Relation between egg size, growth, and natural mortality of larval
fish. Journal of the Fisheries Research Board of Canada, 32: 2503-2512.

79

Weise, J .G. and Pajos, TA. 1998. Inn‘aspecific competition between larval sea lamprey
year-classes as Salem Creek was recolonized, 1990-1994, after a larnpricide
application. North American Journal of Fisheries Management, 18: 561-568.

Wirgin, I. I. and Waldman, J .R. 1994. What DNA can do for you. Fisheries, 19(7): 16-
27.

Wootton, R.J. 1990. Ecology of teleost fishes. Chapman and Hall Ltd., London.

Yap, MR. and Bowen, S.H. In review. Feeding by Northern Brook Lamprey
(Ichthyomyzon fossor) on sestonic biofilrn fi'agments: habitat selection results in
ingestion of a higher quality diet. '

80

      

  

MII.H|(1AN E IATL rIHI‘APIE‘n
‘ . I
i I I

3 1293 02314 7246