MAGEE - DEC 1999

EFFECT OF ACID AND ALUMINUM ON THE PHYSIOLOGY AND MIGRATORY
BEHAVIOR OF ATLANTIC SALMON SMOLTS IN MAINE
By
John A. Magee
B.A. St. Mary’s College of Maryland, 1993
A THESIS
Submitted in Partial Fulfillment of the
Requirements for the Degree of
Master of Science
(in Zoology)
The Graduate School
University of Maine
December, 1999
Advisory Committee:
Terry Haines, Professor of Zoology, Advisor
John R. Moring, Professor of Zoology
Irv Kornfield, Professor of Zoology

LIBRARY RIGHTS STATEMENT
In presenting this thesis in partial fulfillment of the requirements for an advanced
degree at the University of Maine, I agree that the Library shall make it freely
available for inspection. I further agree that permission for “fair use” copying of
this thesis for scholarly purposes may be granted by the Librarian. It is
understood that any copying or publication of this thesis for financial gain shall
not be allowed without my written permission.

EFFECT OF ACID AND ALUMINUM ON THE PHYSIOLOGY AND MIGRATORY
BEHAVIOR OF ATLANTIC SALMON SMOLTS IN MAINE
By John A. Magee
Advisor: Dr. Terry Haines
An Abstract of the Thesis Presented
in Partial Fulfillment of the Requirements for the
Degree of Master of Science
(in Zoology)
December, 1999
In recent years, the numbers of adult Atlantic salmon, Salmo salar,
returning to the rivers of eastern Maine have remained lower than that needed to
sustain their populations. Many attempts have been made to determine the
cause(s) of these declines, but investigations have not yielded sufficient answers.
Because it has been confirmed that the smolt stage is very sensitive to acidic
run-off and increased aluminum concentrations (common in eastern Maine rivers
during the time when smolts migrate to seawater), I investigated the effect that
these abiotic variables have on smolt physiology and migratory behavior.
Atlantic salmon, Salmo salar, smolts of hatchery origin were held for 5 to
16 days in ambient (pH 6.35, labile Al = 60 µg L-1
), limed (pH 6.72, labile Al =
58.4 µg L-1
), or acidified (pH 5.47, labile Al=96 µg L-1
) water from the
Narraguagus River in eastern Maine, USA. I measured gill Na+
/K+
ATPase
activity, hematocrit, and blood chloride concentration, and examined gill
morphology of smolts in freshwater and after 24-hour exposure to seawater.

Actively migrating wild smolts caught during this time were held for up to three
days in ambient river water. Hatchery smolts exposed to acidic water and wild
smolts displayed sub-lethal ionoregulatory stress both in fresh and seawater with
mortalities of wild smolts in seawater. Smolts exposed to limed water had a
greater proportion of chloride cells on the gill filaments but did not seem to
physiologically benefit from the liming.
Using ultrasonic transmitters and stationary receivers, I tracked 22
hatchery-reared and 26 wild smolts as they migrated through freshwater (5
kilometers) and estuarine (8 kilometers) sections of the river. The proportion of
wild smolts migrating during daylight hours was higher than that for hatchery-
reared smolts. Migrating wild smolts remained in the freshwater and the salinity
mixing zone of the estuarine portions of the river longer than hatchery smolts
exposed to either ambient or acid water, although survival during migration to
seawater was similar for all three treatments. Wild smolts were under
ionoregulatory stress that may have affected their migratory behavior, but not
their survival for the time and area in which I tracked them.

ii
Acknowledgements
I thank Terry Haines for the responsibility of being my advisor, and John
Moring and Irving Kornfield for serving on my committee. This thesis would not
have been possible without them.
Funding for this project was provided by U. S. Geological Survey,
Biological Resources Division, Leetown Science Center, Aquatic Ecology
Laboratory, Orono Field Station.
I would like to thank K. Beland, N. Dube, M. Martin, P. Rucksznis, and G.
Horton from the Maine Atlantic Salmon Commission, Bangor, Maine, and J.
Kocik and T. Sheehan of the National Marine Fisheries Service, Woods Hole,
Massachusetts, for allowing me to collaborate with them, performing surgeries,
field assistance, and giving essential advice. I thank M. Tabone for displaying
unsurpassed courage in the field and preparing samples for electron microscopy;
A. Amirbahman for jamming chemistry down my throat; W. Halteman for
statistical advice; and Mr. And Mrs. Frank Gross for access to the study site.
Special thanks goes to my wife, Donna, for her keen listening skills, and my
brother, Matt, for showing me how to objectively see the world.

iii
Table of Contents
Page
Acknowledgements…………………………………………………………………...ii
List of Tables……..…………………………………………………………………...v
List of Figures…..……………………………………………………………………..vi
Introduction……..……………………………………………………………………..1
Materials and Methods.……………………..………………………………………..5
Study area……………………………………………………………………..5
Treatment tanks.………………………………………………………………5
Water chemistry………………………………………………………………9
Physiology tests………………………………………………………………10
Sampling procedure.…………………………………………………………11
Ultrasonic tracking……………………………………………………………12
Surgical procedures…………………………………………………………..14
Analyses and statistics……………………………………………………….15
Results…………………………………………………………………………………19
Water chemistry………………………………………………………………19
Physiology…………………………………………………………………….24
Mortality……………………………………………………………….24
Na+
/K+
ATPase activity………………………………………………27
Blood chloride concentration………………………………………..27
Hematocrit…………………………………………………………….29
Electron microscopy…………………………………………………30

iv
Migratory behavior……………………………………………………………33
Residence time……………………………………………………….33
Timing of movement…………………………………………………35
Survival during migration……………………………………………54
Discussion…………………………………………………………………………….60
Physiology…………………………………………………………………….60
Migratory behavior and residence time……………………………………65
Conclusions…………………………………………………………………………..71
References……………………………………………………………………………72
Appendix………………………………………………………………………………80
Biography……………………………………………………………………………..81

v
List of Tables
Page
Table 1. Physiochemical parameters of each river section during May,
1998……………………………………………………………………………..13
Table 2. Water Chemistry of Narraguagus River from March to May
1998……………………………………………………………………………..21
Table 3. Physiological parameters of smolts held in freshwater and each
seawater challenge test……………………………………………………….28
Table 4a. Residence time of smolts in each river section…….………..………...34
Table 4b. Total residence time of smolts in the study area. …………..…………34
Table 4c. The number of tidal cycles needed for smolts to leave the
estuarine sections of the river………………………………………….…….34
Table 5. Time of day at which smolts left each section…………………………...43
Table 6. Hours past high tide at which smolts left each section……..……….….44
Table 7a. Time of day that smolts returned to Site C, and either
successfully or unsuccessfully left the upper estuary (passed
Site D)…………………………………………………………………………..56
Table 7b. Time after high tide that smolts returned to Site C, and either
successfully or unsuccessfully left the upper estuary (passed
Site D)…………………………………………………………………………..56
Table 8. Survival estimates of migrating wild and hatchery-reared
smolts………………………………………………………………………...…59

vi
List of Figures
Page
Figure 1. Narraguagus River, Maine, showing location of study area for
ultrasonic tracking of Atlantic salmon smolts in 1998…………………….…6
Figure 2. Temperature and pH of ambient Narraguagus River water from
4 March to 22 May 1998…………………………….………………………..20
Figure 3. Temperature (A), pH (B), and specific conductance (C) of test
tanks from 2 May to 22 May 1998….………………………………………..22
Figure 4a. LAl concentration in the test tanks and Narraguagus River from
4 May to 21 May 1998…………………….……………..………………..…..25
Figure 4b. LAl speciation in the acidified treatment tank (freshwater) at
the start of each seawater challenge test…….………………………….....25
Figure 5a. Electron micrograph of gill filaments from a hatchery-reared
smolt held in ambient Narraguagus River water…………………………...31
Figure 5b. Electron micrograph of gill filaments from a hatchery-reared
smolt held in limed Narraguagus River water……………………………...31
Figure 5c. Electron micrograph of gill filaments from a hatchery-reared
smolt held in acidified Narraguagus River water…..……………………...32
Figure 5d. Electron micrograph of primary filament of a hatchery-reared
smolt held in limed Narraguagus River water…………..……………….…32
Figure 9. Discharge of Narraguagus River from 2 May to 22 May 1998……….36
Figure 10. Time of day at which smolts passed the freshwater stationary
receiver at site A………………………………………………………………37
Figure 11. Time of day at which smolts passed the freshwater stationary
receiver at site B……………………………………………………..….…….39
Figure 12. Time of day and time after high tide at which smolts passed
the stationary receiver at site C………..……………………………..……..41

vii
Figure 13. Time of day and time after high tide at which smolts left the
upper estuary……………………………………………………………….…45
middle estuary………………..……………………………………………….47
lower estuary..……………….………………………………………………..49
Figure 16a. Temperature and salinity recorded at the middle of the
middle estuary from 10 May to 22 May 1998………………………..……..52
Figure 16b. Temperature and salinity recorded at the E Array from
10 May to 22 May 1998…………..…………………………………….…....52
Figure 17. Time after high tide at which smolts successfully (A) or
unsuccessfully (B) left the upper estuary…..……………………………….57

1
Introduction
The last naturally-occurring populations of Atlantic salmon in the United
States are in seven rivers in eastern Maine. Local populations were extirpated in
some rivers in the 19th
and 20th
centuries due to land and river management
practices (Haines, 1987). Restocking programs were implemented in the early
part of this century and continue to this day. At least one, and usually multiple,
life stages are stocked into sixteen rivers in Maine. Although the rivers are
heavily stocked (e.g., from 1989-1994, 29,000 fry, 9,500 0+parr, 7,000 age 1
parr, 54,100 age 1 smolts, and 4,900 age 2 smolts were stocked into the
Narraguagus River), adult returns remain quite low (only 325 wild adults returned
to the Narraguagus River from 1991-1996). Smolt-to-adult return rates ranged
from 0.075% to 0.575% for the Penobscot River from 1987 to 1994 (Beland and
Dube, 1998) and 0.5% to 0.93% for the Narraguagus River from 1989 to 1991 (K.
Beland, Maine Atlantic Salmon Commission, Bangor, Maine). Similarly, angler
catches have declined precipitously since the mid-1980s in the Narraguagus
River (Beland and Dube, 1998). The decrease in adult returns may be due to
commercial fishing mortality, which was estimated to be 35% to 50% from 1987
to 1993 (Friedland, 1995). However, returns of wild adult salmon in the five
eastern Maine rivers have declined since the 1980s despite the fact that
curtailment of the marine fishery has significantly decreased high-seas mortality
(K. Beland, Maine Atlantic Salmon Commission, Bangor, Maine).

2
Atlantic salmon populations are known to be adversely affected by both
acidic water (pH<6) and high inorganic aluminum concentrations (Farmer et al.,
1989; Staurnes et al., 1993, 1995, 1996), both of which are common in the rivers
of Maine (Haines et al., 1990). In-river mortality can be high under these
conditions (Staurnes, et al., 1993), but it is the time of parr-smolt transformation
at which the species is most susceptible to acid/Al stress (Rosseland and
Skogheim, 1984; Henricksen et al., 1984; Skogheim and Rosseland, 1986;
Staurnes et al., 1993).
The gills seem to be the target organ affected by low pH and high Al
concentrations, and effects on the morphology of fish gills are well known.
Mueller et al., (1991) reported major damage to and fusion of the secondary
lamellae, and proliferation of the mucous cells in response to pH 5.2 and 150 µg
total Al L-1
. Other effects include lifting of the epithelium of the secondary
lamellae (Daye and Garside, 1976), lesions (Smith and Haines, 1995), increased
volume density of chloride cells and diffusion distance over secondary lamellae
(Tietge et al., 1988). In freshwater, fish lose the ability to regulate ion content
(Booth et al., 1988), blood pH decreases (acidosis), oxygen uptake declines
(Walker et al., 1988), and feeding behavior is altered (Cleveland et al., 1991).
When anadromous salmonids are transferred to seawater after exposure to
acidic water, they often suffer major ionoregulatory stress. Blood chloride is
elevated, hematocrit values are low, and death may result (Lacroix, 1985; Farmer
et al., 1989; Pauwels, 1990; Staurnes et al., 1993).

3
Aluminum increases acid toxicity in fish, and aluminum speciation is very
important in this process. Total Al concentrations may not be indicative of
toxicity. Labile Al (LAl), that fraction of the total dissolved Al that is available to
organisms (i.e., not bound to organics), has been shown to be the form to which
fish are most susceptible (Driscoll et al., 1980; Rosseland and Skogheim, 1984).
In Norway, Atlantic salmon populations seem to be affected at pH 6.2 and
50 µg LAl L-1
(Staurnes et al., 1993), fish have depressed osmoregulatory
capabilities at pH 5.8 to 6.2 and 15 to 20 µg LAl L-1
(Kroglund and Staurnes,
unpublished data), and mortalities occur at pH<5.8 and 30 to 90 µg LAl L-1
.
Staurnes et al. (1996) correlated pH and Al concentration at the time of release
of hatchery-reared smolts to the number of adult returns in two rivers in southern
Norway. Those released into acidic waters (pH 4.8, LAl = 30 to 120 µg L-1
) were
physiologically compromised. Many died upon transfer to seawater, and few
were recaptured as returning adults. Smolts released into limed waters (pH >
5.8, LAl < 35 µg L-1
) were physiologically well prepared for seawater survival, and
significantly more adults returned two years later to spawn. This suggests that
declines in adult Atlantic salmon returns in southern Norway are due to
significant mortality of migrating smolts.
In addition to mortality from impaired physiology, smolts are confronted
with numerous predators while in the estuary of their parent river (Anthony,
1995), particularly double-crested cormorants (Phalacrocorax auritus) in North
America (Moring, 1987; Blackwell and Krohn, 1997). To minimize predation, it is

4
considered advantageous for smolts to minimize the time spent in this region of
the river. Because osmotic stress caused by exposure to seawater has been
shown to affect anti-predatory behavior of smolts in a laboratory setting (Jarvi,
1989, 1990; Handeland et al., 1996), it is of interest to learn what effect this has
on smolts migrating in a natural setting.
In an effort to understand why adult Atlantic salmon returns remain low in
Maine, I studied the effect that acidic water and aluminum have on the
physiology and migratory behavior of Atlantic Salmon smolts in the Narraguagus
River, Maine in May 1998. The overall study was divided into two separate, but
related, studies. The first investigated the effect of acid and aluminum on the
physiology of Atlantic salmon smolts. Hatchery-reared smolts were held in
stream-side tanks of three different pH values and aluminum concentrations.
Wild smolts were collected using rotary-screw type traps and held in ambient
river water. Three physiological parameters were examined while in freshwater
and after 24 hours in seawater. The second study investigated the migratory
behavior of hatchery-reared smolts, held in both ambient and acidified river water
and wild-collected smolts. Ultrasonic transmitters were surgically implanted into
smolts, which were released into the river and subsequently followed by
stationary receivers placed within the fresh and seawater sections of the river. I
hypothesized that exposure to acidified water would lead to a change in
migratory behavior and a decrease in survival during seaward migration.

5
Materials and Methods
Study area
The study was conducted on the Narraguagus River in eastern Maine
(Figure 1). The Narraguagus River is low in alkalinity (generally <150 µeq L-1
),
mildly acidic (typically pH 5-7), and prone to seasonal and short term
depressions in pH (Haines et al., 1990). Historically, Atlantic salmon were
relatively abundant in this river, with several hundred adults returning to spawn
each year (Beland and Dube, 1998). The river is approximately 70 km long and
contains at least 600,000 m2
of suitable salmon fry and parr habitat. The
watershed is approximately 600 km2
.
Treatment tanks
For the physiology study, smolts were held in four, 385 L plastic tanks
located approximately 10 river km upstream of the head-of-tide, and about 100 m
downstream of the upper smolt traps (Figure 1). Water was supplied to the tanks
via 122 m of 3.8 cm diameter plastic tubing, the upstream end of which was
placed at the top of a large riffle area. The vertical drop to the top of the
experimental tanks was 2.1 m. Flow rate in all tanks ranged from 5.7 L min-1
on
1 May to 5.5 L min-1
on 21 May 1998. Ambient river water flowed into four

6
Figure 1. Narraguagus River, Maine, USA. Smolts were held approximately 100
meters downstream of the Upper Smolt Trap. Direction of freshwater
current is from the Upper Smolt Trap to Site C.
F2 Array
F1 Array
Site D
E Array
30 1.5
Kilometers
Lower Smolt Trap
Upper Smolt Trap
Site A
Site B
Site C
Detector
Salinity & Temperature
Smolt Trap
Narraguagus River
Maine

7
mixing tanks (75 L), one for each exposure tank. Water in one tank was acidified
and aluminum added via a 20 L container held above the mixing tank, with a drip
rate of 2.2 mL min-1
of 3.17 x 10-3
M Al(OH)3
.
6H2O and 5 mL l-1
concentrated
HCl. Because eastern Maine rivers supporting Atlantic salmon populations are
commonly in the range of pH 5.0 to 5.7 in the spring (Haines et al., 1990), this
was my target pH range. From 7 May to 12 May, I altered the pH in the limed
tank by passing ambient river water through a mesh bag of CaCO3 chips. This
method produced a few spikes of very high pH, and was stopped on 12 May.
After 12 May, limed water was added in the manner of the acidic water, but with
a drip rate of 5.6 mL min-1
of 2M NaHCO3. The target pH range was 6.6 to 7.
Hydrochloric acid and NaHCO3 solutions were made using deionized water, and
the volume of the containers was maintained between 16 and 20 L by
replenishment twice daily.
The residence time of water in each mixing tank and experimental tank
was 13.7 and 78 min, respectively. Oxygen from a single compressed oxygen
container was delivered to each tank with standard aquarium airline tubing and
one 15 cm long airstone. Dissolved oxygen was measured in all tanks on 12, 17,
and 20 May using a YSI Model 50B dissolved oxygen meter, and was always
greater than 7.5 mg L-1
, and normally greater than 8.5 mg L-1
. Ammonia
concentration was checked every other day using a commercially available test
kit for freshwater (Aquarium Pharmaceuticals, Inc., Chalfont, PA). Ammonia was
always below the detection level of 0.5 mg L-1
.

8
On 2 May 1998, 153 one-year-old Atlantic salmon smolts of Penobscot
River stock (mean total length ± 1SE = 19.15 ± 2.73 cm; mean weight ± 1SE =
67.27 ± 2.74 g; n=24) were obtained from the Green Lake National Fish Hatchery
in Ellsworth, Maine. Water at the hatchery is of circumneutral pH, moderately
low alkalinity, and low aluminum concentration (Appendix 1). Smolts were
transferred to the study site by a truck equipped with compressed oxygen and a
water circulation device (the trip was 1.5 hours, and water temperature was
10ºC).
I planned to investigate the physiological responses of the Machias River
stock of salmon in addition to the Penobscot River stocks. On 3 May, 88 two-
year-old smolts of Machias River stock were transferred as above (except that
the trip duration was 25 minutes) and added to the four tanks. On 4 May, many
smolts were dead in all tanks, and all Machias River smolts were removed. The
cause of death was determined to be lack of oxygen due to overcrowding. On 6
May, an additional 69 smolts of Penobscot stock were transferred as before and
distributed equally among the three tanks for a final count of 50 smolts per tank
except that one tank (that which was to also hold wild smolts) received only 25
Penobscot stock smolts.
Wild smolts were collected from the Narraguagus River using rotary-screw
type smolt traps (Thedinga et al., 1994) positioned 150 m upstream of the study
site (for physiology experiments) and 4 km downstream from the study site (for
migratory behavior experiments). The traps were checked for smolts between
0700 and 0900 hours each day. For physiology experiments, I collected five wild

9
smolts on 7, 8, and 9 May, and eight and seven smolts on 14 and 15 May,
respectively. The wild smolts were placed into the tank containing only 25
Penobscot smolts to keep the density of smolts equal in each tank. All smolts
were fed earthworms ad libitum once daily. I observed wild smolts eating only
rarely, whereas hatchery-reared smolts fed without hesitation in all treatments.
Water chemistry
Exposure of smolts to treatment water began on 7 May. Water samples
were taken daily from the river and exposure tanks for analysis of total and
inorganic labile aluminum (USEPA, 1987), and on 10, 16, and 21 May for
analysis of anions, cations, dissolved organic carbon (DOC), alkalinity, and
flouride. Temperature, specific conductance, and pH were measured every three
hours using automatic recording units and verified twice daily with a calibrated
thermometer, portable pH meter (Orion combination electrode 9165) and
portable conductivity meter (Orion model 120). Light intensity measurements
were taken at the upper and lower smolt traps every 15 minutes with a HOBO®
light intensity meter. These data were used to determine under what light
conditions smolts migrated. Discharge of the river at the ice dam in Cherryfield,
Maine (1 km upstream of head of tide) was recorded by the United States
Geological Survey.

10
At least 15 scales were removed from just posterior of the dorsal fin on the
right side all wild smolts used in the physiology studies to determine their age.
Scales were read on a microfiche reader.
Physiology tests
Three to five smolts were sampled at random from all tanks on 10, 16, and
21 May to determine Na+
/K+
ATPase activity, blood chloride level, hematocrit,
and gill morphology. To assess the ability of smolts to tolerate the transition to
salt water, I conducted three seawater challenge tests (SWCT) using modified
methods of Clarke (1982). Smolts were removed from the treatment tanks on 10,
16, and 21 May, taken by truck (duration of 10 minutes at ambient treatment
temperature and pH) to a building in Cherryfield owned by the State of Maine,
and placed into artificial seawater. The SWCT were conducted here to keep the
temperature between 8 and 10ºC. Seawater of 34‰ was made by adding Instant
Ocean®
salt to two tanks of 160 L of Narraguagus River water at least 24 hours
before introducing smolts. Each tank was divided in half and smolts from a
single treatment were placed into each half. Aeration was provided to each tank
by two large air pumps and dissolved oxygen never fell below 9 mg L-1
. Duration
of the seawater challenge tests was 24 hours and Na+
/K+
ATPase activity, blood
chloride level, hematocrit, and gill morphology of surviving smolts were assessed
at the end of each test. Wild smolts used on 10 and 16 May were collected
during 7 to 9 May and 14 to 15 May, respectively. Condition factor was

11
calculated as 100
.
(weight)
.
length-3
, with weight in g and length (fork length) in
cm.
Sampling procedure
Smolts were anesthetized with enough NaHCO3-buffered (to the pH of the
treatment tank from which they were removed) tricaine methanesulfonate (MS-
222, Finquel®, Argent Chemical Laboratories) to immobilize them within two
minutes. To determine the Na+
/K+
ATPase activity, approximately 6 to 10 gill
filaments were removed from the second gill arch from the left side of each smolt
and placed into 100 µL ice-cold SEI buffer (150 mM sucrose, 10 mM EDTA, 50
mM imidazole, pH 7.3) and frozen at -12ºC. Measurement of Na+
/K+
ATPase
activity was made within 7 weeks by the methods of McCormick (1993). Smolts
were bled from the caudal vein using a 3 mL pre-heparinized syringe and a 26 -
gauge needle, and the blood placed into two, 1.5 mL pre-heparinized
microcentrifuge tubes. Two 100 µL hematocrit tubes were filled with the blood,
centrifuged at 3,000 RPM for 5 minutes, and hematocrit read immediately. The
difference between the two readings was generally <1%. The remaining 1 to 3
mL of blood was centrifuged at 3,000 RPM for 3 minutes. The supernatant was
carefully removed, placed into a new 1.5 mL pre-heparinized microcentrifuge
tube, and frozen at -20ºC. Serum chloride concentration was determined by ion
chromatography.

12
The second gill arch from the right side of each smolt was removed and
immediately placed into a fixative consisting of 4% glutaraldehyde, 1%
formaldehyde, and 5% sucrose HEPES buffer. Within six weeks, the excised
gills were treated with osmium (1% OSO4), dehydrated with a graded ethanol
series (50%, 70%, 85%, 95%, 100% ethanol), dried in a critical point dryer, and
sputter coated with gold to a thickness of 250 nm. The gills were examined with
an AMR 1000 scanning electron microscope at 5kV at a magnification of 100x to
2000x. I examined the gross morphology of the gills for characteristic
acid/aluminum damage and the presence of externally visible structures (e.g.,
chloride cells).
Ultrasonic tracking
Smolt migratory behavior was monitored with the use of ultrasonic
transmitters (pingers) and stationary ultrasonic receivers placed in the river
system (Figure 1). Pingers and receivers were purchased from Vemco(®)
Limited (Voegeli et al., 1998). Pingers were Vemco(®) model V8-SC-L1, which
were 8x25 mm in size 4.3 g weight in air, 2.7 g weight in water. The expected
battery life was 21 days.
Receivers were located in three regions within the freshwater (FW)
sections (sites A, B, and C, Figure 1) and four regions in the estuarine sections
(site D, and the E, F1, and F2 arrays, Figure 1). These locations are boundaries
between distinct physical zones of the river (Table 1). Site B is at the end of 3

13
km of slow-moving water. Between sites B and C is 1 km of fast-moving water
with many, small (<0.5 m) falls and riffles. The head of tide is approximately 200
m upstream from Site C. Between sites C and D is the 'Upper Estuary'. A salt
wedge extended to approximately 3 km downstream of the head of tide and
moved to as far as 9 km downstream of the head of tide at low tide. The salinity
in this section ever exceeded 10‰. The 'Middle Estuary' between sites D and E
is characterized by widely ranging salinities and temperatures resembling that of
the ambient ocean at high tide and FW at low tide. Tidal flow, at mid tide, in this
section is visibly faster than in other sections but this was not quantified. The
'Lower Estuary' is characterized by high (>25 ‰) salinity, visibly less tidal flow,
and lower (8 to 12º C) temperatures. In all estuarine sections, the water column
is vertically well mixed with respect to temperature and salinity. The mean tidal
amplitude during the study was 3.7 m. Temperature was monitored at the
release site, and sites B, C, D, and the F1 and F2 array. Temperature and
salinity were monitored in the middle of the middle estuary and at the E array
using automatic recording units.
Table 1. Physiochemical parameters of each river section during May, 1998.
River
Section
Length
(km)
Width
Range (m)
Depth
range (m)
Salinity
range (‰)
Temperature
range (ºC)
Freshwater 10 10-30 0.5-4 0 10-21
Upper
Estuary
3 10-60 1-6 0-10 8-21
Middle
Estuary
6 10-1400 2-8 0-31 5-19
Lower
Estuary
2 1400-1800 8-14 21-31 5-13

14
Surgical procedures
Smolts were anesthetized with sufficient buffered MS-222 to immobilize
them within two minutes. Fish were weighed and measured, then were placed
onto a damp sponge with an area cut out to accommodate the fish. This formed
a “cradle” in which the smolts could be comfortably positioned and held
stationary. All surgical tools were soaked in a 1:1000 v:v bath of the germicide
Benzalkonium chloride. An incision approximately 1.5 cm long was made in the
abdomen slightly posterior to the pectoral fins. The pinger was inserted into the
peritoneal cavity and three stitches were made using 4-0 suture (CE-6) wire.
Vetbond (™) (3M Animal Care Products) was applied to complete the closure of
the incision and allowed to dry (generally 4-5 sec.). A small amount of the
fungicide nitrofurazone (0.2%) was applied to the area around the incision. The
hatchery-reared smolts were immediately placed into a 20 L plastic bucket of
ambient river water for at least five min. and then into a holding cage in the river
for about 20 min. They were then released into the river in groups of three to
four. The wild smolts were allowed to recover for at least 5 min. in a 20 L bucket
of ambient river water and were released singly. This method is similar to that of
Moore et al. (1995), Lacroix and McCurdy (1996) and Moore et al. (1998), and
has been shown to have a negligible effect on physiology and behavior of the
smolts (Moore et al., 1990).
All receivers were deployed between 24 and 28 April, 1998. Each unit can
detect a pinger up to 100 m distant; therefore, an array of multiple units was used

15
in the middle and lower estuary areas to optimize the probability of detecting all
migrating smolts. In addition to passive tracking, seven days of active tracking
were completed using an omnidirectional hydrophone and portable receiver. The
active searches were conducted using boats and wading, and were concentrated
in the FW and upper estuary sections to ascertain mortalities of smolts in these
areas. Many pingers (either expelled or from smolts that had died several days
earlier) were recovered in these areas during preliminary studies conducted in
1997.
This study was done in cooperation with another investigation of the
migratory behavior of smolts in which smolt migration was monitored from 29
April to 28 May. Because physiochemical parameters of the river changed
markedly during this time, I used only data from wild smolts that were tagged and
released between 10 to 16 May and passed by site A from 12 to 16 May.
Hatchery-reared smolts were tagged and released on 13 and 14 May and passed
site A from 13 to 16 May.
Analyses and Statistics
All chemical analyses were done by standard methods (USEPA, 1987).
Cations were analyzed by flame atomic absorption spectroscopy using a Perkin-
Elmer model 3300 spectrophotometer. Total aluminum was analyzed using
graphite furnace atomic absorption spectroscopy using a Perkin-Elmer model
4100ZL spectrophotometer. LAl was measured by first passing a water sample

16
through an ion exchange column containing Dowex cation exchange resin HCR-
W2 H+
, 16-40 mesh, and then analyzed as above. The difference between total
and this value is LAl concentration. Anions were analyzed by ion
chromatography using a Dionex model DX 500. Dissolved organic carbon was
analyzed by persulfate oxidation followed by IR detection on an OI model 700
carbon analyzer. Alkalinity was analyzed by automated Gran titration using 0.02
N H2SO4.
LAl speciation of the acid treatment on the SWCT dates was calculated
using the computer program Mineql©
(Westall et al., 1976). The complexation of
Al with Cl-
, SO4
2-
, F-
, and NO3
-
was calculated using the measurements of pH,
temperature, cations, anions, DOC, and alkalinity.
For physiological parameters, SAS Institute (1990) software was used to
perform a one-way ANCOVA, using condition factor as the covariate. To detect
differences between treatments, Scheffe’s multiple range test was used because
of unequal replications among treatments. Because there was an interaction
between date of sampling and all physiological parameters, comparisons were
made within each sampling date only.
Previous studies have indicated significant relationships between the
physiological parameters examined here (McCormick et al., 1985). To ascertain
that the physiology of smolts used in the present study was similar to those in
previous studies, I calculated Pearson correlation coefficients between all
parameters that I measured. Preliminary research on the migratory behavior of
smolts in the Narraguagus River (1997 data) indicated a relationship between

17
fish length and survival in the estuary. In order to correlate physiology with
migratory behavior, I calculated the Pearson correlation coefficient between
condition factor and all physiological parameters measured.
Because residence time and tidal cycle data could not be log transformed
to produce a normal distribution, nonparametric Mann-Whitney tests were used
to determine differences between treatments in each section of the river, and the
number of tidal cycles needed to leave the estuary. Rayleigh's test was used to
determine if movement of smolts was random with respect to time of day or tidal
cycle (Batschelet, 1981), and standard deviations were calculated by the method
of Mardia (1972). The Watson-Williams test was used to determine differences
between treatments with respect to time of movement (Batschelet, 1981). These
two tests and the calculation of standard deviation are parametric and have been
used in previous studies on the migratory behavior of Atlantic salmon smolts
(Lacroix and McCurdy, 1996; Moore et al., 1995, 1998). Because many smolts
passed through multiple sections in only one night or one tidal cycle, I did not
calculate the correlation coefficient between the time of day or tide when smolts
passed each detector.
Mortality in each river section was estimated using the Mark computer
program with the Cormack-Jolly-Seber (CJS) model (White, 1998). This model
estimates survival between intervals (sections) and recapture (detection)
efficiency based on the detection of individual smolts in each river section. It
assumes an initial tagging location, which I have defined as site A. Details on the
statistical analysis of this data are in Cormack (1992). I calculated the correlation

18
coefficient of the relationship of the residence time within each section and all
other sections and between condition factor and residence time in each section.

19
Results
Water chemistry
Because of a battery failure, the automatic recording unit collected no
water quality data from 2 April to 14 April 1998, and data reported for this time
are from field monitoring with portable meters only. At other times, water quality
data are from this automatic recording unit (Figure 2; Table 2). The pH fell below
6 several times prior to 1 May, and a severe rainstorm lowered the pH to 4.95 on
11 March. After 24 April, the pH remained above 6, increasing to above 6.5 from
17 to 21 May.
Temperatures of the treatment tanks never differed by more than 0.5ºC
from the river temperature, and were similar to each other (Figure 3a). The pH of
the acid treatment decreased on 8 May and remained approximately 0.7 to 1.2
pH units lower than the ambient and limed treatments until the end of the
experiment. An accidental increase in the dosing rate on 18 May lowered the pH
to 4.4 but this was corrected within 2 hours. Because of problems with the
operation of the liming doser, the smolts in the limed treatment were held under
ambient conditions until 12 May (Figure 3b). The addition of ions to alter the pH
of the limed and acid treatment tanks led to a slight increase in specific
conductance (Figure 3c).
From 8 May to 21 May, the LAl concentration of the acid treatment was
between 30 and 165 µg LAl L-1
higher than for the other treatments which were

20
Figure 2. Temperature and pH of ambient Narraguagus River water from 4
March to 22 May 1998. Data from 4 March to 13 March are from 37 km
upstream of upper smolt trap. Data from 14 March to 22 May are from the
physiology study site. The lack of data points between 2 April and 14 April
is due to the failure of the automatic recording unit. Data on 8 April is from
measurements using portable pH and conductivity meters.

21
Table 2. Water Chemistry of Narraguagus River from March to May 1998.
Date 9 Mar 24 Mar 31 Mar 7 Apr 17 Apr 23 Apr 10 May 16 May 21 May
pH 6.02 6.13 6.27 6.08 6.4 6.29 6.32 6.51 6.64
Ca++
(µeq/L)
38.5 34.5 36.75 29.75 33 34.5 35.5 39.5 45.5
Mg++
(µeq/L)
13.99 15.64 14.81 14.4 15.64 16.87 15.9 16.89 18.11
Na+
(µeq/L)
62.17 86.96 87.39 81.3 97.39 94.35 85.62 102.36 120.87
K+
(µeq/L)
8.5 9.75 9.75 10.25 10.25 11 10.25 12.25 13.08
Cl-
(µeq/L)
39.7 74.1 74.1 65.1 74.1 68.2 72.7 74.5 82.7
NO3
-
(µeq/L)
2.26 2.57 2.57 1.69 2.36 2.03 <1.0 <1.0 <1.0
SO4
=
(µeq/L)
56.9 50.8 50.8 49 47.6 47.6 45.1 46.5 41.3
F-
(µg/L)
62 84 77 73 89 82 66 74 87
ANC
(µeq/L)
70 64 69 60 86 82 89.1 98.3 140
DOC
(mg/L)
6.63 5.96 5.36 6.47 5.84 6.31 7.62 7.31 6.15
Al (total)
(µg/L)
93 120 92 118 148 137 139 125 90
LAl
(µg/L)
3 10 3 3 58 43 73 58 0

22
Figure 3. Temperature (A), pH (B), and specific conductance (C) of test tanks
from 2 May to 22 May 1998.

24
similar to each other (Figure 4a). Between 9 May and 16 May, the total and LAl
concentration of the river decreased, which led to decreases of both species in
all treatments. On 16 May, LAl concentration was 99 and 70 µg LAl L-1
in the
acid and ambient treatments. Thereafter, the concentrations steadily increased
to a maximum of 176 and 75 µg LAl L-1
in acid and all other treatments,
respectively, until the end of the experiment on 21 May. Aluminum speciation in
the acid treatment was dominated by hydroxide and flouride species in the first
and third SWCT and by flouride in the second SWCT (Figure 4b).
Physiology
Mortality
No hatchery-reared smolts died in FW or SW except that four died
between 13 and 15 May in the tank that received ambient river water and held
wild smolts prior to testing. I noticed a fungus on those hatchery-reared fish that
died and on several others. From 16 May to 19 May, I removed six more smolts
that were still active but showed signs of fungus. No wild smolts showed signs of
the fungus at any time. After 24 hr exposure to SW in the second SWCT (17
May), another wild smolt died, one could not move and lay on the bottom of the
tank, and another was very lethargic and could not remain upright.

25
Figure 4a. LAl concentration in the test tanks and Narraguagus River from 4 May
to 21 May 1998.
Figure 4b. LAl speciation in the acidified treatment tank (freshwater) at the start
of each seawater challenge test.

27
Na+
/K+
ATPase activity
Enzyme activity in FW decreased with time in all treatments (Table 3).
Exposure to the acid treatment for three days (10 May) led to a 16% decrease
from 11.74 µmol ADP.
mg protein-1.
hr-1
in ambient-exposed smolts to 9.83 µmol
ADP.
mg protein-1.
hr-1
enzyme activity in acid-exposed smolts. Wild smolts had a
significantly greater enzyme activity than ambient and limed but not acid smolts
on 16 May. On 21 May, all hatchery groups of smolts had nearly identical
enzyme activities.
Following transfer to SW, the enzyme activity of wild smolts increased
(103% and 125% on 11 May and 17 May, respectively) whereas that of hatchery
smolts decreased (Table 3). There were no significant differences in mean
enzyme activity between the groups of hatchery smolts for any SWCT. Wild
smolts had significantly higher enzyme activities than did acid and ambient
smolts (except for that of ambient smolts in the first SWCT). In the second
SWCT, the enzyme activity of wild smolts was more than twice that of any
treatment of hatchery smolts.
Blood Chloride concentration
There were no significant differences among hatchery smolts in freshwater
on any date (Table 3). Blood chloride concentration of wild smolts on the 16 May
was the lowest recorded during the study, and was significantly lower than that of

28
Table 3. Physiological parameters of smolts held in freshwater and each
seawater challenge test. Mean ± 1SE; number in parentheses indicates number
of observations. A p<0.05 level was used to determine significance between
treatments within a sampling date. Treatments with the same letter are not
significantly different from each other.
Date Treatment Na+
/K+
ATPase
activity
(μmol ADP.
mg
protein-1.
hr-1
)
Blood [Cl-]
(meq L-1
)
Hematocrit
(%)
10 May Ambient 11.74 ± 0.82 (4) a
148 ± 7.31 (3) a
53.5 ± 1.9 (5)ab
FW Limed . . .
Acid 9.83 ± 1.43 (3) a
158.75 ± 3.68 (4) a
56.8 ± 2.2 (4) a
Wild 9.87 ± 0.56 (4) a
152.25 ± 15.98 (4) a
48.1 ± 1.8 (4) b
11 May Ambient 8.09 ± 0.74 (10) ab
203.4 ± 5.52 (10) a
51.6 ± 2.1 (10) a
SW Limed . . .
Acid 7.79 ± 0.87 (10) a
210.9 ± 12.85 (10) a
44.3 ± 1.3 (10) b
Wild 10.2 ± 0.75 (6) b
207.9 ± 9.35 (7) a
46.7 ± 1.1 (7) b
16 May Ambient 3.61 ± 0.43 (4) a
148.67 ± 10.73 (3) ab
45.8 ± 1.7 (4) a
FW Limed 4.12 ± 0.5 (6) a
172 ± 6.08 (3) b
46.4 ± 1.9 (6) a
Acid 4.67 ± 0.76 (5) ab
153 ± 5.57 (3) ab
50.1 ± 3.3 (5) a
Wild 6.17 ± 0.5 (5) b
126 ± 9.21 (4) a
51.7 ± 0.3 (5) a
17 May Ambient 2.56 ± 0.23 (8) a
197 ± 10.15 (7) a
41.6 ± 1.6 (8) a
SW Limed 3.12 ± 0.23 (10) a
216.2 ± 13.7 (10) a
42.7 ± 1.4 (10) a
Acid 3.11 ± 0.29 (5) a
224.8 ± 23.7 (5) a
39.5 ± 1.9 (5) a
Wild 7.72 ± 0.29 (8) b
214.1 ± 7.54 (8) a
38.1 ± 2.0 (8) a
21 May Ambient 2.83 ± 0.29 (5) a
148.4 ± 8.13 (5) a
47.6 ± 0.8 (5) a
FW Limed 2.73 ± 0.28 (9) a
156.25 ± 4.33 (8) a
49.9 ± 1.2 (9) a
Acid 2.91 ± 0.09 (5) a
151.67 ± 7.8 (6) a
50 ± 1.5 (6) a
Wild . . .
22 May Ambient 2.42 ± 0.2 (10) a
198.5 ± 6.19 (10) a
48.9 ± 1.9 (10) a
SW Limed 2.46 ± 0.28 (10) a
220.5 ± 9.1 (10) ab
44.6 ± 1.0 (10) b
Acid 2.40 ± 0.15 (10) a
236.2 ± 11.53 (9) b
40.8 ± 1.5 (10) b
Wild . . .

29
lime-exposed smolts on that date. Upon transfer to SW, ambient smolts always
had a lower blood chloride concentration than all other treatments. In the third
SWCT, acid smolts had significantly higher blood chloride than ambient smolts.
In general, acid smolts had higher blood chloride than ambient smolts, and
showed the second greatest increase (147%) in blood chloride when transferred
to SW (17 May). Blood chloride of wild smolts increased by 170% from 16 May
to 17 May.
Hematocrit
In freshwater, the hematocrit of wild smolts on 10 May was significantly
lower than that of acid, but not ambient smolts (Table 3). There were no
differences between treatments on 16 May and 21 May. Exposure to SW led to
significant decreases in hematocrit (compared to ambient smolts) in both the acid
and wild smolts in the first SWCT, and acid and limed smolts in the third SWCT.
Wild smolts in the second SWCT displayed the greatest difference between
mean FW and SW hematocrit values (a 37% increase). Acid smolts in the third
SWCT displayed the second largest difference between FW and SW hematocrit
values (a 56% increase).
In FW, hematocrit was positively correlated with Na+
/K+
ATPase activity
(r2
=0.57, p<0.05, n=50), and negatively correlated with blood chloride level
(r2
=0.72, p<0.001, n=43). Upon transfer to SW, hematocrit was positively
correlated with ATPase activity (r2
=0.48, p<0.05, n=87) and negatively correlated

30
with blood chloride level (r2
=0.62, p<0.001, n=86). Condition factor was not
significantly correlated with any physiological parameter.
Electron microscopy
The gross morphology of gills in all treatments was remarkably similar.
Primary and secondary lamellae of gill samples taken on all test dates were
generally thin and unswollen (Figures 5a-c). Although a few showed signs of
possible acid/aluminum damage, the damage was slight and there were no clear
treatment effects. The gills of four ambient, four limed, three acid, and five wild
smolts sampled on 16 and 17 May had small lesions and slight lifting of the
epithelium. Because this was ubiquitous among smolts from all treatments on
these dates, I suspect that the lesions and epithelial damage were caused by
errors in the preparation of these samples. Only limed smolts from the third
SWCT had noticeably more chloride cells on the primary and secondary lamellae
(Figure 5d).

31
A
B
Figure 5a. Electron micrograph of gill filaments from a hatchery-reared smolt
held in ambient Narraguagus River water. Sample taken on 10 May 1998.
Scale bar = 100um.
Figure 5b. Electron micrograph of gill filaments from a hatchery-reared smolt
held in limed Narraguagus River water. Sample taken on 16 May 1998.
Scale bar = 100um.

32
C
D
Figure 5c. Electron micrograph of gill filaments from a hatchery-reared smolt
held in acidified Narraguagus River water. Sample taken on 22 May 1998.
Scale bar = 100um.
Figure 5d. Electron micrograph of primary filament of a hatchery-reared smolt
held in limed Narraguagus River water. Sample collected on 22 May
1998. Large circular structures on epidermis of primary filament are
chloride cells. Scale bar = 10um.

33
Migratory behavior
Residence time
There were clear differences between hatchery and wild smolts for the
residence time in FW (Table 4a). Wild smolts remained in FW a mean of 1.92
days, significantly longer than for ambient smolts (mean=0.53 days, p=0.012).
The residence time of acid smolts was not significantly different from that of wild
smolts (mean=0.70 days, p=0.076), even though wild smolts took approximately
2.75 times longer to leave the FW section. Only 48% (12) of the wild smolts
were able to leave the FW sections in one day, whereas 91% (9) and 82% (8) of
the ambient and acid smolts, respectively, did so. Residence time in the upper
estuary was similar for all three groups. In the middle estuary, the residence time
of wild smolts was significantly longer than for acid-exposed smolts (p=0.041),
but similar to that of ambient smolts (p=0.416). Condition factor was correlated
with the residence time of ambient (r2
=0.52, p<0.05, n=8) and wild (r2
=0.24,
p<0.05, n=18) smolts in the middle estuary. Those smolts with a higher condition
factor remained in this section longer.
In all river sections except the lower estuary, the mean residence time of
wild smolts was longer than for either group of hatchery smolts (Table 4a). The
total residence time that wild smolts spent in the FW and estuarine sections of
the river was significantly longer than that of ambient (p<0.001) and acid
(p=0.029) smolts (Table 4b). Ambient and acid smolts spent a mean of 2.78 ±

34
Table 4a. Residence time of smolts in each river section. Mann-Whitney tests
were used to determine significance (p<0.05). Treatments with the same letter
are not significantly different from each other.
Residence Time in days
(mean ± 1SE; number of observations in parentheses)Treatment
Freshwater Upper Estuary Middle Estuary Lower Estuary
Ambient
Acid
Wild
0.53 ± 0.107 (11)
a
0.703 ± 0.20 (9)a
1.92 ± 0.42 (26)b
0.58 ± 0.13 (11)
a
0.89 ± 0.38 (9) a
1.43 ± 0.34 (23) a
0.43 ± 0.108 (8)
ab
0.23 ± 0.13 (7) a
0.72 ± 0.18 (20) b
0.34 ± 0.16 (8)
a
0.28 ± 0.06 (8) a
0.23 ± 0.1 (20) a
Table 4b. Total residence time of smolts in the study area. Mann-Whitney tests
were used to determine significance (p<0.05). Treatments with the same letter
are not significantly different from each other.
Table 4c. The number of tidal cycles needed for smolts to leave the estuarine
sections of the river.
Residence Time in days
(mean ± 1SE; number of observations in parentheses)
Treatment Total days in river system
Ambient
Acid
Wild
1.76 ± 0.274 (9)
a
2.43 ± 0.51 (9) a
4.04 ± 0.50 (24) b
Number of tidal cycles (mean ± 1SE; number of observations in parentheses)
Treatment
Ambient
Acid
Wild
2.78 ± 0.40 (9) a
2.78 ± 0.73 (9) ab
4.86 ± 0.83 (22)
b

35
0.4 and 2.78 ± 0.70 tidal cycles, respectively, in tidal water before leaving the
estuary (Table 4c). Wild smolts remained for significantly more tidal cycles
(mean=4.86 ± 0.83) than ambient (p=0.045) but not acid (p=0.140) smolts. Only
one smolt (from ambient conditions) was able to successfully pass through the
entire estuary in only one tidal cycle.
There was a significant relationship between date of release and FW
residence time for all treatments (r2
=0.114, p<0.02, n=44), with smolts released
later spending more time in FW (although this was not true for every individual).
However, this regression only accounts for 11% of the variance in residence time
and is probably of no practical significance. This coincided with decreased river
discharge from 32.4 m3
sec-1
on 9 May to 6.4 m3
sec-1
on 19 May (Figure 9).
There was no significant correlation between FW residence time and estuarine
residence time for any of the three treatments.
Timing of movement
Movement of smolts in all groups in FW was non-random with respect to
time of day, and occurred primarily at night (Figures 10-12; Table 5). Few
hatchery-reared smolts moved during the brightest hours of the day (1000 to
1700 hours), although one ambient and three acid-exposed smolts passed site A
about noon. Smolts from all three treatments entered the upper estuary at times
that were random with respect to the tidal cycle (Figure 12; Table 6).

36
Figure 9. Discharge of the Narraguagus River from 2 May to 22 May 1998.
Discharge measurements were taken at the Ice Dam in Cherryfield,
Maine, at site B.

37
Figure 10. Time of day at which smolts passed the freshwater stationary receiver
at site A. Smolts that did not return (o) and those that returned (•) to site
C after having left the upper estuary are indicated. Large arrow and bar
indicate mean time and standard deviation of movement.

39
Figure 11. Time of day at which smolts passed the freshwater stationary receiver
at site B. Smolts that did not return (o) and those that returned (•) to site
C after having left the upper estuary are indicated. Large arrow and bar
indicate mean time and standard deviation of movement.

41
Figure 12. Time of day and time after high tide at which smolts passed the
stationary receiver at site C. Smolts that did not return (o) and those that
returned (•) to site C after having left the upper estuary are indicated.
Large arrow and bar indicate mean time and standard deviation of
movement. Small arrow in graph indicates time of high water.

42
Time of Day Time after High Tide

43
Table 5. Time of day at which smolts left each section. Mean ± 1SE given in
hours and minutes. Rayleigh’s ‘r’ value is a measure of the length of the mean
vector on the unit circle. A value of 1.0 indicates no variance in mean time. A
probability value <0.05 indicates that the time of day at which smolts passed
each receiver was nonrandom with respect to time of day.
Treatment Mean time leaving section ±±±± 1SD Rayleigh’s r value Probability
Freshwater (Site A)
Ambient 20:32 ± 3:22 0.68 p=0.05
Acid 20:31 ± 4:57 0.43 p>0.1
Wild 231:12 ± 3:32 0.65 p<0.001
Freshwater (Site B)
Ambient 23:56 ±3:19 0.68 p<0.005
Acid 01:05 ± 4:56 0.43 p>0.1
Wild 01:41 ± 4:20 0.52 p<0.002
Freshwater (Site C)
Ambient 0:42 ± 3:22a
0.59 p<0.02
Acid 0:34 ± 2:47a
0.77 p<0.005
Wild 01:00 ± 3:05a
0.72 p<0.001
Upper Estuary (Site D)
Ambient 07:04 ± 4:25 a
0.51 p=0.055
Acid 07:06 ± 4:46 a
0.46 p>0.1
Wild 08:00 ± 6:22 a
0.25 p>0.2
Middle Estuary (Site E)
Ambient 19:23 ± 4:37 a
0.48 p>0.1
Acid 22:40 ± 6:31 a
0.23 p>0.5
Wild 03:03 ± 6:42 a
0.21 p>0.2
Lower Estuary (Sites F1, F2)
Ambient 23:00 ± 7:04 a
0.18 p>0.5
Acid 22:25 ± 7:05 a
0.18 p>0.5
Wild 06:40 ± 5:28 a
0.36 0.1 > p > 0.05

44
Table 6. Hours past high tide at which smolts left each section. Mean ± 1SE
given in hours and minutes. Rayleigh’s ‘r’ value is a measure of the length of the
mean vector on the unit circle. A value of 1.00 indicates no variance in mean
time. A probability value <0.05 indicates that the time after high tide at which
smolts passed each receiver was nonrandom with respect to time of day.
Treatment Mean time leaving section ±±±± 1SD Rayleigh’s r value Probability
Freshwater
Ambient 5:08 ± 4:55ab
0.052 p>0.5
Acid 8:24 ± 4:06a
0.13 p>0.5
Wild 1:12 ± 3:50b
0.17 p>0.5
Upper Estuary
0.81 p<0.001
Acid 3:52 ± 2:02a
0.60 p=0.5
Wild 2:25 ± 1:55a
0.64 p<0.001
Middle Estuary
0.65 0.05 > p > 0.02
Acid 4:45 ± 1:20a
0.80 p<0.01
Wild 4:01 ± 1:50a
0.67 p=0.001
Lower Estuary
0.56 0.1 > p > 0.05
Acid 3:05 ± 1:43a
0.70 p<0.01
Wild 4:45 ± 2:06a
0.58 p<0.005

45
Figure 13. Time of day and time after high tide at which smolts left the upper estuary.
Smolts that did not return (o) and those that returned (•) to site C after having left
the upper estuary are indicated. Large arrow and bar indicate mean time and
standard deviation of movement. Small arrow in graph indicates time of high
water.

46

47
Figure 14. Time of day and time after high tide at which smolts left the middle
estuary. Smolts that did not return (o) and those that returned (•) to site C
after having left the upper estuary are indicated. Large arrow and bar
indicate mean time and standard deviation of movement. Small arrow in
graph indicates time of high water.

48

49
Figure 15. Time of day and time after high tide at which smolts left the lower estuary.
Smolts that did not return (o) and those that returned (•) to site C after having left
the upper estuary are indicated. Large arrow and bar indicate mean time and
standard deviation of movement. Small arrow in graph indicates time of high
water.

50

51
Once in tidal water, the majority of smolts in all treatments moved
downstream during hours of low light intensity and with an ebb tide. The time of
movement with respect to time of day and tidal cycle was similar among
treatments in all river sections (Figures 13–15; Tables 5-6). Few hatchery, but
many wild smolts moved during the brightest hours of the day while in tidal
sections (Figures 13-15). Those smolts that left the three estuarine sections
during bright daylight hours (approximately 1000-1700 hours) did so on random
days (i.e., 14 May to 19 May), and there seemed to be no correlation between
the time of day a smolt left adjacent sections. Smolts often left one section at
night and then departed the next section during the day and vice-versa.
Smolts moved downstream with low salinity water. Smolts passed
through the upper and middle estuary primarily between two and six hours after
high tide. The salinity at this stage of the tidal cycle was lower (generally <10‰)
and temperature was higher (>13ºC) than at high tide, indicating that smolts were
moving in a section of water more closely resembling FW than SW (Figure 16a).
Once in the lower estuary, the smolts were subjected to much higher, more
stable salinities and lower temperatures (Figure 16b). Because salinity was
always high (>25‰) at the E array of detectors, the smolts must have
encountered high salinity water as they moved out of the middle estuary.
The downstream migration of smolts was not always a continuous
process. Smolts in all treatments often passed through multiple sections in one
night or tidal cycle and then remained in a given section for up to 3 days before
resuming migration. Many smolts were able to pass through multiple river

52
Figure 16a. Temperature and salinity recorded at the middle of the middle
estuary from 10 May to 22 May 1998.
Figure 16b. Temperature and salinity recorded at the E Array from 10 May to 22
May 1998.

54
sections in one ebb tide, but some smolts returned upstream to near the head of
tide shortly after entering the middle estuary. One ambient (11%), three acid-
exposed (33%), and seven wild (29%) smolts made upstream movements at this
time. Although the upstream movement of these smolts was random in terms of
time of day (Table 7a), it was significantly related to the tidal cycle (Table 7b).
Smolts that returned upstream to site C did so with a flood tide (mean time =
10:27 ± 1:47 hours after high tide). When these smolts successfully left the
upper estuary, the time of movement was non-random with respect to both time
of day and tidal cycle (Figure 17; Tables 7a,b). Smolts returned downstream
approximately 1.5 hours before sunset (mean time = 1825 hours). This occurred
1-5 hours after high tide on these days, and these smolts migrated with the ebb
tide. In all cases, the smolts moved back downstream with the next ebb tide
regardless of the time of day. Those movements in which smolts were
unsuccessful at leaving the upper estuary were random with respect to time of
day and time of tide (Figure 17; Table 7b).
Survival during migration
The model predicted using program Mark was that of equal survival
among all three treatments, unequal survival rate among river sections, and
equal detection rate (>85%) for all treatments in all river sections. Survival was
lowest in the middle estuary, where salinity and temperature fluctuated the most
(Figure 16a; Table 8). Survival was greatest in the lower estuary, with all smolts

55
entering this section surviving to leave it. Salinity and temperature varied less in
this section than in any other tidal section (Figure 16b).

56
Table 7a. Time of day that smolts returned to Site C, and either successfully or
unsuccessfully left the upper estuary (passed Site D). Data are from smolts that
had already returned upstream to site C.
Time of Day
Action
Mean time ±±±± 1SD Rayleigh’s r value Probability
Return migration 18:23 ± 6:04 0.28 p>0.5
Successful exit 18:25 ± 2:10 0.85 p<0.001
Unsuccessful exit 06:10 ± 5:39 0.33 p>0.5
Table 7b. Time after high tide that smolts returned to Site C, and either
successfully or unsuccessfully left the upper estuary (passed Site D). Data are
from smolts that had already returned upstream to site C.
Time after High Tide
Action
Mean time ±±±± 1SD Rayleigh’s r value Probability
Return migration 10:27 ± 1:47 0.68 p<0.001
Successful exit 1:41 ± 1:07 0.86 p<0.001
Unsuccessful exit 4:14 ± 2:34 0.45 p>0.05

57
Figure 17. Time after high tide at which smolts successfully (A) or unsuccessfully
(B) left the upper estuary. Letters indicate ambient (A), acid-exposed (H),
and wild (W). Numbers indicate individual smolts. Large arrow and bar
indicate mean time and standard deviation of movement. Small arrow in
graphs indicates time of high water.

59
Table 8. Survival estimates of migrating wild and hatchery-reared smolts.
Estimates were calculated using Program Mark with the Cormack-Jolly-Seber-
model.
River Section Survival (%) Standard Error (%) Upper 95% CL Lower 95% CL
Freshwater 95.8 2.8 84.8 98.95
Upper estuary 93.4 3.6 81.6 97.9
Middle estuary 76.7 6.4 61.9 87
Lower estuary 100 0 100 100

60
Discussion
Physiology
Numerous studies have reported decreased Na+
/K+
ATPase activity,
decreased blood chloride, and increased hematocrit of pre-smolts and smolts
which had been exposed to acidic fresh water, with and without added LAl
(Saunders et al., 1983; Farmer et al., 1989; Pauwels, 1990; Staurnes et al.,
1993, 1996). The general trends of physiological effects from acid and Al are the
same in this study as found in previous studies, but the overall effects were not
as pronounced. Although pH values in this study were similar to those recorded
by Lacroix (1985) and Pauwels (1990), mortalities were high after only 24 hours
of exposure in the former study and mortality occurred after 21 days of exposure
in the latter study. In the present study, there were no mortalities due to acid/Al
exposure at any time in fresh water. Also, there were no differences in Na+
/K+
ATPase activities between any of the three hatchery smolt treatments. Saunders
et al. (1983) and Farmer et al. (1989) both reported reductions in Na+
/K+
ATPase
activity for fish undergoing the parr-smolt transition exposed to pH 4.2-5.0
(values lower than used in the present study) and lower concentrations of LAl
(15-37 µg L-1
). Low pH was probably more responsible than was Al for the
decreased enzyme activity in the two former studies. All hatchery smolts
survived the third SWCT even though the enzyme activities were as low as that
of parr (McCormick et al., 1995) and the acid smolts had been exposed to an

61
average of 140 µg LAl L-1
for 13 days. In previous studies, parr and/or smolts
with Na+
/K+
ATPase activities similar to those smolts in the third SWCT died
when exposed to SW for 24 hours (Saunders et al. 1983; Farmer, et al., 1989;
Staurnes et al., 1993).
Blood chloride levels for control and acid treatments in both fresh water
and seawater were higher than in other studies (Kroglund and Staurnes,
unpublished data; Pauwels, 1990; Staurnes et al., 1993, 1996; Lysfjord and
Staurnes, 1998). Although I moved smolts as quickly as possible from the study
site to the SWCT site, this activity may have caused an increase in blood
chloride. Such a response is common in stressed fish when in seawater
(Fletcher, 1975). Nevertheless, compared to ambient smolts, those in the acid
group on all three test dates had reduced and elevated blood chloride
concentrations when in fresh and seawater respectively. Smolts from the limed
tank in the second and third SWCT had higher blood chloride levels in SW than
ambient smolts, and a lower mean hematocrit in seawater in the third SWCT.
This seems contrary to knowledge of the effects of acidic water. It is possible
that these smolts were under some stress from the variance in pH caused by
liming the water with solid CaCO3.
The effect of acid and aluminum on hematocrit values in the present study
was nearly identical to that in other studies (Kroglund and Staurnes, unpublished
data; Farmer et al., 1989; Pauwels, 1990; Staurnes et al., 1993, 1996).
Hematocrit values of smolts held in acidic conditions increased in freshwater and
decreased when transferred to seawater. This indicates that smolts exposed to

62
low pH and high LAl conditions have reduced and increased plasma volume
while in fresh and seawater, respectively.
The lack of treatment effects on gill morphology indicates that conditions
in this study were not as stressful to the fish as have been previously reported.
In contrast to my study, other researchers have shown numerous morphological
effects of acid/aluminum stress, including lamellar thickening (Tietge et al.,
1988), loss of surface microridges (Jagoe and Haines, 1983), and fusing of
secondary filaments (Hamilton and Haines, 1995) in fish exposed to similar pH
values but lower LAl concentrations. Smolts sampled from the limed treatment
on 21 and 22 May had noticeably more chloride cells than in other treatments
from any date, and may indicate better physiological preparedness for seawater
survival. However, more chloride cells implies there is a higher Na+
/K+
ATPase
activity, which these smolts did not have. Several studies have reported
increased numbers of chloride cells with exposure to acid/aluminum conditions
(Karlsson-Norrgren et al., 1986a, 1986b; Ingersoll et al., 1990; Jagoe and
Haines, 1990, 1997) and it has been suggested that this occurs to compensate
for acid-induced ion losses (Jagoe and Haines, 1997). The smolts in the limed
treatment may have been responding to the stress of widely fluctuating pH.
Although wild smolts had Na+
/K+
ATPase activities that were at least 32%
greater than hatchery smolts on 16 May, they displayed the least amount of
hyposomoregulatory ability of any treatment. The effects were not statistically
significantly different from ambient smolts, but those wild smolts that survived
had the lowest blood chloride concentration in freshwater and the largest

63
increase (70%) when exposed to seawater. They also had the highest fresh
water and lowest seawater hematocrit, respectively. These conditions are
characteristic of smolts with impaired osmoregulatory ability (Kroglund and
Staurnes, unpublished data; Lacroix, 1985; Farmer et al., 1989; Pauwels, 1990;
Staurnes et al., 1993, 1995, 1996). In light of this, it is not surprising that three of
the wild smolts either died or were close to death after only 24 hours in seawater.
Based on other studies, it is unlikely that these smolts would have been able to
tolerate seawater for much longer.
The specific reasons for the compromised physiology of wild fish are not
known. In the Narraguagus River, Atlantic salmon are periodically exposed to
low pH and high LAl concentrations, but the pH in May 1998 was always above
6.0. However, LAl concentrations were 125 to 180 µg L-1
from 8 to 11 May, the
time at which those wild smolts used in the second SWCT were most likely
migrating towards, but had not yet reached, the smolt traps. The LAl from early
March to 7 May was always lower than 130 µg L-1
. Salmonids can acclimate to
high LAl, but not to low pH (Wood et al., 1988a, 1988b; Mueller et al., 1991; Reid
et al., 1991). Acclimation to Al by brook trout (Salvelinus fontinalis) takes about
10 days (McDonald et al., 1991; Mueller et al., 1991). The wild smolts tested on
16 and 17 May may not have had enough time to acclimate to the high pulse of
LAl that occurred between the 8 and 11 May. If this were the case, one would
expect that the ambient smolts would have shown a similar response, which they
did not. It is possible that the pulses of low pH water during the early
smoltification process (February through early April) negatively affected that

64
process. Because the wild smolts had not been exposed to high LAl
concentrations for several weeks, they may have not been acclimated to the high
LAl concentrations from 8 to 11 May and this may have led to their compromised
physiology. Those wild smolts in the first SWCT were not exposed to high LAl
concentrations for several weeks and their physiology was uncompromised.
The effects of acidic water and elevated Al concentrations on smolt
physiology reported here are not unique, although the specific chemical
conditions that elicited effects are slightly different from other studies. Studies
from Norway have shown that effects (decreased Na+
/K+
ATPase activity,
decreased blood chloride, and increased hematocrit) are elicited at pH<6.2 and
15-20 µg LAl L-1
(Kroglund and Staurnes, unpublished data; Staurnes et al.,
1993, 1996). However, many studies in North America have revealed that much
lower pH values and higher LAl concentrations are needed before smolts are
under osmoregulatory stress (Farmer et al., 1989; Pauwels, 1990; this study).
There are a few likely reasons for this. First, some North American studies have
used water with calcium concentrations of 75-129 µeq L-1
(Pauwels, 1990; this
study). Calcium may provide protection against Al toxicity (Booth et al., 1988;
Cleveland et al., 1991). Second, flouride concentrations are typically much
higher in North America than in Norway (Henricksen et al., 1984), with much of
the Al forming inorganic complexes with it. Aluminum in this form is much less
toxic (Rosseland and Skogheim, 1984) especially at <100 µg F L-1
(Hamilton and
Haines, 1995); the flouride concentration in this study was generally 60–90 µg

65
L-1
. Third, C concentrations in North American studies have been high, usually
above 6 mg C L-1
(Lacroix, 1985; Pauwels, 1990; Farmer et al., 1989; this study),
whereas rivers in Norwegian studies have generally been below 2 mg DOC L-1
(Kroglund and Staurnes, unpublished data). The toxic effects of aluminum can
be ameliorated by complexation with organic acids (Peterson et al., 1989).
Fourth, in addition to the differences in chemistry, the possibility of genetic
factors also exists. Adult salmon generally return to their natal river to spawn
and transoceanic straying has never been reported. Therefore, it is possible that
populations have adapted to their unique chemical environments. Northern
European stocks of Atlantic salmon may simply be less tolerant of low pH and
elevated LAl.
Migratory behavior and residence time
Mean residence time of wild smolts in FW was almost four and three times
that of ambient and acid smolts, respectively. This may be related to handling or
surgical stress; the smolt is the most delicate life stage with respect to handling
(Carey and McCormick, 1998). Because hatchery smolts have been raised in
relatively small volumes with much handling over the course of their lives, they
may be less prone to physiological stress from handling and surgery than are
wild smolts. However, 12 of the 26 wild smolts left the fresh water section in the
first night after surgery, thus suggesting that they did not respond differently to
handling and surgery stress than the hatchery smolts. The fact that the river

66
discharge dropped markedly during the release period of the wild smolts (10 May
to 16 May) may be important. Migration speed has been shown in one study to
be positively correlated to river discharge in the Penobscot River, Maine (Fried et
al., 1978), although other studies have shown no such relationship for both wild
and hatchery-reared smolts (Moore et al., 1995, 1998; Spicer et al., 1995). The
days that wild and hatchery smolts were migrating in FW overlapped, with only a
few wild smolts migrating one day later than hatchery-reared ones. The relation
of river discharge to migration speed has not previously been investigated in the
Narraguagus River.
A number of studies have been completed on Atlantic salmon smolt
migratory behavior (e.g., McCleave, 1978; Hansen and Jonsson, 1985; Moore et
al., 1995; Spicer et al., 1995; Lacroix and McCurdy, 1996; Moore et al., 1998)
and, for the most part, the results reported here are similar. Smolts move
downstream primarily at night and, once in the estuary, travel with an ebb tide.
However, hatchery-reared smolts in this study moved past receivers almost
entirely at night in freshwater. They moved during hours of low or no light when
in the estuary and when subjected to the tidal cycle (which had at least one ebb
tide during daylight hours). This is in contrast with the results of McCleave
(1978), Hansen and Jonsson (1985), Spicer et al., (1992), Lacroix and McCurdy
(1996), and preliminary results of work on the Narraguagus River in 1997 (John
Kocik, National Marine Fisheries Service, Woods Hole, Mass., personal
communication) which found that hatchery-reared smolts often traveled during
the day. Migrating at night may have originated to avoid the increased chance of

67
predation during the day (Solomon, 1982), but it is unclear why hatchery-reared
smolts in my study moved at times different than those documented in other
studies. One explanation may be that, in a few of the previous studies, smolts
were not kept in ambient conditions of the river for more than a few hours, or not
at all, prior to release (Fried et al., 1978; Tyler et al., 1978; Hansen and Jonsson,
1985). In the present study, hatchery-reared smolts were held in Narraguagus
River water under ambient temperature and light for 7-12 days before release.
The majority of smolts in all treatments passed detectors during an ebb
tide indicating that they take advantage of the strong ‘downstream’ tidal flow
(McCleave, 1978; Moore et al., 1995; Lacroix and McCurdy, 1996; Moore et al.,
1998). Those smolts making large upstream movements from the middle to
upper estuary entered the middle estuary an average of only one hour later in the
day which was not significantly different from those smolts that did not make
large upstream movements. Lacroix and McCurdy (1996) reported that smolts
that moved back upstream through a channel did so after leaving the channel
during a flood tide.
The fact that more wild and acid-exposed smolts made large upstream
movements from the middle to upper estuary than did ambient smolts is
important. The lower part of the upper estuary section is where the smolts first
encounter water of >8‰, and this is the zone of complex aluminum chemistry,
which can be even more toxic than in freshwater (Rosseland et al., 1992).
Assuming that wild smolts were physiologically compromised (as in the second

68
SWCT), their movements upstream at this time may have been in search of a
less physiologically stressful environments (lower salinity). Assuming high
predation rates for smolts in estuarine areas (Larsson, 1985; Dube and Godin,
1987; Hvidsten and Mokkelgjerd, 1987; Hvidsten and Lund, 1988; Blackwell et
al., 1997), it is important for the smolts to minimize the time spent there.
Therefore, wild and acid smolts may have been more prone to predation because
they spent more time in the upper and middle estuary. Handeland et al. (1996)
and Jarvi (1989, 1990) found that predation rates were higher on smolts suffering
from osmoregulatory stress after transfer to SW. Predation rates were higher
because of a change in antipredator behavior (smolts schooled less frequently
and allowed predators to get closer before fleeing).
It is interesting to note that all smolts that made repeated movements
upstream in tidal water survived to leave the lower estuarine area thus
suggesting that the survival rate was no different, if not higher, for them.
However, wild and ambient-reared smolts that did not make upstream
movements also tended to spend an extended period in estuarine sections of the
river.
Because I was limited in the number of active searches, I cannot
determine how many smolts made upstream movements that were shorter than
the distance between detectors stationed in adjacent sections. Other studies
have shown that smolts do indeed make numerous, short upstream and
downstream movements following the flow of tidal water (Fried et al., 1978;
LaBar et al., 1978; McCleave, 1978; Tyler et al., 1978; Moore et al., 1998). The

69
fact that only one smolt moved from the upper to the lower estuary in only one
tidal cycle (a distance of 9 km) suggests that they made many, short upstream
movements or simply may have paused for long periods of time. In fact, smolts
sometimes remained in a given river section for multiple tides, then passing
through multiple sections in only one tide.
The ocean survival of Atlantic salmon varies widely by year, latitude, and
river (Bley and Moring, 1988). The specific causes of large-scale, high-seas
mortality are unknown, although the size of winter habitat (based on temperature)
in the North Atlantic has been strongly correlated with the number of adult returns
to rivers across eastern North America in the same year (Friedland et al., 1993).
The coherence in return rate of different stocks of salmon suggests that the
factors that contribute to high-sea mortality act equally on the stocks (Friedland,
1995, 1998). That is, mortality of the stocks likely occurs at temporally and
spatially equal times and locations.
Ocean temperature at the time smolts enter seawater may also have an
effect on the survival of salmon stocks. Friedland (1998) found a strong
correlation between survival and the area of 8-10ºC sea surface temperatures in
May for a Norwegian and for a Scottish stock. In years when cool surface waters
were dominate on the Norwegian coast, salmon survival was poor. When the
sea surface temperatures were above 8ºC, survival was good. Friedland (1998)
suggested that water temperature below 8ºC depresses growth (by both
decreasing metabolic rate and aquatic productivity) and leads to a higher
predation rate.

70
The effect of acid and aluminum on the imprinting process has not yet
been investigated in Atlantic salmon. Atlantic salmon imprint on the native river
during or just prior to smoltification (Morin et. al., 1997), a time at which their
olfactory senses are enhanced (Morin et. al., 1989; Morin and Doving, 1992;
Morin et. al., 1997). Moore (1994) reported that pH<5.5 decreased the olfactory
responses of mature male Atlantic salmon parr to the presence of testosterone
and the presence of urine from ovulated females. Low pH (4.7) and 135 µg total
Al L-1
disrupted the olfactory organs of rainbow trout (Oncorhynchus mykiss) and
resulted in a loss of receptor cell cilia (Klaprat et. al., 1988). Low pH has also
been shown to inhibit the olfactory response on rainbow trout to fifteen amino
acids to which the olfactory bulbs are sensitive (Hara, 1976). Therefore, it is
possible that the acid/aluminum conditions common in Maine rivers may be
sufficient to disrupt the imprinting process of Atlantic salmon smolts, ultimately
decreasing the number of adults returning to spawn. In addition, the pesticide,
Diazinon, has also been shown to inhibit the olfactory response of mature
Atlantic salmon parr (Moore and Waring, 1996). The Narraguagus River is
bordered by commercial blueberry agriculture on which numerous herbicides,
pesticides, and fungicides are applied. The effect of these chemicals on the
physiology and imprinting process in Atlantic salmon is unknown.

71
Conclusions
Based on data in the present study, it seems that the acidic and LAl
conditions that I used (common in rivers in eastern Maine in the spring) had a
slightly negative effect on the physiology but not the migratory behavior nor the
survival of smolts during their seaward migration. Wild smolts in the second
SWCT were physiologically compromised, and the migratory behavior of wild
smolts during this time was slightly different from the hatchery ambient and
acid-exposed treatments only in the time spent in the four river sections, although
more wild smolts moved during the brightest daylight hours. Seaward migratory
survival of all three treatments was equal, suggesting that the low adult return
rates are not likely to be due to acid and aluminum toxicity.

72
References
Anthony, V.C. 1995. The significance of predation on Atlantic salmon. p. 240-
284 In: A hard look at some tough issues. Proceedings of the New
England Atlantic Salmon Management Conference 1994. Eds. Silvio
Calabi and Andrew Stout. New England Salmon Association,
Newburyport, MA.
Batschelet, E. 1981. Circular statistics in biology. Eds. Robin Sibson and Joel E.
Cohen. Academic Press, New York.
Beland, K.F., and N.R. Dube. 1998. Atlantic salmon research addressing issues
of concern to the National Marine Fisheries Service and the Maine Atlantic
Salmon Authority. Final Project Report, Grant NA57FL0149. Maine
Atlantic Salmon Authority, Bangor, Maine, USA.
Blackwell, B.F. and W. B. Krohn. 1997. Spring foraging distribution and habitat
selection by double-crested cormorants on the Penobscot River, Maine
USA. Colonial Waterbirds 20:66-76.
Blackwell, B.F., W.B. Krohn, N.R.Dube, and A.J.Godin. 1997. Spring prey use
by double-crested cormorants on the Penobscot River, Maine, USA.
Colonial Waterbirds 20:77-86.
Bley, P.W., and J.R. Moring. 1988. Freshwater and ocean survival of Atlantic
salmon and steelhead: A synopsis. United States Fish and Wildlife
Service. Biological Report 88(9).
Booth, C.E., D.G. McDonald, B.P. Simons, and C.M. Wood. 1988. Effects of
aluminum and low pH on net ion fluxes and ion balance in the brook trout
(Salvelinus fontinalis). Can. J. Fish. Aquat. Sci. 45:1563-1574.
Carey, J.B., and S.D. McCormick. 1998. Atlantic salmon smolts are more
responsive to an acute handling and confinement stress than parr.
Aquaculture 168:237-253.
Clarke, W.C. 1982. Evaluation of the seawater challenge test as an index of
marine survival. Aquaculture 28:177-183.

73
Cleveland, L., E.E. Little, C.G Ingersoll, R.H. Wiedmeyer, and J.B. Hunn. 1991.
Sensitivity of brook trout to low pH, low calcium and elevated aluminum
concentrations during laboratory pulse exposures. Aquat. Toxicol. 19:303-
318.
Cormack, R. M. 1992. Interval estimation for mark-recapture studies of closed
populations. Biometrics 48:567-576.
Daye, P.G., and E.T. Garside. 1976. Histopathologic changes in surficial tissues
of brook trout, Salvelinus fontinalis, from acidified lakes. Can. J. Zool.
63:2062-2070.
Driscoll, C.T., J.P Baker, J.J. Bisogni Jr., and C.L. Schofield. 1980. Effects of
aluminum speciation on fish in dilute acidified waters. Nature 284:161-164.
Dube, N.R., and A.J.Godin. 1987. Evaluation of double-crested cormorant
predation upon hatchery-reared Atlantic salmon smolts released in the
Penobscot River, Maine. International Council for the Exploration of the
Sea, Working Paper, Working Group on North Atlantic Salmon.
Farmer, G.J., R.L Saunders, T.R. Goff, C.E.Johnston, and E.B. Henderson.
1989. Some physiological responses of Atlantic salmon (Salmo salar)
exposed to soft, acidic water during smolting. Aquaculture 82:229-244.
Fletcher, G.L. 1975. The effects of capture “stress” and storage of whole blood
on red blood cells, plasma proteins, glucose and electrloytes of the winter
flounder (Pseudopleuronectes americanus). Can. J.Zool. 53:197-206.
Fried, S.M., J.D. McCleave, and G.W. LaBar. 1978. Seaward migration of
hatchery-reared Atlantic salmon, Salmo salar, smolts in the Penobscot
River Estuary, Maine: riverine movements. J.Fish.Res. Board Can. 35:76-
87.
Friedland, K.D. 1995. Marine survival of restoration stocks. p. 223-239 In: A
hard look at some tough issues. Proceedings of the New England Atlantic
Salmon Management Conference 1994. Eds. Silvio Calabi and Andrew
Stout. Published by New England Salmon Association, Newburyport, MA.
Friedland, K.D. 1998. Ocean climate influences on critical Atlantic salmon
(Salmo salar) life history events. Can. J. Fish. Aquat. Sci. 55(Suppl.
1):119-130.
Friedland, K.D., D.G. Reddin, and J.F. Kocik. 1993. Marine survival of North
American and European Atlantic salmon: effects of growth and
environment. ICES. J. Mar. Sci. 50:481-492.

74
Haines, T.A. 1987. Atlantic salmon resources in the northeastern United States
and the potential effects of acidification from atmospheric deposition.
Water Air Soil Poll. 35:37-48.
Haines, T.A., S.A. Norton, J.S. Kahl, C.W. Fay, and S.J. Pauwels. 1990.
Intensive studies of stream fish populations in Maine. Ecological Research
Series. U.S. Environmental Protection Agency. Washington, D.C. 354 pp.
Hamilton, S.J., and T.A. Haines. 1995. Influence of flouride on aluminum toxicity
to Atlantic salmon (Salmo salar). Can. J. Fish. Aquat. Sci. 52:2432-2444.
Handeland, S.O., T. Jarvi, A. Ferno, and S.O. Stefansson. 1996. Osmotic stress,
antipredator behaviour, and mortality of Atlantic salmon (Salmo salar)
smolts. Can. J. Fish. Aquat. Sci. 53:2673-2680.
Hansen, L.P., and B. Jonsson. 1985. Downstream migration of hatchery-reared
smolts of Atlantic salmon (Salmo salar L.) in the River Imsa, Norway.
Hara, T.J. 1976. Effects of pH on the olfactory responses to amino acids in
rainbow trout, Salmo gairdneri. Comp. Biochem. Physiol. 54A:37-39.
Henricksen, A., O.K. Skogheim, and B.O. Rosseland. 1984. Episodic changes in
pH and aluminum-speciation kill fish in a Norwegian salmon river.
Vatten/Water 40(2):255-260.
Hvidsten, N.A., and R.A. Lund. 1988. Predation on hatchery-reared and wild
smolts of Atlantic salmon, Salmo salar L., in the estuary of River Orkla,
Norway. J. Fish Biol. 33:121-126.
Hvidsten, N.A., and P.I. Mokkelgjerd. 1987. Predation of salmon smolts, Salmo
salar L., in the estuary of the River Surna, Norway. J. Fish Biol. 30:273-
280.
Ingersoll, C.G., D.D. Gulley, D.R. Mount, M.E. Mueller, J.D. Fernandez, J.R.
Hockett, and H.L. Bergman. 1990. Aluminum and acid toxicity to two
strains of brook trout (Salvelinus fontinalis). Can. J. Fish. Aquat. Sci.
47:2451-2460.
Jagoe, C.H., and T.A. Haines. 1983. Alterations in gill epithelial morphology of
yearling Sunapee trout exposed to acute acid stress. Trans. Amer. Fish.
Soc. 112:689-695.
Jagoe, C.H., and T.A. Haines. 1990. Morphometric effects of low pH and limed
water on the gills of Atlantic salmon (Salmo salar). Can. J. Fish. Aquat.
Sci. 47:2451-2460.

75
Jagoe, C.H., and T.A. Haines. 1997. Changes in gill morphology of Atlantic
salmon (Salmo salar) smolts due to addition of acid and aluminum to
stream water. Environ. Poll. 97:137-146.
Jarvi, T. 1989. Synergistic effect on mortality in Atlantic salmon, Salmo salar,
smolt caused by osmotic stress and presence of predators. Environ. Biol.
Fishes 26:149-152.
Jarvi, T. 1990. Cumulative acute physiological stress in Atlantic salmon smolts:
the effect of osmotic imbalance and the presence of predators.
Karlsson-Norrgren, L., I Bjorkland, O. Ljingberg, and P. Runn. 1986a. Acid water
and aluminum exposure: experimentally induced gill lesions in brown trout
Salmo trutta L. J. Fish Diseases 9:11-25.
Karlsson-Norrgren, L., W. Dickson, O. Ljingberg, and P. Runn. 1986b. Acid
water and aluminum exposure: gill lesions and aluminum accumulation in
farmed brown trout Salmo trutta L. J. Fish Diseases 9:1-9.
Klaprat, D.A., S.B. Brown, and T.J. Hara. 1988. The effect of low pH and
aluminum on the olfactory organ of rainbow trout, Salmo gairdneri.
Environ. Biol. Fishes 22:69-77.
Labar, G.W., J.D. McCleave, and S.M. Fried. 1978. Seaward migration of
hatchery-reared Atlantic salmon, Salmo salar, smolts in the Penobscot
River estuary, Maine: open water movements. J. Cons. Int. Exp. Mer.
38:257-269.
Lacroix, G.L. 1985. Plasma ionic composition of the Atlantic salmon (Salmo
salar), white sucker (Catostomus commersoni), and alewife (Alosa
pseudoharengus) in some acidic rivers of Nova Scotia. Can. J. Zool.
63:2254-2261.
Lacroix, G.L., and P. McCurdy. 1996. Migratory behaviour of post-smolt Atlantic
salmon during initial stages of seaward migration. J. Fish Biol. 49:1086-
1101.
Larsson, P.O. 1985. Predation on migrating smolts as a regulating factor in
Baltic salmon, Salmo salar L., populations. J. Fish Biol. 26:391-397.

76
Lysfjord, G and M. Staurnes. 1998. Gill Na+
/K+
-ATPase activity and
hypoosmoregulatory ability of seaward migrating smolts of anadromous
Atlantic salmon (Salmo salar), sea trout (Salmo trutta) and Arctic charr
(Salvelinus alpinus) in the Hals River, northern Norway. Aquaculture
168:279-288.
Mardia, K.V. 1972. Statistics of directional data. Academic Press, London and
New York.
McCleave, J.D. 1978. Rhythmic aspects of estuarine migration of hatchery-
reared Atlantic salmon (Salmo salar) smolts. J. Fish Biol. 12:559-570.
McCormick, S.D. 1993. Methods for non-lethal gill biopsy and measurement of
Na+
/K+
-ATPase activity. Can. J. Fish. Aquat. Sci. 50:656-658.
McCormick, S.D., and B. Bjornsson. 1994. Physiological and hormonal
differences among Atlantic salmon parr and smolts reared in the wild and
hatchery smolts. Aquaculture 121:235-244.
McCormick, S.D., B. Bjornsson, M. Sheridan, C. Eilertson, J.B. Carey, and M.
O’Dea. 1995. Increased daylength stimulates plasma growth hormone
and gill Na+
/K+
-ATPase in Atlantic salmon (Salmo salar). J. Comp. Phys.
165:245-254.
McCormick, S.D., R.J. Naiman, and E.T. Montgomery. 1985. Physiological smolt
characteristics of anadromous and non-anadromous brook trout
(Salvelinus fontinalis) and Atlantic salmon (Salmo salar). Can. J. Fish.
Aquat. Sci. 42:529-538.
McDonald, D.G., C.M. Wood, R.G. Rhem, M.E. Mueller, D.R. Mount, and H.L.
Bergman. 1991. Nature and time course of acclimation to aluminum in
juvenile brook trout (Salvelinus fontinalis). I. Physiology. Can. J. Fish.
Aquat. Sci. 48:2006-2015.
Moore, A. 1994. An electrophysiological study on the effects of pH on olfaction in
mature male Atlantic salmon (Salmo salar) parr. J. Fish Biol. 45:493-502.
Moore, A., S. Ives, T.A. Mead, and L. Talks. 1998. The migratory behaviour of
wild Atlantic salmon (Salmo salar L.) smolts in the River Test and
Southampton Water, southern England. Hydrobiologia 371/372:295-304.
Moore, A., E.C.E. Potter, N.J. Milner, and S. Bamber. 1995. The migratory
behaviour of wild Atlantic salmon (Salmo salar) smolts in the estuary of
the River Conway, North Wales. Can. J. Fish. Aquat. Sci. 52:1923-1935.

77
Moore, A., I.C. Russell, and E.C.E. Potter. 1990. The effects of interperitoneally
implanted dummy acoustic transmitters on the behaviour and physiology
of juvenile Atlantic salmon, Salmo salar L. J. Fish Biol. 37:713-721.
Moore, A., and C.P. Waring. 1996. Sublethal effects of the pesticide Diazinon on
olfactory function in mature male Atlantic salmon parr. J. Fish Biol. 48:758-
775.
Morin, P.-P, J.J. Dodson, and F.Y. Dore. 1989. Cardiac responses to a natural
odorant as evidence of a sensitive period for olfactory imprinting in young
Atlantic salmon, Salmo salar. Can. J. Fish. Aquat. Sci. 46:122-130.
Morin, P.-P and K.B. Doving. 1992. Changes in the olfactory function of Atlantic
salmon, Salmo salar, in the course of smoltification. Can. J. Fish. Aquat.
Sci. 49:1704-1713.
Morin, P.-P., T.J. Hara, and J.G. Eales. 1997. Thyroid function and olfactory
responses to L-alanine during induced smoltification in Atlantic salmon,
Salmo salar. Can. J. Fish. Aquat. Sci. 54:596-602.
Moring, J.R. 1987. Restoration of Atlantic salmon to Maine: overcoming physical
and biological problems in the estuary. p. 3129-3140 In: Proceedings of
Ocean Zone 1987, the fifth symposium on coastal and ocean
management. Eds. O.T. Magoon, H. Converse, D. Miner, L.T.Tobin, D.
Clark, and G. Domurat. American Society of Civil Engineers, New York.
Mueller, M.E., D.A. Sanchex, H.L. Bergman, D.G. McDonald, R.G. Rhem, and
C.M. Wood. 1991. Nature and time course of acclimation to aluminum in
juvenile brook trout (Salvelinus fontinalis). II. Gill histology. Can. J. Fish.
Aquat. Sci. 48:2016-2027.
Peterson, R.H. Bourbonniere, R.A. Lacroix, G.L. Martin-Robichaud, D.J. Takats,
P. Brun, G. 1989. Responses of Atlantic salmon (Salmo salar) alevins to
dissolved organic carbon and dissolved aluminum at low pH. Water Air
Soil Poll. 46:399-413.
Pauwels, S. 1990. Some effects of exposure to acid and aluminum on several life
stages of the Atlantic Salmon (Salmo salar). Ph.D. dissertation, thesis,
University of Maine, Orono, Maine, USA.
Reid, S.D., D.G. McDonald, and R.R. Rhem. 1991. Acclimation of sublethal
aluminum: Modifications of metal-gill surface interactions of juvenile
rainbow trout (Oncorhynchus mykiss). Can. J. Fish. Aquat. Sci. 48:1996-
2005.

78
Rosseland, B.O., I.A. Blakar, A. Bulger, F. Kroglund, A. Kvellstad, E. Lydersen,
D.H. Oughton, B. Salbu, M. Staurnes, and R. Vogt. 1992. The mixing
zone between limed and acidic river waters: complex aluminum chemistry
and extreme toxicity for salmonids. Environ. Poll. 78:3-8.
Rosseland, B.O. and O.K. Skogheim. 1984. A comparative study on salmonid
fish species in acid, aluminium-rich water. 2. Physiological stress and
mortality of one- and two-year-old fish. Rep. Inst. Freshwat. Res.,
Drottningholm 61:187-194.
SAS Institute. 1990. SAS Institute, Inc., Cary, NC
Saunders, R.L., E.B. Henderson, P.R. Harmon, C.E. Johnston, and J.G. Eales..
1983. Effects of low environmental pH on smolting of Atlantic salmon
(Salmo salar). Can. J. Fish. Aquat. Sci. 40:1203-1211.
Skogheim, O.K. and B.O. Rosseland. 1986. Mortality of smolt of Atlantic salmon,
Salmo salar L., at low levels of aluminum in acidic softwater. Bull. Environ.
Contam. Toxicol. 37:258-265.
Smith, T.R., and T.A. Haines. 1995. Mortality, growth, swimming activity and gill
morphology of brook trout (Salvelinus fontinalis) and Atlantic salmon
(Salmo salar) exposed to low pH with and without aluminum. Environ. Poll.
90:33-40.
Solomon, D.J. 1982. Smolt migration in Atlantic salmon (Salmo salar L.) and sea
trout (Salmo trutta L.) p. 196-203 In: Salmon and trout migratory
Behaviour Symposium, Eds. E.L. Brannon and E.O. Salo. University of
Washington, Seattle, WA.
Spicer, A.V., J.R. Moring, and J.G. Trial. 1995. Downstream migratory behavior
of hatchery-reared. radio-tagged Atlantic salmon (Salmo salar) smolts in
the Penobscot River, Maine USA. Fish. Res. 23:255-266.
Staurnes, M., P. Blix, and O.B. Reite. 1993. Effects of acid water and aluminum
on parr-smolt transformation and seawater tolerance in Atlantic salmon,
Salmo salar. Can. J. Fish. Aquat. Sci. 50:1816-1827.
Staurnes, M., L.P. Hansen, K. Fugelli, and O. Haraldstad. 1996. Short-term
exposure to acid water impairs osmoregulation, seawater tolerance, and
subsequent marine survival of smolts of Atlantic salmon (Salmo salar).
Can. J. Fish. Aquat. Sci. 53:1695-1704.
Staurnes, M., F. Kroglund, and B.O. Rosseland. 1995. Water quality requirement
of Atlantic salmon (Salmo salar) in water undergoing acidification or liming
in Norway. Water Air Soil Poll. 85:347-352.

79
Thedinga, J.F., M.L. Murphy, S.W. Johnson, J.M. Lorenz, and K.V. Koski. 1994.
Determination of salmonid smolt yield with rotary-screw traps in the Situk
River, Alaska, to predict effects of glacial flooding. North Amer. J. Fish.
Manage. 14:837-851.
Tietge, J.E., R.D. Johnson, and H.L Bergman. 1988. Morphometric changes in
gill secondary lamellae of brook trout (Salvelinus fontinalis) after long-term
exposure to acid and aluminum. Can. J. Fish. Aquat. Sci. 45:1643-1648.
Tyler, P., J.E. Thorpe, and W.M. Shearer. 1978. Ultrasonic tracking of the
movements of Atlantic salmon smolts (Salmo salar L.) in the estuaries of
two Scottish rivers. J. Fish Biol. 12:575-586.
USEPA. 1987. Handbook of methods for acid deposition studies: laboratory
analysis for surface water chemistry. USEPA 600/4-87/026. Sept. 1987.
Voegeli, F.A., G.L. Lacroix, and J.M. Anderson. 1998. Development of minature
pingers for tracking Atlantic salmon smolts at sea. Hydrobiologia
371/372:35-46.
Walker, R.L., C.M Wood, and H.L. Bergman. 1988. Effects of low pH and
aluminum on ventilation in the brook trout (Salvelinus fontinalis). Can. J.
Fish. Aquat. Sci. 45:1614-1622.
Westall, J.C., J.L. Zachary, and F.M.M. Morel. 1976. Mineql, a computer program
for the calculation of chemical equilibrium composition of aqueous
systems. Civil Engineering, Massachusetts Institute of Technology.
White, G. 1998. Program Mark.
http://www.cnr.colostate.edu/~gwhite/mark/mark.htm (16 Jan 1999).
Wood, C.M., R.C. Playle, B.P. Simons, G.G. Goss, and D.G. McDonald. 1988a.
Blood gases, acid-base status, ions, and hematology in adult brook trout
(Salvelinus fontinalis) under acid-aluminum exposure. Can. J. Fish. Aquat.
Sci. 45:1575-1586.
Wood, C.M., B.P. Simons, D.R. Mount, and H.L. Bergman. 1988b. Physiological
evidence of acclimation to acid/aluminum stress in adult brook trout
(Salvelinus fontinalis). 2. Blood parameters by cannulation. Can. J. Fish.
Aquat. Sci. 45:1597-1605.

80
Appendix
Water chemistry of Green Lake National Fish Hatchery, Ellsworth, Maine, USA.
ANC = Acid neutralizing capacity. DOC = dissolved organic carbon.
11 March 1998 6 May 1998
pH 6.4 6.7
Ca++
(µeq/L) 45.75
Mg++
(µeq/L) 19.34
Na+
(µeq/L) 96.52
K+
(µeq/L) 9.49
Cl-
(µeq/L) 99.8
NO3
-
(µeq/L) 2.56
SO4
=
(µeq/L) 64.7
F-
(µeq/L) 3.59
ANC (µeq/L) 73
DOC(mg/L) 3.31
Al (total) (µg/L) 179 82
LAl (µg/L) 147 46

81
Biography
John Andrew Magee was born in Baltimore, Maryland on September 29,
1971. He graduated from Centennial High School, Ellicott City, Maryland, in
1989. In 1986, his brother, Matt, gave him an aquarium as a present. Aquatic
ecology became a keen interest at this time.
He entered St. Mary’s College of Maryland in 1989 and graduated in 1993
concentrating on Aquatic Ecology. He began a rewarding career in aquatic
biology at the University of Maryland Chesapeake Biological Laboratory in May
1993. While there, he worked on projects investigating heavy metal toxicity on
freshwater and estuarine organisms, and long-term saltwater tolerance of zebra
mussels.
He was enrolled for graduate study in Zoology at the University of Maine
in September 1997 and served as both a teaching assistant for the Department
of Biological Sciences and a research assistant for the United States Geological
Survey, Biological Resources Division. He is a candidate for the Master of
Science degree in Zoology from the University of Maine in December, 1999.

MAGEE - DEC 1999

Recommended

Recommended

More Related Content

Similar to MAGEE - DEC 1999

Similar to MAGEE - DEC 1999 (20)

MAGEE - DEC 1999