CONTENTS
Page
Nitrogen Gas Bubble Disease Related to a Hatchery Water Supply from the
Forebay of a High-head
Re-regulating Dam ..............ELLIS J. W Y A T and KIRK T. BElNlNGEN
3
Estimates o f Maturation and Ocean Mortality for Columbia River Hatchery Fall
Chinook Salmon and the Effect of N o Ocean
E
N A . HENRY
Fishing on Yield ..................................................K
13
Investigation of Scale Patterns as a Means of lden tifying Races o f
Spring Chinook Salmon in the
Columbia River ................BURNELL R. BOHN and H O W A R D E. JENSEN
28
Growth of Juvenile Spring Chinook Salmon in Lookingglass
Creek ..........................................................................W A Y N E A . BURCK
37
Depth Distribution of Some Small Flatfishes o f f the Northern OregonSouthern Washington Coast ....................................ROBERT L. DEMORY
44
Nearshore Ocean Currents o f f Clatsop Beach, Oregon, and Their Relation t o the
Distribution of Razor Clam Larvae ......................DARRELL E. DEMORY
49
Algae Growth as a Cause o f Tag Loss from Juvenile Fall Chinook Salmon
in Sixes River Estuary, Oregon ........................................ PAUL E. RE1MERS
56
The Use o f Oxytetracycline Marks in Vertebra o f A d u l t Salmon to
Determine Smolt Size ......................................................D O N F. SWARTZ
59
A Note on the Coastal Movement o f Shad....................T. EDWI N CUMMINGS
60
NITROGEN GAS BUBBLE DISEASE RELATED TO A HATCHERY WATER SUPPLY
FROM THE FOREBAY OF A HIGH-HEAD RE-REGULATING DAM
Ellis J. Wyatt
and
Kirk T. Beiningen
Fish Commission o f Oregon, Hatchery Biology Section
Clackamas, Oregon
Abstract
Studies of the water supply system at the Fish Commission of Oregon's South Santiam River Hatchery were
made fallowing the loss of nearly a half million salmonid fishes due t o gas embolism. Hatchery water from the
forebay of adjacent Foster Dam had become supersaturated with dissolved nitrogen as the result of air entering
the supply system. Experiments revealed that fish were exposed to concentrations i n excess of 150% of saturation
and rapidly died from the effects of gas bubble diseaw. Biological assay data regarding supersaturation rates,
exposure times, death rates, and gross pathology are reported.
Introduction
D u r i n g the n i g h t o f September 5, 1969,
nearly a half m i l l i o n salmonid fingerlings
were lost a t t h e Fish Commission of Oregon's South Santiam Hatchery as a result
o f nitrogen (N2)gas embolism. T h e loss
comprised approximately 75% o f t h e station's spring chinook
(Oncorhynchus
tshawytscha Walbaum) , fingerlings a n d
99% of t h e summer steelhead (Salmo
gairdneri Richardson .
T h e hatchery water supply is delivered
b y gravity f l o w via t w o pipelines w h i c h
originate i n the forebay o f adjacent Foster Dam (Figure 1 ) . At the t i m e of t h e
loss, water f r o m t h e upper intake ( a t the
17-18°C epilimnion) and lower intake
( a t t h e 10-1 1 O C hypolimnion was being
mixed t o provide a mean water temperat u r e o f 13.5OC. T h e grate guarding t h e
upper intake aperture became partially
plugged w i t h submerged debris and,as t h e
forebay water level was lowered, the
blockage increased when surface debris
also lodged against t h e grating. Apparently water n o longer flowed smoothly i n t o
t h e upper pipeline b u t entered around
t h e periphery o f t h e grating a t a reduced
rate o f f l o w and became hiahlv aerated as
it tumbled i n t o t h e throat df t h e pipe.
It was suspected t h a t entrained air' was
forced i n t o solution a t a n accelerated rate
as the m i x t u r e f l o w e d t o a lower elevation
under increased pressure. D u r i n g t h e
morning o f September 5, t h e reduced f l o w
o f epilimnion water caused t h e headbox
water temperature t o drop f r o m 13.5 t o
10.5"C. T h e following morning hatchery
personnel observed dead and dying fish i n
all rearing ponds. Nearly all t h e fish examined showed subcutaneous air bubbles
symptomatic o f nitrogen gas bubble disease.
Previous investigators have reported o n
N, gas supersaturation in hatchery ponds
(Rucker and Hodgeboom, 1953; Harvey
and Smith, 1961 and natural environments (Harvey and Cooper, 1962; Ebel,
1 9 6 9 ) . T h i s paper reports o n experiments
a t the South Santiam Hatchery i n late
September and early October 1969 where
t h e conditions o f September 5 were simulated i n order t o develop criteria for safe
operation o f t h e water supply system.
W a t e r chemistry and bioassay results regarding supersaturation rates, exposure
times, death rates, and gross pathology
are given.
Materials and Methods
Experiments were conducted f r o m September 29 t o October 2 a t various reso-r-
[31
LEGEND
m
UATCUFRY
PONDS
Iucv = UPPER INTAKE CONTROL VALVE
I l c v = LOWER INTAKE CONTROL VALVE
Cvap- CONTROL VALVE ADULT POND
Figure 1. Simplified drawing of water intake system at Foster Dam.
voir elevations, f l o w rates, and under d i f ferent combinations o f intake valve settings. From the morning o f October 2 t o
the morning o f October 3, the reservoir
was a t 192.3 m elevation w h i c h equaled
the situation existing o n September 5. A
1 m by 1 m piece o f plywood was placed
over the grating t o simulate submergent
debris, and manipulation of mixing-valve
settings and f l o w rates established parameters w i t h i n w h i c h supersaturated water
conditions were achieved.
The upper pipeline, the center of w h i c h
is 192.0 m above mean sea level, is 0.8 m
i n diameter w i t h a 1 m bell flange a t i t s
opening i n t o the reservoir. W a t e r f r o m
the upper and lower pipelines merge a t
175.7 m elevation, and f l o w 259 m via
[4I
a single 5 0 . 8 c m diameter pipe t o a screeni n g box a t t h e hatchery (Figure 2 ) .A control valve regulates the rate o f f l o w f r o m
t h e screening box t o the hatchery buildi n g and 10 rearing ponds. W a t e r temperature is controlled by manipulation o f
b u t t e r f l y valves t o regulate f l o w rates o f
t h e upper a n d lower intake pipelines.
W a t e r samples were taken in t h e forebay o f Foster Dam near t h e upper intake
orifice, i n t h e screening box, and i n t h e
rearing ponds. Other samples were taken
i n t h e forebay a t t h e surface and near t h e
b o t t o m a t 175.7 m elevation (near t h e
lower intake o r i f i c e ) . Samples were collected w i t h a V a n Dorn water sampler.
Temperature o f each water sample was
recorded t o the nearest 0.1 "C. T h e samples were stored i n standard 3 0 0 m l glassstoppered bottles and immediately placed
on ice. Analyses were performed as soon
as possible t o m i n i m i z e the e f f e c t of
changes in water quality. Dissolved Np
concentrations were measured w i t h a V a n
Figure 2. Water supply system on hatchery grounds.
Slyke-Neil1 manometric blood gas analyzer modified for water determinations.
Saturations are believed t o be correct t o
w i t h i n 2 2 9 6 o f their true value. Dissolved
oxygen ( 0 2 ) values were obtained b y the
modified W i n k l e r technique.
Biological assays were conducted w i t h
1968-brood W i l l a m e t t e River spring c h i nook fingerlings, averaging 4 4 fish per
kilogram, from t h e M a r i o n Forks Hatchery. Prior t o testing, control and test fish
were maintained in live boxes in the adult
holding pond w h i c h is supplied w i t h water
via a separate line f r o m the lower intake
and therefore was n o t affected b y events
o n September 5 o r tests performed i n this
study. Bioassay fish were held in 175 x
8 4 c m wire-mesh cages, suspended i n 61
c m o f water i n four d i f f e r e n t rearing
ponds. On September 3 0 and October 1,
185 fish each were placed in test cages
i n Ponds 7 and 6, for 3 0 and 5 4 hours before being subjected t o 134 and 1 5 2 %
concentration o f N,. A t h i r d group, placed
i n Pond 9 (N=106), was immediately
subjected t o a 134% concentration, followed b y an increase t o 152% T h e
f o u r t h group, tested in Pond 8 (N=476),
was exposed immediately t o a 1 52% concentration. T h e levels used i n the tests
were determined b y the physical structure
o f t h e water supply system w h e n the conditions w h i c h caused mortality were simulated. Prior t o the tests the different
holding periods were used t o investigate
the influence o f handling stress on the
fish. Dead fish were removed a t 3 0 - m i n u t e intervals u n t i l the tests were completed, and pathological examinations
were promptly conducted.
reservoir elevations o f 192.8 m or greater,
the upper intake orifice remained submerged and excessive supersaturation o f
the water supply did n o t occur a t any
combination o f valve settings and f l o w
rates. W h e n the forebay was drawn down
t o 192.3 m elevation, the top 20 c m of
the upper intake orifice became exposed.
A plywood board placed over the lower
portion o f the grating t o simulate submerged debris resulted in an inadequate
water f l o w t o keep t h e upper intake line
full. W i t h i n 10 minutes dissolved N2concentrations rose f r o m 103 t o 153% o f
saturation i n the screening box and t o
124% i n the rearing ponds (Sample No.
3 on October 2 ) . T h e plywood barrier
was then removed for an extended period,
the f l o w of water entering the screening
box was doubled t o 0.85 m3/sec., t h e l o w er intake valve was closed, and only water
f r o m the upper intake was permitted t o
enter t h e system. A volume o f water less
than 0.85 m3/sec. flowed through t h e
system, resulting in a frothy air-water
m i x t u r e that surged i n t o the screening
box. Sample 4, taken w h e n t h e pond
water temperature rose t o the same level
as the forebay water temperature, contained 13496 dissolved Ng. T h e plywood
board was then replaced and all other conditions were set as described for Sample
N o . 3 for the remainder o f the test.
By October 3, dissolved N, concentrations in the screening box had increased
f r o m 103 t o 175% o f saturation. T h e
colder water temperature o f this sample
indicated t h a t most of t h e water i n t h e
system came from the hypolimnion.
Behavior
Results
Hydraulic Observations and Supersaturation
Rates
Results o f nitrogen concentrations
found in water samples during experiments are presented in Table 1. W i t h
All groups o f fish were under close observation w h e n the f i r s t mortalities occurred. There was n o indication t h a t the
fish were aware o f the potentially lethal
nature o f t h e environment. T h e following
sequence describes a typical behavior pattern during the first 60 seconds of abnor-
.
Table 1.
Approx.
Pool
Ele*.
Date
(m)
9/26
9/29
9/29
9/29
9/29
9/29
9/29
9/30
9/30
9/30
9/30
194.2
194.2
194.2
194.2
194.2
194.2
194.2
193.6
193.6
193.6
193.6
Nitrogen and Oxygen Concentrations Under Various Experimental Conditions
Flow Rate
at
Screening
sample Box Valve
No.
ms/sec.
Butterfly Volve
Setting ( % Open)
lntoke
1
-
-
-
1
0.43
0.43
0.43
0.43
0.43
0.43
0.43
0.43
0.43
0.43
40
40
40
100
100
100
40
40
40
100
100
100
100
0
0
0
100
100
100
0
1
1
2
2
2
1
1
1
2
Sample Stations@
Forebay 0 m
( C ) (ppm)
15.7
Forebay 2.1 m@ 16.7
13.5
13.6
Forebay 2.1 m@ 16.5
Screening box
15.9
Pond # 6
15.8
Foreboy 1.5 m@ 16.0
Screening box
13.6
Pond #6
13.4
Forebay I .5 m@ 16.1
Screening box
Pond #6
N~
O,
Temp.
Intoke
%
(%m)
%
9.8
9.8
9.4
9.5
9.9
10.0
9.9
9.7
9.5
9.7
9.7
98.7
100.8
90.6
91.7
101.4
101.5
99.8
98.8
91.6
92.8
98.6
16.8
16.8
18.0
18.3
16.9
17.1
17.2
16.9
18.0
18.1
16.9
100
102
103
105
102
103
103
101
103
103
102
9.5
9.4
10.0
10.0
12.1
10.7
12.7
9.4
13.8
12.3
12.2
90.4
90.0
100.5
100.1
1 10.9
103.1
127.1
83.7
127.0
113.9
112.3
18.1
18.1
17.0
17.5
27.7
21.6
22.6
20.0
31.8
27.4
26.9
103
103
101
104
153
124
134
108
175
152
149
Screening Box
Pond #6
Forebay 0.91~10
Screening box
Pond #6
Foreboy 0.9 m@
Screening box
Pond $6
Pond #6
Forebay 0.3 m@
10/2
10/2
1012
10/2
10/2
10/2
10/2
10/3
10/3
10/3
10/3
192.3
192.3
192.3
192.3
192.3
192.3
192.3
192.3
192.3
192.3
192.3
1
1
2
2
30
30
4
1
0
1
0
1
0
1
0
0.43
0.43
0.43
0.43
0.850
0.850
0.85
0.43
0.43
0.43
0.43
44
44
100
100
44
44
100
44
44
44
44
100
100
0
0
100
100
0
100
100
100
100
13.2
13.2
Screening box
15.4
Pond #6
15.2
Screening box
11.5
Pond #6
13.5
Pond #6
15.2
Forebay 16.8 m@ 10.2
Screening box
11.5
Pond #6
11.8
Pond #I
11.7
Screening box
Pond #6
@ Figures denote depths o t which water somple was token.
@ Forebay sample token in front of upper intake orifice.
0 Upper intake orifice partially blocked.
@ Valve opened to supply this amount but plywood and trash impeded flow so that less woter was delivered.
@ Forebay sample taken in front of lower intake orifice.
ma1 reaction : ( 1 sudden loss of ability
to swim into the current, ( 2 ) loss of ability to avoid other fish, cage sides, or a
flashlight beam, ( 3 ) loss of equilibrium
and vertical station in the water, and (4)
lateral movement near the surface with-
out a sense of direction. During the next
60 seconds fish exhibited violent writhing
movements, interspersed with moments
of inactivity as they lay at various depths
with ventral surfaces upward. After 3
minutes fish became moribund and show-
ed l i t t l e or no reaction t o handling. Nearly
all fish died w i t h i n 4 minutes after the
onset of abnormal swimming behavior. I n
a f e w instances, the fish exhibited leaping
and writhing movements i n an attempt
t o clear the water surface. None attempted to seek a greater depth before or during this activity.
Some salmon and steelhead that survived the catastrophe of September 5 remained in Ponds 8 and 9 during the test.
They averaged about the same size as the
test fish and died at the same rate. There
were n o survivors i n these ponds after the
tests were completed.
Mortali,ty Rates
Cumulative percentage mortality is
plotted against time f o r the four test
groups in Figure 3. These data show that
after 5 hours exposure, mortality was
much greater i n Ponds 6 and 7 a t 152
than a t 134%, and that prior exposure to
water w i t h a N , content of 134% accelerated the death rate when the fish were
subjected t o 152% levels. Croup 8exposed
directly to 152% N2 died rapidly but, as
was expected, required a longer period
before the first deaths than those previously placed in 134% water.
I n testing for the effect of handling
stress and subsequent susceptibility of
fish to high levels of nitrogen gas, it was
found that fish tested immediately after
handling did not show a significant d i f ference i n death rate when compared to
those held i n cages for 3 0 and 5 4 hours.
Pathology
Gas emboli i n the vascular elements of
the fins (most frequently the caudal)
were the most prominent external symptoms. I n some fish, emboli were seen
only in the dorsal lobe of the caudal fin;
i n others only the lower lobe was affected,
and i n a f e w cases emboli could n o t be
found i n any fins. Observations on Sep-
[8
tember 5 showed emboli t o be most numerous i n the dorsal fin, followed by lesser amounts i n the caudal, anal, and paired fins. I n addition, some fish contained
prominent gas bubbles i n the intercostal
vascular elements behind the peritoneum
along the lateral body wall.
External gill damage was noted i n all
fish checked. The central vascular element of gill filaments contained numerous tiny, bead-like vesicles and the
lamellae fed by these vascular elements
were ruptured and minutely hemorraged.
Several filaments i n a series damaged in
this manner were followed by a group of
intact filaments. Severity of the damage
varied. One fish died showing gill damage
b u t none of the other symptoms. I n another, an embolus was seen underneath
the epithelium in the anterior part of the
roof of the mouth. Emboli were present
i n the dorsal aorta of most fish examined;
however, search under magnification, b u t
w i t h o u t dissection, for emboli i n and
around the eyes proved negative.
Discussion
Hydraulic Observations and Supersaturation
Rates
Supersaturation of the hatchery water
supply system occurred only when the
upper intake line was n o t f u l l of water.
Data from Samples 1 and 2, on October 2,
showed no increase in saturation levels
even though the upper intake grating was
partially exposed for several hours. With
the plywood barrier i n place, just prior t o
taking Sample 3 on October 2, the water
level was 4 or 5 inches higher outside the
upper intake grating than inside, forcing
an agitated air-water mixture into the
throat of the intake. This highly aerated
water obviously came under considerable
pressure as it fell 17.4 m and joined the
higher-velocity flow from the lower intake. The screening box valve restriction
further increased back pressure on the
system, which probably accelerated the
CUMULATIVE PERCENTAGE MORTALITY
GUMUIATIVE PERCENTAGE MORTALITY
solubility of water-borne bubbles.
Partial occlusion of the upper intake
grating was not the only means of supersaturating the water supply. However, i t
was the only way under normal operating
procedures ( f l o w rates and valve settings) that superaturation could be induced. The exception was achieved while experimenting w i t h valve combinations w i t h
the upper intake partially exposed. It was
found that w e could produce dissolved
nitrogen concentrations of 1 34% w i t h o u t
having t o impede the flow o f water
through the grating. It was apparent that
w i t h a delivery rate which does not satisfy
the capacity of the system, due to valves
being 100% open a t the hatchery and
w i t h l i t t l e or no back pressure on the line,
there was ample opportunity for a watergas mixture t o occur because the pipe was
not full.
The epilimnion of Foster Dam forebay
was warmer in early September than in
early October, requiring less surface water
t o achieve a pond temperature of 1 3.S°C.
Sample 1, on October 3, was taken after
hydraulic conditions existing on September 5 were simulated (except as noted
above) and the system had operated at
stability for several hours. The resulting
nitrogen concentration of 175% in the
screening box was the highest measured
i n over 4 years o f research on nitrogen
supersaturation in the Columbia River and
its tributaries@ W e believe that supersaturated water, by passing through a
closed pipeline for an extended period,
combined w i t h a 21.3 m head pressure at
the discharge end, created this unusually
high concentration.
There was considerable reduction i n
concentrations of dissolved gases in pond
water compared t o water in the screening
@ Kirk T. Beiningen and Wesley J. Ebel. 1970. Supersaturation of dissolved nitrogen i n the Columbia and Snake
rivers. Data Report, 1965-69. U.S. Fish and Wildlife
Service, National Marine Fisheries Service, Biological
Laboratory, Seattle, Washington. Manuscript submitted
for publication.
box. This agrees w i t h the findings of
Marsh and Gorham I 1 904) w h o demonstrated that exposure of supersaturated
water t o the open atmosphere removed
excess gasses at a rate dependent on the
degree of exposure. A f t e r water upwells
into the screening box enclosure, it passes
over and through a perforated plate before entering a pipeline for distribution t o
the hatchery ponds. Exposure of the water
mass t o the atmosphere at this point perm i t t e d rapid dissipation of some of the
dissolved gases - from 175 t o 152%
(Table 1 1 . The subsurface position of entrance nozzles in the Burrows-type rearing ponds does not produce any significant
aeration.
Behavior
The behavior of fish affected by highly
supersaturated water a t South Santiam
Hatchery appears t o be characteristic.
Rukavina and Varenika ( 1956) reported
a reaction sequence among trout (Salmo
sp.) dying from gas bubble disease in the
River Boxna, Yugoslavia, similar t o that
observed during our tests.
Our observations also indicated these
fish were unaware of the lethal effects of
a nitrogen-supersaturated environment.
I n another experiment Ebel@ reported
dead and distressed fish w i t h gas bubble
disease symptoms i n water supersatured
t o 135% w i t h dissolved N,. These fish,
held in a vertical cage, were permitted t o
seek any depth between the surface and
4.6 m. Within 7 days 26% died and half
the survivors exibited gas bubble disease
symptoms. Even w i t h the opportunity t o
avoid stress, nearly two-thirds failed t o
do so.
Mortality Rates
Recent findings by Ebel (personal com@ Wesley J. Ebel, Notional Marine Fisheries Service, Biological Laborotorv.
.. Seattle. Washington, personal communication.
blood cells. Fibrous scar tissue had replaced a portion o f the gill filament in one
case and a hydropic degeneration o f the
squamous epithelium had occurred. I t is
believed t h a t the more severe damage observed i n our tests was related t o a very
high concentration o f dissolved nitrogen.
W e found a gill fungus i n all broods o f
fish reared a t the South Santiam H a t c h ery. It is k n o w n t h a t 1969-brood f r y were
exposed t o 1 1 5 % levels o f saturation
during the f r y rearing period (Beiningen
and W y a t t , 1970.) Harvey and Cooper
( 1962) suggest t h a t injuries m i g h t appear
when alevins exposed t o supersaturated
water reach the f r y stage. W h e t h e r this
could produce damage sufficient t o allow
the invasion o f Saprolegnia sp. on the gills
is n o t known. Westgard ( 1964) reported
damage t o the eyes o f adult chinook salm o n due t o supersaturated water. Examination of survivors 18 days after the loss
on September 5, showed t h a t some had
exopthalmia. T h e i r eyes were hemorrhaged, as reported i n adults b y Westgard.
There were also survivors w i t h fungus on
the caudal fin, gills, snout, and opercula.
Marsh and Corham (1 9 0 4 ) were also the
f i r s t t o report emboli i n the dorsal aorta
and heart. T h e dorsal aorta i n the fish w e
examined was found heavily occluded
w i t h emboli. T h i s was considered t o be
the second major cause o f death.
munication) i n laboratory experiments i n dicate that prior exposure t o supersaturation merely shortens t h e t i m e when the
test group w i l l die i n a higher concentration. These data are similar t o ours.
O n September 5, records a t t h e screening box showed a decrease i n water t e m perature 6.75 hours before the fish loss
was discovered. Our experiments indicated t h a t it takes about 3 hours f o r equalization o f screening box and pond temperatures. It thus appears t h a t the LT 100
(lethal t i m e for 1 0 0 % mortality) on
September 5 was about 3.75 hours. T h i s
implies t h a t nitrogen supersaturation was
152% or higher, based o n the LT 100 of
Croup 8 after only 5 hours during these
tests.
T h e f e w survivors o f the September 5
loss provided an unexpected control.
These fish showed t h a t confinement t o a
live box did n o t alter death rates. They
also provided evidence t h a t survivors of
exposure t o highly supersaturated water
apparently do n o t have i n inherent physiological tolerance t o high concentrations
o f N2.
Pathology
Cas emboli w i t h associated hemorrhages i n the gill tissues were the most
important cause o f death. Marsh and Corham ( 1904) f i r s t described gas emboli i n
the gills o f marine fishes exposed t o
supersaturated sea water, and their observations have since been confirmed b y
other investigators. Woodbury ( 1942)
reported severe gill damage due t o gas emboli w h i c h he thought was caused by oxygen supersaturation f r o m an algae bloom
i n a fresh-water lake. Pauley and Nakatani (1967) were the f i r s t t o report o n
the histopathology o f gills and other tissues o f fish affected b y gas bubble disease. T h e specimens used for their study
showed a l i g h t b u t detectable stage o f t h e
disease, Gill filaments were swollen and
edematous, and contained hemolyzed red
Summary
Severe losses o f juvenile salmon and
steelhead a t t h e South Santiam Hatchery
due t o gas bubble disease prompted studies o f i t s water supply system. Hatchery
water is obtained f r o m the epilimnion and
hypolimnion o f the forebay behind highhead Foster Dam. It was determined t h a t
when t h e reservoir level was a t the middle
o f the upper supply line, restriction o f
water f l o w b y trash caused a f r o t h y airwater m i x t u r e a t the entrance t o the
pipe, leading t o high N2 supersaturation
( 1 5 2 % ) i n hatchery ponds. Fish placed
1
I
in this water began to die w i t h i n 2 hours
and a l l were dead in 5 hours. T h e fish did
not appear t o be aware of the lethal environment u n t i l minutes before they died.
Typical pathology of gas bubble disease
was found in the affected fish.
Acknowledgments
Grateful appreciation is extended to
members of several agencies w h o provided valuable assistance in this study. H y draulic observations were made b y Emery
Wagner, Paul Ingram, and W i l l i a m Pilkington, Fish Commission of Oregon. The
hatchery manager, Ted Williams, and his
crew were helpful throughout the study.
Members o f the A r m y Corps o f Engineers
including Donald Westrick, Project Engineer, H . P. Toovoren, Hydraulic Engineer,
and Phli Moon, Chief, Fish Planning Section, provided their expertise; J. C. Archibald made the drawings f r o m which Figures 1 and 3 were adapted. Wesley Ebel,
National Marine Fisheries Services, assisted in solving technical problems as they
arose and reviewed the manuscript. John
Conrad and Kenneth Johnson of the Fish
Commission of Oregon also critically reviewed the manuscript.
Literature Cited
Beiningen, K i r k T., and Wesley J. Ebel.
1970. Effects of John Day Dam on dissolved nitrogen concentrations and salmon i n the Columbia River, 1968.
Trans. Amer. Fish Soc. 99 (4) :664-671.
Beiningen, K i r k T., and Ellis J. W y a t t .
1970. Supersaturation o f dissolved nitrogen in Foster Reservoir, South Santiam Hatchery, Dexter Reservoir, and
Dexter Holding Ponds, Willamette
River, Oregon. Fish Comm. o f Oreg.,
unpublished report.
Ebel, Wesley J. 1969. Supersaturation of
nitrogen in the Columbia River and its
effect on salmon and steelhead trout.
U.S. Fish W i l d l . Serv. Fish Bull., Vol.
67, No. 1, 1 1 p.
Harvey, Harold H., and S. B. Smith. 1961.
Supersaturation o f the water supply
and occurrence of gas bubble disease
a t Cultus Lake T r o u t Hatchery. Can.
Fish. Cult. NO. 30:39-47.
.
Harvey, Harold H., and A . C. Cooper.
1962. Origin and treatment of a supersaturated river water. Intern. Pacific
Salmon Fish. Comm., Progress Report
No. 9, 1 9 p. (Processed 1 .
Marsh, M. C., and F. P. Gorham. 1904.
The gas disease in fishes. U.S. Bur. Fish.
Report, 343-376.
Pauley, Gilbert B., and Roy E. Nakatani.
1967. Histopathology o f gas bubble
disease in salmon fingerlings. J. Fish.
Res. Bd. Canada, 24 (4) :867-87 1.
Rucker, R. R., and K. Hodgeboom. 1953.
Observations on gas bubble disease o f
fish. Prog. Fish. Cult. N o . 15 ( 1 1 :2426.
Rukavina, J., D. Varenika. 1956. A i r Bubble disease o f trout a t the source o f the
River Boxna (Mjehuricavost Pastrva na
Vrelu Bosne). A c t a lchthyologica Bosniae e t Hercegovinae, Editium 1, No.
7; Sport Fish. Abstr., 2 ( 4 ) , October
1 957.
Westgard, R. L. 1964. Physical and biological aspects of gas bubble disease in
impounded adult chinook salmon a t
McNary spawning channel. Trans.
Amer. Fish. Soc. 9 3 : 306-309.
Woodbury, L. A . 1942. A sudden mortality of fishes accompanying supersaturation of oxygen i n Lake Waukesha, W i s consin. Trans. Amer. Fish. Soc. 7 1 : 1 12117.
ESTIMATES OF MATURATION AND OCEAN MORTALITY FOR COLUMBIA RIVER
HATCHERY FALL CHINOOK SALMON AND THE EFFECT OF N O
OCEAN FISHING ON YIELD
Kenneth A. Hen@
Abstract
Nearly 31 million fall chinook salmon, Oncorhynchus tshawytscha (Walbaum), were marked and released in
1962-65 from 12 Columbia River hatcheries. All the major f~sheries that caught these fish were systematicolly
sampled in 1964-69.
The estimated fishing intensities and maturity schedules for the 1962-brood releases were calculated and whereever possible comparisons made with the 1961 brood data. A general model that was developed for the 1961 brood
data, and which equates oceon catch and river entry with recruitment, mortality, and maturation rates, was utilized
in this work. The colculated 6-month fishing intensity rates were higher for 1962 than 1961 brood for both general
marks and Kalama Hatchery marks. Since it was not possible to make estimates for the 1962 brood Spring Creek
fish, no direct comparisons for the two broods from this hatchery were attempted. Estimates of fishing mortality and
proportions maturing changed relatively little when large changes in natural mortality were used. Apparently only
about 0.3% of all marked fish released were alive at the beginning of their third year, when most of them become
available to the fisheries.
Mortality rates for marked Kalama fish were used for unmarked Kalama Hatchery fish, but maturation rates
were adjusted for poss~bledelayed maturity of marked fish. The total catch of 1962-brood chinook from the Kalama
Hatchery was estimated to be 134,409 kg, compared to 203,331 kg for the 1961 brood. Elimination of the estimated
loss due to marking would have raised the 1962 catch to 153,045 kg. Ocean fishing apparently reduced the number
of Kalama fish that returned t o the Columbia River by about 70% and the weight yield of hatchery fish by 25%.
This compares with a reduction of about one-half in number and 19% in weight for 1961-brood Kalama Hatchery fish.
Introduction
A cooperative marking experiment was
undertaken, beginning in 1962, by various federal and state fisheries agencies
along the Pacific Coast to estimate the
contribution of hatchery-reared fall chinook salmon (Oncorhynchus tshawytscha
Walbaum) to the various fisheries. Hatcheries included in the study are shown in
Figure 1. Marking continued over 4 years
(1962-65) and included the 1961-64
broods. Data collection was completed at
the end of 1969. Worlund et al. ( 1969)
published preliminary estimates of the
contribution to the fisheries bv the 1961
brood of hatchery fall chinodk salmon;
Rose and A r ~ 970) did the same
the 1962 brood. Cleaver (1969) made
additional analysisof the 1961 brood, ineluding estimates of ocean mortality, maturity schedules based on mark recoveries, and ocean catch of Spring Creek and
Kalama Hatchery fish that would have returned to the Columbia River in the absence of ocean fishing. Using Cleaver's
O
National Marine Fisher~es Service Biological
equations, similar estimates (except for
effects of ocean fishing on Spring Creek
fish) are made in this report for the
1962 brood and compared with results for
the 1961 brood. Effects of ocean fishing
on 1962 brood Spring Creek fish could
not be evaluated because there were no
river recoveries of 5-year-olds. The numbers marked from each hatchery for the
1962 brood are listed in Table 1.
Estimates of Ocean Mortality Rates
and Maturity Schedules for
Marked Fish
Estimated Capture of Marked Fish
Table 2 summarizes
the capture of
marked fish from the 1962-brood fall
chinook as calculated by ~ a r t i c i ~ a n tin
s
the hatchery
program.
totals represent the complete recoveries
of marks during the years 1964-67 with
the exception of tributary streams in 1967
which were not examined for marked fish.
Laboratory, Seattle, Washington.
1131
Table 1. Estimated numbers of fall chinook salmon released from study hatcheries for the
1962-brood year
Mark
Hatchery Source
Ad-LM
10 hatcheries @
Ad-LM
Spring Creek
Ad-LM
Kalama
Ad-LV-LM
Spring Creek
Ad-RV-LM
Kalama
RV-LM
Cascade
LV-LM
Marked and
Released
(No.)
Ratio o f
Marked to
Unmarked
Marked and Unmarked
Released
(No.)
Grays River
Total Marked
@ Includes all study hatcheries except Spring Creek and
Kalama.
P, = proportion of survivors maturing in nth year ( P , = 11
C,, = ocean catch in nth year
E, = number entering the river in nth
However, this is believed to have involved
few, if any, marked 1962 brood fish.
Ocean and river catch includes sport and
commercial fisheries, and river total includes catch and number of spawning
fish. Although Cleaver used only fish on
which the complete
mark was readily
recognizable, I have included partial
marks ( ~ d A
, ~ - L V ,and A ~ - R V )which
appear to have been due to regeneration
of maxillary bone. Furthermore, there
were considerably fewer mark recoveries
from the 1962 brood than from previous
brood, so these additional recoveries are
desirable. For the three marks that indicate maxillary regeneration, estimated
regeneration is slightly over 18% and is
remarkably consistent: Ad-LM, 18.1 %,
year
The following estimates of
P4 were de-
rived for the general (Ad-LM) and Kalama River (Ad-RV-LMI marks. Because
the other three
Ej
= O
equation ( 2, could not be used
for them.
General mark
P4 = 1.849
Kalama River mark P, = 0.91 3
l-he first valueof p, is tM, large.-rhe
proportion maturing in the 4th year cannot exceed1.0 since
no fist, would remain
at sea into the 5th year, The P4 value for
Ad-LV-LM~ 8 . 6 % ~ and Ad-RV-LM~ Kalama River fish also appears to be high,
18.6%. This compares with 17% calcu- particularly in comparison
with estimates
lated by Cleaver for the l g 6 1
calculated later in this paper. This folProportions of Fish Maturing
lows the general pattern for the 1961
Using Cleaver's equation ( 121, and brood which led Cleaver ( 1 969) to state,
"The best explanation seems to be that
assuming F and M constant,
a l l marks were not equally available for
C5
E4
P4 = - where
capture, and that no single mark was subc
4
E5
jected to a constant fishing rate during
F, = instantaneous ocean fishing the last3 years of life."
mortality rate in nth year
Since the assumption of F and M con-
M , = instantaneous ocean natural
mortality rate in nth year
stant did not produce meaningful results
for the 1962 brood, other possible as-
[ 141
I
1
1
1
I
1
LEWIS
- - - - - ---
l ic katat Hotchery
O R E G O N
Figure 1. Location of Lower Columbia River hutcheries involved in marking studies (adapted from Cleaver, 1969).
Table 2. Number of Columbia River fall chinook salmon of the 1962 brood, by age and
mark@, captured by all sources in years 1964-67
Age at
Capture
Ocean
catch
(C)
River
catch
River
toral
Number
spawning
(E)
General mark (AD-LM)@; 5,249,079 fish released, 0.1 5% recovered
2
3
4
5
Total
Spring Creek Hatchery mark (AD-LV-LM) @; 866,892 fish released, 0.1 2% recovered
2
3
4
5
Total
Kalama Hatchery mark (AD-RV-LM)@; 437,669 fish released, 0.15% recovered
2
3
4
5
Total
Grays River Hatchery mark (LV-LM); 241,494 fish released, 0.0796 recovered
0
0
0
0
106
0
2
2
64
3
4
1
.
-
173
0
5
0
-
0
-
4
--
7
3
Cascade Hatchery mark (RV-LM) ; 541, I 58 fish released, 0.03% recovered
-
2
0
0
0
0
3
4
66
43
5
24
0
6
5
30
5
Total
@ AD = Adipose fin, LV = left ventral fin, RV
8 Includes partial marks Ad-LV (RV).
= right
sumptions need to be examined. If it is
assumed t h a t M = 0 and P = 1, the highest values of e-Z,, ( w h e r e 2, = total instantaneous mortality rate, and e-Z,,=
survival rate), P, and F,, and lowest values of 2, can be calculated using Cleaver's
ventral fin, and L M
= left
maxillary bane.
equations (1969, p.
e-ZJ
--
E5
53)
.@
For example,
. - -45 = 0.263
c5+ES
171
for general marks. The results, using data
from Table 2 for the general and Kalama
Table 3. Maximal values of e-Z and P, and
minimal values of Z from 1962 brood
data, assuming M = O@
Data
General
mark
Kalama
mark
(ADAM)
(AD-RV-LM)
t h a t w h i c h matures a t the end o f t h e
fishing season and n o t necessarily t h e
same as t h a t a t t h e beginning o f t h e fishingyear. If ocean fishing mortalities w e r e
h e same f o r b o t h m a t u r e and i m m a t u r e
fish, t h e n P, w o u l d n o t be changed b y
fishing. However, it is i n t r i g u i n g t h a t in
spite o f a considerable difference i n estimated ocean fishing intensities, t h e P,,'s
are q u i t e similar for t h e t w o brood years
o f Kalama fish.
Effect of Arbitrary Values of M on F and P
Assuming t h a t M, = M 4 = MEand
P, = 1, arbitrary values o f M, ( m o r t a l i t y
@
Z, values
are also the highest possible values of
F,.
marks, are listed in Table 3. These survival rates suggest t h a t loss f r o m mortali t y is a t least 6 7 - 7 4 % i n t h e 5 t h year, 5 6 5 9 % in t h e 4th and 4 6 % i n the 3 r d year.
These mortality levels are generally higher than estimates calculated b y Cleaver
f o r t h e 1961 brood (38-61 % i n t h e 5 t h
year, 2 7 - 4 5 % i n t h e 4th, and 3 0 - 5 3 % i n
t h e 3 r d year.) Furthermore, there is n o
indication t h a t t h e loss may have been
lower i n t h e 4th year, as occurred for t h e
1 96 1 brood.
T h e highest values o f P4shown i n Table
3 are m u c h larger than f o r P, (0.646 t o
0.846 compared w i t h 0.083 t o 0 . 3 8 8 ) .
T h e proportion o f fish t h a t mature i n
their 2 n d year is relatively unimportant.
T h e P4values could be either a n overestimate or underestimate, depending o n
whether t h e C4 fish w o u l d have matured
mainly as 4 - o r 5-year-olds. Furthermore,
as pointed o u t b y Charles 0. Junge o f t h e
Fish Commission o f Oregon (personal
communicaton) , t h e proportion m a t u r i n g
(P,) calculated b y this method is really
@ It should be noted that in Cleaver's paper this equation
is printed with C, i n the denominator;
C, as shown here.
it should be
constant) were n e x t used t o investigate
relationships between P, F, and M, and
t o see h o w great t h e e f f e c t o f error i n M ,
u p o n estimated values o f F, a n d P,,. T o b e
comparable w i t h t h e analysis for t h e previous brood year, instantaneous yearly
values o f M, were taken as 0.24, 0.45,
0.48, 0.60, 0.72, and 0.96. For values of
M,, F, and P, and also f o r R, (recruits t o
t h e n t h year o f l i f e ) , each year o f l i f e was
considered t o begin o n October 1. F r o m
October t o M a r c h 31, only natural m o r t a l i t y was assumed t o be effective. D u r i n g
t h e entire fishing season ( A p r i l 1 t o September 301, b o t h natural and fishing
m o r t a l i t y were assumed t o affect f i s h a t
sea. M o r t a l i t y rates were held constant
throughout t h e periods i n w h i c h they
were acting.
T h e 1962-brood data i n Table 2 were
again used w i t h equations ( 2 0 ) t o ( 2 5 )
f r o m Cleaver ( 1 969, P. 5 5 ) .@ T h e results
are given i n Table 4 and Figures 2 a n d 3.
Similar data have been recalculated f o r
t h e 1961 brood, using partial marks and
more complete recovery data (Table 5 ) .
T h i s inclusion generally caused an i n crease i n t h e calculated fishing intensities
except for the F5value for Spring Creek
fish w h i c h is noticeably reduced. A s w i t h
t h e 1961 brood, t h e changes i n F, and P,
are nearly linear w i t h t h e change in M f o r
t h e range o f values used. Also, i n n o case
Table 4. F , P , and R values for marked@ fall chinook salmon of the 1962 brood; M is
summed for 12 months and F for 6 months
Natural
mortality
Fishing intensity
M
F5
-
F4
F3
Proportion maturing
p3
p2
p4
Recruitment
R5
R4
3,216
3,912
R3
General mark (AD-LM)
.24
1.346
.45
1.304
.781
.733
,535
,460
,810
.332
.290
.009
,007
220
260
10,441
13,681
.48
1.298
,727
.449
.783
.779
.284
,007
266
4,028
14,238
.60
.72
.96
1.275
1.251
1.206
.698
.67 1
,410
,761
.262
.006
293
4,516
16,736
,372
,301
,743
.705
.240
.616
.005
.003
323
393
5,076
6,438
19,831
28,320
.24
1.115
,830
,008
58
41 6
92 1
,199
Kalama mark (AD-RV-LM)
,485
.591
.065
,
@ Includes partial marks.
does t h e change in F, or P , equal t h e
changes in M,.
Because o f t h e relatively small change
in F, and P, w i t h changes in M,, estimates
in Table 4 f o r values between M, = 0.24
and 0.72 have a relatively narrow range.
Consequently, a modest error in estimati n g M, w i l l have l i t t l e e f f e c t o n t h e other
estimates.
Table 4 shows t h a t even if M were constant f o r t h e 3-year period, F is n o t constant f r o m age t o age nor hatchery t o
hatchery f o r t h e 1962 brood. It is u n fortunate t h a t n o analysis could be made
for Spring Creek marks o f t h e 1962 brood.
However, a number o f interesting comparisons can b e made between recoveries
o f general and Kalama marks o f t h e 1 961
and 1962 broods, as shown in Tables 4
and 5. T h e F, values w e r e m u c h greater
for t h e 1962 brood recoveries f o r b o t h
groups. I n fact, all F values w e r e higher
for this brood. On t h e other hand, t h e P
@ Equation (21 ) i n Cleaver's paper should have P, i n the
denominator as was subsequently noted i n an errata
sheet.
uc
=instontoneour natural mortolity
Fn I n s t m t o n ~ u atiahing mortolity
Pn = proportion maturing
Figure 2. Change in F, and P , with arbitrary
values of M,, general mark (AD-LM and
A D ) , 1962 brood.
Table 5. P, P, and R values for marked@ fall chinook salmon of the 1961 brood; M is
summed for 12 months and F for 6 months
-
Natural
mortality
M
Fishing intensity
F5
F4
Proportion maturing
F3
p4
p3
PI
-
Recruitment
R5
R4
R,
General mark (AD-RM)
,443
,413
,409
,392
,376
,343
.499
.421
,411
.370
.331
,261
.843
318
,814
,798
,781
,744
.281
,240
.235
,213
.I93
.I56
Spring Creek mark (AD-LV-RM)
Kalama mark (AD-RV-RM)
3
Includes partial marks.
values were fairly comparable for the two
broods. The 1962 brood Kalama data indicated a very small percentage maturing
before age 4, as was the case with the
1961 brood. The ranges of Kalama P4
values for the two broods were very similar (0.443-0.591 for 1962; 0.467-0.626
for 1961 1 . For the general mark (Ad-LM,
Ad-RM), the P2 values were fairly comparable for the two broods. The phenomenon of F 4 being greater than Fpfor the
Kalama marks of the 1961 brood at M
values of less than about 0.6, did not
occur with the 1 962-brood recoveries.
Finally, from the number of marked fish
released (Table 1 ) and the estimated
number recruited to the 3rd year of life,
it appears that survival was 0.003 or less
(at M = 0.45) for the 1962 brood in the
first 18 months after liberation. This compares with 0.01 1 for the 1 961 brood.
Survival of Marked Spring Creek and
Kalama Fish
The 1961 and 1962 broods are compared in Table 6. As mentioned previously, it was not possible to make a similar
computation for the 1962 brood Spring
Creek fish. I t is apparent from these data
that a much smaller proportion (R3/N) of
the marked 1962 brood fish survived the
first 18 months following release. Futhermore, a greater percentage of those that
survived and entered the ocean fishery
was caught. Although a comparison between the two brood years for Spring
Creek fish is not possible, certain speculations may be made. In most instances,
Table 6. Estimated number of recruits
( R 3 ) to
the ocean fishery, ocean catch
( C ) , and
other data for marked Columbia River fall chinook salmon of the 1961 and 1962 brood,
M=0.45;
based on recovery of Kalama, Spring Creek, and general marks
General mark
Data
1961 brood
Spring Creek mark
1962 brood
1961 brood
Kalama mark
1961 brood
1962 brood
No. marked
released ( N )
R,
R,/N
F,
F4
F5
p,
p
3
p,
Calculated ocean catch i n
t h e 3rd, 4th, a n d 5 t h year
1C )
C/R,
M c = instantaneous noturol mortolity
1.2
Fn =instantaneous fishing mortolity
Pn =proportion maturing
Figure 3. Change in F, and P, with arbitrary
values of M,, Kalama mark (AD-RV-LM
and AD-RV) , 1962 brood.
data for the general mark for 1961 returns
were intermediate between those for
Kalama and Spring Creek (i.e., R d N :
Kalama 0.01 1, General 0.008, Spring
Creek 0.007). On this basis, the data
suggest that 1962 Spring Creek fish survived and entered the ocean fishery in
about the same proportions as Kalarna fish
(Kalama 0.003, General 0.003), but in
much smaller proportions than the 1961
Kalama brood (Kalama 0.01 1, Spring
Creek 0.007). Likewise, it appears that
the ocean fishery caught a smaller percentage ( C / R 3 )of the 1962 Spring Creek
recruits than Kalama recruits (Kalama
0.409, General 0,373 ) ,whereasamong the
1961 brood, a greater percentage of the
Spring Creek recruits had been taken by
the ocean fishery ( Kalama 0.3 12, Spring
Creek 0.371 1 . This conclusion-that the
ocean fishery was less intense on the 1962
brood Spring Creek fish-is also indicated
(Table 2) by the ratio of ocean catch to
river escapement (574/459 = 1.25 for
Spring Creek and 514/135 = 3.81 for
Kalama . On the other hand, the percentage of the river escapement caught by the
river fishery was roughly similar for the
1961 BROOD
20r
FEMALES
MALES
(N) = no. weighed
(3
0
Y
5
I-
I
2
0
W
+
W
(3
a
I
I
Kolamo
Spring Creek
1 9 6 2 BROOD
15
FEMALES
[r
W
>
a
lo
5
0
2
3
4
5
AGE
2
(YEARS)
3
4
5
Figure 4. Average weights by age and sex of 1961- and 1962-brood Kalama and Spring
Creek hatchery marked chinook salmon cought in the Columbia River gill-net fishery.
t w o hatcheries: 80% for Spring Creek and
7 2 % for Kalama.
A s m i g h t be expected, there is a close
relationship between the size o f returning
fish and proportion o f each age group
maturing. N o t only did a consistently l o w er proportion o f Kalama fish mature a t t h e
younger ages, compared w i t h Spring
Creek, b u t Kalama fish t h a t returned t o
the Columbia River each year also weighed less. This is apparent f r o m Figure 4
where average weights are depicted by age
and sex for marked fish f r o m these t w o
hatcheries caught in the Columbia River
gill-net fishery. For b o t h brood years and
b o t h sexes, the 2, 3, and 4-year-old
Kalama chinook salmon caught in t h e
river were consistently smaller than
Spring Creek fish. These differences probably reflect d i f f e r e n t g r o w t h patterns
for the t w o stocks. These data also indicate t h a t after t h e faster growing,
earlier maturing Spring Creek fish are
eliminated, t h e residual, slower growing
5-year-old Spring Creek chinook salmon
are slightly smaller than t h e Kalama
stocks w h i c h are fished less intensively a t
t h e younger ages.
Delayed Maturity
Cleaver (1969, Figure 5 ) reported a
difference i n delay o f return t o t h e hatcheries f o r the Kalama a n d Spring Creek
differences between t h e t w o estimates o f
individual age groups t h a t conclusions based
o n t h e combined data may b e questionable.
In discussing t h e change in m a t u r i t y
schedule, Cleaver ( 1 969, Table 3 4 )
shows t h e average age composition o f returns f o r marked Spring Creek fish f r o m
t h e Columbia River gill-net fishery for a
number o f selected years. T h e age composition o f fhe 1939, 1942, and 1943 brood
from Spring Creek was very close t o that
o f t h e 1922 brood, but f r o m t h e 1961
brood there were relatively fewer 4th
and 5 t h year fish. Cleaver made his analysis before a l l the sampling data were
available. A l t h o u g h his conclusions are
n o t materially altered, there is a significant change in the percentage age composition for the 1961 brood when t h e complete data are included. T h e corrected
figures for Spring Creek for the 1961 and
1962 broods are listed in Table 8. T h e
increased proportion o f younger fish is
evident for t h e 1961 and 1962 broods.
These differences perhaps reflect gear
changes and fishing selectivity over the
years. T h e more intense ocean fishery in
recent years w o u l d tend t o reduce the
numbers of older fish returning t o t h e
river. Also, the comment b y Jensen
(1971 in his study o f brook t r o u t populations m i g h t apply t o chinook salmon:
"The brook t r o u t populations even matured a t an earlied age under heavy exploitation."
En are so variable and inconsistant for
AD - LM
'Spring Creek )
AD-LV-LM
(Spring Creek)
I
I
/
AD-RV-LM
(Kalama)
2
3
4
5
AGE I N YEARS AT RETURN
Figure 5. Percentage of 1962-brood marked
fall chinook salmon, by age and mark, in
returns to Kalama and Spring Creek
hatcheries.
1961 brood fish. Similar data (Figure 5 )
indicate t h a t t h e relationship is q u i t e
d i f f e r e n t f o r t h e 1962 brood. In fact, t h e
Spring Creek marked fish ( A d - L M ) o f
t h e 1 9 6 2 brood exhibit an almost identical
curve w i t h t h e one shown for 1961-brood
Kalama fish ( A D - R M ) .
For t h e 1961 -brood recoveries, Cleaver
( 1969, Table 3 1 ) used m a t u r i t y schedules (P,) for unmarked fish t o estimate
returns o f marked fish and, when compared w i t h observed values, concluded
t h a t the additional t i m e a t sea, due t o
delayed maturity, caused a reduction o f
1 6 % or less in fish returning t o the river.
Similar calculations for t h e 1962 brood
(Table 7 ) did n o t indicate any major
overall loss f r o m this factor. However, t h e
Effect of Ocean Fishing on Unmarked
Kalarna Hatchery Fish
Assuming common distribution patterns and mortality rates, it is possible t o
estimate h o w many fish were caught a t
sea f r o m hatchery stocks and w h a t t h e
return t o t h e river w o u l d have been in the
absence of a n ocean fishery. Sufficient
data are available f o r only the Kalama
stock. T h e P, values f r o m Table 7 for un-
Table 7. Estimates of proportion
maturing (P,) for unmarked fall chinook salmon of the
.
.
1962 brood and reduction in numbers entering the Columbia River due to delayed
maturity
Hatchery group
and age
Number of
unmorked
hatchery
returns
Survival
in river
fishery
Estimated
number
entering
river
253
3,760
4,242
280
.44
.24
.27
.47
575
15,667
15,711
596
(unmarked)
Adiurted En
(marked)
,003
.200
,820
39
1,101
1,107
P
Estimated
octuol En
(marked)
General Hatchery
2
3
4
5
8,535
Total
-
-
32,549
Precentage of marked fish due to delayed maturity..
.. . . .. .. . .0.0
Kalarno Hatchery
2
3
4
5
2
37 1
953
162
-
.14
.28
.29
.33
14
1,325
3,286
491
1,488
Total
,0003
.067
.592
-
5,116
Percentage loss of marked fish due to delayed maturity.. .
marked fish were used.
Estimated Number in the Ocean Catch
Estimated ocean catches for unmarked
chinook of the 1962 Kalama brood are
given in Table 9. Details of the calculations are given by Cleaver ( 1969, p. 64).
The estimated catch of 1962 brood un-
. .. . . ...2.2%
marked Kalama Hatchery fish at sea
( 1 9,375 were considerably greater than
the number estimated to have entered the
Columbia River as maturing fish (5,l 16).
I f none had been caught i n the ocean, the
number returning would have still been
less than the sum of river entry and ocean
Table 8. Age composition of marked Spring Creek hatchery fall chinook salmon caught in
the Columbia River fishery from selected brood years
Brood
lwr
Mark
2
Yeor of return
4
3
5
@
6
Number
of fish
Percent
1939
AN-RV and
AN-LV
.OO
.33
.42
.24
0
33
1942
AD-RV and
AD-LV
.OO
.I2
.70
.17
0
673
1943
AD-LP and
AD-RP
.I5
.68
.17
0
47 1
.14
.15
.42
.75
.68
.66
.56
.23
.I7
.18
.01
.QO
0
0
0
0
1,177
349
1,090
302
Average
1939-43
1922
1961
1962
AD-LV
AD-LV-RM
AD-LV-LM
.00
.01
.02
.02
@ All data except for 1961 and 1962 broods from Cleaver
( 1x9,Table
34). .
Table 9. Estimated ocean catch in numbers and weight of unmarked Kalama Hatchery fall
chinook salmon@, 1962 brood
Estimated ocean catch
Number of
fish weighed
Average
weight ( k g )
Number o f
fish caught
Weight of
catch ( k g )
2
3
4
5
Total
@ Seasonal average weight of marked AD-RV-LM troll caught hatchery fish landed a t Vancouver, Canada, used for
ages 3, 4 ond 5. Weight of troll caught fish adjusted b y 100/85 to compensate for loss in dressing.
c a t c h d u e t o deaths a t sea ( T a b l e 10).
A greater p r o p o r t i o n o f t h e n u m b e r o f
u n m a r k e d 1 9 6 2 Kalama f i s h caught came
f r o m t h e ocean fisheries t h a n was t h e
case with t h e 1 9 6 1 brood. For example,
t h e calculated ocean c a t c h f r o m t h e 1 9 6 1
b r o o d was a b o u t 7 0 % of those caught,
whereas t h e ocean c a t c h f r o m t h e 1962
b r o o d was a b o u t 8 4 % .
Estimated Returns with No Ocean Fisheries
T a b l e 10 gives t h e estimates o f numbers a n d w e i g h t o f u n m a r k e d 1 9 6 2 K a l ama f i s h t h a t w o u l d have r e t u r n e d t o t h e
C o l u m b i a River h a d there been n o ocean
fishing. N u m b e r s e n t e r i n g t h e river w e r e
calculated f r o m
Rne-Zn
( s t a r t i n g with n = 3. Then,
En =
Pn
Table 10. Estimated number and weight of unmarked Kalama Hatchery fall chinook salmon,
1962 brood, that would have entered the Columbia River, using different values of M
and assuming
Values of
F and M
F = 0 for ocean fisheries
Age
2
3
4
-
5
Total
Number
Actual
F=O,M=.24
F=O, M z . 4 5
F=O, M=.48
F=O, M=.60
F = O , M=.72
F=O, M = . 9 6
14
12
14
14
15
16
19
Weight in Kilograms
Actual
F=O,
F=O,
M = .24
M z . 4 5
F=O,M=.48
F=O, M = . 6 0
F=O, M = . 7 2
F=O, M = . 9 6
10
9
10
4,604
7,477
6,869
33,894
143,942
107,159
6,949
107,075
64,604
45,457
258,503
178,642
10
6,765
6,456
6,202
5,743
102,466
86,695
73,894
53,8 12
59,948
44,990
34,008
19,487
169,189
138,151
114,116
79,056
11
12
14
Table 1 1. Comparison of estimated Columbia River catch of unmarked 1962-brood Kalama
Hatchery fall chinook with the potential catch, assuming no ocean fishery
Number
Age spowning
2
3
4
5
2
37 1
953
162
Total
1,488
Average
weight
(kg)
0.73
3.47
10.31
14.15
Number x
weight
(kg)
Estimated
1962 run
t o river
(kg)
1
1,289
9,830
2,293
10
4,604
33,894
6,949
--
-
13,413
45,457
Estimated kg values
with F = 0, M = .45
Catch
in river
(kg)
9
3,315
24,064
4,656
-
Return to
river
10
6,869
107,159
64,604
--
Potentiol
cotch in river
(Col. 7
Col. 41
-
9
5,580
97,329
62.31 I
32,044
178,642
165,229
Apparent reduction in yield due to ocean fishing ( 165,229- 32,044- 92,080)/165,229= .249
R 4 = R3e-Z3 ( 1 - P3);
R3 was obtained in a similar manner.
T h e estimated returns a t M , = .45,
F = 0 were then used t o calculate d i f f e r ences in yield due t o ocean fishing (Table
11 ) . T h e number spawning and rate of
fishing o n each age group were taken f r o m
Table 7. T h e estimated catch i f there had
been n o ocean fisheries was obtained b y
assuming t h a t the number escaping t o
spawn w o u l d remain constant. Y i e l d is t h e
estimated weight o f the potential r e t u r n
t o t h e river with n o ocean fisheries less
t h e w e i g h t o f t h e 1962 brood spawners.
T h e data in Table 1 0 show that, with
M = .45 and n o ocean fishing, more than
three times (3.31 ) as many unmarked
maturing
" Kalama fish w o u l d have returned t o t h e Columbia River as actually entered. Using river survival rates f r o m
Table 7, t h e returns t o the Kalama H a t c h ery w o u l d have increased 3.41 times. T h e
calculated potential increase i n weight o f
fish returning t o t h e river under the assumption o f n o ocean fisheries is 3.93
times greater than the increase i n n u m bers. Assuming M = .45 and n o ocean
fishery, the calculated potential return o f
1962 unmarked Kalama fish w o u l d have
been 178,642 k g (Table 1 1 ) . A l l o w i n g
for the same magnitude o f spawning escapement as actually occurred ( 13,4 13
k g ) , there w o u l d have been a potential
river harvest o f 165,229 kg. Since t h e
combined ocean and river fisheries actually caught only 124,124 k g (river, 32,044
k g and ocean, 92,080 k g ) , there was an
apparent reduction in potential yield in
t h e river o f about 2 5 % due t o ocean fishing.
Estimated catches are considered t o be
t o o l o w because weights o f marked fish
were used. W e i g h t s f o r unmarked fish
were n o t available. Furthermore, only
losses due t o capture were considered. N o
adjustment was made for other losses
such as hooking m o r t a l i t y w h i c h tend t o
increase t h e apparent M.
Total Yield of Marked and
Unmarked Fish
T h e foregoing data can be used t o estimate hatchery yields, w h i c h are t h e sum
o f river a n d ocean catches o f marked and
unmarked fish. T h e catch o f marked
Spring Creek and Kalama fish is calculat e d i n Table 12. Numbers o f fish were
taken f r o m Table 2 and average w e i g h t
f r o m Table 1 1. Because origins o f fish in
t h e catch with t h e general Ad-LM and
regenerated marks were n o t indentifiable
b u t should be considered as a valid component o f hatchery yield, t h e following
relationship was used t o estimate the
w e i g h t contribution o f non-assignable
Table 12. Number and weight (kilograms) of marked Spring Creek and Kalama Hatchery
fall chinook salmon, 1962 brood, taken by all fisheries
Age
Ocean
catch
(number)
Avemge
weight
(kg)
Ocean
catch
(kg)
River
catch
number
Average
weight
(kg)
10
272
85
2.52
8.88
10.78
River
catch
(kg)
Spring Creek
2
3
4
5
Total
34
376
150
14
-
1.86
5.56
8.21
10.660
574
63
2,124
1,232
149
3,568
Non-assignable marks = 1.635 x assignable marks
Total yield t o the fisheries = 3,568
-
-
-
= 1.635
+ 3,356 +
367
x (3,568
3,356) = 11,321 kg.
11,321 = 18,245 kg.
+
25
2,415
916
-
3,356
Kalarna
-
2
3
4
5
0
293
190
31
-
-
2.39
6.69
10.66
6
21
60
10
847
1.27 1
330
-
A
514
.73
3.47
10.31
14.15
2,448
97
Non-assignable morks = 2.130 x assignable marks = 2.1 130 x (2,448
838) = 6,999 kg.
Total yield t o the fisheries = 2,448
838
6,999 = 10,285 kg.
+
+
+
4
73
619
142
838
0 From Kalama data.
marks originating from a specific hatchery:
No. o f non-assignable marks a t
W t . o f non-assignable marks in catch hatchery
No. o f assignable
Wt. of assignable
marks in catch
marks a t hatchery
It is estimated that the fisheries caught a
total of 18,245 k g of marked Spring Creek
fish and 10,285 k g o f marked Kalama
fish. It was impossible t o estimate the
catch o f unmarked Spring Creek fish.
However, the river catch o f unmarked
Kalama fish was estimated a t 32,044 k g
and the ocean catch a t 92,080 kg. Thus,
the total estimated yield t o the fisheries
f r o m the 1962 brood fingerlings released
a t the Kalama Hatchery was 134,409 kg.
This compares with 203,321 k g for the
196 1 brood.
It is possible t o compute the potential
yield f r o m the Kalama Hatchery if n o
fish had been marked and n o marking
mortality occurred. If the catch of u n -
marked Kalama fish is adjusted upward in
the same proportion as the number unmarked is increased t o account f o r the
total number released, the catch o f unmarked fish w o u l d be multiplied by 1.233
and the potential yield becomes 153,045
kg. O n this basis, the cost of the marking
program includes a loss o f 18,636 k g o f
Kalama fish f r o m the 1962 brood. It was
n o t possible t o compute a comparable
figure for Spring Creek fish.
Cleaver also computed the likelihood o f
survival t o the spawning stage for fish
maturing a t 3, 4, and 5 years o f age,
mainly t o show the difference between
Spring Creek and Kalama fish. In Table
13, similar information is listed for the
1962 brood Kalama fish. It was assumed
t h a t mortality rates were the same for
mature and immature members o f the
same age group, and also for marked fish.
Survival t o the hatchery was considerably
lower for the 1962 than 1961 brood, as
would be expected in view o f the higher
Table 13. Expected survival to the hatchery for Kalama fall chinook salmon, 1962 brood,
alive at the beginning of their third yea@
Age
Survival
in
river
fishery
Third
year
survival
a t sea
Fourth
year
survival
a t sea
@ River survival from Table 6; ocean survival with M
From Cleaver (1969, Table 40).
Fifth
year
survival
a t sea
= .45 from
Survival lo
hatchery
15-62 brood
/
1961 brood@
3.
3
fishing mortality.
Summary
The ocean fishery had a significant
effect upon the yield of 1962-brood
hatchery fall chinook in the Columbia
River. Calculated fishing intensity rates
for both general and Kalama Hatchery
marks for the 1962 brood were higher
than for the 1 9 6 1 brood. Only about 0.3 %
of a l l marked fish were estimated to be
alive a t the beginning of their third year.
Under the assumptions stated, the effect
of the ocean fishery was to reduce the
number of Kalama Hatchery fish returning t o the Columbia River by about 7 0 %
and the total yield i n weight by at least
2 5 % . The yield t o all fisheries from the
Kalama Hatchery for 1962-brood fall
chinook was estimated to have been
134,409 kg; without marking loss, it
would have been 153,045 kg. Similar losses probably occurred for the Spring Creek
fish, b u t could not be calculated.
Literature Cited
Cleaver, F. C. 1969. Effects of ocean fishing on 1961 -brood fall chinook salmon
from Columbia River hatcheries. Fish
Comm. Oregon, Res. Rep. 1 ( 1 ) , 7 6 p.
Jensen, A. L. 1971. Response of brook
trout (Salvelinus fontinalis) populations t o a fishery. Fish. Res. Bd. Can.,
Journ. 28 ( 3 : 458-60.
Rose, J. H. and A. H. Arp. 1970. Contribution of Columbia River hatcheries to
harvest of 1962 brood fall chinook salmon (Oncorhynchus tshawytscha) . U.
S. Fish Wildl. Serv., Bur. Commer.
Fish., Columbia Fish. Progr. Off., 2 5 p.
(Processed. )
Worlund, D. D., R. J. Wahle, and P. D.
Zimmer. 1969. Contribution of Columbia River hatcheries to harvest of fall
chinook salmon (Oncorhynchus tshawytscha). U.S. Fish Wildl. Serv., Fish.
Bull. 67 (2) : 361 -391.
INVESTIGATION OF SCALE PATTERNS AS A MEANS OF IDENTIFYING RACES
OF SPRING CHINOOK SALMON IN THE COLUMBIA RIVER
Burnell R. Bohn and Howard E. Jensen
Fish Commission o f Oregon, Research Division
Clackamas, Oregon
Abstract
A n attempt was mode to identify races of spring chinook, Oncorhynchus tshawytscha (Walbaum), from
scale samples collected from adult fish in major tributaries of the Columbia R~ver system dur~ng 1966, The
( 1 ) number o f circuli to the first annulus, ( 2 ) number of circuli to
following chamcteristics were compared:
the end o f fresh-water growth, and (3) mean distances between circuli. Tributary streams were grouped into four
lower Columbio River, ( 2 ) middle Columbia River, ( 3 ) upper Columbia River, and ( 4 )
geographic areas: ( 1
Snake River. It was concluded that races could not be separated by the methods used. Although counts differed
somewhat, the distinction between groups was insufficient for pract~col separation.
Counts to the first annulus did provide a means of identifying wild and hatchery fish in the Willamette
River. Examination of scales from known wild and hatchery fish indicated that an error of about 20% would
result when 16 or less circuli were used to identify wild fish and 17 or more to identify hatchery stocks.
Introduction
Knowledge o f the racial composition o f
a salmon r u n is important i n regulating
the harvest t o insure adequate escapem e n t o f racial components. Some races i n
the Columbia River have been identified
as t o t i m e o f migration. Studies using tagged f i s h show t h a t upriver races of spring
chinook are intermingled in their migrat i o n through t h e lower Columbia (Galbreath, 1 9 6 6 ) . T h i s m i x i n g o f races in an
area where a commercial a n d sport harvest
takes place indicates a need f o r discoveri n g a characteristic peculiar t o each racial
group t o serve as a means o f identification. In this study the use o f scale patterns was investigated for identifying
races o f spring chinook in t h e Columbia.
The International Pacific Salmon Fisheries Commission has successfully used
scales t o identify and manage races o f
sockeye salmon (0.
nerka) i n the Fraser
River (Henry, 196 1 .
T h e use o f scales in identifying races is
based on t h e assumption t h a t numbers o f
circuli and spacing between circuli are
functions o f t h e environment. Good
g r o w t h conditions (i.e., adequate food
supply, o p t i m u m temperatures, etc. re-
sult i n circuli being more numerous and
more widely separated; a n d the opposite
is true w i t h poor growing conditions.
Therefore, w e hypothesized t h a t if environmental differences exist between
major tributaries, they m i g h t be detected
i n t h e fresh-water area o f the scales. W e
also assumed t h a t differences i n g r o w t h
o f hatchery and w i l d fish exist and w o u l d
likewise be indicated in the scale patterns.
Methods
Collec#ingScale Samples
Scales were collected in 1966 f r o m
a d u l t spring chinook in major tributaries
o f the Columbia River system (Figure 1 1 .
T h e data were grouped i n t o four geographic areas: ( 1 ) t h e lower Columbia
tributaries f r o m the mouth upstream t o
W i n d River; ( 2 ) middle Columbia tributaries f r o m above Wind River t o the Snake
River; ( 3 ) upper Columbia River tributaries; a n d ( 4 ) Snake River tributaries.
W e f e l t the environmental conditions
were similar among tributaries in these
areas, thus providing a logical grouping
for analysis. Samples were obtained f r o m
b o t h w i l d a n d hatchery fish.
Scales were taken f r o m t h e area i m mediately above t h e lateral l i n e and
slightly posterior t o t h e origin o f t h e dorsal fin. Koo ( 1 9 5 5 ) indicated scales f i r s t
start t o g r o w i n this area and a f f o r d t h e
most complete record o f growth. W e took
t w o o r more scales f r o m each fish but
o f t e n discarded all scales collected because of regenerated nuclear areas. W e
used 1,787 scales f r o m 3,432 fish sampled
(Table 1 1 . Scales were placed o n g u m m e d
tapes and pressed i n t o plastic w i t h a scale
press. These impressions were projected
a t 1 0 0 X magnification.
Counting and Measuring Circuli
T h e following characteristics were
compared f o r all scale samples: ( 1 n u m ber o f circuli f r o m t h e focus t o t h e f i r s t
A
annulus; ( 2 ) number o f circuli f r o m t h e
focus t o t h e end o f fresh-water growth;
( 3 ) mean distance between circuli t o t h e
f i r s t annulus; a n d ( 4 ) mean distance b e tween circuli t o t h e e n d of fresh-water
growth.
A clear plastic overlay w i t h t h e 20"
dorso-radial line marked o f f in 2 - m i l l i m e t e r intervals was placed over t h e projected scale image. Circulus counts a n d
measurements were made f r o m t h e focus
t o t h e first annulus and f r o m t h e focus t o
t h e end of fresh-water g r o w t h along t h e
20" dorso-radial line of t h e scale (Figure
2 ) . C l u t t e r and W h i t e s e l ( 1 9 5 6 ) f o u n d
t h a t o n sockeye scales circuli in this area
were more complete a n d more w i d e l y
spaced. T h i s also appeared t o b e t r u e o f
spring chinook scales.
COLLECTION
SITES
1I
Figure 1. Map of the Columbia River system.
[ 29
I
Table 1
.
Number of scale samples collected and usable scales. by geographic area and
tributary
Number of scalei
Area and tributary
Collected
LOWER COLUMBIA RIVER
................................................
Cowlitz River
Willamette River
Willamette sport catch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Clockamas River (Eagle Cr . Hatchery) .....................
North Santiam River (Minto Pond) .......................
McKenzie River
Below Leaburg Dam
.................................
...................................
Middle Fork (Dexter Pond) ...................
.... .........
Wind River (Carson Hatchery) ...............................
Area Total ......................................
Above Leaburg Dam
MIDDLE COLUMBIA RIVER
.
-
Deschutes River . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
(Warm Springs River and Sheoror's Falls)
Area Total
......................................
SNAKE RIVER
Tucanon River
................................................
Grand Ronde River
Minam River . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
..............................................
.....................
..................
Catherine Creek . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Lostine River
Lookingglass Creek
Upper Grande Ronde River . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
lmnaha River
.................................................
Salmon River
Upper Solmon River
........................................
Middle Fork Salmon River . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Lemhi River ...............................................
Area Totol
......................................
UPPER COLUMBIA RIVER
Yakima River
................................................
..............................................
.................................................
Wenatchee River
Entiat River
Methow River
................................................
......................................
................................................
Area Total
GRAND TOTAL
Usable
DORSO-VENTRAL
K
~ ~ s
OMRLAY
Figure 2. Position of plastic overlay and 20"
dorso-radial line on outline of magnified
scale image.
T h e procedure used in counting and
measuring circuli follows:
( 1 ) O r i e n t t h e plastic overlay w i t h
t h e focus and vertical and horizontal axes o f t h e scale (Figure
2).
( 2 ) Locate t h e intersection o f t h e
annulus and 20" dorso-radial line.
( 3 ) Determine t h e last complete c i r culus before t h e annulus.
( 4 ) Measure t h e distance i n m i l l i meters f r o m t h e focus t o t h e m i d p o i n t o f t h e last circulus precedi n g t h e annulus.
( 5 ) C o u n t the circuli f r o m t h e focus
t o and including t h e last circulus
preceding t h e annulus.
( 6 ) Measure t h e distance i n m i l l i meters f r o m the focus t o the m i d p o i n t o f t h e last fresh-water c i r culus.
( 7 ) C o u n t the circuli from t h e focus
t o and including t h e last freshwater. circulus.
W e defined t h e annulus as t h e area o f
incomplete, closely spaced circuli followed b y complete, widely spaced circuli.
Since counts o f these incomplete c i rculi
varied, even on scales o f the same fish,
w e decided t o count through t h e last complete circulus before t h e annulus. Despite
our attemps t o standardize this technique,
certain decisions such as the location of
t h e annulus and end o f fresh-water
g m w t h were sometimes rather subjective.
Some scales also had unusual shapes and
~ placement
l c
o f t h e plastic overlay was arbitrary.
Results and Discussion
One assumption o f our study is t h a t
circulus counts f o r spring chinook f r o m a
given tributary show similar patterns annually. This assumption was partially tested for t h e 1961-62 brood years i n most of
the streams sampled b y comparing fish
m a t u r i n g as 4 and 5 year olds i n 1966.
T h e comparisons showed close agreement.
T h e test was extended t o the 1 9 6 0 brood
year a t Lookingglass Creek, tributary t o
t h e Grande Ronde River, b y examining
scales f r o m adults returning t o this stream
as 4-year-old fish in 1964-66. T h e mean
number o f circuli t o t h e f i r s t annulus f o r
1964, 1965, and 1966 was 12.5, 12.9,
and 12.7, respectively, and t h e frequency
distributions o f counts were almost identical (Figure 3 ) . W e found n o evidence t o
justify splitting t h e samples b y brood year
or age a t maturity, and consequently all
age groups were combined f o r analysis.
Scale samples f r o m males and females also
showed similar patterns and were combined.
Since t h e hatchery environment can
d i f f e r substantially f r o m t h a t i n natural
streams, only w i l d stocks were compared
among geographic areas. Results of these
comparisons are presented in t h e order in
which we believe the measurements
w o u l d prove most useful for racial identification: circulus counts t o the f i r s t annulus are considered first; circulus counts
t o t h e end of fresh-water growth second;
and mean distances between circuli third.
:'."I
"
6O
0
20-
0
1
j 4<
UPPER COLUMBIA RIVER
4
I
i , , I
SNAKE RIVER
NZ544
6
8
10 12 14 16 18 2 0 2 2 2 4 2 6 2 8 30
NUMBER OF ClRCULl
Figure 4. Percentage distribution and means
(XIof circulus counts to the first annulus
by geographic areas.
may indicate that significant environmental effects occur over smaller areas
I
than those w e chose. I n some cases small
I
10 l i
h 2 0 - 2 2 24' samples may n o t be representative o f an
NUMBER OF ClRCULl
individual tributary such as the C o w l i t z
Figure 3. Percentage distribution and means or Yakima.
The percentage distribution and means
(XIof circulus counts to the first annulus
for scale samples from Lookingglass o f circulus counts t o the f i r s t annulus
f o r the four geographic areas are presentCreek, 1964-66.
ed i n Figure 4. T h e mean counts f o r the
lower Columbia, middle Columbia, upper
Circulus Counts to the First Annulus
Columbia, and Snake river areas were
Table 2 shows the frequency distribu- 14.3, 14.2, 12.1, and 11.1, respectively.
tion and means o f circulus counts t o the These means were tested, each against
f i r s t annulus for w i l d fish i n tributaries the others, using a ranking test. At the
w i t h i n the four geographic areas. Gener- 5 % significance level, the lower and midally, scale samples f r o m the lower and dle Columbia areas d i d n o t d i f f e r statistimiddle Columbia River tributaries had cally. All the other comparisons were
the highest circulus counts t o the f i r s t found t o be significantly different b y the
annulus, and the upper Columbia and test. Although statistical differences exSnake river tributaries had lower counts. ist between mean counts, separation o f a
There are some variations between means mixed population o f races f r o m each area
for tributaries w i t h i n these areas w h i c h b y circulus counts t o the f i r s t annulus
"
0'4
lk
Table 2.
Frequency distribution and means of circulus counts to the first annulus for scale samples of spring chinook from wild
stocks in tributaries of four areas in the Columbia River system
Middle
Columbia
Lower Columbia
No. of
circuli
6
7
8
9
10
1
2
3
4
W
W
U
15
6
8
9
20
1
2
3
4
25
Total
Scales
Mean
Count
Cawlitz
River
Willamette
River
-.
Total
Deschutes
River
Snake River
--
Upper Columbim
Yakima Wenatchee
River
River
Entiat
River
Methow
River
Tot01
Grande
Ronde
River
lmnaha
River
Sohnon
River
Total
does n o t appear t o be feasible because of
the extensive overlap i n t h e distributions.
Lower Columbia River. Scales were collected f r o m t h e C o w l i t z and W i l l a m e t t e
rivers, the major spring chinook producing
tributaries o f t h e lower Columbia. T h e
overall mean for w i l d fish o f t h e lower
Columbia area was 14.3 circuli t o the f i r s t
annulus. Since only 13 usable scales were
obtained f r o m the Cowlitz, the mean
number o f circuli t o t h e f i r s t annulus
( 1 3.6) may n o t be representative. A s of
1966, the r u n i n the C o w l i t z was composed entirely o f w i l d stocks, but i n f u t u r e
years stocks f r o m a recently completed
hatchery w i l l contribute t o t h e run.
In t h e W i l l a m e t t e drainage, scales were
collected f r o m b o t h hatchery and w i l d
stocks. T h e only sample f r o m k n o w n w i l d
fish was 99 scales f r o m the McKenzie
River above Leaburg Dam. T h e w i l d fish
averaged only 14.4 circuli t o the f i r s t
annulus, similar t o w i l d fish f r o m the
Cowlitz. Samples were n o t obtained f r o m
other k n o w n w i l d populations i n the W i l lamette drainage.
Middle Columbia River. In this section
o f t h e Columbia, the three most importa n t spring chinook tributaries are the
Deschutes, Klickitat, and John Day rivers.
W e sampled Deschutes River fish only.
Samples were collected f r o m the l ndian
dip-net fishery a t Shearar's Falls about 71
km upstream f r o m the mouth, and f r o m
the W a r m Springs River 135 k m above
t h e mouth. T h e mean number o f circuli
t o t h e f i r s t annulus was 14.2, similar t o
w i l d fish f r o m t h e lower Columbia. This
is intermediate between t h e high mean
for hatchery fish i n the lower river and the
l o w means for w i l d fish f r o m the upper
Columbia and Snake rivers.
Snake River.
Scales were examined
f r o m the Crande Ronde, Imnaha, and
Salmon rivers and their tributaries. These
streams produce most o f the spring chinook in the Snake River drainage. T h e
overall mean for the Snake was 1 1.1 cir-
c u l i t o t h e f i r s t annulus.
I n t h e Crande Ronde system, samples
were obtained f r o m t h e Lostine, Minam,
and Upper Crande Ronde rivers, Catherine
Creek, and Lookingglass Creek. Fish i n
t h e drainage exibited an overall mean o f
12.2 circuli t o the f i r s t annulus, w i t h fish
f r o m all tributaries showing t h e same
basic pattern except for t h e upper Crande
Ronde where they had a mean o f 10.3
circuli from a small sample o f 2 5 scales.
Samples f r o m the lmnaha and Salmon
rivers had mean circulus counts of 10.5
and 10.1. Subsamples f r o m the upper Salmon, M i d d l e Fork o f t h e Salmon, and
Lemhi rivers had mean circulus counts o f
9.8,8.9, and 12.0, respectively. T h e upper
and M i d d l e Fork o f the Salmon had fewer
circuli t o the first annulus than any t r i butaries i n t h e Columbia River system.
Upper Columbia River. T h e mean count
o f circuli t o the f i r s t annulus for spring
chinook f r o m the upper Columbia t r i butaries combined was 12.1, similar t o
the mean for the Crande Ronde drainage.
T h e Yakima, Wenatchee, Entiat, and
M e t h o w rivers had mean counts o f 15.5,
10.8, 12.8, and 12.4, respectively. T h e
mean count for fish f r o m the Yakima
River was the highest o f any tributary but
this may have been due t o the small n u m ber o f scales collected.
Circulus Counts to the End of Fresh-Water
Growth
Total counts o f fresh-water circuli
were made f o r all major tributaries. T h e
means and frequency distributions were
compared by the same method as the
circulus counts t o t h e f i r s t annulus.
W e had d i f f i c u l t y locating t h e last
fresh-water circulus. This problem o f
indistinct termination o f fresh-water
growth and beginning o f ocean circuli
occurred t o some degree in all samples
collected. Also, counting circuli through
t h e annulus area introduced inconsistencies i n repeated readings and between
readers.
T h e relationship shown by circulus
counts t o t h e end o f fresh-water growth
was similar t o that o f counts t o t h e f i r s t
annulus. However, differences were less
pronounced. Mean total circulus counts
for fish f r o m the lower Columbia, middle
Columbia, upper Columbia, and Snake
River were 18.6, 20.4, 17.2, and 17.6,
respectively (Figure 5 ) . W e believe t h a t
counts t o t h e end o f fresh-water growth
w o u l d n o t provide a practical method o f
racial separation o f Columbia River spring
chinook.
Mean Distances Between Circuli
Mean distances between circuli were
computed in millimeters f o r counts t o the
first annulus and t o the end o f freshwater growth. T h e overall mean distances
between circuli for each o f the geographic
areas, beginning with the lowermost,
were 2.3 mm, 2.1 mm, 2.5 mm, and 2.4
m m t o the f i r s t annulus and 2.2 mm, 2.1
mm, 2.3 mm, and 2.3 m m t o the end o f
fresh water. All four areas also showed
LOWER COLUMBIA RIVER
nearly the same frequency distributions.
Because o f this similarity, mean distances
w o u l d n o t be usable as the criterion f o r
separatng races.
Separation of Hatchery and Wild Fish
/
Our data showed t h a t extreme environmental differences are required t o create
indentifiable scale patterns i n wild fish.
Such differences can occur between t h e
wild environment and modern hatcheries
where larger juveniles are produced.
I n the W i l l a m e t t e River system w e
sampled three o f five spring chinoqk
hatcheries and assumed these scale samples t o be representative o f hatchery fish
in the system. W e also assumed the M c Kenzie River sample f r o m above Leaburg
D a m is representative o f t h e wild populat i o n i n t h e system. Distributionsof the c i r culuscounts t o t h e f i r s t annulus presented
i n Figure 6 show the relative differences
between wild and hatchery fish. Examinat i o n o f t h e distributions indicates t h a t
the selection of 1 6 circuli or less t o identif y wild fish and 17 o r over t o i d e n t i f y
hatchery fish will m i n i m i z e improper
classification. In this case 2 1 % o f t h e
w i l d and 17% o f the hatchery samples
w o u l d be improperly classified.
W e feel t h a t this level o f error w o u l d
rsl
3 WICLAMETTE RIVER
WATGHERIES
N'S28
P
McKENZlE RIVER ABOVE
LEABURG DAM-WILD
-1
0
1
SNAKE RIVER
1
I
A
10 12 14 16 18 2 0 2 2 24 2 6 2 6 30 3 2 34
NUMBER OF ClRCULl
Figure 5. Percentage distribution and means
12of circulus counts to the end of freshwater growth by geographic areas.
I
NUMBER O F ClRCULl
Figure 6. Comparison of percentage distribution of circulus counts to the first onnulus for hatchery and wild fish in the
Willamette River system, 1966.
[ 35 I
~~
-~
n o t prevent this criterion f r o m serving as
a practical method f o r determining proportions o f hatchery and w i l d fish i n populations n o t heavily weighted toward one
group, e i t h e r w i l d or hatchery. I f the population is heavily weighted toward one
group, the error i n classification o f the
larger group exaggerates the size of the
smaller group.
W h e n this classification b y circulus
counts was applied t o a sample of 185
scales f r o m the 1966 lower W i l l a m e t t e
River sport fishery, 3 1 % o f the fish were
assigned t o hatchery origin and 6 9 % t o
w i l d origin. This m i g h t b e slightly biased
i n favor o f hatchery fish (smaller group).
T a k i n g this i n t o consideration, the results
compare favorably w i t h the W i l l a m e t t e
Falls count and hatchery returns in 1966
w h e n 28,200 were counted a t the falls
and about 6,200, or 2 2 % ) were subsequently handled at upstream hatcheries.
This is a m i n i m u m percentage because
n o t a l l hatchery fish i n t h e r u n were actual l y received i n t o the hatchery collecti n g systems.
T h i s circulus counting technique was
applied t o fish w h i c h were juveniles i n
the early 19601s, and since t h a t time,
hatchery rearing programs have been
pointed toward raising larger fish w h i c h
emphasizes the difference between w i l d
and hatchery stocks. In addition, selected
reservoirs i n the W i l l a m e t t e system have
recently been u t i l i z e d t o rear spring chinook. l ndica tions are t h a t these reservoir
fish w i l l have circulus counts intermediate
between stream-reared w i l d fish and
hatchery stocks. W e believe t h a t using
n e w reference values for hatchery and
reservoir stale patterns w o u l d provide
criteria for identifying hatchery, reservoir,
and w i l d stocks.
Conclusions
A n a t t e m p t was made t o separate t r i butary races o f spring chinook in the Columbia River b y analyzing scale patterns
as follows: ( 1 ) circulus counts t o the
f i r s t annulus, ( 2 ) circulus counts t o t h e
end o f fresh-water growth, and ( 3 ) mean
distances between circuli. T h e same
methods were used t o separate hatchery
and w i l d fish i n the lower Columbia River.
It was concluded t h a t w i l d stocks could
n o t be separated b y these methods. Although differences among races were i n dicated the overlap was too great f o r practical separation when all f o u r groups were
intermingled. However, counts t o the
f i r s t annulus did provide a means o f separa t i n g wild and hatchery fish i n t h e W i l lamette River system.
Acknowledgments
T h e authors appreciate t h e help of individuals f r o m t h e Fish Commission
o f Oregon, Oregon Came Commission,
Washington Department o f Fisheries, and
t h e Idaho Department o f Fish and Game
in collecting the scale samples analyzed i n
this paper. W i t h o u t their cooperation this
study w o u l d n o t have been possible.
Literature Cited
Clutter, R. I. and L. E. Whitesel. 1956.
Collection and interpretation o f sockeye salmon scales. International Pac.
Salmon Fisheries Comm. Bull. IX. 1 5 9
P.
Galbreath, J. L. 1966. T i m i n g o f tributary
races of chinook salmon through t h e
lower Columbia River based on analysis
o f tag recoveries. Fish Comm. Oregon
Res. Briefs 1 2 ( 1 ) :58-80.
Henry, K. A. 1961. Racial identification
o f Fraser River sockeye salmon b y
means o f scales and its applications t o
salmon management. lnternational Pac.
Salmon Fisheries Comm. Bull. XII. 97
P.
Koo, S. Y. and A. P. lsaran Kura. 1962
Scale studies o f Columbia River chinook
salmon. Fisheries Research Institute,
processed report f o r U.S. Fish and
W i l d l i f e Service, Contract 14-170001 - 3 9 3 . 3 2 p.
GROWTH OF JUVENILE SPRING CHINOOK SALMON
IN LOOKINGGLASS CREEK
Wayne A. Burck
Fish Commission o f Oregon, Research Division
Elgin, Oregon
Abstract
I studied growth of juvenile sprlng ch~nooksalmon (Oncarhynchus tshawytscha Walbaum) in Lookingglass
Creek during 1964-68. Mean lengths of fish sampled increased from about 35 mm in March to over 80 m m in October.
Growth was rapid and relatively uniform from April to September and averaged between 7.2 and 9.6 mm per month.
Condition factor was near 0.8 in April and ranged between 1.0 and 1.1 in most samples collected from June through
November. Condition factor increosed with f ~ s hlength.
The length-weight regression equation calculated from the sample data is log W
log
= - 5.44454
+ 3.25386
L.
Fish reared in separate areas of the stream showed
of differences i n water temperature in these areas.
borne
Introduction
In
the Fish Commission of Oregon
began a study designed t o investigate production, early life history, and ecology of
juvenile spring chinook salmon (Oncorhynchus tshaw~tschaW a l b a u m ) . One objective was to measure growth of juvenile
salmon t o provide a basis for understandi n g its relationship t o size a t migration
and numbers of adults which subsequentIY return t o spawn. This report Presents
the results obtained from the g r o w t h
studies. I t describes growth as evidenced
b y length-freq'-'enc~ distributions, average
lengths, and condition factors; presents a
length-weight relationship; and compares
growth o f fish reared i n separate areas o f
the stream.
Materials and Methods
Lookingglass Creek, a tributary o f the
Grande Ronde River in Union County,
Oregon, was the study stream. I t is approximately 27 k m ( 16 miles) long and
has one major tributary, Little Lookingglass Creek, w h i c h is about 16 km ( 10
miles) long and enters the main stream 6 1/2 k m ( 4 miles) above the m o u t h (Figure 1 1 .
[ 37
differences in linear growth which may be a reflection
I seined samples o f juvenile chinook
f r o m 1964-68. Samples were obtained
once eachmonth as early as March and as
late as November and generally consisted
the first 50 fish seined. "Standard Samp l e swere
~ obtained in mainLookingglass
Creek between mileposts 4.50 and 4.75
(standardarea) . Additional samples were
obtained during 1 9 6 6 - 6 8 between mileposts 0.00 and 0.50 (lower
sampling
lowersamples),
during 1966 and 1967
between mileposts 1 0 . 0 0 and 1 0.25
(upper sampling area, upper samples) ,
and during 1967 and 1968 near milepost
1.75 i n L i t t l e Lookingglass Creek. Locationof the four sampling
areas is shown
on ~i~~~~
1 and samplingactivityis summarized in ~
~1 . b
l
~
Fork length o f each fish was measured
t o the nearest millimeter and weight t o
the nearest 0.1 gram. Excess water was
shaken off before weight was measured
Mean length o f each sample was calculated from u n g r o u ~ e d data. However, I
grouped the length data i n t o 5 - m m
intervals t o simplify graphic presentation
of the length-frequency distributions.
Condition factor ( K ) was calculated
for every fish i n the samples according t o
the formula ( W ) ( 1 05)/L3, where W =
I
OREGON
LOCATION MAP
AREA
Figure 1. Lookingglass Creek system.
weight in grams and L = fork length in
millimeters. Sample conditon factor is the
arithmetic mean of the condition factors
of a l l individuals in the sample.
I calculated the length-weight relationship according to the equation log W =
log a
b log L, where W = weight in
grams and L = fork length in millimeters.
Measurements were grouped into 5-mm
class intervals. Logarithmic transformation was a t the midpoint of each of the
resulting 13 intervals, using the mean
weight from the corresponding interval
as the dependent variable. The constants
a and b were determined by least squares.
+
Results
Linear Growth
Chinook fry emerged from the gravel of
Lookingglass Creek as early as mid-January a t a minimum length of 32 mm.
Maximum length of downstream migrants
seldom exceeded 1 10 mm.
Length range 0.f fish in samples collected during March and April was narrow,
and distributions were skewed right, reflecting recruitment of emerging fry.
Length ranges in May remained narrow,
but length distributions began losing
skewness. After May, growth rates and
length ranges increased and distributions
approximated the normal curve (Figure
2).
Mean lengths ranged between 34.8 mm
in March ( 1967) and 86.1 mm in October (1964) and September (1965)
(Figure 3 ) . Annual variation in mean
lengths within the same calendar month
Table 1. Summary of length and length-weight samples of juvenile spring chinook salmon
collected in Lookingglass Creek
Little
Lookingglass
Area and
Standard
1964
1965
1966
1967
March
1968
1966
L
1967
LW
L
L
LW
May
June
LW
LW
LW
LW
LW
LW
LW
LW
LW
LW
LW
LW
LW
LW
LW
LW
LW
LW
LW
August
LW
September
LW
LW
October
LW
November
LW
July
L
= length
only; L W
= length
LW
LW
LW
LW
1967
1968
LW
LW
LW
LW
L
L
L
L
LW
L
LW
LW
LW
LW
LW
LW
LW
was small from April through August ex.
cept for 1967. Mean length of the fish
sampled in July 1967 was 8 mm less than
in July of other years, and subsequent
mean lengths during 1967 were correspondingly low (Figure 3 ) . Mean length
decreased from October to November.
Growth of juvenile spring chinook was
rapid and relatively uniform from April
through September when the average
monthly growth increments ranged between 7.2 and 9.6 mm (Figure 4 ) . Length
increase during October averaged 4 mm,
or about half the average from April
through September. The sampling data
show negative growth during November.
This result will be considered in the Discussion section.
MAY 2 0
N=50
JUNE 20
s 20
0
8
P
:20
9
1967
and weight
0
IL
1966
L
APRIL 21
N = 114
u,
1968
L
April
LW
LW
LW
LW
Upper
Lower
0
N = 50
20
0
Condition
N =50
I
O
OCTOBER 20
N = 50
I
30 4 0 50 60 70 80 9 0 100
FORK LENGTH (MM)
Figure 2. Seasonal growth pattern of juvenile
spring chinook salmon in Lookingglass
Creek based on standard samples, AprilOdober 1966.
Mean condition factor of the juvenile
chinook was lowest during April (approximately 0.8). Condition factor values increased sharply from April to June and
gradually but steadily from June through
September, and decreased after September
(Figure 5 ) . Mean condition factor ranged
between 1.0 and 1.1 in most samples collected between June and November.
The data plotted in Figure 6 show that
within each sample longer fish were rela-
-- - - -
-APPROXIMATE LENGTH AT--EMERGENCE
30
I
I
,
I
I
I
I
I
I
MAY
JULY
SEPT
20
20
20
NOMINAL SAMPLING DATE
MARCH
20
I
NW.
20
!
I
I
Figure 3. Mean length of juvenile spring chinook salmon in Lookingglass Creek based
on standard samples, 1964-68 (dots
represent undesignated years; circles represent 1967)
I
I
0.8
I
I
-
.
tively heavier than shorter ones, thus suggesting that condition is a function of
length as well as time.
Length-weight Relationship
I calculated a length-weight relation-
.
20
= SAMPLE INCREMENT
X = AVERAGE INCREMENT
0.7
I
MARCH
20
1
MARCH
20
,
1
,
1
,
1
MAY
JULY
SEPT.
20
20
20
NOMINAL SAMPLING DATE
1
1
NOV.
20
I
Figure 4. Growth of juvenile spring chinook
salmon in Lookingglass Creek based on
standard samples, 1964-68 (dots represent undesignated years; circles represent
1967).
I
I
l
I
I
MAY
JULY
SEPT
20
20
20
NOMINAL SAMPLING DATE
.
I
I
NOV.
20
Figure 5. Mean condition factor of juvenile
spring chinook salmon in Lookingglass
Creek, 1964-68 (dots represent undesignated years; circles represent 1967).
ship based on all the data from standard
samples grouped into thirteen 5 m m class
intervals. A total of 1,420 measurements
were involved and no class interval had
fewer than 2 4 measurements. The resulting regression equation is log W = -5.
3.25386 log L. The f i t obtained
44454
is excellent ( r 2= 0.9993), and empirical
and calculated sample weights agree
closely (Table 2 ) .This equation is believed t o be reliable for predicting weight,
when length is known, of juvenile chinook
i n all areas of Lookingglass Creek.
The observed mean weights used in
calculating the length-weight relationship
are compared w i t h predicted weights in
+
I
,
"r
Table 3.
Differences in Growth Among Areas
I found that juvenile salmon reared in
the stanaard sampling area were shorter
than those in the lower area but longer
than in the upper area (Figure 7 ) . Length
ranges were greater and distributions had
less distinct modes in samples from the
lower area. Juvenile chinook were not
abundant in the upper area and samples
were not obtained there after July 1967.
Results of sampling in Little Lookingglass
Creek during 1968 showed that juvenile
chinook in this tributary were longer, on
the average, than in the standard area.
Little variation was found in condition
factors of fish in a given length interval
regardless of where they were reared or
when they reached the given length.
g
2
--
4
H MAY
g
O''
&-A
JUNE
JULY
AUGUST
SEPTEMBER
OCTOBER
0.8
FORK LENGTH (MM)
Figure 6. Relationship, within month, between
length and condition of juvenile spring
The smallmean length of the
967
chinook salmon
in Lookingglass Creek
standard sample, and the
small
(from standard samples
in ,966).
gain in growth between June and July
1967 samples (Figures 3 and 4 ) could
Discussion
Table 2. Observed and calculated weight of samples of juvenile spring chinook salmon in
Lookingglass Creek
Total Weight of Samples (g)
Source of samples
Obsewed
Calculated@
Standard area
1964
1965
1966
1967
1968
Lower area
1966
1967
Upper area
1966
Little Lookingglass
1967
1968
@ Log W
= -5.44454
+ 3.25386 log L .
[41
I
Difference
(%
rable 3. Observed and predicted mean
weight of juvenile spring chinook salmon
in Lookingglass Creek
Meon Weight (91
Fork Length (mm)
Interval
Midpoint
28- 32
33- 37
38- 42
43- 47
48- 52
53- 57
58- 62
63- 67
68- 72
73- 77
78- 82
83- 87
88- 92
93- 97
98-102
30
35
40
45
50
55
60
65
70
75
80
85
90
95
-
0.23
0.36
0.57
0.89
1.24
1.70
2.25
2.93
3.71
4.50
5.54
6.7 1
8.16
9.4 1
0.38
0.59
0.86
1.21
1.65
2.19
2.85
3.62
4.54
5.60
6.82
8.21
9.79
100
-
11.57
@ Log W
= -5.44454
Observed
+ 3.25386
Log
Predicted@
L.
have been the result of sampling error.
But if sampling error had been a factor in
July, the August sample should have
shown a large growth increment and an
average length approximating other August samples. This did not happen. I believe
population density was responsible for the
small increase in growth measured in July
1967. An estimated 98,000 1966-brood
chinook (sampled for growth during
1967
emigrated from Lookingglass
Creek, and 45,000 of these left during
July and August 1967. This compares
with estimates of 49,000 emigrants for
the entire 1965 brood (sampled for
growth during 1966) and only 41,000 for
the 1967 brood (sampled for growth during 1968) (unpublished dataj . I believe
competition for space, food, or both probably retarded growth during June and
July
After many of these fish emigrated, growth of those remaining in the
Stream returned to normal,^ but mean
lengths continued to be low in comparison
with other years.
I
20
20
20
20
20
no
20
NOMINAL SAMPLING OAT
Figure 7. Seasonal variation in mean length
of juvenile spring chinook salmon collected from three areas of Lookingglass
Creek.
An apparent anomaly occurred when
mean lengths of the November samples
were less than mean lengths in October
(Figure 3 ) and negative growth was indicated (Figure 4 ) . 1 doubt that sampling
error was a factor since this condition prevailed whenever and wherever November
samples were obtained (Figure 7 ) . However, November samples were collected
from a single habitat type. Larger fish may
undergo a change in habitat preference a t
this time of year and not be represented
[ 42 I
in the November samples I obtained. I t
is also possible that a selective emigration
of larger, individuals takes place at this
time of year. My observations of the timing of downstream migration show that a
small "peak" of migratory activity occurs
during October and November. A series
of length samples indicates that mean
length of migrating juveniles increased
with time during this period (unpublished
data).
Water temperature may be an important, although perhaps an indirect factor
retarding growth in upper Lookingglass
Creek. Comparative water temperature
data are few, but those available suggest
that water was colder in the upper and
warmer in the lower area and in Little
Lookingglass Creek than in the standard
sampling area. Differences were as great
as 5"C, but mostly ranged from 1.7" to
3.3"C (unpublished data).
DEPTH DISTRIBUTION OF SOME SMALL FLATFISHES OFF THE NORTHERN
OREGON-SOUTHERN WASHINGTON COAST
Robert 1. Demory
Fish Commission o f Oregon, Research Division
Newport, Oregon
Abstract
The depth distribution of small flatfish o f six species common to the Oregon and Washington coasts is shown.
Depth distribution varies with season in that small fish o f most species inhabit deeper water in winter than in
spring or summer. There is l ~ t t l eseparation by depth of small f ~ s hfrom their larger kin.
Generally, smoll flatfish are constantly exposed t o encounters with commercial trawl gear since about 95%
of commercial trawl effort is expended on the continental shelf.
Introduction
A sustained groundfish fishery has operated o f f the northern Oregon-southern
Washington coast since about 1942. Research e f f o r t was largely l i m i t e d t o market-size fish. W o r k o n juveniles was lacui n g except for a study on juvenile English
sole, Parophrys vetulus, i n Yaquina Bay,
Oregon (Westrhein, 1955). I n the f a l l o f
1956 and summer o f 1957, an e f f o r t was
made by Fish Commission personnel t o
collect juvenile Dover sole, Microstomus
pacificus, b u t lack o f suitable gear precluded the capture o f smaller juveniles.
This report deals w i t h depth distribution information on juvenile and a d u l t
flounders collected d u r i n g a series o f
cruises f r o m July 1966 t o February 1968.
M y objectives in analyzing the catch data
f r o m these cruises were ( 1 ) t o determine
the e x t e n t o f overlap between juvenile
and adult stocks w i t h i n the depth range o f
the present t r a w l fishery, and (2) t o determine i f nursery areas exist w h i c h could
be closed t o trawling.
Methods
I fished from a chartered commercial
trawler using t w o nets, one of w h i c h was
a G u l f o f Mexico semi-balloon shrimp
trawl, w i t h 3.8 c m mesh (stretch measure) i n the body. The cod end had 3.8 c m
mesh w i t h a 1.3 c m liner. Foot rope and
head rope measured 15.8 m and 12.5 m,
respectively. T h e other n e t was a smallfish t r a w l with 6.3 c m mesh i n the body.
T h e cod end had 3.8 c m mesh, w i t h 2.86
c m mesh liner. T h e foot rope and head
rope measured 18.9 m and 14.3 m, respectively.
Both nets had a 0.8 c m tickler chain
attached t o t h e end o f each w i n g and
rigged t o fish approximately 0.5 m ahead
o f the f o o t rope. Gear type was constant
w i t h i n a cruise.
I used data f r o m four groundfish cruises (July 1966, M a y and August 1967, and
February 1968) and one shrimp cruise
(November 1966). T h e general area of
investigation extended f r o m Willapa Bay,
Washington t o Cape Lookout, Oregon
(Figure 1 ) .
I collected depth distribution data o n
Dover sole, Microstomus pacificus, English
sole, Parophrys vetulus, Rex sole, Glyptocephalus zachirus, Pacific sand dab, Citharichthys sordidus, slender sole, Lyopsetta
exilis, and butter sol e, Isopsetta isolepsis.
Limited catches o f some species o n some
cruises precluded analysis. N o cruise provided sufficient information o n small petrale sole, Eopsetta jordani, t o p e r m i t analysis.
T o w duration ranged from 6 to 3 5 m i n utes, b u t was usually 15 minutes. Since
a 15 m i n u t e t o w usually covers a linear
distance of six Loran micro-seconds, o r
one-half nautical mile. I adjusted catch
,,.
WASHINGTON
I
k!
WILLAPA
BAY
460
CAPE F A U O N
terns between years, even though 17
months elapsed between the f i r s t and last
cruise.
Tows w i t h i n a season were stratified b y
10 f a t h o m depth intervals. T h e percentage occurrence of fish b y depth was calculated separately for each size group.
Thus, for Dover sole i n February, the summation o f percentages i n each depth i n terval for fish equal t o or less than 18 c m
equals 100% as i t does for fish greater
than 18 cm.
Results
-
no
O
NAUTICAL MILES
j
OREGON
45"
Figure 1. General area of study.
rates t o conform t o a 0.8 kilometer tow.
Depth o f fishing ranged f r o m 6 t o 175
fathoms. I measured fish t o the nearest
lower m m and summarized lengths b y 2
c m intervals. On the f o u r groundfish
cruises, I measured all Dover sole and
usually measured the entire catch o f other
soles. I f I had t o subsample the catch o f
other soles, I expanded the subsample t o
the entire catch o f the respective species.
On t h e shrimp cruises only Dover sole
were sampled. These were processed as
t i m e permitted, or preserved i n a 1 0 %
formalin solution and processed later. I
adjusted the numbers o f fish making u p
the catch f r o m each tow, t o the numbers
of fish w h i c h w o u l d have been caught i n
a 0.8 kilometer tow.
A s used here, small fish implies juvenile, but rather than relate t o sizes a t
maturity, "small fish" means fish equal
t o o r less than 18 c m i n length f o r Dover,
English and Rex sole, and fish equal t o or
less than 14 c m i n length f o r sand dabs,
b u t t e r sole, and slender sole.
Data are arranged b y season o n t h e
assumption o f similar distribution pat-
Occurrence and relative abundance are
shown i n Figures 2 through 7, but I caution the reader t o note sample sizes w h e n
reviewing these figures.
Dover Sole
Small Dover sole occurred o n the continental shelf mostly between 3 0 and 80
fathoms, although they are found a t least
I
I
FEBRUARY
20
518 CM. n = 361
>I8 CM,n. 1,957
MAY
60
40 20 O .
w
q
.
Inn n n
JULY
.
* *
(
b
b..?
Slewn
>IBCM,n*
793
1,378
DEPTH IN FATHOMS
Figure 2. Depth distribution of two size
groups of Dover sole by 10-fathom interval and season.
to the 100 fathom contour. There was a
definite seasonal movement also. In February most small fish were at depths
greater than 70 fathoms, but during May,
July, and August the bulk of the small
fish occurred between 30 and 49 fathoms.
In November the depth of greatest abundance of small fish was in the 50-59 fathom stratum. Small fish were not caught in
the 20-29 fathom stratum where they
had occurred in May, July, and August.
In a l l sampling periods there was a
noticeable overlap in the depth distribution of large and small fish. This was most
pronounced in February and was only
slightly less in July.
during February and August stocks are
well blended.
Rex Sole
Small Rex sole occurred from the 20-29
fathom interval to at least the 130-139
fathom range. An inshore movement i s
evident with Rex sole also. In February,
small fish were not found inshore of 40
fathoms, but in May and August they
occurred in the 20-29 fathom stratum.
Larger fish showed a similar pattern of
movement.
Overlap in depth distribution between
small and large fish was present at almost
all depths.
FEBRUARY
English Sole
Small English sole occurred only from
the most inshore stratum to the 40-49
fathom range regardless of sampling period. There is some indication of movement
from deeper water in winter to shallower
water in summer. About 25% of the
small fish caught in February were caught
in the 40-49 fathom stratum, but none
were caught in this depth in May or August. The bulk of the small fish caught
within any sampling period occurred in
the 20-29 fathom range.
There was some separation of small fish
from larger fish, especially in May, but
80
40
60
20
NO FISHING
0
AUGUST
40
CATCH NOT SAMPLED
<I8 CM n :
966
>I8 CM: n :2.065
DEPTH IN FATHOMS
Figure 4. Depth distribution of two size
groups of Rex sole by 10-fathom interval
and season.
FEBRUARY
-
Sand Dabs
-
Sand dabs are unique among the species
studied, in that during February, small
fish were more widely distributed than in
May or August. They were most abundant
between 20 and 49 fathoms. There was no
separation between small and large fish.
>I8 CM, n = 413
::hahj
0
I
518 CM, n = 1,250
CM, n: 303
o >I8
do
-94
IQ
19
"
%
.
5999
'79
DEPTH IN FATHOMS
Slender Sole
3 9
7?9
Figure 3. Depth distribution of two size
groups of English sole by 10-fathom interval and season.
Small slender sole were found from the
30-39 fathom stratum to the 90-99 fathom depth, but were most abundant between 50 and 100 fathoms. Even though
sampling during the August period was
60
n
t
I
1<I4 CM, n = 192
0 >I4 CM n: 399
60
<I4 CM, n = 6 5 0
>I4 CM, n = 1,195
60
>I4 CM, n = 1,102
n
?
~
s
-
t
t
~
q
Figure 5. Depth distribution of two size
groups of sand dab by 10-fathom interval
and season.
sketchy, slender sole were caught closer
inshore than during the other sampling
periods. Both large and small fish were
present i n catches f r o m almost all depths.
Butter Sole
Butter sole are unique in t h a t b o t h
small and large fish occupy a narrow inshore depth range. They were n o t caught
beyond the 40-49 f a t h o m stratum. Small
b u t t e r sole were found a t this depth in
February, b u t n o t deeper than the 20-29
fathom stratum i n May. There was n o
separation o f small and large fish.
1 514 CM, n = 4 6 0
O > I 4 CM,n= 5 0 1
20
n
C14CM n:
61
;14 CM: n = 741
CATCH NOT
SAMPLED
*
0
514 CM, n: llB
> I 4 CM, n = 4 5 0
20
19
399
5$9
7y9
93g
DEPTH I N FATHOMS
o
NONE 'CAUGHT
NO FISHING
1
DEPTH I N FATHOMS
DEPTH I N FATHOMS
6 40
2U 2 0
1514 CM, n: 151
O >I4 CM. n = 301
'19,
Figure 6. Depth distribution of two size
groups of slender sole by 10-fathom interval and season.
,
Figure 7. Depth distribution of two size
q
~
groups of butter sole by 10-fathom interval and season.
Composite Distribution
Figure 8 shows a composite picture o f
depth distribution o f the species studied
and their relationship t o commercial t r a w l
effort. T h e depth distribution o f small
fish is by 30 fathom intervals and thereby
conforms w i t h t h e present method o f
compiling catch and e f f o r t statistics by
depth. M a j o r e f f o r t occurs between 30
and 59 fathoms, as do the b u l k o f t h e
small fish. I n fact, nearly 90% o f t h e
total trawl e f f o r t occurs a t depths under
90 fathoms.
Discussion
Small flounder o f the species studied
generally occur over the entire continental
shelf and nowhere does the small fish
complex escape the influence o f commercial t r a w l gear. Thus, the creation o f exclusive nursery areas t o protect small fish
w o u l d result i n a reduced adult harvest
by the fishery since it appears t h a t t h e
majority o f adult fish protected in nursery
areas w o u l d remain there and n o t cont r i b u t e t o fisheries elsewhere.
For certain species, estuaries provide
some protection f r o m commercial t r a w l i n g b y acting as rearing areas for juveniles.
English sole is a case i n pont. Westrheim
( 1 955) working i n Yaquina Bay found
large numbers o f 0 age English sole, but
relatively f e w age 1 juveniles. By contrast,
40
SMALL FISH
t
30
-
20
-
10
-
Figure 8. Composite of depth distritbution for
all species caught during February, May,
and August sampling periods superimposed on commercial trawl effort for
years 1966-68.
"
CEAN
2cRuIsEs
n = 315
X = 13.0
"
I
..
,
I
AUGUST
20
10
0
my ocean sampling (Figure 9 ) revealed
no 0 age fish. This suggests that, at least
during the initial phase of the rearing
period, juvenile English sole are not available to trawling gear.
The absence of smaller English sole in
ocean waters was apparently not a function of gear. During the course of my investigation, the smallest English sole
caught was 8 cm and very rarely were individuals less than 10 cm taken; however,
numerous other flounder measuring 6 cm
in length were caught.
The impact of commercial trawling on
the stocks of juvenile flatfish remains to
be determined. The importance of estuaries as sanctuary areas will be directly related to the seriousness of this impact,
and also to the degree of rearing which
X.6.3
"
0
1
FEBRUARY
-
-
n
-
= I 020
-
n
- x = h.9
I
2
nn
#
4
6
8
-
I
0 12 14
LENGTH (CM)
16
= 17R
1.
18
20
Figure 9. Length frequency distribution of
juvenile English sole caught in Yaquinb
Bay (Westrheim, 1955) and in the ocean
off the northern Oregon coast.
takes place in the estuarine environment.
Recent observations have indicated that
flounder other than English sole make use
of estuaries, but the extent of use is unknown.
Literature Cited
Westrheim, S. J. 1955. Size composition,
growth, and seasonal abundance of
juvenile English sole (Parophrys vetulus) in Yaquina Bay. Res. Briefs, Fish
Comm. of Oregon, 6 ( 2 ) :4-9.
NEARSHORE OCEAN CURRENTS OFF CIATSOP BEACH, OREGON, AND
THEIR RELATION TO THE DISTRIBUTION OF RAZOR CLAM LARVAE
Darrell Demory
Fish Commission o f Oregon, Research Division
Newport, Oregon
their effect on the distribution of razor
clams. The study was done during May to
Clatsop Beach is the most productive September, the peak spawning and setting
area for Pacific razor clams (Siliqua pat- time.
ula) in Oregon. Both commercial and recA pilot study was first conducted i n
reational fisheries have been active for
1963 todeterminewhich methodsto use
many years, taking 1 to 2 million clams and how t o apply them. Some suppleannually.
mental work was done i n 1965 t o learn
The 29 k m (18 mile) beach extends the effect of longshore d r i f t on bottles
from Tillamook Head (lat. 4S057'N, long. once they landed on the beach.
123'57'W) t o the South jetty of the Col~~~tstudies in the vicinity of Clatsop
umbia River (lat. 46'1 ZN, 10%. 124"W)
Beach pertain to conditions several miles
exposed t o wind and wave offshore. B u r t a n d W y a t t (1964) investiand is
action from the west. According t o Bas- gated the Davidsoncurrent about 6 4 k m
com ( 1 9 5 1 ) , C l a t s o ~ B e a c h i s c h a r a c t e r - ( 4 0 m i l e s ) offshore. Someof their d r i f t
ized by flat slopes ( 1 :70) of the beach bottles were recovered from Clatsop
face and small median diameter values of ~
~Morse,
~ c ~ ~and
h~
Barnes
. ~ ( 1967)
~ ,
sand (0.2 m m ) . The beach undergoes an studied the bottom currents near the
annual cycle as described by Shepard mouth of the Columbia River by the use
(1950) where high waves erode the sand of sea-bed drifters. The results showed
from the beach i n the fall and winter and that bottom currents within 10 km from
small waves move the sand shoreward i n the mouth of the river
and south of the
the spring and summer.
jetty were southeasterly toward the shore.
The sporadic and sometimes isolated Measurements taken by the University of
occurrence o f razor clams on Clatsop Washington Department o f Oceanagraphy
Beach suggests that the nearshore ocean show a southerly current along the coast
currents are an important factor i n the of Vancouver Island, Washington, and
distribution of the planktonic larvae. Oregon as reported by Barnes and PaquetW e ~ m o u t h , McMillian, and t-lolmes
te ( 1 9 5 4 ) . A study by Ballard ( 1 9 6 4 )
(1925) report that veliger larvaeof razor dealt w i t h marine sediments from the
clams are free swimming for about 8 Columbia River. Some o f these sediments
weeks and d r i f t w i t h the ocean currents are deposited along Clatsop Beach, indicanear the surface. A shoreward movement ting that the currents near the river a t
of these currents would tend to deposit times are southerly toward the beach.
the larvae i n the beach area accessible to
diggers. Conversely, a seaward set o f the
Methods
current would move the larvae away from
the beach.
Anchored styrofoam buoys marked the
I n 1963-64, a d r i f t bottle study was drop stations along the 5-fathom curve
conductedoff Clatsop Beach to determine about 1.5 k m ( 1 mile) offshore a t approxithe pattern of nearshore currents and mately 1.5 k m intervals between Tilla-
Introduction
[ 49
I
mook Head and t h e south jetty o f the Columbia River. Three additional stations
were included later due west o f station
16 o f f Seaside a t 0.8 k m (1/2 m i l e ) intervals.
T h e d r i f t bottles were 3 2 5 cc ( 1 1
ounce) beer bottles ballasted w i t h sand so
t h a t about 2.5 c m ( 1 inch) o f the neck
was above water and then corked and
sealed w i t h roofing cement. A numbered
identification card w i t h instructions t o
t h e finder was enclosed i n each. The f i n d er was instructed t o r e t u r n the b o t t l e t o
the Astoria Laboratory o f the Fish Commission, or t o the Seaside Aquarium. N o
reward was offered, b u t a letter of explanation was sent t o each finder.
Five d i f f e r e n t releases, excluding the
p i l o t study, were made between M a y
1963 and September 1964. Five bottles
were released a t each station, using a
10.7-m ( 3 5 - f o o t ) boat o n t h e f i r s t four
release dates, and a helicopter f l y i n g a t
an altitude o f 3 0 m (100 f e e t ) and a
speed o f 1 1 1 k m per hour (60 knots) for
t h e last release. T h e date, time, and n u m ber o f bottles released a t each station are
given in Table 1 .
I made most o f the recoveries by driving the beach, b u t a f e w were found by
beachcombers. A b o u t 7 5 % o f t h e bottles
were retrieved t h e day o f release.
Twelve sea-bed drifters were released
June 26, 1964, a t stations 4, 9, and 15.
A sea-bed d r i f t e r resembles an open u m brella w i t h a weight fastened t o the stem
w h i c h keeps it o n or near the bottom. I t s
use in determining residual b o t t o m currents has been described b y Lee, Bumpus,
and Lauzier ( 1 9 6 5 ) . N o n e o f these were
recovered.
Supplemental w o r k i n 1965 consisted
o f throwing d r i f t bottles i n t o t h e surf b y
hand and n b t i n g the t i m e and distance o f
d r i f t in t h e longshore trough. Once o n
t h e beach, t h e p a t h o f movement was
noted as succeeding waves washed the
bottles u p the beach.
Results and Discussion
Recovery data are shown in Table 1. A
total o f 570 bottles were released and 3 6 7
were recovered, a r e t u r n o f 64.4%. This
contrasts w i t h returns o f 4 . 6 % (Schwartzlose, 1 9 6 3 ) ) 12.9%. ( B u r t and W y a t t ,
1965)) 3.2% (Tibby, 1 9 3 9 ) , and 3.3%
(Dodimead and Hallister, 1958) f o r high
sea's studies, and 4 4 % (Lauzier, 1964)
and 41 % (Waldichuk, 1958) f o r nearshore studies.
T h e returns are p l o t t e d i n Figures 1-3
w i t h the p i l o t study included for comparison. N o a t t e m p t was made t o infer t h e
paths o f returns so t h e velocities calculated are minimal.
I n general, the surface currents w i t h i n
1 mile o f the beach are southerly and toward t h e beach d u r i n g t h e spring and
summer months. T h i s pattern w o u l d be
favorable t o deposition o f razor clam larvae in the area o f beach accessible t o diggers. However, there are some noteworthy
exceptions t o t h e general pattern.
Table 1. Number and percentage of drift
bottles released and recovered
Release
Date
Time
Released
3-7-63
5-3 1-63
0855-1013
0609-0729
140
80
59
67
42.10
83.8
6-28-63
6-26-64
8-4-64
9-14-640
0746-0925
1005-1 142
1006-1 151
1336-1 351
80
95
66
71
82.5
74.7
95
80
53
51
55.8
63.8
367
64.4
Total
Number Number Percentage
Released Recovered Recovered
.570
-
@ Pilot study; 71 % of floaters and 12%
recovered.
3 Helicopter drop.
-
o f sinkers
Returns f r o m t h e p i l o t study (Figure
1 ) show a pronounced southerly d r i f t
w i t h bottles concentrated a t Seaside and
some d r i f t i n g south o f Tillamook Head
u p t o 220 k m ( 1 3 7 miles) f r o m p o i n t o f
release. A southwest current or upwelling
near t h e beach n o r t h o f station 6, w i t h
SOUTH JETTY COLUMBIA RIVER
-O
MILE
I
JUNE 28, 1963
(A1
JUNE 2 6 1964
( 0 ,'
Figure 2. ( A ) Drift bottle returns from June 28, 1963 release and
(0) Returns from June 26, 1963 release.
[ 52 I
SOUTH JETTY COLUMBIA RIVER
AUG. 4, 1964
(A)
SEPT. 14, 1964
(B)
Figure 3. ( A ) Drift botlle returns from August 4, 1964 release and
( B) Returns from September 14,1964 release.
[ 53
I
the possible aid of an east wind, moved
the bottles seaward into the southerly
offshore current.
The results generally show a slight
southerly drift except for a marked northward movement off Seaside. According to
Tibby ( 1 9391, this dispersal suggests an
eddy. Minor dispersal from stations 5-9
(Figures 2A and 381 might indicate eddies or rip currents. Bascom ( 1964) states
that as waves recede from the beach the
longshore trough i s overfilled and excess
water flows seaward over the lowest part
of the longshore bar. A rip channel forms
as this part of the bar is eroded, and strong
rip currents flow through the channel and
disperse seaward. Drift bottles caught in
a rip current might be carried some distance seaward.
Both eddies and rip currents help explain the apparent crossing of lines of drift
of the plotted returns. The hydrographic
charts referred to by Tibby ( 1 939) and
Dodimead ( 1 958) show that currents do
not cross, but there may be exceptions
near the beach under local conditions.
Most recoveries were made the day of
but some were found l t'
days
later and were subject t o nearshore wave
action of both flood and ebb tides. On the
ebb tide, the bottles are not cast up on the
but drift in the longshore
until the tide begins to flood. Bottles
thrown into the longshore trough b~ hand
near lowtide drifted
km ( 3 miles)
from the release point before being cast
On the
and others were
recovered. I t was possible at times to visfollow those bottles which were
a few yards offshore.
Figure 2B illustrates a strong longshore
current. Minimum velocities of 1.3 to
1.85 km per hour (0.7 to 1 .O knot) were
recorded while 0.4 to 0.9 km Per hour
(0.2 to 0.5 knot) prevailed for all other
releases.
A lack of drift bottle returns in August
1964 from stations 9-1 5 (Figure 3 A ) and
a drop in water temperature of 50 C
(Wyatt, Still, and Hoag, 1965) suggests
upwelling. Two returns from the south
coast, 241 and 428 km ( 1 50 and 266
miles) from stations 8 and 1 1, indicate
that upwelling was significant and probably transported the bottles beyond the
north counter current into the dominant
south current offshore,
One drift bottle was
"orth of the Columbia River. On August
291 19681 one was found on the west
of
Island near TofinO.
The bottle had been at liberty 34 days.
Wind effect on drift bottles is negligible according to Dodimead (1958), but
Burt and Wyatt (1964) point out that
local wind stress appears to be important.
Wind roses constructed from U.S. Weather Bureau records at Clatsop airport for
a 30-day period prior to release dates
showed little correlation with recovery
location.
Summary and Conclusions
Drift bottles released off Clatsop Beach
indicate that the nearshore surface currents are southerly and toward the shore
during the spawning and setting time of
the razor clam.This pattern of drift is
favorable to the deposition of larvae
on
the beacharea accessible to clam diggers.
A prominent southerly longshorecurrent and a large eddy at Seasideare evident. These two factors probably enhance
the population of clams
at
Sea side by concentrating the larvae in the
beach area, Razor clam production at
Seasidehasbeen stable in past years and
accounted for 75% of the total Clatsop
Beach harvest in 1965. Harvests from
beaches north of Seaside fluctuate greatly.
Upwelling was indicated north of Gearhart in August 1964. This
set of
surface currents would displace clam larvae away from the beach, depending upon
[ 54 I
the timing of upwelling and larvae setting.
Digging was poor i n this area in 1966.
Acknowledgments
Grateful acknowledgment is made to
the Oregon State Police for use of their
patrol boat; Andy Olsen, skipper of the
Lela-E for use of his fishing boat; the
staff of the Seaside Aquarium for acting
as depository; the U.S. Coast Guard for
use of a helicopter; and to those interested persons who found and reported d r i f t
bottles.
Literature Cited
Ballard, R. L. 1964. Distribution of beach
sediment near the Columbia River.
Univ. Washington Dept. o f Oceanog.
Tech. Report 9 8 . 8 2 ~ .
Barnes, C. A. and R. G. Paquette. 1954.
Circulation near the Washington Coast.
Univ. Wash. Dept. of Oceanog. Tech.
Report 17. 3 1 p.
Bascom, W. N . 1951. The relationship
between sand-size and beach-face
slope. Trans. Am. Ceophys. Union. 3 2
( b ) : 866-874.
Bascom, W. N . 1964. Waves and Beaches.
Doubleday and Co., Garden City, N.Y.
p. 170- 173.
Burt, W. V. and B. Wyatt. 1964. D r i f t
bottle observations of the Davidson
current o f f Oregon. Dept. of Oceanog.
Oregon State Univ., collected reprints,
286-295.
Dodimead, A. J. and H . J. Hallister. 1958.
Progress report o f d r i f t bottle releases
i n the northeast Pacific Ocean. Jour.
Fish. Res. Bd. Canada, 15 ( 5 ) :851-865.
Lauzier, A. M. 1964. D r i f t bottle observations i n Northumberland Straight, Gulf
of St. Lawrence, Jour. Fish Res. Bd.
Canada, 2 2 ( 2 ) : 353-368.
Lee, A . J., D. F. Bumpus, and L. M . Lauzier. 1965. Sea-bed drifters. Int. Comm.
for NW Atlantic Fisheries, Res. Bull.
2. p. 42-47.
Morse, B. A., M . G. Grass, and C. A.
Barnes. 1967. Movement of sea-bed
drifters near the Columbia River. A.S.
C.E. Structural Engineering Conference
reprint 498.
Schwartzlose, R. A . 1963. Nearshore currents of the western United States and
Baja, California, as measured by d r i f t
bottles. Calif. Coop. Fish. Invest. Progress Report 1, July 1960-June 1962.
Marine Res. Comm., Calif. Fish and
Game. p. 15-22.
Shepard, F. P. 1950. Beach cycles in
southern California. U.S. Army Corps of
Engineers, Tech. Memo. 20. 2 6 p.
Talbot, G, 6. 1964. D r i f t bottle modification for air drops. Trans. Am. Fish. Soc.
93 ( 2 ) : 203-204.
Tibby, R. B. 1939. Report on returns of
d r i f t bottles released off southern California, 1937. Calif. Fish and Came.
Fish Bull. 55.
Waldichuk, M. 1958. D r i f t bottle observations i n the Straight o f Georgia.
Jour. Fish. Res. Bd. Canada, 1 5 ( 5 ) :
1,065-1,102.
Weymouth, F. W., H . C. McMillan, and
H. B. Holmes. 1925. Growth and age
a t maturity of the Pacific razor clam,
Siliqua patula. U.S. Bur. of Fish. Bull.
41 , 2 0 1 -236.
Wyatt, B., R. Still, and C. Hoag. 1965.
Surface temperature and salinity observations a t Pacific northwest shore
stations for 1963 and 1964. Dept. of
Oceanog., Oregon State Univ. Data
Report 21. 18 p.
ALGAE GROWTH AS A CAUSE OF TAG LOSS FROM JUVENILE FALL
CHINOOK SALMON IN SIXES RIVER ESTUARY, OREGON
Paul E. Reimers
Fish Commission of Oregon, Research Division
Port Orford, Oregon
During the summer of 1966, about
2,500 vinyl thread and pennant tags were
applied to juvenile fall chinook salmon,
Oncorhynchus tshawytscha ( Wa lbaum ) ,
in Sixes River estuary to determine various
population statistics. The tag was previously developed for work in fresh-water
impoundments (Korn, et al., 1967) and
consisted of a numbered pennant 4.7 mm
long, 2.4 mm wide, and 0.4 mm thick.
The pennant was attached under the insertion of the dorsal fin using vinyl thread
either 0.41 or 0.48 mm in diameter. I
later learned from Korn (personal communication) that his tags were attached
under the origin of the dorsal fin with
the trailing part of the tag riding on the
fin. He also indicated that juvenile salmonids tagged in fresh water showed
little tag loss and negligible effect by the
tag. Some of his fish retained tags after
migrating to sea and returning to the
spawning grounds. However, Pyle ( 1965)
observed that tagged brook trout, Salvelinus fontinalis ( Mitchill ) , retained this
tag poorly and the vinyl irritated the surrounding tissue.
The vinyl thread and pennant tag was
of only limited value in Sixes River estuary because of growth of the algae, Enteromorpha sp., and several diatoms, Schizonema sp., Melosira sp., and Fragilaria sp.
on the tag and thread. Most attachment
occurred at the overhand knot in the
thread. Enteromorpha sp. probably caused
the greatest problem because filaments
were up to 90 mm long (Figure 1 1 . MaxiThis work was conducted in cooperation with the U.S.
Department of Interior, Bureau o f Commercial Fisheries
under The Anadromons Fisheries Act (PL 89-304), Project
26-1, Contract Number 14-17-0007-897.
mum length of the diatom masses was
about 15 mm.
The effect of the additional weight and
resistance of the algae was to gradually
pull the tag out of the fish. Tagged fish
in fresh water and those without algae in
the estuary were not observed to be losing
their tags. Many tags with algae were
found only superficially attached by a
layer of skin. Other fish were found just
after the tag had pulled out leaving an
open wound. As the tag pulled back and
out, the path of migration left a characteristic "V" extending from the last point
of attachment to the original thread hole
on each side of the fish. As the wound
healed the "V" remained as a black scar.
The rate of tag loss could not be measured because the rate of mortality on
fish about to lose their tags was not
known, and fish were thought to be continually emigrating to the ocean.
An increased proportion of recovered
tags possessed algae at longer intervals
after tagging (Figure 2 ) . Most fish recaptured within 15 days had only small
amounts of algae. The decreasing percentage of recoveries at longer intervals
may have resulted from a combination of
tag loss, mortality, and emigration.
A comparison of rate of growth was
made for fish with and without algae on
their tags and no difference was found.
Quite possibly tag loss would have been
less had the thread been attached under
the origin instead of the insertion of the
dorsal fin. On adult salmon, placing Petersen disc tags under the origin of the dorsal
fin provided greater tag retention
(Thompson, et al., 1958.) Had the vinyl
thread and pennant tag been attached un-
I
,
I
Figure 1. Comparison of a tag without algae at time of tagging and a
tag with algae at time of recovery.
der the origin of the dorsal fin, action of
the tag against the fin might have served
to reduce the growth of algae.
Others have reported animal and plant
growth on tags as possible cause of tag
loss. Bonner ( 1965) reported growth of
algae as a cause of loss of spaghetti tags
from striped bass, Roccus saxatilis ( Walbaum), in Chesapeake Bay. Lasater
( 1 966) reported algae growth and high
tag loss on yearling chinook salmon in
Bowman's Bay, Washington, but did not
necessarily link them. (Chadwick ( 1963)
found barnacles, hydroids, and algae on
tags being tested for striped bass. Allee
(personal communication) tagged juvenile salmonids in Big Beef Creek, Washington, and noted accumulations of algae. He
used a small plastic acetate tag attached
under the insertion of the dorsal fin using
nylon fishing line.
I f the vinyl thread and pennant tag is
5o
,40
ALGAE PRESENT
2
W
gg 30
k20
10
"
O
1-5
6-10
11-15
16-20 21-25 26-30 31-35
NUMBER OF DAYS AFTER TAGGING
36-40
Figure 2. Change in number of tags recovered at 5-day intervals after tagging and
increase in proportion of recovered tags
possessing algae.
used in estuarine studies again, I suggest
that tagging under the origin of the dorsal
fin be tested before using the tag extensively.
Acknowledgments
I appreciate the help of Mr. K. Crenshaw in putting out the tags. Dr. A. Stein,
Department of Botany, the University of
British Columbia, identified the plant
growth on the tag. Mr. Lawrence Korn
supplied some of the tags and was very
helpful. Mr. Alan McGie reviewed the
manuscript.
Literature Cited
Bonner, Rupert R., Jr. 1965. Observation
on tag loss and comparative mortality
in striped bass. Chesapeake Sci., 6 ( 3 :
1 97- 1 98.
Chadwick, Harold K. 1963. An evaluation
of five tag types used in striped bass
mortality rate and migration study.
Calif. Fish and Game, 49, ( 2 ) : 64-83.
Korn, L., L. H. Hreha, R. G. Montagne,
W. G. Mullarkey, and E. J. Wagner.
1967. The effect of small impoundments on the behavior of juvenile anadromous salmonids. Fish Comm. Oregon, Clackamas. Processed Report,
127 p.
Lasater, J. E. 1966. Tag loss and mortalities of small chinook salmon. Washington Dept. of Fish., Fish. Res. Papers, 2,
( 4 ) : 90-93.
Pyle, Earl A. 1965. Comparative tests of
three types of vinyl tags on growth and
swimming performance of brook trout
held in hatchery troughs. Fisheries Report Bulletin No. 28, The nutrition of
trout. Cortland Hatchery Report No.
33 for 1964: 48-55.
Thompson, R. N., J. B. Haas, L. M. Woodall, and E. K. Holmberg. 1958. Results
of a tagging program to enumerate the
numbers and to determine the seasonal
occurrence of anadromous fish in the
Snake River and i t s tributaries. Fish
Comm. Oregon, Clackamas. Processed
Report. 200 p.
THE USE OF OXYTETRACYCLINE MARKS IN VERTEBRA OF
A W L T SALMON TO DETERMINE SMOLT SIZE
Don F. Swartz
Fish Commission of Oregon, Hatchery Biology Section
Clackamas, Oregon
Observations made during examination
of oxytetracycline (OTC) marks on vertebra from large numb.ers of adult salmon
led t o the belief that there is a relationship between the OTC mark size and the
size of the fish when it was'marked. The
OTC mark i n a fish vertebra corresponds
to the size and shape of that bone a t the
time the mark was produced much the
same as a growth ring in a tree indicates
the dimension of the t r u n k at a particular
age. Thus, i f hatchery smolts are OTC
marked immediately prior to release, the
marked vertebra from resultant adults
w i l l provide information as t o the size a t
release of those smolts that survived.
An experiment was carried out t o determine if fish of similarsize (fork
length) acquire OTC marks of similar
size i n vertebra from the caudal peduncle
and further t o establishthe relationship
between OTC mark size and fork length.
From .the production stocks at t w o W i l lamette River hatcheries, 133 yearling
spring chinook (Oncorhynchus tshawytscha, Walbaum) were selected a t the conclusion of OTC marking treatments i n
March 1970. Fork lengths ranged from
10 t o 24 c m giving 15 length groups ( 1
cm intervals). The minimum group size
was 4 and the largest was 23. With a
fin clip and freeze brand employed t o identify length groups, the selected fish
were reared an additional 10 weeks in a
6-foot circular tank. The additional growth was considered necessary t o provide
clear definition of the OTC mark i n the
vertebra.
A vertebra from the caudal peduncle of
each fish was examined under ultraviolet
illumination w i t h a binocular microscope
a t 20X. The diameter of the OTC mark
on the vertical axis of the vertebra was
measured by use of an optic micrometer
located i n the eyepiece of the microscope.
L i t t l e variation in OTC mark diameter
was noted w i t h i n a given length group.
W h e n OTC mark diameters were compared w i t h fork length a t time o f marking,
a straight line relationship was observed
(Figure 1 ) . The correlation was significant a t the 1 % level.
The 1967-brood spring chinook a t the
Fish Commission o f Oregon's Trask River
Hatchery received an OTC marking treatment just prior t o release i n t o that river
i n October 1968, and February 1969. In
1970, 37 male 3 - p returnees
~
bearing
OTC marks were recovered a t the hatche r y Comparing OTC mark diameters w i t h
adult fork length, a fair, though highly
~ i g n i f i c a n t correlation
,
Was found t o exist.
24
22
i
?
i20
16
;
14
12
o
GROUP MEAN
10
25
30
35
40
45
50
55
60
65
DIAMETER IN EYEPIECE UNITS
Figure 1.
[ 59 I
Relationship between OTC mark
diameter and fish size at time of OTC
treatment for spring chinook from Willamette River hatcheries.
(Figure 2 ) . Small returnees tended to
have smaller OTC mark diameters than
larger returnees. Further investigation of
this relationship will be made when 4and 5-year fish return to the hatchery.
The observed relationship appears to
provide a means throughout the life of
the fish of accurately estimating fork
length at the time of the OTC treatment.
The Fish Commission will employ this
method to determine the smolt size of
those fish which are contributing most
heavily to our adult returns.
-
75
5
,70
5
'j
2 65
2
60
55
40
45
50
55
60
DIAMETER IN EYEPIECE UNITS
Figure 2. Relationship between OTC mark
diameter and fork length of 3-year-old
Trask River spring chinook.
A Note on Coastal Movement of Shad
There are from 400 to 500 thousand
pounds of shad Alosa sapidisima harvested
commercially from five streams along the
central coast of Oregon each year. Our
knowledge of the life history of these
particular stocks is limited. However, we
do know that shad on the east coast spend
a major portion of their life in the ocean,
and that shad have been taken incidentally in trawl nets fished for shrimp and
bottomfish off Oregon.
Some insight into the coastal movement of shad comes from the recovery of
a tagged shad in 70 fathoms of water off
Willapa Bay, Washington, on September
9,1969. The fish was captured in a shrimp
trawl fished by the commercial boat
Owners Joy. The fish had been tagged
some 180 nautical miles to the south, on
April 4, 1969, as part of a Fish Commission of Oregon project to estimate the
numbers of shad available to commercial
nets in the Umpqua River system.
T. Edwin Cummings
Fish Commission of Oregon
Tillamook, Oregon