AN ABSTRACT OF THE THESIS OF
Gordon Henry Kruse
Master of Science
for the degree of
in
Fisheries and Wildlife
Title:
Relationships between shelf temperatures,
the
coastal uplling index, and
17 April 1981
on
presented
English sole
sea
coastal
(Parophrys
level,
vetulus)
spawning activity ofT Oregon
Redacted for privacy
Abstract approved:
Ty1er
Albert V.
The purpose of this study was to caupile records of
(Parophrys
spawning activity in Oregon waters and to attempt
vetulus)
to identify the environmental factors responsible for
subobjective
A
spawning.
relationships
exist
sole
English
between
was
Oregon
to
determine
coastal
the
control
whether
of
statistical
shelf temperatures,
sea
level, and the upwelling index.
waters
Spawning times for English sole in Oregon
for
13
were
estimated
years using data of adult female gonadal condition and surveys
of pelagic larvae and newly-settled berithic juveniles of known
The
breeding
variable.
season
of
this
species is
ages.
extremely protracted and
Most spawning occurs between September and
spawning lasting 1-3 months within this period.
doci.nented to occur in all seasons of the year.
April, with
peak
Some spawning has been
ta sets of barometric pressure, the upwelling index, photoperioa
and continental shelf temperatures along the Oregon coast were obtained
to exanine possible interrelationships with the English sole spawning
records. Because data of near-bottom temperature were not sufficiently
complete,
I investigated the relationships between the available
sporadic records of temperature and two regularly measured variables,
coastal sea level and the upwelling index.
temperature
at
a moored
of deep
current meter (Poinsettia), sea level at
Monthly means
Newport, Oregon, and the coastal upwelling index at
to
found
be
significantly
(P<O.05)
5°N,
correlated over
125% were
1972-1974.
Significant correlations were also found between both directly observed
values and calculated anomalies of deep temperatures at two repeated
hydrographic stations (NH-5 and NH-iS), and monthly mean sea
Neah
Bay,
1959-1969.
Washington and
level at
the upwelling index at 145°N, 125°W during
The geometric mean estimate of the functional
regression
best describes each linear relationship between variable pairs.
Since
the relationships between temperature and sea level were slightly
stronger than those between temperature arid the upwelling index, Neah
Bay sea level data were used to generate time
series of a
bottom
temperature index, which was then incled into the environmental data
set.
Of
the environmental variables considered, shelf temperature
(inferred fran the relationship with Neah Bay sea level)
appeared to best account for variations in Friglish sole spawning
dynamics
activity.
Three hypotheses about temperature control of spawning were
proposed
and
described
by
relationships were incorporated into
driven
by time
of derived
series
a
between
simulated
and
historical
which
simulation model,
bottom
Hypotheses were parameterized arid evaluated,
These
relationships.
mathematical
spawning
estimates.
temperature
based
agreement
the
on
The
records.
hypotheses which collectively explained most of the
was
three
variability
in
observed spawning are: (1) the rate of gonadal development is inversely
related to simmer bottom temperatures; (2)
temperatures below about
activity
also be used
in
inhibited
by
The model can
be
used
to
simul ate
years of no empirical spawning records and can
for
the
is
7.8 C; and (3) spawning is delayed by rapid
increases in bottom temperature.
spawning
spawning
design of laboratory experiments
validating the proposed hypotheses.
capable
of
Relationships between
elf Temperatures, Coastal Sea Level,
the Coastal Upwelling Index, and
glish sole (Parophrys vetulus)
Spawning Activity off Oregon
by
Gordon Henry Kruse
A THESIS
submitted to
Oregon State University
in partial fulfillment of
the requirnents for the
degree of
ster of Science
Completed April 1981
Cornmencenent June 1981
APPROVED:
Redacted for privacy
Associate Professor of fisheries
in charge of major
Redacted for privacy
iead of Dertznent of Fisheries and Wildlife
Redacted for privacy
Dean of
aduate Schoqi
]te thesis is presented
17 April 1981
Typed by researcher for
Gordon Henry Kruse
Acknowi edgnents
I wish to express my sincere appreciation to my advisor,
Tyler,
and my minor
professor,
Adriana
Huyer,
contributions of tine, advice and encouragenent.
have been
possible without their efforts.
for his inputs as a committee manber.
valiable assistance
I
Katherine
for their generous
This thesis would not
I also thank Howard Horton
am
truly grateful
for
the
offered by William Barsa, George Boehlert, Robert
Demory, James Hell, Robert Hayman, Gary Hewitt,
Laroche,
Albert
Myers,
Bruce
Earl
Krygier,
William Pearcy,
t'bndy,
Joanne
Sally
Richardson, Andrew Rosenberg and Carl Sebreak.
William Gilbert constructed
Pittock
supplied
the
Fig.
1
and
)4
of part
II.
Henry
sea level data and Indrew Eakun provided recent
values of the upwelling index.
I am indebted to the
following
people
who made part III possible by making their data available to me during
the courses of their
Laroche,
Sally
studies:
Richardson,
Gary
William
Hewitt,
Pearcy,
Earl
Krygier,
Joanne
Andrew Rosenberg, Bruce
tindy and Katherine Myers.
of
&ipport for this project was provided by the Pleuronectid
Project
the
Program
Oregon
(Ok8MO11it).
State
University
Sea
Grant
College
TABLE OF CONTENTS
I.
Introduction
II.
The Relationships between Shel f Temperatures,
.
1
Coastal Sea Level arid the Coastal Uplling
Index off Newport, Oregon (coauthor: Adriana I-Juyer) ......... 3
Abstract .................................................
Introduction............................................. 5
Description of Observations .............................. 7
DataAnalysis ........................................... 12
Discussion.............................................. 23
III. Influence of' Physical Factors on the
glish sole
(Paroptirys vetul us) spawning season (coauthor:
Albert V.
1'yler) .......................................... 26
Abstract ................................................ 26
introduction............................................
Compilation of Spawning Times ............................
Deve1o1ent of Spawning Simulation ?"bdel ................
Discussion..............................................
28
30
37
50
IV.
Conclusions ................................................ 54
V.
References ................................................. 55
VI.
Appendix ................................................... 62
LIST OF FIGURES
II. The Relationships between ielf temperatures, Coastal
Sea Level and the Coastal UpweJ.ling Index off Newport, Oregon
Figure
1
2
Page
Location of oceanographic data collection sites.
9
Daily values of (a) 80 m temperature at station Poinsettia;
(b) Newport sea 1 ev el; and (c) the upl 1 irig index at 145°N,
10
Time series of observed and inferred near-bottom temperatures
during 1959-1969 for (a) NH-5 and (b) NH-15 using Eqs. (3)
and (5) of Table 5.
21
Inferred mean monthly temperature near the bottom at NH- 15
from Neah Bay sea level using Eq. (5) and observed mean monthly
22
125% for 16 December 1972 to 3 April 19714.
3
U
temperature at Poinsettia.
III. Influence of Physical Factors on the English sole
(Parophrys vetul us) Spawning Season
Figure
1
Location of data collection sites.
Page
ta used in this study
came fran gonads of females landed at Astoria and Newport,
larval surveys along transects between Astoria and Cape
lanco, and juvenile surveys in Yaquina Bay.
2
Inferred spawning times for English sole using data of adult
female gonadal condition arid surveys of' pelagic larvae and
32
36
benthic j uvenu]. es.
3
Relationship between spawning records and the mean bottom
temperature index for the sunmer (May-August) prior to
The proposed relationships beten mean sunnier
temperature and the completion date of gona1 deve1oent
for (a) most individuals and (b) the most rapidly developing
individuals are also shown.
spawning.
142
Figure
14
Hypothesized relationship between the rate of increase
in mean monthly tnperature and the delay in the spawning
Page
245
time.
5
Flow chart of spawning simulation model.
46
6
Comparison between observed (0) spawning records and those
predicted (F) nen all three temperature hypotheses are
incorporated into a simulation model driven with bottom
temperature estimates.
148
LIST OF TABLES
II. The Relationships between elf Temperatures, Coastal
Sea Level and the Coastal Uplling Index off
Newport, Oregon
Table
1
Page
Descriptions of oceanographic data used in this stxiy.
8
2
Correlation coefficients among daily and monthly means of
80 m temperature at Poinsettia, sea level at Newport and
the upwelling index at 45°N, 125°W.
13
3
Correlation coefficients between monthly observations of
16
near-bottom temperatures at NH-5 and NH-iS and monthly
means of sea level at Neah Bay and the uplling index at
145°N, 125°W over 1959-1969 for both the observed monthly
means and the monthly anomalies.
!
Long-tenn means of temperature observations, Neah Bay sea
level and the upwelling index at 45°N, 125°W by month.
17
5
Regression equations between observations of temperature,
sea level and the upwelling index based on 1959-1969 data
for both the observed monthly means and the monthly mean
20
ancinal ies.
III. Thfluence of Physical Factors on the English sole
(Parophrys vetulus) Spawning Season
Table
Page
Sources of data used to infer English sole spawning times.
31
2
Percentage of spent English sole females landed at Astoria
as determined by the Oregon Department of Fish and Wildlife.
33
3
Total range of values of barometric pressure, day length
and indices of upwelling and bottom temperature over
all months and over only those months in 4nich spawning
38
1
occurred.
Relationships between &ielf Temperatures, Coastal Sea Level,
the Coastal Upwelling Index, and English sole (Parophrys vetulus)
Spawning Activity off' Oregon
Introduction
The reproductive
believed
to
be
strategies
by a given
utilized
species
those which result in rnaxthrn.m reproductive success by
growth and survival.
favorable
conditions
best ensuring that the young will encounter
which
Since reproductive
success
is
of
investigations
biological basis for the
stocks.
an
important
this
management
regard
this
In
the
reproductive strategies and survival are related.
fecundity,
productivity,
for
Annual variations in reproductive success (i.e
year-class strength) reflect deviations of those factors to
species'
are
canponent
of fishery
aspect constitute part of the
of
cctmnercially important
studies of
spawning
activity and
fish
its
envirormtal control can be of value to the management of species with
variable
spawning times.
This is particularly true for those in stnich
year-class strength is determined by factors operating
time
period
during the early stages of develonent.
the spawning time partly determines whether the
young
over
a
short
For these fishes
will
encounter
favorable conditions.
English sole,
species
to
the
?arophrys
bottom
United States and Canada.
trawl
vetulus,
is
fisheries
a
cctnmercially important
along the west coast of the
The breeding season of this species is quite
2
1959; Jow 1969)
protracted (dd 19140; Harry
Richardson 1979; Hayman and Tyler
the
survival
rate of young
1980).
English
and variable (Laroche and
The spawning time may affect
soles, because year-class strength
appears to be largely determined by environmental conditions prevailing
prior
to
spawning
(Hayman
history stages (Ketchen 1956).
variations in
English
sole
and Tyler
1980) and
during the early life
investigation of the
For this reason
activity may lead to a better
spawning
understanding of their recruitment process and to improved management.
The purpose of this study was three fold.
to
The first objective was
canpile records of English sole spawning activity in Oregon waters
using data of gonadal condition from commercially caught females landed
at
Oregon
ports, size distributions and known growth rates of pelagic
larvae collected off the Oregon coast,
juveniles in Yaquina Bay, Oregon.
and
surveys
of newly-settled
The second objective was to attempt
data
to identify interrelationships between these spawning records and
of'
barcinetric
pressure,
the
upwelling
index,
temperature on the Oregon continental shelf.
observations
re
day length, and deep
temperature
Because deep
not sufficiently complete, a third objective was to
determine whether temperature is statistically related to two regularly
measured variables (coastal sea level and the
uplling
index), so that
a continuous temperature index could be developed.
This thesis is comprised of two independent papers.
The first
is
coauthored by Adriana Huyer and is designed to meet the third objective
of this sti.r.ly.
The second paper is coauthored by Albert V.
is directed towards the first two objectives.
Tyler
and
3
The Relationships between
iel f Temperatures, Coastal Sea Level and the
Coastal tjpwelling Index off Newport, Oregon
(Coauthor: Adriana Huyer)
Abstract
Sparse bottc4n temperature records have hampered
with
denersal
fish
dealing
species living on the continental shelf along the
northwest coast of North America.
between deep temperatures
on
regularly measured variables,
index.
studies
investigated
We
the
the
relationships
Oregon continental shelf and two
coastal
sea
level
and
the
upwelling
Monthly means of deep temperature at a moored current meter
(Poinsettia), sea level at Newport, Oregon and
the
coastal
upwell ing
index at L5°N, 125°W were found to be significantly (P<O.05) correlated
over 1972-1974.
Significant correlations were also found between
both
directly observed values and calculated anomalies of deep temperatures
at two repeated hydrographic stations (NH-5
mean
sea level
Lt.5°N, 125°W
at
during
temperature and
Neah
Bay,
1959-1969.
sea
level
and
NH-is),
and monthly
Washington and the upwelling index at
Since
the
relationships
between
were slightly stronger than those between
temperature and the upwelling index, sea level is a more reliable deep
temperature indicator.
The geometric mean estimate of the functional
regression best describes each linear
pairs.
relationship between variable
The regression equations between temperature and sea level (or
the upwelling index) and between
their
anomalies can be
useful
to
reveal
bottctn
observations.
tnperature
dynamics
during
periods
of sparse
5
Introduction
Fishery stock-recruitment models must
factors in order
incorporate
sufficiently account
to
variations of many fish species.
for
environmental
year-class strength
The most cctnmonly considered physical
factor is temperature, owing to its known effects on many physiological
processes and its linkage to other factors such as
and
productivity.
Temperature
appears
to
salinity,
currents
significantly affect the
recruitment success of nunerous fishes (Ketchen 1956; Lett et al. 1975;
Sutcliffe et al. 1977; and others). Temperature may act on reproductive
success by controlling the amount of energy directed
gonadal
in maturing
products
and ivis 1967)
processes associated
production
influence
conditions.
(Bagenaj.
spawning
therefore the timing of eggs and larvae
abiotic
and
with growth
and gonadogenesia (Weibe 1968) or indirectly
through a linkage with food
variations may also
growth
This control may be exercised
adul ts.
directly through the physiological
(Warren
into
1967).
Temperature
dates (de Vianing 1972) and
with
flucttating
biotic
and
In addition temperature may regulate the survival
of' the young by influencing their food availability (Lett et al. 1975),
growth
(Zweifel
(Ketchen 1956).
and
Lasker 1976)
and
For these reasons bottom
duration of pelagic
life
temperatures can be very
important to demersal fish species.
Unfortunately,
northwest
sporadic
bottom
records
along
the
coast of North America have Impeded fishery correlation and
simulation stiies in this region.
to
temperature
Therefore it
uld be advantageous
identify a regularly measured environmental factor 'thich is well
6
correlated with deep temperatures on the continental shelf. Of f Oregon
surface temperatures are not representative of deep temperatures (at
least not seasonally), in part because of the seasonal influences of
coastal upwelling and the Columbia river plume (Huyer
1977).
Previous
sti.iles have exanined relationships between surface or near-surface
temperatures and such factors as sea level, wind, atmospheric pressure
and upwelling (Roden 1960; Bourke
1969;
Fisher
1970;
and
others).
To
our knowledge no stixties have found relationships between bottom or
near-bottcm shelf temperatures and these or other factors.
The apparent similarity between time series of coastal sea level
and
40 in temperature at a current meter mooring located over the 100 in
isobath off Newport, Oregon (Huyer and nith 1978: Fig. 2) suggested to
us
that perhaps sea level might be a good index of bottom temperature.
A temperature-sea level relationship would be usefti because sea level
measurements are readily avail able to researchers and are recorded at
regular intervals.
We
also considered the possibility that deep
temperatures are related to the coastal upwelling index (Eakun
1975),
1973;
because this index is proportional to an important driving force
in the Oregon oceanographic system, the al ongshore wind stress.
index is already
this region.
widely
This
used in fishery and oceanographic stties in
Values of the upwelling index are available at regular
intervals to provide information about upwelling system dynamics.
The
purpose of this sti.y was to determine thether coastal sea level or the
upwelling index can be used to infer near-bottom temperatures on the
continental shelf off Oregon.
7
Description of Observations
period
We obtained a data set for the
2 April
(Table 1),
197'4
which
17 December
through
1972
daily means of the 80 m
included
temperature at a mid-shelf current meter mooring
Poinsettia
named
(Fig. 1), the sea level at nearby Newport, Oregon and daily and monthly
values of the coastal upwellirig
index
at
current
The
45°N, 125°W.
observations fran this mooring and the simultaneous sea level data have
been described by Huyer et al.
recorded
at
was recorded
on an
was
Using
the
6-hourly records
abridged
and
we
per day).
(
calculated
to
daily means of
of the upwelling index were reported by Bakun (1973; 1975).
computation of this index involves
(computed
6-hourly
Daily (Fig. 2) and monthly values
temperature and sea level (Fig. 2).
monthly values
Sea
filtered to remove tides,
basis,
hourly
adjusted for the inverted barometer effect
values.
80 in
intervals of 10 or 20 mm, filtered to remove diurnal and
shorter oscillations and reduced to 6-hourly values
level
at
The temperature
(1 979).
a
nonlinear
tenn
(wind
Since the
stress),
fran monthly mean pressure fields) are not
identical to the monthly means of the daily values (based
on 6-hourly
computations), although they are highly correlated (Bakun 1973).
A second data set spans June 1959 through November 1969
and
(Table 1)
includes observations of near-bottom temperatures fran repeated
hydrographic stations at NH-5 and
hourly
(not
Washington
hydrographic
NH-iS
(Fig. 1),
monthly means of
adjusted for atmospheric pressure) sea level at Neah Bay,
and monthly values
of
Bakun's
upwelling
stations were described by Huyer (1 977).
index.
The
Temperature was
Table 1.
Location
Descriptions of oceanographic data used in this study.
Observation Total depth Sample depth
(m)
(a)
Poinsettia
Temperature*
Dates
Data Type
No. of values
80
17 Dec 72
-
2 Apr 711
Daily means
'400
-
-
17 Dec 72
-
2 Apr 714
Daily means
472
-
-
17 Dec 72
-
2 Apr 711
Daily values
472
100
(44°115'tl,
124° 17. 5W)
Newport, Oil
Sea levelt
115°N 12W
Upwelling
index
Temperature
NH-S
(44°39.1'N,
55
50
29 Jun 59
-
60 - 90
29 Jun 59
-
140 -
Monthly values
17
21 Nov 69
Approx.
78
21 Nov 69
ApprOx.
82
12i1°10.6W)
Temperature
NH-IS
100
('14°39.1N,
8 per yr.
8 per yr.
124°27. '4 W)
Mesh Bay, WA
Sea levelt
145°N 125<'W
IJpwelllng
index
-
.-
Jun 59
-
Nov 69
Monthly means
126
-
-
Jun 59
-
Nov 69
tlonthiy values
126
*This record contains two gaps (23 March 1973 through 12 April 1973 and 25 June 1973 through 14 August 1973). These
gaps were also placed into the sea level and upwelling records for all statistical analyses involving temperature
and these other variables.
t5ublract 1.1433
in
from Newport sea level to refer to mean lower low water.
+Subtract 0.701 in froni Neah Bay sea level to refer to mean lower low water.
1
Figure 1.
Location of stations Poinsettia (P), NH-5 and NH-15 in
relation to Newport, Oregon and Neah Bay, 1ashington. This figure is a
Lambert conforia1 conic proj ectiori made from the Juan de Fuca plate
map, published by the Pacific
Geoscience Center, Sidney, British
Colunbia.
"-I
YT
!Z5
123a
10
tily values of (a) 80 m temperature at station Poinsettia;
Figure 2.
(b) Newport sea level; and (c) the upwelling index at 45°N, 125°t4 for
16 December 1972 to 3 April 19714.
2
6
U
0
w
0
2
L&i
k
33r
3.'
-J
LA.i
>
2.9
2.7
2.5
2.3
200- C
2OO
LJ
-400-
0..
I!
600DJFMAMJJASONDJFMA
I
I
I
MONTH
I
I
I
recorded an average of eight times per year at each station.
than
one
averaged.
observation
was
available
for
a
given month, they were
A total of six monthly values at each
actually means of 2it observations.
When more
NH-5
and
NH-15
are
12
Data P1nalysis
Time series plots (Fig. 2)
arid
scatter diagrams (not
shown)
suggested the existence of linear relationships between Poinsettia deep
temperature, Newport sea level and the upwelling index at
45°N, 125°W
for 1972-197. To verify this we calculated simple linear correlation
coefficients for daily and monthly mean values of the variables
(Table 2).
Since time
series of each variable are autocorrelated,
consecutive observations are not independent and the significance of
these crosscorrelations cannot be determined by simply consulting
standard statistical tables with N1 -2 degrees of freedaii, where N1 is
the number of observations. In order to document statistical
significance we could calculate normalized correlation coefficients
(Sciremanmano 1979) or adjust the number of degrees of freedan by first
estimating
the
actual
of
number
independent
observations
latter
procedure and used the Bayl ey and Hainmersi ey formulation because it is
more conservative, i.e.
results in fewer degrees of freedcm. The
(Bayley and Rarnmersley l946;
Davis 1976).
chose
We
the
Bayley and Hammersley formulation can be written as
N
N-1
N.+ 2
3:1
(N1-j) p (j.t) 2
(1)
where n! is the number of independent observations estimated fran the
autocorrelation (p(jat)) at
lag jt of variable i for i=1,2.
number of degrees of freedan (DF) can then be taken as
DF : mn(n,n) - 2.
(2)
The
13
Table 2. Correlation coefficients (nunber of observations and degrees
of freedan in parentheses)
among daily and monthly means of 80 in
temperature (T) at Poinsettia (P), sea level (SL) at Newport (NP) and
the uplling index (UI) at 145°N, 125°W.
a. Daily means
UI
SL(NP)
_0.609(1472,12)*
O.80k6(1400,12)**
UI
T(P)
O.140k0()4OO,12)
b. Monthly means
SL(NP)
UI
T(P)
UI
_0.8909(17,5)**
0.9085(15,5)** _0.8255(15,5)*
*P<0. 05
**p<0Q2
114
In practice the suimatiori in (1) is carried out for j:1,2,...,K,
where
K is some nunber large relative to the lag at which the autocorrelation
beectnes zero. We took K to be a value such that
DF
reached an
asymptote.
Since the 80 in temperature series at Poinsettia has two gaps,
is
it
not possible to calculate ri' for the entire tnperàture record.
the longest noninterrupted
Over
2 April
period
through
(15 August 1973
the values of n* calculated for the temperature and sea
1 9714)
level records were nearly identical. Therefore we assuned the value of
n
for temperature over the entire record
through 2 April
19714)
length
(17 December 1972
to be equal to the value of n! calculated for sea
level by Eq. (2).
Correlations among Poinsettia temperature, Newport sea level and
the upwelling index (Table 2) were significant in all but one case at
the
95%
confidence level using
tables of critical values of
the
correlation coefficient (Zar 1 9714: Table D. 21). The exception (daily
is not surprising because the short
fluctuations are large relative to those of
temperature versus upwe].ling)
period upwelling index
temperature, particularly during the winter months (Fig. 2a,c).
Observations
at.
hydrographic
stations
made
repeatedly
(approximately 8 per year) at NH-5 and NH-i 5 (Fig. 1) between June 1959
and November
1969
allowed us to determine whether the associations
suggested by the Poinsettia data remained true for longer time scales.
For this time period, we had access to monthly mean sea level data from
Neah
Bay (the Newport sea level record does not start until
1967),
and
15
monthly values of Bakunz upwelling index at 145°N, 125°W.
period deep temperature
individual
fluctuations
are
observations of temperature
approximations
to
the monthly means.
short
Since
usually small (Fig. 2a),
should generally be
On this basis
good
corresponded
NH-5 and NH-15 temperature observations with monthly means of sea level
and
the uplling index.
Because lagged correlations on daily values
during 1973 indicated that the upwelling index leads sea level by about
2
days and deep temperature by 5 days, no lagging of monthly means was
necessary.
Scatter diagrsms (not shown) again suggested linear
between monthly values of each variable pair.
positive correlation exists between
relationships
A significant (P<O.02)
temperature
and
sea
level
and
significant negative correlations exist between these ti
variables and
the coastal uplling index
freedan
by the
calculated
(Table
procedure
used
3a).
Degrees of
earlier.
were
The conservatism of the
Bayley and Haxmnersley (1946) formulation is reflected by the fact
there
approximately only one
is
that
degree of freedan per year of data
(Table 3a).
In order to
correlations,
eliminate
we
variable (Table U).
estimated
first
the
effect of
seasonal
For temperature the long-term monthly means
by averaging
the
cc*nputed the long- term monthly means for each
individual
observations by month
duration of the records (June 1959 - November 1969).
the
cycles on
were
for
the
For sea level and
upwelling index, a ccinparable period was chosen sh that an equal
number of values (11) were used for each month (June 1959 - May
1970).
16
Correlation coefficients (nunber of observations and degrees
of freedi in parentheses) between monthly observations of near-bottom
temperatures (T) at NH-5 and NH-iS and monthly means of sea level
(SL)
at Neah
ay (NB) and the upwelling index (UI) at 45°N, 125°W over
1959-1969 for both the observed monthly means and the
monthly
Table 3.
anana]. ies.
a. Observed values
UI
SL(NB)
UI
T(NH-5)
T(NH-15)
T(NH-5)
_O.8563(126,9)**
o.7k25(78,9)**
O.7168(82,9)**
O.71O3(78,9)**
_O.6791(82,9)**
o.8579(78,9)**
b. InomaJ.ies
SL(NB)
UI
T(NH-5)
_O.6198(126,65)**
0.14662(78, 65)**
T(NH-15)
0.35)414(82,65)**
*P <0.05
**p<002
UI
T(NFI-5)
0. 3)432(78, 65)**
_O.2)428(82,65)* 0.7156(78,65)**
17
Table U.
Long-term means of temperature (°C) observations, Neah Bay
at
sea level0(m) and the upwelling index (m3/s/100 m of coastline)
45°N, 125 W by month.
Monthly means for temperature are based on
Means for sea level and
observations during June 1959 - November 1969.
uplling are based on a comparable period consisting of 11 values for
each month (June 1959 - May 1970).
T(NH- 5)
T(NH-15)
SL(NB)
UI
Jan
Feb
Mar
Apr
9.94
9.12
9.35
8.89
9.59
-92.55
-58.00
-19.73
May
8.143
7.149
2.11
2.07
2.01
1.93
1.88
1.86
1.85
1.89
1.92
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Annual mean
9.149
7.69
10.10
9.29
8.73
8.72
7.78
7.60
7.52
7.76
9.05
11 . 61
9. 87
2. 07
10.68
10.26
2.13
52.55
20.82
-27.18
-65.82
-88.73
9.03
8.81
1.98
-7.09
7.19
7.143
2.00
12.145
39.73
57.00
814.36
1I
We
then
subtracted
these
long-term monthly means frau the
values to obtain monthly anomalies.
all
pairwise
time
Simple linear correlations between
canbinations of anomalies
confidence level (Table 3b).
observed
are
significant at the 95%
The nunber of degrees of freedom
between
series of ananalies is much higher than between absolutes, since
the elimination of seasonal cycles reduces the autocorrelation of each
time series.
Based on these results, it is reasonable to attempt to infer
deep
temperatures fran either sea level or the upwelling index, but slightly
more confidence
Although
uld be placed upon those
variables
were
inferred
fran
sea
level.
clearly definable as dependent
not
independent, we initially conducted simple linear
regression
or
anal yses
on the 1959-1969 records with temperature as the dependent variable and
sea level or the upwelling index as the independent variable.
we
found these regressions to yield undesirable biases.
However,
Plots of time
series (not shown) conspicuously revealed that at high sea levels
upwelling
index)
inferred
temperatures
observed temperatures and at low sea
inferred
temperatures
(low
were considerably lower than
levels
(high
upwelling
were much higher than those observed.
index)
Reversal
of dependent and independent variables resulted in the converse.
This
phenomenon exists since a regression of Y on X minimizes the stin of the
squares of the vertical distances of the
while
a
regression of X on
Y
observations
to
the
line,
minimizes the sun of squares of the
horizontal distances of the points to the line.
19
To circunvent the obvious biases caused by either
simple linear
regressions, we ccmputed the "geometric mean estimate of the functional
regression", hereafter "GM regression", strongly advocated
(1973)
flicker
particularly in situations in *iich variability due to natural
causes outweighs that due to measurement.
line
by
which minimizes
vertical
distances
intermediate
the
fran
sun
The GM regression yields a
of' the products of the horizontal and
each observation
to
the
to the regressions of Y on X and X on Y.
equation is easily calculated
Iran
simple
anal yses
linear
regression
regression equations are presented
statistics
in
is
This regression
normally provided
by
Computed
GM
(Ricker
Table 5
and
line,
1973).
with approximate
95%
confidence intervals for the slopes.
To visually demonstrate the utility of the
temperature
and
sea
level,
relationship between
plots were made of observed and inferred
mean monthly near-bottani temperatures at NH-.5 and NH-i 5
and
(5),
respectively (Fig. 3).
using
A test of this relationship would be
to use sea level to estimate temperature for a different
Since stations NH-iS and Poinsettia are close (15
over the same isobath (100 in),
using
data
one
Eq. (3)
would
expect
time
period.
n apart) and located
Eq. (5)
(developed
from 1959-1969) to reasonably estimate deep temperature at
Poinsettia during 1972-1974.
A plot (Fig. k) of inferred monthly mean
temperatures
observed
at
NH-iS
and
monthly mean
Poinsettia indicates that this is the case.
temperatures
The average error is
0.56 C, an error which should be acceptable for most purposes.
at
only
20
Table 5.
Regression equations with 95% confidence intervals of the
slopes between observations of temperature (00), sea level (m) and the
upwelling index (rrL3/s/1 00 m of coastline) based on 1959-1969 data for
both the observed monthly means and the monthly mean anomalies.
a. Observed values
T(NH-5)
T(NH-5)
T(NH-15)
T(NH-15)
T(NH-15)
-15.8133 + (12.5108 ± 2.17)43) X SL(NB)
8.8278 - (0.0228 0.00)42) X UI
(3)
-9. 1761 + (9.08)41
1.6019) X SL(NB)
8.6901 - (0.0168 ± 0.0031) X UI
(5)
(6)
(7)
2.2118
+
(0.729)4 ± 0.0972) X T(NH-5)
b. 4noma1ies
T(NH-5)
T(NH-5)
T(NH-15)
T(HH-15)
T(NH-15)
-0.0863 + (10.6825 ± 2. 1659) X SL(NB)
-0.0319 - (0.0205 ± 0.00)4)4) X UI
-0.0213 + (8.7)402 ± 1.8257) X SL(NB)
-0.0010 - (0.0176 0.0038) X UI
=
0.0000 + (0.8811 ± 0.1)411) X T(NH-5)
(8)
(9)
(10)
(11)
(12)
21
Time series of observed (dots) and inferred (solid line)
Figure 3.
near-bottom tanperature during 1959-1969 for (a) NH-.5 and (b) NH-iS
using Eqs. (3) and (5) of Table 5.
aS
ri
is:
C)
w
b
a.
w12
I-
LI
1959 1960 $961
1962 $963 $964 $965 $966 196T $968 1969
YEAR
22
Figure L
mean monthly temperature near the bottom at NH-i 5
from Neah Bay sea level using Eq. (5) (dashed line) and observed mean
monthly temperature at Poinsettia (solid line) with ±1 SD of the
Monthly
predicted (light stippled) and observed means (dark stippled).
means and standard deviations of Poinsettia temperature are calculated
using the full complement of daily values for each month except
December 1972 (15 days), and five months in 1973: March (22 days),
April (18 days), June (2 days), July (0 days) and August (17 days).
Inferred
12
310
Ui
4
7
6
D
J
F
M AM
J
J
1973
A SO N
0
J FM
1974
23
Discussion
Studies investigating the dynamics of marine demersal fishes have
established a demand for long- tenil continuous bottom temperature
records.
Since
attempted
such records are not
usually available,
have
we
to meet this need by establishing relationships between
near-bottom temperature
(measured only
sporadically)
other
and
variables which are routinely measured on a continuing basis.
e found that near-bottom temperatures over the continental shelf
off central Oregon are significantly correlated with coastal sea level,
and also (though to a lesser extent) with the coastal upwelling index
from
computed
the large scale wind field.
th of these correlations
hold for monthl y mean data. The correlation with sea level al so holds
for daily means of' low-pass filtered sea level (i.e., with the diurnal
and semi-diurnal tides removed from the sea level data before the daily
means
are computed).
The
correlation between temperature and the
upwelling index does not hold on this shorter time scale.
found
significant correlations
We
between the monthly anomalies
also
of
near-bottom temperature and those of sea level and the upwelling index.
This correlation among
anomalies indicates
actually correlated with each other, instead of
that the variables are
only
appearing to be
correlated because they all have seasonal cycles.
The observed correlations were not entirely unexpected.
upwelling
Coastal
is well known to affect shore temperatures along the Oregon
coast, and one would naturally expect it to also affect the near-bottom
temperatures over the inner continental shelf and perhaps over the
24
mid-shelf.
over
Coastal sea level varies with the
density
changing
field
the continental shelf and slope on both the longer (monthly mean)
time scale (Reid and Mantyla 1976) and the shorter (days-to-weeks) time
scale (Buyer et al.
and
one might
salinity,
correlated
Since density is a function of temperature
1979).
with
coastal
expect
temperatures
near-bottom
sea level.
to
be
Our results show that the bottom
temperatures are actually better correlated with sea
with
than
level
Bakun's upwelling index.
We chose linear equations, tenried Q4 regressions, to describe
relationships
between
variables
the
Simple linear regression
(Table 5).
eapations were not used due to biases which were particularly expressed
at
extreme values of each variable.
near-bottom
temperature
relationships
with
and (10)
be
can
locations.
sea
used
at
NH-5
or
For
estimating monthly mean
one
NH-15,
should
use
level: Eq. (3) or (5), respectively.
for
estimating monthly anomalies
the
Eq. (8)
at
these
The upwelling index is also useful to infer monthly values
of temperature, particularly since gaps
in
the
Neah
Bay
sea
level
record do exist in some years.
Al though our
quantitative
results
(the
constants
in
Tables
! and 5) are probably valid only for the locations from which they were
derived (i.e , NH-5 and NH-iS over the continental shelf off
Oregon),
we
applicable.
expect
the
qualitative results
Anomalies of coastal sea
distances (Enfield and Allen 1980).
correlated
with
sea
1 evel,
level
are
to
Newport,
be more generally
coherent
over
long
Since near-bottom temperatures are
anomalies of near-bottom
temperature
25
probably also have high alongshore coherence.
If so, the Neah Bay sea
level will be a qualitative indicator of bottom temperatures over
shelf
along
California.
much
of
shing ton,
Oregon,
and
perhaps
Our quantitative results will certainly prove
simulation models of fisheries off central Oregon.
the
northern
useful
for
Influence of Physical Factors on the
&glish sole (Parophrys vetulus) Spawning Season
Tyler)
(Coauthor: Albert V.
Abstract
Spawning tines for &iglish
waters
were
sole
(Parophrys
vetulus)
in
Oregon
estimated for 13 years using data of adult female gonadal
condition and surveys of pelagic larvae and benthic juveniles of known
ages.
The spawning season for this species is extremely protracted and
variable.
peak
Most spawning occurs between September and April, with
spawning lasting 1-3 months within this period.
Some spawning has been
docunented to occur in all seasons of the year.
Variations in the
largely attributable
glish
season
spawning
sole
about
temperature
described by mathematical
incorporated
into
be
to
to variations in a continental shelf temperature
index (developed fran a relationship with Neah Bay sea
hypotheses
appear
control
of
relationships.
Three
level).
spawning were proposed and
These
relationships
were
a simulation model, which was driven by tine series
of the bottom temperature index
for
those
years
in
which
spawning
records
were
based on
the
records.
The three temperature hypotheses which collectively explained
available.
Hypotheses were parameterized and evaluated,
agreement between
simulated
much of the variablility in observed spawning
and
are:
historical
(1)
the
spawning
rate
of
gonadal developnent is inversely related to sunnier bottom tanperatures;
27
(2) spawning is inhibited by temperatures below about
spawning
is
delayed
model can be used
empirical
spawning
to
by rapid
simulate
records
and
and
(3)
increases in bottom temperature.
The
spawning
can
activity
also
be
7.8 C;
for
years of no
used in the design of
laboratory experiments capable of validating the proposed hypotheses.
28
Introduction
veral general patterns in the reproductive strategies of marine
fishes have been proposed by
sim (1956), Q.ishing (1969; 1975) and
others. Fishes common to low latitudes or upwelling areas are often
serial spawners and tend to have long
seasons.
In nonupwelling
convergences spring
(and
intermittent breeding
and
regions polet..ard
of
subtropical
the
perhaps some fall) spawners generally breed
only once per year over short (about three months) fixed seasons, while
sunmer spawners tend to be intermediate in these characteristics.
The reproductive strategies utilized
believed
by a
given species are
to be those which result in maxirnun reproductive success by
best ensuring that the young will encounter conditions favorable for
survival.
growth and
Annual variations in reproductive success (i.e
year-class strength) reflect deviations of those factors to which the
species' fecundity, reproductive strategies and survival are related.
Investigations of reproductive success constitute part of the
basis for the managnent of commercially important fish
biological
stocks.
In
this regard studies of spawning activity
and
its
environmental control can be of value to the managnent of species with
This is particularly true for those in which
year-class strength is determined by factors operating over a short
variable spawning times.
time period during the early stages of develoinent. For these fishes
the spawning time partly determines whether the young
favorable conditions.
will
encounter
29
English sole,
species
to
the
Paropkrys
bottom
consistent
fishes
common
suggests
that
is
a
along the west coast of the
because
pleuronectid
with at least one of the general patterns observed for
to
upwelling
areas.
While
examination of ovaries
Hewitt, Oregon
individuals are not serial spawners (G.
quite
comm.),
protracted
time may affect
Harry
(&idd 19)40;
year-class
the
strength
season of this
breeding
the
variable (Laroche and Richardson 1979; Hayman
spawning
commercially important
The spawning strategy of this
State University (OSU), pers.
species
is
fisheries
trawl
United States and Canada.
is
vetulus,
and
1959; Jow 1969) and
Tyler
1980).
The
survival rate of young Fnglish soles,
appears
to
be
largely detennined
by
hydrographic events occurring prior to spawning (Hayman and Tyler 1980)
and during the early stages of develonent (Ketchen
reason
investigation of the
activity may lead
process
to
a better
variations
in
understanding
and to improved managnent.
English
of
sole
their
this
spawning
recruitment
The objectives of this sttdy were
to compile records of English sole spawning activity in
and
For
1956).
Oregon waters
to attampt to identify the environmental factors which control the
time of spawning.
30
Compilation of Spawning Times
Generally the most accurate procedure for the determination of
teleost
spawning
times involves the
representative sample of adult
examination of ovaries in a
fema]. es.
Temporal
variations
in
intensity can be quantified using measures such as the
proportion of females ioh are fully spawned (Harry 1959),
spawning
gonadosornatic
indices ('brse
follicles (Hunter and Goldberg
or the incidence of postovulatory
1980)
1980).
times
Spawning
can
also be
back-calculated fran larval or juvenile size composition data if the
growth rates of these stages are accurately known.
We
access to
had
each of these types of data (Table 1) for the purpose of inferring past
English sole spawning times.
condition of adult P. vetulus females landed at
The gonadal
Astoria (Fig. 1)
Oregon
(Harry
as monitored by the Oregon Fish Commission, now the
Department of Fish and Wildlife
1959).
during
1947-1 951
Since data have not been published on a year-by-year
basis, we present then in Table 2.
Ovaries were also sampled from
individuals landed at Newport from October
by Hewitt
(ODFW),
(1980).
1977
through February
1978
With these data, we used changes in the percentage
of spent females between consecutive sampling dates as a measure of
spawning activity.
Pelagic larval surveys were conducted along a transect off Yaquina
Bay (Fig. 1) during June
1969
1977;
Richardson and Pearcy
Cape
Blanco and
through August
1977),
Astoria in
and
along
1972
(Mundy MS; Richardson
various transects between
rch and April of
1972-1975
(Laroche and
31
Table 1. Sources of' data used to infer English sole spasning times.
Method
Years
Adult goriadal
1947-51
Harry (1959)
1977-78
Hewitt (1980)
1969-72
Mundy (MS)
1971-72
Richardson (1977),
Richardson and Pearcy (1977)
1972-75
Laroche and Richardson (1979)
Source
Condition
Fl anktonic
larval surveys
Benthic juvenile 1970-72
surveys
1977-78
1977-79
Krygier and Pearcy (MS)
Myers (1980)
Krygier and Pearcy (MS)
Rosenberg (1980)
32
Figure 1.
Location of data collection sites. tta used in this study
came fran gonads of females landed at Astoria and Newport, larval
surveys along transects between Astoria and Cape Blanco, and juvenile
surveys in Yaquina Bay.
460
200m
100Dm
)
200Dm
Newport
D
0
Pacific
Yaquina
Ocean
44
Oregon
Cape Olanco
42°N
I
127°W
125°
123°
33
Table 2.
Percentage of spent English sole females landed at Astoria on
each sample date from ODFW data (Harry 1959).
te
Sample Size
18 JAN 19118*
18 MAR 19118
25 OCT
10
16
21
7
14
NOV
19)48
19118
19148*
DEC
DEC 19118*
JAN 19)49
JAN 19)49
MAR
8 MAR
17 MAY
18 NOV
6 FEB
1
37J4
91
9
90
1115
59.0
100.0
0.0
0.0
31.0
210
112.9
19119
167
178
67
19119
73
59.9
72.5
95.5
98.6
99.0
9.2
80.9
96.7
19119
99
19)49
1714
5 APR
21 APR
1950
1950*
1950*
18 NOV 1950*
19 DEC 1950*
378
112
722
255
JAN 1951
9 JAN 1951
31 JAN 1951
158
62
79
1
Percent Spent
1111
95.7
16.0
143.1
514.11
69.14
77.2
*Sample size and percentage of spent females represent averages for
data obtained on several closely spaced collection dates.
314
Richardson
1979).
Laroche
Richardson
and
back-calculated
(1979)
English sole spawning times for several years based on the presence or
absence of larvae in plankton samples and
growth
rates.
it
their estimates of larval
is now known that these growth estimates are
inaccurate and better ones have been obtained using
techniques (Laroche et al. MS).
aging
With these revised growth rates we
were able to back-calculate spawning times from
distributions.
improved
reported larval
size
ae to the variability of ocean currents and of larval
growth and survival rates, spawning dates calculated fran iall larvae
are more accurate than those obtained fran large larvae. For this
reason spawning times which
were
inferred
from
larval
size
distributions were based primarily on the presence of newly-hatched
larvae.
Juvenile Ehglish sole have been collected in Yaquina Bay by a
rnxnber
of studies.
However
these are of limited value for the
back-calculation of spawning times due to the large variation in
juvenile growth rates, particularly for individuals larger than about
25 mm standard length (SL) (Rosenberg 1980: Fig. 14).
we
For
this reason
considered only those juvenile surveys which appeared to adequately
sample newly-settled juveniles,
20 mm SL.
which are individuals
of
about
Growth equations developed for both the pelagic (Laroche et
al. MS) and benthic stages (Rosenberg
1980)
of 0-age English soles show
that individuals of this size average 2.5 months of age. The majority
of 20 mm SL P. vetulus are 2-3 months old, while the youngest
individual found of this size
s 1.7 months old (Laroche et al. MS)
35
and the oldest was 6 months old (Rosenberg 1980).
Using each data
aiglish
type
and
inference
procedure,
we
sunrnarized
sole breeding activity in Oregon waters for 13 years (Fig. 2).
Trends in spawning activity,
probably rather
accurate,
particularly peak
because
sources were usually consistent.
inferences
spawning
from
Certainly some errors
are
times,
different data
exist
due
to
the likelihood that all methods of data collection do not yield samples
representative of the entire Oregon stock.
encountered,
When
we relied most upon inferences from data of adult gonadal
condition, next upon newly-batched larvae, and least upon older
or
were
inconsistencies
newly-settled
juveniles.
larvae
Because subjectivity cannot be entirely
eliminated, we describe the bases for our conclusions in the Appendix.
36
Inferred spawning times for English sole using data of adult
female gonadal condition and surveys of pelagic larvae and benthic
juveniles. Spawning is indicated by a solid line and peak spawning is
shown by the solid shaded areas. Stippled areas denote that data are
inconclusive as to peak spawning times and dotted lines signify that
data are inadequate for any spawning inferences.
Figure 2.
.4:
'947
'949
'949
1950
1950
'95'
1970
1970
1971
'97,
1972
1972
'973
'973
1974
'974
1975
1976
'977
'977
1978
'979
JASON DJ FMAMJ
MONTH
37
Developnent of Spawning Simulation tbdel
cogenous factors which may be involved in the reproductive timing
of marine fishes include photoperiod, temperature, salinity, ocean
currents and food quantity and quality (Brett
1970).
These
variables
may be classified into ultimate and proximate environnental factors
(Sadleir 1973). tlLti.mate factors are those which control reproductive
success through their effects on the survival and growth of the young,
and influence spawning timing through evolutionary adaptation.
These
factors are often investigated by laboratory experiments dealing with
egg and larval survival and by correlation or modelling studies
involving time
series of year-class strength and other fishery and
envirorinenta]. variables. Proximate factors are the envirormental cues
initiating gemetogenesis and subsequent spawning behavior. Temperature
and photoperiod have been found to be the principle proximate factors
in most fishes (de Vlaniing 1972). In addition, the grunion (Leuresthes
tenuis) (Clarke 1925), threadfin (Polydactylus sexfilis) (y 1979),
and many others (Johannes 1978) have spawning closely linked to
lunar-driven tidal cycles.
We attempted to identify the proximate factors responsible for
timing English sole reproduction by comparing the total annt.al range in
magnitude of several envirormental variables with their range of values
over only those months in which spawning occurred (Table 3).
For this
purpose we had access to monthly mean data of barometric pressure at
146°N,
12t°W,
the coastal uplling index at 145°N, 1250W (Bakun
1973),
an index of bottom temperature (derived from a relationship with sea
38
Table 3. Total range of values of barometric pressure,
daylength, and indices of upwelling and bottom temperature
over al]. months and over only those months in which
spawning occurred.
Factor
Barometric Pressure
46°N, 124°W
Upwelling Index
)45°N, 125°W
Bottom Temp.
Index at NH-15
1 'N,
1011.1-1027.0 mb
8.75-15.5 h
Dayl ength
45°N
)4)4039
Total
Range
12)4°27.)4'W
-212 - +106
m3/s/100 m
6.60-11.19 C
Range during
Spawning
1012.3-1027.0 mb
< 14.0 h
most <+3m3/s/100 m
most > 8.00 C
39
level)
at
a station (NH-15) over the 100 m isobath off Newport (Kruse
and Huyer MS), and bimonthly values of dayl ength at
45°N
(Beck 1968).
We chose these variables because they were most likely to be related to
English
sole
temperature
reproductive
cycling.
previously
As
mentioned,
and photoperiod are the common proximate timing factors of
other fishes.
Upwelling could be important to English sole, due to its
relationships
with
current velocity.
it
is
related
variables such as food produciton, temperature and
to local weather patterns.
photoperiod, upwelling, and bottom
reproduction,
since
Comparisons suggested that
temperature might
at
related
be
to
spawning activity occurred only over a portion of
the total range of these variables (Table 3).
occurred
because
Baranetric pressure could also be involved,
daylengths greater
than
Little spawning activity
of light, which at 145°N
h
14
corresponds to the period spanning late April to late August.
Possible
inhibition of spawning activity by cold temperatures is suggested not
only because most spawning seems to occur at temperatures greater
than
8 C, but also since little spawning occurred at values of the upwelling
index greater than 3 m3/s/100 m of coastline.
We
searched
for
other
associations such as correlations between the intensity of spawning and
the magnitodes of these variables, but no
further
relationships
were
English
sole
revealed.
Next we took a more analytical
spawning
approach.
activity is so variable, it is improbable that photoperiod is
responsible for its interannual variation.
is most
Since
likely
to
Instead, shelf temperature
be the primary proximate timing factor.
In other
fishes temperature
can have
at least
two
regulatory effects on
reproduction. It may control the rate of maturation (Hela and Laevastu
1961;
Brett
1970;
and others) or it may act as a releasing factor
(Brett 1970; Hempel 1979;
and others).
Laboratory experiments are
ccmmoni. y
used
to
investigate the
temperature control of spawning. However, since it is desirable to be
able to account for the variable spawning activity of
glish soles in
their natural envirorinent, we attempted a different approach. First we
suggested various hypotheses about temperature control of spawning and
described them by mathematical relationships.
Then we incorporated
these relationships into a simulation model which could be driven by a
time
series of a
bottom
temperature index for those years in which
empirically determined spawning times were available. This
uld allow
us to determine whether spawning times simulated using the temperature
hypotheses are consistent with the spawning times docunented earlier.
Since other physical
factors are correlated with deep temperature
(Kruse and Huyer MS), conclusions about these hypotheses cannot be made
except to evaluate whether or not they are reasonable based on
historical records of spawning activity and
bottom
temperature
dynamics.
The first hypothesis considered is that temperature controls the
rate of gametogenesis. Kreuz et al.
(MS) found a negative correlation
between annual variations in marginal interopercul un growth (an index
of body growth) of
young
glish soles and mean shelf temperature
estimates during the growth season (sunnier). This implies an inverse
41
relationship between
growth rate.
We
suggest that the gonadal develoient rate is also inversely related
to
and
A plot of annual
temperature.
index of bottom
(Fig. 3)
temperature
does
the
a close
reveal
riot
glish sole spawning records and an
during
temperatures
somatic
the
sumner
prior
relationship, although the three
earliest spawning seasons followed three of the four
Even
if maturation
was
solely temperature
spawning time versus temperature
if
temperature must
coldest
a
beyond
plot of
relationship
or increase at a
threshold
certain rate in order for spawning to be initiated.
sufltners.
a
controlled,
uld not reveal a close
increase
spawning
to
From trends in the
dates of the start of peak spawning and the start of minor spawning, we
proposed
relationships
completion date
between mean
summer
temperatures and
the
of gonadal develonent for most individuals (Fig. 3,
line a) and the most rapidly developing individual s (Fig. 3, line b).
We next considered a temperature threshold hypothesis suggested to
us
by
Table 3.
That is, we proposed that even if gonadal maturation
was compi eted, spawning would riot occur until temperature exceeded 8 C.
Both temperature
computer
model.
hypotheses
to
incorporated
Each month of simulation
subintervals of equal length.
temperature
were
was
into
divided
We assigned the estimated
the middle of each month.
FORTRAN
a
into four
mean monthly
The temperature during any
subinterval was estimated by linear interpolation between the values at
the
midpoints of s.ccessive months.
of the bottom
maturity would
temperature
index
and
be completed (lines
We drove this model with records
inferred
the
dates on
/nich
a and b of Fig. 3) and the dates
In
W.ç
__
I
-.
LJ
I
I
I
A
I
M
...............................................................................
_
.............................................
I
S
..............
co.
0
rr
_oo_
111
-u
(J)-.j
A
SPAWMNG ACTIVITY BY MONTH
F
M
J
D
N
0
ctcflC-t
1)
i.o
0 C)
p.
C)CD
c- 0
çt
- U)
I
CD
CD
CD
ct0
ct
CDC
-10
CDp
C).C)..CnCD
çtQ.
Ct
Ca
0(1):3 C)uI).0 C)
0
:30) C1.
-
0P.ctp.
:3p:3
)
i :31-
CD P)
CDCO
)
aCD
'
CD
:3(p.
Z
CD
W(DCI)
.0 Ca -:3
0
0'-
.0
ct:jp.
I-
I-J.1-
CD
CD 0
CD
00 00
0oq-4
CD CD
)'H<.0
(I)
.0
CD 1.4:3 :3
:31.
(DC).
c;ip) c1(D CD
CD(DFc P) 0 i- :3
(DC))
g.g.0 1
0_:31-.
p)o
0-Cct
0
CD
1D,CDCDEF.
CD
CD
BI- 0-
CDCn.CD-.
xJ
CD(I1
01D
c-I.
:3
ctP)
o
g
(DCt 00:3
Ct
0
WOctta
Ct CD
P
HCDI- 0.w
0CDP)CD
CDI-
I-LCD g1) ) 1) Xj
30
FI-'
CD
çt
r
143
thereafter on which spawning
bottom
temperatures
above
uld be started based on the
inferred in this manner were generally in
inferred
fran
records
enpirical
lines a and b
or
improve
correspondence.
the
poor
described
did
clear
became
It
hypotheses were not valid or else that
agreement
with
significantly
not
that
either
environmental
the
those
Adjustments to
earlier.
threshold
temperature
the
of
dates of spawning initiation
The
8 C.
wanning
these
control
of
English sole spawning was more canpi ex.
We next recalled that
the
of change of temperature can
rate
significantly influence fish physiology and behavior.
our data
rapidly
strongly suggested
sets
that
when
Reexamination of
increased
temperature
(>0.95 C per month) just prior to the inferred completion time
of gonadal maturation, spawning activity was delayed.
For example,
in
19147_i 9148 the calculated (using a of fig. 3) compi etion date of gonadal
maturation for most individuals was early
December.
The
temperature
index increased 1 .68 C between November and December, and peak spawning
apparently did not occur until late January.
increases
occurred
1 974-i 975)
just
prior
other
in
in
to
years
which
and
calculated maturation completion dates
(1950-1951,
December
activity.
1972-1973,
and
Large temperature increases (1 .36 C between
1969
and
1 .69 C between October and November
doctinented interruptions in peak
1977) also coincided with the only t
spawning
1970-1971,
peak spawning did not start until sometime later
than the maturation dates.
November
the
Similar large temperature
In
general,
the
magnitude of
increase (>0.95 C) and the number of these
increases
the temperature
between
several
consecutive months appear to be related to the duration of the delay.
In all other years for
spawning
which spawning
data were available, peak
activity began near the calculated dates of completed gonadal
developuent and pre-spawning temperatures increased less
(<0.95 C per month).
rapidly
this we constructed a third temperature
From
hypothesis which incli.des these rate effects (Fig. ).
All three temperature hypotheses were incorporated into a revised
for which a flow chart
is
(Fig. 5). For
expediency we have used shortened notation as follows: t - present
date, t - completion date of gonadal developnent, T(t) - present
spawning model
shown
temperature, Dt - delay in spawning caused by a temperature increase
froni
month
t,
t-1 to month
and
- the actual delay expected
D
considering temperature increases over the past several months.
equation such as t
Pn
t1 is analogous to a FORTRAN assigrnent statement
and replaces the present value of t with the value t+1. The model is
a monthly basis, but spawning dates are interpolated to the
precision of .25 months. In the model the question "Is Spawning
run on
completed?" is answered in to ways.
Peak spawning is completed if it
has occurred for tw months or if the date is later than early
For
of lower intensity, spawning is completed if the present
spawning
date is later than early
y.
The spawning model including
resulted in
y.
good
dates of peak
parameterizat ion
all three temperature
simulated and observed starting
agreement between
spawning.
were
In
necessary
hypotheses
fact,
to
only
minor
changes
in
achieve the best possible
45
Hypothesized relationship between the rate of increase in
mean monthly temperature and the duration of the delay in the spawning
Figure 4.
time.
t+3
=
=
=
C,,
=
;;
t+1
0.0
2.0
1.0
Increase in Temperature between Previous
Month, t-1, and Present Month, t
(in°C)
3.0
Flow ch
Figure 5.
in the text.
147
predictions fran these hypotheses. By adding a constraint that peak
spawning could last no longer than two months for any one spawning year
(July to June), we could also account for the dates when peak spawning
was
terninated.
The
set of hypotheses yielding the best agreement
between simulated and observed spawning activity (Fig. 6)
re: (1) the
earliest possible spawning time is given by b and the earliest possible
peak spawning time is given by a in Fig. 3;
below
7.8 C;
(2)
no
spawning
occurs
and (3) spawning does not occur in a particular month if
the mean monthly temperature increases by more
previous month.
The
length
of the
than
0.95 C
fran the
delay in spawning is linearly
related to the rate of temperature increase (Fig.
14).
Whether or not
spawning actually occurs at this later time would again depend upon (2)
and (3) and the current temperature. Rapid increases in temperature
also cause a cessation of spawning activity if spawning had
would
already begun.
In such a case spawning would restart later according
to these same rules.
We
added
the rule that if peak spawning is
interrupted (by a large temperature rise), it
until one month later.
In our simulations
(lines
a
spawning.
and b
would
not begin again
found that the maturation rate hypothesis
of Fig. 3) set the limits to the earliest dates of
The hypothesis about rapidly increasing temperatures was of
primary importance as it frequently delayed or interrupted spawning.
It also was responsible for the correct mod eli ing of the apparent
bimodal
spawning
seasons.
The
activity in the
1969-1970
and 1977-1978 spawning
threshold hypothesis was usually
responsible
for
L8
Figure 6.
Comparison between observed (0) spawning records and those
predicted (F) when all three tnperature hypotheses are incorporated
into a simulation model driven with bottom temperature estimates. A
distinction is made between spawning (solid line), peak spawning (solid
shaded areas) and probable peak spawning (stippled areas). A dotted
line indicates no data.
-1
948
947
01
948
1
or
1
950
1949:E
1950
-
.
1-
1
1951
970
1
I970F
97!
97IF
l972
972
973
F
1974
973
1
I974Pf
..
977
976
1977
-
1978
1
l978p
11979
A SO M DJ FMAMJ
MONTH
49
terminating spawning activity in the spring or summer.
50
Discussion
Off Oregon Parophrys vetulus spawn intermittently over an extended
breeding
season.
We
have
searched
extensively
for
the proximate
factors regulating English sole reproductive cycling by comparing
of
series
factors.
seems
spawning
activity off Oregon
and
several envirorinental
Barometric pressure, which is related
to play no such role.
time
weather
to
patterns,
Although day lengths greater than 114 h of
of gametogenesis,
light might preclude spawning or cue the initiation
it does not appear that photoperiod can be used to explain large shifts
in English sole spawning times.
accountable
by
bottom
temperature
rel ationship with Neab Bay sea
hypotheses
of temperature
Spawning instead seems to be primarily
Support
level).
control,
dynamics
because
(inferred
is
they
given
result
Iran
to
a
three
in rather
accurate simulation of spawning when incorporated into a computer model
driven
by
records of a bottom temperature index (Fig. 6).
However,
since bottom temperature is correlated to other physical factors (Kruse
and
Huyer
MS),
it
is possible that a simulation model incorporating
hypotheses about roles of other
result
factors
in
in satisfactory spawning simulation.
reproduction
To
also
It is almost certain that
other factors operate in concert with temperature
control.
will
to
affect
spawning
conclusively distinguish the effects of temperature fran
these other factors, laboratory experiments should be conducted.
It does seem reasonable that the rate of maturation
is
inversely
related to the temperatures experienced during the sunnier (May-August).
Gonadal develonerxt probably occurs during this time and
it
is
known
51
that
somatic
growth
is most
temperatures (Kreuz et al.
been
MS).
of
season
cool
Negative correlations have previously
found between annual variations in marginal interoperculxn growth
in young English soles and mean summer
In
this
during
rapid
other
species
on
estimates.
temperature has various effects on rates of gonadal
deveJ. oçnent (de Vlaming 1972),
effects
temperature
bottom
different
having
sometimes
stages of gametogenesis
(Phsan 1966; Weibe 1968).
compl etel y opposite
in
Since anomalies of bottom
any one species
temperature
and
the coastal upwelling index are inversely related (Kruse and Huyer M3),
colder summer temperatures and stronger upwelling may lead
primary and
greater
to
secondary production and increased availablity of English
sole prey items, including polychaetes, amphipods, molluscs, ophiuroids
and crustacea (Kravitz et a.l. 1977).
Relationships between feeding and
maturation rates have been identified
Kinne
1960;
and
others),
so
it
in
other
possible
is
fishes
(Gross
1949;
that food production
operates with temperature to determine the maturation rates of English
soles.
In
any case this species has presumably evolved physiological
mechanisms providing for optimal somatic
environmental
conditions existing
and
during
gonadal
growth
at
the
the summer (including cold
temperatures), when productivity is high.
We also proposed that few or no P. vetulus spawn
colder
than
Schroeder 1953)
(Ware 1977;
Lett 1978),
detenninant
of
spawning
temperatures
In some species such as Atlantic cod, Gadus
about 7.8 C.
morhua (Bigel ow and
at
a
and
Scomber
scombrus
threshold may be the primary
temperature
activity.
mackerel,
Ahlstrom
(1965)
felt
that
52
temperatures
between
13-18 C were responsible for the time and length
of the California sardine (Sardinops sagax) spawning
temperatures
Perhaps
season.
out of this range may operate by inhibiting production or
release of honnones responsible for initiating breeding behavior.
Finally we hypothesized that
temperature
greater
in monthly mean
increases
bottom
than 0.95 C delay or interrupt spawning activity.
The magnitude of the increase and the length of the delay are suggested
be
to
linearly
related.
clear, although it
is
The tuechanins behind this response are not
widely known
that
deviations of
short-term
temperature can produce physiological and behavioral changes in fishes.
It is possible that
relatively
small
English
increases
soles
physiologically
are
course
temperature variations of 14 C or less over the
year
(Huyer 1977;
Kruse
and
Huyer MS).
of
the
(1 972)
found
that
rising
temperature
retard it.
a
general
sudden
increases
Alternatively, sharp increases in Oregon continental
shelf temperatures may be associated with shifts
and
Leggett and
accelerated upstream
American shad (Alosa sapidissima) migration but that
could
entire
Rapid temperature increases
might also alter English sole spawning migration patterns.
itney
by
since adults experience
temperature,
in
stressed
deterioration of
the
set
in
weather
patterns
of conditions required by
English soles for spawning.
Before our hypothesized relationships between
shelf
temperatures
and reproductive cycling are accepted, they should be validated through
controlled laboratory experiments.
are
However, in their present form they
certainly useful for simulating English sole spawning activity, as
53
we have
shown.
This
is
significant,
because
year-class
strength
appears to be largely determined by environmental conditions prevailing
prior to spawning (Hayman and Tyler 1980) and
during
history
spawning
stages
(Ketchen
1956).
Thus
our
valuable to investigations of the mechanisms by which
factors
operate.
Once
the
early life
model
these
will be
physical
these mechanisms are understood, forecasts of
iglish sole availability can be made in
advance
incorporated into future managnent strategies.
of
recruitment and
5Lt
Conci usions
The first study presented here adds to our
of the
understanding
oceanographic processes occurring along the Oregon coast.
Although the
Bakun's
relationships between deep temperatures, coastal sea level and
upwelling
index
are
not
totally
the
unexpected,
significance of these relationships is docunented for the
statistical
first
time.
It is felt that the relationship between sea level and temperature will
be
useful
estimates,
to
researchers
in
and
I
have
found
glish sole spawning model,
(Microstomus pacificus).
fishes and
their
those
glish soles and Dcver
should
help managers
fishery productivity
in
activity and those related
to
to
distinguish
related
to fishing
the environment.
The second part of this thesis contributes
glish
soles
n understanding of such interactions between
environment
changes
Albert
temperature index to be correlated to
this
annual variations in the groith of young
between
temperature
shelf
particularly during periods of sporadic empirical records.
In addition to being useful to the
Tyler
of regular
need
to
our
knowledge of
sole life history and perhaps of similar species living in the
Oregon upwelling system.
It
is
hoped
that
future
studies
can be
conducted to validate the proposed hypotheses of temperature-controlled
spawning and to assess the coupled effects of other
food
avail ability and photoperiod.
factors,
such as
Predictions made from the spawning
model developed here have already been valuable to an investigation of
the
Tyler.
glish
sole
recruitment
process conducted by myself and Albert
55
References
Ahistran, E.H.
A review of the effects of the environment of
1965.
the Pacific sardine. Inter.
Pub.
Fish. Spec.
Northwest Ati.
Comm.
6:53_7LL
Ahsan, S.N.
Effects of tnperature and light
1966.
changes in the
sper'matogenetic
activity of
J.
)44:161-171.
Couesius plctnbus.
Bagenal, T.B.
S.D.
on
Can.
Zool.
the cyclical
the lake chub,
A short review of fish fecundity, p. 89-111.
1967.
(ed.),
Gerking
basis of freshwater fish
biologival.
The
In
production. John Wiley and Sons. New York.
Bakun, A.
Iinerica,
Bakun, A.
Coastal
1973.
1946-1971.
Tech.
NOAA.
Rep.
i].y and weekly upweij.ing
1975.
North
uplling indices,
America,
west coast of North
NMFS.
103 p.
indices, west coast of
NOAA Tech.
1967-1973.
SSRF-671.
Rep.
NMFS.
SSRF-693.
11)4 p.
effective
niinber
of
independent observations in an autocorrelated time series.
J.
Bayley, G. V., and J.M. Haxnrnersley.
Roy.
Beck, S.D.
Stat. Soc. Ser. B
The
19)46.
8:18)4-197.
Acadnic Press,
Insect photoperiodimn.
1968.
New
York.
288 p.
Bigelow, H.B., and W. C.
Maine.
Bourke, R.H.
Schroeder.
Fish. Bull., U.S.
1969.
514
an estuary.
p.
Fishes of the Gulf of
53(7k):577 p.
Monitoring coastal
effects within
Corvallis.
1953.
M.S.
upwelling
by measuring
its
thesis, Oregon State Univ.,
56
Brett, J.II.
Temperature - fishes,
1970.
rine ecology, vol. I.
(ed.),
Budd, P.L.
Wiley - Interscience.
and
Develornent of the eggs
l94O.
California
fishes.
Calif.
New York.
early larvae of six
Fish
Dep.
0. Kinne
In
p. 515-616.
Game,
Bull.
Fish.
56:53 p.
Clarke, F.N.
fish
The life history of Leuresthes tenuis, an atherine
1925.
with
Fish.
tide
Bull.
controlled spawning habits.
10:1-22.
O.ishing, D.H.
1969.
The regularity of the
fishes.
Cons.
Inter.
O.ishing, D.H.
1975.
Marine ecology and
Press.
1vis, R.E.
Fish Game,
Calif.
New York.
spawning
Mer., J.
Explor.
Cons.
fisheries.
of some
season
33:81-92.
Cambridge
278 p.
and
Predictability of sea surface temperature
1976.
level pressure anomalies over the North Pacific Ocean.
Oceanogr.
de Vlaming, V.L.
1972.
Eifield, D.B., and J.5.
of monthly mean
Allen.
1970.
A
during
1949.
Biol.
Fish.
J.
1980.
:131-10.
On the structure
dynamics
and
sea level anomalies along the Pacific coast of
Phys.
J.
statistical
Oregon
Oceanogr.
study
coastal
Oregon State Univ., Corvallis.
Gross, F.
Phys.
J.
reproductive
&ivirorinental control of teleost
North and South America.
temperature
sea
6: 2k9-266.
cycles: a brief review.
Fisher, C.W.
Univ.
of'
winds
10: 557-578.
and
upwelling.
J.
ter
thesis,
M.S.
67 p.
Further observations on fish growth in
sea loch (Loch Graiglin).
sea
Mar.
Biol.
a
fertilized
Assoc., U.K.
28:1-8.
57
Harry, G. Y., Jr.
Time of
1959.
fecundity of the
vetulus,
respectively).
Hayman, R. A., and A. V.
of
Dover
Fish Comm.
Tyler.
sole
and
Microstomus
and
jordani,
Oreg.
and
maturity,
at
petrale, and Dover soles (Parophrys
English,
Eopsetta
length
spawning,
Briefs 7:5-13.
Res.
Environment and cohort
1980.
English
pacificus,
Trans.
sole.
strength
Fish.
Am.
Soc.
109: 5k-68.
Hela, I., and I.
Laevastu.
News (Books) Ltd.
Hempel, G.
London.
Washington Sea Grant Publ.
Hewitt, G.R.
Seattle.
Oregon State University, Corvallis.
Goldberg.
Hunter, J. R., and S. R.
U.S.
Huyer, A.
stage.
70 p.
of the inshore trawl fishery off Oregon.
fecundity
egg
Seasonal changes in English sole distribution:
1980.
analysis
Fishing
137 p.
Early life history of marine fish - the
1979.
Univ.
oceanography.
Fisheries
1961.
1980.
M.S.
an
Thesis.
59 p.
batch
Spawning incidence and
in northern anchovy, Fngraulis mordax
.
Fish.
Bull.,
77:641-652.
Seasonal
1977.
density
over
Oeeanogr.
variation
the
continental
shelf
off
salinity,
Oregon.
and
Limnol.
22: k42-)453.
Huyer, A., and R.L.
nith.
1978.
Physical characteristics of Pacific
Northwestern coastal i.aters, p.
marine plant biomass of the
State Univ.
temperature,
in
Press.
37-55.
Pacific
Corvallis.
In R.
Northwest
Krauss (ed.), The
coast.
Oregon
58
Huyer,
E.J.C.
A.,
transition
in
J. Geophys.
Res.
Johannes, R.E.
currents
Mar.
Fish.
Ketchen, K.S.
sole
continental
shelf.
Biol.
Fishes 3:65.-84.
glish sole tagging off California.
Pacif.
7:15-33.
Factors inf1i.ncing the
survival
of
the laon
(Parophrys vetulus) in Hecate Strait, British Colunbia.
Fish.
Kinne, 0.
&ivir.
Comm., Bill.
1956.
Oregon
the
over
spring
The
1979.
8:6995-7011.
Results of
1969.
nith.
R.L.
Reproductive strategies of coastal marine fishes
1978.
in the tropics.
Jow, T.
and
Sobey,
Res.
Board Can.
1960.
euryplastic
salinities.
Kravitz, M.J., W.G.
13:647-694.
Growth,
food
fish
exposed
Physiol.
J.
in
a
tnperatures and
different
to
33:288-317.
Zool.
Pearcy, and
conversion
food
and
intake,
Gum.
M.P.
1977.
Food
of
five
species of cooccurring flatfishes on Oregon's continental shelf.
Fish.
Kreuz,
Bull., U.S.
K.L.,
A.V.
Variation
74:98)4-990.
Tyler,
Thans.
Kruse, G. H., and A.
temperatures,
Pmer.
Huyer.
coastal
off Newport, Oregon.
Krygier, E.E., and W.G.
waters.
R.L.
Demory.
MS.
in growth of ]ver sole and English sole as related to
upwelling.
glish
and
Kruse,
G.H.
MS.
Fish.
&c.
The relationships between near-bottom
sea level and the coastal upwelling index
Lirnnol.
Pearcy.
MS.
Oceanogr.
(in press).
Distribution
sole (Parorys vetulus)
(in prep.)
(in press).
in
and
abundance of
Yaquina Bay and offshore
59
Laroche, J.L., and S.L.
English
larval
Richardson.
sole,
Winter-spring abundance of
1979.
River and Cape Blanco, Oregon during
occurrences
Sd.
between the Columbia
vetulus,
Paropirys
of three other pleuronectids.
notes
with
1972-1975
Estuar.
on
Mar.
Coast.
8:455-476.
Laroche, J.L., S.L.
growth
of
Richardson,
and
Rosenberg.
A.
and
Age
MS.
a pleuronectid, Parophrys vetulus, during the pelagic
larval period in Oregon coastal waters.
Bull., U.S.
Fish.
(in
press).
Laroche, W.A., and R.L.
Holton.
1979.
Occurrence
English
0-age
of
sole,
Parophrys vetul us, along the Oregon coast: an open nursery
area?
Northwest Sd.
Leggett, W.C., and R.R.
53:94-96.
Whitney.
migrations of American shad.
Lett, P.F.
Water
1972.
temperature
the
70:659-670.
Bull., U.S.
Fish.
and
A comparative study of the recruitment mechanins of
1978.
cod and mackerel, their interaction, and its implication for dual
stock management.
Lett, P.F., A.C.
biomass
Ph.D.
thesis, ]lhousie Univ.
Kohler, and D.N.
and
Fitzgerald.
125p.
Role
1975.
of
stock
in recruitment of southern Gulf of St.
temperature
Lawrence Atlantic cod, Gadus rnorhua.
J.
Fish.
Res.
Board Can.
32:1613-1627.
May, R.C.
Lunar
1979.
sexfilis, in Hawaii.
rers, K.W.W.
areas
1980.
in
spawning
Fish.
of
the
Bull., U.S.
threadfin,
Polydactylus
76:900-9O4.
An investigation of the utilization of four study
Yaquiria
Bay,
Oregon,
by hatchery and wild juvenile
60
salmonids.
Morse, W.M.
M.S.
1980.
thesis. Oregon State Univ. Corvallis. 234 p.
Spawning and fecundity of Atlantic mackerel,
Fish.
Scomber japonicus.
M.ndy, B.C.
Bull., U.S.
78:103-108.
Yearly variation in the abundance and distribution of
MS.
fish 1 arvae in the coastal zone off Yaquina Head, Oregon, from
June
1969
to August
M.S.
1972.
thesis, Oregon State University,
Corvallis. (in prep.)
Qasim, S.Z.
Time and duration of the spawning season in
1956.
marine teleosts in relation to their distribution.
Explor.
Mer, J.
Reid, J.L., and A.W.
Cons.
J.
Ocean.
Richardson, S.L.
Oregon:
August
1976.
The effect of the geostrophic
sea elevations in the Northern North Pacific
Geophys.
1977.
Res.
81:3100-3110.
Larval fisries in ocean waters off Yaquina Bay,
abundance, distribution and seasonality, January
1972.
Sea
Inter.
Cons.
21:1)414_155.
Mantyla.
flow upon coastal
some
Grant Ccll.
Frog.
Fubl.
1971 to
ORESU-T-77-0O3.
73 p.
Richardson, S.L., and W.G.
Pearcy.
1977.
Coastal
and
oceanic
fish
larvae in an area of upwelling off Yaquina Bay, Oregon. Fish.
Bull., U.S.
Bicker, W.E.
Res.
Roden, G.I.
the
1973.
75:125_1L5.
Linear regression in fishery research. J.
Board Can.
1960.
west
65: 2809-2826.
30:kO9_314.
On the nonseasonal variations in sea level
coast
Fish.
of
North
Pmerica.
J.
Geophys.
along
Res.
61
Rosenberg, A.A.
Thesis.
Sadleir, H.
estuarine
in
vetulus,
of
Growth
1980.
English
juvenile
and
nursery areas.
coastal
open
Oregon State University, Corvallis.
New York.
M. S.
51 p.
Press,
Academic
The reproduction of vertebrates.
1973.
Parophrys
sole,
227 p.
Sciranaiimano, F., Jr.
suggestion
A
1979.
presentation
tne
for
correlations and their significance levels.
Oceanogr.
Phys.
J.
of
9:1273-1276.
Sutcliffe,
Jr.,
W.H.
Correlations
of
Ware, D.M.
ine.
Fish.
Board Can.
G.E.
production.
1968.
reproductive
1974.
t.vis.
J.
mackerel,
Fish.
The
Laboratory studies
1967.
and
growth of
biological
Jobn Wiley and Sons.
basis of
on
p. 175_21L1.
fish,
freshwater
Res.
the
In
fish
New York.
The effects of temperature and day length
on
the
physiology of the viviparous seaperch, Cymatogaster
aggregata Gibbons.
Zar, J.H.
of Atlantic
in relation to the plankton.
bioenergetics,
S.D. Gerking (ed.),
Weibe, J.P.
size
34: 2308-2315.
Warren, C.E., and
feeding,
1977.
3:19-30.
Board Can.
Res.
Spawning time and egg
scombrus,
Scctnber
Muir.
B.S.
of fish catch and envirormental factors in the Gulf
J.
1977.
and
Drinkwater,
K.
Can.
J.
Zoo]..
Biostatistical analysis.
CLiffs, New Jersey.
)46:1207-1219.
Prentice
Hall.
Englewood
620 p.
Zweifel, J.R., and R. Lasker.
1976.
Prehatch and posthatch growth
fishes-a general model.
Fish.
&tll., U. S.
74:609-621.
of
appendix
62
Appendix.
Spawning activity by season.
19147-1 948
The gonadaJ. condition of English sole females
were
landed
We had access
sampled by the ODFW during 19147_1951 (Harry 1959).
to the raw data collected
during
this
time
period
Astoria
in
(Table
No
2).
samples were obtained in 19147 and collections were made only in January
and March of 19148.
Approximately 140% of the
spawning
total
activity
Spawning was terminated by
occurred between mid-January and mid-March.
the middle of March.
19148-1 9Lt9
Adult females landed in Astoria were sampled frequently during the
19148_1949
spawning
season
(Table
middle of November and early March.
2).
Spawning occurred between the
was
period
The heaviest spawning
fran early December to late January.
19149-1950
Spawning began prior to the first sampling
date
in
mid-November
19149 and terminated in about the middle of April 1950 (Table 2).
of peak spawning activity cannot be
accurately determined,
samples were taken in December or January.
Ltes
since
no
About 70% occurred sometime
between mid-November and early February.
1950-1951
Gonadal
condition
duration of the
females appear
during
1950-1951
was not
spawning season (Table 2).
to
have
spawned
prior
to
sampled
for
the
Approximately 16% of the
the
first
sampling
in
63
mid-Novanber
and only
1950
31 January 1951.
About 60%
77%
had spawned by the last sampling date,
of the spawning activity occurred between
the middle of Novnber and late January.
1969-1 970
Using revised growth estimates (Laroche et al. MS), collections of
2-3 mm SL English
sole larvae
(Mundy MS)
spawning season began in early Noviber
1969.
show that the
1969-1970
Spawning appears to have
stopped in mid-Decanber, resiined in mid-January and finished compi etel y
by the end of March.
Based on larval abundances, peaks in spawning
intensity are indicated for mid-Novnber
In 1970
beam
1969
and February
1970.
juvenile surveys were also conducted in Yaquina Bay using
trawls (Krygier and Pearcy MS).
Individuals approximately 20 mm
in length were collected in mid-January to early February.
about
Allowing
2-2.5 months to reach this length, this confinns spawning during
the month of November.
Juveniles of this age were also collected (and
in greater abundance) from late March to late May, which indicates that
additional heavy spawing occurred from mid-January through late March.
However,
we
place more confidence in spawning inferred from larval
data, which clearly shows spawning to be heavier in February than in
March.
1970-1 971
nall (2-3 mm SL) larvae first appeared in late
(t4mdy MS),
December
which means that spawning began in mid-December.
data (Mndy M5; Richardson 1977; Richardson and Pear'cy
1977)
also
1970
Larval
show
6)4
that
spawning
early March.
intensity increased many fold frcm late January through
Spawning activity seems to have stopped
late
in
April,
since 5-8 imn SL individuals were collected no later than May.
A survey using beam trawl s in Yàquina
obtained
y (Krygier and Pearcy
This
a few juveniles less than 20 mm SL in mid-January 1971.
to mid-Novnber,
means that some spawning must have occurred in early
contrary to
the
MS)
Significant
lack of such evidence Iran larval data.
nunbers of small juveniles were obtained Iran late February through the
end
of June.
Spawning times inferred fran these later collections are
Since
essentially identical with those inferred fran larval data.
greatest
the
nunbers of these young soles were observed between mid-March
and mid-May, the peak spawning times inferred Iran
juvenile
both
and
larval data also closely agree.
197 1-1 972
1977)
indicate that spawning began in mid-Septnber 1971 and continued
through late February 1972.
been
Pearcy
and
Larval studies 04.indy MS; Richardson 1977; Richardson
in
tcember.
Peak spawning
Al though
no
samples
activity appears
were
taken
to
have
in January or
February 1972, heavy spawning likely did not occur during these months,
since the nunbers of 1-2 month old larvae collected in March were low.
Beam trawis (Krygier and Pearcy
fish
MS)
less than 20 mm SL in late November.
early March when saznpl ing was terminated.
first
encountered
juvenile
They were collected through
This suggests spawning
from
the middle of September 1971 through at least the end of January 1972.
The greatest numbers
of
juveniles
were
taken
Iran
early
to
late
65
February,
indicating
peak
intensity in
spawning
Thus
December.
from
juvenile data are highly supportive of conclusions drawn
larval
samples.
1972-1973
Information
available
regarding
only one
fran
for
spawning
data
Based
1973.
during
17-29
early
March
and
on revised growth rates (Laroche et al.
presence of larvae up to 20 mm SL in March samples
activity from
season
1972-1973
is
source (Laroche and Richardson 1979).
Larvae were surveyed using bongo tows
April
the
MS), the
spawning
No statement can be made
January through March.
about possible spawning prior to January.
indicated
1 7-26
Fran the
presence
of very
large nunbers of 7-1)4 mm SL larvae in March samples and expected ages
of these individuals (Laroche et al.
early February
for
MS), peak spawning
to early March.
is
indicated
Spawning appears to have tapered
off by late April, although it is possible that some spawning
occurred
after this last sample date.
1973-1 97)4
large
Laroche and Richardson (1 979) reported that
of running
ripe
20 October 1973.
March
197)4
and
concentrations
spent adults were found off the Oregon coast on
Larval surveys were conducted only in the middle of
(Laroche and Richardson 1979).
At this time newly-hatched
soles were sparse and in the 2-8 mm SL size range.
This indicates that
minor spawning occurred from mid-February to mid-March arid little or no
spawning fran at least mid-January to mid-February.
t
size
groups
of
0-age
juveniles
along
the
The collection
of
coast in 197)4 also
suggested fall and winter-spring periods of spawning (Laroche and
Hol ton
1979).
Based
this information, it can be inferred that
on
spawning was intense in early October, absent fran mid-January to
mid-February and minor fran mid-February to mid-March.
19714_i 975
Larvae were
Richardson
1979),
only on
surveyed
V4-16
March
1975
(Laroche and
and large numbers were obtained on these dates.
The
presence of newly-hatched and 60 day old larvae indicated that spawning
began at 1 east by mid-February
mid-March.
The predctninance of
mid-February to early March.
and continued at least through
1975
day olds suggest a peak
8-25
from
No statement can be made about possible
spawning prior to mid-January or after mid-March.
1976-1977
A juvenile sampling program (Krygier and Pearcy MS) was conducted
in Yaquina
Bay
from
24
April to
2
July
1977.
Individuals less than
20 mm SL were collected on all dates suggesting that spawning began at
least by late February and continued through early May. The greatest
numbers of larvae less than
20 mm SL
period were taken from 22 May to
14
collected during any sampling
June.
This
uld suggest a peak in
spawning during late March or early April. aie to the shortness of the
sampling period, however, one cannot conclude that this would represent
the peak spawning time over the entire season.
A beach seining program (Myers
initiated in late June 1977,
of 19-22 mm fish. None were
1980)
collected only very slight numbers
collected after late June indicating that spawning did not extend past
67
early
No conclusion can be drawn about possible spawning prior to
y.
mid-January.
1977-1978
Hewitt
(1980)
adult female gonads fran October
sampled
1977
on the percentage of ripe stage III (the
most advanced stage of ripeness) or spent females, his data show that
through
spawning
rch
Based
1978.
began at least by early October, continued moderately through
January, and peaked strongly in February.
were spent by 1 March
of the females
Almost 95%
1978.
Beam trawl sampi es (Ho senb erg 1
Krygier and
Pearc y
MS)
of
small juveniles in Yaquina Bay indicate spawning fran mid-October
through early April. Although trends are difficult to discern from
these samples, spawning can be inferred to be intense in the middle to
late October and heaviest in January.
Beach seine samples in Yaquiria Bay (Myers
individuals less than 20 nn SL cm mid-December
through late April
1978.
1980)
1977
first
encountered
and collected them
A very small ntznber were also obtained
late July to early August.
from
This infonnation implies that spawning
occurred fran early October through the end of February, with a minor
amount
in late May or so.
A peak in spawning activity appears to have
occurred fran mid-December through early March.
Each of the
1977-1 978
studies indicate similar spawning times.
Clearly spawning began at least by early October and continued in
varying intensity through early May.
and
The precise dates for the start
finish of the major peak for this season are uncertain.
However,
68
each study indicates heavy spawning in January and/or February.
1978-1979
Beach seine samples (Myers 1980)
spawning
na.11
began at least by mid-August 1978.
SL were caught as early as 28 October.
occurred
of
between mid-September
and
show that
juveniles
Some fish less than 20 mm
Peak spawning appears
Young
mid-October.
have
to
soles were
caught through the last sampl e date (10 December 1 978).
Beam
indicate
trawl
spawning
samples
(Rosenberg
1980;
Krygier
and
Pearcy
MS)
to have occurred fran early September and continued
through mid-March. Peak spawning is indicated for mid-September through
mid-November.
Both of these juvenile studies yield compatible
spawning
times.
Using
both,
it
seems
that
inferences
spawning
about
began
in
mid-August, peaked from mid-September through mid-November and ended in
mid-March.