Content uploaded by Philippe M Cury
Author content
All content in this area was uploaded by Philippe M Cury on Nov 05, 2014
Content may be subject to copyright.
3
'Y
.U
Reprinted from
Canadian
Journal
of
-+
Fisheries
and
Aquatic
Sciences
Réimpression du
Journal
canadien
des
sciences
halieutiques et
aquatiques
Optimal environmental window and pelagic
fish
recruitment success
in
upwelling areas
-
P.
CURY
AND
CYROY
-
Volume46
0
Number4
0
1989
Pages 670-680
Cansaä
P
--
Fonds
Documentaire
ORSTOM
-
Fisheries
Pêches
m.(;crl-
and
Oceans
et
Océans
Printed in
Canada
by The
Runge
Press
Limited
L
Optimal Environmental Window and Pelagic Fish Recruitment
Success
in
Upwelling Areas‘
Philippe
Cury and Claude Roy
Centre de Recherches Océanographiq~es de Dakar-Thiaroye, lnstitut Sénegalais de Recherches Agricoles,
B.P.
224
1,
Dakar, Sénégal
Cury,
P.,
and C. Roy. 1989. Optimal environmental window and pelagic fish recruitment success in upwelling
Food availability and physical constraints such as turbulence are now considered as important factors that affect
larval survival and pelagic fish recruitment. In Ekman-type upwelling, vertical advection, new inputs of nutrients
and turbulence are linked to wind speed. According to the literature, food availability for larvae is related to
processes (turbulence generated by wind mixing). This limitation does not exist for non Ekman-type upwelling
where upwelling intensity is not correlated with wind speed. We hypothesize that relations between annual
recruitments and upwelling intensity are dome shaped in Ekman-type upwellings and linear for non Ekman-type
upwellings. A statistical method is used to analyse the form of the relationships between recruitments and upwell-
ing indices or wind mixing. The recruitment of the Peruvian anchoveta (Engraulis ringens), of the Pacific sardine
(Sardinops sagax caerulea) and
of
the West African sardines and sardinellas are thereby examined. Results show
that for Ekman-type upwelling the annual recruitment increases with upwelling intensity until wind speed reaches
a value of roughly 5-6 mes-’ and decreases for higher values. For a non Ekman-type upwelling the relationship
-
between recruitment and upwelling intensity is linear. These results confirm the existence of an optimal envi-
ronmental window for recruitment.
La disponibilité en nourriture et des contraintes physiques comme la turbulence sont des facteurs importants pour
la survie des larves et le recrutement des espèces pélagiques. Dans un upwelling d’Ekman, les mouvements
verticaux, les apports en sels minéraux et la turbulence sont liés
à
la vitesse du vent. D’après la littérature, la
disponibilité en nourriture pour les larves est associée
à
des processus biologiques (production primaire) qui
peuvent être perturbés par des processus physiques (turbulence). Ce facteur limitant disparait quand l’intensité
de I’upwelling est indépendante du vent local. Nous proposons une relation en forme de dôme entre le recru-
tement et l’intensité de I‘upwelling pour un upwelling d’Ekman et linéaire pour les autces types d’upwellings.
Une méthode statistique est utilisée pour analyser la forme des relations entre recrutement, les indices d’upwelling
ou
la turbulence. Le recrutement de l’anchois du Pérou (Engraulis ringens) de la sardine du Pacifique (Sardinops
sagax caerulea) et des sardines et sardinelles ouest-africaines est étudié. Les résultats montrent que, dans les
upwellings d‘Ekman, le recrutement annuel s‘accroît avec l’intensité des upwellings jusqu’à ce que le vent atteigne
--
une vitesse proche de 5-6
mes-’
et décroît ensuite pour des vitesses plus élevées. Quand l‘intensité des upwellings
est indépendante des vents locaux et qui la turbulence est faible, la relation entre recrutement et upwelling est
linéaire. Ces résultats confirment l’existence d’une fenêtre environnementale optimale pour le recrutement.
areas. 46: 670-680.
-
L
-
-
biological dynamics (primary production) up to a point where the biological processes are disturbed by physical
Received March
22,
7
988
Accepted November
2
7,
7
988
(39649)
arge variability in pelagic fish recruitment is frequent and
may have an important effect
on
fisheries (Smith 1985).
L
Larvae survival is especially variable and the role of dif-
ferent sources of mortality at the prerecruit stage has been inten-
sively studied. The main causes of larval mortality appear to
be starvation and predation (Blaxter and Hunter 1982) and sev-
eral ef;;vironmental factors have a determinant effect
on
recruit-
ment (Shepherd et al. 1984). Currently
two
theories have
emerged to explain recruitment success
in
relation to environ-
ment. With the match-mismatch hypothesis, Cushing (1975)
emphasizes that the annualpioduction
om
larvae is matched
or mismatched to the production of their food.
In
other words,
a stock releases its larvae into the annual production cycle at
the best time to secure good survival
on
average. The impor-
i
‘This work
is
dedicated
to
Dr.
Reuben Lasker, a generous
man
and
a
pioneer
in recruitment studies.
670
n
Reçu /e
22
mars 1988
Accepté le
2
7
novembre 1988
tance of food availability for larvae is the core of this energetic
approach. The second hypothesis is based
on
Hjort’s (1914,
1926) suggestion; early first feeding for larvae is the most vul-
nerable stage in the life history of fish. Some authors (Lasker
1975, 1981a, 1985; Peterman and Bradford 1987) provide evi-
dence that turbulence in the euphotic layer increases larval mor-
tality during “critical periods” (May 1974).
A
stable environ-
ment is usually needed to allow aggregations of food organisms
to be formed and maintained. This stability hypothesis (Lasker
1981b) takes into account dynamic physical processes even
though food availability is again the crucial factor for larval
survival.
The time and space scales used for these studies
are
mostly
microscales (Lasker 1978; Methot 1983; Peterman and Brad-
ford 1987). It is interesting to
try
to reconcile the ideas which
have been developed within a fine scale using a broader scale.
In
this paper we analyze the relationships that may exist between
.
Can.
J.
Fish.
Aguar.
Sci.,
Vol.
46, 1989
"
n
annual recruitment indices and upwelling intensity or wind mix-
ing for some pelagic fish stocks in upwelling areas. The com-
parison between areas where upwelling intensity and wind mix-
ing
are closely related and areas where they are independent
will illustrate the potential effect of turbulence
on
recruitment.
Analysis
of
the Relationships between Recruitment
*
4
and Upwelling
Upwelling and Turbulence in Pelagic Fish Habitats.
The main coastal upwelling areas are located on the eastern
boundaries of the oceans where the equatorward trade winds
induce offshore Ekman transport. Cold, nutrient rich subsurface
waters
are
brought to the euphotic layers enhancing primary
production. Weak winds reduce primary productivity because
they disrupt the upwelling process and the renewal of nutrients
in the surface layers (Huntsman and Barber 1977). In the
classical Ekman scheme the magnitude of the offshore transport
in the upper layer is considered to be an indication of the amount
-of water upwelled along the coast into the surface layers (Bakun
1973). Higher wind induces higher offshore transport and
increases upwelling. Therefore upwelling intensity and nutrient
input into the euphotic layers could be estimated using offshore
Ekman transport calculated from the wind component parallel
to the coast. Ekman-type upwellings are found off Peru,
California, Morocco, and Senegal.
Off Ivory Coast and Ghana, trade winds are weak and the
strong cooling of the sea surface temperature during the boreal
summer cannot be interpreted as classical Ekman-type
upwelling (Bakun 1978; Picaut 1983). All attempts
to
correlate
the intensity and duration of this upwelling with local winds
have failed (Houghton 1976). Many mechanisms have been
proposed as explanations of this upwelling among which are
internal waves generated in the western part of the Atlantic
(O'Brien et al. 1978), upward thermocline slope at the coast
due to
,the
intensification of the eastward Guinea Current
(Ingham'T979), or local cooling downstream of a cape (Marcha1
and Picaut 1977). Since local winds are
not
the driving force
of this non-Ekman type upwelling, Cury and Roy (1987) used
interannual anomalies of coastal sea surface temperature to
estimate its intensity. The sign of anomalies was changed in
their presentation
so
that a positive value was associated with
a strong upwelling intensity.
The energy transferred through the water column by the wind
creates turbulence in the surface layers. The rate at which
turbulent kinetic energy of the wind is added to the surface layer
is roughly proportional to the cube of the wind speed (Niiler
and Kraus 1977; Elsbery and Ganvood 1978). Therefore a wind
mixing index that estimates turbulence in the upper layer is
usually given by wind speed cubed (Bakun and Panish 1980;
Husby and Nelson 1982). Wind mixing indices, when available,
were used (Peru, Morocco); otherwise wind speed (Senegal) or
upwelling indices (proportional to wind speed squared) were
used (California).
water movement and also generates turbulence in the surface
layers. Therefore, off Peru, Morocco, and Senegal annual wind
mixing index and upwelling intensity are positively correlated
(Fig.
1).
In
non
Ekman-type upwelling off Ivory Coast, Ghana
where wind is not the driving force, annual wind mixing and
upwelling intensity are independent variables (Fig.
2).
-
-In Ekman-type upwelling, high wind speed-enhances upward
-
-
-
Can.
J.
Fish.
Aquat.
Sci.,
Vol.
46,
1989
Ir
I
m'
E
w
100-
o
Y
i
2
!o
m
m
-
=O1
MOROCCO
0
oØ
t.
0.
0.
I I
UPWELLING
INOEX
tm3
a
s-I
-
IOO~I
;
"""1
SENEGAL
UPWELLINS
INDEX
tm3-
s-I
-
IOO~I
FIG.
1. Relationship between annual upwelling (m3*s-'.100 m-l
coastline) and turbulence (m3.sW3) indices
for
Eliman-type upwellings
in
Peru
from
1953 to 1985 (Mendo et al. 1987), in Morocco
from
1968
to
1981 (Belvèze 1984) and Senegal from 1964 to 1986
(C.
Roy,
unpubl. data).
Theoretical Approach
Acceptable food concentrations associated with stable ocean
conditions must be present in the larvae's environment for sur-
vival (Lasker 1981a). Strong turbulence generated by high wind
speed has a neggve effect
on
larval survival by desegregating
food and larvae patches (Saville 1965; Peterman and Bradford
1987) and on the recruitment (Lasker 1981a; Mais 1981). In an
Ekman-type -upwelling, vertical advection, new inputs of
nutrients and turbulence (wind mixing) are linked with wind
speed. Therefore in
an
=:Ekman-typë-upwelling,7&creasing
upwelling intensity from weak to moderate should have a pos-
itive effect
on
recruitment since increased primary production
would enhance food availability, wind mixing remaining low.
Strong upwelling should have a negative effect
on
recruitment
because wind mixing is high even if the primary production
increases. This limitation should not exist in the case where the
67
1
IVORY
COAST
O. O
O
c)
I
m
E
Y
A
I
I
-
UPWELLING INTENSITY
i
;
100
2
W
O
+
I
I
i
IT
J
o!
B
u
I
t
-
4a
-
20
O
20
40
k
UPWELLING
INDEX
C1/10pC3
FIG.
2.
Relationship between annual upwelling (1llO"C) and
turbu-
lence
(m3-se3)
indices for a
non
Ekman-type upwelling
in
Ivory
Coast,
from
1966
to
1981
(1971
and
1972 are missing, Cury and Roy (1987)
and ship
of
opportunity data).
strength of the upwelling is not correlated with wind inteñsity.
We hypothesize that the relationships between recruitment var-
iability and ansual upwelling indices are dome shaped in
Ekman-type upwellings (Fig.
3)
and linear for non Ekman-type
upwellings.
There are two limiting factors that explain the nonlinearity
of the curve for Ekman-type upwelling.
On
the left side of the
curve wind mixing is weak and the limiting factor is the pro-
duction of food due to the low intensity of the upwelling; on
the right side of the curve, the upwelling is strong and turbu-
lence is then the limiting factor. There is therefore an "optimal
environmental window" for moderate upwellings where the
effects of the limiting factors are minimized (Fig.
3).
-
,
Statistical Method
-
In analyzing the relationship between recruitment and envi-
ronmental factors most of the statistical methods are linear or
an a priori transformation is used (essentially
a
logarithmic
transformation) (Parrish and MacCal; 1978; Anthony and
Fogarty 1985; Stocker et al. 1985; Crecco et al. 1986). We
applied a statistical technique developed by Breiman and Fried-
man (1985) that empirically estimates optimal transformations
for multiple regressions. The response variable
Y
and the pre-
dictor variables
XI,
...
Xp
are replaced by functions
Tl(Y)
and
T2(Xl),
. . .
,
Tp+
l(Xp).
A
procedure estimates these functions
Ti
by minimizing
An iterative algorithm (ACE: Altemating Conditional Expec-
tation) permits the calculation of these transformation functions
which do not belong to a particular parameterized family and
which are even not monotone. It also differs from other empir-
ical statistical methods usually used in that the transformations
are unambiguously defined and estimated without use of ad hoc
heuristics, restrictive distributional assumptions, or restriction
of the transformation to a particular parametric family. If we
fix
the values of all but one variable and solve the problem of
what new transformation will minimize the normalized residual
sum of squares, then the solution is a conditional expectation
that can be estimated empirically using a smoothing algorithm.
The algorithm converges to an optimal solution and does not
produce a given equation, but rather an empirical smoothed
transformation of each of the data points for each of the vari-
ables. The transformation is not expressed in a particular unit
(unless a functional transformation can be discerned from the
plot) and its shape is found by plotting the transformed values
of a variable versus the original values. This procedure, pre-
viously used in fishery studies by Mendelssohn and Cury (1987)
and Mendelssohn and Mendo (1987) thus also provides a
method for estimating maximum correlation, and gives new
insights into the relationship between the response and predictor
mables (i.e. it allows the identification of discontiñuities in
the relationship).
612
-
Can.
J.
Fish.
Aquat.
Sci.,
Vol.
46,
1989
TABLE
1. Recruitment (number of 3-mo-old fish) and mean annual
parent stock (adult biomass of
3-mo
or
older fish) of the Peruvian
anchoveta (Pauly et al. 1987). Annual turbulence and upwelling indices
off Trujillo (Mendo et al. 1987).
Adult Turbulence Upwelling index
No.
Recruits biomass index (m3.s-'.lo0
m-'
Year
(
x
io9) (tons
x
IO6) (m3-sb3) coastline
1953
1954
1955
1956
1957
1985
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
260
252
156
106
141
172
487
573
554
394
65 1
364
72 1
635
484
352
609
568
73
105
42
22
115
90
188
209
66
392
33 1
5.965
9.378
8.195
5.017
2.828
3.661
7.163
11.618
16.428
14.688
12.945
14.183
12.902
15.175
18.739
12.859
13.265
14.428
13.708
3.106
3.122
3.446
2.761
4.420
1.155
3.878
1.421
2.504
9.003
179
229
25 1
259
29 1
221
212
199
195
188
185
185
180
22 1
218
273
225
250
206
239
243
148
252
23 8
190
163
178
20 1
138
1 92
236
256
26
1
277
239
233
220
219
217
212
217
214
250
248
278
245
268
233
254
25
8
177
25 1
250
216
202
207
181
136
Review
of
some Pelagic Fish Stocks
of
Upwelling Areas
Peruvian Anchoveta
-
The idea that strong winds and turbulence in the upper layer
can be detrimental to the survival of the Peruvian anchoveta
(Engraulis
ringens)
was presented by Walsh et al. (1980). Using
monthly anchoveta recruitment estimates, Mendelssohn and
Mendo (1987) reinforced the idea of an effect of turbulence on
short-term recruitment fluctuations. However, anchoveta
recruitment
also
depends on the adult biomass level (Csirke
1980) and the dome. shaped stock-recruitment relationship
suggests a strong effect of parental cannibalism
on
anchoveta
eggstanding stocks (Santander 1987). The recruitment depends
both on the parent stock and on environmental fluctuations.
We used the data updated in a recent synthesis on the Peruvian
anchoveta stock (Pauly and Tsukayama 1987). These authors
estimated a recruitment index (number of 3-mo-old
fish)
and
adult biomass using virtual population analysis from 1953 to
1982 (Table 1). An annual turbulence index was calculated for
one of the major anchoveta spawning areas located off Trujillo
usi@ monthly turbulence indices-(wind speed cubed) of Mendo
et
al.
(1987) (Table 1). Upwelling indices were not included in
the calculation as they are strongly correlated with the
turbulence indices (see Fig, 1) and do not improve fit.
Optimal empiricâl transformations (Tl,
T2,
T3) for the
multiple regression were calculated using the method previously
described.
Can.
J.
Fish.
Aquat.
Sci.,
Vol.
46,
1989
o
2*o
t
c
-
-1.5
I
I
I
I
l
recruitment
O
200
400
600
800
o.
2
a-
--0.2
c
a
O
LL
u)
z
U
t-
a
*
4-
1
o
'
'
'
'
t
'
'
s
'
I
'
-
0.6
300
-
100
200
turbulence
1.0
1
@
3
'
'
O
'
'
'
-
I
'0
'O
0.
'
-
1
.o
I
--
o'
'i
Id
1b.5
20
-
adult biaxnass
FIG.
4. Optimal empirical transformations for recruitment
(No.
of
recruits
X
10')
(Tl),
turbulence index (m3-s-') (T2) and adult bio-
mass (t
X
lo6) (T3) for
the
Peruvian anchoveta.
(1) T1 (Recruitment)
=
T2
(turbulence)
+
T3 (adult biomass).
613
TABLE 2.
Recruitment (number of year class at age
2)
and parent stock
(adult biomass of 2-yr-old fish) of
the
Pacific sardine
(MacCalll979).
Annual upwelling index off Monterey (Bakun
1973).
Upwelling Index
No.
Recruits Adult biomass
(m3.s-'-100
m-I)
Year
(X
104
(tons.
lo3)
coastline
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
,
1957
1958
1959
1960
1961
1962
1963
1964
1965
1.625
1.667
3.875
4.261
3.690
0.290
0.397
0.972
1.197
0.382
0.264
0.588
1.586
0.905
0.288
0.111
O.
074
0.056
0.01
1
566
405
740
793
780
277
136
202
239
170
108
90
177
122
88
54
27
21
11
3
92.8
84.8
78.1
68.8
70.1
77.4
68.7
90.8
81.3
160.5
139.8
113.4
97.1
162.3
91.6
88.8
94.8
76.4
152.8
The plot of the transformed values of the data against the
original values are shown in Fig.
4.
The estimated
transformation of the recruitment is almost linear (Fig.
4,
Tl).
Turbulence is transformed to a nearly dome shaped curve
(Fig.
4;
T2). It first increases to a value around 200 m3-sP3
(which corresponds to a wind speed between
5
and 6 mas-')
and then decreases strongly. The transformation of parent stock
is almost linear in shape with a small decrease for high values
of adult biomass (higher than 12.5
-
lo6
t) (Fig.
4,
T3).
The
resulting transformed model (1) explains 70% of the observed
variance in the recruitme& data. Recruitment is positively
.
correlated with stock bhass; however, the relationship
between recruitment and turbulence is non-linear. It suggests
that the upwelling is beneficial for the recruitment until the wind
speed reaches values of 5-6 mes-' and that for higher wind
speed, turbulence has a negative effect on recruitment.
These transformations suggest that both high turbulence and
low adult biomass may have played an important role in the
collapse of the Peruvian anchoveta.
In
1972 and 1973 the parent
stock was low (Table 1) and produced few recruits which had
difficulties surviving in an environment where turbulence was
higher than 200 m3C3 (Table 1; Fig.
4,
T2). Consequently,
overfishing was apparently not the only factõr preventing a
recovery
of
the stock.
Pacific Sardine
The Pacific sardine
(Sardinops
sagax
caerulea)
fishery like
the Peruvian anchoveta fishery is well documented. Analyses
of the sardine stock-recruitment relationship (Clark and Marr
1955; Radovitch 1962; Murphy 1967) showed density depend-
-
ence, often by assuming a Ricker functional relation-
ship. Following Cushing (197 l), who concluded that clupeoid
stocks tend not to have strong density dependent regulatory
mechanisms, MacCall (1979) showed that the stock-recruit-
ment relationship presented no curvature (density-dependent
,
regulation of the recruitment). We used new population esti-
I
-
614
1.5
L
o
e
e*
-Os1
t
"e
I.'
-
-
-
-1.7
'.
*-
I
o12345
-
*u
recruitment
e
a
e
8
8
8
-
e
e
O
e
o
e
e
1
,il
,
,
a
-
0.3
2-
a
u
-1.5
I-
O
100
upwelling
r
-
0.1
I'
200
e*
e*
e*
8
e
-
e
0
-2.5
O
JO0
800
adult biomass
__
FIG.
5.
Optimal empirical transformations for recruitment
(No.
of
recruits
X
10') (Tl),
upwelling index
(m3.s-'-100
m-' coastline)
(T2)
and adult biomass (t
X
IO3)
(T3)
for the Pacific sardine.
Can.
J.
Fish.
Aquut.
Scì.,
Vol:
46,
1989
TABLE
3.
Recruitment (CPUE of age
O
group) of the Moroccan sardine
(Belvèze
1984).
Seasonal turbulence and upwelling indices off Tantan
during the reproductive
period
(Belvèze
1984).
.
Upwelling index Turbulence
CPUE
(m3*s-'-100
m-') index
'e
'
Year (tons-d-
')
coastline
(m3.s-')
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
2.08
2.15
5.80
1.38
1.78
0.89
0.64
0.96
0.55
0.46
2-25
0.63
2.74
0.78
81.7
42.1
81.4
56.1
73.6
73.6
92.1
120.3
121.0
116.6
99.1
69.1
108.6
105.6
103.5
61.0
111.3
60.2
87.6
71.1
105.5
155.8
178.0
152.9
146.3
79.1
140.7
128.7
cn
z
O
I-
d
-
U
O
u,
cn
2
4
O
80
O
O
O
O
O
-1.6
OS2t
I
-*
I
I
I
-
O
2
4
6
-
recruitment
-0.3
I-
O
-0.9
I
I
I
J
EO
100
140
180
t
urb
u
le
ncIe
--
-
FIG.
6.
Optimal empirical transformations for recruitment (t-d-') (Tl)
and turbulence index
mates of the recruitments and Pacific sardine biomass calcu-
lated by MacCall(l979) (Table 2). The annual recruitment (year
class at age 2) was estimated using a cohort analysis of aged
(T2)
for the Moroccan sardine.
-
landings data from 1945 to 1964. The adult biomass was cal-
culated as all fish of age 2 or older. The only environmental
data available were monthly upwelling indices off Monterey
(36"N, 122"W) from Bakun (1973) that were averaged over the
whole year (Table 2).
Optimal transformations (T1
,
T2, T3) for the multiple regres-
sion were calculated using the current year upwelling index and
adult biomass, and year class at age 2 for recruitment 2 yr later.
For example recruitment index in 1948 is associated with annual
upwelling index and adult biomass calculated in 1946
(2) T1 (recruitment)
=
T2 (upwelling index)
+
T3 (adult biomass).
The transformation for the recruitment index presents a sharp
increase for low values (under 0.5
lo9
fish), it decreases for
values between 0.5 and 1
.O
-
lo9
and increases slowly for higher-
-
values (Fig. 5, Tl). The upwelling index transformation is
dome shaped with a breaking point around a value of 100-120
m3-s- per 100 m of coastline (this value corresponds roughly
to an alongshore wind speed of
5
m-s-') (Fig.
5,
T2). The
transformation of adult biomass is typically a Beverton and Holt
stock-recruitment relationship (Fig.
5,
T3); it shows a curva-
ture for sardine biomass of 200
-
lo3
metric tons. Model
(2)
explains 87% of the recruitment variance.
These relationships provide some complementary informa-
tion on the waning years of the Pacific sardine fishery. After
1954, the adult biomass was under 200
lo3
metric tons
(Table 2) and the recruitment-adult biomass relationship was
on a slope where a minor stock produced a minor recruitment.
After 1954, the upwelling intensity was higher than
in
the past
(Table 2), and may have had a negative effect
on
recruitment.
MacCall(1983) suggested that the rate of decline was sustained
because the fishery consistently exceeded sustainable yields.
The transformations of the Ekman index and the adult biomass
suggest that this was not the only factor producing the collapse.
It may have been due to a conjunction of several depressive
factors on recruitment; excessive upwelling, or- too depressed
biomass associated with a high exploitation rate.
West African Sardines and Sardinellas
Morocco
For Morocco, a recruitment index for sardine
(Sardina
pil-
chardus)
may be obtained by using CPUE (Catch Per Unit
of
Effort) of age
O
and age
1
of the following year (Belvèze and
Erzini 1983; Belvèze 1984) (Table 3). Monthly turbulence
indices at 28"N, 13"W from October to April (Belvèze 1984)
weFe used to cal_culate an annual wind mixing index during the
reproductive and larval growth periods.
Optimal transformations (Tl, T2) are estimated for the sim-
ple regression between recruitment and turbulence.
(3) T1 (recruitment)
=
T2 (turbulence).
The model explains only 21% of the observed variance in
recruitment. An estimation of stock size is not available and it
wõuld certainly contribute to an explanation of a much greater
part of the variance if included in the model. However, the
transformation of recruitment (Fig.
6,
T1) is very close to a log
transformation. The transformation of the turbulence appears
to increase very slowly
to
a value of 120 m3-s-3 (which cor-
responds
to
a wind speed close to 5 m-s-'); for higher values
of wind speed it decreases strongly (Fig. 6, T2).
-
-
-
Can.
J.
Fish.
Aquat.
Sci.,
Vol.
46,
1989
675
a
TABLE
4. CPUE, fishing effort for the Senegalese round sardinella fishery (Fréon, 1983). Seasonal
wind speed, turbulence, and upwelling indices off Yoff
(Fréon
1983; C Roy, unpubl. data).
CPUE Effort Wind speed Turbulence index Upwelling index
Year
(tons.10
h-')
io
h-103
(mos-')
(
m3-s
-
3,
(m3.s-"0
m-I)
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
13.54
11.62
12.70
9.86
7.56
10.68
16.32
11.38
9.78
7.22
8.06
8.61
9.15
6.68
7.01
6.09
2.30
2.00
0.607
0.673
0.880
1.325
1.469
1.262
1.455
2.590
3.509
4.062
4.820
5.157
4.913
5.996
6.203
7.773
10.206
10.841
4.90
4.66
4.46
4.37
4.29
5.01
6.00
5.30
5.99
5.50
5.93
5.61
5.01
4.53
5.21
5.03
4.77
4.51
152
178
145
144
105
120
133
176
255
21-1
241
218
26 1
202
172
115
186
135
142
106
135
265
194
97
113
90
.
84
85
83
87
109
137
117
144
122
134
111
103
78
115
-
103
93
82
104
144
119
-
TABLE
5.
CPUE, fishing effort for the Ivoiro-ghanaian sardinellas fish-
ery (Cury and Roy 1987, unpubl. data). Annual turbulence (C. Roy,
unpubl. data) and upwelling index off Abidjan (Cury and Roy 1987).
CPUE Effort index Upwelling index
Turbulence
Year (tonsad-')
(d.103)
(m3*s-3)
(1/
10°C)
1965 29.7
1966 7.33 3.420 91 -9.3
1967 8.51 3.746 I1 1 15.8
1968 7.35
-
4.073 70 ~10.6
___
1969 5.30
1970 3.81
1971 4.64
1972 6.47
1973 3.29
1974 4.87
1975 7.96
1976 10.64
1977 12.23
1978 9.26
1979 6.53
1980 5.75
1981 9.52
3.616
5.716
5.484
3.930
2.483
2.024
1.768
1.824
2.003
2.200
2.681
2.409
2.207
117
200
113
155
158
120
103
105
142
151
-
18.3
7.5
7.8
4.5
-37.8
-
6.8
0.8
36.1
1.2
9.4
-
13.2
-
10.7
-9.9
Senegal and
Ivory
Coast-Ghana
In Senegal and Ivory Coast-Ghana, recmitment indices are
not available for Sardinellas
(Sardinella aurita
and
Sardinella
maderensis).
Fréon (1983) and Cury and Roy (1987) studied
pelagic fish stocks in these upwelling areas and showed that
annualnUE is a function of fishing effort and of the upwelling
intensity during the fishing year and
1
yr before. The upwelling
intensity
1
yr before the fishing year appears to have an impor-
tant effect
on
recruitment (sardinellas are recruited to the fish-
ery after 1 yr). In Senegal, the upwelling is an Ekman-type
upwelling and wind speed is used as an upwelling index (Fréon
1983). The monthly averages of the wind speed over the upwell-
676
!
ing season (November
to
May) were used to estimate interan-
nua1 upwelling intensity and wind mixing. In Ivory Coast-
Ghana, the upwelling is not caused by the local wind and sea
surface temperature anomalies (deviation from a mean cycle
during the upwelling seasons) were used as an upwelling index
(Cury and Roy
1987).
CPUE, fishing effort, upwelling or tur-
bulence indices are presented in Table 4 for Senegal and
Table
5
for Ivory Coast-Ghana.
Optimal transformations (Tl, T2, T3, T4) for the multiple
regression between CPUE and fishing effort, environmental
indices during the fishing year and
1
yr before are calculated
(4) Tl(CPUE,)
=
T2(fishing efforti)
t
T3(ind.,)
+
T4(ind.,_
where
i
=
year index, ind.
=
upwelling index (Ivory Coast-
Ghana) or wind speed (Senegal).
For Senegal and Ivory Coast-Ghana the models explain
respectively 97 and 94% of the observed variance in CPUE.
Empirical transformations of CPUE for Senegal and Ivory
Coast-Ghana
are
curved and suggest that a log transformation
is suitable (Fig. 7, T1 and Fig.
8,
Tl). The transformations of
effort are nearly linear and have a negative slope. The relation-
ship between CPUE and effort is negative and
can
be approx-
imated with a linear model (Fig.
7,
T2 and Fig.
8,
T2). The
transformation of upwelling index or wind speed during the
fishing year increases for Ivory Coast and Senegal and shows
a platform for Senegal (Fig. 7, T3 and Fig.
8,
T3). The con-
was interpreted as the effect of availability of the fish; fish seem
less available during strong upwelling. The models used by
Fréon (1983) and by Cury and Roy (1987) to analyze CPUE in
Senegal and in Ivory Coast empirically integrate the ttpwelling
index
1
yr before fishing to evaluate recruitment. Therefore
analyzing the form of the transformation of this parameter that
maximizes the correlation in the model allows us to identify
tribution of this index to the explication of the CPUE variance
--
--
-
-
Can.
J.
Fish.
Aquat.
Sci.,
Vol.
46,
I989
L
@
2Do
r
O
L
1.6
O.
-1.6
1
o.
2
-J.6
I
I
I
O
6
12
18
cpue
O
6
12
I
O
.
9
%'
Z
t
O
I
-1.5
III01
I
O
2 '4
6
a
z
effort
a
O2 46 810
E
effort
M
O
u,
tn tn
2
-
a
-0.5
,
I
e
z
-0.1
L
-.
ü
I-
-
2.5
-40 -20
O
+20 +40
UPW. ind. year
i
-0.5
I-
I
I
I
4
5
6
wind speed year
i
-Os2/
-0.6
1
I
I
1
1
__
6
___
-
4-
5
windspeed year
¡-i
FIG.
7.
Optimal empirical transfomations for
CPUE
(t-d-') (Tl),
fishing effort
(10
h
X
lo3) (T2), wind speed
(ms-
')
during the fishing
year
(T3)
and wind speed during the previous year
("4)
for the Sen-
egalese round sardinella.
.
-
-1.5
io
I
I
I
-40--20
O
+20
+40
U
pw.
ind.
year
¡-i
FIG.
8.
Optimal empirical transformations for
CPUE
(t-d-') (Tl),
fishing effort (d-103) (T2), upwelling index (l/lO"C) during the fishing
year
(T3)
and upwelling index during the previous year
(T4)
for the
Ivoirian sardinellas.
'
-
Can.
J.
Fish.
Aquat.
Sci.,
Vol.
46,
1989-
611
300
i
*
O
O.
f
?
-
t.
PERU
.
.
O
50
100
200
NUMBER
OF
‘LASKER EVENTS’
SENEGAL
I
.
u)
.
m
..
.5
I
..
g
200
=
w
1
.O.
i
the relationship between the environmental parameter and
recruitment. For Senegal, the transformation of wind speed
1
yr before the fishing year is dome shaped with a breaking point
centered at
5
mes-’ (Fig. 7,
T4).
For the Ivory Coast-Ghana
the transformaton of the upwelling index is nearly linear with
just a platform for values around zero (Fig.
8,
Table 4). This
suggests that recruitment and upwelling are positively corre-
lated in Ivory Coast-Ghana.
The transformations of the indices that evaluate upwelling
intensity in Senegal and in Ivory Coast-Ghana are consistent
with
our
hypothesis that a linear relationship exists between
recruitment and upwelling in a
non
Ekman-type upwelling and
that a dome shaped relationship exists in an Ekman-type
upwelling.
Discussion:
“5-6
m-s-‘ Wind Speed” as an “Optimal
Environmental Window”
“Lasker Events” and Average Seasonal Wind Speeds
Peterman and Bradford (1987) and Mendelssohn and Mendo
(1987) used an index reflecting Lasker’s hypothesis (1978),
678
called the “Lasker event”. It measured the number of
4-d
periods during which the wind speed did not exceed
5
mes-’.
While arrived at from different time scale studies,
our
results
are also consistent with these previous studies which present
evidence of the importance of this criterion and its impact on
Husby and Nelson (1982) noted that
...“
the average intensity
of turbulent wind mixing over a spawning season is not likely
to be well correlated with interannual variability in recruitment.
Rather, the existence of sufficient time-space windows within
which. turbulence does not exceed critical values may be the
relevant factors.” The validity of this assumption based
on
Lasker’s (1978) hypothesis is confiied by the result of
-
Peterman and Bradford (1987); these authors show that it is the
succession of calm periods more than the mean wind speed that
is correlated with daily larval mortality rates.
Our
results are in
agreement with these observations; Fig.
9
shows that the
number of Lasker events during an upwelling season is
negatively correlated with the average wind speed cubed for
Peru and Senegal (data not available for the other areas).
Therefore, wind speed
on
aveIage over a spawning season could
be used as a rough index of the number of low turbulence events
during the spawning season.
Physical and Biological Significance of
5-6
m-s-’
What is the significance of the
5-6
mes-’ value considering
oceanographic features and enrichment processes? From a
physical point of view, the threshold wind speed of
5
m-s-’ is
a value at which wind stress begins to exert a measurable mix-
ing effect on the surface layer in near-shore waters (Kullenberg
1971,1972,1974,1976, and 1978). Also, when the wind speed
is greater than about 7 mes-’, wave breaking becomes obvious
(Pond and Pickard 1978) and generates strong turbulence. From
a biological point
of
view, wind speed
of
approximately
5
m.s-’
has been found to be a “threshold” value above which wind
mixing tencLs to desegregate phytoplankton patchiness (Ther-
riault and Platt 1981; Demers et al. 1987). These authors dem-
onstrate that if the winds are strong enough to surpass this
threshold for surface layer mixing, wind mixing doEnates all
other potential sources
of
variance of the phytoplankton patch-
iness; below this threshold the phytoplankton patchiness can be
ascribed to biological causes. The threshold effect
on
recruit-
ment success is consistent with this observed dynamic of
phytoplankton.
Dispersion might not be the only factor affected by strong
wind mixing. Huntsman and Barber (1977) showed that pri-
mary production and zooplankton biomass in the Northwest
African upwelling is also affected by strong wind mixing. They
show that strong winds produce a strong mixed layer and a light
limited phytoplankton population. Therefore larval survival in
the case of strong wind mixing could also be affected by the
reduction of primary production.
Increased offshore transport of eggs and larvae with increased
upwelling intensity is often cited as a cause of larval mortality
(Bakun
and Parrish 1980; Parrish et al. 1983). At this stage of
this analysis it is impossible to know the relative importance of
this detrimental factor. Since reproductively active fish avoid
areas with strong offshore Ekman transport (Parrish et al. 1983;
Husby and Nelson 1982; Roy et al. 1989), we think that tur-
bulence in a reproductive area is perhaps a dominant factor.
’
larvae survival.
I
,,
,i
,
-
Validity of the Theory
Review of information
on
some of the most important and
well studied pelagic fish stocks of upwelling areas indicates that
CanTJ.
Fish.
Aquat.
Sci.,
Vol. 46,
1989
c
the facts
are
consistent with the theory. When calculating the
transformations of the different variables, taken one by one,
like recruitment with parent stock or recruitment with upwelling
(plots not presented here),
it
appears that each variable explains
a significant percentage of recruitment variance and also that
the transformations are similar to those obtained when simul-
taneously analysing recruitment, parent stock, and upwelling.
Stock-recruitment and recruitment-upwelling relationships are
both important for recruitment success. For an Ekman-type
upwelling, the optimal transformations of turbulence
(or
upwelling index) in the Peruvian, Califomian, Moroccan, and
Senegalese ecosystems are very close to those predicted by the-
ory.
A
dome shaped relationship exists between recruitment and
upwelling intensity estimated from wind data. The non-linearity
always appears for values of wind speed around
5-6
m-s-’.
This value is common for all the transformations and suggests
that for different Ekman-type upwelling ecosystems there is a
common and optimum wind mixing level in the stable layers
of the upper ocean.
The validity of the theory is reinforced by the results obtained
in Ivory Coast-Ghana. In this ecosystem local trade winds are
weak and not correlated with upwelling intensity. Our theory
suggests that strong wind mixing becomes a limiting factor even
if upwelling intensity enhances primary production, but that if
this limiting factor is nor present, primary production and avail-
ability of food is the only limiting factor and recruitment should
increase with upwelling intensity. The transformations,
obtained with Ivory Coast-Ghana data, clearly illustrate that
recruitment and upwelling intensity are almost linear and pos-
itively correlated.
Conclusion
Upwelling intensity differs from one area to another. In a
given upwelling area, pelagic fish reproductive strategy tends
to reach the optimal environmental window (as defined by
our
theory) by locally optimizing physical constraints. For exam-
ple, in a weak upwelling area fish tend to reproduce in the most
productive time-space areas. In the case of a moderate upwell-
ing, fish reproductive strategies have to compromise between
high productivity and strong turbulence. In the case of a strong
upwelling, the turbulence is the only limiting factor for recruit-
ment. Thus, local optimal environmental parameters may differ
from one area to another and recruitment can be sometimes
positively
or
sometimes negatively correlated with upwelling
intensity. In upwelling areas the “match-mismatch”
or
“sta-
bility hypothesis” theories should both be valid considering our
general relationship between recruitment success and environ-
mental limiting factors.
-
Acknowledgments
ISRA
(Institut Sénégalais
de
Recherches
Agricoles)
and ORSTOM
(Institut
français
de
recherche scientifique
pour
le
développement
en
coopération) provided support
for
this
study.
We
wish
to thank
Dr.
R.
Mende_lssohnfocproviding
us
help
in
statistics
and
Dr.
A.
Fon-
tana (Director
of
CRODT/ISRA)
for
his
encouraging
remarks.
We
are
grateful to
A.
Bakun,
D.
Binet,
P.
Cayré,
D.
H.
Cushing,
A.
Fonte-
neau,
P.
Fréon,
R.
Lasker,
R.
H.
Parrish,
M.
Sinclair,
J.
P.
Trodaec,
the
two
referees
and
our
colleagues
of
the
CRODT/ISRA
for helpful
comments
on
this
paper.
Special
thanks to
E
Laloe
for
his continuously
helpful remarks
on
statistics.
Mrs.
Viveca
Fonteneau assisted with
translation.
References
ANTHONY, V. C.,
AND
M.
J.
FOGARTY. 1985. Environmental effects
on
recruit-
ment, growth, and vulnerability of Atlantic herring
(Clupea
harengus
hur-
engus)
in the Gulf of Maine region.
Can.
J.
Fish. Aquat. Sci. 42 (Suppl.
BAKUN, A. 1973. Coastal upwelling indices, west coast of North America,
1946-1971. U.S. Dept. Commer., NOAATech. Rep. NMFS SSRF-671,
103 p.
1978. Guinea current upwelling. Nature (Cond.) 271: 147-150.
BAKUN, A.,
AND
R. H. PARRISH. 1980. Environmental inputs to fishery pop-
ulation models for Eastern boundary current regions, p. 67-104.
In
G.
D.
Sharp [ed.] Workshop
on
the effects of environmental variation on the
survival
of
larval pelagic fishes. IOC Workshop Rep. 28.
BELvZZE, H. 1984. Biologie et dynamique des populations de sardine
(Sardina
pilchardus)
peuplant les côtes Atlantiques marocaines et propositions pour
un aménagement des Fcheries. Thèse de Doctorat d’état. Université de
Bretagne Occidentale, 532 p.
BELV~ZE. H.
AND
K.
ERZINI. 1983. The influence of hydroclimatic factors
on
the availability of the sardine
(Sardina
pilchardus,
Walbaum) in the
moroccan atlantic fishery.
In
G.
D. Sharp, and
J.
Csirke [ed.] Proceedings
of the expert consultation to examine changes in abundance and species
composition of neritic fish resources, FAO, FIRM/R 291 (2): 285-327.
BLAXTER,
J.
H.
S.,
AND
J.
R. HUNTER. 1982. The biology of the clupeoid fishes.
Adv. Mar. Biol. Academic Press, London 20: 1-223.
BREIMAN,
L.,
AND
J.
H. FRIEDMAN. 1985. Estimating optimal transformations
for multiple regression and correlation.
J.
Am. Stat. Assoc. 80 580-619.
CLARK, F. N.,
AND
J.
C. MARR. 1955. Population dynamics of Pacific sardine.
Calif. Coop. Oceanic. Fish. Invest., Prog. Rep., July
1,
1953-March 31,
1955: 11-48.
CRECCO, V., T. SAVOY,
AND
W. WHITWORTH. 1986. Effects of density-depend-
ent and climatic factors
on
American shad,
Alosa
sapidissima,
recruit-
ment: a predictive approach. Can.
J.
Fish. Aquat. Sci. 43: 457-463.
CSIRKE,
J.
1980. Recruitment of the Peruvian anchovy and its dependence
on
the adult population, p. 307-313.
In
A. Saville [ed.] The assessment and
management of pelagic fish stocks. Rapp. P.-V. Réun. Cons. Int. Explor.
Mer. 177: 307-313.
CURY, P.,
AND
C. ROY. 1987. Upwelling et pêche des espèces pélagiques
côtières de Côte d’Ivoire: une approche globale. Océanol. Acta, Vol. lO(3):
347-357.
CUSHING, D.
H.
1971. The dependence of recruitment
on
parent stock in dif-
ferent groups of fishes.
J.
Cons. Int. Explor.
Mzr.
33: 340-362.
1975. Marine ecology and fisheries. Cambridge University Press,
Cambridge. 278 p.
DEMERS,
S.,
J.
C. THERRIAULT, E. BOURGET,
AND
A. BAH. 1987. Resuspension
in
theTallow sublittoral zone of a macrotidal estuarine environment: wind
influence. Limnol. Oceanogr. 32(2): 327-339.
ELSBERY, R.
L.,
AND
R. W. GARWOOD, JR. 1978. Sea-surface temperature
anomaly generation in relation to atmospheric storms. Bull. Am. Meteor.
Soc. 59(7): 786-789.
FRBON,
P.
1983. Production models as applied to sub-stocks depending
on
upwelling fluctuations.
In
G.
D. Sharp, and
J.
Csirke [ed.] Proceedings
of the expert consu1tation;to examine changes in abundance and species
composition of neritic fish resources, FAO FIWR 291 (3): 1047-1064.
HJORT,
J.
1914. Fluctuations in the great fisheries of northern Europe. Rapp. P.-
V.
Réun. Cons. Perm. Int. Explor. Mer. 20 1-228.
1926. Fluctuations in the year classes of important food fishes.
J.
Cons.
Int.
Explor. Mer.
1:
5-38.
HOUGHTON, R. W. 1976. Circulation and hydrographic structure over the Ghana
continental shelf during the 1974 upwelling.
J.
Phys. Oceanogr. 6: 909-
924.
HUNTSMAN,
S.
A.,
AND
R. T. BARBER. 1977. Primary production off northwest
Africa: the relationship to wind and nutrient conditions. Deep-sea Res.
HUSBY, D. M.,
AND
C.
S.
NELSON. 1982. Turbulence and vertical stability in
the California current. CalCOFI Rep. Vol.
XXIII:
113-129.
INGHAM, M. C. 1970. Coastal upwelling
in
the northwestem of Gulf of Guinea.
Bull. Mar. Sci. 20 2-34.
KULLENBERG,
G.
1971. Vertical diffusion in shallow waters. Tellus. 23: 129-
135. 1972. Apparent horizontal diffusion
in
stratified vertical shear flow.
Tellus 24: 17-28.
1974. Effects of environmental factors, part
1,
p. 1-9.
In
A. L. H.
Gameson [ed.
J
Discharge ofsewage fromseaoutfalls. Int. Symp. Pap.
33.
Pergamon Press,
NY.
1):
158-173.
24: 25-33.
-
-
Can.-J.
Fish. Aquat. Sci.,
Vol.
46,
1989
679
-
1976.
On
vertical mixing and the energy transfer from the wind to
the water. Tellus 28: 159-165.
1978. Vertical processes and the vertical-horizontal coupling, p. 43-
71.
In
J.
H.
Steele [ed.] Spatial pattern in plankton communities. Plenum
Press, NY.
LASKER, R. 1975. Field criteria for survival of anchovy larvae: the relationship
between inshore chlorophyll maximum layers and successful frst feeding.
U.S. Fish. Bull. 73: 453462.
1978. The relation between oceanographic conditions and larval
anchovy food in the California Current: identification of factors contrib-
uting to recruitment failure. Rapp. P.-V.-Réun. Cons. Int. Explor. Mer.
1981a.
Factors contributing to variable recruitments of the Northern
anchovy
(Engraulis morda)
in the California current: contrasting years,
1975 through 1978. Rapp. P.-V. Réun. Int. Explor. Mer. 178: 375-388.
[ed.]
1981b.
Marine fish larvae. Morphology, ecology, and relation
to fisheries. University of-Washington Press, Seattle, WA. 131 p.
1985. What limits clupeoid production?
Can.
J.
Fish. Aquat. Sci. 42
MACCALL, A. 1979. Population estimates for the waning years of the Pacific
sardine fishery. Calif. Coop. Oceanic. Fish. Invest. Rep. 20 72-82.
1983. Variability of pelagic fish stocks off California.
In
G. D. Sharp,
and
J.
Csirke [ed.] Proceedings of the expert consultation to examine
changes in abundance and species composition of neritic fish resources.
MAIS,
K.
F.
1981. Age-composition changes in the anchovy,
Engrdis mor-
da,
central population. Calif. Coop. Oceanic. Fish. Invest. Rep. 22 82-
87.
MARCHAL, E.,
AND
J.
PICAUT.
1277. Répartition et abondance évaluées par
échointégration des poissons du plateau continental ivoiro-ghanéen en
relation avec les upwellings locaux.
J.
Res. Océanogr. 2: 39-57.
MAY, R. C. 1974. Larval mortality in marine fishes and the critical period
concept, p. 3-9.
In
J.
H.
S.
Blaxter [ed.] The early life history of fish.
Springer-Verlag, New York, Heidelberg-Berlin.
MENDELSSOHN, R.,
AND
P. CURY. 1987. Fluctuations of afortnightly abundance
index of the Ivoirian coastal pelagic species and associated environmental
conditions. Can.
J.
Fish. Aquat. Sci.
44:
408421.
MENDELSSOHN, R.,
AND
J.
MENDO. 1987. Exploratory analysis of anchqveta
recruitment off Peru and related environmental series, p. 294-306.
In
D.
.
Pauly and
I.
Tsukayama [ed.] The Peruvian anchoveta and its upwelling
ecosystem: three decades of change. ICLARM studies and Reviews 15:
351 p. IMARPE, Callao, Peru; GTZ, Eschbom, Federal Republic of Ger-
many; and ICLARM, Manila, Philippines.
MENDO,
J.,
PIZARRO, L.,
AND
S.
CASTILLO. 4987. Monthly turbulence and
Ekman transport indexes, 1953 to 1985, based
on
local wind records from
Trujillo and Callao, Peru, p. 75-88.
In
D.
Pauly and
I.
Tsukayama [ed.]
d6he Peruvian anchoveta and its upwelling ecosystem: three decades of
change. ICLARM studies and Reviews 15,351 p. IMARPE, Callao, Peru;
GTZ. Eschbom, Federal Republic of Germany; and ICLARM, Manila,
173: 212-230.
(Suppl.
1):
31-38.
FAO FIRWR 291(2): 101-112.
--
Philippines.
METHOT. R.
D.
1983. Seasonal variation in survival of larval
Enaraulismorda
estimated from the age distribution of juveniles. U.S. Nitl. Mar. Fish.
Serv. Fish. Bull. 81.
MURPHY, G. 1967. Vital statistics of the Pacific sardine
(Sardinops
caerulea).
NIILER, P. P.,
AND
E. B. muss. 1977.One-dimensional models of the upper
ocean, p. 143-172.
In
E. B. Krauss [ed.] Modelling and prediction of the
upper layers of the ocean. Pergamon Press, New York.
Ecology 48: 731-736.
O’BRIEN,
J.
J.,
D. ADAMEC,
AND
D. W. MOORE. 1978. A simple model of
equatorial upwelling in the Gulf of Guinea. Geophys. Res. Lett. 5:
641-
644.
PARRISH, R.
H.,
A. BAKUN, D. M. HUSBY, ANDC.
S.
NELSON. 1983. Com-
parative climatology of selected environmental processes in relation to
Eastem boundary current pelagic fish reproduction.
In
G. D. Sharp and
J.
Csirke [ed.] Proceedings of the expert consultation to examine changes
in abundance and species composition of neritic fish resources. FAO
FIRM/
PARRISH, R.
H.,
AND
A.
D.
~CCALL. 1978. Climatic variation and exploi-
tation in the Pacific mackerel fishery. Calif. Dept. Fish Game Fish. Bull.
PAULY,
D.,
M. L. PALOMARES,
AND
F.
C.
GAYANILO. 1987. VPA estimates of
the monthly population length composition, recruitment, mortality, bio-
mass, and related statistics of Peruvian anchoveta, 1953 to 1981, p. 142-
166.
In
D. Pauly and
I.
Tsukayama [ed.] The Peruvian anchoveta and its
upwelling ecosystem:
three
decades of change. ICLARM studies and
Reviews
15. 351
p. IMARPE, Callao, Peru;
GTZ,
Eschbom, Federal
Republic of Germany; and ICLARM, Manila, Philippines.
PAULY,
D.,
AND
I.
TSUKAYAMA
[ed.]
1987. The Peruvian anchoveta and its
upwelling ecosystem: three decades of change. ICLARM studies and
Reviews
15.
351 p. IMARPE, Callao, Peru; GTZ, Eschborn, Federal
Republic of Germany; and ICLARM, Manila, Philippines.
PETERMAN,
M. R.,
AND
M.
J.
BRADFORD.
1987. Wind speed and mortality rate
of a marine fish, the northern anchovy
(Engraulis morda).
Science
PICAUT,
J.
1983. Propagation
of
the seasonal upwelling in the eastem equatorial
Atlantic.
J.
Phys. Oceanogr. 13, 1, 18-37.
POND,
S.,
AND
G.
L.
PIC-.
1978. Introductory dynamic oceanography.
Pergamon Press, 241 p.
RADOVICH,
J.
1962. Effects of sardine spawning iock size and environment
on
year class population. Calif. Fish. Game. 48: 123-140.
ROY, C., P. CURY,
AND
H.
BELVÈZE. 1989. Stratégie spatio-temporelle de la
reproduction des clupéidés des zones d’upwelling d’Afrique de l’ouest.
Aquat. Living Resour. (In press)
SANTANDER,
H.
1987. Relationship between anchoveta egg standing stock and
parent biomass off Peru,
4-143,
p. 179-207.
In
D. Pauly and
I.
Tsu-
kayama [ed.] The Peruvian anchoveta and its upwelling ecosystem: three
decades of change. ICLARM studies and Reviews 15, 351 p. IMARPE,
Callao, Peru; GTZ, Eschborn, Federal Republic of Germany; and
ICLARM, Manila, Philippines.
SAVILLE, A. 1965. Factors controlling dispersal of the pelagic stages of fish
and their influence
on
survival. Int.
Conun.
for N.W. Atl. Fish. Spec.
SHEPHERD,
J.
G.,
J.
.G. POPE,
AND
R.
D.
COUSENS. 1984. Variations
in
fish
stocks aaypotheses concerning their links with climate. Rapp. P.-V.
Réun. Cürrs. Int. Explor. Mer. 185: 255-267.
SMITH, P. E. 1985. Year-class strength and survival of O-group clupeoids. Can.
J.
Fish. Aquat. Sci. 42 69-82.
STOCKER, M.; V. HAIST,
AND
D.
FOURNIER.
1985. Environmental variation and-
recruitment of Pacific herring
(Clupea harengus pallasi)
in the Strait of
Georgia. Can.
J.
Fish. Aquat. Sci. 42 (Suppl. 1): 174-180.
THERRIAULT,
J.
C.,
AND
T.
PLAT^.
1981. Environmental control of phytoplank-
ton patchiness.
J.
Fish. Res. Board Can. 38(6): 368-641.
WALSH,
J. J.,
T. E. WHITLEDGE, W.
E.
Esus,
R. L. SMITH,
S.
A. HUNTSMAN,
H.
SANTANDER,
AND
B.
R. DEMENDIOLA. 1980. The spawning habitat of
the Peruvian anchovy,
Engraulis ringens.
Deep-sea Res. 27: 1-28.
R 291(3): 731-777.
167:
1-110.
-
(Wash., DC) 235: 354-356.
Publ.
6
335-348.
680
Can.
J.
Fish.
Aquat.
Sci.,
Vol.
46,
I989
