Document SCRS/87/76 INTERACTIONS BETWEEN TUNA...
Document SCRS/87/76
INTERACTIONS BETWEEN TUNA FISHERIES
A CRITICAL REVIEW BASED ON SOME ATLANTIC EXAMPLES
by
ALAIN FONTENEAU
ORSTOM scientist
CRODT
BP 2241
DAKAR
SENEGAL
S U M M A R Y
This
paper
describes
some
interactions
observed in
the
Atlantic Ocean between tropical tuna fisheries.
Seversl types of
interactions
are
distinguished. It
is first
shown
th a t
for
tropical
tunas,
there
is not presently a
clear
relationship
between for fisheries adults and for fisheries juvenile.
This is
probably because, despite of the high level of fishing effort r a
recruitment
overfishing never has been observed in the ?\\tlantic,
probably
because of the high fecundity
and wide spawning
areas
of these species.
The interactions between small tuna fisher.ies
and
large
tunas fisheries are also analyzed for
yellowfin
<lnd
biyeye.
For
both
species this interaction is estimateà
ter be
significant at the ocean level,
because of the important catches
of juveniles.
The interaction between gears catching large tunas
such a s
longline
and
purse
Seine,
in
analyzed.
T h e
Larticularities
of the vertical stock structure are hypothesized
for
both
species
from
the catches and
cpue b y
gear.
As a
consequence of this vertical stock structure,
the purse Seine is
the
gear which
cari take full profit of the
yellowfin
biomass,
and
longline
the only gear capable to exploit
intensively
t h e
deep
bigeye
stock.
Some other short term
interactions
within
small
time
and
area units are also described and
discussed ;
several
exemples of those interactions
are
presented. A
mo r e
detailed
discussion
concerning
the
short
term
interactions
between fisheries operating in different areas is presented using
a "boxes model".
Such type of boxes mode1 which are
developped
needs
a huge amount of data
(detailed catch,
effort
and sise
statistics by small time and area strata, intensive tagging), but
seems to
be a key tool to evaluate the short
and
medium
term
potential
interactions between fisheries exploiting a
migratçry
resource
considered
as a
stock.
Al1
the
present
wcrk
4 6
demonstrates,
at least,
that the interaction study needs,
much
more than any other stock assesment work , very good and complete
statistics associated with intensive tagging.
RESUME
Cet
article
décrit quelques interactions
observées
dans
1’Atlantique entre des pêcheries de thonidés tropicaux, Plusieurs
types
d'interactions
sont
distingués.
L'absence
de
re lat ion
claire
entre la
taille du stock reproducteur et le
n iv e au
du
recrutement
est
tout
d'abord
montrée.
Cette
stabilité
du
recrutement et l'absence de chute du recrutement qui est observée
malgré
les
niveaux
de pêche élevés correspond à
l'absence
de
"recrutement overfishing".
Cette situation est interprêtee comme
@tant
due a la forte fécondité et à l'existence de vastes
z 0 ri e s
de
ponte
qui sont des caractères généraux de
ces
especes.
L a
compétition
entre les pêcheries qui exploitent les juvér,iles
et
c e 11 e s
qui exploitent les adultes d'albacores et de
patudo
2 s t.
examinée. Celle-ci semble significative pour les deux especes, au
moins
un niveau de l'océan,
par suite des prises importantes de
juvéniles qui sont observées. La compétition entre les engins qui
capturent
les
individus de
grande
taille,
par
exemple
I e s
senneurs et les palangriers, est analysée. Pour les deux espèces,
cette compétition est interprêtée à partir d'une hypothèse sur la
structure
verticale
des stocks reposant sur les
tendances
des
prises et
des efforts par engin.
Il apparait que la senne
t! s t
l'engin
qui
est le
mieux à
même
d'exploiter
pleinement la
biomasse
d'albacore,
alors
que la palangre est le
seul
engin
capable
d'exploiter intensément les gros patudos de
profondeur.
u n
autre type de compétitions,
celles observées
à c 0 u r t.
terme
entre
des pêcheries pêchant dans de petites
zones,
sont
a u s s 1
examinées et
plusieurs
exemples
sont
présentés.
Enfin
1.1 n e
discussion
concernant
les
interactions à
court
terme
entre
pêcheries
exploitant
des
zones de
pêche
différentes
est
presentée.
Un modèle informatique "à boites" a été établi à ce,t
effet .
Ce type de modèle qui reste en cours de
développement
utilise une grande masse de données détaillées : prises et prises
par
unité
d'effort
par tailles, selon des
zones et
périodes
réduites,
marquages
intensifs.
Cette approche semble toutefois
être du plus grand intérêt futur pour évaluer les interactions a
court et
moyen
terme
entre
des
pêcheries
exploitant
'3 I: C?
ressources
migratrice.
Le
présent
travail
montre
bien
la
nécessité,
pour conduire à bien les études sur les
interackions
entre
pêcheries,
d'excellentes statistiques de pêche,
fines et
complètes, associées à des marquages intensifs.
48
1 - INTRODUCTION
The
concept of
interactions between
fisheries is
quite
complex
and cari be studied from different points of view.
Among
others, three major types of interactions cari be distinguished :
(a). .
TO what extent
large size tuna fisheries cari compete
with small size tuna fisheries,
for long living species? (In the
medium and long term).
a
(b)..
TO what extent fisheries catching small size tunas cari
compete with large size tuna fisheries ?
(In the medium term, in
general during the exploited phase of the fis'h ).
(c).. T O
what extent two or several fis‘heries catching
the
same
sizes
of fishes are competing for this common ressource ?
(In the short term ).
In
this type of interaction ,two cases need to be
analyzed
separately :
- Those
exploiting small tunas (in general a
plurispecific
mixture of small yellowfin, bigeye and skipjack ) ,
- Those exploiting large size tunas .
This
distinction seems necessary because small
size
tunas
live in plurispecific surface schools, while :Large size tuna tend
to be
in both deep and surface
layers,
with a
differential
geographical
distribution between species .
These
studies of interactions between fisheries need
several
types of data which cari be listed as follows :
(1) .A
correct
species identification of catches in
the
statistics.
When
small
bigeye are called yellowfin
and
both
species often registered as skipjack in the statistics, it
Will
be
impossible
to estimate any interaction for any of the
three
species. Those problems of species identification are often found
for small size tunas and difficult ta solve.
(2).A good size sampling for a11
f'isheries ,
necessary to
study the trends of age specific cpue and to conduct VPA .
(3).Catch
and effort statistics established for each
gear,
with
detailed time and area stratification,
which are necessary
to
calculate
series
of cpue which
cari be
representative of
- . abundance of each year class .
:
(4).Good
knowledge upon stock structure,
p r e f e r a b l y b a s e d -
u-f?on
direct
evidence,
such a s
those
obtained
bY
intensive
tagging.
(5).Good
biological
studies
especially on
growth,
reproduction
processes
and
fecundity.
Growth
knowledge is
necessary to
keep
track
of cohorts of
fi.shes
exploited by
fisheries ;
spawning
biology is
necessary
to
estimate
potentials
of the
spawning stocks.
4 9
(G).Preferably high level of fishing efforts,
covering
various areas and a11 sizes of fishes, SO that the complete .range
O f
geographical
distribution
and
a11
ages of
f i s h e s a r e
exploited.
(7) a,
Stock
assesment
studies
especially
stock
size
estimates by age, based on Virtual Population Analysis (VPA).
(8).
Yield
per
recruit analysis based upon
a9e
specific
fishing mortalities of each gear.
‘
(9).
Mathematical models must be developped on computers in
order to analyse the interactions between fisheries. Those models
are
still quite new in the field of tuna stock managment ,
but
are
to be
developped in
order to
better
understand
the
interaction problems ;
Al1 these results must be combined and compared.
When
t:ley
are
obtained
during
a long period of time,
they
cari
provide
reasonably
good answers on the different types of
interactions
between fisheries .
The
Atlantic
ocean offers good exemples for
a11
possible
types of
interactions(type a to c);
furthermore
the s e
interactions
cari be welï analyzed because of intensive statistics
and
research
conducted in the Atlantic during long
periods of
time
under the coordination of the International Commission
fOL
the Conservation of Atlantic Tunas ( ICCAT).
Al1 researches from
type
1 to
~type 9 have been done in the Atlantic at
different
degrees,
depending on species .
Subsequently
the
goal of thi-s paper Will be to review
a II d
discuss
the present knowledge and researches on the interactions
in the Atlantic.
2 .DATA USED.
The
data used for this study are extracted from the ICCAT data
base
available at
the end of the year
1986.
Those
d,ata
are
basically
the
catch,
effort
and size
data ,
for
yellowfin
skipjack and bigeye tuna,
available during different periods of
time
for each fieet,
with a variable degree of
precision.
For
more details upon the nature of those data, the ICCAT Data Record
num.26 cari be usefully consülted.
In the &se of some fleet where
the
data
was net available,
the standard-strata
substitutions
done
by SCRS scientists were performed when necessary
(sec
the
reports of ICCAT ad hoc working groups). --
,,-,.-_
“~--“-.-
.c<
_-
..”
-.
.--
-‘.,----
“-
,_-_-
..-_-^,-,
__-.%,‘ml-
----
^.I-
<S.<I_
--
5 0
3 .
HOW
MUCH LARGE TUNA FISHERIES
CAN AFFECT SMALL
SIZE
TUNA FISHERIES ?
In
most
exploited fish stocks- at least small pelagic
and
demersal species - there is some relationship between adult stock
of spawners and number of recruits generated by
this
spawning
stock .
In
most
fish
stocks,
this
relationship
between
recruitment and :stock is highly variable,
but a reduction of the
spawning stock will tend either to decrease
(usually the case),or
to increase (Ricker type model) the recruitment level (sec figure
la).
However
this
typical stock recruitment mode1
dont
really
amly
for tropical tunas :
in any case ,
at least for tropical
species , it seems that recruitment shows only some moderate year
to
year
variability,
but
no trend ,
even at
high
level of
exploitation of the adult stock (see figure lb) .
However
this stability of recruitment over a large range
of
spawning stocks cannot be indefinitely extrapolated :
if the
number
O f
spawners
tends to
become
equal to
zero
the
recruitment is
necessarily affected .
This critical
leiel o f
spawners beyond which the recruitment could be affected has
been
hypothetized to a 10 % level of the virgin
stock, but
has never
been observed in tuna
fisheries , at least for tropical species.
In fact a decreasing recruitment trend due
to overfishing -
i-e. a
recruitment
overfishing- has
never been
observed
for
tropical tunas , even for highly exploited
ones .
Consequently
the
present paradigm for
tropical
tunas is
that, at
moderate or relatively high level of fishing efforts,
there
Will
not be
any effect or
interaction of
the
adult
fisheries on the juvenile ones .
This
specificity
of tunas (at least the tropical
species)
seems to
be related to the extremely high
fecundity of
_
these
individuals ;
each
female of
mature
yellowfin , skipjack or
bigeye Will spawn several millions of eggs,
several
times
each
year,
covering extensive areas and periods of time (Cayre et al.
in press FAO) (sE?e figure 2 to 4) , where their larvae cari always
(statistically) find good conditions for their survival . In such
conditions of
spawning
the recruitment Will be
limited in
general,
much
more
by the carrying capacity of
the
nursery
areas , than by the size of the spawning stock a
4 .
HOW FISHERIES CATCHING SMALL TUNAS
WILL AFFECT-LARGE ~TUNAS
FISHERIES ?
4.1.The problem:
This
type of
interaction concerns
mainly
yellowfin
and
bigeye
tunas,
which
are
exploited at small sizes by
surface
fisheries
and later at large sizes by longline and
purse
Seine
fisheries.
5 1
How
the fisheries catching juveniles Will affect
fisheries
catching adults is consequently an important question to salve .
The
Atlantic
ocean offers good observations
and
relevant
analysis in that field.
This
type
of study is based primarily on the
analysis of
fishery data.
4.2. Atlantic bigeye tuna .
It should be first noted that the fisheries of juveniles are
located in
areas
which are different from the areas
0 f
adult
concentrations (sec figures 5 to 6).
This observation is
easily
explained by
t.he
physiology of the species :
the young bigeye
tunas are physic'logically a tropical species,
living in
surface
warm
waters;
it becomes a temperate deep species when
becoming
adult ,
except during seasonal spawning when they corne back to
tropical waters (see figure 3).
The
juveniles of
bigeye
tuna
are
quite
diff icult
t 0
recognize from
the
juvenile yellowfin . In
the Atlantic as in
other
oceans
the small bigeye have been
widely
misidentified
with
yellowfin for many years.
This species composition problem
O f
juveniles
has
been
treated
seriously
only
since
1 $7 7 0
(FONTENEAU 1975). A systematic control of the species composition
0 f
the small tuna catches has been developped in the At1anti.c: on
a
routine
basis
only since 1979.
Historical
data
have
btren
tentatively
corrected b y
ICCAT in
1984,
but
are
sti11
questionable. A
review of
the pending
problems
for
species
identification
is given in the report of the ICCAT working group
held in 1987 on this subject.
The
catch
of small (less than 90 cm) bigeye taken in
the
Atlantic
since 1955,
beginning of the industrial
fisheries,are
shown in .figure 7, together with the catch of large bigeye.
The
cpue o f
small bigeye
for
surface
fisheries
shows
fluctuations,
but
without
trend (sec figure 8 ).
The cpue of
large bigeye,
given by the longline Honma index (sec figure
9) )
shows a moderate and regular decline since
the
beginning of the
fisheries.
The
stock nssessment analysis on bigeye concludes that this
decline cari be explained by the increased catch of the longliners
and by increased catches of juvenile;
the great numbers of small
bigeye
taken
by surface gears correspond to a
relatively
low
fishing
mortality
(sec
figure
10)
and
has
had
only
a
moderate
effect o n
adult stock size.
This
is shown
bY
the
results of the VPA,
which give estimates of the recruitment
in
t b-e
adult- fishery.
The decrease of this adult recruitment is
-_
12 le ar ,
at a rate-of approximately 20 % between the virgin
s teck
and recent. years (sec figure 11 ),
but
is relatively moderate.
:IX
cari be noticed also that present
yield
per recruit analysis
indicates that the juvenile fishery affects the longline :fishery,
but does
not significantly reduce the overall yields, at
least
a t
present
levels o f
fishing mortalities
(sec
figure
12)
(PEKEIRA 1986).
5 2
4.3. Atlantic yellowfin.
The
fishing
zones of small yellowfin are
uite similar to
the
adult
ones(see
figure 13
and 14 1, at
i! east
for
purse
seiners.
However
it is noticed that small yellowfin in
general
stay in
more coastal waters that the adults .
The
adults 'of
yellowfin
taken by
longliners
also
show a
wider
area o f
distributiontsee
figure 14b 1,
scattered between 10' north
and
south, from America to Africa.
The
numbers
of small (-90 cm) and large(+90 cm)
yellowfin
taken by atlantic fisheries are given in figure 15.
This
figure
shows
well
the
dramatic
increase of small
yellowfin in
the
catches du:ring recent years (by baitboats and purse seiners), and
the increase of catches of large yellowfin (by purse Seine).
The cpue for small yellowfin,
(baitboats and purse seiners),
shows some year to year variability, but no definite trend, sug-
gesting a relatively stable recruitment (figure 16). The cpüe for
large yellowfin is given for two gears :: longliners catching only
large fishes, and purse seiners (sec figure 17). In
order to ob-
tain surface index for adults, the
usual
overall
purse
Seine
index has been corrected by the amount of small yellowfin in the
catches. Those two adults cpue show, for recent
years a similar
serious decline. This decline of the adult cpue corresponds to ,a
decline of the adult abundance. The VPA analysis shows that this
decline of adult stock is a consequence of two factors:
l.the
development
of juvenile fisheries which
reduces
the
recruitment to the adult stock .
2.the increased effort and increased catches on adults
which
reduces directly the adult biomass.
Present analysis suggests that the first effect, interaction
between
j uvenile
and
adult
fisheries,
would be
relatively
significant.
This is
shown by
t he
decline o f
the
adult
recruitment
(figure 18)estimated by VPA at approximately 50 %,
decline due to the increased catch of juveniles.
However subsequent yield per recruit analysis shows that the
juvenile
fishery
does
not seriously decreases
the
yield
wr
recruit
of
the
overall fishery because the juvenile
catch
in
weight is
only sligthly less than the
subsequent
theoretical
loss of the adult fisheries (FONTENEAU 1986).
4.4.
Conclusion :
juveniles
against
adults fisheries.
For
bath
species,
the
catch of juvenile seems to
reduce
significantly
the catches of the adult fisheries of about 20 %
for
Bigeye and 45 % for yellowfish this interaction seems to be
moderate
compared to the millions of small flrshes -taken ,because
of the relatively low fishing rates estimated on
juveniles. .-
It should also be noticed that the present yield per recruit
analysis
indicates for both species that the--overall
yield
per
recruit
h.as
suffered
only a minor reduction
due to
juvenile
fisheries.
5 . HOW LONGLINE AND PURSE SEINE SHARE THE BIG TUNA BIOMASS ?
5 3
5.1.Catch versus effort for bigeye and yellowfin for the two
gears.
HOW the two gears compete in the short term to exploit l.acge
yellowfin and bigeye tuna is an important problem.
The
relationship between fishing efforts of purse Seine ‘2nd
longline,
and
their
respective catches of large
fishes
shows
completely different patterns for the two species :
. for yellowfin, the increase of longline effort has produced
a decreasing yield curve, while a simultaneous increase of effort
by
purse seiners has increased the catches at a level much
more
higher than the catches by longliners (sec figure 20).
* for
bigeye
tuna there is a
reverse
observation :
the
increase of longline effort is still producing an increase of the
catches,
while
the
increase
of surface effort has produced a
stable or even decreasing level of catches (sec figure 21).
These
observations are in contradiction with the yield
per
recruit
analysis ,
which indicates that increase of effort by
purse
seiners should have increase the catches of
large
bigeye
and
that
the longliners should have increased
their
yellowfin
catches.
This
contradiction
between the expected and
observed
yields fior the two gears and the two species is summarized on the
figure ;!4.
5.2.The vertical stock structure heterogeneity hypothesis.
The
m o s t
probable
explanation
for
this
observation
i. s
related to a vertical stock structure heterogeneity:
- Yellowfin is a surface tropical tuna, even at large sizes.
The purse seiners cari exploit the ressource with great efficiency
a n d
high fishing mortality .
The longliners cari exploit only Ü
rnargi.nal
and
fragile fraction of the
stock,
the
deeper
one.
Con s e que n t 1. y
the longliners cari never exert a real high
fishing
rnortality
on the total stock because of the excessive
reduction
0 f
the deep biomass.
Consequently the yield per recruit concept
does not
really apply to the longline yellowfin fishery
because
the longline adul.t recruitment does not show the same trend
that
the purse Seine adult recruitment .
--Bigeye is an opposite case : being basically a deep and
tem-
perate
tuna,
the tropical purse seiners Will always exploit tne
ressource
with a
very low
fishing
mortality,
even
at
h iqh
nominal fishing effort.
Any increase of fishing effort does
r-1 0 t
produce an increase of F. The
longliners,
-on the contrary,
c an
easi.ly concentrate their effort on deep-bigeye,
and an increased
longline
effort
produces
a corresponding in-crease in
fishing
mortality,
especially with the introduction of the more efficient
deep longlining.
This
general hypothesis has been summarized
graphically
2. n
figure 25.
Under
the
present
hypothesis
the sharing of
the
adult
5 4
biomass between surface and deep gears would be related much more
to a
vertical
stock structure heterogeineity,
than to a
real
interaction :
- for bigeye,
the decline of surface adult cpue is not
due
to
the
longline
catches or to
a real decrease of
the
total
biomass.
- for yellowfin,
the early decline of the longline cpue is
not a consequence of interaction with purse Seine fisheries,
but
would be related to a vertical stock structure heterogeneity .
5.3.0ther
more direct interactions between purse seiners and
longliners to catch large tunas.
There
is still probably some real and direct
interactions
-between the two gears on the adults ;
these interactions cari be
of two types :
.
The "urn"
interaction,
as described in chapter 6 ,where,
in summary, a tuna taken by one gear cannot be taken by the other
gear,
which gives an advantage to the more efficient gear.
.
The geographical interaction,
where two gears exploiting
the
same
area
are
more in
interaction
than i f
they
are
exploiting
distant
areas,
even in the case of a unique
stock.
This
is related with the probability of migratory
distances
0 f
tunas : the probability
of moderate migration (for instance less
than
500 miles),
is always greater than the probability of very
large
ones( for instance more than 2000 miles).
The
situations
are different depending on the species:
(a)Yellowfin
In- that field it is noticed that the areas where the highest
purse Seine catches of large yellowfin are observed are the areas
where
the
highest
longline
catches
rates
were
historically
observed (figure 26).
Presently it is noticed that the
longline
cpue
are lower (figure 27) that the average in this fishing zone
for the purse seiners (see figure 14a).
This is
probably a local consequence of high
purse
Seine
catches
taken
in this area.
This geographical
interaction is
similar
to the interaction described in the simple "urn
model".
However
the
migration of large tunas are much
less
known
and
probably
more
important
that
those of
small
tunas; as a
consequence
the
strong
hypothesis of a close system
does
not
awly
well
to the adult yellowfin or bigeye
fisheries in
an y
area.
-_
-
(b)Bigeye :
It is
clear for the adults of this species that the
purse
Seine
fishery
has
no sigificant
geographical
effect o n
the --
longline
fishery. This is due to the low purse Seine catches and
to the differences in fishing areas of the two gears (see
figure
6).
6. SHORT TERM COMPETITION IN THE SAME AREA : THE "URN" mode1 :
i) 5
6.1 *Concept:
Tunas are clearly migratory ; however duri;: limited perinds
and
in
relatively
small
areas
they
cari
considered as
sedentary,
and
considired as a locally and
temporaly
isolated
fraction of stock.
In
this
case,
the
term
"fishing area"
cari
describe a
relatively small geographical unit in which tuna cari have
ra:ncom
movements.
As an exemple ,
an area of S" by 5' square, i.e. of
approximately
3600 nautical square miles,
would be a reasonable
unit to study.
In such type of interaction,
only the short term
interactions,
such as from 1 to 3 months,
are concerned ; since
tunas are migratory species,
it is quite obvious that a group of
small tunas staying in a given area Will migrate in an other area
if the elapsed time is long.
In
theory, if
the
fraction of a stock is stable
i 11
the
area(e.g. n o
emigration and no
immigration),
the
interaction
problem
becomes quite simple.This stability of the sub stock cari
often be assumed during limited periods of time.
In such case the studies are simplified because:
-the growth of the fish cari be neglected.
-the
natural mortality becomes much less important that
i n
the long term interaction studies;
-fishing
mortality
rates cari be easily calculated if
the
total
catches
a r e k n o w n ,and if there is a
measure of
local
abundance(loca1 cpue for instance).
In
this simple case,
the recruitment in the area
w i 11
b e
ishared
bctween simultaneous fisheries.
In such a simple
c a s e ,
the
interaction
between
fishing
units
Will
depend o f
the
available
total biomass and on the relative effort and catch
3f
each gear.
This type of competition wiI1 be the "competition"' of
type
1 as described by Ricker 1975:
the catch of each day by a
boat reduces the catches of the other units later.
The abundance
of fish is lower and lower as the fishing is developping, and the
rate
of this decrease is proportionna1 to the number of
boats.
This
has
been
expressed in
figure 28b
which
is a
local
production type curve,
within a limited time and area stratum :
it shows the expected relationship between the local catch during
a
fishing seasor.
and the total fishing effort.
In this example
the
initial biomass and potential catch is 10000 tons wh;ch
c 3x-l
be taken only with a very high fishing effort. If a local fishery
working on this substock takes 5000 tons per fishing season
, t'îe
addition
of a
new
fleet
catching
5000
tons
Will
r e du 12 e
significantly the catch
rates and catches of t.he first fleet.
6. 2. Possible exemples of the "urn" mode1 :
6.2.1.Introduction:
-
It is
quite difficult to find in the real
world
of
tuna
fisheries
a true example of this urn model.
This difficulty
.i s
primarily
due to the migratoxy nature of the tunas.
However
(3 t
least
three
type of examples,
belonging more or less to
this
“ u r n " conce,pt, cari be described in the Atlantic ocean.
5 6
6.2.2. The concentrations :
Concentrations
have
been described in
the
Atlantic as
important
groups
O f
schools which are fished in a
small
area
during a short period of time.(FONTENEAU 1986).
Several examples of those concentrations have been analyzed
in
the
Atlantic,
The
figure 30
summarizes
some
important
characteristics of
those
groups of tunas(
yellowfin in
that
case).
The
analysis shows that the initial biomass cari be very
high(
possibly more than 10000 tons)
; tunas are concentrated in a very
small
area
(with an area equal to 3 % or 30 % of a one degree
square).
When
the
fishing effort is high (for instance 20 to 40
purse
seiners),
the
daily
catches
are
very high, at
least at
the
beginning of
the
exploitation.
This
high
catch
produces a
reduction o f
the catch rates,
and a complete
removal o f
the
biomass
within a
short
period of
time
(10 to 30 days).
Consequently the analysis of this phenomenon needs very
detailed
statistics
with daily catches for a11 boats,
corresponding size
composition,
and
detailed satellite positions (to
the
nearest
mile). This type of concentration seems to be very often a closed
systemtlike the "urn" model), where there is only some day to day
movement of
the
fish but little or no
emigration
from ,
0 r
immigration to the system.
6.2.3.The Cape Lopez summer fishery : (figure 30)
This
area
has been extensively fished from May to
August,
since
1962, by
a11 types of fleetstlongliners
,baitboats
and
purse
seiners).
This
fishing zone and season is related to
an
oceanographical front
with highly
productive
waters
(STRETTA
1977)
which is stabilized in this area during this season. There
is probably some input and output of tunas into this strata which
are
well
shown
by tagging done in this
fishery
(HARD 1984).
However,
those movement may be limited and it is
interesting to
relate
the total catches in the fishing season to
the
fishing
effort
(expressed
in purse Seine fishing days).
The result
iS
shown in figure 31.
It cari be seen that, as in the theorical urn
model,
there is some maximal production in the local catches
0 f
tunas.
However,
when
the
fishing e-ffort becomes high (as in
recent
years) j
the
catches
are
higher
during
" good"
years(1974,
1978,
19 81--, 1982).
The
catch
is stable or
low
during
other
years(1975,
1976,
1977,
1980,
1983),
quite
independently of
fishing
efforts. In
Süch a
case, it
cari
reasonably
be assumed that
the potential maximum catch is
not
really proportional to the fishing
effort ; however an increased
fishing effort cari produce a
higher average catch , because it
cari
more
easily remove nearly a11 the existing
local
biomass,
especially during "good" years.
ii 7
6.2.4. The Canary Island skipjack fishery.
The
Canary
Island skipjack fishery is a
summer
a 1: t i s a n a 1
fishery,
where
approximately 500 baitboats catch,
among
ç!tLher
species,
an average of 3000 tons of skipjack each year (197":
t 0
1985).
This
fishery
operates
exclusively
around
t he
Cun 3.rl
Islands. It has been extensively studied by spanish scientists off
IEO
who
conducted
tagging cruises on this species
every
;' e a r:
since 1979.
Of the 3515 skipjack tagged during this period,
the
average recovery rate is very high :
20 %.
Most of the tags +1re
recovered
in the local
fishery,
and very few in
o ther
a r 2 a s
(Açores,
Madeira
and
Senegal)
after the end of
the
fishinc,
season.
Those
high local recovery rates strongly suggest
t.Elat.
t. h e re
is a limited local skipjack fraction of
stock,
which is
more or
less
sedentary during the summer
time
season,
‘3 n d
which
suffers a relatively high exploitation rate.
This fishery cari be considered,
at least to some extent., tc
belong
to the "urn" type,i.e.
a close system,
where
durinq a
limited time and in a ïimited space,
there is a high removal rate
bY
the fisheries and a direct interaction between fishing
u n I.. t s
to catch this limited resource.
6.3.
The
statistical
evidence:
cpue and
effort
relationship
within small time and area strata .
The analysis oT fishery statistics, cpue versus effort, at a
detailed time and area scale, for instance 1" square and 15 days,
is
also an
interesting
way to show
the
local
instantaneous
interactions between fishing units.
Following that idea, it has been shown for Atlantic skipjack
(Fonteneau 1986) that :
-when the effort is low within one square during 15 days,the
average cpue is often high or very high.
-when
the effort is high in the same strata,
the
cpue is
most of the time low.
This is
also observed on the frequency of the
overall
catch
rates of FISM and spanish purse seiners during the period 1980 ta
1986 as shown by figure 33.
In
this case of high local effort,
it is clear that
there
was
probably,
at least at the beginning of the period, a
high
abundance which is in contradiction with the low cpue.
The easiest explanation for the final low cpue,
is .probably
related t o
the
urn concept:
the interaction between a
great
number
O f
boats on-a limited local resource Will
increase
the
catch,
but
Will
reduce the cpue,
more quickly
than .a
small
fishing effort.
^l.--.“.-.-.-~-l”-.lll-~~~~..l-.--~~”-l-l.-L
_...”
”
,,.~
~---~I-.~-~“-“_--“r.-...”
1
,-_,.llr
,.
. _ _
I..I-
^_._
_-----_
-_.
5 8
7. GEOGRAPHICAL COMPETITION IN THE MEDIUM TERM
(less t]lan 2 years) BETWZ?BN Ffl#%NG BREBlt %#E BOXE8 MODa%*
7.1. The concept: migrations treated as a diffusion process;
AnY
tuna
fishery
cari be divided in
time(ex.month)
and
area boxes (ex. 10' to 10°),
which are fished or
unfished. In
every area there is at any time,
a gjven number of fishes( often
unknown....).
When the area is exploited , the abundance in each
box cari be easily estimated using an unbiaised cpue index. Aerial
survey
and
direct
estimates of biomass cari
also
measure
the
abundance in each box, especially for the virgin stocks.
From
one period to the other,
the number of fishes in each
box Will change because of :
l.Input to the box from other areas (immigration),
positive
factor.
2.Out:put from the box toward other areas(emigration).
3.Removal b:y the natural mortality .
4.Removal by fishing mortality.
These three last factors are negative ones,
and reduce
the
local population.
Migrations <oan be measured by intensive and repeated tagging
results,
associated
to significant fisheries covering as
many
areas as possible.
Natural
mortality
is not estimated precisely for
most of
tuna species,
but this is not necessarily critical in the
short
and medium term studies .
The
fishing mortality in each exploited box Will depend on
the fishing effort, and more directly on the total catches.
7.2. An example of the boxes model.
Computer
models
cari
easily
be designed to
describe i n
equations
a11
the
previous
concepts of
boxes,
underlying
populations,
migrations
between
boxes,
fisheries and
natural
mortalities.
Such a
mode1 is presently under testing
for
the
eastern
Atlantic skipjack fisheries:
its framework is shown in
figure 34.
This
mode1 has been built in order to
analyze
the
geographical
interactions
between
skipjack
fisheries in
the
Eastern Atlantic (Figure 35).s:This mode1 is a catenary reversible
differential
model,
with
533oxes,
the unit of time being
the
month .
A sixth box,
adjacent to
the 5
others corresponds to
the
unexploited
fraction of the stock.
This sixth box is
the
"cryptic
nursery"
for
thejuveniles during
their
progressive
recruitment ,
and
the "black hole" concept
(Fonteneau 1983),
where
adults skipjack become unavailable and disappear from
the
fisheries.
Such
type
of differential linear mode1 is widely
used in
5 9
many
research
fields in
order
to
describe
diffusion
processes(Sheppard 1962,Jaquez 1972).
In our present model, the status of the system at the end of
each
monthly interval Will depend of eh@ 18 got@Rtaal
migrk3thon
rates between boxes .
7.3.
Example
of skipjack data used for developping and
testing
the model:
The
data
used in this mode1 belong to
the
for
following
types: (a) The monthly cpue by age is givea in number of skipjack
taken by
fishing day of FIS and spanish purse
seiners
(Figure
36).
It has been calculated monthly for the 5 selected areas of
the
model.
The
cpue is calculated on the catch by
size
file,
assuminq
the
same growth pattern as the one used by
FONTENEAU
1986, i.e.
an average growth of 12 cm per year (figure 36). In
those
calculation the the group of fishes measures from 37 cm to
50 cm in August 81.
Some cpue data are missing in 2 cases, first
when
size
data
are
missing or when the
catch is
only
from
baiboats.
(b) The catch is also calculated in numbers,
from the same
size file and for the same group of fishes,
but is estimated for
a11
fleets (baitboat and purseiners) I
in order to
measure
the
total removal on the cohort (VPA concept).
(cl
The
recoveries of tagged skipjack are extracted
from
the
best
ICCAT recovery file presently available
for
the
two
tagging
conducted
during
skipjack year (1981) in
Senegal
and
Ghana areas (Figure 38 and 39).
(d)
The total fishing mortality estimated on
the
eastern
Atlantic
stock
by age follows the general pattern estimated by
FONTENEAU 1986 (Figure 40).
7.4. Estimating the underlying populations and migration rates :
The mode1 will follow numerically a qroup of fish,
from its
early recruitment until the death of the oldest fish.
This basic
qeneral
concept is
similar to the VPA or
yield
per
recruit
calculations.
The method used to adjust the parameters of the mode1 to the
existing data Will be the following :
l.Knowing
the
cpue
by age in each
box,
the
theoretical
underlying
local population in each box must
be proportionaï to
the
local cpue of the age (at each time).
---
2.Knowing the catch by age in each box, the-local underlying
population
must be at least equal to the
observed
catch.
IThe
removal o f
fishes by the fisheries is exactly-known by
fishery
statistics.)
3.The total fishing mortality on the cohort (i.e. on the sum
of
individuals in
a11 the boxes), is
the
fishing
mortality
previously calculated on the overall stock by standard VPA.
6 0
4.The 1 8
monthly
mixing rates between boxes
must be in
agrement
with the observed diffusion of tagged fish.
Only
some
intensive
tagging cruise cari be used to that purpose,
each
one
with
several thousands of tags applied within a very short
time
in a small area (see table 1).
The
migration
parameters estimated by the mode1
should be in
agrement with the observed diffusion shown in figures 38 and 39.
Unfortunately,
this type of adjustment by the mode1 does not
give a unique solution for migration rates,
unless
in an
ideal
and
hygothetic
case with intensive,
generalized and
permanent
tagging in a11 strata.
However,the
present
adjustment of the mode1 to the
actual 4
sets of
data cari give a range of possible
mixing
rates
which
Will be more realistic and quite narrow compared to the
previous
complete uncertainties.
This work is still under progress on the CRODT computer,
and
a
the
possible
range of
mixing
rates
has
not
yet
been
established.
6.4.
Estimating
interactions
between
fisheries with
the
box
model.
Using
the boxes mode1 and its parameters,
it becomes
quite
simple to
calculate
some theoretical
short
and
medium
term
interactions between fisheries operating in different areas.
For instance,the effect of reducing or increasing the fishing
effort in any area,
upon the catches and cpue in a11 other areas
cari
easily be tested by simulations,
Also,
catches and cpue in
each box under various local fishing mortality schemes,
cari b e
tested .
The boxes mode1 is still in its development phase,
especially
in
order to explore the full range of possible
migration
rates
between boxes.
Several of those parameters Will probably
remain
undetermined because of unsufficient data.
However,
some preliminary results may be of some interest:
- most
of
the
time
the
interactions
are
universely
proportionna1 to the distance.
For instance*
the Ghana
skipjack
fishery
has some significant effect
on the Cape Lopez
(average
distance = 600 naut.miles)and Liberia
fisheries (900 miles). The
Senegal
fishery
has very little or no effect on the Cape
Lope2
(1800 miles) gr Angola fisheries (2500 miles), but some effect on -
the Liberia fishery (600 miles) .
- a
short
distance
between two fishing areas is
not an
evidence for strong potential interactions: the Cape Verde Island
and Senegal fisheries are located at a short distance,
200 miles
only,
but
show
very
little or no
mixing
between
the
two
fisheries.
This
absence of
interactions
has
been
clearly
demonstrated by
the
recoveries of two
simultaneous
intensive
tagging cruises conducted in the 2 areas.
- only intensive fisheries and high catches in one area
c an
affect the othes ones,
because of the intensity (in general) and
the
complexity of mixing processes,
at least for skipjack.
For
instance
the
Angola fishery which operates presently at a
10 w
level,
compared to the local biomass, has no effect on any other
fishery
because
the
Angola catch is small
and
the
skipjack
biomass
transient
from Angola
is diluted
very
quickly
among
skipjack
originated
from
other
areas,
especially
with
the
unfished
fraction of
the
skipjack stock
migrating
from
th'e
"cryptic nursery" to the major fishing areas..
7.5. Discussion of the boxes model.
This
type
of mode1 is a potentially
serious
progress
in
analyzing
the stock structure problem ,
the migration
patterns
and the interactions between fisheries.
However
it
requires a huge amount of detailed
statistics,
especially catches and cpue statistics by sizes. Those statistics
must caver a11 fleets(VPA concept).
Intensive tagging with
good
recovery statistics must also be available. Furthermore this type
of
mode1 in fundamentally based on the existence of
significant
fisheries covering preferably most of the stock and operating at
relatively
high
level of fishing mortality (at least
:Locally).
Even i n
the
optimum case when mixing rates
cari be
estimated
duxing
a limited period of intensive researches (such as
ISYP),
it is
not
yet clear to what degree
the -calculated
migration
pattern
cari be extrapolated to an other period.
More
generally
the
year
to
year
variability of
tuna
migrations
a II d
i t s
determinism is still a great unknown .
Also
the
mode1
rely on
several
underlying
hypothesis,
especially on
the homogeneity of boxes and on the
box to
box
relationship,
which
cari probably be more analyzed and improved,.
The
heterogeneity
of the senegalese box has for
instance
been
clearly demonstrated by the low exchange rate between Senegal and
Cape Verde islands.
Furthermore the analysis assumes that
5vow
of
fishes
is followed (catch and cpue) when the individual
are
growing.
The variante of growth between individuals and
between
area
being
apparently
large
(Bard et Antoine,
19861,
this
hypothesis
is not strictly justified.
However, this type of research is probably a very interesting
field to be developped.
8. CONCLUSION.
The present paper has tried to demonstrate the complexity and
the
heterogeneity of
the
interaction
problem
between
tuna
fisheries. It
makes a
provisional
review of
the
present
observations done upon the atlantic tropical tunas in this field.
1t
is clear
that the Atlantic ocean offers good conditions
for
those studies,
because of intensive fisheries,
good
statistics
over a long period of time and some intensive tagging .
6 2
Many
of the Atlantic results cari probably be extrapolated
at least to some extent,
from one ocean to the other,
even when
the ecological conditions are quite different.
This
type of worldwide analysis of the interactions
betqween
tuna fisheries seems presently a key action to conduct.
In
any case this paper well shows the fundamental importance
for any interaction study of :
-Detailed fishery statistics(catch,effort and sizes taken by
gear),with a detailed time and area stratification.
-Intensive tagging.
The tagging of a great number of fish done simultaneously in the
same spot,
such as the senegalese
,cap verdian or japanese
ISYP
tagging,
seems
to be a key action towards the estimation of the
short term interaction between fisheries,
at least those
taking
small fishes.
-
6 3
BIBLIOGRAPHY
BARD F.X.
1984 . Le listao de l'Atlantique.Résultat des campagnes
de
marquages
effectués
de
1980 à
1982.
La
Peche
Maritime,n01275,juin 1984,p.319-324.
BARD F.X. et
ANTOINE L.
1986.
Croissance du
lista0
dans
l'Atlantique est.
Proc.
of ICCAT Conférence en the ISYP. p. 301-
3 0 8 .
CAYRE P. et
BARD
F,X.(en cours
publication
FAO)Biologie
e t
reproduction
des thonidés de l'Atlantique.Dans synthèse FAO
s UIC
les
thonidés de l'Atlantique centre est-FAO technical
document,
Fonteneau et Marcille éditeurs.
FONTENEAU A.
3.975
Note sur les problèmes
d'identification
du
patudo dans les statistiques de pêche.Rec.Doc.Scient. I.C.C.A.T.,
Vol.5,n" l.,pp 168-171.
FONTENEAU A.
1981 . Note sur le mode de calcul de la
prise
p a. ir
unité d'effort des senneurs FISM.
Rec.
Dot. Scient. I.C,C.A.T.,
vo:L.15(1),
pp 407-411.
FONTENEAU A.
1.984.
Analyse de
l'état des
stocks
d'albacore
(Thunnus
albacares) de l'Atlantique au 30 Mai
1984 .Rec.
DO~..
Scient. ICCAT, Vol. 21(2), pp 80-101.
FONTENEAU A.
E' t
ROY C.
1987.
Pêche ~thonière e t
anomalies
climatiques
de l'environnement dans l'Atlantique tropical centre
est en 1984. Rec. Dot. Scient. ICCAT ,Vo1.26(l),p.228-236,
FONTENEAU A.
1.986 a
Analyse de
l'exploitation de
que Iques
concentrations
d'albacore
par les senneurs durant la
période
1980-1983 dans l'Atlantique est .
Rec.
Scient. ICCAT , Vol,, 25,,
pp 81-98.
FONTENEAU A.
1986.
Etat du stock d'albacore de l'Atlantique est
au 30 Octobre 1986.Doc.
ICCAT SCRS/86/60.
En cours
publication
Rec.Doc. Scient. ICCAT.
HUNTER J.,
ARGUE A.W. ,BAYLIFFF W.H.,DIZON A.E.,FONTENEAU A-H.,
GOODMAN D.
and SECKEL G. 1986. The dynamics of tuna movements:an
evaluation of past and present research.
JAQUEZ J-A,
1972.
- Compartimenta1
analysis. in
biology
an Ci
medecine. Elsevier .Amsterdam.,l972.
LAUREC A.
et FONTENEAU A. 1979 u Estimation de l-abondance d'une
classe
d'age.
Utilisation
des
cpue de
plusieurs
enginsy
en
différentes zones et saisons .Rec, Dot. Scient. ICCAT ,Vo1.28(1),
pp 79-100.
64
MIYAKE M.
and
HAYASI S.
1978 . Field manual for statistics and
sampling
SO f
tunas and related species in
the
Atlantic
Ocean.
149p.
PEREIRA J.
1986 Analyse de l'état du stock de patudo atlantique.
Dot.
ICCAT SCRS/86/62.
En cours publication Rec.
Dot.
Scient.
ICCAT.
RICKER W.E.
19'75
Computation and interpretation of
biological
statistics of fish populations.
Bull.
Fish.
Res.
Board Can. :
382 p.
SHEPPARD C.W. 1962 . Basic principles of the tracer method.
Introduction to mathematical tracer kinetics. Wiley,New York,1962.
STRETTA
J.M.
1977 .
Températures de surface et pêche
thonière
dans la zone frontale du Cap Lopez (Atlantique tropical oriental)
en juin et juillet 1972,
1974 et 1975.
Cah. O.R.S.T.O.M. , sér.
Océanogr.,Vol 15 ,n02,1977:pp163-180.
6 5
Table l.- Skipjack taggings done during recent years which cari be
used
for diffusion analysis because of a significant
number of
tags released during a single cruise in the same place.
____-________ ------------------------------ ------ ------------.-.--.--
YEAR
ZONE
COUNTRY
TAGGED
SKIPJACK
-----------------------------------------------------------------
80
GHANA
JAPON
5976
81
GHANA
JAPON
7000
81
SENEGAL
SENEGAL
1391
81
SENEGAL
CAP VERT
2672
82
SENEGAL
CAP VERT
4552
82
SENEGAL
SENEGAL
2794
_____-____- -_.--- ---.- -------------I- ------ ---. -~~~~~--~.“~~I--._--~~--__.-___-_
TOTAL ZONE SENEGAL
11409
TOTAL ZONE GHANA
12976
_____ - ---_-- ----.-------*------ ---- ---- ----.-- ------_.--_-__--- --.- --.-_,___
RECR”ISYENT
/
UNKNOWN
RANGE 01
--
I
I
f-----
’
’50
1 0 Y ” FECUNOITY
O F S T O C K
Figure l.- The
classical
stock
recruitment
relationship
(a)Ricker type and (b) Beverton and Holt,
and the Atlantic stock
relationship
with
the
present level of
stocks
estimated
for
yellowfin, bigeye and skipjack.
,
(max. janv. juin)
-L-l---.-
&~OS d e rsprodue
dstsrminh
m g o n a d e s m û r e s
,--&r”es’to”ts
rannce
(faibles quatités l
Figure 2.- Spawning areas of Atlantic yellowfin (from
Cayre et
al. in press in FAO synthesis on eastern Atlantic tuna)
0
““?h
Twt. Imvlda 0
(1
\\?+a-
- -. ._:- \\ c;
‘<0
Figure 3.- Spawning areas of Atlantic bigeye (from Cayré et al.,
in press in FAO synthesis on eastern Atlantic tuna).
*
sp’
y
70’
60’
SO’
4;
30°
20’
3’k
I
4
1
I
1
8
NO
40
‘C
I
!14
-
. .
I
10'
30’
20'
20’
10’
10'
0’
0'
tQ’
10'
w’
20*
SO’
30’
. .
3
J
f
w’
1IO’
i
20’ l
Figure 4.- Spawning
areas of Atlantic skipjack (from
Cayre et
al.., in press in FAO synthesis on eastern Atlantic tuna).
Figure 5.- Areas of catches of small bigeye (less than 90 cm) by
purse
seiners
(1979-1983) (5a) and by Tema baitboats (1975-1982)
(5b).
.
.
.
.
.
.
*
Y
-10
-
-0
-
-.
Figure 6.- Areas of catches of large bigeye (more than 90 cm), by
purse
seiners
(1979-1983) (6a) and hy longliners
(1978-1982)
(6b).
6 9
SMALL
BIGEYE.
LARGE
BIGEYE
Figure 7.- Catches of small (less than 90 cm) and of large bigeye
(more than 90 cm), from 1955 to 1983.
CP U E
1970
75
80
ANNÉE
CPU E
1965
70
75
80
A N N É E
Figure 8.- Catch per unit of effort of bigeye (predominantly less
than 90 cm), by surface fleets.
-,_“.. ._ . . . 1 . ..-. _ . . . . . “1- -_... .., -.<., _.” _.,._,._, _ _-
7 i,
-..
BIOMASSE
(1000 t)
P U E
5(
PALANGRE
-
-
---
:Oe6
4 0 0
300 Ir--
‘h
---h---\\\\
,
- 0 . 5
2 0 0 -
- - d
lOO-
- 0 . 2
- 0 . 1
t
-
.
.
1
-
-
’
- 1
-
“.
’ 1 -
’
-
’ I
’
’
1960
1965
1970
1975
1 9 6 0 A N N É E
Figure 9.- Index
of abundance (calculated for bigeye
tuna on
japanese
longline fisheries by Honma method,
corrected for deep
longlining) , and adult biomass c-alculated by VPA.
-..-.a 1972-1975
D-----.0 1076-1979
-+ 1960-1063
o::3:;“-
6
7
A G E ( A N N É E S )
Figure lQ.- Average
fishing mortalities by age,
during
recent
years on bigeye tuna.
Q----+
VARIABLE
RECRUITMENT
NUMBERS 1
(a)
BIGEYE
.-. C O N S T A N T
OF BIGEYE
AT .AGE 3.0I
A
Figure ll.- Estimated numbers of bigeye tunas recruited in
the
adult fisheries (result of VPA, from Pereira 1984).
r- ----_ .--.-- --
’
i-ï\\
)r3
9
12a
0
/
FP
l
VI
1 2 b
Figure 12.- Multigear
yield
per re cruit
analysis
f 0 1:
bigeye
small
bigeye fisheries against large bigeye fishery 12 (a).
Th e
Fishing
mortalities by age of each g ear are given in
figure
l:&(b) il
13a
Figure 13.- Fishing areas of small yellowfin (less than 90 cmi,by
pur-se
seiners(13
a) during recent years
(1979-1983) and by
Tema baitboats (1975-1982) (13b)
Figure 14.~. Fishing areas of lar
lowfin (more than 90 cm) by
purse
seiners
(14a) during rec
ars
longliners (1978-1982) (14 b).
and fishing
areas o f
SMALL
YELLOWFIFI
i
--rllr
1960
1965
1970
1975
1980
1985
Figure 15.- Numbers of small (less than 90 cm) and
large
(more
than 90 cm) yellowfin taken in the Atlantic by a11 fisheries.
n-
1980
YEAfl O F 6ECAd,TnÇti?e
Figure 16.- Catch per unit of effort of yel1owfi.n (from Fonteneau
1984)
at age 1 (FISM purse seiners,
:LAUREC and FONTENEAU
1977
method), measuringg the recruitment.
'bOasb
?. u. fi.
. '1
b-9)
-
4
BlOMASSE
:
- 3
lOO,-
4
I
I
I
I
r *
.
6 0
Figure 17.- (a) Catch per u6iit ofeffor?of the japanese longline
fishery
(Honma
index,
uncorrected for
deep
longlining),
and
biomass
calculated by VPA (from FONTENEAU 1984) and 18(b)
catch
per unit of effort on large yellowfin by purse seiners (Fonteneau
1986,
cpue
index corrected for large fishes),and longline
cpue
index during the same period 1969-1983.
/
3
NUMBERS
OF YELLOWFIN
AT AGE 3.0
Figure 18.- Changes in
the
recruitment levels
yel.lowfin fisheries estimated by VPA.
t0
the
adult
F
2. BE
1.50
/4
P.dd$
1
1
6
5
,
‘3 !Q 11 1:: l! lû 1 5 l b 17 18 19 ia ?l :z î! ;-
RfE (TII* )
Figure 19.- Average age specific fishing mortality
for
e a ci ter n
Atlantic yellowfin (period 1980-1983) (from FONTENEAU 1986).
SURFACE (standard days x 1,000)
,50000
Figure 20.- Catches by purse seiners and by longliners on
Large
ye:Llowfin
as a function of their respective fishing effort iFrorn
HUNTER et al. 1986).
100 L /
1
10
dl
3b
46--
ISO
6à FISHING EFFORT
by SURFACE FISHERIES
21b
P
_ 1
ldoo
1 00
ZdOO EFFECTIVE EFFOFV
O N .BIGEYE
by Lonp Lin.rs
Figure
21..- Catches
by purse seiners (2la)
and by
longliners
(21b)
on large bigeye as a function of their respective
efforts
fishing
(curves are adjusted by eye).
Figure 22.- Yield
per
recruit of yellowfin
eastern
Atlantic,
period 1980-83,
Ricker mode1 corresponding to F vector of figure
19 (from FONTENEAU 2986).
7 5
1
t
1
Figure 23.- Yield per recruit of bigeye, a11
Atlantic,
p e IT io d
1980 to 1984 , Ricker mode1 (from PEREIRA 1986).
-
-
THEOAETICAL
YIELO
f’ER RECR~IT
G N AOULT
10
- - - - - OESERVED YIELL) O F FISHERIES
5
I
70 -74
80-83
F
PUR-SE S E I N E
LONGLINE
YELLOWFIN
F
62!65
71’-75
F
PURSE S E I N E
LONGLINE
Figure 24- Theoretical and observed yield of purse seiners
and
longliners,for
yellowfin and bigeye tunas.The theoretical
yield
is a yield per recruit analysis result,
Ricker model,
where the
yield per recruit is multiplied by the recruitment estimated
for
the
adult fishery . In this Y/R analysis,each gear is supposed to
h n
f i chincl ~-l,Y~~
7 6
e
KIRGIN STOCK
'ED STCICK
VIRGIN STOCK
ZXPLOITED STOCK
-'-----.-L--B ,-._
-
--
.
.
t
-
:
. .
.
'
_
,_
\\
I
\\
:
\\
.
.,
.
<
.
\\
,
.
-.
.
~
Figure 25.- The
hypothetical
vertical
stock
structure
of
yellowfin
and
bigeye for the virgin stocks and
for
the
fully
exploited stocks.
7 7
Figure
26.- Catch
per
unit of effort of
japanese
lc: n g 1 i ri e r s
during
the first quarter e-g.
the better fishing season in
the
Gulf
of Guinea (in tons/1000 hooks),
for the historical fishery
1961
to
1969
(Each
month and 5 to 5
degree
square
cpue i s
represented randomly in the corresponding 5 degree square).
Figure 27.- Catch per unit of effort of japanese longliners during
the
first quarter ( tons/1000 hooks),
during the recent
period
1970
to
1982 (Each month- 5 to 5 degree
cpue i s
represerted
randomly .in the corresponding 5 degree square).
7 8
28a :
THE "URN" MODEL CONCERT
VIRGIN POPULATION
,PHASE 1 OF EXPLOITATION
_ FINAL PHASE
TQTAL CATCH PHASE 1
8 ,
-’
TOTAL CATCH
28b
C A T C H , E F F O R T j
C P U E R E L A T I O N S H I P
CATCH -
LOCAL AVERAGE
BIOMASS
A V E R A G E M A X I M U M C A T C H
- - - - - --_-- _--_-_ - - - --.
---_--- -.---.-
___--
5000.
XX-
l
-
-
d
LOCAL
-
EFFO
FIRST FISHEAY
EFFORT
II
E F F O R T O F A D O I T I O N A L FISHERY
Figure 28.- The "urn model" concept:
(28a)The general
simple idea of a closed and limited
exploited
resource.
(28b)The catch and catch rate trends when an additionna1
fishery
is
added to a first fishery:(a) is the cpue of the first fishery
before the arriva.1 of the second one and (b)is the cpue after -
1st concentration
2 nd concentration
3rd concentration
Cp”e
UNDERLYING
ESTIMATED
POPULATION
I
-.^ I
\\
POP
o - - - - o
?--y
9
(a)
3 DAYS INTEVAL
CC)
5 0
OW8"
weight
I
1
I
1
1
-F--
I NTERVAL
Figure 29.- Exqmple of parameters concerning
the exploitation; of
three yellowfin concentraCions in the eastern Atlantic.
fa)
observed
cpue
of--purse seiners( in tons by 10
hours of
searching),and
estimated
underlying
populations by 3
tiays
-
periods(From Fonteneau 1986).
(b)Numbers o f
positive
sets
by 3 days
periods(Which
cari b e
considered
as
the
number of
schools
removed
from
,t h e
concentration)
(c)Average weight of the yellowfin caught.
y e a r
Figure 30.- The
fishing
area o f
Cap2 Lopez :
geographical
distribution
of the catches by purse seiners (FIS and Spanish),
by 15 days period from 1978 to 1984 (from FONTENEAII and ROY 1987).
-
F I S H I N G E F F O R T
( F I S P S F I S H I N G DAYS)
Figure 31.- The relation observed between total catch (yell<:lwfin,
skipjack and bigeye) and fishing effort in the area of (:ape ~,~>pez
during
the fishing season from may to july,
from 1969
(Curve adjusted by eye).
to
1984
Figure 32.- Average
fishing
zones of skipjack by
on e
decrree
squares, a11 fleets, period 1975-1982.
8 2
e f f o r t
-5
5-9.9
,
+ 55 qq
1 ’
.
I
0
0
D
00
00
0
0
0
0
O
0:
0
0
0
100gbservatiOns
0
o
l
0
O
0
0
0
0
0
0
0
Figure 33.-Frequency of
total cpue as a
function o f
fishing
effort
for
spanish and FISM purse seiners _fleets,
during
the
period 1980-1986.
Each circle is proportionna1 ta the number of
catch
rates observed in the l'squares during 1.5 days periods for
increasing intensity of fishing efforts. Frequencies greater than
200
have been limited arbitrarily%t the 200 level in
order to
facilitaté the overall representation.
8 3
1
-.- --t-- j
---T---Ii “!’
-- -.--IL--- r v-jii
i l
.- -~~I---~--
J
-f--l“.’
Figure
34.- The concept of the boxes
mode1 applied to
skipjack
in the eastern Atlantic ; each arrow shows a monthly tranfer rate
of individuals from one box to the other.
Figure
35.- Hypothetical
migrations
of skipjack
tuna
in
the
eastern
Atlantic,
based on tagging results (from Cayre et al.,
FAO tuna synthesis under press).
----
.-..
--._-
---..--.-*--*P-ms
.^,_
_‘.,,_
-
--z..il/.-
19---.IXX*-I,.
_ .
,l”__
.,II,
,......
.-“-
--.-
--
---.-.__
8 4
Figure 36.- Monthly cpue for an age group of skipjack, by
area,
during
the
period
1979-1982
(each
circle
has an
area
proportionna1
to the cpue);
a blank means a zero cpue,and an x
means
no purse Seine effort data or no size
data.This
skipjack
group is
from
37cm to 50 cm in august 1981 (date of
japanese
tagging) i
a11
individual are growing at a constant rate of
Icm
per month before and after this date.
0 1000 T.
-
Figure 37.- Monthly catch of skipjack,
by area,. during the period
1980-1983 (each circle has an area proportionna1 to the catch).
months
liberty
<’
0 0 0 c/ c>
I
I
.
--
--
8 6
I
l-------~--T---1
3
4
-’ A G E (?mnées)
Figure 4Q.-Vector
of fishing mortalities by age
estimated
for
skipjack (From Fonteneau 1986).
INTERACTIONS BETWEEN TUNA FISHERIES
A CRITICAL REVIEW BASED ON SOME ATLANTIC EXAMPLES
by
ALAIN FONTENEAU
ORSTOM scientist
CRODT
BP 2241
DAKAR
SENEGAL
S U M M A R Y
This
paper
describes
some
interactions
observed in
the
Atlantic Ocean between tropical tuna fisheries.
Seversl types of
interactions
are
distinguished. It
is first
shown
th a t
for
tropical
tunas,
there
is not presently a
clear
relationship
between for fisheries adults and for fisheries juvenile.
This is
probably because, despite of the high level of fishing effort r a
recruitment
overfishing never has been observed in the ?\\tlantic,
probably
because of the high fecundity
and wide spawning
areas
of these species.
The interactions between small tuna fisher.ies
and
large
tunas fisheries are also analyzed for
yellowfin
<lnd
biyeye.
For
both
species this interaction is estimateà
ter be
significant at the ocean level,
because of the important catches
of juveniles.
The interaction between gears catching large tunas
such a s
longline
and
purse
Seine,
in
analyzed.
T h e
Larticularities
of the vertical stock structure are hypothesized
for
both
species
from
the catches and
cpue b y
gear.
As a
consequence of this vertical stock structure,
the purse Seine is
the
gear which
cari take full profit of the
yellowfin
biomass,
and
longline
the only gear capable to exploit
intensively
t h e
deep
bigeye
stock.
Some other short term
interactions
within
small
time
and
area units are also described and
discussed ;
several
exemples of those interactions
are
presented. A
mo r e
detailed
discussion
concerning
the
short
term
interactions
between fisheries operating in different areas is presented using
a "boxes model".
Such type of boxes mode1 which are
developped
needs
a huge amount of data
(detailed catch,
effort
and sise
statistics by small time and area strata, intensive tagging), but
seems to
be a key tool to evaluate the short
and
medium
term
potential
interactions between fisheries exploiting a
migratçry
resource
considered
as a
stock.
Al1
the
present
wcrk
4 6
demonstrates,
at least,
that the interaction study needs,
much
more than any other stock assesment work , very good and complete
statistics associated with intensive tagging.
RESUME
Cet
article
décrit quelques interactions
observées
dans
1’Atlantique entre des pêcheries de thonidés tropicaux, Plusieurs
types
d'interactions
sont
distingués.
L'absence
de
re lat ion
claire
entre la
taille du stock reproducteur et le
n iv e au
du
recrutement
est
tout
d'abord
montrée.
Cette
stabilité
du
recrutement et l'absence de chute du recrutement qui est observée
malgré
les
niveaux
de pêche élevés correspond à
l'absence
de
"recrutement overfishing".
Cette situation est interprêtee comme
@tant
due a la forte fécondité et à l'existence de vastes
z 0 ri e s
de
ponte
qui sont des caractères généraux de
ces
especes.
L a
compétition
entre les pêcheries qui exploitent les juvér,iles
et
c e 11 e s
qui exploitent les adultes d'albacores et de
patudo
2 s t.
examinée. Celle-ci semble significative pour les deux especes, au
moins
un niveau de l'océan,
par suite des prises importantes de
juvéniles qui sont observées. La compétition entre les engins qui
capturent
les
individus de
grande
taille,
par
exemple
I e s
senneurs et les palangriers, est analysée. Pour les deux espèces,
cette compétition est interprêtée à partir d'une hypothèse sur la
structure
verticale
des stocks reposant sur les
tendances
des
prises et
des efforts par engin.
Il apparait que la senne
t! s t
l'engin
qui
est le
mieux à
même
d'exploiter
pleinement la
biomasse
d'albacore,
alors
que la palangre est le
seul
engin
capable
d'exploiter intensément les gros patudos de
profondeur.
u n
autre type de compétitions,
celles observées
à c 0 u r t.
terme
entre
des pêcheries pêchant dans de petites
zones,
sont
a u s s 1
examinées et
plusieurs
exemples
sont
présentés.
Enfin
1.1 n e
discussion
concernant
les
interactions à
court
terme
entre
pêcheries
exploitant
des
zones de
pêche
différentes
est
presentée.
Un modèle informatique "à boites" a été établi à ce,t
effet .
Ce type de modèle qui reste en cours de
développement
utilise une grande masse de données détaillées : prises et prises
par
unité
d'effort
par tailles, selon des
zones et
périodes
réduites,
marquages
intensifs.
Cette approche semble toutefois
être du plus grand intérêt futur pour évaluer les interactions a
court et
moyen
terme
entre
des
pêcheries
exploitant
'3 I: C?
ressources
migratrice.
Le
présent
travail
montre
bien
la
nécessité,
pour conduire à bien les études sur les
interackions
entre
pêcheries,
d'excellentes statistiques de pêche,
fines et
complètes, associées à des marquages intensifs.
48
1 - INTRODUCTION
The
concept of
interactions between
fisheries is
quite
complex
and cari be studied from different points of view.
Among
others, three major types of interactions cari be distinguished :
(a). .
TO what extent
large size tuna fisheries cari compete
with small size tuna fisheries,
for long living species? (In the
medium and long term).
a
(b)..
TO what extent fisheries catching small size tunas cari
compete with large size tuna fisheries ?
(In the medium term, in
general during the exploited phase of the fis'h ).
(c).. T O
what extent two or several fis‘heries catching
the
same
sizes
of fishes are competing for this common ressource ?
(In the short term ).
In
this type of interaction ,two cases need to be
analyzed
separately :
- Those
exploiting small tunas (in general a
plurispecific
mixture of small yellowfin, bigeye and skipjack ) ,
- Those exploiting large size tunas .
This
distinction seems necessary because small
size
tunas
live in plurispecific surface schools, while :Large size tuna tend
to be
in both deep and surface
layers,
with a
differential
geographical
distribution between species .
These
studies of interactions between fisheries need
several
types of data which cari be listed as follows :
(1) .A
correct
species identification of catches in
the
statistics.
When
small
bigeye are called yellowfin
and
both
species often registered as skipjack in the statistics, it
Will
be
impossible
to estimate any interaction for any of the
three
species. Those problems of species identification are often found
for small size tunas and difficult ta solve.
(2).A good size sampling for a11
f'isheries ,
necessary to
study the trends of age specific cpue and to conduct VPA .
(3).Catch
and effort statistics established for each
gear,
with
detailed time and area stratification,
which are necessary
to
calculate
series
of cpue which
cari be
representative of
- . abundance of each year class .
:
(4).Good
knowledge upon stock structure,
p r e f e r a b l y b a s e d -
u-f?on
direct
evidence,
such a s
those
obtained
bY
intensive
tagging.
(5).Good
biological
studies
especially on
growth,
reproduction
processes
and
fecundity.
Growth
knowledge is
necessary to
keep
track
of cohorts of
fi.shes
exploited by
fisheries ;
spawning
biology is
necessary
to
estimate
potentials
of the
spawning stocks.
4 9
(G).Preferably high level of fishing efforts,
covering
various areas and a11 sizes of fishes, SO that the complete .range
O f
geographical
distribution
and
a11
ages of
f i s h e s a r e
exploited.
(7) a,
Stock
assesment
studies
especially
stock
size
estimates by age, based on Virtual Population Analysis (VPA).
(8).
Yield
per
recruit analysis based upon
a9e
specific
fishing mortalities of each gear.
‘
(9).
Mathematical models must be developped on computers in
order to analyse the interactions between fisheries. Those models
are
still quite new in the field of tuna stock managment ,
but
are
to be
developped in
order to
better
understand
the
interaction problems ;
Al1 these results must be combined and compared.
When
t:ley
are
obtained
during
a long period of time,
they
cari
provide
reasonably
good answers on the different types of
interactions
between fisheries .
The
Atlantic
ocean offers good exemples for
a11
possible
types of
interactions(type a to c);
furthermore
the s e
interactions
cari be welï analyzed because of intensive statistics
and
research
conducted in the Atlantic during long
periods of
time
under the coordination of the International Commission
fOL
the Conservation of Atlantic Tunas ( ICCAT).
Al1 researches from
type
1 to
~type 9 have been done in the Atlantic at
different
degrees,
depending on species .
Subsequently
the
goal of thi-s paper Will be to review
a II d
discuss
the present knowledge and researches on the interactions
in the Atlantic.
2 .DATA USED.
The
data used for this study are extracted from the ICCAT data
base
available at
the end of the year
1986.
Those
d,ata
are
basically
the
catch,
effort
and size
data ,
for
yellowfin
skipjack and bigeye tuna,
available during different periods of
time
for each fieet,
with a variable degree of
precision.
For
more details upon the nature of those data, the ICCAT Data Record
num.26 cari be usefully consülted.
In the &se of some fleet where
the
data
was net available,
the standard-strata
substitutions
done
by SCRS scientists were performed when necessary
(sec
the
reports of ICCAT ad hoc working groups). --
,,-,.-_
“~--“-.-
.c<
_-
..”
-.
.--
-‘.,----
“-
,_-_-
..-_-^,-,
__-.%,‘ml-
----
^.I-
<S.<I_
--
5 0
3 .
HOW
MUCH LARGE TUNA FISHERIES
CAN AFFECT SMALL
SIZE
TUNA FISHERIES ?
In
most
exploited fish stocks- at least small pelagic
and
demersal species - there is some relationship between adult stock
of spawners and number of recruits generated by
this
spawning
stock .
In
most
fish
stocks,
this
relationship
between
recruitment and :stock is highly variable,
but a reduction of the
spawning stock will tend either to decrease
(usually the case),or
to increase (Ricker type model) the recruitment level (sec figure
la).
However
this
typical stock recruitment mode1
dont
really
amly
for tropical tunas :
in any case ,
at least for tropical
species , it seems that recruitment shows only some moderate year
to
year
variability,
but
no trend ,
even at
high
level of
exploitation of the adult stock (see figure lb) .
However
this stability of recruitment over a large range
of
spawning stocks cannot be indefinitely extrapolated :
if the
number
O f
spawners
tends to
become
equal to
zero
the
recruitment is
necessarily affected .
This critical
leiel o f
spawners beyond which the recruitment could be affected has
been
hypothetized to a 10 % level of the virgin
stock, but
has never
been observed in tuna
fisheries , at least for tropical species.
In fact a decreasing recruitment trend due
to overfishing -
i-e. a
recruitment
overfishing- has
never been
observed
for
tropical tunas , even for highly exploited
ones .
Consequently
the
present paradigm for
tropical
tunas is
that, at
moderate or relatively high level of fishing efforts,
there
Will
not be
any effect or
interaction of
the
adult
fisheries on the juvenile ones .
This
specificity
of tunas (at least the tropical
species)
seems to
be related to the extremely high
fecundity of
_
these
individuals ;
each
female of
mature
yellowfin , skipjack or
bigeye Will spawn several millions of eggs,
several
times
each
year,
covering extensive areas and periods of time (Cayre et al.
in press FAO) (sE?e figure 2 to 4) , where their larvae cari always
(statistically) find good conditions for their survival . In such
conditions of
spawning
the recruitment Will be
limited in
general,
much
more
by the carrying capacity of
the
nursery
areas , than by the size of the spawning stock a
4 .
HOW FISHERIES CATCHING SMALL TUNAS
WILL AFFECT-LARGE ~TUNAS
FISHERIES ?
4.1.The problem:
This
type of
interaction concerns
mainly
yellowfin
and
bigeye
tunas,
which
are
exploited at small sizes by
surface
fisheries
and later at large sizes by longline and
purse
Seine
fisheries.
5 1
How
the fisheries catching juveniles Will affect
fisheries
catching adults is consequently an important question to salve .
The
Atlantic
ocean offers good observations
and
relevant
analysis in that field.
This
type
of study is based primarily on the
analysis of
fishery data.
4.2. Atlantic bigeye tuna .
It should be first noted that the fisheries of juveniles are
located in
areas
which are different from the areas
0 f
adult
concentrations (sec figures 5 to 6).
This observation is
easily
explained by
t.he
physiology of the species :
the young bigeye
tunas are physic'logically a tropical species,
living in
surface
warm
waters;
it becomes a temperate deep species when
becoming
adult ,
except during seasonal spawning when they corne back to
tropical waters (see figure 3).
The
juveniles of
bigeye
tuna
are
quite
diff icult
t 0
recognize from
the
juvenile yellowfin . In
the Atlantic as in
other
oceans
the small bigeye have been
widely
misidentified
with
yellowfin for many years.
This species composition problem
O f
juveniles
has
been
treated
seriously
only
since
1 $7 7 0
(FONTENEAU 1975). A systematic control of the species composition
0 f
the small tuna catches has been developped in the At1anti.c: on
a
routine
basis
only since 1979.
Historical
data
have
btren
tentatively
corrected b y
ICCAT in
1984,
but
are
sti11
questionable. A
review of
the pending
problems
for
species
identification
is given in the report of the ICCAT working group
held in 1987 on this subject.
The
catch
of small (less than 90 cm) bigeye taken in
the
Atlantic
since 1955,
beginning of the industrial
fisheries,are
shown in .figure 7, together with the catch of large bigeye.
The
cpue o f
small bigeye
for
surface
fisheries
shows
fluctuations,
but
without
trend (sec figure 8 ).
The cpue of
large bigeye,
given by the longline Honma index (sec figure
9) )
shows a moderate and regular decline since
the
beginning of the
fisheries.
The
stock nssessment analysis on bigeye concludes that this
decline cari be explained by the increased catch of the longliners
and by increased catches of juvenile;
the great numbers of small
bigeye
taken
by surface gears correspond to a
relatively
low
fishing
mortality
(sec
figure
10)
and
has
had
only
a
moderate
effect o n
adult stock size.
This
is shown
bY
the
results of the VPA,
which give estimates of the recruitment
in
t b-e
adult- fishery.
The decrease of this adult recruitment is
-_
12 le ar ,
at a rate-of approximately 20 % between the virgin
s teck
and recent. years (sec figure 11 ),
but
is relatively moderate.
:IX
cari be noticed also that present
yield
per recruit analysis
indicates that the juvenile fishery affects the longline :fishery,
but does
not significantly reduce the overall yields, at
least
a t
present
levels o f
fishing mortalities
(sec
figure
12)
(PEKEIRA 1986).
5 2
4.3. Atlantic yellowfin.
The
fishing
zones of small yellowfin are
uite similar to
the
adult
ones(see
figure 13
and 14 1, at
i! east
for
purse
seiners.
However
it is noticed that small yellowfin in
general
stay in
more coastal waters that the adults .
The
adults 'of
yellowfin
taken by
longliners
also
show a
wider
area o f
distributiontsee
figure 14b 1,
scattered between 10' north
and
south, from America to Africa.
The
numbers
of small (-90 cm) and large(+90 cm)
yellowfin
taken by atlantic fisheries are given in figure 15.
This
figure
shows
well
the
dramatic
increase of small
yellowfin in
the
catches du:ring recent years (by baitboats and purse seiners), and
the increase of catches of large yellowfin (by purse Seine).
The cpue for small yellowfin,
(baitboats and purse seiners),
shows some year to year variability, but no definite trend, sug-
gesting a relatively stable recruitment (figure 16). The cpüe for
large yellowfin is given for two gears :: longliners catching only
large fishes, and purse seiners (sec figure 17). In
order to ob-
tain surface index for adults, the
usual
overall
purse
Seine
index has been corrected by the amount of small yellowfin in the
catches. Those two adults cpue show, for recent
years a similar
serious decline. This decline of the adult cpue corresponds to ,a
decline of the adult abundance. The VPA analysis shows that this
decline of adult stock is a consequence of two factors:
l.the
development
of juvenile fisheries which
reduces
the
recruitment to the adult stock .
2.the increased effort and increased catches on adults
which
reduces directly the adult biomass.
Present analysis suggests that the first effect, interaction
between
j uvenile
and
adult
fisheries,
would be
relatively
significant.
This is
shown by
t he
decline o f
the
adult
recruitment
(figure 18)estimated by VPA at approximately 50 %,
decline due to the increased catch of juveniles.
However subsequent yield per recruit analysis shows that the
juvenile
fishery
does
not seriously decreases
the
yield
wr
recruit
of
the
overall fishery because the juvenile
catch
in
weight is
only sligthly less than the
subsequent
theoretical
loss of the adult fisheries (FONTENEAU 1986).
4.4.
Conclusion :
juveniles
against
adults fisheries.
For
bath
species,
the
catch of juvenile seems to
reduce
significantly
the catches of the adult fisheries of about 20 %
for
Bigeye and 45 % for yellowfish this interaction seems to be
moderate
compared to the millions of small flrshes -taken ,because
of the relatively low fishing rates estimated on
juveniles. .-
It should also be noticed that the present yield per recruit
analysis
indicates for both species that the--overall
yield
per
recruit
h.as
suffered
only a minor reduction
due to
juvenile
fisheries.
5 . HOW LONGLINE AND PURSE SEINE SHARE THE BIG TUNA BIOMASS ?
5 3
5.1.Catch versus effort for bigeye and yellowfin for the two
gears.
HOW the two gears compete in the short term to exploit l.acge
yellowfin and bigeye tuna is an important problem.
The
relationship between fishing efforts of purse Seine ‘2nd
longline,
and
their
respective catches of large
fishes
shows
completely different patterns for the two species :
. for yellowfin, the increase of longline effort has produced
a decreasing yield curve, while a simultaneous increase of effort
by
purse seiners has increased the catches at a level much
more
higher than the catches by longliners (sec figure 20).
* for
bigeye
tuna there is a
reverse
observation :
the
increase of longline effort is still producing an increase of the
catches,
while
the
increase
of surface effort has produced a
stable or even decreasing level of catches (sec figure 21).
These
observations are in contradiction with the yield
per
recruit
analysis ,
which indicates that increase of effort by
purse
seiners should have increase the catches of
large
bigeye
and
that
the longliners should have increased
their
yellowfin
catches.
This
contradiction
between the expected and
observed
yields fior the two gears and the two species is summarized on the
figure ;!4.
5.2.The vertical stock structure heterogeneity hypothesis.
The
m o s t
probable
explanation
for
this
observation
i. s
related to a vertical stock structure heterogeneity:
- Yellowfin is a surface tropical tuna, even at large sizes.
The purse seiners cari exploit the ressource with great efficiency
a n d
high fishing mortality .
The longliners cari exploit only Ü
rnargi.nal
and
fragile fraction of the
stock,
the
deeper
one.
Con s e que n t 1. y
the longliners cari never exert a real high
fishing
rnortality
on the total stock because of the excessive
reduction
0 f
the deep biomass.
Consequently the yield per recruit concept
does not
really apply to the longline yellowfin fishery
because
the longline adul.t recruitment does not show the same trend
that
the purse Seine adult recruitment .
--Bigeye is an opposite case : being basically a deep and
tem-
perate
tuna,
the tropical purse seiners Will always exploit tne
ressource
with a
very low
fishing
mortality,
even
at
h iqh
nominal fishing effort.
Any increase of fishing effort does
r-1 0 t
produce an increase of F. The
longliners,
-on the contrary,
c an
easi.ly concentrate their effort on deep-bigeye,
and an increased
longline
effort
produces
a corresponding in-crease in
fishing
mortality,
especially with the introduction of the more efficient
deep longlining.
This
general hypothesis has been summarized
graphically
2. n
figure 25.
Under
the
present
hypothesis
the sharing of
the
adult
5 4
biomass between surface and deep gears would be related much more
to a
vertical
stock structure heterogeineity,
than to a
real
interaction :
- for bigeye,
the decline of surface adult cpue is not
due
to
the
longline
catches or to
a real decrease of
the
total
biomass.
- for yellowfin,
the early decline of the longline cpue is
not a consequence of interaction with purse Seine fisheries,
but
would be related to a vertical stock structure heterogeneity .
5.3.0ther
more direct interactions between purse seiners and
longliners to catch large tunas.
There
is still probably some real and direct
interactions
-between the two gears on the adults ;
these interactions cari be
of two types :
.
The "urn"
interaction,
as described in chapter 6 ,where,
in summary, a tuna taken by one gear cannot be taken by the other
gear,
which gives an advantage to the more efficient gear.
.
The geographical interaction,
where two gears exploiting
the
same
area
are
more in
interaction
than i f
they
are
exploiting
distant
areas,
even in the case of a unique
stock.
This
is related with the probability of migratory
distances
0 f
tunas : the probability
of moderate migration (for instance less
than
500 miles),
is always greater than the probability of very
large
ones( for instance more than 2000 miles).
The
situations
are different depending on the species:
(a)Yellowfin
In- that field it is noticed that the areas where the highest
purse Seine catches of large yellowfin are observed are the areas
where
the
highest
longline
catches
rates
were
historically
observed (figure 26).
Presently it is noticed that the
longline
cpue
are lower (figure 27) that the average in this fishing zone
for the purse seiners (see figure 14a).
This is
probably a local consequence of high
purse
Seine
catches
taken
in this area.
This geographical
interaction is
similar
to the interaction described in the simple "urn
model".
However
the
migration of large tunas are much
less
known
and
probably
more
important
that
those of
small
tunas; as a
consequence
the
strong
hypothesis of a close system
does
not
awly
well
to the adult yellowfin or bigeye
fisheries in
an y
area.
-_
-
(b)Bigeye :
It is
clear for the adults of this species that the
purse
Seine
fishery
has
no sigificant
geographical
effect o n
the --
longline
fishery. This is due to the low purse Seine catches and
to the differences in fishing areas of the two gears (see
figure
6).
6. SHORT TERM COMPETITION IN THE SAME AREA : THE "URN" mode1 :
i) 5
6.1 *Concept:
Tunas are clearly migratory ; however duri;: limited perinds
and
in
relatively
small
areas
they
cari
considered as
sedentary,
and
considired as a locally and
temporaly
isolated
fraction of stock.
In
this
case,
the
term
"fishing area"
cari
describe a
relatively small geographical unit in which tuna cari have
ra:ncom
movements.
As an exemple ,
an area of S" by 5' square, i.e. of
approximately
3600 nautical square miles,
would be a reasonable
unit to study.
In such type of interaction,
only the short term
interactions,
such as from 1 to 3 months,
are concerned ; since
tunas are migratory species,
it is quite obvious that a group of
small tunas staying in a given area Will migrate in an other area
if the elapsed time is long.
In
theory, if
the
fraction of a stock is stable
i 11
the
area(e.g. n o
emigration and no
immigration),
the
interaction
problem
becomes quite simple.This stability of the sub stock cari
often be assumed during limited periods of time.
In such case the studies are simplified because:
-the growth of the fish cari be neglected.
-the
natural mortality becomes much less important that
i n
the long term interaction studies;
-fishing
mortality
rates cari be easily calculated if
the
total
catches
a r e k n o w n ,and if there is a
measure of
local
abundance(loca1 cpue for instance).
In
this simple case,
the recruitment in the area
w i 11
b e
ishared
bctween simultaneous fisheries.
In such a simple
c a s e ,
the
interaction
between
fishing
units
Will
depend o f
the
available
total biomass and on the relative effort and catch
3f
each gear.
This type of competition wiI1 be the "competition"' of
type
1 as described by Ricker 1975:
the catch of each day by a
boat reduces the catches of the other units later.
The abundance
of fish is lower and lower as the fishing is developping, and the
rate
of this decrease is proportionna1 to the number of
boats.
This
has
been
expressed in
figure 28b
which
is a
local
production type curve,
within a limited time and area stratum :
it shows the expected relationship between the local catch during
a
fishing seasor.
and the total fishing effort.
In this example
the
initial biomass and potential catch is 10000 tons wh;ch
c 3x-l
be taken only with a very high fishing effort. If a local fishery
working on this substock takes 5000 tons per fishing season
, t'îe
addition
of a
new
fleet
catching
5000
tons
Will
r e du 12 e
significantly the catch
rates and catches of t.he first fleet.
6. 2. Possible exemples of the "urn" mode1 :
6.2.1.Introduction:
-
It is
quite difficult to find in the real
world
of
tuna
fisheries
a true example of this urn model.
This difficulty
.i s
primarily
due to the migratoxy nature of the tunas.
However
(3 t
least
three
type of examples,
belonging more or less to
this
“ u r n " conce,pt, cari be described in the Atlantic ocean.
5 6
6.2.2. The concentrations :
Concentrations
have
been described in
the
Atlantic as
important
groups
O f
schools which are fished in a
small
area
during a short period of time.(FONTENEAU 1986).
Several examples of those concentrations have been analyzed
in
the
Atlantic,
The
figure 30
summarizes
some
important
characteristics of
those
groups of tunas(
yellowfin in
that
case).
The
analysis shows that the initial biomass cari be very
high(
possibly more than 10000 tons)
; tunas are concentrated in a very
small
area
(with an area equal to 3 % or 30 % of a one degree
square).
When
the
fishing effort is high (for instance 20 to 40
purse
seiners),
the
daily
catches
are
very high, at
least at
the
beginning of
the
exploitation.
This
high
catch
produces a
reduction o f
the catch rates,
and a complete
removal o f
the
biomass
within a
short
period of
time
(10 to 30 days).
Consequently the analysis of this phenomenon needs very
detailed
statistics
with daily catches for a11 boats,
corresponding size
composition,
and
detailed satellite positions (to
the
nearest
mile). This type of concentration seems to be very often a closed
systemtlike the "urn" model), where there is only some day to day
movement of
the
fish but little or no
emigration
from ,
0 r
immigration to the system.
6.2.3.The Cape Lopez summer fishery : (figure 30)
This
area
has been extensively fished from May to
August,
since
1962, by
a11 types of fleetstlongliners
,baitboats
and
purse
seiners).
This
fishing zone and season is related to
an
oceanographical front
with highly
productive
waters
(STRETTA
1977)
which is stabilized in this area during this season. There
is probably some input and output of tunas into this strata which
are
well
shown
by tagging done in this
fishery
(HARD 1984).
However,
those movement may be limited and it is
interesting to
relate
the total catches in the fishing season to
the
fishing
effort
(expressed
in purse Seine fishing days).
The result
iS
shown in figure 31.
It cari be seen that, as in the theorical urn
model,
there is some maximal production in the local catches
0 f
tunas.
However,
when
the
fishing e-ffort becomes high (as in
recent
years) j
the
catches
are
higher
during
" good"
years(1974,
1978,
19 81--, 1982).
The
catch
is stable or
low
during
other
years(1975,
1976,
1977,
1980,
1983),
quite
independently of
fishing
efforts. In
Süch a
case, it
cari
reasonably
be assumed that
the potential maximum catch is
not
really proportional to the fishing
effort ; however an increased
fishing effort cari produce a
higher average catch , because it
cari
more
easily remove nearly a11 the existing
local
biomass,
especially during "good" years.
ii 7
6.2.4. The Canary Island skipjack fishery.
The
Canary
Island skipjack fishery is a
summer
a 1: t i s a n a 1
fishery,
where
approximately 500 baitboats catch,
among
ç!tLher
species,
an average of 3000 tons of skipjack each year (197":
t 0
1985).
This
fishery
operates
exclusively
around
t he
Cun 3.rl
Islands. It has been extensively studied by spanish scientists off
IEO
who
conducted
tagging cruises on this species
every
;' e a r:
since 1979.
Of the 3515 skipjack tagged during this period,
the
average recovery rate is very high :
20 %.
Most of the tags +1re
recovered
in the local
fishery,
and very few in
o ther
a r 2 a s
(Açores,
Madeira
and
Senegal)
after the end of
the
fishinc,
season.
Those
high local recovery rates strongly suggest
t.Elat.
t. h e re
is a limited local skipjack fraction of
stock,
which is
more or
less
sedentary during the summer
time
season,
‘3 n d
which
suffers a relatively high exploitation rate.
This fishery cari be considered,
at least to some extent., tc
belong
to the "urn" type,i.e.
a close system,
where
durinq a
limited time and in a ïimited space,
there is a high removal rate
bY
the fisheries and a direct interaction between fishing
u n I.. t s
to catch this limited resource.
6.3.
The
statistical
evidence:
cpue and
effort
relationship
within small time and area strata .
The analysis oT fishery statistics, cpue versus effort, at a
detailed time and area scale, for instance 1" square and 15 days,
is
also an
interesting
way to show
the
local
instantaneous
interactions between fishing units.
Following that idea, it has been shown for Atlantic skipjack
(Fonteneau 1986) that :
-when the effort is low within one square during 15 days,the
average cpue is often high or very high.
-when
the effort is high in the same strata,
the
cpue is
most of the time low.
This is
also observed on the frequency of the
overall
catch
rates of FISM and spanish purse seiners during the period 1980 ta
1986 as shown by figure 33.
In
this case of high local effort,
it is clear that
there
was
probably,
at least at the beginning of the period, a
high
abundance which is in contradiction with the low cpue.
The easiest explanation for the final low cpue,
is .probably
related t o
the
urn concept:
the interaction between a
great
number
O f
boats on-a limited local resource Will
increase
the
catch,
but
Will
reduce the cpue,
more quickly
than .a
small
fishing effort.
^l.--.“.-.-.-~-l”-.lll-~~~~..l-.--~~”-l-l.-L
_...”
”
,,.~
~---~I-.~-~“-“_--“r.-...”
1
,-_,.llr
,.
. _ _
I..I-
^_._
_-----_
-_.
5 8
7. GEOGRAPHICAL COMPETITION IN THE MEDIUM TERM
(less t]lan 2 years) BETWZ?BN Ffl#%NG BREBlt %#E BOXE8 MODa%*
7.1. The concept: migrations treated as a diffusion process;
AnY
tuna
fishery
cari be divided in
time(ex.month)
and
area boxes (ex. 10' to 10°),
which are fished or
unfished. In
every area there is at any time,
a gjven number of fishes( often
unknown....).
When the area is exploited , the abundance in each
box cari be easily estimated using an unbiaised cpue index. Aerial
survey
and
direct
estimates of biomass cari
also
measure
the
abundance in each box, especially for the virgin stocks.
From
one period to the other,
the number of fishes in each
box Will change because of :
l.Input to the box from other areas (immigration),
positive
factor.
2.Out:put from the box toward other areas(emigration).
3.Removal b:y the natural mortality .
4.Removal by fishing mortality.
These three last factors are negative ones,
and reduce
the
local population.
Migrations <oan be measured by intensive and repeated tagging
results,
associated
to significant fisheries covering as
many
areas as possible.
Natural
mortality
is not estimated precisely for
most of
tuna species,
but this is not necessarily critical in the
short
and medium term studies .
The
fishing mortality in each exploited box Will depend on
the fishing effort, and more directly on the total catches.
7.2. An example of the boxes model.
Computer
models
cari
easily
be designed to
describe i n
equations
a11
the
previous
concepts of
boxes,
underlying
populations,
migrations
between
boxes,
fisheries and
natural
mortalities.
Such a
mode1 is presently under testing
for
the
eastern
Atlantic skipjack fisheries:
its framework is shown in
figure 34.
This
mode1 has been built in order to
analyze
the
geographical
interactions
between
skipjack
fisheries in
the
Eastern Atlantic (Figure 35).s:This mode1 is a catenary reversible
differential
model,
with
533oxes,
the unit of time being
the
month .
A sixth box,
adjacent to
the 5
others corresponds to
the
unexploited
fraction of the stock.
This sixth box is
the
"cryptic
nursery"
for
thejuveniles during
their
progressive
recruitment ,
and
the "black hole" concept
(Fonteneau 1983),
where
adults skipjack become unavailable and disappear from
the
fisheries.
Such
type
of differential linear mode1 is widely
used in
5 9
many
research
fields in
order
to
describe
diffusion
processes(Sheppard 1962,Jaquez 1972).
In our present model, the status of the system at the end of
each
monthly interval Will depend of eh@ 18 got@Rtaal
migrk3thon
rates between boxes .
7.3.
Example
of skipjack data used for developping and
testing
the model:
The
data
used in this mode1 belong to
the
for
following
types: (a) The monthly cpue by age is givea in number of skipjack
taken by
fishing day of FIS and spanish purse
seiners
(Figure
36).
It has been calculated monthly for the 5 selected areas of
the
model.
The
cpue is calculated on the catch by
size
file,
assuminq
the
same growth pattern as the one used by
FONTENEAU
1986, i.e.
an average growth of 12 cm per year (figure 36). In
those
calculation the the group of fishes measures from 37 cm to
50 cm in August 81.
Some cpue data are missing in 2 cases, first
when
size
data
are
missing or when the
catch is
only
from
baiboats.
(b) The catch is also calculated in numbers,
from the same
size file and for the same group of fishes,
but is estimated for
a11
fleets (baitboat and purseiners) I
in order to
measure
the
total removal on the cohort (VPA concept).
(cl
The
recoveries of tagged skipjack are extracted
from
the
best
ICCAT recovery file presently available
for
the
two
tagging
conducted
during
skipjack year (1981) in
Senegal
and
Ghana areas (Figure 38 and 39).
(d)
The total fishing mortality estimated on
the
eastern
Atlantic
stock
by age follows the general pattern estimated by
FONTENEAU 1986 (Figure 40).
7.4. Estimating the underlying populations and migration rates :
The mode1 will follow numerically a qroup of fish,
from its
early recruitment until the death of the oldest fish.
This basic
qeneral
concept is
similar to the VPA or
yield
per
recruit
calculations.
The method used to adjust the parameters of the mode1 to the
existing data Will be the following :
l.Knowing
the
cpue
by age in each
box,
the
theoretical
underlying
local population in each box must
be proportionaï to
the
local cpue of the age (at each time).
---
2.Knowing the catch by age in each box, the-local underlying
population
must be at least equal to the
observed
catch.
IThe
removal o f
fishes by the fisheries is exactly-known by
fishery
statistics.)
3.The total fishing mortality on the cohort (i.e. on the sum
of
individuals in
a11 the boxes), is
the
fishing
mortality
previously calculated on the overall stock by standard VPA.
6 0
4.The 1 8
monthly
mixing rates between boxes
must be in
agrement
with the observed diffusion of tagged fish.
Only
some
intensive
tagging cruise cari be used to that purpose,
each
one
with
several thousands of tags applied within a very short
time
in a small area (see table 1).
The
migration
parameters estimated by the mode1
should be in
agrement with the observed diffusion shown in figures 38 and 39.
Unfortunately,
this type of adjustment by the mode1 does not
give a unique solution for migration rates,
unless
in an
ideal
and
hygothetic
case with intensive,
generalized and
permanent
tagging in a11 strata.
However,the
present
adjustment of the mode1 to the
actual 4
sets of
data cari give a range of possible
mixing
rates
which
Will be more realistic and quite narrow compared to the
previous
complete uncertainties.
This work is still under progress on the CRODT computer,
and
a
the
possible
range of
mixing
rates
has
not
yet
been
established.
6.4.
Estimating
interactions
between
fisheries with
the
box
model.
Using
the boxes mode1 and its parameters,
it becomes
quite
simple to
calculate
some theoretical
short
and
medium
term
interactions between fisheries operating in different areas.
For instance,the effect of reducing or increasing the fishing
effort in any area,
upon the catches and cpue in a11 other areas
cari
easily be tested by simulations,
Also,
catches and cpue in
each box under various local fishing mortality schemes,
cari b e
tested .
The boxes mode1 is still in its development phase,
especially
in
order to explore the full range of possible
migration
rates
between boxes.
Several of those parameters Will probably
remain
undetermined because of unsufficient data.
However,
some preliminary results may be of some interest:
- most
of
the
time
the
interactions
are
universely
proportionna1 to the distance.
For instance*
the Ghana
skipjack
fishery
has some significant effect
on the Cape Lopez
(average
distance = 600 naut.miles)and Liberia
fisheries (900 miles). The
Senegal
fishery
has very little or no effect on the Cape
Lope2
(1800 miles) gr Angola fisheries (2500 miles), but some effect on -
the Liberia fishery (600 miles) .
- a
short
distance
between two fishing areas is
not an
evidence for strong potential interactions: the Cape Verde Island
and Senegal fisheries are located at a short distance,
200 miles
only,
but
show
very
little or no
mixing
between
the
two
fisheries.
This
absence of
interactions
has
been
clearly
demonstrated by
the
recoveries of two
simultaneous
intensive
tagging cruises conducted in the 2 areas.
- only intensive fisheries and high catches in one area
c an
affect the othes ones,
because of the intensity (in general) and
the
complexity of mixing processes,
at least for skipjack.
For
instance
the
Angola fishery which operates presently at a
10 w
level,
compared to the local biomass, has no effect on any other
fishery
because
the
Angola catch is small
and
the
skipjack
biomass
transient
from Angola
is diluted
very
quickly
among
skipjack
originated
from
other
areas,
especially
with
the
unfished
fraction of
the
skipjack stock
migrating
from
th'e
"cryptic nursery" to the major fishing areas..
7.5. Discussion of the boxes model.
This
type
of mode1 is a potentially
serious
progress
in
analyzing
the stock structure problem ,
the migration
patterns
and the interactions between fisheries.
However
it
requires a huge amount of detailed
statistics,
especially catches and cpue statistics by sizes. Those statistics
must caver a11 fleets(VPA concept).
Intensive tagging with
good
recovery statistics must also be available. Furthermore this type
of
mode1 in fundamentally based on the existence of
significant
fisheries covering preferably most of the stock and operating at
relatively
high
level of fishing mortality (at least
:Locally).
Even i n
the
optimum case when mixing rates
cari be
estimated
duxing
a limited period of intensive researches (such as
ISYP),
it is
not
yet clear to what degree
the -calculated
migration
pattern
cari be extrapolated to an other period.
More
generally
the
year
to
year
variability of
tuna
migrations
a II d
i t s
determinism is still a great unknown .
Also
the
mode1
rely on
several
underlying
hypothesis,
especially on
the homogeneity of boxes and on the
box to
box
relationship,
which
cari probably be more analyzed and improved,.
The
heterogeneity
of the senegalese box has for
instance
been
clearly demonstrated by the low exchange rate between Senegal and
Cape Verde islands.
Furthermore the analysis assumes that
5vow
of
fishes
is followed (catch and cpue) when the individual
are
growing.
The variante of growth between individuals and
between
area
being
apparently
large
(Bard et Antoine,
19861,
this
hypothesis
is not strictly justified.
However, this type of research is probably a very interesting
field to be developped.
8. CONCLUSION.
The present paper has tried to demonstrate the complexity and
the
heterogeneity of
the
interaction
problem
between
tuna
fisheries. It
makes a
provisional
review of
the
present
observations done upon the atlantic tropical tunas in this field.
1t
is clear
that the Atlantic ocean offers good conditions
for
those studies,
because of intensive fisheries,
good
statistics
over a long period of time and some intensive tagging .
6 2
Many
of the Atlantic results cari probably be extrapolated
at least to some extent,
from one ocean to the other,
even when
the ecological conditions are quite different.
This
type of worldwide analysis of the interactions
betqween
tuna fisheries seems presently a key action to conduct.
In
any case this paper well shows the fundamental importance
for any interaction study of :
-Detailed fishery statistics(catch,effort and sizes taken by
gear),with a detailed time and area stratification.
-Intensive tagging.
The tagging of a great number of fish done simultaneously in the
same spot,
such as the senegalese
,cap verdian or japanese
ISYP
tagging,
seems
to be a key action towards the estimation of the
short term interaction between fisheries,
at least those
taking
small fishes.
-
6 3
BIBLIOGRAPHY
BARD F.X.
1984 . Le listao de l'Atlantique.Résultat des campagnes
de
marquages
effectués
de
1980 à
1982.
La
Peche
Maritime,n01275,juin 1984,p.319-324.
BARD F.X. et
ANTOINE L.
1986.
Croissance du
lista0
dans
l'Atlantique est.
Proc.
of ICCAT Conférence en the ISYP. p. 301-
3 0 8 .
CAYRE P. et
BARD
F,X.(en cours
publication
FAO)Biologie
e t
reproduction
des thonidés de l'Atlantique.Dans synthèse FAO
s UIC
les
thonidés de l'Atlantique centre est-FAO technical
document,
Fonteneau et Marcille éditeurs.
FONTENEAU A.
3.975
Note sur les problèmes
d'identification
du
patudo dans les statistiques de pêche.Rec.Doc.Scient. I.C.C.A.T.,
Vol.5,n" l.,pp 168-171.
FONTENEAU A.
1981 . Note sur le mode de calcul de la
prise
p a. ir
unité d'effort des senneurs FISM.
Rec.
Dot. Scient. I.C,C.A.T.,
vo:L.15(1),
pp 407-411.
FONTENEAU A.
1.984.
Analyse de
l'état des
stocks
d'albacore
(Thunnus
albacares) de l'Atlantique au 30 Mai
1984 .Rec.
DO~..
Scient. ICCAT, Vol. 21(2), pp 80-101.
FONTENEAU A.
E' t
ROY C.
1987.
Pêche ~thonière e t
anomalies
climatiques
de l'environnement dans l'Atlantique tropical centre
est en 1984. Rec. Dot. Scient. ICCAT ,Vo1.26(l),p.228-236,
FONTENEAU A.
1.986 a
Analyse de
l'exploitation de
que Iques
concentrations
d'albacore
par les senneurs durant la
période
1980-1983 dans l'Atlantique est .
Rec.
Scient. ICCAT , Vol,, 25,,
pp 81-98.
FONTENEAU A.
1986.
Etat du stock d'albacore de l'Atlantique est
au 30 Octobre 1986.Doc.
ICCAT SCRS/86/60.
En cours
publication
Rec.Doc. Scient. ICCAT.
HUNTER J.,
ARGUE A.W. ,BAYLIFFF W.H.,DIZON A.E.,FONTENEAU A-H.,
GOODMAN D.
and SECKEL G. 1986. The dynamics of tuna movements:an
evaluation of past and present research.
JAQUEZ J-A,
1972.
- Compartimenta1
analysis. in
biology
an Ci
medecine. Elsevier .Amsterdam.,l972.
LAUREC A.
et FONTENEAU A. 1979 u Estimation de l-abondance d'une
classe
d'age.
Utilisation
des
cpue de
plusieurs
enginsy
en
différentes zones et saisons .Rec, Dot. Scient. ICCAT ,Vo1.28(1),
pp 79-100.
64
MIYAKE M.
and
HAYASI S.
1978 . Field manual for statistics and
sampling
SO f
tunas and related species in
the
Atlantic
Ocean.
149p.
PEREIRA J.
1986 Analyse de l'état du stock de patudo atlantique.
Dot.
ICCAT SCRS/86/62.
En cours publication Rec.
Dot.
Scient.
ICCAT.
RICKER W.E.
19'75
Computation and interpretation of
biological
statistics of fish populations.
Bull.
Fish.
Res.
Board Can. :
382 p.
SHEPPARD C.W. 1962 . Basic principles of the tracer method.
Introduction to mathematical tracer kinetics. Wiley,New York,1962.
STRETTA
J.M.
1977 .
Températures de surface et pêche
thonière
dans la zone frontale du Cap Lopez (Atlantique tropical oriental)
en juin et juillet 1972,
1974 et 1975.
Cah. O.R.S.T.O.M. , sér.
Océanogr.,Vol 15 ,n02,1977:pp163-180.
6 5
Table l.- Skipjack taggings done during recent years which cari be
used
for diffusion analysis because of a significant
number of
tags released during a single cruise in the same place.
____-________ ------------------------------ ------ ------------.-.--.--
YEAR
ZONE
COUNTRY
TAGGED
SKIPJACK
-----------------------------------------------------------------
80
GHANA
JAPON
5976
81
GHANA
JAPON
7000
81
SENEGAL
SENEGAL
1391
81
SENEGAL
CAP VERT
2672
82
SENEGAL
CAP VERT
4552
82
SENEGAL
SENEGAL
2794
_____-____- -_.--- ---.- -------------I- ------ ---. -~~~~~--~.“~~I--._--~~--__.-___-_
TOTAL ZONE SENEGAL
11409
TOTAL ZONE GHANA
12976
_____ - ---_-- ----.-------*------ ---- ---- ----.-- ------_.--_-__--- --.- --.-_,___
RECR”ISYENT
/
UNKNOWN
RANGE 01
--
I
I
f-----
’
’50
1 0 Y ” FECUNOITY
O F S T O C K
Figure l.- The
classical
stock
recruitment
relationship
(a)Ricker type and (b) Beverton and Holt,
and the Atlantic stock
relationship
with
the
present level of
stocks
estimated
for
yellowfin, bigeye and skipjack.
,
(max. janv. juin)
-L-l---.-
&~OS d e rsprodue
dstsrminh
m g o n a d e s m û r e s
,--&r”es’to”ts
rannce
(faibles quatités l
Figure 2.- Spawning areas of Atlantic yellowfin (from
Cayre et
al. in press in FAO synthesis on eastern Atlantic tuna)
0
““?h
Twt. Imvlda 0
(1
\\?+a-
- -. ._:- \\ c;
‘<0
Figure 3.- Spawning areas of Atlantic bigeye (from Cayré et al.,
in press in FAO synthesis on eastern Atlantic tuna).
*
sp’
y
70’
60’
SO’
4;
30°
20’
3’k
I
4
1
I
1
8
NO
40
‘C
I
!14
-
. .
I
10'
30’
20'
20’
10’
10'
0’
0'
tQ’
10'
w’
20*
SO’
30’
. .
3
J
f
w’
1IO’
i
20’ l
Figure 4.- Spawning
areas of Atlantic skipjack (from
Cayre et
al.., in press in FAO synthesis on eastern Atlantic tuna).
Figure 5.- Areas of catches of small bigeye (less than 90 cm) by
purse
seiners
(1979-1983) (5a) and by Tema baitboats (1975-1982)
(5b).
.
.
.
.
.
.
*
Y
-10
-
-0
-
-.
Figure 6.- Areas of catches of large bigeye (more than 90 cm), by
purse
seiners
(1979-1983) (6a) and hy longliners
(1978-1982)
(6b).
6 9
SMALL
BIGEYE.
LARGE
BIGEYE
Figure 7.- Catches of small (less than 90 cm) and of large bigeye
(more than 90 cm), from 1955 to 1983.
CP U E
1970
75
80
ANNÉE
CPU E
1965
70
75
80
A N N É E
Figure 8.- Catch per unit of effort of bigeye (predominantly less
than 90 cm), by surface fleets.
-,_“.. ._ . . . 1 . ..-. _ . . . . . “1- -_... .., -.<., _.” _.,._,._, _ _-
7 i,
-..
BIOMASSE
(1000 t)
P U E
5(
PALANGRE
-
-
---
:Oe6
4 0 0
300 Ir--
‘h
---h---\\\\
,
- 0 . 5
2 0 0 -
- - d
lOO-
- 0 . 2
- 0 . 1
t
-
.
.
1
-
-
’
- 1
-
“.
’ 1 -
’
-
’ I
’
’
1960
1965
1970
1975
1 9 6 0 A N N É E
Figure 9.- Index
of abundance (calculated for bigeye
tuna on
japanese
longline fisheries by Honma method,
corrected for deep
longlining) , and adult biomass c-alculated by VPA.
-..-.a 1972-1975
D-----.0 1076-1979
-+ 1960-1063
o::3:;“-
6
7
A G E ( A N N É E S )
Figure lQ.- Average
fishing mortalities by age,
during
recent
years on bigeye tuna.
Q----+
VARIABLE
RECRUITMENT
NUMBERS 1
(a)
BIGEYE
.-. C O N S T A N T
OF BIGEYE
AT .AGE 3.0I
A
Figure ll.- Estimated numbers of bigeye tunas recruited in
the
adult fisheries (result of VPA, from Pereira 1984).
r- ----_ .--.-- --
’
i-ï\\
)r3
9
12a
0
/
FP
l
VI
1 2 b
Figure 12.- Multigear
yield
per re cruit
analysis
f 0 1:
bigeye
small
bigeye fisheries against large bigeye fishery 12 (a).
Th e
Fishing
mortalities by age of each g ear are given in
figure
l:&(b) il
13a
Figure 13.- Fishing areas of small yellowfin (less than 90 cmi,by
pur-se
seiners(13
a) during recent years
(1979-1983) and by
Tema baitboats (1975-1982) (13b)
Figure 14.~. Fishing areas of lar
lowfin (more than 90 cm) by
purse
seiners
(14a) during rec
ars
longliners (1978-1982) (14 b).
and fishing
areas o f
SMALL
YELLOWFIFI
i
--rllr
1960
1965
1970
1975
1980
1985
Figure 15.- Numbers of small (less than 90 cm) and
large
(more
than 90 cm) yellowfin taken in the Atlantic by a11 fisheries.
n-
1980
YEAfl O F 6ECAd,TnÇti?e
Figure 16.- Catch per unit of effort of yel1owfi.n (from Fonteneau
1984)
at age 1 (FISM purse seiners,
:LAUREC and FONTENEAU
1977
method), measuringg the recruitment.
'bOasb
?. u. fi.
. '1
b-9)
-
4
BlOMASSE
:
- 3
lOO,-
4
I
I
I
I
r *
.
6 0
Figure 17.- (a) Catch per u6iit ofeffor?of the japanese longline
fishery
(Honma
index,
uncorrected for
deep
longlining),
and
biomass
calculated by VPA (from FONTENEAU 1984) and 18(b)
catch
per unit of effort on large yellowfin by purse seiners (Fonteneau
1986,
cpue
index corrected for large fishes),and longline
cpue
index during the same period 1969-1983.
/
3
NUMBERS
OF YELLOWFIN
AT AGE 3.0
Figure 18.- Changes in
the
recruitment levels
yel.lowfin fisheries estimated by VPA.
t0
the
adult
F
2. BE
1.50
/4
P.dd$
1
1
6
5
,
‘3 !Q 11 1:: l! lû 1 5 l b 17 18 19 ia ?l :z î! ;-
RfE (TII* )
Figure 19.- Average age specific fishing mortality
for
e a ci ter n
Atlantic yellowfin (period 1980-1983) (from FONTENEAU 1986).
SURFACE (standard days x 1,000)
,50000
Figure 20.- Catches by purse seiners and by longliners on
Large
ye:Llowfin
as a function of their respective fishing effort iFrorn
HUNTER et al. 1986).
100 L /
1
10
dl
3b
46--
ISO
6à FISHING EFFORT
by SURFACE FISHERIES
21b
P
_ 1
ldoo
1 00
ZdOO EFFECTIVE EFFOFV
O N .BIGEYE
by Lonp Lin.rs
Figure
21..- Catches
by purse seiners (2la)
and by
longliners
(21b)
on large bigeye as a function of their respective
efforts
fishing
(curves are adjusted by eye).
Figure 22.- Yield
per
recruit of yellowfin
eastern
Atlantic,
period 1980-83,
Ricker mode1 corresponding to F vector of figure
19 (from FONTENEAU 2986).
7 5
1
t
1
Figure 23.- Yield per recruit of bigeye, a11
Atlantic,
p e IT io d
1980 to 1984 , Ricker mode1 (from PEREIRA 1986).
-
-
THEOAETICAL
YIELO
f’ER RECR~IT
G N AOULT
10
- - - - - OESERVED YIELL) O F FISHERIES
5
I
70 -74
80-83
F
PUR-SE S E I N E
LONGLINE
YELLOWFIN
F
62!65
71’-75
F
PURSE S E I N E
LONGLINE
Figure 24- Theoretical and observed yield of purse seiners
and
longliners,for
yellowfin and bigeye tunas.The theoretical
yield
is a yield per recruit analysis result,
Ricker model,
where the
yield per recruit is multiplied by the recruitment estimated
for
the
adult fishery . In this Y/R analysis,each gear is supposed to
h n
f i chincl ~-l,Y~~
7 6
e
KIRGIN STOCK
'ED STCICK
VIRGIN STOCK
ZXPLOITED STOCK
-'-----.-L--B ,-._
-
--
.
.
t
-
:
. .
.
'
_
,_
\\
I
\\
:
\\
.
.,
.
<
.
\\
,
.
-.
.
~
Figure 25.- The
hypothetical
vertical
stock
structure
of
yellowfin
and
bigeye for the virgin stocks and
for
the
fully
exploited stocks.
7 7
Figure
26.- Catch
per
unit of effort of
japanese
lc: n g 1 i ri e r s
during
the first quarter e-g.
the better fishing season in
the
Gulf
of Guinea (in tons/1000 hooks),
for the historical fishery
1961
to
1969
(Each
month and 5 to 5
degree
square
cpue i s
represented randomly in the corresponding 5 degree square).
Figure 27.- Catch per unit of effort of japanese longliners during
the
first quarter ( tons/1000 hooks),
during the recent
period
1970
to
1982 (Each month- 5 to 5 degree
cpue i s
represerted
randomly .in the corresponding 5 degree square).
7 8
28a :
THE "URN" MODEL CONCERT
VIRGIN POPULATION
,PHASE 1 OF EXPLOITATION
_ FINAL PHASE
TQTAL CATCH PHASE 1
8 ,
-’
TOTAL CATCH
28b
C A T C H , E F F O R T j
C P U E R E L A T I O N S H I P
CATCH -
LOCAL AVERAGE
BIOMASS
A V E R A G E M A X I M U M C A T C H
- - - - - --_-- _--_-_ - - - --.
---_--- -.---.-
___--
5000.
XX-
l
-
-
d
LOCAL
-
EFFO
FIRST FISHEAY
EFFORT
II
E F F O R T O F A D O I T I O N A L FISHERY
Figure 28.- The "urn model" concept:
(28a)The general
simple idea of a closed and limited
exploited
resource.
(28b)The catch and catch rate trends when an additionna1
fishery
is
added to a first fishery:(a) is the cpue of the first fishery
before the arriva.1 of the second one and (b)is the cpue after -
1st concentration
2 nd concentration
3rd concentration
Cp”e
UNDERLYING
ESTIMATED
POPULATION
I
-.^ I
\\
POP
o - - - - o
?--y
9
(a)
3 DAYS INTEVAL
CC)
5 0
OW8"
weight
I
1
I
1
1
-F--
I NTERVAL
Figure 29.- Exqmple of parameters concerning
the exploitation; of
three yellowfin concentraCions in the eastern Atlantic.
fa)
observed
cpue
of--purse seiners( in tons by 10
hours of
searching),and
estimated
underlying
populations by 3
tiays
-
periods(From Fonteneau 1986).
(b)Numbers o f
positive
sets
by 3 days
periods(Which
cari b e
considered
as
the
number of
schools
removed
from
,t h e
concentration)
(c)Average weight of the yellowfin caught.
y e a r
Figure 30.- The
fishing
area o f
Cap2 Lopez :
geographical
distribution
of the catches by purse seiners (FIS and Spanish),
by 15 days period from 1978 to 1984 (from FONTENEAII and ROY 1987).
-
F I S H I N G E F F O R T
( F I S P S F I S H I N G DAYS)
Figure 31.- The relation observed between total catch (yell<:lwfin,
skipjack and bigeye) and fishing effort in the area of (:ape ~,~>pez
during
the fishing season from may to july,
from 1969
(Curve adjusted by eye).
to
1984
Figure 32.- Average
fishing
zones of skipjack by
on e
decrree
squares, a11 fleets, period 1975-1982.
8 2
e f f o r t
-5
5-9.9
,
+ 55 qq
1 ’
.
I
0
0
D
00
00
0
0
0
0
O
0:
0
0
0
100gbservatiOns
0
o
l
0
O
0
0
0
0
0
0
0
Figure 33.-Frequency of
total cpue as a
function o f
fishing
effort
for
spanish and FISM purse seiners _fleets,
during
the
period 1980-1986.
Each circle is proportionna1 ta the number of
catch
rates observed in the l'squares during 1.5 days periods for
increasing intensity of fishing efforts. Frequencies greater than
200
have been limited arbitrarily%t the 200 level in
order to
facilitaté the overall representation.
8 3
1
-.- --t-- j
---T---Ii “!’
-- -.--IL--- r v-jii
i l
.- -~~I---~--
J
-f--l“.’
Figure
34.- The concept of the boxes
mode1 applied to
skipjack
in the eastern Atlantic ; each arrow shows a monthly tranfer rate
of individuals from one box to the other.
Figure
35.- Hypothetical
migrations
of skipjack
tuna
in
the
eastern
Atlantic,
based on tagging results (from Cayre et al.,
FAO tuna synthesis under press).
----
.-..
--._-
---..--.-*--*P-ms
.^,_
_‘.,,_
-
--z..il/.-
19---.IXX*-I,.
_ .
,l”__
.,II,
,......
.-“-
--.-
--
---.-.__
8 4
Figure 36.- Monthly cpue for an age group of skipjack, by
area,
during
the
period
1979-1982
(each
circle
has an
area
proportionna1
to the cpue);
a blank means a zero cpue,and an x
means
no purse Seine effort data or no size
data.This
skipjack
group is
from
37cm to 50 cm in august 1981 (date of
japanese
tagging) i
a11
individual are growing at a constant rate of
Icm
per month before and after this date.
0 1000 T.
-
Figure 37.- Monthly catch of skipjack,
by area,. during the period
1980-1983 (each circle has an area proportionna1 to the catch).
months
liberty
<’
0 0 0 c/ c>
I
I
.
--
--
8 6
I
l-------~--T---1
3
4
-’ A G E (?mnées)
Figure 4Q.-Vector
of fishing mortalities by age
estimated
for
skipjack (From Fonteneau 1986).